Inflammasome activation of IL-18 results in endothelial progenitor cell dysfunction in systemic lupus erythematosus

Abstract

Systemic lupus erythematosus (SLE) is an autoimmune disease with heterogeneous manifestations including severe organ damage and vascular dysfunction leading to premature atherosclerosis. IFN-α has been proposed to have an important role in the development of lupus and lupus-related cardiovascular disease, partly by repression of IL-1 pathways leading to impairments in vascular repair induced by endothelial progenitor cells (EPCs) and circulating angiogenic cells (CACs). Counterintuitively, SLE patients also display transcriptional upregulation of the IL-1β/IL-18 processing machinery, the inflammasome. To understand this dichotomy and its impact on SLE-related cardiovascular disease, we examined cultures of human and murine control or lupus EPC/CACs to determine the role of the inflammasome in endothelial differentiation. We show that caspase-1 inhibition improves dysfunctional SLE EPC/CAC differentiation into mature endothelial cells and blocks IFN-α-mediated repression of this differentiation, implicating inflammasome activation as a crucial downstream pathway leading to aberrant vasculogenesis. Furthermore, serum IL-18 levels are elevated in SLE and correlate with EPC/CAC dysfunction. Exogenous IL-18 inhibits endothelial differentiation in control EPC/CACs and neutralization of IL-18 in SLE EPC/CAC cultures restores their capacity to differentiate into mature endothelial cells, supporting a deleterious effect of IL-18 on vascular repair in vivo. Upregulation of the inflammasome machinery was operational in vivo, as evidenced by gene array analysis of lupus nephritis biopsies. Thus, the effects of IFN-α are complex and contribute to an elevated risk of cardiovascular disease by suppression of IL-1β pathways and by upregulation of the inflammasome machinery and potentiation of IL-18 activation.

Figure 1. Inflammasome upregulation occurs in SLE EPCs/CACs by IFN-α signaling in a JAK-dependent manner

Real-time PCR was performed in triplicate on mRNA from control and SLE EPC/CACs (n=11/group) cultured under proangiogenic stimulation for 72 hours, followed by culture in the presence or absence of 1000 U/ml recombinant IFN-α for 6 hours. A. IFN-α significantly upregulates mRNA of inflammasome components in control and lupus EPCs. Control (left) and SLE (right) EPCs/CACs treated with IFN-α were compared to untreated control or untreated SLE cells respectively. B. Inflammasome component transcripts in untreated or IFN-α-treated SLE EPCs/CACs are increased when compared to control EPCs/CACs and their upregulation correlates with type I IFN exposure. B.Untreated SLE (left) cells were compared to untreated control cells and IFN-α treated SLE (right) were compared to IFN-α treated control cells (n-11/group). C. Association of type I IFN-regulated genes (IFN-RG) in SLE patients (n=7)with various inflammasome components. Left: IFI44 and IFIT levels significantly correlated with caspase-1 (p=0.0459 and 0.008, respectively), while PRKR did not reach statistical significance (p=0.184). Right: IFN-RGs significantly correlated with AIM2 levels: PRKR p=0.036, IFI44 p=0.0042, IFIT 0=0004. Bottom: IFI44 (p=.02) and IFIT (p=0.02) significantly correlated with IL-18 levels, while PRKR did not reach statistical significance (p=0.08). Levels of IFN-RGs in SLE are reported as their increase over levels in healthy controls (n=5). D. Western-blot of pro-caspase-1 (top) or β-actin (bottom) as a loading control. Blots are representative of 3 controls and 3 SLE patients. E. Control and SLE EPC/CACs were pre-incubated with either vehicle (DMSO) or inhibitors of JAK (50μM) or PI3K (50 nM) for 30 minutes prior to addition of recombinant IFN-α as above. Fold change was expressed as compared to vehicle-treated cells(n=5). Results represent mean ±SEM. *=p<0.05, **=p<0.01, ***=p<0.001.

Figure 2. Inhibition of caspase-1 improves the capacity of human and murine lupus EPCs/CACs to differentiate into mature ECs in an IFN-α dependent manner

A–C. EPCs/CACs from controls (n=7) or SLE patients (n=10) were incubated with either vehicle (DMSO), 10 μM of the caspase-1 inhibitor ac-YVAD-cmk (A and B) or 1μM of the caspase-3 inhibitor ac-DEVD-cmk (B) in proangiogenic media for 14 days. ECs were quantified as those displaying dual uptake of acetylated-LDL (red) and UEA-1-lectin (green). Photomicrographs represent phase contrast (top) or fluorescent (bottom) images at 100x magnification (C). D. PBMCs were treated overnight with 10 ng/ml LPS to activate caspase-1 in the presence of absence of 10 μM of ac-YVAD-cmk. Cells were then lysed in SDS-page buffer and subjected to Western Blot analysis using an anti-caspase-1 antibody that recognizes both the pro (inactive) and p20 (active) forms. E. EPC-containing bone marrow cells from lupus prone NZM2328 (male n=7, female n=5) and NZM2328 lacking a functional type I IFN receptor (INZM) (male n=5, female n=4) were cultured in proangiogenic media in the presence of vehicle or 10μM ac-YVAD-cmk for 7 days. ECs were quantified as those displaying dual expression of acetylated-LDL (red) and BS-1-lectin (green). Cells were visualized by fluorescent microscopy. Results are expressed as percent improvement in EC differentiation in the presence of YVAD/vehicle alone. Bar graphs represent mean ± SEM of individual cultures performed in triplicate for each patient or mouse. *=p<0.05, **=p<0.01.

