AgBioView Special: 'Cisgenic' as a Product Designation - Nature Biotechnology Debate
November 20, 2006, http://www.agbioworld.org
'Cisgenic' as a Product Designation
- David Schubert & David Williams, Nature Biotechnology v.24, p1327 - 1329; November 2006. www.nature.com; Reproduced in AgBioView with the permission of the editor
To the editor: In a correspondence in the July issue (Nat. Biotechnol. 24, 753, 2006) and a publication in EMBO Reports in August1, Henk Schouten, Frans Krens and Evert Jacobsen discuss a new category of genetically engineered (GE) crop plants known as cisgenics. A cisgenic plant is a plant that has been genetically modified using genes and regulatory elements exclusively from plants to which it can be crossed by normal breeding2, 3.
Because of the similarity of the introduced genes to those of the host plant, Schouten et al. argue that cisgenic plants should not be regulated as transgenic plants that contain genes from noncrossable organisms. Instead, they propose that cisgenic plants should be free of any regulation, and food derived from them should not be labeled as GE. Although the authors recognize that the same basic technology is required to make both cisgenic and transgenic plants, they state that because the introduced gene is already present in a related plant, "cisgenesis does not add an extra trait" and is therefore both safe for consumers and poses no environmental hazard. Are these conclusions valid based upon our understanding of plant GE technology or is this simply a semantic argument designed to elude regulatory oversight and food labeling?
The production of cisgenic plants is still susceptible to a major limitation of transgenic plants. It is widely documented that the process of transfecting any gene into a plant leads to large-scale translocations of the plant DNA, scrambling and fragmentation of the transgene, and frequent random insertions of the plasmid DNA4. Schouten et al. suggest that cisgenic plants provide an advantage over breeding because of the problem with linkage drag of undesirable traits with the latter. However, they fail to point out that the insertion of any transgene is very likely to cause a mutation that cannot be removed by breeding for the added trait. Therefore, cisgenic plants will be susceptible to a more deleterious form of linkage drag.
In addition to genetic alterations, a cisgenic plant would likely lack rigorous, tissue-specific expression of the introduced gene, thereby allowing aberrant secondary modifications of proteins, such as glysosylation, that can cause serious immunogenic responses in animals5. The claim that the expression of the introduced gene would be controlled normally by cotransfection of its regulatory elements is extremely dubious because these elements can be located vast distances from the gene, and rarely, if ever, is the complete regulatory repertoire of a gene known (e.g., see ref. 6). In addition, regardless of the presence of regulatory elements, the pattern and level of gene expression can vary greatly depending upon its insertion site7, 8, 9.
Realizing that genetic engineering is a highly mutagenic process, Schouten et al. state in EMBO Reports1 that normal plant breeding is equivalently mutagenic and therefore there should be no concern about the new technology. Essentially the same argument was used previously to support the view that all transgenic plants should be deregulated10. This argument is not, however, a reflection of scientific reality in the context of modern food crops, for many of the types of genetic alterations cited by these authors occurred in the distant past or in the laboratory where the parent plant was mutagenized11.
For example, the genetic alterations cited by Schouten et al. for gene translocations by transposons may have occurred millions of years ago and no complete genes were translocated12. Even if these mutagenic events do occur in nature, they are very rare and are subject to long-term natural selection as well as human selection as a safe food. Never do they occur naturally in large numbers contemporaneously in the same plant as occurs with genetic engineering.
Schouten et al. also argue that because crops derived from mutagenized material have not been regulated, cisgenic plants should not be regulated either. The comparison between cisgenic and mutagenized plants is weak, however, because the mutations produced by trans- or cisgenic technology are fundamentally different from mutations generated by irradiation (mostly deletions) or chemicals (typically point mutations)13. Irradiation and chemical mutagenesis are not major effectors of gene addition or insertional mutagenesis14. Some modern crops have been derived from mutagenized material, but simply stating that there is no evidence for any deleterious effects from these crops in the absence of any epidemiological studies to support this conclusion does not address the safety question. We believe, however, that future food crops produced by mutagenesis should be regulated as transgenics.
In addition to possessing many of the problems of transgenic plants, cisgenic plants have additional problems of their own. If it is indeed possible to create a truly cisgenic plant free of all transgenic DNA sequences, then it would be difficult to characterize such a plant at the genomic level because of the lack of markers and, more importantly, it would be problematic to track the inserted genes once the plants are released. For example, a common characteristic of trans- and cisgenic plants is the inclusion of more than one copy of the gene in tandem. Given the possibility of breeding to homozygosity, and consequential uneven crossover events, it would be important to test for gene amplification or loss in subsequent generations. From a different viewpoint, organic farmers must have a mechanism to monitor for contamination. Schouten et al. may consider this unnecessary, but the consumer marketplace would argue otherwise and effective monitoring is a necessary aspect of any new technology.
In their EMBO Reports article1, Schouten et al. initially present a very narrow definition of cisgenic plants but then later acknowledge that the ideal cannot be achieved because there will always be some GE plasmid sequences in the cisgenic plant. As there is also ambiguity in the definition of a cisgenic plant, such as allowing the introduction of genes from 'related' plants, the semantics open up the possibility for the use of many of the gene manipulations currently employed to make transgenic plants under the guise of cisgenic plants.
