crystallization communications\(\def\hfill{\hskip 5em}\def\hfil{\hskip 3em}\def\eqno#1{\hfil {#1}}\)

Journal logoSTRUCTURAL BIOLOGY
COMMUNICATIONS
ISSN: 2053-230X

Preliminary X-ray crystallographic analysis of the secreted chorismate mutase from Mycobacterium tuberculosis: a tricky crystallization problem solved

aDepartment of Chemistry and Bioscience, Chalmers University of Technology, PO Box 462, SE-40530 Göteborg, Sweden, bDepartment of Chemistry, University of Oslo, PO Box 1033, Blindern, N-0315 Oslo, Norway, and cLaboratory of Organic Chemistry, ETH Zurich, CH-8093 Zurich, Switzerland
*Correspondence e-mail: ute.krengel@kjemi.uio.no

(Received 5 December 2005; accepted 3 April 2006; online 12 April 2006)

Chorismate mutase catalyzes the conversion of chorismate to prephenate in the biosynthesis of the aromatic amino acids tyrosine and phenylalanine in bacteria, fungi and plants. Here, the crystallization of the unusual secreted chorismate mutase from Mycobacterium tuberculosis (encoded by Rv1885c), a 37.2 kDa dimeric protein belonging to the AroQγ subclass of mutases, is reported. Crystal optimization was non-trivial and is discussed in detail. To obtain crystals of sufficient quality, it was critical to initiate crystallization at higher precipitant concentration and then transfer the drops to lower precipitant concentrations within 5–15 min, in an adaptation of a previously described technique [Saridakis & Chayen (2000[Saridakis, E. & Chayen, N. E. (2000). Protein Sci. 9, 755-757.]), Protein Sci. 9, 755–757]. As a result of the optimization, diffraction improved from 3.5 to 1.3 Å resolution. The crystals belong to space group P21, with unit-cell parameters a = 42.6, b = 72.6, c = 62.0 Å, β = 104.5°. The asymmetric unit contains one biological dimer, with 167 amino acids per protomer. A soak with a transition-state analogue is also described.

1. Introduction

The biosynthesis of amino acids is an essential process in living cells as it provides the building blocks for polypeptides and important primary and secondary metabolites. The shikimate pathway is the anabolic pathway employed by bacteria, fungi and plants to synthesize the aromatic amino acids phenylalanine (Phe), tyrosine (Tyr) and tryptophan (Trp) (Haslam, 1993[Haslam, E. (1993). Shikimic Acid: Metabolism and Metabolites. New York: John Wiley & Sons.]). Chorismate mutase (CM; EC 5.4.99.5) is located at the branch point of the shikimate pathway and channels chorismate into the Tyr/Phe-specific branch. The enzyme catalyzes the conversion of chorismate to prephenate, which is formally a Claisen rearrangement. Chorismate mutases typically accelerate the uncatalyzed rearrangement by roughly one million-fold (Andrews et al., 1973[Andrews, P. R., Smith, G. D. & Young, I. G. (1973). Biochemistry, 12, 3492-3498.]). As one of the few enzymes known to catalyze a pericyclic process, it has generated considerable interest among enzymologists and bioorganic and computational chemists, specifically with respect to the origins of the catalytic efficiency (Andrews et al., 1973[Andrews, P. R., Smith, G. D. & Young, I. G. (1973). Biochemistry, 12, 3492-3498.]; Görisch, 1978[Görisch, H. (1978). Biochemistry, 17, 3700-3705.]; Chook et al., 1994[Chook, Y. M., Gray, J. V., Ke, H. & Lipscomb, W. N. (1994). J. Mol. Biol. 240, 476-500.]; Haynes et al., 1994[Haynes, M. R., Stura, E. A., Hilvert, D. & Wilson, I. A. (1994). Science, 263, 646-652.]; Wiest & Houk, 1995[Wiest, O. & Houk, K. N. (1995). J. Am. Chem. Soc. 117, 11628-11639.]; Lee, Stewart et al., 1995[Lee, A. Y., Stewart, J. D., Clardy, J. & Ganem, B. (1995). Chem. Biol. 2, 195-­203.]; Kast, Asif-Ullah & Hilvert, 1996[Kast, P., Asif-Ullah, M. & Hilvert, D. (1996). Tetrahedron Lett. 37, 2691-2694.]; Kast, Asif-Ullah, Jiang et al., 1996[Kast, P., Asif-Ullah, M., Jiang, N. & Hilvert, D. (1996). Proc. Natl Acad. Sci. USA, 93, 5043-5048.]; Ma et al., 1998[Ma, J., Zheng, X., Schnappauf, G., Braus, G., Karplus, M. & Lipscomb, W. N. (1998). Proc. Natl Acad. Sci. USA, 95, 14640-14645.]; Mattei et al., 1999[Mattei, P., Kast, P. & Hilvert, D. (1999). Eur. J. Biochem. 261, 25-32.]; Gustin et al., 1999[Gustin, D. J., Mattei, P., Kast, P., Wiest, O., Lee, L., Cleland, W. W. & Hilvert, D. (1999). J. Am. Chem. Soc. 121, 1756-1757.]; Kast et al., 2000[Kast, P., Grisostomi, C., Chen, I. A., Li, S., Krengel, U., Xue, Y. & Hilvert, D. (2000). J. Biol. Chem. 275, 36832-36838.]; Kienhöfer et al., 2003[Kienhöfer, A., Kast, P. & Hilvert, D. (2003). J. Am. Chem. Soc. 125, 3206-3207.]; Guo et al., 2003[Guo, H., Cui, Q., Lipscomb, W. N. & Karplus, M. (2003). Angew. Chem. Int. Ed. Engl. 42, 1508-1511.]; Guimarães et al., 2003[Guimarães, C. R. W., Repasky, M. P., Chandrasekhar, J., Tirado-Rives, J. & Jorgensen, W. L. (2003). J. Am. Chem. Soc. 125, 6892-6899.]; Štrajbl et al., 2003[Štrajbl, M., Shurki, A., Kato, M. & Warshel, A. (2003). J. Am. Chem. Soc. 125, 10228-10237.]; Zhang et al., 2005[Zhang, X., Zhang, X. & Bruice, T. C. (2005). Biochemistry, 44, 10443-10448.]).

