Introduction

In the European Union (EU), the import and use of genetically modified organisms (GMOs) is strictly regulated. Currently, March 2016, 66 GM crop plant varieties are authorised in the EU (10 × cotton, 37 × maize, 6 × canola, 12 × soybean and 1 × sugar beet) [19]. EU-authorised GMOs and derived materials have to be labelled as such [19]. Some additional rules apply for GMO-derived feed materials. Under Regulation (EC) No 619/2011 (European [17], GMOs that are in the process of being authorised (currently 20 GM plants) are allowed up to 0.1 % in feed if they have already been authorised in another country and if reference material and a detection method are available. When a product or ingredient is not labelled as GMO-derived, EU-authorised GMOs are allowed to be present in the ingredient up to 0.9 %, if this presence is unintended and cannot be technically avoided. Some GMOs that are withdrawn from the market or where the market approval has expired are allowed up to 0.1 % in feed (European Commission, [1719]. All other GMOs are not authorised for the EU market (0 % tolerance).

Specific detection methods are needed for enforcement of these labelling requirements. TaqMan detection methods that target the authorised GM events are delivered by the respective applicants, verified by the European Union Reference Laboratory for Genetically Modified Food and Feed (EURL-GMFF) and interlaboratory validated by the EURL-GMFF together with the European Network of GMO Laboratories (ENGL) [7]. Specific targets for detection strategies are GMO-related genetic elements (coding sequences (CS-), promoters (P-) or terminators (T-)), constructs (the junction sequences bridging adjacent elements) and events (area spanning the 5′ or 3′ insertion site and the flanking plant region). Because already many GMOs are allowed on the EU market, screening strategies targeting GMO elements and constructs are applied to limit the number of event-specific methods that need to be performed [14, 37, 43]. Also, pre-spotted screening and event plates have been developed [34, 36] to facilitate the screening and identification of GMOs in food and feed products.

Depending on the type of samples to be tested, more or fewer elements and constructs can be selected for screening. Screening with, for example, five detection methods as described by Waiblinger et al. [43] would be a good option to cover the majority of known GMOs in samples for which the absence of known GMOs has to be confirmed. For samples that are labelled as GMO and that may contain several different GMOs (e.g. feed), an extended element and construct screening can be more informative, as here the focus will be primarily on the detection and identification of unauthorised GMOs [13, 37]. If specific detected elements cannot be explained by the subsequently identified events, this can be an indication for the presence of unauthorised GMOs in the sample. The EUginius GMO database (www.euginius.eu) provides an analysis tool that can aid the analyses based on the GMO elements that have been identified and those that have been found negative in the screening. The JRC GMO-Matrix event finder [1] provides a comparable identification strategy. Screening strategies are also described in a CEN Technical Standard NPR-CEN/TS 16707:2014 en [12], and a review on this subject has been written by Fraiture et al. [20].

Where the detection of promoter P-35S from Cauliflower mosaic virus (CaMV) or terminator T-nos from Rhizobium radiobacter was once sufficient to detect most authorised GMOs, there now is a tendency to use a wide variety of promoters, coding genes and terminators in more recent GMOs. Also, the use of plant-derived GMO elements requires the use of construct-specific methods instead of element-specific methods to avoid false positives in feed samples containing multiple species. Therefore, additional detection methods are sometimes necessary in order to target specific (groups of) GMOs. Some recent examples of additional qPCR screening methods are the element bar and construct ctp2/cp4epsps [21], the elements cry1A.105 and cry2Ab2 [15], promoters P-FMV, P-nos, P-SSuAra, P-TA29, P-ubi, P-rice actin and terminators T-35S, T-E9, T-ocs, T-g7 [14], element vip3A [29], and element cry1Ab/Ac and construct P-ubi/cry [22]. Recent examples of SYBR® Green detection methods are P-nos and P-FMV [9], cry3Bb and gat/T-pinII [10], and the application of Combinatory SYBR® Green PCR Screening (CoSYPS) [4, 30, 42]. Detection and identification of GMOs is not static. Progressive insight, changes in GMO authorisations, development of novel GMOs with sometimes (plant-derived) elements for which no suitable detection methods are available, regularly require adjusted adequate detection strategies. Nevertheless, there are still GMOs for which no suitable screening elements are available.