For example, as most plants can be infected by cauliflower mosaic virus (CaMV), then can CaMV promoters be used in conjunction with the normal regulatory elements to promote gene expression in cisgenic plants? Although Schouten et al. state that they can "directly improve an existing variety without disturbing the genetic makeup," in the next paragraph, they say that this could be done by "stacking cloned resistance genes" and claim that this would not change the genetic makeup of the plant1. These concerns are heightened by the well-documented misuse of language by the plant biotechnology industry to promote their agenda15. Excellent examples of such use in this journal are defining recombination as a mutagenic event and phenotype as only a plant's agricultural characteristics10, 11, 16.
In conclusion, cisgenic plants suffer not only from the major shortcomings of other GE organisms, but they also have some problems of their own. Although Schouten et al. claim that cisgenic crop plants are just like those produced by classical breeding and therefore both consumer safe and environmentally friendly, their stated aim is to use this new category to avoid regulatory scrutiny and food labeling. When arguing for less regulation, the potential risk relative to any possible benefits should not be forgotten.
There has been little direct benefit for the consumer from current GE crops in the form of either improved nutritional quality or better taste, and a dubious claim of less than one dollar per year per person in cost benefit17. However, release into the environment and the human food supply can have wide-ranging consequences. Although lowering the regulatory hurdle may increase profits in the short term, it could place the long-term potential of improved agriculture through GE in jeopardy. We would prefer to see plant molecular biologists focus their attention on developing more sophisticated methodologies such as a targeted gene knock-in strategy18 or genomics-assisted breeding19 rather than on schemes to evade regulatory mechanisms with products that are still generated by relatively crude transgenic technology.
1. Schouten, H.J., Krens, F.A. & Jacobsen, E. EMBO Rep. 7, 750-753 (2006).
2. Schaart, J.G. Towards consumer-friendly cisgenic strawberries which are less susceptible to Botrytis cinerea. in Plant Research International, vol. 128 (Wageningen University, Wageningen, The Netherlands, 2004).
3. Rommens, C.M. et al. Plant Physiol. 135, 421-431 (2004).
4. Wilson, A., Latham, J. & Steinbrecher, R. Genome scrambling-myth or reality. Transformation-induced mutations in plants (EcoNexus, Brighton, UK, 2004).
5. Prescott, V.E. et al. J. Agric. Food Chem 53, 9023-9030 (2005).
6. de Folter, S. & Angenent, G.C. Trends Plant Sci. 11, 224-231 (2006).
7. Tetko, I.V. et al. PLoS Comput. Biol. 2, e21 (2006).
8. Lessard, P.A., Kulaveerasingam, H., York, G.M., Strong, A. & Sinskey, A.J. Metab. Eng. 4, 67-79 (2002). | Article | PubMed | ChemPort |
9. Brodersen, P. & Voinnet, O. Trends Genet. 22, 268-280 (2006).
10. Bradford, K.J., Van Deynze, A., Gutterson, N., Parrott, W. & Strauss, S.H. Nat. Biotechnol. 23, 439-444 (2005).
11. Schubert, D. Nat. Biotechnol. 23, 785-787 (2005).
12. Lai, J., Li, Y., Messing, J. & Dooner, H.K. Proc. Natl. Acad. Sci. USA 102, 9068-9073 (2005).
13. Cecchini, E., Mulligan, B.J., Covey, S.N. & Milner, J.J. Mutat. Res. 401, 199-206 (1998).
14. Vizir, I.Y., Thorlby, G. & Mullian, B.J. in Gene Isolation, Principles and Practice (eds. G.D. Foster & D. Twell) 215-245 (Wiley, London, 1996).
15. Cook, G.W.D. Genetically Modified Language. The Discourse of Arguments for GE Crops and Food (Routledge, London, 2005).
16. Bradford, K.J., Gutterson, N., Parrott, W., Van Deynze, A. & Strauss, S.H. Nat. Biotechnol. 23, 787-789 (2005). | Article | ChemPort |
17. Fernandez-Cornejo, J. & Caswell, M. Adoption of bioengineered crops (United States Department of Agriculture, Economic Research Service, Washington, 2006).
18. Kumar, S., Allen, G.C. & Thompson, W.F. Trends Plant Sci. 11, 159-161 (2006).
19. Varshney, R.K., Graner, A. & Sorrells, M.E. Trends Plant Sci. 10, 621-630 (2005).
1 Cellular Neurobiology Laboratory, The Salk Institute, 10010 N. Torrey Pines Road, La Jolla, California 92037, USA. schubert.at.salk.edu
2 UCSD Department of Pharmacology, 9500 Gilman Drive # 0912, La Jolla, California 92093-0912, USA.
'Cisgenic' as a Product Designation
L. Val Giddings, Nature Biotechnology, v.24, p 1329, November 2006. www.nature.com; Reproduced in AgBioView with the permission of the editor.
In their recent articles in July issue (Nat. Biotechnol. 24, 753, 2006) and EMBO Reports1, Schouten et al. propose a rationale for a new subcategory of genetically modified (GM) plants. This new category should, in their view, be regulated not according to the rules applied to transgenics, but instead, no differently from the products of conventional breeding. This is a well-intended and clever proposal; but in my opinion, too clever by half. And although the last thing any reasonable person should discourage is anything that would get Europe out of the regulatory/political quagmire of a corner into which it has painted itself, this is a dog that won't hunt.
