Local microbleeding facilitates IL-6- and IL-17-dependent arthritis in the absence of tissue antigen recognition by activated T cells

Abstract

Cognate antigen recognition by CD4(+) T cells is thought to contribute to the tissue specificity of various autoimmune diseases, particularly those associated with class II MHC alleles. However, we show that localized class II MHC-dependent arthritis in F759 mice depends on local events that result in the accumulation of activated CD4(+) T cells in the absence of cognate antigen recognition. In this model, transfer of in vitro polarized Th17 cells combined with the induction of experimental microbleeding resulted in CCL20 production, the accumulation of T cells in the joints, and local production of IL-6. Disease induction required IL-17A production by transferred T cells, IL-6 and CCL20 expression, and STAT3 signaling in type I collagen-expressing cells. Our data suggest a model in which the development of autoimmune disease in F759 mice depends on four events: CD4(+) T cell activation regardless of antigen specificity, local events that induce T cell accumulation, enhanced sensitivity to T cell-derived cytokines in the tissue, and activation of IL-6 signaling in the tissue. This model provides a possible explanation for why tissue-specific antigens recognized by activated CD4(+) T cells have not been identified in many autoimmune diseases, especially those associated with class II MHC molecules.

Figure 1.

F759 mice expressing a single TCR variant that recognizes a nonjoint antigen develop arthritis. (a and b) Clinical arthritis scores for RAG2^−/− F759-P25 mice (black circles in a; n = 76, 73, 62, 56, 46, 30, 14, and 7 for 1, 2, 3, 4, 5, 6, 7, and 8 mo, respectively) and RAG2^−/− F759-OT2 (black circles in b; n = 5, 5, 5, 7, 8, and 2 for 1, 2, 3, 4, 5, and 6 mo, respectively) were determined as described in Materials and methods and compared with scores for control RAG2^−/− F759, RAG2^−/− F759^+/− P25, and RAG2^−/− F759^+/− OT2 mice. Control RAG2^−/− F759 (open squares in a and b; n = 15), RAG2^−/− F759^+/− P25 mice (open circles in a; n = 47, 47, 40, 37, 31, 20, 10, and 6 for 1, 2, 3, 4, 5, 6, 7, and 8 mo, respectively), and RAG2^−/− F759^+/− OT2 mice (open circles in b; n = 36, 36, 28, 22, 17, and 9 for 1, 2, 3, 4, 5, and 6 mo, respectively) were also examined (RAG2^−/− F759^+/− TCR Tg mice vs. RAG2^−/− F759-TCR Tg mice; *, P < 0.05; ***, P < 0.005). The mean scores ± SEM were determined as described in Materials and methods. P-values were calculated using Student’s t tests.

Figure 2.

F759 mice expressing a single TCR variant that recognizes a nonjoint antigen show homeostatic induction of Th17 cell development. Total cells (a), CD4⁺ T cells (b), memory/activated CD4⁺ T cells (c), and Th17 cells (d) were counted in spleen and lymph node tissues isolated from 6-mo-old RAG2^−/− F759-P25 mice (n = 14) or control mice (n = 11: RAGKO-P25 [n = 3] and F759^+/− RAGKO-P25 [n = 8]). (e and f) Serum concentrations of IL-17A and IL-6 were analyzed in 6-mo-old RAG2^−/− F759-P25 mice (n = 9 and 10, respectively) and control mice (n = 15 and 10, respectively). The number of total cells (***, P < 0.005), CD4⁺ T cells (*, P < 0.05), memory/activated CD4⁺ T cells (*, P < 0.05), and Th17 cells (*, P < 0.05), as well as the serum concentrations of IL-17A (*, P < 0.05) and IL-6 (*, P < 0.05), were all significantly higher in 6-mo-old RAG2^−/− F759-P25 mice. P-values were calculated using Student’s t tests. Horizontal bars indicate mean.

Figure 3.

