Tissue targeting of anti-RNP autoimmunityEffects of T cells and myeloid dendritic cells in a murine model.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 60, No. 2, February 2009, pp 534–542 DOI 10.1002/art.24256 © 2009, American College of Rheumatology Tissue Targeting of Anti-RNP Autoimmunity Effects of T Cells and Myeloid Dendritic Cells in a Murine Model Eric L. Greidinger,1 YunJuan Zang,2 Irina Fernandez,3 Mariana Berho,4 Mehdi Nassiri,2 Laisel Martinez,2 and Robert W. Hoffman1 Objective. To explore the role of immune cells in anti-RNP autoimmunity in a murine model of pneumonitis or glomerulonephritis, using adoptive transfer techniques. Methods. Donor mice were immunized with 50 g of U1–70-kd small nuclear RNP fusion protein and 50 g of U1 RNA adjuvant. Whole splenocytes as well as CD4ⴙ cell and dendritic cell (DC) subsets from the immunized mice were infused into naive syngeneic recipients. Anti-RNP and T cell responses were assessed by immunoblotting, enzyme-linked immunosorbent assay, and flow cytometry. Development of renal or lung disease was assessed by histology and urinalysis. Results. Unfractionated splenocytes from donor mice without proteinuria induced predominantly lung disease in recipients (8 [57%] of 14 versus 2 [14%] of 14 developing renal disease; P ⴝ 0.046). However, infusion of CD4ⴙ cells from donors without proteinuria induced renal disease more frequently than lung disease (7 [70%] of 10 versus 2 [20%] of 10; P ⴝ 0.01); adoptive transfer of RNPⴙCD4ⴙ T cells from short-term culture yielded similar results (renal disease in 8 [73%] of 11 recipients versus lung disease in 3 [27%] of 11). Cotransfer of splenic myeloid DCs and CD4ⴙ T cells from immunized donors prevented induction of renal disease in all 5 recipients (P ⴝ 0.026 versus recipients of fresh CD4ⴙ cells alone), although lung disease was still observed in 1 of 5 mice. Transfer of myeloid DCs alone from immunized donors induced lung disease in 3 (60%) of 5 recipients, without evidence of nephritis. Cotransfer of splenocytes from mice with and those without nephritis led to renal disease in 4 of 5 recipients, without evidence of lung disease. Conclusion. These findings indicate that RNPⴙ CD4ⴙ T cells are sufficient to induce anti-RNP autoimmunity, tissue targeting in anti-RNP autoimmunity can be deviated to either a renal or pulmonary phenotype depending on the presence of accessory cells such as myeloid DCs, and DC subsets can play a role in both propagation of autoimmunity and end-organ targeting. Autoimmunity to RNP autoantigens is frequently seen in systemic autoimmune diseases, including lupus and mixed connective tissue disease (MCTD) (1). The induction of anti-RNP autoantibodies is closely linked, in terms of time to onset, with the initial presentation of clinical manifestations of autoimmune disease in patients, suggesting that anti-RNP responses may have direct pathogenic roles in autoimmune diseases (2,3). We recently demonstrated that in mice that were not otherwise disease-prone, a single direct immunization with an RNP peptide plus adjuvant can induce a persistent autoimmune response, including development of autoantibodies and end-organ injury, both of which are consistent with the characteristics of human anti-RNP syndromes (4). The findings from this model also suggest that the pattern of tissue injury that emerges could Dr. Greidinger’s work was supported by the US Department of Veterans Affairs (Merit Review grant), the NIH (grant AI-1842), and the Lupus Research Institute. Dr. Hoffman’s work was supported by the US Department of Veterans Affairs and the NIH (grants AR-43308 and AR-48805). 1 Eric L. Greidinger, MD, Robert W. Hoffman, DO: Miami VA Medical Center, and University of Miami Miller School of Medicine, Miami, Florida; 2YunJuan Zang, MD, Mehdi Nassiri, MD, Laisel Martinez, MS: University of Miami Miller School of Medicine, Miami, Florida; 3Irina Fernandez, MS: Miami VA Foundation for Medical Research, Miami, Florida; 4Mariana Berho, MD: Cleveland Clinic Foundation, Weston, Florida. Dr. Nassiri holds a patent for a molecular fixative, for which he receives royalties through Sakura FineTek USA. Address correspondence and reprint requests to Eric L. Greidinger, MD, Division of Rheumatology and Immunology, University of Miami, 1400 NW 10th Avenue, Suite 602, Miami, FL 33136. E-mail: firstname.lastname@example.org. Submitted for publication July 1, 2008; accepted in revised form October 17, 2008. 534 TISSUE TARGETING OF ANTI-RNP AUTOIMMUNITY develop into either MCTD-like lung disease or lupuslike nephritis, depending on the immune context present. We have since found that serum transfer from immunized mice could exacerbate lung injury in recipients in the setting of acute inflammation but causes minimal tissue injury in normal physiologic conditions (5). Prior studies have indicated that the functions of T cells may be sufficient to convey anti-RNP autoimmunity (6). This led us to investigate the role of cellular immunity in mediating anti-RNP autoimmunity. Thus, in the present study, we examined the ability of immune cells from anti-RNP–immunized mice to induce disease after adoptive transfer into naive syngeneic animals. Our results show that transfer of whole splenocytes conveys anti-RNP autoimmunity and lung disease from donors to recipients, whereas transfer of only CD4⫹ splenocytes conveys anti-RNP autoimmunity but changes the clinical phenotype from lung disease to nephritis