The crux of their argument is that "deliberate release and market introduction of cisgenic plants is as safe as the release and market introduction of traditionally bred plants." They define a 'cisgenic' plant as "a crop plant...genetically modified with one or more genes...isolated from a crossable donor plant" and construe this to establish that the environmental and health consequences are no different from those associated with the products of traditional breeding. They conclude therefore that 'cisgenic' plants should be exempt from the regulations applied to transgenics. To indict them in their own words, "this is merely a semantic adaptation, rather than a means of controlling risk."
No one familiar with the relevant facts and global experience with transgenic plants over the past two decades would disagree with the conclusion that the plant products generated thus far using recombinant technology are generally very safe indeed and that risks, although in theory perhaps not totally absent (though conspicuously missing to date), are in fact quite low, if not altogether infinitesimal. All of the arguments marshaled by Schouten et al. about the relative risk associated with 'cisgenic' plants seem reasonable, though exceptions can be found or conjured to each. But the logical flaws and factual errors in their rationale are numerous, and fatal. Critically, none of their arguments succeeds in drawing a sustainable distinction between 'cisgenic' and transgenic plants based on any defensible, risk associated criterion. What is proposed, in fact, is a regime no less indefensibly rooted in manufacturing process than other regimes that have been justifiably blistered for the same intellectually bankrupt foundation.
It has been a long time, but I think I remember witnesses testifying before hearings of the US Congress' House of Representatives Committee on Science, Space, and Technology, perhaps in 1983 or 1984 (chaired by Representative Al Gore), in which the proposition was explored as to whether or not the phylogenetic distance between DNA donor and recipient was necessarily an indicator of risk. It was not then, and it is not today. In fact, if there is any correlation between risk and phylogenetic distance, it may be inverse! But about the only rock solid universal truth on which one can rely in this arena is that the risk associated with a novel plant, whether it is genetically modified by in vitro recombinant DNA techniques, classical breeding or any other mechanism, is critically dependent on the encoded trait and the expression patterns of that trait (that is, the phenotype) in the recipient. Other details may be interesting, but are essentially irrelevant to the question of risk.
En route to their well-intended but ill-considered proposal, Schouten et al. misunderstand or mischaracterize a number of facts. Dismissing all regulatory regimes but Canada's as process based, they fairly damn the European regime they hope to ameliorate, while incidentally slandering those in Australia, New Zealand and the United States ("When it gives trouble, they profane even the beautiful and the good."-Goethe, in Faust). They also misleadingly imply the existence of "current international GMO [genetically modified organism] regulations." They offer no citation here, but surely they cannot be referring to the ill-conceived but purely hortatory language of the Biosafety Protocol to the Convention on Biological Diversity? There are excellent guidelines from the Organization for Economic Cooperation and Development (OECD; Geneva) and numerous national bodies, and any number of national regulatory regimes, but nothing that could fairly be described as "international GMO regulations" as invoked. But no matter-these are incidental errors; the take-home message here is clear.
Numerous authoritative bodies on both sides of the pond and around the world have concluded that the risks of transgenic plants are no different from the risks associated with the products of conventional breeding. This finding does indeed have implications for current regulatory regimes: they are all, even the best of them, disproportionate to the level of risk actually posed by the transgenic plants we have seen to date. Those regimes that are easily adapted need to be updated to take this juggernaut truth into account. Regimes that are not easily adapted should be junked and replaced with something (or nothing) that does less damage to reason, common sense and those billions who desperately need agricultural innovations around the world.
1. Schouten, H.J., Krens, F.A. & Jacobsen, E. EMBO Rep. 7, 750-753 (2006).
--Consultant, and Former Vice President for Food and Agriculture, Biotechnology Industry Organization, Washington, DC, USA. lvg.at.prometheusab.com
'Cisgenic' as a Product Designation
- Tjard de Cock Buning, Edith T Lammerts van Bueren, Michel A Haring, Huib C de Vriend & Paul C Struik, Nature Biotechnology, v. 24, 1329 - 1331, November 2006. www.nature.com; Reproduced in AgBioView with the permission of the editor.
Cisgenesis (more commonly called intragenesis1) and transgenesis are two technically similar approaches to create genetic variability through gene-splicing technology. Cisgenic plants are defined as plants that have been genetically modified with one or more genes (including introns and flanking regions such as native promoter and terminator regions in a sense orientation) isolated from a crossable donor plant; that is, of the same or a closely related species2, 3 or isolated from within the existing genome1. Transgenic plants can be described as plants that contain recombined DNA from unrelated organisms1. Thus, the sources of the genes used to genetically modify the plant are different.
