Acetoin Fermentation by Citrate-Positive Lactococcus lactis subsp.

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Sept. 199, p. 2644-2649 99-224/9/92644-6$2./ Copyright C) 199, American Society for Microbiology Vol. 56, No. 9 Acetoin Fermentation by Citrate-Positive Lactococcus lactis subsp. lactis 322 Grown Aerobically in the Presence of Hemin or Cu2+ TSUTOMU KANEKO,* MASAHIRO TAKAHASHI, AND HIDEKI SUZUKI Central Research Institute, Meiji Milk Products Co., 1-21-3 Sakae-cho, Higashimurayama, Tokyo 189, Japan Received 3 April 199/Accepted 6 June 199 Citr+ Lactococcus lactis subsp. lactis 322 produced more biomass and converted most of the glucose substrate to diacetyl and acetoin when grown aerobically with hemin and Cu2". The activity of diacetyl synthase was greatly stimulated by the addition of hemin or Cu2+, and the activity of NAD-dependent diacetyl reductase was very high. Hemin did not affect the activities of NADH oxidase and lactate dehydrogenase. These results indicated that the pyruvate formed via glycolysis would be rapidly converted to diacetyl and that the diacetyl would then be converted to acetoin by the NAD-dependent diacetyl reductase to reoxidize NADH when the cells were grown aerobically with hemin or Cu2+. On the other hand, the YGIU value for the hemincontaining culture was lower than for the culture without hemin, because acetate production was repressed when an excess of glucose was present. However, in the presence of lipoic acid, an essential cofactor of the dihydrolipoamide acetyltransferase part of the pyruvate dehydrogenase complex, hemin or Cu2+ enhanced acetate production and then repressed diacetyl and acetoin production. The activity of diacetyl synthase was lowered by the addition of lipoic acid. These results indicate that hemin or Cu2+ stimulates acetyl coenzyme A (acetyl-coa) formation from pyruvate and that lipoic acid inhibits the condensation of acetyl-coa with hydroxyethylthiamine PP;. In addition, it appears that acetyl-coa not used for diacetyl synthesis is converted to acetate. It has been shown that diacetyl and acetoin in lactic acid bacteria are not produced from carbohydrate unless an additional source or pyruvate, such as citrate, is present (6, 1, 19). This is because in homofermentative organisms, most of the pyruvate is converted to lactate by NADdependent lactate dehydrogenase to regenerate NAD. In addition, lactococci are generally considered to be facultative anaerobes, albeit with a preference for anaerobic conditions. However, we recently showed that citrate-positive (Citr+) Lactococcus lactis subsp. lactis 322 (Streptococcus lactis subsp. diacetylactis 322) cells grown aerobically could produce a considerable amount of diacetyl in MRS medium without citrate and that diacetyl production was further increased by the addition of Cu2+, which is an effective stimulator of diacetyl synthase activity (13, 15). These results indicated that under aerobic conditions, diacetyl synthase could be effectively induced and then the cells could produce diacetyl from pyruvate formed via the glycolytic pathway. Several workers (3-5, 29) also found that acetoin production was enhanced under aerobic conditions. However, it has not yet been demonstrated whether most of the glucose is converted to diacetyl and acetoin. To enhance the production of these flavor compounds instead of lactate, further stimulation of the activity of diacetyl formation from pyruvate (diacetyl synthase) is required. In addition, an alternative system to reoxidize NADH formed via glycolysis is required, because lactate production from pyruvate is repressed. In this study we found that hemin, a chloride of heme, could greatly stimulate the activity of diacetyl synthase in Citr+ L. lactis subsp. lactis 322 in the same manner as Cu2+ and that the activity of NAD-dependent diacetyl reductase in the cells was at a markedly high level. This * Corresponding author. 