Interleukin 10 reduces the release of tumor necrosis factor and prevents lethality in experimental endotoxemia.

Because of its ability to efficiently inhibit in vitro cytokine production by activated macrophages, we hypothesized that interleukin (IL) 10 might be of particular interest in preventing endotoxin-induced toxicity. We therefore examined the effects of IL-10 administration before lipopolysaccharide (LPS) challenge in mice. A marked reduction in the amounts of LPS-induced tumor necrosis factor (TNF) release in the circulation was observed after IL-10 pretreatment at doses at low as 10 U. IL-10 also efficiently prevented the hypothermia generated by the injection of 100 micrograms LPS. Finally, pretreatment with a single injection of 1,000 U IL-10 completely prevented the mortality consecutive to the challenge with 500 micrograms LPS, a dose that was lethal in 50% of the control mice. We conclude that IL-10 inhibits in vivo TNF secretion and protects against the lethality of endotoxin in a murine model of septic shock.

In vivo induction of interleukin 10 by anti-CD3 monoclonal antibody or bacterial lipopolysaccharide: differential modulation by cyclosporin A.

We investigated the in vivo effects of cyclosporin A (CsA) on the production of interleukin (IL) 10, a cytokine with major immunosuppressive properties. To elicit IL-10 production in vivo, BALB/c mice were injected either with the anti-mouse CD3 145-2C11 monoclonal antibody (mAb) (25 micrograms) or with bacterial lipopolysaccharide (LPS) (20 micrograms). A systemic release of IL-10 was observed in both models, IL-10 serum levels reaching 1.60 +/- 0.32 U/ml (mean +/- SEM) and 0.67 +/- 0.09 U/ml 6 h after injection of 145-2C11 mAb and LPS, respectively. Experiments in nude mice indicated that T cells are involved in the induction of IL-10 by anti-CD3 mAb, but not by LPS. Pretreatment with CsA (total dose: 50 mg/kg) before injection of 145-2C11 mAb completely prevented the release of IL-10 in serum as well as IL-10 mRNA accumulation in spleen cells. In contrast, CsA markedly enhanced LPS-induced IL-10 release (IL-10 serum levels at 6 h: 8.31 +/- 0.43 vs. 0.71 +/- 0.15 U/ml in mice pretreated with CsA vehicle-control, p < 0.001), as well as IL-10 mRNA accumulation in spleen. We conclude that CsA differentially affects IL-10 production in vivo depending on the nature of the eliciting agent. This observation might be relevant to clinical settings, especially in organ transplantation.

Epidermal-growth-factor-dependent transformation by a human EGF receptor proto-oncogene.

The epidermal growth factor (EGF) receptor gene EGFR has been placed in a retrovirus vector to examine the growth properties of cells that experimentally overproduce a full-length EGF receptor. NIH 3T3 cells transfected with the viral DNA or infected with the corresponding rescued retrovirus developed a fully transformed phenotype in vitro that required both functional EGFR expression and the presence of EGF in the growth medium. Cells expressing 4 x 10(5) EGF receptors formed tumors in nude mice, while control cells did not. Therefore, the EGFR retrovirus, which had a titer on NIH 3T3 cells that was greater than 10(7) focus-forming units per milliliter, can efficiently transfer and express this gene, and increased numbers of EGF receptors can contribute to the transformed phenotype.

Functional heterogeneity of proto-oncogene tyrosine kinases: the C terminus of the human epidermal growth factor receptor facilitates cell proliferation.