Figure 3. Caspase-1 blockade abrogates the inhibition of EPC/CAC differentiation induced by exogenous IFN-α

A. Control (n=3) or SLE (n=6) EPCs/CACs were cultured in proangiogenic media in the presence of vehicle or 10 μM YVAD for 30 minutes and then treated with 1000 U/ml of recombinant human IFN-α. Media was changed after 3 days followed by 11 more days of growth in proangiogenic media without additional inhibitors or IFN-α. On day 14 of culture, ECs were quantified as in Figure 2. B. Bone marrow cells from NZM (n=7) and Balb/c (n=8) mice were isolated and cultured in pro-angiogenic media under similar conditions as in Figure 3A with or without 1000 U/ml of recombinant murine IFN-α. ECs were quantified as in Figure 2. Results represent mean ±SEM of experiments completed in triplicate for each patient or mouse. *=p<0.05, **=p<0.01.

Figure 4. Circulating IL-18 levels correlate with EPC/CAC dysfunction in SLE

A. The percentage of lupus mature ECs as compared to control mature ECs obtained after 14 days of EPC/CAC culture is plotted versus the presence or absence of detectable IL-18 in SLE serum (n=30). B. Levels of IL-18 in SLE serum (n=32) were plotted against the ratio of SLE ECs after 14 days of culture as compared to age and gender matched controls; r²=0.226. C. Serum IL-18 concentrations were plotted against SLEDAI disease activity scores determined on the day of blood draw; r²=0.09. D. The average number of lupus mature ECs per power field after 14 days of EPC/CAC culture is plotted against SLEDAI disease activity scores determined on the day of blood draw; r²=0.002. E. Lupus serum IL-18 levels are plotted against the percentage of improvement in EC differentiation (number of mature ECs in the presence of YVAD per field/number of mature ECs in the presence of vehicle per field) observed after caspase-1 inhibition with YVAD; r²=0.3103; (n=15). *p=<0.05.

Figure 5. Exogenous IL-18 inhibits the capacity of control EPCs/CACs to differentiate into mature ECs

A. Graded concentrations of recombinant IL-18 were added in triplicate to healthy control EPCs/CACs (n=6) and cultured under proangiogenic stimulation. Media was changed after 3 days followed by 11 days of growth in proangiogenic media without further addition of IL-18. On day 14 of culture, ECs were quantified as those binding both acetylated-LDL (red) and UEA-lectin (green). Results represent mean ±SEM of mature ECs/high power field. B. Representative images from figure 5A showing decreased endothelial differentiation in healthy control EPCs/CACs in the presence of 1ng/ml IL-18. Photomicrographs represent phase contrast (top) or fluorescent (bottom) images at 100x magnification. *=p<0.05.

Figure 6. Neutralization of IL-18 in SLE EPC/CAC cultures restores endothelial differentiation and this is enhanced with addition of IL-1β

A. EPCs/CACs from control (n=8) or SLE (n=6) patients were cultured under proangiogenic stimuli in the presence of neutralizing anti-IL-18 (1μg/ml), anti IFN-α (2 μg/ml) or IgG1 isotype control (1μg/ml) antibodies. Media was changed after 3 days followed by 11 more days of culture in proangiogenic media without additional antibodies. On day 14 of culture, ECs were quantified as those which dually bound acetylated-LDL (red) and UEA-lectin (green). Bar graph represents mean ±SEM. *p=<0.05. All patient or control samples were treated in triplicate for each experimental variable. B. Representative image of an SLE EPC/CAC culture after treatment with isotype or anti-IL-18 antibodies as mentioned above. Photomicrographs represent phase contrast (top) or fluorescent (bottom) images at 100x magnification. C. EPCs/CACs from healthy controls (n=5) were cultured under proangiogenic stimuli in the presence of 1 ng/ml IL-18 with or without 1ng/ml IL-1β. Media was changed on day three and mature ECs were quantified on day 14 as in 6A. D. EPCs/CACs from SLE patients (n=9) were cultured under proangiogenic stimuli in the presence of neutralizing anti-IL-18(1μg/ml)or same concentration of IgG1 isotype control, both in the presence or absence of 1ng/ml IL-1β. Media was changed on day three and mature ECs were quantified on day 14 as in 6A.

Figure 7. The presence of anti-Ro antibodies in SLE serum correlates with circulating IL-18 levels and is negatively associated with lupus EPC/CAC differentiation capacity

A. Lupus serum IL-18 levels (n=69) are plotted according to the presence or absence of specific autoantibodies in autologous sera. B. Endothelial differentiation of lupus EPC/CACs was quantified as in other figures, and the percentage of mature endothelial numbers, compared to those of healthy controls, was plotted in relationship to the presence (n=16) or absence (n=13) of anti-Sm, presence (n=12) or absence (n=13) of anti-RNP or presence (n=12) or absence (n=19) of anti-Ro antibodies. *=p<0.05.

Figure 8. IFN-α mediates EPC/CAC dysfunction via inflammasome upregulation and skewing towards IL-18 production

Exposure of EPCs/CACs to IFN-α results in repression of IL-1β but upregulation of IL-18, caspase-1 and AIM2. In SLE, yet undefined priming and activation steps possibly influenced by the presence of anti-Ro immune complexes, result in activation of the inflammasome and processing of IL-18 into its mature and active form. IL-18 then exerts inhibitory effects on endothelial differentiation leading to decreased vascular repair and an increased risk of atherosclerotic disease. The inhibitory effects of IL-18 can be overcome by the presence of IL-1β.