Skip to main content

Full text of "Year book - Carnegie Institution of Washington"

See other formats




Year Book 82 


Library of Congress Catalog Card Number 3-16716 

International Standard Book Number 0-87279-657-4 

Composition by Harper Graphics, Inc., Waldorf, Maryland 

Printing by Automated Graphic Systems, White Plains, Maryland 

December 1983 


Officers and Staff v 

Report of the President 1 

Reports of Departments and Special Studies 1 

Department of Plant Biology 3 

Department of Embryology 131 

Geophysical Laboratory 229 

Department of Terrestrial Magnetism 439 

Mount Wilson and Las Campanas Observatories 591 

Developmental Biology Research Group 677 

Administrative Reports 693 

Office of Administration 695 

Bibliography 697 

Report of the Executive Committee 699 

Abstract of Minutes of the Eighty-Sixth Meeting of the Board of 

Trustees 701 

Financial Statements 703 

Articles of Incorporation 719 

By-Laws of the Institution 723 

Index of Names 729 

Digitized by the Internet Archive 

in 2012 with funding from 

LYRASIS Members and Sloan Foundation 

President and Trustees 

James D. Ebert 

William R. Hewlett 

William C. Greenough 
V ice-Chairman 

William T. Golden 

Philip H. Abelson 
Robert 0. Anderson 1 
Lewis M. Branscomb 
William T. Coleman, Jr. 
Edward E. David, Jr. 
John Diebold 
Gerald M. Edelman 
Carl J. Gilbert 2 
Robert G. Goelet 
Crawford H. Greenewalt 
Caryl P. Haskins 
Richard E. Heckert 
George F. Jewett, Jr. 3 
John D. Macomber 
William McChesney Martin, Jr. 4 
Antonia Johnson Morner 
Franklin D. Murphy 
Robert M. Pennoyer 
Richard S. Perkins 
Robert C. Seamans, Jr. 
Frank Stanton 
Charles H. Townes 
Sidney N. Weinberg, Jr. 3 

Garrison Norton 
Charles P. Taft 5 
Trustees Emeriti 

designed May 6, 1983 

2 Trustee Emeritus as of May 6, 1983; died November 13, 1983 

3 Elected May 6, 1983 

4 Trustee Emeritus as of May 6, 1983 

5 Died June 24, 1983 


Former Presidents and Trustees 

Daniel Coit Gilman 
Robert Simpson Woodward 
John Campbell Merriam 

Alexander Agassiz 
Lord Ashby of Brandon 
J. Paul Austin 
George J. Baldwin 
Thomas Barbour 
James F. Bell 
John S. Billings 
Robert Woods Bliss 
Amory H. Bradford 
Lindsay Bradford 
Omar N. Bradley 
Robert S. Brookings 
Vannevar Bush 
John L. Cadwalader 
William W. Campbell 
John J. Carty 
Whitefoord R. Cole 
John T. Connor 
Frederic A. Delano 
Cleveland H. Dodge 
William E. Dodge 
Charles P. Fenner 
Michael Ference, Jr. 
Homer L. Ferguson 
Simon Flexner 
W. Cameron Forbes 
James Forrestal 
William N. Frew 
Lyman J. Gage 
Walter S. Gifford 
Carl J. Gilbert 
Cass Gilbert 
Frederick H. Gillett 
Daniel C. Gilman 
Hanna H. Gray 
Patrick E. Haggerty 
John Hay 

Barklie McKee Henry 
Myron T. Herrick 
Abram S. Hewitt 
Henry L. Higginson 
Ethan A. Hitchcock 
Henry Hitchcock 
Herbert Hoover 
William Wirt Howe 
Charles L. Hutchinson 
Walter A. Jessup 
Frank B. Jewett 
Samuel P. Langley 
Ernest 0. Lawrence 
Charles A. Lindbergh 
William Lindsay 
Henry Cabot Lodge 
Alfred L. Loomis 
Robert A. Lovett 


1902-1904 Vannevar Bush 

1904-1920 Caryl P. Haskins 

1921-1938 Philip H. Abelson 

























































Seth Low 
Wayne MacVeagh 
Keith S. McHugh 
Andrew W. Mellon 
John Campbell Merriam 
Margaret Carnegie Miller 
Roswell Miller 
Darius O. Mills 
S. Weir Mitchell 
Andrew J. Montague 
Henry S. Morgan 
William W. Morrow 
Seeley G. Mudd 
William I. Myers 
William Church Osborn 
Walter H. Page 
James Parmelee 
Wm. Barclay Parsons 
Stewart Paton 
George W. Pepper 
John J. Pershing 
Henning W. Prentis, Jr. 
Henry S. Pritchett 
Gordon S. Rentschler 
David Rockefeller 
Elihu Root 
Elihu Root, Jr. 
Julius Rosenwald 
William M. Roth 
William W. Rubey 
Martin A. Ryerson 
Henry R. Shepley 
Theobald Smith 
John C. Spooner 
William Benson Storey 
Richard P. Strong 
Charles P. Taft 
William H. Taft 
William S. Thayer 
Juan T. Trippe 
James W. Wadsworth 
Charles D. Walcott 
Frederic C. Walcott 
Henry P. Walcott 
Lewis H. Weed 
William H. Welch 
Andrew D. White 
Edward D. White 
Henry White 
James N. White 
George W. Wickersham 
Robert E. Wilson 
Robert S. Woodward 
Carroll D. Wright 



Under the original charter, from the date of organization until April 28, 1904, the following were 
ex officio members of the Board of Trustees: the President of the United States, the President of 
the Senate, the Speaker of the House of Representatives, the Secretary of the Smithsonian Institution, 
and the President of the National Academy of Sciences. 



1530 P Street, N.W., Washington, D.C. 20005 

James D. Ebert President 
Margaret L. A. Mac Vicar Vice President 1 

James W. Boise Bursar; Executive Secretary to the 

Finance Committee 

Christopher Wright Staff Member (Science Policy and 

Institutional Development) 

Ray Bowers Publications Officer; Editor 

John C. Lawrence Assistant Bursar 

Kenneth R. Henard Business Manager 

Susan Y. Vasquez Assistant to the President 

Joseph M. S. Haraburda Accounting Manager 

Patricia Parratt Assistant Editor 

Marshall Hornblower Counsel 
Roy J. Britten 

Barbara McClintock 

Effective July 1, 1983. 


Carnegie Institution of Washington adheres in all phases 
of its operations, including employment and educational 
programs, to a policy barring discrimination on the basis 
of race, religion, color, national or ethnic origin, sex, or 
physical handicap. In its educational programs it admits 
qualified students as fellows without regard to race, reli- 
gion, color, national or ethnic origin, sex, or physical 
handicap to all the rights, privileges, programs, and ac- 
tivities generally accorded or made available to fellows at 
the Institution. It does not discriminate on the basis of 
race, religion, color, national or ethnic origin, sex, or 
physical handicap in administration of its educational 
policies, admissions policies, fellowship programs, and 
other Institution-administered programs. 

Report of 
the President 

Nobelists Alfred Hershey and Barbara McClintock of the Carnegie Institu- 
tion's former Department of Genetics, Cold Spring Harbor, New York 

In nature's infinite book of secrecy 
A little I can read. 

Shakespeare: Antony and Cleopatra 

Until very recently Dr. McClintock's work was so far out of 
the mainstream as to border on the heretical. While molecular 
biologists were marveling at the mathematical orderliness of the 
genetic world they were discovering, the genes in Dr. Mc- 
Clintock's maize were jumping about and changing the charac- 
teristics of the cells they inhabited in disconcerting fashion. Only 
the timely intervention and continuing support of the Carnegie 
Institution of Washington made it possible for her to pursue — 
for more than four decades — her intricate and solitary re- 
search .... 

To get the government grants needed to support the expensive 
tools of modern science, researchers typically must work in groups 
and on projects in the mainstream of current theory. This is 
generally a very fruitful process. But there are exceptions to that 
rule, and it's worth remembering that Dr. McClintock — working 
patiently and painstakingly in her laboratory, alone and unrec- 
ognized over the decades — is one of them. 

"Nobel for a Loner" 
The Washington Post 
October 13, 1983 

It was with the deepest pride that I congratulated Barbara McClintock 
upon her selection by the Nobel Assembly of the Karolinska Institute 
in Stockholm. Her richly deserved Nobel Prize caps a long list of honors 
and awards, beginning with the Kimber Genetics Award from the National 
Academy of Sciences in 1967. Her newest and greatest recognition an- 
nounces to all the world what her colleagues in the Carnegie Institution 
of Washington and the Cold Spring Harbor Laboratory, and in the field 
of maize genetics everywhere, have long known — that she is a truly bril- 
liant, devoted, and inspirational scientist. 

Of the thousands of words written in the many articles that appeared 
in the weeks after her selection, none captured the uniqueness of the 
person and her contributions to science better than the editorial in The 
Washington Post. It was further pleasing that the Post's writer recognized 
the crucial role of the Carnegie Institution in McClintock's work: 



Only the timely intervention and continuing support of the Car- 
negie Institution of Washington made it possible for her to 
pursue — for more than four decades — her intricate and solitary 

But the editorial was also perceptive in other ways. In the concluding 
paragraph, the writer noted that in today's world, to obtain necessary 
government grants, "researchers typically must work in groups and on 
projects in the mainstream of current theory." Embedded in this comment 
are the seeds of a major concern of the Institution and of thoughtful 
scientists everywhere. Let me present the matter from our own viewpoint: 
Is the Carnegie Institution still able to identify and then protect other 
unique scientists who, like McClintock, work away from the mainstream 
"patiently and painstakingly, alone and unrecognized"? 

Before proceeding with this main theme, let me emphasize that I have 
deliberately avoided specifying the sex of those "unique scientists." Often, 
tributes to the Institution for recognizing McClintock's genius have made 
special mention of the courage of Milislav Demerec — the director of the 
Institution's Department of Genetics in 1942 — in appointing a woman. I 
believe, however, that McClintock's sex had very little to do with it. To 
put it simply, Demerec was an astute judge of scientific talent, and with 
the support of presidents Vannevar Bush and Caryl Haskins and the 
Institution's trustees, he fashioned a department at Cold Spring Harbor 
that for a period of years had no equal. Moreover, there was ample prec- 
edent in the Institution for the appointment of women scientists. In 1914, 
director Franklin P. Mall appointed Margaret Reed Lewis — a true pioneer 
in the application of cell culture techniques to studies of development and 
cancer — to the Department of Embryology; during that same period, the 
Department of Embryology enjoyed an intimate association with Florence 
Sabin in the Department of Anatomy at Johns Hopkins Medical School. 
Later, the Department of Embryology provided a happy environment for 
Elizabeth M. Ramsey, whose studies of the anatomy and physiology of 
the placenta are nonpareil, and for the gifted experimental morphologist 
Mary Rawles. 

What these women had in common with their male counterparts was 
the constant support of Carnegie Institution — unflagging, consistent sup- 
port at a level sufficient to enable them to choose problems, in Ramsey's 
words, "not for their ease, but for their importance." And, I must add, 
important problems are not confined to those in the mainstream of current 

Moreover, there is not necessarily a direct relation between the im- 
portance of research and its cost. Costs are related in part to the size of 
research groups, though Carnegie scientists do not tend to work in large 
groups; interaction between scientists within our own departments, and 


with scientists in other institutions, is uncommonly high and productive, 
but "established groups" are the rare exception rather than the rule. Our 
heavy costs arise primarily from the ever-accelerating rate at which new 
instrumentation comes into being, and from the constant change in data- 
handling processes. In recent years, our resources have been insufficient 
to grasp the full promise of new instrumentation, and we have fallen behind 
in this area. But today, we are stretching our funds, upgrading our in- 
strumentation on all fronts, and I confidently anticipate further significant 
improvements in 1984-1985. 

But in view of such budgetary pressures, how does the Institution 
preserve its commitment to protect individuals whom we consider to be 
uncommonly promising, but who for some reason are reluctant to apply 
for or unable to secure government (or other restricted) grants? In the 
year ending June 30, 1983, only 67 percent of the Institution's budget was 
derived from either the Institution's endowment or unrestricted gifts or 
grants; thus restricted grants play a significant role in the Institution's 
economy. Can we be confident, then, in our ability to evaluate Staff and 
prospective Staff Members entirely apart from their success in obtaining 
outside grants? 

The problem is not ours alone, nor is it, save for ourselves, a new one. 
Universities around the nation have long been wrestling with the question 
of how to fairly evaluate faculty members apart from their success in 
competing for grants. Increasingly, advertisements in journals announcing 
faculty positions include a phrase like the following: "Candidates must be 
fundable and funded for his/her research." Unfortunately, it is a grim fact 
of life that most research universities are unable to support gifted but 
unfunded, or unfundable, scientists, except possibly for very brief periods. 

This problem is but a fragment of a larger national issue. Last year in 
this Report, I wrote of the difference between a national spirit of science 
and the national science policy. A strong national spirit of science seeks 
to encourage the bold, the unusual, the breaker of rules — in Jerome Wies- 
ner's words, "preserving the anarchy of science." But our current national 
policy seems to emphasize micromanagement and targeting. 

In a recent address, Lewis Thomas — eloquent as always — pointed out 
the growing consequences of such emphasis: 

Now, with intense competition for grants, especially among 
the youngest and brightest investigators just getting under way, 
there is a growing tendency to design one's plans for basic re- 
search problems as though one were doing applied science. The 
protocols are written out in endless and scrupulous detail just as 
though there were genuine certainty in the researcher's mind as 
to the outcome. The time for the work is meticulously pro- 
grammed to conform to the mountainous stacks of paper in the 
grant proposals almost always calculated so that publishable re- 
sults will be ready at hand when the next application of a renewal 


of a grant comes due, sometimes in a year or two. And then, 
worst of all, the lucky investigator settles down with his grant 
and does exactly what he said he was going to do, undeviatingly. 

In concluding, Thomas comes directly to the heart of the problem: 

The system guarantees two kinds of scientific catastrophe. First, 
the best of the youngest investigators will tend to take on only 
those problems that carry certainty of quick results, and a great 
many talented minds will expend themselves on safe and sound 
projects. Second, the implicit agreement to do what is promised 
in those stacks of grant papers will have the effect of excluding 
surprise from the research process. 

But if uncertainty is the hallmark of basic, or "undifferentiated," re- 
search, how is performance in research to be evaluated? Trails of research 
are complex, often labyrinthine. In his 1945 report, Science — The Endless 
Frontier, Vannevar Bush wrote: 

One of the peculiarities of basic science is the variety of paths 
which lead to productive advance. Many of the most important 
discoveries have come as a result of experiments undertaken with 
very different purposes in mind. Statistically it is certain that 
important and highly useful discoveries will result from some 
fraction of the undertakings in basic science; but the results of 
any one particular investigation cannot be predicted with accu- 

In its report "The Quality of Research in Science," the Subcommittee 
on Postperformance Evaluation of Research of the Committee on Science, 
Engineering and Public Policy wrote that assessment of research requires 
understanding of the inner logic of what is going on in any particular 
research and how it fits into a larger pattern within the field. Thus, the 
group concluded, valid methods for judging research outcomes depend 
directly or indirectly on the judgments of other scientists who are active 
in the particular field. 

It is true that peer judgment permeates the scientific endeavor; this is 
how science regulates itself. Articles are accepted for publication on the 
basis of judgments by peer reviewers. Peer judgments are the basis whereby 
funds are allocated by federal agencies. And, to an ever-increasing extent, 
the standing of a scientist within his or her own institution depends on 
the opinions of peers outside it. The usual criteria in evaluating proposed 
research are past research performance and results, and the relationship 
of the proposed work to other work in the same field. Such criteria are 
likely to be successful in judging the work of groups and projects in the 
mainstream of current theory. But they are far less likely to be successful 
in evaluating, prospectively, proposals of beginning investigators and the 
proposals of established scientists who wish to change fields or who wish 
to pursue ideas that are far out of the mainstream. In short, then, though 


much of the evidence is anecdotal, I firmly believe that today's mechanisms 
of peer review are strongly biased toward the safe and sound. 

It is thus of supreme importance that Lewis Thomas's cry be heard. 
Especially at the Carnegie Institution, opportunities must continue to be 
provided for Staff Members and fellows willing to run scientific risks for 
high stakes. I believe that our own unique system of evaluation is an 
important means to this end. 

Within our Institution, evaluation of research is primarily the respon- 
sibility of the directors of the individual departments. The directors may, 
and do, rely upon outside sources of information, including the assessments 
of granting agencies in which our Staff Members compete. (Happily, the 
Institution's scientists have a highly successful record in such competition.) 
But we continue to insist that the assessments of external reviewers should 
not substitute for informed judgment within the department itself. Mean- 
while, a department's Visiting Committee plays another role; its members 
assess the needs and opportunities in the department as a whole, serving 
both as friendly critic and as advocate for the department. Theirs is not 
the primary responsibility for judging the performance of individuals. 

Our directors, and the president and trustees, must walk a very fine 
line, balancing the need to encourage Staff Members whose discoveries 
have taken them into the mainstream, where outside funds are readily 
available, to seek such support, against the need to channel the Institu- 
tion's unrestricted funds to higher-risk, and less-recognized, ventures. 

The continuing review of performance of scientists looms larger perhaps 
in Carnegie Institution than in universities and in larger research orga- 
nizations. We must continue to innovate, but we do not wish to expand 
significantly in size. Thus our Staff Members do not have tenure in the 
sense that many universities afford. Periodic review, leading to gradual 
change, must be the rule, with emphasis placed on preserving opportun- 
ities for the uncommon man or woman. 

Meanwhile, we must periodically remind ourselves that an investigator's 
ability to attract outside grants, while desirable, does not necessarily 
reflect the Institution's raison d'etre. Future generations of Carnegie sci- 
entists should look back upon Barbara McClintock's Nobel Prize as tes- 
timony of this truth. 

"A Spectroscopic Telescope" 

The Institution has embarked upon an ambitious, yet we believe real- 
istic, program of upgrading instrumentation at Mount Wilson and Las 
Campanas Observatories, looking toward the development by the end of 
the decade of a new ground-based, low-cost telescope of large aperture. 

Such an undertaking is not new for us. As George Preston observes, 
in the first half of this century Carnegie scientists participated in the 


design, construction, and successful operation of three major optical tel- 
escopes, each of which was, successively, the world's largest in its time. 
The increased light grasp of the great instruments at Mount Wilson and 
Palomar revolutionized mankind's perception of the universe. From Har- 
low Shapley's work in the 1910's at the Mount Wilson 60-inch telescope 
on the system of globular clusters, the Sun-centered Kapteyn universe 
was replaced by a Galaxy of stars wherein the Sun with its attendant solar 
system was an "orbiting outrider." Little more than a decade later, Edwin 
Hubble, working with the Mount Wilson 100-inch telescope, presented 
evidence for an expanding universe that has preoccupied philosophers, 
theorists, and experimentalists ever since. And for more than three de- 
cades, a host of scientists have used the 200-inch Hale telescope at Cal- 
tech's Palomar Observatory to vastly expand our knowledge of the structure 
and evolution of stars, galaxies, and the universe. 

During the quarter-century after 1950, emphasis shifted from the design 
of larger telescopes to the development of improved photon detection 
methods. Available telescope technology had been pushed to the limit with 
the completion of the Hale reflector, but a revolution in electronic methods 
for detecting and storing light signals increased the power of existing 
telescopes up to a hundredfold. Currently, such efforts at the Observa- 
tories are centered on the incorporation of new detectors, the CCD's 
(charge-coupled devices), into existing instruments and on the acquisition 
of new instruments that take advantage of the remarkable properties of 
CCD's. The CCD is a two-dimensional array of silicon diodes. The diodes 
have quantum efficiencies that approach unity in the red and near- 
infrared — that is, they are a hundred times more sensitive than photo- 
graphic emulsions and up to five times more sensitive than the cathodes 
in existing image intensifiers. Because the diodes can record 200,000 pho- 
tons before saturating, their dynamic range is several hundred times that 
of photographic emulsions. Because they count individual photons, they 
are true linear photometric detectors, unlike photographic emulsions. Be- 
cause the output is digital, many CCD frames can be co-added in a com- 
puter to obtain an arbitrarily high signal-to-noise ratio in an image where 
brightness varies from place to place by a factor of several thousand. 

These devices provide many rich new opportunities. Let me cite just a 

(1) CCD's for the first time make it possible to measure accurately the 
color gradients and spatial arrangements of unresolved stars in the faint 
outer regions of galaxies. These data should provide valuable clues about 
the distributions of chemical compositions in these systems and the dy- 
namical processes that took place during the collapse of the protogalaxies. 

(2) CCD's can be used to obtain high-resolution spectra of faint stars in 
the halo of our own Galaxy and in the Magellanic Clouds. Detailed chemical 


compositions can be determined and calibrations of color measurements 
can be refined. 

(3) The CCD is well suited to determine the spatial structure and spec- 
tral characteristics of the "fuzz" surrounding many quasars. These faint 
nebulosities are generally believed to constitute evidence that quasars are 
embedded in underlying galaxies. If true, this opens a new field — the study 
of galaxies at very early times. 

(4) Spectra of many galaxies in extremely remote clusters can be ob- 
tained simultaneously with a CCD. Because these objects are so remote, 
we see them at early times in the history of the universe; thus, we can 
undertake another new kind of investigation — the evolution of clusters of 

While detectors are the heart of instrumentation in astronomy, there 
are other important items that need attention as well: the devices that 
analyze light (resolve images and measure color, polarization, and other 
properties), the data systems that store and rectify recorded photon events, 
and the laboratory computers that are used to manipulate the raw data 
to obtain meaningful results. Finally, TV systems capable of integrating 
incoming photons over many seconds are required to locate the faint ob- 
jects that can be studied with modern detectors. 

Given the ever-accelerating pace of technological change in science, the 
late 1980's and beyond promise to be even more fruitful. While anticipating 
the space telescope, we also envision important changes in ground-based 
astronomy, including the incorporation of new detectors on existing tele- 
scopes and the design and construction of new-technology telescopes. As 
Preston has written, the thirst for information about ever-fainter celestial 
objects requires renewed search for techniques to increase the collecting 
area of telescopes. The potentials of large, light-weight mirrors, seg- 
mented mirrors, active optics, and optical telescope arrays are all being 
explored in other institutions, and it is likely that one or more of these 
concepts will lead to construction of telescopes with effective apertures 
larger than 5 meters. 

It is in the context of these developments that Stephen Shectman has 
advanced optical designs for a specialized, low-cost telescope of large ap- 
erture. Access to the whole sky, large field, and desirable photometric 
characteristics are deliberately sacrificed for economy, but the end-prod- 
uct — a "spectroscopic telescope" — would be a powerful tool for the Las 
Campanas Observatory, and one that is within the means of an institution 
with modest resources. 

Shectman opens his discussion by referring to the Arecibo radio tele- 
scope, which has the largest collecting area of any telescope in the world. 
The primary reflecting surface, suspended on cables above a Puerto Rican 
valley, is 1000 feet in diameter. The primary surface itself cannot move, 



Schematic diagram for a possible 5-meter telescope with a fixed, 
spherical primary mirror. The primary mirror (shown at the bottom) 
has a radius of curvature of 8 meters. The movable secondary optics 
(consisting of secondary and tertiary mirrors of diameters 98 and 
38 cm, respectively) pivot about the center of curvature. Light from 
the sky (entering at top of drawing) is reflected upward at the fixed 
primary, then downward at the secondary mirror, and finally up- 
ward at the tertiary. The two-element secondary optics is used to 
cancel spherical aberration and coma induced by the fixed primary 

but the telescope is made to point in different directions on the sky by 
adjusting the position of a small secondary "feed" near the focus of the 

Because most of the telescope is immobile, the cost per collecting area 
of the Arecibo telescope is a factor of ten less than the cost for a conven- 
tional radio telescope. The penalty for the low cost is that the telescope 
is restricted to point within 18 degrees of the zenith. Nevertheless, the 
Arecibo radio telescope is one of the most useful and productive in exis- 

Available observing time on existing optical telescopes limits opportun- 
ities for spectroscopic observations. But in spectroscopic work, the re- 
strictions imposed by a large but low-cost optical telescope, patterned 
upon the Arecibo radio telescope, will often be acceptable. Such a telescope 
has been discussed for at least thirty years, but a complete investigation 
of the optical design has never been published. 

The primary mirror for such a telescope must be spherical, because only 
a sphere has complete symmetry about the center of curvature. As the 
secondary optics will pivot about this point, different sections of the pri- 
mary surface are illuminated, but the surface figure is always the same 
sphere. The primary focal ratio must also be very fast, so that as the 
secondary optics are moved away from the zenith, the motion of the il- 


luminated patch on the primary is minimized. Because the primary is 
completely stationary, it may be possible to construct it from hundreds 
of identical, smaller segments mounted on a very stable base. The cost 
per unit area of the small segments is much lower than that for large 
conventional mirrors. 

There are problems, of course, and Shectman is addressing them. At 
this writing it appears likely that preliminary experiments to test the 
viability of this concept and the feasibility of building such a telescope will 
be performed, beginning during 1984-1985. 

Galaxies and Cosmology 

Globular clusters have played a prominent role in astronomical research 
throughout this century. As I have already remarked, more than sixty 
years ago Shapley, working at Mount Wilson, used their distribution on 
the sky to identify the center of our Galaxy at a location remote from the 
Sun. The properties of globular clusters continue to be used to study the 
dynamics and chemical evolution of the Galactic halo. As families of old 
coeval stars, they provide the means to test and then apply theories of 
stellar structure and evolution to such diverse problems as age-dating of 
the Galaxy and reconstruction of the nucleosynthetic history in the inte- 
riors of post-main-sequence stars. Several studies along these lines are 
contained in the Observatories Report. Now, somewhat unexpectedly, 
globular clusters have been used by Leonard Searle as probes of the 
interstellar dust in another galaxy. 

Searle conceived a program to investigate the chemical compositions, 
kinematics, and spatial arrangements of globular clusters in M31, the great 
galaxy in Andromeda. Initially, Searle regarded dust, located in the sym- 
metry plane of M31, merely as a nuisance that dims and reddens the light 
of clusters on the far side of the galaxy. His observations consisted of 
low-resolution spectral scans obtained with the multichannel spectrometer 
of the Palomar 5-meter telescope. By matching spectral features he could 
identify pairs of similar clusters, one of which was in front of and the 
other behind the absorbing layer. From these pairs he could derive the 
wavelength dependence of the absorption (the reddening law) in various 
lines of sight through M31. The notion of a radial gradient in the reddening 
law of a galaxy opens new avenues of exploration into an unexpected 
characteristic of the interstellar medium in stellar systems. 

Astronomers at the Department of Terrestrial Magnetism (DTM) have 
also reported new results bearing on several basic scientific problems — 
the time scale of the universe, differences between the early universe and 
the present, the total mass of the universe, and the way in which certain 
galaxies are formed. 

Vera Rubin, Norbert Thonnard, and W. Kent Ford, in collaboration 



NGC 5139, Omega Centauri, a globular cluster of stars. A globular cluster is a tightly 
packed, symmetrical group of thousands of very old stars. Omega Centauri is the largest 
globular cluster in our Galaxy. Such clusters can be age-dated from measured properties 
of their stars. As such clusters are the oldest stars in the Galaxy, such ages provide 
the age of the Galaxy itself. This photograph was obtained by former Carnegie fellow 
Gary Da Costa with the Institution's 2.5-meter du Pont telescope at Las Campanas, 


with David Burstein of Arizona State University, Tempe, report a new 
determination of the Hubble constant — the fundamental quantity that 
relates the redshift of a galaxy to distance and hence defines the age of 
the universe. The new value derives from the group's earlier discovery 
of the relationship between the intrinsic luminosity of a galaxy and its 
rotational dynamics. This relationship contains the Hubble constant as a 
parameter. Therefore if the luminosity of even a single galaxy can be 
independently determined, a value of the Hubble constant can be calcu- 
lated. The luminosity of the nearby spiral galaxy M33 is known well enough 
to be used in this way; two other galaxies with less accurately determined 
luminosities were also included in the calculation. The resulting value for 
the Hubble constant is about 60 km sec" 1 Mpc -1 , which falls near the low 
end of the 50-100 km sec -1 Mpc -1 range found by other workers. 

The findings of Rubin, Ford, Thonnard, and their collaborators on the 
"missing mass" in spiral galaxies has been discussed extensively in earlier 
Reports. The fundamental principle involved in these measurements is 
that the velocities of stars near the edge of a visible galaxy can serve as 
a probe to test the gravitational field and thereby the mass of the galaxy 
interior to these stars. Linda Schweizer describes the use of gravitation- 
ally bound pairs of galaxies as mutual probes of one another's mass. Be- 
cause only one component of velocity can be measured (along the line of 
sight), the orbits and thereby the masses of individual pairs cannot be 
uniquely determined. Measurements of a large sample of paired galaxies, 
however, permit probabilistic inferences regarding the radial distributions 
of mass. She finds that the ratio of nonluminous, invisible mass to visible 
mass based on galaxy pairs is about a factor of ten greater than that 
estimated for single galaxies using rotational velocity data. When com- 
bined with the further result that the orbits of these pairs are on the 
average highly eccentric, it seems likely that the two galaxies actually 
pass through each other's invisible "halo." 

Until recently it was thought that the division between spiral and el- 
liptical galaxies was primordial in origin. Following earlier theoretical 
studies by Alar Toomre at MIT, Frangois Schweizer has been gathering 
observational evidence for an alternative hypothesis for the origin of el- 
liptical galaxies — that they are formed by the collision and merger of spiral 
galaxies. This year, Schweizer reviews in a more general way the evidence 
for galactic mergers of all kinds, and he proposes a classification scheme 
that systematizes this evidence. So far, about fifty mergers have been 
classified; a larger sample, planned in the near future, should help to clarify 
the similarities and differences between galaxy mergers of various kinds 
and form the observational basis for more-secure inferences as to their 
significance in galactic evolution. 

In a related investigation, F. Schweizer, Rubin, and Bradley Whitmore 
searched photographs of "peculiar" galaxies for examples similar to the 


unusual "spindle" galaxy described last year. Such galaxies appear to have 
a cigar-shaped central spindle ringed by a circular band of stars in a plane 
perpendicular to the spindle's axis. They concluded that the spindle de- 
scribed last year was only apparent, being a disk galaxy seen edge on. 
Nevertheless the associated ring, orbiting the disk galaxy in a plane close 
to its poles, represents a remarkable phenomenon. The ring was also 
useful, as its rotational velocities were used to identify the presence of 
an invisible halo outside the plane of the galaxy. 

The present survey resulted in the discovery of 22 new examples of 
such polar-ring galaxies, including several face-on cases. It now appears 
that polar rings are not always associated with edge-on disks. Observa- 
tional bias simply favors their detection edge on. 

Another important result of this survey is the discovery that most polar 
rings are in binary or multiple systems. The investigators propose that 
these rings are formed by tidal mass transfer following close encounters 
between orbitally bound galaxies. They explain how this provides a natural 
way to understand the otherwise surprising observation that most polar- 
ring galaxies belong to the relatively uncommon Hubble type SO. In the 
words of the authors, "galaxies do not evolve in splendid isolation," but 
continually interact throughout their evolutions. 

Studies of field elliptical galaxies have disclosed relations between lu- 
minosity (and by inference, mass) and velocity dispersion on the one hand 
and between luminosity and metallicity on the other. These results have 
proven invaluable in theoretical investigations of galaxy formation. The 
correlation of luminosity with velocity dispersion requires the existence 
of a mass-radius relation and indicates that galaxies were formed in a quite 
homogeneous way according to rather simple scaling laws. Scatter in the 
relations can be intrinsic; it also may be due to errant distances derived 
from redshifts because of perturbations of the Hubble flow by, for example, 
the gravitational attraction of parts or all of the Local Supercluster. To 
eliminate the effects of distance errors, Alan Dressier of the Observatories 
has constructed luminosity-velocity dispersion and luminosity-metallicity 
relations for 53 elliptical galaxies in the Coma and Virgo clusters. The 
relations are tight and similar in both clusters. The slope of the luminosity- 
velocity dispersion relation is steeper in the clusters than it is for field 
galaxies, and the correlation of metallicity with luminosity is much stronger 
in the Coma- Virgo sample. This result suggests that distance errors are 
a problem in the field sample. Further, the magnitude differences between 
the Coma and Virgo samples imply that Coma is 6.0 times farther away 
from us than Virgo, while the ratio of recessional velocities requires the 
distance ratio to be 7.1. Dressier notes that the discrepancy can be re- 
moved if the Local Group (of which our Galaxy is a member) has a velocity 
perturbation of 230 km sec ~ : toward the Virgo cluster which could be 
attributable to the gravitational attraction of the Local Supercluster. Dres- 



A number of astronomers worldwide are investigating the nature of galaxy clusters, 
where several or many galaxies interact with one another in a coherent system adhering 
to the laws of physics. Shown above is part of the Hercules cluster, Abell-2151, which 
is located 700 million light years away. A concentration of galaxies of various sizes and 
shapes can be recognized superposed on a myriad of foreground stars. Alan Dressier of 
the Mount Wilson and Las Campanas Observatories has been investigating the internal 
dynamics of galaxies in certain clusters; he is thereby developing a new method for 
obtaining the ratio of distances between such clusters. This photograph was obtained 
by Dressier using the 2.5-meter du Pont telescope at Las Campanas. 


sler's value agrees well with those reported last year by Sandage and is 
much smaller than the value of 410 km sec -1 determined from the direc- 
tional asymmetry of the cosmic microwave background. 

Allan Sandage has continued this year to address aspects of several 
closely related problems — the distance scale, the expansion rate, and the 
age of the universe. In his first essay, he reexamines the ages of the 
globular clusters M92 and M15 — two extremely metal poor systems that 
in all current concepts of galaxy formation must be among the Galaxy's 
oldest inhabitants. His result: the age of both clusters is (18 ± 2) x 10 9 
years, in good agreement with a determination reported by Sandage last 
year. In his second essay, Sandage reports progress in establishing dis- 
tances to several nearby galaxies by use of stellar standard candles. In a 
final report, with Gustav Tammann, he reconsiders the calibration of the 
infrared Tully-Fisher relation, yielding H = 59 ± 10 km sec -1 Mpc -1 . 
Tammann and Sandage raise the question of whether the close agreement 
of the Hubble time (for H = 50), the age of the globular clusters, and 
the ages of actinide elements (as recently reported) can be regarded as 

History of the Earth and Solar System 

How was the Earth formed? George Wetherill suggests that the obvious 
approach to the question is to study the rocks of the Earth itself, partic- 
ularly the most ancient rocks one can find. But, as he writes, because the 
Earth is an active planet, this is not easy. As one seeks to penetrate 
farther back in time, the effects of erosion, metamorphism, large-scale 
fault motion, continental rifting, and subduction all hide the earliest his- 
tory of the Earth behind an almost impermeable veil. James Hutton's 
pronouncement in 1785 that in the rocks of the Earth one can find "no 
vestige of a beginning" remains nearly correct. 

Yet it is not entirely correct. The isotopic ratios of lead convey clear 
evidence that the Earth was formed at a quite finite and measurable time 
in the past — 4.55 ± .05 billion years ago. The occurrences of rocks as old 
as 3.7 billion years is well established, and subtle but probably real dif- 
ferences between the isotopic and trace element chemistry of these ancient 
rocks and their modern counterparts is emerging. Lunar exploration has 
provided a window on an even earlier period, characterized by massive 
bombardment by planetesimals that were themselves remnants of the time 
when the Earth and Moon were formed. Today, meteorites provide a 
tangible and detailed record of some aspects of this earlier epoch in solar 
system history. 

In addition, there are observational and theoretical constraints on the 
way in which the Earth was formed. The formation of the Earth was a 
byproduct of the formation of a star, and observational data relevant to 


star formation, particularly in the infrared, are accumulating rapidly. 
Meanwhile, recent efforts by several investigators in the Department of 
Terrestrial Magnetism have increased understanding of the formation of 
the solar system, the initial state of the Earth, and the way this initial 
state constrains its subsequent geological evolution. 

There is little doubt that the Sun and planets originated in a dense 
interstellar cloud. A body of isotopic evidence indicates that certain me- 
teorites still contain distinguishable relict signatures of heterogeneities 
within this primordial material. One of these meteorites is the carbona- 
ceous chondrite Allende. Previous work with this meteorite has shown 
that about 5% of its 27 A1 was produced by a "last-minute" source of newly 
synthesized aluminum nuclei, i.e., within about one million years of the 
formation of the solar system. Measurements on chromium isotopes in 
Allende reported this year by Typhoon Lee and Fouad Tera show that if 
this last-minute source produced a similar fraction of the solar system's 
chromium, this late source was astrophysically different from those that 
produced the bulk of the material in the solar system. This late source 
was much more neutron-rich, probably represents the deep interior of a 
massive star, and may be responsible for the production of the excess 
quantities of neutron-rich isotopes 48 Ca and 50 Ti previously found in this 

Alan Boss is carrying forward his program of numerical studies of the 
ways whereby a dense interstellar cloud collapses under its own gravity 
to form stars. He has extended his numerical techniques to include the 
effects of increasing central temperatures during the late stages of col- 
lapse. The goal of this work is to identify the conditions under which an 
interstellar cloud fragment can collapse into a single star like the Sun, 
instead of fragmenting further into a multiple star system. 

He finds that even for very favorable initial conditions, rotational in- 
stability leads to fragmentation into a binary or multiple system before 
the collapse is complete. He also finds that if one member of a multiple 
system is gravitationally perturbed enough to cause it to escape from the 
multiple system, it will still fail to form a single star. Instead, it will 
fragment further into a new multiple system. The possibility still remains 
open that these late-stage binary stars can gravitationally transfer enough 
of their angular momentum to more-distant parts of the interstellar cloud, 
and thereby merge into a single star. If this can be shown not to occur, 
it will demonstrate the need for other mechanisms of angular momentum 
transfer, such as viscosity or magnetic fields. Such mechanisms may have 
played a part in the formation of gas-rich planets, such as Jupiter and 
Saturn, from the collapsing cloud. 

A principal unresolved question concerning the formation of the ter- 
restrial planets of our solar system is the timing of the removal of the gas 
from the solar nebula wherein the planets formed. If the Earth and other 


terrestrial planets accumulated in a gas-rich nebula, the orbital dynamics 
of the accumulating planetesimals would differ in important ways from 
those expected for the alternative, gas-poor accumulation process. One 
way to attack this question is to work out the consequences of a particular 
mode of accumulation in sufficient detail to permit comparison with what 
is actually observed on the planets. 

One clear predicted consequence of Earth accumulation in a gaseous 
nebula is the gravitational capture of a primordial atmosphere equal in 
mass to several percent of the Earth's present mass. It is possible that 
this atmosphere could have been removed within about 100 million years 
by solar ultraviolet radiation, so its present absence is not necessarily a 
problem. However, such an atmosphere may not have been possible with- 
out incorporating within the Earth enough neon of solar isotopic compo- 
sition to conflict with the nonsolar isotopic composition of neon today. 
Hiroshi Mizuno and Wetherill report a rather detailed investigation aimed 
at defining the conditions under which one can hope to avoid this neon 
problem. The key factor is the grain opacity of the primordial atmosphere, 
because this controls atmospheric radiative loss, atmospheric structure, 
and surface temperature and pressure. They find that avoidance of the 
excess neon problem requires rather restrictive conditions as to the he- 
liocentric orbital eccentricities of the accumulating planetesimals, as well 
as the efficiency with which dust grains are produced by planetesimals as 
they pass through the atmosphere. 

Another "neon problem" relating to the formation of the terrestrial 
planets is the similarity of the 36 Ar/Ne ratio on Venus, Earth, and Mars, 
despite the strong difference between this ratio and its original solar value. 
In an article discussing several aspects of the Venus atmosphere, Thomas 
Donahue, at DTM while on sabbatical leave from the University of Mich- 
igan, describes a mechanism that could be at least partly responsible for 
this result. As planetesimals grow to "lunar size" in the solar nebula, in 
about one million years they will capture significant primordial atmos- 
pheres. After removal of the solar nebula gas at this stage, these atmos- 
pheres will partially escape; the lighter inert gases will escape more readily 
than the heavier ones. Donahue finds that small planetesimals will lose all 
their gases, while others will retain only xenon, krypton, and argon. Larger 
bodies will retain neon as well. Planets assembled from similar mixtures 
of these bodies could thereby achieve similar proportions of the various 
inert gases without requiring that temperature-sensitive fractionation 
mechanisms operate in nearly the same way on all three planets. In this 
article, Donahue discusses other calculations, also published elsewhere, 
showing that it is quite likely that early in its history Venus had an ocean 
possibly as large as that of the Earth. This is implied by the observed 
high ratio of the heavy isotope of hydrogen (deuterium) to ordinary hy- 
drogen in the present atmosphere of Venus. 


Richard Carlson is investigating isotopic and geochemical characteris- 
tics of the oldest available terrestrial rocks, seeking evidence as to the 
extent and nature of magmatism and associated chemical differentiation 
in the primitive Earth. In his report, he describes isotopic data obtained 
on rocks from the ancient gneiss complex in Swaziland. The 143 Nd/ 147 Sm 
age of 3417 ± 34 million years found for these rocks is younger than the 
age of 3526 ± 48 million years found by workers at Columbia University 
for the Onverwacht volcanics, the surficial member of this ancient section 
of the African continent. This age relationship agrees with the hypothesis 
that the formation of early buoyant crust was initiated by the formation 
of a thick volcanic pile, the roots of which soon remelted to form the less- 
dense granitic gneiss terrane. 

The initial neodymium isotopic composition of the rocks studied by Carl- 
son shows that 3400 million years ago igneous rocks, like modern volcanic 
rocks, were derived from a mantle source depleted in light rare earth 
elements by earlier partial melting. Carlson and Diana Diez de Medina 
reported the same result last year for 3000-million-year-old gneiss from 
the Big Horn mountains in Wyoming, and similar results have been ob- 
tained by other workers on other ancient rocks. This evidence of early 
differentiation is consistent with the expected thermal history of the hot 
initial Earth predicted by current models of terrestrial planet formation. 
The abundant data on young volcanic rocks, however, cannot be reconciled 
with a simple model involving only early differentiation. It appears that 
an "antidifferentiating" process must be operating throughout geological 
history; previously depleted mantle material is thereby remixed with its 
enriched counterpart. Several such mechanisms have been proposed, but 
it is not yet clear that they will turn out to be quantitatively satisfactory. 

Accelerator Mass Spectrometry: Beryllium- 10 Studies 

The past five years have witnessed the development of a major new 
technique for the measurement of extremely small quantities of naturally 
occurring isotopes. This technique, termed accelerator mass spectrome- 
try, makes use of a nuclear accelerator, such as a tandem Van de Graaff 
accelerator or a cyclotron, to form ion beams of a sample's constituent 
isotopes. At the resulting high energies of tens of millions of electron- 
volts per ion, the usual mass spectrometric techniques of magnetic and 
electrostatic isotope separation are replaced by techniques, more familiar 
in nuclear physics, where different nuclei are identified by their energy 
loss in absorbing materials. Also, because of the differing stability of 
negative ions of different chemical species during their flight through the 
accelerator, very rare isotopes (such as 10 Be and 2G A1) can be separated 
from the much more abundant species of the same atomic mass ( 10 B and 
26 Mg). Thus the naturally occurring radioisotopes 14 C, 10 Be, 26 A1, and 36 C1, 


heretofore measurable only with radiation-counting techniques, can now 
be measured with much greater sensitivity by mass spectrometry. 

For several years, Louis Brown, Fouad Tera, and I. Selwyn Sacks at 
DTM have been collaborating with Roy Middleton and Jeffrey Klein at 
the University of Pennsylvania Tandem Accelerator Laboratory in de- 
veloping and applying this new technique. They have concentrated on 
measuring 10 Be and applying the measurements to problems of earth sci- 
ence. During the past year, the work was extended to include measure- 
ments of 26 A1. These studies have attracted further collaborators: Gary 
Stensland of the Illinois State Water Survey, Milan Pavich of the U. S. 
Geological Survey, and Nathalie Valette-Silver of the University of Per- 
pignan (France). 

The naturally occurring radioactive isotope 10 Be is produced in the Earth's 
atmosphere by cosmic ray bombardment of the nuclei of atmospheric gases. 
It falls to the surface attached to aerosols in raindrops, and is absorbed 
onto soil particles. Soil is thereby labeled with a 10 Be signature that can 
be used to trace its subsequent erosion, transport, sedimentation, and 
even its subduction at oceanic trenches into the Earth's mantle. The 1.5- 
million-year half-life of 10 Be confers an additional advantage to its use as 
a tracer. Because it "self-destructs" by radioactive decay, the natural 
systems under study are automatically purged of the tracer after a few 
million years, and are not "polluted" by its accumulation over all of Earth 

In all 10 Be studies, a fundamental parameter is the rate at which it is 
produced in the atmosphere. Because of practical experimental difficulties 
this value is not yet well known. The most straightforward way to measure 
it is by measuring the 10 Be content of collected rainwater. Earlier attempts 
to do this, however, are suspect because of contamination of the rain 
sample by wind-blown soil containing 10 Be. This year, new rain measure- 
ments were made using new sampling devices that are open only during 
rainstorms. (These measurements also exploited other chemical and iso- 
topic techniques to detect contamination by soil.) The 10 Be production rate 
found is about 40% lower than the best previous result. This value may 
still be too high, and plans are being made to repeat the measurements 
on islands where soil contamination should be much lower, as well as at 
a tropical site where the rainfall is much greater. 

Probably the most dramatic application of 10 Be is in tracing the contri- 
bution of oceanic sediments to island-arc volcanism. Bound to surficial 
oceanic sediments, 10 Be is carried down to depths of 100 km or more in 
subduction zones bordering oceanic island arcs such as the Aleutians. Last 
year it was first shown that a small but clearly measurable portion of this 
subducted 10 Be found its way into the magmas now erupting as lavas on 
volcanic islands. This year the quantity of data has been doubled, and the 
earlier results have been amply confirmed. Furthermore, these newer 


data show that the 10 Be content of island-arc volcanic rocks is highly 
variable; high values are found in Aleutian and Central American volca- 
noes, low values in the western Pacific. As before, very low values are 
found for volcanoes not associated with subduction zones. The work has 
now evolved from the discovery into the exploration stage. The combi- 
nation of more-extensive 10 Be measurements with other geochemical and 
geophysical data may be expected to contribute much to our understanding 
of the fundamental geologic process of lithosphere subduction. 

It also rains on the continents, and therefore ordinary soil contains 
significant concentrations of absorbed 10 Be — about 2 x 10 ~ 15 grams per 
gram of soil (10 8 atoms/gram). Accordingly, as described in this year's 
Report, 10 Be measurements are contributing to our ability to determine 
the ages of soils, their history of erosion, their transport by rivers, and 
their deposition as sediments. In addition, very interesting but yet poorly 
understood data have been obtained on the migration of 10 Be from soils 
into the underlying bedrock and on how this relates to changes in the 
depth of the water table. 

In one study, 10 Be was measured in sediments deposited near the mouths 
of rivers. These data were used to calculate the amount of 10 Be carried 
annually by the river, which was then compared with the amount of 10 Be 
deposited each year by rainfall over the river's watershed. Except for the 
radioactive decay of 10 Be, in the steady state one would expect this output 
and input to balance. Because some soils are millions of years old and 
some of the 10 Be has decayed, the steady state 10 Be balance should actually 
indicate less output than input. In contrast, in every one of the nine rivers 
studied the output exceeds the input by factors ranging from as low as 
1.2 for the Susquehanna, to 14 for the Huang Ho, to 24 for the Narbada 
(India). The obvious interpretation of these results is that a steady state 
does not exist and that soils are now being eroded much more rapidly 
than they are being formed. It is possible that agriculture is responsible 
for this increased erosion. Geological effects, particularly those associated 
with the relatively recent ice age, may also be in large part responsible. 

The Earth's Upper Mantle 

The upper mantle underlying the Earth's crust begins at a depth about 
5 km below the ocean floor, and at a depth of about 35 km in continental 
regions. But despite its nearness to us, not many years ago the mantle 
seemed an almost imaginary place whence basalts arose. Penologists and 
geochemists argued whether the mantle was made of essentially pure 
olivine or the high-pressure form of basalt called eclogite. 

But in recent years, as Wetherill observes, we have learned much more 
about the upper mantle, principally that it is far more complicated than 
we used to think. "Mantle geology" resembles conventional geological 


mapping in that an enormous quantity of arduous work must be done, and 
much information must be gathered and examined before the general 
unifying principles will become clear. It differs from conventional geology 
in that we cannot visit the field areas directly but must rely on "remote 
sensing" techniques, such as seismology and geochemical investigations 
of volcanic rocks derived from the mantle. 

In spite of these difficulties, Selwyn Sacks, with J. Arthur Snoke of 
VPI, is advancing knowledge of the structure of the continental and sub- 
oceanic uppermost mantle. From the seismological perspective, the sub- 
oceanic mantle seems relatively simple to understand. The upper layer, 
or lithosphere, consists of strong solid rock that migrates away from the 
oceanic ridges where it formed. Beneath the lithosphere and moving along 
with it is the weaker asthenosphere, also formed at oceanic spreading 
centers. As the asthenosphere cools, its conversion into lithosphere can 
be understood in terms of physical principles of heat conduction. In con- 
trast, the nature and position of the continental lithosphere-asthenosphere 
boundary is variable and much more difficult to define; in some ancient 
"shield" areas of the continents, the distinction may indeed have little 

In a collaborative investigation with Thomas Boyd, also of VPI, and 
Anibal Rodriguez B. of the Universidad de San Agustin in Arequipa, Peru, 
Sacks and Snoke report new data on the base of the continental lithosphere 
beneath western South America. In this region of continental-oceanic plate 
collision, the continental lithosphere is truncated by a slab of subducted 
oceanic lithosphere. Extending earlier work by Akira Hasegawa and Sacks, 
the precise location of about 2500 earthquakes occurring within the sub- 
ducted slab during three one-year periods were determined. These earth- 
quake sources "decorate" the migrating slab and permit detailed mapping 
of its position beneath the continental lithosphere. The boundary between 
these regions is found to be highly deformed yet continuous. Over much 
of the region, the slab of oceanic upper mantle moves nearly horizontally 
at a depth of 125 km. Ultimately this horizontally moving slab must en- 
counter the great "root" of the South American continental shield. At this 
juncture, the slab is believed to plunge sharply to much greater depths. 

In related laboratory work, Sacks and Tsutomu Murase carried out an 
experimental study on the physical nature of the asthenosphere. One way 
to identify an asthenospheric region is by its value of "Q," a parameter 
describing the attenuation suffered by seismic waves in traversing the 
region. High Q corresponds to little attenuation, low Q to much atten- 
uation. It is often hypothesized that the low Q of the asthenosphere is a 
consequence of it being partially melted — an idea that fits in well with 
the belief that the asthenosphere is the source of volcanic magmas. These 
experiments show that upon heating, the Q of the rock studied drops to 
low values at temperatures well below its melting point. Thus, if there 


are partial melts in the asthenosphere, the extent of melting seems to be 
very small. 

The suboceanic upper mantle may appear simple to a seismologist, but 
the detailed information obtained by geochemical and isotopic measure- 
ments reveals that it is actually a quite complex and heterogeneous place. 
The isotopic composition of strontium and neodymium in oceanic volcanic 
rocks shows that the suboceanic upper mantle is predominantly a "de- 
pleted" mantle, having lost much of its content of elements with large 
ionic radii — potassium, rubidium, light rare earths, and uranium. Much 
of the lost material now resides in the continental crust, where these same 
elements are highly concentrated. 

The oceanic mantle is not uniformly depleted, however. It has been 
known for quite a while that lavas erupting from the mid-ocean ridge 
system are derived from mantle sources more depleted than those sup- 
plying volcanic magmas to oceanic islands such as Hawaii. David James 
and Emi Ito present evidence from Samoa for a rarer type of suboceanic 
mantle source, which is actually enriched relative to the bulk Earth in 
large ion elements. The possibility that this enrichment is only apparent, 
resulting from assimilation of crustal material by the ascending magma, 
is rejected on the basis of oxygen isotope data indicating a mantle rather 
than a crustal signature. 

Another type of mantle source, also presumably originally suboceanic, 
is discussed by Barbara Barreiro. In New Zealand, certain volcanic rocks 
appear to have originated in a source that is enriched in the large-ion 
elements neodymium and uranium, but depleted in rubidium. It is pro- 
posed that this is the result of "mantle metasomatism," where carbon 
dioxide-rich fluids have enriched those large-ion elements soluble in car- 
bon dioxide, but not those (such as rubidium) that are insoluble. Similar 
data have existed for some time from the oceanic islands of St. Helena in 
the Atlantic and Tubuai in the South Pacific. 

It is interesting that the same mechanism can explain the results re- 
ported this year by Meen on rocks from the Absaroka mountains in Mon- 
tana. In this case the rocks are continental, rather than oceanic, and the 
metasomatism took place 2800 million years ago, rather than only a few 
hundred million years ago, as is the case in New Zealand. Both areas 
seem to be exhibiting evidence for a particular geochemical process that 
has played a role in chemically differentiating the Earth's mantle over 
much of earth history. 

The identification of the chemical nature of the mantle is particularly 
complex and challenging in regions where magmas derived from oceanic- 
like mantle sources must penetrate continental crust as they ascend to 
the surface. James reports the use of oxygen, strontium, and neodymium 
isotopes and other chemical data to unravel the complexities of the sources 
of volcanic rocks in the northern Andes. Although at first glance the 


isotopic data from these rocks do not appear to provide evidence for crustal 
involvement, James shows that upon closer examination, it is likely that 
not one but two crustal sources, as well as a mantle source, are involved. 
One type of "crustal contamination" occurs by assimilation of overlying 
crustal rocks as the magmas rise to the surface. The other is contamination 
of the mantle source itself by subduction of crustal material, including 
sediments of continental origin. 

The effects described by James occur in a region of convergent tectonics 
between a continental and oceanic plate. The opposite tectonic mode — 
rifting and crustal extension — can also occur in a continental or near- 
continental environment. Richard Carlson and William Hart describe their 
work in Oregon, where rocks now part of the continent were formed only 
15 million years ago, probably in an oceanic environment near the conti- 
nent. Their source was depleted mantle, similar to the source of oceanic- 
island and island-arc basalts. In the western portion of the area studied 
there is little evidence for crustal involvement. However, farther to the 
east, the mantle signature is overprinted by a contribution that may rep- 
resent continental crustal contamination or subcontinental mantle. Similar 
results are found by Hart in Afar in Ethiopia. Here, rifting is occurring 
on an existing continent. Although clearly affected by crustal assimilation, 
these magmas seem to have also tapped a mantle source quite similar to 
that reported in Oregon. 

The generation of volcanic rocks at the mid-ocean ridges and in oceanic 
islands must be intimately related to solid-state convective motions in the 
mantle. In particular, the formation of volcanic islands such as the Ha- 
waiian islands, the Samoan islands, and Iceland are usually thought to be 
associated with the formation and up welling of low- viscosity mantle plumes, 
the surface expression of which are "hot spots" of increased volcanic ac- 
tivity. During the past year, Boss and Sacks have carried out numerical 
experiments on the generation of mantle plumes at the core-mantle bound- 
ary; their results compare satisfactorily with theoretical results obtained 
by Stephen Morris of the University of California, Berkeley, who spent 
several months visiting DTM this winter. 

These numerical calculations show that a moderately large temperature 
perturbation at the core-mantle boundary can disrupt the preexisting man- 
tle flow pattern and lead to the generation of a plume. The plume rises 
to the top of the mantle on a time scale of about 100 million years. A down- 
welling "negative plume" then develops, and after about 350 million years 
a new episode of plume upwelling begins. Although highly idealized, these 
calculations indicate that the time scale for the growth and decay of mantle 
plumes is much longer than the approximately million-year episodicity 
observed in oceanic-island chains. This short-period episodicity must 
therefore have a more surficial and local cause. 


Mantle Metasomatism 

Metasomatic rocks are those whose chemical compositions have been 
changed in the solid state by reactions with various fluids. The chemical 
changes may be modifications in the composition of primary minerals or 
the replacement of these minerals by new ones. Metasomatizing fluids can 
be mixtures of volatiles, especially C0 2 and H 2 0, that carry mineral com- 
ponents in solution. Alternatively, the fluids are melts, commonly charged 
with volatiles. Many ore deposits have been produced by metasomatism, 
and the fluids carrying ore components are usually volatiles that infiltrate 
wall rocks enveloping a magmatic intrusion. Metasomatic rocks can also 
be produced by reactions between beds of contrasting composition (e.g., 
limestone and shale) during regional metamorphism. The scale of meta- 
somatic events that have occurred in the Earth's crust can often be de- 
termined by straightforward field investigation. 

The nature and scale of mantle metasomatism, however, is much less 
easily understood, according to Francis R. Boyd of the Geophysical Lab- 
oratory, even after an extensive study (with Richard A. Jones and Peter 
Nixon) of xenoliths in kimberlites. The geologic record contained in xeno- 
liths carried to the surface in volcanic eruptions from great depths is 
fragmentary, and the scale is difficult to determine because few xenoliths 
exceed 20 cm in diameter. The primary spatial relations between xenoliths 
are commonly obscured because of the turbulent processes by which they 
are emplaced. Ambient temperatures in the mantle are higher than those 
in the crust; for this reason alone, metasomatism might be more wide- 
spread in the mantle than in the crust. 

Depletion of mantle peridotites by partial fusion and extraction of liquid 
can be partially compensated by metasomatic introduction of elements, 
including the alkalis Fe and Ti. Thus, "sterile" peridotite (a residue of 
magma-forming events) can be "fertilized" by metasomatism. Further- 
more, the processes of depletion and enrichment may be repetitive. Some 
alkali-rich magmas, including kimberlites, are believed to have formed by 
partial fusion of rocks that were first depleted by magma removal and 
then enriched by metasomatism. Some mantle rocks found in kimberlites, 
however, appear to have been metasomatized during the early stages of 
the volcanic eruptions that brought them to the surface. 

High-Pressure, Fourier-Transform Infrared Studies 

The temperature and dynamics of earth and planetary interiors are 
calculated from models based on geophysical and experimental data. Ther- 
modynamic properties, such as the Gruneisen parameter, heat capacity, 


and entropy of planetary materials, are the most crucial information for 
these calculations; yet in the past these properties could not be measured 
directly at pressures above 50 kbar. Extrapolations were made on esti- 
mates greatly dependent on the assumed hypothesis. As a result, most 
thermal models of the Earth and planets are at best ambiguous. With the 
recent development by Ji-an Xu, Ho-kwang Mao, Ke-nan Weng, and Peter 
Bell at the Geophysical Laboratory of techniques for measuring infrared 
and Raman spectra of lattice phonons at ultrahigh pressure, it is now 
possible to measure these properties under relevant conditions and to 
construct more-reliable planetary models. 

Research with the new Fourier-transform infrared facility at the Lab- 
oratory has resulted in the measurement of the intrinsic thermodynamic 
properties of solidified gases and minerals under the same high-pressure 
conditions to which they are subjected in planetary interiors. Experiments 
were performed with the diamond- window, high-pressure cell on solid 
hydrogen, the olivines (forsterite and fayalite), a pyroxene (hypersthene), 
and the ultrahigh-pressure phase, silicate perovskite. Unusually high qual- 
ity Fourier-transform interferometric spectra were obtained, which can 
now be used to determine mineral phase relations and seismic properties. 

The new data were obtained mostly in the mid-infrared region between 
4000 and 400 wave numbers. The ultimate goal of this research will prob- 
ably be realized in the far-infrared region between 400 and 50 wave num- 
bers, where the phonon spectra can be measured directly. (Current 
experiments indicate that this spectral region is now attainable.) The 
Gruneisen parameter, which relates thermodynamic and elastic proper- 
ties, can be determined from the mid-infrared and the phonon spectra. 
Many attempts have been made in the past to obtain the Gruneisen pa- 
rameter, with but limited success. The prospect of obtaining such data 
directly for the first time is a dream shared by most geophysicists. Con- 
siderable effort is still needed to refine the technique, but the design 
parameters now appear to be feasible. 

Heat Transfer in Metamorphism and Magma Generation 

The large volumes of magma that issue from volcanoes are believed to 
be generated from crystalline crust or mantle. Of the many conceivable 
processes in melting, one involves the transfer of heat from a refractory 
mass to a contiguous mass of lower melting temperature. At the Geo- 
physical Laboratory, Hatten Yoder has devised a laboratory experiment 
to measure the relationship between heat transfer and the rate of liquid 
production and to define the physicochemical constraints inherent in the 
process. In preliminary experiments, it was necessary to measure the 
heat transfer in the absence of melting so that the heat effects of the 


melting process itself could be evaluated. Yoder discovered that a thermal 
gradient could be imposed on the sample at a rate greatly exceeding the 
known rates of metamorphic reactions and melting of common rock-form- 
ing minerals. Such decoupling of the thermal from the chemical processes 
has broad implications. It appears that metamorphic reactions may not 
be progressive, and only those at the highest grades in the thermal gra- 
dient will be observed. Furthermore, the melting of the coarse-grained 
rocks of the lower crust and mantle may be highly rate-dependent, with 
widespread melting occurring simultaneously and through a large volume. 
The quantitative experiments, carried out in the temperature range of 
750°-1150°C, are also of interest to ceramists and mechanical engineers 
concerned with heat transfer in industrial processes. 

Fluid-Determined Reactions 

In the view of Mark Barton at the Geophysical Laboratory, almost all 
geological processes involve fluids. Fluids composed predominantly of car- 
bon, oxygen, hydrogen, and sulfur ("COHS fluids") are very important in 
metamorphism, the formation of ore deposits, the crystallization of vol- 
atile-rich magmas, and other earth processes. Theoretical calculations of 
the compositions of COHS fluids in equilibrium with minerals yield con- 
siderable insight into the nature of rock-forming processes. Most studies, 
however, approach the problem from the point of view of the rocks in- 
volved; that is, the composition of the fluid is calculated from the observed 
mineral association. An alternative way to consider the interaction be- 
tween minerals and a fluid is to assume a fixed composition for the fluid 
and then calculate which minerals should be stable. This second approach 
is more realistic than the first for studying the many geologic environments 
where the reactivity of the fluids exceeds that of the solids. More impor- 
tant, the second approach lends itself to accounting for mass balance, which 
is required for the quantitative interpretation of the formation of ore 
deposits and other rocks formed by mass transfer. 

Barton's calculations show that geologically reasonable fluids may con- 
trol the oxidation and possibly the sulfidation states in many rocks. This 
process may be particularly important in ore deposits where large quan- 
tities of fluid pass through small volumes of rock. In addition, the calcu- 
lations unexpectedly show that the presence of COHS fluids can determine 
which minerals are formed in some volatile-rich igneous rocks and in meta- 
morphic rocks. Heretofore, the minerals in a rock commonly were thought 
to have controlled the fluid compositions. The new calculations demon- 
strate that fluids can dominate in some rocks, a conclusion that opens the 
way for new insights into the petrogenesis of ores and other rocks. 


Molecular Fossils 

In the past two years at the Geophysical Laboratory, Thomas Hoering 
has been investigating the mechanisms by which insoluble organic matter, 
or kerogen, in marine sedimentary rocks, undergoes decomposition into 
smaller, lower-molecular-weight products. This process, called catage- 
nesis, is an important step in the geochemistry of organic matter and is 
responsible in part for the generation of the components of petroleum. 
Hoering used the novel technique of pyrolysis in the presence of excess 
liquid water to break down the kerogen of Recent sea-floor sediment. He 
found that in many ways the products resembled those in ancient rocks. 
The assumption is made that in these rate-limited reactions, slightly el- 
evated temperatures can proxy for long periods of geological time. Hence, 
chemical processes in sediments, whose rates in nature are too slow to be 
observed, become measurable at elevated temperatures in the laboratory. 
The ultimate test of this assumption is how well the products of the syn- 
thetic reactions match those that occur naturally. Many of these products 
have molecular structures that can be related back simply to their bio- 
logical and biochemical precursors. Furthermore, when the reaction is 
carried out in heavy water (deuterium oxide), the products exhibit ex- 
tensive substitution of heavy hydrogen (deuterium) for normal hydrogen. 
In addition, when pure organic substances are added to the sediment and 
heated, transfer of hydrogen from the kerogen to the added substrate 
occurs probably by means of free-radical chain reactions involving free 
hydrogen atoms. 

These experiments are significant from several standpoints. First, they 
demonstrate a simple method for mildly breaking down the refractory 
material kerogen into recognizable fragments that retain some facets of 
their biogeochemical history. Second, the experiments illustrate how the 
principles of modern physical-organic chemistry and isotopic tracer tech- 
niques can be applied to the study of exceedingly complex chemical re- 
actions that take place in natural organic catagenesis. 

The New Plant Biology 

Darwin and his successors showed us that man shares his ancestry 
with all other living things, and physiology and biochemistry have 
shown us that the processes of life are built up from the same 
chemical reactions that take place between non-living sub- 
stances — those processes are indeed marvellous, but they are no 
longer miraculous. A deeper lesson, learnt more readily by the 


working scientist than by the layman, is the provisional and in- 
complete nature of our knowledge at any moment of time. 

Sir Andrew Huxley 

Supplement to Royal Society News 

Volume 2, Number 6, November 1983 

One of the things that has most pleased me since I became president 
of the Institution in 1978 has been the emergence of the plant sciences as 
a focal field of research — in the Institution as well as in the larger scientific 
community. Not only have disciplinary lines become blurred in our De- 
partments of Embryology and Plant Biology, but also, in the midst of 
what may be the greatest biological revolution of all time, so have the 
lines between undifferentiated and applied research. 

In the Department of Embryology, Nina Fedoroff is studying the mo- 
lecular basis of transposable genetic elements in maize, genetic elements 
discovered decades earlier by Barbara McClintock. Fedoroff and her col- 
leagues have cloned both the Activator (Ac) and Dissociator (Ds) elements 
by making use of their integration into and excision from two maize loci — 
Shrunken and Waxy. Transposable elements in maize provide a special 
interest for developmental biologists, since alleles can have very different 
effects on gene expression. We can now expect explanations for these 
phenomena at the molecular level. In addition, transposable elements can 
be used as markers for isolating genes into which they have been inserted. 

Like Fedoroff, Sondra Lazarowitz, also in the Department of Embryol- 
ogy, is making the very best use of the special independence and freedom 
that our departments offer their scientists. Trained in animal virology, 
she decided to change her field of research to plant viruses. As pathogens, 
plant viruses are of enormous societal interest, but they receive compar- 
atively little attention from scientists. Lazarowitz has elected to study a 
class of plant viruses called geminiviruses — the only known single-stranded 
DNA viruses that infect eukaryotes. In just a few months she isolated 
the DNA from one of these viruses and now is in the process of charac- 
terizing it. 

Two years ago, William Thompson and his associates in the Department 
of Plant Biology reported that the DNA of genes undergoing transcription 
is more susceptible to nuclease digestion than is DNA that is not being 
transcribed. This year, Thompson and Richard Flavell obtained evidence 
from wheat suggesting that active groups of ribosomal genes are less 
methylated than are nonactive groups. Although it is still too early to 
conclude that this correlation holds at the level of the individual gene, the 
experiments provide a promising lead to our understanding of gene reg- 

Another finding at the Department of Plant Biology deserving of special 





H H H H H H 

Application of the techniques of the molecular biologist to prob- 
lems in more-traditional plant genetics is increasingly evident in the 
work of the Department of Plant Biology. Joseph Berry and col- 
leagues, interested in whether C0 2 uptake of certain plants could 
be improved by manipulation of genes coding for a critical enzyme, 
crossed two species of a much-studied plant, Atriplex. The gel elec- 
trophoresis (above) is of chloroplast DNA from Atriplex rosea (a C 4 
plant, capable of high C0 2 uptake; labeled R) and Atriplex hastata 
(labeled H). Hybrid offspring are labeled R x H. Note that in the 
five lanes at left, the hydrid DNA is similar to that of the A. rosea 
(female) parent in lacking bands at 7 and 12 (kilobases) and in having 
a band at 14. In the right-hand lanes (digested with a different 
enzyme), the offspring have the rosea band at 8.5 and lack the 
hastata band at 15. The chloroplast DNA of the hybrids is thus 
shown to have been derived from the C 4 (maternal) parent. With 
other results linking chloroplast DNA to the critical enzyme, the 
study confirmed the feasibility of genetic modification of critical- 
enzyme properties and identified the chloroplast genome as the tar- 
get for such modifications. 


mention is the observation by David Stern and Jeffrey Palmer that sig- 
nificant regions of plant mitochondrial DNA are homologous to chloroplast 
DNA. In corn, for example, the gene sequence for the large subunit of 
RuBP carboxylase (chloroplast DNA-coded) is represented in mitochon- 
drial DNA sequences. The authors speculate on the possibility that DNA 
has been exchanged between these two organelles during their evolution- 
ary histories. In another study, Palmer and Shields obtained evidence that 
the mitochondrial DNA of the genus Brassica occurs in three circular 
DNA molecules, and that the two smaller circles are simply the conse- 
quence of intramolecular recombination within the larger circle. 

Neil Polans and Thompson report steady progress in a study of the 
variation in nucleotide sequences in several phytochrome-controlled genes 
among different pea accessions. The wealth of phenotypic variation makes 
this system a particularly favorable tool for studying the multilocus phy- 
tochrome system. John Watson, Palmer, and Thompson also report for 
the first time that higher-plant chloroplasts contain a gene for the elon- 
gation factor tufA required for chloroplast protein synthesis. 

Arthur Grossman, Lawrence Talbott, and Thomas Egelhoff have con- 
ducted detailed investigations to determine the location of the DNA coding 
for the different proteins that form pigmented granules called phycobili- 
somes. Phycobilisomes contain the light-harvesting accessory pigments in 
red and blue-green algae. In the eukaryotic algae, where the phycobili- 
somes are contained within chloroplasts (or, in Cyanophora, within a 
specialized chloroplast-like organelle called the cyanelle), the pigment- 
binding proteins are synthesized within the organelle itself, as is the pro- 
tein that anchors the phycobilisome to the photosynthetic membrane. The 
relationships among the pigmented polypeptides suggest that a gene fam- 
ily encoding both the pigment-binding and anchor proteins is located in 
the chloroplast. On the other hand, the so-called linker proteins, which 
bind the pigmented proteins to the anchor protein, are clearly synthesized 
in the cytoplasm. Thus phycobilisomes, like RuBP carboxylase, represent 
a division of labor between different genomes in the construction of a 
single functional unit. 

Within the photomorphogenesis group, fellows and graduate students 
have been very active. Using a variety of inhibitors, auxin analogs, 
antiauxins, and ionophores, Terri Lomax has studied the uptake of the 
plant auxin indole-3-acetic acid by pH-tight vesicles from zucchini squash 
hypocotyls. She has shown that the uptake does not represent simple 
diffusion of the neutral (protonated) form of the acid through the vesicle 
membranes, but is, rather, highly specific for this active growth hormone. 
Using electron paramagnetic spin resonance techniques in collaboration 
with Mehlhorn at Berkeley, she has been able both to characterize the 
pH gradient from outside to inside of these vesicles and to calculate the 
accumulation ratio (an impressive 100-fold) for the hormone. 


In a continuing study of auxin-red light interactions, James Shinkle has 
shown that physiologically effective concentrations of auxin can sensitize 
sections from etiolated oat coleoptiles to red light by four orders of mag- 
nitude. The auxin treatment converts the elongation reaction to red light 
from a low fluence (LF) to a very low fluence (VLF) response. Somewhat 
lower auxin concentrations yield a red light fluence-response curve show- 
ing two distinct steps. The response is very similar to those obtained 
earlier by Dina Mandoli with intact plants. It will be interesting to de- 
termine whether the hormone treatment has any effect on light-induced 
changes in mRNA transcript abundance. 

In several elegant studies on the role of light in regulation of devel- 
opment, Lon Kaufman has successfully merged the photomorphogenic 
interests of the Briggs laboratory with the molecular approaches of the 
Thompson laboratory. Mandoli had shown previously that completely- 
dark-grown plants had at least two growth responses to red light, both 
probably mediated by the pigment phytochrome. One of these was po- 
tentiated by extremely small light fluences (VLF response), while the 
other required four or more orders of magnitude more light (the LF 
response). This year, in studies of the light regulation of the abundance 
of specific mRNAs, Kaufman has resolved these responses . In peas grown 
in total darkness, he finds that the abundance of messenger RNA used to 
synthesize one of the chlorophyll a/6-binding proteins (an important struc- 
tural component of the light-gathering machinery of photosynthesis) is 
increased by light fluences in the VLF range and is then further increased 
by light treatments in the LF range. By contrast, the messenger abun- 
dance for the small subunit of RuBP carboxylase, the C0 2 -fixing enzyme, 
is unaffected by VLF treatments and increases only in response to light 
doses in the LF range. (Both of these mRNAs are coded by nuclear DNA.) 

Plants Under Stress. The Department of Plant Biology continues to 
probe the influence of various kinds of stress on higher plants. Using 
simple techniques, Jacob Levitt has shown that damage occurring during 
moderate wilting of cabbage leaves can be repaired during water reab- 
sorption even if the process takes several days. Thus, conditions during 
the rehydration period are perhaps as important as the extent of dehy- 
dration in determining the recovery of a wilted plant. C. Eduardo Vallejos 
and Olle Bjorkman report that a single night of low-temperature treatment 
(5°C) significantly reduces the subsequent photosynthetic performance of 
chilling-sensitive tomato plants, even without damage to photosystem II. 
Thus, photoinhibition is not the only way whereby low but not freezing 
temperatures can damage chilling-sensitive plants. 

In yet another study, Mervyn Ludlow and Bjorkman have investigated 
the influence of paraheliotropic leaf movement (turning edgewise to the 


light) in preventing drought damage to the photosynthetic reactions of a 
Mexican legume, siratro (Macroptilium atropurpureum). When this plant 
is water stressed, its leaves move from a position of maximum to one of 
minimum interception of incident light. If the leaves are physically con- 
strained, they show clear evidence of photoinhibition, particularly on the 
upper leaf surfaces. Fluorescence studies show that these leaf movements 
protect the leaves from photoinhibition, from high-temperature damage, 
and from interactive effects of high temperature and light. Siratro clearly 
provides an excellent model system for study of important protective 
responses found in many other species, including some (e.g., soybean) of 
economic importance. 

Susanne von Caemmerer, John Coleman, and Joseph Berry used the 
unicellular green flagellate Chlamydomonas reinhardtii to probe the role 
of ribulose bisphosphate in regulating the rate of photosynthesis. If Chla- 
mydomonas is grown under low-C0 2 conditions, it develops the capacity 
to accumulate carbon (from bicarbonate) to provide its ribulose bisphos- 
phate carboxylase with sufficient substrate for effective photosynthesis. 
Quantitative comparison of photosynthetic behavior clearly supports the 
hypothesis that C0 2 concentrations inside low-C0 2 -grown cells are higher 
(at the same external C0 2 concentration) than they are in high-C0 2 -grown 
cells. These studies also show that ribulose bisphosphate itself plays an 
important role in the regulation of the carboxylation reaction. There was 
excellent agreement between responses measured in vivo and those pre- 
dicted by in vitro studies of the carboxylase and from theoretical models 
published earlier by Graham Farquhar, Berry, and von Caemmerer. 

In a reenactment of an historic experiment (see Year Books 68, 69, and 
70), Malcolm Nobs successfully repeated a cross between a C 3 and a C 4 
species of Atriplex. Because the ribulose bisphosphate carboxylases from 
these two species have clearly distinguishable kinetic properties, Berry, 
with Nobs, Bernardita Osorio, Palmer, James Tepperman, and Thompson 
addressed the question as to which subunit of this enzyme — the small one 
(coded for by the nucleus) or the large one (coded for by the chloroplast) — 
determines these properties. The hybrid yielded a carboxylase with the 
kinetic properties of the C 4 (maternal) parent. Restriction analysis of the 
chloroplast DNAs from both parents and three hybrids indicated a ma- 
ternal inheritance for the chloroplasts, whereas the nuclear DNAs were 
clearly biparentally inherited. Since it is well known that chloroplast DNA 
codes for the large subunit of the carboxylase, it follows that the large 
subunit determines the kinetic properties of these enzymes. The study 
did not define a function for the small subunit, but it is still conceivable 
that some other genetically based differences in kinetics might be related 
to it. All the same, future efforts at kinetic modification of the carboxylase 
to improve photosynthetic yield should focus on the chloroplast genome, 
rather than on the far-more-complicated nuclear genome. 


Developmental Genetics in Animals 

It was just a year ago that I wrote of Allan Spradling and Gerald Rubin's 
new and powerful use of transposable genetic elements as a vector for 
introducing genes into Drosophila. In the first application of this method, 
they demonstrate that a specialized gene, termed rosy, can be transferred 
into many different chromosomal loci without affecting its developmental 
regulation. This is an important preliminary result if the gene transfer 
method is to be used to define functional regions in genes. In another set 
of experiments, the DNA region that accounts for amplification of chorion 
genes is being delimited systematically by transforming defined DNA 
segments into Drosophila. 

Donald Brown's research on the 5S RNA genes evolves toward under- 
standing the molecular details of what he calls the transcription complex — 
the structure that is formed between a gene and a minimum of three 
factors (presumably proteins) required to direct accurate transcription 
initiation by RNA polymerase III. One of these factors has been identified, 
and its interaction with the gene is being elucidated in detail. He believes 
that the transcription complex that can be formed in vitro closely resem- 
bles the state of an active 5S RNA gene in a living cell. Thus Brown and 
his colleagues have turned their attention to the biochemistry of chro- 

Daniel Burke and Samuel Ward have analyzed genes for the major 
sperm-specific proteins in the nematode Caenorhabditis elegans. These 
proteins are encoded by a complex family of related but not identical genes. 
The proteins are major constituents of the pseudopod that moves the 
sperm toward the egg, and they are likely candidates to account for the 
ameboid movement, which is required for fertility. 

The Biology of the Cell Surface 

The ectoplasm, standing between the protoplasmic system's inner 
substance and the outside world, reacts first to environmental 
stimuli and thus conditions the responses of the whole system. 
Its rapidly occurring, highly active structural changes portray 
self-regulation and self-differentiation. Thus by its location and 
by its peculiar attributes, the ectoplasm becomes the most tan- 
gible expression of life-processes. 

E. E. Just, 1939 

The Biology of the Cell Surface 

The Department of Embryology has attached special importance to the 
study of the biology of the cell surface. As Donald Brown observes, until 
recently lipids, which abound in cell surfaces, have been the province of 


biochemists, who studied the chemistry of lipids and the enzymatic re- 
actions that synthesized and degraded them. New methodology is ushering 
in a new era of lipid research which might be called lipid cell biology, 
where the location and function of lipids can be investigated in living cells 
and tissues. Richard Pagano and his colleagues have learned how to tag 
lipids with fluorescent labels that do not interfere with their normal me- 
tabolism. The use of this methodology is demonstrated by a beautiful set 
of studies designed to trace the unique pathways of different lipids through 
the cell cytoplasm after introduction at the cell surface. For his role in 
the discovery of such techniques and related research, Pagano was hon- 
ored this year as recipient of a Joseph Meyerhoff fellowship to study at 
the Weizmann Institute in Rehovat, Israel. 

The departmental interest in cell surfaces has been strengthened by 
the arrival of Martin Snider, who studies the cellular location of inter- 
mediates in glycoprotein synthesis. Complex sugars are added to a poly- 
peptide in precise order within cells, and specific intermediates move from 
the cytoplasmic side of a membrane to its luminal surface, where biosyn- 
thesis is completed. 

In parallel studies, Richard Rotundo describes his recent observations 
on the metabolism of the important enzyme acetylcholinesterase. This 
enzyme has a surprisingly complex metabolism, generating at least two 
different populations of molecules metabolized by the cell in very different 

In the past several years, Douglas Fambrough's research has encom- 
passed an increasingly diverse set of proteins, all of which in some way 
influence how cells sense their chemical environments. One of these surface 
proteins, the (Na,K)-ATPase, controls the flow of ions into cells. A mono- 
clonal antibody has been isolated that reacts specifically with this protein; 
Fambrough and his colleagues are thus able to correlate cytologically the 
protein's location with that of a class of intermembranous particles. The 
central importance of (Na,K)-ATPase to conductivity of nerves makes the 
elucidation of its detailed chemistry of great interest. 


With regret I report the deaths this year of trustees emeriti Charles 
P. Taft and Carl J. Gilbert. Mr. Taft served on the Board from 1936 until 
1975. A devoted friend of the Institution, he took an active role in Carnegie 
affairs even after he officially stepped down from the Board. He spent 
most of his life in his native city, Cincinnati, where he was an original 
member of the City Charter Committee, mayor (1955-1957), city coun- 
cilman, lawyer, and leader in the religious community. He died on June 
24 at the age of 85. Carl Gilbert, a Boston civic leader, died on November 
13, 1983, at the age of 77. He was the retired president and chairman of 


the Gillette Corporation and a former Special Representative for Trade 
Negotiations under President Nixon. He served on Carnegie's Board from 
1962 until May 1983, at which time he was designated trustee emeritus. 
He had been a member of the Executive Committee and had served as 
chairman of the Nominating Committee, the Retirement Committee, and 
the Employee Benefits Committee. 

Several former Staff Members and associates at the Department of 
Embryology passed away during the past year. Staff Member Samuel 
Reynolds (1941-1955), an internationally renowned student of uterine 
physiology, died in September 1982. Nearly a year later, in August 1983, 
the Department lost another pioneer — experimental morphologist Mary 
Rawles Spurbeck. Rawles served as Staff Member from 1957 until her 
retirement in 1966. And finally, with the death in August 1983 of modeler 
Osborne 0. Heard, who served from 1914 until 1956, the field of human 
embryology lost yet another dedicated friend. 

Ralph E. Gibson, former Staff Member of the Geophysical Laboratory 
(1924-1946) and former director of the Applied Physics Laboratory of 
Johns Hopkins University (1948-1969), died in February 1983 at the age 
of 81. 0. Frank Tuttle, another former Staff Member at the Laboratory 
(1947-1953), died in December 1983. 

Three archaeologists of Carnegie's former Division of Historical Re- 
search (later named Department of Archaeology) died this year. Gustav 
U. K. Stromsvik (1926-1957) died in January 1983; Robert E. Smith (1931- 
1960) died in February 1982; and Robert S. Chamberlain (1936-1946) 
passed away during the summer of 1982. Agnes Fisher, former editor and 
secretary at the Department of Genetics (1942-1974), died in January 

Edward Orozco, former accountant at the Geophysical Laboratory (1928- 
1966) died in October 1983. 

After serving as a trustee for eighteen years, William McChesney Mar- 
tin, Jr. , stepped down from the Board this year. He was appointed trustee 
emeritus. Mr. Martin most recently served Carnegie as a member of the 
Executive Committee, the Finance Committee, and the Visiting Com- 
mittee to the Department of Plant Biology. From 1973 to 1976, he was 
chairman of the Board. Robert 0. Anderson, a trustee since 1976, resigned 
from the Board in May. 

Retiring from active service at Plant Biology this year were Malcolm 
Nobs, Ruth Fischer, and Richard Hart. Nobs came to the Department in 
1939 as a technical assistant to work with the experimental taxonomy 
group. In 1958 he was appointed Staff Member. Ruth Fischer served the 
Department as secretary-accountant for thirteen years, and Dick Hart 
worked as a machinist for 27. 

Master craftsman Fred O'Neil retired from the Observatories after 
sixteen years of service. 


Christopher Wright departed in June 1983 after spending four years at 
the Institution as Staff Member (Science Policy and Institutional Devel- 
opment). At the Department of Embryology, Staff Members Gerald Rubin 
and Kenneth Muller both left this year, Rubin after three years to become 
MacArthur Professor of Biochemistry at the University of California, 
Berkeley, and Muller after eight, to join the University of Miami School 
of Medicine. And three retired but still active Staff Members departed 
from their respective departments. Cosmic ray expert Scott Forbush, a 
Staff Member at DTM from 1927 to 1969, and since then a Staff Member 
Emeritus, left DTM for Charlottesville, Virginia. Emeritus isotope geo- 
chemist Gordon Davis (Staff Member from 1941 to 1978) and Emeritus 
Distinguished Professor Elburt Osborn (Staff Member from 1938 to 1945 
and then from 1973 to 1977) departed from the Geophysical Laboratory. 


We are very fortunate in the recent addition to Carnegie's leadership 
of Margaret L. A. Mac Vicar. Concurrently professor of physical science 
and Cecil and Ida Green Professor of Education at the Massachusetts 
Institute of Technology, Mac Vicar assumed her duties as vice president 
of the Carnegie Institution in July 1983. Mac Vicar has been a faculty 
member at MIT since 1970. She founded and directed MIT's highly suc- 
cessful Undergraduate Research Opportunities Program, and has been 
affiliated with the Institute's Division for Study and Research in Education 
and its Department of Physics. Her awards include the 1982 Cecil and 
Ida Green Distinguished Lectureship; the 1979 Chancellor's Distinguished 
Professorship, Berkeley; the 1978 Vollmer W. Fries Lectureship, Rens- 
selaer Polytechnic Institute; and the 1976 Young Faculty Award, General 
Electric Foundation. 

We also welcome new Staff Member Joseph G. Gall, who joined the 
Department of Embryology in September 1983. Gall comes to Carnegie 
from Yale University, where he was Ross Granville Harrison Professor 
of Biology. He plans to continue his investigations into how the genes and 
chromosomes of higher organisms change during embryonic development, 
an area in which he has made several important discoveries. 

In December 1983, William F. Kunkel joined the Observatories as res- 
ident scientist/administrator of Las Campanas. Kunkel, who received his 
Ph.D. in astronomy from the University of Texas in 1967 and who is 
interested in galaxy environments, comes to the Observatories most re- 
cently from the Centro Astronomico Hispano Aleman, which is a Spanish 
field station of the Max Planck Institut fur Astronomie in Heidelberg. At 
Carnegie, he is responsible for the management of Las Campanas Obser- 
vatory facilities and personnel. 

In January 1983, Arturo E. Urquieta joined the Observatories as tech- 


nical assistant to the director, responsible for the technical performance 
of the telescopes and instruments, especially those at Las Campanas. 


Distinguished Service Member Barbara McClintock's Nobel Prize in 
Physiology or Medicine, bestowed in Stockholm on December 10, 1983, 
caps a very special year for Carnegie. McClintock's Prize — which honors 
her discovery of movable genetic elements nearly forty years ago — is a 
well-deserved award for a remarkable scientist, but it is also very signif- 
icant to the Carnegie Institution, who recognized McClintock's extraor- 
dinary talents early in her career, and who offered her, in 1942, a position 
as Staff Member (see pp. 3-h). 

The lifetime achievement of another Carnegie scientist was honored this 
year in the naming of the nation's planned orbiting optical observatory. 
The Edwin P. Hubble Space Telescope, scheduled for launch aboard the 
Space Shuttle in 1986, will be able to see objects many times fainter than 
can now be seen by any ground-based observatory. Edwin Hubble was a 

Edwin P. Hubble, 1889-1953 


Staff Member at the Mount Wilson Observatory from 1919 until his death 
in 1953. With colleague Milton Humason, his observations with the 100- 
inch Hooker telescope confirmed that the spiral nebulae were distant gal- 
axies beyond our Milky Way and that these nebulae were receding from 
us at velocities proportional to their distances. In short, he discovered 
that the universe was expanding. 

The year has also been rich in awards to our other scientists. Winslow 
Briggs, director of the Department of Plant Biology, received a Senior 
United States Scientists Award (Humboldt Award) from the German- 
based Alexander von Humboldt Foundation. This distinguished award 
offers him the possibility of an extended research stay in the Federal 
Republic of Germany. Briggs was also honored this year by the Botanical 
Society of America, who awarded him a Certificate of Merit for his con- 
tributions to the advancement of botanical sciences as researcher, pro- 
fessor, and director. 

Joseph Gall, new Staff Member at the Department of Embryology, 
received the American Society of Cell Biology's most distinguished award, 
the E. B. Wilson Medal, for 1983. Gall was cited for his "exciting studies 
on the structure and function of the genome." In addition, Gall was honored 
by the American Cancer Society, which named him an American Cancer 
Society Research Professor of Developmental Genetics. 

For their development of a gene transfer technique in Drosophila, De- 
partment of Embryology Staff Members Allan Spradling and Gerald Rubin 
were the junior recipients of the 1983 Passano Award of the Passano 
Foundation. Spradling was later named Distinguished Young Scientist of 
the Year by the Maryland Academy of Sciences. 

Observatories Staff Member Alan Dressier was selected to receive the 
1983 Newton Lacy Pierce Prize from the American Astronomical Society. 
The award, which recognizes outstanding achievement in observational 
astronomy over the past five years, is given to an astronomer who has 
not yet attained age 36. 

Geophysical Laboratory Staff Member Ikuo Kushiro was elected to the 
National Academy of Sciences as a foreign associate in April 1983. 

In May, Plant Biology Staff Member Olle Bjorkman was elected to the 
American Academy of Arts and Sciences. 

In October, former Plant Biology Staff Members William Hiesey and 
David Keck received Fellow's Medals from the California Academy of 

Department of Embryology Staff Member Richard Pagano was awarded 
a Meyerhoff Fellowship for summer study at the Weizmann Institute of 
Science in Israel (see p. 85). 

Peter Bell, Staff Member at the Geophysical Laboratory, received the 
Sherman Fairchild Distinguished Scholar Award, presented by the pres- 
ident of the California Institute of Technology. 


Hatten S. Yoder, Jr., director of the Geophysical Laboratory, was elected 
a member of the Council of the American Philosophical Society for a three- 
year term in April 1983. 

Robert Howard, Observatories Staff Member, received a certificate of 
recognition from NASA in April 1983 for technical innovation in instru- 
mentation for solar oscillation measurements. 

Paul Silver, a new DTM Staff Member, was awarded the 1983 Eckart 
Prize from the Scripps Institution of Oceanography for the outstanding 
thesis of 1982. 

Observatories postdoctoral fellow Nicholas Suntzeff received the Robert 
J. Trumpler Award for 1983 from the Astronomical Society of the Pacific. 
The award is given each year to a recent recipient of the Ph.D. whose 
dissertation is considered of unusual importance to astronomy. 

Stanley Hart, a Staff Member at DTM from 1961 until 1975 and now a 
professor at MIT, was elected to membership in the National Academy 
of Sciences in April. Thomas Donahue, a visiting investigator at DTM this 
year, was also elected to the Academy. 

John Gurdon, charter member of the Visiting Committee and former 
fellow at the Department of Embryology, was elected this year to the 
American Philosophical Society as a foreign member. 

Several trustees received special honors this year. Charles H. Townes 
was awarded a National Medal of Science at a White House ceremony in 
May. Earlier, he was nominated to become a member of the Pontifical 
Academy of Sciences by Pope John Paul II. 

The government of Japan conferred the award of the First Class of the 
Order of the Sacred Treasure on Franklin D. Murphy for his distinguished 
service in strengthening the bonds of friendship between Japan and the 
United States. He was also awarded the Officer's Cross of the Order of 
Merit of the Federal Republic of Germany for his furtherance of German 
science, literature, language, and culture. 

William R. Hewlett received an honorary doctorate of engineering at 
Dartmouth College in June and an honorary doctorate of laws at Mills 
College in Oakland, California, in September. Lewis M. Branscomb re- 
ceived honorary doctor of science degrees from the University of Alabama 
in May and from Pratt Institute in June. 

Edward E. David was awarded the 1982 New Jersey Science/Technol- 
ogy Medal on December 8. 

James D. Ebert 




Stanford, California 


Winslow R. Briggs 

Director Emeritus 
C. Stacy French 

William M. Hiesey 
Malcolm A. Nobs 

Staff Members 
Joseph A. Berry 
Olle Bjorkman 
Jeanette S. Brown 
David C. Fork 
Arthur R. Grossman 
William F. Thompson 

Research Associates 

Lon S. Kaufman 
Jeffrey D. Palmer 

Visiting Investigators 

Murray Badger 
Jacob Levitt 

Senior Fellows 

Jerry Brand 
Satoshi Hoshina 
Mervyn M. Ludlow 
Prasanna Mohanty 
Kazuhiko Satoh 

John R. Coleman 
Graham D. Farquhar 
Moritoshi lino 
Neil 0. Polans 
Eberhard Schafer 
Susan C. Spiller 
C. Eduardo Vallejos 

Susanne von Caemmerer 
John C. Watson 


J. Timothy Ball 
Tobias I. Baskin 
Helen E. Edwards 
Thomas Egelhoff 
David Hollinger 
Alan P. Maloney 
Dina F. Mandoli 
Elizabeth Newell 
Terri L. Lomax 
James R. Shinkle 
David B. Stern 
Mary Todd 
Philippe Tacchini 
Lawrence D. Talbott 

Baltimore, Maryland 


Donald D. Brown 

Staff Members 

Douglas M. Fambrough 
Nina V. Fedoroff 
Kenneth J. Muller 
Richard E. Pagano 
Gerald M. Rubin 
Allan C. Spradling 
Samuel Ward 

Staff Associates 

Sondra G. Lazarowitz 
Richard L. Rotundo 
Martin Snider 

Research Associates and Fellows 
M. John Anderson 
Karen Bennett 
Matthias Chiquet 
Mary Collins 



Diane de Cicco 
Ellen Elliott 
Kathy French 
Tulle Hazelrigg 
Ian Jackson 
Laura Kalfayan 
Roger Karess 
Robert Levis 
Naomi Lipsky 
Adrian Mason 
Wy lie Nichols 
Kevin O'Hare 
Heli Roiha 
David Setzer 
Mavis Shure 
Richard Sleight 
Douglas Smith 
William Taylor 
Barbara Wakimoto 
Susan Wessler 
Barry Wolitzky 
Michael Wormington 


Paula Adams 
Daniel Burke 
Barbara Kirschner 
Darlene Marshall 
Vincent Morgese 
Suki Parks 
Mark Schlissel 
Jennifer Schwartz 

Visiting Investigators and 
Extramural Collaborators 

Richard Hallberg 

Herbert Koenig 

Jeannine Koenig 

Martin Schwartz 

Michael Simmons 

Washington, D.C. 


Hatten S. Yoder, Jr. 


Gordon L. Davis 
Elburt F. Osborn 

Staff Members 
Peter M. Bell 
Francis R. Boyd, Jr. 
Felix Chayes 
Marilyn L. F. Estep 
Larry W. Finger 
John D. Frantz 
P. Edgar Hare 
Robert M. Hazen 
Thomas C. Hoering 
T. Neil Irvine 
Ikuo Kushiro 
Ho-kwang Mao 
Bj0rn My sen 
Douglas Rumble III 
David Virgo 

Staff Member (Temporary) 
Kenneth A. Goettel 

Postdoctoral Associates 

Ji-an Xu 
Guangtian Zou 


Mark D. Barton 
Paul A. Danckwerth 
Michael H. Engel 
Andrew P. Gize 
David London 
Stephen A. Macko 
Daniel J. Schulze 

Predoctoral Fellows 

Edward F. Duke 
Andrew P. Jephcoat 
Joseph W. E. Mariathasan 
Charles M. Schlinger 



Washington, D.C. 


George W. Wetherill 

Staff Members 

Thomas Aldrich 
Alan Paul Boss 
Louis Brown 



Richard W. Carlson 
W. Kent Ford, Jr. 
David E. James 
Typhoon Lee 
Alan T. Linde 
Vera C. Rubin 
I. Selwyn Sacks 
Frangois Schweizer 
Paul Silver 
Fouad Tera 
Norbert Thonnard 

Visiting Investigators 

Thomas M. Donahue 
Kyoichi Ishizaka 
Jack J. Lissauer 
Stephen Morris 
Douglas G. Mose 
Milan J. Pavich 
David Schwartzman 
Nathalie Valette-Silver 
J. Arthur Snoke 
Ragnar Stefansson 
Richard T. Williams 

Research Associates 

Charles Angevine 
Hiroshi Mizuno 

Senior Fellows 

Stanley A. Mertzman 
Arvind S. Tamhane 


Barbara Barreiro 
William K. Hart 
Mizuho Ishida 
David Koo 
Tetsu Masuda 
Linda Y. Schweizer 
Bradley C. Whitmore 

Predoctoral Fellows and Students 

Michael So Chyi 
Diana Diez de Medina 
James K. Meen 
Julie Morris 


Pasadena, California 


George W. Preston 

Director Emeritus 
Horace W. Babcock 


Olin C. Wilson 

Assistant Director for 
Mount Wilson 

Robert F. Howard 

Resident Scientist/ Administrator 
for Las Campanas 

Wojciech A. Krzeminski 

Staff Members 

Halton C. Arp 
Alan M. Dressier 
Jerome Kristian 
S. Eric Persson 
Allan R. Sandage 
Paul L. Schechter 
Leonard Searle 
Stephen A. Shectman 

Adjunct Staff Members 

W. Kent Ford 
Vera C. Rubin 
Frangois Schweizer 

Staff Associates 

Jean Lorre 
Arthur H. Vaughan 

Research Associates 

Todd A. Boroson 
Douglas K. Duncan 
Ian B. Thompson 




David H. Bruning 
Peter J. McGregor 
Geoffrey W. Marcy 
David G. Monet 
Herschel B. Snodgrass 
Nicholas B. Suntzeff 
Peter B. Stetson 

Carnegie-Chile Fellows 

Guido Garay 
Fernando J. Selman 

Visiting Fellows 

Bruno Binggeli 
George A. Carlson 

Visiting Associates 

Gerard Gilmore 
Rita E. M. Griffin 
Roger F. Griffin 
Joseph L. Snider 
Gustav A. Tammann 

Judith G. Cohen 

G. Edward Danielson 

Edward L. Dunbar Jr. (JPL) 

Richard Edelson 

Alexei V. Filippenko 

Richard Gomer 

Jules Halpern 

Gary Heiligman 

Keith Home 

David Jewitt 

Dayton Jones 

Peter N. Kupferman (JPL) 

Arthur L. Lane (JPL) 

Michael E. Morrill (JPL) 

Robert M. Nelson (JPL) 

Alain Porter 

Steven H. Pravdo (JPL) 

R. Michael Rich 

Abhijit Saha 

Wallace L. W. Sargent 

Michael Summers 

Richard J. Terrile (JPL) 

John T. Trauger 

David Tytler 

California Institute of Technology 

Schelte J. Bus (JPL) 

JPL: Jet Propulsion Laboratory, Pasadena. 

Reports of Departments and 
Special Studies 

Department of Plant Biology 

Department of Embryology 

Geophysical Laboratory 

Department of Terrestrial Magnetism 

Mount Wilson and Las Campanas Observatories 

Developmental Biology Research Group 

Department of Plant Biology 

Stanford, California 

Winslow R. Briggs 



Introduction (Briggs) 

Cytosine methylation and the activity 
of ribosomal RNA genes in wheat 
(Flavell and Thompson) 


Phytochrome control of specific mRNA 
levels in developing pea buds: The 
presence of low and very low fluence 
responses (Kaufman, Thompson, and 

Widespread presence of chloroplast 
DNA sequences in plant mitochon- 
drial genomes (Stern and Palmer) . . . 

Tripartite organization of the Brassica 
mitochondrial genome (Palmer and 

Variation in pea restriction endonu- 
clease fragment patterns (Polans and 

Chloroplast genes for components of 
the translational apparatus (Watson, 
Palmer, and Thompson) 

A general model for chloroplast gen- 
ome evolution (Palmer and Thomp- 

Auxin uptake by sealed plant plasma 
membrane vesicles is a specific, sat- 
urable transport (Lomax and Briggs) 

Quantitation of sealed vesicle volume, 
pH gradient, and auxin uptake of 
zucchini hypocotyl membrane prepa- 
rations (Lomax, Mehlhorn, and 

Auxin increases coleoptile section sen- 
sitivity to red light (Shinkle and 





Fiber-optic capacity of plant tissues 
varies as a function of tissue water 
status (Mandoli, Boyer, and Briggs) 


Growth distribution during first posi- 
tive phototropic curvature of maize 
coleoptiles (lino and Briggs) 42 

Phytochrome-mediated phototropism in 
maize seedling shoots (lino, Briggs, 
and Schafer) 45 

Spectroscopy of chlorophyll in photo- 
synthetic membranes (Brown) 46 

Is the red absorption band of chloro- 
phyll a skewed? (Brown) 53 

Adaptation by the red alga Porphyra 
perforata to changes in the intensity 
and quality of light — a new mecha- 
nism (Fork and Satoh) 55 

22 Characteristics of the state II-state III 
transitions in the red alga Porphyra 
perforata (Satoh and Fork) 58 

24 A reaction that may protect the red 

alga Porphyra perforata against pho- 
toinhibition (Satoh and Fork) 61 

The room-temperature, long-wave- 
2^ length emission band of Porphyra 
perforata and its association with 
photosystem II (Mohanty 
and Fork) 65 

30 Interruption of photosynthetic pro- 
cesses of the red alga Porphyra per- 
forata by salinity (Smith, Satoh, 
and Fork) 


Reversible inhibition of photosystem II 
photochemistry in Anacystis nidu- 
layis by removal of Ca 2+ (Brand, Mo- 
hanty, and Fork) 72 

Calcium depletion affects energy trans- 
fer and alters the fluorescence yield 

and emission of photosystem II in 
Anacystis nidulans (Mohanty, 
Fork, and Brand) 


Preparation of chlorophyll-protein com- 
plexes from the cyanobacterium An- 
acystis nidulans by SDS-sucrose 
density gradient centrifugation 
(Hoshina and Fork) 


Analysis of absorption spectra of chlo- 
rophyll a in phosphatidylcholine lipo- 
somes at different temperatures 


Injury and repair during rehydration of 
droughted cabbage leaves (Levitt) . . 87 

Partial inhibition of photosynthesis by 
low night temperatures (Vallejos and 
Bjorkman) 88 

Paraheliotropic leaf movement as a 
protective mechanism against 
drought-induced damage to primary 
photosynthetic reactions (Ludlow and 
Bjorkman) 89 

Control of photosynthesis by RuP 2 con- 
centration: studies with high- and 
low-C0 2 -adapted cells of Chlamydo- 
monas reinhardtii (von Caemmerer, 
Coleman, and Berry) 91 

Genetic control of the kinetic parame- 
ters of RuP 2 carboxylase: studies of 
a C 3 and a C 4 species and their Fi 

hybrid (Berry, Nobs, Osorio, Pal- 
mer, Tepperman, and Thompson) . . 

Location and identification of carbonic 
anhydrase in Chlamydomonas rein- 
hardtii (Coleman, Berry, Togasaki, 
and Grossman) 



Effect of inhibitors of protein synthesis 
and glycosylation on the induction of 
carbonic anhydrase (Coleman and 
Grossman) 105 

Regulation of protein synthesis during 
adaptation of Chlamydomonas rein- 
hardtii to low C0 2 (Coleman and 
Grossman) 109 

Biosynthesis of phycobilisome polypep- 
tides of Porphyridium aerugineum 
and Cyanophora paradoxa (Gross- 
man, Talbott, and Egelhoff) 112 

A rapid procedure for the isolation of 
intact phycobilisomes (Grossman and 
Brand) 116 

High-molecular-weight aggregates of 
phycobiliproteins on denaturing poly- 
acrylamide gels (Grossman) 120 

Bibliography 122 

Speeches 125 

Personnel 128 


The Department of Plant Biology has proaches of the Thompson laboratory, 

continued its good fortune in attracting Mandoli had shown previously that com- 

unusually gifted predoctoral and post- pletely-dark-grown plants had at least two 

doctoral students and senior investiga- growth responses to red light, both prob- 

tors. The strength of the current group ably mediated by the pigment phyto- 

of fellows and co-workers is measured chrome. One of these was potentiated by 

not just by the articles that follow, but extremely small light fluences (the very 

also by the Department's unusual record low fluence response, VLF), while the 

of outside publication. It merits note that other required four or more orders of 

even though two faculty members were magnitude more light (the low fluence 

on sabbatical leave a fair part of the year, response, LF). This year, in studies of 

the science continued unabated. the light regulation of the abundance of 

In past years it has been possible to specific mRNAs, Kaufman has resolved 
organize this Introduction into relatively these responses. In peas grown in total 
neat packages around the various sub- darkness, he finds that the abundance of 
disciplines in the Department — e.g., messenger RNA used to synthesize one 
photosynthesis, molecular biology, de- of the chlorophyll a/6-binding proteins (an 
velopment, physiological ecology. This important structural component of the 
year, however, the lines have become se- light-gathering machinery of photosyn- 
riously blurred, as interdisciplinary ac- thesis) is increased by light fluences in 
tivities have increased and molecular the VLF range, and then further in- 
techniques have been applied to prob- creased by light treatments in the LF 
lems in other areas. But while the dis- range. By contrast, the messenger abun- 
ciplinary edges have been blurred, the dance for the small subunit of RuBP car- 
caliber of research has remained high, boxylase, the C0 2 -fixing enzyme, is 
and the year's efforts have produced a unaffected by VLF treatments, and in- 
number of significant findings. The fol- creases only in response to light doses in 
lowing paragraphs represent an attempt the LF range. (Both of these mRNAs are 
to highlight them. coded by nuclear DNA.) 

Two years ago, Thompson and his as- This year, plant mitochondrial DNA is 

sociates reported that DNA of genes un- mentioned for the first time. Stern and 

dergoing transcription is more susceptible Palmer report the surprising finding that 

to nuclease digestion than is DNA that significant regions of plant mitochondrial 

is not being transcribed. This year, Flav- DNA are homologous to chloroplast DNA. 

ell and Thompson obtained evidence from In corn, for example, the gene sequence 

wheat suggesting that active groups of for the large subunit of RuBP carbox- 

ribosomal genes are less methylated than ylase (chloroplast DNA-coded) is rep- 

nonactive groups. Although it is still too resented in mitochondrial DNA 

early to conclude that this correlation sequences. The results raise the very real 

holds at the level of the individual gene, possibility that DNA has been ex- 

the experiments provide a promising lead changed between these two organelles 

to our understanding of gene regulation, during their evolutionary histories. In 

In several elegant studies on the role another study, Palmer and Shields ob- 

of light in the regulation of development, tained evidence that the mitochondrial 

Kaufman has successfully merged the DNA of the genus Brassica occurs in three 

photomorphogenic interests of the Briggs circular DNA molecules, and that the two 

laboratory with the molecular ap- smaller circles are simply the conse- 



quence of intramolecular recombination is very similar to those obtained earlier 

within the larger circle. by Mandoli with intact plants. It will be 

Polans and Thompson report steady interesting to determine whether the 

progress in a study of the variation in hormone treatment has any effect on light- 

nucleotide sequences in several phyto- induced changes in mRNA transcript 

chrome-controlled genes among different abundance. Here is another potential in- 

pea accessions. The wealth of phenotypic terface between the molecular and the 

variation makes this system a particu- developmental approaches, 

larly favorable tool for studying the mul- In her continuing studies of light pip- 

tilocus phytochrome system. Watson, ing in plants, Mandoli, in collaboration 

Palmer, and Thompson also report for with Boyer from the University of Illi- 

the first time that higher-plant chloro- nois, demonstrated that there is an ex- 

plasts contain a gene for the elongation ponential decline in light-piping capacity 

factor tufA required for chloroplast pro- with water loss in soybean hypocotyls, 

tein synthesis. Finally, Palmer and as well as an increase in the acceptance 

Thompson present a model for chloro- angle for maximum transmission of 

plast genome evolution based on their obliquely incident light. These optical 

extensive studies (many with various col- properties may provide important tools 

laborators) of the structure and varia- for the nondestructive assessment of plant 

bility of chloroplast DN A in various plant water status. 

taxa. Since 1934, it has been assumed that 
Graduate students and fellows have phototropic curvatures of higher plants 
been very active in photomorphogenesis are mediated exclusively by blue light 
studies. Using a variety of inhibitors, photoreceptor molecules. This year lino 
auxin analogs, antiauxins, and iono- and Schafer discovered and character- 
phores, Lomax has studied the uptake of ized phototropic responses of corn me- 
the plant auxin indole-3-acetic acid by pH- socotyls to red light. These responses are 
tight vesicles from zucchini squash hy- almost certainly phytochrome-mediated. 
pocotyls. She has shown that the uptake lino has also shown by careful growth 
does not represent simple diffusion of the measurements that blue light-induced 
neutral (protonated) form of the acid phototropic curvature of coleoptiles, at 
through the vesicle membranes, but is, least in the most-sensitive range, is me- 
rather, highly specific for this active diated by differential growth: a decrease 
growth hormone. Using electron para- of growth on the illuminated side is quan- 
magnetic spin resonance techniques in titatively matched by an increase on the 
collaboration with Mehlhorn at Berke- shaded side. He also found that the growth 
ley, she has been able both to character- differential migrated down the coleoptile 
ize the pH gradient from outside to inside at rates consistent with those for auxin 
of these vesicles and to calculate the ac- transport. Taken together, lino's results 
cumulation ratio (an impressive thirty- strongly affirm the hypothesis that this 
fold) for the hormone. particular coleoptilar phototropic re- 
in a continuing study of auxin-red light sponse is mediated by light-induced lat- 
interactions, Shinkle has shown that eral transport of auxin to create an auxin 
physiologically effective concentrations differential, which is then transported 
of auxin can sensitize sections from etiol- down the coleoptile. 
ated oat coleoptiles to red light by four Brown continues her long-standing in- 
orders of magnitude. The auxin treat- terest in the nature of photosynthetic 
ment converts the elongation reaction to pigments and pigment complexes in vivo, 
red light from an LF to a VLF response. She reports significant progress in frac- 
Somewhat lower auxin concentrations donating spinach chloroplasts into three 
yield a red light fluence-response curve major complexes with little alteration of 
showing two distinct steps. The response spectral properties. Curve analysis by the 


RE SOL program indicates that all higher- previously by Fork and his collabora- 
plant chlorophyll a can not be repre- tors), the alga has three other mecha- 
sented by four spectral components, as nisms to prevent overload of system II 
was generally believed. She also docu- reaction centers by absorption of exces- 
ments the sometimes very subtle and sive system II light. The first of these 
misleading effects that certain deter- involves some change in the system II 
gents such as sodium dodecylsulfate and pigment bed leading to dissipation of the 
Triton can have on the components of absorbed energy by a mechanism other 
absorption spectra. Various technical ad- than transfer to system I. Satoh and Fork 
vances, as well as an understanding of refer to this change as a state II to state 
the pitfalls involved in attempts to so- III transition. Evidently, light-driven 
lubilize pigment-protein complexes, have proton translocation across the thylakoid 
led to real progress in the isolation and membranes plays some role in this tran- 
characterization of absorption spectral sition. System I light can drive the tran- 
components from primitive green algae sition back from state III to state II. 
and the notoriously refractory flagellate Next, Porphyra seems to be able to cope 
Euglena. Brown's initial studies of a group with excess system II light by harm- 
of primitive phytoplankton species rep- lessly cycling electrons around system II, 
resented by Mantoniella, which contains preventing the formation of excessive 
a protochlorophyll- or chlorophyll c-like amounts of oxidants or reductants. Us- 
porphyrin in addition to chlorophylls a ing fluorescence techniques, Satoh and 
and b and unusual xanthophylls, are lead- Fork readily distinguished between this 
ing to new insights both in pigment for- mechanism and the others. Finally, Mo- 
mation and in the evolution of eukaryotic hanty and Fork showed the presence of 
algae. a fourth mechanism: When there is too 
Brown has also modified the RE SOL much system II light, the alga can dis- 
program so that different half-band- sipate some of the energy through a long- 
widths can be applied to the left and right wavelength fluorescence associated with 
sides of Gaussian or Lorentzian compo- system II. (Long- wavelength fluores- 
nents in the analysis of complex spectra, cence is normally associated almost ex- 
An analysis of the red absorption band clusively with system I.) 
of chlorophyll a (in acetone) from Ana- Smith, Satoh, and Fork find that Por- 
cystis nidulans indicated that this spec- phyra has yet another bag of tricks. Un- 
trum cannot be matched completely by a der conditions of severe desiccation, this 
single Gaussian-Lorentzian component, facile alga is able to inhibit electron flow 
skewed or not. Obviously there is still on the water side of system II, inhibit 
much to learn about the basis of chloro- photoactivation of electron flow on the 
phyll spectra. reducing side of system I, and inhibit en- 
Fork and his associates have continued ergy transfer between pigment mole- 
to unravel the multitude of mechanisms cules. Inhibition of electron flow at these 
whereby the intertidal red alga For- three different sites seems necessary to 
phyra perforata copes with unfavorable prevent permanent damage by photo- 
conditions. These mechanisms range from inhibition. Evidently electron flow with 
a simple imbalance in the amount of light only part of the system blocked can lead 
reaching photosy stems I and II to the to photoinhibitory damage, 
crisis of extreme desiccation in high-in- Brand, Mohanty, and Fork have shown 
tertidal plants at low tide on a hot day. that normal functi ning of photosystem 
It appears that Porphyra has at least II in the blue-green alga Anacystis ni- 
four ways of coping with light imbalance, dulans requires the divalent cation Ca 2 + . 
Besides the ability to redistribute energy It is likely that this cation is needed for 
between the two photosystems through the stabilization of charges separated at 
the state I to state II transitions (studied the reaction center of photosystem II. In 


addition, Mohanty, Fork, and Brand drought damage to the photosynthetic 
showed that Ca 2+ plays a role in the reg- reactions of a Mexican legume, siratro 
ulation of energy distribution between (Macroptilium atropurpureum). When 
the two photosystems by a mechanism this plant is water stressed, its leaves 
associated with the state I-state II tran- move from a position of maximum to one 
sitions and, further, that a long-term Ca 2+ of minimum interception of incident light, 
depletion leads to disruption of energy If the leaves are physically constrained, 
transfer from phycobilin pigments to they show clear evidence of photoinhi- 
chlorophyll a. bition, particularly on the upper leaf sur- 
Hoshina and Fork report progress in faces. Fluorescence studies show that 
the preparation of three chlorophyll-pro- these leaf movements protect the leaves 
tein complexes from Anacystis by a com- from photoinhibition, from high-temper- 
bination of detergent treatment and ature damage, and from interactive ef- 
sucrose density gradient centrifugation. fects of high temperature and light. 
This method produces large amounts of Siratro clearly provides an excellent model 
purified chlorophyll-protein complexes system for study of important protective 
that are photochemically active. Finally, responses found in many other species, 
Hoshina used the RESOL program to including some (e.g., soybean) of eco- 
investigate the spectral consequences of nomic importance, 
temperatures above or below the phase Von Caemmerer, Coleman, and Berry 
transition for chlorophyll a in liposomes used the unicellular green flagellate 
of phosphatidylcholine. The results sug- Chlamydomonas reinhardtii to probe the 
gest that chlorophyll a 662 is specifically role of ribulose bisphosphate in regulat- 
converted to chlorophyll a 670 when the ing the rate of photosynthesis. If Chla- 
lipids pass from the liquid crystalline to mydomonas is grown under low C0 2 
the gel state, and that this change may conditions, it develops the capacity to ac- 
reflect an alteration in the aggregation cumulate carbon (from bicarbonate) to 
state of chlorophyll a. provide its ribulose bisphosphate car- 
Three reports this year deal with the boxylase with sufficient substrate for ef- 
influence of various kinds of stress on fective photosynthesis. Quantitative 
higher plants. Using simple techniques, comparison of photosynthetic behavior 
Levitt has shown that damage occurring and ribulose bisphosphate concentration 
during moderate wilting of cabbage leaves between high- and low-C0 2 -grown cells 
can be repaired during water reabsorp- clearly support the hypothesis that C0 2 
tion even if the process takes several days, concentrations inside low-C0 2 -grown cells 
Thus, conditions during the rehydration are higher (at the same external C0 2 con- 
period are perhaps as important as the centration) than they are in high-C0 2 - 
extent of dehydration in determining the grown cells. These studies also show that 
recovery of a wilted plant. In the second ribulose bisphosphate itself plays an im- 
study, Vallejos and Bjorkman report that portant role in the regulation of the car- 
a single night of low-temperature treat- boxylation reaction. There was excellent 
ment (5°C) significantly reduces the sub- agreement between responses measured 
sequent photosynthetic performance of in vivo and those predicted by in vitro 
chilling-sensitive tomato plants, even studies of the carboxylase and from the- 
without damage to photosystem II. Thus, oretical models published earlier by Far- 
photoinhibition is not the only way quhar, Berry, and von Caemmerer. 
whereby low but not freezing tempera- In a re-enactment of an historic ex- 
tures can damage chilling-sensitive plants, periment (see Year Books 68, 69, and 70), 
In the third study, Ludlow and Bjork- Nobs successfully repeated a cross be- 
man have investigated the influence of tween a C 3 and a C 4 species of Atriplex. 
paraheliotropic leaf movement (turning Because the ribulose bisphosphate ear- 
edgewise to the light) in preventing boxylases from these two species have 


clearly distinguishable kinetic proper- lation. Finally, they noted how transfer- 
ties, Berry, with Nobs, Osorio, Palmer, ring plants from high to low inorganic 
Tepperman, and Thompson, addressed carbon affects the synthesis of several 
the question as to which subunit of this additional proteins, including both sub- 
enzyme — the small one (coded for by the units of the ribulose bisphosphate car- 
nucleus) or the large one (coded for by boxylase (which may be affected 
the chloroplast) — determines these differently). 

properties. The hybrid yielded a carbox- Grossman, Talbott, and Egelhoff have 
ylase with the kinetic properties of the conducted detailed investigations to de- 
C 4 (maternal) parent. Restriction anal- termine the location of the DNA coding 
ysis of the chloroplast DNAs from both for the different proteins that form pig- 
parents and three hybrids indicated a mented granules called phycobilisomes. 
maternal inheritance for the chloro- Phycobilisomes contain the light-har- 
plasts, whereas the nuclear DNAs were vesting accessory pigments in red and 
clearly biparentally inherited. Since it is blue-green algae. In the eukaryotic al- 
well known that chloroplast DNA codes gae, where the phycobilisomes are con- 
fer the large subunit of the carboxylase, tained within chloroplasts (or, in 
it follows that the large subunit deter- Cyanophora, within a specialized chlo- 
mines the kinetic properties of these en- roplast-like organelle called a cyanelle), 
zymes. The study did not define a function the pigment-binding proteins are syn- 
for the small subunit, but it is still con- thesized within the organelle itself, as is 
ceivable that some other genetically based the protein which anchors the phycobi- 
differences in kinetics might be related lisome to the photosynthetic membrane, 
to it. All the same, future efforts at ki- The relationships among the pigmented 
netic modification of the carboxylase to polypeptides suggest that a gene family 
improve photosynthetic yield should fo- encoding both the pigment-binding and 
cus on the chloroplast genome, rather than anchor proteins is located in the chloro- 
on the far-more-complicated nuclear gen- plast. On the other hand, the so-called 
ome. linker proteins, which bind the pig- 
Several years ago, Berry and his col- mented proteins to the anchor protein, 
leagues reported the carbon-accumula- are clearly synthesized in the cytoplasm, 
tion phenomenon mentioned above in Thus phycobilisomes, like RuBP carbox- 
Chlamydomonas reinhardtii, on trans- ylase, represent a division of labor be- 
fer to low C0 2 . In a recent, discipline- tween different genomes in the 
bridging study, Coleman, Berry, and construction of a single functional unit. 
Grossman, in collaboration with Toga- Grossman and Brand report a new 
saki from Indiana University, have shown method for the rapid isolation of intact 
that the induction of C0 2 -concentrating phycobilisomes from certain algae, and 
capacity is accompanied by the de novo Grossman reports a technique for ob- 
appearance of a polypeptide (molecular taining macromolecular aggregates of 
weight 37,000) identified as the enzyme phycobiliproteins on denaturing gels. Both 
carbonic anhydrase. The enzyme is se- methods should be of considerable help 
creted into the space outside the plasma in continuing efforts to elucidate the mo- 
membrane but inside the cell wall, where lecular architecture of phycobilisomes. 
it may play an important role in main- This brief summary fails to do justice 
taining bicarbonate at equilibrium levels, to the reports that follow, nor does it give 
permitting more-rapid solubilization of full sense of the depth and breadth of the 
gaseous C0 2 and providing inorganic car- research. The following articles, and more 
bon for the transport system. Coleman importantly the published papers that 
and Grossman then showed that the pro- have or will arise from them, provide both 
tein is synthesized on cytoplasmic ribo- the details and the perspective needed 
somes and that it fails to be excreted in to measure the Department's accom- 
the presence of an inhibitor of glycosy- plishments. 