Previous reports have indicated that the C termini of the membrane-associated tyrosine kinases encoded by c-src and c-fms proto-oncogenes have a negative effect on their biological activity and that this effect is mediated by their C-terminal tyrosine residue. To determine whether this was true for the human epidermal growth factor (EGF) receptor, which is also a membrane-associated tyrosine kinase proto-oncogene, we have constructed two premature termination mutants, dc19 and dc63, that delete the C-terminal 19 and 63 amino acids, respectively, from the human full-length receptor (hEGFR). The smaller deletion removes the C-terminal tyrosine residue, while the larger deletion removes the two most C-terminal tyrosines; similar deletions are found in v-erbB. As previously shown for the gene encoding the full-length EGF receptor, the two C-terminal mutants induced EGF-dependent focal transformation and anchorage-independent growth of NIH 3T3 cells. However, both dc19 and dc63 were quantitatively less efficient than the gene encoding the full-length receptor, with dc63 being less active than dc19. Although the C-terminal mutants displayed lower biological activity than the gene encoding the full-length receptor, the mutant receptors were found to be similar in several respects to the full-length receptor. These parameters included receptor localization, stability in the absence of EGF, receptor half-life in the presence of EGF, EGF binding, extent of EGF-dependent autophosphorylation in vitro, and EGF-dependent phosphorylation of an exogenous substrate in vitro. Therefore, the C-terminal 63 amino acids of the human receptor have no detectable influence on EGF-dependent early events. We conclude that in contrast

The bovine papillomavirus E5 oncogene can cooperate with ras: identification of p21 amino acids critical for transformation by c-rasH but not v-rasH.

We have previously used a series of insertion-deletion mutants of the mutationally activated v-rasH gene to identify several regions of the encoded protein that are dispensable for cellular transformation (B. M. Willumsen, A. G. Papageorge, H.-F. Kung, E. Bekesi, T. Robins, M. Johnsen, W. C. Vass, and D. R. Lowy, Mol. Cell. Biol. 6:2646-2654, 1986). To determine if some of these amino acids are more important for the biological activity of c-rasH, we have now tested many of the same insertion-deletion mutants in the c-rasH form for their ability to transform NIH 3T3 cells. Since the transforming activity of c-rasH is low, we have used cotransfection with the bovine papillomavirus (BPV) genome to develop a more sensitive transformation assay for c-rasH mutants. The increased sensitivity of the assay, which is seen both in focal transformation and in anchorage-independent growth, is mediated by cooperation between the BPV E5 gene and ras. E5-dependent cooperation was seen for v-rasH as well as for c-rasH, which suggests that the major effect of E5 was to increase the susceptibility of the cell to transformation to a given level of ras activity. The cooperation assay was used to test the potential importance, in c-rasH, of codons 93 to 108, 123 to 130, and 166 to 183, which were nonessential for v-rasH transformation. Relative to the respective transforming activity of wild-type c-rasH and v-rasH, mutants with lesions in codons 102 and 103 were significantly less active in their c-rasH forms than in their v-rasH forms. We conclude that a region including amino acids 102 and 103 encodes a function that is more critical to c-rasH than to v-rasH. Guanine nucleotide exchange is one function that is compatible with such a phenotype.

Skin biopsies in DC vaccines for stage III-IV melanoma patients: role of neutrophils?

Dendritic cell (DC) vaccines are used for the induction of anti-tumor T cell reaction in melanoma patients. DC are generated in vitro, pulsed with antigen and matured prior to injection. They are supposed to migrate to lymph nodes and to present the processed antigen to naive T cells allowing activation of tumor-specific lymphocytes. It has been suggested that intradermal injection allows a superior migration to the lymph node. Eight HLA-A2 positive patients with stage III or IV melanomas expressing NA 17 antigen were collected. They were included in a pilot trial of vaccination in which they received IL3/INFb DC presenting the NA17 A2 antigen. In each patient, a skin biopsy was performed at the injection site, 24 h after inoculation. The striking features of the biopsies were the presence of a perivascular CD3+/CD8+ T cell infiltrate with a slight population of CD4+ cells and the presence of a massive neutrophilic infiltrate associated with the injected DC still present, realizing a suppurative granuloma. The persistence of DC 24 h after the injection suggests that migration in the lymph node is not necessary for the induction of the immune response. The skin itself could be the location of a reaction starting with a massive recruitment of neutrophils.

