structural communications
Structure of Rhodococcus equi virulence-associated protein B (VapB) reveals an eight-stranded antiparallel β-barrel consisting of two Greek-key motifs
aDepartment of Chemistry, Bielefeld University, Universitaetsstrasse 25, 33615 Bielefeld, Germany, and bInstitute for Cell Biology, University of Bonn, Ulrich-Haberland Strasse 61a, 53121 Bonn, Germany
*Correspondence e-mail: hartmut.niemann@uni-bielefeld.de
Members of the virulence-associated protein (Vap) family from the pathogen Rhodococcus equi regulate virulence in an unknown manner. They do not share recognizable sequence homology with any protein of known structure. VapB and VapA are normally associated with isolates from pigs and horses, respectively. To contribute to a molecular understanding of Vap function, the of a protease-resistant VapB fragment was determined at 1.4 Å resolution. The structure was solved by SAD phasing employing the anomalous signal of one endogenous S atom and two bound Co ions with low occupancy. VapB is an eight-stranded antiparallel β-barrel with a single helix. Structural similarity to avidins suggests a potential binding function. Unlike other eight- or ten-stranded β-barrels found in avidins, bacterial outer membrane proteins, fatty-acid-binding proteins and lysozyme inhibitors, Vaps do not have a next-neighbour arrangement but consist of two Greek-key motifs with strand order 41238567, suggesting an unusual or even unique topology.
Keywords: antiparallel β-barrel; Greek-key motif; Vap protein family; single-wavelength anomalous dispersion; Rhodococcus equi.
PDB reference: R. equi VapB, 4cv7
1. Introduction
Rhodococcus equi is a Gram-positive facultative intracellular pathogen mainly of young foals, but also of other animal species and of human AIDS patients. R. equi is closely related to mycobacteria such as Mycobacterium tuberculosis and causes life-threatening pulmonary and extra-pulmonary infections (Vázquez-Boland et al., 2013; von Bargen & Haas, 2009).
Epidemiological studies have shown that infection of different animal species is closely coupled to the possession of certain types of virulence-associated plasmids by the infecting bacteria. The plasmid type typically found in equine isolates mediates production of the immunodominant secreted protein VapA (virulence-associated protein A; Sekizaki et al., 1995). Porcine isolates typically produce the allelic variant VapB (Takai et al., 2000). vapA and vapB have never been described together in a single bacterial isolate. VapA production is absolutely required for virulence phenotypes in a mouse infection model and in murine macrophages (Jain et al., 2003) and it is also a prerequisite for the neutralization of phagosome and lysosome pH in infected macrophages (von Bargen et al., 2009). An additional ten vap genes and three vap pseudo-genes identified on the vapA- and vapB-encoding virulence plasmids seem to have only a minor role (if any) in virulence (Jain et al., 2003). The amino-acid sequence similarities between all Vap proteins according to ClustalW are typically in the 30–50% identity range, with the 78% identity score between VapA and VapB being particularly high (Letek et al., 2008).
The precise mechanisms of how Vaps support R. equi virulence is unknown, particularly as there is apparently no protein to be found in current amino-acid sequence databases that has a meaningful sequence identity to VapA and for which the function has been annotated. We reasoned that establishing the three-dimensional structure of a Vap protein would possibly hint at the putative function(s) of the Vap family. Here, we present the at 1.4 Å resolution of the 12 kDa C-terminal protease-resistant core of recombinantly produced VapB.
