Overexpression of membrane-bound fas ligand (CD95L) exacerbates autoimmune disease and renal pathology in pristane-induced lupus

Abstract

Loss-of-function mutations in the Fas death receptor or its ligand result in a lymphoproliferative syndrome and exacerbate clinical disease in most lupus-prone strains of mice. One exception is mice injected with 2,6,10,14-tetramethylpentadecane (TMPD), a hydrocarbon oil commonly known as pristane, which induces systemic lupus erythematosus-like disease. Although Fas/Fas ligand (FasL) interactions have been strongly implicated in the activation-induced cell death of both lymphocytes and other APCs, FasL can also trigger the production of proinflammatory cytokines. FasL is a transmembrane protein with a matrix metalloproteinase cleavage site in the ectodomain. Matrix metalloproteinase cleavage inactivates membrane-bound FasL and releases a soluble form reported to have both antagonist and agonist activity. To better understand the impact of FasL cleavage on both the proapoptotic and proinflammatory activity of FasL, its cleavage site was deleted through targeted mutation to produce the deleted cleavage site (ΔCS) mouse line. ΔCS mice express higher levels of membrane-bound FasL than do wild-type mice and fail to release soluble FasL. To determine to what extent FasL promotes inflammation in lupus mice, TMPD-injected FasL-deficient and ΔCS BALB/c mice were compared with control TMPD-injected BALB/c mice. We found that FasL deficiency significantly reduced the early inflammatory exudate induced by TMPD injection. In contrast, ΔCS mice developed a markedly exacerbated disease profile associated with a higher frequency of splenic neutrophils and macrophages, a profound change in anti-nuclear Ab specificity, and markedly increased proteinuria and kidney pathology compared with controls. These results demonstrate that FasL promotes inflammation in TMPD-induced autoimmunity, and its cleavage limits FasL proinflammatory activity.

FIGURE 1. Fas/FasL interactions promote TMPD-induced peritoneal inflammation

(A) Phenotype of peritoneal cavity washout cells (PWC) from untreated mice, and day 14 peritonal exudate cells (PEC) from TMPD-injected mice, obtained from the indicated genotypes. Top panel depicts all peritoneal cells gated on live/singlets events stained with a combination of F4/80 and CD11b. The lower panel depicts the CD11b⁺ cells stained for Ly6C and Ly6G. (B) Total number of cells collected from the peritoneum of untreated BALB/c mice (white bar), or TMPD-treated BALB/c, ΔCS, FasL^−/−, or gld mice (black bars), as well as cell numbers in the CD11b⁺, R0, R1, R2, and R3 gates and total number of CD8⁺ and CD4⁺ T cells (+/− SEM, n=4).

FIGURE 2. Relative levels of Fas expression by PWC and PEC reflects the extent of cell death in TMPD-injected mice

(A) Fas expression from cell types identified in Figure 1, obtained from untreated BALB/c and BALB/c^lpr/lpr mice (R0) and day 14 TMPD-treated mice BALB/c^lpr/lpr (R1, R2, R3, shaded histograms), BALB/c mice (R1 and R2, line), and FasL^−/− (R3, hatched histogram). (B) Fas expression of Ly6G⁺ neutrophil subsets based on high Ly6C (R2*) and intermediate Ly6C (R2^#) expression. (C) Number of dead cells in total, R1, and R2 PEC populations from untreated (white bars) or day 14 TMPD-treated BALB/c, ΔCS or FasL^−/− mice (black bars).

