Search

The cytokine interleukin-1beta (IL-1beta) is a potent activator of human astrocytes, inducing or modulating expression of multiple proinflammatory genes via activation of the transcription factors nuclear factor-kappaB (NF-kappaB) and activator protein-1 (AP-1). In this study, we examined whether IL-1beta signaling is regulated in these cells by extracellular nucleotides that are released at high concentrations under inflammatory conditions and act as ligands for members of the P2 receptor family. Using reporter constructs and electromobility shift assays, we found that cotreatment of astrocyte cultures with ATP (1-100 microm) significantly potentiated IL-1beta-mediated activation of NF-kappaB and AP-1 and that ATP alone activated AP-1. These effects were blocked by the P2 receptor antagonists XAMR 0721, periodate-oxidized ATP, and suramin. A role for ATP in modulating IL-1beta-mediated inflammatory gene expression was supported further by the observation that ATP potentiated the IL-1beta-induced expression of IL-8 mRNA and protein but strongly downregulated IP-10 expression. Reverse transcription-PCR and cloning demonstrated expression of the ATP-responsive P2 receptor subtypes P2Y(1), P2Y(2), and P2X(7), as well as the ATP-insensitive receptor P2Y(4). ADP, a selective agonist for P2Y(1), produced results similar to or greater than those obtained using ATP, whereas 2'-3'-O-(4-benzoyl-benzoyl)-ATP, a selective agonist for P2X(7), was less effective than ATP. In contrast, UTP, a selective agonist for P2Y(2) and P2Y(4), was ineffective. These studies indicate that different P2 receptor subtypes play distinct roles in the modulation of IL-1beta-mediated signal transduction in human astrocytes, and that signaling via P2 receptors may fine-tune the transcription of genes involved in inflammatory responses in the human CNS.

Vertebrate tissues use multiple junctional types to establish and maintain tissue architecture, including gap junctions for cytoplasmic connectivity and tight junctions (TJs) for paracellular and/or cell polarity barriers. The integral membrane proteins of gap junctions are connexins, whereas TJs are a complex between occludin and members of a recently characterized multigene family, the claudins. In normal brain, astrocytes are coupled by gap junctions composed primarily of connexin43 (Cx43), whereas TJs have not been detected in these cells. We now show that treatment of primary human astrocytes with the cytokine interleukin-1beta (IL-1beta) causes rapid induction of claudin-1, with an expression pattern reciprocal to loss of Cx43. Treatment also led to protracted downregulation of occludin but no change in expression of zonula occludens proteins ZO-1 and -2. Immunofluorescence staining localized claudin-1 to cell membranes in IL-1beta-treated astrocytes, whereas freeze-fracture replicas showed strand-like arrays of intramembranous particles in treated cells resembling rudimentary TJ assemblies. We conclude that in human astrocytes, IL-1beta regulates expression of the claudin multigene family and that gap and tight junction proteins are inversely regulated by this proinflammatory cytokine. We suggest that in pathological conditions of the human CNS, elevated IL-1beta expression fundamentally alters astrocyte-to-astrocyte connectivity.

In human astrocytes, interleukin-1beta (IL-1beta) is a potent inducer of genes associated with inflammation. In this study, we tested the hypothesis that in primary cultures of human fetal astrocytes signaling by the P2 purinergic nucleotide receptor pathway contributes to, or modulates, cytokine-mediated signal transduction. Calcium imaging studies indicated that most cells in culture responded to ATP, whereas only a subpopulation responded to UTP. Pretreatment of astrocytes with P2 receptor antagonists, including suramin and periodate oxidized ATP (oATP), resulted in a significant downregulation of IL-1beta-stimulated expression of nitric oxide, tumor necrosis factor (TNFalpha), and IL-6 at both the protein and mRNA levels, without affecting cell viability. In cells transiently transfected with reporter constructs, IL-1beta demonstrated more potent activation of the transcription factors nuclear factor -kappaB (NF-kappaB) and activator protein-1 (AP-1) than TNFalpha. However, pretreatment with oATP downregulated activation of NF-kappaB and AP-1 by IL-1beta or TNFalpha. Electromobility shift assays using oligonucleotides containing specific NF-kappaB binding sequences confirmed that pretreatment with oATP or apyrase attenuated cytokine-mediated induction of this transcription factor. From these data, we conclude that P2 receptor-mediated signaling intersects with that of IL-1beta and TNFalpha to regulate responses to cytokines in the CNS. Because inflammation, trauma, and stress all lead to the release of high levels of extracellular nucleotides, such as ATP and UTP, signaling via P2 receptors may provide a mechanism whereby cells can sense and respond to events occurring in the extracellular environment and can fine tune the transcription of genes involved in the inflammatory response.

