- Email: [email protected]
An improved method for the determination of NADH oxidase in the presence of NADH peroxidase in lactic acid bacteria
Journal of Microbiological Methods 52 (2003) 333 – 339 www.elsevier.com/locate/jmicmeth
An improved method for the determination of NADH oxidase in the presence of NADH peroxidase in lactic acid bacteria Akshat Talwalkar, Kaila Kailasapathy *, Jim Hourigan, Paul Peiris, Rama Arumugaswamy Centre for Advanced Food Research, University of Western Sydney-Hawkesbury, Locked Bag 1797, NSW 1797, Australia Received 14 May 2002; received in revised form 30 August 2002; accepted 3 September 2002
Abstract The complexity of the coupled NADH oxidase – NADH peroxidase enzyme system in lactic acid bacteria makes it difficult to simultaneously determine the individual levels of both these enzymes spectrophotometrically. This study describes an improved assay to accurately determine low concentrations of NADH oxidase from enzyme suspensions containing NADH oxidase and NADH peroxidase. For the standardisation of the assay, pure NADH oxidase and NADH peroxidase were mixed in various proportions and the percentage recovery was estimated by both the currently available assay as well as by the improved assay reported in this study. The recovery of NADH oxidase using the currently available assay ranged from as low as 200% to as high as + 102% as against 90 – 102% in the improved assay. The recovery percentage of NADH peroxidase ranged from 91% to 112% in both assays. The slopes of NADH oxidation by cell-free extracts of six lactic acid bacteria were also measured by both assays for the estimation of NADH oxidase and NADH peroxidase levels. The improved assay can further distinguish between NADH – H2O oxidase and NADH – H2O2 oxidase and was successfully applied to identify the type of NADH oxidase in the lactic acid bacteria tested. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Lactic acid bacteria; NADH oxidase; NADH peroxidase; Spectrophotometric assay
1. Introduction Anaerobic lactic acid bacteria rely on non-haem flavoproteins that act as NADH oxidases and peroxidases to protect against oxygen toxicity for better survival (Dolin, 1961; Condon, 1987). NADH-oxi-
* Corresponding author. Tel.: +61-2-45701231; fax: +61-245701954. E-mail address: [email protected] (K. Kailasapathy).
dising enzymes catalyse the one-, two-, or fourelectron reduction of O2 to O2 , H2O2, or H2O (Higuchi et al., 2000). It is widely accepted that a typical assay of NADH oxidase measures the initial linear slope of NADH oxidation at 340 nm in the presence of cell-free extract and air-saturated buffer (Schmidt et al., 1986; Higuchi et al., 1993; Shin and Park, 1997; Yi et al., 1998; Marty-Teysset et al., 2000). Though this assay is suitable for lactic acid bacteria having only NADH oxidase, it is insufficient for estimating the levels of NADH oxidase in organ-
0167-7012/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 7 0 1 2 ( 0 2 ) 0 0 1 8 9 - 6
334
A. Talwalkar et al. / Journal of Microbiological Methods 52 (2003) 333–339
isms in which NADH peroxidase is also present. As the product of an NADH – H2O2 oxidase reaction, i.e. H2O2 is also the substrate for NADH peroxidase, the slope of NADH oxidation (oxidase activity) is actually a sum of the total NADH oxidised by the activities of both oxidase and peroxidase. While this has not been reported in some published literature, other researchers have had to perform amperometric methods in order to determine individual levels of NADH oxidase based on the oxygen uptake (Carlsson et al., 1983; Thomas and Pera, 1983; Smart and Thomas, 1987; Cox and Marling, 1992; Shimamura et al., 1992). Considerable variation also exists in the assays reported to measure NADH peroxidase. Shimamura et al. (1992) have estimated activities of NADH peroxidase by measuring the consumption of H2O2 under anaerobic conditions. Others have assayed NADH peroxidase activity independently by measuring the slope of NADH oxidation under anaerobic conditions (Anders et al., 1970; Carlsson et al., 1983; Thomas and Pera, 1983; Smart and Thomas, 1987; Shin and Park, 1997). Uesugi and Yajima (1978) and de Vries and Stouthamer (1969) estimated NADH peroxidase as the slope difference in presence and absence of H2O2 under aerobic conditions, whereas the same slope difference obtained under anaerobic conditions was used by Higuchi et al. (1993) for the measurement of NADH peroxidase. Under aerobic conditions and in absence of H2O2, the activity of NADH peroxidase will be dependent solely on the rate of production of H2O2 by NADH oxidase. This introduces a substrate limitation step for NADH peroxidase. As against this, under anaerobic conditions and in excess H2O2, the reaction velocity of NADH peroxidase would be maximal. For the subtraction method to be accurate (Uesugi and Yajima, 1978; de Vries and Stouthamer, 1969; Higuchi et al., 1993), the reaction velocities of NADH peroxidase in the presence as well as absence of excess H2O2 need to be at their maximum, or else, it would lead to inaccurate estimations of NADH oxidase. As it is evident, the interconnectedness of the coupled NADH oxidase– NADH peroxidase enzyme system makes it difficult to simultaneously determine the individual levels of both these enzymes. A standard spectrophotometric assay for accurately determining the levels of NADH oxidase and NADH
peroxidase from such a coupled oxidase –peroxidase system has not been reported yet. To overcome this problem, we developed an improved assay containing excess H2O2 in the reaction system of both aerobic and anaerobic assays. This improved spectrophotometric assay was able to accurately and simultaneously determine the individual levels of NADH oxidase and NADH peroxidase from enzyme suspensions having different proportions of pure oxidase and pure peroxidase. Slopes of NADH oxidation by cell-free extracts of six different lactic acid bacteria strains were also determined by this improved assay.
2. Materials and methods 2.1. Enzymes Pure NADH oxidase and NADH peroxidase (E.C. 1.11.1.1) were obtained from Calbiochem and Sigma– Aldrich, respectively. Stock solutions of 1.0 U ml 1 of each enzyme were prepared in appropriate diluents as provided in the manufacturers’ instructions. Suspensions of oxidase and peroxidase units mixed in different proportions were used for the assays. One unit of NADH oxidase was defined as the amount of enzyme catalysing the oxidation of 1 nmol NADH per min at 30 jC. One unit of NADH peroxidase was defined as the amount of enzyme catalysing the oxidation of 1 nmol H2O2 per min at 30 jC. 2.2. Enzyme assay 2.2.1. Estimation of NADH oxidase by the currently available assay (Smart and Thomas, 1987; Cox and Marling, 1992; Shin and Park, 1997; Marty-Teysset et al., 2000) The reaction system consisted of NADH (67 AM), FAD (67 AM) and Bis – Tris buffer (0.1 M), pH 6.0, in a total volume of 3 ml. The reaction mixture contained 5, 10, 15, or 20 U of NADH oxidase and NADH peroxidase combined in different proportions (Table 1). The assays were conducted at 30 jC under aerobic conditions. The decrease in the absorbance of NADH at 340 nm was measured for a period of 3 min using a Biochrom 4060 spectrophotometer.
