MicroRNA-434-3p regulates age-related apoptosis through eIF5A1 in the skeletal muscle

Increased activation of catabolic pathways, including apoptosis causes sarcopenia. However, the precise molecular mechanism that initiates apoptosis during aging is not well understood. Here, we report that aging alters miRNA expression profile in mouse skeletal muscle as evidenced by miRNA microarray and real-time PCR. We identified miR-434-3p as a highly downregulated miRNA in the skeletal muscle of aging mice. Myocytes transfected with miR-434-3p mimic prevents apoptosis induced by various apoptotic stimuli, and co-transfection of miR-434-3p antagomir abolishes the inhibitory role of miR-434-3p. We found that miR-434-3p inhibits apoptosis by targeting the eukaryotic translation initiation factor 5A1 (eIF5A1). Overexpression of miR-434-3p in myocytes reduces the loss of mitochondrial transmembrane potential, and activation of caspases-3, −8 and −9 by suppressing eIF5A1 in response to various apoptotic stimuli whereas inhibition of miR-434-3p reversed this scenario. Skeletal muscles from aging mice exhibit low levels of miR-434-3p and high levels of eIF5A1, suggesting a possible role for miR-434-3p in the initiation of apoptosis in aging muscle. Overall, our data identified for the first time that miR-434-3p is an anti-apoptotic miRNA that may be therapeutically useful for treating muscle atrophy in various pathophysiological conditions, including sarcopenia.


