Tracing Transgenic Maize As Affected by Breadmaking Process and Raw Material for the Production of a Traditional Maize Bread, Broa ⇑ Telmo J.R

Food Chemistry 138 (2013) 687–692

Contents lists available at SciVerse ScienceDirect

Food Chemistry

journal homepage: www.elsevier.com/locate/foodchem

Tracing transgenic maize as affected by breadmaking process and raw material for the production of a traditional maize bread, broa ⇑ Telmo J.R. Fernandes, M. Beatriz P.P. Oliveira, Isabel Mafra

REQUIMTE, Departamento de Ciências Químicas, Faculdade de Farmácia, Universidade do Porto, Rua de Jorge Viterbo Ferreira, 228, Porto 4050-313, Portugal article info abstract

Article history: Broa is a maize bread highly consumed and appreciated, especially in the north and central zones of Por- Received 15 June 2012 tugal. In the manufacturing of broa, maize flour and maize semolina might be used, besides other cereals Received in revised form 28 September 2012 such as wheat and rye. Considering the needs for genetically modified organism (GMO) traceability in Accepted 24 October 2012 highly processed foods, the aim of this work was to assess DNA degradation, DNA amplification and Available online 12 November 2012 GMO quantification along breadmaking process of broa. DNA degradation was noticed by its decrease of integrity after dough baking and in all parts of bread sampling. The PCR amplification results of Keywords: extracted DNA from the three distinct maize breads (broa 1, 2 and 3) showed that sequences for maize Maize bread invertase gene and for events MON810 and TC1507 were easily detected with strong products. Real-time GMO detection DNA degradation PCR revealed that quantification of GMO was feasible in the three different breads and that sampling Processing location of baked bread might have a limited influence since the average quantitative results of both Traceability events after baking were very close to the actual values in the case of broa 1 (prepared with maize sem- Quantitative real-time PCR olina). In the other two maize breads subjected to the same baking treatment, the contents of MON810 maize were considerably underestimated, leading to the conclusion that heat-processing was not the responsible parameter for that distortion, but the size of particle and mechanical processing of raw maize play also a major role in GMO quantification. Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction (EU), maize has the highest number of authorised GM events (21) for food and feed (GMO Compass, GMO Database, Genetically Mod- Maize (Zea mays L.) plays a major role in nutrition in many ified Food, 2012). Examples of authorised transgenic maize events countries. It is a cereal widely used for human nutrition as a source are the MON810 maize, which contains the cry1Ab gene inserted to of oil, flour and starch. Maize is the basis for the production of sev- confer insect resistance, and the TC1507 maize, which contains the eral foods, such as polenta, bread, tortillas, snacks, cornflakes, etc. cry1Fa2 and pat genes inserted to confer insect resistance and tol- In the production of bread, it is also used as wheat flour erance to the herbicide glufosinate ammonium, respectively. replacement. MON810 was introduced as an authorised maize event in the EU Broa is a Portuguese traditional maize bread highly consumed, in 1998, being actually under a renewal process for authorisation especially in the north and central zones of Portugal and still plays as food and feed and cultivation. The maize event TC1507 is actu- an important economic and social role in rural communities of the ally authorised under the scope of application for food and feed country (Brites, Trigo, Santos, Collar, & Rosell, 2010; Lino, Silva, since 2005, but not yet for cultivation (GMO Compass, 2012). Both Pena, Férnandez, & Mañes 2007; Vaz Patto, Moreira, Carvalho, & events have been reported as the most frequently found in foods Pego, 2007). The traditional manufacturing of broa consists of add- available in the Portuguese retail market (Fernandes, 2011; Fer- ing sieved maize flour (mixed with wheat or rye flours), hot water, nandes, Costa, Oliveira, & Mafra, 2012). yeast and leavened dough. After mixing, resting and proofing, the In EU, under Regulations (EC) No. 1829/2003 and 1830/2003 dough is baked in an oven (temperature over 200 °C) resulting in regarding the authorisation and traceability of food and feed de- a highly accepted foodstuff, mostly due to its distinctive sensorial rived from genetically modified organisms (GMO), labelling for characteristics (Brites et al., 2010). food products containing more than 0.9% authorised GM material Maize is the second most cultivated genetically modified (GM) is mandatory. To monitor and verify the compliance of labelling, crop, corresponding to 32% of the global biotechnological area of appropriate testing methods to detect and identify GM events in 160 million hectares in 2011 (James, 2011). In the European Union processed foodstuffs are needed. Genomic DNA extraction followed by amplification of target sequences (endogenous and ⇑ Corresponding author. Tel.: +351 220428640; fax: +351 226093390. event/construct specific) by qualitative and quantitative polymer- E-mail address: [email protected] (I. Mafra). ase chain reaction (PCR) are the most widely accepted methods

