Angiogenic and astroglial responses to vascular endothelial growth factor administration in adult rat brain.

The effects of exogenous vascular endothelial growth factor (VEGF) on angiogenesis, blood-brain barrier permeability and astroglial proliferation in the adult rat CNS in situ were investigated. Recombinant human VEGF(165) (25 or 50 ng/ml) was delivered for up to 1 week using either intracerebral osmotic minipumps or less traumatic subdural gelatin sponge placement. By 3 days, VEGF delivery caused significantly increased cerebral angiogenesis (25 ng/ml was most effective) in both experimental models when compared to saline controls; VEGF infusion resulted in a 100% increase in an index of vascular proliferation, and gelatin sponge delivery produced a 65% increase. The blood-brain barrier hallmark endothelial glucose transporter-1 was not present in nascent vascular sprouts. Infusion of VEGF produced extensive protein leakage that persisted after saline-induced permeability was mostly resolved, while gelatin sponge administration caused milder barrier dysfunction. Administration of the angiogenic factor had unexpected proliferative effects on astroglia in both models, resulting in an 80-85% increase in mitotically active astroglia when compared to controls. Immunohistochemical results and semi-quantitative reverse transcriptase-polymerase chain reaction indicated that the VEGF receptors flk-1 and flt-1 were up-regulated in response to the infusion trauma; flt-1 was localized to reactive astroglia, while flk-1 was expressed in vascular endothelium but predominantly in neuronal somata and processes adjacent to the delivery site. mRNA for the VEGF(121), VEGF(165) and VEGF(188) isoforms was also increased after delivery of the recombinant protein. These data show that VEGF application has substantial proliferative effects on CNS endothelium and astroglia and causes up-regulation of its own message. Flt-1 and flk-1 receptor mRNAs and proteins are up-regulated in both vascular and non-vascular cell types following infusion trauma. From these results we suggest that administered VEGF has heretofore unanticipated pleiotrophic effects in the adult CNS.

Protein synthesis inhibition in neocortical grafts evaluated by systemic amino acid uptake autoradiography.

The temporal pattern of protein synthesis inhibition was examined in grafted neocortical neurons using [(3)H]valine in vivo autoradiography. Neuronal uptake levels of systemically administered (3)H-labeled amino acids which cross the blood-brain barrier (BBB) via endothelial cell neutral carriers have long been a hallmark in studies of experimental ischemic pathology; there is likely a strong correlation between persistent protein synthesis inhibition and the progression of cell damage. Because the grafting procedure involves the loss of blood flow and the subsequent reperfusion of the donor tissue there are, mechanistically, important similarities to reversible ischemia models. The effects of ischemic injury on grafted CNS neurons are not fully understood. Quantitative analysis of grain distribution in individual graft or control (adjacent host cortex) neurons indicated an initial breakdown of the amino acid barrier system, subsequent recovery, and progressive reduction of amino acid uptake by 1 year. Up to 3 weeks after surgery grafts were flooded with the [(3)H]valine tracer but individual neurons contained relatively few silver grains. After this time, the tracer was normally distributed within graft neurons but at significantly lower levels than in controls. Grain density gradually decreased over time such that 12-month grafted neurons had approximately half that compared to control and only 58% of that in 2-month grafts; the 12-month levels were comparable to those observed at early (10 days) postoperative times. Autoradiography of immunostained sections for MAP-2, SMI 311 (neurofilament marker), and neuron-specific enolase showed reduced expression of these proteins in neurons coupled with weak amino acid tracer uptake. The results further suggest that grafted neurons bear intriguing similarities to neurons placed at ischemic risk, particularly "penumbral" neurons, which are affected by reduced blood flow and are metabolically weakened. The loss of BBB properties in early grafts may also extend to the endothelial cell amino acid carrier system, and the delayed revascularization process could affect neuronal uptake mechanisms.

Vascular, glial and neuronal effects of vascular endothelial growth factor in mesencephalic explant cultures.

