J. Biochem. 94, 1053-1059 (1983)

Purification and Some Properties of from Clostridium perfringens1

Satoshi SEKIGUCHI,- Sachiko SEKI , and Makoto ISHIMOTO

Department of Chemical Microbiology , Faculty of Pharmaceutical Sciences, Hokkaido University, Kita-ku, Sapporo , Hokkaido 060

Received for publication, April 11, 1983

Nitrite reductase from Clostridium perfringens was purified by chromatographies on DEAE-cellulose, DEAE-Sephadex, Sephadex G-150, and hydroxylapatite and by isoelectric focussing to a homogeneous state, showing essentially a single protein band in disc gel electrophoresis and a single immuno-precipitation line in double diffusion against antiserum obtained from immunized rabbits. The reductase was induced in the presence of nitrate. It had a molecular weight of 54,000 and showed no absorption peak in the visible region. The pH optimum was 6.2 and Km for nitrite was 5 mm. Ferredoxin, as well as viologen dyes, was found to be an electron donor. The of nitrite reduction was hydroxylamine. This reductase was inhibited by o-phenanthroline and azide but not by cyanide or diethyldithiocar bamate.

Nitrate reduction in bacteria is carried out by of bacteria (6), fungi (7), and higher plants (8) molybdenum-containing reductase to produce ni reduce nitrate to the level of ammonia and contain trite, but further reduction of nitrite leads to the nitrite reductase which produces ammonia directly formation of nitric oxide, ammonia, or other from nitrite [EC 1.6.6.4, 1.7.7.1]. These products by different sorts of enzymes (1). In are siroheme proteins with iron-sulfur centers and denitrifying bacteria, two classes of nitrite reduc flavins, and use reduced pyridine nucleotide as an tase have been described, one containing c- and electron donor. d-type hemes [EC 1.9.3.2] (2, 3) and the other A strictly anaerobic bacterium, Clostridium copper [EC 1.7.2.1, 1.7.99.3] (4, 5), both of which perfringens, reduces nitrate to ammonia, promot form nitric oxide. On the other hand, a variety ing the oxidation of organic substrates (9). Ni trate reductase from this organism was purified and ferredoxin was found to play a role as an 1 This study was supported in part by a Grant-in-Aid electron donor (10). for Scientific Research from the Ministry of Education, The present paper describes the purification Science and Culture of Japan and by a grant from and characterization of nitrite reductase from C. Yakult Institute. 2 Present address: Department of Agricultural Chemis perfringens; this uses ferredoxin as an electron donor and forms hydroxylamine from try, Faculty of Agriculture, The University of Tokyo, nitrite as the main product. Yayoi, Bunkyo-ku, Tokyo 113.

Vol. 94, No. 4, 1983 1053 1054 S. SEKIGUCHI, S. SEKI, and M. ISHIMOTO

x g for 3 h was subjected to chromatography on

EXPERIMENTAL PROCEDURES a DEAE-cellulose column (3.2 x 38 cm) previously equilibrated with TM buffer. The column was Organism and Cultivation-Clostridium per washed with TM buffer containing 0.1 M NaC1 and fringens, strain HM-l, isolated by Prof. H. Iida, then nitrite reductase was eluted with TM buffer was used throughout this work. containing 0.2 M NaCl. Culture medium contained, per liter; Poly Active fractions were pooled, diluted with pepton (Daigo Eiyo), 7.5 g; yeast extract (Daigo 200 ml of TM buffer, and adsorbed on a DEAE- Eiyo), 5.0 g; glucose, 2.5 g; thioglycolic acid, 0.5 Sephadex A-50 column (2.0 x 30 cm) equilibrated ml; and KNO,, 0.02 g. The pH was adjusted to with TM buffer containing 0. 15 M NaCl. Protein 7 by adding NaOH. This medium was inoculated was eluted with 300 ml of a linear gradient system with 0.4% of subculture and incubated without of 0.15-0.3 M NaCl in TM buffer. agitation in deep layer culture at 37°C. After a The active fractions thus eluted were pooled, 10-11 h incubation, cells in the early stationary concentrated to 6 ml by ultrafiltration with a Toyo phase were harvested by continuous centrifugation, membrane filter UK-10, and passed through a washed twice with 20 mm Tris-HCl buffer, pH 7.2, Sephadex G-150 column (2.8 x 88 cm) equilibrated containing 1 mm 2-mercaptoethanol (TM buffer) with TM buffer-0.2 M NaCl. and stored at - 80°C. Usually 500 g of wet cells Active fractions were then adsorbed on a were obtained from a 60 liters culture. hydroxylapatite column (1.2 x 9 cm) equilibrated Preparation of Crude Extract-The bacterial with 50 mm potassium phosphate buffer, pH 7.3, cells were suspended in TM buffer and disrupted containing 1 mm 2-mercaptoethanol. The column by sonic treatment in an Insonator (Kubota, was washed with the same buffer and then nitrite 200M) at 20 kHz at 180 W for 20 min. The reductase was eluted with 50 ml of a linear con supernatant after centrifugation at 17,000 x g was centration gradient system of 50-300 mm phos used as crude extract. phate buffer containing 1 mm 2-mercaptoethanol. Assay of Nitrite Reductase-The assay of The fractions with the highest activity contained nitrite reductase was run in open test tubes with proteins that could not be removed in the next chemically reduced methyl viologen as an electron purification step, so those fractions were not used donor. The reaction mixture contained, in 1 ml; for further purification; the fractions eluted before

