The effort to explore the specific autoimmune mechanisms of urinary bladder has long been hindered due to a lack of proper animal models. To better elucidate this issue, we developed a novel line of transgenic (Tg) mice, designated as URO-OVA mice, that express the model Ag OVA as a “self”-Ag on the bladder epithelium. URO-OVA mice are naturally tolerant to OVA and show no response to OVA stimulation. Adoptive transfer of naive OVA-specific T cells showed cell proliferation, activation, and infiltration but no bladder histopathology. In contrast, adoptive transfer of activated OVA-specific T cells induced OVA-mediated histological bladder inflammation. Increased mast cells and up-regulated mRNA expressions of TNF-α, nerve growth factor, and substance P precursor were also observed in the inflamed bladder. To further facilitate bladder autoimmunity study, we crossbred URO-OVA mice with OVA-specific CD8+ TCR Tg mice (OT-I mice) to generate a dual Tg line URO-OVA/OT-I mice. The latter mice naturally acquire clonal deletion for autoreactive OT-I CD8+ T cells (partial deletion in the thymus and severe deletion in the periphery). Despite this clonal deletion, URO-OVA/OT-I mice spontaneously develop autoimmune cystitis at 10 wk of age. Further studies demonstrated that the inflamed bladder contained infiltrating OT-I CD8+ T cells that had escaped clonal deletion and gained effector functions before developing histological bladder inflammation. Taken together, we demonstrate for the first time that the bladder epithelium actively presents self-Ag to the immune system and induces CD8+ T cell tolerance, activation, and autoimmune response.
Induction of tolerance is essential for the maintenance of immune homeostasis and the prevention of autoimmune diseases. Multiple mechanisms are involved in the establishment of both central and peripheral tolerance (1, 2). T cell tolerance is first established in the thymus through a tightly controlled process, resulting in the elimination of potentially autoreactive T cell clones and the establishment of a diverse T cell repertoire responsive to foreign pathogens (1). Those deletion-escaped autoreactive T cells are further controlled by additional regulatory mechanisms in the periphery (2). Although these mechanisms have been demonstrated in several transgenic (Tg)3 models that express defined “self” Ags in a tissue-specific manner (3, 4, 5, 6, 7), it is not clear how tolerance is maintained to urinary bladder-associated Ags and how this anatomically unique organ interacts with the immune system.
The bladder epithelium normally resides in a bacteria-free and Ag-free environment and, as such, the bladder mucosa develops acute inflammatory responses whenever the bladder epithelium encounters infectious pathogens (8, 9). Like many other organs, the bladder is also capable of developing autoimmune inflammation under certain pathological conditions. For example, interstitial cystitis is largely considered to be a bladder autoimmune disease that is associated with a localized autoimmune process (10, 11, 12, 13, 14, 15, 16). However, the etiology of interstitial cystitis currently remains unknown. A model with bladder epithelial expression of a defined Ag and its corresponding TCR specificity will facilitate studying bladder autoimmune pathogenesis.
Due to the availability of OVA-specific TCR Tg mice (17, 18), mice that express the model Ag OVA in a tissue-specific manner currently represent one of the most useful tools for study of autoimmune diseases. Mice expressing OVA in the pancreatic islets, intestine, skin, and prostate have shed light on the development of autoimmune disorders in these organs (3, 4, 5, 6, 7). In this study, we also applied this OVA system and developed a novel line of Tg mice (URO-OVA mice) that express OVA as a “self” Ag on the bladder epithelium. We used the uroplakin II (UPII) gene promoter to drive the bladder epithelial expression of OVA. The UPII gene promoter has been used to develop a number of Tg lines that have provided a wealth of information pertaining to the understanding of bladder tumorigenesis (19, 20, 21, 22, 23, 24, 25, 26, 27). In addition, we also generated URO-OVA/OT-I mice that express both OVA and the OVA-specific CD8+ TCR. Both URO-OVA and URO-OVA/OT-I mice develop bladder inflammation upon the introduction of OVA-specific OT-I T cells. Strikingly, URO-OVA/OT-I mice also spontaneously develop autoimmune cystitis when they are 10-wk old. We have investigated OVA-mediated CD8+ T cell tolerance, activation, and autoimmune response in these mice and have demonstrated that they are useful models for the study of bladder-specific autoimmune mechanisms.
