The GMO jubilee has come and gone. The predicted financial boom has not materialised, nor the promised benefits to agriculture and health. Dr. Mae-Wan Ho and Prof. Joe Cummins identify some of the fundamental reasons for the failures of the GMO enterprise.
Genetic engineering – artificial tampering with the genetic material or DNA of organisms - was invented around the mid-1970s. DNA from different organisms are spliced together in new combinations, and foreign genes are inserted into genomes of organisms to make "genetically modified organisms", or "GMOs".
In the course of more than 25 years, GMOs have been created from species throughout the living world. Yet the financial boom never happened, and appears unlikely to happen (see "Biotech fever burning, burning out", this series). The vast majority of the uses (and abuses) of GMOs are still under development, with the only real commercial products being GM crops and a handful of pharmaceuticals such as insulin and growth hormone, and enzymes harvested from GM bacteria. This alone should be enough to give warning signs of fundamental hurdles, and there are [1].
Viruses
Vaccines*, biowarfare agents (including those targeting crops and livestock), biological control of insects (baculovirus with scorpion toxin)
Bacteria
Vaccines*, biowarfare agents (including those targeting crops and livestock, also agents that corrode buildings, plastic and metal [2]) biological control of insects (Bacillus thruingiensis), fermentation, enzymes**, pro-biotics, food processing, pharmaceuticals** under contained use, bio-remediation, nitrogen-fixing, etc.
Yeasts
Beer, bread, fermentation, food processing
Plants
Food crops made herbicide tolerant**, insect resistant**, disease resistant, delayed ripening, and for pollen or seed sterility (‘terminator crops’)* (eg, soya*, maize*, oilseed rape*, rice, potato, wheat, barley, oats, tomato, cassava, papaya, coffee)
Non-food crops as for food crops (eg tobacco, horticulture, trees)
Textile crops as for food crops (eg cotton**)
Bio-reactors for antibodies, cytokines, other drugs, vaccines (many food crops used)
Bio-remediation crops for clearing heavy metals
Animals
Models of disease and other experimental purposes (transgenic mice*, rhesus monkey)
Bio-reactors to produce pharmaceuticals, or textile (spider silk) in milk, urine, blood & semen, eggs also targeted (sheep*, goat*, cow*, chicken*)
Xenotransplantation (‘humanised’ pigs for spare organs*)
Food (‘humanised’ milk, improved meat, fish*)
** commercialised
*major effort
In conventional breeding, only individuals from the same species or related species can be mated to produce offspring. The offspring will have genes from both parents, but the genes are just different variants of the same ones coding for the same functions. A GMO, however, contains completely new genes with new functions, as well as new combinations of genes, which will interact with the organism’s own genes in unpredictable ways.
Conventional breeding involves crossing many individuals belonging to different varieties within a species or belonging to closely related species. The result is a population that preserves much of the initial genetic diversity, and selection occurs in successive generations until the desired results are achieved. It is therefore more controllable and predictable.
A transgenic line, in contrast, results from gene insertion events in a single original cell, out of which the entire line is produced. It is genetically very impoverished. In plants, the cells transformed are often kept in tissue culture, a procedure known to generate uncontrollable (somaclonal) variations that frequently change the plant genome. This is one major source of unpredictability.
Another big hurdle is the transgenic ‘technology’ for making GMOs. ‘Technology’ is a misnomer, for the process is uncontrollable unreliable and unpredictable, and has hardly improved since the first GMOs were made.
To begin with, the transgenic construct – the artificial combination of genetic material from different sources that is to be introduced into the organism – tends to be unstable. It often rearranges, duplicates or loses parts on being inserted. Furthermore, insertion into the genome is random and error-prone, and depending on where the insert lands, it will have entirely different and unpredictable effects on the host genes and genome. It can inactivate genes or cause them to over-express, scramble the host genome and destabilise it. The instability of the transgenic construct is such that even after insertion into the genome, it can still become rearranged or lost. The genes inserted can also become inactive (silenced). Instability can arise in later generations of propagation of GM plants. There have not been as many studies in GM animals, but researchers find evidence of instability whenever studies have been carried out [3].
