Mycobacterium bovis Molecular Typing and Surveillance 67
and environmental features. The phylogenetic
relatedness of MLVA types can be determined
and used to elucidate the evolutionary history of
various molecular types and clusters. Such
approaches clearly show the ongoing generation
of new variants, a substantial number of which
are identified each year, comprising ~50% of the
total MLVA types (not isolates) recovered.
M. bovis genotyping in conjunction with
comprehensive cattle movement databases and
structured wildlife surveillance offers a powerful
tool for investigating bovine TB source, mainte-
nance and spread (Skuce et al., 2010). The popu-
lation structure of M. bovis and the performance
characteristics of molecular typing support its
use to answer detailed epidemiological questions
of direct policy relevance.
5.7 Pathogen Genotype–Phenotype
AssociationsThe striking phylogeography disclosed recently
for the major lineages of the MTC has important
implications for lineage–lineage phenotypic dif-
ferences (Caws et al., 2008; Hershberg et al.,
2008). However, these observed inter-strain dif-
ferences are more convincing in experimental
studies than they are at the population
scale (Coscolla and Gagneux, 2010). Whether
M. bovis molecular types also display different
detectable, reproducible and relevant pheno-
types is being investigated.
In a Northern Ireland study, two proxy
measures for skin test detectability by pathogen
molecular type (MLVA) were developed: (i)
whether tuberculin test results differed signifi-
cantly by MLVA type; and (ii) whether the distri-
bution of MLVA types was significantly different
between TB reactors and tuberculin-negative
abattoir cases. Subtle, but ultimately non-
significant, differences were detected in the rela-
tive detectability by MLVA type. In addition, no
significant difference was detected in the disclo-
sure rate of non-reactor abattoirs (Wright et al.,
2013a). In a further statistical analysis, an
MLVA type effect was detected on virulence and
pathogenesis but not on outbreak size (Wright
et al., 2013b). The M. bovis population in North-
ern Ireland is comprised exclusively of the Euro-
pean 1 clonal complex and shows very limited
diversity at that scale, so it is not unexpected that
no significant differences were detected in the
phenotypes investigated. Recent studies in North
America (Waters et al., 2014) did not detect a
significant genotype–phenotype difference in
the isolates tested. However, differences at the
spoligotype scale were reported in Argentina
(Garbaccio et al., 2014). It remains to be deter-
mined whether phenotypic differences, if detect-
able, are expressed between different clonal
complexes, or within them. Additionally, pheno-
type may be influenced by RNA regulation and
differences in epigenetic profile may be seen in
M. bovis lineages (Drewe and Smith, 2014).
Genomics has been used to investigate pheno-
typic differences between M. bovis and M. caprae
and between M. bovis isolates of the same spoli-
gotype and provides evidence for potential cor-
relates of bacterial viability and virulence (de la
Fuente et al., 2015).
Northern Ireland has also investigated the
extent of M. bovis infection in road-kill badgers.
Where M. bovis was confirmed, isolates were also
MLVA typed. In all cases the MLVA types identi-
fied in badgers were also found in local cattle.
M. bovis MLVA types in badgers also showed
strong geographical clustering to regions and
this pattern was very similar to that disclosed in
cattle herds. M. bovis MLVA types in both cattle
and badgers were mostly clustered to the same
geographical regions. This is indirect evidence of
an ‘association’ between TB infections in cattle
and badgers. However, this association does not
indicate the direction of transmission, or the
relative importance of badger-driven versus
cattle- driven transmission in generating this
association whether on an individual animal/
herd, regional or province-wide basis. Similar
findings and interpretations have been reported
from studies in other countries, including Ire-
land, England and Wales (Olea-Popelka et al.,
2005; Goodchild et al., 2012).5.8 Whole-Genome Sequencing and
Genomic EpidemiologyAlthough there remain significant blind spots
and missing data, molecular epidemiology stud-
ies have improved the understanding of bovine
TB epidemiology, ecology and evolution. Here