- Email: [email protected]
Reciprocal regulation by hypoxia-inducible factor-2α and the NAMPT-NAD+-SIRT axis in articular chondrocytes is involved in osteoarthritis
Osteoarthritis and Cartilage xxx (2015) 1e9
Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPTNADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis H. Oh y a, J.-S. Kwak y a, S. Yang y, M.-K. Gong y, J.-H. Kim y z, J. Rhee y, S.K. Kim y, H.-E. Kim y, J.-H. Ryu y x, J.-S. Chun y * y School of Life Sciences, Gwangju Institute of Science and Technology, Gwangju 500-712, South Korea z Department of Biological Sciences, Seoul National University, Seoul 151-747, South Korea x Research Center for Biomineralization Disorders and Dental Science Research Institute, School of Dentistry, Chonnam National University, Gwangju 500757, South Korea
a r t i c l e i n f o
s u m m a r y
Article history: Received 27 April 2015 Accepted 15 July 2015
Objective: Hypoxia-inducible factor-2a (HIF-2a) transcriptionally upregulates Nampt in articular chondrocytes. NAMPT, which exhibits nicotinamide phosphoribosyltransferase activity, in turn causes osteoarthritis (OA) in mice by stimulating the expression of matrix-degrading enzymes. Here, we sought to elucidate whether HIF-2a activates the NAMPT-NADþ-SIRT axis in chondrocytes and thereby contributes to the pathogenesis of OA. Methods: Assays of NAD levels, SIRT activity, reporter gene activity, mRNA, and protein levels were conducted in primary cultured mouse articular chondrocytes. Experimental OA in mice was induced by intra-articular (IA) injection of adenovirus expressing HIF-2a (Ad-Epas1) or NAMPT (Ad-Nampt). The functions of SIRT in OA were examined by IA co-injection of SIRT inhibitors or adenovirus expressing individual SIRT isoforms or shRNA targeting specific SIRT isoforms. Results: HIF-2a activated the NAMPT-NADþ-SIRT axis in chondrocytes by upregulating NAMPT, which stimulated NADþ synthesis and thereby activated SIRT family members. The activated NAMPT-SIRT pathway, in turn, promoted HIF-2a protein stability by negatively regulating its hydroxylation and 26S proteasome-mediated degradation, resulting in increased HIF-2a transcriptional activity. Among SIRT family members (SIRT1e7), SIRT2 and SIRT4 were positively associated with HIF-2a stability and transcriptional activity in chondrocytes. This reciprocal regulation was required for the expression of catabolic matrix metalloproteinases (MMP3, MMP12, and MMP13) and OA cartilage destruction caused by IA injection of Ad-Epas1 Ad-Nampt. Conclusion: The reciprocal regulation of HIF-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in OA cartilage destruction caused by HIF-2a or NAMPT. © 2015 The Authors. Published by Elsevier Ltd and Osteoarthritis Research Society International. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/ 4.0/).
Keywords: Hypoxia-inducible factor (HIF) NADþ biosynthesis Sirtuin Chondrocytes Osteoarthritis
Introduction Osteoarthritis (OA) is caused by mechanical stresses, including joint instability and injury, and factors that predispose toward OA, such as aging. These factors lead to the activation of biochemical pathways in chondrocytes that result in degradation of the
* Address correspondence and reprint requests to: J.-S. Chun, School of Life Sciences, Gwangju Institute of Science and Technology, Buk-Gu, Gwangju 500-712, South Korea. Tel: 82-62-715-2497; Fax: 82-62-715-3304. E-mail address: [email protected] (J.-S. Chun). a These authors contributed equally to this work.
extracellular matrix by matrix metalloproteinases (MMPs) and aggrecanases (ADAMTS)1e3. We have previously shown that among potentially relevant biochemical pathways, the zinc-ZIP8-MTF1 axis4 and hypoxia-inducible factor (HIF)-2a (encoded by Epas1)5e7 are important catabolic regulators of OA cartilage destruction that act by upregulating chondrocyte expression of matrix-degrading enzymes. Our previous study also demonstrated that the Nampt is a direct target gene of HIF-2a in OA chondrocytes8. NAMPT functions as both an intracellular form (iNAMPT) and an extracellular form (eNAMPT or visfatin). eNAMPT acts as an adipokine, whereas the nicotinamide phosphoribosyltransferase enzymatic activity of iNAMPT regulates salvage pathways of NADþ synthesis9,10. Both
http://dx.doi.org/10.1016/j.joca.2015.07.009 1063-4584/© 2015 The Authors. Published by Elsevier Ltd and Osteoarthritis Research Society International. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009
2
H. Oh et al. / Osteoarthritis and Cartilage xxx (2015) 1e9
iNAMPT and eNAMPT upregulate expression of the matrixdegrading enzymes MMP3, MMP12, and MMP13 in chondrocytes, an action that is necessary for HIF-2a-induced OA cartilage destruction8. Gain-of-function studies involving intra-articular (IA) injection of an adenoviral expression construct for NAMPT (AdNampt) and those using Col2a1-Nampt transgenic mice, as well as loss-of-function studies (i.e., IA injection of the NAMPT inhibitor FK866), have demonstrated that iNAMPT is an essential catabolic regulator of OA pathogenesis8. NAMPT enzymatic activity stimulates the synthesis of NADþ, which is an essential cofactor of sirtuin (SIRT) protein deacetylases9,10. SIRT regulates gene expression by modulating chromatin function through direct deacetylation of histones and transcription factors10e12. Therefore, it is likely that HIF-2a-induced NAMPT expression causes NADþ synthesis and thereby stimulates SIRT activity. The mammalian SIRT family is composed of seven members, SIRT1e7, with different subcellular localization patterns and enzymatic activities11. In this study, we examined whether HIF-2a regulates the NAMPT-NADþ-SIRT axis and thereby contributes to the pathogenesis of OA. We report that reciprocal regulation of HIF-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in the OA cartilage destruction caused by HIF-2a or NAMPT. Materials and methods Experimental OA and histology All animal experiments were approved by the Gwangju Institute of Science and Technology Animal Care and Use Committee. Experimental OA was induced by IA injection in 10-week-old male mice (once weekly for 3 weeks) of adenovirus expressing HIF-2a (Ad-Epas1) or NAMPT (Ad-Nampt) (1 109 plaque forming units [PFUs]), with or without the SIRT inhibitor nicotinamide (300 mg/kg body weight) or EX527 (1.25 mg/kg body weight)12,13, or adenovirus expressing SIRT1 (Ad-Sirt1) or SIRT2 (Ad-Sirt2); IA injection of empty adenovirus (Ad-C) was used as a control4e8. Mice were sacrificed 21 days after the first IA injection for histological and biochemical analyses. Cartilage destruction was scored using Mankin's method14. Cartilage sections were immunostained for the indicated proteins using antibodies described in Supplementary Data. Primary chondrocyte culture, adenovirus infection, expression vector transfection, and reagent treatment Chondrocytes were isolated from femoral condyles and tibial plateaus of mice, as described previously4e8. The cells were maintained in Dulbecco's Modified Eagle's Medium supplemented with 10% fetal bovine serum and antibiotics. Cells on culture day 3 were infected with adenoviruses or were treated with the indicated reagents. Chondrocytes were infected for 3 h with adenoviral expression vectors for mouse HIF-2a (Ad-Epas1), NAMPT (AdNampt), SIRT1 (Ad-Sirt1) or SIRT2 (Ad-Sirt2) (Vector Biolabs, Philadelphia, PA) at the indicated multiplicity of infection (MOI), or were transfected with expression vectors for Epas1, individual Sirt expression vectors (Origene, Rockville, MD) or DEpas1 (Addgene, Cambridge, MA), a mouse Epas1 mutant containing substitutions at three hydroxylation sites (P405A, P530V, N851A), using Lipofectamine 2000 (Invitrogen). Cells were cultured in complete medium for 24 h prior to further use. To evaluate a role of 26S proteasomal pathway in the regulation of HIF-2a protein stability, Ad-Epas1 infected chondrocytes were cultured for 36 h, treated with NAMPT inhibitor (FK866) or SIRT inhibitors (nicotinamide and EX527)12,13 for additional 36 h, and the cells were exposed for the last 8 h of culture to MG132 or PS-341 to inhibit 26S proteasomal pathway or dimethyloxalylglycine (DMOG) to inhibit hydroxylase activity.
Reverse transcription-polymerase chain reaction (RT-PCR) and siRNA Conventional RT-PCR and quantitative RT-PCR (qRT-PCR) were performed as described previously4e8. PCR primers and conditions are described in Supplementary Table 1. The siRNAs for Sirt1, Sirt2, Sirt3, and Sirt4 genes were purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA). Non-silencing (scrambled) siRNA was used as a negative control. Chondrocytes cultured for 60 h were transfected for 6 h with siRNA using Lipofectamine 2000 (Invitrogen)4e8. Immunoprecipitation and Western blotting Total cell lysates were prepared in lysis buffer (150 mM NaCl, 1% NP-40, 50 mM Tris, 5 mM NaF) and used for Western blotting. For nuclear proteins, lysates were prepared using the above-described lysis buffer additionally containing 0.5% sodium deoxycholate. For detection of acetylated proteins, deacetylase inhibitors (5 mM trichostatin A, 5 mM nicotinamide) were included in lysis buffers. All lysis buffers contained a cocktail of protease inhibitors and phosphatase inhibitors. The acetylation status of HIF-2a was detected using HIF-2a immunoprecipitated from nuclear extracts, prepared using a nuclear extraction kit (Thermo Scientific, Hudson, NH). For immunoprecipitations, extracts were incubated with an anti-HIF2a antibody for 12 h at 4 C and precipitated with Dynabeads Protein G (Invitrogen). Aliquots were immunoblotted for HIF-2a and acetyl lysine. Ubiquitinated HIF-2a was detected by first lysing cells in the above lysis buffer supplemented with 0.5% sodium deoxycholate and 10 mM N-ethylmaleimide to inhibit ubiquitinconjugating enzymes. HIF-2a was then immunoprecipitated from 200 mg of total cell lysates, and proteins in immunoprecipitates were resolved by SDS-PAGE, transferred to nitrocellulose membranes, and then incubated in denaturing buffer (6 M guanidineHCl, 20 mM TriseHCl pH 7.5, 5 mM b-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride) for 30 min at 4 C. HIF-2a and ubiquitin were detected by Western blotting. Antibody sources are described in Supplementary Data. HIF-2a reporter gene assay The HIF-2a promoter-reporter gene was constructed in a pGL3 vector by inserting four tandem repeats of the HIF-2a-binding oligonucleotide, 50 -GATCGCCCTACGTGCTGTCTCA-30 , into the upstream region of the SV40 promoter. Chondrocytes were transfected for 3 h with the HIF-2a promoter-reporter gene (1 mg) and a b-galactosidase expression vector (0.1 mg) using Lipofectamine 2000. Transfected cells were infected with Ad-Epas1 for 3 h and then treated with the indicated concentrations of FK866, nicotinamide or EX527. Cells were harvested 36 h after treatment, and luciferase activity was measured and expressed relative to bgalactosidase activity to normalize for transfection efficiency5e7. Assays of NAD production and total SIRT deacetylase activity Intracellular NADþ and total NAD (NADþ and NADH) levels were determined using a colorimetric NAD/NADH quantification kit (BioVision Inc., Milpitas, CA). Briefly, chondrocytes were extracted in NAD/NADH extraction buffer. One half of the lysate was used to determine total NAD concentration; the remainder was heated to 60 C for 30 min and used to determine NADH concentration. The reactions were prepared in 96-well plates and read at an optical density of 450 nm. NADþ concentration was determined by subtracting NADH concentration from total NAD concentration. Total SIRT deacetylase activity was determined using a colorimetric assay
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009
H. Oh et al. / Osteoarthritis and Cartilage xxx (2015) 1e9
