Rigging the Science of GMO Ecotoxicity

by Jonathan Latham

by Jonathan Latham, PhD

Researchers who work on GMO crops are developing special “artificial diet systems”. The stated purpose of these new diets is to standardise the testing of the Cry toxins, often used in GMO crops, for their effects on non-target species. But a paper published last month in the journal Toxins implies a very different interpretation of their purpose. The new diets contain hidden ingredients that can mask Cry toxicity and allow them to pass undetected through toxicity tests on beneficial species like lacewings (Hilbeck et al., 2018). Thus the new diets will benefit GMO crop developers by letting new ones come to market quicker and more reliably. Tests conducted with the new diets are even being used to cast doubt on previous findings of ecotoxicological harm.

GMO Cry toxins

Cry toxins are a family of highly active protein toxins originally isolated from the gut pathogenic bacterium Bacillus thuringiensis (Latham et al., 2017). They confer insect-resistance and up to six distinct ones are added to GMO corn, cotton, and other crops (Hilbeck and Otto, 2015).

Green Lacewing (Photo credit: Joseph Berger)
Green Lacewing (Photo credit: Joseph Berger Bugwood.org)

The resulting crops are usually called Bt crops. Cry toxins kill insects that eat the GMO crop because the toxin punches a hole in the membranes of the insect gut when it is ingested, causing the insect to immediately stop feeding and eventually die of septicaemia.

Cry toxins are controversial. Although the biotech industry claims they have narrow specificity, and are therefore safe for all organisms except so-called ‘target’ organisms, plenty of researchers disagree. They suspect that Cry toxins may affect many non-target species, even including mammals and humans (e.g. Dolezel et al., 2011; Latham et al., 2017; Zdziarski, et al., 2018).

Off-target toxicity

The Cry toxin mode of action, we and others have noted, does not necessarily discriminate between species. Any organism with a membrane-lined gut is, in principle, vulnerable if it consumes the GMO Bt crop. In these Bt crops the leaves, straw, roots, nectar, and pollen, all typically contain Cry toxins. Therefore, most organisms in agricultural landscapes will at some point in their life-cycle be exposed to GMO plant material. As pollinator declines and a more generalised insect apocalypse have revealed, the question of the effects of such crops on biodiversity is far from trivial.

The biotech industry is also very much aware of the steady stream of research, from evidence of allergenicity, to toxicity, of their Cry proteins towards so-called ‘non-target’ organisms. Organisms affected by Cry toxins include monarch butterflies, swallowtail butterflies, lacewings, caddisflies, bees, water fleas, and mammals (Losey et al., 1999; Bøhn et al., 2008; Ramirez-Romero et al., 2008; Schmidt et al., 2008; Sabugosa-Madeira et al., 2008; Mezzomo et al., 2015; Zdziarski, et al., 2018). Much of this research does not get the attention it deserves (e.g. COGEM 2014), but if swallowtail butterflies can succumb to just 14 pollen grains of Syngenta’s BT-176 corn (Lang and Vojtech, 2006) the industry is aware it can hardly truthfully market GMOs as environmentally beneficial.

As we have reported, one response of the biotech industry has been to try to bake approval into regulatory decisions. That is, make regulatory processes operate such that no possible future findings of unexpected harm by Cry toxins towards non-target organisms, no matter where in the risk assessment process they are observed, can derail approval. Thus, in “The Biotech Industry Is Taking Over the Regulation of GMOs from the Inside” we showed how “tiered risk assessment”, a regulatory procedure being promoted by the crop biotech industry, functions, in practice, as an “approval” system (Romeis et al. 2008). That is, regulatory denial of an application, under tiered risk assessment is nigh impossible.

How to mask Cry toxicity

The latest development, which adds further to the improbability that GMO Cry toxins will fail risk assessment, comes to light thanks to Hilbeck and colleagues (Hilbeck et al., 2018).

They report that these new “artificial diet systems” for raising non-target organisms contain surprisingly large amounts of antibiotics (Li et al., 2014; Ali et al., 2016a; and Ali et al. 2016b). The significance of this is that antibiotics are known to act as antidotes to Cry toxins (Broderick et al., 2006, Mason et al., 2011). By masking the harm caused by the toxin, antibiotics can give the unsuspecting reader a false impression of Cry harmlessness.

