- Email: [email protected]
X-ray structure of Clostridium perfringens sortase B cysteine transpeptidase
Biochemical and Biophysical Research Communications 493 (2017) 1267e1272
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
X-ray structure of Clostridium perfringens sortase B cysteine transpeptidase Eiji Tamai a, b, Hiroshi Sekiya a, Jun Maki a, Hirofumi Nariya c, Hiromi Yoshida b, Shigehiro Kamitori b, * a b c
Department of Infectious Disease, College of Pharmaceutical Science, Matsuyama University, 4-2 Bunkyo-cho, Matsuyama, Ehime 790-8578, Japan Life Science Research Center and Faculty of Medicine, Kagawa University, 1750-1 Ikenobe, Miki-cho, Kita-gun, Kagawa 761-0793, Japan Graduate School of Biosphere Science, Hiroshima University, 1-4-4 Kagamiyama-cho, Higashihiroshima, Hiroshima 739-8528, Japan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 22 September 2017 Accepted 26 September 2017 Available online 28 September 2017
The pathogenesis and infectivity of Gram-positive bacteria are mediated by many surface proteins that are covalently attached to peptidoglycans of the cell wall. The covalent attachment of these proteins is catalyzed by sortases (Srts), a family of cysteine transpeptidases, which are classified into six classes, A e F, based on their amino acid sequences and biological roles. Clostridium perfringens, one of the pathogenic clostridial species, has a class B sortase (CpSrtB) with 249 amino acid residues. X-ray structures of CpSrtB and its inactive mutant form were determined at 2.2 Å and 1.8 Å resolutions, respectively. CpSrtB adopts a typical sortase-protein fold, and has a unique substrate-binding groove formed by three b-strands and two helices creating the sidewalls of the groove. The position of the catalytic Cys232 of CpSrtB is significantly different from those commonly found in Srts structures. The modeling study of the CpSrtB/ peptide complex suggested that the position of Cys232 found in CpSrtB is preferable for the catalytic reaction to occur. Structural comparison with other class B sortases demonstrated that the catalytic site likely converts between two forms. The movement of Cys232 between the two forms may help His136 deprotonate Cys232 to be activated as a thiolate, which may the catalytic Cys-activated mechanism for Srts. © 2017 Elsevier Inc. All rights reserved.
Keywords: Bacterial cell wall Clostridium perfringens Sortase cysteine transpeptidase X-ray structure
1. Introduction Gram-positive bacteria possess a thick cell wall that surrounds their cytoplasmic membranes and provides physical protection. The bacterial cell wall is a mesh polymer of peptidoglycans, in which linear glycan backbones are cross-linked by species-specific peptide side chains [1,2]. The pathogenesis and infectivity of Grampositive bacteria are mediated by many surface proteins that are covalently attached to the peptidoglycan of the cell wall [3]. The covalent attachment of these proteins is catalyzed by sortases
Abbreviations: Srt, sortase; CWSS, C-terminal cell wall sorting signal motifs; SrtA, class A sortase; SrtB, class B sortase; SaSrtB, Staphylococcus aureus SrtB; Isd, iron-responsive surface determinant; BaSrtB, Bacillus anthracis SrtB; SpSrtB, Streptococcus pyogenes SrtB; SrtC, class C sortase; CdSrtB, Clostridium difficile SrtB; CpSrtB, Clostridium perfringens SrtB; CpSrtB_C232S, mutant form of CpSrtB with the replacement of Cys232 by Ser; CdSrtB_C209A, mutant form of CdSrtB with the replacement of Cys209 by Ala. * Corresponding author. E-mail address: [email protected] (S. Kamitori). https://doi.org/10.1016/j.bbrc.2017.09.144 0006-291X/© 2017 Elsevier Inc. All rights reserved.
(Srts), a family of cysteine transpeptidases [4,5]. Srts recognize and cleave the C-terminal cell wall sorting signal motifs (CWSS) with five amino acid residues of substrate proteins to form an acylenzyme adduct with the substrate protein. Subsequently, Srts catalyze the formation of amide bonds between the substrate protein and the amino group of the peptidoglycan of the cell wall (Supplemental Fig. S1A). Recently, Srts have attracted attention as potential therapeutic targets [6]. Srts are classified into six classes, A e F, based on their amino acid sequences and biological roles. The class A sortases (SrtAs) are found in many Gram-positive bacteria, and are well known to anchor a large number of proteins with the CWSS of LPXTG to the cell wall, referred to as housekeeping Srts. The other five classes of Srts carry out their biological functions with specific substrate proteins. The class B sortases (SrtBs) are found in low GC content Grampositive bacteria, including Bacillus anthracis, Clostridium difficile, Clostridium perfringens, Listeria monocytogenes, Staphylococcus aureus and Streptococcus pyogenes. Staphylococcus aureus SrtB (SaSrtB) is the best-characterized SrtB enzyme that anchors the
1268
E. Tamai et al. / Biochemical and Biophysical Research Communications 493 (2017) 1267e1272
heme-transport protein IsdC with the CWSS of NPQTN, which is a member of the iron-responsive surface determinant (Isd) system, as the primary heme-iron uptake pathway [7,8]. The X-ray structure of SaSrtB in complex with an analogue of NPQT (SaSrtB/NPQT*) (PDB ID: 4FLD) was reported, providing important clues about the CWSS-recognition mechanisms of SaSrtB [9]. Bacillus anthracis SrtB (BaSrtB) and Listeria monocytogenes SrtB are also involved in hemeacquisition, recognizing the CWSS of NPKTG and NPKSS/NAKTN, respectively [10,11]. Whereas, Streptococcus pyogenes SrtB (SpSrtB) was found to be involved in pilus assembly, which is usually catalyzed by the class C sortases (SrtCs) [12], and Clostridium difficile SrtB (CdSrtB) was demonstrated to anchor the putative adhesion protein CD0386 with the CWSS of SPKTG [13]. Clostridium perfringens is one of the pathogenic clostridial species that causes gas gangrene and food poisoning [14]. Clostridium perfringens SrtB (CpSrtB) (CPE0513) has 249 amino acid residues with 38% similarity with SaSrtB (Supplemental Fig. S1B), and its substrate protein is unknown. Clostridium perfringens has the putative heme-transport protein of CPE0221 with 24% similarity with Staphylococcus aureus IsdC, in which two CWSS-like sequences of SDETG and SEATG are at the C-terminal site. The preliminary biochemical characterization of CpSrtB with CPE0221 we performed revealed that CPE0221 is unlikely to be a substrate of CpSrtB (Supplemental Fig. S2). To identify the substrate protein and examine the biological function of CpSrtB, the three-dimensional structure of CpSrtB is useful. We here report X-ray structures of CpSrtB and its mutant form with the replacement of the catalytic Cys232 with Ser (CpSrtB_C232S), and the modeling study of the enzyme/substrate complex, providing new insights into the catalytic Cys-activated mechanism for Srts.
1 ml of protein solution (18.4 mg/ml in 20 mM Tris-HCl, pH 7.5) and 1 ml of reservoir solution (200 mM sodium nitrate, 20% (w/v) PEG3350, pH6.8) with 50 ml of the reservoir solution, using the sitting drop vapor diffusion method. Crystals of CpSrtB_C232S were prepared by using protein solution (30.2 mg/ml with 2 mM hexapeptide in 20 mM Tris-HCl, pH 7.5) and reservoir solution (200 mM ammonium phosphate monobasic, 20% (w/v) PEG3350, pH4.6). The synthesized hexapeptides of GSDETG and GSEATG were purchased (Eurofins Genomics, Tokyo, Japan). X-ray diffraction data for CpSrtB and CpSrtB_C232S were collected at 100 K using an ADSC QUANTUM 315R area detector system on the PF BL5A beam line in the KEK (Tsukuba, Japan). Diffraction data were processed using the program XDS [17], and the CCP4 program suite [18]. The data collection and scaling results are listed in Table 1. The initial phases of CpSrtB were obtained by molecular replacement with the program MOLREP [19] using the structure of a mutant form of CdSrtB with the replacement of Cys209 with Ala (CdSrtB_C209A, 4UX7) as a probe model [13]. Further model building was performed with the program Coot [20]. Using the structure of CpSrtB, the structure of CpSrtB_C232S was solved by molecular replacement, because the crystal systems are different from each other (Table 1). The structures were refined using the programs Refmac5 [21] and CNS [22]. Refinement statistics are listed in Table 1. Figures were drawn with the program PyMol [23]. 2.3. Modeling of the CpSrtB/APATG complex structure The detailed procedure Supplemental Fig. S3.
