Transgenes monitoring in an industrial soybean processing chain by DNA-based conventional approaches and biosensors
ABSTRACT The development of analytical methods for genetically modified organisms (GMO) screening is of great interest. In particular, since even highly processed GMO-derived food products are covered by new European legislations, a great effort has been devoted to the application of the analytical tests to these products.This work describes a polymerase chain reaction-based qualitative screening assay and a biosensor-based approach to detect transgenes in a Roundup Ready® soybean processing line. Roundup Ready® soybean was specifically analyzed in eight types of processed materials – seeds, crushed seeds, expander, crude flour, proteic flour, crude oil, degummed oil and lecithin – all derived from the same initial source and produced during the manufacturing process. Specific combinations of primers were used to differentiate sequences from the whole insert. The amplification of “marker” fragments with a maximum length of 500 bp was successfully achieved both in raw material (seeds) and in partially (crushed seeds, crude and proteic flours) and highly (crude and degummed oils and fluid lecithin) processed materials.Moreover, the extraction procedure was optimised and the polymerase chain reaction-electrophoresis analysis has been implemented by a biosensor-based approach.
- [Show abstract] [Hide abstract]
ABSTRACT: The authenticity of edible oils has recently attracted increased attention, particularly due to the emergence of swill-cooked dirty oil. There has not been an effective method to detect the authenticity of edible oils due to the diversity of edible oils, the trace DNA in oils and serious DNA degradation. In this study, an efficient method was developed and optimized to detect the authenticity of edible oils. The stability and efficiency of detecting the authenticity of edible oils were dramatically increased by the DNA extraction method, amplification fragment selection and nano-real time PCR. The DNA extraction method was optimized with regard to the type and the amount of organic reagent, the types and the amounts of extraction buffer, and the DNA carriers. The types of organic reagents and DNA carriers determined whether DNA could be successfully extracted from different edible oils. The amount of DNA extracted from edible oils and adulterated edible oils with the optimized extraction method was evaluated by real-time PCR with different amplification fragments and nano-real time PCR. The results showed that 10 mL soybean oil can be detected against a background of 40 mL sesame oil. Additionally, the new nano real-time PCR technology was applied to amplify the DNA in edible oils at the first time. The efficiency and the precision of the PCR were dramatically increased by gold colloid. The results from this research are beneficial for detecting the authenticity of edible oils.Food Control 05/2013; 31(1):71–79. · 2.82 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: The suitability of DNA markers in providing unequivocal identifiers for authentication and traceability of food has been a subject of an increasing number of reports. Even in complex food matrices such as vegetable oils, the use of molecular markers as diagnostic tools has been exploited. Considering the wide variety of vegetable oils available for consumers and the differences in prices, especially among premium olive oil and other oils, species adulteration leading to economic losses and loss of consumer confidence can arise. In this review, the advances of DNA extraction protocols are emphasised as a crucial step to overcome. Specific identification of several plant oils as potential adulterants of olive oil has been a subject of very recent progresses. When the oilseed crops are the source for vegetable oil production, additional concerns due to the presence of genetically modified organisms have prompted to further improvements in DNA analysis. In the specific case of olive oil, the use of genetic markers has provided analytical tools to assess authenticity regarding cultivar identification as independent markers from environmental fluctuations.Trends in Food Science & Technology 07/2012; 26(1):43-55. · 4.65 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: In order to detect presence and quantity of Roundup Ready (RR) soybean in crude oil extracted from soybean seed with a different percentage of GMO seed two extraction methods were used, CTAB and DNeasy Plant Mini Kit. The amplifications of lectin gene, used to check the presence of soybean DNA, were not achieved in all CTAB extracts of DNA, while commercial kit gave satisfactory results. Comparing actual and estimated GMO content between two extraction methods, root mean square deviation for kit is 0.208 and for CTAB is 2.127, clearly demonstrated superiority of kit over CTAB extraction. The results of quantification evidently showed that if the oil samples originate from soybean seed with varying percentage of RR, it is possible to monitor the GMO content at the first stage of processing crude oil.Food Chemistry 02/2014; 145C:1072-1075. · 3.26 Impact Factor
Transgenes monitoring in an industrial soybean processing chain
by DNA-based conventional approaches and biosensors
Patrizia Bogania, Maria Minunnib,*, Maria M. Spiritia, Michele Zavagliab, Sara Tombellib,
Marcello Buiattia, Marco Mascinib
aDipartimento di Biologia Evoluzionistica ‘‘Leo Pardi”., Università degli Studi di Firenze, Via Romana 17-19, Firenze, Italy
bDipartimento di Chimica, Università degli Studi di Firenze, Polo Scientifico, Via della Lastruccia 3, 50019 Sesto F.No., Firenze, Italy
a r t i c l ei n f o
Received 26 March 2008
Received in revised form 10 July 2008
Accepted 19 July 2008
Food processing chain
a b s t r a c t
The development of analytical methods for genetically modified organisms (GMO) screening is of great
interest. In particular, since even highly processed GMO-derived food products are covered by new Euro-
pean legislations, a great effort has been devoted to the application of the analytical tests to these prod-
This work describes a polymerase chain reaction-based qualitative screening assay and a biosensor-
based approach to detect transgenes in a Roundup Ready?soybean processing line. Roundup Ready?soy-
bean was specifically analyzed in eight types of processed materials – seeds, crushed seeds, expander,
crude flour, proteic flour, crude oil, degummed oil and lecithin – all derived from the same initial source
and produced during the manufacturing process. Specific combinations of primers were used to differen-
tiate sequences from the whole insert. The amplification of ‘‘marker” fragments with a maximum length
of 500 bp was successfully achieved both in raw material (seeds) and in partially (crushed seeds, crude
and proteic flours) and highly (crude and degummed oils and fluid lecithin) processed materials.
