Nmnat2 Protects Cardiomyocytes from Hypertrophy Via Activation of SIRT6

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

FEBS Letters 586 (2012) 866–874

journal homepage: www.FEBSLetters.org

Nmnat2 protects cardiomyocytes from hypertrophy via activation of SIRT6 ⇑ Yi Cai a,1, Shan-Shan Yu b,1, Shao-Rui Chen a, Rong-Biao Pi a, Si Gao a, Hong Li a, Jian-Tao Ye a, , ⇑ Pei-Qing Liu a,

a Department of Pharmacology and Toxicology, School of Pharmaceutical Sciences, Sun Yat-Sen University, Guangzhou 510006, China b Department of Pharmaceutical Science, Zhujiang Hospital, Southern Medical University, Guangzhou 510280, China

article info abstract

Article history: The discovery of sirtuins (SIRT), a family of nicotinamide adenine dinucleotide (NAD)-dependent Received 25 October 2011 deacetylases, has indicated that intracellular NAD level is crucial for the hypertrophic response of Revised 4 January 2012 cardiomyocytes. Nicotinamide mononucleotide adenylyltransferase (Nmnat) is a central enzyme Accepted 12 February 2012 in NAD biosynthesis. Here we revealed that Nmnat2 protein expression and enzyme activity were Available online 20 February 2012 down-regulated during cardiac hypertrophy. In neonatal rat cardiomyocytes, overexpression of Edited by Zhijie Chang Nmnat2 but not its catalytically inactive mutant blocked angiotensin II (Ang II)-induced cardiac hypertrophy, which was dependent on activation of SIRT6 through maintaining the intracellular NAD level. Our results suggested that modulation of Nmnat2 activity may be beneﬁcial in cardiac Keywords: Nmnat2 hypertrophy. Cardiac hypertrophy Ó 2012 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. Sirtuin 6 NAD Ang II

1. Introduction ure. However, the molecular mechanisms underlying sirtuin-mediated cardioprotection and how the functions of sirtuins are A variety of stress-responsive signaling pathways promote car- regulated under the circumstance of cardiomyopathy remain to be diac hypertrophy, but the mechanisms that link these pathways clarified. are only beginning to be unveiled. The sirtuin (SIRT) family, which Nicotinamide adenine dinucleotide (NAD) is a classic coenzyme belongs to class III of histone deacetylases (HDACs), is implicated with a well-established role in cellular redox reactions. Recently, in the control of critical cellular processes such as differentiation, several lines of evidence have implicated NAD biochemistry in a proliferation, apoptosis, metabolism and aging [1,2]. In mammals, broad range of biological functions [6,7]. It’s now known that the there are seven sirtuin family members named SIRT1-7. Recently, activity of sirtuins, as NAD-dependent histone deacetylases, is sus- SIRT1, SIRT3, and SIRT7 have been proved to play important roles ceptible to the availability of NAD content, and reduced cellular in physiological and pathological cardiac growth [3–5]. Our previ- levels of NAD attenuate the SIRT deacetylase activity [8]. A recent ous study have also demonstrated that SIRT6 is inactivated during study reported that NAD could repress the agonist-induced cardiac Ang II-induced cardiac hypertrophy, and increased activity of SIRT6 hypertrophy by mediating the activation of SIRT3 but not SIRT1 [9]. protects cardiomyocytes from hypertrophic response by suppress- Consistently, we found that the deacetylase activity of SIRT6 drops ing NFjB activation (Supplementary Fig. S1). These findings conver- along with the intracellular NAD level during cardiac hypertrophy, gently suggest that sirtuins may be potential targets for the indicating a close relationship between NAD and SIRT6 in heart prevention and/or treatment of cardiac hypertrophy and heart fail- (Supplementary Fig. S1). Therefore, it gives rise to the hypothesis that artificial modulation of NAD levels may be used to regulate the function of sirtuins including SIRT6. Abbreviations: SIRT, sirtuin; NAD, nicotinamide adenine dinucleotide; Nmnat, nicotinamide mononucleotide adenylyltransferase; Ang II, angiotensin II; HDACs, In mammalian cells, the biosynthesis of NAD occurs through histone deacetylases; Nampt, nicotinamide phosphoribosyltransferase; AAC, both salvage and de novo pathways [10]. Nicotinamide phosphori- abdominal aortic constriction; ANF, atrial natriuretic factor; BNP, brain natriuretic bosyltransferase (Nampt) is the rate-limiting enzyme for NAD sal- polypeptide; b-MHC, myosin heavy chain b vage from nicotinamide. Previous study has demonstrated that ⇑ Corresponding authors. Fax: +86 20 3994 3026. overexpression of Nampt significantly increased NAD synthesis in E-mail addresses: [email protected] (J.-T. Ye), [email protected] (P.-Q. Liu). primary cultures cardiomyocytes [11]. However, it was reported 1 These authors contributed equally to this work. that overexpression of Nampt in cardiomyocytes stimulates

