Ablation of the gut microbiota alleviates high-methionine diet-induced hyperhomocysteinemia and glucose intolerance in mice.

A high-methionine (HM) diet leads to hyperhomocysteinemia (HHcy), while gastrointestinal tissue is an important site of net homocysteine (Hcy) production. However, the role of the gut microbiota in host HHcy remains obscure. This study aimed to determine whether gut microbiota ablation could alleviate host HHcy and glucose intolerance and reveal the underlying mechanism. The results showed that the HM diet-induced HHcy and glucose intolerance in mice, while antibiotic administration decreased the plasma level of Hcy and reversed glucose intolerance. HM diet increased intestinal epithelial homocysteine levels, while antibiotic treatment decreased intestinal epithelial homocysteine levels under the HM diet. Gut microbiota depletion had no effect on the gene expression and enzyme activity of CBS and BHMT in the livers of HM diet-fed mice. The HM diet altered the composition of the gut microbiota with marked increases in the abundances of Faecalibaculum and Dubosiella, which were also positively correlated with plasma Hcy concentrations. An in-depth analysis of the bacterial cysteine and methionine metabolism pathways showed that the abundances of two homocysteine biosynthesis-related KEGG orthologies (KOs) were markedly increased in the gut microbiota in HM diet-fed mice. Hcy was detected from Dubosiella newyorkensis-cultured supernatant by liquid chromatography-tandem mass spectrometry (LC‒MS) analysis. In conclusion, these findings suggested that the HM diet-induced HHcy and glucose intolerance in mice, by reshaping the composition of the gut microbiota, which might produce and secrete Hcy.

ADAMTS7-Mediated Complement Factor H Degradation Potentiates Complement Activation to Contributing to Renal Injuries.

BACKGROUND: The dysfunction of complement factor H (CFH), the main soluble complement negative regulator, potentiates various complement-induced renal injuries. However, insights into the underlying mechanism of CFH dysfunction remain limited. In this study, we investigated whether extracellular protease-mediated degradation accounts for CFH dysfunction in complement-mediated renal injuries. METHODS: An unbiased interactome of lupus mice kidneys identified CFH-binding protease. In vitro cleavage assay clarified CFH degradation. Pristane-induced SLE or renal ischemia-reperfusion (I/R) injury models were used in wild-type and ADAMTS7-/- mice. RESULTS: We identified the metalloprotease ADAMTS7 as a CFH-binding protein in lupus kidneys. Moreover, the upregulation of ADAMTS7 correlated with CFH reduction in both lupus mice and patients. Mechanistically, ADAMTS7 is directly bound to CFH complement control protein (CCP) 1-4 domain and degraded CCP 1-7 domain through multiple cleavages. In mice with lupus nephritis or renal I/R injury, ADAMTS7 deficiency alleviated complement activation and related renal pathologies, but without affecting complement-mediated bactericidal activity. Adeno-associated virus-mediated CFH silencing compromised these protective effects of ADAMTS7 knockout against complement-mediated renal injuries in vivo. CONCLUSION: ADAMTS7-mediated CFH degradation potentiates complement activation and related renal injuries. ADAMTS7 would be a promising anticomplement therapeutic target that does not increase bacterial infection risk.

Peptide Vaccine Against ADAMTS-7 Ameliorates Atherosclerosis and Postinjury Neointima Hyperplasia.

