Structural Biology Communications Crystallization and Preliminary X-ray Diffraction Analysis of the Periplasmic Domain of the Escherichia Coli Aspartate Receptor Tar and Its Complex with Aspartate

The cell-surface receptor Tar mediates bacterial chemotaxis toward an attractant, aspartate (Asp), and away from a repellent, Ni 2+. To understand the molecular mechanisms underlying the induction of Tar activity by its ligands, the Escherichia coli Tar periplasmic domain with and without bound aspartate (Asp-Tar and apo-Tar, respectively) were each crystallized in two different forms. Using ammonium sulfate as a precipitant, crystals of apo-Tar1 and Asp-Tar1 were grown and diffracted to resolutions of 2.10 and 2.40 A ˚ , respectively. Alternatively, using sodium chloride as a precipitant, crystals of apo-Tar2 and Asp-Tar2 were grown and diffracted to resolutions of 1.95 and 1.58 A ˚ , respectively. Crystals of apo-Tar1 and Asp-Tar1 adopted space group P4 1 2 1 2, while those of apo-Tar2 and Asp-Tar2 adopted space groups P2 1 2 1 2 1 and C2, respectively.


Introduction
The aspartate receptor Tar is a transmembrane protein of 553 aminoacid residues in length consisting of a first membrane-spanninghelix, a periplasmic domain, a second membrane-spanning -helix and a cytoplasmic domain (Krikos et al., 1983). Tar mediates bacterial chemotaxis towards attractants including aspartate (Asp) and maltose, and away from repellents such as nickel and cobalt ions (Reader et al., 1979;Wang & Koshland, 1980). Transmembrane signalling by Tar requires a homodimeric structure of the receptor (Milligan & Koshland, 1988). How the attractants and repellents regulate the activity of Tar during chemotaxis is still unknown.
The first crystal structures of the Tar periplasmic domain were determined using cysteine cross-linked dimers in the absence and presence of bound aspartate in Salmonella typhimurium (Milburn et al., 1991). Other crystal structures of the Tar periplasmic domain in the Asp-bound and unbound forms were subsequently reported (Bowie et al., 1995;Chi et al., 1997;Scott et al., 1993;Yeh et al., 1993Yeh et al., , 1996Yu & Koshland, 2001). Comparison of the two structures in the absence and presence of bound aspartate demonstrated a $1.0 Å vertical shift of the second transmembrane -helix relative to the first (Milburn et al., 1991;Chervitz & Falke, 1996;Ottemann et al., 1999). It has been proposed that in transmembrane signalling, Asp binding to the periplasmic domain of Tar induces a piston-like displacement of the second transmembrane -helix (reviewed in Falke & Hazelbauer, 2001). While the crystal structure of the apo form of the Tar periplasmic domain has been determined (Bowie et al., 1995;Chi et al., 1997), there has been no report of the crystal structure of the Aspbound form of Tar from Escherichia coli. Maruyama et al. (1995) previously proposed an alternative model in which the repellents stabilize the second transmembrane -helix in a different rotational orientation from those of the apo and Aspbound forms through a rotation/twist of the second transmembrane -helix parallel to the plane of the cytoplasmic membrane. The model also predicts that apo-Tar and Asp-bound Tar have similar structures. Structural analysis of the HAMP domains, which are located immediately downstream of the second transmembrane -helix, implies domain rotation in transmembrane signalling (Hulko et al., 2006). To test the rotation/twist model, the E. coli Tar periplasmic domain was crystallized with and without bound aspartate (Asp-Tar and apo-Tar, respectively). Here, the expression, purification, crystallization and preliminary X-ray diffraction studies of Asp-Tar and apo-Tar are described.