Materials and Methods
C57BL/6 (B6) mice were purchased from the National Cancer Institute/Frederick Cancer Research Animal Facility (Frederick, MD). OT-I mice were kindly provided by Dr. W. R. Heath (The Walter and Eliza Hall Institute of Medical Research, Melbourne, Australia). OT-I mice express the transgenic CD8+ TCR (Vα2Vβ5) specific for the OVA257–264 peptide (H2-Kb) (17, 18) and were maintained on a recombinase activating gene 1 knockout (Rag-1−/−) background. B6/OT-I mice (F1 generation) were generated by crossing OT-I mice with B6 mice. These mice are heterozygous for the OVA-specific CD8+ TCR. The urothelial OVA Tg mice, i.e., URO-OVA and URO-OVA/OT-I mice, were generated in our laboratory (see below). All mice were housed in a pathogen-free facility at the University of Iowa (Iowa City, IA). All animal studies were approved by the University of Iowa Animal Care and Use Committee and were in compliance with the Guide for the Care and Use of Laboratory Animals prepared by the National Institutes of Health (Bethesda, MD).
Transgenic OVA construction and generation of urothelial OVA Tg mice
The schematic structure of the transgenic OVA construct is shown in Fig. 1⇓A. A plasmid containing the UPII gene promoter was provided by Dr. T. Sun at the New York University School of Medicine (New York, NY) (19). The 3.6-kb fragment containing the UPII gene promoter was excised and placed upstream to a chimeric transgene (hTrfR-OVA; 1.9 kb) consisting of the human transferrin receptor (hTrfR) transmembrane domain and the OVA coding sequence (28). An intron sequence was inserted between the UPII gene promoter and the hTrfR-OVA transgene to facilitate discriminating OVA mRNA from genomic DNA in RT-PCR amplification (29). The 6.1 Kb DNA fragment containing the above-mentioned sequences and a poly(A) additional site was microinjected into the fertilizing eggs (B6/SJL background). Transgenic founders were backcrossed with B6 mice for six generations to generate B6 congenic URO-OVA mice. PCR was performed for tail genotyping for each generation. The sequences of PCR primer pair used were 5′-GTCGAGACAGAGAAGACT-3′ for the upstream primer and 5′-ACAGCAAGTTTCATCTCCAC-3′ for the downstream primer. To obtain URO-OVA/OT-I mice, URO-OVA mice were crossbred with OT-I mice and selected by PCR tail genotyping and flow cytometry for the TCR (Vα2Vβ5) as described (7). URO-OVA/OT-I mice (F1 generation) were heterozygous for both transgenic OVA and OVA-specific CD8+ TCR.
TaqDNA polymerase. In addition to the OVA primers mentioned above, the following primer pairs were used: 5′-GTTCCAGTATGACTCCACT-3′ and 5′-GTGCAGGATGCATTGCTG-3′ for GAPDH (321 bp); 5′-CGTCAGCCGATTTGCTATCT-3′ and 5′-CGGACTCCGCAAAGTCTAAG-3′ for TNF-α (206 bp); 5′-CACTGAGAACTCCCCCATGT-3′ and 5′-CTGTGGACCCCAGACTGTTT-3′ for nerve growth factor (NGF) (194 bp); and 5′-GCCAATGCAGAACTACGAAA-3′ and 5′-GCTTGGACAGCTCCTTCATC-3′ for tachykinin-1 (substance P precursor; 280 bp). GAPDH was amplified for 25 cycles, whereas the other molecules of interest were amplified for 40 cycles.
