Abstract
Age-associated loss of muscle mass and function is a major cause of morbidity and mortality in the elderly adults. Muscular atrophy can also be induced by disuse associated with long-term bed rest or disease. Although miRNAs regulate muscle growth, regeneration, and aging, their potential role in acute muscle atrophy is poorly understood. Furthermore, alterations in circulating miRNA levels have been shown to occur during aging but their potential as noninvasive biomarkers for muscle atrophy remains largely unexplored. Here, we report comprehensive miRNA expression profiles by miRNA-seq analysis in tibialis anterior muscle and serum of a disuse-induced atrophy mouse model, mimicking the acute atrophy following long-term bed rest, as compared to those of young and old mice. Comparative analysis and validation studies have revealed that miR-455-3p was significantly decreased in muscle of both induced-atrophy model and old mice, whereas miR-434-3p was decreased in both serum and muscle of old mice, as compared to young mice. Furthermore, upregulation of miR-455-3p in fully differentiated C2C12 myoblasts induced a hypertrophic phenotype. These results suggest that deregulation of miR-455-3p may play a functional role in muscle atrophy and miR-434-3p could be a candidate serum biomarker of muscle aging.
Keywords: Sarcopenia, Muscle aging, MicroRNA, Disuse muscle atrophy, Serum
It is well known that muscle mass and its function gradually decrease with age, a process known as sarcopenia (1). Several studies revealed that various factors, such as motor neuron death, nutritional imbalance, and inflammation, contribute to sarcopenia (2–4). Because loss of muscle mass can cause not only physical disability but also contribute to other age-related diseases, such as cardiovascular diseases and diabetes (5–7), understanding the underlying mechanism of the muscle aging process is important for promoting healthy aging in the elderly adults.
The elderly adults are at increased risk of age-related diseases or injuries and are more likely to be in long-term bed rest. Even though bed rest is essential for recovering from disease or injury, prolonged bed rest can exacerbate loss of muscle mass in elderly individuals already suffering from sarcopenia (8,9). The rate of muscle wasting is greater in bed-ridden older adults, with their muscle strength and power being severely decreased following bed rest (10). Because the elderly adults lose muscle tissue more rapidly following physical inactivity than the young, we hypothesized that common intrinsic factors might be involved in muscle wasting in both aged muscle and disuse-induced muscle atrophy. There are several experimental models to induce skeletal muscle atrophy, including denervation, tail suspension, or hind limb immobilization (11–13). Denervation causes atrophy of muscle tissue, which results from destruction of the motor nerve linked to the muscle. Tail suspension induces atrophy through musculoskeletal unloading. Hind limb immobilization, using a surgical staple, causes rapid loss of muscle mass in the immobilized limb compared to the contralateral hind limb muscle. Among these models, immobilization is the most suitable model for disuse-induced skeletal muscle atrophy, as it mimics the atrophy caused by prolonged bed lest.
With the advent of omics technologies, many investigators have examined the expression profiles of mRNA, miRNA, and proteins in skeletal muscle with age in order to identify intrinsic factors contributing to sarcopenia (14–16). Several studies have shown that miRNAs are involved in the homeostasis of muscle stem/progenitor cells and in muscle diseases such as skeletal muscle hypertrophy (17–20). In addition, miRNAs have been shown to be differentially expressed in skeletal muscle with age (14,15,21). Recently, we also reported that the miRNA-mRNA regulatory network is deregulated with age in skeletal muscle (22) and that age-associated miRNAs regulate the myogenic capability of muscle stem/progenitor cells with age (23). MiRNAs are also present and remarkably stable in circulating plasma (24). Alterations in their extracellular levels have been shown to occur during aging and in response to cellular damage and tissue injury in several disease states (25), offering their potential as noninvasive biomarkers.
Although there is increasing evidence that miRNAs regulate muscle growth, regeneration, and aging, their potential role in acute muscle atrophy remains largely unexplored. In the present study, we performed a comparative analysis of miRNA expression profiles by miRNA-seq in tibialis anterior (TA) muscle and serum isolated from disuse-induced atrophy model in young mice, mimicking the acute atrophy following long-term bed rest, compared to those of young and old mice. We identified a total of 23 differentially expressed miRNAs in old TA muscle as compared to young TA muscle, whereas only one miRNA (miR-455-3p) was found to be significantly differentially expressed between young mice and the disuse atrophy model. Similarly in serum, only one miRNA (miR-129-5p) was found to be differentially expressed in the disuse atrophy model as compared to young mice, while 9 differentially expressed miRNAs were identified in old mice as compared to young mice. Interestingly, miR-455-3p and miR-129-5p were significantly decreased both in disuse-induced atrophy model and aged mice as compared to young mice. Our results indicate that while disuse atrophy model generates distinct miRNA profiles in both TA muscle and serum as compared to young or old mice, miR-455-3p and miR-129-5p may be potential biomarkers of muscle atrophy.
Methods
Mice
Young (6-month-old) and old (24-month-old) C57BL/6 mice were purchased from the Laboratory Animal Resource Center (Korea Research Institute of Bioscience and Biotechnology, KRIBB). All animal experiments were performed according to protocols approved by the Animal Care and Use Committee of KRIBB. Blood was collected from infraorbital plexus of the young and old mice (n = 5, each) and stored on ice for at least 30 minutes to induce clotting. Serum was then obtained by centrifugation. Their TA muscle tissues were collected just after blood collection. TA muscle mass was recorded and RNA was immediately isolated from TA muscle and serum for miRNA-seq analysis. For induction of disuse muscle atrophy, hind limb of young mice (6-month-old, n = 5) was immobilized by stapling the foot exploiting normal dorsotibial flexion using an autosuture (Autosuture Royal 35W skin stapler, Tyco Healthcare) as previously described (13). At 14 days postimmobilization, both serum and TA muscle were collected as described above and TA muscle was used for miRNA sequencing.
