Monitoring the prevalence of genetically modified (GM) maize in
Accepted Article
commercial animal feeds and food products in Turkey
Aydin Turkeca,1,*, Stuart J. Lucasb,1,* and Elif Karlıkb,2
a
Uludag University, Mustafa Kemalpasa Vocational School, Department of Plant and Animal Production, 16500 Bursa, Turkey
b
Sabanci University, Nanotechnology Research and Application Centre, Orhanlı, 34956 Tuzla, Istanbul, Turkey
*Corresponding authors. Phone: +90 224 613 3102 Fax: +90 224 613 9666 E-mail:
[email protected] (A. Turkec)
Phone: +90 216 483 9522 Fax: +90 216 483 9885 E-mail:
[email protected] (S.J.Lucas)
1
These two authors contributed equally to this work
2
This work carried out on secondment from: Fatih University, The Institute of Science
and Engineering, 34500 Büyükçekmece, Istanbul
This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1002/jsfa.7496 This article is protected by copyright. All rights reserved.
ABSTRACT BACKGROUND: EU legislation strictly controls use of genetically modified (GM) crops
Accepted Article
in food and feed products, and requires them to be labelled if the total GM content is greater than 9 g.kg-1 (for approved GM crops). We screened maize-containing food and feed products from Turkey to assess the prevalence of GM material.
RESULTS: With this aim, 83 food and feed products – none labelled as containing GM material – were screened using multiplex real-time polymerase chain reaction (PCR) for four common GM elements (35S/NOS/bar/FMV). Of these, 18.2% of feeds and 6% of food samples tested positive for one or more of these elements, and were subjected to event-specific PCR to identify which GMOs they contained. Most samples were negative for the approved GM events tested, suggesting that they may contain adventitious GM contaminants. One sample was shown to contain an unapproved GM event (MON810, along with GA21) at a concentration well above the statutory labelling requirement.
CONCLUSION: Current legislation has restricted the penetration of GM maize into the Turkish food industry but not eliminated it, and the proliferation of different GM events is making monitoring increasingly complex. Our results indicate that labelling requirements are not being followed in some cases.
Keywords Zea mays L., Maize, Genetically modified organism (GMO), Real-time PCR, Food safety, GMO quantification
This article is protected by copyright. All rights reserved.
Introduction
The cultivation of biotechnological crops has been increasing steadily worldwide,
Accepted Article
reaching a global area of 181.5 million hectares in 28 countries by 2014, from which soybean and maize accounted for 50% and 30% of the total area respectively1. Additionally, an increasing number of a new genetically modified (GM) events are in the approval process or already being marketed2. Maize has the greatest diversity of approved GM events globally (a total of 130 varieties approved in one or more countries), and is one of the most widely used staple food and feed ingredients, highly integrated into food and feed supply chains1–3. Consumers and policy-makers have significant concerns about the possible impact of GM crops on health, the environment and socio-economic systems4. In consequence, the spread of GM crops worldwide has resulted in increasing the regulatory requirements for safety assessment of food and feed containing GMOs, before the crops from which they are produced become established in cultivation and trade5. Nevertheless, there is no international consensus on GM crop cultivation and use, but different countries have established their own biosafety laws and regulations6,7. In the European Union (EU), the legislative framework for GMO cultivation and trade requires mandatory labeling and traceability of GMOs in food and feed products when they exceed a threshold of 9 g.kg-1 GM content at all stages of the supply chain8,9. Furthermore, the EU has recently adopted a zero-tolerance policy for the low level presence (LLP) of unauthorized GMOs in imports for animal feed10. In September 2010, Turkey passed GMO legislation similar to that of the EU, with the same labelling threshold (9 g.kg-1) for authorized GM events in foods and feeds, and 1 g.kg-1 for unapproved GM material in feeds only. The production of biotech crops is banned in Turkey, but GM events that have been approved by the Ministry of Food, Agriculture
This article is protected by copyright. All rights reserved.
and Livestock may be imported use in animal feeds11. To date, no GM events have been approved for food use in Turkey. With the global increase in the variety of commercial GM events, the penetration of
Accepted Article
GM maize DNA into food and feed supply chains in the EU and Turkey is highly probable2. Turkey imports large quantities of feed crops for the poultry and livestock sectors each year11. Monitoring the GMO content of food and feed is becoming increasingly challenging. Accurate analytical tests are essential for enforcement of GM regulations at each step in the food and feed supply chain. PCR-based methods are commonly used for GMO detection and identification due to their high sensitivity and relative speed and simplicity. In particular, approaches based on quantitative real-time PCR (qPCR) are the primary means of quantification of GMO content12,13. In recent years, numerous studies have used PCR methods to detect the presence of GMO DNA in food and feed supply chains in different parts of the world7,14–26. On the other hand, relatively few studies have used qPCR to enable quantitative detection of GM maize events in food supply chains27–31. Therefore, further monitoring of the GMO content of maize-containing feeds and foodstuffs is still required to obtain a more complete picture of the prevalence of GM maize in Europe, and assess compliance with biosafety regulations. In the case of Turkey, to our knowledge there is no quantitative data available on the GMO content of maize-containing foodstuffs in Turkey. The aims of present study were to measure the GMO content of maize-containing foodstuffs and feeds in Turkey, and to evaluate compliance with labelling requirements.
