Contact   About this website

BELGIAN BIOSAFETY SERVER

[Home Page] [Previous Menu] [Search]

GM Plants containing antibiotic resistance genes
(Last revised: December 4, 2002)

In Europe, several genetically-modified plants (GMPs) have been authorised for placing on the EU market.
Among these, two are authorised for the food chain: the RoundUp®-tolerant soybeans of Monsanto and one line of insect-tolerant maize of Novartis (Ciba-Geigy).

Other GMPs have undergone risk assessment and are currently in the pipe of the EC decision procedure. Several of these GMPs contain antibiotic resistance marker genes in their genome (see Table 2).

The presence of such type of gene must be submitted to risk assessment according to Annex II of directive 90/220/CEE.
Obviously, the primary concern is the protection of the human therapeutic potential of corresponding antibiotics. But other concerns may have interferred with the risk assessment or the authorisation procedure itself, for example the protection of economical interests linked to the use of antibiotics as a feeding industrial practice.
Since end of 1996, the weak acceptance of transgenic plants by the "Public" has been strengthened by the presence of these antibiotics-resistance genes.  This led in 1999 to a practical ban of their use as gene markers in Europe, as far as the concerned plants are intended for feed and food uses

The present chapter aims at providing technical information about these genes in transgenic plants placed -or intended to be placed- on the EU market.

Types of antibiotics-resistance genes in GMPs
Two types of antibiotic resistance marker genes can be discriminated in GMPs.
The first type relates to bacterial genes that have been adapted for expression in plants (both the promoter and 3’ regulatory regions do function in plant cells).
These marker genes are used in the laboratory to select single genetically-modified plant cells at the initial steps of their transformation procedure. The most commonly used marker gene in GMPs is the nptII or aph(3’)II gene encoding the expression of kanamycin/neomycine resistance.
Another type of marker genes can be present in plants as a function of the type of transformation technology by which a engineered DNA is directly transfered to the plant (i.e. particle bombardment, PEG-mediated). These genes, present on the plasmid used for plant transformation, are transferred and integrated into the genome together with the gene of interest.
After integration to the plant genome, these genes often remain under control of their adjacent prokaryotic promoter. As such, they  have been used as selectable markers during the cloning manipulations preceeding protoplast or callus transformation. This is for example the case of the bla_TEM-1gene coding for a ß-lactamase which is present on the classic pBR and pUC cloning vectors.

The "risks" of antibiotic gene resistance markers (AGRM)
The presence of the bla_TEM-1 gene in the chromosomal genome of Novartis’ maize has opened debates in the EU on the risks for human health of a possible transfer of the Maize chromosomal gene to bacteria, for example to the intestinal microflora as a consequence of feed or food use.
Such possible transfer was described as a risky way to jeopardize human antibiotic therapies (Greenpeace - Courvalin et al., La Recherche, May 1998).

How far relevant could this scenario be?

The transfer of genomic AGMR occurs at very low frequencies (10^-13 to 10^-18) under optimized conditions (Nielsen et al.(1997) Theor. Appl. Genet. 95:815-821, Berche P. Med.Ther. (1998) 9,709). This means in the absence of any other competing antibiotic resistance gene in the experiment.

Legal criteria for the appropriate risk assessment of ARGM are outlined in Annex II of directive 90/220, items I.C.2.i)ii) and v), B9 and IVA, B.1. Consequently, the prevalence of the antibiotic resistance gene in nature must also be taken into account for the risk assessement of ARGM in GM-plants. In addition, for placing on the market, the"product" must be assessed and not only the "GMO" as an academic concept.