Vascular endothelial growth factor is a highly conserved, heparin-binding protein which mediates a number of critical developmental processes in both vertebrates and invertebrates, including angiogenesis, vasculogenesis and hematopoiesis. We employed an organotypic rat explant model (produced from embryonic day 17 fetuses) to assess the effects of vascular endothelial growth factor on brain microvasculature in general and the ventral midbrain specifically. Immunohistochemistry using antisera to rat endothelial cell antigen and laminin demonstrated a robust, dose-dependent effect of vascular endothelial growth factor, resulting in increased vessel neogenesis, branching and lumen size by three days in vitro. This effect was blocked by addition of an anti-vascular endothelial growth factor antibody. At higher doses of vascular endothelial growth factor, the effect was attenuated, though a statistically significant increase in both astrocyte, and neuronal density was observed using antisera to glial and neuronal markers. Tyrosine hydroxylase-immunoreactive (i.e. dopaminergic) neurons, particularly, exhibited increased survival in response to vascular endothelial growth factor application. Vascular endothelial growth factor had a mitogenic effect on endothelial cells and astrocytes, but not dopaminergic neurons, as demonstrated by the addition of [3H]thymidine to the cultures 2 h after the cultures were established. Similarly, results of a radioreceptor assay indicated that specific vascular endothelial growth factor binding sites were present on blood vessels and astrocytes, and were up-regulated by exposure to vascular endothelial growth factor. We conclude that, in explants of the ventral mesencephalon, exogenously applied vascular endothelial growth factor is mitogenic for endothelial cells and astrocytes, and promotes growth/survival of neurons in general and dopaminergic neurons in particular.

Transient coexpression of nestin, GFAP, and vascular endothelial growth factor in mature reactive astroglia following neural grafting or brain wounds.

The spatial and temporal immunoexpression of the intermediate filament (IF) protein nestin and its relationship to glial fibrillary acidic protein (GFAP), vascular endothelial growth factor (VEGF), and its receptor flt-1 (VEGF-R1) in reactive astroglia was examined following stab wounds or transplants of fetal CNS tissue into the adult brain. Since developmentally regulated proteins and gene transcripts can be reexpressed in reactive astroglia following certain brain injuries, we analyzed the nestin profile in these experimental paradigms in order to more fully understand the nature of the gliotic "scar." Nestin expression was transiently up-regulated in some but not all astrocytes which often had a different morphology than the typical stout, stellate GFAP (+) cells; the processes of the nestin (+) cells tended to be slender and elongated. In reactive astroglia from the mature brain, nestin expression was robust but generally localized to the wound or graft site, peaked at 7-10 days postoperative, and was absent by 28 days, whereas GFAP (+) astrocytes were far more widespread and persisted for many months. Only nestin was strongly expressed immediately adjacent to early stab wounds, whereas GFAP (+) cells were located further from the wound sites. In contrast, there was marked nestin/GFAP colocalization at the graft/host interface. Semiquantitative analysis combined with confocal microscopy revealed a unique compartmentalization of protein expression; processes from single astrocytes could be entirely nestin (+), GFAP (+), or could show coexpression. At 4, 7, and 14 days postoperative, 41, 58, and 32% of the immunoexpression, respectively, was accounted for by nestin at the graft/host interface, and it was essentially undetectable at 28 days postoperative. In situ hybridization studies showed nestin transcripts within GFAP (+) cells primarily between 4 and 10 days postoperative and absent by 28 days. Many nestin (+) astrocytes, as shown by electron microscopy, were closely related to the vasculature. Therefore we further examined the expression of vascular endothelial growth factor (VEGF), an endothelial cell mitogen associated with angiogenesis. Nestin colocalized with VEGF in some astrocytes (7%) but far more prominently with the VEGF flt-1 receptor (25%). Early astroglial activation may involve several different IF components and possibly a distinct astrocytic population that shows a rapid, transient nestin expression adjacent to injury sites. Expression of the nestin IF phenotype within affected astrocytes in the surgical vicinity may be indicative of a reversion to an immature phenotype that might be less susceptible to attendant hypoxia after injury. Since injured astrocytes are well known to express many bioactive compounds, such transient reexpression of early, developmentally regulated proteins may be a hallmark for the elaboration of growth factors such as VEGF.

