Cadmium-induced crystallization of proteins: II. Crystallization of the Salmonella typhimurium histidine-binding protein in complex with L-histidine, L-arginine, or L-lysine

Sergei Trakhanov ^1,2, D. I. Kreimer ¹, Sean Parkin ², G. Ferro-Luzzi Ames ¹ and Bernhard Rupp ^2,*

¹ Department of Molecular and Cellular Biology, Division of Biochemistry and Molecular Biology, University of California, Berkeley, CA 94720

² Lawrence Livermore National Laboratory, Biology and Biotechnology Research Program, L-452, University of California, Livermore, CA 94550

ABSTRACT

To further investigate favorable effects of bivalent cations on the formation of protein crystals, three complexes of Salmonella typhimurium histidine-binding protein were crystallized with varying concentrations of Cadmium salts. For each of the three histidine-binding protein complexes, Cadmium cations were found to promote or improve crystallization. The optimal Cadmium concentration is ligand specific and falls within a narrow concentration range. In each case, crystals grown in the presence of Cadmium diffract to better than 2.0 Å resolution and belong to the orthorhombic space group P2₁2₁2₁. From our results and from the analysis of Cadmium sites in well- refined protein structures, we propose that Cadmium addition represents a generally useful technique to modify crystal morphology and to improve diffraction quality.

INTRODUCTION

The search for favorable crystallization conditions yielding well diffracting crystals still remains a mostly empirical, often time-consuming procedure in the course of a crystallographic structure determination (Durbin and Feher, 1996). It has long been recognized that bivalent cations in general can facilitate protein crystallization (McPherson, 1982, 1990). In the particular case of Cadmium, its use to promote crystallization of ferritin dates back even before the advent of protein crystallography itself. The addition of a 10% CdSO₄ solution to horse spleen tissue juice immediately yields cubic ferritin crystals, whose growth can be watched under a cover slide (Michaelis, 1947, and earlier references cited therein).

A first systematic investigation of the influence of bivalent Cations has been published recently (Trakhanov & Quiocho, 1995). Using the leucine-, isoleucine-, valine-binding protein (LIVBP) in complex with L-leucine as a model system Cadmium cations (Cd ²⁺) were found to exhibit the most profound effect on crystallization of the 10 bivalent metal ions examined. A favorable effect of Cd ²⁺ has been also observed in the crystallization of several other bacterial periplasmic binding proteins: the histidine-binding protein from Eschericia coli in complex with L-histidine (Yao et al., 1994); leucine-, isoleucine-, valine-binding protein in complex with L-isoleucine (Trakhanov et al., unpublished results); the leucine-specific protein in complex with L-leucine (Trakhanov & Quiocho, 1995); the dipeptide-binding protein (Nickitenko et al., 1995); and the oligopeptide-binding protein in complex with a tripeptide (Nickitenko &Trakhanov, unpublished results). In this communication we investigate the ability of Cd²⁺ to facilitate the crystallization of another member of the bacterial binding protein family, the histidine-binding protein from Salmonella typhimurium (HisJ) in complex with three high-affinity ligands.

HisJ is the water-soluble component of the histidine permease, one of the most extensively characterized traffic ATPases (ABC transporters, Liu et al., 1997). HisJ is located in the periplasm and specifically binds and delivers exogenous amino acids (L-histidine, L-arginine and L-lysine) to the membrane-bound complex of the permease for subsequent translocation of the ligand into the cytoplasm. HisJ assumes slightly different conformations in complex with each of its ligands, as revealed by difference absorbance and fluorescence spectroscopy (Trakhanov et al., 1989; Wolf et al., 1996), time resolved fluorescence spectroscopy (Malak & Lakowicz, unpublished observation), and different recognition by a conformation-specific antibody (Wolf et al., 1994; 1996).

Structures of HisJ complexes will also provide a unique opportunity for structural comparative studies. The high-resolution structures (1.7-1.9 Å) of the complexes of lysine-, arginine-, ornithine-binding protein (LAO) in the apo form and with four ligands, L-lysine, L-arginine, L-ornithine and L-histidine, have been solved (Oh et al., 1993). LAO shares 70% sequence identity with HisJ and is functionally closely related, but the structure of a HisJ complex is only available at a lower resolution of 2.5 Å (Oh et al., 1994). It is of great interest therefore to obtain crystals of HisJ in complex with other ligands at high resolution.