A. Talwalkar et al. / Journal of Microbiological Methods 52 (2003) 333–339
335
Table 1 Comparison between percentage recoveries of NADH oxidase and NADH peroxidase by the currently available assay and the improved assay NADH oxidase units (U)
NADH peroxidase units (U)
Percent recovery of oxidase Currently available assay
Improved assay
Currently available assay
Improved assay
5
5 10 15 20 5 10 15 20 5 10 15 20 5 10 15 20
91.6 F 10.1 9.6 F 13.6 94.8 F 16.6 200.9 F 11.2 101.2 F 6.8 86.0 F 1.9 41.8 F 2.4 2.4 F 6.6 93.7 F 1.3 92.1 F 5.6 102.3 F 6.2 73.4 F 5.1 102.5 F 3.9 99.2 F 6.7 102.8 F 5.2 98.8 F 3.7
99.6 F 15.7a 93.2 F 7.8b 106.1 F 17.2b 98.0 F 3.9b 90.8 F 6.4a 94.0 F 6.6a 102.9 F 5.8b 93.2 F 7.8b 98.6 F 4.3a 93.7 F 4.7a 94.8 F 5.6a 94.3 F 5.2b 100.0 F 3.3a 96.4 F 4.5a 92.8 F 5.0a 92.4 F 4.7a
99.6 F 4.9 93.2 F 2.4 98.0 F 4.4 97.6 F 2.5 101.2 F 5.2 102.8 F 3.9 104.5 F 2.6 96.4 F 2.6 91.6 F 5.2 98.0 F 7.8 96.4 F 3.5 98.4 F 3.2 107.7 F 14.1 97.2 F 9.3 97.5 F 4.8 96.8 F 2.8
91.6 F 5.2a 100.4 F 7.0a 99.6 F 4.0a 94.8 F 1.9a 104.5 F 9.4a 101.2 F 6.8a 97.5 F 5.6a 96.8 F 2.8a 112.5 F 9.9a 100.4 F 8.3a 101.8 F 3.3a 96.4 F 2.1a 98.0 F 9.4a 101.2 F 7.4a 98.6 F 3.8a 98.0 F 3.2a
10
15
20
Percent recovery of peroxidase
Means F S.D. (n = 6). a Nonsignificant difference. b Significant difference.
Using a Reaction Kinetics software (Biochrom), the initial linear slope of NADH oxidation was recorded. The molar extinction coefficient of NADH at 340 nm (6.22 103 M 1 cm 1) was used for calculating the enzyme units.
subtracting the slope of anaerobic assay from that of the aerobic assay and converted to enzyme units/ cuvette using the following formula:
2.2.2. Estimation of NADH peroxidase by the currently available assay (Thomas and Pera, 1983; Smart and Thomas, 1987; Shin and Park, 1997) H2O2 (1 mM) was incorporated in the reaction mixture given above and the assay was conducted under anaerobic conditions.
These recovered enzyme units were then compared to the actual NADH oxidase enzyme units introduced. NADH peroxidase was estimated by converting the slope of the anaerobic assay into enzyme units/cuvette by the abovementioned formula. This was compared with the number of NADH peroxidase units added. The percentage recovery was then calculated for both enzymes. For the anaerobic assay, the reactants were prepared in an anaerobic workstation (Coy, USA) containing 95% N2 and 5% H2, and kept in an anaerobic condition for 24 h prior to the assay. Nitrogen gas was bubbled through the reactants before the determination. Using a Clark-type oxygen electrode (Microelectrodes, USA), the dissolved oxygen in the reactants was ensured to be zero. No increase in oxygen was recorded within 5 min in the cuvette containing the anaerobic reaction mixture.
2.2.3. Estimation of NADH oxidase and NADH peroxidase by the improved assay The reaction mixture was the same as reported for estimating NADH peroxidase by the currently available assay except that the assay was conducted under both aerobic and anaerobic conditions. In both the currently available assay as well as in the improved assay, NADH oxidase was estimated by converting the slope of the aerobic assay into enzyme units/cuvette. In addition, a separate estimation of NADH oxidase was also performed by
Units=cuvette ¼ ðD A340 3Þ=6:22
336
A. Talwalkar et al. / Journal of Microbiological Methods 52 (2003) 333–339
Additionally, the respective blanks were performed before conducting the assays. The concentration of the reactants in the blanks was the same as that of the actual assay. The mean of six individual determinations was used for calculation and a Student’s t-test was performed (a = 0.05).
conditions when 1 mM H2O2 was incorporated in the free assay mix containing no NADH oxidase. In the absence of H2O2, no decrease in absorbance was observed in both aerobic and anaerobic assays.
2.3. Microorganisms, culture conditions, preparation of cell-free extract, and measurement of the slope of NADH oxidation
When the levels of NADH oxidase were determined from just the aerobic assay slope, all combinations of NADH oxidase with NADH peroxidase showed significantly greater ( p < 0.05) recovery levels of NADH oxidase than what was introduced in the cuvette. This was noted in both the currently available assay as well as in the improved assay. The recovery of NADH oxidase determined from the subtraction of the anaerobic assay slope from the aerobic assay slope is shown in Fig. 1. Although no significantly ( p>0.05) elevated recovery was detected, considerable variation was observed in the recovery of NADH oxidase by the currently available assay. When suspensions combining 5 U NADH oxidase with 15 and 20 U NADH peroxidase were assayed, the recovery obtained by subtracting the peroxidase slope from the oxidase slope gave negative values. In suspensions containing 10 U NADH oxidase and 15 U NADH peroxidase, the subtraction of the slopes resulted in a value of 4.18 U, whereas in suspensions containing 15 U NADH oxidase and 20 U NADH peroxidase, 11.01 U was obtained after calculation. For the abovementioned enzyme combinations, however, the improved assay suggested in this paper demonstrated no significant difference ( p>0.05) between the values of NADH oxidase introduced and that calculated from the slope of NADH oxidation. The percentage recovery for NADH oxidase determined by subtracting the anaerobic assay slope from the aerobic assay slope is listed in Table 1. In the currently available assay, the recovery of NADH oxidase changed with differing oxidase –peroxidase ratios. When the enzyme suspension contained lower amounts of NADH oxidase than NADH peroxidase units, the calculated recovery ranged from 73% to as low as 200%. For enzyme suspensions having equal or higher amounts of NADH oxidase than NADH peroxidase, the calculated recovery ranged from 86% to 102%. However, in the improved spectrophotometric assay, the percentage recovery for
Lactobacillus acidophilus 2400, L. acidophilus 2409, Bifidobacterium infantis 1912, B. lactis 1941, and B. pseudolongum 1944 were obtained from the Commonwealth Scientific Industrial Research Organization (CSIRO), Australia. B. longum 55815 was received from American Type Culture Collection (ATCC), USA. All strains were grown anaerobically for 24 h at 37 jC in 200 ml MRS broth (Oxoid, Australia). Cells were harvested by centrifugation for 10 min at 10,000 g at 4 jC and the cell pellet was washed thrice with 0.1 M phosphate buffer, pH 7. Cells were disrupted by passing them through a FrenchR Pressure Cell (Thermospectronic, USA) and cell-free extract was obtained by removing the cell debris by centrifugation for 15 min at 12,000 g at 4 jC. Pure enzymes in reaction system of all the abovementioned assays were replaced by an appropriate volume of cell-free extract and the slope of NADH oxidation was recorded. Previously boiled cell-free extract was used to negate the possibility of nonenzymatic oxidation of NADH.