In this study, the temporal and spatial expression of the chemotactic factor monocyte chemotactic protein-1 (MCP-1) was examined in the rabbit retina after challenge with the proinflammatory cytokine interleukin-1beta (IL-1). In these tissues, IL-1 induces an acute inflammatory response of the epiretinal vessels that peaks approximately 24 h postintraocular injection (pi) with the cytokine. At 2 h after challenge with IL-1, MCP-1 mRNA was expressed by perivascular microglial cells and astrocytes that form the glial limitans. Protein analysis at 3 h pi with IL-1 confirmed these sites of MCP-1 expression. The intensity of the mRNA and protein signals increased at 6 h and at 24 h. At these time points, MCP-1 message and protein also were detected in infiltrating macrophages and, at the latest time point, in endothelial cells as well. These data support the conclusion that IL-1 provides a strong stimulus for the rapid expression of MCP-1 mRNA and protein in retinal tissues, and they further support the role of endogenous glial cells as important sources of mediators involved in the regulation of inflammation occurring within the nervous system., (Copyright 2000 Wiley-Liss, Inc.)

In mammalian astrocytes, calcium waves are transmitted between cells via both a gap junction-mediated pathway and an extracellular, P2 receptor-mediated pathway, which link the cells into a syncytium. Calcium waves in astrocytes have also been shown to evoke calcium transients in neurons, and activity in neurons can elicit calcium waves in astrocytes. In this study, we show that in primary human fetal astrocytes, the P2 receptor-mediated and gap junction-mediated pathways are differentially regulated by the cytokine IL-1beta. Confocal microscopy of astrocytes loaded with Indo-1 demonstrated that intercellular calcium wave transmission in IL-1beta-treated cultures was potentiated compared with controls. However, transmission of calcium waves via the gap junction-mediated pathway was strikingly reduced. The major component of functional gap junctions in human fetal astrocytes was demonstrated to be connexin43 (Cx43), and there was a marked reduction of junctional conductance, loss of dye coupling, loss of Cx43 protein, and down-regulation of Cx43 mRNA expression after IL-1beta treatment of cultures. Conversely, transmission of calcium waves via the P2 receptor-mediated pathway was potentiated in IL-1beta-treated cultures compared with controls. This potentiation was associated with an increase in the number of cells responsive to UTP, and with a transient increase in expression of the P2Y(2) purinoceptor mRNA. Because in inflammatory conditions of the human central nervous system IL-1beta is produced both by resident glia and by invading cells of the immune system, our results suggest that inflammatory events may have a significant impact on coordination of astrocytic function and on information processing in the central nervous system.

In this work, we studied the expression of type II nitric oxide synthase (NOS) in primary cultures of human astrocytes and microglia. Cytokine-activated human fetal astrocytes expressed a 4.5-kb type II NOS mRNA that was first evident at 8 h, steadily increased through 48 h, and persisted through 72 h. The inducing signals for astrocyte NOS II mRNA expression were in the order IL-1beta + IFN-gamma > IL-1beta + TNF-alpha > IL-1beta. SDS-PAGE analysis of cytokine-stimulated astrocyte cultures revealed an approximately 130-kDa single NOS II band that was expressed strongly at 48 and 72 h (72 h > 48 h). Specific NOS II immunoreactivity was detected in cytokine-treated astrocytes, both in the cytosol and in a discrete paranuclear region, which corresponded to Golgi-like membranes on immunoelectron microscopy. In human microglia, cytokines and LPS failed to induce NOS II expression, while the same stimuli readily induced TNF-alpha expression. In cytokine-treated human astrocytes, neither NOS II mRNA/protein expression nor nitrite production was inhibited by TGF-beta, IL-4, or IL-10. In contrast, IL-1 receptor antagonist exerted near complete inhibition of NOS II mRNA and nitrite induction. Monocyte chemoattractant peptide-1 mRNA was induced in TGF-beta-treated astrocytes, demonstrating the presence of receptors for TGF-beta in astrocytes. These results confirm that in humans, cytokines stimulate astrocytes, but not microglia, to express NOS II belonging to the high output nitric oxide system similar to that found in rodent macrophages. They also show that the regulation of type II NOS expression in human glia differs significantly from that in rodent glia. A crucial role for the IL-1 pathway in the regulation of human astrocyte NOS II is shown, suggesting a potential role for IL-1 as a regulator of astrocyte activation in vivo.

