Purification and Characterization of Glutamate Decarboxylase from Neurospora crassa Conidia*

L-Glutamate decarboxylase, an enzyme under the control of the asexual developmental cycle of Neurospora crassa, was purified to homogeneity from coni- dia. The purification procedure included ammonium sulfate fractionation and DEAE-Sephadex and cellu- lose phosphate column chromatography. The final preparation gave a single band on sodium dodecyl sul-fate-polyacrylamide gels with a molecular weight of 33,200 f 200. A single band coincident with enzyme activity was found on native 7.5% polyacrylamide gels. The molecular weight of glutamate decarboxylase was 30,500 as determined by gel permeation column chro- matography at pH 6.0. The enzyme had an acidic pH optimum and showed hyperbolic kinetics at pH 5.5 with a K , for glutamic acid of 2.2 mM and a K,,, for pyridoxaL5’-phosphate of 0.04 PM.

L-glutamic acid is metabolized (6). GAD is apparently responsible for the metabolism of L-glutamic acid during germination because there is almost a one-to-one correlation with the loss of L-glutamic acid and the transient appearance of GABA during the early stages of conidial germination (6,8).
GAD has been found to be a developmentally regulated enzyme in many organisms. In vertebrates and invertebrates, GAD has been associated with specific neurons of the mature central nervous system (9,(10)(11)(12), and GABA has been implicated in the regulation of neural growth during insect embryogenesis (13). GAD is apparently involved in the germination of seeds (14-16) and bacterial endospores (17,18). A mutant strain of Bacillus megaterium with low levels of GAD activity in its endospores required GABA for germination (19). In Escherichia coli GAD can be induced by growth on glutamic acid in acidic media (20). Thus, GAD is associated with differentiated cells from vertebrates, invertebrates, plants, and bacteria and appears to play a special role in the germination of some plant seeds and bacterial endospores.
GAD appears to play a role in the germination of N. crassa conidia. This enzyme is under the control of the asexual developmental cycle in N. crassa. Of the three cell types in a conidiating culture, only conidia have high levels of GAD activity.' Even though GAD activity is high in dormant conidia, it does not catalyze the metabolism of glutamic acid until germination (5,6,8). GAD activity disappears as conidia differentiate and develop into vegetative hyphae (5). The loss of GAD activity occurs more rapidly than the synthesis of new cellular protein indicating that there is a special mechanism for inactivating GAD during germination. However, the synthesis of new cellular protein is necessary for the loss of GAD activity because cycloheximide prevents the disappearance of GAD activity during germination ( 5 ) . The regulatory mechanisms that control GAD activity during conidiation and germination in N. crussa are not known. In order to investigate these regulatory mechanisms we purified GAD to homogeneity from dormant conidia. This paper describes the purification and initial characterization of this conidial-specific, developmentally regulated enzyme.

EXPERIMENTAL PROCEDURES
Materials-The wild type strain of N. crmsa (FGSC 988) used in these studies was obtained from the Fungal Genetics Stock Center, Department of Microbiology, University of Kansas Medical Center (Kansas City, KS). ~-[l-'~C]Glutamic acid (46.0 mCi/mmol) was purchased from Du Pont-New England Nuclear. DEAE-Sephadex A50-120, Sephadex G100-120, L-glutamic acid, pyridoxal 5'-phosphate, phenylmethylsulfonyl fluoride, and pepstatin A were obtained from Sigma. Cellulose phosphate P11 was purchased from Whatman. All other chemicals were reagent grade.
Glutamate Decarboxylase Assay-GAD activity was assayed by V. Parikh and J. C. Schmit, manuscript in preparation.

Fraction Number
FIG. 1. Elution profile of GAD from a DEAE-Sephadex A-50 column. For details see under "Experimental Procedures." For this figure, GAD specific activity was normalized to the fraction with the highest specific activity. 0, GAD (relative specific activity); m, protein.  measuring the amount of Con released from ~-[l-'~C]glutamic acid (5). Disposable plastic tubes (12 X 75 mm) were used as assay vessels. A scintillation pad (Arthur H. Thomas Co.) with 250 pmol of KOH was inserted in the tube 6 cm above the reaction mixture. The assay mixture contained 0.9 mM pyridoxal 5'-phosphate, 0.9 mM EDTA, 45 mM sodium piperazine-N,N'-bis(2-ethanolsulfonic acid), 0.9 mM 2-mercaptoethanol, and 10 p1 of enzyme from different purification steps. The reaction was started by adding 30 mM L-glutamic acid and 0.055 pCi of ~-[l-"C]glutamic acid which had been adjusted to pH 5.5. The total assay volume was 110 pl. After 60-min incubation a t 37 "C in a shaking water bath, 200 pl of 2 N HzS04 was injected into each tube to stop the reaction. The CO? was allowed to absorb into the KOH on the scintillation pads for a t least 2 h. The scintillation  These values were calculated based on the frequency of histidine. b T h e molecular weight of GAD was taken as 33,200 from the results of SDS-PAGE.

