GM Free Cymru

Nature Biotechnology’s Dummy Proof

The document reproduced below has been converted without other modification from a PDF document into a Word document.  It illustrates how Dr Irina Ermakova was misled by a formerly reputable journal into the belief that the paper entitled “GM soybeans and health safety—a controversy reexamined” would be published under her authorship.  This is the most serious piece of professional malpractice by a journal that we have ever seen.

On 20 August the Senior Production Editor of Nature Biotechnology sent Irina Ermakova a "dummy proof" in PDF format, with the title "GM Soybeans and health safety -- a controversy reexamined" and with her name beneath the title as author.  Eight of the original 12 references submitted by her had been deleted.  In the introductory paragraph (presumably written by the Editor) were the words "Nature Biotechnology approached Ermakova to ask for a detailed account of her work in her own words.  Her answers are presented below together with comments solicited from  a group of researchers working in the field."  The comments from the group of researchers were NOT included in the dummy proof, which was referred to as a "publication proof."



GM soybeans and health safety—a controversy reexamined

Irina V Ermakova

An unprecedented study claiming that transgenic soybeans compromise the fertility of rats and the survival and growth of their offspring has garnered widespread media and political attention but remains unpublished in the peer-reviewed literature. Here, an account of the work from the principal investigator, Irina Ermakova, is appended with comments from researchers in the field.

Neurologist Irina Ermakova of the Institute of Higher Nervous Activity and Neurophysiology of the Russian Academy of Sciences in Moscow made news headlines two years ago when she reported that rats fed diets containing glyphosate-tolerant genetically modified (GM) soybeans gave birth to pups with low survival rates or stunted growth1. Though these findings have yet to appear in a peer-reviewed journal and contradict publications in the literature, they have been widely disseminated and discussed over the media and internet and already cited by >500 organizations as evidence of the potential toxicity of GM products. They’ve also prompted the American Academy of Environmental Medicine Wichita, KS, USA) to call for additional independent studies of food safety for GM crops2, been referred to in a state Australian parliamentary debate as a reason to ban GM crop cultivation3 and motivated regulatory agencies in several countries to review their approvals of GM organisms or to comment on the work4,5. Nature Biotechnology approached Ermakova to ask for a detailed account of her work in her own words. Her answers are presented below together with comments solicited from a group of researchers working in the field.

Briefly describe your experimental design and methods.

Irina Ermakova. My experiments were designed to study the influence of a diet containing genetically modified (GM) soy beans (Roundup Ready (RR) line 40.3.2) on the physiological state and behavior of Wistar rats and their offspring. In addition to laboratory chow, one group of female rats were fed soy flour or seeds for 2 weeks before mating, during mating and pregnancy, and were fed an increased daily amount for every pup during lactation  At the same intervals, a second group of female rats receiving chow was fed conventional soy flour or seeds and a third group received protein isolated from RR GM soy. A fourth group of rats received only the laboratory chow and was considered to be a positive control. We analyzed the physiological state (weight, size and so forth), reproductive functions, rate of mortality and behavior of rats and their offspring. Experiments were repeated five times using soy flour, soy seeds, standard chow and chow mixed with GM soy (~14%) in different groups of rats. Standard chow contained wheat, wheat bran, sunflower, meat flour, animal fat, barley, fodder yeast, microelements and vitamins. RR soy flour genetically modified with the transgene 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS) obtained from Agrobacterium sp. strain CP4 (Monsanto; St. Louis, MO, USA), its protein isolate and conventional soy flour (Arcon SJ 91-330), which has a similar composition and nutritional value to RR GM soy, were obtained from the Netherlands supplier of Archer Daniels Midland (ADM; Decatur, IL, USA). Analysis of soy flour by PCR showed the presence of the EPSPS transgene in all samples of RR GM soy. The chow was administered as dry pellets from a special container placed on the top of their cages and the (GM, GM protein isolate or conventional) soy flour mixed with water (20 g soy paste in 40 ml water) in a small container placed inside their cage for three rats. Each rat thus received 6–7 g flour every day. A similar scheme was used for soy seeds, which were kept in water for 1 day before feeding and then put into a small container inside the cage: four seeds for one female and six seeds for one male.