Intra-vitreal injection of 300 U of interleukin (IL)-1 beta into the rabbit eye induces an inflammation of the retina characterized by hemorrhage, monocyte and neutrophil infiltration, and an increase in vascular permeability that peaks 24 h post-injection. Since the epiretinal vessels involved in this inflammation form part of the blood-retina barrier, we used this model to investigate the effects of the immunosuppressive cytokine TGF beta 1 on inflammation within the context of the central nervous system. We found that intra-vitreal injection of 1 microgram rh TGF beta administered concomitantly with rh IL-1 beta significantly reduced IL-1 beta-induced hemorrhage by 78%, and monocyte and neutrophil infiltration by 53% and 62%, respectively. In contrast, TGF beta did not reduce the IL-1 beta-induced increase in vascular permeability. However, TGF beta by itself caused a statistically significant increase in serum proteins in perfused tissues of the eye, to give a 3.1 +/- 0.4 fold increase in protein content over control values. No cellular inflammation accompanied this alteration in vascular permeability. These data indicate that whereas the local administration of TGF beta may be an effective inhibitor of cellular inflammation in the CNS, the effects on alterations in vascular permeability and accumulation of serum proteins may be more complex.

In this study we have shown that an antibody to CD18 identified a population of cells in the rabbit retina that resembled the perivascular macrophage found in other regions of the central nervous system. In the normal retina these cells possessed a ramified morphology and presented in an ordered array on the vitreal surface in association with the epiretinal vessels. Approximately 50% of the perivascular macrophages constitutively expressed MHC class II. In response to interleukin-1 beta (IL-1 beta)-induced inflammation, these cells became activated, as evidenced by a change from a ramified to an ameboid morphology and increased expression of MHC class II, and migrated away from the vessels. These changes were first detected around 3 h post-intraocular challenge coincident with the onset of inflammation. At the peak of the inflammatory response (approximately 24 h post-challenge), many activated perivascular macrophages were no longer associated with the vessels and formed long "cord" of MHC class II+ cells associated with underlying deposits of fibrin. In eyes challenged with heat-inactivated IL-1, no change in the morphology or distribution of the perivascular macrophage was noted. At 3 weeks post-challenge with IL-1, the number and distribution of the perivascular macrophages were restored to baseline values, although with a reduced cell size. Since these changes closely resemble those that occur in non-lymphoid dendritic cells in the skin, heart, and/or kidney following activation with cytokines or bacterial products, the results suggest that the perivascular macrophage represents the dendritic cell of the retina and may thus play an important role in immune surveillance in the eye and maintenance of the blood-retina barrier.

Recent work in many laboratories has revealed that cytokines are important mediators of inflammation, host defense, and tissue injury in a variety of neurological diseases. A role for astrocytes and microglia in these diseases has been considered pivotal, since both cell types readily produce and respond to cytokines in vitro and show morphologic and immunocytochemical evidence for activation in vivo. Although much of the work documenting these events has been generated in rodent systems, our laboratory has focused on human cell culture systems to define the nature of the activating signals for human microglia and astrocytes and their responses to activating cytokines and growth factors and evidence for activation. The results have shown that interleukin-1 (IL-1) is a potent activator of human astrocytes and induces cytokines such as tumor necrosis factor alpha and interleukin-6, and is a potent activator of nitric oxide generation in astrocytes. Astrocytes also promote microglial growth and differentiation through production of colony-stimulating factors, an activity that is enhanced following activation with IL-1. This review will summarize the human glia data generated in this and other laboratories and present hypotheses how glia may participate in certain human central nervous system diseases.

In human astrocyte cultures established from second-trimester fetal brain tissue, approximately 5-10% of total astrocyte population in unstimulated cultures were GD3+/glial fibrillary acidic protein (GFAP)+. The GD3+ cells were always GFAP+ and grew as flat, highly spread cells but changed to process-bearing cells after interleukin-1 beta (IL-1 beta) stimulation. It is interesting that IL-1 beta, a known mitogen for rat astrocytes, suppressed human fetal astrocyte proliferation as determined by [3H]thymidine incorporation, bromodeoxyuridine (BrdU) labeling, and cell counting. The GD3+ population, however, consistently increased in absolute number after IL-1 beta stimulation, in a dose- and time-dependent manner. The IL-1 beta-mediated increase in number of GD3+ astrocytes was independent of initial cell density or serum concentration. By flow cytometry, IL-1 beta enhanced both the mean fluorescence intensity and the percentage of GD3+ cells. To investigate whether the increase in GD3+ astrocyte cell number was due to proliferation of preexisting GD3+ astrocytes or due to conversion of GD3- to GD3+ cells, we performed BrdU/GD3 double immunocytochemistry. BrdU/GD3 double-positive cells were extremely rare in both control and IL-1 beta-stimulated cultures. Moreover, an increase in number of GD3+ astrocytes was still observed in control and IL-1 beta-stimulated cultures were GD3+ cells had been initially eliminated by cell sorting. These results indicate that GD3+ astrocytes in human fetal culture may represent a postmitotic, differentiated, distinct phenotype.