Structures have been determined for chorismate mutases from Bacillus subtilis (Chook et al., 1993[Chook, Y. M., Ke, H. & Lipscomb, W. N. (1993). Proc. Natl Acad. Sci. USA, 90, 8600-8603.], 1994[Chook, Y. M., Gray, J. V., Ke, H. & Lipscomb, W. N. (1994). J. Mol. Biol. 240, 476-500.]; Ladner et al., 2000[Ladner, J. E., Reddy, P., Davis, A., Tordova, M., Howard, A. J. & Gilliland, G. L. (2000). Acta Cryst. D56, 673-683.]), Escherichia coli (Lee, Karplus et al., 1995[Lee, A. Y., Karplus, P. A., Ganem, B. & Clardy, J. (1995). J. Am. Chem. Soc. 117, 3627-3628.]), Saccharomyces cerevisiae (Xue et al., 1994[Xue, Y., Lipscomb, W. N., Graf, R., Schnappauf, G. & Braus, G. (1994). Proc. Natl Acad. Sci. USA, 91, 10814-10818.]; Sträter et al., 1996[Sträter, N., Håkansson, K., Schnappauf, G., Braus, G. & Lipscomb, W. N. (1996). Proc. Natl Acad. Sci. USA, 93, 3330-3334.], 1997[Sträter, N., Schnappauf, G., Braus, G. & Lipscomb, W. N. (1997). Structure, 5, 1437-1452.]) and the catalytic antibody 1F7 with chorismate mutase activity (Haynes et al., 1994[Haynes, M. R., Stura, E. A., Hilvert, D. & Wilson, I. A. (1994). Science, 263, 646-652.]). Recently, coordinates have also been deposited in the PDB for chorismate mutases from Thermus thermophilus (PDB codes 1ode , 1ufy , 1ui9 ) and Clostridium thermocellum (PDB code 1xho ; Xu et al., 2005[Xu, H., Yang, C., Chen, L., Kataeva, I. A., Tempel, W., Lee, D., Habel, J. E., Nguyen, D., Pflugrath, J. W., Ferrara, J. D., Arendall, W. B. III, Richardson, J. S., Richardson, D. C., Liu, Z.-J., Newton, M. G., Rose, J. P. & Wang, B.-C. (2005). Acta Cryst. D61, 960-966.]). Interestingly, these examples comprise three completely unrelated folds: while the structurally characterized enzymes from B. subtilis, T. thermophilus and C. thermocellum have a trimeric pseudo-α/β-barrel fold typical of the so-called AroH class of chorismate mutases (Chook et al., 1994[Chook, Y. M., Gray, J. V., Ke, H. & Lipscomb, W. N. (1994). J. Mol. Biol. 240, 476-500.]; MacBeath et al., 1998[MacBeath, G., Kast, P. & Hilvert, D. (1998). Biochemistry, 37, 10062-10073.]), 1F7 exhibits the characteristic immunoglobulin β-sheet topology. The chorismate mutases of the AroQ class adopt yet another fold, which is dominated by α-helices. Its prototype, the chorismate mutase domain (AroQp) of the bifunctional E. coli chorismate mutase-prephenate dehydratase (also known as EcCM), is an intertwined homodimer that comprises three α-helices per monomer (Lee, Karplus et al., 1995[Lee, A. Y., Karplus, P. A., Ganem, B. & Clardy, J. (1995). J. Am. Chem. Soc. 117, 3627-3628.]). This fold has recently been classified as AroQα (Ökvist et al., 2006[Ökvist, M., Dey, R., Sasso, S., Grahn, E., Kast, P. & Krengel, U. (2006). J. Mol. Biol. 357, 1483-1499.]). The allosteric yeast chorismate mutase represents the AroQβ subclass (Ökvist et al., 2006[Ökvist, M., Dey, R., Sasso, S., Grahn, E., Kast, P. & Krengel, U. (2006). J. Mol. Biol. 357, 1483-1499.]), a more elaborate version of the basic AroQ fold, which shares weak sequence similarity with EcCM and possesses a structurally related active-site domain (Xue et al., 1994[Xue, Y., Lipscomb, W. N., Graf, R., Schnappauf, G. & Braus, G. (1994). Proc. Natl Acad. Sci. USA, 91, 10814-10818.]; Xue & Lipscomb, 1995[Xue, Y. & Lipscomb, W. N. (1995). Proc. Natl Acad. Sci. USA, 92, 10595-10598.]; Sträter et al., 1996[Sträter, N., Håkansson, K., Schnappauf, G., Braus, G. & Lipscomb, W. N. (1996). Proc. Natl Acad. Sci. USA, 93, 3330-3334.], 1997[Sträter, N., Schnappauf, G., Braus, G. & Lipscomb, W. N. (1997). Structure, 5, 1437-1452.]).

