Part 2 of an interview with Prof Michael Antoniou by Stacy Malkan of US Right to Know

In the US, more than 90% of corn plants are genetically modified to withstand weed-killing chemicals like glyphosate, to make their own insecticides, or both. For soybeans, over 94% are genetically modified, almost all to tolerate glyphosate, with newer varieties also resistant to the herbicides dicamba and 2,4-D. Are these foods safe to consume? 

In an interview with US Right to Know, a leading molecular geneticist warns that regulators are ignoring the risks of these “stacked-trait” GMOs – which have never been properly tested for safety, even though animal studies have shown signs of liver, kidney, immune and digestive system damage from earlier single-trait GMOs. His research has also found liver and kidney damage in rats exposed to mixtures of herbicides used on GMOs (glyphosate, 2,4-D, dicamba) — at levels regulators say are safe. 

In part two of our interview posted here, Professor Michael Antoniou, head of the Gene Expression and Therapy Group at King’s College London, explains the findings of his team’s extensive research on glyphosate and glyphosate-based herbicide formulations, which he said shows that regulators are “clearly wrong” that current exposure levels are safe for our health. 

Professor Antoniou also explains how his research challenges the claim that GMOs are “substantially equivalent” to non-GMOs — the foundation of US regulations for GMOs — and how regulators are ignoring modern scientific tests and methods that can help us better understand the risks to our health from GMOs and pesticides.

Major changes between GMO and non-GMO

Stacy Malkan: As you mentioned (in part 1), the US is not conducting safety assessments to figure out health risks of GMOs. US law dates back to the 1990s during the deregulatory program of the first Bush Administration. It is based on the idea that GM foods are substantially equivalent to regular foods. But you say your research disproves this. Can you explain?

Prof Michael Antoniou: Yes, I had an interesting experience when we published our work demonstrating that the major glyphosate-tolerant GMO corn variety, Monsanto’s NK603, was not substantially equivalent to a non-GM equivalent corn variety. 

First, let me define substantial equivalence. Early on, as part of the health risk assessment of GMOs, industry and regulators came up with this idea of substantial equivalence. Basically that means you do a fairly gross superficial compositional analysis of the GMO and compare it to an equivalent non-GMO variety. If the levels of, say, total protein, total fat, total carbohydrate, total mineral content are roughly the same, you arrive at the conclusion that the GMO is substantially equivalent to the non-GMO. And if it’s substantially equivalent, then why should you be worried?

This kind of compositional analysis would be equivalent to what you might read on the side of your breakfast cereal box, where you see protein X percent, fat X percent, and so on – very superficial. But the devil is in the details, it’s not how much protein is there, but what kinds of proteins; not how much fat, but the spectrum of fats; not total carbohydrate and minerals, but the spectrum of carbohydrates and minerals and other nutrients. 

We conducted one of the first-of-its-kind studies – what is known as a molecular profiling analysis of GMO glyphosate-tolerant GMO corn – and we compared it to the nearest genetically equivalent GMO corn we can lay our hands on. Just to make sure that things were as equivalent as possible, the GMO and the non-GMO were grown at the same time and in the same location. We conducted what’s known as proteomics analysis, which is the actual profile of the proteins – not just total protein, but the profile of the many hundreds of different proteins that were detectable; and also what’s known as metabolomics analysis, which is a small molecule biochemical analysis. We are able to analyze 600 to 700 different biochemicals within these corn samples. 

The GMO corn sample was with glyphosate Roundup spray, and without Roundup spray, so we were able to also see if the spraying of the Roundup did anything to the composition of the corn. What we found was that, both at the level of the protein profile and the biochemical metabolite profile, there were major – dozens, hundreds of – changes in protein and metabolite profiles between the GMO and the non-GMO. Some of the changes we found were attributable to the spraying of the Roundup on the GMO crop. So clearly, when you do your analysis properly – as we felt we did by doing an in-depth composition analysis, not gross compositional analysis – it was very clear that the GMO corn NK603 was not substantially equivalent to non-GMO corn.