2. Materials and methods
2.1. Protein expression and purification
Mature R. equi VapB (residues 36–197 lacking the signal sequence) was generated by PCR from the sequenced virulence-associated plasmid of strain PAM 1593 (Letek et al., 2008). The PCR primers were constructed in such a way that they generated an N-terminally His6-tagged VapB protein when cloned into the pETite N-His Kan vector (BioCat, Heidelberg, Germany) by `in-fusion reaction', according to the manufacturer's manual. The primers were VapB_his-6_forward, 5′-CATCATCACCACCATCACGTGCTGGATTCCGGAGGCGGC-3′, and VapB_his-6_reverse, 5′-GTGGCGGCCGCTCTATTATTATGCAACCTCCCAGTTGTG-3′. Protein expression was induced in Escherichia coli BL21 (DE3) cells at an OD600 of 0.76 with 1 mM IPTG for 4 h at 37°C. Protein purification was carried out at 4°C. The cell pellet from 2 l of bacterial culture was resuspended in 40 ml lysis buffer (50 mM NaH2PO4, 300 mM NaCl, 10 mM imidazole pH 7) plus one EDTA-free cOmplete protease-inhibitor tablet (Roche) and 3 µl DNase I (40 mg ml−1) and lysed by three passes through a French pressure cell. The lysate was cleared by centrifugation and the supernatant was applied onto 10 ml Ni–NTA Sepharose equilibrated in lysis buffer. The resin was washed with 3 × 30 ml lysis buffer (pH 7) containing 10 mM, 20 mM or 40 mM imidazole. The target protein was eluted with lysis buffer including 500 mM imidazole (pH 7). The was dialyzed against 2 l phosphate-buffered saline (PBS) twice. The yield was roughly 80 mg protein per litre of bacterial culture. For crystallization, the protein was digested with proteinase K (Roth, product No. 7528.1). 40 mg of VapB were digested with 80 µg proteinase K (500:1 protein:protease mass ratio) in 40 ml PBS for 90 min at 37°C. The reaction was stopped with 400 µl PMSF (phenylmethylsulfonyl fluoride) (17 mg ml−1 in ethanol; 100 mM) on ice for 20 min and then the solution was dialyzed against 1 l buffer A (25 mM Tris, 20 mM NaCl pH 7), plus 1 ml PMSF (100 mM in ethanol) overnight with a protein recovery of about 50% (20.6 mg). Proteinase-K-digested VapB was further purified by anion-exchange over Source Q (GE Healthcare) resin equilibrated in buffer A. The protein was eluted by a salt gradient to 1 M NaCl. SDS–PAGE analysis revealed that proteinase K digestion had resulted in two fragments of very similar molecular mass, of which the slightly heavier fragment eluted earlier. Early peak fractions were pooled, dialyzed against buffer A overnight, concentrated to 16.8 mg ml−1 using Vivaspin concentrators (Sartorius) and frozen in small aliquots. The recovery after proteinase K digestion and purification was about 12.5% (5 mg).
2.2. Crystallization
Crystals grew in a sitting-drop vapour-diffusion setup in the initial sparse-matrix screen (condition G4 of The JCSG Core Suite II; 10 mM CoCl2, 100 mM sodium acetate pH 4.6, 1.0 M hexanediol) with 10 mg ml−1 protein in buffer A at 20°C with a drop consisting of 1 µl protein solution and 0.5 µl reservoir solution. Crystals were observed after 23 weeks (the previous inspection being at 15 weeks), presumably well after the crystallization drop had reached equilibrium. Single crystals could easily be isolated for mounting from a bundle of thick (40–50 µm) needles reaching 200–300 µm in length. The crystal used for phasing was cryoprotected with 10 mM CoCl2, 100 mM sodium acetate pH 4.6, 1.1 M 1,6-hexanediol, 20% glycerol. The crystal used for high-resolution data collection was cryoprotected with 10 mM CoCl2, 100 mM sodium acetate pH 4.6, 2.3 M 1,6-hexanediol. Both crystals were flash-cooled in liquid nitrogen.
