research papers
The MORPHEUS protein crystallization screen
aMRC Laboratory of Molecular Biology, Hills Road, Cambridge CB2 0QH, England
*Correspondence e-mail: fgorrec@mrc-lmb.cam.ac.uk
A 96-condition initial screen for protein crystallization, called MORPHEUS, has been developed at the MRC Laboratory of Molecular Biology, Cambridge, England (MRC-LMB). The concept integrates several innovative approaches, such as chemically compatible mixes of potential ligands, new buffer systems and precipitant mixes that also act as cryoprotectants. Instead of gathering a set of crystallization conditions that have already been successful, a selection of molecules frequently observed in the Protein Data Bank (PDB) to co-crystallize with proteins has been made. These have been put together in mixes of similar chemical behaviour and structure, and combined with buffers and precipitant mixes that were also derived from PDB searches, to build the screen de novo. Observations made at the MRC-LMB and many practical aspects were also taken into account when formulating the screen. The resulting screen is easy to use, comprehensive yet small, and has already yielded a list of crystallization hits using both known and novel samples. As an indicator of success, the screen has now become one of the standard screens used routinely at the MRC-LMB when searching initial crystallization conditions for biological macromolecules.
Keywords: macromolecular crystallography; macromolecular crystallization; crystallization screening; crystallization additives.
1. Introduction
https://www.pdb.org; Berman et al., 2000) now holds nearly 60 000 coordinate sets. Approximately 80% of those have been determined by X-ray crystallography, and the method, since its first application to biological macromolecules more than 50 years ago (Kendrew et al., 1958; Perutz et al., 1960), has continued to improve. Recently, the atomic structure of the complete 70S ribosome was determined using X-ray crystallography (Selmer et al., 2006). Given the obvious successes, one might be forgiven for assuming that the basis of the method, the crystallization of a protein, DNA or RNA and their complexes, must be an easy process. In fact, crystallization is now rate limiting and a typical project trying to elucidate the structure of a biological macromolecule of interest will spend most time trying to obtain a sample of biological interest that can be crystallized (Chayen & Saridakis, 2008). The underlying problem is that at the time of the crystallization experiment the structure of the molecule is not known and hence a rational approach cannot be taken.
of biological macromolecules has been tremendously successful over recent years. The Protein Data Bank (PDB,To circumvent this problem, crystallization screens are utilized which try to sample the vast number of possible variables in a manageable and efficient way, either systematically or randomly (McPherson, 2004). Development of an effective search strategy depends on determining how parameter variations influence crystal formation and crystal quality (Kingston et al., 1994). The protein itself can be considered as the main variable (Dale et al., 2003). However, the correct composition of the initial crystallization screen is necessary, although by no means sufficient, for success.
Nowadays, vapour diffusion with 50–200 nl drops is the most widespread crystallization technique and many different commercial screening kits are available to initiate experiments (Berry et al., 2006). Many screens are systematic variations of the concentrations or chemical nature of the components and others employ so-called sparse-matrix approaches that are essentially collections of conditions (mixes of reagents used for protein crystallization) that have been found to work previously with other samples (Jancarik & Kim, 1991).
The increasing number of structures deposited in the PDB has motivated some statistical analyses of the crystallization conditions employed (Hennessy et al., 2000; Kantardjieff & Rupp, 2004), together with attempts to rationalize protein crystallization screens (Zhu et al., 2006; Newstead et al., 2008). Rationalization has led to screens with a minimal number of conditions in sparse matrices and footprint screens (Brzozowski & Walton, 2001; Radaev & Sun, 2002; Tran et al., 2004; Newman et al., 2005). This is logical if overall efficiency is the main goal, such as in structural genomics.
At the MRC Laboratory of Molecular Biology (Cambridge, England), protein samples, DNA–protein complexes and RNA-containing complexes are regularly screened using standard procedures with more than 40 commercial initial screen kits (Stock et al., 2005) and over 1500 conditions, assembled into pre-filled MRC 96-well crystallization plates. This large number is still not large enough because many samples fail to crystallize or give only a very few hits. Amongst others, this could be due to two main reasons. Firstly, the vast number of possible conditions is under-sampled (which is surely true). Secondly, crystallization can be critically dependent on the component(s) in the screen (St John et al., 2008) that make proteins behave differently (more stable or rigid, for example). The latter reason is the rationale behind classical additive screening (Cudney et al., 1994) and a recent development called Silverbullets (McPherson & Cudney, 2006).
