Assessment of the fate of plant DNA fragments in the ruminant digestive tract

The fate of transgenic DNA from glyphosate-tolerant (Roundup Ready RTM; R) canola upon ingestion by ruminants was studied in comparison to native canola DNA (i.e., Rubisco gene). Whole and cracked seeds, canola meal, and diets containing canola meal, both from R and a near-isogenic parental line, were used in examining the effects of feed processing and exposure to mixed ruminal culture and duodenal fluid (DF). DNA fragments up to 23 Kbp and the entire 1363-bp cp4 epsps transgene from RR canola were detectable throughout feed processing. The presence of transgene fragments was determined using conventional and real-time PCR and nine primer pairs amplifying gene fragments of 62 to 527 bp, as well as the whole transgene. No evidence of transfer of transgenic fragments to ruminal microorganisms was found. Ruminal persistence of transgene fragments and of Rubisco from R or parental canola were indistinguishable, both associated with plant debris, and inversely related to degree of feed processing. The fate of DNA transported to the duodenum in intact plant cells was studied in vitro. Free DNA was rapidly degraded in DF at neutral pH (simulating distal duodenum), suggesting that uptake into tissues via Peyer's patches is unlikely. Transgene was not detected in blood or feces of sheep fed RR canola in forage- or concentrate-based diets, nor was DNA transfer to duodenal microbes detected. Dietary fibre level (i.e., passage rate) did not affect persistence of ingested DNA. In all respects studied, transgenic DNA was indistinguishable from endogenous DNA. Inclusion of data from testing the digestive fate of transgenes present in R canola was suggested to be unwarranted as part of the feed/food safety assessment. Real-time PCR used to quantify transgene fate and persistence revealed a potential correlation with feed degradation. Thus, the technique was used in developing a novel method for estimating dry matter disappearance for individual feed components in a mixed ration. The method was relatively accurate, yielding a correlation of 0.87 between predicted and actual dry matter disappearances, and for the first time, allowed direct quantification of individual plant digestion in a mixed ration.