(“WWWlWlllUlUlWlWIMHHI‘U(WI 7—! 18°13 yap-co “— w...“ A A“ LIBRARY Michigan State " ~Univcrsity ' TH E8" This is to certify that the thesis entitled PROLINE DbHYDROGENASE FROM CLOSTRIDIUM SPOROGENES: PURIFICATION AND PARTIAL CHARACTERIZATION presented by Daniel J. Monticello has been accepted towards fulfillment of the requirements for Master' Schree in Microbiology Alf/1'19? d Maj - professor Date June 7, 1979 07539 OVERDUE FINES ARE 25¢ PER DAY PER ITEM Return to book drop to remove this checkout from your record. PROLINE DEHYDROGENASE FROM CLOSTRIDIUM SPOROGENES: PURIFICATION AND PARTIAL CHARACTERIZATION By Daniel Joseph Monticello A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Microbiology and Public Health l979 ABSTRACT PROLINE DEHYDROGENASE FROM CLOSTRIDIUM SPOROGENES: PURIFICATION AND PARTIAL CHARACTERIZATION By Daniel Joseph Monticello A single enzyme catalyzing the L-proline-dependent reduction of NAD+ and the A-pyrroline 5-carboxylic acid (PCA)-dependent oxi- dation of NADH has been purified from extracts of Clostridium sporogenes. The enzyme has a molecular weight of 217,000, based on calculations from linear sucrose gradient centrifugation data, and is composed of two subunits, each weighing 108,000, based on $05 analytical disc gel electrophoresis. The L-proline-dependent NAD reducing activity of the enzyme was found to be more sensitive to incubation in low ionic strength buffer than the PCA-dependent NADH oxidizing activity of PDH. Proline dehydrogenase is inhibited by glutathione, cysteine, copper sulfate, p-chloromercuribenzoate and adenine nucleotides. Inhibition of the PCA-dependent NADH oxidizing activity of PDH by hydroxylamine was shown to be due to a reaction of the inhibitor with the substrate (PCA) and not to inactivation of the enzyme. The conversion of proline to PCA by PDH is noncompeti- tively inhibited by L-glutamate (Ki = 0.23 mM at pH 7.4, and 0.65 mM at pH 10.2). PDH activity in the reverse direction (PCA to proline) is not affected by 100 mM L-glutamate. DEDICATION To my family and friends, who saw me as I might be, and spent the time to help me there, and Elizabeth, who saw me as I am, and understands as we keep changing. ii ACKNOWLEDGMENTS I wish to express my sincere gratitude to my major professor, Dr. R. N. Costilow, for his guidance throughout the course of this investigation and the preparation of this thesis. I would also like to express my thanks to Dr. H. L. Sadoff and Dr. R. R. Brubaker for the use of their laboratory facilities. TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES INTRODUCTION LITERATURE REVIEW The Interconversion of Ornithine, Glutamate and Proline . . The Conversion of Ornithine to Proline PCA Reductase . Proline Oxidase Proline Dehydrogenase MATERIALS AND METHODS Culture and Cultural Methods . Growth Media . . Buffers . . Enzyme Assays . . Preparation of AI-Pyrroline 5- -Carboxylic Acid Gel Electrophoresis . Sucrose Density Centrifugation Column Chromotography Chemicals . RESULTS . Purification of the Enzyme . Characterization of the Enzyme Induction of PDH by L-Proline DISCUSSION . LIST OF REFERENCES iv Page vi Table LIST OF TABLES Page Purification scheme for proline dehydrogenase . . . 25 Inhibition of proline dehydrogenase and PCA reductase activities . . . . . . . . . . . . . . . 45 Effect of PCA concentration on inhibition by hydroxyl- amine . . . . . . . . . . . . . . . . 47 The effect of L-proline concentration in the growth medium on the specific activity of proline dehydrog- enase in crude extracts of C, spprogenes . . . . . 54 Figure LIST OF FIGURES Postulated relationship of ornithine, proline, and glutamate in Clostridium sporogenes . . Elution profiles from preliminary experiments with DEAE-cellulose columns . . . . . . Elution profiles of PDH activity and GDH activity from the DEAE-cellulose column used in the final purification scheme . . . . Elution profile from hydroxylapitite column chromatog- raphy . . . . . . . . . . . . . Gel scan of the purified enzyme . Profile of catalase, proline dehydrogenase, and alcohol dehydrogenase activity after sucrose density centrifugation . . . . . . . Comparison of the mobility of proline dehydrogenase and four protein standards on sodium dodecyl sulfate analytical disc gels . . . . . . . . Effects of freezing and thawing on PCA red activity Eddie-Schattard plots of L-glutamate inhibition data . Dixon plot of the reciprocal of