0308-8146/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.foodchem.2012.10.068 688 T.J.R. Fernandes et al. / Food Chemistry 138 (2013) 687–692 for detecting and quantifying levels of GM traits in foodstuffs (Gry- Table 1 son, Dewettinck, & Messens, 2007; Shokere, Holden, & Jenkins, Composition of the three different maize breads prepared. 2009; van der Colff & Podivinsky, 2008). Real-time quantitative Ingredients (g) Broa 1 Broa 2 Broa 3 PCR represents the most powerful means of quantifying GM mate- GM maize semolina 400 – – rial in agricultural and food products, due to its high sensitivity, GM maize flour – 100 150 specificity and reproducibility (Gryson, 2010; Mafra, Ferreira, & Wheat flour (type 65) 300 150 150 Oliveira, 2008). However, applicability of DNA-based methods for Maize flour (type 70) – 100 50 Rye flour (type 130) 100 50 50 GMO detection and quantification depends on the quality and Bread improver 8 4 4 quantity of DNA. Salt 12 6 6 Food processing and consequent DNA degradation, can affect Sugar 8 4 4 DNA recovery and the efficiency of PCR and, therefore, the quality Dry yeast 24 12 12 of analytical results, leading to under- or overestimation of GMO Total + Water 1000 500 500 content (Bergerová, Hrncˇírová, Stankovská, Lopašovská, & Siekel, 2010; Gryson, 2010). The degree of sample processing and the DNA extraction method used may lead to DNA fragment damage, was manually moulded in balls and baked in the oven at 240 °C, which can be caused by exposure to heat or pressure, enzymatic during 30 min. degradation by nucleases, chemical agents, water activity, shear During the preparation of maize bread two samples of dough forces and pH values (Gryson, 2010; Gryson, Messens, & Dewet- were taken: before and after leavening. In the final baked breads, tinck, 2008). Food manufacturing comprises a large number of three samples were taken from different bread location: crust, un- complex processing steps, which contribute to a particular envi- der crust and middle soft part of the bread. Each sample was sep- ronment where DNA undergoes deterioration and proteins are arately ground and homogenised in a Grindomix GM200 (Retsch, denatured (Vijayakumar, Martin, Gowda, & Prakash, 2009). Cook- Haan, Germany). ing and baking are essentially heat processes that may break down DNA, whereas the thermal treatment with temperatures over 2.3. DNA extraction 200 °C considerably reduces the size of the extracted DNA fragments (Bergerová, Godálová, & Siekel, 2011; Gryson, 2010). DNA was extracted using the Wizard method as described by Considering the few number of reports assessing the influence Mafra, Silva, Moreira, Silva, and Oliveira (2008), with minor modi- of the thermal treatment on DNA degradation and PCR amplifica- fications. To each sample (100–150 mg) transferred to a 2 mL ster- tion for GMO detection and quantification in maize-derived food- ile reaction tube, 860 lL of extraction buffer (10 mM Tris, 150 mM stuffs, this study aimed at evaluating the effect of breadmaking NaCl, 2 mM EDTA, 1% SDS), 100 lL of 5 M guanidine hydrochloride of maize bread manufactured with known amounts of transgenic solution and 40 lL of proteinase K solution (20 mg/mL) were maize. Therefore, different traditional maize breads (broa) were added. The mixtures were vigorously vortexed and incubated in a prepared and samples were taken before and after the steps of thermal block (Thermomixer Compact, Eppendorf AG, Hamburg, sourdough leavening and baking to assess DNA degradation and Germany) at 60 °C for 3 h with continuous stirring. After incuba- amplification, with special focus on real-time quantitative PCR tion, the mixtures were centrifuged (15 min, 18,514g at 4 °C), for GMO quantification. and 500 lL of supernatant were collected and mixed with 1 mL of WizardÒ DNA purification resin (Promega, Madison, WI, USA). Ò 2. Material and methods Each mixture was then pushed through a Wizard column (Prome- ga, Madison, WI, USA) mounted with a 2 mL syringe. The DNA-re- 2.1. GM maize material sin mix was washed twice with 2 mL isopropanol solution (80% v/ v). The column was dried for 5 min at room temperature and Certified reference materials from the Institute for Reference placed in a new sterile reaction tube. After incubation for 1 min Materials and Measurements (IRMM, Geel, Belgium) were used with 100 lL of Tris–EDTA buffer (10 mM Tris, 1 mM EDTA) at as standards (Fluka, Buchs, Switzerland) containing 1% and 5% for 70 °C, the column was eluted by centrifugation (1 min, 10,000g). the evaluation of MON810 maize event and 1% and 10% for All the extracts were kept at 20 °C until further analysis. TC1507 event. To prepare incurred maize breads with GM maize, two different 2.4. DNA purity and quality types of maize were used: maize semolina, containing MON810 (20%) and TC1507 (11%) events, which was purchased from a local The yield and purity of extracts were assessed by UV spectro- market in Porto (Portugal); MON810 maize kernels from crops cul- photometry using a spectrophotometer UV1800 Shimadzu (Kyoto, tivated in Alentejo region (Portugal). Maize kernels were ground Japan). The DNA concentration was determined by UV absorbance using a MFC-90D micro-hammer mill at particle size of 0.1 mm, at 260 nm (1 absorbance unit corresponds to 50 lg/mL of dsDNA). (Culatti, Zürich, Switzerland). The GM content of both samples The purity of the extracted DNA was determined by the ratio of the was previously determined. absorbances at 260 and 280 nm. The quality of extracted DNA was further analysed by electrophoresis in a 1.0% agarose gel contain- 2.2. Breadmaking process and sampling ing Gel Red 1x (Biotium, Hayward, CA, USA) for staining and carried out in SGTB 1x (GRiSP, Research Solutions, Porto, Portugal) Three different maize breads were prepared in a bakery in the for 20 min at 200 V. The agarose gel was visualised under UV light region of Porto according to the traditional process, using the and a digital image was obtained using a Kodak Digital Science™ amounts in Table 1. Hot water was added to the maize semolina/ DC120 (Rochester, NY, USA). flours and blended until a paste was obtained. The blend was then cooled at 4 °C. Wheat and rye flours were added along with bread 2.5. Oligonucleotide primers improver and blended with cold water. Afterwards, salt, sugar and yeast were mixed and the dough was worked up and leavened at The oligonucleotide primers and probes used in this work are room temperature during 20 min. After fermentation, the dough presented on Table 2 (qualitative PCR) and Table 3 (quantitative T.J.R. Fernandes et al. / Food Chemistry 138 (2013) 687–692 689