BACKGROUND: The metalloprotease ADAMTS-7 (a disintegrin and metalloproteinase with thrombospondin type 1 motif 7) is a novel locus associated with human coronary atherosclerosis. ADAMTS-7 deletion protects against atherosclerosis and vascular restenosis in rodents. METHODS: We designed 3 potential vaccines consisting of distinct B cell epitopic peptides derived from ADAMTS-7 and conjugated with the carrier protein KLH (keyhole limpet hemocyanin) as well as aluminum hydroxide as an adjuvant. Arterial ligation or wire injury was used to induce neointima in mice, whereas ApoE-/- and LDLR-/- (LDLR [low-density lipoprotein receptor]) mice fed a high-fat diet were applied to assess atherosclerosis. In addition, coronary stent implantation was performed on vaccine-immunized Bama miniature pigs, followed by optical coherence tomography to evaluate coronary intimal hyperplasia. RESULTS: A vaccine, ATS7vac, was screened out from 3 candidates to effectively inhibit intimal thickening in murine carotid artery ligation models after vaccination. As well, immunization with ATS7vac alleviated neointima formation in murine wire injury models and mitigated atherosclerotic lesions in both hyperlipidemic ApoE-/- and LDLR-/- mice without lowering lipid levels. Preclinically, ATS7vac markedly impeded intimal hyperplasia in swine stented coronary arteries, but without significant immune-related organ injuries. Mechanistically, ATS7vac vaccination produced specific antibodies against ADAMTS-7, which markedly repressed ADAMTS-7-mediated COMP (cartilage oligomeric matrix protein) and TSP-1 (thrombospondin-1) degradation and subsequently inhibited vascular smooth muscle cell migration but promoted re-endothelialization. CONCLUSIONS: ATS7vac is a novel atherosclerosis vaccine that also alleviates in-stent restenosis. The application of ATS7vac would be a complementary therapeutic avenue to the current lipid-lowering strategy for atherosclerotic disease.

Nidogen-2 is a Novel Endogenous Ligand of LGR4 to Inhibit Vascular Calcification.

BACKGROUND: Vascular calcification is closely related to the all-cause mortality of cardiovascular events. Basement membrane protein nidogen-2 is a key component of the vascular extracellular matrix microenvironment and we recently found it is pivotal for the maintenance of contractile phenotype in vascular smooth muscle cells (VSMCs). However, whether nidogen-2 is involved in VSMCs osteochondrogenic transition and vascular calcification remains unclear. METHODS: VSMCs was treated with high-phosphate to study VSMC calcification in vitro. Three different mice models (5/6 nephrectomy-induced chronic renal failure, cholecalciferol-overload, and periadventitially administered with CaCl2) were used to study vascular calcification in vivo. Membrane protein interactome, coimmunoprecipitation, flow cytometric binding assay, surface plasmon resonance, G protein signaling, VSMCs calcium assays were performed to clarify the phenotype and elucidate the molecular mechanisms. RESULTS: Nidogen-2 protein levels were significantly reduced in calcified VSMCs and aortas from mice in different vascular calcification model. Nidogen-2 deficiency exacerbated high-phosphate-induced VSMC calcification, whereas the addition of purified nidogen-2 protein markedly alleviated VSMC calcification in vitro. Nidogen-2-/- mice exhibited aggravated aorta calcification compared to wild-type (WT) mice in response to 5/6 nephrectomy, cholecalciferol-overload, and CaCl2 administration. Further unbiased coimmunoprecipitation and interactome analysis of purified nidogen-2 and membrane protein in VSMCs revealed that nidogen-2 directly binds to LGR4 (leucine-rich repeat G-protein-coupled receptor 4) with KD value 26.77 nM. LGR4 deficiency in VSMCs in vitro or in vivo abolished the protective effect of nidogen-2 on vascular calcification. Of interest, nidogen-2 biased activated LGR4-Gαq-PKCα (protein kinase Cα)-AMPKα1 (AMP-activated protein kinase α1) signaling to counteract VSMCs osteogenic transition and mineralization. CONCLUSIONS: Nidogen-2 is a novel endogenous ligand of LGR4 that biased activated Gαq- PKCα-AMPKα1 signaling and inhibited vascular calcification.

PHB2 Maintains the Contractile Phenotype of VSMCs by Counteracting PKM2 Splicing.