0.2 M potassium phosphate buffer, pH 7.2, or Tris and after those fractions were pooled and dialyzed

- HCl buffer, pH 6.9, 1 or 2 mm NaNO2, 6 mm against TM buffer. The dialyzed preparation was methyl viologen, 5 mm Na2S2O,, 10 mm KHCO3, subjected to isoelectric focussing on a 110 ml and enzyme. The reaction was started by adding column containing Pharmalyte giving a pH gra

Na2S2O4-KHCO3. After 3 or 5 min of incuba dient of 4.5-6.0, and was electrophoresed for 47 h tion at 37°C the reaction was stopped by vigorous at a constant voltage of 1,000 V at 4°C. Enzyme

shaking to oxidize reduced methyl viologen and fractions that gave a single protein band in disc the residual nitrite was determined colorimetrically gel electrophoresis were pooled. by the method of Nicholas and Nason (11). Dis- Other Assays-Hydroxylamine was deter-

appearance of nitrite in the absence of enzyme or mined by Csaky's method (12). Ammonia was

in the presence of boiled enzyme was also mea assayed by Chaney and Marbach's method (13).

sured as a control. The activity was obtained Protein was determined by the method of Lowry

from the difference between the amount of nitrite et at. (14) or of Bradford (15) with bovine serum decrease in the presence of enzyme and that in the albumin as a standard.

absence of enzyme or in the presence of boiled Molecular Weight Estimation-Molecular

enzyme. One unit of nitrite reductase was defined weight was estimated by gel filtration as described as the amount of enzyme which catalyzed the by Laurent and Killander (16) using a Sephadex

reduction of 1 ƒÊmol of nitrite per min. G-150 column (2.7 x 86 cm) equilibrated with TM Purification of Nitrite Reductase-All opera buffer containing 0.2 M NaC1. As markers, cyto tions were carried out at 4°C. A soluble fraction chrome c (molecular weight 12,500), chymotryp isolated by centrifuging the crude extract at 63,000 sinogen A (25,000), ovalbumin (45,000), and bo-