Patterns of brain angiogenesis after vascular endothelial growth factor administration in vitro and in vivo.

Vascular endothelial growth factor (VEGF) is a secreted endothelial cell mitogen that has been shown to induce vasculogenesis and angiogenesis in many organ systems and tumors. Considering the importance of VEGF to embryonic vascularization and survival, the effects of administered VEGF on developing or adult cerebrovasculature are unknown: can VEGF alter brain angiogenesis or mature cerebrovascular patterns? To examine these questions we exposed fetal, newborn, and adult rat cortical slice explants to graduated doses of recombinant VEGF. The effects of another known angiogenic factor, basic fibroblast growth factor (bFGF), were evaluated in a comparable manner. In addition, we infused VEGF via minipump into the adult cortex. Significant angiogenic effects were found in all VEGF experiments in a dose-responsive manner that were abolished by the addition of VEGF neutralizing antibody. Fetal and newborn explants had a highly complex network of branched vessels that immunoexpressed the flt-1 VEGF receptor, and flk-1 VEGF receptor expression was determined by reverse transcription-PCR. Adult explants had enlarged, dilated vessels that appeared to be an expansion of the existing network. All bFGF-treated explants had substantially fewer vascular profiles. VEGF infusions produced both a remarkable localized neovascularization and, unexpectedly, the expression of flt-1 on reactive astrocytes but not on endothelial cells. The preponderance of neovascularization in vitro and in vivo, however, lacked the blood-brain barrier (BBB) phenotype marker, GLUT-1, suggesting that in brain the angiogenic role of VEGF may differ from a potential BBB functional role, i.e., transport and permeability. VEGF may serve an important capacity in neovascularization or BBB alterations after brain injury.

Developmental expression of calcium-binding protein-containing neurons in neocortical transplants.

The present study examined the development of calcium binding protein-containing neurons in a timed series of fetal neocortical transplants. The immunoexpression of parvalbumin and calbindin, which are subpopulations of GABAergic neurons, have been widely studied in normal development and in disease and injury states. Because of their purported resistance to oxidative injury by their ability to buffer Ca++ influx, these neurons have been particularly studied following ischemia. Because it is likely that oxidative stress is associated with the grafting procedure, we sought to determine if these neurons displayed enhanced survival characteristics. Normally, parvalbumin and calbindin represent about 5-10% of cortical neurons. Within 2-4 wk after grafting the expression of both proteins increased markedly in that a relatively larger number of neurons (27% for parvalbumin) were immunopositive. This increase was transitory, however, and by 4 mo and beyond, confocal microscopic data showed a reduction of over 50% of parvalbumin (+) neurons and processes. Calbindin (+) processes showed a qualitative change in that they were smaller with less terminal branching. Electron microscopy confirmed a substantial reduction in parvalbumin synaptic contacts. Interestingly, in older grafts, remaining parvalbumin neurons were those that were strongly NSE (+) suggesting a link between normal metabolism and Ca++ buffering in grafted neurons. It is possible that in early grafts certain neuronal populations transiently upregulated calcium binding proteins as a defensive mechanism against Ca++ influx associated with oxidative stress. Over time, however, following physiological normalization within grafts, the calcium binding protein (+) neurons are diminished, possibly due to lack of appropriate afferent input to the interneuronal pool.

VEGF mRNA and its receptor flt-1 are expressed in reactive astrocytes following neural grafting and tumor cell implantation in the adult CNS.

Significant angiogenesis occurs only after injury in the adult mammalian brain; capillaries proliferate and astrocytes are activated by presently unresolved cellular mechanisms. Because of the intimate relationship between astrocytes and brain capillaries we examined the expression of the specific endothelial mitogen vascular endothelial growth factor (VEGF) in reactive astrocytes following CNS trauma models: neural grafting, stab wounds, and glioma implantation. In situ hybridization was combined with GFAP immunohistochemistry to delineate VEGF mRNA expression in reactive astrocytes. In addition, VEGF and its receptor flt-1 protein expression were detected immunohistochemically. In all three models we found unexpectedly that only reactive astrocytes, not endothelium, expressed the VEGF receptor flt-1, VEGF mRNA, and VEGF protein in a spatiotemporal manner, suggesting that activated astroglia may have a direct role in the induction of angiogenesis or permeability in mature brain. In addition, secreted VEGF may play a part in astroglial signalling by the induction of its own receptor in reactive astroglia following injury. These findings may have significant implications with regard to growth and reparative mechanisms of the adult cerebrovasculature.