BACKGROUND: Phenotypic transition of vascular smooth muscle cells (VSMCs) accounts for the pathogenesis of a variety of vascular diseases during the early stage. Recent studies indicate the metabolic reprogramming may be involved in VSMC phenotypic transition. However, the definite molecules that link energy metabolism to distinct VSMC phenotype remain elusive. METHODS: A carotid artery injury model was used to study postinjury neointima formation as well as VSMC phenotypic transition in vivo. RNA-seq analysis, cell migration assay, collagen gel contraction assay, wire myography assay, immunoblotting, protein interactome analysis, co-immunoprecipitation, and mammalian 2-hybrid assay were performed to clarify the phenotype and elucidate the molecular mechanisms. RESULTS: We collected cell energy-regulating genes by using Gene Ontology annotation and applied RNA-Seq analysis of transforming growth factor-ß or platelet-derived growth factor BB stimulated VSMCs. Six candidate genes were overlapped from energy metabolism-related genes and genes reciprocally upregulated by transforming growth factor-ß and downregulated by platelet-derived growth factor BB. Among them, prohibitin 2 has been reported to regulate mitochondrial oxidative phosphorylation. Indeed, prohibitin 2-deficient VSMCs lost the contractile phenotype as evidenced by reduced contractile proteins. Consistently, Phb2SMCKO mice were more susceptible to postinjury VSMC proliferation and neointima formation compared with Phb2flox/flox mice. Further protein interactome analysis, co-immunoprecipitation, and mammalian 2-hybrid assay revealed that prohibitin 2, through its C-terminus, directly interacts with hnRNPA1, a key modulator of pyruvate kinase M1/2 (PKM) mRNA splicing that promotes PKM2 expression and glycolysis. Prohibitin 2 deficiency facilitated PKM1/2 mRNA splicing and reversion from PKM1 to PKM2, and enhanced glycolysis in VSMCs. Blocking prohibitin 2-hnRNPA1 interaction resulted in increased PKM2 expression, enhanced glycolysis, repressed contractile marker genes expression in VSMCs, as well as aggravated postinjury neointima formation in vivo. CONCLUSIONS: Prohibitin 2 maintains VSMC contractile phenotype by interacting with hnRNPA1 to counteract hnRNPA1-mediated PKM alternative splicing and glucose metabolic reprogramming.

MFN2 Prevents Neointimal Hyperplasia in Vein Grafts via Destabilizing PFK1.

BACKGROUND: Mechanical forces play crucial roles in neointimal hyperplasia after vein grafting; yet, our understanding of their influences on vascular smooth muscle cell (VSMC) activation remains rudimentary. METHODS: A cuff mouse model was used to study vein graft hyperplasia. Fifteen percent to 1 Hz uniaxial cyclic stretch (arterial strain), 5% to 1 Hz uniaxial cyclic stretch or a static condition (venous strain) were applied to the cultured VSMCs. Metabolomics analysis, cell proliferation and migration assays, immunoblotting, co-immunoprecipitation, mutagenesis, pull-down and surface plasmon resonance assays were employed to elucidate the potential molecular mechanisms. RESULTS: RNA-sequencing in vein grafts and the controls identified changes in metabolic pathways and downregulation of mitochondrial protein MFN2 (mitofusin 2) in the vein grafts. Exposure of VSMCs to 15% stretch resulted in MFN2 downregulation, mitochondrial fragmentation, metabolic shift from mitochondrial oxidative phosphorylation to glycolysis, and cell proliferation and migration, as compared with that to a static condition or 5% stretch. Metabolomics analysis indicated an increased generation of fructose 1,6-bisphosphate, an intermediate in the glycolytic pathway converted by PFK1 (phosphofructokinase 1) from fructose-6-phosphate, in cells exposed to 15% stretch. Mechanistic study revealed that MFN2 physically interacts through its C-terminus with PFK1. MFN2 knockdown or exposure of cells to 15% stretch promoted stabilization of PFK1, likely through interfering the association between PFK1 and the E3 ubiquitin ligase TRIM21 (E3 ubiquitin ligase tripartite motif [TRIM]-containing protein 21), thus, decreasing the ubiquitin-protease-dependent PFK1 degradation. In addition, study of mechanotransduction utilizing pharmaceutical inhibition indicated that the MFN2 downregulation by 15% stretch was dependent on inactivation of the SP1 (specificity protein 1) and activation of the JNK (c-Jun N-terminal kinase) and ROCK (Rho-associated protein kinase). Adenovirus-mediated MFN2 overexpression or pharmaceutical inhibition of PFK1 suppressed the 15% stretch-induced VSMC proliferation and migration and alleviated neointimal hyperplasia in vein grafts. CONCLUSIONS: MFN2 is a mechanoresponsive protein that interacts with PFK1 to mediate PFK1 degradation and therefore suppresses glycolysis in VSMCs.