Winslow R. Briggs 




Richard B. Flavell and William F. Thompson 

The genes for ribosomal RNAs con- have been made on protein-coding genes 

stitute a prominent multigene family in transcribed by RNA polymerase II rather 

all eukaryotic organisms (reviewed by than on ribosomal genes, which are tran- 

Long and Dawid, 1980). These genes oc- scribed by RNA polymerase I. However, 

cur in tandemly repeating units, usually there are indications that a similar cor- 

at or near the sites of nucleolus organ- relation may hold for ribosomal genes in 

izers (NOR). Although the nucleotide se- at least some animal systems (e.g., Bird 

quences coding for ribosomal RNA are et al., 1981a,b). 

highly conserved during evolution, the In wheat, the relation between meth- 

copy number or repetition frequency of ylation and rDNA activity has been ex- 

the genes can vary quite rapidly within amined in a series of chromosome 

a species from one genome to another, substitution and addition lines containing 

For example, among varieties of hex- widely varying numbers of ribosomal 

aploid wheat, ribosomal gene number genes. The number of genes on each of 

varies between about 3,000 and 15,000 the four pairs of NOR chromosomes was 

per 2C nucleus. Such large variations in previously determined by hybridizing 

gene number among otherwise very sim- rRNA to DNA from a series of aneu- 

ilar organisms indicate that many, or even ploids in which the dosage of each of the 

most, of the genes are not essential for 21 chromosomes was varied, and from an 

survival. Indeed, it is likely that many additional set of intervarietal substitu- 

of the ribosomal genes are transcription- tion lines based on the original aneu- 

ally inactive (e.g., Phillips, 1978). Our ploids. Using these lines and a cytological 

interest centers on how the cell selects assay for nucleolar activity, Flavell and 

which genes should be active and which O'Dell (1979) were able to establish that 

ones should remain (or become) inactive, the activity of a given nucleolar organizer 

It seems likely that the nucleotide se- depends in large measure on the fraction 

quences of active and inactive genes are of the cell's potentially active rRNA genes 

identical in most cases, so we predict that which are contained in that organizer, 

the selection process probably involves Either the rate of transcription or the 

modifications of the DNA or chromatin number of active genes in the organizer 

structure at levels above that of the pri- thus appears to be modulated by changes 

mary nucleotide sequence. in total rDNA dosage. Data from Scheer 

One type of DNA modification that has e t al. (1976) indicate that more genes are 

received much attention in recent years transcribed as rRNA synthesis increases 

is the methylation of cytosine residues, in Triturus. 

Changes in the degree of cytosine meth- The methylation state of rDNA in wheat 

ylation have been correlated with changes lines with different numbers of rRNA 

in expression of various genes in a num- genes has been determined using the Hpa 

ber of animal systems (reviewed by Ra- II/Msp I technique of Waalwijk and 

zin and Friedman, 1981). Although there Flavell (1978). Both Hpa II and Msp I 

are many apparent exceptions, the most cleave the sequence CCGG. However, 

frequent observation is that active (or Hpa II will not do so if the internal cy- 

potentially active) genes are less meth- tosine residue is methylated. The inabil- 

ylated at one or more sites than their ity of Hpa II to cut at sites cleaved by 

inactive counterparts. Most observations Msp I is therefore an index of the degree 



of methylation at CG dinucleotides in 
CCGG sequences. From mapping exper- 
iments with cloned rDNA, and from Msp 
I digests of genomic rDNA, we know that 
there are many (often about fourteen) 
CCGG sites in each wheat rDNA re- 
peating unit. When all these sites are cut 
the rDNA is reduced to fragments of low 
molecular weight. 

Figure 1 shows an example of an ex- 
periment where wheat DNA was di- 
gested with Eco RI and Hpa II. Eco RI 
cuts once in the rDNA repeating unit, 
producing monomer fragments of about 
9 kb if there is no further cutting by Hpa 
II. A very high degree of methylation is 
indicated by the large amount of rDNA 
remaining at the 9-kb position in the gel. 
The many faint bands below 9 kb result 
from Hpa II digestion at a few unme- 
thylated sites in some of the molecules. 
Partial-digestion bands are particularly 


Fig. 1. Digestion of rDNA with Hpa II and 
Eco RI. Wheat DNA was digested with Hpa II and 
Eco RI, fractionated by agarose gel electropho- 
resis, transferred to nitrocellulose, and hybridized 
with 32 P-labeled DNA from pTA71, a clone con- 
taining an entire rRNA repeat unit. Duplicate tracks 
with slightly different amounts of DNA are shown. 
Sizes of major fragments are indicated in kilobase 

prominent at about 6.2 and at about 2.7 
+ 2.9 kb, indicating especially frequent 
undermethylation at a particular site 6.2 
kb from the left-hand Eco RI site, near 
the 5' end of the nontranscribed spacer, 
and the two smaller bands at 2.7 and 2.9 
reflect spacer-length heterogeneity within 
the rDNA complement. 

The fraction of rDNA with at least one 
Hpa II site unmet hylated was measured 
by quantitating the decrease in signal in 
the 9-kb Eco RI monomer upon digestion 
with Hpa II. Figure 2 shows that the 
number of genes cut at least once by Hpa 
II increases with the total dosage of ri- 
bosomal genes until the total dosage 
reaches about 9,000 and then levels off. 
The percentage of genes cut steadily de- 
creases from about 80% to 40% with in- 
creasing gene number, so that the number 
of genes fully methylated is increasing 
more rapidly than the number of under- 
methylated genes even before the "sat- 
uration" dose of 9,000 is reached. This 
effect of gene number on methylation is 
similar to the effect of gene number on 
cytologically detectable nucleolar activ- 
ity, in that both the average nucleolar 
activity and the fraction of undermethy- 
lated rDNA are highest when the total 
number of rRNA genes is low, and both 
decrease with increasing gene dosage. 
Thus there is a correlation between un- 
dermethylation (at least at the fre- 
quently unmethylated Hpa II site) and 
gene activity. If we assume that the con- 
trol of nucleolar activity is exerted through 
modulation of the fraction of genes sub- 
ject to transcription, this correlation 
would suggest that undermethylation and 
transcription might be causally related. 

In Fig. 2, genes are scored as under- 
methylated if they are cut at least once 
by Hpa II. Since we know that under- 
methylation occurs preferentially at a 
single site, a question arises as to whether 
or not the level of methylation at other 
Hpa II sites also varies with gene dos- 
age. The degree of overall methylation 
was estimated from the ratio of intensity 
of the 2.7 + 2.9-kb bands to the 6.2-kb 
band. Since the 6.2-kb fragment has many 






■ ^^" 

.— — -"i 


BY 4000 

Hpa II 


/ * 



J 1__ 


i . ,. 

4000 8000 12000 


Fig. 2. The number of wheat ribosomal genes cut by Hpa II as a function of total rRNA gene dosage. 
DNA from substitution and addition lines with widely varying amounts of rDNA was analyzed by 
measuring the difference in hybridization to the Eco RI 9-kb monomer band between duplicate samples 
digested either with Hpa II and Eco RI or with Eco RI alone. A single cut by Hpa II is sufficient to 
remove a given repeat unit from the 9-kb band. No distinction is made between genes cut only once and 
those cut many times. 

more Hpa II sites than the smaller frag- 
ments, a similar degree of random un- 
dermethylation will have a much greater 
effect on the 6.2-kb band. Increases in 
the ratio (2.7 + 2.9)/6.2 thus indicate de- 
creases in the overall degree of methy- 
lation. The results show that increases 
in this ratio correlate well with increases 
in the percentage of genes cut at least 
once, which we interpret as an indication 
that changes in methylation occur in par- 
allel at many sites, although the per- 
centage of molecules methylated at any 
given site may vary. 

A further correlation between rDNA 
methylation and gene activity can be de- 
duced from experiments with wheat plants 
containing chromosome 1 from Aegilops 
umbellulata in addition to the normal set 
of wheat chromosomes. Cytological. ob- 
servations by Martini, O'Dell, and Flav- 
ell (1982) showed that the nucleolus 
organizers on the A. umbellulata chro- 
mosome formed large nucleoli, while those 
on the wheat chromosomes were less ac- 
tive than in control plants, forming only 
micronucleoli. In a recent series of ex- 
periments it has also been shown that the 
A. umbellulata rDNA, which is distin- 
guishable from that of wheat by virtue 

of its longer nontranscribed spacer re- 
gion, is specifically undermethylated in 
these interspecific hybrids. In addition 
and perhaps even more significantly, the 
wheat rDNA is heavily methylated. Thus 
the suppression of wheat nucleolar or- 
ganizer activity by the A. umbellulata 
chromosome is accompanied by an in- 
crease in the level of methylation of the 
wheat ribosomal genes. 

The data obtained so far indicate that 
increases in the average level of activity 
of the ribosomal genes in a given plant 
or at a given nucleolar organizer site are 
associated with reductions in the average 
level of rDNA methylation. One inter- 
pretation of the data might be that active 
genes are less methylated than inactive 
ones. However, we do not yet have any 
way of knowing that the relationships we 
see in populations of ribosomal genes still 
hold at the level of an individual gene. 
To conclude that active genes are un- 
dermethylated, we would have to mea- 
sure the methylation status of genes 
somehow fractionated according to tran- 
scriptional activity. 

Since it may be difficult to fractionate 
ribosomal genes in this way, we have 
chosen first to examine the methylation 



status of genes showing different de- 
grees of sensitivity to digestion when iso- 
lated nuclei are treated with an 
endonuclease such as DNase I. Sensitiv- 
ity to nuclease digestion is a well-known 
property of active, or potentially active, 
genes in animal cells, and is believed to 
reflect a more-open or extended confor- 
mation of the chromosomal regions con- 
taining active genes (reviewed by 
Weisbrod, 1982). We have recently shown 
that the same general phenomenon also 
occurs in plant chromosomes (Year Book 
80, 76-79, and Spiker et aL, 1983). As in 
the case of methylation, the best-studied 
examples of DNase sensitivity involve 
protein-coding genes transcribed by RNA 
polymerase II, but there are some data 
showing similar sensitivity/activity cor- 
relations in ribosomal DNA (e.g., Bird 
etal, 1981a). 

Work is currently in progress to see 
whether the fraction of DNase-sensitive 
ribosomal genes varies in parallel with 
the fraction showing undermethylation 
of Hpa II sites, and whether the rDNA 
from Aegilops umbellulata is more sen- 
sitive than wheat rDNA in the chromo- 
some addition line where the Aegilops 
nucleolar organizer expression is domi- 

nant over that of the wheat organizers. 
Preliminary experiments indicate that 
useful data should be obtainable in these 
systems. Regardless of the outcome with 
respect to methylation, studies at the level 
of chromosome structure seem essential 
to an ultimate understanding of ribo- 
somal gene expression. 


Bird, A. P., M. H. Taggert, and C. A. Gehring, /. 

Mol. Biol. 152, 1-17, 1981a. 
Bird, A. P., M. H. Taggert, and D. Macleod, Cell 

26, 381-390, 1981b. 
Flavell, R. B., and M. O'Dell, Chromosoma 71, 

135-152, 1979. 
Long, E. D., and I. B. Dawid, Annu. Rev. Biochem. 

1+9, 727-764, 1980. 
Martini, G., M. O'Dell, and R. B. Flavell, Chro- 
mosoma 8h, 687-700, 1982. 
Phillips, R. L., in Genetics and Breeding of Maize, 

711-741, D. B. Walden, ed., John Wiley and Sons, 

Razin, A., and J. Friedman, Prog. Nucl. Acids 

Res. Mol. Biol. 25, 33-52, 1981. 
Scheer, U., M. F. Trendelenburg, and W. W. 

Franke, /. Cell Biol. 69, 465-489, 1976. 
Spiker, S., M. G. Murray, and W. F. Thompson, 

Proc. Nat. Acad. Sci. USA 80, 815-819, 1983. 
Waalwijk, C, and R. A. Flavell, Nucl. Acids Res. 

5, 3231-3236, 1978. 
Weisbrod, S., Nature 297, 289-295, 1982. 




Lon S. Kaufman, William F. Thompson, and Winslow R. Briggs 

Previous Annual Reports and a sub- 
sequent publication by Thompson et al. 
(1983) describe a group of pea transcripts 
characterized by changing abundances 
during light-regulated bud development. 
Th steady-state level for several of these 
transcripts is under the control of phy- 
tochrome. The protocols leading to these 
initial observations required three irra- 
diations, one on each of three successive 
days. We have since attempted to find a 

single day or the fewest number of days 
on which single irradiations would result 
in red light fluence-response curves ex- 
hibiting low fluence responses and pos- 
sibly very low fluence responses. 

Initially, we assayed for chlorophyll and 
fresh weight of the bud. Six-day-old dark- 
grown peas were irradiated with a single 
pulse of red light. Seedlings were re- 
turned to the dark for an additional 24 h 
before receiving 24 h of white light. Red 



light fluence-response curves measuring 
mg chlorophyll/bud, mg fresh weight/bud, 
and mg chlorophyll/mg fresh weight bud 
exhibit both low fluence responses and 
very low fluence responses (Mandoli and 
Briggs, 1981). The red light fluence curve 
for mg fresh weight/bud is shown in Fig. 
3A. The very low fluence response has 
its threshold between 10 ~ 5 and 10 ~ 4 nmol 
cm -2 red light and becomes saturated 
between 10 " 2 and 10 ~ 3 nmol cm -2 red 
light. The low fluence response has its 
threshold at 10 -1 nmol cm" 2 red light 
and does not become saturated by 10 3 
nmol cm -2 red light. These results are 
in excellent agreement with those ob- 



; a 



i i i 


/ - 



£ E 






Li_ # C7> 










i i i 

; b 



i i i 


• / ~ 


> c 


AB/. _ 

'^ -2 

o y. 


/ p _ 

~zi c 




* — • s 


o— 5 

y—o OO 




I I I 




-1 1 3 




nmol cm -2 

Fig. 3. (A) Red light fluence-response curve 
for fresh weight of the bud. Dark-grown pea seed- 
lings were irradiated on day 6, returned to the dark 
for 24 h, and given 24 h of white light. (D is dark 
control.) (B) Red light fluence-response curve for 
small subunit (SS) and a/6-binding protein (AB). 
Dark-grown pea seedlings were irradiated on day 
6 and returned to the dark for an additional 24 h. 
(D is dark control.) 

served for oat coleoptile stimulation and 
mesocotyl suppression (Mandoli and 
Briggs, 1981). 

Prior to initiating a similar set of ex- 
periments measuring transcript abun- 
dance, it was necessary to address the 
question of how to quantitate the RNA 
data. To this end, slot blots have re- 
placed dot blots. The relative density of 
each slot may be easily and reproducibly 
obtained with a densitometer. The area 
under the resulting trace may then be 
determined by a computerized digitizer. 
Construction of standard curves from a 
DNA dilution series present on each slot 
set allows us to assign a relative abun- 
dance to each density. 

Six-day-old dark-grown peas were ir- 
radiated with a single pulse of red light. 
Seedlings were returned to the dark for 
24 h, after which the buds were har- 
vested and total RNA extracted. Red light 
fluence-response curves were measured 
using hybridization techniques and the 
previously mentioned means of quanti- 
tation. Transcripts assayed thus far fall 
into one of three categories: those exhib- 
iting a low fluence response only, those 
exhibiting both a low and a very low flu- 
ence response, and those exhibiting nei- 
ther response. cDNA probes specific to 
the mRNA for the small subunit of RuBP 
carboxylase show a low fluence response 
only. cDNA probes specific to the mRNA 
for the chlorophyll alb binding protein 
show both a low fluence and a very low 
fluence response. The fluence response 
curves for these probes are shown in Fig. 
3B. The very low fluence response has 
its threshold at 10 " 4 nmol cm" 2 and be- 
comes saturated between 10 ~ 2 and 10 ~ 3 
nmol cm -2 . The low fluence response has 
its threshold between 10 " l and 10° nmol 
cm -2 red light and does not become sat- 
urated by 10 3 nmol cm" 2 red light. These 
values are in excellent agreement with 
those observed for mg chlorophyll/bud, 
mg fresh weight/bud, and mg chloro- 
phyll/mg fresh weight. 

Far red light will reverse the low flu- 
ence response measured for either fresh 
weight or transcript abundance. Rever- 


sal is saturated at 7.5 x 10~ 2 J cm" 2 far 
red light. 

In summary, peas irradiated with sin- 
gle pulses of red light exhibit both very 
low and low fluence responses for a va- 
riety of physiological parameters. Light- 
regulated transcripts may show low 
fluence responses only (i.e., the small 
subunit of RuBP carboxylase), both low 
and very low fluence responses (i.e., the 

chlorophyll alb binding protein), or nei- 
ther response. 


Thompson, W. F., M. Everett, N. 0. Polans, R. A. 

Jorgensen, and J. D. Palmer, Planta, in press, 

Mandoli, D. F., and W. R. Briggs, Plant Physiol. 

67, 733-739, 1981. 


David B. Stern ayid Jeffrey D. Palmer 

Plant mitochondrial DNAs (mtDNA) 
are quite large in comparison to their 
fungal and mammalian counterparts and 
are also highly variable in size, ranging 
from 215 kilobase pairs (kb) in the genus 
Brassica to about 2,400 kb in muskme- 
lon. Despite their large size, plant 
mtDNAs appear to encode only a few 
more mitochondrially synthesized poly- 
peptides than do the smallest known 
mtDNAs, those of metozoan animals, 
which are approximately 16 kb in length. 
In seems likely, therefore, that higher 
plant mtDNA consists largely of non- 
coding sequences. 

An unusual aspect of plant mitochon- 
drial genome structure is the presence 
of sequences which have homology to 
chloroplast DNA (ctDNA). Corn mtDNA, 
for example, has been demonstrated to 
contain a 12-kb portion of the corn ctDNA 
inverted repeat (Stern and Lonsdale, 
1982). We have since extended these ear- 
lier results, and have shown that nu- 
merous sequence homologies exist 
between cloned mung bean and spinach 
ctDNA restriction fragments and 
mtDNAs from corn, mung bean, spinach, 
and pea (Stern et al., 1983; Stern and 
Palmer, 1983). Our experimental ap- 
proach has been to use 32 P-labeled ctDNA 
clones spanning most of the mung bean 
chloroplast genome (Palmer and Thomp- 
son, 1981) as hybridization probes against 
Southern blots of restriction endonu- 

clease digestions of the mitochondrial and 
chloroplast DNAs from each species. The 
ctDNA is included in a lane next to 
mtDNA prepared from the same plant 
as a control, since cross-contamination 
between organellar DNAs is an inevi- 
table artifact of their isolation. In inter- 
preting hybridization results, then, one 
can distinguish between mtDNA frag- 
ments with homology to ctDNA and 
ctDNA fragments contaminating the 
mtDNA preparation. 

The major finding from these studies 
is that the presence of ctDNA sequences 
in the mitochondrion is a pervasive phe- 
nomenon in plants. All the ctDNA clones 
tested hybridized to at least one mtDNA 
restriction fragment (Fig. 4), although 
the strongest cross-homologies are be- 
tween clones derived from the ctDNA 
inverted repeat and mtDNA from corn 
and pea. We have also found that the 
chloroplast gene for the beta subunit of 
the chloroplast ATPase has strong ho- 
mology to a segment of mung bean 
mtDNA, and that the gene sequence 
(rbcL) for the large subunit of RuBP car- 
boxylase is contained in corn mtDNA. 

The presence of ctDNA sequences is a 
novel feature of plant mitochondrial gen- 
omes. One cannot rule out, however, that 
some of these sequences actually origi- 
nated in the mitochondrion or its pro- 
genitor and were then transferred to the 
chloroplast. This is unlikely in the case 



MB20-5 MB16-2 




J MB 136 J | T t MB12-8 fMB9-7 f MB 188 f 

MB 17-2 

MB7-5 I MB70 MB&6 


16S 23S 

23S 16S 

psbA atpBE 

i j 

i i 

- . CORN 

'---- 1 MUNG BEAN 


l_.j PEA 

Fig. 4. Top: Map of the mung bean chloroplast genome adapted from Palmer and Thompson (1981). 
Pst I sites (arrows) and Sal I sites (boxes) are indicated. The clones used in this study are shown on the 
map: for example, MB 18.8 is a clone of an 18.8-kb Pst I fragment. The positions of the genes are 
according to Palmer and Thompson (1981), Palmer et al. (1982; unpublished data). The heavy lines just 
beneath the ctDNA map indicate the inverted repeats. Bottom: Schematic representation of the hy- 
bridization strength of the 32 P-labeled ctDNA clones to mtDNA. Hybridizations were classified as very 
weak (dotted lines), weak (thin solid lines), strong (medium lines), or very strong (heavy lines). Clones 
not tested were MB 12.8, MB 17.2, and MB 20.5, whose sequences are largely contained within the 
inverted repeat and were mostly represented by MB 18.8 and MB 16.2, and MB 1.2. Hybridizations are 
shown directly under the cloned mung bean ctDNA fragment used as a probe. 

of sequences that are clearly necessary 
for chloroplast survival, such as the 
transfer and ribosomal RNA (rRNA) 
genes, some of which are contained within 
the ctDNA inverted repeat, and rbcL. 

Two types of mechanisms could ac- 
count for random and widespread se- 
quence transfer between cytoplasmic 
organelles. One type would require di- 
rect physical contact. Outer membrane 
continuities between chloroplast and mi- 
tochondrion have been observed in thin 
sections of the fern Pteris vittata (Crotty 
and Ledbetter, 1973). Also, enclosure of 
several mitohondria by a single chloro- 
plast was observed in the grass Panicum 
schenckii (Brown et al., 1983). Mem- 
brane continuities might facilitate inter- 
molecular recombination, whereas 
enclosure of mitochondria by a chloro- 
plast suggests transformation of the mi- 
tochondrion as a likely mechanism for 
ctDNA uptake. On the other hand, ex- 
change of DNA sequences between or- 
ganelles may not require direct physical 
contact. DNA released into the cyto- 
plasm from broken or lysed chloroplasts 
may be taken up by transformation of the 
mitochondria. Alternatively, there may 
exist in the cytoplasm specific vector 
molecules capable of transferring se- 

quences between organelles. If such a 
vector could be isolated, it could prove 
invaluable for introducing genetic ma- 
terial simultaneously into chloroplasts and 
mitochondria of the same plant. Such a 
system, however, might exhibit se- 
quence-specific selectivity. This could ac- 
count for the apparently higher frequency 
of transfer of certain ctDNA sequences, 
such as parts of the inverted repeat. 

We feel it unlikely that these inte- 
grated ctDNA sequences play any bio- 
logical role in the mitochondrion. 
Significantly, the strongest cross-homo- 
logies we have observed are to ctDNA 
sequences unlikely to have a function in 
the mitochondrion: rbcL (corn) and the 
chloroplast rRNA genes (corn and pea). 
Rather, we feel that the widespread 
presence of ctDNA sequences in plant 
mitochondrial DNAs is best regarded as 
a dramatic demonstration of the dynamic 
nature of interactions between the chlo- 
roplast and the mitochondrion, similar to 
the ongoing process of interorganellar 
DNA transfer already documented be- 
tween the mitochondrion and the nucleus 
in fungi and animals (Van den Boogaart 
et al. , 1982; Wright and Cummings, 1983; 
Gellissen et al., 1983; Farrelly and Bu- 
tow, 1983; Jacobs et al., 1983). 



Brown, R. H., L. L. Rigsby, and D. E. Akin, Plant 

Physiol. 71, 437-441, 1983. 
Crotty, W. J., and M. C. Ledbetter, Science 182, 

839-841, 1973. 
Farrelly, F., and R. A. Butow, Nature 301, 296- 

301, 1983. 
Gellissen, G., J. Y. Bradfield, B. N. White, and 

G. R. Wyatt, Nature 301, 631-634, 1983. 
Jacobs, H. T., J. W. Posakony, J. W. Grula, 

J. W. Roberts, J.-H. Xin, R. J. Britten, and E. H. 

Davidson, /. Mol. Biol. 165, 609-632, 1983. 
Palmer, J. D., and W. F. Thompson, Proc. Nat. 

Acad. Sci. USA 78, 5533-5537, 1981. 


Palmer, J. D., H. Edwards, R. A. Jorgensen, and 
W. F. Thompson, Nucl. Acids Res. 10, 6819- 
6832, 1982. 

Stern, D. B., and D. M. Lonsdale, Nature 299, 698- 
702, 1982. 

Stern, D. B., and J. D. Palmer, Proc. Nat. Acad. 
Sci. USA, in press, 1983. 

Stern, D. B., and J. D. Palmer, W. F. Thompson, 
and D. M. Lonsdale, UCLA Symposia on Mo- 
lecular and Cellular Biology, New Series, Vol- 
ume 12, Robert B. Goldberg, ed., Alan R. Liss, 
Inc., New York, in press, 1983. 

Van den Boogaart, P., J. Samallo, and E. Agster- 
ibbe, Nature 298, 187-189, 1982. 

Wright, R. M., and D. J. Cummings, Nature 302, 
86-88, 1983. 


Jeffrey D. Palmer and Clark R. Shields* 

By far the largest mitochondrial DNAs 
known are those of higher plants, which 
range in size and sequence complexity 
from about 200 kilobase pairs (kb) to over 
2,000 kb (Leaver and Gray, 1982). At the 
present time, the physical organization 
of higher-plant mitochondrial DNA is un- 
clear, although various models have been 
proposed which postulate a complex or- 
ganization of the genome into a collection 
of circles differing in size and sequence 
information (Leaver and Gray, 1982). 

We have chosen to study the structure 
and function of one of the smallest known 
plant mitochondrial DNAs, that of var- 
ious species in the genus Brassica (cab- 
bage, turnip, mustard, rapeseed oil). 
Restriction fragment analysis of the 
Brassica mitochondrial genome indicates 
a size of just over 200 kb (Lebacq and 
Vedel, 1981; Palmer et al., 1983). We have 
already described an unusual, linear ex- 
trachromosomal plasmid found in the mi- 
tochondria of certain fertile and male- 
sterile Brassica cytoplasms (Palmer et 
al., 1983) and will not discuss this mol- 
ecule any further in this report. Instead, 

* Department of Vegetable Crops, University of 
California, Davis. 

we shall briefly summarize our recent 
studies on the physical organization of 
the main mitochondrial genome in Bras- 

Digestion of mitochondrial DNA from 
Brassica campestris (Chinese cabbage, 
turnip) with Pst I and Sal I produced a 
series of 29 and 26 restriction fragments, 
respectively. For each enzyme, all frag- 
ments have the same stoichiometry except 
for one pair of fragments which is at ap- 
proximately one-third the stoichiometry of 
the rest and one pair which is at approx- 
imately two-thirds stoichiometry. The Vs 
and % stoichiometry pairs sum to the same 
molecular weight for each enzyme (26 kb 
for Pst I and 36 kb for Sal I), and sub- 
traction of one pair from each total frag- 
ment summation (243 kb and 251 kb) gives 
a corrected genome size estimate of 217 
kb (Pst I) and 215 kb (Sal I). Restriction 
maps of all four substoichiometric Pst I 
fragments reveal that they share a re- 
peated sequence of approximately 3 kb, as 
defined by common Bam HI, Bgl I, and 
Pvu II sites (Fig. 5). Furthermore, re- 
striction mapping and filter hybridizations 
indicate that each Pst I fragment is ho- 
mologous on one side of the repeat to one 
of the other three fragments and, on the 


Don k I egm fji ef If Imlg ee i f 


g k 

P 4.8 

kh hef Ik 


kh he f If Imlg ee i f g 

J I I 

P10.1 , 

I egm fji e f 

e -BamHI 
f -Bgl I 
i -Kpn I 

j-Nru I 
k-Pst I 
I -Pvull 
m-Sal I 

Fig. 5. Restriction maps of the four substoichiometric Pst I fragments of B. campestris mitochondrial 
DNA. Fractions Vs and % at left indicate fragment stoichiometry. The 3-kb repeat sequence (efl) is flanked 
by four paired combinations of four unique sequences (klegmfjie, lflmlgeeifgk, khhe, Ik). Solid bar indicates 
the minimum extent of the repeat sequence; open extensions of the bar indicate its maximum extent. 

other side of the repeat, is homologous to 
a second one of the other three (Fig. 5). 
Thus, the Ys stoichiometry pair (P21.1 and 
P4.8) is equivalent in sequence complexity 
to the % pair (P15.7 and P10.1), and the 
two pairs are interconvertible by a single 
reciprocal crossover within the 3-kb re- 

The relative order of these four sub- 
stoichiometric fragments and of the re- 
maining fragments, of normal stoi- 
chiometry, was determined by hybrid- 
izing each Pst I or Sal I fragment (cloned 
into the plasmid vector pUC8) to a rep- 
lica nitrocellulose filter containing total 
mitochondrial DNA digested with Pst I, 
Pst I-Sal I, and Sal I. Together with 
double digestions of the individual clones 
with Pst I and Sal I, these hybridizations 
enable construction of a physical map 
which accounts for all of the observed 
restriction fragments (Fig. 6). The sim- 
plest model consistent with both the link- 
age relationship data and the homologies 
among the four substoichiometric Pst I 
fragments (Fig. 5) is that the 215-kb B. 
campestris mitochondrial genome con- 
sists of three circular chromosomes (Fig. 
6). One of the circles (215 kb) contains 
the entire genome, including two copies 
of the 3-kb element present as direct re- 

peats separated from one another by 80 
kb. We postulate that reciprocal recom- 
bination occurs between intramolecu- 


Fig. 6. Three circular linkage groups of the B. 
campestris mitochondrial genome. Arrows denote 
positions and relative orientations of the 3-kb re- 
peat sequence (Fig. 5). The three molecules are 
drawn to scale at their outside circumferences. 



Fig. 7. Model to explain the origin of the three circular linkage groups (Fig. 6) of B. campestris 
mitochondrial DNA via recombination within the 3-kb repeat sequence. 

larly paired repeat sequences (Fig. 7) to 
generate two smaller circles of 80 kb and 
135 kb (Fig. 6). Based on the stoichi- 
ometry of the four repeat-containing Pst 
I fragments, it is estimated that the two 
small circles are present in roughly equal 
amounts and at twice the molarity of the 
215-kb circle. 

The process of recombination between 
repeated sequences to give alternate 
states of Brassica mitochondrial genome 
organization — either two small circles or 
one large cointegrate circle — is analo- 
gous to the process of lambda phage in- 
tegration-excision into the E. coli 
chromosome (Nash, 1981) and to that of 
bacterial transposable-element cointe- 
grate formation-resolution (Kostriken et 
at. 1981). By analogy to these and 
mechanistically related specialized site- 
specific recombination systems (Simon et 
al., 1980; Broach, 1982; van der Putte et 
al.j 1980), we hypothesize the existence 
of a specialized recombination system, 
possibly encoded within the Brassica mi- 
tochondrial genome and physically linked 
to the 3-kb repeat sequence, which me- 
diates recombination between specific 
recombination sites embedded within the 
3-kb repeats. Further analogy to pro- 
cesses of low-frequency recombination 
between lambda DNA and related sec- 
ondary sites in the E. coli chromosome 

(Nash, 1981) lead to the prediction that 
recombination may also occur between 
the strong recombination sites located in 
the Brassica 3-kb repeats and the weaker 
recombination sites located elsewhere in 
the genome. This possibility will be stud- 
ied in Brassica relative to the high levels 
of small, circular mitochondrial DNAs 
observed in suspension cultures of other 
plants (Sparks and Dale, 1980; Dale et 
al., 1981; Dale, 1981; Brennicke and Blanz, 


Brennicke, A., and P. Blanz, Mol. Gen. Genet. 187, 

461-466, 1982. 
Broach, J. R., Cell 28, 203-204, 1982. 
Dale, R. M. K., Proc. Nat. Acad. Sci. USA 78, 

4453-4457, 1981. 

Dale, R. M. K., J. H. Duessing, and D. Keene, 

Nucl. Acids Res. 9, 4583-4593, 1981. 
Kostriken, R., C. Morita, and F. Heffron, Proc. 

Nat. Acad. Sci. USA 78, 4041-4045, 1981. 
Leaver, C. J., and M. W. Gray, Annu. Rev. Plant 

Physiol. 33, 373-402, 1982. 
Lebacq, P., and F. Vedel, Plant Sci. Lett. 20, 1- 

9, 1981. 
Nash, H. A., Annu. Rev. Genet. 15, 143-167, 1981. 
Palmer, J. D., C. R. Shields, D. B. Cohen, and 

T. J. Orton, Nature 301, 725-728, 1983. 
Simon, M., J. Zieg, M. Silverman, G. Mandel, and 

R. Doolittle, Science 209, 1370-1374, 1980. 
Sparks, R. B., Jr., and R. M. K. Dale, Mol. Gen. 

Genet. 180, 351-355, 1980. 
van der Putte, P., S. Cramer, and M. Giphart- 

Gassler, Nature 286, 218-221, 1980. 




Neil 0. Polans and William F. Thompson 

Fundamental to an understanding of or Bam HI) and fractionated them ac- 
the multilocus associations that charac- cording to fragment size on agarose gels, 
terize complex genetic systems is the ex- After transferring these DN A fragments 
istence of appropriate marker loci to serve to reusable membranes (Southern, 1975), 
as experimental tools. A potential source specific sequences were localized by hy- 
of such loci is the collection of cDNA clones bridization with a radioactive pea cDNA 
previously described in connection with clone belonging to the previously de- 
phytochrome regulation of pea transcript scribed collection, 
levels {Year Book 81, 101-105). Several In place of more conventional nitro- 
of these clones represent nuclear genes cellulose filters, rehybridizable transfer 
that encode messenger RNAs whose membranes are used to maximize the ef- 
abundance is regulated by light acting ficiency and comparability of the screen- 
through the phytochrome system ing process. In this way, the same set of 
(Thompson et al., 1983); as such, these DNAs, located on a single blot, can be 
clones may present a unique opportunity hydridized perhaps as many as a dozen 
to investigate the relationships among a times, with different probes recognizing 
group of genes expressing regulatory and different DNA sequences. Total-leaf DN A 
possibly functional associations. can be prepared in a rapid, small-scale 

Historically, genetic studies have used procedure (Jorgensen, unpublished mod- 
marker loci encoding morphological, ification of procedure in Murray and 
quantitative, and, more recently, bio- Thompson, 1980). If appropriate, one or 
chemical traits. In particular, protein two cycles of CsCl banding may be added, 
electrophoresis has been widely used in The "quick-prep" DNA yields acceptable 
such investigations because of the ease results, especially when pea DNA is used 
with which unambiguous genetic assign- in conjunction with pea probes. As ho- 
ment can be made (e.g., see Marshall and mology between DNA and probe de- 
Allard, 1969). An analogous method is to creases, however, it may become 
analyze DNA restriction fragment length necessary to include the additional CsCl 
polymorphisms (see Botstein et al. , 1980), purification steps which also act to re- 
which defines presumptively neutral, co- duce general background, 
dominant marker loci but which is not Preliminary results indicate substan- 
restricted to genes encoding soluble en- tial restriction fragment length poly- 
zymes. DNA "alleles" are not only useful morphism across a range of Pisum 
as classical Mendelian segregates but can accessions including several wild annual 
extend analyses to variations occurring species in addition to commercial vari- 
in different classes of DNA, including eties, experimental lines, subspecies, land 
noncoding DNA or sequences flanking a races, and primitive cultivars of Pisum 
gene of interest. sativum. Although DNAs from different 

To examine the potential of using re- sources did occasionally display similar 

striction fragment length polymorphisms band-pattern phenotypes, the general 

in the study of phytochrome-regulated result was widespread variation among 

pea loci, we carried out digestion of DNAs lines, varieties, and populations. In fact, 

from various Pisum accessions and pea where interlineal similarity did occur, it 

relatives with an appropriate restriction was usually restricted to a single probe, 

endonuclease (usually Eco RI, Hind III, Members of the same line, variety, or 


-- > 

Fig. 8. Genomic blot of experimental pea line DNAs hybridized with pea cDNA clone, pAB96. DN As 
extracted from the leaves of single pea plants and banded through CsCl were digested with Eco RI, 
fractionated on agarose gels, and bound to Gene-Screen membranes. These were then probed with pAB96, 
a pea cDNA clone corresponding to a portion of the alb light-harvesting complex. Lanes 1 and 2 (from 
left to right), which contain DNAs from different plants belonging to the same line, show identical band 
patterns. The remaining DNAs are all derived from different lines. Lanes 5 and 10, and lanes 7 and 11, 
also show what may be identical band patterns. Lanes 5 and 7, and lanes 10 and 11, display similar but 
not identical patterns. Lane 6 is blank. Seed material was obtained through the courtesy of G. A. Marx. 
R. A. Jorgensen kindly provided the particular DNA samples presented. From left to right, the lanes 
contain DNA from the following accessions (G. A. Marx, NYS Agr. Exp. Sta.): B77-288U), B77-288(2), 
B77-289, B77-259, B77-266, blank lane, B77-273, B77-276, B77-279, B78-245, B78-246, B78-248, B78-249. 

population, however, displayed identical with pAB96 (a cDNA clone obtained from 
restriction fragment banding patterns — N.-H. Chua), which corresponds to poly- 
a result consistent with the highly selfing peptide 15 of the chlorophyll alb light- 
nature of pea. harvesting complex (Broglie et al. , 1981). 
An example of the variation observed Although both similar and identical pat- 
in pea restriction fragments is presented terns are in evidence among the DNA 
in Fig. 8. The DNAs pictured were ex- samples (see Fig. 8 legend), the com- 
tracted from individual plants belonging plexity of the patterns and the high level 
to a series of experimental lines, digested of polymorphism displayed is represen- 
with Eco RI, and bound to Gene Screen tative of each combination of DNA, re- 
rehybridizable membranes (New Eng- striction endonuclease, and probe 
land Nuclear). They were then probed examined to date. 



The preliminary evidence, then, indi- 
cates a wealth of phenotypic variation for 
pea restriction endonuclease fragments 
corresponding to the phytochrome-reg- 
ulated genes examined, and suggests a 
good potential source of marker loci for 
studies involving this multilocus system. 
We hope to extend these studies to all of 
the pea cDNA clones of interest, using 
DNAs prepared from a series of exper- 
imental lines for which other morpho- 
logical, quantitative, and biochemical 
markers already exist. After experimen- 
tally verifying the segregational integ- 
rity of these restriction endonuclease 
fragments with extant F 2 lines, we hope 
to use a few of the most variable parental 
lines and corresponding tester strains to 

map these loci (i.e., the restriction frag- 
ments they generate) with respect to the 
numerous traditional marker loci previ- 
ously mentioned. 


Botstein, D., R. L. White, M. Skolnick, and R. W. 

Davis, Amer. J. Hum. Genet. 32, 314-331, 1980. 
Broglie, R. G., G. Bellemare, S. G. Bartlett, N.- 

H. Chua, and A. R. Cashmore, Proc. Nat. Acad. 

Sci. USA 78, 7304-7308, 1981. 
Marshall, D. R., and R. W. Allard, J. Hered. 60, 

17-19, 1969. 
Murray, M. G., and W. F. Thompson, Nucl. Acids 

Res. 8, 4321-4326, 1980. 
Southern, E. M., J. Mol. Biol. 98, 503-517, 1975. 
Thompson, W. F., M. Everett, N. O. Polans, R. A. 

Jorgensen, and J. D. Palmer, Planta, in press, 



John C. Watson, Jeffrey D. Palmer, and William F. Thompson 

Chloroplasts contain the complete 
translational and transcriptional machi- 
neries necessary for chloroplast genome 
expression. Chloroplast DNA encodes 
only a portion of the translational appa- 
ratus; the remainder of the required genes 
are located in the nucleus. The ribosomal 
RNA and transfer RNA genes of chlo- 
roplasts have been widely investigated. 
However, study of the genes encoding 
protein components of the translational 
system is only beginning. The gene that 
encodes elongation factor Tu {tufA) has 
recently been mapped on the chloroplast 
genomes of Chlamydomonas reinhardtii 
(Watson and Surzycki, 1982) and Eu- 
glena gracilis (Passavant et al., 1983). 
The tufA gene was located in C. rein- 
hardtii by heterologous DNA:DNA hy- 
bridizations, using cloned genes from 
Escherichia coli as hybridization probes. 
Homology between chloroplast DNA and 
E. coli genes for translational and tran- 
scriptional components is not limited to 
the tufA gene. Indeed, C. reinhardtii 

chloroplast DNA also shares homology 
with genes of both the S10 and alpha ri- 
bosomal protein operons, as well as with 
the genes encoding the beta and beta' 
subunits of RNA polymerase (Watson and 
Surzycki, 1983). It seemed possible that 
cloned C. reinhardtii or E. coli gene 
probes might be similar enough to the 
corresponding higher-plant chloroplast 
genes to allow their identification in 
DNAiDNA hybridizations. Here we have 
concentrated on the chloroplast genomes 
of pea, mung bean, and spinach using as 
probes the C. reinhardtii tufA gene and 
part of the E. coli S10 r-protein operon. 
The tufA gene probe used here, en- 
coding approximately one-half of the 3' 
end of C. reinhardtii tufA sequences, was 
isolated from plasmid pCpl6 (pCpl6 was 
the gift of J.-D. Rochaix; Rochaix, 1978). 
To begin an analysis of tufA expression, 
this cloned fragment, Eco 25', was used 
to examine tufA transcripts in Chla- 
mydomonas. After electrophoresis of 
Chlamydomonas RNA (gift of Arthur 


Grossman) on a denaturing agarose gel, identified. The coding regions have been 
filter blots of the gel were reacted with further localized by analyzing cloned Pst 
the tufA probe. An abundant transcript, I fragments (Palmer and Thompson, 
primarily present in the nonpolyadeny- 1981b) digested with Pst I and a second 
lated RNA fraction and about 1600 bases enzyme. The restriction fragments to 
in length, is the major transcript of the which the tufA probe hybridizes in the 
Chlamydomonas tufA gene. The E. coli three different species have been found 
tufA gene is about 1200 base pairs in to share homology in previous chloro- 
length, so the Chlamydomofias gene may plast DNA rearrangement studies (Pal- 
be approximately 400 base pairs longer, mer and Thompson, 1982), and the same 
presuming the absence of large, untrans- situation pertains to the fragments hy- 
lated regions in the Chla?riydo?nonas bridizing to the S10 operon probe, 
transcript. The E. coli S10 r-protein op- In mung bean and spinach, tufA is lo- 
eron probe, about 2.3 kb in length, was cated within the large single-copy region 
cloned from specialized transducing phage of the chloroplast genome. About 30 kb 
lambda fusS (Watson and Surzycki, 1983) separates tufA from the junction of the 
and carries genes for r-proteins L2, L3, inverted repeat and large single-copy re- 
L4, L23, and either L22 or S19 (Watson gion. In pea, tufA is separated by 5-6 
and Surzycki, 1983). It is the gene for kb from the 3' end ofthe23SrRNA gene, 
either r-protein L22 or S19 which hy- This placement suggests that, as in Chla- 
bridizes to C. reifihardtii chloroplast mydoinonas, tufA exists in a single copy 
DNA, about 10 kb from the tufA gene, per chloroplast chromosome in pea, mung 
Regions of homology to these two bean, and spinach. However, we have 
probes were located on the physical maps not yet been able to demonstrate de- 
of pea, mung bean (Palmer and Thomp- tectable homology between the E. coli 
son, 1981a), and spinach (Herrmann et tufA gene and higher-plant chloroplast 
al., 1980) chloroplast DNAs by probing DNAs, unlike the situation in C. rein- 
filter blots of agarose gels containing re- hardtii (Watson and Surzycki, 1982). In 
striction digests of these chloroplast mung bean and spinach, the smallest 
DNAs. This analysis identified a single fragments that we have found to react 
Pst I fragment in each species that re- with the E. coli S10 operon probe are 1- 
acted with the tufA probe (Table 1). The 4 kb in length and are clearly present in 
S10 operon probe hybridized to a single two copies per chromosome. That is, the 
Pst I fragment in pea, while in both mung "S10 region" maps within or near the ends 
bean and spinach two Pst I fragments of the inverted repeats which border the 
reacted with this probe (Table 1). Re- large single-copy region, about 20 kb from 
striction fragments produced by other the 5' end of the 16S rRNA gene. The 
enzymes, overlapping the Pst I frag- "S10 region" is about 45-46 kb from the 
ments listed in Table 1, have also been 23S rRNA gene in pea. Recently, the 

nucleotide sequence of a region near the 

end of the inverted repeat in tobacco 

TABLE l. Hybridization of tufA md S10 (Sugita and Sugiura, 1983) revealed the 

Operon Probes to Chloroplast DNAs ^ » » ' 

_ presence of a gene encoding a protein 

Filter-bound Pst I Fragment with substantial homology with E. coli 
Hybridized* r-protein S19. This evidence, together 
Sio Operon with the fact that the gene for either L22 
Species tufA Probe Probe r S19 hybridizes to Chlamydomo?ias 
P ea 57 9^2 cpDNA, and the analogous position of 
Mung Bean 5.6 16.2, 17.2 the fragments hybridizing to the S10 op- 
Spinach 13.5 8.9, 8.1 eron probe described here, suggests that 
* Individual Pst I fragments are identified by size it is the gene for r-protein S19 (rpsS) 
in kb. which we have detected here. Because 



of its location, it is possible that tran- 
scripts through this region are altered 
due to the recombinational events lead- 
ing to chloroplast DNA isomerization 
(Palmer, 1983). 

A study of the transcripts arising from 
tufA and the "S10 region" in pea and mung 
bean has been initiated. Cloned subfrag- 
ments (of the Pst I fragments) overlap- 
ping these two regions hybridize to 
multiple transcripts in gel blots of total 
RNA from both plant species. When RNA 
from dark-grown and light-grown plants 
are compared in this way, it appears that 
light-grown plants have a higher steady- 
state level of these transcripts than plants 
grown in the dark. A preliminary anal- 
ysis of the red light fluence response of 
tufA transcript abundance indicates the 
presence of both very low fluence and low 
fluence responses (L. Kaufman, unpub- 
lished data). These data suggest that tufA 
transcript abundance may be regulated 
by phytochrome, although we have not 
yet demonstrated far red reversibility. 

In summary, the single gene encoding 
elongation factor Tu has been located on 
chloroplast genome maps of pea, mung 
bean, and spinach; regions homologous 
to the E. coli S10 r-protein operon have 

likewise been located. In the coming year, 
we plan to investigate further the struc- 
tures of the genes identified. Knowledge 
of gene structure will facilitate analyses 
of transcriptional organization and light 
regulation of transcript accumulation. The 
elongation factor and presumptive r-pro- 
tein genes provide an important contrast 
to virtually all chloroplast protein-encod- 
ing genes now under study, since they 
are not directly involved in photosyn- 
thesis but rather are involved in the 
expression of chloroplast genes. 


Herrmann, R. G., P. R. Whitfeld, and W. Bottom- 
ley, Gene 8, 179-191, 1980. 

Palmer, J. D., and W. F. Thompson, Proc. Nat. 
Acad. Sci. USA 78, 5533-5537, 1981a. 

Palmer, J. D., and W. F. Thompson, Gene 15, 21- 
26, 1981b. 

Palmer, J. D., and W. F. Thompson, Cell 29, 537- 
550, 1982. 

Palmer, J. D., Nature 301, 92-93, 1983. 

Passavant, C. W., G. L. Stiegler, and R. B. Hal- 
lick, J. Biol. Chem. 258, 693-695, 1983. 

Sugita, M., and M. Sugiura, Nucl. Acids Res. 11, 
1913-1918, 1983. 

Rochaix, J.-D., J. Mol. Biol. 126, 597-617, 1978. 

Watson, J. C, and S. J. Surzycki, Proc. Nat. Acad. 
Sci. USA 79, 2264-2267, 1982. 

Watson, J. C, and S. J. Surzycki, Curr. Genet., 
7, 201-210, 1983. 



Jeffrey D. Palmer and William F. Thompson 

One of the main thrusts of our research 
over the last several years has been to 
compare the physical organization of var- 
ious angiosperm chloroplast genomes in 
order to provide an evolutionary descrip- 
tion of chloroplast DNA rearrangement 
processes (Palmer and Thompson, 1981a, 
1981b, 1982; Palmer, 1982). The major 
conclusions from these studies follow. 
(1) The chloroplast genomes of represen- 
tatives of all 40 families of angiosperms 
examined contain a large, ribosomal 

RNA-encoding inverted repeat some 
20,000-25,000 base pairs in size. (2) The 
only documented loss of this ancient, in- 
verted-repeat structure has occurred 
within one section of the family Leg- 
uminosae, including such species as pea 
and broad bean. (3) There exists a strong 
correlation between the presence of the 
inverted repeat and a stable chloroplast 
genome in which major sequence rear- 
rangements are rare. (4) Conversely, 
numerous rearrangements have oc- 


curred in those legume species that have broad bean allows us to choose between 

lost the inverted-repeat structure. the two hypotheses previously advanced 

We have postulated two equally ten- (see second paragraph) to explain why 

able hypotheses to explain the last two the pea and broad bean genomes are so 

observations, above. (1) The inverted re- rearranged. It now appears unlikely that 

peat may stabilize the chloroplast gen- the deletion of the inverted repeat oc- 

ome against rearrangements. In this case curred simply as one of many rearrange- 

the loss of the inverted repeat in certain ments resulting from a generally 

legumes would lead directly to an in- destabilized chloroplast genome. Rather, 

creased frequency of rearrangements, the evident antiquity of the inverted re- 

(2) The chloroplast genome of certain peat deletion relative to other pea-broad 
legumes somehow acquired the ability to bean rearrangements considerably 
rearrange, and the deletion of the in- strengthens the possibility that there is 
verted repeat simply represents one of a direct, causal link between these two 
many rearrangements following this ac- sets of mutations. At the same time, the 
quisition. possibility still remains that the in- 

During the last year we have concen- verted-repeat deletion occurred as a very 
trated our studies on several additional early event soon after some other gen- 
legumes which either retain or lack the ome destabilization event, such as a 
inverted repeat (Palmer, 1983; Palmer et "transposon infection" or an alteration in 
al., 1983a, 1983b, 1983c). The major find- chloroplast recombination systems, 
ings of these studies follow. (1) Legume We have developed a general model for 
chloroplast DNAs that retain the in- chloroplast genome evolution which at- 
verted repeat (from mung bean, soy- tempts to explain the origin of the var- 
bean, and common bean) are entirely co- ious chloroplast genome types that we 
linear in sequence arrangement. Thus, have described. The starting point for 
the legumes as a family are not charac- this model is the recent demonstration 
terized by an increased chloroplast gen- that inverted repeat-containing chloro- 
ome instability relative to other groups plast DNAs from common bean (Palmer, 
of angiosperms. (2) Alfalfa chloroplast 1983), soybean (Palmer et al., 1983a), 
DNA has a "missing link" status among three fern species (our unpublished data), 
legume chloroplast DNAs in the sense and from the cyanelle of Cy anaphora 
that it has lost one entire segment of the paradoxa (Bohnert and Loffelhardt, 1982) 
inverted repeat but is otherwise unrear- all exist as two equimolar populations of 
ranged relative to legume chloroplast molecules differing only in the relative 
DNAs that retain the repeat. Together orientation of their single-copy se- 
with our previous findings, this result al- quences. From analogies to a number of 
lows the following additional conclusions, well-characterized systems possessing 

(3) The loss of the inverted repeat oc- similar or mechanistically related types 
curred prior to all the other rearrange- of orientation heterogeneity to that found 
ments observed in the pea and broad bean in the chloroplast genome, we can make 
lineages. (4) Broad bean chloroplast DNA two major predictions about the mech- 
probably has evolved from an alfalfa-like anism of chloroplast DNA isomerization 
ancestral genome by two specific inver- (Palmer, 1983; Palmer etal., 1983b). First, 
sions. Thus, all characterized major chlo- there is very likely a specific sequence, 
roplast DNA rearrangements (five so far) perhaps as little as 14 or 15 base pairs in 
are inversions. (5) There is a broad range size, located within the inverted repeat 
of rearrangement frequencies in those and which serves as the site for a con- 
genomes that lack the inverted repeat. servative homologous recombination 

Our conclusion that the inverted-re- event that inverts the two single-copy 

peat loss occurred before all the other regions. Second, inversion is probably 

rearrangements observed in pea and mediated by a specialized recombination 


system which acts independently of any et al., 1981), the frequency of these events 

generalized cellular recombination sys- will fall off dramatically as the distance 

tern. The site-specific recombination sys- between the two sites increases. Recom- 

tems reviewed in Palmer (1983) and bination between inversely oriented sites 

Palmer et al. (1983b) include specific re- will produce inversions without the loss 

combination genes closely linked to the of DNA. However, many of these events 

recombination site itself. Thus, one might will also be lethal, since on a probabilistic 

predict that the chloroplast genome, and basis most of the sites will lie within tran- 

probably the inverted repeat itself, en- scribed or otherwise functional regions, 

codes its own site-specific recombination Inversions between strong sites and those 

system. However, the possibility must weak sites that do not interrupt func- 

also be considered that, as in the case of tional domains may also be disfavored, 

many other originally chloroplast-en- These would generate direct repeats of 

coded chloroplast functions (Weeden, a portion of the inverted repeat, allowing 

1981), the genes for this system may have subsequent excision of intervening se- 

been transferred to the nucleus. quences, and would also destroy one of 

Essentially all the site-specific recom- the two strong sites, abolishing the high- 

bination activity in the chloroplast is pos- frequency inversion system that is prob- 

tulated to occur between two strong ably necessary for chloroplast function 

recombination sites located in the in- when the inverted repeat is present (see 

verted repeat (Fig. 9). These are not only below). 

the two strongest recombination sites in There are two types of recombination 

sequence but are also embedded within events that will be stable and nonlethal 

extremely large (10,000-25,000 base pairs) and that appear to have been detected 

inverted repeats which will pair intra- by our comparative studies on chloro- 

molecularly at high frequency. In addi- plast genome organization. A certain 

tion, we postulate that there are several percentage of weak-site inversions will 

sites scattered throughout the genome not interrupt functional regions. Puta- 

which differ slightly in sequence from the tive examples of these in inverted-repeat 

two strong sites and which serve as sub- genomes are the spinach-mung bean and 

strates for the site-specific recombina- spinach-corn inversions of 50 and 20 

tion reaction at considerably reduced kilobase pairs, respectively (Fig. 9). 

frequencies (Fig. 9). In this case we make Among genomes that lack the inverted 

a direct analogy to the integration of repeat, broad bean appears to be derived 

lambda phage DNA into the E. coli chro- from alfalfa by two inversions, while a 

mosome. Deletion of the primary bac- smaller inversion (2-6 kb) has been 

terial attachment site for integrative mapped between two closely related pea 

recombination leads to a low frequency ecotypes (unpublished data), 

of lambda integration at specific second- There is a broad range of rearrange- 

ary sites possessing varying levels of ho- ment frequencies in those genomes that 

mology to the 15-base pair core lack the inverted repeat, so that in pea, 

recombination site (Nash, 1981). and to a lesser extent in broad bean, 

Chloroplast recombination will occur rearrangements have occurred signifi- 
much more rarely between one strong cantly more often than in genomes that 
and one weak site or between two weak contain the inverted repeat (Fig. 9). We 
sites. Most of these events are likely to postulate that in inverted-repeat-con- 
create lethal alterations in the chloro- taining genomes, essentially all the site- 
plast genome. Recombination between specific specialized recombination activ- 
directly repeated sites will lead to exci- ity is channeled into the two strong sites 
sion of the sequences between the sites; embedded within the inverted repeat, 
given the tight organization of the chlo- Upon loss of the inverted repeat this ac- 
roplast genome into coding regions (Oishi tivity is redistributed more evenly 



Fig. 9. Model for chloroplast genome evolution. Molecules b and b' represent the two orientation 
isomers (Palmer, 1983) of an inverted-repeat-containing chloroplast genome, such as that of mung bean, 
thought to be derived from a spinach-like ancestral genome (molecule a) by a major inversion within the 
large single-copy region (Palmer and Thompson, 1982). Molecule c represents the loss of one-half of the 
inverted repeat from a mung bean-like ancestor, with subsequent evolution yielding the present-day 
alfalfa (molecule d), broad bean (molecule e), and pea (molecule f) genomes. Closed circles represent 
strong recombination sites located within the inverted repeat. Open circles represent weak recombination 
sites scattered throughout the genome. Arrows indicate the relative orientation of recombination sites. 
For simplicity, only one orientation is shown for both the linear and circular configurations of molecule 

throughout the genome, with the result 
that the overall level of permitted recom- 
bination events — inversions and small 
deletions — is markedly increased. The 
decreased rate of rearrangements in al- 
falfa, and to a lesser extent in broad bean, 
is then explained by some combination 
of the stochastic loss of recombination 
sites and, perhaps more importantly, by 

a decrease or even complete loss of re- 
combination activity. Although the evo- 
lutionary conservation of recombination 
activity in inverted-repeat genomes in- 
dicates its maintenance (Palmer, 1983; 
Bohnert and Loffelhardt, 1982), loss of 
the inverted repeat may markedly re- 
duce these pressures. It is thus logical 
to suppose that a marked decrease in re- 



combination function may have taken 
place soon after the deletion event in a 
lineage leading to alfalfa and, later on, 
after two inversions, in the broad bean 

We plan to test this model both in 
higher-plant systems, where sequence 
studies at the ends of very recent inver- 
sions should yield valuable clues con- 
cerning the nature of possible 
recombination sites, and in the green alga 
C hlamydomonas reinhardtii, where 
powerful genetic approaches are avail- 
able that should allow the isolation of in- 
verted-repeat recombination sites and 
recombination genes. In addition, we are 
currently using pea and mung bean as 
model systems to explore the relation- 
ship between chloroplast DNA sequence 
rearrangements and the organization of 
the genome into discrete transcription 


Bohnert, H. J., and W. Loffelhardt, FEBS Lett. 
150, 403-406, 1982. 

Nash, H. A.,Annu. Rev. Genet. 15, 143-167, 1981. 

Oishi, K., T. Sumnicht, and K. K. Tewari, Biochem. 
20, 5710-5717, 1981. 

Palmer, J. D., Nucl. Acids Res. 10, 1593-1605, 

Palmer, J. D., Nature 301, 92-93, 1983. 

Palmer, J. D., and W. F. Thompson, Proc. Nat. 
Acad. Sci. USA 78, 5533-5537, 1981a. 

Palmer, J. D., and W. F. Thompson, Gene 15, 21- 
26, 1981b. 

Palmer, J. D., and W. F. Thompson, Cell 29, 537- 
550, 1982. 

Palmer, J. D., B. Osorio, J. C. Watson, H. Ed- 
wards, J. Dodd, and W. F. Thompson, Photo- 
synthesis, UCLA Symposia on Molecular and 
Cellular Biology, Alan R. Liss, Inc., New York, 
in press, 1983a. 

Palmer, J. D., K. J. Aldrich, and W. F. Thompson, 
submitted for publication, 1983b. 

Palmer, J. D., G. P. Singh, and D. T. N. Pillay, 
Mol. Gen. Genet. 190, 13-19, 1983c. 

Weeden, N. F., /. Mol. Evol. 17, 133-139, 1981. 



Terri L. Lomax and Winslow R. Briggs 

One of the major hurdles in under- 
standing the mechanism of auxin action 
is a knowledge of how the hormone is 
actually transported through the plant. 
While auxin is produced only at the apical 
tip of the plant, it is involved in the con- 
trol of cell division, growth, differentia- 
tion, and regeneration throughout the 
entire plant. Thus, the directed move- 
ment of auxin plays a major regulatory 
role in the development of plants. 

Auxin is transported through the in- 
tact plant in a polar, specific manner. That 
its movement is facilitated is demon- 
strated by the fact that the natural auxin, 
indole-3-acetic acid (IAA), has been found 
to move through a plant segment much 
more quickly (1-2 cm h" 1 ) than do other 
weak acids, which are not active auxins, 
with similar pK's. Auxin also moves at a 

steady rate through tissue over extended 
periods of time: its movement does not 
decelerate, as would be expected if the 
transport were by simple diffusion. When 
auxin is applied to the basal end of an 
inverted segment, it is not transported, 
thereby demonstrating the exclusively 
polar transport of the molecule. 

The current concept of auxin transport 
is that undissociated IAA is taken up into 
the cytoplasm when the pH of the cy- 
toplasm is higher than that of the walls. 
The IAA dissociates in the more-basic 
environment and accumulates as a result 
of the reduced permeability of the anionic 
form. The IAA~ is secreted at the lower 
end of each cell by a localized, selective 
anion carrier which provides the polar 
nature of the transport. This theory is 
known as the chemiosmotic theory of po- 



lar auxin transport (Rubery and Shel- 
drake, 1974). 

Many elegant in vivo experiments have 
led to this model for the cellular mech- 
anism for auxin transport, but until re- 
cently no in vitro system has existed with 
which to test its predictions. We recently 
presented evidence for such a system: 
Tightly sealed membrane vesicles, which 
are prepared from zucchini hypocotyls, 
can transport auxin in vitro using a pH 
gradient — inside basic, outside acidic — 
as the driving force (Year Book 81; Her- 
tel et at. , 1983). This ability to accumu- 
late [ 14 C]IAA was shown to be sensitive 
to the protonophore FCCP (carbonyl cy- 
anide p-trifluoromethoxyphenyl hydra- 
zone) as well as to osmotic shock, 
indicating that sealed vesicles are capa- 
ble of holding a pH gradient. By sepa- 
ration on a linear Dextran T-70 gradient, 
IAA uptake was demonstrated to be re- 
stricted to vesicles of plasma membrane 
origin even though other sealed vesicles 
were present in the preparation (Year 
Book 81, 18-21). 

It remained to be shown that this ac- 
cumulation of [ 14 C]IAA by sealed vesi- 
cles was facilitated by auxin influx and 
efflux via saturable, specific carriers, as 
predicted by Rubery and Sheldrake (1974) 
and Goldsmith (1982), rather than by the 
simple accumulation of a weak acid across 
a pH gradient. Here we present further 
evidence for the specific uptake of IAA 
by means of a saturable IAA " /H + sym- 
port for uptake and an anion efflux car- 

Materials and Methods 

Seeds of zucchini squash (Cucurbita 
pepo L. cv. Dark Green, Ferry Morse 
Seed Co., Mt. View, CA) and maize (Zea 
mays, cv. WF9 x Bear 38, Bear Hybrid 
Corn Co., Decatur, IL) were planted in 
moist vermiculite and grown in plastic 
boxes at 26°C and 95% relative humidity 
for five days in total darkness. Mem- 
brane particles were prepared from 2-m 
sections of hypocotyl (zucchini, cut 5 mm 
below the hook) or coleoptile (maize) and 

were tested following the procedures de- 
scribed in Hertel et al. (1983). Briefly, 
the vesicles were made by homogeniza- 
tion of the tissue in pH 7.9 medium (250 
mM sucrose, 50 or 200 mM Tris base, 3 
mM Na 4 EDTA, and 0.1 mM MgS0 4 , ti- 
trated with 12 N HC1), followed by a se- 
ries of centrifugations, after which the 
resulting microsomal pellet was resus- 
pended in medium buffered at pH 5.5 
(300 mM sucrose, 10 mM Na 3 Citrate, and 
0.1 mM MgS0 4 , titrated with 12 N HC1). 
This creates a pH gradient across any 
sealed membranes. 

To start a transport assay, the vesicles 
were diluted 1:1 with pH 5.5 buffer con- 
taining radioactive auxin and any addi- 
tions. The uptake was then stopped at 
any given time by addition of an aliquot 
(200 (xl) of vesicle suspension to a large 
volume (10 ml) of ice-cold buffer which 
was immediately filtered over a Millipore 
filter under vacuum before any signifi- 
cant efflux could occur, the entire process 
taking <8 s. Alternatively, a centrifu- 
gation assay was used where many sam- 
ples could be stopped simultaneously (see 
Hertel ebal., 1983). 

Results and Discussion 

When the vesicles were given a 5-min 
pulse of radioactive auxin at various times 
after resuspension (after establishment 
of the pH gradient), there was an initial 
rapid rate of accumulation which fell off 
and became stable after approximately 4 
min (Fig. 10). This initial instability may 
be the result of an initial rapid rate of 
disruption of vesicles, movement of 
[ 14 C]IAA in and out of vesicles which are 
not as pH-tight, or a flux of buffer ma- 
terials. Accordingly, all other experi- 
ments were started at least 5 min after 
resuspension, when a stable rate of up- 
take had been reached. 

In Fig. 11, both FCCP, an electrogenic 
protonophore, and nigericin, an electro- 
neutral H + /K + exchanger, are seen to 
induce very similar effects on the uptake 
of IAA, with both having a similar con- 
centration dependence, becoming satu- 




5 min pulse/chase 


44 pmoles/somple "C-IAA 


E H 



\ D 
a N^ 

D > S^__ — 



^^"~ " — — -^ D 


1 1 1 


1 1 1 1 1 ! 1 

4 6 8 10 12 14 16 If 

Time after resuspension, in minutes 

20 22 

Fig. 10. Uptake of a 5-min pulse of [ 14 C]IAA 
given at various times after resuspension of zuc- 
chini membrane vesicles in pH 5.5 buffer. Each 
sample contained 44 pmoles [ 14 C]IAA (2 x 10 7 
M) and 2.5 g fresh wt equivalents ml" 1 of mem- 
branes, prepared at pH 7.9. The filtration assay 
was used. 

rated at about 10 fiM, the level used in 
all subsequent experiments. The fact that 
the effects of FCCP and nigericin are 
similar even though they work by quite 
different mechanisms suggests that the 
reduction in IAA uptake is attributable 
to the dissipation of the proton motive 
force, not to an artifact specific to one 
ionophore or the other. 

-7 -6 -5 -4 

log concentration ionophore 

Fig. 11. The inhibition of [ 14 C]IAA association 
by increasing concentrations of FCCP (circles) or 
nigericin (diamonds). Each sample contained 2 x 
10~ 7 M [ 14 C]IAA, 3 x 10" 6 M TIBA, and 0.8 g 
fresh wt equivalent ml 1 membranes prepared from 
zucchini hypocotyls. Measured by centrifugation 
assay (30 min x 50,000#) after 40-min incubation 
on ice. 

We had previously stated (Year Book 
81, 18-21) that although we were able to 
demonstrate in vitro auxin transport by 
vesicles from several members of the Cu- 
curbitaceae, we were unable to show pH- 
driven accumulation by membrane prep- 
arations from maize coleoptile tissue. We 
have since been able to resolve this dis- 
crepancy by demonstrating FCCP-sen- 
sitive [ 14 C]IAA uptake not only by maize 
coleoptile vesicles but also, to a lesser 
extent, by vesicles prepared from pea 
epicotyl tissue (data not shown). These 
results were achieved by optimizing the 
buffer conditions and increasing the size 
of the pH gradient. The preparations from 
maize and pea may contain a smaller 
number of sealed vesicles, the vesicles 
may be sealed less tightly, or both. How- 
ever, the fact that they exhibit a pH- 
dependent accumulation of IAA demon- 
strates that this is not a phenomenon re- 
stricted merely to the Cucurbit aceae. 

Benzoic acid (BA), like IAA, is a li- 
pophilic weak acid which has a pK and 
structure similar to that of IAA but which 
is inactive as an auxin. In studies with 
intact plant segments, benzoic acid has 
been shown to be taken up to an even 
greater extent than IAA, but it does not 
inhibit the polar movement of [ 3 H]IAA 
through the tissue, as does IAA (Gold- 
smith, 1982). This result is consistent with 
the specificity of IAA transport in vivo 
and supports the hypothesis that trans- 
port consists of a saturable flux of auxin 
anions in parallel with a nonsaturable flux 
of undissociated IAA. In a similar ex- 
periment with membrane vesicles from 
zucchini hypocotyls we find that the up- 
take of [ 14 C]IAA is indeed saturable by 
low concentrations of unlabeled IAA but 
is not inhibited by excess benzoic acid 
until quite high concentrations are reached 
(Fig. 12). The half-maximal level of IAA 
saturation of [ 14 C]IAA accumulation by 
these vesicles appears to be approxi- 
mately 2-4 x 10 " 7 M. In comparison, 
the uptake of [ 14 C] benzoic acid is inhib- 
ited by neither IAA nor BA until milli- 
molar concentrations are present, 
consistent with competition for simple 





£ 1600 


C 1400 





800 h 

"C-BA + IAA or BA '\£\. 
M C-IAA + IAA XI \ \ 


O 20 

15 - 

q io 

c 5 




o o 




'/' 1 *\ « ""♦ « 

7 ™\. 

— / / \"C-Benzoic acid 


/ "^^^^ 

— / 50 pmoles/sample "C— IAA 


85 pmoles/sample "C-benzoic acid 

1 1 1 1 1 1 




-en -7 -6 -5 -4 

log concentration IAA or benzoic acid (BA) 

Fig. 12. Dose-response curve for the saturation 
of [ 14 C]IAA or [ 14 C]BA uptake by either IAA or 
BA. Each 1-ml sample of the centrifugation assay 
contained 0.8 g fresh wt equivalents of zucchini 
membrane vesicles, 3 x 10 ~ ( ' M TIB A, and either 
1.25 x 10 7 M[ 14 C]IAAor2.2 x 10 7 M[ 14 C]BA. 
Open circles = [ 14 C]IAA + IAA; closed circles = 
[ 14 C]IAA + BA; open diamonds = [ 14 C]BA + IAA; 
closed diamonds = [ 14 C]BA + BA. 

diffusion of a weak acid across a pH gra- 
dient. This result provides further evi- 
dence that, while accumulation by 
diffusion is certainly a component, IAA 
accumulation in these vesicles is via a 
specific, saturable mechanism. 

Figure 13 presents a time course for 
the FCCP-sensitive accumulation of both 
[ 14 C]IAA and [ 14 C]BA by zucchini hy- 
pocotyl membrane vesicles. Although the 
initial uptake kinetics appear to be sim- 
ilar, the [ 14 C]IAA is retained far longer 
at a high level than is the [ 14 C]BA. There 
are specific inhibitors of auxin transport 
such as TIBA (2,3,5-triiodobenzoic acid) 
and NPA (naphthylphthalamic acid) which 
block its movement through the tissue 
by blocking the efflux carrier (Goldsmith, 
1982). While TIBA is present in both 
treatments here, it would appear to block 
only the efflux of IAA, not that of BA. 
This specificity is consistent with the ob- 
servations of Sussman and Goldsmith 
(1981) that the transport inhibitor NPA 
stimulates the retention of IAA by intact 

20 40 60 80 100 120 HO 160 

Time in minutes 

Fig. 13. Time course for the uptake of both 
[ 14 C1IAA and [ 14 C]BA by zucchini membrane ves- 
icles in the presence of 3 x 10 6 M TIBA. Each 
sample for the filtration assay contained 2.5 g fresh 
wt equivalents ml l membranes and either 50 pmoles 
[ 14 C]IAA or 85 pmoles [ 14 C]BA and were done in 
duplicate in the presence and absence of FCCP (2 
x 10 " 5 M). The data are given as the percentage 
of total dpm sample 1 represented by the FCCP- 
sensitive uptake. All additions were made at time 
= 0, 5 min after the vesicles, which had been pre- 
pared at pH 7.9, were resuspended in pH 5.5 buffer. 

cells but has no effect on the accumula- 
tion of benzoic acid. Since NPA and TIBA 
inhibit only the saturable efflux compo- 
nent of auxin transport (Goldsmith, 1982), 
the increased retention of auxin by the 
vesicles in the presence of TIBA is pre- 
sumably due to the anion carrier being 
blocked. Thus, the anion carrier in these 
vesicles is not saturable at low concen- 
trations by benzoic acid and is specific 
for the facilitated transport of IAA. The 
following article presents additional ev- 
idence for the facilitated nature of the 


Goldsmith, M. H. M., Planta 155, 68-75, 1982. 
Hertel, R., T. L. Lomax, and W. R. Briggs, Planta 

157, 193-201, 1983. 
Rubery, P. H., and A. R. Sheldrake, Planta 118, 

101-121, 1974. 
Sussman, M. R., and M. H. M. Goldsmith, Planta 

151, 15-21, 1981. 





Ten~i L. Lomax, Rolf J. Mehlhorn* and Winslow R. Briggs 

Having established that auxin uptake vesicles prepared as just described. Even 

by zucchini hypocotyl membrane prepa- in the presence of the protonophore 

rations is the result of a specific, satur- FCCP, there is an initial small incorpo- 

able, and pH-dependent transport of auxin ration of auxin, indicating that FCCP acts 

into sealed plasma membrane vesicles (see slowly. However, after 40 min, a low, 

Lomax and Briggs, this Report), we equilibrium-level association of auxin in 

wished to dissect the various components the presence of FCCP is reached. The 

of this transport in order to elucidate its presence of the auxin transport inhibitor, 

mechanism. TIBA, results in a higher level of accu- 

The transport system is postulated to mulation than that seen when neither 
operate via an uptake IAA-H + symport FCCP nor TIBA is present. This result 
and an anion efflux carrier, as well as by indicates that TIBA is blocking the efflux 
diffusion — all of which are dependent upon site, as it is known to do in intact tissue 
the size and direction of the pH gradient. (Goldsmith, 1977), thus increasing the 
Intact plant cells have previously been amount of accumulation of IAA within 
demonstrated to have a pH gradient the vesicles. The level of IAA association 
across the plasma membrane, with the with the vesicles both in the presence 
cytoplasm being basic or neutral (pH 7) and absence of TIBA declines after about 
and the wall space acidic (pH 5-6). The 15 min (-TIBA) or 30 min ( + TIBA), 
zucchini membrane vesicle system de- presumably because of the decay of the 
scribed here has a pH gradient of the pH gradient. The decline approximates 
same direction established by preparing an exponential decay, as might be ex- 
the vesicles at pH 7.9 and testing them pected to result from a linear decay of 
with pH 5.5 buffer on the outside (Hertel the pH gradient. The addition of FCCP 
et al., 1983). Since these vesicles have after 1 h (Fig. 14, open symbols) to ves- 
been shown to perform all of the essential icles not previously exposed to the pro- 
functions of auxin transport according to tonophore resulted in a more-rapid release 
the chemiosmotic theory of auxin trans- of [ 14 C]IAA. Additional FCCP did not, 
port (Hertel et al., 1983), they provide a however, change the level of IAA asso- 
model system for approximating the con- ciation for vesicles that had been main- 
tributions of the various transport com- tained in FCCP. Thus, these data indicate 
ponents. However, although the zucchini that a pH gradient does indeed still exist 
vesicles have been demonstrated to after one hour and that the accumulation 
maintain a pH gradient for an extended of auxin is maintained by that pH gra- 
period of time, they are most likely not dient. 

as tightly sealed as an intact plant cell The possibility existed that the decline 

is. after 15-30 min in the association of 

Figure 14 illustrates a typical time [ 14 C]IAA with the vesicles was a result 

course for the accumulation of radioac- of gradual deterioration of their sealed 

lively-labeled IAA by zucchini hypocotyl nature over time. However, when the 

reaction was started by adding [ 14 C]IAA 

, T n . . T , , -. to vesicles after 2 h, they were found to 

* Lawrence Berkeley Laboratory and Depart- , , .„ , , ~ 7 , , . T A A ..-. 

ment of Physiology and Anatomy, University of be stl11 capable of accumulating IAA With 

California, Berkeley. kinetics similar to those started at min 



E 14 

V 12 

% 10 

"5 8 

c 6 
5 ^ 


a 2 

I oh 




♦//\ + 10" 5 M FCCP 

: i \ X 

// \-fccp ••■•-, .; 


//a \ \ \ 
^ ^—, \ >■- ♦ 


- * +FCCP at minX^_ -..afi^^^ 

— • 

i i i i i l. 

20 40 

100 120 

Time in minutes 

Fig. 14. Time course of [ 14 C]IAA uptake in the 
presence and absence of TIBA and FCCP. Each 
filtration assay sample contained 2 g fresh wt equiv- 
alents ml" 1 zucchini membrane vesicles prepared 
in pH 7.9 buffer (250 mM sucrose, 50 mM Tris base, 
3 mM EDTA, and 0.1 mM MgS0 4 , titrated with 
12 N HC1), and resuspended in pH 5.5 buffer (300 
mM sucrose, 10 mM Na-jCitrate, and 0.1 mM 
MgS0 4 ). At min, 2 x 10 " 7 M [ 14 C]IAA (6500 
dpm sample" 1 ) was added along with either 10 5 
M FCCP (circles), 3 x 10 " 6 M TIBA (diamonds), 
or no additions (triangles). Additional FCCP (10 " 5 
M) was added to portions of each treatment at 1 h 
(open symbols). 

(Fig. 15). Another possibility was that 
the decline in IAA accumulation was due 
to the metabolism or degradation of 
[ 14 C]IAA. When additional [ 14 C]IAA was 
added after 3 h in the experiment shown 
in Fig. 15, the same fraction of total dpm 
accumulated was reached as in those 
without the addition, indicating that the 


12 - 
10 - 

% 2 


- J *•♦. +TIBA 

■ • 


V / 

s ^ 

- ; \« 

- ♦ *\+FCCP 



i i 






80 120 160 200 240 

Time in minutes 

Fig. 15. Effect of late addition of [ 14 C]IAA. 
Conditions as for Fig. 14 except that the pH 7.9 
buffer contained 200 mM Tris rather than 50 mM. 
One portion of the vesicle resuspension was incu- 
bated on ice along with the other treaments but 
without additions until 2 h, then started as for 
min (squares). 

vesicles were accumulating the auxin at 
their maximum capability at that pH gra- 
dient (data not shown). Thus, degrada- 
tion of IAA is not likely a significant 
contributor to these kinetics. The in- 
creased level of IAA accumulation and 
the longer retention time in Fig. 15 over 
those shown in Fig. 14 occurred because 
the vesicles used for the assay in Fig. 15 
were made with 200 mM Tris buffer rather 
than with 50 mM, as for Fig. 14 (see be- 

To calculate accurately the energetics 
and kinetics of auxin transport, and, most 
importantly, to demonstrate that the ac- 
cumulation in these vesicles is a facili- 
tated transport and not merely the result 
of diffusion across the pH gradient, re- 
liable measurements of both vesicle vol- 
ume and the size of the pH gradient were 
needed to correspond with the measure- 
ments of auxin uptake under various con- 
ditions. This has been accomplished using 
electron spin resonance (ESR). Nitrox- 
ide spin labels are used to measure a va- 
riety of biological properties, including 
cell volumes, pH and electrical gra- 
dients, and membrane and surface po- 
tentials (Mehlhorn et al, 1982). The 
method is especially advantageous be- 
cause of the ease of separating the ESR 
spectra of membrane-bound from aqueous 
probes, the availability of many different 
nitroxide compounds which can be ap- 
plied to different problems or conditions, 
and the very small amount of material 
required. (Determinations can be made 
with only 40 (jlI of a suspension containing 
less than 1 mg ml" 1 protein.) 

The volume of sealed vesicles within a 
solution can be determined by quanti- 
tating the aqueous line heights of a mem- 
brane-permeable, yet highly water- 
soluble, nitroxide probe, Tempone, in the 
presence and absence of an impermeable 
paramagnetic quenching agent, 
Na 2 MnEDTA. The ratio of quenched to 
unquenched signal is directly related to 
the volume, which is sealed and unavail- 
able to the quencher. Using zucchini 
membrane preparations identical to those 
used for IAA accumulation assays, we 



have demonstrated that increasing 
amounts of osmoticum in the external 
medium will reduce the volume of the 
vesicles. Conversely, decreasing the ex- 
ternal osmoticum causes an increase in 
vesicle volume (data not shown). These 
experiments demonstrate that the vesi- 
cles are capable of expanding and con- 
tracting in response to external osmotic 
levels and provide further evidence as to 
their sealed nature. 

Other nitroxide probes are either weak 
acids (Tempacid) or weak bases (Temp- 
amine), which partition preferentially into 
basic or acidic environments, respec- 
tively. The concentration ratios of the 
amine and weak acid on both sides of the 
membrane can be measured and used to 
calculate the size of the pH gradient. Us- 
ing only one probe, the "effective" vol- 
ume can be calculated and related to the 
actual volume determined with Tem- 
pone, and this can again be related to the 
ApH. Using these methods we have de- 
termined both the vesicle volume and the 
size of the pH gradient in zucchini vesi- 
cles made with either 50 mM or 200 mM 
Tris buffer at pH 7.9. All vesicle prep- 
arations were resuspended in pH 5.5 
buffer, exactly as for the auxin uptake 
assays described previously. 

Figure 16 shows that the decay of the 
pH gradient over time is slow (approxi- 

Buffer = 10 mM Na/citrate 
(350 mOsm) 

200 mM Tris, 

(600 mOsm) 

50 mM Tris, 
(350 mOsm) 

10 20 30 40 50 60 

Time after resuspension, in minutes 

Fig. 16. Time course of ApH in vesicles pre- 
pared as in Figs. 14 and 15. Calculated from aqueous 
line heights of nitroxide spin probes measured by 
ESR. Samples contained 1 mM Tempone or Tem- 
pacid in the presence or absence of 0.15 M 
Na 2 MnEDTA and included 2 x 10 ~ 7 M IAA and 
3 x 10 " 6 M TIB A as for an auxin uptake assay. 

mately 0.03 pH units min" 1 ), and is sim- 
ilar under both of these conditions. The 
size of the pH gradient in both cases was 
about 0.8 pH units directly after resus- 
pension, smaller than expected. This is 
caused in part by the elevated pH of the 
surrounding medium. Although the ves- 
icles were resuspended in pH 5.5 medium 
orginally, some breakage or leakage ev- 
idently took place during resuspension. 
Thus, the internal pH 7.9 buffer of higher 
strength (50 or 200 mM vs. 10 mM) was 
released from broken vesicles, raising the 
external pH to 5.73 for those vesicles 
prepared with 50 mM buffer and 6.09 for 
those prepared with 200 mM buffer. 
Nevertheless, the magnitude of the pH 
gradient observed here closely approxi- 
mates that found in the intact plant. 

The ESR measurements providing the 
data depicted in Fig. 16 were made at 
4°C, the same temperature used for the 
auxin uptake assays. When the assay 
temperature was raised to 25°C the rate 
of decline of the pH gradient increased 
markedly (data not shown). This result 
suggests that the vesicles become less- 
tightly sealed at higher temperatures and 
is in agreement with Hertel et al. (1983) 
who found that increased temperature 
decreased auxin uptake by zucchini 
membrane vesicles. 

From our calculations, we found that 
the sealed vesicles made with 50 mM 
buffer were about 0.7% of the total vol- 
ume of the membrane resuspension, or 8 
|jd mg" 1 protein. Those prepared with 
200 mM buffer were half again as large, 
1% of the total volume, or 12 fxl mg _1 
protein. This difference is most likely be- 
cause, while the 50 mM buffer is in os- 
motic balance with the external buffer, 
the osmotic concentration of the 200 mM 
buffer is higher than that of the external 
buffer (600 mOsm vs. 350 mOsm), which 
causes the vesicles to swell. This differ- 
ence in vesicle size when loaded with dif- 
ferent buffer strengths may explain the 
differences in total IAA accumulation be- 
tween Figs. 14 and 15. The difference is 
illustrated more clearly in Fig. 17, where 
both conditions are compared in a single 



- 3 

200 mM Tris, (600 mOsm) 

50 mM Tris, (350 mOsm) 

100 200 

Time in minutes 



Fig. 17. Kinetics of FCCP-sensitive IAA ac- 
cumulation at 50 mM and 200 mM internal buffer 
strength. Conditions as described in Figs. 14 and 

experiment. The maximal FCCP-sensi- 
tive IAA accumulation with vesicles 
loaded with 200 mM buffer is approxi- 
mately twice that of those containing 50 
mM buffer. The retention time is also 
longer, at least 6 h rather than 3-4 h. 

The determinations of vesicle volume 
allow the calculation of the actual IAA 
concentration gradient generated by the 
sealed vesicles. With either internal buffer 
concentration, the maximal [IAA]; is 5- 
10 (jlM. Thus, the differences observed 
in Fig. 17 are a consequence of variable 
vesicle volume and do not represent a 
difference in concentration of IAA ac- 
cumulated. Since only 0.1-0.2 fxM IAA 
was present in the external solution, this 
concentration represents at least a fifty- 
fold accumulation of the auxin. The ac- 
tual figure is most likely much higher be- 
cause we used the total sealed volume 

for these calculations and it has been 
shown that only one population of the 
sealed vesicles actually transports IAA 
(Year Book 81, 18-21). The maximal ac- 
cumulation of IAA at diffusional equilib- 
rium with a pH gradient of the magnitude 
measured here would be only five times 
the external concentration of IAA (Gold- 
smith, 1977), rather than the fifty fold ac- 
cumulation obtained here. This difference 
is convincing evidence that there is in- 
deed facilitated transport of IAA by these 
zucchini membrane vesicles and that the 
transport can enhance the concentration 
of IAA at least tenfold over the equilib- 
rium distribution. These large accumu- 
lations are still within the physiological 
concentration range for IAA. 

We have now demonstrated that we 
have an in vitro system with which to 
study auxin transport at pH gradients 
and IAA concentrations approximating 
conditions found in the intact plant, in- 
cluding all of the components of in vivo 
transport. Using ESR data for volumes 
and ApH in combination with [ 14 C]IAA 
uptake and efflux kinetics, we will now 
be able to analyze in detail the cellular 
mechanism of auxin transport. 


Goldsmith, M. H. M., Annu. Rev. Plant Physiol. 

28, 439-478, 1977. 
Hertel, R., T. L. Lomax, and W. R. Briggs, Planta 

157, 193-201, 1983. 
Mehlhorn, R. J., P. Candau, and L. Packer, Meth. 

Enzymol. 88, 751-761, 1982. 



James R. Shinkle and Winslow R. Briggs 

Last year (Year Book 81, 39-43), we 
reported preliminary findings on the sen- 
sitivity of subapical oat coleoptile sec- 
tions, excised from dark-grown seedlings, 
to red light-induced increases in growth. 
We found that adding IAA to the incu- 
bation buffer before irradiation in- 

creased the apparent sensitivity of the 
sections to red light by a factor of at least 
1000. We have confirmed this finding, and 
have discovered that the nature and 
magnitude of the change in sensitivity to 
red light depends on the concentration 
of IAA applied. For example, certain 



concentrations of I AA cause the sections 
to respond to red light in a manner very 
similar to the response of coleoptiles left 
intact on whole seedlings. This response 
is biphasic, consisting of a very low flu- 
ence response (VLF) and a low fluence 
response (LF), separated by a plateau 
(see Mandoli and Briggs, 1981). In sec- 
tions, both responses appear as step 
functions rather than continuous gradual 
increases in growth. 

Fluence-response curves for red light- 
induced increases in growth were per- 
formed on subapical coleoptile sections 
and coleoptiles on intact seedlings. Seed- 
lings were grown in total darkness for 72 
h after imbibition, as in Mandoli and 
Briggs (1981) and Year Book 78, 140- 
144. All manipulations other than actual 
irradiations on whole seedlings and sec- 
tions were carried out in complete dark- 

For whole seedlings, plates of approx- 
imately 25 seedlings were placed, one at 
a time, in a foil-lined coffee can attached 
to a light source such that only diffuse 
light reached the seedlings. Seedlings 
were irradiated in an upright position. 
Irradiation with red light took place over 
1-5000 seconds, depending on the flu- 
ence applied. All fluences of 1 nmol/cm 2 
or less were given in a one-second irra- 
diation. Each fluence was given to du- 
plicate plates. Plants were returned to a 
dark cabinet after irradiation. Dark con- 
trol plants were moved to and from the 
coffee can without being irradiated. Im- 
mediately after other plants were irra- 
diated, two plates of seedlings were 
removed and measured for initial coleop- 
tile length. At 24 h after irradiation, plants 
were removed from cabinets, placed on 
clear glass plates, and photocopied. The 
facsimile plants were measured for co- 
leoptile length using a computer-assisted 
digitizer, as described in Mandoli and 
Briggs (1981). Growth from the initial 
length was determined, and the red light 
effect on growth was expressed as per- 
centage of maximal red light-induced in- 
crease in length between initial and 
treated plants. 

Fluence-response curves for sections 
were obtained for 5.7-mm coleoptile sec- 
tions cut from 3 mm below the tip of seed- 
lings 72 h after imbibition. Sections were 
harvested into petri dishes with 9-12 ml 
of 5 mM potassium phosphate buffer, pH 
5.9, with 5% sucrose and 80 (jlM chlor- 
amphenicol. IAA, if present, was also in 
solution at the time of harvest. Each har- 
vest yielded 18-30 sections per treat- 
ment dish, and 5-6 dishes were obtained 
from each harvest. Two harvests in se- 
ries were performed for each experi- 
ment. Each harvest and subsequent 
irradiations took one h or less. Dishes 
were irradiated with the same source used 
for whole plants, but irradiation was uni- 
lateral from above. Exposure times were 
similar to those for whole plants. After 
irradiation, dishes were placed on a gy- 
rotory shaker operating at 100 rpm in a 
totally dark cabinet. At 12 h after irra- 
diation, dishes were removed from the 
shaker, and the sections were blotted dry, 
placed on clear glass plates, and photo- 
copied. Facsimile sections were digitized 
for length as for coleoptiles. Red light- 
induced increase in growth was deter- 
mined and expressed as the percentage 
of maximal increment of increase induced 
by red light. 

The red and far red light was obtained 
from an incandescent projector lamp, as 
described by Mandoli and Briggs (1981). 

As seen in Fig. 18, the addition of 6 x 
10 ~ 6 M IAA to the incubation medium 
decreases the fluence of red light re- 
quired for both threshold and saturation 
of the R-induced increase in section 
growth by a factor of 10,000. In the whole- 
plant photobiology described by Mandoli 
and Briggs (1981), the two curves would 
be equivalent to an exclusively LF re- 
sponse (minus IAA) and an exclusively 
VLF response (plus IAA), respectively. 
Consistent with this interpretation, the 
R-induced increase in growth is far red 
reversible in the absence of added IAA, 
but is not FR reversible with IAA added. 
Also, the growth response is inducible 
by far red light alone in the presence of 
added IAA (data not shown). 







8 50 

7 25 


6 X 10~ 6 M IAA ' 

oo -6 -5 -4-3-2-10 1 2 3 
Log fluence (nrnol cm -2 ), Red 

100 - 





oo -6 -5 -4 -3-2-10 1 2 
Log fluence (nmol cm -2 ), Red 

Fig. 18. Fluence-response curves for red light- 
induced increase in growth of oat coleoptile sec- 
tions harvested at 72 h, plus and minus 6 x 10 6 
M IAA. Sections were floated on 5 mM K-phos- 
phate buffer, 5% sucrose, 80 jxM chloramphenicol, 
pH 5.9, for 12 h. Response is expressed as per- 
centage of maximum response. Error bars indicate 
average standard error for three replicate exper- 

Fig. 19. Comparison of fluence-response curves 
for red light-induced increase in growth for sec- 
tions as in Fig. 18 treated with 6 x 10 " 7 M IAA 
(solid line, circles), and for coleoptile growth in 
whole plants measured 24 h after irradiation (dashed 
line, diamonds). Response is expressed as % of 
maximal response. Error bars represent average 
standard error for three replicate experiments. 

The R-induced increase in coleoptile 
growth in whole seedlings irradiated 72 
h after imbibition shows a biphasic flu- 
ence dependency, very similar to that 
found for seedlings irradiated at 56 h af- 
ter imbibition, as in Mandoli and Briggs 
(1981). The biphasic fluence-response be- 
havior is shown in Fig. 19. 

Coleoptile sections cut from 72-h-old 
seedlings can be made to show a biphasic 
fluence-response curve for R-stimulated 
growth by the application of 6 x 10 _7 M 
IAA to the incubation medium. Figure 
19 shows this treatment in comparison 
to the whole-tissue response. This lower 
applied IAA concentration is probably 
closer to the internal IAA level in whole 
tissue (see Bandurski and Schulze, 1974). 
The threshold for R-stimulated growth 
appears to be an order of magnitude 
higher for sections than for whole tissue. 
Otherwise, the ranges and magnitudes 
of the two responses to red light are quite 

Mandoli and Briggs explained both the 
LF and VLF red light-induced increases 
in coleoptile growth in terms of phyto- 
chrome action. The data presented here 

give no contrary indication. There are 
two current models to account for the 
separation of phytochrome responses into 
two phases. VanDerWoude (1983) has 
postulated that the two phases are the 
result of different binding affinities of a 
postulated phytochrome receptor/trans- 
ducer for dimers of phytochrome which 
have either one or both molecules con- 
verted to P fr . BlaauwJansen (1983) ac- 
counts for the two phases in terms of the 
attainment of a plateau level of P fr by 
phytochrome destruction; further in- 
creases in P fr are possible at fluences pro- 
ducing more P fr than the destruction 
process reduces to the plateau level. It 
may now be possible to distinguish be- 
tween these two models using the ap- 
plied auxin to switch between the two 

VanDerWoude (1983) and Fredericq et 
al. (1983) have both shown that chemical 
treatment of seeds can lead to increased 
sensitivity to red light, similar to the shift 
found in Fig. 18. VanDerWoude also re- 
ported that prechilling of lettuce seeds 
could lead to a biphasic fluence-response 
curve for increased germination. Fred- 



ericq et al. reported that the plant hor- 
mone gibberellin GA 3 was active in 
sensitizing the germination of seeds of 
Kalanchoe both to red light and far red 
light. These two reports indicate that al- 
terations in phytochrome response by 
external factors may be common, but seed 
germination experiments do not allow for 
the precise changes in external condi- 
tions possible in the excised coleoptile 


Bandurski, R. S., and A. Schulze, Plant Physiol. 
5.4, 257-262, 1974. 

Blaauw-Jansen, G., Plant Cell Environ. 6, 173- 
179, 1983. 

Fredericq, H., R. Rethy, H. van Onckelen, and 
J. A. de Greef, Physiol. Plant. 57, 402-406, 1983. 

Mandoli, D. F., and W. R. Briggs, Plant Physiol. 
67, 733-739, 1981. 

VanDerWoude, W. J., in Strategies of Plant Re- 
production, 234-244, W. J. Meudt, ed., Belts- 
villeSymp. Agric. Res. Series 6, Allanheld, Osmun 
and Co., New Jersey, 1983. 


Dina F . Mandoli, John S. Boyer* and Winslow R. Briggs 

The light-carrying capacity of plant 
tissues is affected by anatomical struc- 
tures, tissue damage (Mandoli and Briggs, 
1982a), and the age of the cells (Mandoli 
and Briggs, 1982b). Since infiltration with 
distilled water causes a 60-80% increase 
in the light-piping capacity of oat meso- 
cotyl tissue sections (Mandoli and Briggs, 
1983), we attempted to discern if water 
loss in these tissues caused a concomitant 
decrease in the light-guiding capacity 
(Year Book 81, 24-25). 

As plant tissue sections become de- 
hydrated (evidenced by a loss in fresh 
weight), they lose the ability to guide 
light axially by total internal reflection 
(Table 2). For example, a 25% loss in the 
fresh weight of oat mesocotyl sections 
results in a 35% loss in their ability to 
guide light (Table 2). From these prelim- 
inary data, the light-guiding ability ap- 
pears to decrease exponentially as fresh 
weight decreases. 

The optimum angle for acceptance of 
a light beam applied to the side of a tissue 
section is species-specific rather than tis- 
sue-specific and is therefore largely in- 
dependent of tissue geometry (Mandoli 
and Briggs, 1982b). The acceptance an- 
gle for oat mesocotyl and coleoptile tissue 

*Botany Department, University of Illinois, Ur- 

is 59°; for corn coleoptile and root it is 
52-54°; and for mung bean hypotocyl it 
is 47° (Mandoli and Briggs, 1982b). Soy- 
bean (Glycine max) hypocotyls, like corn 
roots and coleoptiles, have a fairly broad 
acceptance angle with a peak centered at 
55° with respect to normal (Fig. 20). When 
fully hydrated (100% water) there is a 
shoulder at 25° in the curve describing 
the angular dependence for soybean (Fig. 
20). To date, the presence of a secondary 
peak in acceptance angle is unique to soy- 
bean. As the tissue becomes dehydrated, 

TABLE 2. Oat Mesocotyl Light Guiding as a 
Function of Tissue Dehydration* 

% Initial 
Light Guided 

% Initial 
Fresh Weight 




*Each value represents five sections each 25 mm 
in length. 100% fresh weight = 13.67 mg/25 mm 
section of mesocotyl. 100% light guiding 9,603 ar- 
bitrary photometer units or about 2.5 mW of mono- 
chromatic (632.8) red light from an He-Ne laser. 
Tissues were cut from 3.5-day-old etiolated Lodi 
oats, which were grown according to Mandoli and 
Briggs (1981). Light guiding of tissue sections was 
measured as in Mandoli and Briggs (1982b) with 
the laser input applied perpendicularly to the cut 
surface of the tissue sections. 
















20 40 60 80 

Incident angle, degrees from normal 

Fig. 20. Acceptance angle in intact Glycine max 
seedlings plotted at three tissue water contents. 
Light emerging from the cut end of a tissue seg- 
ment was measured with a photomultiplier (Man- 
doli and Briggs, 1982b). The maximum light output 
for a given tissue segment was designated as 100%, 
and that at an angle of incidence of zero as 0%. Ten 
seedlings were measured at each angle for all de- 
grees of hydration and the mean and percent stan- 
dard deviation were calculated for each angle for 
each hydration state. Fully hydrated tissues showed 
a peak of light-guiding capacity at 55 degrees. Overall 
light-guiding capacity of tissues dehydrated to 63% 
and 39% of their initial fresh weight decreased in 
their light-guiding capacity to 23% (maximum shifted 
to 63 degrees) and 3.9% (maximum shifted to 65 
degrees) of the fully hydrated control. Curves have 
been adjusted vertically for clarity. 

the major peak of this curve changes by 
10-12° so that light impinging on the side 
of the tissue at more-oblique angles is 
accepted more readily than before. This 
change is effected during loss of the first 
37% of fresh weight (Fig. 20). The cause 
of the species specificity of tissue accep- 
tance angle and the change in the peak 
acceptance angle in soybean as tissue 
water is lost or gained is unknown. 

The sensitivity of the fiber-optic prop- 
erties of plants to changes in water sta- 
tus (Table 2) may provide a means to 
measure water movements through plant 
tissues. Our initial attempts to discern 
the path of water through etiolated soy- 
bean hypocotyls and to correlate these 
movements with the light-guiding capac- 
ity of the cells was complicated by the 
apparent existence of two possible routes 
for water movement during tissue re- 
hydration. Tissues rehydrated via the 
cuticle (ends occluded) regain their initial 
fresh weight roughly ten times faster than 
those allowed to rehydrate via the ends 
alone (sides occluded) (Table 3). The time 
required to rehydrate correlates roughly 
with the relative surface area of the ex- 
posed region. 

The extent of tissue water recovery 
was clearly a function both of the degree 
of dehydration attained and the path of 
rehydration. Tissues dehydrated to 60% 
of their initial fresh weight recovered fully 
if allowed to rehydrate via the cuticle 
whereas even a decrease to 80% of initial 
fresh weight could not be fully regained 


Path of Rehydration 

in Soybeans 


Ends Occluded 

Cuticle Occluded 

Dehydrated to Time 
(approx. %) (min) 

Recovered to 



Recovered to 


80 20-40 
60 45 
40 150-300 








*From three to four 10-cm etiolated hypocotyl sections were allowed to de- 
hydrate to 80, 60, or 40% of their initial fresh weight. The ends or sides of these 
sections were then coated with vaseline and a layer of Saran Wrap and were 
immersed in distilled water for two days. The increase in fresh weight over this 
time was monitored. 


over 48 h if rehydration occurred through known. However, if these changes have 
the tissue ends (Table 3). minor effects, light guiding may provide 
In summary, the endogenous light- a means to study the kinetics of hydra- 
guiding ability of plant tissues may prove tion and dehydration remotely and con- 
useful as a quantitative measure of water tinuously . 
movement and/or water status within the 
tissue. The promise of this concept de- 
pends on the uniformity of change in tis- p 
sue light guiding as water is lost. The References 

effects of possible local changes in hy- Mandoli, D. F., andW. R. Briggs, Plant Cell En- 

dration, such as more rapid dehydration viron. 5, 137-146, 1982a. 

of tissue ends than of the center or al- "^ F 7Q ' s ^'JtP I ^l Proc - Nat Acad ' 

,. . _ .. . . Set. UbA 79, 2902-2906, 1982b. 

teration in dimensions of radial tissues Mandoli, D. F., and w. R. Briggs, Photochem. 

relative to the vascular system are un- Photobiol., in press. 


Moritoshi lino and Winslow R. Briggs 

Phototropic curvature in higher plants Materials and Methods 
is the consequence of growth differential 

induced across the plant axis. Despite Maize seeds were sown on moist Kim- 

this obvious aspect of the phototropic re- pack paper. Seedlings were selected for 

sponse, the nature of growth distribution uniformity two days after sowing and 

during phototropic curvature has not been transplanted into moist vermiculite. 

clearly characterized (see Firn and Digby, Seedlings were grown for another day 

1980). Recent reinvestigation of this for use in experiments. The growth of 

problem by Franssen et al. (1981; 1982) plants took place under red light (R, 0.15, 

with oat coleoptiles or other plant ma- fxmol m" 2 s" 1 ) from the time of sowing 

terials suggested that growth inhibition at 24°-25°C. Plants also received some 

on the irradiated side of the plant axis is R from above during phototropic induc- 

the primary growth response causing tion and curvature development. The 

phototropic curvature. The authors in- length of the coleoptile at the time of the 

terpreted their results in favor of Blaauw's experiments was about 2 cm. The entire 

hypothesis (1915) that the curvature is coleoptile or the tip of the coleoptile was 

induced by differential growth inhibi- unilaterally irradiated with blue light 

tion. However, the results do not exclude (Corning blue glass filter, 5-60). When 

the possibility that the observed growth the coleoptile tip was to be irradiated, 

responses during phototropic curvature the coleoptile was covered during pho- 

are a mixture of growth redistribution, totropic induction with a cylindrical tube 

which is the basis of phototropic re- made of black photographic tape, leaving 

sponses, and more-general growth re- the 1-mm tip portion uncovered. Dura- 

sponses to light, which cannot result in tion of the irradiation was always 30 s, 

a growth differential. We obtained evi- and the longest transverse axis of the 

dence that the first positive curvature of coleoptile was parallel to the direction of 

the maize coleoptile is caused by a re- the incident light, 

distribution of growth without change in To obtain increment in length for ir- 

net growth. radiated and shaded sides of the coleop- 



tile, the coleoptile was marked 1.5 cm 
from the tip, just before phototropic in- 
duction, with India ink. In the fluence- 
response studies, the coleoptile was ex- 
cised at the mark 100 min after photo- 
tropic induction, and the excised 
coleoptiles were photocopied. Images of 
coleoptiles excised immediately after the 
marking and those excised after 100 min 
without phototropic induction (control) 
were also obtained. In the time-course 
studies, the coleoptile was photographed 
using R-sensitive film (Kodak). In this 
case, the images in developed film were 
enlarged by means of a slide projector, 
and the enlarged images were traced on 
paper. Length measurements were ob- 
tained with a digitizer directly on line 
with a computer. 

Results and Discussion 

The phytochrome-mediated growth 
responses in oat or maize seedlings have 
been shown to be so sensitive that even 
short exposure to dim green or blue light 
can cause substantial growth changes 
(e.g., Blaauw et al., 1968; Mandoli and 
Briggs, 1981; lino and Carr, 1981; lino, 
1982). In the present study, growth of 
the seedlings, phototropic induction with 
blue light, and subsequent curvature de- 
velopment all took place under R. In this 
way, we hoped to minimize any effect of 
phytochrome phototransformation by the 
blue light itself. The coleoptiles of these 
seedlings showed a peak-shaped photo- 
tropic fluence-response curve, charac- 
teristic of the classical first positive cur- 
vature of the coleoptile. The threshold 
and peak fluences were also similar to 
those obtained by Chon and Briggs (1966) 
with dark-grown maize seedlings given 
only a brief exposure to R two hours prior 
to phototropic induction. 

As illustrated in Fig. 21, phototropic 
curvature could be ascribed to growth 
stimulation on the shaded side and com- 
pensating growth inhibition on the irra- 
diated side. At the high fluences, there 
was some evidence for overall stimula- 

Fluence, Log /jmol rrr 2 

Fig. 21. Phototropic fluence-response curves 
for curvature (A) and growth increments on irra- 
diated and shaded sides (B). Whole coleoptile ir- 
radiation. Each datum point is the mean obtained 
from 7-9 plants. Different symbols indicate sepa- 
rate experiments. 

tion of growth. Irradiation of the coleop- 
tile tip alone showed a similar phototropic 
fluence-response curve, with identical 
threshold and peak fluences to those ob- 
tained for whole-coleoptile irradiation 
(Fig. 22). Growth stimulation on the 
shaded side and growth inhibition on the 
irradiated side were also apparent. In 
this case, the stimulation of overall growth 
appeared to be shifted to higher fluences. 
The above results for phototropic flu- 
ence-response curves obtained by irra- 
diating the entire coleoptile as well as 
only the coleoptile tip agrees with the 
view that the extreme tip of the coleop- 
tile is the site of photoreception for the 
first positive curvature of the coleoptile 
(Lange, 1927). The fact that the increase 
in the overall growth is shifted to higher 
fluences when only the tip of the coleop- 
tile is irradiated indicates that this re- 
sponse to blue light is mediated by a 



60 80 

Time, min 


Fig. 22. Phototropic fluence-response curves 
for curvature (A) and growth increments on irra- 
diated and shaded sides (B). Coleoptile tip irradia- 
tion. Other details as in Fig. 21. 

different mechanism than that mediating 
the phototropic response. (The photo- 
reception site of the latter response is 
not restricted to the tip of the coleoptile.) 
Time courses of the growth increment 
at shaded and irradiated sides were ob- 
tained for the blue light fluence causing 
the peak of the first positive curvature, 
together with the growth increment for 
the control plants that received no pho- 
totropic induction. As shown in Fig. 23, 
irradiation of the entire coleoptile and 
that of the coleoptile tip produced essen- 
tially identical time courses. Stimulation 
of growth on the shaded side and the 
inhibition of growth on the irradiated side 
began simultaneously about 30 min after 
the phototropic induction. The growth on 
the irradiated side almost ceased. How- 
ever, since the growth rate on the shaded 
side is doubled relative to the control rate, 
it appears that the overall growth during 
phototropic curvature was not much af- 

Fig. 23. Time courses of growth increments on 
irradiated and shaded sides of the coleoptile. The 
whole coleoptile (squares) or the coleoptile tip (tri- 
angles) was unilaterally irradiated at time with 
blue light (5.0 ixmol m 2 ). Control plants (circles) 
received no phototropic induction. Each time-course 
point is the mean obtained from ten plants. 

The results obtained for the fluence- 
response relationship and the time courses 
show that the phototropic growth differ- 
ential is essentially the result of growth 
redistribution and is not necessarily ac- 
companied by changes in the net growth. 
The results are consonant with the view, 
known as the Cholodny-Went theory of 
tropism, that phototropic curvature is the 
consequence of a lateral redistribution of 
auxin (Went and Thimann, 1937). 


Blaauw, A. H., Z. Bot. 7, 465-532, 1915. 

Blaauw, O. H., G. BlaauwJansen, and W. J. van 
Leeuwen, Planta 82, 87-104, 1968. 

Chon, H. P., and W. R. Briggs, Plant Physiol, bl, 
1715-1724, 1966. 

Firn, R. D., and J. Digby, Annu. Rev. Plant Phys- 
iol., 31, 131-148, 1980. 

Franssen, J. M., S. A. Cooke, J. Digby, and R. D. 
Firn, Z. Pfianzerphysiol. 103, 207-216, 1981. 


Franssen, J. M., R. D. Firn, andJ. Digby, Planta Lange, S., Jb. Wiss. Bot. 67, 1-51, 1927. 

155, 281-286, 1982. Mandoli, D. F., and W. R. Briggs, Plant Physiol. 

lino, M., Planta 156, 388-395, 1982. 67, 733-739, 1981. 

lino, M., and D. J. Carr, Plant Sci. Lett. 23, 263- Went, F. W., and K. V. Thimann, Phytohor- 

268, 1981. mones, Macmillan, New York, 1937. 



Moritoshi lino, Winslow R. Briggs, and Eberkard Schafer 

Action spectra for phototropic curva- in the dark for one day prior to curvature 
ture in higher plants show characteristic induction. The fluence-response curve for 
shapes in blue light (B) and near-ultra- R-induced mesocotyl curvature, ob- 
violet regions (e.g., Thimann and Curry, tained by measuring curvature 100 min 
1960). Ultraviolet-sensitive phototrop- after phototropic induction, shows peaks 
ism, the response site distinct from that in two fluence ranges, designated first 
for the B-type phototropism, has also been positive range (from the threshold to the 
evidenced (Curry et al., 1956). A general trough, 10" 4 -10" 02 |jimol m" 2 ) and sec- 
conclusion reached earlier is that the ond positive range (above the trough, 
wavelengths longer than about 520 nm 10~ 02 -10 2 - 5 (xmol m~ 2 ). The fluence-re- 
are normally phototropically inactive in sponse curve for B is similar to that for 
higher plants (Galston, 1959). R but is shifted two orders of magnitude 

More recently, the question whether to higher fluences. Although the classical 
or not phytochrome-mediated inhibition first positive curvature of the coleoptile 
of hypocotyl growth can result in pho- is apparent with B, this response is not 
totropic curvature was tested by Shrop- found with R. Positive mesocotyl cur- 
shire and Mohr (1970) in etiolated vature induced by either R or B is elim- 
seedlings of Sinapis and Fagopyrum. mated by immediate preirradiation with 
Under continuous irradiation (24 h) of the vertical R, unlike the B-induced coleop- 
seedlings with unilateral R or far red (FR) tile curvature, which is not eliminated, 
light, they were able to produce gra- Mesocotyl curvature appears to be in- 
dients in anthocyanin production, known duced by R-sensitive photosy stems, 
to be controlled by phytochrome, across Both R and B lead to negative cur- 
the hypocotyl tissue, especially of Fa- vature of the coleoptile; the B-induced 
gopyrum. However, no phototropic cur- negative curvature is found at fluences 
vature could be observed, although the higher than those showing positive co- 
light fluence rates used were such not to leoptile curvature. Clinostat experi- 
result in saturation of growth inhibition, ments show that the negative coleoptile 

While studying phototropic curvature curvature induced by either R or B is 

of maize seedling snoots, we obtained ev- gravitropic compensation for positive 

idence that the mesocotyls of these seed- mesocotyl curvature, 

lings can bend positively toward R after On the clinostat, the R-induced me- 

a P fr gradient is established across the socotyl curvature develops after a lag and 

plant axis. Results are summarized be- continues through two successive phases 

low. having different curvature rates, the late 

Unilateral irradiation with R or B elic- phase slower than the early phase. Du- 

its positive curvature of the mesocotyl of rations of the lag and the early curvature 

maize seedlings (Zea mays L.) raised un- phase become shorter as the fluence is 

der R for two days from sowing and kept increased in the first positive range, 


whereas both are prolonged as the flu- first positive range could also be brought 
ence is further increased to the second about by the FR. Red light-induced me- 
positive range. socotyl curvature in the first positive 
Mesocotyl curvature in the second pos- range is inhibited by vertical FR given 
itive range is reduced by vertical FR ap- either before or after the phototropic in- 
plied after phototropic induction with R, duction. This result indicates that the 
but it is not affected by FR applied be- vertically applied FR reduces the re- 
fore. Unilateral irradiation with FR fol- sponse by adding P fr at both sides of the 
lowing vertical irradiation with a high R plant axis, 
fluence leads to negative curvature of the 

mesocotyl. We conclude that mesocotyl References 

curvature in the second positive range 

results from a gradient in the amount of Galston, A. W., in Encyclopedia of Plant Physi- 

P fr established across the plant axis. Me- #W» Vo1 - i 7 " 1 - 492-529, W. Ruhland, ed., 

,1 , ji n>^ 4. -i- Springer, Berlin, 1959. 

socotyl curvature m the first positive Cur ^ % ^ K y Thimann> and p M Ray? 

range is induced by unilateral irradiation Physiol. Plant. 9, 429-440, 1956. 

with FR. This result is not inconsistent Shropshire, W., Jr., and H. Mohr, Photochem. 

with the view that the curvature re- Photobiol. 12, 145-149, 1970. 

sponse is mediated by phytochrome, since Th ™> K - v -> an <J G. M Curry in Compara live 
.* „ > i> i , i Biochemistry, Vol. 1, 243-309, H. S. Mason and 

the very small amount of phytochrome M Florkin> eds>> Academic Press, New York, 

phototransformation induced by R in the i960. 


