High-resolution crystal structure of Streptococcus pyogenes β-NAD+ glycohydrolase in complex with its endogenous inhibitor IFS reveals a highly water-rich interface
aDepartment of Chemistry, College of Natural Sciences, Seoul National University, Seoul 151-747, Republic of Korea, bDepartment of Biophysics and Chemical Biology, Seoul National University, Seoul 151-747, Republic of Korea, and cCollege of Pharmacy, Seoul National University, Seoul 151-742, Republic of Korea
*Correspondence e-mail: firstname.lastname@example.org
One of the virulence factors produced by Streptococcus pyogenes is β-NAD+ glycohydrolase (SPN). S. pyogenes injects SPN into the cytosol of an infected host cell using the cytolysin-mediated translocation pathway. As SPN is toxic to bacterial cells themselves, S. pyogenes possesses the ifs gene that encodes an endogenous inhibitor for SPN (IFS). IFS is localized intracellularly and forms a complex with SPN. This intracellular complex must be dissociated during export through the cell envelope. To provide a structural basis for understanding the interactions between SPN and IFS, the complex was overexpressed between the mature SPN (residues 38–451) and the full-length IFS (residues 1–161), but it could not be crystallized. Therefore, limited proteolysis was used to isolate a crystallizable SPNct–IFS complex, which consists of the SPN C-terminal domain (SPNct; residues 193–451) and the full-length IFS. Its crystal structure has been determined by single anomalous diffraction and the model refined at 1.70 Å resolution. Interestingly, our high-resolution structure of the complex reveals that the interface between SPNct and IFS is highly rich in water molecules and many of the interactions are water-mediated. The wet interface may facilitate the dissociation of the complex for translocation across the cell envelope.
The gram-positive bacterium Streptococcus pyogenes causes a variety of human diseases such as superficial infections (pharyngitis and impetigo) and life-threatening conditions (toxic shock syndrome and necrotizing fasciitis) (Cunningham, 2000; Sachse et al., 2002). The virulence of S. pyogenes is enhanced by its toxin β-NAD+ glycohydrolase (SPN; also known as Nga) (Sumby et al., 2005). SPN possesses only β-NAD+ hydrolytic activity but no ADP-ribosyltransferase or ADP-ribosyl cyclase activity (Ghosh et al., 2010; Stevens et al., 2000). SPN is injected across the host cell membrane into the cytoplasm through streptolysin O (SLO), a member of a large family of pore-forming toxins and cholesterol-dependent cytolysins. Once translocated into the cytoplasm of the host cell, SPN contributes to virulence by depleting the intracellular NAD pool and producing the potent vasoactive compound nicotinamide, but not by ADP-ribosylation of protein substrates (Ghosh et al., 2010; Stevens et al., 2000; Bricker et al., 2005).
The β-NAD+ glycohydrolase hydrolyzes β-NAD+, an important cofactor in numerous redox and energy-producing biological reactions, to produce nicotinamide and adenosine diphosphoribose (ADP-ribose) (Tatsuno et al., 2007; Michos et al., 2006). Strict β-NAD+ glycohydrolases are incapable of further catalysis of the products from the initial reaction (Ghosh et al., 2010). Other enzymes that can hydrolyze β-NAD+ are multifunctional and can be classified into two on the basis of additional reactions (Ghosh et al., 2010). ADP-ribosyltransferases catalyse the hydrolysis of β-NAD+ and the transfer of an ADP-ribose moiety onto target proteins (Holbourn et al., 2006). ADP-ribosyl cyclases convert β-NAD+ into cyclic ADP-ribose, a potent second messenger for calcium mobilization (Karasawa et al., 1995).