This difference has been exploited in two articles by Schouten et al.2, 3 to suggest that cisgenic plants pose fewer environmental risks, evoke less moral objection and should thus warrant mitigated requirements in the biosafety regulations for testing and use of genetically modified organisms (GMOs). Schouten et al. argue that "cisgenic plants are fundamentally different from transgenic plants" because the gene and promoter introduced into cisgenic plants are already present in the species, or in crossable relatives, and therefore do not add an extra trait into the species. Consequently, neither the fitness nor the environmental risk posed by a cisgenic plant changes beyond what may occur through traditional breeding. As in both cisgenic breeding and in breeding through mutagenesis sometimes unwanted mutations and rearrangements occur, they conclude that "cisgenic plants are similar to and as safe as traditionally bred plants" and do not require the type of special regulatory oversight applied to transgenic plants. In addition, they suggest that the origin of the introduced gene is considered an important determinant for "acceptability of genetically modified plants by the public."
Here, we challenge three of the central arguments made by Schouten et al. that downplay certain differences between cisgenesis and traditional breeding and exaggerate certain similarities between cisgenic and transgenic plants to place the former in a positive light: first, their contention that cisgenic plants are fundamentally different from transgenic plants; second, their assertion that cisgenesis is equivalent to traditional breeding; and third, their assumption that cisgenesis will gain wider ethical and societal acceptance than transgenesis. Finally, we propose an alternative approach for the risk assessment of 'trait-enhanced' plants.
To emphasize the 'fundamental difference' between cisgenesis and transgenesis, Schouten et al. build a one-sided image from the arguments often used by environmentalists against transgenesis; for example, "that novel traits [in transgenic plants] can affect the fitness in ways new to the species, potentially leading to increased invasiveness." As in all rhetoric, this is merely half of the truth, stressing a single aspect of transgenesis, which is not necessarily limited to the domain of transgenic plants: in a cisgenic scenario, toxic potatoes or tomatoes can be engineered by inserting extra copies of species-specific alkaloid genes that raise the concentration of solanines (e.g., for pest resistance). In this case, one would certainly regard a consistent risk management regime still relevant for cisgenic plants.
It is also conceivable that a cisgenic plant could be equipped with a desirable trait from a wild relative that has not been present in a crop plant before; thus, a 'novel' trait would be added. Many of the desired 'cisgenes' will be resistance genes, which certainly will affect the invasiveness of a genetically modified plant species. The argument of novelty and associated invasiveness applies to both cis- and transgenic plants and loses therefore its power as a categorical and trustful discriminator between safe cisgenic plants and risky transgenic plants. More specifically, the pitfall is hidden in the proposed extension of the definition of cisgenes: "from a crossable donor plant, i.e. of the same or a closely related species"2, 3. Note the difference with the more precise term 'intragenic plant'1. Juiciness, enhanced taste, color, pest resistance and improved storability or processing quality are novel traits compared with the traditional crop, but the moment that the traits originate from a "closely related species," these potentially 'risky' novel genotypes are defined away as safe cisgenic plants without possessing an 'extra trait'. Extra traits, novel genotype or just novelty are not well defined and tend to serve the needs of the one who hijacks them. Salesmen will market the novelties and the regulatory officer will play them down. To avoid these ambiguities, one should strive for a clear and restricted definition to describe the category of GMOs that bears fewer risks and warrants a lighter risk assessment. We regard this current demarcation definition between cisgenic and transgenic plants inadequately wide and ambiguous.
Apart from distinguishing cisgenesis from transgenesis, Schouten et al. also propose that, at the regulatory level, cisgenic plants should not be treated differently from traditionally bred plants because of their equivalence. They assert that plants derived from cisgenesis are similar to plants derived from mutagenesis and subsequent traditional breeding. The process of genetic modification itself can lead to (unwanted) mutations and rearrangements, similar to those observed in mutation breeding. The long history of plant breeding has shown, however, that there are no adverse effects of mutagenic breeding on the environment or on food. Schouten et al. therefore plead for the same regulatory process for cisgenic plants.
But there are two problems with this line of thinking. First, the authors are stretching the definition of traditional plant breeding beyond its limit; although plants produced via mutagenesis or traditional breeding are not subjected to GMO regulations, in the European Union regulations, mutagenesis is still defined as a type of genetic modification. Second, cisgenesis relies on the random introduction of large genomic fragments using the Agrobacterium tumefaciens T-DNA transfer system. This introduces the cisgene in novel positions in the genome, which it has never occupied before. Experiments with resistance genes have shown that 'cisgenic' plants do not always express the resistance trait, whereas progeny of crosses with the original wild relative do4. Furthermore, it is likely that the insertion will be close to a genic region5, which may affect the behavior of a cisgenic plant in an unpredictable manner. One should be careful not to overstate the extent of genome reorganization in crop plants due to mutational breeding and downplay the effects of random insertion of a cisgene in the plant genome.
This brings us to the final point made by Schouten et al.: will cisgenic plants diminish moral concerns? At least one historical case indicates no. One of the first experiments with cis- and transgenesis in animals, known as the Beltsville's pigs experiment, illustrated the irrelevance of the origin of the inserted genes, the relevance of the unexpected phenotype, the strong reaction of the public and the major setback of R&D investment in genetically modified animals. The pigs received extra copies of the gene coding for porcine somatotropin6 or human somatotropin7 to stimulate muscle growth in pigs. Genes from either source induced similar crippled animals whose bones could not bear the weight. The animals were suffering from all kinds of illnesses related to an infringed hormonal system. In their phenotype, these pigs showed a new 'body building' trait. As a result, in both the United States and Europe, genetic modification of animals was heavily criticized by the public.
