One reason to expect the three genes used in transgenic mustard to be safe is that they aren’t new at all. They have been in the human food-chain for around two decades now.
Since transgenic mustard began to be evaluated for commercial release in India in September 2015, an acrimonious debate has broken out on the pros and cons of the technology. A major flashpoint in this debate has been biosafety. Opponents to the introduction of transgenic mustard insist that the new genes it brings into our food supply are likely to be unsafe, and could trigger unforeseen short- and long-term toxic reactions in the people who eat them. The concern is not for human consumers alone. Environmentalist groups have also argued that transgenic mustard is a big threat to Indian mustard biodiversity; if the genes in transgenic mustard somehow escape into local mustard species, they could create super-weeds by giving these wild varieties a survival advantage. Even worse, they could kill out local species that need to be conserved.
The cryptic silence of India’s Genetic Engineering Approval Committee (GEAC), the body that regulates genetically modified crops in India, hasn’t helped allay these fears. Even though a 2007 Supreme Court ruling asked for transparent sharing of data on the testing of transgenic foods, the GEAC has chosen not to comply. This means all the biosafety data that has been generated so far on transgenic mustard is inaccessible to people. So what are potential future consumers of transgenic mustard to make of the new technology? Can it hurt human health? Will it really be the ruin of Indian mustard varieties? The answer to both these questions is: unlikely.
Old genes
One reason to expect the three genes used in transgenic mustard to be safe is that they aren’t actually new at all. Bar, barnase and barstar, as they are called, have been in the human food-chain for around two decades now. This is because most of the canola grown in the USA, Canada and Australia is modified using these genes. What’s more, India imports canola oil from these countries, meaning Indians too have been consuming food from transgenic canola, a relative of mustard. A lot can be inferred about the impacts of these genes from two decades of experience, and it tells us that these not-so-new kids on the blocks are safe.
To understand why these genes are in mustard, one needs to understand a basic challenge of mustard breeding. The challenge arises because mustard bears perfect flowers, meaning each of its flowers has both male anthers and female pistils, unlike, say maize, which has male and female organs in different flowers. Perfect flowers make it hard for plant breeders to use a common technique of hybridisation – sprinkling pollen from male flowers on female flowers – because the proximity of the anthers to the pistils make it hard to keep the mustard from pollinating itself.
DMH-11, the transgenic hybrid developed by Deepak Pental, a plant geneticist at Delhi University, gets around this problem by jumping the species barrier. It borrows a gene called barnase from the soil bacterium Bacillus amyloliquefacien, which can disrupt the development of pollen when inserted into a crop. Meanwhile, another gene called barstar, also found in Bacillus amyloliquifaciens, has the useful ability to reverse this barnase-induced sterility in crops. So when one parent of a mustard hybrid containing barnase is planted alongside another containing barstar, the pollen from the barstar parent fertilises the eggs in the barnase parent, giving rise to a hybrid. But because the hybrid inherits the barnase and barstar genes from its parents, and because barstar neutralises the actions of barnase, the hybrid is fertile.
The third gene, called bar, has a supporting role to play in DMH-11. Bar makes crops resistant to the herbicide glufosinate and is linked to barnase when the latter is inserted into mustard. This is because pollen-free parents are not just needed for making one generation of hybrids but also for making more pollen-free crops for the next generation. This army of sterile mustard crops is grown by crossing sterile barnase crops with normal mustard crops (that have no barstar gene to reverse sterility). The offspring of these crosses are sterile half the time, because only half of them receive the barnase gene. But because these sterile plants also have the bar gene, they are resistant to glufosinate. This allows a plant breeder to recover all the sterile crops by killing off the fertile ones with a glufosinate spray. The process of hybridisation can now be repeated to produce new seeds.