in the recipients. We were able to identify a population of nonplasmacytoid splenic dendritic cells (DCs), referred to as myeloid DCs, that can prevent the induction of nephritis, when cotransferred with CD4⫹ cells, or can induce anti-RNP immunity with lung disease, when transferred alone. However, splenocytes from mice without nephritis, when transferred along with splenocytes from mice with established nephritis, failed to prevent nephritis, and a small number of plasmacytoid DCs from the immunized mice with renal disease was sufficient to convey nephritis to naive recipients. These results suggest that adaptive and innate immune cells collaboratively shape the tissue expression of systemic autoimmune disease. Moreover, we are able to identify distinct steps in the development of the autoimmune responses and the differentiation of autoimmune tissue targeting. MATERIALS AND METHODS Mice. Experiments were performed using C57BL/ 6NTac H2-Abl–knockout DR4-transgenic mice (Taconic, Germantown, NY). These C57BL/6 mice (herein referred to as DR4 mice) are transgenic for the expression of a chimeric human/mouse class II major histocompatibility complex in which the extracellular antigen-presentation domains of HLA– DR4 have replaced the native murine class II regions, with the remainder of the native murine molecule kept intact, as we have previously reported (4). Earlier studies used the same mouse strain successfully in adoptive transfer experiments involving immunization with type II collagen (CII), in which transfer of anti-CII immunity and induction of arthritis were successfully achieved in transfer recipients, without indications of lupus or MCTD-like immunity (7). All experiments were conducted in accordance with Institutional Animal Care and 535 Use Committee–approved protocols and were carried out in facilities certified by the Association for Assessment and Accreditation of Laboratory Animal Care. DR4 mice, at ages 8–10 weeks, were subcutaneously immunized once in the flank with 50 g of U1–70-kd small nuclear RNP fusion protein (70K) and 50 g of U1 RNA adjuvant. As previously reported, a high proportion of the directly immunized mice developed anti-RNP autoantibodies (4), RNP-specific T cells (8), and interstitial pneumonitis that persisted for months after the single immunization, and this was associated with anti-70K T cell epitope spreading (4). In order for a mouse to be qualified as immunized for use in the current study, anti-RNP antibodies were required to be demonstrated by enzyme-linked immunosorbent assay (ELISA) and/or immunoblotting. Adoptive transfer. To assess the contribution of immune cells to this process, we harvested splenocytes from anti-RNP–immunized mice 2 months after immunization; splenocytes from naive syngeneic mice were used as controls. Whole spleens were mechanically disrupted and passed through a fine wire mesh in cold phosphate buffered saline (PBS) to create a single-cell suspension. The cells were treated with red blood cell (RBC) lysis buffer, washed, and resuspended in sterile PBS for infusion or further cell separation. We infused cells (10 ⫻ 106 whole splenocytes, 2 ⫻ 106 CD4⫹ cells, and 0.1 ⫻ 106 myeloid DCs or 0.1 ⫻ 106 plasmacytoid DCs, in a final volume of 50 l of sterile PBS) into 8–10-weekold naive syngeneic (nonirradiated) recipients via the tail vein; thereafter, the recipient mice were followed up for 2 months. As we have previously reported, recipient mice in the whole splenocyte transfer experiments developed anti-RNP antibodies that were typically of lower titer than that found in donor mice (4,8). Moreover, the anti-RNP T cells from recipients had T cell fine specificity and also had T cell receptor (TCR) V-gene usage similar to that of the anti-RNP–specific T cells derived from donor mice (8). Cell separation. CD4⫹ cells were obtained from RBCdepleted splenocytes by positive selection with anti-CD4 microbeads, using AutoMACS (Miltenyi, Auburn, CA). Myeloid DCs were obtained from RBC-depleted splenocytes, using OptiPrep density centrifugation in accordance with the manufacturer’s instructions (Sigma-Aldrich, St. Louis, MO) to isolate the DC layer, followed by negative selection with anti-murine plasmacytoid dendritic cell antigen 1 (PDCA-1) microbeads to deplete plasmacytoid DCs, and positive selection with anti-CD11c microbeads to select myeloid DCs (AutoMACS; Miltenyi). Plasmacytoid DCs were obtained using OptiPrep density centrifugation as described above, followed by use of the Miltenyi murine Plasmacytoid Dendritic Cell Negative Selection Kit, then positive selection with antimurine PDCA-1 microbeads (Miltenyi). Cell culture. Murine T cell lines were generated and characterized using an approach similar to one that has been used extensively by our group for generating human autoantigen-specific T cell lines and one used more recently to generate murine autoantigen-specific T cell lines (8–12). Briefly, spleen and lymph node cells were obtained at necropsy, in sterile conditions, and then mechanically disrupted, filtered through a sterile 100-m nylon mesh filter, and subjected to density-gradient centrifugation using Histopaque (Sigma-Aldrich). Cells obtained from immunized mice were 536 used immediately for the generation of T cell lines. Cells obtained from naive mice were irradiated with 30 Gy for use as antigen-presenting cells (APCs) to stimulate T cells and for use in proliferation assays. Approximately 5 ⫻ 106 cells were