Metabolic regulation of the proteasome under hypoxia by Poldip2 controls fibrotic signaling in vascular smooth muscle cells

Abstract

The polymerase delta interacting protein 2 (Poldip2) is a nuclear-encoded mitochondrial protein required for oxidative metabolism. Under hypoxia, Poldip2 expression is repressed by an unknown mechanism. Therefore, low levels of Poldip2 are required to maintain glycolytic metabolism. The Cellular Communication Network Factor 2 (CCN2, Connective tissue growth factor, CTGF) is a profibrogenic molecule highly expressed in cancer and vascular inflammation in advanced atherosclerosis. Because CCN2 is upregulated under hypoxia and is associated with glycolytic metabolism, we hypothesize that Poldip2 downregulation is responsible for the upregulation of profibrotic signaling under hypoxia. Here, we report that Poldip2 is repressed under hypoxia by a mechanism that requires the activation of the enhancer of zeste homolog 2 repressive complex (EZH2) downstream from the Cyclin-Dependent Kinase 2 (CDK2). Importantly, we found that Poldip2 repression is required for CCN2 expression downstream of metabolic inhibition of the ubiquitin-proteasome system (UPS)-dependent stabilization of the serum response factor. Pharmacological or gene expression inhibition of CDK2 under hypoxia reverses Poldip2 downregulation, the inhibition of the UPS, and the expression of CCN2, collagen, and fibronectin. Thus, our findings connect cell cycle regulation and proteasome activity to mitochondrial function and fibrotic responses under hypoxia.

Keywords: CCN2; Cell cycle; Hypoxia; Mitochondria; Poldip2; Proteasome.

Figure 1. Hypoxia induces the activation of CDK2 and EZH2 phosphorylation in HASMCs.

HASMCs were cultured under normoxia or hypoxia for 48 hours. Total lysates were prepared and analyzed by western blot using specific antibodies (Left). β-actin served as a loading control. Densitometric analysis of protein signals normalized to β-actin (Right). Bar graphs represent mean ± SE from 4 independent experiments.

Figure 2.. A CDK2/CyclinA-dependent mechanism mediates EZH2 phosphorylation and Poldip2 downregulation under hypoxia.

HASMCs were transfected with control siRNA (siControl) or siRNA against (A) CDK2 (siCDK2) or (B) Cyclin A2 (siCCNA2). HASMCs were cultured under normoxia or hypoxia for 48 h. Total lysates were prepared and analyzed by western blot using specific antibodies (Left). β-actin served as a loading control. Densitometric analysis of protein signals normalized to β-actin (Right). Bar graphs represent mean ± SE from 4 independent experiments.

Figure 3.. Poldip2 is downregulated under hypoxia by an EZH2-dependent mechanism.

HASMCs were transfected with control siRNA (siControl) or siRNA against EZH2 (siEZH2). After 24 hours, cells were cultured under normoxia or hypoxia for 48 hours. Total lysates were prepared and analyzed by western blot using specific antibodies (Left). β-actin served as a loading control. Densitometric analysis of protein signals normalized to β-actin (Right). Bar graphs represent mean ± SE from 4 independent experiments.

Figure 4.. Proteasome inhibition under hypoxia requires Poldip2 downregulation.

(A) HASMCs were transduced with empty vector or Poldip2-myc-expressing adenovirus and cultured under normoxia or hypoxia for 48 hours. Lysates were prepared, and UPS activity was measured using a fluorogenic assay described in the Methods section. Results were expressed as a percentage of the control. (B) Representative western blot to analyze the accumulation of Poly-ubiquitinated proteins and the reporter Ub-GFP signal. Bar graphs represent mean ± SE from 4 independent experiments.

Figure 5.. Hypoxia-increased flux through the HBP and OGT activity requires Poldip2 downregulation.

HASMCs were transduced with empty vector or Poldip2-myc-expressing adenovirus and cultured under normoxia or hypoxia for 48 hours. Lysates were prepared and analyzed by western blots using specific antibodies. Representative western blots show the expression of key components of the HBP and the accumulation of O-linked GlcNAcylated proteins. Bar graphs are presented as mean ± SE from 4 independent experiments.

Figure 6.. UPS inhibition under hypoxia is OGT-dependent.

HASMCs were transfected with siRNA control (siControl) or siRNA against OGT (siOGT). After 24 hours, cells were cultured under normoxia or hypoxia for 48 hours. (A) Proteasome activity was measured using a fluorogenic assay described in the Methods section. (B) Representative western blot to analyze the accumulation of Poly-ubiquitinated proteins and the reporter Ub-GFP signal. Bar graphs represent mean ± SE from 4 independent experiments.

Figure 7.. Hypoxia-induced profibrotic gene expression requires Poldip2 downregulation.

