2016 Volume 22 Issue 5 Pages 673-678
Tofu-misozuke is a traditional fermented food in Japan, and contains biogenic amines that function as neurotransmitters, but are toxic when consumed at high levels. Analysis of the amine production of bacterial communities is important for controlling the amine content in fermented foods. We investigated the bacterial community and biogenic amine content of tofu-misozuke. The lactic acid bacteria (LAB) ratio in the community and tyramine content were highly related. LAB were isolated to identify the biogenic amine-producing species. Isolated strains from tofu-misozuke were identified, and their tyramine formation potential was evaluated by PCR detection of the tyrosine decarboxylase gene and by cultivation tests. We detected three tyramine-producing bacteria, Enterococcus faecium, Weissella viridescens, and Lactobacillus curvatus, in tofu-misozuke.
Fermented soybean products, such as miso, soy sauce, natto, tempeh, and fermented tofu, are common in Asian households (Chen et al., 2012). Tofu-misozuke is a traditional fermented soybean product originating in Kumamoto, Japan. It is made by soaking tofu with miso paste, and has a flavor and texture similar to cheese. Recently, a new type of tofu-misozuke was produced by soaking in liquid seasoning containing koji (Aspergillus oryzae) during the manufacturing process. The flavor and taste of this novel tofu-misozuke is different from that traditionally produced. Fungal proteases involved in protein degradation and peptide production generate the unique flavors and textures associated with tofu-misozuke (Funaki et al., 1996, Funaki et al., 1997). In miso paste, a raw material of tofu-misozuke, lactic acid bacteria (LAB) and bacteria of the genus Bacillus have been observed (Kim et al., 2010). LAB are the most common bacteria isolated from traditional Asian fermented foods (Murooka and Yamashita, 2008). It is important to analyze the microbial communities in fermented foods to effectively regulate flavor and taste.
However, the bacterial community in tofu-misozuke has not yet been characterized.
Biogenic amines (BAs) are low-molecular weight nitrogenous compounds of biological importance in microorganisms, plants, and animals (Bai et al., 2013). They are mainly formed by decarboxylation of amino acids via substrate-specific decarboxylases derived from microbes in foods (Santos, 1996). The most well known BAs, such as serotonin, dopamine, noradrenaline, and histamine, are neurotransmitters and/or allergens. Those less well known include tyramine, tryptamine, and β-phenylethylamine, and are abundant in fermented foods. Furthermore, some fermented foods contain putrescine, which belongs to a group of ubiquitous polycationic amines (Shukla et al., 2011). Putrescine can be synthesized either directly from ornithine by ornithine decarboxylase (ODC) or indirectly from arginine via arginine decarboxylase (ADC) (Costantini, 2013, Pereira et al., 2009). ADC converts arginine to agmatine, then agmatine deiminase (AgDI) biosynthetically converts agmatine to putrescine via the AgDI pathway (Nakada and Itoh, 2003). Therefore, arginine contained in soybeans can be a precursor of putrescine (Chen et al., 2012). The consumption of foods containing high amounts of BAs may have toxicological effects, as observed in a number of food-related incidents. In particular, tyramine consumption causes headaches, heart palpitations, hyper/hypotension, and several allergic disorders (Kim and Kim, 2014). Therefore, understanding the relationship between BA production and the bacterial community in fermented foods is important to improve the safety of the manufacturing process.
The objective of this study was to identify the bacterial community in tofu-misozuke and to evaluate the tyramine-producing ability of LAB isolated from tofu-misozuke.
Samples Five tofu-misozuke samples (A to E) were purchased from companies in Kumamoto Prefecture, Japan. Samples A and E were produced by the traditional handmade method of soaking in miso paste, while the others were produced by soaking in liquid seasoning.
Chemicals Peptone and tryptic soy broth (TSB) were purchased from Difco (Sparks, MD, USA). Meat extract was purchased from Merck KGaA, Darmstadt, Germany. Yeast extract and casein were purchased from Nacalai Tesque, Kyoto, Japan. The other chemicals were purchased from Nacalai Tesque and Wako Pure Chemical Industries, Osaka, Japan.