Cloning
A plasmid, pET-Tar, that expresses the periplasmic domain, residues 26-193 (TAR-E), of E. coli Tar was constructed using the expression vector pET-28a(+) (Novagen). This construct encodes an N-terminal methionine, a six-His tag and thrombin-recognition and enterokinase-recognition sites, followed by a Tar periplasmic sequence encompassing Gly26-Gln193. A DNA fragment encoding the periplasmic domain of Tar was amplified by PCR from E. coli DH5 genomic DNA using the forward primer 5 0 -ATG GCT AGC GAT GAC GAC GAC AAG GGC AGC CTG TTT TTT TCT TC-3 0 (NheI site in bold) and the reverse primer 5 0 -CTC GAA TTC TCA TTA TTG CCA CTG GGC AAA TC-3 0 (EcoRI site in bold). The resulting PCR product was digested with NheI and EcoRI, and was cloned into pET-28a(+) using a DNA ligation kit (Takara Bio, Tokyo, Japan). This construct was termed pET-Tar.

Expression
An overnight culture of E. coli BL21 (DE3) cells (Invitrogen) harbouring pET-Tar was made by inoculating a single colony in 10 ml LB medium containing 30 mg ml À1 kanamycin. This overnight culture was poured into 1000 ml LB medium containing 30 mg ml À1 kanamycin, which was then grown at 37 C with agitation at 250 rev min À1 to an OD 600 of 0.3. To this culture, isopropyl -d-1-thiogalactopyranoside (IPTG; Nacalai Tesque, Kyoto, Japan) was added to a final concentration of 0.1 mM to induce the production of TAR-E. Cultivation was continued for an additional 4 h and cells were harvested by centrifugation at 6000g for 5 min.

Purification
The harvested cell paste was resuspended in 10 ml suspension buffer (20 mM sodium phosphate buffer pH 7.4, 500 mM NaCl, 30 mM imidazole) with the addition of 100 ml protease-inhibitor cocktail (Nacalai Tesque) and was sonicated on ice for 40 min with an ultrasonic disruptor (Tomy Seiko, Tokyo, Japan). The resulting cell suspension was centrifuged at 6000g for 10 min at 4 C and resuspended in 10 ml resuspension buffer (8.0 M urea, 20 mM sodium phosphate pH 7.4, 500 mM NaCl) with the addition of a further 50 ml protease-inhibitor cocktail. A second sonication was carried out on ice for 1 min and the resulting suspension was pelleted by centrifugation at 20 000g for 15 min at 4 C. The resulting supernatant was cleared by filtration with a 0.45 mm HPLC filter (Pall Corp.).
From the cleared lysate, TAR-E was first purified by affinity chromatography with two tandemly connected 5 ml HisTrap FF columns (GE Healthcare). Before application of the sample, the columns were washed with three column volumes of washing buffer (8 M urea, 50 mM Tris-HCl pH 8.0, 200 mM NaCl). The cleared lysate was applied onto the HisTrap columns equilibrated with washing buffer at a flow rate of 1 ml min À1 for 40 min. Refolding of the bound TAR-E was performed by applying a linear urea gradient from 8.  refolded proteins was performed by applying a linear imidazole gradient ranging from 0 to 500 mM in buffer consisting of 50 mM Tris-HCl pH 8.0, 200 mM NaCl at a flow rate of 1 ml min À1 for 60 min. Eluted TAR-E was further purified by gel-filtration chromatography with a HiLoad column (26/600 Superdex 75 pg; GE Healthcare) equilibrated with 1.0 mM EDTA, 10 mM Tris-HCl pH 8.0, 100 mM NaCl by FPLC at a flow rate of 1.0 ml min À1 . Fractions containing TAR-E, $30 ml, were dialyzed against buffer consisting of 0.6 M NaCl, 10 mM HEPES-NaOH pH 7.5 and were concentrated by ultrafiltration using Vivaspin 6 (Vivaproducts, Massachusetts, USA). Protein purity was assessed by SDS-PAGE (14%), followed by staining with Coomassie Brilliant Blue. Concentrated TAR-E at 11.6-15.9 mg ml À1 was immediately used for crystallization without freezing. During cleavage with thrombin or enterokinase in order to remove the tag, His-tagged TAR-E precipitated for unknown reason(s). Therefore, TAR-E protein was directly used for crystallization without protease digestion.

