Analysis of microrna expression during the torpor-arousal cycle of a mammalian hibernator, the 13-lined ground squirrel

Transcription

1 Physiol Genomics 48: , First published April 15, 2016; doi: /physiolgenomics Analysis of microrna expression during the torpor-arousal cycle of a mammalian hibernator, the 13-lined ground squirrel Cheng-Wei Wu, Kyle K. Biggar,* Bryan E. Luu,* Kama E. Szereszewski, and Kenneth B. Storey Institute of Biochemistry and Department of Biology, Carleton University, Ottawa, Ontario, Canada Submitted 6 January 2016; accepted in final form 4 April 2016 Wu CW, Biggar KK, Luu BE, Szereszewski KE, Storey KB. Analysis of microrna expression during the torpor-arousal cycle of a mammalian hibernator, the 13-lined ground squirrel. Physiol Genomics 48: , First published April 15, 2016; doi: /physiolgenomics Hibernation is a highly regulated stress response that is utilized by some mammals to survive harsh winter conditions and involves a complex metabolic reprogramming at the cellular level to maintain tissue protections at low temperature. In this study, we profiled the expression of 117 conserved micrornas in the heart, muscle, and liver of the 13-lined ground squirrel (Ictidomys tridecemlineatus) across four stages of the torpor-arousal cycle (euthermia, early torpor, late torpor, and interbout arousal) by real-time PCR. We found significant differential regulation of numerous micrornas that were both tissue specific and torpor stage specific. Among the most significant regulated micrornas was mir-208b, a positive regulator of muscle development that was found to be upregulated by fivefold in the heart during late torpor (13-fold during arousal), while decreased by 3.7-fold in the skeletal muscle, implicating a potential regulatory role in the development of cardiac hypertrophy and skeletal muscle atrophy in the ground squirrels during torpor. In addition, the insulin resistance marker mir-181a was upregulated by 5.7-fold in the liver during early torpor, which supports previous suggestions of hyperinsulinemia in hibernators during the early stages of the hibernation cycle. Although microrna expression profiles were largely unique between the three tissues, GO annotation analysis revealed that the putative targets of upregulated micrornas tend to enrich toward suppression of progrowth-related processes in all three tissues. These findings implicate micrornas in the regulation of both tissue-specific processes and general suppression of cell growth during hibernation. hibernation; metabolic depression; stress resistance; cryobiology; noncoding RNAs * K. K. Biggar and B. E. Luu contributed equally to this work. Address for reprint requests and other correspondence: K. B. Storey, Inst. of Biochemistry and Dept. of Biology, Carleton Univ., 1125 Colonel By Dr., Ottawa, ON, K1S 5B6, Canada ( kenneth_storey@carleton.ca). HIBERNATION IS AN ADAPTIVE response utilized by many mammals when confronted with the harsh and unfavorable conditions of winter. Hibernators enter into a state of dormancy, and this phenotypic plasticity is characterized by drastic physiological changes, which includes the reduction of body temperature (T b )to near ambient (0 5 C), metabolic depression to 5% of resting euthermic rate, and reduced heart (from to 3 5 beats/ min) and respiratory rates (from to 4 6 breaths/min) (10). At the cellular level, global control of cellular energetics is coordinated toward a state of hypometabolism, evident by the reduction in ATP-driven transmembrane ion pumping, inhibition of global mrna transcription and protein translation, and a switch toward reliance on lipid as the primary source of metabolic fuel (13, 26, 30, 38). The regulation of these cellular functions has been shown to be controlled at the posttranslational level, with reversible protein phosphorylation shown to play an integral role in the regulation of key glycolytic enzymes, histone modifications, RNA polymerase II activity, and protein translation initiation (17, 30, 39). Recent discoveries of micrornas (mirnas) have introduced a new dimension of cellular regulation that is highly conserved among species ranging from nematodes, fruit flies, to human (3). MiRNAs are small noncoding RNA transcripts that are 22 nucleotides in length and are known to exert posttranscriptional control by binding to target mrnas near the 3=-untranslated region (UTR) to promote translational silencing through either sequestration or degradation. Transcripts targeted by mirnas have been shown to localize to cytoplasmic foci, which can serve as sites for mrna storage or degradation leading to translational repression (23). We have recently shown evidence for the formation during hibernation of stress-induced granules that comprised RNA-binding proteins, and these could serve as potential mrna storage foci that would complement the regulatory roles of mirnas during torpor (40). A single mirna can regulate hundreds of genes, and a single gene can be targeted by multiple mirnas, creating a complex network that is thought to regulate up to 60% of all protein-coding genes in human (21). We have previously reported the regulatory roles of mirnas during hibernation and have begun to show mirna regulation as part of a global response to other environmental stressors that include estivation, anoxia, and freezing, with select mirnas shown to have a regulatory influence on stress-regulated cellular processes that include apoptosis, cell cycle progression, signal transduction, and lipid metabolism (4 6, 42). The hibernation cycle in ground squirrels is characterized by multiple torpor-arousal bouts that typically last 2 4 wk in duration, with each