Memorandum by LGC Limited (formerly the
Laboratory of the Government Chemist)
1. DNA methods based on PCR amplification of
specific DNA markers of genetic modification have been successfully
developed and applied to the identification of a variety of GM
foods including processed products derived from GM soya, GM maize,
GM tomatoes and GM potatoes.
2. Some, more highly processed, foodstuffs may,
however, contain very degraded DNA and/or contain PCR inhibitors,
both of which factors may affect the PCR reaction such that there
may be substantially decreased assay efficiency, or even no reaction
at all. This could result in the reporting of false negative results
for GM foodstuffs if appropriate controls and reference standards
are not employed. To a certain extent these effects may be overcome
by modification of the DNA extraction process and PCR assay design
and conditions, at additional analytical time and cost. However,
if the GM DNA is removed from the product as with some refined
foodstuffs or totally degraded, detection is not possible even
if the foodstuff originates entirely from a GM crop.
3. Lower cost "routine" GM food screening
options are more reliably applied to raw or moderately processed
foodstuffs or ingredients. Typically, batch analysis of these
type of foods is offered by analytical laboratories for approximately
£100/sample (+VAT). Considerable extra costs are incurred
for the analysis of more "complex" samples, generally
determined on a food by food basis.
4. Biotechnology developments, including antibiotic
resistance gene removal, and the development of alternative GM
control elements, coupled with the ongoing drive to market of
new GM crops and foods will necessitate further research and development
of new GM tests to ensure the continued validity of the PCR testing
5. The introduction of a "de minimis"
threshold for GM material in foodstuffs will necessitate the development
of validated quantitative PCR assays; at present these could only
be considered semi quantitative. The design of suitable PCR quantitative
assays for GM foodstuffs which will have the accuracy and precision
to be applied for enforcement of labelling regulations, will post
significant technical challenges, and would require Government
6. In conclusion, the significant challenges
to GM food detection by PCR, posed by processed foodstuffs, are
such that the food labelling statement "does not contain"
in reality should actually state in many cases "cannot be
detected and/or quantified using currently available technology".
What are genetically modified (GM) foods?
7. "GM foods" originate from organisms,
generally plants, which have had their DNA altered, or "foreign"
DNA introduced by the process of genetic engineering usually for
the purpose of:
Enhanced product quality
Increased pest resistance
Introduced agronomic trait
8. The "foreign" DNA introduced during
the genetic engineering process (Figure 1) can act as a "tag"
or marker for genetically modified (GM) plants. Detection of these
"foreign" genetic markers in food can therefore be used
as the basis for development of tests for GM foods.
Plant genetic engineering
Figure 1. Illustration of a method of production
of a genetically modified crop plant by infection with transformed
agrobacterium (adapted from Biotechnology for Crop Improvement
p7 in Recent Advances in Plant and Microbial Biotechnology, AFRC/NCBE).
9. In a typical genetically modified organism
there are usually several DNA markers that may be used for detection:
the gene(s) introduced into the organism
to elicit the required new characteristic; e.g., the gene introduced
into Monsanto's Roundup Ready soya to make the crop resistant
to glyphosate herbicide.
the "start" (promotes)
and "stop" (terminator) DNA sequences, which flank the
introduced gene and act as "molecular switches" to ensure
that the introduced gene functions properly;
chemical resistance markets, e.g.,
antibiotic resistance, which are introduced into the plant to
aid selection and development of the GM plant.
GM crops approved for field release and food use
10. Table 1 illustrates the great variety of
genetically modified plants grown for food production which have
received regulatory approval for field release by various competent
authorities worldwide (predominantly in the US). Those marked
with an asterisk have received EU approval. It is likely that
most of these will also be approved for food useseveral
already have been, as indicated in the table. GM crops depicted
in bold type may be considered as commodity crops which are likely
to be mixed with unmodified crops, and/or partially processed
in the country of production, as is currently the case the GM
soybeans and maize.
11. These foods can be readily identified as
GM in their raw state (see Detection), but most of them
are likely to be processed to varying degrees and mixed as ingredients
in complex foodstuffs which pose more challenging analytical problems
12. As identified in table 1, there are several
crops now approved for food use within the EU. GM soybean and
maize have resulted in the greatest level of public debate. Issues
with respect to detection of these GM commodity crops in foodstuffs
are discussed below.
