GM Free Cymru

GM maize reduces fertility in mice

Citation: Velimirov A, Binter C and Zentek J. (2008)
Biological effects of transgenic maize NK603xMON810 fed in long term reproduction studies in mice.
Report, Forschungsberichte der Sektion IV, Band 3. Institut für Ernährung, and Forschungsinttitut für biologischen Landbau,
Vienna, Austria, November 2008.


The aim of the study was to examine effects of the stacked GM crop NK603 x MON810 in different models of long term feeding studies. So far no negative effects of GM corn varieties have been reported in peer-reviewed publications. But the hypothesis, that effects after long term exposure might become evident in multi-generation studies has rarely been investigated. In this study three designs were used, including a multi-generation study (MGS), a reproductive assessment by continuous breeding (RACB) and a life- term feeding study (LTS), all performed with laboratory mice (strain OF1). The test diets differed only as to the inclusion of 33% NK603 x MON810 corn (GM) versus non- GM corn of a near isogenic line (ISO), both grown under identical conditions in Canada. The MGS also included one group with a non GM corn cultivated in Austria (A REF). All corn varieties used in the MGS and LTS were harvested in 2005, the transgenic and isogenic corn for the RACB were harvested in Canada in 2007. No Austrian corn was used in this case. In the MGS microscopic and ultrastructural investigations were performed to detect changes at the organ and cell level. Gene expression patterns were compared by micro array expression profiles of the intestine as feed-animal interface and by real time PCR. The results of the MGS showed no statistically significant differences concerning parental body mass. The number of females without litters decreased with time in the GM and ISO group, especially in the 4th generation. In the group fed with A REF corn fewer females were without litters, and accordingly more pups were weaned. The production parameters average litter size and weight as well as number of weaned pups were in favour of the ISO group. These differences were also seen in the RACB design and were statistically significant in the 3rd and 4th litters. In addition, the inter-individual variability was higher in the GM group as compared to the other groups. The LTS showed no statistically significant differences in the survival of 3 groups of mice fed the different maize varieties. In the MGS the continuative investigations revealed differences between the GM and ISO groups. The comparison of organ weights did not indicate directed dietary effects, except for kidneys. The electron histological investigation of the cell nuclei revealed differences as to fibrillar centres, dense fibrillar components and the pore density in hepatocytes. This could point to an effect of the GM crop on metabolic parameters. Immunohistochemistry revealed no systematic differences in CD3, CD20 positive cells and macrophages in gut tissue. The microarrays showed differences between the feeding groups. When the data of both non-GM feeding groups from MGS were combined and compared to the GM feeding group, the discrimination became more evident. Analyses of metabolic pathways indicated, that the groups differed regarding some important pathways, including interleukin signalling pathway, cholesterol biosynthesis and protein metabolism. Summarizing the findings of this study it can be concluded, that multi-generation studies, especially based on the RACB design are well suited to reveal differences between feeds. The RACB trial showed time related negative reproductive effects of the GM maize under the given experimental conditions. The RACB trial with its specific design with the repeated use of the parental generation is a demanding biological factor for the maternal organism. Compared to the findings in the RACB trials it can be assumed that the physiological stress was considerably lower in the MGS trial. The trial design of using “new” parental generations instead of continuous breeding with the same generation has to be considered as being obviously less demanding. This might have masked the impact of dietary factors on reproductive performance. However, this part of the experiment is valuable as such because it underlines the need for different experimental designs for the assessment of dietary effects that have an unknown impact on animals. The outcome of this study suggests that future studies on the safety of GM feed and food should include reproduction studies. Physiological and genomic traits and depending on the nature of the genetic modification proteomic and metabolomic methods might be taken into consideration as additional tools to the tests performed in this study.


Commentary by Dr. Mae-Wan Ho

ISIS Press Release 19/11/08 GM Maize Reduces Fertility & Deregulates Genes in Mice Comprehensive long term studies commissioned by the Austrian government reveal that mice fed GM maize produced fewer and smaller litters with many genes affected compared to controls.