2644 paper reports the effects of hemin or Cu2" on acetoin fermentation by Citr' L. lactis subsp. lactis 322 cells. MATERIALS AND METHODS Microorganisms and growth conditions. All experiments were made with Citr+ L. lactis subsp. lactis 322 isolated from cream cheese. MRS medium (9) without triammonium citrate and containing.3% silicon (by weight) as a foam breaker was used as the basal medium. In some experiments, 1,uM hemin,.1 mm CUC12, 73 U of catalase per ml,.5 mm antimycin A, and 1. mm lipoic acid were added to the basal medium. The cells were grown in a Sakaguchi flask on a reciprocal shaker at 12 strokes per min at 3 C. Molar growth yields (YGIU values) were determined by dividing the dry weights of cells by the number of moles of glucose utilized. Growth was measured by determining the optical density at 58 nm. Bakers' yeast (Saccharomyces cerevisiae 84) cells were also prepared for the assay of cytochromes by incubation in a basal medium at 3 C for 24 h and by three washes with distilled water. Preparation of cell extract and enzyme assay. Preparation of cell extract and assays of the activities of NADH oxidase and diacetyl reductase (EC 1.1.1.5) were carried out as described previously (15). The activity of diacetyl formation from pyruvate (diacetyl synthase) was assayed as follows. The reaction mixture contained 2 mm sodium pyruvate,.2 mm thiamine PP1,.2 mm MgCl2,.1 M phosphate buffer (ph 6.), and.1 ml of cell extract in a total volume of 1. ml. After incubation at 3 C for 3 min, the amount of diacetyl produced in the reaction mixture was determined. One unit of diacetyl synthase activity was defined as the amount of enzyme which catalyzes the production of 1,umol of diacetyl from pyruvate per min at 3 C. Lactate dehydrogenase (EC 1.1.1.27) activity was measured as follows. The reaction mixture contained 5 mm sodium pyruvate, 1 mm NADH,.1 M phosphate buffer

VOL. 56, 199 (ph 6.), and.5 ml of cell extract in a total volume of 5.5 ml. After incubation at 3 C for 3 min, the reaction was stopped by heating in a boiling-water bath for 1 min, and the amount of lactate produced in the reaction mixture was determined by high-pressure liquid chromatography (see below). One unit of lactate dehydrogenase activity was defined as the amount of enzyme which catalyzes the production of 1,umol of lactate per min at 3 C. The protein concentration for the assay of specific activities was determined by the method of Lowry et al. (17) with bovine serum albumin as the standard. Hydrogen peroxide. Hydrogen peroxide was determined by the luminol chemiluminescence technique (14) on a Biocounter M-21 (Lumac, The Netherlands). A 1-,ul volume of.1 ptm luminol in.2 M glycine-naoh buffer (ph 9.9) was added to 1 pl of the sample. After 1,ul of 2 mm K3Fe(CN)6 solution had been added to the mixture as a catalyst, the generated light signal was integrated for 1 s. Relative light units were recorded digitally, and the H22 content of the sample was calculated from a standard curve. Assays of cytochromes. Citr+ L. lactis subsp. lactis 322 cells and Saccharomyces cerevisiae 84 cells (3 g [wet weight] each), which were washed three times with distilled water, were finally suspended in 7 ml of distilled water. The cells were analyzed for the presence of cytochromes by obtaining the absorption spectra from the cell suspension, using a spectrophotometer (model 624; Hitachi, Tokyo, Japan) equipped with opalescent plates, as described by Shibata et al. (24). Analyses. Glucose and end products such as lactate, acetate, formate, and ethanol were analyzed by an enzymatic analysis method with test kits from Boehringer GmbH, Mannheim, Federal Republic of Germany. Lactate in the reaction mixture for the assay of lactate dehydrogenase activity was determined by a high-pressure liquid chromatography system (model L-62; Hitachi) equipped with a refractometer (model RI SE-61; Shodex, Tokyo, Japan). The column used was a Polyspher