Systemic lupus erythematosus (SLE) is associated with a dramatic increase in atherothrombotic cardiovascular disease (CVD), which is not explained by the Framingham risk equation (1–3). To date, no drug has been proven effective in decreasing CV risk in SLE. Therefore, it is crucial to establish the key drivers of vascular insult and atherosclerosis progression in lupus, in order to identify effective therapeutic and preventive targets.
While immune dysregulation may play a major role in premature CVD in SLE, the etiology of accelerated atherosclerosis in this disease remains unclear. Nevertheless, in vitro studies from our group and others indicate that type I interferons (IFNs) could play crucial roles in CVD development in SLE. This may be due to the capacity of IFNα, and potentially other type I IFNs, to impair vascular repair (4), enhance foam cell formation (5), and activate platelets through changes in the megakaryocyte transcriptome (6). IFNβ, another type I IFN, may promote atherosclerosis by promoting macrophage recruitment to arteries (7).
IFNα leads to an imbalance of vascular damage and repair by impairing the phenotype and function of bone marrow (BM)–derived endothelial progenitor cells (EPCs) (4) and by inducing transcriptional repression of angiogenic factors (8). However, while a prominent role for IFNα in lupus pathogenesis and clinical manifestations is supported by human and murine studies (9), it is unclear if type I IFNs are major inducers of vascular damage in vivo in SLE.
Increased endothelial cell (EC) damage in SLE is associated with perturbations in endothelium-dependent vasorelaxation (10), an established predictor of atherosclerosis progression (11). While most mouse strains (including lupus-prone mice) are resistant to diet-induced atherosclerosis (12), endothelial dysfunction develops in various models, including the lupus-prone (NZB × NZW)F1 strain (13). Therefore, impairments in endothelium-dependent vasorelaxation can be used as a surrogate marker of vascular damage in murine models.
In this study, we demonstrate a prominent in vivo role of type I IFNs in the development of endothelial dysfunction, aberrant vascular repair, and atherothrombosis in murine models of lupus and atherosclerosis.
- Top of page
- MATERIALS AND METHODS
- AUTHOR CONTRIBUTIONS
Broad activation of type I IFN pathways in SLE may play crucial roles in disease pathogenesis and clinical manifestations (36). Recently, in vitro work from several groups has suggested a putative link between type I IFNs, vascular damage, and atherosclerosis progression in SLE (4–6). However, whether these effects contribute significantly to endothelial dysfunction in vivo had not been determined.
Our observations suggest that type I IFNs have notable pleiotropic effects on the vasculature, from early EC damage and aberrant vascular repair, to plaque development and destabilization, followed by development of ACS through thrombosis enhancement. We show that type I IFNs play a direct, important role in CV damage and endothelial dysfunction development in vivo in lupus-prone mice and atheroma-prone mice, as well as in mouse strains not typically associated with lupus or enhanced atherothrombotic risk.
The nomenclature for vascular progenitor cells is undergoing constant revision. For the purposes of this work, we continue to refer to these cells as EPCs. IFNα is toxic to BM and blood EPCs in vitro (4, 8), and impairments in their phenotype and function may promote the endothelial dysfunction that has been reported in murine and human lupus (10, 13). While endothelial function and EPC numbers are reduced and functionally impaired in chronic kidney disease (37), the modulation of endothelial and EPC function by type I IFNs appears to be independent of renal dysfunction or of other immune abnormalities characteristic of SLE. Indeed, male NZM mice, spared from overt lupus development and glomerulonephritis, still exhibit significant differences in vascular parameters when compared to male INZM mice. This is consistent with the results of previous studies of human EPCs/circulating angiogenic cells, where lupus disease activity did not explain the profound vascular phenotypic and functional abnormalities (4). These observations support a specific role for type I IFNs in endothelial dysfunction and loss of EPC numbers and function in vivo.
Our findings indicate that chronic and enhanced exposure to type I IFNs may be needed to induce significant decreases in EPC numbers in vivo, since acute exposure to IFNα did not lead to alterations in these parameters, although it significantly impaired EPC differentiation. It is also possible that these effects on the vasculature are mediated, at least in part, by other IFNs besides IFNα, since the knockout system used in this study deleted signaling by all type I IFNs (14, 15). Indeed, while the differential effects that various type I IFNs play in the vasculature and in endothelial repair remain unclear, recent evidence suggests a role for type I IFNs other than IFNα in antiangiogenic responses and atherosclerosis modulation (7, 38).
