Genomic Epidemiology of MAP

6.3.1 Host preference

There appear to be epidemiological trends as­sociated with MAP-C and MAP-S strains with respect to transmission, host preference and susceptibility to infection.

MAP-S strains have been isolated predom­inantly, but not exclusively, from sheep and goats, suggesting a preference for these host species. Genetically distinct Type III MAP-S isolates have also been found in Arabian cam­els, though only a limited number of isolates have been genetically characterized to date (Ghosh etal., 2012). Reports of the occurrence of MAP-S strains in wildlife are relatively un­common. One MAP-S isolate was obtained from a fallow deer (Dama dama) (Machackova et al., 2004) and another from a house mouse (Mus musculus) (Florou et al., 2008). MAP-S isolates were also recovered from tissue sam­ples of kangaroos (Macropus fuliginosus fuligi- nosus) and wallabies (Macropus eugenii decres) grazing with infected sheep on Kangaroo Island, though faecal shedding of MAP by the macropods was not observed (Cleland et al., 2010).

MAP-C isolates have a broad host range and are commonly isolated from both domes­ticated and wildlife species, including non­ruminants (Table 15.1 Fox et al., Chapter 15, this volume), although clinical disease has only been observed in ruminants, camelids, rabbits and hares. MAP-C is by far the most common MAP lineage isolated from cattle. MAP isolates from human Crohn's disease patients that have been typed have all been classified as MAP-C (Whittington et al., 2000; Bull et al., 2003; Ghadiali et al., 2004; Wu et al., 2006; Griffiths et al., 2008; Paustian et al., 2008; Singh et al., 2009; Wynne et al., 2011; Bannantine et al., 2014; Timms et al., 2015) and appear to be most closely related to isolates from cattle or bison in the same geographical region (Bryant et al., 2016).

Table 6.3. Mycobacterium avium subsp. paratuberculosis (MAP) strain-specific gene clusters that may be associated with differences in virulence and pathogenic traits.

Continued

6.3.2 Transmission

Cross-species transmission of MAP-S strains has been observed in natural settings. There are some reports of cattle naturally infected with MAP-S strains, including a pigmented MAP strain (Watt, 1954), and these infections have typically been associated with bullfighting breeds (de Juan et al., 2006) or cases where there has been direct or indirect contact of calves with sheep infected with MAP-S strains (Whittington et al., 2001b). The most convincing demon­stration of natural cross-species transmission of MAP-S strains between sheep and cattle oc­curred in Iceland. Paratuberculosis was intro­duced in 1938 via infected Karakul sheep from Germany and in 1944 spread to the local cattle population, in which it subsequently became en­demic (Fridriksdottir et al., 2000). The Icelandic strains were classified retrospectively as MAP-S (Whittington et al., 2001b). MAP-S strains have also been isolated from farmed red deer in New Zealand (de Lisle et al., 1993), but such isola­tions appear to be rare (Verdugo et al., 2014) and experimental infection of deer calves has suggested that deer are less susceptible to MAP-S than to MAP-C strains (O'Brien et al., 2006). Calves experimentally infected with either pig­mented or Icelandic ovine strains developed clinical paratuberculosis (Taylor, 1953) and the pigmented strain isolated from cattle by Watt (1954) also could be transmitted experimen­tally to sheep.

Data are accumulating regarding the geo­graphical distribution of MAP strains, which has probably been influenced by many factors, including animal movements, strain virulence and farm management systems. MAP-S strains are prevalent in Australian sheep and, despite the fact that MAP-C strains have been isolated from Australian cattle, they are rarely, if ever, isolated from Australian sheep (Whittington et al., 2000; Windsor, 2015). This is in contrast to Europe and New Zealand, where both MAP-S and C strains are isolated from sheep (Stevenson et al., 2009; Verdugo et al., 2014). This differ­ence is possibly linked to differences in manage­ment practices between some parts of Europe and Australia, the scale of the farming practices, and relative proportions of sheep and cattle co- or sequentially grazing. MAP-S pigmented iso­lates have been isolated most commonly from animals in the UK and the Faroe Islands (Taylor, 1951; Stevenson etal., 2002). Pigmented strains have also been reported in Spain (Sevilla et al., 2007) and Morocco (Benazzi et al., 1996) but are relatively uncommon in these settings. The first confirmed MAP-C pigmented strain (C4A4) was recently isolated from goat faeces in Azores, Portugal (Barbosa et al., 2017). In North India, Type B MAP isolates are most commonly isolat­ed, though there are some regional differences, with non-Type B isolates more common in New Delhi and Agra (Singh et al., 2009).

