The application of digital PCR to characterize genomic DNA certified reference materials of genetically modified papaya YK1601

ABSTRACT

Certified reference materials (CRMs) are essential in the detection of genetically modified organisms (GMOs), as they ensure the accuracy, comparability, and traceability of test results across time and laboratories, while also supporting international trade and market regulation. In this study, fresh leaves of genetically modified (GM) papaya YK1601 were used to prepare genomic DNA (gDNA) CRMs with transgenic contents of 50% and 5%. The certified values for three property values of this batch of gDNA CRMs were jointly determined by nine qualified laboratories using digital PCR. Homogeneity assessment confirmed good uniformity, with a minimum sample intake of 2 μL. Stability assessment demonstrated that the gDNA CRMs are stable when transported at temperatures below 37°C (preferably under cold-chain conditions) and can be stored at 4°C or −20°C for up to six months, with no more than 10 freeze-thaw cycles permitted during use. This study establishes a metrological benchmark for detecting GM papaya YK1601 and its derived products, while also providing measurement support for evaluating new detection methods and instruments for GM papaya.

1. Introduction

The global cultivation and commercialization of genetically modified (GM) crops have expanded rapidly, with planting areas reaching 209.8 million hectares in 2024 (ISAAA).¹ Although these crops provide significant agricultural, economic, and environmental benefits, public concerns regarding the safety of GM crops and their derived products remain substantial.² To safeguard consumers’ right to know and strengthen the regulation of genetically modified organisms (GMOs), many countries and regions have implemented GMO labeling policies. These regulations mandate clear labeling when the proportion of approved GM components in food or feed exceeds legally defined thresholds – for example, 0.9% in the European Union (EU), 5% in Japan and Canada, and 3% in South Korea.^3–5 In 2023, China also solicited public comments on establishing a 3% labeling threshold for selected GM crops.⁶ The effective implementation of such labeling policies relies on standardized quantitative detection methods and the availability of certified reference materials (CRMs), which ensure accurate calibration and reliable quantification of GM components.^7–9 CRMs are materials with one or more property values that have been determined using rigorous metrological procedures and that are traceable to the International System of Units (SI) with defined uncertainties.¹⁰ Owing to their high accuracy and traceability, CRMs are widely applied in chemistry, biology, medicine, and food safety, both for quantitative analysis and for calibrating new methods and instruments.¹¹ With the continuous advancement of transgenic technology and its growing industrial application, CRMs for GMO detection are expected to play an increasingly important role in biosafety regulation and international trade standardization.

At the international level, the primary organizations developing GMO reference materials (RMs) are the Institute for Reference Materials and Measurements (IRMM) in the EU and the American Oil Chemists’ Society (AOCS) in the United States. IRMM has produced both matrix CRMs and plasmid DNA (pDNA) CRMs, including 158 matrix CRMs covering 38 GM events approved in the EU, as well as pDNA CRMs for maize events MON810, NK603, and 98,140, and for soybean event 356,043 (https://crm.jrc.ec.europa.eu).¹² Meanwhile, AOCS has produced 62 matrix CRMs and 21 genomic DNA (gDNA) CRMs, covering major GM crops currently commercialized worldwide (https://www.aocs.org/crm?SSO=True). Since 2008, China has also advanced the independent development of GM plant RMs. To date, over 60 RMs for GM plants have been certified as national standards and are publicly available through the National Standard Material Resource Sharing Platform (https://www.ncrm.org.cn), covering major crops such as rice, corn, soybean, cotton, and rapeseed.

It is noteworthy that within both domestic and international CRM systems, there are currently no reported CRMs specifically developed for GM papaya. This absence stands in sharp contrast to the significant economic importance of papaya and the commercialized status of its transgenic event, such as YK1601. The YK1601 event was developed through Agrobacterium-mediated transformation, involving the introduction of the coat protein gene from a dominant strain of papaya ringspot virus (PRSV-YK) from Taiwan into the papaya cultivar Tainung No. 2 (TN2).¹³ This modification confers resistance to PRSV, and, in 2018, YK1601 was granted a biosafety certificate for cultivation in suitable ecological zones of southern China. However, the absence of commercially available YK1601 CRMs both domestically and internationally has resulted in a lack of unified and authoritative metrological benchmarks for its detection, making it difficult to meet the demands of safety supervision and testing. In this study, we developed gDNA CRMs for GM papaya YK1601 (Figure 1). These CRMs can be used for both qualitative identification and quantitative analysis of YK1601 and its derived products, serving as calibration and quality control materials for testing laboratories. By ensuring traceability, reliability, and comparability of test results, the CRMs developed in this study provide essential support for GMO biosafety management and the implementation of labeling policies.

Figure 1.

Flowchart for the development of genomic DNA (gDNA) certified reference materials (RMs) of genetically modified papaya YK1601. gDNA was extracted from the leaves of papaya plants, and its copy number concentration was accurately determined using droplet digital PCR (ddPCR). The gDNA of YK1601 and non-transgenic papaya was diluted and mixed in specific ratios to prepare YK1601 gDNA RMs with copy number ratios of 50% and 5%. After homogeneity and stability assessments, as well as collaborative laboratory characterization, three property values of the gDNA RMs were ultimately determined.

