Potent and Prolonged Innate Immune Activation by Enzyme-Responsive Imidazoquinoline TLR7/8 Agonist Prodrug Vesicles.

Synthetic immune-stimulatory drugs such as agonists of the Toll-like receptors (TLR) 7/8 are potent activators of antigen-presenting cells (APCs), however, they also induce severe side effects due to leakage from the site of injection into systemic circulation. Here, we report on the design and synthesis of an amphiphilic polymer-prodrug conjugate of an imidazoquinoline TLR7/8 agonist that in aqueous medium forms vesicular structures of 200 nm. The conjugate contains an endosomal enzyme-responsive linker enabling degradation of the vesicles and release of the TLR7/8 agonist in native form after endocytosis, which results in high in vitro TLR agonist activity. In a mouse model, locally administered vesicles provoke significantly more potent and long-lasting immune stimulation in terms of interferon expression at the injection site and in draining lymphoid tissue compared to a nonamphiphilic control and the native TLR agonist. Moreover, the vesicles induce robust activation of dendritic cells in the draining lymph node in vivo.

Inhibition of antigen-specific immune responses by co-application of an indoleamine 2,3-dioxygenase (IDO)-encoding vector requires antigen transgene expression focused on dendritic cells.

We have previously shown that particle-mediated epidermal delivery (PMED) of plasmids encoding ß-galactosidase (ßGal) under control of the fascin-1 promoter (pFascin-ßGal) yielded selective production of the protein in skin dendritic cells (DCs), and suppressed Th2 responses in a mouse model of type I allergy by inducing Th1/Tc1 cells. However, intranasal challenge of mice immunized with pFascin-ßGal induced airway hyperreactivity (AHR) and neutrophilic inflammation in the lung. The tryptophan-catabolizing enzyme indoleamine 2,3-dioxygenase (IDO) has been implicated in immune suppression and tolerance induction. Here we investigated the consequences of co-application of an IDO-encoding vector on the modulatory effect of DNA vaccination by PMED using pFascin-ßGal in models of eosinophilic allergic and non-eosinophilic intrinsic airway inflammation. IDO-encoding plasmids and pFascin-ßGal or pCMV-ßGal were co-applied to abdominal skin of BALB/c mice without, before or after sensitization with ßGal protein. Immune responses in the lung were analysed after intranasal provocation and airway reactivity was determined by whole body plethysmography. Co-application of pCMV-IDO with pFascin-ßGal, but not pCMV-ßGal inhibited the Th1/Tc1 immune response after PMED. Moreover, AHR in those mice was attenuated following intranasal challenge. Therapeutic vaccination of ßGal-sensitized mice with pFascin-ßGal plus pCMV-IDO slightly suppressed airway inflammation and AHR after provocation with ßGal protein, while prophylactic vaccination was not effective. Altogether, our data suggest that only the combination of DC-restricted antigen and ubiquitous IDO expression attenuated asthma responses in mice, most probably by forming a tryptophan-depleted and kynurenine-enriched micromilieu known to affect neutrophils and T cells.

Feed pellets containing chitosan nanoparticles as plasmid DNA oral delivery system for fish: In vivo assessment in gilthead sea bream (Sparus aurata) juveniles.

The aim of this study was the assessment of preloaded feed pellets as a delivery system for plasmid DNA (pDNA), with the purpose of evaluating the potential administration of DNA vaccines orally in aquacultured fish. Pellets were made up by usual feed ingredients, which were mixed with chitosan nanoparticles entrapping a model plasmid (pCMVß) expressible in eukaryotic cells before being elaborated. The plasmid is characterized by the insertion of the reporter gene lacZ, encoding for the bacterial enzyme ß-galactosidase (ß-gal). The possible in vivo expression of the exogenous gene was measured in different fish tissues of gilthead sea bream (Sparus aurata) juveniles by two different procedures. On the one hand, the activity of the enzyme ß-gal was detected and quantified in muscle, liver and intestine; on the other, specific IgM against ß-gal antigen was titrated in blood samples. Intramuscular (i.m.) injection of equal amounts of plasmid was also carried out for the purpose of comparison with oral administration. The expression of the reporter gene was detected in fish tissues following both oral and i. m. administration of pDNA up to 60 days. However, organ distribution of the gene expression was more evident after oral (ß-gal activity measured in gut, liver and muscle) than after parenteral administration (restricted to adjacent muscle tissues). In agreement, specific IgM titration indicated that humoral immune response was more intense and sustained throughout the experimental period after oral than after i. m. delivery of equal amounts of pDNA. These results suggest that feed pellets containing chitosan nanoparticles might enable efficient oral delivery of pDNA, a fact that might imply valuable applications in terms of on-farm mass immunization purposes, especially with regard to DNA-based vaccines and small size fish, in which i. m. administration remains unfeasible.

