Although it is known that a part of lactic acid bacteria can produce carotenoid, little is known about the regulation of carotenoid production. The objective of this study was to determine whether aerobic growth condition influences carotenoid production in carotenoid-producing Enterococcus gilvus. Enterococcus gilvus was grown under aerobic and anaerobic conditions. Its growth was slower under aerobic than under anaerobic conditions. The decrease in pH levels and production of lactic acid were also lower under aerobic than under anaerobic conditions. In contrast, the amount of carotenoid pigments produced by E. gilvus was significantly higher under aerobic than under anaerobic conditions. Further, real-time quantitative reverse transcription PCR revealed that the expression level of carotenoid biosynthesis genes crtN and crtM when E. gilvus was grown under aerobic conditions was 2.55–5.86-fold higher than when it was grown under anaerobic conditions. Moreover, after exposure to 16- and 32-mM H2O2, the survival rate of E. gilvus grown under aerobic conditions was 61.5- and 72.5-fold higher, respectively, than when it was grown under anaerobic conditions. Aerobic growth conditions significantly induced carotenoid production and the expression of carotenoid biosynthesis genes in E. gilvus, resulting in increased oxidative stress tolerance.
Oxygen and byproducts such as reactive oxygen species (ROS), which are generated during metabolism, can damage the cells of lactic acid bacteria (LAB) and cause a loss of LAB activity (Miyoshi et al., 2003). To avoid cell damage due to oxidative stress, LAB have developed various antioxidant mechanisms. These mechanisms have been studied for a long time (van de Guchte et al., 2002; Miyoshi et al., 2003) as they are important for improving the tolerance to oxidative stress in LAB. To quench ROS, numerous genes, such as trxB1 (thioredoxin reductase), ahpC (Alkyl hydroperoxide reductase), gshR (glutathione reductase) and sodA (superoxide dismutase), are associated with antioxidative activity in LAB (Serrano et al., 2007; Pedersen et al., 2008). In addition, cytochrome oxidase (cyd) in Lactococcus lactis has been reported to alleviate oxidative stress during respiratory growth with hemin (Pedersen et al., 2008).
Carotenoids, such as β-carotene and zeaxanthin, can act as antioxidants because of their unsaturated double bonds (Young & Lowe, 2001). A wide range of organisms, such as green plants, algae, fungi, and bacteria, can produce carotenoids (Nelis & De Leenheer, 1991). LAB belonging to Lactobacillus and Enterococcus, which comprise nonphotosynthetic bacteria, can produce the C30 carotenoid 4,4'-diaponeurosporene (Taylor & Davies, 1974; Marshall & Wilmoth, 1981; Takaichi et al., 1997; Garrido-Fernández et al., 2010). LAB have carotenoid biosynthesis genes, such as crtN and crtM, which are widely conserved in carotenoid-producing LAB. crtN and crtM encode squalene/dehydrosqualene desaturase and phytoene/dehydrosqualene synthase, respectively; they are the key enzymes responsible for carotenoid biosynthesis (Umeno et al., 2002).
In our previous study, crtN and crtM were cloned from Enterococcus gilvus and expressed in L. lactis MG1363, resulting in the production of 4,4'-diaponeurosporene. Furthermore, we found that carotenoid production significantly increased the tolerance to oxidative stress in L. lactis MG1363 (Hagi et al., 2013). These results indicated that 4,4'-diaponeurosporene in LAB is related to a stress tolerance mechanism in common with other antioxidants, such as superoxide dismutase and glutathione reductase, implying that carotenoid production may be regulated by stress responses. However, not much is known about stress-induced carotenogenesis in LAB.
