Type I interferons modulate vascular function, repair, thrombosis, and plaque progression in murine models of lupus and atherosclerosis

Abstract

Objective: Patients with systemic lupus erythematosus (SLE) have a notable increase in atherothrombotic cardiovascular disease (CVD) which is not explained by the Framingham risk equation. In vitro studies indicate that type I interferons (IFNs) may play prominent roles in increased CV risk in SLE. However, the in vivo relevance of these findings, with regard to the development of CVD, has not been characterized. This study was undertaken to examine the role of type I IFNs in endothelial dysfunction, aberrant vascular repair, and atherothrombosis in murine models of lupus and atherosclerosis.

Methods: Lupus-prone New Zealand mixed 2328 (NZM) mice and atherosclerosis-prone apolipoprotein E- knockout (apoE(-/-) ) mice were compared to mice lacking type I IFN receptor (INZM and apoE(-/-) IFNAR(-/-) mice, respectively) with regard to endothelial vasodilatory function, endothelial progenitor cell (EPC) function, in vivo neoangiogenesis, plaque development, and occlusive thrombosis. Similar experiments were performed using NZM and apoE(-/-) mice exposed to an IFNα-containing or empty adenovirus.

Results: Loss of type I IFN receptor signaling improved endothelium-dependent vasorelaxation, lipoprotein parameters, EPC numbers and function, and neoangiogenesis in lupus-prone mice, independent of disease activity or sex. Further, acute exposure to IFNα impaired endothelial vasorelaxation and EPC function in lupus-prone and non-lupus-prone mice. Decreased atherosclerosis severity and arterial inflammatory infiltrates and increased neoangiogenesis were observed in apoE(-/-) IFNAR(-/-) mice, compared to apoE(-/-) mice, while NZM and apoE(-/-) mice exposed to IFNα developed accelerated thrombosis and platelet activation.

Conclusion: These results support the hypothesis that type I IFNs play key roles in the development of premature CVD in SLE and, potentially, in the general population, through pleiotropic deleterious effects on the vasculature.

Figure 1. Lupus-prone NZM 2328 (NZM) mice exhibit improved endothelial function and increased EPC numbers in the absence of type I IFN signaling

(A) Female NZM IFNAR^−/− (INZM) mice display increased aortic endothelium-dependent vasorelaxation to acetylcholine (Ach) at 30-weeks of age, when compared to female NZM mice (n=8 mice/group); (B) and increased numbers of BM and blood (Bld) EPCs (n=13 mice/group). (C) Male INZM mice also display improved endothelium-dependent vasorelaxation at 40-weeks of age when compared to age- and gender-matched NZM mice (n=6 mice/group). (D) Male INZM mice display increases in EPC numbers in the spleen (Spl) at 40-weeks, compared to matched NZM mice (n=6 mice/group). A and C represent mean ± SEM of % of EC75 contraction. B and D represent mean percentage ± SEM of EPCs of lineage-negative cells; *p<0.05, **p<0.01, ***p<0.001.

Figure 2. Lupus-prone NZM mice exhibit improved EPC function and in vivo neoangiogenesis in the absence of type I IFN signaling

Increased capacity of EPCs to differentiate into mature ECs is observed in female INZM mice (A) at 20-weeks of age (left), and at 30-weeks of age (right) when compared to age- and gender-matched NZM mice (n=6–8 mice/group). (B) Male INZM mice display significant improvement in BM-EPC differentiation into mature ECs at 40-weeks of age (n=9 mice/group). (C) Female and male INZM mice display enhanced in vivo neoangiogenesis, compared to age- and gender-matched NZM mice, as assessed by total mg hemoglobin (Hb)/g of s.c.Matrigel plug (n=16 mice/group, 25–30 weeks in females; n=14 mice/group, 30–40 weeks in males). (D) Representative images of Matrigel plugs demonstrating increased numbers of endothelial von Willebrand Factor⁺ cells (green) in INZM females mice, compared to matched NZM mice (n=3 mice/group; 25–30 week old females).There was approximately a 7-fold increase in wWF+ cells in the INZM female compared to NZM female images, when adjusted for DAPI+ cells. Results in A and B represent mean number of cells/high power field ± SEM; results in C are mean mg of Hb/g of matrigel plug ± SEM; *p<0.05. Scale is 200µM. Magnification is 40×.