Co-delivery of indoleamine 2,3-dioxygenase prevents loss of expression of an antigenic transgene in dystrophic mouse muscles.

A significant problem affecting gene therapy approaches aiming at achieving long-term transgene expression is the immune response against the protein product of the therapeutic gene, which can reduce or eliminate the therapeutic effect. The problem is further exacerbated when therapy involves targeting an immunogenic tissue and/or one with a pre-existing inflammatory phenotype, such as dystrophic muscles. In this proof-of-principle study, we co-expressed a model antigen, bacterial ß-galactosidase, with an immunosuppressive factor, indoleamine 2,3-dioxygenase 1 (IDO1), in muscles of the mdx mouse model of Duchenne muscular dystrophy. This treatment prevented loss of expression of the transgene concomitant with significantly elevated expression of T-regulatory (Treg) markers in the IDO1-expressing muscles. Moreover, co-expression of IDO1 resulted in reduced serum levels of anti-ß-gal antibodies. These data indicate that co-expression of genes encoding immunomodulatory enzymes controlling kynurenine pathways provide a viable strategy for preventing loss of transgenes targeted into dystrophic muscles with pre-existing inflammation.

T cells targeting a neuronal paraneoplastic antigen mediate tumor rejection and trigger CNS autoimmunity with humoral activation.

Paraneoplastic neurologic diseases (PND) involving immune responses directed toward intracellular antigens are poorly understood. Here, we examine immunity to the PND antigen Nova2, which is expressed exclusively in central nervous system (CNS) neurons. We hypothesized that ectopic expression of neuronal antigen in the periphery could incite PND. In our C57BL/6 mouse model, CNS antigen expression limits antigen-specific CD4+ and CD8+ T-cell expansion. Chimera experiments demonstrate that this tolerance is mediated by antigen expression in nonhematopoietic cells. CNS antigen expression does not limit tumor rejection by adoptively transferred transgenic T cells but does limit the generation of a memory population that can be expanded upon secondary challenge in vivo. Despite mediating cancer rejection, adoptively transferred transgenic T cells do not lead to paraneoplastic neuronal targeting. Preliminary experiments suggest an additional requirement for humoral activation to induce CNS autoimmunity. This work provides evidence that the requirements for cancer immunity and neuronal autoimmunity are uncoupled. Since humoral immunity was not required for tumor rejection, B-cell targeting therapy, such as rituximab, may be a rational treatment option for PND that does not hamper tumor immunity.

Administration of DNA Encoding the Interleukin-27 Gene Augments Antitumour Responses through Non-adaptive Immunity.

DNA-mediated immunization of a tumour antigen is a possible immunotherapy for cancer, and interleukin (IL)-27 has diverse functions in adaptive immunity. In this study, we examined whether IL-27 DNA administration enhanced antitumour effects in mice vaccinated with DNA encoding a putative tumour antigen, ß-galactosidase (ß-gal). An intramuscular injection of cardiotoxin before DNA administration facilitated the exogenous gene expression. In mice received ß-gal and IL-27 DNA, growth of ß-gal-positive P815 tumours was retarded and survival of the mice was prolonged. Development of ß-gal-positive Colon 26 tumours was suppressed by vaccination of ß-gal DNA and further inhibited by additional IL-27 DNA administration or IL-12 family cytokines. Nevertheless, a population of ß-gal-specific CD8(+) T cells did not increase, and production of anti-ß-gal antibody was not enhanced by IL-27 DNA administration. Spleen cells from mice bearing IL-27-expressing Colon 26 tumours showed greater YAC-1-targeted cytotoxicity although CD3(-)/DX5(+) natural killer (NK) cell numbers remained unchanged. Recombinant IL-27 enhanced YAC-1-targeted cytotoxicity of IL-2-primed splenic NK cells and augmented a phosphorylation of signal transducer and activator of transcription 3 and an expression of perforin. These data collectively indicate that IL-27 DNA administration activates NK cells and augments vaccination effects of DNA encoding a tumour antigen through non-adaptive immune responses.

