Original papers
Development of an Ultra-sensitive Detection Method for Genetically Modified Soybeans in Natto, a Traditional Japanese Fermented Food
2018 年 24 巻 6 号 p. 1121-1128
詳細
2018 年 24 巻 6 号 p. 1121-1128
When extracting soybean DNA from natto, a traditional Japanese fermented food, Bacillus subtilis var. natto DNA is commonly present despite fastidious washing of samples. We confirmed that DNA extracted from B. subtilis var. natto inhibited PCR amplification of the soybean endogenous lectin gene (Le1) as well as the recombinant nopaline synthase terminator sequence (Nos-ter), which is used to identify genetically modified soybean (GMS) in natto. To mitigate B. subtilis var. natto DNA contamination, we developed a novel method for the specific extraction of Le1 and Nos-ter genes using DNA probes immobilized on streptavidin-coated magnetic beads with hybridization enhancement blockers, which are capable of enhancing the affinity between the target gene DNA and the probe DNA. By using the proposed extraction method, both Le1 and Nos-ter genes could be detected in samples containing B. subtilis var. natto DNA. DNA extraction from commercial nattos and identification of GMS were also successfully accomplished.
At present, the Ministry of Health, Labor and Welfare of Japan has approved 318 types of genetically modified (GM) crops such as soybeans, corn and potatoesi). In Japan, GM foods are not well accepted by consumers despite the results of scientific safety evaluations. Therefore, there are growing demands for the detection of GM crops in processed foods labeled as “genetically modified crops are not used in this product”.
The self-sufficiency rate for soybeans in Japan is particularly low. Ninety-three percent of domestically consumed soybeans were imported in 2016ii), mainly from the United States, Brazil and Canada, and the majority of soybeans (70%) were obtained from the United States. In the United States, GM soybeans (GMS) comprised 94% of all soybeans cultivated in 2016iii). Thus, it has been suggested that GMS might be present in products thought to be produced with non-GM soybeans. In fact, products regarded as “non-use of genetically modified crops” are often contaminated with GM crops. In cases of contamination of 5% or less, in which “non-use of genetically modified crops” can be legally stated on the label, contamination during the distribution route is taken into consideration. At present, however, an amendment is scheduled in which the manufacturer can indicate “non-use of genetically modified crops” only if the GM crops are not at detectable levels. In this respect, there is an increasing need for more precise methods for the detection of GM crops and products.
Natto, which is made of soybeans, is a very popular healthy food and is widely consumed as a national food in Japan. However, as mentioned above, the possibility exists that natto labeled as “non-GM” contains GMS. The Japanese government defines the standard method of recombinant gene detection according to the “Manual of the genetically modified food inspection and analysis (3rd edition)” (iv), which is carried out using the polymerase chain reaction (PCR), for processed foods such as natto. However, it is difficult to determine the GM content of natto (Matsuoka and Hino, 2004, Kitamura et al., 2016) because of the degradation of soybean DNA during the manufacturing process, i.e., steaming under high-temperature conditions, as well as the fragmentation and consumption of soybean genes during the fermentation process. In addition, the presence of B. subtilis var. natto DNA interferes with PCR detection of GMS target sequences. In this study, we confirmed that the presence of B. subtilis var. natto DNA interferes with PCR analysis of GMS. To obtain PCR amplification of the target DNA of GMS without interference, we developed a specific extraction method to purify and concentrate the target DNA, i.e., the soybean endogenous lectin gene (Le1) and the recombinant sequence of the nopaline synthase terminator (Nos-ter) (Kuribara et al., 2002), as illustrated in Figure 1, using biotinized probe DNA attached to streptavidin-coated magnetic beads through biotin-streptavidin linkage and hybridization enhancement blockers (HEBs) (Okumura et al., 2014). HEBs are complementary oligonucleotides upstream and downstream of the target DNA, which are capable of increasing the affinity between the probe and the target DNA.
Schematic diagram of the specific extraction method.
The target DNA is hybridized to biotin-labeled probe DNA attached to streptavidin-coated magnetic beads. Hybridization enhancement blockers (HEBs) hybridize upstream and downstream of the complementary sequence of the target to increase the affinity between the probe and the target. After the target was captured by the probe on the beads, the DNA but the target is washed away. Finally the target DNA are removed from the probe by heating at 100°C. It is then used for PCR as a template to detect the target gene.