J. Biochem. NITRITE REDUCTASE FROM Clostridium perfringens 1055 vine serum albumin (68,000) were used. obtained from Pharmacia, DEAE-cellulose from Polyacrylamide Gel Electrophoresis-Disc Brown, and hydroxylapatite from Bio Rad (HTP). polyacrylamide get electrophoresis was performed by the method of Williams and Reisfeld (17) using RESULTS 7.5 % polyacrylamide gel (5 x 75 mm) at 2 mA/ tube at 4°C. Protein was stained with Coomassie Induction of Nitrite Reductase in C. perfringens brilliant blue G250 in perchloric acid according -When C. perfringens was grown in a medium to the procedure described by Reisner et al. (18). containing 2 mm or 20 mm nitrate, nitrite accumu Nitrite reductase activity in gels was detected by lated at a concentration of 1 mm at the end of activity staining as described by Vega and Kamin exponential growth and then decreased. While (8). activity reached its maximum at Preparation of Antibody and Immunological the time of maximal accumulation of nitrite and Tests-Antiserum against nitrite reductase was then decreased sharply, nitrite reductase had its raised in rabbits by subcutaneous injection of 0.22 maximum after the peak of nitrite concentration mg of purified enzyme per rabbit, in emulsion and decreased only partly. Since nitrite reductase containing 50% Freund's complete adjuvant. Im activity was 7-fold higher in the presence of nitrate munization was repeated twice with intervals of added to the medium, cells grown in the medium two and three weeks, A week after the last containing nitrate were used for the preparation treatment antisera were taken. A y-globulin frac of nitrite reductase. tion precipitated at 40% saturation of ammonium Purification of Nitrite Reductase-Nitrite re sulfate was used as the antibody preparation. ductase causing a decrease of nitrite in the presence Double immuno-diffusion was carried out by of dithionite and methyl viologen was extracted the method of Ouchterlony (19). For inhibition from cells by sonic treatment. This enzyme was tests, 0.2 ml of the supernatant at 63,000 x g from soluble, and was purified by DEAE-cellulose, crude extract and 0.8 ml of phosphate buffer- DEAE-Sephadex A-50, Sephadex G-150, and hy saline containing various amounts of antibodies droxylapatite chromatographies and by isoelectric were mixed and incubated at 4°C for 16 h. The focussing. The purification process is summarized precipitate formed was centrifuged down and en in Table I. Nitrite reductase was purified 18-fold zymatic activity in the supernatant was measured. from the supernatant at 63,000 x g, and the yield Materials-Ferredoxin from C. perfringens was 1 %. The specific activity of the final prepa was prepared according to the purification pro ration was about 50-fold higher than that of crude cedure of Seki et al. (20). extract in another experiment. The rather poor Trizma base, bovine serum albumin, FAD, yield was due to the omission of the fractions with lower specific activity in each step of purifica NAD•{, and NADP- were purchased from Sigma. tion for the following step. There were no other Methyl viologen was a product of British Drug House. DEAE-Sephadex A-50, Sephadex G-150, peaks of activity in any chromatography step. The molecular activity of the enzyme was 300 per Pharmalyte, and molecular weight markers were

TABLE I. Summary of purification of nitrite reductase from C. perfringens.

Vol. 94, No. 4, 1983 1056 S. SEKIGUCHI, S. SEKI, and M. ISHIMOTO min. We did not use the optimal conditions to a spur in double diffusion against antibodies to avoid non-enzymatic reduction of nitrite as de the purified enzyme. Occasionally another very scribed below. Under the optimal conditions, the faint precipitation line was also observed with the molecular activity of this enzyme would be at crude extract, but never with the purified enzyme; least 5,000 per min. the enzyme preparation seems to be almost pure. Purity-To remove Pharmalyte, we dialyzed Activity of nitrite reduction was inhibited by the purified nitrite reductase against 0.85% NaCl the antibodies not only in the purified preparation for two days at 4°C and subjected it to disc elec but also in the crude extract (data not shown). trophoresis. As shown in Fig. 1, the preparation The enzyme appears to be the main nitrite-reducing gave a major protein band and a faint minor one. enzyme in the extract, if not the only one. Activity staining showed that both bands had Properties-The molecular weight of nitrite nitrite reductase activity. A protein band at the reductase was estimated to be 54,000 by gel filtra position of bromophenol blue seems to be due to tion through a Sephadex G-150 column. The Pharmalyte. absorption spectrum of partially purified nitrite An activity band with the same mobility as reductase (obtained by hydroxylapatite chromatog that of purified enzyme was always found in crude raphy) is shown in Fig. 3. There was a peak at extracts from different cultures. Although weak 280 nm but no absorption maxima were found in bands with different relative mobilities were some- times observed, the enzyme as purified is probably the main nitrite reductase in the cell extract. For further examination of the purity of the preparation, Ouchterlony immuno-diffusion anal ysis was carried out (Fig. 2). The purified enzyme and the soluble fraction of cell-free extract showed a single precipitation line fusing together without

Fig. 2. Immunodiffusion of nitrite reductase and crude extract from C. perfringens. The center well con tained antibodies against purified nitrite reductase and wells A, purified enzyme (1.6 munit); B, crude extract (2.3 munits); C, pre-immune serum; and D, phos phate-buffered saline.

Fig. 1. Polyacrylamide gel disc electrophoresis of purified nitrite reductase. The purified enzyme (5.5 ƒÊg protein) was electrophoresed on 7.5% polyacrylamide gel at pH 8.0. The experimental details are shown in the text. A, protein staining and B, activity staining.

The arrows indicate the position of a marker, bromo Fig. 3. Absorption spectrum of nitrite reductase. phenol blue. Hydroxylapatite fraction was used.