MATERIALS AND METHODS

Purification of HisJ

HisJ was extracted from Salmonella typhimurium periplasm by osmotic shock, and purified by two-step ammonium sulfate precipitation and DEAE-HPLC (Nikaido & Ames, 1992) on a Bio-Gel DEAE-5-PW column (Bio-Rad Laboratories, Richmond, CA). The flow rate was 5 ml/min, and approximately 15 mg of 85% pure HisJ in 5 mM Tris-HCl buffer (pH 8.3) were injected. The column was washed for 1 h with the same buffer prior to elution with a NaCl gradient from 0 to 30 mM. This procedure yields highly purified HisJ essentially free of endogenous ligands, as confirmed by fluorescent titration (Wolf et al., 1995; 1996). Unliganded HisJ was concentrated to 75 mg/ml in 5 mM Tris-HCl buffer pH 8.3 using Centricon-10 concentrators (Amicon, Inc., Beverly, MA), yielding stock protein solution. Protein stock was stored on ice for up to two months retains its full binding activity and ability to produce crystals.

Crystallization

A five-fold dilution of the protein stock was prepared with 10 mM Tris-HCl buffer (pH 8.0) to which 1 mM L-histidine, 1 mM L-arginine, or 5 mM L-lysine were added. Crystals were grown at room temperature using the hanging drop vapor diffusion method (McPherson, 1982) using 1 ml wells and drops of varying sizes (7-12 ml). The drops consisted of 5ml of protein stock and 2-7ml of well solution.

RESULTS AND DISCUSSION

Crystallization

The search for crystallization conditions was performed in two steps. First, the concentrations of Cd ²⁺ and PEG-400 were screened widely while the initial volume of a crystallization drop was kept constant (4 µl of the stock protein solution were mixed on a siliconized cover slip with 4 µl of the well solution). In a second step, the best conditions found in the initial PEG/Cd ²⁺ screen (well solutions containing 22-25% (v/v) PEG-400, 50 mM Bis-tris propane at pH 8.5, and 20 mM CdSO4) were used as a starting point for optimization of the drop size and the Cadmium concentration.

The Cadmium concentration was varied down the columns of a Linbro matrix by adding another 10-100 µl of 1 M Cd(OAc)₂ to the well while along the rows the drop size and the ratio of well solution to protein stock was varied as follows: 5 m l of protein stock solution were mixed on a cover slide with increasing amounts of well solution (2 to 7 µl in 1 m l increment). This setup effectively screened for unique crystallization conditions in each of the 24 protein drops in the crystallization plate.

The nature of the anion in the Cadmium salt appears to have no effect on the crystallization. Similar results are obtained with either CdSO₄ alone, or a mixture of CdSO₄ and Cd(OAc)₂. In experiments with LIVBP, we achieved a similar improvement in crystal morphology using CdI₂ and CdCl₂. The optimal Cd ²⁺ concentations were determined individually for each of the three HisJ complexes. In the course of the final optimization glycerol was screened in an effort to improve cryo-protection. However, an increase in glycerol concentration above 5% had a detrimental effect on the crystal morphology (yielding thin plates), especially for the L-arginine and L-lysine complexes. Therefore, 5 % glycerol was used in all trials for these two complexes.

The HisJ complex with L-histidine crystallized from a solution containing 1mM L-histidine, 25 % PEG-400, 20 mM CdSO4, 50 mM Bis-tris propane (pH 8.5) and 5-15% glycerol within 2-3 days. In the presence of 15% glycerol, the crystals grew as plates with average dimensions of 0.5 ´ 0.4 ´ 0.08 mm3 (Fig. 1 A). Their morphology changed favorably to rectangular rods with a decrease in glycerol concentration to 5% and an increase in the initial concentration of Cd ²⁺ in the protein drop to 35-40 mM (Fig. 1 B). These crystals, with average dimensions 0.3 ´ 0.3 ´ 1.0 mm3, were used for data collection.

The HisJ-L-arginine complex (Fig. 1 C) crystallized from a solution of 25% PEG-400, 50 mM Bis-tris propane pH 8.5, 30 mM Cd ²⁺, and 5% glycerol within 2-3 days. The crystals formed long thick prisms with average dimensions of 0.3 ´ 0.3 ´ 0.7 mm3. For the complex of HisJ with L-lysine the best results were obtained in the presence of an initial concentration of 17.5 mM Cd ²⁺ in the protein drop (Fig. 1D). The crystals grew in the shape of single prisms to their final size 0.3 ´ 0.3 ´ 0.8 mm3 in 3 to 4 days at room temperature. Two HisJ-L-Lysine crystals were used to collect a data set complete to 1.8 Å resolution.

Figure 1. Crystals of histidine-binding protein from S. typhimurium in complex with L-histidine (A,B), L-arginine (C) and L-lysine (D), magnification 40 ´ (A,B,C) or 60 ´ (D). Concentration of glycerol in the well solution is 5% (B, C, D) and 15% (A).

Crystallographic analysis

Due to the high PEG content and the glycerol present in the mother liquor, the crystals could be flash cooled directly in liquid nitrogen and were transferred with cryo-tongs into the cold nitrogen stream of a modified Siemens LT-2 low-temperature apparatus attached to a ADSC dual multi-wire detector system. All three HisJ complexes investigated in this study crystallize in the orthorhombic space group P2₁2₁2₁ (Table 1).