Carotenoid production can be induced by environmental stresses. In algae, oxidative stress induces the production of astaxanthin (Kobayashi et al., 1993). In case of a nonphotosynthetic bacterium, Thermus thermophilus, Takano et al. (2011) reported that light could induce the production of thermozeaxanthin, which contributed to a DNA repair system. Induced carotenoid biosynthesis confers protection against environmental stresses, such as oxygen or light stress. For purposes of commercial carotenoid production, the mechanisms underlying carotenogenesis induced by environmental factors, such as temperature, metals, and salts, have been investigated (Bhosale, 2004). Further, it has been indicated that carotenoid-producing LAB can serve as a reliable source of food and antioxidants (Garrido-Fernández et al., 2010).
Determining the relationship between carotenoid production and oxidative stress is important as it can provide novel insights into LAB physiology and opportunities to improve oxidative stress tolerance and carotenoid production efficiency. In this study, to determine whether carotenoid production can be induced by oxidative stress, carotenoid production and the expression of carotenoid biosynthesis genes were compared between E. gilvus grown under aerobic and anaerobic culture conditions. We also assessed the survival rate of E. gilvus grown under aerobic and anaerobic conditions after H2O2 exposure.
Materials and methods
Bacteria and growth conditions
Enterococcus gilvus CR1 (AB742448), isolated from raw milk, was used in this study. To prepare a preculture, E. gilvus was anaerobically incubated at 30 °C for 24 h in M17 medium (Difco Laboratories, Detroit, MI) supplemented with 0.5% glucose (GM17 medium). This preculture was inoculated into 200 mL of GM17 medium (0.5%, v/v) in an Erlenmeyer flask (500 mL). For aerobic conditions, an inoculated Erlenmeyer flask was shaken at 110 r.p.m. at 30 °C. For anaerobic conditions, the culture was statically incubated at 30 °C in an AnaeroPack system (AnaeroPack, Mitsubishi Gas Chemical, Tokyo, Japan). Dissolved oxygen was measured using a PSt3 oxygen meter equipped with a Fibox 3 trace fiber optic device (PreSens, Regensburg, Germany). Bacterial growth (OD600 nm) and culture pH were monitored using a UV mini-1240 UV-VIS Spectrophotometer (Shimadzu, Japan) and a SevenEasy pH metre (Mettler Toledo, Switzerland), respectively. The concentrations of lactic acid and glucose in the cultures were determined, respectively, using F kit for dl-lactic acid (Roche Diagnostics, Tokyo, Japan) and a Glucose CII test kit (Wako Pure Chemical Industries, Osaka, Japan).
The yellow pigments of E. gilvus were extracted and purified as previously described (Hagi et al., 2013). In brief, the cultures of E. gilvus incubated under aerobic or anaerobic conditions were harvested every 2 h. After centrifugation, the bacterial precipitates were washed twice with saline. These washed bacterial cells were extracted with methanol. An equal volume of hexane and a half-volume of distilled water were added to the methanol extract that contained the yellow pigments. After centrifugation at 1500 g for 5 min, the organic phase that contained carotenoids was collected. After evaporation of this organic phase under N2 gas, carotenoids were resuspended with 1 mL of petroleum ether (cat. no. 26619-75, Nacalai tesque, Kyoto, Japan). The amount of pigment in an extract was determined from the maximal A470 nm using a spectrophotometer. After separating the yellow pigments on TLC silica gel 60 plates (Merck, Darmstadt, Germany) with petroleum ether and acetone (9 : 1, v/v), the spectra of purified pigments in petroleum ether were determined with a UV–VIS spectrophotometer.
