FUNCTIONAL ROLES OF MICROGLIA IN NEURAL TISSUE TRANSPLANTATION By NATHAN ADAM PENNELL A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1998 ACKNOWLEDGMENTS Although it is in the nature of graduate school for the student to feel he is on his own and independent much of the time, I realize while looking back over the last five years or so that a great many people have contributed both to my overall learning experience as well as to this finished dissertation. 1 make no pretense of acknowledging every person who made contributions to my graduate career, but I have attempted to thank most of the people without which my life would have been a lot more painful than it was. 1 want to thank my family and my mother in particular for supporting me in my career and life decisions, and for always being there for me emotionally and financially. 1 also want to thank all the wonderful friends 1 have made over the last five years, too numerous to name, who helped me keep my sanity and made being here fun. I would like to thank the members of my graduate committee for all their time and effort molding me into a neuroscientist. Sue Semple-Rowland was always there with helpful advice and frequent kicks in the backside to get me motivated. She also taught me a great deal about science in general and molecular biology in particular. Dr. Joel Schiffenbauer was most helpful in all aspects of this dissertation related to immunology, and generously allowed me to invade his lab to perform mixed lymphocyte reactions. Doug Anderson was an invaluable resource in regards to all aspects of spinal cord injury and transplantation, both by providing me access to his laboratory and expertise, and with his enthusiasm and insights into my projects. I would like to thank Marieta Heaton for allowing me to use her digital imaging equipment. Minnie Ann Smith and Bunny Powell for all their help with histology, Sasha Rabchevsky, who taught me everything I know about cell culture, and Dan Theele for many great conversations that helped to shape my understanding of graft rejection. I thank Sharon Walter for always being around to help, and generally making lab a fun place to be. 1 also want to thank Sean Hurley for taking me under his wing and showing me the ropes around the Streit lab. For financial assistance I would like to thank the Paralyzed Veterans of America, the State of Florida Brain and Spinal Cord Injury Rehabilitation Trust Fund, and the National Institute of Mental Heath for supporting me tlirough a National Research Service Award. Finally, special thanks go to my friend and mentor, Jake Streit. He has been a truly wonderful teacher and a mentor in the best sense of the word. 1 want to thank him for always being there when I needed to pester him and always being willing to drop what he was doing and look at slides. He let me run with my projects in whatever direction 1 thought best and pulled me back on track when they weren't working out, and for this experience I will be forever grateful. TABLE OF CONTENTS page ACKNOWLEDGMENTS ii ABSTRACT vii CHAPTERS L INTRODUCTION 1 Neural Tissue Transplantation 1 Glial Cell Transplantation 2 Grafting of Microglia/Macrophages 4 Microglia as Immunocompetent Cells 5 The Immunology of Transplantation to the CNS 8 Immune Privilege Within the CNS 8 Graft Rejection 9 Angiogenesis 13 Angiogenic Cytokines (VEGF, bFGF, TGF-P) 15 Angiogenic Properties of Microglia 17 2. THE TRACING OF FLUORO-GOLD PRELABELED MICROGLIA INJECTED INTO THE ADULT RAT BRAIN 20 Introduction 20 Materials and Methods 21 Preparation of Purified Microglial Cultures 21 Labeling and Transplantation of Microglia 22 Tissue Processing 23 Results 23 Discussion 25 3. SINGLE-CELL SUSPENSION GRAFTS OF FETAL CNS TISSUE UNDERGO DELAYED REJECTION WHICH IS NOT ALTERED BY DEPLETION OF MICROGLIA 34 Introduction 34 Materials and Methods 36 Preparation of Whole and Microgha-Depleted Fetal CNS Suspensions 36 Grafting Procedures 37 Tissue Processing 38 Antigen Presentation Assays 39 Results 41 Fetal Cell Suspension and Microglial Depletion 41 Analysis of Intrastriatal Xenografts 44 Intraspinal Allografts 45 Antigen Presentation Assays 53 Discussion 56 Xenograft Rejection 57 Allograft Rejection 58 Microglia as Antigen Presenting Cells 60 Conclusions 61 COLONIZATION OF NEURAL ALLOGRAFTS BY HOST MICROGLIAL CELLS: RELATIONSHIP TO GRAFT NEOVASCULARIZATION 63 Introduction 63 Materials and Methods 65 Preparation of Fetal Tissue Suspension 65 Depletion of Microglia and Endothelia from Fetal CNS Suspensions 65 Intrastriatal Grafting Procedure 66 Tissue Processing and Lectin Histochemistry 67 Image Analysis 67 Results 68 Depletion of Microglia and Endothelia from Fetal CNS Suspensions 68 Morphometric Analysis of Graft Vascularization 68 Development of Microglia and Vasculature Within Striatal Suspension Grafts 69 Discussion 81 Microglia and Neovascularization 81 The Origin of Microglia and Vasculature Within Suspension Grafts 83 Conclusions 84 GLIAL AND VASCULAR DEVELOPMENT WITHIN INTRASPINAL SYNGRAFTS: PRODUCTION OF ANGIOGENIC CYTOKINES 86 Introduction 86 Materials and Methods 88 Surgery and Transplantation 88 Tissue Processing 88 Results 90 Microglia, Astrocytes, and Vasculature 90 Cytokine Immunostaining 92 Discussion 106 Glial and Vascular Development 107 Angiogenic and Angiostatic Cytokine Production 109 6. CONCLUSIONS 114 The Tracing of Transplanted Microglia 114 Single-cell Suspension Grafts, Microglial Depletion, and Rejection 115 Graft Neovascularization 119 LIST OF REFERENCES 124 BIOGRAPHICAL SKETCH 141 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy FUNCTIONAL ROLES OF MICROGLIA IN NEURAL TISSUE TRANSPLANTATION By Nathan A. Pennell August, 1998 Chairman: Wolfgang J. Streit Major Department: Department of Neuroscience Transplantation of fetal neural tissue is being evaluated as a treatment for neuropathological disorders ranging from Parkinson's disease to spinal cord injury. This dissertation addresses some fundamental aspects of fetal transplant development with a primary focus on microglial cells. First, fluorescently prelabeled microglia were traced for two weeks after injection into the adult rat brain. The labeled cells survived in vivo and differentiated into ramified microglia. This study showed that, unlike perivascular cells, resting endogenous microglia are not actively pino- or phagocytic. This finding sheds some doubt upon previous ideas of microglial immunocompetence. Second, to test the hypothesis that donor microglia induce alio- and xenograft rejection, fetal rat and mouse suspensions were depleted of microglia prior to transplantation into the either the injured spinal cord or the intact striatum. This treatment did not alter the pattern or time course of xenograft rejection. Both depleted and non- depleted allografts, however, survived ftDr three weeks without signs of rejection. Between three and four weeks post-transplantation (pt) rejection started and proceeded rapidly with destruction of all grafts by 45 days. Rejection did not correlate with graft major histocompatability complex (MHC) antigen expression, suggesting that minor histocompatability antigens were involved. The ineffectiveness of microglial depletion in preventing rejection, along with in vitro results showing that allogeneic microglia are not functional antigen presenting cells, indicates that microglia do not initiate allograft rejection. Finally, the development of microglia, astrocytes, and vasculature within fetal intrastriatal and intraspinal transplants was followed for the first month pt, as was the expression of angiogenic cytokines within the grafts. Microglia, of primarily host origin, invaded the grafts prior to the onset of neovascularization and were often associated with developing vessels. Astrocytes and blood vessels were intimately associated with each other during the early time points p/. The grafts expressed high levels of vascular endothelial growth factor (VEGF) throughout the first month pt, while basic fibroblast growth factor (bPGF) and transforming growth factor-beta (TGF-P) were primarily expressed after 2 1 days pt. These cytokines were not primarily expressed by glial cells, however, indicating that the relationship between glia and angiogenesis may be indirect. CHAPTER 1 BACKGROUND Neural Tissue Transplantation Although it was once thought that damage to the central nervous system was irreversible and irreparable, it is now understood that the CNS's lack of regenerative ability is surmountable (Richardson et al., 1980; Savio and Schwab, 1990; Davies et al., 1997). The use of fetal neural transplants is currently under investigation as a possible treatment for a wide range of neuropathies, ranging from neurodegenerative disorders such as Parkinson's Disease to traumatic disorders such as spinal cord injury (Bjorklund, 1991; Dunnett and Richards, 1990; Gage and Buzsaki, 1989; Gash and Sladek, 1988; Reier et al., 1992). In the case of Parkinson's disease, research demonstrating functional recovery in animals following fetal neural transplantation (Brundin et al., 1988; Freeman et al., 1995; Lindvall et al., 1990) has precipitated the movement of this treatment into clinical trials, with similar results (Freed et al.. 1993; Kordower et al., 1997; Lindvall , 1991). The evidence for functional recovery following transplantation into the injured spinal cord is less abundant, although there are indications that recovery of function is possible (Reier et al., 1994; Anderson et al., 1995). In light of these indications, pilot studies are currently underway to determine the beneficial effects of intraspinal transplants in humans (Falci et al., 1997; D.K. Anderson, personal communication). However, there are still gaps in our basic knowledge of neural transplant biology. For 1 instance, although a great deal of attention has been paid to the phenomenon of transplant rejection (Widner and Brundin, 1988; Nicholas and Arnason, 1989; Lawrence et al., 1990; Theele et al., 1996), the mechanisms underlying this phenomenon are as yet poorly understood. Another example is the increased focus on the development of vasculature within allografts (for review, see Horner et al., 1994), although understanding of this process is still limited. The purpose of this dissertation was to further our knowledge of functional roles of microglial cells regarding these two aspects of neural transplant development. Glial Cell Transplantation Researchers have recognized the necessity of studying the interactions between neural transplants and cellular elements within the host CNS. However, the transplantation of fetal neural tissue involves a mixture of cellular elements, including various populations of neurons as well as the complete gamut of glial cells and mesodermal cells. In order to isolate the functions of defined cellular elements, purified populations of neuronal (Freed et al., 1992; Onifer et al.. 1993; Rostaing-Rigatfieri et al., 1997) or glial elements (Zhou and Lund, 1992; Gout and Dubois-Dalco, 1993; Vignais et al., 1993; Wang et al., 1995; Xu et al., 1995; Pundt et al., 1995) have been grafted and traced. The study of glial cell transplantation has been of particular interest. One of the deleterious events which occurs following traumatic spinal cord injury, as well as in multiple sclerosis, is the loss of myelin in otherwise intact white matter tracts (Bunge et al., 1993; Waxman et al., 1994; Olby and Blakemore, 1996). Grafted populations of either oligodendrocytes or Schwann cells are able to remyelinate, to some extent, regions of myelin loss. Transplanted oligodendrocytes can migrate considerable distances through the spinal cord towards demyelinating lesions, and can form new myelin sheaths (Rosenbluth et al., 1990; Gout and Dubois-Dalco, 1993), even when the cells are derived from a different species (Crang and Blakemore, 1991). Schwann cells are one of the glial cell populations which show the most promise in transplantation paradigms. These cells have been shown to remyelinate areas of demyelination following X-ray treatment (Blakemore et al., 1987). More importantly, Schwann cells have been shown to promote CNS neurite growth and regeneration both in vitro and in vivo (for review, see Guenard et al., 1993), a finding in keeping with their known role in peripheral nerve regeneration. Another glial cell which has received a great deal of attention in transplantation experiments is the astrocyte. Although the astrocytic scar may be a barrier to regeneration (Reier et al., 1983; Reier et al., 1988), astrocytes may also play a beneficial role. In vitro studies have shown that astrocytes can create a permissive environment for axonal elongation (Noble et al., 1984; Fallon, 1985; Assouline et al., 1987; Ard et al., 1991) and that transplanted astrocytes can reduce scarring and increase neurite ingrowth into gelfoam implants (Wang et al., 1995). Co-grafting of striatal astrocytes with fetal mesencephalic neural tissue has been shown to have neurotrophic and neurotropic effects on dopaminergic neurons in vivo (Pierret et al.. 1998), and astrocytic grafts have also been shown to reduce memory deficits caused by experimental lesioning of the rat septal regions (Bradbury et al., 1995). These investigations into the use of glial cell grafts have broadened our knowledge of the possible effects these cell populations might have within fetal neural transplants into the brain and spinal cord. Nevertheless, much work is still to be done before we understand the interactions between these cells and the host environment. Grafting of Microglia/Macrophages Despite representing nearly 10% of the total cells in the adult CNS, the cell type that has been most often ignored in transplantation paradigms is the microglial cell. A great deal of study has focused on transplantation of neurons and glial cells such as astrocytes, oligodendrocytes, and Schwann cells (see above), but we know next to nothing about the role of grafted microglia. Although several studies of graft rejection have reported microglia/macrophage infiltration (Lawrence et al.. 1990; Finsen et al., 1991 ; Duan et al., 1995; Poltorak and Freed, 1989), none have addressed a possible role for these cells beyond phagocytosis of debris and the speculation that they might be antigen presenting cells (APCs). There has, however, been a recent interest in the use of microglia/macrophages as tools to promote regeneration. Polymeric tubes filled with purified microglia, when placed into the acutely injured spinal cord, were shown have increased neuritic ingrow1;h over gelfoam alone or astrocytic grafts. Even neuritic ingrowth into tubes devoid of grafted microglia was associated with host macrophage infiltration (Rabchevsky and Streit, 1997). Microglia/macrophages transplanted into the injured spinal cord, in conjunction with fetal neural transplants, have been shown to increase the regeneration of DRG sensory fibers (Prewitt et al., 1997) Macrophages alone placed into the injured spinal cord had a similar effect that was attributed to their degradation of myelin debris and the promotion of neovascularization (Franzen et al., 1998). A similar study used injections of activated peripheral macrophages to overcome the inherent non-permissive nature of the optic nerve for regeneration (Lazarov-Speigler et al., 1996). The role of macrophages in peripheral nerve regeneration has been known for some time (for review, see Griffin et al., 1993). However, until recently these attributes had not been tested within the CNS. Despite indications that grafted microglia/macrophages may potentially be beneficial to the injured CNS, however, there remained no clearly defined method for the reliable tracing of grafted cells. Chapter 2 details such a method, which should be of benefit to future studies regarding fijnctional roles of transplanted microglial cells. There are many benefits to working with transplants of purified microglial cells. However, to truly understand how these cells interact with fetal neural transplants as well as the host environment, studies must be designed to look at microglia within complete fetal grafts. As described in the following paragraphs, there is ample evidence to suggest beneficial roles for microglia in graft development as well as a detrimental role in the immunology of graft rejection. The research in this dissertation has attempted to delineate some of these functions. Microglia as Immunocompetent Cells Although astroglial cells have been shown to express properties similar to antigen presenting cells (APCs) in vitro (Fontana et al., 1984; Wong et al., 1985), the preponderance of evidence has led to the current belief that microglia are the resident immunocompetent cells of the CNS (Frei et al., 1987; Hickey and Kimura, 1988; Poltorak and Freed, 1989; Lawrence et al., 1990; Graeber and Streit, 1990). Numerous immunohistochemical studies, conducted independently by various laboratories, have examined the expression of antigens of the Major Histocompatability Complex (MHC) in both normal and injured CNS tissue. These studies have shown conclusively that the principal cell type in the CNS expressing MHC antigens is the microglial cell (Matsumoto et al., 1986; McGeer et al., 1988; Konno et al., 1989; Streit et al., 1989; Morioka et al., 1992a; Popovich et al., 1993). These immunophenotypic studies on tissue sections have been confirmed by flow cytometric characterization of microglial cells from nonnal and diseased CNS tissue (Sedgwick et al., 1991). Furthermore, continuous systemic administration of interferon-gamma (IFN-y), a potent MHC inducer, results in widespread MHC expression specifically on microglia and not astrocytes (Steiniger et al., 1988). This provides additional support for microglial immune competence. Similarly, upon treatment with IFN-y, cultured microglia have been found to express MHC class II antigen and present anfigen to MHC class II restricted T-cells (Frei et al., 1987). An important finding in understanding microglial immunological activity concerns their secretory activity with regard to the elaboration of cytokines and growth factors. This work, which was largely derived from studies on microglia in tissue culture, and more recently from in vivo invesfigafions (Kiefer et al., 1993), has provided evidence for a second critical component for defining microglial immunocompetence, namely the production and secretion of substances such as interleukin- 1 , interleukin-6. tumor necrosis factor-alpha (TNF-a), and transforming growth factor-beta (TGF-P) (Giulian et al., 1986; Remick et al., 1988; Sawada et al., 1989). Cytokines/growth factors are known to mediate a number of immunological cell-cell interactions, including anfigen presentation and cytotoxicity, as well as T lymphocyte proliferation and acfivafion. Many of the tissue culture studies on microglial cytokine producfion remain to be confirmed by in vivo localization studies. Nevertheless, they have provided an essential counterpart to studies on MHC expression by revealing that both membrane surface constituents (MHC antigens), as well as secretory components (cytokines) facilitate the immune competence of microglia. In regards to functional roles of microglia in graft rejection, this evidence indicates that grafted microglia may be a major cellular source of cytokines which would normally contribute towards activating the host immune system. Conversely, the infiltration of a developing graft by host microglial cells provides a large population of cells capable of both antigen presentation and pro-inflammatory cytokine production. New and conflicting evidence, however, paints a different picture of the state of microglial immunocompetence. Ford and colleagues (1995) have shown that microglia can be separated from other macrophages within the CNS, i.e. perivascular cells, by flow cytometry and that purified microglia do not function in vitro as antigen presenting cells. This finding was supported by another study which characterized microglia as poor APCs, similar to other tissue-specific macrophages during development and prior to full immunocompetence (Carson, 1998). This can be partly attributed to the lack of expression of appropriate costimulatory molecules, in particular B7.1 . Microglia in vitro have been shown to express B7.1 at low levels which increase upon exposure to IFN-y (Williams et al., 1994; De Simone et al., 1995), and in vivo in acute multiple sclerosis lesions (De Simone et al., 1995; Dangond et al., 1997). However, these same authors confirm that microglia in vivo do not express B7.1 in the normal, non-pathologic CNS nor following normal trauma. A more telling study has demonstrated that purified microglia actually induce apoptosis, or programmed cell death, in T-cells, rather than stimulating them to divide as would be expected of an APC (Ford et al., 1996). The experiments outlined in Chapter 3 were designed to shed new light on the state of microglia immune function by examining their role in transplant rejection. The Immunology of Transplantation to the CNS Immune Privilege Within the CNS The central nervous system has long been regarded as a likely target for transplantation because of it's perceived status as an immunologically privileged site (Medawar, 1948; Barker and Billingham, 1977). This term implies that the CNS is somehow shielded from the immune system in ways that other regions of the body are not. There are three primary observations which contributed to this conclusion. First, most cells of the CNS do not express antigens of the Major Histocompatability Complex (MHC), the primary molecules which define immunological self from non-self (Nicholas and Amason, 1989). Second, it was thought that lymphocytes did not enter the CNS, eliminating both the possibility of detection of transplanted tissue and the effector cells needed to cause rejection (Widner and Brundin, 1988). Finally, and perhaps most importantly, it was thought that the brain had no lymphatic drainage (Yoffey and Courtice, 1 970). effectively isolating the CNS from lymphatic tissue and thus the immune system. It is well known that transplants of neural and non-neural tissue grafted into the CNS survive better than they would be expected to based upon their histocompatability (Murphy and Sturm, 1923; Ridley and Cavanaugh, 1969; Mason et al., 1986; Widner et al., 1989), but it is also clear that the brain is not as well shielded from immune attack as was once thought. It has been shown that although most cells of the CNS do not express MHC antigens under normal conditions (Widner and Brundin, 1988), many are capable of MHC expression under conditions such as trauma or inflammation (Traugott et al., 1985; Steiniger et al., 1988; Streit et al., 1989). Of greater interest is a subpopulation of microglia and perivascular cells which constitutively express MHC Class II antigens, constituting a possible pool of cells capable of antigen presentation (Graeber et al, 1992). It has also been shown that T-cells can in fact enter the CNS during pathological conditions such as experimental allergic encephalomyelitis (EAE) (Wekerle et al., 1986; Cross et al., 1990). Lastly, in regards to lymphatic drainage, it is clear that there is drainage from the CNS into the deep cervical lymph nodes, and that this drainage occurs primarily through the arachnoid villi and across the cribriform plate (Bradbury and Westrop, 1983; Widner et al., 1988; Harling-Berg et al., 1989). Together, these facts indicate that the immune privilege of the CNS is a relative and not absolute quantity, and that transplants placed into the brain and spinal cord are at risk of immunological rejection. Graft Rejection According to Lawrence et al. (1990), it is possible to distinguish at least three phases of the graft rejection process: 1) The immune induction phase (also known as the afferent limb of an immune response); 2) the phase of immune attack (efferent limb of immune response); and 3) a quiescent phase. During the induction phase it is believed that expression of MHC antigens on cells from within the grafted tissue represents the primary inducing stimulus. There are two primary pathways through which this may occur: direct and indirect antigen presentation (see Fig. 1-1). In the direct pathway, thought to be the 10 primary pathway for the initiation of allograil rejection (VanBuskirk et al., 1994), MHC- expressing cells acting as donor APCs migrate out of the graft and present foreign antigen directly to host T-cells. In the indirect pathway, foreign antigen within the graft, of which MHC antigens are a large part, is phagocytosed by host APCs and then presented to host T- cells. With this in mind, the removal of all donor cells capable of MHC expression should interrupt these pathways and prevent or at least delay immune induction (Nicholas and Amason. 