ROS disrupt COP9 signalosome-mediated ABCA1 protection and trigger its ubiquitination and degradation by cullin3 inhibiting cholesterol efflux and promoting foam cell formation in response to GPCR agonists

Abstract

In this study, we investigated the intricate mechanisms underlying thrombin and Ang II-induced depletion of ABCA1. Under basal conditions, the COP9 signalosome interacts with ABCA1 as a whole complex rather than as individual subunits. In the presence of GPCR-agonists, thrombin or angiotensin II (Ang II), ABCA1 was phosphorylated and dissociated from COP9 signalosome, paving the way for its cullin3-mediated ubiquitination and degradation. Furthermore, forced expression of CSN5, the catalytic core subunit of the COP9 signalosome, inhibited thrombin or Ang II-induced ubiquitination and degradation of ABCA1, thereby restoring cholesterol efflux and suppressing foam cell formation. In addition, xanthine oxidase-dependent H₂O₂ production was required for both thrombin and Ang II-induced downregulation of ABCA1 levels and inhibition of cholesterol efflux promoting foam cell formation. Additionally, we identified the involvement of Gα12-Pyk2-Gab1-PKCθ signaling downstream to thrombin and Ang II receptors, namely, Par1 and AT1R in the modulation of these effects. Corroborating these observations, we found that while CSN5 and ABCA1 were found to be colocalized in human non-stenotic coronary artery sections and the aortic root cross sections of CD-fed ApoE^−/− mice, ABCA1 levels were diminished in advanced human atherosclerotic lesions and in the aortic root cross sections of WD-fed ApoE^−/− mice. Collectively, these findings reveal a novel role for xanthine oxidase-dependent H₂O₂ production in the GPCR agonists-induced destabilization of ABCA1 resulting in reduced cholesterol efflux and increased foam cell formation.

Graphical abstract

Highlights

• •COP9 signalosome complex by protecting ABCA1 ensures cholesterol efflux.
• •Phosphorylation of ABCA1 leads to its dissociation from COP9 signalosome.
• •Cullin3 by binding to phospho-ABCA1 causes its ubiquitination and degradation.
• •GPCR agonists trigger the above events and inhibits cholesterol efflux.
• •XO-derived H₂O₂ mediates GPCR agonist-induced ABCA1 destabilization.

Nonstandard abbreviations and acronyms

ABCA1

ATP binding cassette subfamily A member 1

ABCG1

ATP binding cassette subfamily G member 1

Ang II

Angiotensin II

AP

Allopurinol

ApoA-I

Apolipoprotein A-I

AT1R

Angiotensin II type 1 receptor

CD

Chow diet

CLRs

Cullin-RING ubiquitin ligases

CSN

COP9 signalosome

GPCR

G protein coupled receptor

HDL

High-density lipoprotein

LDL

Low-density lipoprotein

LDLR

LDL receptor

MASMC

Mouse aortic smooth muscle cell

NEDD8

Neutral precursor cell expressed developmentally downregulated 8

oxLDL

Oxidized low-density lipoprotein

Par 1

Proteinase-activated receptor 1

ROS

Reactive oxygen species

SMC

Smooth muscle cell

SR-B1

Scavenger receptor class B type 1

WD

Western diet

XO

Xanthine oxidase

1. Introduction

Atherosclerosis is a disease in which lipid-laden plaques build up in the arteries, narrowing their lumen and restricting blood flow, leading to the development of various cardiovascular pathologies [1,2]. Accumulation of cholesterol, specifically, low-density lipoprotein (LDL) cholesterol in macrophages and smooth muscle cells (SMCs) of the arterial walls contributes to the origin and progression of atherosclerotic plaques [3,4]. Thus, the dynamic regulation between cholesterol uptake by and removal from these cells is critical in the pathophysiology of atherosclerosis [4,5]. ATP-binding cassette (ABC) transporters play an important role in the removal of cholesterol from peripheral tissues, a cellular process known as cholesterol efflux, and its transport back to the liver where it is metabolized into bile acids [6,7]. ABC subfamily A member 1 (ABCA1), a member of the ABC transporters, protects against atherosclerosis by transporting cholesterol and phospholipids from peripheral tissues/cells to lipid-poor apolipoprotein A-I (apoA-I), the primary protein component of high-density lipoprotein (HDL), that carries cholesterol from plasma to the liver [8,9]. Dysregulation of ABCA1 results in the accumulation of cholesterol in macrophages and SMCs where it gets oxidized and transforms these cells into foam cells, a hallmark of early atherosclerosis [10,11]. Earlier reports have shown that while disruption of the ABCA1 gene enhances atherosclerosis [12], its stabilization restores cholesterol efflux and reduces atherosclerosis [13,14]. In line with these observations, our previous reports showed that thrombin, a serine protease and a GPCR agonist, by inducing ABCA1 degradation via cullin3-mediated ubiquitination inhibits cholesterol efflux and enhances foam cell formation leading to atherogenesis [15]. We also reported that COP9 signalosome subunit 3 (CSN3) protects ABCA1 from cullin3-mediated ubiquitination and degradation, restores cholesterol efflux and inhibits foam cell formation and atherosclerosis [16].

The COP9 signalosome is a multi-subunit protein complex that acts as a crucial regulator of cullin-RING ubiquitin ligases (CRLs) [17]. CRLs are a large family of ubiquitin ligases that control various cellular processes by ubiquitination and proteasomal degradation of specific proteins [18]. COP9 signalosome inhibits CRL activity by removing a ubiquitin-like protein known as NEDD8 from cullins, a process called deneddylation [19]. Thus, while addition of NEDD8 (neddylation) activates CRLs, deneddylation inactivates CRLs [19]. The COP9 signalosome is typically composed of nine subunits in higher eukaryotes (CSN1-CSN6, CSN7a/b & CSN8), and all CSN subunits are essential for the full function of the COP9 signalosome [19]. In fact, earlier studies have reported a role for different CSN subunits in the pathophysiology of atherosclerosis [[20], [21], [22]]. Whereas some studies have shown that CSN5, the catalytic core subunit of the COP9 signalosome [19,20], was upregulated in human atherosclerotic arteries, and its deletion in mice exacerbates atherosclerotic lesion formation [20,21], other reports showed that CSN2 and CSN5 interact with each other, and forced expression of CSN2 decreases the ubiquitination and degradation of ABCA1 [22]. In line with these studies, we reported that thrombin or western diet (WD) via triggering ABCA1 phosphorylation induces its disassociation from CSN3 and association with cullin3 leading to its ubiquitination and degradation [16]. We also reported that forced expression of CSN3 protects ABCA1 from thrombin or WD-induced ubiquitination and degradation and restores cholesterol efflux [16]. Although the above studies from others and our laboratories have shown a role for different CSN subunits in atherogenesis, it is unclear whether the whole COP9 signalosome is involved in this phenomenon. In the present study, we, for the first time, report that the COP9 signalosome as a whole complex, rather than individual subunits, mediates ABCA1 stabilization. Mechanistically, we show that xanthine oxidase-dependent H₂O₂ production triggers ABCA1 depletion via inducing its phosphorylation and dissociation from COP9 signalosome, thereby facilitating cullin3-mediated ubiquitination and degradation leading to decreased cholesterol efflux and enhanced foam cell formation in response to GPCR agonists.