INTRODUCTION
Sarcopenia is an advanced age-related loss of skeletal muscle mass as well as loss of its function, which limits the independence living and quality of life, thereby it underlies morbidity and mortality in elderly individuals [1,2]. Sarcopenia also reduces the amount of metabolically active tissue; thus, it increases the risk for metabolic diseases [3]. Initiation of sarcopenia involves complex processes that are controlled by both extrinsic and intrinsic factors [3,4], many of which converge on a decline in the ability of muscle stem cells (satellite cells) to replace unhealthy muscle fibers during aging [5][6][7]. Although the mechanism that causes sarcopenia are largely unknown, a progressive decline in anabolism mostly due to a reduction in protein synthesis with an increase of catabolism mainly due to an enhanced activation of pathways like apoptosis, initiate sarcopenia [2].
Apoptosis is essential for organ development, tissue homeostasis, and the elimination of defective cells in multi-cellular organisms; however, accelerated apoptosis in skeletal muscle is a potential mechanism of sarcopenia [8][9][10][11][12][13]. Different apoptotic stimuli such as oxidative stress, calcium, and TNF-α, may be seen as initiators of the apoptotic signaling in skeletal muscle during aging [8,9,14].
MicroRNAs (miRNAs) are a class of small noncoding RNAs that regulate gene expression at the posttranscriptional level. These noncoding RNAs have recently emerged as crucial regulators of aging process [15][16][17]. Emerging evidence has shown alteration in AGING miRNA expression profile in the skeletal muscles of both human and mouse during aging [18][19][20][21][22][23][24][25]. For example, elevated let-7 family members in skeletal muscle contribute to reduced cellular proliferation and regenerative capacity in aged human [22]. Aging alters the expression of 57 miRNAs in mouse quadriceps muscles and few of them are associated with reduced cell proliferation and favored the terminal differentiation of myogenic precursor [20]. Furthermore, resistance exercise, caloric restriction, or nutrientrelated hormones such as the adipokine leptin reverse the expression of age-regulated miRNAs [18,21,26]. Interestingly, it has been shown that miR-210 can mediate hypoxia-induced apoptosis of neuroblastoma cells by targeting expression of the anti-apoptotic protein Bcl-2 gene [27]. The above studies suggest the possibility that dysregulation of miRNAs in skeletal muscle during aging may induce sarcopenia by activating catabolic pathways, including apoptosis. However, no study to date has demonstrated a role for miRNA in skeletal muscle apoptosis, especially during aging. In the current study, we show that aging dysregulates many miRNAs in skeletal muscle, including the highly downregulated miRNA miR-434-3p that inhibits apoptosis by targeting eIF5A1 that promote apoptosis by the intrinsic mitochondrial pathway.
Overall, these data provide experimental evidence that eIF5A1 is a direct target gene of miR-434-3p. Previous studies have shown that eIF5A1 promotes apoptosis in a variety of cells [29][30][31][32]. Our prediction pathway analysis for miR-434-3p identified apoptosis as one of the skeletal muscle-specific pathways. Because eIF5A1 is a direct downstream target of miR-434-3p, we sought to determine if modulation of miR-434-3p would control apoptosis through eIF5A1 in myotubes. TPEN (a potent inducer of apoptosis) has been shown to induce apoptosis in myocytes [33]. Myotubes were transfected with miR-434-3p mimics or NS-miR for 36 h before treatment with TPEN for 24 h. Our data show that myotubes transfected with NS-miRNA before TPEN exposure had no protection against TPENinduced apoptosis (Fig. 4A). In contrast, myotubes transfected with miR-434-3p mimics before TPEN treatment had an about 70% reduction in the percentage of apoptotic cells compared with myotubes transfected with NS-miRNA mimics. The transfection of miR-434-3p antagomir along with miR-434-3p mimics restored the TPEN-induced myotubes death (Fig. 4A). Western blot analysis indicated that TPEN-treated myotubes displayed an increased levels of eIF5A1 protein whereas myotubes transfected with NS-miR or myotubes that were not transfected had no effect on myotubes survival. On the other hand, transfection of myotubes with miR-434-3p mimics strongly reduced eIF5A1 protein levels and the cotransfection of myotubes with miR-434-3p antagomir reinstated eIF5A1 protein levels (Fig. 4B). To further confirm the role of miR-434-3p in apoptosis, we used staurosporine (Stsp), a well-known transcription inhibitors, has been shown to induce apoptosis in several cell types including myocytes [33,34]. While myotubes treated with Stsp had a higher percentage of death, transfection of those myotubes with miR-434-3p mimic increased the proportion of myotubes survival and co-transfection of myotubes with miR-434-3p antagomir reversed the effect of miR-434-mimic (Fig. 4C). Furthermore, loss of myotubes survival due to Stsp treatment correlated with a significant increase in the levels of eIF5A1 protein.
Such altered expression of elF5A1 was reduced when those myotubes were transfected with miR-434-3p antagomir (Fig. 4D). In addition, knockdown of eIF5A1 AGING by siRNA in myotubes significantly reduced the TPEN or Stsp -induced cell death (Fig. 4E), suggesting that eIF5A1 is directly responsible for the TPEN or Stspinduced apoptotic effect. Furthermore, we found that GA muscle from aging mice displayed lower expression levels of miR-434-3p when compared to that in the GA muscle of young mice (Fig. 4F). In contrast, the levels of eIF5A1 mRNA and protein were higher in the GA muscle of aging mice when compared to that in the GA muscle of young mice (Fig. 4F). These data confirm that miR-434-3p protects myocyte from apoptosis induced by different apoptotic stimuli via eIF5A1 and a negative correlation between the expression of miR-434-3p and eIF5A1 in aging muscle.