SPN is comprised of two domains. The amino-terminal 190 residues of SPN are required for translocation of SPN into the host cell via cytolysin-mediated translocation (CMT) pathway (Ghosh & Caparon, 2006). The C-terminal domain (residues 191−451) alone is active as the β-NAD+ glycohydrolase but it is also indispensible for translocation (Ghosh & Caparon, 2006; Ghosh et al., 2010). SPN is also toxic to bacterial cells; therefore S. pyogenes encodes the ifs gene, which encodes the immunity factor for SPN (IFS) as an endogenous antitoxin (Meehl et al., 2005; Kimoto et al., 2006). IFS functions as a competitive inhibitor of the substrate β-NAD+ by blocking the active site in the C-terminal domain of SPN. It was reported that increasing concentrations of IFS decreased the rate of β-NAD+ hydrolysis, with KI,app of 2.0 ± 0.3 nM (Meehl et al., 2005). IFS is essential for the viability of SPN-producing strains of S. pyogenes by protecting the bacterium from the toxic β-NAD+ glycohydrolase activity of SPN that fail to be secreted (Meehl et al., 2005).
In order to provide structural details of the interactions of SPN with IFS and thus the inhibition mode, we have employed limited proteolysis to isolate a crystallizable complex of SPN and IFS, which consists of the SPN C-terminal domain (residues 193–451; SPNct) and the full-length IFS (residues 1–161). We have solved the crystal structure of this SPNct–IFS complex by single anomalous diffraction and refined the model at 1.70 Å resolution. The overall complex structure is highly similar to the previously reported structure that was refined at 2.80 Å (PDB entry 3pnt ; Smith et al., 2011). Much higher resolution of the data used in this study allowed us to identify a lot more water molecules bound to the protein complex, in particular at the interface between SPNct and IFS. Unexpectedly, our higher resolution structure reveals that the interface between SPNct and IFS is highly rich in water molecules, placing it as one of the wettest protein–protein interfaces. Many interactions between SPNct and IFS are water-mediated. As an example, the protruding Arg40 of IFS blocks the β-NAD+ binding site of SPNct through extensive water-mediated interactions. If the IFS-inhibited SPN has to dissociate from IFS before the free SPN is translocated across the cell envelope, the wet SPNct–IFS interface may facilitate such dissociation (Ghosh & Caparon, 2006).
The contiguous spn (SpyM3_0128) gene covering the residues 38–451 and the full-length ifs (SpyM3_0129) gene of S. pyogenes M3 were PCR-amplified, and cloned into the pET-28b(+) vector (Novagen), using the NdeI/XhoI restriction enzymes. This construct added a hexahistidine-containing 21-residue tag (MGSSHHHHHHSSGLVPRGSHM) at the N-terminus of SPN. The two proteins were co-expressed in Escherichia coli Rosetta2 (DE3) cells using Terrific Broth culture medium. Protein expression was induced by 0.5 mM isopropyl β-D-thiogalactopyranoside and the cells were incubated for an additional 18 h at 303 K following growth to mid-log phase at 310 K. The cells were lysed by sonication in a lysis buffer [20 mM Tris-HCl at pH 8.5, 500 mM NaCl, and 5% (v/v) glycerol] containing 5 mM imidazole followed by centrifugation to remove cellular debris. The supernatant was applied to an affinity chromatography column of HiTrap Chelating HP (GE Healthcare). The protein was eluted with the lysis buffer containing 300 mM imidazole and the eluted sample was further purified by size-exclusion chromatography using a HiLoad 16/60 Superdex 200 prep-grade column (GE Healthcare). The elution buffer was 20 mM Tris-HCl at pH 8.5, 200 mM NaCl and 0.1 mM tris(2-carboxyethyl)phosphine. We could confirm the complex formation of the two proteins by SDS-PAGE. However, we noticed that the 49 kDa band corresponding to SPN was degraded slowly. Thus, a limited proteolysis experiment was carried out to obtain a proteolysis-resistant core of the complex. After extensive testing of various combinations of proteases (trypsin and chymotrypsin) at different concentrations (at a mole ratio of 1:100, 1:1000 and 1:10000) and incubation time (30 min, 1 h, 3 h, 6 h and 20 h) and temperature (295 K and 310 K), the best condition was established to be α-chymotrypsin (Sigma catalog No. C4129) at a mole ratio of 1:1000 for 20 h at 310 K. After the α-chymotrypsin treatment, the complex was purified by size-exclusion chromatography using a HiLoad 16/60 Superdex 200 prep-grade column.