1989; Bartlett et al., 1990). It is important to point out, however, that minor histocompatability antigens alone are capable of inducing immunological rejection (Steinmuller. 1983; LaRosa and Talmage, 1985). Since normal embryonic CNS tissue contains no cells that constitutively express MHC antigens, it is thought that MHC expression is induced following transplantation. From the literature regarding in vivo expression of MHC antigens, it is evident that MHC II expression in the rodent CNS is limited to two cell types: microglia and perivascular cells or pericytes (Streit et al., 1989a). Endothelial cells are the other cellular component within the rodent CNS capable of MHC expression, but this expression is limited to Class I antigen (Streit et al., 1989b). We know that microglial and endothelial cells are present in the developing CNS as early as E12 (Ashwell et al., 1989) while perivascular cells do not appear before the first postnatal week (Mato et al., 1985). Together with their MHC- expressing capabilifies, this makes it very likely that transplantation-induced MHC expression occurs on just microglia and endothelial cells. The research described in Chapter 3 is designed to test the hypothesis that the removal of the cell types capable of expressing MHC antigens, i.e. microglia and endothelial cells, will delay or prevent the phase of immune induction. By correlating the Figure 1-1. The Induction of Graft Rejection. 1. The implantation procedure and/or the inflammation resulting fi-om a preexisting lesion causes a breakdown of the blood-brain barrier. 2. This trauma activates host microglia and infiltrating macrophages, which produce pro-inflammatory cytokines such as TNF-a, IL-6. and IL-1|3. 3. The cells within the graft capable of expressing MHC antigens, i.e. microglia, begin to do so. They may migrate out into peripheral lymphatic tissue and present foreign antigen directly to host T-cells. This is known as the direct pathway. 4. Foreign antigen, either shed from living cells or debris from dead cells, is taken up by host APCs and presented to host T-cells. This is the indirect pathway. 5. The activated T-cells initiate the rejection cascade. 12 Graft Rejection Cascade Eiidothelium (2) Host Microglial Activation (jj Implantation Trauma Graft Antigens ♦ mHC (4) Migration ** Antigen Presentation DIRECT PATHWAY INDIRECT PATHWAY 13 temporal expression of donor MHC Class 11 antigens with graft rejection, this study will shed flirther light on the relationship between MHC expression and graft rejection. This experiment also has clinical implications. Currently, immunosuppressive agents, in particular cyclosporin A and FK 506, are widely employed to prevent the immune response against a neural transplant. There are at least two major problems associated with the use of these agents: increased susceptibility to secondary infections and rejection of the graft upon withdrawal of the immunosuppressant. Thus, the administration of immunosuppressants during the early post-transplantation period does not provide long-lasting protection from immune attack but only delays the onset of rejection. Immunosuppression is probably the least desirable regimen for a human transplant patient who is very susceptible to numerous infections. This research will explore a potential new method of inducing tolerance of allo- and xenografted tissue without otherwise compromising the immune system of the host. Angiogenesis The survival of any adult tissue depends on the presence of a vascular network. The availability of nutrients, oxygen, and waste disposal through the blood is crucial once tissues reach a certain minimum size. This is illustrated by the lack of viability of animals deficient in factors that regulate angiogenesis (Beck and D'Amore, 1997) and in tumors in which angiogenesis has been inhibited (Folkman, 1990). In addition, the efficacy of the immune system depends on the ability of bloodborne leukocytes to reach their targets by way of the vasculature, as well as on the ability of endothelial cells to express the necessary adhesion molecules for leukocyte binding and migration (Abbas et al., 1994). For these reasons, the development of new blood vessels within developing neural 14 transplants is an important topic, both for its role in graft survival as well as for its contribution to graft rejection. Blood vessel formation during development is carried out by a combination of two different processes: vasculogenesis and angiogenesis. Vasculogenesis is a process in which endothelial cell precursors, termed angioblasts, migrate into developing tissues and form vessel tubes. Most endodermal tissues are vascularized by vasculogenesis. The development of vasculature within the central nervous system, as with most tissues of ectodermal origin, is carried out by angiogenesis, or the formation of vessels by budding off of already existing vessels (Beck and D'Amore, 1997). One important exception to this rule is the developing retina, which is thought to be vascularized through vasculogenesis despite being a part of the CNS (McLeod et al.. 1987; Beck and D'Amore, 1997). Angiogenesis is also the process by which healing wounds, growing tumors, and developing neural transplants form new blood vessels (Iruela-Arispe and Dvorak, 1997; Homer et al., 1994). The common role of angiogenesis in the formation of new vessels both in neural transplants and in the developing CNS illustrates the usefulness of fetal transplantation as a model for recapitulating development, in addition to its potential therapeutic value. There are several distinct processes involved in angiogenesis The first is the degradation of extracellular matrix by released proteases. The next steps are the induction of endothelial cell proliferation and the migration of the cells towards the angiogenic stimulus. Following migration, the cells differentiate and form capillary tubes, a new basement membrane, and recruit pericytes to the perivascular space (Potgens et al., 1995). All of these processes are controlled through various angiogenic factors and cytokines. 15 Three major cytokines known to be involved in angiogenesis are examined in this dissertation; vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), and transforming growth factor-beta (TGF-P). Angiogenic Cytokines (VEGF. bFGF. and TGF-P) Vascular endothelial growth factor (VEGF; also known as Vascular Permeability Factor) is expressed as a 34-46 kDa homodimeric glycoprotein. Alternative splicing of the VEGF mRNA gives rise to four splice variants of 121, 165, 189, and 206 amino acids. However, the 206 and 189 aa variants are rare, and remain bound to the expressing cells by heparin-like molecules or bound to heparin in the ECM. being released only by plasmin or heparinase cleavage. These variants have little or no mitogenic activity on cultured endothelial cells (Houck et al., 1991), and their biological significance is not known. The more common 165 and 121 aa variants have little and no affinity, respectively, for heparin and are freely secreted by the producing cells (Potgens et al., 1995). Binding of VEGF is restricted to the high and low affinity VEGF receptors {fit and flk-1), which are primarily restricted to endothelial cells (Ferrara et al., 1992; Potgens et al., 1995). VEGF is potently mitogenic exclusively for vascular endothelial cells, in contrast to bFGF which also induces the proliferation of a wide range of additional cells. VEGF is, by itself, capable of initiating all of the events necessary for blood vessel formation (see above; Ferrara et al., 1992). VEGF production by astrocytes is thought to be responsible for vascular development in the retina (Stone et al., 1995) as well as lesion-induced neovascularization following brain injury (Papavassiliou et al., 1997). 16 Basic fibroblast growth factor (bFGF) is expressed as a 16-24 kDa protein characterized by its strong ability to bind to heparin. bFGF also exists as several different splice variants, although the 155 aa variant predominates. All the variants have similar or identical biological properties. bFGF is a pleiotropic molecule which induces proliferation of a number a cells, from astrocytes to smooth muscle cells to fibroblasts. More to the point, it affects proliferation, migration, and differentiation of vascular endothelial cells and is potently angiogenic in vitro and m vivo (Potgens et al., 1995). Interestingly, it also induces neuronal differentiation (Vlodavsky et al., 1991). bFGF also acts through high and low affinity receptors (Jig and bek). Like the VEGF receptors, these are both transmembrane receptor tyrosine kinases. It is thought that bFGF can only act through simultaneous binding of both the high and low affinity receptors (Potgens et al., 1995). In in vitro models of angiogenesis, VEGF and bFGF act synergistically, with much stronger effects than either has alone (Pepper et al., 1992). Both induce expression of proteases required for ECM degradation (i.e. collagenase and constituents of the plasminogen activator pathway) as well as induce migration and capillary tube formation in collagen gels (Potgens et al., 1995). However, despite bFGF's unquestioned potency in inducing angiogenesis in model systems, its role in normal angiogenesis in vivo is a matter of some dispute (Horner et al., 1994; Beck and D'Amore, 1997). Transforming Growth Factor - beta (TGF-P), in active form, is a disulfide-linked homodimer with two chains of 1 12 amino acids. There are five isoforms of TGF-p, three of which (TGF-P 1-3) are expressed in mammals, and all of which are 65-80% 17 homologous with the same basic structure (Flanders et al., 1998). TGF-p is normally secreted in an inactive form, bound to a so-called latency associated protein (LAP), and the in vivo method of activation is unclear. TGF-P binds to a heteromeric receptor complex consisting of two subunits termed Rl and Rll, which are serine/threonine kinases. TGF-P also binds with high affinity to the type 111 receptor, a high molecular weight protein with no signaling domain that is thought to aid in the binding of TGF-P to Rll (Flanders et al.. 1998). TGF-P has diverse functions ranging from immunosuppression to up-regulation of NGF production by astrocytes (Horner et al., 1994). It also is involved in the inhibition of endothelial cell proliferation, the deposition of matrix, and in vascular differentiation (Horner et al., 1994; Beck and D'Amore, 1997; Flanders et al., 1998). TGF-pi is thought to be expressed by microglia in the aduk rat brain, while TGF-p2-3 are expressed in neurons and astrocytes (Flanders et al., 1998). Its inhibitory effects, in contrast to the angiogenic effects of VEGF and bFGF, may make it an important player in the overall regulation of neovascularization. Angiogenic Properties of Microglia The idea that macrophages are involved in wound healing and inflammatory angiogenesis is not a new one. Leibovitch and Ross discovered in 1975 that macrophages were necessary for undisturbed wound repair. Other groups showed both a temporal correlation between endothelial DNA synthesis and mononuclear cell infiltration (Polverini et al., 1977a) and that macrophages isolated from wounds induced neovascularization in vitro and in in vivo assays (Clark et al., 1976; Greenburg and Hunt, 1978). It was soon discovered that macrophage conditioned medium alone promoted 18 vascular proliferation (Polverini, 1977b), implicating some diffusable substance. Macrophages have also been implicated in tumor angiogenesis. Mice depleted of monocytes showed decreased tumor vascularization (Evans. 1 977), and neoplastic tissue explants demonstrate angiogenic properties only when macrophages are present (Mostafa et al., 1980). We now know that activated macrophages produce a number of angiogenic cytokines, the most potent among them being VEGF, bFGF, TNF-a, TGF-(3, and IL-lp (Sunderkotter et al., 1994), although IL-ip and TNF-a likely act through indirect, pro- inflammatory pathways. Regardless of other functions attributed to them, it is well established that microglia represent the brain's tissue-specific macrophage population. Although the subject remains controversial, there is ample evidence that microglia are derived from fetal macrophages which migrate into the developing CNS at early prenatal timepoints and give rise to the adult microglial population (Hurley and Streit, 1996). Following injury, resting microglia become activated and upregulate typical macrophage cell surface molecules, such as MHC antigens. In pathological situations resulting in cell death, activation of microglia results in the formation of brain macrophages, phenotypically indistinguishable from blood derived macrophages (Graeber et al., 1988; Streit et al., 1988, 1989a). In tissue culture, microglia have been demonstrated to produce bFGF (Shimojo et al.. 1991) and TGF-P (Lindholm et al., 1990). It is not known whether microglia are also capable of VEGF production, although peripheral macrophages are known to be a source of VEGF (Sunderkotter et al., 1994). Horner and colleagues (1994) speculate that microglia play a large role in intraspinal transplant neovascularization, but 19 there has been surprisingly little work done to demonstrate this phenomenon. Chapters 4 and 5 examine the hypothesis that microglia are a principal mediator of graft neovascularization. The relationship between microglia and vasculature is examined both morphologically and from the standpoint of cytokine production. CHAPTER 2 TRACING OF FLUORO-GOLD PRELABELED MICROGLIA INJECTED INTO THE ADULT RAT BRAIN Introduction Since microglial cells were first identified as a distinct glial population by del Rio-Hortega (1932), the prevailing view of microglial function has been that they serve as macrophages of the central nervous system (CNS) (Rio-Hortega. 1932; Giulian and Baker, 1986; Kreutzberg, 1996). It is well known that injury to the CNS resuUing in neuronal death induces resting microglia to become phagocytic cells that are morphologically indistinguishable from blood-derived macrophages (Streit et al., 1988). In recent years, this traditional view of microglial function has been expanded by studies showing that microglia secrete neuronal growth factors and angiogenic cytokines in vitro (Mallat et al.. 1989; Elkabes et al., 1996; Lindholm et al., 1990; Shimojo et al., 1991). This suggests that microglia have important functions in supporting neuronal health and promoting vascular development in the normal as well as the pathological CNS (Horner et al., 1994; Streit and Kincaid-Colton, 1995; Pennell and Streit, 1997). Additional support for such beneficial activities of microglia is derived from studies using transplanted peripheral macrophages and microglial cells, indicating that both of these cell populations can enhance the limited regenerative capabilities of injured CNS neurons 20 21 (Lazarov-Speigler et al., 1996; Rabchevsky and Streit, 1997). Taken together, these findings paint a picture of microglial function that is different from their traditional role as the brain's scavenger cell and suggest that microglia perform functions critical for the repair and reconstruction of injured CNS tissue. The recent interest in transplantation of microglial cells and macrophages raises the question of a reliable method for tracing the grafted cells and distinguishing grafted microglia ftom resident microglia. One common method for the labeling of glial cells prior to transplantation is the use of fluorescent labels such as fluorescein-conjugated lectin (Goldberg and Bernstein, 1987) and vital dyes such as Dil, Hoechst 33342, and Fluoro-Gold (Rabchevsky and Streit, 1997; Xu et al., 1995; Andersson et al., 1993). In the present study, we have investigated the use of Fluoro-Gold (FG) to prelabel cultured microglial cells before injecting them into the normal rat brain. Materials and Methods Preparation of Purified Microglial Cultures Isolated whole brains (cerebellum and tectum removed) from newborn Wistar rats (Harlan Sprague-Dawley, Inc.) were stripped of meninges in a 35 mm Petri dish while immersed in "Solution D" (0.8% NaCl, 0.04% KCl, 0.1% Glucose, 2.2% Sucrose, 0.01% Streptomycin sulfate, 0.025% Fungizone, and 0.006% Penicillin G in dHjO). Clean fragments were mechanically minced, transferred to a 15 ml conical tube, and incubated under bi-directional rotation in 0.05% trypsin in "Solution D" for 20 minutes at 37°C. The suspension was then triturated, DNAase added to a final concentration of 5800 units/ml. and the suspension incubated for an additional 10 min. at 37°C. An equal volume of 22 complete medium (DMEM + 10% FBS) was added to the suspension to quench the reaction, and the tissue was triturated again and passed thi'ough a 130 \xm Nitex filter before being pelleted (400 x g, 10 min.). The pellet was then resuspended in 10 ml of complete medium and passed through a 40 i^m Nitex filter before being plated in poly-L-lysine coated 75 cnr flasks. The plating density was one brain per 1.5 flasks. Cultures were incubated in a 37°C incubator set at 8% CO^ for 3 days. The medium was then changed and the cultures incubated for an additional 7 days after which microglia were harvested every three or four days. The flasks were shaken on an orbital shaker in a 37° incubator for 1 hour at 200 rpm. After shaking, the supernatant yielded a high number of microglia which were pelleted, resuspended in complete medium, and plated on 100 mm Petri dishes. The medium was changed after one hour to remove non-adherent cells. Microglia were plated at a density of 5 x 10*" cells/ 100mm dish, and maintained in culture overnight prior to FG- labeling and injection. Labeling and Transplantation of Microglia Five ml of 0.1% Fluoro-Gold (Fluorochrome Intemafional) in DMEM + 10% FBS was added to the cultured microglia followed by incubation at 37° for 2 hours. Dishes were washed with fresh medium several times to remove excess FG. Labeled microglia were removed from the dish using a sterile rubber cell scraper, after which the cells were pelleted, resuspended in complete media, and counted prior to injections. Under ketamine/xylazine anesthesia, cells were injected stereotactically into either the right corpus callosum or the right lateral ventricle of adult male Wistar rats as a total volume of 3 |il (50,000 cells/|al in DMEM + \0% FBS), using a 25 gauge Hamilton syringe. 23 Injections were made slowly over a period of several minutes, and the needle allowed to remain in place for several minutes after injection to prevent reflux. Control rats received injections of 1 ^1 of 2% FG in either sterile H,0 or DMEM + 10% FBS into the right corpus callosum. Stereotaxic coordinates were as follows: 2 mm lateral of Bregma. 3 mm ventral to the dural surface. Tissue Processing Following survival times of 1 , 2, 7, and 14 days, rats were sacrificed by transcardiac perfusion with 100 ml of 0.9% saline followed by 200 ml of 4%) paraformaldehyde in phosphate buffered saline, pH 7.4. Brains were removed and post-fixed in 4% paraformaldehyde for at least 48 hours prior to sectioning on a vibratome at 60 |Lim thickness. Using a slightly modified procedure from previously described protocols (Streit, 1990), selected sections were incubated with biotinylated isolectin from Griffonia simplicifolia (Sigma L-2140; 10 |.ig/ml in PBS containing 0.4% Triton X-100 and 0.1 mM CaClj, MgCU, and MnCU) for 3 hours at room temperature, followed by streptavidin- Texas Red (Molecular Probes S-872; 10 |.ig/ml in PBS) for 45 min. at room temperature. Sections were mounted on gelatin-coated slides, allowed to air dry, and then either dehydrated in ascending ethanols and coverslipped with Permount (FG only) or coverslipped with fluorescent mounting media. Results Following a two hour incubation with 0.1%) FG, all microglial cells in a given culture dish were brightly labeled (Fig. 2-1 A). The cells maintained the typical 24 morphology of microglia in culture, having either a rod-like or flattened shape with few processes. Twenty-four hours after injection, the FG prelabeled cells were seen grouped together in a small cavity created by injection into the corpus callosum (Fig. 2- IB). Grafted microglia remained brightly labeled in vivo, with no apparent labeling of endogenous ramified cells. Twenty-four hours after control injections of free FG there was extensive labeling of axons within the corpus callosum, as well as retrograde labeling of neurons near the injection site. There appeared to be abundant free FG at the injection site, as evidenced by hazy staining that was not localized to any particular cell type. Importantly, there was no apparent labeling of resident microglial cells. Seven days after injection into the right lateral ventricle, bright FG-labeled cells were found within the choroid plexus and ependymal layers of both lateral ventricles. These cells extended processes (Fig. 2-lC), and were easily distinguishable against an unlabeled background. Seven days after control injection of free FG into the corpus callosum, there was no hazy staining indicative of free FG at the injection site, suggesting that most of the FG had been cleared from the extracellular space. Neurons throughout both hemispheres were labeled, and brightly labeled perivascular cells were evident throughout the injected hemisphere (Fig. 2-2). Perivascular cell labeling was brightest near the injection site and became less pronounced farther away, where it was found primarily in white matter containing FG-labeled axons (Fig. 2-2B). Brains injected with free FG in DMEM + 10% FBS also showed labeling of cells with oligodendrocyte morphology (not shown). However, this labeling was present only in the corpus callosum at the injection site. There was no detectable labeling of oligodendrocytes in the animals 25 injected with FG in HjO. Resident microglia that were visualized with fluorescent lectin staining revealed no FG labeling in these control experiments (Fig. 2-3). Two weeks post-injection FG-labeled microglia were still brightly labeled in vivo, and some had migrated out along the corpus callosum, although most of them were still clumped together at the injection site. Some of the grafted cells had also migrated into the surrounding gray matter and extended processes, resembling normal resting microglia (Fig. ID). There was moderate staining of neurons and perivascular cells in the immediate area of the injection site, indicating that these cells had taken up the FG released from prelabeled microglia that died after the injection. Discussion FG has traditionally been used as a fluorescent tracer for the mapping of axonal projections in the PNS and CNS (Schmued and Fallon, 1986; Zhou and Rush, 1995). It is retrogradely transported by axons, and has several advantages over other fluorescent dyes. The human retina has its highest sensitivity at the wavelength emitted by FG (Thanos et al., 1994), and FG does not fade appreciably over time even under frequent exposure to light. It is ideal for double-labeling studies as the dye is visible in vibratome, frozen, paraffin, and plastic sections, and the light emitted does not cross over into the wavelengths of fluorescein or rhodamine (Sclimued and Fallon, 1986). It has also been demonstrated in neuronal labeling studies that the dye does not leak from healthy, intact cells (Rinaman et al, 1991; Crews and Wigston, 1990). However, h can be expected that after transplantation a certain percentage of the cells die and release FG into the Figure 2-1. Fluoro-Gold (FG) labeling and tracing of cultured microglial cells. (A) Microglia in vitro following a 2 hour incubation with 0.1% FG. All cells in the field are labeled and display the characteristic morphology of cultured microglial cells, x 1 12. (B) FG labeled microglia 24 hours after injection into the adult rat corpus callosum. Notice that despite robust labeling of the injected cells there is very little background labeling of neurons by free FG. x 1 12. (C) FG prelabeled microglial cells are present in the ependymal layer of an adult rat one week after intraventricular injection of prelabeled cells. V = ventricular space, x 225. (D) FG prelabeled microglial cell exhibits ramified morphology two weeks after injection into the adult rat corpus callosum. This shows that cultured cells with an ameboid morphology differentiate into process-bearing cells when exposed to the CNS enviromnent. x 560. 