How many animals were studied and how many experiments were pooled into your final results?

I.E. We repeated the experiments five times with different groups of animals and with the four RR GM soy supplementations (that is, GM flour, GM seeds, protein-isolate GM soy or chow with GM soy). Rats in control groups received conventional soy (as flour or seeds). In the first three repeats of the experiments, 30 females, 40 males and 221 pups were investigated. In total, for the five repeats of the experiments, we examined 48 females, 52 males and 396 rat pups. Similar results were obtained in all the different repeat experiments.

How were the animals housed and observed during the study?

I.E. Rats, weighing from 180 g to 200 g, were kept in a vivarium with a reversed light-dark cycle (12 a.m. to 12 p.m.). Each day, females and males in every cage received dry pellets from a special container placed on the top of their cage. Animals were also provided with 200 ml of drinking water per rat per day. After 2 weeks on the different diets, three females from each group were mated with two healthy males of the same age, who had not been exposed to the soy flour supplements. First one male was placed with a female in the cage for 3 days, and then another for 3 days. To minimize infection risk to females, invasive tests to determine sperm count and quality were not determined. Upon delivery, all females were transferred to individual cages, and the amount of soy supplement was increased by an additional 1 g for every pup born. Laboratory chow and water were available ad libitum during the experimental period, for all animals. When rat pups could feed themselves, the daily dose of soy supplement was increased to 2–3g for each pup. All rats ate their soy portions well.

What methods were used to assess animal health and behavior?

I.E. Adult animals were weighed before feeding and 2 weeks following commencement of the feeding experiments. Weights and sizes of pups from the different experimental groups born at the same time (±1–2 days) were recorded 2 weeks after birth. We also determined the weight of some internal organs (e.g., brain, liver, spleen, heart, lungs, kidneys and testes) and analyzed the morphology of the liver and testes. We examined the explorative behavior in the open field, determined the level of anxiety using a light/dark test and observed rat behavior in home cages. Experiments were performed with male and female rats 2 weeks after commencement of feeding and when pups were 2 months old. All experiments were conducted in the second half of the day when rats were more active (starting at 5 p.m.). Each group contained 9–10 animals. The open field was represented by [AU:OK?] a round platform, 100 cm, in diameter divided into zones restricted by sector rays and concentric circles. The platform was surrounded by  a wall, 30 cm high. The center of the open field was illuminated by a frosted bulb (40 W). The session was conducted in a sound- and lightproof room. A rat was placed in the center of the open field and the number of horizontal translocations, vertical positioning, grooming, number of boluses (defecation) and freezing were recorded over 6 min. For each parameter, the relative value of extinction was estimated as the following ratio: difference in activity between the second and the first 3-min intervals divided by integral activity. The level of anxiety was investigated using a light-dark test (Intertex, Multiscreen [AU: Location]) for 5 min. This model included two boxes: dark and light (four 3.5-W lamps). The number of rat entries into the light box, time spent in the light box, duration of rat rearing on hind legs in the light box, the latency before a rat first entered the light box, the number of times a rat looked out from the dark box, the vertical activity of rats in the light box, urinations, defecations and grooming were all recorded. We analyzed the level of mortality in each of the test groups using one-way ANOVA verified using Newman-Keuls share distribution test. Pup weight was analyzed by Mann-Whitney and its distribution by Chi-square using StatSoft (Moscow) Statistical version 6.0.

Briefly describe the main findings from your study.

I.E. Our data demonstrate a high level of mortality in pups born to mothers receiving RR GM soy–supplemented diets during the 3 weeks following birth compared with pups from control groups over the same period. Many (more than one-third) of the surviving pups born to mothers receiving GM soy had a stunted size and low weight compared with pups born to mothers from controls. A similar number of pups were born to mothers receiving GM soy, traditional soy and control groups (10–11 pups per female) but fewer pups were born to rats receiving soy protein isolate (8 pups per female). Behavioral studies indicated a high level of anxiety and aggression in males, females and young pups fed on the different groups fed GM material.[AU: Word/words missing? Or, should “fed GM material” be “of GM material”?] Morphological analysis of internal organs indicated marked pathological changes in the blood supply to testes and vacuolization in the livers of male rats fed GM soy seeds. We also failed to breed second-generation (F2) pups from matings of first-generation (F1) females and males fed material based on GM soy.