of
the
modeling
is
given
in
3. Results 2. Materials and methods 3.1. X-ray structure determination 2.1. Protein preparation In order to construct the expression vector for C-terminal Histagged and N-terminal transmembrane region-lacking CpSrtB, standard PCR was performed using the F-primer and R-primer listed in Supplemental Table S1, and genomic DNA of Clostridium perfringens strain 13 as a template. PCR products were digested with NdeI and XhoI, and then cloned into the expression vector, pColdAcdH (unpublished) which contains the XhoI-x6 His codon after the NdeI site of pColdIV (Takara Bio Inc., Shiga, Japan). The resultant plasmid was designated as pColdSrtBHDN. For the construction of the plasmids expressing the mutant SrtBHDN, overlap extension PCR [15] was performed using the F-primers and R-primers listed in Supplemental Table S1, and pColdSrtBHDN as a template. The PCR product was cloned into the vector in a similar manner. BL21CodonPlus-RIL were transformed with the constructed plasmids for protein expression. Protein expression was carried out as previously described [16]. The protein solution was applied to an affinity column (HisTrap™ HP, GE Healthcare Bio-Sciences Corp., Piscataway, NJ, USA) for purification. The column was washed with buffer A (50 mM Tris-HCl, 500 mM NaCl and 20 mM imidazole, pH 7.5), and the protein was eluted with a linear gradient of 20e400 mM imidazole in buffer A. The eluate from the resin was dialyzed with buffer B (20 mM TriseHCl, pH 7.5) and the dialysate was loaded onto a HiTrap Q HP column (GE Healthcare Bio-Sciences Corp., Piscataway, NJ, USA). The column was washed with buffer B, and the protein was eluted with a linear gradient of 0e500 mM NaCl in buffer B. The eluate from the resin was dialyzed with buffer B. 2.2. X-ray crystallography Crystals of CpSrtB were grown at 293 K in a droplet mixed with
CpSrtB without the N-terminal transmembrane peptide of 23 amino acid residues was crystallized in space group P21212, and the structure was refined to R-factor of 0.227 at 2.2 Å resolution. There are two molecules, Mol-A and Mol-B, in an asymmetric unit related by non-crystallographic 2-fold symmetry. In both Mol-A and Mol-B, the 43 amino acid residues at the N-terminal site were invisible in the electron density map. The refined structures of Mol-A and MolB have 183 amino acid residues each. In order to examine whether the protein was proteolytically cleaved or not, the protein sample for crystallization was sequenced, and the data demonstrated the protein to be intact. Although the preliminary biochemical characterization of CpSrtB with CPE0221 revealed that CPE0221 is unlikely to be a substrate of CpSrtB (Supplemental Fig. S2), co-crystallization with CWSS-like peptides of CPE0221, GSDETG and GSEATG, was performed using an inactive mutant, CpSrtB_C232S. Crystals were obtained in space group C2 with a molecule in an asymmetric unit, and the structure was refined to R-factor of 0.184 at 1.8 Å resolution. There was no electron density for the added peptides, and the cocrystallization of the protein/peptide complex produced no result. However, addition of the peptide may have affected the molecular packing in the crystals, causing the crystal to diffract well. The refined structure of CpSrtB_C232S has 211 amino acid residues (Tyr33 e Ser50, and Ser57 e Asp249) which is more than the 183 amino acid residues of CpSrtB. 3.2. Overall structure of CpSrtB_C232S The structures of Mol-A and Mol-B of CpSrtB, and CpSrtB_C232S are almost equivalent in their root-mean-square deviations of the main-chain atoms of 0.44e0.49 Å. The overall structure of
E. Tamai et al. / Biochemical and Biophysical Research Communications 493 (2017) 1267e1272
1269
Table 1 Data collection and refinement statistics.
Data collection Beamline Temperature (K) Wavelength (Å) Resolution range (Å) No. of measured refs. No. of unique refs. Redundancy Completeness (%) Mean Io/s(Io) Rmerge (%) Space group Unit cell parameters Refinement Resolution range (Å) No. of refs. Completeness (%) Rfactor (%) Rfree (%) RMSD bond lengths (Å) RMSD bond angles ( ) Ramachandran plot Most favoured region (%) Allowed region (%) B-factor (Å2) Protein Solvent PDB code
CpSrtB
CpSrtB_C232S
PF-BL5A 100 1.0 48.20e2.20 (2.26e2.20) 175,050 (12,460) 24,588 (1813) 7.1 (6.8) 99.8 (99.8) 19.29 (4.66) 6.3 (41.3) P21212 a ¼ 58.34, b ¼ 94.87, c ¼ 64.81 Å
PF-BL5A 100 1.0 47.99e1.80 (185e1.80) 72,087 (5384) 19,621 (1455) 3.7 (3.7) 99.7 (99.8) 20.6 (3.2) 3.9 (44.8) C2 a ¼ 74.67, b ¼ 59.73, c ¼ 56.22 Å b ¼ 121.39
48.20e2.20 (2.26e2.20) 24,588 (1813) 8.9 (100.0) 22.6 (28.2) 27.5 (34.7) 0.008 1.1
47.99e1.80 (1.85e1.80) 19,261 (1455) 99.7 (99.8) 18.4 (27.0) 22.8 (33.7) 0.008 1.0
86.6 13.4
87.8 12.2
41.8 43.4 5B23
30.3 34.5 5YFK
Rmerge ¼ Sh Si [jIi(h)j/Sh Si Ii(h)], where Ii is the ith measurement and is the weighted mean of all measurements of I(h). Values in parentheses are for the high resolution bin.
CpSrtB_C232S is shown in Fig. 1A with a topological diagram of secondary structure elements. CpSrtB_C232S adopts a typical sortase-protein fold with eight b-strands and seven helices. There is a distorted b-barrel at the center of the protein. The two a-helices (H1 and H2) form the a-bundle at the N-terminal site. The connected loop between H1 and H2 in CpSrtB_C232S is invisible, and this a-bundle, including two a-helices, is invisible in CpSrtB due to the highly disordered structure. A long loop including a short 310 helix (H3) is inserted between B2 and B3. The 310 helix (H4) and ahelix (H5) between B4 and B5 are across the b-barrel. The two ahelices (H6 and H7) between B6 and B7 are arranged in an antiparallel manner, contacting the b-barrel. The conserved catalytic residues His136, Ser232 (Cys) and Arg240 are located at the loop between B4 and H4, the end of B7, and the beginning of B8, respectively. A groove for substrate-binding is formed by three bstrands (B4, B7 and B8) and two helices (H3 and H6) creating the sidewalls of the groove. The X-ray structures of four SrtBs were reported; SaSrtB (1NG5, 4FLD) [9,24], BaSrtB (1RZ2) [24], SpSrtB (3PSQ) [12], CdSrtB (CdSrtB, 5GYJ) [25] and a mutant CdSrtB_C209A (4UX7) [13]. For structural comparison, the structures of CdSrtB_C209A, SaSrtB and BaSrtB are shown in Fig. 1B. As the structures of CdSrtB and SpSrtB have many missing loops, they were not included in the structural comparison. The central b-barrel with H4, H5 and H7 are well conserved among SrtB structures. H3 is conserved in CdSrtB_C209A, and it is as a 310 helix-like loop in SaSrtB and BaSrtB. H6 creating a sidewall in the substrate-binding groove is unique to CpSrtB, and the corresponding regions of the other SrtB structures are the extended loops. Structural variations are found in the N-terminal a-bundle. In CpSrtB_C232S, the a-helix (H1) with eight amino acid residues and the long a-helix (H2) with 17 amino acid residues forms the abundle, and six invisible amino acid residues between H1 and H2 are highly disordered. In the a-bundle, Tyr 40, Leu43 and Leu46 of
H1, and Leu71 and Ile74 of H2 form a large hydrophobic cluster with Pro92 of B2 and Phe107 of the loop between H3 and B3. In CdSrtB_C209A, the long a-helix (H1) is connected to the short 310 helix (H2) by two amino acid residues, forming a hydrophobic cluster with Leu58, Leu68, Ile71, Pro89 and Phe104. In SaSrtB, H1 and H2 are arranged in an anti-parallel manner, forming a hydrophobic cluster with Tyr39, Leu42, Phe46, Leu49, Phe61, Leu64, Ile66, Pro85 and Phe100. In BaSrtB, H1 and H2 are arranged in the same manner as in SaSrtB, forming a hydrophobic cluster with Met46, Ala49, Ile52, Tyr53, Phe71, Leu74, Ile77, Pro95 and Tyr110, and a long loop including a 310 helix connects H1 and H2.