J. Biochem. NITRITE REDUCTASE FROM Clostridium perfringens 1057 the visible region. The effect of pH on the en The effect of nitrite concentration on nitrite zyme activity was examined with reduced methyl reduction was tested. Nitrite decrease was mea viologen as an electron donor (Fig . 4). The opti sured by using the standard assay method with mal pH was 6.2. However , nitrite was chemically different concentrations of nitrite. As shown in reduced below pH 7 in the absence of enzyme . Fig. 5, non-enzymatic reduction of nitrite increased Therefore, standard assay was performed in potas as the nitrite concentration was increased. The sium phosphate buffer, pH 7.2, though the enzyme rate of the enzymatic reaction was obtained from is not fully active at this pH. the difference of nitrite decrease in the presence and absence of enzyme. The apparent Km was estimated to be 5 mm. To eliminate the non enzymatic nitrite disappearance, we used 1 or 2 mm nitrite in the standard assay method. In order to test the availability of pigments as electron donors, methyl viologen in the assay mixture was replaced by various electron transfer cofactors and nitrite decrease was measured in the presence of the purified nitrite reductase (Table II). It was found that ferredoxin from C. perfringens as well as methyl viologen could mediate the elec tron transfer from dithionite to the purified nitrite reductase. However, NAD+, NADP+, or FAD had no activity. C. perfringens is known to reduce nitrate to Fig. 4. Effect of pH on the activity of nitrite reductase. The solid line and open symbols indicate nitrite de- ammonia (21). When crude extract was used to crease per min due to nitrite reductase; the dotted reduce nitrite in the presence of methyl viologen line and closed symbols indicate that due to chemical and dithionite, ammonia was produced, but the reaction. Buffers used were sodium maleate (triangles), purified nitrite reductase gave no ammonia. Mass potassium phosphate (circles), and Tris-HCl (squares). spectrometry showed small amounts of gaseous products including nitric oxide, nitrous oxide, and nitrogen (date not shown). Thus, the reaction mixture with purified enzyme was analyzed for hydroxylamine.

TABLE II. Utilization of various electron carriers as electron donors for nitrite reductase. The reaction mix ture for enzyme assay was used, but methyl viologen was replaced by the electron carrier indicated. Activity as an electron donor is given as the amount of nitrite de- crease per min.

Fig. 5. Effect of nitrite concentration on nitrite re ductase activity. Decreases in nitrite in the presence

(1, -• -) and absence (2, -•¢-) of nitrite reduc tase are shown, as well as the difference between them

(1-2,-*-).

Vol. 94, No. 4, 1983 1058 S. SEKIGUCHI , S. SEKI, and M. ISHIMOTO

As shown in Table III, hydroxylamine was When ferredoxin was used as an electron found; the amount was almost the same as that donor in place of methyl viologen, hydroxylamine of nitrite reduced. Hydroxylamine appears to be was also found to be produced (data not shown). the main product of the nitrite reduction , if not The reaction mixture containing various rea the sole one. gents at 1 mm was preincubated for 10 min at 37°C and the reaction was started. As shown in

TABLE III. Formation of hydroxylamine in nitrite Table IV, nitrite reduction was inhibited by o- reduction by nitrite reductase. The composition of the phenanthroline and azide, while cyanide and di reaction mixture is indicated in " EXPERIMENTAL ethyldithiocarbamate had only a slight effect.

PROCEDURE." Nitrite used was 2 ƒÊmol .

DISCUSSION

In this paper, we show that nitrite reductase from C. perfringens can use ferredoxin as an elec tron carrier. Nitrate reductase of this bacterium is also linked to ferredoxin (20). An electron transfer system from organic substrates to nitrate through dehydrogenases, NAD+, NAD-ferredoxin TABLE ‡W. Effect of some reagents on nitrite re reductase, ferredoxin, and nitrate reductase was ductase. suggested (22). In nitrite reduction a similar sys- tem with nitrite reductase in place of nitrate re ductase seems to work for the oxidation of organic substances. Ferredoxin also functions in nitrite reduction in photosynthetic organisms: higher plants and algae. As shown in Table V, which lists several sorts of nitrite reductases from different sources , these reductases as well as NAD(P)H-dependent nitrite reductases are assimilatory, contain siro heme, and form ammonia in a 6-electron reduction . a Concentration of inhibitor was 1.0 mm. The absorption spectrum of the purified enzyme

TABLE V. Nitrite reductase in various organisms (Fd , ferredoxin; Cyt, cytochrome).

