1228doi:10.1107/S2053230X14014629 Acta Cryst. (2014). F70, 1228–1231
Acta Crystallographica Section F
Structural Biology Communications
Purification, crystallization and preliminary X-ray
crystallographic analysis of glycosyltransferase-1
Yin-Cheng Hsieh,a‡ Hsi-Ho Chiu,b‡ Yen-Chieh Huang,a Hoong-Kun Fun,c,dChia-Yu Lu,b Yaw-Kuen Lib* and Chun-Jung Chena,c,e,f*
aLife Science Group, Scientific Research Division, National Synchrotron Radiation Research Center, Hsinchu 30076, Taiwan, b
Department of Applied Chemistry, National Chiao Tung University, Hsinchu 30010, Taiwan,cDepartment of Pharmaceutical Chemistry, College of Pharmacy, King Saud University, Riyadh 11451, Saudi Arabia,dX-ray Crystallography Unit, School of Physics, Universiti Sains Malaysia, 11800 USM, Malaysia,eDepartment of Physics, National Tsing Hua University, Hsinchu 30013, Taiwan, andfInstitute of Biotechnology and the Center for Bioscience and Biotechnology, National Cheng Kung University, Tainan City 701, Taiwan
‡ These authors contributed equally to this work.
Received 11 April 2014 Accepted 20 June 2014
Glycosyltransferases (GTs), which are distributed widely in various organisms, including bacteria, fungi, plants and animals, play a role in synthesizing biological compounds. Glycosyltransferase-1 from Bacillus cereus (BcGT-1), which is capable of transferring glucose to small molecules such as kaempferol and quercetin, has been identified as a member of the family 1 glycosyl-transferases which utilize uridine diphosphate glucose (UDP-glucose) as the sugar donor. BcGT-1 (molecular mass 45.5 kDa) has been overexpressed, purified and crystallized using the hanging-drop vapour-diffusion method. According to X-ray diffraction of BcGT-1 crystals to 2.10 A˚ resolution, the crystal belonged to space group P1, with unit-cell parameters a = 54.56, b = 84.81, c = 100.12 A˚ , = 78.36, = 84.66, = 84.84. Preliminary analysis
indicates the presence of four BcGT-1 molecules in the asymmetric unit with a solvent content of 50.27%.
Glycosyltransferases (GTs) have been found in a broad range of organisms such as microbes, plants and animals (Campbell et al., 1997; Paquette et al., 2003; Bowles et al., 2006). These enzymes, which contain two binding sites to carry out their functions by transferring the sugar from a donor to an acceptor, are involved in several important antibiotic biosynthesis pathways, such as those of chloro-eremomycin, vancomycin and balhimycin (Mulichak et al., 2001; Kahne et al., 2005; Pelzer et al., 1999). In contrast to the highly diverse primary structures of GTs, which are presently classified into 94 families (CAZY database), their tertiary structural folds are divided into two major groups: GT-A and GT-B (Breton et al., 2006; Lairson et al., 2008). The GT-A fold contains one major globular structure that associates with the closely abutting binding sites of the sugar donor and acceptor, whereas the GT-B fold comprises two separated globular structures that face each other to bind the sugar donor and acceptor. Although GT-A and GT-B exhibit distinct overall structural folds, they share a general nucleotide-binding domain: the // Rossmann fold. The most common donors for GTs are nucleoside-diphosphate sugars such as UDP-galactose and GDP-mannose, whereas their diverse acceptors include proteins, lipids, steroids and small molecules (Hashimoto et al., 2009). In addition, the sugar may be conjugated onto the carbon, nitrogen, sulfur or oxygen of their acceptors. Although many proteins have been classified into the GT family, the mechanism of glycoside-bond formation in the glycosyl-transferases remains unclear.
Based on sequence alignment, glycosyltransferase-1 (BcGT-1) from Bacillus cereus is classified into the uridine diphosphate (UDP)-containing family 1 glycosyltransferases (UGT), which utilize UDP as a cofactor to transfer sugars to acceptors. Proteins in the UGT family commonly utilize their C-terminal regions to bind to sugar donors such as UDP-glucose, UDP-galactose and UDP-rhammose, whereas the N-terminal region binds to sugar acceptors such as alkaloids,
#2014 International Union of Crystallography All rights reserved
flavonoids, antibiotics and plant hormones. Four Bcgt genes have previously been identified by genome analysis, among which Bcgt-1 has been cloned and studied (Ko et al., 2006). BcGT-1 employs glucose as a major sugar to conjugate onto the O-linking sites of products. Several flavonoids have been identified to activate BcGT-1, among which kaempferol and quercetin generate the greatest activity of the enzyme. BcGT-1 is capable of transferring three sugar mole-cules to three distinct hydroxyl groups of kaempferol (Ko et al., 2006). Mutations of some residues on the protein surface have been proposed to convert the product triglucoside kaempferol to O-monoglucoside kaempferol. The product transformation was also related to alteration of the secondary structure of the protein (Jung et al., 2010).