We couldn’t make any health claims about it. We could say, look, you’re saying it’s substantially equivalent, and on that basis you’ve passed this as safe. But what we’re saying is, they’re not substantially equivalent compositionally, and you’ve got to follow that up with long-term toxicity studies in an appropriate animal model system, such as the rat, to prove that it is substantially equivalent in terms of toxicity too. 

Having said that, we actually have industry data indicating that even in these 90-day rat feeding studies with NK603, exactly the same variety of GMO corn that we showed was not substantially equivalent, Monsanto’s own studies – when analyzed in depth – showed that the urine and blood biochemistry measures conducted by Monsanto as part of its application for market approval actually showed many statistically significant differences in multiple measures indicative of kidney and liver structural and functional damage.

So where does that leave us? It leaves us in a situation where the safety of GMOs needs to be reassessed properly, with a proper analysis of composition, and proper toxicity studies in the animals, and, if possible, conducted in such a way that if we see some kind of toxic effect, we try to identify the cause. What caused it? Is it the Bt toxin, is it the glyphosate Roundup residues, or other pesticide residues? Or is it the biochemical disturbances from the process that was used to generate the GMO? Because in our study comparing the GMO and the non-GMO corn, our data was very clear that it was the GM transformation process, because it has this mutagenic effect where it inadvertently damages the DNA of the plant, and as a result of that, it’s changed gene function. And as a result of that inadvertent change in gene function, this has led to all these biochemical changes that I’ve just mentioned, which clearly show that the GMO is not substantially equivalent to the non-GMO parent.

And that’s just one study. Ours was probably the most in-depth and comprehensive conducted along these lines, but others have found similar findings, though more limited in the range of analyses they did. But there are multiple studies that have shown that when you do proper in-depth, molecular profiling analysis, using these what we call omics methods — proteomics, metabolomics — you find significant differences between the GMO and non-GMO, especially with corn, as most studies have been done with corn. And that means this idea of substantial equivalency, the way it’s applied, I’m not saying it’s a bad idea, but if you’re going to apply it, do the analysis properly. Regulators have simply not asked industry to carry out that analysis properly, to prove whether it’s really substantially equivalent or not.

Stacy Malkan: Are any governments in the world using these molecular profiling analyses or omics methods, or doing long-term feeding studies of the type that you mentioned, to assess safety?

Prof Michael Antoniou: I’m afraid not. Regulatory agencies of all governments in the world, to the best of my ability to know, are digging their heels in about adopting these omics methods. They just don’t want to use them. They come up with various excuses as to why they shouldn’t use them, it’s just so unscientific it’s unbelievable. Omics analyses have revolutionized the way we carry out science, the way diagnosis of diseases and therapeutic interventions are measured. They have literally revolutionized science, and we’ve become incredibly good at predicting the consequences. Any changes that we see to, say, gene expression patterns, profiles — what’s called transcriptomics or proteomics and metabolomics. We’ve become very good, and we’re getting increasingly better at, predicting what the health consequences may be from changes in these omics measures resulting from an intervention of one sort or another in the context here of, say, a pesticide exposure or a GMO consumption. And so why regulators would be so reticent at adopting them is really incomprehensible because adopting them would revolutionize toxicology as well. 

They come up with excuses, oh but these methods are not validated. Okay, come up with a program that will validate them and convince you of their worth. But don’t just dismiss them as being too complex, or that you get a lot of data that you can’t interpret. That’s nonsense. We interpret our omics data as others do, and we make very good sense of it, because we have the bioinformatics computational tools to make sense of these large molecular profiling data sets. And from that we can make very good predictions as to whether something has or does not have any health consequences. 

So to me, regulators, by not adopting omics methods, are being negligent. And my feeling is that the reason why they’re not using it is that they can see what kind of results they could be faced with from adopting these methods. It’s just going to show that GMOs are never going to be substantially equivalent to non-GMO parent varieties, and so it means they can never use this idea of substantial equivalence as a starting point, or for claiming that a certain GMO is completely safe as a non-GMO and therefore you shouldn’t worry about it. I think at the end of the day, this is really what it boils down to. They just don’t want to scrap their regulatory risk assessment process as it is right now. It would certainly make life incredibly difficult for industry. But it would certainly serve the public better if they did adopt it.