the Vmax versus con- centrations of inhibitor at pH 7.4 and pH l0.2 . vi Page 28 3O 33 36 38 41 44 49 51 INTRODUCTION The conversion of A1-pyrroline 5-carboxylic acid (PCA) to proline by PCA reductase [L-proline:NAD(P)+ 5-oxidoreductase ECl.5.l.2] has been observed in animal tissue (29, 41, 42) and a number of microorganisms (9, 33). This activity is irreversible, and is believed to function in the biosynthesis of proline from glutamate or ornithine (which is converted to PCA via ornithine-6— transaminase). In these organisms, proline is oxidized to PCA by an irreversible proline oxidase (4, 12, 18). Laycock and Costilow (7) observed that crude extracts of cells of Clostridium sporogenes and Clostridium botulinum catalyse the nicotinamide adenine dinucleotide (NAD)-dependent interconversion of L-proline and PCA. Costilow and Cooper (6) demonstrated that NAD-linked proline oxidation and NADH-linked PCA reduction in g, sporogenes is probably catalyzed by the same enzyme. MAD-dependent proline dehydrogenases (PDH) have also been identified in a number of plants (21, 22, 3l) and in Chlorella (23). There is some evi- dence that these enzymes are reversible. Clostridium sporogenes (26) and some higher plants (24) convert ornithine to proline via Al-pyrroline 2-carboxylic acid (P2CA). Obviously, these organisms do not require PCA reductase for this conversion. It has been suggested that PDH may function in the oxidation of proline to glutamate in g, sporogenes (6). This is supported by the observation that the oxidation of proline to PCA is inhibited strongly by L-glutamate (Costilow and Cooper, unpublished data). The principle objectives of this investigation were: 1. Purification of PDH to unequivocally determine if the NAD—linked oxidation of proline and the NADH-linked reduction of PCA are catalyzed by the same protein. 2. Determination of the molecular parameters of the enzyme such as molecular weight, number and size of subunits, and inhibitors. 3. Quantitative determination of the effects of L-glutamate on PDH activity. LITERATURE REVIEW The Interconversion of Ornithine, Glutamate and Proline The structural similarity of the amino acids ornithine, proline and glutamate has provoked much experimentation and specu- lation as to their possible metabolic interrelationships. The suggestion that animals can convert glutamate to proline was first made by Abderhalden (l) who, in 1912, demonstrated in dogs that protein hydrolysates rich in glutamate but with greatly reduced amounts of proline (which had been alcohol extracted from the hydrolysates) were as nutritionally effective as whole hydrolysates. As early as 1910, Neubaur and his colleagues (27) presented experimental evidence that natural amino acids were oxidized to their corresponding keto acids and ammonia in rat liver perfusion studies. In 1936, Bernheim, Bernheim and Gillaspie (3) used rat kidneys to examine the oxidation of amino acids. They were able to follow the course of the oxidations manometrically, and to assay for the keto acids fOrmed from various amino acids by precipitation with either phenylhydrazine or sodium bisulfite. Such treatment led to the formation of yellow crystals which were easily visualized. Additional of proline to a purified preparation, followed by addi- tion of bisulfate after the completion of oxidation led to the formation of crystals. However, when proline was added to tissue slices, no crystals were formed on addition of bisulfite. This led the authors to hypothesize: . . A scheme fitting all these facts would involve the loss of one hydrogen atom from the nitrogen and one from the adjacent carbon to which the carboxy group is not attached. This would leave a double bond which might hydrolyze to the corresponding aldehyde. The purified preparation evidently takes the oxidation no further and this would account for the formation of the bisul- fate compound. With the tissue slices, however, this aldehyde could be oxidized to the acid, thus giving glutamic acid. This hypothesis was supported by the work of Heil-Malherbe and Krebs (42), who observed the jn_vitro conversion of proline to glutamate in rat tissue in 1935, and by Krebs (14), who demonstrated the oxidation of D-proline to AI-pyrroline-Z-carboxylic acid (P2CA) in 1939. These observations were the beginning of an investigation into the interconversion of glutamate and proline which would include species as diverse