d-Serine dehydratase purified from chicken kidney was crystallized by the hanging-drop vapour-diffusion method. The crystals diffracted to 2.09 Å resolution.
Keywords: d-serine dehydratase, pyridoxal 5′-phosphate, Schiff base, β-elimination
Abstract
d-Serine dehydratase purified from chicken kidney was crystallized by the hanging-drop vapour-diffusion method using PEG 4000 and 2-propanol as precipitants. The crystal belonged to space group P422, with unit-cell parameters a = 105.0, c = 81.89 Å, and diffracted to 2.09 Å resolution. An attempt to solve the structure using the MAD method is in progress.
1. Introduction
d-Serine is known to activate N-methyl-d-aspartate (NMDA) receptors as a co-agonist of l-glutamate (Labrie & Roder, 2010 ▶). Since the NMDA receptor is likely to be involved in higher brain functions such as learning and memory, the metabolism of d-serine has been studied in order to understand the molecular mechanism of higher brain functions. In mammals, d-serine is mainly degraded by d-amino-acid oxidase (DAO; Verrall et al., 2010 ▶; Pollegioni & Sacchi, 2010 ▶) and is partly dehydrated by a bifunctional serine racemase/dehydratase (Pollegioni & Sacchi, 2010 ▶). Although d-serine is widely believed to be degraded by DAO in vertebrate brains, our recent study showed that there was no detectable DAO activity in chicken brain. Rather, d-serine in chicken brain was degraded by d-serine dehydratase (DSD; Tanaka et al., 2007 ▶), which catalyzes the β-elimination of water from d-serine to form pyruvate and ammonia. We also found that chicken DSD (chDSD) was highly expressed in the kidney. In order to understand the molecular mechanism of d-serine degradation in chicken brain, biochemical and biological analysis of chDSD was initiated.
Examination of the chDSD gene sequence revealed that chDSD belongs to an enzyme family distinct from that of well known bacterial DSDs (Tanaka et al., 2008 ▶). chDSD shows weak amino-acid sequence similarity to alanine racemases from bacteria and exhibits 29% amino-acid sequence identity to DSD derived from Saccharomyces cerevisiae (scDSD; Tanaka et al., 2008 ▶; Ito et al., 2008 ▶). Biochemical analyses showed that chDSD is a pyridoxal 5′-phosphate (PLP)-dependent enzyme like bacterial alanine racemases and scDSD. chDSD is expected to be a member of the fold-type III enzymes such as alanine racemase (Grishin et al., 1995 ▶).
It is known that the PLP–d-serine Schiff base is primed for five distinct reactions: decarboxylation, racemization, transamination, Cα—Cβ cleavage (retro-aldol cleavage) and β-elimination (dehydration) (Toney, 2005 ▶). After forming the Schiff base, a carbanion is formed by heterolytic cleavage of any one of the bonds to Cα (except for the C—N bond). Racemization, transamination and β-elimination of serine follow cleavage of the Cα—H bond. Dunathan’s hypothesis proposed that the cleavage site is determined by the stereoelectronic configuration of the Schiff base in the enzyme active site (Dunathan, 1966 ▶). Biochemical analysis showed that chDSD only catalyzes the dehydration of d-Ser; although the primary structural similarity between chDSD and bacterial alanine racemases suggests that their active-site structures are also similar, chDSD does not show a detectable level of racemase activity (Tanaka et al., 2008 ▶). The molecular mechanism of the high reaction specificity of chDSD could therefore provide novel insights into the molecular evolution of the PLP-dependent enzymes, particularly the evolution of their active-site structures. However, this mechanism remains elusive owing to the lack of three-dimensional structural information for chDSD. To address this problem, we initiated a crystal structure analysis. Here, we report the crystallization and preliminary X-ray crystallographic data of chDSD.
2. Methods and results
2.1. Protein expression, purification and crystallization
ChDSD was purified from chicken kidneys as described previously (Tanaka et al., 2008 ▶). Briefly, chicken kidneys were homogenized in 10 mM potassium phosphate pH 7.2 containing 50 µM PLP. The homogenates were centrifuged at 17 700g for 20 min. The resultant supernatant was purified by four-step chromatography with DEAE-Sepharose (Amersham Pharmacia, Chicago, Illinois, USA), hydroxyapatite (Nacalai Tesque, Kyoto, Japan), Mono Q (Pharmacia) and TSKgel SuperSW3000 (Tosoh, Tokyo, Japan). The purified chDSD was concentrated to 14.7 mg ml−1 in 20 mM potassium phosphate pH 7.5, 1 mM dithiothreitol, 10 µM PLP and 10 mM (R)-(−)-2,3-diaminopropionic acid hydrochloride (d-2,3-DAP) as an inhibitor.
Initial screening for crystallization was performed by the hanging-drop vapour-diffusion method using Crystal Screen (Hampton Research, Laguna Niguel, California, USA; Jancarik & Kim, 1991 ▶) at 293 K. A hanging drop was prepared by mixing 1.0 µl each of the protein and reservoir solutions and was equilibrated against 500 µl reservoir solution. Clusters of needle-shaped crystals appeared in 24 h under several conditions: Crystal Screen condition Nos. 2, 17, 22, 40 and 41. All of the reservoir conditions except for No. 2 contained polyethylene glycol (PEG) 4000 as a precipitant. After 3 d, thin plate-shaped crystals of approximately 10 µm in length were obtained using Crystal Screen condition No. 41 [10%(v/v) 2-propanol, 0.1 M HEPES–NaOH pH 7.5 and 20%(w/v) PEG 4000]. In order to obtain larger plate-shaped crystals, the pH value (4.6–8.5), the PEG 4000 concentration [10–30%(w/v)] and the protein concentration (4–14 mg ml−1) were systematically optimized. After several trials, the optimal reservoir conditions were determined to be 12–15%(w/v) PEG 4000, 50 mM MES–NaOH pH 6.5 and 10%(v/v) 2-propanol using 4–8 mg ml−1 protein solution. Larger plate-shaped crystals (approximately 0.2 × 0.15 × 0.05 mm) were obtained in 2–3 weeks (Fig. 1 ▶). Although chDSD is a PLP-dependent enzyme and thus is typically yellow in colour, the crystals obtained were nearly colourless, suggesting that PLP had dissociated from chDSD during crystallization. In order to incorporate PLP into chDSD in the crystal, the obtained crystal was soaked in PLP solution [5 mM PLP, 30%(w/v) PEG 4000, 10%(v/v) 2-propanol, 100 mM MES–NaOH pH 6.5], resulting in yellow crystals.
