Structural Genomics of Mycobacterium tuberculosis

Abstract

In 1998 the genome sequence of Mycobacterium tuberculosis H37Rv was published1. M. tuberculosis is the primary causative agent of tuberculosis, a disease with a long history in humans, which still has a great impact on human mortality today. As part of the M. tuberculosis Structural Genomics Consortium we selected nine target genes (Rv0534c (menA); Rv0548c (menB); Rv0553 (menC); Rv0555 (menD); Rv0542c (menE); Rv3853 (menG); Rv0558 (ubiE); Rv0989c (grcC2) and Rv0990c) from M. tuberculosis, including all known members of the menaquinone biosynthesis pathway, for structural studies. All nine genes were taken through the structural genomics “pipeline”, either becoming stuck at various “bottlenecks” or continuing successfully to structure solution. At the initial bioinformatics analysis step, eight of the nine targeted genes were deemed suitable for further study. PCR amplification and cloning of these genes into several different expression vectors followed. Expression of the gene products for the seven successfully cloned genes was undertaken in an E. coli expression host, followed by experiments (refolding, lysis buffer and expression temperature screens) aimed at obtaining soluble protein in sufficient quantities for crystallisation. Of the seven proteins successfully overexpressed, five remain at this stage as they could not be obtained in soluble form. The remaining two, Rv3853 (MenG), solubilised by refolding, and MenB, solubilised by 24ºC expression, were purified and both successfully produced diffracting crystals. The crystal structure of Rv3853 was determined by isomorphous replacement (SIRAS) and refined at 1.9 Å resolution (R = 19.0% and Rfree = 22.0%). The structure of several different crystal forms of MenB, were determined by molecular replacement. Refinement of two of these structures, MenB_P43212 at 2.15Å resolution (R = 20.3% and Rfree = 23.1%) and MenB_C2-NCoA at 2.3 Å resolution (R = 19.7% and Rfree = 22.5%), has been completed. The structure of Rv3853, combined with the discovery that UbiE was more likely to catalyse the final, S-adenosylmethionine-dependent, methyltransfer step of menaquinone biosynthesis, led to the conclusion that Rv3853 had been misannotated as MenG. Combined with further bioinformatics analysis the Rv3853 structure has been useful in providing new ideas as to the real function of Rv3853. In contrast, the structure of MenB confirmed its place as a member of the crotonase superfamily although the C-terminus was located in a position not observed in other crotonase superfamily structures. Several flexible regions likely to be important in MenB function have been identified by examination of the various MenB structures