Experimental
Research materials
This article is protected by copyright. All rights reserved.
A total of 83 maize-containing samples included 33 maize-based animal feeds (in the form of flakes, pellets, small particles or whole grains), which were obtained from feed manufacturing companies located in several different regions of Turkey (Aegean,
Accepted Article
Anatolian, Central Anatolian, Marmara, and Mediterranean). The 50 food samples, including maize flour, maize starch, maize-based breakfast cereals, canned maize, corn chips, and biscuits and snacks containing maize, were randomly purchased from Turkish retailers in Istanbul and Bursa in 2015. Certified reference materials (CRMs) consisting of ground dried maize kernels for events bt11, DAS59122, GA21, MON810, NK603 & TC1507 were obtained from Sigma-Aldrich (St. Louis, MO, USA), at a range of GMO concentrations from 1-100 g.kg-1, along with GMO-free controls. Pure ground kernels of MON88017 and MON89034 were obtained from the American Oil Chemists’ Society (AOCS; Urbana, Il, USA) and used as positive controls for these events. Except where stated below, other chemicals were obtained at molecular biology grade from Sigma-Aldrich.
Genomic DNA extraction All samples apart from flour and starch were homogenized by grinding in liquid nitrogen and stored at -80oC prior to use. Total DNA was extracted from 200 mg of the majority of the samples using the Foodproof GMO Sample Preparation Kit (Biotecon Diagnostics GmbH, Potsdam, Germany) according to the manufacturers’ instructions, except that the initial extraction step was routinely extended from 30 to 60 minutes, and at the final step 40 µl Elution Buffer was used in 2x20 µl elutions, to recover DNA at the highest possible concentration. For the tinned sweet corn samples, the cetyltrimethylammonium bromide (CTAB) precipitation method was used as previously described5. In the case of the maize starch and breakfast cereal samples, the procedure was modified as follows: 200 mg of ground sample were incubated with 1 ml Edward’s buffer (17.3 mM SDS, 250 mM NaCl, 25 mM EDTA, 200 mM Tris pH 8.0) for 5 minutes at 80oC32. They were then spun down at top speed in a microcentrifuge for
This article is protected by copyright. All rights reserved.
10 minutes, and the supernatant extracted twice with chloroform to remove protein. The aqueous (upper) phase was then incubated with 2 volumes of CTAB precipitation solution, after which the CTAB protocol was followed as previously described5. As
Accepted Article
necessary, samples were extracted in duplicate or more repetitions to improve DNA yield. Spectrophotometric analysis of the DNA yield and purity was carried out at 230, 260 & 280 nm using a NanoDrop 2000c UV/Vis spectrophotometer (Thermo Scientific, Wilmington, DE, USA), and degradation of the isolated DNA was assessed by agarose gel electrophoresis, in which 25-100 ng of purified DNA samples were separated on 10g.l-1 agarose gels containing GelRed nucleic acid stain (Biotium, Hayward, CA, USA) in 0.5x TBE buffer.
PCR primers and probes
Primer pair IVR1-F/IVR1-R was used to amplify the invertase gene as a taxonspecific indicator for maize genomic DNA33. The specific detection of GM maize events bt11, DAS59122, GA21, MON810, MON88017, and NK603 was carried out using the primers described in Table 1, with PCR conditions for each primer pair as described in previous studies15,29,34–37. Fluorescently labelled hydrolysis probes were used for event-specific quantitation of GM maize events by qPCR. Primers and probes, which were dual-labelled with FAM and TAMRA and purified by HPLC, were produced by Macrogen (Seoul, Korea). Primer and probe sequences are summarized in Table 2, along with final concentrations used in qPCR reactions, as determined by previous studies38–42.
GMO screening and detection of specific GM maize events
This article is protected by copyright. All rights reserved.
