Quantitative real-time PCR analysis of four genetically modified soybean events using plasmid and genomic DNA calibrators

Abstract

A calibration curve is required for reliable and accurate quantitative real-time PCR analysis of genetically modified (GM) organisms, necessitating the use of reference materials as calibrators. In this study, two types of DNA calibrators—plasmid DNA (pDNA) and genomic DNA (gDNA)—were used to quantify four GM soybean events (SYHT0H2, MON87751, DAS-44406-6, and DAS-81419-2). The PCR efficiency and linearity for the calibrators adhered to the CODEX guidelines. The conversion factor (Cf) was calculated as the ratio of copies of GM events to those of endogenous genes using the pDNA calibration curve. To assess the accuracy and repeatability of these assays, quantification at GM levels of 3% and 1% was performed. Based on our results, we believe that the pDNA calibrator assessed in this study can be used as a reference material for GMO quantitative analysis and can replace gDNA, especially considering the ease of management and advantages of mass production.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10068-023-01392-0.

Introduction

Genetically modified (GM) crops have been steadily developed for commercialization. The area under GM cultivation has increased 113-fold from 1.7 million hectare in 1996 to 191.7 million hectare in 2018. GM soybean has the largest planted area (95.9 million hectare), accounting for approximately 50% of the area under GM cultivation, globally (ISAAA, 2018). According to the GM Approval Database of the International Service for the Acquisition of Agri-biotech Applications (ISAAA), more than 500 events involving 32 crops were authorized globally in 2020. In Korea, seven crops (soybean, maize, cotton, canola, alfalfa, sugar beet, and potato) and 188 events, including stacked events, have been approved for food in accordance with the Ministry of Food and Drug Safety information (MFDS, 2023).

In many countries (e.g., in the European Union (EU), Korea, and Japan), threshold levels have been established, and the labeling of GM crops and their products is mandatory for consumer information (Kamle and Ali, 2013). To implement labeling requirements, the development and application of reliable quantitative detection methods is essential (Huang and Pan, 2005). Real-time PCR is the most widely accepted quantification technique for analysis and determination of GM content in crops and products. Reference materials (RMs), which are used as calibrators, are necessary for reliable and accurate results in quantitative real-time PCR analysis. Two types of RMs—dried powder and plasmid RMs—have been developed and are commonly used to prepare calibration curves in the analysis of genetically modified organisms (GMO). Dried powder RMs have been generally used owing to properties similar to those of test samples, but they have some limitations, such as limited quantitative range, uncomfortable preparation procedures, and high costs (Meng et al., 2012; Zhang and Guo, 2011). Unlike genomic DNA (gDNA) extracted from dried powder, plasmid RMs can be easily constructed and stored stably for a long time when frozen. Furthermore, they have high productivity and can be recombined for multiple GM sequences into one plasmid (Chaouachi et al., 2013; Li et al., 2017; Lievens et al., 2010). Although plasmid RMs have contamination problems, they have been extensively used in GMO analysis owing to their handiness and low costs (Meng et al., 2012). Comparison of gDNA and plasmid DNA (pDNA) as calibrators for quantitative analysis of GMO showed that pDNA is a good alternative for gDNA in real-time PCR systems (Burns et al., 2006; Capioara-Buda et al., 2012; Toyota et al., 2006a; 2006b). In addition, detection methods validated for newly constructed plasmid standard molecules have been steadily developed for GM soybean, maize, canola, cotton, alfalfa, and potato events (Ballari et al., 2013; Guan et al., 2011; Guertler et al., 2019; Kuribara et al., 2002; Li et al., 2017; Lievens et al., 2010; Meng et al., 2012; Rho et al., 2004; Taverniers et al., 2005; Wang et al., 2013; Yang et al., 2007).

In this study, the performance of pDNA as a calibrator vis-à-vis that of gDNA was evaluated for four GM soybean events (SYHT0H2, MON87751, DAS-44406-6, and DAS-81419-2). It is pertinent to mention here that the commercial availability of certified reference materials (CRMs) with specified GM content (%) for GM soy events, SYHT0H2 and MON87751, is limited. To overcome this limitation, we developed RMs with different copy numbers. Additionally, we determined the conversion factor (Cf) using a pDNA calibration curve and 100% GM seeds. The suitability of pDNA for real-time PCR was confirmed through quantification assays, and the obtained results were subjected to rigorous statistical analysis.