A third AroQ subclass, initially known as *AroQ (Xia et al., 1993[Xia, T., Song, J., Zhao, G., Aldrich, H. & Jensen, R. A. (1993). J. Bacteriol. 175, 4729-4737.]; Calhoun et al., 2001[Calhoun, D. H., Bonner, C. A., Gu, W., Xie, G. & Jensen, R. A. (2001). Genome Biol. 2, 30.01-30.16.]; Sasso et al., 2005[Sasso, S., Ramakrishnan, C., Gamper, M., Hilvert, D. & Kast, P. (2005). FEBS J. 272, 375-389.]) and recently classified as AroQγ (Ökvist et al., 2006[Ökvist, M., Dey, R., Sasso, S., Grahn, E., Kast, P. & Krengel, U. (2006). J. Mol. Biol. 357, 1483-1499.]), comprises a group of chorismate mutases that exhibit convincing sequence similarity to other AroQ sequences only in the N-terminal moiety (Sasso et al., 2005[Sasso, S., Ramakrishnan, C., Gamper, M., Hilvert, D. & Kast, P. (2005). FEBS J. 272, 375-389.]; Ökvist et al., 2006[Ökvist, M., Dey, R., Sasso, S., Grahn, E., Kast, P. & Krengel, U. (2006). J. Mol. Biol. 357, 1483-1499.]). These enzymes are about twice the size of the classical AroQα domain and feature a cleavable signal peptide responsible for the export of these chorismate mutases from the cytoplasm (Xia et al., 1993[Xia, T., Song, J., Zhao, G., Aldrich, H. & Jensen, R. A. (1993). J. Bacteriol. 175, 4729-4737.]; Calhoun et al., 2001[Calhoun, D. H., Bonner, C. A., Gu, W., Xie, G. & Jensen, R. A. (2001). Genome Biol. 2, 30.01-30.16.]; Sasso et al., 2005[Sasso, S., Ramakrishnan, C., Gamper, M., Hilvert, D. & Kast, P. (2005). FEBS J. 272, 375-389.]). It has been speculated that AroQγ enzymes may play some role in the pathogenicity of its producer organism (for a discussion and references, see Sasso et al., 2005[Sasso, S., Ramakrishnan, C., Gamper, M., Hilvert, D. & Kast, P. (2005). FEBS J. 272, 375-389.]). Primary structure comparisons suggested that AroQγ proteins also have other structural peculiarities (MacBeath et al., 1998[MacBeath, G., Kast, P. & Hilvert, D. (1998). Biochemistry, 37, 10062-10073.]; Sasso et al., 2005[Sasso, S., Ramakrishnan, C., Gamper, M., Hilvert, D. & Kast, P. (2005). FEBS J. 272, 375-389.]). In particular, a critical amino-terminal segment carrying a conserved active-site arginine is lacking in AroQγ chorismate mutases. Instead, a homologous helical segment is present in the middle of the polypeptide chain, indicating that the structure of the exported AroQγ proteins must be topologically distinct from those of the other known AroQ prototypes (Sasso et al., 2005[Sasso, S., Ramakrishnan, C., Gamper, M., Hilvert, D. & Kast, P. (2005). FEBS J. 272, 375-389.]). To investigate this issue and to gain insights into the structure–function relationship of these enzymes, which have the potential to serve as formidable specific drug targets, we have set out to determine the high-resolution crystal structure of the secreted chorismate mutase from the model pathogen Mycobacterium tuberculosis (*MtCM; encoded by open reading frame Rv1885c in strain H37Rv), which represents the first characterized example of an AroQγ fold (Ökvist et al., 2006[Ökvist, M., Dey, R., Sasso, S., Grahn, E., Kast, P. & Krengel, U. (2006). J. Mol. Biol. 357, 1483-1499.]).

However, crystallization of this enzyme turned out to be non-trivial. While small well ordered crystals could readily be obtained within less than half an hour of setting up the drops, it seemed impossible to obtain (fewer and) larger crystals with better diffraction quality. Common strategies to increase crystal size are to lower the precipitant or protein concentration, thereby moving in the phase diagram from a condition of high supersaturation (where nucleation occurs) to a concentration of lower supersaturation (thus reducing the number of spontaneously formed nuclei, but still staying within the `labile' zone where nucleation can occur; Mikol & Giegé, 1992[Mikol, V. & Giegé, R. (1992). Crystallization of Nucleic Acids and Proteins, edited by A. Ducruix & R. Giegé, pp. 219-239. Oxford University Press.]). Another well known factor that can influence nucleation is the temperature, which has an impact both on the phase diagram itself and on equilibration kinetics. As an alternative to temperature variation, the equilibration time might be influenced more predictably by applying a layer of oil on top of the reservoir of a vapour-diffusion setup (Chayen, 1997[Chayen, N. E. (1997). J. Appl. Cryst. 30, 198-202.]). In this way, equilibration is slowed down, warranting controlled entry into the `labile' nucleation zone.

