Relation of rumen microbiome and blood transcriptome with the genetic merit in Italian Simmental and Italian Holstein cows

Phenotypic selection of livestock over last 50 years, in particular of dairy cows, has been effective in increase productions and improve animal nutrients utilization. The effects of the phenotypic improvement, in particular on the rumen microbial profile and on the gene expression remains largely unknown. In this study a picture of the effects of the estimated breeding value for a productive trait on cows rumen metagenome and blood transcriptome of healthy Italian Holstein and Italian Simmental lactating cows is presented. Preliminarily, two Italian Holstein and two Italian Simmental commercial farms were selected in the Friuli Venezia Giulia region for having similar management and feeding conditions. Within each farm healthy cows, ranging from two to five calvings and from 70 to 180 days in milk were selected. Within each farm cows were then grouped in three classes of Estimated Breeding Value for milk protein (EBVp). The EBV classes were defined within each breed and farm by sorting the EBV values into quartiles, where the first (Q1) and the third (Q3) quartiles correspond to the ±25% extremes EBV values versus the mean of the distribution. Three quartiles (Q1, Q2 and Q3) resulted in each farm. Milk yields and compositions form 135 cows were obtained from the official recording service and the same day blood, urine and rumen content were sampled during the morning meal and subjected to biochemical analysis. These data were used to asses the relationship between EBVp and productive traits and metabolic condition. Based on these results, two subsets of cows were chosen to investigate rumen microbiome and blood gene expression. For rumen microbiome composition two EBVp minus and two EBVp plus variant cows for each farm and each breed (overall 16 cows) were selected. The massive parallel High Throughput Sequencing of the whole rumen contents and the annotation procedure of FASTQ base called files allowed to describe the classify the taxonomy of the bacterial and archaeal communities. Statistical analysis applied to the abundance profiles at the different taxonomic levels indicated a significant effect of breed and EBVp. Within Bacteria domain Italian Simmental cows had significant (P<0.05) higher presence of bacteria belonging to Actinobacteria and lower bacteria of the Clostridia class in comparison to Italian Holstein cows. Rumen contents of Italian Simmental cows were also significantly richer (P<0.001) of archaeal Methanobacteria than Italian Holstein cows. Overall, the EBVp resulted positively related with Bacteroidetes (P<0.01) and negatively with methanogen Euryarchaeota phyla. The investigation of gene expression profile at the blood level was performed in eight cows for each EBVp quartile (eight Q1, eight Q2 and eight Q3) for one Italian Simmental and one Italian Holstein farm (overall 24 Italian Simmental and 24 Italian Holstein cows). For the dual purpose Italian Simmental cows the most relevant affected pathways were related to glutathione metabolism and intestinal IgA production. In the Q3 group GSS and TGF1B genes were significantly (P<0.05) over expressed in comparison to the Q1 group. For the Italian Holstein cows the most relevant affected pathways were related to T cell receptor and adipocytokine signalling pathways. In the Q3 group CD3G, CD4 and ACSL1 genes were significantly (P<0.05) over expressed in comparison to the Q1 group. At the best of our knowledge, this approaches has never been used and a direct comparison with published studies is not straight full, but some preliminary considerations from the data gathered from our study can be drawn. Despite of the limited number of cows included in the study, the rumen metagenomic data indicate that microbial population is significantly affected by genetic merit other than breed. The analysis of blood gene expression showed that also specific metabolic and immunitary pathways are directly related to the genetic merit, and can be considered a signature of the quantitative selection. Starting from this preliminary observations, the approaches are promising since they can identified phenotypic biomarkers that can be used for genomic selection, based on rumen microbial profile and differentially expressed genes more than on DNA variants.