3. Results 3.1. Results of assay blanks A blank containing NADH, FAD, H2O2, and buffer showed no decrease in absorbance over the time of the assay under both aerobic and anaerobic conditions. The activity of NADH oxidase alone under aerobic conditions was not affected in the presence of 1 mM H2O2. Under anaerobic conditions, however, no NADH oxidase activity was noticed. NADH peroxidase alone demonstrated the same activity under both aerobic and anaerobic
3.2. Recovery of NADH oxidase
A. Talwalkar et al. / Journal of Microbiological Methods 52 (2003) 333–339
NADH oxidase remained very high regardless of the proportion of NADH oxidase and NADH peroxidase units and ranged between 90% and 102% even at lower concentrations of oxidase. The means of the percentage recovery from the currently available assay and our improved assay were found to differ significantly ( p < 0.05) in enzyme suspensions where the amount of NADH oxidase units was less than that of NADH peroxidase units. In all the remaining enzyme suspensions where the proportion of NADH oxidase was either equal to or greater than NADH peroxidase, no significant difference ( p>0.05) was observed among the means of the two assays. 3.3. Recovery of NADH peroxidase It was interesting to note that in both assays, the calculated values of NADH peroxidase approximated the number of units of NADH peroxidase introduced in the cuvette. The anaerobic conditions of the assay and the abundance of substrate (H2O2) ensured maximum activity of NADH peroxidase. Consequently, the values obtained through calculation showed similarity with the actual peroxidase units introduced. The proportion of oxidase and peroxidase units in the various enzyme suspensions did not affect the recovery of NADH peroxidase. No significant difference ( p>0.05) was found between the means of the two assays for NADH peroxidase. The means ranged from 91% to 107% for the currently available assay and from 91% to 112% in the improved assay (Table 1). 3.4. Slope of NADH oxidation in cell-free extracts of lactic acid bacterial strains No oxidation of NADH was observed when boiled cell-free extract was used in the assays. Cell-free extracts of all six bacterial strains oxidised NADH when assayed under anaerobic conditions and in presence of H2O2 (Table 2). The slope of NADH oxidation by the currently available assay differed from that obtained by the improved assay. Negative values were observed in B. infantis 1912 and B. pseudolongum 1944 when the slope of NADH peroxidase assay was subtracted from the slope of NADH oxidase assay (currently available assay). With the NADH oxidase-improved assay, however,
337
Table 2 Differences in the estimation of NADH oxidases of six lactic acid bacteria strains by the currently available assay and the improved assay Strain
B. infantis 1912 B. lactis 1941 B. pseudolongum 1944 B. longum 55815 L. acidophilus 2400 L. acidophilus 2409
Slope of NADH oxidase assay (a)
Slope of NADH peroxidase assay (b)
Difference in slopes (a b)
CAAa
IAb
CAAa
IAb
0.10 0.15 0.12
0.22 0.15 0.24 0.11 0.22 0.13
0.05 0.07 0.04 0.13 0.01 0.09
0.16 0.70
0.21 0.10 1.10 0.52
0.06 0.11 0.18 0.58
0.39
0.60 0.30
0.09 0.30
Mean (n = 6). a CAA = currently available assay. The reaction system of cellfree extract, NADH (67 AM), FAD (67 AM) and Bis – Tris buffer (0.1 M), pH 6.0, in a total volume of 3 ml was assayed for 3 min at 30 jC under aerobic conditions. b IA = improved assay. The reaction system of cell-free extract, NADH (67 AM), FAD (67 AM), H2O2 (1 mM) and Bis – Tris buffer (0.1 M), pH 6.0, in a total volume of 3 ml was assayed for 3 min at 30 jC under aerobic conditions.
the difference in the slopes gave positive values for all the six bacterial strains.
4. Discussion The percentage recovery in the currently available assay was found to depend solely on the ratio of NADH oxidase and NADH peroxidase, and change as their proportions differed. Some researchers have estimated NADH oxidase from just the aerobic assay slope (de Vries and Stouthamer, 1969; Uesugi and Yajima, 1978; Shin and Park, 1997). The slope of NADH oxidation in the aerobic assay, however, is actually a sum of the total NADH oxidised by the activities of both oxidase and peroxidase. Enzyme units calculated from this slope would, therefore, result in elevated levels of NADH oxidase. This was confirmed by the significantly elevated recoveries of NADH oxidase obtained when its levels were determined by this method as also by elevated slopes of NADH oxidation by cell-free extracts of the six bacterial strains (Table 2). This, there-
338
A. Talwalkar et al. / Journal of Microbiological Methods 52 (2003) 333–339
fore, suggests that the reported values of NADH oxidase where levels were determined from just the slope of aerobic assay may have been overestimated. Smart and Thomas (1987) have reported that their amperometric estimation of NADH oxidase correlated well with that obtained from the subtraction of the anaerobic assay slope from the slope of the aerobic assay. This suggests that one can subtract the slope of peroxidase (anaerobic assay slope) from the oxidase – peroxidase slope (aerobic assay slope) to accurately determine the levels of NADH oxidase spectrophotometrically. The difference in the reaction velocities of NADH peroxidase in these two assays, however, can give rise to inaccurate estimations of NADH oxidase. This is confirmed in the negative recovery percentages of NADH oxidase obtained using the currently available assay (Fig. 1) and by the negative slope differences in some of the bacterial strains tested (Table 2). As against this, in our improved assay, the uniformity of the reactants in the aerobic and anaerobic assay ensured oxygen as the only variable-affecting enzyme activities between these two assays. This guaranteed accurate estimations of NADH oxidase when the slope of NADH peroxidase was subtracted from the slope of
the aerobic assay and was reflected in the high-percentage recoveries of NADH oxidase as well as NADH peroxidase in all the different enzyme proportions we tested (Table 1). This was further confirmed by positive slope differences in all the cell-free extracts assayed (Table 2). In many reports of NADH oxidases in lactic acid bacteria, the assay system used was based on the consumption of NADH; however, the end product was not measured. This does not distinguish between H2O- and H2O2-forming NADH oxidases. This is further complicated by the fact that the activity of an H2O2-forming NADH oxidase combined with that of an excess of NADH peroxidase is similar to an H2O-forming NADH oxidase (Condon, 1987). Although our improved assay is best suited for NADH – H2O2 oxidase/NADH peroxidase system, it can also help in distinguishing between NADH – H2O2 and NADH – H2O oxidases. This can be achieved by performing an additional aerobic assay without the addition of any H2O2 in the reaction system. If the slopes of the aerobic assay in the presence and absence of H2O2 are similar, then the enzyme in question is an NADH – H2O oxidase, regardless of the presence of
Fig. 1. Recovery of NADH oxidase in the presence of NADH peroxidase.