All of these protocols work relatively well if the nucleation rate increases gradually upon increasing precipitant or protein concentration. However, if instead there is an abrupt increase in nucleation rate, it can be very difficult to hit the point optimal for both nucleation and crystal growth (if such a point exists at all). In these cases, the application of seeding techniques may be attempted (Stura & Wilson, 1992[Stura, E. A. & Wilson, I. A. (1992). Crystallization of Nucleic Acids and Proteins, edited by A. Ducruix & R. Giegé, pp. 99-126. Oxford University Press.]; Stura, 1999[Stura, E. A. (1999). Protein Crystallization: Techniques, Strategies, and Tips, edited by T. M. Bergfors, pp. 141-153. La Jolla, CA, USA: International University Line.]), which approach the problem from the opposite end by placing crystal nuclei in the form of small seeds into the metastable zone optimal for crystal growth. Finding conditions appropriate for seeding has its own inherent difficulties, however, as the metastable zone can be rather narrow and excursions to either side of the phase diagram will inevitably result in either extensive secondary nucleation or dissolution of the seeds. If only little protein is available, this adds to the challenge. A few years ago, two new approaches to the problem were described (Saridakis et al., 1994[Saridakis, E. E. G., Shaw Stewart, P. D., Lloyd, L. F. & Blow, D. M. (1994). Acta Cryst. D50, 293-297.]; Saridakis & Chayen, 2000[Saridakis, E. & Chayen, N. E. (2000). Protein Sci. 9, 755-757.]). The first method involves diluting microbatch drops after incubating them for several hours under spontaneous nucleation conditions (Saridakis et al., 1994[Saridakis, E. E. G., Shaw Stewart, P. D., Lloyd, L. F. & Blow, D. M. (1994). Acta Cryst. D50, 293-297.]), while the second approach relies on transferring cover slips of hanging-drop vapour-diffusion setups from nucleation conditions to the metastable zone (Saridakis & Chayen, 2000[Saridakis, E. & Chayen, N. E. (2000). Protein Sci. 9, 755-757.]). In the present report, we describe the crystal optimization of a protein target that crystallizes very rapidly and show that this problem can be solved by adapting the approach by Saridakis & Chayen (2000[Saridakis, E. & Chayen, N. E. (2000). Protein Sci. 9, 755-757.]) to very short crystallization times.

2. Experimental procedures, results and discussion

2.1. Crystallization

Two different protein batches were used for crystallization. In both cases, recombinant *MtCM was prepared from E. coli strain KA29 as described in Sasso et al. (2005[Sasso, S., Ramakrishnan, C., Gamper, M., Hilvert, D. & Kast, P. (2005). FEBS J. 272, 375-389.]) and concentrated to 10–20 mg ml−1. Immediately prior to crystallization setups, the protein solutions were centrifuged at 10 000 rev min−1 for 10 min at 277 K in order to spin down any floating particles and aggregated insoluble protein. Standard initial crystallization tests were then performed by the hanging-drop vapour-diffusion method in 24-well tissue-culture plates (Göteborgs Termometerfabrik, Sweden) using siliconized glass cover slips from Hampton Research (CA, USA).

For the first crystallization trials, we used a 175-residue version of *MtCM (encoded on plasmid pKTU3-HCT) devoid of the cleavable signal peptide but containing a C-terminal His tag (Sasso et al., 2005[Sasso, S., Ramakrishnan, C., Gamper, M., Hilvert, D. & Kast, P. (2005). FEBS J. 272, 375-389.]). The protein sample (12.6 mg ml−1) was dissolved in 20 mM potassium phosphate buffer pH 7.5. Structure Screen 1 (Molecular Dimensions Ltd, England) gave a few leads, which were optimized, with the best preliminary condition consisting only of 15% PEG 8000 (without buffer). In order to explore different conditions which were not limited by the use of phosphate as buffer, we then switched to protein (18.6 mg ml−1) buffered with 20 mM Tris–HCl pH 8.0. This time, the leaderless untagged 167-residue version of *MtCM encoded by plasmid pKTU3-HT was used (Sasso et al., 2005[Sasso, S., Ramakrishnan, C., Gamper, M., Hilvert, D. & Kast, P. (2005). FEBS J. 272, 375-389.]). The first screen at room temperature (293 K) gave a hit for condition No. 18 (0.2 M sodium acetate, cacodylate pH 6.5, 30% PEG 8000). A shower of small, well shaped and strongly birefringent single crystals appeared within 20 min of the first crystallization setup (Fig. 1[link]a). These crystals diffracted to 3–4 Å resolution. Subsequently, an intense effort was invested into optimizing the crystallization conditions and particularly into slowing down crystallization in order to obtain fewer but larger crystals. However, none of the modifications, such as lowering the protein or precipitant concentration, changing the volume ratios and temperature, using additives or layering the reservoirs with various oils etc., proved successful: we either still obtained a shower of small crystals within 20 min of setting up the drops (using a reservoir concentration of ≥20% PEG) or we did not obtain any crystals at all. Even seeding did not help. The final breakthrough came when we recalled a publication by Saridakis & Chayen (2000[Saridakis, E. & Chayen, N. E. (2000). Protein Sci. 9, 755-757.]) in which nucleation was initiated at a high precipitant concentration (which is more likely to produce nucleation centres) followed by transferring the drops before crystals became visible macroscopically to reservoirs with a lower precipitant concentration (which is favourable for crystal growth). In our case, crystallization times were extremely short, so the drops needed to be transferred within less than 20 min. After initial tests showed dramatic improvements upon implementing this strategy, we adopted a more systematic approach and devised a grid screen in which both the incubation period before transfer and the concentration of the final reservoir solution were varied. The best results were obtained when suspending the drops (1.5 µl protein + 1.5 µl reservoir solution) over a starting reservoir (1 ml) containing 20% PEG 4000 (in addition to 0.2 M sodium acetate and cacodylate buffer pH 6.5) and then transferring them after 5–15 min to reservoirs containing only 5–15% PEG 4000 (other ingredients constant). In this way, huge crystals (∼1 mm in the longest dimension) appeared overnight which did not look perfect but diffracted to beyond 1.6 Å resolution (Fig. 1[link]b). In general, lower reservoir concentrations and shorter incubation times coincided with fewer (and larger) crystals, with the reservoir concentration having the larger impact (particularly for concentrations of <10% PEG).

