HLA-DRB1*0402 (DW10) Transgene Protects Collagen- Induced Arthritis-Susceptible H2A^q and DRB1*0401 (DW4) Transgenic Mice from Arthritis ¹

Certain HLA-DR\/DQ alleles are associated with susceptibility to autoimmune diseases such as insulin-dependent diabetes mellitus (1, 2), multiple sclerosis (3), and systemic lupus erythematosus (4). Human rheumatoid arthritis (RA) ³ predisposition has been associated with expression of some HLA-DR alleles, with HLA-DR4 probably being the most extensively studied linkage (5, 6, 7, 8). Other alleles, notably DR1, DR6, and DR10, have been reported with predisposition to RA in certain ethnic populations (9, 10, 11). Gregersen and coworkers (12) and Winchester et al. (13) formulated the ‘shared epitope’ hypothesis to account for the association of certain DR alleles with RA in various ethnic groups. According to their hypothesis, the third hypervariable region comprising residues 67–74 is shared among susceptibility alleles and is a critical region for selection of RA relevant autoreactive T cells. Importantly, the sequence motif of I/D/E/A expressed at positions 67, 70, 71, and 74 (as expressed in DRB1*0402) confers resistance to RA. Although the mechanisms by which alleles affect degrees of RA susceptibility is not understood, it is thought that the T cell selection (positive selection of potentially autoreactive T cells by susceptibility alleles or negative selection by resistant alleles) and the resultant effect on shaping T cell repertoire has significant clinical outcomes (14, 15). The shared epitope hypothesis, however, may not account for all DR alleles linked to RA. For example, HLA-DR3 and HLA-DR9, linked in some populations, have been exceptions to a shared epitope hypothesis (16, 17).

RA is a chronic inflammatory disease that is characterized by synovial inflammation and erosion of bone and cartilage leading to destruction of joints. Patients with RA have been shown to produce anti-collagen type II (CII) Abs, T cell reactivity to CII, and accumulation of CII-reactive T cells in RA synovial fluid suggesting that autoreactivity to collagen might be important in pathogenesis (18, 19, 20). Collagen-induced arthritis (CIA) in mice has served as a classical model for human RA. The advent of mouse class II knockout mice expressing human HLA-DR and HLA-DQ transgenes has significantly advanced the understanding of the role of individual HLA class II molecules in various clinical conditions including RA. We have reported that transgenic mice expressing HLA-DQA1*0301 and DQB1*0302, but lacking endogenous class II molecules, are highly susceptible to CIA (21) and can present multiple human CII peptides (22). These observations contributed to the concept that DQ may be a major contributing factor in human RA while HLA-DR may modulate RA severity. We have proposed that both the DQ and DR alleles contribute to pathogenesis of RA (23). HLA class II molecules determine the T cell repertoire in the thymus by presenting self-peptides (24). This is supported by the findings that MHC-derived peptides constitute some of the naturally processed peptides eluted from class II molecules (25, 26). Binding studies have demonstrated that RA-associated HLA-DR alleles bind fewer human CII peptides compared with HLA-DQ8 allele (22, 27), suggesting HLA-DQ molecules may be a major factor in conferring susceptibility to develop arthritis. HLA-DQ occurs in linkage disequilibrium with DR genes and is inherited en bloc as haplotypes (28). HLA-DQ8 is found in linkage disequilibrium with DRB1*0401 and in increased frequency in RA patients of some ethnic groups (9, 29, 30).