torpor bouts consisting of an entry and arousal from a full metabolically depressed state. The euthermic transition to and from torpor is a multistage process, although many studies have focused on the molecular changes between euthermia and deep torpor, little is known about molecular regulations between other transitional torpor stages. In the present study, we profiled the expression of 117 known and conserved mammalian mirnas during four distinct stages of the torpor-arousal cycle in the heart, muscle, and liver of the 13-lined ground squirrel Ictidomys tridecemlineatus (Fig. 1). We show that although the patterns of mirna regulation are largely unique between the different tissues examined, enrichment analysis of predicted mirna target genes indicates that common cellular processes such as cell growth and proliferation are among the top mirna-targeted processes during torpor in all tissues. The results gathered here provide strong evidence for mirnas as an important molecular regulators of hibernation in the ground squirrel /16 Copyright 2016 the American Physiological Society

2 microrna REGULATIONS IN THE TORPOR-AROUSAL CYCLE 389 MATERIALS AND METHODS Fig. 1. Model of the torpor-arousal cycle in the hibernating ground squirrels. Four distinct torpor stages were analyzed for mirna expression profile. Euthermic control group were torpor-ready animals with a body temperature (T b) of C; early and late torpor animals both had a T b of 5 8 C, with the former maintaining this temperature for 24 h and latter for 5 days. Interbout arousal animals underwent a successful torpor bout and recovered at at b of C for 18 h after 3 days of torpor. Animal treatments. Protocols for animal treatment were the same as previously described in detail (28). The 13-lined ground squirrels were captured by USDA-licensed personnel (TLS Research, Bartlett, IL) and transported to the National Institute of Neurological Disorder and Stroke (NINDS) at the National Institutes of Health (Bethesda, MD). All protocols used for animal care were approved by the NINDS animal care and use committee (protocol no. ASP ). Using a temperature transponder (IPTT-300, Bio Medic Data System) injected in the intrascapular region of the ground squirrels, we monitored T b of individual animals to determine the stages of torpor. Euthermic control (EC) animals used as control group were kept in a cold room (4 5 C) for 3 days while maintaining a T b of C. EC animals used were capable of entering full torpor but had not done so in the past 72 h. For experimental animals, early torpor (ET) was indicated byat b of 5 8 C for 24 h, late torpor (LT) was indicated by a T b of 5 8 C for 5 days, and interbout arousal (IA) was indicated by at b that returned to C for at least 18 h after 3 days of torpor. Ground squirrel heart, muscle, and liver tissues were collected and shipped on dry ice to Carleton University and stored at 80 C until use. MiRNA expression analysis by quantitative PCR. Extraction of total RNA from heart, muscle, and liver of EC, ET, LT, and IA ground squirrels was carried out as previously described in detail (44). Polyadenylation of mirnas was carried out with the Poly (A) Polymerase Tailing kit (Epibio, Madison, WI) according to the manufacturer s instructions. cdna synthesis of polyadenylated mirnas was performed as previously described using a universal stem-loop RT adapter (7). Real-time PCR was performed with a Bio-Rad MyiQ2 Detection System (Bio-Rad, Hercules, CA) using mirna-specific primers (available upon request) and the universal primer (7). Melting curve analyses were performed for each mirna as quality check. MiRNA identification and functional annotation. Highly conserved mirnas were identified from the I. tridecemlineatus reference genome using a downloaded database of known mirna sequences from Mus musculus (mirbase v.21) as a search template. First, the precursor mirna (pre-mirna) sequences from M. musculus were searched against I. tridecemlineatus WGS traces downloaded from the National Center for Biotechnology Information using megablast. The resulting sequences were refined for sequences with nucleotide mismatches 10, sequence gaps 3, sequence identity 90%, and an e-value 6.00E 20. Next, the mature mirnas corresponding to the identified squirrel pre-mirnas were downloaded from the mature mirna database for M. musculus (mirbase v.21) and used as a search template. This new database was searched against a database consisting of the identified I. tridecemlineatus pre-mirna sequences using megablast (optimized for short sequences). The resulting sequences were searched for mature mirna sequence matches with a sequence start within the first 2 nt of the template sequence and an e-value Data reported are averages SD. To predict I. tridecemlineatus-specific targets, mature mirnas were searched against the 3=-UTR sequences from the I. tridecemlineatus reference genome available in the UCSC table browser (Broad/speTri2) using miranda (v.3.3a) (15). To identify mirna target genes, miranda software was used within the following parameters and conditions: a gap opening penalty of 9, a gap extension penalty of 4; match with minimum score threshold 160, target duplex with maximum threshold free energy 25 kcal/mol, scaling parameter 4 for complementary nucleotide match score, counting from the mirna 5=-end, and demand for a strict 5=-seed pairing on between 2 and 9 nt. To determine the functional targeting of mirna that have altered expression in response to any specific torpor stage, all identified targets were classified based on biological processes defined by gene ontology (GO) annotation for biological processes (11). Reactome