[Figure 1]
Figure 1
(a) Photograph of *MtCM crystals obtained straight from initial screens (crystal size ∼20 µm). (b) Photograph of *MtCM crystals from optimized crystallization conditions (crystal size ∼1 mm in the longest dimension).

To our knowledge, this is the first time that the technique of Saridakis & Chayen (2000[Saridakis, E. & Chayen, N. E. (2000). Protein Sci. 9, 755-757.]) has been applied to such short crystallization times. Obviously, in time spans as short as a few minutes, equilibration with the reservoir solution cannot occur. Nevertheless, it turned out to be critical for our experiments to initiate crystallization at a PEG concentration of 20% (before transferring the drop to lower concentrations). The reason for this is presumably a rather steep concentration gradient of PEG in the drop immediately after pipetting the reservoir solution onto the protein droplet. In the mixing zone at the interface between protein and precipitant solution, the PEG concentration will be very close to 20%, enabling the formation of crystal nuclei (`shock' nucleation). The locally high PEG concentration will subsequently decrease by diffusion, establishing conditions favorable for crystal growth, provided that the vapour pressure from the reservoir is sufficiently high to prevent rapid dehydration of the drop, thus avoiding further nucleation.

2.2. Preparation of a transition-state analogue inhibitor complex

In order to localize and investigate the active site of *MtCM, we performed soaks of the unliganded protein crystals in a solution containing an endo-oxabicyclic dicarboxylic acid (referred to here as `TSA'; Bartlett & Johnson, 1985[Bartlett, P. A. & Johnson, C. R. (1985). J. Am. Chem. Soc. 107, 7792-7793.]; Bartlett et al., 1988[Bartlett, P. A., Nakagawa, Y., Johnson, C. R., Reich, S. H. & Luis, A. (1988). J. Org. Chem. 53, 3195-3210.]). This compound is known to be an excellent mimic of the geometry of the transition state of the chorismate mutase reaction and is currently the best inhibitor for a broad range of chorismate mutases (Mandal & Hilvert, 2003[Mandal, A. & Hilvert, D. (2003). J. Am. Chem. Soc. 125, 5598-5599.]), including *MtCM (Sasso et al., 2005[Sasso, S., Ramakrishnan, C., Gamper, M., Hilvert, D. & Kast, P. (2005). FEBS J. 272, 375-389.]). Consequently, it has been used as a ligand in many structural studies of chorismate mutases (Chook et al., 1993[Chook, Y. M., Ke, H. & Lipscomb, W. N. (1993). Proc. Natl Acad. Sci. USA, 90, 8600-8603.], 1994[Chook, Y. M., Gray, J. V., Ke, H. & Lipscomb, W. N. (1994). J. Mol. Biol. 240, 476-500.]; Haynes et al., 1994[Haynes, M. R., Stura, E. A., Hilvert, D. & Wilson, I. A. (1994). Science, 263, 646-652.]; Lee, Karplus et al., 1995[Lee, A. Y., Karplus, P. A., Ganem, B. & Clardy, J. (1995). J. Am. Chem. Soc. 117, 3627-3628.]; Sträter et al., 1997[Sträter, N., Schnappauf, G., Braus, G. & Lipscomb, W. N. (1997). Structure, 5, 1437-1452.]).

Two different soaking techniques were tested. In order to minimize consumption of the precious compound, we first tried to dissolve a tiny amount of solid TSA directly in a drop containing *MtCM crystals. However, all crystals in the drop dissolved immediately. We then took 0.5 µl of this TSA-containing drop and added it to a 1.5 µl drop containing more *MtCM crystals. After 20 min soaking time, the crystals were cryoprotected by pipetting a drop of mother liquor containing 10% PEG 400 on top of the soak. Thereafter, the crystals were flash-frozen in liquid nitrogen and stored for data collection. Crystal structure analysis (Ökvist et al., 2006[Ökvist, M., Dey, R., Sasso, S., Grahn, E., Kast, P. & Krengel, U. (2006). J. Mol. Biol. 357, 1483-1499.]) confirmed the presence of TSA in the active site of *MtCM.

