Molecular epidemiology of mycobacteria : development and refinement of innovative molecular typing tools to study mycobacterial infections

One approach of molecular epidemiology of mycobacteria is the genotyping and comparison
of DNA of infectious strains in order to monitor the transmission pathways of diseases. It is
based on the assumption that patients infected with clustered strains are epidemiologically
linked. Such results may help in understanding the modes of transmission and therefore in
putting in place an adapted control strategy. To perform molecular epidemiological studies
appropriate genotyping tools are a basic requirement. For M. tuberculosis they are well
developed but their appropriateness has to be evaluated in the geographical area of interest.
Like M. tuberculosis, M. bovis is also a member of the M. tuberculosis complex (MTC) and
causes bovine tuberculosis in cattle, humans and a wide variety of other hosts. However,
compared to M. tuberculosis, it is generally much more homogenic which renders the choice
of an appropriate genotyping tool much more challenging. M. ulcerans appears to be even
less diverse as, so far, strains have only been differentiated between but not within continents
(with the exception of Australia).
Therefore the overall aim of this study was to contribute to the development and refinement
of innovative molecular typing tools in order to study Mycobacterium tuberculosis, bovis and
ulcerans infections.
Variable Number Tandem Repeats (VNTR) typing is a genotyping tool which evaluates the
number of repeats at different loci distributed throughout the genome. We performed VNTR
typing of 12 Mycobacterial Interspersed Repetitive Units (MIRU) and 3 Exact Tandem
Repeats (ETR) for 40 M. tuberculosis strains from Chad. This revealed a similar
discriminatory power to spoligotyping, which evaluates the presence or absence of 43 spacer
DNA sequences between the 36 bp direct repeats (DRs) in the genomic DR region. Therefore,
VNTR typing for M. tuberculosis is as valid a genotyping tool as spoligotyping. However, in
contrast to spoligotyping, VNTR typing could also be useful in evaluating mixed infections
within different members of the M. tuberculosis complex members in the future.
Additionally, the use of both spoligotyping and VNTR typing could provide additional
valuable information for future micro-epidemiological studies of the possible highly virulent
Cameroon family clone. This clone is most prevalent in Nigeria, Cameroon and Chad, and is
defined by the lack of spoligo spacers 23-25 and by the loss of characteristic chromosomal
deletions.
We also performed spoligotyping and VNTR typing based on 16 known loci (12 MIRUs, 3
ETRs and VNTR 3232) for 67 M. bovis strains collected sequentially at the slaughterhouse of
N’Djaména, Chad. The strains originated from two different zebu breeds of which the
Mbororo was found to be more susceptible than the Arabe breed.
Genotyping of Chadian M. bovis strains confirmed the usual characteristically high
homogenetic population structure of M. bovis. We could even identify that the 67 strains are
members of only 2 clones. The clones were defined by spoligotyping (lack of spacer 30 vs.
lack of spacers 20-22) and the finding of characteristic chromosomal deletions, indicating that
the strains derived from two ancestral, single cells in the past. However, ETR A, B, C and
MIRU 26, 27 were most appropriate for first line typing of M. bovis strains from Chad and
superior than spoligotyping. This finding could help in identifying risk factors for inter
animal and also zoonotic transmission and therefore have important public health
implications.
As VNTR-typing is very attractive for M. tuberculosis complex members, attempts for using
VNTR typing for M. ulcerans have also recently been made. However, the presented
resolution was not higher than other genotyping tools. During this thesis, we identified a new
VNTR locus, designated ST1, which did not have any orthologues in the M. tuberculosis
genome. In combination with a previously published MIRU locus, we were able to identify
three different genotypes within Ghanaian M. ulcerans strains and therefore demonstrate
diversity in African strains for the first time. We further showed that DNA sequencing of the
different VNTR loci can refine the discriminatory power if the loci are analyzed separately
but, if analyzed commonly, doesn’t improve the overall discriminatory power. In the latter,
agarose gel electrophoresis of the amplification products of all polymorphic VNTR loci is
normally sufficient and sequencing does not result in further refinement.