2. Materials and methods

2.1. Plant materials

GM papaya YK1601 and its recipient parent TN2 were cultivated at the Teaching Experimental Base of South China Agricultural University. Leaf samples were collected during the vegetative growth stage and used for large-scale gDNA extraction.

2.2. Primer and probe design and synthesis

For the genotyping of GM papaya YK1601, three primer sets were employed: papaya endogenous reference gene primers (Papain-F/Papain-R), YK1601 event-specific primers (YK-F/YK-R), and flanking sequence primers (YK-F/Tai-R) (Figure 2). For quantitative detection, the primer–probe sets Papain-QF/Papain-QR with Papain-QP were used to amplify the papaya endogenous reference gene Papain, while YK-QF/YK-QR with YK-QP were designed for event-specific detection of YK1601 (Table 1). All primers and probes were synthesized by Sangon Biotech Co., Ltd. (Shanghai, China). The same primer–probe sets were applied in both digital PCR (dPCR) and quantitative real-time PCR (qPCR) analyses.

Figure 2.

Schematic diagram of primer design positions for genotype identification of GM papaya YK1601. The blue rectangle represents the gDNA, the black rectangle represents the inserted T-DNA, the red vertical bar indicates the insertion site of the T-DNA, and the black arrows mark the positions of the primers. Among these, YK-F and Tai-R are located on the gDNA, while YK-R is located on the T-DNA. The primer pair YK-F/YK-R is used to detect the YK1601 event-specific sequence, and the primer pair YK-F/Tai-R is used to detect the flanking sequences of the insertion site.

Table 1.

Primer and probe used in this study.

Target	Primer/probe	Sequence (5’−3’)	Amplicon size (bp)	Source
Papain	Papain-F	GGGCATTCTCAGCTGTTGTA	211	Jiang et al., 202014
	Papain-R	CGACAATAACGTTGCACTCC
	Papain-QF	TGCTTGACTGCGACAGAC	144
	Papain-QR	CTTTCTCCCTTGAGCGAC
	Papain-QP	FAM-AGCTACGGGTGCAATGGAGGTTACC-BHQ1
YK1601	YK-F	GGTATCACCCCAGTCTTAGTTTG	250	Li et al., 202015
	YK-R	CTTCAACGTTGCGGTTCTGT
	YK-QF	GGATACTATCTCATTCAACACTG	90	This study
	YK-QR	ATAACGTGACTCCCTTAATTCTC
	YK-QP	HEX-TGTCGTTTCCCGCCTTCAGTTTA-BHQ1		Li et al., 202015
Insert site	YK-F	GGTATCACCCCAGTCTTAGTTTG	193	This study
	Tai-R	CAGAAAGACATATATCATCAAGACCA

2.3. Preparation of YK1601 gDNA RMs

Fresh leaves of YK1601 and TN2 were ground in liquid nitrogen, and gDNA was extracted and purified using the DNeasy® Plant Mini Kit (Qiagen, Dusseldorf, Germany). The extracted DNA was dissolved in 1×TE buffer. First, the purity of gDNA was assessed using ultraviolet spectrophotometry. Then, its integrity and semi-quantitative concentration were evaluated by agarose gel electrophoresis with DNA standards of known concentrations. Finally, the copy number concentrations of YK1601 and TN2 gDNA were precisely measured using droplet digital PCR (ddPCR), with each group independently tested at least three times. Based on these results, mixtures of YK1601 and TN2 gDNA were prepared to obtain RM solutions with 50% and 5% copy number ratios. To assess initial homogeneity, subsamples (n = 9) were collected from the upper, middle, and lower layers of each RM solution at 6, 12, and 18 h post-preparation and analyzed by ddPCR.

2.4. Conventional PCR

Conventional PCR was conducted on a T100™ Thermal Cycler (Bio-Rad, Hercules, CA, USA) in a 25 μL reaction mixture containing 1×PCR Buffer (Mg²⁺ plus), 0.625 U Taq™ HS DNA polymerase (Takara Bio Inc, Osaka, Japan, 5 U/μL), 0.4 μmol/L of each primer, 0.1 mmol/L dNTP mixture, and 20 ng of DNA template. The amplification program consisted of an initial denaturation at 94°C for 3 min, followed by 35 cycles of 94°C for 30 s, 58°C for 30 s, and 72°C for 30 s, and a final extension at 72°C for 5 min. The PCR products were analyzed by 2% agarose gel electrophoresis.

2.5. qPCR

qPCR analysis was conducted according to the method of Zhu et al. using the CFX Connect™ Real-Time System (Bio-Rad, Hercules, CA, USA).¹⁶ Each 20 μL reaction mixture contained 10 μL 2×Premix Ex Taq™ (Probe qPCR) (Takara Bio Inc., Osaka, Japan), 0.4 μmol/L of each primer, 0.2 μmol/L probe, and 2 μL DNA template. The thermal cycling program consisted of an initial denaturation at 95°C for 2 min, followed by 45 cycles of 95°C for 15 s and 60°C for 60 s. Fluorescence signals were collected and analyzed with CFX Manager software.