An RA-susceptible DR allele would presumably positively select an autoreactive T cell during thymic education while a nonsusceptible allele would negatively select autoreactive T cells or would be neutral. To test this hypothesis we introduced DRB1*0401 and DRB1*0402 transgenes in B10.RQB3 to generate DRB1*0401.B10.RQB3 and DRB1*0402.B10.RQB3 transgenic mice. B10.RQB3 mice develop severe arthritis following immunization with bovine type II collagen (BII) but not with porcine type II collagen (PII). Introduction of DRB1*0401 gene into B10.RQB3 mice leads to development of severe arthritis in these mice following immunization with PII, indicating gene complementation of HLA-DR and H2-A is required for predisposition to develop disease (31). We report that DRB1*0402 can modulate the H2-A-restricted response to CII and protect *0402.B10.RQB3 transgenic mice from arthritis. In vivo, depletion with anti-DR Abs suggests that the autoreactive T cells are selected in thymus by DR polymorphism. We generated double transgenic mice expressing both DRB1*0401 and *0402 molecules. These mice develop milder arthritis and have a reduced incidence of CIA.

The HLA-DRB1*0401 gene construct was made as previously described (32). DRB1*0401 gene was mutated in β2 domain (amino acids 110 and 139 DRB1*0401 (NT)) for better interaction with murine CD4. The overlap PCR method and DRB1*0401 cDNA template was used to generate the DRB1*0402 construct. The differences of the nucleotide sequence between DRB1*0401 and DRB1*0402 are: DRB1*0401, CAG AAG GAC CTC CTG GAG CAG AAG CGG GCC GCC; DRB1*0402, — — — A– — –A G-C G– — — —.

Briefly, the 5′ end and 3′ end complementary oligo primers of DR4β (DRB1*0402) were synthesized as DW10–5′ (5′-CCGGAATTCATGGTGTGTCTGAAGTTC-3′); Dw10-3′ (5′-GTGGAATTCTCAGCTCAGGAATCCTG-3′).

A pair of DRB1*0402-specific oligos were synthesized as follows: C1 (5′-AAGGACATCCTGGAGGACGAGCGGGCC-3′); C2 (5′-GGCCCGCTCGTCT TCCAGGATGTCC TT-3′).

The first two fragments were generated by PCR using DW10–5′ plus C2 and DW10-3′ plus C1 as primers and DRB1*0401 (NT) cDNA as template. The two fragments were purified and used in overlap extension to generate the full-length DRB1*0402 gene. This full-length gene was sequenced to exclude unexpected mutations, and then subcloned into pDOI-5 expression vector at EcoRI site downstream of the H2 Eα promoter and rabbit β-globulin intron (33).

The DRB1*0402 β(nontransgenic)/pDOI-5 construct was double digested with NruI and XbaI to remove the plasmid sequence and microinjected into fertilized eggs from (SWR×B10)F₁ mice. Viable embryos were reimplanted into oviducts of pseudopregnant foster mothers. Mice carrying the transgene were identified by Southern blot analysis using DRB1*0402 β cDNA construct as a probe. Founders were intercrossed and back-crossed to B10.RQB3 mice, which lack endogenous Eβ but express Eα intracytoplasmically. Thus the β-chain of DR molecule pairs with Eα to express DRB1*0402. For double transgenic mice, B10RQB3 transgenic mice expressing *0401 and *0402 were mated to generate a line expressing both DRB1*0401/*0402 B10.RQB3 mice. For convenience, DRB1*0402.B10RQB3 transgenic mice will be referred to as DW10 and DRB1*0401.B10RQB3 transgenic mice as DW4 and double transgenic mice expressing both DR4 molecules as DW4/DW10 in the study.

The expression of DRβ, CD4, TCR Vβ-chains on PBLs of transgenic mice were analyzed by flow cytometry using mAbs: L227 (anti-DR), MKD6 (anti-A^q), 14-4-4s (anti-Eα), GK1.5 (anti-murine CD4), B20.6 (anti-Vβ2), KT4-10 (anti-Vβ4), MR9-4 (anti-Vβ5.1.2), MR9-8 (anti-Vβ5.1), 44-22-1 (anti-Vβ6), TR310 (anti-Vβ7), KJ-16 (anti-Vβ8.1.2), F23.2 (anti-Vβ8.2), MR10-2 (anti-Vβ9), KT11 (anti-Vβ11), 14.2 (anti-Vβ14), KJ23a (anti-Vβ17) as previously described (34).