For GMO screening, the Foodproof GMO Screening Kit (Biotecon Diagnostics GmbH, Potsdam, Germany) was used according to the manufacturer’s instructions, using 80 ng of template DNA in each 10 µl volume. All real-time PCR reactions were
Accepted Article
performed using a LightCycler 480 II Instrument (Roche Diagnostics GmbH, Mannheim, Germany) as described previously43 with multi-color detection in 4 channels (FAM; VIC/HEX; Rox; Cy5). A color compensation object was first created using Color Compensation Set 3 (Biotecon Gmbh, Potsdam, Germany) and all samples were then screened at least twice on different days, alongside appropriate CRMs as controls. For the specific detection of GM maize events, and the taxon-specific invertase gene, qualitative PCR reactions were performed a final volume of 20 µl in a Mastercycler 384 Gradient thermocycler (Eppendorf AG, Hamburg, Germany) alongside CRMs and no template controls.
PCR products were analyzed by
electrophoresis on 12-20 g.l-1 agarose gels with the GeneRuler 100 bp DNA ladder (Thermo Scientific, Waltham, MA, USA) and visualized as described above (see ‘Genomic DNA extraction’).
Quantification of GM maize events
The qPCR amplifications were carried out using the LightCycler 480 II Instrument and the primers listed in Table 2. For each sample, 100 ng of template was used in a final reaction volume of 10 µl, containing 1x LightCycler 480 Probes Master reaction mix (Roche Diagnostics Gmbh, Mannheim, Germany). A common amplification program was used for all of the elements tested as follows: initial denaturation at 95oC for 10 minutes, followed by 45 cycles of 10s at 95oC, 30s at 60oC, and 10s at 72oC, with fluorescence detection during the third step of each cycle. For real-time quantifications, the double standard curve method was used, as this is typically more robust than Ct comparison44. The maize invertase standard curve was set up using the GMO-free CRM for GA21, with 100 ng template in the most concentrated standard,
This article is protected by copyright. All rights reserved.
followed by 5 serial fourfold dilutions. Standard curves for GM events were set up similarly, except that the serial dilutions were made from 100 ng of the most concentrated CRM available (100 g.kg-1, except for GA21, which was 43 g.kg-1).
Accepted Article
Standard curve dilutions were prepared and assayed in duplicate on at least two separate occasions. Test samples were assayed in triplicate for maize invertase and each GMO event; every qPCR plate also included the CRM of each event at 10 g.kg-1 GMO concentration as a positive control/calibrator. The crossing point (Cp/Ct value) of each amplification curve was calculated using the Second Derivative Maximum method, and copy number of each element estimated by comparison with the relevant standard curve.
The
estimated copy number of each GMO event was divided by that of maize invertase to derive the GMO concentration.
Results and Discussion
Sample collection and DNA extraction
In this study, 83 samples of maize-containing foods and feeds were acquired in 2015 in Turkey. Genomic DNA was extracted from each of these samples.
As
mentioned in our previous report45, selection of the appropriate isolation method for each food matrix is necessary in order to obtain good quality, amplifiable DNA. Accordingly, in this study we found that while the Foodproof GMO Sample Preparation Kit gave good yields and adequate purity (A260/A280 ≥ 1.8) for the majority of sample types, it did not perform as well for tinned corn, breakfast cereals or corn starch. When these samples were isolated using the CTAB method5 yield and/or purity were improved in most cases. We also tested several modifications of the CTAB protocol, particularly for maize starch, which is difficult to recover genomic DNA from as most of
This article is protected by copyright. All rights reserved.
the cell nuclei are removed during its production.
We found that using Edwards
buffer32 instead of CTAB during the initial lysis & protein precipitation steps of the protocol improved yields both for maize starch and breakfast cereal, suggesting that
Accepted Article
the negatively charged SDS detergent in Edwards buffer is more effective than the positively charged CTAB at separating genomic DNA from the contaminating biomolecules present in these samples. The integrity of DNA from all of our samples was checked by agarose gel electrophoresis.
While the least processed samples (flour, feeds) gave a high
molecular weight band of intact chromosomal DNA, there was some smearing in all cases, which shifted to lower molecular weight in the more highly processed samples. DNA degradation can reduce the effectiveness of PCR analysis if the majority of fragments are shorter than the region to be amplified46.
However, all of the DNA
samples used here were confirmed to contain amplifiable DNA by PCR for the invertase gene (data not shown).
Incidence of GM elements in foodstuffs and feed samples
All the samples were screened for four genetic elements of non-plant origin that are widely used in GM constructs (Cauliflower Mosaic Virus 35S promoter (CaMV 35S), Agrobacterium tumefaciens Nopaline synthase terminator (tNOS), Figwort Mosaic Virus 35S promoter (FMV 35S) and the bar gene from Streptomyces hygroscopicus) by a multiplex real-time PCR method, the Foodproof GMO Screening Kit. Among the 7 maize GMO events approved for feed use in Turkey, 6 (bt11, DAS59122, MON88017, MON89034, NK603 & TC1507) contain the CaMV 35S element, while 5 (bt11, GA21, MON88017, MON89034 & NK603) include tNOS47. The other 2 elements are not present in any of these approved events, but bar is found in non-approved maize varieties such as bt176, while FMV 35S is present in an approved soya GMO (MON89788) and so is a useful test for cross-contamination in industrial processes.