Identification of the bacterial community of tofu-misozuke Twenty-five grams of each sample was homogenized in 225 mL of sterilized saline. The homogenized solution was centrifuged at 8,000 × g at 4°C for 15 min. The precipitate was suspended in sterilized water and the suspension was centrifuged again under the same conditions. The precipitate was then suspended in 200 µL of sterilized water. DNA was extracted from 50 µL of each sample solution using the Fast DNA SPIN Kit for Soil (MP Biomedicals, Solon, OH, USA) according to the protocol. The extracted DNA was used as a template for PCR targeting the partial 16S rRNA gene, using the primer set of 27F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 518R (5′-GTATTACCGCGGCTGCTGG-3′). AmpliTaq Gold (Applied Biosystems, Carlsbad, CA, USA) was used for PCR analysis. After preheating at 95°C for 5 min, a cycle of heating at 95°C for 1 min, annealing at 50°C for 1 min, and extension at 72°C for 2 min was repeated 25 times. PCR products were purified using the Ultra Clean PCR Clean-up Kit (MO-BIO, Carlsbad, CA, USA) according to the manufacturer's protocol. Purified PCR products were ligated with the pT7Blue vector (Novagen, Darmstadt, Germany) using the DNA Ligation Kit (Takara Bio, Kusatsu, Japan) according to the manufacturer's protocol. Then, 10 µL of the ligation mixture was transformed to 100 µL of Escherichia coli DH5α (Takara Bio) as competent cells by standard methods. After colonies were formed on an LB-ampicillin plate, white colonies were selected and plasmids were extracted using the Wizard SV Minipreps DNA Purification System (Promega, Madison, WI, USA). Extracted plasmids were digested by EcoRI and PstI (Takara Bio) to determine the size of the inserted DNA. After electrophoresis, DNA bands were visualized by staining the gels in 5 mg/L ethidium bromide solution for 20 min. FAS-III (Toyobo, Osaka, Japan) was used to view the DNA bands. A sequence analysis of the inserted DNA was completed by Takara Bio. Data were compared with the National Center for Biotechnology Information (NCBI) database using a Basic Local Alignment Search Tool (BLAST) search to identify the clones. The ratio of LAB was determined by the number of LAB clones to that of total bacterial clones.
Isolation of LAB from tofu-misozuke MRS medium (peptone 10 g/L, meat extract 20 g/L, yeast extract 5 g/L, glucose 20 g/L, Tween80 1 g/L, K2HPO4 2 g/L, sodium acetate 5 g/L, diammonium hydrogen citrate 2 g/L, MgSO4·7H2O 0.2 g/L, MnSO4·nH2O 0.05 g/L), and LM17 medium (casein 5 g/L, peptone 5 g/L, meat extract 5 g/L, yeast extract 2.5 g/L, ascorbic acid 0.5 g/L, MgSO4·7H2O 0.25 g/L, disodium-β-glycerophosphate 10 g/L, lactose 0.25 g/L) were used to cultivate and isolate LAB. Amounts of 1 g/L each of sodium azide and cycloheximide were added to MRS or LM17 medium at a final concentration of 0.01 g/L each. Agar (15 g/L) was added to prepare the solid medium. Tofu-misozuke was cut and inoculated into MRS or LM17 medium in a flask. The flask was incubated statically at 30°C for 2 days. The culture broth was diluted and spread on the MRS plate. Colonies grown on the MRS plate were observed based on color, size, and shape. Various colonies were picked and inoculated onto an MRS slant, and incubated at 30°C for 2 days. They were subsequently stored at 4°C.
Identification of isolated LAB based on partial 16S rRNA gene sequences Isolated strains were analyzed taxonomically by partial 16S rRNA gene sequencing. Part of the 16S rRNA gene was amplified by colony PCR with the 27F and 518R primer set. GoTaq Master Mix (Promega) was used for colony PCR. The PCR reaction contained the picked colony as a template, 2 µL of forward and reverse primers (25 µM each), 25 µL of GoTaq green, and 20 µL of nuclease-free water. After preheating at 95°C for 5 min, a cycle of heating at 95°C for 1 min, annealing at 50°C for 1 min, and extension at 72°C for 2 min was repeated 25 times. A sequence analysis of the amplified DNA from isolates was completed by Takara Bio. Both PCR primers were used as the sequence primer for the sequence analysis. The obtained partial 16S rRNA gene sequences were compared to those in the NCBI database using BLAST.
