EPIDEMIOLOGICAL ANALYSIS AND CHARACTERIZATION OF MICROBIOME ASSOCIATED WITH BOVINE MASTITIS

Abstract

Mastitis, the inflammation of the mammary gland and/or quarters is one of the most prevalent diseases in the dairy industry worldwide with economic significance. The milk of the lactating cows presents a complex ecosystem of interconnected microbial communities which can impose a significant influence on the pathophysiology of mastitis. Bovine mastitis is caused by a wide range of apparently resident microbes including bacteria, viruses, and archaea. Our hypothesis- the possible dynamic shifts of microbiome compositions with the progress of different pathophysiological states of mastitis are determined by its favoring genomic potentials. In order to address the hypothesis, whole metagenome sequencing (WMS) is carried out to compare the microbiomes of clinical mastitis (CM = 5), recurrent clinical mastitis (RCM = 6), subclinical mastitis (SCM = 4), and healthy (H = 5) milk samples. The metagenomics data analyzed to characterize the microbiomes associated with bovine mastitis and their cross-talk in respect to disease progression, virulence factorsassociated genes (VFGs), antibiotic resistance genes (AGRs), resistomes, and metabolic functional potentials. The WMS generated 416.64 million reads (with an average of 20.83 million reads/sample) from the samples of four metagenomes. PathoScope (PS) and MGRAST (MR) analyses mapped the WMS data to 442 bacterial, 58 archaeal, and 48 viral genomes with distinct variation in microbiome composition, and abundances across these metagenomes (CM>H>RCM>SCM). Significant variations observed in species richness (i.e., alpha-diversity; P = 0.003, Kruskal–Wallis test), and microbial community structure (i.e., beta-diversity; P = 0.001, Kruskal–Wallis test) among the samples of four metagenomes. These diversities differ across the CM, RCM, SCM and H metagenomes, and numerically dominated by phyla Proteobacteria, Firmicutes, Actinobacteria and Bacteroidetes. Through PS analysis, we detected 385, 65, 80 and 144 bacterial strains in CM, RCM, SCM, and H milk, respectively, with an inclusion of 67.19% previously unreported opportunistic strains in mastitis milk metagenomes. The MR pipeline detected 56, 13, 9 and 46 archaeal, and 40, 24, 11 and 37 viral genera in CM, RCM, SCM and H-milk metagenomes, respectively. Of these, 12.06% archaeal and 20.83% viral genera found to be shared in CM, RCM, SCM, and H metagenomes. Furthermore, we identified 333, 304, 183 and 50 VFGs, and 48, 31, 11 and 6 AGRs in CM, RCM, SCM, and H-microbiomes, respectively, showing a significant correlation between the relative abundances of VFGs (P = 0.001, Pearson test), ARGs (P = 0.0001, Pearson test), and associated bacterial taxa.

We also detected correlated variations in the presence and abundance of several metabolic functional genes related to bacterial colonization, proliferation, chemotaxis, motility and invasion to mammary epithelial cells (P = 0.001, Kruskal–Wallis test) across these metagenomes. Furthermore, genes coding for oxidative stress, immune-diseases, twocomponent regulatory systems, regulation and cell signaling, virulence and pathogenicity, phage integration and excision, biofilm-formation, and quorum-sensing also varied significantly (P = 0.001, Kruskal–Wallis test) in different episodes of mastitis. In addition, we found a significant association between the resistomes and microbiome composition of the CM milk with no apparent association with cattle breeds, despite significant differences in microbiome diversity among the breeds. The in vitro investigation revealed 76.2% of six selected CM pathogens that are considered ―biofilm formers‖, and found to be highly resistant to tetracycline, doxycycline, nalidixic acid, ampicillin, chloramphenicol while being sensitive to five heavy metals (Cr, Co, Ni, Cu, Zn) at varying concentrations. In a separate experiment, fecal microbiota transplantation (FMT) from mastitic cows to pregnant mice resulted in visible mastitis symptoms in mice mammary glands as validated through histopathologic changes in mammary glands and gut tissues, and microbiome characterization. We also observed significant (P = 0.012, Kruskal–Wallis test) microbiome dysbiosis between mastitis and healthy mice fecal samples after FMT. Therefore, profiling the dynamics of microbiomes in different states of mastitis, concurrent VFGs, ARGs, and genomic functional correlations contribute to developing microbiome-based diagnostics and therapeutics for bovine mastitis, and carries significant implications on curtailing the economic fallout from this disease. Furthermore, the cows-to-mouse FMT might shed new light on rational selection of animal models to study and interpret host-tropism to interrogate role of microbiota in this important economic disease of dairy industries.