Methods for detection of GM foods
13. Various methods have been developed for
the detection of genetically modified organisms. Labelling regulations
state that GM protein and/or DNA based methods may be used for
Protein based methods
14. Protein based methods (immunological and
enzymic) detect the gene product or metabolites whose production
is influenced by the gene product. These methods have two significant
limitations so are not routinely employed by most laboratories
for GM food detection. Antibodies have to be raised to the specific
proteins produced as a result of the genetic modification, and
very few relevant antibodies are available. More critically, proteins
degrade on processing, so even if antibodies exist, the methods
are generally only applicable to fresh, raw foodstuffs.
DNA based methods
15. DNA based methods are the most reliable
for the identification of genetic modifications, and have been
most widely used. At a practical level, DNA based testing for
GM food involves first extracting DNA from the food sample. Some
foodstuffs such as tomato puree are more challenging, and it may
be necessary to try several extraction and purification procedures
in order to extract DNA suitable for analysis. Genetic markers
indicative of the genetic modification (as described above) are
then detected using a very specific and sensitive DNA amplification
and detection technique called the polymerase chain reaction (PCR)
(see appendix 1). DNA sequences called primers can be synthesised
in the laboratory that are designed to bind specifically to the
genetically modified DNA if extracted from a food sample. The
PCR reaction copies, or amplifies, the DNA designated by the primers,
which is then typically identified by simple gel analysis and
visual detection of specific bands as illustrated in Figure 2.
Figure 2 GM soya detection by PCR analysis
using primers specific for GM "Roundup Ready" soya.
Lanes 2-4 nonGM soya samples, lanes 5-8 GM flour samples, lane
9 GM soyabean, lane 10 oilseed rape seeds
16. DNA is a remarkably stable molecule and
often survives food manufacturing processes. Tests can therefore
be carried out not only on raw foods but on cooked and processed
17. An ideal PCR screening method is designed
primers selected are specific for
genetic elements in a number of GM crops (see Appendix 2 for a
table of typical genetic markers used in PCR analysis);
the genetic element/marker should
not occur naturally in the plant or in likely contaminating micro-organism;
only a small DNA fragment needs to
be PCR amplified to allow application to processed samples in
which the DNA is likely to be very fragmented.
18. At present, using a limited number of PCR
reactions the majority of commercially released GM crops/foodstuffs
may be identified (see appendix 3). This generic approach to GM
food detection, employing the most commonly used markers for plant
genetic modification, has significant advantages in that a specific
assay does not have to be individually designed for every single
19. However, in order for a GM food testing
regime to remain valid, GM targets selected would have to be continually
updated to include new/changed "second generation" marker
20. It is likely that, in future, plant regulatory
sequences will be used to control expression of introduced DNA.
Similarly pressure on companies to replace/delete genes conferring
antibiotic resistance and diversification of transcription terminators
will reduce the utility of widely employed markers.
21. Experienced analytical laboratories will
be able to continually match the new generation of markers with
new PCR tests for GM food identification. However, it is likely
that funding will need to be provided by Government for the ongoing
R&D necessary to maintain the validity of such PCR based tests
in support of GM food labelling regulations. Furthermore, this
will increasingly necessitate extended knowledge of the genetic
modification, particularly as more organisms are brought into
the food chain, e.g., GM fish, GM micro-organisms. Commercial
sensitivities may also limit access to the required information,
a potential constraint which may have to be overcome through regulatory
Quality controlensuring valid experimental
22. When carrying out tests on genetically modified
organisms it is important to minimise the potential for incorrect
analysis. This can be achieved by carrying out appropriate imitation
control PCR reactions, using DNA primers designed to recognise
whether any plant DNA is present and/or the DNA from the specific
crop, e.g., soya or maize is present in addition to the specific
genetic modification of interest. These additional PCR tests help
guard against false negative results (as may occur through PCR
inhibition, see below) and increase confidence in the analysis
by confirming the presence of DNA from these crops in the extracted
PCR application to a variety of foodstuffs
23. At LGC we have successfully applied PCR
to the analysis of a wide variety of foodstuffs:
|Examples of processed foods from which DNA has been extracted and PCR amplified at LGC
|Soya protein isolates||Tomato soup
|Soya grits||Tomato puree1
||Processed meat products|
||Canned meat products1|
|Maize gluten||Potato salad
|Pasta and noodles||Smash
|Flour samples (maize and soya)||Oxo cubes
|1 Indicates sample types where DNA extraction and PCR are highly variable, and success cannot be guaranteed.
24. As indicated with 1 in the above table many processed
foodstuffs give highly variable results with PCR. The potential
reasons for this are discussed below.
