Bovine tuberculosis

Mycobacterium bovis Molecular Typing and Surveillance 61

metagenomics (Doughty et al., 2014; Votintseva
et al., 2015) may further improve the prospects
for real-time molecular surveillance.

5.4.3 Molecular typing methodology

Prior to the DNA-based studies, much of what
was known about mycobacterial ecology, patho-
genesis and epidemiology came from early meth-
ods used to identify these challenging bacteria.
Such methods relied on strain phenotypic
characteristics, including colony morphology,
susceptibility to antimicrobials, biochemical and
serological reactivity and mycobacterial phage
typing. While some are still of value, such phe-
notypic methods have largely been superseded
by methods that index structural genetic diver-
sity in mycobacterial genomes (Jagielski et al.,
2016). Due to the high degree of genetic similar-
ity and extreme clonality in modern MTC bacte-
ria and the consequent co-linearity of their
genomes, these methods are applicable and
remarkably congruent across the ecotypes that
comprise the MTC. The methods do vary in some
performance characteristics, for example their
discriminatory power, and they do have various
strengths and weaknesses (Schurch and van
Soolingen, 2012). Structural genetic markers
that tend to evolve unidirectionally (e.g. SNPs
and deletions) are well suited to phylogenetic
and evolutionary studies, whereas markers that
evolve more quickly and bi-directionally (e.g.
mini-satellites and repeat variations) are prone
to convergent evolution (homoplasy) but are
better suited to outbreak investigations.
Prior to DNA amplification technologies,
molecular epidemiology was based on gel electro-
phoretic analysis of bacterial whole-genome
restriction enzyme analysis (REA) or the subse-
quent restriction fragment length polymorphism
(RFLP) analysis, which probed REA profiles with
repetitive DNA sequences. REA (Collins, 1999)
and RFLP typing (van Soolingen, 2001) were
standardized and became routine procedures in
several public health and some veterinary
research laboratories. While considered discrimi-
nating at the time, these techniques were techni-
cally demanding, requiring expensive software
for archiving complex banding patterns and
inter-laboratory reproducibility was not trivial

(Heersma et al., 1998). Most M. bovis isolates

sampled worldwide turned out to belong to the

same clonal complex and tended to have only one

copy of the IS 6110 RFLP target (Smith, 2012).

Due to the landmark achievements of the

first genome sequences for M. tuberculosis (Cole

et al., 1998) and M. bovis (Cole, 2002; Garnier

et al., 2003), the mycobacterial diagnostics and

research community now has a tool box of

molecular genetic tests that discriminate the

MTC at different geographic and evolutionary

scales (Pepperell et al., 2011, 2013; O’Neill et al.,

2015). The first of the more portable PCR-

enabled methods was spoligotyping (Kamerbeek

et al., 1997). This technique indexes unidirec-

tional possession of short repeat sequences at a

locus since shown to be a clustered regularly

interspersed polymorphic repeat (CRISPR), a

locus now finding game-changing applications

in gene-drive and genome-editing technologies

(Carlson et al., 2016; Wright et al., 2016). Spoli-

gotyping methodology, including nomenclature,

is internationally accepted (Mbovis.org, SIT-

VIT2) and provides a moderate level of discrimi-

nation between isolates.

Significantly higher discrimination was

achieved by configuring molecular typing tests

based on multi-locus tandem repeat variation

in mycobacterial interspersed repetitive units

(MIRUs) and variable number of tandem repeats

(VNTRs), the so-called MIRU-VNTR or multi-

locus VNTR analysis (MLVA) techniques. Meth-

odology and applications are reviewed elsewhere

(Drewe and Smith, 2014; Robbe-Austerman

and Turcotte, 2014) and will not be repeated in

detail here. Unfortunately, nomenclature for

naming M. bovis molecular types using these

tests has not been settled internationally.

5.5 International Findings of M. bovis

Molecular Typing

Here we summarize some general findings from

deployment of genome-enabled molecular typ-

ing tests internationally. Systematic analysis of

M. bovis sampled worldwide has disclosed strik-

ing region-to-region difference in the M. bovis

clonal complexes that dominate within regions

(Rodriguez-Campos et al., 2012a; Smith, 2012;

Allen et al., 2013).