kit (Abcam 156915). Briefly, whole-cell extracts were prepared using lysis buffer (2 mM TriseHCL pH 8.0, 100 mM NaCl, 1 mM EDTA, 0.5% NP-40) supplemented with 1 mM trichostatin A, 10 mM dithiothreitol, and phosphatase and protease inhibitors. Lysates (2.5 mg) were incubated with 1 ml of SIRT substrate at 37 C for 1 h. The samples were sequentially incubated with capture and detection antibodies, and allowed to undergo a 10-min colorimetric reaction with tetra-methylbenzidine substrates. Following incubation, the reaction was stopped by the addition of H2SO4, and absorbance at 450 nm was measured using a VERSAmax microplate reader (Molecular Devices, Sunnydale, CA), with reference absorbance at 655 nm. Statistical analysis For cell-based in vitro study, a set of litters (generally 10) was sacrificed and the collected femoral condyles and tibial plateaus were used for each independent primary culture of chondrocytes. Any experimental data from this culture was counted as one independent trial. Therefore, the n numbers in each experiment represent the number of independent primary culture performed, which was used to calculate the average and standard error and perform statistical analysis. Normal distribution of acquired data was confirmed with the ShapiroeWilk test. For all qRT-PCR data and enzymatic activity assays, two-way analysis of variance (ANOVA) with post-hoc tests (LSD) was used. Significance was accepted at the 0.05 level of probability (P < 0.05). For animal-based in vivo experiments, each independent trial was conducted from an individual mouse. For each animal, at least three slides spanning the medial tibial plateaus in knee joints were used for histological analysis. The most advanced lesion in these serial sections of medial tibial plateaus was selected for further data quantification. Cartilage destruction and synovitis were scored based on discrete grading systems such as Mankin's score. Collected data from laboratory experiments on mice were analyzed using a non-parametric statistical method, the Mann-Whitney U test or KruskaleWallis H-test. Results HIF-2a activates the NAMPT-NADþ-SIRT axis in articular chondrocytes To examine whether HIF-2a regulates the NAMPT-NADþ-SIRT axis, we used primary cultured mouse articular chondrocytes overexpressing HIF-2a through Ad-Epas1 infection. Consistent with our previous report8, HIF-2a overexpression in articular chondrocytes caused upregulation of NAMPT at both transcript and protein levels [Fig. 1(A)]. Adenoviral-mediated overexpression of HIF-2a or NAMPT also caused a significant increase in NADþ and total NAD levels, effects that were markedly reduced by inhibition of NAMPT with FK866 [Fig. 1(B)]. Overexpression of HIF-2a or NAMPT also increased total SIRT deacetylase activity; this increase was also significantly reduced by inhibition of NAMPT with FK866 or SIRT with nicotinamide or EX527 [Fig. 1(C)]. Thus, our results confirm HIF-2a activation of the NAMPT-NADþ-SIRT axis in articular chondrocytes. The NAMPT-NADþ-SIRT axis regulates HIF-2a protein stability and transcriptional activity Among SIRT family members, SIRT1 is known to deacetylate HIF2a, and thereby stimulate its transcriptional activity, in Hep3B and HEK293 cells15,16. However, we found no evidence in chondrocytes for a change in acetylation status of overexpressed HIF-2a, with or without inhibitors/activators of NAMPT or SIRT [Fig. 2(A)]. In
3
contrast, HIF-2a transcriptional activity was significantly attenuated by inhibition of NAMPT with FK866 or inhibition of SIRT with nicotinamide or EX527, and was enhanced by activation of SIRT with resveratrol [Fig. 2(B)]. Western blot analysis [Fig. 2(C)] and immunofluorescence microscopy [Fig. 2(D)] revealed that changes in HIF2a transcriptional activity were correlated with HIF-2a protein levels: inhibition of NAMPT or SIRT reduced HIF-2a protein levels, whereas activation of SIRT increased them. HIF-2a mRNA levels were not affected by NAMPT or SIRT inhibitors [Fig. 2(C)], indicating post-transcriptional regulation of HIF-2a protein levels. Although resveratrol is nonspecific and not a direct activator of SIRT12,13, its effects are opposite to SIRT inhibitor EX527 or NAMPT inhibitor FK866, suggesting that its effect may due to activation of SIRT. To elucidate the mechanisms involved in regulating HIF-2a protein levels, we examined the proteasomal degradation pathway. Inhibition of the 26S proteasome pathway with MG132 caused a dramatic increase in overexpressed HIF-2a [Fig. 3(A)], indicating that HIF-2a protein levels are regulated by the 26S proteasomal degradation pathway. Additionally, inhibition of the 26S proteasome with MG132 [Fig. 3(B)] or PS-341 [Fig. 3(C)] prevented the reduction in HIF-2a protein levels caused by inhibition of NAMPT or SIRT without affecting HIF-2a mRNA levels. Furthermore, HIF-2a was ubiquitinated, an effect that was further enhanced by inhibition of NAMPT or SIRT [Fig. 3(D)]. This is consistent with the observed changes in HIF-2a protein levels. HIF-2a proteasomal degradation is regulated by prolyl-4hydroxylase domain (PHD) proteins, which hydroxylate HIF-2a leading to its ubiquitination17. Consistent with this, inhibition of hydroxylase activity with DMOG18 reversed the effect of NAMPT or SIRT on HIF-2a, restoring HIF-2a protein to levels comparable to those in controls [Fig. 4(A)]. Furthermore, whereas inhibition of NAMPT or SIRT caused proteasomal degradation of wild-type HIF2a [Fig. 4(B)], HIF-2a mutated at hydroxylation sites (DEpas1) was not degraded following inhibition of SIRT with nicotinamide or EX527 [Fig. 4(C)]. Interestingly, however, inhibition of NAMPT with FK866 caused effective degradation of the mutated HIF-2a [Fig. 4(C)], suggesting that SIRT, but not NAMPT, regulates hydroxylation-dependent HIF-2a degradation. In unstimulated chondrocytes, mRNA levels of the PHD2 isoform were the most abundant, whereas expression of PHD3 was minimal [Fig. 4(D)]. HIF-2a overexpression did not modulate the expression of PHD1. PHD2 expression was slightly increased by HIF-2a overexpression, whereas PHD3 expression was significantly enhanced [Fig. 4(D)]. The increase in PHD3 protein levels induced by HIF-2a overexpression was reduced by inhibition of NAMPT with FK866 or SIRT with nicotinamide or EX527 [Fig. 4(D)]. This increase in PHD3 levels following HIF-2a overexpression is consistent with previous reports of negative feedback regulation during proteasomal degradation of HIF-a19,20. In contrast to PHD3, PHD2 protein levels were markedly increased by inhibition of NAMPT with FK866 or inhibition of SIRT with nicotinamide or EX527 [Fig. 4(D)]. The increase in PHD2 protein is likely involved in the increased hydroxylation, and thereby enhanced proteasomal degradation, of HIF-2a. Indeed, other reports indicate that PHD2 is involved in HIF-a hydroxylation20,21. Therefore, it is likely that PHD2 is the main isoform that regulates HIF-2a hydroxylation in articular chondrocytes. Taken together, these results indicate that NAMPT and SIRT activities are necessary for HIF-2a protein stability by virtue of their ability to negatively regulate hydroxylase activity. SIRT activity is necessary for the OA pathogenesis caused by HIF-2a and NAMPT The significance of NAMPT and SIRT regulation of HIF-2a stability and transcriptional activity in OA pathogenesis was
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009
4
H. Oh et al. / Osteoarthritis and Cartilage xxx (2015) 1e9
Fig. 1. HIF-2a activates the NAMPT-NADþ-SIRT axis. (A) Representative RT-PCR gels and Western blots from more than six independent experiments using different sets of primary cultures of chondrocytes untreated (None) or infected with empty virus (Ad-C; 800 MOI) or Ad-Epas1 at the indicated MOI. (B and C) Chondrocytes were left untreated (None) or were infected with Ad-C (800 MOI) or Ad-Epas1 or Ad-Nampt at the indicated MOI, and incubated with or without FK866 (100 mM), nicotinamide (NIC; 15 mM) or EX527 (100 mM). NADþ and total NAD levels (B) and total SIRT activity (C) were determined (n 6). Values are presented as means ± S.E.M (*P < 0.01, **P < 0.001).
monitored by IA injection of SIRT inhibitors during the pathogenesis of experimental OA induced by Ad-Epas1 or Ad-Nampt injection. Similar to the inhibition of cartilage destruction by FK8668, IA injection of the SIRT inhibitor nicotinamide or EX527 significantly blocked the induction of cartilage destruction by Ad-Epas1 or AdNampt [Fig. 5(A)]. Immunostaining of cartilage sections revealed that the increases in MMP3 and MMP13 protein levels induced by Ad-Epas1 or Ad-Nampt were also blocked by IA co-injection of EX527 [Fig. 5(B)]. Consistent with this, qRT-PCR analyses revealed that inhibition of SIRT with nicotinamide or EX527 blocked the upregulation of matrix-degrading enzymes (MMP3, MMP9, MMP12, and MMP13) in primary cultured chondrocytes induced by HIF-2a or NAMPT [Fig. 5(C)]. IA injection of the SIRT inhibitor also
blocked the synovial inflammation induced by Ad-Epas1 or AdNampt [Supplementary Fig. 1]. Taken together, these results indicate that SIRT activity is necessary for the OA cartilage destruction caused by HIF-2a and NAMPT. Differential role of SIRT isoforms in the regulation of HIF-2a stability and transcriptional activity We next determined which SIRT isoforms are involved in HIF-2a regulation. Among SIRT family members, SIRT2 was found to be the most abundant in chondrocytes [Fig. 6(A)]. Although HIF-2a and NAMPT activated SIRT deacetylase activity [Fig. 1(C)], mRNA levels of SIRT family members were not changed by overexpression of
Fig. 2. The NAMPT-NADþ-SIRT axis regulates the stability and transcriptional activity of HIF-2a. (A) Chondrocytes were left untreated (None) or were infected with Ad-C or AdEpas1 at the indicated MOI and then treated with FK866 (FK; 100 mM), nicotinamide (NIC; 15 mM), resveratrol (Res; 100 mM), Ad-Nampt (Ad-N; 800 MOI), or MG132 (MG; 1 mM) for the last 8 h of culture. HIF-2a and acetylated lysine (Ac-Lys) were detected by Western analysis of HIF-2a immunoprecipitates. (BeD) Chondrocytes were infected with Ad-C or AdEpas1 (800 MOI) and then treated with the indicated concentrations of FK866, nicotinamide (NIC), EX527, or resveratrol (Res) for 36 h. HIF-2a reporter gene activity (B), mRNA and protein levels (C), and HIF-2a immunofluorescence microscopy (D) are shown. Representative images are presented in A, C and D from more than six independent experiments using different sets of primary cultures. Values in B are presented as means ± S.E.M (n 6; *P < 0.0005, **P < 0.00005).
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009
H. Oh et al. / Osteoarthritis and Cartilage xxx (2015) 1e9
5
Fig. 3. The NAMPT-SIRT pathway regulates proteasomal degradation of HIF-2a. (A) Chondrocytes were infected with Ad-C or Ad-Epas1 at an MOI of 800. HIF-2a protein was detected by Western blotting in chondrocytes treated with MG132 for 8 h. (B and C) Chondrocytes were infected with Ad-Epas1 for 36 h and then treated with FK866, nicotinamide (NIC), or EX527 for additional 36 h. MG132 (B) or PS-341 (C) was included for the last 8 h of culture to inhibit the 26S proteasomal pathway. HIF-2a mRNA and protein levels were detected by RT-PCR and Western blotting, respectively. (D) Chondrocytes were left untreated (None) or were infected with Ad-C or Ad-Epas1 at an MOI of 800 for 16 h in the presence or absence of the indicated concentrations of FK866 or nicotinamide (NIC), and then treated with MG132 for an additional 8 h. HIF-2a and ubiquitinated (Ub) HIF-2a were detected by Western analysis of HIF-2a immunoprecipitates. Representative images are presented from more than six independent experiments using different sets of primary cultures.