The products of GTs have been reported in therapies for several diseases. For example, flavonoids, which scavenge free radicals, have served in the treatment of congestive heart failure and the improvement of blood circulation (Middleton et al., 2000; Perez-Vizcaino & Duarte, 2010). Etoposide, which inhibits DNA topo-isomerase, is a participant in anticancer therapy (Pommier et al., 2012). Anthracyclines, which inhibit DNA synthesis by intercalation, are common drugs for the treatment of many cancers such as leukaemias, lymphomas and breast, uterine, lung and ovarian cancers (Minotti et al., 2004). GT is hence an important enzyme in drug development.
To elucidate the structure and function of BcGT-1, we have obtained pure protein in a large quantity for crystallization. Here, we report the overexpression, purification, crystallization and preli-minary characterization by X-ray diffraction of glycosyltransferase-1 from B. cereus.
2. Materials and methods
2.1. Protein expression and purification
The synthetic gene of BcGT-1 (GenBank accession No.
AAS41089.1) was inserted into the expression vector pRSET A (Invitrogen) between NdeI and HindIII restriction sites and was expressed in Escherichia coli ECOS21 (Yeastern, Taiwan). The
bacteria were cultured in Luria–Bertani (LB) medium with ampicillin (1 mM) at 289 K and induced by 1 mM isopropyl -d-1-thiogalacto-pyranoside (IPTG). After induction for 16–18 h, the cells were
harvested by centrifugation (8000 rev min1); the cell pellet was
washed with water and resuspended in 25 mM sodium phosphate
buffer pH 7.0 containing 5 mM dithiothreitol (DTT) and 1 mM
phenylmethanesulfonyl fluoride (PMSF). After disrupting the cells with a French press, the cell debris was removed by centrifugation
(18 000 rev min1) at 277 K for 30 min. The supernatant was loaded
onto an ion-exchange DEAE column (Toyopearl 650M) which had been pre-equilibrated with 25 mM sodium phosphate buffer pH 7.0 containing 5 mM DTT. After elution with 100 mM NaCl in pre-equilibration buffer, the fractions that showed positive results with the Bradford assay were collected. These fractions were further purified with a size-exclusion column (Superdex 200, GE Healthcare) and eluted with 25 mM Tris buffer pH 7.0 containing 5 mM DTT. A 12.5% SDS–PAGE analysis showed a single band near 45.5 kDa, which was further identified to be BcGT-1 by in-gel trypsin digestion and peptide mapping with MALDI–TOF–MS (Bruker Autoflex III). For the purposes of crystallization, the concentration of purified
BcGT-1 was further enhanced to 12 mg ml1 by ultrafiltration
(Vivaspin, 10 000 MWCO, GE Healthcare). 2.2. Crystallization
Crystallization was performed by the hanging-drop vapour-diffu-sion method at 291 K using 48-well VDX plates. Small crystals were observed from a condition consisting of 25%(w/v) PEG 3350, 200 mM sodium chloride, 100 mM Tris buffer pH 8.5 within 3 d of initial screening using the Crystal Screen kit (Hampton Research). This condition was further optimized to produce larger BcGT-1 crystals using hanging drops (2 ml) consisting of 1 ml protein solution and 1 ml reservoir solution equilibrated against 150 ml reservoir solution consisting of 28%(w/v) PEG 3350, 200 mM sodium chloride, 100 mM Tris buffer pH 8.5. Crystals of diffraction quality were used to collect X-ray diffraction data.
2.3. X-ray data collection and processing
The protein crystals were initially screened and characterized with synchrotron radiation on the SPXF beamline BL13B1 equipped with a CCD detector (Q315r, ADSC) at the National Synchrotron Radiation Research Center (NSRRC), Taiwan. Data collection was
Acta Cryst. (2014). F70, 1228–1231 Hsieh et al. Glycosyltransferase-1
Single crystals of BcGT-1 grown by the hanging-drop method.
