MicroRNA-539 Is Up-regulated in Failing Heart, and Suppresses O-GlcNAcase Expression*

Background: Protein O-GlcNAcylation is nearly ubiquitous; however, regulation of the expression of key enzymes remains unknown. Results: miR-539 is up-regulated in the failing heart, binds to the 3′UTR, and negatively regulates O-GlcNAcase expression. Conclusion: Protein O-GlcNAcylation can be regulated by post-transcriptional mechanisms. Significance: miR-539 regulates one of the two enzymes responsible for O-GlcNAcylation in multicellular eukaryotes. Derangements in metabolism and related signaling pathways characterize the failing heart. One such signal, O-linked β-N-acetylglucosamine (O-GlcNAc), is an essential post-translational modification regulated by two enzymes, O-GlcNAc transferase and O-GlcNAcase (OGA), which modulate the function of many nuclear and cytoplasmic proteins. We recently reported reduced OGA expression in the failing heart, which is consistent with the pro-adaptive role of increased O-GlcNAcylation during heart failure; however, molecular mechanisms regulating these enzymes during heart failure remain unknown. Using miRNA microarray analysis, we observed acute and chronic changes in expression of several miRNAs. Here, we focused on miR-539 because it was predicted to target OGA mRNA. Indeed, co-transfection of the OGA-3′UTR containing reporter plasmid and miR-539 overexpression plasmid significantly reduced reporter activity. Overexpression of miR-539 in neonatal rat cardiomyocytes significantly suppressed OGA expression and consequently increased O-GlcNAcylation; conversely, the miR-539 inhibitor rescued OGA protein expression and restored O-GlcNAcylation. In conclusion, this work identifies the first target of miR-539 in the heart and the first miRNA that regulates OGA. Manipulation of miR-539 may represent a novel therapeutic target in the treatment of heart failure and other metabolic diseases.

fine-tuning gene expression by ultimately inhibiting translation of their target mRNAs and such post-transcriptional regulation has been associated with various cardiovascular diseases (12)(13)(14). Emerging evidence indicates that the miRNA expression profile changes in a region-specific manner during myocardial ischemia (15). Several diseases, at least experimentally, are improved by oligonucleotide-based technologies, making them attractive therapeutic candidates. Here we hypothesize that miRNAs up-regulated in heart failure may suppress OGA. The combination of miRNA microarray and bioinformatics revealed miR-539 as a potential novel regulator of OGA expression. These findings have important implications for cardiovascular disease, diabetes, metabolism, and other conditions in which alterations in O-GlcNAcylation may be implicated.