The bladders were fixed in 2% buffered formalin for 24 h, embedded in paraffin, and cut into 8-μm sections. After deparaffinization with xylene, slides were rehydrated with graded alcohol washes and subjected to Ag retrieval in a steamer for 30 min in sodium citrate buffer (pH 6.0). The slides were then blocked for endogenous peroxidase activity using 3% hydrogen peroxide and for nonspecific binding using normal goat serum (1/20 dilution in PBS (pH 7.4)). After blocking, the slides were incubated with a rabbit anti-OVA Ab (Cortex) for overnight. After initial rinsing, the slides were treated with HRP-labeled goat-anti-rabbit IgG Ab (Pierce) for 30 min followed by the addition of a substrate solution diaminobenzidine substrate kit; BD Pharmingen). After rinsing again, the slides were counterstained with a H&E solution and viewed with an Olympus microscope.
Flow cytometric analysis
In various experiments, single-cell suspensions of the thymus, spleen, bladder-draining lymph nodes (BLNs), and bladder were prepared by mechanical disruption. For splenocyte preparation, ammonium chloride lysing buffer lysis and Ficoll-Paque gradient centrifugation were performed as described (30, 31). For bladder cell preparation, bladders were gently disrupted in Sigma-Aldrich cell dissociation solution and 0.25% trypsin as described (32). Cells were washed with staining buffer (1% FBS and 0.09% (w/v) NaN3 in Mg2+- and Ca2+ 257–264
In vivo cytotoxicity assay
Splenocytes of B6 mice were prepared as described above and pulsed with either the OVA257–264 peptide (20 μg/ml) or a control peptide of random sequence (20 μg/ml) for 2 h. The OVA257–264 peptide-pulsed cells were further labeled with 5 μM CFSE (CFSEhigh cells; Molecular Probes), whereas the control peptide-pulsed cells were labeled with 0.5 μM CFSE (CFSElow cells). After labeling at 37°C for 10 min, CFSEhigh and CFSElow cells were mixed at a 1:1 ratio and 1 × 107 cells were injected i.v. into each mouse. Splenocytes of recipient mice were prepared 15 h later and analyzed for CFSE-labeled cells.
Adoptive transfer of OT-I T cells
Naive splenocytes of OT-I/Rag-1−/− mice were prepared and labeled with 5 μM CFSE as described above. After washing, 1 × 107 cells per mouse were injected i.v. into URO-OVA mice. At days 1, 3, and 7 after injection, the spleen and BLNs were collected for flow cytometric analysis of CFSE-labeled dividing cells. The i.v. injection of naive splenocytes was also used to investigate T cell activation in the BLNs and T cell infiltration and cytokine production in the bladder at certain time points.
In some experiments, OT-I CD8+ T cells were prepared using the MACS magnetic isolation system (Miltenyi Biotec) and anti-CD8α (Ly-2) microbeads according to the manufacturer’s instruction. The purity of isolated CD8+ T cells was 96%. The OVA257–264 peptide (10 μg/ml) was used to activate CD8+ T cells in the presence of syngeneic D1 dendritic cells for 2 days (33). After Ficoll-Paque gradient centrifugation, 5 × 106 OT-I CD8+ T cells per mouse were injected i.v. into URO-OVA mice. At days 1, 4, 7, and 14 postinjection, mice were sacrificed and the bladders were collected for inflammation analysis.
Bladder tissue cytokine microarray
The bladder was homogenized in 300 μl of preservative buffer containing 0.2 M Tris-HCl (pH 7.6), 0.5% BSA, 0.01% sodium azide, and Complete protease inhibitor tablet (1/10 volume; Roche). After centrifugation at 12,000 rpm, the supernatant was collected for a cytokine microarray assay provided by Allied Biotech. Briefly, the service used an Ab-based protein multiplex apparatus that allowed the simultaneously measurement of a panel of 13 cytokines including IL-1β, IL-2, IL-4, IL-5, IL-6, IL-10, IL-12 p40, IL-12 p70, IL-13, IFN-γ, TNF-α, vascular endothelial growth factor (VEGF), GM-CSF, and MCP-1. A streptavidin-Cy5 conjugate was used for fluorescence detection in the system (34).