OAKC (Cica-Merck, Tokyo, Japan) packed with a cation-exchange resin in the hydrogen ionic form and protected by a Polyspher OAKC guard column (Cica-Merck). The column was run at 35 C, and the flow rate of the mobile phase (.2 N H2SO4) was set at.4 ml/min. Diacetyl and acetoin were determined by headspace gas chromatography on a dual flame ionization detector chromatograph (Sigma 3; The Perkin-Elmer Corp., Norwalk, Conn.) with a steel column (1.8 m by 3 mm [outer diameter]) packed with Unicarbon A2, 6/8 mesh (12). A 1-g portion of the sample was transferred into a headspace vial (Perkin- Elmer), which was closed with a rubber septum and an aluminum cap. After equilibration at 8 C for 3 min, diacetyl and acetoin were analyzed under the following conditions. After the column temperature had been held at 7 C for 4 min, the temperature was programmed to increase at a rate of 3 C/min from 7 to 13 C. The nitrogen flow rate was 3 ml/min, and the temperature of the injection port and detector was set at 15C. RESULTS Effects of hemin on growth, glucose utilization, and end product formation. There are great differences in growth, ph, glucose utilization, and end product formation by Citr+ L. lactis subsp. lactis 322 aerobic cultures with and without hemin (Fig. 1). In a culture with hemin, 91.8% (89.2 mmol) of the glucose was consumed after a 24-h incubation. In a ACETOIN FERMENTATION BY L. LACTIS 2645 culture without hemin, however, the cells could utilize only 38.8% (37.6 mmol) of the glucose after a 48-h incubation. The growth rate and the production of diacetyl and acetion were also markedly affected by the addition of hemin. The cells in hemin-containing culture grew faster and produced more biomass than those in the culture without hemin (2.27 and.87 mg [dry weight] per ml, respectively, after a 15-h anaerobic incubation at 3 C), although the YG1U value of the former culture was lower than that of the latter (37.5 and 47.2 g [dry weight] per mol, respectively). (Fig. 1). Furthermore, the production of diacetyl and acetoin in the culture with hemin after a 48-h incubation was about 1-fold higher than in the culture without hemin (Table 1). On the other hand, lactate production per mole of glucose was inhibited in the culture with hemin, and, interestingly, the consumption of accumulated lactate and the production of acetate were very rapid after the disappearance of glucose from the medium (Fig. 1). In hemin-containing culture, acetate production was repressed when an excess of glucose was present. In addition, the increase in ph from 4.73 at 24 h of incubation to 5.31 at 48 h of incubation was accompanied by the consumption of lactate (Fig. 1; Table 1). In this study, formate and ethanol were not detected in either culture. Influence of additives on growth, glucose utilization, and end product formation. The influence of Cu2' and lipoic acid on glucose utilization and end product formation were studied when Citr+ L. lactis subsp. lactis 322 cells were grown aerobically (Tables 1 and 2). Cu2+ stimulated growth, glucose utilization, and diacetyl and acetoin production when the cells were grown aerobically in the medium with silicon used as a foam breaker. Growth, glucose utilization, and diacetyl and acetoin production were further increased when both Cu2+ and hemin were added to the basal medium. In this case, 99.8 mmol of glucose was converted to 1.19 mmol of diacetyl and 15.4 mmol of acetoin after 48 h of incubation. In contrast, lactate production in the culture with both Cu2' and hemin was lower than that in the culture with just one these additions. Furthermore, acetate production at 48 h of incubation in the culture with both Cu2+ and hemin was also lower than in the cutlure supplemented with only hemin. However, when lipoic acid was added to the medium with hemin or Cu2+, acetate production was markedly enhanced and diacetyl and acetoin production was repressed (Table 2). Lipoic