Moreover, the extraction procedure was optimised and the polymerase chain reaction-electrophoresis
analysis has been implemented by a biosensor-based approach.
? 2008 Elsevier Ltd. All rights reserved.
In recent years a great effort has been devoted to the develop-
ment of new methods for the qualitative and quantitative detec-
tion of transgenic sequences in food. The European Union (EU)
has elaborated a legislation for genetically modified (GM) food
control, which establishes both the legal basis for the approval pro-
cedure of GMOs and the post market traceability and labelling
requirements for GMOs and GMO-derived food and feeds (Euro-
pean Commission, 2003a, 2003b).
Most of the developed analytical methods for GMO detection
are DNA-based, since protein-based assays are not suitable for
processed food. Polymerase chain reaction (PCR) and real time
PCR-based methods have been generally accepted for regulatory
compliance. They are widely used to amplify target DNA fragments
corresponding to marker genes, promoter and terminator se-
quences, or to transgene-host integration junctions (Ahmed,
2002; Anklam, Gadani, Heinze, Pijnenburg, & van den Eede, 2002;
Corbisier et al., 2007; Miraglia et al., 2004; Rodriguez-Lazaro
et al., 2007).
reproducible methods have been recently developed for the identi-
fication of different GMO events (Bordoni, Germini, Mezzelani,
Marchelli, & De Bellis, 2005; Germini et al., 2005; Kim, Chae, Chang,
& Kang, 2005; Peano et al., 2005; Rossi et al. 2007; Xu et al., 2006).
Among these, DNA biosensors based on several transduction princi-
ples(Gambari&Feriotto,2006;Giakoumakiet al., 2003;Kalogianni,
Koraki, Christopoulos, & Ioannou, 2006; Leimanis et al., 2006;
Mannelli, Minunni, Tombelli, & Mascini, 2003; Minunni et al.,
2005; Passamano & Pighini, 2006) are methods able to detect target
sequences in complex media and in real time.
The official methods for the detection of GMO events in food
and the quantification of material deriving from authorised trans-
genic crops are provided by the European Commission reference
laboratory, but when dealing with surveillance testing of GMOs
in highly processed food, researchers must face with major analyt-
ical challenges. Moreover, since each food and each processing
phase contribute to a unique environment, more data on the kinet-
ics of DNA degradation and amplificability along the whole indi-
vidual food processing chain are necessary.
The traceability of transgenes in processed food has been stud-
ied by selecting different set of primers to amplify regions of the
endogenous genes varying in length (Germini et al., 2005) co-
amplifying, as internal control, also a specie-specific housekeeping
0308-8146/$ - see front matter ? 2008 Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: +39 0554573314; fax: +39 0554573384.
E-mail address: firstname.lastname@example.org (M. Minunni).
Food Chemistry 113 (2009) 658–664
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/foodchem
gene, naturally present in the genome of the analyzed species (i.e.,
zein in Zea mays and lectin in Glycine max) (Meyer, Chardonnens,
Hubner, & Luthy, 1996; Meyer & Jaccaud, 1997). Despite the inter-
esting aspects of these studies, most of the reported work refers to
raw material, such as flours or seeds, or to the final product of food
processing (Aarts, van Rie, & Kok, 2002; Ahmed, 2002; Di Pinto et
al., 2007; Engel, Moreano, Ehlert, & Busch, 2006; Germini et al.,
2005; Passamano & Pighini, 2006;Peano et al., 2005). A systematic
study on the traceability of GM DNA along a food processing chain
has never been reported, and only laboratory model studies have
been conducted on different processing procedures (Bauer, Weller,
Hames, & Hertel, 2003).
The correct traceability of the transgenic events in food process-
ing can be achieved only by monitoring target sequences during
the whole manufacturing process, where the target analytes are
followed from the initial source to the end product, in a systematic
and controlled manner passing through a series of intermediate
phases. It should be mentioned that the analysis of processed food
encounters a certain number of complications, negatively affecting
the detection of transgenes, on dependence of the considered
matrices. Complications can be ascribed to manufacturing pro-
cesses involving enzymes or chemical reactions, as well as simple
mechanical procedures, such as milling, or thermal treatments.
Moreover, the use of harsh conditions in the processing steps could
impact on the integrity of genomic DNA, which represents the
starting point of the whole analysis (Moreano, Busch, & Engel,
2005). It is thus of primary importance to verify the possibility to
amplify (amplificability) the extracted DNA. The DNA degradation
can be caused, for example, by heating treatments (Ahmed, 2002)
and exposure to pH changes (Aarts et al., 2002). Finally, another
important factor which could affect the amplification step, leading
also to false negative results should be taken into account, which is
the presence of different inhibitors of the enzyme Taq polymerase
such as proteins, fats, polysaccharides, polyphenolics and other
compounds that may be present in DNA extracted from food matri-
ces (Corbisier et al., 2007; Di Pinto et al., 2007; Lüthy, 1999; Peano
et al., 2005).