Evaluating crtNM expression by real-time quantitative reverse transcription PCR (qRT-PCR)
Total RNA was extracted from harvested E. gilvus cells that had been grown aerobically and anaerobically using a RNeasy Protect Bacteria Mini Kit (QIAGEN Valencia, CA) according to the manufacturer's instructions. Total RNA was extracted from three independent cell cultures of E. gilvus in each condition. To completely eliminate DNA, we used DNase treatment with an RNase-Free DNase Set (QIAGEN). Reverse transcription PCR was carried out using a PrimeScript® II 1st strand cDNA Synthesis Kit (TAKARA, Otsu, Japan). Denaturation used the C1000 thermal cycler (BIO-RAD, Hercules) at 65 °C for 5 min in a 10 μL solution containing 1 μL of random 6 mers, 1 μL of a dNTP mixture, 200 ng of total RNA, and sterile distilled water up to 10 μL. Then, 10 μL of RT-PCR mixtures containing 4 μL of 5× PrimeScript II buffer, 0.5 μL of RNase inhibitor, 1 μL of PrimeScript II RTase, and 4.5 μL of sterile distilled water was added to the denatured samples. Reverse transcription PCR for a total volume of 20 μL of these mixtures used the C1000 thermal cycler with the following conditions: initial 10 min pretreatment at 30 °C, latter 42 °C for 60 min, and finally 72 °C for 15 min. The resulting cDNA solutions were diluted with 60 μL of sterile distilled water and used for templates for real-time PCR experiments.
Real-time qPCR was performed using THUNDERBIRD SYBR qPCR Mix (TOYOBO, Osaka, Japan) and the C1000 Thermal Cycler CFX96 Real-Time System (BIO-RAD, Hercules). To specifically amplify the crtN and crtM genes of E. gilvus, we used the primer set crtNqF1 5′-GTAGAGGCGTTGCATTCGATTC-3′/crtNqR1 3′-TCTGGTACAGGCACTAACGCAT-5′ (amplicon length: 135 bp), and crtMqF1 5′-GTAATGCGTGACCAATTGGA-3′/crtMqR1 5′-GCCAATTTGCAGAGAGAATG-3′ (amplicon length: 227 bp). For both assays, 1 μL of diluted cDNA solution was added to 19 μL of PCR mixture containing 10 μL of THUNDERBIRD SYBR qPCR Mix, 1.2 μL of each primer (10 pmol), and 6.6 μL of sterile distilled water. The thermal cycling conditions were as follows: 95 °C for 10 s, followed by 40 cycles of 95 °C for 5 s, and 60 °C for 30 s. The melting program was 95 °C for 10 s, 65 °C for 5 s, and 95 °C for 50 s. For calibration, a serial dilution series (109–105 copies per reaction) of pRC expression plasmid with full-length crtN and crtM genes of E. gilvus, constructed previously (Hagi et al., 2013), was used as the standard template. mRNA levels were normalized to 16S rRNA gene transcript levels as reported previously (Tasara & Stephan, 2007; Chen et al., 2010). To amplify the 16S rRNA gene, we used the primer set 341F 5′-CCTACGGGAGGCAGCAG-3′ and 518R 5′-ATTACCGCGGCTGCTGG-3′ (amplicon length: 161 bp). For calibration, we prepared serial dilutions for a pGEM-T Easy vector (105–109 copies per reaction; Promega, Madison, WI) containing the partial length of the 16S rRNA gene of E. gilvus (AB742448; designated pGCR1) and used this as a standard template. To maintain sample copy numbers within calibration values, 10-fold diluted cDNA solutions were used to amplify the 16S rRNA gene.
Standard curves (slope, efficiency = 10−1/slope−1, y-intercept, correlation coefficient = r2) and the gene expression levels in E. gilvus were analyzed using cfx manager software, version 2.1. Real-time qPCR was run in triplicate wells for each sample. To ensure that DNA was removed after RNA extraction, the total RNA samples untreated with reverse transcriptase were also used as templates for real-time PCR with each primer set.