Purified native BII, PII, and human type II collagen (HII) were obtained by multiple step purification (35). To induce CIA, 8- to 12-wk-old transgenic mice and negative littermates were immunized with 100 μg of CII emulsified 1:1 with CFA H37Ra (Difco Laboratories, Detroit, MI) intradermally at the base of the tail. Animals received a booster of 100 μg of CII emulsified with IFA 28 days later. DW4 transgenic mice were used as controls. Mice were monitored for the onset and progression of CIA from 3 to 12 wk postimmunization. The arthritic severity of mice was evaluated as previously described with a grading system for each paw of 0–3 (36). The mean arthritic score was determined using arthritic animals only.

Mice were sacrificed 10–12 wk postimmunization, and paws were decalcified and fixed. Sections were stained with H&E and examined histologically for mononuclear infiltration and bone erosion.

L227 (250 μg), specific for HLA-DRB1, was i.v. administered to DW10 transgenic and negative littermate mice at different time points beginning when mice were 8 wk of age. A total of 1.5 mg L227 was injected in each mouse.

Levels of anti-bovine and anti-mouse CII IgG were detected in sera obtained 35 days following CII immunization by ELISA (37). Briefly, microtiter plates were coated overnight with CII (6 μg/well in KPO₄, pH 7.6) at 4°C, washed and blocked with 1% BSA in PBS (0.05% Tween 20). Sera were added in 4-fold dilutions (1/100, 1/400, 1/1600, and 1/65,000) and incubated overnight at 4°C. The plates were washed and peroxidase-conjugated goat anti-mouse IgG (Organon Teknika, West Chester, PA) was added for another overnight incubation at 4°C. After washing, o-phenylenediamine was added, and the colorimetric change was measured at 410 nm. All assays were performed in duplicate and were quantified against a standard curve obtained with known positive sera, arbitrarily determined to equal 100 U/ml Ab units.

Serum samples of PII immunized DW10, DW4, and negative littermates obtained 35 days postimmunization were tested for Ab reactivity with cyanogen bromide (CB) peptides as described (31). Rheumatoid factor was measured by ELISA in transgenic mice as previously described (34).

Mice were immunized with 200 μg of CII emulsified 1:1 in CFA (Difco) intradermally at the base of the tail and one hind foot pad. Ten days postimmunization, draining popliteal, caudal, and lumbar lymph nodes were removed and prepared for in vitro culture. Lymph node cells (1 × 10⁶) were challenged by adding 100 μl of medium (negative control), Con A (20 μg/ml, positive control), and native collagen (50 μg/ml). For inhibition experiments, culture supernatant containing mAb (25 μg/ml) GK1.5 (anti-CD4), anti-A^q (MKD6), or Lyt2 (anti-CD8) was added to the cells challenged in vitro with CII at 50 μg/ml. The cells were incubated for 48 h at 37°C. During the last 18 h the cells were pulsed with [³H]thymidine and the tritium incorporation was determined by liquid scintillation counting. Results are calculated as the mean count per minute (Δ cpm) of triplicate cultures containing Ag mean cpm of medium.

Transgenic and negative littermates were also tested for T cell response to self-DR peptide DW10 (65–79) and control peptide DW4 (65–79), synthesized and purified at Mayo Clinic Peptide Facility. The mice were primed with 200 μg of peptide and challenged in vitro with 100 μg/ml of the peptide. To determine the restriction molecule for the in vitro response, blocking studies using L227 (anti-DR) and MKD6 (anti-H2A) Abs were performed.

Capture ELISA was done for measuring cytokines IFN-γ, IL-2, -6, -10, -13, -18, and TNF-α, using kits according to manufacturer’s instructions (BD PharMingen, San Diego, CA).