as rates, fungi, bacteria and plants. The first experiments demonstrating the jfl_vivo conversion of proline to glutamate were performed by Stetten and Schoenheimer (38) in 1943. Deuterium labeled proline (produced by shaking a-pyridone, an organic precursor of proline, in 99.6 atm of deuterium gas at 100°C fin~7 hours) and 15N-proline were used in this study to determine the fate of proline fed to rats. These experiments demonstrated that the nitrogen in proline remained with the molecule in its conversion to glutamate. This was strong experimental support for the hypothesis proposed earlier by Benheim et al. (3). Concurrently, Blanchard et al. demonstrated jg_vitro that a single amino acid oxidase from rat liver was responsible for the oxidation of at least 11 naturally occurring amino acids. In 1949, Taggert and Krakaur (40) demonstrated in rabbit kidney preparations that the intermediate between proline and l-pyrroline-S-carboxylic acid (PCA), and that it glutamate was A was in spontaneous equilibrium with glutamic y-semialdehyde. Lang and Schmid (l7) reproduced these results in 1951. The metabolic pathway between proline and glutamate in Escherichia coli was being examined at the same time by Vogel and Davis (41), by means of metabolic studies on mutants. In 1952, they demonstrated that PCA was the intermediate in this pathway, and showed that some proline auxotrophs were able to grow when supplied with PSCA. They obtained similar results in proline auxotrophs of Neurospora grease: In 1957, Meister, et a1. (25) suggested that PZCA and PCA might be involved in alternate pathways between glutamate and proline. They demonstrated that these compounds were reduced to proline by two different enzymes in Neurospora crassa and Aerobacter aerogenes. In addition, they found an enzyme capable 1-piperidine 2-carboxy1ic acid to pipecolic acid of reducing both A and P2CA to proline, in rat tissue and in Pisum sativum (garden peas). One of the early problems in the experiments with P5CA was the difficulty of preparing the compound in a pure form. In 1960, Strecker (39) overcame this problem and was able to purify and to some extent characterize the compound. In this work, he notes that P5CA does not make a very stable addition product with bisulfite, but will react with 2,4-dinitrophenylhydrazine to make a stable compound. Both of these reagents were believed to be attack- ing the free aldehyde (glutamic y-semialdehyde) which is in spontan- eous equilibrium with PCA. The Conversion of Ornithine to Proline Muth and Costilow (26) demonstrated that ornithine is con- verted to proline in g, sporogenes via o—keto—a-aminovaleric acid and P2CA intermediates. They utilized uniformly labeled ornithine (14C), also labeled in the G-amino group with 15N, and demonstrated that this nitrogen atom was conserved in the enzymatic cyclization of ornithine to proline. In addition, they showed that this con- version was catalyzed by a single enzyme, ornithine cyclase (deaminating). g, sporogenes does not convert ornithine to PCA via an ornithine-é-transaminase pathway. If glutamate is produced from ornithine in this species, the most likely pathway is via proline and PCA, as outlined in Figure 1. Recently (1979), Mestichelli, et al. demonstrated that ornithine is converted to proline via P2CA in several higher plants (Nicotiana tabacum, Datura stramonium and Lupinus angustifolius). 3 Using tracer methods with ornithine labeled with H and 14C, these investigators demonstrated that the conversion to proline takes place with the maintenance of the 6-hydrogen atoms but with the loss of the a-hydrogen atoms. This indicated a route via a-keto-G- aminovaleric acid and P2CA, and disproved the accepted route via glutamate NH 2 COOH COOH \ glutamic 8-semialdehyde "H2 0” \H COOH A'-pyrroline 5—carboxylic acid \N coow H2 [ 1 MHz OOH N coow ornithine PFOIIHe Figure 1.