3.3. Catalytic site structure and modeling of the CpSrtB/APATG complex The three catalytic residues of His136, Cys232 and Arg240 are located at the end of the substrate-binding groove, and are arranged side-by-side (Fig. 2A). Cys232 directs its Sg atom to the Nd atom of His136 at a distance of 4.7 Å, suggesting that they do not form a hydrogen bond. The position of the catalytic Cys residue of CpSrtB is significantly different from those in other SrtBs. The Cys residues of CdSrtB (Ala), SaSrtB and BaSrtB emerge toward the outside of the protein compared with Cys232 of CpSrtB with an average Ca-Ca deviation of 5.0 Å, but the positions of His and Arg residues are mostly conserved (Fig. 2A). Cys223(Sg) and His130(Nε) of SaSrtB do not form a hydrogen bond at a distance of 4.6 Å (Cys223 e His130) (Fig. 2A). This location of three catalytic residues is mostly found in Srt structures, and designated as Form 1. Cys232 of CpSrtB is depressed into the inside of the protein to form a dent on the molecular surface, which is designated as Form 2 (Fig. 2A and B). Form 2 was found in the structures of SpSrtB (3PSQ) [12], group B Streptococcus SrtC1 (3TB7) [27] and Streptococcus pneumonia cognate sortase (4Y4Q) [28]. The loop structures between
1270
E. Tamai et al. / Biochemical and Biophysical Research Communications 493 (2017) 1267e1272
Fig. 1. Overall structures of CpStrB and structural comparison with other SrtBs. (A) The overall structures of CpSrtB_C232S (left) and the topological diagram of the secondary elements (right) are illustrated. Three catalytic residues and hydrophobic residues involved in the formation of the hydrophobic cluster with H1 and H2 are shown with a stick model. The invisible loop between H1 and H2 is indicated with a dashed line. (B) The overall structures of CdSrtB_C209A, SaSrtB and BaSrtB are illustrated in the same manner as CpSrtB_C232S in (A).
catalytic Cys and Arg residues vary among CpSrtB, CdSrtB and SaSrtB. The loop region of BaSrtB was invisible in the X-ray structure (Fig. 2A). As the crystal of the CpSrtB/peptide complex was not obtained, the CpSrtB/peptide complex was modeled to examine the proteinsubstrate interactions. The sequence of APATG was used for the peptide in the modeling structure. The second Pro and fourth Thr residues are usually conserved in CWSS, and most Srts recognize the hydrolyzing site of Thr-Gly. To avoid unexpected proteinsubstrate interactions, Ala was assigned as the first and third residues. The modeling was carried out based on the structure of human apopain (cysteine protease) in complex with tetrapeptide DVAD (1CP3) [26] because the location of catalytic Cys and His residues in human apopain is very similar with that of CpSrtB in Form 2 (Supplemental Fig. S3). As shown in Fig. 2B, the peptide APATG was modeled in the substrate-binding groove of CpSrtB. Subsite numbers for amino acid units are defined as P10, P1, P2, P3 and P4, and the hydrolyzing site was between P1 and P1’ (Fig. 2B).
The modeled APATG was well fitted to the substrate-binding groove, in which Thr_P1 and Pro_P3 direct their side-chain groups to the bottom of the substrate-binding groove, whereas those of Ala_P2 and Ala_P4 face outward, suggesting that the amino acid residues with bulky side-chain groups possibly occupy the positions of P2 and P4. Cys232 directs its Sg atom to the C atom of Thr_P1 at a distance of 3.2 Å. His136 forms hydrogen bonds with the O atoms of Leu103 and Thr_P1. The side-chain group of Thr_P1 makes a hydrogen bond with Arg240 (Thr (Og) - Arg (Nε)) and hydrophobic interactions with Leu103. Pro_P3 forms hydrophobic interactions with Leu103. The amino acid residues at P2 and P4 may be recognized by His104, Lys184 and Glu187 directing their sidechains to P2 and/or P4. 4. Discussion The structural comparison of CpSrtB (CpSrtB_C232S) with other SrtBs demonstrated a significant difference in structure in the N-
E. Tamai et al. / Biochemical and Biophysical Research Communications 493 (2017) 1267e1272
1271
Fig. 2. Catalytic site structure of CpSrtB. (A) Catalytic site structures of CpSrtB (yellow), CdSrtB_C209A (cyan), SaSrtB (green) and BaSrtB (pink) are superimposed with three catalytic residues. (B) Substrate-binding groove of CpSrtB with the modeled peptide APATG is illustrated. The putative amino acid residues interacting with the bound modeled peptide are shown with a stick model. The loop between Cys232 and Arg240 is shown in magenta. (C) The proposed catalytic reaction mechanism for CpSrtB is shown. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
terminal a-bundle. This a-bundle is characteristic to SrtB, and was believed to determine the protein orientation on the cytoplasmic membrane [6]. Although the structures of the loops between H1 and H2 vary, the relative orientation of H1 with the b-barrel is almost equivalent due to formation of the conserved hydrophobic cluster by H1, H2, B2, and the loop between H3 and B3. (Fig. 1). H1 may dominantly determine the protein orientation, and the flexible loop between H1 and H2 may interact with the species-specific cell wall component to help the protein maintain proper orientation on the cytoplasmaic membrane. The structure of the substrate-binding groove of CpSrtB is unique in that H6 overlaps B7 and B8, making the substrate-binding groove narrower than in other SrtBs. Although the true substrate with the CWSS for CpSrtB remains unknown, the narrow substratebinding groove likely fits with the CWSS of the substrate, and H6 may be involved in substrate-recognition by Lys184 and/or Glu187. The structures and amino acid sequences of the loop between catalytic Cys and Arg residues vary among SrtBs (Fig. 2A and B, and Supplemental Fig. S1B). In the X-ray structure of SaSrtB in complex with the tripeptide GGG that is the peptidoglycan fragment of the Staphylococcus aureus cell wall, the loop recognized the GGG [29]. This loop may recognize the species-specific peptidoglycan
fragment to which the substrate protein attaches. The positions of the catalytic Cys residue differs between Form 1 and Form 2. In both forms, the Cys residue does not form a hydrogen bond with the His residue (Fig. 2A), suggesting that the His residue is unlikely to deprotonate and activate the Cys residue as a thiolate. Thus, a reverse protonation mechanism was proposed as the catalytic reaction mechanism for Srts by combined kinetic observations [30]. In this mechanism, a small proportion of the enzyme (ca. 0.06%) is in the active form with the deprotonated Cys residue and the protonated His residue. The thiolate of Cys232 nucleophilically attacks the carbonyl carbon of Thr_P1 to form a tetrahedral intermediate that is stabilized by Arg240, and His136 donates a proton to Gly_P10 to form an acyl-enzyme adduct. The amine leaving group is released from the enzyme and replaced by the amino group of the peptidoglycan of the cell wall, and then the reversal reactions occur to anchor the substrate protein to the cell wall. (Fig. 2C (2)e(4)). This mechanism is consistent with the model structure of the CpSrtB/APATG complex in Form 2. Cys232 directs its Sg atom to the C atom of Thr_P1 at a distance of 3.2 Å, and His136 and Arg240 are in close proximity with Gly_P10 and Thr_P1. In addition to these three catalytic residues, Leu103 plays important roles in determining the proper side-chain conformation of His136
1272
E. Tamai et al. / Biochemical and Biophysical Research Communications 493 (2017) 1267e1272