Local "on-demand" generation and function of antigen-specific Foxp3+ regulatory T cells.

Extrathymically derived regulatory T cells (iTregs) protect against autoimmunity to tissue-specific Ags. However, whether Ag-specific iTreg generation and function is limited to secondary lymphoid tissue or whether it can occur within the tissue-specific local environment of the cognate Ag remains unresolved. Mice expressing ß-galactosidase (ßgal) on a retina-specific promoter (ßgal mice) in conjunction with mice expressing GFP and diphtheria toxin (DTx) receptor (DTR) under control of the Foxp3 promoter, and ßgal-specific TCR transgenic (BG2) mice were used to examine this question. Local depletion (ocular DTx), but not systemic depletion (i.p. DTx), of ßgal-specific iTregs enhanced experimental autoimmune uveoretinitis induced by activated ßgal-specific effector T cells. Injections of small amounts of ßgal into the anterior chamber of the eye produced similar numbers of ßgal-specific iTregs in the retina whether the mouse was depleted of pre-existing, circulating Tregs. Taken together, these results suggest that protection from tissue-specific autoimmunity depends on the function of local Ag-specific iTregs and that the retina is capable of local, "on-demand" iTreg generation that is independent of circulating Tregs.

Effects of antigen-expressing immunostimulatory liposomes on chemotaxis and maturation of dendritic cells in vitro and in human skin explants.

PURPOSE: Antigen-Expressing Immunostimulatory Liposomes (AnExILs) represent a novel DNA vaccination platform based on the production of protein antigens from DNA templates inside liposomes mediated by an in vitro transcription and translation (IVTT) mix. The aim of this study was to analyze the effects of AnExILs on different dendritic cells (DCs) models and to better understand the role of the different components of this formulation on its adjuvanticity. METHODS: The effect of ß-galactosidase-expressing AnExILs on maturation and particle uptake by murine DC cell line, fresh human monocyte-derived DCs or human dermal DCs in skin explants was investigated and compared to the effects of either plain liposomes or IVTT mix alone. RESULTS: AnExILs induced efficient DC chemotaxis and promoted up-regulation of maturation markers on murine DCs, due to the presence of IVTT in the formulation. Furthermore, the amount of active ßGal associated with DCs was higher for AnExILs than for free ßGal expressed in IVTT or ßGal encapsulated into non-adjuvanted liposomes. Most interestingly, the same trend was observed with human DCs. CONCLUSIONS: Both IVTT mix and liposomal vehicles were shown to be key components of the AnExIL formulation responsible for its adjuvanticity. AnExILs combine antigen production, adjuvanticity and delivery in one system, and can efficiently activate both murine and human DCs.

In vivo assessment of specific cytotoxic T lymphocyte killing.

The direct killing of target cells by cytotoxic T lymphocytes (CTLs) plays a fundamental role in protective immunity to viral, bacterial, protozoan and fungi infections, as well as to tumor cells. In vivo cytotoxic assays take into account the interaction of target and effector cells in the context of the proper microenvironment making the analysis biologically more relevant than in vitro cytotoxic assays. Thus, the development, improvement and validation of in vivo methods are necessary in view of the importance of the results they may provide. We describe and discuss in this manuscript a method to evaluate in vivo specific cytotoxic T lymphocyte killing. We used as model system mice immunized with human recombinant replication-deficient adenovirus 5 (HAd5) containing different transgenes as the trigger of a CTL-mediated immune response. To these mice, we adoptively transferred syngeneic cells labeled with different vital fluorescent dyes. Donor cells were pulsed (target) or not (control non-target) with distinct CD8 T-cell epitopes, mixed in a 1:1 ratio and injected i.v. into immunized or non-immunized recipient mice. After 18-24h, spleen cells are collected and analysed by flow cytometry. A deviation from the 1:1 ratio of control and target cell populations indicates antigen specific lysis of target cells.

