Maria Masucci and Nico Dantuma reply

In response to Prof. Magnani, it is well established that the ubiquitination cascade exhibits a hierarchical organization, where a single ubiquitin-activating enzyme (E1) interacts with a large family of ubiquitin-conjugating enzymes (E2), which in turn cooperate with even larger groups of ubiquitin ligases (E3). Furthermore, current knowledge suggests that substrates targeted for degradation converge to a common proteolytic machine composed of the catalytic core of the proteasome (20S) and different regulatory subunits1. Thus, the reporters described in Dantuma and colleagues2 may allow the identification of inhibitors acting at the early stages of ubiquitination, by selective inhibition of E1, as well as downstream of ubiquitination, including protein unfolding and proteolysis. It is true that the reporters will not detect inhibitors specific for a multitude of E2s or E3s. Yet, reporters recognized by different families of ubiquitin–protein ligases would suffer the same type of limitations—being of no use in the identification of inhibitors targeting individual enzymes. The easy establishment of stable GFP transfectants in numerous model systems suggests that the long-term toxicity of GFP in transgenic animals has no direct bearing on the use of the reporters in screening assays.

The accumulation of processed GFP in HeLa cells expressing Ub-M-, Ub-L-, or Ub-R-GFP confirms that these chimeras are efficiently recognized by ubiquitin hydrolases and are therefore bona fide N-end rule substrates in vivo. The observation that similar chimeras are recognized as UFD substrates in the slime mold Dictyostelium discoideum does not detract from their applicability in mammalian cells. Finally, it seems unlikely that highly conserved recognition strategies, such as the N-end rule and UFD signals, would have limited biological significance. We hope that the reporters described in Dantuma et al. will be useful also in dissecting this conundrum of regulated proteolysis.