0014-5793/$36.00 Ó 2012 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2012.02.014 Y. Cai et al. / FEBS Letters 586 (2012) 866–874 867

Fig. 1. Tissue distribution and age-dependent expression of Nmnat2 in rats. The mRNA (A) and protein (B) expression of Nmnat2 in different tissues of SD rats were detected by qRT-PCR and Western blot, respectively. The mRNA expressions of Nmnat isoforms and Nmnat2 protein level in rat hearts at indicated ages were also determined (C and D). ⁄P < 0.05, ⁄⁄P < 0.01 vs. 1d group, n =6.

neither Akt/p70S6K signaling in vitro nor cardiac hypertrophy mutant (H24D) plasmids were purchased from LAND Company. in vivo [12]. Nmnat is another central enzyme in NAD synthesis Plasmids were transfected into cells using Lipofectamine 2000 pathway, which catalyzes both de novo and salvage pathways of (Invitrogen) according to the manufacturer’s protocol. NAD synthesis [13]. Up to now, three Nmnat isoforms have been Additional materials and methods including echocardiography identified in mammals which showed differences in subcellular and morphometric measures, quantitative real-time polymerase localization and tissue distribution [14], but very little is known chain reaction (qRT-PCR), Western blot and immunoprecipitation, about the effects of the Nmnat family members on cardiac func- SIRT6 histone deacetylase activity assay, NAD determination, mea- tion. Among the three Nmnat isoforms, Nmnat2 is reported to be surement of cell surface area, immunofluorescence assay, Nmnat2 most sensitive to NAD and can act as a sensor to intracellular enzyme activity assay and RNA interference are listed in Supple- NAD level [13], which might partly explain the facts that high lev- mentary data. Data are presented as mean ± S.E. Statistical analyses els of Nmnat2 are detected in the organs with high demand for en- between two groups were performed by unpaired Student’s t-test. ergy such as heart, brain and skeletal muscle [14]. Using a positive Differences among groups were tested by one-way analysis of var- screen, we have identified that the mRNA level of Nmnat2, but not iance (ANOVA) with Tukey’s post hoc test. In all cases, differences Nmnat1 and Nmnat3, is significantly upregulated in response to were considered statistically significant with P < 0.05. hypertrophic stimulation. However, to our best knowledge, there are no investigations on the potential roles of Nmnat2 in cardiac 3. Results hypertrophy. In this study, by using gain- and loss-of-function approaches in 3.1. Tissue distribution and age-dependent expression of Nmnat2 in cardiomyocytes we show that Nmnat2 is a negative regulator of rats cardiac hypertrophy. Our results also reveal that the anti-hypertrophic effect of Nmnat2 is dependent on its NAD synthesis activity As shown in Fig. 1A and B, Nmnat2 was highly expressed in and is possibly mediated via activation of SIRT6. These findings liver, heart, brain and muscle, while it could hardly be detected give the first look into the relationship between Nmnat2 and in spleen, lung, kidney and testis. We further investigated the SIRT6-mediated cardioprotection, and may provide a potentially expression of Nmnat isoforms in hearts of rats at indicated ages. novel strategy against cardiac hypertrophy by modulating Nmnat2. The results showed that the mRNA and protein expression of Nmnat2 (Fig. 1C and D) was increased in an age-dependent man- 2. Materials and methods ner, which reached peak at 12 weeks (P < 0.01, compared with 1d group). In contrast, the expression of Nmnat1 and Nmnat3 showed SD rats (male, weighing 180–200 g, SPF grade, certification No. no significant age-related changes. 0067950) were supplied by the Experimental Animal Center of Sun Yat-Sen University (Guangzhou, China). Pressure overload 3.2. Expression and NAD synthase activity of Nmnat2 in neonatal rat was induced by abdominal aortic constriction (AAC) according to cardiomyocytes treated with Ang II procedure reported by others [15]. Primary cultures of neonatal rat cardiomyocytes were prepared as we previously described Ang II is an well-established neurohumoral factor that can stim- [16]. The EGFP-Nmnat2 (Nmnat2) and the catalytically inactive ulate cardiac hypertrophy [17]. As shown in Fig. 2A, the mRNA 868 Y. Cai et al. / FEBS Letters 586 (2012) 866–874