Both assumptions were a driving force behind my attempts to formulate the new screen MORPHEUS that could enhance the chances of crystallization. The most important feature of MORPHEUS is the inclusion of mixes containing potential ligands and additives that can promote crystallization through specific interactions. This strategy includes the risk that one component of a mix might have a deleterious effect on crystal growth (or complex association) and thereby mask the positive contribution of another (Larson et al., 2007). By selecting components that have been seen to be ordered in crystal structures in the PDB, the chances of incorporating molecules playing a positive role should increase.
An extensive search of the PDB was performed and small molecules and ions that bind to biological macromolecules were selected. The molecules are stable, commercially available, have a molecular weight below 250 Da and are easy to handle. Components found abundantly in the PDB are potentially good crystallization agents for two reasons. Firstly, they can be stabilizers. For example, some sugars are well known for their thermodynamic stabilization of macromolecules (Arakawa & Timasheff, 1982). Stabilization can also mean `rigidifying' the protein or the and thus improving diffraction quality. Secondly, ligands can create crystallization variants by changing possible interactions on the molecular surface, hence increasing the chances of obtaining different crystals. From this perspective, small counter-anions like nitrate, phosphate and sulfate, with a multitude of possible binding modes via different spatial arrangements of O atoms, are ideal components. For the same reason, small organic salts with carboxylic acid groups can facilitate crystal growth (McPherson, 2001). Additional agents found frequently in the PDB include halides that promote different crystal forms (Lim et al., 1998) and can help with crystallographic phase determination (Dauter et al., 2000). It has been shown that polyethylene (PEGs) tend to form linear binding patterns in clefts on protein surfaces (Hasek, 2006). Therefore, a selection of six PEGs completes the formulation of MORPHEUS.
MORPHEUS provides 96 original conditions made from innovative mixes of potential ligands that have been found with high frequency in the PDB. Will MORPHEUS, like the Greek god of dreams, take different forms, especially those in the shape of crystals? Here, ideas about the formulations and the results from crystallization experiments using test proteins and novel samples are described, proving the high usability and efficiency of MORPHEUS.
2. Materials and methods
The complete formulation of MORPHEUS is shown in Table 1. Fig. 1 is a schematic representation of the screen layout.
|
2.1. Selection of PDB-derived ligands
The set of 47 PDB-derived ligands is listed in Table 2. Initially, structures with ligand(s) were tabulated (July, 2008). Data were then filtered with a molecular weight cut-off of 250 Da. The resulting list was filtered again to keep only ligands seen with at least five unrelated protein structures.
|
Not included in MORPHEUS because of chemical incompatibility are all i.e. additives) that can be integrated into 96 conditions. Concentrations must be high because low affinities should be considered (Sauter et al., 1999).
heavy atoms and detergents. Many divalent cations and some were discarded in later tests because of problems with stability and false positives. Also, there is a limit to the number of ligands (2.2. Additive mixes
Thirty-eight of the selected PDB-derived ligands have been grouped into families depending on their chemical nature to form eight additive mixes. For example, one of the additive mixes is composed of n-ethylene (n = 2–5). By grouping the additives based on chemical nature, the possibility of cross-reaction is avoided and stock solutions are stable. When additives were salts with an acid or base form, the salts were selected so that the final pH of the mix was as neutral as possible. A compound-to-protein ratio of 10:1 is commonly adopted for co-crystallization with small molecule ligands (Danley, 2006) and hence the final concentration of each additive in MORPHEUS is 0.02 M minimum, representing ten times the concentration of a 10 kDa protein at 20 mg ml−1. The recipes for preparing the eight MORPHEUS additive mixes can be found in Table 3.
|
2.3. Precipitant mixes
Precipitants can be mixed to have a synergistic effect (Majeed et al., 2003) and/or to provide cryoprotection (Mitchell & Garman, 1994; McFerrin & Snell, 2002). To take advantage of these findings, four precipitant mixes were integrated in the formulation of MORPHEUS. Three of the mixes have been observed to be more successful in the crystallization of MRC-LMB samples than expected from their under-sampling in our initial screens, as described previously. A fourth mix was designed from scratch with components not found in the other three mixes. Principally, the precipitant mixes have been chosen so that the final conditions produce vitrified ice when frozen. It should be noted, however, that the optimal concentration of cryoprotectant is sample dependent and may need optimization later (Chinte et al., 2005). Recipes for preparing the four MORPHEUS stock solutions with precipitants can be found in Table 4. The table includes the frequency of similar mixes in our MRC-LMB standard initial screens.