software was used to determine the torpor-responsive mirna targets that were enriched into pathway networks of regulation (29). The Reactome analysis used the available M. musculus network annotations. Data and statistical analysis. Cycle threshold (Ct) of each mirna was normalized to the Ct of endogenous control 5S rrna amplified from the same sample. The comparative Ct method was used to determine relative expression each mirna. Data presented in the heat map are expressed as mean expression from n 4 independent samples from different animals from each sampling point. Statistical analysis of mirna expression was determined by one-way ANOVA in the SigmaPlot 12.0 statistical package; multiple test comparisons were applied with the Bonferroni correction, with P considered a significant change compared with EC. RESULTS MiRNA expression patterns are unique between tissues during the torpor-arousal cycle. To investigate the involvements of mirnas in the regulation of the torpor-arousal cycle, we profiled 117 known mammalian mirnas at four distinct torpor stages of the ground squirrel hibernation cycle including EC, ET, LT, and IA (Fig. 1) in three tissues using real-time PCR. The two torpor stages (ET and LT) represent the lowest point of metabolic rate and T b reduction compared with the euthermic control yet are distinct from one another with ET just entering a metabolic depressed state and LT considered a fully established state of hypometabolism. While ground squirrels at the IA stage exhibit a stabilized metabolic rate and T b that are similar to EC, it is considered a recovery stage as euthermia has only recently been established for a short duration after a torpor period of 3 days. These four sampling points were chosen as the most contrasting stages of the torpor-arousal cycle. In the heart, ANOVA revealed that a total of 29 mirnas were differentially expressed during ET, 36 during LT, and 34 during IA, compared with the EC stage (Fig. 2A). In skeletal muscle, a total of 21 mirnas were differentially expressed during ET, 39 during LT, and 33 during IA torpor stages (Fig. 2B). Compared with the heart and muscle tissues, liver showed a different overall pattern of mirna expressions (Fig. 2C). During the ET and IA stages, 28 and 26 differentially expressed mirnas were found respectively, with only five differentially expressed during LT (Fig. 2C). Interestingly, we identified several mirnas that were highly overexpressed. In the heart, mir (fold-change of ), mir-208b (fold-change of ), and mir-375 (fold-change of ) were all upregulated by 10-fold during IA compared with EC (Supple-

3 390 microrna REGULATIONS IN THE TORPOR-AROUSAL CYCLE

4 microrna REGULATIONS IN THE TORPOR-AROUSAL CYCLE 391 Fig. 3. Unique mirnas are regulated in each tissue during torpor. Number of shared mirnas significantly upregulated (A) or downregulated (B) between the heart, muscle, and liver tissue at each torpor stage of the ground squirrel. mental Table S1). 1 In the muscle, mir-148b, mir-152, and mir- 193a were found to be upregulated by , , and fold, respectively during ET, while mir-152 was found to be upregulated by fold during IA (Supplemental Table S1). Surprisingly there were very few overlaps in mirnas that were differentially expressed in all three tissues during the torpor stages examined (Fig. 3, A and B), with only a few downregulated mirnas that were found to be shared between the skeletal muscle and the heart during LT (Fig. 3B). The absence of a common torpor-responsive mirna illustrates the specificity of mirna expression during hibernation between tissues. To determine if each torpor stage is regulated by different mirnas, we analyzed the overlaps in differentially expressed mirnas between the torpor stages within each tissue (Fig. 4). In this analysis, we also found that each torpor stage show unique expression patterns of mirnas, with the exception of downregulated mirnas in the heart, which had 1 The online version of this article contains supplemental material. 16 shared mirnas at all three time points. Interestingly, although both ET and LT are similar in terms of their relative T b and metabolic depression, there were very few overlaps in uniquely shared mirnas between these two time points in all three tissues (Fig. 4). Upregulated mirnas target genes enriched for positive regulation of metabolic processes. MiRNAs, when upregulated, have generally been shown to have an inhibitory effect on expression of the target genes through posttranscriptional controls and, when downregulated, could provide a potential relief in transcript repression of the target gene (3). To better understand the potential roles for mirna in the torpor-arousal cycle, we performed a prediction analysis for all the up- and downregulated mirnas using the miranda algorithm to identify its potential target genes in the ground squirrel (Supplemental Table S2) and implemented a GO enrichment analysis of the putative target genes to identify biological processes that were regulated by the differentially expressed mirnas (15, 41). Several of the predicted upregulated mirna target genes from our analysis have previously been shown to be down- Fig. 2. Summary heat maps of microrna (mirna) fold-changes in the ground squirrel during the torpor-arousal cycle. Relative levels of mirnas are shown as a heat map for skeletal muscle (A), heart (B), and liver (C) at the 3 experimental stages of torpor (early torpor, late torpor, interbout arousal) compared with euthermic control. Differentially expressed mirnas are highlighted in brackets; n 3 4 independent samples from different animals, P were considered significant with 1-way ANOVA.