What was the level of mortality of the pups you found in the control and test groups?

I.E. In first three repeats of experiments, up to five times higher mortality was observed in newly born pups whose mothers had received the GM soy flour supplementation compared with pups from rats receiving GM soy protein isolate, traditional soy or laboratory chow (controls) (see Table 1). Pups from rats that had been fed a GM soy diet died during the 3 weeks following birth; pups from rats fed laboratory chow (positive control) died during the 2 weeks postpartum; and pups from those fed traditional soy died during the first week after birth.

What was the weight of the control and  test group animals?

I.E. We did not find any significant differences in the weights of adult rats fed the different diets. Even so, for 2 weeks following birth, the weights of pups from mothers fed GM soy supplement were lower than those of pups from rats in the positive control (laboratory chow) group or from the conventional soy flour–supplemented group. We also found that 33% of pups from rats fed GM soy had smaller sizes and lower weights than pups from rats fed laboratory chow, traditional soybeans or soybean protein isolate (Table 3). A crude anatomical analysis revealed that the organs of pups from rats fed GM soy were much smaller (except the brain mass) than those from pups born to rats fed other diets (Table 4). Thus, age-matched pups in the test and control groups show differences in the development of internal organs.

How was animal behavior and fertility affected?

I.E. Behavioral experiments showed very slight differences between groups in open field behavior. Even so, both anxiety in the ‘light-dark’ test and aggression were higher in females, males and offspring receiving GM soy in their home cages than in rats from other groups. Aggression was more frequent in females and pups; not only toward one another, but also toward the laboratory personnel caring for them. Some (~20%) of the females, fed by GM soy, failed to care for their pups (instead scattering them around the cage without nesting). For rats fed GM soy, we failed to breed second-generation pups from F1 males (n = 24) and females (n = 24). In marked contrast, the crossing of F1 females (n =12) receiving the GM soy diet with F1 males (n = 12) from the positive control group (laboratory chow) resulted in 72 pups (Table 5). Even here, however, the number of pups per female was fewer than in the other groups (8 pups per female instead of 10–11 pups per female) and 25% of females didn’t deliver pups at all. These results indicate that GM soy had a deleterious effect on the reproductive function especially of F1 males, but also female rats.

What do you conclude from your findings and what are your plans for future research?

I.E. As it is well established that raw soybean contains several antinutrients (e.g., lectins and trypsin inhibitors)1 and female hormone-like substances (e.g., phytoestrogens), our experiments both used a positive control (laboratory chow alone) and fed rats experimental and control diets 2 weeks before mating, during mating, through pregnancy and until the litters were weaned. The very high rate of pup mortality in litters of mothers on a diet supplemented with RR GM soy flour was very unexpected. The lower weight of surviving pups from rats receiving GM soy was also notable, particularly because the higher mortality resulted in (~50%) smaller litters, which should have doubled the amount of milk available. These pups should have had a better chance to grow than pups from other groups with larger litters, unless the amount and/or the quality of the milk is deleteriously affected by consumption of GM soy flour. We concluded that RR GM soy appears to have a strong negative influence on Wistar rats and their offspring, causing high levels of pup mortality, infertility in surviving pups, decreased weight gain in some pups, pathological changes in internal organs and deleterious effects on behavior. My opinion is that GM soy’s effect on Wistar rats and their offspring should be relevant to all mammals, including humans. It would have been instructive to compare the effect on rats and their offspring of RR GM soy with another GM soy line or with a completely different kind of GM plant. I hope to perform these experiments in future. We plan to compare the influence of different GMOs [genetically modified organisms] (not only RR soybeans) on the physiological state and behavior of rats and their offspring. We are also planning to analyze the reason of pup’s death and attempt to detect the presence of foreign DNA in white blood cells, brain, liver and other internal organs of adult animals and pups.

Do you feel that the translation/interpretation of your work has been accurate?