Overexpression of mouse IsK protein fused to green fluorescent protein induces apoptosis of human astroglioma cells.

Intracellular K(+) plays an important role in controlling ion homeostasis for maintaining cell volume and inhibiting activity of pro-apoptotic enzymes. Cytoplasmic K(+) concentration is regulated by K(+) uptake via Na(+) -K(+) -ATPase and K(+) efflux through K(+) channels in the plasma membrane. The IsK (KCNE1) protein is known to co-assemble with KCNQ1 (KvLQT1) protein to form a K(+) channel underlying the slowly activating delayed rectifier K(+) outward current which delays voltage activation. In order to further study the activity and cellular localization of IsK protein, we constructed a C-terminal fusion of IsK with EGFP (enhanced green fluorescent protein). Expression of the fusion protein appeared as clusters located in the plasma membrane and induced degeneration of both transiently or stably transfected cells.

The biological activity of the human epidermal growth factor receptor is positively regulated by its C-terminal tyrosines.

The epidermal growth factor receptor (EGF-R) C-terminus contains three conserved tyrosines (Y-1068, Y-1148, Y-1173) which are phosphorylated upon EGF activation. To clarify the functional role of these tyrosines, each has been mutated to phenylalanine and studied as single, double and triple mutants in the full length receptor. EGF-dependent transforming ability of the single point mutants is similar to that of the wild type, while that of double mutants is decreased and an even lower activity is present in the triple mutant. In each bioassay, including EGF-dependent focal transformation, growth in agar and growth in low serum, mutant receptors display a similar hierarchy of activity. The lower activity is intrinsic in the mutants since they are expressed at similar level as the wild type and bind EGF with similar affinity. Deletion mutants lacking the last 19 or 63 amino acids (Velu et al., 1989a) show a similar decline in biological activity when compared to the corresponding point mutants, although the reduction is more pronounced than with the point mutants. Deletion of the last 123 aa, which removes all three tyrosines (Dc123), results in a receptor that is almost inactive biologically. The EGF-R kinase activity is affected by tyrosine substitution since in vitro phosphorylation of exogenous substrates is reduced in the double and triple mutants. Autophosphorylation, in vivo and in vitro, is also reduced, but not totally abolished in the triple point mutant and Dc123 indicating the existence of other autophosphorylation sites. A new site of autophosphorylation is found in the Dc123 mutant. We conclude, therefore, that the tyrosines at the extreme C-terminus positively regulate the biological and transforming activity of the EGF-R, probably via autophosphorylation.

Transient expansion of peptide-specific lymphocytes producing IFN-gamma after vaccination with dendritic cells pulsed with MAGE peptides in patients with mage-A1/A3-positive tumors.

Assessment of T-cell activation is pivotal for evaluation of cancer immunotherapy. We initiated a clinical trial in patients with MAGE-A1 and/or -A3 tumors using autologous DC pulsed with MAGE peptides aimed at analyzing T-cell-derived, IFN-gamma secretion by cytokine flow cytometry and ELISPOT. We also tested whether further KLH addition could influence this response favorably. Monocyte-derived DC were generated from leukapheresis products. They were pulsed with the relevant MAGE peptide(s) alone in group A (n=10 pts) and additionally with KLH in group B (n=16 pts). A specific but transient increase in the number of peripheral blood T lymphocytes secreting IFN-gamma in response to the vaccine peptide(s) was observed in 6/8 patients of group A and in 6/16 patients of group B. We conclude that anti-tumor vaccination using DC pulsed with MAGE peptides induces a potent but transient anti-MAGE, IFN-gamma secretion that is not influenced by the additional delivery of a nonspecific, T-cell help.

Spontaneous and cycloheximide-induced interleukin-10 mRNA expression in human mononuclear cells.