cultured in Dulbecco’s modified Eagle’s medium with 2 mM L-glutamine (complete medium) supplemented with 20 g/ml gentamicin/15% fetal calf serum and containing fusion protein at a final concentration of 50 g/ml. The 70-kd fusion protein was used as antigen, as described previously (8–12). Cells in a final volume of 5 ml were placed in a 25-cm2 flask and incubated in 5% carbon dioxide at 37°C. Cells were restimulated with 5 ⫻ 106 murine APCs, which had been irradiated with 30 Gy, together with antigen in fresh medium on days 7–10. Following 1 or 2 cycles of stimulation, cells were harvested for adoptive transfer, in vitro proliferation assays, and/or TCR analyses. For TCR analyses, cells were stimulated with plate-bound antiCD3 in the absence of APCs, and at 48 hours, RNA was extracted from the cells and analyzed using RNeasy (Invitrogen, Carlsbad, CA). Disease assessment. The presence of active urinary sediment was determined by urinalysis on urine specimens obtained at the time the mice were killed. The presence of histologic abnormalities was determined in a blinded manner by assessing UM-Fix–prepared, hematoxylin and eosin–stained tissue, as previously reported (4,13). Additional sections of the kidney, obtained after the mice were killed, were embedded in OCT compound and snap-frozen for immunofluorescence assays with fluorescein isothiocyanate–conjugated anti-mouse C3 and anti-mouse IgG antibodies. Serum C3 levels were measured in mouse serum samples, using a commercial ELISA (Immunology Consultants, Newberg, OR). Flow cytometry. Fluorochrome-conjugated antibodies to CD3, CD4, CD8, CD11c, and HLA–DR4 were purchased from Becton Dickinson (Franklin Lakes, NJ), 5,6carboxyfluorescein succinimidyl ester (CFSE) was obtained from Molecular Probes (Eugene, OR), and fluorochromeconjugated anti-murine PDCA-1 was from Miltenyi. For flow cytometric analyses, purified myeloid DCs were labeled with CFSE in a 10-minute incubation, in which the cells were incubated in complete medium containing a final concentration of 2.5 M CFSE. Cells were washed and resuspended in cold sterile PBS prior to infusion into the tail vein. Cells were quantitated on a BD LSRII instrument, and data were analyzed using BD FACSDiva software (Becton Dickinson). Anti-RNP assays. Anti-RNP autoantibodies were detected by immunoblotting against intact and apoptotic Jurkat cell lysates and by an ELISA against purified 70K fusion protein as antigen, both of which have been previously described in detail (14). Anti-RNP T cells were detected using a previously described method (8,12), in which we noted elevations in the stimulation indices in response to known DR4restricted 70K T cell epitopes that were presented by the irradiated APCs from syngeneic naive mice. T cell lines proliferating in response to RNP peptides were generated as described above. The V␤ regions were sequenced using primers for the TCR heavy-chain V␤ region, as previously described (8), to isolate the recombined area of TCRs from RNPresponsive cells. GREIDINGER ET AL Statistical analysis. Categorical variables were compared using Fisher’s exact test. Results were calculated using the Prism program, version 3.0 (GraphPad, San Diego, CA). RESULTS Disease induction with RNPⴙ splenocyte adoptive transfer. Direct immunization of DR4 mice (n ⫽ 80) with a single subcutaneous dose of 50 g of 70K peptide along with 50 g of U1 RNA adjuvant induced anti-70K immune responses in the mice, as confirmed by ELISA and/or immunoblotting. Of these mice, 47 (59%) were determined, by urinalysis, to be without proteinuria, while 33 (41%) had at least trace proteinuria. Among the 47 mice without proteinuria, 18 (38%) developed histologic manifestations of interstitial lung disease, whereas none of these mice developed histologically evident renal disease. In contrast, among the 33 mice with proteinuria, 5 (15%) developed histologic manifestations of lung disease (P ⫽ 0.027 versus mice without proteinuria, by Fisher’s exact test) and 9 (27%) developed histologic manifestations of renal disease (P ⫽ 0.0002 versus mice without proteinuria, by Fisher’s exact test). We therefore hypothesized that the presence or absence or proteinuria may be a marker of immunologic conditions that could favor the induction of nephritis over the induction of lung disease. We performed adoptive transfer studies using serum transfer of splenocytes from immunized DR4 mice into naive syngeneic recipients. Two months after transfer of 10 million RBC-depleted whole splenocytes from mice without proteinuria, 8 (57%) of 14 recipients developed interstitial lung disease, while only 2 (14%) of 14 mice showed evidence of renal disease without lung disease (Fisher’s exact P ⫽ 0.046) (Table 1 and Figure 1). Examination of the joints, heart, liver, spleen, skin, esophagus, salivary glands, and intestines revealed no gross or histologic differences between the directly immunized mice and the recipients of the splenocyte adoptive transfer. In contrast, 2 months after transfer of 10 million splenocytes from mice with proteinuria to naive syngeneic recipients, none of the 5 recipient mice developed lung disease, and 3 (60%) of 5 developed renal disease. Similar transfer of 10 million splenocytes from naive syngeneic donors led to no identifiable lung or renal disease in any of the 5 recipients. We could identify anti-RNP antibodies and anti-RNP T cells in the recipients of cells from immunized mice but not in the recipients