2.4. Buffer systems
Six of the selected PDB-derived ligands described before have been used to build three buffer systems within a physiological pH range, namely 6.5, 7.5 and 8.5. The common advantage of buffer systems is that no titration with concentrated acid or base is required (Newman, 2004). Each MORPHEUS buffer system includes an acid and base pair of buffers with similar pKa values. This way, the systems combine the characteristics of two different Good buffers for biological research (Good et al., 1966).
Recipes for preparing 50 ml of the three MORPHEUS buffer systems can be found in Table 5. Non-titrated stock solutions of the individual buffers (at a concentration of 1 M) were mixed at different ratios for optimization purposes.
|
The chemicals used for making the buffer systems were MES [2-(N-morpholino)ethanesulfonic acid; Sigma, M8250, pH 2.7], imidazole (1,3-diazacyclopenta-2,4-diene; BDH, 286874D, pH 9.9), MOPS [3-(N-morpholino)propanesulfonic acid; BDH, 4438321, pH 2.9], HEPES-Na [sodium 4-(2-hydroxyethyl)piperazine-1-ethanesulfonate; Melford, B2001, pH 10.4], bicine [N,N-bis(2-hydroxyethyl)glycine; Fluka, 14871, pH 4.9] and Trizma base [proprietary Tris, 2-amino-2-(hydroxymethyl)-1,3-propanediol; Sigma, T1503, pH 10.6]. The pH was measured at 294 K with an InLab 490 solid-state probe (Mettler–Toledo) to avoid inaccuracies with Tris-containing buffers.
2.5. Stability tests
The stability of the conditions during their development was assessed by checking the turbidity and pH after one week at 293 K, one week at 277 K and another week at 293 K.
2.6. Proteins
For details of the proteins used, please refer to Table 6.
|
2.7. Crystallization trials
MRC crystallization plates (Swissci) containing MORPHEUS (85 µl in the main wells) were prepared on a Mosquito (TTP labtech) or ScreenMaker (Innovadyne) nanolitre liquid handler. Our standard setup for initial screens is to mix equal-volume aliquots of the protein and condition at 297 K, with a 200 nl final volume of drops, and to store the plates at 292 K. Final assessments were made after one week by manual inspection using a high-powered Leica MX-12 stereomicroscope. A drop was considered a crystallization hit when it contained protein crystals larger than 20 µm, so that they could be mounted in a cryoloop for X-ray diffraction.
2.8. Optimization of conditions
Finally, all three components, the ligand mixes, the precipitant mixes and the buffers, are combined using a fixed ratio,
This simple recipe facilitates easy follow-up optimization experiments. As an initial approach, one can simply change the above ratios of the stock solutions. The composition of the buffer systems may be altered during optimization experiments to change the pH. Obviously, all of these optimization experiments are very amenable to automation (Hennessy et al., 2009).
3. Results and discussion
Both well known test proteins and novel samples were tried with MORPHEUS. Table 6 shows all the details and results of the crystallization trials performed for 16 samples. Fig. 2 shows the different crystal morphologies observed. All the crystals shown represent initial hits, except for Scc3 (domain of sister chromatid cohesion protein 3) and PI3K-I (pi3-kinase p110 in complex with isoform-specific inhibitors) which involved optimization.
Importantly, three samples have crystallized exclusively in MORPHEUS and produced no hits from any other screen tried (over 1500 conditions): Scc3, PI3K-I and TriUb-D (triubiquitin in complex with a ubiquitin-binding domain).
The possible specificity of ligand mixes can be spotted easily because of the systematic screen layout: when there are several hits in the same row of MORPHEUS, it means there is specificity to ligands used in the conditions of that row (see samples PI3K-I, ParR, PAK4G and THM). In the same way, specificity to precipitant(s) and pH can easily be noticed (see Fig. 1). For example, most of the hits with the test sample BAR were in conditions that integrate the mix of precipitants developed for MORPHEUS (mix found in columns 4, 8 and 12: 12.5% PEG 1000, 12.5% PEG 3350 and 12.5% MPD).