Transgenic instability is something the industry does not want to discuss. There are, up to now, no molecular data supporting the genetic stability of any transgenic line of plants and animals that has been produced for commercial use.
In micro-organisms, such as bacteria and yeast, it is relatively easy to grow and select repeatedly for the stable colonies, but this is much more difficult in other organisms. Consequently, there has been little success in GM animals, except for transgenic mice in the laboratory that are routinely created to serve as disease models or other experimental purposes, often by manipulating embryonic stem cells in culture that can be injected into an early mouse embryo to develop a mouse. And mice breed faster than other mammals.
The failure of GM crops is now widely acknowledged, except by the proponents of biotechnology. GM crops are plagued by lower yields, poor and inconsistent performance in the field, increased use of pesticides and reduced profit for farmers [4]. Transgenic instability certainly contributes to poor agronomic performance and continues to hamper development of GM crops [5].
One of the hardest problems to solve is to target genes to precise positions in the genome, thereby minimising untoward effects as well as instability. The best reported results of targeted gene insertion come from human cell cultures transformed with a viral vector containing a human gene sequence into which the foreign genes are spliced. This still gave random gene insertions that outnumber targeted insertions 10 to 1 [6]. There has been no report of success in targeted gene insertion in any GM plant or animals.
The next hurdle is the safety of GMOs. Despite many assertions to the contrary by the pro-biotech scientific establishment, the safety of GMOs is unproven and open to question. There are reasons to be very cautious about GMOs. New genes and gene products, many from bacteria, viruses and other non-food species are being introduced that we have never eaten before, nor have they been part of our food chain. They may well be toxic or allergenic, and may also adversely affect other organisms in the ecosystem.
To give just one example, the bt-toxins from Bacillus thuringiensis, widely incorporated into both food and non-food crops, has been found to be harmful to mice [7], butterflies [8, 9] and lacewings up the food chain [10]. Bt toxins are active against insects in the Order of Coleoptera (beetles, weevils and styloplids) which contains some 28,600 species, far more than any other Order [11]. The bacterial spores were allergenic for farm workers [12, 13], while the purified protoxin was a "potent systemic and mucosal immunogen in mice" [14]. These toxins should never have been approved for use in GM crops, least of all those used for food.
A neglected source of hazard – in GM crops, though not in gene therapy (see "Genetically modified humans, for what and for whom?" this series) - is DNA from bacteria and their viruses, which have a high frequency of the CpG dinucleotide. These CpG motifs are immunogenic and can cause inflammation [15, 16], septic arthritis [17] and promotion of B cell lymphoma [18]. Yet many genes introduced into GMOs are from bacteria and their viruses.
Dangerous genes, including those from human beings, are incorporated into our food and non-food crops as pharmaceuticals in ‘next generation’ GM crops. These include cytokines, such as interleukin-10, known to disarm the immune system, making people more susceptible to viral infection [19].
Another is the glycoprotein gene gp120 of the AIDS virus HIV-1, incorporated into GM maize as a ‘cheap, edible oral vaccine’. There is a lot of evidence indicating that this gene can interfere with the immune system, as it has homology to the antigen-binding variable regions of the immunoglobulins, and has recombination sites similar to those of the immunoglobulins. Furthermore, these recombination sites are also similar to the recombination sites present in many viruses and bacteria, with which the gp120 can recombine to generate deadly pathogens [20].
We have long called for such ‘pharm’ crops to be banned from the open fields, as they will surely contaminate our food supply with dangerous consequences not only for human beings but also for all other organisms in the food web.
More than 75% of all GM crops are engineered to be tolerant to broad-spectrum herbicides manufactured by the same companies that make most of their profits from the sales of the herbicides. These broad-spectrum herbicides not only kill plants indiscriminately, but are also harmful to practically all species of animal wild-life and to human beings [21, 22]. They have also resulted in the evolution of multi-herbicide resistant volunteers and other weeds.