I.E. My experiments were published first in Russian and then in English. There were several incorrect (some even funny) interpretations of my work. One of the most serious critiques of my work was published in the “Statement on the effect of GM soy on newborn rats” from the UK’s Advisory Committee of Novel Foods and Processes (ACNFP; London)2. The Committee compared my research with only one (!) published article by Brake and Evenson3. But my study is not comparable with the work by Brake and Evenson for several reasons. First, the focus of the two investigations was completely different. Our experiments analyzed the effect of GM soy on mortality, physiological state and behavior of pups; in contrast, the studies of Brake and Evenson investigated the effect of GM soy on fetal, postnatal, pubertal and adult testicular development. Second, we used several different schemes of feeding; we commenced feeding 2 weeks before mating, which suggests that foreign genes ingested by these animals can penetrate and affect the sexual cells and/or organs. In the experiments of Brake and Evenson “pregnant mice were fed a transgenic soybean or a nontransgenic (conventional) diet through gestation and lactation....Multi-generational studies were conducted in the same manner.” Thus, in their study, foreign genes could influence only embryonic cells in the womb and not sexual cells or organs before mating. And third, Brake and Evenson used only a very small number of pups in their study: “At each point, three male mice were killed, the testes surgically removed and the cell populations measured by flow cytometry.” And they also mated a smaller number of animals: “Two C3H/HeJ males and two C3H/HeJ females were bred to keep that strain pure.” In our experiments, more females and males were mated and 10–20 times more pups were obtained in each group. Thus, it is clear that my investigation and that of Brake and Evenson’s are quite different and should not be compared.

Why have you so far forgone publishing your work in a peer-reviewed journal?

I.E. I first presented the data at the 11th Russian Gastroenterological Week (in a section on Nutrition and GMOs organized by the Moscowbased National Association for Genetic Safety) at the Russian Academy of State Service in Moscow, October 10–12, 2005. I was perplexed by my data and I appealed to scientists at this conference to repeat my experiments. This drew the attention of a journalist, Dmitry Starostin, and a note was published by the Russian federal news agency Regnum4. In December 2005, I spoke at a conference “Epigenetics, Transgenic Plants and Risk Assessment” in Frankfurt am Main, Germany. The paper detailing my preliminary results was published in the Proceedings of this conference1. Several papers have subsequently been published in different journals and proceedings. I have submitted a paper to a Russian peer-reviewed journal and am currently preparing other papers for consideration by peer-reviewed scientific journals in English.


COMPETING INTERESTS STATEMENT The authors declare competing financial interests: details accompany the full-text HTML version of the paper at

1. Ermakova, I.V. Influence of genetically modified soya on the birth-weight and survival of rat pups. in Proceedings of the Conference Epigenetics, Transgenic Plants and Risk Assessment, Frankfort am Main, Germany, December 1, 2005 (ed. Moch, K.) 41–48 (Öko-Institut, Freiburg, 2006). < application/octet-stream/download. php?id=277>

2. The Advisory Committee on Novel Foods and Processes. ACNFP statement on the effect of GM soya on newborn rats. ACNFP <> (2005).

3. Brake, D.G. & Evenson D.P. A generational study of glyphosate-tolerant soybeans on mouse fetal, postnatal, pubertal and adult testicular development. Food Chem. Toxicol. 42, 29–36 (2004).

4. Genetically modified soy affects posterity: results of Russian scientists’ studies. Regnum [online] published online 10 December 2005 <>.




Irina V. Ermakova is at the Institute of Higher Nervous Activity and Neurophysiology of the Russian Academy of Sciences, Butlerova Street 5A, Moscow 117485, Russia.

Irina Ermakova, the author of controversial studies reporting soybeans genetically modified for resistance to glyphosate may be dangerous to newborns, agreed to provide details of her work to Nature Biotechnology.



Table 1 Mortality of rat pups by the end of the 3rd week of lactation

Table 2 Comparison of different kinds of chow on rat pup mortalityb

Table 3 Distribution of weights of pups in 2 weeks after birth

Table 4 Examples of absolute values of organ massa in pups 3 weeks after birth

Table 5 Success of mating of first-generation (F1) offspring receiving GM soy

 ((Table data omitted here because of loss of formatting when converting from PDF to Word.))
  To see Tables in pdf format Click Here