acid also slightly repressed the glucose utilization and lactate production. On the other hand,.5 mm antimycin A did not reduce the stimulatory effect of hemin on growth, glucose utilization, and acetoin production. In addition, no increases in growth rate, glucose utilization, and end product formation (particularly diacetyl and acetoin) were observed after the addition of 73 U of catalase per ml (T. Kaneko, unpublished results). Production and reduction of H22 by cells. Citr+ L. lactis subsp. lactis 322 cells that had been washed three times in distilled water produced H22 when incubated aerobically at 35 C for 6 min (3,87 and 5,67,ug of H22 per liter before and after incubation, respectively). In this case, H22 accumulation was repressed by the addition of 1,uM hemin to the cell suspension (3,87 and 4,26 pug/liter before and after incubation, respectively). However, catalase activity was not observed in the cells grown with hemin. Cytochrome spectra. Cytochrome spectra of S. cerevisiae 84 cells and of Citr+ L. lactis subsp. lactis 322 cells grown aerobically with hemin were determined (Fig. 2). Peaks at 52, 548, 559, and 6 nm, presumably due to

2646 KANEKO ET AL. APPL. ENVIRON. MICROBIOL. 1Ir A B _ I 81- X 1.5 6 21-,-- ao a 1. o5 A--A --- 1 12 24 36 48 12 24 36 48 2.5 :..4)w 4 F. 4) *~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I= to OL O L ao ZO T e 1t5 3:. U I.'v - 4rt-T Or 6 Time (h) 5 2 O._. 21-5 5 ~~~~~ 4 16 4) N 4 F 3 12 "A 4) Go I 6 F 8-45 F IOO L 413 L 12 24 36 48 A N A-& iza A "' I o --A 12 24 36 48 2 1 4 4 Time (h) FIG. 1. (A and B) Growth of and diacetyl and acetoin production by Citr+ L. lactis subsp. lactis 322 grown aerobically at 3 C in a basal medium without (A) and with (B) 1,uM hemin. (C and D) Glucose utilization, ph change, and production of lactate and acetate by Citr+ L. lactis subsp. lactis 322 grown aerobically at 3 C in a basal medium without (C) and with (D) 1,uM hemin. cytochromes, were detected for S. cerevisiae 84 cells, but no trace of cytochrome peaks was recognized for Citr+ L. lactis subsp. lactis 322 cells grown aerobically with or without hemin. Effect of hemin on the activities of several enzymes. To investigate the reasons for the enhancement of diacetyl and acetoin production in the culture with hemin, we compared the activities of NADH oxidase, diacetyl synthase, and diacetyl reductase for the cells grown aerobically in the medium with and without hemin (Table 3). There was a great difference in diacetyl synthase activity under the two different conditions: diacetyl synthase activity in the cells grown with hemin was 4.7-fold higher than that in the cells grown without hemin. The diacetyl reductase activity in the former cells was slightly lower than that in the latter, although both the activities were high. No great difference in NADH oxidase activity was observed. The influence of hemin on the activities of several enzymes was also studied by using a cell extract prepared from the cells grown aerobically in a basal medium. The activities of NADH oxidase, diacetyl reductase, and lactate dehydrogenase were not influenced by hemin. However, diacetyl synthase activity was increased 3.5-fold by the addition of 5 p.m hemin to the reaction mixture. In contrast, lipoic acid lowered the activity of diacetyl synthase, regardless of the presence of hemin or Cu2" (Table 4). DISCUSSION The ability to utilize 2 as an electron acceptor and to produce H22 is widespread among lactic acid bacteria (2, 7, 8, 2). An inhibitory effect due to the formation of H22 has been indicated by decreases in growth rate and acid production under aerobic conditions (21). Since catalase containing