Fig. 2. The protein expression and NAD synthase activity of Nmnat2 was decreased in neonatal rat cardiomyocytes treated with Ang II. (A) The mRNA levels of Nmnat isoforms were determined by qRT-PCR after treatment with 100 nM Ang II for 24 h. The time and concentration-dependent effect of Ang II on Nmnat2 mRNA and protein expression was detected by qRT-PCR (B and C) and Western blot (D and E). (F) The Nmnat2 enzyme activity in cardiomyocytes treated with 100 nM AngII. ⁄P < 0.05, ⁄⁄P < 0.01 vs. control, n =5. expression of Nmnat2, but not the other two Nmnat isoforms, was weight (HW/BW) ratio and left ventricle weight to body weight enhanced after treatment with 100 nM Ang II (P < 0.05, compared (LVW/BW) ratio were both increased in AAC rats comparing with with control). Furthermore, the Nmnat2 mRNA expression in- sham-operated animals (Supplementary Fig. S2E and Table S3). And creased in a dose and time dependent manner responding to Ang the expression of hypertrophy marker genes including ANF, BNP and II stimulation (Fig. 2B and C). However, Ang II downregulated b-MHC mRNA was signiﬁcantly increased in AAC rats (Supplementary Nmnat2 protein expression in a dose and time dependent manner Fig. S2F). These results convergently suggested the successful induc- (Fig. 2D and E), and Nmnat2 NAD synthase activity dropped grad- tion of cardiac hypertrophy by AAC surgery. In addition, as shown in ually after the treatment with 100 nM Ang II (Fig. 2F). Fig. 3A and B, both protein expression and enzyme activity of Nmnat2 were signiﬁcantly downregulated (45.8 ± 5.76% and 51.2 ± 4.96%, 3.3. Expression and NAD synthase activity of Nmnat2 in AAC rats respectively, P < 0.01, compared with Sham group) after 8 weeks of pressure overload induced by AAC. Abdominal aortic constriction surgery was performed to produce pressure overload hypertrophy in rats. At 8 weeks after surgery, the 3.4. Nmnat2 protected cardiomyocytes from Ang II-induced hearts of AAC rats were obviously larger than that of sham-operated hypertrophy in neonatal rat cardiomyocytes animals (sham) and presented with typical hypertrophic changes as demonstrated by HE staining (Supplementary Fig. S2A–C). The echo- To investigate the potential effect of Nmnat2 on cardiac hyper- cardiography revealed that IVSd, IVSs, LVPWd and LVPWs were signif- trophy, neonatal rat cardiomyocytes were transfected with Nmnat2 icantly increased, whereas LVIDs was decreased after AAC or H24D, which lacked NAD synthase activity (Fig. 4A and B). Subse- (Supplementary Fig. S2D and Table S2).Theheartweighttobody quently, the cells were treated with 100 nM Ang II for 24 h, and Y. Cai et al. / FEBS Letters 586 (2012) 866–874 869

transfection with, no significant decrease of NAD content was observed following Ang II treatment (Fig. 6B). Among all seven sirtuins, SIRT6 mRNA expression was increased 2.5-fold in response to Ang II stimulation, and SIRT1 also showed slight increase (Fig. 6C). The protein expression and enzyme activity of SIRT6 were obviously elevated after transfection with Nmnat2 (Fig. 6D and E). Furthermore, confocal immunofluorescence microscopy revealed that SIRT6 was predomi- nantly localized in the nucleus, while Nmnat2 was localized in the cytoplasm (Fig. 6F),suggestingthattherewasnodirectinteractionbe- tween Nmnat2 and SIRT6. As shown in Fig. 6G and H, overexpression of Nmnat2, but not H24D, could increase the SIRT6 protein expression and enzyme activity, and similar effects were observed with NAD supplement. However, in cells transfected with si003, knockdown of Nmnat2 significantly decreased NAD content, SIRT6 protein expression and enzyme activity (Fig. 7A–C). As shown in Fig. 7D, when SIRT6 but not SIRT1 was knocked down, cardiomyocytes with Nmnat2 overexpression showed a significantly increase in hypertrophic response as determined by the expression of cardiac fetal gene. It suggested that Nmnat2 overexpression was unable to block Ang II-induced cardiac hypertrophy without the participation of SIRT6. In addition, SIRT6 overexpression could attenuate hypertrophic responses induced by Nmnat2 knockdown (Supplementary Fig. S6)as indicated by decreased expression of hypertrophic biomarkers as well as cell-surface area.

