BackgroundThe growth and development of skeletal muscle directly impacts the quantity and quality of pork production. Chinese indigenous pig breeds and exotic species vary greatly in terms of muscle production and performance traits. We present transcriptome profiles of 110 skeletal muscle samples from Tongcheng (TC) and Yorkshire (YK) pigs at 11 developmental periods (30, 40, 55, 63, 70, 90, and 105 days of gestation, and 0, 1, 3, and 5 weeks of age) using digital gene expression on Solexa/Illumina’s Genome Analyzer platform to investigate the differences in prenatal and postnatal skeletal muscle between the two breeds.ResultsMuscle morphological changes indicate the importance of primary fiber formation from 30 to 40 dpc (days post coitus), and secondary fiber formation from 55 to 70 dpc. We screened 4,331 differentially expressed genes in TC and 2,259 in YK (log2 ratio >1 and probability >0.7). Cluster analysis showed different gene expression patterns between TC and YK pigs. The transcripts were annotated in terms of Gene Ontology related to muscle development. We found that the genes CXCL10, EIF2B5, PSMA6, FBXO32, and LOC100622249 played vital roles in the muscle regulatory networks in the TC breed, whereas the genes SGCD, ENG, THBD, AQP4, and BTG2 played dominant roles in the YK breed. These genes showed breed-specific and development-dependent differential expression patterns. Furthermore, 984 genes were identified in myogenesis. A heat map showed that significantly enriched pathways (FDR <0.05) had stage-specific functional regulatory mechanisms. Finally, the differentially expressed genes from our sequencing results were confirmed by real-time quantitative polymerase chain reaction.ConclusionsThis study detected many functional genes and showed differences in the molecular mechanisms of skeletal muscle development between TC and YK pigs. TC pigs showed slower muscle growth and more complicated genetic regulation than YK pigs. Many differentially expressed genes showed breed-specific expression patterns. Our data provide a better understanding of skeletal muscle developmental differences and valuable information for improving pork quality.Electronic supplementary materialThe online version of this article (doi:10.1186/s12864-015-1580-7) contains supplementary material, which is available to authorized users.

Tongcheng (TC) and Yorkshire (YK) are two pig breeds with distinctive muscle morphology. Porcine microRNAome (miRNAome) of the longissimus muscle during five developmental stages (40, 55, 63, 70, and 90 days post coitum (dpc)) was explored by Solexa sequencing in the present study to find miRNAs involved in the different regulation of skeletal muscle development between the two breeds. A total of 320 known porcine miRNAs, 64 miRNAs corresponding to other mammals, and 224 potentially novel miRNAs were identified. Principal component analysis (PCA) and hierarchical cluster analysis (HCA) suggested that the factor “pig breed” affected the miRNA expression profiles to a lesser extent than the factor “developmental stage”. Fifty-seven miRNAs were differentially expressed (DE) between the neighbor developmental stages in TC and 45 such miRNAs were found in YK, 34 in common; there were more down-regulated stage-DE miRNAs than up-regulated. And a total of 23, 30, 12, 6, and 30 breed-DE miRNAs between TC and YK were identified at 40, 55, 63, 70, and 90 dpc, respectively, which were mainly involved in cellular protein modification process, protein transport, and metabolic process. As the only highly expressed breed-DE miRNA found in no less than four developmental stages, and also a stage-DE miRNA found both in TC and YK, miR-499-5p could bind the 3’-UTR of a myofibrillogenesis regulator, destrin/actin depolymerizing factor (DSTN), as validated in dual luciferase reporter assay. The results suggested that miR-499-5p possibly play a noteworthy role in the breed-distinctive porcine muscle fiber development associated with the regulation of DSTN.

Obese and lean pig breeds show obvious differences in adipose metabolism/fat deposition; however, the molecular mechanism underlying phenotype variation remains unknown. In order to understand it, we analyzed the differences of gene expression in backfat between Meishan (a typical Chinese indigenous obese breed) and Large White (a lean Western breed) pigs. Here, we cloned porcine β subunit of IDH3 (IDH3B) and 2447 bp 5'-flanking sequence of this gene, and determined the genomic structure. Porcine IDH3B contains three isoforms, IDH3B ( 1 ), IDH3B ( 2 ) and IDH3B ( 3 ). Real-time RT-PCR revealed that these three isoforms were prevalently up-regulated in backfat of western commercial pigs, Large White, Landrace and Duroc, compared with Chinese indigenous breeds, Meishan and Tongcheng pigs. A 304 bp insertion/deletion variant was found in the 5'-flanking region. Dual-luciferase reporter assays showed that in vitro the promoter of IDH3B gene with the insertion had higher luciferase activity as compared with the wild type. Three genotypes AA, AB and BB, due to this insertion, were detected, and the frequency of allele A was dominant in western commercial pigs, whereas allele B predominated in Chinese indigenous breeds. IDH3B mRNA expression in Meishan pigs was more abundant with genotype AA than with genotype AB or BB, as in Large White pigs. In addition, the polymorphism was detected in 317 pigs of a Large White × Meishan F2 resource population. Association analysis showed that pigs with genotype AA possessed higher backfat thickness at buttocks than those with genotype AB (P < 0.05) or BB. These data suggested that the 304 bp insertion mutation in promoter region increased the expression of porcine IDH3β transcripts and this mutation might be a candidate marker for marker assistant selection in swine breeding.

To detect the respiratory disease through pig cough sound in the early stage, a novel method based on Deep Neural Networks-Hidden Markov Model (DNN-HMM) was proposed to construct an acoustic model for continuous pig cough sound recognition. Noises in the continuous pig sounds were eliminated by the Wiener algorithm based on wavelet thresholding the multitaper spectrum, and the experimental corpus was obtained from the denoised continuous pig sounds. The 39-dimensional Mel Frequency Cepstral Coefficients (MFCC) extracted from the corpus were considered as feature vectors. Sounds in pig farms were divided into pig coughs, non-pig coughs, and silence segments. In the HMM, the number of hidden states of pig cough, non-pig cough and silence segments were 5, 5 and 3 respectively, and the observation states represented the feature vectors of the continuous pig sound signal. Based on experiments and empirical theory, the DNN model with 3 hidden layers and 100 nodes per layer was used to describe the correspondence between hidden states and observation serials. Through experiments, the context frames of DNN input were set to 5. Under the condition of optimal parameter setting, the traditional acoustic model Gaussian Mixture Model-Hidden Markov Model (GMM-HMM) was compared with DNN-HMM through a 5-fold cross-validation experiment. It was found that the Word Error Rate (WER) of each group in DNN-HMM was lower than that in GMM-HMM, and the average WER was 3.45% lower. At the same time, the best result of the DNN-HMM model was obtained with the lowest WER of 7.54%, and the average WER was 8.03%. The results showed that the method of DNN-HMM based acoustic model for continuous pig cough sound recognition was stable and reliable.

No abstract