--Postulated relationship of ornithine, proline, and glutamate in Clostridium sporogenes. glutamic y-semialdehyde and PCA. Enzymes for the conversion of proline to PCA has been found in several higher plants (21, 22, 31). It seems likely that the pathway for the conversion of orni- thine to glutamate outlined in Figure 1 may be operational in some higher plants. PCA Reductase In 1964, the nature of the regulation of proline biosyn- thesis in E, gglj_was further characterized. Baich and Pierson (2) showed that in proline biosynthesis, control is localized in the first reduction step from glutamic acid to glutamic y-semialdehyde. This reaction is strongly inhibited by small amounts of proline in growing and resting cells. The second reaction, the reduction of PCA (which is in spontaneous equilibrium with glutamic y-semi- aldehyde) to proline by PCA reductase (PCAred) proceeds unrestrained in the presence and absence of proline. They isolated a proline excreting mutant, and in this strain the reduction of gluamic acid to glutamic y-semialdehyde is relatively insensitive to proline. In 1977, PCA reductase from g, £911 was partially purified and characterized by Rossi et a1. (33). The enzyme, with an approx- imate molecular weight of 320,000, was found to have a similar Km for PCA regardless of whether NADH or NADPH was used as a cofactor. The Km of the enzyme was much higher for NADH than NADPH (0.23 and 0.03, respectively). They observed that this non-particulate enzyme was not repressed by growth in the presence of proline, but was inhibited by the reaction and products, proline and NADP. This enzyme is not reversible even at very high substrate concen- trations. Costilow and Cooper (6) identified a PCAred in extracts of Q, sporogenes cells which was specific for L-PCA and NADH. The Km for L-PCA of this enzyme was found to be 0.33 mM at pH 8.0, and 0.2 mM at pH 6.5. Based on a sedimentation coefficient of 10.38 they estimated the molecular weight of the enzyme to be 200,000 daltons. In addition, the PCAred activity capurified extensively with proline dehydrogenase which catalyzes the reverse reaction. This evidence suggested that the activities were located on the same protein moiety. However, a completely purified preparation was not obtained. Praline Oxidase In 1962, Johnson and Strecker (12) demonstrated that the oxidation of proline in rat liver required oxygen and cytochrome c. In addition, nan-physiological oxidants could serve as electron acceptors. They found that pyridine nucleotides were not required for activity, nor were they reduced in the presence of proline and the liver preparation. They suggested that the proline oxidase was linked to the respiratory chain, and did not require dissociable coenzymes. Soon after this, Peisach and Strecker (29) demonstrated a non-reversible, pyridine nucleotide linked PCAred (Al-pyrroline 5-carboxylic acid specific) was responsible for the reduction of PCA to proline in calf liver extract. Both NAHD and NADPH served as electron donors, although NADPH was the preferred substrate. 10 These observations added support to the growing belief that the conversion of glutamate to proline is catalyzed by enzymes which differ from those used in the oxidation of proline. In 1964, Ling and Hendrick examined proline oxidase in the yeast Hansenula subpelliculosa (18). They found two distinct enzymes. One was constitutive, and it converted very small amounts of proline to P2CA. They speculated that this enzyme was associated with the use of proline as a nitrogen source. The second enzyme was inducible with proline, and yielded PCA. They also reported the presence of NAD-linked enzymes in the crude extracts able to oxidize both P2CA and PCA. In the same year, Frank and Ranhand (11) examined proline catabolism by g, 9911, They noted that proline was oxidized in the same fashion as had been previously noted in animal tissue. Praline was oxidized to PCA by a membrane bound oxidase, and the PCA oxi— dized to glutamate by a pyridine nucleotide linked PCA dehydro- genase. Analysis of mutants demonstrated that the proline oxidase and PCA reductase were unrelated. Mutants with one activity and not the other were readily isolated. However, they also observed that it was impossible to separate the proline oxidase and the PCA dehydrogenase, and suggested that the entire proline oxidase complex may exist as a single physical entity. DeHauwer et a1. (9) in 1964, and Dendinger and Brill (10) in 1970, studying Bacillus subtilis and Salmonella typhmurium respectively, demonstrated that these bacteria had pathways for proline degradation identical to E, coli. Laishley and Bernlohr 11 (16), in 1968, presented some evidence to suggest that PCA nay be the inducer of proline catabolism in B, licheniformis. The proline oxidase from E, gglj_was purified and charac- terized in 1978, by Scarpulla and Soffer (34). They were able to separate the enzyme from PCA dehydrogenase, to which