This article is protected by copyright. All rights reserved.
All samples tested positive in the plant gene-specific assay included in the kit. The results of the GMO screening assay indicated that nine out of 83 maize samples (10.8%) were positive for at least one GM element, with 7 samples (8.4%) positive for
Accepted Article
CaMV 35S and 8 samples (9.6%) for tNOS; 6 of the samples contained both of these elements. None of the samples were positive for the bar gene or FMV 35S promoter (Table 3). The majority of GM positive samples were among the maize feeds (6/33; 18.2%), with relatively fewer in the foodstuffs tested (3/50; 6%), these comprising a single sample each of maize starch, corn chips and biscuits. Although this assay is not quantitative, the Cp values (PCR cycle at which fluorescence is increasing most rapidly; analogous to Ct value) obtained for different samples provide a qualitative indication of the relative abundance of the elements detected, with lower cycle numbers suggesting higher template concentration.
Notably, the high Cp values
observed for GM elements in most of the samples (Table 3) suggest that these elements may only be present in very low concentrations. These results are comparable to those obtained in our previous study45 wherein 4/49 maize samples (8.2%) tested positive for 1 or more GM elements, with both animal feed and maize starch among the GM positive samples.
These levels of
incidence are lower than 2 earlier studies conducted using samples collected before 2010, which detected GM event Bt11 in 32.6%48 and 35%21 of maize-containing food and feed products tested. In contrast, a more recent Turkish study26 did not detect GM maize in any of 11 feeds tested, although occurrences could have been missed due to the small sample size. Taking these studies together with our results, it seems that the Biosafety Law that came into force in September 2010 has reduced the prevalence of GM maize in the Turkish food and feed industries, although it can still be found in the marketplace. Considering a broader picture, Greiner and Konietzny27 also detected the presence of GM maize in 8-11% of samples collected in Brazil, varying depending on the year. Comparable levels of GM incidence were also recently reported in maize-containing
This article is protected by copyright. All rights reserved.
foodstuffs from Iran (12%)25, Jordan (5.4%)28 and the United Arab Emirates (13%)7. However, studies in some countries detected GM maize at much greater frequency; for example, 30% of 119 maize products tested in Portugal were positive for CaMV 35S,
Accepted Article
and 10% for tNOS29, while 44% of those from Saudi Arabia14 were positive for GM maize. Conversely, a recent study in Serbia detected no GM material in 48 maizebased food samples20. Clearly, the penetration of GM maize into national markets varies considerably depending on the regulations and supply networks in each country.
Detection and quantitation of specific GM maize events
There are currently 14 GM maize varieties approved for import into Turkey for feed use only, including Bt11, GA21, NK603, DAS59122, MON88017 and TC1507, along with several varieties generated by stacking of these events through cross-breeding. The event MON810 was also approved in 2010 but has subsequently had its authorization rescinded11. The results from the screening of GMO elements (above) provide some indication of which GMOs may be present in each sample; DAS59122, MON810 and TC1507 contain the CaMV 35S promoter but not tNOS, GA21 contains tNOS only, while Bt11, MON88017 & NK603 include both of these elements. With the aim of determining which specific GMO(s) were present in our samples, PCR amplifications using primers specific for each of these events (Table 1) were carried out for all of the GMO-positive samples, along with GMO positive and negative controls. Only one of the samples (Feed #09) was positive in any of the qualitative PCR tests; in this sample, both MON810 and GA21 were present, consistent with the previous screen in which CaMV 35S and tNOS were both detected. The remaining 8 samples gave no amplification in any of the event-specific qualitative PCRs, even though they had tested positive in the initial screen. Conceptually, this could be explained by the CaMV 35S and tNOS elements in these samples originating from GM
This article is protected by copyright. All rights reserved.
events other than those tested. Alternatively, they could be false negatives resulting from lower sensitivity of the qualitative PCR amplification compared to qPCR using fluorescent probes. To address this possibility, all the samples that had tested positive
Accepted Article
in the initial screen were also tested in event-specific qPCR reactions (below). Compliance with labelling regulations also requires the concentration of GMO to be known; this was determined by qPCR using standard curves, as this is usually more robust than the relative Ct method44. Standard curves for the maize invertase gene (taxon-specific) and GM maize events DAS59122, GA21, MON810 & MON88017 were set up using dilution series from CRMs of each of these events, and the primers and probes described in Table 2. For all elements, the PCR efficiency indicated by the standard curves were within recommended limits (90-105%) as was the observed error in the data points (