ßIII-Gal is involved in galactan reduction during phloem element differentiation in chickpea stems.

ßIII-Gal, a member of the chickpea ß-galactosidase family, is the enzyme responsible for the cell wall autolytic process. This enzyme, whose activity increases during epicotyl growth, displays significant hydrolytic activity against cell wall pectins, and its natural substrate has been determined as an arabinogalactan from the pectic fraction of the cell wall. In the present work, the localization of ßIII-Gal in different seedling and plant organs was analyzed by using specific anti-ßIII-Gal antibodies. Our results revealed that besides its possible role in cell wall loosening and in early events during primary xylem and phloem fiber differentiation ßIII-Gal acts on the development of sieve elements. Localization of the enzyme in this tissue, both in epicotyls and radicles from seedlings and in the different stem internodes, is consistent with the reduction in galactan during the maturation of phloem elements, as can be observed with LM5 antibodies. Thus, ßIII-Gal could act on its natural substrate, the neutral side chains of rhamnogalacturonan I, contributing to cell wall reinforcement allowing phloem elements to differentiate, and conferring the necessary strengthening of the cell wall to fulfill its function. This work completes the immunolocation studies of all known chickpea ß-galactosidases. Taken together, our results reflect the broad range of developmental processes covered by different members of this protein family, and confirm their crucial role in cell wall remodeling during tissue differentiation.

Comparison of innate immune activation after prolonged feeding of milk fermented with three species of Lactobacilli.

The present investigation aimed at identifying the abilities of three different species of probiotic lactobacilli to modulate cellular immune responses in mouse neutrophils and macrophages in vivo over a study period of 60 days. Neutrophil respiratory burst enzymes (cytochrome c reductase and MPO) showed remarkable increased activity (P ≤ 0.01) after consumption of milks fermented by different species of probiotics over 30 and 60 days of feeding trials. Enzyme activities (ß-galactosidase and ß-glucuronidase) and nitric oxide production also increased considerably (P ≤ 0.01) in macrophages, both in peritoneal fluid and in enriched cell cultures. The effects of enhanced enzyme activities were corroborated by simultaneous increases in the phagocytic activities of neutrophils and macrophages. The increases in cellular functions were invariably maximal during the first 30 days of study and were maintained, but did not increase, over the next 30 days. Further, Lactobacillus helveticus-fed groups were most effective at modulating neutrophil functions whereas Lactobacillus paracasei-fed groups were more potent at enhancing macrophage functions. Together, our results indicate that probiotics have strain specific effects on stimulating cellular functions while not causing excessive stimulation of the immune system over longer feeding periods, thereby resulting in maximum and stable health benefits.

Incomplete tumour control following DNA vaccination against rat gliomas expressing a model antigen.

BACKGROUND: Vaccination against tumour-associated antigens is one approach to elicit anti-tumour responses. We investigated the effect of polynucleotide (DNA) vaccination using a model antigen (E. coli lacZ) in a syngeneic gliosarcoma model (9L). METHODS: Fisher 344 rats were vaccinated thrice by intramuscular injection of a lacZ-encoding or a control plasmid in weekly intervals. One week after the last vaccination, lacZ-expressing 9L cells were implanted into the striatum. RESULTS: After 3 weeks, in lacZ-vaccinated animals the tumours were significantly smaller than in control-vaccinated animals. In cytotoxic T cell assays lysis rates of >50 % could only be observed in a few of the lacZ-vaccinated animals. This response was directed against lacZ-expressing and parental 9L cells but not against syngeneic MADB 106 adenocarcinoma cells. In Elispot assays interferon-γ production was observed upon stimulation with 9LlacZ and 9L wild-type but not MADB 106 cells. This response was higher for lacZ-immunized animals. All animals revealed dense infiltrates with CD8+ lymphocytes and, to a lesser extent, with NK cells. CD25-staining indicated cells possibly associated with the maintenance of peripheral tolerance to self-antigens. All tumours were densely infiltrated by microglia consisting mostly of ramified cells. Only focal accumulation of macrophage-like cells expressing ED1, a marker for phagocytic activity, was observed. CONCLUSION: Prophylactic DNA vaccination resulted in effective but incomplete suppression of brain tumour formation. Mechanisms other than cytotoxic T cell responses as measured in the generally used in vitro assays appear to play a role in tumour suppression.