Materials and methods

Materials

Seeds of non-GM and GM soybean (SYHT0H2, MON87751, DAS-44406-6, and DAS-81419-2) were kindly supplied by developers. CRMs of 10% and 1% GM soybean (DAS-44406-6 (ERM-BF436e and ERM-BF436d) and DAS-81419-2 (ERM-BF437e and ERM-BF437d)) were purchased from the Institute for Reference Materials and Measurements (IRMM, Geel, Belgium). pUC-GM soy_C (which contains sequences of lectin gene, SYHT0H2, MON87751, DAS-44406-6, and DAS-81419-2) was used as a reference plasmid calibrator for the standard curve (Fig. 1). This reference plasmid was developed as part of a Ministry of Food and Drug Safety (MFDS) R&D project and serves as one of the four reference plasmids for quantification of 17 GM soybeans that have been approved for use in Korea.

Fig. 1.

Schematic diagrams of the standard reference plasmid, pUC-GM soy_C (A), and gene constructs for SYHT0H2, MON87751, DAS-44406-6, and DAS-81419-2 (B). The arrows indicate the location of the event-specific primers. LB, left border; RB, right border; P-SMP, synthetic minimal plant promoter; avhppd, p-hydroxyphenylpyruvate dioxygenase gene derived from oat; T-NOS, terminator sequence from the nopaline synthase gene of Agrobacterium tumefaciens; P-35S, promoter region of cauliflower mosaic virus; pat, Streptomyces viridochromogenes gene encoding the selectable marker PAT; P-CMP, promoter and leader sequence from the Cestrum yellow leaf curling virus; P-Act2, promoter from act2 gene of Arabidopsis thaliana; TS-CTP2, targeting sequence of the ShkG gene from Arabidopsis thaliana; cry2Ab2, coding sequence for the Cry2Ab2 protein of Bacillus thuringiensis; T-Mt, terminator of metallothionein like gene from Oryza sativa; P-RbcS4, promoter and leader sequences from Arabidopsis thaliana rbcS gene; TS-RbcS4, targeting sequence from Arabidopsis thaliana rbcS gene; cry1A.105, coding sequence for the Cry1A.105 protein from Bacillus thuringiensis; T-Pt1, terminator of phosphate transporter gene from Medicago truncatula; T-histon_At, termination sequence from H.4 gene of Arabidopsis thaliana; mepsps, sequence encoding the modified EPSP (5-enolpyruvylshikimate-3-phosphate) enzyme from Zea mays; P-Ph4a748, sequence including the promoter region of the histone H4 gene of Arabidopsis thaliana; P-ubi10, protmoter region from Arabidopsis thaliana UBQ10 polyubiquitin gene; aad-12, gene sequence that encodes the AAD-12 protein from Delftia acidovorans; T-ORF23, terminator and polyadenylation site of open reading frame 23 (ORF23) of Agrobacterium tumefaciens; P-CsVMV, constitutive promoter derived from Cassava vein mosaic virus; T-ORF1, terminator and polyadenylation site of open reading frame 1 (ORF1) of Agrobacterium tumefaciens; cry1F v3, synthetic version of the cry1F gene from Bacillus thuringiensis subsp. azaiwai strain PS811; cry1Ac, synthetic version of the cry1Ac gene of Bacillus thuringiensis subsp. kurstaki strain HD73

DNA extraction

Non-GM and GM soybean (SYHT0H2, MON87751, DAS-44406-6, and DAS-81419-2) seeds were milled using a grinder. Genomic DNA was extracted from ground materials and powdered CRMs using the PowerPrep™ DNA Extraction from Food and Feed kit (Kogenebiotech, Korea) following the manufacturer’s instructions. The concentration of the extracted gDNA was measured using the QUBIT™ 3.0 Quantitation Starter Kit (Thermo Fisher Scientific, Wilmington, DE, USA) and the stock concentration was adjusted to 20 ng/µL by diluting in nuclease-free water. The DNA samples were stored at − 20 °C until use.