In this study, we first confirmed that the presence of B. subtilis var. natto DNA interferes with the amplification of soybean DNA by real-time PCR. Then, using the novel specific extraction method proposed here, Le1 and Nos-ter as the two targets in GMS were detected from a DNA mixture of GMS and B. subtilis var. natto as well as from DNA extracted from commercial natto samples labeled as “non-use” or “use” of GM crops.
Samples A natto sample, “Natto no susume™”, made of GMS (Roundup Ready), was purchased from A-Hit bio (Sapporo, Japan). GMS and non-GMS were obtained from SDIX (Newark, DE, USA). A control plasmid containing parts of the genes necessary for the detection of GMS was purchased from Nippon Gene (Tokyo, Japan). The plasmid contains three parts of the recombinant region, a 35S ribosomal RNA promoter derived from the cauliflower mosaic virus (P35S), Nos-ter, and the functional gene of Roundup Ready soybean (RRS) (Kuribara et al., 2002). It also contains a part of the endogenous gene region Le1, encoding a lectin, of the Roundup Ready soybean produced by Monsanto. A 528 bp region including the above four parts of the plasmid was amplified by PCR with the primer set RRS 01-5′ and P35S 1-3′ designed by Kuribara et al. (2002) (Fig. 2), and was then purified using the Wizard SV Gel PCR Clean-up System (Promega Corporation, Madison, WI, USA) and used as a positive control for subsequent experiments.
Positions of primers, HEBs and probes on the positive control.
The figure shows the 528 bp DNA region of the positive control plasmid purchased from Nippon Gene (Tokyo, Japan) for detection of endogenous and recombinant genes in GM soybean. Arrows indicate primers used in this study. The black primers, RRS 01-5′ and P35S 1-3′, are used to amplify the 528 bp region of the positive control. The red and blue primers are used to amplify the soybean endogenous lectin gene (Le1) and the recombinant sequence of the nopaline synthase terminator (Nos-ter), respectively. Red shading indicates the Le1 target and blue shading the Nos-ter target. Red letters indicate the biotin-conjugated probe Le1, which is complementary to part of the Le1 target, and blue letters indicate the probe Nos-ter. Pink letters indicate HEBs to enhance the affinity between the Le1 probe and target, and sky blue letters indicate HEBs for Nos-ter.
Oligonucleotides Table 1 summarizes the three sets of primers used for PCR and real-time PCR. The positions of the primers in the positive control sequence are shown in Figure 2. All primer sets were designed by Kuribara et al. (2002).
| Primer name | Sequence (5′→3′) | Product size |
|---|---|---|
| RRS 01-5′ | CCTTTAGGATTTCAGCATCAGTGG | 528 bp |
| P35S 1-3′ | CCTCTCCAAATGAAATGAATCCTT | |
| Le1n02-5′ | GCCCTCTACTCCACCCCCA | 118 bp |
| Le1n02-3′ | GCCCATCTGCAAGCCTTTTT | |
| Nos ter 2-5′ | GTCTTGCGATGATTATCATATAATTTCTG | 151 bp |
| Nos ter 2-3′ | CGCTATATTTTGTTTTCTATCGCGT |
All of them were designed by Kuribara et al. (2002).
Figure 2 also shows the sequences of two target regions, Le1 and Nos-ter, two probes that are complementary to the target regions, and four HEBs that increase binding between the target and the probe. These are summarized in Table 2 and were purchased from Sigma-Aldrich Japan (Tokyo, Japan).
