Caspases play in traffic

The unfolding of apoptosis involves the orchestrated proteolysis of hundreds of proteins, and to achieve efficacy the cell has evolved a three-step activation cascade. First, initiator caspases (caspases 8, 9, 10) are activated either by external death ligands triggering the formation of the death-inducing signaling complex (DISC; caspases 8 and 10) or by the release of cytochrome c from mitochondria provoking the assembly of the apoptosome (caspase-9). Once activated, the initiator caspases directly cleave executioner caspases 3 and 7, leading to their activation. The third step is the caspase-3-mediated activation of the executioner caspase-6, a peptidase with a substrate preference similar to that of initiator caspases. In this manner, the cell amplifies the proteolytic activity of caspases to swiftly cause cell demise. As caspases become active, proteolysis of various death substrates happens with the general goal of stopping crucial cellular processes, but for a limited number of proteins a gain in function occurs, often via the loss of a regulatory domain or the separation of domains carrying different functions. These substrates are difficult to identify as they necessarily require the study of individual fragments generated by caspases and because their proteolysis does not occur in isolation. Earlier studies have established that most caspases recognize a five-amino-acid motif in their substrates and cleave following an aspartate residue. However, recent reports have muddied the water in respect to these simple rules for substrate recognition and cleavage, by showing that caspases can also cut after a glutamate residue and employ exosites to improve catalysis. It is therefore not trivial to assign a role to a specific cleavage event and a substrate to a specific caspase. In a recent publication, we identified two novel caspase substrates, sorting nexin (SNX) 1 and 2, involved in endosomal sorting (Figure 1). The two proteins are cleaved by initiator caspases in vitro and in cellulo during apoptosis. We further showed that SNX1 contains multiple cleavage sites, including following glutamate residues. Interestingly, cellular repression of caspase-6, which also cleaves SNX2, did not fully abolish SNX2 proteolysis, suggesting that proteases upstream of this caspase may also participate in the cleavage of SNX2 in cells. Furthermore, only initiator caspases were able to cleave SNX1, reinforcing the idea that initiators contribute to the cleavage of SNX proteins in cell death. This possibility is of particular interest as few substrates have been assigned to initiator caspases. SNX1 and SNX2 are two early endosome-localized proteins belonging to the SNX family of proteins, whose members are critical for protein trafficking stepswithin the endocytic pathway. Both SNX1 and SNX2 regulate the endosome-to-TGN (transGolgi network) retrograde transport of lysosomal receptors as part of the retromer complex (Vps26, Vps29, Vps35) and the endosomal sorting of various plasma membrane receptors. Our work demonstrates that SNX2 cleavage leads to dissociation from Vps35 and the delocalization of Vps26, suggesting the uncoupling of the endosomal cargo recognition machinery from the transport machinery itself. Consequently, the cleavage of SNX2 and by extension that of SNX1 affect the trafficking of several proteins. One aspect that may be affected by SNX cleavage is apoptosis via death receptors that requires their trafficking in endosomes, where the associated initiator caspase-8 activates the bulk of the executioner caspase-3 (reviewed in Guicciardi et al.). Moreover, studies have demonstrated that caspase-3 immobilization via palmitoylation at the plasma membrane, where caspase-8 is initially recruited at the DISC, enhances its activation. This suggests that co-localization of a caspase with specific substrates can result in more efficacious proteolysis. Although we have not yet assessed the role of SNX cleavage in modulating caspase-8-mediated apoptosis, this avenue is appealing in light of our findings because SNX1 and SNX2 promote endosomal trafficking. Thus, we speculate that SNX1 and SNX2 cleavage by caspase-8, which we demonstrated are best performed by this caspase in vitro, is a means to stall internalized DISC in vesicles and further promote the activation of caspases. A second way SNX1 and SNX2 cleavage may affect the cell is by modulating the endosomal sorting of many signaling receptors, including receptor tyrosine kinases (RTKs). Significantly, we showed that cellular depletion of SNX2 increased hepatocyte growth factor (HGF) RTK (MET) phosphorylation and its capacity to support Erk1/2 activation,