Genome-wide phylodynamic approach reveals the epidemic dynamics of the main Mycoplasma bovis subtype circulating in France

Abstract

Mycoplasma bovis is a major aetiological agent of bovine respiratory disease worldwide. Genome-based analyses are increasingly being used to monitor the genetic diversity and global distribution of M. bovis, complementing existing subtyping schemes based on locus sequencing. However, these analyses have so far provided limited information on the spatiotemporal and population dynamics of circulating subtypes. Here we applied a genome-wide phylodynamic approach to explore the epidemic dynamics of 88 French M. bovis strains collected between 2000 and 2019 in France and belonging to the currently dominant polC subtype 2 (st2). A strong molecular clock signal detected in the genomic data enabled robust phylodynamic inferences, which estimated that the M. bovis st2 population in France is composed of two lineages that successively emerged from independent introductions of international strains. The first lineage appeared around 2000 and supplanted the previously established antimicrobial-susceptible polC subtype 1. The second lineage, which is likely more transmissible, progressively replaced the first M. bovis st2 lineage population from 2005 onward and became predominant after 2010. Analyses also showed a brief decline in this second M. bovis st2 lineage population in around 2011, possibly due to the challenge from the concurrent emergence of M. bovis polC subtype 3 in France. Finally, we identified non-synonymous mutations in genes associated with lineages, which raises prospects for identifying new surveillance molecular markers. A genome-wide phylodynamic approach provides valuable resources for monitoring the evolution and epidemic dynamics of circulating M. bovis subtypes, and may prove critical for developing more effective surveillance systems and disease control strategies.

Keywords: bacteria; cattle; fitness; lineage replacement; respiratory disease; surveillance.

Fig. 1.

Geographical and temporal distribution of Mycoplasma bovis samples from this study and cattle density. (a) Map of France showing department sub-divisions. The colour gradient indicates cattle density km⁻² extracted from Gilbert et al. [61]. (b, c) Sampling location and temporal distribution of M. bovis isolates from the two most cattle-populous regions of France, i.e. North-West and Centre.

Fig. 2.

Fitness-annotated time-scaled phylogeny of Mycoplasma bovis subtype 2 in France. A maximum-credibility clade (MCC) estimated from a Bayesian molecular clock analysis using the core-genome alignment of 88 French M. bovis subtype 2 isolates (87 genomes generated in this study and the L15762 reference genome). The phylogeny is temporally rooted. Branch colours indicate relative fitness between lineages in the phylogeny. Black circles at internal nodes denote posterior clade probabilities >0.75. Black arrows indicate internal branch links with samples outside France (Fig. S2). The top-left panel shows the regression analysis between the sampling dates and the root-to-tip distances in the maximum-likelihood phylogenetic tree of 88 French subtype 2 M. bovis isolates (see Fig. S3).

Fig. 3.

Population dynamics of Mycoplasma bovis subtype 2 in France. (a) Effective population size (N _e) through time, estimated using a Bayesian SkyGrid approach. The black line and grey hatched area represent the posterior median estimate of N _e and its 95 % posterior highest credible density intervals, respectively. Blue and red areas indicate the relative N _e proportion of lineages A and B, estimated using lineage-through-time analysis. (b) Effective reproductive number (R _e) through time, estimated using a birth–death skyline approach. The blue and red lines and shadings represent posterior median R _e estimates of lineages A and B and their 75 % posterior highest credible density intervals, respectively. Black arrows indicate the first detection of M. bovis subtypes 2 (in 2000) and 3 (in 2011) in France through the Vigimyc epidemiological surveillance network.

Fig. 4.

Main non-synonymous SNP changes identified among the French Mycoplasma bovis subtype 2 isolates. Left panel shows the French M. bovis subtype 2 phylogeny with lineages A and B ancestral node-annotated. Right panel shows non-synonymous SNP changes comparatively to the ancestral isolate 2357, with corresponding gene names given at the bottom of the panel. Superscript letters associated with gene names correspond to several mutations at different site positions: ‘a’ denotes mutations in the four site positions identified in topA, ‘b’ denotes the three site positions identified in mutM and ‘c’ and ‘d‘ denote site positions 1550 and 1852 identified in MIB, respectively. White blocks indicate gene absences (or possibly non-recovery in sequencing) in certain lineages due to gene deletion in genome isolates. Non-synonymous SNPs that could have an impact on gene functions are detailed in Table S2.