it is tightly coupled in the bacterial membrane (11). They demonstrated that the fbrmation of PCA was more than 99% dependent on the presence of an artificial electron transport system (consisting of phenazine methosulfate and p-iodonitrotetrazolium) in the assay preparation. This indicates that molecular oxygen does not function as a proxi- mate electron acceptor. The authors suggested that the enzyme be termed a dehydrogenase, to distinguish its activity from the amino acid oxidases which are directly linked to molecular oxygen. In a subsequent paper by these authors in 1979 (35), they examined the role of 1eucyl-, phenylalanyl-tRNA:protein transferase in the regulation of the membrane bound proline oxidizing enzyme. In mutants lacking the transferase, an increase in the level of proline oxidation was observed. In this investigation, the authors exclude the possibility of modification of the proline oxidase demonstrating that the increased activity is due to more of the enzyme in the mutants than the wild-type cells. They suggest rather than modifying the enzyme by direct covalent modification, the transferase, or an acceptor substrate(s) may perhaps be a regu- latory molecule which controls the biosynthesis of proline oxidase (dehydrogenase) at the transcriptional or translational level. 12 Praline Dehydrogenase In 1969, Costilow and Laycock (7) reported that Q, sporogenes and g, Botunlinum Type A contained a pyridine nucleotide linked PCAred. In addition, they demonstrated far the first time a pyr- idine nucleotide linked proline oxidation. The oxidation was spe- cifically NAB-dependent and the product was PCA. Subsequently, NAD-linked proline oxidation has been observed in several higher plants and in algae. In 1971, Mazeliis and Fowden found a proline dehydrogenase in peanut seedlings (22). This non-particulate enzyme was specific for L-proline and NAD. The pH optimum of the enzyme was 10.3, and NADP acted as a competitive inhibitor. The product of the oxidation was not identified, but it was not believed to be P2CA or PCA. However, a proline dehydrogenase (PDH) from wheat germ was identified in 1974 by Mazelis and Creveling (21) which converted proline to PCA. This enzyme was reversible. The dehydrogenase activity was specific for L-proline and NAD, and the reductase activity for PCA and NADH. NADP and NADPH were good competitive inhibitors. They estimated the molecular weight of the molecule to be 200,000 daltons. Shortly after these initial findings, PDH was reported as being present in two other systems. Rena and Splittstoesser (31) demonstrated an NAD-linked proline oxidation, and an NADH-linked PCA reduction in pumpkin cotyledons. Although unable to purify the enzyme, their data suggested that the two activities were catalyzed by the same protein molecule. They noted that the ratio of PDH:PCAred changed during the storage of the enzyme, and that 13 PCAred activity was inhibited to some extent by sodium bisulfite, and to a much greater extent by hydroxylamine. McNamer and Stewart (23) found PDH activity in Chlorella. The activity had a pH optimum of 10.2, and was specific for L-proline and NAD. They estimated the molceular weight as ". . . greater than 100,000. . . ." The reversibility of the reaction was apparently not studied. In 1978, Costilow and C00per reported on the partial puri- fication of PDH from Q, sporogenes. They demonstrated that the two activities coelute from diethylaminoethyl (DEAE)-ce11ulose, hydroxylapatite and Sephadex G-200 columns. Both have identical sedimentation coefficients and isoelectric points and are heat- stabilized by high ionic strength buffer. The activities of both PDH and PCAred could be reduced by 50% when glucose was added to the growth medium. The pH optima for the two activities were very different, pH 10.2 for PDH and 6.5-7.5 for PCAred. A small increase in the pH resulted in large shifts of the reaction equilibrium toward PCA. However, even at pH 8.6 the equilibrium constant for PDH was about 2.5 x 10's. The authors proposed that L-proline and L-PCA are interconverted by ". . . either a single enzyme or an enzyme complex . . ." in extracts of Q, sporogenes cells. Some doubt has been cast on the significance of PDH in plants. Boggess et a1. (5) in 1978, demonstrated that mitochondria isolated from etiolated shoots of corn, wheat, barley, soybean and mung bean exhibited a proline dependent uptake of oxygen