Table 2 Oligonucleotide primers for qualitative PCR.

Primer Sequence (50–30) Target Amplicon (bp) References IVR1-F CCGCTGTATCACAAGGGCTGGTACC Maize invertase gene 226 ISO 21569 (2005) IVR1-R GGAGCCCGTGTAGAGCATGACGATC VW01 TCGAAGGACGAAGGACTCTAACG Maize genome/35S promoter (MON810) 178 ISO 21569 (2005) VW03 TCCATCTTTGGGACCACTGTCG TC-F GACGTCTCAATGTAATGGTTAACGA Maize genome/ubiZM1 promoter 190 La Paz et al. (2006) TC-R GGGTAACCGCTCTTCCAGTTGT

Table 3 Oligonucleotide primers and probes for real-time PCR.

Primer Sequence (50–30) Concentration Target Amplicon References (lM) (bp) SSIIb 1–50 CTCCCAATCCTTTGACATCTGC 375 Maize starch synthase IIb 151 ISO 21570 (2005) SSIIb 1–30 TCGATTTCTCTCTTGGTGACAGG 375 SSIIb-Taq FAM-AGCAAAGTCAGAGCGCTGCAATGCA-BHQ2 150 MON810 2–50 GATGCCTTCTCCCTAGTGTTGA 500 hps70/Cry1Ab 113 ISO 21570 (2005) MON810 2–30 GGATGCACTCGTTGATGTTTG 500 MON810-Taq FAM-AGATACCAAGCGGCCATGGACAACAA-BHQ2 200 TC-F GACGTCTCAATGTAATGGTTAACGA 300 Maize genome/ubiZM1 promoter 190 La Paz et al. (2006) TC-R GGGTAACCGCTCTTCCAGTTGT 300 TC-P FAM-CCGCGTTAACAAGCTTACTCGAGGTCG AGGTCATTC-BHQ2 150

