Numbers in parenthesis correspond to the highest resolution shell.
Structure of the bacillus subtilis YYCN protein: A putative N-acetyltransferase
Article first published online: 22 OCT 2003
Copyright © 2003 Wiley-Liss, Inc.
Proteins: Structure, Function, and Bioinformatics
Volume 53, Issue 4, pages 950–952, 1 December 2003
How to Cite
Taneja, B., Maar, S., Shuvalova, L., Collart, F., Anderson, W. and Mondragón, A. (2003), Structure of the bacillus subtilis YYCN protein: A putative N-acetyltransferase. Proteins, 53: 950–952. doi: 10.1002/prot.10561
- Issue published online: 17 NOV 2003
- Article first published online: 22 OCT 2003
- Manuscript Accepted: 12 JUN 2003
- Manuscript Received: 10 JUN 2003
- National Institute of Health. Grant Number: GM62414
YYCN is a conserved hypothetical protein from Bacillus subtilis that belongs to Pfam domain PF00583. It was selected for study because it did not show significant sequence similarity to proteins of known three-dimensional (3D) structures and may contain a previously unobserved fold. The crystal structure of the YYCN protein was determined at 2.2 Å resolution as part of the ongoing Midwest Structural Genomics initiative (http://www.mcsg.anl.gov/). One of the major aims of the initiative is to try to complete the remaining repertoire of protein folds.
Materials and Methods.
Cloning, protein expression, and purification:
The open reading frame (ORF) of YYCN was amplified, cloned, and the protein was purified as described.1, 2 However, the His-tag and the TEV protease cleavage site were retained, because the protein was not processed with TEV protease after purification over a Nickel column. Hence, the protein used for crystallization contained 24 additional residues at the N-terminus.
Protein crystallization, data collection and structure determination:
Initial crystallization trials with Seleno-methionine derivatized YYCN were done using Screen 1 and Screen 2 of Hampton Research. Diffraction-quality crystals were grown at 22°C using the sitting drop vapor diffusion method, by mixing 3 μL of 5 mg/mL YYCN with 3 μL of reservoir solution (900 mM ammonium phosphate and 190 mM NaCl in 100 mM imidazole buffer, pH 8.0). Crystals were flash-frozen in liquid nitrogen with 25% glycerol in the crystallization buffer as a cryoprotectant.
Crystals were monoclinic with unit cell parameters, a = 87.3 Å, b = 88.6 Å, c = 55.8 Å, β = 98.8°, and contained two molecules in the asymmetric unit. Two-wavelength multiple-wavelength anomalous dispersion (MAD) data were collected at beam line 5ID-B at the Dupont Northwestern Dow Collaborative Access Team at the Advanced Photon Source in Argonne National Laboratory (Table I). Phasing calculations were carried out using SOLVE3 and SHARP.4 Solvent flattening with the refined phases yielded good-quality maps suitable for automatic model building of 76% of amino acid residues by RESOLVE.5 The remaining structure was built manually using O.6 Iterative cycles of model building and refinement using REFMAC57 resulted in an R-factor of 0.233 and Rfree of 0.259. Atomic coordinates have been deposited into the Protein Data Bank (PDB) with ID 1UFH.
|Peak energy||High energy|
|Wavelength||0.9792 Å||0.9565 Å|
|Resolution (Å)a||26.08–2.20 (2.32–2.20)||26.26–2.20 (2.32–2.20)|
|Number of unique reflections||21,192||21,206|
|Completeness (%)||99.3 (99.3)||99.4 (99.1)|
|Data redundancy||3.8 (3.5)||3.8 (3.5)|
|Rsym (%)b||8.5 (23.8)||8.9 (26.1)|
|I/σ(I)||6.8 (2.6)||6.5 (2.4)|
|No. of protein atoms||2543|
|No. of water atoms||81|
|RMSD bond lengths||0.012 Å|
|RMSD bond angles||1.26°|
|Most favored regions (%)||95.2|
|Disallowed regions (%)||0|
The structure of YYCN contains a mixed α/β fold [Fig. 1(a)]. The overall structure consists of a 5-stranded β-sheet forming the core of the structure, with strands 1 to 4 antiparallel to one another, while strand 5 runs in a parallel direction to strand 4. The central β-sheet core is flanked on either side by two α-helices. A small cavity is formed where the C-termini of parallel strands 4 and 5 diverge and is enclosed by helix 1 and the N-termini of helices 3 and 4. The asymmetric unit contains two molecules of YYCN, but they are not related by a simple 180° rotation. The residues of the YYCN sequence observed in the structure include residues 2 through 156 in one subunit and residues 3 to 156 in the other subunit. No noncrystallographic symmetry restraints were applied during refinement. The root-mean-square deviation (RMSD) between the Cα atoms for all residues of the two subunits is 0.7 Å, indicating that although the two molecules are similar, they are not identical. Deviations between the two molecules occur in the loop regions connecting different secondary structure elements, for instance, the loop between the second and the third β-strands, the loop between the third and the fourth strands, and the loop connecting the fifth strand and the third α-helix. The additional residues corresponding to the his-tag and the TEV site at the N-terminus were disordered in both subunits, and no density was observed for them.