4. Conclusions
The advantages of designing an initial screen de novo have been demonstrated. MORPHEUS delivers a screen that is easy to make and the conditions are easy to optimize. It contains components that have been selected from crystallized complexes of previously published structures. It also contains a limited number of precipitant mixes that have been selected using local data from the MRC-LMB. MORPHEUS has been successful in crystallizing both known proteins and important new samples.
Ideally, more small molecules with interesting characteristics that are not used in commercially available screens should be investigated, like some polyols (Cohen et al., 1993). An extensive set of amine derivatives, including well known polyamine additives (Ding et al., 1999) and aminated amino acids (Matsuoka et al., 2007), could form an excellent additive screen with frozen solutions for storage. Also, protein chaperones could be added for some challenging crystallizations (Ostermeier et al., 1995; Tereshko et al., 2008). In the same spirit, it would be interesting to investigate what could be done with molecules designed to mimic protein–protein interactions (Allen et al., 1998).
Acknowledgements
I would like to thank Andrew Turnbull (Cancer Research Technology Ltd.) and Evangelos Papagrigoriou (Insight Research Group) for giving early advice. The development of the screen was sponsored by the MRC-LMB under the supervision of Jan Löwe and Olga Perisic. The frequency of the precipitant mixes was revealed by the LMB screen database (https://www2.mrc-lmb.cam.ac.uk/groups/JYL/WWWrobots/robot.html) developed in collaboration with Paul Hart (MRC-LMB). Data on triubiquitin, pi3-kinase and plk1 complexes were given by Yogesh Kulathu, Alex Berndt and Ana Julia Narvaez, respectively. I would also like to thank everyone at the MRC-LMB for their assistance in trying MORPHEUS and providing test samples for two years before its commercialization. Conflicting commercial interest: I hereby state that I have a conflicting commercial interest, in that the MORPHEUS crystallization screen has been commercialized by Molecular Dimensions Ltd (https://www.moleculardimensions.com) under an exclusive licence to MRC Technology.
References
Allen, J. V., Horwell, D. C., Lainton, J. A. H., O'Neill, J. A. & Ratcliffe, G. S. (1998). Lett. Pept. Sci. 5, 133–137. CAS Google Scholar
Arakawa, T. & Timasheff, S. N. (1982). Biochemistry, 21, 6536–6544. CrossRef CAS PubMed Web of Science Google Scholar
Berman, H. M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T. N., Weissig, H., Shindyalov, I. N. & Bourne, P. E. (2000). Nucleic Acids Res. 28, 235–242. Web of Science CrossRef PubMed CAS Google Scholar
Berry, I. M., Dym, O., Esnouf, R. M., Harlos, K., Meged, R., Perrakis, A., Sussman, J. L., Walter, T. S., Wilson, J. & Messerschmidt, A. (2006). Acta Cryst. D62, 1137–1149. Web of Science CrossRef CAS IUCr Journals Google Scholar
Brzozowski, A. M. & Walton, J. (2001). J. Appl. Cryst. 34, 97–101. Web of Science CrossRef CAS IUCr Journals Google Scholar
Chayen, N. E. & Saridakis, E. (2008). Nat. Methods, 5, 147–153. Web of Science CrossRef PubMed CAS Google Scholar
Chinte, U., Shah, B., DeWitt, K., Kirschbaum, K., Pinkerton, A. A. & Schall, C. (2005). J. Appl. Cryst. 38, 412–419. Web of Science CrossRef CAS IUCr Journals Google Scholar
