Author + information
- Received April 11, 2013
- Revision received August 5, 2013
- Accepted August 6, 2013
- Published online January 7, 2014.
- Lingfeng Qin, MD∗,†,
- Qunhua Huang, MD, PhD∗,
- Haifeng Zhang, PhD∗,
- Renjing Liu, PhD∗,
- George Tellides, MD, PhD∗,
- Wang Min, PhD∗∗ ( and )
- Luyang Yu, PhD∗,†∗∗ ()
- ∗Interdepartmental Program in Vascular Biology and Therapeutics, Departments of Pathology and Surgery, Yale University School of Medicine, New Haven, Connecticut
- †Institute of Genetics, College of Life Sciences, Zhejiang University, Hangzhou, Zhejiang, China
- ↵∗Reprint requests and correspondence:
Dr. Wang Min, Interdepartmental Program in Vascular Biology and Therapeutics, Department of Pathology, Yale University School of Medicine, New Haven, Connecticut 06520.
- ↵∗∗Dr. Luyang Yu, Institute of Genetics, Zhejiang University, Hangzhou, Zhejiang 310058, China.
Objectives The aim of this study was to determine the role of suppressor of cytokine signaling 1 (SOCS1) in graft arteriosclerosis (GA).
Background GA, the major cause of late cardiac allograft failure, is initiated by immune-mediated endothelial activation resulting in vascular inflammation and consequent neointima formation. SOCS1, a negative regulator of cytokine signaling, is highly expressed in endothelial cells (ECs) and may prevent endothelial inflammatory responses and phenotypic activation.
Methods Clinical specimens of coronary arteries with GA, with atherosclerosis, or without disease were collected for histological analysis. SOCS1 knockout or vascular endothelial SOCS1 (VESOCS1) transgenic mice were used in an aorta transplant model of GA. Mouse aortic ECs were isolated for in vitro assays.
Results Dramatic but specific reduction of endothelial SOCS1 was observed in human GA and atherosclerosis specimens, which suggested the importance of SOCS1 in maintaining normal endothelial function. SOCS1 deletion in mice resulted in basal EC dysfunction. After transplantation, SOCS1-deficient aortic grafts augmented leukocyte recruitment and neointima formation, whereas endothelial overexpression of SOCS1 diminished arterial rejection. Induction of endothelial adhesion molecules in early stages of GA was suppressed by the VESOCS1 transgene, and this effect was confirmed in cultured aortic ECs. Moreover, VESOCS1 maintained better vascular function during GA progression. Mechanistically, endothelial SOCS1, by modulating both basal and cytokine-induced expression of the adhesion molecules platelet/endothelial cell adhesion molecule-1, intercellular adhesion molecule-1, and vascular cell adhesion molecule-1, restrained leukocyte adhesion and transendothelial migration during inflammatory cell infiltration.
Conclusions SOCS1 prevents GA progression by preserving endothelial function and attenuating cytokine-induced adhesion molecule expression in vascular endothelium.
Pathological vascular remodeling is the major cause of cardiovascular diseases. A prototypical example of pathological vascular remodeling is graft arteriosclerosis (GA), also referred to as cardiac allograft vasculopathy, which leads to clinical failure of organ allografts after the first year post-transplantation (1,2). The development of GA arises from the recruitment of acute inflammatory cells (e.g., macrophages) and subsequent alloantigen-responsive T cells on recognition of graft endothelial cells (ECs). The local production of proinflammatory cytokines, in turn, drives vascular smooth muscle cell migration and proliferation within the neointima, leading to luminal obstruction and allograft ischemia. The critical initiating events in this process are EC activation and the consequent induction of endothelial adhesion molecules, including E-selectin, P-selectin, intercellular adhesion molecule (ICAM)-1, vascular cell adhesion molecule (VCAM)-1, and platelet/endothelial cell adhesion molecule (PECAM)-1. The enhanced expression of endothelial adhesion molecules provides an environment for initial adhesion and subsequent migration of inflammatory cells into the vessel wall. Leukocyte adhesion to activated ECs is a sequential, multistep process consisting of tethering, rolling, firm adhesion, and transmigration (3). Each adhesion molecule has specific functions in the adhesion cascade of inflammatory cell recruitment; selectins are associated with cell rolling, ICAM-1 and VCAM-1 correlate to firm adhesion, and PECAM-1 is responsible for paracellular transmigration. The major stimuli for induction of endothelial adhesion molecules are proinflammatory cytokines, which typically include interleukin (IL)-6 and interferon (IFN)-γ in GA.