A. Talwalkar et al. / Journal of Microbiological Methods 52 (2003) 333–339
any peroxidase. If peroxidase is detected and the slope of the aerobic assay in the absence of H2O2 is less than in presence of H2O2, then it is an NADH – H2O2 oxidase. NADH peroxidase activity was detected in all the bacterial strains tested, and the slope of NADH oxidation in absence of H2O2 was less than in presence of H2O2 (Table 2). Accordingly, it can be concluded that all six strains possessed NADH – H2O2 oxidase. In lactic acid bacteria containing NADH oxidase and NADH peroxidase, the proportion of these two enzymes can vary from strain to strain. In this paper, 16 different proportions were tested and our improved assay was found to demonstrate high accuracy in the recovery of both NADH oxidase (especially low levels) and NADH peroxidase, regardless of the enzyme proportions. In comparison, the currently available assay was suitable only for determining individual levels of NADH peroxidase. When levels of NADH oxidase were low in comparison to NADH peroxidase, this assay gave inaccurate estimations of NADH oxidase. It is also clear that estimating the level of NADH oxidase from just the slope of the aerobic assay may lead to over estimation of the enzyme units. In addition, cell-free extracts of six lactic acid bacteria did not interfere with the measurement of the slope of NADH oxidation by our improved assay. The proposed assay reported in this paper, therefore, is a suitable technique for determining the individual levels of NADH peroxidase from a suspension containing NADH oxidase and NADH peroxidase in lactic acid bacteria. Acknowledgements This research was supported by the Australian Research Council (SPIRT) and the Dairy Farmers, Australia. References Anders, R.F., Hogg, D.M., Jago, G.R., 1970. Formation of hydrogen peroxide by group N streptococci and its effect on their growth and metabolism. Appl. Microbiol. 19, 608 – 612.
339
Carlsson, J., Iwami, Y., Yamada, T., 1983. Hydrogen peroxide excretion by oral streptococci and effect of lactoperoxidase – thiocyanate – hydrogen peroxide. Infect. Immun. 40, 70 – 80. Condon, S., 1987. Responses of lactic acid bacteria to oxygen. FEMS Microbiol. Rev. 46, 269 – 281. Cox, R.P., Marling, N., 1992. High-affinity oxygen uptake by Bifidobacterium bifidum. Antonie Van Leeuwenhoek 62, 291 – 297. de Vries, W., Stouthamer, A.H., 1969. Factors determining the degree of anaerobiosis of Bifidobacterium strains. Arch. Mikrobiol. 65, 275 – 287. Dolin, M.I., 1961. Cytochrome-independent electron transport enzymes of bacteria. In: Gunsalus, I.C., Stanier, R.Y. (Eds.), The Bacteria, vol. 2. Academic Press, New York, pp. 425 – 460. Higuchi, M., Shimada, M., Yamamoto, Y., Hayashi, T., Koga, T., Kamio, Y., 1993. Identification of two distinct NADH oxidases corresponding to H2O2-forming oxidase and H2O-forming oxidase induced in Streptococcus mutans. J. Gen. Microbiol. 139, 2343 – 2351. Higuchi, M., Yamamoto, Y., Kamio, Y., 2000. Molecular biology of oxygen tolerance in lactic acid bacteria: functions of NADH oxidases and Dpr in oxidative stress. J. Biosci. Bioeng. 90, 484 – 493. Marty-Teysset, C., de la Torre, F., Garel, J.-R., 2000. Increased production of hydrogen peroxide by Lactobacillus delbrueckii subsp. bulgaricus upon aeration: involvement of an NADH oxidase in oxidative stress. Appl. Environ. Microbiol. 66, 262 – 267. Schmidt, H.-L., Sto¨cklein, W., Danzer, J., Kirch, P., Limbach, B., 1986. Isolation and properties of an H2O-forming NADH oxidase from Streptococcus faecalis. Eur. J. Biochem. 156, 149 – 155. Shimamura, S., Abe, F., Ishibashi, N., Miyakawa, H., Yaeshima, T.A., Tomita, M., 1992. Relationship between oxygen sensitivity and oxygen metabolism of Bifidobacterium species. J. Dairy Sci. 75, 3296 – 3306. Shin, S.-Y., Park, J.-H., 1997. Activities of oxidative enzymes related with oxygen tolerance in Bifidobacterium sp. J. Microbiol. Biotechnol. 7, 356 – 359. Smart, J.B., Thomas, T.D., 1987. Effect of oxygen on lactose metabolism in lactic streptococci. Appl. Environ. Microbiol. 53, 533 – 541. Thomas, E.L., Pera, K.A., 1983. Oxygen metabolism of Streptococcus mutans: uptake of oxygen and release of superoxide and hydrogen peroxide. J. Bacteriol. 154, 1236 – 1244. Uesugi, I., Yajima, M., 1978. Oxygen and ‘‘strictly anaerobic’’ intestinal bacteria: 2. Oxygen metabolism in strictly anaerobic bacteria. Z. Allg. Mikrobiol. 18, 593 – 601. Yi, X., Kot, E., Bezkorovainy, A., 1998. Properties of NADH oxidase from Lactobacillus delbrueckii ssp. bulgaricus. J. Sci. Food Agric. 78, 527 – 534.