27 ^'^^f.# iP ji-nn^^miigii^ i- :^ ,■ r^% 1 y; 1 ^i^ ki 1 A * F #^^ * • . Figure 2-1 Figure 2-2. Cellular labeling one week following injection of free FG into the adult rat corpus callosum. (A) Numerous FG labeled cells are evident, notably perivascular cells (arrows) and neurons. No cells with microglial morphology are labeled, x 125. (B) Higher power micrograph demonstrates specific labeling of perivascular cells by free FG (arrow). FG is also evident within axons passing through the corpus callosum. x 250. (C) Pyramidal neurons immediately dorsal to the injected corpus callosum exhibit bright labeling. Note that there are no non-neuronal cells labeled, demonstrating that the dye is retained within healthy neurons, x 250 29 » ♦ ^ f • t ;fXf*^.^^ Figure 2-2 Figure 2-3. Control brain injected with 2% FG in DMEM + 10% FBS, one week after injection. The section was stained with biotinylated GS I-B4and streptavidin-Texas Red to visualize resident microglial cells. (A) FG labeled neurons in the cortex near the injection site. (B) The same field shows Texas Red-labeled resident microglial cells (arrows show location of microglial cell nucleus). Note the lack of FG labeling of resident microglial cells. X 800 31 Figure 2-3 32 extracellular space. It was therefore necessary to determine which cells in the CNS are capable of taking up free FG. Based on both morphology and double labeling with GS I-B4, we have demonstrated that resting, endogenous microglia do not take up free FG, and thus do not seem to be actively pino- or phagocytic. This finding is in contrast to the known phagocytic activity of activated microglia, illustrating the functional plasticity of microglial cells and providing direct in vivo evidence that ramified microglia are functionally distinct from reactive or phagocytic microglia (Giulian and Baker, 1986; Bocchini et al., 1988; Streit at al., 1988). Ward and colleagues (1991) hypothesized that ramified microglia in the brain served the function of interstitial fluid cleansing, cleaning out excess neurotransmitters and unwanted wastes. This hypothesis was based on the selective pinocytotic activity of microglia in whole-brain cultures. Our current findings, however, argue against such a function in vivo, as resident microglia did not sequester free FG in the manner of cultured microglia. Instead it appears that perivascular cells are the predominant pinocytotic cells of the uninjured CNS, as has been reported previously (Kida et al., 1993; Streit and Graeber, 1993). In addifion, our findings suggest that oligodendrocytes may become pinocytotic under certain conditions, although their capacity for uptake of FG was much less than that of perivascular cells. Aside from leakage, a concern with the use of any diffusable substance as a cellular label is the problem of dilution following repeated division of cells. Although microglia are known to proliferate following neuronal injury (Graeber et al., 1988), they do so only minimally under normal conditions (Korr et al., 1983). Furthermore, when fluorescently prelabeled microglia respond to optic nerve transection by phagocytosing 33 dying retinal ganglion cells, they remain specifically labeled afterwards, indicating that even injury-induced proliferation is insufficient to dilute out their fluorescence (Thanos, 1991). Another concern involves the duration of FG-labeling. Although we did not follow the FG-labeled cells longer than two weeks post-transplantation, previous studies have shown that microglia labeled through phagocytosis of retrogradely FG-labeled neurons are identifiable as long as six weeks after labeling, the longest time point examined (Crews and Wigston, 1990). Studies using other fluorescent dyes have traced microglia in vivo for over twelve months (Thanos et al., 1994), demonstrating that microglia are not readily replaced in the adult brain, and suggesting that microglia labeled with a dye as persistent as FG would be traceable for extended periods of time (Schmued and Fallon, 1986). In summary, we have demonstrated that cultured microglial cells can be robustly labeled with FG prior to transplantation, that FG prelabeled microglia are readily identifiable within the host brain up to two weeks, and that some of these transplanted cells undergo differentiation into ramified microglia. We have also shown that ramified microglia in vivo do not take up free FG, emphasizing their resting nature and supporting the concept of functional plasticity. Our findings indicate that FG pre-labeling is appropriate for the tracing of transplanted microglia. CHAPTER 3 SINGLE-CELL SUSPENSION GRAFTS OF FETAL CNS TISSUE UNDERGO DELAYED REJECTION WHICH IS NOT ALTERED BY DEPLETION OF MICROGLIA Introduction Fetal neural transplants are currently being tested as a possible treatment for neurodegenerative disorders such as Parkinson's Disease as well as traumatic disorders such as spinal cord injury (Bjorklund, 1991; Durmett and Richards, 1990; Gage and Buzsaki, 1989; Gash and Sladek, 1988; Reier et al., 1992). Since the use of isogeneic neural tissue is not clinically feasible, investigators and clinicians have focused instead on allogeneic and xenogeneic donor tissue, which raises the question of histocompatability and the possibility of graft rejection (Nicholas and Amason, 1989). To date, the majority of strategies used to prevent immunological rejection involve either immunosuppression (Duan et al., 1996; Pederson et al., 1997; Nicholas and Arnason, 1989) or the alteration of the host immune system in such a way as to induce tolerance (Theele and Reier, 1996; Okura et al., 1997; Wood et al., 1996). An alternative approach would be to alter the donor tissue in some way as to make it unrecognizable to the host immune system. It has been speculated that the presence of antigen presenting cells (APCs) within the donor tissue is the primary stimulus for the onset of graft rejection, as these cells may present foreign antigen directly to host T-cells (Lawrence et al., 1990; Nicholas and Arnason, 1989; VanBuskirk et al.. 1994; Hirota et al., 1997). 34 35 These cells are primarily defined as cells which express Class II antigens of the major histocompatability complex (MHC II), as well as appropriate co-stimulatory molecules such as B7 (Abbas et al, 1994) . From the literature regarding in vivo expression of MHC antigens, it is evident that MHC II expression in the rodent CNS is limited to two cell types: microglia and perivascular cells (Streit et al., 1989a). Endothelial cells are the other cellular component within the rodent CNS capable of MHC expression, but only Class I antigen (Streit et al.. 1989b). As microglial and endothelial cells are present in the developing rodent CNS as early as E12 (Ashwell et al., 1989) while perivascular cells do not appear in significant numbers before the first postnatal week (Mato et al., 1985), it is very likely that donor-derived MHC II expression occurs primarily on microglia within fetal transplants. Thus, if donor-derived MHC II expression were truly responsible for initiating immunological rejection, the removal of microglia from the donor tissue prior to transplantation should delay or prevent this response. The removal of endothelial cells may also be beneficial by removing an additional source of foreign antigen, i.e. MHC Class I antigens. In the present study, we describe the use of single-cell suspensions of microglia and endothelia-depleted fetal neural tissue in two common transplantation paradigms in which rejection has been well characterized: xenografts of El 4 mouse neural tissue placed into the intact rat striatum (Takei et al., 1990; Finsen et al., 1991 ; Duan et al., 1995); and allografts of E14 rat neural tissue (RTl .A/B" MHC haplotype) placed into the injured spinal cords of histo-incompatible rats (RTl.A/B" MHC haplotype) (Theele et al., 1996). The depletion of microglia and endothelial cells from fetal cell suspensions has been previously described (Pennell and Streit, 1996), using the specific affinity of the Bj-isolectin from Griffonia 36 simplicifoUa for these cell types and negative cell selection with magnetic beads. Both the growth and survival of these grafts as well as the pattern of immunological rejection is followed over time both morphologically and immunohistochemically. In addition, to ascertain the immunogenic potential of the allogeneic donor tissue both cultured microglia and whole and microglia-depleted fetal suspensions from donor rats were tested for their ability to present antigen to host T-cells in vitro using a modified mixed lymphocyte reaction. Materials and Methods Preparation of Whole and Microglia-Depleted Fetal CNS Suspensions The preparation of fetal CNS single-cell suspensions and the microglia depletion procedure have been previously described (Pennell and Streit, 1997). Briefly, E14 fetuses were obtained from either timed-pregnant female ACI rats or ICR mice (Harlan Sprague- Dawley, Inc.). Isolated fetal CNS (cortex, brainstem, and spinal cord) was stripped of meninges in a 35 mm Petri dish while immersed in 2 ml of ''Solution D" (0.8% NaCl, 0.04% KCl, 0.1% Glucose. 2.2% Sucrose, 0.01% Streptomycin sulfate, 0.025% Fungizone, and 0.006% Penicillin G in dHjO). Clean fragments were then incubated in 0.05% trypsin in "Solution D" for 20 min., triturated, and incubated an additional 10 min.. in DNAase (5800 units/ml). An equal volume of complete medium (DMEM + 10%) FBS) was then added to stop the reaction and the suspension was filtered through a 130 ^m Nitex filter. The filtrate was pelleted (400 x g, 10 min..), resuspended in 5 ml of complete medium, and filtered through a 40 fim Nitex filter. The cells were then counted using a hemocytometer 37 and their viability and degree of dissociation assessed using an acridine orange/ethidium bromide viability stain (live cells fluoresce green while dead cells fluoresce red; Fig. 3-lc). GS I-B4 was conjugated to tosylactivated Dynabeads® (Dynal, Inc.) according to the manufacturer's specifications. Fetal CNS suspensions (1x10^ cells) were added to 4x10' GS I-B4-conjugated Dynabeads® in 5 ml test tubes. The tubes were then incubated at 4°C under bi-directional rotation for one hour, after which the tubes were placed in a magnetic particle concentrator which pulls bound cells to the side of the tube. The suspension was transferred to a new tube and the process repeated a second time. The freshly depleted suspension, as well as a control, undepleted (whole) suspension, was smeared on gel-subbed slides, allowed to dry, and stained with GS I B^-HRP (Sigma L5391). By forming a ratio of GS I- B4-positive cells in experimental suspensions and GS I-B4-positive cells in control suspensions, we obtained a measure of depletion effectiveness (Fig. 3-la-c). Grafting Procedures A total of 79 adult female Wistar-Furth (WF) rats (Harlan Sprague-Dawley, Inc.) weighing 200-250 g were used as graft recipients. For the striatal xenografts, approximately 1x10* cells (10 |il total volume) from whole or depleted murine suspensions were injected stereotactically into the intact right striatum under ketamine/xylazine anesthesia, using a 25 gauge Hamilton syringe. The stereotaxic coordinates were 1 mm rostral to bregma, 3 mm lateral to bregma, and 4.5 mm ventral to the dural surface. The injection was performed slowly over a period of five minutes and the needle was allowed to remain in place for 38 several min. afterwards to avoid reflux. The number of animals and time points examined are described in Table 1 . Allograft recipients received low thoracic spinal contusion injuries under ketamine/xylazine anesthesia. Spinal cord injury was induced using the New York University (NYU) weight drop contusion device (Gruner et al., 1992). Seven to ten days after injury, approximately 1x10'' cells (15 |j.l total volume) from whole or depleted fetal rat suspensions were injected into the lesion cavity at the Tl 1 spinal cord level, using a 25 gauge Hamilton syringe. The number of animals used and time points examined is described in Table 1 . Tissue Processing The animals were sacrificed by transcardiac perfusion following the indicated survival periods after transplantation (see Table 1). Two animals from each group were perfused with 100 ml of 0.9% saline and the tissue frozen in cold isopentane for immunocytochemistry. The remaining animals at each timepoint were perfused with 100 ml saline followed by 200 ml Bouin's fixative. The Bouin's-fixed tissue was post-fixed for 24 hours and embedded in paraffin. Immunostaining was carried out on the fresh-frozen tissue using a panel of monoclonal antibodies (Harlan Sera-Lab). F4/80 (1 TOO dilution) was used to visualize murine microglia within the xenografts, OX-6 (1 :2000 dilution) was used to visualize MHC Class II antigens from both donor and host cells. F 17-23-2 (T.IOOO dilution) was used to visualize allograft donor (ACI) haplotype-specific MHC Class II antigens, and W3/25 39 (1 :150 dilution) was used to visualize CD4* T-cells, macrophages, and microglia. Frozen sections were cut on a cryostat at a thickness of 25 |j,m, mounted onto gel-subbed slides, and allowed to air dry for 30 min.. The sections were fixed for 5 min. at room temp, using a modified periodate-lysine-paraformaldehyde (0.75% paraformaldehyde) fixative (McLean and Nakane, 1974; Theele et al., 1996), followed by 2 min. in cold acetone. After rinsing in PBS, the sections were blocked in 10% normal goat serum for one hour at room temp., rinsed briefly, and incubated overnight at 4° C in primary antibody diluted in PBS + 1% BSA + 0.2%) Triton X-100. Sections were then rinsed three times in PBS and incubated for two hours at room temp, with biotinylated goat anti-mouse secondary antibody (Vector Laboratories, Inc.; 1 :300 dilution in PBS). The sections were rinsed three times in PBS and incubated for 45 min. at room temp, with avidin D-HRP (Vector; 1 :400 dilution in PBS). The sections were then rinsed three times with PBS and the antibody binding sites visualized using diaminobenzidine-H202 (DAB; Sigma) as a peroxidase substrate. The sections were dehydrated in an ascending series of ethanols, cleared in xylene, and coverslipped in Permount. Some sections from each specimen were stained with cresyl violet. To elucidate general graft morphology, paraffin embedded tissue was cut on a microtome at a sectional thickness of 7 \Jim for hematoxolin/eosin (H&E) staining. Antigen Presentation Assays The spleens from adult female WF (host/responder) and ACI (donor/stimulator) rats were removed and homogenized in 5 ml of "Solution D" using a 15 ml Bounce 40 homogenizer. 5 ml of homogenate was layered onto 3 ml of Ficoll solution (Histopaque 1083, Sigma) and centrifuged at 800 x g for 30 min. at 4°C. The interface was collected, spun down (250 x g, 10 min.), and resuspended in Iscove's modified Delbecco's medium (IMDM) containing 10% FBS. To enrich the responder T-cell population, the WF splenocytes were panned to remove B-cells using anti-rat IgG-coated 100 mm Petri dishes. The cells were added to the coated dishes and the IgG-positive cells were allowed to bind for 30 min. at room temp., after which the process was repeated twice using fresh plates. The remaining non-adherent cells were collected and counted for antigen presentation assays. The ACI-derived splenocytes were counted and used as positive control stimulator cells. Purified microglial cultures were prepared as previously described (Pennell and Streit, 1998). Stimulated microglia were treated with 100 units/ml of recombinant rat IFN-y (R&D Systems, Inc.) and 10 |J.g/ml LPS (Sigma) in complete medium for 72 hours, while unstimulated microglia were cultured in complete medium alone. Stimulated and unstimulated microglia were scraped from their culture dishes using a disposable rubber cell scraper, pelleted, resuspended in IMDM + 10 % FBS, and counted. Stimulator cells were irradiated with '"Cs to prevent proliferation. ACI splenocytes as well as whole and depleted fetal rat CNS suspensions were exposed to 1 500 rad of radiation, while microglial cells were exposed to 2000 rad. 4 x 10'^ stimulator cells were added to 8 x 10' WF splenocytes, in triplicate, in round bottomed 96 well plates. WF splenocytes alone served as a control for non-specific proliferation. The cells were maintained in an incubator with 8% CO, at 37° C for three days before being pulsed with 1 41 |aCi/well of [""HJ-methylthymidine (NEN, Inc.). Approximately 1 8 hours after being pulsed, the cells were collected using a semi-automated cell hairvester and the incorporation of ['H]- methylthymidine was assessed using a liquid scintillation counter. TABLE 3-1 Timepoints and Number of Animals XENOGRAFTS ALLOGRAFTS Time (days) Depleted Non-depleted Depleted Non -Depleted 7 3 3 6 6 14 5 5 3 3 21 5 5 3 4 30 - - 4 4 35 5 5 - - 45 - - 5 5 Results Fetal Cell Suspension and Microglial Depletion Suspensions of El 4 rat and mouse brain consisted of single cells with a few scattered clumps of five cells or fewer (Fig. 3- Id). Viability stains consistently demonstrated >95% viability of whole and depleted suspensions. Lectin staining of non- depleted suspensions indicates that microglia and endothelial cells together make up no more than 1% of the total cells within the suspensions. As has been previously described (Pennell et al., 1995; Pennell and Streit, 1997a), we were able to remove approximately 98% of microglia from the depleted suspensions (Fig. 3-lc). This represents approximately one microglial or endothelial cell per 5,000 total cells within the suspensions. Figure 3-1, Analysis of the fetal CNS suspensions. (A) Whole (non-depleted) El 4 mouse CNS suspension, stained with GS-I-B4-HRP. Notice the darkly stained cells (arrows), which represent donor microglia and/or endothelial cells. (B) Depleted El 4 mouse CNS suspension, also stained with the lectin. Notice lack of stained cells, x 200 (C) Acridine orange/ethidium bromide viability stain of living whole E14 mouse CNS suspension, illustrating both the high degree of dissociation and the high viability. (D) Graphical representation of microglial depletion from both E14 rat and mouse CNS suspensions. Approximately 98% of lectin-positive cells have been removed from the depleted suspensions. Student's t-test indicates significance of this depletion with a p<0.0001 (n=3 for the mouse and 4 for the rat suspensions). 43 Whole B Depletion Effectiveness %L«c(in+CrilslJ6 Depleted ■ RJIWIK* ■ Mouse IWote [JMou« DapMed Figure 3-1 44 Analysis of Intrastriatal Xenografts By 7 days post-transplantation (pt), intrastriatal xenografts in all animals appeared as large areas of undifferentiated cells (Fig. 3-2a,b). Perivascular cuffing, the accumulation of mononuclear cells within the perivascular space and a hallmark of graft rejection, was evident in blood vessels within grafts of both whole and microglia-depleted tissue. There were no F4/80-positive cells within the grafts in either group, indicating a lack of donor microglial cells within the grafts. OX-6 immunostaining indicated that grafts from both groups were populated with MHC Class Il-positive cells with the ramified morphology typical of activated microglia. Perivascular cells within and around the grafts were also positive for OX-6. There was little or no W3/25 immunoreactivity within the seven day grafts, although CD4-positive microglia and perivascular cells were evident throughout the rest of the brain. By 14 days/?/ the grafts were larger and more organized, with many cells within the grafts taking on a distinctly neuronal morphology (Fig. 3-2c,d). Perivascular cuffing was more prominent than at seven days within both whole and depleted grafts, with most inflammatory cells localized to the periphery of the grafts. There were still no F4/80- positive cells within any of the grafts examined. All grafts were filled with OX-6-positive cells with both ramified and rounded morphology. W3/25 staining indicated that both CD4- positive microglia and a scattering of lymphocytes were present within all grafts by this time. By 21 days after transplantation, the grafts were much larger than at 14 days, with small islands of apparenfly healthy tissue scattered among larger areas of clearly rejecting graft (Fig. 3-2e,f). Both whole and depleted grafts were filled with mononuclear cells, and 45 even in regions of surviving graft the neurons often appeared shrunken and unheahhy. No F4/80-positive cells were apparent in any graft at this time, and OX-6 inimunoreactivity was spread evenly throughout the grafts, making identification of individual cells difficult. CD4-positive lymphocytes were present in large numbers in all grafts examined, indicating the presence of a large antigen-specific immune response. By 35 daysp^, there was no identifiable surviving graft tissue in any animal examined (Fig. 3-2g,h). Surrounding microglia were still OX-6 and W3/25-positive. but there were no lymphocytes remaining within the area of the grafts. Intraspinal Allografts At 7 days/7/, the grafts in both groups were small and appeared relatively undifferentiated (not shown). The grafts were located at the periphery of the lesion site, often at the edges of cysfic cavities formed as a result of the contusion injury. There were numerous macrophages in and around the grafts, but no signs of perivascular cuffing in the surrounding blood vessels. There were numerous OX-6-positive cells throughout the injured spinal cord, mostly activated microglial cells and macrophages, and there were numerous large, round MHC Il-positive cells within the grafts. There were no donor- derived MHC Il-positive cells within the whole or depleted grafts, as evidenced by F17-23- 2 immunostaining. CD4-positive cells were limited to the injured spinal cord surrounding the grafts. By 14 days pt, the grafts were large, filling most of the center of the lesioned spinal cords and extending rostral and caudal of the injury site (Fig. 3-3a,b). Both whole and depleted graft tissue appeared more organized and less densely cellular than at 7 days. Figure 3-2. H&E stained sections of whole and depleted intrastriatal xenografts at 7, 14, 21, and 35 days pt. (A, B) At 7 days pt both whole and depleted grafts were fairly large. Notice that large blood vessels near each graft contain numerous darkly staining mononuclear cells i.e. perivascular cuffing. (C, D) At 14 days pt the grafts are larger and contain many neurons. Both whole and depleted grafts are bordered by blood vessels containing inflammatory cells, but the grafts themselves are relatively free of immune cells. (E, F) By 21 days pt all grafts in both groups are undergoing rigorous rejection, with many foci of densely packed mononuclear cells in each graft. (G, H) By 35 days pt there was no recognizable graft left in any animal examined, although the site of injection was readily visible, sometimes surrounded by pockets of inflammatory cells. Scale bar=l 10 |im. 47 Figure 3-2 . ./>,> ^ •//:•;. v;v,;vVyhol$ ;/.