Among IFN-γ–induced genes, suppressor of cytokine signaling 1 (SOCS1) is an attractive candidate as a putative regulator of GA pathogenesis. SOCS1, a member of the SOCS family of proteins, was first identified as a negative feedback inhibitor of cytokine signaling. So far, SOCS1 has been implicated in the regulation of dozens of cytokines, including IL-6 and IFN-γ. It not only functions as a powerful attenuator of JAK-STAT signaling, but also has been shown to disrupt other inflammatory pathways by regulating nuclear factor κB (4) and ASK1 degradation, based on our previous work (5,6). In addition to those in vitro studies, the physiological role of SOCS1 is extensively established in immune function and tumor progression. SOCS1 is abundantly expressed in the thymus and spleen and is induced in all immune cells by a large panel of cytokines and pathogens. SOCS1-deficient mice die approximately 2 to 3 weeks post-natally because of overactive inflammatory responses (7,8). On the other hand, SOCS1 overexpression in vivo results in a reduction of inflammatory cytokines produced by infiltrating immunocytes (9,10). The distinct role of SOCS1 in inflammation regulation is correlated with several immunological diseases, such as multiple sclerosis (11), inflammatory arthritis (12), and diabetes (13,14). Additionally, SOCS1 functions as a tumor suppressor in carcinogenesis. Decreased SOCS1 expression in cancer cells and increased expression in stromal cells have been observed in a variety of tumor samples. In clinical practice, SOCS1 has been set up as a diagnostic biomarker of tumors at both the transcriptional (15,16) and post-translational levels (17,18). Although we have previously dissected part of the SOCS1 signal transduction pathway in cultured ECs (5,6), little is known about the pathophysiological role of SOCS1 in the cardiovascular system. In the present study, we investigated the role of SOCS1 in vascular ECs within GA based on both analysis of clinical specimens and experimental research in mouse models of disease.
An expanded Methods section is provided in an Online Appendix that includes descriptions of clinical specimens, generation of SOCS1 transgenic mice, mouse aortic transplantation model, graft analyses, cell culture, immunoprecipitation and immunoblotting, leukocyte-endothelial adhesion and transmigration assays, statistical analyses, and Online Figures S1 to S6.
Loss of endothelial SOCS1 expression correlates with arterial disease in clinical samples
SOCS1 expression in human vascular cells has been described in several in vitro studies. To determine the expression pattern of SOCS1 in clinical specimens of vascular disease, human coronary arteries with GA from transplanted hearts and with atherosclerotic plaque, or no disease from nontransplanted hearts, were collected for histological examination. SOCS1 was abundantly expressed in nondiseased arteries, whereas it was dramatically reduced in diseased arteries, especially within the luminal endothelial layer identified by PECAM-1 staining (Figs. 1A and 1B, with quantification in 1B). However, no significant differences were detected at the messenger ribonucleic acid (mRNA) level between healthy and diseased arteries (Online Fig. S1), suggesting diminished SOCS1 protein synthesis. Histological analysis also demonstrated obvious intimal expansion with plentiful leukocyte infiltration, as shown by CD45 immunostaining of diseased arteries (Figs. 1C and 1D, with quantification in Fig. 1D). Therefore, to further determine a link between endothelial SOCS1 expression and vascular inflammation, we assessed the adhesion molecules and the markers of EC activation, ICAM-1 and VCAM-1, by immunofluorescence. In contrast to the pattern for SOCS1, expression levels of ICAM-1 and VCAM-1 were markedly enhanced, whereas that of PECAM1 was slightly increased in the endothelium of diseased arteries compared with nondiseased controls (Fig. 1E). These results suggest a protective role of SOCS1 in pathological vascular remodeling by preventing endothelial activation and vascular inflammation.
