ACNFP additional data on the nutritional impact assessment of T25 maize

1. In the original ACNFP safety assessment of T25 maize, it was requested that seed composition be monitored over time to demonstrate the stability of the line and to determine any possible effect of genetic drift on the plant’s metabolism.

2. In September 2002, Bayer CropScience (previously Aventis) provided the ACNFP with these data. The Committee considered them by post.

3. The data represent compositional and nutritional analysis using raw agricultural commodity grain generated from 15 field sites over a two-year period. Three different treatment groups of maize were used in the analysis: 1) non-transgenic, unsprayed, 2) transgenic, unsprayed and 3) transgenic, sprayed with Liberty herbicide. A total of 135 samples were taken from each site, with 92 components being analysed in each sample.

4. Statistical analysis was performed by analysis of variance and a 95% confidence interval. In each case, the first statistical comparison carried out was between the three treatment groups grown within the same site (site-by-site analysis). Following this, comparisons were made with data collected from across all 15 sites (over-all-sites analysis).

5. To determine compositional equivalence between the transgenic line and the non-transgenic line, pooled over-all-sites data were compared to published reference ranges. The non-transgenic line was also used as a reference value since it provides the most accurate control, being the same variety of maize grown in the same field trial as its GM counterpart.

Proximate analysis (moisture, fat, protein, fibre, ash and carbohydrates)

6. Homogeneity within the non-transgenic, unsprayed reference line grown across all sites, could not be demonstrated for proximate components.

7. For most of the 15 sites, site-by-site analysis showed no significant differences between the three treatment groups for moisture, fat, protein, ash, total and available carbohydrates.

8. There were statistically significant differences at several sites in the level of total dietary fibre in the sprayed, transgenic line. However, these differences were not seen in the over-all-sites analysis using pooled data.

9. With the exception of moisture and total dietary fibre, the levels of the proximate components across all three treatment groups fell within published reference range. The three treatment groups had moisture levels averaging 27.4% (non-transgenic), 27.5% (transgenic, non-sprayed) and 27.1% (transgenic, sprayed), compared with the published range of 7-23%. For total dietary fibre, only one published reference figure was available.

Minerals

10. The analysis of minerals in the non-transgenic, unsprayed reference line did not show homogeneity across all of the sites.

11. With the exception of chloride, consistently higher levels of minerals were found in the sprayed transgenic lines when compared to the reference group within the same site. This was also the case when the sprayed and unsprayed transgenic lines were compared, thus indicating that this was not a result of the genetic modification event.

12. Using pooled data from all 15 sites, no statistically significant differences were found for the minerals phosphorous, potassium, magnesium, manganese, zinc, chloride, calcium and iron between the three treatment groups. Higher levels of copper were present in the transgenic lines: 1.8 mg/kg and 2.0 mg/kg respectively in the unsprayed and sprayed transgenic groups, compared with 1.5 in the reference group. However, all three groups were within the published range of 0.8-10 mg/kg.

13. Most of the levels of minerals found in the three treatment groups fell within published reference ranges, except for magnesium and calcium. The levels of calcium were significantly lower than those reported in the published reference ranges, but were consistent within the three treatment groups, including the non-transgenic line.

Vitamins

14. The levels of most of the 11 vitamins measured in the study revealed considerable variation within each site for all three treatment groups.

15. Although there were some statistical differences when individual vitamins were compared between the three treatment groups grown at individual sites, these were not consistent across all sites, and no uniform pattern was evident.

16. Across all sites, no statistically significant differences were seen between the transgenic and reference groups for 10 of the 11 vitamins, namely vitamin B1, vitamin B2, Niacin, pantothenic acid, folic acid, alpha tocopherol, beta tocopherol, gamma tocopherol, alpha tocotrienol vitamin E activity.

17. The mean level of delta-tocopherol was significantly lower in the non-transgenic group than in the unsprayed transgenic group (0.16 and 0.18mg/100g dry matter respectively). The value for the sprayed transgenic group was intermediate.

18. The levels of all the vitamins in the pooled over-all-sites data fell within published reference ranges, with the exception of folic acid. The levels of folic acid in all three treatment groups were higher than the upper figure for the published reference range for this vitamin, with the highest level being evident in the non-transgenic unsprayed reference group. However, there were no significant differences in folic acid between the non-transgenic and two transgenic groups, in either the site-by-site or the over-all-sites analyses.

Anti-nutrients (phytic acid)

19. The phytic acid levels were very variable in all three treatment groups and the non-transgenic reference group did not show homogeneity across all of the sites. A slightly higher level of phytic acid was found in the transgenic samples at a number of the individual sites. However, no statistically significant differences were found between the three treatment groups in the over-all sites analysis.

20. The levels of phytic acid in the over-all-sites pooled data were found to fall within the published reference ranges.

Total amino acids (after hydrolysis)

21. Homogeneity of the levels of total amino acids was demonstrated for the non-transgenic reference group across all sites. The site-by-site analysis showed no differences for all the amino acids for the three treatment groups.

22. Although no significant differences were observed among the field data, small differences were observed in 7 of the 18 amino acids included in the analyses when compared with published reference ranges. In all cases, the differences were represented across all three treatment groups, and were not considered to be significant.

Free amino acids

23. The site-by-site analysis did not show homogeneity in the non-transgenic reference group for these components, with the exception of asparagine, tryptophan and the sum of all 18 free amino acids.

24. No consistent differences were seen in the statistical analysis carried out on the transgenic plants either in the site-by-site or over-all-sites analyses.

25. No published reference ranges are available for free amino acids, and so a comparison was not possible.

Total fatty acids (after hydrolysis)

26. Homogeneity in the non-transgenic reference group could not be established for total fatty acids across all of the sites.

27. Significant differences between the three treatment groups were only observed for 2 of the 10 fatty acids, namely palmitoleic acid and eicosenoic acid. The differences between treatment groups were mixtures of increases and decreases depending on the site and no consistent changes were observed. The three treatments can be considered equivalent with respect to the over-all-sites analysis.

28. Total fatty acids in the non-transgenic and transgenic lines were in good correspondence with published reference ranges, where these were available.

Free fatty acids

29. Levels of 12 free fatty acids were analysed in all of the samples but 5 of these were consistently below the limit of detection. The remaining 7 – myristic, palmitic, palmitoleic, oleic, linoleic, linolenic and behenic acids – were detectable in only a minority of the samples and at very low levels. Where statistical comparisons were possible, there was no uniform tendency in the differences between the three treatment groups. No published reference values were available for free fatty acids.

30. As anticipated, the results of this study have demonstrated that there is considerable variability in the composition of crops grown on different sites or at different times.

31. Most of the values obtained from the transgenic and non-transgenic lines in this study fall within the published reference values. Where they do not, this may either reflect differences in the measurement method or the limited range of the analyses reported in the literature.

32. The Committee is content that the data presented demonstrates stable inheritance of the introduced trait and do not suggest that genetic changes are affecting the modified plant’s metabolism.

33. The Committee is satisfied that the majority of components measured in the T25 seed fall within standard reference ranges for maize and that any statistically significant differences found between the three different test groups are not of biological significance when viewed in the context of the normal ranges.

February 2003