3.6. Proteasome participated in Ang II-induced Nmnat2 expression in cardiomyocytes Fig. 3. Nmnat2 protein expression was downregulated in AAC rats. (A) The protein levels of Nmnat2 were detected by Western blot. (B) Nmnat2 enzyme activity in AAC and sham rats. ⁄P < 0.05 vs. sham, n =8. Cardiomyocytes were pretreated with 1 lM proteasome inhibitor MG132 for 1 h followed by stimulation with 100 nM Ang II. In Fig. 8A, Nmnat2 protein level dropped after Ang II stimulation for 48 h, which could be attenuated by MG132 pretreatment. Nmnat2 protein expression was measured by Western blot. The re- And MG132 alone could significantly upregulate Nmnat2 protein sults showed that after Ang II stimulation, the Nmnat2 and H24D expression as compared with control group. However, as shown protein levels were higher in Nmnat2-tranfected cardiomyocytes in Fig. 8B, Nmnat2 mRNA expression increased after Ang II treat- than in cells without transfection. It indicated that overexpression ment. MG132 pretreatment could block this Ang II-induced effect, of Nmnat2 and H24D could significantly attenuate Ang II-induced and MG132 alone did not affect the mRNA expression of Nmnat2 as the downregulation of Nmnat2 protein level (Supplementary compared with control. Fig. S5). Then, the cellular hypertrophy response was indicated by increase in cell surface area and activation of fetal gene expression. As shown in Fig. 4C and D, overexpression of Nmnat2, but not the 4. Discussion H24D, could significantly attenuate Ang II-induced hypertrophic response. Nmnat catalyzes the last step in the biosynthesis of NAD. In After that, we took the approach of RNA interference to knock contrast to Nmnat1 and Nmnat3, which appear to be almost ubiq- down the endogenous Nmnat2 in cardiomyocytes. Quantitative uitously expressed, it was reported that Nmnat2 mRNA mainly ex- RT-PCR and Western blot was performed to evaluate the effect of pressed in high energy consumption organs including brain, heart three independent siRNAs, marked si001, si002, and si003, respec- and skeletal muscle [14], showing a close relationship between tively. As shown in Fig. 5A and B, si003 reduced the mRNA and pro- Nmnat2 and energy metabolism. Here, we revealed that Nmnat2 tein expression of Nmnat2 by 83% and 81%, respectively (P < 0.01, protein was highly expressed in brain, heart, liver and skeletal compared with control), and did not interfere with Nmnat1 and muscle of SD rats, while there was nearly no Nmnat2 detected in Nmnat3 (Fig. 5C), exhibiting potential efficacy and selectivity for kidney, lung, spleen and testis (Fig. 1A). Moreover, the expression Nmnat2 silence. Therefore, it was used in the following experi- of Nmnat2 significantly increased with age in rat hearts (Fig. 1B ments. The cell surface area and mRNA levels of ANF, BNP and and C). It suggests that the upregulation of Nmnat2 during aging b-MHC were measured after transfection. The results revealed that may be a compensatory mechanism to maintain the cardiac level the downregulation of Nmnat2 by siRNA003 could mimic the of NAD. effects of Ang II as indicated by increased cell surface area as well Recently, it has been recognized that maintaining intracellular as high expression of hypertrophic biomarkers (Fig. 5D and E). levels of NAD is essential for many crucial cell processes, including energy metabolism, intracellular Ca2+ regulation and the activity of 3.5. The anti-hypertrophic effect of Nmnat2 was dependent on sirtuins [11,18,19]. Moreover, researchers have proved that patho- activation of SIRT6, but not SIRT1 logic cardiac hypertrophy is associated with depletion of cellular NAD, and NAD treatment can block the agonist-triggered cardiac As shown in Fig. 6A, overexpression of Nmnat2 markedly increased hypertrophic response [9]. Thus, enzymes involved in NAD biosyn- cellular level of NAD (1.52-fold, P < 0.01, compared with vector group). thetic pathways have drawn increased attentions. Nampt catalyzes Ang II treatment and transfection with vector or H24D significantly re- the first reversible step in NAD biosynthesis and nicotinamide duced the NAD level in a time-dependent manner. In contrast, after (NAM) salvage [20]. It has been demonstrated that upregulation 870 Y. Cai et al. / FEBS Letters 586 (2012) 866–874

