Association of Increased F4/80high Macrophages With Suppression of Serum-Transfer Arthritis in Mice With Reduced FLIP in Myeloid Cells

Abstract

Objective: Macrophages are critical in the pathogenesis of rheumatoid arthritis (RA). We recently demonstrated that FLIP is necessary for the differentiation and/or survival of macrophages. We also showed that FLIP is highly expressed in RA synovial macrophages. This study was undertaken to determine if a reduction in FLIP in mouse macrophages reduces synovial tissue macrophages and ameliorates serum-transfer arthritis.

Methods: Mice with Flip deleted in myeloid cells (Flip^f/f LysM^c/+ mice) and littermate controls were used. Arthritis was induced by intraperitoneal injection of K/BxN serum. Disease severity was evaluated by clinical score and change in ankle thickness, and joints were examined by histology and immunohistochemistry. Cells were isolated from the ankles and bone marrow of the mice and examined by flow cytometry, real-time quantitative reverse transcriptase-polymerase chain reaction, or Western blotting.

Results: In contrast to expectations, Flip^f/f LysM^c/+ mice developed more severe arthritis early in the clinical course, but peak arthritis was attenuated and the resolution phase more complete than in control mice. Prior to the induction of serum-transfer arthritis, the number of tissue-resident macrophages was reduced. On day 9 after arthritis induction, the number of F4/80^high macrophages in the joints of the Flip^f/f LysM^c/+ mice was not decreased, but increased. FLIP was reduced in the F4/80^high macrophages in the ankles of the Flip^f/f LysM^c/+ mice, while F4/80^high macrophages expressed an antiinflammatory phenotype in both the Flip^f/f LysM^c/+ and control mice.

Conclusion: Our observations suggest that reducing FLIP in macrophages by increasing the number of antiinflammatory macrophages may be an effective therapeutic approach to suppress inflammation, depending on the disease stage.

Figure 1. Flip^f/fLysM^c/+ KO mice exhibit enhanced early but attenuated late phase of STIA

(A). The course of the arthritis was documented as the clinical activity score (left panel) and the ankle swelling (Δ ankle thickness in mm for hind ankles, right panel). Data were combined from 5 independent experiments, employing ≥ 40 Flip^f/fLysM^c/+ or control mice. After sacrificing to harvest samples at days 2 and 4, 18–19 remaining mice were scored on day 9, and 7 mice on day14 post arthritis induction. (B). Representative H&E histology imaging from the ankles of control and Flip^f/fLysM^c/+ mice harvested at day 4 and day 9. (C). The histological score of ankle joints for joint and extra-articular inflammation, PMN infiltration, pannus formation and bone erosion over the course of STIA (n= 7–16 ankles/group). (D). IL-1β, IL-6 and IL-10 in ankles determined by ELISA. Data are presented as the mean ±1 SE. Significance between the groups was determined by unpaired 2-sided t-test. * represents p< 0.05, ** p < 0.01 and *** p< 0.001 between the indicated groups.

Figure 2. Attenuated STIA is associated with decreased osteoclastogenesis

(A). Representative TRAP-stained ankle sections from mice that were sacrificed on day 9 post arthritis induction. Numerous TRAP+ osteoclasts (indicated by the arrows) were detected in the ankle joints of control mice (left panel). The frequency of osteoclasts were scored (0–4) and analyzed (right panel, n=7 Flip^f/fLysM^c/+ mice and 8 controls). (B). The expression levels of RANKL and OPG in ankles at day 9 post arthritis induction mice was quantified by ELISA (n=17 Flip^f/fLysM^c/+ and 20 control ankles). Data are presented as the mean ±1 SE. Significance was analyzed by unpaired 2-sided t-test for score and OPG analysis, and by Mann-Whitney test for RANKL. *represents p < 0.05 and ** p < 0.01.