Spleen cells were collected by centrifugation at 200 × g for 5 min and washed with cold PBS. Whole cell lysates were prepared using cold lysis buffer (50 mM Tris, pH 7.2, 100 mM NaCl, 1 mM DTT, 1% NP-40 and protease inhibitor mixture). Lysates were centrifuged at 10,000 × g and the supernatants were analyzed for the presence of BclII, Bcl-x_L, c-Jun N-terminal kinase (JNK), JNK-P, caspase-3 and, poly(ADP-ribose) polymerase (PARP) by immunoblotting using Abs against BclII, Bcl-x_L (Santa Cruz Technology, Santa Cruz, CA), caspase-3, PARP, and JNK and its phosphorylated form JNK-P (New England Biolabs, Beverly, MA).

For cell division, pooled lymph node cells from two to three primed mice were collected and stained with cell division tracking dye, CFSE, and cultured in vitro for 48 h with PII. Cells were stained with CD3-PE and analyzed by FACS.

To study the sensitivity of the proliferating cells to CII-induced AICD, CD3⁺ cells were stained with annexin-V conjugated with FITC after in vitro stimulation and analyzed by FACS.

The difference in the incidence of arthritis between groups was analyzed using the chi square test. Ab levels, onset of arthritis, and mean scores for arthritic mice were compared using Student’s t test.

Expression of DRB1*0402 molecule on the cell surface was studied by flow cytometry. All transgenic mice showed 20–40% cells positive for expression of DRB1*0402 molecule in peripheral blood. As a criterion for functional expression of DW10, we determined Vβ T cell repertoire profile. Fewer number of Vβ8.1.2- and Vβ8.2-positive T cells were selected by DW10 mice compared with negative littermates suggesting DW10 to be functional in transgenic mice (Table I). All mice showed partial deletion of Vβ5.1.2, Vβ5.1, and Vβ11 compared with B10 mice (21).

DW4, DW10 transgenic mice, and negative littermate controls were immunized with BII to ascertain whether the introduction of DW10 molecule alters CIA susceptibility of B10.RQB3. As shown in Fig. 1,A, mice expressing DW10 were protected against arthritis (p < 0.05) following immunization with BII. Introduction of DW4 in B10.RQB3 mice increases the incidence of CIA following immunization with PII (20–57%) and HII (17–60%). In contrast, DW10 did not significantly increase CIA incidence following PII or HII administration (Fig. 1,B and Table II). However, DW10 expression did result in a delayed onset (40 vs 72 days, p < 0.005) following HII injection (Fig. 1,C, Table II). In comparison with DW4 mice, HII-immunized DW10 mice showed significantly lower incidence (p < 0.05) and delayed onset (p < 0.001) with milder disease (p < 0.05).

DW10 mice were treated with anti-DR Ab to deplete DW10 in vivo (Fig. 2,A). Expression of HLA-DR was determined by flow cytometry before collagen administration and at different time points thereafter (Fig. 2 B). Mice exhibiting at least 40–50% decrease in HLA-DR expression were immunized with BII following one week of Ab treatment. Only two of seven mice (28%) developed one swollen digit in one hind paw after 7 wk of immunization, consistent with a requirement for T cell selection.

All arthritic mice developed anti-collagen Abs to heterologous and self-collagen. A comparison of levels of anti-CII Abs showed that DW10 mice, negative littermates, and DW4 (positive control) mice produced an equal amount of anti-PII and anti-mouse CII Abs (Fig. 3,A). In addition, arthritic mice primed with BII or HII also produced high levels of anti-CII and self-CII Abs (data not shown). We analyzed sera for reactivity to PII CB-digested fragments (CB 9/7, 8, 10, 11, 12). Expression of either DW4 or DW10 molecules did not significantly alter the pattern of Ab binding measured against PII CB polypeptides. Sera from both transgene positive and negative littermates showed a similar heterogeneous pattern of reactivity against major CB 12, CB 11, and CB 9/7 fragments (Fig. 3 B). No difference was observed in reactivity against CB fragments in CIA⁺ vs CIA⁻ transgenic mice.