Figure 1.
Crystals of chDSD purified from chicken kidney.
2.2. X-ray diffraction analysis
Diffraction data were collected at 95 K using an ADSC CCD detector on beamline BL-5A of Photon Factory (PF; Tsukuba, Japan; Fig. 2 ▶). Prior to data collection, the crystals were soaked in a cryoprotectant solution consisting of 30%(v/v) glycerol, 20%(w/v) PEG 4000, 100 mM MES–NaOH pH 6.5 and 10%(v/v) 2-propanol containing 5 mM PLP for 20 s and flash-cooled in an N2 stream. The exposure time per frame was 8 s.
Figure 2.
X-ray diffraction pattern of a chDSD crystal. The diffraction data were collected on beamline BL-5A of PF using an ADSC Quantum 315 CCD detector.
The diffraction data were processed and scaled at 2.09 Å resolution using the programs XDS and XSCALE (Kabsch, 2010 ▶). The crystals belonged to space group P422 (Table 1 ▶). Assuming one subunit of chDSD (40.4 kDa) per asymmetric unit of the crystal, the Matthews coefficient (V M) was calculated to be 2.79 Å3 Da−1, corresponding to a solvent content of 56.0% (Matthews, 1968 ▶). This result suggests that the subunits of the dimeric chDSD molecule are related by a crystallographic twofold axis.
Table 1. Data-collection statistics.
Values in parentheses are for the outermost resolution shell.
| Beamline | BL-5A (PF) |
| Oscillation angle (°) | 1 |
| Crystal-to-detector distance (mm) | 284.3 |
| Exposure time (s) | 8 |
| Wavelength (Å) | 1.0000 |
| Temperature (K) | 95 |
| Space group | P422 |
| Unit-cell parameters (Å) | a = 105.0, c = 81.89, |
| Resolution (Å) | 17.0–2.09 (2.21–2.09) |
| Observations | 387030 (59770) |
| Unique reflections | 27631 (4172) |
| Completeness (%) | 99.6 (100.0) |
| Multiplicity | 14.0 (14.3) |
| Average I/σ(I) | 33.1 (5.1) |
| Rmerge | 0.057 (0.597) |
| Mosaicity (°) | 0.315 |
To prepare heavy-atom derivatives, crystals were soaked in artificial mother liquor [30%(w/v) PEG 4000, 100 mM MES–NaOH pH 6.5 and 10%(v/v) 2-propanol] containing 1 mM zinc chloride for 25 min. The crystal diffracted to approximately 2.15 Å resolution. MAD data were collected at four different wavelengths and four Zn atoms were located in the asymmetric unit of the cell using the MAD data. MAD phasing and structure determination are in progress.
Acknowledgments
This study was supported in part by the New Energy and Industrial Technology Development Organization (NEDO) of Japan, a Grant-in-Aid for Cooperation of Innovative Technology and Advanced Research in City Area Program in the Southern Area of Lake Biwa from the Ministry of Education, Culture, Sports, Sciences and Technology of Japan and a Grant-in-Aid (Heisei era 19, 20 and 21) from the Shiga University of Medical Science.
References
- Dunathan, H. C. (1966). Proc. Natl Acad. Sci. USA, 55, 712–716. [DOI] [PMC free article] [PubMed]
- Grishin, N. V., Phillips, M. A. & Goldsmith, E. J. (1995). Protein Sci. 4, 1291–1304. [DOI] [PMC free article] [PubMed]
- Ito, T., Hemmi, H., Kataoka, K., Mukai, Y. & Yoshimura, T. (2008). Biochem. J. 409, 399–406. [DOI] [PubMed]
- Jancarik, J. & Kim, S.-H. (1991). J. Appl. Cryst. 24, 409–411.
- Kabsch, W. (2010). Acta Cryst. D66, 125–132. [DOI] [PMC free article] [PubMed]
- Labrie, V. & Roder, J. C. (2010). Neurosci. Biobehav. Rev. 34, 351–372. [DOI] [PubMed]
- Matthews, B. W. (1968). J. Mol. Biol. 33, 491–497. [DOI] [PubMed]
- Pollegioni, L. & Sacchi, S. (2010). Cell. Mol. Life Sci. 67, 2387–2404. [DOI] [PMC free article] [PubMed]
- Tanaka, H., Yamamoto, A., Ishida, T. & Horiike, K. (2007). Anal. Biochem. 362, 83–88. [DOI] [PubMed]
- Tanaka, H., Yamamoto, A., Ishida, T. & Horiike, K. (2008). J. Biochem. 143, 49–57. [DOI] [PubMed]
- Toney, M. D. (2005). Arch. Biochem. Biophys. 433, 279–287. [DOI] [PubMed]
- Verrall, L., Burnet, P. W. J., Betts, J. F. & Harrison, P. J. (2010). Mol. Psychiatry, 15, 122–137. [DOI] [PMC free article] [PubMed]