Jeanette S. Brown 

I. Spectra of Three Native , . ,, TT ™ X , . , . , ., 

Chlorophyll-Proteins P rot T (L ^S n whlch m ^ A ^he most 

rapidly. Other minor green bands were 

Our goal for many years has been to also found, but these were similar to one 

measure the absorption spectrum of chlo- or the other of the three major bands. 

rophyll a in its photosynthetically func- We repeated the procedure with spin- 

tional state. The problem has been to ach chloroplasts, and measured absorp- 

separate the chlorophyll-proteins from tion and fluorescence spectra at 77 K of 

chloroplast membranes without altering the pigment-proteins extracted from the 

the state of the pigment molecules. Ear- separate gel bands (Brown, 1983). Those 

lier this year, Picaud et al. (1982) found spectra of the chlorophyll-proteins lo- 

a separation procedure which meets these cated close to the reaction centers of pho- 

requirements. They used digitonin for tosy stems I and II and of the antenna 

solubilizing chloroplast membranes and are shown in Fig. 24. Spectra of washed 

sodium deoxycholate (DOC) for a charged spinach chloroplast membranes are shown 

carrier during poly aery lamide gel elec- at the top of the figure for comparison. 

trophoresis (PAGE). Three bands were After these membranes were solubilized 

identified in the gel: a phot osy stem I with digitonin, their absorption spectra 

chlorophyll-protein (CPI), which was the were unchanged but their fluorescence 

slowest migrating band; a photosystem emission spectra (not shown) resembled 

II band (CPa); and an antenna Chi a-b the spectrum of LHCP, because energy 




650 700 

Wavelength, nm 


Fig. 24. Absorption and fluorescence-emission 
spectra (77 K) of spinach chloroplasts and three 
chlorophyll-protein complexes isolated by DOC- 
PAGE. Excitation at 438 nm, slit-width = 10 nm; 
emission slit-width = 3 nm. 

transfer from the antenna to the other 
chlorophyll-proteins was disrupted. When 
absorption spectra of the three isolated 
chlorophyll-proteins were added to- 
gether in proportions corresponding to 
their density in the gel (35% CPI, 15% 
CPa, and 50% LHCP), the sum nearly 
matched the spectrum of the thylakoids. 
This result shows that the spectra of these 
three complexes account for essentially 
all of the chlorophyll absorption in the 
membranes, and each spectrum repre- 
sents the absorption of the native state 
of each chlorophyll-protein. 

This is the first time that such spectra 
of native chlorophyll-proteins isolated 
from the same chloroplasts in one pro- 
cedure have been measured (Fig. 24). The 
maximum near 650 nm in LHCP is from 
Chi b. Obviously the shape of the Chi a 
absorption differs considerably between 
the three spectra. Because we had pre- 
viously postulated that Chi a absorption 
in most plants may be represented by 
four major Gaussian-Lorentzian-shaped 
component bands, the RESOL curve-fit- 
ting computer program was applied to 

these spectra with input components 
similar to those previously found with 
both whole and partially fractionated 
thylakoid particles (French et al., 1972). 
Only the LHCP spectrum could be 
matched well with the four "universal" 
components, or "forms," at approxi- 
mately 660, 670, 677, and 684 nm. CPa 
has the 660-nm and 670-nm components; 
it also has a large, broad component at 
675 nm and another at 682 nm. CPI has 
components at 662, 668, 679, and 687 nm; 
the 679 band is very large and all com- 
ponents are wider than those in the other 
spectra. Thus we may no longer consider 
that only four major absorbing forms of 
Chi a exist. The antenna chlorophylls 
dedicated to each photosystem have their 
own groups or pools of pigment mole- 
cules. Probably these differences reflect 
differences in the proteins and their 
binding to chlorophyll. 

II. Detergent Action on 
Chlorophyll-Protein Absorption 

Because sodium dodecylsulfate (SDS) 
and Triton X-100 are used so commonly 
for the solubilization and separation of 
intrinsic membrane proteins, the effects 
of these detergents on chlorophyll ab- 
sorption should again be documented 
(Year Book 78, 189-194; 81, 38-40). Both 
the antenna Chi a-b protein (LHCP) iso- 
lated from higher plants and a similar 
complex from Mantoniella (see section 
III, below) have a sharp absorption max- 
imum near 676 nm when they are isolated 
with digitonin, DOC, or Triton. Some 
years ago we isolated LHCP from spin- 
ach chloroplasts by solubilization with 
Triton followed by separation by cen- 
trifugation in sucrose density gradients 
(Year Book 78, 193-194). LHCP was 
precipitated from the sucrose gradient 
band by the addition of cations, lyophi- 
lized, and stored at -20°C. The absorp- 
tion of this LHCP, resuspended in buffer 
by sonication, still retained the 676-nm 
maximum (Fig. 25). The Mantoniella 
LHCP was prepared with digitonin and 
DOC, as described in section III, below. 



630 640 650 660 670 680 690 700 

Wavelength, nm 

Fig. 25. Absorption spectra (77 K) of spinach 
and Mantoniella light-harvesting chlorophyll-pro- 
teins before and after treatment with 0.1% SDS or 

Aliquots of each LHCP were diluted 
with 50 nM Tris buffer at pH 7.3 to about 
30 |xg/ml of chlorophyll and 0. 1% of either 
SDS or Nonidet P40 (a detergent similar 
to Triton X-100). After mixing with the 
detergent, the samples were incubated 
for 10 min at room temperature in dark- 
ness before absorption spectra were re- 
corded at 77 K. The spectra before and 
after detergent treatments are shown in 
Fig. 25. The most obvious change is a 
decrease in the 676-nm peak. That this 
change was caused by Nonidet is sur- 
prising because the spinach LHCP was 
isolated originally with this detergent. 
Probably the chlorophylls in the isolated 
complex were more exposed to detergent 
action than they were in the thylakoid 
membrane. The view that the 676-nm peak 
is an artifact caused by cation-induced 
aggregation of LHCP and therefore is 
lowered when the complexes are disag- 

gregated by detergent action may be 
countered by the observation that the 
most rapidly migrating LHCP after elec- 
trophoresis with DOC is certainly not ag- 
gregated, but does have the 676-nm 
absorption maximum (Fig. 24). Spinach 
LHCP was also treated wth 0.1% DOC 
in this experiment, but there was no 
change in its spectrum (not shown). 

Another example of SDS action can be 
seen in Fig. 26 and Table 4, where spec- 
tral analyses of Mantoniella antenna 
chlorophyll-protein isolated by either su- 
crose gradient centrifugation with 0.05% 
DOC or by PAGE with 0.07% SDS are 
compared. Like the curves in Fig. 25, 
SDS caused a decrease in absorption at 
676 relative to 670 nm. In addition, the 
curve resolutions show that several of 
the components changed. The long- 
wavelength bands at 680 and 687 nm 

620 630 640 650 660 670 680 

Wavelength, nm 


700 710 

Fig. 26. RESOL analyses of absorption spec- 
tra of Mantoniella antenna chlorophyll-protein iso- 
lated either by sucrose density gradient 
centrifugation with DOC or by PAGE with SDS. 
The measured data are plotted as points, while the 
line through them is the sum of the component 
curves. Band maxima and half-bandwidths are given 
in Table 4. 


TABLE 4. The Band Maxima, Half-widths, and Proportion (%) of Total Area from 
RESOL Spectral Analyses in Fig. 26* 



Band No. 

Max (nm) 

HBW (nm) 


Max (nm) 

HBW (nm) 

































































*Spectra are of antenna chlorophyll-proteins prepared from Mantoniella membranes solu- 
bilized by either digitonin and DOC or by SDS only. 

nearly disappeared. The 676-nm and 669- should be used with material that has not 

nm components retained the same peak been exposed to SDS or Triton, 
positions and widths, but the band at 676 

nm decreased by 1%, and the band at 669 m Chlorophyll-Protein 

increased by 3% The 662-nm band shifted CoMpLEXES FR0M Primitive Green 


to 663 nm and became 1 nm wider — a 
small change but evidently enough to re- 
move the shoulder near 660 nm on the Many times in the past we have com- 
spectrum. The 652-nm component be- pared absorption spectra of diverse algal 
came narrower by 6 nm but occupied 2% species to gain a better understanding of 
more of the total area under the curves the essential nature of chlorophyll in vivo, 
after SDS treatment; also, it changed from Because we now know that most of the 
85% to 67% Gaussian in shape. The 643- chlorophyll is attached to protein, our goal 
nm Chi b band became narrower by 4 nm has become the extraction of native pig- 
but accounted for nearly the same total ment-protein complexes from the differ- 
area. The 632-nm band from the Chi c- ent algae. This year we have begun the 
like pigment and the composite 621-nm study of a primitive group of green mi- 
band remained essentially unchanged, croalgae called Prasinophytes (Norris, 
Except for the 676-nm band, the pro- 1980). This group contains a number of 
portions of Gaussian- and Lorentzian- motile genera thought to be the anteced- 
shaped curves changed somewhat with ents of the better-known Chlamydo- 
all of the components, most notably with monas and Dunaliella. Two genera, Mi- 
the 652-nm component. cromonas and Mantoniella, are at the 
In summary, SDS action caused subtle bottom of the evolutionary scale, and both 
changes in most of the component bands contain a chlorophyll c-like or protochlo- 
comprising the chlorophyll absorption rophyll-like pigment — magnesium 2,4 
spectrum. If these component bands do divinylphaephorphyrin a 5 monomethyl 
indeed represent discrete groups of chlo- ester — in addition to chlorophylls a and 
rophyll molecules bound to different sites b (Ricketts, 1966a). Micromonas and 
on a protein, it is not surprising that a Mantoniella also have an unusual caro- 
strong detergent can disturb several of tenoid, which has been named micro- 
these sites. These results emphasize that none, or microxanthin in its reduced form 
studies designed to investigate the state (Ricketts, 1966b). Ricketts studied Mi- 
of chlorophyll in vivo using such tech- cromonas pusilla only, but we have found 
niques as resonance Raman, dichroism, that both Micromonas pusilla (Culture 
or fluorescence polarization spectroscopy Collection, University of Texas, UTEX 



No. 991) and Mantoniella squamata 
(UTEX No. 990) have identical absorp- 
tion and fluorescence spectra. Cells of both 
species are highly motile spheres about 
one micron in diameter. We have found 
that Mantoniella grows much better than 
Micromonas in an enriched seawater 
medium, making it more useful for bio- 
chemical studies. 

We routinely culture Mantoniella in 3- 
liter batches at about 20°C over a bank 
of daylight fluorescent lamps (continuous 
light) with air bubbling through the me- 
dium. It is our intention to try several, 
more-defined growth conditions in the fu- 
ture. The cells are harvested by centrif- 
ugation and broken by passage through 
a French press. The cell membranes are 
washed with 5 nM EDTA, pelleted, and 
stored frozen at -80°C. 

An absorption spectrum of the cell 
membranes shows several features that 
may be unique to Mantoniella and Mi- 
cromonas (Fig. 27). The peak of Chi b 
near 643 nm is shifted downward by nearly 
5 nm compared to its maximum in spec- 
tra of higher green algae and plants. The 

band at 632 nm is probably from the Chi 
c-like pigment. The band near 520 nm 
may be one of the peaks of microxanthin. 
To estimate the chlorophyll concentra- 
tion we used the equations of Jeffrey and 
Humphrey (1975), which were designed 
to measure Chls a, b, and c in mixed 
phytoplankton populations. The cells each 
had approximately 4 x 10 -8 (xg chloro- 
phyll, with ratios of Chi alb = 3 and al 
c = 10. These ratios have varied be- 
tween 2.6 and 3.4 and between 8.9 and 
11 because of yet undefined culture con- 

We are trying various detergents for 
solubilization and sucrose density gra- 
dient centrifugation or gel electropho- 
resis for isolation of chlorophyll-proteins 
from the cell membranes. Thus far we 
have separated a green fraction in su- 
crose gradients from membranes solu- 
bilized in 0.5% Nonidet. This fraction is 
highly enriched in photosystem I (Chi/ 
P700 = 89). Another brown fraction in 
the same gradient contained most of the 
antenna pigments (Chi alb = 1.8). Spec- 
tra of these fractions are shown in Fig. 

1 — i — i — r 

t — i — i — r 

400 450 500 550 600 650 

Wavelength, nm 



Fig. 27. Absorption (solid lines), fluorescence excitation (400-600 nm), and emission (650-750 nm) 
spectra measured at 77 K of chloroplast membranes, antenna pigments, and PS I fractions isolated from 
Mantoniella. Excitation at 438 nm for emission spectra, slit-width = 3 nm; excitation slit-width = 4 
nm for emission at 680 nm. 


27. We have not yet been able to sub- cies for study by biologists, and at least 

divide further the brown antenna pig- two books are devoted exclusively to this 

ment fraction. organism (Wolken, 1967; Buetow, 1968). 

An important observation from fluo- In our laboratory, French and Elliott 

rescence excitation spectra of both the (Year Book 57, 278-286) first observed 

membranes and fractions is the appear- the unique shape of the chlorophyll a ab- 

ance of a band near 520 nm. Possibly, sorption spectrum of Euglena with its 

microxanthin can transfer absorbed light three distinct maxima at about 672, 680, 

energy to Chi a and may be in a class of and 695 nm. Later we found (Year Book 

truly photosynthetic carotenoids similar 59, 330-333) that the relative heights of 

to peridinin and fucoxanthin. Finding ex- these maxima as well as the proportions 

citation bands at 470 nm from Chi b and of Chi a and Chi b depended upon the 

near 520 nm from a carotenoid in spectra light intensity during growth of the cells; 

of the PS I fraction was unexpected. I the brighter the light, the more Chi b 

have observed these bands even in PS I and the less longer-wavelength-absorb- 

fractions separated by gel electropho- ing Chi a forms. 

resis, which should be less contaminated Euglena was included in our first com- 
by antenna fractions than those sepa- parisons of isolated chlorophyll-proteins 
rated in sucrose gradients. The emission from diverse plants (Year Book 73, 694- 
spectrum of Chi a in PS I fractions is also 706). Solubilization of Euglena chloro- 
more like that in antenna-pigment frac- plasts with Triton X-100 or sodium do- 
tions. The emission spectrum of either decylsulfate, followed by separation of 
intact or broken cells does not have the fractions with hydroxylapatite chroma- 
long-wavelength bands seen in most green tography or polyacrylamide gel electro- 
algae or higher plants at low tempera- phoresis (PAGE), respectively, yielded 
ture (Brown, 1969). Obviously Manton- chlorophyll-proteins that were only partly 
iella does have long-wavelength analogous to those from other algae and 
absorption bands (Fig. 27). Possibly the higher plants. One of these fractions, 
fluorescence yield of photosystem I in this called CPa x , has been observed only from 
alga is so low that a small amount of an- Euglena and has an unusual absorption 
tenna pigment can mask its emission. peak at 683 nm in addition to the broad 

Micromonas-like phytoplankton have maximum at 670 nm. Later we were able 
been isolated from the open ocean, some- to isolate three different chlorophyll-pro- 
times at considerable depth (Jeffrey and teins from Euglena using Triton X-100 
Hallegraeff, 1980). The little light that to solubilize the chloroplasts, and sucrose 
penetrates to these depths is mostly in density gradient centrifugation to sepa- 
the blue-green spectral region. Our spec- rate the fractions (Brown, 1980). By this 
tral studies show that these organisms procedure the lightest gradient band 
are well adapted by their pigment con- contained the Chi a-b antenna protein, a 
tent to harvest this light. middle band was highly enriched in pho- 

We have begun to study two other tosystem I (Chl/P700 = 50), and the 
Prasinophytes, Platymonas sp. (UTEX heaviest fraction contained both photo- 
No. 634) and Tetraselmis chui (CSIRO system I (CM/P700 = 108) and CPa x . 
Culture Coll. No. CS26), which may ac- In our early experiments with SDS and 
tually be in the same genus (Norris, 1980). PAGE (Year Book 73, 694-706), we could 
Their spectra are similar to each other distinguish the photosystem I band (CPI) 
and to those of most other green algae near the top of the gel, but the other 
containing Chls a and b. chlorophyll bands quickly broke down to 
IV. Chlorophyll-Protein fr u f Pig™**- Ortiz and Stutz (1980) were 
Complexes From Euglena able u *? separate the three major chlo- 

y rophyll-protems (CPI, CPa, and LHCP) 

Euglena has long been a favorite spe- with PAGE after solubilization of Eu- 



glena chloroplasts with SDS. However, 
because SDS destroys the sharp absorp- 
tion peak at 683 nm in CPa l5 and Triton 
destroys the long-wavelength-absorbing 
Chi a (Brown, 1980), we tried the new 
procedure of Picaud et al. (1982), which 
uses digitonin for solubilization and so- 
dium deoxycholate for electrophoresis. 
We still found considerable free pigment, 
but absorption and fluorescence spectra 
of three green bands isolated from the 
gel strongly suggested that they are CPI, 
CPa, and LHCP (Fig. 28). 

The slowest migrating band in Fig. 28 
(#1) has the long- wavelength absorption 
and emission characteristics of CPI in 
most higher plants. Band #2 has the ab- 
sorption typical of CPa, with its promi- 
nent shoulder near 680 nm (Fig. 24) and 
emission band near 695 nm, but the emis- 
sion near 720 nm may be unique to Eu~ 
glena. Band #3 not only has most of the 
Chi b, like LHCP from other plants, but 
also has a sharp absorption band near 683 
nm, like CPa^ This last result was un- 
expected because the CPa x is isolated af- 
ter Triton treatment and hydroxylapatite 
chromatography has no Chi b. Possibly, 
CPa x is co-migrating with LHCP in this 
electrophoretic system. Futher experi- 
ments will be necessary to clarify this 
last point, but the results suggest that 
definitive characterization of the chlo- 
rophyll-proteins of this very interesting 
alga will soon be possible. 


Brown, J. S., Biophys. J. 9, 1542-1552, 1969. 
Brown, J. S., Biochim. Biophys. Acta 591, 9-21, 

Buetow, D. E., ed., The Biology of the Euglena, 

Vol. II, Academic Press, New York, 1968. 

I i 1 r 



650 700 

Wavelength, nm 


Fig. 28. Absorption (solid lines) and fluores- 
cence emission spectra (77 K) of Euglena chloro- 
plast fragments and fractions obtained after 
solubilization with digitonin and PAGE with de- 
oxycholate. Excitation at 438 nm, slit-width = 10 
nm; emission slit-width = 3 nm. 

French, C. S., J. S. Brown, and M. C. Lawrence, 

Plant Physiol. U9, 421-429, 1972. 
Jeffrey, S. W., and G. F. Humphrey, Biochem. 

Physiol. Pflanzen 167, 191-194, 1975. 
Jeffrey, S. W., and G. M. Hallegraeff, Mar. Ecol. 

Prog. Ser. 3, 285-294, 1980. 
Norris, R. E., in Phytoflagellates, 85-145, E. R. 

Cox, ed., Elsevier/North Holland, New York, 

Ortiz, W., and E. Stutz, FEBS Lett. 116, 298-302, 

Picaud, A., S. Acker, and J. Duranton, Photosyn. 

Res. 3, 203-213, 1982. 
Ricketts, T. R., Phytochemistry 5, 223-229, 1966a. 
Ricketts, T. R., Phytochemistry 5, 571-580, 1966b. 
Wolken, J. J., Euglena, Appleton-Century Crofts, 

New York, 1967. 





Jeanette S. Brown 

In 1968, we first used a computer pro- 
gram called RE SOL to deconvolute ab- 
sorption spectra of chloroplast and algal 
fragments into Gaussian-shaped compo- 
nent curves (Year Book 67, 536-546). 
When mixed Gaussian-Lorentzian com- 
ponents were used, the fit was much bet- 
ter, particularly in the long-wavelength 
tail region. Symmetrical component bands 
have always been assumed until now 
(French et al., 1972). However, it was 
noted in 1968 that the red absorption 
bands of Chi a dissolved in 80% acetone- 
water and, later, bacteriochlorophyll a 
dissolved in 80% acetone-water (Phil- 
ipson and Sauer, 1972) were not sym- 
metrical but were skewed towards shorter 
wavelengths. This observation sug- 
gested that the complex spectra of chlo- 
rophyll in vivo might better be matched 
with asymmetric component bands. 

This year we contracted with David 
Pasta of DMA Corporation to modify 
RESOL to permit the use of skewed 
components. In effect, the program can 
now vary the ratio between the half- 
widths of the left (short-wavelength) and 
right sides of each component band. We 
can set this ratio or allow the program 
to find the best one; the ratio is 1 for a 
symmetrical band. 

To test the modified program, we re- 
corded the absorption spectrum at 20°C 
of an 80% acetone-water extract of An- 
acystis nidulans, a blue-green alga in 
which Chi a is the only chlorophyll (Year 
Book 80, 15-16). The optical density at 
the peak, 664 nm, was 0.45. We at- 
tempted to match the absorption band of 
this spectrum between 630 and 700 nm 
with either two or three components; the 
component added near 630 nm repre- 
sented lower vibrational levels of the ma- 
jor band. Different peak wavelengths, 
band widths, Gaussian-Lorentzian ratios, 

630 640 650 660 670 680 

Wavelength, nm 

690 700 

Fig. 29. Absorption spectrum of Chi a in 80% 
acetone-water recorded at 20°C resolved into 
(A) two or (B) three components. Band No. 1, that 
at the longest wavelength, represents lower vi- 
brational levels of the major band. The measured 
data are plotted as points, while the line through 
them is the sum of the component curves. The error 
of fit is shown below each spectrum on a scale with 
the designated magnification. 


TABLE 5. Four RESOL Analyses of an Absorption Spectrum (630-700 nm) of Chi a in 80% 


Band No. 

Max (nm) 

HBWf (nm) 

Gaussian (%) 

Skew Factor 

Total Area (%) 


































































*The curve was analyzed with two or three components and with or without skewing. The S.E. of 
each analysis is expressed as percentage of peak height, 
f Bandwidth at half height. 
tFigure 29A. 
§ Figure 29B. 

and half-bandwidth ratios (skew factors) sorption spectrum of Chi a cannot be 
were tried for input to RESOL. matched completely with a single Gaus- 
Figure 29A shows the spectrum re- sian-Lorentzian component, skewed or 
solved with two components, and Fig. not. Perhaps another type of skewed 
29B shows the spectrum with three. Ta- function may give a better approxima- 
te 5 shows the band parameters for the tion of chlorophyll absorbance. 
components when the RESOL program The continuing interest and help of 
was allowed to skew each component (Fig. Glenn Ford for modifying RESOL is much 
29) and also for corresponding analyses appreciated, 
where skewing was not allowed. The data 
show that the addition of only 1% of a 
band at 648 nm improved the match over 

that with one component, but skewing References 

the components did not change the re- _ , „ n 

suits significantly. The main component ™; £&£ & *^«** c - L ™, 

at 664 nm IS about 85% Gaussian and 155 Philipson, K. D., and K. Sauer, Biochemistry 11, 

Lorentzian in shape. Apparently, the ab- 1180-1185, 1972. 



ADAPTATION BY THE RED ALGA Porphyra perforata 

David C. Fork and Kazuhiko Satoh 


Even though light is required for pho- 
tosynthesis, the absorption of too much 
light can damage the photosynthetic ma- 
chinery by photoinhibition (Powles and 
Thorne, 1981). By contrast, plants that 
live under low-light conditions often ab- 
sorb light of a quality (color) that is not 
optimal for photosynthesis, thereby pro- 
ducing unbalanced excitation of the two 

Plants apparently have several mech- 
anisms to aid them in overcoming limi- 
tations from absorption of too much light 
or light of an unfavorable wavelength. 
Last year, we reported (Year Book 81, 
45-58) that in a blue-green and in a green 
alga we could clearly observe one of these 
mechanisms, the so-called state I-state 
II transition (Bonaventura and Myers, 
1969; Murata, 1969; Wang and Myers, 
1974). This year, we found two other 
mechanisms available to the red alga 
Porphyra perforata in overcoming un- 
favorable conditions of illumination. These 
mechanisms appear related to the ability 
of Porphyra to thrive under conditions 
of high light intensity, high salinity, and 

The plants obtained in these studies 
were collected at Half Moon Bay, Cali- 
fornia, and maintained under illumina- 
tion in open dishes of sea water at 13°C. 
Samples were kept in the dark for 1 h 
before use. 

Fluorescence spectra at 77 K and time 
courses at room temperature of fluores- 
cence at 685 nm were measured with a 
fiber-optic system to excite and collect 
the fluorescence (Year Book 78, 196-199). 
Light-induced absorbance changes were 
measured with a single-beam spectro- 

We could see marked changes in the 
intensity of chlorophyll fluorescence 

(Kautsky transients) when dark-adapted 
Porphyra was illuminated with light ab- 
sorbed largely by photosystem II (light 
II, data not shown). Figure 30 shows flu- 
orescence emission spectra measured at 
77 K before and after illumination of dark- 
adapted Porphyra thalli with light II. The 
difference spectrum, given in the bottom 
curve of the figure, shows that light II 
produced a fluorescence decrease of both 

-i — i — i — i — I — I - 1 — n — I — i — i — i — r 


650 700 750 

Wavelength, nm 


Fig. 30. Fluorescence emission spectra of Por- 
phyra perforata before and after illumination of 
thalli with 235 fxW/cm 2 of green light (light II) for 
2 min at 21°C. The dark-adapted thallus (solid line) 
or preilluminated thallus (dashed line) were cooled 
to 77 K and the fluorescence spectra measured. 
Green excitation light was obtained by passing the 
white light from a 150W, 21.5V projector lamp (type 
DLS) through Corning glass filters 4-96 and 3-96. 
For fluorescence measurements, the light had an 
intensity of 58 |xW/cm 2 . The dot-dashed line rep- 
resents the difference spectrum of the dark-adapted 
minus the preilluminated sample. 



the photosystem II (PS II) fluorescence 
bands (F685 and F695) and a small de- 
crease of the photosystem I (PS I) fluo- 
rescence band (F736). This result suggests 
that light II leads to some kind of pig- 
ment rearrangement within the thyla- 
koid membranes of Porphyra. Such light 
II-induced fluorescence changes were not 
seen in the blue-green alga Synechococ- 
cus (Year Book 81, 45-49) or in the green 
alga Scenedesmus (Year Book 81, 54-58). 

Illumination of dark-adapted Por- 
phyra with light I for 30 min produced a 
fluorescence increase that was larger at 
695 nm than at 685 nm, and PS I fluo- 
rescence decreased (Fig. 31). These 
changes are characteristic of a state II- 
state I transition (Year Book 81, 45-49) 
and suggest that rearrangements of pig- 
ment-proteins within the thylakoid mem- 
branes produced by light I are different 
from those induced by light II. 

We examined whether the changes we 
observed in the fluorescence spectra were 
related to redox changes of Q (Duysens 

and Sweers, 1963) by measuring fluores- 
cence spectra under conditions where Q 
was either all oxidized (F ) or all reduced 
(F m ). The results (not shown) indicated 
that the fluorescence spectra induced by 
redox changes of Q differed from those 
produced either by light I or light II ex- 
cept that the two fluorescence bands at 
685 and 695 nm changed in parallel, as 
they did with light II illumination. 

We measured activities of PS I and 
PS II in cells that were preilluminated 
with light I or II, as described above. 
For a PS I reaction we measured the in- 
itial rates of oxidation of cytochrome c- 
553. Figure 32 gives the kinetics of the 
photooxidation of cytochrome c-553 on a 
semilogarithmic plot before and after il- 
lumination of a dark-adapted Porphyra 
with light I or light II (states II, I, and 
III, respectively; see below). The data 

i — I — i — r 

\ / 

state I -state II \. 7 ? 9/ 

\ I / 

J — i — i — i — i__J i i i a. I i i i l 

J L 


7 , ! ! 


"-. 1.6 

^\^ Q -\ state I (□) 



o 1.2 


state II (o)\_ 

t Q8 

state III c^) ^\ 

s. - 

3 0.4 



i i i i 

1 - 


0.1 0.15 
Time, s 

0.2 0.25 

650 700 750 

Wavelength, nm 


Fig. 31. Fluorescence emission spectra of Por- 
phyra before (solid line) and after (dashed line) 
illumination of thalli with far red light (\>690 nm) 
for 30 min at 21 °C. The samples were cooled to 
77 K and fluorescence emission spectra were mea- 
sured as described in Fig. 30. 

Fig. 32. Photooxidation of cytochrome c-553 
(measured at 419 nm) in states I, II, and III (see 
text) in Porphyra. State I (squares) was obtained 
by illuminating the sample with far red light for 30 
min; state II (circles) was obtained by dark-adapt- 
ing the thallus for 1 h; state III (triangles) was 
obtained by illuminating the dark-adapted sample 
with light II for 3 min. The actinic light was 632.8 
nm from an He-Ne laser (128 (xW/cm 2 ). DCMU (10 
jjlM) and DBMIB (10 llM) and ascorbate (5 mM) 
were added soon after the pretreatment and ab- 
sorbance changes measured after 1 min of incu- 






<D 8 





u_ 4 _ 

2 - 





state 1 

l l 


— — 

state II 

— / / 



state III 





_ J 






i i 


Time, sec 




Fig. 33. Time courses of fluorescence induction in Porphyra in the presence of DCMU in states I, 
II, and III (see text). DCMU (10 |xM) was added soon after the pretreatment, and the time courses 
were measured after 1 min of incubation. State III was obtained by illuminating the thallus with light 
II for 2 min. The other states were obtained as described in Fig. 32. 

of Fig. 32 show that the rate of cyto- 
chrome oxidation was decreased by light 
I preillumination. By contrast, light II 
preillumination did not change the rate 
of photooxidation of the cytochrome, and 
the rate remained the same as in the dark- 
adapted sample. 

A PS II reaction was followed by mea- 
suring the time courses of fluorescence 
induction in Porphyra in the presence of 
DCMU. The data of Fig. 33 show that 
light I treatment induced a large increase 
in F v (variable fluorescence) without a 
large change in F fluorescence. The flu- 
orescence rise curve seemed to be com- 
posed of a fast and a slow component 
(Melis and Homann, 1976). Light I in- 
creased the extents of both of the two 
components, but the rate of the fast com- 
ponent was slightly higher. By contrast, 
light II induced a marked decrease in F 
fluorescence, a decrease in the fast com- 
ponent, and an increase in the slow com- 
ponent of the fluorescence time course. 
Since light II induced a state that is ap- 

parently different from states I or II, we 
have defined it as state III. 

Light II preillumination did not induce 
a change in the transfer of energy from 
PS II to I, but light II apparently did 
give rise to some type of pigment rear- 
rangements in PS II so that less energy 
was delivered to the reaction centers of 
PS II (Fig. 33, state III). It is also pos- 
sible that light energy arriving at the 
reaction centers of PS II is quenched by 
a different mechanism than the one in- 
volving Q (Duysens and Sweers, 1963) 
and that this mechanism is put into op- 
eration after a state II-state III tran- 
sition has taken place. 

Since Porphyra lives in the high in- 
tertidal zone, where it is exposed to high 
light intensities that can often occur in 
combination with desiccation and high salt 
concentrations, it appears that the state 
II-state III transition may provide a 
mechanism to eliminate excess light en- 
ergy and prevent damage to the photo- 
synthetic machinery by photoinhibition. 




Bonaventura, C, and J. Myers, Biochim. Biophys. 

Acta 189, 366-383, 1969. 
Duysens, L. N. M., and H. E. Sweers, in Studies 

on Microalgae and Photosynthetic Bacteria, 353- 

372, Japanese Society of Plant Physiologists, ed., 

University of Tokyo Press, 1963. 

Melis, A., and P. H. Homann, Photochem. Pho- 

tobiol. 23, 343-350, 1976. 
Murata, N., Biochim. Biophys. Acta 172, 242-251, 

Powles, S. B., and S. W. Thorne, Planta 152, 471- 

477, 1981. 
Wang, R., and J. Myers, Biochim. Biophys. Acta 

31,7, 134-140, 1974. 


Kazuhiko Satoh and David C. Fork 

As described in the previous report, 
excess light absorption by photosystem 
II in the red alga Porphyra perforata 
transforms it into a newly found state, 
termed state III. Ater conversion to this 
state, system I activity was unchanged, 
but there was a decrease in the light en- 
ergy reaching photosystem II. Light ab- 
sorbed by photosystem I was found to 
have a more antagonistic effect than sys- 
tem II light on the state II-state III 
transitions. The state II to III transition 
was related to the formation of ApH, and 
the state III to II transition was related 
to the formation of a membrane potential 
across the thylakoid membrane. 

The Porphyra used in the experiments 
here reported was collected locally and 
maintained as described in the previous 
report. Green light (light II) was ob- 
tained by filtering the white light from a 
150W, 21.5V projector lamp (DLS) 

TABLE 6. Effects of NH 4 C1, DCMU, DCCD, 

and Valinomycin on the State II to III Transition 

in Porphyra* 

through Corning glass filters 4-96 and 3- 
69 and a Calflex C heat-reflecting filter 
(Balzers). For light I, far red or blue light 
absorbed mainly by chlorophyll a of PS I 
was used. The output of the lamp was 
filtered through a Schott RG10 (X > 690 
nm) cutoff filter (for far red light) or 
through Corning filters 4-96 and 5-60 (for 


F695, relative extent 

No addition 
NH 4 C1 
Val. + KC1 



*The samples were illuminated with light II for 
3 min at 21°C, and the extents of the state II to 
III transition were measured by the decrease of 
F695 at 77 K. The concentrations of NH 4 C1, DCMU, 
DCCD, and Valinomycin and KC1 were 30 mM, 20 
[xM, 10 fxM, 10 ^M, and 10 mM, respectively. 

i—i — 1 — 1 — 1 — 1 — r~ 1 — 1 — 1 — " — 1 — 1 — 1 — 1 — 1 — 1 — 1 — r 


J I I L 

650 700 750 

Wavelength, nm 


Fig. 34. The continuation in the dark of the 
state II to state III transition after exposure of 
Porphyra perforata to green light absorbed by pho- 
tosystem II. Green light (235 |xW/cm 2 was given 
for 5 s (at 21°C), after which the sample was cooled 
to 77 K and the emission spectrum measured (solid 
curve) using the same green light attenuated to 58 
|xW/cm 2 . The dashed curve was obtained from a 
sample given 5 s of green light, 30 s dark, and then 
cooled to 77 K. 


i r 


4 6 

Time, min 

Fig. 35. The kinetics of the state III to II tran- 
sition in the dark and in light I. State III was 
obtained by illuminating Porphyra with light II at 
21°C for 3 min. The recovery to state II in the dark 
or in light I was then measured as the increase of 
F695 at 77 K. The intensity of the far red light was 
1.40 mW/cm 2 (K > 690 nm). 

blue light). A Nicolet Signal Averager 
(Model 1010) was used as a transient time 
converter for measurements of fluores- 
cence in the ms time range. 

We found in the previous report that 
Porphyra exists in state II in the dark. 
Conversion of Porphyra from state II to 
state III upon absorption of light II gives 
rise to a decrease of system II fluores- 
cence at 695 nm (F695). We measured 
fluorescence spectra at 77 K and studied 
the effects of several substances. The data 
presented in Table 6 show that the un- 
coupler of photophosphorylation NH 4 C1 
inhibited the state II to III transition. A 
similar effect was found using CCCP 
(carbonylcyanide m-chlorophenylhydra- 
zone) (not shown). DCMU also inhibited 
the transition. The inhibitor of ATP for- 
mation, DCCD (N,N'-dicyclohexyl car- 
bodiimide), and the ionophore valinomycin 
+ KC1 did not have an effect on the tran- 
sition. These results suggest that proton 
translocation across the thylakoid mem- 


700 750 

Wavelength, nm 


Fig. 36. The effects of light I on the fluores- 
cence spectra in Porphyra preilluminated with light 
II. The sample was preilluminated at 21°C with 
green light absorbed by PS II (G, dashed line) and 
cooled to 77 K. The solid line was obtained with a 
sample given 2 min of green light followed by 1 
min of blue light (662 fxW/cm 2 ) absorbed by pho- 
tosystem I (B, solid line). 

branes is related to the state II to III 

In order to convert dark-adapted Por- 
phyra from state II to state III, system 
II light was given for 5 s and the sample 
was then immediately cooled to 77 K (Fig. 
34, solid line). Another sample was also 
illuminated for 5 s, but it was then given 
a period of 30 s darkness before being 
frozen to 77 K (Fig. 34, dashed line). It 
is clear from this result that the state II 
to III transition proceeds in the dark af- 
ter illumination with light II. Ried and 
Reinhardt (1977) also observed a dark 
step in light II effects in other red algae. 

Figure 35 shows that the recovery from 
state III to II occurred slowly in the dark 
and that light I accelerated this transi- 
tion. It is unlikely that the increase of 
F695 shown in Fig. 35 is produced by a 
state II to I transition, since we saw that 
this transition occurred very slowly and 



required 30 min to show the maximum 
effect in Porphyra (see previous report). 
In order to check further whether the 
state III to II transition was accelerated 
by light I, we measured fluorescence 
spectra at 77 K. Figure 36 shows the flu- 
orescence spectrum obtained when Por- 
phyra was illuminated for 2 min with 
green light absorbed by system II and 
then frozen to 77 K (dashed line, labeled 
G). The solid line (B) of Fig. 36 repre- 
sents the spectrum obtained when Por- 
phyra was illuminated with green light 

for 2 min as before, but it was then ex- 
posed for 1 min to blue light absorbed by 
system I. The difference spectrum (B- 
G) shows that system I light brought about 
the state III to II transition and not the 
state II to I transition, since there was 
a parallel increase of both system II flu- 
orescence bands, at 685 and 695 nm, and 
an increase of the long- wavelength sys- 
tem I band — features characteristic of 
the state III to II transition (see Fig. 
30). By contrast, the state II to I tran- 
sition induces changes specific to one of 






o 12 



2 10 


_g -, 



a. No addition 


b. NH 4 CI 

30 sec 

c. Valinomycin + KCI 


o 1 - 




d. Antimycin A 

Light II off 
Light I on 

Light I off 
Light II on 

Fig. 37. The kinetics of the transitions from states II to III and from III to II in Porphyra at 21°C 
and the effects of NH 4 C1, valinomycin plus KCI, and antimycin A. The arrows indicate the times when 
light I and light II were switched. The concentrations of NH 4 C1, valinomycin, KCI, and antimycin A 
were 5 mM, 10 |xM, 10 mM, and 50 [xM, respectively. Blue light (662 jxW/cm 2 ) was used as light I. 


the two system II fluorescence bands to III transition continued until reaching 

(F695) and a decrease of system I flu- a certain level (Fig. 35). Prolonged stor- 

orescence. age in the dark brought about the slow 

Since light I accelerates the state III conversion back to state II. The accel- 
to II transition, an exposure to light I eration of the state III to II transition 
following an exposure to light II should by light seems to be related to the for- 
increase the fluorescence level excited by mation of a membrane potential change 
system II light. The fluorescence kinetics across the thylakoid membranes pro- 
given in Fig. 37 demonstrate these tran- duced by cyclic electron flow around PS I 
sitions and show the effects of various (Fig. 37), since antimycin A, an inhibitor 
substances. After 3 min illumination with of cyclic flow around PS I, strongly in- 
light II, the light was switched to light hibited the transition, as did the potas- 
I for 1 min and then switched back to siurn ionophore valinomycin. 
light II. The transitory high level of flu- Light I and light II have antagonistic 
orescence following an exposure to light effects on the state II to state III tran- 
I is usually considered to result from a sition (Fig. 37). Thus the antagonistic ef- 
state II to state I transition in light I fects of light I and II on the changes of 
(Year Book 81 , 45-49). But the transient chlorophyll fluorescence do not necessar- 
high fluorescence level following an ex- ily indicate that the change is related to 
posure to light I is an indication of the a state I to II transition, 
transition from state III to state II in It was shown in the previous article 
light I, as shown in Figs. 35 and 36. The that the state II to III transition reduced 
subsequent decline of fluorescence when the light energy reaching photosystem 
light I was turned off and light II was II. This process would take place when 
turned on indicates the state II to III there was an excess amount of light II. 
transition. Figure 37 also shows that the If, however, light I was present, the re- 
uncoupler NH 4 C1 did not prevent the state verse process (state III to II transition) 
III to II transition, but valinomycin + was accelerated. It would thus appear 
KC1 or antimycin A strongly inhibited that the state II to state III transition 
the change. may function to balance the light energy 

The state II to state III transition has reaching photosystems I and II and 

both a light and a dark process (Fig. 34). thereby avoid photodestruction by ab- 

Proton translocation across the thylakoid sorption of excess amounts of light II. 
membranes induced by light seems to 

promote the transition (Table 6). The pH p 
difference induced by the light reaction 

must decrease during the subsequent dark Ried, A., and B. Reinhardt, Biochim. Biophys. 

period. This dark process of the state II Acta * 60 > 25 " 35 > 1977 - 


Kazuhiko Satoh and David C. Fork 

Studies reported previously (Year Book priate activity ratio by controlling the 

81 , 45-58) described characteristics of the transfer of light energy from pigment 

state I-state II transitions. These tran- system II to system I. 
sitions appear to be mechanisms helping Since in its natural environment Por- 

the red alga Porphyra perforata to main- phyra is exposed periodically to desic- 

tain its two photosystems in an appro- cation and high light intensities as well 



as to high salinity, it can be expected that 
this alga might have developed several 
other coping mechanisms. The effects of 
salinity on the photosynthesis of this alga 
are discussed by Smith et al. (this Re- 
port, p. 68). A new mechanism, found in 
Porphyra and termed the state II to state 
III transition, is also discussed in this 
Report, by Fork and Satoh (p. 55) and 
by Satoh and Fork (p. 58). This mecha- 
nism apparently decreases the amount of 
light energy reaching the reaction cen- 
ters of PS II without causing any change 
in photosystem I (PS I) activity. Both 
mechanisms appear to protect the pho- 
tosynthetic apparatus against damage by 
photoinhibition when Porphyra is ex- 
posed to high light intensities. If Por- 
phyra becomes desiccated or exposed to 
high salt concentrations, conditions that 
occur commonly in the intertidal habitat, 
its water potential can become very low 
and the activity of PS II would be ex- 
pected to be strongly inhibited. This re- 
port describes a third mechanism by which 
Porphyra can avoid photodamage to its 
reaction centers under conditions of high 
light and low water potential. 

The Porphyra used in these studies 
was collected at Half Moon Bay, Califor- 
nia, and maintained in the laboratory as 
described by Fork and Satoh on p. 55. 
Samples were kept in the dark for 1 h or 
more before use. 

Figure 38 shows the kinetics of fluo- 
rescence when Porphyra was illumi- 
nated in the presence of DCMU with blue 
light absorbed by PS I and with green 
light absorbed by PS II. The blue and 
green actinic lights were adjusted ini- 
tially to excite PS II equally. The initial 
reduction of Q was too rapid to be seen 
clearly on the time scale used in Fig. 38. 
Surprisingly, after the initial, rapid flu- 
orescence increase attributable to Q re- 
duction (Duysens and Sweers, 1963), 
fluorescence during illumination de- 
creased more rapidly in blue light (PS I) 
than in green (PS II). This result sug- 
gests that PS I was responsible for this 
decrease. The decrease of fluorescence 
cannot be attributed to the reoxidation 





a. 433 nm light 

b. 550 nm light 

2 min 



Fig. 38. Time courses of Chi a fluorescence at 
685 nm in Porphyra perforata treated with 20 |xM 
DCMU and illuminated with system I (433 nm) and 
system II light (550 nm). The blue and green actinic 
lights were obtained using interference filters and 
had intensities of 27.6 and 4.4 jjiW/cm 2 , respec- 

of Q by a PS I reaction because, at a 
concentration of 20 (jlM DCMU, this re- 
action was completely inhibited, as was 
2 evolution. 

Porphyra was illuminated for 20 s and 
7 min, respectively, with relatively weak 
blue light, and fluorescence kinetics were 
then measured in each case using a strong 
actinic light and on a faster time scale 
than in Fig. 38. 

Figure 39 shows that 20 s of preillu- 
mination with PS I light in the presence 
of DCMU caused Q to become almost 
completely reduced, since the initial flu- 
orescence (F ) was high and there was 
very little variable fluorescence (F v ). This 
result indicates that Q was almost com- 
pletely reduced after the 20-s period of 
preillumination. But after 7 min of preil- 
lumination, Q was largely in its oxidized 



Fig. 39. Induction of Chi a fluorescence in P. perforata treated with DCMU as in Fig. 38 and measured 
after 20 s and 7 min of preillumination of the sample with blue light (433 nm). The actinic light to excite 
fluorescence was obtained from an He-Ne laser (632.8 nm) and had an intensity of 124 |xW/cm 2 . 

state, since F was low and F v large. It 
is important to note that both traces 
reached the same maximum fluorescence 
level. Thus the fluorescence decline dur- 
ing illumination in the presence of DCMU 
was produced by an oxidation of Q by a 
mechanism other than oxidation via plas- 
toquinone because this reaction was com- 
pletely inhibited by the DCMU. Since 
fluorescence spectra were the same be- 
fore and after the fluorescence decline, 
it is clear that neither a state I-state II 
transition nor a state I I-state III tran- 
sition was responsible for this effect. 

Three mechanisms may be responsible 
for the oxidation of Q: (1) oxidation by 
PS I via plastoquinone (not possible, 
however, because DCMU inhibits this 
reaction, as pointed out previously); (2) 
a back reaction of PS II where oxidation 
of Q is induced by oxidants produced on 
the water side of PS II (Bennoun, 1970; 
Homann, 1971; Ikegami and Katoh, 1973); 
or (3) donation of electrons from Q to 
unspecified redox substances. 

Figure 40 shows the results of an ex- 
periment designed to distinguish the rea- 

tion(s) responsible for Q oxidation in the 
presence of DCMU. Trace b of Fig. 40 
shows that NH 2 OH prevented Q oxida- 
tion in the presence of DCMU. If Q were 
being oxidized by other unspecified re- 
dox substances (mechanism 3, previous 
paragraph), then this oxidation should be 
increased by NH 2 OH since NH 2 OH at 1 
mM is an inhibitor of PS II (Bennoun and 
Joliot, 1969; Izawa et al., 1969). When 
functioning, NH 2 OH keeps Q reduced. 
The uncoupler, CCCP (carbonylcyanide 
m-chlorophenylhydrazone), strongly in- 
hibited the fluorescence decline (Fig. 40c). 
At the concentration of CCCP used (5 
|jlM), it acts as an uncoupler of photo- 
phosphorylation. Again, the oxidation of 
Q seems not to proceed via activity of 
PS I (mechanism 1) because the uncou- 
pling action of CCCP would be expected 
to accelerate rather than retard the re- 
action. We also tested the effects of CCCP 
on the back reaction of PS II (mechanism 
3) and found that low concentrations of 
CCCP and NH 2 OH inhibited this reac- 
tion, as was noted previously by Homann 
(1971) and Ikegami and Katoh (1973). 




< 1 1 1 1 — 

a. Control 

i i i 
b. NH 2 OH 

























d Antimycin A 






















1 2 3 1 2 3 

Time, min 

Fig. 40. Effects of NH 2 OH, CCCP, and anti- 
mycin A on the fluorescence decline in the light in 
the presence of DCMU in P. perforata. Blue actinic 
light (433 nm) was used to excite PS I. The con- 
centrations of DCMU, NH 2 OH, CCCP, and anti- 
mycin A were 20 |xM, 1 mM, 5 |xM, and 50 |xM, 

Figure 40d shows that antimycin A, 
which inhibits cyclic flow around PS I 
(Tagawa et al., 1963), strongly acceler- 
ated the fluorescence decline in the pres- 
ence of DCMU. Inhibition of cyclic 
electron flow may give rise to an accu- 
mulation of either oxidants or reductants 
produced by PS I. 

Since the results of Figs. 38 and 40 
suggested that PS I-produced oxidants 
or reductants accelerated the back re- 
action of PS II, we tested the effect of 
artificial electron donors to PS I on the 

fluorescence decline. Reductants such as 
reduced DCIP (2,6-dichlorophenol indo- 
phenol) or DAD (diaminodurol) de- 
creased the rate of the fluorescence decline 
in the light. This result suggests, there- 
fore, that an oxidant produced by PS I 
may give rise to an acceleration of the 
back reaction of PS II. 

These results indicate that still an- 
other mechanism is available to Por- 
phyra to consume the excess light energy 
absorbed by PS II: the acceleration of 
the back reaction of PS II. Such a mech- 
anism could serve to prevent the for- 
mation of oxidants with high redox 
potentials in desiccated plants when ox- 
idation of water is impossible. This mech- 
anism, as well as other mechanisms 
reported previously (Year Book 80, 39- 
43; Fork and Satoh, and Satoh and Fork, 
this Report), apparently serve to pre- 
vent photoinhibition under unfavorable 
physiological conditions. We can specu- 
late that cytochrome 6 559 may be the sub- 
stance controlling the back reaction of 
PS II, since it is located close to the re- 
action center of PS II, and it can be pho- 
tooxidized by both PS II and PS I. 
Further experiments will be necessary 
to verify this suggestion. 


Bennoun, P., Biochim. Biophys. Acta 216, 357- 

363, 1970. 
Bennoun, P., and A. Joliot, Biochim. Biophys. Acta 

189, 85-94, 1969. 
Duysens, L. N. M., and H. E. Sweers, in Studies 

on Microalgae and Photosynthetic Bacteria, 353- 

372, Japanese Society of Plant Physiologists, ed. , 

University of Tokyo Press, 1963. 
Homann, P. H., Biochim. Biophys. Acta 2^5, 129- 

143, 1971. 
Ikegami, I., and S. Katoh, Plant Cell Physiol. Ik, 

837-850, 1973. 
Izawa, S., R. L. Heath, and G. Hind, Biochim. 

Biophys. Acta 180, 388-389, 1969. 
Tagawa, K., H. Y. Tsujimoto, and D. I. Arnon, 

Proc. Nat. Acad. Sci. USA k9, 567-572, 1963. 




EMISSION BAND OF Porphyra perforata AND ITS 


Prasanna Mohanty and David C. Fork 

Most fluorescence at room tempera- 
ture emanates from photosystem II 
(PS II), while photosystem I (PS I) is very 
weakly fluorescent (Goedheer, 1972). Fork 
et al. (1982) reported an intense long- 
wavelength emission band at room tem- 
perature in the marine red alga Por- 
phyra perforata, produced upon excitation 
of either PS I or PS II pigments. 

In this study, we examined in some 
detail the changes in the yield and emis- 
sion of this long- wavelength emission band 
at 715-730 nm in P. perforata and stud- 
ied its possible associations with PS II 
and PS I. 

P. perforata was collected from Jan- 
uary to May at the Hopkins Marine Sta- 
tion, Pacific Grove, California, and at 
Pigeon Point, California, and was main- 
tained in the laboratory as described ear- 
lier (Fork et al., 1982). Fluorescence 
emission spectra and transients were 
measured according to Fork et al. (1982). 
Delayed luminescence (DL) was mea- 
sured in the ms time range. The cycle 
and the duration of the actinic light of 
the phosphoroscope were 5.9 and 0.8 ms, 
respectively. Light from an He-Ne laser 
(632.8 nm) was used as the actinic light. 

Fluorescence emission of Porphyra at 
room temperature. Figure 41 shows the 
emission spectra obtained upon illumi- 
nation of Porphyra with (A) low-inten- 
sity green light (550 nm) or (B) blue light 
(442 nm). Excitation with green light in 
the absence of DCMU (Fig. 41, curve a) 
induced a peak emission at 655 nm (al- 
lophycocyanin). A small hump was seen 
near 683 nm that probably represents the 
composite emission from allophycocyanin 
and Chi a. A long-wavelength emission 
of Chi a appeared as a broad shoulder 
around 720 nm. Addition of 50 \xM DCMU 
(Fig. 41, curve b) enhanced the fluores- 

cence yield at 683 nm approximately five- 
fold; this emission became the dominant 
peak in the spectrum. The long-wave- 
length band was also increased approx- 
imately threefold upon closure of PS II 
traps after the addition of DCMU. Curve 
d of Fig. 41 shows an emission spectrum 



700 750 

Wavelength, nm 



Fig. 41. Room-temperature emission spectra 
of Porphyra perforata excited with green (A) or 
blue actinic (B) light in the presence or absence of 
50 |xM DCMU. Green light was obtained by using 
a Corning glass filter CS 4-96 in combination with 
a 550-nm interference filter having a half-band- 
width of 5 nm. The green light had an intensity of 
5.6 |xmol quanta/m 2 s. The blue light (442 nm) was 
obtained from an He-Cd laser (6.75 fimol quanta/ 
m 2 s). The sample was incubated with DCMU for 
4 min in the dark and measured at the same sen- 
sitivity that was used for spectra obtained with 
green light. 



obtained by using monochromatic blue 
light (442 nm) absorbed mostly by Chi a. 
Again, low-intensity excitation was used 
such that all PS II reaction centers re- 
mained open. This spectrum has the two 
characteristic peaks at 683 and 725 nm, 
and the ratio of long- wavelength to short- 
wavelength (F725/F683) bands was about 
1.45. Addition of DCMU (Fig. 41, curve 
e) increased the fluorescence yield 3.75- 
fold at 683 nm and 2.5-fold at 725 nm. 
The differences in the emission spectra 
produced by DCMU show that the long- 
wavelength emission was relatively higher 
in blue light (Fig. 41, curve f) than in 
green light (Fig. 41, curve c). It is clear 
from these results that Porphyra exhib- 
its an appreciable amount of a long- wave- 
length Chi a emission whose yield 
increases significantly upon closure of 
PS II traps after DCMU addition. 

Since the long-wavelength emission 
band and the short-wavelength 683-nm 
emission band showed similar sensitivity 
to DCMU, especially when excited with 
blue light, we examined the effect of the 
inhibitor DBMIB (2,5-dibromo-3-methyl- 
6-isopropyl-p-benzoquinone), a quencher 
of Chi a fluorescence, on the room-tem- 
perature emission spectrum of DCMU- 
treated Porphyra. The difference spec- 
trum obtained using blue excitation (Fig. 
42) shows that DBMIB mainly quenched 
the fluorescence at 683 nm and had very 
little effect on the long- wavelength band. 
Thus DBMIB seems to affect preferen- 
tially the PS II-sensitized variable flu- 
orescence at 683 nm. 

Fluorescence and delayed lumines- 
cence transients. Most algae exhibit flu- 
orescence transients (Kautsky effect) 
when dark-adapted cells are excited with 
bright PS II light. Porphyra, like other 
algae, shows complex transients depend- 
ing on the physiological state and length 
of dark adaptation (Satoh and Fork, 1983). 
Fluorescence-yield changes measured at 
685 or 725 nm showed similar Kautsky 
transients (data not shown). Addition of 
DCMU increased the yield both at 685 
and at 725 nm. We found that the flu- 
orescence transients measured at 725 nm 


700 750 

Wavelength, nm 


Fig. 42. Fluorescence emission spectra at room 
temperature of Porphyra treated with DCMU 
measured in the presence (dashed line) or absence 
(solid line) of 10 |xM DBMIB. The sample was ex- 
cited with blue actinic light (442 nm; 342 jxmol quanta/ 
m 2 s). The lower part of the figure shows the dif- 
ference in the spectrum that was induced by DBMIB. 

were similar whether they were obtained 
using blue actinic light absorbed mostly 
by Chi a of PS I, red (633-nm) light ab- 
sorbed mostly by phycocyanin, or green 
light (PS II) absorbed mostly by phy- 
coerythrin (data not shown). 

Millisecond-delayed luminescence (DL) 
originates from re-excitation of PS II an- 
tenna Chi a as a consequence of charge 
recombination at the PS II reaction cen- 
ters (Govindjee and Jursinic, 1979). As 
with Chi a fluorescence transients, dark- 
adapted Porphyra exhibits complex DL 
kinetics. Figure 43 shows the induction 
kinetics of DL measured at 685 nm and 
at 730-735 nm when Porphyra was ex- 
cited with red light (632.8 nm) in the 
presence or absence of DCMU. The ki- 
netics of DL observed at the long-wave- 
length band (720-730 nm) were similar 
to those observed at 685 nm, and the 
extent of DL measured in the 730-nm 
region was appreciable even in the pres- 
ence of DCMU. We have observed sim- 
ilar DL transients using blue (442-nm) 



at 735 nm 

Time, s 

Fig. 43. Kinetics of delayed luminescence mea- 
sured at 685 and 735 nm in Porphyra in the 
(A) absence or (B) presence of 50 \xM DCMU. The 
sample was kept in darkness for 5-6 min before 
illumination. DL was detected using interference 
niters that had half-bandwidths of 5 nm. The ex- 
citation light (632.8 nm) was obtained from an He- 
Ne laser (Spectra Physics, Model 124B) and had 
an intensity of 382 (xmol quanta/m 2 s. 

excitation (data not shown). Thus light 
observed mostly by PS II or PS I pro- 
duced DL at the long-wavelength emis- 
sion band of Porphyra. 

The long-wavelength fluorescence 
emission band in Porphyra is quite broad; 
we found that its relative magnitude var- 
ies depending upon growth conditions. 
This broad band suggests that it is com- 
prised of more than one component and 
that the relative intensity varies as a re- 
sult of variations in the amounts of dif- 
ferent components. Goedheer (1981) has 
reported two forms of long- wavelength- 
emitting Chi a in cyanobacteria, one of 
which is closely linked to PS II. It ap- 
pears that Porphyra, and possibly the 
other marine red algae, also possess such 
forms of long- wavelength-emitting Chi a 
forms associated with PS II. 

The similarity in the fluorescence ki- 
netics measured at the long-wavelength- 
emission band with those at 685 nm, and 
the extent of fluorescence-yield enhance- 
ment upon trap closure (Fig. 41) in blue 
light absorbed mostly by PS I, indicate 
that the long-wavelength emission pos- 
sesses a component which is linked to the 
PS II antenna pigments. This idea is fur- 
ther supported by observation of a sim- 
ilarity in the kinetics at the long 
wavelengths and at 685 nm in the ms DL 
originating from charge recombination at 
the reaction centers of PS II (Fig. 43). 
The differential sensitivity of the 683 and 
725 nm bands to DBMIB quenching, 
however, suggests that the long-wave- 
length emission also possesses a major 
component of Chi a of PS I, as suggested 
by Fork et al. (1982). 

Although the spectral distinction be- 
tween these two (or more) forms of the 
long-wavelength-fluorescing forms of Chi 
a in Porphyra remains to be elucidated, 
the results presented in this report argue 
in favor of the existence of a long- wave- 
length-emitting PS II band in the inter- 
tidal alga Porphyra. The presence of such 
a far red-emitting form of Chi a asso- 
ciated with PS II suggests that it may 
function as an antenna by way of reverse 
("uphill") energy transfer and/or as a 
protective device against photooxidation 
of PS II centers in high light during pe- 
riods of desiccation and other stresses, 
since excess light energy absorbed by the 
PS II reaction centers would be funneled 
to the long- wavelength-emitting forms. 


Fork, D. C, G. Oquist, G. Hoch, Plant Sci. Lett. 

2U, 249-254, 1982. 
Goedheer, J. C, Annu. Rev. Plant Physiol. 23, 

87-112, 1972. 
Goedheer, J. C, Photosynthesis Res. 2, 49-60, 1981. 
Govindjee, and P. A. Jursinic, Photochem. Pho- 

tobiol. Rev., Vol. 4, 125-205, K. E. Smith, ed., 

Satoh, K., and D. C. Fork, Biochim. Biophys. Acta 

722, 190-196, 1983. 



THE RED ALGA Porphyra perforata BY SALINITY 

Celia M. Synith, Kazuhiko Satoh, and David C. Fork 

The intertidal marine red alga Por- 
phyra perforata lives in an environment 
that, for most plants, is considered ex- 
tremely unfavorable for efficient photo- 
synthesis. Porphyra can tolerate severe 
drying whereby up to 91% of its water 
is lost (Smith, 1983). Under these con- 
ditions the photosynthesis of this alga can 
be completely inhibited. The alga also 
tolerates high light intensities and high 
salinity (from evaporative water loss). 
Even when exposed to high light inten- 
sities and when its photosynthesis is in- 
hibited by desiccation, Porphyra is able 
to avoid permanent damage from pho- 

Earlier studies on Porphyra dealt with 
the effects of desiccation on the photo- 
chemical reactions of photosynthesis (Year 
Book 72, 384-388) and on excitation en- 
ergy distribution (Year Book 80, 39-43). 
Other studies reported in this Report 
(Fork and Satoh, Satoh and Fork, Mo- 
hanty and Fork) describe a newly found 
mechanism available to this alga for 
adapting to changes of light intensity and 

In the present work we investigated 
the effects of salinity on the primary pro- 
cesses of photosynthesis in Porphyra. By 
varying the salinity from normal sea water 
to saturating concentrations of NaCl we 
could impose an osmotic stress that mim- 
icked the effects of air-drying, and thereby 
attain discrete levels of dehydration. 

At least three sites of photosynthesis 
were affected by high salinity. The site 
most strongly affected was the photoac- 
tivation of electron flow on the reducing 
side of photosystem I (PS I). 

The algae used in these experiments 
were collected at Monterey Bay, Cali- 
fornia, and maintained in open dishes of 
sea water at 13°C under illumination (8 
fxmol quanta/m 2 s). Samples were dark- 

incubated for 1 h or more in normal sea 
water and then incubated for 5 min in 
normal sea water or in NaCl solutions of 
various concentrations. 

As shown in Fig. 44, Porphyra in nor- 
mal sea water exhibits a typical fluores- 
cence time course (Kautsky transient) 
with several maxima and minima (des- 
ignated 01 DPS by Govindjee and Pa- 
pageorgiou, 1971). The to I rise is 
thought to represent the reduction of Q 
by photosystem II (PS II) (Duysens and 
Sweers, 1963; Kautsky et al., 1960). The 
dip (D) has been proposed to be produced 
by photooxidation by PS I of Q that was 

a, control 

b, 70 %o 


c, 115 %o 

d, 160 %o 



v 12 



Fig. 44. Effects of salinity on the time course 
of chlorophyll fluorescence in dark-adapted Por- 
phyra. The samples were incubated (a) in normal 
sea water, (b) in 70% o NaCl, (c) in 115%o NaCl, or 
(d) in 160%c NaCl for 5 min, and fluorescence time 
courses were measured under the same salinity. 
Low-intensity actinic light (632.8 nm), 55 (xW/cm 2 , 
was used for excitation. 



initially reduced by PS II (Schreiber and 
Vidaver, 1974; Satoh and Katoh, 1981). 
The DPS transient is related to elec- 
tron flow on the reducing side of PS I, 
since it can be eliminated by methyl viol- 
ogen, which accepts electrons from PS I 
(Munday and Govindjee, 1969; Satoh et 
al.j 1977). The DP rise was proposed to 
result from accumulation of photored- 
uced Q caused by the presence of a dark- 
inactivated site on the reducing side by 
PS I (Satoh et aL, 1977). Photoactivation 
of this site (probably ferredoxin-NADP + 
reductase, see Satoh, 1981) gives rise to 
the PS decline. The DP rise, therefore, 
is a reflection of the activity of PS II, and 
the PS decline in fluorescence corre- 
sponds to the photoactivation of PS I that 
was previously dark-inactivated. 

The fluorescence kinetics shown in Fig. 
44b demonstrate that the D to P rise is 
decreased by doubling the NaCl concen- 
tration of normal sea water from 32 to 
70 parts per thousand (%o). This result 
suggests that PS II activity was re- 
tarded by this treatment. (See also Wil- 
tens et aL, 1978.) Increasing the NaCl 
concentration did not seem to cause a 
further inhibition of the DP rise, but the 
P to S decline started to slow down at 
about 70 and stopped at 160% o NaCl. The 
I to D dip disappeared between 70 and 
115%o NaCl. This latter change may re- 
sult from a salt-induced retardation of 
electron flow between the two photosys- 
tems as suggested by Wiltens et aL (1978). 

Figure 45 presents a summary of data 
obtained on the effect of NaCl on fluo- 

40 80 120 160 200 240 

NaCl, %o 

Fig. 45. The effects of salinity on the rate of the P to S decline and on the extents of the I to P rise 
(triangles and circles) of the fluorescence transients in dark-adapted Porphyra. The samples were in- 
cubated with various concentrations of NaCl for 5 min, and fluorescence time courses were measured 
under low light (same conditions as for Fig. 44) and high light (235 |xW/cm 2 ). Circles and triangles were 
measured under the low light, and squares under high light. 



rescence time courses. The PS decay 
(photoactivation of ferredoxin-N ADP + 
reductase) was decreased by one-half at 
a salt concentration of around 100%c and 
was zero at 160%c. 

As mentioned earlier, the level of the 
peak (P) is determined by two factors: 
reduction of Q by PS II, and oxidation 
of Q by PS I. The effect of NaCl concen- 
tration on the difference between fluo- 
rescence at the P and I levels was 
measured at low or high light intensity. 
At a low light intensity, as used for Fig. 
45 (triangles) and for Fig. 44, the F P - Fj 
extent was already noticeably reduced 
by an NaCl concentration of 44%o. With 
higher light intensities (Fig. 45, squares), 
the decrease of F P - Fj was seen at higher 
NaCl concentrations. This difference may 
occur because photoactivation of electron 
flow on the reducing side of PS I is slower 
at the lower light intensity. In this case 
photoactivation of PS I proceeds during 
the course of the reduction of Q, so that 
the level of P is highly sensitive to PS II 
activity. Since the I to P rise decreased 
and then increased with increasing salt 
concentrations, it is clear that salt inhib- 
its electron flow on the water side of PS II 
and inhibits even more strongly electron 
flow on the reducing side of PS I by sup- 
pressing the photoactivation reaction. 

Figure 46 shows the effects of NaCl 
concentration on the time courses of P700 
oxidation (a PS I reaction). Illumination 
of dark-adapted Porphyra in sea water 
produced a rapid oxidation of P700 (de- 
crease of absorbance). In this sample, 
electron flow on the reducing side of PS I 
would be initially dark-inactivated. 
Therefore, when the electron carriers 
between the inactivated site and the re- 
action centers of PS I are fully reduced, 
PS II becomes inactive. This results in a 
photoreduction of P700 by PS II (Fig. 46, 
trace a, increase of absorbance following 
initial decrease). After some period of 
illumination, photoactivation of the dark- 
inactivated site took place and photoox- 
idation of P700 occurred once again. The 
rate of the dark reduction of P700 was 
determined by the activity of PS II. At 

a, control 

b, 115 %o 

c, 160 %o 

Fig. 46. Effects of salinity on the time course 
of photooxidation of P700 in dark-adapted Por- 
phyra. Two layers of Porphyra were used. The 
P700 measuring beam was chopped at a frequency 
of 170 Hz and the oscillating signals were detected 
with a lock-in amplifier (PARC EG & G 128A). A 
continuous blue actinic light was obtained using 
Corning glass filters 4-96 and 5-60, and had an in- 
tensity of 331 |xW/cm 2 . For explanation of traces 
a, b, c, see text. 

a high salt concentration the initial tran- 
sient P700 oxidation occurred as in nor- 
mal sea water but the re-oxidation during 
illumination and dark reduction of P700 
were both inhibited (Fig. 46b). At 160%o 
NaCl, photoactivation was almost com- 
pletely inhibited and P700 stayed re- 
duced in the light (Fig. 46c). These results 
also suggest that electron flow on the 
water side of PS II and photoactivation 
of electron flow on the reducing side of 
PS I were inhibited by high salinity, the 
latter being the most strongly inhibited. 
We measured a DPS transient in nor- 
mal sea water and, after a dark interval 
of 5 min, measured the DPS transient 
again. During this period, dark inacti- 
vation occurred so that the extent of the 
second DPS transient was little more than 



50% of the transient seen in the com- 
pletely dark-adapted sample. If the sam- 
ple was placed in 160%o NaCl during the 
5-min dark interval, there was almost no 
DPS transient (data not shown). This re- 
sult suggests that dark inactivation of 
electron flow on the reducing side of PS I 
is inhibited by high salinity. In these ex- 
periments the steady-state levels of flu- 
orescence were about the same for both 
samples, which indicates that the high 
salinity did not interfere with electron 
flow on the reducing side of PS I. 

By measuring fluorescence induction 
in the presence of DCMU, we found that 
high salinity has another effect on the 
primary processes of photosynthesis. 
Figure 47 shows that a high concentra- 
tion of salt did not affect the initial (F ) 
fluorescence level. It did, however, de- 
crease the variable fluorescence level (F v ) 
and retard the rate of reduction of Q. The 
ratio of fluorescence at its maximum level 
(F m ) to its F level decreased linearly 
with increasing salinity (data not shown). 
This result could be explained by assum- 
ing that the back reaction of PS II is ac- 
celerated (re-oxidation of Q by electron 

donation to an electron acceptor on the 
water side of PS II). However, this pos- 
sibility does not seem likely because we 
found that Q oxidation was slowed rather 
than accelerated by high salinity. 

It seems likely that high salinity gives 
rise to a decrease of the light energy that 
is transferred to the reaction centers of 
PS II, since these algae were shown to 
have an increased energy transfer from 
PS II to I under adverse conditions (Year 
Book 79, 193-197; 80, 39-43). A decrease 
of the quantum yield of the photochem- 
ical activity of the reaction center of PS II 
itself by high salinity could not be ruled 
out by the present experiments. 

The data presented here show that 
there are at least three sites in the pri- 
mary processes of photosynthesis that are 
inhibited by high salinity: photoactiva- 
tion of electron flow on the reducing side 
of PS I, electron flow on the water side 
of PS II, and transfer of light energy be- 
tween pigment molecules. Perhaps an in- 
hibition of electron flow at the different 
sites is necessary to prevent permanent 
damage to Porphyra by photoinhibition, 
since electron-flow inhibition is known to 

40 60 80 
Time, ms 

100 120 140 

Fig. 47. The induction of chlorophyll fluorescence in the presence of DCMU in Porphyra in normal 
sea water or in 250%c NaCl. The samples were incubated in normal sea water or in 250%c NaCl for 5 
min in the presence of 20 |xM DCMU, and the time courses of chlorophyll fluorescence were measured. 
Arrows indicate the half-increase times. The green actinic light was obtained using Corning glass filters 
4-96 and 3-69, and Calflex-C heat-reflecting filter, and had an intensity of 235 |xW/cm 2 . 



occur when only one of these sites is in- 
hibited. Inhibition of electron flow at three 
sites may help to prevent accumulation 
of large amounts of free reductants with 
very low redox potentials generated by 
PS I and of oxidants with very high re- 
dox potentials from PS II, both of which 
can quickly destroy the photosynthetic 
apparatus. A decrease in the light energy 
reaching the reaction center of PS II may 
also be a useful mechanism to avoid pho- 


Duysens, L. N. M., and H. E. Sweers, in Studies 
on Microalgae and Photosynthetic Bacteria, 353- 

372, Japanese Society of Plant Physiologists, ed., 

University of Tokyo Press, 1963. 
Govindjee, and G. Papageorgiou, in Photophysiol- 

ogy, Vol. 6, 1-46, A. C. Giese, ed., Academic 

Press, New York, 1971. 
Kautsky, H., W. Appel, and H. Amann, Biochem. 

Z. 332, 277-292, 1960. 
Munday, J. J., and Govindjee, Biophys. J. 9, 1- 

21, 1969. 
Satoh, K., Biochim. Biophys. Acta 638, 327-333, 

Satoh, K, and S. Katoh, Plant Cell Physiol. 22, 

11-21, 1981. 
Satoh, K., A. Yamagishi, and S. Katoh, in Pho- 
tosynthetic Organelles, Special Issue of Plant 

and Cell Physiol., 75-86, Japanese Society of 

Plant Physiologists, ed., University of Tokyo 

Press, 1977. 
Smith, C. M., Ph.D. Thesis, Stanford University, 

Schreiber, U., and W. Vidaver, Biochim. Biophys. 

Acta 368, 97-112, 1974. 
Wiltens, J., U. Schreiber, and W. Vidaver, Can. 

J. Bot. 56, 2787-2794, 1978. 


OF Ca 2 + 

Jerry J. Brand, Prasanna Mohanty, and David C. Fork 

Intact cells of Anacystis nidulans lose 
the ability to evolve 2 when Ca 2+ is 
depleted from the growth medium (Becker 
and Brand, 1982). The Ca 2+ -depleted cells 
do not photoreduce benzoquinone cou- 
pled to 2 evolution, but do carry out 
photooxidation of reduced diaminodu- 
rene (DAD) in the presence of methyl 
viologen. Thus Ca 2+ depletion specifi- 
cally inhibits photosystem II (PS II). 
Since the addition of Ca 2+ to the de- 
pleted cells restores the capacity to evolve 
2 it appears that Ca 2+ is required for 
the functioning of PS II. In this report 
we provide evidence that Ca 2+ ions are 
involved in the primary photochemical 
charge separation of PS II reaction cen- 
ters of A. nidulans. 

A. nidulans was grown photoauto- 
tropically in continuous culture at 39°C. 
Cells at late log phase were harvested 
and transferred to Cg-10 medium of Van 

Baalen (1967) lacking Ca 2+ and supple- 
mented with 1.5 times the normal amount 
of glycylglycine. As Ca 2+ depletion re- 
quires light, the cells were incubated in 
red light (300 |jimol, m~ 2 s -1 ) for several 
hours. Control cultures were treated 
identically except they contained 0.35 mM 
Ca(N0 3 ) 2 . Aliquots of Ca 2+ -depleted and 
control cells were taken for 2 evolution 
and fluorescence assays. 

Chlorophyll a (Chi a) fluorescence-yield 
changes are indicative of photochemical 
functioning of PS II. The upper traces of 
Fig. 48 show that control cells exhibited 
typical Kautsky transients when dark- 
adapted cells were illuminated with bright 
monochromatic blue light (442 nm). We 
note that the fast Kautsky (OIDPS) tran- 
sient is followed by a slow S-to-M rise in 
fluorescence yield in control cells (Mo- 
hanty and Govindjee, 1973). The Ca 2+ - 
depleted cells, however, did not show any 











9 + 

p Ca^ depleted 



Ca^ depleted 



Time, s 

Fig. 48. Time course at 39°C of Chi fluorescence measured at 685 nm in control or Ca 2 + -depleted 
cells of A. nidulans illuminated with monochromatic blue light (442 nm) in the absence (upper traces) 
or presence (lower traces) of DCMU (20 jjlM). Cells were incubated for 8 h in red light in the presence 
(solid line) or absence (dashed line) of Ca 2 + . Curves were normalized to correct for small differences in 
Chi concentration between control and Ca 2 + -depleted cells. The arrow indicates the time when illumi- 
nation began. Chi a = 4-6 |xg/ml. 

fluorescence transient; the fluorescence 
yield remained invariable throughout the 
period of illumination. The lower traces 
of Fig. 48 show the fluorescence tran- 
sients of control and Ca 2+ -depleted cells 
in the presence of the inhibitor DCMU. 
The control cells showed a rise in fluo- 
rescence from a low (F ) level to a max- 
imal (F m ) level upon illumination with light 
absorbed almost exclusively by Chi a, 
while Ca 2+ -depleted cells showed no time- 
dependent change in fluorescence yield. 
The fluorescence yield remained at the 
F level. Table 7 shows that the time- 
dependent loss of 2 evolution parallels 
the loss of variable fluorescence (F m —F ), 
as a result of the withdrawal of Ca 2 + 
from the growth medium. These results 
thus suggest that Ca 2+ depletion stops 
the flow of electrons from H 2 to PS II. 
Hydroxylamine (NH 2 OH) at high con- 
centration feeds electrons near the re- 
action center of PS II. Addition of 10 
mM NH 2 OH induced a rapid rise of Chi 
a fluorescence in control cells, but it did 
not change the fluorescence yield of Chi 

a in Ca 2 + -deprived cells (Fig. 49). These 
results suggest that Ca 2+ depletion 
probably does not affect the 2 -evolving 
system of Anacystis but does affect the 
photochemical reaction of PS II itself. The 
loss of delayed luminescence measured 
in the millisecond region (Table 8) also 
confirms the suggestion that Ca 2+ de- 
pletion interferes with the primary pho- 
tochemistry of PS II. 

TABLE 7. Progressive Loss of 2 Evolution 

Activity and Variable Fluorescence Yield in 

Ca 2 + -depleted Cells* 

Time (h) 

o 2 


(% of Control) 


% of Control 













* Calcium depletion ad fluorescence measure- 
ments were done as described in Fig. 48. Chi con- 
centration was 3 |xg ml" 1 . Rates of 2 evolution 
expressed in fxmol 2 -mg Chi" 1 h _1 . 2 evolution 
was measured at 39°C in saturating red light. F v 
was calculated from the fluorescence transients as 
the difference in yield of F„, and F . 



-NH 2 OH 

/\^^~-^- ____ *NH 2 OH 
// *NH 2 OH 

+ Ca 2+ 
- - r.2 + 

. — 








^^ -NH 2 OH 

15 sec 






Fig. 49. Time course of Chi fluorescence in presence or absence of NH 2 0H. Control and Ca 2 + -depleted 
cells (20% of control 2 evolution) were illuminated, and fluorescence was measured, as described in Fig. 
48. NH 2 0H (10 mM) was added directly to the sample and equilibrated in complete darkness prior to 
fluorescence measurements. Solid lines, no addition; dotted lines, NH 2 0H addition. 

Fluorescence-yield changes at 77 K re- 
flect the photoreduction of the primary 
acceptor of PS II (Butler, 1978). In Fig. 
50, it can be seen that illumination of 

TABLE 8. Effect of Ca 2+ Depletion on Delayed 
Fluorescence (DF) Emission in A. nidulans* 



F v 


% of Control 

% of Control 









*The ms-delayed fluorescence was measured at 
685 nm using an He-Ne laser (Spectra-Physics Model 
124B) and a phosphoroscope, the rotating sector of 
which provided alternating light-dark cycle of 0.8 
and 5. 1 ms. The resulting signal was amplified by 
a lock-in amplifier tuned to the chopper. Height of 
the transient peak was measured and expressed as 
percentage of the height in control cells of the same 
chlorophyll concentration. Rates of 2 evolution 
given in (xmol 2 -mg Chi -1 h _1 . 

control cells, frozen in the dark to 77 K, 
caused an increase in fluorescence yield 
measured at 695 nm. However, in Ca 2 + - 
depleted cells the fluorescence yield re- 
mained invariable. The extent of variable 
fluorescence yield measured at 77 K di- 
minished in parallel with the decrease in 
2 evolution (Table 9). These results 
strongly suggest that Ca 2+ plays a role 
in the primary photochemical charge sep- 
aration of PS II in Anacystis. 

Brand (1979) has previously shown that 
in order to obtain membrane fragments 
with high 2 -evolving capacity from A. 
nidulans, the cells required Ca 2+ during 
French-press breakage. Recently En- 
gland and Evans (1983) showed that Ca 2+ 
addition enhanced the photooxidation of 
the artificial electron donor diphenylcar- 
bazide (DPC) by Anacystis submem- 
brane preparations pretreated with Pb 2+ , 
the latter treatment being known to in- 
hibit electron flow from H 2 to PS II. 



■■ ' ! 




Ca depleted 






i i i 

10 20 30 

Time, s 


Fig. 50. Time course at 77 K of Chi fluores- 
cence at 695 nm in A. nidulans. Cells were incu- 
bated in red light for 8 h in the presence (solid line) 
or absence (dashed line) of Ca 2 + , during which time 
the 2 -evolving capacity of the minus-Ca 2+ cells 
decreased to zero. Fluorescence was measured in 
undiluted cells in growth medium with no other 
addition. Fluorescence was excited by blue-green 
light (30 jxmol m" 2 s" 1 ). Samples were kept in the 
dark for 10 min prior to cooling to 77 K for the 
fluorescence measurements. 

Our results with whole cells indicate that 
the Ca 2+ requirement in isolated mem- 
brane preparations is a physiological one, 
and that the site of Ca 2 + function is near 
the reaction center of PS II. It appears 
likely that Ca 2+ is required for the initial 
charge-separation reactions of PS II. 

TABLE 9. Progressive Loss of 2 Evolution 

Activity and Variable Fluorescence at 77 K in 

Control and Ca 2 + -depleted Cells* 

o 2 


F v 

Time (h) 


% of Control 

% of Control 


















*Calcium-depletion and fluorescence experi- 
ments were done as described in Fig. 49. Chi con- 
centration was 5 ixg ml _1 . Rates of 2 evolution 
given in jxmol 2 -mg Chi" 1 h" 1 . 2 measured in a 
Clark-type electrode at 39°C, and the samples were 
illuminated with saturating red light (1000 jxmol 
m" 2 s _1 ). F v was calculated as F m (all traps closed) 
- F (all traps open) (see Fig. 50) from the fluo- 
rescence transient measured at 77 K. 


Becker, D. W., and J. J. Brand, Biochem. Bio- 

phys. Res. Commun. 109, 1134-1139, 1982. 
Brand, J. J., FEBS Lett. 101, 110-112, 1979. 
Butler, W. L., Annu. Rev. Plant Physiol. 20, 345- 

378, 1978. 
England, R. R., and E. H. Evans, Biochem. J. 

210, 473-476, 1983. 
Mohanty, P., and Govindjee, Plant Cell Physiol. 

U, 611-629, 1973. 
Van Baalen, C, J. Phycol. 3, 154-157, 1967. 



Prasanna Mohanty, David C. Fork, and Jerry J. Brand 

Incubation of cells of Anacystis ni- 
dulans in light in a growth medium lack- 
ing Ca 2+ and supplemented with a Ca 2 + 
chelator, glycylglycine, has been found 
to induce a loss of 2 evolution catalyzed 
by photosystem II (PS II), while pho- 
tosystem I (PS I) was unaffected (Becker 
and Brand, 1982). In the previous report 
we presented data showing that Ca 2+ de- 
pletion also induced loss of chlorophyll a 
(Chi a) fluorescence of variable yield both 
at the growth temperature (39°C) and at 
77 K. We present here evidence that Ca 2+ 

is required not only for efficient photo- 
chemistry but also for efficient energy 
transfer from phycobilins to Chi a and 
for efficient energy distribution between 
the two photosystems in Anacystis ni- 

The phycobilins constitute the major 
light-harvesting pigments for PS II in 
cyanobacteria. We have therefore stud- 
ied the effect of Ca 2+ depletion on energy 
transfer from phycobilins to Chi a. Fig- 
ure 51 shows the time-dependent changes 
in fluorescence yield in control, Ca 2 + -de- 




Ca -depleted 










30 sec 


200 400 600 

2 evolution, ^moles/mg Chl,h 

Fig. 51. Time course of Chi a fluorescence measured at 685 nm at 39°C in control, Ca 2+ -depleted, 
and partially restored cells of Anacystis nidulans illuminated with monochromatic red light (633 nm) 
obtained from He-Ne laser. Cells were incubated in Ca 2 + -deprived medium in red light (>600 nm, 300 
ixmol m~ 2 s \ as described earlier by Brand et al., 1983. Aliquots of cells were dark adapted for 4 min 
at 39°C and then illuminated with 633-nm red light obtained from the He-Ne laser (15 mW) attenuated 
with a 25% neutral density filter. Fluorescence at 685 nm was measured with a narrow interference 
filter. The partial restoration of Chi a fluorescence transient shown in the curve labeled "Partially 
restored" was obtained by adding Ca(N0 3 ) 2 (0.35 mM) to the Ca 2+ -depleted cells and incubating in red 
light for 1 h. The insert shows the change in the steady-state level of fluorescence yield at 685 nm at 
39°C excited by the phycobilin-absorbing red light as a function of loss in the rate of 2 evolution resulting 
from Ca 2+ depletion. 

pleted, and partially restored cells of An- 
acystis nidulans when the cells were 
excited with a monochromatic red light 
(633 nm) obtained from an He-Ne laser. 
The control cells showed the character- 
istically slow fluorescence rise and de- 
cline, while the Ca 2 + -depleted cells 
exhibited a rapid rise of fluorescence to 
high level; the fluorescence yield re- 
mained unchanged with the time of il- 
lumination. Thus with light primarily 
absorbed by phycobilins, the Ca 2+ -de- 
pleted cells produced a greater fluores- 
cence yield than the control cells. As the 
cells were depleted of Ca 2+ , the rate of 
2 evolution declined and the fluores- 

cence yield, measured at 685 nm upon 
excitation with light absorbed by phy- 
cocyanin, increased (Fig. 51, inset). Ad- 
dition of Ca 2+ to the depleted cells lowered 
the fluorescence level and restored the 
fluorescence transient (Fig. 51) as well 
as 2 evolution. We note that the extent 
of restoration depended very much upon 
the extent of depletion (as measured in 
terms of loss of 2 -evolution capacity and 
time of incubation in Ca 2+ -depleted me- 
dium) as well as the time of incubation 
in light after addition of Ca 2+ to the de- 
pleted cells. 

Addition of 20 |xM DCMU increased 
the fluorescence yield in the control cells, 



and the slow time-dependent change in 
fluorescence yield persisted in the pres- 
ence of DCMU (Mohanty and Govindjee, 
1973). However, in the case of the Ca 2 + - 
depleted cells, the fluorescence yield re- 
mained insensitive to the addition of 
DCMU; the depleted cells showed no slow- 
fluorescence transients (data not shown). 
The large increase in fluorescence yield 
in Ca 2 + -depleted cells when excited with 
light mainly absorbed by phycocyanin and 
not with light absorbed by Chi a thus 
suggests that Ca 2+ depletion affects the 
energy transfer from phycobilins to Chi a. 
We therefore investigated the effect of 
Ca 2+ depletion on the emission charac- 
teristics of Anacystis excited by light ab- 
sorbed either mostly by phycobilins or 
by Chi a. Ca 2 + -depleted cells with no 
ability to evolve 2 did not show a change 
in the absorption spectrum measured ei- 
ther at 39°C or at 77 K (data not shown). 
Figure 52A shows emission spectra at 
39°C of both control and Ca 2 + -depleted 
cells using monochromatic blue excita- 
tion (442 nm) from an He-Cd laser. Con- 
trol and Ca 2 + -depleted cells showed 
almost identical emission spectra. How- 
ever, excitation of cells with a broad band 
of blue-green light which excited both 
phycocyanin (PC) and Chi a produced an 
emission spectrum that showed an en- 
hanced emission at 665 nm in Ca 2 + -de- 
pleted cells (Fig. 52B). Furthermore, a 
peak around 700 nm with a shoulder 
around 720 nm was also seen in the spec- 
trum of the Ca 2 + -depleted cells. Control 
cells, however, showed a typical emis- 
sion spectrum with a peak at 655 nm pro- 
duced by PC and a peak at 685 nm 
produced by Chi a (Goedheer, 1968). The 
results given in Fig. 52 show that exci- 
tation of PC, and not Chi a, brought about 
the changes in the emission spectrum. 
Thus, Ca 2+ depletion not only prevents 
Q reduction (seen as a loss of transients 
in fluorescence of variable yield in light 
absorbed by PC or by Chi a) but also 
causes the change in the emission spec- 
trum seen only with illumination ab- 
sorbed mostly by PC. It appears that 
Ca 2+ depletion alters the energy trans- 

~\ — I — i — i — " — i — | — i — i — i — i — | — i — i — i — r 


A - 

, Control 


700 750 

Wavelength, nm 


Fig. 52. Fluorescence emission spectra of con- 
trol and Ca 2 + -depleted cells of A. nidulans mea- 
sured at 39°C. Control cells (solid line) and cells 
depleted of Ca 2+ (dashed line) to inhibit all 2 - 
evolving activity were prepared; then fluorescence 
emission spectra were measured as described in 
text. (A) Cells were illuminated with monochro- 
matic blue light from the He-Cd laser. (B) Cells 
were illuminated with a broad band of blue-green 
light. Slit- width of the emission monochromator was 
3.3 nm. The concentration of Chi a was 6 jxg ml -1 . 

fer from phycobilins to Chi a. Excitation 
spectra measured at 39°C also showed 
that there was a relative increase in the 
effectiveness of 620-nm light in exciting 
fluorescence emission at 685 nm (F685) 
in Ca 2+ -depleted cells, as compared to 
control cells (data not shown). This sug- 
gests that Ca 2 + depletion affects the en- 
ergy transfer. 

The emission spectra of control and 
Ca 2 + -depleted cells excited with blue- 






green light and measured at 77 K are 
presented in Fig. 53. The control cells 
exhibited three characteristic emission 
bands, peaking at 655 nm (PC emission), 
684 nm, 696 nm (Chi a 2 emission), and 
722 nm (Chi a x associated with PS I) 
(Goedheer, 1968). In Ca 2+ -depleted cells, 

650 700 750 

Wavelength, nm 


Fig. 53. Fluorescence emission spectra of 
(a) control, (b) Ca 2+ -depleted, and (c) partially re- 
stored cells of A. nidulans at 77 K. Control cells 
and cells depleted of Ca 2+ for 6.5 h, which retained 
15% of their original 2 -evolving capacity, were 
preincubated in darkness at 39°C, then plunged 
into liquid nitrogen for fluorescence measurements. 
Ca(N0 3 ) 2 (0.35 mM) was added to the Ca 2+ -de- 
pleted cells and incubated in red light for 2 h to 
allow recovery (60%) of 2 evolution and restora- 
tion of the previously seen emission characteristics 
(c). Fluorescence spectra were determined from 
samples as described for Fig. 52 and in text. Chi 
a concentration was 6.0 fig ml ~ J for both samples. 
The bottom part of the figure represents difference 
spectra of the (d) depleted minus control and 
(e) restored minus control. 

the PC and PS I emission bands re- 
mained almost unchanged, while a new 
band appearing at 681-683 nm (F683) be- 
came very distinct. The difference spec- 
trum shown in Fig. 53 depicts positive 
peaks at 663 and 681 nm. Addition of 
calcium to the Ca 2 + -depleted cells sup- 
pressed this F683 seen at 77 K; the emis- 
sion spectra resembled that of control 

The measurement of the development 
of this new low-temperature 683-nm 
emission peak in relation to the loss of 
2 evolution due to Ca 2+ depletion showed 
that the loss of 2 evolution and of var- 
iable fluorescence ensued much earlier 
than did the development of F683. How- 
ever, when cells were allowed to remain 
in the medium deprived of Ca 2+ for a 
long-enough period, the F683 became very 
intense (Fig. 54). This band is seen only 
when Ca 2 + -depleted cells were excited 
with light absorbed by PC and not with 



~\ — i — r 

i i i i r 

683 A 

1 / • 

r. ' » 

i i i i i i r 


/ \ 
I i 


\ / p* l Blue-green actinic 
\ Blue actinic 


/ \ i 697 

,' Control 

- / 685 

_ / 

- / 

-r-1 i i LJ 1 L_l l_l I I I L_l I L 

650 700 750 800 

Wavelength, nm 

Fig. 54. Effect of long-term Ca 2+ depletion on 
fluorescence emission of A. nidulans at 77 K. The 
spectra were measured using either monochro- 
matic blue light (442 nm) or a broad band of blue- 
green light. The cells at a concentration of 4 |xg 
Chi ami' 1 were depleted of Ca 2+ for 15 h and used 
at this concentration for spectral measurements. 
These cells completely lacked both 2 evolution and 
the DCMU-induced increase of fluorescence yield. 



monochromatic blue light, which is ab- 
sorbed by Chi a. The latter illumination 
produced no intense emission at 682 nm, 
and the spectrum of the Ca 2+ -depleted 
cells looked identical to that of control 
cells (Fig. 54). 

Blue-green algae exhibit characteristic 
state transitions in that the cells in the 
dark remain in low-fluorescent (high- 
spillover) state II, and change to highly 
fluorescent (low-spillover) state I upon 
adaptation in light absorbed by PS I 
(Murata, 1969; Fork and Satoh, 1983). 
Blue-green algae show these state tran- 
sitions even in the presence of DCMU 
(Mohanty and Govindjee, 1973; Satoh and 
Fork, 1983). In the presence of DCMU, 
the cells remain in state II in the dark 
and adapt to state I in light absorbed by 
either PS I or PS II (Mohanty and Gov- 
indjee, 1973; Satoh and Fork, 1983). Be- 
sides changes in yield produced by dark 
or light adaptation, the cells ofAnacystis 
also show characteristic changes in the 
emission at 77 K that are produced by 
each state (Ley and Butler, 1980; Fork 
and Satoh, 1983). Table 10 shows the re- 
sults of dark or light adaptation on the 
yields of fluorescence in control and Ca 2+ - 
depleted cells in the presence of 20 |jlM 

TABLE 10. Effect of Adaptation to Blue or 

Blue-green Light on State Transitions in Control 

and Ca 2 + -depleted A. nidalans Cells* 

Slow-Fluorescence Rise (Relative Yield) 
at 685 nm 


Adapted to 
Adapted to blue-green 
blue (PS I) (mostly PS II) 
light light 

+ Ca 2+ 24.0 
-Ca 2+ 2.0 

41.0 33.7 
2.0 2.0 

*The state transitions were measured in the 
presence of 20 |a,M DCMU as the slow-fluorescence 
rise (the fluorescence intensity after 5 s of illumi- 
nation minus the fluorescence level (F m ) attained 
after 200 ms of illumination) that is seen after Q 
reduction has taken place. The samples were in- 
cubated with 20 ijlM DCMU at 30°C in darkness 
for 4 min or they were preilluminated in mono- 
chromatic blue or blue-green light for 4 min. The 
preilluminating light was turned off and the fluo- 
rescence transient was recorded as described in 

DCMU. Adaptation of control cells both 
to monochromatic blue light (PS I) or to 
a broad-band blue-green light mostly ab- 
sorbed by PS II produced the slow rise 
in fluorescence yield. The extent of in- 
crease in slow fluorescence yield at 39°C 

Depleted, state I- 1 


J I l_l I I l_J — I — L 

650 700 750 800 

Wavelength, nm 

Fig. 55. The effects of state transitions on the 
low-temperature (77 K) fluorescence emission 
spectra of (6) control and (a) Ca 2 + -depleted A. ni- 
dulans. Cells were incubated in dark or in light 
(broad-band blue-green light) at 39°C for 4 min 
before freezing the cells in liquid nitrogen. Emis- 
sion spectra were recorded with a broad band of 
blue-green light as described in text. The differ- 
ence spectra given in the bottom of the figure show 
(d) the characteristic increase of the F696 band in 
the control cells and (c) an absence of such increase 
in Ca 2 + -depleted cells. The difference curves were 
obtained by subtracting the spectrum for cells in 
state II from that in state I. 



upon adaptation to light is slightly higher 
for blue than for blue-green light (Table 
10). We note that light adaptation caused 
a substantial increase in fluorescence 
yield, suggesting that, in darkness, An- 
acystis cells remain in state II (Satoh and 
Fork, 1983) but are transferred to state 
I in the light even in the presence of 
DCMU. (See also Year Book 81, 50-54.) 
No such state changes were seen in the 
Ca 2+ -depleted cells. Figure 55 shows the 
increase in the F696 band in the emission 
spectrum when control cells were adapted 
to state I before freezing to 77 K. No 
such change was observed when Ca 2 + - 
depleted cells were used. Addition of Ca 2+ 
to depleted cells revived the capacity for 
the state adaptations, as was evidenced 
by the enhanced F696 band upon light 
adaptation (Fig. 55). 

The results presented in this report 
thus confirm and amplify our observation 
that Ca 2+ depletion affects the primary 
photochemistry of PS II (Brand et al., 
1983). These results also suggest that a 
long-term absence of Ca 2+ from intact 
cells of Anacystis disrupts energy trans- 
fer from PC to Chi a (Figs. 52, 53). Ca 2+ 
depletion induced a large alteration of 
the emission spectrum measured at 77 K 
when cells were excited with broad-band 
blue-green light; the 695-nm peak was 
suppressed and a new peak appeared at 

a shorter wavelength (683 nm), its in- 
tensity increasing with time of depletion. 
Since this intense F683 band is absent 
with blue excitation absorbed by Chi a, 
it is evident that Ca 2+ depletion disrupts 
energy transfer from phycobilins to Chi 
a. The fact that the F683 band is not seen 
either in control cells excited by light ab- 
sorbed by PC or Chi a, or in Ca 2 + -de- 
pleted cells in blue light absorbed by Chi 
a suggests that this band originates from 
the emission of phycobilin. This long- 
wavelength emission indicates the fluo- 
rescence in vivo from the terminal form 
of allophycocyanin which transfers en- 
ergy to Chi a (Gantt, 1980). 


Becker, D. W., and J. J. Brand, Biochem. Bio- 

phys. Res. Commun. 109, 1134-1139, 1982. 
Brand, J. J., P. Mohanty, and D. C. Fork, FEBS 

Lett. 155, 120-124, 1983. 
Fork, D. C, and K. Satoh, Photochem. Photobiol. 

37, 421-428, 1983. 
Gantt, E., Int. Rev. Cytol. 66, 45-80, 1980. 
Goedheer, J. C, Biochim. Biophys. Acta 153, 903- 

906, 1968. 
Satoh, K., and D. C. Fork, Photobiochem. Pho- 

tobiophys., in press, 1983. 
Ley, A. C, and W. L. Butler, Plant Physiol. 65, 

714-722, 1980. 
Mohanty, P., and Govindjee, Biochim. Biophys. 

Acta 305, 95-104, 1973. 
Murata, N., Biochim. Biophys. Acta 189, 171-181, 





Satoshi Hoshina and David C. Fork 

Following the initial work of Ogawa et 
al. (1966), who showed two distinct chlo- 
rophyll-protein complexes of spinach 
thylakoid membranes by using SDS- 
PAGE (sodium dodecylsulfate-polyacry- 
lamide gel electrophoresis), several kinds 
of chlorophyll-protein complexes have 
been prepared from thylakoid mem- 

branes of higher plants and algae. In a 
recent study of chlorophyll-protein com- 
plexes of cyanobacteria, six chlorophyll- 
protein complexes were separated from 
Anacystis nidulans by a method em- 
ploying LDS (lithium dodecylsulfate)- 
PAGE (Guikema and Sherman, 1981, 
1983). By contrast, Takahashi et al. (1982) 



resolved eight chlorophyll-proteins from 
another cyanobacterium, Synechococcus 
sp., by SDS-PAGE. 

Argyroudi-Akoyunoglou and Thomou 
(1981) reported the separation of chlo- 
rophyll-proteins from thylakoids of 
Phaseolus vulgaris by SDS-sucrose 
density gradient centrifugation. They 
showed that the SDS-sucrose density 
gradient centrifugation procedure could 
be used for all types of SDS-solubilized 
thylakoids and that this method provided 
a convenient alternative to SDS-PAGE, 
particularly when large amounts of chlo- 
rophyll-proteins for biochemical work 
were required. 

In this report, therefore, we used the 
SDS-sucrose density gradient centrifu- 
gation method to separate three chloro- 
phyll-proteins in high yield from Anacystis 
nidula?is and characterized their spec- 
tral properties. 

Continuous cultures of A. nidulans 
were grown at 39°C (Brand et al., this 
Report). Cells in the logarithmic growth 
phase were harvested, suspended in 30 
mM Na,K phosphate buffer (pH 7.0), and 
washed with 0.6 M sucrose and 30 mM 
Na,K phosphate buffer (pH 6.8). The 
washed cells were incubated with lyso- 
zyme (5 mg/30 ml) containing 2 mM 
EDTA, 0.6 M sucrose, and 30 mM Na,K 

phosphate buffer (pH 6.8) for 2 h at 30°C, 
as described by Murata et at. (1981). The 
spheroplasts were suspended in 10 mM 
NaCl and 10 mM Tricine-NaOH (pH 7.8), 
and centrifuged at 10,000(7 for 30 min. 
The pellet was resuspended in the same 
medium and disrupted by passage through 
a French pressure cell at 14,000 psi. The 
disrupted cells were cooled in an ice- water 
mixture and centrifuged at 30,000# for 
30 min at 4°C to remove unbroken cells 
and debris. The supernatant (25 ml) was 
layered onto a medium (8 ml) containing 
0.4 M sucrose, 10 mM NaCl, and 10 mM 
Tricine-NaOH (pH 7.8), and centrifuged 
at 117,000# (ave) (T-865 rotor, 40,000 rpm) 
for 1 h at 4°C. The pellet (thylakoid mem- 
branes) was resuspended with 0. 1 M Tris- 
HC1 (pH 8.6). 

The thylakoids were solubilized in 0. 1 
M Tris-HCl (pH 8.6)-4.5% SDS (750 |jig 
Chi a/ml, SDS/Chl a = 60) for 30 min at 
room temperature. This solubilized sam- 
ple (0.5 ml) was layered onto a linear 
sucrose density gradient [11 ml, 5-35% 
sucrose in 50 mM Tris-borate (pH 9.5)- 
0.1% SDS] and centrifuged at 258,000a 
(max) (SW 41 rotor, 38,000 rpm) for 16 
h at 15°C. 

Figure 56 shows the separation pat- 
tern obtained using sucrose density gra- 
dient centrifugation. Five bands and an 












1 1 ■ 1 


1 I 


• i 


\ A iv 


la lb 1 

p - 

' i l i I 

■ i 


10 20 30 40 

Fraction number (0.2 ml/frac) 



Yellow green 




Dark green 






|| HI || 








Fig. 56. Separation of chlorophyll-protein complexes by centrifugation on an SDS-sucrose density 



orange-yellow pellet were observed after 
centrifugation. These bands have been 
designated as la, lb, II, III, and IV (from 
top to bottom). In this condition, band II 
constituted the major fraction, compris- 
ing 40-50% of total chlorophyll. When 
the thylakoids were solubilized at a ratio 
of SDS/Chl a = 40 (3.0% SDS, 750 |jig 
Chi a/ml), the major band was band III, 
and at SDS/Chl a = 20 (1.5% SDS, 750 
jjug Chi a/ml), band IV became the major 
fraction. The orange pellet seems to orig- 
inate from cell envelopes contained in the 
thylakoid preparation. 

Bands II, III, and IV exhibited P700 
photooxidation and contained 80-90 Chi 
a molecules per P700, as shown in Table 
11. No photoresponse of P700 was de- 
tected in bands la and lb. 

Figure 57 shows the absorption spec- 
trum of those bands and thylakoids mea- 
sured at room temperature. The spectra 
of bands II, III, and IV showed a max- 
imum at 675 nm, and the spectra of bands 
la and lb had peaks at 671 nm. By con- 
trast, thylakoids showed a maximum ab- 
sorbance at 678 nm. Absorption spectra 
indicate that bands II, III, and IV have 
a very small amount of carotenoids. 

TABLE 11. P700 Content in Thylakoids and 
Chlorophyll-Proteins from A. nidulans* 

Chi a/P700 (Molar Ratio) 

Band la 
Band lb 
Band II 
Band III 
Band IV 

242 ± 6 

79 ± 3 
81 ± 4 
89 ± 3 

*Photooxidation of P700 was determined with a 
Perkin-Elmer 356 spectrophotometer. Sample and 
reference beams were 698 nm and 725 nm, respec- 
tively. The sample was illuminated with blue-green 
light (370 fjimol quanta/m 2 s obtained by using two 
Corning 4-96 filters. The photomultiplier was pro- 
tected with a Schott RG-2 filter. The activity was 
measured at room temperature in a reaction mix- 
ture containing 10 mM Tricine-NaOH (pH 7.8), 10 
mM NaCl, 2 jjlM DCPIP, 1 mM sodium ascorbate, 
1 mM methyl viologen, and thylakoids or complexes 
(about 5 jig Chi a/ml). An extinction coefficient of 
64 mM -1 cm -1 (Hiyama and Ke, 1972) was used. 
Mean values of 4-9 measurements are given. 






1 1 i i 

at 25°C 

400 450 500 550 600 650 

Wavelength, nm 

700 750 






Wavelength, nm 

Fig. 57. Absorption spectra of chlorophyll-pro- 
teins and thylakoids (T). Spectra were measured 
at room temperature using a Perkin-Elmer 356 
spectrophotometer. Chlorophyll concentrations were 
4-11 |xg/ml. 

Figure 58 shows fluorescence emission 
spectra of thylakoids and bands la and 
II measured at 25°C. The spectra of thy- 
lakoids showed three peaks — at 682, 725, 
and 750 nm — and a shoulder near 695 
nm. The band at 750 nm may be attrib- 
uted to an emission from P750 of the cell 
envelopes. In band II, two peaks ap- 
peared at 682 and 725 nm with a shoulder 
around 695 nm. The spectra of bands III 
and IV were very similar to that of band* 
II. Spectra of bands la and lb showed 
two peaks at 680 and 730 nm. These bands 
are highly fluorescent when compared 
with bands II, III, or IV. 

Figure 59 shows the fluorescence emis- 
sion spectra measured at 77 K. Thyla- 
koids showed three emission bands 
peaking at 683, 694, and 715 nm, and a 
shoulder near 750 nm. Band II had two 
peaks, at 683 (F683) and 720 nm (Fl); 



650 700 750 

Wavelength, nm 


Fig. 58. Fluorescence emission spectra of bands 
la, II, and thylakoids (T) measured at 25°C using 
the apparatus described in Year Book 78, 196-199. 
Samples excited by monochromatic blue light (442 
nm, He-Cd laser, Liconix Model 4240). Chlorophyll 
concentrations were 5 fxg/ml. 

similarly, bands III and IV showed two 
peaks, at 683 (F683) and 715 nm (Fl). 
The ratio of the peak height (F683/F1) 
varied among those bands, and showed 
the lowest ratio for band II. By contrast, 
bands la and lb both showed peak emis- 
sions at 680 nm and 730 nm instead of at 
683 and 715 nm. 

The short-wavelength (F680) emis- 
sions of bands la and lb and the lack of 
photochemical activity indicate that these 
bands contain mainly free pigments. 
Bands II, III, and IV are related to pho- 
tosystem I. Thus, SDS-sucrose density 
gradient centrifugation is a useful prep- 










\ \ 

II 1 1 | 1 

A 730 

mi ii 


715 \ 

Sv ^— — -JiL 

^ lb 

\ ' / 



/ 694/ 

/683 | / 

^\. IV 


1 Mill 

i i i i I i 

Ml II 

650 700 750 800 

Wavelength, nm 

Fig. 59. Fluorescence emission spectra of chlor- 
ophyll-proteins and thylakoids (T) measured at 77 K. 
Samples were excited at 442 nm (He-Cd laser). 
Chlorophyll concentrations were 5 |xg/ml. 

aration method for separating the three 
photochemically active chlorophyll-pro- 
teins associated with photosystem I from 
Anacystis nidulans. Further study is 
under way to clarify the interrelation- 
ships of these three chlorophyll-proteins, 
and to compare them with chlorophyll- 
proteins prepared by Guikema and Sher- 
man (1981, 1983). 




Argyroudi-Akoyunoglou, J. H., and H. Thomou, 
FEBS Lett. 135, 177-181, 1981. 

Guikema, J. A., and L. A. Sherman, Biochim. Bio- 
phys. Acta 637, 189-201, 1981. 

Guikema, J. A., and L. A. Sherman, Arch. Biochem. 
Biophys. 220, 155-166, 1983. 

Hiyama, T., and B. Ke, Biochim. Biophys. Acta 

267, 160-171, 1972. 
Murata, N., N. Sato, T. Omata, and T. Kuwabara, 

Plant Cell Physiol 22, 855-866, 1981. 
Ogawa, T., F. Obata, and K. Shibata, Biochim. 

Biophys. Acta 112, 223-234, 1966. 
Takahashi, Y., H. Koike, and S. Katoh, Arch. 

Biochem. Biophys. 219, 209-218, 1982. 




Satoshi Hoshina 

Several absorbing forms of chlorophyll 
a can be detected in the thylakoid mem- 
branes at liquid nitrogen temperature: 
Chi a-662, Chi a-670, Chi a-677, and 
Chi a-684, where the numbers represent 
the wavelengths in nm of the peak po- 
sition (Brown, 1972). Several forms of 
chlorophyll a have also been detected in 
vitro, such as chlorophyll a in nonpolar 
solvents (Cotton et al., 1974), chlorophyll 
a in aqueous dispersions of lipids (Murata 
and Sato, 1978), or water-soluble chlo- 
rophyll-proteins (Sugiyama and Murata, 

It was previously reported that the 
difference spectrum (B - A) between 
chlorophyll a in phosphatidylcholine li- 
posomes at the phase transition temper- 
ature (A) and Chi a at a lower temperature 
(B) had a negative peak at 657-668 nm 
and a positive peak at 675-685 nm. A 
pronounced change in absorbance of chlo- 
rophyll a was seen when the lipid passed 
through the phase-transition tempera- 
ture (Hoshina, 1981). 

The purpose of the present study was 
to apply the curve-fitting method to the 
absorption spectrum of chlorophyll a in 
liposomes at temperatures above and be- 
low the phase transition of the lipid. The 
results suggest that the band with a peak 
at 662-663 nm is transformed into the 
band at 670-671 nm upon cooling below 
the phase-transition temperature of the 

Spinach leaf chlorophyll a and phos- 
pholipid liposomes containing chlorophyll 
a were prepared as described previously 
(Hoshina, 1981). The molar ratios of the 
lipid to chlorophyll a were 49. The ab- 
sorption spectra at different tempera- 
tures were measured with a Shimadzu 
Recording Spectrophotometer UV-300 
and digitally recorded with a Shimadzu 
Spectral Data Processor Sapcom-1. The 
curve analysis of the absorption spec- 
trum was performed by using the RE- 
SOL Program (French et al., 1972). 

The absorption spectra of chlorophyll 
a in phosphatidylcholine liposomes were 
analyzed at temperatures above and be- 
low the lipid phase transition. Figure 60 
shows the results of the curve-fitting 
analysis on the absorption spectrum of 
chlorophyll a in liposomes having an 
equimolar mixture of dipalmitoyl and di- 
myristoyl phosphatidylcholine at 41°C. 
The absorption spectrum was composed 
of one major band with a peak at 669.9 
nm and minor bands with peaks at 650.6, 
662.6, and 684.8 nm. This result is es- 
sentially identical to the analysis of the 
absorption spectrum of chlorophyll a in 
an aqueous dispersion of a mixture of sul- 
foquinovosyldiglyceride and monogalac- 
tosyldiglyceride reported by Murata and 
Sato (1978). The results of curve analysis 
of the absorption spectrum of the same 
sample measured at different tempera- 
tures are shown in Table 12. At lower 









Wavelength, nm 

Fig. 60. Curve-fitting analysis of the absorption spectrum of chlorophyll a in liposomes having an 
equimolar mixture of dipalmitoyl and dimyristoyl phosphatidylcholine. Absorption spectrum was mea- 
sured at 41°C. Dotted line, observed spectra; solid lines, analyzed component bands. 

temperatures, the width of the band at 
662-663 nm tended to be narrower and 
that at 684-686 nm tended to be broader 
than width at higher temperatures, al- 
though the peak positions of bands were 
reasonably similar in the results at other 
temperatures. No change in the area of 
any band was detected upon cooling from 

41°C to 31°C. Upon cooling from 31°C to 
21°C, the area of the band at 662-663 nm 
decreased significantly and the band at 
670-671 nm increased. Further cooling 
of the sample from 21°C to 6°C induced 
a little change in the bands at 662-663 
nm and 670-671 nm. No change in the 
areas of the bands at 650-652 nm and 

TABLE 12. Curve Analysis of Absorption Spectra of Chlorophyll a in Liposomes having an 
Equimolar Mixture of Dipalmitoyl and Dimyristoyl Phosphatidylcholine* 

Band Component (nm) 


















HW Area 

(nm) (%) 















17.6 16.2 
17.1 16.1 

16.7 11.7 
15.4 10.8 





* Absorption spectra were measured at 41°, 31°, 21°, and 6°C. HW indicates half-bandwidth. The 
percentage of each band was calculated from the sum of the areas of the bands. Gaussian curves make 
up 77-100%, and Lorentzian curves the remainder. 



TABLE 13. Curve Analysis of Absorption Spectra of Chlorophyll a in Liposomes of Dipalmitoyl 

Phosphatidylcholine * 

Band Component (nm) 







































* Spectra were measured at 48° and 4°C. HW indicates half-bandwidth. The percentage of each band 
was calculated from the sum of the areas of the bands. Gaussian curves make up 84-100%, and Lorentzian 
curves the remainder. 

684-686 nm was observed. According to 
the phase diagram for the mixtures of 
lipids measured with ESR or fluores- 
cence method (Lee, 1977), the lipid in an 
equimolar mixture of dipalmitoyl and di- 
myristoyl phosphatidylcholine is in the 
liquid crystalline state at 41°C. When the 
temperature is reduced, lipid in the gel 
state appears at 30-33°C. Upon further 
cooling, the regions of the gel state of 
lipid increase and those of the liquid crys- 
talline state of lipid decrease; finally, only 
the gel state of lipid is present. All of the 
lipid is in the gel state at 21° and 6°C. 

Table 13 shows the results of the curve 
analysis of the absorption spectrum of 
chlorophyll a in dipalmitoyl phosphati- 
dylcholine liposomes, measured at 48° and 
4°C. At 4°C (the gel state of the lipid), 
the proportion of the band at 662-663 nm 
decreased and that at 670-671 nm in- 
creased, compared with results obtained 
at 48°C (the liquid crystalline state of the 
lipid). No change in the area of the other 
bands was detected. 

Lee (1975) and Knoll et al. (1980) re- 
ported from results obtained by fluores- 
cence measurements of chlorophyll a in 
phosphatidylcholine liposomes that both 
monomeric (fluorescent) and aggregated 
(nonfluorescent) forms of chlorophyll a 
were present in the gel phase as well as 
in the liquid crystalline phase of the lipid, 
and that the proportion of aggregated 
chlorophyll a increased at the expense of 

monomer when the lipid was in the gel 
state. The present study showed that the 
band at 670-671 nm increased at the ex- 
pense of the band at 662-663 nm in the 
gel state of lipid. One of the possible causes 
of this interconversion of the band at 662- 
663 nm to the band at 670-671 nm may 
be a change in the aggregation state of 
chlorophyll a, as the lipids form a gel 
state at the temperature below the phase 
transition of the lipids. If this view is 
correct, the results suggest that Chi a- 
662 is a less-aggregated form (monomer) 
and that Chi a-670 is a more-aggregated 
form (dimer). However, further evi- 
dence will be required to test this hy- 


Brown, J. S., Annu. Rev. Plant Physiol. 23, 73- 

86, 1972. 
Cotton, T. M., A. D. Trifunac, K. Ballschmitter, 

and J. J. Katz, Biochim. Biophys. Acta 368, 181— 

198, 1974. 
French, C. S., J. S. Brown, and M. C. Lawrence, 

Plant Physiol. J>9, 421-429, 1972. 
Hoshina, S., Biochim. Biophys. Acta 638, 334-340, 

Knoll, W., J. Baumann, P. Korpiun, and U. Thei- 

len, Biochem. Biophys. Res. Commun. 96, 968- 

974, 1980. 
Lee, A. G., Biochemistry U, 4397-4402, 1975. 
Lee. A. G., Biochim. Biophys. Acta 472, 285-344, 

Murata, N., and N. Sato, Plant Cell Physiol. 19, 

401-410, 1978. 
Sugiyama, K., and N. Murata, Biochim. Biophys. 