In this work we report for the first time the study of transgene
traceability along a complete industrial soybean processing chain.
The analysed samples were collected from the different levels of an
industrial soybean processing plant working with 80% GM Round-
up Ready?soybean. The genetic modification is the Roundup
Ready?soybean event GTS 40-3-2 (notification C/UK/94/M3/1 of
Monsanto). In particular, the analysis was conducted aiming to
study the traceability of different parts of the inserted cassette in
all the different matrices, derived from the same source, produced
during the manufacturing process (Fig. 1). For this purpose, an
optimised extraction procedure was applied and various primer
couples were set for PCR amplification of the event cassette (Fig. 2).
The ‘‘marker” fragment which should be followed along the soy-
bean processing procedure has been identified and the electropho-
retic analysis was at the end replaced by a biosensor-based
approach. Biosensor development aims at making the analytical
approach simpler, faster and less costly. Moreover, biosensors ex-
hibit several advantages with respect to electrophoresis which is
not automatable and provides no sequence confirmation. The find-
ings confirms the suitability of biosensor analysis not only for raw
materials (Kalogianni et al., 2006; Mannelli et al., 2003; Minunni
et al., 2005) but also for process control and traceability of GM
crops in industrial processing plants.
2. Materials and methods
2.1. Soybean samples
The soybean samples were kindly provided by Dr. R. Onori (Isti-
tuto Superiore di Sanità, Rome, Italy). The samples were collected
from the different levels of an industrial soybean processing plant
working with 80% GM Roundup Ready?soybean. The genetic mod-
tion C/UK/94/M3/1 of Monsanto) which was developed to be
of a gene encoding the glyphosate tolerant 5-enol-pyruvylshiki-
Fig. 1. Flow chart of a soybean processing chain. The products analysed in this
study are highlighted with an asterisk.
Fig. 2. (A) Schematic diagram of part of heterologous DNA event RR GTS-40-3-2.
Arrows indicate the positions of different primers used for PCR qualitative analysis.
(B) List of primers used for the PCR analysis.
P. Bogani et al./Food Chemistry 113 (2009) 658–664
mate-3-phosphate synthase from Agrobacterium sp. CP4 (CP4
The soybean processing steps along the industrial chain are
illustrated in Fig. 1. The samples used in this work for the extrac-
tion of DNA cover all the different stages and they consisted in
seeds, crushed seeds, expander, crude flour, proteic flour, crude
oil, degummed oil and lecithin.
PCR qualitative analysis was performed by using the following
Certified Reference Materials: non-GM soybean powder and soy-
bean powder containing 0.1%, 0.5%, 1%, 2% and 5% of powder pre-
pared from genetically modified
(IRMM-410S, Fluka, Sigma–Aldrich, Milan, Italy).
2.2. DNA extraction
DNA was extracted from all the soybean samples by using a
‘‘Wizard?Magnetic DNA Purification System for Food” kit, Prome-
ga (Leiden, The Netherlands). The method was slightly modified to
improve the yield of the extraction from highly processed prod-
ucts. One hundred milligram per sample of seeds, flours, expander
and fluid lecithin were ground in a mortar with liquid nitrogen to a
fine powder before their suspension in the extraction buffer. DNA
extraction was then performed according to the manufacturer
instructions. With crude and degummed oils, 500 lL of the relative
sample were mixed with 500 lL of chloroform instead of hexane
(this latter, suggested by the manufacturer), 500 lL of extraction
buffer A (Promega kit), 250 lL of extraction buffer B (Promega
kit) and then centrifuged at 11,000g for 5 min. The supernatant
was removed, thoroughly remixed with 500 lL of chloroform, cen-
trifuged at 11,000g for 5 min, and transferred into a new tube con-
taining 750 lL of precipitation buffer (Promega kit) and 500 lL of
chloroform. The mixture was mixed thoroughly for 1 min and cen-
trifuged at 11,000g for 5 min. The aqueous phase was then mixed
with 50 lL of Magnesil
After extraction, the DNA was further purified by adding an
(50:49:1). The suspension was gently mixed for 1 min and then
centrifuged at 4000g for 5 min. Finally the extracted DNA was puri-
fied by precipitation at ?20 ?C overnight, after mixing with abso-
lute ethanol, 3 M sodium acetate (pH 5.2) in a 1:2.5:0.1 volume
ratio and centrifugation for 30 min at 11,000g (4 ?C). The resulting
pellet was washed with 500 lL 70% ethanol, air dried and re-sus-
pended in 20 lL of distilled water until use.
The DNA concentration was measured by UV absorption at
260 nm, while DNA purity was evaluated on the basis of the UV
absorption ratios at 260/280 nm and 260/230 nm. The concentra-
tion of the DNA samples was evaluated by DyNA Quant 200 fluo-
rometer (Höefer, Pharmacia Biotech, Uppsala, Sweden).