as an acid/base catalyst, and recognizing Thr_P1 and Pro_P3 by hydrophobic interactions. The X-ray structure of the SaSrtB/NPQT* complex (4FLD) in Form 1 revealed that the catalytic Cys223 formed a disulfide link with the Thr moiety of the peptide ligand to form an analogue of acyl-enzyme adduct, but that His130 is too far from Thr_P1 at a distance of 8.4 Å (His130(Nd) e Thr_P1(C)) (Supplemental Fig. S4) [9]. These results suggest that the position of Cys232 in Form 2 is preferable for the catalytic reaction to occur. As Form 1 and Form 2 were found in the X-ray structures of Srts, the catalytic site is expected to convert between two forms. Indeed, in the X-ray structure of SpSrtB (3PSQ), the catalytic sites of two molecules in an asymmetric unit adopted Form 1 and Form 2, respectively [12]. During the conversion between the two forms, Cys232 moves between His136 and Arg240, and it may form a hydrogen bond with His136 in the intermediate form (Fig. 2C (1)). Cys232 was found to form a hydrogen bond with His136 in the intermediate form modeled by setting the main-chain conformations of Ile230, Thr231 and Cys232 to the intermediate values between Form 1 and Form 2 (Supplemental Fig. S5). The movement of Cys232 between the two forms may help His136 deprotonate Cys232 to be activated as a thiolate, which may be the catalytic Cysactivated mechanism for Srts. Depending on substrate-binding, Srts may adopt Form 2 for the catalytic reaction to occur. Form 1 may be favorable for stabilizing an acyl-enzyme adduct with many proteinsubstrate interactions (Fig. 2 (4)). Funding This work was supported in part by JSPS KAKENHI, Grant Numbers JP15K06973 and JP15K08482, from the Japan Society for the Promotion of Science (JSPS). This research was performed with approval from the Photon Factory Advisory Committee and National Laboratory for High Energy Physics, Japan. Disclosure summary The authors have nothing to disclose. Acknowledgements The authors are grateful to Dr. A. Itoh for technical assistance. Transparency document Transparency document related to this article can be found online at https://doi.org/10.1016/j.bbrc.2017.09.144. Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.bbrc.2017.09.144. References [1] M. Leyh-Bouille, R. Bonaly, J.M. Ghuysen, R. Tinelli, D. Tipper, LL-diaminopimelic acid containing peptidoglycans in walls of Streptomyces sp. and of Clostridium perfringens (type A), Biochemistry 9 (1970) 2944e2952. [2] D.L. Popham, Visualizing the production and arrangement of peptidoglycan in Gram-positive cells, Mol. Microbiol. 88 (2013) 645e649. €o € k, Surface protein adhesins of Staphylococcus aureus, Trends [3] T.J. Foster, M. Ho
Microbiol. 6 (1998) 484e488. [4] T. Spirig, E.M. Weiner, R.T. Clubb, Sortase enzymes in Gram-positive bacteria, Mol. Microbiol. 82 (2011) 1044e1159. [5] O. Schneewind, D. Mihaylova-Petkov, P. Model, Cell wall sorting signals in surface proteins of gram-positive bacteria, EMBO J. 12 (1993) 4803e4811. [6] W.J. Bradshaw, A.H. Davies, C.J. Chambers, A.K. Roberts, C.C. Shone, K.R. Acharya, Molecular features of the sortase enzyme family, FEBS J. 282 (2015) 2097e2114. [7] S.K. Mazmanian, H. Ton-That, K. Su, O. Schneewind, An iron-regulated sortase anchors a class of surface protein during Staphylococcus aureus pathogenesis, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 2293e2298. [8] J.C. Grigg, G. Ukpabi, C.F. Gaudin, M.E. Murphy, Structural biology of heme binding in the Staphylococcus aureus Isd system, J. Inorg. Biochem. 104 (2010) 341e348. [9] A.W. Jacobitz, J. Wereszczynski, S.W. Yi, B.R. Amer, G.L. Huang, A.V. Nguyen, M.R. Sawaya, M.E. Jung, J.A. McCammon, R.T. Clubb, Structural and computational studies of the Staphylococcus aureus sortase B-substrate complex reveal a substrate-stabilized oxyanion hole, J. Biol. Chem. 289 (2014) 8891e8902. [10] A.W. Maresso, T.J. Chapa, O. Schneewind, Surface protein IsdC and sortase B are required for heme-iron scavenging of Bacillus anthracis, J. Bacteriol. 188 (2006) 8145e8152. [11] H. Bierne, C. Garandeau, M.G. Pucciarelli, C. Sabet, S. Newton, F. Garcia-del Portillo, P. Cossart, A. Charbit, Sortase B, a new class of sortase in Listeria monocytogenes, J. Bacteriol. 186 (2004) 1972e1982. [12] H.J. Kang, F. Coulibaly, T. Proft, E.N. Baker, Crystal structure of Spy0129, a Streptococcus pyogenes class B sortase involved in pilus assembly, PLoS One 6 (2011) e15969. [13] C.J. Chambers, A.K. Roberts, C.C. Shone, K.R. Acharya, Structure and function of a Clostridium difficile sortase enzyme, Sci. Rep. 5 (2015) 9449. [14] G.K. Adak, S.M. Long, S.J. O'Brien, Trends in indigenous foodborne disease and deaths, England and Wales: 1992 to 2000, Gut 51 (2000) 832e841. [15] S.N. Ho, H.D. Hunt, R.M. Horton, J.K. Pullen, L.R. Pease, Site-directed mutagenesis by overlap extension using the polymerase chain reaction, Gene 77 (1989) 51e59. [16] E. Tamai, H. Sekiya, E. Goda, N. Makihata, J. Maki, H. Yoshida, S. Kamitori, Structural and biochemical characterization of the Clostridium perfringens autolysin catalytic domain, FEBS Lett. 591 (2017) 231e239. [17] W. Kabsch, XDS, Acta Crystallogr. D. Biol. Crystallogr. 66 (2010) 125e132. [18] Collaborative computational project 4, the CCP4 suite: programs for protein crystallography, Acta Crystallogr. D. Biol. Crystallogr. 50 (1994) 760e763. [19] A. Vagin, A. Teplyakov, MOLREP: an automated program for molecular replacement, Acta Crystallogr. D. Biol. Crystallogr. 30 (1997) 1022e1025. [20] P. Emsley, K. Cowtan, Coot: model-building tools for molecular graphics, Acta Crystallogr. D. Biol. Crystallogr. 60 (2004) 2126e2132. [21] G.N. Murshudov, A.A. Vagin, E.J. Dodson, Refinement of macromolecular structures by the maximum-likelihood method, Acta Crystallogr. D. Biol. Crystallogr. 53 (1997) 240e255. [22] A.T. Brunger, Version 1.2 of the crystallography and NMR system, Nat. Protoc. 2 (2007) 2728e2733. [23] W.L. DeLano, The PyMOL Molecular Graphics System, DeLano Scientific, San Carlos, CA, USA, 2002. http://www.pymol.org. [24] R. Zhang, R. Wu, G. Joachimiak, S.K. Mazmanian, D.M. Missiakas, P. Gornicki, O. Schneewind, A. Joachimiak, Structures of sortase B from Staphylococcus aureus and Bacillus anthracis reveal catalytic amino acid triad in the active site, Structure 12 (2004) 1147e1156. [25] J.C. Yin, C.H. Fei, Y.C. Lo, Y.Y. Hsiao, J.C. Chang, J.C. Nix, Y.Y. Chang, L.W. Yang, I.H. Huang, S. Wang, Structural insights into substrate recognition by Clostridium difficile sortase, Front. Cell Infect. Microbiol. 6 (2016) eCollection 2016. [26] P.R.E. Mittl, S. Di Marco, J.F. Krebs, X. Bai, D.S. Karanewsky, J.P. Priestle, K.J. Tomaselli, M.G. Grutter, Structure of recombinant human CPP32 in complex with the tetrapeptide acetyl-Asp-Val-Ala-Asp fluoromethyl ketone, J. Biol. Chem. 272 (1997) 6539e6547. [27] B. Khare, Z.Q. Fu, I.K. Huang, H. Ton-That, S.V. Narayana, The crystal structure analysis of group B Streptococcus sortase C1: a model for the “lid” movement upon substrate binding, J. Mol. Biol. 414 (2011) 563e577. [28] M.M. Shaik, C. Lombardi, D. Maragno Trindade, D. Fenel, G. Schoehn, A.M.D. Guilmi, A. Dessen, Structural basis of pilus anchoring by the ancillary pilin RrgC of Streptococcus pneumonia, J. Biol. Chem. 289 (2014) 16988e16997. [29] Y. Zong, S.K. Mazmanian, O. Schneewind, S.V. Narayana, The structure of sortase B, a cysteine transpeptidase that tethers surface protein to the Staphylococcus aureus cell wall, Structure 12 (2004) 105e112. [30] B.A. Frankel, R.G. Kruger, D.E. Robinson, N.L. Kelleher, D.G. McCafferty, Staphylococcus aureus sortase transpeptidase SrtA: insight into the kinetic mechanism and evidence for a reverse protonation catalytic mechanism, Biochemistry 44 (2005) 11188e11200.