Detection of tyrosine decarboxylase (tdc) gene GoTaq Master Mix (Promega) was used. PCR was performed as described above, and included 5 min at 95°C followed by 35 cycles of 30 s at 95°C, 30 s at 55°C, and 90 s at 72°C. The primer set of TD5 (5′-CAAATGGAAGAAGAAGTWGG-3′) and TD2 (5′-ACATAGTCAACCATRTTGAA-3′) was used for Enterococcus and Weissella. TD2-Lac (5′-GTCCAATCGACCATATTGAA-3′), instead of TD2, was used for Lactobacillus. Five microliters of PCR products and 1 µL of loading buffer were mixed and analyzed by electrophoresis to determine the size of the amplified fragments.
Determination of tyramine content in tofu-misozuke samples and the biogenic amine-producing ability of isolates An analysis of biogenic amines in tofu-misozuke was performed following the procedure described by Byun and Mah (2012). Briefly, 20 mL of 0.4 M perchloric acid was added to 5 g of sample, and the mixture was homogenized using a vortex mixer. The sample was kept in a cold chamber at 4°C for 2 h, and centrifuged at 3000 × g at 4°C for 10 min. The supernatant was collected, and the residue was extracted again with an equal volume of 0.4 M perchloric acid. Both supernatants were mixed, and the final volume was adjusted to 50 mL with 0.4 M perchloric acid. The extract was filtered through Whatman paper No. 1.
Biogenic amines in the culture broth were determined according to the procedure developed by Mah et al. (2003) and modified by Burdychova and Komprada (2007). A loopful of the isolated strain was inoculated into 5 mL of TSB with 0.25% (w/v) L-tyrosine disodium salt hydrate, L-phenylalanine, L-arginine, or L-ornithine hydrochloride (pH 5.8) with 0.0005% pyridoxal-HCl. After incubation at 30°C for 24 h, 100 µL of the culture broth was transferred to a new tube containing 5 mL of the same broth and incubated at 30°C for 24 h. After cultivation, the broth was taken up using a sterile syringe and filtered through a 0.2-µm membrane. Then, 9 mL of 0.4 M perchloric acid was added to 1 mL of the filtrate and mixed well using a vortex mixer. The mixture was kept in a cold chamber at 4°C for 2 h, and centrifuged at 3000 × g at 4°C for 10 min. The extract was filtered through Whatman paper No. 1. Standard solutions of biogenic amines were separately prepared at a concentration of 100 or 1000 mg/L to a final volume of 10 mL.
Determination of biogenic amines in extracts and standards An analysis of biogenic amines was carried out according to the procedures described by Eerola et al. (1993) with minor modifications. One milliliter of each solution was mixed with 200 µL of 2 M sodium hydroxide and 300 µL of saturated sodium bicarbonate. Then, 2 mL of dansyl chloride solution (10 mg/mL) dissolved in acetone was added to the mixture and incubated at 40°C for 45 min. Residual dansyl chloride was removed by adding 100 µL of 25% ammonium hydroxide. After 30 min of incubation at room temperature, the volume of the amine solution was adjusted with acetonitrile to 5 mL. Finally, the mixture was centrifuged at 3000 × g for 5 min, and the supernatant was filtered through a 0.20-µm syringe filter. The filtered supernatant was kept at −25°C until it was assayed by high-performance liquid chromatography (HPLC). An HPLC system (Shimadzu, Kyoto, Japan) equipped with an SPD-M20A detector and LC solution software was employed. L-column ODS (4.6 mm I.D. × 150 mm; Chemicals Evaluation and Research Institute, Tokyo, Japan) was used with H2O (solvent A) and acetonitrile (solvent B) as the mobile phases at a total flow rate of 1 mL/min. The program was set for a linear gradient starting from 50% solvent B to reach 90% after 30 min. A sample volume of 20 µL was injected, and the sample was monitored at 254 nm.