Acta 503, 107-119, 1978. 






Jacob Levitt 

When onion bulbs are frozen, injury 
increases or decreases during subse- 
quent thawing and post-thawing periods 
depending on the temperatures at which 
the bulbs were frozen (Palta et al. , 1977). 
This kind of freezing is extracellular, re- 
sulting in cell dehydration. It is conceiv- 
able, therefore, that dehydration induced 
by other stresses, such as drought, may 
also be followed on rehydration by an 
increase or decrease in the injury. 

In order to investigate this possibility, 
young (30-50 days from sowing) potted 
cabbage plants, raised from seed in a 
growth chamber, were allowed to wilt by 
withholding water. Leaves were excised 
after moderate to severe degrees of wilt- 
ing, and floated on water. The rate and 
degree of dehydration were determined 
by weighing. 

Moderate wilting, with a net loss of 
about 35% of the leafs water, had no 
harmful effects, and the leaf recovered 
its full turgor within about an hour (curve 
3(4), Fig. 61). Severe wilting (loss of 78% 
of the leafs water) killed the leaves es- 
sentially completely, and there was little 
reabsorption of water even after floating 
on water for three days (curve 9(1), Fig. 
61). Intermediate degrees of wilting (loss 
of 60-65% of the leafs water) yielded the 
most interesting results [curves 6(5) and 
7(5), Fig. 61]. Damage must have oc- 
curred during the wilting, since the reab- 
sorption of water was much slower (3-4 
days) than by the moderately wilted leaf 
(1-2 hours). 

These results suggest that moderate 
injury resulting from moderate wilting is 

20 40 60 80 

Recovery time (h) 


Fig. 61. Reabsorption of water by cabbage 
leaves floated on distilled water after wilting by 
withholding water for four days [leaf 3(4)], six days 
[leaves 6(5) and 7(5)], and ten days [leaf 9(1)]. 

repairable during the slow reabsorption 
of water. Sometimes, however, further 
injury occurs even after turgor is re- 
gained, for the turgor is later lost again 
(Fig. 61) even though the leaf is still float- 
ing on water. The rehydration period may 
therefore be as important as the dehy- 
dration period, as the reabsorption of 
water seems to involve either repair or 
additional injury. These results indicate 
that the same kinds of responses occur 
during the rehydration of dehydrated 
cells, whether the dehydration is drought- 
induced or freeze-induced. 


Palta, J. P., J. Levitt, and E. J. Stadelmann, Plant 
Physiol., 60, 398-401, 1977. 




200 300 . 400 500 

C0 2 concentration, ppm 




C. Eduardo Vallejos and Olle Bjorkman 

Tomato plants from early spring plant- 
ings are exposed to extreme diurnal tem- 
perature fluctuations only where night 
temperatures are generally below the 
critical chilling temperature (approxi- 
mately 12°C). The effect of one-night, low- 
temperature treatment was evaluated in 
tomato plants (cv. VF36) grown in a phy- 
totron cell at 25717°C. The distal leaflet 
of the first fully expanded leaf was se- 
lected for gas-exchange measurement. 
C0 2 -dependence curves for assimilation 
were obtained before and after a single 
12-h, low-temperature (5°C) treatment 
in the dark. In order to eliminate the 
possibility of water stress caused by cool- 
ing of the roots, the pot was wrapped 
with heating tape and insulated with dry 
sand in a styrofoam container. The tem- 
perature of the roots was kept at 17°C. 

A single 12-h, low-temperature treat- 
ment at 5°C decreased the rate of C0 2 
assimilation at C0 2 -saturation conditions 
(Fig. 62) and lowered the mesophyll con- 
ductance. Fluorescence measurements 
at liquid-nitrogen temperature (Powles 
et at. , 1983) indicated that no damage had 
occurred to PS II. These results dis- 
agree partially with those of Martin et 
at. (1981), who reported that, while the 
saturation rate decreased, the mesophyll 
conductance remained the same. In ad- 
dition to strong reduction of stomatal 
conductance, an oscillation in the rate of 
assimilation was observed for more than 
nine hours after treatment. As time 
passed, the oscillations decreased in in- 
tensity. Such oscillations have not been 
reported before. One reason might be that 


Fig. 62. C0 2 -dependence curves before and af- 
ter a single low-temperature treatment (5°C) in the 

although they are easily detectable in a 
continuous-flow system, used here, they 
are difficult to detect in a closed system. 
Although it is tempting to assign these 
oscillations to changes in stomatal ap- 
erture, due to induction of ABA synthe- 
sis by the treatment (Daie and Campbell, 
1981), no direct evidence is available to 
support this assertion. Further experi- 
ments with a tomato mutant, "flacca," 
which lacks ABA (Bradford et al., 1983), 
will probably help elucidate this ques- 


Bradford, K. J., T. D. Sharkey, and G. D. Far- 

quhar, Plant Physiol. 72, 245-250, 1983. 
Daie, J., and W. F. Campbell, Plant Physiol. 67, 

26-29, 1981. 
Martin, B., D. Ort, and J. S. Boyer, Plant Physiol. 

68, 329-334, 1981. 
Powles, S. B., J. A. Berry, and O. Bjorkman, Plant 

Cell Environ. 6, 117-123, 1983. 






Mervyn M. Ludlow and Olle Bjorkman 

Previous investigations with Nerium 
oleander have shown that water stress 
dramatically increases the susceptibility 
of the photosynthetic system of the leaves 
to photoinhibitory injury (Year Book 80, 
57-59; Year Book 81, 76-77). It follows 
from these studies that any mechanism 
that minimizes the interception of ra- 
diant energy under conditions of water 
stress should alleviate injury to the pho- 
tosynthetic system caused by accumu- 
lation of excess excitation energy. 

Siratro (M acroptilium atropurpu- 
reum), a legume native to Mexico, is 
known to maximize the interception of 
radiant energy under conditions of ample 
water supply by continuously adjusting 
the leaf angle so that the leaves remain 
essentially normal to incident sunlight 
throughout the day (diaheliotropic leaf 
movement). However, under conditions 
of water stress the leaves of this species 
instead become paraheliotropic, i.e., the 
leaf angle adjusts so that the leaves re- 
main parallel to the sun, thereby dra- 
matically reducing the interception of 
radiant energy. Hence both the light ab- 
sorbed by the photosynthetic system and 
the leaf temperature are reduced. 

In our present studies, potted siratro 
plants were grown in the Stanford gar- 
den under full summer daylight (maxi- 
mum photon fluence rate 2000 fxmol m" 2 
s" 1 ). After establishment under well- 
watered conditions, plants were gradu- 
ally water-stressed to a leaf water po- 
tential (v|/) of - 1.5 MPa by restricting 
the water supply. This resulted in com- 
plete stomatal closure and parahelio- 
tropic leaf movements. Well-watered 
plants (i|> = 0.5 MPa) maintained a high 
stomatal conductance (400-500 mmol m~ 2 
s _1 ), and the leaves continued to move 
diaheliotropically. When desired, leaf 

movements were prevented by restrain- 
ing the leaves to a horizontal position with 
a grid of fine nylon wires. The light in- 
tensity received by the lower surfaces of 
horizontally restrained leaves was only 
about 8% of that impinging on the upper 
leaf surfaces. 

To assess the damage to the photo- 
synthetic system caused by the different 
treatments, chlorophyll fluorescence of 
the leaves at 77 K was followed at 692 
nm and 734 nm, as described by Powles 
and Bjorkman (1982). All fluorescence 
measurements reported here were made 
in the early morning before sunrise. 
Hence, only those effects of the treat- 
ments that persisted after a full night of 
recovery are shown. 

Figure 63 compares the fluorescence 
characteristics at 692 nm for the upper 
and the lower leaf surfaces of well- watered 
and water-stressed siratro leaves. Water 
stress had little or no effect on the 


u 80 


a 60 

g 40 


L_~ 20 \ 





Fig. 63. Fluorescence emission characteristics 
of upper (left-hand bars) and lower (right-hand bars) 
surfaces of well- watered, unrestrained leaves, and 
of both restrained and unrestrained water-stressed 
leaves of siratro. Intrinsic fluorescence F OJ solid 
area; maximum fluorescence F,„, total height of 
bar; variable fluorescence F v , open and hatched 



fluorescence characteristics of either the 
upper or lower leaf surface as long as the 
leaves were permitted to move para- 
heliotropically. However, when the water- 
stressed leaves were restrained to a hor- 
izontal position, drastic quenching of the 
maximum (F m ) and the variable (F v ) flu- 
orescence took place, indicating severe 
damage to the photosynthetic system. 
Although most pronounced for the upper 
leaf surfaces, the fluorescence character- 
istics of the lower leaf surfaces were also 
affected. This fact, and the fact that the 
intrinsic (F ) fluorescence of the lower 
leaf surface increased, suggest that the 
damage to the photosynthetic system 
caused by restraining the water-stressed 
siratro leaves resulted not only from ex- 
cessive light but also from excessive heat. 
This may not be surprising, since re- 
straining these leaves to a horizontal po- 
sition resulted in a maximum leaf 
temperature of 44°C (in comparison with 
33° and 30°C for the unrestrained water- 
stressed and well- watered plants, re- 
spectively). Although the restrained, 
water-stressed leaves and the unre- 
strained, well-watered leaves inter- 
cepted similar amounts of radiant energy, 
the loss of latent heat by high transpir- 
ation rates prevented the temperature of 
the well-watered leaves from rising to 
the point at which heat damage begins. 
To permit a separation of the effects 
of excessive heat and of excessive light, 
leaves of water-stressed siratro leaves, 
restrained in a horizontal position, were 
subjected to a high light level over a range 
of leaf temperatures in a temperature- 
controlled leaf chamber. Each treatment 
lasted for 5.5 h, and the fluorescence 
characteristics of the upper and the lower 
leaf surfaces were measured the follow- 
ing morning. As shown in Fig. 64, there 
was no high-temperature damage be- 
tween 31° and 42°C because F m ,692 of 
the lower surface was unaffected by leaf 
temperature over this range. However, 
F m ,692 fell sharply as temperature in- 
creasd above 42 °C and high temperature 
became progressively more severe. In 
contrast, F m ,692 of the upper, exposed 

80 - 

.> 60 










1 1 1 

1 1 ' 



1 ' I 








n^ upper 










1 1 1 

1 1 1 

i i i 1 


, 1 1 

30 35 40 45 

Leaf temperature (°C) 

Fig. 64. Interaction between light and leaf 
temperature on fluorescence emission character- 
istics (F m , 692) of restrained, water-stressed sir- 
atro leaves. 

leaf surface fell linearly with tempera- 
ture between 31° and 42°C. Over this 
range of temperature, therefore, there 
was a marked interaction between light 
and leaf temperature, such that in- 
creased temperature exacerbated pho- 
toinhibition. Moreover, at temperatures 
above 42°C, which cause high-tempera- 
ture damage in the dark, the exacerba- 
tion by temperature was even greater, 
probably reflecting functional and phys- 
ical dissociation of the pigment-protein 
complexes of the photosynthetic mem- 

The interaction between temperature 
and light on fluorescence characteristics 
was also determined by enclosing leaves 
in a temperature-controlled chamber and 
exposing them to a range of leaf tem- 
peratures and photon flux densities for 
5.5 h. Fluorescence characteristics de- 
termined the following morning are shown 
in Fig. 65. Even at a quite moderate tem- 
perature of 31°C, variable fluorescence 
(F v ,692) falls, and hence photoinhibition 
increases as photon flux density in- 
creases. Moreover, the sensitivity of 
photoinhibition to photon flux density in- 
creases with temperature until 42°C, when 
damage to the pigment-protein com- 
plexes of the photosynthetic membranes 



Fig. 65. Interactive effects of photon fluence 
rate and leaf temperature (°C) on the variable flu- 
orescence (F v , 692) of the upper surfaces of water- 
stressed siratro leaves. 

by high temperature exacerbates the ef- 
fect of photon flux density on photoinhi- 

bition. Thus over the range of 
temperatures normally experienced by 
siratro leaves in the field, the probability 
of photoinhibitory damage is high unless 
paraheliotropic leaf movements reduce 
the light incidence-fluence ratio to a low 

Paraheliotropic leaf movements there- 
fore protect water-stressed leaves of sir- 
atro from (1) high-temperature damage, 
(2) photoinhibition, and (3) the interac- 
tive effects of high temperature and ex- 
cess light. 


Powles, S. B., and 0. Bjorkman, Planta 156, 97- 
107, 1982. 

ADAPTED CELLS OF C hlamydomovias reinhardtii 

Susanne von Caemmerer, John R. Coleman, and Joseph A. Berry 

The photosynthetic fixation of C0 2 by 
C 3 plants is dependent upon internal pro- 
duction of the substrate ribulose-1, 
5-bisphosphate (RuP 2 ). The concentra- 
tion of this substrate is a function of the 
balance between light-dependent reac- 
tions that produce it and the reactions of 
RuP 2 carboxylase/oxygenase that con- 
sume it. These latter reactions are a 
function of the concentrations of the sec- 
ond substrates C0 2 or 2 , the amount of 
enzyme present, and the state of acti- 
vation of the enzyme as influenced by 
cofactors and effectors such as H + , Mg 2+ , 
and phosphorylated compounds (Lori- 
mer, 1981). Berry and Farquhar (1978) 
and Farquhar et al. (1980) developed 
models for the kinetics of C 3 photosyn- 
thesis which make the assumption that 
the rate might alternatively be limited 
either by the rate of production of RuP 2 
or by the enzymatic reactions that con- 
sume it, the choice depending upon en- 
vironmental factors such as temperature, 

C0 2 concentration, and light intensity, 
and upon plant characteristics that de- 
termine the capacity of these reactions. 
As pointed out by Farquhar (1979), the 
steady-state concentration of RuP 2 pres- 
ent in vivo should provide an index of the 
nature of the limiting step. When the rate 
of supply of RuP 2 is limiting, it follows 
that the concentration of RuP 2 [RuP 2 ] 
present in vivo should be such that the 
enzyme active sites are not fully occu- 
pied. The concentration of enzyme active 
sites [E t ] in the chloroplast is quite high 
(>1 mM) and the Michaelis constant of 
RuP 2 for these sites is very low. There- 
fore, when [RuP 2 ] > [E t ] the active sites 
should be nearly saturated with RuP 2 
molecules. This leads to the interpreta- 
tion that at any particular C0 2 and 2 
concentration when [RuP 2 ] < [EJ, the 
reaction is limited by the supply of RuP 2 , 
and when [RuP 2 ] > [E t ], all of the active 
sites of the enzyme present in the chlo- 
roplast are probably occupied by RuP 2 


molecules, and the rate of catalysis is lim- condition, cells were drawn into a sy- 

ited by the amount of enzyme, or its state ringe containing HC10 4 ~ (final concen- 

of activation. tration 10%) to sample the steady-state 

Collatz (Year Book 78, 248-251; 1982) RuP 2 concentration. The RuP 2 was as- 
reported simultaneous measurements of sayed as described by Collatz {Year Book 
photosynthetic 2 exchange and [RuP 2 ] 77, 248-251). RuP 2 carboxylase active- 
of isolated spinach leaf cells and Chla- site density was assayed as described by 
mydomonas reinhardtii in response to Collatz et al. (Year Book 78, 171-175), 
changes in C0 2 , 2 , light intensity, and and catalytic activity was assayed as de- 
temperature. In the dark, [RuP 2 ] was scribed by Seemann and Berry (Year Boo k 
undetectable and increased to concentra- 81 , 78-83) with cell lysates obtained with 
tions exceeding the estimates of binding- a French pressure cell. Carbonic anhy- 
site concentration at high light intensity drase was assayed as described by Cole- 
and low concentrations of C0 2 and 2 , man et al. (this Report), 
providing evidence that the photosyn- 
thesis of these organisms was limited by Results and Discussion 
the RuP 2 supply under some conditions Cells used in these experiments were 
and by the amount of activated RuP 2 car- routinely analyzed for chlorophyll, RuP 2 
boxylase/oxygenase under other condi- carboxylase activity, RuP 2 carboxylase 
tions. Other investigators (Perchorowicz protein, and carbonic anhydrase activity 
et al., 1982; Perchorowicz and Jensen, (Table 14). The activity of carbonic an- 
1983) show that [RuP 2 ] is often higher hydrase was about 40 times higher in cells 
than the presumed active site concentra- cultured at low rather than high C0 2 con- 
tion of wheat leaves; they argue that light- centrations, but there were only small 
and C0 2 -dependent regulation of the ac- differences in the amount of chlorophyll 
tivation state of RuP 2 carboxylase plays per cell or in the amount of RuP 2 car- 
an important role in regulating the rate boxylase on a chlorophyll basis. Kinetic 
of photosynthesis even when light is studies of the carboxylase obtained from 
strongly limiting for photosynthesis. The freshly lysed cells indicated a higher k cat 
present studies were conducted to reex- (6.6 s" 1 per active site vs. about 3 s _1 ) 
amine and extend the studies of Collatz, and a higher K m (55 |jlM vs. about 10 \xM) 
and to investigate the influence of the for the algal as opposed to the higher- 
inducible C0 2 -concentrating system of the plant enzyme. As reported by Badger et 
alga C. reinhardtii (see Coleman et al., al. (1980), there was no detectable dif- 
this Report) on the local environment of ference in the kinetic properties of RuP 2 
the RuP 2 carboxylase reaction in these carboxylase from high- or low-C0 2 -grown 
cells as indicated by changes in the en- cells. 

vironmental control of [RuP 2 ]. Studies of photosynthetic exchange of 

oxygen (Fig. 66, bottom) verified that at 

Materials and Methods constant and rate-saturating illumination 

Chlamydomonas reinhardtii 2137 mt + there was about a tenfold difference be- 
was grown as described (Coleman et al., tween high- and low-C0 2 -grown cells in 
this Report) and bubbled either with air the concentration of total inorganic car- 
or with 5% C0 2 in air. Cells were har- bon required to achieve half-saturation 
vested during log-phase growth and re- of photosynthesis. As discussed else- 
suspended in 30 mM MOPS buffer, pH where in this Report (Coleman et al.), 
7.2. Assays of net 2 production were this difference in the apparent require- 
conducted in a Clark-type oxygen elec- ment for carbon is thought to be related 
trode (Rank Brothers) with 2.0 ml of cell to the existence in low-C0 2 -adapted cells 
suspension containing 20-30 fxg of chlo- of a transport system that provides C0 2 
rophyll ml -1 . After a constant rate of to the site of the RuP 2 carboxylase re- 
photosynthesis was achieved at a given action at a concentration several times 


TABLE 14. Comparison of High- and Low-C0 2 -Grown Cells of Chlamydomonas reinhardtii 


High C0 2 

Low CO.? 

Chlorophyll content, \xg 10 7 cells 

RuP 2 Carboxylase 
activity, (xmol mg l Chi hr _1 
active site density, nmol mg" 1 Chi 
protein content, g/g Chi 

Carbonic Anhydrase 
activity, Wilbur-Anderson units 
ratio to RuP 2 carboxylase 

















336 ± 60 
14.1 ± 3.5 
1.0 ± 0.2 

2194 ± 991 

higher than that in the surrounding me- 
dium. The high-C0 2 -grown cells, lacking 
this mechanism, have internal C0 2 con- 
centrations similar to that in the me- 

The intracellular pool of RuP 2 present 
in algae rapidly killed while photosyn- 
thesizing at a constant rate (Fig. 66, top) 
also responded strongly to the carbon 

100 « 

12 3 4 

Total carbon, mM 

Fig. 66. The dependence of net 2 production 
(bottom) and the RuP 2 pool size (top) on the total 
inorganic carbon concentration with high- (circles) 
and low-C0 2 -grown (triangles) cells of Chlamy- 
domonas reinhardtii. The measurements were 
conducted in 21% 2 -saturated buffer (30 mM 
HEPES, pH 7.2) at 25°C, and at an irradiance of 
1000 ixmol quanta (400-700 nm) m" 2 s _1 . 

concentration. The highest [RuP 2 ] was 
observed at low carbon concentrations, 
and [RuP 2 ] fell as carbon was increased. 
This pattern is consistent with a chang- 
ing balance between the rates of pro- 
duction and consumption of RuP 2 . At high 
carbon concentrations the RuP 2 carbox- 
ylase reaction can occur rapidly and draws 
the steady-state [RuP 2 ] down to a low 
level. At low carbon concentration the 
rate of the carboxylation reaction is much 
slower, and the potential rate of RuP 2 
synthesis should remain about the same 
as it was at high carbon. Consequently 
the steady-state pool of RuP 2 builds up 
to a high level. 

These studies at saturating light in- 
tensity should have permitted the max- 
imum rate of RuP 2 production to occur, 
while the potential rate of carboxylation 
was altered by changing the carbon con- 
centration. At low carbon concentrations 
[RuP 2 ] (Fig. 66) was greater than [E t ] 
(17.3 nmol mg" 1 Chi, Table 14), indicat- 
ing that the concentration of activated 
RuP 2 carboxylase was limiting the rate 
of carboxylation under this condition. As 
the carbon concentration was increased, 
the [RuP 2 ] fell, reaching equivalence with 
the active-site concentration at a total 
carbon concentration of about 1.5 mmol 
and 0.5 mmol for the high- and low-C0 2 - 
grown cells, respectively. Above these 
carbon concentrations, [RuP 2 ] was lower 
than [E t ] indicating that carboxylation of 
RuP 2 was limited by the capacity for RuP 2 

If, as proposed, the low-C0 2 -grown al- 
gae have a higher internal concentration 
of C0 2 than do the corresponding high- 



C0 2 -grown algae, then the [RuP 2 ] should 
be lower at any given carbon concentra- 
tion in the low-C0 2 cells, since the car- 
boxylation reaction would be limited to 
a lesser extent by C0 2 . This is indeed 
what was observed (Fig. 66, top), a re- 
sult providing strong additional support 
for the inference that the low-C0 2 cells 
have a C0 2 -concentrating mechanism. One 
additional observation was a tendency for 
[RuP 2 ] to fall at very low external carbon 
concentration (Fig. 66) with the low-C0 2 
cells, but [RuP 2 ] did not fall to a concen- 
tration that would be rate-limiting. 

Another way to study the balance be- 
tween RuP 2 production and consumption 
is by controlling the light intensity while 
keeping the carbon concentration con- 
stant. With high-C0 2 cells at 0.2 mM car- 
bon, [RuP 2 ] varied from about 2 to about 
70 nmol mg -1 Chi when the light was 
increased from darkness to 1600 ixmol 
quanta (400-700 nm) m -2 s _1 . Data col- 
lected over this range of light intensity 
are plotted (Fig. 67) as rate of 2 pro- 
duction vs. [RuP 2 ]. It is clear that the 
data are biphasic, with a region of strong 
dependence of photosynthesis on [RuP 2 ] 
and another region where rate appears 
to be independent of [RuP 2 ]. The tran- 

sition between these regions seems to be 
at about 20 nmol mg " l Chi, which is close 
to the estimate of [EJ. 

This pattern of response is similar to 
that predicted (see fig. 1 of Farquhar, 
1979) from a model for the kinetics of the 
reaction of RuP 2 with the carboxylase 
when the concentration of active sites is 
high, as occurs in vivo. We should, also, 
be able to predict the rate of photosyn- 
thesis in vivo from the model and from 
measurements of the quantity of enzyme 
present in these algae and its kinetics in 
vitro. For this comparison we selected 
measurements of photosynthesis for which 
[RuP 2 ] > E t , as these rates should be 
rate-saturated with respect to RuP 2 . For 
the calculation we used E t = 17.3 nmol 
mg" 1 Chi and k cat = 6.6 s _1 (from this 
study); K m (C0 2 ) = 55 ijlM (Badger et al. , 
1980); ^(0 2 ) = 480 |jlM (Jordan and 
Ogren, 1981), and a Michaelis-Menten 
expression for the RuP 2 -saturated rate 
of the carboxylation reaction as a func- 
tion of 2 and C0 2 (equation 5, Far- 
quhar, 1979). The C0 2 and 2 
concentrations were 20 and 250 fxM, re- 
spectively. We also took into account the 
finding of Kaplan and Berry (1981) that 
net 2 production under these conditions 









o o 

CO j_ 

O o 

CO J= 







*± en 





£ o 






1 ° 


r E 





20 40 60 80 

[RuP 2 ] , nmol mg Chh 1 

Fig. 67. The dependence of net 2 production on the RuP 2 pool size. Inorganic carbon concentration 
was constant at 0.2 mM, and light intensity was varied to cause variation in [RuP 2 ]. Cells were grown 
at high C0 2 , oxygen was 21%, and the temperature was 25°C. 



was only about 80% of the rate of gross 
C0 2 uptake (presumably because these 
high-C0 2 -grown cells form gly collate as 
an end product of photosynthesis). These 
calculations yield a theoretical estimate 
of the rate of 2 production (ignoring res- 
piration) of 63.4 |jimol mg ~ 1 Chi h ~ 1 which 
is slightly higher than the actual rate 
measured in vivo (mean = 42.4 |xmol 
mg -1 Chi h -1 ). The rate of respiration 
observed with these cells in the dark (30 
jjimol mg ~~ 1 Chi h ~ *) is similar to the dif- 
ference between the actual and the cal- 
culated rates. If there was substantial 
respiratory C0 2 production under the 
measurement condition, then the esti- 
mated and actual rates may agree nearly 
perfectly. Taken together, the good 
quantitative and qualitative agreement 
obtained in this study with predictions 
based on the model of Farquhar (1979), 
provide strong experimental support for 
the validity of that model as a basis for 
analyzing the kinetics of C 3 photosyn- 
thesis in vivo. 

Studies of the control of [RuP 2 ] by car- 
bon concentration conducted with low- 
C0 2 -grown cells (data not shown) yielded 
about a fivefold steeper initial depend- 
ence of 2 production on [RuP 2 ] at the 
same external carbon concentration as 
that shown for high-C0 2 -grown cells (Fig. 
67). This might be explained by assuming 
that the k cat for the carboxylase was much 
higher than for the corresponding high- 
C0 2 cells (not supported by measure- 
ments conducted in vitro), or that the 
C0 2 concentration at the site of the car- 
boxylase reaction was much higher. High- 
C0 2 -grown cells measured at tenfold 
higher external carbon concentration 
yielded a similarly steeper slope, sup- 
porting the latter interpretation. 


The data obtained in these studies with 
C. reinhardtii provide additional support 

for the postulate that [RuP 2 ] is a factor 
regulating the rate of the RuP 2 carbox- 
ylase reaction, and the observed de- 
pendence of photosynthesis upon [RuP 2 ] 
was qualitatively similar to that pre- 
dicted from the analysis by Farquhar 
(1979). Quantitative analysis of the re- 
sponses showed reasonably good agree- 
ment between the responses observed in 
vivo and those predicted from measure- 
ments of the RuP 2 carboxylase of these 
cells in vitro. The comparison of highl- 
and low-C0 2 -adapted cells gave results 
completely consistent with the proposal 
that, given the same external C0 2 con- 
centrations, the C0 2 concentration at the 
site of the RuP 2 carboxylase reaction was 
substantially higher than in air-adapted 
cells in high-C0 2 -grown cells of Chla- 


Badger, M. R., A. Kaplan, and J. A. Berry, Plant 
Physiol. 66, 407-413, 1980. 

Berry, J. A., and G. D. Farquhar, Proceedings of 
the Fourth International Congress of Photosyn- 
thesis, Reading, 119-131, D. 0. Hall, J. Coombes, 
and T. Goodwinn, eds., The Biochemical Society, 
London, 1978. 

Collatz, G. J., Ph.D. Thesis, Stanford University, 

Farquhar, G. D., Arch. Biochem. Biophys. Acta 
193, 456-468, 1979. 

Farquhar, G. D., S. von Caemmerer, and J. A. 
Berry, Planta U9, 78-90, 1980. 

Jordan, D. R., and W. L. Ogren, Nature 291, 513- 
515, 1981. 

Kaplan, A., and J. A. Berry, Plant Physiol. 67, 
229-232, 1981. 

Lorimer, G. H., Annu. Rev. Plant Physiol. 32, 
349-384, 1981. 

Perchorowicz, J. T., D. A. Raynes, and R. G. Jen- 
sen, Proc. Nat. Acad. Sci. USA 78, 2985-2989, 

Perchorowicz, J. T., and R. G. Jensen, Plant 
Physiol. 71, 955-960, 1983. 

Seemann, J. R., Ph.D. Thesis, Stanford Univer- 
sity, 1982. 





Joseph A. Berry, Malcolm Nobs, Bernardita Osorio, Jeffrey D. Palmer, James Tepperman, and 

William F. Thompson 

The enzyme ribulose-l,5-bisphosphate spinach. Thus, in order to extend this 
(RuP 2 ) carboxylase (EC 4. 1. 1.39), which line of investigation we chose to examine 
catalyzes the initial C0 2 -fixation reac- the RuP 2 carboxylases of C 3 and C 4 spe- 
tion, is a major limiting factor in the pho- cies of Atriplex and their F x hybrids. A 
tosynthetic metabolism of C 3 plants, previous study (Yeoh et aL, 1981) dem- 
Previous reports (Year Book 80, 67-72; onstrated that the K m (C0 2 ) of RuP 2 car- 
Fear Book 81 , 78-83) have shown that boxylase from C 4 plants was consistently 
significant differences may exist be- about double that of C 3 plants, and the 
tween species in the catalytic efficiency C 3 species Atriplex hastata L. (synon- 
of this enzyme on a protein basis, sug- yms, A. patula ssp. hastata Hall and 
gesting that it might be possible to im- Clements, A. triangularis Wild.) could 
prove the photosynthetic performance (at be hybridized with the C 4 species A. ro- 
least on a protein basis) of some plants sea, yielding a plant with many charac- 
through genetic manipulation of the genes teristics intermediate between the two 
coding for this enzyme. For example, the parents but lacking a functional C 4 pho- 
rate of photosynthesis of leaves of spin- tosynthetic pathway (Year Book 68, 620- 
ach under C0 2 -limiting conditions was 633; Year Book 69, 624-649; Year Book 
about 1.7 times higher per unit RuP 2 car- 70, 507-511). We reasoned that these 
boxylase protein than it was for similar species of Atriplex might provide the op- 
leaves of soybean plants under identical portunity to test the heritability of 
conditions, and, since this enzyme is a kinetic differences. Preliminary ex- 
large fraction of the total leaf protein of periments verified that the kinetic prop- 
these species (about 20%), the rate of erties of the RuP 2 carboxylase protein 
photosynthesis per unit total leaf nitro- from A. rosea and A. hastata were in- 
gen also differed by a factor of about 1.5. deed different, and Malcolm Nobs was 
These differences in the efficiency of pho- persuaded to repeat the tedious hybrid- 
tosynthesis could be related quantita- ization process, 
tively to a corresponding difference in 

the rate of C0 2 uptake, predicted to oe- Materials and Methods 

cur from studies of the kinetics of the 

corresponding enzymes assayed in vitro. Atriplex rosea and A. hastata seeds 

In this instance, the K m (C0 2 ) was about were collected from the vicinity of the 

the same but the k cat was higher for the Palto Alto yacht harbor, and seedlings 

spinach enzymes than for the soybean were maintained in a greenhouse. Hy- 

enzymes. bridizations were conducted as previ- 

If these differences in the apparent ously described (Year Book 69, 624-629). 

catalytic efficiency of these enzymes are Hybrid seeds were germinated in moist 

indeed intrinsic and not the result of an sand and in chambers, as described above, 

artifact, then it should be possible to ma- These plants were morphologically and 

nipulate the enzymatic properties genet- cytologically similar to the F 1 hybrids ob- 

ically. Genetic barriers between species tained in the earlier experiments (Year 

of different families prevent the testing Book 69, 624-649). RuP 2 carboxylase was 

of this hypothesis with soybeans and extracted and assayed essentially as de- 


scribed previously (Year Book 81, 78-83) concentrating systems accrues from the 

except that RuP 2 (0.4 mM) was enzy- higher k cat . The change in K m is not likely 

matically generated in the assay medium to have any significant effect at the high 

immediately prior to the assays from ATP C0 2 concentration provided to these en- 

and ribose-5 phosphate using phosphor- zymes. 

ibose isomerase (Sigma) and phosphori- The hybrid plants had kinetic prop- 
bulokinase (RuP 2 carboxylase-free) erties (Table 15) closely resembling those 
generously supplied by George Lorimer. of the C 4 parent, A. rosea. Highly rep- 
Nuclear and chloroplast DNAs were ex- Heated studies with different hybrid plants 
tracted and purified from leaves of ma- indicated no significant difference be- 
ture plants as described in Palmer (1982). tween individual plants or between the 
Chloroplast and nuclear DNA hetero- hybrid and the C 4 . The hybridization was 
geneities were analyzed as described accomplished with the C 4 species as the 
previously (Year Book 81, 96-97; Year female parent. The fact that the hybrid 
Book 81, 98-101). enzyme properties resembled those of the 

female parent suggests that these traits 

Results and Discussion are inherited maternally. It is well es- 

tablisned that the gene specifying the 

The K m (C0 2 ) of RuP 2 carboxylase from large subunit of RuP 2 carboxylase is lo- 

the C 4 species was about double that of cated on the DNA of the chloroplast and 

the C 3 species (Table 15), as reported for that this subunit contains the active site 

other C 3 and C 4 species (Yeoh et al. , 1981). of the enzyme (Kung, 1977). In contrast, 

In addition, we observed that the k caf of the small subunit is coded by the nuclear 

the C 4 species was also significantly genome. No function is yet known for the 

higher. This is in agreement with the small subunit, and no additional insight 

findings of a recent survey of other plants into this puzzle is provided from these 

(M. R. Badger and J. R. Seemann, per- studies. 

sonal communication). The higher k cat of In order to check that the purported 

the enzyme from C 4 species would, in the hybrid was indeed a hybrid and that the 

presence of the high concentration of C0 2 chloroplast of the hybrid was derived from 

thought to occur within the bundle sheath the maternal parent as expected, DNA 

cells of these plants, yield a higher rate was prepared from the two parents and 

of carboxylation per unit protein than three hybrid individuals, and analyzed 

would the corresponding enzyme from the for restriction-site heterogeneity. The 

C 3 plant. It is of interest that the car- fragments obtained after digesting the 

boxylase of the green alga Chlamydo- chloroplast DNAs with Nru 1 and Sac 1 

monas reinhardtii, which is thought to (Fig. 68) showed a great deal of homol- 

possess a C0 2 -concentrating system ogy between the two genomes, but A. 

(Coleman et al., this Report), also has hastata had fragments at about 2 and 7 

higher k cat and K m than the carboxylases kb (Nru 1) and at 15 kb (Sac 1) that were 

of C 3 higher plants. It seems likely that lacking in the genome of A. rosea, while 

the changes in K m and k caf are linked (see A. rosea had fragments at about 9 and 

Ulmer, 1983), and that the major advan- 14 kb (Nru 1) and at 8.5 kb (Sac 1) lacking 

tage to these organisms possessing C0 2 - in A. hastata. The genomes of the hy- 

TABLE 15. Kinetic Constants of RuP 2 Carboxylase from a C 3 and a C 4 
Species of Atriplex and Their F l Hybrid. 

K m (|xM) k ca i (s _1 per active site) 

A. hastata 13.7 ± 0.5 4.4 ± 0.2 

A. rosea 22.3 ± 1.0 6.0 ± 0.2 

F! hybrid 19.6 ± 0.7 6.0 ± 0.2 



R R R R R R 


H H H H H H 


RxxxH RxxxH 


Fig. 68. Electrophoresis in 0.8% agarose gels 
of Nru 1 and Sac 1 digests of chloroplast DNA from 
Atriplex rosea (R), A. hastata (H), and three F x 
hybrid plants (R x H). Fragment size differences 
attributable to restriction site changes between the 
parental genomes are evident at approximately 2, 
7, 12, and 14 kilobases (kb, left axis) with the Nru 
1 digest, and at 8.5 and 15 kb with the Sac 1 digest. 
The hybrids possessed all of the fragments char- 
acteristic of A. rosea and lacked those character- 
istic of A. hastata. 

brids were each identical to that of A. 
rosea. These differences in the restric- 
tion sites of chloroplast DN As permitted 
a clear distinction of the two parental 
species and confirmed that the chloro- 
plast DNA of the hybrid was derived from 
the maternal parent. 

The nuclear DNA was much more 
complex (Fig. 69), and no heterogeneity 
could be resolved by fragment size alone. 
Increased resolution was obtained by 
Southern blot analysis of heterogeneity 
in those fragments that showed homol- 

Fig. 69. Electrophoresis (left) of Eco RII di- 
gests of nuclear DNA from A. rosea (R), A. hastata 
(H), and three F 1 hybrid plants (R x H), and South- 
ern blot analysis (right) of duplicate nitrocellulose 
filter blots of the gel shown at the left using a cloned 
rDNA repeat unit from Pisum sativum. A restric- 
tion site heterogeneity between the parental gen- 
omes is indicated by fragment size differences at 
approximately 1.9 and 2.2 kb. Hybrid individuals 
possessed DNA characteristic of both parents. 

ogy to the rDNA repeat unit of Pisum 
sativum. A clear difference in the pa- 
rental genomes was resolved with a frag- 
ment of about 2.3 kb (A. rosea) and 1.9 
kb (A. hastata). The hybrids contained 
both fragments, indicating biparental in- 
heritance. These studies confirm that hy- 
brids were obtained and that the 
chloroplast s were maternally inherited. 
Since the small subunit of RuP 2 carbox- 
ylase is coded by the nuclear DNA, it 
seems likely that the enzyme present in 
the hybrids was heterogeneous with a 
mixture of small subunits derived from 
either parent (Rhodes et al., 1980). The 



large subunit, however, could only be from 
A. rosea. Attempts to achieve the hy- 
bridization with the C 3 species as the fe- 
male parent, the reciprocal cross, have 
not been successful. 


A hybrid between A. rosea and A. has- 
tata possessed an RuP 2 carboxylase that 
was clearly similar to the maternal par- 
ent. The large and easily resolved dif- 
ferences in the kinetic properties of the 
parental types and their hybrid provide 
strong evidence that these differences are 
under genetic control. Examination of the 
nuclear and chloroplast DNAs confirmed 
that these were inherited in the hybrid 
according to a biparental and a maternal 
pattern, respectively. Since the large 
subunit of the hybrid enzyme was de- 

rived only from the maternal parent, these 
results indicate that the major determi- 
nants of the catalytic properties are res- 
ident on the large subunit and are coded 
for by the chloroplast DNA. Not only 
have these studies confirmed the feasi- 
bility of genetic modification of RuP 2 car- 
boxylase kinetic properties, but the 
chloroplast genome has been identified as 
the relevant target for such modifica- 


Kung, S.-D., Annu. Rev. Plant Physiol. 28, 401, 

Palmer, J. D., Nucl. Acids Res. 10, 1593-1605, 

Rhodes, P. R., S.-D. Kung, and T. Marsho, Plant 

Physiol. 65, 69-73, 1980. 
Ulmer, K., Science 219, 666-671, 1983. 
Yeoh, H.-H., M. R. Badger, and L. Watson, Plant 

Physiol. 67, 1151, 1981. 

ANHYDRASE IN Chlamydomonas reinhardtii 

John R. Coleman, Joseph A. Berry, Robert K. Togasaki,* and Arthur R. Grossman 

The photosynthetic characteristics of 
the green alga Chlamydomonas rein- 
hardtii are dependent upon the C0 2 con- 
centration experienced by the alga during 
growth (Year Book 75, 423-432). Cells 
grown at air levels of C0 2 (0.03%) have 
a much higher affinity for C0 2 during 
photosynthesis than do algae grown at 
high C0 2 concentrations (3-5%). Since 
the C0 2 concentration at which the algae 
are grown has no effect on either the 
mechanism of photosynthetic C0 2 fixa- 
tion or the K m (C0 2 ) of the principal C0 2 - 
fixing enzyme, ribulose-l,5-bisphosphate 
(RuBP) carboxylase (Year Book 75, 423- 
432), another process for improving the 
efficiency for C0 2 utilization must be 
present in air-grown cells and absent in 
high-C0 2 -grown cells. We now know that 

*Department of Biological Sciences, Indiana 
University, Bloomington. 

the appearance both of carbonic anhy- 
drase activity and of the mechanisms for 
the active transport and accumulation of 
bicarbonate enables air-grown algae to 
photosynthesize efficiently at low levels 
of inorganic carbon (Nelson et al., 1969; 
Badger et al., 1980). Growth at high lev- 
els of C0 2 suppresses the activity of car- 
bonic anhydrase and the bicarbonate 
transport system, and if these cells are 
transferred to low C0 2 concentrations 
they are unable to fix carbon photosyn- 
thetically. The presence of carbonic an- 
hydrase and bicarbonate transport enables 
air-grown cells to form a large, intracel- 
lular inorganic carbon pool, which pro- 
vides sufficient C0 2 for saturation of 
RuBP carboxylase and elimination of the 
oxygen inhibition of photosynthesis (Bir- 
mingham et al., 1981; Coleman and Col- 
man, 1980). C. reinhardtii is also able to 
adapt to changes in the C0 2 concentra- 



tion during growth. When high-C0 2 - 
grown cells are transferred to air levels 
of C0 2 and illuminated, their photosyn- 
thetic capacity at these low levels of car- 
bon gradually increases such that 5 h after 
the transfer the algae display photosyn- 
thetic characteristics similar to those ob- 
served in air-grown cultures. During this 
adaptation period the induction both of 
carbonic anhydrase activity and of bicar- 
bonate transport occurs. Exactly how 
carbonic anhydrase and the transport 
system coordinate inorganic carbon ac- 
cumulation and, indeed, where this pro- 
cess is located in the cell are not fully 

We have studied the induction of car- 
bonic anhydrase activity in C. reinhard- 
tii in an attempt to localize and identify 
the protein responsible for this activity 
and to gain some understanding of the 
mechanism of its regulation. We have also 
examined the effect of the change in C0 2 
concentration on the regulation of syn- 
thesis of other proteins. 

Materials and Methods 

Chlamydornonas reinhardtii 2137 mt + 
(obtained from M. Spalding, Michigan 
State University) and the cell wall-less 
mutant, CW-15 (obtained from R. K. To- 
gasaki, Indiana University), were cul- 
tured axenically in the minimal medium 
described by Spalding et al. (1982) at 28°C 
and a light intensity of 300 |jimol m " 2 s " 1 
(400-700 nm). Cultures were vigorously 
shaken and bubbled with either 5% C0 2 
in air or with air alone. All experiments 
were performed with cells in early-to- 
midphase exponential growth. 

Carbonic anhydrase activity, ex- 
pressed in Wilbur- Anderson units (WA), 
in cell pressates was determined electro- 
metrically, as previously described (Wil- 
bur and Anderson, 1948). 

For the determination of if'/~>(C0 2 ), as 
well as for the maximal rate of photo- 
synthesis, P majn algae grown under the 
appropriate conditions were harvested 
by centrifugation (4,000#), resuspended 
in C0 2 -free 20 mM MOPS buffer, pH 7.2, 

and the rates of 2 evolution at varying 
HC0 3 ~ concentrations were measured at 
saturating light intensity and 25°C with 
a Clark-type 2 electrode. 

Autolysin was isolated according to 
Tamaki et al. (1981). Titrations were per- 
formed with the crude autolysin prepa- 
ration to determine both the time of 
incubation and the concentration most ef- 
fective in removing the Chlamydornonas 
cell wall. Since the effectiveness of these 
preparations varied, titrations were re- 
quired in each case. 

Purification of carbonic anhydrase re- 
leased in the medium by air-grown CW- 
15 cells or autolysin-treated wild-type cells 
was achieved by a combination of DE AE 
cellulose and agarose gel column chro- 
matography. Proteins synthesized dur- 
ing growth at 5%, 0.03%, and after 
transfer from 5% to 0.03% C0 2 , were 
analyzed following in vivo labeling with 
35 S0 4 ~" . After a labeling period of 3-5 h 
in the light, the cells were quickly broken 
in a pre-chilled French press (25,000 psi) 
and centrifuged briefly at 2,000# to re- 
move unbroken cells and debris. The su- 
pernatant was then centrifuged at 43,000# 
for 30 min to pellet the thylakoid frac- 
tion. The resulting supernatant was sub- 
jected to high-speed centrifugation 
(150,000#) for 3 h to produce a pellet and 
supernatant fraction. The supernatant 
was made 10% trichloroacetic acid to pre- 
cipitate the protein. The three frac- 
tions — thylakoids, high-speed super- 
natant, and pellet — were treated with 
sodium dodecylsulfate (to 1.7%), and the 
proteins were analyzed by sodium do- 
decylsulfate polyacrylamide gel electro- 
phoresis (12-18% linear acrylamide 
gradient containing 8 M urea). The gels 
were stained with Coomassie brilliant blue 
G-250, dried, and the newly synthesized 
polypeptides were visualized by autora- 

Results and Discussion 

As shown in Table 16, the transfer of 
C. reinhardtii from growth at high C0 2 


TABLE 16. Adaptation of 5% COr-Grown Chlamydomonas to Air Levels 

(0.03%) of C0 2 

Time (h) 

Ki M (CO,) 


1 max 


(|j.mol-mg Chi 

















h' 1 ) 


concentrations to air levels results in a 
substantial increase in the whole-cell af- 
finity for C0 2 . The inorganic carbon con- 
centration required for half-saturation of 
photosynthesis by high-C0 2 -grown cells 
is similar to that required for RuBP car- 
boxylase isolated from the alga (Year Book 
75, 423-432). These results suggest that 
high-C0 2 -grown Chlamydomonas lacks 
a C0 2 -concentrating mechanism. In con- 
trast, the ^(C0 2 ) of photosynthesis for 
high-C0 2 -grown cells transferred to air 
decreases rapidly with time; after a 5-h 
exposure to low C0 2 , the ^ A (C0 2 ) is 
twenty times lower than either that of 
high-C0 2 -grown cells or the K nt (C0 2 ) of 
the isolated carboxylase. Carbonic an- 
hydrase activity was measured in Chla- 
mydomonas at various times after 
transferring cultures from high C0 2 to 
air (Fig. 70). After a 5-h induction pe- 
riod, the level of carbonic anhydrase ac- 
tivity approaches 50% of the total activity, 
measured after 24 h of induction. A com- 
parison of the increase in carbonic an- 
hydrase activity with the increase in 
affinity for C0 2 during photosynthesis 
(Table 16) suggests that only a portion 
of the total activity induced is needed for 
the initial decline in the K>/>(C0 2 ) ob- 
served after a 5-h exposure to air. The 
continued increase in carbonic anhydrase 
activity past this point may be required 
for maximal reduction in the if'/XC0 2 ) 
attained after extended periods of air ex- 

The induction of carbonic anhydrase 
activity after transfer from high to air 
levels of C0 2 was also examined in the 

cell wall-less Chlamydomonas mutant, 
CW-15. In this mutant the kinetics of 
carbonic anhydrase induction were sim- 
ilar to that observed in the wild-type cells 
(data not shown). However, the enzyme 
does not remain in the cell but is excreted 
into the medium. The export of carbonic 
anhydrase into the medium by this mu- 
tant suggests that most of this enzyme 
is normally located within the peri- 
plasmic space (an area between the cell 
wall and the plasmalemma) or within the 
cell wall itself. In the absence of the cell 
wall, the enzyme is lost to the surround- 
ing medium. Further evidence for the 
location of the enzyme was obtained by 
treating air-adapted wild-type cells with 
autolysin, a cell wall-degrading enzyme 


1 1 ' 1 ' 1 

1 1 ' 1 ' 1 




"5 900 






E 700 

/ • 


< 500 






/ , 1 . 1 . 1 


4 8 12 16 20 

Time, h 


Fig. 70. Induction of carbonic anhydrase activ- 
ity in Chlamydomonas reinhardtii after transfer 
of cultures from 5% C0 2 to air levels of C0 2 . 



synthesized by Chlamydomonas during 
mating. This treatment resulted in the 
release of more than 80% of the total as- 
sayable carbonic anhydrase activity into 
the medium. Carbonic anhydrase was not 
being released as a result of cell lysis, 
since less than 0.1% of the total RuBP 
carboxylase, a soluble intracellular pro- 
tein, could be detected in the medium 
following autolysin treatment. These data 
indicate that although some carbonic an- 
hydrase may be located within the cell 
(presumably inside the chloroplast), most 
of the enzyme passes through the plas- 
malemma, its appearance regulated by 
the C0 2 concentration in the growth me- 

Since the cell wall-less mutant of 
Chlamydomonas excretes carbonic an- 
hydrase into the medium during growth 
at air levels of C0 2 , this cell type was 
used as a source of the protein for iden- 
tification and purification. Following a 24- 
h period at air levels of C0 2 , the CW-15 
cells were removed from the medium by 
low-speed centrifugation. The carbonic 
anhydrase-containing medium was then 
fractionated by column chromatography 
(DEAE cellulose and agarose gel filtra- 
tion), and the proteins in fractions con- 
taining carbonic anhydrase activity were 
separated by sodium dodecylsulfate po- 
lyacrylamide gel electrophoresis (Fig. 71). 
Carbonic anhydrase was also identified 
by reacting an aliquot of ammonium sul- 
fate-concentrated growth medium with 
dansylamide (5-dimethyaminonapthal- 
ene-1 -sulfonamide), which forms a highly 
fluorescent complex with carbonic an- 
hydrase (Drescher, 1978). This sample 
was applied to a native polyacrylamide 
gel and electrophoresed at 4°C. A single 
fluorescent band was located by expo- 
sure of the gel to ultraviolet illumination. 
This band was excised from the gel, elec- 
troeluted, and electrophoresed on a so- 
dium dodecylsulfate polyacrylamide 
gradient gel (Fig. 71). Analysis of the 
fluorescent band in this manner reveals 
a single polypeptide (Fig. 71, lane 1) which 
co-migrates with the major polypeptide 
in the fractions containing the highest 

1 23456789 10 






Fig. 71. Isolation and identification of carbonic 
anhydrase (CA) by sodium dodecylsulfate, poly- 
acrylamide gel electrophoresis (as described in text). 
The sample in lane 1 was prepared from a band 
exhibiting carbonic anhydrase activity on a non- 
denaturing Laemmli gel. The activity on the gel 
was monitored by ultraviolet fluorescence of dan- 
sylamide, which associates tightly with carbonic 
anhydrase. The nondenatured sample was excised 
from the gel, electroeluted, and re-electro- 
phoresed. (Sodium dodecylsulfate polyacrylamide 
gel electrophoresis is described in text.) Lanes 2- 
10 show stained polypeptides from fractions exhib- 
iting carbonic anhydrase activity after a combina- 
tion of DEAE cellulose and agarose gel column 
chromatography. The peak activity was found in 
the fractions of lanes 3-6. The molecular weight 
standards indicated in this figure were phospho- 
rylase b (92,500), bovine serum albumin (66,200), 
ovalbumin (45,000), carbonic anhydrase (31,000), 
soybean trypsin inhibitor (21,500), and lysozyme 

carbonic anhydrase activity from the 
agarose column (Fig. 71, lanes 2-10). 
These data demonstrate that the mon- 
omer of carbonic anhydrase in Chlamy- 
domonas is a polypeptide with a molecular 
weight of 37,000. One earlier report has 



described carbonic anhydrase of Chla- 
mydomonas as a hexamer in the native 
state (Bundy and Cote, 1980). The mon- 
omer was reported to be 27,000. While 
it is difficult to reconcile our results re- 
lating to monomer size with those re- 
ported by Bundy and Cote, the monomelic 
species might be susceptible to proteo- 
lytic degradation. 

In vivo labeling of proteins of wild-type 
Chlamydomonas cells with 35 S0 4 ~ has 
demonstrated the synthesis of this same 
37,000 polypeptide when the cells are 
grown continuously in air or allowed to 
adapt for 5 h to air levels of C0 2 . This 
polypeptide is not synthesized when the 
cells are maintained at high C0 2 concen- 
trations (Fig. 72, compare lanes 1, 2, and 
3). The 37,000 protein is most easily vi- 
sualized in the high-speed pellet fraction 
where approximately 60% of the total 
carbonic anhydrase activity is located. 
This band is also visible in the high-speed 
supernatant fraction which contains ap- 
proximately 40% of the carbonic anhy- 
drase activity. However, in this latter 
fraction, the labeled protein profile is much 
more complex, and carbonic anhydrase 
represents only a minor component. The 
distribution of carbonic anhydrase be- 
tween the pellet and the supernatant may 
be due to the high molecular weight of 
the multimeric enzyme, its association 
with other cellular components, or non- 
specific adhesion to other proteins (per- 
haps to RuBP carboxylase, the major 
protein of the high-speed pellet fraction). 
In both the high-speed supernatant and 
pellet, carbonic anhydrase is such a mi- 
nor species it is only apparent in the au- 
toradiogram and not in the profile of 
stained polypeptides. These results are 
not surprising, since carbonic anhydrase 
is a very efficient enzyme (Tobin, 1970) 
and therefore little protein is required 
for high levels of catalysis. 

Also shown in Fig. 72, the transfer of 
Chlamydomonas from high to low C0 2 
concentrations affects the rate of syn- 
thesis of a number of polypeptides and, 
in particular, RuBP carboxylase (com- 
pare lanes 1 and 2). A more-detailed ex- 

1 2 3 4 5 6 







*< ss 


Fig. 72. Newly synthesized polypeptides dur- 
ing the transfer of cultures of Chlamydomonas 
reinhardtii from 5% C0 2 to air. Samples prepared 
from the high-speed pellets (lanes 1-3) and high- 
speed supernatants (lanes 4-6) following in vivo 
labeling were electrophoresed on polyacrylamide 
gels (as described in text). The gels were dried and 
exposed to Kodak XAR-5 film to visualize the newly 
synthesized polypeptides. Cells were labeled for 3 
h during growth at 28°C on 5% C0 2 (lanes 1 and 
4), air (lanes 3 and 6), or following a transfer from 
5% C0 2 to air (lanes 2 and 5). All lanes in the gel 
received equal protein loads, as determined by 
staining with Coomassie brilliant blue G-250. Mo- 
lecular weight markers are described in Fig. 71. 
The large subunit (LS) and small subunit (SS) of 
RuBP carboxylase and the carbonic anhydrase 
monomer (CA) are indicated in the figure. 



amination of this response is presented 
elsewhere in this Report (pp. 109-111). 

Additional evidence that the 37,000 
polypeptide is carbonic anhydrase has 
been obtained by analyzing 35 S-labeled 
proteins secreted into the medium by CW- 
15 cultures grown either in air or high 
levels of C0 2 , and comparing the se- 
creted proteins with those released from 
wild-type cells (also grown in air or high 
levels of C0 2 ) by autolysin treatment. 
The results of such an experiment are 
presented in Fig. 73. With either cell type, 
the 37,000 polypeptide is synthesized only 
by air-grown cells or by those exposed 
to low levels of C0 2 for 5 h. Maintenance 
of the cells at high levels of C0 2 inhibits 
the synthesis of this polypeptide. In the 
samples prepared from the autolysin- 
treated wild-type cells, a small amount 
of cell lysis, as indicated by the release 
of RuBP carboxylase into the medium, 
occurs. However, the prominence of the 
37,000 protein after autolysin treatment 
is in complete agreement with the dis- 
tribution and regulation of carbonic an- 
hydrase activity demonstrated in the 
experiments with cultures of CW-15. 

It is interesting to speculate on the 
role of carbonic anhydrase in Chlamy- 
domonas reinhardtii. The enzyme acts 
to maintain the concentrations of the var- 
ious inorganic carbon species in rapid 
equilibrium with one another; the rela- 
tive abundance of each species is a func- 
tion of the relative pH of the system. The 
rapid influx of bicarbonate into the cell 
by a transport system on the plasma 
membrane may require the action of car- 
bonic anhydrase to keep the bicarbonate 
concentration at equilibrium levels. The 
location of carbonic anhydrase in the 
periplasmic space would be of particular 
advantage to a soil organism, such as 
Chlamydomonas, where the aqueous en- 
vironment is limited to a thin layer of 
water. The maintenance of the bicarbon- 
ate concentration at equilibrium levels 
would allow for a more-rapid solubiliza- 
tion of gaseous C0 2 and would provide 
sufficient inorganic carbon for the trans- 
port system. 

1 2 3 4 5 6 7 




;; s 


31 000 

21 500»- 



Fig. 73. Newly synthesized polypeptides re- 
leased to the medium by the cell wall-less mutant 
(CW-15) of Chlamydomonas reinhardtii, or follow- 
ing autolysin treatment of wild-type cells. Condi- 
tions for growth, labeling, and autoradiography were 
as described in the legend of Fig. 72. Lane 1 shows 
newly synthesized polypeptides in the high-speed 
pellet of air-grown wild-type cells. Lanes 2-4 show 
newly synthesized polypeptides released into the 
growth medium by CW-15 grown on high C0 2 (lane 
2), air (lane 4), and following transfer from high 
C0 2 to air (lane 3). Lanes 5-7 show newly synthe- 
sized polypeptides released into the growth me- 
dium after autolysin treatment of wild-type cells 
grown on high C0 2 (lane 5), air (lane 7), or following 
transfer from high C0 2 to air (lane 6). Molecular 
weight markers are described in the legend of Fig. 
71. The large subunit (LS) and small subunit (SS) 
of RuBP carboxylase and the carbonic anhydrase 
monomer (CA) are indicated in the figure. 



Badger, M. R., A. Kaplan, and J. A. Berry, Plant 
Physiol. 66, 407-413, 1980. 

Birmingham, B. C, J. R. Coleman, and B. Col- 
man, Plant Physiol. 69, 259-262, 1981. 

Bundy, H. F., and S. Cote, P histochemistry 19, 
2531-2534, 1980. 

Coleman, J. R., and B. Colman, Plant Physiol. 65, 
980-983, 1980. 


Drescher, D. G., Anal. Biochem. 90, 349-358, 1978. 
Nelson, E. B., A. Cenedella, and N. E. Tolbert, 

Phytochemistry 8, 2305-2306, 1969. 
Spalding, M. H., and W. L. Ogren, FEBS Lett. 

U5, 41-44, 1982. 
Tamaki, S., Y. Matsuda, and Y. Tsubo, Plant Cell 

Physiol. 22, 127-133, 1981. 
Tobin, A. J., /. Biol. Chem. 2^5, 2656-2666, 1970. 
Wilbur, K. M., and N. G. Anderson, J. Biol. Chem. 

176, 147-154, 1948. 



John Coleman and Arthur Grossman 

The expression of carbonic anhydrase 
activity in Chlamydomonas and other 
chlorophytes is regulated by the C0 2 
concentration of the growth medium 
(Berry et al., Year Book 75, 423-432; 
Findenegg, 1976; Hogetsu and Miyachi, 
1977). Cells grown at or adapted to air 
levels of C0 2 exhibit both carbonic an- 
hydrase activity and the capacity to 
transport actively and accumulate inor- 
ganic carbon. In contrast, algae grown 
at high levels of C0 2 (3-5%) have little 
carbonic anhydrase and only a limited ca- 
pacity to transport inorganic carbon. The 
concomitant expression of carbonic an- 
hydrase and bicarbonate transport activ- 
ities enable air-grown algae to 
photosynthesize much more efficiently 
than high-C0 2 -grown algae at low levels 
of inorganic carbon. In the previous re- 
port we demonstrated that synthesis of 
a specific protein, which appears follow- 
ing the transfer of C. reinhardtii from 
high to low C0 2 concentrations, is re- 
sponsible for carbonic anhydrase activ- 
ity. This protein is a diffuse band of 
approximately 37,000 on sodium dode- 
cy lsulfate poly aery lamide gels, and is lo- 
calized in the periplasmic space (the area 
where the majority of the carbonic an- 
hydrase activity is found). In this study 
we examined the effect of inhibitors of 
protein synthesis and protein glycosy- 
lation on the expression of carbonic an- 

Materials and Methods 

Chlamydomonas reinhardtii 2137 mt + 
and the cell wall-less mutant, CW-15, 
were grown as described previously 
(Spalding and Ogren, 1982). For in vivo 
labeling studies, the cells were har- 
vested, resuspended in S0 4 ~-free growth 
medium containing the appropriate in- 
hibitor, and preincubated in the dark for 
15 min prior to the addition of 35 S0 4 ~. 
The cultures were then placed at the ap- 
propriate C0 2 concentration and incu- 
bated under growth conditions for 5 h. 
The concentrations of inhibitors used 
were: 100 jjig/ml chloramphenicol, 2 fxg/ 
ml cycloheximide, 2 |xg/ml tunicamycin. 
After labeling, the cells were fraction- 
ated (Coleman et al., this Report), and 
the newly synthesized polypeptides were 
analyzed by sodium dodecylsulfate po- 
lyacrylamide gel electrophoresis fol- 
lowed by autoradiography. Proteins 
released into the medium by CW-15 cells 
were concentrated on DEAE cellulose 
prior to their preparation for electropho- 
resis. Carbonic anhydrase activity was 
assayed electrometrically as described 
previously (Wilbur and Anderson, 1948). 

Results and Discussion 

As shown in Table 17, an inhibitor of 
translation on 80S cytoplasmic ribo- 
somes, cycloheximide, blocks the indue- 



TABLE 17. Effect of Inhibitors of Protein Synthesis and Glycosylation on the 
Induction of Carbonic Anhydrase Activity 

Growth Conditions 

Carbonic Anhydrase Activity 
WA-mgChr 1 

High C0 2 


5 h on Air 

5 h on Air + Chloramphenicol 

5 h on Air + Cycloheximide 

5 h on Air + Tunicamycin 






12 3 4 5 6 7 

8 9 1011121314 


66,200 * 

45,000 * 







Fig. 74. Effect of chloramphenicol and cycloheximide on the synthesis of carbonic anhydrase in C. 
reinhardtii. Cells were labeled with 35 S0 4 ~ for 5 h during growth at 5% C0 2 (lanes 1, 2, 3 and 8, 9, 10) 
after transfer from 5% C0 2 to air (lanes 4, 5, 6 and 11, 12, 13) and during growth at air levels of C0 2 
(lanes 7 and 14). Chloramphenicol (lanes 2, 5 and 9, 12) was used at 100 (xg/ml while cycloheximide (lanes 
3, 6 and 10, 13) was used at 2 jxg/ml. After labeling, the cells were lysed and fractionated as described 
in text. Samples prepared from the high-speed pellet were subjected to sodium dodecylsulfate poly- 
acrylamide gel electrophoresis. The profiles of the Coomassie blue-stained polypeptides obtained after 
each treatment are shown in lanes 1-7. The gels were then dried and exposed to Kodak XAR-5 film for 
visualization of the newly synthesized polypeptides (lanes 8-14). Carbonic anhydrase (CA) and the large 
(LS) and small (SS) subunits of RuBP carboxylase are indicated in the figure. Molecular weight markers 
were phosphorylase b (92,500), bovine serum albumin (66,200), ovalbumin (45,000), carbonic anhydrase 
(31,000), soybean trypsin inhibitor (21,500), and lysozyme (14,400). 



tion of carbonic anhydrase activity in 
Chlamydomonas. Chloramphenicol, an 
inhibitor of translation on the 70S plastid 
ribosomes, has no effect on the expres- 
sion of activity. Since we have demon- 
strated that a 37,000 polypeptide is 
responsible for carbonic anhydrase activ- 
ity, we examined the effect of inhibitors 
on its synthesis (Fig. 74). As shown pre- 
viously, synthesis of both the 37,000 
polypeptide and carbonic anhydrase ac- 
tivity is repressed by growth in high C0 2 
(Fig. 74, lane 8) but induced if the cul- 
tures are grown continuously in air (Fig. 
74, lane 14) of if they are transferred 
from high C0 2 to air (Fig. 74, lane 11). 
Cycloheximide inhibits the synthesis of 
this polypeptide (Fig. 74, lane 13), while 
chloramphenicol has no significant effect 
(Fig. 74, lane 12). The impact of these 
inhibitors on the synthesis of the large 
subunit (LS) and small subunit (SS) of 
RuBP carboxylase indicates that they are 
working properly. Chloramphenicol blocks 
the translation of LS (Fig. 74, lanes 9 
and 12) but does not inhibit SS synthesis, 
while cycloheximide blocks the transla- 
tion of SS (Fig. 74, lanes 10 and 13) but 
does not eliminate LS synthesis. These 
data demonstrate that carbonic anhy- 
drase is synthesized on cytoplasmic ri- 
bosomes and is therefore encoded in the 
nuclear genome. 

Since most of the carbonic anhydrase 
is localized to the periplasmic space, it 
must traverse the plasmalemma to reach 
its site of function. This fact, in conjunc- 
tion with the observation that the poly- 
peptide responsible for carbonic 
anhydrase activity is a diffuse band on a 
poly aery lamide gel, suggests that gly- 
cosylation may be an important step in 
the enzyme's biosynthesis. Tunicamycin, 
a potent inhibitor of protein glycosyla- 
tion (Ericson et al. y 1977), limits the in- 
duction of carbonic anhydrase activity 
after a 5-h exposure to air to approxi- 
mately 20% of the control level (Table 
17). Examination of the polypeptides 
synthesized during the induction period, 
in the presence or absence of the tuni- 
camycin, revealed that this inhibitor 

blocks the synthesis of the 37,000-molec- 
ular-weight polypeptide (Fig. 75, lanes 5 
and 7). No synthesis of this polypeptide 

1 2 3 4 5 6 7 





, > 

:,,,: y- ■ 



mm**** ""* SS 


Fig. 75. Effect of tunicamycin on the synthesis 
of carbonic anhydrase by C. reinhardtii. Cells grown 
at the appropriate C0 2 concentration were incu- 
bated with 35 S0.4~ and tunicamycin (2 |xg/ml) for 5 
h. The newly synthesized polypeptides of the high- 
speed pellet fraction were determined as described 
in Fig. 74. A profile of the Coomassie blue-stained 
polypeptides is in lane 1. Inhibitor treatment did 
not lead to noticeable alterations in this profile. 
Newly synthesized polypeptides from C0 2 -grown 
cells (lanes 2 and 3), air-grown cells (lanes 6 and 
7), and cells transferred from high C0 2 to air levels 
of C0 2 (lanes 4 and 5) are presented. Cultures treated 
with tunicamycin are in lanes 3, 5, and 7. Untreated 
cultures are in lanes 2, 4, and 6. Carbonic anhy- 
drase (CA) and large (LS) and small (SS) subunits 
of RuBP carboxylase are indicated in the figure. 
Molecular weight markers are described in the leg- 
end of Fig. 74. 



occurs at high levels of C0 2 in either the 
presence or absence of tunicamycin (Fig. 

75, lanes 2 and 3). The synthesis of other 
polypeptides in the high-speed pellet 
fraction is not affected by tunicamycin. 

Since tunicamycin inhibits the induc- 
tion of carbonic anhydrase, a major 
periplasmic protein, we examined the ef- 
fect of this inhibitor on the synthesis and/ 
or export of the general population of 
proteins of the periplasmic space. For 
this purpose we employed the cell wall- 
less Chlamydomonas mutant, CW-15. 
Electrophoretic analysis of polypeptides 
released into the medium by CW-15 is 
presented in Fig. 76. The 37,000 poly- 
peptide is not found in the medium of 
cultures maintained on high C0 2 in the 
presence or absence of tunicamycin (Fig. 

76, lanes 1, 2, 4, and 5). Air-grown cells 
excrete large amounts of carbonic an- 
hydrase (Fig. 76, lanes 3 and 7) unless 
tunicamycin is included in the culture 
medium (Fig. 76, lanes 4 and 8). In ad- 
dition to blocking the appearance of car- 
bonic anhydrase, tunicamycin causes a 
reduction in the total number of exported 
polypeptides, regardless of the C0 2 con- 
centration. In cultures grown under high 
C0 2 almost no polypeptides seem to be 
exported, while air-grown cultures do 
export some polypeptides in the pres- 
ence of tunicamycin. However, many of 
these polypeptides do not co-migrate with 
exported polypeptides from cultures 
grown in the absence of tunicamycin. 
These results suggest that even if some 
export is occurring, the exported poly- 
peptides may not be properly or com- 
pletely processed. In yeast, tunicamycin 
also blocks the synthesis of several ex- 
ported glycoproteins such as invertase 
and acid phosphatase, but it has no effect 
on the synthesis of intracellular polypep- 
tides (Kuo and Lampen, 1974). 

The reason for the absence of 37,000 
protein in the presence of tunicamycin is 
unclear. The addition of the carbohy- 
drate moiety may be required for export 
or resistance of the newly synthesized 
polypeptide to proteolytic degradation 
(either within the cell or in the peri- 

1 2 3 4 5678 






Fig. 76. Effect of tunicamycin on the export of 
polypeptides from CW-15. Lanes 1, 2 (stained gel) 
and 5, 6 (autoradiogram) show profiles from the 
media of high-C0 2 -grown cells in the presence (lanes 
2 and 6) and the absence (lanes 1 and 5) of tuni- 
camycin. Lanes 3, 4 (stained gel) and 7,8 (autora- 
diogram) show profiles from the media of cells 
transferred from high C0 2 to air in the presence 
(lanes 4 and 8) and absence (lanes 3 and 7) of tuni- 
camycin. Labeling was for 5 h. Carbonic anhydrase 
(CA) is indicated in the figure. Molecular weight 
markers are described in the legend of Fig. 74. 

plasmic space). The low levels of carbonic 
anhydrase activity found following in- 
duction in the presence of tunicamycin 
may be the result of another form of the 
enzyme which does not require glyco- 
sylation (perhaps localized to the chlo- 
roplast), or may represent the activity 
of the nascent polypeptide (localized 
within the cell prior to degradation). 




Kuo, S.-C, and J. 0. Lampen, Biochem. Biophys. 
™ „ r m r, ** i a r^ r^iu Res. Commun. 58, 287-295, 1974. 

E T£', m rZl\l ^LTl^ Elbem ' Spalding, M. a, and W. L. Ogren, FEES Lett. 

J. Biol. Chem. 252, 7431-7433, 1977 
Findenegg, G. R., Z. Pfiayizenphysiol. 79, 428-437, 

Hogetsu, D., and S. Miyachi, Plant Cell Physiol. 

18, 342-352, 1977. 

U5, 41-44, 1982. 
Wilbur, K. M., and N. G. Anderson, J. Biol. Chem. 
176, 147-154, 1948. 

ADAPTATION OF Chlamydomonas reinhardtii TO LOW C0 2 

John Coleman and Arthur Grossman 

The transfer of Chlamydomonas rein- 
hardtii from high to low concentrations 
of C0 2 during growth results in marked 
changes in the photosynthetic character- 
istics of the organism (Berry etal., Year 
Book 75, 423-432). The induction both of 
carbonic anhydrase and of a bicarbonate 
transport capacity following the trans- 
fer, enable the alga to photosynthesize 
efficiently at C0 2 concentrations well be- 
low that required for saturation of either 
isolated RuBP carboxylase or most ter- 
restrial plants. 

In this study we examined the effect 
of adaptation to low C0 2 concentrations 
on the synthesis of polypeptides other 
than carbonic anhydrase. We note tran- 
sient changes in the synthesis of a few 
polypeptides, including both the small and 
large subunits of RuBP carboxylase. 

Materials and Methods 

Chlamydomonas reinhardtii 2137 mt + 
was grown at the desired C0 2 concen- 
tration, as previously described (Cole- 
man etal., this Report). 35 S0 4 _ was used 
for in vivo labeling of proteins. Cells grown 
at high C0 2 concentrations were trans- 
ferred to air levels of C0 2 at the appro- 
priate time and labeled for 2 h. The cells 
were then harvested, broken in a pre- 
chilled French pressure cell, and frac- 
tionated, as described previously (Cole- 
man et at., this Report). Proteins in each 
fraction were separated by sodium do- 

decylsulfate polyacrylamide gel electro- 
phoresis, and the newly synthesized 
polypeptides were visualized by autora- 

Results and Discussion 

Chlamydomonas reinhardtii grown at 
high C0 2 concentrations was allowed to 
adapt in air for 2, 4, 6, 12, and 24 hours. 
Prior to the final 2 h of adaptation, the 
cells were harvested and resuspended in 
sulfate-free medium to which 35 S0 4 ~ was 
added. Controls were maintained at high 
C0 2 concentrations over the labeling pe- 
riod (Fig. 77, lanes 1 and 7). In this man- 
ner we were able to examine newly 
synthesized polypeptides from cells at 
different stages of adaptation. The stained 
gel and the autoradiogram of the poly- 
peptides in the high-speed pellet fraction 
are shown in Fig. 77. This high-speed 
pellet contains much of the RuBP car- 
boxylase holoenzyme plus other, less- 
prominent polypeptides. Although each 
lane received equal amounts of protein 
(as shown by the stained gel profile, Fig. 
77, lanes 1-6), the autoradiogram shows 
substantial differences in the rate of syn- 
thesis of both the large (LS) and the small 
(SS) subunits of RuBP carboxylase over 
the adaptation period. Synthesis of RuBP 
carboxylase is substantially reduced 
after a 2-h exposure to air (lane 8) and 
is almost eliminated after 4 h of adap- 
tation (lane 9). Following the 4-h time 



I 2 3 4 5 6 7 8 91011 12 

66 200) 






Fig. 77. Newly synthesized polypeptides in the 
high-speed pellet fraction after transfer of Chla- 
mydomonas from high C0 2 to air for various times. 
Stained protein profiles are presented in lanes 1- 
6, while labeled polypeptides are shown in lanes 
7-12. Cells were incubated in 35 S0 4 during growth 
at 5% C0 2 (lanes 1 and 7) or during exposure to 
air levels of C0 2 for 2 h (lanes 2 and 8), 4 h (lanes 
3 and 9), 6 h (lanes 4 and 10), 12 h (lanes 5 and 11), 
and 24 h (lanes 6 and 12). Labeling was for 2 h. 
Breakage and fractionation are described in text. 
Samples were prepared from the high-speed pellet 
fraction and subjected to sodium dodecylsulfate po- 
lyacrylamide gel electrophoresis followed by au- 
toradiography. Large subunit (LS) and small subunit 
(SS) of RuBP carboxylase are indicated on the fig- 
ure. Molecular weight markers are phosphorylase 
b (92,500), bovine serum albumin (66,000), oval- 
bumin (45,000), carbonic anhydrase (31,000), soy- 
bean trypsin inhibitor (21,500), and lysozyme 

point, synthesis of both subunits begins 
to increase. Other polypeptides in this 
fraction, in a similar manner, show a 
transient decrease in synthesis (2-4 h) 
during the induction period. Carbonic an- 
hydrase, a diffuse band of approximately 
37,000 in the high-speed pellet, is not 
easily seen in this autoradiogram be- 
cause of its low rate of synthesis with 

respect to other newly synthesized poly- 

Changes in protein synthesis during 
adaptation are also observed in the high- 
speed supernatant fraction from Chla- 
mydomonas (Fig. 78). As indicated on 
the stained gel (Fig. 78, lanes 1-6), each 
lane received equal amounts of protein. 
However, the autoradiogram clearly 
shows transient changes in the rates of 
synthesis of both large and small sub- 
units of RuBP carboxylase (this enzyme 
is found in both the high-speed pellet and 
supernatant), as well as other soluble- 
proteins (indicated by arrows). After a 
2-h exposure to air, two prominent poly- 
peptides are synthesized (lane 8, indi- 
cated by arrows on the left-hand side of 
the autoradiogram). The rate of synthe- 
sis of these polypeptides quickly de- 
creases and essentially no new synthesis 
occurs after a 6-h exposure to air. A poly- 
peptide of approximately 20,000 (indi- 
cated by arrow) is also synthesized during 
the induction period. This species is sim- 
ilar in molecular weight to the precursor 
of the small subunit of RuBP carboxylase 
and may represent the immature poly- 
peptide. A pre-sequence (or transit pep- 
tide) of approximately 4,000 is present 
at the N-terminal of the primary trans- 
lation product of the small subunit. The 
pre-sequence is thought to enable the 
small subunit to traverse the double 
membrane of the chloroplast envelope. 
(For a review of this subject see Gross- 
man et al. t 1982.) The observed decline 
in the rate of synthesis of RuBP carbox- 
ylase holoenzyme could be the result of 
a specific decrease in the translation of 
the large subunit messenger RNA. (Some 
evidence, not presented here, does in- 
dicate that the control of large subunit 
synthesis is at the level of translation and 
not transcription.) A decreased rate of 
large subunit synthesis may result in ele- 
vated levels of the small subunit precur- 
sor if the unassembled large subunit is 
required for the processing and/or trans- 
port of the immature small subunit. Im- 
munological characterization of this 
presumptive precursor is in progress. 



12 3 4 5 6 

7 8 9101112 

92 500 
66,'200 ! 

45,000 * 


21 500 >- 

-« LS 



Fig. 78. Newly synthesized polypeptides in the high-speed supernatant fraction after transfer of 
Chlamydomonas from high C0 2 to air for various times. Stained protein profiles are presented in lanes 
1-6, while labeled polypeptides are shown in lanes 7-12. The cells were labeled in vivo (see legend of 
Fig. 77), and the high-speed supernatant fraction was prepared as described in text. Cells grown at 5% 
C0 2 (lanes 1 and 7) or exposed to air levels of C0 2 for 2 h (lanes 2 and 8), 4 h (lanes 3 and 9), 6 h (lanes 
4 and 10), 12 h (lanes 5 and 11), and 24 h (lanes 6 and 12), were used for labeling. Large subunit (LS) 
and small subunit (SS) of RuBP carboxylase as well as other polypeptides with altered rates of synthesis 
during the induction are indicated in the figure. Molecular weight markers are described in the legend 
to Fig. 77. 

In conclusion, the rates of synthesis of 
a number of polypeptides are affected by 
the transfer of high-C0 2 -grown algae to 
air levels of inorganic carbon. There is a 
transitory decline in the rate of synthesis 
of pre-existing proteins such as the 
holoenzyme of RuBP carboxylase, with 
perhaps a decline specifically in the rate 
of synthesis in the large subunit. There 
is the transient, rapid synthesis of poly- 
peptides with undetermined function. And 
finally, the decreased synthesis of the 
large subunit might lead to the accu- 

mulation of the small subunit precursor 
in the cytoplasm of the cell. This last point 
is important to our understanding of the 
biosynthesis of plastid components and 
suggests a lack of coordination between 
the synthesis of the large and small sub- 
units of RuBP carboxylase. 


Grossman, A. R., S. G. Bartlett, G. W. Schmidt, 
J. E. Mullet, and N.-H. Chua, J. Biol. Chem. 
257, 1558-1563, 1982. 



Porphyridium aerugineum AND Cyanophora paradoxa 

Arthur Grossman, Lawrence Talbott, and Thomas Egelhoff 

Phycobilisomes are major light-har- phora paradoxa. P. aerugineum is a red 
vesting complexes in both prokaryotic and alga thought to have evolved following 
eukaryotic algae (Bogorad, 1975; Gantt, the invasion of a unicellular protist by a 
1981). The pigmented polypeptides, or cyanobacterium (Harington and Thorn- 
phycobiliproteins, of this complex (phy- ley, 1982). An intermediate in this evo- 
coerythrin or phycoerythrocyanin, phy- lutionary process may be represented by 
cocyanin, and allophycocyanin) serve to Cyanophora paradoxa (Jaynes and Ver- 
harvest light energy and transfer it to non, 1982). The plastid of Cyanophora 
the photosynthetic reaction centers. Each paradoxa closely resembles a cyanobac- 
of these pigmented polypeptides is com- terium. This organelle, termed "cy- 
posed of an a and a p subunit (Gantt, anelle," is surrounded by a thin 
1981). The a subunits of the different peptidylglycan cell wall and is located in 
phycobiliproteins are related, as are the the vacuole of the host organism (Aitkin 
p subunits (Glazer, 1977). Furthermore, and Stanier, 1979). While this arrange- 
the a and p subunits are related to each ment suggests a symbiotic association, 
other and probably arose via a gene du- the cyanelle cannot be cultured indepen- 
plication (Glazer, 1977). In addition to dently, and the DNA within the cyanelle 
phycobiliproteins, nonpigmented poly- is the size of chloroplast DNA and not 
peptides are integral constituents of the the size of blue-green algal DNA (Herd- 
phycobilisome (Tandeau de Marsac and man and Stanier, 1977; Mucke et at. , 1980). 
Cohen-Bazire, 1977). These nonpig- The reduced size of the DNA within the 
mented polypeptides (linker proteins) cyanelle suggests that the functional as- 
function in the maintenance of phycobi- pects of this organelle depend upon ge- 
lisome structure (Glazer, 1982). The linker netic information located in the nucleus 
proteins have been most thoroughly of the host organism. We have compared 
studied in cyanobacteria (Lundell et al. y the origin of phycobilisome constituents 
1981a, 1981b). In Synechococcus 6301 in P. aerugineum and Cyanophora par- 
there are four linker polypeptides (see adoxa, and we find that these two or- 
Glazer, 1982, for a review). These poly- ganisms exhibit many similarities with 
peptides are located in specific positions respect to the sites of synthesis of phy- 
along the phycobilisome and serve to hold cobilisome polypeptides, 
the hexameric arrays of pigmented poly- 
peptides together. 

We are studying the biosynthesis of Mfthods 

the phycobilisomes in eukaryotic algae 

and hope to determine the compartment Porphyridium aerugineum (UTEX 

in which each phycobilisome constituent 755) was grown in a modified Bristol's 

is synthesized, the arrangement of the medium (Egelhoff and Grossman, 1983), 

genes encoding these constituents on both and Cyanophora paradoxa, kindly pro- 

the plastid and nuclear genomes, and the vided by L. Provasoli, Yale University, 

factors important in the regulation of these was grown in Schenk's medium (Schenk, 

genes. In this report, we identify the sites 1977). Both organisms were grown at 22°C 

of synthesis of the specific phycobilisome and bubbled continuously with a mixture 

polypeptides in two eukaryotic algae, of 5% C0 2 and air. Illumination was from 

Porphyridium aerugineum and Cyano- fluorescent tubes (100 |xmol m" 2 s" 1 ). 


For in vivo labeling of C. paradoxa, tween 15,000 and 95,000. The band with 

100 ml of mid-log-phase cells were cen- the highest molecular weight (band 1) is 

trifuged at 4,000 rpm (Sorvall SS34 ro- thought to be the anchor protein (Red- 
tor) and the pellets washed with 20 ml 

of 20 mM MOPS, pH 7.7, 10 mM NaCl, 
1 mM MgCl 2 , and resuspended in 40 ml 
of the wash medium. P. aerugineum cul- 
tures were labeled in modified Bristol's 
medium minus sulfate and soil extract. 
We used 300 fxg/ml chloramphenicol for 
inhibition of translation on 70S ribosomes 
and 1 (jug/ml cycloheximide for inhibition 
of translation on 80S ribosomes. The cul- 
tures, both with and without inhibitors, 
were bubbled with air for 10 min in the 
light at 22°C prior to the addition of la- 
beled S0 4 ~. We used 0.5 mCi of 35 S0 4 " 
per tube of P. aerugineum and 1.0 mCi 
of 35 S0 4 ~ per tube of C. paradoxa. 

Following in vivo labeling, the cells 
were washed in 0.65 M phosphate buffer, 
pH 7.0, and the phycobilisomes isolated 
according to the method of Williams et 
al. (1980), as modified (see preceding re- 
port). After collecting the intact phyco- 
bilisome band from sucrose gradients, the 
phycobilisomes were diluted 4-5 times 
with 0.75 M phosphate buffer, pH 8.0, 
and pelleted for 2 h at 45,000 rpm in a 
50 Ti rotor. The pellets were solubilized 
as before, and treated with Vz vol of 
5% sodium dodecylsulfate, 30% sucrose, 
0.1% bromphenol blue, and then electro- 
phoresed on 12-18% polyacrylamide gra- 
dient gels containing 8 M urea. Molecular 
weight standards were also run. After 
staining with Coomassie brilliant blue G- 
250, destaining, and fluorography (Bon- 
ner and Laskey, 1974), the gels were dried 
and exposed to Kodak XAR-5 film. 

12 3 4 5 


Several polypeptide constituents of 
phycobilisomes of both P. aerugineum 
and C. paradoxa were resolved by so- 
dium dodecylsulfate, polyacrylamide gel 
electrophoresis. The profile of phycobi- 
lisome polypeptides of P. aerugineum is 
presented in Fig. 79, lane 1. These poly- 
peptides have molecular weights be- 

Fig. 79. Synthesis of phycobilisome polypep- 
tides of Porphyridium aerugineum. Growth of P. 
aerugineum, in vivo labeling with and without in- 
hibitors of translation, phycobilisome isolation, and 
polyacrylamide gel electrophoresis and fluorogra- 
phy are described in the text. (1) Stained profile of 
P. aerugineum phycobilisome polypeptides. 

(2) Phycobilisome polypeptides labeled in vivo. 

(3) As in (2) but in the presence of cycloheximide. 

(4) As in (2) but in the presence of chloramphenicol. 

(5) As in (2) but in the presence of cycloheximide 
and chloramphenicol. 



linger and Gantt, 1982), while the major 
bands toward the bottom of the gel (bands 
9-12) are the pigmented components of 
the complex (a and (3 subunits of the phy- 
cobiliproteins). The anchor protein is in- 
volved in the attachment of the 
phycobilisome to the external surface of 
the thylakoid membranes, while the low- 
molecular-weight pigmented polypep- 
tides are the major light-absorbing com- 
ponents. Since the pigmented 
polypeptides have similar molecular 
weights, some of these broad bands may 
represent more than one species. All of 
the stained polypeptides of the phycobili- 
somes of P. aerugineum do incorporate 
label in vivo. Some bands that do not 
correspond to phycobilisome polypep- 
tides are also labeled. Generally, these 
polypeptides are less-intensely labeled 
than phycobilisome constituents and may 
represent contamination of the isolated 
light-harvesting complex with heavily la- 
beled cytoplasmic polypeptides. Minor 
stained bands such as bands 3, 4, and 5 
of Fig. 79 may also be cytoplasmic con- 
tamination of phycobilisome prepara- 
tions. Some bands may also result from 
a small amount of proteolysis (in spite of 
the inclusion of protease inhibitors in all 
solutions used in the isolation). For ex- 
ample, band 2 of P. aerugineum phyco- 
bilisomes may result from proteolytic 
degradation of the anchor protein. This 
band, like band 1, has very faint pig- 
mentation (both are blue), and while we 
generally find that band 1 predominates 
over band 2, in other published profiles 
the lower band (band 2) appears to pre- 
dominate (Morschel, 1982). 

Phycobilisome polypeptides labeled in 
the presence of cycloheximide and chlor- 
amphenicol are presented in lanes 3 and 
4 of Fig. 79, respectively, while lane 5 
shows labeling in the presence of both 
inhibitors. Cycloheximide inhibits trans- 
lation on 80S cytoplasmic ribosomes, while 
chloramphenicol inhibits translation of 70S 
chloroplast ribosomes. We observe a re- 
ciprocal labeling pattern of the phycobili- 
some polypeptides in the presence of these 
inhibitors. If both inhibitors are included 

during in vivo labeling, no phycobilisome 
constituents are synthesized (Fig. 79, lane 
5). Comparing lanes 3 and 4 suggests that 
the anchor protein (band 1) and the major 
pigmented bands (bands 9-12) are trans- 
lated on chloroplast ribosomes, while band 
6, a linker polypeptide (Morschel, 1982), 
is synthesized in the cytoplasm of the 
cell. While the synthesis of polypeptides 
7 and 8 is lowered in the presence of ei- 
ther inhibitor, longer exposures of the 
fluorographed gel reveal that both of these 
polypeptides are made on the 70S ribo- 
somes of the chloroplast. The decreased 
labeling of those polypeptides that we 
observe in the presence of cycloheximide 
may be the result of poor integration of 
these components into intact phycobili- 
somes without the synthesis of cyto- 
plasmic phycobilisome polypeptides. 

In Fig. 80, the polypeptide profiles of 
the phycobilisomes of C. paradoxa are 
presented. The profile of the stained phy- 
cobilisome polypeptides is presented in 
lane 1, while labeled polypeptides syn- 
thesized in the absence and presence of 
protein synthesis inhibitors are in lanes 
2-5. The anchor protein of C. paradoxa 
appears to be degraded into bands 1 and 
2 during the isolation procedure, and the 
amount of each species varies depending 
upon the speed of isolation. The major 
pigmented polypeptides are bands 7-10. 
At least some of the polypeptides local- 
ized between the anchor protein and the 
pigmented polypeptides, and ranging in 
molecular weight from 34,000 to approx- 
imately 55,000, may represent linker 
polypeptides. Although there appears to 
be contamination of the C. paradoxa 
phycobilisome preparation with some 
stainable bands (note the high-molecu- 
lar-weight band above the anchor pro- 
tein), the origin of phycobilisome 
polypeptides of C. paradoxa seems to be 
similar to those of P. aerugineum as well 
as to those of P. omentum and C. cal- 
darium (Egelhoff and Grossman, 1983). 
The anchor protein is synthesized in the 
chloroplast (represented by band 1 and 
probably a proteolytic product of this 
polypeptide, band 2), since it is made in 



12 3 4 5 

sized in the chloroplast, as determined 
by inhibitor sensitivity. On the other hand, 
the possible linker polypeptides (bands 
3-6) located between the anchor protein 
and the pigmented polypeptides are syn- 
thesized on 80S ribosomes. 



Fig. 80. Synthesis of phycobilisome polypep- 
tides of Cyanophora paradoxa. Procedures used 
are in the legend of Fig. 79. (1) Stained profile of 
C. paradoxa phycobilisome polypeptides. 

(2) Phycobilisome polypeptides labeled in vivo. 

(3) As in (2) but in the presence of cycloheximide. 

(4) As in (2) but in the presence of chloramphenicol. 

(5) As in (2) but in the presence of cycloheximide 
and chloramphenicol. 

the presence of cycloheximide but not in 
the presence of chloramphenicol (com- 
pare Fig. 80, lanes 3 and 4). The other 
labeled polypeptides in the region of bands 
1 and 2 do not correspond to stained poly- 
peptides and probably represent cyto- 
plasmic contamination of the 
phycobilisome preparation. The major 
pigmented polypeptides are also synthe- 


Recently we examined the sites of syn- 
thesis of phycobilisome-containing eu- 
karyotic algae (Egelhoff and Grossman, 
1983); we concluded that the major pig- 
mented polypeptides are synthesized in 
the chloroplast, as is the high-molecular- 
weight protein that holds the phycobili- 
some to the external surface of the thy- 
lakoid membranes. Linker proteins (other 
than the anchor protein), which are those 
polypeptides of the phycobilisome which 
hold the arrays of pigmented components 
together, are synthesized in the cyto- 
plasm. Since polypeptides synthesized on 
the 70S ribosomes of the chloroplast have 
thus far been found to be encoded by 
chloroplast DNA (Ellis, 1981), we con- 
clude that the gene for the anchor protein 
of the phycobilisome as well as the genes 
for the pigmented polypeptides are lo- 
calized to the chloroplast DNA. Fur- 
thermore, since the phycobiliproteins are 
encoded by closely related genes (see 
Glazer, 1977, for a more complete dis- 
cussion), we suggest that a phycobilipro- 
tein gene family is on the chloroplast 
DNA. The arrangement of these genes 
on the chloroplast genome and the co- 
ordination of their regulation with those 
phycobilisome constituents made in the 
cytoplasm are of primary interest to us. 

Cyanophora paradoxa is a unicellular 
protist with an organelle having features 
similar to a cyanobacterium. The char- 
acteristics of the photosynthetic appa- 
ratus of the cyanelle are strikingly similar 
to those observed for cyanobacteria (Klein 
et al. f 1981). On the other hand, the cy- 
anelle cannot be cultured as an indepen- 
dent organism, and the size of the cyanelle 
DNA is similar to the size of DNA found 
in plastids of higher plants (Jaynes et al. , 



1981). Therefore, the cyanelle has lost 
the genetic potential to exist indepen- 
dently of its biflagellate host. Since it has 
been suggested that C. paradoxa may 
represent an evolutionary bridge be- 
tween cyanobacteria and chloroplasts (see 
Jaynes and Vernon, 1982, for review), 
we thought that the sites of synthesis of 
the phyobilisome polypeptides might be 
different from the sites of synthesis of 
phycobilisome constituents of red algae 
and Cyanidium caldarium. In this re- 
port we show that the biosynthesis of C. 
paradoxa phycobilisomes does not differ 
in general aspects from the biosynthesis 
of phyobilisomes of other eukaryotic al- 
gae (Egelhoff and Grossman, 1983). In 
spite of serious proteolysis of the anchor 
protein, our data indicate that both the 
anchor protein and the major pigmented 
polypeptides are synthesized within the 
cyanelle and are probably encoded on the 
plastid DNA. Other polypeptides possi- 
bly analogous to linker polypeptides (ex- 
tensively studied in blue-green algae) are 
synthesized in the cytoplasm of the or- 
ganism. The location and arrangement of 
the genes for these polypeptides and the 
specific changes that these genes have 
undergone during their transfer from the 
invading endosymbiont to the nucleus of 
the host organism will strengthen our 
understanding of gene transfer and the 
development of eukaryotic chloroplasts. 


Aitkin, A., and R. Stanier, J. Gen. Microbiol. 112, 

219-233, 1979. 
Bogorad, L., Annu. Rev. Plant Physiol. 26, 369- 

401, 1975. 
Bonner, W. M., and R. A. Laskey, Eur. J. Biochem. 

U6, 83-88, 1974. 
Egelhoff, T., and A. R. Grossman, Proc. Nat. Acad. 

Sci. USA. 80, 3339-3343, 1983. 
Ellis, R. J., Annu. Rev. Plant Physiol. 32, 111— 

137, 1981. 
Gantt, E.,Annu. Rev. Plant Physiol. 32, 327-347, 

Glazer, A. N., Annu. Rev. Microbiol. 36, 171-196, 

Glazer, A. N., Photochem. Photobiol. Rev. 1, 71- 

115, 1977. 
Harington, A., and A. L. Thornley, J. Mol. Evol. 

18, 287-292, 1982. 
Herdman, M., and R. Y. Stanier, FEBS Lett. 1, 

7-12, 1977. 
Jaynes, J. M., and L. P. Vernon, Trends Biochem. 

Sci. 7, 22-25, 1982. 
Jaynes, J. M., L. P. Vernon, S. M. Klein, and G. A. 

Strobel, Plant Sci. Lett. 21, 345-356, 1981. 
Klein, S., J. M. Jaynes, S. S. Kent, and L. P. Ver- 
non, Plant Physiol. 68, 407-410, 1981. 
Lundell, D. J., R. C. Williams, and A. N. Glazer, 

J. Biol. Chem. 256, 3580-3592, 1981a. 
Lundell, D. J., G. Yamanaka, and A. N. Glazer, 

J. Cell. Biol. 91, 315-319, 1981b. 
Morschel, E., Planta 151>, 251-258, 1982. 
Mucke, H., W. Loffelhardt, and H. J. Bohnert, 

FEBS Lett. Ill, 347-352, 1980. 
Redlinger, T., and E. Gantt, Proc. Nat. Acad. Sci. 

USA. 79, 5542-5546, 1982. 
Schenk, H., Arch. Microbiol. 1U, 261-266, 1977. 
Tandeau de Marsac, N., and G. Cohen-Bazire, Proc. 

Nat. Acad. Sci. USA. 7k, 1635-1639, 1977. 
Williams, R. C., J. C. Gingrich, and A. N. Glazer, 

/. Cell Biol. 85, 558-566, 1980. 



Arthur Grossman and Jerry Brand 

Only a few different methods have been 
employed in the isolation of intact phy- 
cobilisomes from blue-green and red al- 
gae. The procedure used most extensively 
involves the breakage of cells in high con- 
centrations of phosphate buffer (0.65-0.75 
M, NaH 2 P0 4 /K 2 HP0 4 , pH 7.0-8.0) fol- 
lowed by incubation in 1-2% Triton X- 

100 and sucrose gradient centrifugation 
(Gantt and Lipschultz, 1972; Gantt et al., 
1979; Williams et al., 1980). The method 
is time-consuming, involves a long ultra- 
centrifugation step, and is limited in the 
amount of sample that can be processed. 
There are other, less frequently used 
methods to isolate phycobilisomes. So- 


dium sulfate has been used in place of Materials and Methods 
phosphate buffer when CaCl 2 was re- 
quired in the isolation medium (Yaman- All chemicals used, unless otherwise 
akaetal., 1982). Another method involves indicated, were reagent grade. The ac- 
the pelleting of phycobilisomes with 15% rylamide (Bio Rad) was electrophoresis 
polyethylene glycol 6000 (Rigbi et al., grade. The N'N-methylenebisacryl- 
1980; Rosinski et al., 1981). This proce- amide was from Kodak. Sequanal-grade 
dure, like the one described below, is rel- sodium dodecylsulfate was from Pierce 
atively rapid and does not require and was further purified with neutral- 
prolonged ultracentrifugation. ized, activated charcoal. Glycine and 

A similarity among these procedures Trizma base used in electrophoresis buff- 
is the reduced water activity to which ers were purchased from Sigma. Ana- 
the phycobilisome is exposed, as has been cystis nidulans (UTEX 20), 
pointed out by Glazer (1982). With in- Porphyridium aerugineum, and Cyan- 
creased water activity, energy transfer idium caldarium were grown under con- 
between the different classes of pig- tinuous illumination and bubbled with 5% 
mented polypeptides is uncoupled as the C0 2 and 95% air. A. nidulans was grown 
phycobilisome begins to dissociate. Low in Kratz and Myers medium (1955) at 37°C. 
temperature also promotes dissociation P. aerugineum was grown in a modified 
of phycobilisomes (Gantt et al., 1979). Bristol's medium (Egelhoff and Gross- 
Exposure of phycobilisomes to low-phos- man, 1983) at 23°C, and C. caldarium 
phate buffer and/or low temperature for was grown in Allen's medium (1959) at 
short periods of time has been used to 39°C. 

cause partial dissociation of the phyco- Phycobilisomes were isolated in two 
bilisome and has yielded information about ways. The first was a standard proce- 
the structural organization of this light- dure, employed for comparative anal- 
harvesting complex (Morschel, 1982; yses. This method (Williams and Glazer, 
Myeong-Hee Yu et al., 1981). Further- 1980) involved the disruption of cells in 
more, by careful adjustment in the con- high-phosphate buffer (0.65 M, pH 7.0) 
centration of phosphate buffer, and the subsequent purification of intact 
reconstitution of phycobilisomes (both phycobilisomes by ultracentrifugation 
homologous and heterologous) from in- through sucrose step gradients. The phy- 
dividual constituents has been achieved cobilisomes, which collected at the 0.75- 
(Canaani and Gantt, 1982). 1.0 M interface, were diluted with 0.75 

By varying the concentrations of phos- M phosphate buffer, pH 7.5, and pelleted 

phate buffer to which the phycobilisomes by centrifugation for 2 h at 130,000#. 

are exposed, we have developed a simple The second method of isolation (rapid- 

and rapid method for the isolation of this pelleting procedure) involved breakage 

light-harvesting complex. The yield ob- of cells in 1 M phosphate buffer with sub- 

tained using this technique is either as sequent low-speed centrifugations to iso- 

high or higher than yields obtained using late the intact light-harvesting complex, 

standard procedures. The isolation can Cells suspended in 1 M NaKP0 4 buffer, 

be performed in 3-4 hours and does not pH 7.5, were lysed by passage through 

require ultracentrifugation. We have the French pressure cell (15,000 psi). To 

employed this technique in the isolation achieve more-complete disruption of C. 

of phycobilisomes from the prokaryotic caldarium, these cells were passed 

organism Anacystis nidulans {Synecho- through the French pressure cell twice. 

coccus 6301) and the eukaryotic organ- Thelysate was made to 1% Triton X-100, 

isms Porphyridium aerugineum and incubated at room temperature for 30 min, 

Cyanidium caldarium, but have not and then centrifuged in the Sorvall RC5B 

found it useful in the isolation of phyco- (SS34 rotor) at 20,000 rpm for 30 min. 

bilisomes from Porphyridium cruentum. The initial centrifugation caused both the 



membranous material and the phycobi- 
lisomes to pellet. The supernatant, con- 
taining the soluble cytoplasmic 
polypeptides as well as the stromal poly- 
peptides, was discarded, and the pelleted 
material was resuspended in 0.6 M phos- 
phate buffer, pH 7.5. For homogeneous 
resuspension the pellet was dispersed with 
a ground-glass homogenizer. This sus- 
pension was incubated for 30 min after 
the addition of Triton X-100 to 1% and 
then centrifuged at 20,000 rpm for 30 min. 
During this centrifugation the mem- 
branes became pelleted while a large 
proportion of intact phycobilisomes re- 
mained in solution. The pellet was dis- 
carded and the supernatant diluted tenfold 
with 1.0 M NaKP0 4 , pH 7.5, and spun 
at 20,000 rpm for 1 h. Intact phycobili- 
somes became pelleted under these con- 

Phycobilisome pellets obtained by ei- 
ther method described above were sol- 
ubilized in 0.1 M Na 2 C0 3 , 0.1 M 
dithiothreitol, and then treated with Vz 
vol of 5% sodium dodecylsulfate, 30% su- 
crose, and 0.1% bromphenol blue. The 
samples were boiled for 1 min prior to 
electrophoresis on 12-18% poly aery 1- 
amide gradient gels containing 8 M urea. 
Molecular-weight standards were phos- 
phorylase b (92,500), bovine serum al- 
bumin (66,200), ovalbumin (45,000), 
carbonic anhydrase (31,000), soybean 
trypsin inhibitor (21,500), and lysozyme 
(14,400). After electrophoresis the gels 
were stained with Coomassie brilliant blue 

Fluorescence emission of phycobili- 
some preparations was analyzed follow- 
ing excitation of phycocyanin with broad- 
band blue-green light, produced by pass- 
ing light from a type DTS 125W tungsten 
lamp (21.5V) through two CS 4-96 (Corn- 
ing) niters and one Calflex C (Balzers) 
heat-reflecting filter. Fluorescence emis- 
sion spectra (600-850 nm) were mea- 
sured with a microprocessor-based 
spectrofluorimeter equipped with a tri- 
furcated fiber-optic bundle (Fork, Ford, 
and Catanzaro, Year Book 78, 196-199) 
which permitted illumination and fluo- 

rescence detection at the same surface 
of the sample. The photomultiplier was 
protected from the exciting light by the 
cutoff filter CS 2-64. 


Polypeptide profiles of the phycobili- 
somes of A. nidulans, P. aerugineum, 
and C. caldarium isolated by sucrose 
gradient centrifugation or via the rapid, 
pelleting procedure are presented in Fig. 
81. The polypeptide profiles obtained for 

1 2 3 4 5 6 




Fig. 81. Polypeptide profiles of phycobilisomes 
isolated from Porphyridium aerugineum, Cyani- 
dium caldarium, and Anacystis nidulans. The iso- 
lation procedures and electrophoretic analyses are 
described in the text. Phycobilisomes of P. aeru- 
gineum, C. caldarium, and A. nidulans were iso- 
lated by sucrose gradient centrifugation (lanes 2, 
4, and 6, respectively) or by the rapid procedure 
described in the text (lanes 1, 3, and 5, respec- 



each organism by the two different meth- 
ods of isolation are similar. For C. cal- 
darium we find that the rapid-pelleting 
method eliminates some of the minor 
bands often observed after isolation by 
ultracentrifugation (compare Fig. 81, 
lanes 3 and 4). Furthermore, the phy- 
cobilisome yield is at least as high as for 
the standard method of isolation (data 
not shown). 

Although the polypeptide profiles of 
phycobilisomes obtained by the two 
methods are similar, we used fluores- 
cence emission spectroscopy to demon- 
strate the integrity of the complex after 
the rapid-pelleting procedure. This 
method of determining phycobilisome in- 
tegrity makes use of the observation that 
the intact light-harvesting complex effi- 
ciently transfers energy from the short- 
wavelength-absorbing biliproteins (phy- 
coerythrin or phycocyanin) to allophy- 
cocyanin and ultimately to the pigmented 
anchor protein or allophycocyanin B (see 
Glazer, 1982, for review). These latter 
two proteins fluoresce between 675 nm 
and 680 nm. Therefore, a phycobilisome 
preparation which absorbs wavelengths 
of light below 600 nm and fluoresces be- 
tween 675 and 680 nm has coupled en- 
ergy transfer. This indicates that the in 
vivo arrangement of the phycobilisome 
polypeptides has not been altered during 
the isolation procedure. 

Fluorescence emission analysis of P. 
aerugineum phycobilisomes obtained by 
the pelleting method is presented in Fig. 
82. Similar preparations from A. nidu- 
lans and C. caldarium gave comparable 
results. A broad-band filter with a cutoff 
at 620 nm (CS 4-96) and very low trans- 
mission above 600 nm was used to excite 
the isolated phycobilisome, and fluores- 
cence emission was measured between 
the wavelengths of 600 and 850 nm. As 
a control, the pelleted phycobilisomes 
were dissociated in low-phosphate buffer 
(30 mM, pH 7.0) prior to the measure- 
ment of fluorescence emission. The 
fluorescence emission from the phyco- 
bilisome preparation maintained in high- 
phosphate buffer had a peak at 677 nm. 


700 750 

Wavelength, nm 


Fig. 82. Fluorescence emission spectra of Por- 
phyridium aerugineum phycobilisomes at 23°C. 
Fluorescence emission analysis is described in the 
text. The phycobilisomes were isolated by the rapid 
isolation procedure and were either maintained in 
high-phosphate buffer (0.6 M, pH 7.0) (curve with 
677-nm peak) or dissociated in low-phosphate buffer 
(0.03 M, pH 7.0) (curve with 658-nm peak). 

Following dissociation in the low-phos- 
phate buffer, the emission broadened and 
peaked at 658 nm. This fluorescence 
emission data, in conjunction with the 
electrophoretic analysis of pelleted phy- 
cobilisomes, clearly indicates that this 
rapid isolation procedure is effective for 
the preparation of intact phycobilisomes 
from certain eukaryotic and prokaryotic 
algae. The phycobilisomes obtained are 
comparable to phycobilisomes isolated via 
sucrose gradient centrifugation and in 
some cases show less cytoplasmic con- 


We have described a method for the 
rapid isolation of intact phycobilisomes 
from both prokaryotic and eukaryotic al- 
gae. This procedure does have some lim- 
itation, however, and we have not been 
able to use it for isolating the larger, more- 
complex phycobilisomes of P. omentum. 
The isolation entails lysis in 1.0 M phos- 
phate buffer followed by a series of low- 
speed centrifugations, resulting in a phy- 



cobilisome pellet. These phycobilisomes 
appear to be intact, as determined by 
their polypeptide profiles and fluores- 
cence emission analysis. For certain ap- 
plications this method has important 
advantages over the widely used isola- 
tion procedure, which employs sucrose 
gradient centrifugation. First, it is very 
rapid and requires no ultracentrifuga- 
tion. It can be used for very-large-scale 
preparations of phycobilisomes. This lat- 
ter characteristic has made it the method 
of choice for the isolation of phycobili- 
somes to be further fractionated into in- 
dividual polypeptides for use in antibody 
preparations. Finally, it facilitates the 
isolation of phycobilisomes from many 
different samples and allows for the rapid 
characterization of this light-harvesting 
complex either from different organisms 
or from the same organism grown under 
different environmental conditions. 


Allen, M. B., Arch. Mikrobiol. 32, 270-277, 1959. 
Canaani, 0., and E. Gantt, Proc. Nat. Acad. Sci. 

USA 79, 5277-5281, 1982. 
Egelhoff, T., and A. R. Grossman, Proc. Nat. Acad. 

Sci. USA 80, 3339-3343, 1983. 
Gantt, E., and C. A. Lipschultz, /. Cell. Biol. 54, 

313-324, 1972. 
Gantt, E., C. A. Lipschultz, J. Grabowski, andB. K. 

Zimmerman, Plant Physiol. 63, 615-620, 1979. 
Glazer, A. N.,Annu. Rev. Microbiol. 36, 171-196, 

Kratz, W. A., and J. Myers, Amer. J. Bot. 42, 282- 

287, 1955. 
Morschel, E., Planta 154, 251-258, 1982. 
Myeong-Hee Yu, A. N. Glazer, and R. C. Wil- 
liams, J. Biol. Chem. 256, 13130-13136, 1981. 
Rigbi, M., J. Rosinski, H. W. Siegelman, and J. C. 

Sutherland, Proc. Nat. Acad. Sci. USA 77, 1961- 

1965, 1980. 
Rosinski, J., J. F. Hainfeld, M. Rigbi, and H. W. 

Siegelman, Ann. Bot. 47, 1-12, 1981. 
Williams, R. C, J. C. Gingrich, and A. N. Glazer, 

J. Cell Biol. 85, 558-566, 1980. 
Yamanaka, G., D. J. Lundell, and A. N. Glazer, 

/. Biol. Chem. 257, 4077-4086, 1982. 




Arthur Grossman 

Phycobilisomes are macromolecular 
aggregates which function in harvesting 
light energy and transferring it to the 
photosynthetic reaction centers. They are 
composed of colored proteins and linker 
proteins. The linker proteins are re- 
quired for assembly of the pigmented 
polypeptides and are essential structural 
components of the complex (Glazer, 1982). 
Phycobilisomes are generally isolated as 
intact complexes on sucrose gradients in 
high-phosphate buffer. Upon electropho- 
resis of the intact phycobilisome on de- 
naturing polyacrylamide gels, the 
constituent polypeptides dissociate and 
the molecular weights of the individual 
components generally range in size from 
12,000 to 95,000 (Gantt, 1981). The high- 
est-molecular-weight polypeptide (an- 
chor protein) serves in the attachment of 

the complex to the thylakoid membranes 
(Redlinger and Gantt, 1981), while the 
low-molecular-weight polypeptides 
(12,000-20,000) are the major pigmented 
polypeptides. In the phycobilisomes of 
Porphyridium omentum, three or four 
other pigmented polypeptides with mo- 
lecular weights between 25,000 and 35,000 
have been visualized (Redlinger and 
Gantt, 1981). While the association of 
various components has been examined 
by several methods (partial dissociation 
followed by sucrose gradient centrifu- 
gation [Morschel, 1982], analysis of phy- 
cobilisome mutants [Gingrich et al., 1982]), 
no report of high-molecular-weight ag- 
gregates of phycobilisome polypeptides 
on sodium dodecylsulfate polyacrylamide 
gels has appeared. Here we note that if 
the large hemi-ellipsoidal phycobilisomes 



of the red alga P. omentum are elec- 
trophoresed on 7.5-15% polyacrylamide 
gradient gels in the cold, large complexes 

containing phycobilisome polypeptides can 
be seen. While only preliminary work has 
been done on the nature of these aggre- 
gates, they appear to contain both the 
major pigmented polypeptides and the 
less-prominent pigmented components, 
which range in molecular weight from 
25,000 to 35,000. 


Phycobilisomes of P. cruentum were 
isolated according to the method of Wil- 
liams et al. (1980) except that protease 
inhibitors (1 mM phenylmethylsulfonyl 
fluoride, 1 mM benzamidine-HCl, 5 mM 
e-amino-n-caproic acid) were added to all 
solutions. The concentrations of sucrose 
used in the step gradients were 0.5 M, 
0.75 M, 1.0 M, and 2.0 M. Centrifugation 
was overnight at 18°C in an SW27 rotor. 
The phycobilisomes, which collected at 
the 1.0/2.0 M sucrose interface, were di- 
luted 3-4-fold with 0.75 M phosphate 
buffer, pH 7.5, and then pelleted for 2 h 
at 45,000 rpm in a Beckman Ti50 rotor. 
The pellet was resuspended in 0.1 M 
Na 2 C0 3 , 0.1 M dithiothreitol, and treated 
with V2 vol of 5% sodium dodecylsulfate, 
30% sucrose, 0.1% bromphenol blue. Some 
samples were boiled for 1 min prior to 
loading them onto the gel. The electro- 
phoresis system used Laemmli buffers 
(Laemmli, 1970) and 7.5-15% polyacry- 
lamide slab gels. The samples were elec- 
trophoresed at 4°C and, at the completion 
of electrophoresis, stained with Coom- 
assie brilliant blue G-250. 

z x 




Fig. 83. Polypeptide profiles of Porphyridium 
cruentum phycobilisomes before and after heating. 
Phycobilisomes were isolated and prepared for 
electrophoresis as described in text. After treat- 
ment of the samples with sodium dodecylsulfate, 
they were either not heated (NH) or boiled for 1 
min (H). 

Results and Discussion 

In Fig. 83 we show a polypeptide pro- 
file of P. cruentum phycobilisomes. There 
is some contamination of the phycobili- 
some preparation with cytoplasmic com- 
ponents. The phycobilisomes were either 
heated (Fig. 83, H) or nonheated (Fig. 
83, NH). Two phycobiliprotein aggre- 

gates are present in the nonheated sam- 
ple toward the top of the gel. These 
aggregates are labeled A and B (Fig. 83, 
NH) and are bright-red bands prior to 
Coomassie-blue staining of the gel. Fre- 
quently, component B can be resolved 
into two distinct bands. Upon heating the 
phycobilisome sample (1 min of boiling), 
these aggregates dissociate. In the figure 



presented, dissociation of B is incomplete 
and the two distinct pigmented compo- 
nents (perhaps even three) of the band 
are resolved. Furthermore, following 
dissociation a number of components (C, 
D, E, F, and G in the figure) intensify. 
These polypeptides are all pigmented, 
with the components between 25,000 and 
35,000 (C and D) being only faintly col- 
ored. Component E contains several 
polypeptides and represents the major 
pigmented polypeptides of the complex. 
Therefore, it appears that at least the 
major pigmented aggregate (component 
B) contains primarily colored polypep- 
tides. Since A represents such a minor 
species, it is uncertain what components 
(other than at least some colored poly- 
peptides) are contained within it. 

While we have not been able to ob- 
serve phycobiliprotein aggregates fol- 
lowing electrophoresis of phycobilisomes 
isolated from a number of other organ- 
isms, it is clear that high-molecular- weight 
phycobiliprotein complexes from P. 
cruentum are stable to electrophoresis 
under these conditions. The apparent 
molecular weights of these aggregates, 
over 200,000, may not be accurate, since 
the protein-protein interactions of each 
complex may cause anomalous detergent 

binding. The nature of these complexes 
has not been clearly defined but the ma- 
jor constituents (at least for the B com- 
plexes) are pigmented polypeptides of 
both low and intermediate molecular 
weights. These aggregates may repre- 
sent trimers and/or hexamers of the phy- 
cobiliproteins, although we cannot 
eliminate the possibility that they are ar- 
tifactual associations which occur during 
sample preparation. The isolation of sta- 
ble aggregates of phycobilisome poly- 
peptides on sodium dodecylsulfate 
polyacrylamide gels and the subsequent 
analysis of these aggregates may prove 
helpful in elucidating the molecular ar- 
chitecture of the phycobilisome. 


Gantt, E., Annu. Rev. Plant Physiol. 32, 327-347, 

Gingrich, J. C, L. K. Blaha, and A. N. Glazer, J. 
Cell Biol. 92, 261-268, 1982. 

Glazer, A. N., Annu. Rev. Microbiol. 36, 173-198, 

Laemmli, U. K., Nature 227, 680-685, 1970. 

Morschel, E., Planta m, 251-258, 1982. 

Redlinger, T., and E. Gantt, Proc. 5th Interna- 
tional Congress Photosynthesis, HI, 257-262, G. 
Akoyunoglou, ed., Balaban International Sci- 
ence Services, 1981. 

Williams, R. C, J. C. Gingrich, and A. N. Glazer, 
J. Cell Biol. 85, 558-566, 1980. 


786 Anderson, J. A., J. S. Brown, and R. 
Malkin, Chlorophyll b: an integral 
component of photosystem I of higher 
plant chloroplasts, Photochem. Pho- 
tobiol., in press. 
Anderson, J. A., see Brown, J. S. 
Armond, P. A., see Badger, M. 

679 Badger, M., O. Bjorkman, and P. A. 
Armond, An analysis of photosyn- 
thesis response and adaptation to 
temperature in higher plants; tem- 
perature acclimation in the desert 
evergreen Nerium oleander L., 
Plant Cell Environ. 5, 85-99, 1982. 