For example, any food distributed on the market is not sterile. Consequently, antibiotic(s)-resistant natural bacteria are members of the microflora living or surviving on non-sterilized  or non well cooked foods. The foods most likely to carry a large numbers of bacteria are raw vegetables and salads. However, the harvesting, collection, storage of grains in the opened environment is well known, millenia ago, to favor spoilage by contaminating micro-organisms.
In our laboratory, washings of maize or soya grains with sterile water showed that antibiotic-resistant bacteria with several phenotypes of antibiotic resistance (not only the ampicilline, kanamycine/neomycine resistances) can be recovered at high frequencies from both Novartis transgenic and several brands of non-transgenic maize or from Monsanto's transgenic soya and non transgenic soya, independently of the transgenic genotype of the grain considered. (Collard et al., 1997, SBB unpublished).
Corpet (1988) also showed that a sterile diet substantially reduced the number of antibiotic-resistant bacilli excreted by human volunteers.
Therefore, it can be accepted that the likelihood of intestinal bacteria to acquire an antibiotic resistant gene from the natural microflora contaminating any food -and thus from its possible mobile elements- is very much higher than the likelihood of intestinal bacteria to acquire it from the plant genome. The above mentionned probabilities are thus only a very small fraction of a very much higher probability.
The fact that the feed or food has a transgenic origin, implicating or not the insertion of transcriptionnaly-functional antibiotic resistance gene, should not mathematically modify significantly the global probability of gene transfer from natural bacteria.

Discussion and Opinion
Without denying the existence of the genetical phenomenon of recombination/transfer between a food genomic DNA -and consequently any transgenic sequences thereof- and the human or animal microflora, the scenario outlined in the public was builded from an incomplete risk assessment. It is consequently most probably exagerated, given the scale of final probabilities involved (Amman Klaus(1999) La recherche 325, Nov 99, 104:107).
Such a scenario does twist the overal picture of antibiotic misuses by pinpointing out a very small aspect of the whole problematics of antibiotics uses management. Furthermore, such a fashionable scenario is ultimately dangerous for Public Health management because it does promote the picture of a tree masking the forest, at least from the Public viewpoint.
On the other side, it is rather politically and technically easier to ban the use of these antibiotics resistance markers in transgenic plants on basis of the precautionary principle than to regulate the feed market and to control its upstream agricultural practices. This has now pushed GMOs designers to set up new methods and new gene constructs allowing to seggregate the antibiotics resistance gene markers along plant breeding after double transfection of when exploiting alternative markers or the CreII-Lox system bordering the marker (Yoder and Goldsbrough, 1994; SFT, 1997).
Several of the plants currently under EU regulatory investigation still contain "undesired" genes.
These GMPs listed in Table 2 were designed and developped from 5 up to 10 years ago and have undergone a complete risk evaluation for human health and the environment before being commercialized or proposed as such.

The problematics of antibiotics production and uses thereof is a complex and dramatic one. It implies not only the clinical or agricultural spheres but also the environment and the waste management. Two non-limitative examples could be given:
1. It is useful to remind that pBR and pUC core plasmids are universally used in contained laboratories. The transformed bacteria or fungi derived thereof are further classified as genetically-modified micro-organisms (MGMs) in the EU. About 70% -as minimal average- of these MGMs are belonging to the biological class of risk ONE (source: Member States Trisannual reports of directive 90/219/EEC). However, there is no legal duty  to destroy  these bacteria before being discarded as waste (See Directive 90/219/EEC or its revised version 98/81/EC). Altough some Member States did impose such destruction when implementing directive 90/219/EC into national laws, it is not an harmonized situation.
2. The waste management of antibiotics producing plants from non-GM micro-organisms may also be an important problem as far as the introduction of tons of antibiotic-resistant genes-containing wastes into the environnement could be considered as safety and/or environmental problem. However in both cases, the leading paradigm in the media is not to consider these problems as comparable to the spreading of antibiotic resistance markers from genetically-modified plants.

Therefore, a wiser attitude would be to address the safety of GM plants in the frame of the global management of antibiotics uses. Such a management should include the scientific knowledge of the prevalence and the fluxes of the antibiotic resistance genes in the human and non-human ecosystems and under the pressure of human activities. Such a knowledge would certainly contribute to scale at the significant level the risks of GMOs containing antibiotics gene resistance markers.

TABLE 2: GMPs carrying an antibiotic resistance gene and notified or approved for marketing in the European Union.

^{1, 2 and 3} refer to aminoglycoside resistance genes

⁴ refers to a ß-lactam resistance gene

^* with regulatory sequences for expression in plants

[Home Page] [Previous Menu] [Search] Copyright ©2005 IPH-SBB  Contact  Privacy