2. Results

2.1. Thrombin triggers dissociation of COP9 signalosome from ABCA1 and its association with cullin3

We previously reported that thrombin by triggering dissociation of CSN3 from ABCA1 and its association with cullin3 induces ubiquitin-dependent proteasomal degradation of ABCA1, thereby inhibiting cholesterol efflux and enhancing foam cell formation leading to atherosclerosis [15,16]. The COP9 signalosome is a multi-subunit protein complex, and it acts as a crucial regulator of cullin-RING ubiquitin ligases and protects several proteins by preventing their proteasomal degradation [[17], [18], [19]]. Therefore, first, we asked whether the COP9 signalosome mediates ABCA1 stabilization as a whole complex or as individual subunits. To this end, we treated RAW264.7 cells and mouse aortic smooth muscle cells (MASMCs) with and without thrombin for different periods, protein extracts were prepared, immunoprecipitated with anti-ABCA1 antibody, and the immunocomplexes were analyzed by Western blotting for different subunits of COP9 signalosome. We found that ABCA1 exists as a complex with CSN1, CSN2, CSN3, CSN4, and CSN5 under basal conditions, and in response to thrombin, it gets dissociated from all these CSN subunits and associates with cullin3 both in RAW264.7 cells and MASMCs (Fig. 1A and B). In addition, thrombin induces ABCA1 phosphorylation in a time-dependent manner that correlates with its dissociation from CSN1-5 and association with cullin3, both in RAW264.7 cells and MASMCs (Fig. 1A and B). These results indicate that COP9 signalosome mediates ABCA1 stabilization as a whole complex but not as individual subunits. Since CSN5 is the catalytic core subunit of the COP9 signalosome and maintains the stability of several proteins [19,20], we focused hereon on CSN5 to explore the mechanisms underlying COP9 signalosome-mediated ABCA1 stabilization. First, we found that thrombin, while depleting ABCA1 levels, did not affect the steady state levels of CSN5 and cullin3, both in RAW264.7 cells and MASMCs (Fig. 1C and D). In addition, thrombin did not affect the steady state levels of other transporters such as ABCG1 and SR-B1 that play a role in of cholesterol efflux either in RAW264.7 cells or MASMCs (Fig. 1C and D). These results indicate that the effect of thrombin on the downregulation of cholesterol transporters is selective to ABCA1. Regarding molecular interactions, only ABCA1 but not cullin3 was found to be coimmunoprecipitated with CSN5 in RAW264.7 cells and MASMCs under basal conditions (Fig. 1E and F). On the other hand, thrombin induces time-dependent dissociation of ABCA1 from CSN5 in RAW264.7 cells and MASMCs (Fig. 1E and F). The interactions between ABCA1, CSN5 and cullin3 were further confirmed by in situ proximity ligation assay (PLA). PLA results showed that ABCA1 and CSN5 exists as a complex in vehicle treated RAW264.7 cells and MASMCs and thrombin disrupts their interactions in both the cell types (Fig. 1G and H). On the other hand, ABCA1 while having no interactions with cullin3 under basal conditions forms complex with cullin3 in response to thrombin in both RAW264.7 cells and MASMCs (Fig. 1G and H). These results indicate that thrombin modulates ABCA1 interactions with COP9 signalosome and cullin3, in a reciprocal manner, facilitating its downregulation via cullin3-mediated ubiquitination and proteasomal degradation.

Fig. 1.

Thrombin triggers ABCA1 dissociation from COP9 signalosome and association with cullin3. A and B, Quiescent RAW264.7 cells (A), or MASMCs (B) were treated with and without thrombin (0.5 U/ml) for the indicated time periods, cell extracts were prepared, immunoprecipitated with anti-ABCA1 antibody and the immunocomplexes were analyzed by Western blotting for the indicated proteins using their specific antibodies and the input protein was immunoblotted for β-actin. C and D, All the conditions were the same as in panels A and B except that protein extracts were analyzed by Western blotting for the indicated proteins using their specific antibodies and quantified. E and F, All the conditions were the same as in panel A and B except that cell extracts were immunoprecipitated with anti-CSN5 antibody and the immunocomplexes were analyzed by Western blotting for the indicated proteins using their specific antibodies and the input protein was immunoblotted for β-actin. G and H, Quiescent RAW264.7 cells (G), or MASMCs (H) were treated with and without thrombin (0.5 U/ml) for 4 h and subjected to in situ PLA as described in Materials and Methods section. The bar graphs represent Mean ± SD values of three biologically independent experiments. ∗p < 0.01 versus control. Scale bar in panels G and H is 10 μm.

2.2. Overexpression of CSN5 restores ABCA1 levels and cholesterol efflux from inhibition by thrombin and reduces foam cell formation

COP9 signalosome maintains the stability of several proteins via deneddylation of CRLs [19,20]. Therefore, we asked whether its overexpression rescues ABCA1 from thrombin-induced downregulation. To this end, we overexpressed CSN5 in RAW264.7 cells and MASMCs, treated with and without thrombin for 4 h, protein extracts were prepared, immunoprecipitated with anti-ABCA1 antibody, and analyzed by Western blotting for ABCA1 phosphorylation, and its dissociation from CSN5 and association with cullin3. The Co-IP results showed that forced expression of CSN5 restores ABCA1-CSN5 interactions from thrombin-induced disruption in RAW264.7 cells and MASMCs (Fig. 2A and B). Additionally, overexpression of CSN5 prevented thrombin-induced ABCA1 phosphorylation, and its association with cullin3 leading to its stabilization both in RAW264.7 cells and MASMCs (Fig. 2A and B). The overexpression of CSN5 had no effect on the steady state levels of the other subunits of COP9 signalosome such as CSN1 and CSN2 either in control or in response to thrombin (Fig. 2A and B). Towards functional aspects, forced expression of CSN5 restored cholesterol efflux from inhibition by thrombin and prevented foam cell formation both in RAW264.7 cells and MASMCs (Fig. 2C–F). Together, these findings demonstrate that COP9 signalosome protects ABCA1 by preventing its phosphorylation and association with cullin3, thus promoting cholesterol efflux and inhibiting foam cell formation.

Fig. 2.

Forced expression of CSN5 restores ABCA1 levels and cholesterol efflux from thrombin-induced downregulation and diminishes foam cell formation. A and B, RAW264.7 cells (A) and MASMCs (B) were transfected with pCMV or pCMV-CSN5 plasmid, treated with and without thrombin (0.5 U/ml) for 4 h, cell extracts were prepared, immunoprecipitated with anti-ABCA1 antibodies and the immunocomplexes were immunoblotted for the indicated proteins using their specific antibodies. The input protein was immunoblotted for CSN1, CSN2, CSN5, Myc-tag and β-actin. C–F, All the conditions were the same as in panel A and B except that cells were assayed for cholesterol efflux (C and D) or foam cell formation (E and F). The bar graphs represent Mean ± SD values of three biologically independent experiments. ∗p < 0.01 versus pCMV + vehicle; ∗∗p < 0.01 versus pCMV + thrombin. Scale bar in panels E and F is 20 μm.

2.3. Role of Par1-Gα12-Pyk2-Gab1-PKCθ signaling axis in thrombin-induced ABCA1 dissociation from CSN5 and association with cullin3 leading to its downregulation

Previously, we have shown that the Par1-Gα12-Pyk2-Gab1-PKCθ signaling axis mediates thrombin-induced disruption of ABCA1 and CSN3 interactions and leads to cullin3-mediated ABCA1 ubiquitination and degradation [16]. Therefore, we have examined whether the same signaling axis operates in thrombin-induced dissociation of CSN5 from ABCA1. To this end, we transfected RAW264.7 cells and MASMCs with control siRNA or siPar1 or siGα12 or siPyk2 or siGab1 or siPKCθ siRNAs (100 pmol/ml), treated with and without thrombin for 4 h, protein extracts were prepared, immunoprecipitated with anti-ABCA1 antibody and the immunocomplexes were analyzed by Western blotting for ABCA1 phosphorylation, and its dissociation from CSN5 and association with cullin3. Depletion of Par1, Gα12, Pyk2, Gab1 or PKCθ levels by their siRNAs suppressed thrombin-induced ABCA1 phosphorylation and its dissociation from CSN5. In addition, siRNA-mediated downregulation of Par1, Gα12, Pyk2, Gab1 or PKCθ levels blunted ABCA1 interaction with cullin3 and restored ABCA1 levels from thrombin-induced downregulation both in RAW264.7 cells and MASMCs (Fig. 3A and B). These findings imply that Par1-Gα12-Pyk2-Gab1-PKCθ signaling axis mediates disruption of COP9 signalosome interaction with ABCA1 leading to its phosphorylation and association with cullin3 facilitating its ubiquitination and degradation in response to thrombin.