Electrostatic protein immobilization using charged polyacrylamide gels and cationic detergent microfluidic Western blotting.

We report a novel protein immobilization matrix for fully integrated microfluidic Western blotting (WB). The electrostatic immobilization gel (EIG) enables immobilization of all proteins sized using cetyl trimethylammonium bromide polyacrylamide gel electrophoresis (CTAB-PAGE), for subsequent electrophoretic probing with detection affinity reagents (e.g., labeled antibodies). The "pan-analyte" capture strategy introduced here uses polyacrylamide gel grafted with concentrated point charges (zwitterionic macromolecules), in contrast to existing microfluidic WB strategies that rely on a sandwich immunoassay format for analyte immobilization and detection. Sandwich approaches limit analyte immobilization to capture of only a priori known targets. A charge interaction mechanism study supports the hypothesis that electrostatic interaction plays a major role in analyte immobilization on the EIG. We note that protein capture efficiency depends on both the concentration of copolymerized charges and ionic strength of the gel buffer. We demonstrate pan-analyte immobilization of sized CTAB-laden model proteins (protein G, ovalbumin, bovine serum albumin, ß-galactosidase, lactoferrin) on the EIG with initial capture efficiencies ranging from 21 to 100%. Target proteins fixed on the EIG (protein G, lactoferrin) are detected using antibody probes with signal-to-noise ratios of 34 to 275. The approach advances protein immunoblotting performance through 200× reduction on sample consumption, 12× reduction in assay duration, and automated assay operation, compared to slab-gel WB. Using the microfluidic WB assay, assessment of lactoferrin in human tear fluid is demonstrated with a goal of advancing toward nonbiopsy-based diagnosis of Sjögren's Syndrome, an autoimmune disease.

Continuous CD8⁺ T-cell priming by dendritic cell cross-presentation of persistent antigen following adeno-associated virus-mediated gene delivery.

Recombinant adeno-associated virus (rAAV) vectors establish persistent transgene expression in the skeletal muscle of mice. How dendritic cells acquire encoded antigens for CD8(+) T-cell priming is unknown. Here we document CD8(+) T-cell priming after lethal irradiation and bone marrow reconstitution of mice treated with an AAV vector several weeks earlier. Temporal separation of vector delivery and successful class I antigen presentation indicated that T-cell priming does not necessarily require antigen synthesis in AAV-transduced dendritic cells. An apparent cross-presentation of antigen acquired from muscle suggests that strategies to limit transgene expression in dendritic cells will not prevent unwanted CD8(+) T-cell responses.

Quality of the transgene-specific CD8+ T cell response induced by adenoviral vector immunization is critically influenced by virus dose and route of vaccination.