Bladder histological analysis
The standard histological H&E staining was performed. Briefly, paraffin- embedded bladder sections were prepared. After deparaffinization, slides were stained with hematoxylin for 5 min followed by eosin staining for 1 min. The bladder inflammation was scored based on cellular infiltration in the lamina propria and interstitial edema as follows: 1+ (mild infiltration with no or mild edema); 2+ (moderate infiltration with moderate edema); and 3+ (moderate to severe infiltration with severe edema) (35). Giemsa staining was also performed to detect mast cells using a mast cell-staining kit according to the manufacturer’s instruction (Electron Microscopy Sciences). Mast cells were counted in 10 consecutive sections with 80 μm apart for each bladder.
Constitutive expression of OVA on the bladder epithelium induces T cell tolerance
To generate a mouse model expressing a defined Ag on the bladder epithelium, we made a DNA construct consisting of the UPII gene promoter and the OVA-containing transgene (hTrfR-OVA) (Fig. 1⇑A). The UPII gene promoter is urothelium-specific and facilitates the expression of transgenic OVA on the bladder epithelium (19). The hTrfR was fused to the OVA coding sequence for anchoring the OVA protein on the epithelial cell membrane (7). This transgenic construct led to bladder epithelial expression of OVA in mice (Fig. 1⇑, B and C). The Tg mice (URO-OVA mice) appear healthy and do not show histological bladder inflammation without inflammatory induction (Fig. 5⇓, left panel). No OVA-tetramer+CD8+ T cells were detected in the thymus and the periphery (data not shown), indicating that those clones with sufficient affinity to bind the OVA-tetramer had been deleted in these mice.
To determine whether URO-OVA mice were responsive to OVA stimulation, mice were infected with an OVA-delivering adenoviral recombinant (Ad-OVA) as described previously (36). B6 mice were used as a control, because these mice readily developed OVA-specific CD8+ T cells after Ad-OVA infection (36). After 14 days, splenocytes were prepared for flow cytometric analysis of OVA-tetramer+CD8+ T cells, and no double-positive cells were detected in URO-OVA mice (Fig. 2⇓A). Splenocytes were further stimulated with the OVA257–264 peptide and evaluated for IFN-γ-producing CD8+ T cells (Fig. 2⇓B). Again, no induction of IFN-γ+CD8+ T cells was observed for URO-OVA mice infected with Ad-OVA. Furthermore, the in vivo cytotoxicity assay showed no killing of OVA257–264 peptide-pulsed targets by URO-OVA mice infected with Ad-OVA (Fig. 2⇓C). Thus, it is clear that URO-OVA mice are naturally tolerant to OVA and do not respond to OVA stimulation.
Adoptively transferred naive OT-I T cells proliferate and activate in URO-OVA mice
Although the expression of bladder epithelial OVA renders URO-OVA mice nonresponsive, the transgenic OVA appears to be antigenic. A clear T cell division was observed by flow cytometric analysis of BLN cells at days 3 (31.03%) and 7 (97.59%) after adoptive transfer of CFSE-labeled naive OT-I/Rag-1−/− splenocytes (Fig. 3⇓A, lower panels). The spleens of the same mice also showed the dividing cells at days 3 (14.3%) and 7 (83.2%) (Fig. 3⇓A, upper panels). These spleen CFSE+ cells likely migrated from the BLNs after T cell activation, because CFSE+ cells were also observed in the bladder by fluorescence microscopy (data not shown). No T cell division was observed for the bladder nondraining inguinal LNs (NLNs) of URO-OVA mice or the BLNs or NLNs of control B6 mice (data not shown). In addition, BLN CD8+ T cells of URO-OVA mice also showed up-regulated expression of the CD69 marker but no changes for CD44, CD45RB, or CD62L at day 4 after the T cell transfer (Fig. 3⇓B). This phenotype indicated that these CD8+ T cells were partially activated T cells similar to the CD8+ T cells identified in the murine intestinal epithelium (37). No T cell activation was observed for the NLNs of URO-OVA mice or the BLNs or NLNs of control B6 mice (data not shown). Our results suggest that the bladder epithelial OVA is antigenic and could gain access to the BLNs for Ag presentation.