5 392 microrna REGULATIONS IN THE TORPOR-AROUSAL CYCLE Fig. 4. Overlaps of mirna regulated in each tissue at the 3 torpor stages. Number of shared mirnas significantly upregulated (A) or downregulated (B) between each torpor stage in the heart, muscle, and liver tissues of the ground squirrel. ET, early torpor; LT, late torpor; IA, interbout arousal. regulated in ground squirrels during hibernation, demonstrating the anticipated negative correlation between mirna levels and the target gene expression. The predicted mirna target genes that were downregulated in the liver during LT included: znf66 (zinc finger protein 644), dnajb9 [DnaJ (Hsp40) homolog], ubr4 (ubiquitin protein ligase E3 component N-recognin 4), zdhhc17 (zinc finger, DHHC-type containing 17), b3gat3 (beta-1,3-glucuronyltransferase 3), brf1 (RNA polymerase III transcription initiation factor 90 KDa subunit), sox9 [SRY (sex determining region Y)-box 9], and ddhd19 (DDHD domain containing 1), all of which were previously reported to be significantly downregulated during torpor in the liver tissue by Xu et al. (45). In addition, similar changes in transcript levels of mirna predicted targets genes were also found for the skeletal muscle and heart in previous genomewide transcriptomic studies (Supplemental Table S3) (19, 45). Interestingly, although different sets of mirnas were found to be upregulated between the heart and muscle tissues, GO analysis of putative mirna targets showed enrichment toward regulation of some similar categories of cellular processes in both tissues (Fig. 5). These include positive regulation of cell growth, cell-cell junction organization, and microtubule cytoskeleton organization during the ET and LT stages. In contrast, for the genes that were targeted by downregulated mirnas, GO analysis showed enrichment toward unique cellular processes between the two tissues. The muscle tissue showed enrichment toward biological processes such as amino acid transport and protein and macromolecule complex assembly, while the heart tissue showed and enrichment toward the regulation of cell cycle, nucleic acid transport, and regulation of nucleic acid processes (Fig. 5B). For liver tissue, upregulated mirnas targeted genes that are involved in cellular processes such as growth factor receptor signaling pathway, regulation of nuclear division, and glycolysis during the torpor stages, while dopamine and hormone receptor signaling along with snrna processing are among the biological processes targeted during IA. Using the Reactome pathway analysis, we further examined the specific signaling pathways that were targeted by the up and downregulated mirnas during torpor in each tissue (Supplemental Table S4). Interestingly, we identified a number of stress-responsive pathways were target by mirnas that were significantly downregulated during ET and LT, suggesting that these pathways could be activated through a relief of mirnamediated suppression. These include the p53 and DNA damage response pathways in the heart and DNA damage reversal and SMAD transcriptional response in skeletal muscle. Interestingly, bile acid and bile salt metabolism-related pathways were identified by the downregulated mirnas in the liver; although a previous study showed that bile acids such as cholesterol and fatty acid concentrations were largely unchanged in ground squirrel during winter hibernation, the levels of bilirubin that

6 microrna REGULATIONS IN THE TORPOR-AROUSAL CYCLE 393 Fig. 5. Gene ontology (GO) enrichment analysis of genes targeted by differentially regulated mirnas. The top 10 GO enriched biological processes targeted by mirnas as determined by analysis of putative target genes of upregulated (A) or downregulated mirnas (B) for each tissue during the torpor-arousal cycle. No enrichments were found for the downregulated mirnas in the liver. can function as ROS scavenger were increased significantly by about fivefold (2). DISCUSSION The emergence of mirnas roles as major regulators of cellular functions in the recent years has led to novel insights into understanding the mechanisms of many biological processes. Most evident in pathological states such as cancer, proper regulation of mirnas have been shown to be critical in maintaining a normal cellular homeostasis, with dysregulated mirna expressions often found in many oncogenic proliferatory states (25). We and others have previously shown that mirnas are differentially regulated during exposures to environmental stress, and this is a conserved response observed for different types of hypometabolic stress that includes hibernation, freezing, and aestivation (4, 24, 42, 47). In the ground squirrel, we have recently reported a potential role for mirnas in regulating adipose tissue metabolism, with downregulation of select mirnas linked to the regulation of adipogenesis and MAPK signaling during torpor (43). In the present study, we expanded our mirna expression profiling to three different tissue types across four stages of the hibernation cycle, measuring the expression of 117 conserved human mature mirnas to determine their regulatory profiles during ground squirrel torpor. Our data show that the majority of the differentially expressed