RNA Isolation
From TA muscle tissues from young, old, and disuse atrophy model mice, total RNA was extracted using Trizol (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions. Quality and quantity of the RNAs were assessed by A260/A280 nm reading using NanoDrop1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE). RNA integrity was determined by running an aliquot of the RNA samples on a denaturing agarose gel stained with SYBR Green I.
For serum, total RNA was purified from 300 μL of serum samples using the miRNeasy serum/plasma kit (Qiagen) according to the manufacturer’s instructions (26). One microliter was used to quantify total RNA using Qubit-iT Ribogreen RNA assay kit (Invitrogen), which correlated well with quantities reported by picogram-scale capillary electrophoresis (Agilent Bioanalyzer).
miRNA Sequencing
Two micrograms of total RNA from each tissue sample and 10 ng of total RNA from each serum sample were used for small RNA cDNA library preparation (26). Using Illumina TruSeq Small RNA sample preparation kit, barcoded 3′-adapters and 5′-adapters were ligated to RNA in each sample. Synthesized cDNA amplified up to 15 cycles using indexes for multiplexing. Amplified DNAs were selected by the excision (145–160 bp) on 10% TBE polyacrylamide gel and purified. Eluted library was qualified using bioanalyzer and quantified using Qubit. Sequencing was performed on an Illumina Hiseq 2000 analyzer.
Data Analysis
As described previously, the sequencing data, in FASTQ format, were trimmed of adapter sequences using QUART (26). Sequences from each of the samples were aligned to the known mouse miRNAs present in mm10 using version 20 of mirBase. Analysis of the resulting read counts was performed in EdgeR (27). In order to control for differing library sizes, read counts were normalized by library size, giving values in reads per million (rpm), as described previously (28). After normalization, any miRNAs with fewer than 10 rpm more than 33% of the samples were excluded. Correlations were calculated with Pearson’s correlation coefficient using the function built-in to R (29). PCA was also performed using the function built-in to R.
Multilinear Regression Analysis
Regularized multilinear regression was performed on normalized miRNA read counts using GNU Octave version 3.8.2 (29,30); a λ value of 105 was empirically determined to provide the best fit for the data while avoiding overfitting. Separate models for serum and tissue were constructed; after the results were calculated, one of the young tissue samples was determined to be an outlier and the analysis was redone with that sample excluded. The models were trained using the fminunc function using the gradient of the cost function and a maximum of 400 iterations. For the disuse data, the models were trained with all young and old samples and then used to predict the age of the disuse samples. For each young and old sample, the models were trained with all other young and old samples and then used to predict the age for that sample. p values for the differences in predicted age for each group were calculated using the two-sample Wilcoxon test in R version 3.1.1 (29).
Quantitative RT-PCR Analysis
Total RNA isolated from tissue and serum was converted to complementary DNA using TaqMan Reverse Transcription kit (Applied Biosystems, Foster City, CA) with microRNA specific RT primer (Applied Biosystems). A TaqMan microRNA assay was performed using AB StepOne real-time PCR system to quantify relative miRNAs expression in these samples. The 20 µL total volume final reaction mixture consisted of 1 µL of TaqMan microRNA-specific primer, 10 µL of 2× Universal Master Mix with no AmpErase UNG (Applied Biosystems) and 1.3 µL of complementary DNA. PCR was performed using the following conditions: 50°C for 2 minutes, 95°C for 10 minutes, 40 cycles of 95°C for 15 seconds, and 60°C for 1 minute. U6 snRNA (Applied Biosystems) was used as an internal control for tissues. For serum, we used miR-30d-5p/miR-92a-3p/miR-320-3p for normalization. Means from different conditions were compared using the Student’s t test. A significance threshold of p less than .05 was used.
Cell Culture
C2C12 cells were purchased from ATCC and cultured in growth medium consisting of DMEM (Gibco) with antibiotics and 10% fetal bovine serum. Differentiation was initiated 24–48 hours after seeding by changing to the differentiation medium containing DMEM (Gibco) and antibiotics and 5% horse serum. After 5 days postinduction of differentiation, miRNA mimics (30 nM) were transfected into differentiated C2C12 myotubes using RNAiMAX (Invitrogen) according to the manufacturer’s protocols. For measurement of the diameters of differentiated myotubes, transfected C2C12 myotubes were fixed with ice cold methanol, and stained using Eosin-Y solution (Thermo Scientific, #6766007). After cells were then washed in phosphate-buffered saline buffer, the measurement of diameters of myotubes was performed using NIS-Elements BR software (Nikon). Quantitative data are presented as the means ± SEM. Differences between means were evaluated using Student’s paired t test. A p value of less than .05 was considered statistically significant.