2.3. Data collection and processing
Data sets were collected from two different crystals from the same crystallization drop. High-redundancy data for SAD (single-wavelength anomalous dispersion) phasing were collected on a Super Nova diffractometer (Agilent) equipped with a 40 W copper sealed-tube microsource and an Atlas CCD detector. High-resolution data for XDS and scaled with XSCALE (Kabsch, 2010). Data-collection and processing statistics are given in Table 1.
were collected on ESRF beamline ID29 at a wavelength of 0.91985 Å on a PILATUS 6M detector. All data were indexed and integrated with
‡Friedel mates merged. §Rmeas as defined in Diederichs & Karplus (1997): Rmeas = . |
2.4. Structure solution and refinement
2.4.1. solution
The SHELXD (Sheldrick, 2010) run through the graphical user interface HKL2MAP (Pape & Schneider, 2004) for several thousand trials with different high-resolution cutoffs, and between two and four requested sites. Equivalence of the resulting substructures was checked with phenix.emma (Adams et al., 2010). Solutions with CCall and CCweak of around 50 and 30, respectively, yielded three consistent sites.
was solved with2.4.2. Density modification and polyalanine model
Density modification was carried out with the β-test version of SHELXE including autotracing. SHELXE (Sheldrick, 2010) was run with 20 cycles of density modification, a solvent fraction of 0.45 and a high-resolution cutoff for the native data of 1.9 Å (the synchrotron data only became available later) for 200 cycles of autotracing. The best autotracing cycle resulted in 84 out of 113 residues in six chains. The resulting structure had a significant amount of secondary structure and revealed a β-barrel with a single helix.
2.4.3. Model completion and and structure analysis
The resulting polyalanine structure and phases were fed into phenix.autobuild (Adams et al., 2010; Terwilliger et al., 2008) along with the amino-acid sequence for automatic model completion using default settings. The resulting model had 96 out of 113 residues (84%) built in sequence and Rwork of 23.51% and Rfree of 26.02%. This model was completed manually in Coot (Emsley et al., 2010). For the high-resolution synchrotron data, free-R flags were copied from the data set used for phasing and extended. The model was refined with REFMAC5 (Murshudov et al., 2011) using TLS with 12 TLS groups that had been determined with the TLSMD server (Painter & Merritt, 2006). are given in Table 2. The oligomeric state was determined with PISA (Krissinel & Henrick, 2007). Structural similarities were identified with the DALI server (Holm & Rosenström, 2010) and PDBeFold (Krissinel & Henrick, 2004). PDBeFold was run with default settings except setting `lowest acceptable match' to 60% for both query and target. Figures were prepared with PyMOL (DeLano, 2002). Sequences were aligned with MAFFT (Katoh & Standley, 2013). Alignment figures were prepared with ESPript (Gouet et al., 2003).
|
3. Results and discussion
3.1. of the highly conserved protease-resistant Vap core
We first tried to crystallize VapA, as its role in virulence has been most clearly documented. Full-length VapA was degraded into a stable fragment, presumably by trace amounts of 1). We purified this proteinase-K-resistant fragment for crystallization, but did not obtain any diffracting crystals. Therefore, we turned to VapB, the Vap most similar to VapA. Again, full-length VapB was unstable and did not crystallize. We used proteinase K digestion to remove the N-terminus. Edman sequencing (Proteome Factory, Berlin) revealed that the resulting 12 kDa stable fragment started at amino acid Glu85. We will refer to this fragment as VapB throughout this paper. Within the proteinase-K-resistant domains, the VapA and VapB amino-acid sequences are 88% identical plus 6% highly conserved. VapB crystallized in a condition containing 10 mM CoCl2 from a commercial screen. We collected a high-redundancy data set on a copper sealed-tube diffractometer (Table 1). The anomalous signal originating from the S atom of the single methionine and two low-occupancy (∼25%) Co2+ ions was sufficient to solve the by SAD. Clear electron density is visible for residues Glu87–Val196 with some density for residues Gln86 and Ala197. Apparently, there was no or incomplete cleavage of the five C-terminal VapB residues although they are accessible and in a loop conformation.