2.3. X-ray diffraction analysis

X-ray data were collected and processed as reported by Ökvist et al. (2006[Ökvist, M., Dey, R., Sasso, S., Grahn, E., Kast, P. & Krengel, U. (2006). J. Mol. Biol. 357, 1483-1499.]). *MtCM crystals diffracted to beyond 1.6 Å resolution (diffraction limit 1.3 Å resolution; Fig. 2[link] and Table 1[link]) and belong to the monoclinic space group P21, with unit-cell parameters a = 42.6, b = 72.6, c = 62.0 Å, β = 104.5° for the free enzyme and a = 42.9, b = 72.8, c = 61.7 Å, β = 104.0° for the TSA complex. The Matthews parameter VM is 2.9 Å3 Da−1 (Matthews, 1968[Matthews, B. W. (1968). J. Mol. Biol. 33, 491-497.]; calculated for a unit cell containing two protomers per asymmetric unit), which corresponds to a solvent content of 57%. The dimeric nature of *MtCM is consistent with previous biochemical data (Sasso et al., 2005[Sasso, S., Ramakrishnan, C., Gamper, M., Hilvert, D. & Kast, P. (2005). FEBS J. 272, 375-389.]; Prakash et al., 2005[Prakash, P., Aruna, B., Sardesai, A. A. & Hasnain, S. E. (2005). J. Biol. Chem. 280, 19641-19648.]). Since sequence alignments (Sasso et al., 2005[Sasso, S., Ramakrishnan, C., Gamper, M., Hilvert, D. & Kast, P. (2005). FEBS J. 272, 375-389.]) suggested homology to the E. coli AroQp domain, phasing was attempted by molecular replacement using the E. coli structure (PDB code 1ecm ; Lee, Karplus et al., 1995[Lee, A. Y., Karplus, P. A., Ganem, B. & Clardy, J. (1995). J. Am. Chem. Soc. 117, 3627-3628.]) as a search model. However, even exhaustive trials did not yield a solution to the problem. A probable reason for the failure is the relatively poor overall sequence homology (16% identity, according to BLASTP v.2.2.6 from NCBI; Altschul et al., 1997[Altschul, S. F., Madden, T. L., Schäffer, A. A., Zhang, J., Zhang, Z., Miller, W. & Lipman, D. J. (1997). Nucleic Acids Res. 25, 3389-3402.]) between the entire 167-­residue *MtCM polypeptide and the structurally characterized 109-residue E. coli AroQp domain, together with significant structural differences of the C-terminal helices (Ökvist et al., 2006[Ökvist, M., Dey, R., Sasso, S., Grahn, E., Kast, P. & Krengel, U. (2006). J. Mol. Biol. 357, 1483-1499.]).

Table 1
Data-collection statistics

Values in parentheses are for the highest resolution shells included in the final processing.

Data set Unliganded *MtCM TSA complex
Synchrotron beamline ESRF, ID14-4 Max II, I711
Resolution limit (Å) 31.01–1.55 (1.63–1.55) 38.92–1.64 (1.73–1.64)
Completeness (%) 99.8 (99.8) 99.5 (96.4)
Redundancy 4.1 (4.1) 4.4 (4.0)
Mean I/σ(I) 10.9 (2.5) 21.4 (7.3)
Rmerge (%) 8.5 (40.0) 5.0 (15.4)
PDB code 2fp1 2fp2
†Crystals diffracted to a maximum resolution of ∼1.3 Å for both data sets (see Fig. 2[link]); however, final data sets only include data to ∼1.6 Å for the following reasons: for unliganded *MtCM, the high-resolution data set was the third of three data sets collected from the same crystal (the preceding two data sets were collected at the MAD peak and edge) and therefore reflections at higher resolution were no longer of acceptable quality and strength. For the TSA complex, data collection was limited to 1.6 Å for technical reasons.
Rmerge = [\textstyle \sum|I_{i}-\langle I\rangle|/][\textstyle \sum I_{i}].
[Figure 2]
Figure 2
Diffraction image from a *MtCM crystal recorded at beamline ID14-4, ESRF, Grenoble. Resolution rings are indicated in (a) and a close-up view of the outer rings (b) shows that the crystal diffracts to a maximum resolution of 1.3 Å (see encircled reflection in outer resolution ring).

We have therefore prepared heavy-atom derivatives (PbII acetate and TlIII acetate soaks) and solved the structure by the MAD phasing technique based on the PbII derivative (Ökvist et al., 2006[Ökvist, M., Dey, R., Sasso, S., Grahn, E., Kast, P. & Krengel, U. (2006). J. Mol. Biol. 357, 1483-1499.]).

3. Concluding remarks

While crystallization of the same protein has been reported very recently by another group (Qamra et al., 2005[Qamra, R., Prakash, P., Aruna, B., Hasnain, S. E. & Mande, S. C. (2005). Acta Cryst. F61, 473-475.]), our results are distinct since the crystal forms are different (space group P21 instead of C2) and we have significantly higher resolution data (≤1.6 Å versus 2.1 Å). Furthermore, we describe the crystallization of free enzyme and a soak with a transition-state analogue, while Qamra and coworkers cocrystallized *MtCM with L-tryptophan.

In addition, this report addresses in detail a general and very challenging crystallization problem and its subsequent solution, which led to the development of an empirical method for crystal optimization for cases with short crystallization times. By applying the optimized protocol, the diffraction limit of the *MtCM crystals could be extended from ∼3.5 to 1.3 Å. This technique, which involves the transfer of crystallization drops to lower concentration reservoirs (varying both reservoir concentration and incubation time before transfer), is particularly useful if the protein of interest crystallizes in a narrow window of conditions and only little protein material is available.