of cells from naive mice. TISSUE TARGETING OF ANTI-RNP AUTOIMMUNITY 537 Table 1. Protective effect of myeloid DCs on CD4⫹ cell transfer– induced nephritis* Transfer experiment Whole splenocytes Fresh CD4⫹ cells alone Cultured CD4⫹ cells Fresh CD4⫹ cells plus myeloid DCs Myeloid DCs alone No. of recipients No. (%) developing lung disease No. (%) developing renal disease 14 10 11 5 8 (57) 2 (20) 3 (27) 1 (20) 2 (14) 7 (70)† 8 (73)‡ 0 (0)§ 5 3 (60) 0 (0) * The development of lung disease and renal disease was assessed in naive syngeneic female mice 2 months after adoptive transfer of 10 ⫻ 106 red blood cell–depleted whole splenocytes, 2 ⫻ 106 purified splenic CD4⫹ cells that were either immediately transferred after purification (fresh) or cultured for 2 weeks in vitro with irradiated naive syngeneic antigen-presenting cells and U1–70-kd small nuclear RNP fusion protein (70K) (cultured), and/or 0.1 ⫻ 106 myeloid dendritic cells (DCs). All donor mice were syngeneic animals killed 2 months after anti-70K immunization, and were confirmed to have anti-70K antibodies by enzyme-linked immunosorbent assay and to have consistently normal results on urinalyses (i.e., no proteinuria). † P ⫽ 0.01 versus renal disease after whole splenocyte transfer, by Fisher’s exact test. ‡ P ⫽ 0.005 versus renal disease after whole splenocyte transfer, by Fisher’s exact test. § P ⫽ 0.026 versus renal disease after fresh CD4⫹ cell transfer alone, by Fisher’s exact test. Renal disease after adoptive transfer of RNPⴙ CD4ⴙ splenocytes from donors without proteinuria. After harvesting cells from the spleens of immunized RNP⫹ DR4 mice without proteinuria, we separated the CD4⫹ cells from the spleens by magnetic bead positive selection. We then adoptively transferred 2 million CD4⫹ cells, approximating the number of CD4⫹ cells recoverable from a single immunized donor mouse Figure 1. Development of lung disease after adoptive transfer of splenocytes from immunized DR4 mice into naive syngeneic recipients. Representative hematoxylin and eosin–stained sections of the lung and kidney from a naive syngeneic recipient are shown. Transfer of whole splenocytes from a U1–70-kd small nuclear RNP fusion protein plus U1 RNA adjuvant–immunized HLA–DR4⫹ donor into the recipient resulted in a mixed connective tissue disease–like lung disease (A), but no evidence of renal disease (B). Results from urinalysis (not shown) were also normal. (Original magnification ⫻ 10 in A; ⫻ 40 in B.) Figure 2. Induction of nephritis, rather than lung disease, after adoptive transfer of CD4⫹ splenocytes from non-nephritic donors. Representative hematoxylin and eosin–stained sections of the lung and kidney from naive syngeneic recipients are shown. CD4⫹ cells were directly isolated from a U1–70-kd small nuclear RNP fusion protein (70K) plus U1 RNA adjuvant–immunized naive syngeneic donor and transferred into the recipient (A and B), or CD4⫹ cells were harvested from a 70K-immunized syngeneic donor and grown in vitro for 2 weeks with 70K-loaded, irradiated antigen-presenting cells from a naive syngeneic donor before being transferred into the recipient (C and D). In both cases, minimal interstitial pulmonary infiltrates were evident, but renal lesions consistent with nephritis were present. Results from urinalysis (not shown) also revealed active urinary sediment in both cases. (Original magnification ⫻ 10 in A and C; ⫻ 40 in B and D.) spleen, to naive syngeneic recipients. In contrast to our findings in recipients of whole splenocytes, renal disease was common in recipients of CD4⫹ cells from donors without proteinuria, since it was identified in 7 (70%) of 10 recipient mice (P ⫽ 0.01 versus whole splenocyte transfer recipients with renal disease, by Fisher’s exact test) (Table 1 and Figure 2). Lung disease was less common, being present in only 2 (20%) of 10 recipients of CD4⫹ cells from immunized mice. Transfer of 2 million CD4⫹ cells from naive mice to syngeneic donors under the same protocol led to no lung or kidney disease in any of the 5 recipients. To ensure that the nephritis induction was mediated by anti-RNP–specific T cells, we isolated CD4⫹ splenocytes in the same manner as described above and cultured the CD4⫹ cells with 70K antigen and irradiated syngeneic APCs from naive donors for 2 weeks, followed by separation of the T cells using Histopaque densitycentrifugation of the nonadherent cells. The T cells thus isolated from the recipients remained homologous to those isolated directly from immunized donors, with regard to RNP epitope responses and TCR V␤-region sequence (results not shown). After adoptive transfer of 2 million of these T cells from short-term culture into recipient mice, we found that the same manifestations of nephritis without 538 GREIDINGER ET AL Figure 3. Persistence of adoptively transferred myeloid dendritic cells (DCs) at 48 hours and 7 days after transfer, with variable tissue specificity to the lungs. Myeloid DCs were purified from the spleens of immunized mice, labeled with 5,6-carboxyfluorescein succinimidyl ester (CFSE), and transferred into 10-week-old naive syngeneic mice by tail-vein injection. At 48 hours and 1 week after cell transfer, single-cell suspensions of the spleens and lungs were prepared, treated with OptiPrep centrifugation to enhance for identification of DCs, stained with fluorochrome-labeled anti-CD11c and allophyocyanin, and assessed by flow cytometry. Histograms show CFSE fluorescence in relation