Oligonucleotide primers and probes

Primers and probes for the event-specific detection of SYHT0H2, MON87751, DAS-44406-6, and DAS-81419-2 were used for quantitative PCR analysis. The lectin gene was selected as the endogenous soybean reference gene. The sequences of the lectin gene and GM soybean event-specific primers and probes (Le1n02-5′ Le1n02-3′/Le1-Taq for lectin gene, FE08316 fwd/FE08317 rev/FE08318 probe for SYHT0H2, MON87751 Q1/Q2/Q3 for MON87751, DAS-44406-5F/5R/5p1 for DAS-44406-6, and DAS81419-f2/r1/p3 for DAS-81419-2) were referenced from the Korean Food Code Notice No. 2019-16 of the MFDS (Table S-1 in supplementary material). All probes were labeled with 6-carboxyfluorescein (FAM) at their 5′-end. The 3′-end contained tetramethyl-6-carboxyrhodamine (TAMRA) or black hole quencher-1 (BHQ1). Primers and TaqMan probes were synthesized by Bioneer (Daejeon, Korea). The schematic diagrams in Fig. 1 show the location of each primer in the flanking region of the four GM soybean events.

Preparation of pDNA and gDNA calibrators

One haploid genome copy of soybean was considered 1.13 pg (Greilhuber and Obermayer, 1997). The plasmid copy number was calculated using Eq. (1) as follows:

$$\text{m}=\left[\text{n}\right]\times [1.096\times 10^{-21}\text{g}/\text{copy}]$$ 1

where m is the mass of DNA and n is the number of base pairs of one copy of a recombinant plasmid or genome. The pDNA calibrator was prepared by tenfold serial dilution in TE buffer (10 mM Tris–HCl, pH 8.0; 1 mM EDTA) consisting of five concentrations ranging from 3 × 10⁵ to 3 × 10¹ copies per reaction.

The final concentration of 100 ng used for real-time PCR analysis was 88,496 copies. The standard curve for gDNA extracted from 100% GM seeds was calibrated using tenfold dilution in nuclease-free water. It consisted of five concentrations ranging from 88,496 to 8.85 copies per reaction.

Preparation of test samples

For quantitative analysis using pDNA and gDNA calibrators, test samples of GM 3% and 1% were used. The 3% samples of DAS-44406-6 and DAS-81419-2 were prepared by mixing non-GM soybean and 10% CRM DNA diluted to 20 ng/µL. For 1% samples, DNA extracted from 1% CRM was used. The 3% and 1% samples of SYHT0H2 and MON87751 were mixed with non-GM soybean, and DNA extracted from 100% GM seeds was diluted to 20 ng/µL.

Real-time PCR

Quantitative assays were performed using an ABI 7500 real-time PCR system (Applied Biosystems, Carlsbad, CA, USA) in a final reaction volume of 25 µL. The PCR mixtures contained 12.5 µL TaqMan™ Universal Master Mix II with UNG (Applied Biosystems), 500 nM of each primer, 200 nM of probe for lectin, 300–600 nM of each primer, 120–300 nM of probe for GM events, and 100 ng of DNA template. The real-time PCR conditions were as follows: 1 cycle of 2 min at 50 °C, 10 min at 95 °C, 45 cycles of 15 s at 95 °C, and 1 min at 60 °C. Data were analyzed using the 7500 software v2.0.6 (Applied Biosystems). All real-time PCR were repeated three times, using triplicate samples.

Measurement of conversion factor and calculation of GMO amounts

The C_f was defined as the ratio of the transgenic gene copy number to the endogenous gene copy number in 100% GM seeds. The final concentration of the gDNA extracted from 100% GM seeds was adjusted to 100 ng. In each experiment, the average C_f value was obtained as the mean of three repeated measurements for each GM event. The final C_f was determined using the average value of the results obtained from three different experiments. It can be calculated according to Eq. (2).

$$Cf=\frac{CopiesofGMeventinthe100\%GMseeds}{Copiesoftaxoninthe100\%GMseeds}$$ 2

The GMO amounts (%) of the test samples in the quantitative analysis using pDNA and gDNA calibrators were calculated using Eqs. (3) and (4), respectively, as follows:

$$GMOamounts(\%)=\frac{CopiesofGMeventintestsample\times 100}{Copiesoftaxonintestsample\times Cf}$$ 3

$$GMOamounts(\%)=\frac{CopiesofGMeventintestsample\times 100}{Copiesoftaxonintestsample}$$ 4

Statistical analysis

The test was independently performed in triplicate on three separate occasions. A t-test was conducted to statistically compare the calibration curves of pDNA and gDNA.