Table 1. Unit cell dimensions and diffraction limits of HisJ in complex with L-histidine, L-arginine and L-lysine at 125 K, and room temperature data for crystals of S. typhimurium HisJ-L-histidine grown in the absence of any Cadmium (Kang et al., 1992). Despite the use of a synchrotron source for data collection (Oh et al., 1994), the latter crystals diffracted to considerably lower resolution (2.5 Å).

The structure of the unliganded E.coli histidine-binding protein (Yao et al., 1994), determined at 1.9 Å resolution, was selected as an initial phasing model for molecular replacement. The molecular replacement solution was unambiguous. The model structures for each of the three HisJ complexes are currently being refinement. There is also an effort underway to crystallize complexes of HisJ with each of two low-affinity ligands, L-ornithine and D-histidine (K_D ca. 1 mM).

The role of Cadmium

The analysis of a number of high-resolution structures of proteins crystallized employing Cadmium ions, namely the histidine-binding protein from E. coli (Yao et al., 1994), the dipeptide-binding protein (Nickitenko et al., 1995), and the leucine-binding proteins (Trakhanov & Quiocho, unpublished data) provides insight into the role of Cadmium in crystal formation. Cadmium atoms are positioned at the interface between two neighboring protein molecules and coordinated by two carboxyl groups of glutamic or aspartic acid side chains each belonging to one of the two molecules. This coordination is often complemented by coordination bonds formed with carbonyl groups of the polypeptide backbone and/or with a water molecule.

Similar Cadmium coordinations at intermolecular contacts are present in crystals of ferritin. In native cubic horse spleen ferritin (Grainier et al, 1997), two Cadmium ions occupy special positions on a 3-fold axis and are coordinated by three glutamine or three aspartic acid residues, repectively. A third Cd ²⁺ occupies a 2-fold axis and is coordinated by Gln and Asp from different molecules. A complex arrangement involving partially occupied Cd sites has been observed in a crystal of cubic L-chain horse spleen ferritin (Hempstead et al, 1997) where a His side chain is involved in the Cd coordination in addition to Glu and Asp.

Whether the imidazole group of a histidine side chain is involved in Cd coordination depends upon the protonation status of the histidine. In the HisJ-His complex, histidine is essentially deprotonated at pH 8.5, thus providing the coordination bond for Cadmium (Figure 2), whereas at pH below 6 protonation of the imidazole group does not allow the formation of a coordination bond. The pH – dependence of the formation of a coordination bond with Cadmium can thus provide an additional means to modify crystal morphology. Since none of the interactions between the amino acid residues and the Cadmium ions are unique to HisJ and related complexes, the formation of intermolecular bridges by Cadmium atoms should be of general use in protein crystallization.

Figure 2. Parallel stereo view of Cd coordination at an intermolecular contact region in the HisJ-His complex in partially refined structure. Cd is coordinated between two molecules by two acidic residues (Asp-187 of molecule 1, Glu-136# from molecule 2), as well as His-129 from molecule 1, and a water molecule, WAT 79. This figure was generated using MOLSCRIPT (Kraulis 1991).

CONCLUSIONS

The addition of Cadmium can provide an advantage in growing well diffracting crystals of a variety of proteins in a form suitable for high resolution crystallographic studies. The HisJ complexes reported in this work provide three examples of the improvement in crystal quality obtained from the addition of Cd ²⁺. From the analysis of refined structures, we have strong evidence that the propensity of Cadmium cations to form coordination bonds enables the development of Cadmium bridges across the interfaces of protein molecules. Side groups of various amino acids, water molecules, and carbonyl groups of the backbone chain can serve as coordination partners. The unique pattern of the negative charge distribution on the surface of any given protein molecule will necessitate the variation of the Cadmium concentration to create the conditions essential for Cd ²⁺ coordination and crystal formation. Cadmium concentrations can range from as low as 1 mM (for the leucine-, isoleucine-, valine-binding protein (Trakhanov & Quiocho, 1995)) to as high as 200 mM (for the histidine-binding protein from E. coli (Yao et.al, 1994)).

It is conceivable that substantial changes could be induced by Cadmium binding, especially at high concentrations. So far, we have no evidence of significant changes in the three-dimensional structure of the leucine-, isoleucine-, valine-binding protein solved with Cadmium compared to these structures without Cd ²⁺ (S.T., unpublished results). In addition, the analysis of seven Cadmium-binding sites in the 1.9 Å crystal structure of the histidine-binding protein from E. coli has confirmed that all Cd sites are located at the molecular interface and not inside the protein, regardless of the high Cd concentration used (Yao et.al., 1994). Additional structural details should become available when the refinements of our HisJ-ligand complexes are completed.