Cohen, S., Marcus, Y., Migron, Y., Dikstein, S. & Shafran, S. (1993). J. Chem. Soc. Faraday Trans. 89, 3271–3275. CrossRef CAS Web of Science Google Scholar
Cordell, S. C., Robinson, E. J. & Löwe, J. (2003). Proc. Natl Acad. Sci. USA, 100, 7889–7894. Web of Science CrossRef PubMed CAS Google Scholar
Cudney, R., Patel, S., Weisgraber, K., Newhouse, Y. & McPherson, A. (1994). Acta Cryst. D50, 414–423. CrossRef CAS IUCr Journals Google Scholar
Dale, G. E., Oefner, C. & D'Arcy, A. (2003). J. Struct. Biol. 142, 88–97. Web of Science CrossRef PubMed CAS Google Scholar
Danley, D. E. (2006). Acta Cryst. D62, 569–575. Web of Science CrossRef CAS IUCr Journals Google Scholar
Dauter, Z., Dauter, M. & Rajashankar, K. R. (2000). Acta Cryst. D56, 232–237. Web of Science CrossRef CAS IUCr Journals Google Scholar
Ding, L., Zhang, Y., Deacon, A. M., Ealick, S. E., Ni, Y., Sun, P. & Coleman, W. G. Jr (1999). Acta Cryst. D55, 685–688. Web of Science CrossRef CAS IUCr Journals Google Scholar
Garcia-Alvarez, B., de Carcer, G., Ibanez, S., Bragado-Nilsson, E. & Montoya, G. (2007). Proc. Natl Acad. Sci. USA, 104, 3107–3112. Web of Science CrossRef PubMed CAS Google Scholar
Good, N. E., Winget, G. D., Winter, W., Connolly, T. N., Izawa, S. & Singh, R. M. (1966). Biochemistry, 5, 467–477. CrossRef CAS PubMed Web of Science Google Scholar
Hasek, J. (2006). Z. Kristallogr. 23, 613–618. CrossRef Google Scholar
Hennessy, D., Buchanan, B., Subramanian, D., Wilkosz, P. A. & Rosenberg, J. M. (2000). Acta Cryst. D56, 817–827. Web of Science CrossRef CAS IUCr Journals Google Scholar
Hennessy, D. N., Narayanan, B. & Rosenberg, J. M. (2009). Acta Cryst. D65, 1001–1003. Web of Science CrossRef IUCr Journals Google Scholar
Jancarik, J. & Kim, S.-H. (1991). J. Appl. Cryst. 24, 409–411. CrossRef CAS Web of Science IUCr Journals Google Scholar
Joerger, A. C., Ang, H. C. & Fersht, A. R. (2006). Proc. Natl Acad. Sci. USA, 103, 15056–15061. Web of Science CrossRef PubMed CAS Google Scholar
Kambach, C., Walke, S., Young, R., Avis, J. M., de la Fortelle, E., Raker, V. A., Luhrmann, R., Li, J. & Nagai, K. (1999). Cell, 96, 375–387. Web of Science CrossRef PubMed CAS Google Scholar
Kantardjieff, K. A. & Rupp, B. (2004). Bioinformatics, 20, 2162–2168. Web of Science CrossRef PubMed CAS Google Scholar
Kendrew, J. C., Bodo, G., Dintzi, H. M., Parrish, R. G., Wyckoff, H. & Phillips, D. C. (1958). Nature (London), 181, 662–666. CrossRef PubMed CAS Web of Science Google Scholar
Kingston, R. L., Baker, H. M. & Baker, E. N. (1994). Acta Cryst. D50, 429–440. CrossRef CAS Web of Science IUCr Journals Google Scholar
Larson, S. B., Day, J. S., Cudney, R. & McPherson, A. (2007). Acta Cryst. D63, 310–318. Web of Science CrossRef CAS IUCr Journals Google Scholar
Leonard, T. A., Butler, P. J. & Löwe, J. (2004). Mol. Microbiol. 53, 419–432. Web of Science CrossRef PubMed CAS Google Scholar
Lim, K., Nadarajah, A., Forsythe, E. L. & Pusey, M. L. (1998). Acta Cryst. D54, 899–904. Web of Science CrossRef CAS IUCr Journals Google Scholar
Low, H. H. & Löwe, J. (2006). Nature (London), 444, 766–769. Web of Science CrossRef PubMed CAS Google Scholar
Majeed, S., Ofek, G., Belachew, A., Huang, C. C., Zhou, T. & Kwong, P. D. (2003). Structure, 11, 1061–1070. Web of Science CrossRef PubMed CAS Google Scholar
Matsuoka, T., Tomita, S., Hamada, H. & Shiraki, K. (2007). J. Biosci. Bioeng. 103, 440–443. Web of Science CrossRef PubMed CAS Google Scholar