SOCS1 deletion in mice exacerbates GA and EC dysfunction
To further investigate the function of SOCS1 in vessel wall cells under pathological conditions, we used a mouse aorta transplantation model of GA that we established in a previous study (19,20). Briefly, a segment of male donor thoracic aorta is interposed into the abdominal aorta of a female recipient of the same background strain. The host then mounts an alloimmune response against the male-specific H-Y minor histocompatibility antigen expressed by the graft in which leukocyte-derived proinflammatory cytokines induce endothelial activation and drive graft neointima formation (19–22). Because SOCS1 knockout (SOCS1-KO) mice die perinatally, as a result of overproduction of inflammatory cytokines, we used SOCS1-KO mice from an IFN-γ–deficient background in which the mice survive normally (23). Transplantation of the IFN-γ–KO male C57BL/6 (B6) to female B6 aorta induced GA, as characterized by graft infiltration with leukocytes and neointima formation (Figs. 2A and 2B, with quantification in Fig. 2B). Strikingly, SOCS1/IFN-γ–KO male B6 donor grafts to female B6 recipients generated significantly larger neointima that contained more CD45-positive infiltrating leukocytes, resulting in lumen loss compared with the IFN-γ–KO donor group (Figs. 2A and 2B, with quantification in Fig. 2B).
Because EC activation and dysfunction has been implicated as the early step for arteriosclerosis progression, we investigated whether SOCS1 deficiency might affect endothelial function of the vessel wall. To this end, aortas from IFN-γ–KO or SOCS1/IFN-γ–KO mice were isolated for vessel function assays. Vascular reactivity of isolated aortic rings was determined by examining their responses to vasoconstrictor phenylephrine (PE), endothelium-dependent vasodilator acetylcholine (Ach), and the nitric oxide synthase inhibitor l-nitroarginine methyl ester (l-NAME). Compared with IFN-γ–KO aortas, those from SOCS1/IFN-γ–KO mice showed increased constriction in response to PE but reduced relaxation in response to Ach (Online Fig. S2). These results are consistent with our clinical observations discussed previously and support a role for SOCS1 in preventing GA pathogenesis in which SOCS1-regulated EC function may be a potential mechanism.
Vascular endothelial SOCS1 overexpression in mice inhibits GA
The preceding conclusions led us to hypothesize that SOCS1 may function as a negative regulator of endothelial activation during GA. To clarify this issue, we generated a VE-cadherin promoter-driven SOCS1 transgenic mouse (VESOCS1) (Online Fig. S3A) and backcrossed it onto the B6 background for more than 10 generations. EC-specific expression of SOCS1 was determined at both mRNA and protein levels (Online Figs. S3B to S3D). Nontransgenic wild-type (WT) littermates and VESOCS1 male mice were used as aorta donors in the GA model described earlier. Donor grafts were harvested at 2 weeks after transplantation for histological analysis and morphometric assessment. Aorta grafts overexpressing SOCS1 generated significantly less neointima and contained fewer CD45+ inflammatory cells than the WT vessels (Figs. 3A and 3B). To assess the effects on inflammation-induced EC signaling, IL-6 and IFN-γ were selected for evaluation as they represent major proinflammatory cytokines produced in GA (24–28). Real-time quantitative polymerase chain reaction results showed no significant differences between cytokine production of the WT group and that of the VESOCS1 group (Online Fig. S4). In contrast, there was decreased cytokine signaling as assessed by immunostaining for phosphorylated JAK2 and phosphorylated STAT3 in aortas overexpressing SOCS1 at 3 days post-operatively (Figs. 3C and 3D). The expression of these activated signaling molecules colocalized with PECAM-1, demonstrating events within the graft ECs. Because JAK-STAT signaling is one of the major pathways in proinflammatory cytokine-mediated EC activation, these results provide clues to the exploration of the effects of SOCS1 in GA progression.
