An investigation into the biosynthesis of proximicins

Abstract

The proximicins are a family of three compounds – A-C – produced by two marine Actinomycete Verrucosispora strains – V. maris AB18-032 and V. sp. str. 37 - and are characterised by the presence of 2,4-disubstituted furan rings. Proximicins demonstrate cell-arresting and antimicrobial ability, making them interesting leads for clinical drug development. Proximicin research has been largely overshadowed by other Verrucosispora strain secondary metabolites (SM), and despite the publication of the V. maris AB18-032 draft, the enzymatic machinery responsible for their production has not been established. It has been noted in related research into a pyrrole-containing homolog – congocidine –due to the structural similarity exhibited, proximicins likely utilise a similar biosynthetic route. The initial aim of this research was to confirm the presumed pathway to proximicin biosynthesis. Following the sequencing, assembly and annotation of the second proximicin producer, Verrucosispora sp. str. MG37, and genome mining of V. maris AB18-032, no common clusters mimicked that of congocidine, casting doubt on the previously assumed analogous biosynthetic routes. A putative proximicin biosynthesis (ppb) cluster was identified, containing non-ribosomal peptide synthetase (NRPS) enzymes, exhibiting some homology with congocidine. NRPSsystems represent a network of interacting proteins, which act as a SM assembly line: crucially, adenylation (A)- domain enzymes act as the ‘gate-keeper’, determining which precursors are included into the elongating peptide. To elucidate the route to proximicins, activity characterisation of the four A-domains present in ppb cluster was attempted. The A-domain Ppb120 was shown to possess novel activity, demonstrating a high promiscuity towards heterocycle containing precursors, in addition to the absence of an apparent essential domain. This discovery refutes previous work outlining the core residues which dictate A-domain activity, while also presenting a facile route to novel heterocycle-containing compounds. Despite extensive work, A-domains ppb195 and ppb210, were ineffectively purified in the active form – informing future work into A-domains activity characterisation. Finally, the ppb220 A-domain which lies at the border of ppb, was inactive suggesting over-estimation of the cluster margins. To confirm ppb220 redundancy and confirm ppb boundaries, CRISPR/Cas gene editing studies were done. The gene responsible for the orange pigment of Verrucosispora strains was initially targeted and successfully deleted, and ppb studies commenced. The research here refutes the previously presumed route to proximicin biosynthesis; the ppb cluster instead comprises enzymes exhibiting unique activity and structure. The findings represent the foundations for allowing exploitation of chemistry exhibited within the proximicin family. The novelty exhibited can be utilised in the search for antimicrobial clinical leads, by allowing the production of compounds containing previously inaccessible heterocycle chemistry.