Fig. 3.

Role of Par1-Gα12-Pyk2-Gab1-PKCθ signaling axis in thrombin-induced downregulation of ABCA1. A and B, RAW264.7 cells (A) and MASMCs (B) that were transfected with the indicated siRNAs were quiesced, treated with and without thrombin (0.5 U/ml) for 4 h, cell extracts prepared, immunoprecipitated with anti-ABCA1 antibody and the immunocomplexes were analyzed by Western blotting for the indicated proteins using their specific antibodies and quantified. The bar graphs represent Mean ± SD values of three biologically independent experiments. ∗p < 0.01 versus siControl + vehicle; ^#p < 0.01 versus siControl + thrombin.

2.4. ROS play a role in thrombin-induced ABCA1 downregulation leading to reduced cholesterol efflux and increased foam cell formation

Reactive oxygen species (ROS) play a key role in upregulation of ubiquitin-proteasomal activity and promote ubiquitination of several proteins [23,24]. Therefore, to investigate the far upstream mechanisms of thrombin-induced ABCA1 destabilization, we examined the role of ROS. First, we tested whether thrombin influences ROS production. We found that thrombin induces H₂O₂ production both in RAW264.7 cells and MASMCs (Fig. 4A and B). Previous studies have reported that allopurinol, a specific inhibitor of xanthine oxidase [25], by acting as a potential ROS scavenger inhibits the progression of atherosclerosis [26]. Based on this clue, we tested the effect of allopurinol on thrombin-induced H₂O₂ production. Allopurinol blunted thrombin-induced H₂O₂ production both in RAW264.7 cells and MASMCs (Fig. 4A and B). These observations infer that thrombin induces H₂O₂ production in a xanthine oxidase-dependent manner. To explore the underlying mechanisms, we studied the time course effect of thrombin on xanthine oxidase expression and observed that thrombin induces xanthine oxidase in a time-dependent manner both in RAW264.7 cells and MASMCs (Fig. 4C and D). Next, we examined the role of xanthine oxidase in thrombin-induced suppression of cholesterol efflux and foam cell formation. Inhibition of xanthine oxidase by allopurinol restored cholesterol efflux from thrombin-induced inhibition and reduced foam cell formation both in RAW264.7 cells and MASMCs (Fig. 4E–H). Corroborating these observations, we found that allopurinol inhibits thrombin-induced ABCA1 phosphorylation, its dissociation from CSN5, and association with cullin3, thereby preventing ABCA1 ubiquitination and degradation both in RAW264.7 cells and MASMCs (Fig. 4I and J). Furthermore, allopurinol blocked thrombin-induced PKCθ phosphorylation in these cells (Fig. 4K and L). These findings indicate that xanthine oxidase, via H₂O₂ production, triggers COP9 signalosome dissociation from ABCA1 and its association with cullin3 leading to its ubiquitination and degradation and resulting in decreased cholesterol efflux and increased foam cell formation. Regarding the spatial position of xanthine oxidase in the Par1-Gα12-Pyk2-Gab1-PKCθ signaling, we found that while siPar1 or siGα12 were inhibiting thrombin-induced xanthine oxidase expression both in RAW264.7 cells and MASMCs, siPyk2, siGab1 or siPKCθ had no effect (Fig. 4M and N). Based on the findings, it is likely that xanthine oxidase acts downstream to Gα12 and upstream to Pyk2 in the Par1-Gα12-Pyk2-Gab1-PKCθ signaling activation by thrombin in the mediation of ABCA1 destabilization.

Fig. 4.

ROS play a role in thrombin-induced ABCA1 downregulation leading to inhibition of cholesterol efflux and increased foam cell formation. A and B, Quiescent RAW264.7 cells (A) and MASMCs (B) were treated with and without thrombin (0.5 U/ml) in the presence or absence of allopurinol (50 μM) for 4 h, and H₂O₂ production was measured using Amplex™ Red Hydrogen Peroxide/Peroxidase Assay Kit. C and D, Quiescent RAW264.7 cells (C), or MASMCs (D) were treated with and without thrombin (0.5 U/ml) for the indicated time periods, cell extracts were prepared and analyzed by Western blotting for xanthine oxidase and β-actin levels using their specific antibodies and quantified. E-H. All the conditions were the same as in panels A and B except that cells were assayed for cholesterol efflux (E and F) or foam cell formation (G and H). I and J, Quiescent RAW264.7 cells (I) and MASMCs (J) were treated with and without thrombin in the presence and absence of allopurinol for 4 h and cell extracts were prepared and immunoprecipitated with anti-ABCA1 antibody and the immunocomplexes were analyzed by Western blotting for the indicated proteins. The input protein was immunoblotted for β-actin. K and L, Quiescent RAW264.7 cells (K) and MASMCs (L) were treated with and without thrombin (0.5 U/ml) in the presence or absence of allopurinol for 4 h, cell extracts prepared and analyzed by Western blotting for phospho and total PKCθ levels. M and N, RAW264.7 cells (M) and MASMCs (N) that were transfected with the indicated siRNAs and quiesced were treated with and without thrombin for 4 h and cell extracts were prepared and analyzed by Western blotting for xanthine oxidase using its specific antibody and the blots were normalized for β-actin. The bar graphs represent Mean ± SD values of three biologically independent experiments. ∗p < 0.01 versus control or siControl; ∗∗p < 0.01 versus thrombin or siControl + thrombin. Scale bar in panels G and H is 20 μm. AP, allopurinol. XO, xanthine oxidase.

2.5. Lack of the effect of uric acid in thrombin-induced ABCA1 destabilization, cholesterol efflux and foam cell formation

Xanthine oxidase converts xanthine to uric acid and produces H₂O₂ as a major byproduct [27]. Therefore, next, we tested the role of H₂O₂ and uric acid in ABCA1 destabilization, cholesterol efflux, and foam cell formation. Intriguingly, H₂O₂ but not uric acid induced ABCA1 phosphorylation, its dissociation from CSN5 and association with cullin3 both in RAW264.7 cells and MASMCs (Fig. 5A and B). Corroborating these observations, while H₂O₂ induced substantial ubiquitination and degradation of ABCA1, uric acid had no effect on these events both in RAW264.7 cells and MASMCs (Fig. 5A and B). Consistent with these observations, H₂O₂ but not uric acid inhibited cholesterol efflux and induced foam cell formation both in RAW264.7 cells and MASMCs (Fig. 5C–F). Together, these findings indicate that downstream to thrombin, xanthine oxidase-dependent H₂O₂ production plays a role in COP9 signalosome dissociation from ABCA1 and its association with cullin3 leading to its ubiquitination and degradation and thereby resulting in reduced cholesterol efflux and increased foam cell formation.

Fig. 5.

H₂O₂ but not uric acid plays a role in thrombin-induced downregulation of ABCA1 and cholesterol efflux and increased foam cell formation. A and B, Quiesced RAW264.7 cells (A) and MASMCs (B) were treated with and without H₂O₂ (200 μM) or uric acid (100 μM) for 4 h, cell extracts prepared, immunoprecipitated with anti-ABCA1 antibody and the immunocomplexes were analyzed by Western blotting for the indicated proteins using their specific antibodies. The input protein was immunoblotted for β-actin. C–F. All the conditions were the same as in panels A and B except that cells were assayed for cholesterol efflux (C and D) or foam cell formation (E and F). The bar graphs represent Mean ± SD values of three biologically independent experiments. ∗p < 0.01 versus vehicle; ∗∗p < 0.01 versus H₂O₂. Scale bar in panels E and F is 20 μm. UA, uric acid.