remaining after purification. In an attempt to completely degrade purified VapA to serve as negative control in cellular assays, we had fortuitously found that VapA has a proteinase-K-resistant core consisting of the highly conserved C-terminus (Supplementary Fig. S13.2. Structure of VapB
VapB forms a monomeric eight-stranded antiparallel β-barrel with a single helix (Fig. 1a). We refer to the molecule as oriented in Fig. 1(a) with both termini and the helix at the bottom and the turn connecting strands 1 and 2 and the loops between strands 5 and 6 or 7 and 8 located at the top. The distribution of hydrophobic and polar residues on the VapB surface is not uniform. The bottom is enriched in polar side chains, including eight negatively and two of the three positively charged residues of VapB, while the top is more hydrophobic (Figs. 1b and 1c). Conserved surface-exposed residues are presumably relevant for the fold and often form hydrogen bonds via their side chains, e.g. Tyr102, Tyr143 or Asn159 and Asn161 (Supplementary Fig. S1). Vaps contain many glycine residues that are very well defined in the electron density. They often locate to β-strands and are conserved for steric reasons, assuming backbone conformations that are not accessible to nonglycine residues, or because the addition of only a Cβ atom would result in clashes (Supplementary Fig. S1). It is questionable whether VapF, another member of the Vap family, can form a stable protein with the same fold, as it lacks highly conserved sequences forming strands 7 and 8 (Supplementary Fig. S2).
A search for similar structures with DALI returns mostly other eight-stranded and some ten-stranded β-barrels but no highly significant hits (Z-scores < 6; sequence identity < 15%). Similar structures include bacterial outer membrane proteins (Schulz, 2002), proteins from the avidin/streptavidin superfamily (Livnah et al., 1993; Weber et al., 1989) and cytoplasmic fatty-acid-binding proteins (Zimmerman & Veerkamp, 2002). PDBeFold identified bacterial lysozyme inhibitors (Leysen et al., 2011; Yum et al., 2009) as top hits.
VapB has a tightly packed core sealed at the bottom by the helix (Fig. 2a). There is neither a central pore typical of outer membrane porins nor is there a central cavity typical for fatty-acid-binding proteins. Functional similarity might be highest to avidins. Avidins bind biotin in a groove at the top of the molecule, when oriented like VapB (Fig. 2b ; Livnah et al., 1993; Weber et al., 1989). The loops at the top of VapB have higher B factors, indicating flexibility (Fig. 2c). The connections between strands 1 and 2 and strands 7 and 8 (residues 99–101 and 181–184, respectively) display double conformations. It is conceivable that these turns/loops move to open a mainly hydrophobic binding site formed by the top of the VapB hydrophobic core. A function of VapB in binding some kind of ligand thus seems possible.
3.3. VapB consists of two Greek-key motifs with unprecedented topology
While VapB is structurally related to bacterial outer membrane proteins, avidins, fatty-acid-binding proteins and lysozyme inhibitors, with root-mean-square deviations of around 3 Å for some 70 aligned Cα atoms, the topology of VapB is quite unusual. Outer membrane proteins, avidins, fatty-acid-binding proteins and lysozyme inhibitors all are antiparallel β-barrels in which each strand is directly linked to its neighbour (Fig. 3a). In contrast, the VapB β-barrel consists of two Greek-key motifs (Hutchinson & Thornton, 1993) with strands arranged in the order 41238567 and a helix connecting strands 4 and 5 (Fig. 3b). Greek-key motifs are common in five- or six-stranded β-barrels and in β-sandwich structures (Zhang & Kim, 2000), for example the β- and γ-crystallins with two Greek-key motifs that form two four-stranded sheets (Jaenicke & Slingsby, 2001). A comprehensive review of Greek-key motifs in β-barrels stated that there is only one eight-stranded β-barrel known that contains Greek keys (Zhang & Kim, 2000). This protein (RNA polymerase subunit RPB8; PDB entry 1a1d; Krapp et al., 1998) has strand order 32148567 (Fig. 3c) and is structurally unrelated to VapB. While we cannot exclude the fact that an eight-stranded β-barrel with the strand order of VapB has been described before, such a topology is not found among the top hits of DALI or PDBeFold.