| Name | Sequence (5′→3′) | Size |
|---|---|---|
| Probea | ||
| Probe Le1 | Biotin-AAAAAA-GCGTTGCCAGCTTCGCCGCTTCCTTC | 32 bp |
| Probe Nos-ter | Biotin-AAAAAA-GTAATGCATGACGTTATTTATGAGATGGGTTTT | 39 bp |
| HEBb | ||
| Le1-left | ACTTCACCTTCTATGCCCCTGACAC | 28 bp |
| Le1-right | TCCACATTTGGGACAAAGAAACCGG | 25 bp |
| Nos-ter-left | ATGATTAGAGTCCCGCAATTATACATTTAAT | 31 bp |
| Nos-ter-right | TTGAATTACGTTAAGCATGTAATAATTAACA | 31 bp |
DNA extraction from natto samples, soybeans and B. subtilis var. natto Template DNA for PCR was extracted from powdered GMS and non-GMS as well as natto samples using a DNeasy Maxi Plant kit (Qiagen, Venlo, Netherlands) according to the manufacturer's protocol. Prior to extraction, natto samples were washed to remove the sticky paste on the surface and then homogenized according to the standard method defined by the governmentiv), in which a silica-based DNA extraction and purification method such as the above kit is recommended to obtain good PCR amplification results; notably the cetyltrimethylammonium bromide (CTAB) method is also listed. In this study, the method with a DNeasy Maxi Plant kit was adopted as the standard method for DNA extraction from natto samples. DNA concentrations were measured using a NanoDrop micro volume UV-Vis spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA).
Genomic DNA from B. subtilis var. natto was also extracted in order to assess its inhibitory effect on PCR. First, B. subtilis var. natto Naruse strain was cultured in 1000 mL of nutrient broth medium (DIFCO Laboratories, Detroit, MI, USA) composed of beef extract (3 g/L), peptone (10 g/L), and NaCl (5 g/L) for 24 h at 37°C, then the cells were harvested by centrifugation and DNA was extracted using the above mentioned kit.
Real-time PCR The Le1 and Nos-ter regions were amplified from the positive control in the presence of various concentrations of B. subtilis var. natto genomic DNA using the LightCycler ST300 (Roche Diagnostics, Basel, Switzerland). The reaction was performed in a total volume of 20 µL, which contained 10 µL SYBR Premix ExTaq (Takara Bio, Shiga, Japan), 0.8 µM of each primer (0.32 µL each), 0.36 µL template solution and 9 µL of a 3-fold serially-diluted solution of extracted B. subtilis var. natto DNA. The thermal amplification profile was as follows: 95°C for 60 s; 75 cycles of 95°C for 10 s, 60°C for 10 s, and 72°C for 10 s. Subsequently, a melting curve was plotted from 65 to 98°C (0.1°C/sec).
Specific extraction of the Le1 or Nos-ter gene using magnetic beads For specific extraction of the Le1 or Nos-ter gene from samples containing a large amount of B. subtilis var. natto DNA, streptavidin-coated magnetic beads (Dynabeads M270 streptavidin, Thermo Fisher Scientific) were used. First, a 12-µL sample solution was prepared by mixing the positive control and the DNA solution extracted from B. subtilis var. natto, or by extracting from natto samples, 1 µL of 100 µM biotin-conjugated probe, 3.75 µL of 50 µM right- and left-HEB, and 20 µL of PBS containing 5 M NaCl. The solution was incubated for 24 h at 4°C. Next, 10 µL of the 10 mg/mL bead solution was dispensed into a 1.5 mL microtube and the liquid was removed by magnetic attraction. Then, the beads and the sample solution were mixed and incubated for 1 h with moderate rotation to recover the probe and target bound to the beads. Next, the beads were washed with PBS containing 1 M NaCl and immersed in 15 µL of ultrapure water. The water and beads were heated at 100°C, and the beads were removed by magnetic attraction. Finally, the Le1 and Nos-ter genes were detected by PCR using the solution as the template. The reaction was performed in a total volume of 25 µL, which consisted of 10 µL of the final sample solution, 12.5 µL Premix ExTaq (Takara Bio), and 0.25 µM of each primer under the following conditions: 94°C for 1 min; 30 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 30 s; and a final extension at 72°C for 7 min. The PCR products were electrophoresed on 20% acrylamide gels (Novex TBE Gels; Thermo Fisher Scientific) to confirm amplification of the target sequences.
