Host and rumen microbiome contributions to feed efficiency traits in Holstein cows.

Of the newer developments are the availability of high-dimensional omics, such as the metagenome, metabolome and transcriptome, which provide the opportunity to incorporate such data, in addition to genomic data, to improve the prediction of feed efficiency. The inclusion of microbial data in genomic models enables unravelling the contribution of a particular host genome and its microbiome to the phenotype of interest. Thus, the heritability (h²) of a trait is decomposed into a direct genetic effect, quantified as direct heritability(h²_d), and an indirect genetic effect mediated by the microbiome. This means that the host genome can contribute to the variation of a phenotype in a direct way, or via modifying the microbial composition in the rumen. The microbiome effect on the phenotype is termed the microbiability (m²). Furthermore, there could be a joint action of the host genome and microbiome on the phenotype that is included in prediction models as an interaction effect, resulting in the phenotype being explained by the joint contribution of the genome, the microbiome, and their interaction. The combination is defined as holobiont and quantified as the holobiability (ho²).

With this as background, the objectives of the study cited were (1) to estimate relevant parameters such as h², h²_d, m², and ho² for three feed efficiency traits, namely DMI, milk energy, and RFI, and (2) to evaluate the predictive ability of different models including the host genome, the rumen microbiome, and their interaction in mid-lactation Holstein cows.

The data consisted of feed efficiency records, SNP genotype data, and 16S rRNA rumen microbial compositions from 448 mid-lactation Holstein cows from two research farms. Three kernel models were fit to each trait: one with only the genomic effect (model G), one with the genomic and microbiome effects (model GM), and one with the genomic, microbiome, and interaction effects (model GMO).

The model GMO, or holobiont model, showed the best goodness-of-fit. The h²_d estimates were always 10% to 15% lower than h² estimates for all traits, suggesting a mediated genetic effect through the rumen microbiome, and m² estimates were moderate for all traits, and up to 26% for milk energy. The ho² was greater than the sum of h²_d and m², suggesting that the genome-by-microbiome interaction had a sizable effect on feed efficiency. Kernel models fitting the rumen microbiome (i.e., models GM and GMO) showed larger predictive correlations and smaller prediction bias than the model G.

In conclusion: incorporating the rumen microbiome information in addition to genomic data allows for revealing the relative effects of the host genome and the microbiome on feed efficiency traits in dairy cattle. Rumen microbiome data can be used to estimate host direct and indirect genetic effects on feed efficiency. Indeed, the differences obtained between the h² and the h²_d strongly suggest that the microbiome mediates part of the host genetic effect. The holobiont model, which incorporates the host genome-by-microbiome interaction, provides further insights into the biological mechanisms underlying dairy cow feed efficiency.