Crystallization using ammonium sulfate as a precipitant
All crystallization experiments were performed in 24-well VDX plates with sealant (Hampton Research, California, USA) against a 1.0 ml reservoir. Previously reported conditions for crystallization (Bowie et al., 1995) were employed with slight modification. Crystals of apo-Tar1 were grown by mixing 1.5 ml purified TAR-E protein solution (0.6 M NaCl, 10 mM HEPES-NaOH pH 7.5, 15.9 mg ml À1 TAR-E) with 1.5 ml crystallization solution [0.4 M ammonium sulfate, 35 mM ammonium formate, 15 mM formic acid, 1.25%(v/v) glycerol pH 3.9] on a siliconized glass cover slide at 10 C. Hanging drops consisting of 1.5 ml each of the protein and crystallization solutions were equilibrated against a 1.0 ml reservoir of crystallization solution. A crystal of apo-Tar1 was sequentially soaked in crystallization solution supplemented with 10, 20 or 30%(v/v) glycerol and was flashcooled in liquid nitrogen (Fig. 1a).
Prior to crystallization of Asp-Tar1, 44 ml of an aspartate solution (0.6 M NaCl, 10 mM HEPES-NaOH pH 7.5, 45 mM aspartate) was mixed with 156 ml purified TAR-E protein solution (0.6 M NaCl, 10 mM HEPES-NaOH pH 7.5, 11.6 mg ml À1 TAR-E). Asp-Tar1 crystals were grown at 10 C by mixing 1.5 ml purified TAR-E solution (0.6 M NaCl, 10 mM HEPES-NaOH pH 7.5, 9.9 mM aspartate, 9.1 mg ml À1 TAR-E) with 1.5 ml crystallization solution (0.8 M ammonium sulfate, 20 mM ammonium formate, 30 mM formic acid pH 3.4) on a siliconized glass cover slide with streak-seeding with previously obtained Asp- Tar1   1.5 ml each of the protein and crystallization solutions were equilibrated against a 1.0 ml reservoir of crystallization solution. A crystal of Asp-Tar1 was sequentially soaked in crystallization solution supplemented with 20, 30 or 35%(v/v) glycerol and was flash-cooled in liquid nitrogen (Fig. 1c).

Crystallization using NaCl as a precipitant
For crystallization of apo-Tar2 and Asp-Tar2, NaCl was also used as a protein precipitant. Crystallization conditions with modifications have previously been reported (Chi et al., 1997). Crystals of apo-Tar2 were grown at 10 C by mixing 1.5 ml purified TAR-E solution (0.6 M NaCl, 10 mM HEPES-NaOH pH 7.5, 13.7 mg ml À1 TAR-E) with 1.5 ml crystallization solution (4.3 M NaCl, 100 mM Tris-HCl pH 8.0) on a siliconized glass cover slide with streak-seeding with crystals of apo-Tar1. Hanging drops consisting of 1.5 ml each of the protein and crystallization solutions were equilibrated against a 1.0 ml reservoir of crystallization solution. A crystal of apo-Tar2 was soaked in crystallization solution supplemented with 33%(v/v) glycerol and was flash-cooled in liquid nitrogen (Fig. 1b). Crystals of Asp-Tar2 were grown at 10 C by mixing 1.5 ml purified TAR-E solution (0.6 M NaCl, 10 mM HEPES-NaOH pH 7.5, 13.5 mg ml À1 TAR-E) with a 1.5 ml reservoir of crystallization solution (3.5 M NaCl, 100 mM Tris-HCl pH 8.0, 25 mM aspartate) on a siliconized glass cover slide with streak-seeding with crystals of apo-Tar1. Hanging drops consisting of 1.5 ml each of the protein and crystallization solutions were equilibrated against a 1.0 ml reservoir of crystallization solution. A crystal of Asp-Tar2 was sequentially soaked in a crystallization solution supplemented with 25 or 33%(v/v) glycerol and was flash-cooled in liquid nitrogen (Fig. 1d).
2.6. X-ray diffraction data collection X-ray diffraction data were collected under cryogenic conditions by flash-cooling with liquid nitrogen after stepwise transfers of all crystals into the crystallization solution supplemented with 10-35%(v/v) glycerol. Diffraction data for apo-Tar1 and Asp-Tar1 were collected using an ADSC Q270 detector on the AR-NE3A beamline and an ADSC Q315 detector on the BL5A beamline at the Photon Factory, Tsukuba, Japan, respectively. Diffraction data for apo-Tar2 and Asp-Tar2 were collected using a MAR 300HE charge-coupled device (CCD) detector on the BL44XU beamline and an ADSC Quantum315 CCD detector on the BL38B1 beamline at SPring-8, Harima, Japan, respectively. Diffraction images were processed with iMosflm and scaled using SCALA in the CCP4 program suite (Battye et al., 2011;Winn et al., 2011). The crystal data statistics are listed in Table 1.