SOCS1 is a negative regulator of endothelial activation in GA
To further determine the function of SOCS1 in GA, we examined samples for markers of EC activation early after transplantation of WT and VESOCS1 grafts prior to discernable neointima formation. At 3 days post-operatively, real-time quantitative polymerase chain reaction revealed decreased mRNA expression of the inducible endothelial adhesion molecules ICAM-1 and VCAM-1 (but not PECAM-1) in VESOCS1 aortas compared to that in the WT group (Fig. 4A). Immunostaining confirmed an inhibitory effect of SOCS1 on the expression of ICAM-1 and VCAM-1 and also of PECAM-1 in the graft endothelium (Fig. 4B). Correspondingly, we observed that numbers of infiltrating leukocytes were decreased in VESOCS1 grafts compared with those in the WT group (Figs. 4C and 4D). On the other hand, vessel function assays revealed reduced constriction of VESOCS1 graft in response to PE but enhanced relaxation in response to Ach when compared with B6 controls (Online Fig. S5). Taken together, these results demonstrate the direct role of SOCS1 in preventing endothelial activation and dysfunction during GA formation.
SOCS1 overexpression restrains expression of adhesion molecules and leukocyte-endothelial adhesion and transmigration
To define how SOCS1 mediates gene expression of PECAM-1, ICAM-1, and VCAM-1 in vascular EC, we further examined the effects of the SOCS1 transgene in isolated primary mouse aortic EC. We observed that the basal level of PECAM-1, which is both a major scaffold protein and also an adhesion molecule of ECs, is clearly suppressed in the presence of the SOCS1 transgene (Fig. 5A). Consistent with in vivo findings, PECAM-1 mRNA was not altered by SOCS1 transgene expression (data not shown), suggesting that SOCS1 may regulate PECAM-1 expression at a post-translational level. We investigated whether SOCS1 mediates PECAM-1 proteasomal degradation as reported previously (29). In aortic ECs from VESOCS1 animals, the low expression level of PECAM-1 was strongly reversed by the presence of MG132, a panproteasome inhibitor (Fig. 5A). Moreover, associations between SOCS1 and PECAM-1 could be detected by a coimmunoprecipitation assay with anti-PECAM-1 followed by Western blotting for SOCS1 in the same primary cells (Fig. 5B). Therefore, these results demonstrate that SOCS1 binds PECAM-1 and mediates its proteasomal degradation.
In contrast to PECAM-1, we observed that both the mRNA (not shown) and the protein (Fig. 5C) levels of ICAM-1 and VCAM-1 were up-regulated by IL-6 and IFN-γ, and this regulation was attenuated by the SOCS1 transgene (Fig. 5C). Because both IL-6 and IFN-γ use JAK-STAT's signaling pathways to induce gene expression of proinflammatory molecules, we examined the signaling responses in aortic ECs. Primary cultured aortic ECs were treated with mouse IL-6 or IFN-γ, and activation of their downstream signaling mediators was determined by Western blotting with phospho-specific antibodies. IL-6 induced activation of JAK2 and STAT3, and the SOCS1 transgene reduced IL-6–activated JAK2-STAT3 signaling (Fig. 5D). IFN-γ induced activation of JAK1/2 and STAT1/3 in aortic ECs, and IFN-γ–activated signaling was attenuated by the SOCS1 transgene in EC (Fig. 5C). These effects of SOCS1 on JAK-STAT activation are consistent with the in vivo observations (Fig. 3). In summary, these results demonstrate that SOCS1 specifically inhibits proinflammatory cytokine responses in vascular EC.
To correlate the role of SOCS1 in proinflammatory cytokine signaling to its effect on vascular inflammation, we determined the effects of SOCS1 overexpression by using in vitro assays of leukocyte-endothelial adhesion and transmigration, 2 critical steps involved in inflammatory cell recruitment (30,31). In an adhesion assay using fluorescently pre-labeled mouse monocytes seeded onto confluent, primary cultured aortic EC, SOCS1 overexpression had no effect on monocyte attachment to resting ECs but prevented proinflammatory cytokine-mediated monocyte attachment compared with the WT group (Online Figs. S6A and S6B). Similarly, an inhibitory effect of the SOCS1 transgene on monocyte transmigration across cytokine-activated ECs was observed in a transendothelial migration assay (30) (Online Figs. S6C and S6D). These data suggest that SOCS1 specifically regulates proinflammatory cytokine-dependent functions in vascular ECs that are relevant for leukocyte trafficking and development of GA.