831 Ball, J. T., J. Keeley, H. Mooney, J. 
Seemann, and W. Winner, Relation- 

ship between form, function, and 
distribution of two Arctostaphylos 
species (Ericaceae) and their puta- 
tive hybrids, Acta Ecol. J+, 153-164, 
692 Belford, H. S., W. F. Thompson, and 
D. B. Stein, DNA hybridization 
techniques for the study of plant ev- 
olution, in Phytochemistry and An- 
giosperm Phytogeny, Phytochemical 
Soc. Symp., 1-18, D. A. Young and 
D. S. Siegler, eds., Praeger, New 
York, 1982. 
Berry, J. A., see Farquhar, G. D.; 
Mooney, H. A.; Powles, S. B.; Rai- 
son, J. 



773 Bjorkman, 0., and S. B. Powles, Pho- 
toinhibition of photosynthesis: effect 
on chlorophyll fluorescence at 77 K 
in intact leaves and in chloroplast 
membranes of Nerium oleander, 
Planta 156, 97-107, 1982. 
Bjorkman, 0., see Badger, M.; Pearcy, 
R. W.; Powles, S. B.; Mooney, H. A. 

799 Blatt, M. R., and W. R. Briggs, Quan- 
titative microphotometry at the cel- 
lular level: A simple technique for 
measuring chloroplast movements 
in vivo, Planta, in press. 

823 Blatt, M. R., The action spectrum for 
chloroplast movements and evi- 
dence for blue-light-photoreceptor 
cycling in the alga Vaucheria, Planta, 
in press. 

810 Brand, J. , P. Mohanty, and D. C Fork, 
Reversible inhibition of photochem- 
istry of photosystem II by Ca 2+ re- 
moval from intact cells of Anacystis 
nidulans, FEBSLett. 155, 120-124, 

816 Briggs, W. R., and M. lino, Blue light- 
absorbing photoreceptors in plants, 
Phil. Trans. Roy. Soc. London Ser. 
B., in press. 
Briggs, W. R., see Hertel, R.; Man- 
doli, D. F.; Blatt, M. R.; Cooke, 
T. J.; Schaer, J.; Shinkle, J. R. 

778 Brown, J. S., J. M. Anderson, and 
L. H. Grimme, Antenna chlorophyll 
a complexes in mutant and devel- 
oping barley, Photosyn. Res. 3, 279- 
291, 1982. 
Brown, J. S, see Anderson, J. A. 
Cohen, D. B., see Palmer, J. D. 

801 Cooke, T. J., R. H. Racusen, and 
W. R. Briggs, Initial events in the 
tip-swelling response of the filamen- 
tous gametophyte of Onoclea sen- 
sibilis L. to blue light, Planta, in 
Dodd, J., see Palmer, J. D. 
Edwards, H., see Palmer, J. D. 

796 Egelhoff, T., and A. Grossman, Cy- 
toplasmic and chloroplast synthesis 
of phycobilisome polypeptides, Proc. 
Nat. Acad. Sci. USA 80, 3339-3343, 
Ehleringer, J., see Mooney, H. A. 
Everett, M., see Thompson, W. F. 

787 Farquhar, G. D., M. H. O'Leary, and 
J. A. Berry, On the relationship be- 
tween carbon isotope discrimination 
and the intercellular carbon dioxide 

concentration in leaves, Aust. J. 
Plant Physiol. 9, 121-137, 1982. 

Field, C, see Mooney, H. A. 
783 Fork, D.C., and K. Satoh, State I- 
state II transitions in the thermo- 
philic blue-green alga (cyanobacter- 
ium) Synechococcus lividus, 
Photochem. Photobiol. 37, 421-427, 

Fork, D. C, see Brand, J.; Satoh, K.; 
Sweeney, B. 

Grimme, H. L., see Brown, J. S. 

Grossman, A., see Egelhoff, T. 
779 Hertel, R., T. Lomax, and W. R. 
Briggs, Auxin transport in mem- 
brane vesicles from Cucurbita pepo 
L., Planta 157, 193-201, 1983. 

827 Hiesey, W. M., and M. A. Nobs, Ex- 
perimental studies on the nature of 
species. VI. Interspecific hybrid de- 
rivatives between facultatively 
apomictic species ofbluegrasses and 
their responses to contrasting envi- 
ronments, Carnegie Institution of 
Washington, Washington, D.C., 119 
pp., 1982. 

795 Hoshina, S., Temperature-induced 
change of chlorophyll a forms in 
phosphatidylcholine liposomes, Plant 
Cell Physiol. 21+, 937-940, 1983. 
lino, M., see Briggs, W. R. 
Jorgensen, R. A., see Palmer, J. D.; 
Thompson, W. F. 

800 Levitt, J., Plasmolysis shape in rela- 
tion to freeze-hardening of cabbage 
plants and to the effect of penetrat- 
ing solutes, Plant Cell Environ. 6, 
465-470, 1983. 
Lomax, T., see Hertel, R. 
Lonsdale, D. M., see Stern, D. 
Malkin, R., see Anderson, J. A. 

789 Mandoli, D. F., and W. R. Briggs, Fi- 
ber-optic plant tissue: Spectral de- 
pendence in dark grown and green 
tissues, Photochem. Photobiol., in 

815 Mandoli, D. F., and W. R. Briggs, 
Physiology and optics of plants in 
the soil, What's New in Plant Phys- 
iology, in press. 
Mandoli, D. F., see Schaer, J. 
Mohanty, P., see Brand, J. 

818 Mooney, H. A., J. Berry, O. Bjork- 
man, and J. Ehleringer, Compara- 
tive photosynthetic characteristics 
of coastal and desert plants of Cal- 



ifornia, Bol. Soc. Bot. Mexico 1+2, 19- 
33 1983 
826 Mooney, H. A., C. Field, W. E. Wil- 
liams, J. A. Berry, and 0. Bjork- 
man, Photosynthetic characteristics 
of plants of a California cool coastal 
environment, Oecologia 57, 38-42, 

Murray, M. G., see Spiker, S. 

Nobs, M. A., see Hiesey, W. M. 

O'Leary, M. H., see Farquhar, G. D. 

Orton, T. J., see Palmer, J. D. 

Osorio, B., see Palmer, J. D. 
782 Palmer, J. D., G. P. Singh, and 
D. T. N. Pillay, Structure and se- 
quence evolution of three legume 
chloroplast DNAs, Mol. Gen. Genet. 
190, 13-19, 1983. 

792 Palmer, J. D., Chloroplast DNA ex- 

ists in two orientations, Nature 301, 
92-93, 1983. 

793 Palmer, J. D., C. R. Shields, D. B. 

Cohen, and T. J. Orton, An unusual 
mitochondrial DNA plasmid in the 
genus Brassica, Nature 301, 725- 
728, 1983. 
777 Palmer, J. D., H. Edwards, R. A. Jor- 
gensen, and W. F. Thompson, Novel 
evolutionary variation in transcrip- 
tion and location of two chloroplast 
genes, Nucl. Acids Res. 10, 6819- 
6832, 1982. 

804 Palmer, J. D., B. Osorio, J. C. Wat- 

son, H. Edwards, J. Dodd, and W. F. 
Thompson, Evolutionary aspects of 
chloroplast genome expression and 
organization, in Photosynthesis, 
UCLA Symp. on Molecular and 
Cellular Biology, R. Hallick, ed., 
Alan R. Liss, Inc., New York, in 

805 Palmer, J. D., C. R. Shields, D. B. 

Cohen, and T. J. Orton, Chloroplast 
DNA evolution and the origin of am- 
phidiploid Brassica species, Theor. 
Appl. Genet. 65, 181-189, 1983. 
Palmer, J. D., see Stern, D.; Thomp- 
son, W. F. 

811 Pearcy, R. W., and 0. Bjorkman, 
Physiological effects, in Plant Re- 
sponses to More Carbon Dioxide, E. 
Leman, ed., AAAS, Washington, 
D.C., in press. 
Pillay, D. T. N., see Palmer, J. D. 
Polans, N. 0., see Thompson, W. F. 
770 Powles, S. B., J. A. Berry, and 0. 

Bjorkman, Interaction between light 
and chilling temperature on the in- 
hibition of photosynthesis in chilling 
sensitive plants, Plant Cell Envi- 
ron. 6, 117-124, 1983. 

Powles, S. B, see Bjorkman, 0. 

Racusen, R. H., see Cooke, T. J. 
754 Raison, J. K., J. K. M. Roberts, and 
J. A. Berry, Correlations between 
the thermal stability of chloroplast 
(thylakoid) membranes and the com- 
position and fluidity of their polar 
lipids upon acclimation of the higher 
plant, Nerium oleander, to growth 
temperature, Biochem. Biophys. 
Acta 688, 221-228, 1982. 

Roberts, J. K. M., see Raison, J. K. 
764 Satoh, K., and D. C. Fork, Chloro- 
phyll fluorescence transients as in- 
dicators of changes in the redox state 
of plastoquinone in intact Bryopsis 
chloroplasts, Plant Sci. Lett. 29, 133- 
144, 1983. 

784 Satoh, K., and D. C. Fork, State I- 

State II transitions in the green alga, 
Scenedesmus obliquus, Photochem. 
Photobiol. 37, 429-434, 1983. 

785 Satoh, K., and D. C. Fork, A mecha- 

nism for adaptation to changes in light 
intensity and quality in the red alga, 
Porphyra perforata. I. Relation to 
state I-state II transitions, Biochim. 
Biophys. Acta 722, 190-196, 1983. 

790 Satoh, K., and D. C. Fork, The rela- 

tionship between state II to state I 
transitions and cyclic electron flow 
around photosystem I, Photosyn. 
Res., in press. 

791 Satoh, K., and D. C. Fork, A new 

mechanism for adaptation to changes 
in light intensity and quality in the 
red alga, Porphyra perforata. III. 
Fluorescence transients in the pres- 
ence of DCMU, Plant Physiol. 71, 
673-676, 1983. 

794 Satoh, K., and D. C. Fork, A new 
mechanism for adaptation to changes 
in light intensity and quality in the 
red alga, Porphyra perforata. II. 
Characteristics of state I I-state III 
transitions, Photosyn. Res. 4, 61- 
70, 1983. 

797 Satoh, K., and D. C. Fork, Changes 
in distribution of light energy be- 
tween the two photosystems in spin- 
ach leaves, Photosyn. Res. 4, 71-79, 



820 Satoh, K., and D. C. Fork, Induction 
of MS delayed luminescence in the 
thermophilic blue-green alga Syne- 
chococcus lividus, in Photosynthetic 
Water Oxidation and PSII Photo- 
chemistry, Riken International 
Symposium, Y. Inoue, N. Murata, 
and Govindjee, eds., in press. 

812 Satoh, K., C. M. Smith, and D. C. 
Fork, Effects of salinity on primary 
processes of photosynthesis in the 
red alga, Porphyra perforata, Plant 
Physiol., in press. 
Satoh, K., see Sweeney, B.; Fork, D. C. 

798 Schaer, J., D. F. Mandoli, and W. R. 
Briggs, Phytochrome-mediated cel- 
lular photomorphogenesis, Plant 
Physiol. 72, 706-712, 1983. 
Shields, C. R., see Palmer, J. D. 

825 Shinkle, J. R., and W. R. Briggs, 
Auxin concentration/growth rela- 
tionship for Avena coleoptile sec- 
tions from seedlings grown in 
complete darkness, Plant Physiol., 
in press. 
Singh, G. P., see Palmer, J. D. 
Smith, C. M., see Satoh, K. 

780 Spiker, S., M. G. Murray, and W. F. 
Thompson, DNAase I sensitivity of 
transcriptionally active genes in in- 
tact nuclei and isolated chromatin of 

plants, Proc. Nat. Acad. Sci. USA 
80, 815-819, 1983. 
Stein, D. B., see Belford, H. S. 

821 Stern, D., J. D. Palmer, W. F. 
Thompson, and D. M. Lonsdale, Mi- 
tochondrial DNA sequence evolu- 
tion and homology to chloroplast 
DNA in angiosperms, Plant Mol. 
Biol. , UCLA Symposium on Molec- 
ular and Cellular Biology, R. B. 
Goldberg, ed., Alan R. Liss, Inc., 
New York, in press. 

781 Sweeney, B., D. C. Fork, and K. Sa- 
toh, Stimulation of bioluminescence 
in dinoflagellates by red light, Pho- 
tochem,. Photobiol. 4, 457-465, 1983. 

788 Thompson, W. F., M. Everett, N. 0. 
Polans, R. A. Jorgensen, and J. D. 
Palmer, Phytochrome control of 
RNA levels in developing pea and 
mung bean leaves, Planta, in press. 
Thompson, W. F., see Belford, H.; 
Palmer, J. D.; Spiker, S.; Stern, 
J. D. 

813 Vallejos, C. E., Enzyme activity 
staining, in Isozymes in Plant Ge- 
netics and Breeding , S. D. Tanksley 
and T. J. Orton, eds., Elsevier Pub- 
lications, Amsterdam, in press. 
Watson, J. C, see Palmer, J. D. 
Williams, W. E., see Mooney, H. A. 


Berry, Joseph A., Enzymatic fixation of car- 
bon dioxide in photosynthesis: Prospects 
for genetic engineering, Lecture to the 
Board of Trustees, Carnegie Institution of 
Washington, Washington, D.C., May 6, 

Bjorkman, Olle, How plants cope with ex- 
tremes of light, Botany Department, Uni- 
versity of California, Berkeley, November 
15, 1982. 

Bjorkman, Olle, How plants cope with ex- 
tremes of solar radiation, J. G. Wood Mem- 
orial Lectures, 23rd Annual Meeting of the 
Australian Society of Plant Physiologists, 
University of Queensland, Brisbane, May 
17, 1983. 

Bjorkman, Olle, Interactions between major 
stress factors on higher plant photosyn- 
thesis, Research School of Biological Sci- 
ences, Australian National University, May 
23, 1983. 

Brand, Jerry, Ca 2+ and Na + in photosyn- 

thesis, Department of Chemical Biodyn- 
amics, University of California, Berkeley, 
November 3, 1982. 

Brand, Jerry, Regulation by calcium of pho- 
tosystem II activity in Anacystis, Stan- 
ford-Carnegie Plant Science Seminar, 
Carnegie Institution, Stanford, California, 
November 10, 1982. 

Brand, Jerry, Calcium requirement for pho- 
tosystem II function in Anacystis nidu- 
lans, Gordon Conference, Physical-Chemical 
Aspects of Photosynthesis, Ventura, Cal- 
ifornia, February 8, 1983. 

Brand, Jerry, Ca 2+ and Na + in photosystem 
II of Anacystis nidulans, Hopkins Marine 
Station, Pacific Grove, California, Febru- 
ary 15, 1983. 

Briggs, Winslow R., Phototropism: A new 
look at an old problem, Biology Depart- 
ment, University of California, Los An- 
geles, February 2, 1983. 

Briggs, Winslow R., Blue light-absorbing 



photoreceptors in plants, Photoreception 
by Plants, Discussion Meeting, The Royal 
Society, London, England, March 9, 1983. 

Briggs, Winslow R., Henry Willard Coe State 
Park — A natural treasure of the inner coast 
range, California Native Plant Society, 
Santa Clara County Chapter, Santa Clara, 
California, May 26, 1983. 

Briggs, Winslow R., Blue light-photorecep- 
tors in higher plants and fungi, American 
Society for Photobiology, Madison, Wis- 
consin, June 29, 1983. 

Briggs, Winslow R., see Lomax, Terri. 

Brown, Jeanette S., Organization of chloro- 
phyll in photosynthetic membranes, Gor- 
don Conference, Ventura, California, 
February 7, 1983. 

Coleman, John R., Induction of carbonic an- 
hydrase in C hlamydomonas , Stanford- 
Carnegie Plant Science Seminar, Carnegie 
Institution, Stanford, California, March 2, 

Fork, David C, Energy distribution in pho- 
tosynthesis, Stanford-Carnegie Plant Sci- 
ence Seminar, Carnegie Institution, 
Stanford, California, May 11, 1983. 

Fork, David C, see Satoh, Kazuhiko. 

Grossman, Arthur, The biosynthesis of chlo- 
roplast components, Department of Biolog- 
ical Sciences, Stanford University, 
California, October 4, 1982. 

Grossman, Arthur, Uptake of polypeptides 
into isolated chloroplasts, Amoco Research 
Center, Naperville, Illinois, October 28, 

Grossman, Arthur, The biosynthesis of chlo- 
roplast components, Bay Area Plant Mo- 
lecular Biology Meeting, Berkeley, 
California, December 8, 1982. 

Grossman, Arthur, How to build a chloro- 
plast, Meeting of Stanford Affiliates, Stan- 
ford University, California, March 14, 1983. 

Grossman, Arthur, Study of molecular sys- 
tems in some eukaryotic algae, Photosyn- 
thesis Group, University of California, 
Berkeley, May 18, 1983. 

lino, Moritoshi, Phototropism in maize seed- 
lings: New perspectives on an old problem, 
Thimann Laboratories, University of Cal- 
ifornia, Santa Cruz, November 6, 1982. 

lino, Moritoshi, Phototropism in maize seed- 
lings: New perspectives on an old problem, 
Stanford-Carnegie Plant Science Seminar, 
Carnegie Institution, Stanford, California, 
February 9, 1983. 

Kaufman, Lon S., Regulation of chloroplast 
development in Euglena by a 600-nm pho- 
tomorphogenic system, Stanford-Carnegie 

Plant Science Seminar, Carnegie Institu- 
tion, Stanford, California, January 12, 1983. 

Levitt, Jacob, Biochemical mechanisms of ad- 
aptation of higher plants to changes in the 
environmental adaptations to cold and 
drought, Bern, Switzerland, March 14-15, 

Lomax, Terri, J. Shinkle, and W. R. Briggs, 
In vitro transport of [ 14 C]IAA by mem- 
brane vesicles from zucchini, 11th Inter- 
national Conference on Plant Growth 
Substances, Aberystwyth, Wales, United 
Kingdom, July 15, 1982. 

Lomax, Terri, In vitro auxin transport by 
sealed plasma membrane vesicles, Stan- 
ford-Carnegie Plant Science Seminar, Car- 
negie Institution, Stanford, California, June 
1, 1983. 

Ludlow, Mervyn M. , Ecological value of para- 
heliotropic leaf movement in water-stressed 
legumes, Department of Range Science, 
Utah State University, Logan, November 
1, 1982. 

Ludlow, Mervyn M. , Ecological value of para- 
heliotropic leaf movement in water-stressed 
legumes, Department of Biological Sci- 
ences, University of Utah, Salt Lake City, 
November 4, 1982. 

Mohanty, Prasanna, Inhibition and stimula- 
tion of photosystem II activity by metal 
ions, Stanford-Carnegie Plant Science 
Seminar, Carnegie Institution, Stanford, 
California, February 2, 1982. 

Mohanty, Prasanna, Calcium requirement for 
photosystem II function in Anacystis ni- 
dulans, Gordon Research Conference on 
Physiochemical Aspects of Photosynthesis, 
Ventura, California, February 8, 1983. 

Palmer, Jeffrey D., Evolutionary dynamics 
of chloroplast genes and genomes, Plant 
Breeding Institute, Cambridge, England, 
September 28, 1982. 

Palmer, Jeffrey D., Chloroplast DNA evo- 
lution: Molecular and phylogenetic studies, 
Department of Biology, Yale University, 
New Haven, Connecticut, October 6, 1982. 

Palmer, Jeffrey D., Evolutionary dynamics 
of chloroplast genes and genomes, Depart- 
ment of Botany, University of Massachu- 
setts, Amherst, November 15, 1982. 

Palmer, Jeffrey D., Evolutionary dynamics 
of chloroplast genes and genomes, Depart- 
ment of Botany, University of Illinois, Ur- 
bana, November 18, 1982. 

Palmer, Jeffrey D., Applications of chloro- 
plast DNA analysis to plant systematics 
and evolution, AIBS-American Society of 
Plant Taxonomists on Nucleic Acids and 



Plant Systematics, State College, Penn- 
sylvania, August 10, 1982. 

Palmer, Jeffrey D., and W. F. Thompson, 
Evolution of chloroplast genome structure, 
NATO Course on Structure and Function 
of Plant Genomes, Porto Portese, Italy, 
August 28, 1982. 

Palmer, Jeffrey D., Rearrangement and re- 
combination of plant cytoplasmic genomes, 
Division of Biological Sciences, University 
of Michigan, Ann Arbor, February 8, 1983. 

Palmer, Jeffrey D., Rearrangement and re- 
combination of plant cytoplasmic genomes, 
Department of Botany, University of Cal- 
ifornia, Berkeley, February 17, 1983. 

Palmer, Jeffrey D., Variation of plant cyto- 
plasmic DNA, Affiliates Meeting, Stanford 
University, California, March 14, 1983. 

Palmer, Jeffrey D., Rearrangement of plant 
cytoplasmic genomes, Department of Bi- 
ology, Washington University, St. Louis, 
Missouri, February 21, 1983. 

Palmer, Jeffrey D., Rearrangement of plant 
cytoplasmic genomes, Department of Bo- 
tany, Duke University, Durham, North 
Carolina, February 23, 1983. 

Palmer, Jeffrey D., Rearrangement and re- 
combination of plant cytoplasmic genomes, 
Stanford-Carnegie Plant Science Seminar, 
Carnegie Institution, Stanford, California, 
March 30, 1982. 

Palmer, Jeffrey D., Evolutionary aspects of 
chloroplast genome organization and 
expression, UCLA Symposia on Molecular 
and Cellular Biology, Keystone, Colorado, 
April 22, 1983. 

Satoh, Kazuhiko, and D. C. Fork, Induction 
of MS delayed luminescence in the ther- 
mophilic blue-green alga Synechococcus 
lividus, The Institute of Physical and 
Chemical Research Wako-shi, Saitama, Ja- 
pan, March 17, 1983. 

Satoh, Kazuhiko, and D. C. Fork, Changes 
in the intensity of fluorescence and state 
transitions in the red alga Porphyra per- 
forata, Ehime University, Ehime, Japan, 
April 2, 1983. 

Schafer, Eberhard, Phytochrome responses 
under continuous light, Stanford-Carnegie 
Plant Science Seminar, Carnegie Institu- 
tion, Stanford, California, October 13, 1982. 

Shinkle, James R., see Lomax, Terri. 

Stern, David, DNA sequence homologies be- 
tween mitochondrial and chloroplast gen- 
omes in angiosperms, Stanford-Carnegie 
Plant Science Seminar, Carnegie Institu- 
tion, Stanford, California, April 27, 1983. 

Stern, David, Organizational features of plant 

mitochondrial genomes, Gordon Confer- 
ence on Plant Molecular Biology, Andover, 
Massachusetts, June 17, 1983. 

Surzycki, Stefan J., see Watson, J. C. 

Thompson, William F., Phytochrome control 
of RNA levels in developing pea and mung 
bean, Plant Breeding Institute, Cam- 
bridge, England, October 15, 1982. 

Thompson, William F., Evolutionary dynam- 
ics of chloroplast DNA, Department of Plant 
Sciences, University College, Cardiff, 
Wales, October 27, 1982. 

Thompson, William F., Evolution of the chlo- 
roplast genome, Department of Biological 
Sciences, University of Warwick, England, 
November 5, 1982. 

Thompson, William F., Evolution of the chlo- 
roplast genome in angiosperms, Botany 
School, University of Cambridge, En- 
gland, February 1, 1983. 

Thompson, William F., Evolution of the chlo- 
roplast genome, John Innes Institute, Nor- 
wich, England, February 11, 1983. 

Thompson, William F., Evolution of chloro- 
plast DNA, Biochemistry Institute, Uni- 
versity of Bergen, Norway, February 23, 

Thompson, William F., Evolution and 
expression of the chloroplast genome, Plant 
Breeding Institute, Cambridge, England, 
March 16, 1983. 

Thompson, William F., Evolution and 
expression of the chloroplast genome, Di- 
vision of Biology and Medicine, Brown Uni- 
versity, Providence, Rhode Island, March 
19, 1983. 

Thompson, William F., Cytosine methylation 
and chromatin structure of ribosomal genes 
in wheat, Plant Molecular Biology, Gordon 
Conference, Andover, New Hampshire, 
June 13-17, 1983. 

Watson, John C, Chloroplast genes for 
translational components, Thimann Labo- 
ratories, University of California, Santa 
Cruz, October 12, 1982. 

Watson, John C, Chloroplast genes for 
translational components, Stanford-Car- 
negie Plant Science Seminar, Carnegie In- 
stitution, Stanford, California, October 13, 

Watson, John C, S. J. Surzycki, Both the 
chloroplast and nuclear genomes of Chla- 
mydomonas reinhardtii share homology 
with Escherichia coli genes for transcrip- 
tional and translational components, UCLA 
Symposium on the Biosynthesis of the Pho- 
tosynthetic Apparatus, Keystone, Colo- 
rado, April 22, 1983. 




Research Staff 

Joseph A. Berry 

Olle Bjorkman 

Winslow R. Briggs, Director 

Jeanette S. Brown 

David C. Fork 

C. Stacy French, Director Emeritus 

Arthur R. Grossman 

William M. Hiesey, Emeritus 

Malcolm A. Nobs, Emeritus 

William F. Thompson 

Research Associates 

Lon S. Kaufman 
Jeffrey D. Palmer 

Visiting Investigators 

Murray Badger, Australian National 
University, Canberra City, Australia 1 

Jacob Levitt, Senior Fellow, University 
of Minnesota, Minneapolis 

Postdoctoral Fellow 
John R. Coleman, NSERC, Canada 

Carnegie Institution of Washington 
Postdoctoral Fellows 

Jerry Brand, Senior Fellow, University 

of Texas at Austin 2 
John R. Coleman 
Graham D. Farquhar, Australian 

National University 3 
Satoshi Hoshina, Senior Fellow, 

Kanazawa University, Japan 
Moritoshi lino 
Mervyn M. Ludlow, Senior Fellow, 

CSIRO, Australia 4 
Prasanna Mohanty, Senior Fellow, 

Jawaharlal Nehru University, New 

Delhi, India 
Neil 0. Polans 
Kazuhiko Satoh, Senior Fellow, 

University of Tokyo, Japan 5 

J To October 31, 1982 
2 To March 24, 1983 
3 To February 14, 1983 
4 To December 31, 1982 
5 To September 30, 1982 

Eberhard Schafer, University of 

Freiburg, West Germany 6 
Susan C. Spiller 
C. Eduardo Vallejos 
Susanne von Caemmerer 7 
John C. Watson 

Graduate Students 

J. Timothy Ball, Stanford University 
Tobias I. Baskin, Stanford University 
Thomas Egelhoff, Stanford University 
David Hollinger, Stanford University 
Alan P. Maloney, Stanford University 
Dina F. Mandoli, Stanford University 8 
Elizabeth Newell, Stanford University 
Terri L. Lomax, Stanford University 
James R. Shinkle, Stanford University 
Celia M. Smith, Stanford University 
David B. Stern, Stanford University 
Mary Todd, Stanford University 
Philippe Tacchini, Stanford University 
Lawrence D. Talbott, Stanford University 

Undergraduate Student 

Helen E. Edwards, Stanford University 

Clerical and Technical Staff 

J. Timothy Ball, Laboratory Technician 9 
Marvin W. Fawley, Laboratory 

Technician 10 
Dorothy Ruth Fischer, Administrative 

Glenn A. Ford, Laboratory Manager 
John A. Gamon, Laboratory Technician 
Karen L. B. Hall, Laboratory Technician 
Richard W. Hart, Mechanical Engineer 11 
Eva L. Huala, Laboratory Technician 12 
Einar C. Ingebretsen, Electrical 

Douglas A. Jones, Laboratory Technician 
Jerome P. Lapointe, Laboratory 

Linda K. Morris, Laboratory Technician 
Frank Nicholson, Senior Technician 

6 To October 19, 1982 
7 To April 15, 1983 
8 To December 31, 1982 
9 To September 30, 1982 
10 To August 13, 1982 
n To March 31, 1983 
12 To June 28, 1983 


Bernardita Osorio, Laboratory Technician Rudolph Warren, Technician 

Norma J. Powell, Typist Aida E. Wells, Department Secretary 

Pedro F. Pulido, Technician Brian M. Welsh, Mechanical Engineer 

Maureen A. Simpson, Typist Nancy Beth White, Laboratory 

Mary A. Smith, Administrative Technician 13 

Assistant- Accountant 

James M. Tepperman, Technician 13 To June 24, 1983 

Department of Embryology 

Baltimore, Maryland 

Donald D. Brown 

+J .co -^ 
O Sh O --5 

r ^ ,J=! ■£ © 

cs o> o> jh s 

* ~ g PQ >§ 
^ ° ^ „roo 

0) cO 

3 « o 
Eh W i-s 

i*f.5 '' 

> X 

O i-r-S rtl 

•in (hW* 

■a >>^ .s s 

1 3 * S3 * S 

ctf <5 co k^j Cd 
*7 03 r==5 

s ^ B W 
ea m K 




« 2 2 


_ a Q> 

r- Sh " CO 

rs: o3 ft, <i .2 

bJO 5h Cj> ^^ 

o ^3 -K 





CD !> 

O O &H 


~ sj CD 0_i i? 
£ 0Q P 1 o >> 

tJ ^ — ■ 

en 3 ^ <i) 
«3 ,g O » 




-* _. 




oo -3 


£ c3 JS 0) 

H3 m -s ^ s 



>5 O 


bJO o 

C ±^ o3 Sh > 


g « JS a ai 

s ' 


i> a ?h -£ 

Q Q g pqtf 


Introduction 135 

Functioning and Repair of the Nervous System 137 

Regeneration of Synapses between Touch Sensory Neurons 138 

Accurate Regeneration of the Synapse between S-cells after Selective Destruction 

of the S-cell's Distal Segment 139 

Survival and Growth of Isolated Axon Segments after Destruction of Their En- 
sheathing Glial Cells 140 

Distinctions between Gap Junctions and Sites of Intermediate Filament Attach- 
ment in the Nervous System 141 

Leech Chemoreception 144 

Large Glia Are Not Dye-Coupled to Microglia 144 

Studies on Cell Surface Glycoproteins 145 

Myotendinous Antigen Ml 145 

Lysosome-Plasma Membrane Dynamics 147 

Correlation of (Na,K)-ATPase and Intramembranous Particles 148 

Synthesis, Assembly, and Processing of Acetylcholinesterase in Muscle 151 

Synthesis and Activation of AChE 152 

Activation and Transport of AChE Requires Glycosylation 153 

Only Catalytically Active AChE Molecules Are Processed in the Golgi 

Apparatus 154 

Transmembrane Movement of Oligosaccharide-Lipids during Asparagine-Linked 

Oligosaccharide Assembly 155 

Intracellular Translocation and Metabolism of Lipids 158 

Internalization of Fluorescent Phosphatidylcholine and 

Phosphatidylethanolamine 159 

Studies on Fluorescent Sphingolipids 162 

Isolation and Characterization of Mutants Using a Fluorescence Photosensitization 

Method 163 

Intracellular Processing of Lipids: A Theory Based on Microinjection of Fluores- 
cent Lipids into Single Cells 165 

The Molecular Characterization of Geminiviruses 167 

Molecular Studies of Squash Leaf Curl Virus 167 

Biological Studies of Beet Curly Top Virus 170 

Molecular Analysis of Mutations at the Shrunken and Waxy Loci of Maize Caused 

by the Controlling Elements Ac and Ds 170 

Structure of the sh-m,59S3 Allele of the Shrunken Locus 172 

The pcWx cDNA Plasmids Are Homologous to the Genetically Identified Waxy 

Locus 173 

Cloning of the Wx Locus and the Ac and Ds Elements 174 

Direction of Transcription of the Wx Locus and Approximate Extent of the Tran- 
scription Unit 178 

Preliminary Analysis of the RNAs Transcribed in Endosperms Carrying the Ac 

wx-m9 and wx-m9 Alleles 178 

Transposable Genetic Elements and the white Locus in Drosophila 180 

DNA Sequence Studies of the white Locus 180 

Transcription of the white Locus 180 

Germ Line Transformation with white Locus DNA 182 

Chromosomal Rearrangements Induced by the FB Transposable Element 183 

In Vitro Mutagenesis and In Vivo Assay of P-Element Functions 185 

Hybrid Dysgenesis and the singed Locus 186 

P-Element-Mediated Gene Transfer: A Practical System for Studying Drosophila 

Gene Regulation 187 

Vectors for P-Element-Mediated Gene Transfer 188 

The Effect of Chromosome Position on Xanthine Dehydrogenase Expression . . . 190 

The Amplification and Expression of Drosophila Chorion Genes 194 

Characterization of Amplification Origins 195 

Analysis of Transcriptional Regulation of the s38 Gene by Transformation 198 

Biochemical and Genetic Studies of Chorion Gene Expression 202 

Sequence Organization of Chorion Gene Regions . . 204 

Genetic Control of Cell Morphology 204 

Molecular Analysis of the Major Sperm Protein Genes 205 

There Are Multiple MSP Gene Products 207 

Asymmetric Segregation of Proteins during Spermatogenesis 209 

Control of 5S RNA Gene Transcription in Xenopus 211 

Domains of the 40-kDa Transcription Factor 212 

In Vitro Labeling of the 40-kDa Transcription Factor 214 

Use of Immobilized 5S-Specific Transcription Factor for the Isolation of Compo- 
nents of Transcription Complexes 215 

Isolation of the Gene Encoding the 5S-Specific Transcription Factor 215 

Analysis of the 5S rRNA Transcription Complex 216 

The State of 5S RNA Genes in Chromatin 218 

An Assay for Discrimination between Somatic and Oocyte 5S RNA Genes: Evi- 
dence for Post-Transcriptional Control 219 

The Collection of Human Embryos 220 

Developmental Stages in Human Embryos 220 

Development of the Nervous System 220 

The Developing Skeletal System 221 

Placentology 221 

Staff Activities 221 

Staff Activities '. 221 

Seminars 222 

Bibliography 223 

Personnel 225 


New and powerful methods of re- For the past several years, Douglas 

search have made possible remarkable Fambrough's research has encompassed 

advances in our understanding of genes an increasingly diverse set of proteins all 

and their protein products. Because of of which in some way influence how cells 

new methods we are entering a new era sense their chemical environments. One 

of lipid research, which might be called of these surface proteins, the (Na,K)- 

lipid cell biology. Whereas in the past ATPase, controls the flow of ions into 

lipids were the province of biochemists, cells. A monoclonal antibody has been 

who studied the chemistry of lipids and isolated that reacts specifically with this 

the enzymatic reactions that synthesized protein; Fambrough and his colleagues 

and degraded them, now it is possible to are thus able to correlate cytologically 

investigate the location and function of the protein's location with that of a class 

lipids in living cells and tissues. A path- of intramembranous particles. The cen- 

finder in the discovery and use of such tral importance of (Na,K)-ATPase to 

techniques is Richard Pagano; he and his conductivity of nerves makes the eluci- 

colleagues at the Department have dation of its detailed chemistry of great 

learned how to tag lipids with fluorescent interest. 

labels that do not interfere with normal In last year's Report, Spradling and 

metabolism. In this Report, they dem- Rubin described their new and powerful 

onstrate their new methodology in a use of transposable genetic elements as 

beautiful set of studies designed to trace a vector for introducing genes into Dro- 

the unique pathways of different lipids sophila. In the first application of this 

through the cell cytoplasm after intro- method, they demonstrate that a spe- 

duction at the cell surface. In recognition cialized gene, termed rosy, can be trans- 

of this and related work, Pagano has been ferred into many different chromosomal 

honored as a recipient of a 1983 Joseph loci without affecting its developmental 

Meyerhoff fellowship for study at the regulation. This is an important prelim- 

Weizmann Institute in Rehovat, Israel, inary result if the gene transfer method 

The departmental interest in cell sur- is to be used to define functional regions 

faces has been strengthened by the ar- in genes. In another set of experiments, 

rival of Martin Snider, who studies the the DNA region that accounts for am- 

cellular location of intermediates in gly- plification of chorion genes is being de- 

coprotein synthesis. Complex sugars are limited systematically by transforming 

added to a polypeptide in precise order defined DNA segments into Drosophila. 
within cells, and specific intermediates Several years ago, Nina Fedoroff 

move from the cytoplasmic side of a changed her field of research altogether, 

membrane to its luminal surface, where She decided to study the molecular basis 

biosynthesis is completed. of transposable genetic elements in maize, 

In parallel studies, Richard Rotundo genetic elements discovered decades 
describes his recent observations on the earlier by Barbara McClintock. Fedo- 
metabolismofthe important enzyme ace- roffs very difficult - id, some would ar- 
ty lcholinesterase. This enzyme has a sur- gue, risky change in direction is 
prisingly complex metabolism, generat- succeeding. She and her colleagues have 
ing at least two different populations of cloned both the Activator (Ac) and Dis- 
molecules metabolized by the cell in very sociation (Ds) elements by making use 
different ways. of their integration into and excision from 



two maize loci — Shrunken and Waxy, turned our attention to the biochemistry 

Studying transposable elements in maize of chromatin. 

has a special interest for developmental Kenneth Muller came to this Depart- 
biologists, since alleles can have very dif- ment in 1975. Combining microscopy and 
ferent effects on gene expression. We can physiology in his analyses of the leech 
now expect explanations for these phe- nervous system, he first demonstrated 
nomena at the molecular level. In addi- that intracellular dye injection methods 
tion, transposable elements can be used would faithfully outline connections be- 
as markers for isolating genes into which tween neurons in this simple system. He 
they have been inserted. then carried out a series of experiments 

Like Nina Fedoroff, Sondra Lazarow- on the fidelity of nerve regeneration and 

itz is making the very best use of the synapse connection to target wherein he 

independence and freedom that our De- analyzed the roles of the cell body, the 

partment offers its scientists. Trained in target itself, and the glial-cell sheath in 

animal virology, she decided to change the specificity of regeneration and nerve 

her field of research to plant viruses. As sprouting. These highly original experi- 

pathogens, plant viruses are of enormous ments were products of extraordinary 

societal interest, but they receive com- skill. In September 1983, Muller will be- 

paratively little attention from scien- come a professor in the Department of 

tists. Lazarowitz has elected to study a Physiology and Biophysics at the Uni- 

class of plant viruses called geminivi- versity of Miami Medical School, known 

ruses — the only known single-stranded for its strength in neurobiology. He will 

DNA viruses that infect eukaryotes. In be missed not only for his many scientific 

the few months that she has been here, skills but also for his unselfish citizenship 

she has already isolated the DNA from and special warmth and humor, 

one of these viruses and is in the process Gerald Rubin also leaves the Depart- 

of characterizing it. ment, to assume the MacArthur profes- 

Burke and Ward have analyzed genes sorship in the Department of Biochemistry 

for the major sperm-specific proteins in at the University of California, Berke- 

the nematode Caenorhabditis elegans. ley. In his slightly less than three years 

These proteins are encoded by a complex here, Rubin has become a world leader 

family of related but not identical genes, in modern molecular genetics. 

The proteins are major constituents of Other honors awarded to Staff Mem- 

the pseudopod that moves the sperm to- bers this year included the Young Sci- 

ward the egg, and they are likely can- entist Award of the Passano Foundation 

didates to account for this ameboid to Rubin and Allan Spradling for their 

movement, which is required for fertil- research on Drosophila. Spradling was 

ity. named Maryland's Distinguished Young 

Research on the 5S RNA genes evolves Scientist of 1982 by the Maryland Acad- 

toward understanding the molecular de- emy of Sciences. Donald Brown received 

tails of what we call the transcription an honorary D.Sci. from the University 

complex — the structure that is formed of Maryland, Baltimore County campus, 

between a gene and a minimum of three in June 1983. 

factors (presumably proteins) required The research of Brown, Fambrough, 

to direct accurate transcription initiation Muller, Pagano, Rubin, Snider, Spra- 

by RNA polymerase III. One of these dling, and Ward has benefited from in- 

factors has been identified, and its inter- dividual grants from the National 

action with the gene is being elucidated Institutes of Health. Fedoroff, Lazarow- 

in detail. We believe that the transcrip- itz, and Rotundo receive support from 

tion complex that can be formed in vitro the National Science Foundation, and 

closely resembles the state of an active Fedoroff also has a grant from the U.S. 

5S RNA gene in a living cell. This has Department of Agriculture. Fambrough 


and Rotundo have grants from the Mus- dation Fellow. Departmental research has 

cular Dystrophy Association. Pagano's been assisted greatly by grants from the 

and Rubin's research receive additional Commonwealth Fund and the Fleisch- 

support from the Whitehall Foundation mann Foundation as well as by a Biomed- 

and the American Cancer Society, re- ical Research Support Grant from the 

spectively. Muller is a recipient of a National Institutes of Health. 
McKnight Neuroscience Development 

Award, and Rotundo is a Sloan Foun- Donald D. Brown 


K. J. Muller, E. J. Elliott, K. A. French, A. Mason, V. J. Morgese, and B. E. Thomas 

The nervous system's complex wiring erate synapses with normal targets in 
places demands not only on neural de- adjacent ganglia, but, until looking with 
velopment but also on the regeneration higher resolution at the regenerated syn- 
of connections, or synapses, between apses, we did not know that mechano- 
nerve cells. After nerve injury, success- sensory neurons reconnect incompletely, 
ful regeneration requires several steps, The distal portion of an axon that has 
including outgrowth of the severed nerve been severed from its cell body often sur- 
fibers, or axons, and specific recognition vives for weeks or months. In the leech, 
and synapse formation between these ax- we have shown that a distal axon seg- 
ons and appropriate targets. For reasons ment can sometimes aid regeneration by 
that remain obscure, regeneration is acting as a functional splice until regen- 
generally not successful in the nervous eration is complete. It is clear that the 
systems of man and other mammals, but regenerating axon recognizes its distal 
it is accurate between individual nerve segment, but does the distal segment ac- 
cells, or neurons, in the medicinal leech, tually guide accurate regeneration? We 
In the past year, we have continued to selectively destroyed the distal segment 
study axon growth and synapse regen- of anonsensory cell, the S-cell, which has 
eration in the leech nervous system. We a distal segment large enough to permit 
also have probed more deeply into some micropipette injection of a lethal dose of 
of the leech nervous system's cellular re- protease. Ensuing axon and synapse re- 
lationships and their structural under- generation proceeded as accurately with- 
pinnings. out the segment as with it. 

The basic building blocks of the leech Isolated axon segments have the sur- 

nervous system — the neurons and non- prising ability to sprout at their ends and 

neuronal glial and microglial cells — are grow for days, either in the animal or in 

similar in appearance and function to their segments of nerve cord in tissue-culture 

counterparts in the brains of higher an- medium. To test whether the glial cells 

imals. However, the small number of cells that ensheathe the segments are an es- 

in the leech nervous system, and their sential source of materials for survival 

stereotyped positions, shapes, connec- and growth, as has been widely hypoth- 

tions, and functional roles make the leech esized, glial cells were selectively de- 

particularly favorable for experimental stroyed. Preliminary results indicate that 

analysis. even without their glial sheath, axon seg- 

Among the 400 neurons that constitute ments will sprout, 

each segmental cell cluster, or ganglion, Among the most puzzling elements of 

are the fourteen mechanosensory neu- the nervous system's "cytoskeleton" are 

rons. We knew these cells could regen- intermediate filaments, believed to sta- 


bilize cell structure. We have found that cells) can reconnect with one class of tar- 
bundles of intermediate filaments in the gets in adjacent ganglia, S-cells, after the 
leech are associated with distinctive sites T-cells' axons were severed between 
in cell membranes. Other investigators ganglia. The synapse between T- and S- 
may have mistaken these sites for gap cells is indirect, however, involving cou- 
junctions. But gap junctions, or electrical pling interneurons called C-cells. To con- 
synapses, differ markedly from these sites firm physiologically that T-cells re- 
in structure and distribution. In studying establish contact with T-cells, we re- 
maturation of the nervous system and its corded voltage changes in the target post- 
ability to regenerate, it will be useful to synaptic T-cell using signal-averaging 
learn whether intermediate filaments and techniques. We thus were able to dem- 
their sites of membrane attachment in- onstrate more rigorously than before, the 
fluence neuronal plasticity. specificity and effectiveness of regener- 

The large glial cells of the leech en- ated connections, 
sheathe the neurons, including their ax- The connection between T-cells in ad- 

ons. Smaller cells, called microglia, which jacent ganglia normally produces a signal 

are scattered within the nerve cord, in 400-600 |jlV in amplitude, about one-half 

some respects resemble glia because they the strength of the connection between 

eventually ensheathe portions of axons T-cells in the same ganglion. T-cells make 

after nerve injury. It is a well-estab- synapses and contact one another at var- 

lished fact, shown first in the leech, that icosities or swellings along axonal 

the large glial cells are coupled to one branches in the neuropile (the central re- 

another by electrical junctions. Dyes of gion of the ganglion). Varicosities are 

low molecular weight pass at such junc- distinctive structures, and, in the light 

tions selectively from one glial cell into microscope, one can normally see about 

neighboring glia. By injecting large glia 50 contacts between the anterior branch 

with dyes and examining sections of nerve of one T-cell and its ipsilateral homolog 

cord at high resolution, we have learned in the anterior adjacent ganglion. Since 

that no similar communication exists be- the recorded synaptic potential repre- 

tween large glia and microglia. sents the sum of synaptic potentials gen- 

From behavioral observations, it is clear erated at all contacts, the average contact 
that taste and smell are among the leech's probably generates a synaptic potential 
better-developed senses. The relative of approximately 10 (jlV. In our studies 
simplicity of the nervous system should of newly regenerated synapses, we find 
make it a favorable preparation for the total number of contacts and aver- 
studying integration of chemical and other aged synaptic potentials to be smaller than 
sensory information. A first step has now normal. In addition, the calculated "syn- 
been made, in the identification of pe- aptic potential per contact" is in the range 
ripheral structures likely to be receptors of 5-7 (jlV, rather than the 10 (jlV nor- 
for chemical stimuli and in the recording mally recorded (Fig. 1). This discrepancy 
of nerve impulses that propagate into the might represent a diminished conduct- 
central nervous system in the presence ance at new contacts, a diminished cur- 
of blood and other substances known to rent-generating capability of regenerating 
stimulate leech responses. fibers, or a greater electrical distance be- 
tween the site of intracellular recording 
Regeneration of Synapses and the contacts - Some supporting evi- 
between Touch Sensory Neurons dence exists for each of these explana- 

K. J. Midler, in collaboration with 
E. R. Macagno and S. A. DeRiemer 


The regenerated preparations fall into 
two categories: partial regenerates and 
Last year (Year Book 81, 129) we re- (a small minority) full regenerates. Ex- 
ported that touch sensory neurons (T- periments reported last year {Year Book 

1 YR 




T 7- '■••_- 



• tur 


'v art. A 

'8 -r ' -. / 

10 nsec 

Fig. 1. Synaptic transmission between touch cells (T-cells) was reduced in amplitude up to one year 
after regeneration. Axon of T 9 (touch cell in ganglion 9), severed a year before the experiment, had 
regenerated 22 contacts with T 8 in the next anterior ganglion. The synaptic potentials in T s that were 
evoked by impulses in T 9 were approximately one-half their normal strength (control). Normally there 
would be about 50 contacts between the cells. Traces are voltage recordings made with microelectrodes 
in the cell pairs and represent the average of 16 sweeps. Preparation was bathed in saline containing 15 
mM Mg + + to block chemical synaptic transmission. 

81, 126-128) demonstrated that T-cell 
axons can, rarely, regenerate by fusion 
with severed distal segments. These re- 
sults suggest that the rare, full regen- 
erates are neurons whose axons have 
fused with severed distal segments. Thus, 
most of those axons that regenerate do 
so by growing the entire distance to their 
targets, usually in the company of a tar- 
get T-cell's axon, and the connection they 
form is less effective than the connection 
before injury. There is no evidence that 
compensatory connections are formed 
with the target cell by other neurons. 

Accurate Regeneration of the 

Synapse between S-cells after 

Selective Destruction of the 

S-cell's Distal Segment 

A. Mason, K. J. Mullet, and B.'E. Thomas 

In each segmental ganglion of the leech 
there is a single cell (S-cell) which, in 
addition to making synapses within the 
ganglionic neuropile, extends an axon ap- 
proximately halfway anteriorly and pos- 
teriorly toward adjacent ganglia to make 
electrical synapses with its neighboring 
homologs. S-cells are particularly favor- 
able for study because of their electrical 

synapses, their axons' relatively large 
size, their stereotyped axon pathways, 
and their ability to reconnect reliably. 
One important question has been whether 
the distal segment of the axon, cut from 
the cell body but viable for months af- 
terward, must be present to guide the 
regenerating axon. (The regenerating 
axon does grow along its distal segment 
and sometimes forms an electrical syn- 
apse specifically with it, indicating that 
the axon can recognize the segment.) 

In our initial attempts to destroy the 
distal segment, which is typically only 7 
|xm in diameter, we filled the segment 
by heavily injecting the target S-cell in 
the adjacent ganglion with Lucifer Yel- 
low dye. The dye travelled down the cell's 
axon and across the electrical synapse 
into the distal segment of the experi- 
mental cell. We then irradiated the seg- 
ment with intense light from a laser. It 
proved difficult, however, to destroy se- 
lectively the segment without consider- 
ably damaging surrounding tissue, so we 
chose a more direct approach. Axons were 
severed and a microelectrode containing 
a protease solution was inserted into the 
S-cell distal segment, which was identi- 
fied by its location and its electrical char- 



acteristics. Protease was then injected 
into the segment, the amount titrated 
visually by co-injection of a dye. Protease 
destroys axons as rapidly and effectively 
as it does whole cells after injection into 
the cell body, so that the distal segment 
was in most cases no longer detectable 
after a few days. Distal segments were 
injected and destroyed in a total of 32 
preparations; these were then examined 
from two days to fourteen weeks later. 
For technical reasons, the animals we used 
were small. Within two or three weeks 
(the usual period for reconnection be- 
tween S-cells in such preparations), the 
cells were again electrically and dye cou- 
pled (Fig. 2). Ordinarily, when we sever 
S-cell axons by crushing the connective, 
80% of S-cells have reconnected one month 
later. In the absence of the distal seg- 
ment, we obtained similar results. Thus, 


.■■■■■■■ ■.■... 

Fig. 2. S-cell axon reconnected with its target 
within seven weeks after the distal segment was 
destroyed with protease. The regenerated axon, 
on the left, was filled with horseradish peroxidase 
(HRP) (Mr -40,000) and Lucifer Yellow dye (Mr 
—500). After the dye crossed specifically into the 
target axon at the synapse, the preparation was 
stained for HRP, which obscured the Lucifer Yel- 
low fluorescence. Upper panel: Fluorescence mi- 
crograph showing regenerated axon (left) and 
fluorescent target (right). Background fluorescence 
of the connectives is visible. Lower panel: Bright- 
field micrograph with only the HRP-stained re- 
generated axon. Calibration bar represents 100 |xm. 

killing the distal segment was no impe- 
diment to regeneration. 

The regenerating axon grows prefer- 
entially along its normal pathway, but it 
remains unclear whether it is guided by 
a single cue, such as a yet-undetected 
component of extracellular space, or by 
multiple cues, including the distal seg- 
ment, all of which must be eliminated to 
interfere with accurate regeneration. 

Survival and Growth of Isolated 
Axon Segments after Destruction 
of Their Ensheathing Glial Cells 

K. A. French 

In previous experiments in our labo- 
ratory, it was found that leech sensory 
cell axon segments, isolated from their 
cell bodies and terminal arborizations, 
could sprout at both ends. We wanted to 
learn the source of materials and energy 
that support sprouting without a cell body. 
One candidate was the large glial cell that 
surrounds axons in each connective. Al- 
though the transfer of materials between 
glial cells and neurons has been seen in 
some other invertebrates, such transfer 
does not normally occur in the leech (E. 
Elliott, personal communication). 

The role of glial cells in axonal sprout- 
ing was tested by observing axon seg- 
ments deprived of their glial wrappings. 
If glial cells were the source of material 
necessary for sprouting, an axon seg- 
ment with no glial wrapping would not 
sprout and might even die. 

As a first step, we destroyed glial cells. 
A single glial cell was impaled in each 
leech with a protease-filled microelec- 
trode. Protease was injected into the cell, 
killing only the impaled cell (Year Book 
81), and the leech was allowed to re- 
cover. After 1-2 weeks, a chain of three 
ganglia, including the connective wherein 
the glial cell had been killed, was re- 
moved from the leech and placed in tis- 
sue-culture medium. The two medial- 
pressure-sensory cells (P-cells) in the 
middle ganglion of the three were im- 

K. J. Midler and B. E. Thomas, in collaboratioyi 


paled and filled with two marker sub- Distinctions between Gap 
stances — fluorescein-labeled dextran and Junctions and Sites of 
horseradish peroxidase (HRP). After the Intermediate Filament 
markers diffused throughout the cells, Attachment in the Nervous 
cuts were made to isolate axon segments System 
in the connectives. With appropriate flu- 
orescence optics, the dextran label showed ™with *D. Pumplin 
the morphology of the P-cells and their 
axons in living preparations. During the 

following week, with the aid of an image The electron microscope is a powerful 
intensifier, P-cell growth was recorded tool for examining the structure of the 
on videotape. The structure of the HRP- nervous system, for it is only with the 
filled axon segments was observed after electron microscope that one can conclu- 
fixation by conventional light-micro- sively identify such important structures 
scopic and electron-microscopic meth- as synapses and fine, unmyelinated ax- 
ods. In each preparation, we compared ons. Unfortunately, electrical synapses, 
sprouting in the segment lacking an in- whose structural correlate is the gap 
tact glial sheath with sprouting in the junction, are difficult to detect in con- 
segments surrounded by glia. ventionally sectioned tissue. If tissue is 

When we observed the desheathed axon instead frozen, usually after fixation, and 
segments immediately after cutting, the rapidly fractured, gap junctions can of- 
segments usually appeared as blunt- ten be detected as distinctive clusters of 
ended, smooth shafts. When we removed particles within expanses of opened 
ganglia from the animal a week after glial membrane. The freeze-fracture tech- 
injection, we observed branching along nique splits membranes into two halves, 
the P-cell axon in three of twenty prep- one half in contact with the inside of the 
arations. When we removed ganglia three cell (the P face), the other in contact with 
weeks after glial injections, we observed the outside (the E face). In vertebrates, 
branching in two of four preparations, gap junction particles are clustered in 
Such side branching by intact axons had patches within the P face, leaving small 
been seen previously in our laboratory pits in the E face. In some invertebrates, 
when the time after injection was about the opposite configuration has been re- 
three weeks, but it was not known that ported. 

branching could begin only a few days We have used freeze-fracture tech- 
after killing the glial cell. To differentiate niques in the leech nerve cord in an at- 
between pre-existing branching and that tempt to characterize gap junctions in 
which occurred after isolating the seg- that organism. Slabs of nerve cord about 
ment, we found "zero-time" records to 50-|xm thick were prepared by cutting 
be essential. agar-embedded tissue with a vibrating 

Preliminary results indicate that iso- blade on a Vibratome; the slabs were fro- 

lated axon segments lacking glial sheath- zen and fractured, and platinum-carbon 

ing were no more likely to degenerate replicas were made. In addition to neu- 

during the week of observation than were rons, the nerve cord contains glia and 

those surrounded by glia. Furthermore, muscle cells. In the samples examined, 

the amount of sprouting at the distal and we did not see definite gap junctions be- 

proximal ends of both glial-ensheathed tween neurons, but at contacts between 

and non-glial-ensheathed axon segments muscle cells and between glia there were 

did not differ noticeably. The frequency arrays of P-face particles having sizes and 

of branching that developed along seg- arrangements characteristic of gap junc- 

ments that were smooth at the start of tional particles (Fig. 3). Glia are electri- 

the culture period was also about the same cally coupled to one another, as are certain 

in sheathed and unsheathed axons. muscle cells in the leech. 




Fig. 3. Complementary views of a gap junction (toward the left) and a hemidesmosome (toward the 
right) on a glial cell of a lateral connective. The membrane faces (E face above; P face below) appear as 
though the membrane were opened like a book along the horizontal axis between the figures. The 
hemidesmosome is seen in the E face as a group of large irregular particles scattered within a sharply 
bounded plaque. The complementary P face bears very shallow pits (arrowhead in B), making the 
membrane appear slightly rough. The gap junction is a plaque of particles in the P face; the particles 
are more regular both in shape and arrangement. The complementary E face bears a distinct group of 
regularly arranged pits. A small number of gap junction particles remain on the opposing face (arrowhead 
in A); this phenomenon also occurs in the hemidesmosome. Both figures, same scale; bar in A indicates 
0.1 jxm. 

Another distinctive particle cluster is 
obvious in fractured membranes. Such 
clusters have been identified as sites of 
attachment of intermediate filaments in 
nerve, glia, and muscle (Figs. 3 and 4). 
Intermediate filaments occur in distinc- 
tive bundles within the leech nervous 
system. Indeed, it was the silver staining 
of these filaments that attracted anato- 
mists to the study of the leech nervous 
system at the turn of the century. Con- 

ventional thin sections have shown that 
intermediate-filament bundles attach to 
the membrane at structures called hemi- 
desmosomes. We now know that hemi- 
desmosomes are composed in part of 
characteristic clusters of E-face parti- 
cles, which can be distinguished from gap 
junctional particles not only by their di- 
mensions and density but also by their 
membrane association. Because the in- 
termediate filaments and their attach- 




4 '"'■■jF >/ ~ -■-— ■<»- ' 




Fig. 4. Within the S-cell axon, in medial connective between ganglia, bundles of intermediate fila- 
ments (arrowheads in A, F in B) approach the membrane and attach opposite extracellular space (E in 
A and B). At the attachment sites, the membrane bulges outward (in A, B, and C), and distinctive 
patches of intramembrane particles appear in the E face of the membrane. B is an enlargement of the 
lower attachment site in A, tilted 26°. A line of electron density (arrowheads in B) lies subjacent and 
parallel to the membrane at an attachment site; cross-fractured small filaments appear in this location 
in freeze-fracture (D). A group of large particles on the P face of the S-cell axon (circled region in E) is 
presumably a gap junction. Some of the constituent particles are poorly resolved due to membrane 
curvature away from the shadowing beam. Cytoplasm of surrounding glial cells is recognizable by its 
relationship to axons and by filaments seen in thin sections (A) and freeze-fracture (G in C). A, C, D, 
and E are same scale; bar in C indicates 0.5 |xm. B is enlarged three times greater; the bar (shown in 
C) represents 0.17 |xm on B. 

ment to membranes are thought to be change during maturation and regener- 

important in stabilizing cell structure, it ation of the nervous system. Electrical 

will be useful to know how the frequency synapses between glia and between mus- 

and distribution of hemidesmosomes cle cells function like those between neu- 



rons, but obviously it is important to 
confirm that gap junctions between neu- 
rons resemble in structure those that we 
have now identified between other cells 
in the nerve cord. 

Leech Chemoreception 

e. J. Elliott 

My studies of chemoreception in Hi- 
rudo medicinalis are aimed, in the short 
run, at determining the basic elements 
of the chemosensory system, including 
the primary sensory receptors and the 
neurons involved in processing chemical 
sensory information. Experiments by 
others have indicated that leeches can 
sense and react to chemicals in the water 
surrounding them and that the capability 
for this chemical sense may be localized 
in the head. I have focused my initial 
anatomical and physiological studies on 
the mouth and lip region of the head. 

In physiological experiments on dis- 
sected head preparations, I attempted to 
record electrical responses by means of 
suction electrodes placed on the first pair 
of ventral roots and the first two dorsal 
roots of the head ganglion while applying 
chemical stimuli to the mouth and lip re- 

I have recorded a response to blood 
from the first dorsal root of the supra- 
esophageal ganglion. The lip region was 
bathed in "artificial spring water" (the 
ionically dilute water in which the leeches 
are maintained). The blood stimulus was 
mixed 1:2 with the water and applied by 
adding it to one of the two reservoirs that 
provided continuous perfusion of the lips. 
The nerve roots and internal side of the 
mouth preparation were kept in Ringer's 
solution and protected from contact with 
the spring water by using a two-com- 
partment chamber and by tying shut the 
mouth opening with fine suture thread. 

Scanning electron microscopy has re- 
vealed that the sensilla of the lip region 
lack the long cilia found in segmental sen- 
silla, and have only structures that re- 
semble the shorter, "grouped" cilia of 

segmental sensilla. These structures also 
resemble olfactory cilia found in verte- 
brates and are good candidates for che- 
moreceptive cilia. Currently in progress 
are experiments attempting to backfill 
these structures from the nerve roots 
leading to the lip sensilla (with HRP or 
cobalt), and to observe them in the trans- 
mission electron microscope. 

Large Glia Are Not Dye-Coupled 
to Microglia 

V. J. Morgese, E. J. Elliott, K. J. Muller, and 
B. E. Thomas 

That gap junctions exist between large 
glial cells can be demonstrated by the 
passage of small, charged dye molecules 
such as Lucifer Yellow, specifically from 
one cell to the next. It was not known, 
however, whether the small cells called 
microglia, which react to injury of the 
nervous system (Year Book 81, 132), are 
dye-coupled to the large glia. Both types 
of cells are evidently capable of en- 
sheathing axons. 

Microglia are too small for microelec- 
trode penetration, which would allow us 
directly to measure electrical coupling to 
the large glia. We were able to test for 
dye-coupling, however, by injecting the 
large glial cells with Lucifer Yellow dye 
and looking for the dye in the microglia. 

To increase locally the density of mi- 
croglia, we used crushed nerve cords in 
our studies. The large glia were heavily 
injected with Lucifer Yellow. The nerve 
was then fixed in 4% paraformaldehyde, 
dehydrated in acetone, stained with 1% 
hafnium tetrachloride in cold acetone for 
one hour, and embedded in Epon. Cross 
sections of nerve cord were cut for light 
microscopy and electron microscopy. Mi- 
croglia identified with the electron mi- 
croscope corresponded perfectly to 
nonfluorescent regions of nerve cord pho- 
tographed earlier although there was 
strong dye-coupling between the large 
glia. Microglia are thus not linked to large 
glia in the way that large glia are linked 
to one another. 



M. Chiquet, J. Lippincott-Schwartz, B. Wolitzky, and D. Fambrough, with the technical assistance of 

D. Somerville 

The cell surface is the major interface appearance of components of the extra- 
between the cell and its environment. In- cellular matrix in the endomysium, that 
teractions at the cell surface are essential space between individual muscle fibers, 
in the uptake of nutrients, in the re- I spent most of the year characterizing 
sponsesofthe cell to various signals from a single component found in chick skel- 
other parts of the body via hormones and etal muscle endomysium, because this 
trophic substances, in the maintenance component exhibits many characteristics 
of osmotic balance, in the correct posi- expected for a developmental and func- 
tioning of the cell in its tissue, and in the tional "link" between tendon cells and 
issuance of signals by the cell to other muscle fibers. This "myotendinous anti- 
cells, to name a few examples. gen" is recognized by monoclonal anti- 

During the past year, our studies of body Ml, whose isolation was described 

cell surfaces have focused on molecules earlier ( Year Book 81, 137-138). The first 

involved in several of the interactions step was to do a more extensive analysis 

mentioned above. Three of these are of the distribution in adult tissue of the 

summarized in this report. The first sec- antigen recognized by Ml antibody, as 

tion describes a component of the extra- well as to study its developmental ap- 

cellular matrix which appears likely to pearance in chick embryos. Ml antigen 

be involved in the attachment of muscle occurs in the endomysium (but not the 

fibers to their tendons and may mediate perimysium and epimysium) of chick ALD 

cell attachment in other cellular contexts muscle (Year Book 81), and is unique 

as well. The second section deals with a among such antigens in that it occurs only 

membrane glycoprotein that shuttles be- in proximal and distal portions of the 

tween the cell surface and the lysosomes muscle fibers near their insertion points, 

(organelles involved in the breakdown of not in medial regions. Myotendinous 

macromolecules). The third section fo- junctions and tendon fascicles are rich in 

cuses on the molecule that plays the cen- Ml antigen (Fig. 5), whereas the loose 

tral role in regulating the sodium and connective tissue surrounding the mus- 

potassium ion concentrations in the cell cles and tendons appears devoid of the 

and in maintaining osmotic balance. antigen, with the exception of blood ves- 
sel walls and perineurium. 
The apparent nonhomogeneous distri- 

Myotendinous Antigen Ml bution of the myotendinous antigen on 

.. ~, . the surface of single muscle fibers raises 

M. Chiquet , . ,, ° . . .,, . . , 

questions as to the origin ol this material 

The extracellular matrix found on the during development. If it were produced 
surface of an adult muscle fiber is prob- by the muscle fibers themselves, the fi- 
ably composed of components made by bers must have been able to restrict dep- 
both muscle and nonmuscle cells, and osition to specialized locations, i.e., to 
might, in part, reflect the developmental their tips. Alternatively, the antigen could 
history of that muscle fiber. A particu- have been deposited by tendon fibro- 
larly interesting possibility is that extra- blasts partially invading the endomy- 
cellular matrix components act as sium. This mechanism is suggested by 
morphogenetic signals between nonmus- the staining patterns found with Ml an- 
cle and muscle cells during muscle mor- tibody during the development of the limb, 
phogenesis. We have focused our studies In four-day-old embryos, myotendinous 
on the cellular origin and developmental antigen appears first in the perichon- 







fPP» > 

"' , s 

:fe * Mt-^^f^^ 



V ^, 

Fig. 5. Longitudinal cryosection through an 18-day-old chick embryo leg muscle stained with Ml 
antibody (10 fxg/ml) and fluorescein-labeled second antibody, t, tendon; m, muscle fibers; mt, myoten- 
dinous junction. 

drium of the long bones of the limb. On 
day five, streaks of material reacting with 
Ml antibody extend from the perichon- 
drium towards the developing premuscle 
masses. When the muscle anlagen split 
into individual muscles (day six), myo- 
tendinous antigen is detected in the ten- 
don primordia located between individ- 
ual muscles and attached to the muscles 
on both their proximal and distal ends. 
The only other structures labeled in chick 
embryos (besides perichondrium, liga- 
ments, and tendons) apparently are 
smooth-muscle tissues of the aorta, the 
lung buds, and the gizzard. We therefore 
hypothesize that myotendinous antigen 
is first accumulated by cells of the peri- 
chondrium and the developing tendons, 
and might later be involved in the at- 
tachment of the arising muscle fibers to 
tendon fascicles. 

Biochemical studies done with meta- 
bolically labeled myotendinous antigen 
suggest that this antigen interacts with 

proteoglycans produced by myotubes. 
Myotendinous antigen released by chick 
muscle or fibroblast cultures into the me- 
dium was shown to be a large glycopro- 
tein complex consisting of at least five 
(or more) disulfide-linked heterogeneous 
subunitsofMr = 150,000-220,000 on SDS 
gels. The differently sized subunits yielded 
very similar proteolytic cleavage pat- 
terns and are therefore related. The an- 
tigenic site recognized by Ml antibody 
resides on the different subunits, as was 
demonstrated by immunoblotting tech- 
niques. The antigen complex is resistant 
to bacterial collagenase (unlike known 
collagen types) and is barely affected by 
chymotrypsin; a pepsin-resistant domain 
was described earlier (Year Book 81). 
Despite having subunits of similar size, 
myotendinous antigen differs from fibro- 
nectin antigenically and structurally; un- 
like fibronectin, it is not present in chick 
serum and does not bind to gelatin. Fi- 
broblasts released newly synthesized 


myotendinous antigen at a rate seven to study the movement, distribution, and 

times that of muscle cells. Interestingly, biochemistry of molecules involved in the 

very-high-molecular-weight material was membrane traffic between cell surface and 

always co-precipitated with myotendin- lysosome. We report here on one of the 

ous antigen by Ml antibody from muscle antibodies, monoclonal antibody CV24, 

culture medium. This material could be which recognizes a membrane antigen that 

labeled heavily by [ 3 H]-glucosamine and we believe participates in the dynamic 

[ 35 S]-sulfate and, to a much lower extent, flow of membrane between cell surface 

by radioactive amino acids; it was di- and lysosome. 

gested by hyaluronidase (which also di- The distribution of CV24 antigen on 
gests chondroitin sulfate). We know that cultured cells is predominantly on intra- 
myotubes (but not fibroblasts) synthe- cellular vesicles, a large portion of which 
size a proteoglycan with very large chon- were determined to be lysosomal by sev- 
droitin sulfate side chains (Carrino and eral criteria. The immunofluorescent dis- 
Caplan, J. Biol. Chem. 257, 14145, 1982). tribution of antigen on cultured cells 
The interesting possibility that myoten- co-localized markedly with acid-phospha- 
dinous antigen specifically interacts with tase staining on the same cells; fixed, 
this muscle-specific proteoglycan awaits permeabilized cells incubated with HRP- 
further investigation. If it does, it is in- CV24 showed immunoperoxidase label- 
deed likely that myotendinous antigen ing of lysosomes at the EM level; and 
represents a functional link between (de- immunoprecipitation of the antigen was 
veloping and adult) muscle and tendon possible from lysosome-enriched subcel- 
cells. lular fractions. In addition to the intra- 
cellular pool, where the antigen appears 
to be distributed between lysosomes and 
Lysosome-Plasma Membrane other small vesicles, there is a cell sur- 
Dynamics face pool of antigen. The relative sizes of 
T T . . „ , ± these two pools of antigen have been es- 

timated using radiolabeled C V24 binding 
The membrane system that directs to fixed (for surface binding) or fixed, 
proteins between the cell surface and ly- permeabilized (for total binding) cul- 
sosomes includes several specialized tured fibroblast cells. Results of binding 
membranous structures (for example, isotherms generated in this manner and 
coated pit, coated vesicle, prelysosomal corresponding Scatchard plots indicate 
compartments). Proteins brought into this that there are about 8,000 surface sites, 
pathway by receptor-mediated endocy- compared to 370,000 internal antigen sites, 
tosis, by cell surface turnover, or by per cell. Thus, only 2% of the CV24 an- 
membrane flow may either recycle to the tigen resides on the cell surface of these 
cell surface or enter the lysosomes. How cells, and 98% exists intracellularly. 
cells target extracellular molecules for The CV24 antigen has been isolated 
degradation in lysosomes and how they from detergent extracts of [ 35 S]-methi- 
regulate their endogenous surface pop- onine-labeled fibroblast and muscle em- 
ulations of molecules during growth and tures by affinity chromatography on 
differentiation remain major mysteries, antibody coupled to Sepharose beads. 
Using hybridoma technology, we have SDS-polyacrylamide gel electrophoresis 
begun to study at the molecular level the analysis of antigen isolated this way shows 
properties and dynamics of the lysosome- one broad band, with apparent molecular 
plasma membrane system in muscle and weight 100,000, corresponding to 0.01% 
associated fibroblast cells. Using purified of all labeled membrane protein. The 
coated vesicles from chick liver as im- CV24 antigen can also be metabolically 
munogen, we have generated a library labeled with [ 3 H]-mannose, and both 
of monoclonal antibodies to use as probes wheat germ agglutinin and Con A bind 


to the antigen. Thus, the CV24 antigen tween the cell surface and a post-syn- 
can be categorized as a membrane gly- thesized intracellular pool, we repeated 
coprotein. Efforts at identifying the the 125 I-CV24 uptake and degradation 
function of the antigen are in progress; experiments in the presence of cyclo- 
preliminary results indicate that purified heximide, a protein-synthesis inhibitor. If 
antigen is capable of cleaving ATP. The surface antigen arrived only from a newly 
possibility that CV24 antigen is the ly- synthesized pool, then in the presence of 
sosomal proton pump is particularly ap- cycloheximide one would expect the cell 
pealing. surface radioactivity to drop as 125 I-C V24 
The distribution of the CV24 antigen is internalized and degraded, since once 
between lysosomes, other vesicular antigen is internalized with antibody, no 
structures, and the cell surface sug- new antigen can replace it. Results of 
gested that it might be involved in the such experiments indicated no qualita- 
membrane flow and sorting of newly syn- tive difference in 125 I-CV24 binding and 
thesized molecules between these com- degradation in the presence of cyclo- 
partments. To begin testing this heximide. We therefore conclude that the 
hypothesis, we used 125 I-labeled CV24 CV24 antigen shuttles to and from the 
antibody to tag surface antigen and fol- cell surface from a post-synthesized in- 
low its movement into the cell. When cell tracellular pool. Since much of the intra- 
surface antigen was labeled with 125 I- cellular pool of CV24 antigen is found on 
CV24 antibody at 4°C (to bind only sur- lysosomes, further characterization of the 
face sites) and then warmed to 37°C to route of transport of this antigen be- 
ensure membrane circulation, the sur- tween cell surface and lysosomes should 
face-bound 125 I-antibody was rapidly in- provide insight into the mechanisms for 
ternalized, since antibody could not be sorting the molecules that move between 
stripped away from low-pH-shifted cells, these compartments. 
After being internalized, 125 I-CV24 an- 
tibody was transported to lysosomes, 

where it was degraded, yielding 125 I-ty- Correlation of (Na,K)-ATPase and 

rosine with a half-time of three hours. To Intramembranous Particles 
demonstrate that such uptake and deg- 

radation of 125 I-C V24 was specific for this ' m r 
monoclonal and not due to antibody-in- When cells are freeze-fractured, the 
duced internalization or fluid-phase up- fracture plane preferentially passes be- 
take of antibody from the medium, we tween the leaflets of the lipid bilayer. 
carried out the same experiments with The bilayer structure is interrupted where 
iodinated monoclonal antibodies directed membrane proteins span it, and these in- 
against other cell surface or nonspecific terruptions are thought to cause pertur- 
markers. Little or no internalization and bations in the fracture plane. Replicas of 
degradation of antibody was observed freeze-fractured cell surfaces, made by 
over the same period for these antibod- depositing a very thin film of metal, often 
ies. Assuming that the CV24 surface an- contain large numbers of these irregu- 
tigen is internalized with the 125 I-CV24 larities, called "particles" and "pits. "While 
antibody during these experiments, then such particles and pits may well signify 
throughout the period of uptake and deg- positions of integral membrane proteins 
radation of 125 I-CV24 antibody, antigen in the bilayer, only a few integral mem- 
continually reappears at the cell surface, brane proteins have been related directly 
since surface binding of 125 I-CV24 re- to such structures. In several cases, 
mained constant. To determine whether membrane proteins reconstituted into 
the CV24 antigen that reappears at the artificial bilayers have been shown to 
cell surface comes from a pool or whether generate intramembrane particles dur- 
it represents shuttling of the antigen be- ing freeze-fracture. In a few other cases, 



tight clustering of some class of intra- 
membrane particle (IMP) is correlated 
with clustering of a functional element in 
the membrane. For example, acetylcho- 
line receptor clusters on myotubes are 
correlated with clusters of very large 
IMP's. Most of the IMP's seen in replicas 
of cells are not clustered and are quite 
variable in size. The density of such IMP's 
is typically on the order of from 500 to 
2000/|jLm 2 . Whether or not these gener- 
ally smaller but quite numerous IMP's 
are correlated with membrane proteins 
was not known. During the past year, in 
collaboration with David W. Pumplin 
(Department of Anatomy, University of 
Maryland School of Medicine), we have 
identified the sodium- and potassium-ion- 
stimulated ATPase (sodium pump) as a 
major correlate of these IMP's. 

To correlate IMP's and the sodium 
pump of chick myotubes, we employed a 
simple strategy. First, we used a mono- 
clonal antibody that binds the sodium 
pump to determine the location of sodium 
pump molecules on the myogenic cells. 
We coupled fluorescein to the antibody 
so that the distribution of antigen (the 
sodium pump) could be viewed as the dis- 
tribution of fluorescence on the cells after 
antibody exposure. The distribution 
proved to be nearly uniform (Fig. 6A). 
However, the distribution could be al- 
tered by exposing the antibody-labeled 
cells to a second antibody, which reacted 
with and cross-linked the first. This sec- 
ond antibody treatment changed the pat- 
tern of fluorescence to one of tiny patches 
(Fig. 6B). (This procedure, incidentally, 
showed that the sodium pump sites are 
not fixed in their positions on the cell 
surface but are at least somewhat free 
to move about in the plane of the mem- 

Next, we compared the distribution of 
IMP's in freeze-fracture replicas of my- 
ogenic cells labeled with fluorescent anti- 
sodium pump antibody with and without 
subsequent second-antibody treatment. 
The directly labeled cells yielded replicas 
indistinguishable from those of totally 
untreated cells. The number and the size 

Fig. 6. Fluorescence micrographs revealing 
distribution of (Na + + K + )- ATPase on chick my- 
otubes in tissue culture. (A) Portion of branched 
myotube labeled with fluorescein-conjugated mono- 
clonal antibody to the (Na + + K + )- ATPase. (B) 
Myotube similar to that in A except that after the 
fluorescent labeling, it was treated with an anti- 
serum to cross-link and cluster the monoclonal an- 
tibody bound to the (Na + + K + )- ATPase. 

distribution of IMP's were the same. The 
cells that had their sodium pumps clus- 
tered by double-antibody treatment, 
however, showed a pattern of clustered 
IMP's (Fig. 7). Quantification of the clus- 
tering phenomenon revealed (1) that the 
total number of IMP's was not altered 
from that of control cells, (2) that the size 
distribution of the clustered particles was 
virtually identical to that of total IMP's, 
(3) that about 80% of the clustered IMP's 
were found in the P face of the replica, 
and (4) that up to 50% of total IMP's were 
clustered by the treatments. 

Several control experiments were done 
to show that the clustering of sodium 
pump sites did not result in a co-cluster- 


Fig. 7. Freeze-fracture replicas of membrane of (A) control and (B) double-antibody-treated chick 
myotubes. Cross-linking of the (Na + + K + )-ATPase molecules by this procedure correlates with for- 
mation of intramembrane particle clusters (outlined in B). 

ing of other membrane proteins, and that myotubes were labeled with fluorescent 

clustering of each of several other mem- antibody and second antibody to cause 

brane proteins by different monoclonal clustering of the sodium pumps into 

antibodies and the same second antibody patches made visible by fluorescence. The 

did not cluster sodium pump sites (and, distribution of these patches was docu- 

incidentally, did not yield freeze-fracture mented by photography. Freeze-frac- 

replicas with IMP clusters). ture replicas of these same cells were 

To further substantiate the correlation then made and the distribution of IMP's 

between IMP's and sodium pumps, we was determined by examining the repli- 

performed single-cell studies. Individual cas in the electron microscope. A one-to- 


one correlation of fluorescent patches and dance from one muscle fiber to another 
clusters of IMP's was found. Thus, we in adult skeletal muscle, suggesting that 
conclude that the intramembrane parti- abundance is possibly a reflection of the 
cles seen in freeze-fracture replicas of history of muscle activity. We would pre- 
chick myotubes correlate largely with the diet that the muscles of patients with 
presence of sodium pumps. muscular dystrophy experience a differ- 
There is possible medical relevance to ent level and kind of usage than those of 
these observations. It has been reported healthy children. Thus, there might well 
by Schotland and colleagues that the be a measurable difference in amount of 
freeze-fracture replicas of muscle fibers sodium pump when these two groups are 
from children with muscular dystrophy compared. Such a difference would also 
have a paucity of IMP's. We have noted be seen in comparison of freeze-fracture 
that the sodium pump is subject to up- replicas of the muscles, since sodium 
and-down regulation in cultured skeletal pumps may account for up to 50% of total 
muscle and is markedly variable in abun- IMP's. 


R. L. Rotundo and A. Mays 

In the nervous system, information is molecules synthesized are neither cata- 
transmitted between neurons and their lytically active nor processed and ex- 
target cells at highly specialized regions ported by the cells, but rather appear to 
of cell-cell contact called synapses. Each be rapidly degraded. Only a small per- 
synapse, in turn, consists of many types centage of the newly synthesized enzyme 
of specific macromolecules necessary for molecules become activated, and only 
transmitting and receiving chemical and these molecules are transported through 
electrical information between cells, as the Golgi apparatus and subsequently 
well as for maintaining the structural and exported, indicating that the cells can 
functional integrity of the synapse itself, discriminate catalytically active from in- 
Our major interest is the synthesis and active molecules. Although we do not yet 
regulation of macromolecules localized at understand the mechanisms of activation 
the neuromuscular junction, the synapse and sorting, our studies show that gly- 
formed between nerves and muscle fi- cosylation is a necessary prerequisite, 
bers. Our research has focused in par- Another aspect of this problem is the rapid 
ticular on the enzyme acetylcholinesterase degradation of inactive AChE molecules. 
(AChE), which is highly concentrated at This interesting observation suggests a 
the regions of nerve-muscle contact and specific and perhaps novel degradation 
is in part responsible for terminating the mechanism for sequestered proteins, 
actions of the neurotransmitter acetyl- Since the relative abundance of the sev- 
choline. eral catalytically active AChE forms in 

One aspect of our research this year cultured muscle is constant, yet different 
has concerned the early events in the from the pattern of total AChE forms 
synthesis and assembly of the several synthesized, the cells must have a spe- 
molecular forms of AChE found in tissue- cific mechanism for selecting the appro- 
cultured muscle. A unique and surprising priate forms to activate and process. At 
result is that the vast majority of AChE this point, we conclude that probably 



several post-translation controls are im- 
portant both for the regulation of AChE 
and for the expression of the multiple 
molecular forms of this enzyme in mus- 

Synthesis and Activation of AChE 

Newly synthesized, catalytically ac- 
tive AChE molecules consist primarily of 
dimers and tetramers. These assembled 
forms appear within minutes after syn- 
thesis, indicating that assembly and ac- 
tivation of the multimeric AChE forms 
are early events in their processing. These 
events most likely occur in the rough en- 
doplasmic reticulum (RER) because the 
time interval within which they occur is 
very short. These AChE molecules are 
still sensitive to endoglycosidase-H (Endo- 
H), which cleaves oligosaccharides of the 
type found on glycoproteins while still in 
the RER. Furthermore, most newly syn- 
thesized AChE molecules are not precip- 
itated by wheat germ agglutinin, which 

binds to a sugar added to glycoproteins 
only after entering the Golgi apparatus. 
When the synthesis of catalytically ac- 
tive AChE molecules is compared to mol- 
ecules incorporating 35 S-methionine, 
major discrepancies are found. Most of 
the catalytically active molecules consist 
of dimeric and tetrameric forms, whereas 
most 35 S-methionine-labeled molecules 
consist of monomers and dimers (Fig. 8). 
Furthermore, once synthesized and as- 
sembled, the catalytically active AChE 
forms are relatively stable, while the 35 S- 
methionine-labeled AChE polypeptides 
are rapidly degraded with an apparent 
half-life of about one hour. The results 
from several pulse-chase studies of en- 
zymatically active molecules and 35 S-me- 
thionine-labeled enzyme molecules 
indicate that only a small fraction, about 
10-20% of the total AChE synthesized, 
is activated and subsequently trans- 
ported through the cell. Clearly, the cell 
can distinguish between active and in- 

36 . 

a- 20 J 

£ 16 


J> 12 







j \ 

. » 



\ /'; 


2 12. 

it '. 




a. 10 



<T> 00 

1 1 



£ 4. 




/\ Jt 




/ ~ V J i 

J/\ I 

i i 

\ \ 
I \ \ 

\ y 

1 1 







Fig. 8. Comparison of newly synthesized, isotopically labeled and catalytically active AChE molecular 
forms. Tissue-cultured muscle cells were treated with an irreversible inhibitor of AChE and allowed to 
synthesize new AChE molecules in the presence of :i5 S-methionine. After one hour, the labeling medium 
was removed and replaced with chase medium containing excess methionine. At the designated time 
intervals, cultures were washed and extracted. The cell extracts were then analyzed by velocity sedi- 
mentation to resolve the individual AChE forms. After assaying for enzyme activity, the AChE in each 
fraction was immunoprecipitated, electrophoresed under denaturing conditions, and fluorographed. In- 
dividual bands of 35 S-methionine-labeled AChE were then excised from the gel and counted in a scin- 
tillating counter. 



active AChE molecules, and it processes 
the two populations differently. 

We intend to pursue several interest- 
ing questions. For example, we would 
like to know where in the cell the sorting 
decisions are made, where the nonex- 
portable AChE molecules are degraded, 
and what role muscle activity plays in 
these processes. 

Activation and Transport of 
AChE Requires Glycosylation 

Although we can demonstrate that most 
AChE molecules in tissue-cultured mus- 



cle are catalytically inactive, we do not 
yet understand the mechanism of acti- 
vation. There are no obvious modifica- 
tions of the AChE polypeptides detectable 
by NaDodS0 4 gel electrophoresis, nor are 
there obvious changes in glycosylation. 
What we do know, however, is that gly- 
cosylation is a necessary prerequisite for 
activation of AChE. Acetylcholinester- 
ase is glycosylated cotranslationally by 
the addition of approximately 12 kilo- 
daltons (kDa) of asparagine-linked oli- 
gosaccharides per polypeptide chain. 
Tunicamycin, a drug that blocks the ad- 
dition of asparagine-linked obligosaccha- 

A 8 





Fig. 9. Synthesis of inactive AChE in the presence of tunicamycin. Left panel: Muscle cultures were 
pre-incubated for three hours in the presence or absence of 1 jxg/ml tunicamycin, treated with an 
irreversible AChE inhibitor, and allowed to recover in the presence or absence of 1 fig/ml tunicamycin. 
For each timepoint, three cultures from each group were assayed for AChE activity. Right panel: Muscle 
cultures were pre-incubated as above, then labeled for five hours with 35 S-methionine in the presence 
or absence of 1 (xg/ml tunicamycin. The labeled AChE molecules in the cell extracts and medium were 
immunoprecipitated using monoclonal anti-chicken AChE antibodies, analyzed by NaDodS0 4 gel elec- 
trophoresis, and fluorographed. (A) Control cell-associated AChE; (B) tunicamycin cell-associated AChE; 
(C) control-medium AChE; (D) tunicamycin-medium AChE. Note that the AChE synthesized in the 
presence of tunicamycin shows the absence of about 12 kDa of asparagine-linked oligosaccharides. The 
major band at about 92 kDa from the tunicamycin-treated cells contains about 40% of the label found in 
the control AChE band. 



rides, also inhibits the appearance of 
newly synthesized catalytically active 
AChE molecules (Fig. 9). This is not due 
to inhibition of protein synthesis, be- 
cause total protein synthesis in treated 
cells is reduced by only 5% and the syn- 
thesis of AChE polypeptides continues 
at 40% the normal rate (Fig. 9, lanes A 
and B). This latter value may in fact be 
an underestimate because additional 35 S- 
methionine-labeled proteins appear that 
may be fragments of the AChE polypep- 
tides. In addition, unglycosylated AChE 
molecules are not exported by the cells. 
Normally, muscle cells secrete substan- 
tial amounts of AChE into the culture 
medium that can be labeled with ^-me- 
thionine, immunoprecipitated, and ana- 
lyzed by gel electrophoresis (Fig. 9, lane 
C). In the presence of tunicamycin, the 
AChE molecules are not secreted (lane 
D) but rather are retained by the cells. 

Preliminary results suggest that the un- 
glycosylated AChE polypeptides are de- 
graded intracellularly at the same rate 
as are the glycosylated (but catalytically 
inactive) AChE molecules. 

Only Catalytically Active AChE 

Molecules Are Processed in the 

Golgi Apparatus 

Intracellular catalytically active AChE 
molecules are stable. Pulse-chase studies 
of enzymatically active AChE, illus- 
trated in Fig. 10B, indicate that the cell- 
associated AChE activity does not change 
until the molecules are secreted by the 
cells. (See also Year Books 79-81 for ad- 
ditional studies.) Most of the stable, cat- 
alytically active AChE is released into 
the culture medium, and all of this AChE 
has been shown to be resistant to en- 
doglycosidase-H {Year Book 81). How- 











50 100 200 

ul WGA Bead sol. 

0.5 1 2 3 

Time Post Puromycin (hrs) 



60 w 






Fig. 10. All catalytically active AChE molecules pass through the Golgi apparatus. Left: Aliquots of 
medium AChE were incubated with varying amounts of wheat germ agglutinin (WGA), and the super- 
natants were assayed for AChE activity. Identical results are obtained with concanavalin A and ricin, 
indicating that all secreted AChE molecules pass through the Golgi apparatus. Right: Muscle cultures 
were treated with DFP, allowed to synthesize new AChE molecules for 30 min, then transferred to 
puromycin-containing medium to inhibit protein synthesis. At the indicated times, sets of three cultures 
were washed and extracted, and portions of the extracts incubated with or without WGA. The super- 
natants were then assayed for AChE activity. These results show that all catalytically active AChE 
molecules pass through the Golgi. The decrease in cell-associated AChE during the second and third 
hour of incubation reflects the secretion of catalytically active molecules into the culture medium. 


ever, the large intracellular pool of AChE vidual carbohydrate molecules linked to 
molecules that is rapidly degraded also glycoproteins. Concanavalin A, for ex- 
remains sensitive to endoglycosidase-H. ample, binds specifically to mannose res- 
These studies indicate that the catalyti- idues and can bind AChE molecules, both 
cally inactive AChE is either not trans- active and inactive, at all stages of syn- 
ported to the Golgi, or, if it is, must reach thesis, processing, and transport. On the 
only the earliest compartment, since en- other hand, N-acetylglucosamine, de- 
zyme molecules passing through this or- tected with wheat germ agglutinin 
ganelle also acquire endo-H resistance. (WGA), and terminal galactose, detected 
An alternative means of determining using ricin (RCA), are sugar residues 
where, within a cell, a given population which are attached to AChE in the cis 
of glycoprotein molecules is located is by and trans regions, respectively, of the 
the composition of their oligosaccha- Golgi. Using these lectins to precipitate 
rides. Specific sugar residues are added AChE, we can show that all catalytically 
in different subcellular compartments, and active AChE molecules pass through the 
the presence of a given sugar residue on Golgi apparatus before being exported 
a glycoprotein can be detected by its abil- by the cells (Fig. 10). This same popu- 
ity to bind a particular lectin. Lectins are lation of AChE molecules also becomes 
plant proteins which have very high af- resistant to Endo-H, another indication 
finities for specific carbohydrate residues that oligosaccharide processing has oc- 
and hence can be used to identify indi- curred. 


Martin D. Snider and Ophelia C. Rogers 

All cells synthesize macromolecular assembled on membrane-bound ribo- 

products, which they export to the cell somes on the rough endoplasmic reticu- 

surface and extracellular space. This ex- lum (ER) and extruded directly into the 

port poses an important problem in eel- lumen of this organelle. Much less is 

lular organization because these known about our area of interest — the 

macromolecules are assembled from cy- mechanism by which the sugar residues 

toplasmic precursors. Thus, the syn- that make up the oligosaccharide portion 

thetic apparatus must be organized so of these molecules are transported from 

that molecules which cannot cross mem- the cytoplasm during synthesis, 

branes when mature are transferred out Asparagine-linked oligosaccharides are 

of the cytoplasm and across a membrane assembled in two distinct stages. In the 

during assembly. first stage, a large lipid-linked precursor 

We have been investigating this prob- oligosaccharide is assembled in the mem- 
lem for the synthesis of oligosaccharides brane of the rough ER and then trans- 
linked to asparagine residues of eukar- ferred as a unit to nascent and newly 
yotic glycoproteins. This important class made polypeptides. The precursor oli- 
of macromolecule comprises a large frac- gosaccharide, Glc 3 Man 9 GlcNAc 2 , which 
tion of both cell surface and secreted pro- is linked to the polyisoprenoid lipid dol- 
teins in all eukaryotes. The transport of ichol pyrophosphate, is assembled by the 
the polypeptide portion of these mole- stepwise transfer of single-sugar resi- 
cules from the cytoplasm during synthe- dues derived from cytoplasmic nucleo- 
sis is well understood; peptides are tide sugars. In the second stage of 



oligosaccharide assembly, protein-linked 
oligosaccharides are extensively modi- 
fied, by both the removal and addition of 
sugar residues, to give the diverse array 
of final products. These modifications oc- 
cur during the transport of glycoproteins 
from the rough ER, through the Golgi 
apparatus, to the cell surface. 

Our investigations have concerned the 
topography of the first stage of aspara- 
gine-linked oligosaccharide synthesis, the 
assembly of the lipid-linked oligosac- 
charide, and its transfer to protein in the 
rough ER. Sugar residues must move 
across the rough ER membrane in some 
chemical form during this process, since 
the lipid-linked precursor is assembled 
from cytoplasmic nucleotide sugars, and 
is transferred to protein at the luminal 
face of the rough ER. We have ap- 
proached this problem by studying the 
orientation in the ER membrane of a 
number of oligosaccharide-lipid inter- 
mediates in the assembly of 
Glc 3 Man 9 GlcNAc 2 -lipid. In these exper- 
iments, we used microsomal vesicles pre- 

pared from cultured hamster fibroblasts. 
These sealed vesicles, which are derived 
from the ER, have a unique orientation: 
The cytoplasmic side of the membrane 
faces the medium. Thus, reagents that 
cannot enter the vesicles can be used to 
probe these membranes, as they will act 
only on the cytoplasmic side of intact ves- 

We have used the plant lectin con- 
canavalin A (Con A), which binds to oli- 
gosaccharide-lipids, to probe the 
orientation of oligosaccharide-lipids. 
Several years ago, we developed an as- 
say to measure Con A binding and were 
able to show that the largest species, 
Glc 3 Man 9 GlcNAc 2 -lipid, faces the micro- 
somal lumen in vesicles prepared from 
[ 3 H]Man-labeled cells. In the past year, 
we have used the same assay to examine 
the orientation of smaller oligosacchar- 
ide-lipids. Figure 11 shows the binding 
of Con A to lipid-linked Man 8 GlcNAc 2 - 
Glc 3 Man 9 GlcNAc 2 in intact and leaky mi- 
crosomal vesicles. All of these species are 
not bound by Con A in intact vesicles, 










50 - 

Fig. 11. Binding of oligosaccharide-lipids by Con A in intact and leaky microsomal vesicles. Micro- 
somal vesicles from [ 3 H]-Man-labeled Chinese hamster ovary cells were incubated with Con A in the 
absence (left-hand bar) or presence (right-hand bar) of nonionic detergent. Binding of lectin to each 
species was measured by the shift of oligosaccharide-lipids that are bound by Con A from a chloroform- 
methanol-soluble pool to an insoluble pool. G, glucose; M, mannose; N, N-acetylglucosamine. 



but are bound by the lectin in leaky ves- 
icles. Because binding occurs only when 
the luminal side of microsomes is exposed 
to lectin, all of these oligosaccharide-lip- 
ids face the microsomal lumen. More- 
over, because these species are adjacent 
intermediates in the pathway, it is likely 
that the synthesis of Glc 3 Man 9 GlcNAc 2 - 
lipid from Man 8 GlcNAc 2 -lipid involves the 
addition of one Man and three Glc resi- 
dues at the luminal face of the rough ER. 
Similar experiments using Con A to 
study Man 5 GlcNAc 2 -lipid suggest that this 
species has the opposite orientation in 
the microsomal membrane (Fig. 12). This 
species is bound by Con A in both intact 
and leaky microsomal vesicles. Figure 12 
also shows that the binding is specific, as 
it is blocked by the competitor yeast 
mannan. The binding of Man 5 GlcNAc 2 - 
lipid by Con A in intact vesicles when 
only the cytoplasmic side of the vesicle 
membrane is accessible to lectin suggests 
that this species faces the cytoplasmic 
side of the microsomal membrane. We 
also have preliminary evidence that 

100 - 










wt 5 2 









Man 3 GlcNAc 2 - and Man 4 GlcNAc 2 -lipids 
have similar orientations. 

These results lead us to propose a model 
for oligosaccharide-lipid assembly which 
involves the initiation of synthesis on the 
cytoplasmic side of the rough ER, trans- 
location of an oligosaccharide-lipid inter- 
mediate from the cytoplasmic to the 
luminal face of the membrane, and com- 
pletion of Glc 3 Man 9 GlcNAc 2 -lipid on the 
luminal side (Fig. 13). The initiation of 
oligosaccharide synthesis on the cyto- 
plasmic side of the rough ER is sug- 
gested by the location of Man 5 GlcNAc 2 - 
lipid (and probably Man 3 . 4 GlcNAc 2 -lipid) 
on this side of the membrane. This lo- 
cation is quite remarkable, as most gly- 
cosyltransferases and substrates involved 
in the synthesis of other glycoproteins 
and glycolipids are found within the lu- 
men of the ER and Golgi. Also of interest 
is the proposed translocation of an oli- 
gosaccharide-lipid intermediate — an am- 
phipathic molecule containing a large 
hydrophilic oligosaccharide — across the 
rough ER membrane during synthesis. 
While we have no direct evidence as to 
how this translocation occurs, a specific 
protein is probably involved, since it has 
been found that dolichol derivatives do 
not move spontaneously across lipid bi- 




■ 2 i 5 

5 2\\ V 


Fig. 12. Binding of Man 5 GlcNAc 2 -lipid by Con 
A in intact and leaky microsomal vesicles. Labeled 
vesicles were prepared from wild-type Chinese 
hamster ovary cells or from B4-2-1, a mutant line 
that contains Man 5 GlcNAc 2 as its major lipid-linked 
oligosaccharide, and incubated with Con A and, 
where indicated, the nonionic detergent Triton X- 
100 (Tx-100) and yeast mannan. Con A binding was 
measured as in Fig. 11. Similar results were ob- 
tained when the vesicles were prepared from the 
two cell lines. 

Fig. 13. Model for the topography of oligosac- 
charide-lipid synthesis in the ER membrane. As- 
sembly of luminal Glc :i Man 9 GlcNAc 2 -lipid begins with 
the synthesis of Man 5 GlcNAc 2 -lipid on the cyto- 
plasmic face of the membrane. After the translo- 
cation of an intermediate across the membrane, 
oligosaccharide-lipid is completed on the luminal 
face. While the translocated intermediate is shown 
as Man 5 GlcNAc 2 -lipid, its identity is not known for 


layers. The identity of the translocated the ER membrane explains how sugar 

intermediate is also unknown, although residues are exported from the cyto- 

it is probably Man 5 _ 7 GlcNAc 2 -lipid, since plasm during the first stage of aspara- 

larger intermediates are found exclu- gine-linked oligosaccharide assembly, and 

sively on the luminal side of the mem- suggests that this translocation may be 

brane. In the coming year, we plan to a major reason for the participation of 

identify the translocated intermediate and dolichol-linked oligosaccharides in gly- 

set up an in vitro system that will allow coprotein synthesis. Moreover, our un- 

us to characterize this translocation in derstanding of the topography of 

detail. The most-likely candidate for oligosaccharide-lipid assembly should be 

translocation is Man 5 GlcNAc 2 -lipid, since applicable to a number of other biosyn- 

different types of sugar donors are used thetic processes that involve polyiso- 

to make this compound and to elongate prenoid lipid carriers. These include the 

it to the mature oligosaccharide-lipid; cy- synthesis of polysaccharides in plants, as 

toplasmic sugar nucleotides are the do- well as the synthesis of cell walls, lipo- 

nors for Man 5 GlcNAc 2 -lipid synthesis, polysaccharide, and teichoic acids in bac- 

while dolichol-linked monosaccharides, teria. Finally, our studies underline the 

which may also be translocated across notion that biological membranes active 

the rough ER membrane in a manner in the synthesis and export of macrom- 

similar to oligosaccharide-lipid, donate olecules are capable of overcoming the 

the last seven residues in the synthesis permeability barrier of the lipid bilayer 

of Glc 3 Man 9 GlcNAc 2 -lipid. Translocation in a specific way. While most membranes 

of Man 5 GlcNAc 2 -lipid offers an explana- are impermeable to large or hydrophilic 

tion for the use of these two different molecules, it is now clear that complex 

types of sugar donors. carbohydrates, peptides, and lipid mol- 

We believe that our studies on oligo- ecules can be specifically translocated 

saccharide-lipid topography make a sig- across the membranes that are involved 

nificant addition to the understanding of in the export of these macro- 

glycoprotein synthesis and export. The molecules from the cytoplasm during 

movement of oligosaccharide-lipid across synthesis. 



N. G. Lipsky, 0. C. Martin, R. G. Sleight, and R. E. Pagano 

Lipid molecules are as much a part of translocation of lipids present sorting and 

membrane biogenesis and membrane dif- targeting problems analogous to those 

ferentiation as are membrane proteins, encountered with cellular proteins, whose 

Although the enzymes responsible for lipid assembly into membranes, turnover, and/ 

biosynthesis reside primarily on the rough or secretion represent one of the most 

and smooth endoplasmic reticulum, lip- actively studied areas of cell biology to- 

ids are found throughout the cell, with day. 

different intracellular membranes often Until now, however, only limited in- 
having different lipid compositions. Ad- formation has been obtained about the 
ditionally, one leaflet of a given mem- mechanisms of lipid translocation within 
brane bilayer (e.g., the plasma membrane) eukaryotic cells. This is because of var- 
may have a different lipid composition ious shortcomings in the traditional 
from the other. Thus, the synthesis and methods used to examine lipid metabo- 


lism at the subcellular level. For exam- certain lipid species. Finally, Pagano 

pie, in the "pulse-chase" experiment, presents the results of microinjection ex- 

which involves the uptake of radiolabeled periments designed to elucidate the 

lipid precursors followed by cell fraction- mechanism(s) by which lipids are trans- 

ation, labeled lipids may exchange be- located from one intracellular compart- 

tween isolated membrane fractions, either ment to another, 
by the action of soluble exchange pro- 
teins or by the tendency of certain lipids 

to transfer spontaneously between mem- Internalization of Fluorescent 

branes. Thus, the appearance with time Phosphatidylcholine and 

of a particular lipid species in a given Phosphatidylethanolamine 

membrane fraction may or may not be D . , , n c , . ,. 

. ,. r» i , 11 • Richard G. Sleight 
indicative of what actually occurs in situ. 

Another major problem faced by the cell Very little is currently known about 
biologist or lipid biochemist interested in the intracellular trafficking of phospho- 
the movement of lipids within cells is that lipids, mainly because of the lack of spe- 
except for electron microscopic autora- cific probes for individual phospholipid 
diographic techniques, which are tedious molecules. As described in previous edi- 
and require special attention to lipid re- tions of this Year Book, our laboratory 
tention during sample processing, the lo- has synthesized a series of fluorescent 
cation of lipids within intact cells cannot derivatives of many common phospho- 
be readily determined. This is because, lipid species. By visualizing the move- 
in contrast to studies of protein trans- ment of these fluorescent molecules in 
location, which can use tagged and spe- living cells, we hope to elucidate the 
cific antibodies or toxins for examining mechanisms that control the transport 
subcellular distribution, no specific probes and sorting of phospholipids. This infor- 
for the major classes of lipids exist. mation should add greatly to our under- 
To circumvent some of the problems standing of the metabolic regulation of 
outlined above, we have continued our living tissue. 

studies with fluorescent lipid analogs (see When V79 Chinese hamster fibro- 
Year Books 78-81). This approach is po- blasts are incubated at 2°C in the 
tentially very powerful, since we can carry presence of either C 6 -NBD-PC or C 6 - 
out traditional experiments in lipid me- NBD-PE, some of the fluorescent lipid is 
tabolism, then correlate these biochem- transferred from the vesicles to the cells' 
ical data with "positional" information on plasma membrane (Fig. 14A and C). As 
the intracellular location of the lipid me- long as the cells are maintained at 2°C, 
tabolites in living cells using conven- the fluorescent phospholipids remain re- 
tional microscopy. In this year's report, stricted to the plasma membrane. If, 
Sleight and Lipsky present the results however, the liposomes are removed and 
of their studies with fluorescent analogs the cells are warmed to 37°C, some of the 
of phosphatidylcholine, phosphatidyl- fluorescent lipid is internalized (Fig. 14B 
ethanolamine, ceramide, sphingomyelin, and D). We have found that the subcel- 
and cerebroside. One of the most inter- lular distributions of the internalized Ch- 
esting and exciting aspects of their work NBD-PC and C 6 -NBD-PE are different, 
is that each analog is metabolized and The majority of the internalized fluores- 
translocated differently within the cell, cent PC accumulates in a perinuclear re- 
Martin presents a recently devised se- gion of the cell, while a smaller amount 
lection procedure based on fluorescence is observed as random, punctate spots, 
photosensitization. We hope to use this These punctate spots are most likely the 
method in the near future to isolate mu- result of the presence of fluorescent en- 
tant cells which are defective in their docytic vesicles. In contrast, C 6 -NBD- 
ability to metabolize and/or translocate PE enters the cells at 37°C and accu- 



Fig. 14. Fluorescent photomicrographs of cells incubated with C 6 -NBD-PC or C 6 -NBD-PE at 2°C 
and then warmed to 37°C. V79 Chinese hamster fibroblasts were incubated for 30 min at 2°C in the 
presence of liposomes containing either 40 mol % C, r NBD-PC or 50 mol % C 6 -NBD-PE. The cells were 
then washed to remove the liposomes and incubated at 37°C for 30 min. To visualize internal fluorescence, 
the warmed cells were incubated at 2°C in the presence of dioleoylphosphatidylcholine vesicles. Most of 
the fluorescence from the cells' plasma membrane is transferred to the lipid vesicles under these "back- 
exchange" conditions. Cells treated with C 6 -NBD-PC at 2°C are displayed in (A) and, after subsequent 
warm-up and back-exchange, in (B). Cells treated with C 6 -NBD-PE at 2°C are displayed in (C) and, 
after subsequent warm-up and back-exchange, in (D). 

mulates in the nuclear envelope and mi- 
tochondria. A small amount of internalized 
fluorescent PE also appears in a peri- 
nuclear region and as punctate spots. 

To determine how endocytosis may 
contribute to the internalization of flu- 
orescent membrane lipids, we can ob- 
serve the distribution of fluorescent lipids 
in cells preincubated in 5 mM sodium azide 
plus 50 mM 2-deoxyglucose. This mix- 
ture of metabolic inhibitors has been 
shown to greatly decrease endocytosis. 
When cells are prelabeled with C 6 -NBD- 
PC at 2°C and then warmed to 37°C for 
30 minutes in buffer containing the in- 
hibitors, no internalization of fluores- 

cence is observed. In contrast, 
internalization occurs when a similar ex- 
periment is performed using C 6 -NBD-PE. 
Although the nuclear envelope and mi- 
tochondria of these cells are fluorescent, 
incubation in the presence of the inhibi- 
tors blocks the appearance of the peri- 
nuclear and punctate staining seen in cells 
incubated in the absence of inhibitors (Fig. 

When V79 fibroblasts are incubated at 
2°C in the presence of rhodamine-con- 
jugated Lens culinaris agglutinin, flu- 
orescent staining is observed only at the 
plasma membrane. After warming the 
cells to 37°C, much of the labeled lectin 



is internalized. A comparison of the in- 
ternal fluorescent patterns of cells incu- 
bated with both C 6 -NBD-PC and 
rhodamine Lens culinaris agglutinin at 
2°C, and then warmed to 37°C for 30 
min, is presented in Fig. 15. Initially, at 
2°C, both the fluorescent lipid and flu- 
orescent protein are located at the plasma 
membrane. The simultaneous internali- 
zation of the two fluorescent molecules is 
followed by their sorting. Although the 
perinuclear area of the cells labeled with 
C 6 -NBD-PC also contains some rhoda- 
mine label, the intensity of the rhoda- 
mine staining is significantly reduced. 

Our current experiments are being de- 
signed to examine the mechanisms that 
result in the different patterns of inter- 
nalization observed using C 6 -NBD-PC and 
C 6 -NBD-PE. Our working hypothesis is 
that the majority of the difference occurs 
because the two lipids move across the 
membrane at different rates. That is, if 
both C 6 -NBD-PC and C 6 -NBD-PE are 
inserted at 2°C only into the outer leaflet 

of the plasma membrane's lipid bilayer, 
we envision that upon warming to 37°C 
only C 6 -NBD-PE crosses the bilayer (i.e., 
flip-flops). Because the fluorescent PE is 
now exposed to the cytosol, it may be 
able to transfer directly to intracellular 
organelles, most likely by a monomer dif- 
fusion mechanism similar to that used by 
the lipids when moving from the lipo- 
somes to the cells' plasma membrane. In 
addition, some of the C 6 -NBD-PE may 
enter by endocytosis, and, if flip-flop of 
C 6 -NBD-PE occurs in endocytic vesicles, 
more of the lipid will have a chance to 
move through the cytosol. In contrast to 
C 6 -NBD-PE, C 6 -NBD-PC may not be able 
to undergo transmembrane movement 
and, therefore, can enter the cells only 
by endocytosis. The endocytic vesicles 
would carry the fluorescent PC mole- 
cules in the inner leaflet of their lipid 
bilayer, shielding them from the cytosol. 
All of the data we have obtained to 
date fits this working hypothesis. Fur- 
ther studies should allow us to determine 

Fig. 15. Co-localization of internalized C 6 -NBD-PC and rhodamine-labeled Lens culinaris agglutinin. 
Cells were incubated for 30 min at 2°C in the presence of both C, r NBD-PC-containing liposomes and 5 
jxg/ml rhodamine-labeled Leyis culinaris agglutinin. The cells were then warmed to 37°C and back- 
exchanged, as described in the legend to Fig. 14. Photomicrographs were taken using optics appropriate 
for viewing either (A) NBD or (B) rhodamine fluorescence. 



more precisely the mechanisms involved 
in the transport and sorting of these 

Studies on Fluorescent 

N. G. Lipsky 

In our previous studies (Year Book 81), 
we found that fluorescently labeled cer- 
amide, a potential sphingolipid precur- 
sor, was metabolized by fibroblasts to a 
fluorescent cerebroside and fluorescent 
sphingomyelin. The distribution of fluo- 
rescence in the cell initially was in the 
mitochondria and ER; later, a region 
identified as the Golgi apparatus became 
fluorescent, and ultimately the plasma 
membrane became markedly labeled. We 
have extended these studies to investi- 
gate (1) the behavior of the individual 
metabolites after their addition to fibro- 
blasts, and (2) the effects of metabolic 
inhibitors on these fluorescently labeled 

In order to study the interactions be- 
tween the cells and the individual flu- 

orescent metabolites, it was necessary to 
synthesize or purify the NBD-com- 
pounds in large quantities. We first de- 
termined that the fluorescent cerebroside 
was glucosylceramide. We purified this 
compound from labeled cells in minute 
quantities sufficient for fluorescence mi- 
croscopy. We were able to chemically 
synthesize larger quantities of fluores- 
cent NBD-galactosylceramide and NBD- 
sphingomyelin for use in more extensive 
studies. When these compounds (NBD- 
glucosylceramide, NBD-galactosylcer- 
amide, and NBD-sphingomyelin) were 
individually incorporated into liposomes 
and incubated with fibroblasts at 4°C, all 
three exhibited the same behavior: They 
remained in the fibroblast plasma mem- 
brane (Fig. 16B). This is in contrast to 
the NBD-ceramide, which under the same 
conditions appears almost entirely within 
the cell (Fig. 16A). In addition, even af- 
ter subsequent incubation at 37°C, very 
little internal labeling was seen. NBD- 
sphingomyelin was not significantly de- 
graded under these conditions. These re- 
sults are consistent with the known 
distribution of endogenous sphingomye- 

Fig. 16. Distribution of NBD-sphingomyelin and NBD-ceramide in cultured Chinese hamster fibro- 
blasts. Cells were incubated at 4°C with liposomes containing either (A) NBD-ceramide or (B) NBD- 



lin and cerebroside, which are found pre- 
dominantly in the plasma membrane. 

We next investigated the effects of a 
variety of inhibitors on the metabolism 
of NBD-ceramide. Neither sodium azide, 
an inhibitor of respiration, nor chloro- 
quine, a lysosomal inhibitor, affected the 
appearance or metabolism of NBD-cer- 
amide in the cells. However, the addition 
of monensin, an ionophore which reduces 
the export of glycoproteins from the Golgi 
to the plasma membrane, caused a dra- 
matic change in the appearance of labeled 
cells. Figure 17 demonstrates the ap- 
pearance of cells labeled with NBD-cer- 
amide and then incubated for 45 min at 
37°C, either in the (A) absence or (B) 
presence of 10 jxM monensin. It can be 
seen that the Golgi in the presence of 
monensin remains extremely bright and 
enlarged relative to the control; plasma 
membrane labeling appears to be re- 
duced in these cells. Biochemical analysis 
of these cells indicates that monensin does 
not cause a significant change in the 
amounts or rates of synthesis of the flu- 
orescent metabolites produced by the fi- 

broblasts. It does seem to change the 
intracellular distribution of the metabo- 
lites in a striking manner. While it has 
been established that monensin inter- 
feres with the cellular transport of gly- 
coproteins, little information is available 
on how it affects glycolipids. 

These studies suggest that monensin 
inhibits the transport of some or all of 
the fluorescent lipids to the plasma mem- 
brane. Further biochemical studies are 
under way to determine whether both 
NBD-cerebroside and NBD-sphingo- 
myelin distributions are altered and, if 
so, if they are altered in parallel. Results 
of such studies will provide valuable in- 
formation on the subcellular sites of syn- 
thesis and routes of translocation of the 

Isolation and Characterization 

of Mutants Using a Fluorescence 

Photosensitization Method 

O. C. Martin and R. E. Pagano 

We are attempting to isolate mutant 
cells that are defective in their ability to 

Fig. 17. Effect of monensin on NBD-ceramide treated cells. Cells were incubated at 4°C with lipo- 
somes containing NBD-ceramide, then rinsed and incubated in medium at 37°C for 45 min. (A) control; 
(B) all steps in the presence of 10 |xM monensin. 



take up and/or translocate various flu- 
orescent lipid analogs. With such mu- 
tants we hope to determine whether (and 
how) metabolic conversion of one lipid 
species to another affects intracellular 
transport, and whether the processes in- 
volved are tightly coupled. In our initial 
studies, we limited our attention to one 
fluorescent molecule, C 6 -NBD-PA — a 
fluorescent analog of phosphatidic acid 
(see Year Books 80, 81). These studies 
were also restricted to one cell type — 
cultured Chinese hamster V79 lung cell 

We first developed an effective selec- 
tion procedure based on the killing of cells 
by fluorescence photobleaching, a method 
used in the isolation of mutants. Our pro- 
cedure is similar to that described by 
Brown, Goldstein, and co-workers, for 
the isolation of mutants in the LDL-re- 
ceptor pathway. Cells are treated with 
C 6 -NBD-PA for 60 min at 2°C, washed, 
and then uniformly irradiated (at 2°C) 
using various intensities of light from an 

argon laser tuned to 457.9 nm. The cells 
are then cultured and the number of sur- 
vivors at each dose are quantified. As 
shown in Fig. 18, C 6 -NBD-PA-treated 
cells are much more susceptible to killing 
than are the non-vesicle-treated con- 
trols, suggesting that the presence of the 
fluorescent lipid in cellular membranes is 
specifically required for photodamage and 
subsequent death. The site of photoda- 
mage in these irradiated cells may be the 
mitochondria, since immediately follow- 
ing irradiation at high intensities, mito- 
chondria no longer take up the vital dye 
rhodamine 3B. 