TMPMPs (Promega kit) according to the man-
2.3. PCR analysis
The extracted DNA was diluted with distilled water to a final
concentration of 50 ng/lL for seeds and flours and of 10 ng/lL
for oils and lecithin. PCR was performed by using a final volume
of 25 lL of these DNA stock solutions. Each reaction contained
1 lL of template DNA, 1.5 U of DNA Taq polymerase (Amersham,
UK), 1? reaction buffer for DNA polymerase, 2.5 or 1.5 mM MgCl2
and 200 lM dATP, dTTP, dGTP, dCTP (Amersham, UK). The detec-
tion of sequences from the transgenic GTS 40-3-2 cassette, was
performed by using 200 nM of each primer. The PCR program con-
sisted of the hot start activation of the polymerase at 95 ?C for
5 min. The amplification profile comprised 40 cycles of 45 s at
94 ?C (denaturation step), 45 s at 55 ?C (annealing step) and 45 s
for short fragments or 1 min and 10 s for long fragments at 72 ?C
(extension step), followed by a final extension at 72 ?C for
10 min. Amplification reactions were performed in a MJPT200
thermocycler and PCR products were separated on a 2% agarose
(Roche Diagnostics, Milan, Italy) gel in 1? Tris acetate EDTA
(TAE) buffer at constant voltage (8 V cm?1) for 30 min. Gels were
visualised with a Gel Doc 2000 gel imaging system (BioRad, Inc.,
The CP4 EPSPS cassette contained in the plasmid vector used
to produce the transgenic event GTS 40-3-2, which was consid-
ered in this work, is reported in Fig. 2A. In the cassette, the CP4
EPSPS coding sequence is linked to a chloroplast transit peptide
sequence designated CTP4, under the control of the 35S cauli-
flower mosaic virus promoter (P-35S) and it is joined to the
nopaline synthase 30non-translated sequence (T-NOS) from Agro-
bacterium tumefaciens. The primers used in the PCR analysis and
listed in Fig. 2B are internal to the cassette and they cover the
P-35S promoter, the CTP4 region and the CP4 EPSPS coding
2.4. Biosensor development
2.4.1. Reagents and instrumentation
AT-Cut (9.5 MHz) quartz crystals (14 mm) with gold evaporated
(42.6 mm2area) on both sides were purchased from International
Crystal Manufacturing (USA). The quartz crystal analyzer used for
the measurements was the QCMagic by Elbatech (Marciana,
11-Mercaptoundecanol and Dextran 500 were from Sigma
(Milan, Italy); (+)/?epichlorohydrin and N-hydroxysucciminide
were purchased from Fluka (Milan, Italy). Ethanol and all the
reagents for the buffers were purchased from Merck (Italy). Two
different buffers, immobilisation buffer (NaCl 300 mM, Na2HPO4
20 mM, EDTA 0.1 mM, pH 7.4) and hybridisation buffer (NaCl
150 mM, Na2HPO420 mM, EDTA 0.1 mM, pH 7.4), were used.
The following oligonucleotides were purchased from MWG Bio-
tech (Milan, Italy): biotinylated probe (35S, 25 mer) 50Biotin-GGC
CAT CGT TGA AGA TGC CTC TGC C-30; complementary target (35S,
25 mer) 50-GGC AGA GGC ATC TTC AAC GAT GGC C-30; non-com-
plementary (Tnos, 25 mer) 50-GAT TAG .AGT CCC GCA ATT AAT
The biotinylated probe was immobilised via biotin-streptavidin
binding on the gold sensor surface, previously modified with thiol/
dextran/streptavidin (Mannelli et al., 2003).
2.4.3. Hybridisation reaction
Sixty microliter of the DNA solution (synthetic oligonucleotides
or PCR amplified samples) in hybridisation buffer were added to
the measuring cell and incubated with the immobilised probe for
10 and 20 min with oligonucleotides or amplified samples, respec-
tively. The surface was then washed with the same hybridisation
buffer to eliminate the unbound oligonucleotides. The reported
frequency shifts (DF) are the difference between this final value
and the value displayed before the hybridisation reaction. The
signal is considered positive when DF > 3 Hz, which represents
three times the blank signal, both for oligonucleotides and PCR
After each cycle of hybridisation, the single stranded probe on
the crystal surface was regenerated by two consecutive treatments
of 30 s with 1 mM HCl allowing a further use of the sensor.
All the experiments are performed at room temperature (25 ?C).
2.4.4. Thermal denaturation
The PCR DNA samples were heated at 95 ?C for 5 min and then
cooled in ice for 1 min and immediately injected in the cell.
P. Bogani et al./Food Chemistry 113 (2009) 658–664
3. Results and discussion
3.1. Optimisation of the PCR procedure for DNA extracted from
samples representing the whole industrial soybean processing chain
3.1.1. Optimisation of DNA extraction
The soybean samples originating from the different steps of an
industrial soybean processing plant (Fig. 1) were all extracted
using the Wizard?magnetic method which had already been suc-
cessfully used to recover DNA even from complex foodstuff (Corbi-
sier et al., 2007; Peano et al., 2005). Although successful DNA
extraction was obtained with CRMs and processed soybean sam-
ples from the whole chain, no amplification of the endogenous lec-
tin gene, used as internal standard, was achieved with DNA
extracted from oils and lecithin (data not reported). As frequently
discussed when dealing with processed materials, the failure of the
amplification procedure is often related to the low quality of the
extracted DNA in terms of fragment length and presence of inhib-
itory factors (Miraglia et al., 2004).