• . , g-^ Depleted • ./ . * '. ..'"W,'-' -.*.• 'Vr'*A' • ;-.''"'• A^'"' '•'*'- >^^'">,- .A' .•■ ■•■ ■■ '■ • ..B-r-:,l ■■■■■■■.': ..-••■ -.1' "' >'.'/«S'* < ' "\-- -> /' • ' .£ V'." ' ' '^ Ai' D '^''•^' ,■ ■"■ ■:m:'xA:'--' ^:•.^^^::■^^*•|;^V%--■■■v. Figure 3-3. H&E stained sections of whole and depleted intraspinal allografts at 14, 21, and 30 days;?/. (A, B) At 14 days/7/ all grafts in both groups (g) were large and well organized, with good integration with the host spinal cord (h). There were no signs of inflammation in any graft examined, and blood vessels were free of adherent mononuclear cells. (C, D) At 21 days grafts in both groups were practically identical to those at 14 days, albeit larger. Again, the grafts seemed to be surviving well with no signs of inflammation. (E, F) By 30 days pt the grafts in both groups had undergone a striking change. In several grafts in both groups there was no longer any grafted tissue evident (E) while in those with surviving grafts there was a robust rejection response underway (F), with the remaining areas of graft inundated with inflammatory cells. Scale bar=l 10 |im. 49 Whole Depleted -H Figure 3-3 *. , D 50 with numerous neuronal profiles. There was no sign of perivascular cuffing in any of the grafts examined in either group, although there were many macrophages within and surrounding the graft tissue. Surprisingly, there were only a few scattered OX-6-posifive cells within the whole and depleted grafts at this time, although the surrounding spinal cord was filled with OX-6-positive microglia and macrophages. The first donor-specific MHC ITpositive cells appeared at this time, although there were only a few round cells apparent in any graft, sometimes located in perivascular spaces. There were no differences in number or appearance of these cells between the whole and microglia-depleted grafts. W3/25 staining was still primarily limited to cells outside the grafts, with a few rounded cells within the grafts, although there was no difference in this pattern between the two groups. By 21 days pt, the grafts were morphologically very similar to the 14 day grafts, although they were significantly larger (Fig. 3-3c, d, 3-4a, c). There was sfill no evidence of perivascular cuffing or inflammatory cell infiltrate in any graft examined in either group. Immunostaining patterns were also virtually identical to those seen at 14 days, with few MHC II or CD4-positive cells within the grafts, and a very few scattered donor-specific MHC Il-positive cells (Fig. 3-4b, d). It is clear that at this fime the grafts are sfill surviving well with no sign of immunological rejection. At 30 days pt. the grafts had undergone striking change (Fig. 3-3e, f). There was no recognizable surviving graft fissue in two of the whole graft recipients and one of the depleted graft recipients, and the remaining grafts in both groups contained a dense cellular infiltrate with widespread perivascular cuffing throughout the surviving graft tissue. The surviving areas of graft were filled with OX-6-posifive cells, while the Figure 3-4. Expression of donor-derived MHC class II antigen in whole and depleted intraspinal allografts. (A, C) Nissl stained sections of 21 day whole (A) and depleted (C) allografts, showing large, healthy grafts free of inflammation. (B, D) Higher power micrographs of the approximate areas outlined in A and C in adjacent sections, stained with F 17-23-2. Only a few isolated donor MHC II expressing cells are present within the grafts. (E,G) Nissl stained sections of 30 day whole (E) and depleted (G) allografts, showing that the grafts at this time were inundated with small, darkly stained inflammatory cells, indicative of a rejection response. (F, H) Higher power micrographs of the approximate areas outlined in E and G in adjacent sections, also stained with F 17-23-2. Areas of graft which are undergoing rejection now contain numerous donor MHC II expressing cells, while areas of graft free of inflammation are still relatively free of donor MHC II expression (not shown). Scale bar=l 10 |um. /■I' "f 52 ^' B ■,., ■■ -'i :"■/'■ • >\ '<»^, , \ ■ •'-'■. %- Figure 3-4 •j .y H 53 periphery of the grafts where the densest areas of mononuclear cell infiltrate were located contained numerous donor MHC Il-positive cells (Fig 3-4f,h). There was no difference in the numbers or patterns of staining of donor MHC Il-positive cells between whole and depleted grafts. By 45 days pt, none of the whole grafted spinal cords and only one of the depleted spinal cords contained surviving graft tissue (Fig. 3-3i,j). The surviving graft was surrounded by a dense cellular infiltrate and many of the neurons within the graft appeared shrunken and unhealthy, indicating that this graft was in the final stages of rejection. All spinal cords still contained numerous OX-6 and W3/25-positive microglia and macrophages, but no CD4-positive lymphocytes were apparent in the spinal cords in which grafts had been completely rejected. There were no remaining cells expressing donor MHC II. Antigen Presentation Assays The results of the antigen presentation assays are presented graphically in Fig. 3-5. Donor (ACI rat)-derived splenocytes (positive control) induced proliferation of host (WF)- derived lymphocytes in a dose-specific fashion, with 4x10'' and 4 x 10' stimulator cells inducing significant proliferation above control values (p<0.002 and p<0.001 respectively; n=4 each). Allogeneic microglia, both normal and stimulated with LPS and IFN-y, actually seemed to suppress lymphocyte proliferation, although the suppression was not quite significant (p<0.087 and p<0.097 respectively; n=3 each). Suspensions of allogeneic E14 rat CNS tissue, whether whole or depleted of microglia, did not induce a Figure 3-5. Graphical representation of 'H-thymidine incorporation by host T-cells in response to various allogeneic stimuli, in counts per minute (CPM). Allogeneic splenocytes induced T-cell proliferation in a dose-dependent manner, with significant responses when 4 X lO** and 4x10' stimulator cells were added (p<0.002 and 0.001, respectively). In contrast, there was no significant proliferative response to allogeneic normal or stimulated microglia or to whole or depleted E14 CNS suspensions. 55 T Cell Proliferation Assay 60000 Q. O 50000 - 40000 - 30000 - 20000 10000 H 0 Control (n=7) 4x10^ ACI Splenocytes (n=3) 4x10"* AC I Splenocytes (n=4) 4x10^ ACI Splenocytes (n=4) Cultured ACI Microglia (n=3) D Stimulated ACI Microglia (n=3) i Whole E14 CNS Suspension (n=5) ] Depleted E14 CNS Suspension (n=3) Figure 3-5 56 significant response from the responder lymphocytes (n=3 each), indicating a lack of functional antigen presenting cells within the graft tissue. Discussion In the present study we have attempted to elucidate the cellular components of allogeneic and xenogeneic fetal neural transplants which are responsible for the initiation of immunologic rejection. We have done this by removing those cells most likely to represent flinctional antigen presenting cells (APCs) within the donor tissue, i.e. microglia. We found that depletion of microglia and endothelial cells from fetal neural alio- and xenografts did not alter the pattern of rejection in any observable way, indicating that these cells are unlikely to play a significant role in the initiation of graft rejection. We also demonstrated that neither allogeneic fetal neural cell suspensions nor purified allogeneic microglia induced T-cell proliferation in vitro, indicating both a lack of functional APCs within the suspensions and that microglia alone do not act as APCs. Together, these findings indicate that the traditional view of the microglial cell as the resident immunocompetent cell of the CNS (Streit et al., 1988) may need to be reevaluated. Surprisingly, we have also shown that single-cell suspensions of allogeneic fetal CNS tissue placed into the sub-acutely (7-10 day post-contusion) injured spinal cord undergo a significant delay in the onset of rejection when compared to reports of fetal spinal cord piece allografts placed into the acutely injured spinal cord, although the final outcome is unchanged. This rejection does not coincide temporally with donor MHC II expression, as has been previously shown in other allograft models (Theele et al., 1996; Lawrence et al., 57 1990). Instead it occurs between one and two weeks after donor MHC II expression was first noted. This finding, along with those described in the previous paragraph, lead us to conclude that the presence of MHC-expressing cells is not the primary stimulus for allograft rejection in suspension grafts, but that other factors must be coming into play, such as indirect presentation of minor histocompatability antigens. Xenograft Rejection In the current study, the rejection of murine striatal xenografts followed a pattern identical to that described in previous studies (Finsen et al.. 1991; Duan et al., 1995). This pattern was not altered by the removal of microglia and endothelial cells from the donor tissue prior to transplantation, for which there could be a variety of reasons. We have demonstrated that microglia and endothelial cells together make up no more than 1% of the total number of cells within the donor cell suspensions, and that we cannot detect any murine microglia within the xenografts at any time point, using species-specific markers. This duplicates results produced by Perry and Lund (1989) in which they were unable to find donor-derived microglia within murine retinae transplanted into the rat brain. A later study which followed human fetal microglia in xenografts placed into the rat brain demonstrated that these cells were few in number and that they disappeared over time (Geny et al., 1995). Taken together with the results described here, it would appear that few if any fetal microglia survive within developing xenografts, and thus their removal prior to transplantation may be unnecessary. In addition, there is evidence that the primary stimulus for xenograft rejection is not direct presentation of foreign antigen to host T-cells by donor APCs, as is likely the case in 58 allograft rejection, but rather indirect presentation of minor histocompatability antigens by host APCs is the initiating event (Kaufman et al., 1995). This is supported by our observation that perivascular cuffing within striatal xenografts appears within seven days, a period of time in which there is not likely to be much graft-specific MHC expression (Takei et al.. 1990). Similarly, bovine chromaffin cells transplanted into the rat CNS do not survive without immunosuppression, even after depletion of MHC expressing cells (Czech et al, 1997), further illustrating that xenogeneic cells which do not express MHC antigens are subject to rejection. Conversely, a recent study has shown that there is no immune response directed against human Schwann cell transplants placed into the adult rat brain, despite constitutive MHC II expression on the implanted cells (Hermanns et al., 1997). Taken together, these findings indicate that MHC antigens are just one among many immunogenic molecules present within xenografts, and are not themselves essential for immunologic rejecfion. Allograft Rejecfion The allograft model used in this study, donor ACI rats paired with recipient Wistar- Furth rats, is known to be a high responder cross. A previous spinal allograft study using these two strains demonstrated induction of donor-specific MHC II expression by 14 days pt which increased until it was widespread at 21 days (Theele et al., 1996). This MHC II expression coincided temporally with the onset of inflammatory cell infiltration and perivascular cuffing within the grafts as well as CD4* and CDS* T-cell infiltration, and the grafts were all destroyed by 30 dayspt. This model clearly illustrates the three stages of rejection outlined by Lawrence and colleagues (1990): 1) the immune induction phase. 59 coinciding with the onset of graft MHC expression; 2) the phase of immune attaclc, coinciding with immune ceil infiltrate and graft tissue destruction; and 3) the quiescent phase, in which inflammation recedes, ft was our goal to interrupt this process at the immune induction phase, by removing the cells from the graft most likely to express MHC antigens. A similar study was performed in which MHC antigen expression was induced on allogeneic neural tissue in vitro by IFN-y, following which all cells expressing MHC were removed using anti-MHC conjugated magnetic beads (Bartlett et al., 1990). Although this strategy prevented rejection of the allografts, the transplants themselves consisted only of a subset of neurons, all the glia and even some neurons having been induced to express MHC. Our donor suspensions are complete in all neuronal and macroglial elements. We have succeeded in delaying the immune induction phase, although not in the manner we intended. Our results also demonstrate low level MHC II expression within the grafts by 14 days pt, but this level does not increase until between 21 and 30 days pt. The observation that this increase is coincident with inflammatory cell infiltration may indicate that the MHC expression is induced by the CD4^ T-cells present within the graft, a likely source of pro-inflammatory cytokines such as IFN-y. Furthermore, this delay in the onset of the phase of immune attack was independent of the removal of microglia and endothelial cells, indicating that these cells are probably not the direct cause of rejection, as was originally hypothesized. Instead, it appears that the suspension itself is the cause of the delay. Previous allograft studies have utilized either whole piece grafts which contain complete blood vessels and probably some residual meningeal tissue (Lawrence et al., 1990; Theele et al., 1996) or "slurry" suspensions which consist of large clumps and 60 aggregates of cells and cellular debris (Duan et al., 1995; Poltorak and Freed, 1991; Reier et al., 1992), while the present study uses only dissociated cells. This process probably results in a much lower antigenic load, filtering out the majority of vascular and meningeal elements as well as cellular debris which could contribute to the pool of alloantigen within the graft site. This may have raised the threshold of antigen necessary to initiate an immune response, or merely increased the amount of time necessary to mount a response. Interestingly, this period between 21 and 30 days is also the window of time in which new blood vessels fully mature within developing grafts (see chapters 4 and 5). It would be interesting to find out when host perivascular cells are recruited to the new vessels, providing a ready pool of non-microglial host APCs within the graft. Another key difference between this and previous studies is the use of sub-acutely injured animals (7-10 days post-injury) as opposed to acutely injured animals (Theele et al., 1996; Theele and Reier, 1996). The sub-acute contusion model reduces the risk that the graft tissue will encounter peripheral and meningeal elements when compared to acute resection lesions, and also allows the damaged spinal cord time to repair vascular damage and reestablish the blood-brain barrier, limiting the graft's exposure to serum products and circulating lymphocytes. Microglia as Antigen Presenfing Cells The prevailing opinion in regards to the role of the microglial cell in transplant biology, and in neuroscience in general, is that microglia represent the resident immunocompetent cells of the CNS (Streit et al., 1988; Guilian, 1995; Poltorak and Freed; 1989). This view is supported by a large body of research demonstrating in vitro and in vivo 61 expression of MHC I and II antigens on microglia (Suzumura et al.. 1987; Matsumoto et al., 1986; Streit et al., 1989; Morioka et al., 1992; Popovitch et al., 1993; Steiniger et al., 1988) as well as their ability to present antigen to syngeneic T-cells (Frei et al., 1987). More recently, however, evidence has indicated that microglia may not be fully immunocompetent APCs after all. Ford and colleagues (1995) separated microglia from other brain-derived macrophages, i.e. perivascular cells, by flow cytometry and discovered that while perivascular cells were potent APCs, microglia were quite poor at stimulating either T-cell proliferation or IL-2 production. A recent study characterized the poor antigen- presenting ability of adult microglia as comparable to that of immature APCs from other tissues (Carson et al., 1998). Taking this idea one step farther, a recent study showed that purified microglia actually stimulate apoptosis in T-cells (Ford et al., 1996). These previous studies support the results in the current study, which demonstrate that purified allogeneic adult microglia do not function as APCs. and that fetal microglia within the cell suspensions are also incapable of antigen presentation. Conclusions It has been thought for some time that MHC antigens expressed by donor tissue, and by microglia in particular, have been primarily responsible for initiating graft rejection. Our results indicate that MHC expression within alio- and xenografts is probably not the primary stimulus in rejection of suspension grafts, and that donor microglia are not functional APCs and thus not responsible for immune recognition of the grafted tissue. We have also shown that spinal allograft rejection can be delayed for more than three weeks by the use of single-cell suspension grafts as opposed to the grafting of pieces of neural tissue. 62 These findings may have an impact on current and future clinical protocols regarding the time course of immunosuppression used in human transplants. CHAPTER 4 COLONIZATION OF NEURAL ALLOGRAFTS BY HOST MICROGLIA: RELATIONSHIP TO NEOVASCULARIZATION Introduction There has been considerable interest over the last decade in the use of fetal neural tissue grafts to treat traumatic and degenerative disorders of the central nervous system (CNS) (Bjorklund, 1991; Dunnett and Richards, 1990; Gage and Buzsaki, 1989; Gash and Sladek, 1988; Reier et al., 1992). Where once neuronal loss was thought irreparable, neural transplantation has reached a point where significant recovery of function is being reported, both in animal models and in clinical studies (Cheng et al., 1996; Fiandaca, 1991; Freed et al., 1992; Freed et al., 1993; Reier et al., 1994). However, despite these successes, there are fundamental aspects of transplant biology which are not yet understood and which will need to be investigated before neural tissue transplantation can be applied clinically to the wide range of neural disorders for which it has shown potential. In the present study, we have paid particular attention to potential interactions occurring between microglial and endothelial cells, as we have hypothesized that microglia are involved in promoting graft neovascularization. Little is known about functional roles of microglial cells in neural transplants, and the source of microglia (i.e. donor or host derived) within fetal grafts has been subject to debate (Geny et al., 1995; Perry and Lund, 1989). Most studies have focused on the 63 64 cells' capacity to act as brain macrophages and thus, as a likely source of antigen presenting cells (Duan et al., 1995; Finsen et al.. 1991; Lawrence et al., 1990). More recently, the limited traditional view of microglia as a mere scavenger cell has been broadened by experimental observations demonstrating that microglia produce neurotrophic factors in vitro (Elkabes et al., 1996; Mallet et al., 1989; Nagata et al., 1993), and can enhance the poor regenerative capacity of the adult CNS in vivo (Rabchevsky and Streit, 1996). Additional support derives from experiments showing that transplants of peripheral macrophages help to overcome the inhibitory nature of the CNS towards axonal regeneration (Lazarov-Speigler et al, 1996). Although a role of macrophages in angiogenesis and neovascularization has been well described (Clark et al.. 1976; Evans, 1977; Greenburg and Hunt, 1978; Mostafa et al., 1980; Polverini et al., 1977a; Polverini et al., 1977b), little attention has been paid to a possible role for microglia in transplant vascularization (for review see Horner et al., 1994). The goal of the current investigation was to study microglia in striatal suspension grafts and to correlate their temporal appearance and morphology with the development of graft vasculature. The B4 -isolectin from Griffonia simplicifolia (GS I-B4) binds specifically to microglial and endothelial cells in the developing and adult rat CNS. making it an ideal tool for visualizing both microglia and blood vessels (Ashwell et al., 1989; Peters and Goldstein, 1979; Streit, 1990; Streit and Kreutzberg, 1987). Furthermore, by depleting the grafts of donor microglia and endothelial cells prior to transplantation, we have isolated the effects of donor and host cells on neovascularization and determined the contribution of donor microglia to the overall population of microglia within the graft. Our results suggest that striatal suspension grafts are populated and vascularized entirely 65 by host-derived microglia and endothelial cells. We hypothesize that these invading microglia play a pivotal role in the neovascularization of the developing grafts. Materials and Methods Preparation of Fetal Tissue Suspension E14 fetuses were obtained from timed-pregnant female Sprague-Dawley rats (Harlan Sprague-Dawley, Inc.). The fetuses were delivered by Cesarean section under deep sodium pentobarbital anesthesia. Isolated fetal CNS (cortex, brainstem, and spinal cord) was stripped of meninges in a 35 mm Petri dish while immersed in 2 ml of "Solution D" (0.8% NaCl, 0.04% KCl, 0.1% Glucose, 2.2% Sucrose, 0.01% Streptomycin sulfate, 0.025%) Fungizone, and 0.006%) Penicillin G in dHiO). Clean fragments were then incubated in 0.05% trypsin in "Solution D" for 20 minutes, triturated, and incubated an additional 10 min. in DNAase (5800 units/ml). An equal volume of complete media (DMEM + 10% FBS) was then added to stop the reaction and the suspension was filtered through a 130 ^im Nitex filter. The filtrate was pelleted (400 x g, 10 min.), resuspended in 5 ml of complete media, and filtered through a 40 i^m Nitex filter. The cells were then counted using a hemocytometer and their viability assessed using an acridine orange/ethidium bromide viability stain (live cells fluoresce green, dead cells fluoresce red). Viability was always >95%o. Deplefion of Microglia and Endothelia from Fetal CNS Suspensions GS I-B4 was conjugated to tosylactivated Dynabeads® (Dynal, Inc.) according to the manufacturer's specifications. Optimal cell selection using Dynabeads® is obtained using a 66 beadxell ratio of 40:1 . Under the original assumption that microglia make up no more than 10% of the total cells in the fetal CNS suspension, IxlO' cells were added to 4x10^ GS I-B^- conjugated Dynabeads® in 5 ml test tubes. The tubes were then incubated at 4°C under bi- directional rotation for one hour, after which the tubes were placed in a magnetic particle concentrator which pulls bound cells to the side of the tube. The suspension was transferred to a new tube and the process repeated a second time. The freshly depleted suspension, as well as a control, undepleted (whole) suspension, was smeared on gel-subbed slides, allowed to dry, and stained with GS 1 B4-HRP (Sigma L5391). By forming a ratio of GS I- Bj-positive cells in experimental suspensions and GS I-B4-positive cells in control suspensions, we obtained a measure of depletion effectiveness. Intrastriatal Grafting Procedure Under ketamine/xylazine anesthesia, approximately 1 x lO*" cells (10 ^1 total volume) from whole or depleted suspensions were injected stereotactically into the right striatum of adult male Sprague-Dawley rats (Harlan Sprague-Dawley, Inc.) weighing approximately 250 g, using a 25 gauge Hamilton syringe. The stereotaxic coordinates were 1 mm rostral to bregma, 3 mm lateral to bregma, and 4.5 mm ventral to the dural surface. The injection was performed slowly over a period of five minutes and the needle was allowed to remain in place for several minutes afterwards to avoid reflux. Four animals were used for each timepoint (3, 7, 10, 14, 21, and 30 days), two receiving whole, non- depleted suspensions and two receiving depleted suspensions. 