Fig. 4. Nmnat2 blocked cardiomyocyte hypertrophy through its NAD synthesis activity. The neonatal rat cardiomyocytes were transfected with Nmnat2 or catalytically inactive mutant H24D. The protein expression (A) and NAD synthesis activity (B) of Nmnat2 were detected. The cells were further treated with 100 nM Ang II for 24 h. The cell surface area (C) and mRNA levels of ANF, BNP and b-MHC (D) were measured respectively. ⁄P < 0.05 vs. control, #P < 0.05 vs. Ang II treatment, n =5. of Nampt significantly increases the NAD content and NAD-depen- mRNA level of SIRT1 could be increased by NAD directly [21]. Pillai dent SIRT1 activity [12]. However, it remains necessary to investi- et al. also revealed that the protein level of SIRT3 was markedly gate the potential effects of other NAD synthases on cellular NAD elevated after NAD treatment [9]. Although the mechanism under- level and NAD-dependent pathways related to cardiomyopathy. lying remains unclarified, it has been suggested that PPARa may In this study, we screened the mRNA expressions of all Nmnat iso- act as a transcriptional factor for SIRT1, which is activated by forms in Ang II-induced cardiomyocyte hypertrophy, and found the NAD and subsequently promotes SIRT1 expression [21]. Certainly, change of Nmnat2 was most significant (Fig. 2A). Our results also whether similar or different signaling pathways were involved in demonstrated that overexpression of wide-type Nmnat2, but not the regulation of SIRT6 still require further investigation. its catalytically inactive mutant, could block Ang II-induced cardiac The roles of sirtuins, a family of NAD-dependent histone deacet- hypertrophy (Fig. 4). Intriguingly, we showed that overexpression ylases, in aging-related diseases have been intensively explored in of Nmnat2 in cardiomyocytes significantly increased NAD content the last decade [22]. The best-characterized sirtuin, SIRT1, has been and upregulated the mRNA and protein levels of SIRT6 (Fig. 6A, C indicated to protect cardiomyocytes from oxidative stress-medi- and D). Recently, the modulatory effect of NAD on expression pro- ated cell death and retard certain cardiac degenerative changes files of sirtuins has begun to be realized. It was reported that the associated with aging [23,24]. Researchers have also revealed that Y. Cai et al. / FEBS Letters 586 (2012) 866–874 871 pharmacological activation of SIRT1 is potentially beneficial in degradation of proteins and ensure pathological consequences models of cardiovascular diseases [25]. However, although moder- [29]. And aberrations in the ubiquitin–proteasome system have ate overexpression of SIRT1 protected against oxidative stress and been implicated in pressure overload-induced cardiac hypertrophy age-dependent cardiac dysfunction, higher expression levels of [30]. By using MG-132, a specific protease inhibitor, we found that SIRT1 could lead to apoptosis and hypertrophy and decrease car- activation of proteasome was required for Ang II induced Nmnat2 diac function, suggesting a rather narrow window of optimal SIRT1 expression, which may also explain that inhibition of proteasomal activity [26]. Moreover, recent study showed that SIRT1 activation degradation increases SIRT6 levels [31]. In agreement with our re- was not involved in the regulation of NAD for cardiac hypertrophy sults, Gilley et al. have also shown that Nmnat2 is the most labile [9]. Here, by using RNA interference, we revealed that SIRT6, but Nmnat isoform that is degraded by the proteasome in HEK cells not SIRT1, participated in the anti-hypertrophic signals of Nmnat2 [32]. Our study gives the extra support to the possibility of using overexpression (Fig. 7D). This result is consistent with our previous proteasome inhibitors against cardiac hypertrophy. finding that knockdown of SIRT6 in primary cardiomyocytes led to Another noteworthy finding of the current study is that expres- obvious cellular hypertrophy (unpublished data). sion at the transcript level does not always correspond to expression In this study, the discrepancy between mRNA and protein levels at the protein level, which highlights the complexity inherent in of Nmnat2 in response to hypertrophic stimuli promoted us to transcriptional and translational regulation. In the current study, explore whether protein degradation pathway participated in the our data showed that the Nmnat2 protein expression dropped in regulation of Nmnat2 expression. It is well known that the ubiqui- cardiomyocytes after Ang II stimulation, while its mRNA level was tin–proteasome system plays a pivotal role in the degradation of increased simultaneously. This unexpected finding may suggest short-lived, regulatory proteins important for a variety of basic the existence of an autoregulatory feedback loop to maintain cellular processes, including regulation of the cell cycle, modulation Nmnat2 protein and mRNA homeostasis. In fact, a number of pro- of cell-surface receptors and ion channels, as well as antigen presen- teins have been reported to act as a regulator of its own expression tation [27,28]. The crucial role of this system in human disease is at the transcript level [33–35]. For instance, PARP-1, by binding to obvious, because its dysregulation will result in inappropriate BUR-related structures in its promoter, down-regulated its own

