|Home | About | Journals | Submit | Contact Us | Français|
CBS domains are small protein motifs consisting of a three-stranded β-sheet and two α-helices that are present in proteins of all kingdoms of life and in proteins with completely different functions. Several genetic diseases in humans have been associated with mutations in their sequence, which has made them promising targets for rational drug design. The C-terminal domain of the Methanococcus jannaschii protein MJ0100 includes a CBS-domain pair and has been overexpressed, purified and crystallized. Crystals of selenomethionine-substituted (SeMet) protein were also grown. The space group of both the native and SeMet crystals was determined to be orthorhombic P212121, with unit-cell parameters a = 80.9, b = 119.5, c = 173.3 Å. Preliminary analysis of the X-ray data indicated that there were eight molecules per asymmetric unit in both cases.
CBS domains are small motifs of approximately 60 amino-acid residues that were originally discovered in the enzyme cystathionine-β-synthase (Bateman, 1997 ). CBS domains are not unique to CBS but can be identified in a wide variety of proteins (Pfam database PF00571; http://www.sanger.ac.uk/Users/agb/CBS/CBS.html), alone or combined with other domains performing different functions. CBS domains are usually associated in tandems, forming a so-called CBS-domain pair or Bateman domain, although some proteins such as 5′-AMP-activated protein kinase (AMPK) contain four CBS domains (Ignoul & Eggermont, 2005 ). The first CBS-domain structure to be solved was the structure of inosine monophosphate dehydrogenase (IMPDH; Sintchak et al., 1996 ; Zhang et al., 1999 ). This structure showed that a single CBS domain consists of a three-stranded β-sheet and two α-helices packed according to the sequence β1-α1-β2-β3-α2. The crystal structures of CBS domains from many proteins are now available, such as the chloride channels ClC-0 from the ray Torpedo marmorata (Meyer & Dutzler, 2006 ), human ClC-5 (Meyer et al., 2007 ) and human ClC-Ka (Markovic & Dutzler, 2007 ), human AMPK (Day et al., 2007 ), yeast AMPK (Amodeo et al., 2007 ; Jin et al., 2007 ; Rudolph et al., 2007 ; Townley & Shapiro, 2007 ), truncated rat–human AMPK (Xiao et al., 2007 ) and the bacterial Mg transporter MgtE from Thermus thermophilus (Hattori et al., 2007 ). Comparison of the CBS domains of the solved structures shows a highly conserved fold despite low sequence similarity. In all these structures, two CBS domains associate to form a compact structure with a cleft between the domains. This cavity has been proven to be the binding site for adenosyl groups in ClC5 and AMPK. The crystal structures of the complex ClC5–ADP/ATP (Meyer et al., 2007 ) and of the complex AMPK–ATP/AMP/ADP/ZMP (Amodeo et al., 2007 ; Day et al., 2007 ; Jin et al., 2007 ; Townley & Shapiro, 2007 ; Xiao et al., 2007 ) raise the question of whether a CBS-domain pair has one or two nucleotide-binding sites. Although there is experimental evidence that the binding of adenosyl groups to the CBS domain of IMPDH (Scott et al., 2004 ), AMPK (Cheung et al., 2000 ) and CBS (Finkelstein et al., 1975 ) has an influence on the catalysis of these proteins, the structural information available does not elucidate how the binding of the adenosyl compound is transduced to the corresponding catalytic domains. It seems that the physiological functions and binding partners of the Bateman domains may vary considerably between different proteins (Pimkin & Markham, 2008 ).
Important human pathologies associated with mutations within the CBS domains have been described as homocystinuria (CBS; Shan et al., 2001 ), autosomal retinitis pigmentosa (IMPDH; Hunter et al., 2002 ), myotonia (ClC-1; Pusch, 2002 ), idiopathic epilepsy (ClC-2; Haug et al., 2003 ), Bartter syndrome (ClC-Kb; Konrad et al., 2000 ), osteopetrosis (ClC-7; Cleiren et al., 2001 ) and familiar hypertrophic cardyomiopathy or Wolff–Parkinson–White syndrome (AMPK; Blair et al., 2001 ), which enhances the chances that these domains will be promising targets for the development of novel drugs.