The selenomethionine (SeMet)-labeled complex protein was expressed and purified as above, except that we used the M9 cell culture medium that contained extra amino acids including SeMet.
The protein complex was concentrated to 50 mg ml−1 for crystallization using an Amicon Ultra-15 centrifugal filter unit (Millipore). Crystals were grown by sitting-drop vapor-diffusion method at 295 K. Each sitting drop prepared by mixing 1 µl each of the protein solution and the reservoir solution was placed over 100 µl of the reservoir solution. Best crystals of both SeMet-labeled and native SPNct–IFS complex were obtained with the reservoir solution of 20% (w/v) tacsimate at pH 4.0 and 20% (w/v) polyethylene glycol 3350. Crystals were transferred to a cryoprotectant solution, which contained 20% (v/v) glycerol in the reservoir solution. Single-wavelength anomalous diffraction (SAD) data were collected from a crystal of the SeMet-substituted SPNct–IFS complex at 100 K on an ADSC Quantum 315 CCD detector system (Area Detector Systems Corporation, Poway, CA, USA) at the BL-4A experimental station of Pohang Light Source, Korea. Raw data were processed using the program suit HKL2000 (Otwinowski & Minor, 1997). The crystal of SeMet-substituted SPNct–IFS complex belongs to the space group P1, with unit-cell parameters of a = 44.71 Å, b = 57.24 Å, c = 91.48 Å, α = 72.34°, β = 81.65° and γ = 79.49°. Native X-ray data were collected at 100 K on an ADSC Quantum 270 CCD detector system at the BL-7A of Pohang Light Source. The native crystal belongs to the space group P1, with a = 43.20 Å, b = 56.88 Å, c = 89.98 Å, α = 72.96°, β = 90.01° and γ = 82.27°. The presence of two molecules of the complex in the asymmetric unit gives a Matthew's parameter and solvent fraction of 2.17 Å3 Da−1 and 43.3%, respectively (Table 1).
‡Values obtained by treating Friedel pairs as separate observations.
§Rmerge = ΣhΣi|I(h)i − 〈I(h)〉|/ΣhΣiI(h)i, where I(h) is the intensity of reflection h, Σh is the sum over all reflections, and Σi is the sum over i measurements of reflection h.
¶R = Σ||Fobs| − |Fcalc||/Σ|Fobs|, where Rfree and Rwork are calculated for a randomly chosen 5% of reflections that were not used for refinement and for the remaining reflections, respectively.
††Values obtained using Refmac.
‡‡Values obtained using MolProbity.
The structure of SPNct–IFS complex was solved by Se SAD phasing. Phase calculation, density modification and initial model building were carried out using PHENIX AutoSol and AutoBuild (Adams et al., 2010). Phenix AutoSol located all 30 expected selenium atoms of two complex molecules in an asymmetric unit. Subsequent manual model building was conducted using the program COOT (Emsley & Cowtan, 2004) and the model was refined with the programs REFMAC (Murshudov et al., 1997) and PHENIX (Adams et al., 2010), including the bulk solvent correction. 5% of the data were randomly set aside as the test data for the calculation of Rfree (Brünger, 1992). Water molecules were added using the program COOT and were manually inspected. The quality of the refined model was assessed by MolProbity (Chen et al., 2010). Crystallographic and refinement statistics are summarized in Table 1. The coordinates and structure factors have been deposited in the Protein Data Bank (PDB) under the accession code 4kt6 .