Austrian scientists carried out long term studies that showed GM maize fed to mice significantly reduced their fertility over three to four breeding cycles within one generation [1]. Similar effects were found in mice fed GM maize and bred over four generations; although the results did not reach statistical significance in any one generation, the trend was unmistakable, more pups lost and smaller litters in the GM-fed mice.

The studies are by far the most meticulous and comprehensive feeding trials to-date, and confirm deleterious reproductive and health impacts obtained by scientists independent of the biotech industry and farmers’ observations in the field. For a recent review, see [2] GM is Dangerous and Futile (SiS 40).

The new research results are a landmark in the safety assessment of GM food. Most feeding trials were short-term and restricted to a single generation or a single breeding cycle. The “multi- generational” study widely cited as evidence of no long term adverse impacts from GM feed is highly misleading as the experiment did not involve trans-generational feeding, but merely breeding mice that were not GM fed for three generations, and carrying out a separate experiment with GM feed for each generation [3] (Letter to Nature Biotechnology: Systematic bias in favour of no adverse impacts from GM feed, SiS 37). There were other serious flaws in that experiment, not least the failure to ascertain by polymerase chain reaction (PCR) that the processed GM feed used actually contained GM soya.

Targeting long term effects

The new studies were commissioned by the Austrian government several years ago, when it became clear that proper feeding trials were very thin on the ground, and regulators all over the world were largely dependent on companies submitting data that were inadequate and unreliable in many ways, but which they accepted without question [4] (GM Food Nightmare Unfolding in the Regulatory Sham, ISIS scientific publication).

Alberta Velimirov at Research Institute on Biological Agriculture (Forschungsinttitut für biologishen Landbau) and Claudia Binter and Jürgen Zentck at the Institute for Nutrition (Institut für Ernährung) both in Vienna, monitored changes in gross morphology, reproductive performance, and sub-microscopic analyses of tissues and cells, as well as gene expression. (They found no immunochemical or other microscopic differences in the tissues.)

Three series of experiments were done. The first was a multigeneration feeding trial in which the mice were fed and bred for four successive generations, beginning with the F0 parents that were fed on the diets from birth. The second was a multi-cycle breeding trial lasting 20 weeks in which breeding pairs of mice were fed beginning 1 week prior to co-habitation until the end of experiment, and allowed to go through four breeding cycles in the same generation. The third was a life-term trial involving feeding the mice without breeding from conception (via the pregnant mothers) to their eventual death.

A laboratory non-inbred strain of mice was used for all experiments, in order to avoid the effects of inbreeding, so that the results would be more generally applicable to natural populations.

The researchers report that it was not possible to obtain a GM test crop plus parental line from the agro-business companies, which was why the test diets consisting of 33 percent GM maize had to be compared with a non-GM maize variety (also at 33 percent) that was closely related to the GM maize. Both were grown under identical conditions in the Organic Agriculture Centre of Canada in Nova Scotia, in 2005 and 2007. The GM maize was the transgene hybrid NK603 x MON810 containing three gene cassettes, two conveying glyphosate herbicide tolerance and one insect resistance coding for endotoxin Cry1Ab. The transgenic protein was estimated to be 0.11-0.24 microgram per gram of fresh grain.

The multigeneration study also included one group with a non-GM maize variety cultivated in Austria.

The herbicides dicamba, atrazine and s-metalochlor were used with non- GM Maize, while glyphosate only was used with the GM maize. Contamination levels of the maize grains with herbicides were determined to be less than 0.01 percent of each of the herbicides. This was important to make sure that effects due to herbicides were not confounded with those from the GM feed.

Main effects on reproduction

In the multigeneration study, the parental generation was fed since birth with either GM or nonGM maize diet, and 4 generations were bred. Less pups were born in successive generations in both control and GM fed mice. But the controls tended to do better than GM fed. The average litter size and weight as well as number of weaned pups were in favour of the non-GM maize group. None of the differences reached statistical significance in any one generation, although the trend was clear.