2.6. Ang II via downregulating ABCA1 inhibits cholesterol efflux and increases foam cell formation

GPCRs are a large family of membrane receptors that play a crucial role in various cellular signaling pathways in response to their agonists [28,29]. To find whether ABCA1 destabilization is a unique response to thrombin or is a common response to other GPCR agonists, we next tested the effect of angiotensin II (Ang II), another potent GPCR agonist which has profound effects in the development of hypertension [30]. First, we studied the time course effect of Ang II on the expression levels of ABCA1, ABCG1, and SR-B1, the key molecules involved in reverse cholesterol transport [31]. Ang II reduced ABCA1 levels while having little or no impact on ABCG1 and SR-B1 levels in RAW264.7 cells (Fig. 6A). Besides, Ang II had no effect on the steady state levels of CSN5 and cullin3 in this cell type (Fig. 6A). Subsequently, the co-IP experiments demonstrated that ABCA1 exists in complex with CSN1, CSN2, CSN3, CSN4, and CSN5 under basal conditions and upon treatment with Ang II, it dissociates from all these COP9 subunits and associates with cullin3 (Fig. 6B). Moreover, we observed that Ang II induces ABCA1 phosphorylation that correlates with its dissociation from CSN1-5 and association with cullin3 leading to its ubiquitination and degradation (Fig. 6B and C). For rigor, we also examined the interactions between ABCA1, CSN5 and cullin3 using in situ PLA. We found that while ABCA1 and CSN5 exist as a complex in RAW264.7 cells under basal conditions they were dissociated in response to Ang II (Fig. 6D). In contrast, ABCA1 while having no interaction with cullin3 in vehicle-treated RAW264.7 cells showed an interaction with cullin3 in response to Ang II (Fig. 6D). Like thrombin, Ang II also induced xanthine oxidase expression and ROS production (Fig. 6E and F). In addition, allopurinol mitigated Ang II-induced ROS production (Fig. 6E). In line with these observations, allopurinol also blocked Ang II-induced ABCA1 phosphorylation, its dissociation from CSN1-5, and association with cullin3, thereby inhibiting ABCA1 ubiquitination and its degradation (Fig. 6B and C). Furthermore, allopurinol restored cholesterol efflux from inhibition by Ang II and decreased foam cell formation (Fig. 6G and H). In addition, forced expression of CSN5 restored ABCA1-CSN5 interactions from disruption by Ang II in RAW264.7 cells (Fig. 6I). Additionally, CSN5 overexpression prevented Ang II-induced ABCA1 phosphorylation and its association with cullin3, thereby restoring ABCA1 levels in these cells (Fig. 6I). The overexpression of CSN5 had no effect on the steady state levels of the other subunits of COP9 signalosome such as CSN1 and CSN2 either in control or in response to Ang II (Fig. 6I).Consistent with these observations, forced expression of CSN5 also restored cholesterol efflux from inhibition by Ang II and prevented foam cell formation (Fig. 6J and K). Together, these findings demonstrate that COP9 signalosome protects ABCA1 from Ang II-induced downregulation by preventing its phosphorylation and association with cullin3, promoting cholesterol efflux and inhibiting foam cell formation. These results also imply that, in addition to thrombin, other GPCR agonists such as Ang II can also disrupt ABCA1 stabilization resulting in decreased cholesterol efflux and increased foam cell formation.

Fig. 6.

Ang II inhibits cholesterol efflux and increases foam cell formation by downregulation of ABCA1 levels. A, Quiescent RAW264.7 cells were treated with and without Ang II (1 μM) for the indicated time periods, cell extracts were prepared and analyzed by Western blotting for ABCA1, ABCG1, SR-B1, CSN5, cullin3, and β-actin levels using their specific antibodies and quantified. B, Quiescent RAW264.7 cells were treated with and without Ang II (1 μM) in the presence or absence of allopurinol (50 μM) for indicated time-periods, cell extracts prepared, immunoprecipitated with anti-ABCA1 antibody and the immunocomplexes were analyzed by Western blotting for the indicated proteins using their specific antibodies. The input protein was immunoblotted for β-actin. C, Quiescent RAW264.7 cells were treated with and without Ang II in the presence or absence of allopurinol for 4 h, cell extracts prepared, immunoprecipitated with anti-ABCA1 antibody and the immunocomplexes were analyzed by Western blotting for the indicated proteins using their specific antibodies. The input protein was immunoblotted for β-actin. D, All the conditions were same as in panel C except that the cells were subjected to in situ PLA. E, Quiescent RAW264.7 cells were treated with and without Ang II (1 μM) in the presence or absence of allopurinol (50 μM) for 4 h and assayed for H₂O₂ production using Amplex™ Red Hydrogen Peroxide/Peroxidase Assay Kit. F, All the conditions were same as in panel A except that cell extracts were analyzed by Western blotting for xanthine oxidase, and β-actin levels using their specific antibodies and quantified. G and H, All the conditions were the same as in panel E except that the cells were assayed for cholesterol efflux (G) or foam cell formation (H). I, RAW264.7 cells were transfected with pCMV or pCMV-CSN5 plasmid, treated with and without Ang II (1 μM) for 4 h, cell extracts were prepared, immunoprecipitated with anti-ABCA1 antibodies and the immunocomplexes were immunoblotted for the indicated proteins using their specific antibodies. The input protein was immunoblotted for CSN1, CSN2, CSN5, Myc-tag and β-actin. J and K, All the conditions were the same as in panel I except that cells were assayed for cholesterol efflux (J) or foam cell formation (K). The bar graphs represent Mean ± SD values of three biologically independent experiments. ∗p < 0.01 versus control; ∗∗p < 0.01 versus Ang II. Scale bar in panel D is 10 μm and panels H and K is 20 μm.

2.7. A role for AT1R-Gα12-Pyk2-Gab1-PKCθ signaling axis in Ang II-induced ABCA1 dissociation from CSN5 and association of cullin3 leading to its ubiquitination and degradation

Ang II exerts its cellular effects majorly through angiotensin II type 1 receptor (AT1R), which belongs to the G-protein coupled receptor family [32]. In the previous studies, we demonstrated that the Par1-Gα12-Pyk2-Gab1-PKCθ signaling axis mediates thrombin-induced disruption of ABCA1-CSN3 interaction resulting in cullin3-mediated ubiquitination and degradation of ABCA1 [16]. Moreover, our current findings indicate that the same signaling axis is involved in the disruption of CSN5-ABCA1 interactions in response to thrombin (Fig. 3). Based on these clues, we investigated whether a similar signaling axis is involved in the Ang II-induced dissociation of ABCA1 from CSN5. First, we examined the time course effect of Ang II on the expression levels of its receptor, AT1R, and downstream signaling molecules [32]. Ang II was found to increase AT1R at protein levels in a time-dependent manner in RAW264.7 cells (Fig. 7A). Additionally, the co-IP results revealed that Ang II, while not affecting the interaction of Gαq, Gα11, and Gα13, induced Gα12 dissociation from AT1R (Fig. 7B). Furthermore, Ang II stimulated the phosphorylation of Pyk2, Gab1, and PKCθ in a time-dependent manner in RAW264.7 cells (Fig. 7C). These findings suggest that Ang II activates AT1R-Gα12-Pyk2-Gab1-PKCθ signaling in RAW264.7 cells. Subsequently, we found that pharmacological inhibition of AT1R using losartan [33], or siRNA-mediated depletion of Gα12, Pyk2, Gab1, or PKCθ levels inhibited Ang II-induced ABCA1 phosphorylation and its dissociation from CSN5 and association with cullin3 restoring its levels (Fig. 7D–F). These results suggest that the AT1R-Gα12-Pyk2-Gab1-PKCθ signaling axis mediates ABCA1 phosphorylation leading to dissociation of COP9 signalosome and association with cullin3 resulting in its ubiquitination and degradation in response to Ang II. To address the spatial position of xanthine oxidase in the AT1R-Gα12-Pyk2-Gab1-PKCθ signaling, we tested the role of these signaling molecules in Ang II-induced xanthine oxidase expression. While inhibition of AT1R by losartan or downregulation of Gα12 levels by its siRNA negated Ang II-induced xanthine oxidase expression, depletion of Pyk2, Gab1 or PKCθ levels by their respective siRNAs had no effect (Fig. 7G). In view of these findings, it is likely that xanthine oxidase acts downstream to Gα12 and upstream to Pyk2 in the AT1R-Gα12-Pyk2-Gab1-PKCθ signaling activation by Ang II in the mediation of ABCA1 destabilization.