Results and discussion
The TAR-E expression plasmid encodes the entire periplasmic domain of Tar from Gly26 to Gln193, which is the longest of the Tar periplasmic domains that have been crystallized to date (Bowie et al., 1995;Chi et al., 1997;Jancarik et al., 1991;Milburn et al., 1991;Scott et al., 1993;Yeh et al., 1993Yeh et al., , 1996Yu & Koshland, 2001). TAR-E was purified to give a single band on SDS-PAGE (Fig. 2). The molecular mass of denatured TAR-E was estimated to be 22 kDa, which is comparable to the calculated mass (22.3 kDa).
Asp-Tar1 and apo-Tar1 were crystallized in the presence and absence of Asp, respectively (Fig. 1). Based on previously reported methods (Bowie et al., 1995), the crystallization conditions for apo-Tar1 were optimized by varying the ammonium sulfate concentration in the range 0.4-1.1 M, by varying the buffer pH from pH 3.6 to 3.9 and by varying the incubation temperature from 10 to 25 C. Crystal nuclei appeared in 7 d in a droplet after crystallization commenced. The best crystal diffracted to 2.10 Å resolution and adopted space group P4 1 2 1 2 (Fig. 3a, Table 1). For crystallization of Asp-Tar1 (Fig.  1c), the conditions were optimized by varying the ammonium sulfate concentration in the range 0.4-0.8 M and the buffer pH from pH 3.4 to 3.9. Crystal nuclei appeared within 2 d in a droplet. The crystal of Asp-Tar1 diffracted to 2.40 Å resolution (Fig. 3c) and adopted space group P4 1 2 1 2 (Table 1). Bowie et al. (1995) reported that sulfate ions occupied aspartatebinding sites in the crystals of apo-Tar when ammonium sulfate was used as a precipitant. To overcome this problem, NaCl was used as a precipitant instead of ammonium sulfate (Figs. 1b and 1d). It took a few months to obtain crystals of apo-Tar2. On the other hand, many crystal nuclei of Asp-Tar2 appeared within 7 d in a droplet after the crystallization commenced. Crystals of apo-Tar2 and Asp-Tar2 diffracted to 1.95 and 1.58 Å resolution, respectively (Figs. 3b and 3d), and adopted space groups P2 1 2 1 2 1 and C2, respectively (Table 1). Structure determinations of all four types of crystals (apo-Tar1, apo-Tar2, Asp-Tar1 and Asp-Tar2) are currently in progress, and we have confirmed that aspartate molecules are bound in the crystals of Asp-Tar1 and Asp-Tar2 (data not shown). Furthermore, crystallization of the periplasmic domain with bound Ni 2+ is advancing.