Regulators say glyphosate is safe; independent science says otherwise

Stacy Malkan: I want to get back to this idea about layers of risk. You described new scientific analyses showing that GM is not the same as its non-GM counterparts. We now have many studies suggesting the hazards of glyphosate. Your group has studied glyphosate for many years and you are considered global leaders in understanding the toxicity. You have said that your work shows that the safety exposure limits set by regulators for glyphosate are not safe for health and need to be revisited. 

Prof Michael Antoniou: This is a huge body of work here. Glyphosate is probably now the most studied pesticide in the world ever; thousands of studies have been conducted on both health and environmental risk … When you consider the toxicity of glyphosate herbicides, it’s extremely important not only to consider the toxicity of just glyphosate on its own but the commercial glyphosate formulations, this complex chemical mixture of glyphosate plus the co-formulants.  A 2014 study by Robin Mesnage et al found that 8 out of 9 pesticide (insecticides, fungicides as well as Roundup) formulations tested were up to 1,000 times more toxic than the active pesticide ingredient (ie, glyphosate) when you exposed human cells to these agents in a laboratory.

The other thing we must consider is the mechanism of how glyphosate actually kills plants. It is a broad spectrum weed killer, which means it kills everything that is a plant. It inhibits a certain biochemical pathway in plants known as the shikimate pathway, which produces certain key essential amino acids. The glyphosate blocks the shikimate pathway, and the production of these amino acids. The plant can’t grow and dies very quickly. 

Because the shikimate pathway doesn’t exist in animals or humans, this was the foundation for claims of it being amazingly safe. But what those claims ignored was that, first, the shikimate pathway is present in certain bacteria and fungi, including bacteria in our gut microbiome. That’s very important, because we all know how important balance of composition and function of our gut microbiome is now to our health. I mean, it’s hardly out of the news.

And, of course, what about other mechanisms of toxicity? It’s not just inhibiting the shikimate pathway that we need to be concerned about; chemicals can bring about toxicity through many different mechanisms such as oxidative stress or direct DNA damage to name just two, perhaps even neurotoxicity. So these are fundamental issues that we need to consider when we consider the potential toxicity from exposure to glyphosate herbicides.

That’s why, in our animal studies, whenever resources permitted, we always compared different doses of glyphosate, and glyphosate-equivalent doses of representative commercial formulations of Roundup or other glyphosate herbicides, including typical ones that are used in the United States, or the UK and Europe. What we found was, first, we confirmed studies going back to the 1980s that showed that glyphosate and, even more so, commercial formulations like Roundup, caused oxidative stress both in the gut and within the body of the animals. When we analyze the blood and the liver, especially, we saw oxidative stress. 

What is oxidative stress? In our bodies, we produce molecules known as reactive oxygen species as part of our oxidative metabolism. We live on oxygen, we breathe air with oxygen, and part of that oxygen is used to produce energy in our body. Part of that process of what we call oxidative metabolism is that our bodies produce what are known as reactive oxygen species. When they’re produced in the right amounts, they perform very important functions in our body. We have mechanisms in our body that checks, so we don’t produce too much, and that neutralizes the reactive oxygen species. If you don’t check them, they will actually damage proteins, cells, and DNA within an organism. This is not a good thing; there needs to be a very fine balance of reactive oxygen species production. 

What we and others are finding, and I think my group has documented this more accurately than others: glyphosate, and more so the typical commercial glyphosate Roundup formulations, elicit an oxidative stress response. What that means is that the body cannot cope; the body in response to exposures to these toxic chemicals is producing excessive amounts of reactive oxygen species. This is not good because excessive reactive oxygen species and oxidative stress is a sign that the person is going to suffer cellular and DNA damage. The most worrying is the DNA damage, because DNA damage is the key risk factor to developing many diseases, but particularly cancer.