VOL. 56, 199 ACETOIN FERMENTATION BY L. LACTIS 2647 TABLE 1. Effects of hemin and Cu2" on growth, ph, glucose utilization, and end product formation by Citr+ L. lactis subsp. lactis 322 cells grown aerobicallya Additionlontime Incubation (h) Loglo ph Amt (mmol) of Amt (mmol) of end product formed: D58 P glucose utilized (%) Lactate Acetate Diacetyl Acetoin None 24.21 4.65 3.5 (31.4) 41.4 5..14 9.7 48.17 4.52 37.6 (38.8) 5.3 4.4.12 5.5 1,uM hemin 24.54 4.73 89.2 (91.8) 34.1 1.3.88 55.9 48.5 5.31 96.8 (99.7).9 16.7 1.14 78.5.1 mm CuCl2 24.39 4.73 85.5 (88.) 35.8 2.2.43 32.7 48.35 4.61 97.1 (99.9) 44.8 2.9.53 35. 1,uM hemin +.1 mm CUC12 24 48.62.63 5. 5.58 96.9 (99.8) 96.9 (99.8) 4.2. 3.3 9.9.83 1.19 85.3 15.4 a Citr+ 1. lactis subsp. lactis 322 cells were propagated in a basal medium at 3 C for 18 h. Next, 1 ml of this was inoculated into 1 ml of fresh basal medium and incubated aerobically at 3 C. Cultivations in basal medium containing 1 plm hemin,.1 mm CuC12, or 1 p.m hemin plus.1 mm CuC12 were also carried out. b OD58, Optical density at 58 nm. heme has been found in some lactic acid bacteria grown in medium with heated blood or hematin (11, 3, 31), one possible explanation for the enhancement of growth in hemin-containing culture is that there is a system to detoxify H22. Formation of H22 in lactic acid bacteria occurs through the action of a flavoprotein, NADH oxidase (7, 8, 18, 27). Citr+ L. lactis subsp. lactis 322 also possessed NADH oxidase, although there was no marked difference in the activity between the cells grown with and without hemin (Table 3). The cells produced H22 when incubated under aerobic conditions. In this case, H22 production was repressed by the addition of hemin to the cell suspension, since hemin can decompose H22 (16). However, catalase activity was not detected in the cells grown aerobically with hemin. The increases in growth, glucose utilization, and acetoin production were not observed by the addition of catalase into the medium. These results indicate that the elimination of H22 formed in the culture is not the main reason for the enhancement of growth, although the decomposition of H22 by hemin may be available for the cells. Another possible explanation for the enhancement of growth is that there is a system for ATP formation from oxidative phosphorylation, because cytochrome induction occurs exclusively under aerobic conditions and is dependent upon the environment for exogenous heme (4, 22, 23, TABLE 2. Effects of lipoic acid on glucose utilization and end product formation by Citr+ L. lactis subsp. lactis 322 cells grown aerobically in the presence of hemin or Cu2+a Addition Amt Amt (mmol) of end products formed: (mmol) glucose of utilized Lactate Acetate Diacetyl Acetoin None 29.2 53.7 7.2.11 <5. 1. mm lipoic acid 28. 44.5 8.9.2 <5. 1,uM hemin 98.7 31.1 9..98 61.3 1 pum hemin + 1. 7.7 19.5 31.2.75 37.6 mm lipoic acid.1 mm CuCl2 34.5 43.3 5.8.31 25.5.1 mm CuCl2 + 1. 31.2 4.9 1Q.2.3 <5. mm lipoic acid a Citr+ L. lactis subsp. lactis 322 cells were grown at 3 C for 24 h (see Table 1, footnote a), except that the medium contained 1. mm lipoic acid, 1 FLM hemin,.1 mm CuC12, and 1 p.m hemin plus 1. mm lipoic acid or.1 mm CuCl2 plus 1. mm lipoic acid. 25). However, no trace of cytochrome peaks was recognized in Citr+ L. lactis subsp. lactis 322 cells grown aerobically with hemin (Fig. 2). Furthermore, antimycin A, which is an inhibitor of oxidative phosphorylation, did not reduce the stimulative effects of hemin on growth, glucose utilization, and acetoin production. These results indicate that the cytochrome-mediated respiration system is absent in the cells grown aerobically with hemin. Therefore, it appeared that the cells grown aerobically with hemin or Cu2+ would have an alternative system to reoxidize NADH which is 46 52 58 64 7 Wave Length (nm) FIG. 2. Light absorption spectra of Citr+ L. lactis subsp. lactis 322 and S. cerevisiae 84 cells. S. cerevisiae 84 cells (A) were grown in a basal medium at 3 C for 24 h. Citr' L. lactis subsp. lactis 322 cells were grown aerobically at 3 C for 24 h in a basal medium with (B) and without (C) 1,uM hemin. The cells were prepared as described in Materials and Methods.

