MicroRNA-33a regulates cholesterol synthesis and cholesterol efflux-related genes in osteoarthritic chondrocytes

Abstract

Introduction: Several studies have shown that osteoarthritis (OA) is strongly associated with metabolism-related disorders, highlighting OA as the fifth component of the metabolic syndrome (MetS). On the basis of our previous findings on dysregulation of cholesterol homeostasis in OA, we were prompted to investigate whether microRNA-33a (miR-33a), one of the master regulators of cholesterol and fatty acid metabolism, plays a key role in OA pathogenesis.

Methods: Articular cartilage samples were obtained from 14 patients with primary OA undergoing total knee replacement surgery. Normal cartilage was obtained from nine individuals undergoing fracture repair surgery. Bioinformatics analysis was used to identify miR-33a target genes. miR-33a and sterol regulatory element-binding protein 2 (SREBP-2) expression levels were investigated using real-time PCR, and their expression was also assessed after treatment with transforming growth factor-β1 (TGF-β1) in cultured chondrocytes. Akt phosphorylation after treatment with both TGF-β1 and miR-33a inhibitor or TGF-β1 and miR-33a mimic was assessed by Western blot analysis. Furthermore, we evaluated the effect of miR-33a mimic and miR-33a inhibitor on Smad7, a negative regulator of TGF-β signaling, on cholesterol efflux-related genes, ATP-binding cassette transporter A1 (ABCA1), apolipoprotein A1 (ApoA1) and liver X receptors (LXRα and LXRβ), as well as on matrix metalloproteinase-13 (MMP-13), using real-time PCR.

Results: We found that the expression of miR-33a and its host gene SREBP-2 was significantly elevated in OA chondrocytes compared with normal chondrocytes. Treatment of cultured chondrocytes with TGF-β1 resulted in increased expression of both miR-33a and SREBP-2, as well as in rapid induction of Akt phosphorylation, whereas TGF-β-induced Akt phosphorylation was enhanced by miR-33a and suppressed by inhibition of miR-33a, as a possible consequence of Smad7 regulation by miR-33a. Moreover, treatment of normal chondrocytes with miR-33a resulted in significantly reduced ABCA1 and ApoA1 mRNA expression levels and significantly elevated MMP-13 expression levels, promoting the OA phenotype, whereas miR-33a's suppressive effect was reversed using its inhibitor.

Conclusions: Our findings suggest, for the first time to our knowledge, that miR-33a regulates cholesterol synthesis through the TGF-β1/Akt/SREBP-2 pathway, as well as cholesterol efflux-related genes ABCA1 and ApoA1, in OA chondrocytes, pointing to its identification as a novel target for ameliorating the OA phenotype.

Figure 1

MicroRNA-33a is elevated in osteoarthritic chondrocytes and is induced by transforming growth factor-β1. (A) Relative expression of microRNA (miR)-33a in normal and osteoarthritic (OA) chondrocytes (n = 9 for normal chondrocytes from 9 different donors, n = 14 for OA chondrocytes from 14 different donors). U6 was used for normalization of the real-time PCR data. The data are expressed as mean and standard error of the mean (SEM) of three independent experiments, each of which was run in duplicate. *P < 0.05 as measured using an unpaired Student’s t-test. (B) Relative expression of sterol regulatory element-binding protein 2 (SREBP-2) in normal and OA chondrocytes (n = 5 for normal chondrocytes from 5 different donors, n = 10 for OA chondrocytes from 10 different donors). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used for normalization of the real-time PCR data. The data are expressed as mean and SEM of three independent experiments, each of which was run in duplicate. *P < 0.05 as measured using an unpaired Student’s t-test. (C) and (D) miR-33a expression levels in cultured normal chondrocytes (n = 4 from 4 different donors) following treatment with 10 ng/ml transforming growth factor (TGF)-β1 for 6 hours, 24 hours (C) and 48 hours (D). U6 was used for normalization of the real-time PCR data. The data are expressed as mean and SEM of two independent experiments, each of which was run in triplicate. *P < 0.05 compared with control. (E) and (F) SREBP-2 expression levels in cultured normal chondrocytes (n = 5 from 5 different donors) following treatment with 10 ng/ml TGF-β1 for 6 hours, 24 hours (E) and 48 hours (F). GAPDH was used for normalization of the real-time PCR data. The data are expressed as mean and SEM of two independent experiments, each of which was run in triplicate. *P < 0.05 compared with control.

Figure 2

Effect of microRNA-33a on transforming growth factor-β1-mediated phosphoinositide 3-kinase/Akt signaling pathway in human chondrocytes. (A) Sterol regulatory element-binding protein 2 (SREBP-2) mRNA expression in human normal chondrocytes (n = 3 from 3 different donors) after treatment with 10 ng/ml transforming growth factor (TGF)-β1 with or without microRNA (miR)-33a inhibitor. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used for normalization of the real-time PCR data. The data are expressed as mean and SEM from triplicates of one representative of three experiments. *P < 0.05 versus negative control, **P < 0.01, #TGF-β1 with miR-33a inhibitor versus TGF-β1. (B) Western blot showing phosphorylated (p-Akt) and total Akt protein expression levels in normal chondrocytes treated with 10 ng/ml TGF-β1 for 0.5, 2 and 24 hours. (C) Diagram showing p-Akt protein expression normalized to total Akt using ImageJ software. The result shown represents the mean from three different blots. *P < 0.05 compared with control. (D) Western blot showing p-Akt and total Akt protein expression levels in normal chondrocytes treated with TGF-β1, TGF-β1 plus miR-33a or negative control. (E) Diagram showing p-Akt protein expression normalized to total Akt using ImageJ software. The result shown represents the mean from three different blots. *P < 0.05 compared with control. #TGF-β1 with miR-33a versus TGF-β1. (F) Western blot showing p-Akt and total Akt protein expression levels in osteoarthritis chondrocytes treated with TGF-β1 and miRNA inhibitor (anti-miR-33a), followed by treatment of TGF-β1 or negative control. (G) Diagram showing p-Akt protein expression normalized to total Akt using ImageJ software. The result shown represents the mean from three different blots. *P < 0.05 compared with control. #TGF-β1 with miR-33a inhibitor versus TGF-β1.