Interleukin-10 (IL-10) is known to be spontaneously produced by human peripheral blood mononuclear cells. In order to define the cell type in which IL-10 gene is spontaneously expressed we used the reverse polymerase chain reaction for IL-10 mRNA expression, which was also used to study the effects of cycloheximide (CHX). First, we found that IL-10 mRNA is spontaneously expressed in monocytes and B cells but not T cells from healthy donors, and second, we demonstrated that CHX superinduces IL-10 mRNA in monocytes and B cells. Experiments including nuclear run-on analyses established that the effects of CHX on IL-10 gene expression involve both gene transcription and mRNA stabilization.

Influence of sequence and size of DNA on packaging efficiency of parvovirus MVM-based vectors.

We have derived a vector from the autonomous parvovirus MVM(p), which expresses human IL-2 specifically in transformed cells (Russell et al., J. Virol 1992;66:2821-2828). Testing the therapeutic potential of these vectors in vivo requires high-titer stocks. Stocks with a titer of 10(9) can be obtained after concentration and purification (Avalosse et al., J. Virol. Methods 1996;62:179-183), but this method requires large culture volumes and cannot easily be scaled up. We wanted to increase the production of recombinant virus at the initial transfection step. Poor vector titers could be due to inadequate genome amplification or to inefficient packaging. Here we show that intracellular amplification of MVM vector genomes is not the limiting factor for vector production. Several vector genomes of different size and/or structure were amplified to an equal extent. Their amplification was also equivalent to that of a cotransfected wild-type genome. We did not observe any interference between vector and wild-type genomes at the level of DNA amplification. Despite equivalent genome amplification, vector titers varied greatly between the different genomes, presumably owing to differences in packaging efficiency. Genomes with a size close to 100% that of wild type were packaged most efficiently with loss of efficiency at lower and higher sizes. However, certain genomes of identical size showed different packaging efficiencies, illustrating the importance of the DNA sequence, and probably its structure.

Loss of tumorigenicity and increased immunogenicity induced by interleukin-10 gene transfer in B16 melanoma cells.

Because interleukin-10 (IL-10) has potent immunosuppressive and anti-inflammatory properties and is produced by some cancers, we hypothesized that its production might play a role in carcinogenesis by inhibiting adequate antitumoral immune responses. To test this hypothesis, retroviral vectors containing the IL-10 cDNA were generated and used to infect B16F1 melanoma cells that were injected subcutaneously in syngeneic mice. Surprisingly, IL-10 gene transfer resulted in a loss of tumorigenicity that was proportional to the amount of IL-10 secreted. Histological analysis showed massive area of necrosis of these tumor cells, with infiltration of polymorphic inflammatory cells. Parental cells simultaneously implanted had decreased tumorigenicity only when mixed with IL10-producing cells, but not when injected contralaterally, suggesting that their eradication is mediated mostly by a local phenomenon. Host T lymphocytes and natural killer (NK) cells were involved in this eradication because IL-10-producing cells grew in nude mice and in CD8+ or NK-depleted mice. Finally, mice injected with IL-10-secreting cells developed an antitumoral systemic immune response able to protect them against a subsequent challenge with parental cells. These results demonstrate that, in some settings, IL10 may have in vivo immunostimulating and proinflammatory properties that need to be considered in its therapeutic development.

Curative potential of herpes simplex virus thymidine kinase gene transfer in rats with 9L gliosarcoma.

The transfer of the gene coding for the thymidine kinase of the herpes simplex virus (HSV-tk), followed by ganciclovir (GCV) administration, has been described for the treatment of several types of cancer, especially brain tumors. We further studied the efficacy of this approach by using the 9L rat gliosarcoma model, and cells producing 5 x 10(3), 9 x 10(4), 3 x 10(5) HSV-tk retroviral particles per milliliter. Their stereotactic injection in 9L brain tumors and GCV treatment did not result in any increase of survival. To study a model of optimal in vivo transduction, we examined the survival of rats with tumors growing from 9L cells that had been previously transduced in vitro with the HSV-tk vectors (9LTk cells). We observed that GCV administration cured 26% (n = 42) of the animals with 9LTk brain tumors, with most of the relapsing tumors remaining HSV-tk positive. The increase of either the dose or the duration of GCV treatment did not improve the survival rate. But the cerebral localization of the tumor played an important role, because this survival rate reached 67% (n = 12) when similar tumors were growing subcutaneously. No or only marginal antitumoral responses were induced by the presence of a selectable marker gene in the HSV-tk vectors. These results demonstrate that in vitro HSV-tk gene transfer in 9L tumor cells, but not in vivo gene transfer, followed by GCV treatment, is able to cure rats at a rate that is higher for subcutaneous than for intracerebral tumors.

