Crystal structure of Escherichia coli L - asparaginase , an enzyme used in cancer therapy ( amidohydrolase / leukemia / active site / aspartate ) AMY

The crystal structure of Escherchia coli asparaginase II (EC 3.5.1.1), a drug (Elspar) used for the treatment of acute lymphoblastic leukemia, has been determined at 2.3 A resolution by using data from a single heavy atom derivative in combination with molecular replacement. The atomic model was refined to an R factor of 0.143. This enzyme, active as a homotetramer with 222 symmetry, belongs to the class of a/P proteins. Each subunit has two domains with unique topological features. On the basis of present structural evidence consistent with previous biochemical studies, we propose locations for the active sites between the Nand C-terminal domains belonging to different subunits and postulate a catalytic role for Thr-89. Two isozymes of L-asparaginase (EC 3.5.1.1), which catalyze the hydrolysis ofL-asparagine to L-aspartate and ammonia, are produced by Escherichia coli (1). Type I E. coli asparaginase (EcAI) is located in the cytosol, whereas type II (EcA) is located in the periplasmic region of the bacteria and has a much higher affinity for asparagine (Km = 11.5 ;LM) than EcAI (2). Production of EcA is dramatically increased (>100-fold) under anaerobic conditions when amino acids are the primary source of carbon for the bacteria. In this case, asparaginase is likely to participate in the oxaloacetate pathway (3). EcA is active as a tetramer of identical subunits (4), each of 35.6 kDa (5). This enzyme has antileukemic activity and is used clinically for the treatment of acute lymphoblastic leukemia (1, 6). Much effort has been made to characterize EcA, but has yielded few details about its mode of neoplastic inhibition or its enzymatic mechanism (7). The enzyme is effective against neoplasias that require asparagine and obtain it from circulating pools (8), presumably because the cancer cells have diminished expression of asparagine synthetase (9). It is uncertain, however, whether neoplastic cell death after asparaginase administration results directly from the depletion of circulating asparagine levels or, indirectly, from some other metabolite of the asparaginase reaction (10). Nevertheless, this enzyme has been in clinical use for many years, affecting complete disease remission in some cases, as well as proving effective in maintenance therapy. Biochemical studies have given little insight into the mechanism of the asparaginase reaction in vitro. Substrate-analog modification studies of different but homologous amidohydrolases have given seemingly conflicting results in attempts to identify residues important for activity (12, 13). In one case, the substrate analog was bound to a residue conserved among all bacterial asparaginases, Thr-12, of Acinetobacter glutaminasificans glutaminase-asparaginase (AGA) (12) and in another case to Ser-120 of EcA (13), a residue that is not conserved. Chemical modification of the enzyme and 1H NMR spectroscopy indicated that a histidine residue and one or two tyrosine residues may be necessary for activity (14), but subsequent mutagenesis studies indicate that none of the histidine residues in EcA are required for catalysis (15). The most convincing evidence for the location of a residue important for activity comes from a recent report that shows loss of asparaginase activity when Thr-12 is mutated to alanine, but preservation of activity when it is mutated to serine (16). Crystallographic investigations ofasparaginase and related amidohydrolases have been going on for over 20 years (17-22), resulting in a partially correct model ofAGA (22) and a very low-resolution map of EcA in which the only recognized structural elements were two long helices (18). In view of the importance of this macromolecule, both as an enzyme and as a drug, and the need to clarify the conflicting biochemical reports, we pursued the study ofits structure at high resolution. We report here the x-ray crystallographic structure§ of EcA, determined by using single isomorphous replacement and anomalous scattering data, combined with information from molecular replacement (MR) by using the preliminary model of AGA (22) as a probe. MATERIALS AND METHODS Crystaliation. EcA was obtained in the form of Elspar concentrate (product 38663, Rx 95885; Merck, Sharp and Dohme). Previously unreported monoclinic crystals were grown in hanging drops from 0.1 M sodium acetate (pH 5) equilibrated against 0.1 M sodium acetate, pH 5/2-methyl2,4-pentanediol/25% polyethylene glycol 3350, 50:33:17 (vol/vol). The space group is P21 with unit cell parameters a = 76.7 A, b = 96.1 A, c = 111.3 A, andfA= 97.10, and each asymmetric unit contains a complete tetramer of EcA. X-Ray Data Collection and Procesing. X-ray intensity data were collected using a Siemens area detector mounted on a three-axis goniometer. The native data are 90% complete to 2.5 A and 74% complete at 2.3 A, with a redundancy of 5.4. The Rmerge for data collected from five different crystals is 0.055, where Rmerge = LIIi -(IlI/X(I). An isomorphous Abbreviations: EcAI, type I E. coli asparaginase; EcA, type II E. coli asparaginase; AGA, Acinetobacter glutaminasificans glutaminaseasparaginase; ErA, Erwinia chrysanthemi asparaginase; MR, molecular replacement. *Present address and on leave from: A. Mickiewicz University, Faculty of Chemistry, PoznadS, Poland. tPresent address: Institut de Biologie Structurale, Laboratoire de Cristallographie et Cristallogdntse des Protdines, Batiment 4, 41, avenue des Martyrs, 38027 Grenoble cedex 1, France. tTo whom reprint requests should be addressed. §The atomic coordinates and structure factors have been deposited in the Protein Data Bank, Chemistry Department, Brookhaven National Laboratory, Upton, NY 11973 (reference 2ECA, 2ECASF). 1474 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. Proc. Natl. Acad. Sci. USA 90 (1993) 1475 derivative was obtained by soaking a native crystal for 3.5 days in a solution of 10 mM uranyl nitrate in the buffer described above. Derivative data were measured from one crystal, resulting in a set with 24,407 unique reflections and 13,399 Friedel pairs {Rmerge = 0.061 with a redundancy of 1.7; the anomalous R factor was Ra = 0.085, where Ra = [Y1jF(hkl)I IF(hi1)II]/[0.5X(lF(hkl)I + IF(hiU1)]}. The Rdiff between native (N) and derivative (D) data in the resolution range of 35-3.5 A is 0.177, where Rdiff = XIFN FDI/IFN. All calculations were made with I 2 1.5o-(I) data. Other statistics for the heavy atom derivative are given in Table 1. The only previously published amidohydrolase structure was that of AGA reported by this laboratory at 2.9 A resolution (22). That model consisted of two polypeptide chains with a total of 331 residues per subunit, and its known shortcomings were identified. In particular, although most of the secondary structure elements were identified correctly, the direction of the chains, as well as the connections between the elements, remained in doubt. This imperfect model was, however, sufficient to locate the EcA tetramer in its unit cell. Several MR techniques were employed for this purpose, all leading to the same solution, and the one used for subsequent analysis was that obtained from X-PLOR (23). A peak of 5o in the resolution range 8-4 A with a 0.20 grid was found at O1 = 191.80, 82 = 29.20, and 83 = 75.60. The translation function showed the position ofthe tetramer to be x = 0.25, z = 0.25 or x = 0.75, z = 0.25 (y was arbitrarily set to 0.0 in the space group P21). A difference-Patterson map for the uranyl derivative of EcA revealed numerous Harker peaks that could not be fully deconvoluted. However, an (FD FN) difference Fourier map, calculated by using phases from the MR solution based on the AGA model discussed above, had the first 11 peaks substantially above background. A search for differencePatterson cross vectors between those 11 sites showed that 9 of them represented a self-consistent set, and only these sites were used for subsequent phasing to 3.5 A. A lack-ofclosure refinement carried out in PROTEIN (24) resulted in the Table 1. Uranyl derivative statistics Step 1: Heavy atom refinement