Overexpression of miR-434-3p reduces activation of caspases 3, 8, and 9 in myotubes
To determine whether miR-434-3p-regulated apoptosis involves inhibition of activation of caspases, the effects of their over-expression on the activation of executioner caspase-3 as well as initiator caspases -8 and -9 were examined in myotubes over a period of 24 h. All three caspases were activated in response to TPEN or Stps treatment; however, caspase 3 activity was dramatically higher when compared to caspases 8 and 9, especially in Stps treated myotubes ( Fig. 5A and B). Myotubes transfected with miR-434-3p mimic significantly suppressed all three caspases activities, and the cotransfection of miR-434-antagomir restored the TPEN or Stps-induced caspases activities whereas AGING myotubes transfected with NS-miR had no effect on all three caspases activities in response to TPEN or Stps treatment ( Fig. 6A and B). One of the earliest events in the progression of apoptosis is the dissipation of the mito-chondrial membrane potential (∆ψm). Our data in Fig 5C and D showed suppression of caspases 3, 8 and 9 activities by miR-434-3p, suggesting the possibility that miR-434-3p may maintain Δψm. JC-1 is a different cationic dye has been used to measure the collapse of the electrochemical gradient across the mitochondrial membrane [34,35]. In normal untreated myotubes, JC-1 is present as a red fluorescent aggregate in mitochondria, and in the green fluorescent monomeric form in the cytosol in myotubes. Upon the treatment with TPEN or Stsp, there is a dissipation of the ∆ψm indicated by the increase in staining in the green filter AGING (530 nm) and a decrease in staining in the red filter (590 nm) in myotubes. Quantification of the fluorescent intensities indicated that myotubes treated with TPEN or Stsp significantly decreased the ∆ψm index while myotubes transfected with miR-434-3p mimic reversed the scenario; in contrast, the cotransfection of antagomir restored the effect of TPEN and Stsp in ∆ψm (Fig. 5C). These data suggest that the diminished depolarization of mitochondrial membrane potential of myotubes in response to TPEN and Stsp may be due to suppression of apoptosis by miR-434-3p through eIF5A1.