H2O2 treatment test was performed as previously reported with some modification (Hagi et al., 2013). Enterococcus gilvus cultures (1 mL each; OD600 nm = c. 1.0) grown aerobically and anaerobically were harvested. After centrifugation at 9100 g for 5 min, bacterial pellets were washed with saline. These washed bacterial pellets were resuspended in 500 μL of saline containing H2O2 (16 and 32 mM) to an OD of 2.0 at A600 nm. The bacterial suspension for each stress condition was incubated for 90 min at 37 °C, after which the bacterial suspension was centrifuged and washed twice with saline. These washed bacterial pellets were resuspended in 500 μL of saline, and then, the serial dilutions of the bacterial suspensions were prepared and plated onto GM17 agar plates to determine the number of viable bacteria after exposure to H2O2. The survival rate was determined as the number of bacteria after exposure to stress compared with the number of bacteria incubated in normal saline for 90 min at 37 °C. Three independent tests were performed in duplicate.
Student's t-test was performed to determine statistical differences.
Growth and carotenoid production of Enterococcus gilvus under aerobic and anaerobic conditions
After incubation for 10 h, the optical densities of E. gilvus cultures grown under anaerobic and aerobic conditions were 3.2 and 2.7, respectively. The concentration of cells (OD600 nm) in E. gilvus under aerobic was lower than that under anaerobic conditions after 10 h of incubation (P < 0.05; Fig. 1a). In addition, the culture pH levels after incubation for 10 h under anaerobic and aerobic conditions were 5.6 and 5.8, respectively.
The effects of different growth conditions on glucose consumption and lactic acid production were examined (Fig. 1b). The decline in the glucose concentration under anaerobic conditions was higher than under aerobic conditions. Lactic acid concentrations under anaerobic conditions were also higher than under aerobic conditions. l-lactic acid was primarily produced by E. gilvus under both aerobic and anaerobic conditions. The oxygen concentrations for aerobic and anaerobic culture were 5.2–7.3 and 0.01–0.07 mg mL−1, respectively (data not shown).
Carotenoid production and crtNM expression levels under aerobic and anaerobic conditions
Figure 2a shows the yellow pigmentation of E. gilvus when grown under anaerobic and aerobic conditions. Pigmentation levels were determined from A470 nm, which was normalized to the OD600 nm values for bacterial growth as previously reported (Umeno et al., 2002). It is apparent that E. gilvus yellow pigmentation significantly increased under aerobic conditions. Figure 2b shows the time course of carotenoid production under anaerobic and aerobic conditions. Under aerobic conditions, carotenoid production considerably increased after incubation for 2 h. At the same time point, carotenoid production level under anaerobic conditions was much lower, which subsequently only increased slightly. Carotenoid production was the highest at 4 h; at this time point, the amount of carotenoids produced under aerobic conditions was 22-fold higher than that under anaerobic conditions. The absorption maxima of purified pigments extracted from E. gilvus grown under aerobic condition for 10 h were at 412, 434.5, and 464 nm; these values are identical to those previously reported for 4,4'-diaponeurosporene (412, 434.5 and 464 nm) (Hagi et al., 2013).
To determine whether increased carotenoid production was related to the increased expression of two carotenoid biosynthesis genes, crtN and crtM (AB742449), the mRNA expression levels of these genes were monitored using qRT-PCR, and 16S rRNA gene transcript level was also monitored as reference gene for normalization. Before assessing the expression levels of crtN and crtM and the 16S rRNA gene, we checked the PCR assay performance. The resulting standard curves with 10-fold serial dilutions of control plasmid DNA were as follows: target gene crtN, slope = −3.595, y-intercept = 41.155, efficiency = 89.8%, r2 = 0.999; crtM, slope = −3.572, y-intercept = 40.798, efficiency = 90.5%, r2 = 0.998; the 16S rRNA gene, slope = −3.551, y-intercept = 41.186, efficiency = 91.2%, r2 = 0.994. All qRT-PCR amplifications with each primer set had a single peak in the melt curve and a detection limit of 1000 copies of target DNA per reaction. Regarding DNA contamination after RNA extraction and DNase treatment, quantitative PCR with each primer set using an RNA template did not cross threshold levels on reaching 40 cycles.
