Improved Functional Prediction of Hypothetical Proteins from Listeria monocytogenes 08-5578

Abstract

Listeria monocytogenes is a foodborne human pathogen responsible for listerosis. The genomes of several L. monocytogenes strains have been recently sequenced. The genome of L. monocytogenes 08-5578, which was in part responsible for a significant listerosis outbreak in 2008, contains an unexpectedly high percentage of protein-encoding genes (1,927 out of 3,161; 60.96%) autonomously annotated as hypothetical proteins. The aim of this study was to test whether a manual annotation strategy could be used to assign more meaningful functional names to the hypothetical proteins of 08-5578. A holistic, manual gene annotation strategy that utilized sequence homology, cellular localization predictions, structure-based evidence, phylogeny, and protein-protein interaction data was used to assign potential cellular roles to 79 out of 100 hypothetical proteins randomly selected from the genome of 08-5578. Of significance, 5 of the 79 hypothetical proteins assigned a more meaningful name may contribute to the virulence of L. monocytogenes 08-5578, by contributing to chemotaxis, cell surface protein sorting, cell wall biosynthesis, and cold adaptation. The findings here support the notion that manual annotations, using a combination of diverse bioinformatics tools, can improve the quality of genomic information provided by automated genome annotation methods alone.