Over all generations, about twice as many pups were lost in the GM group as compared with the control group (14.59 percent vs 7.4 percent). More litters with 8 or more pups were seen in the control compared with GM group. And a greater number of pups were lost at weaning in the GM fed.

Comparison of organ weights did not indicate direct dietary effects in the multigeneration study, except for the kidneys. Kidney weight of females in the GM-fed group were significantly lower in the F2, F3 and F4 generations than controls; and males in the GM-fed group also had significantly lower kidney weight than controls in the F2 generation

The electron microscope investigations revealed differences in the liver cells indicative of reduced core metabolism in the GM-fed mice. In addition, DNA microarray analyses showed important differences in gene expression between both groups fed non-GM maize and the group fed GM maize.

In the multi-cycle breeding trial, the same differences between GM- fed and controls were evident. and reached statistically significant levels in the 3rd and 4th litters. There were clearly fewer and smaller litters in the GM-fed mice.

The average number of pups born was always lower in the GM fed but did not reach statistical significance before the 3 rd and 4th deliveries. The number of pups at weaning was also always smaller in the GM-fed group. Over all the deliveries, more pups were born in the controls than in the GM group (1035 vs 844).

Consistent with these findings, the life-term feeding trial showed no significant differences in the average life-span of the GM-fed mice compared with controls.

Epigenetic effects of GM maize feed

In the F3 generation of the multigeneration trial, DNA microarray analyses were performed on the lower small intestine. This identified 2 374 genes that were significantly abnormally expressed in GM fed compared with non GM fed mice; with 421 of these showing a 2-fold or greater change from controls. This was more than 3.2 percent of the total 13 034 genes expressed in the lower small intestine. The reproductive and other effects observed could be just the tip of the iceberg as far as the epigenetic changes are concerned. The impacts could take more generations of GM feeding to become fully manifest.

The genes were functionally classified and found to predominate in the pathways of protein biosynthesis and protein metabolism and modification, interleukin signalling and cholesterol biosynthesis.

Epigenetics – the study of heritable changes in gene expression that do not involve alterations in DNA sequence – is a maturing discipline with growing applications in toxicology, cancer, nutrition, and brain and behavioural sciences [5-8]. A change in our diet as far-reaching and profound as GM food cannot be entertained without detailed long term studies of the kind carried out by the Austrian scientists together with analyses using DNA microarrays, proteomics, and metabolic profiling, which are now routine in laboratory and field studies.


1. Velimirov A, Binter C and Zentek J. Biological effects of transgenic maize NK603xMON810 fed in long term reproduction studies in mice. Report, Forschungsberichte der Sektion IV, Band 3. Institut für Ernährung, and Forschungsinttitut für biologischen Landbau, Vienna, Austria, November 2008.

2. Ho MW. GM is dangerous and futile. Science in Society 40 (in press).

3. Ho MW. Letter to Nature Biotechnology, systematic bias in favour of finding no adverse impacts of no adverse impacts from GM feed. Science in Society 37, 10, 2008.

4. Ho MW, Cummins J and Saunders PT, GM food nightmare unfolding in the regulatory sham. Microbial Ecology in Health and Disease 2007, 19, 66-77.

5. Newbold RR, Padilla-Banks E and Jefferson WN. Adverse effects of the model environmental estrogen diethylstilbestrol are transmitted to subsequent generations. Endocrinology 2006, 147, S11- S17.

6. Reamon-Buettner SM, Mutschler V and Borlak J. The next innovation cycle in toxicogenomics: environmental epigenetics. Mutation Res 2008, 639, 158-65.

7. Weidman JR, Dolinoy DC, Murphy SK and Jirtle RL. Cancer susceptibility: epigenetic manifestation of environmental exposures. Cancer J 2007, 13, 9-16.

8. McGowan PO, Meaney MJ and Szyf M. Diet and the epigenetic (re) programming of phenotypic differences in behaviour. Brain Research 2008, 1237, 12-24.