Results
Discovery of miRNAs Expressed in Muscle Tissue and Serum of Young, Old, and Disuse Atrophy Model Mice
To analyze miRNA expression profiles in TA muscle or serum of young, old, and disuse atrophy model mice, we isolated them from 6-month-old (hereafter designated “young”), 24-month-old (hereafter designated “old”) mice, and 6-month-old mice with hind limbs immobilized by surgical stapling (hereafter designated “disuse”). Consistent with the previous reports (13,22), we found that TA muscle isolated from both old and disuse model mice showed a significant loss of muscle mass (Supplementary Figure 1). We next generated small RNA libraries from TA muscle tissues and serum of young, old, and disuse atrophy model mice. Sequencing of these libraries by an Illumina technology platform yielded a total of 13.6 × 106 reads from young tissue, 35.7 × 106 reads from old tissue, and 22.7 × 106 reads from disuse tissue and 1.33 × 106 reads from young serum, 0.714 × 106 reads from old serum, and 1.17 × 106 reads from disuse serum (Supplementary Table S1). miRNAs with less than 10 reads per million in more than 33% of samples (low-abundance miRNAs) were discarded due to the error rate of Illumina sequencing and stochastic variation in gene expression (31). For all comparisons, the number of reads for a given miRNA was normalized by the rpm method, yielding a normalized read count for each miRNA. In muscle tissue, we identified a total of 290 known miRNAs after removal of low-abundance miRNAs with a wide dynamic range of read counts, from 10 to over 100,000 reads per million (Figure 1A). About 6% of these miRNAs had a copy number greater than 10,000 and about 82% a copy number less than 1,000. The normalized read count for each miRNA indicated that miR-133a-3p was the most abundant miRNA detected in muscle tissue of young, old, and disuse atrophy model mice (Figure 1B).
Figure 1.
MiRNAs discovered in tibialis anterior (TA) muscle of young, old, and disuse atrophy model. (A) Distribution of the miRNAs expressed in TA muscle according to their sequence counts in young, old, and disuse atrophy model. The y-axis indicates the number of miRNAs with sequence read counts within the given range in the x-axis. (B) Top 10 known miRNAs expressed in TA muscle of young, old, and disuse atrophy model, which constitute nearly 80% of all miRNAs expressed in the tissue.
In the serum samples, we identified a total of 287 known miRNAs after removal of low-abundance miRNAs with a wide dynamic range of read counts, from 10 to over 100,000 reads per million (Figure 2A). About 7% of these miRNAs had a copy number greater than 10,000 and about 80% a copy number less than 1,000. The normalized read count for each miRNA indicated that miR-486-5p was the most abundant miRNA detected in serum of young, old, and disuse atrophy model mice (Figure 2B). The top 10 miRNAs by read count comprised approximately 80% of all sequence reads.
Figure 2.
MiRNAs discovered in serum of young, old, and disuse atrophy model. (A) Distribution of the miRNAs expressed in serum according to their sequence counts in young, old, and disuse atrophy model. The y-axis indicates the number of miRNAs with sequence read counts within the given range in the x-axis. (B) Top 10 known miRNAs expressed in serum of young, old, and disuse atrophy model, which constitute nearly 80% of all miRNAs expressed in the serum.
Correlation of miRNA Expression Between Aged Versus Disuse Muscle and Serum Versus Muscle
To test if miRNA expression levels are correlated between each group (young, old, and disuse) in serum and muscle tissue, we analyzed the correlation coefficient from miRNA scatter plots of each groups (Figure 3A). Figure 3A shows a set of scatter plots of each miRNA between different sample types (the average of each sample type was used). We found that within each sample type (ie, serum vs serum, tissue vs tissue), the correlation between different groups was high (median r = .99). By contrast, analysis between sample types (serum vs tissue) indicates a positive but much lower correlation (median r = .13). The significant difference (p = .009, Wilcoxon test) in correlation coefficients in different tissues suggests that the miRNAs induced or repressed in response to aging or disuse vary greatly depending upon the tissue in which that process occurs.
Figure 3.
Correlation analysis for miRNA signature between each pair of serum and muscle of young, old, and disuse atrophy model and age prediction by multilinear regression analysis in muscle tissue and serum. (A) Numbers are the Pearson’s correlation coefficient for each pairwise comparison. (B) Predicted age of young, old, and disuse mice in muscle tissue. (C) Predicted age of young, old, and disuse mice in serum. (D) Predicted age of disuse mice in tissue and serum. Error bars indicate the standard error of the mean. *p value < .05; **p value < .01.
The miRNA signature of age/disuse is highly cell type-specific. Within the serum samples, miRNA signature in old mice was more correlated with in disuse atrophy model than it is with young mice (Figure 3A). Within the tissue, we again found that miRNA expression in old mice is more correlated with the disuse atrophy model than with young mice. These data suggest that miRNAs in disuse atrophy model may represent a state analogous to accelerated aging although this interpretation is highly speculative due to the small magnitude of the difference in correlation coefficients (Figure 3A: .9996 vs .9983 in tissue and .9759 vs .9681 in serum).
Predicted Age of Young, Old, and Disuse Muscle and Serum
To predict the extent by which the disuse mimics aging, we predicted the biological age of the disuse mice using a multilinear regression model trained on the data from the young and old mice (Figure 3B–D). Separate models for serum and tissue were constructed. In order to gauge the extent to which disuse mimicked aging, we predicted the biological ages of the disuse samples after training the models using the data from the young and old samples. To validate our predictions, we separately predicted the age of each young and old sample after training the models using the data from all other young and old samples (Figure 3B,C). We found that the predicted age of muscle in disuse average around 16 months, which substantially departs from their chronological age (6 months) and closely approaches to the chronological age of the old mice (24 months; Figure 3B). These data suggested that the muscle tissue of disuse model might mimic the miRNA signature of aging. In the serum samples, the predicted age of disuse mice was 5.7 months, almost exactly the same as their chronological age of 6 months, suggesting that serum in the disuse model did not mimic aging of old mice. The difference in predicted biological age between the disuse muscle samples and disuse serum samples was statistically significant (p = .008, Figure 3D).