ACKNOWLEDGMENTS

We thank Dr. Alex Nickitenko for assistance in data collection of HisJ in the complex with L-histidine and Prof. F.A. Quiocho for stimulating discussion and advice. The work on the utilization of Cadmium ions in protein crystallization was initiated by one of us (S.T.) while in Prof. F.A. Quiocho’s laboratory at Baylor College of Medicine, Houston. Part of the work was performed under the auspices of the U.S. Department of Energy at Lawrence Livermore National Laboratory under Grant No. W-7405-ENG-48. Brent Segelke has helped us with the manuscript preparation.

REFERENCES

Durbin SD, Feher G. 1996. Protein crystallization. Annu Rev Phys Chem 47:171-204.

Hempstead PD, Yewdall SJ, Fernie AR, Lawson DM, Artyniuk PJ, Rice DW, Ford GC, Harrison PM. 1997. Comparison of the Three-dimensional Structures of Recombinant Human H and Horse L Ferritins at High Resolution. J Mo. Biol 268: 424-448.

Granier T, Gallois B, Dautant A, Langlois D'Estaintot B, Precigoux G. 1997. Cubic Crystal Structure Of Native Horse Spleen Ferritin. Brookhaven Protein Data Bank entry 1EIR.

Kang CH, Gokcen S, Ames, GF-L. 1992. Crystallization and Preliminary X-ray Studies of the Liganded Lysine, Arginine, Ornithine-binding Protein from Salmonella typhimurium. J Mol Biol 225:1123-1125.

Kraulis PJ. 1991. MOLSCRIPT: a program to produce both detailed and schematic plots of protein structures. J Appl Crystallogr 24:946-950.

McPherson A. 1990. Current approaches to macromolecular crystallization. Eur J Biochem 189: 1-23.

McPherson A. 1982. Preparation and Analysis of Protein Crystals. New York, USA: John Wiley and Sons.

Michaelis L. 1947. Ferritin and Apoferritin. Adv Prot Chem 33: 53-66.

Nickitenko AV, Trakhanov S, Quiocho FA. 1995. 2 Å Resolution Structure of DppA, a Periplasmic Dipeptide Transport/Chemosensory Receptor. Biochem 34:16585-19595.

Nikaido K, Ames G F-L. 1992. Purification and characterization of the periplasmic lysine-, arginine-, ornithine-binding protein (LAO) from Salmonella typhimurium. J Biol Chem. 267:20706-20712.

Oh B-H, Kang C-H, De Bondt H, Kim S-H, Nikaido K, Joshi AK, Ames GF-L. 1994. The Bacterial Periplasmic Histidine-binding Protein. Structure/Function Analysis of the Ligand -binding Site and Comparison with Related Proteins. J Biol Chem 269:4135-4143.

Oh B-H, Pandit J, Kang C-H, Nikaido K, Goksen S, Ames GF-L, Kim S-H. 1993. Three-dimensional Structures of the Periplasmic Lysine/Arginine/Ornithine-binding Protein with and without a Ligand. J Biol Chem 268:11348-11355.

Parkin S. 1997. Low Temperature Methods in Biocrystallography That Work. ACA Annual Meeting, July 19-25 1997, St.Louis, American Crystallographic Association Annual Meeting Series 2:24, 67.

Trakhanov SD, Chirgadze NY, Yusifov EF. 1989. Crystallization and preliminary X-ray crystallographic data of a histidine-binding protein from Escherichia coli. J Mol Biol 207: 847-849.

Trakhanov S, Quiocho FA. 1995. Influence of divalent cations in protein crystallization. Protein Sci 4:1914-1919.

Wolf A, Lee KC, Kirsch JF, Ames GF-L. 1996. Ligand-dependent conformational plasticity of the periplasmic histidine-binding protein. J Biol Chem 271: 21243-21250.

Wolf A, Shaw EW, Oh B-H, De Bondt H, Joshi AK, Ames GF-L. 1995. Structure/function analysis of the periplasmic histidine-binding protein. J Biol Chem 270:16097-16106.

Wolf A, Shaw EW, Nikaido K, Ames GF-L. 1994. The histidine-binding protein undergoes conformational changes in the absence of ligand as analyzed with conformation-specific monoclonal antibodies. J Biol Chem 269:23051-23058.

Yao N, Trakhanov S, Quiocho FA. 1994. Refined 1.89 Å Structure of the Histidine-Binding Protein Complexed with Histidine and Its Relationship with Many Other Active Transport/Chemosensory Proteins. Biochem 33:4769-4779.

Back to X-ray Facility Introduction
LLNL Disclaimer
This World Wide Web site conceived and maintained by Bernhard Rupp (br@llnl.gov)
Last revised April 06, 1999 11:05
UCRL-MI-125269