Some hyperthermophilic microorganisms from archaea, including Methanococcus jannaschii (Bult et al., 1996 ), possess a large number of proteins with CBS domains in their genome. Therefore, they are excellent models for the study and characterization of the binding sites for different types of adenosyl groups in CBS domains, as well as for the study of the specificity of these proteins for their ligands. In this work, we present the cloning, expression, purification and crystallization of the CBS-domain pair of M. jannaschii protein MJ0100. The ORF of gene mj0100 (UniProtKB/Swiss-Prot Q57564) codes for a polypeptide chain of 509 amino acids with a molecular weight of 56 458 Da. Its sequence contains two domains: a DUF39 (Pfam database PF01837) domain (residues 15–320) of so far unknown function found in bacteria and archaea and a CBS-domain pair (residues 392–509) (Fig. 1 ). There are 15 sequences belonging to archaea with DUF39 + CBS architecture (Fig. 1 b).
An initial attempt to purify full-length MJ0100 protein showed that it was insoluble under the conditions assayed. Therefore, we decided to study only the CBS-domain pair of the protein (MJ0100c; MW = 14 426 Da) consisting of residues 381–509. Plasmid pML1 carries the CBS-domain pair of MJ0100 (residues 381–509) under a T7 promoter. This plasmid was constructed by amplifying from M. jannaschii genomic DNA the 397 bp fragment that corresponds to the CBS-domain pair with oligonucleotides MJ0100-381F (CACCATGAAGCCAATGAAGTCACCAATAAC) and MJ0100R (TCATTTTTTCCCTCCGAATAATC). The PCR product was cloned in pET101D plasmid using the Champion pET Directional TOPO Expression Kit (Invitrogen). The DNA sequence was verified by sequencing and then transformed into Escherichia coli strain BL21-Codon Plus (Stratagene).
MJ0100c was purified starting from a 2 l BL21-Codon Plus (Stratagene) culture containing plasmid pML1 grown at 310 K with 100 µg ml−1 ampicillin and 25 µg ml−1 chloramphenicol. Overexpression of MJ0100c was induced by the addition of IPTG to a final concentration of 0.5 mM when the optical density of the culture at 600 nm was 0.6. Expression was allowed to proceed for 3 h. Cells were harvested by 10 min centrifugation at 3000g and 277 K and stored at 193 K. Cells were thawed at room temperature and resuspended in 80 ml lysis buffer (50 mM HEPES pH 7.0, 1 mM EDTA, 1 mM DTT, 1 mM benzamidine, 0.1 mM PMSF). Bacterial lysis was carried out by ultrasonication in a Labsonic P instrument (Sartorius). After centrifugation at 145 000g for 30 min at 277 K, protein MJ0100c remained mainly as a soluble protein. The resulting supernatant was loaded onto a P11-phosphocellulose (Whatman) column equilibrated with buffer A1 (50 mM HEPES pH 7.0, 200 mM NaCl, 1 mM EDTA, 1 mM DTT). The column contained 15 ml P11-phosphocellulose previously activated according to the manufacturer’s specifications and packed in a K9 column (GE Healthcare). Unbound proteins were washed off with 60 ml buffer A1. Elution of the target protein was achieved with buffer B1 (50 mM HEPES pH 7.0, 750 mM NaCl, 1 mM EDTA, 1 mM DTT). 5 ml fractions were collected. The fractions containing MJ0100c were pooled and dialyzed against buffer A2 (50 mM HEPES pH 7.0, 50 mM NaCl, 1 mM EDTA, 1 mM DTT). The dialyzed protein was applied onto a 1 ml MonoS column (GE Healthcare) equilibrated with buffer A2. After washing with 15 column volumes of buffer A2, MJ0100c was eluted with a linear 30 ml NaCl gradient (50–600 mM NaCl) using buffer B2 (50 mM HEPES pH 7.0, 600 mM NaCl, 1 mM EDTA, 1 mM DTT) at a flow rate of 1 ml min−1 and collecting 1.0 ml fractions. In a final polishing step, concentrated MJ0100c was applied onto a high-resolution gel-filtration column (Superdex-75 10/300, GE Healthcare) equilibrated in buffer A1. The chromatography was performed at a flow rate of 0.5 ml min−1, collecting 1.0 ml fractions. The protein eluted as a single peak with a molecular weight of 26 kDa, which is in concordance with the formation of a dimer (28.8 kDa). Dynamic light-scattering (DLS) analysis confirmed that MJ0100c (at a protein concentration of 1 mg ml−1) is a dimer in solution in the pH range 5–9 (data not shown). The concentration of the purified MJ0100c protein was determined by UV absorption at 280 nm using the theoretical extinction coefficient computed from the amino-acid sequence (280 = 8250 M −1 cm−1; Gill & von Hippel, 1989 ). The typical yield from a 2 l expression was 20 mg purified protein. The MJ0100c protein was flash-frozen in liquid nitrogen and stored at 193 K. SDS–PAGE (Laemmli, 1970 ) was used to analyze the protein purity (Fig. 2 ).