Even so, Myskja8 has made the argument that the public should have fewer moral objections against cisgenic plants merely because inserting genes from the same genome reduces risks and because scientists will be better able to predict and control the effects of human intervention in cisgenesis than in transgenesis. Although part of society may have fewer moral problems with cisgenesis than with transgenesis8, another part will continue to perceive cisgenesis as unnatural, artificial manipulation and not in harmony with their world view9. Especially for the latter, who in Europe protected their rights to exert a freedom of choice in food products-a situation that has led to repeated clashes in the World Trade Organization (Geneva) between the United States (and others) and Europe-no perceptions will be changed by the introduction of 'cisgenic' plants.
Even if many citizens consider cisgenic plants different from transgenic plants, this is no guarantee of improved acceptance. Citizens' trust in regulation of genetic engineering might also be an important factor in the acceptance of the technology. Analysis of the results of the most recent Eurobarometer on biotech10 demonstrates that the lack of trust of citizens in regulation of genetic engineering and negative opinions about encouraging genetically modified foods are related. Lack of trust in regulation, which prevails in Greece and-to a lesser extent-in Germany, corresponds with relatively low levels of support for the application of genetic engineering in food. Relatively high levels of trust in regulation in countries like the Netherlands, Belgium and the Czech Republic correspond with relatively high levels of support. Therefore, it is questionable that relaxing regulatory oversight on cisgenic plants will enhance acceptance.
Governmental administrations tend to find a fair balance between safety for the citizens, economic growth and political expediency. Biosafety assessments are costly for all parties. Currently, competent authorities investigate options to classify types of genetic modifications based on risk levels. Low-risk level modifications might then be assessed with a light (and less costly) dossier. However, in our analysis, cisgenesis is not a suitable candidate for a low-risk category.
To sum up, Schouten et al.'s cisgenic proposal is flawed on several levels but is instructive in highlighting several aspects that might enable the definition of a subcategory of gene-spliced plants that could be both scientifically defensible and societally acceptable.
Such products-which, for the sake of argument, we will term 'enhanced trait products' (similar to cisgenic products, but defined on the basis of the phenotype)-would be acknowledged (unlike cisgenic products) to still fall within the various legal definitions of genetic modification. They would be defined as products created by the insertion of original gene fragments (containing introns and flanking regions, such as native promoter and terminator regions in a sense orientation, excluding other regulatory elements) that enhance traits that are already expressed in naturally crossable plants. A full description of the insertion site of the novel cisgene and the effect on the expression of nearby located genes in such products would also be provided.
The existing risk assessment process could then be supplemented with a flexible expert-based, gene-by-gene approach to allow legal shortcuts based on advancing knowledge of risks related to the 'expression' and 'nature' of the gene to acquire simultaneous approval for field experiments and market introduction.
And finally, to acknowledge freedom-of-choice for consumers and create new opportunities for new niche-markets, such products could be promoted under a new label (e.g., for 'enhanced traits').
1. Nielsen, K.M. Nat. Biotechnol. 21, 227-228 (2003).
2. Schouten, H.J., Krens, F.A. & Jacobsen, E. Nat. Biotechnol. 24, 753 (2006).
3. Schouten, H.J., Krens, F.A. & Jacobsen, E. EMBO Rep. 7, 1-4 (2006).
4. Simons, G. et al. Plant Cell 10, 1055-1068 (1998).
5. Alonso, J.M. et al. Science 301, 653-657 (2003).
6. Ebert, K.M., Smith, T.E., Buonoma, F.C., Overstrom, E.W. & Low, M.J. Animal Biotechnol. 1, 145-159 (1990)
7. Pursel, V.G. et al. Vet. Immunol. Immunopathol. 17, 303 (1987).
8. Myskja, B.K. J. Agric. Environ. Ethics 19, 225-238 (2006).
9. Lammerts van Bueren, E.T., Struik, P.C., Tiemens-Hulscher, M. & Jacobsen, E. Crop Sci. 43, 1922-1929 (2003).
1 ATHENA Institute of the Free University, Earth & Life Sciences, De Boelelaan 1085, Amsterdam, 1081 HV, The Netherlands. tjard.de.cock.buning.at.falw.vu.nl
2 Plant Sciences Group, Wageningen University, Droevendaalsesteeg 1, PO Box 386, 6700 AJ Wageningen, The Netherlands.
3 Plant Physiology, University of Amsterdam, Kruislaan 318, 1098 SM Amsterdam, The Netherlands.
4 LIS Consult, Laan van Hoornwijck 1, 2289 DG Rijswijk, The Netherlands.
5 Plant Sciences Group, Crop and Weed Ecology, Wageningen University, PO Box 430, 6700 AK Wageningen, The Netherlands. To the editor:
Nature Biotechnology - 24, 1331 - 1333 (2006) doi:10.1038/nbt1106-1331
Reply to 'Cisgenic' as a product designation - Schouten and colleagues respond:
Schubert and Williams contend that genetic engineering is a highly mutagenic and imprecise process, an argument that they have previously touted in a correspondence to this journal1 relating to an article by Strauss and colleagues2 published in the April 2005 issue. Readers who are interested in the scientific evidence rebutting their old arguments-with respect to (i) comparison of the 'lack of precision' of genetic engineering with 'lack of precision' in conventional breeding (including wide crosses), (ii) genetic changes that do occur naturally due to the inherent dynamic nature of the plant's genome and (iii) the elaborate screening in the laboratory, greenhouse and field for abnormal, unstable or undesirable genetically modified (GM) genotypes during cultivar development-are referred to the responses provided by Strauss and colleagues3, as well as the literature cited in our articles4, 5. Here, we respond only to the new arguments raised by Schubert and Williams.