Adenoviral vectors have been widely used for experimental gene therapy and vaccination, yet there is a surprising lack of knowledge connecting the route and dose of adenovirus administration to the induced transgene-specific immune response. We have recently demonstrated polyfunctional CD8(+) T cells and protective memory responses using adenoviral vectors, which seem to contrast with recent reports suggesting that an exhausted CD8(+) T cell phenotype is induced by inoculation with adenoviral vectors. Accordingly, we investigated the route and dose interrelationship for transgene-specific CD8(+) T cells using adenoviral vectors encoding beta-galactosidase applied either s.c. or i.v. Irrespective of the route of inoculation, most of the adenoviral inoculum was found to disseminate systemically as the dose was raised beyond 10(9) particles. The number of transgene-specific CD8(+) T cells correlated positively with dissemination, whereas the functional capacity of the generated T cells correlated inversely with vector dissemination. A comparison of the immune response to s.c. or i.v. administration at moderate doses revealed that inoculation by both routes induced a transient peak of IFN-gamma-producing CD8(+) T cells 2 to 3 wk postinfection, but following i.v. administration, these cells were only detected in the liver. Two to four months after systemic, but not peripheral, immunization, dysfunctional transgene-specific CD8(+) T cells impaired in both cytokine production and important in vivo effector functions, accumulated in the spleen. These findings indicate that the localization of the adenoviral inoculum and not the total Ag load determines the quality of the CD8(+) T cell response induced with adenoviral vaccines.

Silencing of T lymphocytes by antigen-driven programmed death in recombinant adeno-associated virus vector-mediated gene therapy.

Recombinant adeno-associated virus (rAAV) vectors are considered promising for human gene replacement because they facilitate stable expression of therapeutic proteins in transduced tissues. Whether the success of gene therapy will be influenced by cellular immune responses targeting transgene-encoded proteins that are potentially immunogenic is unknown. Here we characterized CD8(+) T-cell activity against beta-galactosidase and enhanced green fluorescent protein, model antigens containing major histocompatibility complex (MHC) class I epitopes that are constitutively produced in murine skeletal muscle after rAAV vector transduction. Antigen-specific CD8(+) T cells were detected in the spleen and liver of mice within 7 days of muscle transduction. CD8(+) T-cell frequencies in these organs were stable, and effector functions were intact for months despite ongoing antigen production in muscle. CD8(+) T cells also infiltrated transduced muscle, where frequencies were at least 5-fold higher than in untransduced spleen and liver. Significantly, the majority of antigen-specific CD8(+) T cells in vector-transduced muscle were not functional. Loss of function in the muscle was associated with programmed death of the effector cells. Stable gene expression therefore depended on selective death of CD8(+) T cells at the site of antigen production, an effective mechanism for subverting immunity that is also potentially reversible.

Enhancement of DNA tumor vaccine efficacy by gene gun-mediated codelivery of threshold amounts of plasmid-encoded helper antigen.

Nucleic acid-based vaccines are effective in infectious disease models but have yielded disappointing results in tumor models when tumor-associated self-antigens are used. Incorporation of helper epitopes from foreign antigens into tumor vaccines might enhance the immunogenicity of DNA vaccines without increasing toxicity. However, generation of fusion constructs encoding both tumor and helper antigens may be difficult, and resulting proteins have unpredictable physical and immunologic properties. Furthermore, simultaneous production of equal amounts of highly immunogenic helper and weakly immunogenic tumor antigens in situ could favor development of responses against the helper antigen rather than the antigen of interest. We assessed the ability of 2 helper antigens (beta-galactosidase or fragment C of tetanus toxin) encoded by one plasmid to augment responses to a self-antigen (lymphoma-associated T-cell receptor) encoded by a separate plasmid after codelivery into skin by gene gun. This approach allowed adjustment of the relative ratios of helper and tumor antigen plasmids to optimize helper effects. Incorporation of threshold (minimally immunogenic) amounts of helper antigen plasmid into a DNA vaccine regimen dramatically increased T cell-dependent protective immunity initiated by plasmid-encoded tumor-associated T-cell receptor antigen. This simple strategy can easily be incorporated into future vaccine trials in experimental animals and possibly in humans.

A lytic mechanism based on soluble phospholypases A2 (sPLA2) and ß-galactoside specific lectins is exerted by Ciona intestinalis (ascidian) unilocular refractile hemocytes against K562 cell line and mammalian erythrocytes.