Targeting macrophage TFEB-14-3-3 epsilon Interface by naringenin inhibits abdominal aortic aneurysm.

Abdominal aortic aneurysm (AAA) is a lethal cardiovascular disease, and there is no proven drug treatment for this condition. In this study, by using the Connectivity Map (CMap) approach, we explored naringenin, a naturally occurring citrus flavonoid, as a putative agent for inhibiting AAA. We then validated the prediction with two independent mouse models of AAA, calcium phosphate (CaPO4)-induced C57BL/6J mice and angiotensin II-infused ApoE-/- mice. Naringenin effectively blocked the formation of AAAs and the progression of established AAAs. Transcription factor EB (TFEB) is the master regulator of lysosome biogenesis. Intriguingly, the protective role of naringenin on AAA was abolished by macrophage-specific TFEB depletion in mice. Unbiased interactomics, combined with isothermal titration calorimetry (ITC) and cellular thermal shift assays (CETSAs), further revealed that naringenin is directly bound to 14-3-3 epsilon blocked the TFEB-14-3-3 epsilon interaction, and therefore promoted TFEB nuclear translocation and activation. On one hand, naringenin activated lysosome-dependent inhibition of the NLRP3 inflammasome and repressed aneurysmal inflammation. On the other hand, naringenin induced TFEB-dependent transcriptional activation of GATA3, IRF4, and STAT6 and therefore promoted reparative M2 macrophage polarization. In summary, naturally derived naringenin or macrophage TFEB activation shows promising efficacy for the treatment of AAA.

Unspliced XBP1 Counteracts ß-Catenin to Inhibit Vascular Calcification.

BACKGROUND: Vascular calcification is a prevalent complication in chronic kidney disease and contributes to increased cardiovascular morbidity and mortality. XBP1 (X-box binding protein 1), existing as the XBP1u (unspliced XBP1) and XBP1s (spliced XBP1) forms, is a key component of the endoplasmic reticulum stress involved in vascular diseases. However, whether XBP1u participates in the development of vascular calcification remains unclear. METHODS: We aim to investigate the role of XBP1u in vascular calcification. XBP1u protein levels were reduced in high phosphate-induced calcified vascular smooth muscle cells, calcified aortas from mice with adenine diet-induced chronic renal failure, and calcified radial arteries from patients with chronic renal failure. RESULTS: Inhibition of XBP1u rather than XBP1s upregulated in the expression of the osteogenic markers Runx2 (runt-related transcription factor 2) and Msx2 (msh homeobox 2), and exacerbated high phosphate-induced vascular smooth muscle cell calcification, as verified by calcium deposition and Alizarin red S staining. In contrast, XBP1u overexpression in high phosphate-induced vascular smooth muscle cells significantly inhibited osteogenic differentiation and calcification. Consistently, smooth muscle cell-specific XBP1 deficiency in mice markedly aggravated the adenine diet- and 5/6 nephrectomy-induced vascular calcification compared with that in the control littermates. Further interactome analysis revealed that XBP1u is bound directly to ß-catenin, a key regulator of vascular calcification, via amino acid (aa) 205-230 in its C-terminal degradation domain. XBP1u interacted with ß-catenin to promote its ubiquitin-proteasomal degradation and thus inhibited ß-catenin/TCF (T-cell factor)-mediated Runx2 and Msx2 transcription. Knockdown of ß-catenin abolished the effect of XBP1u deficiency on vascular smooth muscle cell calcification, suggesting a ß-catenin-mediated mechanism. Moreover, the degradation of ß-catenin promoted by XBP1u was independent of GSK-3ß (glycogen synthase kinase 3ß)-involved destruction complex. CONCLUSIONS: Our study identified XBP1u as a novel endogenous inhibitor of vascular calcification by counteracting ß-catenin and promoting its ubiquitin-proteasomal degradation, which represents a new regulatory pathway of ß-catenin and a promising target for vascular calcification treatment.