to CD11c staining. Identical gating conditions were used for all experiments. A, Transfer of myeloid DCs from immunized donors without proteinuria resulted in ⬎30% of the gated cells from the lung (versus 1% from the spleen) showing CD11c⫹/CFSE⫹ staining consistent with that of donor-derived myeloid DCs, at both 48 hours and 1 week after transfer, with approximately equal intensity of CFSE staining at each time point, suggesting that there was minimal proliferation or other dilution of these CD11c⫹ cells between 48 hours and 7 days. B, Transfer of myeloid DCs from immunized nephritic donors resulted in fewer than 4% CD11c⫹/CFSE⫹ cells appearing in the lungs at 48 hours or at 1 week after transfer, while the migration of labeled myeloid DCs to the spleen was similar to that shown in A. lung disease had developed as was observed with direct CD4⫹ cell transfers (Table 1 and Figure 2, showing results representative of 1 of 2 separate experiments). The results revealed that CD4⫹ cells from 70K⫹ donors without renal disease expanded significantly in vitro, after a total of 11 recipients (6 from one experiment and 5 from the other) had received 2 million T cells each. Lung disease was observed in 3 (27%) of 11 recipients, while renal disease was seen in 8 (73%) of 11 recipients (P ⫽ 0.005 versus whole splenocyte transfer recipients with renal disease, by Fisher’s exact test). The renal lesions seen after CD4⫹ cell transfers were typified by histologic findings of glomerular proliferation, mesangial hypertrophy, and hematoxylin body TISSUE TARGETING OF ANTI-RNP AUTOIMMUNITY formation, which are characteristic of lupus nephritis, although immune complex deposition and complement depletion were generally not observed (results not shown). Among both the recipients of fresh CD4⫹ cells and the recipients of short-term cultured cells, the small group of mice with lung disease was equally divided between those without renal disease (1 of 10 fresh CD4⫹ cell recipients and 1 of 11 cultured CD4⫹ cell recipients) and those that also had renal disease (1 of 10 fresh CD4⫹ cell recipients and 2 of 11 cultured CD4⫹ cell recipients). Recipients of CD4⫹ cells from immunized mice showed anti-RNP T cell reactivity, as previously described (8). These results suggest that the effects of antiRNP–reactive T cells are sufficient to transfer anti-RNP autoimmunity from immunized donors to otherwise naive recipients. Moreover, the results demonstrate that in the absence of other signals, RNP-reactive T cells can direct the induction of glomerulonephritis. Effects of myeloid DC adoptive transfers from RNPⴙ donors. The divergence in tissue targeting between recipients of whole splenocyte transfers (frequent lung disease, less renal disease) and CD4⫹ cell transfers (frequent renal disease, less lung disease) from donor mice without proteinuria suggests that non-CD4⫹ splenocytes participate in the development of the lungtargeting pattern observed with whole splenocyte transfers. Based on our previous observations that Toll-like receptor 3 (TLR-3)–null mice were more susceptible to renal disease and less susceptible to lung disease, as compared with TLR-3–intact mice, after direct immunization with 70K plus U1 RNA (4,13), we considered whether, in a similar manner, TLR-3–expressing splenocytes could protect against nephritis induction. We therefore investigated myeloid DCs, a major known TLR-3–expressing immune cell type in the spleen (15). We separated myeloid DCs from RBC-depleted spleens by a 3-step process of OptiPrep density-gradient centrifugation, negative selection for plasmacytoid DCs with anti-murine PDCA-1 magnetic beads, and then positive selection for CD11c surface expression with magnetic beads. We found that this yielded ⬃200,000 cells per spleen from the immunized mice, with a population of myeloid DCs that was ⬎95% pure, as determined by flow cytometry using the CD11c and anti-murine PDCA-1 selection markers (results not shown). To establish that myeloid DC adoptive transfers from immunized mice would lead to engraftment in syngeneic naive recipients, we adoptively transferred naive mice with 100,000 myeloid DCs from immunized mice without proteinuria; the cells were labeled with CFSE prior to infusion. Recipient mice were then killed 539 Figure 4. Effect of myeloid dendritic cell (DC) adoptive transfers. Syngeneic naive recipient mice received 100,000 myeloid DCs from RNP-immunized mice without renal disease (A and B) or the same cells along with 2 million CD4⫹ cells from immunized mice with renal disease (C and D). Representative hematoxylin and eosin–stained sections of the lung (A and C) and kidney (B and D) of 2 different mice are shown. Adoptive transfer of myeloid DCs alone was associated with development of pulmonary infiltrates but not renal disease, whereas adoptive transfer of myeloid DCs plus CD4⫹ cells typically resulted in neither lung disease nor renal disease. (Original magnification ⫻ 10 in A and C; ⫻ 40 in B and D.) either 48 hours or 7 days after cell transfer. In analyses of cultures at 48 hours and 7 days after cell transfer, substantial and stable numbers of CFSE-expressing CD11c⫹ cells could be detected in the lungs (but not the spleens) of recipient mice (Figure 3A). However, the accumulation of lung myeloid DCs was insufficient to manifest as histologically significant lung inflammation at either the 48-hour or 7-day time points (results not shown). Using CFSE-labeled myeloid DCs from