Genomic epidemiology of Mycobacterium avium subsp. paratuberculosis isolates from Canadian dairy herds provides evidence for multiple infection events

Abstract

Mycobacterium avium subsp. paratuberculosis (MAP) is the pathogen responsible for paratuberculosis or Johne's Disease (JD) in ruminants, which is responsible for substantial economic losses worldwide. MAP transmission primarily occurs through the fecal-oral route, and the introduction of an MAP infected animal into a herd is an important transmission route. In the current study, we characterized MAP isolates from 67 cows identified in 20 herds from the provinces of Quebec and Ontario, Canada. Whole genome sequencing (WGS) was performed and an average genome coverage (relative to K-10) of ∼14.9 fold was achieved. The total number of SNPs present in each isolate varied from 51 to 132 and differed significantly between herds. Isolates with the highest genetic variability were generally present in herds from Quebec. The isolates were broadly separated into two main clades and this distinction was not influenced by the province from which they originated. Analysis of 8 MIRU-VNTR loci and 11 SSR loci was performed on the 67 isolates from the 20 dairy herds and publicly available references, notably major genetic lineages and six isolates from the province of Newfoundland and Labrador. All 67 field isolates were phylogenetically classified as Type II (C-type) and according to MIRU-VNTR, the predominant type was INMV 2 (76.1%) among four distinct patterns. Multilocus SSR typing identified 49 distinct INMV SSR patterns. The discriminatory index of the multilocus SSR typing was 0.9846, which was much higher than MIRU-VNTR typing (0.3740). Although multilocus SSR analysis provides good discriminatory power, the resolution was not informative enough to determine inter-herd transmission. In select cases, SNP-based analysis was the only approach able to document disease transmission between herds, further validated by animal movement data. The presence of SNPs in several virulence genes, notably for PE, PPE, mce and mmpL, is expected to explain differential antigenic or pathogenetic host responses. SNP-based studies will provide insight into how MAP genetic variation may impact host-pathogen interactions. Our study highlights the informative power of WGS which is now recommended for epidemiological studies and to document mixed genotypes infections.

Keywords: John’s disease; MIRU-VNTR analysis; MLSSR typing; molecular epidemiology; mycobacterium avium subsp. paratuberculosis; phylogenetic SNP based analysis; strain typing; whole genome sequencing (WGS).

FIGURE 1

Graphical plots showing the relationship of fecal shedding of MAP (qPCR Ct value) and blood ELISA (Sample/Positive Ratio) to animal age. Panels (A) and (B) reflect the profile of phenotypes collected in “high shedding” animals (Ct < 27), panels (C) and (D) reflect the trends present in “moderately shedding” animals (27 < Ct < 33), and panels (E) and (F) reflect the trends present in “low shedding animals” (Ct > 33). The arrows present on each graph represent the timepoint in each animal that fecal samples were selected for the culture of MAP. The grey line with points represented by “O” represents the S/p-value of the serum ELISA at that point in time. The black line with points represented by “X” represent the Ct results (mean value) obtained from qPCR at that point in time. The dashed line along the bottom represents the minimum ELISA threshold for isolates to be suspected of being MAP positive.

FIGURE 2

Average number of SNPs detected in the strains isolated within each herd. The number of SNPs in each strain was recorded (Supplementary Table S4A). The average number of SNPs detected in stains isolated in each herd with different letters differ significantly (p < 0.05) after application of a Tukey correction, as described in Material and Methods. The number of strains analyzed by whole genome sequencing (WGS) is reported for each herd. Each strain is derived from a single isolated colony from cow.

FIGURE 3

Core SNP phylogeny of Type II MAP strains. Phylogenetic tip labels are colored by herd. Tip labels in grey reflect type II strains used as reference sequences downloaded from the NCBI database. The isolates are divided into two major clades, represented by the main axis that is divided into two major branches. Arrows leading from one isolate to another are indicative of a confirmed transmission event, as described within Section 3.2. Arrows with arrowheads at both ends are present herds QC-2 and QC-3 indicate that while transfer between herds has likely occurred, the exact transmission pattern of MAP between hosts is unknown.

FIGURE 4

Visualization of the distribution of shedding status, SSR type and MIRU-VNTR type across different herds in ON and QC. Isolates are grouped by herd, with colors in each column reflecting a different value within their respective category. In the province column, isolates from QC are labeled in blue, while isolates from ON are labeled in red. Herd colors are the same as those in Figures 2, 3 and are labelled along the side of the figure. The shedding status column identifies isolates from animals that were high shedders (red), moderate shedders (orange) and low shedders (yellow). Isolates with no shedding data were colored grey. Both the SSR type and MIRU-VNTR type columns use colors to represent patterns, with unknown patterns colored white.