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
X-ray structure of Clostridium perfringens sortase B cysteine transpeptidase Eiji Tamai a, b, Hiroshi Sekiya a, Jun Maki a, Hirofumi Nariya c, Hiromi Yoshida b, Shigehiro Kamitori b, * a b c
Department of Infectious Disease, College of Pharmaceutical Science, Matsuyama University, 4-2 Bunkyo-cho, Matsuyama, Ehime 790-8578, Japan Life Science Research Center and Faculty of Medicine, Kagawa University, 1750-1 Ikenobe, Miki-cho, Kita-gun, Kagawa 761-0793, Japan Graduate School of Biosphere Science, Hiroshima University, 1-4-4 Kagamiyama-cho, Higashihiroshima, Hiroshima 739-8528, Japan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 22 September 2017 Accepted 26 September 2017 Available online 28 September 2017
The pathogenesis and infectivity of Gram-positive bacteria are mediated by many surface proteins that are covalently attached to peptidoglycans of the cell wall. The covalent attachment of these proteins is catalyzed by sortases (Srts), a family of cysteine transpeptidases, which are classified into six classes, A e F, based on their amino acid sequences and biological roles. Clostridium perfringens, one of the pathogenic clostridial species, has a class B sortase (CpSrtB) with 249 amino acid residues. X-ray structures of CpSrtB and its inactive mutant form were determined at 2.2 Å and 1.8 Å resolutions, respectively. CpSrtB adopts a typical sortase-protein fold, and has a unique substrate-binding groove formed by three b-strands and two helices creating the sidewalls of the groove. The position of the catalytic Cys232 of CpSrtB is significantly different from those commonly found in Srts structures. The modeling study of the CpSrtB/ peptide complex suggested that the position of Cys232 found in CpSrtB is preferable for the catalytic reaction to occur. Structural comparison with other class B sortases demonstrated that the catalytic site likely converts between two forms. The movement of Cys232 between the two forms may help His136 deprotonate Cys232 to be activated as a thiolate, which may the catalytic Cys-activated mechanism for Srts. © 2017 Elsevier Inc. All rights reserved.
Keywords: Bacterial cell wall Clostridium perfringens Sortase cysteine transpeptidase X-ray structure
1. Introduction Gram-positive bacteria possess a thick cell wall that surrounds their cytoplasmic membranes and provides physical protection. The bacterial cell wall is a mesh polymer of peptidoglycans, in which linear glycan backbones are cross-linked by species-specific peptide side chains [1,2]. The pathogenesis and infectivity of Grampositive bacteria are mediated by many surface proteins that are covalently attached to the peptidoglycan of the cell wall [3]. The covalent attachment of these proteins is catalyzed by sortases
Abbreviations: Srt, sortase; CWSS, C-terminal cell wall sorting signal motifs; SrtA, class A sortase; SrtB, class B sortase; SaSrtB, Staphylococcus aureus SrtB; Isd, iron-responsive surface determinant; BaSrtB, Bacillus anthracis SrtB; SpSrtB, Streptococcus pyogenes SrtB; SrtC, class C sortase; CdSrtB, Clostridium difficile SrtB; CpSrtB, Clostridium perfringens SrtB; CpSrtB_C232S, mutant form of CpSrtB with the replacement of Cys232 by Ser; CdSrtB_C209A, mutant form of CdSrtB with the replacement of Cys209 by Ala. * Corresponding author. E-mail address: [email protected] (S. Kamitori). https://doi.org/10.1016/j.bbrc.2017.09.144 0006-291X/© 2017 Elsevier Inc. All rights reserved.
(Srts), a family of cysteine transpeptidases [4,5]. Srts recognize and cleave the C-terminal cell wall sorting signal motifs (CWSS) with five amino acid residues of substrate proteins to form an acylenzyme adduct with the substrate protein. Subsequently, Srts catalyze the formation of amide bonds between the substrate protein and the amino group of the peptidoglycan of the cell wall (Supplemental Fig. S1A). Recently, Srts have attracted attention as potential therapeutic targets [6]. Srts are classified into six classes, A e F, based on their amino acid sequences and biological roles. The class A sortases (SrtAs) are found in many Gram-positive bacteria, and are well known to anchor a large number of proteins with the CWSS of LPXTG to the cell wall, referred to as housekeeping Srts. The other five classes of Srts carry out their biological functions with specific substrate proteins. The class B sortases (SrtBs) are found in low GC content Grampositive bacteria, including Bacillus anthracis, Clostridium difficile, Clostridium perfringens, Listeria monocytogenes, Staphylococcus aureus and Streptococcus pyogenes. Staphylococcus aureus SrtB (SaSrtB) is the best-characterized SrtB enzyme that anchors the
1268
E. Tamai et al. / Biochemical and Biophysical Research Communications 493 (2017) 1267e1272
heme-transport protein IsdC with the CWSS of NPQTN, which is a member of the iron-responsive surface determinant (Isd) system, as the primary heme-iron uptake pathway [7,8]. The X-ray structure of SaSrtB in complex with an analogue of NPQT (SaSrtB/NPQT*) (PDB ID: 4FLD) was reported, providing important clues about the CWSS-recognition mechanisms of SaSrtB [9]. Bacillus anthracis SrtB (BaSrtB) and Listeria monocytogenes SrtB are also involved in hemeacquisition, recognizing the CWSS of NPKTG and NPKSS/NAKTN, respectively [10,11]. Whereas, Streptococcus pyogenes SrtB (SpSrtB) was found to be involved in pilus assembly, which is usually catalyzed by the class C sortases (SrtCs) [12], and Clostridium difficile SrtB (CdSrtB) was demonstrated to anchor the putative adhesion protein CD0386 with the CWSS of SPKTG [13]. Clostridium perfringens is one of the pathogenic clostridial species that causes gas gangrene and food poisoning [14]. Clostridium perfringens SrtB (CpSrtB) (CPE0513) has 249 amino acid residues with 38% similarity with SaSrtB (Supplemental Fig. S1B), and its substrate protein is unknown. Clostridium perfringens has the putative heme-transport protein of CPE0221 with 24% similarity with Staphylococcus aureus IsdC, in which two CWSS-like sequences of SDETG and SEATG are at the C-terminal site. The preliminary biochemical characterization of CpSrtB with CPE0221 we performed revealed that CPE0221 is unlikely to be a substrate of CpSrtB (Supplemental Fig. S2). To identify the substrate protein and examine the biological function of CpSrtB, the three-dimensional structure of CpSrtB is useful. We here report X-ray structures of CpSrtB and its mutant form with the replacement of the catalytic Cys232 with Ser (CpSrtB_C232S), and the modeling study of the enzyme/substrate complex, providing new insights into the catalytic Cys-activated mechanism for Srts.
1 ml of protein solution (18.4 mg/ml in 20 mM Tris-HCl, pH 7.5) and 1 ml of reservoir solution (200 mM sodium nitrate, 20% (w/v) PEG3350, pH6.8) with 50 ml of the reservoir solution, using the sitting drop vapor diffusion method. Crystals of CpSrtB_C232S were prepared by using protein solution (30.2 mg/ml with 2 mM hexapeptide in 20 mM Tris-HCl, pH 7.5) and reservoir solution (200 mM ammonium phosphate monobasic, 20% (w/v) PEG3350, pH4.6). The synthesized hexapeptides of GSDETG and GSEATG were purchased (Eurofins Genomics, Tokyo, Japan). X-ray diffraction data for CpSrtB and CpSrtB_C232S were collected at 100 K using an ADSC QUANTUM 315R area detector system on the PF BL5A beam line in the KEK (Tsukuba, Japan). Diffraction data were processed using the program XDS [17], and the CCP4 program suite [18]. The data collection and scaling results are listed in Table 1. The initial phases of CpSrtB were obtained by molecular replacement with the program MOLREP [19] using the structure of a mutant form of CdSrtB with the replacement of Cys209 with Ala (CdSrtB_C209A, 4UX7) as a probe model [13]. Further model building was performed with the program Coot [20]. Using the structure of CpSrtB, the structure of CpSrtB_C232S was solved by molecular replacement, because the crystal systems are different from each other (Table 1). The structures were refined using the programs Refmac5 [21] and CNS [22]. Refinement statistics are listed in Table 1. Figures were drawn with the program PyMol [23]. 2.3. Modeling of the CpSrtB/APATG complex structure The detailed procedure Supplemental Fig. S3.