Here, we describe the development of TaqMan PCR detection methods for the coding sequence cry3A and the construct gat/T-pinII to facilitate the detection of present and future GMOs containing these elements. The element cry3A and construct gat/T-pinII were selected as targets since they are present in GMOs that harbour zero or one element that can be detected in the element screening, which is not sufficient for the identification of the GMO(s) present. The cry3A method was initially intended for the screening for MIR604 maize (SYN-IR6Ø4-5) and 5307 maize (SYN-Ø53Ø7-1), but may also detect other GMOs containing cry3A. At least 27 potato events that are deregulated in the USA [2], the Russian 2904/1 kgs (ISAAA’s GM Approval Database [24]), a GM potato developed in Bulgaria [26] and two GM potatoes developed in China [23, 31] contain cry3A or variants thereof. For other GMOs like the currently unauthorised MZIR098 (SYN-ØØØ98-3) containing ecry3.1Ab (chimaera of mCry3A and cry1Ab), the cry3A detection method may be suitable, although this GMO also contains the CaMV 35S promoter and the nos terminator that will already be used in most screening approaches. Also, in experimental lines of tobacco [35], poplar [16], alfalfa [41], Norway spruce [8] and rice [27], cry3A was applied, conferring resistance against Coleoptera. The gat/T-pinII method can be used to screen for canola 61061 (DP-Ø61Ø61-7) and 73496 (DP-Ø73496-4), maize 98140 (DP-Ø9814Ø-6) for which events no GMO element screening methods are available yet, and soybean 356043 (DP-356Ø43-5) that can be screened for only using the promoter element P-35S. In the present paper, we describe the newly developed methods and evaluate these in the light of established ENGL method performance requirements [25].

Materials and methods

Samples

MIR604 and Event 5307 were used as positive reference materials for the cry3A detection method. A range of GMOs (including those containing variants of the cry gene) were used as negative control. For the construct gat/T-pinII, maize 98140 and soybean 356043 were used as positive controls. A range of GMOs (including those containing either gat or pinII) were used as negative control. In Table 1, all reference materials are described.

Table 1 GMO reference materials obtained from the ERM (European Reference Materials: IRMM, Geel, Belgium) or AOCS (American Oil Chemist’s Society: Urbana, IL, USA)

DNA extraction

DNA from reference material was isolated from 100 mg material using a CTAB DNA isolation method [33], the Maxwell® 16 Tissue DNA Purification Kit (Promega) used with the Maxwell® 16 machine (Promega), or with CTAB/Qiagen (CTAB lysis step, followed by the DNeasy Plant Mini Kit (Qiagen) isolation. Optical density was measured on a Nanodrop spectrophotometer (ND-1000, Montchanin, DE, USA), which was used to quantify and assess the purity of the DNA. DNAs (extracted from reference materials, or provided as DNA) were diluted with water (Gibco distilled water, DNase/RNase free, Life Technologies, Grand Island NY, USA) to 10 ng/μl final concentration and stored at 4 °C, or at −20 °C for long-term storage.

Primers and probes

Primers and probe for cry3A were designed using Beacon Designer 7.0 (PREMIER Biosoft, Palo Alto, CA, USA). Settings for the amplicons were Ta = 60 °C ± 2 °C; amplicon length 70–120 bp (base pairs).

Primers and probe for gat/T-pinII were designed using Beacon Designer 8.14 (PREMIER Biosoft, Palo Alto, CA, USA). Settings for the amplicons were Ta = 60 °C ± 5 °C; amplicon length 60–120 bp. The detection method was partially based on the SYBR Green qPCR CoSYPS method described by Broeders et al. [10]. Primers and probes were synthesised by Biolegio (Nijmegen, Netherlands) with a high-purity method. Probes were purified by HPLC.

Sequence analysis was performed at the GenBank database by using the BLASTn tool (Basic Local Alignment Search Tools Nucleotide) of the National Center for Biotechnology Information (NCBI). The alignments among similar sequences were performed by using the software Clustal Omega [38] version 1.2.1.

Mastermix and PCR programme

The following mastermix was used in 25-µl final volume: 1 × Diagenode mastermix with UNG (Diagenode, Belgium) to prevent amplification of previously amplified products, 400 nM forward and reverse primer each, 200 nM probe, 5 µl water, 5 µl DNA (10 ng/µl, or dilutions thereof). PCR programme in Bio-Rad CFX96: initial decontamination with UNG (uracil-N-glycosylase) 120 s at 50 °C, denaturation and activation 600 s at 95 °C, then 45 cycles with 15 s at 95 °C, 60 s at 60 °C (hold at 20 °C indefinitely). qPCR trace files were analysed with the Bio-Rad CFX manager software version 3.1.