DW10 mice and negative littermates were immunized with 200 μg of PII, HII, or BII and draining lymph node cells challenged in vitro. T cells from primed DW10 mice mounted a strong response to PII (Δ cpm 9356), although responses to BII and HII (Δ cpm 3104 and 2618, respectively) were much lower in DW10 mice compared with DW4 (Δ cpm 10,116, PII; 5861, BII, and 9815, HII) and transgene negative mice (Fig. 3 C). The in vitro response to PII was mediated by CD4 cells and H2-A restricted as observed by inhibition of response (90% with anti-CD4 Ab and 80% with anti-A^q Ab).

Transgenic mice were immunized with a 15mer peptide comprising residues 65–79 of DW10 and DW4 molecules. DW10 transgenic mice responded to non-self DW4 peptide (65–79) but not to self-peptide DW10 consistent with self-tolerance (Fig. 3 D). Addition of anti-H2A^q Ab (MKD6) in vitro resulted in inhibition of response in DW10 (80% inhibition of response to DW4 and 65% to DW10) and transgene negative (70% inhibition of response to DW4 and DW10 peptides) mice suggesting the response to be H-2A restricted. DW4 mice mounted a minimal response to DW10 but a stronger response to DW4 peptide. These responses were also H2-A restricted as observed by inhibition assay (85% and 75% to DW4 and DW10 peptides, respectively).

Draining lymph nodes from primed mice were cultured for 48 h in vitro in presence or absence of CII. At which time culture supernatants were harvested and assessed for the presence of IFN-γ and TNF-α (Th1), IL-13 and IL-10 (Th2), and IL-6 (regulatory) cytokines. DW10 transgenic mice produced high amounts of IFN-γ, IL-10, and IL-13 but undetectable amounts of TNF when immunized with PII (Fig. 4). Negative littermates produced similar amounts of IFN-γ in response to PII. Transgenic mice immunized with BII produced lower amounts of IFN-γ and undetectable levels of IL-10 and IL-13 (data not shown). Lymph node cells harvested from DW10 mice after 12 wk of immunization with BII or PII produced IL-6 and IL-10 with barely detectable IFN-γ when challenged in vitro with PII or BII (data not shown). DW4 mice develop severe CIA following immunization with onset of the heterologous type II collagen and produce high amounts of IFN-γ and TNF-α to PII while lower amounts of IL-10 were produced. Both strains of transgenic mice and negative littermates primed mice produce IL-6 (Fig. 6 A).

Mice primed with PII were sacrificed after 2 wk or 8–10 wk of immunization and spleen cells were collected. Whole cell extracts were prepared and analyzed for the expression of BclII. Levels of BclII were similar among the transgenic mice, negative littermates and positive controls 2 wk after immunization (data not shown). Lysates prepared following 8–10 wk of immunization clearly showed lower levels of BclII and Bcl-x_L in cells from DW10 mice compared with those from DW4 transgenic mice while similar levels were observed in negative littermates (Fig. 5, A and B). To further define the role of apoptosis in disease development in all three strains we also studied apoptosis markers, caspase-3 and PARP. There was an increased activation of PARP and caspase-3, as apparent by enhanced cleavage of both proteins in resistant mice compared with DW4 (Fig. 5, C and D). A stress-activated kinase, JNK, involved in apoptosis and proliferation is known to be induced by TNF-α. Because our results showed that TNF-α levels were undetectable in DW10 mice, we decided to determine the levels of JNK and its phosphorylation. Lysates from splenic cells showed similar amounts of JNK (Fig. 5,E) but lower levels of its phosphorylation in DW10 mice compared with DW4 mice following immunization (Fig. 5 F).

To determine whether DW10 can protect DW4 transgenic mice from arthritis, we generated double transgenic mice expressing both DR4 molecules, DW4/DW10. Cell characterization demonstrated that T and B cell profiles were similar in all mice (Table III). For CIA studies, single and double transgenic mice were immunized with PII and monitored for development of arthritis. A delayed onset with lower incidence of arthritis was observed in DW4/DW10 mice compared with DW4 mice, p < 0.05 (Table IV). Because mice with human transgenes are known to produce rheumatoid factor in CIA model (34), we analyzed rheumatoid factor in all transgenic mice. All the DW4 and 40% of DW4/DW10 mice produced rheumatoid factor, although no significant difference was observed in level of autoantibody. None of the DW10 mice produced rheumatoid factor (data not shown).

