Genomic and transcriptomic investigations into the feed efficiency phenotype of beef cattle

Abstract

Improving the feed efficiency of cattle is a method to increase profits while simultaneously reducing the environmental impact of beef production. Residual feed intake (RFI) is a measure of feed efficiency, calculated as the difference between actual and predicted feed intake. Low-RFI (feed efficient) cattle consume less feed than their high-RFI (inefficient) counterparts while maintaining their growth rate and emitting less methane. Therefore, incorporation of RFI into breeding programmes represents an opportunity to improve profitability while reducing the environmental impact of beef production. However, RFI is a difficult trait to measure, requiring expensive and time consuming feeding trials. Identification of biomarkers for RFI would enable genomic-assisted selection for this trait, circumventing the need for continual feeding trials. There is considerable variability associated with RFI whereby cattle have been observed to re-rank in terms of this trait when offered varying diets and different breeds display inherent variation in RFI. Consequently, robust biomarkers for RFI must be identified which are applicable across breed and diet. The difficulty in identifying robust biomarkers for this trait has been an impediment to the adoption of RFI in genomic assisted breeding programmes. The aims of this thesis were: (i) To identify SNPs associated with RFI in a multi-breed and crossbred reference population of Irish beef cattle for inclusion in the Irish genomic assisted breeding programme. (ii) To identify differentially expressed genes associated with RFI-divergence in two breeds of steers offered three diets throughout their lifetime, which may be candidate genes for interrogation for the discovery of biomarkers for RFI. (iii) To identify key regulatory genes and biological processes associated with variation in RFI based on gene expression data across two breeds and three dietary phases, which may be candidates for further work while enhancing our understanding of the biology underlying RFI-divergence.

In my first study, seven, fourteen and three SNPs associated with RFI, average daily gain and feed intake, respectively, were identified in a multi-breed reference population of Irish beef cattle (n = 1,492). To investigate the effects of associated variants on nearby genes, expression quantitative trait loci (eQTL) analysis was carried out. One eQTL, between rs43555985 and GFRA2, was identified for RFI. rs43555985 was a SNP associated with RFI following meta-analysis. xi

Following this, RNA-Seq analysis was performed to investigate the hepatic transcriptome of Charolais and Holstein-Friesian steers divergent for RFI. These steers were offered three differing diets throughout their lifetime. A total of 355 differentially expressed genes were identified across all diet-breed combinations. Three genes, GADD45G, HP and MID1IP1, were differentially expressed across two dietary phases for the Charolais steers. No gene was differentially expressed across all three dietary phases for either breed, however several physiological processes such as immune function and lipid metabolism, were enriched across all diet and breed combinations.

In order to gain insight into gene networks and key regulatory genes implicated in RFI-divergence, weighted gene co-expression network analysis (WGCNA) of the RNA-Seq data was carried out. WGCNA allows identification of modules of genes associated with RFI. These modules are genes which have similar expression profiles and may work in unison to affect the phenotype. Similarly, WGCNA can identify hub genes which are postulated to be master regulators of gene expression within their modules. These genes and modules may be candidates for further study in order to identify SNPs associated with RFI across diet and breed. This analysis identified a total of ten modules significantly associated with RFI. Within these modules, 349 hub genes were identified. Thirty-seven of these have previously been associated with feed efficiency in livestock while ten hub genes were identified as differentially expressed in the earlier RNA-Seq analysis. Hub genes were found to play roles in protein turnover and mitochondrial efficiency amongst other physiological processes.

Overall, these genomic and transcriptomic analyses have provided insight into the biology of feed efficiency and can form the basis of biomarker discovery for genomic selection for RFI. Results reported within this thesis support previous findings with regards to the role of physiological processes such as lipid metabolism and protein turnover in RFI-divergence. This thesis also reports novel findings such as the first eQTL identified for RFI in beef cattle. These results together have utility to inform future genomic assisted breeding programmes.