HASMCs were transduced with empty vector or Poldip2-myc-expressing adenovirus and cultured under normoxia or hypoxia for 48 hours. Lysates were prepared and analyzed by western blot using specific antibodies against profibrotic proteins (Left). β-actin served as a loading control. Densitometric analysis of protein signals normalized to β-actin (Right).

Figure 8.. The media of arteries of Poldip2 deficient animals display exacerbate expression of CCN2.

Wild type and Poldip2 deficient mice were treated with AngII for 30 days, hearts were harvested and process for histology. A. Expression of CCN2 and HIF1a in coronary arteries acquired by Confocal microcopy. B. Bar graphs of mean fluorescence intensity ± SE calculated using Image J software (NIH) from 10 independent experiments. C. Representative images of Masson’s Trichrome staining of heart tissue and Bar graphs of mean staining intensity ± SE calculated using Image J software (NIH) from 10 independent experiments.

Figure 9.. Hypoxia-induced profibrotic gene expression requires OGT-mediated inhibition of the UPS.

(A) HASMCs were transfected with control siRNA (siControl) or siRNA against OGT (siOGT) and cultured under normoxia or hypoxia for 48 hours. Lysates were prepared and analyzed by western blots using specific antibodies (Left). β-actin served as a loading control. Densitometric analysis of protein signals normalized to β-actin (Right). (B) HASMCs were transduced with an empty vector or adenovirus expressing PSMC1-myc and cultured under normoxia or hypoxia for 48 hours. Lysates were prepared and analyzed by western blots using specific antibodies (Left). β-actin served as a loading control. Densitometric analysis of protein signals normalized to β-actin (Right). Bar graphs are presented as mean ± SE from 4 independent experiments.

Figure 10.. Hypoxia-induced profibrotic gene expression requires CDK2.

HASMCs were transfected with control siRNA (siControl) or siRNA against CDK2 (siCDK2). HASMCs were cultured under normoxia or hypoxia for 48 h. Total lysates were prepared and analyzed by western blot using specific antibodies (Left). β-actin served as a loading control. Densitometric analysis of protein signals normalized to β-actin (Right). Bar graphs are presented as mean ± SE from 4 independent experiments.

Figure 11.. Hypoxia-induced profibrotic genes expression requires EZH2.

A. HASMCs were transfected with control siRNA (siControl) or siRNA against EZH2 (siEZH2) and cultured under normoxia or hypoxia for 48 hours. Lysates were prepared and analyzed by western blots using specific antibodies (Left). β-actin served as a loading control. Densitometric analysis of protein signals normalized to β-actin (Right). Bar graphs are presented as mean ± SE from 4 independent experiments.

Figure 12.. Stabilization of SRF by UPS inhibition downstream Poldip2 repression is required for profibrotic signaling.

A. HASMCs were transduced with empty vector or Poldip2-myc-expressing adenovirus and cultured under normoxia or hypoxia for 48 hours. Lysates were prepared and analyzed by western blot using specific antibodies (Left). β-actin served as a loading control. Densitometric analysis of protein signals normalized to β-actin (Right). B. HASMCs were transduced with an empty vector or adenovirus expressing PSMC1-myc and cultured under normoxia or hypoxia for 48 hours. Lysates were prepared and analyzed by western blots using specific antibodies (Left). β-actin served as a loading control. Densitometric analysis of protein signals normalized to β-actin (Right). C. HASMCs were transfected with control siRNA (siControl) or siRNA against SRF (siSRF) and cultured under normoxia or hypoxia for 48 hours. Lysates were prepared and analyzed by western blots using specific antibodies (Left). β-actin served as a loading control. Densitometric analysis of protein signals normalized to β-actin (Right). Bar graphs are presented as mean ± SE from 4 independent experiments.

Figure 12.. Stabilization of SRF by UPS inhibition downstream Poldip2 repression is required for profibrotic signaling.

A. HASMCs were transduced with empty vector or Poldip2-myc-expressing adenovirus and cultured under normoxia or hypoxia for 48 hours. Lysates were prepared and analyzed by western blot using specific antibodies (Left). β-actin served as a loading control. Densitometric analysis of protein signals normalized to β-actin (Right). B. HASMCs were transduced with an empty vector or adenovirus expressing PSMC1-myc and cultured under normoxia or hypoxia for 48 hours. Lysates were prepared and analyzed by western blots using specific antibodies (Left). β-actin served as a loading control. Densitometric analysis of protein signals normalized to β-actin (Right). C. HASMCs were transfected with control siRNA (siControl) or siRNA against SRF (siSRF) and cultured under normoxia or hypoxia for 48 hours. Lysates were prepared and analyzed by western blots using specific antibodies (Left). β-actin served as a loading control. Densitometric analysis of protein signals normalized to β-actin (Right). Bar graphs are presented as mean ± SE from 4 independent experiments.

Figure 13.

Proposed signaling pathway leading to increased expression of profibrotic genes under hypoxia.