Chapter 3: Results
3.1 Constituents of the cobalamin biosynthetic pathway in Mtb
The whole-genome sequences of Mtb, together with the related mycobacterial pathogens, M.
marinum, M. kansasii and M. leprae, revealed diversity amongst genes associated with cobalamin-related metabolism as a consequence of gene deletion events, differential acquisition of genes by horizontal transfer, and single nucleotide polymorphisms with predicted impact on protein function and transcriptional regulation (Young et al., 2015).
Differences in cobalamin synthesis, methionine biosynthesis, fatty acid catabolism, and DNA repair and replication are consistent with adaptations to different environmental niches and pathogenic lifestyles (Young et al., 2015). Multiple comparative genomics studies have predicted that mycobacteria can assimilate or utilize cobalamin (Gopinath, Venclovas, et al., 2013; Minias et al., 2018; Rodionov et al., 2003; Shelton et al., 2019; Vitreschak et al., 2003;
Zhang et al., 2009). M. leprae (like Mtb, an obligate pathogen) lacks a complete set of genes
55
for cobalamin biosynthesis but retains those genes required for transporting exogenous cobalamin precursors (Rodionov et al., 2003). Mtb, on the other hand, possesses a near complete cobalamin biosynthetic pathway (Gopinath, Moosa, et al., 2013; Rodionov et al., 2003), whereas M. marinum and M. kansasii possesses complete cobalamin biosynthetic pathways (Young et al., 2015).
Previous studies in the Molecular Mycobacteriology Research Unit have identified putative homologues of most cobalamin biosynthetic enzymes in Mtb and Msm, enabling the construction of a predicted cobalamin biosynthetic pathway (Gopinath, Moosa, et al., 2013).
The proposed pathway is suggestive of the aerobic type because the CobG mono-oxygenase, which contains an iron-sulphur centre and is responsible for converting precorrin-3A into precorrin-3B, is of the class that requires molecular oxygen for activity (Debussche et al., 1993). Also, Mtb possesses aerobic-type cobN-Rv2850c-encoded subunits. In addition, it is predicted that cobalt insertion occurs late in the Mtb cobalamin biosynthetic pathway, and the Mtb CobK and CobJ orthologues contain residues commonly found in aerobic bacteria (Warren et al., 2002). Notably, Mtb also appears to possess some features that are characteristic of anaerobic biosynthesis. For example, Rv0259c exhibits homology to CbiX, a cobalt chelatase (Raux et al., 1998), that was identified in Bacillus megaterium – possibly implying that mycobacteria might possess partial capacity to use the anaerobic pathway (Gopinath, Moosa, et al., 2013).
As noted above, a distinctive feature of Mtb is that the cobF-encoded precorrin-6a synthase was lost during evolution from the M. canettii-like ancestor (Supply et al., 2013; Young et al., 2015). Instead, it has been suggested that CobF function in Mtb is provided by an alternative methyltransferase: either Rv2067c, which contains a C-terminal methyltransferase domain, or Rv0391 (designated as metZ) (Gopinath, Moosa, et al., 2013; Rodionov et al., 2003). In addition, it has been postulated that Mtb may rely on the host environment as a source of cobamides (Young et al., 2015).
Cobalt insertion into hydrogenobyrinic-acid a,c-diamide is predicted to be catalyzed by a CobN-CobS-CobT fusion protein (Debussche et al., 1992; Rodionov et al., 2003). CobS and CobT are thought to form a complex that interacts with CobN to generate cob(II)yrinic acid a,c-diamide (Moosa, 2013). The gene encoding the cob(II)yrinic acid a,c-diamide reductase is yet to be identified in the mycobacterial genome. In Mtb, BluB, which also catalyses the formation of DMB, was predicted to encode the cobalt reductase function (Gopinath, Moosa,
56
et al., 2013; Rodionov et al., 2003) based on sequence homology to CobR, a predicted cob(II)yrinate a,c diamide reductase (Lawrence et al., 2008).
In contrast to Mtb, Msm appears to encode all enzymes required for de novo cobalamin biosynthesis and the predicted pathway is also suggestive of the aerobic type. Additionally, Msm encodes five other putative cobalamin biosynthesis proteins (MSMEG_1123, MSMEG_2607, MSMEG_4934, MSMEG_6048 and MSMEG_6069). Although their enzymatic functions have not been confirmed, their predicted gene annotations are suggestive of roles in cobalamin metabolism. Furthermore, Msm encodes putative homologues of additional cobalamin-dependent enzymes that are not encoded in Mtb: specifically, glutamate mutase (MSMEG_0969), ethanolamine ammonium lyase made up of small and large subunits (MSMEG_1553-1554), and three glycerol dehydratases made up of small and large subunits (MSMEG_0496-0497, MSMEG_6320-6321 and MSMEG_1547-1548). Therefore, while the genome of Msm does not encode the same cobalamin-dependent enzymes that characterizes the genome of Mtb, this non-pathogenic mycobacterium appears to contain the full machinery required for de novo cobalamin biosynthesis. Moreover, in vitro biosynthesis of cobalamin was recently confirmed microbiologically and biochemically (Kipkorir et al., 2021).
In this study, a genome-scale approach was adopted to yield detailed genetic maps of de novo cobalamin biosynthesis and salvage in Msm. Msm is convenient model for the general study of mycobacteria because it has a relatively fast doubling time of approximately 3 hr and requires a biosafety level 2 laboratory. On the other hand, Mtb is a slow grower, has doubling time of 24 hr in standard broth medium and requires biosafety level 3 facilities (James et al., 2000).
Moreover, different aspects of mycobacterial physiology and metabolism have been studied using non-pathogenic saprophyte, Msm strain mc2155 as a model mycobacterium (Barry, 2001;
Reyrat & Kahn, 2001). However, the comparison between an environmental mycobacterium (Msm) and an obligate human pathogen (Mtb) might be more useful than simply regarding Msm as a “model”. Differences observed in cobalamin-dependent metabolism between Msm and Mtb might be more informative in understanding the evolution of Mtb into a human pathogen, as well as understanding the role of cobalamin in mycobacterial pathogenesis (Kipkorir et al., 2021). The approach involved selection of conditionally essential genes using whole-genome transposon (Tn) mutagenesis combined with next-generation sequencing, called transposon insertion sequencing (TnSeq) (Cain et al., 2020; Chao et al., 2016). In this study, the frequencies of mutants under-represented or over-represented in pools of an Msm
57
ΔmetE Tn mutant library we investigated during growth in defined medium with or without CNCbl or cobalt supplementation. Following negative selection under nutrient limited or nutrient-starved conditions, mutants with insertions in genes that are conditionally essential should fail to survive the selection pressure and will, therefore, be underrepresented. In this study, TnSeq was applied to elucidate the full complement of de novo cobalamin biosynthetic genes in Msm. It was predicted that disruption of any gene required for cobalamin biosynthesis would result in growth impairment of the ΔmetE mutant – a strain dependent on the alternative, cobamide-dependent methionine synthase, MetH, for methionine production (Kipkorir et al., 2021) – when grown in medium without CNCblsupplementation.