Improved detection of microbial risk of releasing genetically modified bacteria in soil by using massive sequencing and antibiotic resistance selection.

High-throughput 16S rRNA gene-targeted pyrosequencing was used with commonly used risk assessment techniques to evaluate the potential microbial risk in soil after inoculating genetically modified (GM) Corynebacterium glutamicum. To verify the risk, reference experiments were conducted in parallel using well-defined and frequently used GM Escherichia coli and wild-type strains. The viable cell count showed that the number of GM bacteria in the soil was reduced to below the detection limit within 10 days, while the molecular indicator for GM plasmids was detected throughout the experiment by using quantitative real-time polymerase chain reactions. Subsequent pyrosequencing showed an insignificant influence of the GM bacteria and/or their GM plasmids on the structure of the soil bacterial community this was similar to non-GM wild-type strains. However, pyrosequencing combined with kanamycin-resistant bacteria selection uncovered a potential risk of GM bacteria on the soil bacterial community and pathogens. The results of the improved methodology showed that the microbial risk attributable to GM C. glutamicum was relatively lower than that attributable to the reference GM E. coli.