This effect of antibiotics was first shown in 2006 by researchers at the University of Wisconsin. They and others showed that gut bacteria are required for Cry toxins to achieve their full effect (Broderick, et al., 2006; Mason et al., 2011). Broderick et al wrote:

“Here, we report that B. thuringiensis does not kill larvae of the gypsy moth in the absence of indigenous midgut bacteria. Elimination of the gut microbial community by oral administration of antibiotics abolished B. thuringiensis insecticidal activity” (Broderick et al., 2006)

What Hilbeck et al., note about the diets, which their inventors claim are “needed” for reproducible testing of Cry toxins on carnivorous insects such as green lacewings and ladybirds, are the very large quantities of antibiotic compared to the amounts necessary to prevent spoilage (Li et al., 2014; Ali et al. 2016a; and Ali et al. 2016b). The diets developed by these authors contain the antibiotics Streptomycin (130mg) and Cephalosporin (50mg) per 100g (Li et al.; 2014); or Streptomycin (400mg) and Penicillin (400mg) per 100g (Ali 2016a); and Streptomycin (400mg) and Penicillin (400mg) per 100g (Ali 2016b).

Such large antibiotic quantities are questionable on several grounds. Previous authors have specifically noted the need to minimise antibiotic use in test diets intended to measure the toxicity of Cry proteins. As Porcar et al. (2010) wrote:

“Antibiotics were deliberately excluded from the diet composition since bacteria occurring in the insect midgut naturally might be critical for sensitivity (Broderick et al., 2006).” (Porcar et al. 2010)

 Furthermore, a very similar diet developed by other authors, also in China, for green lacewings (reared for different reasons), used no antibiotics (Cheng et al., 2017). Third, Li et al. (2014) claimed their diet is a development of one from Cohen and Smith (1998). Yet the diet developed by Li et al. contains almost six times as much streptomycin and two and a half times the level of a second antibiotic, cephalosporin (though Cohen and Smith used tetracycline). No mention or explanation of the raised antibiotic levels was made.

In other words, the problem of antibiotics acting as antidotes to Cry proteins is widely known to Cry toxin researchers but is ignored by the authors of these three papers. Less surprisingly, using their antibiotic-laden diet Li et al. in 2014 found “no detrimental impact of these Cry proteins on any of the C. sinica (Green lacewing) life-table parameters measured”, as also did Ali et al. (Ali et al., 2016a)

Such papers as these have multiple harmful effects. Their results contradict (in strong probability erroneously) previous findings that GMO Cry toxins harm non-target insects like ladybeetles and lacewings; thus placing earlier findings in doubt. Second, future experimenters who adopt these diets will also likely, wittingly or unwittingly, obtain falsely negative results. Third, if the claimed “need” for them is any guide, these artificial diet systems will be promoted (with industry help) to be adopted as gold-standards for future research, just like tiered risk assessment.

Unsurprisingly perhaps, researching the background of these authors one finds that Joerg Romeis is one of them (Li et al., 2014). Romeis is a Swiss academic who has for many years been closely associated with the biotech industry. Romeis has called studies that find harmful effects of Cry toxins “bad science”; he has published many papers with industry authors that he believes refute evidence of effects on non-target organisms, and he led their tiered testing project (Romeis et al., 2008; Romeis et al., 2013).

Who is really doing bad science?


Ali, I.; Zhang, S.; Cui, J.J. (2016a) Bio-safety evaluation of Cry1Ac, Cry2Ab, Cry1Ca, Cry1F and Vip3Aa on Harmonia axyridis larvae. J. Appl. Entomol. 141, 53–60.

Ali, I.; Zhang, S.; Luo, J.-Y.; Wang, C.-Y.; Lv, L.-M.; Cui, J.-J. (2016) Artificial diet development and its effect on the reproductive performances of Propylea japonica and Harmonia axyridis. J. Asia-Pac. Entomol. 19, 289–293.

Bøhn T., Primicerio R., Hessen D.O., Traavik T. (2008) Reduced Fitness of Daphnia magna Fed a Bt-Transgenic Maize Variety. Arch. Environ. Contam. Toxicol. 55: 584-92.

Broderick, N.A.; Raffa, K.F.; Handelsman, J. Midgut bacteria required for Bacillus thuringiensis insecticidal activity. Proc. Natl. Acad. Sci. USA 2006, 103, 15196–15199.

Cheng, Y., Zhi, J., Li, F., Jin, J., & Zhou, Y. (2018). An artificial diet for continuous maintenance of Coccinella septempunctata adults (Coleoptera: Coccinellidae). Biocontrol Science and Technology, 28(3), 242–252.

Cohen A C, L K Smith (1998) A New Concept in Artificial Diets for Chrysoperla rufilabris: The Efficacy of Solid Diets.Biological Control Volume 13, Issue 1, September 1998, Pages 49-54.