To investigate whether the modulation of disease was due to changes in cytokine profile, inflammatory cytokines IL-2, IL-6, IL-18, and TNF-α were measured by ELISA from the supernatants of splenic cells cultured in the presence of PII. Lower amounts of IL-2 and IL-18 were produced by DW10 and DW4/DW10 mice compared with DW4 mice (Fig. 6,A). To study the kinetics of cell division and proliferation, lymph node cells CD3⁺ cells were isolated from primed mice and challenged in vitro with PII after staining them with CFSE. Higher number of dividing CD3⁺ T cells were observed in DW4 compared with DW10 and DW4/DW10 mice after 48 h in response to PII (Fig. 6,B). We further investigated whether increased apoptosis was observed in double transgenic mice. Splenic cells were isolated from all the primed transgenic mice and cultured in the presence of PII for 72 h and then stained with annexin-V. Equal percent of CD3⁺ cells were found to be annexin-positive in DW10 (42%) and B10.RQB3 mice (38%) while higher numbers of apoptotic cells were observed in DW4/DW10 mice (52%) compared with that of DW4 mice (25%), (Fig. 6 C).

DRB1*0402 (DW10) has been associated with protection from RA in human. In this study we report that DW10 can protect CIA susceptible mice against development of arthritis. B10.RQB3 mice develop severe arthritis following BII immunization but exhibit a milder and a lower incidence following PII challenge. However, introduction of a DW4 transgene renders them susceptible to PII-induced arthritis (31). No increase in incidence and severity of arthritis was observed in DW10 mice compared with parental mice after immunization with PII. Double transgenic mice expressing both DW4 and DW10 molecules had a lower incidence of CIA with PII compared with DW4 mice. DW10 mice are also protected from developing arthritis following immunization with BII. The present and previous data suggest that development of arthritis in B10.RQB3 mice is modulated by DR transgenes. Primed mice were capable of mounting a T cell response to PII in vitro and producing Abs to self and heterologous CII. However, the fine specificity of the Abs as measured against CB polypeptides was not different among the mouse groups in this study suggesting that the anti-collagen Ab response was primarily determined by H2-A molecule. These results confirm our previous studies with HLA-DR/DQ and DR/H-2A^q transgenic mice suggesting that complementation between DR and DQ/H2-A genes is required for the development of arthritis (31, 38).

MHC molecules have been shown to be important in selection of T cell repertoire in thymus and various Vβ TCRs (39). Our results are consistent with this hypothesis in that a difference in Vβ repertoire was observed between the two transgenic strains. Expression of DW10 transgene leads to negative selection of Vβ8.2 T cell population. Previous studies have shown that Vβ8.2 is important in induction of CIA in H2A^q mice. Absence of Vβ8.2 TCR population in FVB mice is one of the reasons for nonsusceptibility to CIA (40). In vivo depletion of DW10 did not make the transgenic mice susceptible to arthritis suggesting that negative/positive selection of autoreactive cells occur in thymus and is probably influenced by polymorphism in MHC. Contribution of DRB1 molecule toward T cell selection is further supported by the fact that DW10 mice mount milder response to CII immunization than DW4 mice and negative littermates although the major response is restricted by A^q molecule. These findings are reminiscent of the MHC linked protection described in diabetes model (41). According to this model, MHC molecules provide dominant resistance to a given autoimmune disease by deleting the most pathogenic autoreactive T cells rather than all autoreactive T cells. Thus the reason that RA patients carrying DRB1*0402 are rare might be explained by the fact that most of the pathogenic autoreactive cells have been deleted in these individuals and even if the other haplotype has susceptible alleles, resistance is dominant. Our results with double transgenic DW4/DW10 mice are in agreement with this hypothesis. In these mice we observed lower incidence and delayed onset of arthritis with only 40% of mice producing RF compared with 100% of DW4 mice tested. In H2-A^q mice pathogenic response to CIA has been mapped to CB11 region of CII with amino acid 260–270 being the immunodominant region (42). Interestingly, both transgenic strains, DW10 and DW4, mount response to peptide 254–270 of human type II collagen although a lower response was observed in DW10 mice (Ref.31 and data not shown). Both DW4 and DW10 have been shown to be able to bind this peptide, albeit with different anchors in P1 pocket (43). Presentation of this peptide in double transgenic mice might lead to increase in proliferation as observed with CFSE staining after in vitro culture with PII thus reaching a threshold for autoreactivity and development of arthritis.

