[CANCER

RESEARCH

33, 1862-1865,

August

1973]

Isolation and Purification of L-Methionine-a-deamino--y mercaptomethane-Lyase (L-Methioninase) from Cbs tridiu m sporo genes'
Willi Kreis and Catherine Hession
Memorial Sloan-Kettering Cancer Center, New York, New York 10021

SUMMARY In an attempt irreversibly to deplete biological systems of the essential amino acid L-methionine, L-methionine a-dcam mo- -y-mercaptomethane-lyase was isolated from Clostridium sporogenes (ATCC 7955), highly purified, and characterized. All the investigations undertaken indicate a double function: release of the a-amino- and the y methanethiol group by a single protein unit. It is possible that this bifunctional activity is brought about by gene fusion. INTRODUCTION This laboratory reported that a methylhydrazin@ de rivative, I-methyl-2-(p-isopropylcarbamoyl) benzylhydra zinc hydrochloride, (NSC 77213), with clinically demon strated activity in Hodgkin's disease and lymphosarcoma, showed characteristic disturbance of the normal methyla tion of RNA, especially tRNA in P815 mouse neoplasms (8). This disturbance consisted of a significant in vivo methylation of RNA-guanine at position 7 and a consid erable reduction of the normal methylation of cytoplasmic RNA of these cells. The effect preceded the inhibition of RNA synthesis and the therapeutic activity of the com pound could be prevented by massive doses of L-methi onine administered 2 hr before the drug (7). If indeed the undermethylation of RNA (and possibly also DNA) con tributes to the inhibition of RNA synthesis, it was con cluded that by depletion of the source of methyl groups, i.e., L-methionine, this effect could be achieved more ef fectively. It was hoped that an L-methionine-degrading enzyme would further the studies in this direction, would help to elaborate on the biological significance and mechanism of DNA and RNA methylation in normal and neoplastic tissues, and might be beneficial in tumor chemotherapy. Clostridium sporogenes was used as the source of an enzyme that degrades 1-methionine to methanethiol, am monia, and ct-ketobutyric acid. Such an enzyme in crude preparations of C. sporogenes was first described by Wie sendanger and Nisman (21). We carried the isolation to
1 This work was supported in part by Grant CA 08748 from the

the point of high purity, as indicated the disc electrophoretic assay. MATERIALS AND METhODS

by a single band in

C. sporogenes (ATCC 7955) was grown for 18 hr in a medium containing Bacto cooked meat medium, 0.35% (Difco Laboratories, Detroit, Mich.); Bacto tryptone, 1.0% (Difco); proteose peptone, 1.0% (Difco); sodium thioglycollate, 0. 1% (Matheson Coleman and Bell, East Rutherford, N. J.); and Ucon lubricant LB625, 0.01% (Union Carbide Corp., Pearl River, N. Y.). Batches of 4 g of cells were suspended in 3.5 ml of 0. 1 M sodium: potassium phosphate buffer, pH 7.4, containing 0.5 mM neutralized 2 (Fluka, Buchs, Switzerland) and 5 mM mercaptoethanol (NKPPM buffer) (Sigma Chemical Co., St. Louis, Mo.) were treated in an ultrasonic disinte grator, Model 60W (Measuring and Scientific Equipment Ltd., London, England) 3 times for 3 mm over 15 mm at 20 kc/sec. The cell suspension was kept in an alcohol: ice mixture throughout the sonic disruptions. All extrac tion steps were performed at 04. disrupted suspen The sion was centrifuged at 30,000 x g for 30 mm, the pre cipitate was washed 2 times with NKPPM buffer, and the supernatant and the 2 washings of the precipitate were then combined. Protamine sulfate (Nutritional Bio chemicals Corp., Cleveland, Ohio), 0.5% in NKPPM buffer, was added to a final concentration of 0.25% and the precipitate was discarded after centrifugation. Am monium sulfate was added to 42% saturation and the en zyme-inactive material was removed by filtration through glass wool. Additional ammonium sulfate (to 75% satura tion) precipitated all enzyme-active material. After cen trifugation the precipitate was dissolved in and dialyzed against NKPPM buffer. Two successive gel filtrations on a Sephadex G-200 (Pharmacia Fine Chemicals, Inc., Piscataway, N. J.) column (50 x 3.0 cm), conditioned and cluted with NKPPM buffer, were then performed. One enzyme-active protein peak was observed. Following filtration through the Sephadex G-200 column, the en zyme-active fractions were pooled, precipitated with
The abbreviations used are: PPH, pyridoxal phosphate; NKPPM, sodium : potassium phosphate buffer, pH 7.4, containing 0.5 m@i neu tralized PPH and 5 mM mercaptoethanol.