Adoptively transferred naive OT-I T cells home to the bladder but fail to induce bladder histopathology in URO-OVA mice
In addition to their proliferation and activation in the BLNs, adoptively transferred naive OT-I/Rag-1−/− T cells also home to the bladder of URO-OVA mice. The infiltrating Vα2+CD8+ T cells were detectable at day 4 (Fig. 4⇓A, right panel), increased to a plateau at days 7–10, and declined to a basal level thereafter (data not shown). The bladder cytokine expression was analyzed by a commercial cytokine microarray assay. Among 13 cytokines tested, elevated IL-5, IL-6, IL-10, IFN-γ, and MCP-1 were observed in the bladders collected at day 4 (Fig. 4⇓B). Elevated MCP-1 was also observed in the bladders collected at day 1 before the detection of other cytokines. The bladder VEGF appeared to have no clear correlation with T cell transfer, because the control bladders (day 0) also showed a comparable level of VEGF. Despite these T cell infiltration and cytokine production, no histopathological changes were found in the day 4 bladders and the bladders collected at the later time points (data not shown). These observations suggest that the bladder epithelial OVA serves as a target for OVA-specific CD8+ T cells but is insufficient to mediate bladder histopathology by the adoptive transfer of naive T cells.
Cystitis induction by the adoptive transfer of activated OT-I T cells in URO-OVA mice
Because naive OT-I T cells showed their responses in the BLNs and subsequent homing to the bladder but were unable to induce bladder histopathology, we sought to induce bladder inflammation by the adoptive transfer of activated OT-I T cells in URO-OVA mice. OT-I CD8+ T cells were isolated and activated in vitro with the OVA257–264 peptide in the presence of syngeneic D1 dendritic cells. The bladders were collected at 1, 4, 7, and 14 days after i.v. injection of activated OT-I CD8+ T cells and processed for paraffin sections and H&E staining. Clear histological bladder inflammation was observed at days 7 (score, 3+) and 14 (score, 2+) (Fig. 5⇓, middle and right panels). Interstitial edema, cellular infiltration, mucosal hyperemia, and epithelial hyperplasia were apparent for the inflamed bladders. Increased mast cells (2-fold) were also observed in the lamina propria and the detrusor of the inflamed bladder at day 14 (Fig. 6⇓). In correlation with the histological bladder inflammation, increased mRNA expression of the mast cell- and sensory nerve-associated inflammatory mediators TNF-α, NGF, and substance P precursor was also observed at days 7 and 14 (Fig. 6⇓C). Among them, elevated NGF was detectable as early as day 4 before the development of bladder histological lesions. Similar bladder histopathology was also observed in URO-OVA mice after the adoptive transfer of activated OT-I splenocytes (data not shown). These observations indicate that URO-OVA mice are capable of developing OVA-mediated bladder inflammation upon the introduction of activated OVA-specific OT-I T cells.
Clonal deletion and spontaneous autoimmune cystitis in URO-OVA/OT-I mice
To facilitate bladder autoimmunity study, we cross-bred URO-OVA mice with OT-I mice to generate URO-OVA/OT-I mice (F1 generation). The latter mouse line was dual transgenic for both OVA and the OVA-specific CD8+ TCR (Vα2Vβ5). Due to the OVA expression, URO-OVA/OT-I mice naturally acquire clonal deletion for potentially autoreactive OT-I CD8+ T cells. Compared with control B6/OT-I mice (F1 generation) that expressed the same OVA-specific CD8+ TCR but no OVA, URO-OVA/OT-I mice showed partial deletion of OVA-tetramer+CD8+ T cells in the thymus (39.2% vs 61.34%) and severe deletion in both spleen (1.33% vs 15.58%) and BLNs (2.12% vs 16.65%) (Fig. 7⇓A).