In order to improve both the quality and the quantity of the ex-
tracted DNA, some modifications in the extraction protocol and
additional purification steps were introduced. In particular, an in-
creased amount of starting material was employed (2- up to 10-
fold increase). Moreover, chloroform was used, replacing hexane
in the DNA purification step, in analogy to the extraction protocol
proposed by Doyle and Doyle (1989) based on the use of CTAB
(Hexadecyl trimethyl ammonium bromide) and further modified
(Bogani et al., 1995). The resulting amount of extracted DNA,
shown in Fig. 3A, varied for the different processed materials.
The lowest amount of DNA was obtained from oils and lecithin,
Fig. 3. (A) Total amount and sample purity values of DNA extracted from differently processed soybean products of an industrial supply chain. (B) Agarose gel electrophoresis
of DNA extracted from different soybean samples using a modified Wizard method: seeds (lane 1), crushed seeds (lane 2), expander (lane 3), crude flour (lane 4), proteic flour
(lane 5), crude oil (lane 6), degummed oil (lane 7) and lecithin (lane 8). M: 1 kb ladder, Gibco BRL.
Fig. 4. (A) Qualitative PCR analysis of the housekeeping lectin gene (a), of gene-specific CTP4 marker (b), screening P-35S marker (c) and construct specific (d–f) in soybean
chain processed materials. M: 50 bp marker (Fermentas); 1: seeds; 2: crushed seeds: 3: expander; 4: crude flour; 5: proteic flour; 6: crude oil; 7: degummed oil; 8: lecithin; 9:
negative PCR control (no DNA). (B) PCR amplification of the gene specific (a), screening specific (b) and construct specific (c and d) GTS 40-3-2 markers in soy flours certified
reference materials. [M: 100 bp ladder, Fermentas. GM-free soy flour (lane 1), 0.1% (lane 2), 0.5% (lane 3), 1% (lane 4), 2% (lane 5), 5% (lane 6), negative PCR control (lane 7).]
P. Bogani et al./Food Chemistry 113 (2009) 658–664
in line with the known difficulties in extracting DNA from oils and
with the fact that such kind of products are characterised by extre-
mely low traces of DNA (Miraglia et al., 2004).
Moreover, despite the applied purification steps, A260/A280and
A260/A230 values still demonstrate the presence of proteins, fat
and polysaccharides in all the DNA samples.
To check the DNA quality in terms of size and degradation level,
the extracted samples (100 ng) were analysed by gel electrophore-
sis (Fig. 3B). The results showed that DNA from proteic flour (lane
5) is highly degraded with an average fragment size below 500 bp.
DNA extracted from oils and lecithin (lanes 6–8), was instead not
visible on the agarose gel due to an even higher degradation level
caused by repeated thermal and chemical treatments in the food
3.1.2. PCR amplification
To assess the amplificability of the DNA extracted with the im-
proved procedure, a fragment of 118 bp, corresponding to a part of
the endogenous lectin gene, was amplified in all the samples using
GM03 and GM04 primer pair (Meyer & Jaccaud, 1997). As shown in
Fig. 4A (set a), PCR products were obtained from all the DNA sam-
ples. To achieve these results, the amount of template DNA to be
used in the amplification procedure has been varied for the differ-
ent samples. In particular, 50 ng of template DNA were used for
seeds, crushed seeds, expander and flours, whereas 10 ng were em-
ployed in the case of DNA extracted from oils and lecithin. The
improvements in amplificability of the DNA extracted from these
latter materials with a lower amount of DNA template, could be re-
lated to the consistent lower amount of PCR inhibitory factors in
these non-pure samples.
After PCR optimisation using the lectin gene, all the samples
were amplified using different primer pairs (Fig. 2B) and analysed
by gel electrophoresis (Fig. 4A (sets b–g)). In addition, as positive
control, CRMs of soybean flours containing 0%, 0.1%, 0.5%, 1%, 2%
and 5% of Roundup Ready?soybean were analysed with the same
selected primer pairs (Fig. 4B).
Fig. 4A reports the PCR results obtained with the different soy-
bean samples for the amplification of the short CTP4 188 bp frag-
ment (set b), of the P-35S promoter 195 bp fragment (set c), of
the construct-specific CP4 EPSPS 644 and 1006 bp fragments (sets
e and g) and the construct-specific P-35S CTP4 470 and 900 bp
fragments (sets d and f).
The amplificability of the fragments was inversely related to
their size since the degree of template DNA integrity affected the
success of the PCR reaction, especially in the case of long ampli-
cons. This could explain the finding that only short fragments
(up to 470 bp) could be detected in highly processed samples,
while both, long and short amplicons can be found into raw or little
processed matrices. These results are in agreement with the work
performed on soybean and maize processed food, where the
amplificability of short, medium and long fragments of endoge-
nous genes was studied (from 169 bp up to 1626 in soybean and
133 up to 1956 in maize, respectively) using different extraction
protocols (Peano et al., 2005).
Finally, the 188, 195, 470, 644, 900 and 1006 bp fragments were
cloned and sequenced to check their identity, which was confirmed
(data not shown).