MiRNAs Differentially Expressed Between Muscle Tissue or Serum of Young, Old, and Disuse Atrophy Model Mice
We analyzed differentially expressed miRNAs with a fold change ≥ 1.5 and false discovery rate (FDR)-adjusted p value of less than .05 between each comparison (young vs old, young vs disuse atrophy model, and old vs disuse atrophy model) (Tables 1–6). A complete list of miRNAs sequenced in each library, normalized read counts, and fold changes in muscle tissue and serum of young, old, and disuse atrophy model mice is provided in Supplementary Tables S2–S7. For visualization, we also generated Venn diagram of miRNAs differentially expressed in young, old, and disuse atrophy model mice and categorized them as Young versus Old (Y vs O), Young versus Disuse (Y vs D), and Old versus Disuse (O vs D) for muscle (Figure 4A) and for serum (Figure 4B). In muscle tissue, we identified a total of 23 differentially expressed miRNAs between young and old mice. Of the 23 differentially expressed miRNAs, 17 miRNAs were significantly decreased and 6 miRNAs were significantly increased in muscle tissues of old mice as compared to young mice (Figure 4A and Table 1). We found one miRNA differentially expressed between young and disuse, miR-455-3p, which was significantly decreased in disuse (Figure 4A and Table 3). Interestingly, miR-455-3p was downregulated in muscle tissue of both of old mice and the disuse atrophy model as compared to young mice (Figure 4A).
Table 1.
Differentially Expressed miRNAs in Tibialis Anterior Muscle of Young and Old Mice (false discovery rate < 0.05, fold change ≥ 1.5)
| Young | Old | Fold Change | P Value | |
|---|---|---|---|---|
| mmu-miR-381-3p | 46.907 | 10.229 | 0.218 | 1.94E−13 |
| mmu-miR-673-5p | 41.483 | 11.617 | 0.280 | 3.84E−09 |
| mmu-miR-341-3p | 13.703 | 3.985 | 0.291 | 4.91E−08 |
| mmu-miR-455-3p | 28.884 | 8.516 | 0.295 | .00000105 |
| mmu-miR-541-5p | 29.571 | 8.745 | 0.296 | 5.32E−08 |
| mmu-miR-434-3p | 414.730 | 124.578 | 0.300 | 2.14E−14 |
| mmu-miR-673-3p | 31.208 | 9.461 | 0.303 | 2.02E−10 |
| mmu-miR-329-5p | 27.330 | 9.674 | 0.354 | .003746829 |
| mmu-miR-127-3p | 6,709.833 | 2,427.742 | 0.362 | 2.56E−13 |
| mmu-miR-431-5p | 33.423 | 13.663 | 0.409 | .0000376 |
| mmu-miR-667-3p | 41.902 | 20.218 | 0.482 | .000147611 |
| mmu-miR-540-3p | 66.517 | 32.726 | 0.492 | .000136101 |
| mmu-miR-485-5p | 40.466 | 20.023 | 0.495 | .004444364 |
| mmu-miR-411-5p | 80.166 | 39.787 | 0.496 | .00000308 |
| mmu-miR-3107-3p | 112.764 | 68.812 | 0.610 | .004267488 |
| mmu-miR-151-3p | 3,609.317 | 2,215.847 | 0.614 | .000869639 |
| mmu-miR-181c-5p | 126.738 | 77.865 | 0.614 | .000498415 |
| mmu-miR-378a-5p | 212.489 | 321.702 | 1.514 | .003601157 |
| mmu-miR-29a-3p | 2,250.372 | 3,825.834 | 1.700 | .000000813 |
| mmu-miR-138-5p | 49.522 | 88.425 | 1.786 | .000456039 |
| mmu-miR-92a-1-5p | 10.283 | 20.409 | 1.985 | .001380018 |
| mmu-miR-29b-2-5p | 14.676 | 32.746 | 2.231 | .001275543 |
| mmu-miR-342-3p | 253.065 | 572.316 | 2.262 | .000300269 |
Table 6.
Differentially Expressed miRNAs in Serum of Old Mice and Disuse Atrophy Model (false discovery rate < 0.05, fold change ≥ 1.5)
| Old | Disuse | Fold Change | p Value | |
|---|---|---|---|---|
| mmu-miR-146a-5p | 48101.592 | 20792.696 | 0.432 | .015397866 |
| mmu-miR-192-5p | 19190.471 | 9114.908 | 0.475 | .015397866 |
Figure 4.
Venn diagram of differentially expressed miRNAs in young, old, and disuse and qRT-PCR validation. (A) Venn diagram of differentially expressed miRNAs in muscle tissues in young, old, and disuse. (B) Venn diagram of differentially expressed miRNAs in serum samples in young, old, and disuse. Fold change ≥ 1.5 and false discovery rate < 0.05. Representative miRNAs were determined by TaqMan qRT-PCR each in muscle (A) and serum (B). Relative expression level of miRNAs in old and disuse as compared to young was shown. The qRT-PCR results were normalized by U6 snRNA for muscle tissue samples and by the average of miR-30d, -92a, and -320 for serum samples. Standard error of mean for qPCR values was shown. Y = young mice; O = old mice; D = disuse atrophy model. *p value < .05; **p value < .01.
Table 3.