McFerrin, M. B. & Snell, E. H. (2002). J. Appl. Cryst. 35, 538–545. Web of Science CrossRef CAS IUCr Journals Google Scholar
McPherson, A. (2001). Protein Sci. 10, 418–422. Web of Science CrossRef PubMed CAS Google Scholar
McPherson, A. (2004). Methods, 34, 254–265. Web of Science CrossRef PubMed CAS Google Scholar
McPherson, A. & Cudney, B. (2006). J. Struct. Biol. 156, 387–406. Web of Science CrossRef PubMed CAS Google Scholar
Mitchell, E. P. & Garman, E. F. (1994). J. Appl. Cryst. 27, 1070–1074. CrossRef CAS IUCr Journals Google Scholar
Møller-Jensen, J., Ringgaard, S., Mercogliano, C. P., Gerdes, K. & Löwe, J. (2007). EMBO J. 26, 4413–4422. Web of Science PubMed Google Scholar
Newman, J. (2004). Acta Cryst. D60, 610–612. Web of Science CrossRef CAS IUCr Journals Google Scholar
Newman, J., Egan, D., Walter, T. S., Meged, R., Berry, I., Ben Jelloul, M., Sussman, J. L., Stuart, D. I. & Perrakis, A. (2005). Acta Cryst. D61, 1426–1431. Web of Science CrossRef CAS IUCr Journals Google Scholar
Newstead, S., Ferrandon, S. & Iwata, S. (2008). Protein Sci. 17, 466–472. Web of Science CrossRef PubMed CAS Google Scholar
Ostermeier, C., Iwata, S., Ludwig, B. & Michel, H. (1995). Nat. Struct. Biol. 2, 842–846. CrossRef CAS PubMed Web of Science Google Scholar
Perutz, M. F., Rossmann, M. G., Af, C., Muirhead, H., Will, G. & North, A. C. (1960). Nature (London), 185, 416–422. CrossRef PubMed CAS Web of Science Google Scholar
Peter, B. J., Kent, H. M., Mills, I. G., Vallis, Y., Butler, P. J., Evans, P. R. & McMahon, H. T. (2004). Science, 303, 495–499. Web of Science CrossRef PubMed CAS Google Scholar
Radaev, S. & Sun, P. D. (2002). J. Appl. Cryst. 35, 674–676. Web of Science CrossRef CAS IUCr Journals Google Scholar
Sauter, C., Ng, J. D., Lorber, B., Keith, G., Brion, P., Hosseini, M. W., Lehn, J. M. & Giegé, R. (1999). J. Cryst. Growth, 196, 365–376. Web of Science CrossRef CAS Google Scholar
Selmer, M., Dunham, C. M., Murphy, F. V. T., Weixlbaumer, A., Petry, S., Kelley, A. C., Weir, J. R. & Ramakrishnan, V. (2006). Science, 313, 1935–1942. Web of Science CrossRef PubMed CAS Google Scholar
Sivanathan, V., Allen, M. D., de Bekker, C., Baker, R., Arciszewska, L. K., Freund, S. M., Bycroft, M., Löwe, J. & Sherratt, D. J. (2006). Nat. Struct. Mol. Biol. 13, 965–972. Web of Science CrossRef PubMed CAS Google Scholar
St John, F. J., Feng, B. & Pozharski, E. (2008). Acta Cryst. D64, 1222–1227. Web of Science CrossRef CAS IUCr Journals Google Scholar
Stewart, M., Kent, H. M. & McCoy, A. J. (1998). J. Mol. Biol. 284, 1517–1527. Web of Science CrossRef CAS PubMed Google Scholar
Stock, D., Perisic, O. & Löwe, J. (2005). Prog. Biophys. Mol. Biol. 88, 311–327. Web of Science CrossRef PubMed CAS Google Scholar
Teo, H., Gill, D. J., Sun, J., Perisic, O., Veprintsev, D. B., Vallis, Y., Emr, S. D. & Williams, R. L. (2006). Cell, 125, 99–111. Web of Science CrossRef PubMed CAS Google Scholar
Tereshko, V., Uysal, S., Koide, A., Margalef, K., Koide, S. & Kossiakoff, A. A. (2008). Protein Sci. 17, 1175–1187. Web of Science CrossRef PubMed CAS Google Scholar
Tran, T. T., Sorel, I. & Lewit-Bentley, A. (2004). Acta Cryst. D60, 1562–1568. Web of Science CrossRef CAS IUCr Journals Google Scholar
Wernimont, A. K. & Weissenhorn, W. (2004). BMC Struct. Biol. 4, 10. Google Scholar
Zhu, D. W., Garneau, A., Mazumdar, M., Zhou, M., Xu, G. J. & Lin, S. X. (2006). J. Struct. Biol. 154, 297–302. 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.