This demonstrates that for a correct traceability of the target
transgenic event along an industrial food (soybean) processing
chain, which means working with raw and highly processed mate-
rials, short PCR fragments should be taken into consideration for
the amplification. The processing chain which was considered is
characterised by the presence of a high percentage (80%) of trans-
genic material, thus, to more generally apply the proposed ap-
proach as traceability method in food processing, it is important
to clarify if the selected fragments are the correct choice also when
the transgenic material is present in low percentage.
Fig. 5. (A) Sequence of the 195 bp PCR fragment containing the two primers (in
grey) and the probe (underlined). (B) Frequency changes during the hybridisation of
the complementary oligonucleotide (0.1 lM). (C) Frequency changes after the
addition of a non-complementary oligonucleotide (1 lM). (D) Frequency changes
during the hybridisation of a PCR fragment after the denaturation (sample:
P. Bogani et al./Food Chemistry 113 (2009) 658–664
With this aim, the same fragments were analysed in soybean
flour containing from 0.1% up to 5% Roundup Ready?soybean
(CRM). Fig. 4B shows that 188 bp (CTP4) and 195 bp (P-35S) frag-
ments were detected in all the samples, while longer fragments
(900 and 1006 bp) could be detected in CRMs with a different sen-
sitivity. In particular, the 900 bp fragment was detectable in all the
tested samples (0.1%, 0.5%, 1%, 2% and 5%) while the 1006 bp frag-
ment was detected only in soybean flour containing 1%, 2% and 5%
The final out-coming of these data was that 188, 195 and 470 bp
fragments can be detected in all the matrices along the whole soy-
bean processing chain, while longer fragments were not detected
in highly processed materials, such as proteic flour, oils and leci-
thin. Moreover, when dealing with CRMs at a low percentage of
transgenic materials, shorter fragments are still the best choice
for a complete traceability.
Thus, for the correct traceability of transgenic materials along
an industrial processing chain, it is important to analyze short frag-
ments (188, 195 or 470 bp), and this finding seems independent
from the transgenic material content, as indicated by the results
conducted on CRMs.
The first step in the development of the biosensor for the trace-
ability of GMOs along the industrial soybean processing chain was
the choice of the probe that should be immobilised onto the sensor.
Since the study conducted on the PCR detection has identified as
‘‘marker” fragment the 195 bp one internal to the Roundup Ready?
soybean P-35S promoter, the probe sequence should be comprised
in this fragment.
A 25 bp probe has been chosen in the P-35S amplified fragment
and immobilised on the sensing surface of piezoelectric quartz
crystals as already reported in other works by this group (Fig.
5A) (Mannelli et al., 2003; Minunni et al., 2005). The good analyt-
ical performances of the biosensor have been confirmed (Mannelli
et al., 2003) by using synthetic 25 bp complementary oligonucleo-
tides. Key characteristics for the further application to the analysis
of complex samples, such as sensitivity, reproducibility and selec-
tivity have been verified (Mannelli et al., 2003). The sensorgrams
reporting the signal during the hybridisation between the immobi-
lised probe and the target DNA, and the interaction between the
probe and the non-complementary DNA are reported in Fig. 5B
and C, respectively.
The calibrated biosensor has been further applied to the analy-
sis of the PCR fragments representing all the different stages along
the soybean industrial processing chain. A simple and fast pre-
treatment (95 ?C 5 and 1 min cooling in ice) to denaturate the
amplicon double helix to allow hybridisation with the probe was
performed. The sensorgram reporting the frequency changes dur-
ing the hybridisation between a PCR fragment (from a soyflour
sample) and the immobilised probe is reported in Fig. 5D. In this
case the kinetic is influenced by the initial temperature of the sam-
ple which is added to the cell after the final step of the denatur-
ation (1 min in ice). The analytical datum is the difference
between the initial frequency value with the buffer and the value
after the washing when the crystal is in contact with the same buf-
fer, which are both at the same temperature.
A negative control consisted in an amplified fragment from the
T-NOS terminator of the transgenic cassette. Moreover, the same
195 bp fragment has been tested after amplification from DNA ex-
tracted from soybean CRMs (0%, 2% and 5% of Roundup Ready?
soybean powder). The ‘‘blank” samples consisted in all the PCR re-
agents, except the template DNA. Fig. 6 reports the frequency shifts
obtained with the different samples of the soybean processing
chain and with the CRMs. All the samples were analysed in tripli-
cate. Negative samples (CRM 0%, T-NOS fragment and blank) were
clearly distinguishable (DF 6 3 Hz) from the positive ones. The va-
lue of 3 Hz, calculated as three time the blank signal, was consid-
ered as cut-off value for discriminating positive from negative
samples. All the samples containing the P-35S fragment were cor-
rectly identified. The heterogeneity in the positive signals is due to
the different concentrations of the amplified fragments when using
DNA extracted from the different matrices, but it is not affecting
the identification of the GM samples. Good reproducibility (aver-
age CV% = 11%) was achieved for all the measurements.
The same complete protocol, from the extraction of DNA to the
biosensor analysis, has been applied to a different lot of the sam-
ples, analysed in triplicate, coming from the same soy processing
chain. The results were in agreement with the reported ones, con-
firming the reliability of the method.