Previous studies in this and other laboratories have shown that interleukin-1 beta (IL-1 beta) is a selective and potent activator of human astrocytes with respect to induction of cytokines and hematopoietic growth factors. To study the effect of recombinant human IL-1 beta (rhIL-1 beta) on astrocyte morphology, glial fibrillary acidic protein (GFAP) and vimentin expression, and actin organization, we conducted a systematic survey using dissociated human fetal astrocyte cultures. Within hours of stimulation with IL-1 beta, the majority of astrocytes converted from flat, polygonal cells to small, contracted, highly branched cells. This change in morphology was more striking when serum was eliminated from the medium. Complete dissolution of filamentous actin occurred simultaneously with the change in cell shape, as demonstrated by fluorescein-phalloidin binding. These "activated" astrocytes displayed intense GFAP and vimentin immunoreactivity in the small perikarya and processes. In contrast, the large, flat astrocytes in control cultures showed diffuse pale immunoreactivity for GFAP and vimentin. To quantify the changes in GFAP and vimentin content with IL-1 beta stimulation, densitometric analyses of northern and western blots were performed. Northern blot analysis of IL-1 beta-stimulated astrocytes revealed a transient, marked decrease in steady-state levels of mRNA for GFAP, vimentin, and microtubule-associated protein 4. The decrease in mRNA levels was evident by 4-8 h and fell to the lowest level at 16-24 h (80-98% decrease by densitometry) with partial recovery by 72 h. By immunoblotting, a significant decrease in both GFAP and vimentin protein content was observed after IL-1 beta stimulation. Furthermore, metabolic labeling studies revealed an almost total loss of GFAP synthesis following stimulation with IL-1 beta for 16 h. These observations are consistent with the idea that increases in immunoreactivity were related to factors such as redistribution of epitope, rather than increases in total protein content. We hypothesize that in IL-1 beta-stimulated astrocytes, synthesis of other proteins, e.g., inflammatory cytokines, occurs at the expense of structural proteins and that the decrease in content of cytoskeletal proteins may reflect an "activated" state of astrocytes.

Background: During inflammatory conditions of the central nervous system (CNS), the protective function of the blood-brain barrier (BBB) may be compromised, resulting in CNS edema. However, it is not well understood how inflammatory cells may increase BBB permeability, since increased transendothelial transport of serum proteins is observed in CNS capillaries that are not directly in contact with inflammatory cells. One possible explanation may be that soluble inflammatory factors may cause BBB changes, since pathologic conditions that increase circulating cytokines produce detectable increases in BBB permeability., Experimental Design: To investigate the role of inflammatory cytokines in induction of endothelial cell changes and inflammation in the CNS, we utilized the rabbit retinal system as a model. This system shows vascularization similar to the BBB, and is termed the blood-retina barrier (BRB). The rabbit visual system allows injection of cytokines, causing minimal trauma, and the contralateral eye serves as an intra-animal control., Results: Ultrastructural morphometric analysis of vesicular content in BRB endothelium showed significant increase at 3 hours postintravitreal injection of interleukin-1 beta (IL-1 beta) or tumor necrosis factor-alpha (TNF-alpha). Increased transport did not correlate with increased vitreal protein. However, intravascular tracer (horseradish peroxidase) revealed that pericytes, Müller cells, and perivascular microglia accumulate serum proteins, thus acting as sinks for extravasated proteins after BRB disruption. The IL-1 beta-induced inflammatory response was characterized by polymorphonuclear and mononuclear cells, whereas the TNF-alpha-induced response was less intense and comprised monocytes and occasional eosinophils. At the height of inflammation, IL-1 beta produced large gaps between endothelial cells that allowed for extensive cellular inflammation and hemorrhage. TNF-alpha induced necrotic changes on endothelial cells, being most severe at 3 hours postintravitreal injection, revascularization was noted at 24 hours postintravitreal injection., Conclusions: These results demonstrate that proinflammatory effects of IL-1 beta and TNF-alpha in the BRB initiate many of the changes associated with inflammation of the CNS vasculature, such as those induced during experimental autoimmune encephalitis and multiple sclerosis. Once the permeability of the BRB endothelium is increased, perivascular phagocytic cells such as perivascular, microglia and Müller cells may act as secondary barriers to extravasated proteins.