Results and discussion

Comparison of the efficiency and linearity of PCR using pDNA and gDNA as calibrators

Accurate and reliable quantification using real-time PCR necessitates the construction of a suitable standard curve. In assessing the standard curve for real-time PCR assays, we examined the two crucial parameters of PCR, namely efficiency and linearity. As is evident from the results presented in Table 1 and Fig. 2, the slope of the pDNA calibrator for lectin and the four GM events ranged from − 3.46 to − 3.53. For the gDNA calibrator, the slope ranged from − 3.54 to − 3.47. The R² values for pDNA and gDNA calibration curves ranged from 0.996 to 1.000, exhibiting no significant differences for lectin or the four GM events (Fig. 2). PCR efficiency was determined by calculating the slope of the standard curve using the formula: E (%) = (10^(−1/slope) − 1) × 100. The PCR efficiency and linearity for both pDNA and gDNA calibrators revealed an acceptable bias adhering to the criteria described in the CODEX guidelines (CAC-GL 74-2010). Statistical analysis using the t-test confirmed that the slope and PCR efficiency for the pDNA and gDNA calibrators were equal.

Table 1.

PCR efficiency and linearity for pDNA and gDNA as calibrators

Target	Parameter	Calibrator	Mean	SD	Variance	p-value*
Lectin	Slope	pDNA	− 3.52	0.02	0.60 × 10–3	0.553
		gDNA	− 3.53	0.05	2.93 × 10–3
	PCR efficiency (%)	pDNA	92.25	0.88	0.77	0.592
		gDNA	91.91	1.97	3.87
SYHT0H2	Slope	pDNA	− 3.51	0.01	3.63 × 10–5	0.226
		gDNA	− 3.53	0.03	9.07 × 10–4
	PCR efficiency (%)	pDNA	92.83	0.22	0.05	0.225
		gDNA	91.93	1.06	1.13
MON87751	Slope	pDNA	− 3.46	0.01	0.61 × 10–4	0.884
		gDNA	− 3.47	0.06	3.90 × 10–3
	PCR efficiency (%)	pDNA	94.54	0.29	0.09	0.904
		gDNA	94.36	2.36	5.57
DAS-44406-6	Slope	pDNA	− 3.53	0.05	2.42 × 10–3	0.834
		gDNA	− 3.54	0.03	0.82 × 10–3
	PCR efficiency (%)	pDNA	92.07	1.73	2.99	0.824
		gDNA	91.80	1.01	1.02
DAS-81419-2	Slope	pDNA	− 3.47	0.01	0.72 × 10–4	0.395
		gDNA	− 3.51	0.07	5.49 × 10–3
	PCR efficiency (%)	pDNA	94.15	0.32	0.10	0.402
		gDNA	92.70	2.66	7.08

SD repeatability standard deviation

^*p-value was calculated at 95% confidence level, using the t-test

Fig. 2.

Standard curves for SYHT0H2, MON87751, DAS-44406-6, and DAS-81419-2 in real-time PCR assays using pDNA and gDNA as calibrators

Repeatability of the pDNA and gDNA calibration curve

To assess the repeatability of the pDNA and gDNA standard curves for the four GM events, the relative standard deviation (RSD) values were evaluated. The RSD was calculated based on the average and repeatability standard deviation (SD) derived from three repeated measurements. The SD and RSD values for both the calibrators were below 25%, well within the acceptable range in the CODEX guidelines (Table 2). Hence, pDNA and gDNA proved to be reliable calibrators for use in the real-time PCR detection method employed in this study.

Table 2.