Differentially Expressed miRNAs in Muscle of Old Mice and Disuse Atrophy Model (false discovery rate < 0.05, fold change ≥ 1.5)
| Old | Disuse | Fold Change | p Value | |
|---|---|---|---|---|
| mmu-miR-6240 | 20.133 | 8.403 | 0.417 | .004601214 |
| mmu-miR-34a-5p | 122.593 | 60.938 | 0.497 | .00507763 |
| mmu-miR-338-3p | 95.478 | 53.345 | 0.559 | .017379069 |
| mmu-miR-182-5p | 21.912 | 12.431 | 0.567 | .027907553 |
| mmu-let-7d-3p | 1,662.157 | 2,571.545 | 1.547 | .017379069 |
| mmu-miR-411-5p | 39.787 | 64.552 | 1.622 | .048082686 |
| mmu-miR-154-5p | 29.645 | 51.115 | 1.724 | .019156282 |
| mmu-miR-667-3p | 20.218 | 38.600 | 1.909 | .004601214 |
| mmu-miR-127-3p | 2,427.742 | 5,245.867 | 2.161 | .00000232 |
| mmu-miR-337-5p | 12.537 | 31.832 | 2.539 | .000758143 |
| mmu-miR-673-3p | 9.461 | 24.208 | 2.559 | .000758143 |
| mmu-miR-673-5p | 11.617 | 31.937 | 2.749 | .00041064 |
| mmu-miR-434-3p | 124.578 | 344.380 | 2.764 | 3.625E−08 |
| mmu-miR-341-3p | 3.985 | 15.439 | 3.874 | .000296525 |
| mmu-miR-381-3p | 10.229 | 40.186 | 3.929 | 6.6265E−08 |
Table 2.
Differentially Expressed miRNAs in Tibialis Anterior Muscle of Young and Disuse Atrophy Model (false discovery rate < 0.05, fold change ≥ 1.5)
| Young | Disuse | Fold Change | p Value | |
|---|---|---|---|---|
| mmu-miR-455-3p | 28.884 | 9.724 | 0.337 | 5.67E−06 |
In serum samples, we identified a total of 9 differentially expressed miRNAs between young and old mice, among which 7 miRNAs were significantly decreased and 2 miRNAs were significantly increased in old mice as compared to young mice (Figure 4B and Table 4). Only one miRNA (miR-129-5p) was found to be differentially expressed with significantly decreased expression in disuse as compared to young mice (Figure 4B and Table 5). This miRNA, miR-129-5p, was downregulated both in old and disuse as compared to young mice. We identified 2 miRNAs (miR-146a-5p and miR-192-5p) differentially expressed between old and disuse and these miRNAs were overlapped with miRNAs differentially expressed between young and old mice (Figure 4B and Table 6).
Table 4.
Differentially Expressed miRNAs in Serum of Young and Old Mice (false discovery rate < 0.05, fold change ≥ 1.5)
| Young | Old | Fold Change | p Value | |
|---|---|---|---|---|
| mmu-miR-211-5p | 316.755 | 17.132 | 0.054 | .00000242 |
| mmu-miR-129-5p | 55.258 | 4.553 | 0.082 | .001096165 |
| mmu-miR-144-5p | 246.772 | 43.876 | 0.178 | .0000836 |
| mmu-miR-127-3p | 1,674.890 | 445.105 | 0.266 | .0000271 |
| mmu-miR-434-3p | 133.424 | 38.345 | 0.287 | .000842903 |
| mmu-miR-98-5p | 244.058 | 76.057 | 0.312 | .000274143 |
| mmu-miR-15b-5p | 372.337 | 159.374 | 0.428 | .000571212 |
| mmu-miR-192-5p | 9,889.315 | 19,190.471 | 1.941 | .00122328 |
| mmu-miR-146a-5p | 20,992.180 | 48,101.592 | 2.291 | .000298506 |
Table 5.
Differentially Expressed miRNAs in Serum of Young and Disuse Atrophy Model (false discovery rate < 0.05, fold change ≥ 1.5)
| Young | Disuse | Fold Change | p Value | |
|---|---|---|---|---|
| mmu-miR-129-5p | 55.258 | 2.942 | 0.053 | .0000435 |
These data suggested that while the disuse model has a distinct miRNA expression profile as compared to young or aged mice, the directional changes in expression levels of some miRNAs, including miR-455-3p and miR-129-5p, are similar between old and disuse atrophy model as compared to young mice. Furthermore, our study indicates that while muscle and serum have distinct age-related changes in miRNA expression, some miRNAs showed the same directionality in expression changes. In particular, miR-127-3p and miR-434-3p were significantly decreased with aging both in muscle and serum, both of which have been reported to be downregulated in aged skeletal muscle tissue (22). We used Taqman real-time reverse transcription-PCR (qRT-PCR) qRT-PCR and validated these miRNAs. qRT-PCR results for miR-455-3p, miR-127-3p, and miR-434-3p in muscle tissue samples were consistent with the miRNA-seq results (Figure 4A). In serum samples, miR-434-3p and miR-146a-5p were validated and showed the consistent results with miRNA-seq (Figure 4B). However, miR-129-5p was not detectable in qRT-PCR due to its low abundance. miR-127-3p and miR-192-5p showed the decreased expression pattern in old but not significant. These data suggest that sequencing is a gold standard to identify differentially expressed miRNAs (31).
The Role of miR-455-3p in Skeletal Muscle Cells
miR-455 has been reported to be highly expressed in muscle and decreased in aerobically adapted human muscle biopsies in contrast to controls (32). miR-455 has also been implicated in differentiation of brown adipocytes derived from a common progenitor, mesenchymal stem cells, of muscle cells (32). Thus, miR-455 might have a crucial role not only in muscle but also in metabolic homeostasis in the elderly adults. To investigate the biological function of miR-455-3p, which we found to be downregulated in muscle tissues of disuse and old mice, we transfected fully differentiated C2C12 myoblasts with miR-455-3p mimic. As shown in Figure 5, eosin stains showed that miR-455-3p transfected C2C12 myotubes had significantly larger diameters, comparing to control myotubes, showing that miR-455-3p induced hypertrophic myotube formation. Indeed, we found that some of the predicted target genes of miR-455-3p including Paired-like homeodomain transcription factor 1 (PITX1) and retinoid X receptor-β (RXRB) are involved in muscle dystrophy and aging, respectively (Supplementary Table S8). Another predicted target gene, RXRB also has been known to be increased in the aged skeletal muscle (33). Future investigation on the targets of miR-455-3p may provide important insight into its role in age-related muscle atrophy.