National Cancer Institute. Received December 27, 1972; accepted

April 18, 1973.

1862

CANCER RESEARCH

VOL. 33

L-MethiOninase ammonium sulfate (5.2 g/lO ml), and dissolved in and dialyzed against NKPPM buffer. This pooled sample was filtered a 2nd time through the Sephadex G-200 column. The enzyme-active fractions from the 2nd Seph adex G-200 were pooled, and the precipitation and dialysis steps were repeated. The resulting enzyme-ac tive sample was applied to a DEAE-Sephadex A-SO (Pharmacia) column (45 x 3. 1 cm). The proteins were eluted with a linear gradient of 0 to 0.4 M KCI in NKPPM without PPH. The 1st of 2 major peaks contained all the enzyme activity. After precipitation with ammonium sul fate (5.2 g/ 10 ml), the protein was dissolved in I mM p0tassium phosphate buffer, pH 6.5, containing 5 mM mer captoethanol. Final purification was achieved on a hy droxylapatite (Bio-Gel HI) (Bio-Rad Laboratories, Rich mond, Calif.) column (9 x 3 cm), using a linear gra dient of I mM to 0.2 M phosphate buffer, pH 6.5, which contained 5 mM 2-mercaptoethanol. For protection of the enzyme activity in the last 2 columns, 2 ml of 0.5 M phosphate buffer containing 2.5 mM PPH were added to all the tubes used for the collection of lO-ml fractions. When methanethiol and ammonia were assayed si multaneously in each individual fraction of the first and last column, there was good qualtitative and quanti tative coincidence. Acceptable coincidence was also found for the ratio of methanethiol to ammonia in all the combined fractions during the isolation procedure (Table I). The overall purification was 120-fold (Table I). About 98% of the enzyme activity was lost during the process. Disc electrophoretic monitoring of the pro teins present at several steps of the extraction procedure gave evidence of the stepwise purification (Chart I). The enzymatically active fractions of the last column (Step 6) produced only I band in the disc electrophoresis. For the assay of methanethiol and ammonia, 50 of L-methionine-methyl- @C .33 zmoles; specific activ (1 ity, 0.0379 mCi/mmole; Amersham/Scarle, Arlington Heights, Ill.); 50 j.zl of sodium PPH (0.472 ,umole); SO to 90 @l f sodium : potassium phosphate buffer, pH 7.4 o (0.5 M); and 10 to 50 zl of enzyme were mixed in pre cooled Pyrex ignition tubes (10 x 70 mm) each containing 1 small glass bead. Each tube was then plugged in dividually with a fluted paper disc (Whatman No. 1, 26-mm diameter) which was moistened with 2 drops of a saturated mercuric chloride solution. The tubes were incubated in a shaking water bath for 15 mm at 37,and the reaction was stopped by cooling in ice. For the comple tion of the release of methanethiol-'4C from the reaction mixture, the tubes were heated to boiling over a Bunsen burner. The paper discs were then removed, and their radioactivity was determined in 10 ml Diotol (5) in a Packard Model 3380 scintillation spectrophotometer. For the ammonia determination 250 @zl 10% trichloroacetic of acid were added after incubation of the tube. Conway dishes and Nessler's reagent were used for trapping and evaluating the ammonia. One unit of enzyme is defined as the amount of enzyme necessary to release from the substrate I @tmoleeach of methanethiol-'4C and/or NH3 per mm. Methanethiol was identified as its mercury salt. Radioactive methanethiol released from L-methionine la beled with either 35S or 4C the methyl group (Amer in sham/Searle) showed good quantitative agreement (33.6 and 33.8 nmoles). No radioactive material was trapped when the substrate was labeled with 4C positions I or in 1, 2, 3, and 4 of L-methionine (Amersham/Searle). The 3rd reaction product, a-ketobutyric acid, was identified as the 2 , 4-dinitrophenylhydrazonc derivative by paper [Whatman No. 1, system, 1-butanol :ethanol : water (4 : I : 5)] and thin-layer chromatography (silica gel, system, 75% phenol) and by its melting point and melting point of mixture with the derivative of authentic a-keto butyric acid. Minor amounts of other reaction products have not yet been analyzed. Methanethiol in the incubation medium was identified by a color reaction for mercaptans ( I 5) which consisted of the addition of 0. 1 ml of sodium nitroprussiate solu tion [1.5 g sodium nitrosopentacyanofcrratc(III) dis solved in 5 ml of 2 N HC1 and followed by the addition of 95 ml methanol and 10 ml concentrated ammonia] to the reaction product after incubation as above and immediate spectrophotometric determination of the developed purple color at 530 nm. The test is, at best, semiquan titative. Disc electrophoresis was performed on polyacrylamide gel at 5according to the methods of Ornstein (13) and Davis (4). Reagents and conditions were as described earlier (9). The molecular weight, as determined by gel filtration on Sephadex G-200 and by ultracentrifugation and sedi mentation analysis with Rayleigh interferometry (16), was about 150,000. Activity of the enzyme was optimal in the pH range of 7_S to 8.5 for both functions and absent below pH 6.0. The Km values under the standard conditions of the enzyme assay were similar for the 2 functions: 90 mM for the re lease of methanethiol and 78 mM for the release of am monia. Maximal enzyme activity was dependent upon the presence of PPH. The approximate Km value for the latter was 10 @M PPH when measured in the methane thiol assay, indicating that L-methionine is bound loosely and the cofactor more strongly to the enzyme. Dialysis of the enzyme against 10 mM hydroxylaminc in NKPPM buffer decreased the activity exponentially to 5% within 3 hr and to less than 2% within 18 hr. Dialysis against plain buffer reduced the enzyme activity within the same time intervals to SO and 5%, respectively. No reactivation was observed after adding PPH. The isoelectric point de termined by electrofocusing on semipurified preparations (LKB Instruments, Inc., Rockville, Md.) was about pH 4.2. The stability of a semipurified enzyme preparation (after step 4) kept frozen between analyses showed a steady loss of activity of about 1.5%/day over a span of 43 days. Pure enzyme preparations (after Step 6) lost 85% of their initial activity within 14 days. The enzyme activity (methanethiol test) increased slightly when heated for IS mm at 50or for 10 mm at 60with rapid loss of activity after further heating at 60.Within 1 mm there was 90%