Effect of B. subtilis var. natto DNA on real-time PCR amplification of Le1 and Nos-ter genes Figure 3a and b show SYBR Green fluorescence charts of real-time PCR of Le1 and Nos-ter with various amounts of DNA extracted from B. subtilis var. natto, respectively. Each solution contains the same amount of template and various amounts of DNA extracted from B. subtilis var. natto, diluted from 30- to 33-fold. For both Le1 and Nos-ter, in the presence of undiluted B. subtilis var. natto DNA, amplification was not observed (Fig 3a and b, solid red line). In the presence of DNA diluted 32- and 33-fold, amplification of the Le1 gene and Nos-ter region was observed and the graphs of 32- and 33-fold dilutions were almost the same. In the presence of B. subtilis var. natto DNA at a 31-fold dilution, amplification was obviously inhibited compared to the 32- and 33-fold dilutions. In a comparison of Le1 and Nos-ter, amplification of the Nos-ter region was observed at a higher cycle number. This is probably because PCR efficiency is low in the Nos-ter region compared to Le1. The results of real-time PCR showed that the reaction was inhibited by the presence of B. subtilis var. natto DNA, revealing one of the reasons for the difficulty in detecting GMS in natto samples. By using PCR, the target region can be in principle amplified from several copies of template DNA, even if a large amount of unrelated DNA, which lacks the amplified region, is present. Thus, it is noteworthy that the PCR was inhibited by the presence of the unrelated DNA. However, we have confirmed the reproducibility of the results with a few additional tests. The inhibition scheme of a large amount of unrelated DNA against the target PCR amplification can be proposed as follows. In the presence of a large amount of unrelated DNA, primers may non-specifically bind to sequences from B. subtilis var. natto aside from the complementary sequence of the target DNA in soybeans, resulting in the lack of primer molecules in the PCR reaction mixture. It has been reported that a lower primer concentration limits the PCR reaction, forcing it to plateau at a lower level of productv).
Effect of the coexistence of B. subtilis var. natto DNA on real-time PCR.
Real-time PCR was performed with the primer sets Le1n02-5′ and -3′ (a), and Nos ter 2-5′ and -3′ (b). The amplification efficiencies were examined in the presence of B. subtilis var. natto DNA solution (18.6 ng/µL) diluted 27 (broad spacing green broken line), 9 (medium spacing blue broken line), 3 times (narrow spacing pink broken line), and without dilution (red line).
In this experiment, we aimed to set the experimental condition of the concentration of unrelated DNA to that of the target to be equivalent to concentrations found in actual natto samples. However, because of the difficulty in multiplex quantification of soybean-derived DNA and B. subtilis var. natto derived-DNA from the DNA extracted from natto samples, we were unable to accurately determine these concentrations. However, it has now become easier to sequence large amounts of DNA with next-generation sequencers, and to estimate the concentrations of DNA from multiple origins. Thus, we aim to quantify these in the future.
These results indicate that DNA from B. subtilis var. natto interferes with the PCR amplification of Le1 and Nos-ter genes of GMS, which was observed to occur in the PCR using natto samples as templates.
Detection of GMS genes in the mixture of positive control and DNA extracted from B. subtilis var. natto using the specific extraction method The mixture of the positive control and B. subtilis var. natto-derived DNA was prepared as a representative solution extracted from natto samples, and the effectiveness of the specific extraction method (Fig. 1) to detect GMS genes was examined. To summarize this method, first the biotin-modified probe is added to the sample, and HEB is added to enhance the affinity between the probe and the target. Following incubation, streptavidin-coated magnetic beads are added to the solution and the biotin-modified probe, along with the target, is captured by the beads. Finally, the beads are washed with PBS to remove non-target DNA, and PCR is performed using the captured target as a template. Experimental conditions, such as the type and amount of beads and the probe amount, etc., of the proposed method were determined by flow cytometric analysis using the FITC-conjugated 528 bp DNA (Fig. 2).
For Le1, four representative samples of a DNA solution extracted from natto samples, which were mixed with two concentrations of the positive control, 3.2×101 and 9.9×107 fg/µL (final concentrations; same as below), and two concentration of the DNA from B. subtilis var. natto, 5.1×104 and 1.6×104 fg/µL, were prepared. The target sequence was extracted from the sample solution using the proposed method, and PCR analysis was examined. The results of gel electrophoresis of the PCR amplification products are shown in Figure 4a. A band considered to be a Le1 amplification product (118 bp) was observed slightly above the 100 bp marker in all four lanes in samples treated with the extraction method (Fig. 4a, lanes 1–4). In contrast, no band was observed in all lanes in samples not treated using the method (Fig. 4a, lanes 5–8). The results show that the target sequence could be extracted effectively and PCR interference from the presence of a large amount of B. subtilis var. natto-derived DNA was mitigated using the proposed method.
