Study on unintended metabolic interactions in genetically modified soybean with stacked events / Salehuddin Asyraf Mohd Najib

Malaysia imports a large amount of maize and soybean, included genetically-modified
(GM) varieties approved by the National Biosafety Board (NBB) for the purpose of
food, feed and processing (FFP). Continuous breeding by seed developers has led to a
large variety of GM maize and soybean with multiple transgene inserts (‘stacked
events’). This pose a new challenge to the risk assessment of these GM crops – the
presence of two or several transgenes in one organism can lead to unintended and
unpredicted effects and interactions, which may not be captured by the current data
requirements. Proponents of biosafety regulation has long requested for the inclusion
of metabolomics profiling data, which can provide a snapshot of a large number of
metabolites, in the risk assessment of stacked events. Such ‘metabolic fingerprints’
techniques can technically be used to compare non-GM and stacked events GM crops
and to uncover irregularities in the metabolism. This study was proposed to investigate
the suitability of using metabolomic profiling to detect unpredicted effects when
transgenes are stacked together, and to evaluate if such data will add value to the risk
assessment process. Experiments were designed to eliminate as many confounding
factors as possible by constructing and maintaining the sample materials under similar
laboratory environment. Using standard molecular biology and plant tissue culture
techniques, GM corn and soybean were constructed carrying the Pat and Cry1Ab
transgenes, singly and in stack. Despite the success in developing the transgenic maize
calli, none of them were able to develop into full leaves and thus were terminated at this
point. On the other hand, it was found that while the presence and levels of certain
metabolites are changed in non-GM, single GM and stacked GM of soybean samples,
the overall metabolomic analysis is not able to clearly differentiate the metabolome
profiles. Principle component analysis (PCA) of the data indicate that the first two
components could at most explain for 25% of the variation between the 4 categories.
Observations indicate that the inherent biological variability within and between the
samples are large and tend to eclipse any variations induced by the genetic
modifications. Thus, metabolomics data do not add value to the risk assessment of GM
stacked events at this stage of technology.