Schubert and Williams state that expression of a cisgene may change when not all relevant regulatory elements are co-inserted with the coding sequence of that gene. Although we concur with this statement, variation in expression of a gene is a natural phenomenon because of, for example, mutations in the promoter region or allelic variation in regulating genes, such as transcription factors or post-transcriptional factors. In addition, in conventional breeding, expression of a specific gene may vary at the phenotypic level, depending on the genetic background, even if the promoter region of the gene is constant6. In our scheme, however, only cisgenic plants that have the desired gene expression and phenotype would be selected for further cultivar development.
Schubert and Williams state it would be difficult to characterize cisgenic plants at the genomic level and monitor them after release. However, this is not the case. As long as the insertion site differs from the native genomic site, an event-specific PCR reaction can be developed with one primer that anneals to the inserted sequence and another primer that anneals to flanking DNA. The flanking DNA can be sequenced by using commonly available genome walking kits.
In our definition of cisgenesis, we used the wording 'crossable' or 'sexually compatible', to link to the wording of the European Commission's Directive 2001/18/EC; that is, "organisms which can exchange genetic material through traditional breeding methods" 7. We did not use the term 'related plants' as Schubert and Williams mistakenly state. 'Related' is too vague and subject to semantic or taxonomical discussions.
The closing argument of Schubert and Williams about direct benefit for the consumer is outside the scope of our texts and intentions. However, there is no reason that future cisgenic or other GM plants will not have clear benefits for the consumer as this is likely to lead to higher acceptance of such products8.
In his letter, Giddings also raises concerns with our scheme. He states there are "numerous and fatal...logical flaws and factual errors" in our rationale, but fails to provide either clear scientific arguments or citations to support his criticisms. Rather, he refers to testimonies of witnesses of a meeting of more than 20 years ago and to Goethe.
Giddings is mistaken in stating that "the crux" of our argument "is that deliberate release of cisgenic plants into the environment is as safe as the deliberate release of transgenic plants." This is not the crux of our argument at all. Rather, our rationale is that deliberate release of cisgenic plants into the environment is as safe as the deliberate release of conventionally bred plants.
Although we have pointed out some differences between cisgenic and transgenic crops, our aim was not to label all transgenic crops as risky, or justify the sometimes disproportional attention paid to possible risks in the regulation of transgenic crops. We just argue that cisgenic crops are similar to traditionally bred crops because of genes used. Cisgenic crops, by our definition, will have no genes or traits other than already available to the breeder within the conventional germplasm, and this should be reflected in the regulation of these crops.
In the third letter, De Cock Buning et al. have structured their reaction to our arguments along three lines and we respond to them in the same order.
With respect to the fundamental differences between cisgenic and transgenic plants, De Cock Buning et al. suggest that toxic cisgenic potatoes or tomatoes could be engineered and introduced onto the market by inserting extra copies of species specific alkaloid genes. This would be an argument to require extra risk-management for cisgenic plants. However, this is not the case at all, for two reasons.
First, in traditional introgression breeding of potato, elevated levels of alkaloids is a frequently occurring problem. The reason for this is linkage drag from wild relatives. For this reason, traditionally bred potato cultivars are evaluated regarding glycoalkaloids to get permission for introduction onto the market. The maximum level of glycoalkaloids for eating potatoes in The Netherlands is 100 mg per kg fresh weight9. If cisgenic potatoes were treated like traditionally bred potatoes, which is what we propose, then this maximum level applies also to cisgenic potato varieties.
Second, it is peculiar that De Cock Buning et al. suggest that undesired genes, such as genes that cause toxicity to humans or cattle, would be used for cisgenesis. On the contrary, cisgenesis excludes such undesired genes that may hitchhike into crops in traditional breeding because of linkage drag. This is the main advantage of cisgenesis. From the fact that cisgenesis prevents hitchhiking of unwanted genes and unwanted traits, such as genes or alleles for enhancement of alkaloid production or for new, unknown, alkaloids, cisgenesis of existing varieties with a history of use is even safer then traditional introgression breeding.
We stated that "no novel genes, and therefore no novel traits are added to the species or its crossable relatives." However, De Cock Buning et al. give different and changing meanings to the word 'novel' in their text. First, they define novel, as novel to a crop plant. A few sentences further novel refers to a species (intragenic). Because of this inconsistency, it is not clear what they mean exactly. In our definition of cisgenes, we have been clear; that is, the genes that are available to the traditional breeder of the crop involved. Genes outside this gene pool are novel to traditional breeding. The borders of cisgenesis are the borders of traditional breeding. The European Directive 2001/18 on genetically modified organisms (GMOs) uses the same borders. Annex 1B of this directive lists techniques that are exempted from the regulation: "(1) mutagenesis; (2) cell fusion (including protoplast fusion) of plant cells of organisms which can exchange genetic material through traditional breeding methods [authors' emphasis]."7 Traditional breeding is the accepted baseline for the GMO regulations, such as Directive 2001/18. We consistently have used the same demarcation. Although we appreciate that Nielsen10 categorized GMOs according to the source of the gene, we consciously decided not to use the word 'intragenic', as this word refers to the species' borders10, not to the germplasm available to the traditional breeder. Traditional breeders commonly use not only the crop species itself as germplasm, but also crossable plants from related species11, 12.
