Fig. 5. Polarized degradation of Cactus-lacZ fusion proteins does not rely on endogenous Cactus. Enzymatic activity was used to determine Cactus-lacZ distributions in the progeny of cact^BQ/cact^E10 females, expressing endogenous mutant Cactus that does not undergo regulated degradation (Bergmann et al., 1996). While the 101Cactus-lacZ variant undergoes polarized degradation (A), the 125Cactus-lacZ variant does not (B). Similarly, 101Cactus-lacZ restored ventral and lateral elements (D) to the cuticles of embryonic progeny of cact^BQ/cact^E10 females (C), while 125Cactus-lacZ did not (E). Two-hybrid prey constructs containing full-length, N terminally deleted and C terminally deleted segments of Cactus fused to a bacterial activator segment in plasmid JG4-5 (Gyuris et al., 1993) were constructed (for extent of Cactus present in each construct, see G) and tested for their interaction with a lexA-Cactus bait (F). An interaction between Pelle and Pellino is shown as a positive control, while empty vector (JG4-5) and Tube and Pellino fusions to the activation domain fail to interact with lexA-Cactus and represent negative controls.