We co-expressed the mature SPN (residues 38–451) and its endogenous inhibitor IFS (residues 1–161) from S. pyogenes but we could not crystallize the whole complex, because the SPN component was degraded slowly. Therefore, we optimized the condition of limited proteolysis to isolate a readily crystallizable SPNct–IFS complex. The SPNct–IFS complex consisted of C-terminal residues (193–451) of SPN and all residues (1–161) of IFS. Under our optimized proteolysis condition, the chymotrysin cleavages occurred only before Gly193. The loss of the SPN N-terminal region was supported by mass analysis of trypsin-digested peptide fragments of the denatured complex. Previously, the SPNct–IFS complex was co-expressed with the full-length IFS (residues 1–161) by identifying a C-terminal enzymatically active domain of SPN (residues 191–451) (Smith et al., 2011). The structure of the SPNct–IFS complex was determined at 2.80 Å (Smith et al., 2011), with the model accounting for residues 196–445 for both chains A and C of SPN, and residues 1–161 or 2–161 for chains B or D of IFS.
Our crystals of the purified SPNct–IFS complex diffracted to high resolution and allowed us to solve the structure by the Se SAD method. The model of the SPNct–IFS complex was refined to yield Rwork and Rfree values of 19.7% and 23.5%, respectively, for 20.0–1.70 Å data. The model includes 830 residues in two copies of the complex (residues 193–446 of SPN and residues 1–161 of IFS) and 596 water molecules. The C-terminal residues 447−451 of SPN (in both chains A and C) are likely disordered in the crystal. Two independent heterodimeric complexes in the P1 unit cell are highly similar to each other, with root-mean-square (r.m.s.) deviations of 0.27 Å for 415 Cα atoms (254 residues of SPNct and 161 residues of IFS) in the model. The two chains of SPNct in the asymmetric unit are highly similar to each other with an r.m.s. deviation of 0.22 Å for 254 Cα atoms; the two chains of IFS in the asymmetric unit are also highly similar to each other with an r.m.s. deviation of 0.22 Å for 161 Cα atoms.
Our 1.70 Å structure of the SPNct–IFS complex is highly similar to the previously reported structure determined at 2.80 Å (Smith et al., 2011), with r.m.s. deviations of 0.62–0.70 Å for 410 Cα atoms for each heterodimeric complex. The SPNct contains the NAD-binding catalytic domain consisting of an α/β fold (Smith et al., 2011). When we compare the SPNct part only between the two structures of the SPNct–IFS complex, the r.m.s. deviations are 0.51–0.55 Å for 250 Cα atoms. For the IFS part only, the r.m.s. deviations are 0.40–0.46 Å for 160 Cα atoms. When we compare the IFS model in our SPNct–IFS complex with the reported structure of the unbound IFS (PDB entry 3qb2 ; Smith et al., 2011), the r.m.s. deviations are 7.4–9.3 Å for 156–161 Cα atoms. Notable differences between them are found in the N-terminal loop (residues 1–6) with r.m.s. deviations of 20.7–22.1 Å and in the C-terminal region containing the SPN interaction loop 2 (SIL2; residues 139–149) and α7b (residues 150–161) with 15.6–19.7 Å. The N-terminal residues (1–6) of IFS are not involved in the interaction with SPN in the complex but they point in an opposite direction from those of the free IFS.
In our high-resolution structure of the SPNct–IFS complex, IFS interacts with SPNct through numerous hydrogen bonds and electrostatic interactions, many of which are water-mediated. The predominant contacts between SPNct and IFS involve an α1–α2 loop, α6, α6–β2 loop, α8–β3 loop and α9–β4 loop of SPNct; α1, α2–α3 loop, α5, α7a–α7b loop and α7b of IFS. The complex buries a large surface area at the interface between SPNct and IFS (3210 Å2 and 3280 Å2 for A:B and C:D interfaces, respectively). Our higher-resolution (1.70 Å) structure reveals that the interface is very rich in water molecules (Fig. 1); 67 and 71 water molecules are identified at the A:B and C:D interfaces, respectively. Many of these water molecules are conserved and common to both interfaces. The interface waters have B-factors ranging from 20.2 to 50.0 Å2 for the A:B interface and from 21.3 to 47.8 Å2 for C:D. The mean B-factor of interface waters (32.2 Å2 for the A:B interface and 33.0 Å2 for C:D) is slightly higher than that of non-hydrogen protein atoms (26.9 Å2 for A/B chains and 30.0 Å2 for C/D chains) but is lower than the overall B-factor of other waters (36.0 Å2). The previous complex structure was determined at 2.80 Å resolution (Smith et al., 2011) and it shows essentially identical buried surface areas at the interface (3260 Å2 and 3270 Å2 for A:B and C:D interfaces, respectively). However, only a small number of water molecules could be located due to insufficient resolution. A total of 153 water molecules were identified per two complex molecules in the asymmetric unit, with only 14 and 10 at the A:B and C:D interfaces, respectively (Fig. 1). Nearly all of these interface water molecules are present in our higher-resolution structure.