Microbleeding contributes to the development of arthritis in F759 mice. (a) Clinical arthritis scores from the left legs of control C57BL/6 mice (open triangles; n = 4) and F759 mice (open circles; n = 4) subjected to microbleeding and intravenously injected with 5 × 10⁶ Th17 cells (*, P < 0.05; **, P < 0.01; ***, P < 0.005 for 3, 5, 6, 7, 9, 10, 13, and 14 d, respectively). (b) Clinical arthritis scores from the right legs of F759 mice (open circles; n = 4) or C57BL/6 mice (open triangles; n = 4) after injections of 5 × 10⁶ Th17 cells without microbleeding induction. (c) Clinical arthritis scores from the legs of F759 mice injected with 5 × 10⁶ Th17 cells and subjected to microbleeding (closed circles; n = 5) or F759 mice injected with 5 × 10⁶ activated nonpolarized CD4⁺ T cells and subjected to microbleeding (closed triangles; n = 5). Results were compared with clinical arthritis scores from the legs of F759 mice not subjected to microbleeding (open circles: mice receiving Th17 cells [n = 5]; **, P < 0.01; ***, P < 0.005 for 12, 14, 15, and 16 d; open triangles: mice receiving activated, nonpolarized CD4⁺ T cells [n = 5]). (d) Clinical arthritis scores from the legs of F759 mice subjected to microbleeding without transfers of effector T cells are also shown (open triangles; n = 4) and compared with scores from the legs of F759 mice injected with 5 × 10⁶ Th17 cells and subjected to microbleeding (open circles; n = 4; *, P < 0.05; ***, P < 0.005 for 6, 7, 10, 11, 13, and 14 d, respectively). (e) Number of total cells, CD4⁺ T cells, CD4⁺CD45.1⁺ T cells, and CD4⁺CD45.1⁺CCR6⁺ T cells from joint tissues of F759 mice that received transfers of 2 × 10⁷ Th17 cells in the presence or absence of microbleeding induction (24 h after transfer) were analyzed (untreated vs. microbleeding: *, P < 0.05; ***, P < 0.005). Horizontal bars indicate mean. (f) Joint tissues of F759 mice that did or did not receive transfers of 5 × 10⁶ Th17 cells in the presence or absence of microbleeding were prepared. IL-6 mRNA expression levels were analyzed (no microbleeding vs. microbleeding in the presence of intravenous injection of Th17 cells: *, P < 0.05). (g) Clinical arthritis scores from the legs of F759 mice after intravenous transfer of Th17 cells derived from C57BL/6 mice (open circles; n = 7) or IL-17A^−/− mice (open triangles; n = 5). All mice were subjected to microbleeding (WT vs. IL-17A^−/− Th17 cells: *, P < 0.05 for 10, 11, 12, and 14 d). (h and i) Clinical arthritis scores from the legs of positive control F759 mice (open circles; n = 8 for IL-6^−/− F759 mouse experiment; n = 10 for type I collagen-Cre STAT3^flox/flox F759 mouse experiment) compared with those from the legs of IL-6–deficient F759 mice (crosses; n = 6 in h) or type I collagen-Cre STAT3^flox/flox F759 mice (closed squares; n = 6 in i) after intravenous transfer of 5 × 10⁶ WT Th17 cells and microbleeding induction (IL-6–deficient F759: *, P < 0.05 for 9, 10, 13, and 15 d; type I collagen-Cre STAT3^flox/flox F759 mice: *, P < 0.05; **, P < 0.01; ***, P < 0.005 for 7, 8, 10, 12, and 15 d). The mean scores ± SEM were determined as described in Materials and methods. P-values were calculated using Student’s t tests. These experiments were performed at least twice independently. Representative data are shown.

Figure 4.

CCL20 contributes to the development of arthritis in F759 mice. (a) Mouse embryonic fibroblast (MEF) cells and synovial fibroblasts were prepared and stimulated with human IL-6 + soluble IL-6R and/or IL-17A for 3 h. Cells were harvested and total RNA were collected and assessed with real-time PCR for CCL20. (***, P < 0.005 for IL-6 and IL-17A). (b and c) Cells from F759 and C57BL/6 joints were prepared and CCL20 mRNA expression levels and CD4⁺CCR6⁺ T cell percentages were analyzed. (*, P < 0.05; ***, P < 0.005 for 8-wk-old [8wo] C57BL/6, 12-mo-old [12mo] C57BL/6, 8-wk-old F759, and 12-mo-old spontaneous arthritic F759). Horizontal bars indicate mean. (d) Clinical arthritis scores from the left legs of F759 (closed circles; n = 8) or C57BL/6 (closed triangles; n = 8) mice that received transfers of 5 × 10⁶ Th17 cells as well as CCL20 injections. Clinical arthritis scores from the left legs of F759 (open circles; n = 8) or C57BL/6 mice (open triangles; n = 8) after CCL20 injections without Th17 cell transfers are also shown (Th17 vs. saline: **, P < 0.01; ***, P < 0.005 for 4, 6, 8, and 11 d). (e) Cells from the joint tissues of F759 mice that received transfers of 5 × 10⁶ Th17 cells in the presence or absence of microbleeding were prepared and CCL20 mRNA expression levels were analyzed (untreated vs. microbleeding: *, P < 0.05). (f) Clinical arthritis scores from the left legs of F759 mice after CCL20 injections and intravenous transfer of Th17 cells derived from C57BL/6 mice (closed circles; n = 8) or IL-17A^−/− mice (open circles; n = 7; WT vs. IL-17A^−/− Th17 cells; *, P < 0.05; ***, P < 0.005 for 4, 9, and 11 d). (g) Clinical arthritis scores from the left legs of F759 mice (closed circles; n = 7) or IL-6^−/− F759 mice (open circles; n = 5) after intravenous transfer of 5 × 10⁶ WT Th17 cells plus CCL20 injection (*, P < 0.05; **, P < 0.01 for 3, 6, and 10 d). (h) Clinical arthritis scores from the left legs of F759 mice (closed circles; n = 6) or type I collagen-Cre STAT3^flox/flox F759 mice (open circles; n = 4) after intravenous transfer of 5 × 10⁶ WT Th17 cells and CCL20 injections (*, P < 0.05; ***, P < 0.005 for 6, 7, and 10 d). (i) Clinical arthritis scores from F759 mice that received transfers of 5 × 10⁶ Th17 cells in the presence of microbleeding and joint injections of lentivirus carrying shRNA specific for CCL20 (open squares; n = 6) or STAT3 (open triangles; n = 5), or a nontarget sequence (open circles; n = 4; nontarget sequence vs. CCL20 specific: ***, P < 0.005 for 11, 12, 13, and 14 d; nontarget sequence vs. STAT3 specific: **, P < 0.01; ***, P < 0.005 for 9, 11, 12, 13, and 14 d). Mean scores ± SEM were determined as described in Materials and methods. P-values were calculated using Student’s t tests. These experiments were performed at least twice independently. Representative data are shown.