EXPERIMENTAL PROCEDURES
Myocardial Infarction-Adult (3-4 months old) C57BL6/J mice were subjected to in vivo coronary ligation to induce heart failure, as described previously (11, 16 -20). In brief, mice were orally intubated, subjected to a thoracotomy, and the left coronary artery visualized and permanently occluded with the aid of a dissecting microscope. Upon recovery of spontaneous respiration, the endotracheal tube was removed and mice were allowed to recover in a temperature-controlled area supplemented with oxygen. All animal procedures were performed in accordance with federal guidelines and approved by the University of Louisville Animal Care and Use Committee.
Echocardiographic Assessment-Transthoracic echocardiography of the left ventricle was performed as previously described (16,21,22). Body temperature was maintained at 36.5-37.5°C using a rectal thermometer interfaced with a servo-controlled heat lamp throughout the procedure. Briefly, mice were anesthetized with 2% isoflurane then maintained under anesthesia with 1.5% isoflurane and examined. Using the Vevo rail system, the mouse was placed chest up on an examination board interfaced with the Vevo 770. Next, depilatory cream was applied to the chest of the mouse and wiped clean to remove all hair in the area of interest. The 707-B (30 MHz) scan head was used to obtain two-dimensional images (100 frames/s) of the parasternal long axis. M-modes images were also acquired. The probe was then rotated to acquire a short axis view of the heart. Beginning at the level of the papillary muscles and moving apically, serial two-dimensional images were taken every millimeter. All measurements were taken using the rail system Vevo 770 to maintain probe placement and allow for minute adjustments of position. Left ventricular inner diameters during diastole and systole were determined from M-modes along with heart rate. Diastolic and systolic volumes were acquired by applying Simpson's rule of discs to the serially acquired short axis images. Stroke volume was calculated as: diastolic volume Ϫ systolic volume. Ejection fraction was calculated as: (stroke volume/diastolic volume) ϫ 100%. Cardiac output was determined by: stroke volume ϫ heart rate.
Masson's Trichrome Staining-After 5 and 28 days (to assess early and late stage heart failure), the infarcted and sham-operated mice were euthanized, and the hearts were rapidly excised and fixed for immunohistochemical analysis or immediately frozen in liquid nitrogen and stored at Ϫ80°C. In brief, after excision, sham and infarcted mouse hearts (5 and 28 days) were fixed in 4% paraformaldehyde for 1 h at 4°C, then kept in PBS with 30% sucrose overnight at 4°C, embedded in optimal cutting temperature compound (Sakura Finetek, Torrance, CA), sectioned (10 m) with cryostat (Leica, Germany), and placed on superfrost slides. The degree of myocardial infarction (MI) was evaluated using Masson's trichrome staining kit according to manufacturer's protocol (Sigma).
miRNA Microarray and Real-time PCR-Total RNA from the 5-and 28-day sham and infarcted mouse hearts (n ϭ 4) was isolated using TRIzol reagent (Invitrogen). Rodent miRNA microarray kit (Applied Biosystems) was used according to the manufacturer's protocol. In brief, 1 g of total RNA was reverse-transcribed with Megaplex RT primers (Megaplex RT Rodent Pool A), followed by a real-time PCR with TaqMan Rodent MicroRNA Array performed on an Applied Biosystems 7900HT System. SDS software version 2.3 and DataAssist version 3.0 (Applied Biosystems) were used to obtain the comparative threshold cycle (C t ) value. U6 small nuclear RNA included in the TaqMan Rodent MicroRNA Array was used as an endogenous control. Quantitative RT-PCR (qRT-PCR) analyses were carried out using TaqMan miRNA assays (Applied Biosystems) according to the manual. Relative expression of miR-539 was calculated using the ⌬⌬C T method normalized to the expression of U6 small nuclear RNA (Applied Biosystems). Relative levels of OGA and OGT mRNA were examined with specific primers (OGA: forward, 5Ј-TGGAAGACCTTGGGT-TATGG-3Ј and reverse, 5Ј-TGCTCAGCTTCTTCCACTGA-3Ј; OGT, forward, 5Ј-CCTGGGTCGCTTGGAAGA-3Ј and reverse, 5Ј-TGGTTGCGTCTCAATTGCTTT-3Ј) using Fast SYBR Green (Applied Biosystems) and normalized to levels of 18 S mRNA. All qRT-PCRs were performed in duplicate.
Cell Culture-Neonatal rat cardiomyocytes (NRCMs) were isolated from 1-2-day-old Sprague-Dawley rats according to the protocol described elsewhere (23). The isolated cardiomyocytes were cultured in DMEM containing 10% fetal bovine serum, penicillin/streptomycin, and vitamin B 12 in the presence of anti-mitotic BrdU (0.1 mM) for 4 days to inhibit fibroblast growth and subsequently grown in the absence of BrdU. HEK293 cells were grown in DMEM containing 10% fetal bovine serum and penicillin/streptomycin. 293FT cells cultured in DMEM Glutamax (Invitrogen) containing 10% fetal bovine serum, penicillin/streptomycin, and Geneticin (Invitrogen) were used for the lentivirus preparation.
Functional Study of miRNA-539 -We examined the effect of miR-539 on OGA and OGT by transducing NRCMs and HEK293 cells with miR-539 lentivirus or scrambled or mCherry control lentiviruses. The cells were maintained in their regular growth medium for 2 days and selected with blasticidine for 5 days. Alternatively, cells were transfected with mirVana TM miR-539 inhibitor or an negative control Oligonucleotides (10 nmol/liter) using Lipofectamine TM RNAiMAX (Invitrogen) after 5 days of lentivirus treatment. Overexpression of miR-539 expression was verified by qRT-PCR using miRNA assay as described above and the loss or gain of OGA expression followed by miR-539 mimic/inhibitor was analyzed by Western blotting.
In Vitro Hypoxia-reoxygenation Injury-To determine whether additional insults, such as hypoxia-reoxygenation alter miR-539 levels, NRCMs were subjected to hypoxia-reoxygenation as described (9). In brief, cells were subjected to 3 h hypoxia in Esumi lethal ischemia medium for glucose and nutrient deprivation (containing 117 mmol/liter of NaCl, 12 mmol/liter of KCl, 0.9 mmol/liter of CaCl 2 , 0.49 mmol/liter of MgCl 2 , 4 mmol/liter of HEPES, 20 mmol/liter of sodium lactate, and 5.6 mmol/liter of L-glucose; pH 6.2) in a sealed humidified hypoxic chamber (Billups-Rothenberg, Inc.) flushed with 5% CO 2 and 95% N 2 , and maintained at 37°C. After 3 h, the cells were switched to Esumi control medium (containing 137 mmol/liter of NaCl, 3.8 mmol/liter of KCl, 0.9 mmol/liter of CaCl 2 , 0.49 mmol/liter of MgCl 2 , 4 mmol/liter of HEPES, and 5.6 mmol/ liter of D-glucose, pH 7.4) and allowed to reoxygenate for 3, 6, and 12 h in the modular incubator. Cells grown in the Esumi control medium for the same durations were used as normoxic controls. The expression of miR-539 and OGA protein level was analyzed as described above.
Statistical Analysis-Unpaired Student's t test was used for the comparison between two groups. Analysis of variance was performed when comparisons were made among three or more groups along with Scheffe's post hoc analysis. p Ͻ 0.05 was accepted as statistically significant.