Analysis of molecular mass of BcGT-1 by 12.5% SDS–PAGE. Lane M, protein marker (labelled in kDa). Lane 1, cellular proteins from whole cells before IPTG induction. Lane 2, cellular proteins from whole cells after IPTG induction. Lane 3, supernatant after cell disruption and centrifugation. Lane 4, purified enzyme from the ion-exchange (DEAE) column. Lane 5, purified enzyme from the size-exclusion (Superdex 200) column.
completed on the Taiwan-contracted beamline BL12B2 for protein crystallography equipped with a CCD detector (Q210r, ADSC) at SPring-8, Japan. The crystal was transferred from a crystallization drop into 10 ml cryoprotectant solution consisting of 28%(w/v) PEG 3350, 200 mM sodium chloride, 20%(v/v) glycerol, 0.1 M Tris buffer pH 8.5 for a few seconds, mounted on a nylon loop (0.3–0.4 mm; Hampton Research) and subsequently flash-cooled in liquid nitrogen
at 100 K. For complete data collection, 360 rotations with 1.0
oscillation were measured with an X-ray wavelength of 1.00 A˚ , an
exposure duration of 15 s and a crystal-to-detector distance of 180 mm at 110 K in a nitrogen stream using a cryo-system (X-Stream, Rigaku/MSC Inc.). All data were indexed, integrated and scaled using HKL-2000 (Otwinowski & Minor, 1997).
3. Results and discussion
The BcGT-1 gene was initially inserted into the vector pRSET A with XhoI and KpnI cutting sites, which contained a 6His tag before the N-terminus of BcGT-1. However, this overexpression did not produce protein in a sufficient quantity for crystallization trials. Fortunately, excluding the 6His tag by inserting the gene into the NdeI and HindIII sites produced BcGT-1 in large quantities with an expression yield of 10 mg per 250 ml. Purification with both DEAE and Superdex-200 columns produced highly pure protein of mole-cular mass 45.5 kDa according to SDS–PAGE analysis (Fig. 1). The purified BcGT-1 was further confirmed by peptide mapping with MALDI–TOF–MS.
An initial crystal screening trial was performed with BcGT-1 in phosphate buffer, which easily generated flat and rectangular salt crystals in the crystallization drops. After buffer exchange from phosphate buffer to 25 mM Tris buffer pH 7.0, crystals of BcGT-1 could be obtained in several independent conditions after initial screening, but only the conditions reported here produced X-ray
diffraction data of high quality with a resolution of up to 2.10 A˚ .
Although the protein purity was fairly good (>95%) after the first DEAE column, crystals were only obtained after another size-exclusion column (Superdex 200) was applied to improve the purity in the final purification step. Protein crystals with a needle shape appeared after 3 d and continued to grow to a final size of 0.06 0.06 0.8 mm within one week in an incubator at 291 K (Fig. 2). A long needle crystal was cut and shortened to a smaller size of about 0.06 0.06 0.3 mm before being mounted on a loop for data collection. The protein crystals were sensitive to variation in precipitant concentration while being transferred to the cryoprotectant solution containing glycerol (20%).
Selected crystals with mosaicity <1 that diffracted to higher
resolution (>2.5 A˚ ) were collected. Radiation damage that caused a
decrease in I/(I) and an increase in Rmerge was observed on
protracted exposure during data collection. Analysis of the diffrac-tion pattern indicates that the crystals exhibit a triclinic symmetry and belong to space group P1 (Fig. 3). Assuming the presence of four
1230Hsieh et al. Glycosyltransferase-1 Acta Cryst. (2014). F70, 1228–1231
The self-rotation function with = 180 shows noncrystallographic twofold
X-ray diffraction image of the BcGT-1 crystal showing diffraction to 2.10 A˚
X-ray diffraction statistics of BcGT-1 crystals.
Values in parentheses are for the highest resolution shell.
Wavelength (A˚ ) 1.00 Temperature (K) 110 Resolution range (A˚ ) 30.0–2.10 (2.18–2.10) Space group P1 Unit-cell parameters (A˚ ,) a = 54.56, b = 84.81, c = 100.11, = 78.36, = 84.66, = 84.84 Unique reflections 96752 Completeness (%) 97.3 (96.8) hI/(I)i 11.2 (1.9) Average multiplicity 2.1 (2.1) Rmerge† (%) 5.6 Mosaicity () 0.71
No. of molecules per asymmetric unit 4 Matthews coefficient (A˚3Da1) 2.47 Solvent content (%) 50.27 † Rmerge= P hkl P ijIiðhklÞ hIðhklÞij= P hkl P
iIiðhklÞ, where Ii(hkl) is the ith measure-ment and hI(hkl)i is the weighted mean of all measuremeasure-ments of I(hkl).
molecules of BcGT-1 per asymmetric unit, the Matthews coefficient is
estimated to be 2.47 A˚3Da1 with a solvent content of 50.27%
(Matthews, 1968), which is within the normal range for protein crystals. The self-rotation function calculated using the Self Rotation Function Server (http://www.doe-mbi.ucla.edu/~sawaya/selfrot/) with
data in the resolution range 15–4 A˚ shows noncrystallographic
twofold symmetry (Fig. 4), suggesting that two sets of homologous dimers exist with the twofold axes parallel to each other in an asymmetric unit. Details of the data statistics of X-ray diffraction are summarized in Table 1.