of
the
modeling
is
given
in
3. Results 2. Materials and methods 3.1. X-ray structure determination 2.1. Protein preparation In order to construct the expression vector for C-terminal Histagged and N-terminal transmembrane region-lacking CpSrtB, standard PCR was performed using the F-primer and R-primer listed in Supplemental Table S1, and genomic DNA of Clostridium perfringens strain 13 as a template. PCR products were digested with NdeI and XhoI, and then cloned into the expression vector, pColdAcdH (unpublished) which contains the XhoI-x6 His codon after the NdeI site of pColdIV (Takara Bio Inc., Shiga, Japan). The resultant plasmid was designated as pColdSrtBHDN. For the construction of the plasmids expressing the mutant SrtBHDN, overlap extension PCR [15] was performed using the F-primers and R-primers listed in Supplemental Table S1, and pColdSrtBHDN as a template. The PCR product was cloned into the vector in a similar manner. BL21CodonPlus-RIL were transformed with the constructed plasmids for protein expression. Protein expression was carried out as previously described [16]. The protein solution was applied to an affinity column (HisTrap™ HP, GE Healthcare Bio-Sciences Corp., Piscataway, NJ, USA) for purification. The column was washed with buffer A (50 mM Tris-HCl, 500 mM NaCl and 20 mM imidazole, pH 7.5), and the protein was eluted with a linear gradient of 20e400 mM imidazole in buffer A. The eluate from the resin was dialyzed with buffer B (20 mM TriseHCl, pH 7.5) and the dialysate was loaded onto a HiTrap Q HP column (GE Healthcare Bio-Sciences Corp., Piscataway, NJ, USA). The column was washed with buffer B, and the protein was eluted with a linear gradient of 0e500 mM NaCl in buffer B. The eluate from the resin was dialyzed with buffer B. 2.2. X-ray crystallography Crystals of CpSrtB were grown at 293 K in a droplet mixed with
CpSrtB without the N-terminal transmembrane peptide of 23 amino acid residues was crystallized in space group P21212, and the structure was refined to R-factor of 0.227 at 2.2 Å resolution. There are two molecules, Mol-A and Mol-B, in an asymmetric unit related by non-crystallographic 2-fold symmetry. In both Mol-A and Mol-B, the 43 amino acid residues at the N-terminal site were invisible in the electron density map. The refined structures of Mol-A and MolB have 183 amino acid residues each. In order to examine whether the protein was proteolytically cleaved or not, the protein sample for crystallization was sequenced, and the data demonstrated the protein to be intact. Although the preliminary biochemical characterization of CpSrtB with CPE0221 revealed that CPE0221 is unlikely to be a substrate of CpSrtB (Supplemental Fig. S2), co-crystallization with CWSS-like peptides of CPE0221, GSDETG and GSEATG, was performed using an inactive mutant, CpSrtB_C232S. Crystals were obtained in space group C2 with a molecule in an asymmetric unit, and the structure was refined to R-factor of 0.184 at 1.8 Å resolution. There was no electron density for the added peptides, and the cocrystallization of the protein/peptide complex produced no result. However, addition of the peptide may have affected the molecular packing in the crystals, causing the crystal to diffract well. The refined structure of CpSrtB_C232S has 211 amino acid residues (Tyr33 e Ser50, and Ser57 e Asp249) which is more than the 183 amino acid residues of CpSrtB. 3.2. Overall structure of CpSrtB_C232S The structures of Mol-A and Mol-B of CpSrtB, and CpSrtB_C232S are almost equivalent in their root-mean-square deviations of the main-chain atoms of 0.44e0.49 Å. The overall structure of
E. Tamai et al. / Biochemical and Biophysical Research Communications 493 (2017) 1267e1272
1269
Table 1 Data collection and refinement statistics.
Data collection Beamline Temperature (K) Wavelength (Å) Resolution range (Å) No. of measured refs. No. of unique refs. Redundancy Completeness (%) Mean Io/s(Io) Rmerge (%) Space group Unit cell parameters Refinement Resolution range (Å) No. of refs. Completeness (%) Rfactor (%) Rfree (%) RMSD bond lengths (Å) RMSD bond angles ( ) Ramachandran plot Most favoured region (%) Allowed region (%) B-factor (Å2) Protein Solvent PDB code
CpSrtB
CpSrtB_C232S
PF-BL5A 100 1.0 48.20e2.20 (2.26e2.20) 175,050 (12,460) 24,588 (1813) 7.1 (6.8) 99.8 (99.8) 19.29 (4.66) 6.3 (41.3) P21212 a ¼ 58.34, b ¼ 94.87, c ¼ 64.81 Å
PF-BL5A 100 1.0 47.99e1.80 (185e1.80) 72,087 (5384) 19,621 (1455) 3.7 (3.7) 99.7 (99.8) 20.6 (3.2) 3.9 (44.8) C2 a ¼ 74.67, b ¼ 59.73, c ¼ 56.22 Å b ¼ 121.39
48.20e2.20 (2.26e2.20) 24,588 (1813) 8.9 (100.0) 22.6 (28.2) 27.5 (34.7) 0.008 1.1
47.99e1.80 (1.85e1.80) 19,261 (1455) 99.7 (99.8) 18.4 (27.0) 22.8 (33.7) 0.008 1.0
86.6 13.4
87.8 12.2
41.8 43.4 5B23
30.3 34.5 5YFK
Rmerge ¼ Sh Si [jIi(h)
CpSrtB_C232S is shown in Fig. 1A with a topological diagram of secondary structure elements. CpSrtB_C232S adopts a typical sortase-protein fold with eight b-strands and seven helices. There is a distorted b-barrel at the center of the protein. The two a-helices (H1 and H2) form the a-bundle at the N-terminal site. The connected loop between H1 and H2 in CpSrtB_C232S is invisible, and this a-bundle, including two a-helices, is invisible in CpSrtB due to the highly disordered structure. A long loop including a short 310 helix (H3) is inserted between B2 and B3. The 310 helix (H4) and ahelix (H5) between B4 and B5 are across the b-barrel. The two ahelices (H6 and H7) between B6 and B7 are arranged in an antiparallel manner, contacting the b-barrel. The conserved catalytic residues His136, Ser232 (Cys) and Arg240 are located at the loop between B4 and H4, the end of B7, and the beginning of B8, respectively. A groove for substrate-binding is formed by three bstrands (B4, B7 and B8) and two helices (H3 and H6) creating the sidewalls of the groove. The X-ray structures of four SrtBs were reported; SaSrtB (1NG5, 4FLD) [9,24], BaSrtB (1RZ2) [24], SpSrtB (3PSQ) [12], CdSrtB (CdSrtB, 5GYJ) [25] and a mutant CdSrtB_C209A (4UX7) [13]. For structural comparison, the structures of CdSrtB_C209A, SaSrtB and BaSrtB are shown in Fig. 1B. As the structures of CdSrtB and SpSrtB have many missing loops, they were not included in the structural comparison. The central b-barrel with H4, H5 and H7 are well conserved among SrtB structures. H3 is conserved in CdSrtB_C209A, and it is as a 310 helix-like loop in SaSrtB and BaSrtB. H6 creating a sidewall in the substrate-binding groove is unique to CpSrtB, and the corresponding regions of the other SrtB structures are the extended loops. Structural variations are found in the N-terminal a-bundle. In CpSrtB_C232S, the a-helix (H1) with eight amino acid residues and the long a-helix (H2) with 17 amino acid residues forms the abundle, and six invisible amino acid residues between H1 and H2 are highly disordered. In the a-bundle, Tyr 40, Leu43 and Leu46 of
H1, and Leu71 and Ile74 of H2 form a large hydrophobic cluster with Pro92 of B2 and Phe107 of the loop between H3 and B3. In CdSrtB_C209A, the long a-helix (H1) is connected to the short 310 helix (H2) by two amino acid residues, forming a hydrophobic cluster with Leu58, Leu68, Ile71, Pro89 and Phe104. In SaSrtB, H1 and H2 are arranged in an anti-parallel manner, forming a hydrophobic cluster with Tyr39, Leu42, Phe46, Leu49, Phe61, Leu64, Ile66, Pro85 and Phe100. In BaSrtB, H1 and H2 are arranged in the same manner as in SaSrtB, forming a hydrophobic cluster with Met46, Ala49, Ile52, Tyr53, Phe71, Leu74, Ile77, Pro95 and Tyr110, and a long loop including a 310 helix connects H1 and H2.
