The placental microbiome during nocardioform placentitis

Abstract

Historically, nocardioform placentitis (NP), has been associated with the presence of Amycolatopsis spp., Crossiella equi, and Streptomyces spp. However, these organisms have not been detected or isolated in all NP cases, suggesting alternative causative agents. Additionally, a previous attempt to experimentally induce NP using Crossiella equi was not successful (Canisso et al. EVJ. 2014; 47(1)). Recent studies suggest the existence of a placental microbiome and suggest that a shift in the relative abundance of community members may alter the outcome of gestation. Here, we investigated the placental microbial population in normal (n=11) and NP cases (n=31). Total RNA was extracted, ribosomal depletion was performed, and sequencing libraries were constructed followed by deep sequencing (>160 million reads/sample, 150bp PE). Host RNA was mapped to the equine reference genome (EquCab3.0) using STAR-2.5.3a, annotated and quantified by featureCount. Microbial RNA was separated from host RNA and mapped to the NCBI prokaryotic database (Kraken2). Additionally, these data were analysed using Phyloseq, vegan, amplicon, microbiome, and WGCNA for compositional analysis. The phylum Actinomycetota, containing the nocardioform bacteria, was more abundant in NP samples than in controls (p<0.01). Even though Amycolatopsis spp., Crossiella equi, and Streptomyces spp. had a high abundancy in NP cases, these microbes were also present in the healthy placentas at a 60-fold (p<0.01), 100-fold (p<0.05) and 2-fold (p>0.05) lower abundancy, respectively. Alpha diversity(inverse Simpson) was similar between the groups, but beta diversity was higher in NPs than in the controls (Bray Curtis and Aitchison distances, p<0.05), suggesting no difference in richness within groups, but a change in diversity between groups that may be linked to disease outcome. Microbial gene expression analysis revealed several genes solely expressed in NP cases, associated with microbial responses to hyperosmotic shocks, protein synthesis during stress, and ATP-dependent zinc metalloproteases and were associated with expression from Actinomycetota and Mycobacteriaceae. In the NP samples, 30 virulence factors (VF) were expressed, while no VF expression was detected in the controls. Interestingly, 70% (21/30) of VFs were expressed by Mycobacterium spp. An abundance of Amycolatopsis (r=0.56) and Mycobacterium (r=0.64) was positively correlated with genes associated with inflammation and immune cells in the placenta (p<0.01). Overall, these results suggest that the shift in the microbial composition during NP was associated with the disease. Additionally, the results suggest involvement of Mycobacterium spp. in NP cases. These bacteria have not been previously linked to NP, however, Mycobacterium avium has been previously linked to equine abortions with similar placental lesions as NP (Kinoshita et al. Vet Med Sci. 2021; 7:621-625). Funding: Special Research Fund (BOF) at UGent, The Foundation for The Horse, Grayson Jockey Club Foundation, Clay Endowment at UKY, and the Center for Equine Health at UCDavis.