Transgenic instability is an important safety issue. There is a large literature on gene silencing, in which the transgenes remain in the genome, but are not expressed. More serious, from the safety point of view, is structural instability, the tendency for the transgenic DNA to come loose, to rearrange or become lost in part or in whole in successive generations [23, 24]. This could change the transgenic line in unpredictable ways in terms of health and environmental risks. And it will increase the chance of transgenic DNA being taken up by unrelated species to make new combinations with their genetic material. That’s referred to as horizontal gene transfer and recombination. Transgenic DNA can spread to every species that interact with the GMO, especially bacteria and viruses, in the soil, in the air, in the mouth and gut and the respiratory tracts of animals including human beings [25]. And there is already experimental evidence for horizontal gene transfer of transgenic DNA from GM plants.
New viruses and bacteria that cause diseases could be generated, and antibiotic resistance marker genes could spread to the pathogens. Transgenic DNA may also get into human cells and insert into the human genome; and a large body of evidence from so-called gene therapy experiments have amply demonstrated this does occur [26]. The constructs used in gene therapy are very similar to those used in transgenic plants, and one side-effect of transgenic DNA inserting into the genome during gene therapy is cancer.
Among the GM crops are those engineered for seed or pollen sterility, called ‘terminator crops’ by critics. As originally conceived, their main purpose was to protect corporate patents on the GM crops. Although the development of seed-sterility was largely abandoned after world-wide protest, pollen-sterility is now been promoted by UK and US governments as a way of preventing gene flow by cross-pollination.
Our own investigations [27] revealed that such pollen sterile crops have been grown in field trials in Europe since 1989, and may even have been commercially grown in Canada and the US. Some such crops actually spread male-sterile genes via pollen. Others rely on genes that are known to scramble genomes. Furthermore, there is no way to prevent such genes from spreading horizontally to other species.
The transgenic contamination of the Mexican maize landraces, reported in Nature last November [28], aroused great concern, not the least of the reasons being that Mexico is the centre of origin and diversity for maize. The paper immediately attracted a storm of criticisms from pro-biotech scientists, leading to its retraction by Nature, which is unprecedented for a paper that has not been proved wrong or fraudulent. The critics are not contesting the fact that transgenic contamination has occurred. In fact, the researchers were able to present new data that firmed up their original conclusion that the landraces were contaminated with the CaMV 35S promoter that’s in practically all transgenic maize. What their critics are contesting is the form in which the promoter has entered the landraces’ genomes [29].
The researchers claimed to have found that the promoter in the landraces’ genomes is linked, not to the original transgenes, but to a variety of other DNA sequences. This is as though the promoter has broken off and joined up at random, or as stated in one of the critiques, "fragmenting and promiscuously scattering throughout genomes". This, the critics are strenuously denying.
But that may be just what to expect, in view of the ‘recombination hotspot’ – a site prone to break and rejoin – associated with the CaMV 35S promoter, discovered years after the GM crops containing the promoter has been widely released! We have called for these crops to be banned in 1999 [30], for which we were fiercely attacked. In the ensuing debate [31, 32], we found, in scientific literature dating to 1989, that the promoter is active in species across the entire living world, bacteria, fungi, algae, plants, animals including humans. That too, was unknown to the plant experts who put it into GM crops. Two years later, one group of scientists who discovered the recombination hotspot finally called for the promoter to be phased out in the internal report of the John Innes Centre [33], not on grounds of safety but on grounds that the instability will compromise agronomic performance. In mid-May 2002, a small notice on John Innes Centre’s website announced that this group has been "disbanded in April 2002".
One serious consequence of the invasion of the Mexican landraces by transgenic DNA is the destabilisation of the landraces’ genomes, and the potential for extinction.
The same transgenic instability dogs the production of GM animals. It is a very inefficient process, and the rate of success is not much better than 1%. Furthermore, the transgenes are often lost or silenced in later generations. That is why cloning is attempted in order to bypass reproduction, but equally to no avail so far [34]. We shall highlight some of the problems and risks, using xenotransplantation as a case study.
Xenotransplantation is the transplant of animal organs, tissues or cells into human beings. One of the main problems is immune rejection of the xenograft, especially hyperacute rejection, which takes place within minutes. Hence, we are told, it is necessary to genetically modify pigs to ‘humanise’ them, so as to overcome this rejection, while dealing with the longer term immune reactions with immune suppressive drugs. Biotech giants like Novartis had invested billions hoping to profit from the sale of organs and drugs.