identically immunized donors with proteinuria (and no lung disease), we observed dramatically less trafficking of donor myeloid DCs to the lungs of naive recipient mice but comparable trafficking to the spleen (Figure 3B). Among the mice receiving 100,000 myeloid DCs from RNP⫹ mice with lung disease that were followed up for 2 months, 3 (60%) of 5 developed lung disease, whereas none of the 5 developed renal disease (Table 1 and Figure 4). These results were similar to the frequency of end-organ manifestations in recipients of whole splenocytes. To assess the ability of myeloid DCs to inhibit CD4⫹ cell–induced nephritis, we performed cotransfers. When the same 100,000 myeloid DCs from RNP⫹ mice were transferred along with 2 million CD4⫹ cells from RNP⫹ mice without renal disease, a dramatic drop 540 GREIDINGER ET AL Figure 5. Induction of nephritis after cotransfer of splenocytes from nephritic and non-nephritic immunized mice. Syngeneic HLA–DR4–transgenic B6 mice were immunized with U1–70-kd small nuclear RNP fusion protein plus U1 RNA adjuvant. CD4⫹ and CD4⫺ splenocytes from a typical mouse with mixed connective tissue disease and lung disease but no renal lesions (A and B) were mixed with CD4⫹ and CD4⫺ splenocytes from a mouse with active urinary sediment and histologic manifestations of nephritis but no lung disease (C and D). Representative sections from an adoptive transfer recipient of mixed cells are shown, demonstrating nephritis in the absence of lung lesions (E and F), which was typical of the findings in recipient mice in this experiment. (Original magnification ⫻ 10 in A, C, and E; ⫻ 40 in B, D, and F.) in the incidence of nephritis (none of 5 mice) occurred, as compared with that in mice receiving CD4⫹ cells alone (7 [70%] of 10; P ⫽ 0.026, by Fisher’s exact test) (Table 1 and Figure 4). In addition, among mice cotransferred with CD4⫹ cells and myeloid DCs, lung disease was observed in only 1 (20%) of 5 recipients. These results support the hypothesis that myeloid DCs from immunized mice can mediate a nephritis-protective effect in anti-RNP autoimmunity. In contrast, when we cotransferred CD4⫹ cells from immunized mice without renal disease along with myeloid DCs from nonimmunized syngeneic mice, we observed that nephritis developed in all 5 (100%) of the recipients, and lung disease developed in only 1 (20%) of the 5 recipients. The protective effect of myeloid DCs on CD4⫹ cell–mediated kidney disease was thus not present with myeloid DCs from nonimmunized mice, which were subjected to the same separation protocol. Adoptive transfer from nephritic donors. Directly immunized RNP⫹ DR4 mice that had active urinary sediment were also studied. The histologic lesions that were identified in these mice, similar to those observed in the T cell recipient mice described above, consisted of glomerular proliferation, mesangial hypertrophy, and hematoxylin body formation, without prominent immune complex deposition or complement depletion (results not shown). For adoptive transfer studies, mice with renal disease and no lung disease (determined on the basis of active urinary sediment and subsequent confirmatory histologic findings) were killed 2 months after immunization, and RBC-depleted splenocytes were purified as described above. Mixing studies were then performed, in which naive syngeneic DR4 recipient mice each received 5 million splenocytes from an RNP⫹ donor with lung disease and no kidney disease, and each also received 5 million splenocytes from an RNP⫹ donor with renal disease but no lung disease. When these recipients were killed 2 months after adoptive transfer, renal disease was found to be present in 4 (80%) of the 5 mice, and lung disease was found in none of the 5 mice (Figure 5). Using DC subsets derived from immunized donors with active urinary sediment, 4 (80%) of 5 recipients of plasmacytoid DCs developed renal disease, and none of these 5 recipients developed lung disease. In contrast, none of the 4 recipients of myeloid DCs from nephritic donors developed renal disease. Notably, lung disease also was not evident in any of the 4 recipients of myeloid DCs from immunized donors with proteinuria and no lung disease. We were able to detect, by ELISA and/or immunoblotting, anti-RNP antibody responses in recipient mice after transfer of immunized myeloid DCs alone from donor mice without proteinuria or after cotransfer of myeloid DCs along with CD4⫹ cells. Similarly, we detected anti-RNP antibodies after transfer of plasma- TISSUE TARGETING OF ANTI-RNP AUTOIMMUNITY cytoid DCs alone from nephritic mice (results not shown). Anti-RNP antibodies were not detected, however, in recipients of myeloid DCs from immunized nephritic donors. DISCUSSION These studies demonstrate that whole splenocytes, CD4⫹ T cells, or purified DC subsets from RNP-immunized mice can be sufficient to convey antiRNP autoimmunity to naive syngeneic mice. Furthermore, anti-RNP–reactive T cells can lead to either lupus-like nephritis or MCTD-like lung disease, depending on the cell types that are transferred or cotransferred. Effects mediated by TLR-3–expressing myeloid DCs on tissue targeting of anti-RNP responses appear to potentially account for the previously reported divergence in pulmonary and renal manifestations in directly immunized mice with anti-RNP immunity (4). Consistent with the reported roles of plasmacytoid DCs in lupus pathogenesis (16), adoptive transfer of a relatively small number