AUGUST

1973

I 863

Willi Kreis and Catherine Hession

143Sephadex sulfate precipitation3 . 848 filtrationlstcolumn6.724.00.9342ndcolumn10.613.20.955DEAE-Sephadex G-200 Gel

22.67.41.09matography6Hydroxylapatite column chro 12O.0'@2.11.13matography column chro

a Evaluation
b In some

by the assay
steps during the

of methanethiol.
extraction procedure, dialysis for removal of ammonium sulfate had

to be restricted
C At this point

for loss of enzyme activity.

several experiments. Table 2 Enzyme activity on different substrates inthe For conditions for enzyme assay see text. All substrates were used saturatedsolution same concentration (1 .33 zmoles) except 1-cystine, where a was used. Purified enzyme was used for these assays. a-Keto-y mercaptomethylbutyric acid was synthesized from i-methionine-methyl C according to the method of Meister (II).N reactionreleased , mercaptoSubstrate compounds1-Methionine +D-Methionine 0DL-Methionine 0L-MethiOnine sulfoxide 0a-Keto-y-mercaptomethylsulfone 0butyric acida-Aminobutyric 0L-Ethionine acid 81.0DL-Homocysteine 206.01-Cysteine 01-Cystine (nmoles) 29.4 0 17.6 14.7 0 Color for