2.6. ddPCR

ddPCR analysis was performed using the QX200™ Droplet Digital PCR system. Each 20 μL reaction mixture contained 10 μL 2×ddPCR™ Supermix for Probes (Bio-Rad, Pleasanton, CA, USA), 0.9 μmol/L of each primer, 0.25 μmol/L probe, and 2 μL DNA template. The 20 μL mixture was combined with 70 μL droplet generation oil in a droplet generation cartridge, sealed with a gasket, and processed in the droplet generator (Bio-Rad, Pleasanton, CA, USA). The resulting droplets were transferred to a 96-well PCR plate and amplified on a T100™ Thermal Cycler under the following conditions: 95°C for 10 min, followed by 40 cycles of 94°C for 30 s and 56.5°C for 60 s, a final step at 98°C for 10 min, and a final hold at 4°C, with a ramp rate of 2 °C/s. After amplification, droplets were read using the droplet reader (Bio-Rad, Pleasanton, CA, USA), and data were analyzed with QuantaSoft v1.7.4 software.

2.7. Homogeneity Assessment

The homogeneity of the RMs was evaluated in accordance with ISO Guide 35:2017.¹⁷ Twelve units each of the 50% and 5% YK1601 gDNA RMs were randomly selected. From every unit, three subsamples were collected and analyzed by ddPCR. Statistical analysis was performed using analysis of variance (ANOVA), calculating the mean square difference between units (MS_between) and the mean square difference within units (MS_within). The ratio of these values (F-value) was compared with the critical value F_α at a significance level of α = 0.05. Homogeneity was confirmed when F < F_α. The relative standard uncertainty due to between-unit heterogeneity (u_{rel (bb)}) was then calculated according to ISO Guide 35:2017.

2.8. Stability Assessment

The stability of the 50% and 5% YK1601 gDNA RMs was evaluated through isochronous stability studies, following ISO Guide 35:2017.^17,18 Three types of stability were examined: short-term, long-term, and freeze-thaw. For the short-term stability assessment, nine tubes of gDNA RMs were randomly selected and divided into three groups based on their storage temperatures (4°C, 25°C, or 37 °C), with three tubes per group. These tubes were stored for 0, 3, 7, or 14 days. For the long-term stability assessment, six additional tubes of gDNA RMs were randomly selected, with three stored at 4°C and three at −20°C, for durations of 0, 1, 3, or 6 months. At each time point, two subsamples (4 µL each) were taken from each tube, transferred into new tubes, and stored at −20°C until analysis. For freeze-thaw stability, five tubes of gDNA RMs were subjected to 0, 5, or 10 freeze-thaw cycles. One cycle consisted of removing samples from −20°C storage, thawing them at 25°C in a water bath, thoroughly mixing, and refreezing at −20°C. Each cycle was completed every two days. After each designated cycle, 4 µL subsamples were taken from each tube, transferred into new tubes, and stored at −20°C until analysis. After collection, all samples were analyzed simultaneously by the same operator using the same ddPCR instrument, with each sample tested in triplicate. Data were statistically analyzed, and stability was assessed using t-tests. The relative standard uncertainties associated with short-term stability (u_{rel (sts)}) and long-term stability (u_{rel (lts)}) were calculated according to ISO Guide 35:2017.

2.9. Collaborative Characterization

Eight laboratories in China, all accredited for GMO detection, participated in a collaborative dPCR-based value assignment of the 50% and 5% YK1601 gDNA RMs. Of these, seven used the Bio-Rad Q200 Droplet Digital PCR system, and one used the chip-based Applied Biosystems™ QuantStudio™ 3D Digital PCR system. The RMs, together with the corresponding primer–probe sets, were distributed to the participating laboratories via dry ice cold-chain transport. Each laboratory prepared its own dPCR reagents but followed a unified protocol to quantify the copy number concentrations of the YK1601 event-specific sequence and the endogenous reference gene Papain, and to calculate their copy number ratio. Each RM was analyzed in at least eight replicates, and each laboratory was required to submit a minimum of eight independent datasets for each property value. The raw dPCR results were centrally compiled and statistically analyzed by the coordinating laboratory.