Figure 3. Flip^f/fLysM^c/+ ankles exhibit reduced synovial tissue resident macrophages

(A). A representative flow cytometric analysis of ankle macrophages under steady state conditions presenting the gating strategy. After exclusion of doublets and debris, and gating out granulocytes, dendritic cells and B cells, synovial macrophages were identified as CD64⁺CD11b⁺F4/80⁺. After removing Ly6C⁺ and MHCII⁺ cells, synovial tissue resident macrophages are identified as CD64⁺CD11b⁺F4/80⁺Ly6C⁻MHCII⁻. (B). Analysis of the total synovial macrophages and the percent of synovial tissue resident macrophages from ankles of Flip^f/fLysM^c/+ mice and the control littermates prior to induction of STIA. (n =13 for Flip^f/fLysM^c/+ mice and 15 for the controls). Data are presented as the mean ±1 SE. Significance was analyzed by unpaired 2-sided t-test. * represents p < 0.05, ** p < 0.01 and *** p < 0.001 between the indicated groups

Figure 4. Flip^f/fLysM^c/+ mice exhibit increased synovial F4/80⁺ macrophages by immunohistochemistry and increased F4/80^hi synovial macrophages in Flip^f/fLysM^c/+ mice which correlate with reduced STIA

(A). Immunohistochemistry employing anti-F4/80 was performed to identify the distribution of macrophages in the ankles of mice day 9 post arthritis induction. The left panels are representative anti-F4/80-staining of ankles from control and Flip^f/fLysM^c/+ mice. The F4/80⁺ cells are brown. Synovial tissue F4/80⁺ cells were counted and presented as the average from 3 fields for each ankle (n = 7 Flip^f/fLysM^c/+ and 8 controls) present at right. (B). Representative flow cytometry images of ankles from control and Flip^f/fLysM^c/+ mice 9 days post arthritis induction. After gating out granulocytes and CD11b⁻ cells, macrophages were identified as CD64⁺CD11b⁺, and further as F4/80^hi and F4/80^lo. (C). The correlations between arthritis severity (Δ ankle thickness, mm) and the number of F480^hi macrophages and the mean florescence intensity (MFI) of F4/80 expression are presented (n = 7 control and 8 Flip^f/fLysM^c/+ mice). Significance was determined by unpaired 2-sided t-test (A,B), * represents p < 0.05 and ** p < 0.01 between the indicated groups. Pearson linear correlation was performed to determine correlations (C).

Figure 5. During STIA, F4/80^hi synovial macrophages exhibit an M2-like phenotype and reduced Flip expression

Macrophages were purified by flow cytometry from the ankles of control and Flip^f/fLysM^c/+ mice 9 days post STIA induction. Macrophages were gated as CD11b⁺, CD64⁺ F4/80^hi and F4/80^lo as in Figure 4B. (A). The expression of IL-10, Resistin-like molecule alpha (Relm-a), Arginase (Arg-1) and inducible nitric oxide synthases (iNos) determined by qRT-PCR, presented as the fold of expression normalized to the F4/80^hi cells of the control mouse group. Data were compared between the F4/80^hi and F4/80^lo populations of the control or Flip^f/fLysM^c/+ mice for each molecule (n=6 control and 3 Flip^f/fLysM^c/+ mice). Significance was determined by unpaired 2-sided t-test (B). The expression of Flip in the same samples was also determined by qRT-PCR, presented as fold of expression normalized to control F4/80^hi cells. Analysis was performed by one-way ANOVA, and Tukey post-test. * represents p< 0.05, ** p < 0.01 and *** p< 0.001 between the indicated groups

Figure 6. No difference in the apoptosis or necrosis of synovial macrophages at day 9 of STIA between Flip^f/fLysM^c/+ and control mice

(A). Apoptotic cells were defined as Annexin V⁺ and 7AAD⁻ and necrotic cells as Annexin V⁺ and 7AAD⁺ in CD64⁺CD11b⁺F4/80^hi and F4/80^lo synovial macrophages, harvested at 9 days post arthritis induction. Representative flow images for apoptosis defined in the F4/80^hi population of control and Flip^f/fLysM^c/+ mice are presented in the panels on the left and the percent of apoptotic or necrotic cells in the panels on the right. This analysis was from 5 control and 7 Flip^f/fLysM^c/+ mice. Significance was determined by unpaired 2-sided t-test. (B). Bone marrow cells from Flip^f/fLysM^c/+ or littermate control mice differentiated for 7 days in the presence of M-CSF, with or without TNFα. FLIP expression were determined by Western blot. The blot presented is representative of 2 independent experiments.