In the present study, we investigated the role of SOCS1 in GA by assessing clinical specimens and using a mouse aorta transplantation model, across the H-Y–dependent minor histocompatibility antigen barrier, which closely mimics GA progression as we have previously described (20). Here we demonstrate that SOCS1 is highly expressed in the luminal endothelium of nondiseased human coronary arteries, whereas its expression is decreased in ECs of chronically rejecting heart grafts and of those of atherosclerotic plaques. Such dramatic loss of expression accompanied by abundant immune cell infiltration in the pathologically remodeled vessel wall suggests a regulatory role for SOCS1 in GA pathogenesis. Clinical observation is supported by enhanced intimal expansion and increased leukocyte infiltration as well as EC dysfunction in SOCS1-deficient murine allografts. Because endothelial activation initiates leukocytic infiltration and neointima formation, a critical role for SOCS1, specifically in the endothelium of transplanted aortas, is implied. To this end, transgenic mice overexpressing SOCS1 in vascular endothelial cells were generated. Grafts from VESOCS1 mice exhibit dramatically decreased endothelial activation, leukocyte infiltration, and neointima formation. ECs from VESOCS1 grafts also blunt proinflammatory cytokine signaling and induction of adhesion molecules. This mirrors the finding of decreased endogenous SOCS1 and enhanced adhesion molecule expression in the endothelium of clinical GA specimens. Mechanistically, we show that SOCS1 overexpression in ECs significantly reduces IL-6– and IFN-γ–induced JAK2-STAT1/3 signaling and leukocyte-endothelial adhesion and transmigration. Furthermore, we show that SOCS1 directly binds to PECAM-1 and interrupts PECAM-1 stability. We conclude that endothelial SOSC1 prevents GA by down-regulating proinflammatory cytokine-induced EC activation and subsequent leukocyte infiltration.
Negative role of SOCS1 in endothelial inflammation
GA progression depends closely on inflammation, the important amplification mechanism of innate and adaptive immunity. Infiltration of immune cells into the vessel wall, including neutrophils, macrophages, and T cells, requires vascular endothelial activation. Characterized by the induction of adhesion molecules, endothelial activation results in adherence of immune cells from the circulation onto endothelium, followed by transendothelial migration into the artery wall. In the present study, clinical analysis suggests a special inhibitory role of SOCS1 on endothelial sensing and responses during vascular rejection. This observation supports our further investigations into the role of SOCS1 in maintaining EC homeostasis. In addition to preserving EC vessel function, we also find that SOCS1 overexpression in vascular EC inhibits the induction of endothelial adhesion molecules by proinflammatory cytokines. To this extent, our results reveal an important role for SOCS1 as a negative regulator of endothelial activation and subsequent arterial inflammation.
Dual regulatory mechanisms of endothelial adhesion molecule expression by SOCS1
Most studies of the SOCS family have focused on the immune system, including the regulation of cytokine production by immune cells in vascular diseases. To determine the effect of endothelial SOCS1 overexpression on inflammatory cytokine generation, we assessed the levels of IL-6 and IFN-γ transcripts typically expressed in GA lesions. No significant differences between the levels of WT and those of the VESOCS1 groups were detected. This suggests that the responses of SOCS1-expressing ECs are critical to understanding GA progression and need to be elucidated. Up-regulation of adhesion molecule expression generally involves transcriptional mechanisms. In addition to nuclear factor κB, STAT family members are major transcriptional factors controlling the expression of inducible adhesion molecules, such as ICAM-1 and VCAM-1, and are canonical signaling mediators for many inflammatory stimuli (32,33). Upon binding of different proinflammatory cytokines to their receptors, intracellular activation of the Janus kinases JAK1 and JAK2 is initiated, which leads to autophosphorylation and downstream STAT recruitment and phosphorylation. Activated STATs subsequently form homodimers and translocate to the nucleus to launch target gene expression. In our present study, inhibition of proinflammatory cytokine-induced JAK-STAT signaling by endothelial SOCS1 is identified as its molecular function on endothelial activation. Of note, IL-6 activates STAT1 while IFN-γ induces broader STAT activation in vascular ECs. More interestingly, our findings illustrate a novel mechanism of SOCS1's effects on endothelial adhesion molecule expression, in addition to transcription. We identify the fact that SOCS1 interacts directly with PECAM-1 to enhance its degradation. The association of PECAM-1 downregulation and its phosphorylation in activated EC has been previously reported (29) but little is known regarding the mechanisms. Here we demonstrate that SOCS1 is a potent negative mediator in this process.