FIGURE 3. ΔCS T cells express higher levels of FasL than non-ΔCS T cells and induce more IL-1β from Fas positive target cells

(A) FasL expression of CD4⁺ T cells collected from untreated or day 14 TMPD-injected mice of the indicated genotypes (left, representative contour plots; right, % of cells in the indicated gate). (B) Histograms for FasL expression on unstimulated (shaded histogram) and anti-CD3 stimulated DO11.10 lpr T cells (line) or DO11.10 lpr ΔCS T cells (broken line). (C) Increasing numbers of DO11.10 lpr T cells (gray bars) or DO11.10 lpr ΔCS T cells (black bars) were co-cultured for 20h in the presence of OVA-peptide with BALB/c WT BMDCs, and IL-1β present in the supernatant was determined by ELISA. White bars depict control co-culture conditions in the absence of OVA-peptide and presence of DO11.10 lpr ΔCS T cells (-ova) or in the complete absence of T cells (- T cells). (D) DO11.10 lpr ΔCS T cells were co-cultured for 20h in the presence of ova-peptide together with BALB/c WT (white bars) or BALB/c^lpr/lpr BMDCs (black bars) and IL-1β present in the supernatant was determined by ELISA. Cell-free membrane-bound FasL-expressing microvesicles (mFasLVes) and poly(deoxyadenylic-thymidylic) acid (dAdT) were used as controls (as described previously (32)).

FIGURE 4. Both inflammatory monocytes and neutrophils from TMPD-injected mice up-regulate ISGs and proinflammatory cytokines

(A) Heat map depicting the gene expression of total peritoneal cells (PC) or indicated peritoneal cell BM populations from uninjected (−) or TMPD-treated (+) BALB/c mice. R0 cells (from untreated mice) or R1, R2 and R3 (from TMPD-injected mice) were sorted according to the gating strategy in Figure 1A after depletion of B cells and T cells. Unactivated neutrophils were isolated from the bone marrow of untreated BALB/c mice. Samples from two different mice for the R1 and R2 groups are shown. (B) H&E stain of cells from the R1, R2 and R3 subsets used for gene expression profiling. (C) IL-18 cytokine concentration of peritoneal wash fluids collected from untreated mice (white bars) or day 14 TMPD-treated mice (black bars) (+/− SEM n=8). (D) Viperin mRNA expression of total splenocytes from day 14 TMPD-injected mice of the indicated genotypes (+/− SEM n=3).

FIGURE 5. TMPD induces significantly greater splenomegaly and neutrophil accumulation in ΔCS mice

(A) Spleen weight, total number of splenocytes, and percent total of the indicated subsets from untreated BALB/c mice or TMPD-injected mice 6 months post-treatment. Each dot represents an individual mouse (untreated BALB/c mice (n=7), TMPD BALB/c (n=13), TMPD ΔCS (n=13)). Line indicates the mean.

FIGURE 6. Antibody and autoantibody differences between TMPD-treated ΔCS and wild-type mice

(A) Total serum levels of IgG1, IgG2a and IgG2b in mice 6 month post TMPD and the calculated IgG2a/IgG1 ratio for BALB/c and ΔCS mice (BALB TMPD n=9 and TMPD ΔCS n=7). Each dot represents an individual mouse. (B) Representative images of HEp2 ANA staining patterns from untreated and TMPD-treated BALB/c and ΔCS mice. (C) SmB/D autoantibody titers of BALB/c and ΔCS mice 6 month post TMPD treatment. (D) MFI (mean +/− SEM, n=4 TMPD BALB/c and TMPD ΔCS n=6) for IgG and IgM autoantibodies against selected autoantigens as quantified by autoantigen array.

FIGURE 7. Lupus nephritis and proteinuria are exacerbated in TMPD-treated ΔCS mice

(A) Urine albumine (mg/ml) measured by ELISA and obtained from TMPD-treated BALB/c (n=17) and ΔCS mice (n=11) at 6 month post–TMPD injection. Statistical comparison was made using a two-tailed Mann-Whitney U test. (B) Interstitial and glomerular scores for renal pathology were determined at 6 months post-TMPD injection, as described in methods, and reported as mean values (+/− SEM, untreated BALB/c mice n=5, TMPD BALB/c n=12, TMPD ΔCS n=13). Each dot represents an individual mouse. (C) Glomerular IgG deposition was quantified (glomerular MFI of the green channel; representative images are depicted in Supplemental Figure 3B) from individual glomeruli of TMPD-injected mice 6 months post-treatment.