An improved method for the determination of NADH oxidase in the presence of NADH peroxidase in lactic acid bacteria Akshat Talwalkar, Kaila Kailasapathy *, Jim Hourigan, Paul Peiris, Rama Arumugaswamy Centre for Advanced Food Research, University of Western Sydney-Hawkesbury, Locked Bag 1797, NSW 1797, Australia Received 14 May 2002; received in revised form 30 August 2002; accepted 3 September 2002
Abstract The complexity of the coupled NADH oxidase – NADH peroxidase enzyme system in lactic acid bacteria makes it difficult to simultaneously determine the individual levels of both these enzymes spectrophotometrically. This study describes an improved assay to accurately determine low concentrations of NADH oxidase from enzyme suspensions containing NADH oxidase and NADH peroxidase. For the standardisation of the assay, pure NADH oxidase and NADH peroxidase were mixed in various proportions and the percentage recovery was estimated by both the currently available assay as well as by the improved assay reported in this study. The recovery of NADH oxidase using the currently available assay ranged from as low as 200% to as high as + 102% as against 90 – 102% in the improved assay. The recovery percentage of NADH peroxidase ranged from 91% to 112% in both assays. The slopes of NADH oxidation by cell-free extracts of six lactic acid bacteria were also measured by both assays for the estimation of NADH oxidase and NADH peroxidase levels. The improved assay can further distinguish between NADH – H2O oxidase and NADH – H2O2 oxidase and was successfully applied to identify the type of NADH oxidase in the lactic acid bacteria tested. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Lactic acid bacteria; NADH oxidase; NADH peroxidase; Spectrophotometric assay
1. Introduction Anaerobic lactic acid bacteria rely on non-haem flavoproteins that act as NADH oxidases and peroxidases to protect against oxygen toxicity for better survival (Dolin, 1961; Condon, 1987). NADH-oxi-
* Corresponding author. Tel.: +61-2-45701231; fax: +61-245701954. E-mail address: [email protected] (K. Kailasapathy).
dising enzymes catalyse the one-, two-, or fourelectron reduction of O2 to O2 , H2O2, or H2O (Higuchi et al., 2000). It is widely accepted that a typical assay of NADH oxidase measures the initial linear slope of NADH oxidation at 340 nm in the presence of cell-free extract and air-saturated buffer (Schmidt et al., 1986; Higuchi et al., 1993; Shin and Park, 1997; Yi et al., 1998; Marty-Teysset et al., 2000). Though this assay is suitable for lactic acid bacteria having only NADH oxidase, it is insufficient for estimating the levels of NADH oxidase in organ-
0167-7012/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 7 0 1 2 ( 0 2 ) 0 0 1 8 9 - 6
334
A. Talwalkar et al. / Journal of Microbiological Methods 52 (2003) 333–339
isms in which NADH peroxidase is also present. As the product of an NADH – H2O2 oxidase reaction, i.e. H2O2 is also the substrate for NADH peroxidase, the slope of NADH oxidation (oxidase activity) is actually a sum of the total NADH oxidised by the activities of both oxidase and peroxidase. While this has not been reported in some published literature, other researchers have had to perform amperometric methods in order to determine individual levels of NADH oxidase based on the oxygen uptake (Carlsson et al., 1983; Thomas and Pera, 1983; Smart and Thomas, 1987; Cox and Marling, 1992; Shimamura et al., 1992). Considerable variation also exists in the assays reported to measure NADH peroxidase. Shimamura et al. (1992) have estimated activities of NADH peroxidase by measuring the consumption of H2O2 under anaerobic conditions. Others have assayed NADH peroxidase activity independently by measuring the slope of NADH oxidation under anaerobic conditions (Anders et al., 1970; Carlsson et al., 1983; Thomas and Pera, 1983; Smart and Thomas, 1987; Shin and Park, 1997). Uesugi and Yajima (1978) and de Vries and Stouthamer (1969) estimated NADH peroxidase as the slope difference in presence and absence of H2O2 under aerobic conditions, whereas the same slope difference obtained under anaerobic conditions was used by Higuchi et al. (1993) for the measurement of NADH peroxidase. Under aerobic conditions and in absence of H2O2, the activity of NADH peroxidase will be dependent solely on the rate of production of H2O2 by NADH oxidase. This introduces a substrate limitation step for NADH peroxidase. As against this, under anaerobic conditions and in excess H2O2, the reaction velocity of NADH peroxidase would be maximal. For the subtraction method to be accurate (Uesugi and Yajima, 1978; de Vries and Stouthamer, 1969; Higuchi et al., 1993), the reaction velocities of NADH peroxidase in the presence as well as absence of excess H2O2 need to be at their maximum, or else, it would lead to inaccurate estimations of NADH oxidase. As it is evident, the interconnectedness of the coupled NADH oxidase– NADH peroxidase enzyme system makes it difficult to simultaneously determine the individual levels of both these enzymes. A standard spectrophotometric assay for accurately determining the levels of NADH oxidase and NADH
peroxidase from such a coupled oxidase –peroxidase system has not been reported yet. To overcome this problem, we developed an improved assay containing excess H2O2 in the reaction system of both aerobic and anaerobic assays. This improved spectrophotometric assay was able to accurately and simultaneously determine the individual levels of NADH oxidase and NADH peroxidase from enzyme suspensions having different proportions of pure oxidase and pure peroxidase. Slopes of NADH oxidation by cell-free extracts of six different lactic acid bacteria strains were also determined by this improved assay.
2. Materials and methods 2.1. Enzymes Pure NADH oxidase and NADH peroxidase (E.C. 1.11.1.1) were obtained from Calbiochem and Sigma– Aldrich, respectively. Stock solutions of 1.0 U ml 1 of each enzyme were prepared in appropriate diluents as provided in the manufacturers’ instructions. Suspensions of oxidase and peroxidase units mixed in different proportions were used for the assays. One unit of NADH oxidase was defined as the amount of enzyme catalysing the oxidation of 1 nmol NADH per min at 30 jC. One unit of NADH peroxidase was defined as the amount of enzyme catalysing the oxidation of 1 nmol H2O2 per min at 30 jC. 2.2. Enzyme assay 2.2.1. Estimation of NADH oxidase by the currently available assay (Smart and Thomas, 1987; Cox and Marling, 1992; Shin and Park, 1997; Marty-Teysset et al., 2000) The reaction system consisted of NADH (67 AM), FAD (67 AM) and Bis – Tris buffer (0.1 M), pH 6.0, in a total volume of 3 ml. The reaction mixture contained 5, 10, 15, or 20 U of NADH oxidase and NADH peroxidase combined in different proportions (Table 1). The assays were conducted at 30 jC under aerobic conditions. The decrease in the absorbance of NADH at 340 nm was measured for a period of 3 min using a Biochrom 4060 spectrophotometer.