real-time PCR). The primers and probes were synthesised by Euro- cycle. Data were collected and processed using a Bio-Rad CFX Man- fins MWG Operon (Ebersberg, Germany). ager 2.0 (Bio-Rad Laboratories, Hercules, CA, USA). For quantitative analysis, standard curves were prepared using certified reference materials from IRMM containing 5% for 2.6. Qualitative PCR MON810 maize and 10% for TC1507 maize serially diluted (1/3) from 200 to 0.3 ng of template DNA. Each set of dilutions and sam- The amplifications by polymerase chain reaction (PCR) were ples was amplified in parallel reactions targeting the event under carried out in 25 lL total reaction volume containing 2 lLof test (MON810 with MON810 20–50/MON810 20–30 primers and DNA extracts, respectively, 15 mM Tris–HCl (pH 8.3), 50 mM KCl, MON810-Taq probe or TC1507 with TC-F/TC-R primers and TC-P 0.5 mM of each primer, 0.2 mM of each dNTP (Invitrogene, Carls- probe) and the endogenous reference gene for maize (with SSIIb bad, CA, USA), 1.5 mM for primers IVR1-F/IVR1-R and 2 mM MgCl2 1–5’/SSIIb 10–30 primers and SSIIb-Taq probe) (Table 3). For control for primers VW01/VW03 and TC-F/TC-R (Table 2), and 1 U of DNA and validation, the 1% reference material for each event was in- Ò polymerase AmpliTaq Gold (Applied Biosystems, Branchburg, NJ, cluded in all assays. In each assay, the standard dilutions were USA). amplified in duplicate and each sample was amplified in triplicate The PCR amplifications were performed in a thermal cycler MJ reactions. Each sample was analysed in at least two independent Mini (Bio-Rad Laboratories, Hercules, CA, USA) using the following assays. program: an initial denaturation at 95 °C for 5 min; 35 cycles of amplification for primers IVR1-F/IVR1-R and 40 cycles for primers VW01/VW03 and TC-F/TC-R, at 95 °C for 30 s, followed by anneal- 3. Results and discussion ing at 64, 65, and 61 °C for primers IVR1-F/IVR1-R, VW01/VW03 and TC-F/TC-R, respectively, for 30 s and extension at 72 °C for 3.1. Influence of processing on DNA degradation 30 s; and a final extension at 72 °C for 5 min. Each extract was amplified at least in duplicate assays. Food processing and consequent DNA degradation can affect The amplified fragments were analysed by electrophoresis in a DNA recovery and the efficiency of PCR and, therefore, the quality 1.5% agarose gel containing Gel Red 1x (Biotium, Hayward, CA, of analytical results, leading to under- or overestimation of GMO USA) for staining and carried out in SGTB 1x (GRiSP, Research Solu- content (Gryson, 2010). In this context, the effect of breadmaking tions, Porto, Portugal) for 20–25 min at 200 V. The agarose gel was processing on the extracted DNA from the three maize breads visualised under UV light and a digital image was obtained using a along the stages of preparation and location of sampling was per- Kodak Digital Science™ DC120 (Rochester, NY, USA). formed by agarose gel electrophoresis (Fig. 1A). The results show that dough samples before and after leavening the breads led to shearing and degradation of DNA of high molecular mass. The 2.7. Real-time PCR appearance of bands with the effect of ladder might be due to enzymatic activity present in both doughs. After oven cooking The amplifications by real-time PCR were carried out in 20 lL the breads (samples 3–5), it was noted a decrease of the DNA containing 2 lL of DNA extract, 1 of iQ™ Supermix (Bio-Rad Lab- amount and integrity by the obtained low intense sheared DNA. oratories, Hercules, CA, USA), primers and probes according to con- These results agree with other reports that observed DNA degrada- centrations in Table 3. The assays were performed on a tion after soybean bread cooking, evidencing a higher degradation fluorometric thermal cycler CFX96 Real-time PCR Detection Sys- on the top crust (Gryson et al., 2008), after baking cookies (Gryson tem (Bio-Rad Laboratories, Hercules, CA, USA) using the following et al., 2007), after baking wheat bread (Tilley, 2004) and after soy conditions: 95 °C for 5 min; 50 cycles at 95 °C for 30 s and 60 °C meal cooking (Murray, Butler, Hardacre, & Timmerman-Vaughan, for 1 min, with collection of fluorescence signal at the end of each 2009). 690 T.J.R. Fernandes et al. / Food Chemistry 138 (2013) 687–692

A Broa 1 Broa 2 Broa 3 M 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 6 10000 bp