subject to respiratory control and independent of NAD concentration. These authors suggest that PDH may not play a role in proline oxidation 14 in.vjxg, based on the facts that (l) ". . . the necessity to assay it at high pH . . .," and (2) . . . that it copurifies with PCA reductase, the NADH-linked proline biosynthetic enzyme that is typically stable and present in relatively high activity. . . ." However, Costilow and Cooper indicate in a paper published at the same time (6) that the high pH optimum of PDH is to be expected if the oxidation of proline and the reduction of PCA are catalyzed by a reversible enzyme. In organisms which convert ornithine to proline via P2CA, proline dehydrogenase may play an important role in the oxidation of ornithine and proline to glutamate. Until recently, only species of the clostridia (7, 26) had been shown to convert ornithine to proline in this manner. In most organisms, ornithine is thought to be converted to proline via a ornithine-B—transaminase enzyme (which converts ornithine to PCA) and then a PCA reductase. Such a pathway would not require the presence of PDH for the conversion of ornithine to glutamate. However, Mestichelli et al. (24) have shown that several species of plants convert ornithine to proline via PZCA. In addition, they argue that the evidence supporting the proposal that ornithine is converted to proline via PCA in other plant, animal and microbial systems can be interpreted to show that this conversion results from the loss of the a-amino group from ornithine (yielding P2CA), not the 6-amino group (which yields PCA). Obviously, this would make PDH very important for the conversion of ornithine to glutamate in these systems. 15 It is clear that despite the large amount of accumulated information on the interconversions of ornithine, proline and glut- amate, this system is still not well understood. The times seems ripe for a re-examination of the metabolic relationship among these amino acids. MATERIALS AND METHODS Culture and Cultural Methods Clostridium sporogenes (ATCC 7955, National Canners Associa- tion PA 3679) was used in all experiments. Large batches of the cells far enzyme purification were regularly grown in 20 liter glass carboys at 37°C. All growth experiments were conducted utilizing a Coy Manufacturing Co. anaerobic chamber. Growth Media The media used in this investigation were: Medium A: 4.0% trypticase, 2 ppm thiamine-hydrochloride and 0.05% sodium thioglycollate, brought to pH 7.4 with sodium hydroxide. Medium 8: Modified Perkins and Tsuji (30) synthetic medium, consisting of salts, vitamins, sodium thioglycollate and lOmM con- centrations of L-arginine, glycine, L-histidine, L-leucine, L-lysine, L-serine, L-methionine, L-phenylalanine, L-threonine, L-valine, L-tyrosine, with 0.5 mM concentrations of L-tryptophan and L-cysteine. Buffers The buffers employed in these studies were: Buffer A: 0.15 M tris(hydroxymethy1)aminomethane-chloride buffer, 2 mM dithiothreitol, 10 mM L-glutamate (free base), 10% glycerin, pH 7.4. 16 17 Buffer B: 20 mM potassium phosphate buffer, 2 mM dithio- threitol, 10% glycerin, pH 7.4 Buffer C: 0.25 M tris(hydroxymethy1)aminomethane-chloride buffer, pH 7.4. Buffer D: 0.25 M potassium phosphate buffer, pH 7.4. Enzyme Assays Praline dehydrogenase (PDH) was assayed by monitoring the proline-dependent reduction of NAD, or the PCA-dependent oxidation of NADH. Assays for proline oxidation referred to as PDH determina- tions were conducted at two pH levels. In both assays the rate of reduction of NAD was fbllowed by monitoring the increase in absorb- ancy at 340 nm (A340). Reaction mixtures of 1 m1 at pH 10.2 con- tained 0.2 M sodium bicarbonate buffer, 10 mM NAD and 50 mM L-proline. At pH 7.4, 0.32 M potassium phosphate buffer was used instead of bicarbonate. In studies of L-glutamate inhibition, 10 mM L-gluta- mate was added to the reaction mixture prior to the addition of L-proline, to ensure that there was no glutamate dehydrogenase activity in the enzyme preparation. The NADH-linked reduction of PCA by PDH (referred to as PCAred in the text) was assayed by moni- toring the loss of absorbancy at 340 nm. Typical reaction mixtures contained 0.32 M potassium phosphate buffer (pH 7.4), 0.75 mM PCA, 0.1 mM NADH and enzyme. Endogenous oxidation of NADH was corrected for when necessary. Glutamate dehydrogenase (GDH) was assayed in reaction mix- tures of 0.25 