of purified plasmacytoid DCs from immunized mice was also able to induce the development of glomerulonephritis in recipients. The lung disease observed in our mice appears to be indistinguishable from cases of lung disease in human MCTD, as determined by conventional hematoxylin and eosin staining. We did not observe substantial fibrosis in the mouse lungs, in contrast to the findings in the majority of cases of interstitial lung disease in patients with MCTD (17). Further studies using immunohistochemistry to characterize the murine and human infiltrates are warranted. The renal lesions in the mice in our study differed from those typically found in lupus nephritis, in that immunoglobulin or complement component deposits at the glomerular basement membrane were rare. Nevertheless, diffuse or focal proliferative patterns, mesangial involvement, and frequent development of hematoxylin bodies were evident in our mice, as in lupus nephritis. Reports have linked TLR-3 signals to exacerbated lung inflammation and, conversely, TLR-7 signals to reduced lung inflammation, in models of asthma, viral infection, and direct TLR agonist challenge (18–23). We hypothesize that in our system, lung-predominant disease similarly develops when TLR-3–activated myeloid DCs direct anti-RNP T cells to pulmonary, rather than renal, targets. The mechanism of action of myeloid DCs in mediating lung-specific tissue targeting has not yet been defined, and further study will be needed to establish the role, if any, of TLR-3 in this process. 541 Conversely, we hypothesize that renal disease develops when anti-RNP autoimmunity is augmented at an early stage by a TLR-7 signal, rather than being modulated by a strong TLR-3 signal. The inability of 70K-immunized splenocytes from non-nephritic mice to inhibit established nephritis in cotransfer experiments may reflect the involvement of immune mediators downstream of the point where myeloid DCs are able to inhibit nephritis induction, or may reflect the elaboration of factors that inhibit (putatively) TLR-3–induced myeloid DC activity. The observation that anti-RNP CD4⫹ T cell–induced nephritis was able to be inhibited by myeloid DCs in our studies suggests that 1) development of anti-RNP immunity can occur prior to a commitment of an autoimmune process to renal injury, and 2) cells other than CD4⫹ T cells emerge after the induction of anti-RNP immunity, which can drive the renal targeting in a manner that is resistant to reversal by activated myeloid DCs. Given the associations identified between lupus nephritis and TLR-7 (24), TLR-7– expressing cells such as plasmacytoid DCs or B cells appear to be logical candidates for playing roles in the development of renal disease downstream of anti-RNP T cells, and myeloid DCs cannot reverse this process. Since MCTD was first proposed clinically as a diagnostic entity, debate has occurred regarding whether MCTD is truly distinct from other established rheumatic diseases, including lupus (25–27). Consistent with this experience, the results of our study indicate that subtle shadings of difference exist in a mouse model of antiRNP autoimmunity between MCTD-like (lung disease) and lupus-like (renal disease) manifestations of tissue injury. In mice, the same T cells appear to be relevant to either lung disease or renal disease induction, with the tissue targeting determined by accessory cell types, including myeloid DCs and plasmacytoid DCs. Our finding that in adoptive transfer mixing studies, cells from mice with established renal disease are dominant over cells from mice with established lung disease may inform the clinical observation that MCTD can evolve into systemic lupus erythematosus, but that systemic lupus erythematosus seldom evolves into MCTD. Although our adoptive transfer studies show that either T cells or DCs can spark persistent autoimmunity in a naive host, there is no guarantee that either of these cell types participate in the early stages of actual spontaneously developing clinical disease. Studies in humans are needed to support the relevance of the observations made in mice. This study demonstrates, for the first time, that elements of the same immune responses to a rheumatic 542 GREIDINGER ET AL disease autoantigen can lead to divergent tissue targeting based on differences in innate immune effectors. Improved understanding of the cell and molecular biology underlying this process may allow for the development of new diagnostic and therapeutic approaches to systemic autoimmune diseases in this spectrum. 11. 12. AUTHOR CONTRIBUTIONS Dr. Greidinger had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study design. Greidinger, Hoffman. Acquisition of data. Greidinger, Zang, Fernandez, Berho, Nassiri, Martinez. Analysis and interpretation of data. Greidinger, Zang, Fernandez, Berho, Nassiri, Martinez, Hoffman. Manuscript preparation. Greidinger, Zang, Fernandez, Martinez, Hoffman. Statistical analysis. Greidinger. 13. 14. 15. REFERENCES 1. Sharp GC, Irvin WS, May CM, Holman HR, McDuffie FC, Hess EV, et al. Association of antibodies to ribonucleoprotein and Sm antigens with mixed connective-tissue disease, systematic lupus erythematosus and other rheumatic diseases. N Engl J Med 1976;295:1149–54. 2. Arbuckle MR, McClain MT, Rubertone MV, Scofield RH, Dennis GJ, James JA, et al. Development of autoantibodies before the clinical onset of systemic lupus erythematosus. N Engl J Med 2003;349:1526–33. 