Dolezel, M., Miklau, M., Hilbeck, A., Otto, M., Eckerstorfer, M., Heissenberger, A., … Gaugitsch, H. (2011). Scrutinizing the current practice of the environmental risk assessment of GM maize applications for cultivation in the EU. Environmental Sciences Europe, 23, 33. doi:10.1186/2190-4715-23-33.

Hilbeck, A., & Otto, M. (2015). Specificity and combinatorial effects of Bacillus thuringiensis Cry toxins in the context of GMO environmental risk assessment. Frontiers in Environment Science. doi:10.3389/fenvs.2015.00071

Hilbeck A, N Defarge, T Bøhn, M Krautter, C Conradin et al. (2018) Impact of Antibiotics on Efficacy of Cry Toxins Produced in Two Different Genetically Modified Bt Maize Varieties in Two Lepidopteran Herbivore Species, Ostrinia nubilalis and Spodoptera littoralis. Toxins , 10: 489;

Latham, J.R.; Love, M.; Hilbeck, A. The distinct properties of natural and GM Cry insecticidal proteins. Biotechnol. Genet. Eng. Rev. 2017, 33, 62–96.

Losey J. E. , L. S. Rayor & M. E. Carter (1999) Transgenic pollen harms monarch larvae. Nature 399: 214.

Li, Y.; Hu, L.; Romeis, J.; Wang, Y.; Han, L.; Chen, X.; Peng, Y. Use of an artificial diet system to study the toxicity of gut-active insecticidal compounds on larvae of the green lacewing Chrysoperla sinica. Biol. Control 2014, 69, 45–51.

Mason, K.L.; Stepien, T.A.; Blum, J.E.; Holt, J.F.; Labbe, N.H.; Rush, J.S.; Raffa, K.F.; Handelsman, J. From commensal to pathogen: Translocation of Enterococcus faecalis from the midgut to the hemocoel of Manduca sexta. mBio 2011, 2, e00065-00011.

Mezzomo, BP, Miranda-Vilela,AL and Grisolia CK (2015) Toxicological Evaluation of a Potential Immunosensitizer for Use as a Mucosal Adjuvant—Bacillus thuringiensis Cry1Ac Spore-Crystals: A Possible Inverse Agonist that Deserves Further Investigation. Toxins (Basel). 2015 Dec; 7(12): 5348–5358.

Porcar, M.; Garcia-Robles, I.; Dominguez-Escriba, L.; Latorre, A. (2010) Effects of Bacillus thuringiensis Cry1Ab and Cry3Aa endotoxins on predatory coleoptera tested through artificial diet-incorporation bioassays. Bull. Entomol. Res. , 100, 297–302.()

Ramirez-Romero R.; Desneux N.; Decourtye A.; Chaffiol A.; Pham-Delègue M.H. (2008) Does Cry1Ab protein affect learning performances of the honey bee Apis mellifera L. (Hymenoptera, Apidae)? Ecotoxicol Environ Saf. 70: 327-33.

Romeis, Jörg; Bartsch, Detlef; Bigler, Franz; Candolfi, Marco P; Gielkens, Marco M C; et al. (2008) Assessment of risk of insect-resistant transgenic crops to nontarget arthropodsNature Biotechnology; 26: 203-8.

J Romeis, MA McLean, AM Shelton (2013) When bad science makes good headlines: Bt maize and regulatory bans. Nature Biotechnology 31: 386–387.

Rosi-Marshall, E.J.; J. L. Tank; T. V. Royer; M. R. Whiles; M. Evans-White; C. Chambers; N. A. Griffiths; J. Pokelsek and M. L. Stephen (2007) Toxins in transgenic crop byproducts may affect headwater stream ecosystems. Proc Natl Acad Sci USA 104: 16204–16208.

Sabugosa-Madeira B.; Abreu I.; Ribeiro H. and Cunha M. (2007) Bt transgenic maize pollen and the silent poisoning of the hive. Journal of Apicultural Research 46: 57-58.

Schmidt J.E.; Braun C.U.; Whitehouse L.P.; Hilbeck A. (2008) Effects of Activated Bt Transgene Products (Cry1Ab, Cry3Bb) on Immature Stages of the Ladybird Adalia bipunctata in Laboratory Ecotoxicity Testing. Arch Environ Contam. Toxicol. 56: 221-228

Zdziarski, I M., J.A. Carman, J.W. Edwards (2018) Histopathological Investigation of the Stomach of Rats Fed a 60% Genetically Modified Corn DietFood and Nutrition Sciences9, 763-796.

If this article was useful to you please consider sharing it with your networks.