2648 KANEKO ET AL. TABLE 3. Effects of hemin on the activities of several enzymes of Citr+ L. lactis subsp. lactis 322a Addition to: Sp act (mu/mg of protein) of: Assay NADH Diacetyl Diacetyl Lactate no. Medium Reaction oxi- syn- reduc- dehydromixture dase thase tase genase 1 None None 4 8.7 2,81 1 2 None 5 F.M hemin 4 3.2 2,59 1 3 1,uM hemin None 3 41.1 1,57 NDb a Cell extracts were prepared after Citr+ L. lactis subsp. lactis 322 cells were grown aerobically in a basal medium with and without 1,uM hemin at 3 C for 15 h. Assays 1 and 2 were performed with the same cell extract which was prepared from the cells grown aerobically in a basal medium without hemin. b ND, Not determined. formed via glycolysis, since the lactate production per mole of glucose was markedly repressed. Acetoin formation from diacetyl by diacetyl reductase is coupled to the oxidation of NADH to NAD. In this study, the diacetyl reductase activity of Citr+ L. lactis subsp. lactis 322 was markedly high. Furthermore, the activity of diacetyl synthase was greatly stimulated by the addition of hemin or Cu2+ to the medium or reaction mixture (Tables 3 and 4). These results indicate that the pyruvate formed via glycolysis would be rapidly converted to diacetyl in aerated culture with hemin or Cu2+ and that the diacetyl would then be converted to acetoin by the NAD-dependent diacetyl reductase. In fact, most of the glucose was converted to diacetyl and acetoin more rapidly in the culture with both hemin and Cu2+ than in the cutlure without these additives (Fig. 1; Table 1). However, growth, glucose utilization, and acetoin production were apparently lower in the culture with both KCN and hemin than in the culture supplemented with only hemin, because KCN reduced the stimulatory effect of hemin on diacetyl synthase activity (T. Kaneko, unpublished results). KCN is known to combine strongly with heme compounds (26). Therefore, the amount of diacetyl needed for the synthesis of acetoin is likely to be lowered by the addition of KCN, and the pathway of acetoin production from diacetyl would be essential to reoxidize NADH formed via glycolysis when Citr+ L. lactis subsp. lactis 322 cells were grown aerobically with hemin or Cu2+. Diacetyl is normally formed enzymatically by condensation of hydroxyethylthiamine PP1 with acetyl coenzyme A (28). We defined the activity of diacetyl formation from pyruvate (diacetyl synthase) as the amount of enzyme which catalyzes the production of 1 p.mol of diacetyl from pyruvate TABLE 4. Effects of lipoic acid on diacetyl synthase activity in the presence of hemin or Cu2+ Diacetyl synthase Addition sp acta (mu/mg of protein) None... 5.9.5 mm lipoic acid... 4.6.1 mm CuC12... 22.8.1 mm CuCl2 +.5 mm lipoic acid... 17.2 1,uM hemin...... 21.2 1,uM hemin +.5 mnm lipoic acid... 18.6 a Diacetyl synthase activities were assayed as described in Materials and Methods, except that the reaction mixture contained lipoic acid, hemin, or cu2+. APPL. ENVIRON. MICROBIOL. in the presence of thiamine PP1 (see Materials and Methods). Therefore, diacetyl synthase catalyzes both the formation of acetyl-coa from pyruvate and the condensation of acetyl- CoA with hydroxyethylthiamine PP1. In this study, formate was not detected when the cells were grown aerobically with and