Fig. 5. Nmnat2 knockdown mimicked the hypertrophic responses in neonatal rat cardiomyocytes. Cells were transfected with siRNAs for Nmnat2 (marked si001, si002 and si003) and negative control (Neg). The mRNA level (A) and protein expression (B) of Nmnat2 were detected respectively. (C) The mRNA expression of Nmnat isoforms after transfection with Neg or si003. (D and E) The cell surface area and mRNA levels of ANF, BNP and b-MHC were measured in cardiomyocytes treated with 100 nM Ang II for 24 h or submitted to Nmnat2 knockdown by si003. ⁄P < 0.05 vs. control, n =5. 872 Y. Cai et al. / FEBS Letters 586 (2012) 866–874 expression [36]. In such cases, the homoeostatic control of endoge- mRNA expression of itself. Conversely, Ang II treatment leads to de- nous protein is efﬁciently and continuously kept in check through crease in Nmnat2 protein, thereby relieving the repressed gene the downregulation of transcript levels by the protein molecules transcription and resulting in elevated Nmnat2 mRNA level. MG- present in excess. Self-regulation of expression becomes particu- 132 blocks the degeneration of Nmnat2 protein induced by Ang II, larly important for proteins whose abundance results in cellular so that Nmnat2 mRNA expression may be remained at a basal level. toxicity. Thus, it may be speculated that a similar mechanism is in- Further studies concerning this point are necessarily warranted. volved in the control of Nmnat2 level and subsequent NAD synthe- In conclusion, the data presented here give the ﬁrst evidence sis. Although Nmnat2 locating mainly in cytoplasm is unable to that Nmnat2 protects cardiomyocytes from Ang II-induced hyper- regulate its own transcription by directly acting as an mRNA pro- trophy, which may involve the activation of SIRT6. It raises the pos- cessing factor, there remains a possibility that overexpressed sibility that drugs modulating Nmnat2 activity may be utilized to Nmnat2 protein, in some unknown or indirect way, suppress the treat or prevent cardiac hypertrophy.

Fig. 6. Overexpression of Nmnat2 increased SIRT6 expression and deacetylase activity in neonatal rat cardiomyocytes. Cells were transfected with Nmnat2 for 24 h. The NAD content (A) was measured. The cells were further treated with 100 nM Ang II. The intracellular NAD levels (B) and mRNA changes of all seven sirtuins (C) were measuredat indicated time points. ⁄P < 0.05 vs. vector, n = 5. (D) Representative immunoreactive bands of SIRT1 and SIRT6 after Nmnat2 transfection. (E) The SIRT6 deacetylase activity was measured in cardiomyocytes transfected with Nmnat2. ⁄P < 0.05 and ⁄⁄P < 0.01 vs. control, n = 5. (F) The intracellular distribution of Nmnat2 (red) and SIRT6 (green) were detected by confocal immunoﬂuorescence microscopy. DAPI was used to mark the nuclei. (G and H) The protein expression and deacetylase activity of SIRT6 were measured in cardiomyocytes transfected with Nmnat2, H24D or treated with 50 lM NAD. ⁄P < 0.05 vs. control, #P < 0.05 vs. Nmnat2 group and $P < 0.05 vs. H24D group, n =5. Y. Cai et al. / FEBS Letters 586 (2012) 866–874 873

Fig. 7. Nmnat2 knockdown decreased NAD content, SIRT6 protein expression and enzyme activity. Neonatal rat cardiomyocytes were transfected with si003 and negative control (Neg) for 24 h. The NAD levels (A), protein expression and deacetylase activity of SIRT6 (B and C) were measured. (D) Cardiomyocytes overexpressed with Nmnat2 after transfection with siSIRT6 or siSIRT1 were treated with Ang II for 24 h. The mRNA levels of hypertrophic biomarkers were measured by qRT-PCR. ⁄P < 0.05 vs. control, #P < 0.05 vs. Ang II treatment and $P < 0.05 vs. Nmnat2 overexpression plus Ang II treatment.

Fig. 8. Proteasome participated in Ang II-induced Nmnat expression. Neonatal rat cardiomyocytes were pretreated with 1 lM proteasome inhibitor MG132 for 1 h followed by stimulation with 100 nM Ang II. The protein and mRNA expressions of Nmnat2 were detected by Western blot (A) and qRT-PCR (B). ⁄P < 0.05 vs. control and #P < 0.05 vs. Ang II treatment, n =5.

