Crystal structure of M. tuberculosis phosphopantetheinyl transferase PptT

Phosphopantetheinyl transferases (PPTases) are essential enzymes that catalyze covalent attachment of the 4'-phosphopantetheine (4'-PP) moiety from coenzyme A (CoA) to a conserved serine residue on acyl (ACP) and peptidyl carrier proteins (PCP) [1]. This post-translational modification converts the inactive apo-carrier proteins to the functional form, shuttling the intermediates of biosynthetic reactions catalyzed by fatty acid synthases (FAS), polyketide synthases (PKS) and non-ribosomal peptide synthetases (NRPSs). In Mycobacterium tuberculosis (Mtb), the causative agent of tuberculosis (TB), two PPTases, AcpS (type-I) and PptT (type-II), are involved in the biosynthesis of essential lipids, virulence factors and siderophores, activating over 20 target proteins [2]. These two proteins have been shown to be independently essential, suggesting that PPTases could be targeted against tuberculosis [2, 3]. We have expressed the Mtb-PptT protein as a fusion protein with maltose binding protein (MBP). The use of the MBP-PptT fusion protein overcame stability and solubility problems and resulted in successful crystallization. The structure of Mtb-PptT in complex with CoA was determined from the crystal of the fusion protein, solved at 1.75 Å resolution. Excellent electron density is present for all parts of the CoA molecule, revealing a conserved CoA-binding mode. Conformational and charge distribution differences in the putative ACP binding cleft, however, suggest a different mode of ACP binding compared to other homologues. This is the first and only three-dimensional structure of a type-II PPTase from pathogenic bacteria, providing structural features that can be exploited in drug development when compared with its human counterpart.

Characterization of proline utilization pathway in Mycobacterium tuberculosis

The proline utilization pathway in Mycobacterium tuberculosis (Mtb) has been recently identified as an important factor in Mtb persistence in vivo, suggesting that this pathway could be a valuable therapeutic target against tuberculosis (TB). In Mtb, two distinct enzymes perform the conversion of proline into glutamate; the first step is the oxidation of proline into Δ1-pyrroline-5-carboxylic acid (P5C) by the flavoenzyme proline dehydrogenase (PruB) and the second reaction involves converting the tautomeric form of P5C (glutamate-γ-semialdehyde) into glutamate using the NAD+-dependent Δ1-pyrroline-5-carboxylic dehydrogenase (PruA). Here we describe three-dimensional structures of Mtb-PruA, determined by X-ray crystallography both in its apo state and in complex with NAD+ at 2.5 and 2.1 Å resolution, respectively. The structure reveals a conserved NAD+ binding mode, common to other related enzymes. Conformational differences in the active site, however, linked to changes in the dimer interface, suggest possibilities for selective inhibition of Mtb-PruA despite reasonably high sequence identity with other PruA enzymes. Using recombinant PruA and PruB, the proline utilization pathway in Mtb has also been reconstituted in vitro. Functional validation using a novel NMR approach has demonstrated that the PruA and PruB enzymes are together sufficient to convert proline to glutamate, the first such demonstration for monofunctional proline utilization enzymes.

Protein crystallography in the 1970s: starting from scratch in a remote outpost

A conversation with Dorothy Hodgkin on a long bus trip, just before I returned to New Zealand in 1970, left me full of determination and optimism. This presentation will recount my experience in starting a protein crystallography lab with only a sealed-source generator and a precession camera for equipment. Crystals had to be large (0.5 - 1.0 mm on edge) and X-ray data, collected at room temperature on a borrowed small-molecule diffractometer, accumulated very slowly. We corrected for absorption and decay and scaled data sets rather crudely. It was very much a do-it-yourself environment. With no CCP4, software had to be borrowed or adapted or written oneself; methods papers in Acta Cryst. were like gold as I pored through them trying to understand. Communications with friends, by airmail, were vital. An electron density map for the cysteine protease actinidin, at 2.8 Å, was immediately interpretable thanks to excellent data and phases from more derivatives than were really necessary. An R factor of 42% to 2.0 Å, for a model built in a Richards box, was really quite astonishing. Later FFT-based least squares refinement at the University of York in 1978 was even more astonishing as the R factor rocketed down [1]. No computer graphics, but the difference electron density told the story - in retrospect there was even evidence that the crystals (prepared from kiwifruit bought at the local shop) contained several genetic variants of the protein! It may not have been the most exciting protein in the world (except to me!) but what a way to learn protein crystallography and protein structure.