Introduction

 

Caspases, or cysteine-aspartic acid proteases represent a unique class of intracellular cysteine proteases which function as critical effectors in apoptosis, necrosis, inflammation, and other important cellular processes such as cell proliferation, cell differentiation, cell migration and receptor internalization. Due to their essential roles in programmed cell death, caspases have been termed as “executioner” or “killer” proteins. The hallmark of caspase is that it contains a cysteine residue at the active site and can cleave the substrates at specific tetrapeptides (canonical P4-P1 sites) with a highly conserved aspartate (D) at the P1 position. To date at least fourteen mammalian caspases have been identified and phylogenetic analysis indicates that they can segregate into two major sub-families: ICE and CED-3. Although almost 400 caspase substrates have been reported to date, there are likely to be hundreds of novel caspase substrates yet to be discovered. Nevertheless, experimental identification and characterization of protease substrates is expensive and time-consuming, which will become increasingly difficult without the prior knowledge of the roles of these enzymes in various cellular pathways. Computational prediction of caspase substrate specificity may provide useful information about potential cleavage sites of candidate substrates for further determination and experimental characterization. In this context, the development of computational tools that can accurately predict the caspase substrate specificity and identify the new cleavage sites of novel substrates would be much more desirable and valuable.

Procleave is an online webserver for the computational identification of caspase substrate cleavage sites from primary substrate sequences. This server takes into consideration the characteristic sequential and structural features surrounding the cleavage and non-cleavage sites of caspase substrates, such as the predicted secondary structure, solvent accessibility and natively disordered regions, based on a novel bi-profile Bayesian feature extraction method. Potential caspase cleavage sites are further infered based on the probability score for each residue of a query substrate sequence predicted by Procleave. It is anticipated that Procleave will be useful for preliminary screening of putative caspase substrates, the identification of potential cleavage sites, as well as providing new insights for a better understanding of the caspase-substrate interaction relationships.