Nitric oxide (NO) is a short-lived, diffusible molecule that has a variety of biological activities including vasorelaxation, neurotransmission, and cytotoxicity. In the central nervous system, a constitutive form of nitric oxide synthase (NOS) has been localized in a subset of neurons and in endothelial cells. In addition, both constitutive and LPS-inducible NOS has been demonstrated in rat astrocytes and microglia in vitro. In this report, we present evidence for the production of NO, as measured by the production of nitrite, in highly enriched human fetal astrocyte cultures stimulated with IL-1 beta. The production of nitrite paralleled the induction of NADPH diaphorase enzyme activity in the perikarya of the majority of stimulated astrocytes. The IL-1 beta-induced nitrite production by astrocytes was markedly enhanced when cells were co-stimulated with IFN-gamma or TNF-alpha (IFN-gamma > TNF-alpha); LPS had no effect used as a single agent or in combination with other cytokines. NGMMA and NG-nitro-arginine, competitive inhibitors of NOS, diminished the accumulation of nitrite, but calmodulin antagonists (trifluoperazine, W-5 and W-7) had little or no inhibitory effect. Human fetal microglia, in contrast to astrocytes, failed to secrete significant amounts of nitrite in response to various stimuli. The results demonstrate the presence of an inducible form of NOS in human fetal astrocytes; human microglia, in turn, may control astrocyte NO production by providing IL-1 beta as an activating signal.

Astrocytes produce factors that control the growth and differentiation of many cell types within the CNS as well as play a role in the generation of the immune response. The extent to which these two functions interact has received less attention. We now report that astrocyte cultures established from rat brain endogenously express mRNA and low levels of secreted biologically active protein for the monocyte growth and differentiation factor colony stimulating factor-1 (CSF-1). Exposure of astrocytes to interleukin-1 (IL-1) and/or tumor necrosis factor (TNF) upregulated the expression of CSF-1 mRNA and protein. Following treatment with 100 U/ml of TNF, IL-1, or TNF+IL-1, maximum CSF-1 mRNA expression was observed at 3 hr. In the presence of IL-1 an increase in biologically active CSF-1 was detected in the astrocyte conditioned medium at 6 hr. These data indicate that the expression of CSF-1 by astrocytes can be modulated by exposure to the cytokines IL-1 and TNF. To determine whether CSF-1 provides a mitogenic signal for microglia during development, mouse spinal cord organotypic cultures were exposed to recombinant mouse CSF-1 (rmCSF-1), resulting in proliferation of microglia by 7 days and an increase in the number of ramified microglia over ameboid microglia by 14 days.

As part of a study on the role of cytokines in central nervous system development and dysfunction, we determined the pattern of cytokine production in highly purified cultures of microglia and astrocytes isolated from second-trimester human fetal brains. Levels of TNF-alpha, IL-1 beta, and IL-6 mRNA and protein were determined by Northern blot analysis and ELISA before and after stimulation with LPS, TNF-alpha, or IL-1 beta. In microglia, LPS induced mRNA for all three cytokines. High protein levels of IL-6 and TNF-alpha were also found in the medium, whereas IL-1 beta protein was mostly cell associated. IL-1 beta also induced message for all three cytokines, in the rank order of IL-1 beta > IL-6 > TNF-alpha. TNF-alpha induced mRNA and protein for IL-1 beta but not for TNF-alpha or IL-6. In contrast, LPS failed to stimulate either mRNA or protein expression for any of the three cytokines in astrocytes. On the other hand, IL-1 beta provided a strong stimulus for astrocytes. IL-1 beta induced mRNA and protein for both TNF-alpha and IL-6, but the kinetics of the response differed for the two cytokines. TNF-alpha mRNA and protein levels peaked early (at 4 h and 16 h, respectively) and were undetectable by 72 h, whereas IL-6 mRNA peaked later (at 16 h) and protein levels continued to accumulate in the medium through 72 h. IL-1 beta did not induce IL-1 beta mRNA or protein in astrocytes. TNF-alpha did not induce expression of any of the cytokines in astrocytes. In conclusion, our results demonstrate that cytokine production can be induced in human fetal microglia and astrocytes but that the stimuli for induction differed significantly for the two cell types. Whereas LPS was a potent stimulus for microglia, astrocytes primarily responded to IL-1 beta. The data further suggest that microglia may be key regulators of astrocyte response, working primarily through the expression of cell-associated IL-1 beta.