DISCUSSION
Sarcopenia is a progressive decline in skeletal muscle mass, strength, and quality during aging. Although the molecular mechanisms involved in sarcopenia development are not entirely understood, it is welldocumented that transcriptomic, proteomic and epigenetic changes during the progression of aging are the main causative factors for sarcopenia. These epigenetic changes could increase oxidative stress and vice versa and lead to muscle loss and function due to activation of multiple catabolic pathways, including apoptosis. The main objective of this study was to determine the effect of aging on miRNA expression profile and identify how dysregulation of a specific miRNA in aging muscle could cause the activation of the catabolic pathway(s). Using microarray analysis, we found that aging dysregulated miRNA expression profile in skeletal muscle and most of the dysregulated miRNAs are suppressed in aging muscle. Using exten-sive bioinformatics analysis, we identified the eukaryotic translation initiation factor 5A1 (eIF5A1) as one of the potential target genes of the most highly downregulated miRNA miR-434-3p. Overexpression or knockdown of miR-434-3p in myotubes validated eIF5A1 as a target mRNA of miR-434-3p that negatively regulated apoptosis through eIFA1. Furthermore, dual-luciferase assay validation method further confirmed miR-434-3p binding site within the 3'UTR mRNA encoding eIF5A1. Interestingly, we found that skeletal muscle from aging mice downregulated of miR-434-3p expression that was negatively correlated with the levels of eIF5A1, suggesting that dysregulation of miR-434-3p in aging muscle may be responsible, in part, for the pathogenicity of sarcopenia.
In the present study, we found that aging dysregulated many miRNAs, including two upregulated and 13 downregulated (≥1.5 fold) miRNAs in skeletal muscle, suggesting diminished miRNA expressions profile in aging muscle, and this is in agreement with previous findings [19,23,25,36]. The differentially expressed miRNAs in earlier studies appeared consistently in mouse skeletal muscle from our study; for example, the expression pattern of miR-146a-5p, miR-146b-5p, miR-434-3p, miR-127-3p, and miR-148a-3p are similar to previous studies. In our study, we found that miR-434-3p was the highly downregulated mRNA (4.08 fold) in the aging muscle; this is also in agreement with the previous study in which aging mice downregulates this miRNA at 5.0 fold in skeletal muscle [20]. This suggesting that miR-434-3p may have a crucial role in the etiology of aging. Using extensive bioinformatics analysis, we identified eIF5A1 as one of the potential target genes of miR-434-3p. We confirmed that miR-434-3p indeed suppressed eIF5A1 mRNA expression by binding on the 3'UTR of eIF5A1, as evidenced by western blot analysis and luciferase assay. Moreover, we for the first time demonstrate that aging muscle upregulated eIF5A1 protein expression, suggesting a negative correlation in the expression levels between miR-434-3p and eIF5A1 in the aging muscle. In agreement with our findings, a previous study demonstrates that overexpression of miR-434-3p in vivo rat skeletal muscle post-transcriptionally suppresses eIF5A1 expression and both miR-434-3p and eIR5A1 are regulated in muscle in an opposite manner after spinal cord injury [37]. Although miR-434-3p was the most highly downregulated miRNA in the skeletal muscle of aged mice, the signal intensity was relatively small (1767) when compare to other miRNAs such as miR-1 (59283) and miR-126-3p (42474) as evidenced by microarray that is not a quantitative assay and signal intensity more than 500 is biologically relevant. Interestingly, our quantitative PCR assay in Figures 2A  and 4F show that miR-434-3p was nearly 4-fold significantly downregulated in aging muscle when compared to that in the muscle of young mice, suggesting that aging significantly affected the expression of miR-434-3p. This significance was confirmed by solution hybridization detection (miR-434-3p) and western blot (eIF5A1). In addition, the miR-434-3p is absence in human according to miRBase 21 that limits the translational significance. Because many miRNAs can target a single mRNA, other human miRNA(s) that target eIF5A1 may have a similar role of miR-434-3p and that specific miRNA(s) may reinstate the translational significance of miR-434-3p. Overall our findings suggest that the dysregulation of the miR-434-3p/eIF5A1 pathway in aging muscle may play a role in the pathogenicity of sarcopenia.
The small protein eIF5A1 promotes the initiation and translation elongation phases of protein synthesis by transient association with 80S ribosome complex [38][39][40][41]. However, numerous other functions have also been identified for eIF5A, including apoptosis in a variety of cells [31,32,42]. In this study, up-regulation of eIF5A1 protein and downregulation of miR-434-3p in aging skeletal muscle suggest that miR-434-3p/eIF5A1 pathway may induce apoptosis in skeletal muscle during aging and that may be one of the epigenetic causative mechanisms for the induction of sarcopenia. Although the etiology of sarcopenia is complex and characterized by the contribution of multiple factors [43], there is growing evidence for a prominent role of accelerated apoptosis in sarcopenia [12,[43][44][45][46][47].
Previous studies identified a role for eIF5A1 in apoptosis; these studies demonstrate that siRNA-mediated suppression of eIF5A1 expression protects cells against apoptosis induced by TNF-α [30], Actinomycin D [31,32], sodium nitroprusside as well as the proteasome inhibitor MG-132 [32]. Conversely, over-expression of eIF5A1 has been shown to induce apoptosis [31,32,42], further supporting the apoptogenic nature of eIF5A1. This study sheds some light to determine how eIF5A1 engages in apoptosis in skeletal muscle during aging. We found that eIF5A1 facilitates activation of the intrinsic mitochondrial apoptotic pathway. We demonstrate that knockdown of miR-434-3p in myotubes induced activation of caspase-3, -8 and -9, and permeabilization of the outer mitochondrial membrane along with upregulation of eIF5A1, suggesting that loss of miR-434-3p promotes apoptosis by facilitating activation of the intrinsic mitochondrial pathway. We confirmed the above data by demonstrating that overexpression of miR-434-3p resulted in inhibition of caspase-3, -8 and -9 activations, and permeabilization of the outer mitochondrial membrane along with downregulation of eIF5A1. Oxidative damage to lipids, proteins, and DNA, especially in post-mitotic tissue like the skeletal muscle of an aged organism, may be severe and ultimately lead to apoptotic or necrotic cell death. We and others have previously shown an elevation of oxidative and apoptosis markers in skeletal muscle from aging mice [48][49][50][51][52]. Future studies should focus on determining the causal mechanism through which aging of skeletal muscles suppresses miR-434p.