What we know about bar, barnase and barstar
Humans have been eating proteins encoded by bar, barnase and barstar for a long time now. Canada and the US introduced canola hybrids using these genes in the late 1990s; Australia followed suit in 2001. Each of these countries has tested transgenic canola for toxicity and allergenicity. These tests typically begin by looking at where exactly in the DNA of the plant the new gene is inserted and what proteins they express. Next, they test the composition of the food from these crops and compare it with food from normal canola. Finally, they look at the structure of the expressed proteins (bar encodes for a protein called PAT, barnase encodes for barnase, and barstar for barstar) and compare them with known databases of toxins and allergens that contains thousands of substances. When no kinship was found between known allergens and any of these proteins, transgenic canola was declared safe by these countries.
As opponents of transgenic foods often point out, none of Australia, the US or Canada have tested transgenic canola for toxicity to animals. The regulatory bodies in these countries believe the heuristics described above are enough to throw up any likelihood of harm. Moreover, the refining that canola oil undergoes usually removes all traces of these proteins from the oil, the regulators say.
But this doesn’t mean animal studies do not exist. Independent scientists have tested transgenes in a range of experiments, from acute toxicity studies, in which animals are fed high doses of the modified food, to generational studies, in which smaller doses are fed to several generations of animals.
Viewed together, this body of experiments shows bar, barnase and barstar to be harmless. Add to this the fact that no humans consuming canola across the world have ever experienced ill-effects on a mass scale, which would have happened if these genes had toxic elements in them. This is compelling given that Canada, which mostly grows transgenic canola, is the largest exporter of canola oil in the world.
India chooses caution
Despite the decisions of the Canadian and US regulators to skip animal studies, the Indian regulator isn’t taking it easy. The GEAC has asked Pental’s team for data from both acute and sub-chronic feeding tests in animals, making India the scene of one of the tighter regulatory regimes for transgenic foods (short of European countries that won’t allow transgenics).
Dinesh Kumar Bharadwaj, a pharmacologist at Hyderabad’s National Institute of Nutrition who carried out the animal testing for transgenic mustard, says the tests were designed to be very stringent.
To test for acute toxicity, for example, Bharadwaj’s team fed Swiss albino mice a dose of the transgenic proteins that was about thousand times the equivalent dose humans consume in mustard leaves. The data for average mustard consumption came from the National Institute of Nutrition’s surveys. For comparison, Bharadwaj’s team also fed another group of mice non-transgenic food, while a third group ate normal lab-mouse food. For fourteen days after this dose was given, all mice were watched closely for changes in behaviour, diet and signs of poisoning. At the end of the test period, the animals were killed and tested for toxicity in their enzymes, blood and tissue.
Even though the team concluded that the mouse study threw-up no evidence of harm, some environmentalist groups have claimed that bits of data from the study point to transgenic mustard’s lethal effects. Kavita Kuruganti, a member of the Coalition for a GM-free India, which is against the introduction of transgenic mustard, told The Wire that the high kidney weight in mice fed the barnase protein in Bharadhwaj’s study was worrying. The GEAC, too, has asked Pental’s group what this increased weight could mean. Bharadwaj says, however, that the higher kidney weight is unlikely to have been due to the barnase diet because no other animal in the barnase group suffered the same. Microscopic testing of the animal’s tissues showed that the swelling was due to nephritis, or inflammation of the kidney, although all other organs were normal. This inflammation was likely because the animal itself was sick even before it became a part of the experiment, Bharadwaj hypothesises. “This happened in one animal and it was only the left kidney,” he told The Wire.
After the acute test, Bharadwaj’s team also carried out a three-month study on a species of lab rats called Sprague Dawley. For this test, the rats were fed a fine powder of mustard leaves every day for the duration of the study. Once again, the transgenic group was benchmarked against control rats, after which all the animals were killed and their vital organs, blood and enzymes tested. Again, as the team reported, the tests threw up nothing out of the ordinary.
But Kuruganti and others opponents of transgenic foods have insisted that the tests required by GEAC aren’t enough because they do not include studies conducted across the entire lifetime of an animal. This is unnecessary, says Bharadwaj, because several parameters examined during sub-chronic testing can be early signs of long-term disease. “We can easily detect this by doing enzymatic studies. Tests like liver function and kidney function test are very early biomarkers that give us an indication of (impending) disease.” Moreover, he argues, even animal-rights groups wouldn’t sanction chronic tests when acute tests and bioinformatics throw up nothing.
