Molecular studies of Mycobacterium ulcerans

Mycobacterium ulcerans causes a serious necrotic skin disease in humans called Buruli ulcer. The third most common mycobacterial infection in man, the disease has been reported worldwide with a major disease burden in Western and Central Africa. The destructive pathology of Buruli ulcer is primarily due to the presence in the bacterium of a cytotoxic lipid called mycolactone. Mycolactone is the product of specific polyketide synthases (PKS) that are encoded by three large plasmid-borne genes (mlsA1, mlsA2 and mlsB). Three accessory genes are also thought to play a role in mycolactone synthesis (mup038, mup045 and mup053) and all are co-located with the mycolactone PKS genes on the pMUM plasmid. All mycobacteria that synthesize mycolactones appear to have a common recent ancestor and have evolved from a Mycobacterium marinum-like ancestor by a process of horizontal gene transfer and reductive evolution. The overall aim of this thesis was to explore the genetic basis for the synthesis of the mycolactones. The first results chapter describes investigations into the regulation of the mycolactone pMUM genes by a combination of primer extension, site-directed mutagenesis and reporter plasmid analysis, revealing a powerful SigA-like promoter driving expression of the mlsA1/mlsB genes. Using a similar approach, the promoters of the putative mycolactone accessory genes mup045 and mup053 were also mapped. Exploitation of the novel SigA-like promoter upstream of mls by transcriptional coupling to a green fluorescent protein reporter gene enabled exploration of the interaction of M. marinum and M. ulcerans in mosquito larvae. M. ulcerans was observed to contaminate the mouth and mid-gut of mosquito larvae whereas M. marinum did not. Despite sharing >98% nucleotide identity with M. ulcerans, M. marinum is unable to produce mycolactones even when equipped with the pMUM plasmid, suggesting that other genetic differences between M. ulcerans and M. marinum (in addition to pMUM) are required for mycolactone synthesis. To explore the genetic basis for this observation by comparative genomics, a third M. ulcerans/M. marinum complex (MuMC) genome was fully assembled and annotated. The second results chapter describes the fully assembled 6.2 Mbp genome of the frog pathogen M. ulcerans Liflandii (MuL). Genome analysis shows that MuL is closer in length and architecture to M. marinum than to M. ulcerans Agy99. However, like M. ulcerans Agy99, MuL also appears to have undergone genome reduction by DNA loss and mutation, but with several specific gene lesions likely rendering this bacterium a phenylalanine, tyrosine and tryptophan auxotroph. While no specific differences were observed that might explain the capacity of M. ulcerans to produce mycolactone, the genomic characteristics of MuL indicate it is also a niche-adapted mycobacterium similar to M. ulcerans isolates from Africa. However, one could also infer that MuL is adapting to a slightly different privileged environment, with somewhat different pressures shaping its genome compared to other M. ulcerans isolates. In addition to exploring these genomic differences, we also assessed the contribution to mycolactone synthesis of a protein constitutively expressed in M. ulcerans but not in M. marinum. The gene MUL_2232 encodes an antigenic 18kDa protein small heat shock protein that is up-regulated in all M. ulcerans strains through a point mutation in the promoter region of its specific repressor, hspR_2. This mutation is not present in M. marinum, which as a consequence, has low Hsp18 expression. Using gel chromatography and insulin protection assays, a chaperone function was attributed to this protein. The hypothesis that Hsp18 was acting as a chaperone for the mycolactone megasynthases was also tested. However, the absence of Hsp18 in M. ulcerans had no impact on mycolactone production. In this thesis the promoters controlling expression of genes required for mycolactone biosynthesis have been characterized, a third, fully assembled reference genome in the MuMC has been established and a function has been identified for the M. ulcerans protein, Hsp18. These advances in our understanding of M. ulcerans represent useful additions to our knowledge of this pathogen and will open new avenues of research in the fight against Buruli ulcer.