Dynamics of the milk microbial community during subacute ruminal acidosis with or without intramammary lipopolysaccharide challenge in dairy cows

Background: The lipopolysaccharides (LPS) released by pathogenic Gram-negative bacteria play a key role in the symptom's development during clinical mastitis. Additionally, subacute ruminal acidosis (SARA) is known to induce rumen dysbiosis in dairy cows, leading to ruminal LPS release and translocation of free LPS, triggering both local and systemic immune responses. Underlying the theory of the gut-mammary gland axis and the environmental influence of faecal microbes, we aimed to investigate to which extent a high-grain induced SARA affects the milk microbial community of dairy cows. Furthermore, we sought to elucidate the impact of intramammarily injected LPS, imitating clinical mastitis without the presence of the causing agent, on the milk microbes. Results: During the experiment, 70.1% of the taxa increased in absolute abundance in the milk of the control group (CON, 40% grain), whereas 65.3% decreased in the SARA group (60% grain; P<0.001). Among the 40 most abundant taxa, representing 90% of the dataset, all but one increased numerically in the CON group. Only a few, such as Staphylococcus, Streptococcus, Bifidobacterium, also increased numerically in the milk of SARA cows. After the intramammary challenge, the SARA cows that received NaCl as placebo (SARA_PLA) did not change this trend, with again 62.4% of the taxa decreasing. The injection of LPS inversed the dietary trends, so that in SARA_LPS cows 79.3% of the taxa increased, while in CON_LPS cows 85.5% of the taxa dropped in absolute abundance (P<0.001). In the 40 most abundant taxa, 18 showed changes by trend (P<0.1) as for example Staphylococcus and Corynebacterium increased in the SARA_LPS group (P<0.1) but decreased in CON_LPS (P>0.1), while Streptococcus decreased in all groups (P<0.05 for CON_LPS). The microbial diversity increased with increasing bacterial cell equivalents, and 22.1% of the taxa correlated positively with Milk-Amyloid-A (MAA), which clearly increased after LPS injection, 21.7% correlated positively with lactose, and 13.4% positively with urea in the milk, but there was no strong correlation with SCC. Conclusion: The increase or decrease patterns of the taxa serves as indicator that the microbial, metabolic, and immunological consequences of diet and LPS injection influence the bacterial community in the milk of dairy cows. Interestingly, the association of the taxa was stronger with MAA than with SCC, questioning the use of SCC to predict microbial occurrence in milk. A roughage-rich diet favours a more abundant microbial community in the udder, in contrast to SARA inducing high-grain diets. Administration of intramammary LPS decreases the absolute abundance even in the CON group, but the combination of SARA and LPS increases abundance. This reversed phenomenon might be interpreted as a host response due to an altered local immune response in the udder caused by the systemic reaction during SARA.