Sample characteristics and tyramine content The characteristics and tyramine content of tofu-misozuke samples are summarized in Table 1. The two types of tofu-misozuke differed with respect to taste and flavor. The flavor and taste of samples A and E (samples soaked in miso paste) were similar to cheese, while the samples of novel tofu-misozuke tasted like seasoning, such as soy sauce, or lightly of sake. Since this involves only soaking in liquid seasoning, it is not thought to promote fermentation during manufacturing. However, the addition of sake is known to prevent bacterial contamination during manufacturing. In addition, the tyramine content differed between the two types. We detected a high concentration of tyramine in samples A and E. β-Phenylethylamine and putrescine were also detected in both samples (as shown in Fig. 1a for sample E). Chromatograms of standard BAs, including tyramine, β-phenylethylamine, histamine, cadaverine, tryptamine, and putrescine, are shown in Fig. 1b. The amine content of other fermented soybean foods showed a similar pattern to that of tofu-misozuke (Byun and Mah, 2012, Yang et al., 2014). It was suggested that tofu-misozuke would be a typical BA-containing fermented soybean food.
| Sample | Process type | Taste | Tyramine content (mg/kg) |
|---|---|---|---|
| A | handmade | cheese like | 158.2 |
| B | manufactured | seasoning | N.D |
| C | manufactured | seasoning | N.D |
| D | manufactured | seasoning | 12.8 |
| E | handmade | cheese like | 136.1 |
N.D., Not detected
Typical HPLC chromatograms of biogenic amines in sample E (a) and standard solution (b) (1, tryptamine; 2, β-phenylethylamine; 3, putrescine; 4, cadaverine; 5, histamine; 6, tyramine).
Bacterial diversity in tofu-misozuke We amplified the partial 16S rRNA gene from the extracted DNA. The concentration of extracted DNA (ng/µL) from samples A, B, C, D, and E was 21.0, 6.5, 6.2, 9.6, and 16.0, respectively. These concentrations may indicate the bacterial concentration in each sample. The amplified PCR product was cloned into E. coli DH5κ competent cells for the construction of a bacterial clone library.
The number of analyzed clones was 30, 74, 24, 24, and 48 for samples A, B, C, D, and E, respectively. We determined the clones with the highest sequence similarity by BLAST searches, and the bacterial distribution of each sample is shown in Fig. 2. The ratio of LAB in samples A and E (soaked in miso paste) were as high as 80% and 92%, respectively. In the case of sample A, Weissella viridescens (36%), Lactobacillus sakei (28%), and L. curvatus (16%) were the dominant species with highly similar sequences. Bacillus megaterium (16%) and Deinococcus geothermalis (4%) were also detected. In sample E, L. homohiochii (50%),
Comparison of bacterial diversity among 5 tofu-misozuke samples
W. viridescens (27%), L. sakei (8%), and L. curvatus (8%) were the dominant LAB. W. viridescens was the major bacterial species identified in samples A and E. W. viridescens was reported to be isolated from fermented dry sausage (Papamanoli et al., 2003), and dry sausage showed high BA content (Suzzi and Gardini, 2003). Therefore, it was suggested that W. viridescens might be involved in BA production. The ratios of LAB to all bacteria detected in the manufactured type (samples B, C, and D) were very low: 5%, 0%, and 0%, respectively. Staphylococcus was dominant in sample B. Staphylococcus spp. are often found in fermented foods such as soy sauce (Wei et al., 2013). B. amyloliquefaciens and B. subtilis were the dominant species in samples C and D, respectively. These Bacillus spp. have been found in miso (Byun and Mah, 2012) and douchi, a fermented soybean curd (Chen et al., 2012). However, Bacillus spp. are likely not involved in the BAs produced in tofu-misozuke.
Figure 3 shows the correlation between the ratio of LAB and tyramine content in tofu-misozuke samples. This result is consistent with the generation of tyramine by LAB. It was suggested that LAB might be involved in BA production in tofu-misozuke as with fermented protein foods.
The linear fitting between tyramine content and LAB ratio in tofu-misozuke samples.