Within- and cross-species transmission of specific MAP strains is difficult to conclusively determine in non-experimental settings. Even with the highest resolution molecular typing tools, epidemiological data are needed to sup­port inference regarding transmission events. This is exemplified by the low mean nucleotide substitution rate in the MAP genome, which was estimated to be below 0.25 substitutions per ge­nome per year, with an upper 95% confidence limit of less than 0.5 substitutions per genome per year (Bryant et al., 2016).

WGS has been particularly useful in iden­tifying the relative number of SNP differences between isolates to derive epidemiological in­ferences at different spatial scales (Ahlstrom et al., 2016a; Bryant et al., 2016). Sequencing of 141 diverse MAP isolates revealed the global diversity, though the authors found no evidence of strong geographical clustering (Bryant et al., 2016). At a national and provincial scale, WGS of 182 MAP isolates from Canada indicated that Canadian isolates represented a subset of the known global diversity, and similarly dem­onstrated weak evidence of geographical clus­tering within Canada (Ahlstrom et al., 2016a). However, a polymerase chain reaction (PCR) as­say designed to target five SNPs that differentiat­ed isolates into four strains relevant to Canadian dairy herds was used to screen 602 MAP iso­lates from Canada (Ahlstrom et al., 2016b) and indicated that some strains appeared to be overrepresented in certain provinces. Comprehensive within-herd level SNP diversity has not been reported to date, though sequenc­ing of multiple isolates per herd has confirmed the co-occurrence of genetically unrelated MAP isolates as well as clonal isolates within a single herd and between herds (Ahlstrom et al., 2015; Davidson et al., 2016).

Human-origin MAP isolates, isolated from patients with Crohn's disease, ulcerative colitis and non-inflammatory bowel disease controls, were found to show high genomic similarity to each other as well as to an isolate from a bovine host with clinical paratuberculosis from the same region in Australia (Wynne et al., 2011). A SNP genotyping approach targeting SNPs unique to isolates derived from human, bovine and ovine hosts was used to screen 52 isolates predominantly from Australia (Wynne et al., 2014). The authors found that some human­origin MAP isolates were highly similar to those derived from bovine hosts, whereas others were more divergent. These studies provide evidence for potential zoonotic transmission of MAP.

The endemic nature of MAP in many coun­tries challenges efforts aimed at understand­ing transmission dynamics, as multiple strains co-circulate in most regions. Molecular targets previously used to estimate the diversity and ge­netic relationship of MAP isolates are often not discriminatory enough, not biologically relevant or insufficiently stable. Indeed, the reliability of traditionally used genotyping schemes to accu­rately reflect genetic similarity and phylogenetic relationships of MAP strains has been assessed using WGS (Ahlstrom et al., 2015; Bryant et al., 2016). For example, MIRU-VNTR typing and IS 1311 typing were each found to have limita­tions in their ability to differentiate strains. As such, the limitations of genotyping tools should be considered when interpreting studies based on those tools, and further comparison using WGS is warranted to determine if currently used molecular markers are appropriate to differen­tiate strains on an epidemiologically relevant scale.

Molecular tools have nevertheless been, and continue to be, instrumental in supplement­ing epidemiological data to infer inter- and in­traspecies transmission of MAP among domestic ruminants (Verdugo et al., 2014; Marquetoux et al., 2016), between domestic ruminants and both ruminant and non-ruminant wildlife (Stevenson et al., 2009; Fritsch et al., 2012) and among wildlife (Forde et al., 2012). Within a herd, short-sequence repeat (SSR) typing of MAP indicated that supershedders may play an important role in spreading MAP to herd mates (Pradhan et al., 2011). Within an individual ani­mal, SSR typing has suggested mixed infection within an individual cow, which was later con­firmed by WGS (Davidson et al., 2016). For more complete descriptions on the molecular epidemi­ology of MAP, see reviews (Stevenson, 2015; Li et al., 2016; Fawzy et al., 2018).

6.4