Fraction Number
ND. not determined.
pads were transferred to scintillation vials containing 5-ml scintillation cocktail, and the amount of I4CO2 was determined with a Beckman scintillation counter. A unit of GAD activity will catalyze the release of 1 pmol of COZ from glutamic acid per min a t 37 "C. Specific activity was defined as units of GAD activity per mg of protein.
Protein Assay-Protein concentrations of the various preparations Purification of glutamate decarboxylase from N. crassa conidia A unit of GAD activity will release 1 pmol of CO, from glutamic acid per min at 37 "C.   (20,(29)(30)(31)  were determined with protein assay kits that were purchased from Bio-Rad. Crystalline bovine serum albumin was used as the standard protein.
Preparation of Conidia-Conidia were obtained from cultures grown in Pyrex baking dishes (38 X 25 X 5 cm) containing 500 ml of minimal medium (2% glucose, 1.5% agar, and 2% Vogel's (21) salts). The baking dish cultures were inoculated with conidia from a slant tube culture and incubated for 5 days at 25 "C. The conidia were lyophilized and stored at -70 "C. The mycelial fragments contaminating the conidia preparation were removed with a sieve.
Crude Extract-The standard buffer used for purifying GAD contained 50 mM potassium phosphate, pH 7.2, 2 mM EDTA, 0.2 mM pyridoxal 5'-phosphate, 2 mM 2-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride, 0.5 p~ pepstatin A, and 10% glycerol. Approximately 15 g of lyophilized conidia were homogenized for 2 min with 0.4-0.6-mm acid-washed glass beads at 4 "C in the standard buffer in a "Bead-Beater" (Biospec product). The homogenate was centrifuged at 600 X g for 10 min at 4 "C, and the pellet was washed twice with the standard buffer. The supernatants were combined and centrifuged at 100,000 X g for 30 min at 4 "C to give the crude extract.
Ammonium Sulfate Fractionation-The crude extract was brought to 25% saturation with the addition of saturated ammonium sulfate solution. The pH was adjusted to pH 7.2 with 1 N KOH. The solution was centrifuged at 100,000 X g for 30 min at 4 "C, and the pellet was discarded. The supernatant was brought to 65% saturation with ammonium sulfate and centrifuged at 100,000 X g for 30 min at 4 "C. The pellet containing GAD was either stored at -20 "C for use at a later time or prepared immediately for DEAE-Sephadex column chromatography.
DEAE-Sephadex A50 Column Chromatography-The ammonium sulfate fraction was dissolved in a minimal amount of standard buffer and was dialyzed three times against 2 liters of standard buffer. The dialyzed protein from 15 g of conidia was applied to two 2.5-X 50cm DEAE-Sephadex A50 columns which had been equilibrated with the standard buffer at 4 "C. The columns were washed with two column volumes of the standard buffer, and GAD was eluted with a linear gradient (0.05 M potassium phosphate, pH 7.2, to 0.5 M potassium phosphate, pH 6.5). The fractions with more than 60% of the specific activity in the peak fraction from the DEAE-Sephadex columns were combined and concentrated by ammonium sulfate precipitation (70% saturation).
Cellulose Phosphate PI1 Column Chromatography-The concentrated protein sample from the DEAE-Sephadex column was dissolved in the standard buffer and dialyzed against the standard buffer in which the potassium phosphate concentration had been reduced to 5 mM. The dialyzed protein was applied to a 2.5-X 50-cm cellulose phosphate P11 column which had been equilibrated with standard buffer prepared with 5 mM potassium phosphate at 4 "C. GAD was eluted from the column with a linear gradient (0.2 M potassium phosphate, pH 7.0, to 1.0 M potassium phosphate, pH 6.0). The peak fractions with constant GAD-specific activity were pooled and concentrated either by the ultrafiltration (PM 10 membrane) at 4 "C or by 70% ammonium sulfate precipitation.
SDS-Polyacrylamide Gel Electrophoresis-SDS-polyacrylamide gel electrophoresis (SDS-PAGE) was performed on a 1.5-mm thick slab gel with a 12.5% resolving gel (22). Protein samples containing GAD activity from various stages of purification were treated with 0.25% 2-mercaptoethanol and 1% SDS prior to loading on the gel (22).
Native Polyacrylamide Gel Electrophoresis-A 7.5% polyacrylamide gel, 2 X 150 X 100 mm, was prepared with Tris-HCl (22). The running buffer contained 192 mM glycine, 25 mM Tris (pH 8.4), 2 mM EDTA, 0.2 mM pyridoxal 5'-phosphate, 2 mM 2-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride, 0.5 p M pepstatin A, and 10% glycerol. 10 pg of purified GAD from the cellulose phosphate column was applied to each of two gel tracks, and the electrophoresis was done at 4 "C at 100 V and 20 mA for 16-18 h. The adjacent tracks were cut from the gel. One gel track was stained with Coomassie Blue R-250 to obtain the protein pattern, and the other track was cut into 1.0cm lengths, chopped into small pieces, and assayed for GAD activity. The gel fragments were incubated with 400 pl of GAD assay mixture for 3 h at 37 "C.
Determination of K,,, for Glutamate and Pyridoxal 5'-Phosphate-The K,,, for glutamic acid was determined by varying the concentration of glutamic acid from 0.15 to 16 mM in the standard assay mixture. The K,,, for pyridoxal 5"phosphate was determined by varying the concentration of pyridoxal 5"phosphate from 0.01 to 0.5 p M in the standard assay mixture. Prior to determining the K,,, for pyridoxal 5'-phosphate, the enzyme solution was dialyzed against the standard buffer lackingpyridoxal5'-phosphate. The assays were done in triplicate, and the data were analyzed using nonlinear least-squares analysis (23).
Amino Acid Analysis of Purified GAD-Protein samples with norleucine as an internal recovery standard were hydrolyzed in 10 mM phenol and 5.8 N HCl (Sequanal grade, Pierce Chemical Co.) for 24, 48, and 72 h at 110 "C (24). The samples were concentrated by lyophilization and dissolved in 0.2 M sodium citrate, pH 2.2. The amino acid composition was determined with a Beckman amino acid analyzer. The concentrations of serine and threonine were determined by extrapolation to zero time of hydrolysis, and the concentrations of valine and isoleucine were determined by extrapolation to infinite time of hydrolysis (25).