The relative standard uncertainty from the characterization process (u_{rel (char)}) consisted of four components: the relative standard uncertainty from precision data (u_{rel (A)}), the relative standard uncertainty associated with the dilution factor (u_{rel (d)}), the relative standard uncertainty due to droplet volume variation (u_{rel (v)}),¹⁹ and the relative standard uncertainty from the method detection limit (u_{rel (LOD)}). These components were combined to calculate the total measurement uncertainty according to Equation (1):

$$u_{\mathit{rel}\mathit{\hspace{0.17em}}\left(\mathit{char}\right)}=\sqrt{{{\mathit{u}}_{\mathit{rel}\mathit{\hspace{0.17em}}\left(A\right)}}^2+{{\mathit{u}}_{\mathit{rel}\mathit{\hspace{0.17em}}\left(d\right)}}^2+{{\mathit{u}}_{\mathit{rel}\mathit{\hspace{0.17em}}\left(v\right)}}^2+{{\mathit{u}}_{\mathit{rel}\mathit{\hspace{0.17em}}\left(\mathit{LOD}\right)}}^2}$$ (1)

2.10. Estimation of Expanded Uncertainty

The combined relative standard uncertainty of the certified value (u_{rel (CRM)}) was derived from four sources: the relative standard uncertainty from the characterization process (u_{rel (char)}), the relative standard uncertainty due to potential between-unit heterogeneity (u_{rel (bb)}), the relative standard uncertainty from potential short-term transport instability (u_{rel (sts)}), and the relative standard uncertainty from potential long-term storage instability (u_{rel (lts)}). These four components were combined according to Equation (2). The expanded uncertainty of the certified value, U_(CRM), was then calculated by multiplying the combined relative standard uncertainty by the coverage factor k (where k =2 at a 95% confidence level) and the overall mean value y obtained from the characterization experiments:

$$U_{\left(\mathit{CRM}\right)}=\mathit{k}\times \sqrt{{\left({\mathit{u}}_{\mathit{rel}\mathit{\hspace{0.17em}}\left(\mathit{char}\right)}\right)}^2+{\left({\mathit{u}}_{\mathit{rel}\mathit{\hspace{0.17em}}\left(\mathit{bb}\right)}\right)}^2+{\left({\mathit{u}}_{\mathit{rel}\mathit{\hspace{0.17em}}\left(\mathit{sts}\right)}\right)}^2+{\left({\mathit{u}}_{\mathit{rel}\mathit{\hspace{0.17em}}\left(\mathit{lts}\right)}\right)}^2}\times \mathit{y}$$ (2)

3. Results

3.1. Identification of Candidate Plant Materials

To detect the presence of GM components in the raw materials, qPCR was performed on GM papaya YK1601 and non-GM papaya TN2. In YK1601, typical amplification curves were generated using both Papain-QF/QR/QP and YK-QF/QR/QP primer and probe sets, whereas in TN2, amplification curves were observed only with Papain-QF/QR/QP (Figure S1). These results confirmed the presence of the expected GM component in YK1601 and its absence in TN2. The zygosity of YK1601 plants was subsequently assessed using conventional PCR and ddPCR. In this study, the reaction conditions of the ddPCR method were systematically optimized, including annealing temperature, cycling parameters, and primer/probe concentrations. The optimization results revealed that the system exhibited optimal performance under the conditions of an annealing temperature of 56.5°C, 40 amplification cycles, a primer concentration of 0.9 μmol/L, and a probe concentration of 0.25 μmol/L (Figures S2, S3, S4). Conventional PCR indicated that Papain-F/R and YK-F/Tai-R amplified the expected bands in both TN2 and YK1601, whereas YK-F/R produced the expected band only in YK1601 (Figure 3), suggesting YK1601 plants were heterozygous. Consistently, ddPCR showed that the copy number ratio of Papain to the event-specific sequence in YK1601 gDNA was approximately 2:1, further confirming the heterozygous status of YK1601 plants (Figure 4).

Figure 3.

PCR electrophoresis profiles for genotype identification of GM papaya YK1601. A: detection of the papaya endogenous reference gene using the Papain-F/R primer pair; B: detection of the YK1601 event-specific sequence using the YK-F/R primer pair; C: detection of the flanking sequences of the insertion site using the YK-F/Tai-R primer pair. M: DNA Marker DL 2000; 1: tainung No. 2 (TN2); 2: YK1601. The results show that both the Papain-F/R and YK-F/Tai-R primer pairs amplified bands of the expected size in both TN2 and YK1601 plants. In contrast, the YK-F/R primer pair only amplified a band of the expected size in YK1601 plants, indicating that the YK1601 plant is heterozygous.

Figure 4.

Genotype identification of GM papaya YK1601 by droplet digital PCR. A: detection of the papaya endogenous reference gene using the Papain-QF/QR/QP primer and probe set; B: detection of the YK1601 event-specific sequence using the YK-QF/QR/QP primer and probe set. The results show that the ratio of the copy number of the Papain gene to that of the event-specific sequence in the gDNA of YK1601 plants is approximately 2:1, further confirming that the YK1601 plant is heterozygous.

3.2. Preparation of gDNA RMs

To prepare gDNA RMs, gDNA was extracted from fresh leaves of TN2 and YK1601 plants. DNA integrity was evaluated using 2% agarose gel electrophoresis, which showed clear bands without smearing or RNA contamination and largely consistent fragment sizes, indicating good integrity (Figure S5). UV spectrophotometry revealed that TN2 gDNA had a concentration of ~ 290 ng/µL, with OD₂₆₀/OD₂₈₀ = 1.96 and OD₂₆₀/OD₂₃₀ = 2.01, whereas YK1601 gDNA had a concentration of ~ 265 ng/µL, with OD₂₆₀/OD₂₈₀ = 1.98 and OD₂₆₀/OD₂₃₀ = 2.00. These results confirmed that both the concentration and quality of the extracted gDNA met the requirements for subsequent experiments.