Nidogen-2 Maintains the Contractile Phenotype of Vascular Smooth Muscle Cells and Prevents Neointima Formation via Bridging Jagged1-Notch3 Signaling.

BACKGROUND: How the extracellular matrix (ECM) microenvironment modulates the contractile phenotype of vascular smooth muscle cells (VSMCs) and confers vascular homeostasis remains elusive. METHODS: To explore the key ECM proteins in the maintenance of the contractile phenotype of VSMCs, we applied protein-protein interaction network analysis to explore novel ECM proteins associated with the VSMC phenotype. By combining in vitro and in vivo genetic mice vascular injury models, we identified nidogen-2, a basement membrane glycoprotein, as a key ECM protein for maintenance of vascular smooth muscle cell identity. RESULTS: We collected a VSMC phenotype-related gene dataset by using Gene Ontology annotation combined with a literature search. A computational analysis of protein-protein interactions between ECM protein genes and the genes from the VSMC phenotype-related gene dataset revealed the candidate gene nidogen-2, a basement membrane glycoprotein involved in regulation of the VSMC phenotype. Indeed, nidogen-2-deficient VSMCs exhibited loss of contractile phenotype in vitro, and compared with wild-type mice, nidogen-2-/- mice showed aggravated post-wire injury neointima formation of carotid arteries. Further bioinformatics analysis, coimmunoprecipitation assays, and luciferase assays revealed that nidogen-2 specifically interacted with Jagged1, a conventional Notch ligand. Nidogen-2 maintained the VSMC contractile phenotype via Jagged1-Notch3 signaling but not Notch1 or Notch2 signaling. Nidogen-2 enhanced Jagged1 and Notch3 interaction and subsequent Notch3 activation. Reciprocally, Jagged1 and Notch3 interaction, signaling activation, and Jagged1-triggered VSMC differentiation were significantly repressed in nidogen-2-deficient VSMCs. In accordance, the suppressive effect of Jagged1 overexpression on neointima formation was attenuated in nidogen-2-/- mice compared with wild-type mice. CONCLUSIONS: Nidogen-2 maintains the contractile phenotype of VSMCs through Jagged1-Notch3 signaling in vitro and in vivo. Nidogen-2 is required for Jagged1-Notch3 signaling.

LOXL4 Abrogation Does Not Exaggerate Angiotensin II-Induced Thoracic or Abdominal Aortic Aneurysm in Mice.

It has been shown that thoracic aortic aneurysm and dissection (TAAD) could be a Mendelian trait caused by a single gene mutation. The LOX gene mutation leads to the development of human TAAD. The LOXL4 gene is a member of the lysyl oxidase gene family. We identified seven variants in the LOXL4 gene in 219 unrelated patients with TAAD by whole-exome sequencing (WES). To further investigate whether LOXL4 is a candidate causative gene for human TAAD, a LOXL4 knockout mouse was generated, and the mutant mice were treated by subcutaneous infusion of angiotensin II. We found that abrogation of LOXL4 did not induce a more severe thoracic or abdominal aortic aneurysm compared with the wild-type C57BL/6J mice. Our results suggest that LOXL4 may not play a major role in the development of angiotensin II-induced aortic aneurysm. The functional study using this animal model system is important for the evaluation of candidate genes of TAAD identified by WES.

Cartilage oligomeric matrix protein is an endogenous ß-arrestin-2-selective allosteric modulator of AT1 receptor counteracting vascular injury.

