Identification of early blood differentially expressed genes between two pig lines divergently selected for feed efficiency: Potential biomarkers for feed efficiency

Abstract Text:

Improving feed efficiency (FE) of domestic animals is of economic, social and environmental significance. However, measuring FE is expensive and time-consuming. Thus, predictive biomarkers for FE are in great need. In this study, we attempted to identify early stage differentially expressed genes (DEGs) between two Yorkshire lines of pigs divergently selected for FE based on residual feed intake (RFI), with an ultimate goal of developing early biomarkers for FE. Blood samples were collected at 35 ~ 42 days of age from Generation 9 piglets of the two lines, which were later randomly assigned to two diets of different fiber and energy content, and each pig’s RFI was measured. We also measured the complete blood count (CBC) of all blood samples. Blood samples of 8 animals of extremely low RFI for the low RFI line and extremely high RFI for the high RFI line on each diet were selected for RNA-sequencing, with one sample dropped due to low RNA quality. Globin transcripts were depleted by an RNase H-mediated method before RNA-seq library construction. After backward feature selection, the model for read counts included line, CBC covariates (neutrophil, lymphocyte, monocyte and basophil) and technical variables (processing batch, RNA concentration before globin depletion, RNA integrity number (RIN) after globin depletion). Using the QuasiSeq package, we found 454 DEGs (q ≤ 0.05) between the two lines, with 50 DEGs with fold change ≥ 2. With the high RFI line as reference, gene ontology-biological process (GO-BP) terms including small molecule biosynthetic process, lipid biosynthetic process, steroid biosynthetic process and positive regulation of immune system process, were overrepresented among up-regulated genes, while signaling, RNA biosynthesis process and cellular response to stimulus were overrepresented among down-regulated genes (q ≤ 0.15). Weighted gene coexpression network analysis showed that the eigengene expression in 4 modules was highly associated with lines (p < 3.9E-07, R² > 0.59) and GO-BP terms of lipid metabolic process, lipid biosynthetic process, steroid biosynthetic process and response to stimulus were overrepresented among the top 2 modules. We will use qRT-PCR to validate the DEGs of ≥ 2-fold change in original samples and new representative samples of the two lines. Taken together, this study suggests that the blood expression level of genes involved in several biosynthetic processes might be related to the difference in FE of the two lines. In the future, the identified DEGs and genes highly associated with line will be further validated and selected to build a biomarker-based classifier for FE. Support from USDA-NIFA-AFRI- 2011-68004-30336 is acknowledged.

Keywords:

 feed efficiency, pig, differentially expressed genes