Investigating the effects of protein patterns on microorganism identification by high-performance liquid chromatography-mass spectrometry and protein database searches

High-performance liquid chromatography-electrospray ionization mass spectrometry (HPLC-ESI-MS) has been employed for separation and detection of protein biomarkers from E. coli samples. LC-MS is suitable for microbial identification because it can couple on-line with sample clean-up devices and is readily amenable to automation. In this work, we have investigated the effects of sample preparation methods on the detection of bacterial proteins by LC-MS. Many factors effect the degree of variations in the protein patterns (i.e. number and masses of proteins). For example, changing the polarity as well as pH of the extraction solvent may control the number of detected proteins. It is also noted that the protein patterns can vary even when the total ion chromatography plots seem to be the same under the same sample preparation conditions. Further, we have tested experimentally the influence of LC-MS-analyzed protein patterns (molecular masses between 2000 and 60,000) on microbial identification by protein database searches. This is in contrast to the current database search approach, where only the masses of smaller proteins (less than or equal to20,000) from direct matrix-assisted laser/desorption ionization MS are used. In spite of the variations in protein patterns, all the database search results show that the best matches come from the correct microorganism. (C) 2002 Elsevier Science B.V. All rights reserved.