Proteomics-based confirmation of protein expression and correction of annotation errors in the Brucella abortus genome

Abstract

Background: Brucellosis is a major bacterial zoonosis affecting domestic livestock and wild mammals, as well as humans around the globe. While conducting proteomics studies to better understand Brucella abortus virulence, we consolidated the proteomic data collected and compared it to publically available genomic data. Results: The proteomic data was compiled from several independent comparative studies of Brucella abortus that used either outer membrane blebs, cytosols, or whole bacteria grown in media, as well as intracellular bacteria recovered at different times following macrophage infection. We identified a total of 621 bacterial proteins that were differentially expressed in a condition-specific manner. For 305 of these proteins we provide the first experimental evidence of their expression. Using a custom-built protein sequence database, we uncovered 7 annotation errors. We provide experimental evidence of expression of 5 genes that were originally annotated as non-expressed pseudogenes, as well as start site annotation errors for 2 other genes. Conclusions: An essential element for ensuring correct functional studies is the correspondence between reported genome sequences and subsequent proteomics studies. In this study, we have used proteomics evidence to confirm expression of multiple proteins previously considered to be putative, as well as correct annotation errors in the genome of Brucella abortus strain 2308.