We previously reported the aerotolerant nature of L. panis PM1 and its ability to use glycerol as the means of NADH recycling in the absence of oxygen (Khan et al. 2013). However, the presence of oxygen prevented 1,3-PDO formation and thus markedly-affected NADH recycling in this strain. In this study, the influence of oxygen on NADH recycling system and the oxidative stress resistance mechanism in its aerotolerance was investigated. Moreover, the metabolic profile was further investigated to understand how oxidative stress resistance mechanisms of L. panis PM1 influenced the profile of metabolic end-products.
During aerobic culture, L. panis PM1 prematurely entered into a stationary phase without depleting glucose (Figure 1a). This early entry into stationary phase was also associated with a ten-fold higher accumulation of H2O2 compared with microaerobic culture (Figure 2a). Therefore, the accumulation of H2O2 in aerobic culture was an apparent reason for the early cessation of growth. Anaerobic metabolism theoretically makes one ethanol per every glucose consumed, but the presence of oxygen altered this pattern to less than 1:1 ratio. These observations suggested that H2O2 could be a main end-product of an alternate pathway for NADH recycling under aerobic conditions, and that this could compete with NAD+-regeneration through ethanol production.
The production of H2O2 by LAB grown under aerobic conditions is commonly the result of flavoprotein oxidases, including NADH oxidase, pyruvate oxidase, α-glycerophosphate oxidase, and superoxide dismutase (Condon 1987). However, candidate genes for these enzymes were not found in the draft genome data of L. panis PM1, with the exception of NADH oxidase. Pyruvate oxidase has been documented in a few species of lactobacilli and is known to convert pyruvate to CO2 and acetyl phosphate, along with the formation H2O2 (Condon 1987). Pyruvate oxidase has its highest activity during the early stationary phase of growth and is induced and repressed by oxygen and glucose, respectively, in L. plantarum (Saxena et al. 2009; Veiga-da-Cunha and Foster 1992). However, the presence of pyruvate oxidase does not adequately explain the early entry into stationary phase observed during the aerobic culture of L. panis PM1. Our results showed that most of pyruvate produced during glucose consumption was used to produce lactate in aerobic culture (Figure 1a), indicating that pyruvate oxidase apparently removed little pyruvate from this pathway. NADH oxidase is the most common enzyme responsible for the production of H2O2 from oxygen and is highly-active in LAB (Condon 1987; Higuchi et al. 2000; Tachon et al. 2011). LAB are known to possess either a NADH: H2O2 or a NADH: H2O oxidase, or sometimes both (Condon 1987; Higuchi et al. 2000). Final products of the reaction of NADH oxidase include either NAD+ and H2O2, or NAD+ and H2O, depending on whether two- or four-electrons are transferred by NADH: H2O2 oxidase or NADH: H2O oxidase (Condon 1987; Higuchi et al. 2000; Miyoshi et al. 2003). Our results showed that the crude extract from L. panis PM1 cultured under aerobic and microaerobic conditions could directly produce H2O2 using oxygen as a substrate, and the activity of the enzyme was found to increase with the addition of FAD+ as well as aeration of the assay mixture (approximately 1.5 fold). These results indicated that the NADH oxidase in L. panis PM1 was a NADH: H2O2 oxidase and a flavoprotein-like NADH oxidase, as seen in other gram-positive bacteria (Komagata 1996; Marty-Teysset et al. 2000; Tachon et al. 2011).
Most LAB can respond (and protect themselves) to high concentrations of H2O2 produced through their oxidase enzymes during sugar fermentation (Higuchi et al. 2000). In fact, most LAB possess NADH peroxidase or pseudocatalase, and superoxide dismutase exists in some LAB (Condon 1987). These enzymes can enable LAB to overcome otherwise-lethal concentrations of hydrogen peroxide. The annotation data of the L. panis PM1 genome sequence and the results of the enzyme assays of NADH oxidase and NADH peroxidase suggest that these enzymes are main factors in oxidative stress resistance. The levels of accumulated H2O2 in the culture media could be accounted for by the differences in the activities of NADH peoxidase and NADH oxidase. Our qRT-PCR analyses showed that oxygen did not regulate nox and npx at the transcriptional-level, and mainly affected enzyme activities in L. panis PM1 (Table 2). While transcription levels were similar, activity assays exhibited that NADH peroxidase was positively-activated by oxygen but required a long induction time to express activity contrary to NADH oxidase. The oxygen-availability analyses indicated that higher oxygen availability in the 3-ml culture could provide higher amounts of substrate (oxygen) for NADH oxidase, resulting in greater accumulation of H2O2 in the first 12 hours. In the subsequent 12 hours, the accumulated H2O2 was decomposed by NADH peroxidase activity. The degree of degradation of H2O2 was dependent on NADH peroxidase activity, and the amount of activity was in proportion with oxygen availability (Figure 3 and Table 3). Therefore, we concluded that a coupled NADH oxidase - NADH peroxidase system, regulated by oxygen availability, was a key oxidative stress resistance mechanism in L. panis PM1.
Accumulation of H2O2 by NADH oxidase has been reported in group I homofermentative lactobacilli, like L. delbrueckii, where approximate 97% of NADH was reoxidized by lactate dehydrogenase and NADH oxidase accounted for only 3% of NADH reoxidation (Marty-Teysset et al. 2000). Thus, NADH recycling in group I LAB depends on a pyruvate supply from glycolysis, rather than oxygen. Unlike homofermentative lactobacilli, the presence of electron acceptors, such as oxygen, citric acid, or glycerol, directly influenced the flux of NADH reoxidation in L. panis PM1. In our other studies, when L. panis PM1 was cultured in mMRS containing citric acid (24 mM) and glycerol (150 mM) under microaerobic conditions, the major changes in end-product formation included a decrease in ethanol, an increase in acetic acid, and the production of succinic acid (19 mM) and 1,3-PDO (88 mM), respectively (unpublished data). The results of HPLC analyses in the present study showed that aerobic conditions negatively-affected the production of ethanol relative to glucose consumption, regardless of the presence of electron acceptors (Figures 1a and 4). Also, when L. panis PM1 was cultured under aerobic conditions in mMRS containing citric acid and glycerol, oxygen was used as the preferred electron acceptor, resulting in a shift of NADH flux along with a significant decrease of the production of succinic acid (4 mM) and 1,3-PDO (7 mM) (Figure 4). This data indicated that the activity of NADH oxidase was a key mechanism for the reoxidation of NADH during growth in aerobic culture.
In addition to oxidative stress responses, NADH oxidase can also help L. panis PM1 use oxygen during energy metabolism, directly. That is, the shift of NADH recycling with molecular oxygen redirected acetyl phosphate, which normally would be used to produce ethanol, to the formation of acetic acid. This acetic acid production via acetate kinase can stoichiometrically generate ATP (Condon 1987). Thus, O2-directed NADH recycling should be advantageous with respect to energy metabolism. However, regeneration of NAD+ via NADH oxidase in L. panis PM1 led to overproduction of H2O2 with subsequent negative effects on growth and end-product formation. Our findings indicate that varied oxygen availabilities of culture environments would greatly affect energy metabolism as well as oxidative stress of L. panis PM1. The formation of 1,3-PDO is a main route for NADH reoxidation in the presence of glycerol under anaerobic conditions; whereas, under aerobic conditions, NADH recycling largely occurs through NADH oxidase activity. The present study indicates that energy metabolism via the NADH oxidase system explains why L. panis PM1 fails to produce 1,3-PDO under aerobic conditions.