Grafts of EGF-responsive neural stem cells derived from GFAP-hNGF transgenic mice: trophic and tropic effects in a rodent model of Huntington's disease.

The present study examined whether implants of epidermal growth factor (EGF)-responsive stems cells derived from transgenic mice in which the glial fibrillary acid protein (GFAP) promoter directs the expression of human nerve growth factor (hNGF) could prevent the degeneration of striatal neurons in a rodent model of Huntington's disease (HD). Rats received intrastriatal transplants of GFAP-hNGF stem cells or control stem cells followed 9 days later by an intrastriatal injection of quinolinic acid (QA). Nissl stains revealed large striatal lesions in rats receiving control grafts, which, on average, encompassed 12.78 mm3. The size of the lesion was significantly reduced (1.92 mm3) in rats receiving lesions and GFAP-hNGF transplants. Rats receiving QA lesions and GFAP-hNGF-secreting grafts stem cell grafts displayed a sparing of striatal neurons immunoreactive (ir) for glutamic acid decarboxylase, choline acetyltransferase, and neurons histochemically positive for nicotinamide adenosine diphosphate. Intrastriatal GFAP-hNGF-secreting implants also induced a robust sprouting of cholinergic fibers from subjacent basal forebrain neurons. The lesioned striatum in control-grafted animals displayed numerous p75 neurotrophin-ir (p75NTR) astrocytes, which enveloped host vasculature. In rats receiving GFAP-hNGF-secreting stem cell grafts, the astroglial staining pattern was absent. By using a mouse-specific probe, stem cells were identified in all animals. These data indicate that cellular delivery of hNGF by genetic modification of stem cells can prevent the degeneration of vulnerable striatal neural populations, including those destined to die in a rodent model of HD, and supports the emerging concept that this technology may be a valuable therapeutic strategy for patients suffering from this disease.

Expression of blood-brain barrier characteristics following neuronal loss and astroglial damage after administration of anti-Thy-1 immunotoxin.

In most regions of the CNS, vascular endothelial cells play an important role in maintaining the composition of the neuronal microenvironment by virtue of their blood-brain barrier (BBB) characteristics. The maintenance of the endothelial BBB phenotype in vitro has been attributed primarily to astrocytes but little attention has been paid the potential role of neurons. In this study we have attempted to injure or destroy neurons and fibers of passage in a circumscribed area while leaving vascular and glial elements intact in order to determine if neurons are involved in BBB maintenance in situ. The immunotoxin OX7-SAP, a conjugate of the Thy-1 antibody OX7 and the ribosome-inactivating protein saporin, was injected into the adult rat striatum to effect neuronal death at the injection site. Although neurons and fibers of passage were destroyed within the lesion, glial cells unexpectedly were also severely injured as determined by immunohistochemical expression of several neuronal and astroglial marker proteins and ultrastructural analysis. The microvasculature remained intact, allowing a qualitative immunohistochemical analysis of several BBB markers at time points ranging from 3 to 28 days postinjection. Despite the loss of both neurons and astroglia within the lesions, the microvasculature continued to express the brain-type endothelial glucose transporter GLUT-1 at all time points examined. In contrast, the barrier to endogenous protein (rat serum albumin) and the expression of endothelial barrier antigen (EBA) decreased initially but recovered even in areas that contained minimal numbers of astroglia and neuronal elements. We conclude that intact neuronal or glial cells do not appear to be necessary for the maintenance in situ of the BBB properties examined herein.

Permeability of the blood-brain barrier to protein and [3H]GABA in intraparenchymal fetal CNS tissue grafts.

