Molecular Breeding 1: 309-317, 1995. © 1995 Kluwer Academic Publishers. Printed in Belgium. 309 Transgenic broccoli expressing a Bacillus thuringiensis insecticidal crystal protein: implications for pest resistance management strategies Timothy D. Metz 1,2, Richard T. Roush 3, Juliet D. Tang 4, Anthony M. Shelton 4 and Elizabeth D. Earle 1,, i Department of Plant Breeding, Cornell University, Ithaca, NY 14853-1902, USA (*author for correspondence)," z Current address: Department of Biological Sciences, Campbell University, Buies Creek, NC 27506, USA," 3Department of Entomology, Cornell University, Ithaca, N Y 14853, USA," 4Department of Entomology, New York State Agricultural Experiment Station, Geneva, NY 14456, USA Received 11 July 1994; accepted in revised form 30 March 1995 Key words: Bacillus thuringiensis, Brassica oleracea, diamondback moth, Plutella xylostella, resistance management, transgenic broccoli Abstract We used Agrobacterium tumefaciens to transform flowering stalk explants of five genotypes of broccoli with a construct containing the neomycin phosphotransferase gene and a Bacillus thuringiensis (Bt) gene [CryIA(c) type] optimized for plant expression. Overall transformation efficiency was 6.4~o; 181 kanamycin-resistant plants were recovered. Of the 162 kanamycin-resistant plants tested, 112 (69~o) caused 100 ~o morality of 1st-instar larvae of a Bt-susceptible diamondback moth strain. Southern blots of some resistant transformants confirmed presence of the Bt gene. Selected plants that gave 100~ mortality of susceptible larvae allowed survival of a strain of diamondback moth that had evolved re- sistance to Bt in the field. F 1 hybrids between resistant and susceptible insects did not survive. Analy- sis of progeny from 26 resistant transgenic lines showed 16 that gave segregation ratios consistent with a single T-DNA integration. Southern analysis was used to verify those plants possessing a single T-DNA integration. Because these transgenic plants kill susceptible larvae and F 1 larvae, but serve as a suitable host for resistant ones, they provide an excellent model for tests of Bt resistance management strategies. Introduction The soil-borne bacterium Bacillus thuringiensis (Bt) produces a family of proteins with insecti- cidal activity [20]. Five classes of insecticidal crystal proteins (ICPs) have been identified, each with activity against one or two orders of insects. The range of activity so far identified extends to many Lepidopteran, Coleopteran and Dipteran insects. Bt ICPs have been used in conventional insecticidal sprays since the 1950s [30]. Genes encoding some of the Bt ICPs have recently been cloned and transferred to crop plants to reduce the need for spraying and to increase the activity and persistence ofBt ICPs [5, 12, 22, 28, 37, 45]. Unfortunately, some insect species have devel- oped resistance to Bt ICPs in laboratory selection experiments [25, 36, 38, 45], and one insect, the