Fig. 4. The SIRT pathway regulates HIF-2a degradation via prolyl-4-hydroxylase. (A) HIF-2a mRNA and protein levels were detected in untreated chondrocytes (None) or chondrocytes infected with Ad-C or Ad-Epas1 (800 MOI) for 36 h in the absence or presence of FK866, nicotinamide (NIC), or EX527. DMOG at the indicated concentrations was added to inhibit hydroxylase activity for additional 36 h. (B) Chondrocytes were left untreated (None) or were infected with Ad-C or Ad-Epas1 (800 MOI) in the absence or presence of FK866, nicotinamide (NIC), or EX527. HIF-2a and lamin B protein levels were detected by Western blotting. (C) Chondrocytes were left untreated (None) or were transfected with empty vector (EV) or vector expressing hydroxylation sites-mutated HIF-2a (Myc-tagged DEpas1) in the absence or presence of the indicated concentrations of FK866, nicotinamide (NIC), or EX527. Myc-tagged HIF-2a and lamin B protein levels were detected by Western blotting. (D) Left: PHD mRNA levels determined by qRT-PCR in chondrocytes infected with Ad-C or Ad-Epas1 (800 MOI) for 24 h. Right: Western blot detection of PHD isoforms in chondrocytes infected with Ad-Epas1 in the absence or presence of FK866, nicotinamide (NIC), or EX527 (right). Representative images are presented from more than seven independent experiments using different sets of primary cultures. Values in D are presented as means ± S.E.M (n ¼ 8; *P < 0.001 vs Ad-C).
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009
6
H. Oh et al. / Osteoarthritis and Cartilage xxx (2015) 1e9
Fig. 5. SIRT activity is necessary for HIF-2a- and NAMPT-induced expression of matrix-degrading enzymes and OA pathogenesis. (A and B) Mice were IA-injected with Ad-C, AdEpas1 or Ad-Nampt (1 109 PFU), and incubated with or without nicotinamide (NIC; 300 mg/kg) or EX527 (1.25 mg/kg). Representative images of cartilage sections stained with safranin-O (left), and Mankin's method (right) from 12 different mice (A). The indicated proteins were detected by immunostaining cartilage sections (n 6). (B and C) Chondrocytes were left untreated (None) or were infected with Ad-C, Ad-Epas1, or Ad-Nampt at an MOI of 800 in the absence or presence of the indicated concentrations of nicotinamide (NIC; left) or EX527 (right). Expression levels of the indicated mRNAs were quantified by qRT-PCR from more than six independent experiments using different sets of primary cultures. Values are presented as means ± S.E.M (*P < 0.01, **P < 0.001; NS, not significant). Scale bar: 100 mm.
HIF-2a or NAMPT [Fig. 6(A)]. To elucidate which SIRT isoforms were involved in the regulation of HIF-2a, we overexpressed individual SIRT isoforms in articular chondrocytes by transfection with the corresponding expression vectors. Overexpression of SIRT2 caused the most marked increase in HIF-2a protein stability and transcriptional activity [Fig. 6(B)]. Overexpression of SIRT4 also increased HIF-2a protein stability and transcriptional activity, albeit to a lesser degree [Fig. 6(B)]. In contrast, SIRT3 overexpression caused a decrease in HIF-2a stability and reporter gene activity, whereas overexpression of SIRT1, SIRT5, SIRT6, or SIRT7 had no effect on HIF-2a stability or transcriptional activity [Fig. 6(B)]. These results indicate that SIRT members display isoform-specific regulation of HIF-2a stability and transcriptional activity. To further elucidate the roles of SIRT isoforms in the regulation of HIF-2a, we used specific siRNAs to knockdown each SIRT isoform. HIF-2a transcriptional activity was significantly reduced by Sirt2 knockdown and, to a lesser degree, by Sirt4 knockdown, whereas knockdown of Sirt3 increased it [Fig. 6(C)]. Changes in HIF-2a transcriptional activity were correlated with HIF-2a protein levels: knockdown of Sirt2 or Sirt4 reduced overexpressed HIF-2a protein levels, whereas knockdown of Sirt3 increased them, without affecting HIF-2a mRNA levels [Fig. 6(C)]. Taken together, these results suggest that both SIRT2 and SIRT4 appear to enhance HIF-2a protein stability and transcriptional activity in chondrocytes.
SIRT2 regulates HIF-2a- and NAMPT-induced OA cartilage destruction Based on the capacity of SIRT2 to regulate HIF-2a protein stability, we examined the role of SIRT2 in HIF-2a- and NAMPTinduced OA cartilage destruction. As expected11, overexpressed SIRT2 was localized to the cytosol [Fig. 7(A)]. Similar to the effects of SIRT2 overexpression produced by transfection with a SIRT2 expression vector, co-infection with Ad-Sirt2 and Ad-Epas1 enhanced HIF-2a protein stability and transcriptional activity
[Fig. 7(B)]. In contrast, knockdown of Sirt2 in chondrocytes by infection with Ad-shSirt2 reduced HIF-2a protein levels and transcriptional activity [Fig. 7(C)]. Overexpression of SIRT2 alone in joint tissues through IA injection of Ad-Sirt2 in mice did not cause cartilage destruction [Fig. 7(D)]. However, IA co-injection of AdshSirt2 with Ad-Epas1 or Ad-Nampt significantly inhibited HIF-2aand NAMPT-induced cartilage destruction [Fig. 7(E)], indicating the involvement of SIRT2 in both HIF-2a- and NAMPT-induced OA cartilage destruction. Notably, conditional knockout of Sirt1 in cartilage tissue did not affect HIF-2a-induced cartilage destruction [Supplementary Fig. 2], and IA injection of Ad-Sirt1, which caused overexpression of SIRT1 in cartilage tissue, also did not cause cartilage destruction [Supplementary Fig. 2]. These results suggest that SIRT1 is not involved in HIF-2a-induced cartilage destruction.
Discussion The results presented here reveal reciprocal regulation of HIF-2a and the NAMPT-NADþ-SIRT axis in chondrocytes. We first demonstrated HIF-2a activation of the NAMPT-NADþ-SIRT axis, showing that HIF-2a upregulates NAMPT, which in turn stimulates NADþ synthesis and SIRT activation. Conversely, NAMPT/SIRT activity is necessary for HIF-2a protein stability and transcriptional activity. Inhibition of NAMPT or SIRT blocked HIF-2a-induced cartilage destruction, demonstrating that this reciprocal regulation is required for HIF-2a-induced OA pathogenesis. HIF-2a activation of the NAMPT-NADþ-SIRT axis is expected, given our previous observation that the Nampt is a direct target of HIF-2a in chondrocytes8. Furthermore, NAMPT upregulation is necessary for HIF-2a-induced expression of matrix-degrading enzymes and OA cartilage destruction8. The main function of NAMPT is to stimulate the synthesis of NADþ, which is an essential cofactor for members of the SIRT deacetylase family9,10. Indeed, overexpression of HIF-2a or NAMPT caused NADþ synthesis and SIRT activation in chondrocytes. More importantly, the HIF-2a-
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009
H. Oh et al. / Osteoarthritis and Cartilage xxx (2015) 1e9
7
Fig. 6. SIRT isoform-specific regulation of HIF-2a protein stability and transcriptional activity. (A) Chondrocytes were left untreated (None) or were infected with Ad-C, Ad-Epas1, or Ad-Nampt at an MOI of 800. mRNA levels were determined by RT-PCR (left). The relative amount of SIRT mRNA in untreated chondrocytes was normalized to that of GAPDH mRNA (right). (B) Individual SIRT isoforms were overexpressed in chondrocytes by transfection with expression vectors. Cells were infected with Ad-C or Ad-Epas1 at an MOI of 800 for 36 h. HIF-2a reporter gene activity and HIF-2a and SIRT protein levels were detected. (C) Chondrocytes were left untreated (None) or were transfected with 100 nM control siRNA (C-siRNA) or the indicated concentrations of siRNA specific for individual SIRT isoforms. Cells were infected with Ad-C or Ad-Epas1 (400 MOI) for 36 h. HIF-2a reporter gene activity and protein levels were detected. Values are presented as means ± S.E.M from more than six independent experiments using different sets of primary cultures (*P < 0.05, **P < 0.005; NS, not significant).
Fig. 7. SIRT2 regulates HIF-2a- and NAMPT-induced cartilage destruction. (A) Representative immunostaining image of SIRT2 in chondrocytes transfected with Sirt2 expressing vector from more than six independent experiments using different sets of primary cultures. (B) Chondrocytes were co-infected with Ad-Epas1 and Ad-Sirt2. Left: mRNA and protein levels of SIRT2 and HIF-2a; right: HIF-2a transcriptional activity (n ¼ 5). (C) Chondrocytes were left untreated or were infected with Ad-C or Ad-Epas1 and Ad-shSirt2. Left: mRNA and protein levels of SIRT2 and HIF-2a; right: HIF-2a transcriptional activity (n ¼ 11). (D) Safranin-O staining (left) and SIRT2 immunostaining (right) in cartilage sections of mice injected with Ad-C or Ad-Sirt2. (E) Safranin-O staining and Mankin score of cartilage sections from 14 mice IA-injected with Ad-C or Ad-Epas1 with or without Ad-shSirt2 (n ¼ 14).