We next obtained a mutagenized pop- 
ulation of cells using ethyl methane sul- 
fonate (EMS). Cells were treated with 
various concentrations of the mutagen for 
24 hours, washed, and allowed to grow 
for five or more doublings to produce sta- 
ble phenotypic expression. The number 
of surviving clones compared to control, 
non-mutagenized cells was then deter- 
mined after fixation and staining. Effi- 








50 100 150 200 

LASER IRRADIATION AT 457.9 nm (mWxtime) 


Fig. 18. Comparison of survival of control and mutant populations of cultured fibroblasts after pho- 
tosensitization with a fluorescent analog of phosphatidic acid. A non-mutagenized population of cells was 
incubated in a simple balanced salt solution (open circles) or with C 6 -NBD-PA (closed circles) for 60 min 
at 2°C. The cells were then washed, irradiated at 2°C at various intensities of light (457.9 nm), and 
cultured, and the percentage of survivors at each dose was quantified. Identical treatment of a muta- 
genized population of cells with C 6 -NBD-PA was also carried out (closed triangles). 


ciency of mutagenesis was assessed by ence of a cellular mechanism which 

testing for thioguanine resistance. To somehow "overrides" the property of 

minimize multiple mutations in the same spontaneous lipid transfer seen with these 

cell, we chose the lowest possible con- analogs in model membranes. A proper 

centration of EMS that would still result understanding of this paradox may yield 

in a sufficient number of mutant cells. In fundamental information about the 

this series of experiments, we found that mechanism or mechanisms whereby lipid 

only about 3% of the cells survive at 650 molecules are translocated from their sites 

fxg EMS/ml. The percentage of thio- of synthesis within cells to other cellular 

guanine-resistant clones among survi- membranes where they are required for 

vors was about four times greater than membrane assembly, 

among control, non-mutagenized cells. One possible explanation for this par- 

We then subjected the mutagenized adox is that the distribution of fluores- 

population of cells to several rounds of cent lipids among intracellular membranes 

selection, using the photosensitization simply reflects compositional differences 

method, and obtained a population of mu- or some physical property of the differ- 

tant cells that was very resistant to pho- ent membranes, and that, as a result of 

tosensitization after treatment with C 6 - these differences, the various lipids "par- 

NBD-PA (Fig. 18). We are currently tition" differently into the various com- 

cloning these cells, and intend to char- partments. In support of this idea are 

acterize any interesting clones in detail data we obtained showing that the rate 

during the coming year. In particular, we at which an NBD-lipid enters or leaves 

will be looking for (1) uptake-defective a liposome depends both on the molecular 

mutants which fail to take up any flu- structure of the NBD-lipid and on the 

orescent lipid, and (2) localization-defec- characteristics (composition) of the do- 

tive mutants, which may result in an nor and acceptor liposomes. While this 

altered intracellular distribution of the may be a partial explanation for our re- 

C 6 -NBD-PA and its fluorescent metab- suits with cells, it is difficult to imagine 

olites. that the nearly complete restriction of 

certain NBD-lipids to certain organelles 

is the result solely of differences in or- 

Intracellular Processing of ganelle composition. 

Lipids: A Theory Based on As an alternative mechanism, one con- 

Microinjection of Fluorescent sistent with our results to date, I suggest 

Lipids into Single Cells that the various fluorescent lipids are 

R F p compartmentalized within the cell and are 

restricted to the luminal surface of an 

In our studies of the metabolism and organelle or cellular vesicle. As a result, 

translocation of fluorescent lipids in cul- these lipids cannot spontaneously trans- 

tured mammalian cells, a paradox has ari- fer between intracellular membranes by 

sen: While fluorescent lipid analogs un- monomer diffusion, but can move only by 

dergo rapid spontaneous transfer be- some kind of transport mechanism. For 

tween model membranes in vitro (see Year example, consider a lipid such as C 6 -NBD- 

Book 81, 159), they do not appear to ex- PC, which can be inserted into the plasma 

hibit this property when incorporated into membrane of cells at 2°C. From the to- 

intracellular membranes of living cells pology of the internalization process it 

(e.g., the plasma membrane, endo- can be argued that fluorescent lipids in- 

plasmic reticulum, or the Golgi appara- itially present in the external leaflet of 

tus). Rather, during cellular metabolism, the plasma membrane become localized 

one fluorescent lipid may specifically move on the inner leaflet (or lumen) of endo- 

to one organelle while another is trans- cytic vesicles. If this NBD-lipid does not 

located elsewhere, suggesting the pres- readily flip-flop in vivo, as in the case of 



our model membrane studies, it could not 
reach the opposite leaflet of these en- 
docytic vesicles and, therefore, could not 
spontaneously transfer into other intra- 
cellular membranes and render them flu- 
orescent. Similar considerations might 
also hold true for C 6 -NBD-PA and C 6 - 
NBD-ceramide. Once localized to a par- 
ticular organelle, the enzymes of lipid 
biosynthesis existing there could operate 
on the fluorescent lipid substrate to con- 
vert it to the observed fluorescent lipid 
metabolites. All the while, because these 
products are restricted to the luminal face 
of the organelle, they cannot readily 
transfer into other membranes, such as 
the plasma membrane, even though 
readily exchangeable lipids such as C 6 - 
NBD-PC are being made. To extend the 
model further, each of the fluorescent lip- 
ids formed will ultimately be translo- 
cated to other compartments only by a 
directed flow or transport process, per- 
haps involving intracellular vesicles. Any 
"sorting" of lipids that occurs might be 
caused by the restriction of certain en- 
zymes in lipid metabolism to particular 
regions of, for example, the endoplasmic 
reticulum (ER) from which lipid trans- 
port originates. 

To test the idea of compartmentaliza- 
tion of lipids, we have begun a series of 
studies wherein fluorescent lipids are mi- 
croinjected directly into the cytoplasm of 
living cells, and their distribution ob- 
served. In the case of C 6 -NBD-PC- 
treated cells, we know that when C 6 -NBD- 
PC present in the plasma membrane is 
internalized, it is not randomized 
throughout the cell, but appears in a dis- 
crete intracellular location. According to 
the model developed above, if this same 
fluorescent lipid were on the opposite side 
(cytoplasmic face) of the presumed en- 
docytic vesicles, it would then be free to 
undergo lipid transfer, randomize 
throughout the cell, and label all internal 
membranes. While we cannot microin- 
ject endocytic vesicles that are NBD-la- 
beled in this way to test this idea, we 
can microinject C 6 -NBD-PC directly into 
the cytoplasm of living cells. When this 

is done, we find prominent labeling of the 
ER, mitochondria, nuclear membrane, 
and other structures (Fig. 19). This ap- 
pearance is in marked contrast to that 
seen when cells internalize plasma mem- 
branes "stained" with C 6 -NBD-PC (see 
Fig. 14B). Microinjections of imperme- 
able but potent quenchers of fluorescence 
are currently being explored as addi- 
tional tools for addressing the sidedness 
of NBD-lipids within living cells. In the 
coming year, we hope to test rigorously 
the working hypothesis posed above and 
perhaps solve this interesting paradox. 

Fig. 19. Microinjection of C 6 -NBD-PC. A sin- 
gle cell was microinjected with fluorescent phos- 
phatidylcholine and then photographed using 
fluorescence and phase optics. 




S. G. Lazarowitz and B. A. Conde 

The geminiviruses are a recently rec- We have chosen to characterize two 

ognized group of plant pathogens (Mat- geminiviruses: (1) squash leaf curl virus 

thews, Intervirology 12, 129, 1979) of (SqLCV), which is transmitted by the 

worldwide economic importance. Their whitefly B. tabaci; and (2) beet curly top 

structural features of a paired icosahed- virus (CTV), which is transmitted by the 

ral capsid (hence gemini, twin moons) and leafhopper C. tenellus. 
a genome of covalently closed, circular, 
single-stranded DNA make them unique 

among eukary otic viruses. They are found MoLECULAR Studies of Squash 
m tropical or subtropical regions and are Lfaf Curt Virus 
transmitted in nature by either white- 
flies or leafhoppers. While most gemi- SqLCV is currently a serious problem 
niviruses, in common with other plant in cucurbits in the Imperial Valley of Cal- 
viruses, have a restricted host range, as ifornia. It has proven to be less amenable 
a group they infect a wide variety of mono- to study using classical approaches of 
cotyledonous and dicotyledonous plants transmission and isolation than have other 
including important crops such as maize, whitefly-transmitted geminiviruses and 
tomatoes, wheat, and beans (see Good- thus, in collaboration with J. A. Dodds 
man, in Handbook of Plant Virus Infec- of the University of California, River- 
tions and Comparative Diagnosis, 879, side, we have undertaken the study of 
Elsevier, North Holland, and Biomedical this agent using recombinant DNA tech- 
Press, N.Y., 1981). niques. The molecular probes generated 

Little is known about the biology of in using this approach will allow us to 
geminiviruses owing largely to the fact analyze the molecular biology of SqLCV 
that they are tightly phloem-associated as a model gemini virus, and help us un- 
and very difficult to grow and isolate, derstand the epidemiology of SqLCV in 
Our knowledge of these pathogens on the the Imperial Valley, 
molecular level comes almost exclusively SqLC V-infected material was supplied 
from studies on whitefly-transmitted to us by Dr. Dodds in the form of lyoph- 
agents (Haber et al., Nature 289, 324, ilized tissue. Four- to five-week-old sum- 
1981; Hamilton et al., Nucl. Acids Res. mer squash plants were inoculated with 
10, 4901, 1982; Stanley and Gay, Nature SqLCV in California using viruliferous 
301, 200, 1983), since these grow to higher whiteflies. Leaves and petioles were har- 
titers and have proven to be easier to vested about two weeks postinfection, at 
transmit mechanically than their leaf- which time the plants displayed severe 
hopper-transmitted counterparts. These symptoms of SqLCV. Tissue was ex- 
studies on whitefly-transmitted gemini- tracted with 0.5 M KH 2 P0 4 , pH 7.2-2.5% 
viruses have demonstrated that the gen- Triton X-100-0.75% Na 2 S0 3 , and a crude 
ome consists of two unique closed, circular 200,000 x g nucleoprotein pellet was pre- 
single-stranded DNAs (ssDNA) of —2.8 pared (Hamilton et al., Nucl. Acids Res. 
kb each, and there appears to be a single 10, 4901, 1982). Following solubilization 
ssDNA encapsidated per paired icosa- in SDS and extraction with phenol: 
hedron capsid. However, recent studies chloroform-1% isoamyl alcohol (1:1) the 
on a leafhopper-transmitted geminivirus nucleic acids were analyzed on agarose 
demonstrate the presence of only a single gels, as shown in Fig. 20 A. Preparations 
circular genomic DNA of -2.8 kb (R. from both infected and uninfected squash 
Symons, personal communication). tissue contained some broken-down host 


a b c d e f 



a b c d e f 

Fig. 20. (A) Ethidium bromide-stained agarose gel of nucleic acid extracts from squash tissue, and 
(B) Southern blot of gel shown in (A) probed with BGMV DNA. (a,b) Six-week-old uninfected squash 
plants incubated (a) with SI nuclease and (b) without enzyme. (c,d,e) SqLCV-infected squash, inoculated 
with viruliferous whiteflies at four weeks and harvested showing severe symptoms at six weeks of age, 
(c) incubated with SI nuclease, and (d and e) without enzyme, (f) Aliquot of (e) incubated with pancreatic 
DNase. All samples were incubated with pancreatic RNase before gel analysis. Arrow shows position 
of marker, supercoiled 2.7-kb dsDNA. 

cell DNA (Fig. 20A and D) and tRNA- 
sized RNA (not shown). In addition, the 
SqLCV-infected tissue was found to con- 
tain a species of the approximate mobil- 
ity expected for geminivirus DNA (Fig. 
20 A and D). The yield of the small viral- 
sized component was —0.5 |JLg/g wet 
weight tissue. As shown in Fig. 20, both 
the host cell and the viral-sized compo- 
nents were resistant to digestion with 
pancreatic RNase, but sensitive to diges- 
tion with pancreatic DNase. The viral- 
sized DNA was sensitive to nuclease SI 
digestion (Fig. 20A and C) and resistant 
to digestion by exonuclease-VII (not 
shown) consistent with its being circular 

SqLCV has genomic homology with 
bean golden mosaic virus (BGMV), an- 
other whitefly-transmitted geminivirus 

(Haber and Goodman, personal commu- 
nication). This has allowed us to confirm 
that the small ssDNA component in in- 
fected squash tissue is SqLCV DNA (Fig. 
20B). Southern blot analysis of extracts 
from infected squash demonstrates ho- 
mology of this ssDNA with cloned BGMV 
DNA under conditions of reduced strin- 
gency (1 M salt, 55°C). Furthermore, at 
this level of sensitivity, we confirm that 
all of the SqLCV DNA is Sl-sensitive 
(Fig. 20B and C). We do not detect any 
other forms of SqLCV DNA. This is in 
contrast to findings with other whitefly- 
transmitted geminiviruses where large 
amounts of double-stranded forms of the 
viral DNA are found in infected tissue 
(Ikegami et al., Proc. Nat. Acad. Sci. 
USA 78, 4102, 1981; Hamilton et al, Nucl. 
Acids Res. 10, 4901, 1982). 



Electron microscopic examination of 
gel-purified material confirmed that the 
SqLCV DNA in infected tissue was in- 
deed circular ssDNA (Fig. 21). The 
SqLCV DNA consisted of predominantly 
single-stranded circles of apparently two 
size classes (Fig. 20A). Double-stranded 
supercoiled forms were found at a fre- 
quency of <1%, and a few broken single- 
stranded linear molecules were also seen. 
Measurements relative to phage M13mpl0 
DNA gave size estimates for the two 
populations of SqLCV DNA circles of 
-3.0 kb and ^2.5 kb (Fig. 20B). 

Gel-purified SqLCV single-stranded 
DNA was converted to a double-stranded 
form using random primers prepared from 
salmon sperm DNA (Taylor et al., 
Biochim. Biophys. Acta U2, 324, 1976) 
and the large fragment of E. coli DNA 
polymerase I (Klenow and Hennigsen, 
Proc. Nat Acad. Sci. USA 65, 168, 1970). 
Restriction of the converted double- 
stranded SqLCV DNA with Hpa II and 
Sal I allows us to estimate its genomic 

complexity (Table 1). The fragments found 
in restriction digests add up to 5.5 kb and 
are consistent with SqLCV having cir- 
cular DNA components of —2.25 kb and 
-3.25 kb. This makes SqLCV different 
from the other whitefly-transmitted 
geminiviruses so far characterized, all of 
which have two components of 2.8 kb each. 
Using this double-stranded DNA, we have 
identified a few convenient restriction 
enzyme sites and are currently in the 
process of cloning the SqLCV genome. 

TABLE 1. Restriction Analysis of SqLCV 

Hpa II s 

Sal r 









Total 5510 


*Fragment sizes in bp generated upon digestion 
of in vitro synthesized SqLCV dsDNA with the 
restriction enzymes indicated. 

Fig. 21. Electron micrographs of gel-purified SqLCV DNA. (A) Typical field showing two populations 
of circular ssDNAs. (B) Spread showing size relative to phage M13mpl0 RF, which is located on the 
right in the micrograph. 



a SR 

t) <D 

•r-j +J 

^C o3 


o Jy 

^ c 

CO -•-» 


« X 

r* a 

Ph o 

'—. 'qj 

o > 

- o> 

CQ T3 

. >s 

-<-> T? 

s £ 

03 Q) 

a = 

s £ 

2 s o 

v JZ 

-^ cO 

^3 co 

5 £ 

a o 

<V Sh 

«4-l 5h 

C as 

£ to 

13 03 


S3 £ 


03 aS 


r- "^ 




O 'O 

, i 

£ « 


«4-H ^ 


OS J* 


<D f -1 


^5 bJD 


>s 0) 


b +s 


^ a 




'C 'S 




^ CO 

T3 ^ 


«4— 1 CJ 


"" CO 

CO -r-i 


133 fli 





> <v 

o ^ 


+J 03 

+-> & 


a bfi 


J5 a) 


CL Sh 


r-T ^ 


P ^ 



+j co 


C -^> 

o c 


O iS 


<~V Cl< 





e pa -S 


. CO w 

cu a CO 

<££ c 




CM g bJD 

. S S3 

fe 3 o 

co co 

■r o 


Biological Studies of Beet Curly petiole into the cut petiole on the recip- 

Top Virus ient plant and supported in place. The 

graft union was complete in a few days 
The investigation of CTV is of partic- with 100% of the grafts taking. Symp- 
ular interest because of its exceptionally toms of epinasty , distension of the upper 
wide host range and the emergence of surface, and dwarfing appeared in the 
strains of increased severity, apparently developing trifoliates and were well de- 
in response to an aggressive breeding veloped by —10-14 days after grafting, 
program aimed at its control in sugar- The affected leaves were indistinguish- 
beets in the western United States (see able in their appearance from those on 
Bennett, Monograph 7, American Patho- bean plants inoculated via viruliferous 
logical Society, 1971). As appears true leafhoppers. The appearance of symp- 
for other leafhopper-transmitted gemi- toms was very similar in time scale to 
niviruses, CTV appears to replicate to those observed following inoculation with 
lower titers in infected plants than do viruliferous leafhoppers. This is not sur- 
whitefly-transmitted geminiviruses, and prising since geminivirus infection ap- 
has thus proven to be correspondingly pears to follow the direction of flow of 
more difficult to work with. Our efforts nutrients in the phloem with the earliest 
to date with CTV have concentrated on most severe symptoms developing at the 
establishing a reliable system for trans- growing points in the newly emerging 
mitting and maintaining the virus in the leaves. In the pinched and grafted plants 
absence of its leafhopper vector. We have we noticed that symptoms appeared in 
used grafting techniques to establish col- the growing point developing on the same 
onies of CTV-infected beans (P. vul- side of the plant as the grafted leaf, 
garis, Bountiful) as a reliable source of whereas the other growing point seemed 
experimental material. Our choice of to develop normal-appearing, symptom- 
source plant was based on considerations less leaves. We therefore pinched the 
of rapid plant growth, ease of manipu- growing points a second time 4-6 days 
lation for grafting, and yields of CTV. after grafting. These "twice-pinched" 
The recipient plants were 9- to 10-day- plants subsequently developed multiple 
old bean seedlings from which the grow- growing points all of which gave rise to 
ing point was pinched 1-2 days prior to emerging leaves showing severe symp- 
grafting. In this stage, the primary leaves toms of CTV infection. This procedure 
had just emerged. One of the primaries thus increased our yield per plant of in- 
was cut from the plant, and a trifoliate fected leaves. Examples of infected bean 
leaf from an infected plant showing clear plants showing symptoms of CTV are 
symptoms of CTV was inserted via its shown in Fig. 22. 




N. Fedoroff, M. Schwartz, M. Shure, and S. Wessler 

Maize controlling elements are trans- Dissociation family studied by Mc- 

posable elements that cause unstable in- Clintock (Brookhaven Symp. Biol. 18, 

sertion mutations, chromosome breakage, 162, 1965). A controlling element family 

and chromosomal rearrangements. Sev- comprises a group of related elements 

eral families of controlling elements have that differ in their capacity for autono- 

been identified, including the Activator- mous transposition as well as in the fre- 

N. Fedoroff 


quency and developmental timing of the locus not only is interrupted by an 
transposition. The interests of our lab- insertion but also is partially duplicated, 
oratory center on the isolation and char- The structure of the disrupted locus is 
acterization of controlling elements, and illustrated diagrammatically in Fig. 23a. 
on the structure and expression of genes The transcription unit contains what ap- 
with controlling element mutations. Dur- pears to be an insertion of approximately 
ing the past year, we have made prog- 30 kb that has inverted repetitions of 7- 
ress in understanding the structure of a 8 kb at its ends. The sequences at each 
complex mutation at the Shrunken locus end of the insertion have a similar re- 
caused by the Dissociation (Ds) element, striction site map, except that a sequence 
We have also isolated the Activator and of approximately 0.5 kb present adjacent 
two different Ds elements inserted at the to the 3' portion of the disrupted locus 
Waxy (Wx) locus and carried out prelim- is missing from the sequence adjacent to 
inary studies on the expression of the Wx the 5' portion of the disrupted locus. In 
locus with Ac and Ds insertion muta- addition, the Sh sequence 5' to the break- 
tions. point and part of the insertion are du- 
plicated nearby on the same chromosome. 
Although the terminal 7-8 kb of the 
Structure of the sh-m593S Allele insertion detected at the Sh locus in the 
of the Shrunken Locus sh-m5933 allele are repeated, it is likely 

that neither the entire insertion nor even 
the entire length of the terminal repe- 
Studies carried out during the past year tition constitutes the Dissociation (Ds) 
on the sh-m5933 allele of the Shrunken element causing the mutation. Recent 
(Sh) locus have substantially clarified the experiments in P. Starlinger's labora- 
nature of the mutation and the structure tory indicate that the 3' breakpoint is 
of the rearranged locus (see Fedoroff et flanked by a complex derivative of the 2- 
al. f Year Book 81). These studies were kb Ds element whose isolation from the 
carried out in collaboration with the lab- Waxy (Wx) locus is described in a sub- 
oratory of Peter Starlinger at the Uni- sequent section. The 4-kb sequence 
versity of Cologne. Genomic clones of the flanking the breakpoint constitutes one 
Sh locus had been isolated in P. Starlin- 2-kb Ds element inserted in inverted or- 
ger's laboratory, and subcloned se- der into the middle of another, identical 
quences derived from the clones were used Ds element. Since the foreign sequence 
to analyze the structure of both the sh- adjacent to the 5' side of the rearranged 
m5933 allele and nine Sh revertants of locus resembles that adjacent to the 3' 
the allele isolated and analyzed in our side of the locus and there is, in addition, 
laboratory over the past few years. The a duplication of the 5' fragment, it fol- 
properties of the revertants were similar lows that there are several Ds elements 
to those of revertants isolated many years involved in this mutation, 
ago by McClintock {Year Book 52). The Reversion of the sh-m5933 allele re- 
revertants have an Sh phenotype but suits in restoration of the original gene 
continue to show Ds-mediated chromo- sequence, although the duplication is re- 
some breakage at the Sh locus, suggest- tained (Fig. 23b). The persistence of the 
ing that reversion of the Ds mutation is duplication and its Ds elements in the 
not accompanied by loss of the Ds ele- revertants accounts for the persistence 
ment from the locus. of Z>s-mediated chromosome breakage at 
Detailed restriction endonuclease the Sh locus in revertants. Direct evi- 
analysis of sh-m5933 genomic DNA, us- dence in support of this inference comes 
ing hybridization probes in the immedi- from the observation that one of the nine 
ate vicinity of the known rearrangement revertants analyzed showed a geneti- 
breakpoint in the strain, revealed that cally stable change in the developmental 



10 kb 






Sh revertants 

I reversion 

5' 3' 


deletion in Sh-r5 

Fig. 23. A diagrammatic representation of the structure of the Sh locus in the sh-m5933 strain and 
revertants derived from it. Solid boxes represent the sucrose synthetase-coding sequence at the Sh 
locus; open boxes represent the foreign DNA introduced by the rearrangement in the sh-m5933 allele. 
As discussed in detail in the text, the Sh locus in the sh-m5933 allele (a) is represented as having an 
insertion of approximately 30 kb within the transcription unit, as well as a partial duplication of unknown 
length which includes the 5' junction fragment between the Sh locus sequences and the foreign sequence 
introduced by the rearrangement. Both copies of the partially duplicated locus are on the same chro- 
mosome, but the distance between them and their relative orientations are not known, as represented 
by the break in the chromosome in the diagram. The structure of the Sh revertants derived from the 
sh-m5933 strain is depicted in (b) and comprises a normal or nearly normal mRNA coding sequence, as 
well as a copy of the duplicate 5' junction fragment present in the sh-m5933 strain on the same chro- 
mosome. The duplicate 5' junction fragment in the sh-m-5933 strain is indistinguishable from that present 
in a majority of the revertant strains. The Sh-r5 strain, however, has a 5' junction fragment that contains 
a deletion of 2-2.5 kb at the junction between Sh locus sequence and the foreign DNA introduced by 
the rearrangement. The arrows within the insertion represent the terminal sequences of the insertion 
having similar restriction endonuclease maps. The arrow adjacent to the 5' side of the rearrangement 
breakpoint is truncated to indicate that sequences present adjacent to the 3' breakpoint are missing at 
the 5' breakpoint. 

timing of Ds-mediated chromosome 
breakage to very late in kernel devel- 
opment. The revertant is designated Sh- 
r5. Analysis of Sh-r5 DNA revealed the 
existence of a 2-kb deletion, probably 
corresponding to a complete 2-kb Ds ele- 
ment, at the junction between the du- 

We described (Year Book 81, 167-171) 
the construction and identification of a 
cDNA clone complementary to Wx mRNA 
sequences. We have subsequently re- 
covered two more chimeric plasmids, 
pcWx0.4 and pcWx0.25, having inserted 
cDNA fragments of 0.4 kb and 0.25 kb, 

plicated portion of the Sh locus and the respectively. To determine whether the 
duplicated portion of the insertion (Fig. cDNA plasmids homologous to Wx RNA 
23b). were also homologous to the genetically 

identified Wx locus, we used the cDNA 
insert from pcWx0.35 as a hybridization 
probe to analyze DNA from maize strains 
with genetically well characterized al- 
leles of the Wx locus. We chose the wx- 
?n6 allele for this purpose, because both 
Our interest in the Waxy locus stems the progenitor Wx allele from which the 
from the fact that there are many con- Ds-induced wx-m6 allele was derived and 
trolling element-induced mutable alleles several newly isolated stable Wx deriv- 
of this locus. Therefore, with the goal of atives were available. DNAs from plants 
isolating controlling element DNA se- homozygous for the progenitor Wx allele, 
quences, we have continued our efforts the mutable wx-m6 allele, and two ger- 
to isolate the Wx locus. minally stable Wx revertants (wx-m6rl 

The pcWx cDNA Plasmids Are 

Homologous to the Genetically 

Identified Waxy Locus 

M. Share and S. Wessler 


and wx-m6r2) were cleaved with various the wx-m6 mutation is accompanied by a 

restriction endonucleases, fractionated restoration in size of the homologous re- 

by agarose gel electrophoresis and trans- striction fragment to that of the frag- 

ferred to nitrocellulose filters for hybrid- ment in DNA from the Wx progenitor 

ization with the labeled cDNA insert. The allele, we conclude that we have cor- 

results of such an experiment are shown rectly identified the Wx gene product, 

in Fig. 24. The cDNA insert in pcWx0.35 and that the cloned Wx cDNA is indeed 

is homologous to a unique genomic DNA homologous to Wx locus DNA sequences, 

fragment in each of the digests. In four We further conclude that the Ds muta- 

of the five digests, the homologous frag- tion is a simple insertion of approxi- 

ment in DNA from wx-m6 plants is larger mately 2.4 kb of DNA, and that 

by 2.4 kb than the fragments in DNAs phenotypic reversion of this mutation is 

from plants homozygous for the progen- accompanied by excision of the inserted 

itor Wx allele or either of the two Wx sequences, 
revertants. Thus, the mutation at the 

Waxy locus in the wx-m6 allele results Cloning of the Wx Locus and the 

in an increase in the sizes of restriction Ac and Ds Elements 

endonuclease fragments homologous to n F H ff 
the cloned Wx cDNA. The wx-m6 allele 

has been characterized as a Z)s-induced Restriction endonuclease fragments 

mutation (McClintock, Year Book 51, 212, with homology to a Wx cDNA clone were 

1952). Because Ac-mediated reversion of isolated from recombinant X phage li- 

+ &r + ike j» wx-m6r 1 

C-l Sh Bz WxDs 9 :; 4AC ,. wx~m6 * AC , < ^ 

"""""* wx-m6r2 

EcoRI Bglfl XmnI EcoRV Pvu 

kb kb kb kb kb 

if %Z m • «. ?3-4- *m 12.5- W 

u, f -p. « «, h.O-^ mm IO.O- # * * 

5.5- m m m m 


*. "f to { ~ f <N 


10 1 — \ CM 

x i (0 1 — ' OJ 

*! <J» 1 — \<M 

3 s f ■ *5 k 

x £ £ 


\ | « 

* f ^ & 

S e *- *~ 

7 U) U3 

A x 

S i £ 

x L 

— X 

x £ c 

x £ e 

x X 

5* !•■: w <>~ 

^ e e 

3S 5 3 3 

Fig. 24. Autoradiograph of filters, following hybridization with the 32 P-labeled cDNA insert from 
pcWx0.35, containing restriction endonuclease-digested DNAs from plants homozygous for the mutable 
wx-m6 allele, the progenitor Wx allele, and two stable germinal Wx revertants, wx-m6rl and wx-m6r2. 
The diagram at the top of the figure outlines the relationship among these alleles. The restriction 
endonucleases used, as well as the sizes (in kb) of the fragments homologous to the cDNA insert, are 
indicated in the figure. 



braries of genomic DNA isolated from 
plants carrying several different alleles 
of the Wx locus. These included the wx- 
m6 allele (a Ds insertion mutation), its 
Wx progenitor, the Ac insertion muta- 
tion Ac wx-m9, wx-m9 (a Ds derivative 
of the Ac wx-m9 allele), and Wx9-rl, a 
newly derived Wx revertant of the Ac 
wx-m9 allele (McClintock, Year Books 51 
and 62). The first fragment isolated was 
an Eco RI fragment of DNA from a plant 
that was homozygous for the Wx pro- 
genitor allele of the wx-m6 mutant allele. 
A restriction map of the fragment ap- 
pears in Fig. 25a. The direction of tran- 
scription and the approximate extent of 
the transcription unit were investigated 
by M. Schwartz (see below) and are in- 
dicated by the arrows. 

Bgl II fragments were cloned from 
genomic DNAs isolated from plants car- 
rying the Ac wx-m9, wx-m9, wx-m6, and 
Wx9-rl alleles. The structure of the Bgl 
II fragments is shown in Fig. 25. The Ac 
wx-m9 and wx-m9 alleles yielded frag- 
ments that differed from those in their 
Wx counterparts by the presence of in- 
sertions of 4.3 and 4.1 kb, respectively, 
near the 3' end of the transcription unit. 
The absence of the insertion in the Wx9- 
rl Bgl II fragment isolated from a Wx 
revertant of the Ac wx-m9 allele iden- 
tifies the insertion in the Ac wx-m9 strain 
as the Ac element. Restriction endonu- 
clease mapping of the original Bgl II 
fragments isolated from Ac wx-m9 and 
wx-m9 DNA indicated that both frag- 
ments have an insertion at precisely the 




N W- 1 

>— ! l-H 

• — • i — i 




CT* CT> in 

1/1 OT 

V) <A 


i — 

— 1 

ao ma- 















« »H 

►-H M 



o> </> 

</> </) 

(O V) 




o> a> 


a. a. 

a. a. 




m m 




1 1 





■ i 

1— 1 


H l-H 

•— i >— i 


>— i 




CJ> m 

m m 

in m 





a. a. 





















l-H »~1 




i— i 





CT> to 

to «/> 

in <" 







m a. 

a. a. 

























1— 1 



l-H M 

.-I l-H 





+— •»- 

,|_ +m 




O 1 10 

(O </) 

in in 





a. a. 

o. a. 












Fig. 25. The structure of the Eco RI and Bgl II fragments of the Wx locus cloned from maize strains 
carrying various alleles of the Wx locus. The relationship between the Eco RI Wx fragment and the Bgl 
II fragments cloned from strains carrying the Ac wx-m9, wx-m9, wx-m6, and Wx9-rl alleles is indicated 
by the vertical alignment of the restriction endonuclease cleavage site maps. The Bgl II fragments contain 
only the 3' end of the transcription unit, but they include the site of insertion of the Ac and Ds elements 
in the mutant strains. The locations and relative sizes of the Ac and Ds insertions, whose structure is 
described in the text, are indicated by the triangles below the fragments. 


same site. The insertions were subcloned the Ac9 element that generated the Ds9 

in plasmids and analyzed by heterodu- element affected expression of a trans- 

plex analysis and restriction mapping. The acting function necessary for transposi- 

insertions are nearly identical, forming tion. 

an almost perfect heteroduplex. The re- The Bgl II fragment with homology to 
striction maps are extremely similar, dif- the Wx cDNA plasmid isolated from wx- 
fering by the absence of a sequence of m6 genomic DNA also differs from its 
less than 0.2 kb from the center of the Wx progenitor allele by the presence of 
wx-7u9 insertion present in the Ac wx- an insertion near the 3' end of the tran- 
m9 insertion (Fig. 26 a and b). The in- scription unit (Fig. 25e). The insertion 
sertion in the wx-m9 Bgl II fragment has site is approximately 0.5 kb 5' to the in- 
therefore been identified as the Ds ele- sertion site in the Ac wx-m9 and wx-m9 
ment whose presence at the locus can be strains, and the insertion is 2.0 kb in 
detected genetically. The elements are length. It has been established by net- 
designated Ac9 and Ds9 to indicate their eroduplex analysis and restriction en- 
isolation from the Ac wx-m9 and wx-m9 donuclease mapping that the insertion in 
alleles, respectively. The structural sim- the wx-m6 allele, designated Ds6, is ho- 
ilarity between the Ac9 and Ds9 ele- mologous to a 1-kb sequence at each end 
ments suggests that the Ds9 element of the Ac9 element (Fig. 26c). 
arose by an internal deletion in the Ac9 To determine the genomic represen- 
element after its insertion at the Wx lo- tation of sequences homologous to the Ac 
cus to give the Ac wx-m9 allele. Because and Ds elements, genomic DNAs from 
Ds elements can transpose in the pres- several different maize strains were di- 
ence, but not in the absence, of an Ac gested with BstE II, an enzyme that does 
element, it is likely that the deletion in not cleave within any of the isolated ele- 

D.C-C o o d c c 
> o. a. > -c n _•- 
CLcocn <x XXI 

^ ii mi Y ii i i i i i l I i_ y — ' Ac 9 

5' 3' 


x —• 

<f> W CD X Q. CD CD \, 0- 

_J I , i I i I L-Kj — ' Ds9 

> Q. Q. >- c -O— ■- 

Q-irxD <X XXX 


Q- XX/ / toX Q.COCA) £ 00<Q-< cnCDCL 

c. ' ^ ' r ( s U UJ 1 LlJ 1^ Ds6 

1.0 kb 

Fig. 26. Restriction endonuclease cleavage site maps of Wx Pst I fragments containing the Ac9, Ds9, 
and Ds6 controlling element insertions. The open boxes represent the Wx locus sequence, bounded by 
Pst I sites, within which the insertion is located. The solid central lines represent the controlling element 
sequences. The orientation of the Wx locus fragment into which the elements have inserted is indicated 
relative to the direction of transcription (Fig. 25a). The Ac9 and Ds9 elements are indistinguishable by 
restriction endonuclease mapping except for the absence of a sequence of less than 0.2 kb (represented 
by the interruption in diagram b) from the Ds9 element that is present in the Ac9 element. The Ds6 
element is inserted in the opposite orientation to the Ac9 and Ds9 elements at a different site in the 
transcription unit. The restriction endonuclease cleavage site map of the element resembles the maps 
of the ends of the Ac9 and Ds9 elements. 


ments, and were probed with either the ies of Ac-like sequences. There are at 

Ds6 element or a central Ava I-Eco RI least twenty discrete bands, some pres- 

fragment from the Ac9 element. The re- ent in multiple copies, that show homol- 

sults are displayed in Fig. 27. Because ogy to Ds6 — itself homologous to the Ac 

equal amounts of cloned DNA fragments termini. The discrete bands appear on a 

were included in each experiment and continuum of hybridizing sequences, 

hybridized to the same extent to the two suggesting the presence of many more 

probes, the hybridization intensities in sequences with limited homology to the 

the two panels are directly comparable, probe. By contrast, the various strains 

It is evident that there are many copies examined had between four and ten cop- 

of Z)s-like sequences in all of the maize ies of sequences homologous to the left 

DNAs examined, but relatively few cop- Ava I-Eco RI fragment from the center 

Genomic DNA 

$ <5 « «» S c - - 

"iC /T\ fc > > 

x p * 2 • 

to f ■$ 2. ° 

X CM X u 

* 3ft '* < X X D. ^ ^ * <t x x 



















Ix 5x Ix 5x 

C!oned DNA s Ac wx-m9 Ac wx-m9 

Probe; pAc9 AvoI/EcoRI pD^ Aval 

fragment fragment 

Fig. 27. Blot hybridization analysis of maize DNAs digested with the restriction endonuclease BstE 
II. Maize DNAs from plants having (Ac wx-m9) and lacking (Wx, W2S x K55, wx-m9) an Ac element 
were digested with BstE II, fractionated on a 0.7% agarose gel, and transferred to nitrocellulose. Mixtures 
of X. laevis DNA, which shows no homology to the probes used, and an amount of \EMBL4 phage DNA 
containing the Ac wx-m9 Bgl II fragment corresponding to one or five genomic copies were included as 
an internal hybridization control in both experiments. Filters were probed with 32 P-labeled restriction 
endonuclease fragments, (a) The Ac probe used was the left Ava I-Eco RI fragment from the center of 
the Ac9 element (Fig. 26a). (b) The Ds probe used was an Ava I fragment from the pDs6 plasmid that 
contains most of the Ds element (Fig. 26c). 

M. Shure 


of Ac9 (Fig. 26a). Thus, the central re- clone lies beyond a Bam HI site 2.0 kb 

gion of the Ac element is very much less from the Eco RI site at the left end of 

abundantly represented in the genome the Wx genomic Eco RI fragment and 

than are the termini, suggesting that de- that the 3' end lies beyond a Pst I site 

fective Ds-like elements are rather abun- 2.1 kb from the right end of the Eco RI 

dant. fragment (Fig. 25a). 

The internal structure of the Ac-like 
sequences in maize DNA has also been 
investigated. The results of the experi- 
ments indicate that the Ac-like se- Preliminary Analysis of the 
quences have similar central regions but RNAs Transcribed in Endosperms 
differ from each other at the ends. It is Carrying the Ac wx-m9 and wx-m9 
not yet clear whether the differences are Alleles 
attributable to modification of the gen- 
omic sequences or to differences in struc- 
ture among the various copies of the Ac- It was of interest to determine how the 
like sequences. However, the single ac- inserted elements Ac and Ds, in the re- 
tive Ac element at the Wx locus can be spective alleles Ac wx-m9 and wx-m9, 
distinguished in genomic blot hybridiza- affect the expression of the Waxy locus, 
tion experiments from other related se- In the absence of the autonomously 
quences, raising the possibility that it is transposing element Ac, wx-m9 endo- 
structurally unique. sperms exhibit a phenotype that, al- 
though uniform and recessive to Wx, is 
actually intermediate between wx and Wx 
Direction of Transcription of the with respect to the level of amylose pres- 
Wx Locus and Approximate Extent ent. In the presence of an active Ac ele- 
of the Transcription Unit ment, wx-m9 endosperms are somatic 

, , , mosaics comprising clonal sectors of Wx 

M. Schwartz ,. i i t n - ± v \l 

tissue on a background of intermediate 

The direction of transcription of the phenotype. 
Wx locus was determined by synthesiz- Poly A + RNA was prepared from im- 
ing a labeled hybridization probe em- mature endosperms of the following gen- 
ploying T4 polymerase. The Eco RI oty pes: Ac wx-m9, wx-m9 + Ac, wx-m9 
genomic clone (Fig. 25a) was cleaved at no Ac, and Wx. The RNA populations 
the unique Eco RV site and labeled with were electrophoretically fractionated 
T4 polymerase. The labeled DNA was through denaturing agarose gels, after 
further digested with Sal I, Bam HI, Xho which the RNA was transferred to ni- 
I, or Eco RI. Following agarose gel elec- trocellulose filters. When Wx DNA is used 
trophoresis, fragments of the following as a hybridization probe, it is apparent 
size (labeled at the Eco RV site) were that in RNA from all three mutant con- 
recovered: 1.1-kb Eco RV/Sal I, 2.3-kb figurations, there is an RNA species sim- 
Eco RV/Bam HI, 3.0-kb Eco RV/Xho I, ilar in abundance and length to the 2.3- 
and 6.6-kb Eco RV/Eco RI. All of these kb Wx RNA (Fig. 28A, I). Despite this, 
fragments were found to hybridize to the much of the RNA must be aberrant in 
mR N A of the Wx locus by the Northern sequence, because when these mutant 
blotting technique. Since these frag- RNA populations are used to direct pro- 
ments were labeled only at the Eco RV tein synthesis in vitro, only low levels of 
site, it follows that the direction of tran- protein are immunoprecipitated using 
scription is from left to right as indicated antiserum directed against the 58-kD Wx 
in Fig. 25a. Similar techniques were em- protein (Fig. 28B). It appears that there 
ployed to show that the 5' end of the are at least two cross-reacting proteins 
transcription unit of the Eco RI genomic synthesized when RNA from these mu- 


a) b) 

i. n. 

12 3 4 

12 3 4 5 6 7 8 

Fig. 28. (A) Autoradiograph of filters containing fractionated RNA samples. Poly A + RNA was 
fractionated by electrophoresis through a formaldehyde-containing 1% agarose gel. The fractionated 
RNA was transferred to nitrocellulose filters, following which the filters were hybridized with either (I) 
the 10.8-kb genomic Eco RI fragment containing Wx sequences or (II) the cloned Ac element, both of 
which had been labeled with 32 P by nick-translation. RNA samples were purified from endosperms of 
the following genotypes: (1 and 5) Wx; (2 and 6) Ac wx-m9; (3 and 7) wx-m9, no Ac; (4 and 8) wx-m9, 
+ Ac. Arrow indicates position of the 2.3-kb Wx mRNA. 

(B) Fluorograph of an SDS-polyacrylamide gel on which immunoprecipitated 35 S-methionine-labeled 
in vitro translation products were fractionated. Synthesis was programmed by total poly A + RNA from 
immature endosperms, and the translation products were precipitated using antiserum directed against 
the 58-kD Wx protein purified from starch granules. Synthesis was programmed by RNA from endo- 
sperms of the following genotypes: (1) Wx; (2) Ac wx-m9; (3) wx-m9, no Ac; (4) wx-m9, + Ac. Arrow 
indicates position of the 65-kD Wx in vitro precursor polypeptide (Year Book 81). 

tants is used; at least one of these poly- be seen, some of which exhibit the same 

peptides differs in mobility from the wild- relative difference in length observed for 

type 65-kD Wx preprotein. the three large transcripts. These ob- 

In all three mutant RNA populations, servations correlate with the finding that 

up to three additional RNA species are the Ds element at wx-m9 is a slightly 

easily detected when Wx DNA is used deleted version of the Ac element at Ac 

as a hybridization probe (Fig. 28A, I). wx-m9, thus suggesting that the homol- 

All three species are much larger than ogous RNAs span at least that portion 

Wx RNA, and the three from wx-m9 en- of each element corresponding to this le- 

dosperms are each slightly shorter than sion. It is not known at present from which 

the corresponding RNAs from Ac wx-m9 strand these large RNAs are tran- 

endosperms. In each case, the same three scribed, or whether transcription is ini- 

large RNA species are detected even more tiated in Wx DNA sequences or in 

easily when a similar filter is hybridized controlling element DNA sequences, 

with labeled Ac DNA sequences (Fig. However, since the intermediate RNA 

28A, II). In addition, in each sample sev- species are detected more easily with the 

eral other intermediate RNA species can Ac DNA probe than with the Wx DNA 


probe, it is not clear that the large RNA quentially removed by splicing to pro- 
species are precursors from which the duce the abundant RNA species of wild- 
controlling element sequences are se- type mobility and abundance. 



M. Collins, T. Hazelrigg, R. Karess, R. Levis, C. Murphy, K. O'Hare, H. Roiha, and G. M. Rubin 

The overall aim of our laboratory is to all the sites of insertion of transposable 
determine certain aspects of the struc- elements causing or associated with 
ture and function of transposable genetic known white mutations. We have deter- 
elements in Drosophila ?nelanog aster. In mined the changes in DNA sequences as- 
previous Year Books, we have described sociated with the mutations listed in Table 
the isolation of the DNA sequences com- 2. 
prising the white locus. This locus con- 
trols both the level and pattern of Transcription of the white Locus 

K. O'Hare, R. Levis, and G. M. Rubin 

pigmentation in the eye. Because white 
is a nonessential gene and has an easily 
scored phenotype, it provides an attrac- We have characterized transcription of 
tive experimental system for studying the white locus in order to further our 
both gene regulation and the effects of understanding of its expression, how it 
transposable element insertions on gene is normally regulated, and how this 
expression. Our initial approaches to this expression is altered by mutations. On 
problem, described in Year Books 80 and the basis of the different phenotypic 
81 have continued during the past year, properties of mutations mapping in the 
We have recently initiated a study of white white locus (Green, Heredity 13, 303, 
locus regulation using a combination of 1959), it has been postulated that the lo- 
in vitro mutagenesis and P-element-me- cus is composed of a centromere-distal 
diated gene transfer methods. These same structural domain and a proximal regu- 
techniques have also been applied to the iatory domain. We and others have iden- 
study of the P element itself. tified DNA insertions or deletions 

associated with many of these mutations, 

thereby positioning these mutations on 

DNA Sequence Studies of the the physical map of the DNA (see Fig. 

white Locus 29 and Table 2). There is a gap of about 

is ^,tt ^ n* i jn n* r> u- 3 kb between the sites of the w a iwhite- 

K. O Hare, C. Murphy, ana G. M. Rubin . x , h , , ., , s 

apricot) and w n (white- honey) mutations, 
These studies are aimed at furthering where no mutations with phenotypic ef- 
our understanding of the white locus, both fects are known. Our analysis of tran- 
as a genetically well characterized eu- scription has more clearly defined the 
karyotic gene and as a model system for relationship between the transcript of 
the mechanisms whereby transposable white and the distal and proximal do- 
elements affect gene expression. We have mains, as well as the mutationally silent 
determined the DNA sequence of the wild- gap between these domains, 
type locus from the strain Canton S be- We have identified a single transcript 
tween coordinates -3.2 and +11.0 (Fig. of about 2.7 kb by hybridizing blots of 
29). This interval includes the sequences gel-fractionated poly(A)-containing RNA 
found to be transcribed ( - 2.2 to + 3.7) of adults or pupae with 32 P-labeled, strand- 
to produce a 2.7-kb RNA (see below) and specific probes derived from cloned gen- 



P copia BI04 P copia 
w* 6 w hd w bf w* 12 w a 








03 CQ 


B ~ 








□ □ 

+ 4 

■ □ I 

□ □ 


+ 5 

+ 6 

+ 7 


+ 10 



Fig. 29. Relationship of white locus transcript and transposable element insertions to the physical 
map. The positions of a number of insertions causing or associated with white mutations are shown on 
the restriction map of the locus. The coordinate system is that of Levis et al. (Proc. Nat. Acad. Sci. 
USA, 79, 466, 1982), where one unit corresponds to 1 kb. The DNA insertions were cloned as parts of 
previous studies in this laboratory (iv a ; w* 6 ; w M,s "' 11 , abbreviated as w M in the figure, and w* u ; w l+A ; 
w DZL ; see previous Year Books) and other laboratories (w bf , w h , and uf p , see Zachar and Bingham, Cell 
30, 529, 1982). Insertions were classified by their restriction enzyme maps and terminal DNA sequences. 
The positions of insertion have all been determined to the nucleotide by DNA sequencing. The phenotypes 
(but not the size of the insertions) are shown by the triangles: a solid triangle indicates that flies with 
this insertion have wild-type colored eyes; a clear triangle indicates bleached-white eyes; a shaded triangle 
indicates eyes with some degree of coloration. The allele and family of the inserted element are shown 
above the positions of insertion. The bars show the extent of sequences in different probes used in RNA 
blotting experiments. Probes that hybridized to the 2.7-kb RNA are shown as solid bars; open bars show 
those probes that did not hybridize. The probes were all M13 subclones of the genomic wild-type locus 
and would hybridize with transcripts whose direction of transcription was proximal to distal. 

omic DNA of the white locus. This tran- 
script is present at low abundance, making 
up only about 0.0005% of the poly(A)- 
containing RNA. The genomic sequences 
represented in the mature transcript have 
been mapped by determining which 
probes hybridize to this transcript band. 
The transcript contains a small 5' exon 
of about 200-500 bp, separated from the 
main body of the transcript by an intron 
of about 2.8 kb. The 5' end of the tran- 
script maps within the same 1.3-kb Hind 

III-Bam HI interval as do a group of 
mutations of the proximal domain. The 
w sp (white-spotted) mutation and three 
other alleles, which all result in a mottled 
pattern of eye pigmentation, are asso- 
ciated with insertions into or deletions of 
sequences at least 1 kb proximal to the 
5' end of the transcript. The distal region 
mutations of white map within the large 
3' transcribed region. The major intron 
is devoid of mutations with phenotypic 
effects. Since many transposable ele- 

TABLE 2. white Locus Mutations 


Associated Structural Change in Locus 





W bf 
W* 12 
W a 
W i + A 

w h 

W sp 

w sp4 

1.1-kb P element inserted at -2.1 
0.5-kb P element inserted at -2.1 
0.6-kb P element inserted at -2.1 
5.0-kb copia element inserted at - 1.4 
9.2-kb B10J> element inserted at - 1.2 
1.6-kb P element inserted at -0.6 
5.0-kb copia element at 0.0 
3.4-kb F element inserted at +0.8 
4.6-kb F element inserted at +3.7 
9.2-kb BlOJf element inserted at +4.9 
13-kb FB element inserted at + 9.8 
1.1-kb deletion from +4.9 to +6.0 


ment insertions with easily recognizable formants with proper eye pigmentation, 

phenotypic effects are known which map dosage compensation, and interaction with 

on either side of this intron, it is likely zeste. Moreover, we have expanded our 

that most insertions into the major intron knowledge of the influence of chromo- 

of white do not alter its phenotypic somal position on the regulation of white, 

expression. We are extending our anal- by comparing the phenotypes of trans- 

ysis of white transcription to mutants formants having the same cloned white 

carrying various transposable element DNA segment present at an array of dif- 

insertions in the white locus. Preliminary ferent chromosomal locations. We have 

results indicate that some insertions also taken an important step in our in 

eliminate the accumulation of the normal vitro mutagenesis analysis by showing 

transcript and cause transcripts of larger that a specific truncation of the white gene 

or smaller size to appear. eliminates its ability to confer any pig- 

We have also investigated the effect mentation on the eyes. 
on the white transcript of the zeste mu- We have introduced the three DNA 
tation, which represses white expression segments of white, labeled A, B, and C 
as judged by eye color. Genetic evidence in Fig. 30, into the germ lines of em- 
has suggested that the wild-type zeste bryos. These segments have the same 3' 
product or products regulate white at a boundary at coordinate —5.1, and have 
transcriptional level (Jack and Judd, Proc. 5' boundaries at coordinates +9, +6.7, 
Nat. Acad. Sci. USA 76, 1368, 1979). We and +3.3, respectively. Segments B and 
have found that the white transcript of C were inserted into the nonautonomous 
adults with zeste-colored eyes was un- P-element vector pVIO (see Spradling and 
changed in size or abundance in compar- Rubin, this Report) in tandem with the 
ison to that of wild-type flies. This rosy gene. These plasmids were microin- 
indicates that if the zeste mutation alters jected into embryos of a rosy mutant 
the transcription of vjhite, it must do so strain. Adults developing from the in- 
only in particular tissues or at a stage in jected embryos were mated to rosy mu- 
development prior to adulthood. It is also tants, and transformants were identified 
possible that zeste alters the structure of by their wild-type eye color. In this way, 
the white transcript without detectably flies carrying the injected white DNA 
altering its overall size. segment could be identified without re- 
quiring its phenotypic expression. Seg- 
ment A was inserted into the P-element 
vectors without rosy and injected into 
Germ Line Transformation with white mutant embryos, and transform- 
white Locus DNA ants with red eyes were recovered from 

rp j u j ■ Dr i n nf i> u- mating of the iniected flies with white 

T. I. Hazelmqq, R. Levis, and G. M. Rubin Aim'* i o i • 

mutants. Transformants were round in 

The recently developed technique of the progeny of 3-4% of the fertile adults 
DNA-mediated germ line transformation derived from injected embryos, irre- 
of Drosophila embryos (see Year Book spective oft he plasmid or host strain used. 
81, 181-188) provides a powerful tool for For each of the three DNA segments 
identification of DNA regions required tested, four initial transformant strains 
for gene regulation. Cloned DNA se- were recovered. Some of these initial 
quences for white can be systematically transformants contained a single site of 
mutated in vitro, then reintroduced into insertion, whereas others carried mul- 
the genome, where the functional con- tiple inserts. In some cases, the trans- 
sequences of these sequence alterations formed segment transposed to a new 
can be assessed. chromosomal site during propagation of 

During the past year we have achieved the initial transformant strain. Figure 30 

our initial goal of producing white trans- lists the total number of single-insert 


i transcribed region i 

* T3 

o c 
o - 




A L 

B L 

C L 

!? t> 


% < O * O 

i i i i i 
v v v v v 



— r 
+ 5 


' 5 * Y 

c o y -P 
- o 
x UJ 

1 I 

+ 10 



£ p c 















Fig. 30. Germ-line transformants carrying segments of white locus DNA. A physical map of the while 
locus is shown with coordinates in kilobases. The positions of a number of restriction enclonuclease 
cleavage sites and DNA insertions associated with mutations are indicated. The transcribed region of 
the locus and direction of transcription are also shown, but intron/exon borders are not demarcated. 
Below the map, the brackets A, B, and C show the DNA segments transformed into embryos. The table 
at the right indicates the number of independent single-insert transformant strains carrying each seg- 
ment, and the eye color of flies carrying this segment and a deletion for the wild-type white locus. 

transformant lines we have isolated for 
each chromosomal segment. 

The phenotypic expression of seg- 
ment-A transformants could be observed 
immediately in Gl flies, since the host 
for these injections was a white mutant. 
For segment-Z? and segment-C trans- 
formants, the wild-type white allele of 
the host had first to be replaced with a 
null mutant allele. Segments A and B are 
able to confer an eye color indistinguish- 
able from wild type, while the eyes of 
segment-C transformants are unpig- 
mented. Therefore, we conclude that the 
segment between coordinates -5.1 and 
+ 6.7 is able to bestow a wild-type eye 
color on flies bearing it at any one of about 
17 chromosomal positions, but that a seg- 
ment 3.4 kb smaller on the 5' end gives 
no detectable phenotypic expression. 

Segment-A and segment-5 transform- 
ants were tested for repression of the 
inserted white gene by the zeste muta- 
tion. For either segment, some transfor- 
mant lines show typical repression by 
zeste, some show partial repression, and 
some show no detectable repression. We 
interpret this to mean that the repres- 
sion of white by zeste is more sensitive 
to its chromosomal environment than is 

white expression in a zeste + genetic 

The white locus exhibits dosage com- 
pensation typical of an X-linked gene; 
males with one copy of white have the 
same amount of eye pigment as do fe- 
males with two copies. The insertion of 
white DNA segments into autosomes 
provides a unique opportunity to define 
the DNA sequences at white which are 
necessary for normal dosage compensa- 
tion. Visual comparisons of male and fe- 
male autosomal transformants suggest 
that dosage compensation is occurring in 
most lines. We are currently assaying 
the pigment levels in male and female 
flies heterozygous for various autosomal 
insertions to confirm this observation. 

Chromosomal Rearrangements 

Induced by the FB Transposable 


M. Collins and G. M. Rubin 

We have used the white-crimson Or' ) 
mutation to screen for phenotypically de- 
tectable chromosomal rearrangements 
induced by an FB transposable element. 
w c , a highly mutable allele of the white 
locus, was isolated as a partial revertant 



of white-ivory (w l ) (Green, Genetics, 56, 
429, 1967). The w l mutation results from 
the duplication of 2.9 kb of DNA within 
the white locus (Karess and Rubin, Cell 
30, 1982). The w c mutation results from 
the insertion of a 10-kb FB element into 
the w l duplication (Collins and Rubin, Cell 
30, 71, 1982; Levis, Collins, and Rubin, 
Cell 30, 551, 1982). 

FB elements are dispersed, middle-re- 
petitive DNA sequences with long, in- 
verted repeats at their termini (Potter, 
et al., Cell 20, 639, 1980; Truett, et al., 
Cell 2J>, 753, 1981). These inverted re- 
peats vary in length and are composed 
of many small, imperfect direct repeats. 
The inverted repeats in the w c FB ele- 
ment, 2.2 kb and 3.4 kb in length, are 
separated by a 4-kb sequence present in 
only a few copies in the Drosophila gen- 
ome (Levis, et al., Cell 30, 551, 1982). 

w c mutates at a frequency of from one 
per 100 to one per 1000 X chromosomes 
to novel phenotypes, including wild-type, 
white-ivory, and white-eyed derivatives. 
We have reported (Collins and Rubin, 

1982; Year Book 81, 173-175 that wild- 
type revertants of w c result from the ex- 
cision of both the w c insertion and one 
copy of the w l duplication, and that w l 
derivatives result from the apparently 
precise excision of the insertion (Fig. 31). 
We have recently sequenced the excision 
point in two w l revertants and have shown 
that this excision is truly precise (Collins 
and Rubin, Nature 303, 259, 1983). 

We are now analyzing the structure of 
ten white-eyed, male- viable derivatives 
of w c by Southern blots of genomic DNA. 
Six derivatives are deletions of at least 
12 kb of DNA, with one endpoint lying 
within the insertion and the other end- 
point within a 3.5-kb fragment beyond 
the region coding for the 3' end of the 
white locus transcript. This suggests that 
there is a hotspot for deletion formation 
beyond the 3' end of the locus. Deletions 
of the 5' side of the locus may be less 
common, or may not result in a white- 
eyed phenotype. We are cloning se- 
quences from these deletion breakpoints 
for fine-structure mapping. 












Fig. 31. Structure of ^-induced chromosomal rearrangements. Sequences duplicated in w' are in- 
dicated by the open bars. The filled triangles represent the uf insertion and related sequences. The 
following structures are represented: (a) w c ; (b) wild-type revertants; (c) w' revertants; (d) deletions 
with one breakpoint within the insertion and one to the left of the insertion; (e) phenotypically unstable 
white-eyed derivative; (f and g) phenotypically stable white-eyed derivatives. 


We have isolated four derivatives that be complemented in trans by a nonde- 

contain rearrangements of sequences fective 2.9-kb element. The factors de- 

within or near the insertion (Fig. 31). termining P cytotype are also probably 

One phenotypically unstable derivative encoded by the 2.9-kb element, although 

now contains a triplication of the se- the evidence is more circumstantial: only 

quences which are duplicated in w 1 ; it flies containing these elements have been 

probably arose by an unequal crossover, found to possess the P cytotype. 

A second mutation is similar in structure. We have constructed several P-ele- 

The third derivative contains a re- ment derivatives for a genetic analysis 

arrangement within the insertion which of P-element-encoded functions. The first 

increases its size to 14 kb. This deriva- of these derivatives is a nondefective P 

tive is apparently phenotypically stable element that bears the phenotypically se- 

despite the presence of the larger inser- lectable gene rosy in a nonessential re- 

tion. The fourth derivative results from gion of the element just internal to the 

the deletion of sequences within the in- right-hand inverted repeat. When trans- 

sertion, accompanied by a recombination ferred to the genome of an/ - fly, this 

event in flanking DNA. The net result of element confers the ry + eye phenotype. 

this double event is a phenotypically sta- Therefore, the presence of this P element 

ble mutation that has lost the w l dupli- can be monitored by following the ry + 

cation, and that now contains a 3.5-kb marker. 

FB element insertion. We are continuing After determining that this ry-P ele- 
our analysis of these rearrangments to ment acts in several measurable respects 
elucidate further their origin and to de- as does the original P element (see be- 
fine structural requirements for muta- low) and that it did not break down upon 
bility. integration into the fly genome, we con- 
sidered the ry-P as our wild-type ele- 
ment. We then created small frameshift 
In Vitro Mutagenesis and In Vivo mutations in each of four open reading 
Assay of P-Element Functions frames of the P element, which were de- 
fined by the sequence study of O'Hare 
and Rubin (Year Book 81, 179-180), and 
We have combined the P-element- which we have found to be represented 
mediated gene transfer system of Sprad- in poly A + RNA of P-strain flies, 
ling and Rubin with in vitro mutage- The ry-P mutant elements were intro- 
nesis to begin defining the functions duced into the fly genome with the aid 
encoded by P elements. of a modified helper element. The P ele- 
From the behavior of these elements ment in pir25.7wc lacks the last 23 bases 
and their role in P-M hybrid dysgenesis of the right-hand inverted repeat and is 
(see Year Book 81, 178-179), at least two apparently incapable of integration; yet 
functions are predicted: transposase and it still produces the transposase activity, 
determinants of cytotype. The trans- Thus, a mutant ry-P element can inte- 
posase is responsible for the excision and/ grate into the fly genome without regard 
or integration of P elements when they to the nature of the mutation and without 
are present in a cell of the M cytotype. the complications of a nondefective un- 
Transposase is most likely encoded in the marked helper element integrating as 
P element, as indicated by the observa- well. 

tion of Rubin and Spradling {Year Book From the wild-type ry-P and each of 
81, 182-183) that the 2.9-kb element can four frameshift mutants, we have de- 
autonomously transpose from its plasmid rived several Drosophila lines, each con- 
donor to the recipient fly genome. Par- taining a single marked P element, and 
tially deleted elements are incapable of we are currently testing the dysgenic 
autonomous transposition, but they can properties of these flies. 

R. Karess mid G. M. Rubin 


The wild-type ry-P element is fully but at least tenfold less frequently at the 
competent in transposase activity by three white (w) locus. The phenotypes of dys- 
assays. (1) This element can autono- genesis-induced sn mutations vary con- 
mously transpose from the vector plas- siderably, from barely discernable to 
mid to the Drosophila genome. (2) In a extreme, while all the dysgenesis-in- 
transformed fly where the ry-P has in- duced w mutations that we have char- 
tegrated onto the X chromosome, we have acterized have similar, null phenotypes. 
found that the element continues to To understand the basis (or bases) for 
transpose, at a low frequency, to the au- these differences, we have undertaken 
tosomes. (3) When crossed into an M- the isolation of the sn locus and the char- 
strain genome homozygous for the sn lv acterization of dysgenesis-induced mu- 
mutation (which is highly mutable in the tations. 

presence of intact P elements), the ry-P In our initial attempts to clone the lo- 

element can induce instability at this lo- cus we used the sn w M-cytotype strain 

cus. (Year Book 81). This strain has a dys- 

We are currently employing assay (3) genesis-induced mutation of the sn locus 

to test the frameshift mutant ry-P ele- which has been genetically isolated from 

ments for their ability to destabilize sn w . all functional P factors. We constructed 

We thus hope to identify the sequences a library of sn w M-genomic DNA and 

encoding the transposase function and also screened this library with fragments from 

to determine if instability at sn w is pro- the w locus that contained P elements, 

portional to the number of P elements Phages from the w locus and a site with 

present. We find that a fly with a single homology to P elements were recovered, 

marked P element or with a homozygous This second site appears to be localized 

pair of elements still acts as if it has the near, but not at, the sn locus at 7D and 

M cytotype. That is, P elements within has a small internally deleted P element, 

the genome continue to transpose. If one Derivatives of sn w which had further mu- 

thinks of the P cytotype as being the tated to an extreme phenotype or had 

presence of a repressor molecule, it may reverted to wild type during dysgenesis 

be necessary for a critical number of ele- were examined by Southern blotting, us- 

ments to produce repressor before cy- ing probes from this region. It was clear 

totype switches from M to P. We hope that there was no consistent change in 

to test this by genetically manipulating the structure of the genome in this region 

ry-P elements integrated on different when genetic changes occurred at the sn 

chromosomes to construct flies contain- locus. We presume that this defective P 

ing from four to six complete elements element is not associated with a pheno- 

in their genomes. In addition, we hope type and is maintained in the strain by 

to employ our mutant P elements to de- virtue of its close linkage with sn. Our 

fine the gene- or genes-encoding cyto- failure to isolate phages from the sn locus 

type. from this library remains unexplained. 

To isolate the sn locus, we used a de- 
rivative of the sn w P-cytotype strain which 
had. simultaneously mutated to sn e and 

Hybrid Dysgenesis and the singed hdp. Mutation to hdp in this strain is as- 

Locus sociated with chromosomal rearrange- 

tt o •, rs ^tt in ■# r> l- ment at the P element localized at 17C 

H. Roiha, K. O Hare, and G. M. Rubin /T -, , ,-,-> , ^ 77 __ . n . -,na-i\ 

(Engels and Preston, Cell 26, 421, 1981). 

The frequency with which mutations This derivative (B337.2) had a cytological 

arise in a given locus during hybrid dys- inversion between 7D and 17C. The P 

genesis varies considerably. For exam- element from 17C had been cloned 

pie, mutations occur at the singed locus (p7r25.l) and was used to isolate this site 

(sn) at rates as high as one in 100 progeny from the Canton S strain (pS25. 1). These 



a) 7T ? 


Eo a 
on o 
QD X co 


b m 


1 I 

- o 




O -Q 


— O == 

b) B337.2 

c) Canton S 

d) sn w 

V ' ' mm— i 












a x 

S "g E 

o - o 










Fig. 32. Restriction site maps are shown as indicated: (a) 17CD region of tt 2 (plasmid ptt25.1), (b) 
7D1-2/17CD junctions of B337.2 (plasmids pK6.1 and pK15.2), (c) 7D1-2 region'of Canton S, (d) 7D1- 
2 region of sn w . Solid bars show P-element sequences; hatched bars show flanking sequences from 17CD; 
open bars show sn sequences. The map of the sn w allele was established by DNA blotting, and the 
insertion is shown as P-element sequences, although non-P-element sequences may also be present. All 
other maps were constructed by analysis of cloned DNA segments. The maps are shown so as to facilitate 
comparisons; the configurations do not show their relative orientations in the chromosomes. 

sequences were used in DNA blotting 
experiments to show that there were two 
novel Bam HI fragments in the strain 
B337.2. These fragments, which presum- 
ably represent junction fragments from 
the chromosomal rearrangement, were 
isolated from a Bam HI library made from 
B337.2 and then used to isolate homol- 
ogous phages from the wild-type Canton 
S library. Maps of the B337.2 fragments 
and of the phages isolated from the Can- 
ton S library are shown in Fig. 32. We 
believe that this region represents the 

sn locus. Probes from these phages have 
been used to show that genetic changes 
at s?i are correlated with changes in the 
structure of the genome in the cloned 
region. A map of the s?i w allele that was 
established by DNA blotting is also shown 
in Fig. 32. The sn w allele appears to be 
associated with a 2-kb insertion whose 
structure is unlike that of any defective 
P element we have characterized. This 
may account for the exquisite sensitivity 
of the sn w allele to P transposase (see 
Year Book 81). 



A. C. Spradling and G. M. Rubin 

P-element-mediated gene transfer 
permits the introduction of cloned DNA 
sequences into the germ-line chromo- 
somes of a complex metazoan organism, 
Drosophila melanogaster (Spradling and 
Rubin, Year Book 81). In principle, this 

should allow one to modify genes in vitro 
and then test their ability to function in 
a normally developing organism. In prac- 
tice, however, the utility of this approach 
depends on several factors, which have 
been studied extensively this year. First, 



the technique for introducing DNA into 
flies had to be further refined so that a 
single copy of virtually any DNA seg- 
ment could be routinely transferred and 
maintained in a host strain. This involved 
the construction and testing of P-element 
vectors that facilitated the assembly of a 
transposon from its component DNA se- 
quences. Other problems were more fun- 
damental. P-element transposons were 
found to integrate at a wide variety of 
sites in the Drosophila genome. Were the 
sites representative of the genome as a 
whole? Would the integrated elements 
remain as stable, functional genomic 
components? To what extent would the 
different flanking DNA sequences at these 
sites influence the expression of genes 
carried on the transposon? If the expres- 
sion of a gene was highly site dependent, 
the usefulness of transformation for 
studying gene regulation would be lim- 
ited. We have studied these questions 
using transposons carrying the rosy gene 
encoding xanthine dehydrogenase (see 
Year Book 81, 181). As described below, 
none of the potential problems listed above 
was found to be significant. Conse- 
quently, we are now convinced that the 
P-element method can be practically ap- 
plied to a wide variety of problems. In- 
deed, work described in other sections 
of this Report (Collins et al.; de Cicco et 
al.) illustrates that such studies are al- 
ready well under way. 

Vectors for P-Element-Mediated 
Gene Transfer 

We describe here several new nonau- 
tonomous P-element vectors. These vec- 
tors are small plasmids having convenient 
unique restriction sites for cloning DNA 
fragments within a P element but which 
lack sites in the plasmid backbone out- 
side the element. We constructed plas- 
mids according to the protocol 
diagrammed in Fig. 33. First, the Sal I 
fragment of p6. 1 (Rubin and Spradling, 
Science 218, 348), which contains a non- 
autonomous P element inserted into white 
locus DNA (Rubin et al, Cell 29, 987), 

Hind IHn 

L SalI 



Digest with BamHI 
Digest with exonuclease Bal 31 
Ligate with T4 DNA ligase 
Transform E. coli 

Sall J 

Hind ID 



Digest with Sail 
Digest with exonuclease Bal 31 
Ligate with T4 DNA ligase 
Transform E. col i 

Hind III- 


Insert EcoRI linker into PvuE 
site to create pCIW4 

Insert poly linker into pCIW4 
between EcoRI and left HindHI site 

Hind HI 
Ca rnegie 1 : [ 

Hindm- Pstl-Sall-BamHI-Smal- EcoRI 
Car negie 2. | 

Car negie 3: | 

'Hindm-Pstl-Sall-BamHI-Kpnl- EcoRI ' 
Ca rnegie 4- | 


Fig. 33. Construction of vectors; steps were 
performed as indicated. DNA sequences derived 
from the pUC8 plasmid are shown stippled, those 
from the P element are solid, and those from the 
D. melanogaster white locus are open. 

was ligated into the Sal I site of the plas- 
mid vector pUC8 (Vierira and Messing, 
Gene 19, 259) to create the plasmid 
pCIWl. Then, two successive treat- 
ments with BAL-31 nuclease were used 



to eliminate undesirable restriction sites 
and reduce the size of the plasmid. Fi- 
nally, the indicated linkers were inserted 
to create the plasmid vectors Carnegie 
1-Carnegie 4. The vector Carnegie 20 
(Fig. 34) was made by inserting a 7.2-kb 
Hind III fragment containing the rosy 
gene into the Hind III site of the Car- 
negie 2 polylinker. We have successfully 
used Carnegie 20 to transfer DNA frag- 
ments that do not themselves confer a 
recognizable phenotype by selecting for 
rosy + transformants. 

Figure 34 summarizes the results of 
testing various vector constructs for their 
ability to mediate gene transfer. For these 
assays (experiments 1-8), we inserted a 
7.2-kb Hind III fragment or an 8.1-kb 
Sal I fragment containing the rosy gene 
into the P element of each vector con- 
struct. The resulting plasmids were mi- 
croinjected into rosy mutant embryos. 
DNA of a plasmid carrying the 2.9-kb P 

element (either pir25.1 or pir25.7) was 
co-injected to provide trans-acting func- 
tions required for transposition of the 
nonautonomous rosy transposons. Adults 
developing from injected embryos (GO 
adults) were mated to rosy" flies. The 
presence of rosy + progeny in the next 
generation indicated that the vector being 
tested was competent for germ line 

Figure 34 shows that Carnegie 1-Car- 
negie 20 were all highly efficient vectors 
for gene transfer. Our attempts to trans- 
fer rosy genes carried on linear rather 
than circular DNA molecules were un- 
successful regardless of whether the gene 
was within a P element or not (experi- 
ments 7 and 8). Furthermore, our initial 
attempts to construct a transposition-de- 
fective P element which was still able to 
provide transposition functions in trans 
to nonautonomous elements did not suc- 
ceed (experiments 9-11). Nonetheless, 


DNAs injected 


pVIO: bmfr7m77777//////////;////)/////////////////////h E55S5Sgg3 I + pEI25.l 


pVIl: k*mr7m77777//////////)//// //////////////////////h ESSSagSSa i + pEI25.l 


pV9: I ■■ I I I +pri25.l 


Pnrnpri.Pi i )M) hi) i) )) )) ) mi /ll in)/l//ll////! /nail// Ul/llhl/m ES5555gS3 I + pil25.7 

H ^X/S-H X^ ^R H RX/SJH 

pVI2: W//V//////////////////)f/////////////////////////?/7J?m l-.v.-.i-:.:-.V-.-:V7?T1 + pI125.7 


Injected Fertile 
(No) adults 

GO ry+ 

expression progeny 

linear rosy SH \jl H ^\? 

7 fragment : W/////////////////////Y//////////////////////////////)\ R 

linear pry I S H X/SH X R hIx/SH 

8 fragment : I ■■ ' >////////W////V///////////// ////////////)/777m- 

+ pI125.7 

+ pEI25.7 
+ pri25.7 

I 23 

I 33 

I 56 
I 71 
I 16 

I 83 

6 ( 5%) 

9 ( 7%) 

5 ( 4%) 

10 ( 6%) 

21 (12%) 

20 (17%) 

9 ( 5%) 

4 (67%) 

6 (67%) 
4 (80%) 
4 (40%) 

7 (33%) 
I I (55%) 

4 (67%) 
6 (67%) 

5 (50%) 

14 (70%) 




10 pn25.7AI: bMJj— h—jl 1 J52Z2SSS£S a 


11 pn25.7A2: 

+ pry I +pry 3 
+ pry I + pry 3 

+ pry I + pry 3 

121 1 7 ( 14%) 

208 21 (10%) 

114 6 ( 5%) 

6 (35%) 5 (29%) 


2 (33%) 

Fig. 34. Testing of vector constructs. The structures of the microinjected DNAs are diagrammed. 
Rosy sequences are shown cross-hatched; P-element, white, and pUC8 sequences are indicated as in Fig. 
33. Cleavage sites for Hind III, Eco RI, Xho I, Sal I, and Bam HI are indicated by H, R, X, S, and B, 
respectively. X/S indicates the product of ligation of Xho I and Sal I ends. All DNAs were microinjected 
as supercoiled plasmids (except for experiments 7 and 8) at the following concentrations: ptt25.1, p-n-25.7, 
and their deletion derivatives, 50 |xg/ml; ros#-containing DNAs in experiments 1-8, 300 (xg/ml; pryl and 
pry3 in experiments 9-11, 150 |xg/ml each. The number of embryos microinjected and the number and 
percentage surviving to become fertile adults are given. Also shown are the number and percentage of 
fertile adults showing GO expression of the rosy gene, and the number and percentage of fertile adults 
giving transformed rosy + progeny. 


nearly any DNA segment could be in- useful for studies of position effect, lines 

troduced into Drosophila chromosomes containing single transposon inserts had 

with a minimum of effort using one of the to be prepared for the initial population 

new Carnegie vectors. of transformants obtained. Since rosy 

gene expression is affected by several 

The Effect of Chromosome unlinked genes, it was necessary to 

Position on Xanthine P ai " ta ! n a u ™°™ f net ! c background 

Dehydrogenase Expression m a11 th . e ^ lvl dual hnes 1S0 iated. Con- 

sequently, all subsequent genetic crosses 

Gene transfer methods provide a pow- involved only the same rosy mutant strain, 

erful technique for determining the ex- ry**, that was used as a host for microin- 

tent to which regulated expression of a jection. Our strategy was first to deter- 

gene depends on its normal chromosomal mine the number and location of the 

environment. An identical segment of integration sites in each population of 

DNA containing the gene of interest is transformants. Then, we derived one or 

introduced into the chromosomes of a se- more sublines containing only one of the 

ries of individuals of identical genetic inserts. Using this approach, 36 isogenic 

background. Lines are then established lines were established which contained 

differing only in the site at which a sin- single transposon insertions. Of these, 23 

gle, intact copy of the gene has inte- lines contained inserts of the chromo- 

grated. If all other potential sources of somal Sal I fragment, while 13 contained 

variability between the lines can be elim- the Hind III segment. Figure 35 illus- 

inated, the ability of the gene to function trates the chromosomal location of the 

in each line would then reflect the influ- rosy gene present in each line. Also shown 

ence of the surrounding chromosomal se- are the locations of all the sites of trans- 

quences. poson insertion detected in the chromo- 

We have used P-element-mediated somes of the initial transformants. 
transformation to construct a series of The distribution of insertions is slightly 
Drosophila strains that contain a single nonrandom with respect to the 100 cy- 
copy of the wild-type rosy gene at di- togenetic chromosome subdivisions 
verse sites. The rosy gene provides a (0.05>p>0.01), but it is clearly very di- 
sensitive assay for position effects, since verse. Only one insertion appeared to have 
its product, xanthine dehydrogenase occurred within heterochromatin. Either 
(XDH), is easily assayed and displays a P elements do not frequently integrate 
distinctive pattern of expression during within the approximately 30% of the gen- 
development. In examining XDH pro- ome represented in the polytene chrom- 
duction in 36 such lines, we found that ocenter or such insertions were not 
tissue-specific rosy gene expression was detected by in situ hybridization. The 36 
not qualitatively influenced by chromo- sites selected for further study are a rep- 
somal position. However, the quantita- resentative sample of the total popula- 
tive level of enzyme activity produced tion. No visible mutant phenotypes were 
was sometimes subject to position effect, detected in flies from any of the lines, 
In particular, rosy gene insertions on the but about 30% of the insertions appeared 
X chromosome exhibit dosage compen- to be associated with recessive lethality, 
sation, a characteristic of normal X-linked This is similar to previous estimates (25- 
genes in Drosophila. 50%) of the frequency of lethality asso- 

We constructed defective P elements ciated with unselected chromosome 

containing either an 8.2-kb Sal I frag- breaks induced by x-ray treatment (see 

ment or a 7.2-kb Hind III fragment of Lefevre, Annu. Rev. Genet. 8, 51, 1974). 

chromosomal DNA that includes the rosy Thus, the distribution of transposon in- 

gene, and injected them into embryos sertions are probably unbiased with re- 

from a rosy mutant strain (ry U2 ). To be spect to the location of vital loci. Finally, 




▼ o* o 



o v 

V V o 


▼ T 




1 1 1 

— C\J _ 




co c 


ro ^ _• 




o o o 



r^- r*- 


<3- <j- r~- 



<r cc 


or cc or 





V o 



V V 


▼ o v 







— ; 


— ■' 





U3 10 



or or 





O ▼ 



. ov * 

' ▼ 

▼ o 



% /)A\ 

1 [ 




KJ to 




o o 




r- ro 



cc cc cc cc cc 

q: cr 







V ▼ 

▼ ▼ 







co ib 



o o 



to r~ 



or or 


O ^O 

o ▼ O n 




▼ ▼ 

vo o 


▼ V 





_/ . CO cj 

• r»' * oo — 






-OOO o 



o o 


cc cc act cc 





10 r- 
or or 

oryl or ry3 v ry 10 ▼ ryll 

Fig. 35. Chromosomal sites of transposon in- 
sertion. All the sites of insertion of the indicated 
rosy transposons identified by in situ hybridization 
to G2 larvae are shown for each of the five major 
chromosome arms. The location of the single trans- 
poson present in each of the 36 isolines is indicated 
by the labels below the lines depicting each chro- 

the use of wild-type eye color phenotype 
to select the initial transformant popu- 
lation probably did not result in the elimi- 
nation of many nonfunctional sites. In all 
five cases where at least three isolines 
were established from lines containing 
two or three inserts, each of the fourteen 
individual sites was shown to be active. 
Previous studies have indicated that 
exogenous DNA, following introduction 
into a variety of mammalian cells, may 
undergo rearrangement and may not be 
transmitted or expressed in an orderly 
manner. We therefore examined the iso- 
lines for evidence of rearrangement, in- 
activation, or secondary transposition of 
introduced P elements, all of which could 
potentially interfere with the detection 
of position effects. During the isolation 
of the isolines, we found no obvious evi- 
dence of high rate of transposition. To 


investigate the possibility of even low 
levels of instability, tests were carried 
out using eight lines containing X-linked 
insertions. Males from these lines were 
crossed to attached-X females that were 
also homozygous for a partial deletion of 
the normal rosy locus. In such lines, males 
will display wild-type eye color, while fe- 
males will have rosy eyes, as long as the 
transposon carrying the rosy gene re- 
mains X-linked. However, movement of 
the transposon to an autosome would be 
detected, since it would give rise to wild- 
type eye color in a female. Furthermore, 
if the rosy transposon is excised or in- 
activated by any mechanism not leading 
to lethality, it would result in the ap- 
pearance of a male with the mutant eye 
color. Among more than 12,000 individ- 
uals scored over four generations, no wild- 
type females resulting from such trans- 
position events were observed, and only 
two rosy' males were detected. Hence, 
these studies show that the frequency of 
transposition or inactivation is not high 
enough to affect measurements of XDH 
gene function. 

Since other potential factors had been 
eliminated, any differences in XDH 
expression between the isolines can be 
attributed to the influence of chromo- 
somal sequences flanking each transpo- 
son. Initially, all the 36 isolines were 
screened to determine if the pattern of 
tissue-specific expression of the rosy gene 
present on the transposon was distin- 
guishable from wild-type. Flies hetero- 
zygous for the transposon were selected 
from each strain and tissues were dis- 
sected and stained for XDH activity. XDH 
activity in all strains was present in the 
adult Malphigian tubules and in larval fat 
body cells — the same tissues which con- 
tain high levels of activity in wild-type 
flies. Little or no staining was seen in the 
adult testis, again paralleling the normal 
situation. Since the XDH activity pres- 
ent in an adult is the sum of activity in 
a variety of tissues, changes in either the 
tissue distribution or level of rosy gene 
expression could alter total XDH-specific 
activity. Extracts of flies from each of 



the 36 isolines were prepared and as- 
sayed for XDH activity. Activity was 
normalized to the protein content of the 
homogenate. To demonstrate that the 
specific activity measured in these ex- 
periments was a heritable property of 
the strain in question, the measurements 
were repeated on flies from subsequent 
generations. The XDH activity of the lines 
was between 30% and 130% of the value 
for an Oregon R strain. The specific ac- 
tivity of each strain was found to be highly 
reproducible in a subsequent generation, 
and in duplicate assays. The distribution 
of activities observed in the 36 lines is 
presented graphically in Fig. 36. 

Comparison of the levels of XDH ac- 
tivity in different strains provides in- 
sight into some of the potential variables 
in evaluating position effects on rosy gene 
expression. For example, no correlation 
existed between XDH activity and the 
orientation of the Hind III fragment con- 
taining the rosy gene relative to the P- 
element vector. (The orientation of the 
Sal I fragment in lines transformed with 

a mixture of ryl- and ry3-containing 
plasmids was not determined in most 
cases). These observations suggest that 
the P element does not play a simple role 
in the expression of the rosy gene con- 
tained on the transposon, for example, 
by providing a strong promoter. They do 
not rule out the possibility that the P 
element could play some other role, such 
as providing an enhancer-like activity. 

In Drosophila, X-linked genes show 
equivalent activity in males and females 
despite the difference in gene copy num- 
ber, a phenomenon called dosage com- 
pensation. It was therefore of interest to 
determine if rosy genes inserted on the 
X chromosome came under such regu- 
lation. The XDH-specific activity in male 
and female flies from all eight lines con- 
taining X-linked rosy genes was mea- 
sured. For comparison, identical 
measurements were carried out on seven 
lines with autosomal insertions. Since re- 
productively active females such as those 
used in these measurements contain large 
amounts of yolk protein, their XDH-spe- 




8.1 kb Sail inserts 




. E 6 


w. 4 




3 Z 




72 kb Hindi inserts 




.6 .8 .2 

XDH specific activity 



Fig. 36. XDH-specific activity in transformed lines. The distribution of mean XDH-specific activities 
(nmol isoxanthopterin/min/mg protein) among lines containing Sal I- or Hind Ill-bearing transposons 
is illustrated. Lines containing ryl or ry3 inserts were not distinguished. However, those containing 
rylO (open bars) are shown separately from those containing ryll (stippled bars) in the right panel. The 
structure of these transposons is summarized at top of figure: P-element sequences are shown in black, 
while sequences from pDm2844S8.5 containing the chromosomal rosy gene are stippled. For comparison, 
a wild-type Ore R line had an XDH-specific activity of 0.60. The width of the bars is approximately equal 
to the within-line standard deviation of measurement for lines with average activity. 


cific activity is lower than that in males 
by an amount that depends on the details 
of the culture conditions, age, etc. These 
variables were controlled so that any dif- 
ferences between the X-linked and au- 
tosomal lines would reflect dosage 
compensation. The results of these mea- 
surements are shown in Fig. 37. 

The average ratio of the activity in 
males to females for the eight autosomal 
lines was 1.6. This was similar to a value 
of 1.7 in an Oregon R line containing ac- 
tive rosy genes at the normal autosomal 
location. In contrast, the activity ratio in 
the X-linked lines ranged from 1.9 to 3.2, 
with a mean of 2.6. Two lines, R404.2 
and R301.2, showed nearly the twofold 
elevation expected for complete dosage 
compensation. In the other lines, a lower 

no dosage 

full dosage 








C 2 

.3 1.5 1.7 1.9 2.1 2.3 2.5 2.7 2.9 

d7 9 XDH activity ratio 

3.1 3.3 

Fig. 37. Dosage compensation of X-linked rosy 
genes. The distribution of ratios of male to female 
XDH-specific activity is shown for eight lines con- 
taining a rosy gene on the X chromosome (stippled 
bars) and for seven lines containing an autosomal 
rosy gene (open bars). The mean for the autosomal 
lines, indicated as "no dosage compensation," is 
similar to the ratio obtained in a wild-type strain 
containing the rosy gene at its normal autosomal 
location. A value of twice the mean autosomal ratio 
is indicated "full dosage compensation." 

degree of elevation occurred, but all lines 
showed some degree of dosage compen- 

The studies reported here represent a 
direct test of the sequence requirements 
for developmentally regulated expres- 
sion of the rosy gene. At least 4.5 kb of 
DNA would be required to code for the 
160-kilodalton XDH polypeptide. There- 
fore, while the exact location of the rosy 
structural gene within the 7.2-kb Hind 
III fragment is unknown, at most 2-3 kb 
of flanking sequences could have* been in- 
cluded in the transposons used. Despite 
the small amount of flanking DNA, flies 
containing a transposon expressed nor- 
mal or nearly normal amounts of xan- 
thine dehydrogenase. XDH activity was 
distributed among larval and adult tis- 
sues in a manner indistinguishable from 
that in wild-type flies. Clearly the rosy 
gene does not require a particular, large 
chromosomal domain for apparently nor- 
mal expression. The absence of strong 
position effects on rosy expression not 
only indicates that a large region sur- 
rounding the normal gene is not neces- 
sary for its expression, but also implies 
that other chromosomal domains capable 
of overriding the local controls on XDH 
expression are rare. 

These results demonstrate that the 
mechanisms responsible for dosage com- 
pensation can spread to adjacent auto- 
somal genes. In each of eight independent 
insertions of a 7-kb or 8-kb autosomal 
DNA fragment into a site on the X chro- 
mosome, partial or complete dosage com- 
pensation of XDH-specific activity was 
observed. Furthermore, the increased 
ratio of XDH-specific activity in male 
compared to female extracts observed in 
the case of X-linked inserts was due to 
an increase in the absolute specific activ- 
ity in males, rather than to decreased 
activity in females. The specific activities 
of the extracts from females were similar 
for autosomal and X-linked lines. Thus, 
the dosage compensation of rosy genes 
we have observed occurs as a "position 
effect" exerted to a differing extent by 
different local X-chromosome sequences; 


it takes place by increasing the activity will have to be made in lines where the 
of the gene in the male X chromosome chromosomal location of the inserted gene 
relative to the female X or relative to the differs. In some cases it may be neces- 
activity of that same gene at most au- sary to include an internal control gene 
tosomal locations. These observations within the transposon to monitor the po- 
suggest the existence of multiple sites tential activity of the particular site at 
located along the X chromosome which which insertion has occurred. However, 
mediate a hyperactivation of gene activ- our results suggest that in many cases it 
ity in a local region. may be possible to make direct conclu- 
For many purposes, it would be useful sions based on the properties of transpo- 
to construct a series of mutant genes in sons inserted at a small number of sites, 
vitro and then test their ability to func- Thus, position effects do not appear to 
tion by reintroducing each into the germ represent a serious obstacle to the anal- 
line using P-element-mediated gene ysis of gene function by in vitro muta- 
transfer. In such a scheme, comparison genesis and P-element-mediated 
of the properties of different constructs transformation. 


D. V. de Cicco, L. J. Kalfayan, J. Levine, S. Parks, B. T. Wakimoto, and A. C. Spradling 

During oogenesis in Drosophila, two sis. Her results are consistent with the 
clusters of chorion genes undergo coor- view that the amplification of an entire 
dinate amplification; individual genes are gene cluster is mediated by a single reg- 
then expressed according to a precise de- ulatory element, which probably serves 
velopmental timetable. This year, we have as the origin of the extra rounds of rep- 
begun to obtain definite evidence as to lication. 

the size and location of the regulatory The expression of the individual cho- 

elements involved in programming this rion genes within a cluster are not reg- 

intricate genetic subroutine, which forms ulated coordinately, however (see Parks 

part of a program leading to the produc- et al., Year Book 80, 188; 81, 190). The 

tion of the beautiful shell that protects simplest model would place individual 

the developing embryo. We owe much of regulatory elements that determine the 

our progress this year to the P-element- timing and tissue specificity of gene 

mediated transformation technique (Year expression, near the coding sequences of 

Book 81, 181), which has progressed to each chorion gene. Wakimoto and Kal- 

the point of becoming a standard labo- fayan have carried out experiments which 

ratory procedure (see Spradling and demonstrate that the s38 chorion gene 

Rubin, this Report). contains closely linked regulatory ele- 

Last year, de Cicco obtained the first ments. This gene was shown to function 

evidence that a relatively small subre- autonomously when introduced at new 

gion of the chorion gene cluster could in- chromosomal sites along with small 

duce amplification at novel chromosomal amounts of flanking DN A sequences; the 

sites into which it had been introduced, tissue specificity and even the precise 

Since that time, she has transferred seg- timing of expression during oogenesis 

ments of DNA which together comprise were retained. In one variant of these 

the central gene-containing portions of studies, the 5' portion of the s38 gene 

both chorion gene clusters into flies, and was fused to the (3-galactosidase gene of 

studied their replication during oogene- E. coli prior to introduction into Dro- 


sophila chromosomes. Strains carrying ectopic sites in the genome is tissue spe- 

such fusion genes synthesized comple- cific and able to spread into adjacent se- 

mentary RNA during the same stages of quences. 

oogenesis wherein s38 mRNA is nor- Genomic chorion DNA fragments (de- 

mally made, and high levels of p-galac- picted in Fig. 38) were cloned into a de- 

tosidase activity were detected in their fective P transposable element vector 

egg chambers. Taken together, these marked with the rosy + gene and intro- 

studies demonstrate that, at most, 1.3 duced into rosy~ M-strain embryos by 

kb of DNA sequence is required in cis to co-injection with nondefective P-element 

specify the tissue and temporal program DNA. Transformed rosy + flies were re- 

of s38 gene expression. Further studies covered in the next generation by scoring 

by Parks on the ocelliless mutation (Year for wild-type eye color; lines with single 

Book 81, 190) have also demonstrated the insertions were established. We tested 

autonomously regulated expression of a the transformed DNA in these lines for 

second chorion region transcript. its ability to induce amplification by com- 

The identification and precise locali- paring the DNA prepared from egg 

zation of the regulatory elements re- chambers of early stages (1-8) and late 

sponsible for the normal functioning of stages (12-13) with the DNA from ovar- 

chorion genes introduced into chromo- ectomized females, 

somes on P-element transposons is the The two constructs made to test three 

first prerequisite to understanding their DNA fragments from the third chro- 

mechanism of action. Two additional areas mosome cluster are shown in Fig. 39. 

of work, described below, are important The plasmid p7710 contains a 7.7-kb Eco 

in reaching this goal. (1) The full reper- RI fragment (see Fig. 38), and p37512 

toire of regulatory activity carried out contains the 1.9-kb and 2.0-kb Eco RI- 

by genes within the two gene clusters Sal I fragments within the 7.7-kb frag- 

must be defined (Parks). (2) The DNA ment, arranged back-to-back. Egg cham- 

sequence of genomic regions which are ber and carcass DNA from line 771030, 

important in specifying chorion gene am- which has a single insertion of the trans- 

plification and expression must be deter- poson containing the 7.7-kb Eco RI frag- 

mined (Levine). ment, was digested with Bam HI, elec- 

trophoresed, transferred to nitrocellulose 

paper, and hybridized successively with 

Characterization of the chorion- and ros?/-DNA-specific 

Amplification Origins probes shown in Fig. 39. The same ex- 

D periment was carried out with three lines 

(375 12L, U, and Y) containing different 

Last year, we reported experiments insertions of the transposon from plas- 

wherein we used the technique of P mid p37512, except that the DNA was 

transposable element transformation to digested with Eco RI. Data are shown 

reintroduce two DNA fragments from the for one line only (Fig. 40), as all three 

chorion gene clusters into the genome of lines were found to be equivalent. Since 

Drosophila. Preliminary results showed endogenous rosy DNA does not amplify, 

that one of these, a 3.75-kb Sal I frag- the intensity of hybridization of the rosy 

ment from the third chromosome cluster, probe to the host-specific rosy DNA frag- 

initiated amplification of DNA sequences ments B in panels a and c gives a measure 

at the site of integration in late but not of the total amount of DNA present in 

in early stages of egg chamber devel- each lane. It is clear for both lines that 

opment. the amount of DNA from egg chambers 

Here, we describe experiments con- of late stages is much lower than that 
firming this initial observation and show- from egg chambers of early stages or car- 
ing that the amplification induced at casses. Not suprisingly, the intensity of 


sl5 sl8 

R R S Bg Ba S R 

I 1 N— ^m-\ 1 1 




1.9 2.0 



s38 s36 

R R R A R A R 

I 1 MM 1 ■■ 1 1 1 1 






Fig. 38. DNA fragments cloned into P-element vectors to test for origin activity. DNA fragments 
from the regions of the third and X chromosomes which undergo maximal amplification were cloned into 
P-element vectors marked with the rosy + gene (pVll and Carnegie 20; see Spradling and Rubin, this 
Report, for details of construction). Each fragment is depicted with its size in kilobases. The positions 
of the four major chorion protein genes on the chromosomes are indicated with a thick line. R = Eco 
RI; S = Sal I; Bg = Bgl II; Ba = Bam HI; A - Ava I. 

hybridization of the chorion probe to the bands of hybridization reveals that the 

host-specific chorion DNA fragments 2 transformed DNA in line 771030 is am- 

and 3 for line 771030 (panel b) and frag- plified about 16-fold, which is about a 

ment 2 for line 37512Y (panel d) is much quarter of the maximal amplification of 

stronger than would be expected from third-chromosome chorion DNA (60-fold), 

the amount of total DNA present. This whereas the transformed DNA in line 

is to be expected, since these endogenous 37512Y is not amplified. This amplifica- 

chorion DNA fragments are known to tion follows the same timing and tissue 

specifically amplify in follicle cells of late- specificity as that of the host chorion 

stage egg chambers (Year Books 80, 81). DNA. In addition, the rosy sequences 

In the case of line 771030, the intensity adjacent to the transformed chorion DNA 

with which the transformant-specific (fragments A and C, panel a) also un- 

chorion DNA fragments 1 and U (panel dergo specific amplification in line 771030, 

b) hybridize indicates that these frag- in a manner reminiscent of the proposed 

ments also undergo specific amplifica- mechanism for normal chorion gene am- 

tion, though not as markedly as do the plification. Therefore, at least some of 

host chorion DNA fragments. The trans- the sequences required for origin activity 

formant-specific chorion DNA fragments must be present in the 3.75-kb Sal I frag- 

2 and 3 of line 375 12 Y, on the other hand, ment flanked by the 1.9-kb and 2.0-kb 

do not appear to exhibit any amplifica- Eco Rl-Sal I fragments within the 7.7- 

tion. Densitometric tracing analysis of the kb Eco RI fragment tested. This frag- 

Chorion Probe 


Rosy Probe 

i • 



\ / 
\ / 


/ \ 

s/ \r/ \sh 



Fig. 39. Constructs used to transform three DNA fragments from the third chromosome cluster. 
The plasmid p7710 contains the 7.7-kb Eco RI fragment (see Fig. 38) cloned in the vector pVll. The 
plasmid p37512 contains the two Eco Rl-Sal I fragments within the 7.7-kb Eco RI fragment cloned 
back-to-back, as depicted, into the vector Carnegie 20. The line drawing represents bacterial and defective 
P-element sequences (flanked by short stretches of Drosophila DNA from the white locus). The filled 
and blank blocks represent chorion and rosy DNA, respectively. The orientation of the rosy + sequences 
are opposite in pVl and Carnegie 20 relative to the P-element sequences. 

771030 line 

DNA: £ L C E L C 

37512Y line 

£ L C E L C 

-1 T 

-A T m* 

A T 

B H 

C T 

-4 T 

m0 -b H 

3 T 

probe : 





Fig. 40. Southern blot analysis of two transformed lines. Line 771030 and line 37512Y DNA were 
digested with Bam HI and Eco RI, respectively, fractionated on 1% agarose gels, transferred to nitro- 
cellulose fillers, and hybridized successively with the chorion and rosy DNA probes p302.77 and pRR47. 
p302.77 contains the 7.7-kb Eco RI fragment (see Fig. 38) cloned in pBR322. pRR47 contains a 4.7-kb 
Eco RI fragment (which includes the 4.1-kb Eco RI-Hind III fragment labeled "Rosy probe" in Fig. 
39) cloned into pAT153. E = early stage (1-8) egg chamber DNA; L = late stage (12-13) egg chamber 
DNA; C = ovarectomized female carcass DNA. The DNAs loaded in each track were prepared from 
10-12 females fed on fresh yeast for at least 24 hours. The number of late-stage egg chambers obtained 
was 60-70. T = transformant-specific; H = host-specific. 


ment is located in a region of the third- in ovarian follicle cells. In addition, these 
chromosome cluster, which exhibits genes are temporally regulated in such 
maximal amplification and contains two fashion that very closely linked genes 
genes (sl5 and sl8) coding for two of the within a cluster can exhibit quite differ- 
most abundant chorion proteins. ent patterns of message accumulation 

Preliminary results with six lines in (Parks, Year Book 80, 188). This second 
which the 3.9-kb Eco RI fragment ad- property suggests that sequences con- 
jacent to the 7.7-kb Eco RI fragment (see trolling the fine level of temporal regu- 
Fig. 38) has been transformed suggest lation of expression lie close to the 
that this fragment does not support am- structural gene sequences. We are in- 
plification. The possibility that the level terested in defining the signals respon- 
of amplification induced by the 7.7-kb Eco sible for both time and tissue-specific 
RI fragment is lower than expected be- regulation of chorion genes. We have be- 
cause of position effect is currently being gun to study this problem, concentrating 
investigated by analyzing lines trans- on the s38 gene. This gene codes for one 
formed with the 3.75-kb Sal I fragment of the more abundant chorion proteins 
presumed to contain the origin. Subfrag- and is located within the region of the X 
ments of the 3.75-kb Sal I fragment are chromosome that shows maximal levels 
also being tested to localize the origin of amplification during oogenesis. More 
activity more precisely. specifically, mapping of the s38 gene 

Analysis of the 6.0-kb and 2.8-kb Eco transcript has shown that the mRNA 
R I and 3.9-kb A va I fragments of the X complementary region, and approxi- 
chromosome (see Fig. 38), has not yet mately 1 kb 5' and 2.5 kb 3' to the mes- 
revealed any origin activity. Lines trans- sage are contained within a 4.7-kb Eco 
formed with the 4.7-kb and 4.0-kb Eco RI fragment (Fig. 38). Transformation 
RI fragments on the same chromosome techniques (Spradling and Rubin, Year 
are being currently tested. Since the Book 81) allow us to test the ability of 
maximal amplification of X-chromosome this and other defined fragments to sup- 
chorion genes is normally only 16-fold, port normal s38 gene transcription when 
compared to 60-fold for the third chro- isolated from the chorion gene cluster at 
mosome, and since the transformation various positions in the genome, 
results so far suggest that an origin may We have constructed a transposon con- 
not be fully active elsewhere in the gen- taining the 4.7-kb Eco RI fragment using 
ome, the origin of the X-chromosome the scheme summarized in Fig. 41 A. Our 
cluster may be more difficult to detect, experiments require that we be able to 
The information gained from the study distinguish the gene product of the trans- 
of the third-chromosome origin should formed gene from that of the normal s38 
help to localize its X-chromosome coun- gene. To accomplish this, we modified 
terpart. the s38 gene contained in the transposon 

by inserting a 572-base pair fragment of 
the phage M13 into the mRNA comple- 
mentary region. The insertion of M13 has 

Analysis of Transcriptional the advantage of allowing us to detect 

Regulation of the s38 Gene by the transcript above the wild-type back- 

Transformation ground by virtue of its larger size and by 

n Tir . . t rr is j a a ji- probing for M13 complementary se- 

B. Wakimoto, L. Kalfayan, and A. Spradling c ° ml on -n/ri n r / 

y quences. The s38-M13 fragment was 

Several features of the chorion gene cloned into a P-element vector containing 

clusters make them ideal subjects for the the ry + marker gene. This transposon 

study of transcriptional regulation. First, was then injected into ry~ embryos, and 

the chorion genes show tissue-specific stable transformants were obtained as 

expression, being selectively transcribed ry + progeny from the injected flies. We 


A B 

Carnegie 1: 






S B , lac z SB B 

Sal or Bam 

S ry* S 


Inject into ry 506 embryos 
with pi i25.1 


s38-lac z 

Fig. 41. (A) Construction of the s38-M13 transposon. The 4.7-kb Eco RI fragment containing the s38 
gene was modified by the insertion of a 572-bp fragment from M13 into the Bam HI site. Arrow indicates 
the approximate position of the insert near the 5' end of the message and the direction of transcription. 
The s38-M13 fragments were then cloned into the Eco RI site of pVll so that the region containing 
chorion and ry + gene sequences was flanked by P-element ends (solid blocks). This plasmid was co- 
injected with ptt25.1 into embryos of the ry 506 strain. (B) Construction of s38-lac z fusion genes. The 
indicated steps were carried out according to standard procedures. Briefly, the 1.3-kb Eco RI-Bam HI 
fragment containing part of s38 was ligated into Eco RI-Bam Hi-digested Carnegie 1 DNA to yield 
plasmid pACSl. pACSl was cleaved with Bam HI and ligated to the lac z-containing fragment of plasmid 
pMC1871, which had been partially cleaved with Bam HI. To obtain the third reading frame, the Bam 
HI site of pACSl was filled in, and the DNA was cleaved with Sal I and ligated to Sal I-digested 
pMC1871. All three DNAs containing the lac z gene in tandem orientation to s38 were then cleaved with 
Sal I and ligated to the 8.1-kb Sal I fragment of pDm2844.S85, which contained the wild-type rosy gene. 
Since this fragment could insert in two orientations relative to the fusion gene, this yielded a total of 
six DNAs: pACSll, pACS14, pACS24, pACS25, pACS33, and pACS36. 

have established five separate lines con- 
taining single s38-M13 transposons. In 
situ hybridization has shown that four of 
these inserts are located at various po- 
sitions on chromosome 3, and one on 
chromosome 2. All five of these lines show 
the presence of a transcript containing 
s38 and M13 sequences in poly A + RNA 
isolated from ovaries. This transcript is 
of the expected size, indicating that the 
transformed gene is expressed in the 
proper tissue in all five genomic posi- 
tions. We further examined the trans- 
formed gene to see if it was capable of 

proper developmental regulation during 
oogenesis. The results obtained (Fig. 42) 
indicate that the expression of the trans- 
formed gene parallels that of the normal 
s38 gene; it produces low levels of com- 
plementary RNA during early stages of 
oogenesis and maximum levels during late 
stages. However, the absolute levels of 
RNA from the transformed gene are lower 
than the s38 mRNA levels, presumably 
because the s38-M13 insert lacks ampli- 

The construction of gene fusions pro- 
vides a powerful tool for analyzing the 




O *~« **** f*i "T 

r~4 ,«-« »■"< tr-4 «-« 

pl03.47 probe 

«« « 

1632 — 



puc 8 probe 

i3 X ^J* 


Fig. 42. Developmental profile of the s38-M13 
transcript. RNA was isolated from staged egg 
chambers of the transformed line 1A25-6 and as- 
sayed by Northern blotting for the presence of the 
s38-M13 transcript. RNA was size-fractionated on 
a 1% agarose-formaldehyde gel. Lanes labeled stl- 
9 contain RNA from early stages of five ovaries; 
other lanes contain RNA of 25 staged chambers. 
RNA was transferred to nitrocellulose filters, then 
probed with 32 P-labeled probes containing (upper) 
the 4.7-kb Eco RI fragment containing the s38 gene 
(p 103. 47) or (lower) sequences homologous to the 
M13 insert (puc8). 

function of specific genetic regulatory 
elements and has been widely used in 
studies of E. coli and yeast. This year, 
we have begun experiments to test the 
feasibility of using this approach in our 
studies of Drosophila chorion genes. At 
the minimum, fusions would simplify 
studies of chorion gene function by mak- 
ing the activity of the reinserted gene 
easier to detect. Eventually, however, 
strains carrying an appropriate chorion- 
fusion gene could simplify the isolation 
of new mutations affecting chorion gene 
expression. They might also facilitate de- 
tection of differences in the program of 
gene expression within individual follicle 

Our initial studies have taken advan- 
tage of the versatility of E. coli (3-galac- 
tosidase. The activity of this protein can 
be detected reliably and with great sen- 
sitivity; furthermore, activity is retained 
even when a foreign protein is attached 
to the amino terminal portion of the 
molecule. Consequently, we expected that 
if the 5' flanking sequences and the amino 
terminal coding portion of the s38 gene 
were fused to the lac z gene (coding for 
E. coli p-galactosidase) in the correct 
reading frame, the resulting fusion pro- 
tein might be synthesized with the tissue 
and temporal specificity appropriate to 
s38 but still retain p-galactosidase activ- 
ity. We therefore constructed such fu- 
sions and used them to study further the 
sequence requirements for the regulated 
synthesis of s38. 

The construction of a series of fusion 
genes is shown in Fig. 41B. The produc- 
tion of the desired gene was simplified 
by a gift from M. Casadaban (University 
of Chicago) of plasmid pMC1871. This 
plasmid contains sufficient lac z gene cod- 
ing sequence for the production of an ac- 
tive protein, flanked by several restriction 
sites which facilitate its linkage to other 
genes. Ligation of appropriate restric- 
tion fragments from s38, pMC1871, the 
P-element vector Carnegie 1, and the rosy 
gene resulted in the production of plas- 
mids pACSll-pACS36 (Fig. 41B). Con- 
structions were carried out in such way 


as to insure that fusion of s38 at its in- egg chambers from strains containing 
ternal Bam HI site was made to the amino genes fused in reading frames 2 or 3. All 
terminal end of the pMC1871 polylinker the strains containing s38-lac genes fused 
in all three reading frames relative to the in reading frame 1 contained greatly el- 
P-galactosidase coding sequences in evated levels of p-galactosidase activity 
pMC1871. Each fusion gene was situated in late-stage egg chambers, however. This 
within the nonautonomous P element from activity was sufficient to cause egg cham- 
Carnegie 1, and each contained a func- bers from these strains to turn blue when 
tional rosy gene at its 3' end. Plasmids incubated in the presence of the chromo- 
were constructed in which the rosy gene genie p-galactosidase substrate X-gal 
was present in either orientation relative (data not shown). Thus, gene fusions to 
to the fusion gene. We assumed that se- lac z provide considerable promise for 
quences present in the 8.1-kb Sal I frag- simplifying the detection of modified cho- 
ment of genomic DNA containing rosy in rion genes after transformation, 
at least one orientation might cause ef- The transposons containing these fu- 
ficient termination of transcripts initi- sion genes have retained only the 1.3-kb 
ated at the s38 promoter and lead to the Eco RI-Bam HI fragment from the 5' 
production of a reasonably stable mRNA. portion of the s38 gene. Is there suffi- 
Each plasmid containing a fusion con- cient information on this DNA to cause 
struct was mixed with DNA containing the fusion gene to be expressed with the 
an autonomous P element and injected same temporal and tissue-specific pat- 
into embryos homozygous for the rosy 506 tern present in a normal s38 gene? RNA 
allele. Rosy + strains were obtained con- from individual egg chamber stages of 
taining transposon insertions with genes the fusion strain 4004 was analyzed ex- 
fused in each of the three possible read- actly as described in Fig. 42, except that 
ing frames. To look for expression of the the probe was the lac z sequences in plas- 
fusion genes in these strains, stage 12- mid pMC1871. A single 4.0-kb band of 
14 egg chambers (normally, 38 expres- poly(A)-containing RNA was detected, 
sion begins in stage 11) were dissected, and it displayed a stage specificity iden- 
and extracts were assayed for p-galac- tical to that of s38 mRNA (data not 
tosidase activity (Table 3). Very low lev- shown). Thus, 3' flanking sequences and 
els of p-galactosidase activity were even most of the region coding for s38 
detected in egg chambers of the rosy 506 are not essential for correct development 
host strain or in an Oregon R wild-type regulation. All the sequences necessary 
strain. Similar levels were observed in for the regulation of the s38 gene are 


P-Galactosidase Activity 

in Egg Chambers of Strains Containing s38-lac ; 

'. Fusion Genes 











Ore R 

. . . 













































*a = fusion gene and rosy gene in same orientation; b = fusion gene and rosy gene in opposite 
tO.D. 420 /50 stage 12-14 egg chambers/18 h at 27°C; n.d. = not determined. 


contained on the 1.3-kb Eco RI-Bam HI 
fragment. Chorion genes are turning out 
to be perhaps even more compact than 
we had originally imagined. 

Biochemical and Genetic Studies 
of Chorion Gene Expression 

5. Parks 

Our efforts this year have been de- 
voted to obtaining a more-detailed tran- 
scription map from the X-chromosome 
chorion gene region at 7F. This gene 
cluster encompasses 17 kb of DMA, cen- 
trally located in an area of the chromo- 
some that differentially replicates in 
follicle cells during oogenesis (Year Books 
79, 80, 81). We have previously shown 
that, in spite of their close physical prox- 
imity to one another, these genes exhibit 
individual patterns of developmental 
expression. This year, study of gene 
expression in wild-type flies and the mu- 
tant ocelliless, using Northern blot anal- 
ysis with single-stranded probes, has 


enabled us to characterize further this 

Northern analysis of the 2.8-kb Eco RI 
fragment and the 4.7-kb Eco RI frag- 
ment (which contains the gene for s38) 
reveals a large polyadenylated transcript 
of approximately 3.1 kb (Fig. 43). This 
transcript is larger than any previously 
seen from the X and third chromosome 
amplified regions. In addition, this tran- 
script fails to demonstrate a dramatic 
stage-specific pattern of accumulation; it 
is present at similar levels throughout 
oogenesis. Blots of poly A + RNA from 
males and ovarectomized females have 
failed to hybridize to the 2.8-kb RI frag- 
ment probe. Thus, in spite of its unique 
character, this transcript seems to be egg 
chamber specific. 

A closer examination of the 4.0-kb RI 
fragment, which contains the gene for 
s36 chorion proteins, shows that an ad- 
ditional transcript approximately fifty 
bases larger than that of s36 is produced 
from this region (Fig. 43). This larger 

1 kb 

In (Doc 

EcoRI Frawmsrrts 


















1.4- g 




<n| m < 

© ,* M M <* 

© «* <S <•» «* 

o ••« <N <*> ■<* 


° 3 S3 2 3 

Fig. 43. Analysis of transcription from the chorion gene region at 7F by Northern blotting. RNA 
from wild-type flies was electrophoresed in denaturing formaldehyde gels and blotted onto nitrocellulose 
filters. The niters were hybridized to DNA restriction fragments and a cDNA, which derive from the 
indicated positions on the map. Blots A-E contain poly A + RNA, and blot F contains poly(A)" RNA. 
Transcript sizes are in kb and differ slightly from the values given in Year Book 81 , presumably because 
of the change in denaturing system from glyoxal to formaldehyde. 



transcript is expressed maximally in stage A powerful approach in the elucidation 

11 and decreases in stage 12, while the of factors necessary for gene expression 

s36 transcript is expressed maximally at has been the study of mutants. This year, 

stages 12 and 13. A cDNA for s36, 5F9, we have continued our study of the fe- 

fails to hybridize to this larger transcript male sterile mutation ocelliless. The 

(Fig. 43). ocelliless mutation is associated with a 

Transcripts from the 6.0-kb Eco RI chromosomal inversion of 7F1,2-8A1,2, 

fragment are transcribed from the same which disrupts the X-chromosome cho- 

strand as those for s36 and s38, yet they rion gene cluster. This disruption results 

exhibit different accumulation patterns, in a loss of amplification for the portion 

Northern blots of staged RNA hybrid- of the cluster remaining at 7F and the 

ized to the 6.0 reveal a progressive de- 
crease in transcript size over 
developmental time (Fig. 43) and a par- 
allel increase in the amount of each mes- 

loss of a transcript of 950 bases (Fig. 44) 
whose gene is interrupted by the distal 
inversion breakpoint. Last year, we 
demonstrated that in ocelliless, both 

sage not selected by oligo-dT columns halves of the gene cluster are expressed 

(Fig. 43). This suggests that these tran- at the proper developmental stages in 

scripts are becoming smaller as a result spite of the loss of amplification for the 

of a decrease in poly(A) tail length. If genes remaining at 7F and the relocation 

true, this would also suggest that gene of the other half of the cluster at 8A( Year 

transcription does not continue through- Books 80, 81). This led us to conclude 

out the entire period of time wherein that amplification plays a primarily quan- 

complementary RNA is present. titative role in gene expression and that 








O *H <N| CO *f* 

^pm| ^^^1 I**! VNI V"! * 

Fig. 44. Northern blots of poly A + stages egg chamber RNA from ocelliless (left panel) and 
ocelliless transformed with the 6.0-kb Eco RI fragment (right panel). 

/. Levine 


the sequences necessary for proper tern- to the third chromosome amplification or- 

poral control are located close to the genes, igin and then transform it into ocelliless 

We postulated that the loss of the 950- flies. These experiments with ocelliless 

base transcript could be the cause of ster- may determine whether amplification is 

ility in ocelliless. This year, we at- necessary for the entire X-chromosome 

tempted to restore the fertility of ocelliless gene cluster or if it is only required for 

by supplying the 950-base transcript via the major chorion genes s36 and s38. 
P-factor transformation. 

Using transformation, a strain con- 
taining the 6.0-kb Eco RI fragment as Sequence Organization of 
part of a rosy + transposon was con- Chorion Gene Regions 
structed (see de Cicco, this Report). This 
fragment encompasses the 950 gene in 
its entirety. This line, 63225, was crossed Many of the structural characteristics 
to ocelliless heterozygous ry~ flies. Af- of a gene can be determined from a 
ter several generations, female progeny knowledge of relevant DNA sequences, 
that were homozygous for ocelliless and for example, by comparing cDNA and 
ry + (the result of the P-factor 6.0-kb in- corresponding genomic regions. DNA 
sertion) were mated and tested for fer- primary structure also yields useful in- 
tility. The 6.0-kb insertion failed to revert formation for the construction of mutant 
the sterility. Northern analysis of RNA genes and provides a basis for inter- 
from developmentally staged egg cham- preting studies of gene function. Ad- 
bers probed with the 6.0-kb Eco RI frag- vances in sequencing techniques have 
ment shows the presence of the 950-base made practical our goal of determining 
transcript in the transformed ocelliless the complete DNA sequence of both ma- 
line (Fig. 44). These experiments also jor chorion gene clusters. This year, we 
show that the 950-base RNA made from sequenced cDNAs derived from all four 
the insertion is produced with the normal major chorion genes. About 3 kb of DNA 
developmental profile. sequence derived from the 3.7-kb Sal I 

In spite of the restoration of the 950- fragment containing the sl5-l and sl8-l 

base RNA at the correct time in the ocel- genes was also determined. Analysis of 

liless transformed line, the sterility of the sequence has provided information 

the mutant was not reverted. It is pos- on chorion protein structure and on gene 

sible that the transcripts produced from structure, including the location of pu- 

the 6.0-kb fragment are not produced in tative initiation sites, intron locations, and 

sufficient amounts to restore fertility be- polyadenylation sites. Furthermore, an 

cause this fragment failed to amplify (see origin of amplification has been localized 

de Cicco, this Report). To test if this is to this 3.7-kb fragment (see de Cicco, this 

truly an amplification defect, we will at- Report). We will present a summary of 

tempt to fuse the 6.0-kb Eco RI fragment this information in a later Year Book. 