To evaluate the presence of potential PCR inhibitors (e.g., organic solvents, RNA, proteins, or other trace impurities), gradient dilutions of TN2 and YK1601 gDNA were analyzed by qPCR. Replicates at each concentration produced consistent Ct values (Table S1) and typical amplification curves (Figure S6A, C). Based on the slope of the standard curve, the amplification efficiencies for the Papain gene detection were calculated as 99.9% and 100.0%, respectively (Figure S6B, D). These results indicated the absence of significant PCR inhibitors and confirmed the suitability of TN2 and YK1601 gDNA for RM preparation.

The initial homogeneity of the prepared gDNA RMs was then assessed by collecting nine subsamples from different locations of the mixture at 6-hour intervals and analyzing them by ddPCR. For both the 50% and 5% YK1601 gDNA RMs, the copy number concentrations of the YK1601 event and Papain, as well as their ratios, yielded F-values lower than the critical F_0.05(8,18), with no significant differences detected across time points (Table S2). These findings demonstrated that the gDNA mixtures exhibited good homogeneity and achieved ideal uniformity after 6 h of mixing. Finally, the 50% and 5% YK1601 gDNA mixtures were aliquoted into 200 vials each (100 μL per vial) and stored at −20°C.

3.3. Homogeneity Assessment

To evaluate the homogeneity of the two RMs after aliquoting, unit-to-unit variation was assessed for the 50% and 5% YK1601 gDNA RMs using three parameters: the copy number concentration of the YK1601 event, the copy number concentration of the Papain gene, and their copy number ratio. Grubbs’ test was first applied to identify outliers in the homogeneity dataset, and no outliers were detected at the 95% confidence level. Then, F-values were calculated as the ratio of MS_between to MS_within. For all parameters in both RMs, the F-values were lower than the critical value F_0.05(11,24) at the 95% confidence level (Table 2), confirming good unit-to-unit homogeneity after aliquoting. The relative standard uncertainty introduced by potential between-unit heterogeneity (u_{rel (bb)}) was also estimated (Table 2).

Table 2.

Results of the homogeneity assessment of 50% and 5% YK1601 gDNA RMs.

Parameter	50% YK1601 gDNA RM			5% YK1601 gDNA RM
	YK1601 (copies/μL)	Papain (copies/μL)	YK1601/Papain (copies/μL)	YK1601 (copies/μL)	Papain (copies/μL)	YK1601/Papain (copies/μL)
Mean	1.24E+05	2.51E+05	4.95E-01	1.23E+04	2.44E+05	5.03E-02
$\mathit{M}{S}_{\mathrm{between}}$	2.39E+06	5.28E+06	2.08E-05	1.05E+05	3.79E+06	2.30E-06
$\mathit{M}{S}_{\mathrm{within}}$	1.86E+06	3.69E+06	1.98E-05	8.81E+04	2.19E+06	1.87E-06
F	1.29	1.43	1.05	1.19	1.72	1.23
F0.05(11,24)	2.18	2.18	2.18	2.18	2.18	2.18
sbb	421.44	727.07	–	74.09	728.22	–
swb	1.36E+03	1.92E+03	4.45E-03	2.97E+02	1.48E+03	1.37E-03
ubb	421.44	727.07	–	74.09	728.22	–
$u_{\mathrm{rel}\left(\mathrm{bb}\right)}$	0.339%	0.290%	0.446%	0.603%	0.298%	0.673%

In addition, during homogeneity assessment, 2 µL of gDNA solution was used per subsample. The results confirmed that both the 50% and 5% YK1601 gDNA RMs exhibited good homogeneity at this sampling volume, establishing 2 µL as the minimum sample intake.

3.4. Stability Assessment

To determine appropriate short-term transport and long-term storage conditions for the 50% and 5% YK1601 gDNA RMs, short-term, long-term, and freeze-thaw stability tests were conducted. Since the RMs were transported in ice boxes and stored in refrigerators as gDNA solutions, the effects of light and radiation were excluded, leaving time and temperature as the primary influencing factors. Grubbs’ test was first applied to identify outliers in the stability dataset. No outliers were detected at the 95% confidence level. Then, the regression slopes of the YK1601 event copy number concentration, Papain copy number concentration, and their ratio against storage time were calculated to assess trends under different temperatures. The statistical significance of these slopes was evaluated using a t-test. In the short-term stability test, both gDNA RMs stored at 4°C, 25°C, and 37°C for 14 days yielded t_β1 values smaller than the critical value t_{0.95, 3} (Table 3). Similarly, in the long-term stability test, both RMs stored at 4°C and −20°C for 6 months also showed t_β1 < t_{0.95, 3} (Table 4). In the freeze-thaw stability test, after 10 cycles, both RMs still exhibited t_β1 < t_{0.95, 3} (Table S3). At the 95% confidence level, no significant slopes were observed under any of the test conditions, indicating that none of the three measured parameters exhibited significant time-dependent changes. Based on these findings, the 50% and 5% YK1601 gDNA RMs were confirmed to be stable for transportation at ambient temperatures below 37°C for up to 14 days and for storage at −20°C for up to 6 months. In addition, 10 freeze-thaw cycles were established as the maximum allowable usage for this batch of RMs.