Table 1 shows the qRT-PCR results. The copy numbers for each gene (crtN, crtM, and the 16S rRNA gene) per μL of cDNA template were 1.66 ± 0.27 (means ± SD) × 106–5.33 ± 0.39 × 106, 7.25 ± 1.52 × 105–2.38 ± 0.33 × 106, and 9.59 ± 4.21 × 109–1.40 ± 0.19 × 1010 under anaerobic conditions, respectively, and 8.69 ± 0.74 × 106–1.40 ± 0.08 × 107, 3.70 ± 0.65 × 106–6.06 ± 0.62 × 106, and 9.61 ± 1.21 × 109–1.23 ± 0.36 × 1010 under aerobic conditions, respectively. No significant differences were observed in the amount of 16S rRNA gene transcript level when RT-qPCR was performed with equal amount of total RNA. This result indicates that expression of 16S rRNA gene in E. gilvus is insusceptible to culture condition, and 16S rRNA gene transcript is appropriate as a reference in this study. After normalizing carotenoid biosynthesis gene levels against the 16S rRNA gene levels, the expression levels for crtN and crtM were 1.26 ± 0.21 × 10−4–4.98 ± 1.47 × 10−4 and 5.49 ± 2.28 × 10−5–2.22 ± 1.07 × 10−4 under anaerobic conditions, respectively, and 7.36 ± 0.16 × 10−4–1.32 ± 0.84 × 10−3 and 3.14 ± 0.73 × 10−4–5.72 ± 3.82 × 10−4 under aerobic conditions, respectively. To compare gene expression levels under aerobic conditions with those under anaerobic conditions, normalized carotenoid gene expression levels under aerobic conditions were divided by those under anaerobic conditions. crtN expression levels in E. gilvus grown under aerobic conditions were 2.65–5.86-fold greater than those under anaerobic conditions. crtM expression levels showed the same trend (2.55–5.71-fold). The expression levels of these genes were significantly greater when E. gilvus was grown under aerobic than under anaerobic conditions (P < 0.05).
Table 1. Comparisons of crtN and crtM expression levels under anaerobic and aerobic conditions using qRT-PCR
Ratios (± SD) of relative crtN and crtM expression levels
5.86 ± 1.94
3.69 ± 0.96
2.65 ± 0.22
3.98 ± 1.45
3.44 ± 1.33
5.71 ± 2.61
3.39 ± 1.18
2.55 ± 0.53
3.81 ± 1.15
2.89 ± 0.16
Survival rates after H2O2 exposure
We previously reported that carotenoid production due to the expression of carotenoid biosynthesis genes (crtN and crtM) derived from E. gilvus significantly increased stress tolerance against H2O2 by L. lactis (Hagi et al., 2013). Because carotenoid production by E. gilvus was better under aerobic conditions, we tested whether carotenoid production under aerobic conditions increased stress tolerance against H2O2 in E. gilvus. Enterococcus gilvus cultures incubated under aerobic and anaerobic conditions were harvested during the exponential phase at an OD600 nm of c. 1.0, followed by the suspension of washed cells in H2O2. After incubation for 90 min at 37 °C, the survivability of H2O2-treated cells was determined. After exposure to 16- and 32-mM H2O2, the survival rate of E. gilvus grown under aerobic conditions was 16.6% and 0.0029%, respectively, whereas the survival rate of E. gilvus grown under anaerobic conditions was 0.27% and 0.00004%, respectively. The survival rate of E. gilvus grown under aerobic conditions was 61.5- (16 mM) and 72.5-fold (32 mM) higher than when it was grown under anaerobic conditions (Fig. 3); these differences in the survival rate were statistically significant (P < 0.05).
Carotenoids can act as antioxidants (Young & Lowe, 2001) and their production can be induced by oxidative- or light-induced stress (Kobayashi et al., 1993; Takano et al., 2011). In LAB, although several antioxidants are induced by oxidative stress (Pedersen et al., 2008), not much is known regarding the relationship between carotenoid production and stress responses. In the present study, to determine whether carotenoid production in LAB could be induced by stress, we examined the effects of oxidative stress on carotenoid production in E. gilvus.