Limit of detection and PCR efficiency

Limit of detection and PCR efficiency were determined according to the minimum performance requirements for analytical methods of GMO testing [25] and the guidelines published by Broeders et al. [11]. The detection limit should at least be 20 haploid genome copies. The TaqMan PCR analysis was evaluated for each primer/probe pair by the serial dilution of CRM for and in water. MIR604 maize was used for cry3A. DP073496 canola and DP356043 soybean were used for gat/T-pinII. The copy numbers in the calibration curve samples are obtained by dividing the amount of sample DNA (picograms) by the published average 1C value for the maize (2.725 pg), canola (1.15 pg) and soybean (1.13 pg) nuclear genome [3].

Results and discussion

TaqMan PCR development

The cry3A gene (BCH abbreviation CS-cry3A-BACTU) is derived from Bacillus thuringiensis subsp. tenebrionis, strain BI 256-82. It codes for a Bt-toxin, conferring resistance to Western corn rootworm (Diabrotica virgifera virgifera), Northern corn rootworm (D. longicornis barberi) and other related Coleopteran species [5]. The gene is often codon-optimised [40] for the target species (mcry3A), or fused to cry1Ab (eCry3.1Ab). At least 30 known potato events contain cry3A or variants thereof. For cry3A in potato, a conventional PCR detection method was already developed [39], but to reduce chances of contamination, qPCR in closed vials is preferred. A loop-mediated isothermal amplification (LAMP) assay for detection of cry3A is also available [28], but for applicability in a TaqMan PCR screening strategy, this is not suitable. The gene is also applied in experimental lines of GM tobacco, poplar, alfalfa, Norway spruce and rice, and maize Event 5307, MIR604 and MZIR098. The latter one is currently pending for the determination of non-regulated status in the USA [2].

The construct gat/T-pinII is a synthetic fusion sequence between the coding sequence of the glyphosate-N-acetyltransferase (gat) gene and the terminator from the proteinase inhibitor II (pinII) gene. The gat gene (CS-gat-BACLI) is derived from Bacillus licheniformis. It codes for glyphosate-N-acetyltransferase, which is able to detoxify glyphosate [6]. There are at least two modified genes present in known GMOs: GAT4621 and GAT4601. The pinII terminator (T-pinII-SOLTU) is derived from Solanum tuberosum [5]. The construct gat/T-pinII is at least present in canola 61061 and 73496, maize 98140 and soybean 356043.

Table 2 describes the primers and probes that were developed for the two TaqMan PCR detection methods, based on the GenBank entries in the reference.

Table 2 Primers and probe designed for the detection of element CS-cry3A and construct gat/T-pinII

Specificity

The specificity of the gat/T-pinII detection method was checked by similarity searches in silico by BLAST analysis against GenBank. BLASTn against the patent database with the gat/T-pinII amplicon from maize 98140 returned 17 different patents in the first 100 hits, of which 4 belonged to either canola 61061 and 73496, or maize 98140. The other hits were GMOs containing the gat/T-pinII construct (Table 3). Patent US 8901377 is a patent on GM sunflower. The last eight patents showed a single difference in the amplicon, between the forward primer and the probe. This difference will not hinder a correct amplification of the amplicon.

Table 3 CLUSTAL Omega (1.2.1) multiple sequence alignment with 9 100 % identical amplicon entries A (represented by 98140: KP784699, containing gat/T-pinII) and 8 100 % identical amplicon entries B (represented by patent WO2010077890, containing cry3A with one difference in the amplicon) shows the difference (at nt 24 in bold) between entries A and B

BLASTn against the nucleotide collection (nr/nt) database with the gat/T-pinII amplicon from maize 98140 returned two entries (both maize 98140).

The specificity of the cry3A detection method was checked in silico by BLAST analysis against GenBank. BLASTn against the patent database with the cry3A amplicon from maize MIR604 returned eleven different patents in the first 100 hits, of which three belonged to either MIR604 or Event 5307. The other hits were GMOs containing the cry3A gene.

BLASTn against the nucleotide collection (nr/nt) database with the cry3A amplicon from maize MIR604 returned three entries. A CLUSTAL Omega multiple sequence alignment returned the alignment shown in Table 4. Amplicons for MIR604 and Event 5307 were included from the patent search.

Table 4 CLUSTAL Omega (1.2.1) multiple sequence alignment of the cry3A amplicon

The synthetic cry3A genes GenBank: JN989558.1 [32] and GenBank: M84650.1, and the two potato cry3A genes are not likely to be amplified with the designed primer combination. eCry3.1Ab (GenBank: GU327680.1) is the cry3A gene from Event 5307.