RESULTS
Purification of GAD-N. crassa conidia were suspended in buffer containing 10% glycerol, 2 mM 2-mercaptoethanol, 2 mM EDTA, and 0.2 mM pyridoxal 5"phosphate to stabilize GAD. The protease inhibitors, phenylmethylsulfonyl fluoride (1 mM) and pepstatin A (0.5 wM) were included to prevent proteolytic degradation of GAD during purification. The conidia were homogenized with glass beads, and cell debris was removed by centrifugation. In the first step of the purification, GAD was precipitated with ammonium sulfate (25-65% saturation). The ammonium sulfate pellet was dissolved in a minimal amount of standard buffer and dialyzed against this buffer before application to DEAE-Sephadex A-50 columns. GAD was eluted from the DEAE-Sephadex columns with a linear potassium phosphate-pH gradient (Fig. 1). GAD began to elute from these columns at a potassium phosphate con-centration of 0.18 M. Only a single peak of GAD activity was detected. The fractions with GAD specific activity of 60% of the peak fraction were combined, concentrated by ammonium sulfate precipitation (70% saturation), dialyzed against 0.05 M potassium phosphate, pH 7.2, containing all of the other components in the standard buffer, and applied to a cellulose phosphate P11 column.
GAD began to elute from the cellulose phosphate P11 column at a potassium phosphate concentration of 0.3 M (Fig.  2). The protein profile from this column showed the same pattern as did GAD activity indicating that there were few contaminating proteins. The specific activity through the peak fractions was constant. The protein in the combined fractions from the cellulose phosphate column gave a single band (Fig. 3) with an apparent molecular weight of 33,200 k 200 on SDS-polyacrylamide gels (Fig. 4). Occasionally, higher molecular weight bands corresponding in size to aggregates of GAD were seen on SDS-PAGE. A single band was also seen on a nondenaturing gel (data not shown). When the nondenaturing gel was cut into sections and assayed for GAD activity, enzyme activity was found associated with the protein band.
Characterization of GAD-Purified GAD exhibited Michaelis-Menten kinetics and with an apparent K, of 2.2 mM for glutamic acid and 0.04 PM for pyridoxal 5"phosphate. GAD had an acidic pH optimum with maximum activity at pH 5 and very little activity at pH 7.0 in citrate-phosphate buffer (26). An acidic pH optimum had been reported previously using other buffer systems with crude preparations of GAD from N. crassa (5).
The molecular size of native GAD was determined by gelpermeation column chromatography. At pH 6, purified GAD was retarded on Sephadex G-100 columns with an apparent molecular weight of 30,500 (Fig. 5). In the standard buffer at pH 7.2, GAD eluted in the void volume from both Sephadex G-100 and Sephacryl S-200 gel-permeation columns (data not shown). Apparently GAD can form high molecular weight aggregates under these conditions. The amino acid composition of purified GAD was determined (Table I). GAD contained 11% acidic amino acids and their amides (Asx and Glx), 11% basic amino acids (Lys and Arg), and 37% hydrophobic amino acids (Pro, Val, Ile, Leu, Tyr, and Phe). A minimum molecular weight from the amino acid composition based on histidine was 34,800 and, based on glutamic acid plus glutamine, was 32,800. These values compared favorably with 33,200 obtained by SDS-PAGE and 30,500 obtained by Sephadex G-100 gel-permeation column chromatography at pH 6.