25. DNA detection methods are not applicable if, in the course
of the food production and/or processing, plant DNA is completely
separated or destroyed e.g., refined sugars or oils. However,
again for these types of food product, the extent of processing
may give rise to variability in GM detection. For example, DNA
may be able to be extracted from raw pressed oils but not the
heavily refined oils, and LGC's practical experience with maize
starch has indicated that GM maize could be identified in samples
from some sources but not others, presumably due to differences
PCR DETECTION FOR
26. Although PCR is a potentially powerful detection assay
for the rapid, sensitive and specific identification of GM foods,
food processing can significantly influence the validity of the
27. Various factors contribute to the degradation of DNA
in processed foodstuffs: chemical, physical and enzymatic e.g.
prolonged heat treatment such as autoclaving used
in the canning process, may result in DNA hydrolysis which fragments
the DNA, or modifies the chemistry of the DNA in such a way that
the PCR process may not work. For this reason canned products
can give inconsistent results.
increased chemical modification and hydrolysis
of DNA at low pH (e.g., vinegar). For example, these factors make
tomato puree (and related tomato products such as ketchup) extremely
difficult to work with; results are inconsistent and we, in common
with many other workers, often fail to achieve successful amplification
even with primers designed to detect any plant DNA.
enzymatic degradation of DNA by nucleases may
also occur on prolonged storage of fresh foodstuffs.
28. A recognised problem in using PCR methods with foods
is the presence of PCR inhibitors that reduce the efficiency of
the genetic amplification process. These include many common food
cations e.g., Ca2+, Fe3+;
salts e.g., NaC1, nitrites.
29. Work at LGC, in addition to studies reported in the literature,
indicates that the degree of PCR inhibition is to a great extent
dependent on the food type, e.g., boiled ham shows little or no
inhibition, whereas various kinds of soft cheese completely inhibit
the reaction. This could lead to potential analytical problems
if for example three pies with fillings of ham, soft cheese, soft
cheese and ham were analysed for GM soya content.
30. Without the proper reference standards and PCR controls,
as outlined above in section 2.2, PCR inhibition may easily lead
to false negative results, particularly if the GM food analysis
is being undertaken by laboratories with insufficient experience
with food analysis by PCR to recognise the problem.
31. It is sometimes possible to overcome these inhibitory
effects by extensive dilution of the DNA extract, however, this
may not be an option when the amount of DNA in the sample is limiting.
In these cases, further purification of the DNA, or the addition
to the reaction of PCR enhancers may reduce the level of inhibition.
Knowledge of likely inhibitory components in foodstuffs can inform
the type of analysis undertaken and modifications to the routine
extraction and PCR procedure. Such special analytical modifications
may prevent false negatives and allow a higher level of confidence
in the result obtained, but add time and cost to the analysis.
32. It should be noted that if a foodstuff labelled "does
not contain [GM material]" becomes subject to forensic analysis
as a result of trading standards enforcement of labelling regulations,
it is likely to be subjected to more exhaustive tests than afforded
by routine screening methods.
Implications for validity of GM food analysis by PCR
33. Many foodstuffs which may need to be labelled will be
subject to varying degrees of processing, frequently with the
addition of ingredients which result in a complex food matrix.
Detection of DNA by PCR will therefore probably be influenced
by the relative extent of PCR inhibition and DNA degradation on
a food by food basis.
34. There are very significant challenges in the detection
and quantification of processed foods and food components even
with complete knowledge of the processing history of the sample,
its origins and purity. The possibility of identification of processed
food as containing GM source ingredients should ideally be elucidated
on a case by case basis. Given the variety and complexity of foodstuffs
available in today's markets this would pose a considerable expense
for the industry.
35. If absolute traceability of ingredients can be ensured
during food preparation, then testing for genetic modification
in less processed or raw ingredients which are being added into
the food product would be a more reliable option for determining
whether the final food product contains GM material. This can
be achieved from the ingredients if traceability exists and, in
fact, a wide range of related products could all be assessed from
the smaller range of initial ingredients, saving time and money.
36. In common with other food labelling regulations a de
minimis threshold has been proposed for the labelling of GM
foodstuffs, to potentially allow for the adventitious "contamination"
of a foodstuff or ingredient through the food supply chain. Such
a threshold, which has been considered at 1-3 per cent, in line
with other food legislation, would necessitate quantitative analysis
of [GM] foodstuffs.