mirnas were controlled in a tissue/torpor-stage specific manner, possibly reflecting the different essential cellular processes regulated between the tissues at various torpor stages. Using miranda microrna prediction analysis, we generated a list of genes that are putatively targeted by the mirnas that were found to be differentially expressed during torpor and applied an enrichment analysis through the use of GO annotations and the Reactome pathway database. Although we were able to validate the correlated expression changes between mirna and a subset of its target genes using available transcriptomic data, it should be noted that not all the predicted genes were found to have a correlated expressional change. This is not uncommon since currently available prediction software has been known to produce false-positive rates of

7 394 microrna REGULATIONS IN THE TORPOR-AROUSAL CYCLE 20 30% (27); this is further complicated by the fact that a positive interaction between mirna and target gene does not always lead to a decrease in target mrna levels, especially in cases where mirna functions to inhibit target mrna from protein translation via repression, rather than degradation (36). Therefore, it should be noted that the correlations between mirna and target transcript levels are to be interpreted cautiously. Ideally, comparison of mirna levels to proteomic changes would likely provide a better validation of the accuracy of mirna target prediction modeling; however, we were unable to find significant correlations between our mirna target expressions using available proteomics data (20, 22, 34), but this was likely biased toward the small amount of proteome data currently available compared with the transcriptome data. Our analysis shows that the heart and muscle tissues shared a limit number of overlapping mirnas; however, their predicted genes targeted by upregulated mirnas were enriched to some similar cellular processes such as regulation of growth and cell size, microtubule cytoskeleton organization, and amino acid transport (Fig. 4). As mirnas generally exert negative regulation of its target expression via translational silencing, it can be suggested that these cellular processes are likely downregulated during torpor. These findings are not surprising as progrowth cellular processes are generally suppressed during hibernation given the extreme hypometabolic state established during torpor (10, 38). Compared with the muscle and heart tissues, liver showed a much different profile in mirna expression patterns (Fig. 2). The liver plays an important role in coordinating a wide range of cellular processes that influence whole body metabolism; these include detoxification reactions, carbohydrate storage, and lipid transport. Like most tissues during torpor, overall functions of the liver are also reduced, and this is supported by the targeting of progrowth processes by overexpressed mirnas in the liver during torpor, which included endosome transport, growth factor receptor signaling, regulation of glycolysis, and regulation of mitosis and nuclear division (Fig. 5). Although the target enrichment provides a broad overview on many of the cellular processes potentially regulated by the mirnas during torpor, we have also identified several wellcharacterized mirnas that were highly expressed in a tissuespecific manner. In our study, we observed an approximately fivefold increase in mir-208b expression in the heart during LT, followed by a further upregulation during interbout arousal by 13-fold. It is known that the cardiac function of hibernating ground squirrels undergoes multiple challenges that include drastic reduction in heart rate from 300 to 10 beats/min. To maintain adequate blood circulation, the heart compensates for a reduced pumping rate with 30% ventricular hypertrophy in addition to elevated expression of cardiac myosin-heavy chain (MHC) to generate a potential increase in contractile power (31). The increase in myocardial activities and strain is also observed during the arousal stages, with cardiac output recorded at nearly 40% of euthermic states with only a 26% active heart rate (31). Previous studies have shown that mir- 208a and mir-208b are two of the most important heartenriched mirnas (9, 32). In the heart, expression of mir-208b is directly linked to the regulation of cardiac arrhythmias, muscle remodeling, and hypertrophic signaling pathways (32). In mice, mir-208b is encoded within the intron of mhc7 gene, with transgenic overexpression of mir-208 leading to the induction of a hypertrophic cardiac growth. Deletion of mir- 208b has been shown to reduce the transcript and protein levels of -MHC, while elevated expression of mir-208 leads to an increase in -MHC levels and a reduced expression of myostatin 2 gene, which serves as a negative regulator of myogenesis (9). Our observation of a highly upregulated mir-208b would suggest that it likely functions in a similar role to positively regulate cardiac gene expression in the ground squirrels and could be important in regulating the development of a hypertrophic heart to maintain cardiac functions during the torpor and arousal stages of hibernation (31). Meanwhile in the