In the present study solid grafts of rat fetal neocortex or substantia nigra were transplanted into young adult rat cortex or striatum. To study the permeability to protein and potential changes in the blood-brain barrier (BBB), HRP was administered intravascularly or, to examine endogenous protein exudation, noninjected animals were evaluated after immunocytochemical stain for rat serum albumin. In addition, [3H]GABA, a neurotransmitter that does not cross the BBB, was systemically administered to cortical graft-bearing hosts in order to determine if a barrier to small potentially bioactive compounds was present. Postoperative periods ranged between 1 week and 15 months for the cortical grafts and between 1 week and 8 months for nigral grafts; the experiments were repeated several times at the same time point in order to determine the potential for variability in the model systems. The results show that a high percentage (80%) of both neocortical and nigral grafts lack a complete BBB to protein up to 2-3 weeks postoperative. After this time the majority of neocortical grafts were impermeable to protein yet 20-30% of longer-term specimens showed variable degrees of permeability to protein; permeability to [3H]GABA, which manifested in the transmitter being sequestered by graft neurons, was observed only up to 4 weeks postoperative. Nigral grafts after 4 weeks postoperative showed negligible permeability to serum protein. Although in all early CNS grafts protein permeation is extensive in both graft and adjacent host, the mature host brain resolved the exudation far more efficiently so that the reaction appeared exclusively in the graft or interface and not in the host by 2-3 weeks, suggesting that the graft tissue may lack appropriate mechanisms for clearance of extracellular materials. It is unclear why this phenomenon occurred consistently but in a minority percentage of cortical specimens. It is suggested that either frank pathology or immunopathology provides a subtle and incomplete rejection process that allows the interface vessel to be permeable over the life of the graft. Since nigral grafts appeared to have significantly less BBB dysfunction variation it is possible that the maturation levels or the ontogeny between different graft sources may play a role in differing responses to injury or vascular problems after transplant surgery.

Autoradiographic distribution of bombesin/gastrin-releasing peptide receptors in fetal cortex transplants.

Bombesin/gastrin-releasing peptide (BB/GRP) receptors were characterized in fetal cortex transplants in the rat. Using in vitro autoradiography techniques, the density of (125I-Tyr4)BB grains was low 2 weeks after transplantation but increased 4 weeks after cortex transplantation into the adult rat fourth ventricle. Densitometry analysis of the autoradiograms indicated that (125I-Tyr4)BB bound with high affinity (Kd = 5 nM) to a single class of sites in the transplant tissue. Specific (125I-Tyr4)BB binding was inhibited with high affinity by (Tyr4)BB but not by NMB (IC50 values of 3 and 100 nM, respectively). These data suggest that GRP may act as a novel growth factor and play a regulatory role in the development of fetal cortex transplants.

Functional fetal nigral grafts in a patient with Parkinson's disease: chemoanatomic, ultrastructural, and metabolic studies.

A patient with Parkinson's disease received bilateral fetal human nigral implants from six donors aged 6.5 to 9 weeks post-conception. Eighteen months following a post-operative clinical course characterized by marked improvement in clinical function, this patient died from events unrelated to the grafting procedure. Post-mortem histological analyses revealed the presence of viable grafts in all 12 implant sites, each containing a heterogeneous population of neurons and glia. Approximately 210,146 implanted tyrosine hydroxylase-immunoreactive (TH-ir) neurons were found. A greater number of TH-ir grafted neurons were observed in the right (128,162) than the left (81,905) putamen. Grafted TH-ir neurons were organized in an organotypic fashion. These cells provided extensive TH-ir and dopamine transporter-ir innervation to the host striatum which occurred in a patch-matrix fashion. Quantitative evaluations revealed that fetal nigral grafts reinnervated 53% and 28% of the post-commissural putamen on the right and left side, respectively. Grafts on the left side innervated a lesser area of the striatum, but optical density measurements were similar on both sides. There was no evidence that the implants induced sprouting of host TH-ir systems. Electron microscopic analyses revealed axo-dendritic and occasional axo-axonic synapses between graft and host. In contrast, axo-somatic synapses were not observed. In situ hybridization for TH mRNA revealed intensely hybridized grafted neurons which far exceeded TH mRNA expression within residual host nigral cells. In addition, gamma-amino butyric acid (GABA)-ergic neurons were observed within the graft that formed a dense local neuropil which was confined to the implant site. Serotonergic neurons were not observed within the graft. Cytochrome oxidase activity was increased bilaterally within the grafted post-commissural putamen, suggesting increased metabolic activity. In this regard, a doubling of cytochrome oxidase activity was observed within the grafted post-commissural putamen bilaterally relative to the non-grafted anterior putamen. The grafts were hypovascular relative to the surrounding striatum and host substantia nigra. Blood vessels within the graft stained intensely for GLUT-1, suggesting that this marker of blood--brain barrier function is present within human nigral allografts. Taken together, these data indicate that fetal nigral neurons can survive transplantation, functionally reinnervate the host putamen, establish synaptic contacts with host neurons, and sustain many of the morphological and functional characteristics of normal nigral neurons following grafting into a patient with PD.