3. Greidinger EL, Hoffman RW. The appearance of U1 RNP antibody specificities in sequential autoimmune human antisera follows a characteristic order that implicates the U1–70 kd and B⬘/B proteins as predominant U1 RNP immunogens. Arthritis Rheum 2001;44:368–75. 4. Greidinger EL, Zang YJ, Jaimes K, Hogenmiller S, Nassiri M, Bejarano P, et al. A murine model of mixed connective tissue disease induced with U1 small nuclear RNP autoantigen. Arthritis Rheum 2006;54:661–9. 5. Keith MP, Moratz C, Egan R, Zacharia A, Greidinger EL, Hoffman RW, et al. Anti-ribonucleoprotein antibodies mediate enhanced lung injury following mesenteric ischemia/reperfusion in Rag-1⫺/⫺ mice. Autoimmunity 2007;40:208–16. 6. Fatenejad S, Mamula MJ, Craft J. Role of intermolecular/intrastructural B- and T-cell determinants in the diversification of autoantibodies to ribonucleoprotein particles. Proc Natl Acad Sci U S A 1993;90:12010–4. 7. Wang D, Hill JA, Jevnikar AM, Cairns E, Bell DA. Induction of transient arthritis by the adoptive transfer of a collagen II specific Th1 clone to HLA-DR4 (B1*0401) transgenic mice. J Autoimmun 2002;19:37–43. 8. Greidinger EL, Zang YJ, Jaimes K, Martinez L, Nassiri M, Hoffman RW. CD4⫹ T cells target epitopes residing within the RNA binding domain of the U1-70kD small nuclear ribonucleoprotein autoantigen and have restricted TCR diversity in an HLA-DR4 transgenic murine model of mixed connective tissue disease. J Immunol 2008;180:8444–54. 9. Hoffman RW, Takeda Y, Sharp GC, Lee DR, Hill DL, Kaneoka H, et al. Human T cell clones reactive against U-small nuclear ribonucleoprotein autoantigens from connective tissue disease patients and healthy individuals. J Immunol 1993;151:6460–9. 10. Holyst MM, Hill DL, Hoch SO, Hoffman RW. Analysis of human 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. T cell and B cell responses against U small nuclear ribonucleoprotein 70-kd, B and D polypeptides among patients with systemic lupus erythematosus and mixed connective tissue disease. Arthritis Rheum 1997;40:1493–503. Greidinger EL, Gazitt T, Jaimes KF, Hoffman RW. Human T cell clones specific for heterogeneous nuclear ribonucleoprotein A2 autoantigen from connective tissue disease patients assist in autoantibody production. Arthritis Rheum 2004:50:2216–22. Greidinger EL, Foecking MF, Schafermeyer KR, Bailey CW, Primm SL, Lee DR, et al. T cell immunity in connective tissue disease patients targets the RNA binding domain of the U1-70kDa small nuclear ribonucleoprotein. J Immunol 2002;169:3429–37. Greidinger EL, Zang YJ, Martinez L, Jaimes K, Nassiri M, Bejarano P, et al. Differential tissue targeting of autoimmunity manifestations by autoantigen-associated Y RNAs. Arthritis Rheum 2007;56:1589–97. Greidinger EL, Foecking MF, Magee J, Wilson L, Ranatunga S, Ortmann RA, et al. A major B cell epitope present on the apoptotic but not the intact form of the U1-70-kDa ribonucleoprotein autoantigen. J Immunol 2004;172:709–16. Edwards AD, Diebold SS, Slack EM, Tomizawa H, Hemmi H, Kaisho T, et al. Toll-like receptor expression in murine DC subsets: lack of TLR7 expression by CD8 alpha⫹ DC correlates with unresponsiveness to imidazoquinolines. Eur J Immunol 2003; 33:827–33. Banchereau J, Pascual V. Type I interferon in systemic lupus erythematosus and other autoimmune diseases. Immunity 2006; 25:383–92. Bodolay E, Szekanecz Z, Devenyi K, Galuska L, Csipo I, Vegh J, et al. Evaluation of interstitial lung disease in mixed connective tissue disease (MCTD). Rheumatology (Oxford) 2005;44:656–61. Le Goffic R, Pothlichet J, Vitour D, Fujita T, Meurs E, Chignard M, et al. Cutting edge: influenza A virus activates TLR3-dependent inflammatory and RIG-I-dependent antiviral responses in human lung epithelial cells. J Immunol 2007;178:3368–72. Rudd BD, Smit JJ, Flavell RA, Alexopoulou L, Schaller MA, Gruber A, et al. Deletion of TLR3 alters the pulmonary immune environment and mucus production during respiratory syncytial virus infection. J Immunol 2006;176:1937–42. Camateros P, Tamaoka M, Hassan M, Marino R, Moisan J, Marion D, et al. Chronic asthma-induced airway remodeling is prevented by Toll-like receptor-7/8 ligand S28463. Am J Respir Crit Care Med 2007;175:1241–9. Demedts IK, Bracke KR, Maes T, Joos GF, Brusselle GG. Different roles for human lung dendritic cell subsets in pulmonary immune defense mechanisms. Am J Respir Cell Mol Biol 2006; 35:387–93. Kato A, Favoreto S Jr, Avila PC, Schleimer RP. TLR3- and Th2 cytokine-dependent production of thymic stromal lymphopoietin in human airway epithelial cells. J Immunol 2007;179:1080–7. Jeon SG, Oh SY, Park HK, Kim YS, Shim EJ, Lee HS, et al. TH2 and TH1 lung inflammation induced by airway allergen sensitization with low and high doses of double-stranded RNA. J Allergy Clin Immunol 2007;120:803–12. Christensen SR, Shupe J, Nickerson K, Kashgarian M, Flavell RA, Shlomchik MJ. Toll-like receptor 7 and TLR9 dictate autoantibody specificity and have opposing inflammatory and regulatory roles in a murine model of lupus. Immunity 2006;25:417–28. Smolen JS, Steiner G. Mixed connective tissue disease: to be or not to be? [review]. Arthritis Rheum 1998;41:768–77. Sharp GC, Hoffman RW. Clinical, immunologic, and immunogenetic evidence that mixed connective tissue disease is a distinct entity: comment on the article by Smolen and Steiner [letter]. Arthritis Rheum 1999;42:190–1. Isenberg D, Black C. Naming names! Comment on the article by Smolen and Steiner [letter]. Arthritis Rheum 1999;42:191–3.