Print Friendly, PDF & Email
Comments 15
  • What are we doing people? Allowing this to happen? We, AND they, must shift our purpose here on planet earth if we are to thrive — as opposed to joining the list of extinctions caused by human activity. Instead of exploiting and abusing nature and its resources, we must learn to work with nature toward full sustainability,food security,and healthier children.

  • Love this — saw it on Counterpunch and will now subscribe to your site. Thank you.

  • This problem was constantly on my mind as I was reading through the Monsanto regulatory applications. It was normal in the feeding/toxicity studies to read the Monsanto self-report that the study complied with “Good Laboratory Practice” with the following exception:
    “Characterization of the test substance was the responsibility of the Sponsor” (ie, Monsanto characterized the substances, rather than the testing house.
    Ok… so only Monsanto knew what was in the substances. How could any regulator take such studies seriously?

  • Dr. Latham, thank you for your latest effort in revealing this deeply misguided science. Do you have any information on whether cry toxins would develop in the kernels of non-GMO corn which has been inadvertently pollinated by GMO BT corn? My corn patch grows downwind of some big corn acreage that is likely GMO.

  • Good question. I don’t know of any research on that question specifically but plant seeds contain parts that are fertilised, such as the endosperm. If the pollen is GMO then those parts will contain DNA and may contain the protein. See this wikipedia page https://en.wikipedia.org/wiki/Endosperm
    It is possible that the pollen DNA doesnt produce Cry protein. Only an experiment can determine that.

  • Jonathan Carr, I looked at a similar question for a while. I wondered at what stage in development a constitutive promoter would actually stop. Would it keep transcribing the cry transgenes even as the seed was going dormant. I didn’t find an answer in respect of GM transcription. I did learn that even when a seed is dormant, its DNA is still transcribing, usually saying “don’t wake up yet”, etc. Someone might have looked at what role the pollenating genetic material plays in the development of the grain – any role, or none?
    At any rate, the pollen coming onto your property from a GM field nearby, and perhaps blown on the high winds by a tornado will contain GM Cry proteins, at least in respect of two GM lines MON810 and MON863 (I just checked) – maybe all of them.

  • I have a number of scientific and scholarly issues with this article. I have detailed these in a medium post: https://medium.com/@niederhuth/omitting-the-science-a-critique-of-rigging-the-science-of-gmo-ecotoxicity-af19b2bcfdca

    • There are several mistakes in your Medium article. To begin with you attempt to argue against adverse effects in monarchs from Bt176 by using studies that actually observe adverse effects from Bt176.
      “1) Cherry-picking studies: the article completely ignores 20+ years of research contradictory to their claims.
      Specific example: The article mentions that Bt corn pollen was lethal when force fed to Monarch larvae[2], but does not mention that field studies have shown little risk to Monarchs[3, 4].”
      You reference for #2 Stanley-Horn, et al. which states:
      “Corn hybrids based on event 176 may be hazardous to susceptible stages of monarch larvae that are present on their host plants within cornfields during anthesis”
      That is a pretty straightforward claim, from your own source, that Bt176 may be hazardous to monarchs. So you appear wrong on that one.
      You go on to state, “It’s not just the claims made about Monarchs either. Similar work in other organisms, like swallowtail butterflies[11], lacewings[5], etc. have also shown little risk to non-target insects[12].”
      Let’s look at the first statement here on swallowtails. In the article here it states, “if swallowtail butterflies can succumb to just 14 pollen grains of Syngenta’s BT-176 corn (Lang and Vojtech, 2006)”
      This is clearly an argument about Bt176. Your reference, which you claim shows little risk to swallowtails, is this, “11. Wraight, C.L., et al., Absence of toxicity of Bacillus thuringiensis pollen to black swallowtails under field conditions. Proceedings of the National Academy of Sciences, 2000. 97(14): p. 7700–7703. ”
      Here is what that study says, “Potted host plants were infested with first instar black swallowtails and placed at intervals from the edge of a field of Bt corn (Pioneer 34R07 containing Monsanto event 810) at the beginning of anthesis.”
      Your reference is about Mon810 and not Bt176 so it is in no way a rebuttal to the claim made in this article.
      I would suggest you revisit your article, because it appears several of your arguments are invalid.

      • Thanks for bringing this up Robert.

        As to BT176, registration for this line ended in 2001 in the US and it has not been grown in the US since then. Why is this important information excluded from the discussion?

        • I’m afraid your critique still has some serious holes in it, Chad.

