The Killing Fields: Terminator Crops at Large

By Drs. Mae-Wan Ho, Joe Cummins and Jeremy Bartlett

The Governments of the United States and the United Kingdom are simultaneously pushing terminator technology in ‘public consultations’. But terminator crops have already been released since the beginning of the 1990s.


Last December, one of us (MWH) was acting as expert witness in defence of citizens who have taken civil disobedience action against GM crops. Among the crops in question were GM oilseed rape varieties used to produce F1 hybrids, as described in the application for release from AgrEvo UK (now Aventis).[1] At the time, we were just preparing our submission to the consultation document, “Guidance on Best Practice in the Design of GM Crops” put out by the UK Government’s Advisory Committee for Release to the Environment (ACRE). One of the main ‘enabling technologies’ for ‘best practice’ – to prevent gene flow – is precisely the seed/pollen sterility system mentioned in AgrEvo’s application.

It soon dawned on us that the GM oilseed rape lines undergoing field trials in the UK are engineered using ‘terminator technology’ – so named by critics because it can render harvested seeds sterile – for no other reason than to enforce corporate patents on GM seeds. Not only that, according to AgrEvo’s application, similar crops produced by the company Plant Genetic Systems (PGS), a subsidiary of AgrEvo, have been undergoing field-trials in France and Belgium since the beginning of 1990, and subsequently on larger scales, also in Sweden and Canada before coming to the UK.

A search on the US database on field trials[2] revealed that similar male sterile lines engineered with the ‘terminator-gene’, barnase (see below), have been tested at least as early as 1992. Since then, there have been 132 field trials, the vast majority of them done without risk assessment, as the first environmental assessment came up with ‘FONSI’ – Finding of No Significant Impact.[3] Crops modified for male sterility include rapeseed, corn, tobacco, cotton, Brassica oleracea, potato, poplar, Cichorium intybus, petunia and lettuce.

Separately, the other genetic component in terminator crops, the site-specific recombinase (see below), has also been engineered into corn and papaya, and there have been 14 field trials between 1994 and 1998. No environmental impact assessment had been carried out at all, as it was deemed unnecessary.

There are more than 150 US patents listing barnase or site-specific recombination or both,[4] the oldest, on site-specific recombinase, going back to 1987.[5]

The first terminator patents that came to the attention of the public were those jointly owned by US Department of Agriculture and Delta and Pine Land Company, which Monsanto had intended to acquire. The novelty in those patents is the proposal to combine the terminator-gene system with the site-specific recombinase system, giving the company complete control over the hybrids as well as proprietary chemicals that control gene expression.

As a result of universal condemnation and rejection by farmers and non-governmental organisations world wide, Monsanto had announced it will not commercialise terminator crops, to everyone’s relief. Research and development, however, have continued unabated, and the technology kept surfacing in different forms.[6] But on the whole, everyone has been duped into thinking that such crops only exist in theory, when they have been out there in one form or another for more than 10 years.

In the UK, ACRE’s consultation can only be seen as an exercise in smoothing the path for commercial development of a technology condemned as contrary to basic human rights, because it prevents farmers from saving, replanting and exchanging seeds, practices going back thousands of years that are essential to food security.

It is no coincidence that simultaneous consultation is going on in the United States on the USDA-Delta and Pine terminator patents. The USDA is indeed considering commercial development of the technology, and also recommends it for preventing GM gene flow. Surely, to require containment of GM genes is to admit that they are unsafe, which is an argument for stopping GM crop development altogether. It is not an excuse for validating a morally bankrupt technology.

What the regulators and the public are not yet aware of is that the technology introduces serious hazards over and above those of GM crops in general.[7] The terminator-gene barnase is a universal poison that breaks down RNA, an intermediate in the expression of all genes. The recombinase, in theory, breaks and rejoins DNA at specific sites, but is far from accurate, so it has the potential to break and rejoin DNA inappropriately, thereby scrambling the genome in unpredictable, lethal ways.