S. Ward, D. J. Burke, B. Kirschner, and E. Hogan 

Our laboratory continues to investi- matode Caenorhabditis elegans (Year 

gate the mechanisms whereby genes con- Books 77-81). During the past year, we 

trol the morphology of cells. Our approach have continued the molecular analysis of 

is to study mutations that alter the mor- the genes encoding the major protein in 

phology of the spermatozoa of the ne- C. elegans sperm. Burke reports that the 


major sperm protein (MSP) is encoded unusual shape. Over the past several 
by a multigene family with more than years, I have attempted to characterize 
thirty genes not tightly linked. Burke and abundant sperm-specific proteins and the 
Ward report that although the MSP genes that encode them. Some of these 
mRNA is homogenous in size, the major proteins should be the products of genes 
sperm protein is not. Instead, it is a fam- that we have already identified by mu- 
ily of proteins with three different iso- tations that alter sperm morphology, 
electric variants. Because each variant is In Year Book 81 , 202, I outlined a gen- 
a primary translation product without eral strategy for identifying the products 
detectable post-translational modifica- of specific mutated genes by direct iso- 
tions, at least three MSP genes must be lation of the cloned genes. Once identi- 
expressed. When the MSP mRNAs are fied, a cloned fragment of DNA can be 
copied into cDNAs and sequenced, we mapped relative to standard genetic 
find that even more genes are ex- markers, and sperm-specific mutants that 
pressed — at least nine. Mutations alter- map nearby can be tested for alteration 
ing these proteins have not yet been of the protein, the mRNA, or the DNA. 
identified. I reported the isolation of genomic clones 
In addition to the molecular analysis corresponding to the major sperm pro- 
of sperm-specific genes, we have ex- tein (MSP), which constitutes 15-20% of 
tended our studies of the mechanism of the total cell protein. The MSP is seg- 
protein segregation to specific parts of regated into the spermatids during sper- 
spermatocytes during spermatogenesis, matogenesis and is concentrated in the 
Ward shows that, in contrast to the pseudopods of spermatozoa. The abun- 
abundant sperm proteins — MSPs and dance and localization of this protein sug- 
spots 10 and 11 {Year Book 81, 207), which gest that it is one of the proteins involved 
are segregated preferentially to the in shape determination and perhaps sperm 
spermatids — the proteins tubulin and ac- motility. 

tin are specifically excluded from the I have continued analysis of the MSP 

spermatid. The mechanism for tubulin genes to determine how many copies of 

exclusion appears to be that the tubulin the gene there are and how these copies 

remains assembled as the meiotic spindle are arranged in the genome. Eleven re- 

which is segregated intact to the residual combinant phages containing genomic 

body, leaving the spermatid devoid of copies of MSP genes have been exam- 

microtubules. ined. Restriction enzyme mapping of these 

phage genes showed no overlaps in the 
restriction maps, suggesting that each 
phage contained a different copy of an 
MSP gene. This was confirmed by locat- 

Molecular Analysis of the Major ing the MSP DNA in each recombinant 

Sperm Protein Genes phage using an MSP cDNA as a probe 

(Fig. 45). The MSP cDNA hybridized to 
a single fragment of DNA after restric- 

The structural proteins comprising the tion enzyme digestion of total phage DNA, 

cytoskeleton of most eukaryotic cells — showing that only one MSP gene was 

actin and tubulin — are virtually absent present in each phage and that the genes 

from C. elegans spermatozoa. Other pro- each came from different regions of the 

teins must have usurped the shape-de- worm genome. The same result was ob- 

termining functions that these proteins tained using three different enzymes; no 

usually perform. Such proteins should be multiple MSP genes were found. Thus, 

abundant because they form the cyto- the MSP genes must be dispersed 

skeleton of the sperm, and they should throughout the genome and separated by 

be sperm-specific because of the sperm's at least 16 kb of DNA. In addition, since 



f g h 



Fig. 45. Mapping the MSP cDNA on Individual Genomic Clones. Phage DNA was digested with Eco 
RI and fractionated on a 0.7% agarose gel by electrophoresis. The gel was stained with ethidium bromide 
(A) prior to transfer to nitrocellulose. The MSP cDNA was labeled by nick translation, denatured, and 
hybridized to the nitrocellulose and autoradiographed (B). (a) XBA2, (b) \BA5, (c) A.BA8, (d) XBA9, (e) 
XBA14, (f) \BA15, (g) \BA16, (h) XBA18. The mobility of fragments for Hind Ill-digested \cI857 DNA 
is shown on the left. 

none of the MSP genes contain internal struct a DNA probe that contained only 
restriction sites for three different en- coding sequences. Nematode DNA was 
zymes, the genes are not interrupted by digested with various restriction en- 
large intervening sequences. zymes, fractionated by agarose electro- 
To determine the precise number of phoresis, transferred to nitrocellulose, 
MSP genes in the genome, a probe that and hybridized with the MSP coding se- 
contained only DNA sequences from quence probe. The results (Fig. 46) show 
within the coding region of an MSP gene that approximately thirty different frag- 
was needed to ensure that only MSP genes ments hybridized. The same result was 
would hybridize. This was obtained by obtained with three additional enzymes, 
first purifying MSP and having Ned Sie- The fragments ranged in size from 1 to 
gel of Monsanto sequence part of the pro- 25 kb and varied in intensity. A recon- 
tein. Then, an MSP cDNA clone was struction experiment demonstrated that 
subcloned into bacteriophage M13 and the fifteen strongly hybridizing bands 
sequenced by the dideoxy method with corresponded to single copies of the MSP 
the help of Kevin O'Hare. The protein gene. The weaker bands probably rep- 
sequence was aligned to the DNA se- resented MSP genes that had diverged 
quence, indicating that the MSP is 127 in sequence, since the MSP genes that 
amino acids long with a molecular weight had been cloned did not contain restric- 
of 14,250. Knowing the coding region of tion sites for the enzymes used for diges- 
the MSP cDNA made it possible to con- tion. 



a b c 

These results show that the MSP is 
encoded in the genome by a large mul- 
tigene family with at least thirty mem- 
bers. These genes are not closely linked, 
and the coding sequences are not inter- 
rupted by large intervening sequences. 
Some of the members of the family have 
diverged in sequence. 


Fig. 46. Hybridization of the cDNA to C. ele- 
gans genomic DNA. 2 \xg of DNA from adult N2 
hermaphrodites was digested with restriction en- 
zymes, fractionated on 0.7% agarose gels, and 
transferred to nitrocellulose. The probe was pre- 
pared from an MSP cDNA subcloned in phage M13. 
(a) Bgl II-digested DNA; (b)Eco Rl-digested DNA; 
(c) Hind Ill-digested DNA. 

There Are Multiple MSP Gene 

D. J. Burke and S. Ward 

Since many MSP genes are dispersed 
in the genome, we would expect their 
sequences to have diverged, producing 
multiple products. To determine if the 
MSP mRNA is heterogeneous, serial two- 
fold dilutions of male and hermaphrodite 
RNA were fractionated by agarose gel 
electrophoresis, transferred to nitrocel- 
lulose, and hybridized with an MSP probe. 
A probe for actin was included at the 
same time as an internal control. We found 
that the MSP mRNA from both males 
and hermaphrodites is homogeneous in 
size with no indication of multiple RNA 
products (Fig. 47). In contrast, the actin 
probe reveals three different actin 
mRNAs that appear identical in males 
and hermaphrodites. This shows that al- 
though there are only four actin genes in 
the genome, they produce three distinct 
messages, whereas the thirty MSP genes 
apparently produce a single-sized mRNA. 

A single-sized mRNA does not assure 
the expression of only one MSP gene. If 
several genes were expressed, multiple 
mRNAs might exist with differing se- 
quences, and there could be more than 
one protein form of MSP. If the proteins 
differed in charge, they should be sepa- 
rable by isoelectric focusing. We used an- 
tibodies prepared to spots cut out of both 
two-dimensional gels (Year Book 81) and 
high-resolution isoelectric focusing gels 
to show the existence of multiple MSP's. 
Antibodies raised to a protein with the 
same size as the MSP but with a different 
isoelectric point (pi = 7.1) (spot 12 in fig. 
57, Year Book 81, 207) were found to 





Fig. 47. The MSP mRNA is homogeneous in 
size, (a) 2 (xg of male poly A + RNA or (b) her- 
maphrodite poly A + RNA, and serial twofold di- 
lutions were fractionated by agarose electrophoresis 
and transferred to nitrocellulose. The MSP mRNA 
was detected using p\BA2,2.0 as a probe. A Dro- 
sophila actin probe, pDMA 2 , kindly provided by 
Eric Fyrberg, is included as a comparison. To max- 
imize detection of the actin mRNA (A), the filters 
were washed under conditions of low stringency. 

cross-react with the major MSP spot (pi 
= 8.7-8.8). This was shown by trans- 
ferring the sperm proteins fractionated 
on two-dimensional gels to nitrocellulose 
sheets, and then adding the antibody to 
see which proteins are antigens. Figure 
48 shows that antibodies directed to the 
major MSP spot react with "spot 12"; 
antibodies against "spot 12" cross-react 
with the major MSP. Thus, this spot is 
antigenically related to the major MSP 
spot and has been renamed MSP (7.1). 
When the major MSP spot is fractionated 
on narrow-range isoelectric focusing gels, 
it separates into two overlapping spots 
of pi = 8.7 and pi = 8.8 (Fig. 49). There- 
fore, at least three MSP forms are sep- 
arable by isoelectric focusing. 

Proteins can be modified in many ways 
after they are translated, so that a poly- 
peptide that is the product of a single 

gene can give multiple spots after iso- 
electric focusing. To determine whether 
the multiple MSP's we detected arose 
from post-translational modifications, we 
first asked if any of them were phospho- 
rylated. We did this by analyzing 32 P- 
labeled sperm proteins. Less than 0.02 
molecules of 32 P was found to be asso- 
ciated with MSP. We next asked if any 
MSP's were glycosylated by using a sen- 
sitive periodate-silver stain. We de- 
tected no glycosylation in MSP's, whereas 
many other, less-abundant proteins were 
found to be glycosylated. Finally, we 
compared the pattern of MSP's found by 
in vitro translation of male poly A + RNA. 
All three MSP forms were found to co- 
migrate exactly with the MSP's found in 
mature sperm, whether they were syn- 
thesized by in vitro translations of RNA 
in reticulocyte lysates that do not carry 
out post-translational modifications or 
were synthesized in frog oocytes that do 
modify translated proteins (Fig. 49 b-d). 
This was true for both male and her- 
maphrodite RNA. 

These results show that the three forms 
of MSP separated by isoelectric focusing 
are each primary translation products and, 
therefore, must be encoded by different 
genes. It is possible that the large, poorly 
resolved MSP's (8.7 and 8.8) are them- 
selves composed of multiple polypeptides 
identical in charge, so that there may be 
more than three different MSP polypep- 

In order to determine the best way to 
identify mutations in the MSP genes, it 
is necessary to know how many different 
forms of MSP exist and how many of the 
MSP genes are active. If there are many 
active genes all producing similar or 
identical proteins, then mutations in any 
one of them might not give a recessive 
mutant phenotype. Since the resolution 
of proteins on gels is not sufficient to de- 
termine the number of classes, we de- 
cided to clone and sequence a large 
number of DNA copies of MSP mRNA 
to see how many different sequences were 
present. Each different sequence should 
represent the product of a different gene. 











Fig. 48. Antisera specificity. All panels are immunostains of sperm proteins fractionated on poly- 
acrylamide gels (a-d, one-dimensional; e-f, two-dimensional, showing only the MSP region) and trans- 
ferred electrophoretically to nitrocellulose, (a) Anti-MSP serum 1:100, 2.5-h exposure, (b) Pre-immune 
serum from the rabbit that produced the anti-MSP, 1:50, 16-h exposure, (c) Anti-7.1 serum 1:100, 11-h 
exposure. The apparent high-molecular-weight band is an artifact. It is found in several rabbit sera that 
do not react with sperm proteins, and it extends all the way across some gels as if caused by something 
in the running buffer, not the sample, (d) Pre-immune serum from the rabbit that produced the anti- 
7.1, 1:50, 11-h exposure. Longer exposures were also blank, (e) Anti-MSP serum 1:200, 2-h exposure, 
(f) Anti-7.1 serum 1:100, 6-h exposure. 

We constructed a cDNA library to male 
mRNA, and isolated and sequenced fif- 
teen MSP cDNAs. In comparing the se- 
quences, we found that at least nine 
different genes were expressed. The un- 
translated portions of the mRNA are 
highly diverged, while the protein coding 
sequences are conserved. Some of the 
proteins differ in their charged amino 
acids, and these differences could explain 
the protein heterogeneity seen on iso- 
electric focusing gels. 

These results show that at least nine 
of the MSP genes in the genome produce 
mRNA that is homogeneous in size but 
varied in sequence. At least three dif- 
ferent proteins are produced. If these 

proteins are functionally interchangeable 
so that a mutation altering one of them 
reduces the total level of MSP only 
slightly, then mutations in MSP genes 
will not be in our present collection of 
recessive mutations. It may be necessary 
to analyze dominant mutations or dele- 
tions to determine how the MSP's func- 
tion in determining sperm morphology 
and motility. 

Asymmetric Segregation of 
Proteins during Spermatogenesis 

S. Ward 

After completing the nuclear events of 
meiosis, the haploid nuclei segregate to 



Fig. 49. Narrow range two-dimensional gels of 
MSP's. All two-dimensional gels are with pH 8.5- 
10 ampholytes. All gels contained 1-2 x 10 7 sperm 
and were stained to identify the MSP spots before 
autoradiography. Only the region around the MSP 
is shown, (a) 35 S-labeled sperm. Note that proteins 
other than MSP give symmetrical spots, (b) La- 
beled proteins from male poly A + RNA translated 
in a rabbit reticulocyte lysate. (c) Protein immuno- 
precipitated by anti-MSP serum from male poly A + 
RNA translated in frog oocytes, (d) Labeled pro- 
teins from hermaphrodite poly A + RNA extracted 
during spermatogenesis and translated in a rabbit 
reticulocyte lysate. 

the perimeter of the spermatocyte (Year 
Book 79, 40, fig. 25). The spermatids form 
around these nuclei by the accumulation 
of sperm organelles. The sperm-specific 
MSP's and the proteins "10 and 11" ac- 
cumulate in these spermatids, presum- 
ably by their assembly into organelles 
known as fibrous bodies (Year Book 80, 
149; 81, 207). We have shown previously 
that after the spermatids have completed 
maturation and separated from the re- 
sidual body, they contain almost no actin 
or tubulin (Year Books 79, 81). Yet, both 
actin and tubulin are abundant in the 
spermatocytes. How does the spermatid 
avoid accumulating these two abundant 
structural proteins? I have used a mono- 
clonal antibody specific for tubulin, and 
a fluorescent probe, rhodamine phall- 
oidin, specific for actin to follow the dis- 
tribution of these two proteins during 

The maturation of spermatocytes pro- 
ceeds from the distal to the proximal end 
of the testis. A fragment of testis con- 
taining spermatocytes developing to form 
spermatids was fixed, permeabilized, 
treated with anti-tubulin, and then stained 
with fluorescently tagged anti-antibody. 
Figure 50 shows that tubulin is found in 
a wispy pattern throughout the cyto- 
plasm of primary spermatocytes, that it 
is formed into spindles in secondary sper- 
matocytes, but that it is undetectable in 
spermatids. This result confirms the bio- 
chemical and electron microscopic evi- 
dence that almost no tubulin and no 
microtubules are detected in spermatids 
or spermatozoa. When a similar testis 
was dissected further to release individ- 
ual spermatocytes and then prepared for 
anti-tubulin immunofluorescence, it was 
again found that spermatids have no de- 
tectable tubulin, but that tubulin is abun- 
dant in the smooth-looking residual bodies 
(RB) and remains organized in spindles 
(Fig. 51). When additional spermato- 
cytes were examined during the process 
of spermatid budding, it was seen that 
the spindle that organized the final meiotic 
segregation of chromosomes to the sper- 
matid remained assembled after the com- 
pletion of meiosis. It is this spindle that 
is segregated intact to the residual body. 
This has been confirmed by electron mi- 
croscopic examination of residual bodies, 
which showed microtubules radiating out 
from the point of separation of the sper- 

Analysis of the actin distribution using 
rhodamine phalloidin reveals a similar 
pattern: Actin is found only in the resid- 
ual body concentrated at the site of sper- 
matid budding. This actin must be 
organized into microfilaments because 
phalloidin only binds to actin in the fila- 
mentous form. These results show that 
actin and tubulin are segregated from the 
spermatid while assembled in their fila- 
mentous forms. Thus, the segregation of 
sperm proteins to the spermatid, and the 
segregation of tubulin and actin away from 
the spermatid, is accomplished while these 



1° Spermatocytes 


«fO L1L i g , o^ 

Spe rma tids 

Fig. 50. Testis fragment stained with anti-tubulin. (a) Nomarski optics, (b) Fluorescence. 

proteins are assembled into specific fi- this results in a spermatid and a residual 
brous structures. The cell appears to body with different protein composi- 
separate these structures as intact units; tions. 



D. D. Brown, I. J. Jackson, E. Jordan, M. Schlissel, D. Setzer, D. R. Smith, and W. L. Taylor 

Our studies of 5S RNA gene control 
emphasize the biochemical and biophys- 
ical details of gene transcription. We be- 
lieve that knowledge of how the various 
factors (presumably proteins) interact 
with the gene and guide RNA polymer- 
ase to initiate synthesis at the correct 
nucleotide will be essential for our un- 
derstanding of how this gene is con- 
trolled. D. Smith and I. Jackson have 

made the important discovery that one 
of the protein factors (variously termed 
5S specific transcription factor, factor A, 
or 40-kDa factor) required for formation 
of a transcription complex on a 5S RNA 
gene and for accurate initiation of tran- 
scription, can be cleaved by proteases 
into domains that have only partial func- 
tion. This new information will make the 
complete structure of this protein even 



Fig. 51. Dissected testis stained with anti-tubulin. (a) Nomarski optics. The round cells on the left 
are late-primary and secondary spermatocytes. RB = residual bodies, (b) Fluorescence. 

more informative, and we are continuing 
our efforts to clone the gene that encodes 
the factor (Taylor and Jackson, below). 
The transcription complex assembled in 
vitro can be studied using a novel method, 
devised by D. Setzer, in which the DNA 
is immobilized on cellulose. Setzer has 
learned something about the order of ad- 
dition of factors to form the complex and 
the stability of the complex. In parallel 
studies, M. Schlissel is investigating the 
5S RNA genes in their chromatin states. 
For the first time, active and repressed 
genes in chromatin are being analyzed 
using as an assay their ability to tran- 
scribe RNA faithfully. 

Domains of the 40-kDa 
Transcription Factor 

D. Smith and I. Jackson 

The transcription factor complexed with 
5S RNA in the form of 7S particles can 
be cleaved with proteolytic enzymes into 
a number of smaller fragments. Among 
these are two fragments, 30 kDa and 20 
kDa in molecular weight, which under 
the appropriate conditions can be gen- 
erated as the major products of proteo- 
lytic digestion. These fragments (and also 
the 40-kDa protein) were purified on a 
Bio Rex-70 column to remove the pro- 
teolytic enzyme and other contaminating 



proteins. The resulting preparations were 
judged to be greater than 95% pure, based 
on SDS gel electrophoresis (Fig. 52). 

Each of the purified proteins was tested 
for its ability to protect 5S DNA from 
DNase I digestion in footprint experi- 
ments (Fig. 53a). The 40-kDa factor pro- 
tects a region of about 50 nucleotides in 
the center of the 5S RNA gene (the con- 
trol region). The pattern produced with 
the 30-kDa fragment was essentially the 
same except that cleavage at hypersen- 
sitive sites near the 5' end was reduced. 
The 20-kDa fragment protected only the 
3' half of the control region. This is in- 
teresting in light of transcription com- 
petition (Wormington, Year Book 79, 86) 
and chemical protection (Sakonju, Year 
Book 80, 207) experiments which show 
that the factor binds most tightly to the 
3' end of the control region, and that vir- 
tually all of the essential contacts be- 


40 kd 

30 kc! 

20 kd 

Fig. 52. SDS-polyacrylamide gel electropho- 
resis of purified transcription-factor preparations. 
Lane 1, 2.5 |xg of 7S particle, crude glycerol gra- 
dient fraction; lane 2, 1.1 fxg of 7S particle, purified 
on DEAE cellulose; lane 3, 1 fxg of 40-kDa tran- 
scription factor, further purified on Bio Rex-70; 
lane 4, 1 (xg of 30-kDa fragment produced by papain 
digestion of 7S particle followed by purification on 
Bio Rex-70; lane 5, 1 |xg of 20-kDa fragment pro- 
duced by trypsin digestion of 7S particle followed 
by purification on Bio Rex-70. 


I 2 3 


Fig. 53. (A) DN Ase I protection of 5S DNA by 
purified factor preparations. Factors were incu- 
bated with a linearized, end-labeled plasmid con- 
taining a somatic 5S RNA gene at a 20:1 protein- 
to-DNA ratio followed by DNAse I digestion of the 
DNA and electrophoresis on a denaturing poly- 
acrylamide gel. Lane 1, no added protein; lane 2, 
40-kDa factor; lane 3, 30-kDa papain fragment; lane 
4, 20-kDa trypsin fragment. (B) Transcription of 
5S DNA in the presence of various factor prepa- 
rations. Factors were incubated with a plasmid 
containing a somatic 5S RNA gene at a 2:1 protein- 
to-DNA ratio, followed by additions of RNA po- 
lymerase III, B, and C phosphocellulose fractions 
from an X. laevis somatic cell extract (see Setzer, 
this Report) and nucleoside triphosphates. 5S RNA 
synthesized in the reaction was extracted and elec- 
trophoresed on a denaturing polyacrylamide gel. 
Lane 1, 40-kDa factor; lane 2, 30-kDa papain frag- 
ment; lane 3, 20-kDa trypsin fragment. 

tween the protein and specific sites on 
the DNA also occur in this region. 

The proteins were also used to pro- 
gram transcription on a somatic 5S RNA 
gene in a fractionated transcription sys- 
tem that was completely dependent upon 
added factor (purified RNA polymerase 
III, plus the B and C phosphocellulose 
fractions from a somatic cell extract; see 
Setzer, this Report) (Fig. 53b). The 30- 
kDa fragment, which gave a complete 
DNase I footprint, was able to stimulate 
5S RNA transcription, but only at 20- 



30% of the level of the 40-kDa factor. The 
20-kDa fragment, which gave only half a 
footprint, was completely inactive for 

Transcriptions were done at a number 
of protein-to-DNA ratios, and the amount 
of 5S RNA synthesized was quantitated 
(Fig. 54). With both 40-kDa and 30-kDa 
proteins, maximal stimulation of tran- 
scription occurred at a ratio of approxi- 
mately 1 molecule of protein per DNA 
molecule. This suggests that only a single 
molecule of factor is required to bind to 
the control region to activate the genes 
for transcription. Quantitation of DNase 
footprint experiments at a 1:1 protein to 
DNA ratio gave a value of 60-70% pro- 
tection, also consistent with a stoichi- 
ometry of 1 molecule per gene. 

We can now construct a more detailed 
model for the interaction of the 40-kDa 
factor with 5S DNA and other molecules 
in a transcription complex. The 20-kDa 
core of the factor molecule contains the 
primary sequence recognition site for the 
3' end of the control region, and there- 
fore must be involved in the sequence- 
specific recognition of all types of 5S RNA 

genes. The 30-kDa fragment contains a 
second domain required for binding to 
the 5' end of the control region and might 
be involved in the discrimination be- 
tween oocyte and somatic 5S RNA genes 
(since the important nucleotide differ- 
ences between these genes occur in the 
5' half of the control region). A third do- 
main in the complete 40-kDa protein is 
required for efficient transcription initi- 
ation and is therefore likely to participate 
in interactions with other molecules in a 
transcription complex, such as factors B 
and C, or RNA polymerase III. Since 
these interactions might be common to 
all RNA polymerase Ill-transcribed 
genes, perhaps other specific factors, such 
as those required for transcription of 
tRNA genes, will contain a homologous 

In Vitro Labeling of the 40-kDa 
Transcription Factor 

D. Smith 

We would like to identify the proteins 
that interact with the 40-kDa factor in a 

















A ^- 



7 */ 






1.0 2.0 

Protein to DNA ratio (molar) 


Fig. 54. Dependence of transcription on factor concentration. Transcriptions were carried out as in 
Fig. 53 except that the amount of factor was varied. The amount of 5S RNA synthesized (measured by 
counting the radioactive bands on the gel) is plotted against the protein-to-DNA ratio in the reaction 
for the 40-kDa factor (circles) or the 30-kDa papain fragment (triangles). 


transcription complex. One approach is for their ability to restore transcription 
to label radioactively the factor molecule, of 5S DN A and tDNA templates in vitro, 
and then to form transcription complexes Transcriptional activity on these tem- 
with a labeled protein and chemically plates could be reconstituted by adding 
cross-link the protein in the complex. The one of these fractions to the material that 
molecules could then be separated on a did not bind to factor-Sepharose in low 
gel and the cross-linked molecules easily salt. The proteins in this fraction were 
identified above the background of un- found to be very complex, indicating a 
labeled proteins. An extension of this high degree of nonspecific binding of pro- 
technique could be used to identify which teins to the factor-Sepharose. Attempts 
domains of the 40-kDa factor are in- to lower this background were unsuc- 
volved in these interactions. cessful. This approach — of isolating com- 
For technical reasons, iodine-125 is the ponents of 5S transcription complexes — 
most convenient of the isotopes available should prove to be useful in conjunction 
for protein labeling, but standard meth- with other fractionation procedures, 
ods for oxidative iodination of tyrosine 
residues inactivate the transcription fac- 
tor. However, the protein can be labeled Isolation of the Gene Encoding 
with 125 I-Bolton Hunter reagent, which THE 5 S-Specific Transcription 
reacts with lysine residues, while still Factor 
maintaining its ability to bind specifically 

to 5S RNA W- Taylor and I. Jackson, in collaboration with 

A. Kumar and N. Siegel 

Our previous attempts to isolate the 

Use of Immobilized 5S-Specific gene encoding the 5S-specific transcrip- 

Transcription Factor for the tion factor centered around the use of a 

Isolation of Components of synthetic 17-nucleotide probe to identify 

Transcription Complexes cDNA clones for the 5S specific factor. 

The sequence of this probe was deduced 
from the first five amino acids of a cyan- 

W. Taylor 

In attempting to isolate the compo- ogen bromide peptide (CB-1) of the 5S- 

nents of 5S DNA transcription com- specific factor (Jackson, Year Book 81, 

plexes, several different schemes have 222). Several cDNA clones were isolated 

been used. (Shastry et al. , /. Biol. Chem. that hybridized to this oligonucleotide 

257, 12979, 1982; Setzer, Year Book 81, probe, but none of them had the pre- 

216). In my own efforts to purify these dieted DNA sequence. We re-isolated the 

components, I sought to use interactions cyanogen bromide peptide CB-1, and it 

of the purified specific transcription fac- was sequenced again by N. Siegel of the 

tor with other components of the tran- Monsanto Corporation. The sequence ob- 

scription complex. tained differed from that previously re- 

The factor was immobilized on Se- ported for CB-1 (Jackson, Year Book 81) 

pharose, and the factor-Sepharose was in two positions. The most important dif- 

assayed for its ability to bind specifically ference was the presence of a lysine rather 

5S DNA using a footprint assay (Sak- than an alanine residue at position 1 of 

onju, Year Book 79, 80). The results in- the CB-1 peptide. We believe that the 

dicated that the immobilized factor retains most recent sequence is correct, which 

specificity for 5S DNA. This factor-Se- means that the probe sequence is incor- 

pharose was then interacted with tran- rect. Recently, we decided to use an oli- 

scriptionally competent extracts that had gonucleotide that encodes a different 

been previously depleted of the 5S spe- portion of the CB-1 peptide sequence, 

cific factor. Fractions were eluted from Synthesis of this 19-nucleotide probe with 

this column with salt and then assayed 32 ambiguities is under way. 



Another approach to isolate the gene 
involves the use of cDNA expression li- 
braries prepared in the plasmid vectors 
pUC8 and pUC9 (Helfman et al., Proc. 
Nat. Acad. Sci. USA 80, 31, 1983). Poly 
A-containing RNA, prepared from im- 
mature X. laevis ovaries, was shown to 
contain sequences corresponding to the 
5S-specific transcription factor by in vi- 
tro translation and precipitation of the 
translation products with antibodies to 
the 5S-specific transcription factor (Fig. 
55). The cDNA expression libraries were 
prepared using this RNA preparation and 
screened in duplicate using antibodies to 
the 5S specific factor and pre-immune 
serum. No colonies were detected which 
interacted specifically with the factor an- 
tibodies and not with the pre-immune 

Analysis of the 5S rRNA 
Transcription Complex 

D. Setzer 

Bogenhagen et al. (Cell 28, 413, 1982) 
have presented evidence that 5S rRNA 
is synthesized in germinal vesicle ex- 
tracts from 5S DNA templates that are 
stably associated with required tran- 
scription factors through multiple rounds 
of transcription. These stable complexes 
have been called transcription com- 
plexes. We previously (Year Book 81) de- 
scribed the use of 5S genes covalently 
attached to DBM-cellulose as templates 
for the assembly of transcription com- 
plexes in vitro, and the recovery of these 
complexes from crude extracts by pel- 
leting and extensive washing of the cel- 
lulose-bound template. We have since 
analyzed the stability and formation of 
complexes prepared in this way and have 
reached the following conclusions. 

(1) RNA polymerase III is the least 
tightly bound component of the tran- 
scription complex and can be removed by 
washing with 0.5 M NaCl or by tran- 
scribing in the presence of excess non- 
specific competitor DNA. Purified RNA 
polymerase III can be added to com- 

A o C D E 

Fig. 55. Immunoprecipitation of 5S-specific 
transcription factor labeled with 35 S cysteine. Im- 
mature Xenopus ovary proteins were labeled with 
35 S cysteine in whole ovaries cultured in oocyte 
Ringer's solution (A and B). These same proteins 
were also labeled in vitro in a wheat germ trans- 
lation system to which (C) 0.5 |xg of poly A-con- 
taining RNA or (D and E) 10 (xg of total RNA from 
immature Xenopus ovaries was added. This labeled 
material was then immunoprecipitated either with 
antibody to the 5S-specific transcription factor (B, 
C, and E) or with non-immune serum (A and D). 



plexes lacking polymerase to reconsti- 
tute transcriptionally active complexes. 

(2) All required components of the 
transcription complex except RNA po- 
lymerase III remain bound in salt con- 
centrations as high as 0.5-1.0 M NaCl. 
In contrast, concentrations as low as 0.25 
M NaCl inhibit formation of complexes 
that subsequently can be transcribed by 
purified RNA polymerase III in low salt. 

Shastry et al. (J. Biol. Chem. 257, 
12979, 1982) have shown that crude ex- 
tracts prepared from Xenopus cultured 
cells can be fractionated on phosphocel- 
lulose columns into two fractions (des- 
ignated B and C) that are capable of 
supporting 5S rRNA synthesis in vitro 
in combination with the 40-kDa 5S-spe- 
cific positive transcription factor (factor 
A) and RNA polymerase III. Beginning 
with crude extracts prepared according 
to Manley et al. (Proc. Nat. Acad. Sci. 
USA 77, 3855, 1980), we have duplicated 
these results. 

It seemed possible that we might be 
able to generate and recover stable par- 
tial transcription complexes using these 
components. To test this, we attempted 
to form functional complexes on DNA- 
cellulose templates (pXbsl-cellulose) by 
adding purified factor A, fraction B, or 
fraction C separately and in succession 
while washing away unbound material 
between steps. Finally, purified RNA 
polymerase III was added, and the pres- 
ence of active transcription complexes was 
assayed by its ability to synthesize 5S 
rRNA from radioactive ribonucleoside 
triphosphates. As seen in Fig. 56, only 
two of the six possible orders of com- 
ponents resulted in the production of an 
active transcription complex. Either fac- 
tor A or fraction C was capable of stably 
interacting directly with the DNA tem- 
plate, but fraction B stably interacted 
with the gene only after it was first ex- 
posed to factor A and fraction C. We think 
it is probable that this represents the 
order of interaction of components in the 
formation of a transcription complex. 

5S genes added to transcribing ex- 
tracts did not instantaneously become 

Fig. 56. Order of interaction of transcription 
factors needed to produce a 5S transcription com- 
plex. pXbsl-cellulose at a concentration of 5 |xg/ml 
DNA was separately and successively incubated 
for 45 min at 20°C with the 40-kDa transcription 
factor A, fraction B, or fraction C in the order given 
below. Finally, purified RNA polymerase III was 
added along with nucleoside triphosphate precur- 
sors including [a- 32 P]GTP and incubated for 10 min. 
Purified radioactive RNA was electrophoresed on 
a 10% polyacrylamide gel containing 8M urea. The 
order of addition of the various components was (a) 
C, B, A; (b) C, A, B; (c) B, C, A; (d) B, A, C; (e) 
A, C, B; (f) A, B, C. 


transcriptionally active. On DNA-cellu- can be activated by washing the template 
lose templates, the lag period preceding in 0.6 M NaCl. Complete transcribing ex- 
the maximal transcriptional rate was tracts (oocyte nuclear supernate) cata- 
about 45 minutes in both oocyte nuclear lyze transcription of the de-repressed 
extracts and somatic cell extracts. We oocyte-type 5S RNA genes. When such 
analyzed the kinetics of interaction of de-repressed chromatin is transcribed 
factor A, fraction C, fraction B, and RNA with purified RNA polymerase, how- 
polymerase III with the appropriate sub- ever, a high background of nonspecific 
strates, beginning with naked DNA and transcripts obscures 5S RNA. It is known, 
assuming an order of interaction of A, C, though, that transcription complexes 
B, pol III. Fraction B was responsible formed on cloned genes are stable to ionic 
for the rate-limiting step in transcription strengths up to 1.0 M (Setzer, this Re- 
complex formation and accounted for the port). One hypothesis explaining this re- 
lag seen in complete extracts. Factor A, suit is that washed chromatin presents 
fraction C, and RNA polymerase III all the purified enzyme with a template highly 
exhibited maximal activity within a few accessible to nonspecific initiation events, 
seconds or minutes after their addition, thereby competing with the preformed 

transcription complexes for limiting 
amounts of RNA polymerase. 

Chromatin prepared from the eryth- 

The State of 5S RNA Genes in rocytes of individual frogs shows differ- 

Chromatin ing degrees of developmental regulation. 

,, a ,,. , Figure 57 presents the results of tran- 

scnption ot three different chromatin 

The 5S rRNA synthesized in somatic preparations with three different kinds 
cells is the product of the 800 somatic- of extracts: (1) oocyte nuclear extract, 
type 5S RNA genes in the Xenopus gen- (2) the same extract depleted of the 5S- 
ome, while 40,000 copies of the oocyte- specific transcription factor, and (3) with 
type gene are transcriptionally inactive, purified RNA polymerase III. Com- 
Wormington has demonstrated (Year pletely regulated preparations of chro- 
Books 80, 81) that this developmental matin synthesize the same amount and 
regulation of 5S RNA gene expression is kinds of 5S RNA with the three types of 
controlled at the level of chromatin in transcription reactions (lanes 1, 2, 3). 
both embryonic and adult somatic cells. Lanes 4, 5, 6 show partially regulated, 
The studies outlined here are designed and lanes 7, 8, 9 show largely de-re- 
to elucidate the biochemical basis of dif- pressed, chromatin preparations. The in- 
ferential 5S RNA gene expression from crease in 5S synthesis in lanes 4 and 7 is 
a chromatin template. the result of de-repression of the oocyte- 

A gentle procedure solubilizes chro- type genes, and the extract depleted of 
matin from somatic nuclei while main- the 40-kDa factor in each case gives a 
taining the developmentally regulated signal identical to that of the fully reg- 
state. Nuclei are digested with small ulated chromatin template (lanes 5 and 
amounts of micrococcal nuclease to de- 8). It is also evident that de-repressed 
crease the size of DNA. This increases chromatin preparations give more non- 
the solubility of chromatin prepared by specific transcription by RNA polymer- 
lysis of the nuclei. This chromatin is then ase III than do well-regulated chromatin 
concentrated by centrifugation and used preparations. We are trying to correlate 
for transcription studies. the state of oocyte-type 5S RNA gene 

In agreement with results reported by expression with the complement of chro- 

Wormington {Year Book 81, 214), the mosomal proteins associated with these 

oocyte-type genes in somatic chromatin genes. 



123 4 56 789 10 11 

i ii in 



Fig. 57. (A) Autoradiogram of the polyacrylamide/urea gel analysis of in vitro transcription using 
erythrocyte chromatin from three different frogs as template. Lanes 1, 2, 3, prep I; lanes 4, 5, 6, prep 
II; lanes 7, 8, 9, prep III. Lanes 10 and 11 were programmed with a mixture of cloned 5S and tRNA 
genes. Reactions 1, 4, and 7 were transcribed in oocyte nuclear extract; 2, 5, 8, and 10 were transcribed 
in the same extract but depleted with a specific antibody for the 40-kDa 5S-specific transcription factor; 
3, 6, and 9 were transcribed by pure RNA polymerase III; 11 was transcribed by depleted oocyte nuclear 
extract to which 0.1 |xg of purified factor was added. (B) An SDS-polyacrylamide gel of chromatin proteins 
from preps I, II, and III. The positions of the core histones and HI are shown. 

An Assay for Discrimination 

between Somatic and Oocyte 5S 

RNA Genes: Evidence for Post- 

Transcriptional Control 

D. Broivn mid E. Jordan 

We reported in Year Book 81, p. 221, 
that a mixture of cloned oocyte and so- 
matic 5S RNA genes injected into early 
Xenopus embryos were transcribed to 
about the same relative extent as they 
are at the midblastula stage, when the 
first transcription of these genes occurs 
during embryogenesis. We screened 
deletion mutants and found that muta- 
tions in the internal control region re- 
duced transcription of the 5S RNA genes 
introduced into embryos (compared to 
their transcription in in vitro extracts or 
when injected into oocyte nuclei). Thus, 

a 1:50 mixture of wild-type and mutant 
cloned 5S RNA genes, when transcribed 
in vitro or when injected into early em- 
bryos, will transcribe RNA with effi- 
ciencies of 1:10 and 1:1, respectively. We 
conclude that mutations in the internal 
control region in cloned genes injected 
into embryos are exaggerated in their 
effect on transcription by an order of 

We then tested the stability of a va- 
riety of 5S RNAs injected into embryos. 
Every labeled 5S RNA that had muta- 
tions in the 5' half of the molecule, in- 
cluding the 5' part of the control region 
(oocyte genes), was more labile than con- 
trol somatic 5S RNA. Mutations in the 
3' part of the molecule did not affect 5S 
RNA stability. This result suggests that 
at least part of the exaggerated tran- 
scription of oocyte 5S RNA genes in- 


jected into embryos may be due to weakened control regions are tran- 

selective post-transcriptional processing scribed an order of magnitude less effi- 

of the 5S RNA. However, genes with ciently than are control genes. 


R. O'Rahilly andF. Miiller 

The human and comparative embryol- which are being produced by the Per- 
ogical collections are housed at the Uni- spektomat. (2) Appropriate data are being 
versity of California, Davis, and are in incorporated into a Digital Equipment 
the process of being moved to the Pri- Corporation computer at the Johns Hop- 
mate Center. Inquiries, as well as re- kins Medical Institutions, in collabora- 
quests for permission for publications, tion with Grover M. Hutchins and G. 
should be addressed to R. O'Rahilly, William Moore. At present, some 100 key 
Carnegie Laboratories of Embryology, developmental features in 168 embryos 
California Primate Research Center, of stages 7-15 are being analyzed. 
Davis, California 95616. As part of a comprehensive study of 

the development of the cranial nerves, a 

Developmental Stages in Human verv , detail f investigation of the early 

p development of the hypoglossal nerve and 

occipital somites in staged human em- 

The revision of stages 10 to 23 is being bryos is awaiting publication. An unu- 

pursued by O'Rahilly and Miiller, who sually precise interpretation has been 

are concentrating on the nervous sys- made possible by examining a large num- 

tem. Data relating to the reproductive ber (105) of specimens, including silver 

and endocrine systems were published preparations, and by preparing 23 graphic 

during the year (O'Rahilly, 1983). reconstructions. For the first time, it has 

been possible to adduce adequate evi- 

Development of the Nervous dence j n f the hu ™ an ^ the ^S 1 ? 3 sal 

q musculature is derived largely, if not en- 
tirely, from occipital myotomes. 

An account of the nervous system at The hypoglossal nucleus is evident at 

stage 8 is now available (O'Rahilly and stage 12 and becomes isolated from other 

Miiller, 1981). The development of the efferent nuclei at stage 14. The first hy- 

brain as seen in intact, staged embryos poglossal nerve fibers appear at stage 12, 

has been detailed (O'Rahilly, Miiller, and and the developing roots become seg- 

Bossy, 1982). A popular account of the mentally related to the occipital my- 

developing brain has also appeared otomes at stage 13. The roots unite at 

(O'Rahilly and Miiller, 1983). stage 14 and the main trunk arrives in 

The development of the nervous sys- the tongue at stage 15. Four occipital 

tern in staged human embryos is being somites can be identified during stage 

studied by O'Rahilly and Miiller along 13, and the sclerotomic material forms 

two main routes. (1) An extremely de- two bilateral masses. The fourth sclero- 

tailed histological examination of serial tome separates in stage 14 and develops 

sections, including silver preparations, is like a vertebra. This and the remaining 

being combined with the preparation of sclerotomic material form the basioccip- 

precise graphic reconstructions, some of ital and exoccipital, which are the first 



parts of the chondrocranium to appear. 
Four occipital myotomes develop and 
grow towards the tongue as the "hypo- 
glossal cord," which arrives prior to the 
hypoglossal nerve. The developmental 
similarity in the hypoglossal region be- 
tween birds and mammals, combined with 
experimental studies in birds, renders it 
extremely likely that the hypoglossal 
musculature in mammals also is derived 
from occipital somites. The present study 
is the first in which this conclusion is ad- 
equately supported in the human. This 
investigation aids in the interpretation 
and timing of origin of variations (e.g., 
bipartite hypoglossal canal) and anom- 
alies (e.g., persistent hypoglossal ar- 

Collaboration continues with J. Bossy 
(Nimes), who has recently published pre- 
liminary findings on the sequence of de- 
velopment of the cutaneous innervation 
of the limbs (Bossy, 1982). 

Collaboration continues also with W. 
Wozniak (Poznari) on the fine structure 
of the developing nervous system, as a 
result of which a study of the fetal mye- 
lination of pyramidal fibers at the level 
of the pyramidal decussation was pub- 
lished recently (Wozniak and O'Rahilly, 

The Developing Skeletal System 

Detailed studies of the embryonic ver- 
tebral column are being continued by 
O'Rahilly, Miiller, and David B. Meyer 
(Detroit). An investigation of the occip- 
itocervical region at stage 23, including 
graphic reconstructions, has been pub- 
lished recently (O'Rahilly, Miiller, and 
Meyer, 1983). 


A new book on placentation in general 
(including classifications) and on the pla- 
centae of selected animals has been writ- 
ten by Elizabeth M. Ramsey. The artwork 
is by Ranice W. Crosby. A cross-index, 
glossary, and bibliography are included. 

Staff Activities 

Ronan O'Rahilly was appointed The 
Sterling Bunnell Memorial Lecturer for 
1982, as a result of which he spoke on 
"The human hand in development" in San 
Francisco in November 1982. Also in No- 
vember 1982 he was the Hermann Pinkus 
Lecturer at Wayne State University, 
Detroit, where he spoke on "Ecto- and 
other derms." 


Staff Members and fellows partici- 
pated as chairpersons and/or speakers at 
the following conferences and symposia: 
Gordon Research Conferences on Nu- 
cleic Acids, Plant Molecular Biology, An- 
imal Cells and Viruses, Biological 
Regulatory Mechanisms, Lysosomes, 
Molecular Pharmacology, and Membrane 
Lipid Metabolism; Sixth Taniguchi Sym- 
posium on the Neurobiology of Plasticity 
and Growth, Kyoto; Neuronal Develop- 
ment; Society of Neuroscience; EMBO 
Workshop, Crete; Expression of Genes, 
Melbourne; 3rd International Congress 
of Recombinant DNA, Philadelphia; Na- 
tional Drosophila Meeting, Asilomar; Ge- 
netic Manipulation, Cologne; Expression 

of Cloned Genes, Bethesda; Role of 
Biomembranes in the Integrity and 
Function of Cells, New York; Lipid in 
Biomembranes, Woods Hole; Brain Re- 
search, Keystone; 4th International C. 
elegans meeting; 5th International Con- 
gress of Parasitology; NATO Sympos- 
ium on Structure and Function of Plant 
Genomes; Plant Molecular Biology, Key- 
stone; Molecular Approaches to the Ner- 
vous System, Cold Spring Harbor; 
International Congress on Neuromus- 
cular Diseases, Marseilles; Ciba meeting 
on Molecular Biology of Egg Maturation, 
London; UCLA Meeting on Gene 
Expression, Park City; Transfer and 
Expression of Eukaryotic Genes, New 


York; Minority Biomedical Support Methods. Members served as organizer 

Symposium, Washington; American So- and co-chairman of the first Princeton Li- 

ciety for Cell Biology; Bicentennial Sym- posome Conference on Liposomes in Cell 

posium of Harvard Medical School. Biology; councillor, Society of General 

Lectures were given at Cold Spring Physiologists; member, scientific advi- 
Harbor Laboratory; Marine Biological sory boards of Searle Scholars Program, 
Laboratory; State University of New Muscular Dystrophy Association, De- 
York at Buffalo, Stony Brook, and partment of Biochemistry, Princeton 
Binghamton; Purdue University; Tokyo University, Department of Cellular and 
Medical School; Columbia University; Developmental Biology, Harvard Uni- 
University of Maryland Medical School; versity (chairman), the Robert Wood 
University of Miami Medical School; Johns Johnson Foundation, the Life Sciences 
Hopkins Medical School and School of Research Foundation (president), and the 
Hygiene and Public Health; University NIH Recombinant DNA Advisory Com- 
of Utah; Massachusetts General Hospi- mittee. 
tal; University of Paris; Upstate Medical 

Center, Syracuse; Syracuse University; Seminars 

Rutgers University; Yale University; 

Eisenhower Medical Center; Salk Insti- The Department offers at least two 

tute; University of Pennsylvania Medical seminars each week to the Baltimore 

School; University of Pittsburgh; Cornell community. In addition, there are weekly 

University; Princeton University; "Progress Reports" by our staff, fellows, 

Northwestern University; University of and students for those interested in cell 

Ghent; Roswell Park Memorial Institute; biology and nucleic acid topics. This year 

Animal Parasitology Institute of the U.S. eight invited scientists from Baltimore 

Department of Agriculture; University lectured. The Sixth Annual Carnegie 

of Miami; Washington University; Har- Minisymposium was entitled "The Ge- 

vard University; University of Mary- netics of Mouse Development." The five 

land; New York University Medical speakers were Virginia Papaiannou, Tufts 

School; University of Virginia School of University; Willys Silvers, University of 

Medicine; The Upjohn Company; Uni- Pennsylvania; Lee Silver, Cold Spring 

versity of Connecticut; Brandeis Uni- Harbor Laboratory; Wallace Rowe, Na- 

versity; University of Iowa; University tional Institutes of Health; and Richard 

of Nebraska; Arizona State University; Mulligan, Massachusetts Institute of 

Stanford University; Fred Hutchinson Technology. This year we inaugurated a 

Cancer Research Center; Max Planck In- "Disease of the Month Club," an occa- 

stitute, Tubingen; Howard University; sional evening lecture mainly on dis- 

Armed Forces Radiobiology Research eases, but also to include agricultural and 

Institute; University of Geneva; Cancer environmental problems. The objective 

Institute, Lausanne; University of Zu- is cross-fertilization and information 

rich; Eleanor Roosevelt Institute for transfer between ourselves and those who 

Cancer Research; Institute for Medical study these important problems. To date 

Research, Camden, N.J.; Medical Re- we have heard about multiple sclerosis, 

search Council, Cambridge; the Sonne- lupus erythemitosis, diseases of the ret- 

born Memorial Lecture, Bloomington; ina, and antibiotic-induced diarrhea. We 

Massachusetts Institute of Technology; host a monthly evening Cell Biology 

University of California, Berkeley. Meeting that draws its audience and 

Members of the Staff served on the speakers from Baltimore and Washing- 
editorial boards of Gene, Developmental ton. Other lecturers during the year in- 
Biology, Journal of Lipid Research, eluded R. Bloodgood and S. Young, 
Neuroscience, Developmental Brain Re- University of Virginia; D. Carroll, Uni- 
search, and Journal of Neuroscience versity of Utah; J. Schell, Max Planck 



Institute, Cologne; R. Kelley and J. 
Rothman, Stanford University; P. Us- 
ter, University of California, Davis; B. 
Cain and J. Chorey, University of Illi- 
nois; G. Leys, Baylor College of Medi- 
cine; R. Klausner, B. Moss, and J. 
Weinstein, National Institutes of Health; 
S. McKnight and R. Reeder, Hutchinson 
Cancer Research Institute; P. Fisher, U. 
Rutishauser, and T. Wong, Rockefeller 
University; J. Newport and M. Kir- 
schner, University of California, San 

Francisco; E. Weishaus, Princeton; M. 
Fuller, A. Mahowald, R. Raff, and M. 
Scott, Indiana University; M. Tamkun, 
University of Washington; C. Montell, 
University of California, Los Angeles; F. 
Laski, 0. Sundin, and C. Tabin, Mas- 
sachusetts Institute of Technology; J. 
Deming, Chesapeake Bay Institute; P. 
Constantinides, Brown University; K. 
Bennett, Roswell Park Memorial Insti- 
tute; D. Bennett, Purdue University; S. 
Malcolm, Queen Elizabeth College. 


Bayne, E. K., see Fambrough, D. M. 

Bossy, J., Sequence du developement de l'in- 
nervation cutanee des membranes chez 
l'embryon humain, Bull. Ass. Anat. 66, 57- 
61, 1982. 

Bossy, J., see O'Rahilly, R. 

Brown, D. D., Gene expression in eukar- 
yotes, Oncodevel. Biol. Med. k, 9-29, 1982. 

Brown, D. D., The Ernst W. Bertner Mem- 
orial Award Lecture: Developmental Con- 
trol of a Simple Gene, in Perspectives on 
Genes and the Molecular Biology of Can- 
cer, D. L. Robberson and G. F. Saunders, 
eds., Raven Press, New York, 7-16, 1983. 

Brown, D. D., see Sakonju, S. 

Carbonetto, S., and K. J. Muller, Nerve fiber 
growth and the cellular response to axo- 
tomy, Curr. Top. Devel. Biol. 17, 33-76, 

Chaleff, D., see Fedoroff, N. 

Chiquet, M., and D. M. Fambrough, Cellular 
origin of extracellular matrix components 
during muscle morphogenesis revealed by 
monoclonal antibodies, in Limb Develop- 
ment and Regeneration , R. 0. Kelly, P. F. 
Goetinck, and J. A. MacCabe, eds., Alan 
R. Liss, Inc., New York, pp. 359-368, 1983. 

Collins, M., and G. M. Rubin, Structure of 
the Drosophila mutable allele, white-crim- 
son, and its white-ivory and wild-type de- 
rivatives, Cell SO, 71-79, 1982. 

Collins, M., see Levis, R. 

Courage-Tebbe, U., H.-P. Doring, N. Fe- 
doroff, and P. Starlinger, The controlling 
element Ds at the Shrunken locus in Zea 
mays: structure of the unstable sh-m593S 
allele and several mutants, Cell, in press, 

Courage-Tebbe, U., see Fedoroff, N. 

Doring, H.-P., see Courage-Tebbe, U., and 
Fedoroff, N. 

Fambrough, D. M., and E. K. Bayne, Mul- 
tiple forms of (Na + + K + )-ATPase in the 
chicken: selective defection of the major 
nerve, skeletal muscle, and kidney form by 
a monoclonal antibody, /. Biol. Chem. 258, 
3926-3935, 1983. 

Fambrough, D. M., J. M. Gardner, and E. K. 
Bayne, Regulation of ACh receptor num- 
ber and distribution in skeletal muscle, in 
Neurotransmitter Receptors: Biochemical 
Aspect and Functional Significance, H. 
Yoshida, ed., John Wiley, pp. 1-12, 1983. 

Fambrough, D. F., see Chiquet, M., Gard- 
ner, J. M., and Rotundo, R. L. 

Fedoroff, N., Controlling elements in maize, 
in Mobile Genetic Elements, J. Shapiro, ed., 
Academic Press, New York, pp. 1-63, 1983. 

Fedoroff, N., and D. Chaleff, The structure 
and expression of the Shrunken locus in 
maize strains with mutations caused by the 
controlling element Ds, in Genetic Engi- 
neering: Applications to Agriculture, L. 
Owens, ed., Rowman and Allenheld Pub- 
lishers, Totowa, New Jersey, in press, 1983. 



Fedoroff, N., J. Mauvais, and D. Chaleff, Mo- 
lecular studies on mutations at the Shrun- 
ken locus in maize caused by the controlling 
element Ds, J. Mol. Appl. Gen. 2, 11-29, 

Fedoroff, N., D. Chaleff, U. Courage-Tebbe, 
H.-P. Doring, M. Geiser, P. Starlinger, E. 
Tillman, E. Week, and W. Werr, Mutations 
at the Shrunken locus in maize caused by 
the controlling element Ds, in Structure and 
Function of Plant Genomes, 0. Ciferri and 
L. Dure, eds., Plenum Press, New York, 
in press, 1983. 

Fedoroff, N., see Courage-Tebbe, U. 

Gardner, J. M., and D. M. Fambrough, Me- 
tabolism of cell surface receptors: possible 
role in cell sensitivity and responses to ac- 
tivators, in Biological Regulation and De- 
velopment, Vol. 5, Hormone Action, K. 
Yamamoto and R. F. Goldberger, eds., 
Plenum Press, New York, 1982. 

Gardner, J. M., and D. M. Fambrough, Fi- 
bronectin expression during myogenesis, 
/. Cell Biol. 96, 474-485, 1983. 

Geiser, M., see Fedoroff, N. 

Hogan, E., see Ward, S. 

Karess, R. E. , and G. M. Rubin, A small tan- 
dem duplication is responsible for the un- 
stable white-ivory mutation in Drosophila, 
Cell SO, 63-69, 1982. 

Klass, M., see Ward, S. 

Levis, R. L., and G. M. Rubin, The unstable 
w dzl mu tation of Drosophila is caused by 
a 13 kilobase insertion that is imprecisely 
excised in phenotypic revertants, Cell SO, 
543-550, 1982. 

Levis, R. L., M. Collins, and G. M. Rubin, 
FB elements are the common basis of the 
instability of the w DZL and w c Drosophila 
mutations, Cell SO, 551-565, 1982. 

Lipsky, N. G., and R. E. Pagano, Sphingo- 
lipid metabolism in cultured fibroblasts — 
microscopic and biochemical studies em- 
ploying a fluorescent analogue of ceramide, 
Proc. Nat. Acad. Sci. USA 80, 2608-2612, 

Longmuir, K. J., see Pagano, R. E. 
Martin, 0. C, see Pagano, R. E. 

Mason, A., and K. J. Muller, Regeneration 
and plasticity of neuronal connections in the 
leech, Trends Neurosci. 6, 172-176, 1983. 

Mauvais, J., see Fedoroff, N. 

Meyer, D. B., see O'Rahilly, R. 

Muller, F., see O'Rahilly, R. 

Muller, K. J., Cell recognition and specific 
synapse formation, in Repair and Regen- 
eration of the Nervous System, J. G. Ni- 
cholls, ed., Dahlem Konferenzen, Berlin, 
Heidelberg, New York, Springer- Verlag, 

Muller, K. J., see Carbonetto, S., and Mason, 

Nelson, G. A., see Ward, S. 

Nichols, J. W., and R. E. Pagano, Resonance 
energy transfer assay of protein-mediated 
lipid transfer between vesicles, /. Biol. 
Chem, 258, 5368-5371, 1983. 

O'Rahilly, R., The timing and sequence of 
events in the development of the human 
reproductive system during the embryonic 
period proper, Anat. Embryol. 166, 247- 

O'Rahilly, R., The timing and sequence of 
events in the development of the human 
endocrine system during the embryonic pe- 
riod proper, Anat. Embryol. 166, 439-451. 

O'Rahilly, R., and F. Muller, The first ap- 
pearance of the human nervous system at 
stage 8, Anat. Embryol. 16S, 1-13, 1981. 

O'Rahilly, R., and F. Muller, Developpement 
prenatal du cerveau, C'est Pour Quand? 6, 
38-42, 1983. 

O'Rahilly, R., F. Muller, and J. Bossy, Atlas 
des stades du developpement du systeme 
nerveux chez l'embryon humain intact, Arch . 
Anat. Histol. Embryol. 65, 57-76, 1982. 

O'Rahilly, R., F. Muller, and D. B. Meyer, 
The human vertebral column at the end of 
the embryonic period proper, 2. The occip- 
itocervical region, J. A?iat. 1S6, 181-195, 

O'Rahilly, R., see Wozniak, W. 

Pagano, R. E., and K. J. Longmuir, Intra- 
cellular translocation and metabolism of flu- 
orescent lipid analogues in cultured 



mammalian cells, Trends Biochem. Sci. 8, 
157-161, 1983. 

Pagano, R. E., K. J. Longmuir, and 0. C. 
Martin, Intracellular translocation and me- 
tabolism of a fluorescent phospholipid an- 
alogue in cultured fibroblasts, J. Biol. Chem. 
258, 2034-2040, 1983. 

Pagano, R. E., see Lipsky, N. G., and Ni- 
chols, J. W. 

Ramsey, E. M., The Placenta, Human and 
Animal, Praeger, New York, 1982. 

Roberts, T. M., and S. Ward, Directed mem- 
brane flow on the pseudopods of C. elegans 
spermatozoa, Cold Spring Harbor Symp. 
Quant. Biol. XLVI, 695-702, 1982. 

Rotundo, R. L., and D. F. Fambrough, Syn- 
thesis, transport, and fate of acetylcholin- 
esterase and acetylcholine receptors in 
cultured muscle, in Membranes in Growth 
and Development, J. Hoffman, G. Gie- 
bisch, and L. Bolis, eds., Alan R. Liss, Inc., 
New York, 1982. 

Rubin, G. M., and A. C. Spradling, Genetic 
transformation of Drosophila with trans- 
posable element vectors, Science 218, 348- 
353, 1982. 

Rubin, G. M., see Collins, M., Karess, R. E., 
Levis, R., and Spradling, A. C. 

Sakonju, S., and D. D. Brown, Contact points 
between a positive transcription factor and 
the Xenopus 5S RNA gene, Cell 31, 395- 
405, 1982. 

Spradling, A. C, and G. M. Rubin, Trans- 
position of cloned P elements into Droso- 
phila germ line chromosomes, Science 218, 
341-347, 1982. 

Spradling, A. C, see Rubin, G. M. 

Starlinger, P., see Courage-Tebbe, U., and 
Fedoroff, N. 

Tillman, E., see Fedoroff, N. 

Ward, S., Genetic analysis of the sensory ner- 
vous system of Caenorhabditis elegans, 
Proc. 5th Int. Congr. Parasitol. VII, 28- 
31, 1982. 

Ward, S., and M. Klass, The localization of 
the major protein in C. elegans sperm and 
spermatocytes, Devel. Biol. 92, 203-208, 

Ward, S., G. A. Nelson, and E. Hogan, The 
initiation of Caenorhabditis elegans sper- 
miogenesis in vivo and in vitro, Devel. Biol. 
98, in press, 1983. 

Ward, S., see Roberts, T. M. 

Week, E., see Fedoroff, N. 

Werr, W., see Fedoroff, N. 

Wozniak, W., and R. O'Rahilly, An electron 
microscopic study of myelination of pyra- 
midal fibres at the level of the pyramidal 
decussation in the human fetus, J. Hirn- 
forsch. 23, 331-342, 1982. 


Year Ended June 30, 1983 

(Including those whose services ended during the year) 

Research Staff 

Donald D. Brown, Director 
Douglas M. Fambrough 
Nina V. Fedoroff 
Kenneth J. Muller 
Richard E. Pagano 
Gerald M. Rubin 
Allan C. Spradling 
Samuel Ward 

Staff Associates 

Sondra G. Lazarowitz 
Richard L. Rotundo 
Martin Snider 

Research Associates (Extramural) 

Bent Boving, Detroit, Michigan 
Igor B. Dawid, Bethesda, Maryland 
Robert L. DeHaan, Atlanta, Georgia 



Arthur T. Hertig, Boston, Massachusetts 
Irwin R. Konigsberg, Charlottesville, Vir- 
Ronan O'Rahilly, Davis, California 
Elizabeth M. Ramsey, Washington, D.C. 
Ronald H. Reeder, Seattle, Washington 
Yoshiaki Suzuki, Okazaki City, Japan 

Postdoctoral Fellows and Grant- 
Supported Associates 

M. John Anderson, 1 Research Associate, 
Muscular Dystrophy (Fambrough) 

Karen Bennett, 2 Fellow of the Carnegie 
Institution of Washington 

Matthias Chiquet, Fellow of the Bay Foun- 

Mary Collins, Fellow of the National In- 
stitutes of Health 

Diane de Cicco, Postdoctoral Fellow of the 
European Molecular Biology Depart- 

Ellen Elliott, 3 Fellow of the National Mul- 
tiple Sclerosis Society 

Kathy French, 4 Fellow, National Insti- 
tutes of Health (Muller) 

Tulle Hazelrigg, Postdoctoral Fellow of the 
Damon Runyon- Walter Winchell Can- 
cer Fund 

Ian Jackson, 5 Fellow of the European Mo- 
lecular Biology Organization 

Laura Kalfayan, Fellow of the National In- 
stitutes of Health 

Roger Karess, Fellow of the National In- 
stitutes of Health 

Robert Levis, Research Associate, Na- 
tional Institutes of Health Grant (Rubin) 

Naomi Lipsky, Fellow of the National In- 
stitutes of Health 

Adrian Mason, 6 Research Associate, Na- 
tional Institutes of Health Grant (Muller) 

Wylie Nichols, 7 Fellow of the National In- 
stitutes of Health 

Kevin O'Hare, Research Associate, Amer- 
ican Cancer Society Grant (Rubin) 

Heli Roiha, 8 Fellow of the British Science 
and Engineering Research Council. 

David Setzer, Fellow of the Jane Coffin 
Childs Memorial Fund 

*To February 1, 1983 
2 From May 1, 1983 
3 To November 30, 1982 
4 From January 1, 1983 
5 To May 1, 1983 
6 To February 1, 1983 
7 To August 1, 1982 
8 From October 1, 1982 

Mavis Shure, Fellow of the Carnegie In- 
stitution of Washington 

Richard Sleight, Fellow of the National In- 
stitutes of Health 

Douglas Smith, Fellow of the Damon Run- 
yon- Walter Winchell Cancer Fund 

William Taylor, 9 Fellow of the National In- 
stitutes of Health 

Barbara Wakimoto, Fellow of the Helen 
Hay Whitney Foundation 

Susan Wessler, 10 Fellow of the American 
Cancer Society 

Barry Wolitzky, 11 Fellow of the Muscular 
Dystrophy Association 

Michael Wormington, 12 Fellow of the Na- 
tional Institutes of Health 


Paula Adams, Undergraduate, Johns Hop- 
kins University 

Daniel Burke, Graduate Student, National 
Research Council of Canada 

Barbara Kirschner, Graduate Student, 
Johns Hopkins University 

Darlene Marshall, Undergraduate, Johns 
Hopkins University 

Vincent Morgese, Undergraduate, Johns 
Hopkins University 

Suki Parks, Graduate Student, Johns Hop- 
kins University 

Mark Schlissel, Medical Science Training 
Program, Johns Hopkins University 

Jennifer Schwartz, Graduate Student, Johns 
Hopkins University 

Visiting Investigators and Extramural 

R. Hallberg, Ames, Iowa 

H. Koenig, Paris, France 

J. Koenig, Paris, France 

M. Schwartz, Baltimore, Maryland 

M. Simmons, Minneapolis, Minnesota 

Clerical and Technical Staff 

Betty Addison, Laboratory Helper 
Michael Barnes, 13 Laboratory Helper 
Paul Blackwell, Custodian (part-time) 
Scott Downing, 14 Technician 
William H. Duncan, Senior Technician 

9 From June 15, 1982 
10 To December 1, 1982 
n From March 15, 1983 
12 To September 1, 1982 

13 To September 1, 1982 
14 From September 22, 1982 



James Fenwick, 15 Laboratory Helper 
Ernestine V. Flemmings, Laboratory 

Richard D. Grill, Photographer 
Virginia Hicks, Laboratory Helper 
Mary E. Hogan, Technician 
John E. Jones, Custodian 
Eddie Jordan, Senior Technician 
Nancy Jordan, Laboratory Helper (part- 
Joseph Levine, Technician 
Alice H. Mabin, 16 Laboratory Helper 
Thomas F. Malooly, Business Manager 
Jeffrey Malter, 17 Librarian (part-time) 
Ona Martin, Technician 
Jeffery Mauvais, 18 Technician 

15 From September 16, 1982 
lt5 To August 2, 1982 
17 From November 22, 1982 
18 To September 30, 1982 

Abbie Mays, Technician 

Thomas F. Miller, Custodian 

Christine Murphy, Technician 

Betty Ortiz, 19 Technician 

John Pazdernik, Building Engineer 

Betty Lou Phebus, Bookkeeper/Clerk 

Ophelia Rogers, 20 Technician 

Susan Satchell, Secretary 

Pat Schmidt, Secretary 

Delores Somerville, Senior Technician 

Rodney Stalks, 21 Laboratory Helper 

Barbara Thomas, 22 Technician 

Diane Thompson, 23 Laboratory Helper 

Joe Vokroy, Machinist 

Gloria Wilkes, Laboratory Helper 

19 From October 1, 1982 
20 From December 16, 1982 
21 To February 9, 1983 
22 To June 10, 1983 
23 From February 14, 1983 

Geophysical Laboratory 

Washington, District of Columbia 

Hatten S. Yoder, Jr. 


.5 - 

O Jh 










PQ £ 


Ph § 

t ^ 

Sh fl) 



bJO >j c 

S "bJD O 

^ £ * 

fl S CD 









;> <d 

O Sh 

Sh TO 

3 • 

9 • 

C Ph 

,-* Sh 

TO O) 

CO &J0 



O r p 


— I Sh CD 




e :«m 


N oi S 
So 4J.S x 

>> s s 
d PP w 4 

o . 

■a cc 

5 <u 

oo r • 

OS (J-H 

J3 b 



£ X3 . 



Sh T3 
5 O 

TO • 

w 2 


o . 
£ En 


. ■ - *h 
TO £, 

■■'■a p 

. TO 1-3 

of PQ "> 








. O £ 

Ph . g 

Q 3 




TO tJ 

O p 
a o 

h£ Sh 

^ .b 


sj: to 



Acknowledgments 233 

Commentary and Overview of the Director . . 235 

Igneous and Metamorphic Petrology 251 

Heat transfer related to metamorphism 
and magma generation: results of ex- 
ploratory experiments (Yoder) 251 

Partitioning of iron and magnesium be- 
tween melilite, olivine, and clinopyrox- 
ene in lavas (Velde and Yoder) 256 

Petrology of olivine melilitites from Salt- 
petre Kop and Sutherland Commage, 
Cape province, South Africa (Boctor 
and Yoder) 264 

Correlation of bimodal 18 values with al- 
teration minerals of synmetamorphic 
granitic rocks, Augusta, Maine (Rum- 
ble, Ferry, and Hoering) 267 

Textural and isotopic variation of graphite 
in the New Hampshire Plutonic Series 
(Duke and Rumble) 270 

Mineralogical and lead isotope studies of a 
polymetallic sulfide deposit at Derhib, 
Egypt: implications for metallogenesis 
in the Red Sea district (Boctor and 
Tera) 273 

Partitioning of nickel between silicate and 
iron sulfide melts (Boctor and Yoder) . . 275 

An IGBA base builder and information 
system (Li and Chayes) 277 

Co-occurrence of mode, mineral assem- 
blage, and bulk chemical composition in 
published petrographic descriptions 
(Chayes) 278 

Data verification procedures in IGBA 
(Chayes, Li, and Stewart) 280 

Inversion of variance relations by the log- 
arithmic transformation (Chayes) 281 

Layered Intrusions 284 

Observations on the origins of Skaergaard 

layering (Irvine) 284 

Skaergaard trough-layering structures 

(Irvine) 289 

Melting relations of two Bushveld chilled 

margin rocks and implications for the 

origin of chromitite (Sharpe and 

Irvine) 295 

Density and viscosity characteristics of 

melts of Bushveld chilled margin rocks 

(Sharpe, Irvine, Mysen, and Hazen) . . . 300 

Structure and Properties of Liquids and 

Glasses 305 

Redox equilibria and the anionic structure 
of Na 2 # xSi0 2 -Fe-0 melts: effect of ox- 

ygen fugacity (Virgo, Mysen, and 
Danckwerth) 305 

The coordination of Fe 3+ in oxidized vs. 
reduced sodium aluminosilicate glasses: 
a 57 Fe Mossbauer study (Virgo, Mysen, 
and Danckwerth) 309 

Redox equilibria, structure, and melt 
properties in the system Na 2 0-Al 2 3 - 
Si0 2 -Fe-0 (Mysen and Virgo) 313 

Iron-bearing alkaline-earth silicate melts: 
relations between redox equilibria of 
iron, melt structure, and liquidus phase 
equilibria (Mysen and Virgo) 317 

Effect of pressure on the structure of 
iron-bearing silicate melts (Mysen and 
Virgo) 321 

Diffusion of oxygen in jadeite and diop- 
side melts at high pressures (Shimizu 
and Kushiro) 325 

Diffusion of network-forming cations in 
jadeitic melts at high pressures 
(Kushiro) 327 

Petrology and Geophysics of the Mantle .... 330 

Mantle metasomatism: the Kimberley 
dunites (Boyd, Jones, and Nixon) 330 

Mineral chemistry of garnet lherzolites 
from the Sudetes, southwest Poland 
(Bakun-Czubarow) 336 

Graphic rutile-olivine intergrowths from 

South African kimberlites (Schulze) .... 343 

Pyroxenites, eclogites, and megacrysts in 
kimberlite from the De Bruyn and Mar- 
tin Mine, Bellsbank, South Africa (Boc- 
tor, Boyd, and Nixon) 346 

High-pressure, Fourier-transform in- 
frared spectra of forsterite and fayalite 
(Xu, Mao, Weng, and Bell) 350 

Preliminary data on the Fourier-trans- 
form infrared-frequency shifts in hyper- 
sthene at high pressure (Xu, Mao, 
Weng, and Bell) 352 

Preliminary Fourier-transform infrared 
spectral data on the Si0 6 8 " octahedral 
group in silicate-perovskites (Weng, 
Xu, Mao, and Bell) 355 

Crystal Structures at High Pressures 357 

High-pressure crystal structures and com- 
pressibilities of bertrandite, beryl, and 
euclase (Hazen, Finger, and Barton) . . . 357 
Compressibilitv and high-pressure pleo- 

chroism of CaCr^SiAo (Hazen) 359 

Anomalous compressibility of zeolite 4A in 
several hydrostatic pressure media (Ha- 
zen and Finger) 361 

Composition of Planets and Meteorites 363 

Present constraints on the composition of 

the mantle of Mars (Goettel) 363 

Pressure-induced infrared spectra of hy- 
drogen to 542 kbar (Mao, Xu, and Bell) 366 

Mineral-Fluid Equilibria 372 

Electrical conductances and ionization 
constants of acids and bases in super- 
critical aqueous fluids: hydrochloric acid 
from 100° to 700°C and at pressures to 

4000 bars (Frantz and Marshall) 372 

Exchange equilibria of alkali feldspars 
with fluoride-bearing fluids (Barton and 

Frantz) 377 

Calculation of C-O-H-S equilibria at con- 
stant bulk composition: some petrologic 
implications (Barton) 381 

Biogeochemistry 386 

Molecular fossils generated by hydrous 
pyrolysis of kerogen (Hoering) 386 

Source of organic nitrogen in Mid-Atlantic 
coastal bays and continental shelf sedi- 
ments of the United States: isotopic ev- 
idence (Macko) 390 

Microbial alteration of stable nitrogen and 
carbon isotopic compositions of organic 
matter (Macko and Estep) 394 

Nitrogen isotope biogeochemistry of ther- 
mal springs (Estep) 398 

Stable nitrogen and carbon isotopic com- 
position of individual amino acids iso- 
lated from cultured microorganisms 
(Macko, Estep, Hare, and Hoering) . . . 404 

Carbon and nitrogen isotopic composition 
of amino acids in modern and fossil col- 
lagens (Hare and Estep) 410 

Mesophase development in a bitumen 
from the Nanisivik Mississippi Valley- 
type deposit (Gize and Rimmer) 414 

New Apparatus and Techniques 419 

High-pressure, Fourier-transform spec- 
troscopy (Mao, Bell, Xu, and Wong) . . . 419 

Automated system for heating and for 
spectral measurements in the diamond- 
window, high-pressure cell (Mao, Hadi- 
diacos, Bell, and Goettel) 421 

Position-sensitive, x-ray diffraction sys- 
tem for high-pressure experiments 
(Mao, Bell, and Weng) 424 

Radiographic determination of the posi- 
tions of platinum spheres in density- vis- 
cosity studies of silicate melts (Hazen 
and Sharpe) 428 

Staff Activities 430 

Seminar Series 430 

Penologists' Club 430 

Washington Crystal Colloquium 431 

Washington Organic Geochemistry 

Colloquium 431 

Lectures, Symposia, and Awards 431 

Field Studies 433 

Bibliography 434 

Personnel 436 


The following individuals and organizations have generously sup- 
plemented the general funds allotted to the Geophysical Laboratory: 

American Selco, Inc. 


Carnegie Corporation of New York 

Charles E. Culpeper Foundation, Inc. 

Charles A. Dana Foundation, Inc. 

The Ambrose Monell Foundation 

National Aeronautics and Space Administration 

National Science Foundation 

Teagle Foundation, Inc. 

U.S. Department of Energy 


National interest in geological prob- all of which contribute to solutions of to- 
lems has grown very rapidly in the last day's practical geological problems, 
few years. The major natural catastro- Because geology effectively integrates 
phes of volcanic eruptions, earthquakes, all the basic sciences and is problem ori- 
and floods, the man-made catastrophes ented rather than discipline oriented, it 
of pollution and toxic waste disposal, the requires all the most modern tools of the 
long-term problem of fresh water deple- basic sciences. Every new tool the chem- 
tion, and potential limitations of access ist, physicist, biologist, or mathemati- 
to strategic and critical minerals, all have cian invents can be applied to earth 
generated a need to know more about the problems; yet each must be greatly mod- 
crust of the earth. As the nation now asks ified by the geologist to accommodate the 
for an unprecedented leap forward in sci- extremes of pressure and temperature 
ence and technology for economic prog- under which earth processes operate and 
ress and national security, the demand minerals are formed or destroyed. In turn, 
for geologically prescribed solutions to the other basic sciences profit from the 
some of these problems must be met. inventions and extensions of instruments 
Some national leaders perceive the de- by geologists. Unfortunately, in many of 
teriorating state of education in the geo- the 126 Ph.D. -degree-granting depart- 
logical sciences, the lack of modern ments of geology in the United States, 
equipment, the need for professional this "technology transfer" either has not 
leadership and for coordination among occurred or is decades behind, 
concerned government departments, and The Geophysical Laboratory has been 
the shrinking mining and mineral-pro- outstanding in sustaining a leadership role 
cessing industries as cause for concern, in the earth sciences. It has been the 
Society is discovering that major efforts wellspring not only of ideas and data but 
are needed to acquire the knowledge to also of exceptionally well-trained inves- 
solve today's geological problems. tigators. The Carnegie Institution has 

It is obvious that practical problems funded geological research because this 
can be resolved more easily and more research is important to mankind, not 
promptly if the general laws and funda- because it is popular or commercially re- 
mental methods of analysis are known, warding. Such support is in the Carnegie 
Development of such principles for solv- tradition — to support those areas of great 
ing today's geologically related problems intellectual importance, 
involves long lead times, especially in view Few other sciences have depended so 
of the complexity of natural phenomena, much on a small number of basic research 
The Geophysical Laboratory is one of the organizations. Whereas the universities 
few organizations able to concentrate on are necessarily preoccupied with teach- 
developing geological principles rather ing, there remain too few institutions 
than on responding to short-term, prac- where the development of principles is 
tical problems. Studies of the origin of the major role. Some organizations (e.g., 
life, the formation of the earth's mantle the U.S. Geological Survey) are even 
and core, the generation of magmas, the being divested of their research capabil- 
distribution of elements in the crust, the ities in order to pursue activities pro- 
structures and physicochemical proper- grammed by nonprofessionals. Strong, 
ties of minerals, and the formation of ocean vigorous institutions like the Geophy si- 
basins are fundamental research areas, cal Laboratory are a necessity, not a lux- 



ury, for the scientific growth of the earth periments involving three different heat 
sciences. It is essential that free, unfet- fluxes, the thermal gradient approached 
tered research by talented individuals be steady state at a rate greater than that 
maintained if economic progress and na- observed for most experimental meta- 
tional security are to be achieved. Re- morphic reactions on very-fine-grained 
quests for such freedom are usually met material. Yoder concludes, with due re- 
with a three-, five-, or ten-year plan that gard for the scale of his experiments, that 
by its very nature destroys the flexibility the thermal process in contact meta- 
and inspirational aspects of basic re- morphism may be decoupled from the 
search. A forefront researcher must have chemical reaction process and will pre- 
the freedom to change direction rapidly, sumably be decoupled even in a process 
take advantage of new opportunities, fol- resulting in melting, 
low the most rewarding avenues, and The cooling of a lava on the surface is 
communicate with a wide diversity of tal- a kinetic process that greatly influences 
ented colleagues in the basic sciences. In the composition of the minerals in the 
turn, the organization in which he or she final assemblage. A perplexing problem 
works must also attempt to provide is to ascertain when a mineral nucleates 
whatever is necessary to enhance the re- and grows relative to other phases formed 
searcher's abilities to achieve his or her or to be formed. Correlation of the zones 
goals. The interrelationships of talented of growth between crystals, for example, 
researchers within the organizational has important bearing on the interpre- 
structure take years to develop and are tation of partition coefficients of coexist- 
among the most valuable assets of the ing phases. Velde and Yoder initiated a 
Institution. study on the nucleation and growth char- 
In the following pages, the reader can acteristics of the minerals in melilite- 
perceive the weave of talents that has bearing lavas because of the advantages 
yielded an exceptionally large number of inherent in their mineralogy (e.g., ab- 
new ideas and techniques. Whether at- sence of plagioclase) and the limited 
tributable to the stimulation of visitors, number of distinct assemblages. They 
chance meetings, vigor of fellows, dedi- found that apparent partition coefficients 
cated help of support personnel, or ma- and crystal sizes are compatible with a 
turity of staff, the products are evidence variety of possible interpretations in- 
of the success of the Carnegie spirit. volving the rate of cooling, extent of in- 
heritance from depth, and initiation and 

T 7 , , , , . n , , closure properties. In two specimens, one 

Igneous and Metamorphw Petrology phage appearg tQ haye jnitiated and ter . 

Igneous rocks form as a result of the minated growth, apparently not as a re- 
partial or complete crystallization of a suit of a reaction, before the other phases 
liquid. Study of such rocks usually fo- grew! It became evident in this study 
cuses on the processes that take place in that many of the traditional interpreta- 
tive liquid or mixture of liquids after gen- tions of the crystallization sequence in 
eration. The generation of the liquid it- lavas will have to be reexamined, 
self, however, imposes many constraints The melilite-bearing rocks were also 
on its behavior. Yoder has initiated a new the focus of attention of Boctor and Yoder 
experimental program to examine the because of the uncertainty surrounding 
melting process in which heat is trans- the relations, if any, between olivine 
ferred from a hot refractory body to a melilitite, kimberlite, and carbonatites. 
mass that undergoes partial melting. In The close association in space and time 
preliminary experiments preparatory to between these rocks led some petrolo- 
observing the thermal characteristics of gists to believe that they are genetically 
melting, he obtained results that are ap- related. Experimental data on the sta- 
plicable to contact metamorphism. In ex- bility of melilite places severe con- 


straints on the pressure regimes at which tural evidence of two generations of 

olivine melilitites may crystallize. Oli- graphite in the plutons is supported by 

vine melilitites from two intrusive sat- 6 13 C analyses; the older, flake graphite 

ellite plugs of the Saltpetre Kop volcano differs by as much as 2%c from the youn- 

in the Cape province, South Africa, are ger, spherulitic graphite. In three vein 

characterized by unusual reaction man- deposits that have not suffered intense 

ties of monticellite, spinels, and mica on deformation, textural relations suggest 

olivine, Ba- and Ti-rich trioctahedral mica, the precipitation of successive layers of 

and Ti-rich garnet, and by groundmass graphite, apparently filling fissures. Dif- 

spinels that mimic in their crystallization ferent graphite layers vary in 8 13 C by as 

trends groundmass spinels in kimber- much as 2%o, plutonic graphites cover the 

lites. Boctor and Yoder suggest that a same range of 8 13 C values as graphites 

carbonate-rich melt enriched in the REE from metasedimentary wall rocks, and 

and other incompatible elements proba- values of vein graphites overlap the up- 

bly evolved in the final stages of frac- per end of the range of plutonic graph- 

tionation of the magmas that produce both ites. It appears likely, therefore, that 

kimberlites and olivine melilitites. They plutonic graphites record evidence of in- 

conclude, however, that this observation tergranular flow of metamorphic fluids 

does not imply a genetic relationship be- from wall rocks into plutons, whereas the 

tween these rocks but rather fractiona- vein deposits demonstrate the existence 

tion toward similar residual liquids. of fracture-controlled fluid flow. 

Rumble, Ferry, and Hoeri?ig are in- Ore deposits of another type were ex- 
vestigating the nature of fluid flow dur- amined by Boctor and Tera. The poly- 
ing regional metamorphism of Paleozoic metallic sulfides in a Precambrian 
sediments and synmetamorphic granitic metavolcanic belt in the Eastern Desert, 
plutons from Augusta, Maine. Ferry had Egypt, are similar to those occurring in 
shown previously, from a study of sec- island-arc settings in other parts of the 
ondary minerals of the granitic rocks, that world. The Egyptian deposits differ from 
the plutons were infiltrated by C0 2 - and those in island arcs because of their oc- 
CH 4 -bearing fluids released by meta- currence in talc and because they occupy 
morphic devolatilization reactions. In the fracture zones that run in an east- west 
present study, a correlation has been dis- direction with respect to the axial trough 
covered between the presence and kind of the Red Sea. The mineralogical and 
of secondary minerals and 8 18 values of lead isotope investigations of Boctor and 
granitic rocks as well as their primary Tera indicate that a single metamorphic 
minerals quartz, feldspar, and muscov- episode cannot account for the wide range 
ite. Rocks of the plutons containing sec- in iron content of sphalerite. They argue 
ondary epidote have whole-rock 8 18 that a later hydrothermal event may ac- 
values of 11.1— 11.8%o, whereas rocks count for such variation and for the homo- 
lacking epidote but containing secondary geneity of lead isotope compositions. Their 
calcite and graphite have 8 18 values of lead isotope data indicate equilibration in 
12.5-14.2%o. The results of isotopic anal- very recent times (0-100 m.y.). The ma- 
ysis confirm Ferry's hypothesis of infil- jor tectonic event in East Africa in that 
tration of the plutons by metamorphic period was the Red Sea ocean-floor 
fluids. Quantitative evaluation of the in- spreading to which the system of frac- 
tensity of fluid flow and its spatial vari- tures bearing the talc and associated sul- 
ation awaits completion of isotopic analysis fide ores may be related, 
of the wall rocks. In a study of the partitioning of Ni 

The disseminated graphite and asso- between silicate and sulfide melts as a 
ciated graphite vein deposits of certain function of temperature, /o 2 >/s 2 ' Ni con- 
granite plutons of New Hampshire have centration, and silicate melt composition, 
been studied by Duke and Rumble. Tex- Boctor and Yoder found that the distri- 


bution coefficient for Nit decreases with every eight that contain bulk chemical 

increase in temperature and increase in analyses also contains a modal analysis, 

/o 2 but is not significantly affected by and in recent years the relative incidence 

changes in/ s . Increase in the mafic na- has been much lower. The result is so 

ture of the melt leads to a decrease in extreme as to leave little doubt of the 

K D m . These results are consistent with acute shortage of modal information, even 

the occurrence in nature of Ni-rich Fe- though no probabilistic appraisal of the 

Ni sulfide ores with komatiites and other reliability of the estimate is currently 

ultramafic rocks rather than with bas- possible. 

altic rocks. The distribution coefficients To assist in the troublesome task of 

determined can now be used in conjunc- proofreading project card-image input 

tion with available numerical models to files, Chayes, Li, and Stewart have been 

simulate the fractionation of basaltic and developing two quite different forms of 

ultramafic magmas. machine-assisted verification. In one of 

For over 35 years, the Geophysical these, now in routine operation, the pro- 
Laboratory has maintained a program in cessor reports all departures from sys- 
systematic petrology that has been the tern grammar or vocabulary; in the other, 
envy of many geological institutions with the processor will compare various parts 
larger staffs and more-extensive re- of a specimen description with each other 
sources. As the program's goals near in a substantive consistency check. Most 
fruition, mainly owing to the efforts of a of the consistency checking is ultimately 
single staff member, a world data base to be incorporated in a new data-transfer 
for igneous petrology is now being de- program whereby all grammar will be 
signed and constructed by a group under supplied and all vocabulary verified as 
the International Geological Correlation the first machine-readable version of the 
Project. Important advances have been data is generated, 
made by Li and Chayes, who report that In the processing of petrographic data 
their FORTRAN-IV base-building pro- arrays characterized by very broad var- 
gram is now operating satisfactorily and iance ranges, the logarithmic transfor- 
has been used to generate a trial base mation is often used to linearize 
containing the 6644 specimen descrip- relationships and reduce variance differ- 
tions that have to date been transferred ences. Chayes points out that in arrays 
to machine-readable form from coding in which raw variances differ greatly in 
sheets submitted by volunteer contrib- size the logarithmic transformation also 
utors. frequently inverts variance ranks, so that 

Chayes has used this base in a model sample descriptions of variance relations 
study aimed at providing an inventory of obtained from transformed and untrans- 
published information bearing on rela- formed data may be sharply contradic- 
tions between mineral assemblage and tory. 
chemical and modal composition. In view 
of the critical importance of these rela- 
tions in many schemes for rock nomen- Layered Intrusions 
clature and classification, the major result 

of this first application of the base is rather The Skaergaard Intrusion, East 

surprising; of the specimen descriptions Greenland, continues to be a major source 

of igneous rocks published in major jour- of ideas regarding the crystallization of 

nals in the period 1916-1980, only one in magma. The vast amount of petrological 

and geochemical information being ac- 

/ v i i v cumulated sparks new directions for re- 

tK D Ni = I — — ) / I — - ) search and often results in the need for 

^ FeSy ^ meit ' ^ Fe0 / silicate meit more detailed mapping. Irvine has com- 



pleted a detailed field study of layering 
in the Skaergaard Intrusion, and he pre- 
sents two reports on this work, one de- 
scribing features that appear to be 
indicative of the general origins of the 
layering, the other giving new observa- 
tions on the characteristics and origin of 
the remarkable trough-layering struc- 
tures. In the first, he illustrates struc- 
tures in which it is evident that the 
common, thin, modally graded layers of 
the intrusion were formed by transport 
and deposition of crystalline materials by 
magmatic density currents, and he shows 
further that at least some of the thicker, 
modally differentiated macrolayers 
probably had similar origins, presumably 
from very large currents. He concludes, 
however, that much of the uniform rock 
that is interstratified with modally graded 
layers in the vicinity of the trough struc- 
tures probably formed by in situ crys- 

In his detailed mapping, Irvine reex- 
amined the previously recognized trough 
structures and discovered additional ma- 
jor and subsidiary troughs. He found that 
the troughs are not convergent, as was 
supposed by previous workers. Rather, 
they have only an apparent convergence 
at the present erosion surface, arising 
because some of the troughs shift later- 
ally with stratigraphic height. Irvine 
presents a graphic analysis of the troughs 
in formation, in which the ridges of mas- 
sive rock between the troughs are shown 
to form by crystallization of liquid con- 
vening in roller cells developed in a thin, 
slightly wedge-shaped liquid layer on the 
cumulate floor of the intrusion. The roll- 
ers are driven by the drag of wall-type 
convection cells in a thicker overlying 
layer of less-dense liquid and by the 
buoyancy effects of crystallization and 
upward heat loss. The trough layers 
themselves are deposited by density cur- 
rents of suspended crystals channeled 
between the ridges. As the bottom liquid 
layer is thinned by crystallization, the 
roller cells gradually shift laterally to- 
ward its thickest part, with the effect 

that the massive ridges and the troughs 
shift with stratigraphic height. 

The Bushveld Intrusion, Republic of 
South Africa, has major economic im- 
portance as the world's major resource 
of chromium, platinum, and other ele- 
ments; its study also contributes in a ma- 
jor way to the understanding of igneous 
processes. In an evaluation of a concept 
where separate and unrelated liquids are 
believed to have mixed to produce the 
Bushveld layered rocks, Sharpe and Ir- 
vine have determined the phase relations 
of 1-atm melting for two chilled margin 
rocks from the Bushveld Complex and 
for a complete series of mixtures be- 
tween them. The chilled margin rocks, 
one compositionally equivalent to an ol- 
ivine boninite and the other, to an alu- 
minous tholeiite, are presumed to be 
representative of these liquids. Field, 
petrographic, and chemical relationships 
indicate that the first belongs to a magma 
series that was parental to the ultramafic 
and noritic cumulates in the Bushveld 
Complex, whereas the second represents 
a different series that evidently yielded 
most of the anorthositic and gabbroic cu- 
mulates. A particularly interesting fea- 
ture of the melting data is that in mixtures 
of the two rocks, the crystallization tem- 
peratures of their respective primary sil- 
icates (olivine and plagioclase) are 
depressed in a eutectic-like relationship, 
whereas that of chromite varies almost 
linearly between its points of appearance 
in the end-member melts. Sharpe and Ir- 
vine point out that this relationship is a 
basis for forming chromitite by mixing 
the two magma series. 

The concept referred to above depends 
in part on knowledge of the viscosity and 
density characteristics of the melts of two 
Bushveld chilled margin rocks. Quanti- 
tative data were obtained by Sharpe, Ir- 
vine, My sen, and Hazen for each rock 
by the falling-sphere method at 5, 10, and 
20 kbar to temperatures as much as 250°C 
above the liquidus. Several innovative, 
additional experiments were performed 
that yielded unequivocal qualitative in- 


formation on the relative densities of the findings to the practical aspects of phase 

melts. Among the quantitative results, equilibria. 

viscosities decrease at liquidus temper- Virgo, Mysen, and Danckwerth pro- 

atures from a few hundred poises at 1 vide a general pattern of the structural 

atm to 25-35 poises at 20 kbar, and den- states of Fe 3 + as a function of both com- 

sity values increase from about 2.6 to 2.9 position and f in the simple system 

g cm" 3 . At constant temperature, the Na 2 0-Si0 2 -Fe-0. They found that under 

viscosity values are almost unaffected by oxidizing conditions Fe 3+ substitutes for 

pressure; at constant pressure, coeffi- Si with four bridging oxygens, and Fe 3 + 

cients of thermal expansion are about an is partitioned between two different Fe- 

order of magnitude larger than those in- rich anionic units depending on the bulk 

dicated by the currently best available composition of the melt. They suggest 

computational methods for estimating that the distinct bands at 980 and 900 

melt densities. In the qualitative exper- cm" 1 due to Fe 3+ found in the Raman 

iments, pairs of compositions were melted spectra for low and high Na 2 0/Si0 2 

together in the same capsule, one at the glasses, respectively, are indicative of 

top, the other at the bottom. After the Fe0 2 1_ species with different average 

melts were quenched, their distributions numbers of nonbridging oxygens. Under 

were determined by mapping the charge reducing conditions the structural state 

with the electron microprobe. Clear ev- of Fe 3+ is complex because the ion not 

idence that the melts inverted position only is partitioned between different Fe- 

was found in those cases where the initial rich anionic units but also occurs as a 

arrangement was gravitationally unsta- network modifier. No evidence is found 

ble, and it appears that, by this method, for an association of Fe 2+ and Fe 3+ in 

density differences so small as to be in the form of a charge-balanced complex 

the third decimal place can be detected of the type Fe 2 + Fe 2 3 + 4 or clusters of 

for basic melts. The results of these ex- Fe 3 4 . 

periments are in complete accord with Next, the same investigators deter- 

the process of double-diffusive convec- mined the structural state of Fe 3+ in ox- 

tion proposed by Irvine et al. in Year idized vs. reduced aluminosilicate glasses 

Book 81 , whereby magmas like the Bush- that contain different network modifiers, 

veld chilled liquids might undergo con- such as Na, Ca, and Mg. They found that 

current crystallization and mixing. i n each of the different base-glass com- 
positions, Fe 3+ is tetrahedrally coordi- 
nated under oxidizing conditions (1550°C, 

Structure and Properties of Liquids f ^ }®~\ at , m) but that at low /o 2 > Fe ^ 

nwd rintmo* 1S octahedrally coordinated in the melt 

ana masses (1550°C, f 0z - 10" 6 - atm). Mixture of 

The oxidation state of iron in a magma both structural states at intermediate 

is especially important because of the values of f is anticipated. Despite the 

variable roles ferrous and ferric ions play change in coordination environment of 

in the structure of a liquid. Ferrous iron Fe 3+ as a function of/o 2 , the depend- 

is a network modifier, whereas ferric iron ences of log Fe 2+ /Fe 3 + on log/o 2 are lin- 

can be both a network former and a net- ear in each of the different systems, as 

work modifier. Small changes in these expected, and the slopes of the correla- 

roles greatly influence the phase equilib- tion are approximately - l A. It is pro- 

ria and physical properties of silicate posed that these findings will have a 

melts. A major effort was made this year significant influence on the interpreta- 

to integrate 57 Fe Mossbauer and Raman tion of certain physical properties of melts, 

spectroscopy to characterize the roles of such as crystal-melt partition coefficients 

the iron ions in simple systems. A group that are linearly related to the number 

of investigators examined the effects of of bridging and nonbridging oxygen ions 

individual variables and then applied their in the structure. 


*The model of Fe-cation-site distri- Virgo suggest that the pyroxene-feld- 

butions presented last year has been ex- spar liquidus boundaries, for example, in 

plored furt her by Danckwerth and Virgo, natural magmas are sensitive to Fe 3+ / 

A comparison of the Mossbauer spectra SFe, expanding the stability field of feld- 

of crystalline and glassy CaFeSi 2 6 shows spar with decreasing Fe 3 + /2Fe of the liq- 

that the individual coordination poly- uid. 

hedra for Fe 2 + in silicate glass may be After the above variables were dealt 
more regular than those in the corre- with, the alkaline earths were chosen for 
sponding crystal. In addition, they show charge-balance of tetrahedrally coordi- 
proofthat ferric iron exhibits amphoteric nated Fe 3+ because their natural abun- 
character in Na 2 0-Al 2 03-Si0 2 glasses de- dance indicates that they are the principal 
pending on the partial pressure of oxy- charge-balancing cations of amphoteric 
gen. They draw the extremely important, oxides (Fe 3+ and Al 3 + ) in magmatic liq- 
if not surprising, conclusion that mod- uids. From Mossbauer and Raman spec- 
eling the solution mechanisms of mixed- troscopic studies, My sen and Virgo found 
valence cations on experiments con- that the Fe 3+ /SFe affects the structural 
ducted in air is likely to have little or no position of Fe 3+ . With less than about 
relevance for natural systems of geologic 50% Fe 3+ (relative to total iron content) 
interest. in the iron-oxide abundance range of nat- 
The structural positions of Al 3+ and ural magmatic liquids, Fe 3+ shows a 
Fe 3 + in natural magmatic liquids are im- transition from tetrahedral to octahedral 
portant because of their abundance and coordination. The Fe 3+ /SFe correspond- 
because the polymerization and distri- ing to this transition appears indepen- 
bution of anionic units in silicate melts dent of the NBO/Si and the type of 
depend on the relative stabilities of alkaline-earth, network-forming cations 
charge-balanced aluminate and ferrite in the melt. The value of Fe 3+ /SFe de- 
complexes. The stability of these com- creases slightly with increasing iron con- 
plexes relates to the type of charge-bal- tent of the melt. The coordination 
ancing cations, A1/(A1 + Si) and Fe 3+ / transformation results in depolymeriza- 
(Fe 3+ + Al +. Si), of the magma. My sen tion of the melt. The spectroscopic data 
and Virgo used the system Na 2 0- A1 2 3 - indicate that network- forming Fe 3 + and 
Si0 2 to isolate the effects of A1/(A1 + Si), Fe 2+ show a preference for nonbridging 
temperature, and degree of polymeri- oxygens in the most depolymerized an- 
zation (NBO/7 1 ). Mossbauer spectro- ionic units in the melts (Si0 3 2 ~, Si 2 7 6 ", 
scopic data show that Fe 3+ /2Fe is a simple and Si0 4 4 ~ ). Because of the increased ac- 
function of NBO/7 1 , with the dependence tivity of these anionic units, the reduc- 
on NBO/T increasing with increasing Al/ tion and coordination transformation of 
(Al + Si) of the melt. (The Al 3+ is in ferric iron results in expansion of the li- 
tetrahedral coordination in all these quidus phase fields of olivines and py- 
melts.) There is also a linear correlation roxenes in natural magma. Mysen and 
of redox ratio with Al/( Al + Si) at the Virgo also suggest that Fe 2 + -Mg 2 + par- 
same NBO/Si and with oxygen fugacity titioning between ferromagnesian sili- 
at the same A1/(A1 + Si) and NBO/Si. cate minerals and silicate melts will be 
The Fe 3+ /SFe range corresponding to the affected by the changes in Fe 3+ /SFe. 
tetrahedral-octahedral coordination Finally, Mysen and Virgo employed 
change appears to be insensitive to the Mossbauer and Raman spectroscopy to 
Al content and degree of polymerization determine the structure of iron-bearing 
of the melt. From these data, Mysen and glasses produced at high pressure. They 

found that the reduction of ferric to fer- 
rous iron at pressures above 10 kbar in 
melts of disilicate composition appar- 

*Abstract only. Article available in open file on ently results from a gradual transfor- 

request. mation of Fe 3+ from tetrahedral to 


octahedral coordination. The pressure increases, whereas that in CaMgSi 2 6 
corresponding to the onset of coordina- melt decreases with increasing pressure, 
tion transformation and the pressure The pressure dependence of oxygen dif- 
range over which the coordination change fusivity in these melt compositions is, 
occurs increase with increasing iron con- therefore, inversely correlated with vis- 
tent of the system. The decreased po- cosity. They also observed that the ac- 
lymerization thus affected results in a tivation energy of oxygen diffusion is 
rapid decrease in relative abundance of similar to that of viscous flow. Shimizu 
three-dimensional and sheet units and an and Kushiro conclude, therefore, that the 
increase in chain and dimer units in the unit that diffuses during viscous flow is 
melts. indeed the individual oxygen ion. Fur- 
As a result of this process, Mysen and thermore, they suggest that the negative 
Virgo conclude that liquidus fields of de- activation volume for NaAlSi 2 6 melt re- 
polymerized minerals such as olivines and suited from a transient collapse of the 
pyroxenes will expand relative to those three-dimensional melt structure asso- 
observed in iron-free systems. Thus, dated with the diffusive jump of the spe- 
fractional crystallization of iron-rich nat- cies. In CaMgSi 2 6 melt the activation 
ural magma at high pressure (deep crust volume is greater than the volume of the 
and upper mantle) will presumably result oxygen ion. Thus, positive dependence 
in silica-enrichment trends. They also of the diffusivity and viscosity would be 
found that the viscosity and density of expected. 

iron-rich magmatic liquids are more sen- In other processes, the diffusion of Si 
sitive to pressure than those of iron-free and Al contribute to the rate-determin- 
melts with the same degree of polymer- m g steps. Kushiro investigated the in- 
ization. terdiffusion of AfcGa and Si^Ge in melts 
*Danckwerth and Virgo have con- of NaAlSi 2 6 composition between 6 and 
ducted an interlaboratory comparison of 20 kbar. He found that the SfeGe in- 
the determination of Fe 2+ /Fe 3+ of nat- terdiffusion coefficient decreases by a 
ural glasses by wet chemistry and by factor of 8-18 and the Ga^Al interdif- 
Mossbauer spectroscopy. They show that fusion coefficient, by a factor of 3-5, in 
at 298 K the two methods are generally the pressure range 6-20 kbar. These re- 
in agreement to within 10% for the var- suits accord with those from molecular 
ious methods of fitting Mossbauer spec- dynamics calculations on similar melt 
tra with lines of Lorentzian shape and compositions. Kushiro notes that these 
are virtually identical when fitting mod- diffusion coefficients show an inverse 
els that employ distributions of the hy- correlation with the viscosity of Na 
perfine fields. The additional structural AlSi 2 6 melt. He suggests that the un- 
information about the coordination en- deriving cause of these changes in trans- 
vironments of iron leads them to believe port parameters is a pressure-induced 
that experiments conducted in air may weakening of T-O-T bonds in the three- 
not be applicable to geologically relevant dimensionally interconnected alumino- 
(i.e., more-reduced) systems. silicate melt network. This interpreta- 
Because oxygen is the principal con- tion is supported by published data on 
stituent of melts and its transfer is the the structure of similar aluminosilicate 
key to most redox phenomena, Shimizu melts at high pressure. The significant 
and Kushiro determined by means of an increase in diffusivity of network-form- 
ion microprobe the oxygen diffusion coef- ing cations at high pressure means that 
ficients for melts of contrasting degrees diffusion data obtained at 1-atm pressure 
of polymerization (CaMgSi 2 6 and can be applied to upper mantle magmatic 
NaAlSi 2 6 ) as a function of pressure. The processes only after correction for pres- 
diffusivity of oxygen in NaAlSi 2 O fi melt sure effects. 



Petrology and Geophysics of the 

Much of the evidence for the mineral 
content of the mantle comes from the 
nodules brought up in kimberlite pipes. 
To these are added the suites of nodules 
found in alkali basalts and the ultramafic 
bodies of deep-seated terranes. As more 
specimens are found, the variation of as- 
sociations and alterations being docu- 
mented appears to expand. The 
complexity of the mantle is probably no 
less than that of the crust. 

Boyd, Jones, and Nixon describe dun- 
ites that have been metasomatized at 
depth in the mantle by introduction of 
Fe, Ti, Al, Ca, K, Na, and possibly Cr 
prior to eruption as xenoliths in kimber- 
lite. The metasomatism occurred in con- 
cert with deformation. Fine-grained, 
recrystallized olivine and some coarse ol- 
ivine in the dunites have been changed 
to a more-Fe-rich composition (Fo 87 ) than 
the primary range of the relict porphy- 
roclasts (Fo 93 _ 89 ). Secondary phlogopite 
and Fe-Ti oxides have crystallized as 
groundmass phases, and diopside may also 
have been introduced. The association of 
metasomatic alteration with a deforma- 
tion characteristic of conduit formation 
and the early stages of eruption is evi- 
dence that the kimberlite magma itself 
was the source of the metasomatizing 

Bakun-Czubarow has investigated the 
mineral chemistry of small bodies of gar- 
net lherzolite occurring in granulites in 
the Middle Sudetes of southwest Poland. 
The rocks occur as small (1-2 m) boudins 
and as part of a larger (ca. 200 m) ultra- 
mafic lens, and there has been consid- 
erable controversy about whether they 
and similar European rocks might have 
been emplaced directly from the upper 
mantle or could have formed in the crust 
under conditions of high-grade regional 
metamorphism. The samples investi- 
gated consist of olivine (~Fo 91 ), chrome 
diopside, and Cr-Al-enstatite, plus sub- 

ordinate amounts of three alumina-rich 
phases — pyrope, chrome spinel, and so- 
dian pargasite. In the garnet grains Mg 
decreases, and Fe, Ca, and Mn increase, 
from core to rim. Both pyroxenes are 
slightly zoned, with Al decreasing, and 
Fe increasing, outward. Bakun-Czuba- 
row has attempted to estimate the range 
of pressures and temperatures under 
which the rocks formed, by assuming that 
the cores and rims of the associated min- 
eral grains represent different conditions 
of equilibration. The indicated conditions 
range from 1030°C at 27 kbar to 800°C 
at 16 kbar. From these results she infers 
that the garnet lherzolites were probably 
pieces of upper mantle material formed 
under the first set of conditions. She sug- 
gests that the bodies were then emplaced 
mechanically during a major tectono- 
metamorphic event, under conditions 
governed by the second set, and finally 
were rapidly uplifted into their present 
geological setting. 

Many types of mantle-derived xeno- 
liths found in kimberlites, including gar- 
net peridotites, eclogites, and even 
diamonds, have at one time or another 
been considered fragments of subducted 
oceanic lithosphere. Schulze may have 
found another possible example in the 
graphically intergrown rutile and olivine 
occurring as xenoliths in the Jagersfon- 
tein and Kampfersdam kimberlites in 
South Africa. Although texturally simi- 
lar to graphic pyroxene-ilmenite inter- 
growths, usually considered igneous 
rocks, the rutile-olivine intergrowths also 
bear a resemblance to intergrowths of 
olivine and ilmenite that are a result of 
prograde metamorphism of titanoclino- 
humite, a minor phase in Alpine serpen- 
tinites. Schulze concludes that, if these 
rare nodules did indeed form from sub- 
ducted titanoclinohumite, some of the 
more common peridotite nodules could 
well be of similar origin. 

Pyroxenite nodules are important be- 
cause variation in their abundance in dif- 
ferent kimberlites probably reflects real 
differences in the constitution of the 


mantle. Boctor, Boyd, and Nixon stud- ity infrared spectra obtained in situ at 

ied a suite of pyroxenites, eclogites, and high pressure. The new pressure coeffi- 

bronzite megacrysts from the De Bruyn cients of spectral line position are the 

and Martin Mine, Bellsbank, South Af- first to be obtained for these minerals, 

rica. They found that garnet websterite It should now be possible to combine data 

xenoliths equilibrated at 1065°C at a depth that were measured in the mid-infrared 

of 160 km, whereas the bronzite mega- region with those in the far-infrared to 

crysts equilibrated at shallower depths calculate thermodynamic properties, 

of 130-140 km and at lower temperatures *Data at high pressure on the com- 

of 810°-950°C. These depths are shal- pressibility of pyrite (FeS 2 ) were ob- 

lower than those at which kimberlite tained for the first time under both 

magma is generated beneath the Kaap- nonhydrostatic and hydrostatic condi- 

vaal craton or the depth range calculated tions by Jephcoat, Mao, and Bell. The 

from discrete-nodule equilibration. On the results provide a comparison of two pres- 

basis of the wide range of equilibration sure-transmitting media. The data may 

temperatures of the pyroxenites, eclo- be useful in evaluating the role of sulfur 

gites, and megacrysts (810°-1140°C), they as a component in the deep earth, 
conclude that nodules from different lev- 

els in the lithosphere have been sampled Q { Structures at mh Pressures 

by erupting kimberlite and that bronzite ° * 

megacrysts apparently were derived from An important long-range research ob- 

disaggregated lherzolite or pyroxenite jective at the Geophysical Laboratory is 

nodules. to gain an understanding of the role of 

*Clinopyroxene and phlogopite me- crystal structure in controlling physical 
gacrysts from the He Bizard diatreme and thermochemical properties of min- 
are compositionally distinct from those erals. The system BeO-Al 2 3 -Si0 2 -H 2 
in kimberlites. Garnet, ilmenite, and or- is well suited to a study of structure vs. 
thopyroxene are absent, and the clino- properties because of the relatively large 
pyroxenes are aluminous (3-7 wt%Al 2 3 ), number of stoichiometric, well-crystal- 
yet less so than those from alkali basalts, lized minerals that are available. As part 
Schulze found that they have a strong of an ongoing study of this system, Ha- 
trend of Fe enrichment with increasing zen, Finger, and Barton determined the 
Ca/(Ca + Mg), unmatched by megacryst compressibilities and high-pressure 
suites other than that of an alnoite from structures of beryl (Be 3 Al 2 Si 6 18 ), eu- 
Malaita in the Solomon Islands. Al- clase (BeAlSi0 4 OH), and bertrandite 
though Mg-Ti magnetite is also a com- [Be 4 Si 2 7 (OH)2]. Each of these minerals 
mon megacryst at He Bizard, the strong is characterized by three-dimensional 
Fe-enrichment trend of clinopyroxene networks of tetrahedrally coordinated Be 
may indicate that these two minerals did and Si. In all three structures the be- 
not precipitate together. ryllium tetrahedra are more compressi- 

Studies of minerals with the olivine, ble than the silicon tetrahedra, and the 

pyroxene, and silicate-perovskite struc- relative linear compressibilities are thus 

tures require a broad variety of experi- dependent on the distribution of these 

mental data to relate their physical and two types of four-coordinated polyhedra. 

thermal properties to seismic and other *The phase-transformation mecha- 

geophysical data on the earth's mantle, nisms in mantle silicates at high pressure 

New results by Xu, Mao, Weng, and Bell are often deduced by studying transi- 

on the olivine end members, forsterite tions in other ionic compounds. Hazen 

and fayalite, and on the pyroxene hy- and Finger investigated the compression 

persthene, as well as those by Weng, Xu, and structural changes of beta-LiI0 3 , 

Mao, and Bell on silicate-perovskite of which undergoes a reversible transition 

MgSi0 3 composition, include high-qual- from tetragonal to monoclinic symmetry 


at 50 kbar. This compound has important channels and pores, thus preventing 

bearing on the transitions in the domi- compression of the "open" portions of the 

nant mantle mineral, perovskite, which zeolite. Larger molecules, on the other 

has similar complex polyhedra. Com- hand, do not enter the crystal and thus 

pressibility of the tetragonal phase is un- compress both framework and channels. 

usual, with the c axis expanding between 

1 bar and 20 kbar. With increasing pres- n ... , D7 , -, A/r , -. 

. i-i • Composition of Planets and Meteorites 

sure, c-axis compression steadily in- r J 

creases, whereas a-axis compression The composition of the mantle of Mars 
decreases; a strong coupling between is constrained most directly by infer- 
compression of a and c, therefore, is in- ences drawn from its density. Significant 
dicated. Behavior of LiI0 3 is consistent seismic data do not exist for the interior 
with a second-order ferroelastic transi- of Mars, and analyses of mantle-derived 
tion. martian rocks, with the possible excep- 
The synthetic compound CaCrSi 4 O 10 is tion of several rare types of achondritic 
a new analog of gillespite, BaFeSi 4 O 10 , meteorites that may have originated on 
which is unusual both for its square-planar Mars, are not yet available. The density 
iron and for a high-pressure phase tran- of the mantle of Mars is, therefore, con- 
sition that is characterized by a striking strained only by inferences drawn from 
red-to-blue color change. Hazen studied the mean density and moment of inertia 
the compressibility and high-pressure of the planet. Goettel determined the 
pleochroism of this new compound and range of mantle densities that are corn- 
observed dramatic, but continuous, patible with these factors by computing 
changes in color to 50 kbar. Measure- more than 125 models for the density dis- 
ments of unit-cell parameters to 40 kbar tribution within Mars in order to explore 
revealed a continuous variation in tetra- systematically the present uncertainties 
gonal dimensions, with no indication of a about the density of the mantle. The zero- 
phase transition. The observed axial pressure density of the mantle of Mars 
compression ratio of c to a is 2.4, which was estimated to be 3.44 ± 0.06 g/cm 2 , 
is identical with that observed in gilles- which is consistent with the solar pro- 
pite; structural changes with pressure are portions of rock-forming elements with 
thus assumed to be similar in the two about 30% of the total Fe present in the 
compounds below the transition in gil- mantle. This density is also consistent 
lespite. with a range of other composition mod- 
It is usually assumed in diamond-anvil, els. Goettel places the probable value for 
pressure-cell studies that hydrostatic the FeO content of the martian mantle 
pressure fluids do not interact with min- between 13 and 17 wt %, and in his view 
eral samples. Zeolite molecular sieves, this high value is the principal difference 
which have large structural channels and between the terrestrial (~8 wt % FeO) 
cavities, are exceptions to this assump- and martian mantles. Needless to say, 
tion. In an ongoing investigation of the better definition of the composition of the 
high-pressure behavior of zeolite min- mantle of Mars requires seismic data for 
erals, Hazen and Finger have observed the martian interior and analyses of man- 
the compressibility of zeolite sieve 4A in tie-derived martian rocks, 
a variety of fluid media. In pure water Because of the major importance of hy- 
the compressibility of zeolite is compa- drogen in planet genesis and its contin- 
rable to that of many close-packed oxides ued presence within the interiors of some 
and silicates. In fluids with larger mol- planets, knowledge of the properties of 
ecules, such as glycerol or organo-fluo- hydrogen is in great demand. Mao, Xu, 
rine compound FC-75, the compressibility and Bell obtained high-quality infrared 
is more than six times greater. Smaller spectra of solid hydrogen in the pressure 
water molecules are assumed to enter range 55-60 kbar at 25°C. It was re- 


markable that hydrogen, which normally the purple variety of ringwoodite may be 

is not infrared active, had strong, well- present in more than one valence state 

resolved absorption spectra, apparently (e.g., Fe 2+ and Fe 3 + ) and that the purple 

the result of a collision-induction process color may be due to intervalence, charge- 

at high pressure. The results are of un- transfer absorption, a process that has 

usual value in determining the thermo- been found to be responsible for color 

dynamic properties of hydrogen in this changes in other minerals. No optical ab- 

extended pressure range. Thermody- sorption or Mossbauer spectral data are 

namic quantities were extracted from the available on natural ringwoodite because 

infrared spectral shifts as a function of of its scarcity and minute grain size. 

pressure in combination with Raman and Spectroscopic studies on synthetic ring- 

Brillouin-scattering data already ob- woodite doped with various transition 

tained. metals would indeed help clarify this 

*The mineralogy and composition of the problem. 

meteorites have a great influence on the 

construction of models of the interiors of ,,. 7E7/ . , „ .,., . 

