Deliverable D2.6 Predictive biomarkers as traits for digestive efficiency in pigs
Objectives
The aim of Feed-a-Gene is to improve and adapt the different components of monogastric (i.e., pigs, poultry, and rabbits) livestock production systems to enhance their overall efficiency and to reduce their environmental impact. One instrument to reach this overall aim is to develop novel indicators of feed efficiency.
Feed efficiency has two components, relating to the digestive efficiency on the one hand and to the post-absorptive efficiency on the other hand. Therefore, new data referring to digestive efficiency and data referring to the post-absorptive efficiency (and to the overall feed efficiency) are needed to improve feed efficiency in animals. This deliverable aims to report findings on predictive biomarkers as new traits for digestive and overall feed efficiency in pigs.
Rationale
Large variations in growth responses and nitrogen efficiency have been observed in populations of pigs. This issue has been addressed using different approaches in the Feed-a- Gene project. Differences in piglet birth weight may cause variation in the digestibility of protein and amino acids. Faecal protein digestibility and N-retention as affected by birth weight was evaluated in an experiment performed at DLO (reported in Deliverable D2.4). However, N-balance studies are time-consuming and involve a limited number of animals. Indirect indicators such as biomarkers in blood or urine would be useful for the in vivo evaluation of differences in protein and amino acid metabolism and amino acid requirements. Another approach is to study feed efficiency by selecting for feed efficiency and study differences between the most and the least efficient animals. This is traditionally done by measuring feed intake and body weight gain in individual animals. However, knowledge of differences and changes in the molecular pathways contributing to digestive or post-absorptive nutrient efficiency through the identification of biomarkers would be of value.
Pigs divergently selected for residual feed intake (RFI), also called net efficiency, can be used as an animal model to obtain large variation in traits related to feed efficiency (e.g., conversion ratio (FCR)) and different tissues and fluids can be collected to propose biomarkers of overall feed efficiency. Because of the genetic background of the animals, biomarkers can be used as (early) predictors of this trait in the population to increase genetic progress (e.g., in connection with WP5).
The measurement of small molecules involved in or generated by metabolic processes in tissues and body fluids (e.g., blood and urine) is a feasible tool to identify possible biomarkers for specific metabolic responses and traits related to feed efficiency. Molecules include mRNA produced by gene transcription, proteins produced by mRNA translation, and metabolites. Metabolites measured with untargeted metabolomics can, because of the sensitivity of the technique, detect subtle alterations in biological pathways and provide insight in the mechanisms that underlie different physiological conditions. Liquid chromatography-mass spectrometry (LC-MS) based untargeted metabolomics analysis was applied to i) blood and urine samples from the study relating to N-utilization (i.e., birth weight x dietary protein supply or genetic capacity to deposit protein x dietary protein supply), ii) to blood samples from a study with genetic lines divergently selected for residual feed intake (RFI) and plasma cortisol concentration, and iii) to blood samples from pigs with feed efficiency extremes.
Taking advantage of large sets of transcriptomics data (mRNA concentrations) that have been acquired in i) muscle, one of the main tissues affected by feed efficiency differences, or ii) blood (a fluid that can be collected at repeated times during the growth period, large sets of molecular predictors of feed efficiency traits, such as RFI, gain-to-feed ratio (overall feed efficiency), and gain-to-energy feed ratio (energy efficiency) were also identified using machine learning methods. The assumption that it is possible to identify molecules in tissues and fluids that are able to predict feed efficiency traits was thus validated. The variety of biological functions represented by the genes and metabolites included in the predictive models confirmed the integrative and complex nature of feed efficiency in growing pigs. The lists of biomarkers are transferred to WP5 as new possible traits to be included in next selection procedures.