Figure 5.

Th17 cell–mediated induction of the IL-6 amplification loop in joints is critical for arthritis development in F759 mice. (a) Clinical arthritis scores from the left legs of F759 mice after left leg joint injections of 5 × 10⁶ Th17 cells (circles; n = 9; *, P < 0.05; ***, P < 0.005 for 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, and 14 d), 5 × 10⁶ Th1 cells (triangles; n = 4; *, P < 0.05 for 7, 8, 9, 10, 11, 12, 13, and 14 d), and 5 × 10⁶ activated nonpolarized CD4⁺ T cells (diamonds; n = 6) compared with those from the left legs of F759 mice after left leg joint injections of saline (squares, n = 9). (b) Clinical arthritis scores from the left legs of C57BL/6 mice after left leg joint injections of 5 × 10⁶ Th17 cells (circles; n = 7), 5 × 10⁶ Th1 cells (triangles; n = 4), 5 × 10⁶ activated nonpolarized CD4⁺ T cells (diamonds; n = 4), or saline (squares; n = 7). (c) Clinical arthritis scores from the left legs of F759 mice after left leg joint injections of Th17 cells derived from C57BL/6 mice (circles; n = 8), IL-17A–deficient mice (triangles; n = 3), or saline (squares; n = 5; WT vs. IL-17A–deficient Th17 cells: *, P < 0.05; **, P < 0.01 for 4, 6, 7, 8, 9, 10, 11, 12, 13, and 14 d). (d) Clinical arthritis scores from the left legs of F759 mice (circles; n = 9) or IL-6–deficient F759 mice (triangles; n = 5) after left leg joint injections of 5 × 10⁶ Th17 cells (*, P < 0.05; **, P < 0.01; ***, P < 0.005 for 4, 6, 7, 8, 9, 10, 11, 12, 13, and 14 d). (e) Clinical arthritis scores from the left legs of F759 mice (circles; n = 8) or type I collagen-Cre STAT3^flox/flox F759 mice (triangles; n = 3) after left leg joint injections of 5 × 10⁶ Th17 cells (*, P < 0.05; **, P < 0.01; ***, P < 0.005 for 5, 6, 7, 8, 9, 10, 11, 12, 13, and 14 d). (f) Clinical arthritis scores from the left legs of F759 mice after left leg joint injections of Th17 cells derived from C57BL/6 mice (closed circles; n = 4), OT2 RAG2^−/− mice (diamonds; n = 3), or saline (open circles; n = 7; OT2 RAG2^−/−–derived Th17 cells vs. saline: *, P < 0.05; **, P < 0.01; ***, P < 0.005 for 2, 3, 4, 7, 8, 9, 10, 11, 12, and 14 d). Mean scores ± SEM were determined as described in Materials and methods. P-values were calculated using Student’s t tests. These experiments were performed at least twice independently. Representative data are shown.

Figure 6.

IL-17A in the joints induced the development of arthritis in F759 mice. (a) Clinical arthritis scores from the left legs of F759 mice after left leg joint injections of 0.1 µg IL-17A on days 0, 1, and 2 (circles; n = 5; *, P < 0.05 for 4, 7, and 10 d) or F759 mice after left leg joint injections of saline (triangles, n = 3). (b) Clinical arthritis scores from the left legs of C57BL/6 mice after left leg joint injections of 0.1 µg IL-17A (circles; n = 4) or saline (triangles, n = 4) on days 0, 1, and 2. These experiments were performed at least three times independently. Representative data are shown.