Kiran Kumar Sharma, who heads the genetic-modification program at the International Crops Research Institute for the Semi Arid and Tropics, adds that it is impossible to be sure if any new food crop is a hundred-percent safe prior to its release. In fact, such demands betray an ignorance of agricultural practices such as plant breeding, he says. “There is a limit to human knowledge for everything. We have been breeding hybrids and varieties for ages. We didn’t know if they are safe or not either.”
It is true that conventional foods carry as many biosafety risks as transgenic crops, as Sharma says. When a new variety of celery, bred to be pest resistant, was introduced in the US in the 1980s, grocery workers and produce handlers across the country began breaking into rashes. An investigation by the Centers of Disease Control revealed that the celery, bred through conventional methods, contained a carcinogen called psoralens at nearly eight times the level of regular celery.
Then there is the story of Lenape, a potato variety bred in the 1960s by crossing a blight-resistant potato with another that yielded high dry matter, so that Lenape was ideal for use by the makers of potato chips. When consumers complained of nausea from eating the potato, the vegetable was analysed and found to contain glycoalkaloids in large amounts. Glycoalkaloids are poisons that protect potatoes against disease but are toxic to humans. Lenape potatoes were subsequently withdrawn from the market even as testing for glycoalkaloids for new potato varieties become mandatory.
When compared with foods such as Lenape and the American celery, which underwent no testing before they were released to farmers, transgenic foods carry very low risk, argue Sharma and others.
The threat to biodiversity
If toxicity to humans is a worry among opponents of transgenics, the fear of harm to local mustard varieties is a bigger one. Environmental groups have argued that India is a centre of diversity for mustard, meaning that the long history of mustard in India has given rise to a variety of wild mustard forms. A transgenic crop that can cross with these local varieties could endanger them, thus hurting biodiversity, they say.
Mexico, where maize was first domesticated almost ten thousand years ago, has resisted the cultivation of GM maize for precisely this reason. Because of its long history of growing maize, Mexico has numerous landraces, or traditional varieties of maize, that show high genetic diversity and are a crucial part of the local cuisine and culture, Kuruganti says. Moreover, the high genetic diversity means they are a valuable source of useful traits to conventional plant breeders too. This is why the mixing of transgenes with local crops on a large scale is a worry.
The extinction of local species due to escaped genes from cultivated crops has happened before, says Norman C. Ellstrand, who studies plant population genetics at the University of Texas in Austin. It happened with Taiwanese wild rice, which disappeared in part due to hybridisation with cultivated rice over many decades. One reason was that the wild rice lost traits that were vital to its survival. The other extreme is the example of common cord grass in Southern England, a hybrid of a local species called smooth cordgrass (Spartina maritima) with an introduced species called small cordgrass (spartina alterniflora). Today, common grass is a weed in southern England that kills off ecosystems in coastal wetlands.
Such risks from cultivated species become magnified when a new plant variety is introduced into a centre of diversity, because of the sheer number of wild crops available for hybridisation. And yet, such instances of extinction or super-weeds remain rare, says Ellstrand. For either possibility to occur, a number of pre-conditions have to be met: the wild and cultivated crop have to be easily crossable, pollinators such as birds and bees must be around, wild varieties must grow close enough to the cultivated ones, and the cultivated crops must transfer traits that are also beneficial in the wild. These are tough conditions to fulfil.
And mustard doesn’t fulfil many of them.
Mustard doesn’t mix very well
First, most taxonomists believe India is a not primary centre of diversity of mustard. Brassica juncea, the variety cultivated in India, is actually a cross between Brassica rapa and Brassica nigra, both of which originated in the Mediterranean region, the primary centre of diversity for these forms. Brassica juncea is thought to have originated in a region where rapa and nigra overlap. Competing theories say juncea either originated in the Middle East or in China. According to K.C. Bansal, who heads India’s National Bureau of Plant Genetic Resources, wild forms of both rapa and nigra are not common across India.