Many scientists have expressed concerns, especially over the possibility of endogenous pig viruses crossing species to infect humans. These endogenous viruses are present in all genomes. Most of them are inactive or dormant. However, some have retained the ability to infect the cells of other species, a phenomenon known as xenotropism. There are many copies of such viruses in the pig genome and it is impossible to breed them out. When we reviewed the topic in 2000, there was already evidence of cross-species viruses from xenografts infecting human subjects [3]. ‘Humanising’ pigs, if anything, will increase the risks of viral infections because the pig viruses will, in effect, be wearing a disguise to escape immune detection.
We found a systematic lack of documentation of the transgenic process and characterisation of the transgenic pigs. There was no quality control. No attempts were made to obtain stable transgenic lines before transplantation experiments were carried out on hundreds of primates, leading to excessive suffering of the animals. And inconclusive studies were being used to push for clinical trials.
In fact, we learned recently that some xenotranplants are routinely done. A friend had an emergency operation before Christmas last year as the result of aneurism of the main aorta of the heart. He received a transplant of a bovine aorta and a porcine aortic valve, and has been recovering
remarkably since, without the need for immunosuppressive drugs, as those tissues are metabolically inert and hence non-immunogenic [35].
The US Food and Drug Administration (FDA) has recently been accused of withholding thousands of documents on side effects and deaths relating to
xenotransplantation. The charge of violating the Freedom of Information Act was brought in January this year by the Campaign for Responsible Transplantation (CRT), a group representing 90 public interest bodies [36].
It claims that Diacrin, a xenotransplantation sponsor, whose clinical testing involved using pig cells to treat Parkinson’s, yielded 232 adverse events in 1997. In 2000, their trials to treat stroke victims had to be halted by the FDA when two patients suffered adverse reactions after having pig cells injected into their brains.
CRT cites 16 patient deaths during or after xenotransplant trials since 1992, and in all cases, death was attributed to "previous medical conditions". Eight patients died in 1997 after having their blood filtered through pig livers at Cedars Sinai Medical Center in Los Angeles.
Can genetic modification ever work? Yes, very rarely. We came across a good paper on the subject, which is also rare.
At Pukyong National University and Hanyang University of South Korea, researchers have been creating transgenic mud loach Misgurnus mizolepis [37]. This is one of the few groups that have taken the trouble to document the stability of their transgenic lines with molecular data. One positive aspect of this experiment is that no extraneous DNA was used, and the gene and promoter both originated from the mud loach itself.
Out of 4 100 eggs injected with mud loach growth hormone gene fused to mud loach b-actin promoter, 7.5% of the hatchlings showed dramatically accelerated growth, with a maximum of 35-fold faster than non-transgenic siblings. Many showed extraordinary gigantism: their body weight and total length - largest 413 gm and 41.5 cm – exceeding even those of 12-yr old normal broodstock (max. 89g and 28 cm).
Of 45 transgenic fish tested by crossing with non-transgenic fish, 30 transmitted the transgene to the next F1 generation at a frequency ranging from 2% to 33%. But only 18 lines retained stable patterns on Southern blot, a technique that cuts the DNA into pieces with restriction enzymes, which are then probed with the transgene.
The growth performance of the F1 was also dramatically accelerated up to 35-fold, although the levels of enhanced growth were variable among transgenic lines. Further instability occurred from F1 to F2 in several transgenic lines.
Based on growth estimates, three transgenic F2 lines showing the greatest growth rates were selected to produce subsequent generations up to F4. Stable inheritance of transgenes up to F4 was confirmed by Southern blot. For Southern blot, the restriction enzyme DraII, that cuts once within the actin promoter was used. This gave multiple bands (3 to 5 or more) in all lines, indicating that multiple copies of the transgenes are present. But there was no segregation of the bands, which means that all were inherited together as a complex single locus.