A. Talwalkar et al. / Journal of Microbiological Methods 52 (2003) 333–339
335
Table 1 Comparison between percentage recoveries of NADH oxidase and NADH peroxidase by the currently available assay and the improved assay NADH oxidase units (U)
NADH peroxidase units (U)
Percent recovery of oxidase Currently available assay
Improved assay
Currently available assay
Improved assay
5
5 10 15 20 5 10 15 20 5 10 15 20 5 10 15 20
91.6 F 10.1 9.6 F 13.6 94.8 F 16.6 200.9 F 11.2 101.2 F 6.8 86.0 F 1.9 41.8 F 2.4 2.4 F 6.6 93.7 F 1.3 92.1 F 5.6 102.3 F 6.2 73.4 F 5.1 102.5 F 3.9 99.2 F 6.7 102.8 F 5.2 98.8 F 3.7
99.6 F 15.7a 93.2 F 7.8b 106.1 F 17.2b 98.0 F 3.9b 90.8 F 6.4a 94.0 F 6.6a 102.9 F 5.8b 93.2 F 7.8b 98.6 F 4.3a 93.7 F 4.7a 94.8 F 5.6a 94.3 F 5.2b 100.0 F 3.3a 96.4 F 4.5a 92.8 F 5.0a 92.4 F 4.7a
99.6 F 4.9 93.2 F 2.4 98.0 F 4.4 97.6 F 2.5 101.2 F 5.2 102.8 F 3.9 104.5 F 2.6 96.4 F 2.6 91.6 F 5.2 98.0 F 7.8 96.4 F 3.5 98.4 F 3.2 107.7 F 14.1 97.2 F 9.3 97.5 F 4.8 96.8 F 2.8
91.6 F 5.2a 100.4 F 7.0a 99.6 F 4.0a 94.8 F 1.9a 104.5 F 9.4a 101.2 F 6.8a 97.5 F 5.6a 96.8 F 2.8a 112.5 F 9.9a 100.4 F 8.3a 101.8 F 3.3a 96.4 F 2.1a 98.0 F 9.4a 101.2 F 7.4a 98.6 F 3.8a 98.0 F 3.2a
10
15
20
Percent recovery of peroxidase
Means F S.D. (n = 6). a Nonsignificant difference. b Significant difference.
Using a Reaction Kinetics software (Biochrom), the initial linear slope of NADH oxidation was recorded. The molar extinction coefficient of NADH at 340 nm (6.22 103 M 1 cm 1) was used for calculating the enzyme units.
subtracting the slope of anaerobic assay from that of the aerobic assay and converted to enzyme units/ cuvette using the following formula:
2.2.2. Estimation of NADH peroxidase by the currently available assay (Thomas and Pera, 1983; Smart and Thomas, 1987; Shin and Park, 1997) H2O2 (1 mM) was incorporated in the reaction mixture given above and the assay was conducted under anaerobic conditions.
These recovered enzyme units were then compared to the actual NADH oxidase enzyme units introduced. NADH peroxidase was estimated by converting the slope of the anaerobic assay into enzyme units/cuvette by the abovementioned formula. This was compared with the number of NADH peroxidase units added. The percentage recovery was then calculated for both enzymes. For the anaerobic assay, the reactants were prepared in an anaerobic workstation (Coy, USA) containing 95% N2 and 5% H2, and kept in an anaerobic condition for 24 h prior to the assay. Nitrogen gas was bubbled through the reactants before the determination. Using a Clark-type oxygen electrode (Microelectrodes, USA), the dissolved oxygen in the reactants was ensured to be zero. No increase in oxygen was recorded within 5 min in the cuvette containing the anaerobic reaction mixture.
2.2.3. Estimation of NADH oxidase and NADH peroxidase by the improved assay The reaction mixture was the same as reported for estimating NADH peroxidase by the currently available assay except that the assay was conducted under both aerobic and anaerobic conditions. In both the currently available assay as well as in the improved assay, NADH oxidase was estimated by converting the slope of the aerobic assay into enzyme units/cuvette. In addition, a separate estimation of NADH oxidase was also performed by
Units=cuvette ¼ ðD A340 3Þ=6:22
336
A. Talwalkar et al. / Journal of Microbiological Methods 52 (2003) 333–339
Additionally, the respective blanks were performed before conducting the assays. The concentration of the reactants in the blanks was the same as that of the actual assay. The mean of six individual determinations was used for calculation and a Student’s t-test was performed (a = 0.05).
conditions when 1 mM H2O2 was incorporated in the free assay mix containing no NADH oxidase. In the absence of H2O2, no decrease in absorbance was observed in both aerobic and anaerobic assays.
2.3. Microorganisms, culture conditions, preparation of cell-free extract, and measurement of the slope of NADH oxidation
When the levels of NADH oxidase were determined from just the aerobic assay slope, all combinations of NADH oxidase with NADH peroxidase showed significantly greater ( p < 0.05) recovery levels of NADH oxidase than what was introduced in the cuvette. This was noted in both the currently available assay as well as in the improved assay. The recovery of NADH oxidase determined from the subtraction of the anaerobic assay slope from the aerobic assay slope is shown in Fig. 1. Although no significantly ( p>0.05) elevated recovery was detected, considerable variation was observed in the recovery of NADH oxidase by the currently available assay. When suspensions combining 5 U NADH oxidase with 15 and 20 U NADH peroxidase were assayed, the recovery obtained by subtracting the peroxidase slope from the oxidase slope gave negative values. In suspensions containing 10 U NADH oxidase and 15 U NADH peroxidase, the subtraction of the slopes resulted in a value of 4.18 U, whereas in suspensions containing 15 U NADH oxidase and 20 U NADH peroxidase, 11.01 U was obtained after calculation. For the abovementioned enzyme combinations, however, the improved assay suggested in this paper demonstrated no significant difference ( p>0.05) between the values of NADH oxidase introduced and that calculated from the slope of NADH oxidation. The percentage recovery for NADH oxidase determined by subtracting the anaerobic assay slope from the aerobic assay slope is listed in Table 1. In the currently available assay, the recovery of NADH oxidase changed with differing oxidase –peroxidase ratios. When the enzyme suspension contained lower amounts of NADH oxidase than NADH peroxidase units, the calculated recovery ranged from 73% to as low as 200%. For enzyme suspensions having equal or higher amounts of NADH oxidase than NADH peroxidase, the calculated recovery ranged from 86% to 102%. However, in the improved spectrophotometric assay, the percentage recovery for
Lactobacillus acidophilus 2400, L. acidophilus 2409, Bifidobacterium infantis 1912, B. lactis 1941, and B. pseudolongum 1944 were obtained from the Commonwealth Scientific Industrial Research Organization (CSIRO), Australia. B. longum 55815 was received from American Type Culture Collection (ATCC), USA. All strains were grown anaerobically for 24 h at 37 jC in 200 ml MRS broth (Oxoid, Australia). Cells were harvested by centrifugation for 10 min at 10,000 g at 4 jC and the cell pellet was washed thrice with 0.1 M phosphate buffer, pH 7. Cells were disrupted by passing them through a FrenchR Pressure Cell (Thermospectronic, USA) and cell-free extract was obtained by removing the cell debris by centrifugation for 15 min at 12,000 g at 4 jC. Pure enzymes in reaction system of all the abovementioned assays were replaced by an appropriate volume of cell-free extract and the slope of NADH oxidation was recorded. Previously boiled cell-free extract was used to negate the possibility of nonenzymatic oxidation of NADH.