67 Tissue Processing and Lectin Histochemistry Animals were sacrificed by transcardiac perfusion with 100 ml of 0.9% saline followed by 200 ml of 4% paraformaldehyde in phosphate buffered saline (PBS), pH 7.4. The brains were removed and post-fixed in 4% paraformaldehyde for at least one week prior to sectioning. 60 ^m sections were cut on a vibratome and soaked in PBSC (PBS containing 0.4% Triton X-100 and 0.1 mM CaCf, MgCf, and MnCf ) at 4°C for 24hrs before being incubated with GS I-B4-HRP (10 |ig/ml) in PBSC overnight at 4°C. The sections were briefly rinsed with PBS and the lectin binding sites were visualized using 3,3' -diaminobenzidine-HjOT Sections were then mounted on gelatin-coated slides, allowed to air dry, dehydrated in ascending ethanols, and coverslipped with Permount. Image Analysis The percent area of graft covered by blood vessels was quantitated using a computerized image-analysis system and NIH Image software. The area covered by blood vessels was measured in four sections from each graft and from three noncontiguous areas of each section, and divided by the total area of graft to give a relative measure of the degree of vascularization. This method gives a better indication of overall graft vascularization than the counting of vessels, as the majority of blood vessels within the grafts are contiguous and joined by branches. 68 Results Depletion of Microglia and Endothelia from Fetal CNS Suspensions The microglial depletion procedure has been described previously (Pennell et al, 1995; Pennell and Streit, 1996). Single cell suspensions of E14 fetal CNS were obtained consistently, with a viability of 95% or higher. GS I-B4 staining of the whole cell suspensions revealed that microglia and endothelial cells make up no more than 1% of the total cell number. Negative selection using GS l-Bj-conjugated Dynabeads® resulted in a ratio within the depleted suspensions of approximately one microglial or endothelial cell per five thousand total cells (0.02%), a depletion effectiveness of approximately 98%) (Fig. 1). Student's t-test indicated statistical significance with p<0.0001 (n=3). The depletion procedure did not greatly affect cell viability, which remained consistently above 90% after depletion. Morphometric Analysis of Graft Vascularization The results from morphometric analyses of graft vascularization are summarized in Figure 3. There was no significant difference in percent area of graft covered by vessels between whole and depleted grafts at any time point. There were no vessels present at three days post-transplantation, and very few present (on average) at seven days. By ten days post-transplantation, however, the graft vascular bed covered over 8% of total graft area. Fourteen days after transplantation the blood vessels in both groups covered approximately 10% of total graft area, and by twenty-one days vascular bed coverage was at approximately 12%. There was no further increase in percent area of graft covered by vessels at thirty days post-transplantation, indicating that neovascularization was complete by twenty-one days. 69 Development of Microglia and Vasculature Within Striatal Suspension Grafts Lectin staining of the grafts provided clear visualization of microglia and microvasculature (Fig. 2). Staining of microglia and vessels was stronger inside the grafts than outside, allowing clear demarcation of the graft/host border. There was no apparent difference in the number of microglial cells between the whole and depleted grafts at any time point examined (Fig. 2A-F), indicating that the relative contribution of donor-derived microglia to the overall population of microglia within the striatal suspension grafts was negligible. Similarly, there was no significant difference between the density of vasculature within whole and depleted grafts at any time point, indicating that donor endothelial cells are not essential for graft neovascularization (Fig. 3). There were no signs of immunological rejection, such as lymphocyte infiltration, at any time examined. Little or no graft tissue was present three days after transplantation in either whole or depleted grafts. The needle track was evident, surrounded by activated microglial cells with an ameboid morphology (Fig. 5A). More ramified cells were present with increasing distance from the injection site. There was no evidence of blood vessels within the small grafts visible at this time. By seven days post-transplantation grafts were larger and evident in all animals. In both whole and depleted grafts there were numerous microglial cells, appearing hypertrophied with short, thick processes (Fig. 5B). Microglia were more numerous at the periphery of the graft, suggesting that they migrated in from the host tissue. Very few, if any, blood vessels were evident at this early time point (Fig. 4A). By ten days the grafts were considerably larger than at seven days, and the microglial cell density within the grafts appeared decreased, leaving large areas of graft free of microglial cells (Fig. 2C, D; Fig. 4B). Most cells were still hypertrophic, similar to those Figure 4-1. Effectiveness of depletion procedure using magnetic beads. The graph shows the percentage of GS I-B4 -positive cells in El 4 CNS cell suspensions before and after depletion. Student's t-test indicates significance at p<0.0001, n=3. 71 1 .ic 1.1 - T 1.0 - 1 1 Whole mm Depleted c: c ■- o 0.9 - (/) -^ ^ -1 0.8 ^ ^ a. 0.7 - + P ^ Q- 0.6 - QQ = _L CD 0.5 - (3 ^ °-^ - •4— ' vo o 0.3 - ^ H 0.2 - 0.1 - 0.0 - //////7/////A Cell Suspension Figure 4-1 72 seen at seven days. At this point, there was considerable vascular development, with numerous vessels growing in from the periphery of the graft (Fig. 4B). The vessels appeared very large in diameter compared to later time points, so that despite the low number of vessels they covered a large percentage of the graft. Interestingly, most microglial cells within the grafts at this time point were associated in some way with the newly formed blood vessels, either directly overlapping or wrapping their processes around them. By fourteen days post-transplantation, microglia were more numerous, maintaining their hypertrophic moi-phology, particularly towards the edges of the graft (Fig. 4C). Although the overall percent total area of vasculature was only slightly higher than at ten days post-transplantation, there appeared to be more branches and small diameter vessels than before. Microglia remained closely associated with blood vessels, and were frequently seen in the proximity of angiogenic sprouts from developing endothelium (Fig 5C). By twenty-one days post-transplantation, most microglia in the graft had assumed a morphology similar to normal, resting microglia in adult rat brain (Fig. 4D). They had spread out in a characteristic fashion, each cell occupying its own distinct, non-overlapping territory. The vascularization of the grafts was complete by 21 days, and microglial cells were no longer found predominantly associated with blood capillaries, although a number continued to wrap their processes around these vessels. At thirty days after grafting there was an increase in the arborization of microglial processes (Fig. 5D), and microglia and vasculature were indistinguishable from normal adult rat brain (Fig. 4E, F). Figure 4-2. Comparison of microglial and vascular staining in whole (A, C, E) and depleted (B, D, F) suspension grafts at 7, 10, and 21 days post-transplantation. Microglia and blood vessels are clearly visualized, and the grafts are demarcated from surrounding host tissue by their lighter background. Graft boundaries are outlined by arrows. Despite relative differences in graft size, there do not appear to be any differences in the degree of microglial colonization or extent of vascularization between whole and depleted grafts at any time point examined. GS I-B4 -HRP; scale bar= lOOjim. 74 \ \ \ * \ t. t 1 ^ ,, •« tar. • - S * ^ 2A B -:■••"' t V '\^ 3 "■ *-^ ' ' *► E*r > •» • "^ \. f" ^ -^'» ' ' ..«-/- 2id a^: * . 2id* • 1 ■• .^ ^ -. - • ^ J- ■I ♦ ■> *v ....'*, I' i» Figure 4-2 Figure 4-3. Graphical representation of graft vascularity in developing striatal suspension allografts over time. Maximum neovascular growth occurs between 7 and 10 days post- transplantation. Thereafter, little additional development takes place. Student's t-test indicates no significant difference between vascularity in grafts derived from whole or microglia-depleted CNS suspensions at any time point examined. 76 5 10 15 20 25 30 35 Days Post-Transplantation Figure 4-3 Figure 4-4. Time course of microglial and vascular development in whole suspension allografts into the striatum. (A), numerous ameboid and poorly differentiated microglial cells are present within the grafts 7 days after transplantation. There are very few blood vessels present at this time. (B), at 10 days there is a striking increase in vascularity within the grafts, consisting mostly of large diameter and poorly branched vessels. Most microglia are found adjacent to vessels. (C), by 14 days the blood vessels are of smaller diameter and more highly branched, and microglia are more numerous. Note that areas unoccupied by microglia are also relatively free of blood vessels. (D), at 21 days the graft vasculature resembles that of normal rat brain, although microglia still appear hypertrophied compared to resting cells. (E), by 30 days post-transplantation, the graft microglia and vasculature are indistinguishable from that of control, normal rat brain depicted in (F). GS I-B4 -HRP; scale bar= 100|a,m. 78 '*« * 7. «. f .. # \ if 4A * «r*-< :~-- .vU" I -^-'''-' ^ V -.7.%f/ 14d *Li - *^ / J •^ 21 d n:. "^' to-' y \ 4 »» * 30d ^. « w J >«■ _ ^ .-* \ ,' 4^ ^ ~ ^ ' **' fc ** % - 95%. Suspensions were counted and maintained at 4°C until transplantation. Spinal surgeries were performed as described in Chapter 3. Female Lewis rats weighing approximately 200 g received lower thoracic (Til) spinal cord injuries using the NYU weight drop device, from a height of 25 cm. One week later, the lesions were exposed and approximately 1x10'' cells (15 )j,l total volume) from the donor suspension was injected into each injury site using a 25 gauge Hamilton syringe. A total of twenty- five animals received transplants. Tissue Processing Under deep pentobarbital anesthesia, five animals each were sacrificed at 7, 10, 14, 21, and 30 days post-transplantation by transcardiac perfusion of 0.9% saline followed by 200 ml of Bouin's fixative (72.5% picric acid, 22.5% formalin, 5% glacial acetic acid). The spinal cords were removed and post-fixed in Bouin's fixative for 2 hours 89 before being embedded in paraffin for sectioning. Sections were cut on a microtome at a thickness of 12 jim and mounted on Fisher Superfrost PlusTf^ sHdes. Slides were baked at 65°C overnight and allowed to cool before being processed. Sections were deparaffmized in xylene for 10 min., rehydrated in descending ethanols, and rinsed twice in PBS. Antigen retrieval was performed by microwaving the slides in 10 mM citrate buffer (pH 6.0) for 10 min. and allowing the slides to cool to room temp. Sections were blocked for 2 hours at 4°C in PBS + 1% BSA + 1% normal goat serum. Primary antibodies were diluted in the blocking solution. Mouse anti-bFGF (Upstate Biotechnology, Inc.), rabbit anti-VEGF (Santa Cruz Biotechnology, Inc.), rabbit anti-TGF-|3 (R&D Systems), and Griffonia simplicifoUa B4- isolectin-peroxidase (GS-1- B4-HRP; Sigma) were used at 1 :100 dilution, while mouse anti-GFAP (Sigma) was used at 1:500. The TGF-P antibody is pan-specific and recognizes TGF-P 1, 2, and 3, while the VEGF antibody recognizes all four splice variants of VEGF. GS-1- B4-HRP was used to visualize microglia/macrophages and blood vessels. Sections were incubated in primary antibodies and lectins at 4°C overnight. Controls in which primary antibodies were omitted were performed in every case. Sections were rinsed three times in PBS and incubated in either biotinylated goat anti-mouse (bFGF, GFAP) or goat anti-rabbit (TGF-P, VEGF) secondary antibody at a dilution of 1 :400 for 2 hours at room temp. The sections were then rinsed three times in PBS and incubated in 1 :400 avidin D-HRP (Vector) for 45 min. at room temp. Sections were again rinsed three times in PBS and the peroxidase visualized using DAB (GS-1- B4-HRP treated sections 90 were visualized immediately after the lectin incubation). Some sections fi-om each animal were stained with hematoxolin/eosin (H&E) to show general morphology. Results Microglia. Astrocytes, and Vasculature By seven days post-transplantation (pt) all spinal cords had recognizable grafts when stained with H&E, although the size varied considerably from animal to animal. The grafts generally consisted of small islands of undifferentiated cells suspended within and along the edges of large cystic cavities in the lesioned spinal cord. GS-1- Bj allows microglia/macrophages and blood vessels to be visualized in the same sections, and typically stains developing tissue, including transplants, much more strongly than adult tissue (Ashwell et al., 1991; Pennell and Streit, 1997). In every case the grafts were surrounded by numerous macrophages, as evidenced by lectin staining, and contained microglia in various stages of ramification (Fig. 5- la). These microglia were few in number and differed from adult ramified microglia in that they were larger and had fewer processes. There were few blood vessels evident at this early time, and those present were of large diameter and extended into the grafts from the adjoining areas of host spinal cord. At this time the grafts were almost entirely free of GFAP staining, except where blood vessels were present (Fig. 5-2a). Astrocytic processes (Fig. 5-3a) always surrounded developing blood vessels. By 10 days pt, the grafts were generally larger than at seven days, although there was still considerable variability. The grafts were still surrounded by macrophages, but also contained many macrophages, possibly engulfed by the growing transplants as the 91 small islands joined to form larger areas of graft (Fig. 5- lb). Microglia within the grafts were still larger and less ramified than resting adult microglia, and were frequently found near developing blood vessels (Fig. 5-3b). The grafts contained numerous large diameter blood vessels at this time, most still appearing to extend from the periphery of the graft (Fig. 5- lb). GFAP staining was still almost exclusively limited to the outlines of blood vessels, although there were a few isolated astrocytes that appeared to be occupying areas free of vasculature (Fig. 5-2b). By two weeks pt the grafts were uniformly much larger than at ten days, frequently filling the entire center of the spinal cord and eliminating the cystic cavities evident at earlier time points. H&E stained sections showed that the grafts now contained many cells that were phenotypically neuronal, whereas earlier grafts consisted of small, undifferentiated neuro-ectodermal cells (data not shown). The grafts continued to be surrounded and populated by macrophages, and contained numerous microglia, some of which were smaller and more ramified than those seen at seven and ten days (Fig. 5-lc). Blood vessels at this time were much more numerous than at ten days, were spread evenly throughout the grafts, and appeared to be of smaller diameter than those at earlier time points. At this time the grafts contained a great many astrocytes, which were no longer associated exclusively with blood vessels (Fig. 5-2c). The astrocytes were smaller and less darkly stained than those in the surrounding host tissue. By three weeks pt all grafts examined had filled the lesion cavity, smoothly integrating with the host spinal cord tissue. There were significantly fewer macrophages surrounding the grafts at this time, and very few were contained within the graft. In contrast, there were many more microglial cells within the grafts at this time, most of 92 which resembled normal adult ramified microglia with small soma and long, branching processes (Fig. 5- Id). The grafts were filled with evenly distributed blood vessels, but the vessels were of much smaller diameter than at earlier timepoints, and appeared more numerous (Fig. 5- Id). GFAP staining at three weeks p/ revealed that the grafts were filled with astrocytes with a density greater than that at 14 days pt, but still less than that of the surrounding host tissue (Fig. 5-2d). These cells were generally smaller than those in the surrounding tissue, although the astrocytes immediately adjacent to the graft were hypertrophied, probably due to the contusion injury, at every time point examined. At thirty days pt the grafts were of a comparable size to those at twenty-one days. As has been previously reported (Pennell and Streit, 1997), GS-1- B,, staining of microglia and blood vessels was much lighter at this time point than at earlier times, probably due to downregulation of glycoproteins associated with the maturation of the graft. The microglia within the graft had differentiated further, with most of them being indistinguishable from normal resting adult microglia (Fig. 5-le). The density and diameter of blood vessels at 30 days pt appeared to be relatively unchanged from 21 days. GFAP staining showed that astrocytes within the graft were morphologically indistinguishable from and approximately as numerous as the astrocytes in the surrounding host tissue (Fig. 5-2e). The graft/host interface was difficult to discern from the astrocytic staining, indicating minimal scarring and good integration of the transplant with the host spinal cord. Cytokine Immunostaining At seven days/??, bFGF staining was absent in all grafts (Fig. 5-4a), although positive staining was evident around blood vessels in the surrounding host spinal cord. Figure 5-1. Development of microglia and blood vessels within intraspinal syngrafts. (A) 7 day pt islands of graft are surrounded by numerous macrophages. The islands contain many semi-ramified microglia but few blood vessels (arrow). (B) 10 days pi' the grafts contain many macrophages and microglia. Notice the significant increase in blood vessel number at this time. (C) By 14 days pt the blood vessels are more numerous and of smaller diameter. The number of microglia within the grafts has also increased. (D) At 21 days pt the grafts are larger and the blood vessels consist primarily of small capillaries. The density of microglia is comparable to that in the surrounding host spinal cord. (E) By 30 days pt. the staining is much fainter, and although the blood vessels seem even smaller than at 21 days, the number of vessels and microglia seems mostly unchanged. GS-I-B4- HRP. Scale bar=l 10 |im. 94 ^ 7d ■"' )*■/;. ♦ lOd ' M ?- .' .» -. "^ ' 1*^ '*^- ^ "'(^^ ^ ^^■'■ ;v.V 'i "' •E-^ '^1, . 'f^ <^. ^. " i*<^^mf:i Figure 5-1 Figure 5-2. The development of astrocytes within intraspinal syngrafts. (A) 7 days pt the grafts are mostly free of astrocytes, with astrocytic processes only present surrounding blood vessels (arrows). (B) At 10 days/?? the staining pattern is the same as at 7 days, allowing for increased number of blood vessels at this time point. (C) By 14 days pt astrocytes are no longer just surrounding blood vessels, but have spread out to fill the entire graft, although at much lower density than in the surrounding host tissue. (D) At 21 days/?? astrocytes have increased in number. (E) By 30 days/?/, the astrocytic density is comparable to the surrounding host tissue, the cells having increased in size and number over those at 21 days. GFAP. Scale bar=l 10 juim. 96 '. \ • .^ '^^. v;- r>.v. i v.^ I' Figure 5-2 Figure 5-3. Morphological relationship between astrocytes, microglia, and developing blood vessels. (A) At 10 days/7? new blood vessels are covered with astrocytic processes (arrows), leaving the non-vascularized areas of graft bare of GFAP staining. (B) At 14 days pt microglia (arrowhead) and macrophages (arrow) are frequently seen in close apposition to ingrowing vessels. GS-I-B4-HRP. Scale bar=46 |j.m. 98 •J 1, Figure 5-3 Figure 5-4. The expression of basic fibroblast growth factor (bFGF) in developing intraspinal syngrafts. (A, B, C) At 7, 10, and 14 days pt, there is little or no expression of bFGF in the grafts. (D, E) By 21 days/?? and continuing at 30 days pt, however, there is widespread expression of bFGF on neurons. Scale bar=l 10 |j,m (F) Higher power micrograph shows expression of bFGF on neurons and astrocytes (arrows) at 30 days pt. Scale bar=27.5 |j,m. 100 7d lOd c . -V'-V^ ''j ■ '1 ^ ',' (.' .". Figure 5-4 B I4d "V- : ,. 7d r% iOA • •. "' " *' •*'• V. ' <_^;|' ' ; ••' ■ >/ ■. \ ;\ , f' ■ "V - i :■ -^ \\ ' , 4. ^ V s ■ ' . • ' '* "^ ' i^j*. ^ ■ ^ ' \ * * . '■'\ ■', • ^\ B 14cl •■ i', '^ '^ J .