After the measurements, the surface of the biosensor can be
regenerated by dissociating the hybrid and making the immobi-
lised probe free again for a new hybridisation. This was achieved
by exposing the surface to a treatment with 1 mM HCl (30 s),
allowing up to 20 measurements on the same sensing surface (data
These findings confirm the suitability of this kind of DNA-based
to the monitoring of an entire industrial soybean processing chain
advantages of biosensors with respect to the electrophoretic ap-
the label-free characteristic and the sequence-specific recognition,
can be transferred also to this kind of complex application.
In this work we have underlined the importance of analytical
methods for transgene traceability along the whole food process-
ing chain together with that most of the reported work refers to
raw material, such as flours and seed or to processed food as end
product. What is missing is a systematic approach where target
transgenes sequences are monitored during the whole manufac-
turing process. In this work a homogenous processing line involv-
ing Roundup Ready?soybean, event GTS 40-3-2 (C/UK/94/M3/1,
Monsanto), along a whole soybean industrial processing chain
has been considered. The analytical steps considered, were tailored
to the homogeneous Roundup Ready?soybean food chain, ranged
from the optimisation of the extraction protocol to allow amplifi-
Fig. 6. Sensor responses recorded with samples consisting in PCR amplified DNA
(195 bp) containing the target sequence (P-35S) complementary to the immobilised
probe. The tested DNA was extracted from different processing phases along the
selected soybean processing chain.
P. Bogani et al./Food Chemistry 113 (2009) 658–664
cability of DNA also from highly processed matrices) to the design
of specific primer couples to amplify fragments of various length
(from 118 up to 1006 bp). The suitable fragment lengths for trans-
gene analysis in all the different matrices considered, were 188,
195, 470 bp indicating that short or medium size fragments are
the choice for monitoring of transgenes over the whole food chain.
Differently long fragments were not detected in highly processed
samples. This finding is in agreement with some work reported
on processed isolated samples analyzing endogenous genes (Peano
et al., 2005).
The importance of this study relies on the combined approach
based on the systematic analysis of the same markers in a homog-
enous food chain at high degree of transgenic content, and suitable
biosensor design. In this way, it was possible study the traceability
of the same targets in different materials, undergoing to different
degree of processing, starting from the same source. In this condi-
tion, difficulties eventually encountered in direct analysis of highly
processed food samples, could be controlled and ascribed to partic-
ular treatments occurring during the processing, such as milling,
thermal or chemical treatments. Our findings showed how the
DNA amplificability was related both to the DNA degradation and
to the presence of different inhibitors of Taq polymerase, particu-
larly significant in highly processed food. Specific primer pairs dif-
ferentiated sequences from the whole insert, corresponding to
screening, gene-specific and construct-specific methods (Fig. 2).
On the base of these findings, biosensor analysis to detect an
amplified fragment of the P-35S promoter (195 bp), present in all
the sample along the industrial food chain, was applied and results
compared with the reference technique. The results emerged from
this approach confirmed the suitability of the biosensor for GMO
analysis not only in raw material but also for monitoring the trans-
genic event along all the steps involved in the considered food pro-
Aarts, H. J. M., van Rie, J. P. P. F., & Kok, E. J. (2002). Traceability of genetically
modified organisms. Expert Review of Molecular Diagnostics, 2, 69–77.
Ahmed, F. E. (2002). Detection of genetically modified organisms in foods. Trends in
Biotechnology, 20, 215–223.
Anklam, E., Gadani, F., Heinze, P., Pijnenburg, H., & van den Eede, G. (2002).
Analytical methods for detection and determination of genetically modified
organisms in agricultural crops and plant-derived food products. European Food
Research and Technology, 214, 3–26.
Bauer, T., Weller, P., Hames, W. P., & Hertel, C. (2003). The effect of processing
parameters on DNA degradation in food. European Food Research and Technology,
Bogani, P., Simoni, A., Bettini, P., Mugnai, M., Pellegrini, M. G., & Buiatti, M. (1995).
Genome flux in tomato auto and auxotrophic cell clones cultured in different
auxin/cytokinin equilibira: I DNA multiplicity and methylation levels. Genome,
Bordoni, R., Germini, A., Mezzelani, A., Marchelli, R., & De Bellis, G. (2005). A
microarray platform for parallel detection of five transgenic events in foods: A
combined polymerase chain reaction-ligation detection reaction-universal
array method. Journal of Agricultural and Food Chemistry, 53(4), 912–918.
Corbisier, P., Broothaerts, W., Gioria, S., Schimmel, H., Burns, M., Baoutin, A. A., et al.
(2007). Toward metrological traceability for DNA fragment ratios in GM
quantification. 1. Effect of DNA extraction methods on the quantitative
determination of Bt176 corn by real-time PCR. Journal of Agricultural and Food
Chemistry, 55, 3249–3257.
Di Pinto, A., Forte, V. T., Corsignano Guastadisegni, M., Martino, C., Schena, F. P., &
Tantillo, G. (2007). A comparison of DNA extraction methods for food analysis.
Food Control, 18, 76–80.
Doyle, J., & Doyle, J. (1989). Isolation of plant DNA from fresh tissue. Focus, 12,
Engel, K., Moreano, F., Ehlert, A., & Busch, U. (2006). Quantification of DNA from
genetically modified organisms in composite and processed foods. Trends in
Food Science and Technology, 17, 490–497.