Identification of isolated LAB In the previous section, LAB constituted a large proportion of the bacteria detected in samples A and E. Therefore, we attempted to isolate LAB from samples A and E. Based on the results of partial 16S rRNA gene sequencing, we identified the isolated strains as 7 strains belonging to 3 genera (Table 2). Other than Enterococcus faecium, almost all of the strains detected in samples A and E (shown in Fig. 2) were isolated using MRS medium. Enterococcus was not detected in Fig. 2. However, Enterococcus was isolated using LM17 medium for lactic streptococci. The number of Enterococcus in the samples was likely very small, but we isolated E. faecium most frequently when using LM17 medium. We isolated Lactobacillus and Weissella using MRS medium. We detected two W. viridescens isolates from sample A; however, the accession numbers of their closest matches were different. We isolated W. hellenica and Weissella sp. from sample A, and W. viridescens, L. homohiochii, L. curvatus, and L. sakei from sample E.
| Isolate No.1) | Closest strains | Accession No.2) | Similarity(%) | tdc gene | Tyramine production |
|---|---|---|---|---|---|
| AL2 | Enterococcus faecium | KM495940 | 100 | + | + |
| ELI | 99 | + | + | ||
| AM1 | Weissella viridescens | KJ580423 | 99 | + | + |
| EM7 | 99 | + | + | ||
| EM8 | 99 | + | + | ||
| AM4 | W. viridescens | JF756284 | 99 | − | − |
| EM11 | 99 | − | − | ||
| AM3 | W. sp. SXVIII2 | HQ728331 | 99 | − | − |
| AM7 | W. hellenica | AB494724 | 100 | − | − |
| EM23 | Lactobacillus homohiochii | JX441600 | 100 | − | − |
| EM4 | L. curvatus | KJ477385 | 99 | + | + |
| EM32 | L. sakei | KC787547 | 99 | − | − |
Determination of tyramine-producing ability of isolated LAB The tyramine-producing ability of the isolated LAB is summarized in Table 2. The tyrosine decarboxylase coding gene (tdc) was detected by PCR, and tyramine-producing ability was examined by cultivation tests and HPLC analysis. E. faecium was positive in both tests. Enterococcus has been isolated from cheese and produces tyramine (Burdychova and Komprada, 2007). Isolates from tofu-misozuke closest to Enterococcus also produced tyramine. On the other hand, isolates close to W. viridescens were divided into two groups, namely, tyramine-producing and non-producing bacteria. Thus, tyramine-producing ability is not species-specific, but strain-specific. Isolates close to W. hellenica and Weissella sp. were negative in both tests. L. curvatus and L. homohiochii have been shown to produce tyramine (Pereira et al., 2001). In this study, the isolate identified as L. curvatus tested positive using both PCR and HPLC. In contrast, the isolate close to L. homohiochii was negative using both methods. In tofu-misozuke samples A and E, isolated bacteria that were genetically close to E. faecium, W. viridescens and L. curvatus showed tyramine-producing ability.
Determination of non-tyramine BA-producing ability of LAB To examine biogenic amines other than tyramine, we performed cultivation tests using the isolated strains. In the analysis of samples A and E, we detected phenylethylamine, putrescine, and tyramine (see sample E in Fig. 1a). We tested for phenylethylamine and putrescine production using phenylalanine and arginine or ornithine as precursors of each amine. Phenylethylamine was produced by the addition of phenylalanine only by E. faecium. Thus, it might be produced via tyrosine decarboxylase, as reported previously (Bargossi et al., 2015). Only isolate AM7 (close to W. hellenica) generated putrescine from agmatine, but not from ornithine, with the supplemented medium (Fig. 4). Therefore, isolate AM7 appeared to produce putrescine via the AgDI pathway. The peak detected at a retention time of 21.5 min in Fig. 4 was confirmed as residual phenylalanine that was added to the test medium.
Typical HPLC chromatogram of the putrescine production test from AM7 (W. hellenica)
BA production was related to the ratio of LAB in the final products. However, the BA producing ability of isolated LAB was dependent on the strain. Therefore, the content of BAs could be reduced by using non-producing LAB as the starter strain for the production of tofu-misozuke.
Acknowledgements We thank Mr. Toshiro Toyama at the main Toyama store for his kind advice regarding tofu-misozuke.