Like their parental URO-OVA mice, URO-OVA/OT-I mice readily develop histological bladder inflammation upon the introduction of OVA-specific OT-I T cells (data not shown). In addition, because they retain deletion-escaped OVA-specific CD8+ T cells and the expression of transgenic OVA is sustained, URO-OVA/OT-I mice spontaneously develop autoimmune cystitis when they are 10 wk old. As an example, the bladder of a mouse at 12 wk showed clear epithelial hyperplasia and cellular infiltration, the characteristics of chronic bladder inflammation (Fig. 7⇑B, right panel; score, 1–2+). Further analysis indicated that the bladder-infiltrating T cells consisted of a large number of OVA-tetramer+Vα2+CD8+ T cells (Fig. 7⇑C). A similar population of up-regulated CD69 and Vβ5+ T cells was also detected in the inflamed bladder. Accordingly, increased mRNA expression of the inflammatory mediators IFN-γ, TNF-α, and NGF was also observed in the inflamed bladder (data not shown). Despite this bladder autoimmune inflammation, URO-OVA/OT-I mice appear to live a normal life span like their parental URO-OVA mice.
Autoreactive OT-I CD8+ T cells gain effector functions before the development of autoimmune cystitis in URO-OVA/OT-I mice
Because URO-OVA/OT-I mice retained deletion-escaped OVA-tetramer+CD8+ T cells (Fig. 7⇑A) and their bladder infiltrating cells were the same phenotypic cells (Fig. 7⇑C), we questioned whether OT-I CD8+ T cells could gain functional activities before the development of autoimmune cystitis. Compared with B6 and URO-OVA mice, URO-OVA/OT-1 mice (8 wk) contained a much smaller population of CD8+ T cells in the BLNs (Fig. 8⇓A, bottom panel). However, among the CD8+ T cells 8.65% were functionally IFN-γ-producing cells, which was much higher than the percentages for B6 (0.94%) and URO-OVA (0.86%) mice. These IFN-γ-producing CD8+ T cells were increased to 12.95% after OVA257–264 peptide stimulation. There was virtually no change for the BLN cells of B6 and URO-OVA mice after stimulation. The in vivo cytotoxicity assay also demonstrated the presence of OVA-specific CTL activity in URO-OVA/OT-I mice at 8 wk of age (Fig. 8⇓B). The killing of OVA257–264 peptide-pulsed, CFSEhigh-labeled target cells was only observed in URO-OVA/OT-I mice (33.33%) and the positive control B6/OT-I mice (57.5%) but not in the negative control B6 and URO-OVA mice. These results indicate that the deletion-escaped OT-I CD8+ T cells gain effector functions before the spontaneous development of autoimmune cystitis in URO-OVA/OT-I mice.
Progress in transgenic technology has provided crucial information for the better understanding of many autoimmune diseases (3, 4, 5, 6, 7). However, little is currently known about bladder autoimmunity due to a lack of proper animal models. In this study, we developed a novel line of urothelial OVA Tg mice and investigated the functional outcomes of Ag-specific CD8+ T cells in these mice. We chose CD8+ T cells, because these cells are effector cells responsible for the pathogenesis of numerous autoimmune diseases (38, 39, 40, 41, 42).