- ? ■ i. ^ :s •* >> n!', ; „^,'-t '5'/ ^>\ --'< Figure 5-6 105 TGF-P immunoreactivity was restricted to blood vessels in the graft and in the surrounding lesion cavity (Fig. 5-5a), with widespread staining of neurons and astrocytes, but not blood vessels, in the surrounding host spinal cord. In contrast, virtually every cell in the graft and lesion area was immunoreactive for VEGF, including most if not all grafted cells, macrophages and astrocytes within and surrounding the grafts, and astrocytes throughout the host spinal cord (Fig. 5-6a). At ten days pt, the patterns of bFGF and TGF-P were identical to those at seven days, with no staining and staining of new blood vessels, respectively (Figs. 5-4b, 5-5b). It is interesting to note that the patterns of TGF-p and GFAP immunoreactivity overlap at this time (Figs. 5-5b, 5-2b). Likewise, VEGF staining was very similar to that at seven days (Fig. 5-6b). with the graft staining so darkly in some cases that individual cells were difficult to pick out. At fourteen days/?/ there was no bFGF staining apparent in four out of five transplants (Fig. 5-4c), with weak staining of neurons in the remaining graft. There was no apparent immunoreactivity for TGF-P within the grafts (Fig. 5-5c), although staining of astrocytes in the surrounding host tissue served as a positive control for the efficacy of the antibody. VEGF staining, however, was still abundant, and could now be localized to the cytoplasm of neurons within the graft (Fig. 5-6c), as well as to astrocytes immediately adjacent to the graft within the host spinal cord. By 2 1 days pt the staining patterns had undergone a striking change. There was strong neuronal bFGF immunoreactivity in every graft examined (Fig. 5-4d). In almost 106 identical patterns, there was also strong TGF-P staining in neurons throughout every graft examined (Fig. 5-5d). as well as in some blood vessels. There was still strong VEGF expression (Fig 5-6d). Staining for all three cytokines remained high at 30 days pt (Figs. 5-4e, 5-5e, 5-6e), and was still limited to neurons within the grafts with the exception of bFGF, in which immunostaining was also localized to astrocytes (Fig. 5-4f). Discussion In the present study we have followed the development of glial and vascular elements within developing intraspinal syngrafts during the first month post- transplantation (pt), and attempted to show the cellular sources of three major cytokines/growth factors known to be involved in angiogenesis, i.e. bFGF, TGF-p, and VEGF. We have shown that neovascularization begins at approximately 7 days pt, is well underway at 1 0 days pt, and continues throughout the first month with progressive differentiation of vessels from large diameter to smaller diameter. Microglial infiltration of developing grafts precedes vessel ingrowth, and microglia are frequently associated with immature vessels. Astrocytes are almost exclusively found associated with blood vessels for the first ten days pt, after which they expand out to populate the entire graft by one monlhpt. bFGF immunoreactivity is absent within the grafts until 21 days pt, after which all grafted neurons and a subset of astrocytes express it. TGF-P is expressed only on blood vessels for the first two weeks pt, after which it is expressed by grafted neurons. VEGF, one of the most potent known angiogenic factors, is highly expressed by the grafted cells as early as 7 days pt, along with host macrophages and astrocytes, and 107 continues to be expressed by graft neurons throughout the first month/)/. Together these findings present an interesting picture of intraspinal syngraft development, from a different viewpoint (non-neuronal) than has been previously addressed. Glial and Vascular Development The early migration of microglia into developing syngrafts is very similar to that which has been described in early intrastriatal allografts (Pennell and Streit, 1997). It has been fairly well established at this point that these cells are derived from host microglia/macrophages and are not primarily graft-derived cells (Perry and Lund, 1989; Geny et al., 1995; Pennell and Streit, 1997). It is interesting to note that the microglia within the graft begin to differentiate into ramified cells as early as one week/?/, despite being suspended within a lesion cavity and being surrounded by numerous macrophages, a likely source of pro-inflammatory cytokines. It has been shown that phenotypic brain macrophages can differentiate into ramified microglia in vivo (Pennell and Streit, 1998), indicating that the large pool of macrophages within the lesion cavity is the likely source of the ramified microglia within the grafts at later time points. The development of astrocytes within the grafts is also very interesting. As has been previously shown in other transplant models (Lawrence et al., 1984; Conner and Bernstein, 1987), the only astrocytic presence within the grafts during the first ten days is associated with ingrowing vessels, which are covered by astrocytic processes. Lawrence and colleagues (1984) have shown an association between streamers of new basal lamina from developing vessels and astrocytic processes at the ultrastructural level, and suggest that the astrocytes play a part in directing basal lamina assembly. Whatever their role is. 108 however, the absence of astrocytes with the grafts prior to blood vessel ingrowth suggests that this interaction is induced by the vasculature itself and not the other way around. The striking increase in astrocytic density within the grafts between ten days pt, when there is little GFAP staining, and fourteen days pt, where there is widespread GF AP immunoreactivity, suggests that immature precursor cells within the graft are induced to differentiate into astrocytes at this time. This makes sense since the second weekp/ corresponds temporally with the first postnatal week of development, the primary period of astrocytic differentiation in the CNS (Skoff, 1980). These early astroglia are smaller and less darkly stained than mature astrocytes, and they gradually increase in number and size until at one month pt they are comparable in size and density to those in the surrounding host spinal cord. Conner and Bernstein (1987) reported that graft astrocyte numbers were increased relative to the host spinal cord, but we must remember that in our case the surrounding spinal cord is injured, which probably accounts for the increased number and size of the host glia. It has been hypothesized that transplants placed into the injured spinal cord, rather than having to rely solely on the inductive faculties of the grafted tissue, would be vascularized earlier than those placed into the uninjured CNS due to the presence of diverse angiogenic factors within the lesion cavity associated with the injury itself (Horner et al., 1994). This is supported somewhat by our data, considering the presence of some large diameter vessels seen growing in from the periphery of the grafts at one week (Fig. 5- la). This is in contrast to previous studies in uninjured animals in which one week grafts were almost entirely free of blood vessels (Pennell and Streh, 1997; Rostataing-Rigattiera et al., 1997). The differences were far from striking, however, as 109 the one week grafts were still relatively fi-ee of blood vessels. In addition, the continued development of vasculature within the intraspinal grafts paralleled that previously shown in other suspension graft models (Pennell and Streit. 1997; Rostataing-Rigattiera et al., 1997), indicating that the influence of the lesion environment was minimal at best. Instead, it is probably the metabolic needs of the developing neurons within the rapidly expanding grafts which is the primary inductive stimulus, as has been hypothesized in other models (Horner et al., 1994; Pennell and Streh, 1997). Overall, it would appear that the development of glial and vascular elements within intraspinal syngrafts is not significantly different from said development in suspension grafts placed into other, uninjured areas of the CNS. Angiogenic and Angiostatic Cytokine Production Basic fibroblast growth factor (bFGF), transforming growth factor beta (TGF-P), and vascular endothelial growth factor (VEGF) are all cytokines/growth factors that have been shown to have potent effects regarding angiogenesis (Beck and D'Amore, 1997). It has been hypothesized that astrocytes and microglia play a role in the induction of neovascularization during graft development (Lawrence et al., 1984; Homer et al., 1994; Pennell and Streit, 1997). Thus, these particular cytokines were selected for this study because they are known to be produced by astrocytes (Ballabriga et al., 1997; Flanders et al., 1998; Stone et al., 1995) and microglia/macrophages (Shimojo et al., 1991; Lindholm etal., 1990; Sunderkotter et al., 1994). We have shown that bFGF is not expressed within the grafts until at least two to three weeks after transplantation. This makes it unlikely to be involved in at least the 110 initial stages of neovascularization, although there is continued differentiation of large vessels into smaller capillaries at this time, so a role in vascular development cannot be ruled out. bFGF is potently angiogenic in in vitro and in vivo models of angiogenesis, with effects on endothelial cell proliferation, migration, and differentiation (Potgens et al., 1995), but it's role in actual angiogenesis in vivo is disputed (Horner et al., 1994; Beck and D'Amore, 1997). Because bFGF also has lots of effects not related to angiogenesis, such as causing proliferation of astrocytes, smooth muscle cells, and of course fibroblasts (Potgens et al., 1995), the upregulation of bFGF between 14 and 21 days pt may partially account for the increase in the number of astrocytes within the grafts. Interestingly, bFGF has been shown to induce the differentiation of neurons (Vlodavsky et al., 1991) and to greatly increase the size of neural transplants (Giacobini et al., 1991), so it is possible that it serves as a neurotrophin within the intraspinal grafts. The expression of TGF-P within the developing grafts is quite illuminating. It is expressed only in new vessels in and around the early (7 and 10 day) transplants, in a pattern strikingly similar to the GFAP staining pattern within the grafts at these times. It is known that TGF-P is both produced by astrocytes (Flanders et al., 1998) and has significant effects upon them. TGF-P inhibits the growth of astrocytes in vitro (Flanders et al., 1993) and, more importantly, induces the deposition of extracellular matrix (ECM) by astrocytes, a crucial step in angiogenesis (Baghdessarian et al., 1993; Flanders et al., 1993). These findings correspond well with Lawrence and colleagues (1984) study which showed astroglial processes intimately associated with newly formed vascular basal lamina in early cortical transplants. TGF-p. in contrast to bFGF and VEGF, actually Ill inhibits endothelial cell proliferation, and is thought to be involved in vascular differentiation (Beck and D'Amore. 1997). This would make sense considering that the onset of TGF-P expression by grafted neurons (21 days pt) corresponds approximately with the end of neovascularization within the grafts. Last but not least, we found that VEGF expression was widespread within the grafts at every time point examined. VEGF is generally considered the most important angiogenic cytokine, both because of its effects on endothelial cell proliferation, migration, and differentiation, and because of the specificity of its actions for vascular endothelial cells in particular (Ferrara et al., 1992; Potgens et al., 1995; Beck and D'Amore. 1997). It is known that VEGF production by astrocytes is crucial to angiogenesis within the CNS following brain injury (Papavassiliou et al., 1997) and to vasculogenesis within the developing retina (Stone et al., 1995). VEGF production by macrophages is also very important to wound healing-related angiogenesis (Sunderkotter et al., 1994), another consideration in our grafts placed into a lesioned environment. At early time points (7 and 1 0 days pt) there was expression of VEGF by macrophages and astrocytes in addition to the grafted cell themselves, suggesting a combination of inflammatory and normal developmentally derived VEGF. VEGF expression after that, however, was limited to neurons within the graft and astrocytes and neurons in the surrounding host spinal cord. It is known that VEGF is expressed at high levels during development and then maintained at low levels even in the adult CNS (Breier et al., 1992), so the continued expression of VEGF by the grafts even after the apparent end of angiogenesis should not be surprising. 112 What is surprising is the relative lack of support for the hypothesis that glial cells are key components in graft neovascularization. Although astrocytes may be playing a role in the deposition and assembly of ECM associated with endothelial cell migration, it remains to be shown conclusively. Furthermore, while macrophages and astrocytes surrounding the grafts do express high levels of VEGF at early time points, this expression is not reflected by glial cells within the grafts and is probably more associated with the wound healing response of the injured spinal cord than it is induced by the graft. Intraspinal graft neovascularization proceeds more or less along the same time scale as in grafts not placed into lesioned environments, downplaying the significance of inflammatory angiogenesis. Instead, angiogenic and angiomodulatoiy cytokine production by the grafted neurons themselves appears to play the largest role in vascular neogenesis. This does not, however, rule out important roles for microglia and astrocytes in new blood vessel growth. For example, there are a number of directly and indirectly angiogenic molecules potentially produced by these cells which have not been examined, such as platelet derived growth factor (PDGF), macrophage colony stimulating factor (M-CSF), interleukin-1 (IL-1), tumor necrosis factor-alpha (TNF-a), and countless others. Microglia have been shown to produce ECM components which both promote and inhibit angiogenesis, such as laminin and thrombospondin (Masuda-Nakagawa et al., 1993; Chamak et al., 1994). Microglia have also been shown to produce constituents of the plasminogen activator pathway (Nakajima et al., 1992) which are responsible for both for activation of latent TGF-P (Schultz-Cherry and Murphy-Ulrich, 1993) and for cleavage of membrane and matrix-bound bFGF and VEGF into soluble forms (Potgens et al., 1995). So we see that there are still a number of 113 potential explanations for the intriguing morphological association between astrocytes, microglia, and developing blood vessels within neural transplants. Elucidation of these mechanisms, however, must wait for future studies. CHAPTER 6 CONCLUSIONS The research in this dissertation consists primarily of three separate projects which are loosely related in that they all have to do with both microglial cells and transplantation. Chapter 2 described a method for the reliable tracing of cultured microglial cells after injection into the adult rat brain. Chapter 3 attempted to test the hypothesis that microglia are responsible for graft rejection. Finally, in chapters 4 and 5, the relationship between microglia and developing graft vasculature was investigated. In this final chapter I shall attempt to summarize what was learned from these projects, propose new directions for research, and point out the common threads which tie these seemingly disparate ideas together. The Tracing of Transplanted Microglia There has been a great deal of recent interest in the transplantation of purified microglia/macrophages into the injured CNS (Rabchevsky and Streit, 1996; Lazarov- Speigler et al., 1996; Prewitt et al., 1997; Franzen et al., 1998). However, these studies all share the common problem of being unable to reliably trace the injected cells and differentiate them from the endogenous host microglia/macrophages. In chapter 2, Fluoro- Gold (FG) prelabeled microglia were traced for up to two weeks after injection into the 114 115 adult rat corpus callosum and lateral ventricle. They remained brightly labeled in this time, with little or no background labeling of endogenous cells. Some of the injected cells differentiated from the macrophage phenotype typical of cultured microglia into ramified cells morphologically indistinguishable from resting in vivo microglia. These results indicate that this method would be useful for the reliable tracing of transplanted microglia/macrophages in future studies. Furthermore, control injections of free FG resulted in widespread labeling of neurons and perivascular cells within the injected hemisphere, but no labeling of endogenous microglia. This would indicate that resting microglia, in contrast to phagocytic microglia, are not actively pino- or phagocytic. Thus microglia represent a population of facultative macrophages within the CNS, while the role of constitutive phagocytes goes to the perivascular cells. This concept will be addressed again later in this discussion. Single-cell Suspension Grafts, Microglial Depletion, and Rejection Using a xenograft model , mouse to rat, and what we thought was a well characterized model of allograft rejection in the spinal cord, the ACI to Wistar-Furth cross (Theele and Reier, 1996), we attempted to test the hypothesis that microglia induce immunologic rejection by removing the microglia from the grafts prior to transplantation. Approximately 98% of microglia and endothelial cells, the primary sources of major histocompatability complex (MHC) antigens present within the prenatal CNS, were removed from single-cell suspensions of E14 mouse and ACI rat CNS prior to injection 116 into the intact striatum or sub-acutely injured spinal cord of Wistar-Furth rats. In the xenografted animals, all the grafts were rejected by 35 days regardless of the presence or absence of microglia. This finding was consistent with what was known about xenograft rejection, and the rejection itself followed a well defined pattern. In the allografted animals, however, we found out that our "well-characterized model" was not so well- characterized after all. We expected the non-depleted grafts to begin rejecfing at about 10- 14 days post-transplantation (pt), at the time of onset of donor MHC II expression. Instead, none of the allografts in either group showed signs of rejecfion in the first three weeks;?/. After that, at 30 days pt, ALL of the grafts were rejecting, and were all destroyed by 45 days pt. What was going on here? The answer is that we had created a new model of intraspinal transplantation. No one had previously investigated single-cell suspension grafts placed into the sub-acutely injured spinal cord, and it was unreasonable to expect graft/host interactions to be the same as they were in other models. Donor MHC II expression began at the same time as in previous models, at about 14 days/?/, but in this case it did not precipitate an immune response from the host. The suspension grafts contained very few (-1%) microglia and endothelial cells even in the non-depleted grafts, and thus the suspensions were almost purely neuroectodermally derived. The whole-piece and slurry suspensions from former studies almost certainly contained a greater amount of vascular and menigial elements, as well as larger amounts of cellular debris that could contribute to the antigenic load of the graft. Thus, even without depleting the microglia, we had removed the vast majority of MHC expressing cells from the grafts prior to injection. In essence, we had succeeded in our goal to interrupt the immune induction stage of graft rejection by eliminating the 117 direct pathway of antigen presentation (see chapter 1). The question that remains, however, is why did the grafts reject at all? And why so consistently between 21 and 30 days pt? Obviously the lack of variability in this window of time indicates a discrete event taking place. However, explanations at this point remain speculative until further studies are done. Although the direct presentation pathway is thought to represent the primary inductive mechanism for allograft rejection (VanBuskirk et al., 1994), it is clear that indirect presentation of foreign antigen can be sufficient in and of itself to induce an immune response (Steinmuller, 1983; LaRosa and Talmage, 1985). In essence, this means that donor-derived antigen presenting cells (APCs) are not always necessary to induce allograft rejection. Instead. APCs from the host can migrate into the graft, process donor (foreign) antigen, and present it to host T-cells. It had been generally thought that the microglial cell represents the CNS's endogenous APC (Frei et al., 1987; Streit et al., 1989). We know that host microglia are present within the graft as early as 7 days pt, presumably providing a large pool of host APCs within the graft. Why then do they not induce rejection through the indirect pathway until four weeks/?/? The answer may be that microglia do not ftinction as APCs at all, or at least not very effectively. We have shown that cultured microglia, even after activation with IFN-y and LPS, do not stimulate allogeneic T-cell proliferation. This confirms previous studies which indicate that microglia are not functional APCs (Ford et al., 1995; Carson, 1998), and that it is instead the perivascular cells, i. e. macrophages within the vascular wall, that actually represent the functional APCs of the adult CNS. This, I believe, is the missing piece of the puzzle. The next logical 118 question is, when during the development of vasculature within grafts do host perivascular cells appear? I would hypothesize that they appear between 21 and 30 days/?A at the time of the onset of rejection. This would be easy to determine, using the ED-2 antibody to selectively stain perivascular cells within developing grafts, and represents perhaps the most important follow-up study to the current work. If a temporal correlation were established between perivascular cell appearance and rejection, mixed lymphocyte reactions could be carried out using FACS-sorted perivascular cells to see if they can present alloantigen to T- cells. This ability could then be compared to that of microglia. In addition, to definitively establish whether or not MHC antigens are crucial to graft rejection, transplants should be carried out in animals differing only at the MHC I and/or II loci, and not at both major and minor histocompatability loci, such as in the animals used in this study. The significance of this research is twofold. First, this new model of transplantafion should have an impact on our understanding of the need to immunosuppress allograft recipients. There would appear to be no real need to administer immunosuppression during the first two to three weeks after transplantation, during the patient's most vulnerable period to post-surgical infections. If immune recognition occurs between three and four weeks pt, this would then be the best period to target for immunosuppressive therapies. Although, since human tissue is vascularized so much slower than is rat tissue, human transplants may not be at risk of rejection for even longer periods of time. While rat tissue is completely vascularized by one month/?