3.3. Catalytic site structure and modeling of the CpSrtB/APATG complex The three catalytic residues of His136, Cys232 and Arg240 are located at the end of the substrate-binding groove, and are arranged side-by-side (Fig. 2A). Cys232 directs its Sg atom to the Nd atom of His136 at a distance of 4.7 Å, suggesting that they do not form a hydrogen bond. The position of the catalytic Cys residue of CpSrtB is significantly different from those in other SrtBs. The Cys residues of CdSrtB (Ala), SaSrtB and BaSrtB emerge toward the outside of the protein compared with Cys232 of CpSrtB with an average Ca-Ca deviation of 5.0 Å, but the positions of His and Arg residues are mostly conserved (Fig. 2A). Cys223(Sg) and His130(Nε) of SaSrtB do not form a hydrogen bond at a distance of 4.6 Å (Cys223 e His130) (Fig. 2A). This location of three catalytic residues is mostly found in Srt structures, and designated as Form 1. Cys232 of CpSrtB is depressed into the inside of the protein to form a dent on the molecular surface, which is designated as Form 2 (Fig. 2A and B). Form 2 was found in the structures of SpSrtB (3PSQ) [12], group B Streptococcus SrtC1 (3TB7) [27] and Streptococcus pneumonia cognate sortase (4Y4Q) [28]. The loop structures between
1270
E. Tamai et al. / Biochemical and Biophysical Research Communications 493 (2017) 1267e1272
Fig. 1. Overall structures of CpStrB and structural comparison with other SrtBs. (A) The overall structures of CpSrtB_C232S (left) and the topological diagram of the secondary elements (right) are illustrated. Three catalytic residues and hydrophobic residues involved in the formation of the hydrophobic cluster with H1 and H2 are shown with a stick model. The invisible loop between H1 and H2 is indicated with a dashed line. (B) The overall structures of CdSrtB_C209A, SaSrtB and BaSrtB are illustrated in the same manner as CpSrtB_C232S in (A).
catalytic Cys and Arg residues vary among CpSrtB, CdSrtB and SaSrtB. The loop region of BaSrtB was invisible in the X-ray structure (Fig. 2A). As the crystal of the CpSrtB/peptide complex was not obtained, the CpSrtB/peptide complex was modeled to examine the proteinsubstrate interactions. The sequence of APATG was used for the peptide in the modeling structure. The second Pro and fourth Thr residues are usually conserved in CWSS, and most Srts recognize the hydrolyzing site of Thr-Gly. To avoid unexpected proteinsubstrate interactions, Ala was assigned as the first and third residues. The modeling was carried out based on the structure of human apopain (cysteine protease) in complex with tetrapeptide DVAD (1CP3) [26] because the location of catalytic Cys and His residues in human apopain is very similar with that of CpSrtB in Form 2 (Supplemental Fig. S3). As shown in Fig. 2B, the peptide APATG was modeled in the substrate-binding groove of CpSrtB. Subsite numbers for amino acid units are defined as P10, P1, P2, P3 and P4, and the hydrolyzing site was between P1 and P1’ (Fig. 2B).
The modeled APATG was well fitted to the substrate-binding groove, in which Thr_P1 and Pro_P3 direct their side-chain groups to the bottom of the substrate-binding groove, whereas those of Ala_P2 and Ala_P4 face outward, suggesting that the amino acid residues with bulky side-chain groups possibly occupy the positions of P2 and P4. Cys232 directs its Sg atom to the C atom of Thr_P1 at a distance of 3.2 Å. His136 forms hydrogen bonds with the O atoms of Leu103 and Thr_P1. The side-chain group of Thr_P1 makes a hydrogen bond with Arg240 (Thr (Og) - Arg (Nε)) and hydrophobic interactions with Leu103. Pro_P3 forms hydrophobic interactions with Leu103. The amino acid residues at P2 and P4 may be recognized by His104, Lys184 and Glu187 directing their sidechains to P2 and/or P4. 4. Discussion The structural comparison of CpSrtB (CpSrtB_C232S) with other SrtBs demonstrated a significant difference in structure in the N-
E. Tamai et al. / Biochemical and Biophysical Research Communications 493 (2017) 1267e1272
1271
Fig. 2. Catalytic site structure of CpSrtB. (A) Catalytic site structures of CpSrtB (yellow), CdSrtB_C209A (cyan), SaSrtB (green) and BaSrtB (pink) are superimposed with three catalytic residues. (B) Substrate-binding groove of CpSrtB with the modeled peptide APATG is illustrated. The putative amino acid residues interacting with the bound modeled peptide are shown with a stick model. The loop between Cys232 and Arg240 is shown in magenta. (C) The proposed catalytic reaction mechanism for CpSrtB is shown. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
terminal a-bundle. This a-bundle is characteristic to SrtB, and was believed to determine the protein orientation on the cytoplasmic membrane [6]. Although the structures of the loops between H1 and H2 vary, the relative orientation of H1 with the b-barrel is almost equivalent due to formation of the conserved hydrophobic cluster by H1, H2, B2, and the loop between H3 and B3. (Fig. 1). H1 may dominantly determine the protein orientation, and the flexible loop between H1 and H2 may interact with the species-specific cell wall component to help the protein maintain proper orientation on the cytoplasmaic membrane. The structure of the substrate-binding groove of CpSrtB is unique in that H6 overlaps B7 and B8, making the substrate-binding groove narrower than in other SrtBs. Although the true substrate with the CWSS for CpSrtB remains unknown, the narrow substratebinding groove likely fits with the CWSS of the substrate, and H6 may be involved in substrate-recognition by Lys184 and/or Glu187. The structures and amino acid sequences of the loop between catalytic Cys and Arg residues vary among SrtBs (Fig. 2A and B, and Supplemental Fig. S1B). In the X-ray structure of SaSrtB in complex with the tripeptide GGG that is the peptidoglycan fragment of the Staphylococcus aureus cell wall, the loop recognized the GGG [29]. This loop may recognize the species-specific peptidoglycan
fragment to which the substrate protein attaches. The positions of the catalytic Cys residue differs between Form 1 and Form 2. In both forms, the Cys residue does not form a hydrogen bond with the His residue (Fig. 2A), suggesting that the His residue is unlikely to deprotonate and activate the Cys residue as a thiolate. Thus, a reverse protonation mechanism was proposed as the catalytic reaction mechanism for Srts by combined kinetic observations [30]. In this mechanism, a small proportion of the enzyme (ca. 0.06%) is in the active form with the deprotonated Cys residue and the protonated His residue. The thiolate of Cys232 nucleophilically attacks the carbonyl carbon of Thr_P1 to form a tetrahedral intermediate that is stabilized by Arg240, and His136 donates a proton to Gly_P10 to form an acyl-enzyme adduct. The amine leaving group is released from the enzyme and replaced by the amino group of the peptidoglycan of the cell wall, and then the reversal reactions occur to anchor the substrate protein to the cell wall. (Fig. 2C (2)e(4)). This mechanism is consistent with the model structure of the CpSrtB/APATG complex in Form 2. Cys232 directs its Sg atom to the C atom of Thr_P1 at a distance of 3.2 Å, and His136 and Arg240 are in close proximity with Gly_P10 and Thr_P1. In addition to these three catalytic residues, Leu103 plays important roles in determining the proper side-chain conformation of His136
1272
E. Tamai et al. / Biochemical and Biophysical Research Communications 493 (2017) 1267e1272