URO-OVA mice constitutively express OVA as a self-Ag on the bladder epithelium (Fig. 1⇑C). The persistence of OVA expression renders mice systemic tolerance to OVA, as we have been unable to detect OVA-tetramer+CD8+ T cells in the thymus and spleen of URO-OVA mice (data not shown). We have also failed to induce OVA-specific CD8+ T cells by Ad-OVA infection in these mice (Fig. 2⇑). These observations are similar to those observed in other OVA Tg models. Vezys et al. (4) reported that IFABP-tOVA (intestinal fatty acid-binding promoter-truncated OVA; a small intestinal enterocyte OVA Tg line) mice failed to develop an OVA-specific CD8+ T cell response after infection with vesicular stomatitis virus-OVA. Liu and Lefrancois (5) reported that oral infection with Listeria monocytogenes-OVA did not elicit an OVA-specific CD8+ T cell response in Fabpl4× at −132-OVA mice (an intestinal epithelial OVA Tg line). Shibaki et al. (6) reported that K14-mOVA mice (an epidermal OVA Tg line) developed only a marginal OVA-specific T cell response after immunization with OVA emulsified with CFA. In contrast, Kurts et al. (43, 44, 45) reported the presence of clonal deletion for OVA-specific CD8+ T cells in rat insulin promoter (RIP)-mOVA Tg mice, whereas Lees et al. (7) reported the similar OVA-specific CD8+ T cell deletion in POET mice (a prostate epithelial OVA Tg line).
Whether naive OT-I T cells can cause a tissue pathological change is controversial in OVA Tg models. IFABP- tOVA, Fabpl4× at −132-OVA, and POET mice were unable to develop histopathology by passive transfer of naive OT-I CD8+ T cells unless they were concomitantly infected with vesicular stomatitis virus-OVA, L. monocytogenes-OVA, or Ad-OVA, respectively (4, 5, 7). In contrast, passively transferred naive OT-I CD8+ T cells induced diabetes and the graft-vs-host disease-like skin lesion in RIP-mOVA and K14-mOVA mice, respectively (6, 44). In URO-OVA mice, we found that exogenous naive OT-I T cells proliferated and were activated in the BLNs of these mice (Fig. 3⇑). We also observed that the naive OT-I T cells homed to the bladder and induced several cytokines such as IL-5, IL-6, IL-10, IFN-γ, and MCP-1 (Fig. 4⇑). However, this phenomenon was transient, as the bladder-infiltrating T cells diminished after reaching to a plateau at days 7–10. No bladder histopathological lesions were observed in these mice. The reasons for naive OT-I T cells to exhibit the different effects may come from the presence of different tissue APCs in these OVA Tg models (46, 47). It has been reported that tissue-resident dendritic cells expressing low levels of costimulatory molecules are quiescent and responsible for the induction of T cell tolerance (2). The different OVA expression levels in these OVA Tg models may also lead T cells to the different fates such as deletion, anergy, or suppression (48, 49, 50). Indeed, we have observed clonal deletion in both BLNs and the spleens of URO-OVA mice 10 days after the adoptive transfer of naive OT-I T cells (data not shown). Thus, clonal deletion likely contributes at least in part to the failure of the induction of cystitis by naive OT-I T cells in these mice. Nevertheless, our results have demonstrated that the bladder epithelial OVA serves as a target for OVA-specific CD8+ T cells and gains access to the BLNs for Ag presentation. How professional APCs acquire the bladder epithelial OVA and present it to CD8+ T cells remains to be determined.
In contrast to naive OT-I T cells, URO-OVA mice develop OVA-mediated bladder inflammation upon the adoptive transfer of activated OT-I CD8+ T cells. The inflamed bladder showed phenotypic characteristics seen in human bladder inflammation diseases such as interstitial cystitis, including epithelial hyperplasia, interstitial edema, cellular infiltration, and mucosal hyperemia (Fig. 5⇑). This observation suggests that the bladder epithelium could present a self-Ag in the context of MHC class I molecules to OT-I CD8+ T cells and be attacked by these T cells upon recognition. Mast cell counts and the mast cell- and sensory nerve-associated inflammatory mediators TNF-α, NGF, and substance P precursor were also increased in the inflamed bladder (Fig. 6⇑). Although increased mast cell counts were only observed in the inflamed bladder at day 14 (the late time point), their inflammatory mediators were readily detectable on days 4 and 7. This result suggests that mast cells actively participate in the process of OVA-mediated autoimmunity even at early stage. Recently, it was reported that mast cells play an important role in amplifying an Ag-specific T cell response at sites via their surface costimulatory molecules and secreted inflammatory mediators (51, 52).