Table 3.

Results of the short-term stability test of 50% and 5% YK1601 gDNA RMs.

RM	Storage Time (Days)	YK1601 (copies/μL)			Papain (copies/μL)			YK1601/Papain (copies/μL)
		4°C	25°C	37°C	4°C	25°C	37°C	4°C	25°C	37°C
50% YK1601 gDNA RM	0	1.28E+05	1.28E+05	1.27E+05	2.53E+05	2.53E+05	2.53E+05	5.05E-01	5.05E-01	5.04E-01
	3	1.28E+05	1.27E+05	1.28E+05	2.54E+05	2.53E+05	2.53E+05	5.04E-01	5.03E-01	5.05E-01
	7	1.28E+05	1.28E+05	1.27E+05	2.55E+05	2.55E+05	2.52E+05	5.01E-01	5.01E-01	5.03E-01
	14	1.28E+05	1.28E+05	1.29E+05	2.53E+05	2.54E+05	2.55E+05	5.05E-01	5.06E-01	5.04E-01
	Mean	1.28E+05	1.28E+05	1.28E+05	2.54E+05	2.54E+05	2.53E+05	5.04E-01	5.04E-01	5.04E-01
	β1	3.03	53.03	74.24	9.09	54.55	54.55	−0.00001	−0.00001	−0.00007
	β0	127773	127432	127263	253612	253423	253048	0.50391	0.50364	0.50449
	s(β1)	18.51	41.84	76.97	96.45	110.79	131.99	0.00020	0.00020	0.00012
	$t_{0.95,\mathit{n}-2}$	4.30	4.30	4.30	4.30	4.30	4.30	4.30	4.30	4.30
	$t_{{\beta }_1}=\frac{\left|{\beta }_1\right|}{\mathit{s}\left({\beta }_1\right)}$	0.16	1.27	0.96	0.09	0.49	0.41	0.06	0.06	0.61
	u(sts)	–	–	1077.60	–	–	1847.92	–	–	–
	urel (sts)	–	–	0.844%	–	–	0.729%	–	–	–
5% YK1601 gDNA RM	0	1.24E+04	1.23E+04	1.25E+04	2.47E+05	2.44E+05	2.45E+05	5.02E-02	5.03E-02	5.08E-02
	3	1.25E+04	1.24E+04	1.23E+04	2.47E+05	2.45E+05	2.45E+05	5.04E-02	5.05E-02	5.01E-02
	7	1.24E+04	1.23E+04	1.24E+04	2.46E+05	2.45E+05	2.46E+05	5.03E-02	5.02E-02	5.05E-02
	14	1.24E+04	1.23E+04	1.22E+04	2.46E+05	2.44E+05	2.41E+05	5.05E-02	5.05E-02	5.05E-02
	Mean	1.24E+04	1.23E+04	1.23E+04	2.46E+05	2.45E+05	2.44E+05	5.04E-02	5.04E-02	5.05E-02
	β1	−0.91	2.58	−16.82	−87.88	31.82	31.82	0.00001	0.00001	0.00000
	β0	12401	12305	12413	246694	244351	243809	0.05027	0.05031	0.05049
	s(β1)	4.82	4.51	10.82	49.52	80.04	283.70	0.00001	0.00001	0.00003
	$t_{0.95,\mathit{n}-2}$	4.30	4.30	4.30	4.30	4.30	4.30	4.30	4.30	4.30
	$t_{{\beta }_1}=\frac{\left|{\beta }_1\right|}{\mathit{s}\left({\beta }_1\right)}$	0.19	0.57	1.55	1.77	0.40	0.11	1.19	1.19	0.13
	u(sts)	–	–	151.50	–	–	2249.46	–	–	–
	urel (sts)	–	–	1.230%	–	–	0.920%	–	–	–

Table 4.

Results of the long-term stability test of 50% and 5% YK1601 gDNA RMs.

