Brucella ceti and Brucella pinnipedialis genome characterization unveils genetic features that highlight their zoonotic potential

Abstract

The Gram-negative bacteria Brucella ceti and Brucella pinnipedialis circulate in marine environments primarily infecting marine mammals, where they cause an often-fatal disease named brucellosis. The increase of brucellosis among several species of cetaceans and pinnipeds, together with the report of sporadic human infections, raises concerns about the zoonotic potential of these pathogens on a large scale and may pose a threat to coastal communities worldwide. Therefore, the characterization of the B. ceti and B. pinnipedialis genetic features is a priority to better understand the pathological factors that may impact global health. Moreover, an in-depth functional analysis of the B. ceti and B. pinnipedialis genome in the context of virulence and pathogenesis was not undertaken so far. Within this picture, here we present the comparative whole-genome characterization of all B. ceti and B. pinnipedialis genomes available in public resources, uncovering a collection of genetic tools possessed by these aquatic bacterial species compared to their zoonotic terrestrial relatives. We show that B. ceti and B. pinnipedialis genomes display a wide host-range infection capability and a polyphyletic phylogeny within the genus, showing a genomic structure that fits the canonical definition of closeness. Functional genome annotation led to identifying genes related to several pathways involved in mechanisms of infection, others conferring pan-susceptibility to antimicrobials and a set of virulence genes that highlight the similarity of B. ceti and B. pinnipedialis genotypes to those of Brucella spp. displaying human-infecting phenotypes.

OrthoPhyl – Streamlining large scale, orthology-based phylogenomic studies of bacteria at broad evolutionary scales

“…The clade comprising the classical Brucella agrees with published phylogenies (Figure 4B) (Suárez-Esquivel et al ., 2020; Ashford et al ., 2020). Brucella vulpis at the base , B. pinnipedialis arising from within B. ceti (rendering B. ceti paraphyletic) (Orsini et al ., 2022), B. canis arising from within B. suis (rendering B. suis paraphyletic), and B. melitensis and B. abortus being sister groups. Interestingly, several Brucella assemblies seem to be mislabeled ( Supplemental Figure 2 , grey tip labels ).…”

“…Whether these placements are accurate or not remains to be determined, however, B. melitensis and B. abortus are very closely related (precluding saturation) and do not show signs of significant interspecies recombination (Vishnu et al ., 2015; Suárez-Esquivel et al ., 2020). The classical Brucella have very short branches at the crown ( B. microti , B.neotomae , B.ovis , B. suis / canis , B. abortus / melitensis and B. pinnipedialis / ceti ), indicating very rapid divergence times for these clades, as seen elsewhere (Ashford et al ., 2020; Suárez-Esquivel et al ., 2020; Orsini et al ., 2022). While the topology of this tree is congruent with several published trees, more analyses are required to create the most robust tree possible (see Discussion).…”

“…For Rickettsiales assemblies, we again used gather_genomes.sh, this time with NCBI taxon number 766. To illustrate a more challenging scenario, filtering of Rickettsiales assemblies was less stringent with cut-offs at > 95% completion, < 0.2% duplication, and < 1.5% contamination, again using CheckM (Parks et al, 2015). This resulted in 1202 assemblies.…”

“…The distribution of taxon number represented in each alignment shows that most orthogroup alignments have all taxa represented, with approximately 1400 OGs having 680-689 taxa (Figure 2 The clade comprising the classical Brucella strongly agrees with published phylogenies (Suárez-Esquivel et al, 2020;Ashford et al, 2020). Brucella vulpis at the base, B. pinnipedialis arising from within B. ceti (rendering B. ceti paraphyletic) (Orsini et al, 2022)…”

“…illustrate a more challenging scenario, filtering of Rickettsiales assemblies was less stringent with cutoffs at > 95% completion, < 0.2% duplication, and < 1.5% contamination, again using CheckM(Parks et al, 2015) and our independent calculation of duplication (Supplemental methods "Gathering and Filtering Assemblies". This resulted in 1201 assemblies.…”