European Commission (2003a). Regulation (EC) No. 1829/2003 of the European
Parliament and of the Council of 22 September 2003 on genetically modified
food and feed. Official Journal of European Communities, L 268, 1–23.
European Commission (2003b). Regulation (EC) No. 1830/2003 of the European
Parliament and of the Council of 22 September 2003 concerning the traceability
and labelling of genetically modified organisms and the traceability of food
and feed products produced from genetically modified organisms and
amending Directive 2001/18/EC. Official Journal of European Communities, L
Gambari, R., & Feriotto, G. (2006). Surface plasmon resonance for detection of
genetically modified organisms in the food supply. Journal of AOAC International,
Germini, A., Rossi, S., Zanetti, A., Corradini, R., Fogher, C., & Marchelli, R. (2005).
Development of a peptide nucleic acid array platform for the detection of
genetically modified organisms in food. Journal of Agricultural and Food
Chemistry, 53(10), 3958–3962.
Giakoumaki, E., Minunni, M., Tombelli, S., Tothill, I. E., Mascini, M., Bogani, P., et al.
(2003). Combination of amplification and post-amplification strategies to
improve optical DNA sensing. Biosensors and Bioelectronics, 19, 337–344.
Kalogianni, D. P., Koraki, T., Christopoulos, T. K., & Ioannou, P. C. (2006).
modified organisms. Biosensors and Bioelectronics, 21, 1069–1076.
Kim, Y., Chae, J., Chang, J., & Kang, S. H. (2005). Microchip capillary gel
electrophoresis using programmed field strength gradients for the ultra-fast
Chromatography A, 1083(1–2), 179–184.
Leimanis, S., Hernandez, M., Fernandez, S., Boyer, F., Burns, M., Bruderer, S., et al.
(2006). A microarray-based detection system for genetically modified (GM)
food ingredients. Plant Molecular Biology, 61, 123–139.
Lüthy, J. (1999). Detection strategies for food authenticity and genetically modified
foods. Food Control, 10, 359–361.
Mannelli, I., Minunni, M., Tombelli, S., & Mascini, M. (2003). Quartz crystal
microbalance (QCM) affinity biosensor for genetically modified organisms
(GMOs) detection. Biosensors and Bioelectronics, 18, 129–140.
Meyer, R., & Jaccaud, E. (1997). Detection of genetically modified soya in processed
food products: Development and validation of a PCR assay for the specific
detection of glyphosphate-tolerant soybeans. In R. Amadò, R. Battaglia (Eds.),
Authenticity and Adulteration of Food – the Analytical Approach (pp. 23–28).
European Food Chemistry IX.
Meyer, R., Chardonnens, P., Hubner, J., & Luthy, Z. (1996). Polymerase chain reaction
(PCR) in the quality and safety assurance of food: Detection of soya in processed
meat products. Lebensm Unters Forsch, 203, 339.
Minunni, M., Tombelli, S., Fonti, J., Spiriti, M. M., Mascini, M., Bogani, P., et al. (2005).
Detection of fragmented genomic DNA by PCR-free piezoelectric sensing using a
denaturation approach. Journal of the American Chemical Society, 127(22), 7966.
Miraglia, M., Berdal, K. G., Brera, C., Corbisier, P., Holst-Jensen, A., Kok, E., et al.
(2004). Detection and traceability of genetically modified organisms in the food
production chain. Food and Chemical Toxicology, 42, 1157–1180.
Moreano, F., Busch, U., & Engel, K. H. (2005). Distortion of genetically modified
organism quantification in processed foods: Influence of particle size
compositions and heat-induced DNA degradation. Journal of Agricultural and
Food Chemistry, 53, 997.
Passamano, M., & Pighini, M. (2006). QCM DNA-sensor for GMOs detection. Sensors
and Actuators, B: Chemistry, B118(1–2), 177–181.
Peano, C., Lesignoli, F., Gulli, M., Corradini, R., Samson, M. C., Marchelli, R., et al.
(2005). Development of a peptide nucleic acid polymerase chain reaction
clamping assay for semiquantitative evaluation of genetically modified
organism content in food. Analytical Biochemistry, 344(2), 174–182.
Rodriguez-Lazaro, D., Lombard, B., Smith, H., Rzezutka, A., D’Agostino, M., Helmuth,
R., et al. (2007). Trends in analytical methodology in food safety and quality:
Monitoring microorganisms and genetically modified organisms. Trends in Food
Science Technology, 18(6), 306–319.
Rossi, S., Lesignoli, F., Germini, A., Faccini, A., Sforza, S., Corradini, R., et al. (2007).
Identification of PCR-amplified genetically modified organisms (GMOs) DNA by
peptide nucleic acid (PNA) probes in anion-exchange chromatographic analysis.
Journal of Agricultural and Food Chemistry, 55(7), 2509–2516.
Xu, J., Miao, H., Wu, H., Huang, W., Tang, R., Qiu, M., et al. (2006). Screening
oligonucleotide microarray. Biosensors and Bioelectronics, 22(1), 71–77.
P. Bogani et al./Food Chemistry 113 (2009) 658–664