Repeatability for the pDNA and gDNA calibration curve

Target	pDNA calibrator				gDNA calibrator
	Copy no	Mean of Ct value	SD	RSD (%)	Copy no	Mean of Ct value	SD	RSD (%)
Lectin	900,000	19.51	0.37	1.92	88,496	23.16	0.20	0.87
	90,000	23.03	0.38	1.63	8,849.6	26.61	0.27	1.02
	9000	26.60	0.40	1.50	884.96	30.14	0.28	0.94
	900	30.11	0.39	1.29	88.5	33.70	0.31	0.92
	90	33.58	0.39	1.17	8.85	37.28	0.38	1.03
SYHT0H2	900,000	19.68	0.21	1.06	88,496	23.33	0.12	0.52
	90,000	23.17	0.17	0.73	8,849.6	26.86	0.10	0.36
	9000	26.79	0.20	0.74	884.96	30.31	0.11	0.35
	900	30.23	0.20	0.65	88.5	33.85	0.04	0.11
	90	33.69	0.17	0.52	8.85	37.49	0.26	0.68
MON87751	900,000	20.49	0.04	0.21	88,496	23.85	0.02	0.10
	90,000	23.94	0.01	0.03	8,849.6	27.46	0.01	0.02
	9000	27.51	0.03	0.09	884.96	30.96	0.00	0.01
	900	30.96	0.02	0.07	88.5	34.38	0.08	0.24
	90	34.28	0.07	0.19	8.85	37.72	0.24	0.63
DAS-44406-6	900,000	18.66	0.18	0.99	88,496	22.39	0.07	0.31
	90,000	22.17	0.14	0.65	8,849.6	25.83	0.05	0.18
	9000	25.77	0.09	0.35	884.96	29.33	0.03	0.11
	900	29.26	0.10	0.34	88.5	32.99	0.10	0.31
	90	32.76	0.07	0.22	8.85	36.49	0.07	0.20
DAS-81419-2	900,000	18.92	0.00	0.02	88,496	22.60	0.06	0.27
	90,000	22.35	0.01	0.04	8,849.6	26.06	0.07	0.29
	9000	25.93	0.01	0.04	884.96	29.52	0.10	0.34
	900	29.33	0.01	0.04	88.5	33.06	0.11	0.33
	90	32.79	0.03	0.10	8.85	36.66	0.26	0.70

SD repeatability standard deviation, RSD repeatability relative standard deviation

Conversion factor

For accurate quantification in real-time PCR analysis using pDNA as a calibrator, the GM content (%) of samples was calculated using Cf to minimize discrepancies in PCR amplification efficiency between pDNA and gDNA (Meng et al., 2012). C_f was determined as the ratio of the copy number of GM events to the copy number of the endogenous gene (lectin) from the pDNA calibration curve, as described in Eq. (2) in the Materials and Methods section. Table 3 presents the average C_f values obtained in three different experiments, which were 0.96, 0.89, 0.97, and 0.99 for SYHT0H2, MON87751, DAS-44406-6, and DAS-81419-2, respectively. The SD and RSD of these C_f values ranged from 0.02 to 0.16 and from 2.33 to 16.11%. These values were significantly below the acceptable range (RSD ≤ 25%) stipulated in the CODEX guidelines. The application of Cf values in the quantitative real-time PCR analysis of the four GM soybean events ensured accuracy and repeatability.

Table 3.

Conversion factor (C_f) values

Target	Copy no			Mean of Copy no	Cf	SD	RSD (%)
	1	2	3
Lectin	25,170	24,458	26,055	25,228	0.96	0.04	4.22
SYHT0H2	24,654	24,324	23,915	24,297
Lectin	35,435	36,731	36,660	32,276	0.89	0.03	2.88
MON87751	32,475	32,229	31,832	32,179
Lectin	23,255	21,873	24,781	23,303	0.97	0.16	16.11
DAS-44406-6	23,092	24,582	20,138	22,604
Lectin	26,834	24,607	25,586	25,676	0.99	0.02	2.33
DAS-81419-2	26,635	24,934	24,748	25,439

SD repeatability standard deviation, RSD repeatability relative standard deviation