The concentrations of cells (OD600 nm) in E. gilvus under aerobic were lower than under anaerobic conditions after 10 h of incubation. In contrast, aerobic conditions significantly increased carotenoid production and the expression levels of carotenoid biosynthesis genes in E. gilvus. Photosynthetic organisms such as plants and photosynthetic bacteria are exposed to oxidative stress during photosynthesis. To alleviate oxidative stress, photosynthetic organisms produce antioxidative enzymes, such as catalase and carotenoids (nonenzymatic antioxidants) in response to photo-oxidative stress. There have been several reports on the relationship between carotenoid production and photo-oxidative stress.
Bouvier et al. (1998) reported that the expression of carotenoid genes was induced by oxidative stress in Capsicum annuum. In Rhodobacter, a photosynthetic bacterium, carotenoid production was modulated by photo-oxidative stress (Ziegelhoffer & Donohue, 2009). In case of nonphotosynthetic bacteria, such as T. thermophilus and Streptomyces coelicolor, carotenoid production could be induced only by light (Takano et al., 2005, 2011). These results indicate that carotenoid production is strongly related to light exposure. However, our results revealed that carotenoid production and the expression of carotenoid biosynthesis genes in E. gilvus were significantly increased on exposure to oxygen under dark conditions. Thus, our results provide novel information on carotenogenesis in organisms without exposure to light.
In addition, we found that the survival rate of E. gilvus when grown under aerobic conditions was significantly higher than under anaerobic conditions when exposed to H2O2. We suggest that inducing carotenoid production upon oxygen exposure is one of the stress tolerance mechanisms in LAB. In LAB, production of some antioxidants has been reported to be induced by aerobic conditions, resulting in oxidative stress (Miyoshi et al., 2003; Serrano et al., 2007; Pedersen et al., 2008). Therefore, both some antioxidants and carotenoids may influence increased oxidative stress tolerance in E. gilvus.
There are various transcriptional regulators in LAB (Ravcheev et al., 2013). An SOS response or oxygen-regulated factors may affect the expression of carotenoid biosynthesis genes. In contrast, our results revealed that the concentration of lactic acid in the culture under aerobic condition was significantly lower than that under anaerobic condition after 10 h of incubation. The conversion rate of glucose to lactic acid in E. gilvus under aerobic condition (44.2%) was significantly lower than that under anaerobic condition (59.1%; P < 0.05). This indicates altered glucose metabolism under aerobic conditions, as observed with LAB (Nordkvist et al., 2003; Quatravaux et al., 2006). C30 carotenoid is synthesized from isoprene, which is derived from acetyl-CoA through the mevalonate pathway (Garrido-Fernández et al., 2010). Acetyl-coA production is influenced by glucose metabolism (Wilding et al., 2000); thus, altered carotenoid production may be related to a change in glucose metabolism caused by carbon catabolite regulators, such as CcpA, which are associated with oxygen (Gaudu et al., 2003).
In E. gilvus, carotenoid production was induced under aerobic conditions. However, carotenoid production is not always regulated by aerobic conditions in carotenoid-producing LAB, because Garrido-Fernández et al. (2010) reported that carotenoid production levels were different among strains of Lactobacillus plantarum when grown under static conditions. The activity of a promoter located up-stream of carotenoid biosynthesis genes may also influence carotenoid production in LAB.
In conclusion, aerobic culture conditions increased carotenoid production and contributed to conferring oxidative stress tolerance in E. gilvus. These results provide a novel insight into the oxidative stress responses of carotenoid-producing microorganisms, including LAB.
The authors thank Dr M. Waki and Dr M. Motoyama (NARO Institute of Livestock and Grassland Science) for help with measurements of dissolved oxygen. The authors would like to thank Enago (www.enago.jp) for the English language review. No author of this work has any conflict of interest.