3. Results 3.1. Results of assay blanks A blank containing NADH, FAD, H2O2, and buffer showed no decrease in absorbance over the time of the assay under both aerobic and anaerobic conditions. The activity of NADH oxidase alone under aerobic conditions was not affected in the presence of 1 mM H2O2. Under anaerobic conditions, however, no NADH oxidase activity was noticed. NADH peroxidase alone demonstrated the same activity under both aerobic and anaerobic
3.2. Recovery of NADH oxidase
A. Talwalkar et al. / Journal of Microbiological Methods 52 (2003) 333–339
NADH oxidase remained very high regardless of the proportion of NADH oxidase and NADH peroxidase units and ranged between 90% and 102% even at lower concentrations of oxidase. The means of the percentage recovery from the currently available assay and our improved assay were found to differ significantly ( p < 0.05) in enzyme suspensions where the amount of NADH oxidase units was less than that of NADH peroxidase units. In all the remaining enzyme suspensions where the proportion of NADH oxidase was either equal to or greater than NADH peroxidase, no significant difference ( p>0.05) was observed among the means of the two assays. 3.3. Recovery of NADH peroxidase It was interesting to note that in both assays, the calculated values of NADH peroxidase approximated the number of units of NADH peroxidase introduced in the cuvette. The anaerobic conditions of the assay and the abundance of substrate (H2O2) ensured maximum activity of NADH peroxidase. Consequently, the values obtained through calculation showed similarity with the actual peroxidase units introduced. The proportion of oxidase and peroxidase units in the various enzyme suspensions did not affect the recovery of NADH peroxidase. No significant difference ( p>0.05) was found between the means of the two assays for NADH peroxidase. The means ranged from 91% to 107% for the currently available assay and from 91% to 112% in the improved assay (Table 1). 3.4. Slope of NADH oxidation in cell-free extracts of lactic acid bacterial strains No oxidation of NADH was observed when boiled cell-free extract was used in the assays. Cell-free extracts of all six bacterial strains oxidised NADH when assayed under anaerobic conditions and in presence of H2O2 (Table 2). The slope of NADH oxidation by the currently available assay differed from that obtained by the improved assay. Negative values were observed in B. infantis 1912 and B. pseudolongum 1944 when the slope of NADH peroxidase assay was subtracted from the slope of NADH oxidase assay (currently available assay). With the NADH oxidase-improved assay, however,
337
Table 2 Differences in the estimation of NADH oxidases of six lactic acid bacteria strains by the currently available assay and the improved assay Strain
B. infantis 1912 B. lactis 1941 B. pseudolongum 1944 B. longum 55815 L. acidophilus 2400 L. acidophilus 2409
Slope of NADH oxidase assay (a)
Slope of NADH peroxidase assay (b)
Difference in slopes (a b)
CAAa
IAb
CAAa
IAb
0.10 0.15 0.12
0.22 0.15 0.24 0.11 0.22 0.13
0.05 0.07 0.04 0.13 0.01 0.09
0.16 0.70
0.21 0.10 1.10 0.52
0.06 0.11 0.18 0.58
0.39
0.60 0.30
0.09 0.30
Mean (n = 6). a CAA = currently available assay. The reaction system of cellfree extract, NADH (67 AM), FAD (67 AM) and Bis – Tris buffer (0.1 M), pH 6.0, in a total volume of 3 ml was assayed for 3 min at 30 jC under aerobic conditions. b IA = improved assay. The reaction system of cell-free extract, NADH (67 AM), FAD (67 AM), H2O2 (1 mM) and Bis – Tris buffer (0.1 M), pH 6.0, in a total volume of 3 ml was assayed for 3 min at 30 jC under aerobic conditions.
the difference in the slopes gave positive values for all the six bacterial strains.
4. Discussion The percentage recovery in the currently available assay was found to depend solely on the ratio of NADH oxidase and NADH peroxidase, and change as their proportions differed. Some researchers have estimated NADH oxidase from just the aerobic assay slope (de Vries and Stouthamer, 1969; Uesugi and Yajima, 1978; Shin and Park, 1997). The slope of NADH oxidation in the aerobic assay, however, is actually a sum of the total NADH oxidised by the activities of both oxidase and peroxidase. Enzyme units calculated from this slope would, therefore, result in elevated levels of NADH oxidase. This was confirmed by the significantly elevated recoveries of NADH oxidase obtained when its levels were determined by this method as also by elevated slopes of NADH oxidation by cell-free extracts of the six bacterial strains (Table 2). This, there-
338
A. Talwalkar et al. / Journal of Microbiological Methods 52 (2003) 333–339
fore, suggests that the reported values of NADH oxidase where levels were determined from just the slope of aerobic assay may have been overestimated. Smart and Thomas (1987) have reported that their amperometric estimation of NADH oxidase correlated well with that obtained from the subtraction of the anaerobic assay slope from the slope of the aerobic assay. This suggests that one can subtract the slope of peroxidase (anaerobic assay slope) from the oxidase – peroxidase slope (aerobic assay slope) to accurately determine the levels of NADH oxidase spectrophotometrically. The difference in the reaction velocities of NADH peroxidase in these two assays, however, can give rise to inaccurate estimations of NADH oxidase. This is confirmed in the negative recovery percentages of NADH oxidase obtained using the currently available assay (Fig. 1) and by the negative slope differences in some of the bacterial strains tested (Table 2). As against this, in our improved assay, the uniformity of the reactants in the aerobic and anaerobic assay ensured oxygen as the only variable-affecting enzyme activities between these two assays. This guaranteed accurate estimations of NADH oxidase when the slope of NADH peroxidase was subtracted from the slope of
the aerobic assay and was reflected in the high-percentage recoveries of NADH oxidase as well as NADH peroxidase in all the different enzyme proportions we tested (Table 1). This was further confirmed by positive slope differences in all the cell-free extracts assayed (Table 2). In many reports of NADH oxidases in lactic acid bacteria, the assay system used was based on the consumption of NADH; however, the end product was not measured. This does not distinguish between H2O- and H2O2-forming NADH oxidases. This is further complicated by the fact that the activity of an H2O2-forming NADH oxidase combined with that of an excess of NADH peroxidase is similar to an H2O-forming NADH oxidase (Condon, 1987). Although our improved assay is best suited for NADH – H2O2 oxidase/NADH peroxidase system, it can also help in distinguishing between NADH – H2O2 and NADH – H2O oxidases. This can be achieved by performing an additional aerobic assay without the addition of any H2O2 in the reaction system. If the slopes of the aerobic assay in the presence and absence of H2O2 are similar, then the enzyme in question is an NADH – H2O oxidase, regardless of the presence of
Fig. 1. Recovery of NADH oxidase in the presence of NADH peroxidase.