/, human graft neovascularization is barely beginning by that time (Geny et al, 1994). The second major significant addition that this research provides is in the area of microglial immunocompetence (see chapter 1 for review). We showed in chapter 2 that 119 brain macrophages are induced to differentiate into ramified microglia which are not actively phagocytic. We observed in chapter 3 that microglia likely do not act as functional APCs. These findings are in agreement with the recent literature that describes microglia as immature immune cells (Carson, 1998), which can actually induce apoptosis in T-cells rather than activate them (Ford et al., 1996). In light of these studies, we must conclude that microglia are not the immunocompetent cells of the CNS. That honor falls to perivascular cells, which have been conclusively shown to be functional APCs (Ford et al., 1995). Graft Neovascularization The vascular networks within neural transplants are vitally important both to graft survival and to immunologic rejection. In chapters 4 and 5 we followed the development of microglia and blood vessels within developing striatal and intraspinal grafts. In both cases, microglia infiltrated the grafts prior to the onset of neovascularization and were infimately associated with ingrowing vessels during the early time points p/. Blood vessels began to form between seven and ten days pt and increased in number while decreasing in diameter until completing their development by approximately 30 daysp?. Microglial differentiation from macrophage-like morphology into ramified, adult microglia took approximately the same amount of time. This temporal correlation between microglial and vascular development led to the hypothesis that microglia play an important role in graft vascular neogenesis. This morphological correlation in mind, we hypothesized that microglia, and possibly astrocytes, produced angiogenic factors such as the cytokines bFGF, TGF-P, and 120 VEGF to regulate neovascularization. TGF-P and bFGF are produced by both types of glial cells, while VEGF is known to be produced by astrocytes and peripheral macrophages. All three are generally considered to be among the primary players in angiogenesis. Thus, we examined the expression of these cytokines by immunohistochemistry during the first month /?r in intraspinal syngrafts. What we found, however, was that bFGF was only expressed after 21 days pt and primarily in neurons. Its role in graft vascular development, if any, remains unclear. TGF-p was expressed by ingrowing blood vessels at early time points and then by neurons at later timepoints, perhaps as a differentiation factor. The early expression may have been a signal to surrounding astrocytes to aid in extracellular matrix (ECM) deposition, although this is only speculation. Finally, VEGF expression was widespread at every time point examined, both in the host spinal cord and in the grafted neurons. These patterns of expression would indicate that the grafted neurons are the primary inducers and regulators of neovascularization, although other roles for glial cells are not ruled out. They may be involved in ECM degradation and/or deposition, for instance. Microglia in particular are known to produce matrix metalloproteases such as MMP2 (Yamada et al., 1995) and can secrete ECM components such as laminin and thrombospondin (Masuda-Nakagawa et al., 1993; Chamak et al., 1994). Glial cells may also produce angiogenic factors other than those examined in this study, such as platelet derived growth factor (PDGF), macrophage colony stimulating factor (M-CSF), interleukin-1 (IL- 1), and tumor necrosis factor-alpha (TNF-a). These studies have been valuable first steps into the understanding of glial and vascular development in neural transplants. It will be a long time before every element 121 involved in graft development is understood, and our investigations have raised more questions than they have answered. For example, what signals induce the migration of microglia into developing grafts, and what induces their differentiation? What is the nature of the interaction between astrocytes and blood vessels in early graft development? What factors, if any, do astrocytes and microglia manufacture that influence angiogenesis? If microglia do not influence angiogenesis, do they play a neurotrophic role in supporting transplant growth, as has been hypothesized (Prewitt et al., 1997)? We can test these questions either by examining the expression of various molecules through immunohistochemistry or in situ hybridization, or by using animals in which the molecules in question have been knocked-out or inhibited. Differential display RT-PCR or subtractive hybridization could be used to identify mRNA species upregulated during early angiogenesis, which could be cloned and identified. Microglia can also be inhibited in vivo using microglia/macrophage inhibitory factor (MIF; Thanos and Mey, 1995). Experiments could be performed in which grafts are treated with MIF to examine effects on angiogenesis and microglial infiltration. The studies in this dissertation have just begun to investigate potential functional roles for microglia in neural tissue transplantation. The fact that most of our hypotheses were either not supported or outright disproven by our results should not be dismaying, for these are really just first steps into a world of research previously dominated by neuronal and behavioral studies. General, descriptive studies must be performed before more specific hypotheses can be made and tested, and we did manage to carve some new ground. A new model of neural transplantation was developed and characterized, light was cast on some long misunderstood ideas about microglial immunocompetence, and some interesting 122 correlations can now be made between graft vascularization and immunologic rejection (see Table 6-1). These investigations should be thought of as pilot studies, starting points for many future studies. Besides, if there is anything that this author has learned over the last five years, it is that no question in science is ever really answered. 123 -^ * * c^ p« T1 H^* 0) o — 3 3 C/5 ■a o 5' (T) c (— ^ ^71 O) ::::^ §■ -P^ U) M ^^ ^^ -J ^ '^ L^l o ^ -t^ O ^ a: ,C5 V si- O § ? £ ^ 1 3 p "■ '£ O a. p 3 S ^ fB p -►) 2 -^ 3 v: _. ft c _ o 3 v: n 3 3 p CL 1 3 3 3 P 1 ft p P ?0 i c era zi. o 3 f^ tt_ s o 3 E "0 3 3- O. P > ■a ■o 3 X ^ o 5 3 ft r-i "O p * 3 o p ft < O cfo' O 3 » t« 3- -. ft O P 3 ~ 3 3 3' 3 -t, ^ p o o -n 3 — ■ o 3- ^ a. ft c 3 ft 3 =i ?2 3 O 1 i P p o ft p 3 ft 3 I 3 n p _- D- O 3 O 4 a. TT 1^. ' — 1 "" ?a ft_ p O ft 3 X T3 2 X to' g; o n n TO 1 a. p v: o' 3 5' 3 ft " O 3 3- I S> 2 2 o g O 3 G. 03 oq' =• I s. T3 p » i O 3" S.n3 5' 3 2. 3 » Co rT ca — P p c^ O f^ 3" K ft 5 3 3 c 3 ft < ft_ o p TD 3 3 p 5' ;:? — . ft O P 3 ~ 3" 1 ^ 3 fU -h D. O 3 O 3 U5 < O ft 5' o o « 7 fT 2 3 O a 03 3 < - • ft_ |S CI. o 3 o ^ o 2 "— ' O O ^ ^ ^ X n = 2 z 2 ^ — ° o t" 3 o X n ^ o3' 2 3 ^ tio' 3 -a ft 2 o OQ -a g < 5' D 3 & 3- P 50 ^ 3 3 ^ p p » <• 2. S o 5; q rv ft P o 3; 3) era' Co 03 o c 03 o ft n < o c_ 01 c 3 c ular density near i mum; **Still no h ascular cells withi 3 p_ fT o 3 ft cr ft 5" Si < P D. ^ 1 ft p 2 ft o c < ft ft J< ation of host mien little or no vascu s point 5=1 il ft 3 re 3 O ^ 3- "^ ft ft v; ft p fJQ ft ^ ere g' p- o 3 3 O REFERENCES Abbas, A.K., Lichtman, A.H., and Pobar, J.S. 1994. Cellular and molecular immunology. W.B. Saunders Co., Philadelphia. Anderson, D.K., Rowland, D.R., and Reier, P.J. 1995. Fetal neural grafts and repair of the injured spinal cord. Brain Pathology. 5: 451-457. Andersson, C, Tyrell, M., and Brunso-Bechtold, J. 1993. Tranplantation of cultured type 1 astrocyte cell suspensions into young, adult and aged rat cortex: cell migration and survival. Int. J. Dev. Neurosci., 1 1 :555-568. Ard, M.D.. Bunge, M.B., Wood, P.M., Schachner, M., and Bunge, R.P. 1991. Retinal neurite growth on astrocytes is not modified by extracellular matrix, anti-Ll antibody, or oligodendrocytes. Glia. 4: 70-82. Ashwell, K.W.S., Hollander, H., Streit, W.J., and Stone, J. 1989. The appearance and distribution of microglia in the developing retina of the rat. Visual Neuro.science 2:437-448. Assouline, J.G., Bosch, P., Lim, R., Kim, I.S., Jensen, R., and Pantagis, N.J. 1987. Rat astrocytes and Scwann cells in culture synthesize nerve growth factor-like neurite- promoting factors. Dev. Brain Res. 31 : 103-1 18. Baghdessarian, D.Toru-Delbauffe, D., Gavaret, J.M., and Pieer, M. 1993. Effects of transforming growth factor-pl on the extracellualr matrix and cytoskeleton of cultured astrocytes. Glia 7:193-202. Ballabriga, J., Pozas, E., Planas, A.M., and Ferrer, I. 1997. bFGF and FGFR-3 immunoreactivity in the rat brain following systemic kainic acid administration at convulsant doses: localization of bFGF and FGFR-3 in reactive astrocytes, and FGFR-3 in reactive microglia. Brain Res. 752(1 -2):3 15-3 18. Barker, C.F. and Billingham, R.E. 1977. Immunologically privileged sites. Adv. Immunol. 25: 1-54. Bartlett, P.F., Rosenfeld, J., Bailey, K.A., Cheesman, H.. Harvey, A.R., and Ken\ R.S.C. 1990. Allograft rejection overcome by immunoselection of neuronal precursor cells. 124 125 In: Progress in Brain Research (S.B. Dunnett and S.J. Richards, eds.) Vol. 82: pp. 153- 160. Elsevier, Amsterdam. Beck, L. and D'Amore, P.A. 1997. Vascular development: cellular and molecular regulation. FASEBJ. 11: 365-373. Bjorklund, A. 1991. Neural transplantation- An experimental tool with clinical possibilities. Trends Neurosci. 14: 319-322. Blakemore, W.F., Crang, A.J., and Patterson, R.C. 1987. Schwann cell remyelination of CNS axons following injection of cultures of CNS cells into areas of persistent demyelination. A^e«ro5c/e«ce Lert. 77: 15-19. Bocchini, V., Artauh, J.C, Rebel, G., Dreyfus, H., and Massarelli, R. 1988. Phagocytosis of polystyrene latex beads by rat brain microglia cell cuhures is increased by treatment with gangliosides. Dev. Neurosci., 10:270-276. Boje, K.M., Arora, P.K. 1992. Microglial-produced nitric oxide and reactive nitrogen oxides mediate neuronal cell death. Brain Res. 587: 250-256. Bradbury, E.J., Kershaw, T.R., Marchbanks, R.M., and Sinden, J.D. 1995. Astrocyte transplants alleviate lesion induced memory deficits independently of cholinergic recovery. Neuroscience. 65(4): 955-972. Bradbury, M.W. and Westrop, R.J. 1983. Factors influencing exit of substances from cerebrospinal fluid into deep cervical lymph of the rabbit. J. Physiol. (Land.) 339: 519-534. Breier, G., Albrecht, U., Sterrer, S., and Risau, W. 1992. Expression of vascular endothelial growth factor during embryonic angiogenesis and endothelial cell differentiation. Development 1 14: 521-532. Broadwell, R.D., Charlton, H.M., Ebert, P., Hickey, W.F., Villegas J.C, Wolf, A.L. 1990. Angiogenesis and the blood-brain barrier in solid and dissociated cell grafts within the CNS. In: Progress in Brain Research (S.B. Dunnett and S.J. Richards, eds.) Vol. 82: pp. 95-109. Elsevier, Amsterdam. Brundin, P., Strecker, R.E., Widner, H., Clarke, D.J., Nilsson, O.G., Astedt, B., Lindvall, O., and Bjorklund, A. 1988. Human fetal dopamine neurons grafted in a rat model of Parkinson's disease: immunological aspects, spontaneous and drug-induced behaviour, and dopamine release. Exp. Brain Res. 70(1): 192-208. Bunge, R.P., Puckett, W.R., Becerra, J.L., Marcillo, A.. Quencer, R.M. 1993. Observations on the pathology of human spinal cord injury. A review and classification of 126 22 new cases with details from a case of chronic cord compression with extensive focal demyelination. Adv. Neurol. 59:75-89. Carson. M.J., Reilly, C.R., Sutcliffe, J.G., and Lo, D. 1998. Mature microglia resemble immature antigen-presenting cells. Glia. 22: 72-85. Chamak B., Morandi, V., and Mallat M. 1 994. Brain macrophages stimulate neurite growth and regeneration by secreting thrombospondin. J. Neurosci. Res. 38: 221-233. Clark, R.A., Stone, R.D., Leung, D.R.K., Silver, I., Hohn, D.C., Hunt, T.K. 1976 . Role of macrophages in wound healing. Surg. Forum 27:17-18. Colton, C.A., Gilbert, D.L. 1993. Microglia, an in vivo source of reactive oxygen species in the brain. Adv. Neurol. 59: 321-326. Connor, J.R. and Bernstein, J.J. 1987. Astrocytes in rat fetal cerebral cortical homografts following implantation into aduU rat spinal cord. Brain Res. 409(1): 62-70. Crang, A.J. and Blakemore, W.F. 1991. Remyelination of demyelinated rat axons by transplanted mouse oligodendrocytes. Glia. 4: 305-313. Crews, L.L. and Wigston, D.J. 1990. The dependence of motoneurons on their target muscle during postnatal development of the mouse. J. Neurosci., 10:1643-1653. Cross A.H., Cannella, B., Brosnan, C.F., and Raine C.S. 1990. Homing to central nervous system vasculature by antigen specific lymphocytes: 1. Localization of 14C-labeled cells during acute, chronic, and relapsing experimental allergic encephalitis. Lab. Invest. 63: 162-170. Czech, K.A., Ryan, J.W., Sagen, J., and Pappas, G.D. 1997. The influence of xenotransplant immunogenicity and immunosuppression on host MHC expression in the rat CNS. Exp. Neurol 147(1): 66-83. Davies, S.J.A., Fitch, M.T., Memberg, S.P., Hall, A.K., Raisman, G., and Silver J. 1997. Regeneration of adult axons in white matter tracts of the central nervous system. Nature. 390: 680-683. De Simone, R., Giampaolo, A., Giometto, B., Gallo, P., Levi, G., Peschle, C, and Aloisi, F. 1995. The costimulatory molecule B7 is expressed on human microglia in culture and in multiple sclerosis acute lesions. J. Neuropathol. Exp. Neurol. 54: 175-187. Duan, W.M., Brundin, P., Grasbon-Frodl, E.M., and Widner, H. 1996. Methylprednisolone prevents rejection of intrastriatal grafts of xenogeneic embryonic neural tissue in adult rats. Brain Res. 712: 199-212. 127 Duan, W.M., Widner, H., Bmndin, P. 1995. Temporal pattern of host responses against intrastriatal grafts of syngeneic, allogeneic, or xenogeneic embryonic neuronal tissue in rats. Exp. Brain Res. 104: 227-242. Duan, W.M., Widner, H., Frodl, E.M., and Brundin, P. 1995. Immune reactions following systemic immunization prior or subsequent to intrastriatal transplantation of allogeneic mesencephalic tissue in adult rats. Neuroscience. 64 (3): 629-641. Dunnett, S.B. and Richards, S.J. 1990. Neural transplantation: From molecular basis to clinical applications. In: Progress in Brain Research (S.B. Dunnett and S.J. Richards, eds.), Vol. 82. Elsevier, Amsterdam. Elkabes, S., Diciccobloom, E.M., and Black, I.B. 1996. Brain microglia macrophages express neurotrophins that selectively regulate microglial proliferation and ftinction. J Neurosci., 16:2508-2521. Evans, R. 1977. Effect of X-irradiation on host-cell infiltration and growth of a murine fibrosarcoma. Br. J. Cancer 35:557-566. Falci, S., Holtz, A., Akesson, E., Azizi, M., Ertzgaard, P., Hultling, C, Kjaedgaard. A., Levi, R., Ringden, O., Westgren, M., Lammertse, D., and Seiger, A. 1997. Obliteration of a posttraumatic spinal cord cyst with solid human embryonic spinal cord grafts: first clinical attempt. J. Neurotrauma. 14(1 1): 875-884. Fallon, J.R. 1985. Peripheral outgrowth of central nervous system neurites on astrocytes and Schwann cells as compared to non-glial cells in vitro. J. Cell. Biol. 100: 198- 207. Ferrara, N., Houck, K., Jakeman, L., and Leung, D.W. 1992. Molecular and biological properties of the vascular endothelial growth factor family of proteins. Endocrine Reviews U{'^): 18-32. Fiandaca, M.S. 1991. Brain grafting for Parkinson's disease- Experimental, clinical, and immunological considerations. Transplantation 51 :549-556. Finsen, B.R., Sorensen, T., Castellano, B., Pedersen, E.B., and Zimmer, J. 1991. Leukocyte infiltration and glial reactions in xenografts of mouse brain tissue undergoing rejection in the adult rat brain. A light and electron microscopical immunocytochemical study. J Neiiroimmimol. 32:159-183. Flanders, K.C., Ludecke, G., Renzig, J., Hamm, C, Cissel, D.S., and Unsicker, K. 1993. Effect of TGF-Ps and bFGF on astroglial cell growth and gene expression in vitro. Mol Cell Neurosci 4: 406-417. 128 Flanders, K.C., Ren, R.F., and Lippa, C.F. 1998. Transforming growth factor-Ps in neurodegenerative disease. Prog. Neurobiol. 54: 76-85. Folkman, J. 1 990. What is the evidence that tumors are angiogenesis dependent? J. Natl. Cancer Biol. 3:65-71. Fontana, A., Fierz, W., and Wekerle, H. 1984. Astrocytes present myelin basic protein to encephalitogenic T-cell lines. Nature 307:273-276. Ford, A.L., Foulcher, E., Lemckert, F.A., and Sedgwick, J.D. 1996. Microglia induce CD4 T lymphocyte final effector function and death. J. Exp. Med 184: 1737-1745. Ford, A.L., Goodsall, A.L., Hickey, W.F., and Sedgwick, J.D. 1995. Normal adult ramified microglia separated from other central nervous system macrophages by flow cytometric sorting. J. Immunol. 154: 4309-4321. Franzen, R., Schoenen, J., Leprince, P., Joosten, E., Moonen, G., and Martin, D. 1998. Effects of macrophage transplantation in the injured adult rat spinal cord: a combined immunocytochemical and biochemical study. J. Neuroscience Res. 51: 316-327. Freed, C.R., Breeze, R.E., Rosenberg, N.L., Schenck, S.A. 1993. Embryonic dopamine cell implants as a treatment for the second phase of Parkinson's disease. Adv. Neurol. 60:72\-72S. Freed, C.R., Breeze, R.E., Rosenberg, N.L., Schneck, S.A., Kriek, E., Qi. J., Lone, T., Zhang, Y., Snyder, J.A., Wells, T.H., Ranig, L.O., Thompson, L., Mazziotta, J.C, Huang, S.C, Grafton, S.T., Brooks, D., Sawle, G., Schroter. G., and Ansari, A.A. 1992. Survival of transplanted fetal dopamine cells and neurologic improvement 12 to 46 months after transplantation for Parkinson's disease. New Eng. J. Med. 327(22): 1549-1 555. Freeman, T.B., Sanberg, P.R., Nauert, G.M., Boss, B.D., Spector, D., Olanow, C.W., and Kordower, S.H. 1995b. The influence of donor age on the survival of solid and suspension intraparenchymal human embryonic nigral grafts. Cell Transplant 4(1): 141- 154. Frei, K., Siepl, C., Groscurth, P., Bodmer, S., Schwerdel, C, and Fontana, A. 1987. Antigen presentation and tumor cytotoxicity by interferon-gamma-treated microglial cells. Eur. J. Immunol. 17:1271-1278. Gage, F.H. and Buzsaki, G. 1989. CNS grafting: Potential mechanisms of action. Neural regeneration and transplantation. Alan R. Liss, Inc., New York. Gash, D.N. and Sladek, J.R. 1988. Tranplantation into the mammalian CNS. In: 129 Progress in Brain Research (D.N. Gash, and J.R. Sladek, eds.), Vol. 78. Elsevier, Amsterdam. Geny, C, Naimi-Sadaoui, S., Belkadi, A.E.M., Jeny, R., Kammoun, M., and Peschanski. M. 1995. Microglial chimerism in human xenografts to the rat brain. Brain Res. Bull. 38(4): 383-391; 1995. Geny. C. Naimi-Sadaoui, S., Jeny, R., Belkadi. A.E.M.. Juliano, S.L., Peschanski, M. 1994. Long-term delayed vascularization of human neural transplants to the rat brain. J. Neurosci. 14(12): 7553-7562. Giacobini, M.M., Hoffer, B.J., Zerbe, G., and Olsen, L. 1991 . Acidic and basic fibroblast growth factors augment growth of fetal brain tissue grafts. Exp. Brain Res. 86(1): 73-81. Giulian, D. 1995. Microglia and neuronal dysfunction. In: Neuroglia (H. Kettenmann and B.R. Ransom, eds.), pp.67 1-684. Oxford University Press, New York. Giulian. D. and Baker, T.J. 1986. Characterization of ameboid microglia isolated from developing mammalian brain. J. Neurosci., 6:2163-2178. Goldberg, W.J. and Bernstein, J.J. 1987. Transplant-derived astrocytes migrate into host lumbar and cervical spinal cord after implantation of E14 fetal cortex into the adult spinal cord. J. Neurosci. Res., 17:391-403. Gout, O. and Dubois-Dalco, M. 1993. Directed migration of transplanted glial cells towards a spinal cord demyelinating lesion. Int. J. Devi. Neuroscience. 1 1(5): 613-623. Graeber, M.B. and Streit, W.J. 1990. Microglia: Immune network in the CNS. Brain Pathol 1:2-5. Graeber M.B.. Streit W.J., Biiringer, D., Sparks, D.L., and Kreutzberg, G.W. 1992. Ultrastructural localization of major histocompatability complex (MHC) Class 11 positive perivascular cells in histologically normal human brain. J. Neuropath. Exp. Neurol. 51(3): 303-311. Graeber, M.B., Tetzlaff, W., Streit, W.J., and Kreutzberg, G.W. 1988. Microglial cells but not astrocytes undergo mitosis following rat facial nerve axotomy. Neurosci. Lett., 85:317-321. Greenburg, G.B., Hunt, T.K. 1978. The proliferative response in vitro of vascular endothelial and smooth muscle cells exposed to wound fluids and macrophages. J. Cell Physiol. 97:353-360. 130 Griffin, J.W., George, R., and Ho, T. 1993. Macrophage systems in peripheral nerves. A review. J. Neuropath. Exp. Neurol. 52(6): 553-560. Gruner. J. A. 1992. A monitored contusion model of spinal cord injury in the rat. J. Neurotrauma. 9: 123-128. Guenard, V., Xu, X.M., and Bunge, M.B. 1993. The use of Schwann cell transplantation to foster central nervous system repair. Sem. Neurosci. 5: 401-41 1 . Harling-Berg, C., Knopf, P.M., Merriam, J., and Cserr, H.F. 1989. Role of cervical lymph nodes in the systemic humoral immune response to human serum albumin microinfusion into rat cerebrospinal fluid. J. Neuroimmunol. 25: 185-193. Hermanns, S., Wunderlich, G., Rosenbaum, C, Hanemann, CO., Muller, H.W., and Stichel, C.C. 1997. Lack of immune responses to immediate or delayed implanted allogeneic and xenogeneic Schwann cell suspensions. Glia. 21: 299-314. Hickey, W.F. and Kimura, H. 1988. Perivascular microglial cells of the CNS are bone marrow-derived and present antigen in vivo. Science 239:290-292. Hirota, T., Hirose, H., Iwata, H., Kanetake, K., Murakawa. S., Sasaki, E., Takagi, H., Bando, M., Hamaoka, T., and Fujiwara, H. 1997. Direct recognition of rat MHC antigens on rat antigen-presenting cells by mouse CD4^ and CD8* T cells and establishment of T cell clones exhibiting a direct recognition pathway. Transplantation. 63 (5): 705-710. Homer, P.J.. Popovich, P.G., Mullin, B.B., and Stokes, B.T. 1996. A quantitative spacial analysis of the blood spinal cord barrier. II. Permeability after intraspinal fetal transplantation. Exp. Neurol. 142(2): 226-243. Homer, P.J., Popovich, P.G., Reier, P.J., and Stokes, B.T. 1994. Fetal spinal transplant vascularity: metabolic and immunologic mechanisms. In: Neural Transplantation, CNS Neuronal Injury, and Regeneration (J. Marwah, H. Teitelbaum, and K.N. Prasad, eds.), pp. 1 19-140. CRC Press, Inc., Boca Raton. Homer, P. J., Reier, P.J., and Stokes, B.T. 1996. Quantitative analysis of vascularization and cytochrome oxidase following fetal transplantation in the contused rat spinal cord. J. Comp Neurol. 364(4): 690-703. Houck, K.A., Ferrara, N., Winer, J., and Leung, D.W. 1991. The vascular endothelial growth factor family: identification of a fourth molecular species and characterization of altemative splicing of mRNA. Mol. Endocrinol. 5: 1806. Hurley, S. D., Streit, W.J. 1996. MicrogUa and the mononuclear phagocyte system. In: Topical Issues of Microglial Research (E.A. Ling and C.K. Tan, eds.), pp. 1-19. Singapore Neuroscience Association, Singapore. 131 Iruela-Arispe, M.L., and Dvorak, H.F. 1997. Angiogenesis: a dynamic balance of stimulators and inhibitors. Thrombosis and Hemostasis. 78(1): 672-677. Kaufman, C.L., Gaines, B.A., and Ildstad, S.T. 1995. Xenotransplantation. Annii. Rev. Immunol. 13: 339-367. Kida, S., Steart, P.V., Zhang, E.T., and Weller, R.O. 1993. Perivascular cells act as scavengers in the cerebral perivascular spaces and remain distinct from pericytes, microglia and macrophages. Acta Neuropathol. Beri, 85: 646-652. Kiefer, R., Lindholm, D. and Kreutzberg, G.W. 1993. lnterleukin-6 and transforming growth factor-Bl mRNAs are induced in rat facial nucleus following motoneuron axotomy. Eur J Neurosci 5:775-781. Konno, H., Yamamoto, T., Iwasaki, Y., Suzuki, H., Saito, T. and Terunuma, H. 1989. Wallerian degeneration induces la-antigen expression in the rat brain. J Neuroimmunol 25:151-159. Kordower, J.H., S. Styren, M. Clarke, S.T. DeKosky, C.W. Olanow, and T.B. Freeman. 