Detection of GM in standard samples.
A standard sample with a mixture of the positive control and B. subtilis var. natto-derived DNA was prepared. Le1 (a) and Nos-ter (b) genes were detected by PCR with (a, lanes 1–4; b, lanes 1 and 2) or without (a, lanes 5–8; b, lanes 3 and 4) the proposed DNA extraction method. For Le1 (a), four samples were prepared, with final concentrations of 3.2×101 fg/µL positive control and 5.1×104 fg/µL DNA from B. subtilis var. natto (a, lanes 1 and 5), 3.2×101 fg/µL positive control and 1.6×104 fg/µL B. subtilis var. natto DNA (a, lanes 2 and 6), 9.9×107 fg/µL positive control and 5.1×104 fg/µL B. subtilis var. natto DNA (a, lanes 3 and 7), and 9.9×107 fg/µL positive control and 1.6×104 fg/µL B. subtilis var. natto DNA (a, lanes 4 and 8). Lane M: 100 bp ladder, lane 9: positive control, lane 10: negative control. For Nos-ter (b), two samples were prepared, with final concentrations of 9.5×10−2 fg/µL positive control and 5.1×104 fg/µL DNA from B. subtilis var. natto (lanes 1 and 3), and 9.5×10−2 fg/µL positive control and 1.6×104 fg/µL B. subtilis var. natto DNA (lanes 2 and 4). Lane M: 100 bp ladder, lane 5: negative control. All negative controls are the PCR product with the sterilized water used as a template.
A non-specific band less than 100 bp was detected below the target band (lanes 1–4). The most likely reason for this observation is that the Le1 probe was carried over to the PCR step and acted as a reverse primer, which amplified a 73 bp product together with the Le1n02-5′ primer. Although the existence of the lower band has no direct influence on the issue of detection, in the case of quantification with real-time PCR, the 73 bp amplification product may represent a potential source of criticism. To avoid this problem, the TaqMan method could be used initially, where the sequences of the reverse primer and the probe could be optimized. In any case, some minor modifications regarding the position and design of the probes and primers are likely needed, and we are considering this as a future subject.
For Nos-ter, two representative samples of a DNA solution extracted from natto samples, which were mixed with DNA concentrations of 5.1×104 and 1.6×104 fg/µL obtained from B. subtilis var. natto and a positive control (9.5×10−2 fg/µL) were prepared. As with Le1, these samples were subjected to the extraction method and PCR analysis. As shown in Figure 4b, a band considered to be a Nos-ter amplification product (151 bp) was observed in two lanes in the case of application of the extraction method (Fig. 4b, lanes 1, 2); whereas, no band was observed in these two lanes without application of the method (Fig. 4b, lanes 3, 4). For Nos-ter, a non-specific band as observed for Le1 was not present.
Kakihara et al. (2002) successfully detected the expected PCR products in GMS and heat-treated GMS using primer sets for lectin 1 and the junction region between Nos and CP4EPSPS, producing a product of 100 –130 bp and thereby overcoming the influence of DNA degradation resulting from heat-treatment and the CTAB method. However, PCR products could not be detected in natto extracts and the need for further investigation of PCR inhibition in natto extracts was discussed.
These results show that the proposed extraction method is a candidate solution and is of practical use for detecting Le1 and Nos-ter as targets in the presence of DNA derived from B. subtilis var. natto.
Detection of GMS in commercial natto Finally, the detection of GMS in commercial natto was conducted using the proposed extraction method. Several commercial natto samples labeled as “made of non-genetically modified soybean” were purchased at a supermarket, whereas the natto, “Natto no susume™”, which is made with GMS was purchased directly from the manufacturer.
First, DNA was extracted from the natto samples using the standard method, then the target DNA was extracted by the proposed extraction method, finally the extracted DNA was subjected to PCR analysis. Figure 5 shows the image of gel electrophoresis analysis of the PCR products. The bands of the PCR amplification products of Le1 were successfully detected at the same position as the positive control (Fig. 5, lane LP) from all seven samples (left lanes of 1-7, 118 bp). Additionally, a distinct band of Nos-ter was detected only from the sample of “Natto no susume™” (right lane of 7, 151 bp) at the same position as the positive control (lane NP).