Compelling evidence has revealed that biased activation of G protein-coupled receptor (GPCR) signaling, including angiotensin II (AngII) receptor type 1 (AT1) signaling, plays pivotal roles in vascular homeostasis and injury, but whether a clinically relevant endogenous biased antagonism of AT1 signaling exists under physiological and pathophysiological conditions has not been clearly elucidated. Here, we show that an extracellular matrix protein, cartilage oligomeric matrix protein (COMP), acts as an endogenous allosteric biased modulator of the AT1 receptor and its deficiency is clinically associated with abdominal aortic aneurysm (AAA) development. COMP directly interacts with the extracellular N-terminus of the AT1 via its EGF domain and inhibits AT1-ß-arrestin-2 signaling, but not Gq or Gi signaling, in a selective manner through allosteric regulation of AT1 intracellular conformational states. COMP deficiency results in activation of AT1a-ß-arrestin-2 signaling and subsequent exclusive AAA formation in response to AngII infusion. AAAs in COMP-/- or ApoE-/- mice are rescued by AT1a or ß-arrestin-2 deficiency, or the application of a peptidomimetic mimicking the AT1-binding motif of COMP. Explorations of the endogenous biased antagonism of AT1 receptor or other GPCRs may reveal novel therapeutic strategies for cardiovascular diseases.

Extracellular matrix dynamics in vascular remodeling.

Vascular remodeling is the adaptive response to various physiological and pathophysiological alterations that are closely related to aging and vascular diseases. Understanding the mechanistic regulation of vascular remodeling may be favorable for discovering potential therapeutic targets and strategies. The extracellular matrix (ECM), including matrix proteins and their degradative metalloproteases, serves as the main component of the microenvironment and exhibits dynamic changes during vascular remodeling. This process involves mainly the altered composition of matrix proteins, metalloprotease-mediated degradation, posttranslational modification of ECM proteins, and altered topographical features of the ECM. To date, adequate studies have demonstrated that ECM dynamics also play a critical role in vascular remodeling in various diseases. Here, we review these related studies, summarize how ECM dynamics control vascular remodeling, and further indicate potential diagnostic biomarkers and therapeutic targets in the ECM for corresponding vascular diseases.

COMP-prohibitin 2 interaction maintains mitochondrial homeostasis and controls smooth muscle cell identity.

Vascular smooth muscle cells (VSMCs) are highly phenotypically plastic, and loss of the contractile phenotype in VSMCs has been recognized at the early onset of the pathology of a variety of vascular diseases. However, the endogenous regulatory mechanism to maintain contractile phenotype in VSMCs remains elusive. Moreover, little has been known about the role of the mitochondrial bioenergetics in terms of VSMC homeostasis. Herein, we asked if glycoprotein COMP (Cartilage oligomeric matrix protein) is involved in mitochondrial bioenergetics and therefore regulates VSMCs homeostasis. By using fluorescence assay, subcellular western blot and liquid chromatography tandem mass spectrometry analysis, we found that extracellular matrix protein COMP unexpectedly localized within mitochondria. Further mitochondrial transplantation revealed that both mitochondrial and non-mitochondrial COMP maintained VSMC identity. Moreover, microarray analysis revealed that COMP deficiency impaired mitochondrial oxidative phosphorylation in VSMCs. Further study confirmed that COMP deficiency caused mitochondrial oxidative phosphorylation dysfunction accompanied by morphological abnormality. Moreover, the interactome of mitochondrial COMP revealed that COMP interacted with prohibitin 2, and COMP-prohibitin 2 interaction maintained mitochondrial homeostasis. Additionally, disruption of COMP-prohibitin 2 interaction caused VSMC dedifferentiation in vitro and enhanced the neointima formation post rat carotid artery injury in vivo. In conclusion, COMP-prohibitin 2 interaction in mitochondria plays an important role in maintaining the contractile phenotype of VSMCs by regulating mitochondrial oxidative phosphorylation. Maintaining the homeostasis of mitochondrial respiration through COMP-prohibitin 2 interaction may shed light on prevention of vascular disease.

