A multi-factor model for caspase degradome prediction

Abstract

Background: Caspases belong to a class of cysteine proteases which function as critical effectors in cellular processes such as apoptosis and inflammation by cleaving substrates immediately after unique tetrapeptide sites. With hundreds of reported substrates and many more expected to be discovered, the elucidation of the caspase degradome will be an important milestone in the study of these proteases in human health and disease. Several computational methods for predicting caspase cleavage sites have been developed recently for identifying potential substrates. However, as most of these methods are based primarily on the detection of the tetrapeptide cleavage sites - a factor necessary but not sufficient for predicting in vivo substrate cleavage - prediction outcomes will inevitably include many false positives.

Results: In this paper, we show that structural factors such as the presence of disorder and solvent exposure in the vicinity of the cleavage site are important and can be used to enhance results from cleavage site prediction. We constructed a two-step model incorporating cleavage site prediction and these factors to predict caspase substrates. Sequences are first predicted for cleavage sites using CASVM or GraBCas. Predicted cleavage sites are then scored, ranked and filtered against a cut-off based on their propensities for locating in disordered and solvent exposed regions. Using an independent dataset of caspase substrates, the model was shown to achieve greater positive predictive values compared to CASVM or GraBCas alone, and was able to reduce the false positives pool by up to 13% and 53% respectively while retaining all true positives. We applied our prediction model on the family of receptor tyrosine kinases (RTKs) and highlighted several members as potential caspase targets. The results suggest that RTKs may be generally regulated by caspase cleavage and in some cases, promote the induction of apoptotic cell death - a function distinct from their role as transducers of survival and growth signals.

Conclusion: As a step towards the prediction of in vivo caspase substrates, we have developed an accurate method incorporating cleavage site prediction and structural factors. The multi-factor model augments existing methods and complements experimental efforts to define the caspase degradome on the systems-wide basis.

Figure 1

Propensity for cleavage sites to occur in secondary structures. A. Propensities for different secondary structure elements (coils, C_p; helices, H_p; β-strands, E_p) were measured for 24-mer subsequences with or without caspase cleavage sites (labelled "cleavage sites" and "non-cleavage sites" respectively). B. Distribution of "cleavage sites" and "non-cleavage sites" subsequences to C_p bins. Each C_p bin (0.10, 0.20...1.00) was allocated a proportion of sequences with C_p scores falling within the bin range (0-0.10, 0.11-0.19...0.91-1.00).

Figure 2

Propensity for cleavage sites to occur in solvent accessible locations. A. Propensities for solvent accessibilities (Sp) were measured for 24-mer subsequences with or without caspase cleavage sites (labelled "cleavage sites" and "non-cleavage sites" respectively). B. Distribution of "cleavage sites" and "non-cleavage sites" subsequences to Sp bins. Each Sp bin (0.10, 0.20...1.00) was allocated a proportion of sequences with Sp scores falling within the bin range (0-0.10, 0.11-0.19...0.91-1.00).

Figure 3

Scatter plots of Sp and Cp values. Each data point corresponds to the Sp and Cp values of a single 24-mer subsequence of A. cleavage sites and B. non-cleavage sites.

Figure 4

Schematic diagram of the multi-factor model for caspase substrate prediction. Step1: A window scans the entire protein (example used: 14-3-3, Uniprot ID: P31946) for potential caspase cleavage sites (bold, underlined, P1 residue to the left of inverted triangle) using a caspase cleavage site prediction tool (example used: a 24 residue scanning window from CASVM). Step 2: 24-mer subsequences containing predicted cleavage sites with flanking ten residues downstream and upstream from the tetrapeptide sequence (P14 to P10') are constructed. Cp, Sp and P-score are calculated for all subsequences. Cleavage sites in subsequences with P-score above the assigned cut-off are selected.

Figure 5

Result of model prediction on test dataset. A. CASVM predicted cleavage sites were assigned to pools containing "true positives" or "false positives". Fractions of cleavage sites in their respective pools (vertical axis) with P-scores above the cut-offs (horizontal axis) were measured. B. GraBCas predicted cleavage sites were assigned to pools containing "true positives" or "false positives". Fractions of cleavage sites in their respective pools (vertical axis) with P-scores above the cut-offs (horizontal axis) were measured.