In conclusion, we have determined the structure of VapB as representative of the R. equi Vap family. The Vaps fold into two Greek-key motifs forming an eight-stranded β-barrel with an unusual, if not unprecedented, topology. The arrangement of strands is different from β-barrels with next-neighbour connections present in outer membrane proteins, avidins, fatty-acid-binding proteins and lysozyme inhibitors and different from crystallins that also consist of two Greek-key motifs. Analysis of the structure and surface properties suggests that Vaps might act as binding proteins involved in interaction with as yet unidentified ligands.
Supporting information
PDB reference: R. equi VapB, 4cv7
Supporting Information. DOI: https://doi.org/10.1107/S2053230X14009911/hv5258sup1.pdf
Acknowledgements
We gratefully acknowledge access to beamline ID29 of the ESRF (Grenoble, France) and coverage of travel expenses through BAG MX-1421. We thank Roman Fedorov (Hannover Medical School) for data collection at ESRF ID29 and Stephan Barden for critical reading of the manuscript and help with figures. This study was supported by the Deutsche Forschungsgemeinschaft (grant SFB 670/TP3 to AH).
References
Adams, P. D. et al. (2010). Acta Cryst. D66, 213–221. Web of Science CrossRef CAS IUCr Journals Google Scholar
von Bargen, K. & Haas, A. (2009). FEMS Microbiol. Rev. 33, 870–891. Web of Science CrossRef PubMed CAS Google Scholar
von Bargen, K., Polidori, M., Becken, U., Huth, G., Prescott, J. F. & Haas, A. (2009). Infect. Immun. 77, 5676–5681. Web of Science CrossRef PubMed CAS Google Scholar
DeLano, W. L. (2002). PyMOL. https://www.pymol.org. Google Scholar
Diederichs, K. & Karplus, P. A. (1997). Nature Struct. Biol. 4, 269–275. CrossRef CAS PubMed Web of Science Google Scholar
Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. (2010). Acta Cryst. D66, 486–501. Web of Science CrossRef CAS IUCr Journals Google Scholar
Gouet, P., Robert, X. & Courcelle, E. (2003). Nucleic Acids Res. 31, 3320–3323. Web of Science CrossRef PubMed CAS Google Scholar
Holm, L. & Rosenström, P. (2010). Nucleic Acids Res. 38, W545–W549. Web of Science CrossRef CAS PubMed Google Scholar
Hutchinson, E. G. & Thornton, J. M. (1993). Protein Eng. 6, 233–245. CrossRef CAS PubMed Web of Science Google Scholar
Jaenicke, R. & Slingsby, C. (2001). Crit. Rev. Biochem. Mol. Biol. 36, 435–499. Web of Science CrossRef PubMed CAS Google Scholar
Jain, S., Bloom, B. R. & Hondalus, M. K. (2003). Mol. Microbiol. 50, 115–128. Web of Science CrossRef PubMed CAS Google Scholar
Kabsch, W. (2010). Acta Cryst. D66, 125–132. Web of Science CrossRef CAS IUCr Journals Google Scholar
Katoh, K. & Standley, D. M. (2013). Mol. Biol. Evol. 30, 772–780. Web of Science CrossRef CAS PubMed Google Scholar
Krapp, S., Kelly, G., Reischl, J., Weinzierl, R. O. & Matthews, S. (1998). Nature Struct. Biol. 5, 110–114. Web of Science CrossRef CAS PubMed Google Scholar