skeletal muscle, the same mir-208b showed an opposing expression, with mir-208b levels downregulated by 3.7-fold during LT. This was consistent with previous reports of mass reduction and atrophy ( 30%) of the gastrocnemius muscle in the ground squirrels that is combined with an undetectable levels of type II -MHC during hibernation compared with active squirrels (35). The differential regulation of mir-208b between both the heart and the muscle tissues during torpor is likely linked to their relative physiological changes during torpor, indicating the importance of tissue-specific expressions of mirnas during hibernation. In addition to mir-208b, we also identified several mirnas that are differently regulated between the heart and skeletal muscle during ET or LT; these include the upregulation of mir-130b, mir-145a, mir-148, mir-184, and mir-449a in heart muscle and mir-127, mir- 144, mir-152, mir-193a, and mir-425 in skeletal muscle. MiR-145a is an important mirna that is highly expressed in the heart and has been demonstrated to be involved in vascular smooth muscle cell differentiation, in addition to regulation of actin remodeling, contractility, with decreased expression found in patients with cardiovascular diseases (33). The expression of mir-145a in the heart is also linked to the regulation of F-actin dynamics and is involved in the control of cardiac chamber functions (14). Interestingly, mir-145a was found to be upregulated at all torpor stages only in the heart compared with the EC, suggesting a potential important and continuous role of mir-145a in regulating heart-specific functions throughout the torpor cycle. The skeletal muscle showed highly upregulated expression of mir-193a by 9.4-fold during ET. A recent study indicates that mir-193a downregulates multiple key activators of the insulin signaling cascade, with confirmed targets including phosphoinositide-3-kinase regulatory subunit (PI3KR), mammalian target of rapamycin (mtor), and ribosomal S6 kinase (S6K) (46). We have previously shown that the insulin signaling network was suppressed to promote translational incompetence in the skeletal muscle during torpor, and this was achieved through posttranslational modifications that includes a phosphorylation cascade (44). However, the upregulation of mir-193a during torpor in the skeletal muscle could function as an additional mechanism to promote protein synthesis inhibition by directly targeting the transcript levels of key regulators in the insulin signaling cascade. Another interesting observation is the 5.7-fold increase in mir-181a in the liver during ET. MiRNA-181a when overexpressed has been shown to cause insulin resistance in the hepatic cells, and diabetic patients show elevated levels of mir-181a in the liver and serum; similar findings were also shown in mice, with mir- 181a found to be upregulated in the Type 2 diabetes db/db mice model (48). Hibernating ground squirrels have been

8 shown to exhibit reversible insulin resistance during the torporarousal cycle; levels of serum insulin are highly elevated during the early months of the hibernation season, and this is combined with upregulation of insulin mrna levels throughout the hibernation months from December to March (8). The upregulation of mir-181a was found to be specific in the liver during ET, with no changes observed in the heart or the muscle. Previous studies have suggested that the onset of hyperinsulinemia and insulin resistance are at its peak when the animals initially enter hibernation (16); it is possible that the overexpression of mir-181a is part of a regulatory response that aims to stimulate insulin resistance during ET and could be crucial in maintaining the storage of excess fat that is utilized as metabolic fuel throughout hibernation. Conclusions As recent studies have shown that global transcription and translation are both suppressed during torpor (17, 30), an emphasis has been placed on understanding the role of posttranscriptional mechanisms as regulatory components of hibernation. Grabek et al. (18) have shown that select groups of transcripts can be protected and stabilized during torpor via differential polyadenylations at the 3=-UTR end of the mrna transcript. This mechanism of dynamic polyadenylation can alter the length of poly (A) tail of select mrna transcripts, with the varying sizes of poly (A) tails directly influencing the rate of mrna translation (18). These findings are complemented by our recent reports of mrna processing factors, which include poly-a binding protein 1 in the formation of subcellular granule-like structures during hibernation (40). The evidence of alternative polyadenylation is intriguing with respect to mirna regulations, as changes to the 3=-UTR region of a transcript through alternative polyadenylation can directly influence its interaction with a target mirna, potentially modulating the mrna susceptibility to mirna-mediated repression (1). Meanwhile, the formation of subcellular granules during torpor could serve as a potential site for mirnamediated mrna processing (40). Our data here show that mirna expressions are dynamically regulated during torpor and are uniquely expressed in a tissue-specific