, , A i i . + ,.., Mineral-r luid Equilibria 

planets. An unusual polymict enstatite * 

achrondrite from Cumberland Falls, The solution chemistry of minerals has 
Kentucky, that contains chondritic clasts advanced to the stage where computa- 
was studied by Boctor and Lee. They note tion of the properties and behavior of 
that the clasts contain olivine and low Ca- complex solutions is now possible. These 
pyroxene that differ in composition from computations, if they are to be more than 
those of ordinary chondrites. Both host just a "house of cards," must be based 
and clasts show various degrees of shock on accurate experimental observations, 
metamorphism, such as deformation la- In a long-term study of the behavior of 
mellae in enstatite, formation of clinoen- acids and bases in supercritical aqueous 
statite, and local vitrification to silica-rich fluids, Frantz and Marshall have as one 
mafic glass. The isotopic ratios for Nd of their objectives the determination of 
and Sm are similar in both the achon- the hydrolysis constants of such salts as 
dritic host and the chondritic clasts and KC1, NaCl, NaF, and KF. These con- 
vary within 1% from those ratios in chon- stants are obtained by measuring the 
drites. It is evident that a chondritic me- electrical conductivity from which the 
teorite has impacted an achondritic parent ionization constants for the correspond- 
body; however, unlike the oxygen iso- ing acids and bases are obtained, 
tope data, the similarity of the Nd-Sm During the past year the electrical con- 
ratios does not provide supportive evi- ductances of 0.002-, 0.005-, and 0.010- 
dence that the host and clast are of dif- molal hydrochloric acid solutions were 
ferent origins. measured by Frantz and Marshall at 
*In L6 chondrites, Boctor found that temperatures between 100° and 700°C and 
ringwoodite, the high-pressure poly- at pressures to 4000 bars. The results of 
morph of olivine, occurs only in shock the measurements were then used to 
veinlets. It is represented by colorless compute limiting equivalent conduct- 
and purple varieties that are similar in ances and ionization constants for the acid, 
their major-element chemistry. Boctor The limiting equivalent conductances were 
studied the distribution of Ni, Co, and found to vary linearly with density and 
Mn, which occur as trace elements in to be temperature dependent above 350°C. 
ringwoodite, by a special, trace-element, The molar ionization constants were found 
computer program on the electron mi- to vary linearly with respect to inverse 
croprobe. He found that the concentra- temperature and the logarithm of the 
tions of these trace elements do not vary density of water. Frantz and Marshall 
significantly in either the colorless or the have used recent information on limiting 
purple variety. He suggests that the iron equivalent conductances to recompute 
and possibly other transition metals in ionization constants of KC1. The ioniza- 


tion constants of KC1, like those of HC1, of organic matter on earth. Although it 
were found to vary linearly with respect is formed by the reaction of organic 
to inverse temperature and the loga- chemicals derived from once-living or- 
rithm of the density of H 2 0. This infor- ganisms, kerogen has undergone exten- 
mation is important not only in the sive structural changes during its 
interpretation of the behavior of ore- formation and subsequent diagenesis. In 
forming fluids but also in the understand- general, its molecular constitution is 
ing of corrosive processes in steam-gen- poorly known, yet kerogen exhibits cer- 
eration plants. tain molecular features, the so-called mo- 
Fluorine is the most abundant halo- lecular fossils, that can be related back 
gen, and fluoride complexes may play a directly to its biochemical precursors. In 
significant role in some aqueous fluids, a continuing study of the molecular con- 
such as those involved in the formation stitution and reactivity of kerogen with 
of ore deposits rich in lithophile ele- isotopic and molecular tracers, Hoering 
ments. Barton and Frantz have begun has used the novel technique of pyrolysis 
study of fluoride complexing by deter- in the presence of excess liquid water to 
mining the partitioning of sodium and po- split kerogen into fractions of lower mo- 
tassium between coexisting alkali lecular weight that are amenable to anal- 
feldspars and a fluoride fluid at temper- ysis by the gas chromatography-mass 
atures from 450° to 600°C. From these spectrometer system, which was put into 
results the relative free energies of operation during the past year. Hoering 
aqueous sodium fluoride and potassium has posed two questions: first, does hy- 
fluoride have been determined. Under drous pyrolysis of kerogen give sufficient 
the conditions of the experiments the ex- yield of identifiable molecular fossils, and 
change properties of alkali fluoride so- second, how does the molecular compo- 
lutions are similar to those of alkali sition of the pyrolysate compare with that 
chloride solutions. of natural petroleums? 

Barton has done calculations on spe- The answer to the first question is 
ciation in mineral assemblages in equi- clearly affirmative. Normal isoprenoid, 
librium with carbon-oxygen-hydrogen- methyl-branched, sterane and triterpane 
sulfur fluids, treating the components of hydrocarbons, whose structure can be 
the fluid as independent variables. These related to known constituents of the lipid 
calculations show that when mass-bal- fraction of living organisms, were pro- 
ance constraints are considered, geolog- duced. Metallo-porphyrins that are di- 
ically reasonable amounts of fluids can rect descendents of the green pigment 
have in some occurrences the dominant chlorophyll are easily detectable. His an- 
role in determining the modes and com- swer to the second question is much more 
positions of coexisting solids. It is sug- guarded. The relative proportions and 
gested that relatively small amounts of molecular-weight distributions of the 
fluid may be the controlling factor in re- constituents generated by hydrous pyr- 
dox reactions (and hence mineral assem- olysis of kerogen have their counterparts 
blages) for some felsic intrusives and a in many natural petroleums. Crude oil is 
factor in prograde metamorphism. The such a complex material, however, that 
sequences of mineral associations seen in simple comparison is inadequate. Never- 
some skarns are consistent with cooling theless, there are sufficient similarities 
of fluids of approximately constant com- to suggest that the catagenetic processes 
position. taking place during hydropyrolysis at el- 
evated temperatures in the laboratory 

D . , . .. are a reasonable, first approximation to 

Bioqeochemistrii ,, , u , , , ' , *T, 

* a those that take place at low tempera- 

The insoluble, high-molecular-weight tures in natural environments. Thus, 

organic matter of sedimentary rocks, tracer and molecular probe experiments 

called kerogen, is the largest reservoir appear to provide a method for under- 


standing the structure, reactivity, and nitrogen was enriched in 15 N (up to 

evolution ofkerogen over geological times. +16%o) when the bacterium was cul- 

The source of the organic matter that tured on either glutamic or aspartic acid, 
eventually yields kerogen was sought by These amino acids enter the metabolic 
Macko. He analyzed the organic com- pathways directly, and the remainder of 
ponents of surficial sediments collected cellular nitrogen is derived via trans- 
along the northeast coast of the United amination. Other amino acids (e.g., ala- 
States for the stable isotopes of nitrogen nine and serine) are immediately deam- 
and carbon. These isotopes serve to iden- inated, and free ammonia is then 
tify the sources of the organic matter in transferred to glutamic acid. Bacteria that 
marine sediments. Sediment sample sites have been cultured on alanine or serine 
included three major bay systems, the are depleted in 15 N by up to - 12%o. The 
New York Bay and Hudson Canyon, the results are important in determining iso- 
Delaware Bay, and the Chesapeake Bay, tope effects in respiration and biosyn- 
as well as the associated continental shelf thesis. Moreover, the results could be 
and slope. A highly significant correla- used to demonstrate that microbes can 
tion was observed between the organic greatly alter the isotopic composition of 
nitrogen and carbon isotopic composi- organic matter during and after deposi- 
tions of the sediments. Results from each tion of the sediment, 
of the three bays, however, had a unique Distinct biochemical pathways for ni- 
position on the correlation curve. The New trogen in living organisms can be de- 
York Bay system generally showed a tected by stable isotope tracers and may 
simple mixing line between two readily be useful in the interpretation of kerogen 
resolvable end members — a marine com- from ancient rocks. For this reason Estep 
ponent and a terrestrial source. All the extended her work on recent stromato- 
stations in the Delaware Bay closely re- lites in hot springs to include the study 
sembled those farther offshore on the of nitrogen metabolism and biochemis- 
continental shelf and slope. The Chesa- try. Natural variations in the inorganic 
peake Bay sediments were more en- environment of the hydrothermal areas 
riched in 13 C and 15 N than were all other affect the cycling of nitrogen through or- 
locations. These results can be used to ganic matter. In general, hot springs have 
indicate the relative accessibility of an very low concentrations of nitrate and 
estuarine system to inputs from the ma- ammonia, the combined nitrogen forms 
rine environment and potentially could available to algae and bacteria. Corn- 
indicate the degree of recycling of or- monly, the nitrogen isotopic composition 
ganic matter occurring within a more re- of stromatolites in neutral pH springs with 
stricted environment. a constant flow of water was similar to 

The organic material that reaches the that of organisms growing in a nutrient- 

seafloor almost immediately undergoes limited steady state. In quiescent pools, 

change through the action of microbes, however, the isotopic composition of sim- 

One of these microbes is Vibrio harveyii, ilar types of organisms was distinctly dif- 

a marine, aerobic, heterotrophic bacter- ferent and may possibly indicate the 

ium. Macko and Estep cultured this bac- importance of internal recycling of am- 

terium on amino acids or amino sugars, monia. The source of inorganic nitrogen 

to learn whether it fractionated the iso- in hydrothermal waters is unknown. Ni- 

topes of nitrogen and carbon. The C-N trogen from the atmosphere is reduced 

ratio of the substrate as well as the bio- to ammonia and metabolized to amino 

synthetic and metabolic pathways are acids by nitrogen-fixing organisms such 

important in determining the isotopic as blue-green algae. Stable nitrogen iso- 

composition of the organism. The bac- tope ratios of these organisms indicate 

terium did indeed fractionate the iso- that this reaction is occurring in hot 

topes uniquely. For example, bacterial springs not only in blue-green algal mats, 


but at higher temperatures (up to 60°C) lagen, however, and as fossils undergo 
by a nonphotosynthetic bacterium. Other diagenesis, the noncollagenous proteins 
factors governing the nitrogen isotopic are more persistent. Hare and Estep de- 
composition of the organic matter include termined the 8 18 C and 5 1;) N of both es- 
the concentration of combined nitrogen sential and nonessential amino acids in a 
and the presence of bacteria that derive modern collagen. The essential amino 
their cellular energy from converting acids had quite variable isotopic compo- 
ammonia to nitrite. sitions, which could be related to an ex- 
Additional studies on the isotopic com- cess or deficiency of these amino acids in 
position of nitrogen and carbon were un- the animal's diet. The stable isotopic 
dertaken by Macko, Estep, Hare, and composition of nonessential amino acids 
Hoering on individual amino acids iso- from the modern collagen were nearly 
lated from cultured microorganisms. Two identical with those of nonessential amino 
classes of prokaryotic organisms were acids of a 10,000-year-old fossil bison bone 
cultured on specific inorganic and organic consisting primarily of collagen. The re- 
substrates. Blue-green algae were raised lationship of carbon and nitrogen iso- 
on molecular nitrogen, nitrate, and am- topes in the total protein hydroly sate and 
monia in the presence of excess carbon OH-proline, an amino acid unique to col- 
dioxide. A bacterium, Vibrio harveyii, lagen, exhibited large differences be- 
was grown on single, pure amino acids tween the modern and fossil proteins. The 
for its only source of nitrogen and car- acid-insoluble proteins of fossil bones 
bon. These cultures were then hydro- contain considerable amounts of chemi- 
lyzed, and the individual amino acids were cally and isotopically distinct "humic acids 1 ' 
isolated and isotopically analyzed by a that have 8 13 C values 5-7%o different from 
method similar to that of Abelson and acid-soluble protein hydrolysates. Com- 
Hoering (Year Book 59, 158-165). The bination of the two techniques of amino 
isotopic compositions of the amino acids acid biogeochemistry and stable isotope 
have a wide range of values, but many analysis has resulted in a new approach 
appear to be associated with kinetic frac- to the study of fossils, 
tionations that follow known metabolic In another unusual combination of 
pathways of amino acids. A better un- techniques, Gize and Rimmer made a pe- 
derstanding of the biosynthesis of amino trographic study of a bitumen that had 
acids and the importance of transami- been heated naturally by diabase dikes, 
nation reactions in organisms may result The bitumen in the host rocks exhibits a 
from this research. Additionally, the rea- lamellar texture with banding on a sub- 
sons for the observed natural abun- micrometer scale. Close to the dikes, an- 
dances of 15 N and 13 C in living organisms isotropic bodies were observed to have 
and organic deposits such as coals and grown from an isotropic pyrobitumen. The 
oils may be explained. similarities between the observed tex- 
Hare and Estep have extended the tures and those reported during the pro- 
studies of the isotopic composition of in- duction of coke from coal strongly suggest 
dividual amino acids to include modern that the anisotropic bodies are an aro- 
and fossil proteins from bones. Collagen matic mesophase (liquid crystals). Their 
is the major structural protein in modern report, the first of a natural occurrence 
bones and has a distinct amino acid com- of mesophase, indicates at least one route 
position that includes high levels of gly- by which sedimentary organic matter can 
cine, glutamic acid, and proline. The be transformed into graphite, 
carbon and nitrogen isotopic composi- 

tions of total bone proteins, notably col- N A tus and Techniques 
lagen, have been used by archaeologists 

in delineating the diets of prehistoric man. Infrared spectra at high pressures have 

Not all of the proteins in bone are col- long been considered essential in relating 


thermodynamic properties of minerals and a position-sensitive detector system to 
solidified planetary gases to their elastic the diamond-window, high-pressure cell, 
geophysical parameters. The application Volume-compression data of minerals are, 
of laboratory experimental data to geo- perhaps, the most essential physical pa- 
physical observations has great potential rameters for deriving geophysical equa- 
in studies of the inaccessible mantles and tion-of-state models of the earth. The 
cores, particularly of the earth, but also demands of experimental accuracy and 
of the moon and planets. The recent de- precision of compressibility are severe, 
velopment of Mao, Bell, Xu, and Wong however, and there exists today only a 
in adapting a diamond-window, high- very small set of usable values. The po- 
pressure cell to a Fourier-transform in- sition-sensitive detector acquires most of 
frared interferometer, eliminates ab- the lines of an x-ray diffraction pattern 
sorption, low-intensity, and stray light simultaneously, and thus the high losses 
factors that impeded measurements in in precision of high-pressure measure- 
the past. The new system has the low- ments that have resulted from mechan- 
noise and high-resolution characteristics ical drift during the days or weeks 
of the interferometric technique. A beam- required for measurement by film meth- 
condensing method results in a high-in- ods have been eliminated. The new sys- 
tensity, infrared signal. With the new tern is accurate and rapid, and, therefore, 
technique it is now possible to study min- the reproducibility can be tested. The 
eral and solidified gas samples under a system has sufficient precision for appli- 
range of temperatures and pressures. The cation to synchrotron x-ray sources of 
system has yielded high-resolution data the future. For the present, the system 
in the entire infrared region at a range is operational on the Geophysical Labo- 
of pressures that can be used to calculate ratory's rotating- anode, x-ray genera- 
directly lattice-dynamic values for min- tor. 
erals. * During melting experiments at 1 atm 

A new computerized system has been under controlled f , silicate charges in 

developed by Mao, Hadidiacos, Bell, and Fe-bearing systems lose Fe to the en- 

Goettel for automated spectral measure- veloping Pt-wire loop. The Fe loss is a 

ments in experiments with the diamond- rate process controlled by temperature, 

window, high-pressure cell. The system, /b 2 , and bulk composition of the charge, 

which involves an automated stage and and produces a Pt-Fe alloy whose equi- 

spectrometer, has been applied to pres- librium composition may be calculated 

sure measurement by the ruby-fluores- from the activity-composition relations 

cence method. The automated stage and in the melt at known T and f . Sharpe 

the computer-storage facility are readily has described a routine process whereby 

adapted to the laser-heating system. The Fe can be electroplated onto preformed 

advantages of this system over manual Pt-wire loops from a solution of ferrous 

adjustment are that the pressure distri- ammonium sulfate under controlled T, pH, 

bution in the cell can be determined more and cell parameters. The amount of Fe 

precisely and rapidly and plotted for deposited on the loop is simply a function 

analysis. Such a system is extremely val- of time. When the loop is annealed at 

uable in experiments at pressures above 1350°C for six hours in a stream of form- 

0.5 Mbar, where a continuous assess- ing gas (95% N 2 , 5% H 2 ), homogeneous 

ment of stress distribution and alignment alloys are produced. Sharpe has melted 

of the high-pressure cell is required. The two natural basic rocks at supraliquidus 

system is part of a controlled laser-heat- temperatures and has shown that no Fe 

ing device under development. loss occurs from the charge to such a loop 

Mao, Bell, and Weng have worked to at times of up to 96 hours. Electron mi- 
improve compressibility measurements croprobe analyses of the loops after ex- 
on minerals at high pressure by adapting periments showed that their compositions 


remained essentially unchanged. The deducing climatic and environmental 
compositions of Pt-Fe alloys produced by conditions. A preparative method for iso- 
experiments with undoped Pt loops were lation of milligram quantities of D-glu- 
comparable with the predicted compo- cosamine for stable isotope measurements 
sitions. Over extended periods, how- has been developed by Schimmelmann 
ever, Na 2 and K 2 loss from charges and Hare in which a cation-exchange col- 
was seen to be significant. umn with 1 M HC1 for the eluent is used. 

The density and viscosity of liquids may Pure D-glucosamine is eluted from the 

be determined by Stokes' law from the column well ahead of the amino acids, 

settling rates of spheres of known radius Concentration by simple evaporation of 

and density. This technique has been ap- the solvent yields the hydrochloride of D- 

plied for several years to the study of glucosamine. 

silicate melts, but the location of tiny *Macko and Engel have combined two 
spheres within a capsule of glass has previously separate techniques of amino 
proved a time-consuming task. Further- acid analysis to develop a potentially 
more, spheres smaller than 0.2 mm in powerful analytical tool. Large-scale 
diameter could not be found by standard separation by ion exchange of a partic- 
cutting and polishing procedures. Hazen ular amino acid (Macko et al., Year Book 
and Sharpe have developed a radio- 81, 417-422) was used to purify isolates 
graphic technique that may be used to of D- and L-amino acids from a chiral mo- 
locate spheres as small as 0.02 mm in bile phase (Engel and Hare, Year Book 
diameter quickly and accurately. This new 80, 394-397). These purified separates 
procedure will result in the study of a were then analyzed for their isotopic 
wider range of silicate melt composi- composition. Isotopic integrity ( 13 C and 
tions. 15 N) was maintained throughout the sep- 

*Chitin, an important biopolymer in aration and analysis. This unification of 

many invertebrate exoskeletons, both two chromatographic methods with iso- 

modern and fossil, can be hydrolyzed to topic analysis of the isolates holds prom- 

D-glucosamine, an amino sugar, and amino ise for future research on kinetic isotope 

acids. The isotopic ratios of carbon, ni- effects, racemization processes, and res- 

trogen, oxygen, and hydrogen in glucos- olution of contaminants from pure com- 

amine from modern and fossil chitin are ponents. 

expected to yield information useful in Ratten S. Yoder, Jr. 


Heat Transfer Related to of metamorphic zones and the rate of 
Metamorphism and Magma magma production, an apparatus has been 
Generation: Results of constructed for obtaining heat-transfer 
Exploratory Experiments properties in advance of melting and dur- 
H „ Y , j ing the partial-melting process. In prep- 
aration for the hot-plate-type melting- 
One of the important modes of heat experiment (Yoder, Year Book 79, 263- 
transfer resulting in metamorphism or 267), the characteristics of the apparatus 
magma generation is from an external itself have been tested, and the prelim- 
planar heat source (Yoder, 1978). To ob- inary observations appear to have some 
tain quantitative information on the rates applications to thermal metamorphic 
of heat transfer that determine the width processes. 



The apparatus for the study of heat- 
transfer properties (Fig. 1) holds a series 
of sample disks (0.55 cm thick and 8 cm 
in diameter) in which fine thermocouples 
(wire diameter = 0.13 mm) are imbed- 
ded in the surfaces. An array of 54 ther- 
mocouples is laid out in a diametric plane 
through a stack of nine disks. The 108 
thermocouple leads are brought up the 
sides of the stack of disks in grooves, cut 
with a fine diamond saw, in the sides of 
the disks (Fig. 2). The fine grooves on 
the surfaces and edges of the disks are 
filled with a minimal amount of high-tem- 
perature cement to insulate the ther- 
mocouple wires and prevent movement 
during assembly. The stack of disks is 
held in a close-fitting platinum crucible, 
which sits on a heat-flux gauge of the 
same diameter as the crucible in a block 
of insulating bricks. The heat-flux gauge 

consists of approximately 750 differential 
platinum-platinum 10 rhodium 90 ther- 
mocouples straddling opposite sides of a 
contiguous series of 0.3-cm 2 bars of boron 
nitride. Immediately under the heat-flux 
gauge is a primary Pt-resistance heater 
of circular plan also of exactly the di- 
ameter of the crucible. Guard heaters are 
imbedded in the Alundum plate holding 
the primary heater, but they were not 
required in the current set of preliminary 

The "sample" used in three prelimi- 
nary experiments was of the same ma- 
terial as the aluminosilicate insulating 
bricks. Three different heat fluxes were 
applied to the bottom of the crucible, and 
in the absence of adequate calibration of 
the heat-flux gauge, it may be assumed 
that, as a first approximation, half of the 
power released entered the base of the 

Fig. 1. Schematic section through heat-transfer apparatus in kiln. 



Fig. 2. Photograph of platinum crucible and 
thermocouple harness (108 leads). 

crucible. On this basis, the presumed heat 
flux at the surface of the primary heater 
for the three experiments was 0.48, 1.07, 
and 1.43 cal cm" 2 sec" 1 , respectively, 
over the 50.3-cm 2 surface. The calculated 
effective heat flux into the basal portion 
of the sample, whose temperature and 
thermal properties are changing with 
time, however, may be as much as an 
order of magnitude less if the mathe- 
matical description of the process, given 
below, can be used for comparison. 

The temperatures recorded for one 
vertical array of thermocouples as a func- 
tion of time for each flux are given in 
Figs. 3-5 as examples. The initial tem- 
perature of each experiment was about 
750°C, and the distribution throughout 
the sample was well within ± 2°C of that 
value. Because a relatively uniform tem- 

perature distribution was recorded 
throughout the entire diametric plane 
measured within the sample, it appears 
that the greater rate of heat transfer up 
the platinum walls of the crucible essen- 
tially equaled the heat losses from the 
sample through the platinum wall. 

Each thermocouple was measured and 
recorded approximately every 48 sec in 
2-sec intervals, usually in sequential or- 
der through the vertical arrays. The total 
elapsed time for measuring and record- 
ing each array from the highest to the 
lowest temperature was 20 sec. All the 
thermocouples were cut from the same 
spools of wire. Although the thermocou- 
ples were not calibrated, the relative dif- 
ferences are considered negligible. Errors 
in the three-dimensional positioning of 
the thermocouples within the sample are 
also considered negligible. The principal 
errors are more likely to arise owing to 
local inhomogeneities in the thermal en- 
vironment of the imbedded thermocou- 
ple junction at the interface of the disks. 
To minimize heat losses that result from 
the imbedded thermocouple leads, wire 
only 0.13 mm in diameter was used, and 
the thermocouple leads were laid along 
the disk surfaces, that is, along approx- 
imately isothermal planes. Heat losses 
were observed by measuring the tem- 
perature rise in twelve thermocouples 
imbedded in the surrounding insulating 

The theoretical temperature profiles 
were calculated through the courtesy of 
Larry W. Finger for the case in which 
the heat flux at the base of the sample is 
assumed to be constant, the initial tem- 
perature is fixed, and the upper bound- 
ary is kept at the initial temperature. In 
the temperature range studied there were 
no heat sinks or sources within the ma- 
terial. According to Carslaw and Jaeger 
(1959, p. 113) the appropriate expression 

\T = 

F x 8F l 




/CTT 2 w =0 (2W + l) 2 

K(2n + l) 2 >n 2 t 


(2 n + 1) 7T x 


1 1 I 

Heat Flux: 0.48 cal cm -2 sec 1 


850 900 

Temperature, °C 


Fig. 3. Temperatures recorded at various times at a vertical series of thermocouple junctions in the 
sample. Heat flux = 0.48 cal cm" 2 sec" 1 . Duration of heat flux, 2 hr + 45 min, after approximately 3 
hr at initial temperature, 754°C. Curves were calculated on the basis of the theory described in the text. 

where ATMs the change in temperature, 
°C; F , heat flux, cal cm -2 sec -1 ; k, ther- 
mal conductivity, cal cm -1 sec -1 deg" 1 ; 
K, thermal diffusivity, cm 2 sec" 1 ; t, time, 
sec; I, sample thickness, cm; and x, dis- 
tance from top of sample. A plot of A7 1 
vs. t l/2 revealed that the effective t at 
AT = was approximately 2.9, 1.0, and 
3.8 min, respectively. Furthermore, the 
effective x , where the kiln temperature 
is maintained, was found to be, not at the 
top of the sample, but near the top of the 
thermocouple harness collar, about 1.55 
cm above the sample. The calculations 
were made on the assumption that T, t, 
and x were known and the terms F /k, 
x Q , and K were unknown. A least-squares 
fit was made first with all the data of a 
particular experiment; however, because 
of the initial changes in heat flux and the 
decrease in thermal conductivity with 
temperature, it was convenient to illus- 

trate single profiles at about 4-min in- 
tervals. Small changes in the effective x 
were not significant, whereas the start- 
up change in properties of the primary 
heater with temperature did result in 
significant changes in the initial heat flux. 
An inspection of Figs. 3-5 reveals that 
the model curves do represent, as a first 
approximation, the form of the data, and 
the fit is quite adequate. The assump- 
tions of constant heat flux and constant 
thermal diffusivity are presumably in- 
appropriate, and other models will have 
to be investigated. 

Although these observations are of an 
exploratory nature, the results signal 
some critical effects. The most pertinent 
to both metamorphism and magma gen- 
eration is the high rate at which a ther- 
mal gradient is established relative to the 
rate of reaction or melting. It seems that 
the heating process may be essentially 




850 900 

Temperature, °C 



Fig. 4. Temperatures recorded at various times at a vertical series of thermocouple junctions in the 
sample. Heat flux = 1.07 cal cm" 2 sec" 1 . Duration of heat flux, 28 min, after approximately 5.75 hr at 
initial temperature, 744°C. Curves were calculated on the basis of the theory described in the text. 

750 800 850 900 950 1000 

Temperature, °C 




Fig. 5. Temperatures recorded at various times at a vertical series of thermocouple junctions in the 
sample. Heat flux = 1.43 cal cm" 2 sec" 1 . Duration of heat flux, 29 min, after approximately 5 hr at 
initial temperature, 748°C. Curves were calculated on the basis of the theory described in the text. 



decoupled from the reaction or melting 
process. In the writer's experience, 
metamorphic reactions in the laboratory 
involving very fine grained materials re- 
quire runs at least 24 hr to 1 week in 
duration to achieve equilibrium in this 
temperature range, and runs even in the 
presence of a liquid require several hours 
to days. Because of these kinetic factors 
some attention will have to be given, with 
due regard for scale, to the decoupling 
of the thermal process from the reaction 
or melting processes.* 

The influence of the change in thermal 
conductivity with temperature is also ev- 
ident from the change in initial slope of 
the curves in Figs. 3-5. The thermal con- 
ductivity appears to be decreasing with 
temperature in accord with the obser- 
vations on most rock-forming minerals. 
This effect will be important when re- 
sults of systems involving an enthalpy of 
reaction or melting are compared be- 
cause such enthalpy may be consumed 
over a range of temperatures. In such 
cases, the enthalpy of reaction or melting 
may appear as an anomalous change in 
the specific heat (C P ), which, along with 
density (p) changes, influences the ther- 
mal diffusivity, K = k/pC P . 

After the apparatus is tested further 
at other heat fluxes, it will then be ap- 
propriate to investigate melting in the 
system sodium disilicate-nepheline 
(Yoder, Year Book 79, 263-267). A syn- 
thetic crystalline sample of NDS 14.2 Ne 
85.8 (wt %), consisting of ten disks 0.5 
cm thick, has been prepared for this pur- 
pose. The system has the appropriate 
melting range (i.e., granite through ba- 
salt liquidus temperatures), melting rates, 
and liquid-density distribution along the 
nepheline liquidus surface to insure con- 
vective overturn in a porous medium (i.e., 
crystal network of nepheline); and no solid 
solutions or polymorphic transitions in 

the temperature region to be investi- 
gated (750°-1150°C) are involved. To avoid 
partial melting of the sample beyond the 
limit of crystal bridging, f it will be nec- 
essary to switch from a constant heat flux 
to a constant temperature condition at 
the limiting degree of melting. It is be- 
lieved at this stage of the study that these 
two limiting conditions will provide in- 
sights into natural melting processes. 


Carslaw, H.S., and J.C. Jaeger, Conduction of Heat 
in Solids, 2d ed., Oxford University Press, New 
York, 1959. 

Yoder, H.S., Jr., The MgO-Al 2 ;r Si0 2 -H 2 system 
and the related metamorphic facies, Amer. J. 
Sci., Bowen Vol., 569-627, 1952. 

Yoder, H.S., Jr., Generation of Basaltic Magma, 
National Academy of Sciences, Washington, D.C., 

Yoder, H.S., Jr., Basic magma generation and ag- 
gregation, Bull. Volcanol, U, 301-316, 1978. 

Partitioning of Iron and 

Magnesium between Melilite, 

Olivine, and Clinopyroxene in 


D. Velde and H. S. Yoder, Jr. 

Petrographic description of lavas has 
been, for the most part, based on visual 
evaluation of the grain sizes and inter- 
relationships of the crystal species pres- 
ent in thin section. The largest crystals 
were presumed in the past to nucleate 
before smaller crystals. Those that are 
inordinately large were commonly inter- 
preted as representing intratelluric con- 
ditions, whereas small-size groundmass 
phases were thought to nucleate and grow 
under surface conditions. Experimental 
studies (e.g., Lofgren et al., 1974), how- 
ever, have revealed that such porphy- 
ritic textures may also result from the 

* Because of the decoupling, metamorphism may 
not be progressive and a metamorphosed bed will 
record only the highest temperatures to which it 
has been exposed, a concept already alluded to by 
Yoder (1952). 

tThe observed limit of crystal bridging is about 
45 wt % liquid (Yoder, 1976, pp. 112-113). Loss of 
crystal bridging would result in collapse of the ther- 
mocouple array, and the position of the thermo- 
couples would be indeterminate with time except 
at the termination of the experiment. 



cooling of liquids, for example, at con- 
stant linear rates. Such kinetic factors 
apparently also influence the partitioning 
of major, minor, and trace elements (see 
Lofgren, 1980, for review) in lavas that 
rapidly cool on the surface. 

The melilite-bearing lavas are partic- 
ularly useful in testing the influence of 
kinetic factors on crystallization because 
of several features: plagioclase, with its 
special nucleation-rate problems, is ab- 
sent; crystallization of melilite is re- 
stricted to the crust; and all the principal 
phases partition iron and magnesium. The 
main objectives of the present study on 
melilite-bearing lavas were to 

(1) determine the apparent partitioning 
of Fe and Mg between melilite and co- 
existing olivine and clinopyroxene, 

(2) ascertain the influence of rapid sur- 
face cooling on the range of partitioning, 

(3) estimate the order of initiation and 
especially the cessation (closure temper- 
ature) of the crystallization of each phase, 
and (4) identify, if possible, those fea- 
tures inherited from depth. 

Experimental Procedure 

Criteria for the selection of samples 
used in this investigation were that the 
samples (1) were free from alteration, 
(2) contained simple mineralogical asso- 
ciations with either prominent olivine + 
melilite or clinopyroxene + melilite pairs, 

and (3) had a distinct range of Fe/(Fe + 
Mg) values. The localities and mineral- 
ogy of the samples studied are listed in 
Table 1. Specimens Y3 and Y23 have been 
melted and crystallized in the laboratorv 
(Tilley et al, Year Book 6J>, 69-82). In 
all samples, analyses were made along 
traverses across several crystals every 
10 to 30 |xm, in order to obtain a reliable 
evaluation of zoning. The NiO content 
was routinely determined in all olivine 
analyses, for use as a discriminant be- 
tween xenocrysts (high NiO content) and 

Fe-Mg Partitioning 

Specimen Y156: The Fe/(Fe + Mg)* 
varies from 0.179 to 0.217 in the olivine. 
The corresponding NiO contents do not 
vary in a regular fashion but always re- 
main less than 0.25 wt %. In the melilite, 
Fe/(Fe + Mg) varies from 0.187 to 0.223. 
These results were confirmed by less de- 
tailed traverses across other crystals, and 
all analyses are summarized in Fig. 6A. 

Specimen Y23: The Fe/(Fe + Mg) is 
apparently constant within an olivine 
crystal with the exception of a small in- 
crease in iron at the outermost rim. The 
pattern of NiO variation is irregular, but 
values remain less than 0.20 wt % NiO. 

*Total Fe is calculated as Fe 2+ in all silicates 
analyzed by electron microprobe in this report. 

TABLE 1. Melilite-Bearing Rocks for which Fe/(Fe + Mg) in Coexisting Phases Were Determined 



Rock Type 


Mineral Content' 

Y156 Melilite-olivine nephelinite Biesiesfontein farm, Ol + Mel + Cpx + Ma + Ne 

South Africa 
Y23 Melilite-olivine nephelinite Moiliili, Honolulu, 01 + Mel + Cpx + Ne + Sodal + Ma 

Y3 Olivine-melilite nephelinite Uvalde County, Ol + Mel + Cpx + Ma + Ne 

Y342 Olivine melilitite Spiegel River, 01 + Mel + Per + Ma + Ne 

South Africa 
Y253 Coppaelite Cupaello, Rieti, Cpx + Mel + 01 + Ks + Phlog + Per + Melanite 


*01, olivine; Mel, melilite; Cpx, clinopyroxene; Ma, magnetite; Ne, nepheline; Ks, kalsilite; Per, per- 
ovskite; Phlog, phlogopite; Sodal, sodalite. 



Melilite crystals also exhibit variations 
of Fe/(Fe + Mg). Plateau values vary 
from 0.194 to 0.199 (Fig. 6B), whereas 
at the outer zones of the crystals the ratio 

5 - 

<c _ 

z o 

I Y156 



: , /. 








h n 


Fe/(Fe + Mg) 


Fig. 6A. Univariate distribution of Fe/(Fe + 
Mg) in olivines and melilites from melilite-olivine 
nephelinite Y156. 

5 - 

I I I I I I I I I 





h i i 

Z 5 








5 - 





Fe/(Fe + Mg) 


reaches 0.292. Values of Fe/(Fe + Mg) 
in the clinopyroxenes vary between 0.241 
and 0.263, the high values being found 
at the rims of the crystals. Clinopyrox- 
ene is the third phase to appear in spec- 
imen Y23, as deduced from textural 
relationships. This deduction is in ap- 
parent contradiction with the experi- 
mental results, according to which olivine 
appears at 1260°C; clinopyroxene, at 
1160°C; and melilite, at 1095°C (Tilley et 
al., Year Book 6J>, 69-82). 

Specimen Y3\ The Fe/(Fe + Mg) val- 
ues in olivine vary from 0.121 to 0.132 
on the plateau with an abrupt 50-|xm- 
wide increase toward the rim, reaching 
a maximum 0.210 value at the outer rim 
of the crystal (Fig. 6C). Melilite yields 

5 - 


D D 






Fe/(Fe + Mg) 



Fig. 6B. Univariate distribution of Fe/(Fe + 
Mg) in olivines, melilites, and clinopyroxenes from 
melilite-olivine nephelinite Y23. 

Fig. 6C. Univariate distribution of Fe/(Fe + 
Mg) in olivines, melilites, and clinopyroxenes from 
olivine-melilite nephelinite Y3. 



values of about 0.170 in the center, 
reaching 0.241 at the outer zones of the 
crystals. Values for Fe/(Fe + Mg) in the 
clinopyroxene cluster between 0.200 and 
0.210. Laboratory results indicate pre- 
cipitation of olivine at 1287°C, then cli- 
nopyroxene at 1190°, and melilite at 
1130°C (Tilley et al., Year Book 64, 69- 

Specimen Y3J>2: This particular sam- 
ple was selected as being typical of cer- 
tain olivine melilitites in which extremely 
abundant, euhedral olivine crystals, large 
and small, are set in a groundmass com- 
prised mostly of abundant but small crys- 
tals of melilite. In the olivines the Fe/ 
(Fe + Mg) varies from 0.150 to 0.177, 
the highest ratio being found at the rim 
of a small crystal (Fig. 6D). Melilites, 
found exclusively in the groundmass, yield 
ratios varying from 0.190 to 0.206. The 
maximum, found in the zone near the rim, 
is 0.253. In this particular lava, the Fe/ 
(Fe + Mg) values found in the olivines 
and melilites do not overlap. 

Specimen Y253: In this unusual lava, 
the type coppaelite, clinopyroxene forms 

phenocrysts, melilite appears as micro- 
phenocrysts, and some very rare, minute 
olivine crystals (previously unrecorded) 
are found in the groundmass. In clino- 
pyroxene, values of Fe/(Fe + Mg) clus- 
ter around 0.060 (Fig. 6E). The analyses 
for melilites show Fe/(Fe + Mg) be- 
tween 0.100 and 0.120. Values found for 
olivine are about 0.100. 

Size of Crystals 

In a pioneer study, Winkler (1949) at- 
tempted to relate the crystal sizes of three 
mineral species (plagioclase, magnetite, 
and clinopyroxene) in a tholeiitic basalt 
dike with the cooling rate of the intru- 






5 - 

I 1 I I 1 I I I 



5 - 

r, n 



Fe/(Fe + Mg) 







Fig. 6D. Univariate distribution of Fe/(Fe + Fig. 6E. Univariate distribution of Fe/(Fe + 

Mg) in olivines and melilites from olivine melilitite Mg) in clinopvroxenes, melilites, and olivines from 
Y342. coppaelite Y253. 



sion. His study showed that sizes vary 
in a regular fashion with respect to the 
distance from the edge of the dike and 
also that coprecipitating phases (at least 
those in a eutectic-like relationship) have 
sizes of the same order of magnitude. 
These results imply that, within reason- 
able limits, the size of coprecipitating 
silicate minerals does not differ widely. 
A corollary is that silicate crystals with 
sizes differing by an order of magnitude 
probably result from crystallization un- 
der different rates of cooling. More re- 
cent experimental studies have shown that 
whereas the shape of liquidus phases 
varies greatly with cooling rates (Don- 
aldson, 1976), the sizes of coprecipitating 
minerals are constrained by the high 
density of nucleation, which results in 
slower growth of phases consuming the 
same materials, in the range of cooling 
rates within a lava flow. 

Figures 7A-E show the distribution of 
maximum dimension of prismatic crys- 
tals of olivine, melilite, and clinopyrox- 
ene in the five samples described above. 
For all samples but Y342, the coprecip- 
itation of melilite and olivine seems likely. 
In specimen Y342, however, there is lit- 
tle overlap in the measured largest di- 
mension of the olivine and melilite 
crystals, and this observation is taken to 
mean that at least a large proportion of 








h , 



Z n 

, 1 , 

1 1 

1 1 1 



i i i I i 

J L 

Length of crystals, fim 



; o 









I I I I 


Length of crystals, /jm 


Fig. 7A. Univariate distribution of the maxi- 
mum dimension of olivine and melilite crystals in 
melilite-olivine nephelinite Y156. 

Fig. 7B. Univariate distribution of the maxi- 
mum dimension of olivine, melilite, and clinopy- 
roxene crystals in olivine-melilite nephelinite Y23. 

the olivine had crystallized before the ap- 
pearance of melilite. In specimen Y253, 
most clinopyroxenes are larger than most 
melilites, which in turn are larger than 
the very minute olivine crystals. 

Composition of the Spinels 

Spinels crystallize, albeit with chang- 
ing compositions, over a large tempera- 
ture interval in melilite-bearing lavas. 
Spinels are found as inclusions in oli- 
vines, as in most minerals that occur in 
the rocks, and, of course, they abound in 
the groundmass. For these reasons spi- 
nel should be helpful in providing a guide 
to cooling rates; however, difficulties arise 
in evaluating and interpreting coexisting 
phases as having crystallized together 
owing to the absence of unequivocal cri- 

Figures 8A-B show the joint distri- 
bution of Ti0 2 and Cr 2 3 in the spinels 





§ 5 





~l I 

n r , 

■ 1 1 1 1 1 j 1 1 

I I I 

•> 1 


o I ■ i i i i 1 1 i t i I n ii 





Length of crystals, ^im 

Fig. 7C. Univariate distribution of the maximum dimension of olivine, melilite, and clinopyroxene 
crystals in olivine-melilite nepheline Y3. 

found in specimens Y156 and Y342. These 
elements have been chosen because they 
show the largest variance in the minerals 
under consideration. Furthermore, they 
are apparently the most sensitive to ki- 
netic factors (Grove and Bence, 1977). At 
fast cooling rates the ratio of Cr to Ti is 
large, diminishing at an increasing rate 
with slower cooling rates and depletion 
of Cr in the liquid. The Cr 2 3 is highly 
concentrated in the early-formed spinels, 
whereas Ti0 2 (as an ulvospinel molecule) 



iru 1 

H^rin ,^i n 

c5 15 


I 10 


Qll I I I 



I I I L 



Z 10 


i I i i I I I I L 

1000 2000 

Length of crystals, /jm 



J I I I I I I 1 L 

1000 2000 

Length of crystals, nm 

Fig. 7E. Univariate distribution of the maxi- 
mum dimension of olivine and melilite crystals in mum dimension of clinopyroxene, melilite, and oli- 

Fig. 7D. Univariate distribution of the maxi- 

olivine melilitite Y342. 

vine crystals in coppaelite Y253. 




C? 10 








i i i 





• • 


• • 





i i 

I I I 



Cr 2 3 , wt% 

Fig. 8A. The joint distribution of Ti0 2 and Cr 2 3 in spinels from specimen Y342. 


is progressively incorporated as crystal- 
lization proceeds. In specimen Y342, in 
which olivine may have crystallized be- 
fore nucleation of melilite, the range of 
spinel compositions extends from 32 to 
4.5% Cr 2 3 . A number of spinels are found 
that contain 0-2% Cr 2 3 and 15-20% Ti0 2 . 
These spinels probably correspond to late- 

Cr 2 o 3 , wt % 

Fig. 8B. The joint distribution of Ti0 2 and Cr 2 3 
in spinels from specimen Y156. 

crystallizing magnetites. The other spec- 
imen, Y156, illustrates a very different 
spread of compositions, similar to that 
found in samples Y3 and Y23. In all sam- 
ples a systematic search was made for 
Cr-rich spinels. Some were indeed found 
in specimens Y3 and Y23 and are pres- 
ent, though rare, in specimen Y156. When 
present, they are exceptional, and no in- 
termediate compositions were found be- 
tween the Cr-rich, Ti-poor spinels and 
the Cr-poor, Ti-rich magnetites. 


Specimens Y156, Y23, and Y3 have the 
same mineralogical assemblage (Table 1), 
whereas clinopyroxene is absent from 
specimen Y342. Coppaelite (Y253) con- 
tains melanite, perovskite, and phlogo- 
pite in addition to olivine, clinopyroxene, 
and melilite. Because element partition- 
ing is influenced by assemblage, the Fe/ 
(Fe + Mg) values for coexisting phases 
are examined in three different sets. 

In specimen Y156, the Fe/(Fe + Mg) 
values of melilite and olivine show con- 
siderable similarity. The crystals are 
comparable in size, but a few are over- 
sized (xenocrysts?). In specimen Y23 the 
lowest values for Fe/(Fe + Mg) increase 
from olivine and melilite to clinopyrox- 
ene, but the mode of Fe/(Fe + Mg) is 
lower for olivines than it is for melilites, 
in accord with the crystallization se- 



quence deduced from textural observa- 
tions. There is considerable overlap in 
the Fe/(Fe + Mg) values for the three 
phases. In specimen Y3 the lowest Fe/ 
(Fe + Mg) is in olivine, then clinopyrox- 
ene, then melilite, and the succession or- 
der conforms with the experimentally 
determined sequence. Sizes of the crys- 
tals are comparable for the three phases. 

In specimen Y342 there is no overlap 
in Fe/(Fe + Mg); olivines have lower Fe/ 
(Fe + Mg) than melilites. The olivine 
crystals show a wide spread of sizes, 
whereas melilite sizes are restricted to a 
narrow domain. Spinels have variable 
compositions in a large field of Ti and Cr 
values, an extensive array, albeit with 
gaps, that is not found in any of the other 

In specimen Y253, the lowest Fe/(Fe 
+ Mg) is found in the clinopyroxenes, 
with no overlap with the melilite and ol- 
ivine Fe/(Fe + Mg), which are within 
the same range. 

It is concluded from these data that 
coprecipitating melilite and olivine, or 
melilite and clinopyroxene, have essen- 
tially the same Fe/(Fe + Mg). The sim- 
ilarity may result from the fact that the 
partitioning within each phase is ap- 
proaching the same value or that each of 
the phases grows at a sufficiently fast 
rate that all the crystals incorporate 
whatever ratio is available in the liquid 
(stable or metastable). That ratio may or 
may not be related to the total MgO and 
FeO in the liquid, the structure of which 
may influence the availability of each ele- 

The reason that one phase (olivine in 
specimen Y342, clinopyroxene in speci- 
men Y253) appears to stop crystallizing, 
on the basis of its Fe/(Fe + Mg), before 
complete crystallization of the other 
phases is not yet fully appreciated. One 
likely possibility is that the presumed 
missing compositions do indeed exist but 
are not amenable to analysis, forming thin 
rims around crystals or very small 
groundmass crystals. Another possibil- 
ity, especially for Y342, is that one or 
more of the phases accumulated from liq- 

uids at different stages of cooling. A third 
possibility is that the fastest-growing 
phase (presumably olivine) develops to 
the fullest extent possible in a short pe- 
riod of time so that when the next phase 
begins to nucleate, it is from a liquid (sta- 
ble or metastable) of a different compo- 
sition, devoid of that portion of the 
components that make up the preceding 
phase(s). Such circumstances may exist 
if nucleation and growth are confined, for 
example, to the temperature region be- 
tween the solidus and the closure tem- 
perature of the glass at the glass 
transition, T g . In short, the lack of over- 
lap of Fe/(Fe + Mg) may have resulted 
from a wide disparity of such kinetic fac- 
tors as nucleation and growth rates. A 
fourth possibility is that the olivine has 
undergone a reaction relation with liquid 
and has ceased crystallizing. Although 
this event is considered unlikely for lack 
of petrographic confirmation in the spec- 
imens studied, such a reaction relation 
has been predicted on the basis of iron- 
free synthetic systems for other melilite- 
bearing magmas (Schairer and Yoder, 
Year Book 63, 65-74). 

It is evident that a scheme for corre- 
lating the contemporaneous growth of 
zones in each of the phases and an ap- 
preciation of kinetic factors are needed 
before the apparent partitioning coeffi- 
cients and tie lines for minerals in lavas 
can be interpreted. 


Donaldson, C.H., An experimental investigation of 
olivine morphology, Contrib. Mineral. Petrol., 
57, 187-213, 1976. 

Grove, T.L., and A.E. Bence, Experimental study 
of pyroxene-liquid interaction in quartz-norma- 
tive basalt 15597, Proc. Lunar Sci. Conf., 8th, 
1549-1579, 1977. 

Lofgren, G., Experimental studies on the dynamic 
crystallization of silicate melts, in Physics of 
Magmatic Processes, R. B. Hargraves, ed., 
Princeton University Press, pp. 487-551, 1980. 

Lofgren, G., C.H. Donaldson, R.J. Williams, D. 
Mullins, Jr., and T.M. Usselman, Experimen- 
tally reproduced textures and mineral chemistry 
of Apollo 15 quartz normative basalts, Proc. Lu- 
nar Sci. Conf, 5th, 549-567, 1974. 



Winkler, H.G.F., Crystallization of basaltic magma 
as recorded by variation of crystal-size in dikes, 
Mineral. Mag., 28, 557-574, 1949. 

Petrology of Olivine Melilitites 

From Saltpetre Kop and 

Sutherland Commage, Cape 

Province, South Africa 

N. Z. Boctor and H. S. Yoder, Jr. 

The close association in space and time 
of olivine melilitites, carbonatites, and 
kimberlites led some investigators (e.g., 
Tuttle and Gittins, 1966; Mclver and 
Ferguson, 1979) to believe that the par- 
ent magmas of these rocks are geneti- 
cally related. Experimental investigation 
of the stability of melilite (Yoder, 1975), 
however, places severe constraints on the 
pressure regimes at which olivine mel- 
ilitites crystallize. All three rock types 
are found at Saltpetre Kop volcano, Cape 
province, South Africa, which consists of 
intrusive and extrusive rocks, such as 
trachytes, carbonatites, olivine melili- 
tites, and pyroclastic kimberlitic erup- 
tives (Mclver and Ferguson, 1979). 
Because of the close association of the 
rock types, samples from two satellite 
intrusives associated with the main Salt- 
petre Kop volcano have been investi- 
gated. One occurrence is a small olivine 
melilitite plug (100 m in diameter) about 
5 km north of Saltpetre Kop, and the 

other is at Sutherland Commage about 
16 km north of Saltpetre Kop volcano 
(Duncan et aL, 1978). The olivine melil- 
itites in both occurrences are character- 
ized by unusual reaction mantles on their 
olivine phenocrysts, Ba- and Ti-bearing 
trioctahedral mica, Ti-rich garnets, and 
groundmass spinels that mimic in their 
crystallization trends the groundmass 
spinels in kimberlites. The results of this 
investigation do not provide evidence for 
a genetic relation between kimberlites 
and olivine melilitites; however, they do 
indicate that a carbonate-rich melt that 
concentrates the rare-earth and other in- 
compatible elements evolves in the final 
stages of crystallization of the magmas 
that produce both kimberlites and olivine 

Petrography and Mineral Chemistry 

At Saltpetre Kop the major phases in 
the olivine melilitites are olivine, melil- 
ite, phlogopite, and members of the spi- 
nel solid solution series, whereas the 
minor phases include Ti-rich garnet, 
nepheline, sodalite, perovskite, apatite, 
and calcite. Olivine (Fig. 9A) occurs as 
large phenocrysts (Fo 85 _ 89 ) that com- 
monly display mantles of monticellite 
(Mg 37 Ca4 8 Fe 1 5-Mg4oCa 47 Fe 1 3); in rare 
cases the latter also occur as veinlets 
transecting the olivine phenocrysts. 


Ca 2 FeSi 2 7 


Ca 2 MgSi 2 7 Ca 2 FeSi 2 7 

Ca 2 MgSi 2 7 

Fig. 9. Ca-Mg-Fe plots of average compositions of olivine and monticellite reaction mantles. (A) Saltpetre 
Kop; (B) Sutherland Commage. 





Fig. 10. Average compositions of melilite plotted in terms of the mol percentages of akermanite, 
ferroakermanite, and soda melilite. (A) Saltpetre Kop; (B) Sutherland Commage. 

Mantles of zoned spinels occur on some 
olivine crystals with inner zones of 
chromite and outer zones of titanomag- 
netite. Trioctahedral Ba- and Ti-bearing 
mica forms a less common type of mantle 
on olivine. Melilite occurs both as phen- 
ocrysts and as a groundmass mineral. The 
phenocrysts show idiomorphic habit even 
where they occur in the immediate vi- 
cinity of the monticellite reaction man- 
tles. The melilite (Fig. 10A) is close in 
composition to those reported by El Go- 
resy and Yoder (Year Book 73, 359-363) 
and Velde and Yoder (Year Book 76, 478- 
483) from the Nyiragongo. Its composi- 
tion is about two-thirds akermanite and 
one-third soda melilite. Phlogopite (BaO 
7.4-16.95 wt %, Ti0 2 1.4-4.3 wt %) oc- 
curs as zoned crystals. The Ba, Ti, and 
Al contents decrease, whereas K, Si, and 
Mg increase, from core to rim. Garnet 
occurs as yellowish brown to opaque 
crystals that range in composition from 
melanite (0.6-3.5 wt % Ti0 2 ) to schor- 
lomite (10.8-15.3 wt % Ti0 2 ). Some gar- 
net crystals display zoning with decrease 
in Ti0 2 and increase in Fe 2 3 and Si0 2 
from core to rim (Fig. 11). In some places 
the zoning pattern is more complex with 
alternating layers of melanite and schor- 
lomite. Spinels occur both as rare ma- 
crocrysts and as an abundant groundmass 
mineral. The spinel macrocrysts are Mg- 

Al chromite (Ti0 2 1.2-2.7 wt %) that is 
mantled by titanomagnetite (4.4-11.3 wt 
% Ti0 2 ). Groundmass spinels occur as 
subhedral to euhedral crystals of titano- 
magnetite (2.7-12.0 wt % Ti0 2 ) with rare 
cores of Mg-Al chromite. Discrete crys- 
tals of perovskite occur in the ground- 
mass as rims on titanomagnetite. 
Nepheline, sodalite, and apatite occur as 
rare minerals in the groundmass, whereas 
calcite and fibrous zeolites are abundant. 
The olivine melilitites from Sutherland 
Commage display many similarities to 

6 8 10 12 

Ti0 2 Weight percent 

Fig. 11. Plot of the Si0 2 and Fe 2 : 5 m garnet 
as a function of weight percentage of Ti0 2 , Salt- 
petre Kop. 


those from Saltpetre Kop. The monti- absence of vapor). The upper tempera- 

cellite and spinel mantles (Fig. 9B) on ture and pressure stability limit of com- 

olivine (Fo 79 _ 90 ) are less common than in plex Ba- and Ti-bearing phlogopite, 

the olivine melilitites from Saltpetre Kop. however, is not known. 

The phlogopite at Sutherland is charac- The enrichment of Ti in garnet is sim- 

terized by lower Ba (BaO 0.91-2.73 wt ilar to that reported for this mineral in 

%) and Ti(Ti0 2 l. 14-2.02 wt%) and higher the nephelinite-ijolite alkaline rock se- 

iron in the tetrahedral sites relative to ries. Investigation of the phase relations 

phlogopite at Saltpetre Kop. Melilite on the join andradite-Ti-garnet (Huck- 

compositions (Fig. 10B) overlap those enholz, 1969) shows that garnet crystal- 

from Saltpetre Kop. lizing as a liquidus phase has a composition 

between andr 86 Ti-gar 14 and andr 71 Ti-gar 29 

ry • and that the maximum degree of solid 

solution of 3CaOFe 2 3 , 3Ti02 in andra- 

The monticellite mantles on olivine dite is 55.5 wt % at 1137°C and 1 atm. 
phenocrysts at both Saltpetre Kop and Although the effect of pressure on the 
Sutherland Commage appear to have stability of Ti-rich garnet has not been 
formed by a reaction between olivine and investigated, its occurrence in associa- 
Ca-rich melt, perhaps prior to the crys- tion with melilite in the groundmass sug- 
tallization of melilite. The occurrence of gests that it formed at low pressure, 
euhedral phenocrysts of melilite in the Yoder (1975) showed that the maximum 
immediate vicinity of the monticellite stability limit of akermanite in an excess 
mantles that show no evidence of cor- of C0 2 is about 6 kbar. 
rosion and the absence of diopside both The crystallization trends of spinels in 
suggest that the mantles are not the the reaction mantle on olivine and in the 
product of the solid-state reaction inves- groundmass are toward a decrease in Mg, 
tigated by Yoder (Year Book 66, 471- Al, and Cr and an increase in total Fe, 
477), Me + Fo ss ^± Mo ss + Di. The com- Fe 3 + , and Ti. These trends mimic those 
position of the monticellite projects within observed in groundmass spinels in some 
the solvus of monticellite-forsterite, as kimberlites. In contrast to kimberlites, 
determined by Warner and Luth (1973). however, significant amounts of Ti are 
The mutual solubility tends to decrease incorporated in some of the silicate phases 
with pressure, so the unusual solid so- in olivine melilitites, such as phlogopite 
lution cannot be attributed to inheritance and garnet. This observation may ac- 
from depth. Because of the presence of count for the relative depletion of Ti in 
13-15 mol % of the kirschsteinite end spinels in olivine melilitites relative to 
member, the solvus may be closed by the those from some kimberlites. 
presence of ferrous iron; however, that The mineral chemistry of the ground- 
conclusion should be tested experimen- mass phases in olivine melilitites indi- 
tally. The monticellite mantles, there- cates that a carbonate-rich melt that 
fore, could be a stable or metastable concentrated the rare-earth and other in- 
product of the crystallizing liquid. compatible elements evolved through the 

Phlogopite at Saltpetre Kop is richer fractionation of their parent magma. The 

in Ba and Ti relative to most of the trioc- carbonate-rich portion of the melt ap- 

tahedral mica reported from melilite- pears to be the same type of residual melt 

bearing igneous rocks (Velde and Yoder, that evolves in the final stages of crys- 

Year Book 76, 478-483; Velde, Year Book tallization of kimberlite magma. Because 

78, 468-475). Forbes and Flower (1974) of the wide range of mineral composi- 

noted that the substitution of Ti in the tions, reversals in trends, and apparent 

phlogopite structure increases its ther- incompatibility relationships, it is con- 

mal stability (1350° or 1500°C, respec- eluded that the primary phases endured 

tively, at 30 kbar in the presence or highly variable conditions involving large 



changes in/ 02 and C0 2 /H 2 0, as well as 
pressure and temperature. In spite of 
these changes the olivine melilitite magma 
appears to be fractionating toward a re- 
sidual liquid similar to that in kimberlite. 
This conclusion does not imply that there 
is a genetic relationship between kim- 
berlites and olivine melilitites even though 
Yoder (1975) demonstrated theoretically 
that an olivine melilite leucitite could be 
converted to a kimberlite by the addition 
of C0 2 and H 2 0. The melilite assemblage 
at Saltpetre Kop and Sutherland is also 
stable with C0 2 and H 2 but only at very 
low pressures, in accord with the labo- 
ratory experiments on individual end 
members. The olivine melilitite and kim- 
berlite magmas may, therefore, have 
originated in different ways in the man- 
tle, yet each appears to be trending to- 
ward the same fractionation goal in the 


Duncan, R.A., R.B. Hargraves, and G.P. Brey, 
Age, palaeomagnetism and chemistry of melilite 
basalts in the Southern Cape, South Africa, Geol. 
Mag., 115, 317-396, 1978. 

Forbes, W.C., and M.F. J. Flower, Phase relations 
of titan-phlogopite K 2 Mg4TiAl 2 Si 6 02o(OH) 4 : a re- 
fractory phase in the upper mantle? Earth Planet. 
Sci. Lett., 22, 60-66, 1974. 

Huckenholz, H.G., Synthesis and stability of Ti 
andradites, Amer. J. Sci., Schairer Vol., 267 A, 
209-232, 1969. 

Mclver, J.R., and J. Ferguson, Kimberlitic, mel- 
ilititic, trachytic and carbonatite eruptions at 
Saltpetre Kop, Sutherland, South Africa, in 
Kimberlite Diatremes and Diamonds: Their Ge- 
ology, Petrology, and Geochemistry, F.R. Boyd 
and H.O.A. Meyer, eds., American Geophysical 
Union, Washington, D.C., pp. 111-128, 1979. 

Tuttle, O.F., and J. Gittins, eds., Carbonatites, 
John Wiley and Sons, Inc., New York, 1966. 

Warner, R.D., and W.C. Luth, Two-phase data on 
the join monticellite (CaMgSi0 4 )-forsterite 
(Mg 2 Si0 4 ): experimental results and numerical 
analysis, Amer. Mineral., 58, 998-1008, 1973. 

Yoder, H.S., Jr., Relationship of melilite-bearing 
rocks to kimberlites: a preliminary report on the 
system akermanite-C0 2 , Phys. Chem. Earth, 9, 
883-894, 1975. 

Correlation of Bimodal 18 

Values with Alteration Minerals 

of Synmetamorphic Granitic 

Rocks, Augusta, Maine* 

Douglas Rumble III, John M. Ferry, 
and T. C. Hoering 

Fluid flow is the chief factor controlling 
both transport of matter and convective 
heat transfer between different rock types 
during dynamic, regional metamorph- 
ism. The changes in mineralogic and iso- 
topic composition of rocks that were out 
of equilibrium with the fluids delineate a 
pattern of the quantity and pathways of 
fluid flow during this process. Comple- 
mentary isotopic and mineralogic data are 
reported from three synmetamorphic 
granitic plutons from a Buchan-type ter- 
rain in south-central Maine where the 
stocks were out of equilibrium with in- 
filtrating fluids derived from contiguous 
metasediments. Initially, the stocks dif- 
fered little in primary mineralogy and Sr 
isotope geochemistry, but after intrusion 
they could be divided into two groups 
based on correlations between secondary 
mineralogy, 18 content, geographic lo- 
cation, and the nature of the wall rocks. 

The three stocks, which occur along a 
common axis, are, from west to east, the 
Hallowell Quartz Monzonite, the Togus 
Quartz Monzonite, and the Three Mile 
Pond Granodiorite. The primary igneous 
mineralogy of the three, described by 
Chayes (1952), Barker (1964), and Ferry 
(1978, 1979), is comparable and consists 
of quartz, K-feldspar, plagioclase, bio- 
tite, and muscovite with accessory gra- 
net, apatite, ilmenite, and sphene. Table 
2 summarizes the Rb-Sr ages and the 
initial 87 Sr/ 86 Sr, which are the same for 
all three bodies. 

In spite of the major, common features 
of the three plutons, they can be divided 
into two groups. In Fig. 12 the two groups 
of analyses of 18 in whole-rock samples 
fall into distinct clusters with no over- 
laps. The isotopic compositions of min- 

*Research supported in part by National Science 
Foundation grant EAR 82-06017. 



TABLE 2. Rb-Sr Whole-Rock Ages and Initial 
Sr Isotope Ratios* 


Age, m.y. 

r Sr/ 86 Sr), 

Hallowell 387 ± 11 

Togus 394 ± 8 

Three Mile Pond 393 ± 15 

0.705 ± 0.002 
0.7049 ± 0.0007 
0.7058 ± 0.005 

*Dallmeyer and Van Breeman, 1981. 

erals separated from the same samples 
also show a bimodal distribution, which 
demonstrates that the grouping is not 
merely a consequence of variations in the 
proportions of primary minerals. The 
minerals quartz, feldspar, and muscovite 
do not record a state of isothermal, iso- 
topic exchange equilibrium; tempera- 
tures calculated from quartz-muscovite 
pairs are at least 100°C higher than those 
from quartz-feldspar. 

Alteration minerals of the plutonic rocks 
account for no more than a few volume 

percent of the total, yet they differ sys- 
tematically in the two groups of rocks 
with similar oxygen isotope ratios. Pla- 
gioclase in Group-I rocks is altered only 
to muscovite + calcite; plagioclase in 
Group II is altered only to muscovite + 
epidote, muscovite + epidote + calcite, 
or, in one unusual sample, muscovite 
alone. The key correlation is between 18 
and the presence or absence of epidote. 
Two-feldspar temperatures of Group-I 
rocks are in the range 376°-450°C; two- 
feldspar temperatures of Group-II rocks 
are in the range 385°-483°C (Ferry, 1978). 
Temperatures recorded by the equilib- 
rium among feldspars, alteration min- 
erals, and C0 2 -H 2 fluid for Group-II 
rocks are in the range 415°-440°C (Ferry, 

Group-I and Group-II rocks are sep- 
arated geographically by a roughly north- 
south line that follows closely, but not 




t T 




5 - 

5 - 

O d 8n H d nXX quartz _ 

M^m&MM Q nfifi R n nn£ 










5 - 

mn n nnx 








5 18 0, 


Fig. 12. Histogram of whole-rock, quartz, alkali feldspar, and muscovite 5 lx O analyses. Western 
pluton is Hallowell Quartz Monzonite, middle pluton is Togus Quartz Monzonite, and eastern pluton is 
Three Mile Pond Granodiorite. Filled symbols denote presence of secondary epidote. Open symbols show 
rocks lacking epidote. 



exactly, the stratigraphic boundary be- 
tween the two major, Silurian-age, me- 
tasedimentary units in the area. Group- 
I rocks lie to the west of the line and are 
associated in outcrop with the Waterville 
Formation. Group II is associated with 
the Vassalboro Formation. The two rock 
groups are associated generally, but not 
exclusively, with different plutons; most 
Group-I rocks occur in the Hallowell body 
or its dikes and satellite intrusions, and 
most Group-I I rocks occur either in the 
Togus or the Three Mile Pond stock. 

The similarity between the tempera- 
tures of oxygen isotope exchange and the 
temperatures calculated from phase 
equilibria indicates that the isotopic and 
mineralogical alteration of the granitic 
rocks occurred in the same event. The 
alteration resulted from the infiltration 
of granite by C0 2 -H 2 0-CH 4 fluids during 
metamorphism after emplacement of the 
stocks. Ferry (1980, 1983) showed that 
the time-integrated flux of fluid through 
metasedimentary rock adjacent to the 
granitic plutons during the regional 
metamorphic event was from two to four 
rock volumes. The differences between 
Group I and Group II can best be ex- 
plained by infiltration by two distinct 
fluids. Whether or not the two groups of 
rocks also differ systematically in the 
amounts of fluid that infiltrated them 
cannot be determined at present because 
of insufficient data on the 18 values of 
wall rocks. Thermochemical calculations 
show that the fluids that infiltrated the 
Group-I rocks contained higher concen- 
trations of C0 2 than the fluids that infil- 
trated the Group-II rocks (Ferry, 1978, 
1979). The isotopic data show that the 
fluids that infiltrated the Group-I rocks 
were richer in 18 than the fluids that 
infiltrated the Group-II rocks. The cor- 
relation of Group-I and Group-II rocks 
with different country rocks suggests a 
reason why the two groups were infil- 
trated by fluids of different chemical and 
isotopic composition. Marble and calc-sil- 
icate are more abundant in the Water- 
ville Formation, which is closely 
associated with Group-I rocks. Calcar- 

eous metasediments are less abundant in 
the Vassalboro Formation, which is 
closely associated with Group-II rocks. 
Fluids generated during metamorphism 
of the Waterville Formation (which in- 
filtrated Group-I rocks), therefore, were 
likely both richer in C0 2 and 18 than 
fluids generated during metamorphism 
of the Vassalboro Formation (which in- 
filtrated Group-II rocks). The geo- 
graphic distribution of Group-I and Group- 
II rocks has important implications for 
the manner in which fluid flowed through 
the study area during regional meta- 
morphism. It is evident that metasedi- 
ments control the chemical and isotopic 
compositions of the fluids they release 
during devolatilization reactions. Evi- 
dence presented herein suggests that the 
direction of intergranular fluid flow is 
controlled by stratigraphic boundaries and 
structural fabric and that fluid flow is 
largely parallel to the boundaries of the 
stratigraphic units in which the fluids are 


Barker, D.S., The Hallowell granite, south-central 
Maine, Amer. J. Sci., 262, 592-613, 1964. 

Chayes, F., The finer-grained calcalkaline granites 
of New England, J. GeoL, 60, 207-254, 1952. 

Dallmeyer, R.D., and 0. Van Breeman, Rb-Sr 
whole-rock and 40 Ar/ 39 Ar mineral ages of the To- 
gus and Hallowell Quartz Monzonite and Three 
Mile Pond Granodiorite plutons, south-central 
Maine, Contrib. Mineral. Petrol., 78, 61-73, 1981. 

Ferry, J.M., Fluid interaction between granite and 
sediment during metamorphism, south-central 
Maine, Amer. J. Sci., 278, 1025-1056, 1978. 

Ferry, J.M., Reaction mechanisms, physical con- 
ditions, and mass transfer during hydrothermal 
alteration of mica and feldspar in granitic rocks 
from south-central Maine, Contrib. Mineral. Pe- 
trol., 68, 125-139, 1979. 

Ferry, J.M., A case study of the amount and dis- 
tribution of heat and fluid during metamorphism, 
Contrib. Mineral. Petrol, 71, 373-385, 1980. 

Ferry, J.M., Regional metamorphism of the Vas- 
salboro Formation, south-central Maine: a case 
study of the role of fluid in metamorphic pedo- 
geneses, J. Geol. Soc. London, in press, 1983. 



Textural and Isotopic Variation 

of Graphite in the New Hampshire 

Plutonic Series* 

Edward F. Duke and Douglas Rumble III 

The carbon isotopic composition of 
graphite from igneous rocks of the New 
Hampshire Plutonic Series (NHPS) and 
contiguous metasedimentary wall rocks 
has been analyzed in order to measure 
the extent to which fluid components were 
exchanged between synmetamorphic 
plutons and wall rocks during regional 
metamorphism. Analyses have also been 
made of graphite from cross-cutting vein 
deposits associated with the plutons. 
Carbon isotopes in igneous rocks usually 
have been discussed in terms of carbon 
reservoirs in the earth's crust and mantle 
and in terms of biogenic or abiogenic or- 

* Research supported in part by National Science 
Foundation grant EAR 82-06017. 

igin of reduced carbon. The values of 8 13 C 
of igneous and vein graphites reported 
in Fig. 13, however, are intermediate be- 
tween the average values of crustal bio- 
genic reduced carbon (-25%o), crustal 
biogenic carbonate (0%c), and mantle car- 
bon (-5%o) (Ohmoto and Rye, 1979). 
Carbon isotope ratios of graphites from 
metamorphosed igneous and sedimen- 
tary rocks are sensitive to isotopic ex- 
change with fluids because of the 
dependence of fluid-graphite fractiona- 
tion on both temperature and carbon spe- 
ciation (Ohmoto and Rye, 1979). Fluid- 
graphite interactions are of particular in- 
terest in the study area, where plutons 
containing desseminated graphite are 
closely associated with hydrothermal 
graphite vein deposits. This report sum- 
marizes the first detailed study of both 
textural and isotopic variations in graph- 
ite from igneous rocks, and shows that 
these variations provide an important re- 


• >80 mesh(0.18 mm) 
A < 230 mesh (0.06mm) 
O unsized 

tie lines connect 

values from same sample 



I L 

J I I L 

J I 

J L 

J I I L 

■ ° , ■ 


I L 

I I I L 

O O 

J I I I I I I L 



J L 

J L 

J L 

J L 




5 13 c 



, %o 

Franklin Pierce Mine 
Mt. Kearsarge Mine 

granodiorite-granite -apl ite 
series (peraluminous) 

tonalite series 

diorite-quartz diorite- 

calc-silicate granofels 
pelitic schist 


Fig. 13. 5 13 C values of graphite from igneous rocks, graphite vein deposits, and metasedimentary 
host rocks related to the New Hampshire Plutonic Series. Data points represent a minimum of two 
determinations; standard deviations (la) of replicate analyses are smaller than size of symbols except 
for one vein sample where error bar is shown. Symbols represent different size fractions of graphite and 
reflect textural differences as explained in the text. 


cord of fluids in the plutonic and hydro- tated from vapor-saturated magma at the 

thermal environment. same time as the other primary minerals. 

It is also possible that type-1 graphite is 

^ 7 . 0jjt . , ~ . , . - a refractory phase that survived partial 

Geologic Setting and Description of mdting Qr assimilation of parental me _ 

Samples tasediments. The association of type-2 
Samples of igneous rocks selected for graphite with retrograde alteration of 
study are from the Spaulding Intrusive primary silicates suggests that the 
Suite, a subseries of the Devonian NHPS. spherulitic graphite grew after crystal- 
Most samples are from the Blackwater lization. Type-2 graphite may have pre- 
Pluton, south-central New Hampshire, a cipitated from residual deuteric fluid or 
semiconcordant sheet approximately 2.5 it may have crystallized when metamor- 
km thick with a late-tectonic age of em- phic fluids infiltrated the pluton from the 
placement (Neilson et al., 1976). The wall rocks, 
temperature of intrusion, estimated from 

the garnet-biotite geothermometer of Mfthod* of Vtvdv 
Ferry and Spear (1978), was 600°-750°C memoas oj ziuay 
with late-magmatic or subsolidus Fe-Mg Previous studies of the isotopic corn- 
exchange persisting to at least 525°C in position of reduced carbon in igneous rocks 
many samples. The pressure during report neither mineralogical nor textural 
emplacement, estimated from garnet- data on carbon (Craig, 1953; Hoefs, 1973; 
plagioclase-orthopyroxene-quartz as- Fuex and Baker, 1973). A primary goal 
semblages (Newton and Perkins, 1982), of the present study was to determine 
was approximately 3.6-4.1 kbar. the isotopic composition of the textural 
Graphite in both plutonic and meta- varieties of graphite that the writers have 
sedimentary rocks displays two general observed in plutonic rocks. Accordingly, 
textural varieties: (1) coarse (0.1-1.0 mm) a scheme was devised to separate coarse 
flakes and (2) fine (0.01-0.1 mm), irreg- (type-1) from fine-grained (type-2) 
ular, poly crystalline aggregates, which graphite. 

grade into graphite spherulites with a Samples were pulverized with a jaw 
well-developed radial internal structure, crusher and a disk mill fitted with por- 
Vein graphite includes coarse (5.0 mm), celain grinding disks. In order to avoid 
bladed crystals and nodular masses with grinding below the natural grain sizes of 
both radial and concentric internal struc- the minerals, samples were crushed to 
tures. Graphite flakes (type 1) occur in <5 mm, at which point graphite was con- 
fresh and altered rocks alike and are fol- centrated by flotation on water. Then, 
iated parallel to primary silicates such as concentrates from samples showing co- 
biotite, indicating growth before or dur- existence of both textural varieties of 
ing deformation. Fine graphite aggre- graphite— coarse flakes and fine spher- 
gates or spherulites (type 2) are generally ulites — were sieved carefully to separate 
restricted to rocks showing some degree the two textural forms so that a test could 
of alteration (hydration with or without be made of isotopic inhomogeneity. Sam- 
carbonation), and like vein graphite, their pies were converted to C0 2 by combus- 
delicate textures and lack of planar ori- tion in evacuated silica tubes with an 
entation require growth in a static en- excess of CuO and Cu metal at 900°C 
vironment. These textural observations (Stump and Fraser, 1973). Carbon diox- 
indicate that coarse-flake graphite grew ide was analyzed for 8 13 C values on an 
before fine, spherulitic graphite. isotope-ratio mass spectrometer with re- 
Interpretation of the textural features suits calibrated to National Bureau of 
has led to multiple working hypotheses Standards isotope reference materials 
for the origin of the two types of graph- Nos. 21 and 22 with 8 13 C PDB values of 
ite. Type-1 graphite may have precipi- -28.130%c and -29.61%o, respectively, 


relative to the Pee Dee Belemnite stan- ing flake (type-1) and spherulitic (type- 

dard. The standard deviation of duplicate 2) graphite. Although the sieving treat- 

or triplicate analyses was 0.1 1% , on av- ment used here is a crude method for 

erage. separating the graphite textural vari- 
eties and provides only a minimum es- 

Discussion of Results tir " ate **** *?? isoto P ic sM \' [t j s 

noteworthy that in several samples the 

The 8 13 C values of graphite are shown shift measures approximately 2%o, far 

in Fig. 13 for various rock types exam- larger than the analytical uncertainty of 

ined in this study. The isotopic compo- 0.1 l%o. Local variations in isotopic com- 

sition of graphite in both metasedimentary position were also found in vein-graphite 

rocks and granitoids is variable, but samples, where values from one vein vary 

granitoids show a correlation between up to 3. 5%o, and two samples 3 cm apart 

8 13 C values and chemical and mineral- differ by over 2%o. The difference in iso- 

ogical subdivisions. Graphite in the per- topic composition within individual sam- 

aluminous tonalite series (characterized pies is comparable to that reported by 

by accessory garnet, muscovite, cordier- Weis (1980) between flake and "skeleton" 

ite, and sillimanite) is isotopically indis- graphite (8 13 C = -2.2% and -4.5%p, 

tinguishable from graphite in associated respectively) in graphitic marble sam- 

metasedimentary rocks. The coarse-flake pies. 

graphite of the tonalite series could have The demonstration of isotopic inho- 

originated through anatexis or assimi- mogeneity in graphite of igneous rocks 

lation of similar sedimentary material, illustrates the need for cautious inter- 

The metaluminous diorite-quartz dior- pretation of 5 13 C values, but it also re- 

ite-tonalite series (amphibole-bearing, veals the potential of combined petrologic 

with or without pyroxene) contains and isotopic studies in reconstructing the 

graphite that appears to be slightly en- history of fluid-rock interactions. Defin- 

riched in 13 C relative to the analyzed me- itive choices between multiple working 

tasediments. Graphite in the peraluminous hypotheses for the origin of graphite must 

granodiorite-granite-aplite series and await the completion of more research, 

graphite veins is enriched in 13 C and re- especially concerning the physicochemi- 

quires either a source of carbon other cal state of retrograde fluids in the plu- 

than wall-rock metasediments or carbon tons, metamorphic fluids in the wall rocks, 

isotope fractionation during the process and graphite vein fluids. Evidence of a 

leading to precipitation of graphite. correlation between different textural 

In addition to the large variation be- generations of graphite in plutonic rocks 
tween samples in Fig. 13, this study doc- and variations in carbon isotope ratios 
uments significant carbon isotope confirms that fluids in the plutons evolved 
inhomogeneity within individual sam- in composition during crystallization and 
pies. In selected granitoid and metase- subsequent retrograde metamorphism. 
dimentary rocks an isotopic shift of up to The inhomogeneity of vein graphites 
2%o was noted between coarse-flake (>80- points toward a history of fluids of dif- 
mesh) graphite and fine (<230-mesh) ferent isotopic and probably different 
graphite from the same sample. (Note chemical composition, becoming mixed 
different symbols in Fig. 13.) The sepa- as they flowed through fissures, 
ration of textural forms achieved by siev- 
ing was accentuated by separating References 
spherical from platy grains on a tilted 

Stainless Steel plate. The observed dif- Crai g' H ' The geochemistry of the stable carbon 

ferences in 5 13 C between coarse- and fine- [ ™^ es > Geochim - Cosmochim - Act ^ *> 53 " 92 > 

Size fractions, therefore, reflect differ- Ferry, 'j.M., and F.S. Spear, Experimental cali- 

ences in isotopic composition of Coexist- bration of the partitioning of Fe and Mg between 



biotite and garnet, Contrib. Mineral. Petrol., 66, 
113-117, 1978. 

Fuex, A.N., and D.R. Baker, Stable carbon iso- 
topes in selected granitic, mafic, and ultramafic 
igneous rocks, Geochim. Cosmochim. Acta, 37, 
2509-2521, 1973. 

Hoefs, J., Ein Beitrag zur Isotopengeochemie des 
Kohlenstoffs in magmatischen Gesteinen, Con- 
trib. Mineral. Petrol, U, 277-300, 1973. 

Neilson, D.L., R.G. Clark, J.B. Lyons, E.J. En- 
glund, and D.J. Borns, Gravity models and mode 
of emplacement of the New Hampshire Plutonic 
Series, Geol. Soc. Amer. Mem., U6, 301-318, 

Newton, R.C., and D. Perkins, Thermodynamic 
calibration of geobarometers based on the as- 
semblages garnet-plagioclase-orthopyroxene 
(clinopyroxene)-quartz, Amer. Mineral., 67, 203- 
222, 1982. 

Ohmoto, H., and R.O. Rye, Isotopes of sulfur and 
carbon, in Geochemistry of Hydrothermal Ore 
Deposits, 2d ed., H. L. Barnes, ed., Wiley-In- 
terscience, New York, pp. 509-567, 1979. 

Stump, P.K. and J.W. Fraser, Simultaneous de- 
termination of C, H, and N in organic com- 
pounds, Nucl. Sci. Abstr., 38, 7848, 1973. 

Weis, P.L., Graphite skeleton crystals — a newly 
recognized morphology of crystalline carbon in 
metasedimentary rocks, Geology, 8, 296-297, 1980. 

mlneralogical and lead isotope 
Studies of a Polymetallic Sulfide 

Deposit at Derhib, Egypt: 

Implications for Metallogenesis 

in the Red Sea District 

N. Z. Boctor and F. Tera* 

Polymetallic Cu-Zn-Pb-Fe sulfide ores 
occur in association with metavolcanic 
rocks and metasediments in different 
parts of the world and range in age from 
Archean to Phanerozoic. Their classifi- 
cation, origin, and geotectonic setting 
have been discussed by Sawkins (1976) 
and Franklin et al. (1981). In the Red 
Sea region in the Eastern Desert of 
Egypt, polymetallic sulfide mineral- 
izations occur in four localities in asso- 
ciation with intermediate to basic 
metavolcanic rocks of Precambrian age 
(825-856 m.y.). These metavolcanic rocks 
are believed to be related to an episode 
of island-arc volcanism (Garson and Shal- 

; Department of Terrestrial Magnetism. 

aby, 1976). A notable difference between 
these deposits and other polymetallic sul- 
fide deposits is their exclusive occur- 
rence in talc deposits that occupy fracture 
zones running roughly east-west with re- 
spect to the axial trough of the Red Sea. 
The question arises how these deposits 
formed. Two alternatives are formation 
(1) by volcanic exhalative processes in an 
island-arc environment or (2) by later lo- 
calized redistribution of stratiform sul- 
fide deposits (Garson and Shalaby, 1976). 
A third possibility is that they are of epi- 
genetic origin, formed by the hydrother- 
mal processes that formed the talc 
deposits. One of these deposits at Derhib 
occurs in a talc shear zone about 600 m 
long and 150 m wide. The shear zone oc- 
curs within a sequence of rhyolitic and 
andesitic flows that overlie unconform- 
ably older schists and granites and are 
intruded by younger granites and dior- 

Petrography and Mineralogy 

The metavolcanic rocks in the imme- 
diate vicinity of the talc shear zone are 
composed mainly of hornblende, plagio- 
clase (Ab 6 2_6sOr 3_ 5 An 3 4_3 7 ), and minor 
amounts of biotite. Opaque minerals are 
represented by magnetite-ilmenite in- 
tergrowths and discrete ilmenite crys- 
tals. In the shear zone, talc is the major 
constituent and may be associated with 
minor amounts of more-calcic plagioclase 
(Ab 517 Or 3 An 48 ). Quartz is present as 
veinlets or as inclusions in porphyro- 
blasts of pyrite up to 1 cm in diameter. 
Tremolite and quartz occur in the talc in 
the mineralized areas of the shear zone. 
The ore minerals fill the interstices be- 
tween tremolite crystal aggregates and 
talc, and in some places, replace tremo- 
lite. The ore is composed mainly of sphal- 
erite and chalcopyrite; pyrite and galena 
are minor phases, though locally abun- 
dant, with rare molybdenite and hessite. 
Sphalerite shows randomly distributed 
inclusions of chalcopyrite of various sizes. 
The FeS content of such sphalerite, where 
it coexists with pyrite and chalcopyrite, 



is quite variable within individual spec- 
imens, as well as from one specimen to 
another, and ranges between 6.0 and 12.7 
mol %. Minor amounts of marcasite and 
covellite are commonly associated with 
the chalcopyrite-sphalerite pyrite as- 
semblage and usually occur at the inter- 
faces between sphalerite and chalcopyrite. 
Rare sphalerite inclusions in pyrite have 
the highest iron content (14.9 ±0.5 mol 
% FeS). The lowest FeS content (1.25- 
1.53 mol %) was in sphalerite where it 
existed in contact with galena and rare 
pyrite and chalcopyrite in galena-rich ore. 
Pyrite occurred in many specimens as 
idiomorphic crystals, in some places in 
close proximity with marcasite, with no 
indication of replacement. Galena is a rare 
mineral except in the specimens with Fe- 
poor sphalerite, where it is locally abun- 
dant. Hessite occurs rarely in mutual 
contacts with galena, sphalerite, or chal- 
copyrite. Molybdenite occurs as de- 
formed acicular crystals disseminated in 
chalcopyrite or sphalerite. 

Lead Isotope Investigation 

The isotopic investigation of the ore 
and country rocks is at a preliminary 
stage. Lead isotopic compositions of the 
country rock, galena, pyrite, and sphal- 
erite were determined by the method de- 
scribed by Tera and Wasserburg (1975). 
The results are shown in Table 3. The 
errors are estimated at 0.05% per mass 
unit. The country rock and sulfide min- 
erals contain lead of identical isotopic 
composition, a result indicating complete 
isotopic equilibration in a very recent time 
(0-100 m.y.). The isotopic data are dif- 


Lead Isotope Data for Whole Rock 
and Sulfide Minerals 


p b 206/p b 204 

p b 207/p b 204 

p b 208/p b 204 





*Metavolcanic host rock close to the talc shear 

ferent from those reported for galena from 
Phanerozoic massive sulfide ores from is- 
land- or continental-arc environments 
(Doe and Zartman, 1979) and from those 
reported for Precambrian massive sul- 
fide ores of the southern Appalachians 
(LeHuray, 1982). 


Application of sphalerite geobarome- 
try to the Derhib ore is hampered by the 
lack of the assemblage hexagonal pyr- 
rhotite-pyrite-sphalerite, by the pres- 
ence of inclusions of chalcopyrite in 
sphalerite, which make it unsuitable for 
geobarometry (Hutchison and Scott, 
1981), and by the low-temperature re- 
actions that affected the ore. The coex- 
istence in the ore of pyrite and marcasite 
as well as covellite and chalcopyrite in- 
dicates metastability (Barton and Skin- 
ner, 1979). Equilibration to temperatures 
below ~145°C is demonstrated by the 
presence of the low-temperature poly- 
morph of hessite (Kracek et al. y 1966). 

Sphalerite inclusions in pyrite are 
known to preserve the high P-T equilib- 
rium under which they formed (Hutchi- 
son and Scott, 1980). The maximum 
pressure obtained for sphalerite encap- 
sulated in pyrite, calculated from the data 
of Hutchison and Scott (1981), is 4.7 ± 
0.5 kbar. This pressure is consistent with 
the low limit of amphibolite-facies me- 
tamorphism. The variable iron content of 
sphalerite cannot be reconciled with a 
single metamorphic episode. A hydro- 
thermal event may have been responsi- 
ble for the wide variation in the iron 
content of chalcopyrite-bearing sphaler- 
ite and the precipitation of iron-poor 
sphalerite in galena-rich samples. These 
observations imply an epigenetic origin 
of the ore. It is interesting to note that 
Zierenberg and Shanks (1983) described 
epigenetic, sulfide-bearing, hydrother- 
mal talc and magnesian smectites from 
veins and layers in the deep zones of the 
metalliferous sediments, Atlantis II Deep, 
Red Sea. They attributed their origin to 
convective circulation of sea water which 



has been heated and has interacted with 
rift-zone volcanic rocks. Although the 
tectonic setting of the Derhib ore is dif- 
ferent, it is likely that hydro thermal fluids 
circulating through the fracture zone were 
responsible for the formation of the talc 
and associated sulfides. Nakhla et al. 
(1973) argued that the trace elements in 
the sulfide minerals in the ore were de- 
rived from the metavolcanic rocks. The 
retarded nature of lead in the metavol- 
canic host rock and the ore (Pb 206 /Pb 204 
= 17.3) in comparison with the average 
value for terrestrial lead (Pb 206 /Pb 204 = 
18.3) is, however, consistent with re- 
mobilization of Precambrian-Cambrian 
lead-bearing deposits. The recent age in- 
dicated by the lead isotopes (0-100 m.y.) 
is significant because within this time span 
the Red Sea ocean-floor spreading, to 
which the east-west system of fractures 
in the Eastern Desert and Sudan may be 
related, took place. The striking con- 
stancy of the lead isotope data may rep- 
resent the isotopic imprint of this 
prominent event. 


Barton, P.B., Jr., and B.J. Skinner, Sulfide min- 
eral stabilities, in Geochemistry of Hydrother- 
mal Ore Deposits, 2ded., H.L. Barnes, ed., John 
Wiley and Sons, Inc., New York, pp. 278-403, 

Doe, R.B., and R.E. Zartman, Plumbotectonics, 
the Phanerozoic, in Geochemistry of Hydrother- 
mal Ore Deposits, H.L. Barnes, ed., John Wiley 
and Sons, Inc., New York, pp. 22-71, 1979. 

Franklin, J.M., J.W. Lydon, and D.F. Sangster, 
Volcanic associated massive sulfide deposits, 
Econ. GeoL, 75th Ann iv. Vol., 485-627, 1981. 

Garson, M.S., and I.M. Shalaby, Precambrian- 
Lower Palaeozoic plate tectonics and metallo- 
genesis in the Red Sea region, in Metallogeny 
and Plate Tectonics, Geol. Assn. Can. Spec. Pap., 
U, 573-596, 1976. 

Hutchison, M.N., and S.D. Scott, Sphalerite geo- 
barometry applied to metamorphosed sulfide ores 
of the Swedish Caledonides and U.S. Appala- 
chians, Norg. Geol. Unders. Skr., 360, 59-71, 

Hutchison, M.N., and S.D. Scott, Sphalerite geo- 
barometry in the Cu-Fe-Zn-S system, Econ. Geol. , 
76, 143-153, 1981. 

Kracek, F.C., C.J. Ksanda, and L.J. Cabri, Phase 
relations in the silver-tellurium system, Amer. 
Mineral., 51, 14-27, 1966. 

LeHuray, A. P., Lead isotopic patterns of galena 
in the Piedmont and Blue Ridge ore deposits, 
southern Appalachians, Econ. Geol., 77, 335-351, 

Nakhla, F.M., M. Rasmy, and A.B. Basily, Con- 
tribution to the sulfide mineralization at Gebel 
Derhib Talc Mine, Egypt, Neues Jahrb. Min- 
eral. Abh., 118, 149-158, 1973. 

Sawkins, F.J., Massive sulfide deposits in relation 
to geotectonics, in Metallogeny and Plate Tec- 
tonics, Geol. Assn. Can. Spec. Pap., 1!+, 221- 
242, 1976. 

Tera, F., and G.J. Wasserburg, Precise isotopic 
analyses of Pb in picomole and subpicomole quan- 
tities, Anal. Chem., 1,3, 2214-2220, 1975. 

Zierenberg, R.A., and W.C. Shanks III, Miner- 
alogy and geochemistry of epigenetic features in 
metalliferous sediment, Atlantis II Deep, Red 
Sea, Econ. Geol, 78, 57-72, 1983. 

Partitioning of Nickel between 
Silicate and Iron Sulfide Melts 

N. Z. Boctor and H. S. Yoder, Jr. 

Iron-nickel sulfide ores are generally 
associated with mafic and ultramafic rocks. 
The Ni content of these deposits as well 
as their Cu/(Cu + Ni) are a function of 
the composition of their parent magmas 
(Naldrett and Cabri, 1976). These sulfide 
deposits are believed by some investi- 
gators to have formed by sulfide-silicate 
liquid immiscibility (Rajamani and Nal- 
drett, 1978; Naldrett, 1981). Study of the 
partitioning of Ni and other transition 
metals between sulfide and silicate melts, 
therefore, is necessary for an under- 
standing of the genesis of magmatic iron- 
nickel sulfide ores. Maclean and Shima- 
zaki (1976) studied the partitioning of Ni 
and other transition metals between sul- 
fide and silicate liquids in the system FeS- 
FeO-Si0 2 at 1150°C. Rajamani and Nal- 
drett (1978) extended the investigations 
of Maclean and Shimazaki to melts of bas- 
altic, olivine basaltic, and andesitic com- 
positions at temperatures between 1255° 
and 1325°C. Because alumina crucibles in 
sealed, evacuated silica tubes were used 
as sample containers, Rajamani and Nal- 
drett were unable to extend the inves- 
tigations to higher temperatures or to 
buffer the fugacities of oxygen and sul- 


In the present study the partitioning durations based on these experiments 
of Ni between sulfide and silicate melts were about 1 hr at 1300°C, 40 min at 
was investigated in the temperature range 1400°C, and 25 min at 1460°C. Rajamani 
1300°-1460°C under conditions in which and Naldrett (1978) found that equilib- 
both / and /g 2 were controlled by the rium between sulfide and silicate melts 
use of the technique described by Boctor in their experiments was reached in 1 hr 
{Year Book 81, 366-369). The initial com- at temperatures between 1250° and 
positions of the silicates were synthetic 1325°C. The silicate and sulfide melts were 
basaltic glass similar to that used by Ra- analyzed by the electron microprobe, and 
jamani and Naldrett (1978) (Si0 2 52.0, the MAGIC computer program of Colby 
A1 2 3 16.0, FeO 11.5, MgO 9.1, CaO 9.1, (1971) was used for data reduction. Nickel 
Na 2 2.3 wt %) and a more-mafic glass concentration in the silicate melt was de- 
containing —35% normative olivine (Si0 2 termined by a trace-element computer 
47.5, Al 2 3 8.9, FeO 14.3, MgO 21.6, CaO program (see Boctor, Year Book 81, 366- 
8.1, Na 2 0.44 wt %) similar to that syn- 369). In the more-mafic glass, a few spi- 
thesized by Arndt (Year Book 76, 553- nifex-textured olivine crystals were 
556). The reaction between silicate and formed on quench. Glass free from any 
sulfide melts was reversed, with Ni- of these crystals was analyzed, and runs 
bearing sulfide in one set of experiments containing more than —3% olivine were 
and Ni-bearing silicate in the reversed discarded. 

experiments, while the bulk composition The experimental results are shown in 

was maintained nearly constant. Table 4. The distribution coefficient, K D , 

Use of the Pt loop-glass bead tech- is independent of the concentration of Ni 

nique proved to be unsuccessful in this in the sulfide melt (1.5-9.8 wt % Ni) or 

investigation. Because of density differ- the silicate melt (0. 1-5 wt % Ni). The K D 

ences between the silicate and sulfide liq- is temperature dependent for basaltic 

uid, the latter tended to separate and melts (7.5-9.8 wt % MgO) but is also 

accumulate in the glass tube at the bot- influenced by changes in/ 02 - Changes in 

torn of the furnace. A thin Pt-5% Au foil / s do not seem to have a significant effect 

(100 |xm) was placed on the loop to pre- on the K D . The K D decreases as the melt 

vent the loss of the sulfide melt. A series changes from basaltic to more-mafic com- 

of experiments were undertaken to de- position (19-22 wt % MgO). 

termine (1) the amount of iron loss to the The distribution coefficients deter- 

foil as a function of time and (2) the times mined in this investigation are difficult 

required for equilibration of the sulfide to compare with those determined by 

and silicate melt. Excess Fe as Fe^^.O Rajamani and Naldrett (1978) because the 

was added to the oxide mix to compen- f and /g 2 in their experiments are not 

sate for the loss of Fe to the foil. Run known. The average K D of 43.1 ± 3.2 


Distribution Coefficients for Ni between Silicate and Iron Sulfide Melt 

No. of 

Glass Type Temp., °C -'°2 ->s 2 K D Ni * 

6 Basaltic 1400 ± 5 10 7 " 10 " 2 - 92 35.9 ±2.5 

6 Basaltic 1300 ± 3 10 7 " 10 " 272 43.1 ± 3.2 

6 Basaltic 1300 ± 5 10 902 10 ~ 270 50.9 ± 2.6 

6 Basaltic 1300 ± 5 10" 897 10 2:54 49.4 ±3.0 

8 Maficf 1460 ± 2 10 ~ 8 - 55 10 ~ 295 22.4 ±2.3 

L D 

^NiS \ / / -^NiO 

^FeS/ ail , fiHo mo it / V^Fe 

sulfide melt ' x reu/ s jij ca t e me it 

tGlass more mafic than basalt with —35% normative olivine. 



(Table 4) at 1300°C, /b 2 10' 7 ", and ./ s , 
j^Q-2.72 j s comparable to their K D value 
of 38 for basaltic melts if it is taken into 
account that the techniques used in their 
experiments and the present investiga- 
tion are quite different. The decrease in 
K D as the melt composition changes from 
basaltic to more-mafic compositions sup- 
ports the suggestions that the Ni content 
of Fe-Ni sulfide ores is controlled by the 
mafic nature of the magma and that these 
ores form by liquid immiscibility from 
magmas that are more enriched in Ni 
relative to basaltic magmas. The distri- 
bution coefficients determined in this in- 
vestigation can be used in conjunction 
with the models developed by Duke and 
Naldrett (1978) and Duke (1979) to sim- 
ulate fractionation of mafic and ultra- 
mafic magmas. These models, however, 
have not taken into account the fraction- 
ation of the oxide phases, such as chrom- 
ites and magnetite, which could act as 
sinks for transition metals during the 
fractionation of mafic and ultramafic 
magmas. The distribution coefficients for 
transition metals between these oxide 
phases and silicate and sulfide melts should 
be determined experimentally. 


Colby, J.D., MAGIC IV, a computer program for 
quantitative electron microprobe analysis, Bell 
Telephone Laboratories, Allentown, Pennsyl- 
vania, 1971. 

Duke, J.M., Computer simulation of the fraction- 
ation of olivine and sulfide from mafic and ultra- 
mafic magmas, Can. Mineral, 17, 507-514, 1979. 

Duke, J.M., and A.J. Naldrett, A numerical model 
of the fractionation of olivine and molten sulfide 
from komatiite magma, Earth Planet. Sci. Lett., 
39, 255-266, 1978. 

Maclean, W.H., and H. Shimazaki, The partition 
of Co, Ni, Cu, and Zn between sulfide and silicate 
liquids, Econ. Geo!., 71, 1049-1057, 1976. 

Naldrett, A.J., Nickel sulfide deposits: classifica- 
tion, composition, and genesis, Econ. GeoL, 75th 
Anniv. Vol., 628-686, 1981. 

Naldrett, A.J., and L.J. Cabri, Ultramafic and re- 
lated mafic rocks: with special reference to the 
concentration of nickel sulfides and platinum-group 
elements, Econ. GeoL, 71, 1131-1158, 1976. 

Rajamani, V., and A.J. Naldrett, Partitioning of 
Fe, Co, Ni and Cu between sulfide and basaltic 
melts and the composition of Ni-Cu sulfide de- 
posits, Econ. GeoL, 73, 82-93, 1978. 

An IGBA Base Builder and 
Information System* 

Shu-Zhong Li and Felix Chayes 

The FORTRAN-IV program for data- 
base construction proposed last year (Year 
Book 81, 316-319) for the base being de- 
veloped by International Geological Cor- 
relation Project 163 (IGBA), is now in 
routine operation locally; if reasonably 
stable provision for completion and main- 
tenance of the base can be made, the en- 
tire system will be prepared for general 
distribution. In the currently operating- 
local version, main programs that re- 
ceive instructions interactively are used 
to build or update the base from IGBA 
card-image data files that follow the vo- 
cabulary and grammatical conventions 
described in project circular IGB832 
(Chayes, 1983). A repertoire of service 
programs facilitates examination and 
testing of the base and its associated in- 
dex and reference tables. Organization 
and information content of the base have 
been described in a recent publication (Li 
and Chayes, 1983), and provisional copy 
of a reference manual is in hand. 

In principle, a petrologist familiar with 
the structure of the base or prepared to 
study the manual in sufficient detail could 
access the experimental version of the 
base directly with his or her own pro- 
grams. A set of search routines, how- 
ever, provides an interface that makes 
such familiarity unnecessary. Three of 
these subroutines return identification 
numbers of specimen descriptions satis- 
fying specified requirements of location, 
rock name, or stratigraphic age. Given 
the identification number, which can also 
be obtained directly from an external ta- 
ble, a fourth brings the entire specimen 

*Research supported in part bv National Science 
Foundation grant EAR 78-0822L 



description from disk into memory. Ad- 
ditional subroutines then return (in var- 
iables accessible to the user program) 
amounts of essential oxides or trace ele- 
ments, stratigraphic or physical age, 
mineral assemblage, petrographic de- 
scriptors, detailed location, modal anal- 
ysis, status indicators, or other nonblank 
information field of the specimen descrip- 
tion. From a reference number encoun- 
tered anywhere in the description, 
another subroutine returns the full bib- 
liographic citation. Only as many of these 
search routines are called by a particular 
applications program as are needed, and 
the calling sequences of all are brief. Full 
documentation is in preparation. 

Ultimately, a query procedure will be 
added so that the user can specify his 
requests interactively in something very 
close to natural language, but at the 
present time the search routines recog- 
nize only symbolic or numerical varia- 


Chayes, F., Grammar, syntax and vocabulary of 
the IGBA card-image file, Circ. IGB832, IGCP 
Proj. 163, 1983 (copies available from the project 

Li, S.-Z., and F. Chayes, A prototype data base 
for IGCP Project 163— IGBA, Comput. Geosci., 
in press, 1983. 

co-occurrence of mode, mineral 

Assemblage, and Bulk Chemical 

Composition in Published 

Petrographic Descriptions* 

Felix Chayes 

Since the development of the petro- 
graphic microscope, the common nomen- 
clature and more widely used 
classifications of plutonic rocks have been 
dominantly mineralogical. In these sys- 
tems, the amount of information re- 
quired for taxonomic purposes varies from 
mere characterization of the assemblage, 

*Research supported in part by National Science 
Foundation grant EAR 78-08221. 

as in the early classifications of Rosen- 
busch (1898) and Zirkel (1893), to the 
quantitative mode or mineral ratios nec- 
essary for proper application of the Jo- 
hannsen (1931) and Shand (1943) systems 
and of the proposals of the current Sub- 
commission on Petrological Classification 
of the International Union of Geological 
Sciences (Streckeisen, 1976). As is ac- 
knowledged and provided for in the Shand 
system but not in that of Johannsen, a 
classification based strictly on quantita- 
tive modal analysis is unsuitable for vol- 
canic rocks; in fact, even one calling only 
for mineral abundance ranking will often 
be little better. For volcanic rocks the 
principal taxonomic criteria must be 
chemical rather than mineralogical. 

If a single system is to be used for both 
types of rock, or if there is to be effective 
translation from the conventions pre- 
scribed in a plutonic classification to those 
prescribed in a volcanic one, there will 
have to be a standard procedure for re- 
casting bulk chemical analyses into com- 
ponents of mineral-like composition. 
Probably there should also be a comple- 
mentary set of standard conventions by 
which a mode could be processed to yield 
an estimate of bulk chemical composi- 

In recent years there has been consid- 
erable discussion of this matter in spe- 
cialized journals, monographs, and 
committee circulars. In reviewing this 
material it is difficult to escape the notion 
that one is tilling soil already well 
ploughed. In fact, however, the technical 
literature contains remarkably little firm 
information about the interrelations of 
bulk chemical and modal composition in 
igneous rocks. This rather unanticipated 
shortage is brought into sharp relief by 
results obtained in pilot tests of the data 
base being developed by IGCP Project 
163, and in work aimed at devising effi- 
cient search procedures for use with that 

A scan of the base will provide an im- 
mediate estimate of the amount of infor- 
mation available for study of the 
interrelation between modal and bulk 



chemical composition. Rather, it would 
do so if the base itself were drawn from 
either an exhaustive or a suitably ran- 
dom sample of the petrographic litera- 
ture. Currently it satisfies neither 
condition, but the scarcity in it of spec- 
imen descriptions containing modal anal- 
yses is so extreme as to suggest that, in 
actual practice, modal composition has 
had little to do with the naming or clas- 
sification of rocks for which bulk chemical 
analyses are available. As is shown in 
column A of Table 5, only about one in 
eight of the specimen descriptions in the 
current version of the base contains a 
quantitative mode. Only in the period 
1926-1940 did the incidence of modes 
consistently surpass this figure. Since 
World War II it has averaged 10.5%, but 
the postbellum average falls to 4.5% if 
the anomalous value of 26.9% for 1966— 
1970 is excluded. Beginning in about 1950, 
there was a strong shift of emphasis in 
published reports from plutonic to vol- 
canic rocks, and this is perhaps partly 
responsible for the apparent decline. Even 

TABLE 5. Relative Incidence of Modes and 

Mineral Assemblages of Chemically Analyzed 

Specimens in IGBA Base* 

Publication Date 












































































at its maximum, however, the rate of 
generation of this kind of information was 
low, and the accumulated stock is far from 
sufficient for the task at hand. Clearly, 
most of the labor both of establishing sys- 
tematic knowledge of the relation be- 
tween modal and bulk chemical 
composition and of persuading the 
profession to use it remains to be done. 
With regard to the broader question 
of relations between mineral assem- 
blage, petrographic properties, and bulk 
chemical composition, the situation is 
considerably better, though probably still 
less favorable than would have been an- 
ticipated by most petrologists, the writer 
included. The relative frequencies, in the 
base, of mineralogical assemblage and 
petrographic descriptors specific to the 
che?nically analyzed material are shown 
in columns B and C of Table 5. An overall 
summary is given in Table 6, from which 
it is evident that about 40% of the de- 
scriptions in the base provide this kind 
of information. The association of "spec- 
ificity" in mineral assemblage with the 
occurrence of petrographic descriptors, 
however, is very high, a result implying 
that two distinct kinds of published rock 
description may be accompanied by bulk 
chemical analysis: in one, the analyzed 
specimen is itself of some intrinsic inter- 
est; in the other, either petrography and 
mineralogy are of no concern whatever 
or interest attaches to the specimen solely 
because it provides information about 
some broader category, the geological 
formation, unit, or complex. Either type 
of description may be useful in some con- 
texts and useless in others. The discrim- 

TABLE 6. Incidence of Mineral- Assemblage 
and Petrographic Descriptors in IGBA Base* 

*A^, number of specimen descriptions in IGBA 
base, May 1983; A, percentage containing modal 
analysis of chemically analyzed sample; B, per- 
centage containing mineral assemblage specific to 
chemically analyzed sample; C, percentage con- 
taining mineral assemblage and petrographic de- 
scriptors specific to chemically analyzed sample. As 
of May 1983. 














* Specific (M, P) and absent or not specific {m, 
p) to chemically analyzed sample, in specimen de- 
scriptions published in the period 1916-1980 and 
contained in Mav 1983 version of IGBA base. 



inating user of a heterogeneous 
retrospective data base will obviously re- 
quire some means of restricting his re- 
trieval to specimen descriptions of 
appropriate type. 


Johannsen, A., A Descriptive Petrography of the 
Igneous Rocks, Vol. 1, University of Chicago 
Press, Chicago, Illinois, pp. 140-157, 1931. 

Rosenbusch, H., Elemente der Gesteinslehre, 
Schweizerbart'sche Verlagshandlung, Stuttgart, 
pp. 65-68, 1898. 

Shand, S.J., Eruptive Rocks, 2d ed., Murby, Lon- 
don, pp. 186-207, 1943. 

Streckeisen, A., To each plutonic rock its proper 
name, Earth Sci. Rev., 12, 1-33, 1976. 

Zirkel, F., Lehrbuch derPetrographie, 2d ed., Vol. 
1, Engelman, Leipzig, pp. 829-842, 1893. 

Data Verification Procedures in 

Felix Chayes, Shu-Zhong Li, and Dion Steivarti 

In the day of the punched card, data 
"verification" was a simple, unambiguous 
concept. Each card contained only and 
exactly the characters shown on some 
line of the coding form. A procedure by 
which, in principle, all departures from 
1:1 correspondence between cards and 
final hard copy could be systematically 
detected and eliminated was known as 
verification, and card decks that had been 
subjected to this procedure were spoken 
of as verified. 

Organization of the data-capture op- 
eration of International Geological Cor- 
relation Project 163 (IGBA) precludes this 
kind of verification, for there is no 1:1 
correspondence between final hard copy 
and first machine-readable form. In fact, 
the coding form may contain no punc- 
tuation whatever, and many of the sym- 
bols that are shown in column sequence 
on it appear in row sequence on the card 

*Research supported in part by National Science 
Foundation grant EAR 78-08221. 

tDepartment of Geology, Bowling Green State 
University, Bowling Green, Ohio. 

image. Yet there is still the same re- 
quirement for assurance that data "qual- 
ity" has been maintained in the transfer 
operation. During the report year, two 
different verification procedures aimed 
at providing or reinforcing this assurance 
have been under development. 

In the first, which is now in routine 
operation, the card-image data file is 
scanned by a processor designed to de- 
tect and report departures from the cur- 
rent grammar and vocabulary of IGBA, 
described in project circular IGB832. In 
the output of the processor a one-line 
message informs the operator of the lo- 
cation and nature of each error and pro- 
vides a verbatim copy of it; the operator 
uses this output as an aid in proofing the 
file, corrects the proofed file with a stan- 
dard text editor, and resubmits it, con- 
tinuing the process until no further errors 
are detected. 

The procedure is spoken of within the 
project as "formal verification" because, 
although it checks all grammar and vo- 
cabulary, except when errors are de- 
tected and corrected it provides no 
systematic assurance of agreement be- 
tween coding form and card image. The 
only tests made of most numerical fields, 
for instance, are for limiting values and 
for the presence of characters that are 
not digits; a digital numerical variable 
that passes both tests is not necessarily 
the same as the corresponding number 
on the coding form. In the same way, the 
fact that a literal symbol is accepted by 
the processor establishes only that it is 
one of the 857 that are legal, not that it 
is indeed the symbol indicated on the cod- 
ing form. For the bulk of the information 
involved in the transfer, assurance of 
agreement between card image and cod- 
ing form thus still depends on ordinary, 
nonmachine proofing. 

The second procedure is essentially a 
test of the reasonableness of the speci- 
men description. In it, various parts of 
each description (e.g., the amounts of 
certain components, the rock name, the 
mineral assemblage) are compared, within 
and across categories, and unlikely or im- 



possible associations are nagged, alert- Inversion of Variance Relations 

ing the operator to the need for by the Logarithmic 

reexamination of the coding form and, Transformation* 

possibly, the source reference. „ ,. n , 

t -l A .c j. i i-i i • Feti* Chayes 
Like the first procedure, the second is 

not machine verification in the conven- The logarithmic plot is a convenient 
tional sense of the term, for it will re- graphical device for linearizing relation- 
quire direct comparison of coding form ships and for displaying compositional 
and card image only when inconsisten- variables of widely differing means in the 
cies and incompatibilities well enough same diagram. In the numerical analysis 
known to be included in the test battery of data the logarithmic transformation is 
of the processor are detected in the data, sometimes used as a device for homog- 
And often there will be no prior assur- enizing variances or normalizing varia- 
ance that the occurrence of an inconsis- bles, and it has recently been argued 
tency indicates error in either the card (Aitchison, 1981) that a ratio of logs may 
image or the coding form. As with the provide an escape from the closure effect 
formal verifier, verification of the bulk of that makes difficult or impossible the se- 
the information involved in the transfer lection of a null value against which to 
will still require proofing of just the kind test the significance of observed corre- 
ordinarily used in preparing typeset ma- lations between proportions. It has also 
terial for publication. been pointed out (Link and Koch, 1975) 
Although neither procedure will re- that the unwarranted assumption of log- 
place conventional proofing of machine- normality may sometimes introduce se- 
readable project data, each will greatly rious estimation biases. It does not seem 
assist the proofreader properly prepared to have been noted previously in the pe- 
to exploit it. Efficient use of the first will trological literature, however, that sam- 
require thorough familiarity with system pie descriptions of variance relations 
grammar and vocabulary. Efficient use obtained from transformed and raw data 
of the second will require a working may be sharply contradictory, 
knowledge of igneous petrology. From the definition of a logarithm it 
It is planned to incorporate some of the is obvious that log a < log b if, and only 
more critical tests used in this second if, a < b, so that although the relative 
procedure in an improved interactive sizes of the elements of a data array may 
transfer module that will also generate differ vastly as a result of the transfor- 
all system punctuation internally. (More mation, the ranks of their sizes are un- 
than half the formal errors in copy gen- affected by it. Because within-variable 
erated by the current transfer module ranks are unaffected by the transfor- 
concern subfield grammar, which the op- mation, it also follows that the rank cor- 
erator is now required to provide via the relation between any set of paired logged 
console.) Although the resulting initial variables is exactly the same as that be- 
copy will be much cleaner, it is still an- tween the raw values. Tests of real and 
ticipated that thorough conventional simulated data indicate that product-mo- 
proofing and editing followed by ex post ment correlations also are commonly very 
facto formal verification will precede in- similar. If the effects under considera- 
corporation of data files in the base. tion can be usefully described in terms 

of ranks and correlations, the discussion 
can probably be carried on as well with 
logs as with raw values, and there may 

*Research supported in part by National Science 
Foundation grant EAR 78-08221. 

Array * 
















































indeed be circumstances in which the logs TABLE 7. Effect of Log Transformation on 

are Dreferable Selected Variance Ranks in Some Harker Arrays 

In much practical work, however, the 
test parameter of initial concern is nei- 
ther mean nor correlation but variance. 
When one speaks of the variation "ac- 
counted for" or "explained" by a regres- 
sion relation, for instance, one is usually 
referring to the proportion of the vari- 
ance of one variable that can be described 
as a function of another, or of a group of *A, raw data, all sets, individually; B, trans- 
others. Similarly, in the commoner forms formed data, Taupo (Steiner, 1958); C, transformed 
Of variance analysis, one partitions var- ^ata, Crater Lake (Williams, 1942); D, transformed 
\ ' K , , , , data, Lassen Park (Williams, 1932); E, transformed 

lance, not covanance, and the test en- data> Medicme Lake (Anderson, 1941); F, trans- 

terion is a ratio of sums of squares, not formed data, Tristan (Dunne, 1941; Baker et al., 

sums of cross products. Again, eigen- 1964); G, transformed data, all sets, pooled. 

values tell what proportion of the total ^Number of specimens. 
variance of the original array finds its 

way into the scores of each of the eigen- displays variance ranks for variables 
vectors, or principal components; these making the largest and smallest contri- 
scores, however, provide information only butions to the total variances of five well- 
about variance, for they are by definition known data sets. 

uncorrelated and their means need bear Each set has been the subject of at 
no simple relation to the means of the least one published Harker diagram, and 
raw variables. As a further example, the in the raw data for each the variance of 
discriminant function is so chosen as to Si0 2 , as is characteristic of such arrays, 
maximize the ratio of between-group to is at least as large as the sum of all other 
within-group variation, the measure of variances. That is to say, in each array 
dispersion again being a ratio of sums of Si0 2 alone "accounts for" or "explains" 
squares or mean squares. These reduc- (more properly, contains) at least 50%, 
tion procedures are perhaps still not as and sometimes very much more, of the 
widely used by geologists as they could total observed variance, a quantity to 
be, but each mirrors and attempts to for- which P 2 5 and MnO contribute negli- 
malize a pattern of thinking about data gibly. After the transformation, how- 
that is common in the earth sciences. ever, the major contributor to the total 
It is thus of some importance that the variance in each array is either MnO or 
effect of the logarithmic transformation P 2 5 , and the variance of Si0 2 is without 
on variances is very different from its exception smaller than that of either. The 
effect on means and correlations. The situation, of course, becomes more corn- 
variances are indeed homogenized, in the plex as other variables are added. In all 
sense that differences between them are arrangements so far treated, however, 
greatly reduced. But that is far from the Si0 2 is the minimum contributor, and MnO 
whole story, for the ranks of variances and P 2 5 are either dominant or major 
and co variances, unlike those of means contributors, to the variance of the log- 
and correlations, do not necessarily sur- transformed array, 
vive the transformation. Indeed, ifinpe- The way in which variance reversals 
trologic compositional data the means of are generated is sufficiently obvious as 
two variables differ sufficiently, the rank to require only passing comment. In the 
of their variances is very likely to be in- same set of objects (rocks, minerals), some 
verted by the transformation, the last components may vary over several or- 
becoming first and the first last. Some ders of magnitude but others are always 
examples are shown in Table 7, which present in considerably more than trace 


amounts. Constituents always present in reductions with Ts rather than X's. The 
amounts of more than 1% cannot vary by product-moment correlations are again 
as much as two orders of magnitude, and unaffected by this transformation, but 
those whose minima are more than 10% variance ranks may often be inverted, 
cannot vary by as much as one. The char- the variances of trace and some minor 
acteristics of the logs, the ranges of which elements ordinarily tending toward un- 
essentially determine the sample vari- ity, those of major elements commonly 
ances after transformation, may thus vary being closer to zero than to one. 
much more widely for the trace and mi- How, then, and with what empirical 
nor constituents than for the majors. Ac- justification, should the multivariate re- 
cordingly, the results of reductions that duction of petrographic compositional data 
are essentially descriptions or reparti- be undertaken? If the raw data are used, 
tionsofthe total variance of an array may it is almost inevitable that the influence 
differ dramatically depending on whether of some of the variables familiarly known 
they are carried out on raw or trans- as "essential" or "major" will greatly out- 
formed data. weigh that of elements known as "minor" 
In the first principal component cal- or "trace." The latter may be much more 
culated from the data of a typical Harker influential than the former, however, if 
array, for instance, the coefficient for Si0 2 a transformation that inverts variance 
will be very large, for only in this way ranks is employed. Further, unless there 
can it "account for" or describe the bulk are powerful cross correlations it does 
of the total variance; coefficients for MnO not seem likely that information about 
and P 2 5 , on the other hand, will be very relations between variables of major (mi- 
small — so small that in factor analysis nor) variance will contribute much to an 
they may often be dropped, for neither understanding of relations between those 
variable contributes materially to the to- of minor (major) variance, 
tal variance of the raw array. In the first Until recently, the problem raised by 
principal component calculated from the data inhomogeneity of the magnitude 
log-transformed data, however, the rel- discussed here has been essentially ac- 
ative sizes of these coefficients will or- ademic, partly because of the paucity of 
dinarily be reversed, for very little of the trace-element data and partly becuse 
sample variance will then be contributed multivariate reductions of petrographic 
by Si0 2 and a substantial part of it will compositional data of any kind were so 
be in MnO and P 2 5 ; a principal compo- time-consuming as to be rarely at- 
nent that did not contain large coeffi- tempted. Trace-element data are no 
cients for MnO and P 2 5 could rarely be longer scarce, and with modern compu- 
the first principal component of the log- tation facilities the amount of human time 
transformed array. and effort required by multivariate re- 
Other commonly used transformations ductions, even those of gr