Cytoskeletal protein immunoexpression in fetal neural grafts: distribution of phosphorylated and nonphosphorylated neurofilament protein and microtubule-associated protein 2 (MAP-2).

The present study examined the immunocytochemical expression of important cytoskeletal proteins within the neurons of an extended series of neocortical grafts and smaller group of ventral mesencephalic (nigral) grafts. Using antibodies that were directed at all three neurofilament (NF) epitopes, NF-L, NF-M, and NF-H, we attempted to determine whether these neurons would have an altered cytoskeletal profile following the stress of transplantation, because previous studies have shown such changes following ischemia or direct brain injury. We studied phosphorylated NF protein, which is found predominantly in axons, nonphosphorylated NF protein, which is found predominantly in the somata-dendritic compartment, and MAP-2, a specific microtubule marker that is localized exclusively in the somato-dendritic compartment. The results show that in all neocortical grafts examined, both phosphorylated and nonphosphorylated NF immunoexpression was significantly downregulated and appeared only in relatively few axons and somatic profiles, respectively, even though there were numerous Nissl-stained neuronal profiles in the grafts. There was no particular pattern to the immunopositive profiles. At later times occasional neuronal profiles were positive for phosphorylated NF protein, suggesting a reaction to cellular injury. In contrast to neocortical grafts, the cytoskeletal profiles of MAP-2 and phosphorylated NF protein in nigral grafts appeared very similar to age-matched control although the nonphosphorylated NF protein expression did appear somewhat lessened at 1-2 mo postoperative. Because cytoskeletal proteins play important roles in neuronal size, shape, and structural stability, they may subserve key cellular issues in neural grafting. These results show a significant loss of cytoskeletal protein expression in neocortical grafts that does not occur in nigral grafts. These results suggest that fetal neurons from different brain regions (i.e., graft source) may respond differently to the grafting procedure insofar as their cytoskeletal makeup is concerned. In addition, a potential lack of appropriate growth substrates or synaptic contacts may also produce cytoskeletal alterations. As such, the cytoskeletal protein profiles in central nervous system (CNS) grafts may be useful markers for functional performance, perhaps reflecting a degree of cellular injury.

An improved autoradiographic coating technique for neurohistopathological stains and electron microscopy.

We describe modifications and improvements to our first report of a new nuclear emulsion coating technique for both light and electron microscopic autoradiography. Although this technique was originally designed to facilitate electron microscopic autoradiography, the methodology also allows pre-staining of plastic-embedded tissue sections prior to coating the slides with nuclear emulsion for light microscopic autoradiography. We now demonstrate that paraffin sections can be autoradiographically processed after being subjected to a combination of immunocytochemical reactions and special neuroanatomical strains. Parlodion film has been found to be more resistant to temperature changes and less prone to contamination than Formvar film. The shape of the double adhesive tape is an important aspect of the electron microscopic technique; it has been modified to minimize contamination and facilitate the removal of the grids from the glass slide. These technical adjustments facilitate the application of the nuclear emulsion and increase radionuclide specificity, thus expanding the investigative horizons of light and electron microscopic autoradiography.

