Field-based validation of a citrandarin buffer system for gene flow containment in genetically modified sweet orange trials

Containment measures to suppress the dispersion of transgenes from plots of genetically modified (GM) citrus into the progeny of surrounding non-GM citrus trees were evaluated under open-field conditions using border plants as a buffer zone. The border consisted of two plant rows each of citrandarin, selected for its superior pollination ability; Clemenules clementine, selected for its monoembryony combined with self-incompatibility and its efficiency as a pollen-mediated gene flow monitoring genotype; and Pera sweet orange, a polyembryonic commercial genotype used to estimate GM seedling frequency, arranged sequentially from the inner GM plot outward. Four assays with GM citrus plots were evaluated over two to four consecutive years for pollen flow, summing 12 seasons. Pollen compatibility among sweet orange cultivars, citrandarin, and clementine flowers was assessed using controlled crosses. Gene flow was assessed by the presence of neomycin phosphotransferase II (nptII) transgene in seedlings from clementine, Pera sweet orange border plants, as well as from clementine control trees located inside the plots. Trifoliate phenotype and microsatellite markers in clementine seedlings were used as indicators for paternity analysis. Transgene frequency in Pera sweet orange border plants was only 0.31%, meanwhile transgene frequency at the clementine border was lower (6.05%) than that in the controls within the plot (24.09%), working as a mitigating strategy to reduce gene flow. Overall, the superiority of citrandarin pollination and fertilization in comparison to other citrus genotypes, the distance among the genotypes indicated by the physical proximity between pollen-receptor and citrandarin, and the genetic barriers of facultative apomixis and self-compatibility in sweet orange genotypes strongly limits, but does not prevent GM pollen dispersion, favoring fertilization from citrandarin instead of the GM sweet orange genotypes into Clemenules.