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009
8
H. Oh et al. / Osteoarthritis and Cartilage xxx (2015) 1e9
stimulated NAMPT-NADþ-SIRT axis, in turn, promoted HIF-2a protein stability and transcriptional activity. Indeed, inhibition of the NAMPT/SIRT pathway promoted degradation of ectopically expressed HIF-2a, indicating that the NAMPT-SIRT pathway negatively regulates the proteasomal degradation of HIF-2a. Interestingly, SIRT regulation of HIF-2a proteasomal degradation depended on HIF-2a hydroxylation, whereas NAMPT action was hydroxylation independent. HIF-2a-simulated SIRT activity is required for HIF-2a- and NAMPT-induced OA pathogenesis, as evidenced by the fact that inhibition of SIRT activity blocked HIF-2a regulation of matrix-degrading enzyme expression in chondrocytes and OA cartilage destruction. Additionally, the catabolic functions of the NAMPT-SIRT pathway are exerted through enhanced HIF-2a stability and transcriptional activity. The mammalian SIRT family is composed of seven members (SIRT1e7) that possess NADþ-dependent deacetylase, deacylase, and ADP-ribosyltransferase activities10e12. They are found in different subcellular locations, including the nucleus (SIRT6 and SIRT7), nucleus and cytosol (SIRT1 and SIRT2), and mitochondria (SIRT3e5). Among SIRT family members, SIRT1 is the best characterized and has been shown to play a protective role in OA pathogenesis22e24. For example, SIRT1 is necessary for chondrocyte survival, and loss of SIRT1 function causes chondrocyte apoptosis25e28. SIRT1 also regulates cartilage-specific gene expression29e31. It was recently reported that chondrocyte-specific conditional knockout of Sirt1 in mice causes transiently accelerated progression of surgically induced OA32. Similarly, Sirt1þ/ mice show increased chondrocyte apoptosis and enhanced OA severity27, and mutant mice carrying a variant of SIRT1 lacking enzymatic activity show elevated rates of cartilage degradation with age33. Collectively, these observations suggest that SIRT1 activity serves a protective role in OA pathogenesis, although this capacity of SIRT1 does not appear to be especially marked, as evidenced by the observation that knockout of Sirt1 promotes only the early stage of OA progression32. In contrast to the reported protective function of SIRT1 in surgically induced OA pathogenesis, our current results demonstrate that SIRT1 is not involved in HIF-2a- or NAMPT-induced OA cartilage destruction in mice. This is demonstrated by the observation that conditional knockout of Sirt1 in cartilage tissue does not affect HIF-2a-induced cartilage destruction. Furthermore, we found that SIRT1 does not affect HIF-2a protein stability or transcriptional activity in chondrocytes. Although SIRT1 is not involved in HIF-2aor NAMPT-induced OA cartilage destruction, we demonstrated that inhibition of NADþ-dependent SIRT deacetylase activity blocked OA cartilage destruction induced by HIF-2a or NAMPT, with a concomitant inhibition of the expression of matrix-degrading enzymes. This indicates that NADþ-dependent SIRT activity promotes HIF-2a- and NAMPT-induced OA pathogenesis by virtue of its capacity to regulate HIF-2a protein stability and transcriptional activity. Among SIRT family members, SIRT2, which is the most abundant in chondrocytes, exerted marked effects on HIF-2a protein stability and transcriptional activity. SIRT2 exhibits deacetylase activity and localizes mainly to the cytosol10e12. This is consistent with the regulatory mechanisms of HIF-2a protein stability. HIF-2a protein is degraded by 26S proteasomal pathway in the cytosol, resulting in minimal transcriptional activity. However, under pathological conditions (i.e., under hypoxic condition), its degradation pathway is blocked and accumulated HIF-2a translocates into nucleus forming heterodimer with HIF-1b to regulate target gene expression17. Therefore, it is likely that the SIRT2 regulates HIF-2a protein stability and transcriptional activity, and suppression of SIRT2 activity is likely responsible for the observed inhibitory effects of SIRT inhibitors on the OA cartilage destruction caused by HIF-2a or NAMPT. Moreover, our demonstration that
knockdown of Sirt2 by IA injection of Ad-shSirt2 inhibits HIF-2aand NAMPT-induced cartilage destruction clearly indicate the role of SIRT2. In contrast to our observation that SIRT2 stabilizes HIF-2a protein without affecting its acetylation status in chondrocytes, recent report indicated that SIRT2 destabilizes HIF-1a by regulating deacetylation in tumor cells34. Despite many similarities between HIF-1a and HIF-2a, these two isoforms show different sensitivity to oxygen tension and display distinct, and sometimes opposing, cellular activities17. Therefore, it is likely that protein stability of HIF-1a and HIF-2a are differentially regulated by SIRT2 depending on cell types. In addition to SIRT2, our results indicate that mitochondrial SIRT3 and SIRT4 also regulate HIF-2a stability. Similar to SIRT2, overexpression of SIRT4 enhanced HIF-2a stability, whereas knockdown of Sirt4 destabilizes HIF-2a. To the best of our knowledge, this is the first evidence that SIRT4 regulates HIF-2a stability. However, it remains to be elucidated whether SIRT4 is also associated with OA pathogenesis caused by HIF-2a and NAMPT. Contrast to cytosolic SIRT2 and mitochondrial SIRT4, SIRT3 which is a major mitochondrial NADþ-dependent deacetylase destabilized HIF-2a stability in chondrocytes. Although we could not find any previous report for the regulation of HIF-2a by SIRT3, many reports indicated that SIRT3 destabilizes HIF-1a in tumor cells35,36. Therefore, the significance of SIRT3 regulation of HIF-2a stability in OA pathogenesis remains to be elucidated. Furthermore, it is of interesting to reveal possible orchestration of SIRT isoforms in the regulation of HIF-2a stability. In summary, we demonstrated reciprocal regulation of HIF-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes. HIF-2a activates the NAMPT-NADþ-SIRT axis, which, in turn, promotes HIF2a protein stability resulting in increased HIF-2a transcriptional activity. We also revealed that many SIRT isoforms, including SIRT2, SIRT3, and SIRT4, are associated with HIF-2a stability regulation. Among them, SIRT2 and SIRT4 are positively associated with HIF-2a stability in chondrocytes. This reciprocal regulation is involved in the expression of catabolic MMPs and OA cartilage destruction caused by HIF-2a or NAMPT.
Author contributions OH and JSK: Study design, data acquisition, data analysis and interpretation, manuscript preparation and approval; SY, MKG, JHK, JR, SKK, HEK and JHR:data acquisition, analysis, and interpretation, and manuscript approval; JSC: funding acquisition, study design, data interpretation,manuscript preparation and approval. JSC ([email protected]) takes responsibility for the integrity of this work.
Conflicts of interest The authors have no conflicts of interest.
Acknowledgment This work was supported by grants from the National Research Foundation of Korea (2007-0056157 and 2013R1A2A1A01009713), the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare (HI14C3484), and the Integrative Aging Research Center of Gwangju Institute of Science and Technology. The funders had no role in study design, conduct of research, or preparation of the manuscript. The content of this manuscript is solely the responsibility of the authors.
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009
H. Oh et al. / Osteoarthritis and Cartilage xxx (2015) 1e9
Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.joca.2015.07.009.
20.
References 1. Heinegard D, Saxne T. The role of the cartilage matrix in osteoarthritis. Nat Rev Rheumatol 2011;7:50e6. 2. Bian Q, Wang YJ, Liu SF, Li YP. Osteoarthritis: genetic factors, animal models, mechanisms, and therapies. Front Biosci 2012;E4:74e100. 3. Little CB, Hunter DJ. Post-traumatic osteoarthritis: from mouse models to clinical trials. Nat Rev Rheumatol 2013;9:485e97. 4. Kim JH, Jeon J, Shin M, Won Y, Lee M, Lee G, et al. Regulation of the catabolic cascade in osteoarthritis by the zinc-ZIP8-MTF1 axis. Cell 2014;156:730e43. 5. Yang S, Kim J, Ryu JH, Oh H, Chun CH, Kim BJ, et al. Hypoxiainducible factor-2a is a catabolic regulator of osteoarthritic cartilage destruction. Nat Med 2010;16:687e93. 6. Ryu JH, Shin Y, Huh YH, Yang S, Chun CH, Chun JS. Hypoxiainducible factor-2a regulates fas-mediated chondrocyte apoptosis during osteoarthritic cartilage destruction. Cell Death Differ 2012;19:440e50. 7. Ryu JH, Yang S, Shin Y, Rhee J, Chun CH, Chun JS. Interleukin-6 plays an essential role in hypoxia-inducible factor 2a-induced experimental osteoarthritic cartilage destruction in mice. Arthritis Rheum 2011;63:2732e43. 8. Yang S, Ryu JH, Oh H, Jeon J, Kwak JS, Kim JH, et al. NAMPT (visfatin), a direct target of hypoxia-inducible factor-2a, is an essential catabolic regulator of osteoarthritis. Ann Rheum Dis 2015;74:595e602. 9. Zhang LQ, Heruth DP, Ye SQ. Nicotinamide phosphoribosyltransferase in human diseases. J Bioanal Biomed 2011;3: 13e25. 10. Imai SI. Nicotinamide phosphoribosyl transferase (NAMPT): a link between NAD biology, metabolism, and diseases. Curr Pharm Des 2009;15:20e8. 11. Morris BJ. Seven sirtuins for seven deadly diseases of aging. Free Radic Biol Med 2013;56:133e71. 12. Carafa V, Nebbioso A, Altucci L. Sirtuins and disease: the road ahead. Front Pharmacol 2012;3:1e4. 13. Lavu S, Boss O, Elliott PJ, Lambert PD. Sirtuins-novel therapeutic targets to treat age-associated diseases. Nat Rev Drug Discov 2008;7:841e53. 14. Mankin HJ, Dorfman H, Lippiello L, Zarins A. Biochemical and metabolic abnormalities in articular cartilage from osteoarthritic human hips. II. Correlation of morphology with biochemical and metabolic data. J Bone Joint Surg Am 1971;53:523e37. 15. Dioum EM, Chen R, Alexander MS, Zhang Q, Hogg RT, Gerard RD, et al. Regulation of hypoxia-inducible factor-2a signaling by the stress-responsive deacetylase sirtuin 1. Science 2009;324:1289e93. 16. Chen R, Dioum EM, Hogg RT, Gerard RD, Garcia JA. Hypoxia increases sirtuin 1 expression in a hypoxia-inducible factordependent manner. J Biol Chem 2011;286:13869e78. 17. Greer SN, Metcalf JL, Wang Y, Ohh M. The updated biology of hypoxia-inducible factor. EMBO J 2012;31:2448e60. 18. Fujita N, Chiba K, Shapiro IM, Risbud MV. HIF-1a and HIF-2a degradation is differentially regulated in nucleus pulpous cells of the intervertebral disc. J Bone Min Res 2012;27:401e12. 19. Minamishima YA, Moslehi J, Padera RF, Bronson RT, Liao R, Kaelin Jr WG. A feedback loop involving the Phd3 prolyl
21. 22. 23. 24. 25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
9
hydroxylase tunes the mammalian hypoxic response in vivo. Mol Cell Biol 2009;29:5729e41. Fujita N, Markova D, Anderson DG, Chiba K, Toyama Y, Shapiro IM, et al. Expression of prolyl hydroxylases (PHDs) is selectively controlled by HIF-1 and HIF-2 proteins in nucleus pulposus cells of the intervertebral disc: distinct roles of PHD2 and PHD3 proteins in controlling HIF-1a activity in hypoxia. J Biol Chem 2012;287:16975e86. Minamishima YA, Kaelin Jr WG. Reactivation of hepatic EPO synthesis in mice after PHD loss. Science 2010;329:407. Asahara H. Switching gears to an arthritis gene expression network by SirT1 cleavage. Arthritis Rheum 2011;63:2161e3. Gabay O, Sanchez C. Epigenetics, sirtuins and osteoarthritis. Joint Bone Spine 2012;79:570e3. Dvir-Ginzberg M, Steinmeyer J. Towards elucidating the role of SirT1 in osteoarthritis. Front Biosci 2013;18:343e55. Takayama K, Ishida K, Matsushita T, Fujita N, Hayashi S, Sasaki K, et al. SIRT1 regulation of apoptosis of human chondrocytes. Arthritis Rheum 2009;60:2731e40. Gagarina V, Gaybay O, Dvir-Ginzberg M, Lee EJ, Brady JK, Quon MJ, et al. SirT1 enhances survival of human osteoarthritic chondrocytes by repressing protein tyrosine phosphatase 1B and activating the insulin-like growth factor receptor pathway. Arthritis Rheum 2010;62:1383e92. Gaybay O, Oppenhiemer H, Meir H, Zaal K, Sanchez C, DvirGinzberg M. Increased apoptotic chondrocytes in articular cartilage from adult heterozygous SirT1 mice. Ann Rheum Dis 2012;71:613e6. Oppenheimer H, Gaybay O, Meir H, Haze A, Kandel L, Lieberggall M, et al. 75-kd sirtuin 1 blocks tumor necrosis factor a-mediated apoptosis in human osteoarthritic chondrocytes. Arthritis Rheum 2012;64:718e28. Dvir-Ginzberg M, Gagarina V, Lee EJ, Hall DJ. Regulation of cartilage-specific gene expression in human chondrocytes by SirT1 and nicotinamide phosphoribosyltransferase. J Biol Chem 2008;283:36300e10. Fujita N, Matsushita T, Ishida K, Kubo S, Matsumoto T, Takayama K, et al. Potential involvement of SIRT1 in the pathogenesis of osteoarthritis through the modulation of chondrocyte gene expressions. J Orthop Res 2011;29: 511e5. Dvir-Ginzberg M, Gagarina V, Lee EJ, Booth R, Gaybay O, Hall DJ. Tumor necrosis factor a-mediated cleavage and inactivation of SirT1 in human osteoarthritic chondrocytes. Arthritis Rheum 2011;63:2363e73. Matsuzaki T, Matsushita T, Takayama K, Matsumoto T, Nishida K, Kuroda R, et al. Disruption of Sirt1 in chondrocytes causes accelerated progression of osteoarthritis under mechanical stress and during ageing in mice. Ann Rheum Dis 2014;73:1397e404. Gabay O, Sanchez C, Dvir-Ginzberg M, Gaqarina V, Zaal KJ, Song Y, et al. Sirt1 enzymatic activity is required for cartilage homeostasis in vivo. Arthritis Rheum 2013;65:159e66. Seo KS, Park JH, Heo JY, Jing K, Han J, Min KN, et al. SIRT2 regulates tumor hypoxia response by promoting HIF-1a hydroxylation. Oncogene 2015;34:1354e62. Finley LW, Carracedo A, Lee J, Souza A, Egia A, Zhang J, et al. SIRT3 opposes reprogramming of cancer cell metabolism through HIF1a destabilization. Cancer Cell 2011;1: 416e28. Bell EL, Emerling BM, Ricoult SJ, Guarente L. SirT3 suppresses hypoxia inducible factor 1a and tumor growth by inhibiting mitochondrial ROS production. Oncogene 2011;30:2986e96.
