Send

Effects of Bt maize containing three Bt proteins on micro-organisms in the soil

(2008 – 2011) Johann Heinrich von Thünen-Institut (vTI - the Federal Research Institute for Rural Areas, Forestry and Fisheries), Institute of Agricultural Ecology

Topic

In the field, maize plants provide a possible habitat for a wide range of micro-organisms (bacteria, fungi and archaea) both inside their roots and on the root surface. The aim of this project was to compare the diversity of the naturally occurring micro-organism communities in the genetically modified Bt maize line MON89034 x MON88017 with the diversity in conventionally bred maize varieties. This GM line produces three different Bt proteins (Cry1A.105, Cry2Ab and Cry3Bb1), which can be measured in the soil using specific highly sensitive detection systems.

The project aimed to measure the quantities of the three Bt proteins in the roots and soil over the course of the three-year field trial to investigate any correlations between potential ecological effects and concentrations of Bt proteins. In addition, the project investigated the impacts of Bt proteins on the diversity of micro-organisms in the intestines of bees.

Summary

The concentrations of Bt proteins in 2008 and 2009 were comparable. They were somewhat lower in 2010 because of drought stress. Higher levels of the Bt proteins Cry1A.105 and Cry2Ab2 per gram of biomass were found in the fine roots than in the large roots. No such difference was observed for Cry3Bb1. The Bt quantities that entered the soil via the roots were extremely small. The Bt proteins did not survive in the soil until the next growing season.

Bt proteins were also found in the bee intestines, but the quantities were very small. The Bt proteins were broken down quickly.

The micro-organism communities in and on the maize roots and in the rhizosphere were specific for each of the four maize variants. The differences between the Bt maize variety and the control varieties lay within the range of natural variability that occurs between conventionally bred varieties.

After feeding with pollen, varietal differences were found in the bacterial community in the bee intestines. The differences between the Bt variety and the control varieties were, however, no greater than the differences between conventional maize varieties.

No differences were found in the colonisation density of the micro-organisms in the rhizosphere or in the bee intestines. There were therefore no indications that the Bt proteins have a negative effect on micro-organisms either in the ‘root and rhizosphere’ habitat or in the ‘bee intestine’ habitat.

Experiment description

Levels of Bt proteins in the soil of a field of GM maize were measured and their impact on microbiological communities was investigated.

The genetically modified variety, its isogenic parent variety and two conventional control varieties were grown as part of a field trial. This meant that it was possible to differentiate between potential effects caused by the Bt proteins and conventional varietal effects. In addition, the isogenic variety in one part of the plot was treated with an insecticide so that the effect of conventional pest control methods could also be included in the study.

Root samples and samples from the soil in the immediate vicinity of the roots (rhizosphere) were taken during the maize-flowering period because the level of Bt protein produced by the plants is relatively high at this time. These samples formed the basis for the study of the micro-organism communities (bacteria, fungi and archaea) and for measuring the Bt protein levels. In addition, the Bt concentrations were measured in the plant remains after the maize harvest and before the start of the next growing season. Unlike earlier projects, the experiments were conducted on individual plants rather than by creating mixed samples. This increased the detection sensitivity. Moreover, fine and large roots were studied separately. Here too, the aim was to achieve a more sensitive measurement of impacts of Bt proteins on micro-organisms.

As well as investigating the impacts of the three Bt proteins on soil micro-organisms, the project studied the bacterial communities in the mid- and hindgut of honeybees (in collaboration with the University of Würzburg). As a pollinator of a large number of plants, the honeybee (Apis mellifera) is of great ecological and economic significance and is therefore considered to be one of the most important non-target organisms when considering the potential risks of GM plants. Since bacterial gut flora are vital for the health both of individual bees and of the entire colony, the project studied bees that were kept in special tents on the field during the maize-flowering period of 2009 and were only able to collect Bt maize pollen. As a comparison, the project also studied bees that had collected only pollen from the isogenic parent variety and from a conventional variety, and free-flying bees from a hive located next to a field of Phacelia.

Quantifying Bt proteins

As part of the ELISA method, marked antibodies were used to measure the concentration of the Bt proteins in the roots, in the rhizosphere and in bee intestines. A specific detection system was used for each of the three proteins formed (Cry1A.105, Cry2Ab2 and Cry3Bb1) that was capable of detecting Bt proteins at concentrations of 5 nanograms per gram of root and 0.1 nanograms per gram of soil.

Characterising the micro-organism communities in and on maize roots and in bee intestines

Culture-independent methods were used to measure the structural diversity of the microbial communities because a large proportion of the micro-organisms do not grow on conventional culture media. A number of different molecular biological methods were used. The T-RFLP technique (Terminal Restriction Fragment Length Polymorphism) was used to create genetic profiles (’finger prints’) of the bacterial, fungal and archaeal communities. It was used to analyse certain ribosomal genes (SSU rRNA genes) because they are found in all organisms but have different DNA sequences depending on the degree of relationship.

A number of molecular-biological methods are used to determine the structural diversity of the microbial communities. The T-RFLP technique (Terminal Restriction Fragment Length Polymorphism) is used to create genetic profiles (’fingerprints’) of the bacterial and fungal communities. It analyses certain genes (ribosomal SSU rRNA genes) that are found in all organisms but which have different DNA sequences depending on the degree of relationship.

In order to measure and identify the complexity of the bacterial communities in the rhizosphere, this biosafety research project for the first time made use of the high-throughput sequencing method (next generation sequencing). With this method it is possible to make much more robust statements about the composition of the bacterial communities based on over 600,000 SSU rRNA genes.