without hemin, probably because pyruvate-formate lyase is sensitive to oxygen (1). Therefore, acetyl-coa formation by Citr+ L. lactis subsp. lactis 322 in aerated culture would be catalyzed by the pyruvate dehydrogenase complex. Hemin or Cu2+ stimulated diacetyl and acetoin production, although acetate production was repressed when excess glucose was present (Fig. 1). However, in the presence of lipoic acid, an essential cofactor of the dihydrolipoamide acetyltransferase part of the pyruvate dehydrogenase complex, hemin or Cu2+ enhanced acetate production and then repressed diacetyl and acetoin production (Table 2). Cogan et al. (5) also showed that acetate production in aerated culture depended on lipoic acid. In this study, the activity of diacetyl synthase was lowered by lipoic acid, regardless of the presence of hemin or Cu2+ (Table 4). However, acetate production from acetyl-coa was not stimulated by the addition of lipoic acid to the reaction mixture (T. Kaneko, unpublished results). These results indicate that hemin or Cu2+ would stimulate acetyl-coa formation from pyruvate and that lipoic acid would inhibit the condensation of acetyl-coa with hydroxyethylthiamine PP1. Therefore, it appeared that acetyl-coa not used for diacetyl synthesis would be converted to acetate. When the cells were grown aerobically with hemin, the amount of acetyl-coa available for acetate synthesis would not be sufficient in the presence of glucose, because most of acetyl-coa is utilized for diacetyl and acetoin production to regenerate NAD. The reduction of acetate production would be the reason why the YGIu value in hemin-containing culture was lower than in the culture without hemin. However, after the disappearance of glucose from the medium, the cells utilized the lactate accumulated in the culture and produced acetate concomitantly when grown in the presence of hemin (Fig. 1). Presumably, since NADH formation is low in the absence of glucose, the regeneration of NAD coupled with the conversion of diacetyl to acetoin is not so strongly required. LITERATURE CITED 1. Abbe, K., S. Takahashi, and T. Yamada. 1982. Involvement of oxygen-sensitive pyruvate formate-lyase in mixed-acid fermentation by Streptococcus mutans under strictly anaerobic conditions. J. Bacteriol. 152:175-182. 2. Borch, E., and G. Molin. 1989. The aerobic growth and product formation of Lactobacillus, Leuconostoc, Brochothrix and Carnobacterium in batch cultures. Appl. Microbiol. Biotechnol. 3:81-88. 3. Bruhn, J. C., and E. B. Collins. 197. Reduced nicotinamide adenine dinucleotide oxidase of Streptococcus diacetilactis. J. Dairy Sci. 53:857-86. 4. Bryan-Jones, D. G., and R. Whittenbury. 1969. Haematindependent oxidative phosphorylation in Streptococcus faecalis. J. Gen. Microbiol. 58:247-26. 5. Cogan, J. F., D. Walsh, and S. Condon. 1989. Impact of aeration on the metabolic end-products formed from glucose and galactose by Streptococcus lactis. J. Appl. Bacteriol. 66:77-84. 6. Cogan, T. M. 1975. Citrate utilization in milk by Leuconostoc cremoris and Streptococcus diacetilactis. J. Dairy Res. 42:139-146. 7. Collins, E. G., and K. Aramaki. 1979. Production of hydrogen peroxide by Lactobacillus acidophilus. J. Dairy Sci. 63:353-357. 8. Condon, S. 1987. Responses of lactic acid bacteria to oxygen.