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009
Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPTNADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis H. Oh y a, J.-S. Kwak y a, S. Yang y, M.-K. Gong y, J.-H. Kim y z, J. Rhee y, S.K. Kim y, H.-E. Kim y, J.-H. Ryu y x, J.-S. Chun y * y School of Life Sciences, Gwangju Institute of Science and Technology, Gwangju 500-712, South Korea z Department of Biological Sciences, Seoul National University, Seoul 151-747, South Korea x Research Center for Biomineralization Disorders and Dental Science Research Institute, School of Dentistry, Chonnam National University, Gwangju 500757, South Korea
a r t i c l e i n f o
s u m m a r y
Article history: Received 27 April 2015 Accepted 15 July 2015
Objective: Hypoxia-inducible factor-2a (HIF-2a) transcriptionally upregulates Nampt in articular chondrocytes. NAMPT, which exhibits nicotinamide phosphoribosyltransferase activity, in turn causes osteoarthritis (OA) in mice by stimulating the expression of matrix-degrading enzymes. Here, we sought to elucidate whether HIF-2a activates the NAMPT-NADþ-SIRT axis in chondrocytes and thereby contributes to the pathogenesis of OA. Methods: Assays of NAD levels, SIRT activity, reporter gene activity, mRNA, and protein levels were conducted in primary cultured mouse articular chondrocytes. Experimental OA in mice was induced by intra-articular (IA) injection of adenovirus expressing HIF-2a (Ad-Epas1) or NAMPT (Ad-Nampt). The functions of SIRT in OA were examined by IA co-injection of SIRT inhibitors or adenovirus expressing individual SIRT isoforms or shRNA targeting specific SIRT isoforms. Results: HIF-2a activated the NAMPT-NADþ-SIRT axis in chondrocytes by upregulating NAMPT, which stimulated NADþ synthesis and thereby activated SIRT family members. The activated NAMPT-SIRT pathway, in turn, promoted HIF-2a protein stability by negatively regulating its hydroxylation and 26S proteasome-mediated degradation, resulting in increased HIF-2a transcriptional activity. Among SIRT family members (SIRT1e7), SIRT2 and SIRT4 were positively associated with HIF-2a stability and transcriptional activity in chondrocytes. This reciprocal regulation was required for the expression of catabolic matrix metalloproteinases (MMP3, MMP12, and MMP13) and OA cartilage destruction caused by IA injection of Ad-Epas1 Ad-Nampt. Conclusion: The reciprocal regulation of HIF-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in OA cartilage destruction caused by HIF-2a or NAMPT. © 2015 The Authors. Published by Elsevier Ltd and Osteoarthritis Research Society International. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/ 4.0/).
Keywords: Hypoxia-inducible factor (HIF) NADþ biosynthesis Sirtuin Chondrocytes Osteoarthritis
Introduction Osteoarthritis (OA) is caused by mechanical stresses, including joint instability and injury, and factors that predispose toward OA, such as aging. These factors lead to the activation of biochemical pathways in chondrocytes that result in degradation of the
* Address correspondence and reprint requests to: J.-S. Chun, School of Life Sciences, Gwangju Institute of Science and Technology, Buk-Gu, Gwangju 500-712, South Korea. Tel: 82-62-715-2497; Fax: 82-62-715-3304. E-mail address: [email protected] (J.-S. Chun). a These authors contributed equally to this work.
extracellular matrix by matrix metalloproteinases (MMPs) and aggrecanases (ADAMTS)1e3. We have previously shown that among potentially relevant biochemical pathways, the zinc-ZIP8-MTF1 axis4 and hypoxia-inducible factor (HIF)-2a (encoded by Epas1)5e7 are important catabolic regulators of OA cartilage destruction that act by upregulating chondrocyte expression of matrix-degrading enzymes. Our previous study also demonstrated that the Nampt is a direct target gene of HIF-2a in OA chondrocytes8. NAMPT functions as both an intracellular form (iNAMPT) and an extracellular form (eNAMPT or visfatin). eNAMPT acts as an adipokine, whereas the nicotinamide phosphoribosyltransferase enzymatic activity of iNAMPT regulates salvage pathways of NADþ synthesis9,10. Both
http://dx.doi.org/10.1016/j.joca.2015.07.009 1063-4584/© 2015 The Authors. Published by Elsevier Ltd and Osteoarthritis Research Society International. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009
2
H. Oh et al. / Osteoarthritis and Cartilage xxx (2015) 1e9
iNAMPT and eNAMPT upregulate expression of the matrixdegrading enzymes MMP3, MMP12, and MMP13 in chondrocytes, an action that is necessary for HIF-2a-induced OA cartilage destruction8. Gain-of-function studies involving intra-articular (IA) injection of an adenoviral expression construct for NAMPT (AdNampt) and those using Col2a1-Nampt transgenic mice, as well as loss-of-function studies (i.e., IA injection of the NAMPT inhibitor FK866), have demonstrated that iNAMPT is an essential catabolic regulator of OA pathogenesis8. NAMPT enzymatic activity stimulates the synthesis of NADþ, which is an essential cofactor of sirtuin (SIRT) protein deacetylases9,10. SIRT regulates gene expression by modulating chromatin function through direct deacetylation of histones and transcription factors10e12. Therefore, it is likely that HIF-2a-induced NAMPT expression causes NADþ synthesis and thereby stimulates SIRT activity. The mammalian SIRT family is composed of seven members, SIRT1e7, with different subcellular localization patterns and enzymatic activities11. In this study, we examined whether HIF-2a regulates the NAMPT-NADþ-SIRT axis and thereby contributes to the pathogenesis of OA. We report that reciprocal regulation of HIF-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in the OA cartilage destruction caused by HIF-2a or NAMPT. Materials and methods Experimental OA and histology All animal experiments were approved by the Gwangju Institute of Science and Technology Animal Care and Use Committee. Experimental OA was induced by IA injection in 10-week-old male mice (once weekly for 3 weeks) of adenovirus expressing HIF-2a (Ad-Epas1) or NAMPT (Ad-Nampt) (1 109 plaque forming units [PFUs]), with or without the SIRT inhibitor nicotinamide (300 mg/kg body weight) or EX527 (1.25 mg/kg body weight)12,13, or adenovirus expressing SIRT1 (Ad-Sirt1) or SIRT2 (Ad-Sirt2); IA injection of empty adenovirus (Ad-C) was used as a control4e8. Mice were sacrificed 21 days after the first IA injection for histological and biochemical analyses. Cartilage destruction was scored using Mankin's method14. Cartilage sections were immunostained for the indicated proteins using antibodies described in Supplementary Data. Primary chondrocyte culture, adenovirus infection, expression vector transfection, and reagent treatment Chondrocytes were isolated from femoral condyles and tibial plateaus of mice, as described previously4e8. The cells were maintained in Dulbecco's Modified Eagle's Medium supplemented with 10% fetal bovine serum and antibiotics. Cells on culture day 3 were infected with adenoviruses or were treated with the indicated reagents. Chondrocytes were infected for 3 h with adenoviral expression vectors for mouse HIF-2a (Ad-Epas1), NAMPT (AdNampt), SIRT1 (Ad-Sirt1) or SIRT2 (Ad-Sirt2) (Vector Biolabs, Philadelphia, PA) at the indicated multiplicity of infection (MOI), or were transfected with expression vectors for Epas1, individual Sirt expression vectors (Origene, Rockville, MD) or DEpas1 (Addgene, Cambridge, MA), a mouse Epas1 mutant containing substitutions at three hydroxylation sites (P405A, P530V, N851A), using Lipofectamine 2000 (Invitrogen). Cells were cultured in complete medium for 24 h prior to further use. To evaluate a role of 26S proteasomal pathway in the regulation of HIF-2a protein stability, Ad-Epas1 infected chondrocytes were cultured for 36 h, treated with NAMPT inhibitor (FK866) or SIRT inhibitors (nicotinamide and EX527)12,13 for additional 36 h, and the cells were exposed for the last 8 h of culture to MG132 or PS-341 to inhibit 26S proteasomal pathway or dimethyloxalylglycine (DMOG) to inhibit hydroxylase activity.
Reverse transcription-polymerase chain reaction (RT-PCR) and siRNA Conventional RT-PCR and quantitative RT-PCR (qRT-PCR) were performed as described previously4e8. PCR primers and conditions are described in Supplementary Table 1. The siRNAs for Sirt1, Sirt2, Sirt3, and Sirt4 genes were purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA). Non-silencing (scrambled) siRNA was used as a negative control. Chondrocytes cultured for 60 h were transfected for 6 h with siRNA using Lipofectamine 2000 (Invitrogen)4e8. Immunoprecipitation and Western blotting Total cell lysates were prepared in lysis buffer (150 mM NaCl, 1% NP-40, 50 mM Tris, 5 mM NaF) and used for Western blotting. For nuclear proteins, lysates were prepared using the above-described lysis buffer additionally containing 0.5% sodium deoxycholate. For detection of acetylated proteins, deacetylase inhibitors (5 mM trichostatin A, 5 mM nicotinamide) were included in lysis buffers. All lysis buffers contained a cocktail of protease inhibitors and phosphatase inhibitors. The acetylation status of HIF-2a was detected using HIF-2a immunoprecipitated from nuclear extracts, prepared using a nuclear extraction kit (Thermo Scientific, Hudson, NH). For immunoprecipitations, extracts were incubated with an anti-HIF2a antibody for 12 h at 4 C and precipitated with Dynabeads Protein G (Invitrogen). Aliquots were immunoblotted for HIF-2a and acetyl lysine. Ubiquitinated HIF-2a was detected by first lysing cells in the above lysis buffer supplemented with 0.5% sodium deoxycholate and 10 mM N-ethylmaleimide to inhibit ubiquitinconjugating enzymes. HIF-2a was then immunoprecipitated from 200 mg of total cell lysates, and proteins in immunoprecipitates were resolved by SDS-PAGE, transferred to nitrocellulose membranes, and then incubated in denaturing buffer (6 M guanidineHCl, 20 mM TriseHCl pH 7.5, 5 mM b-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride) for 30 min at 4 C. HIF-2a and ubiquitin were detected by Western blotting. Antibody sources are described in Supplementary Data. HIF-2a reporter gene assay The HIF-2a promoter-reporter gene was constructed in a pGL3 vector by inserting four tandem repeats of the HIF-2a-binding oligonucleotide, 50 -GATCGCCCTACGTGCTGTCTCA-30 , into the upstream region of the SV40 promoter. Chondrocytes were transfected for 3 h with the HIF-2a promoter-reporter gene (1 mg) and a b-galactosidase expression vector (0.1 mg) using Lipofectamine 2000. Transfected cells were infected with Ad-Epas1 for 3 h and then treated with the indicated concentrations of FK866, nicotinamide or EX527. Cells were harvested 36 h after treatment, and luciferase activity was measured and expressed relative to bgalactosidase activity to normalize for transfection efficiency5e7. Assays of NAD production and total SIRT deacetylase activity Intracellular NADþ and total NAD (NADþ and NADH) levels were determined using a colorimetric NAD/NADH quantification kit (BioVision Inc., Milpitas, CA). Briefly, chondrocytes were extracted in NAD/NADH extraction buffer. One half of the lysate was used to determine total NAD concentration; the remainder was heated to 60 C for 30 min and used to determine NADH concentration. The reactions were prepared in 96-well plates and read at an optical density of 450 nm. NADþ concentration was determined by subtracting NADH concentration from total NAD concentration. Total SIRT deacetylase activity was determined using a colorimetric assay
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009
H. Oh et al. / Osteoarthritis and Cartilage xxx (2015) 1e9