After a report in the Scottish press, a spokesperson from the UK Department of the Environment, Transport and the Regions (DETR) denied that the enzyme barnase was in the crops undergoing field trials. The DETR spokesman was reported to have said that it was the barnase gene and not the enzyme which was present in “a few oil seed rape crops currently being trialled.” and that “where the enzyme would be poisonous, the gene was not harmful.”[8] Obviously, he did not know that the barnase gene had to be expressed to make the barnase enzyme in order to have male sterility. Furthermore, the barnase gene could spread, either by crossing with related species, or by the genetically modified DNA being taken up and integrated into the genome of unrelated species, and it may become expressed in other cells and tissues, with potentially fatal consequences.

On seeing our press releases,[9] Dr. Ian Woiwod of Rothampstead, a scientist involved in overseeing the UK field trials, indicated that he had no knowledge of such crops in the field trials.[10] Indeed, in a correspondence describing the trials published in Nature in 1999,[11] there was no mention of the male sterile spring and winter oilseed rape. Have our regulators been kept in the dark? During a workshop at the first meeting of the Intergovernmental Committee on the Cartagena Protocol on Biosafety held in Montpellier last December,[12] the UK Government delegate from the DETR actually thanked Mae-wan Ho for providing the information on terminator crops.


There are two key components to terminator technology, which is being widely used, not only in plants but in animals as well, as revealed by the 150 plus patents filed in the US alone (see above). The first component is ‘site-specific recombination’, carried out by a recombinase, an enzyme that recognises specific ‘sites’, or short DNA sequences, labelled ‘s’ in the diagram below. Any stretch of DNA sequence flanked by two such sites will be spliced out by the recombinase.

…s-any DNA sequence-s…

The other key element is literally the ‘terminator’. It is barnase, an enzyme that breaks down RNA. RNA is an intermediate in the expression of all genes, and that is why barnase is lethal to all cells in which it is expressed, unless its specific inhibitor, barstar, is also present. Both barnase and barstar are produced by a soil bacterium, Bacillus amyloliquefaciens. Inside the bacterial cell, barstar binds to barnase in a one-to-one complex, disarming the latter so it can do no harm. However, when barnase is secreted outside, it is no longer bound to barstar and is thus harmful to other cells.

To engineer pollen sterility, the barnase gene is placed under the control of a promoter that allows the gene to be expressed only during anther development, i.e., in the male part of the flower. The barnase with its anther-specific promoter is stitched next to the transgene of interest, say, a gene coding for herbicide tolerance, also with its own promoter. Theoretically, there will be no fertile pollen from this transgenic crop. In the case of crops that are normally self-fertilised, there will be no seeds set. In out-crossing plants, the only fertile seeds set will be those fertilised by non-genetically modified varieties nearby, which will not be herbicide tolerant; so farmers who want the herbicide tolerant trait will have to buy fresh seeds from the company every season. The problem is that such a male-sterile line by itself cannot be propagated; it does not breed true.

To propagate the line, the company makes use of site-specific recombination. For example, the promoter of the barnase could normally be blocked by a sequence flanked by sites recognised by a recombinase

.. anther-specific promoter-s-blocking sequence-s-barnase gene..

The recombinase can be engineered into the same genetically modified line with the barnase gene for male sterility, or it could be introduced by crossing the genetically modified line containing barnase with another that contains the recombinase to generate a hybrid. The recombinase is placed under the control of a promoter that responds to an external chemical, say, the antibiotic tetracycline.

..tet-specific promoter-recombinase gene…

When tetracycline is applied, the recombinase is expressed, and splices out the blocking sequence in the barnase promoter, so barnase is expressed. By treating harvested seed with tetracycline before they are sold to the farmer, the company can ensure that the plants grown from the seeds will be pollen sterile.

If female-sterility is required, the barnase gene could be placed under the control of a promoter that works only during ovule development, i.e., in the female part of the flower, and the rest is similar.