A very frequent goal of a traditional breeder is introgression of traits from these crossable plants, such as resistance to biotic or abiotic stress, or enhanced quality. These traits and their underlying genes or alleles can be novel to the cultivars; however, they are not novel to the germplasm that is available to the traditional breeder. Also, such alleles may spread around between crossable plants of related species in nature, possibly causing increased fitness. We never stated that cisgenes cannot cause increased fitness; however, cisgenes do not cause increased fitness beyond the fitness enhancement that can be achieved by traditional breeding or beyond the fitness enhancement that can occur in nature by means of gene or allele flow between crossable wild plants.
In relation to the similarities between cisgenic plants and traditionally bred plants, De Cock Buning et al. state that we are stretching definitions, but in fact we are not. We have made clear in the final paragraph and Figure 1 of our EMBO Reports paper that mutagenesis is considered a type of genetic modification in the European regulations, but that it is exempt from these GMO regulations. We have proposed the same for cisgenesis.
Furthermore, De Cock Buning et al. stress the fact that a cisgene will land in a new position in the genome compared with its native position, which may affect its expression and may lead to a mutation in a functional gene. This is not a new point; it merely repeats the discussion already put forward and addressed in our paper in EMBO Reports, in particular page 751 and 752 regarding the insertion of a cisgene at unpredictable sites in the genome of the recipient plant. Cisgenesis is similar to natural mini-translocations, frequently found in molecular genome evolution studies within the species13.
De Cock Buning et al. suggest that we overstate the extent of genome reorganization due to mutation breeding and downplay the effect of random insertion of a cisgene in the plant genome. However, we are convinced that mutation breeding generally leads to at least several mutations in different loci in the genome of a plant, whereas cisgenesis generally leads to mutations in one locus only. Spontaneous mutations in existing cultivars occur also frequently, which may even lead to new registered cultivars, called essentially derived cultivars. This phenomenon is well known, for example in apple cultivars, which are clonally propagated by means of grafting. Numerous spontaneous mutants of apple cultivars are on the market, such as nine mutants of the cultivar Jonagold14, each phenotypically distinguishable and each with its own officially registered cultivar name. This is a result of the dynamic nature of the plant genome. In reality, mutations appear far more frequently in apple, but by far the majority of mutations either does not lead to an improved cultivar or are not recognized at the phenotypic level. Recent research of Soleimani et al.15 also shows the dynamic nature of plant genomes and the small time-scale within which genomic changes can occur in traditional breeding.
De Cock Buning et al. also address the issue of whether cisgenesis will gain wider ethical and societal acceptance than transgenesis. They mention experiments with pigs that received extra copies of a gene for muscle growth, the resulting illness of these pigs and the heavy criticism by the public. This argument, however, is not appropriate for two reasons.
The first reason is that in the experiment of Ebert et al.16 transgenic, not cisgenic, pigs were made because the gene for the porcine growth hormone received a foreign promoter from a virus, with the intention and result of boosting the expression of the hormone gene. Combining a gene with another nonnative promoter or even a foreign promoter is outside our definition of cisgenesis. In our definition, the gene should be flanked by its native promoter and terminator and should contain its native introns. This is not the case at all in the aforementioned paper16. What's more, in the experiments described by Pursel et al.17 not only a foreign promoter was used (a human promoter), but also a foreign gene (that is, a gene from a mouse). The fusion product was microinjected into pigs and sheep. These pigs and sheep were therefore not cisgenic, but transgenic. So, these cases do not illustrate public rejection of cisgenic organisms.
The second reason is that our cisgenesis proposal does not refer to animals, only to plants. This can be very clearly read from the titles of our contributions that contains three times the word 'plants' ("Cisgenic plants warrant less stringent oversight" and "Cisgenic plants are similar to traditionally bred plants").
De Cock Buning et al. doubt whether cisgenic plants are better accepted than transgenic plants, but they do not underpin this with literature on plants. In this context, we feel it is important to mention a survey carried out by Lusk and Sullivan18 in the US. In their survey, 17% of the respondents was willing to consume a GM vegetable with a gene from a microorganism (transgenic plants), but this figure increased to 81% if the vegetable was genetically modified with an extra gene from the same vegetable (cisgenic plants). This research indicates that cisgenic plants are better accepted than transgenic plants.
We agree with Myskja19 that cisgenic plants evoke less moral concerns than transgenic plants. As indicated by De Cock Buning et al., there will still be a group of consumers that rejects cisgenesis, perceiving it as unnatural and not in harmony with their worldview. This is a deontological reason that refers to the process (that is, the unnatural way cisgenic plants have been treated). It does not refer to the result, such as safety of the product. We do not endorse this deontological argument; however, we do respect it.