An initial attempt to solve the crystal structure of BcGT-1 with molecular replacement was performed using the monomer structures of glycosyltransferase from Micromonospora echinospora (sequence identity 29%; PDB entry 3rsc; Chang et al., 2011) and Streptomyces antibioticus (sequence identity 27%; PDB entry 2iyf; Bolam et al., 2007), but all the calculations failed to yield an obvious solution. From the low sequence identities and several gaps based on sequence alignments, the structure of BcGT-1 is expected to differ markedly from the available structures. A search for useful heavy-atom deri-vatives and preparation of selenomethionine derideri-vatives for SAD or MAD phasing for structure determination is in progress.
We are grateful to our colleagues, Dr Yuch-Cheng Jean and supporting staff for technical assistance with and discussion of the use of the synchrotron-radiation X-ray facility for data collection at BL13B1 of the NSRRC, Taiwan, Masato Yoshimura for technical assistance at BL12B2 and Osaka beamline BL44XU of SPring-8, Japan under proposal Nos. 201A4009 and 2012A6760, and Ms Ko-Jen Su and Dr Yu-Kuo Wang at NCTU for help in protein preparation and MALDI–TOF–MS peptide-mapping analysis. This work was supported in part by National Science Council (NSC) grants
101-2628-B-213-001-MY4, 102-2627-M-213-001-MY3 and National
Synchrotron Radiation Center (NSRRC) grants 1013RSB02 and 1023RSB02 to C-JC and NSC 101-2113-M-009-012 -MY3 to Y-KL.
The authors extend their appreciation to the Deanship of Scientific Research at King Saud University for funding the work through research group project No. RGP-VPP-207.
Bolam, D. N., Roberts, S. M., Proctor, M. R., Turkenburg, J. P., Dodson, E. J., Martinez-Fleites, C., Yang, M., Davis, B. G., Davies, G. J. & Gilbert, H. J. (2007). Proc. Natl Acad. Sci. USA, 104, 5336–5341.
Bowles, D., Lim, E. K., Poppenberger, B. & Vaistij, F. E. (2006). Annu. Rev. Plant Biol. 57, 567–597.
Breton, C., Sˇnajdrova´, L., Jeanneau, C., Kocˇa, J. & Imberty, A. (2006). Glycobiology, 16, 29R–37R.
Campbell, J. A., Davies, G. J., Bulone, V. & Henrissat, B. (1997). Biochem. J. 15, 929–939.
Chang, A., Singh, S., Helmich, K. E., Goff, R. D., Bingman, C. A., Thorson, J. S. & Phillips, G. N. (2011). Proc. Natl Acad. Sci. USA, 108, 17649–17654. Hashimoto, K., Tokimatsu, T., Kawano, S., Yoshizawa, A. C., Okuda, S., Goto,
S. & Kanehisa, M. (2009). Carbohydr. Res. 344, 881–887.
Jung, N. R., Joe, E. J., Kim, B.-G., Ahn, B. C., Park, J. C., Chong, Y. & Ahn, J.-H. (2010). J. Microbiol. Biotechnol. 20, 1393–1396.
Kahne, D., Leimkuhler, C., Lu, W. & Walsh, C. (2005). Chem. Rev. 105, 425– 448.
Ko, J. H., Kim, B. G. & Ahn, J.-H. (2006). FEMS Microbiol. Lett. 258, 263–268. Lairson, L. L., Henrissat, B., Davies, G. J. & Withers, S. G. (2008). Annu. Rev.
Biochem. 77, 521–555.
Matthews, B. W. (1968). J. Mol. Biol. 33, 491–497.
Middleton, E. Jr, Kandaswami, C. & Theoharides, T. C. (2000). Pharmacol. Rev. 52, 673–751.
Minotti, G., Menna, P., Salvatorelli, E., Cairo, G. & Gianni, L. (2004). Pharmacol. Rev. 56, 185–229.
Mulichak, A. M., Losey, H. C., Walsh, C. T. & Garavito, R. M. (2001). Structure, 9, 547–557.
Otwinowski, Z. & Minor, W. (1997). Methods Enzymol. 276, 307–326. Paquette, S., Møller, B. L. & Bak, S. (2003). Phytochemistry, 62, 399–413. Pelzer, S., Su¨bmuth, R., Heckmann, D., Recktenwald, Z., Walsh, C. T. &
Garavito, R. M. (1999). Antimicrob. Agents Chemother. 43, 1565–1573. Perez-Vizcaino, F. & Duarte, J. (2010). Mol. Aspects Med. 31, 478–494. Pommier, Y., Leo, E., Zhang, H. & Marchand, C. (2012). Chem. Biol. 17, 421–