Gene therapy: socioeconomic and ethical issues. A roundtable discussion.

Gene therapy research has the potential to revolutionize the way in which many human diseases are treated. Despite its enormous potential, roundtable panelists concluded that the field needs time to mature scientifically without pressure to develop a marketable therapeutic product. In addition, health care decision makers, physicians, and the lay public need to be educated on the future medical, economic, and ethical ramifications of gene therapy.

Lack of intrathyroidal tumor necrosis factor alpha in Graves' disease.

In vitro tumor necrosis factor alpha (TNF alpha) exerts a synergistic action on HLA class II expression and a cytotoxic action in FRTL-5 cells. Therefore, a role for TNF alpha as a local mediator of cell destruction in thyroid autoimmunity has been postulated. To elucidate the in vivo significance of these and other in vitro findings for the pathophysiology of Graves' disease we investigated 11 thyroid glands of patients suffering from Graves' disease for TNF alpha. In situ hybridization was done with a TNF alpha probe synthesized with T7 polymerase on a 750-base pair EcoRI fragment of the coding region. Primers at positions 152 and 854 in the TNF alpha copy DNA sequence were used for reverse polymerase chain reaction (PCR) amplification of TNF alpha in 5 patients. Immunohistological staining for TNF alpha was done with a mouse monoclonal antibody. We could not detect any TNF alpha and TNF alpha messenger RNA in thyroid tissue of 11 patients suffering mostly from relapsing Graves' disease by immunohistology and in situ hybridization as well as reverse PCR, respectively. A faint signal could be detected by reverse PCR in control thyroid tissue from a patient with recurrent goiter. This lack of intrathyroidal TNF alpha in relapsing Graves' disease is in accordance with a lack of increased TNF alpha production by T cell clones isolated from Graves' disease thyroid glands and contrasts with previous in vitro results. Since protective TNF actions have been demonstrated in other autoimmune diseases it could therefore be envisaged that the lack of intrathyroidal TNF alpha may be associated with the relapse of Graves' disease in our patients.

The role of cellular- and prodrug-associated factors in the bystander effect induced by the Varicella zoster and Herpes simplex viral thymidine kinases in suicide gene therapy.

To investigate the factors influencing the bystander effect--a key element in the efficacy of suicide gene therapy against cancer--we compared the effect triggered by four extremely efficient gene/prodrug combinations, i.e., VZVtk/BVDU, the thymidine kinase of Varicella zoster virus associated with (E)-5-(2-bromovinyl)-2'-deoxyuridine; VZVtk/BVaraU, the same enzyme associated with (E)-5-(2-bromovinyl)-1-beta-D-arabinofuranosyluracil; HSVtk/BVDU, the association of the Herpes simplex virus thymidine kinase with BVDU; and the classical HSVtk/GCV (ganciclovir) paradigm. The cells used, the human MDA-MB-435 breast cancer, and the rat 9L glioblastoma lines were equally sensitive in vitro to these four associations. In both cell types, the combinations involving pyrimidine analogues (BVDU, BVaraU) displayed a smaller bystander killing than the combination involving the purine analogue (GCV). In addition, the bystander effect induced by all the tk/prodrug systems was reduced in MDA-MB-435 cells in comparison to 9L cells; albeit, the viral kinases were produced at a higher level in the breast cancer cells. All systems induced apoptotic death in the two cell types, but the MDA-MB-435 cells, deprived of connexin 43, were noncommunicating in striking contrast with the 9L cells. That functional gap junctions have to be increased in order to improve the breast cancer cell response to suicide gene therapy was demonstrated by transducing the Cx43 gene: this modification enhanced the bystander effect associated in vitro with GCV treatment and, by itself, decreased the tumorigenicity of the untreated cells. However, the noncommunicating MDA-MB-435 cells triggered a significant bystander effect both in vitro and in vivo with the HSVtk/GCV system, showing that communication through gap junctions is not the only mechanism involved.