We also agree with De Cock Buning et al. that people should have their freedom of choice regarding cisgenic plants. In parallel, people that do not want to use plants that received chemical fertilizers or synthetic pesticides, because of the unnaturalness of the process20, also have the freedom to avoid such plants, because of specific labeling of plants that have been 'ecologically' or 'organically' grown. However, this does not require that all other plants and derived products that have been treated with chemical fertilizers should be labeled as 'treated with chemical fertilizers'. Similarly, we do not suggest labeling cisgenic plants and derived products as genetically modified or 'trait enhanced', but possibly labeling of products that are not derived from cisgenic plants.
De Cock Buning et al. deduce from the Eurobarometer survey that there is a correlation between trust of citizens in the GMO regulations and support for GM food. From this correlation, they question whether relaxing regulatory oversight on cisgenic plants will enhance acceptance. This reasoning contains the assumption that strict GMO regulation does encourage trust of the citizens, which we regard as very disputable on the basis of three arguments. First, the mentioned differences in trust in GM food in European countries are not caused by differences in strictness of the GMO regulations, as all citizens of the EU deal with the same EU regulations on GMOs. Second, the very strict EU regulatory framework has not led to increased trust in GM food in Europe, despite the fact that European regulations are stricter than those in countries, such as the United States, where there is more trust in GM food. Finally, other 'unnatural' breeding technologies for food crops, such as induced mutation breeding by means of irradiation or chemicals, have hardly raised mistrust from citizens, in spite of more relaxed regulatory oversight. We conclude that the supposition that relaxing regulatory oversight would decrease acceptance is equivocal.
For gaining trust from citizens, governments should be reliable and trustworthy. They should not maintain bureaucratic, long-lasting, expensive procedures that hamper safe innovations, such as cisgenesis. Government should focus instead on real risks and present dangers that may threaten their citizens.
De Cock Buning et al. state that cisgenesis is not a suitable candidate for a low-risk category, but they fail to provide any scientific evidence that cisgenic plants are less safe than traditionally bred plants or plants from mutation breeding. However, in our papers we extensively provided evidence for the similarity of cisgenic plants to traditionally bred plants or plants from mutagenesis, and their similar safety levels. This argument can also be taken a step further: as cisgenesis prevents linkage drag of unwanted genes (e.g., genes or alleles encoding toxins to humans), cisgenic improvement of varieties that already have a history of use could be even safer than traditional introgression breeding with linkage drag.
In summary, we feel that we rebutted all arguments from Schubert and Williams, Giddings and DeCock Buning et al. This has increased our confidence that we have presented sound arguments to consider cisgenic plants similar to traditionally bred crops. Therefore, we propose that the EU Directive 2001/18/EC and other GMO regulations be changed so that cisgenic crops are exempt, as has been done in the mentioned directive for products produced by mutagenesis and by protoplast fusion between crossable plants.
1. Schubert, D. et al. Nat. Biotechnol. 23, 785-787 (2005).
2. Bradford, K.J., Van Deynze, A., Gutterson, N., Parrott, W. & Strauss. S.H. Nat. Biotechnol. 23, 439-444 (2005).
3. Bradford, K.J., Gutterson, N., Parrott, W., Van Deynze, A. & Strauss. S.H. Nat. Biotechnol. 23, 787-789 (2005).
4. Schouten, H.J., Krens, F.A. & Jacobsen, E. Nat. Biotechnol. 24, 753 (2006).
5. Schouten, H.J., Krens, F.A. & Jacobsen, E. EMBO Rep. 7, 750-753 (2006).
6. Gardiner, S.E. et al. Theor. Appl. Genet. 93, 485-493 (1996).
7. European Parliament. Off. J. Eur. Comm. L 106, 1-38 (2001). 8.
9. Anonymous. Field Crops. Recommended List of Varieties 2005 (GSB, Naarden, The Netherlands, 2004).
10. Nielsen, K.M. Nat. Biotechnol. 21, 227-228 (2003).
11. Ross, H. Potato Breeding: Problems and Perspectives (Parey, Berlin, 1986).
12. Chahal, G.S., Gosal, S.S. Principles and Procedures of Plant Breeding: Biotechnological and Conventional Approaches (Alpha Science International, Pangbourne, 2002).
13. Schmidt, R., Plant Mol. Biol. 48, 21-37 (2002).
14. Anonymous. Rassenlijst voor Fruitgewassen. (Nederlandse Fruittelers Organisatie, The Hague, 1999).
15. Soleimani, V.D., Baum, B.R. & Johnson, D.A. Genome 49, 389-396 (2006).
16. Ebert, K.M., Smith, T.E., Buonoma, F.C., Overstrom, E.W. & Low, M.J. Animal Biotechnol. 1, 145-159 (1990).
17. Pursel, V.G. et al. Vet. Immonol. Immunopathol. 17, 303-312 (1987).
18. Lusk, J.L. & Sullivan, P. Food Technol. 56, 32-37 (2002). 19. Myskja, B.K. J. Agric. Environ. Ethics 19, 225-238 (2006).
20. Lammerts van Bueren, E.T., Struik, P.C., Tiemens-Hulscher, M. & Jacobsen, E. Crop Sci. 43, 1922-1929 (2003).