Fig. 7.

A role for AT1R-Gα12-Pyk2-Gab1-PKCθ signaling axis in Ang II-induced ABCA1 dissociation from CSN5 and association with cullin3 leading to its ubiquitination and degradation. A, Quiescent RAW264.7 cells were treated with and without Ang II (1 μM) for the indicated time periods, cell extracts were prepared and analyzed by Western blotting for AT1R and β-actin levels using their specific antibodies and quantified. B, All the conditions are same as in panel A except that cell lysates were immunoprecipitated with anti-AT1R antibody and the immunocomplexes were analyzed by Western blotting for the indicated proteins. The input protein was immunoblotted for β-actin. C, Cell lysates in panel A were analyzed by Western blotting for the indicated proteins using their specific antibodies and quantified. D, Quiescent RAW264.7 cells were treated with and without Ang II (1 μM) in the presence or absence of losartan (10 μM) for 4 h, cell extracts prepared, immunoprecipitated with anti-ABCA1 antibody and the immunocomplexes were analyzed by Western blotting for the indicated proteins using their specific antibodies. The input protein was immunoblotted for β-actin. E and F, RAW264.7 cells that were transfected with the indicated siRNAs were quiesced, treated with and without Ang II (1 μM) for 4 h, cell extracts prepared, immunoprecipitated with anti-ABCA1 antibody and the immunocomplexes were analyzed by Western blotting for the indicated proteins and quantified. G, The cell extracts from panels D, E and F were analyzed by Western blotting for xanthine oxidase using its specific antibody and the blots were normalized for β-actin. The bar graphs represent Mean ± SD values of three biologically independent experiments. ∗p < 0.01 versus control or siControl; ^#p < 0.01 versus Ang II or siControl + Ang II.

2.8. CSN5-ABCA1 interactions were disrupted in mouse and human atherosclerotic lesions

To extrapolate our in vitro findings to mouse atherosclerotic model, we investigated the interactions of ABCA1-CSN5-cullin3 in the aorta of CD or 16 weeks of WD-fed ApoE^−/− mice. First, we studied the effect of WD on the expression levels of ABCA1, ABCG1, and SR-B1 in the mouse aorta. WD, while downregulating ABCA1 levels, had no effect on ABCG1 and SR-B1 levels in the mouse aorta (Fig. 8A). WD did not affect the steady state levels of CSN5 and cullin3 (Fig. 8A). Second, the co-IP experiments have shown that ABCA1 exists as a complex with CSN1, CSN2, CSN3, CSN4, and CSN5 in CD-fed mouse aorta, and upon WD-feeding, it dissociated from all these COP9 signalosome subunits and associated with cullin3 (Fig. 8B). In addition, WD induced ABCA1 phosphorylation, its ubiquitination and degradation (Fig. 8B). Consistent with the co-IP observations, coimmunofluorescence staining revealed that while CSN5-ABCA1 interactions were diminished, the tendency between cullin3 and ABCA1 interactions were slightly enhanced in the aortic root cross-sections of WD-fed ApoE^−/− mice as compared to CD-fed ApoE^−/− mice (Fig. 8C and D). Towards translational relevance, we also examined the interactions of ABCA1-CSN5-cullin3 in human non-stenotic and stenotic (stage V) coronary artery sections using coimmunofluorescence staining. In line with our findings in the mouse model, while CSN5-ABCA1 interactions were completely lost, there was a noticeable increase between cullin3 and ABCA1 interactions in human coronary artery atherosclerotic sections as compared to non-stenotic human coronary artery sections (Fig. 8E and F).

Fig. 8.

CSN5-ABCA1 interactions were disrupted in mouse and human atherosclerotic lesions. A, Aortic lysates from CD or 16 weeks of WD-fed ApoE^−/− mice were analyzed by Western blotting for ABCA1, ABCG1, SR-B1, CSN5, cullin3, and β-actin levels using their specific antibodies and quantified. B, Aortic lysates in panel A were immunoprecipitated with anti-ABCA1 antibody and the immunocomplexes were analyzed by Western blotting for the indicated proteins using their specific antibodies. The input protein was immunoblotted for β-actin. C and D, Aortic root cross-sections of CD or WD-fed ApoE^−/− mice were coimmunostained for CSN5 and ABCA1 (C) or cullin3 and ABCA1 (D). E and F, Non-stenotic and stenotic human coronary artery sections were coimmunostained for CSN5 and ABCA1 (E) or cullin3 and ABCA1 (F). ∗p < 0.01 versus CD. Scale bar in panels C–F is 200 μm.

3. Discussion

ABCA1 plays an important role in mitigating the risk of atherogenesis by facilitating cholesterol efflux [8,9]; [[34], [35], [36]]. While the function of ABCA1 in the modulation of cholesterol efflux is well studied [9,34], the biochemical processes that govern the steady-state levels of ABCA1 and its activity under normal and pathological conditions remain poorly understood. Previous studies have reported that disruption of the ABCA1 gene exacerbates atherosclerosis, whereas its stabilization enhances cholesterol efflux and diminishes atherosclerosis [[12], [13], [14]]. In line with these reports, previously we have reported that thrombin, a GPCR agonist, via inducing ABCA1 proteasomal degradation inhibits cholesterol efflux and promotes foam cell formation, thereby contributing to atherosclerosis [15]. We have also shown that thrombin-induced ABCA1 downregulation is dependent on cullin3-mediated ubiquitination and degradation [15]. In addition, we showed that while cullin3 was found to be associated with ABCA1, other cullins such as cullin1 and cullin2 although exist in complex with ABCA1 under basal condition appear to be dissociated from ABCA1 in response to thrombin [15]. These observations infer that cullin3 but not cullin1 or cullin2 was involved in thrombin-induced ubiquitination and degradation of ABCA1. In addition, our recent study has revealed that under physiological conditions, CSN3, a subunit of COP9 signalosome, protects ABCA1 from cullin3-mediated ubiquitination and degradation which in turn maintains cholesterol efflux and hinders foam cell formation [16]. The COP9 signalosome is a multi-subunit protein complex that stabilizes several proteins by inhibiting their proteasomal degradation via regulating the activity of CRLs [[17], [18], [19]]. It is composed of nine subunits – CSN1, CSN2, CSN3, CSN4, CSN5, CSN6, CSN7a/b and CSN8 and all are essential for the functional activity of the complex [19,37].

