ABSTRACT: Background. Steer progeny suckled by cows fed a dried distillers grains and solubles (DDGS) diet the first three months of lactation were heavier during feedlot finishing and had significantly lower marbling and larger longissumus muscles than steer suckled by cows fed a control diet (CON). These differences were profound in that progeny were managed and fed identically from weaning until finishing, and suggested that the developmental program of muscle composition was established during the suckling period. Purpose. Transcriptomes of longissimus muscle were measured using next generation sequencing to investigate whether there were any developmental clues to the differences in marbling scores and muscle content between steers suckled by DDGS (n=5) versus control (CON; n=5) diet fed cows during lactation. Methods. Multiparous Angus × Simmental cows with male progeny were fed one of two diets supplemented with either DDGS or soybean meal (CON), from calving until day 129 postpartum (PP). After conclusion of the treatments at 129 days postpartum, cow–calf pairs were comingled and managed as one group until weaning at 219 days postpartum. Steers were then transitioned to a common diet. Longissimus muscle (LM) biopsies were obtained from steers suckled by cows fed DDGS (n=5) and CON (n=5) 10 d prior to slaughter. RNA was extracted from LM. Total RNA isolated and Library Prep Kit was used to generate library for measuring messenger RNA molecules. Library was sequenced on an Illumina HiSeq2500 platform. FastQC (v 0.10.1) was used to assess sequence quality, and quality trimming was done using FASTX toolkit (v 0.0.13). Bases with less than Phred33 score of 30 were removed and resulted in reads of at least 50 bases. Reads were mapped against the bowtie2-indexed Bos taurus genome using Tophat (v 2.0.9) with default parameters, and the genome annotation file (.gtf) from Ensembl was used to associate the genome with annotated features. Raw read counts of each sample for each gene feature were generated with HTSeq (v 0.5.3p7) using Tophat output and Bos taurus annotations. Counts were merged using custom perl scripts and used for downstream differential gene expression (DE) analysis. DE pairwise analysis between samples CON and DDGS was done with ‘R’ (Version 2.15.2). Counts matrix, library sizes, and experimental design were combined to create an object using edgeR (v 3.0.3) package. Results. Number of reads per sample ranged from 79,530,352 to 58,600,824. After trimming there were 77,569,706 to 57,307,614 per sample. The percent of reads that mapped to bovine genome per sample ranged from 77.3 to 92.3%, and averaged 86.3% across all samples. Number of reads per gene ranged from one to over two-hundred ninety thousand counts per million (CPM) reads. There were 809 genes differentially expressed (P-adj<0.1) between CON and DDGS muscle. Of these 637 were upregulated and 172 downregulated in DDGS relative to CON. Overall the DDGS versus CON muscle transcriptomic signature was pro-myogenic and anti-adipogenic. In particular, myokines/satellite cell maintenance factors were found among upregulated (LIF, CNTF, FGFB1, EPHB1) genes. The anti-adipogenic signature was typified by the upregulation of anti-inflammatory cytokines and receptors (IL1RAP, IL1RL2, IL13RA2, IL1F10,), and downregulation of expression of inflammation/inflammatory cytokines and receptor (TNF, IL6R CXCL9), which suggests a selection of differentiation pathways away from adipogenic line. While the upregulation of TGFB, SPP1 and INHBA supports selection of fibroblast lineage of cells. Conclusion. The lactation phase of production can effect meat quality by affecting transcriptional signatures that favor myogenesis and depress inflammation.