Why do neural transplants survive? An examination of some metabolic and pathophysiological considerations in neural transplantation.

Neural transplantation continues to be a growing field that has advanced beyond animal experimentation and into the clinic where trials in patients with Parkinson's disease are moving forward (14, 30). The remarkable continuity of the brain grafting paradigm owes to the fact that experimentation and data collection may lend themselves to many disciplines and up-to-date technical analyses. Such procedures can represent significant advances in our knowledge of brain development and disorders but extensive investigation remains to be done to ascertain more precisely the cellular and molecular mechanisms by which neural grafts function. There are still many issues in transplantation that need to be resolved, not the least of which is cell survival and enhancement of graft functional capacity. There have been reports that only about 5-10% of dopaminergic neurons in mesencephalic grafts survive (22) and while the numbers of surviving neurons in cortical grafts is not known, recent studies have indicated that surviving neurons may have abnormal cellular aspects such as immature dendrites (32), inhibited axonal outgrowth (47), or reduced glycolytic activity (43). In the following commentary, possible cellular mechanisms of graft survival will be explored. It will be suggested that based on what is known about both cerebrovascular pathology and normal brain metabolic development, coupled with the mechanisms of graft vascularization, it is unlikely that neural grafts should survive or, at best, survive exhibiting only a modicum of activity. It will be further suggested that the answer "...because it is fetal brain ..." to the question--why (or how) do neural grafts survive?--is an inadequate and presently unexplained one.(ABSTRACT TRUNCATED AT 250 WORDS)

Diminished expression of microtubule-associated protein (MAP-2) and beta-tubulin as a putative marker for ischemic injury in neocortical transplants.

The present study examined the immunoexpression of the neuronal cytoskeletal proteins, MAP-2 and beta-tubulin within a timed series of rat fetal neocortical transplants. beta-tubulin is a major component of microtubules and MAP-2 regulates the assembly and stability of neuronal microtubules and is a major site for the phosphorylation cAMP dependent protein kinase in neurons. Both proteins are strongly expressed in the soma and dendrites of normal neurons. MAP-2 has been shown to be a sensitive marker for ischemia in neurons and is downregulated in this form of injury. Immunoexpression of both MAP-2 and beta-tubulin in grafted cortical neurons was markedly reduced when compared to age-matched or even perinatal specimens at all post-operative times. Dendritic staining was confined to random, thin processes with no laminar patterns and staining within somata was very weak. In some specimens, somatic expression was increased and dendrites were more robustly stained when a portion of the graft was juxtaposed to a fiber tract even though in other regions of the same graft there was very weak immunostaining. The present results corroborate previous studies of cortical transplants indicating an immature structure and metabolism, and it is suggested here that the primary factor is a sublethal form of ischemic injury. Another possibility for the relative paucity of cytoskeletal protein expression could be that transplanted neurons undergo a new developmental scheme (neodevelopment) that is brought about by truncated migration patterns and abnormal synaptic connections.

Immunocytochemical expression of the blood-brain barrier glucose transporter (GLUT-1) in neural transplants and brain wounds.