Cartilage oligomeric matrix protein prevents vascular aging and vascular smooth muscle cells senescence.

Aging-related vascular dysfunction contributes to cardiovascular morbidity and mortality. Cartilage oligomeric matrix protein (COMP), a vascular extracellular matrix protein, has been described as a negative regulatory factor for the vascular aging-related processes including atherosclerosis and vascular calcification. However, whether COMP is implicated in the process of vascular aging remains unclear. Here, we identified a novel function of COMP in preventing vascular aging and vascular smooth muscle cells (VSMCs) senescence. Firstly, vascular COMP expression was decreased in three different senescence-accelerated mouse models and was also declining with age. COMP(-/-) mice displayed elevated senescence-associated markers expression, including p53, p21 and p16, in the aortas compared with their wild type (WT) littermates. In accordance, COMP deficiency induced aging-related vascular dysfunction as evidenced by the significantly reduced phenylephrine-induced contraction and increased vascular stiffness as evaluated by pulse wave velocity. The aortic wall of COMP(-/-) mice was susceptible to senescence by displaying senescence-associated ß-galactosidase (SA ß-gal) activity induced by periadventitial application of CaCl2 to the abdominal aorta. In vitro, COMP knockdown by small interfering (si) RNA led to the elevation of p53, p21 and p16 as well as SA ß-gal activity in VSMCs after H2O2 stimulation. VSMCs isolated from COMP(-/-) mice showed elevated senescence-associated markers expression and supplement of COMP adenovirus to COMP-deficient VSMCs greatly rescued cellular senescence. Taken together, these findings revealed the essential role of COMP in retarding the development of vascular aging and VSMC senescence.

Shift of Macrophage Phenotype Due to Cartilage Oligomeric Matrix Protein Deficiency Drives Atherosclerotic Calcification.

RATIONALE: Intimal calcification is highly correlated with atherosclerotic plaque burden, but the underlying mechanism is poorly understood. We recently reported that cartilage oligomeric matrix protein (COMP), a component of vascular extracellular matrix, is an endogenous inhibitor of vascular smooth muscle cell calcification. OBJECTIVE: To investigate whether COMP affects atherosclerotic calcification. METHODS AND RESULTS: ApoE(-/-)COMP(-/-) mice fed with chow diet for 12 months manifested more extensive atherosclerotic calcification in the innominate arteries than did ApoE(-/-) mice. To investigate which origins of COMP contributed to atherosclerotic calcification, bone marrow transplantation was performed between ApoE(-/-) and ApoE(-/-)COMP(-/-) mice. Enhanced calcification was observed in mice transplanted with ApoE(-/-)COMP(-/-) bone marrow compared with mice transplanted with ApoE(-/-) bone marrow, indicating that bone marrow-derived COMP may play a critical role in atherosclerotic calcification. Furthermore, microarray profiling of wild-type and COMP(-/-) macrophages revealed that COMP-deficient macrophages exerted atherogenic and osteogenic characters. Integrin ß3 protein was attenuated in COMP(-/-) macrophages, and overexpression of integrin ß3 inhibited the shift of macrophage phenotypes by COMP deficiency. Furthermore, adeno-associated virus 2-integrin ß3 infection attenuated atherosclerotic calcification in ApoE(-/-)COMP(-/-) mice. Mechanistically, COMP bound directly to ß-tail domain of integrin ß3 via its C-terminus, and blocking of the COMP-integrin ß3 association by ß-tail domain mimicked the COMP deficiency-induced shift in macrophage phenotypes. Similar to COMP deficiency in mice, transduction of adeno-associated virus 2-ß-tail domain enhanced atherosclerotic calcification in ApoE(-/-) mice. CONCLUSIONS: These results reveal that COMP deficiency acted via integrin ß3 to drive macrophages toward the atherogenic and osteogenic phenotype and thereby aggravate atherosclerotic calcification.