To further facilitate bladder autoimmunity study, we crossbred URO-OVA mice with OT-I mice to generate URO-OVA/OT-I mice. The latter mouse line was dual transgenic for both bladder epithelial OVA and the OVA-specific CD8+ TCR. Similarly as POET/OT-I and RIP-mOVA/OT-I mice (7, 53), URO-OVA/OT-I mice naturally acquire clonal deletion for autoreactive OT-I CD8+ T cells. Partial deletion was observed in the thymus, whereas severe deletion was observed in the periphery (Fig. 7⇑A). However, despite this clonal deletion, URO-OVA/OT-I mice spontaneously develop OVA-mediated autoimmune cystitis at 10 wk of age. The inflamed bladders showed phenotypic characteristics of chronic bladder inflammation, including epithelial hyperplasia and cellular infiltration (Fig. 7⇑B). Like other OVA Tg models (7, 54), autoreactive OT-I CD8+ T cells that escape clonal deletion are likely the causal factor for autoimmune cystitis in URO-OVA/OT-I mice. There are three lines of evidence supporting this assumption: 1) the bladder-infiltrating T cells consisted of a large number of OVA-specific OT-I CD8+ T cells (Fig. 7⇑C); 2) the BLNs contained functionally active OVA-specific CD8+ T cells (Fig. 8⇑A); and 3) in vivo OVA-specific CTL activity was demonstrated (Fig. 8⇑B). Because no cystitis has been found in URO-OVA/OT-I mice before 10 wk, the development of spontaneous autoimmune cystitis appears to be an aging-related process. Probably, the naturally acquired OVA tolerance diminishes over time during this developmental stage, which leads to the activation of deletion-escaped OT-I CD8+ T cells in the BLNs where APCs present OVA Ag to them. Once activated, autoreactive CD8+ T cells home to the bladder and attack the bladder epithelium that also displays the same OVA Ag in the context of MHC class I molecules. However, the actual mechanisms controlling this autoimmune process and the causes for loss of the controlling need to be further investigated.
In summary, we show here for the first time that the bladder epithelium actively presents self-Ag to the immune system and causes CD8+ T cell tolerance, activation, and autoimmune response. These urothelial OVA Tg mice may provide an ideal model for understanding the bladder-originated autoimmune mechanisms and for developing effective interventions for bladder autoimmune diseases.
We thank Dr. Tung-Tien Sun at the New York University School of Medicine for providing the UPII gene promoter-containing plasmid. We also thank Dr. Curt D. Sigmunt for generating URO-OVA Tg founders and Dr. Barry R. DeYoung for assisting with bladder histological analysis. Thanks also to Drs. Timothy L. Ratliff and Michael A. O’Donnell for constructive discussions and to Mitchell L. Rotman for helping edit the manuscript.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵1 This work was supported by National Institutes of Health Grant RO1DK066079 and Department of Defense Peer Reviewed Medical Research Program Award W81XWH-04-1-0070 (to Y.L.).
↵2 Address correspondence and reprint requests to Dr. Yi Luo, Department of Urology, University of Iowa, 3202 Medical Education and Research Facility, 375 Newton Road, Iowa City, IA 52242-1087. E-mail address:
↵3 Abbreviations used in this paper: Tg, transgenic; Ad-OVA, OVA-delivering recombinant adenovirus; B6, C57BL/6 mice; BLN, bladder-draining lymph node; hTrfR, human transferrin receptor; NGF, nerve growth factor; NLN, nondraining inguinal lymph node; RIP, rat insulin promoter; UPII, uroplakin II; VEGF, vascular endothelia growth factor.
- Received August 28, 2006.
- Accepted October 18, 2006.
- Copyright © 2007 by The American Association of Immunologists