The genes used to create the genetic profiles are also used to determine the population density (abundance) of the micro-organisms. Quantitative PCR methods (real-time PCR) are used to measure the number of copies of these genes. The results enable researchers to draw conclusions about the number of micro-organisms in the sample.

Investigating the chemical soil characteristics of the rhizosphere

The soil from the rhizosphere was analysed in terms of its nitrogen and sulphur levels, pH value and total carbon so that these parameters could also be correlated to any changes in the micro-organism communities.

Statistical analyses

Cutting-edge bioinformatic methods were used to evaluate the microbial diversity. Using various statistical and, in particular, multivariate analyses, it is possible to assess the importance of various factors, such as soil parameters or Bt concentrations, for the response of the micro-organism communities. The various parameters can be correlated, making it possible to assess whether there is a genetic engineering effect on the micro-organisms, or whether the effect lies within the range of normal variation for conventionally bred varieties.

Results

Quantifying Bt proteins

Three highly sensitive systems were set up to selectively detect different Bt proteins in root and rhizosphere extracts. The protein concentrations in the samples from all three years were measured.

Cry protein levels in the roots (top) and in the rhizosphere (above) in 2009

The concentrations of Bt proteins in 2008 and 2009 were comparable. They were somewhat lower in 2010 because of drought stress. The 2009 samples were found to contain an average of around 15 micrograms of Cry1A.105, 50 micrograms of Cry2Ab2 and 60 micrograms of Cry3Bb1 per gram of root. The Cry1A.105 and Cry2Ab2 levels were higher in the fine roots than in the large roots. No such differences were found for Cry3Bb1. The rhizosphere was found to contain less than 0.5 nanograms of Cry1A.105 and Cry3Bb1 on average per gram of soil. The Cry2Ab2 levels in the soil were below or close to the detection limit of 0.1 nanograms per gram.

In bees, the levels of Cry1A.105 and Cry3Bb1 in the mid- and hindgut were investigated. On average, the midgut was found to contain 0.7 nanograms of Cry1A.105 and 0.4 nanograms of Cry3Bb1 per gut segment. The hindgut was found to contain 1.0 nanogram of Cry1A.105 and 0.3 nanograms of Cry3Bb1 per gut segment. It should be noted here that the hindgut contained significantly more pollen than the midgut. Comparable quantities of Bt proteins were found in individual hindgut samples of bees that had definitely not come into contact with Bt maize pollen. This can be explained by the sporadic appearance of naturally occurring Bacillus thuringiensis.

In general, the three-year trial showed that the quantities of Cry1A.105 and Cry3Bb1 proteins entering the soil via the roots were extremely small and that they did not survive in the soil until the next growing season. The quantity of Bt proteins (Cry1A.105 and Cry3Bb1) in the bee intestines was also very low. Calculations based on the total quantity of Bt proteins released from the ingested Bt maize pollen show that the Bt proteins are broken down quickly in the bees’ intestines. So far there are no indications of particular impacts caused by the Bt maize varieties studied in this project on soil micro-organisms or bacteria in the intestines of honeybees.

Characterising the micro-organism communities in and on maize roots and in bee intestines

The T-RFLP technique (see above) was used to establish genetic ‘fingerprints’ to show the bacterial diversity in the roots (endophytes) and in the bee intestines, and the bacterial, fungal and archaeal diversity in the rhizosphere.

The endophytic bacteria in the fine roots displayed significant differences between the different maize varieties in almost all cases. However, no difference was found between the isogenic maize and the Bt variety, probably because of how closely they are related. There were significant differences between the maize varieties in the large roots as well. Treatment with tefluthrin, an insecticide, had no impact on the diversity of the micro-organisms.

The diversity of the bacterial and fungal communities in the rhizosphere was also very specific for each of the four varieties. The differences between the Bt maize variety and the control varieties lay within the range of natural variability for conventional varieties. There were significant differences between large and fine roots for both the bacterial and fungal communities, which also highlighted the sensitivity of the detection system. The archaeal communities were less diverse and presented a very similar composition in all years and for all varieties.

The bacterial communities in the rhizosphere were analysed in detail using high-throughput sequencing. Calculations showed that 99 per cent of the bacterial diversity present was identified. All maize varieties had a significant effect on the frequency of individual species of bacteria. According to the findings to date from ongoing bioinformatic analyses, the Bt maize showed no greater difference from the conventional varieties than these varieties displayed between themselves.

In the bees there were characteristic bacterial communities in the mid- and hindgut and clear varietal differences. The differences between the Bt variety and the control varieties were, however, no greater than the variability between conventional varieties or between maize and Phacelia.

Quantitative analyses showed no differences in the colonisation density of the micro-organisms in the rhizosphere or in the bee intestines. There were therefore no indications of a negative effect caused by the Bt proteins on the micro-organisms.

An analysis of the data does not provide any indications of specific effects on micro-organisms in the soil or on bacterial gut flora in bee intestines caused by the Bt maize under investigation.

Investigating the chemical soil characteristics of the rhizosphere

The soil parameters measured in the rhizosphere – total carbon, total nitrogen, total sulphur and pH value – were within the normal range for typical agricultural soils. The heterogeneity of the field was very low for these parameters. Multivariate statistical analyses showed only a low impact (approx. 5%) on the variability of the structural diversity of the bacterial communities. There was no indication that the effect of the Bt maize on the soil communities assessed here is any different from the effects of conventionally bred maize.