RM	Storage Time (Months)	YK1601 (copies/μL)		Papain (copies/μL)		YK1601/Papain (copies/μL)
		4°C	-20°C	4°C	-20°C	4°C	-20°C
50% YK1601 gDNA RM	0	1.26E+05	1.27E+05	2.53E+05	2.54E+05	5.00E-01	5.01E-01
	1	1.26E+05	1.27E+05	2.52E+05	2.54E+05	5.01E-01	4.99E-01
	3	1.26E+05	1.27E+05	2.53E+05	2.54E+05	4.98E-01	5.00E-01
	6	1.27E+05	1.27E+05	2.54E+05	2.54E+05	5.01E-01	4.98E-01
	Mean	1.26E+05	1.27E+05	2.53E+05	2.54E+05	5.00E-01	5.00E-01
	β1	146.83	−103.17	214.29	3.97	0.000158	−0.000426
	β0	126091.27	127091.27	252380.95	253865.08	0.500	0.501
	s(β1)	74.77	62.74	82.48	42.55	0.000313	0.000238
	$t_{0.95,\mathit{n}-2}$	4.30	4.30	4.30	4.30	4.30	4.30
	$t_{{\beta }_1}=\frac{\left|{\beta }_1\right|}{\mathit{s}\left({\beta }_1\right)}$	1.96	1.64	2.60	0.09	0.50	1.79
	u(lts)	897.21	–	989.74	–	–	–
	urel (lts)	0.709%	–	0.391%	–	–	–
5% YK1601 gDNA RM	0	1.25E+04	1.23E+04	2.45E+05	2.45E+05	5.10E-02	5.01E-02
	1	1.26E+04	1.23E+04	2.46E+05	2.44E+05	5.13E-02	5.05E-02
	3	1.25E+04	1.24E+04	2.47E+05	2.46E+05	5.08E-02	5.03E-02
	6	1.25E+04	1.24E+04	2.45E+05	2.46E+05	5.10E-02	5.04E-02
	Mean	1.25E+04	1.23E+04	2.46E+05	2.45E+05	5.10E-02	5.04E-02
	β1	−0.40	15.48	79.37	166.67	−0.00002	0.00003
	β0	12530.16	12307.14	245384.92	244750.00	0.05108	0.05028
	s(β1)	11.13	6.63	199.60	172.52	0.00005	0.00004
	$t_{0.95,\mathit{n}-2}$	4.30	4.30	4.30	4.30	4.30	4.30
	$t_{{\beta }_1}=\frac{\left|{\beta }_1\right|}{\mathit{s}\left({\beta }_1\right)}$	0.04	2.33	0.40	0.97	0.41	0.75
	u(lts)	133.59	–	2395.20	–	–	–
	urel (lts)	1.066%	–	0.975%	–	–	–

The relative standard uncertainties introduced by potential instability during short-term transport (u_{rel (sts)}) and long-term storage (u_{rel (lts)}) were also estimated (Tables 3 and 4). The value of u_{rel (sts)} was based on experimental data obtained from 14 days of storage at 37°C, while u_{rel (lts)} was estimated from 6 months of storage at 4°C.

3.5. Collaborative Characterization of the Property Values

The property values of the 50% and 5% YK1601 gDNA RMs were collaboratively assigned by our laboratory in conjunction with eight other qualified laboratories using the dPCR method. Statistical analysis of the results showed no intra-laboratory outliers and no anomalies in inter-laboratory means or standard deviations. At the 95% confidence level, all datasets followed a normal distribution, indicating good consistency among the participating laboratories. Therefore, the arithmetic mean of the results from all nine laboratories was adopted as the certified value for both gDNA RMs (Table S4). All relevant components of relative standard uncertainty in the characterization process (u_{rel (A)}, u_{rel (d)}, u_{rel (v)}, and u_{rel (LOD)}) were comprehensively evaluated. These components were then combined according to Equation (1) to obtain the total relative standard uncertainty of characterization (u_{rel (char)}) (Table 5).

Table 5.

Estimation of the uncertainty contributions in the characterization of 50% and 5% YK1601 gDNA RMs.

RM	Property Value	urel (A)	urel (d)	urel (v)	urel (LOD)	urel (char)
50% YK1601 gDNA RM	YK1601 (copies/μL)	4.546%	1.446%	1.000%	0.004%	4.874%
	Papain (copies/μL)	5.135%	1.446%	1.000%	0.002%	5.427%
	YK1601/Papain (copies/μL)	6.858%	1.446%	1.000%	0.004%	7.080%
5% YK1601 gDNA RM	YK1601 (copies/μL)	5.006%	1.446%	1.000%	0.043%	5.306%
	Papain (copies/μL)	3.895%	1.446%	1.000%	0.002%	4.273%
	YK1601/Papain (copies/μL)	6.343%	1.446%	1.000%	0.043%	6.582%

3.6. Evaluation of Expanded Uncertainty

The four relative standard uncertainty components—u_{rel (char)}, u_{rel (bb)}, u_{rel (sts)}, and u_{rel (lts)}—were combined to derive the relative standard uncertainty of the certified value (u_{rel (CRM)}). Based on Equation (2), the expanded uncertainty (U_CRM) was then calculated (Table 6). The final certified values were as follows: for the 50% YK1601 gDNA RM, the YK1601 event copy number concentration was (1.25 ± 0.13) × 10⁵ copies/μL, the Papain copy number concentration was (2.50 ± 0.28) × 10⁵ copies/μL, and their ratio was (5.01 ± 0.72) × 10⁻¹. For the 5% YK1601 gDNA RM, the YK1601 event copy number concentration was (1.17 ± 0.13) × 10⁴ copies/μL, the Papain copy number concentration was (2.23 ± 0.21) × 10⁵ copies/μL, and their ratio was (5.06 ± 0.70) × 10⁻² (Table 6).