Alternatively, the recombinase may be engineered into a genetically modified line with the gene coding for barstar, which, when crossed with the male sterile genetically modified line containing barnase, will produce a hybrid. The hybrid, treated with tetracycline, will produce plants that will still set seed, at least in theory, because the barstar inactivates the barnase. However, if the farmer tries to re-sow the harvested seeds, he or she will find that only about half (7/16)[13] of the seeds will have the same characteristics as those bought from the company, and about one fifth (3/16) of the seeds may be completely sterile. But it could be considerably worse.

In AgrEvo’s application for field trials, only two lines are mentioned. These are:

a.. the ‘male sterile oilseed rape line’ engineered with barnase under the anther-specific promoter and a gene for phosphinothricin (glufosinate herbicide) tolerance; and
b.. the ‘restorer oilseed rape line’ engineered with barstar, also under the anther-specific promoter plus the same gene for glufosinate tolerance.
No detailed genetic map or other molecular genetic data were supplied with the document, as it was clearly intended for the public register. Companies are currently allowed to conceal molecular genetic data under ‘commercially sensitive information’, and most of them do so. Does either of these lines contain the site-specific recombinase? Does the barnase gene exist in a blocked form in another line in which male fertility can in principle be indefinitely maintained?

If barnase is not blocked, then the ‘male-sterile’ line cannot possibly be a true-breeding, uniform line, as it must be fertilised by pollen originating from non-male sterile oilseed rape. A male-sterile line can only be heterozygous for barnase and herbicide tolerance.


The system is ineffective for preventing gene flow for the following reasons:

a. All gene control systems are known to be ‘leaky’ in the sense of not being 100% effective, and the proposed system is no exception, particularly as so many elements have to be engineered perfectly, which is beyond current capability. As a result, some fertile pollen/seeds are likely to be produced.

b. Pollen sterile genetically modified (GM) plants can still be fertilised by non-GM pollen, just as GM pollen from ovule-sterile plants can cross with non-GM plants, thus enabling gene escape.

c. This system does not at all prevent horizontal gene transfer, a process whereby the GM DNA is taken up directly into cells of unrelated species and incorporated into the cell’s genome. If anything, horizontal gene transfer may be enhanced due to the increased structural instability of the complicated constructs involved. Transfer to bacteria and viruses in all environments can be envisaged. Plant residues, dust and pollen may all contribute. Transfer to insect pollinators or feeders could take place; and these may also become vectors for further horizontal gene transfer.

Significant hazards are introduced by this system, over and above those due to GM crops in general. First, barnase is a potent RNAse that breaks down RNA indiscriminately, and is known to be harmful, if not lethal, to all cells, animals and humans included. When perfused into rat kidneys, barnase causes kidney damage.[14] IT SHOULD NOT BE PERMITTED IN ANY GM CROP, LET ALONE GM CROP INTENDED FOR ANIMAL FEED OR HUMAN FOOD.

Second, the ‘site-specific’ recombinases are known not to be 100% specific. There is already evidence suggesting that unintended rearrangements and deletions of genomic sequences have resulted from the use of such recombinases. In other words, the recombinases have the potential to scramble genomes in unpredictable, harmful ways (see Note 7). This has now been demonstrated for the first time, basically because some researchers have finally cared to look for it.

The recombinase Cre is part of the ‘site-specific recombination’ Cre/lox system originally isolated from the bacteriophage (bacterial virus) P1. Cre catalyses recombination between two lox sites, splicing out any stretch of DNA in between.

The system is not only used in plants, but also extensively exploited in transgenic mice. Studies in the test-tube have shown that Cre recombinase can catalyze recombination between DNA sequences found naturally in yeast and mammalian genomes. These ‘illegitimate sites’ often bear little similarity to the lox element. However, there have been no reports on such illegitimate recognition in the animals or plants themselves. And there have even been pilot studies using the Cre/lox system in human gene therapy.