1000 bp

200 bp

B Broa 1 Broa 2 Broa 3 M 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 6 N

C Broa 1 Broa 2 Broa 3 M 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 6 N

D Broa 3 M 1 2 3 4 5 7 N

Fig. 1. Agarose gel electrophoresis of extracted DNA (A) and respective PCR amplification products for the specific detection of invertase gene (B), MON810 event (C) and TC1507 event (D) from the 3 maize breads along breadmaking and sampling location. Legend: lane 1, dough (before yeast); lane 2, dough (after yeast); lane 3, top crust; lane 4, under crust; lane 5, middle soft part; lane 6, 100% MON810 maize flour; lane 7, positive control (reference material TC1507 1%, IRMM, Geel, Belgium); M, DNA ladder; N, negative control.

3.2. Influence of processing on PCR amplification ments (Vijayakumar et al., 2009), refined (Costa, Mafra, Amaral, & Oliveira, 2010) and crude vegetable oils (Costa, Mafra, & Oliveira, To evaluate the degree of DNA degradation on PCR amplifica- 2012). Vijayakumar et al. (2009) demonstrated that amplification tion, extracted DNA from the three different breads along the pro- of fragments <200 bp are most suited for the detection of GMO duction and sampling part were amplified targeting sequences of in highly processed foods, while Costa et al. (2010) demonstrated the maize invertase gene and the events MON810 and TC1507. that fragments of 6100 bp are highly recommended to detect The obtained PCR fragments were strong bands for all samples DNA in vegetable oils. (Fig. 1B–D), regardless the target sequence, although those of crust were, generally, less intense, agreeing with previous findings and 3.3. Performance of real-time quantitative PCR Gryson et al. (2008). The other PCR fragments in all the tested samples displayed similar amplification characteristics, though prod- The quantification of GM traits was performed targeting starch ucts from maize flour gave the most intense bands (Fig. 1B and synthetase IIb (SSIIb) as taxon-specific maize gene (ISO21570, C, lane 6). The lower level of amplification for the samples taken 2005) for both events, construct-specific sequence for MON810 from the top crust might be explained by the heat exposure that (ISO21570, 2005) and event-specific sequence for TC1507 maize decreased DNA integrity. Baking is one of the heat processes that (La Paz et al., 2006). The double calibration curve method with a most degrades DNA, making it sheared into smaller fragments set of seven DNA concentrations (100–0.137 ng/lL of maize DNA) and, therefore, reducing PCR sensitivity (Gryson, 2010). was used for estimating DNA amount from the unknown samples. It is important to highlight that all the samples were positively The prerequisites for the evaluation and comparison of the real- amplified because the target DNA fragments were lower or not time PCR systems were based on the available document of the much longer than 200 bp. Several studies have demonstrated that definition of minimum performance requirements for analytical the choice of relatively small DNA fragments is critical when ana- methods of GMO testing (ENGL (European Network of GMO Labo- lysing processed foods like potato (van der Colff & Podivinsky, ratories). Definition of minimum performance requirements for 2008), cooked and extruded maize (Murray, Butler, Hardacre, & analytical methods of GMO testing. Brussels: European Commis- Timmerman-Vaughan, 2007), processed soy-based foods (Murray sion, 2008). The presented real-time PCR results were according et al., 2009), soybean and maize subjected to several heat treat- to those requirements since the correlation coefficient (R2)of T.J.R. Fernandes et al. / Food Chemistry 138 (2013) 687–692 691

Table 4 Quantitative real-time PCR results of MON810 and TC1506 events along breadmaking processing and at different locations in the maize bread.

Sample Maize event (%) (mean ± SD)* Broa 1 Broa 2 Broa 3 MON810 TC1507 MON810 MON810 Dough Before yeast 33.2 ± 0.8a 11.0 ± 0.5b 31.0 ± 1.0a 48.8 ± 2.0b After yeast 29.5 ± 2.1a 14.7 ± 0.6a 21.8 ± 0.1bc 73.1 ± 0.8a Baked bread Top crust 22.0 ± 0.4bc 14.6 ± 0.4a 23.0 ± 2.0b 39.8 ± 0.1c Under crust 26.6 ± 0.9ab 6.2 ± 0.4d 18.4 ± 0.8 cd 50.6 ± 2.6b Middle soft part 19.6 ± 0.8c 8.2 ± 0.5c 16.6 ± 0.8d 41.9 ± 1.3c Mean parts 22.7 9.7 19.3 44.1 Actual values 20 11 44 67 Errore 0.14 0.12 0.56 0.34