Myocardial Infarction and Cardiac Function-Masson's
trichrome staining showed distinct blue collagen fibers and red myocytes in the sham and infarcted mouse hearts (Fig. 1, A-D). At 5 days the infarcted heart showed significantly decreased wall thickness with increased collagen deposition, which was greater at 28 days compared with the sham heart (Fig. 1, B and  D). Echocardiography was performed at 5 and 28 days post-MI to demonstrate the severity of cardiac dysfunction. At 5 days after MI, hearts exhibited significant dilation (Fig. 1, E-H) and severely depressed cardiac function, as evidenced by significantly decreased stroke volume, ejection fraction, and cardiac output (Fig. 1, I-L) when compared with sham-operated mice. Indeed, this defect persisted; echocardiography confirmed dysfunction 28 days after MI (Fig. 1, E-L).
miRNA-539 Is Up-regulated in Failing Heart-Global miRNA profiling analysis revealed that several miRNAs are differentially expressed (Ͻ0.7-or Ͼ2-fold) in 5-and 28-day failing hearts compared with sham heart (Table 1). To identify the miRNAs that target HBP, we performed target prediction anal-

miR-539 Suppresses O-GlcNAcase
ysis (TargetScan and MiRanda algorithms) and found a predicted miR-539 binding site in OGA-3ЈUTR (26). The miR-539 binding sites in the OGA-3ЈUTR are conserved among mouse, rat, and human ( Fig. 2A). We performed qRT-PCR to confirm the miRNA microarray data. Specifically, miRNA-539 was upregulated ϳ5and ϳ3-fold (p Ͻ 0.05) in 5 and 28 days, respectively, in the infarcted heart compared with sham heart (Fig.  2B). Further in situ hybridization analysis with the miR-539 LNA probe revealed increased expression of miR-539 in the remote zone (non-infarcted) of the failing heart compared with the sham heart (Fig. 2C). Interestingly, a strong signal of miR-539 was observed in 5-day compared with 28-day failing heart and the sham hearts showed weak signals at both 5 and 28 days. Histomorphological observation also showed the perinuclear localization of miR-539 in the cardiomyocytes. Collectively these data demonstrate that miR-539 was significantly up-regulated in the failing heart by 5 days and through at least 28 days.