kit (Abcam 156915). Briefly, whole-cell extracts were prepared using lysis buffer (2 mM TriseHCL pH 8.0, 100 mM NaCl, 1 mM EDTA, 0.5% NP-40) supplemented with 1 mM trichostatin A, 10 mM dithiothreitol, and phosphatase and protease inhibitors. Lysates (2.5 mg) were incubated with 1 ml of SIRT substrate at 37 C for 1 h. The samples were sequentially incubated with capture and detection antibodies, and allowed to undergo a 10-min colorimetric reaction with tetra-methylbenzidine substrates. Following incubation, the reaction was stopped by the addition of H2SO4, and absorbance at 450 nm was measured using a VERSAmax microplate reader (Molecular Devices, Sunnydale, CA), with reference absorbance at 655 nm. Statistical analysis For cell-based in vitro study, a set of litters (generally 10) was sacrificed and the collected femoral condyles and tibial plateaus were used for each independent primary culture of chondrocytes. Any experimental data from this culture was counted as one independent trial. Therefore, the n numbers in each experiment represent the number of independent primary culture performed, which was used to calculate the average and standard error and perform statistical analysis. Normal distribution of acquired data was confirmed with the ShapiroeWilk test. For all qRT-PCR data and enzymatic activity assays, two-way analysis of variance (ANOVA) with post-hoc tests (LSD) was used. Significance was accepted at the 0.05 level of probability (P < 0.05). For animal-based in vivo experiments, each independent trial was conducted from an individual mouse. For each animal, at least three slides spanning the medial tibial plateaus in knee joints were used for histological analysis. The most advanced lesion in these serial sections of medial tibial plateaus was selected for further data quantification. Cartilage destruction and synovitis were scored based on discrete grading systems such as Mankin's score. Collected data from laboratory experiments on mice were analyzed using a non-parametric statistical method, the Mann-Whitney U test or KruskaleWallis H-test. Results HIF-2a activates the NAMPT-NADþ-SIRT axis in articular chondrocytes To examine whether HIF-2a regulates the NAMPT-NADþ-SIRT axis, we used primary cultured mouse articular chondrocytes overexpressing HIF-2a through Ad-Epas1 infection. Consistent with our previous report8, HIF-2a overexpression in articular chondrocytes caused upregulation of NAMPT at both transcript and protein levels [Fig. 1(A)]. Adenoviral-mediated overexpression of HIF-2a or NAMPT also caused a significant increase in NADþ and total NAD levels, effects that were markedly reduced by inhibition of NAMPT with FK866 [Fig. 1(B)]. Overexpression of HIF-2a or NAMPT also increased total SIRT deacetylase activity; this increase was also significantly reduced by inhibition of NAMPT with FK866 or SIRT with nicotinamide or EX527 [Fig. 1(C)]. Thus, our results confirm HIF-2a activation of the NAMPT-NADþ-SIRT axis in articular chondrocytes. The NAMPT-NADþ-SIRT axis regulates HIF-2a protein stability and transcriptional activity Among SIRT family members, SIRT1 is known to deacetylate HIF2a, and thereby stimulate its transcriptional activity, in Hep3B and HEK293 cells15,16. However, we found no evidence in chondrocytes for a change in acetylation status of overexpressed HIF-2a, with or without inhibitors/activators of NAMPT or SIRT [Fig. 2(A)]. In
3
contrast, HIF-2a transcriptional activity was significantly attenuated by inhibition of NAMPT with FK866 or inhibition of SIRT with nicotinamide or EX527, and was enhanced by activation of SIRT with resveratrol [Fig. 2(B)]. Western blot analysis [Fig. 2(C)] and immunofluorescence microscopy [Fig. 2(D)] revealed that changes in HIF2a transcriptional activity were correlated with HIF-2a protein levels: inhibition of NAMPT or SIRT reduced HIF-2a protein levels, whereas activation of SIRT increased them. HIF-2a mRNA levels were not affected by NAMPT or SIRT inhibitors [Fig. 2(C)], indicating post-transcriptional regulation of HIF-2a protein levels. Although resveratrol is nonspecific and not a direct activator of SIRT12,13, its effects are opposite to SIRT inhibitor EX527 or NAMPT inhibitor FK866, suggesting that its effect may due to activation of SIRT. To elucidate the mechanisms involved in regulating HIF-2a protein levels, we examined the proteasomal degradation pathway. Inhibition of the 26S proteasome pathway with MG132 caused a dramatic increase in overexpressed HIF-2a [Fig. 3(A)], indicating that HIF-2a protein levels are regulated by the 26S proteasomal degradation pathway. Additionally, inhibition of the 26S proteasome with MG132 [Fig. 3(B)] or PS-341 [Fig. 3(C)] prevented the reduction in HIF-2a protein levels caused by inhibition of NAMPT or SIRT without affecting HIF-2a mRNA levels. Furthermore, HIF-2a was ubiquitinated, an effect that was further enhanced by inhibition of NAMPT or SIRT [Fig. 3(D)]. This is consistent with the observed changes in HIF-2a protein levels. HIF-2a proteasomal degradation is regulated by prolyl-4hydroxylase domain (PHD) proteins, which hydroxylate HIF-2a leading to its ubiquitination17. Consistent with this, inhibition of hydroxylase activity with DMOG18 reversed the effect of NAMPT or SIRT on HIF-2a, restoring HIF-2a protein to levels comparable to those in controls [Fig. 4(A)]. Furthermore, whereas inhibition of NAMPT or SIRT caused proteasomal degradation of wild-type HIF2a [Fig. 4(B)], HIF-2a mutated at hydroxylation sites (DEpas1) was not degraded following inhibition of SIRT with nicotinamide or EX527 [Fig. 4(C)]. Interestingly, however, inhibition of NAMPT with FK866 caused effective degradation of the mutated HIF-2a [Fig. 4(C)], suggesting that SIRT, but not NAMPT, regulates hydroxylation-dependent HIF-2a degradation. In unstimulated chondrocytes, mRNA levels of the PHD2 isoform were the most abundant, whereas expression of PHD3 was minimal [Fig. 4(D)]. HIF-2a overexpression did not modulate the expression of PHD1. PHD2 expression was slightly increased by HIF-2a overexpression, whereas PHD3 expression was significantly enhanced [Fig. 4(D)]. The increase in PHD3 protein levels induced by HIF-2a overexpression was reduced by inhibition of NAMPT with FK866 or SIRT with nicotinamide or EX527 [Fig. 4(D)]. This increase in PHD3 levels following HIF-2a overexpression is consistent with previous reports of negative feedback regulation during proteasomal degradation of HIF-a19,20. In contrast to PHD3, PHD2 protein levels were markedly increased by inhibition of NAMPT with FK866 or inhibition of SIRT with nicotinamide or EX527 [Fig. 4(D)]. The increase in PHD2 protein is likely involved in the increased hydroxylation, and thereby enhanced proteasomal degradation, of HIF-2a. Indeed, other reports indicate that PHD2 is involved in HIF-a hydroxylation20,21. Therefore, it is likely that PHD2 is the main isoform that regulates HIF-2a hydroxylation in articular chondrocytes. Taken together, these results indicate that NAMPT and SIRT activities are necessary for HIF-2a protein stability by virtue of their ability to negatively regulate hydroxylase activity. SIRT activity is necessary for the OA pathogenesis caused by HIF-2a and NAMPT The significance of NAMPT and SIRT regulation of HIF-2a stability and transcriptional activity in OA pathogenesis was
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009
4
H. Oh et al. / Osteoarthritis and Cartilage xxx (2015) 1e9
Fig. 1. HIF-2a activates the NAMPT-NADþ-SIRT axis. (A) Representative RT-PCR gels and Western blots from more than six independent experiments using different sets of primary cultures of chondrocytes untreated (None) or infected with empty virus (Ad-C; 800 MOI) or Ad-Epas1 at the indicated MOI. (B and C) Chondrocytes were left untreated (None) or were infected with Ad-C (800 MOI) or Ad-Epas1 or Ad-Nampt at the indicated MOI, and incubated with or without FK866 (100 mM), nicotinamide (NIC; 15 mM) or EX527 (100 mM). NADþ and total NAD levels (B) and total SIRT activity (C) were determined (n 6). Values are presented as means ± S.E.M (*P < 0.01, **P < 0.001).
monitored by IA injection of SIRT inhibitors during the pathogenesis of experimental OA induced by Ad-Epas1 or Ad-Nampt injection. Similar to the inhibition of cartilage destruction by FK8668, IA injection of the SIRT inhibitor nicotinamide or EX527 significantly blocked the induction of cartilage destruction by Ad-Epas1 or AdNampt [Fig. 5(A)]. Immunostaining of cartilage sections revealed that the increases in MMP3 and MMP13 protein levels induced by Ad-Epas1 or Ad-Nampt were also blocked by IA co-injection of EX527 [Fig. 5(B)]. Consistent with this, qRT-PCR analyses revealed that inhibition of SIRT with nicotinamide or EX527 blocked the upregulation of matrix-degrading enzymes (MMP3, MMP9, MMP12, and MMP13) in primary cultured chondrocytes induced by HIF-2a or NAMPT [Fig. 5(C)]. IA injection of the SIRT inhibitor also
blocked the synovial inflammation induced by Ad-Epas1 or AdNampt [Supplementary Fig. 1]. Taken together, these results indicate that SIRT activity is necessary for the OA cartilage destruction caused by HIF-2a and NAMPT. Differential role of SIRT isoforms in the regulation of HIF-2a stability and transcriptional activity We next determined which SIRT isoforms are involved in HIF-2a regulation. Among SIRT family members, SIRT2 was found to be the most abundant in chondrocytes [Fig. 6(A)]. Although HIF-2a and NAMPT activated SIRT deacetylase activity [Fig. 1(C)], mRNA levels of SIRT family members were not changed by overexpression of
Fig. 2. The NAMPT-NADþ-SIRT axis regulates the stability and transcriptional activity of HIF-2a. (A) Chondrocytes were left untreated (None) or were infected with Ad-C or AdEpas1 at the indicated MOI and then treated with FK866 (FK; 100 mM), nicotinamide (NIC; 15 mM), resveratrol (Res; 100 mM), Ad-Nampt (Ad-N; 800 MOI), or MG132 (MG; 1 mM) for the last 8 h of culture. HIF-2a and acetylated lysine (Ac-Lys) were detected by Western analysis of HIF-2a immunoprecipitates. (BeD) Chondrocytes were infected with Ad-C or AdEpas1 (800 MOI) and then treated with the indicated concentrations of FK866, nicotinamide (NIC), EX527, or resveratrol (Res) for 36 h. HIF-2a reporter gene activity (B), mRNA and protein levels (C), and HIF-2a immunofluorescence microscopy (D) are shown. Representative images are presented in A, C and D from more than six independent experiments using different sets of primary cultures. Values in B are presented as means ± S.E.M (n 6; *P < 0.0005, **P < 0.00005).
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009
H. Oh et al. / Osteoarthritis and Cartilage xxx (2015) 1e9
5
Fig. 3. The NAMPT-SIRT pathway regulates proteasomal degradation of HIF-2a. (A) Chondrocytes were infected with Ad-C or Ad-Epas1 at an MOI of 800. HIF-2a protein was detected by Western blotting in chondrocytes treated with MG132 for 8 h. (B and C) Chondrocytes were infected with Ad-Epas1 for 36 h and then treated with FK866, nicotinamide (NIC), or EX527 for additional 36 h. MG132 (B) or PS-341 (C) was included for the last 8 h of culture to inhibit the 26S proteasomal pathway. HIF-2a mRNA and protein levels were detected by RT-PCR and Western blotting, respectively. (D) Chondrocytes were left untreated (None) or were infected with Ad-C or Ad-Epas1 at an MOI of 800 for 16 h in the presence or absence of the indicated concentrations of FK866 or nicotinamide (NIC), and then treated with MG132 for an additional 8 h. HIF-2a and ubiquitinated (Ub) HIF-2a were detected by Western analysis of HIF-2a immunoprecipitates. Representative images are presented from more than six independent experiments using different sets of primary cultures.
