N-myristoyltransferase from Leishmania donovani: structural and functional characterisation of a potential drug target for visceral leishmaniasis

Abstract

N-Myristoyltransferase (NMT) catalyses the attachment of the 14-carbon saturated fatty acid, myristate, to the amino-terminal glycine residue of a subset of eukaryotic proteins that function in multiple cellular processes, including vesicular protein trafficking and signal transduction. In these pathways, N-myristoylation facilitates association of substrate proteins with membranes or the hydrophobic domains of other partner peptides. NMT function is essential for viability in all cell types tested to date, demonstrating that this enzyme has potential as a target for drug development. Here, we provide genetic evidence that NMT is likely to be essential for viability in insect stages of the pathogenic protozoan parasite, Leishmania donovani, causative agent of the tropical infectious disease, visceral leishmaniasis. The open reading frame of L. donovani NMT has been amplified and used to overproduce active recombinant enzyme in Escherichia coli, as demonstrated by gel mobility shift assays of ligand binding and peptide-myristoylation activity in scintillation proximity assays. The purified protein has been crystallized in complex with the non-hydrolysable substrate analogue S-(2-oxo)pentadecyl-CoA, and its structure was solved by molecular replacement at 1.4 A resolution. The structure has as its defining feature a 14-stranded twisted beta-sheet on which helices are packed so as to form an extended and curved substrate-binding groove running across two protein lobes. The fatty acyl-CoA is largely buried in the N-terminal lobe, its binding leading to the loosening of a flap, which in unliganded NMT structures, occludes the protein substrate binding site in the carboxy-terminal lobe. These studies validate L. donovani NMT as a potential target for development of new therapeutic agents against visceral leishmaniasis.

Fig. 1

The NMT gene codes for an essential enzyme in L. donovani promastigotes. (a) Restriction maps of the wild-type NMT locus and targeted single alleles containing replacement by either the HYG or PAC genes. 5′Flank and 3′flank boxes represent NMT flanking regions used for gene targeting. 5′DHFR and 3′DHFR boxes represent 5′ and 3′ dihydrofolate reductase flanking regions. Arrows show positions of primers used in PCR analysis of clones in (c). Position of the DIG-labeled probe hybridized to membranes in (b) indicated by stippled box. (b) Southern blot analysis of ScaI and BamHI digests of parasite genomic DNA hybridized with 5′U probe, which hybridizes to the 5′ flanking region. Arrows indicate the following bands: 5.0 kb, HYG replacement at NMT locus; 4.6 kb, PAC replacement at NMT locus; 3.7 kb, Wt NMT locus. Upper panel: wild type (Wt), three single allele HYG replacement (+/ΔNMT::HYG H1, H4, H16) and 11 clones generated from H4 +/ΔNMT::HYG parent clone by attempted replacement with PAC (H4 +/ΔNMT::HYG + PAC clones 1–12). Lower panel: 12 clones generated from an H4 +/ΔNMT::HYG parent clone overexpressing NMT by attempted replacement with PAC (+/ΔNMT::HYG [pTEX NEO NMT] +  PAC) and a single allele PAC replacement (+/ΔNMT::PAC, P1). (c) PCR analysis of the wild-type NMT locus (using primers Ufor and Nrev) and NMT ORF (using primers Nfor and Nrev). Clones are as described in (b). (d) Groups of five BALB/c mice were inoculated with 4 × 10⁷L. donovani metacyclic promastigotes (either wild-type parasites or one of the +/ΔNMT::HYG clones [NMT SKO]) intravenously. Parasite burdens in the spleen (left panel) and liver (right panel) were determined at 28 days post-infection by examination of methanol-fixed, Giemsa-stained tissue imprints, as previously described . Data are presented as Leishman Donovan units (LDU), in which LDU represents the number of intracellular amastigotes/1000 host cell nuclei  ×  organ weight (mg). Statistical analysis was performed by a paired Student's t-test.

Fig. 2

Protein mobility shifts on native electrophoresis. Altered mobility of LdNMT protein on non-denaturing PAGE (7.5% polyacrylamide; run conditions, 90 V for 100 min) corresponds to complex formation with cofactor. 3 μg of protein (lane 1) was incubated with myristoyl-CoA in 10-fold molar excess (lane 2) or 2-fold excess of ARF peptide (lane 3). Low-mobility material is visible when both cofactor and peptide are present (lane 4), which does not appear when the non-hydrolysable cofactor NHM is used (lane 5, protein + NHM; lane 6, +NHM + peptide).