J. Biochem. NITRITE REDUCTASE FROM Clostridium perfringens 1059 of C. perfringens does not show the presence of 4. Kakutani, T., Watanabe, H., Arima, K., & Beppu, heme; this enzyme also differs from nitrite reduc T. (1981) J. Biochern. 89, 453-461 tases containing c- or d-type heme. It does not 5. Iwasaki, H. & Matsubara, T. (1972) J. Biochern. seem to contain copper as an active center, since 71, 645-652 6. Jackson, R.H., Bowden, A.C., & Cole, J.A. (1981) the activity was not inhibited by diethyldithiocar Biochem. J. 193, 861-867 bamate, a strong inhibitor of copper-containing 7. Lafferty, M.R. & Garrett, R.H. (1974) J. Biol. nitrite reductases. Although structural charac Chem. 249, 7555-7567 terization of the active center remains incomplete, 8. Vega, J.M. & Kamin, H. (1977) J. Biol. Chem. 252, C. perfringens nitrite reductase is probably a new 896-909 type of nitrite reductase. 9. Ishimoto, M., Umeyama, M., & Chiba, S. (1974) It was surprising that the product was hy Z. allg. Mikrobiol. 14, 115-121 droxylamine; this compound was found to be a 10. Seki-Chiba, S. & Ishimoto, M. (1977) J. Biochern. 82,1663-1671 product of nitrite reductase from Bacillus pumilus 11. Nicholas, D.J.D. & Nason, A. (1957) in Methods twenty-five years ago (26), but there was no fur in Enzymology (Colowick, S.P. & Kaplan, N.O., ther report. The assay method for hydroxylamine eds.) Vol. III, pp. 983-984, Academic Press, Inc., is based on the oxidation of this compound with New York iodine back to nitrite (12), so other partially 12. Csaky, T.Z. (1948) Acta Chem. Scand. 2, 450-454 oxidized nitrogen compounds might be products 13. Chaney, A.L. & Marbach, E.P. (1962) Clin. Chein. of nitrite reduction. Nitric oxide might also be 8, 130-132 such a product since nitric oxide reacts with re 14. Lowry, O.H., Rosebrough, N.J., Farr, A.L., & duced methyl viologen, producing hydroxylamine Randall, R.J. (1951) J. Biol. Chem. 193, 265-275 (27). Although this cannot be excluded, hydrox 15. Bradford, M.M. (1976) Anal. Biochern. 72, 248-254 16. Laurent, T.C. & Killander, J. (1963) J. Chromatogr. ylamine may be the real product, since living cells 14,317-330 as well as crude extract form ammonia in nitrate 17. Williams, D.E. & Reisfeld. R.A. (1964) Ann. N.Y. or nitrite reduction and there is no evidence of Acad. Sei. 121, 373-381 denitrification in this bacterium. The purified 18. Reisner, A.H., Nemes, P., & Buchotz, C. (1975) preparation had little activity to reduce hydrox Anal. Biochern. 64, 509-516 ylamine, but the crude extract showed the presence 19. Ouchterlony, O. (1948) Acta Pathol. Microbiol. of hydroxylamine reductase in activity staining of Scand. 25, 186-191 the gel after disc electrophoresis. The relative 20. Seki, S., Hagiwara, M., Kudo, K., & Ishimoto, M. mobility seems to be different from that of nitrite (1979) J. Biochem. 85, 833-838 reductase (Seki, S., unpublished data). Hydroxyl- 21. Wood, D.D. (1938) Biochem. J. 32,2000-2012 amine formed by nitrite reductase would be further 22. Chiba, S. & Ishimoto, M. (1973) J. Biochern. 73, reduced to ammonia in the cells by hydroxylamine 1315-1318 23. Zumft, W.G. (1972) Biochim. Biophys. Acta 276, reductase. 363-375 24. Prakash, O. & Sadana, J.C. (1972) Arch. Biochern. We thank Dr. H. Iwasaki for valuable advice and Mrs. Biophys. 148, 614-632 A. Yamazaki for mass-spectrometric measurement. 25. Liu, M.-C. & Peck, H.D., Jr. (1981) J. Biol. Chem. 256,13159-13164 REFERENCES 26. Taniguchi, S., Asano, A., lida, K., Kono, M., 1. Stouthamer, A.H. (1976) Adv. Microbiol. Physiol. Ohmachi, K., & Egami, F. (1958) in Proceedings of 14,315-375 the International Symposiwn on Enzyme Chemistry, 2. Yamanaka, T. (1964) Nature 204, 253-255 Tokyo and Kyoto (Ichihara, K., ed.) pp. 238-244, Maruzen, Tokyo 3. Iwasaki, H. & Matsubara, T. (1971) J. Biochein. 69,847-857 27. Iwasaki, H. (1983) J. Plant Cell Physiol. in press

Vol. 94, No. 4, 1983