Hemocytes from the ascidian Ciona intestinalis exert in vitro Ca²+-dependent cytotoxic activity toward mammalian erythrocytes and K562 cells. To examine the lytic mechanism, hemocyte populations were separated (B1-B6 bands) through a Percoll discontinuous density gradient, the hemocyte cytotoxic activity (HCA) and the lytic activity of the hemocyte lysate supernatant (HLS) were assayed. In addition the separated hemocytes were cultured and the cell-free culture medium (CFM) assayed after 3 h culture. Results support that unilocular refractile hemocytes (URGs), enriched in B5, are cytotoxic. The B5-HLS contains lysins and the activity of B5-CFM shows that lysins can be released into a culture medium. The B5 activity was blocked by D-galactose, α-lactose, lactulose, LacNAc, thiodigalactoside (TDG), L-fucose, D-mannose, D-glucose, sphingomyelin (SM), and soluble phospholipase A2 (sPLA2) inhibitors (dibucain, quinacrine). Accordingly, HLS chemico-physical properties (alkaline medium, high thermostability, Ca²+-dependence, trypsin treatment, protease inhibitors) and SEM observations of the affected targets suggested that sPLA2 could be responsible for changes and large alterations of the target cell membrane. An apoptotic activity, as recorded by a caspase 3, 7 assay, was found by treating K562 cells with very diluted HLS. A lytic mechanism involving sPLA2 and lectins promptly released by URGs and morula cells respectively is suggested, whereas target cell membrane SM could be a modulator of the enzyme activity.

Peripheral induction of tolerance by retinal antigen expression.

The contribution of peripheral expression of tissue-specific CNS Ags to the generation of tolerance is uncertain. To study this question, we examined mice transgenic (Tg) for expression of beta-galactosidase (beta gal) on the retinal photoreceptor cell arrestin promoter, in conjunction with TCR Tg mice producing CD4(+) T cells specific for beta gal (beta galTCR). Several strategies were used to test the hypothesis that betagal expressed in the retina supported thymus-independent tolerance and regulatory T cell development. Retinal expression generated an immunoregulatory response that depressed development of immune responses to beta gal following systemic immunization with beta gal. This regulation was transferable to naive mice by CD3(+)4(+)25(+) T cells from naive retinal beta gal(+) donors. Experiments that removed the beta gal(+) retina by enucleation showed that subsequent development of a regulatory response was lost. Adoptive transfer of CD25(-) beta galTCR T cells into retinal beta gal Tg mice on the Rag(-/-) background led to regulatory activity that limited lymphopenia-induced proliferation of beta galTCR T cells in mice with retinal expression of beta gal and inhibited the ear-swelling assay for delayed type hypersensitivity. These results show that retinal expression of very small amounts of a tissue-specific Ag can generate tolerance that includes regulatory T cells.

Lymphopenia-induced proliferation is a potent activator for CD4+ T cell-mediated autoimmune disease in the retina.

To study retinal immunity in a defined system, a CD4+ TCR transgenic mouse line (betagalTCR) specific for beta-galactosidase (betagal) was created and used with transgenic mice that expressed betagal in retinal photoreceptor cells (arrbetagal mice). Adoptive transfer of resting betagalTCR T cells, whether naive or Ag-experienced, into arrbetagal mice did not induce retinal autoimmune disease (experimental autoimmune uveoretinitis, EAU) and gave no evidence of Ag recognition. Generation of betagalTCR T cells in arrbetagal mice by use of bone marrow grafts, or double-transgenic mice, also gave no retinal disease or signs of Ag recognition. Arrbetagal mice were also resistant to EAU induction by adoptive transfer of in vitro-activated betagalTCR T cells, even though the T cells were pathogenic if the betagal was expressed elsewhere. In vitro manipulations to increase T cell pathogenicity before transfer did not result in EAU. The only strategy that induced a high frequency of severe EAU was transfer of naive, CD25-depleted, betagalTCR T cells into lymphopenic arrbetagal recipients, implicating regulatory T cells in the T cell inoculum, as well as in the recipients, in the resistance to EAU. Surprisingly, activation of the CD25-depleted betagalTCR T cells before transfer into the lymphopenic recipients reduced EAU. Taken together, the results suggest that endogenous regulatory mechanisms, as well as peripheral induction of regulatory T cells, play a role in the protection from EAU.