Purification of selenomethionine (SeMet) labelled MJ0100c was performed using E. coli B834 (DE3) strain (Novagen; Wood, 1966 ) and New Minimal Medium supplemented with seleno-l-methionine (SeMet; Budisa et al., 1995 ). The purification was performed according to the previously described procedure. The yield was slightly lower than that of unlabelled material.
SDS–PAGE gel bands were subjected to in-gel tryptic digestion according to Shevchenko et al. (1996 ) with minor modifications. The gel piece was swollen in a digestion buffer containing 50 mM NH4HCO3 and 12.5 ng µl−1 trypsin (Roche Diagnostics) in an ice bath. After 30 min, the supernatant was removed and discarded, 20 µl 50 mM NH4HCO3 was added to the gel piece and digestion was allowed to proceed at 310 K overnight. Prior to mass-spectrometric analysis, the sample was acidified by adding 5 µl 0.5% TFA. 0.5 µl digested sample was directly spotted onto the MALDI target and then mixed with 0.5 µl α-cyano-4-hydroxycinnamic acid matrix solution [20 µg µl−1 in acetonitrile, 0.1% TFA, 70:30(v:v)]. Peptide mass fingerprinting was performed on a Bruker Autoflex III mass spectrometer (Bruker Daltonics). Positively charged ions were analyzed in reflector mode using delayed extraction. The spectra were obtained by randomly scanning the sample surface. About 600–800 spectra were averaged in order to improve the signal-to-noise ratio. Spectra were externally calibrated, resulting in a mass accuracy of <50 p.p.m. when external calibration was performed and typically <20 p.p.m. in the case of internal calibration. Protein identification was performed by searching in a nonredundant protein database (NCBI) using the MASCOT search engine (http://matrixscience.com). The following parameters were used for database searches: one missed cleavage with allowed modifications carbamidomethylation of cysteine (complete) and oxidation of methionine (partial). Ultimately, MJ0100c expressed in E. coli was wild type with no mutations in the amino-acid sequence according to the M. jannaschii genome sequence database. Mass spectrometry indicated that the extent of SeMet labelling was >95% (data not shown).
Purified protein was dialyzed against 50 mM HEPES pH 7.0 and concentrated to 100–150 mg ml−1 using Vivaspin columns (Sartorius) with a molecular-weight cutoff of 5 kDa. Crystallization trials were set up using the hanging-drop vapour-diffusion technique in 24-well VDX plates (Hampton Research) at 293 K. Drops consisting of 1 µl protein solution mixed with 1 µl reservoir solution were equilibrated against 500 µl reservoir solution. Suitable protein concentrations were determined with the native protein using the Pre-Crystallization Screen (PCT, Hampton Research). Initial crystallization conditions were found using Crystal Screen (Hampton Research). Crystals were obtained when the precipitant was 8% PEG 4000 in Tris–HCl pH 8.5 (condition No. 36). These crystallization conditions were optimized with respect to PEG and pH. The best crystallization conditions were 10–15% PEG 6000, 8000 or 10 000 in 100 mM Tris–HCl. No differences were observed in the pH range of the Tris–HCl buffer between 8.4 and 9.4. The crystals appeared overnight and grew to maximum dimensions of about 1 × 0.7 × 0.5 mm within 2–3 d (Fig. 3 a). Crystals which appeared to have two different habits grew in the same crystallization drop, but both had identical unit-cell parameters and space-group symmetry. Crystals of SeMet-substituted MJ0100c (Fig. 3 b) were obtained using the same crystallization conditions as used for the native crystals.