M potassium phosphate buffer, pH 7.4, 0.1 mM NADH, 18 110 mM ammonium chloride, 50 mM a-ketoglutarate and enzyme. The a-ketoglutarate-dependent oxidation of NADH was monitored at 340 nm. Catalase activity in the sucrose density gradient experiments was assayed with 20 mM hydrogen peroxide in 50 mM potassium phosphate buffer at pH 7.4. Loss of absorbancy at 240 nm was monitored. Reaction mixtures for alcohol dehydrogenase contained 320 mM ethanal, 10 mM NAD, in 50 mM sodium pyrophosphate buffer, pH 8.8. For all of the enzymes, one unit of enzyme activity is defined as that amount of enzyme necessary to convert l pmole of substrate to 1 pmole of product in 1 minute. Specific activity is defined as units of enzyme per milligram of protein. Protein con- centrations were routinely assayed by the method of Kalb and Bernlohr (13) and in some cases the method of Lowry et a1. (19). In the assays utilizing NAD and NADH, the millimolar extinction coefficient of reduced nicotinamide adenine dinucleotide employed was 6.22 mM"1 cm']. Preparation of Al—Eyrroline 5-Carboxy1ic Acid DL-PCA was produced by the method of Williams and Frank (43), which involves the peroxidation of 6-hydroxy1ysine to glutamate-semialdehyde, which is in equilibrium with PCA. Quanti- tative assays of PCA were performed by measuring the color farmed with o-aminobenzaldehyde, using a millimolar extinction coefficient 1 of 2.94 mM'] cm" at 444nM (43). This assay is based on the forma- tion of yellow dihydroquinolinium salts when cyclic imines react 19 with o-aminobenzaldehyde. Preparation were stored until use in 1 N hydrochloric acid at -20°C. Gel Electrophoresis Native protein disc gel electrophoresis was accomplished by the method of Davies (8). Gels were stained far protein using Coomassie blue G, and destained in 7% acetic acid. PDH activity in the gels was detected by incubating the gel at 37°C in23 ml of 25 mM tris-chloride buffer (pH 7.5) containing 300 mM L-proline, 2 mM NAD, 12 mg of phenazine methosulfate and 2.5 mg of nitroblue tetrazolium. GDH activity in the gel was located in a similar fashion, replacing proline with 10 MM L-glutamate. Subunit molecular weight was determined by the sodium dodecyl sulfate (SDS) disc gel electrophoresis method of Laemmli (15). Samples of the purified protein were mixed with buffer containing 3% SDS and 5% 2-mercaptoethanol, and subsequently heated at 90°C for 15 minutes. Samples were layered onto 7% acrylamide gels (10 cm in length) with a 1 cm 3% acrylamide stacking gel. The electrode buffer consisted of 0.1% $05 in 25 mM tris-193 mM glycine buffer, pH 8.3. A current of 1 ma per tube was applied through the gels until the dye (bromphenol blue) front just entered the lower gel, at which time the current was increased to 2 ma per gel. The gels were immersed overnight in a 10% trichloroacetic acid, 33% methanol solution, to fix the protein and extract the SDS. Protein bands were developed with Coomassie blue 6, and destained in 7% acetic acid. 20 Sucrose Density Centrifugation A value for the molecular weight of PDH was obtained using linear sucrose gradients (6 to 30%) by the method of Martin and Ames (20). Samples of the purified enzyme (2 pg) were combined with alcohol dehydrogenase (yeast) and catalase (bovine liver) standards in 50 mM potassium phosphate buffer (pH 7.4) with a total volume of 150 pl, and layered on the gradient. The gradient was centrifuged in an SN50.1 rotor in a Spinco model L ultracentri- fuge (Beckman-Spinco) at 35,000 rpm for 10.5 h. Two-drop fractions were collected (43 total fractions), and assayed far the enzyme activities. Column Chromatography In preliminary experiments, a variety of chromatographic systems were analyzed to determine their utility in purifying the enzyme, and in the separation of PDH from traces of GDH activity. The methods used in the final preparation are described below. DEAE-Cellulose Chromatography Diethylaminoethyl(DEAE)-ce11ulose (Cellex D, Biorad Labora- tories) was equilibrated with buffer A as per the manufacturer's instructions, and used to prepare a 3.0 x 45 cm column. A sample previously equilibrated by dialysis against buffer A was layered onto the column. The column was washed with buffer A, and the effluent monitored at 280 nm with an Isco model UA-2 absorbance recorder and optical unit. After an initial absorbance peak was observed and the recorder returned to baseline, a linear gradient 21 