In the gut of the rats that we expose glyphosate and Roundup to, we found the shikimate pathway in our gut bacteria was inhibited by glyphosate and Roundup. So that’s already one disturbance that you don’t want happening, because if you disturb the shikimic pathway in your gut bacteria, you could disturb the balance of their functioning. But over and above that, we found biochemical changes indicative of oxidative stress. And that could lead to inflammation in the gut and leaky gut as well. And we all know again how bad a leaky gut syndrome can be in terms of bringing about food intolerances and allergies, especially in the young.

Then, when we looked inside of the bodies of the same animals, we found signs in the blood of oxidative stress and liver damage. Again we were using omics. The key to what we found in our studies: in order to see these signs of negative health outcomes, we had to use these in-depth omics – molecular profiling assays. If you use the regulatory standard gross measures, you could easily miss them. Looking at the omics biochemical analysis of the blood, we saw oxidative stress and signs of liver damage. When we actually looked at the liver in these same animals, sure enough, we saw oxidative stress. We saw changes in gene expression and other measures suggestive of DNA damage. And when we looked at the DNA directly for damage, we found that glyphosate alone caused DNA damage, which again, is not a good thing.  Ultimately we found that the liver was also suffering from what we call lipotoxicity — non-alcoholic fatty liver disease. 

This is a little bit of a digression, but I want to mention this because it’s such an important observation. In a 2017 study, rats were exposed for two years to a typical Roundup formulation that you could buy in France, at an incredibly low dose – the amount of glyphosate the rats were ingesting was 125,000 times lower than what is permitted in Europe and the UK, and 250,000 times lower than what’s permitted in the United States — staggering figures. But even though the level of ingestion of this Roundup was so incredibly low, I called it an ultra low dose in our publications, using the omics measures — not only just histology and gross biochemical measures but using gene expression, protein, and metabolite profiling methods – we showed that the livers of these animals were suffering from non-alcoholic fatty liver disease and its progression to the more serious non-alcoholic steatohepatitis: basically liver structure and functional damage that can ultimately lead to cirrhosis and liver cancer.

Why is this observation so important? This study to date is the only one that has tested Roundup in a rat toxicity model system at what we call an environmentally relevant dose that somebody could typically be exposed to. So is non-alcoholic fatty liver disease a problem in the United States? Well, actually, it’s a problem everywhere, virtually. Twenty-five percent of citizens in the United States suffer from non-alcoholic fatty liver disease, and its progression to more serious fatty liver conditions, as in Europe, as in many other nations as well.

What was discovered in this study, and proved later in our own subsequent investigations, was that even at doses of glyphosate and Roundup that regulators say we should see no negative outcomes in our animals – what’s known as no observed adverse effect level (NOAEL) and the acceptable daily intake – we were seeing fatty liver disease. We were seeing DNA damage, oxidative stress that was resulting in DNA damage. So this was not good news. 

The exposure in the rats started with animals in young adulthood – 5 or 6 weeks of age. We stuck with that because that’s the regulatory standard. It’s not ideal, but it’s the regulatory standard, and we wanted to be in compliance with regulatory guidelines. But in further follow-up investigations, we and others have started exposure prenatally, when the animals are developing in the wombs of their mothers. That, of course, is far more realistic. So it’s very important to try to replicate that in our laboratory toxicity study.

And that’s what the Global Glyphosate Study, led by the Ramazzini Institute in Italy, has done in their investigation. My group has been involved; we found that even after just 90 days of exposure post weaning, there were even more acute disturbances in the bacterial populations in the gut of these animals that we didn’t see when we started the exposure in young adulthood. In other words, it was far greater in this more real-world scenario with exposure from mid pregnancy onwards.

In parallel to that, the fungal population was markedly increased in these animals — again, something we didn’t see when exposure was from young adulthood onwards. So that was telling us that bacterial populations have been suppressed, and fungal populations then found themselves a niche to grow and expand. Now this is not good news, because many fungal species can be pathogenic; they can cause disease. So you don’t want to give room for pathogenic fungi to grow in your gut, which is what this investigation showed.  