Acknowledgments and Technology Major Project of China ‘‘Key New Drug Creation and Manufacturing Program’’ (No. 2009ZX09102-152, 2011ZX09401-307). This work was supported by grants from the National Natural Sci- ence Foundation of China (No. 81072641); NSFC-CIHR CHINA-CAN- Appendix A. Supplementary data ADA Joint Health Research Initiative Proposal (No. 330811120434); Major Project of Guangdong Province (No. 2008A030201013); Major Supplementary data associated with this article can be found, in Project of Guangzhou City (No. 2008Z1-E571) and the National Science the online version, at doi:10.1016/j.febslet.2012.02.014. 874 Y. Cai et al. / FEBS Letters 586 (2012) 866–874

References [18] Canto, C. et al. (2009) AMPK regulates energy expenditure by modulating NAD+ metabolism and SIRT1 activity. Nature 458, 1056–1060. [19] Zima, A.V., Copello, J.A. and Blatter, L.A. (2004) Effects of cytosolic NADH/ [1] Michan, S. and Sinclair, D. (2007) Sirtuins in mammals: insights into their NAD(+) levels on sarcoplasmic reticulum Ca(2+) release in permeabilized rat biological function. Biochem. J. 404, 1–13. ventricular myocytes. J. Physiol. 555, 727–741. [2] Sauve, A.A., Wolberger, C., Schramm, V.L. and Boeke, J.D. (2006) The [20] Revollo, J.R. et al. (2007) Nampt/PBEF/Visfatin regulates insulin secretion in biochemistry of sirtuins. Annu. Rev. Biochem. 75, 435–465. beta cells as a systemic NAD biosynthetic enzyme. Cell Metab. 6, 363–375. [3] Planavila, A., Iglesias, R., Giralt, M. and Villarroya, F. (2011) Sirt1 acts in [21] Hayashida, S., Arimoto, A., Kuramoto, Y., Kozako, T., Honda, S., Shimeno, H. and association with PPARalpha to protect the heart from hypertrophy, metabolic Soeda, S. (2010) Fasting promotes the expression of SIRT1, an NAD+- dysregulation, and inflammation. Cardiovasc. Res. 90, 276–284. dependent protein deacetylase, via activation of PPARalpha in mice. Mol. [4] Sundaresan, N.R., Gupta, M., Kim, G., Rajamohan, S.B., Isbatan, A. and Gupta, Cell. Biochem. 339, 285–292. M.P. (2009) Sirt3 blocks the cardiac hypertrophic response by augmenting [22] Polito, L., Kehoe, P.G., Forloni, G. and Albani, D. (2010) The molecular genetics Foxo3a-dependent antioxidant defense mechanisms in mice. J. Clin. Invest. of sirtuins: association with human longevity and age-related diseases. Int. J. 119, 2758–2771. Mol. Epidemiol. Genet. 1, 214–225. [5] Vakhrusheva, O., Smolka, C., Gajawada, P., Kostin, S., Boettger, T., Kubin, T., [23] Hsu, C.P., Odewale, I., Alcendor, R.R. and Sadoshima, J. (2008) Sirt1 protects the Braun, T. and Bober, E. (2008) Sirt7 increases stress resistance of heart from aging and stress. Biol. Chem. 389, 221–231. cardiomyocytes and prevents apoptosis and inflammatory cardiomyopathy [24] Zheng, T. and Lu, Y. (2011) Changes in SIRT1 expression and its downstream in mice. Circ. Res. 102, 703–710. pathways in age-related cataract in humans. Curr. Eye Res. 36, 449–455. [6] Lin, S.J. and Guarente, L. (2003) Nicotinamide adenine dinucleotide, a [25] Zhang, C. et al. (2011) Resveratrol attenuates doxorubicin-induced metabolic regulator of transcription, longevity and disease. Curr. Opin. Cell cardiomyocyte apoptosis in mice through SIRT1-mediated deacetylation of Biol. 15, 241–246. p53. Cardiovasc. Res. 90, 538–545. [7] Imai, S., Armstrong, C.M., Kaeberlein, M. and Guarente, L. (2000) [26] Alcendor, R.R. et al. (2007) Sirt1 regulates aging and resistance to oxidative Transcriptional silencing and longevity protein Sir2 is an NAD-dependent stress in the heart. Circ. Res. 100, 1512–1521. histone deacetylase. Nature 403, 795–800. [27] Schwartz, A.L. and Ciechanover, A. (1999) The ubiquitin-proteasome pathway [8] Borradaile, N.M. and Pickering, J.G. (2009) NAD(+), sirtuins, and cardiovascular and pathogenesis of human diseases. Annu. Rev. Med. 50, 57–74. disease. Curr. Pharm. Des. 15, 110–117. [28] Predmore, J.M. et al. (2010) Ubiquitin proteasome dysfunction in human [9] Pillai, V.B. et al. (2010) Exogenous NAD blocks cardiac hypertrophic response hypertrophic and dilated cardiomyopathies. Circulation 121, 997–1004. via activation of the SIRT3-LKB1-AMP-activated kinase pathway. J. Biol. Chem. [29] Lecker, S.H., Goldberg, A.L. and Mitch, W.E. (2006) Protein degradation by the 285, 3133–3144. ubiquitin-proteasome pathway in normal and disease states. J. Am. Soc. [10] Rongvaux, A., Andris, F., Van Gool, F. and Leo, O. (2003) Reconstructing Nephrol. 17, 1807–1819. eukaryotic NAD metabolism. BioEssays 25, 683–690. [30] Depre, C. et al. (2006) Activation of the cardiac proteasome during pressure [11] Pillai, J.B., Isbatan, A., Imai, S. and Gupta, M.P. (2005) Poly(ADP-ribose) overload promotes ventricular hypertrophy. Circulation 114, 1821–1828. polymerase-1-dependent cardiac myocyte cell death during heart failure is [31] Kanfi, Y. et al. (2008) Regulation of SIRT6 protein levels by nutrient mediated by NAD+ depletion and reduced Sir2alpha deacetylase activity. J. availability. FEBS Lett. 582, 543–548. Biol. Chem. 280, 43121–43130. [32] Gilley, J. and Coleman, M.P. (2010) Endogenous Nmnat2 is an essential [12] Hsu, C.P., Oka, S., Shao, D., Hariharan, N. and Sadoshima, J. (2009) survival factor for maintenance of healthy axons. PLoS Biol. 8, e1000300. Nicotinamide phosphoribosyltransferase regulates cell survival through [33] Lareau, L.F., Inada, M., Green, R.E., Wengrod, J.C. and Brenner, S.E. (2007) NAD+ synthesis in cardiac myocytes. Circ. Res. 105, 481–491. Unproductive splicing of SR genes associated with highly conserved and [13] Lau, C., Niere, M. and Ziegler, M. (2009) The NMN/NaMN adenylyltransferase ultraconserved DNA elements. Nature 446, 926–929. (NMNAT) protein family. Front. Biosci. 14, 410–431. [34] Ni, J.Z. et al. (2007) Ultraconserved elements are associated with homeostatic [14] Berger, F., Lau, C., Dahlmann, M. and Ziegler, M. (2005) Subcellular compartmentation control of splicing regulators by alternative splicing and nonsense-mediated and differential catalytic properties of the three human nicotinamide mononucleotide decay. Genes Dev. 21, 708–718. adenylyltransferase isoforms. J. Biol. Chem. 280, 36334–36341. [35] Wollerton, M.C., Gooding, C., Wagner, E.J., Garcia-Blanco, M.A. and Smith, C.W. [15] Phrommintikul, A., Tran, L., Kompa, A., Wang, B., Adrahtas, A., Cantwell, D., (2004) Autoregulation of polypyrimidine tract binding protein by alternative Kelly, D.J. and Krum, H. (2008) Effects of a Rho kinase inhibitor on pressure splicing leading to nonsense-mediated decay. Mol. Cell 13, 91–100. overload induced cardiac hypertrophy and associated diastolic dysfunction. [36] Vidakovic, M., Gluch, A., Qiao, J., Oumard, A., Frisch, M., Poznanovic, G. and Am. J. Physiol. Heart Circ. Physiol. 294, H1804–H1814. Bode, J. (2009) PARP-1 expression in the mouse is controlled by an [16] Li, R., Zheng, W., Pi, R., Gao, J., Zhang, H., Wang, P., Le, K. and Liu, P. (2007) autoregulatory loop: PARP-1 binding to an upstream S/MAR element and to Activation of peroxisome proliferator-activated receptor-alpha prevents a novel recognition motif in its promoter suppresses transcription. J. Mol. Biol. glycogen synthase 3beta phosphorylation and inhibits cardiac hypertrophy. 388, 730–750. FEBS Lett. 581, 3311–3316. [17] Wollert, K.C. and Drexler, H. (1999) The renin-angiotensin system and experimental heart failure. Cardiovasc. Res. 43, 838–849.