Recent molecular studies, however, show that India and China could be secondary centres of diversity for Brassica juncea – meaning that while this crop did not originate in these countries, it did migrate here a long time ago. One theory holds that Brassica juncea entered India through the northeast, making this region a potential home to diversity, while another holds that it entered from the Middle east through the north-west.
How much of a risk does transgenic mustard present if this is true?
According to most mustard breeders The Wire interviewed, the risk is low because wild species do not hybridise easily with Brassica juncea. “I don’t buy this argument of risk just because some wild species are there; this does not mean that there will be gene flow,” says S.R. Bhat, a plant biologist who created several hybridisation techniques for mustard at the National Research Centre for Plant Biotechnology. “We struggle to make even simple crosses in Brassica. Even in centres of diversity, species isolation is often maintained.” According to Bhat, a majority attempts at crossing wild forms of mustard with juncea have met with failure because of the unwillingness of this species to cross.
K.R. Shivanna, a plant biologist who spent over ten years crossing wild mustard with cultivated mustard to capture the handy traits of the former, told The Wire such crosses are unlikely to occur in nature. Even in the laboratory, such hybrids require complicated technological interventions such as embryo rescue and tissue culture. “I used embryo rescue and thousands of pollinations, and I was able to get fifty hybrids… From that point of view, brassica transgenics are comparatively safe compared to others crops, where we have many other wild species in our country,” Shivanna says. Shivanna, Bhat and other mustard researchers, such as Trilochan Mohapatra, who heads the Indian Agricultural Research Institute, insist that there aren’t many wild forms of mustard growing in areas where mustard is cultivated in India.
The story of Mexico and maize isn’t a good parallel for India and mustard, Bhat adds. The ancestor of maize, the teosinte species, grows wildly across Mexico and crosses very easily with cultivated maize, putting it at a special risk from transgenic contamination. Mustard is a far-cry from this.
Vibha Ahuja, a microbiologist and chief general manager of the Biotech Consortium of India Limited, who helped design regulatory guidelines for genetically modified foods in India, says there are many barriers to gene flow – such as the reproductive compatibility of species and the presence of pollinators. She likens this to a complicated pathway in which a failure at any point can prevent hybridisation, “Suppose both of us have to watch a movie together: first, I have to come to where you are, we have to buy a ticket, we have to take a taxi, go to cinema hall, sit down, and watch the movie. There are five or six steps in this pathway, and all of these steps have to take place for the final result of us watching a movie together.
“Similarly, for gene flow to happen, there are a whole lot of complicated biological steps than need to take place. Even when all of them take place, the issue is: will these crosses survive in nature? Assuming this too can happen, we have to then ask: what are consequences of this to the environment?”
Canada is a good case study of what happens when gene flow does occur. The country has several wild species of Brassica rapa which cross easily with the transgenic Brassica napus. Sure enough, transgenes have escaped and hybridised with wild varieties here. And yet, because the bar, barnase and barstar genes do not confer a significant advantage in the wild, they haven’t stayed in circulation. According to Loren Rieseberg, an ecologist who studies hybridisation and weed evolution at the University of British Columbia, in Vancouver, “Herbicide resistance will only increase the fitness of plants growing on the edges of crop fields that are exposed to herbicides. Thus, the spread will mainly be restricted to these areas and the formation of so-called superweeds is unlikely. This is what we see in Canada. Feral transgenic Brassicas do exist outside of farmer’s field, but they are not troublesome weeds,” he says.
Responding to the criticism of GEAC’s reticence to sharing data, Ramesh Sonti, a member of the GEAC told The Wire that the data on DMH-11 would be shared after it is finalised, something that hasn’t happened yet, because the committee is still seeking information from Pental. In the absence of this data, however, there is little reason to assume transgenic mustard will be unsafe. The fear-mongering about mustard’s impact on biodiversity has little basis, says Bhat. “Let’s decide these things on a scientific basis. I am not saying nobody should evaluate transgenic crops. But it should be scientifically understood,” says Bhat.