Pathophysiological function of SOCS1 in pathological vascular remodeling
As a major member of the SOCS family, SOCS1 has received the most attention as a critical negative regulator of immune cells. In pathological vascular remodeling, infiltration of immune cells is the major source of proinflammatory cytokines that activate vascular ECs and vascular smooth muscle cells as well as other vessel wall cells. In response to proinflammatory cytokines, SOCS1 is induced in these inflammatory infiltrating cells and negatively regulates the generation of these cytokines (34). Based on both clinical and experimental data, the present study reveals a critical function of SOCS1 in vascular EC activation, dysfunction, and homeostasis. Our results shed light on the role of SOCS1 in vessel wall cells, which expands the understanding of the SOCS family beyond their traditional role in the immune system. To our surprise, endothelial SOCS1 expression is inhibited and not induced by the occurrence of both clinical GA and atherosclerosis. The fact that SOCS1 mRNA was not reduced in these specimens suggests that protein degradation may represent a novel regulation mechanism for SOCS1 expression. However, we cannot exclude the expression of endogenous SOCS1 transcripts in vessel wall cell types other than ECs, such as infiltrating leukocytes. Recent other studies report that SOCS1 expressed in smooth muscle cells is a key regulator of vascular cell responses in atherosclerosis (35). Our observation is also consistent for SOCS1 induction in the underlying neointima and shoulder regions of plaques beyond endothelium (Fig. 1A). Herewith, SOCS1 is an important regulator in maintaining normal function of the vasculature in pathological vascular remodeling. Modulation of endothelial SOCS1 expression and activity may represent a novel strategy for the treatment of cardiovascular diseases, such as GA and atherosclerosis.
Dr. Min is supported by National Institutes of Health grants R01 HL085789 and R01 HL109420. Dr. Yu is supported by Natural Science Foundation of China (81270357), Fundamental Research Funds for the Central Universities, and American Heart Association Scientist Development grant 12SDG9320033. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- endothelial cell
- graft arteriosclerosis
- intercellular adhesion molecule
- Janus kinase
- platelet/endothelial cell adhesion molecule
- suppressor of cytokine signaling 1
- signal transducers and activators of transcription
- vascular cell adhesion molecule
- Received April 11, 2013.
- Revision received August 5, 2013.
- Accepted August 6, 2013.
- American College of Cardiology Foundation
- He Y.,
- Zhang W.,
- Zhang R.,
- Zhang H.,
- Min W.
- Yu L.,
- Min W.,
- He Y.,
- et al.
- Naka T.,
- Matsumoto T.,
- Narazaki M.,
- et al.
- Starr R.,
- Metcalf D.,
- Elefanty A.G.,
- et al.
- Yu C.R.,
- Mahdi R.R.,
- Oh H.M.,
- et al.
- Balabanov R.,
- Strand K.,
- Kemper A.,
- Lee J.Y.,
- Popko B.
- Chong M.M.,
- Chen Y.,
- Darwiche R.,
- et al.
- Flodstrom-Tullberg M.,
- Yadav D.,
- Hagerkvist R.,
- et al.
- Roman-Gomez J.,
- Jimenez-Velasco A.,
- Castillejo J.A.,
- et al.
- Galm O.,
- Yoshikawa H.,
- Esteller M.,
- Osieka R.,
- Herman J.G.
- Yu L.,
- Qin L.,
- Zhang H.,
- et al.
- Plenz G.,
- Eschert H.,
- Erren M.,
- et al.
- Tellides G.,
- Pober J.S.
- Zehnder J.L.,
- Hirai K.,
- Shatsky M.,
- McGregor J.L.,
- Levitt L.J.,
- Leung L.L.
- Roebuck K.A.,
- Finnegan A.
- Ortiz-Munoz G.,
- Martin-Ventura J.L.,
- Hernandez-Vargas P.,
- et al.