Krissinel, E. & Henrick, K. (2004). Acta Cryst. D60, 2256–2268. Web of Science CrossRef CAS IUCr Journals Google Scholar
Krissinel, E. & Henrick, K. (2007). J. Mol. Biol. 372, 774–797. Web of Science CrossRef PubMed CAS Google Scholar
Letek, M., Ocampo-Sosa, A. A., Sanders, M., Fogarty, U., Buckley, T., Leadon, D. P., González, P., Scortti, M., Meijer, W. G., Parkhill, J., Bentley, S. & Vázquez-Boland, J. A. (2008). J. Bacteriol. 190, 5797–5805. Web of Science CrossRef PubMed CAS Google Scholar
Leysen, S., Van Herreweghe, J. M., Callewaert, L., Heirbaut, M., Buntinx, P., Michiels, C. W. & Strelkov, S. V. (2011). J. Mol. Biol. 405, 1233–1245. Web of Science CrossRef CAS PubMed Google Scholar
Livnah, O., Bayer, E. A., Wilchek, M. & Sussman, J. L. (1993). Proc. Natl Acad. Sci. USA, 90, 5076–5080. CrossRef CAS PubMed Web of Science Google Scholar
Murshudov, G. N., Skubák, P., Lebedev, A. A., Pannu, N. S., Steiner, R. A., Nicholls, R. A., Winn, M. D., Long, F. & Vagin, A. A. (2011). Acta Cryst. D67, 355–367. Web of Science CrossRef CAS IUCr Journals Google Scholar
Painter, J. & Merritt, E. A. (2006). J. Appl. Cryst. 39, 109–111. Web of Science CrossRef CAS IUCr Journals Google Scholar
Pape, T. & Schneider, T. R. (2004). J. Appl. Cryst. 37, 843–844. Web of Science CrossRef CAS IUCr Journals Google Scholar
Schulz, G. E. (2002). Biochim. Biophys. Acta, 1565, 308–317. Web of Science CrossRef PubMed CAS Google Scholar
Sekizaki, T., Takai, S., Egawa, Y., Ikeda, T., Ito, H. & Tsubaki, S. (1995). Gene, 155, 135–136. CrossRef CAS PubMed Web of Science Google Scholar
Sheldrick, G. M. (2010). Acta Cryst. D66, 479–485. Web of Science CrossRef CAS IUCr Journals Google Scholar
Takai, S., Anzai, T., Fujita, Y., Akita, O., Shoda, M., Tsubaki, S. & Wada, R. (2000). Vet. Microbiol. 76, 71–80. Web of Science CrossRef PubMed CAS Google Scholar
Terwilliger, T. C., Grosse-Kunstleve, R. W., Afonine, P. V., Moriarty, N. W., Zwart, P. H., Hung, L.-W., Read, R. J. & Adams, P. D. (2008). Acta Cryst. D64, 61–69. Web of Science CrossRef CAS IUCr Journals Google Scholar
Vázquez-Boland, J. A., Giguère, S., Hapeshi, A., MacArthur, I., Anastasi, E. & Valero-Rello, A. (2013). Vet. Microbiol. 167, 9–33. Web of Science PubMed Google Scholar
Weber, P. C., Ohlendorf, D. H., Wendoloski, J. J. & Salemme, F. R. (1989). Science, 243, 85–88. CrossRef CAS PubMed Web of Science Google Scholar
Yum, S., Kim, M. J., Xu, Y., Jin, X. L., Yoo, H. Y., Park, J.-W., Gong, J. H., Choe, K.-M., Lee, B. L. & Ha, N.-C. (2009). Biochem. Biophys. Res. Commun. 378, 244–248. Web of Science CrossRef PubMed CAS Google Scholar
Zhang, C. & Kim, S.-H. (2000). Proteins, 40, 409–419. Web of Science CrossRef PubMed CAS Google Scholar
Zimmerman, A. W. & Veerkamp, J. H. (2002). Cell. Mol. Life Sci. 59, 1096–1116. Web of Science CrossRef PubMed CAS Google Scholar
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.