Quantitative analysis of test samples

The quantitative results were evaluated based on accuracy and repeatability of the calculated GM content (%). Accuracy was determined using the formula: Accuracy (%) = tested value/true value × 100, where trueness was measured by assessing bias (%). RSD was computed by dividing the SD by the mean tested GM content (%) for each assay. To evaluate the performance of pDNA and gDNA as reference calibrators in the quantitative analysis of the four GM soybean events, GM samples containing 3.0% and 1.0% of each event were analyzed. DNA from the test samples was obtained either from CRMs or prepared by mixing GM soybean gDNA with NGM soybean gDNA. The GM content (%) of each sample used in the quantitative analysis with pDNA and gDNA calibrators was calculated using Eqs. (3) and (4), respectively, as described in the Materials and Methods section. The quantitative results were evaluated based on accuracy and repeatability (RSD) of the calculated GM content (%). Accuracy was determined using the formula: Accuracy (%) = tested value/true value × 100, where trueness was measured by assessing bias (%). RSD was computed by dividing the SD by the mean tested GM content (%) for each assay. The GM content (%) ranged from 2.74 to 3.55% and from 0.88 to 1.16%, being very close to the actual values of 3.0% and 1.0%, respectively (Table 4). All parameters related to the results of quantitative analysis, including accuracy, repeatability, bias, precision, and RSD, were evaluated and met the criteria outlined in the CODEX guidelines. Notably, there were no significant differences between the calibrated samples, pDNA, and gDNA (Table 4). These findings underscore the suitability and reliability of plasmid calibrators as reference materials for quantifying the four GM soybean events. We successfully demonstrate the efficacy of pDNA calibrators as reliable reference materials for quantitative real-time PCR analysis of four GM soybean events (SYHT0H2, MON87751, DAS-44406-6, and DAS-81419-2). The PCR efficiency and linearity of the test samples, pDNA, and gDNA were comparable, yielding accurate and repeatable quantification results. Plasmid DNA references offer several advantages over gDNA, including ease of amplification by inserting multiple target sequences into organisms, such as Escherichia coli, which results in cost-effective large-scale production. Conversely, obtaining large quantities of gDNA from raw materials is challenging, which leads to potential inconsistencies in quality across different batches and requires substantial effort and time for proper management. Thus, pDNA could be a better alternative to gDNA for quantitative analysis of GMO. In conclusion, the pDNA calibrator and detection method employed in this study can be effectively used for the quantitative analysis of GM soybean events. Moreover, considering that there are no content-specific commercially available CRMs for SYHT0H2 and MON87751 events, the pDNA used in this study could serve as a valuable resource for generating RMs with varying copy numbers. Ongoing efforts are dedicated to developing pDNA RMs for other GMO events.

Table 4.

Quantification results of GM soybean events using pDNA and gDNA calibrators

Target	Calibrator	Reference value (%)	Trueness		Accuracy	Repeatability
			Mean of measured value	Bias (%)	Recovery (%)	RSD (%)
SYHT0H2	pDNA	3.0	3.08	2.56	102.56	8.78
		1.0	1.03	3.33	103.33	8.99
	gDNA	3.0	2.93	− 2.44	97.56	7.18
		1.0	1.00	0.33	100.33	6.48
MON87751	pDNA	3.0	2.74	− 8.67	91.33	4.48
		1.0	0.89	− 10.67	89.33	5.74
	gDNA	3.0	2.83	− 5.67	94.33	0.93
		1.0	0.97	− 3.00	97.00	2.73
DAS-44406-6	pDNA	3.0	3.55	18.22	118.22	11.38
		1.0	1.18	18.00	118.00	3.69
	gDNA	3.0	3.25	8.33	108.33	5.12
		1.0	1.16	16.33	116.33	7.99
DAS-81419-2	pDNA	3.0	3.07	2.33	102.33	0.33
		1.0	0.88	− 11.67	88.33	2.36
	gDNA	3.0	3.11	3.78	103.78	3.83
		1.0	0.95	− 4.67	95.33	6.74

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

MKS and SMJ contributed equally to this work.

Funding

This study was funded by the Ministry of Food and Drug Safety of Korea (Grant Number: 19161MFDS052).

Declarations

Conflict of interest

The authors declare no conflict of interest.