manner. Putative target analyses suggest that upregulated mirnas, although varied between tissues, tend to target similar progrowth processes during torpor. In addition, we identified several highly differentially expressed mirnas during the torpor-arousal cycle, which includes mir-208b and mir-181a, that may regulate key physiological changes in hibernators such as cardiac hypertrophy and insulin resistance. Our expression profiling has provided new insights into the regulation patterns of mirnas during the torpor-arousal cycle and highlights the importance of mirna regulation during hibernation. The combination of alternative polyadenylation and mirna regulations could help to explain the disconnect that is observed between mrna and protein levels during hibernation (37) and, more broadly, emphasizes the important roles posttranscriptional regulatory mechanisms play during hibernation. Future functional studies will be needed to fully characterize and validate the regulatory nature of these mir- NAs and determine their relative requirement for a hibernation phenotype. microrna REGULATIONS IN THE TORPOR-AROUSAL CYCLE ACKNOWLEDGMENTS Thanks go to Dr. J. M. Hallenbeck for providing the ground squirrel tissues and J. M. Storey for editorial review of the manuscript. Current address for C.-W. Wu: Dept. of Biology, Genetics Inst., Univ. of Florida, Gainesville, FL GRANTS This research was funded by a Discovery grant from the Natural Sciences and Engineering Research Council (NSERC) of Canada to K. B. Storey (#6793) and by NSERC doctoral scholarships to C. W. Wu, K. K. Biggar, and B. E. Luu. DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the author(s). AUTHOR CONTRIBUTIONS C.-W.W., K.K.B., and K.B.S. conception and design of research; C.-W.W. and B.E.L. performed experiments; C.-W.W., K.K.B., B.E.L., and K.E.S. analyzed data; C.-W.W., K.K.B., and K.E.S. interpreted results of experiments; C.-W.W. and B.E.L. prepared figures; C.-W.W. drafted manuscript; C.-W.W. and K.K.B. edited and revised manuscript; C.-W.W., K.K.B., B.E.L., and K.B.S. approved final version of manuscript. REFERENCES An J, Zhu X, Wang H, Jin X. A dynamic interplay between alternative polyadenylation and microrna regulation: implications for cancer (Review). Int J Oncol 43: , Baker JA, van Breukelen F. Bile constituents in hibernating goldenmantled ground squirrels (Spermophilus lateralis). Comp Hepatol 8: 2, Bartel DP. MicroRNAs: target recognition and regulatory functions. Cell 136: , Biggar KK, Dubuc A, Storey K. MicroRNA regulation below zero: differential expression of mirna-21 and mirna-16 during freezing in wood frogs. Cryobiology 59: , Biggar KK, Kornfeld SF, Maistrovski Y, Storey KB. MicroRNA regulation in extreme environments: differential expression of micrornas in the intertidal snail Littorina littorea during extended periods of freezing and anoxia. Genomics Proteomics Bioinformatics 10: , Biggar KK, Storey KB. Evidence for cell cycle suppression and microrna regulation of cyclin D1 during anoxia exposure in turtles. Cell Cycle 11: , Biggar KK, Wu CW, Storey KB. High-throughput amplification of mature micrornas in uncharacterized animal models using polyadenylated RNA and stem-loop reverse transcription polymerase chain reaction. Anal Biochem 462: 32 34, Buck MJ, Squire TL, Andrews MT. Coordinate expression of the PDK4 gene: a means of regulating fuel selection in a hibernating mammal. Physiol Genomics 8: 5 13, Callis TE, Pandya K, Seok HY, Tang RH, Tatsuguchi M, Huang ZP, Chen JF, Deng Z, Gunn B, Shumate J, Willis MS, Selzman CH, Wang DZ. MicroRNA-208a is a regulator of cardiac hypertrophy and conduction in mice. J Clin Invest 119: , Carey HV, Andrews MT, Martin SL. Mammalian hibernation: cellular and molecular responses to depressed metabolism and low temperature. Physiol Rev 83: , Consortium GO. Gene Ontology Consortium: going forward. Nucleic Acids Res 43: D1049 D1056, Croft D, O Kelly G, Wu G, Haw R, Gillespie M, Matthews L, Caudy M, Garapati P, Gopinath G, Jassal B, Jupe S, Kalatskaya I, Mahajan S, May B, Ndegwa N, Schmidt E, Shamovsky V, Yung C, Birney E, Hermjakob H, D Eustachio P, Stein L. Reactome: a database of reactions, pathways and biological processes. Nucleic Acids Res 39: D691 D697, Dark J. Annual lipid cycles in hibernators: integration of physiology and behavior. Annu Rev Nutr 25: , Deacon DC, Nevis KR, Cashman TJ, Zhou Y, Zhao L, Washko D, Guner-Ataman B, Burns CG, Burns CE. The mir-143-adducin3 pathway is essential for cardiac chamber morphogenesis. Development 137: , 2010.