38 . 28 .0
was noticed
both

0
the release
color

a A positive

color

reaction
and

indicating
gave a

of mer
reaction

captoethanol.
b DL-Homocysteine 1-cysteine positive

by themselves. Chart I . Disc electrophoresis of protein at different purification steps. Disc I, supernatant after sonic disruption and centrifugation of cells (105
14 of protein on gel); Disc 2, combined enzyme-active fractions after

DISCUSSION The data indicate the occurrence of a single protein with dual enzymatic functions. All the simultaneously per formed evaluations of methanethiol and ammonia of the crude extracts and the individual fractions from all column chromatograms up to the final purified product revealed close quantitative ratios within the limits of the assay procedure used. Also Km values, loss of enzyme activity over 6 weeks, and pH optima indicated dual functions of a single enzyme. Such a phenomenon has been ob served previously. An enzyme has been described (18) that degrades D-mcthionine but not L-methionine into

2nd G-200 column [54 zg of protein on gel (Table I Step 4)]. Disc 3, combined enzyme-active fractions after DEAE-Sephadex column [26 @g of protein on gel (Table I, Step 5)]; Disc 4, combined enzyme-active fractions after hydroxylapatite column [37 @.tgf protein on gel (Table I, o Step 6,)]. Cathode at top: I left, 4 right. For procedure see text.

inactivation at 70. The purified enzyme did not act cx clusively on L-methionine but, as demonstrated in Table 2, deaminated DL-methionine sulfoxide, L-methionine sul fone, L-ethionine, L-cysteine, L-cystine, and DL-homo cysteine.

1864

CANCER RESEARCH

VOL. 33

L-MethiOfliflase similar products. Rechler and Bruni (14) reported that an enzyme in Salmonella typhimurium had a histidinol de hydrogenase and imidazolyl acetol phosphate amino transferase activity in the same protein molecule. Similar observations were reported by Yourno et a!. (22). This dual-function enzyme was thought to be the product of gene fusion resulting in the covalent linkage of the de hydrogenase and the aminotransferase (14). Currently, gene fusions are of wide interest (2). In our case, it is likely that this enzyme represents another example of the same phenomenon. Less likely is the 2-step mechanism for the double ef fect of the enzyme. The deamination of L-methionine leads to ct-keto- -y-mercaptomethylbutyric acid. This product was synthesized according to the method of Meis ter (I 1). Under the conditions of our assay method, it did not show any spontaneous decomposition or release of methanethiol, which was contrary to the findings of Waelsch and Borek (20) for far more vigorous treatment. Furthermore, the commercially available Crotalus ada manteus ct-amino acid oxidase (Sigma), when reacted under the identical conditions of our enzyme prepara tion, showed deamination of L-methionine exclusively. The lack of specificity of the reported purified enzyme (Table 2) is expressed in the release of ammonia also from DL-homocysteine, L-cysteine, L-cysteine, L-ethionine,
DL-methionine sulfoxide, and L-methionine sulfone,