A search using the DALI server8 (http://www2.ebi.ac.uk/dali/) for similar folds revealed several structural homologues for YYCN. Despite very low sequence identity (<20%), significant structural similarities were found with several members of the GCN5-related N-acetyltransferase (GNAT) superfamily,9 with a Z score between 12.2 and 5.8 [Fig. 1(b)]. The closest structural homologue was the yeast Hpa2 acetyltransferase10 (PDB ID: 1QSM, Z score = 12.2, RMSD = 1.9 Å for 107 equivalent Cα atoms and an overall 14% sequence identity). Strong structural similarities to other acetyltransferases were observed as well: For example, (1) histone acetyltransferase (HAT) domain of yeast GCN511 (PDB ID: 1YGH, Z score = 10.2, RMSD = 1.8 Å for 83 equivalent Cα atoms, and an overall sequence identity of 13%), (2) Enterococcus faecium aminoglycoside 6′-N-acetyltransferase12 (PDB ID: 1B87, Z score = 10.0, RMSD = 1.8 Å for 92 equivalent Cα atoms, and sequence identity = 11%), and (3) sheep serotonin N-acetyl transferase13 (PDB ID: 1CJW, Z score = 9.6, RMSD = 1.4 Å for 81 equivalent Cα atoms, and sequence identity = 12%). The comparison of YYCN with the other members of the GNAT family revealed a very similar core domain, with a central β-sheet flanked by α-helices [Fig. 1(b)]. This suggests a possible acetyltransferase function for YYCN similar to the other structural homologues. Although no potential substrate or cofactor was observed to be bound in the C-terminal cavity of YYCN, the cavity appears to correspond to the cofactor-binding site of acetyltransferases and may be important for substrate and/or cofactor binding. YYCN showed a poor conservation of residues in this putative active site when compared to the structurally related proteins. However, tyrosine140 of YYCN was found to be present at a structurally equivalent position to the proposed catalytic tyrosine of GNAT acetyltransferases: Tyr139 in Hpa2 acetyltransferase,10 Tyr168 in serotonin N-acetyltransferase,13 and Tyr143 in glucosamine-6-phosphate N-acetyltransferase.14
In the observed crystal structure, YYCN is a dimer. Analysis with the PITA server (http://www.ebi.ac.uk/thornton-srv/databases/pita) suggested that the observed interactions were consistent with crystal packing rather than an oligomeric protein. Another recently available structure of YYCN in the PDB (ID: 1ON0) suggests a tetrameric form for the protein. To investigate the oligomeric state of the protein in solution, a gel filtration column was used to determine the apparent molecular weight by comparison with known standards. The gel filtration experiment indicated the oligomeric form to be a dimer (data not shown). A structural homologue of YYCN, yeast Hpa2, has been shown to be present as a dimer in solution but forms a stable tetramer in the presence of its cofactor, acetyl CoA.10 A similar ligand-induced oligomerization might occur with YYCN. However, since the dimer arrangement of the present structure has a very weak interface between the two subunits and does not match any pair of subunits of the other crystal form of YYCN (1ON0), the structural arrangement might be a crystallographic artifact. It is not clear whether either of the two structures resembles the true biological dimer.
In conclusion, the structure of YYCN protein of B. subtilis indicates that despite low sequence similarity with acetyltransferases, the protein belongs to the previously observed GNAT protein fold for acetyltransferases. Comparison of the structure with other members of the GNAT family, along with conservation of Tyr140, suggests that YYCN may act as an acetyltransferase. However, the nature of the substrate or its biological role is not clear.
Our thanks to Justina Lee for assistance in the expression and purification of YYCN, and to Andrey Krasilnikov and George Minasov for help and suggestions. Portions of this work were performed at the DuPont–Northwestern–Dow Collaborative Access Team (DND-CAT) Synchrotron Research Center at the Advanced Photon Source (APS). DND-CAT is supported by DuPont, Dow, and the NSF, and use of the APS is supported by the DOE.