Figure 5.
Biological function of miR-445-3p in skeletal muscle cells. (A) Representative eosin-stained images of control and miR-455-3p mimic-treated C2C12 myotubes. Scale bar, 200 µm. (B) The diameter of myotubes was calculated and then represented by the percentage of fiber diameter. Six different views were randomly selected for measurement of the diameter.
Discussion
Several studies have reported the alternation of miRNA expression during muscle aging (21,22) but no comparative study has been conducted between disuse-induced muscle atrophy and muscle aging. To our knowledge, this is the first study to comprehensively and comparatively profile miRNA expression both in muscle and serum of disuse atrophy model as compared to those of young and old mice.
We identified 23 (muscle) and 9 (serum) miRNAs differentially expressed in old mice as compared to young mice (Figure 4A,B). Interestingly, we found only one miRNA each in muscle (miR-455-3p) and serum (miR-129-5p), differentially expressed in disuse atrophy model as compared to young mice. These were downregulated not only in disuse mice but also in old mice as compared to young mice (Figure 4A,B). These data indicate that these miRNAs can be common factors in muscle and serum respectively indicating muscle aging. When we analyzed differentially expressed miRNAs in muscle between old mice and disuse atrophy mice model, we identified 15 miRNAs (Table 3). Among them, eight miRNAs were differentially expressed between young and old mice (Figure 4A) and seven miRNAs were found to be differentially expressed only between old mice and disuse atrophy model (Figure 4A) suggesting that muscle tissue in disuse atrophy model induces unique miRNA signatures compared to aged muscle. This may provide important information in the study of muscle aging using disuse atrophy model. In serum, we found all differentially expressed miRNAs (miR-146a-5p and miR-192-5p) between old mice and disuse atrophy model were also differentially expressed between young and old mice (Figure 4B) indicating that miRNA signatures in serum in disuse atrophy model may be more related with that in young mice. These two miRNAs have been reported to be increased in serum of aged mice and reversed in the expression level after calorie restriction (34).
Of the differentially expressed miRNAs, several are particularly interesting. We found that miR-455-3p was downregulated in muscle tissues of aged mice and disuse atrophy model mice as compared to young mice (Figure 4A). Interestingly, we found that miR-455-3p can induce a skeletal muscle hypertrophic phenotype in differentiated C2C12 cells (Figure 5). We also analyzed the predicted targets of miR-455-3p using Targetscan and DAVID functional annotation bioinformatics and found that some genes including Paired-like homeodomain transcription factor 1 (PITX1) and retinoid X receptor-β (RXRB) have known to be involved in muscle dystrophy and aging, respectively (Supplementary Table S8). PITX1 has been reported to be upregulated in patients with facioscapulohumeral muscular dystrophy and muscle-specific Pitx1 transgenic mouse model showed significant loss of body weight, muscle mass, and reduction of muscle fiber diameters (35). Another predicted target gene, RXRB also has been known to be increased in the aged skeletal muscle (33). These data will promote further studies for underlying mechanism in miR-455-3p-dependent myotube hypertrophy.
In our study, the expression of miR-127-3p and miR-434-3p was downregulated during aging in muscle tissues and interestingly in serum as well (Figure 4). These data are consistent with our previous study demonstrating that miR-127-3p and miR-434-3p are downregulated in skeletal muscle of aged mice (22). The correlation of miRNA expression between muscle tissue and serum indicates that the expression of some miRNAs changes with age in both muscle and serum, suggesting their potential contribution to the aging-related phenotypes such as muscle dysfunction (36). Whereas the global pattern of correlation coefficients showed distinct age-related miRNA signatures between muscle and serum (Figure 3A), these two miRNAs might be explored as universal biomarkers of muscle aging. McCarthy and colleagues profiled soleus muscle miRNAs differentially expressed between normal and hind limb-suspended rats, which had their limbs suspended for 2 or 7 days (37) but none of the miRNAs they found to be differentially expressed were identified as having significant expression changes in our study. This lack of overlap may be due to the different species of model organism used, or short-term (2 or 7 days, rat) versus long-term (14 days, mice) immobilization of muscle.
One of the major questions of our study was whether disuse-induced muscle atrophy phenocopies aging-related muscle dysfunction in terms of the miRNA profiles. From the coefficient correlation data (Figure 3A), we found that disuse-induced atrophy tends to be an intermediate state in muscle and serum between young and aged mice but there was a low correlation between miRNA expression in aged mice and disuse atrophy model mice. This is in line with the recent studies in model organisms that have indicated that sarcopenia, a major age-related phenotype, fundamentally differs from the rapid atrophy of muscles observed following disuse and fasting (38). In mice, the maximum isometric force (an indicator of muscle strength) has been reported to be decreased to a greater extent than muscle mass during aging, suggesting that muscle function is a more important health parameter than muscle mass (39). Similarly, the decrease in maximal strength in aging humans has been reported to be three times greater than the decline in muscle mass, suggesting that muscle mass and strength are independently regulated during aging (3). Moreover, muscle strength has been reported to be a better predictor of mortality during aging (40). Thus, the change in miRNA expression associated with the age-related loss of muscle function in aged mice could be derived from different mechanisms than the changes associated with the disuse-induced atrophy model, as reflected by their distinct miRNA signatures. Nevertheless, the common miRNA (miR-434-3p) that we showed to decrease with age both in muscle and serum provide an opportunity to develop potential biomarkers of age-related muscle atrophy and to elucidate the molecular mechanisms underlying muscle dysfunction with age. Also, we found the predicted biological age of muscle after disuse-induced atrophy was closer to that of old mice than that of young mice (Figure 3B–D) although predicted biological age of serum in disuse showed almost same age with young mice, suggesting the serum may not be a good surrogate for muscle with the alteration of circulating miRNAs during aging or disuse-induced atrophy. Nevertheless, there is no doubt that the use of animal models with disuse-induced muscle atrophy, mimicking the atrophy following long term bed rest which is more likely to occur in the elderly adults, will provide a part of important evidence to understand the mechanism of muscle aging.