e Error, ((mean value-true value)/true value). * Values represent mean and standard deviation (SD). In each column, different letters indicate significant differences between sampling zone (p<0.05). Fig. 2. Comparison of the estimated GMO content of maize bread samples (4, broa 1, s, broa 2 and 3) with the actual values. The dotted line is 1:1 correspondence. standard curves was, generally, P0.98, while PCR efficiencies ranged from 91% to 98%, indicating the adequacy of the standard (Murray et al., 2007), on soybean endogenous and transgenic curves for quantification. Moreover, the validation results using DNA (Chen, Wang, Ge, & Xu, 2005) and on potato endogenous CRM containing 1% of each tested event confirmed the acceptabil- and transgenic DNA (van der Colff & Podivinsky, 2008). Demonstra- ity of the methods to estimate MON810 and TC1507 events in tions that several processing treatments are responsible to cause maize since the estimated values were 0.97 ± 0.21% and degradation of endogenous and transgenic DNA have been fully de- 1.1 ± 0.3%, respectively. scribed. However, those studies lack information about the extent and accuracy of quantification, though real-time PCR has been used to assess maize endogenous DNA (Murray et al., 2007) and potato 3.4. Estimation of GMO content in maize breads endogenous and transgenic DNA (van der Colff & Podivinsky, 2008). Real-time quantitative PCR results of maize materials used in Another recent study performed by Bergerová et al. (2010) sta- the breadmaking confirmed the absence of GM material under test ted that heat-processing parameters have no practical influence on in other non-GM flours used in all breads, the 100% maize grains the quantification of transgenic maize and soybean in different (103.0 ± 5.7%) used in broa 2 and 3 and the simultaneous presence parts of bread loaf. Analysing the average quantitative results of of MON810 and TC1507 in maize semolina used in broa 1, with the broa 1, one can say that baking did not affect much the quantifica- proportions of 11% (10.9 ± 1.5%) and 20% (19.9 ± 2.2%), respec- tion, which agreed with Bergerová et al. (2010). However, some tively. Considering the described amounts of each raw material significant differences could be found in the GM contents of differ- used for breadmaking and respective GM contents (Table 1), we ent parts of breads, even in broa 1(Table 4). Additionally, those calculated the proportion of each GM event expected to be found authors did not analyse the influence of different particle size in the three prepared breads, which are represented as the actual material on the quantification of transgenic material used in bread- values in Table 4. making, although they tested the DNA extractability and quality of In broa 1, the overall estimated contents for maize MON810 and flour with different particle size and boiling times. TC1507 were very close to the actual values (Table 4). In broa 2 and Regarding the quantification of processed matrices and accord- 3, the contents of MON810 maize were generally underestimated, ing to Moreano, Busch, and Engel (2005) the size of amplicons for especially in bread samples after oven baking, indicating that the transgenic and reference genes should be equal. If the transgenic decrease of DNA integrity affected quantification. The overall re- amplicon was 25 bp longer than the reference gene, the GMO con- sults for all breads indicate that the higher estimates corresponded tent was underestimated, reflecting the favoured degradation of mostly to samples before cooking, and also to top crust in some the longer sequence used for the detection of the transgene. In cases. The underestimated values of broa 2 and 3, in comparison the present work, the amplified sequence of TC1507 was much lar- to broa 1, might be related to the particle size of the maize added, ger than the reference (Table 3), but this difference did not cause since broa 1 contained GM maize semolina while the other two considerable underestimation of the mean value of baked parts, breads contained GM maize flour (smaller particle size and higher although the under crust and middle soft part were considerably level of mechanical processing than maize semolina). By plotting reduced and all parts had significantly different GM material con- the estimated GM contents in the three prepared breads versus tents. It is important to stress that this typical bread after baking the actual values, it becomes more evident that the contents of forms a very hard crust, which is a very heterogeneous material broa 2 and 3 were, generally, underestimated and with a higher to analyse. This might explain the large variability of results. To variability among samples (Fig. 2). It is important to