A. Talwalkar et al. / Journal of Microbiological Methods 52 (2003) 333–339
any peroxidase. If peroxidase is detected and the slope of the aerobic assay in the absence of H2O2 is less than in presence of H2O2, then it is an NADH – H2O2 oxidase. NADH peroxidase activity was detected in all the bacterial strains tested, and the slope of NADH oxidation in absence of H2O2 was less than in presence of H2O2 (Table 2). Accordingly, it can be concluded that all six strains possessed NADH – H2O2 oxidase. In lactic acid bacteria containing NADH oxidase and NADH peroxidase, the proportion of these two enzymes can vary from strain to strain. In this paper, 16 different proportions were tested and our improved assay was found to demonstrate high accuracy in the recovery of both NADH oxidase (especially low levels) and NADH peroxidase, regardless of the enzyme proportions. In comparison, the currently available assay was suitable only for determining individual levels of NADH peroxidase. When levels of NADH oxidase were low in comparison to NADH peroxidase, this assay gave inaccurate estimations of NADH oxidase. It is also clear that estimating the level of NADH oxidase from just the slope of the aerobic assay may lead to over estimation of the enzyme units. In addition, cell-free extracts of six lactic acid bacteria did not interfere with the measurement of the slope of NADH oxidation by our improved assay. The proposed assay reported in this paper, therefore, is a suitable technique for determining the individual levels of NADH peroxidase from a suspension containing NADH oxidase and NADH peroxidase in lactic acid bacteria. Acknowledgements This research was supported by the Australian Research Council (SPIRT) and the Dairy Farmers, Australia. References Anders, R.F., Hogg, D.M., Jago, G.R., 1970. Formation of hydrogen peroxide by group N streptococci and its effect on their growth and metabolism. Appl. Microbiol. 19, 608 – 612.
339
Carlsson, J., Iwami, Y., Yamada, T., 1983. Hydrogen peroxide excretion by oral streptococci and effect of lactoperoxidase – thiocyanate – hydrogen peroxide. Infect. Immun. 40, 70 – 80. Condon, S., 1987. Responses of lactic acid bacteria to oxygen. FEMS Microbiol. Rev. 46, 269 – 281. Cox, R.P., Marling, N., 1992. High-affinity oxygen uptake by Bifidobacterium bifidum. Antonie Van Leeuwenhoek 62, 291 – 297. de Vries, W., Stouthamer, A.H., 1969. Factors determining the degree of anaerobiosis of Bifidobacterium strains. Arch. Mikrobiol. 65, 275 – 287. Dolin, M.I., 1961. Cytochrome-independent electron transport enzymes of bacteria. In: Gunsalus, I.C., Stanier, R.Y. (Eds.), The Bacteria, vol. 2. Academic Press, New York, pp. 425 – 460. Higuchi, M., Shimada, M., Yamamoto, Y., Hayashi, T., Koga, T., Kamio, Y., 1993. Identification of two distinct NADH oxidases corresponding to H2O2-forming oxidase and H2O-forming oxidase induced in Streptococcus mutans. J. Gen. Microbiol. 139, 2343 – 2351. Higuchi, M., Yamamoto, Y., Kamio, Y., 2000. Molecular biology of oxygen tolerance in lactic acid bacteria: functions of NADH oxidases and Dpr in oxidative stress. J. Biosci. Bioeng. 90, 484 – 493. Marty-Teysset, C., de la Torre, F., Garel, J.-R., 2000. Increased production of hydrogen peroxide by Lactobacillus delbrueckii subsp. bulgaricus upon aeration: involvement of an NADH oxidase in oxidative stress. Appl. Environ. Microbiol. 66, 262 – 267. Schmidt, H.-L., Sto¨cklein, W., Danzer, J., Kirch, P., Limbach, B., 1986. Isolation and properties of an H2O-forming NADH oxidase from Streptococcus faecalis. Eur. J. Biochem. 156, 149 – 155. Shimamura, S., Abe, F., Ishibashi, N., Miyakawa, H., Yaeshima, T.A., Tomita, M., 1992. Relationship between oxygen sensitivity and oxygen metabolism of Bifidobacterium species. J. Dairy Sci. 75, 3296 – 3306. Shin, S.-Y., Park, J.-H., 1997. Activities of oxidative enzymes related with oxygen tolerance in Bifidobacterium sp. J. Microbiol. Biotechnol. 7, 356 – 359. Smart, J.B., Thomas, T.D., 1987. Effect of oxygen on lactose metabolism in lactic streptococci. Appl. Environ. Microbiol. 53, 533 – 541. Thomas, E.L., Pera, K.A., 1983. Oxygen metabolism of Streptococcus mutans: uptake of oxygen and release of superoxide and hydrogen peroxide. J. Bacteriol. 154, 1236 – 1244. Uesugi, I., Yajima, M., 1978. Oxygen and ‘‘strictly anaerobic’’ intestinal bacteria: 2. Oxygen metabolism in strictly anaerobic bacteria. Z. Allg. Mikrobiol. 18, 593 – 601. Yi, X., Kot, E., Bezkorovainy, A., 1998. Properties of NADH oxidase from Lactobacillus delbrueckii ssp. bulgaricus. J. Sci. Food Agric. 78, 527 – 534.