Fig. 3

Sequence alignment of NMTs. Structure-based amino acid sequence alignment of NMT proteins of known structure (Ld, L. donovani; Hs, human; Ca, C. albicans; Sc, S. cerevisiae). Strictly conserved residues are coloured red and well-conserved residues are coloured yellow. The secondary-structure elements of LdNMT and ScNMT are shown above and below the sequence alignment, respectively. This figure was generated using ESPript.

Fig. 4

The three-dimensional structure of LdNMT. (a) Stereo ribbon representation of LdNMT in its complex with NHM, which is shown in cylinder format and coloured by atom type: carbon, cyan; oxygen, red; nitrogen, blue, sulfur, yellow; phosphorus, magenta. The protein chain is colour ramped from residue 11 at the N-terminus (red) to residue 421 at the C-terminus (magenta). The secondary-structure elements are labelled with uppercase letters for α-helices and lowercase letters for β-strands. The ' and ” superscripts indicate elements additional to those observed in the first NMT structures to be described. (b) Topology diagram of LdNMT with secondary-structure elements labelled and with α-helices represented as circles and β-strands as triangles. Lobe 1 is shown above lobe 2 with the light blue shading indicating the tandemly duplicated region. The italicised N and C indicate the chain termini. The additional helices C′ and C ″ appear as an insert coloured in orange in lobe 1 at a region that is disordered in lobe II. (c) Electrostatic surface representation of LdNMT. The molecule is shown in an orientation similar to that in (a). Blue and red surface colouring represents positive and negative electrostatic potential, respectively. The NHM ligand is shown in cylinder format and coloured by element as above except that carbon atoms are in green. (d) Comparison of the structures of LdNMT and ScNMT. The structures of LdNMT–NHM and ScNMT–NHM–GLYASKLA (PDB code 1IIC³⁵) were superposed using the SSM superpose routine in CCP4MG. The structures are displayed as ribbons, LdNMT (blue) and ScNMT (green), with the ligands represented in cylinder format. The Ab loop from the superposed ScNMT–myristoyl-CoA binary complex (1IIC) is shown in red.

Fig. 5

Fatty acyl-CoA binding to LdNMT. (a) 2F_o − F_c electron density contoured at 3σ and displayed in the neighbourhood of the NHM species in the refined LdNMT model. (b) Orthogonal views of the fatty acyl-CoA in the LdNMT binary complex and ScNMT binary and ternary complexes. Carbon atoms are coloured according to structure: LdNMT–NHM, cyan; ScNMT–myristoyl-CoA, green; ScNMT–NHM–GLYASKLA, coral. The fatty acyl-CoA species were superposed by least-squares methods. (c) Stereo view of the NHM and neighbouring residues in LdNMT. The fatty acyl-CoA is shown with thicker bonds relative to the protein and its carbon atoms are coloured in cyan rather than green. The remaining atoms are coloured as follows: oxygen, red; nitrogen, blue, sulfur, yellow; phosphorus, magenta. The backbone of the protein is indicated as a light blue ribbon. Polar interactions between the protein and the fatty acyl-CoA are indicated by dashed lines. The figure was prepared with CCP4MG.

Fig. 6

Conservation of residues in the peptide binding pocket of LdNMT and HsNMT. Stereo view of the GLYASKLA peptide from the ternary ScNMT complex (1iid.pdb) displayed in the context of the LdNMT peptide binding groove. The LdNMT structure (displayed as a cyan ribbon) was superposed onto that of the yeast NMT using the SSM routine in CCP4MG and is displayed together with the GLYASKLA peptide (thick light green cylinders) from ScNMT. The peptide runs N to C, bottom to top in the figure. Side chains of LdNMT residues in the vicinity of the peptide are displayed in thin cylinder format and coloured in coral for residues that are conserved in HsNMT and in ball-and-stick format with atoms coloured by element for positions where the LdNMT and HsNMT sequences diverge. Selected residues are labelled in black at conserved positions and red at divergent positions. The NHM species has been omitted for clarity.