In a study just published,[15] researchers in the United States showed that high levels of Cre expression in the spermatids of heterozygous transgenic mice leads to 100% sterility in the males, despite the absence of any lox sites. Heterozygous mice carry only one copy of the Cre recombinase gene.

The sterility is caused directly by the recombinase enzyme scrambling the genome, essentially by breaking and rejoining DNA at inappropriate sites on the same or different chromosomes. The researchers have pinpointed the genome-scrambling event to the time at which the two ‘daughter’ spermatids and their paired chromosomes have just separated from each other, but are still joined by a ‘cytoplasmic bridge’. This is enough to allow the enzyme to pass from the spermatid containing the recombinase gene to the other that does not, thereby to scramble up the chromosomes of both the transgenic and non-transgenic spermatid. The result is 100% sterility. Embryos fertilized by these sperms arrest predominantly at the 2-cell stage, and do not go beyond the 4-cell stage.

The researchers warn: “These results indicate that Cre can catalyze illegitimate recombination having overt pathological consequences in animals.” A similar recombination system is found in animals containing the RAG recombinases. Illegitimate recombinations in somatic cells are linked to human leukemias.

The greatest danger of terminator crops stems from the spread of the genes and constructs, not only to related species by out-crossing but by horizontal transfer to unrelated species. The increased complication of the genetically modified constructs involved will only increase structural instability and hence the tendency to horizontal gene transfer and recombination. Transfer of both the terminator gene barnase and the recombinase will have drastic, potentially fatal effects on agriculture and on biodiversity.

It is high time to stop these killing crops once and for all.

End Notes:

[1] Application for field trials from AgrEvo (now Aventis) March 1999 “Part B: Information about the release application to be included on the public register”.


[3] “Environmental Assessment and Finding of No Significant Impact” Prepared by Biotechnology Permits, Biotechnology, Biologics, and Environmental Protection Animal and Plant Health Inspection Service, U.S. Department of Agriculture, Permit Number 92-017-01: rapeseed; male sterility; restorer gene


[5] US04673640 06/16/1987 Regulated protein production using site-specific recombination.

[6] See “Terminator in different guises” ISIS News #3, December 1999

[7] ISIS has warned of this more than once. See “Why patents on life-forms and living processes should be rejected from TRIPS – Scientific briefing on TRIPS Article 27.3(b)” by Mae-Wan Ho And Terje Traavik, TWN and ISIS Report, 1999; “Terminator in different guises” by Mae-Wan Ho, ISIS News #3, December 1999; “Terminator gene product alert” by Joe Cummins, ISIS News#6, September 2000

[8] “GM ‘poison’ allegation denied”

[9] “Terminator alert: UK GM field trials contain ‘terminator’ crops” ISIS Press Release 6.12.2000; “Terminator recombinase does scramble genomes” ISIS Press Release 8.12.2000

[10] Personal communication by e-mail to ISIS from Dr. Ian Woiwod, 8.12.2000.

[11] Firbank, L.G. et al, (1999). Nature 399, 727-8.

[12] Critical Issues on Biosafety, Third World Network Workshop during the First Meeting of the Intergovernmental Committee on the Cartagena Protocol on Biosafety, 11 December 2000, Le Corum, Montpellier.

[13] It deviates from the usual 9/16 Mendelian ratio the inheritance of two genes, barnase and barstar, because there is a one-to-one complex between barnase and barstar; hence plants in which there are two copies of barnase and only one copy of barstar will be expected to be partially sterile.

[14] Ilinskaya,O and Vamvakas,S (1997). Nephrotic effect of bacterial ribonucleases in the isolated and perfused rat kidney. Toxicology 120, 55-63

[15] Schmidt, E.E., Taylor, D.S., Prigge, J.R., Barnett, S. and Capecchi, M.R. (2000). Illegitimate Cre-dependent chromosome rearrangements in transgenic mouse spermatids. PNAS 97, 13702-13707.

articles post