Comparative in vitro and in vivo cytotoxic activity of (E)-5-(2-bromovinyl)-2'-deoxyuridine (BVDU) and its arabinosyl derivative, (E)-5-(2-bromovinyl)-1-beta-D-arabinofuranosyluracil (BVaraU), against tumor cells expressing either the Varicella zoster or the Herpes simplex virus thymidine kinase.

The inhibitory effects of (E)-5-(2-bromovinyl)-2'-deoxyuridine (BVDU) and its arabinosyl derivative (E)-5-(2-bromovinyl)-1-beta-D-arabinofuranosyluracil (BVaraU) on the growth of both MDA-MB-435 human breast carcinoma and 9L rat gliosarcoma cells expressing the thymidine kinase (tk)-encoding gene of the Varicella zoster virus (VZV) or the Herpes simplex virus (HSV) were evaluated. In vitro, BVDU and BVaraU effectively killed both cell types expressing VZVtk, with 50% inhibitory concentration values ranging from 0.06 to 0.4 microM, whereas ganciclovir (GCV) lacked activity. On HSVtk+ cells, BVDU had high cytotoxic activity, with 50% inhibitory concentration values that were similar to those of GCV, whereas BVaraU was inactive. In vivo, BVDU applied intraperitoneally caused a 50% tumor growth inhibition in nude mice inoculated subcutaneously with VZVtk+ as well as HSVtk+ mammary tumor cells. In mice and at variance with the in vitro results, BVaraU had very little activity against the VZVtk+ mammary cells; GCV had the highest activity on the HSVtk+ cells, resulting in a 50% eradication of the tumors. With the 9L rat gliosarcoma model, the VZVtk/BVDU system completely failed to inhibit the development of VZVtk+ glioma tumors induced subcutaneously in syngeneic rats, although BVDU had a similar 45-minute half-life in both rats and mice. Factors other than degradation of the prodrug and related to the mode of action of these analogs are possibly involved in the observed discrepancies between the in vitro and in vivo results.

Neuroprotective gene therapy for Parkinson's disease.

Parkinson's disease (PD) is a neurodegenerative disease characterised by a progressive loss of the dopaminergic neurones in the substantia nigra pars compacta. Accumulating evidence indicates that apoptosis contributes to neuronal cell death in PD patients' brain. Excitotoxicity, oxidative stress, and mitochondrial respiratory failure are thought to be the key inducers of the apoptotic cascade. Even though the initial cause and the mechanism of degeneration are poorly understood, neuroprotection can be achieved by interfering with neuronal cell death either directly or by preventing neuronal dysfunction. Potential agents for neuroprotection are neurotrophic factors, inhibitors of apoptosis or anti-oxidative agents. However, the existence of the blood-brain barrier precludes systemic delivery of these factors. In situ gene delivery provides strategies for local and sustained administration of protective factors at physiologically relevant doses. Viral vectors mediating stable gene expression in the central nervous system exist and are still under development. Efficacy of these vectors has repeatedly been demonstrated in the animal models both ex vivo and in vivo. Ex vivo gene delivery could furthermore be combined with cell replacement therapies by transplanting genetically modified cells compensating for the lost neuronal cell population in order to provide neuroprotection to both the grafted cells and degenerating host neurones. However, several aspects of gene transfer, such as uncontrolled diffusion, axonal transport, unpredictable site of integration and immunological responses, still raise safety concerns and justify further development of viral and non-viral vectors as well as genetic elements with tightly controlled gene expression. Various relevant animal models for Parkinson's disease are available for the evaluation of gene therapy strategies. These include induction of cell death in specific neurone population through administration of toxins either directly in the brain or systemically, as well as transgenic mice expressing human disease-associated mutations.