Reduced OGA Expression and Increased O-GlcNAcylation in the Failing
Heart-Use of qRT-PCR analysis showed significant down-regulation of OGA mRNA and protein levels in both 5-and 28-day infarcted hearts compared with sham hearts (Fig. 3, A and  B). A significant increase in OGT mRNA expression was observed FIGURE 2. Up-regulation of miR-539 in the failing heart. A, miRNA-539 binding sites in the OGA-3ЈUTR are conserved among mouse, rat, and human. TargetScan 6.2, where human OGA-3ЈUTR has two putative miR-539 binding sites. B, expression of miR-539 was measured by qRT-PCR and normalized to the expression of U6 in each sample. The failing heart shows significant up-regulation of miR-539 in 5 (p Ͻ 0.001) and 28 day (p Ͻ 0.05) compared with sham heart. C, probing of miR-539 LNA in the cryosections using in situ hybridization shows increased miR-539 expression (in purple) in the remote zone of the failing heart (II and IV) compared with the sham heart (I and III). Histological observation also reveals perinuclear localization of miR-539 in the cardiomyocytes (insets). Low magnification images were taken at ϫ20; insets are at ϫ60 magnification. Scale bar ϭ 10 m. OCTOBER 24, 2014 • VOLUME 289 • NUMBER 43 in 5-and 28-day infarcted hearts, which was accompanied by a significant increase in OGT protein at 5 days, but not 28 days (Fig.  3, C and D). In addition, there was a significant increase in protein O-GlcNAcylation at both 5 and 28 days in infarcted hearts (Fig.  3E).

miR-539 Suppresses O-GlcNAcase
miRNA-539 Targets OGA-3ЈUTR-To test whether miR-539 binds directly to the OGA-3ЈUTR as predicted, we co-transfected reporter plasmid containing luciferase upstream to the OGA-3ЈUTR or miR-539 binding site mutant (Fig. 4A) with scrambled control or miRNA-539 expression plasmids in HEK293 cells. Co-transfection of wild-type OGA-3ЈUTR with miR-539 significantly reduced luciferase activity (Ͼ40%) compared with scrambled control. In contrast, co-transfection of the miR-539 binding site mutant with miR-539 did not affect the luciferase activity (Fig. 4B). These results suggest that miR-539 binds to the OGA-3ЈUTR, thereby providing a potential mechanism to reduce OGA expression.
miRNA-539 Modulates OGA Expression in Vitro-To determine whether overexpression of miR-539 suppresses OGA expression, we transduced NRCMs and HEK293 cells with miR-539 lentivirus to overexpress miR-539. Real-time PCR analysis showed significant up-regulation of miR-539 in both NRCMs (ϳ170-fold) and HEK293 cells (ϳ35,000-fold) transduced with miR-539 lentivirus compared with scrambled virus (Fig. 4, C and E). Immunoblot analysis showed ϳ40% reduction in OGA expression in miR-539 overexpressed NRCMs compared with scrambled control or mCherry-transduced NRCMs (Fig. 4D), whereas HEK293 showed ϳ30% reduction in OGA expression in miR-539 overexpressed cells compared with scrambled control (Fig. 4F). Furthermore, transfection with the miR-539 inhibitor rescued OGA protein expression in miR-539 overexpressing NRCMs (Fig. 5, A and B) and HEK293 cells (Fig.  6, A and B) compared with negative control inhibitor-treated cells. Such findings indicate that overexpression of miR-539 significantly suppresses OGA expression in NRCMs. Interestingly, we found the OGT protein levels were also significantly diminished following miR-539 overexpression in this in vitro system (Figs. 5, C and D, and 6, C and D).
Effect of miRNA-539 on O-GlcNAcylation-To assess the functional consequence of miR-539-mediated OGA regulation, miR-539 overexpressed or expression-silenced NRCMs and HEK293 cells were assessed for O-GlcNAc levels. Overexpres- A, OGA expression was measured by qRT-PCR and normalized to the expression of 18 S rRNA, which shows significant down-regulation in the failing heart compared with the sham heart at 5 and 28 days; p Ͻ 0.0001. B, Western blot analysis from the 5-and 28-day failing heart shows significant down-regulation of OGA at the protein level (p Ͻ 0.01). C, qRT-PCR analysis shows significant up-regulation of OGT in the failing heart compared with sham heart at 5 and 28 days; p Ͻ 0.0001. D, Western blot analysis shows and increased OGT protein level in 5-day failing heart (p Ͻ 0.0001) but not at 28 days. E, significant increase in protein O-GlcNAcylation compared with sham heart; p Ͻ 0.001. Data are expressed as the mean Ϯ S.E. (n ϭ 4/group). IB, immunoblot.

