We monitored pollen-mediated transgene dissemination from commercial transgenic squash CZW-3 into its wild relative Cucurbita pepo ssp. ovifera var. texana (C. texana). Transgenic squash CZW-3 expresses the neomycin phosphotransferase II (nptII) gene and the coat protein (CP) genes of Cucumber mosaic virus (CMV), Zucchini yellow mosaic virus (ZYMV), and Watermelon mosaic virus (WMV); thereby, it is resistant to these three aphid-borne viruses. The rate of NPT II and CP transgene introgression increased with overlapping flowering patterns and a high ratio of transgenic F1 hybrids (C. texana × CZW-3) to C. texana. Transgene transfer also readily occurred from transgenic F1 hybrids into C. texana over three generations in field settings where test plants grew sympatrically and viruses were not severely limiting the growth, and fruit and seed production of C. texana. In contrast, introgression of the transgenes into C. texana was not sustained under conditions of high viral disease pressure. As expected, C. texana progeny that acquired the CP transgenes exhibited resistance to CMV, ZYMV, and WMV. This is the first report on transgene dissemination from a transgenic crop that exhibits disease resistance and hybridizes with a wild plant species without loss of fertility.

We compared some fitness components of the wild squash species Cucurbita pepo spp. ovifera var. texana (C. texana) and three generations of hybrids (F1, BC1, and BC2) between C. texana and commercial transgenic squash CZW-3 over three consecutive years under field conditions of low (LDP) and high disease pressure (HDP) by Cucumber mosaic virus (CMV), Zucchini yellow mosaicvirus (ZYMV) and Watermelon mosaic virus (WMV). Transgenic squash CZW-3 expresses the coat protein (CP) genes of CMV, ZYMV, and WMV, and is resistant to these three aphid-borne viruses. Across all HDP trials, transgenic BC1 and BC2 hybrids expressing the three CP genes grew more vigorously, displayed resistance to CMV, ZYMV, and WMV, and produced a greater number of mature fruits and viable seeds than nontransgenic hybrid segregants and C. texana. Transgenic F1 hybrids behaved similarly to BC1 and BC2 hybrids but grew less vigorously than C. texana. In contrast, across all LDP trials, C. texana outperformed the transgenic and nontransgenic hybrid segregants. Further, only one back cross was necessary to recover individuals with most of the C. texana characteristics and yet maintain virus resistance. Our data suggest that C. texana acquiring CP transgenes upon hybridization and introgression could have a selective advantage if CMV, ZYMV, and WMV are severely limiting the growth and reproductibility of wild squash populations.