Summarising, the BLASTn analyses against nr/nt and patent databases show that the element-specific method for cry3A is specific for cry3A, and only returns hits with MIR604, Event 5307 and some other entries that contain cry3A. BLASTn analyses against nr/nt and patent databases show that the construct-specific method for gat/T-pinII is specific for the gat/T-pinII construct, and only returns hits with canola 61061 (DP-Ø61Ø61-7) and 73496 (DP-Ø73496-4), and maize 98140 (DP-Ø9814Ø-6). No hit with soybean 356043 (DP-356Ø43-5) could be found that confirmed the entry as soybean 356043. Some other entries that contain gat/T-pinII were found, including a GM sunflower. One group of entries (all from Du Pont de Nemours and Company) from GenBank had one difference in the amplicon, without affecting its amplification.

Experimental results from testing the methods with non-target transgenic events, non-transgenic material and target material showed that all reference materials described in Table 1 were negative for the cry3A detection method, except MIR 604 and Event 5307. For the construct-specific method gat/T-pinII, all reference materials described in Table 1 were negative for gat/T-pinII, except canola 73496 (50 ng 100 %: Cq = 23), maize 98140 (50 ng 10 %: Cq = 28) and soybean 356043 (50 ng 10 %: Cq = 30). Since the pinII terminator originates from S. tuberosum, both S. tuberosum and S. lycopersicum were tested as well. GMOs containing T-pinII (maize DAS59122 and 4114) were also included. All were found negative for the gat/T-pinII construct-specific detection method. All results were in agreement with the expected outcome and confirmed the specificity of the two methods with the reference materials tested.

Efficiency and correlation coefficient

The efficiency and correlation coefficient R 2 were calculated with CFX software using five serial dilutions in duplicate (two reactions per concentration). In Table 5, the average values are given. Efficiency and R 2 are in accordance with ENGL minimum performance requirements that specify an efficiency range between 90 and 110 % and an R 2 higher than or equal to 0.98, in order for a PCR method to be used for quantification. The method performed in a linear quantitative manner in the full range DNA concentrations tested.

Table 5 Average values for efficiency (%) and R2 of the cry3A and gat/T-pinII detection methods

Limit of detection

For gat/T-pinII, a first estimate of the LOD10 (10 repeats of different copy levels where the LOD is at the amount of copies that were detected 10/10) was performed, ranging from 40 (10/10), 20 (10/10), 10 (10/10), 5 (7/10) and 1 copies (6/10) per reaction in 10 repetitions (with 73496 canola and 356043 soybean). The estimated LOD10 was below 10 copies. The 10 copies dilution was repeated 60 times to obtain an estimate of the LOD95 %. At 10 copies, 93 % (57/60) of the reactions were positive. Next, 20 copies were repeated 60 times, resulting in 60/60 positives (100 %) for 73496 canola and 356043 soybean. The estimated LOD95 % is below 20 copies.

For cry3A, the initial LOD estimation was performed with 10 reactions per concentration, ranging from 25 (10/10), 10 (10/10), 5 (9/10), 2.5 (7/10) and 0.5 copies (2/10) (MIR604) per reaction. LOD10 was below 10 copies. The 10 copies level was repeated 60 times to verify the LOD95 % at 10 copies, resulting in 59/60 positives (98.3 %). The estimated LOD95 % is below 10 copies.

Conclusion

Two new methods were developed to improve the screening approach for known and unknown GMOs containing either the element cry3A or the construct gat/T-pinII. The cry3A detection method was specific for the cry3A element. No false positives were found with the cry3A detection method, including those GMOs containing cry-related genes, and no false negatives were found since all the GMOs containing cry3A were positive. The gat/T-pinII detection method was specific for the gat/T-pinII construct. No false positives and no false negatives were found with the gat/T-pinII detection method, because only the GMOs with gat4621/T-pinII and gat4601/T-pinII constructs were positive. Both methods were found satisfactory with regard to their specificity, PCR efficiency, linearity and sensitivity in the light of criteria for such methods as have been established by the European Network of GMO Laboratories. The applicability and practicability is expected to be similar as for other TaqMan screening methods. After full validation, the methods can be used in GMO element screening approaches to facilitate the detection and identification of maize 5307, MIR604 and 98140, soybean 356043, canola 73496 and potentially present currently unknown GMOs containing cry3A or gat/T-pinII.