37. In consideration of a de minimis threshold for
the presence of DNA [or protein] resulting from genetic modification
the following issues should be carefully considered:
PCR detection limits
38. Detection limits were determined for GM soya using serial
dilutions of purified GM soya DNA with non GM soya DNA to stimulate
different mixtures of GM and non GM soybeans. Detection limits
were determined to increase from 1 per cent to 0.01 per cent as
the extent of PCR amplification was increased above standard practise
(Wurz and Willmund, 1997). At LGC we have also conducted experiments
to determine the limit of detection for GM soya flour in admixture
with non-GM flour. Using standard PCR methodology a 0.1 per cent
level of GM soya flour DNA was detected.
39. Food matrix effects on the relative limits of detection
possible with PCR were demonstrated in experiments carried out
by Greiner and Konietzny (1997). These researchers reported experiments
in which they introduced "foreign" (E. coli)
DNA into a baking process. DNA was extracted at different stages
of processing, and PCR employed to detect the introduced E.
|Experimental results demonstrating that the limit of detection by PCR|
of "foreign DNA" is significantly increased by food processing
|Processing stage||Limit of detection by|
PCR (gene "copies")
|DNA added to rye flour||100-150
40. These results clearly demonstrate that at each successive
stage of the baking process it becomes more difficult to detect
the inserted "foreign DNA", even though the same percentage
of DNA is present as an ingredient at the end of the process as
at the start. It is expected that the same situation would apply
in the case of detection of "foreign" GM soya flour
in bakery products.
41. Therefore, the evidence suggests that the ability to
detect GM residues in food significantly decreases as the level
of food processing increases. If the requirement in the proposed
labelling regulations would be for a 1 per cent cut off for the
particular food product, then in more highly processed foods which
may actually contain a high percentage of for example GM soya,
the level of detection may be below this threshold value, and
the product could potentially be labelled "does not contain
42. There is therefore a requirement to establish the relative
limits of detection (under standardised PCR conditions) for different
types of foods, to inform the proposed legislation.
Quantification of amount of GM material by PCR
43. Future EU legislation is likely to set limits for the
percentage of GM material below which food can be labelled as
not containing GM Derived material. Enforcement of such legislation
will require development of testing procedures which can accurately
measure the amount of GM material in food samples.
44. Precise and accurate quantification of the amount of
GM material in any given sample is analytically very demanding.
However, this would be required for enforcement of labelling regulations,
which would necessitate, the confident determination of the level
of GM material in a food for instance, 0.5 per cent may be below
the threshold value for labelling whereas 1.5 per cent may be
above. Exact PCR quantitation is essentially still at the developmental
stage for food analysis, and is not yet routinely possible.
45. Current quantitative PCR methods are of two major types:
(1) Threshold detection analysis is really an extension
of limits of detection analysis and involves PCR application of
standards containing known amounts of GM material alongside unknown
samples. Instead of looking for presence or absence of bands against
known standards, the relative yields of amplified product from
the standards and unknown samples are compared to enable quantitation.
(2) Quantitative competitive PCR (QCPCR) relies on the
co-amplification of known amounts of a second DNA target or mimic.
In this competitive amplification, both targets have the same
PCR priming sites and compete for the available reagents in the
PCR reaction. The result of the competitive nature of the reaction
is that the relative yield of target and mimic can be related
to the starting ratios of the two. By competing an unknown amount
of one target with a known dilution series of a suitable mimic,
the amount of target GM DNA can be calculated.
46. An advantage of QCPCR is that amplification of mimic
and target are carried out in the same tube providing an internal
PCR control. The disadvantage of QCPCR is that development of
PCR mimics can be technically challenging and new mimics will
have to be generated for each new target.In contrast, quantitation
using threshold detection analysis only requires access to the
PCR primers used for standard detection and standard DNA samples
of known composition.
LGC final report on MAFF project number 2B065 "The Detection
of Genetically Modified Organisms in Food." (1998)
LGC partner reports to EU-project SMT4-CT96-2072 "Development
of Methods to Identify Foods Produced by Means of Genetic Engineering."
Greiner, R and Konietzny, U (1997) Is there a possibility to identify
processed foods as produced through genetic engineering by PCR
technology? pp. 100-102 in Foods Produced by Means of Genetic
Engineering. Second Status Report. Bg VV Hefte, Berlin.
Wurz, A and Willmund, R (1997) Identification of transgenic glyphosate-resistant
soybeans pp. 115-117 in Foods Produced by Means of Genetic Engineering.
Second Status Report. BgVV Hefte, Berlin.