Type I IFNs can impair the ability of stem cells to repopulate the BM niche. Therefore, the negative impact of IFNs on EPC function could indicate a more general stem cell impairment (39). Furthermore, type I IFNs can induce apoptosis and senescence of mature ECs (40). Therefore, long-term exposure to type I IFNs, as is observed in SLE, could promote sustained vascular damage and reduced local repair, leading to enhanced EPC and mature EC consumption, as shown in atherosclerosis models (33). Impaired EPC differentiation by type I IFNs may be associated with induction of antiangiogenic signatures in these cells and in other tissue, as we previously described in human lupus in vivo (8). This could create a cytokine profile that enhances antiangiogenic responses, promotes vasculopathy, and accelerates atherosclerosis.
Lupus-prone mice are resistant to florid diet-induced atherosclerosis, unless crossed with proatherosclerotic models (41–44), or used to generate chimeras following lupus BM transplantation into atherosclerosis-prone mice (45). While those studies demonstrated that lupus immune dysregulation accelerates plaque development, they have not addressed the specific roles of type I IFNs and have been performed primarily in murine models that do not appear to depend on these cytokines for disease progression (46, 47).
Recent evidence points to a potentially crucial role for type I IFNs in atherosclerosis due to the ability of IFNα to increase foam cell formation (5), its increased expression in areas of unstable human arterial plaques (48), and the capacity of myeloid IFNβ to worsen atherosclerosis (7). Strong associations of type I IFN signatures in human SLE with impaired peripheral arterial tone (49) and carotid intima-media thickness have been found (50). In the present study, we present direct in vivo evidence that type I IFNs significantly modulate plaque development and the severity of arterial macrophage and T cell infiltration in atherosclerosis-prone murine systems. Under proatherogenic conditions, monocytes migrate to vessel walls and differentiate into macrophages that can become lipid-laden foam cells. The biologic properties of atherosclerotic plaque macrophages and T lymphocytes determine lesion size, composition, and stability (45, 51). Decreased macrophage infiltration in apoE−/−IFNAR−/− mice may be secondary to impaired monocyte recruitment and/or differentiation into foam cells, as supported by recent observations (7, 52).
The proatherogenic effects of type I IFNs may be partially mediated by deleterious modulation of lipoproteins. While HDL is atheroprotective through roles in reverse cholesterol transport and antiinflammatory properties, it may become proinflammatory and promote atherothrombosis (26). SLE patients have increased proinflammatory HDL (34, 35), an abnormality we have now observed in female NZM mice. While oxidized HDL levels were decreased in female INZM mice, we could not elucidate if this was due to abrogation of type I IFN signaling on lipoprotein oxidation, or related to lack of lupus development in these mice. Since male NZM mice do not develop increased oxidized HDL, we used these mice as a reliable comparison. Furthermore, AdIFNα administration did not increase HDL oxidation, even if it exacerbated lupus. It is possible that chronic, rather than acute, differences in type I IFN signaling are required for lipoprotein dysregulation. Nevertheless, our results suggest a putative role of these cytokines in lipoprotein oxidation and the redox environment in vivo, which requires further investigation.
While platelets are activated in SLE and display evidence of increased exposure to type I IFNs (6), the functional relevance of this phenomenon in vivo had not previously been addressed. In our study, NZM and apoE−/− mice exposed to IFNα had a shorter time to occlusive thrombosis upon photochemical injury. This was associated with enhanced platelet activation, independent of a lupus-prone genetic background. Since circulating activated platelets can exacerbate atherosclerosis in murine systems (53), it is possible that this type I IFN–induced phenomenon also contributed to plaque formation. Future studies should address how type I IFNs modulate endothelial and platelet activation in autoimmune and nonautoimmune backgrounds, and the interplay of these molecules with lipoproteins and other factors. Furthermore, our results shed light on previous reports that infections may precede or be associated with enhanced atherothrombotic risk (54, 55). Indeed, this may be mediated via type I IFNs, following exposure to microbial products (48).
The results of the present study also support the notion that treatments disrupting type I IFN signaling, which are currently being tested in various autoimmune diseases, could promote additional benefits by hampering CV risk. Therefore, it will be important to include biomarkers of vascular damage and functional studies of endothelial health as end points of efficacy analyses for patients receiving these agents. Whether inhibition of these pathways in the general population would also decrease atherosclerosis progression should be investigated.
- Top of page
- MATERIALS AND METHODS
- AUTHOR CONTRIBUTIONS
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Kaplan had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Thacker, Koch, Pennathur, Davidson, Eitzman, Kaplan.
Acquisition of data. Thacker, Zhao, Smith, Luo, Wang, Vivekanandan-Giri, Rabquer, Pennathur.
Analysis and interpretation of data. Thacker, Smith, Luo, Vivekanandan-Giri, Rabquer, Koch, Pennathur, Eitzman, Kaplan.