Finding optimal cryogenic conditions to freeze MJ0100c crystals was not straightforward and required significant experimental effort. Owing to their size, the largest crystals grown by the hanging-drop or sitting-drop techniques were not suitable for freezing and usually (in 95% of the trials) fractured just after opening the cover slip of the reservoir or, in the best case, after a brief exposure (1–2 s) to various cryoprotectants including MPD, low-molecular-weight PEGs, sucrose, erythritol, glycerol or ethylene glycol. In order to prevent freezing problems arising from physical tension between the inner and outer regions of the sample, we only froze crystals in the size range 0.1–0.5 mm (in the largest dimension). The best result was obtained by soaking the crystals in crystallization solution containing a final glycerol concentration of 25% and a slight increase (5%) in the precipitant. Unfortunately, this approach only yielded optimal results in 5% of the samples tested. In parallel, crystals were also harvested with stepwise transfer via a series of solutions containing increasing concentrations of the cryoprotective agent for varying times with no further success. Dehydration of the crystals by slowly increasing the PEG concentration (in steps of 5%) in the reservoir solution for varying times (from 1 h to one week) prior to flash-cooling in liquid nitrogen and/or to the addition of the cryoprotectants was also tested, with no further success. In parallel, we tried a different approach to freeze the crystals, consisting of covering the hanging-drop crystallization drops with paraffin oil, Paratone-N or silicon oil DC200 (Fluka) just after opening the cover slip of the crystallization well. This technique yielded optimal results and avoided crystal dehydration and fracture (Pflugrath, 2004 ). Once the crystallization drop had been covered with oil, the crystals were carefully moved from the inner aqueous solution to the surrounding silicon oil, where they seemed to be stable. Water surrounding the crystal surface was then carefully removed by gently displacing the crystal within the silicon oil with the help of a loop. The crystals could then be flash-cooled by immersion in liquid N2 prior to data collection. To check whether the cryoprotectants (glycerol and silicone oil) might be responsible for the limited resolution of the data sets, we analyzed the diffraction images at both room temperature and after freezing the crystals. Similar results were obtained, suggesting that the resolution limit might be intrinsic to the crystals despite their apparent external beauty.
Derivative crystals were also very fragile but could be frozen by transferring them for few seconds into a cryosolution containing 27.5% glycerol, 7.5% PEG 6000, 50 mM Tris–HCl pH 8.6. Longer soaking periods resulted in crystal breakage. The crystals were then flash-cooled in liquid nitrogen.
Crystals were pre-screened using in-house X8 Proteum equipment (Bruker). A native data set was collected to 3.1 Å resolution at 100 K on beamline ID14-2 at the European Synchrotron Research Facility (ESRF), Grenoble, France from a crystal grown in 15% PEG 10 000, 100 mM Tris–HCl pH 8.8 (Fig. 4 a). The oscillation range was 1.0° and 200 images were collected on an ADSC Q4R CCD detector. Two-wavelength SeMet MAD data were collected to 3.4 Å resolution at 100 K at station BM16 at the ESRF synchrotron (Grenoble, France) from a crystal grown in 10% PEG 8000, 100 mM Tris–HCl pH 8.6, 1 mM DTT. The oscillation range was 1.0° and 180 images were collected at peak and inflection-point wavelengths on an ADSC Quantum-210 detector (Fig. 4 b). Data sets were processed using the HKL-2000 package (Otwinowski & Minor, 1997 ).
Native crystals of MJ0100c diffracted to 3.1 Å resolution (Fig. 4 a) and belonged to space group P212121, with unit-cell parameters a = 80.9, b = 119.5, c = 173.3 Å. The presence of six, eight or ten molecules within the asymmetric unit give a Matthews coefficient of 4.86, 3.67 and 2.92 Å3 Da−1, respectively (Matthews, 1968 ) and a solvent content of 74.7, 66.5 and 57.85%, respectively. After careful analysis of the self-rotation function (Fig. 5 ) and considering the fragility of MJ0100c crystals and their limited diffraction power, we estimated that eight molecules per asymmetric unit was the most probable value. The crystals of MJ0100c labelled with SeMet diffracted to 3.4 Å resolution (Fig. 4 b) and belonged to space group P212121, with unit-cell parameters a = 82.4, b = 120.6, c = 175.5 Å; the asymmetric unit was also likely to contain eight monomers. This is consistent with a Matthews coefficient of 3.7 Å3 Da−1 and a solvent content of 67.1%. The data-collection statistics are summarized in Table 1 .
We thank Professor Sung-Hou Kim from University of California at Berkeley for providing us with the genomic DNA from M. jannaschii and the staff of ESRF beamlines BM16 and ID14-2 for support during synchrotron data collection. We also thank Beatriz González Callejas from CIC bioGUNE for maintenance of the in-house X-ray equipment. This research was supported by program grants from the Basque Government (ETORTEK IE05-147, IE07-202), Diputación Foral de Bizkaia (Exp. 7/13/08/2006/11 and 7/13/08/2005/14) and the Spanish Ministry of Education (SAF2005-00855), as well as a postdoctoral fellowship from CIC bioGUNE.