CSF-1 is a growth factor that selectively promotes the proliferation, survival, and differentiation of cells of the mononuclear phagocyte series. As part of a study on the role of cytokine and hematopoietic growth factors in central nervous system (CNS) development and inflammation, we examined the expression of CSF-1 in dissociated glial cells cultured from human fetal CNS tissue. CSF-1 mRNA and protein were constitutively expressed by astrocytes. The steady state level of CSF-1 mRNA was markedly up-regulated by both IL-1 beta and TNF-alpha in a time- and dose-dependent manner, whereas only a minimal increase was detected after stimulation with LPS. In unstimulated astrocyte cultures, CSF-1 protein levels gradually increased to 3.5-fold base-line values by 96 h and were significantly increased by all three stimulants in the order of IL-1 > or = TNF > LPS. Low levels of CSF-1 mRNA and protein were also detected in unstimulated microglia cultures. In contrast to astrocyte cultures, CSF-1 mRNA and protein increased significantly after stimulation with LPS, but changed only minimally after exposure to TNF-alpha or IL-1 beta. The effect of CSF-1 on cell proliferation, morphology, and class II MHC Ag expression was determined in highly enriched cultures of microglia and astrocytes. Microglia treated with CSF-1 showed a modest level of proliferation and differentiation into rod-shaped cells, whereas neither cell number nor shape was changed in astrocyte cultures. Interestingly, marked inhibition of both basal and IFN gamma-induced class II MHC Ag expression was observed in microglial cells cultured in the presence of CSF-1, whereas no effect was detected in astrocytes. These results suggest the possibility that in situ production of CSF-1 in the CNS may regulate normal glial cell development and contribute to the immunologic status of the CNS through the down-regulation of class II MHC expression.

The chronic effects of human recombinant IL-1 (hrIL-1) on the specialized vasculature of the central nervous system (CNS) and on the CNS itself have been examined over a 35-day period in the rabbit retina. A single intraocular injection of physiological levels of hrIL-1 (300 units) induced a biphasic inflammatory reaction with well-defined acute and chronic phases in the challenged eye. Quantitative histopathological examination of the vascularized portion of the retina in the IL-1-challenged eye documented a persistent mononuclear (MN) cell response that peaked 7-14 days post-challenge. Included in the MN cell count were perivascular plasma cells. Elevated protein levels in the vitreous persisted throughout the time points studied and alterations in vascular permeability of the epiretinal vessels were demonstrated by tracer leakage at 2 weeks post-challenge. The results show that exposure of the CNS-vasculature to IL-1 induces long-lasting inflammatory changes typical of a chronic inflammatory reaction.

Interleukin-1 is a potent immunomodulator and has been shown to initiate many aspects of the inflammatory response. To determine the effects of IL-1b in the central nervous system (CNS), the rabbit retina was used, adjacent to which factors can be injected with minimal trauma and both pathologic and physiologic effects can be monitored. Intravitreal injection of 300 units of IL-1b induced an alteration in the visual evoked potentials (VEP) that was associated with marked intravascular red blood cell accumulations, hemorrhage, and cellular inflammation of the epiretinal vessels. Analysis of these events showed slowing and occasional hyper-excitability of the compound action potential of the optic tract and of the cortical VEP that correlate with the maximum inflammatory response. Histologic studies show the following: no apparent response occurs within the first 1.5 hours after intraocular challenge; and between 3 and 6 hours after injection an extensive intravascular red blood cell accumulation and progressive hemorrhage is accompanied by an increase in the number of mononuclear (MN) cells and the appearance of polymorphonuclear (PMN) cells. Polymorphonuclear cells continue to increase with time to give a single wave of inflammation that peaks 24 hours after injection, while the number of MN cells steadily increases. These events are associated with changes in the permeability of the blood-brain barrier and correlate with the electrophysiologic dysfunctions. Forty-one hours after injection, MN inflammation, reactive gliosis, and residual PMN inflammation are evident. Neutralization with specific antibody inhibited the responses through 6 hours after injection. It is concluded that the rabbit retina provides a valuable model for the in vivo analysis of CNS inflammation.