REFERENCES
1. Binkley, F. A Note on the Specificity of the Enzymatic Cleavage of Thio-esters. J. Biol. Chem., 192: 20921 1951. 1, 2. Biochemical Society Symposium. Two Genes, One Polypeptide Chain. Federation Proc., 31: 176209,1972. 3. Chen, S. S., Walgate, J. H., and Duerre, J. A. Oxidative Deamina tion of Sulfur Amino Acids by Bacterial and Snake Venom L-Amino Acid Oxidase. Arch. Biochem. Biophys., 146: 5463,1971. 4. Davis, B. J. Disc Electrophoresis II. Method and Application to Human Serum Proteins. Ann. N. Y. Acad. Sci., 121: 404-427, 1964. 5. Herberg, R. J. Determination ofCarbon-l4 and Tritium in Blood and Other Whole Tissues. Liquid Scintillation Counting of Tissues. Anal. Chem., 32: 42-46, 1960. 6. Kallio, R. E., and Larson, A. D. Methionine Degradation by a Species of Pseudomonas. In: W. D. McElroy and H. B. Glass (eds.), A Symposium on Amino Acid Metabolism, pp. 616-631. Baltimore: Johns Hopkins Press, 1955. 7. Kreis, W. Mechanism of Action of Procarbazine. In: S. K. Carter (ed), Proceedings of the Chemotherapy Conference on Procarba zinc, pp. 3544.Bethesda; National Cancer Institute, 1970. 8. Kreis, W. Metabolism of an Antineoplastic Methyl-hydrazine Dc rivative in a P815 Mouse Neoplasm. Cancer Res., 30: 8289, 1970. 9. Kreis, W., Drahovsky, D., and Borberg, H. Characterization of Protein and DNA in P815 Cells Sensitive and Resistant to l-@9-DArabinofuranosylcytosine. Cancer Res., 32: 696-701, 1972. 10. Kreis, W., and Hession, C. Biological Effects of Enzymatic Depriva tion of L-MethiOnine in Cell Culture and an Experimental Tumor. Cancer Res., 33: 18661869, 1973. 11. Meister, A. Enzymatic Preparation of a-Keto Acids. J. Biol. Chem., 197: 309317, 952. 1 12. Onitake, J. On the Formation of Methylmercaptan from i-Cystine and 1-Methionine by Bacteria. J. Osaka Med. Assoc., 37: 263270, 1938. 13. Ornstein, L. Disc Electrophoresis I. Background and Theory. Ann. N. Y. Acad. Sci., 121: 321-349, 1964. 14. Rechler, M. M., and Bruni, C. B. Properties of a Fused Protein Formed by Genetic Manipulations. J. Biol. Chem., 246: 18061813, 1971. 15. Reid, E. E. Organic Chemistry of Bivalent Sulfur, p. 158. New York: Chemical Publishing Corp., 1958. 16. Richards, E. G., and Schachman, H. K. Ultracentrifuge Studies with Rayleigh Interference Optics. I. General Applications. J. Phys.Chem., 63: 1578-1591,1959. 17. Ruiz-Herrera, J:, and Starkey, R. L. Dissimilation of Methionine by Fungi. J. Bacteriol., 99: 544-551, 1969. 18. Ruiz-Herrera, J., and Starkey, R. L. Dissimilation of Methionine by a Demethiolase of Aspergillus Species. J. Bacteriol., 99: 764-770, 1969. 19. Segal, W., and Starkey, R. L. Microbial Decomposition of Methio nine and Identity of the Resulting Sulfur Products. J. Bacteriol., 98: 908-913, 1969. 20. Waelsch, H., and Borek, E. The Stability of the Keto Acid from Methionine. J. Am. Chem. Soc., 61: 2252, 1939. 21. Wiesendanger, S., and Nisman, B. La L-Methionine De-mercapto desaminase: un Nouvel Enzyme A Pyridoxal-phosphate. Compt. Rend, 237: 764-765, 1953. 22. Yourno, J., Kohno, T., and Roth, J. R. Enzyme Evolution: Genera tion of a Bifunctional Enzyme by Fusion of Adjacent Genes. Na ture, 228: 820-824, 1970.

whereas D-methionine and a-aminobutyric acid did not act as substrates for the deaminase function of the enzyme. L-Ethionine did act as substrate for both deamination and demercaptoethylation. Other microorganisms have been reported to catabolize methionine such as Proteus rettgeri (3), Escherichia co/i (12), Pseudomonas (6), and various fungi (17) by de amination and/or demercaptomethylation. Not all of these microorganisms release methanethiol from methio nine (19) or a-keto-7-mercaptomethylbutyric acid. Equi molar release of ammonia and methanethiol from D methionine was reported only with the purified fungal enzyme (17). Rat liver tissue has also been found to re lease methanethiol and ammonia from L-methionine (1). However, our experiments did not indicate the presence of an enzyme in homogenized rat liver that catalyzes these reactions. The enzyme described might be useful for studies in which a complete removal of 1-methionine is required. It might also be applied for the quantitative assay of L methionine in biological samples. The successful use of the enzyme for the inhibition of growth of a tissue culture and an experimental tumor has been described (10). ACKNOWLEDGM ENTS

The authors acknowledge the interest and help of Dr. Dorris J. Hutchi son, Dr. R. Barclay, Dr. J. Gill, and J. Begel and the skillful technical assistance of Patsy Stroble.

AUGUST

1973

1865