Supplementary Material
Supplementary data are available at The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences online.
Funding
This work was supported by NIH grants AG017242, GM104459, and CA180126 (awarded to Y.S.) and by a grant from KRIBB Research Initiative Program (awarded to K-S.K. and Y.S.) and and NRF grant 2013M3A9B6076413 (awarded to K-S.K.).
Conflict of Interest
The authors have no conflict of interest to declare.
Supplementary Material
References
- 1. Marcell TJ. Sarcopenia: causes, consequences, and preventions. J Gerontol A Biol Sci Med Sci. 2003;58:M911–M916. [DOI] [PubMed] [Google Scholar]
- 2. Frontera WR, Hughes VA, Lutz KJ, Evans WJ. A cross-sectional study of muscle strength and mass in 45- to 78-yr-old men and women. J Appl Physiol (1985). 1991;71:644–650. [DOI] [PubMed] [Google Scholar]
- 3. Goodpaster BH, Park SW, Harris TB, et al. The loss of skeletal muscle strength, mass, and quality in older adults: the health, aging and body composition study. J Gerontol A Biol Sci Med Sci. 2006;61:1059–1064. [DOI] [PubMed] [Google Scholar]
- 4. Hughes VA, Frontera WR, Wood M, et al. Longitudinal muscle strength changes in older adults: influence of muscle mass, physical activity, and health. J Gerontol A Biol Sci Med Sci. 2001;56:B209–B217. [DOI] [PubMed] [Google Scholar]
- 5. von Haehling S, Morley JE, Anker SD. An overview of sarcopenia: facts and numbers on prevalence and clinical impact. J Cachexia Sarcopenia Muscle. 2010;1:129–133. doi: 10.1007/s13539-010-0014-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Afilalo J, Karunananthan S, Eisenberg MJ, Alexander KP, Bergman H. Role of frailty in patients with cardiovascular disease. Am J Cardiol. 2009;103:1616–1621. doi:10.1016/j.amjcard.2009.01.375 [DOI] [PubMed] [Google Scholar]
- 7. Morley JE. Sarcopenia: diagnosis and treatment. J Nutr Health Aging. 2008;12:452–456. [DOI] [PubMed] [Google Scholar]
- 8. English KL, Paddon-Jones D. Protecting muscle mass and function in older adults during bed rest. Curr Opin Clin Nutr Metab Care. 2010;13:34–39. doi:10.1097/MCO.0b013e328333aa66 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Kortebein P, Ferrando A, Lombeida J, Wolfe R, Evans WJ. Effect of 10 days of bed rest on skeletal muscle in healthy older adults. JAMA. 2007;297:1772–1774. doi:10.1001/jama.297.16.1772-b [DOI] [PubMed] [Google Scholar]
- 10. Gill TM, Allore H, Guo Z. The deleterious effects of bed rest among community-living older persons. J Gerontol A Biol Sci Med Sci. 2004;59:755–761. [DOI] [PubMed] [Google Scholar]
- 11. Morey-Holton ER, Globus RK. Hindlimb unloading rodent model: technical aspects. J Appl Physiol (1985). 2002;92:1367–1377. doi:10.1152/japplphysiol.00969.2001 [DOI] [PubMed] [Google Scholar]
- 12. Frimel TN, Kapadia F, Gaidosh GS, Li Y, Walter GA, Vandenborne K. A model of muscle atrophy using cast immobilization in mice. Muscle Nerve. 2005;32:672–674. doi:10.1002/mus.20399 [DOI] [PubMed] [Google Scholar]
- 13. Caron AZ, Drouin G, Desrosiers J, Trensz F, Grenier G. A novel hindlimb immobilization procedure for studying skeletal muscle atrophy and recovery in mouse. J Appl Physiol (1985). 2009;106:2049–2059. doi: 10.1152/japplphysiol.91505.2008 [DOI] [PubMed] [Google Scholar]
- 14. Hamrick MW, Herberg S, Arounleut P, et al. The adipokine leptin increases skeletal muscle mass and significantly alters skeletal muscle miRNA expression profile in aged mice. Biochem Biophys Res Commun. 2010;400:379–383. doi:10.1016/j.bbrc.2010.08.079 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Drummond MJ, McCarthy JJ, Sinha M, et al. Aging and microRNA expression in human skeletal muscle: a microarray and bioinformatics analysis. Physiol Genomics. 2011;43:595–603. doi: 10.1152/physiolgenomics.00148.2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Hwang CY, Kim K, Choi JY, et al. Quantitative proteome analysis of age-related changes in mouse gastrocnemius muscle using mTRAQ. Proteomics. 2014;14:121–132. doi: 10.1002/pmic.201200497. [DOI] [PubMed] [Google Scholar]
- 17. Thum T, Galuppo P, Wolf C, et al. MicroRNAs in the human heart: a clue to fetal gene reprogramming in heart failure. Circulation. 2007;116:258–267. doi:10.1161/CIRCULATIONAHA.107.687947 [DOI] [PubMed] [Google Scholar]