The present study examined the immunocytochemical expression of the blood-brain barrier glucose transporter (GLUT-1) in a series of fetal neocortical transplants, autonomic tissue transplants, and stab wounds to the rat brain. GLUT-1 is one of a family of different glucose transporters and is found exclusively on barrier-type endothelial cells. In the brain it is responsible for the regulated facilitative diffusion of glucose across the blood-brain barrier. This investigation is the first to determine if this important molecule is altered during the process of angiogenesis that occurs following neural transplantation procedures or direct brain injury. Beginning in late fetal brain, e.g., E18 and continuing into maturity, GLUT-1 was strongly and exclusively expressed on normal cerebral vessels. In solid fetal central nervous system (CNS) transplants up to around 3 weeks postoperative, GLUT-1 was only weakly expressed, particularly as exemplified by colloidal gold immunostaining when compared with the host. At later times examined, up to 15 months postoperative, GLUT-1 immunoexpression was comparable with the normal adjacent brain. In autonomic tissue transplants, where the vessels do not have a blood-brain barrier, as expected, GLUT-1 was not expressed. In stab wounds, at 1 week there was extensive gliosis, and the injured vessels appeared fragmented and collapsed but still expressed GLUT-1, although to a somewhat lesser extent than normal brain. Between 3 and 6 weeks, GLUT-1 was expressed on tortuous vessels and in apparently fibrillar processes in the wound vicinity with a similar pattern to astrocyte (GFAP) reactivity. These results suggest the occurrence of a down-regulation of GLUT-1 in early transplants, perhaps related to reduced glycolytic activity or transient ischemia, or possibly due to the utilization of alternative energy sources. That GLUT-1 expression was not entirely lost in stab wounds to the mature brain suggests that the protein may be more labile in fetal or perinatal brain than in the adult and may not be affected by direct injury. Coupled with previous transplantation studies that have shown reduced neuronal glycolysis and potential barrier alterations, the reduction of GLUT-1 activity within nearly the identical time frame could indicate a relatively early critical period in cellular metabolism following transplantation of CNS tissue.

Immunocytochemical distribution of the brain glucose transporter (GLUT 1) in experimental gliosis.

The present study has examined the immunocytochemical expression of the blood-brain barrier glucose transporter GLUT 1 as compared with GFAP in models of experimental gliosis. In neocortical grafts, gliosis was prominent at the graft-host interface mostly associated with blood vessels. Consecutive sections examined for anti-GLUT 1 showed that the protein was distributed in nearly an identical pattern to the anti-GFAP, staining fibrillar processes and all vessels and also appeared extracellularly. In stab wounds, GLUT 1 immunoexpression was similar to GFAP reactivity and stained injured vessels and glial-like processes that were reminiscent of astrocytic end-feet. Normal glial cells and processes in unaffected neuropil, however, were never stained. This report suggests that GLUT 1 protein may be upregulated in non-endothelial components, such as reactive astroglia or possibly microglia, that are associated with injured or angiogenic vessels.

Developmental expression of neuron-specific enolase immunoreactivity and cytochrome oxidase activity in neocortical transplants.

The present study has examined certain metabolic markers in fetal neocortical tissue transplanted to the cortex, hippocampus, striatum, or ventricle. Particularly, the immunocytochemical expression of neuron-specific enolase (NSE) was studied in a series of host rats ranging between 10 days and 15 months postoperative. NSE is a major glycolytic pathway enzyme found in all neurons. The antibody to NSE is a very reliable marker for neuronal functional metabolic activity and developmental status and its onset has been shown to coincide with synaptic connections. In some grafts oxidative metabolic status was investigated using cytochrome oxidase (CO) histochemistry. In addition, the normal development of NSE expression in rat neocortex was also examined. In normal development, NSE was weakly expressed in fetal brain, but by 1-2 weeks postnatal the enzyme was strongly expressed in all neurons. Typical cortical laminar patterns were evident at 30 days with neurons in layer V and scattered interneurons the most strongly stained. In cortex-cortex transplants NSE expression was very weak; at 1-3 weeks postoperative, it was practically nonexistent; and at all later times only a minority of neurons had normal expression when compared to that in normal development even though by Nissl staining standards in adjacent sections they appeared "normal." Labeling indices ranged between 30 and 49%. Intraventricular grafts had consistently low NSE expression with labeling indices ranging between 18 and 46%. However, when the neocortical tissue was placed in other regions, neuronal NSE appeared only slightly below normal. CO histochemistry corroborated the NSE activity with regards to graft placement. Several possibilities that may account for reduced NSE profile in transplanted neurons include incomplete migration patterns, reduced synaptic connectivity, and potential ischemia causing lowered protein synthesis during reestablishment of vascular connections. If neuronal glycolysis is weakened, it is possible that neurotransmitter production or axonal transport are reduced. Since most energy capacity in brain is dependent on the glycolytic sequence for oxidative metabolism, reduced glycolytic capacity, as depicted by NSE expression, may suggest the presence of transplanted neurons that have adapted to their new environment with a relatively immature profile.