Detection of Le1 and Nos-ter genes from seven packs of commercial natto.
The endogenous Le1 and recombinant Nos-ter genes were identified using the beads and probe extraction method for commercial natto. Each numbered area contains two samples of amplicons, Le1 on the left lane and Nos-ter on the right lane from each natto. Six of them are purchased at s supermarket, and all of them were indicated as “made of non-genetically modified soybean” (1 –6). The remaining one is “Natto no susume™” and it was indicated as “made of recombinant soybean” (7). Lanes LP and LN are positive and negative control with Le1 primers, respectively, lanes NP and NN are positive and negative control with Nos-ter primers, respectively. Positive control is the PCR product with the 528 bp DNA (Fig 3) used as a template, and negative control is that with the sterilized water. Lane M indicates 100 bp ladder marker.
When the target band was present, non-specific bands less than 100 bp were also detected under the target bands for both Le1 and Nos-ter, representing amplicons generated by the probe, which had dissociated from the beads and acted as a reverse primer. As mentioned in the previous section, although this non-specific band has no influence on GMS detection, in the case of quantification with real-time PCR, the non-specific products may be a problem. Thus, efforts are needed to improve the quantification method.
Although it is difficult to detect GMS genes from natto samples using a combination of the standard DNA extraction method and PCR amplification, the present results indicate that the proposed extraction method, using beads and probes in conjunction with HEBs, can effectively detect the DNA of GMS in commercial natto.
Further precise observation revealed a minor band associated with Nos-ter amplification for all six natto samples labeled as “made of non-genetically modified soybean” (right lane of 1–6). These are thought to result from trace contamination of GMS in the non-GMS as a raw material.
It is of interest how the proposed purification method with beads and probes alters the individual concentrations of the target and unrelated DNA. In regards to this point, the relevant data could be acquired using next-generation sequencing in future studies. In addition, we aim to further improve this method to be semi-quantitative, by reducing the influence of non-specific amplified products through the use of the TaqMan method as well as by reviewing the positions of the probes and primers.
In natto samples, soybean DNA is degraded during the manufacturing process, i.e., steaming under high-temperature conditions, and is decomposed and consumed during the fermentation process. Moreover, despite fastidious washing, soybean DNA is commonly contaminated with B. subtilis var. natto DNA. As demonstrated in this study, the presence of B. subtilis var. natto DNA interferes with PCR amplification, making it more difficult to detect GMS by the conventional method, a combination of the standard DNA extraction method and PCR amplification.
GMS genes, both the recombinant Nos-ter and the endogenous Le1, were extracted from the mixture of template DNA, positive control, and genomic DNA from B. subtilis var. natto using the specific extraction method of streptavidin-coated magnetic beads and biotinized probe with simultaneous use of HEBs. As a result, both genes were successfully amplified by PCR, indicating that interference from the presence of DNA in the mixture was mitigated. The genes were also extracted from DNA solutions of commercial nattos made with GMS and non-GMS, and GMS were successfully detected using PCR amplification.
Recently, the legislation for the labeling of GM crops has become stricter. Using the present method, it will be possible to reduce “inconclusive results” due to non-detection of Le1 in natto samples. Furthermore, it will be possible to respond to the increasing demand for GM detection in the future. We will continue to investigate avenues to improve gene detection and quantification of GM crops for the practical application of this technology.
Acknowledgments This work was supported in part by a grant for the Development of Bio-Ventures from the public benefit corporation of Kakihara Science and Technology Research Foundation. We thank Robbie Lewis, MSc, from Edanz Group (www.edanzediting.com/ac) for editing a draft of this manuscript.
cetyltrimethylammonium bromide
dsdouble strand
FITCfluorescein isothiocyanate
GMSgenetically modified soybeans
HEBhybridization enhancement blocker
Le1endogenous lectin gene from soybean
Nos-ternopaline synthase terminator
P35S35S RNA promoter derived from cauliflower mosaic virus
PBSphosphate buffer saline (pH 7.4)
PCRpolymerase chain reaction
RRSRoundup Ready soybean
sssingle strand