DISCUSSION
This paper is the first report of the purification of GAD from fungi, and the first report of the purification to homogeneity of an intracellular enzyme from N . crassa conidia.
GAD was purified 115-fold from crude protein extracts of conidia (Table 11). The purification procedure consisted of ammonium sulfate precipitation, DEAE-Sephadex column chromatography, and cellulose phosphate P11 column chromatography. Cellulose phosphate P11 column chromatography was very effective for purifying GAD. GAD apparently has a high affinity for the phosphate groups on the column.
About 1 mg of GAD was obtained from 15 g of dry conidia with a 11% yield. The specific activity of purified GAD was 2.7 units per mg of protein, GAD was a relatively abundant protein constituting about 1% of the protein in the crude extract of conidia.
The molecular weight, subunit composition, specific activ-ity, K,, and pH optima of purified GADs from N . crassa snd several other organisms are summarized in Table 111. GAD has been purified from vertebrates (9, 10, 27), invertebrates (12), squash (28), and Escherichia coli (20,29). The K, for glutamic acid for the N . crassa conidial enzyme is intermediate among those reported for GAD from various sources (Table  111). The pH optimum for purified N . crassa GAD was in the acid range. GAD with acidic pH optima have been found in E. coli (30,31) and in plants (15,16,28). The active form of N. crassa conidial GAD is apparently a monomeric protein. SDS-PAGE under reducing conditions gave a molecular weight of 33,200 (Fig. 4), and gel-permeation column chromatography at pH 6 gave a molecular weight of 30,500 (Fig. 5). Since there is little GAD activity when this enzyme is assayed at pH 7.0 (5, and this paper), the monomeric form of N. crassa GAD is presumably the active species.
In contrast, purified GADs from other organisms exist as multimeric proteins of identical (29) and nonidentical (9-12) subunits (Table 111). Purified N. crassa conidial GAD can aggregate to higher molecular weight forms. Aggregates of GAD were detected on some of our SDS gels, and on gel-permeation columns when these columns were run at pH 7.2 (see "Results"). GAD from bacteria (29) and rat brain (9) also form aggregates. The purification of GAD was done largely at pH 7.2. The aggregation of GAD at this pH did not apparently adversely affect purification. The high ionic strengths and lowered pH of the elution buffers for the DEAE-Sephadex and cellulose phosphate columns would presumably favor the dissociation of GAD, though this has not been tested. We found no indication that there are additional forms of GAD in N. crassa conidia other than the 33,200-molecular weight form and its aggregates. Multiple molecular forms of GAD have been observed in mammals (9-11) and plants (15,28). Whether these multiple forms represent distinct species of GAD or different aggregational states remains to be resolved (13).