miR-539 Suppresses O-GlcNAcase
sion of miR-539 significantly increased O-GlcNAc levels, whereas inhibition of miR-539 in miR-539 stably expressing NRCMs and HEK293 cells reduced O-GlcNAc to normal levels (Figs. 5, E and F, and 6, E and F).
Hypoxia-reoxygenation Increases miR-539 and Decreases OGA Expression-To determine whether miR-539 might play a role in regulating OGA expression in an in vitro proxy for a pathologic condition, we subjected NRCMs to hypoxia-reoxygenation. Indeed, the level of miR-539 was significantly increased after 3 h of reoxygenation with a decrease in the OGA protein level after 6 and 12 h of reoxygenation (Fig. 7, A and B). Thus, miR-539 is up-regulated in NRCMs following hypoxia, and is associated with a reduction in OGA expression.

DISCUSSION
More than 1,000 nuclear and cytoplasmic proteins are known to be O-GlcNAcylated, including targets important for transcription, translation, signal transduction, and cell cycle control (27)(28)(29)(30). In response to various stresses, global augmentation of protein O-GlcNAcylation confers a survival advantage at the cellular level. Several in vitro and in vivo analyses showed O-GlcNAc-mediated cardioprotection against ischemia-reperfusion, myocardial infarction, and oxidative stress (4,7,9,(31)(32)(33). The present study indicates that down-regulation of OGA may be involved, at least partially, in increased O-GlcNAcylation in non-reperfused myocardial infarction. . miR-539 regulates OGA expression. A, reporter plasmid containing luciferase upstream to the OGA-3ЈUTR was obtained from Genecopoeia and the miR-539 binding site seed sequence (AUUUCUC) was mutated (ACGCGAG) by site-directed mutagenesis using mutated forward and reverse primers. B, luciferase activity was measured after 48 h from HEK293 lysates co-transfected with OGA-3ЈUTR or miR-539 binding site mutant with scrambled or miRNA-539 expression plasmids. Luciferase activity was normalized with internal Renilla control. Significant down-regulation of luciferase activity was determined by miR-539 when co-transfected with OGA-3ЈUTR, whereas the miR-539 binding site mutation was unaffected; p Ͻ 0.05. C and E, qPCR analysis a shows significant level of miR-539 expression in NRCMs and HEK293 cells transduced with lentivirus encoding miR-539 (n ϭ 3/group) compared with mCherry-transduced cells; p Ͻ 0.0001. D and F, Western blot analysis and respective quantitative analysis shows around 40 (p Ͻ 0.005) and 30% (p Ͻ 0.05) reduction of OGA expression in miR-539 overexpressing NRCMs and HEK293 cells, respectively, compared with scrambled controls (n ϭ 3/group). IB, immunoblot.

miR-539 Suppresses O-GlcNAcase
Numerous gain-and loss-of-function studies have uncovered prominent roles for miRNAs in human cardiovascular disorders including myocardial infarction, cardiac hypertrophy, heart failure, fibrosis, and pressure overload-induced remodeling (34 -38). Few studies focused on miRNAs implicated in metabolic disorders (39,40), yet, there are no prior studies   OCTOBER 24, 2014 • VOLUME 289 • NUMBER 43 demonstrating miRNA-mediated regulation of OGA (or OGT) expression. In this study, we report a novel paradigm of miRNA-mediated down-regulation of OGA with subsequent augmentation of O-GlcNAcylation in the failing heart. Use of miRNA microarray followed by qRT-PCR analysis clearly demonstrated that miR-539 was up-regulated in the failing heart.