stress that overcome this problem, we used a special laboratory mill to obtain though the underestimated contents corresponded to the higher complete fragmentation and homogenisation of each part of the values, none of them was out of the range of quantifications per- bread. Another point to highlight is the fact that although the top formed by real-time PCR. crust apparently exhibits lower amplification rate in qualitative This result suggests the possibility of detecting and tracing the PCR, that does not seem to affect the quantification by real-time GM material by both qualitative and real-time PCR in baked prod- PCR. ucts, but the ability of accurately quantify it, might be compro- Analysing the quantitative results of MON810 maize in all mised by processing. This finding agrees with the results of other breads, distinct results can be envisaged in broa 1 compared to authors about the effect of processing on maize endogenous DNA the other two breads. According to Moreano et al. (2005), if the re- 692 T.J.R. Fernandes et al. / Food Chemistry 138 (2013) 687–692 combinant amplicon was 16 bp shorter than the reference gene, Commission Regulation (CE) No. 1830/2003 of the European Parliament and of the the GMO content was overestimated. That was the case of the Council of 22 September 2003 concerning the traceability and labelling of genetically modified organisms and the traceability of food and feed products MON810 quantification in the present work since MON810 se- produced from genetically modified organisms and amending Directive 2001/ quence was 38 bp shorter than the reference gene. However, after 18/EC. Official Journal of the European Union, L 268, 24–28. baking broa 1 the mean MON810 content was only slightly overes- Costa, J., Mafra, I., Amaral, J. S., & Oliveira, M. B. P. P. (2010). Monitoring genetically modified soybean along the industrial soybean oil extraction and refining timated and very close to the actual. In opposition, in broa 2 and 3 processes by polymerase chain reaction techniques. Food Research International, the contents were considerably underestimated indicating that the 43, 301–306. sizes of amplified sequences were not so relevant in these cases. Costa, J., Mafra, I., & Oliveira, M. B. P. P. (2012). Advances in vegetable oil authentication by DNA-based markers. Trends in Food Science and Technology, Probably due to the use of maize flour with a substantial lower par- 26, 43–55. ticle size and higher degree of mechanical processing than maize ENGL (European Network of GMO Laboratories). Definition of minimum semolina was the most prominent effect in these two breads. Sem- performance requirements for analytical methods of GMO testing. Brussels: European Commission, 2008. URL: http://gmo-crl.jrc.ec.europa.eu/ olina is the result of incomplete milling of the grains and it is guidancedocs.htm. Accessed 19.03.12. appropriate and frequently used for the production of traditional Fernandes, T. J. R. (2011). Detecção e quantificação de milho geneticamente breads. modificado em alimentos comerciais e em broas de milho ao longo da produção. M.Sc., Thesis, Faculty of Pharmacy, University of Porto, Porto, Portugal. Fernandes, T. J. R., Costa, J., Oliveira, M. B. P. P., & Mafra, I. (2012). Organismos 4. Conclusions geneticamente modificados em alimentos: Conseguimos detectá-los? Agrotec, 2, 20–23. GMO Compass. GMO Database. Genetically Modified Food and Feed: Authorization In the present work, a traditional Portuguese bread with distinct in the EU, 2012. URL: http://www.gmo-compass.org/eng/gmo/db. Accessed textural and sensorial characteristics was the object of study for 07.05.12. the first time regarding DNA degradation, detection and quantifica- Gryson, N. K., Dewettinck, K., & Messens, K. (2007). Detection of genetically modified soy in doughs and cookies. Cereal Chemistry, 84, 109–115. tion of transgenic maize along breadmaking. As expected, DNA Gryson, N., Messens, K., & Dewettinck, K. (2008). PCR detection of soy ingredients in degradation was noticed after baking dough and in all parts of bread. European Food Research and Technology, 227, 345–351. bread sampling. The results of PCR amplification of extracted Gryson, N. (2010). Effect of food processing on plant DNA degradation and PCR- DNA from the three different breads along the breadmaking and based GMO analysis: A review. Analytical and Bioanalytical Chemistry, 396, 1–20. ISO 21569 (2005). Foodstuffs – Methods of analysis for the detection of genetically sampling parts of maize bread showed that the sequences for modified organisms and derived products- Qualitative nucleic acid based maize invertase gene and for events MON810 and TC1507 were methods, Switzerland. easily detected with strong bands. Real-time PCR showed that ISO 21570 (2005). Foodstuffs – Methods of analysis for the detection of genetically modified organisms and derived products- Quantitative nucleic acid based the sampling location of bread might had a limited influence on methods, Switzerland. quantification since the average contents of MON810 and TC1507 James, C. (2011). Global status of commercialized biotech/GM crops 2011. Ithaca, New events after baking were very close to the actual values in broa 1 York, USA: ISAAA. ISAAA brief no. 43. La Paz, J.