1997. Fetal grafting for Parkinson's disease: expression of immune markers in two patients with functional fetal nigral implants. Cell Transplant. 6(3): 213-219. Korr, H., Schilling, W.D., Schultze, B., and Maurer, W. 1983. Autoradiographic studies of glial proliferation in different areas of the 14-day-old rat. Cell Tissue Kinet., 16:393-413. Kreutzberg, G.W. 1996. Microglia: a sensor for pathological events in the CNS. Trends Neurosci. . 19:312-318. Krum, J.M. and Rosenstein, J.M. 1988. Patterns of angiogenesis in neural transplant models: II. Fetal neocortical tranplants. J. Comp. Neurol. 271: 331-345. LaRosa, F.G. and Talmage, D.W. 1985. Synergism between major and minor histocompatability antigens in rejection of cultured allografts. Transplantation 39:480-485. Lawrence, J.M., Huang, S.K., and Raisman, G. 1984. Vascular and astrocytic reactions during establishment of hippocampal transplants in adult host brain. Neuroscience 12(3): 745-760. Lawrence, J.M., Morris, R.J., Wilson, D.J., and Raisman, G. 1990. Mechanisms of allograft rejection in the rat brain. Neuroscience. 37: 431-462. Lazarov-Speigler, O., Soloman, A.S., Zeev-Brann, A.B., Hirschberg, D.L., Lavie, 132 v., and Schwartz, M. 1996. Transplantation of activated macrophages overcomes central nervous system regrowth failure. F^Siifi J., 10: 1296-1302. Leibovitch, S.J., Ross, R. 1975. The role of the macrophage in wound repair. A study with hydrocortisone and antimacrophage serum. J. Pathol. 78: 71-100. Leigh, K., Elisevich, K., Rogers, K.A. 1994. Vascularization and microvascular permeability in solid versus cell-suspension embryonic neural grafts. J. Neurosurg. 81 : 272- 283. Lindholm, D., Heumann, R., Meyer, M., Thoenen, H. 1987. Interleukin-1 regulates synthesis of nerve growth factor in non-neuronal cells of rat sciatic nerve. Nature 330: 658- 659. Lindholm, D., Heumann, R., Meyer, M. and Thoenen, H. 1990. Transforming growth factor-pi stimulates expression of nerve growth factor in the rat CNS. Neuroreport, 1:9-12. Lindvall, O. 1991. Transplants in Parkinson's disease. Eur. Neurol. 31 (Suppl 1):17- 27. Lindvall, O., Brundin, P., Widner, H., Rehncrona, S., Gustavii, B., Frackowiak, R., Leenders, K.L., Sawle, G., Rothwell, J.C, Marsden, CD., et al. 1990. Grafts of fetal dopamine neurons survive and improve motor function in Parkinson's disease. Science. 247(4942):574-577. Mallat, M., Houlgatte, R., Brachet, P., and Prochiantz, A. 1989. Lipopolysaccharide-stimulated rat brain macrophages release NGF in vitro. Dev. Biol., 133:309-311. Mason, D.W., Charlton, H.M., Jones, A.J., Lavy, C.B., Puklavec, M., and Simmonds, S.J. 1986. The fate of allogeneic and xenogeneic neuronal tissue transplanted into the third ventricle of rodents. Neuroscience. 19: 685-694. Masuda-Nakagawa, L. M., Muller, K.J., and Nicholls, J.G. 1993. Axonal sprouting and laminin appearance after destruction of glial sheaths. Proc. Natl. Acad. Sci. 90: 4966- 4970. Mato, M., Ookawara, S., Mato, T.K., and Namiki, T. 1985. An attempt to differentiate further between microglia and fluorescent granular perithelial (FGP) cells by their capacity to incorporate exogenous protein. Am. J. Anat. 172: 125-140. Matsumoto, Y., Hara, N., Tanaka, R., and Fujiwara, M. 1986. Immunohistochemical analysis of the rat central nervous system during experimental 133 allergic encephalomyelitis with special reference to la-positive cells with dendritic morphology. J. Immunol. 136:3668-3676. McGeer, P.L., Itagaki, S., and McGeer, E.G. 1988. Expression of the histocompatibility glycoprotein HLA-DR in neurological disease. Acta Neuropathol (Bed) 76:550-557. McLean, I.W. and Nakane, P.K. 1974. Periodate-lysine-paraformaldehyde fixative. A new fixative for immunoelectron microscopy. J. Histochem. Cytochem. 22: 1077-1083. McLeod, D.S., Lutty, G.A., Wajer, S.D., and Flower, R.W. 1987. Visualization of a developing vasculature. Microvasc. Res. 33: 257-269. Medawar, P.B. 1948. Immunity to homologous grafted skin. III. The fate of skin homografts transplanted to the brain, to subcutaneous tissue, and to the anterior chamber of the eye. Br. J. Exp. Pathol. 29: 58-69. Morioka, T., Kalehua, A.N., and Streit, W.J. 1992. Progressive expression of immunomolecules on microglial cells in rat dorsal hippocampus following transient forebrain ischemia, ^cto Neuropathol. (Berl.) 83:149-157. Mostafa, L.K., Jones, D.B., Wright, D.H. 1980. Mechanism of the induction of angiogenesis by human neoplastic lymphoid tissue: studies on the chorioallantoic membrane (CAM) of the chick embryo. J. Pathol. 132:191-205. Murphy, J.B. and Sturm, E. 1923. Conditions determining the transplantability of tissues in the brain. J. Exp. Med. 38: 183-197. Nagata, K., Takei, N., Nakajima, K., Saito, H., Kohsaka, S. 1993. Microglial conditioned medium promotes survival and development of cultured mesencephalic neurons from embryonic rat brain. J. Neurosci. Res. 34:357-363. Nakajima K., Tsuzaki, N., Shimojo, M., Hamanoue, M., and Kohsaka, S. 1992. Microglia isolated from rat brain secrete a urokinase-type plasminogen activator. Brain Res. 577: 285-292. Neuhaus, J., Fedoroff, S. 1994. Development of microglia in mouse astroglia cultures. Glia 11: 11-17. Nicholas, M.K. and Amason, E.G. 1989. Immunologic considerations in transplantation to the central nervous system. In: Neural Regeneration and Transplantation. pp. 266-269. Alan R. Liss, Inc., New York. Noble, M., Fok-Seang, J., and Cohen, J. 1984. Glia are a unique substrate for the in vitro growth of central nervous system neurons. J. Neurosci. 4: 1892-1903. 134 Okura, Y., R. Tanaka, K. Ono, S. Yoshida, N. Tanuma. and Y. Matsumoto. 1997. Treatment of rat hemiparkinson model with xenogeneic neural transplantation: tolerance induction by anti-T-cell antibodies. J. Neurosci. Res. 48: 385-396. Olby, N.J. and Blakemore, W.F. 1996.Primary demyelination and regeneration of ascending axons in the dorsal funiculus of the rat spinal cord following photochemically induced injury. J. Neurocytol. 25(8):465-480 Onifer, S.M., Whittemore, S.R., and Holets, V.R. 1993. Variable morphological differentiation of a raphe-derived neuronal cell line following transplantation into the adult ratCNS. Exp. Neurol. 122: 130-142. Papavassiliou, E.. Gogate, N., Proescholdt, M., Heiss, J.D., Walbridge, S., Edwards, N.A., Oldfield, E.H., and Merril, M.J. 1997. Vascular endothelial growth factor (vascular permeability factor) expression in injured rat brain. J. Neurosci. Res. 49:451-460. Pedersen, E.B., Zimmer, J., Finsen, B. 1997. Triple immunosuppression protects murine intracerebral, hippocampal xenografts in adult rat hosts: effects on cellular infiltration, major histocompatibility complex antigen induction and blood-brain barrier leakage. Neuroscience. 78(3):685-701. Pennell, N.A., Rabchevsky, A.G., Streit, W.J. 1995. Depletion of major histocompatability complex (MHC)- bearing cells from embryonic rat spinal cord. Soc. Neurosci. Abstr. Pennell, N.A., Streit, W.J. 1996. Depletion of MHC-bearing cells from embryonic rodent nervous tissue: an alternative to immunosuppression. American Society for Neural Transplantation A bstr. Pennell, N.A. and Streit, W.J.. 1997. Colonization of neural allografts by host microglia: relationship to graft neovascularization. Cell Transplant. 6(3): 221-230. Pennell, N.A. and Streit, W.J.. 1998. Tracing of fluoro-gold labeled microglia injected into the adult rat brain. Glia 23: 84-88. Pepper, M.S., Ferrara, N., Orci, L., and Montesano, R. 1992. Potent synergism between vascular endothelial growth factor and basic fibroblast growth factor in the induction of angiogenesis in vitro. Biochem. Biophys. Res. Commun. 189: 824-831. Perry, V.H., and Lund, R.D. 1989. Microglia in retinae transplanted to the central nervous system. Neuroscience. 31 (2): 453-462. Peters, B.P., Goldstein, I.J. 1979. The use of fluorescein-conjugated Bandeiraea 135 simplificolia B4-isolectin as a histochemical reagent for the detection of alpha-D-galactopyranosyl groups. Ex/?. Cell Res. 120:321-334. Piani, D., Frei, K., Do, K.Q., Cuenod, M., Fontana, A. 1991 . Murine brain macrophages induce NMDA receptor mediated neurotoxicity in vitro by secreting glutamate. A^ewro.s'cz. Lett. 133: 159-162. Pierret, P., Quenneville, N., Vandaele, S., Abbaszadeh, R., Lanctot, C, Crine, P., and Doucet, G. 1998. Trophic and tropic effects of striatal astrocytes on cografted mesencephalic dopamine neurons and their axons. J. Neurosci. Res. 51(1): 23-40. Poltorak, M. and Freed, W.J. 1991 . BN rats do not reject F344 brain allografts even after systemic sensitization. yi«n. Neurol. 29: 377-388. Poltorak, M. and Freed, W.J. 1989. Immunological reactions induced by intracerebral transplantation: evidence that host microglia but not astroglia are the antigen-presenting cells. Exp. Neurol. 103:222-233. Polverini, P.J., Cotran, R.S., Gimbrone, M.A., Unanue, E.M. 1977a. Activated macrophages induce vascular proliferation. Nature 269:804-806. Polverini, P.J., Cotran, R.S., Sholly, M.M. 1977b. Endothelial proliferation in the delayed hypersensitivity reaction: an autoradiographic study. J. Immunol. 118: 529-532. Popovich, P.G., Streit, W.J., and Stokes, B.T. 1993. Differential expression of MHC class II antigen in contused rat spinal cord. J. Neurotrauma. 10:37-46. Potgens, A.J.G., Westphal, H.R., de Waal, R.M.W., and Ruiter, D.J. 1995. The role of vascular permeability factor and basic fibroblast growth factor in tumor angiogenesis. Biol. Chem. Hoppe-Seyler. 376:57-70. Prewitt, C.M.F., Niesman, I.R., Kane, C.J.M., and Houle, J.D. 1997. Activated macrophage/microglial cells can promote the regeneration of sensory axons into the injured spinal cord. Exp. Neurol 148(2): 433-443. Pundt, L.L., Kohdoh, T., and Low, W.C. 1995. The fate of human glial cells following transplantation in normal rodents and rodent models of neurodegenerative disease. Brain Res. 695: 25-36. Rabchevsky, A.G. and Streit, W.J. 1997. Grafting of cultured microglial cells into the lesioned spinal cord of adult rats enhances neurite outgrowth. J. Neurosci. Res.. 47:34- 48. Reier, P.J., Anderson, D.K., Thompson, F.J., and Stokes, B.T. 1992. Neural tissue transplantation and CNS trauma: Anatomical and ftinctional repair of the injured spinal 136 cord. J. Neurotrauma 9:S223-S247 Reier, P.J., Anderson, D.K., Schrimsher, G.W., Bao, J., Friedman, R.M., Ritz, L.A., and Stokes, B.T. 1994. Neural cell grafting: Anatomical and functional repair of the spinal cord. In: The neurobiology of central nervous system trauma (S.K. Salzman and A.I. Faden, eds.). pp. 288-3 1 1 . Oxford University Press, New York. Reier, P.J., Eng, L.F., and Jakeman, L. 1988. Reactive astrocyte and axonal outgrowth in the injured CNS: is gliosis really an impediment to regeneration? In: Neural Regeneration Research for the Clinician (F.J. Sell, ed.), pp. 183-209. Alan R. Liss, New York. Reier, P. J., Stensaas, L.J., and Guth, L. 1983. The astrocytic scar as an impediment to regeneration in the central nervous system. In: Spinal Cord Reconstruction (C.C. Kao, R.P. Bunge, and P.J. Reier, eds.), pp. 163-195. Raven Press, New York. Remick D.G., Scales W.E., May M.A., Spengler M., Nguyen D. and Kunkel S.L. 1988. In situ hybridization analysis of macrophage-derived tumor necrosis factor and interleukin-1 mRNA. Lab Invest 59:809-816. Richardson, P.M., McGuinness, U.M., and Aguayo, A.J. 1980. Axons from CNS neurones regenerate into PNS grafts. Nature. 284: 264-265. Ridley, A. and Cavanagh, J.B. 1969. The cellular reactions to heterologous, homologous, and autologous skin implanted into brain. J. Pathol. 99: 193-203. Rinaman, L., Milligan, C.E., and Levitt, P. 1991. Persistence of Fluoro-Gold following degeneration of labeled motoneurons is due to phagocytosis by motoneurons and macrophages. Neurosci., 44:765-776. Rio-Hortega, P. del. 1932. Microglia. In: Cytology and cellular pathology of the nervous system (W. Penfield, ed.), Vol. 2: pp. 481-584. Paul B. Hoeber, New York. Rosenbluth, J., Hasegawa, M., Shirasaki, N., Rosen, C.L., and Liu, Z. 1990. Myelin formation following transplantation of normal fetal glia into myelin-deficient rat spinal cord. J Neurocytology. 19: 718-730. Rostaing-Rigattieri, S., Flores-Guevara, R., Peschanski, M., and Cadusseau, J. 1997. Glial and endothelial cell response to a fetal transplant of purified neurons. Neuroscience. 79(3): 723-734. Savio, T. and Schwab, M.E. 1 990. Lesioned corticospinal tract axons regenerate in myelin-free rat spinal cord. Proc Nat Acad Sci 87:4130-4133. 137 Sawada, M., Kondo, N., Suzumura, A., Marunouchi T. 1989. Production of tumor necrosis factor-alpha by microglia and astrocytes in culture. Brain. Res. 491 :394-397. Schmued, L.C. and Fallon, J.H. 1986. Fluoro-Gold: a new fluorescent retrograde axonal tracer with numerous unique properties. Brain Res., 377:147-154. Schultz-Cherry, S., and Murphy-Ulrich, J.E. 1993. Thrombospondin causes activation of latent transforming growth factor-beta secreted by endothelial cells by a novel mechanism. J, Cell Biol. 122: 923-932. Shimojo, M., Nakajima, K., Takei, N., Hamanoue, M. and Kohsaka, S. 1991. Production of basic fibroblast growth factor in cultured rat brain microglia. Neurosci. Lett., 123:229-231. Skoff, R.P. 1980. Neuroglia: a reevaluation of their origin and development. Pathol. Res. Pract 168:279-300. Steiniger, B. and van der Meide, P.H. 1988. Rat ependyma and microglia cells express class II MHC antigens after intravenous infusion of recombinant gamma interferon. J. Neuroimmunol. 19:1 1 1-1 18. SteinmuUer D. 1983. Skin-specific histocompatability antigens. In: Traumatic Injury, Infection, and other Immunologic Sequelae, pp. 181-196. Baltimore: University Park Press. Stone, J., Itin, A., Alon, T., Pe'er. J., Gnessin, H.. Chan-Ling, T., and Keshet, E. 1995. Development of retinal vasculature is mediated by hypoxia-induced vascular endothelial growth factor (VEGF) expression by neuroglia. J. Neurosci. 15(7): 4738-4747. Streit, W.J. 1990. An improved staining method for rat microglial cells using the lectin from Griffonia simplicifolia (GSA I-B4). J. Histochem. Cytochem., 38:1683-1686. Streit, W.J. and Graeber, M.B. 1993. Heterogeneity of microglial and perivascular cell populations: insights gained from the facial nucleus paradigm. Glia., 7:68-74. Streit, W.J., Graeber, M.B., and Kreutzberg, G.W. 1988. Functional plasticity of microglia: a review. Glia 1: 301-307. Streit, W.J., Graeber, M.B., and Kreutzberg, G.W. 1989a. Expression of la antigen on perivascular and microglial cells after sublethal and lethal motor neuron injury. Exp. Neuro.l 105:115-126. Streit, W.J., Graeber, M.B., and Kreutzberg, G.W. 1989b. Peripheral nerve lesion 138 produces increased levels of major histocompatability complex antigens in the central nervous system. J Neuroimmunology. 21: 117-123. Streit, W.J. and Kincaid-Colton, C.A. 1995. The brain's immune system. Sci. Am., 273:54-61. Sunderkotter, C, Steinbrink, K., Goebeler, M., Bhardwaj, R., Sorg, C. 1994. Macrophages and angiogenesis. J. Leukocyte Biol. 55:410-422. Suzumura, A., Mezitis, S.G.E., Gonatas, N.K., and Silberberg, D.H. 1987. MHC expression on bulk isolated macrophage-microglia from newborn mouse brain: induction of la antigen expression by y-interferon. J. Neuroimmunology. 15: 263-278. Takei, K., Nakano, Y., Shinozaki, T., Toya, S., Tsukada, Y., and Kohsaka, S. 1990. Immunological rejection of grafted tissue in xenogeneic neural transplantation. Prog. Brain Res. 82: 103-143. Thanos, S., Kacza, J., Seeger, J., and Mey, J. 1994. Old dyes for new scopes: the phagocytosis-dependent long-term fluorescence labeling of microglial cells in vivo. Trends Neurosci., 17:177-182. Thanos, S. and Mey, J. 1995. Type-specific stabilization and target-dependent survival of regenerating ganglion cells in the retina of adult rats. J. Neurosci. 1 5(2): 1057- 1079. Thanos, S., Pavlidis, C, Mey, J., and Thiel, H-J. 1991. Specific transcellular staining of microglia in the adult rat after traumatic degeneration of carbocyanide-filled retinal ganglion cells. Exp. Eye Res., 55:101-1 17. Theele, D.P. and Reier, P.J. 1996. Immunomodulation with intrathymic grafts or anti-lymphocyte serum promotes long-term intraspinal allograft survival. Cell Transplant. 5(2): 243-255. Theele, D.P., Schrimsher, G.W., and Reier, P.J. 1996. Comparison of the growth and fate of fetal spinal iso- and allografts in the adult rat injured spinal cord. Exp. Neurol. 142: 128-143. Traugott, U., Scheinberg, L., and Raine, C. 1985. On the presence of la-positive endothelial cells and astrocytes in multiple sclerosis lesions and its relevance to antigen presentation. J Neuroimmunol. 8: 1-14. VanBuskirk, A.M., Brown, D.J., Adams, P.W., and Orosz, C.G. 1994. The MHC and allograft rejection. In: The Role of MHC and Non-MHC Antigens in Allograft Immunity (T. Mohanakumar, ed.). pp. 27-72. R.G. Landes Company, Austin. 139 Vignais, L., Nait Oumesmar, B., Mellouk, F., Gout, O., Labourdette, G., Baron-Van Evercooren, A., and Gumpel, M. 1993. Transplantation of oligodendrocyte precursors in the adult demyelinated spinal cord: migration and remyelination. Int. J. Dev. Neurosci.\\(5):603-6\2. Vlodavsky, 1., Fuks, Z., Ishai-Michaeli, R., Bashkin, P.. Levi, E., Komer, G., Bar- Shavit, R., and Klagsbmn, M. 1991. Extracellular matrix-resident bascic fibroblast growth factor: implication for the control of angiogenesis. J Cell. Biochem. 45: 167-176. Wang, J.J., Chuah, M.I., Yew, D.T.W., Leung, P.C., and Tsang, D.S.C. 1995. Effects of astrocyte implantation into the hemisected adult rat spinal cord. Neuroscience. 65(4): 973-981. Ward, S.A., Ransom, P.A., Booth, P.L., and Thomas, W.E. 1991. Characterization of ramified microglia in tissue culture: pinocytosis and motility. J. Neurosci. Res.. 29:13- 28. Waxinan, S.G., Utzschneider, D.A., Kocsis, J.D. 1994. Enhancement of action potential conduction following demyelination: experimental approaches to restoration of function in multiple sclerosis and spinal cord injury. Prog. Brain Res. 100:233-243 Wekerle, H., Linington, C, Lassman, H., and Meyerman. 1986. Cellular immune reactivity within the CNS. Trend Neurosci. 9: 271-277. Widner, H. and Brundin, P. 1988. Immunological aspects of grafting in the mammalian central nervous system. A review and speculative synthesis. Brain Res. Reviews. 13:287-324. Widner, H. and Brundin, P., Bjorklund, A., and Moller, E. 1989. Survival and immunogenicity of dissociated allogeneic fetal dopaminergic-rich grafts when implanted into the brains of adult mice. Exp. Brain Res. 76: 187-197. Widner, H., Moller, G., and Johansson, BB. 1988. Immune response in deep cervical lymph nodes and spleen in the mouse after antigen deposition in different intracerebral sites. Scand. J. Immunol. 28: 563-571. Williams, K., Ulvestad, E., and Antel, J.P. 1994. B7/BB-1 antigen expression on adult human microglia studied in vitro and in situ. Eur. J. Immunol. 24(12):3031-3037. Wong G.H.W., Bartlett P.P., Clark-Lewis 1., McKimm-Breschkin J.L. and Schrader J.W. 1985. Interferon-gamma induces expression of H-2 and la antigens on brain cells. J Neuroimmunol 7:255-278. 140 Wood, M.J.A., Sloan, D.J., Wood, K.J., and Charlton, H.M. 1996. Indefinite survival of neural xenografts induced with anti-CD4 monoclonal antibodies. Neuroscience. 70(3): 775-789. Xu, X.M., Guenard, V., Kleitman, N., and Bunge, M.B. 1995. Axonal growth into Schwann cell-seeded guidance channels grafted into transected adult rat spinal cord. J. Comp. Neurol. 351:145-160. Yoffey, J.M. and Courtice, F.C. 1970. Lymphatics, lymph and the lymphomyeloid complex. Academic, New York. Zhou, H.F. and Lund, R.D. 1992. Neonatal host astrocyte migration into xenogeneic cerebral cortical grafts. Brain Res. Dev. Brain Res. 65(1):1 27-131. Zhou, X.F. and Rush, R.A. 1995. Peripheral projections of rat primary sensory neurons immunoreactive for neurotrophin 3. J. Comp. Neurol., 363:69-77. BIOGRAPHICAL SKETCH Nathan Adam Pennell was bom on March 21, 1971, in Latrobe, Pennsylvania. An only child, Nathan was raised there by his mother and his grandparents until the age of seven when his mother and he moved to St. Petersburg, Florida. There he spent the next seven years growing up. In 1985 Nathan and his mother moved to Plantation, Florida, where Nathan attended high school. There the formerly chubby and bespectacled boy joined the junior varsity football team, lost weight, received his first pair of contact lenses, and decided the world had to call him "Nate" now. Four years later Nate was accepted to the University of North Carolina at Chapel Hill, a place that seemed like Heaven on Earth to a poor South Florida boy, and proceeded to have the time of his life. At UNC Nate was a varsity sabre fencer, making it all the way to the NCAA Championships his senior year. Nate was also a biology major, and discovered an interest in research while working as an undergraduate in Dr. Ken Lohmann's lab. In this lab Nate developed an intense dislike for all things behavioral. In the fateful summer of 1992 Nate was working at the C.V. Whitney Marine Laboratory in Marineland, Florida. Here he was deciding that he hated anything to do with biochemistry, and was wondering if science was the way to go (although beach volleyball made up for a lot). Fortunately, Nate made a trip to the UF campus to meet with various departmental heads in the graduate school, and discovered his calling while 141 142 reading the Department of Neuroscience brochure. Nate appUed, was accepted, and thus began his graduate career with no idea what he wanted to research. After a couple of uneventftal rotations, Nate heard a seminar dealing with graft rejection and was hooked. He approached Dr. Wolfgang "Jake" Streit with his interest and was told that there simply was no funding available. Here Fate smiled as Dr. Streit soon received a Paralyzed Veterans of America grant to look at graft rejection, but had no one to do the work. Not having to be asked twice, Nate took the reins and hasn't looked back since. Along the road culminating in this dissertation, Nate received a NIMH pre- doctoral fellowship and published several papers dealing with microglia and neural transplantation. While working in spinal cord injury research, Nate also discovered an intense interest in the clinical aspects of neuroscience, and decided to attend medical school. He was accepted into the University of Florida College of Medicine, and will attend UF this Fall, with the intent of pursuing a career in academic medicine. I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy/ j J /j M' I Wolfgang J. Streit, Chair Associate Professor of Neuroscience I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequ^teijn scope and quality, as a dissertation for the degree of Doctor of Philosophy .\ Douglas Kk Anderson CM. and K.E. Overstreet Eminent Scholar and Professor of Neuroscience and Neurological Surgery I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Susan Iv^Semple-l^A Associate Professor of Neuroscience I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. \^J- A^tyU vjK,. Jo^l^L. Schiffenbauer \ Associate Professor of Molecular Genetics and Microbiology This dissertation was submitted to the Graduate Faculty of the College of Medicine and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August, 1998 Dean, Graduate Scnool IHIIMl'lV °'' '''■°"'°'* 3 1262 08666 362 1