miR-539 Suppresses O-GlcNAcase
Significant down-regulation of OGA in miR-539 overexpressing NRCM (rat) and HEK293 (human) cells indicates that OGA is a potential conserved target of miR-539. Although OGT does not have target sites for miR-539 binding, reduction of the OGT protein level by miR-539 overexpression reveals that there could be indirect mechanisms involved in the regulation of OGT by OGA or other targets of miR-539. Increased OGA expression with concomitant reduction in O-GlcNAcylation using the miR-539 inhibitor in both NRCMs and HEK293 cells overexpressing miR-539 indicates that miR-539 may be a potential marker/target of disease or other prognostic indicator.
MicroRNA-539 is present in the miRNA-rich intragenic region of mouse chromosome 12 (NC_000078.6 (109728129 to 109728202)) between Rian (RNA imprinted and accumulated in nucleus) and Mirg (miRNA containing gene). There are few reports of verified targets of miR-539. Recently, Milosevic et al. (41) observed TGF-␤1-mediated stimulation of the cluster of microRNAs from chr14q32 (mouse analog; chr12qF1) in normal human lung fibroblasts where miR-539 is one among them; however, the exact mechanism of miR-539 expression per se requires further investigation. Interestingly, a recent study indicates that miR-539 regulates mitochondrial fission and apoptosis by targeting prohibitin 2 (PHB2) in cardiomyocytes (42). It is possible that miR-539 plays an even larger role than simply regulating OGA expression. In mast cells, CD117 (i.e. KIT or c-kit) represses miR-539 expression, thereby de-repressing microphthalmia-associated transcription factor expression and promoting proliferation (43). In other cell types, miR-539 expression may be biotin-sensitive, and miR-539 targets the mRNA of holocarboxylase synthase, which participates in genetic stability (44). Such collective findings create interesting implications for a more detailed understanding of the molecular interactions governing O-GlcNAc-dependent cell function; however, there is no evidence, at present to directly link the aforementioned observations to one another, at least at the level of O-GlcNAcylation.
It is also tempting to speculate whether miR-539 plays a role in other cardiovascular diseases in which O-GlcNAc appears to be involved. For example, during hypertension, inhibition of OGA increases arterial vascular reactivity (45). Another study demonstrated that OGA inhibition improved cardiac function during reperfusion (46). Our results demonstrate that up-regulation of miR-539 might be a possible molecular mechanism that suppresses OGA expression in the infarcted, failing heart, which, in turn favors O-GlcNAcylation. The increase in miR-539 followed by a decrease in OGA protein level during hypoxia-reoxygenation mimics the phenomenon seen in the present in vivo myocardial infarction data. Future studies will directly interrogate the role of miR-539 in the evolution of disease and subsequent recovery.
Understanding the regulation of O-GlcNAcylation holds critical importance not only in heart failure but also in multiple diseases; no disease exemplifies the potential impact better than diabetes. Several studies suggest that elevated O-GlcNAc levels contribute to diabetic cardiomyopathy (47). In humans, the OGA/MGEA5 chromosome locus 10q24.1 is associated with late-onset Alzheimer disease (48). Experimental observations demonstrated that inhibition of OGA decreases phosphorylation of Tau and protects against Tau-mediated neurodegeneration as well as prevents amyloid-␤ load by increasing the amount of secreted amyloid precursor protein (sAPP␣) (49,50). This line of evidences suggests that increased O-GlcNAcylation by OGA inhibition improves neuronal outcome.
In addition, numerous studies showed hyper O-GlcNAcylation, increased expression of OGT, and decreased OGA expression in various cancers (51,52). In part, hyper O-GlcNAcylation was also observed as a mechanism that promotes cancer cell survival and stress resistance. Similar to the approach of targeting FIGURE 7. Hypoxia-reoxygenation induces miR-539 expression followed by a decrease in OGA expression. NRCMs subjected to 3 h of hypoxia were reoxygenated for 3, 6, and 12 h. qRT-PCR analysis shows a significant increase in miR-539 expression at 3 h reoxygenation; p Ͻ 0.05. Western blot analysis shows a significant decrease in OGA protein level at both 6 and 12 h of reoxygenation (n ϭ 4/group); p Ͻ 0.001. IB, immunoblot.

miR-539 Suppresses O-GlcNAcase
kinases, targeting OGA and OGT could be a valuable approach in many cancer therapies. Thus, the implications for our findings with miR-539 could be broad and manifold.
In light of results from this study, it will be interesting to see whether miR-539 expression is altered in these diseases and the use of miR-539 mimic or inhibitor may contribute a new and broad therapeutic approach to modulate O-GlcNAc signaling. Moreover, it is time to more carefully investigate the transcriptional and post-transcriptional regulation of OGT and OGA in relevant disease models.