Water is often indispensable for specific recognition of two proteins as an integral part of protein–protein interfaces (Ladbury, 1996; Levy & Onuchic, 2004). Analyses of water molecules at the protein–protein interfaces showed that, on average, the interfaces of complexes and homodimers contain about ten water molecules per 1000 Å2 of interface area, and crystal packing interfaces, about 15 (Rodier et al., 2005; Reichmann et al., 2008). Moreover, interfaces of weak and highly transient complexes contain more waters than found in high-affinity complexes (Rodier et al., 2005; Reichmann et al., 2008). In our SPNct–IFS complex structure, ∼21 water molecules are found per 1000 Å2, making it one of the wettest protein–protein interfaces. The water-rich interface may be advantageous for facilitating the dissociation of IFS from the complex immediately before translocation across the cell envelope.
Our high-resolution crystal structure of the SPNct–IFS complex reveals that many interactions between SPNct and IFS are water-mediated. A prominent example is the α2–α3 loop of IFS, which points toward the NAD binding cavity of SPN. Compared with the unbound IFS structure (PDB entry 3qb2 ; Smith et al., 2011), the α2–α3 loop of IFS bound to SPNct is considerably moved toward the active site cavity of SPN in our SPNct–IFS complex, with an r.m.s. deviation of 0.52–0.94 Å for 20 Cα atoms. The side-chain of Arg40 on the IFS α2–α3 loop protrudes into the NAD binding cavity of SPNct (Fig. 2), blocking the binding of the substrate β-NAD+. Arg40 of IFS interacts with SPNct through extensive water-mediated interactions (Fig. 2). It makes an extensive water-mediated hydrogen-bond network with the residues located on α2, α8, α8–β3 loop and α9–β4 loop of SPNct (Gln216 on α2; Ile328 and Lys329 on α8; Gly330 and Asp332 on α8–β3 loop; Gly368, Asn370, Asn373, Ile374, Gln378, Thr379, Trp380, Glu389 and Glu391 on α9–β4 loop).
SPNct was shown to be an atypical member of the ADP-ribosyltransferase superfamily with the characteristic Arg/His (R/H) motif (His273) and the ADP-ribosylating turn–turn (ARTT) motif. The Ser-Thr-Ser (STS) motif of the ADP-ribosyltransferase superfamily is missing in SPN. It has an α-helical linker subdomain, which is absent in other ADP-ribosyltransferase superfamily enzymes (Smith et al., 2011). The ARTT motif is important for the substrate specificity and recognition of the ADP-ribosyltransferase superfamily (Han & Tainer, 2002). The Q/E–X–E sequence of the ARTT motif provides the key catalytic glutamic acid to stabilize an oxocarbenium ion intermediate (Han & Tainer, 2002; Holbourn et al., 2006). The second Gln or Glu (Q/E), located two positions upstream from the catalytic Glu in the ARTT loop, is essential for the ribosyltransferase activity of ADP-ribosylating toxins. It may be important for recognizing the target residue of substrate proteins (Nagahama et al., 2000; Wilde et al., 2002; Han et al., 2001). It was suggested that the different conformation of the ARTT loop in SPN as well as SPN's unique α-helical linker subdomain does not allow accommodation of protein substrates in the canonical mode of other ADP-ribosyltransfereases (Smith et al., 2011).