Acknowledgements

We would like to thank Rosalino Pulido for TSA synthesis and Elin Grahn for help with data collection. At the synchrotron beamlines ID14-4, ESRF, Grenoble and I711, Max II, Lund, we had excellent support from Raimond Ravelli and Yngve Cerenius, respectively. This work has been supported by grants from the Carl Trygger foundation (research grant No. 02:158 to UK and postdoctoral fellowship of RD) from the Glycoconjugates in Biological Systems program of the Swedish National Foundation for Strategic Research (research position of UK), from Novartis Pharma (to SS) and the ETH Zürich (PK).

References

First citationAltschul, S. F., Madden, T. L., Schäffer, A. A., Zhang, J., Zhang, Z., Miller, W. & Lipman, D. J. (1997). Nucleic Acids Res. 25, 3389–3402.  CrossRef CAS PubMed Web of Science Google Scholar
First citationAndrews, P. R., Smith, G. D. & Young, I. G. (1973). Biochemistry, 12, 3492–3498.  CrossRef CAS PubMed Web of Science Google Scholar
First citationBartlett, P. A. & Johnson, C. R. (1985). J. Am. Chem. Soc. 107, 7792–7793.  CrossRef CAS Web of Science Google Scholar
First citationBartlett, P. A., Nakagawa, Y., Johnson, C. R., Reich, S. H. & Luis, A. (1988). J. Org. Chem. 53, 3195–3210.  CrossRef CAS Web of Science Google Scholar
First citationCalhoun, D. H., Bonner, C. A., Gu, W., Xie, G. & Jensen, R. A. (2001). Genome Biol. 2, 30.01–30.16.  CrossRef Google Scholar
First citationChayen, N. E. (1997). J. Appl. Cryst. 30, 198–202.  CrossRef CAS Web of Science IUCr Journals Google Scholar
First citationChook, Y. M., Gray, J. V., Ke, H. & Lipscomb, W. N. (1994). J. Mol. Biol. 240, 476–500.  CrossRef CAS PubMed Web of Science Google Scholar
First citationChook, Y. M., Ke, H. & Lipscomb, W. N. (1993). Proc. Natl Acad. Sci. USA, 90, 8600–8603.  CrossRef CAS PubMed Web of Science Google Scholar
First citationGörisch, H. (1978). Biochemistry, 17, 3700–3705.  CrossRef PubMed CAS Google Scholar
First citationGuimarães, C. R. W., Repasky, M. P., Chandrasekhar, J., Tirado-Rives, J. & Jorgensen, W. L. (2003). J. Am. Chem. Soc. 125, 6892–6899.  Web of Science CrossRef PubMed CAS Google Scholar
First citationGuo, H., Cui, Q., Lipscomb, W. N. & Karplus, M. (2003). Angew. Chem. Int. Ed. Engl. 42, 1508–1511.  Web of Science CrossRef PubMed CAS Google Scholar
First citationGustin, D. J., Mattei, P., Kast, P., Wiest, O., Lee, L., Cleland, W. W. & Hilvert, D. (1999). J. Am. Chem. Soc. 121, 1756–1757.  Web of Science CrossRef CAS Google Scholar
First citationHaslam, E. (1993). Shikimic Acid: Metabolism and Metabolites. New York: John Wiley & Sons.  Google Scholar
First citationHaynes, M. R., Stura, E. A., Hilvert, D. & Wilson, I. A. (1994). Science, 263, 646–652.  CrossRef CAS PubMed Web of Science Google Scholar
First citationKast, P., Asif-Ullah, M. & Hilvert, D. (1996). Tetrahedron Lett. 37, 2691–2694.  CrossRef CAS Web of Science Google Scholar
First citationKast, P., Asif-Ullah, M., Jiang, N. & Hilvert, D. (1996). Proc. Natl Acad. Sci. USA, 93, 5043–5048.  CrossRef CAS PubMed Web of Science Google Scholar
First citationKast, P., Grisostomi, C., Chen, I. A., Li, S., Krengel, U., Xue, Y. & Hilvert, D. (2000). J. Biol. Chem. 275, 36832–36838.  Web of Science CrossRef PubMed CAS Google Scholar
First citationKienhöfer, A., Kast, P. & Hilvert, D. (2003). J. Am. Chem. Soc. 125, 3206–3207.  Web of Science CrossRef PubMed Google Scholar
First citationLadner, J. E., Reddy, P., Davis, A., Tordova, M., Howard, A. J. & Gilliland, G. L. (2000). Acta Cryst. D56, 673–683.  Web of Science CrossRef CAS IUCr Journals Google Scholar
First citationLee, A. Y., Karplus, P. A., Ganem, B. & Clardy, J. (1995). J. Am. Chem. Soc. 117, 3627–3628.  CrossRef CAS Web of Science Google Scholar
First citationLee, A. Y., Stewart, J. D., Clardy, J. & Ganem, B. (1995). Chem. Biol. 2, 195–­203.  CrossRef CAS PubMed Web of Science Google Scholar
First citationMa, J., Zheng, X., Schnappauf, G., Braus, G., Karplus, M. & Lipscomb, W. N. (1998). Proc. Natl Acad. Sci. USA, 95, 14640–14645.  Web of Science CrossRef CAS PubMed Google Scholar