The time required to attain marketable size (10g) in these transgenic lines was only 30 to 50 days after fertilisation, while non-transgenic fish required at least 6 months. There was also significantly improved feed-conversion efficiency, up to 1.9 fold. The sizes of many transgenic were larger, some reaching maximum size at around 1 year of age. Unlike transgenic fish obtained in other experiments, there appeared to have been no gross abnormalities except for the size increase. However, most transgenic individuals tend to die when they exceeded 400g body weight, although the time required to attain this size varied among transgenic lines.
The growth hormone is attached to a promoter that is expressed at high levels continuously in many tissues, which is hardly normal for the species. That may be why the transgenic giants die prematurely. What other effects it has on the metabolism of the fish and how it will affect animals including humans who eat the fish are among the questions that have to be addressed. But at least, they seem to have managed to produce some stable transgenic fish: three lines stable to F4 out of 4 100 eggs injected.
Genetic engineering has been going on for more than a quarter of a century. If it is not a failure, then it certainly has not delivered the promises. The same failures are documented in gene therapy (see next article in this series).
One of us predicted - in a short paper "Genetic engineering, hope or hoax" published in Third World Resurgence in 1994 - that genetic engineering would fail, because it was based on ideas of genes that were totally at odds with a wealth of scientific findings dating back to the late 1970s.
The genetic engineer sees a static and constant genome, with genes determining the characteristics of organisms in linear causal chains. Scientific findings have revealed however that genes and genomes are dynamic and flexible, and that the environment strongly affects both the structure and function of the genetic material. By the early 1980s, geneticists were inspired to coin the term, "the fluid genome" to contrast with the previous conception.
It took more than ten years of the human genome project to finally explode the myth of genetic determinism, by none other than Craig Venter, the maverick scientist who set up a private company to race the public consortium to the finishing line.
His words sent shock waves around the genomics stock market [38]. "We simply do not have enough genes for this idea of biological determinism to be right," he said, "The wonderful diversity of the human species is not hard-wired in our genetic code. Our environments are critical."
One of the possible explanations for why the gene count is "so discordant with our predictions", says Barry Commoner [39], is that up to 59% of human genes are alternatively spliced [40] (see below). Alternative splicing of a single gene in cells of the inner ear of chicks (and of humans) could give rise to 576 variant proteins [41]. This number is dwarfed by a single gene in the fruit fly, which could potentially generate up to 38,016 variants [42].
But surely, Venter and other scientists in the public human genome consortium must have known that the ‘genetic message’ encoded in the gene is subject to numerous modifications within the cell. These findings were all published before the human genome project began. ‘Interrupted genes’ were discovered in the 1970s, as the first major surprise to hit the genetics community. Genes carrying ‘genetic messages’ are in reality split into numerous coding regions (exons) interrupted by non-coding regions (introns), and these are spliced into different messages in different cells at different times, and translated into different proteins.
And there was a string of other findings. The same gene can be read in different frames to give entirely different proteins. The RNA transcript of the gene can be edited, by base modification or additions, to give a completely different message. Several gene families undergo rearrangements in immune cells in order to generate a vast diversity of antibodies.
In addition, different portions of the genome can undergo gross amplifications, or reductions and deletions. Genes can jump in and out of the genome, and multiply themselves in the process. Genes can enter cells and convert copies in the genome. In short, genes and genomes are just as adaptable and fluid as the rest of the organism [1].
So, was it a conspiracy of silence in the molecular genetics establishment? Barry Commoner, a critic of genetic determinism going back to the 1970s and earlier, says,
"By the mid 1980s, therefore, long before the $3 billion Human Genome Project was funded, and long before genetically modified crops began to appear in our fields, a series of protein-based processes had already intruded on the DNA gene’s exclusive genetic franchise…..By any reasonable measure, these results contradict the central dogma’s cardinal maxim: that a DNA gene exclusively governs the molecular processes that give rise to a particular inherited trait.
"The credibility of the Human Genome Project is not the only casualty of the scientific community’s stubborn resistance to experimental results that contradict the central dogma. Nor is it the most significant casualty. The fact that one gene can give rise to multiple proteins also destroys the theoretical foundation of a multibillion-dollar industry".
Article first published 04/07/02
Got something to say about this page? Comment