Fig. 4. The SIRT pathway regulates HIF-2a degradation via prolyl-4-hydroxylase. (A) HIF-2a mRNA and protein levels were detected in untreated chondrocytes (None) or chondrocytes infected with Ad-C or Ad-Epas1 (800 MOI) for 36 h in the absence or presence of FK866, nicotinamide (NIC), or EX527. DMOG at the indicated concentrations was added to inhibit hydroxylase activity for additional 36 h. (B) Chondrocytes were left untreated (None) or were infected with Ad-C or Ad-Epas1 (800 MOI) in the absence or presence of FK866, nicotinamide (NIC), or EX527. HIF-2a and lamin B protein levels were detected by Western blotting. (C) Chondrocytes were left untreated (None) or were transfected with empty vector (EV) or vector expressing hydroxylation sites-mutated HIF-2a (Myc-tagged DEpas1) in the absence or presence of the indicated concentrations of FK866, nicotinamide (NIC), or EX527. Myc-tagged HIF-2a and lamin B protein levels were detected by Western blotting. (D) Left: PHD mRNA levels determined by qRT-PCR in chondrocytes infected with Ad-C or Ad-Epas1 (800 MOI) for 24 h. Right: Western blot detection of PHD isoforms in chondrocytes infected with Ad-Epas1 in the absence or presence of FK866, nicotinamide (NIC), or EX527 (right). Representative images are presented from more than seven independent experiments using different sets of primary cultures. Values in D are presented as means ± S.E.M (n ¼ 8; *P < 0.001 vs Ad-C).
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009
6
H. Oh et al. / Osteoarthritis and Cartilage xxx (2015) 1e9
Fig. 5. SIRT activity is necessary for HIF-2a- and NAMPT-induced expression of matrix-degrading enzymes and OA pathogenesis. (A and B) Mice were IA-injected with Ad-C, AdEpas1 or Ad-Nampt (1 109 PFU), and incubated with or without nicotinamide (NIC; 300 mg/kg) or EX527 (1.25 mg/kg). Representative images of cartilage sections stained with safranin-O (left), and Mankin's method (right) from 12 different mice (A). The indicated proteins were detected by immunostaining cartilage sections (n 6). (B and C) Chondrocytes were left untreated (None) or were infected with Ad-C, Ad-Epas1, or Ad-Nampt at an MOI of 800 in the absence or presence of the indicated concentrations of nicotinamide (NIC; left) or EX527 (right). Expression levels of the indicated mRNAs were quantified by qRT-PCR from more than six independent experiments using different sets of primary cultures. Values are presented as means ± S.E.M (*P < 0.01, **P < 0.001; NS, not significant). Scale bar: 100 mm.
HIF-2a or NAMPT [Fig. 6(A)]. To elucidate which SIRT isoforms were involved in the regulation of HIF-2a, we overexpressed individual SIRT isoforms in articular chondrocytes by transfection with the corresponding expression vectors. Overexpression of SIRT2 caused the most marked increase in HIF-2a protein stability and transcriptional activity [Fig. 6(B)]. Overexpression of SIRT4 also increased HIF-2a protein stability and transcriptional activity, albeit to a lesser degree [Fig. 6(B)]. In contrast, SIRT3 overexpression caused a decrease in HIF-2a stability and reporter gene activity, whereas overexpression of SIRT1, SIRT5, SIRT6, or SIRT7 had no effect on HIF-2a stability or transcriptional activity [Fig. 6(B)]. These results indicate that SIRT members display isoform-specific regulation of HIF-2a stability and transcriptional activity. To further elucidate the roles of SIRT isoforms in the regulation of HIF-2a, we used specific siRNAs to knockdown each SIRT isoform. HIF-2a transcriptional activity was significantly reduced by Sirt2 knockdown and, to a lesser degree, by Sirt4 knockdown, whereas knockdown of Sirt3 increased it [Fig. 6(C)]. Changes in HIF-2a transcriptional activity were correlated with HIF-2a protein levels: knockdown of Sirt2 or Sirt4 reduced overexpressed HIF-2a protein levels, whereas knockdown of Sirt3 increased them, without affecting HIF-2a mRNA levels [Fig. 6(C)]. Taken together, these results suggest that both SIRT2 and SIRT4 appear to enhance HIF-2a protein stability and transcriptional activity in chondrocytes.
SIRT2 regulates HIF-2a- and NAMPT-induced OA cartilage destruction Based on the capacity of SIRT2 to regulate HIF-2a protein stability, we examined the role of SIRT2 in HIF-2a- and NAMPTinduced OA cartilage destruction. As expected11, overexpressed SIRT2 was localized to the cytosol [Fig. 7(A)]. Similar to the effects of SIRT2 overexpression produced by transfection with a SIRT2 expression vector, co-infection with Ad-Sirt2 and Ad-Epas1 enhanced HIF-2a protein stability and transcriptional activity
[Fig. 7(B)]. In contrast, knockdown of Sirt2 in chondrocytes by infection with Ad-shSirt2 reduced HIF-2a protein levels and transcriptional activity [Fig. 7(C)]. Overexpression of SIRT2 alone in joint tissues through IA injection of Ad-Sirt2 in mice did not cause cartilage destruction [Fig. 7(D)]. However, IA co-injection of AdshSirt2 with Ad-Epas1 or Ad-Nampt significantly inhibited HIF-2aand NAMPT-induced cartilage destruction [Fig. 7(E)], indicating the involvement of SIRT2 in both HIF-2a- and NAMPT-induced OA cartilage destruction. Notably, conditional knockout of Sirt1 in cartilage tissue did not affect HIF-2a-induced cartilage destruction [Supplementary Fig. 2], and IA injection of Ad-Sirt1, which caused overexpression of SIRT1 in cartilage tissue, also did not cause cartilage destruction [Supplementary Fig. 2]. These results suggest that SIRT1 is not involved in HIF-2a-induced cartilage destruction.
Discussion The results presented here reveal reciprocal regulation of HIF-2a and the NAMPT-NADþ-SIRT axis in chondrocytes. We first demonstrated HIF-2a activation of the NAMPT-NADþ-SIRT axis, showing that HIF-2a upregulates NAMPT, which in turn stimulates NADþ synthesis and SIRT activation. Conversely, NAMPT/SIRT activity is necessary for HIF-2a protein stability and transcriptional activity. Inhibition of NAMPT or SIRT blocked HIF-2a-induced cartilage destruction, demonstrating that this reciprocal regulation is required for HIF-2a-induced OA pathogenesis. HIF-2a activation of the NAMPT-NADþ-SIRT axis is expected, given our previous observation that the Nampt is a direct target of HIF-2a in chondrocytes8. Furthermore, NAMPT upregulation is necessary for HIF-2a-induced expression of matrix-degrading enzymes and OA cartilage destruction8. The main function of NAMPT is to stimulate the synthesis of NADþ, which is an essential cofactor for members of the SIRT deacetylase family9,10. Indeed, overexpression of HIF-2a or NAMPT caused NADþ synthesis and SIRT activation in chondrocytes. More importantly, the HIF-2a-
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009
H. Oh et al. / Osteoarthritis and Cartilage xxx (2015) 1e9
7
Fig. 6. SIRT isoform-specific regulation of HIF-2a protein stability and transcriptional activity. (A) Chondrocytes were left untreated (None) or were infected with Ad-C, Ad-Epas1, or Ad-Nampt at an MOI of 800. mRNA levels were determined by RT-PCR (left). The relative amount of SIRT mRNA in untreated chondrocytes was normalized to that of GAPDH mRNA (right). (B) Individual SIRT isoforms were overexpressed in chondrocytes by transfection with expression vectors. Cells were infected with Ad-C or Ad-Epas1 at an MOI of 800 for 36 h. HIF-2a reporter gene activity and HIF-2a and SIRT protein levels were detected. (C) Chondrocytes were left untreated (None) or were transfected with 100 nM control siRNA (C-siRNA) or the indicated concentrations of siRNA specific for individual SIRT isoforms. Cells were infected with Ad-C or Ad-Epas1 (400 MOI) for 36 h. HIF-2a reporter gene activity and protein levels were detected. Values are presented as means ± S.E.M from more than six independent experiments using different sets of primary cultures (*P < 0.05, **P < 0.005; NS, not significant).
Fig. 7. SIRT2 regulates HIF-2a- and NAMPT-induced cartilage destruction. (A) Representative immunostaining image of SIRT2 in chondrocytes transfected with Sirt2 expressing vector from more than six independent experiments using different sets of primary cultures. (B) Chondrocytes were co-infected with Ad-Epas1 and Ad-Sirt2. Left: mRNA and protein levels of SIRT2 and HIF-2a; right: HIF-2a transcriptional activity (n ¼ 5). (C) Chondrocytes were left untreated or were infected with Ad-C or Ad-Epas1 and Ad-shSirt2. Left: mRNA and protein levels of SIRT2 and HIF-2a; right: HIF-2a transcriptional activity (n ¼ 11). (D) Safranin-O staining (left) and SIRT2 immunostaining (right) in cartilage sections of mice injected with Ad-C or Ad-Sirt2. (E) Safranin-O staining and Mankin score of cartilage sections from 14 mice IA-injected with Ad-C or Ad-Epas1 with or without Ad-shSirt2 (n ¼ 14).
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009
8
H. Oh et al. / Osteoarthritis and Cartilage xxx (2015) 1e9
stimulated NAMPT-NADþ-SIRT axis, in turn, promoted HIF-2a protein stability and transcriptional activity. Indeed, inhibition of the NAMPT/SIRT pathway promoted degradation of ectopically expressed HIF-2a, indicating that the NAMPT-SIRT pathway negatively regulates the proteasomal degradation of HIF-2a. Interestingly, SIRT regulation of HIF-2a proteasomal degradation depended on HIF-2a hydroxylation, whereas NAMPT action was hydroxylation independent. HIF-2a-simulated SIRT activity is required for HIF-2a- and NAMPT-induced OA pathogenesis, as evidenced by the fact that inhibition of SIRT activity blocked HIF-2a regulation of matrix-degrading enzyme expression in chondrocytes and OA cartilage destruction. Additionally, the catabolic functions of the NAMPT-SIRT pathway are exerted through enhanced HIF-2a stability and transcriptional activity. The mammalian SIRT family is composed of seven members (SIRT1e7) that possess NADþ-dependent deacetylase, deacylase, and ADP-ribosyltransferase activities10e12. They are found in different subcellular locations, including the nucleus (SIRT6 and SIRT7), nucleus and cytosol (SIRT1 and SIRT2), and mitochondria (SIRT3e5). Among SIRT family members, SIRT1 is the best characterized and has been shown to play a protective role in OA pathogenesis22e24. For example, SIRT1 is necessary for chondrocyte survival, and loss of SIRT1 function causes chondrocyte apoptosis25e28. SIRT1 also regulates cartilage-specific gene expression29e31. It was recently reported that chondrocyte-specific conditional knockout of Sirt1 in mice causes transiently accelerated progression of surgically induced OA32. Similarly, Sirt1þ/ mice show increased chondrocyte apoptosis and enhanced OA severity27, and mutant mice carrying a variant of SIRT1 lacking enzymatic activity show elevated rates of cartilage degradation with age33. Collectively, these observations suggest that SIRT1 activity serves a protective role in OA pathogenesis, although this capacity of SIRT1 does not appear to be especially marked, as evidenced by the observation that knockout of Sirt1 promotes only the early stage of OA progression32. In contrast to the reported protective function of SIRT1 in surgically induced OA pathogenesis, our current results demonstrate that SIRT1 is not involved in HIF-2a- or NAMPT-induced OA cartilage destruction in mice. This is demonstrated by the observation that conditional knockout of Sirt1 in cartilage tissue does not affect HIF-2a-induced cartilage destruction. Furthermore, we found that SIRT1 does not affect HIF-2a protein stability or transcriptional activity in chondrocytes. Although SIRT1 is not involved in HIF-2aor NAMPT-induced OA cartilage destruction, we demonstrated that inhibition of NADþ-dependent SIRT deacetylase activity blocked OA cartilage destruction induced by HIF-2a or NAMPT, with a concomitant inhibition of the expression of matrix-degrading enzymes. This indicates that NADþ-dependent SIRT activity promotes HIF-2a- and NAMPT-induced OA pathogenesis by virtue of its capacity to regulate HIF-2a protein stability and transcriptional activity. Among SIRT family members, SIRT2, which is the most abundant in chondrocytes, exerted marked effects on HIF-2a protein stability and transcriptional activity. SIRT2 exhibits deacetylase activity and localizes mainly to the cytosol10e12. This is consistent with the regulatory mechanisms of HIF-2a protein stability. HIF-2a protein is degraded by 26S proteasomal pathway in the cytosol, resulting in minimal transcriptional activity. However, under pathological conditions (i.e., under hypoxic condition), its degradation pathway is blocked and accumulated HIF-2a translocates into nucleus forming heterodimer with HIF-1b to regulate target gene expression17. Therefore, it is likely that the SIRT2 regulates HIF-2a protein stability and transcriptional activity, and suppression of SIRT2 activity is likely responsible for the observed inhibitory effects of SIRT inhibitors on the OA cartilage destruction caused by HIF-2a or NAMPT. Moreover, our demonstration that
knockdown of Sirt2 by IA injection of Ad-shSirt2 inhibits HIF-2aand NAMPT-induced cartilage destruction clearly indicate the role of SIRT2. In contrast to our observation that SIRT2 stabilizes HIF-2a protein without affecting its acetylation status in chondrocytes, recent report indicated that SIRT2 destabilizes HIF-1a by regulating deacetylation in tumor cells34. Despite many similarities between HIF-1a and HIF-2a, these two isoforms show different sensitivity to oxygen tension and display distinct, and sometimes opposing, cellular activities17. Therefore, it is likely that protein stability of HIF-1a and HIF-2a are differentially regulated by SIRT2 depending on cell types. In addition to SIRT2, our results indicate that mitochondrial SIRT3 and SIRT4 also regulate HIF-2a stability. Similar to SIRT2, overexpression of SIRT4 enhanced HIF-2a stability, whereas knockdown of Sirt4 destabilizes HIF-2a. To the best of our knowledge, this is the first evidence that SIRT4 regulates HIF-2a stability. However, it remains to be elucidated whether SIRT4 is also associated with OA pathogenesis caused by HIF-2a and NAMPT. Contrast to cytosolic SIRT2 and mitochondrial SIRT4, SIRT3 which is a major mitochondrial NADþ-dependent deacetylase destabilized HIF-2a stability in chondrocytes. Although we could not find any previous report for the regulation of HIF-2a by SIRT3, many reports indicated that SIRT3 destabilizes HIF-1a in tumor cells35,36. Therefore, the significance of SIRT3 regulation of HIF-2a stability in OA pathogenesis remains to be elucidated. Furthermore, it is of interesting to reveal possible orchestration of SIRT isoforms in the regulation of HIF-2a stability. In summary, we demonstrated reciprocal regulation of HIF-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes. HIF-2a activates the NAMPT-NADþ-SIRT axis, which, in turn, promotes HIF2a protein stability resulting in increased HIF-2a transcriptional activity. We also revealed that many SIRT isoforms, including SIRT2, SIRT3, and SIRT4, are associated with HIF-2a stability regulation. Among them, SIRT2 and SIRT4 are positively associated with HIF-2a stability in chondrocytes. This reciprocal regulation is involved in the expression of catabolic MMPs and OA cartilage destruction caused by HIF-2a or NAMPT.