It appears that the distinct side-chain orientation of SPN Glu389 (Fig. 3), the second Q/E in the ARTT motif, is responsible for the lack of ADP-ribosyltransferase activity in SPN. In our present complex structure, as well as in the previously reported structure (Smith et al., 2011), the conformation of the ARTT loop of SPNct is considerably different from ADP-ribosyltransferases. Furthermore, the side-chain of Glu389 is stretched into the interior and is surrounded by α-helices α7 and α8 (Fig. 3b). The side-chain orientation of Glu389 is in an almost opposite direction from corresponding residues of ADP-ribosyltransferases (Fig. 3a). When we modeled an NAD molecule of B. cereus VIP2 (PDB entries 1qs2 ) into the active site of SPNct by superimposing the two structures, a water lies in the SPNct structure between Glu389 and the susceptible glycosidic bond of NAD+ (Fig. 3b). This water may be activated by Glu389 to act as a nucleophile for the hydrolytic reaction catalyzed by SPNct (Ghosh et al., 2010; Robertus et al., 1998). Unlike Glu389, the side-chain of the catalytic Glu391 of SPNct overlaps well with the corresponding residues of ADP-ribosyltransferases (Fig. 3a).
We thank the beamline staff at Pohang Light Source, Korea (BL-4A and BL-7A), and Photon Factory, Japan (BL-1A and BL-5A), for assistance during X-ray diffraction experiments. This work was supported by Korea Ministry of Education, Science and Technology, National Research Foundation (NRF) of Korea, Innovative Drug Research Center for Metabolic and Inflammatory Disease (grant No. R11-2007-107-00000-0), and Korea Healthcare Technology R&D Project, Ministry of Health, Welfare & Family Affairs, Republic of Korea (grant No. A092006). This research was also supported by the NRF grant funded by the Korean government (2011-0013663) to HJY.
Adams, P. D., Afonine, P. V., Bunkóczi, G., Chen, V. B., Davis, I. W., Echols, N., Headd, J. J., Hung, L.-W., Kapral, G. J., Grosse-Kunstleve, R. W., McCoy, A. J., Moriarty, N. W., Oeffner, R., Read, R. J., Richardson, D. C., Richardson, J. S., Terwilliger, T. C. & Zwart, P. H. (2010). Acta Cryst. D66, 213–221. Web of Science CrossRef CAS IUCr Journals
Bricker, A. L., Carey, V. J. & Wessels, M. R. (2005). Infect. Immun. 73, 6562–6566. Web of Science CrossRef PubMed CAS
Brünger, A. T. (1992). Nature (London), 355, 472–475. PubMed Web of Science
Chen, V. B., Arendall, W. B., Headd, J. J., Keedy, D. A., Immormino, R. M., Kapral, G. J., Murray, L. W., Richardson, J. S. & Richardson, D. C. (2010). Acta Cryst. D66, 12–21. Web of Science CrossRef CAS IUCr Journals
Cunningham, M. W. (2000). Clin. Microbiol. Rev. 13, 470–511. Web of Science CrossRef PubMed CAS
DeLano, W. L. (2002). PyMOL, http://www.pymol.org .