-L., García-Muniz, N., Nadal, A., Esteve, T., Puigdomènech, P., & Pla, M. prepared with maize semolina. In the other two prepared maize (2006). Interlaboratory transfer of a real-time polymerase chain reaction assay breads subjected to the same baking treatment, the contents of for quantitative detection of genetically modified maize event TC-1507. Journal MON810 maize were considerably underestimated, leading to the of AOAC International, 85, 1347–1352. Lino, C. M., Silva, L. J. G., Pena, A., Férnandez, M., & Mañes, J. (2007). Occurrence of conclusion that heat-processing was not the responsible parameter fumonisins B1 and B2 in broa, typical Portuguese maize bread. International for that distortion, but the particle size and mechanical processing Journal of Food Microbiology, 118, 79–82. of raw maize play also a major role in GMO quantification. Mafra, I., Ferreira, I. M. P. L. V. O., & Oliveira, M. B. P. P. (2008). Food authentication by PCR-based methods. European Food Research and Technology, 227, 649–665. Mafra, I., Silva, S. A., Moreira, E. J. M. O., Silva, C. S. F., & Oliveira, M. B. P. P. (2008). Acknowledgments Comparative study of DNA extraction methods for soybean derived food products. Food Control, 19, 1183–1190. The authors acknowledge to Cristina Santos and Vítor Melo for Moreano, F., Busch, U., & Engel, K.-H. (2005). Distortion of genetically modified organism quantification in processed foods: Influence of particle size their contribution in the breadmaking of samples. This work has compositions and heat-induced DNA degradation. Journal of Agricultural and been supported by Fundação para a Ciência e a Tecnologia (FCT) Food Chemistry, 53, 9971–9979. through Grant no. PEst-C/EQB/LA0006/2011. Murray, S., Butler, R. C., Hardacre, A. K., & Timmerman-Vaughan, G. M. (2007). Use of quantitative real-time PCR to estimate maize endogenous DNA degradation after cooking and extrusion or in food products. Journal of Agricultural and Food References Chemistry, 55, 2231–2239. Murray, S., Butler, R. C., Hardacre, A. K., & Timmerman-Vaughan, G. M. (2009). Bergerová, E., Hrncˇírová, Z., Stankovská, M., Lopašovská, M., & Siekel, P. (2010). Quantitative real-time PCR assays to detect DNA degradation in soy-based food Effect of thermal treatment on the amplification and quantification of products. Journal of the Science of Food and Agriculture, 89, 1137–1144. transgenic and non-transgenic soybean and maize DNA. Food Analytical Shokere, L. A., Holden, M. J., & Jenkins, G. R. (2009). Comparison of fluorometric and Methods, 3, 211–218. spectrophotometric DNA quantification for real-time quantitative PCR of Bergerová, E., Godálová, Z., & Siekel, P. (2011). Combined effects of temperature, degraded DNA. Food Control, 20, 391–401. pressure and low pH on the amplification of DNA of plant derived foods. Czech Tilley, M. (2004). PCR amplification of wheat sequences from DNA extracted during Journal of Food Sciences, 29, 337–345. milling and baking. Cereal Chemistry, 81, 44–47. Brites, C., Trigo, M. J., Santos, C., Collar, C., & Rosell, C. M. (2010). Maize-based van der Colff, L., & Podivinsky, E. (2008). Cooking DNA: The effect of ‘domestic’ gluten-free bread: Influence of processing parameters on sensory and cooking methods on detection of GM potato. International Journal of Food Science instrumental quality. Food and Bioprocess Technology, 3, 707–715. and Technology, 43, 2105–2112. Chen, Y., Wang, Y., Ge, Y., & Xu, B. (2005). Degradation of endogenous and Vaz Patto, M. C., Moreira, P. M., Carvalho, V., & Pego, S. (2007). Collecting maize (Zea exogenous genes of Roundup-Ready soybean during food processing. Journal of mays L. convar. mays) with potential technological ability for bread making in Agricultural and Food Chemistry, 53, 10239–10243. Portugal. Genetic Resources and Crop Evolution, 54, 1555–1563. Commission Regulation (CE) No. 1829/2003 of the European Parliament and of the Vijayakumar, K. R., Martin, A., Gowda, L. R., & Prakash, V. (2009). Detection of Council of 22 September 2003 on genetically modified food and feed. Official genetically modified soya and maize: Impact of heat processing. Food Chemistry, Journal of the European Union, L 268, 1–37. 117, 514–521.