Deciphering the microRNA transcriptome of skeletal muscle during porcine development

MicroRNAs (miRNAs) play critical roles in many important biological processes, such as growth and development in mammals. Various studies of porcine muscle development have mainly focused on identifying miRNAs that are important for fetal and adult muscle development; however, little is known about the role of miRNAs in middle-aged muscle development. Here, we present a comprehensive investigation of miRNA transcriptomes across five porcine muscle development stages, including one prenatal and four postnatal stages. We identified 404 known porcine miRNAs, 118 novel miRNAs, and 101 miRNAs that are conserved in other mammals. A set of universally abundant miRNAs was found across the distinct muscle development stages. This set of miRNAs may play important housekeeping roles that are involved in myogenesis. A short time-series expression miner analysis indicated significant variations in miRNA expression across distinct muscle development stages. We also found enhanced differentiation- and morphogenesis-related miRNA levels in the embryonic stage; conversely, apoptosis-related miRNA levels increased relatively later in muscle development. These results provide integral insight into miRNA function throughout pig muscle development stages. Our findings will promote further development of the pig as a model organism for human age-related muscle disease research.

After sacrifice, all muscle tissues were fixed in 10% neutral buffered formalin solution, 94 embedded in paraffin using a TP1020 semi-enclosed tissue processor (Leica), sliced at a thickness 95 of 6 μm using RM2135 rotary microtome (Leica), and stained with hematoxylin and eosin (H&E).
96 The mean diameter of muscle cells was calculated as the geometric average of the maximum and 97 minimum diameter, and 100 cells were measured for each sample in randomly selected fields using 98 a TE2000 fluorescence microscope (Nikon) and Image Pro-Plus 7.0 software.

123
The potential targets of a certain miRNA were predicted by PicTar (Krek et al. 2005)and 124 TargetScan Table S1. Three endogenous control genes (U6 snRNA, 18S rRNA, and 5S rRNA) 147 were used in this assay. The ΔΔCt method was used to determine the expression level differences 148 between surveyed samples. 149

Results and Discussion
150 Phenotypic measurements

151
In this study, we investigated muscle fiber diameter and body weight change during the postnatal 152 muscle development stage. As shown in Fig. S1A-C, we found that muscle fiber diameter and 153 body weight were significantly increased (P < 0.001, one-way analysis of variance) from 0 d to 7 154 y, which is consistent with results from previous studies in pigs Fiedler et al. 155 1997). Notably, we found that muscle fiber diameter and body weight growth rate significantly 156 slowed with age (P < 0.05) (Fig. 1A, B). This age-related decrease of muscle fiber growth rate is 157 likely due to decreases in protein synthesis and increases in protein degradation (Jackman & 158 Kandarian 2004). These phenotypic differences indicated possible underlying intrinsic differences 159 in molecular mechanisms.  (Fig. S2). Of these high-quality reads (Fig. S3), the majority 166 (90.35 ± 1.34%, n = 9) of the small RNAs were approximately 21-24 nt in length. More than half 167 of the reads were 22 nt in length (57.03 ± 7.53%, n = 9), followed by 23 nt (16.40 ± 4.34%, n = 168 9), 21 nt (12.37 ± 5.26%, n = 9), and 24 nt (4.55 ± 1.49%, n = 9), which are typical lengths of 169 Dicer-processed products (Bartel 2009;Berezikov 2011). This result indicates that small RNA 170 sequencing is a reliable approach to generate miRNA reads for further analysis. Manuscript to be reviewed The high-quality reads (sequences) were compared against the pig genome and divided into 173 three groups (Table 1) 176 Specifically, 286 miRNAs were known, but 118 were not previously identified porcine miRNAs 177 (Table S2). Of the porcine conserved miRNAs, 101 corresponded to 71 other known mammalian 178 species except for porcine pre-miRNAs in miRBase 19.0, which were further mapped to the pig 179 genome (Table S3). Of the porcine putative new miRNAs, 431 corresponded to 325 candidate pre-180 miRNAs that were not mapped to mammalian pre-miRNAs in miRBase 19.0, but predicted RNA 181 hairpins derived from the pig genome, and the extended genome sequences form hairpins, which 182 were labeled PN (Table S4). Therefore, 629 pre-miRNAs that encode 936 mature miRNAs were 183 identified. Because there are distinct pre-miRNAs and genomic loci that code identical mature 184 sequences, 903 unique miRNAs were finally identified (Table S5).  (Fig. 2). Of these miRNAs, six (miR-133a-1/-2-3p, let-7a-1/-2-5p, miR-27b-3p, miR-193 26a-5p, miR-1-3p, and let-7f-1/-2-5p) were shared by all five stages and were closely related to 194 myogenesis, cell growth, myocyte proliferation, and cell apoptosis. For example, miR-133a is a 195 muscle-enriched miRNA that inhibits proliferation of progenitor cells and promotes myogenesis 296 in relatively older animals . Moreover, it has been demonstrated that muscle 297 atrophy induced in middle-aged or older animals is associated with higher amounts of cell 298 apoptosis and death (Dirks & Leeuwenburgh 2005). This may be the primary reason these miRNAs 299 that play vital roles in apoptosis are highly expressed in middle-aged stage. Particularly, miR-1 300 showed significantly increasing expression level in 180d and 7y stages (Fig. 5A and Table S7). By 301 using both C2C12 myotubes and dex-induced muscular atrophy mouse models, Kukreti et al. 312 Conclusions

313
In this study, we performed a comprehensive investigation of miRNA expression pattern in 314 skeletal muscle across various developmental processes, from embryonic to adult stages, in pigs. 315 We identified a number of DE miRNAs that were associated with porcine muscular development.