Crystal structure of Helicobacter pylori pseudaminic acid biosynthesis N-acetyltransferase PseH: implications for substrate specificity and catalysis

Abstract

Helicobacter pylori infection is the common cause of gastroduodenal diseases linked to a higher risk of the development of gastric cancer. Persistent infection requires functional flagella that are heavily glycosylated with 5,7-diacetamido-3,5,7,9-tetradeoxy-L-glycero-L-manno-nonulosonic acid (pseudaminic acid). Pseudaminic acid biosynthesis protein H (PseH) catalyzes the third step in its biosynthetic pathway, producing UDP-2,4-diacetamido-2,4,6-trideoxy-β-L-altropyranose. It belongs to the GCN5-related N-acetyltransferase (GNAT) superfamily. The crystal structure of the PseH complex with cofactor acetyl-CoA has been determined at 2.3 Å resolution. This is the first crystal structure of the GNAT superfamily member with specificity to UDP-4-amino-4,6-dideoxy-β-L-AltNAc. PseH is a homodimer in the crystal, each subunit of which has a central twisted β-sheet flanked by five α-helices and is structurally homologous to those of other GNAT superfamily enzymes. Interestingly, PseH is more similar to the GNAT enzymes that utilize amino acid sulfamoyl adenosine or protein as a substrate than a different GNAT-superfamily bacterial nucleotide-sugar N-acetyltransferase of the known structure, WecD. Analysis of the complex of PseH with acetyl-CoA revealed the location of the cofactor-binding site between the splayed strands β4 and β5. The structure of PseH, together with the conservation of the active-site general acid among GNAT superfamily transferases, are consistent with a common catalytic mechanism for this enzyme that involves direct acetyl transfer from AcCoA without an acetylated enzyme intermediate. Based on structural homology with microcin C7 acetyltransferase MccE and WecD, the Michaelis complex can be modeled. The model suggests that the nucleotide- and 4-amino-4,6-dideoxy-β-L-AltNAc-binding pockets form extensive interactions with the substrate and are thus the most significant determinants of substrate specificity. A hydrophobic pocket accommodating the 6'-methyl group of the altrose dictates preference to the methyl over the hydroxyl group and thus to contributes to substrate specificity of PseH.

Fig 1. The CMP-pseudaminic acid biosynthesis pathway in H. pylori [10].

Fig 2. The overall fold of H. pylori PseH.

(A) Stereo diagram of the structure of the PseH monomer. β-strands and α-helices are represented as arrows and coils and each element of the secondary structure is labeled and numbered as in text. The bound AcCoA molecule is shown in black. (B) The topology of secondary structure elements PseH. The α-helices are represented by rods and β-strands by arrows. Residue numbers are indicated at the start and end of each secondary structure element. (C) The molecular surface representation of PseH showing the AcCoA-binding tunnel between strands β4 and β5, which is a signature of the GNAT fold.

Fig 3. Comparisons of PseH with other GNAT superfamily enzymes.

(A) Stereo ribbon diagram of the superimposed structures of PseH from H. pylori (black), RimL from S. typhimurium (red) and the acetyltransferase domain of MccE from E. coli (green). The side chains of the conserved tyrosine in PseH and serine in MccE and RimL, likely to be implicated in deprotonation of the leaving thiolate anion of CoA in the reaction, are shown using a stick representation. (B) A sequence alignment of PseH, RimL, MccE and WecD from E. coli. The elements of the secondary structure and the sequence numbering for PseH are shown above the alignment. Conserved residues are highlighted in red. (C) Comparison of dimers observed in the crystal structures of PseH (in which the two halves of the dimer are drawn in black and grey) and RimL (red/salmon). (D) Comparison of the structures of PseH and WecD. Like PseH, WecD catalyses transfer of an acetyl group from AcCoA to the 4-amino moiety of the nucleotide-linked sugar substrate. Structurally equivalent domains are drawn in the same colour. The additional N-terminal domain in WecD is shown in yellow.

Fig 4. The stereoview of the electron density for AcCoA bound in the active site of PseH.

The cofactor molecule is shown in CPK representation and coloured according to atom type, with carbon atoms in orange, nitrogen in blue, oxygen in red, phosphorus in magenta and sulphur in yellow. Only the protein residues that form hydrogen bonds or van der Waals contacts with the cofactor molecule are shown for clarity. Protein carbon atoms are coloured black. The hydrogen bonds important for recognition of the cofactor are shown. The map was calculated at 2.3 Å resolution with coefficients |F_obs| − |F_calc| and phases from the final refined model with the coordinates of AcCoA deleted prior to one round of refinement. The map is contoured at 3.0-σ level.

Fig 5. The structural similarity between the nucleotide-binding pocket in MccE and the putative nucleotide-binding site in PseH.

The positions of the protein side-chains that form similar interactions with the nucleotide moiety of the substrate and with AcCoA are shown in a stick representation. The 3'-phosphate AMP moiety of CoA is omitted for clarity. (A) Key interactions between the protein and the nucleotide in the complex of the acetyltransferase domain of MccE with AcCoA and AMP. The protein backbone is shown as ribbon structure in light green for clarity of illustration. The AMP and AcCoA molecules are shown in ball-and-stick CPK representation and coloured according to atom type, with carbon atoms in black, nitrogen in blue, oxygen in red, phosphorus in magenta and sulphur in yellow. (B) The corresponding active-site residues in PseH and the docked model for the substrate UDP-4-amino-4,6-dideoxy-β-L-AltNAc. The protein backbone is shown as ribbon structure in light grey for clarity of illustration. AcCoA and modeled UDP-sugar are shown in ball-and-stick CPK representation and coloured according to atom type, with carbon atoms in black, nitrogen in blue, oxygen in red, phosphorus in magenta and sulphur in yellow.

Fig 6. Interactions between the docked substrate UDP-4-amino-4,6-dideoxy-β-L-AltNAc, acetyl moiety of the cofactor and protein residues in the active site of PseH in the modeled Michaelis complex.

The protein backbone is shown as ribbon structure in light grey for clarity of illustration. The substrate and AcCoA molecules are shown in ball-and-stick CPK representation and coloured according to atom type, with carbon atoms in black, nitrogen in blue, oxygen in red, phosphorus in magenta and sulphur in yellow. Only the protein side-chains that interact with the substrate are shown for clarity. The C4-N4 bond of the substrate (labeled) is positioned optimally for the direct nucleophilic attack on the thioester acetate, with the angle formed between the C4 of the amino-altrose, N4 of amino-altrose and the thioester carbonyl carbon being approximately 120°. The water molecule that is hydrogen bonded to the side-chains of Ser78 and Thr80, and is located within a hydrogen-bond distance of the 3’-hydroxyl of the modeled 4’-amino-altrose, is represented as a grey-blue ball. Deprotonation of the substrate’s amine group may occur via the 3’-hydroxyl of the altrose and this intervening water molecule.