as an acid/base catalyst, and recognizing Thr_P1 and Pro_P3 by hydrophobic interactions. The X-ray structure of the SaSrtB/NPQT* complex (4FLD) in Form 1 revealed that the catalytic Cys223 formed a disulfide link with the Thr moiety of the peptide ligand to form an analogue of acyl-enzyme adduct, but that His130 is too far from Thr_P1 at a distance of 8.4 Å (His130(Nd) e Thr_P1(C)) (Supplemental Fig. S4) [9]. These results suggest that the position of Cys232 in Form 2 is preferable for the catalytic reaction to occur. As Form 1 and Form 2 were found in the X-ray structures of Srts, the catalytic site is expected to convert between two forms. Indeed, in the X-ray structure of SpSrtB (3PSQ), the catalytic sites of two molecules in an asymmetric unit adopted Form 1 and Form 2, respectively [12]. During the conversion between the two forms, Cys232 moves between His136 and Arg240, and it may form a hydrogen bond with His136 in the intermediate form (Fig. 2C (1)). Cys232 was found to form a hydrogen bond with His136 in the intermediate form modeled by setting the main-chain conformations of Ile230, Thr231 and Cys232 to the intermediate values between Form 1 and Form 2 (Supplemental Fig. S5). The movement of Cys232 between the two forms may help His136 deprotonate Cys232 to be activated as a thiolate, which may be the catalytic Cysactivated mechanism for Srts. Depending on substrate-binding, Srts may adopt Form 2 for the catalytic reaction to occur. Form 1 may be favorable for stabilizing an acyl-enzyme adduct with many proteinsubstrate interactions (Fig. 2 (4)). Funding This work was supported in part by JSPS KAKENHI, Grant Numbers JP15K06973 and JP15K08482, from the Japan Society for the Promotion of Science (JSPS). This research was performed with approval from the Photon Factory Advisory Committee and National Laboratory for High Energy Physics, Japan. Disclosure summary The authors have nothing to disclose. Acknowledgements The authors are grateful to Dr. A. Itoh for technical assistance. Transparency document Transparency document related to this article can be found online at https://doi.org/10.1016/j.bbrc.2017.09.144. Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.bbrc.2017.09.144. References [1] M. Leyh-Bouille, R. Bonaly, J.M. Ghuysen, R. Tinelli, D. Tipper, LL-diaminopimelic acid containing peptidoglycans in walls of Streptomyces sp. and of Clostridium perfringens (type A), Biochemistry 9 (1970) 2944e2952. [2] D.L. Popham, Visualizing the production and arrangement of peptidoglycan in Gram-positive cells, Mol. Microbiol. 88 (2013) 645e649. €o € k, Surface protein adhesins of Staphylococcus aureus, Trends [3] T.J. Foster, M. Ho
Microbiol. 6 (1998) 484e488. [4] T. Spirig, E.M. Weiner, R.T. Clubb, Sortase enzymes in Gram-positive bacteria, Mol. Microbiol. 82 (2011) 1044e1159. [5] O. Schneewind, D. Mihaylova-Petkov, P. Model, Cell wall sorting signals in surface proteins of gram-positive bacteria, EMBO J. 12 (1993) 4803e4811. [6] W.J. Bradshaw, A.H. Davies, C.J. Chambers, A.K. Roberts, C.C. Shone, K.R. Acharya, Molecular features of the sortase enzyme family, FEBS J. 282 (2015) 2097e2114. [7] S.K. Mazmanian, H. Ton-That, K. Su, O. Schneewind, An iron-regulated sortase anchors a class of surface protein during Staphylococcus aureus pathogenesis, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 2293e2298. [8] J.C. Grigg, G. Ukpabi, C.F. Gaudin, M.E. Murphy, Structural biology of heme binding in the Staphylococcus aureus Isd system, J. Inorg. Biochem. 104 (2010) 341e348. [9] A.W. Jacobitz, J. Wereszczynski, S.W. Yi, B.R. Amer, G.L. Huang, A.V. Nguyen, M.R. Sawaya, M.E. Jung, J.A. McCammon, R.T. Clubb, Structural and computational studies of the Staphylococcus aureus sortase B-substrate complex reveal a substrate-stabilized oxyanion hole, J. Biol. Chem. 289 (2014) 8891e8902. [10] A.W. Maresso, T.J. Chapa, O. Schneewind, Surface protein IsdC and sortase B are required for heme-iron scavenging of Bacillus anthracis, J. Bacteriol. 188 (2006) 8145e8152. [11] H. Bierne, C. Garandeau, M.G. Pucciarelli, C. Sabet, S. Newton, F. Garcia-del Portillo, P. Cossart, A. Charbit, Sortase B, a new class of sortase in Listeria monocytogenes, J. Bacteriol. 186 (2004) 1972e1982. [12] H.J. Kang, F. Coulibaly, T. Proft, E.N. Baker, Crystal structure of Spy0129, a Streptococcus pyogenes class B sortase involved in pilus assembly, PLoS One 6 (2011) e15969. [13] C.J. Chambers, A.K. Roberts, C.C. Shone, K.R. Acharya, Structure and function of a Clostridium difficile sortase enzyme, Sci. Rep. 5 (2015) 9449. [14] G.K. Adak, S.M. Long, S.J. O'Brien, Trends in indigenous foodborne disease and deaths, England and Wales: 1992 to 2000, Gut 51 (2000) 832e841. [15] S.N. Ho, H.D. Hunt, R.M. Horton, J.K. Pullen, L.R. Pease, Site-directed mutagenesis by overlap extension using the polymerase chain reaction, Gene 77 (1989) 51e59. [16] E. Tamai, H. Sekiya, E. Goda, N. Makihata, J. Maki, H. Yoshida, S. Kamitori, Structural and biochemical characterization of the Clostridium perfringens autolysin catalytic domain, FEBS Lett. 591 (2017) 231e239. [17] W. Kabsch, XDS, Acta Crystallogr. D. Biol. Crystallogr. 66 (2010) 125e132. [18] Collaborative computational project 4, the CCP4 suite: programs for protein crystallography, Acta Crystallogr. D. Biol. Crystallogr. 50 (1994) 760e763. [19] A. Vagin, A. Teplyakov, MOLREP: an automated program for molecular replacement, Acta Crystallogr. D. Biol. Crystallogr. 30 (1997) 1022e1025. [20] P. Emsley, K. Cowtan, Coot: model-building tools for molecular graphics, Acta Crystallogr. D. Biol. Crystallogr. 60 (2004) 2126e2132. [21] G.N. Murshudov, A.A. Vagin, E.J. Dodson, Refinement of macromolecular structures by the maximum-likelihood method, Acta Crystallogr. D. Biol. Crystallogr. 53 (1997) 240e255. [22] A.T. Brunger, Version 1.2 of the crystallography and NMR system, Nat. Protoc. 2 (2007) 2728e2733. [23] W.L. DeLano, The PyMOL Molecular Graphics System, DeLano Scientific, San Carlos, CA, USA, 2002. http://www.pymol.org. [24] R. Zhang, R. Wu, G. Joachimiak, S.K. Mazmanian, D.M. Missiakas, P. Gornicki, O. Schneewind, A. Joachimiak, Structures of sortase B from Staphylococcus aureus and Bacillus anthracis reveal catalytic amino acid triad in the active site, Structure 12 (2004) 1147e1156. [25] J.C. Yin, C.H. Fei, Y.C. Lo, Y.Y. Hsiao, J.C. Chang, J.C. Nix, Y.Y. Chang, L.W. Yang, I.H. Huang, S. Wang, Structural insights into substrate recognition by Clostridium difficile sortase, Front. Cell Infect. Microbiol. 6 (2016) eCollection 2016. [26] P.R.E. Mittl, S. Di Marco, J.F. Krebs, X. Bai, D.S. Karanewsky, J.P. Priestle, K.J. Tomaselli, M.G. Grutter, Structure of recombinant human CPP32 in complex with the tetrapeptide acetyl-Asp-Val-Ala-Asp fluoromethyl ketone, J. Biol. Chem. 272 (1997) 6539e6547. [27] B. Khare, Z.Q. Fu, I.K. Huang, H. Ton-That, S.V. Narayana, The crystal structure analysis of group B Streptococcus sortase C1: a model for the “lid” movement upon substrate binding, J. Mol. Biol. 414 (2011) 563e577. [28] M.M. Shaik, C. Lombardi, D. Maragno Trindade, D. Fenel, G. Schoehn, A.M.D. Guilmi, A. Dessen, Structural basis of pilus anchoring by the ancillary pilin RrgC of Streptococcus pneumonia, J. Biol. Chem. 289 (2014) 16988e16997. [29] Y. Zong, S.K. Mazmanian, O. Schneewind, S.V. Narayana, The structure of sortase B, a cysteine transpeptidase that tethers surface protein to the Staphylococcus aureus cell wall, Structure 12 (2004) 105e112. [30] B.A. Frankel, R.G. Kruger, D.E. Robinson, N.L. Kelleher, D.G. McCafferty, Staphylococcus aureus sortase transpeptidase SrtA: insight into the kinetic mechanism and evidence for a reverse protonation catalytic mechanism, Biochemistry 44 (2005) 11188e11200.