Previous studies from other laboratories have reported the involvement of various CSN subunits in protecting against the pathophysiology of atherosclerosis [[20], [21], [22]]. Specifically, it was shown that CSN5, the catalytic core subunit of the COP9 signalosome [19,20], was upregulated in human atherosclerotic arteries, and its depletion in mice worsens the formation of atherosclerotic lesions [20,21]. Moreover, interactions between CSN2 and CSN5 have been reported, and forced expression of CSN2 results in a reduction of ABCA1 ubiquitination and its degradation [22]. Consistent with these observations, our findings showed a role for CSN3 in maintaining ABCA1 stability and promoting cholesterol efflux [16]. Essentially, our work revealed that CSN3 interacts with ABCA1 and safeguards it against ubiquitination and degradation by atherogenic stimulants such as thrombin or WD [16]. Additionally, we have reported that thrombin or WD by inducing ABCA1 phosphorylation renders its dissociation from CSN3 and subsequently leads to its association with cullin3 resulting in its ubiquitination and degradation [16]. Despite all these reports demonstrating the involvement of various CSN subunits against the development of atherosclerosis, the role of the whole COP9 signalosome complex in this process remains unknown. In the present study, we report that the COP9 signalosome, as a whole complex rather than as individual subunits, is responsible for the protection of ABCA1 from ubiquitination-mediated downregulation. We found that, besides CSN3, CSN1, CSN2, CSN4, and CSN5 also exist in complex with ABCA1 under basal conditions, and all these subunits dissociate from the transporter in response to thrombin, allowing its ubiquitination and degradation by cullin3. Thus, these results infer that the COP9 signalosome, as a whole complex, mediates ABCA1 stabilization. Given the fact that CSN5, the catalytic core subunit of the COP9 signalosome, plays a pivotal role in the regulation of cullin's function through its isopeptidase activity [20,21,38], we further focused on investigating its role in the maintenance of ABCA1 stability. We found that forced expression of CSN5 recovers COP9 signalosome interaction with ABCA1 and restores cholesterol efflux from inhibition by thrombin and attenuates foam cell formation. Although, CSN5 overexpression did not increase the expression of other COP9 signalosome subunits, it appears that it enhances the formation of the COP9 holocomplex and its stability maintaining its overall deneddylation activity. In line with this view, the overexpression of CSN5 has also been reported to enhance the COP9 holocomplex formation and its stability with increased deneddylation activity without influencing the expression levels of other CSN subunits [39]. Corroborating these findings, CSN5 has also been shown to protect against atherosclerosis [20,21].

ROS play an important role in ubiquitin-proteasome-mediated protein ubiquitination and degradation [23,24]. Specifically, ROS upregulate ubiquitination pathways by enhancing the expression and/or activity of ubiquitination enzymes [23]. ROS can also influence signaling pathways that control E3 ligases [40]. Besides, a large body of data showed a role for ROS in NFκB activation and inflammation [41,42]. Interestingly, CSN5 was implicated in modulating NFκB activation [43], suggesting a potential interplay between COP9 signalosome and ROS as well [43]. Therefore, we investigated the ROS involvement in thrombin-induced ABCA1 destabilization. Our findings showed that H₂O₂ induces ABCA1 phosphorylation, its dissociation from CSN5 and association with cullin3 leading to its ubiquitination and degradation. Besides, H₂O₂ inhibited cholesterol efflux and induced foam cell formation. Moreover, the time course of xanthine oxidase expression and H₂O₂ production was correlated temporally with ABCA1 phosphorylation and its dissociation from CSN5 and association with cullin3. Together, these findings infer that H₂O₂ acts downstream to thrombin in mediating ABCA1 downregulation affecting cholesterol efflux and enhancing foam cell formation. Since H₂O₂ is a major byproduct of xanthine oxidase activity [27], and because we found that thrombin induces xanthine oxidase expression abundantly, next we tested its role in thrombin-induced ABCA1 destabilization. Pharmacological inhibition of xanthine oxidase by allopurinol blocked thrombin-induced ABCA1 destabilization. Besides, allopurinol restored cholesterol efflux from inhibition by thrombin and blunted foam cell formation both in RAW264.7 cells and MASMCs. In line with these observations, previous studies have reported that allopurinol treatment can slowdown or even reduce various cardiovascular diseases, including atherosclerosis [26,[44], [45], [46], [47], [48]]. Interestingly, our finding also revealed that xanthine oxidase acts downstream to Gα12-and upstream to Pyk2 in the Par1-Gα12-Pyk2-Gab1-PKCθ signaling in the mediation of ABCA1 diabatization by thrombin. The possible mechanism underlying Pyk2 activation by H₂O₂ was due to the capacity of oxidants to inhibit protein tyrosine phosphatase activity via oxidation of its catalytic cysteine residue and thereby enhancing tyrosine phosphorylation of protein tyrosine kinases [49,50]. Similarly, oxidants have been reported to inhibit protein phosphatase 2A (PP2A) activity by its tyrosine phosphorylation, that in turn, triggers the serine/threonine phosphorylation of serine/threonine kinases such as PKCs [51,52]. Thus, H₂O₂ might be involved in the activation of Pyk2 downstream to Gα12 and PKCθ downstream to Gab1.

GPCRs are a large family of receptors that were implicated in various diseases [28,29]. Therefore, we wanted to find whether ABCA1 destabilization is a unique response to Par-thrombin axis, or it can be influenced by other GPCR-agonists such as Ang II. To test this, we studied the effect of Ang II, a potent proatherogenic GPCR agonist [30]. Intriguingly, while showing minimal effects on ABCG1 and SR-B1 levels, Ang II downregulated ABCA1 substantially in RAW264.7 cells. Besides, like thrombin, Ang II induced ABCA1 phosphorylation, its dissociation from CSN5, and association with cullin3 leading to its ubiquitination and degradation, thereby inhibiting cholesterol efflux and promoting foam cell formation of these cells. Furthermore, Ang II induced xanthine oxidase expression and H₂O₂ production, and allopurinol blocked Ang II-induced H₂O₂ production. Additionally, forced expression of CSN5 restored cholesterol efflux from inhibition by Ang II and reduced foam cell formation in RAW264.7 cells. Furthermore, we found that AT1R-Gα12-Pyk2-Gab1-PKCθ signaling plays a role in Ang II-induced ABCA1 phosphorylation and its dissociation from CSN5 and association with cullin3 leading to its downregulation. Our observations also reveal that xanthine oxidase acts downstream to Gα12 and upstream to Pyk2 in the AT1R-Gα12-Pyk2-Gab1-PKCθ signaling in the mediation of ABCA1 diabatization by Ang II. These findings indicate that besides thrombin, other GPCR-agonists such as Ang II can also disrupt ABCA1 stabilization, leading to inhibition of cholesterol efflux and foam cell formation. Supporting these findings, previous reports showed that Ang II plays a role in the function of the COP9 signalosome complex [53,54]. Specifically, it was shown that Ang II increases the expression of CSN6 in hypertrophic mouse hearts and rat cardiomyocytes [53]. In addition, depletion of CSN6 levels decreases cardiomyocyte size and inhibits expression of the genes associated with Ang II-induced hypertrophy [53]. Conversely, overexpression of CSN6 inhibits Ang II-induced myocardial hypertrophy [53]. Supporting our in vitro observations, we found that while CSN5 and ABCA1 were found to be colocalized in aortic root cross sections of CD-fed ApoE^−/− mice and non-stenotic human coronary artery sections, ABCA1 levels were diminished in aortic root cross sections of WD-fed ApoE^−/− mice and advanced human atherosclerotic lesions. Although cullin3 was associated with ABCA1 in cells in response to GPCR agonists, their colocalization was found to be minimal in aortic root cross sections of WD-fed ApoE^−/− mice and stenotic human coronary artery sections. This could be due to a substantial reduction of ABCA1 levels in these sections. The above observations were in line with our previous reports where we showed that depletion of CSN3 leads to downregulation of ABCA1 levels and cholesterol efflux, exacerbating atherosclerosis in ApoE^−/− mice [16]. In conclusion, our observations reveal for the first time that the COP9 signalosome as a whole complex is required for ABCA1 protection in the modulation of cholesterol efflux and preventing foam cell formation and atherosclerosis. In addition, our findings showed a novel role for xanthine oxidase-dependent ROS production in GPCR-agonists, thrombin and Ang II-induced ABCA1 destabilization leading to diminished cholesterol efflux and increased foam cell formation. Furthermore, the present findings show for the first time that xanthine oxidase acts downstream to Gα12 and upstream to Pyk2 in the Par1/AT1R-Gα12-Pyk2-Gab1-PKCθ signaling by thrombin/Ang II in the mediation of ABCA1 destabilization leading to inhibition of cholesterol efflux and promoting foam cell formation. Since both thrombin and Ang II had no effect on the steady state levels of ABCG1 or SR-B1, it is unlikely that these transporters can compensate the downregulation of cholesterol efflux caused by the depletion of ABCA1 in response to these GPCR agonists. In view of all these findings, it may be speculated that stroke or coronary artery disease patients with increased circulating levels of thrombin and/or Ang II may be vulnerable for the development of atherosclerosis [30,[55], [56], [57], [58]]. If this is the case, use of xanthine oxidase inhibitors should protect ABCA1 from ROS-induced destabilization restoring cholesterol efflux and thereby may prevent the development of atherosclerosis.