9 396 microrna REGULATIONS IN THE TORPOR-AROUSAL CYCLE 15. Enright AJ, John B, Gaul U, Tuschl T, Sander C, Marks DS. MicroRNA targets in Drosophila. Genome Biol 5: R1, Florant GL, Lawrence AK, Williams K, Bauman WA. Seasonal changes in pancreatic B-cell function in euthermic yellow-bellied marmots. Am J Physiol Regul Integr Comp Physiol 249: R159 R165, Frerichs KU, Smith CB, Brenner M, DeGracia DJ, Krause GS, Marrone L, Dever TE, Hallenbeck JM. Suppression of protein synthesis in brain during hibernation involves inhibition of protein initiation and elongation. Proc Natl Acad Sci USA 95: , Grabek KR, Diniz Behn C, Barsh GS, Hesselberth JR, Martin SL. Enhanced stability and polyadenylation of select mrnas support rapid thermogenesis in the brown fat of a hibernator. Elife: 4, Hampton M, Melvin RG, Kendall AH, Kirkpatrick BR, Peterson N, Andrews MT. Deep sequencing the transcriptome reveals seasonal adaptive mechanisms in a hibernating mammal. PLoS One 6: e27021, Hindle AG, Karimpour-Fard A, Epperson LE, Hunter LE, Martin SL. Skeletal muscle proteomics: carbohydrate metabolism oscillates with seasonal and torpor-arousal physiology of hibernation. Am J Physiol Regul Integr Comp Physiol 301: R1440 R1452, Lewis BP, Burge CB, Bartel DP. Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microrna targets. Cell 120: 15 20, Li H, Liu T, Chen W, Jain MR, Vatner DE, Vatner SF, Kudej RK, Yan L. Proteomic mechanisms of cardioprotection during mammalian hibernation in woodchucks, Marmota monax. J Proteome Res 12: , Liu J, Valencia-Sanchez MA, Hannon GJ, Parker R. MicroRNAdependent localization of targeted mrnas to mammalian P-bodies. Nat Cell Biol 7: , Liu Y, Hu W, Wang H, Lu M, Shao C, Menzel C, Yan Z, Li Y, Zhao S, Khaitovich P, Liu M, Chen W, Barnes BM, Yan J. Genomic analysis of mirnas in an extreme mammalian hibernator, the Arctic ground squirrel. Physiol Genomics 42A: 39 51, Lu J, Getz G, Miska EA, Alvarez-Saavedra E, Lamb J, Peck D, Sweet-Cordero A, Ebert BL, Mak RH, Ferrando AA, Downing JR, Jacks T, Horvitz HR, Golub TR. MicroRNA expression profiles classify human cancers. Nature 435: , MacDonald JA, Storey KB. Regulation of ground squirrel Na K - ATPase activity by reversible phosphorylation during hibernation. Biochem Biophys Res Commun 254: , Mazière P, Enright AJ. Prediction of microrna targets. Drug Discov Today 12: , McMullen DC, Hallenbeck JM. Regulation of Akt during torpor in the hibernating ground squirrel, Ictidomys tridecemlineatus. J Comp Physiol B 180: , Milacic M, Haw R, Rothfels K, Wu G, Croft D, Hermjakob H, D Eustachio P, Stein L. Annotating cancer variants and anti-cancer therapeutics in reactome. Cancers (Basel) 4: , Morin P Jr, Storey KB. Evidence for a reduced transcriptional state during hibernation in ground squirrels. Cryobiology 53: , Nelson OL, Rourke BC. Increase in cardiac myosin heavy-chain (MyHC) alpha protein isoform in hibernating ground squirrels, with echocardiographic visualization of ventricular wall hypertrophy and prolonged contraction. J Exp Biol 216: , Oliveira-Carvalho V, Carvalho VO, Bocchi EA. The emerging role of mir-208a in the heart. DNA Cell Biol 32: 8 12, Rangrez AY, Massy ZA, Metzinger-Le Meuth V, Metzinger L. mir- 143 and mir-145: molecular keys to switch the phenotype of vascular smooth muscle cells. Circ Cardiovasc Genet 4: , Rose JC, Epperson LE, Carey HV, Martin SL. Seasonal liver protein differences in a hibernator revealed by quantitative proteomics using whole animal isotopic labeling. Comp Biochem Physiol Part D Genomics Proteomics 6: , Rourke BC, Yokoyama Y, Milsom WK, Caiozzo VJ. Myosin isoform expression and MAFbx mrna levels in hibernating golden-mantled ground squirrels (Spermophilus lateralis). Physiol Biochem Zool 77: , Ruike Y, Ichimura A, Tsuchiya S, Shimizu K, Kunimoto R, Okuno Y, Tsujimoto G. Global correlation analysis for micro-rna and mrna expression profiles in human cell lines. J Hum Genet 53: , Shao C, Liu Y, Ruan H, Li Y, Wang H, Kohl F, Goropashnaya AV, Fedorov VB, Zeng R, Barnes BM, Yan J. Shotgun proteomics analysis of hibernating arctic ground squirrels. Mol Cell Proteomics 9: , Storey KB. Out cold: biochemical regulation of mammalian hibernation - a mini-review. Gerontology 56: , Storey KB. Regulation of liver metabolism by enzyme phosphorylation during mammalian hibernation. J Biol Chem 262: , Tessier SN, Audas TE, Wu CW, Lee S, Storey KB. The involvement of mrna processing factors TIA-1, TIAR, and PABP-1 during mammalian hibernation. Cell Stress Chaperones 19: , Thomas PD, Campbell MJ, Kejariwal A, Mi H, Karlak B, Daverman R, Diemer K, Muruganujan A, Narechania A. PANTHER: a library of protein families and subfamilies indexed by function. Genome Res 13: , Wu CW, Biggar KK, Storey KB. Dehydration mediated microrna response in the African clawed frog Xenopus laevis. Gene 529: , Wu CW, Biggar KK, Storey KB. Expression profiling and structural characterization of micrornas in adipose tissues of hibernating ground squirrels. Genomics Proteomics Bioinformatics 12: , Wu CW, Storey KB. Regulation of the mtor signaling network in hibernating thirteen-lined ground squirrels. J Exp Biol 215: , Xu Y, Shao C, Fedorov VB, Goropashnaya AV, Barnes BM, Yan J. Molecular signatures of mammalian hibernation: comparisons with alternative phenotypes. BMC Genomics 14: 567, Yu T, Li J, Yan M, Liu L, Lin H, Zhao F, Sun L, Zhang Y, Cui Y, Zhang F, He X, Yao M. MicroRNA-193a-3p and -5p suppress the metastasis of human non-small-cell lung cancer by downregulating the ERBB4/PIK3R3/mTOR/S6K2 signaling pathway. Oncogene 34: , Yuan L, Geiser F, Lin B, Sun H, Chen J, Zhang S. Down but not out: the role of micrornas in hibernating bats. PLoS One 10: e , Zhou B, Li C, Qi W, Zhang Y, Zhang F, Wu JX, Hu YN, Wu DM, Liu Y, Yan TT, Jing Q, Liu MF, Zhai QW. Downregulation of mir-181a upregulates sirtuin-1 (SIRT1) and improves hepatic insulin sensitivity. Diabetologia 55: , 2012.