Modulation of the release of cytokines and reduction of the shock syndrome induced by anti-CD3 monoclonal antibody in mice by interleukin-10.

Since IL-10 was recently shown to inhibit several T cell functions in vitro, we investigated the effects of IL-10 on the cytokine release syndrome induced in mice by the 145-2C11 anti-CD3 mAb. As OKT3 in man, this mAb induces a massive polyclonal T cell activation before to induce immunosuppression. First, we found that administration of 1000 U of recombinant mouse IL-10 (mIL-10) 30 min before injection of 10 micrograms of the 145-2C11 antimouse CD3 mAb markedly reduced the systemic release of IFN-gamma and TNF. In contrast, IL-10 pretreatment did not significantly modify the release of IL-6. To determine the effect of IL-10 pretreatment on the endogenous secretion of IL-10 induced by the 145-2C11 mAb, mice were injected with human IL-10 (hIL-10) which does not cross-react in the ELISA for mIL-10 determination. While hIL-10 was as efficient as mIL-10 in reducing TNF and IFN-gamma release, it did not modify peak serum levels of IL-10. The modulation of cytokine production by mIL-10 was associated with a significant reduction of the toxicity of the 145-2C11 mAb, as assessed by the attenuation of hypothermia and by the reduced lethality in D-galactosamine-sensitize mice. We conclude that IL-10 differentially regulates the in vivo production of cytokines and decreases the systemic toxicity induced by the 145-2C11 mAb. These observations suggest potential therapeutic applications of IL-10 in organ transplantation, especially in association with anti-CD3 mAb.

Role and expression of CD40 on human retinal pigment epithelial cells.

PURPOSE: To examine the CD40 costimulatory molecule expression on normal resting or activated adult human retinal pigment epithelium (hRPE) cells and to evaluate its role as an activation molecule considering the potential antigen presentation functions of hRPE cells. METHODS: Expression of HLA-DR and costimulatory (CD40, B7.1, B7.2, CD54, and CD58) molecules on hRPE cells was analyzed by flow cytometry. CD40 triggering was performed using soluble CD40L or cocultures with CD40L transfected fibroblasts. Interleukin (IL)-6, -8, -10, and -12 secretions were measured by enzyme-linked immunosorbent assay. Antigen presentation function of hRPE cells was assessed by coculturing hRPE cells with allogeneic T cells. T-cell proliferation was measured by [(3)H]-thymidine incorporation, and T-cell apoptosis by measurement of caspase-3 activity. RESULTS: Interferon (IFN)gamma-activated hRPE cells expressed CD40, but not B7.1 or B7.2. Although interferongamma enhanced IL-6 and IL-8 production, CD40 triggering of IFNgamma-activated hRPE cells did not induce IL-12 secretion. hRPE cells did not stimulate allogeneic resting T cells and downregulated phytohemagglutinin-activated allogeneic T cells via a cell-to-cell contact-dependent mechanism. Some induction of apoptosis was detected. CONCLUSIONS: CD40 is expressed on IFNgamma-activated hRPE cells. Its ligation leads to an increased production of IL-6 and IL-8 but fails to induce B7.1 or B7. 2 expression, or to induce IL-12 secretion. Accordingly, hRPE cells do not activate allogenic T cells but inhibit T-cell proliferation, partly through induction of apoptosis. These results suggest that hRPE cells could be implicated more in a deviant antigen presentation. If the exact molecular mechanisms are unclear, it is likely that CD40-CD40L interaction could play a role in this process.