Emsley, P. & Cowtan, K. (2004). Acta Cryst. D60, 2126–2132. Web of Science CrossRef CAS IUCr Journals
Ghosh, J., Anderson, P. J., Chandrasekaran, S. & Caparon, M. G. (2010). J. Biol. Chem. 285, 5683–5694. Web of Science CrossRef CAS PubMed
Ghosh, J. & Caparon, M. G. (2006). Mol. Microbiol. 62, 1203–1214. Web of Science CrossRef PubMed CAS
Han, S., Arvai, A. S., Clancy, S. B. & Tainer, J. A. (2001). J. Mol. Biol. 305, 95–107. Web of Science CrossRef PubMed CAS
Han, S. & Tainer, J. A. (2002). Int. J. Med. Microbiol. 291, 523–529. CrossRef PubMed CAS
Holbourn, K. P., Shone, C. C. & Acharya, K. R. (2006). FEBS J. 273, 4579–4593. Web of Science CrossRef PubMed CAS
Karasawa, T., Takasawa, S., Yamakawa, K., Yonekura, H., Okamoto, H. & Nakamura, S. (1995). FEMS Microbiol. Lett. 130, 201–204. CrossRef CAS PubMed
Kimoto, H., Fujii, Y., Hirano, S., Yokota, Y. & Taketo, A. (2006). J. Biol. Chem. 281, 9181–9189. Web of Science CrossRef PubMed CAS
Ladbury, J. E. (1996). Chem. Biol. 3, 973–980. CrossRef CAS PubMed Web of Science
Levy, Y. & Onuchic, J. N. (2004). Proc. Natl Acad. Sci. USA, 101, 3325–3326. Web of Science CrossRef PubMed CAS
Meehl, M. A., Pinkner, J. S., Anderson, P. J., Hultgren, S. J. & Caparon, M. G. (2005). PLoS Pathog. 1, e35. CrossRef PubMed
Michos, A., Gryllos, I., Håkansson, A., Srivastava, A., Kokkotou, E. & Wessels, M. R. (2006). J. Biol. Chem. 281, 8216–8223. Web of Science CrossRef PubMed CAS
Murshudov, G. N., Vagin, A. A. & Dodson, E. J. (1997). Acta Cryst. D53, 240–255. CrossRef CAS Web of Science IUCr Journals
Nagahama, M., Sakaguchi, Y., Kobayashi, K., Ochi, S. & Sakurai, J. (2000). J. Bacteriol. 182, 2096–2103. Web of Science CrossRef PubMed CAS
Otwinowski, Z. & Minor, W. (1997). Methods Enzymol. 276, 307–326. CrossRef CAS Web of Science
Reichmann, D., Cohen, M., Abramovich, R., Dym, O., Lim, D., Strynadka, N. C. & Schreiber, G. (2007). J. Mol. Biol. 365, 663–679. Web of Science CrossRef PubMed CAS
Reichmann, D., Phillip, Y., Carmi, A. & Schreiber, G. (2008). Biochemistry, 47, 1051–1060. Web of Science CrossRef PubMed CAS
Robertus, J. D., Monzingo, A. F., Marcotte, E. M. & Hart, P. J. (1998). J. Exp. Zool. 282, 127–132. CrossRef CAS PubMed
Rodier, F., Bahadur, R. P., Chakrabarti, P. & Janin, J. (2005). Proteins, 60, 36–45. Web of Science CrossRef PubMed CAS
Sachse, S., Seidel, P., Gerlach, D., Günther, E., Rödel, J., Straube, E. & Schmidt, K. H. (2002). FEMS Immunol. Med. Microbiol. 34, 159–167. CrossRef PubMed CAS
Smith, C. L., Ghosh, J., Elam, J. S., Pinkner, J. S., Hultgren, S. J., Caparon, M. G. & Ellenberger, T. (2011). Structure, 19, 192–202. Web of Science CrossRef CAS PubMed
Stevens, D. L., Salmi, D. B., McIndoo, E. R. & Bryant, A. E. (2000). J. Infect. Dis. 182, 1117–1128. Web of Science CrossRef PubMed CAS
Sumby, P., Porcella, S. F., Madrigal, A. G., Barbian, K. D., Virtaneva, K., Ricklefs, S. M., Sturdevant, D. E., Graham, M. R., Vuopio-Varkila, J., Hoe, N. P. & Musser, J. M. (2005). J. Infect. Dis. 192, 771–782. Web of Science CrossRef PubMed CAS
Tatsuno, I., Sawai, J., Okamoto, A., Matsumoto, M., Minami, M., Isaka, M., Ohta, M. & Hasegawa, T. (2007). Microbiology, 153, 4253–4260. Web of Science CrossRef PubMed CAS
Wilde, C., Just, I. & Aktories, K. (2002). Biochemistry, 41, 1539–1544. Web of Science CrossRef PubMed CAS
This is an open-access article distributed under the terms of the Creative Commons Attribution (CC-BY) Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are cited.