VOL. 56, 199 FEMS Microbiol. Rev. 46:269-28. 9. De Man, J. C., M. Rogosa, and M. E. Sharpe. 196. A medium for the cultivation of lactobacilli. J. Appl. Bacteriol. 23:13-135. 1. Drinan, D. F., S. Tobin, and T. M. Cogan. 1976. Citric acid metabolism in hetero- and hemofermentative lactic acid bacteria. Appl. Environ. Microbiol. 31:481-486. 11. Johnston, M. A., and E. A Delwiche. 1965. Distribution and characteristics of the catalase of Lactobacillaceae. J. Bacteriol. 9:347-351. 12. Kaneko, T., H. Suzuki, and T. Takahashi. 1986. Diacetyl formation and degradation by Streptococcus lactis subsp. diacetylactis 322. Agric. Biol. Chem. 5:2639-2641. 13. Kaneko, T., H. Suzuki, and T. Takahashi. 1987. The effects of metal ions on diacetyl production by Streptococcus lactis subsp. diacetylactis 322. Agric. Biol. Chem. 51:2315-232. 14. Kaneko, T., H. Suzuki, and T. Takahashi. 1987. Influences of cellular components and redox potential of liquid concentrated culture of Lactobacillus bulgaricus on acid-producing activity and viability. J. Dairy Sci. 7:1128-1133. 15. Kaneko, T., Y. Watanabe, and H. Suzuki. 199. Enhancement of diacetyl production by a diacetyl-resistant mutant of citratepositive Lactococcus lactis subsp. lactis 322 and by aerobic conditions of growth. J. Dairy Sci. 73:291-298. 16. Kremer, M. L. 1989. The reaction of hemin with H22. Eur. J. Biochem. 185:651-658. 17. Lowry,. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265-275. 18. Lucey, C. A., and S. Condon. 1986. Active role of oxygen and NADH oxidase in growth and energy metabolism of Leuconostoc. J. Gen. Microbiol. 132:1789-1796. 19. Marshall, V. M. 1987. Lactic acid bacteria: starters for flavours. FEMS Microbiol. Rev. 46:327-336. 2. Murphy, M. G., and S. Condon. 1984. Correlation of oxygen utilization and hydrogen peroxide accumulation with oxygen induced enzymes in Lactobacillus plantarum cultures. Arch. ACETOIN FERMENTATION BY L. LACTIS 2649 Microbiol. 138:44 48. 21. Murphy, M. G., and S. Condon. 1984. Comparison of aerobic and anaerobic growth of Lactobacillus plantarum in a glucose medium. Arch. Microbiol. 138:49-53. 22. Pritchard, G. G., and J. W. T. Wimpenny. 1978. Cytochrome formation, oxygen-induced proton extrusion and respiratory activity in Streptococcus faecalis var. zymogenes grown in the presence of haematin. J. Gen. Microbiol. 14:15-22. 23. Ritchey, T. W., and H. W. Seeley, Jr. 1974. Cytochromes in Streptococcus faecalis var. zymogenes grown in a haematincontaining medium. J. Gen. Microbiol. 85:22-228. 24. Shibata, K., A. A. Benson, and M. Calvin. 1954. The absorption spectra of suspensions of living microorganisms. Biochim. Biophys. Acta 15:461-47. 25. Sijpesteijn, A. K. 197. Induction of cytochrome formation and stimulation of oxidative dissimilation by hemin in Streptococcus lactis and Leuconostoc mesenteroides. Antonie van Leeuwenhoek 36:335-348. 26. Slater, E. C. 1967. Application of inhibitors and uncouplers for a study of oxidative phosphorylation. Methods Enzymol. 1:48-57. 27. Smart, J. B., and T. D. Thomas. 1987. Effect of oxygen on lactose metabolism in lactic streptococci. Appl. Environ. Microbiol. 53:533-541. 28. Speckman, R. A., and E. B. Collins. 1968. Diacetyl biosynthesis in Streptococcus diacetilactis and Leuconostoc citrovorum. J. Bacteriol. 95:174-18. 29. Thomas, T. D., and K. W. Turner. 1981. Carbohydrate fermentation by Streptococcus cremoris and Streptococcus lactis growing in agar gels. Appl. Environ. Microbiol. 41:1289-1294. 3. Whittenbury, R. 1964. Hydrogen peroxide formation and catalase activity in the lactic acid bacteria. J. Gen. Microbiol. 35:13-26. 31. Wolf, G., and W. P. Hammes. 1988. Effect of hematin on the activities of nitrite reductase and catalase in lactobacilli. Arch. Microbiol. 149:22-224.