THE JOURNAL OF BIOLOGICAL CHEMWTRY Vol. 254, No. 11, Issue of June 10, pp. 4915-4918, 1979 Prrnted in U.S.A. Crystallographic Studies of Glutamate Dehydrogenase PRELIMINARY CRYSTAL DATA* (Received for publication, August 7, 1978, and in revised form, November 10, 1978) Jens J. Birktoft, F’umio Miake, Leonard J. Banaszak, and Carl Frieden From the Department of Biological Chemistry, Division of Biology and Biomedical Sciences, Washington University School of Medicine,-St. Louis, Missou& 63110 Native and pyridoxal phosphate modified rat liver glutamate dehydrogenase crystals have been obtained and used for a preliminary x-ray crystallographic anal- ysis. The space group is P6222 (P6422) having unit cell dimensions a = b=101Ij,c=724BLandy=12Q~.The unit cell contains 36 subunits (six hexameric molecules) of molecular weight 56,000 and there is one half- molecule, ie. three subunits, in the asymmetric unit. Packing considerations suggest that the glutamate de- hydrogenase molecule has the point group symmetry 32 and that each subunit can be represented as a par- ticle with approximate dimensions of 45 x 45 X 60 A. Glutamate dehydrogenase catalyzes the reductive amina- tion of a-ketoglutarate to glutamate. The kinetic and molec- ular properties of the enzyme from a wide variety of sources have been extensively studied (l-5). The enzymes isolated from animal sources utilize either NADH or NADPH nearly equally well and are subject to control by a wide variety of metabolites, particularly purine nucleotides (1). It has been suggested that these kinetic characteristics are relevant to the role that the enzyme serves in controlling carbohydrate and protein metabolism (1, 6). Among the animal enzymes, the bovine liver enzyme has been the most extensively studied. This enzyme undergoes a reversible self-association to high molecular weight forms and it has been suggested that this polymerization could be one reason for the failure to grow crystals suitable for x-ray crystallographic analysis (7). Rat liver glutamate dehydrogenase, although having kinetic prop- erties similar to those of the bovine liver enzyme, does not polymerize to any great extent (8). Both the rat and bovine liver enzymes consist of six subunits of molecular weight 56,000 and comparison of the complete amino acid sequence of the bovine enzyme (3) with the available amino acid se- quence data of the rat liver enzyme (Coffee, C. J. and Frieden, C, as quoted in Ref. 3) suggests that the two enzymes are about 95% homologous. Glutamate dehydrogenases in general have a high propen- sity towards polymerization. Our efforts, therefore, were di- rected towards an enzyme form for which this tendency is negligible (7). Initial experiments used the pyridoxal phos- phate derivative of rat liver glutamate dehydrogenase. This modification prevents polymerization of the bovine liver en- zyme (8) and it was felt that this might eliminate the slight tendency towards polymerization displayed by the rat liver * These studies have been supported by Grants AM 13332 and GM 13925 from the National Institutes of Health and by Grant P-CM 76- 81481 from the National Science Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. enzyme (9, 10). Crystals of both native and modified enzyme have been prepared. While the data presented here have been obtained with the pyridoxal phosphate derivate, preliminary results for the native enzyme show it to be isomorphous or nearly isomorphous with the modified form. Rat liver glutamate dehydrogenase was prepared by the method of King and Frieden (8). Pyridoxal phosphate modi- fication was carried out at 25°C in 0.05 M phosphate buffer, pH 7.4, at 1 mg/ml of enzyme and 0.4 mu pyridoxal phosphate as described elsewhere (11). After 75 min, 2 mu NaBH4 was added to attach irreversibly the pyridoxal phosphate to the enzyme and the solution made 60% with respect to saturated (NH&SO+ After this treatment, the enzyme contained 1.2 groups of pyridoxal phosphate/subunit and had about 10% of the activity of the native enzyme. Modification of the bovine liver enzyme under similar conditions leads to depolymeriza- tion and loss of most of the catalytic activity (11,12). It should be noted that, for the bovine liver enzyme, Talbot et al. (13) have concluded that these two properties are a consequence of modification of two different lysines/subunit rather than one as suggested by other workers (14). The bovine liver enzyme, when modified with pyridoxal phosphate under the same conditions as used for the rat liver enzyme, did not yield crystals under the conditions described below. Crystallization of pyridoxal phosphate-modified rat liver glutamate dehydrogenase was accomplished by the hanging drop method. On a silanized cover glass a lo-$ droplet of 30 mg/ml of protein solution in 0.05 M phosphate, 1 mu EDTA, pH 7.0, was mixed with 10 ~1 of 14% polyethylene glycol (MI = 6,000) solution in the same buffer. The cover glass then was inverted over a well containing 1 ml of the 14% polyethylene glycol solution, and sealed air tight with vacuum grease to prevent evaporation. After 2 to 14 days at room temperature, crystals with dimensions up to 0.5 x 0.25 x 0.25 mm developed. They displayed a hexagonal prismatic habit with well devel- oped faces along (loo), (OlO), and (001). Typical examples are shown in Fig. 1. Similar crystals were obtained for native rat liver glutamate dehydrogenase when grown under these con- ditions in the presence of 4% dioxane. The crystals were mounted in thin walled quartz capillaries together with a droplet of mother liquor for x-ray diffraction analysis. Seven- or eight-degree precession photographs were obtained with a crystal to fii distance of 75 mm. The exposure times were approximately 24 h at 18°C using CuKa radiation from a fine focus x-ray tube run at 0.8 kilowatts. Despite the fairly large size of the crystals the x-ray diffraction pattern was rather weak, although on still photographs indi- vidual x-ray reflections could be seen corresponding to Bragg spacings of at least 2.9 A. Precession photographs yielded measurable intensities only to about 6 A. Furthermore, after approximately 24 h of x-ray radiation, the crystals had lost most of their ability to diffract. Preliminary precession pho- 4915 by guest on April 28, 2018 http://www.jbc.org/ Downloaded from