Both transgenic strains produce high amounts of IFN-γ and IL-10 in vitro when challenged with PII. IL-10 has been associated with protection from arthritis in most of the experimental models studied. IL-10 deficiency in a CIA resistant mouse leads to severe disease (44). IL-10 has been shown to be produced by regulatory cells that have become anergic in vivo (45) or by CD4⁺ T cells in response to autoantigen presented by B cells (46). DW4 mice produce higher amounts of proinflammatory cytokines, IL-18 and TNF-α, than DW10 and DW4/DW10 mice thus indicating that a balance of Th1/Th2 cytokines is required for development of disease. IL-18 is known to be induced by TNF-α and has recently been shown to be important in CIA (47). Inflammatory cytokines TNF-α and IL-1 have been shown to induce activation of JNK, a stress activated kinase. Using JNK inhibitor and JNK knockout mice, it has been shown that it is required for the expression of metalloproteinase and joint destruction in inflammatory arthritis (48). JNK has also been implicated in signaling cell survival (49). We observed increased phosphorylation of JNK in cells of DW4 mice compared with DW10 mice.

Our data using annexin support the previously described results. Double transgenic and DW10 mice had higher number of CD3 cells positive for annexin. Increased apoptosis in DW10 mice was confirmed by enhanced cleavage of PARP and caspase-3. From our results of apoptotic proteins, it appears that early on there is no difference in apoptosis after immunization in DW4 and DW10 mice. However, a longer duration following immunization resulted in a different pattern in transgenic mice. A comparison of various proteins associated with apoptosis after 8–10 wk of immunization showed decreased BclII and Bcl-x_L levels in DW10 transgenic mice compared with DW4 transgenic mice. Increased levels of BclII and phosphorylation of JNK in DW4 mice might contribute to decreased cell death and to clonal expansion in these mice. Decreased proliferation with increased apoptosis might result in protecting the DW10 expressing mice from developing arthritis. CFSE staining of cultured lymph node cells confirmed decreased cell division in DW10 mice compared with both DW4 and DW4/DW10. This might explain stronger in vitro response to CII immunization and development of arthritis in DW4 mice and not in DW10 mice. Recent studies have shown oligoclonal expansion of T cells in periphery and joints of RA patients, which is thought to be driven in part by CII, suggesting that T cells reactive to CII may be mediators of RA pathogenesis (50, 51). In double transgenic mice resistant allele is able to modulate the development of arthritis by lower amounts of inflammatory cytokines and increased apoptosis leading to decreased incidence. Thus even though our results suggest that the major response to immunizing collagen in vivo is H2-A^q restricted, it is apparent that cytokine milieu produced by T cells selected in thymus lead to pathogenesis.

Thus on the basis of our observations, following mechanisms occurring alone or in combination can be suggested: 1) DRB1 polymorphism leads to selection of TCR in thymus thus controlling the selection of autoreactive cells available in periphery, 2) presentation of a peptide by DR molecule might enhance the autoantigenic immune response, and 3) variability in DRB1 loci is associated with differential AICD or apoptosis thus leading to protection/susceptibility.

We acknowledge Dr. Chris Krco for critical comments on the manuscript, Julie Hanson for maintaining transgenic mice, and Michele Smart for screening the transgenic mice.