Feature-based reappraisal of the Bacillus subtilis exoproteome

Proteomics-based verification of computer-assisted predictions on bacterial protein export have indicated that problems occur with the distinction between (Sec-type) signal peptides that govern protein secretion, and lipoprotein signal peptides or amino-terminal membrane anchors that cause protein retention in the membrane. Therefore, the main aim of this study was to investigate whether feature-based predictions by the SecretomeP (SecP) algorithm will aid the proteomics-based analysis of protein export in Bacillus subtilis. The SecP algorithm is trained to recognize features such as secondary structure and disordered regions, which are generally present in secreted proteins. The results showed that membrane-retained proteins receive, in general, high SecP scores, similar to the scores of secretory proteins. Importantly, the SecP algorithm aided in the re-evaluation of a class of previously identified proteins that remain attached to the membrane despite the presence of an apparent Sec-type signal peptide. These so-called 'Sec-attached' proteins receive on average a lower SecP score, and several of these proteins could be unmasked as transmembrane proteins by combined SecP and signal peptide analyses. Finally, the present study suggests that feature-based outlier analysis may provide leads towards the discovery of novel special-purpose pathways for bacterial protein export.