# Plasma biomarkers for residual feed intake prediction in beef bulls

## TL;DR

This study identifies blood compounds like choline and specific fatty acids that could serve as biomarkers for predicting feed efficiency in beef cattle.

## Contribution

The study discovers potential plasma biomarkers for residual feed intake in beef bulls, offering new insights for breeding programs.

## Key findings

- Choline was identified as a biomarker for RFI at day 0 with an AUC > 0.7 and p-value < 0.05.
- Cer(d18:1/23:0) and TG(18:1_30:0) were identified as biomarkers for RFI at day 56.
- The primary bile acid biosynthesis and sphingomyelin metabolism pathways were enriched in low RFI bulls.

## Abstract

Residual feed intake (RFI) is a measure of feed efficiency (FE) independent of growth and body weight (BW), calculated as the difference between actual and expected feed intake (FI) based on mean metabolic weight (MW) and weight gain. We hypothesized that bulls with contrasting RFI differ in plasma concentration of different compounds. RFI was evaluated in 302 bulls from 3 different ranches. After adaptation, bulls consumed the same diet for 56 d, individual FI was daily recorded, BW was measured every 2 weeks, and blood samples were taken at d 0 and 56. The bulls were ranked as low RFI (LRFI) and high RFI (HRFI), and the top and bottom 30 were used for metabolomic, hormone, and isotope analysis. Multivariate and pathway analysis were conducted with MetaboAnalyst, and univariate analysis was conducted with SAS using Mixed procedures. Additionally, for Biomarker analysis, Receiver Operating Characteristic (ROC) curves were constructed with MetaboAnalyst. Sparse Partial Least Squares–Discriminant Analysis (sPLSDA) showed partial cluster separation evidenced by acceptable R2 values. The model showed poor predictive power, reflected by low Q2 values. For LRFI animals (more feed efficient), the most abundant metabolites (P = 0.05) at d 0 were Cer(d18:1/24:1), TG(20:2_32:1), SM (OH) C22:1, SM (OH) C22:2, SM C24:0, TG(20:2_34:3), GCA, Choline, TG(20:2_34:4), and Hex2Cer(d18:1/14:0), while at d 56 were Cer(d18:1/18:0), Cer(d18:1/23:0), TG(14:0_34:2), C16:2, HexCer(d18:1/20:0), C16:1, lysoPCaC16:0, TG(18:3_38:5), C9, SM C24:1, SM C16:0, Carnosine, Cer(d18:2/12:0), SM C18:0, and Cer(d18:1/24:0). Primary bile acid biosynthesis pathway was enriched at d 0 (P = 0.008) and sphingomyelin metabolism at d 56 in LRFI (P = 0.041). Compounds that were identified as variable importance of projection (VIP) and were also statistically different in the univariate analysis, were used to construct ROC curves to identify their potential as biomarker for RFI. With an Area Under the Curve (AUC) value > 0.7 and p-value < 0.05 as the criteria for diagnostic potential, choline was identified as biomarker of RFI at d 0, and Cer(d18:1/23:0) and TG(18:1_30:0) were identified as biomarkers of RFI at d 56. Although multivariate model showed a poor predictive value, further exploration of individual metabolites could provide insights into the mechanisms contributing to FE differences.

Blood compounds such as choline and certain fatty acids may help explain why some cattle are more feed efficient than others, offering potential biomarkers to guide future selection and breeding programs.