4. Materials and methods

4.1. Materials

Anti-β-actin (sc-47778), anti-ABCG1 (sc-20795), anti-AT1R (sc-515884), anti-Gα12 (sc-409), anti-Gα13 (sc-293424), anti-Gab1 (sc-6292), anti-Par1 (sc-13503), anti-SRB1 (sc-67099), and anti-XO (sc-398548) antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-ABCA1 (ab18180), anti-CSN1 (ab4535), anti-CSN2 (ab155774), anti-CSN3 (ab229807), anti-CSN4 (ab12322), anti-CSN5 (ab75851), anti-Cullin3 (ab75851), anti-Gαq (ab153951), anti-Pyk2 (ab32571) and anti-pSer/Thr (ab17464) antibodies were obtained from Abcam (Cambridge, MA). Anti-CSN5 (6895), anti-phospho PKCθ (9377), anti-PKCθ (13643) and HRP-conjugated light chain-specific anti-mouse (58802 S) and anti-rabbit IgGs (93702 S) antibodies were purchased from Cell Signaling Technology (Beverly, MA). Goat anti-rabbit HRP (31460) and goat anti-mouse HRP (31437) antibodies and lipofectamine 3000 transfection reagent (L3000-015) were bought from Thermo Fisher Scientific (Waltham, MA). Ang II (Cat # SCP0020), allopurinol (Cat # PHR1377), Duolink in situ PLA kit (Cat # DUO92101), H₂O₂ (Cat #H1009), losartan (Cat # SML3317), thrombin (Cat #T8885) and uric acid (Cat #U2875) were procured from Sigma-Aldrich (St. Louis, MO). OxLDL (BT910) was purchased from Alfa Aesar (Tweksbury, MA). MycDDK-tagged pCMV-CSN5 plasmid (MR204941) was bought from Origene (Rockville, MD). Anti-ubiquitin antibody (P4D1) was purchased from Enzo Life Sciences (Farmingdale, NY). Control non-targeting siRNA (D-001810-10), mouse Gα12 siRNA (L-043467-00) and mouse PKCθ siRNA (L-048426-00) were obtained from Dharmacon RNAi Technologies (Chicago, IL). Mouse CSN5 Silencer Select (ID-s77284), mouse Gab1 Silencer Select (ID-s66350), mouse Par1 Silencer Select (ID-s65790), mouse Pyk2 Silencer Select (ID-s72406) were purchased from Ambion (Waltham, MA). The Vectashield mounting medium (H-1000) was obtained from Vector Laboratories (Burlingame, CA). [³H]-Cholesterol (S.A. 53 Ci/mmol) was purchased from PerkinElmer (Waltham, MA). ApoA-I (BT927) was obtained from Biomedical Technologies (Stoughton, MA). Enhanced chemiluminescence Western blotting detection reagents (RPN2106) were obtained from GE Healthcare (Pittsburg, PA).

4.2. Animals

All the experiments involving animals were performed according to the guidelines of the Institutional Animal Care and Use Committee of the University of Tennessee Health Science Center (UTHSC), Memphis, TN. ApoE^−/− mice (stock number 002052) were purchased from The Jackson Laboratory. Mice were maintained at UTHSC laboratory animal care unit. The UTHSC laboratory animal care unit facility maintains a 12/12 h light/dark cycle and the animals had ad libitum access to food and water. Mice were bred and maintained according to the institutional animal care and use committee guidelines. Eight-week-old ApoE^−/− mice (male and female) were kept on either CD (Teklad Irradiated LM-485 mouse/rat diet, Envigo, Catalog number 7912) or WD containing 21% fat and 0.2% cholesterol (Cat # TD.88137, Harlan Teklad, Harlan Laboratories) for 16 weeks. At the end of the experimental period, mice were euthanized by ketamine/xylazine overdose and aortas were collected for further analysis as required. All the mice used in the study were confirmed by genotyping.

4.3. Human normal and atherosclerotic artery sections

Collection of human non-stenotic and stenotic coronary arteries and preparation of tissue sections were described previously [16]. Briefly, the postmortem samples were obtained from both the genders with age group ranging from 21 to 71 years. The plaque severity was determined by H&E staining and scoring by two independent pathologists using a modified Stary classification system [16]. These samples were considered to be nonhuman subjects, as they were postmortem samples.

4.4. Cell culture

RAW264.7 cells (Cat # TIB-71) and MASMCs (Cat # CRL-2797) were purchased from American Type Culture Collection, and Cell Biologics Inc, respectively. RAW264.7 cells and MASMCs were cultured and maintained in DMEM containing 10% FBS, 100 units/ml penicillin, and 100 μg/ml streptomycin at 37 °C in a humidified 95% air and 5% CO₂ atmosphere. The cells were growth-arrested overnight in basal medium without growth supplements and used unless otherwise indicated.

4.5. Transfections

RAW264.7 cells and MASMCs were transfected with scrambled or targeted siRNA at a final concentration of 100 pmol/ml using Lipofectamine 3000 transfection reagent according to the manufacturer's instructions. In the case of plasmid vectors, cells were transfected with plasmid DNAs at a final concentration of 2.5 μg/60 mm dish or 5 μg/100 mm dish. After 6 h of transfection, cells were recovered in complete medium for 24 h, growth-arrested overnight in serum-free medium and used as required.

4.6. Western blotting

Cells with appropriate treatments were washed with PBS and scraped into RIPA buffer containing protease inhibitors (PBS containing 1% Nonidet P40, 0.5% sodium deoxycholate, 0.1% SDS, 100 μg/ml PMSF, 100 μg/ml aprotinin, 1 μg/ml leupeptin, and 1 mM sodium orthovanadate) and lysed by sonication at 45% amplitude for 15 s with 10 s intervals for 2 min on ice (Branson Sonifier 450). To prepare aortic extracts, it was cleaned from perivascular fat, chopped, and minced in RIPA buffer using a hand homogenizer and then sonicated at 45% amplitude for 15 s with 10 s intervals for 4 min on ice. Cell extracts were cleared by centrifugation at 12,000 rpm for 10 min at 4 °C and protein concentration was determined using Mico bicinchoninic acid kit (Thermo Fisher Scientific). Cell extracts containing equal amounts of protein from control and the indicated treatments were resolved by electrophoresis on 0.1% (w/v) SDS and 10% or 12% (w/v) polyacrylamide gels. The proteins were transferred electrophoretically onto a nitrocellulose membrane. After blocking in 5% (w/v) nonfat dry milk or 5% bovine serum albumin (BSA), the membranes were incubated with the appropriate primary antibodies at 1:1000 dilution overnight at 4 °C, followed by incubation with HRP-conjugated secondary antibodies at 1:5000 dilution for 1 h at room temperature (RT). After washing in Tris-buffered saline containing Tween 20 (TBST) three times, the membranes were incubated with enhanced chemiluminescence detection reagent, and the antigen-antibody complexes were detected by exposure to X-Ray film and developing it in a AFP Mini-Med 90 X-Ray film processor (AFP Imaging Corp, Elmsford, NY). The band intensities were quantified by densitometry using NIH ImageJ software.