Figure 3

ΜicroRNA-33a affects Smad7 expression in human chondrocytes. (A) TargetScan 6.2 and miRanda identified one conserved sequence in the 3′ untranslated region of human Smad7 mRNA that was completely complementary to microRNA (miR)-33a. (B) Relative expression levels of miR-33a in normal human chondrocytes (n = 3 from 3 different donors) transfected with miR-33a mimic or negative control. U6 was used for normalization of the real-time PCR data. The data are expressed as mean and standard error of the mean (SEM) of three independent experiments, each of which was run in duplicate. *P < 0.05 versus negative control. (C) Relative expression levels of Smad7 mRNA 24 hours after transfection of 30 nM miR-33a or negative control in normal human chondrocytes (n = 6 from 6 different donors). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used for normalization of the real-time PCR data. The data are expressed as mean and SEM of three independent experiments, each of which was run in triplicate. *P < 0.05 versus negative control. (D) Relative expression levels of miR-33a in human osteoarthritis (OA) chondrocytes transfected with miR-33a inhibitor or negative control (n = 3 from 3 different donors). U6 was used for normalization of the real-time PCR data. The data are expressed as mean and SEM of three independent experiments, each of which was run in duplicate. *P < 0.05 versus negative control. (E) Relative expression levels of Smad7 mRNA 24 hours after transfection of 50 nM anti-miR-33a or negative control in human OA chondrocytes (n = 4 from 4 different donors). GAPDH was used for normalization of the real-time PCR data. The data are expressed as mean and SEM of three independent experiments, each of which was run in triplicate. *P < 0.05 versus negative control.

Figure 4

ΜicroRNA-33a affects ATP-binding cassette transporter A1 and apolipoprotein A1 expression in human chondrocytes. (A) TargetScan 6.2, miRanda and miRDB identified three conserved sequences in the 3′ untranslated region of human ATP-binding cassette transporter A1 (ABCA1) mRNA that were completely complementary to microRNA (miR)-33a. (B) and (C) Relative expression levels of ABCA1 (B) and apolipoprotein A1 (ApoA1) (C) mRNA at 6, 24 and 48 hours after transfection of 30 nM miR-33a or negative control in normal human chondrocytes (n = 6 from 6 different donors). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used for normalization of the real-time PCR data. The data are expressed as mean and standard error of the mean (SEM) of two independent experiments, each of which was run in triplicate. *P < 0.05 versus negative control. (D) Relative expression levels of matrix metalloproteinase (MMP)-13 mRNA at 6, 24 and 48 hours after transfection of 30 nM miR-33a or negative control in normal human chondrocytes (n = 4 from 4 different donors). GAPDH was used for normalization of the real-time PCR data. The data are expressed as mean and SEM of two independent experiments, each of which was run in triplicate. *P < 0.05 versus negative control. (E) and (F) Relative expression levels of liver X receptor α (LXRα) (E) and LXRβ (F) mRNA at 6, 24 and 48 hours after transfection of 30 nM miR-33a or negative control in normal human chondrocytes (n = 4 from 4 different donors). GAPDH was used for normalization of the real-time PCR data. The data are expressed as mean and SEM of two independent experiments, each of which was run in triplicate. NS, Not significant.

Figure 5

Anti-microRNA-33a affects ATP-binding cassette transporter A1 and apolipoprotein A1 expression in human chondrocytes. (A) Relative expression levels of ATP-binding cassette transporter A1 (ABCA1) mRNA 24 and 48 hours after transfection of 50 nM anti-microRNA (miR)-33a or negative control in human osteoarthritis (OA) chondrocytes (n = 5 from 5 different donors). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used for normalization of the real-time PCR data. The data are expressed as mean and standard error of the mean (SEM) of two independent experiments, each of which was run in triplicate. *P < 0.05 versus negative control. (B) Relative expression levels of apolipoprotein A1 (ApoA1) mRNA 24 and 48 hours after transfection of 50 nM anti-miR-33a or negative control in human OA chondrocytes (n = 5 from 5 different donors). GAPDH was used for normalization of the real-time PCR data. The data are expressed as mean and SEM of two independent experiments, each of which was run in triplicate. *P < 0.05 versus negative control. (C) Relative expression levels of matrix metalloproteinase (MMP)-13 mRNA 24 and 48 hours after transfection of 50 nM anti-miR-33a or negative control in human OA chondrocytes (n = 5 from 5 different donors). GAPDH was used for normalization of the real-time PCR data. The data are expressed as mean and SEM of two independent experiments, each of which was run in triplicate. *P < 0.05 versus negative control.