In the carcinogenicity arm of the Global Glyphosate Study, they allowed the animals to grow all the way to two years of age after birth, with glyphosate exposure (starting in the womb) continuing through life to reflect the real world exposure scenario — and they looked at levels of leukemia in these animals. What was found: both the glyphosate alone, all the way down to this really low dose of the acceptable daily intake — European acceptable daily intake, which is half the USA acceptable daily intake — and going up from there, the animals suffered from different types of leukemia. It was very, very clear data, highly statistically significant, excellent well-executed study.

So, in other words, what the regulator says you and I can be exposed to – just glyphosate alone – levels of leukemia and the different types of leukemia were greater with exposure to the two Roundup formulations. But even with just glyphosate alone, it was carcinogenic. How can we explain this? Well, we go back to our earlier studies, especially from my own group. We showed that glyphosate and the Roundup formulations especially cause oxidative stress. Oxidative stress can cause DNA damage and alter patterns of gene expression, both of which can lead to cancer. So we have a logical kind of mechanism here of how the cancer, the leukemias in this case, could have arisen: oxidative stress, DNA damage, cancer leukemia.

An excellent epidemiological study conducted by Lianne Shepherd, who’s based at Washington University in Seattle, divided groups of people into a low-exposure group and a high-exposure group, and found that in the higher exposure group there was a 41% increase in risk of coming down with non-Hodgkin lymphoma.

Do we have any evidence for Roundup exposure causing oxidative stress and DNA damage in humans? Yes we do. In another epidemiological study, led by Vicky Chang, who works with Jonathan Hoffman at the NIH, they looked at the urine of agricultural workers and measured a readout for oxidative stress. They found a positive correlation between the frequency and duration of use of Roundup herbicide, and the levels of oxidative stress markers in the urine of these farm workers. The more glyphosate Roundup they were using, and the longer they were using it for, the higher the levels of oxidative stress markers in their urine were detected. Most tellingly for me is that one of those oxidative stress markers, which is what is known as 8 hydroxyguanine, not only is a marker of oxidative stress, but is also a marker of DNA damage.

It’s one of the classical measures of DNA damage that you can research in an epidemiological context. So here we have a marker that’s telling you, yes, the person suffered oxidative stress, and that was also leading to DNA damage.

One other study conducted by Vicki Chang and Jonathan Hoffman – and this was a blood analysis – showed that the higher the level of Roundup use, the higher the level of what is known as Y chromosome loss in the blood cells, which is another marker of DNA damage. So here we have two studies from human populations that are reflective of what we see in our animal studies: oxidative stress and DNA damage. Therefore, should we be surprised that long term frequent long-term use of glyphosate herbicides leads to cancer as a major contributor? I’m not saying it’s the exclusive cause, but it’s clearly a major contributor to at least this blood cancer. 

Stacy Malkan: The FDA says that glyphosate is safe but they’re not looking at most of the evidence you just described. So where does that leave us with US assurances of safety?

Prof Michael Antoniou: There was an excellent review published by Chuck Benbrook, asking how did EPA and academic scientists arrive at opposite views as to the DNA damaging potential of glyphosate herbicides? The EPA, when it looked at the data, said, no, it’s not DNA damaging, and we don’t have to worry about it being a carcinogen — whereas when you look at all the academic studies, not just from my lab, but well over a hundred conducted by independent academics, more than 70% of them found clear evidence that glyphosate Roundup were DNA damaging.

How is it that you get diametric opposing views? The EPA went purely by the industry data, industry-sponsored work, either directly or indirectly; only one out of a hundred studies found any evidence of DNA damage. Basically the EPA ignored the independent research and just focused almost exclusively on the work conducted by the industry as part of its application for market approval. And I’m afraid this is the problem we suffer all over with regulation of chemicals, including glyphosate herbicides. 