4.7. Pull-down assay

Cell or tissue extracts were prepared by lysis in 400 μl of lysis buffer for 20 min on ice. The extracts were cleared by centrifugation at 12000 rpm for 20 min at 4 °C. The cleared cell extracts containing an equal amount of protein (400 μg) from control and the indicated treatments were incubated with the indicated primary antibodies (4 μg) overnight at 4 °C, followed by incubation with protein A/G-Sepharose CL4B beads for 4 h with gentle rocking. The beads were collected by centrifugation at 1000 rpm for 2 min at 4 °C and washed 4X with lysis buffer and once with PBS. The immunocomplexes were released by heating the beads in 40 μl of Laemmli sample buffer and analyzed by Western blotting for the indicated proteins using their specific antibodies.

4.8. Proximity ligation assay

The interaction between endogenous ABCA1, CSN5 and cullin3 was analyzed by using in situ proximity ligation assay kit (Duolink PLA kit; Sigma) as described previously [16,59]. In brief, after appropriate treatments, cells were fixed with 4% paraformaldehyde for 15 min, permeabilized in 0.3% Triton X100 for 15 min, and blocked with 3% BSA for 1 h. The cells were then incubated with mouse anti-ABCA1 (1:100) and rabbit anti-CSN5 or cullin3 antibodies (1:100) overnight at 4 °C followed by incubation with goat anti-mouse secondary antibody conjugated with oligonucleotide PLA probe Plus and goat anti-rabbit secondary antibody conjugated with oligonucleotide PLA probe Minus for 1 h at 37 °C. The cells were then incubated with a ligation solution for 30 min at 37 °C followed by a rolling-circle amplification of ligated oligonucleotide probes. A fluorescently labeled complementary oligonucleotide detection probe was used to amplify the oligonucleotides conjugated to the secondary antibodies. After washing with the wash buffer, the slides were mounted with a mounting medium. PLA signals were examined using a Zeiss inverted fluorescence microscope (Axiovision Observer.z1; 40X/NA 0.6) and the fluorescence images were captured using Zeiss AxioCam MRm camera and analyzed using microscope operating and image analysis software ZEN 2.6 (blue edition).

4.9. H₂O₂ production assay

Cell lysates from RAW264.7 cells or MASMCs with appropriate treatments were incubated with 50 μl of 100 μM Amplex Red along with HRP (0.2 U/ml) for 30 min at 37 °C in the dark. Following incubation, the fluorescence intensity was measured in SpectraMax Gemini XPS Spectrofluorometer (Molecular Devices) with excitation at 530 nm and emission at 590 nm. The H₂O₂ production was expressed as RFU.

4.10. Cholesterol efflux assay

RAW264.7 cells or MASMCs that were transfected with the indicated plasmids and/or with indicated treatments were plated in 12-well plates at a density of 2 × 10⁵ cells/well. Cells were incubated with oxLDL (20 μg/ml) and [³H]-cholesterol (1 μCi/ml) for 24 h followed by extensive washings with PBS. Cells were then equilibrated in serum-free DMEM containing 0.2% fatty acid free-BSA for 2 h. After equilibration, the medium was replaced with fresh DMEM containing 0.2% fatty acid free-BSA and ApoA-I (10 μg/ml) and incubation was continued for 4 h in the presence and absence of indicated agents [thrombin (0.5 U/ml) or Ang II (1 μM) or allopurinol (50 μM) or H₂O₂ (200 μM) or uric acid (100 μM)]. An aliquot of the efflux medium (100 μl) was removed for radioactivity determination. Cells were then rinsed with PBS, dried, and isopropanol was added for cholesterol extraction overnight at RT. An aliquot of the extract (100 μl) was used for radioactivity counting by LS6500 Multipurpose Scintillation Counter (Beckman Coulter). Cholesterol efflux was measured as % of total cellular radioactivity released into the medium.

4.11. Foam cell formation assay

RAW264.7 cells or MASMCs that were transfected with the indicated plasmids and/or with indicated treatments [thrombin (0.5 U/ml) or Ang II (1 μM) or allopurinol (50 μM) or H₂O₂ (200 μM) or uric acid (100 μM) for 4 h] were incubated with oxLDL (10 μg/ml) for 6 h at 37 °C. Cells were then fixed with 4% paraformaldehyde for 30 min, stained with BODIPY solution (2 μM) for 15 min and counterstained with 4’,6-diamidino-2-phenylindole, mounted with a cover slip using ProLong Gold antifade reagent and observed under a Zeiss inverted microscope (AxioObserver.Z1; X10/0.25 NA). The fluorescence images were captured using Zeiss AxioCam MRm camera and analyzed by ZEN (Blue edition) V2.0 software.

4.12. Aortic root cross sections

Aortic root cross sections were made from the hearts of CD-fed or 16 weeks of WD-fed ApoE^−/− mice for immunofluorescence staining. Briefly, anesthetized mice were perfused with 4% PFA followed by PBS and the hearts were dissected out, fixed in 4% PFA overnight and then placed in OCT. Sequential 10 μm aortic root sections were cut from the point of appearance of the aortic valve leaflets with a Leica CM3050Scryostat machine (Leica Biosystems).

4.13. Immunofluorescence staining

The mouse aortic root sections were fixed with acetone/methanol (1:1), permeabilized with Triton X100, and blocked with 3% BSA containing 5% goat serum. In the case of human non-stenotic and stenotic coronary artery sections, first they were deparaffinized with xylene and then treated with antigen unmasking solution for 15 min at 95 °C. The sections were then permeabilized with 0.5% Triton X100 for 15 min and blocked with normal goat serum. After blocking, sections were incubated with the indicated antibodies, at 1:100 dilution, followed by incubation with Alexa Fluor 488–conjugated goat anti-mouse and Alexa Fluor 568–conjugated goat anti-rabbit secondary antibodies at 1:500 dilution. The sections were washed with PBS, counterstained with Hoechst, mounted with a cover slip using ProLong Gold antifade reagent and observed under a Zeiss inverted microscope (AxioObserver.Z1; X10/0.25 NA). The fluorescence images were captured using Zeiss AxioCam MRm camera and analyzed by ZEN (Blue edition) V2.0 software.

4.14. Statistical analysis

All the experiments were repeated 3 times. Normality of the data was analyzed using Anderson-Darling Normality Test and F-test was used to assess the equality and group variance. The treatment effects were analyzed by student t-test for two group comparisons and one-way analysis of variance (ANOVA) followed by Bonferroni post hoc test for multiple comparisons. All the statistical tests were performed using GraphPad Prism v 8.00 software. In the case of in vivo experiments, a minimum of 7 mice for each gender were included. Since we did not observe statistical differences between male and female mice in our experiments on platelet activation and hemostasis, we did not express the results separately. The data are presented as Mean ± SD and the p < 0.05 were considered statistically significant.

Ethics approval

All the experiments involving animals were performed according to the guidelines of the Institutional Animal Care and Use Committee of the University of Tennessee Health Science Center (UTHSC), Memphis, TN.

Source of funding

This work was supported in part by NEI of National Institutes of Health grant EY034425 to GNR and National Institutes of Health grants GM121307, HL133497 and HL173972 to AWO.

CRediT authorship contribution statement

Suresh Govatati: Data curation, Formal analysis, Investigation, Methodology, Validation, Writing – original draft. Chandra Sekhar Yadavalli: Data curation, Formal analysis, Investigation. James G. Traylor Jr: Resources, Writing – review & editing. A. Wayne Orr: Resources, Writing – review & editing. Gadiparthi N. Rao: Conceptualization, Funding acquisition, Project administration, Resources, Supervision, Validation, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request.