          The Bt176 references support some of the main ideas in the article if you look at them in context.
          One of the main ideas of the article is, “Cry toxins are controversial. Although the biotech industry claims they have narrow specificity, and are therefore safe for all organisms except so-called ‘target’ organisms, plenty of researchers disagree.“
          Here the idea is that people are claiming that all Cry toxins only impact target species and all NTOs are not harmed by any Cry toxins. All one would need to refute this claim is one example of a Bt crop harming an NTO, and Bt176 impact on monarchs and swallowtails are good examples. Other examples are given as well in the article which are not about Bt176 such as the impact of Mon810 of water fleas. Although it is understandable that you may have missed that since the hyperlink seems to go to the wrong paper which was probably supposed to be: Bøhn, T., Primicerio, R., Hessen, D. O., & Traavik, T. (2008). Reduced Fitness of Daphnia magna Fed a Bt-Transgenic Maize Variety. Archives of Environmental Contamination & Toxicology, 55(4).
          Another main idea of this article is, “As we have reported, one response of the biotech industry has been to try to bake approval into regulatory decisions. That is, make regulatory processes operate such that no possible future findings of unexpected harm by Cry toxins towards non-target organisms, no matter where in the risk assessment process they are observed, can derail approval.”
          The idea here is that there is a failure in the regulatory process. Bt176 is again a good example of this. Bt176 was deregulated by the EPA and was being grown commercially for years before adverse effects were observed in experiments. Here Bt176 is being used as evidence of a failure in the regulatory process.
          You argue, correctly, that no adverse effects were observed from Mon810 in the studies you referenced within the parameters of the experiments. That, however, does not refute a claim that some NTOs are adversely impacted by some Bt crops, or even that all butterflies are not adversely impacted by some Bt crops. There is evidence that some butterflies, including monarchs, may be adversely impacted by Mon810 and other Bt crops on the market.
          Here are just a few examples:
          Niels Holst, Andreas Lang, Gabor Lövei, Mathias Otto (2013) Increased mortality is predicted of Inachis io larvae caused by Bt-maize pollen in European farmland Ecological Modelling 250 (2013) 126–133
          Paula DP, Andow DA, Timbó RV, Sujii ER, Pires CS, Fontes EM (2014) Uptake and transfer of a Bt toxin by a Lepidoptera to its eggs and effects on its offspring. PLoS One. 2014 Apr 18;9(4):e95422.
          Paula DP, Andow DA, Timbó RV, Sujii ER, Pires CS, Fontes EM (2014) Uptake and transfer of a Bt toxin by a Lepidoptera to its eggs and effects on its offspring. PLoS One. 2014 Apr 18;9(4):e95422.
          The point is that this really isn’t a case of cherry-picking or omitting. If someone claimed all cars are red. All I would need is 1 example of a blue car, green car, etc. to reject their claim. That isn’t cherry-picking or omitting. Here the claim is that some people say “all” Cry toxins only adversely impact target organisms, so presenting any evidence of NTO harm from any Cry toxin is a valid argument. You can’t argue that they cherry-picked or omitted evidence in this case. It may be that most experiments with most Cry toxins don’t have any observable effects, that isn’t demonstrated here due to the limited number of references in your article, but even if that were true “most” doesn’t mean “all”.
          If someone claims only tests used during the initial regulatory process are valid all you would need is 1 example of a Bt crop that adversely impacted an NTO after it was deregulated, like Bt176, to reject that claim. That isn’t cherry-picking or omitting either.

  • You probably have to be in the middle of an insect population eruption to notice it, but BT demonstrates beyond doubt that insects can suffer. They suffer for about a week before they die. Humans are *normally* immune to BT, but some have had painful symptoms similar to what insects show…and our vulnerability may rise with repeated exposure. It’s not a disease we want to start having, or watching our children have.

  • I am not a scientist and I have a question that has been on my mind since I first read about bt crops. An article about the use of bt toxins on organic crops states that in bt toxins “The ICPs are responsible for insect toxicity. ICPs are usually biologically inactive within hours or days.” Is this not a very different thing than putting bt that never becomes biologically inactive into every cell of a plant? Or am I not understanding something?

    • You are absolutely right. In their natural bacterial state Cry1 protoxins form into crystals, that by report require a highly alkaline environment to dissociate, and action by proteases and DNases to bring the protoxins down to their active state. Cry1-type toxins/toxicants in GM crops do not form crystals, and have either been specifically engineered to be produced in shortened active state, and/or are reduced down to active state by proteases in the plants. Most of the claimed “safety mechanisms” associated with the Cry crystals used in organic agriculture have been bypassed in GM crops. Indeed some of these inventions were called “supertoxins” in a Monsanto patent document. This is discussed at length in this ISN article https://www.independentsciencenews.org/environment/have-monsanto-and-the-biotech-industry-turned-natural-bt-pesticides-into-gmo-super-toxins/ from a paper written Jonatham Latham et al.

Leave a comment