We suffer from it in Europe, we suffer from it here in the UK: regulators heavily, if not exclusively, rely on data put to them by the industry, which for me makes no sense. I’m not saying you ignore it, but it makes no sense just to focus on that because anybody can see there’s a conflict of interest there. When glyphosate herbicides were reapproved for another 10 years in Europe, a year or two ago, I was shocked by the evidence that was ignored by the European Food Safety Authority in arriving at its conclusion of safety for glyphosate herbicides. They certainly ignored all the animal studies that my group conducted that showed evidence of harm. And the other way they get away with passing glyphosate as safe is they ignore the clear evidence of heightened toxicity from commercial formulations. They only look at studies and data derived from just glyphosate alone and not from the greater toxicity that arises with the far more realistic exposure from the commercial formulations. 

We suffer from that everywhere. So where does that leave us? I’m afraid it leaves us in a very bad space. Because until regulators do two things: one is, they must consider toxicity of formulations as well as just active ingredients, and they’ve got to ask for long-term toxicity studies of the formulations as well as the active ingredient, which at the moment they’re not asking for at all — or if they’re asking for it, like in Europe, no one’s doing it.

Academics have had to do it for them. And now that academics are doing it, we are indeed finding that levels of exposure which they’re saying for glyphosate should be completely safe, we’re finding really serious negative health outcomes, whether it’s fatty liver or or even cancers.

And two, they need to adopt these omics methods. They’re missing a great opportunity here. One of the reasons we used omics molecular profiling methods was to see if we can actually reduce the numbers of animals and duration of exposure of Roundup in the animals. Anything  you can do to reduce numbers and reduce duration of exposure would be a good thing. Omics analysis, and the great predictive power of health or disease using these analytical methods, could be exploited to study smaller numbers of animals exposed for shorter periods of time. 

So it’s a missed opportunity, because regulators are always talking about even today, and at the moment worryingly, they want to replace animal studies with cell culture experiments, or even worse, just computer prediction. I’m sorry we just can’t do that. You can’t replace direct toxicity studies at the moment, at least, with any computer prediction models. But you can reduce the number and duration of your animal studies by exploiting the omics methods of molecular profiling, and the high predictive power of health and disease that they offer.

How can we protect ourselves? 

Stacy Malkan: If regulators are not guided by science, how can we, as consumers, take what we’ve learned to make better choices for our health?

Prof Michael Antoniou: What our work and others have shown is that the levels of glyphosate and Roundup exposure that regulators say are safe, clearly that’s wrong. The animal studies now are repeatedly showing that is clearly wrong. Minimally, if you’re going to continue using this, you’ve got to reduce the safe limit of exposure. And to be honest right now, I don’t know what that safe limit would be, because the long term exposure study hasn’t been done with lower and lower amounts to find out a level of exposure in our rat model system which says the animals are fine. Because right now, we’ve gone down to very low levels, and the animals are still suffering.

So, I don’t know. My guess would be, minimally, you need to reduce the acceptable daily intake or chronic reference dose by at least tenfold, but probably at least a hundred fold, maybe even more. I don’t know where to put a number to be honest.

So that’s one thing. Now, how can our listeners here protect themselves and their loved ones? Well, they can do what my family does, which is to source their food as much as possible organically grown.

I’m afraid organic farms tend to be surrounded by non-organic farms, so you can find pesticide residues even on organic foods, but they are a small fraction of what you find in conventionally grown foods. So my recommendation to people is to be as much as possible on an organic whole food diet and avoid GMOs and foods that can contain GMO ingredients. And clean up your water as well, because there can be pesticide residues in your water. So use some kind of filtration unit that you know will remove pesticides. I know people say organic food is more expensive, but it doesn’t have to be more expensive if you shop wisely. And at the end of the day ill health is far more expensive than what you may pay for organic food. Our health is the most important thing in our lives.  

Thank you for asking that question, because it’s very important that we do take our own steps, because at the moment regulatory agencies are favoring industry needs rather than properly protecting the public. That’s evidently clear throughout the world. In the European Union, there was a move to seriously reduce the amount of pesticide used. And now they’ve made a U-turn on that. Why? Because of industry lobbying. They didn’t want to inconvenience industry too much.

So our governments and our regulatory agencies are indeed failing us by not undertaking the kind of risk assessment that protects the public and puts the public health and the environment first. They’re putting money and industry first.


This article was first published by US Right to Know and is republished on GMWatch with their kind permission.