was developed from 0.15 to 0.30 M tris-chloride, pH 7.4 (250 ml resevoirs). Concentrations of DTT, glutamate and glycerin were as in buffer A. In preliminary experiments NAD or L-proline were sub- stituted for L-glutamate in the elution buffers. Fractions of 3 ml were collected, with a flow rate of 40 ml/h, and assayed for PCAred and GDH. Fractions with a PCAred:GDH ratio of greater than one were pooled, and concentrated to a final volume of 5 ml by ultra- filtration through a Diaflo PM-lO (Amicon Corp.) membrane using nitrogen as the pressurizing gas. Sephadex G-10 Desalting Column A 1 x 20 cm column of Sephadex G-10 (Pharmacia Fine Chemicals) column was employed to quickly desalt enzyme prepara- tions, and equilibrate the protein in a new buffer system. Samples of 5 ml applied to the column eluted off in 7 ml, after a void volume of 8 ml. Hydroxylapatite Column Chromatography A 1.5 x 30 cm hydroxylapatite column (Biogel HTP, Biorad Labs) was prepared and poured as per the manufacturer's instruc- tions, and equilibrated with buffer 8. Enzyme preparations equilib- rated into this buffer with a Sephedex G-10 column were applied to the column and washed with a sufficient volume of the starting buffer to return the A280 recorder to baseline. The enzyme was eluted with a linear gradient from 0.02 to 0.30 M potassium phosphate buffer, pH 7.4 (150 m1 of each). Concentrations of DTT and glycerin 22 were as in buffer B. Fractions of 2.5 ml were collected with a flow rate of 30 ml/h, and assayed for PCAred and GDH. Fractions containing more than 10% of the PCAred activity observed on the peak fraction were pooled and concentrated through a Diaflo PM-lO membrane to a final volume of 5 ml. Biogel A-5M Column Chromatography A 2.5 x 45 cm column of Biogel A-SM (Biorad Labs) was equilibrated with buffer C. An elution volume of 50 ml was deter- mined for blue dextran (molecular weight 2,000,000). A 5 ml sample was applied to the column, and after 80 m1 had eluted off, 0.5 ml fractions were collected (flow rate of 10 ml/h), and assayed for PCAred and GDH. Fractions with PCAred activity and no detectable GDH activity were pooled and concentrated to a final volume of 5 m1. Sephadex G-200 Column Chromatography A 2.5 x 45 cm column Sephadex G-200 (Pharmacia) was equil- ibrated overnight in buffer C. A void volume of 70 ml was deter- mined with blue dextran. A 5 ml sample was layered on the moist column bed, and after it was moved into the column, and washed in with small volumes of buffer, a head pressure of 10 cm was applied. Fractions with PCAred activity were pooled, concentrated to a volume of 5 ml, and stored in 0.5 m1 aliquots at -21°. 23 Chemicals o-Aminobenzaldehyde, y-hydroxylysine (a mixture of hydroxy- DL-lysine and allohydroxy-lysine), NAD and NADH were obtained from Sigma Chemical 00. These and all other chemicals used were of the highest standards of purity available. RESULTS Purification of the Enzyme A summary of the effectiveness of the purufication procedure is given in Table 1. Whenever possible, all enzyme preparations were maintained in high ionic strength buffer and held at refriger- ator temperatures. Procedures used in individual steps are outlined below. Growth of Cells For the batchwise purification of PDH, 18 liters of medium A in a 20 liter glass carboy were inoculated with 2 liters of exponen- tially growing 9, sporogenes previously cultured in an anaerobic chamber (see Materials and Methods). The carboy was equipped with air locks and stirred slowly with a magnetic stirrer at 37°C. After 8-10 hours, the cells were harvested with a Sharples continuous flow centrifuge, model AS-lO. The collected cells were suspended in buffer 0, 10 ml per gram wet weight of cells. Preparation of Crude Extracts Although sonication was used in some early experiments, crude extracts were usually prepared with a French press, which was fbund to be more effective. Cell debris was removed by centrifugation at 20,000 x g for 20 minutes. 24 25 .mpwmpmv Poucmswcmnxm com moogumz new mpewcmumz mom .¢.~ In pm umzmmmm om? 8m: :8 use 335$ :9. mo 5.588. 0239. 9: .80an ed :a pm umammmm mm: :9? o 8 _ o e: m; 9.. ad 83 332233 w o mm P o mo. NF.F mspa.o 2m-< Pamowm A m.~ oo_ 2 m.~ o.om mm.o as 0.0 meccaaaFAXOLuaz o mm mm _ ~._m m.mm 0a.? ma P.P ewe m- mo .uaoaaz m AN mm _ a.m e.m o~.o as _.o_ amapapeau-mwpo< mawpw>wpu< umcwuaam a.mmmcmmoguxsmc mcwpoca cow mamcom covumowmmczauu.— m4m