First citationMacBeath, G., Kast, P. & Hilvert, D. (1998). Biochemistry, 37, 10062–10073.  Web of Science CrossRef CAS PubMed Google Scholar
First citationMandal, A. & Hilvert, D. (2003). J. Am. Chem. Soc. 125, 5598–5599.  Web of Science CSD CrossRef PubMed CAS Google Scholar
First citationMattei, P., Kast, P. & Hilvert, D. (1999). Eur. J. Biochem. 261, 25–32.  Web of Science CrossRef PubMed CAS Google Scholar
First citationMatthews, B. W. (1968). J. Mol. Biol. 33, 491–497.  CrossRef CAS PubMed Web of Science Google Scholar
First citationMikol, V. & Giegé, R. (1992). Crystallization of Nucleic Acids and Proteins, edited by A. Ducruix & R. Giegé, pp. 219–239. Oxford University Press.  Google Scholar
First citationÖkvist, M., Dey, R., Sasso, S., Grahn, E., Kast, P. & Krengel, U. (2006). J. Mol. Biol. 357, 1483–1499.  Web of Science CrossRef PubMed Google Scholar
First citationPrakash, P., Aruna, B., Sardesai, A. A. & Hasnain, S. E. (2005). J. Biol. Chem. 280, 19641–19648.  Web of Science CrossRef PubMed CAS Google Scholar
First citationQamra, R., Prakash, P., Aruna, B., Hasnain, S. E. & Mande, S. C. (2005). Acta Cryst. F61, 473–475.  Web of Science CrossRef IUCr Journals Google Scholar
First citationSaridakis, E. & Chayen, N. E. (2000). Protein Sci. 9, 755–757.  Web of Science CrossRef PubMed CAS Google Scholar
First citationSaridakis, E. E. G., Shaw Stewart, P. D., Lloyd, L. F. & Blow, D. M. (1994). Acta Cryst. D50, 293–297.  CrossRef CAS Web of Science IUCr Journals Google Scholar
First citationSasso, S., Ramakrishnan, C., Gamper, M., Hilvert, D. & Kast, P. (2005). FEBS J. 272, 375–389.  Web of Science CrossRef PubMed CAS Google Scholar
First citationŠtrajbl, M., Shurki, A., Kato, M. & Warshel, A. (2003). J. Am. Chem. Soc. 125, 10228–10237.  Web of Science CrossRef PubMed Google Scholar
First citationSträter, N., Håkansson, K., Schnappauf, G., Braus, G. & Lipscomb, W. N. (1996). Proc. Natl Acad. Sci. USA, 93, 3330–3334.  CrossRef CAS PubMed Web of Science Google Scholar
First citationSträter, N., Schnappauf, G., Braus, G. & Lipscomb, W. N. (1997). Structure, 5, 1437–1452.  CrossRef CAS PubMed Web of Science Google Scholar
First citationStura, E. A. (1999). Protein Crystallization: Techniques, Strategies, and Tips, edited by T. M. Bergfors, pp. 141–153. La Jolla, CA, USA: International University Line.  Google Scholar
First citationStura, E. A. & Wilson, I. A. (1992). Crystallization of Nucleic Acids and Proteins, edited by A. Ducruix & R. Giegé, pp. 99–126. Oxford University Press.  Google Scholar
First citationWiest, O. & Houk, K. N. (1995). J. Am. Chem. Soc. 117, 11628–11639.  CrossRef CAS Web of Science Google Scholar
First citationXia, T., Song, J., Zhao, G., Aldrich, H. & Jensen, R. A. (1993). J. Bacteriol. 175, 4729–4737.  CAS PubMed Web of Science Google Scholar
First citationXu, H., Yang, C., Chen, L., Kataeva, I. A., Tempel, W., Lee, D., Habel, J. E., Nguyen, D., Pflugrath, J. W., Ferrara, J. D., Arendall, W. B. III, Richardson, J. S., Richardson, D. C., Liu, Z.-J., Newton, M. G., Rose, J. P. & Wang, B.-C. (2005). Acta Cryst. D61, 960–966.  Web of Science CrossRef CAS IUCr Journals Google Scholar
First citationXue, Y. & Lipscomb, W. N. (1995). Proc. Natl Acad. Sci. USA, 92, 10595–10598.  CrossRef CAS PubMed Web of Science Google Scholar
First citationXue, Y., Lipscomb, W. N., Graf, R., Schnappauf, G. & Braus, G. (1994). Proc. Natl Acad. Sci. USA, 91, 10814–10818.  CrossRef CAS PubMed Web of Science Google Scholar
First citationZhang, X., Zhang, X. & Bruice, T. C. (2005). Biochemistry, 44, 10443–10448.  Web of Science CrossRef PubMed CAS Google Scholar

© International Union of Crystallography. Prior permission is not required to reproduce short quotations, tables and figures from this article, provided the original authors and source are cited. For more information, click here.

Journal logoSTRUCTURAL BIOLOGY
COMMUNICATIONS
ISSN: 2053-230X
Follow Acta Cryst. F
Sign up for e-alerts
Follow Acta Cryst. on Twitter
Follow us on facebook
Sign up for RSS feeds