Table 6.

The certified values and their uncertainties for the three property values of 50% and 5% YK1601 gDNA RMs.

RM	Property Value	Certified Value	urel (char)	urel (bb)	urel (sts)	urel (lts)	urel (CRM)	UCRM
50% YK1601 gDNA RM	YK1601 (copies/μL)	1.25E+05	4.874%	0.339%	0.844%	0.709%	5.009%	1.25E+04
	Papain (copies/μL)	2.50E+05	5.427%	0.290%	0.729%	0.391%	5.497%	2.75E+04
	YK1601/Papain (copies/μL)	5.01E-01	7.080%	0.446%	1.115%	0.810%	7.227%	7.24E-02
5% YK1601 gDNA RM	YK1601 (copies/μL)	1.17E+04	5.306%	0.603%	1.230%	1.066%	5.583%	1.31E+03
	Papain (copies/μL)	2.32E+05	4.273%	0.298%	0.920%	0.975%	4.488%	2.08E+04
	YK1601/Papain (copies/μL)	5.06E-02	6.582%	0.673%	1.536%	1.445%	6.944%	7.03E-03

4. Discussion

The rapid advancement of transgenic technology and the increasing commercialization of GM crops have elevated concerns about the potential risks of food and feed products derived from these crops. To safeguard consumers’ right to know and to choose, many countries and regions have implemented labeling regulations for GM products, requiring mandatory disclosure when the content of approved GM components exceeds legally defined thresholds.^3–5 A key element in the effective implementation of these labeling policies is the availability of standardized detection technologies capable of accurately quantifying GM components. The reliability of such detection systems depends on ensuring that test results are accurate, comparable, and traceable. Within this framework, the development of CRMs has become a cornerstone for standardizing GM detection technologies.²⁰ To meet the technical requirements for GM product labeling, several CRMs for quantitative detection of GM components have been developed and certified. For example, Li et al. prepared gDNA CRMs using transgenic rice Kefeng 6 as calibration standards,²¹ while Chen et al. developed gDNA CRMs from glyphosate-resistant GM soybean ZUTS-33 for detecting the g10evo-epsps gene in this soybean and its products.²² Although the raw materials for these CRMs had obtained safety certification, they had not yet been applied commercially. In contrast, GM papaya YK1601 has been cultivated and applied in suitable ecological regions of South China since 2018; however, corresponding CRMs for its detection remain unavailable domestically. Addressing this gap is critical to meet the urgent needs of regulatory authorities in safety supervision and testing, and to provide technical support for the biosafety management of GMOs.

In this study, we developed gDNA CRMs for GM papaya YK1601 using heterozygous materials. We are confident that the heterozygous state does not affect the accuracy or applicability of the CRM as a calibrant under the calibration strategy adopted. The core reason is that we employed high-precision dPCR to directly determine the copy number ratio of the exogenous gene to the endogenous reference gene. Based on this precisely determined ratio, we prepared CRMs with nominal GMO contents of 50% and 5%. The essence of this calibration method lies in the determination of copy number ratios, a measurement that is fundamentally independent of the zygosity of the material. To ensure the scientific validity and reliability of the value assignment, we employed dPCR as the core method. dPCR is recognized as a benchmark measurement method for determining DNA copy number concentration. Through micro-partitioning, dPCR enables absolute quantification without relying on standard curves, thereby significantly reducing systematic errors. This makes it particularly suitable for accurately quantifying low-abundance samples, such as those with less than 1% GM components.^23,24 Based on this, we adopted a rigorous metrological approach involving a multi-laboratory collaboration. Using dPCR, these laboratories precisely determined the copy number ratio of the exogenous gene to the endogenous reference gene and comprehensively evaluated the associated measurement uncertainty, thereby ensuring the accuracy, validity, and traceability of the certified values and their uncertainties.

Therefore, this batch of CRMs, whose certified values were determined through the optimized ddPCR method, not only provides a traceable calibration benchmark for the current PCR-based detection systems but also possesses broader application potential. With the rapid development of disruptive nucleic acid detection technologies such as CRISPR (e.g., the VTD-CRISPR platform), detection methods are evolving toward being faster, simpler, and more suitable for on-site applications.^25–27 For these emerging technologies to transition from laboratory research to standardized application, the accuracy and quality control of their results must be addressed. The CRMs developed in this study can directly serve as calibrators for the performance evaluation and calibration of such emerging platforms, ensuring that they produce accurate, comparable, and traceable data even within simplified procedures. Currently, this batch of gDNA CRMs has been made available on the National Standard Material Resource Platform. Their application in the calibration and validation of YK1601 quantitative detection further enhances the scientific value and utility of reference materials in supporting GMO labeling management.

Funding Statement

This work was supported by the Biological Breeding-Major Projects [2023ZD04062].

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Supplementary Material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/21645698.2026.2620130