- 18. Liu N, Williams AH, Maxeiner JM, et al. microRNA-206 promotes skeletal muscle regeneration and delays progression of Duchenne muscular dystrophy in mice. J Clin Invest. 2012;122:2054–2065. doi:10.1172/JCI62656 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Cheung TH, Quach NL, Charville GW, et al. Maintenance of muscle stem-cell quiescence by microRNA-489. Nature. 2012;482:524–528. doi:10.1038/nature10834 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Eisenberg I, Eran A, Nishino I, et al. Distinctive patterns of microRNA expression in primary muscular disorders. Proc Natl Acad Sci USA. 2007;104:17016–17021. doi:10.1073/pnas.0708115104 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Mercken EM, Majounie E, Ding J, et al. Age-associated miRNA alterations in skeletal muscle from rhesus monkeys reversed by caloric restriction. Aging (Albany NY). 2013;5:692–703. doi:10.18632/aging.100598 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Kim JY, Park YK, Lee KP, et al. Genome-wide profiling of the microRNA-mRNA regulatory network in skeletal muscle with aging. Aging (Albany NY). 2014;6:524–544. doi:10.18632/aging.100677 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Lee KP, Shin YJ, Panda AC, et al. miR-431 promotes differentiation and regeneration of old skeletal muscle by targeting Smad4. Genes Dev. 2015;29:1605–1617. doi:10.1101/gad.263574.115 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Creemers EE, Tijsen AJ, Pinto YM. Circulating microRNAs: novel biomarkers and extracellular communicators in cardiovascular disease? Circ Res. 2012;110:483–495. doi:10.1161/CIRCRESAHA.111.247452 [DOI] [PubMed] [Google Scholar]
- 25. Olivieri F, Spazzafumo L, Santini G, et al. Age-related differences in the expression of circulating microRNAs: miR-21 as a new circulating marker of inflammaging. Mech Ageing Dev. 2012;133:675–685. doi:10.1016/j.mad.2012.09.004 [DOI] [PubMed] [Google Scholar]
- 26. Ho GY, Jung HJ, Schoen RE, et al. Differential expression of circulating microRNAs according to severity of colorectal neoplasia. Transl Res. 2015;166:225–232. doi:10.1016/j.trsl.2015.02.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Robinson MD, McCarthy DJ, Smyth GK. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010;26:139–140. doi:10.1093/bioinformatics/btp616 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Milholland B, Gombar S, Suh Y. SMiRK: an automated pipeline for miRNA analysis. Source J Genom. 2015;1:1–8. [PMC free article] [PubMed] [Google Scholar]
- 29. Team RC. R: A Language and Environment for Statistical Computing. 2014. http://www.R-project.org/ [Google Scholar]
- 30. Eaton JW BD, Hauberg S, Wehbring R. GNU Octave Version 3.8.1 Manual: A High-Level Interactive Language for Numerical Computations. 2014. https://www.gnu.org/software/octave/doc/octave-4.0.1.pdf [Google Scholar]
- 31. Gombar S, Jung HJ, Dong F, et al. Comprehensive microRNA profiling in B-cells of human centenarians by massively parallel sequencing. BMC Genomics. 2012;13:353. doi:10.1186/1471-2164-13-353 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Walden TB, Timmons JA, Keller P, Nedergaard J, Cannon B. Distinct expression of muscle-specific microRNAs (myomirs) in brown adipocytes. J Cell Physiol. 2009;218:444–449. doi:10.1002/jcp.21621 [DOI] [PubMed] [Google Scholar]
- 33. Giresi PG, Stevenson EJ, Theilhaber J, et al. Identification of a molecular signature of sarcopenia. Physiol Genomics. 2005;21:253–263. doi:10.1152/physiolgenomics.00249.2004 [DOI] [PubMed] [Google Scholar]
- 34. Dhahbi JM, Spindler SR, Atamna H, et al. Deep sequencing identifies circulating mouse miRNAs that are functionally implicated in manifestations of aging and responsive to calorie restriction. Aging (Albany NY). 2013;5:130–141. doi:10.18632/aging.100540 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Pandey SN, Cabotage J, Shi R, et al. Conditional over-expression of PITX1 causes skeletal muscle dystrophy in mice. Biol Open. 2012;1:629–639. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Koutsoulidou A, Kyriakides TC, Papadimas GK, et al. Elevated muscle-specific miRNAs in serum of myotonic dystrophy patients relate to muscle disease progress. PLoS One. 2015;10:e0125341. doi:10.1371/journal.pone.0125341 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. McCarthy JJ, Esser KA, Peterson CA, Dupont-Versteegden EE. Evidence of MyomiR network regulation of beta-myosin heavy chain gene expression during skeletal muscle atrophy. Physiol Genomics. 2009;39:219–226. doi:10.1152/physiolgenomics.00042.2009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Demontis F, Piccirillo R, Goldberg AL, Perrimon N. Mechanisms of skeletal muscle aging: insights from Drosophila and mammalian models. Dis Model Mech. 2013;6:1339–1352. doi:10.1242/dmm.012559 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. González E, Delbono O. Age-dependent fatigue in single intact fast- and slow fibers from mouse EDL and soleus skeletal muscles. Mech Ageing Dev. 2001;122:1019–1032. [DOI] [PubMed] [Google Scholar]
- 40. Metter EJ, Talbot LA, Schrager M, Conwit R. Skeletal muscle strength as a predictor of all-cause mortality in healthy men. J Gerontol A Biol Sci Med Sci. 2002;57:B359–B365. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.