Figure 3. Acute exposure to IFN-α impairs endothelium-dependent vasorelaxation and EPC function in lupus-prone and non-lupus prone mice

(A) NZM female mice exposed to IFN-α adenovirus (AdIFN-α) a t 1 2-weeks of age display decreased endothelium-dependent vasorelaxation (left), and reduced EPC differentiation (right) at 3 weeks post-injection, compared to age- and gender-matched NZM mice exposed to control adenovirus (AdControl) (n=10–11 mice/group). (B) Male NZM mice similarly exposed to AdIFN-α also demonstrate decreased endothelium-dependent vasorelaxation, compared to AdControl-treated mice (left), while EPC function was not significantly altered (right) (n=5–8 mice/ group). (C) Female non-lupus-prone BALB/c mice exposed to AdIFN-α at 12 weeks of age also exhibit a decrease in endothelium-dependent vasorelaxation (left) and EPC differentiation (right) at 3 weeks following AdIFN-α (n=12 mice/group). Results represent mean ± SEM, *p<0.05, **p<0.01.

Figure 4. Atherosclerosis severity, and inflammatory arterial infiltration are sensitive to perturbations in type I IFN signaling in ApoE^−/− mice

(A) ApoE^−/−IFNAR^−/− mice (18–20-wk old) display significantly reduced atherosclerotic lesion size when compared to age- and gender-matched ApoE^−/− mice (n=13 mice/group). No differences in atherosclerosis severity are observed when comparing females to males in either group. ApoE^−/− mice administered AdIFN-α at 12 wks of age display a trend towards increased lesion size at 20 wks of age (p<0.05, n=9 mice/group and p=0.06, n=13 mice/group, respectively). (B) Representative en face views of aortic trees stained with Oil Red O from one ApoE^−/− mice and one ApoE^−/−IFNAR^−/− mice, demonstrate decreased lesion size in ApoE^−/− mice lacking type I IFN-R. Percentage of atherosclerosis lesions for representative images in (B) are 25.8% for ApoE^−/− and 14.1% for ApoE^−/−IFNAR^−/− (C) ApoE^−/−IFNAR^−/− mice (18–20-wk old) display significantly reduced MØ and T cell infiltration of aortic valves, compared to ApoE^−/− mice. Results represent % of area infiltrated by MØs or T cells, after counting 6 random fields (n=6 mice/group).

Figure 5. Modulation of neoangiogenesis and HDL oxidation by type I IFN signaling

(A) Representative images and (B) quantification of cells costaining for Sca-1 and CD31 in the aortic root of ApoE^−/−IFNR^−/− and ApoE^−/− mice (original magnification 400X). Sca-1 staining was graded according to intensity on a 0 to 2+ scale (0, no staining; 2+, maximum intensity staining). The cells stained with maximum-intensity were counted in three different representative fields per slide (n=5 mice/group). Bar graphs in (B) represent the number of double positive cells ±SEM. The highlighted box represents a high magnification region displaying Sca-1⁺CD31⁺ cells in the ApoE^−/−IFNR^−/−. C) ApoE^−/− IFNAR^−/− mice (20–26 weeks of age) have increased neoangiogenesis, when compared to gender- and age-matched ApoE^−/− mice. Neoangiogenesis is assessed as mg hemoglobin (Hb)/g matrigel plug (n=11 mice/group); *p<0.05. (D) Lupus-prone female NZM have increased levels of proinflammatory HDL, which significantly decreases in the absence of type I IFN signaling. Results represent mean±SEM of HDL nitrotyrosine µmol/mol tyrosine of 8–10 mice/group; age 30 week-old mice).

Figure 6. Type I IFNs modulate thrombosis, endothelial and platelet activation in atherosclerosis-prone and lupus-prone mice

(A) ApoE^−/− and NZM mice administered AdIFN-α display significantly shorter time to occlusive thrombosis in the mid-common carotid artery, compared to age- and gender-matched mice administered AdControl (n=10–14 mice/group; age=10–14 week old mice), while lack of type I IFN signaling in these mice has no effect on time to occlusive thrombosis (n=10–14 mice/group; age=20–22 week-old mice). (B) Shorter time to occlusive thrombosis correlates with increased serum soluble P-selectin in ApoE^−/− (n=8 mice/group; age=10–11 weeks) and NZM (n=6 mice/group; age=14 weeks) mice. AdIFN-α also increases serum P-selectin in BALB/c mice (n=6 mice/group; age 17 weeks). (C) INZM mice have significantly decreased serum P-selectin compared to NZM mice (n=10–15 mice/group; age=30 week old). Type I IFNs also modulate platelet activation as assessed by (D) P-selectin expression on platelets (left) and percentage of leukocyte:platelet aggregates (right) (n=9–10 mice/group and n=11–16 mice/group, respectively; age=20–25 week old mice). Results represent mean ± SEM, *p<0.05, **p<0.01.