Author contributions OH and JSK: Study design, data acquisition, data analysis and interpretation, manuscript preparation and approval; SY, MKG, JHK, JR, SKK, HEK and JHR:data acquisition, analysis, and interpretation, and manuscript approval; JSC: funding acquisition, study design, data interpretation,manuscript preparation and approval. JSC ([email protected]) takes responsibility for the integrity of this work.
Conflicts of interest The authors have no conflicts of interest.
Acknowledgment This work was supported by grants from the National Research Foundation of Korea (2007-0056157 and 2013R1A2A1A01009713), the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare (HI14C3484), and the Integrative Aging Research Center of Gwangju Institute of Science and Technology. The funders had no role in study design, conduct of research, or preparation of the manuscript. The content of this manuscript is solely the responsibility of the authors.
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009
H. Oh et al. / Osteoarthritis and Cartilage xxx (2015) 1e9
Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.joca.2015.07.009.
20.
References 1. Heinegard D, Saxne T. The role of the cartilage matrix in osteoarthritis. Nat Rev Rheumatol 2011;7:50e6. 2. Bian Q, Wang YJ, Liu SF, Li YP. Osteoarthritis: genetic factors, animal models, mechanisms, and therapies. Front Biosci 2012;E4:74e100. 3. Little CB, Hunter DJ. Post-traumatic osteoarthritis: from mouse models to clinical trials. Nat Rev Rheumatol 2013;9:485e97. 4. Kim JH, Jeon J, Shin M, Won Y, Lee M, Lee G, et al. Regulation of the catabolic cascade in osteoarthritis by the zinc-ZIP8-MTF1 axis. Cell 2014;156:730e43. 5. Yang S, Kim J, Ryu JH, Oh H, Chun CH, Kim BJ, et al. Hypoxiainducible factor-2a is a catabolic regulator of osteoarthritic cartilage destruction. Nat Med 2010;16:687e93. 6. Ryu JH, Shin Y, Huh YH, Yang S, Chun CH, Chun JS. Hypoxiainducible factor-2a regulates fas-mediated chondrocyte apoptosis during osteoarthritic cartilage destruction. Cell Death Differ 2012;19:440e50. 7. Ryu JH, Yang S, Shin Y, Rhee J, Chun CH, Chun JS. Interleukin-6 plays an essential role in hypoxia-inducible factor 2a-induced experimental osteoarthritic cartilage destruction in mice. Arthritis Rheum 2011;63:2732e43. 8. Yang S, Ryu JH, Oh H, Jeon J, Kwak JS, Kim JH, et al. NAMPT (visfatin), a direct target of hypoxia-inducible factor-2a, is an essential catabolic regulator of osteoarthritis. Ann Rheum Dis 2015;74:595e602. 9. Zhang LQ, Heruth DP, Ye SQ. Nicotinamide phosphoribosyltransferase in human diseases. J Bioanal Biomed 2011;3: 13e25. 10. Imai SI. Nicotinamide phosphoribosyl transferase (NAMPT): a link between NAD biology, metabolism, and diseases. Curr Pharm Des 2009;15:20e8. 11. Morris BJ. Seven sirtuins for seven deadly diseases of aging. Free Radic Biol Med 2013;56:133e71. 12. Carafa V, Nebbioso A, Altucci L. Sirtuins and disease: the road ahead. Front Pharmacol 2012;3:1e4. 13. Lavu S, Boss O, Elliott PJ, Lambert PD. Sirtuins-novel therapeutic targets to treat age-associated diseases. Nat Rev Drug Discov 2008;7:841e53. 14. Mankin HJ, Dorfman H, Lippiello L, Zarins A. Biochemical and metabolic abnormalities in articular cartilage from osteoarthritic human hips. II. Correlation of morphology with biochemical and metabolic data. J Bone Joint Surg Am 1971;53:523e37. 15. Dioum EM, Chen R, Alexander MS, Zhang Q, Hogg RT, Gerard RD, et al. Regulation of hypoxia-inducible factor-2a signaling by the stress-responsive deacetylase sirtuin 1. Science 2009;324:1289e93. 16. Chen R, Dioum EM, Hogg RT, Gerard RD, Garcia JA. Hypoxia increases sirtuin 1 expression in a hypoxia-inducible factordependent manner. J Biol Chem 2011;286:13869e78. 17. Greer SN, Metcalf JL, Wang Y, Ohh M. The updated biology of hypoxia-inducible factor. EMBO J 2012;31:2448e60. 18. Fujita N, Chiba K, Shapiro IM, Risbud MV. HIF-1a and HIF-2a degradation is differentially regulated in nucleus pulpous cells of the intervertebral disc. J Bone Min Res 2012;27:401e12. 19. Minamishima YA, Moslehi J, Padera RF, Bronson RT, Liao R, Kaelin Jr WG. A feedback loop involving the Phd3 prolyl
21. 22. 23. 24. 25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
9
hydroxylase tunes the mammalian hypoxic response in vivo. Mol Cell Biol 2009;29:5729e41. Fujita N, Markova D, Anderson DG, Chiba K, Toyama Y, Shapiro IM, et al. Expression of prolyl hydroxylases (PHDs) is selectively controlled by HIF-1 and HIF-2 proteins in nucleus pulposus cells of the intervertebral disc: distinct roles of PHD2 and PHD3 proteins in controlling HIF-1a activity in hypoxia. J Biol Chem 2012;287:16975e86. Minamishima YA, Kaelin Jr WG. Reactivation of hepatic EPO synthesis in mice after PHD loss. Science 2010;329:407. Asahara H. Switching gears to an arthritis gene expression network by SirT1 cleavage. Arthritis Rheum 2011;63:2161e3. Gabay O, Sanchez C. Epigenetics, sirtuins and osteoarthritis. Joint Bone Spine 2012;79:570e3. Dvir-Ginzberg M, Steinmeyer J. Towards elucidating the role of SirT1 in osteoarthritis. Front Biosci 2013;18:343e55. Takayama K, Ishida K, Matsushita T, Fujita N, Hayashi S, Sasaki K, et al. SIRT1 regulation of apoptosis of human chondrocytes. Arthritis Rheum 2009;60:2731e40. Gagarina V, Gaybay O, Dvir-Ginzberg M, Lee EJ, Brady JK, Quon MJ, et al. SirT1 enhances survival of human osteoarthritic chondrocytes by repressing protein tyrosine phosphatase 1B and activating the insulin-like growth factor receptor pathway. Arthritis Rheum 2010;62:1383e92. Gaybay O, Oppenhiemer H, Meir H, Zaal K, Sanchez C, DvirGinzberg M. Increased apoptotic chondrocytes in articular cartilage from adult heterozygous SirT1 mice. Ann Rheum Dis 2012;71:613e6. Oppenheimer H, Gaybay O, Meir H, Haze A, Kandel L, Lieberggall M, et al. 75-kd sirtuin 1 blocks tumor necrosis factor a-mediated apoptosis in human osteoarthritic chondrocytes. Arthritis Rheum 2012;64:718e28. Dvir-Ginzberg M, Gagarina V, Lee EJ, Hall DJ. Regulation of cartilage-specific gene expression in human chondrocytes by SirT1 and nicotinamide phosphoribosyltransferase. J Biol Chem 2008;283:36300e10. Fujita N, Matsushita T, Ishida K, Kubo S, Matsumoto T, Takayama K, et al. Potential involvement of SIRT1 in the pathogenesis of osteoarthritis through the modulation of chondrocyte gene expressions. J Orthop Res 2011;29: 511e5. Dvir-Ginzberg M, Gagarina V, Lee EJ, Booth R, Gaybay O, Hall DJ. Tumor necrosis factor a-mediated cleavage and inactivation of SirT1 in human osteoarthritic chondrocytes. Arthritis Rheum 2011;63:2363e73. Matsuzaki T, Matsushita T, Takayama K, Matsumoto T, Nishida K, Kuroda R, et al. Disruption of Sirt1 in chondrocytes causes accelerated progression of osteoarthritis under mechanical stress and during ageing in mice. Ann Rheum Dis 2014;73:1397e404. Gabay O, Sanchez C, Dvir-Ginzberg M, Gaqarina V, Zaal KJ, Song Y, et al. Sirt1 enzymatic activity is required for cartilage homeostasis in vivo. Arthritis Rheum 2013;65:159e66. Seo KS, Park JH, Heo JY, Jing K, Han J, Min KN, et al. SIRT2 regulates tumor hypoxia response by promoting HIF-1a hydroxylation. Oncogene 2015;34:1354e62. Finley LW, Carracedo A, Lee J, Souza A, Egia A, Zhang J, et al. SIRT3 opposes reprogramming of cancer cell metabolism through HIF1a destabilization. Cancer Cell 2011;1: 416e28. Bell EL, Emerling BM, Ricoult SJ, Guarente L. SirT3 suppresses hypoxia inducible factor 1a and tumor growth by inhibiting mitochondrial ROS production. Oncogene 2011;30:2986e96.
Please cite this article in press as: Oh H, et al., Reciprocal regulation by hypoxia-inducible factor-2a and the NAMPT-NADþ-SIRT axis in articular chondrocytes is involved in osteoarthritis, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.07.009