Animals
The Institutional Animal Care and Use Committee from the Baylor College of Medicine approved all experimental procedures. We used a total of 44 mice to complete the experiments mentioned in this study. These include three-month-old young (n=20), and 26 months old aged (n=24) male C57BL/6J mice (The Jackson Laboratory, Bar Harbor, ME). All mice were kept in a temperature-controlled room on a 12-h light/dark cycle, with a temperature of 23°C, humidity of 40-60%, and food and water ad libitum.

Cell culture
For in vitro studies, we used primary myoblasts isolated from young mouse hind limb muscles and cultured them to induce myotubes as described previously [53].

miRNA microarray analysis
We isolated total RNA from the skeletal muscles of young and aged mice and used for microarray analysis as mentioned earlier [54,55]. The fold difference values of altered miRNAs were converted into log 2 scale.

Gene functional analysis
To identify the predicted biological pathways regulated by miRNAs (≥1.5-fold), we used the DIANA miRPath v2.0 Web-based computational tool with a threshold pvalue of 0.05 and a MicroT threshold value of 0.8 [56]. miRNAs and their predicted targets were identified using the miRWalk Web-based computational tool, which provides miRNA targets from at least eight established miRNA prediction programs [57].

Reverse transcription and quantitative PCR (RT-qPCR)
To validate differentially regulated miRNA data from microarray analysis, we performed miRNA RT-qPCR array as described earlier [54,55]. To normalize mRNA and miRNA levels in qPCR, we used glyceraldehyde-3phosphate dehydrogenase (GAPDH) and U6, respectively.

Solution hybridization detection analysis
To confirm microarray data, we measured the expression levels of mature miRNAs by a solution hybridization detection method with mirVana miRNA and Bright-Star BioDetect kits as described previously [54].

Western blot
Western blots were performed as documented earlier [53]. eIF5A1 and GAPDH antibodies were purchased from Cell Signaling, Danvers, MA.

Induction of apoptosis
Apoptosis was induced in myotubes by treated with TPEN or staurosporine (Stsp) (Santa Cruz Biotechnology, Inc. Dallas, TX). Briefly, cells were treated for seven hours with TPEN (10 μM -100 μM) or Stsp (0.5-2 μM) to induce apoptosis. We observed that about a 50-60% cell death occurred at 100 μM TPEN or 1 μM Stsp at eight hours (data not shown) and therefore we used this concentration for further experiments.

Cell viability assay
Cell survival was measured using the standard MTT cell viability assay protocol. Briefly, primary myoblasts were plated in 24-well plates in medium containing 10% FBS and allowed to adhere for 24 hours. Then, the differentiation of primary myoblasts was induced as described previously [53]. After treatment of myotubes, cell viability was quantified by MTT (obtained from Sigma-Aldrich) and expressed as a percentage as mentioned earlier [59]. All experiments were repeated at least three times, with each experimental condition repeated at least in triplicate per experiment.

Assessment of mitochondrial membrane potential (Δψm)
Because a decline in Δψm causes the escape of proapoptotic proteins that regulate both caspase-dependent and independent apoptosis from mitochondria into the cytosol, we measured Δψm as an early event in the initiation of apoptosis. After appropriate treatments of myotubes, Δψm was assessed using the cationic dye, JC-1 (5,5′, 6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimidazolylcarbocyanine iodide). JC1 stains the healthy myotubes with bright red due to accumulation of the dye within the mitochondria whereas the dye stains the apoptotic myotubes with green because of the collapse of the mitochondrial membrane potential resulting in the dye unable to accumulate within the mitochondria. After treatment, myotubes were rinsed with PBS, followed by incubation with JC-1 reagent 1:100 at 37°C for 30 min. After rinsed myotubes twice with PBS, images were obtained using TRITC (red, 590 nm) and GFP (green, 530 nm) filters on a fluorescent microscope (Carl Zeiss). Care was taken to obtain pictures with identical exposure times, and pictures were analyzed using Carl Zeiss software. The automatic measurement program with same user defined parameters for densitometric and geometric variables was used to determine fluorescent intensity for both AGING filters. The ratio of the sum of intensities of red over green fluorescence was identified and expressed as the Δψm index.