Mass Spectrometry Proteotyping-Based Detection and Identification of Staphylococcus aureus, Escherichia coli, and Candida albicans in Blood

Abstract

Bloodstream infections (BSIs), the presence of microorganisms in blood, are potentially serious conditions that can quickly develop into sepsis and life-threatening situations. When assessing proper treatment, rapid diagnosis is the key; besides clinical judgement performed by attending physicians, supporting microbiological tests typically are performed, often requiring microbial isolation and culturing steps, which increases the time required for confirming positive cases of BSI. The additional waiting time forces physicians to prescribe broad-spectrum antibiotics and empirically based treatments, before determining the precise cause of the disease. Thus, alternative and more rapid cultivation-independent methods are needed to improve clinical diagnostics, supporting prompt and accurate treatment and reducing the development of antibiotic resistance. In this study, a culture-independent workflow for pathogen detection and identification in blood samples was developed, using peptide biomarkers and applying bottom-up proteomics analyses, i.e., so-called “proteotyping”. To demonstrate the feasibility of detection of blood infectious pathogens, using proteotyping, Escherichia coli and Staphylococcus aureus were included in the study, as the most prominent bacterial causes of bacteremia and sepsis, as well as Candida albicans, one of the most prominent causes of fungemia. Model systems including spiked negative blood samples, as well as positive blood cultures, without further culturing steps, were investigated. Furthermore, an experiment designed to determine the incubation time needed for correct identification of the infectious pathogens in blood cultures was performed. The results for the spiked negative blood samples showed that proteotyping was 100- to 1,000-fold more sensitive, in comparison with the MALDI-TOF MS-based approach. Furthermore, in the analyses of ten positive blood cultures each of E. coli and S. aureus, both the MALDI-TOF MS-based and proteotyping approaches were successful in the identification of E. coli, although only proteotyping could identify S. aureus correctly in all samples. Compared with the MALDI-TOF MS-based approaches, shotgun proteotyping demonstrated higher sensitivity and accuracy, and required significantly shorter incubation time before detection and identification of the correct pathogen could be accomplished.

Multiplexing the Identification of Microorganisms via Tandem Mass Tag Labeling Augmented by Interference Removal through a Novel Modification of the Expectation Maximization Algorithm

“…To meet this condition, we implemented the approach we proposed in MiCId. 21,35,57,62 MiCId is a fast workflow that assigns accurate statistical significance (E value) to identified microorganisms and has high sensitivity and specificity for the identifications of microorganisms from samples composed of single and multiple microorganisms. 31,35 A major computational difficulty encountered when using TMT labeling is caused by co-fragmentation of several precursor ions, resulting in interference associated with the measured intensities of TMT reporter ions.…”

“…As the proposed method is complementary to existing methods, it has the potential to have a significant impact in assisting with the identification of typical and atypical microorganisms, identification of microorganisms in polymicrobial samples, and providing protein expression information. 21,58,62,82 In addition, the proposed method can also be used to assist with the identifications of Journal of the American Society for Mass Spectrometry pubs.acs.org/jasms Article microorganisms in other research areas such as culturomics studies, where many unknown microorganismal isolates are grown under hundreds of culture conditions and then submitted for microorganismal identification. 11,12 Another potential limitation of our proposed approach is the cost associated with TMT labeling reagents.…”

“…A prerequisite of our study was to have accurate identification of microorganisms as the starting point. To meet this condition, we implemented the approach we proposed in MiCId. ,,, MiCId is a fast workflow that assigns accurate statistical significance ( E value) to identified microorganisms and has high sensitivity and specificity for the identifications of microorganisms from samples composed of single and multiple microorganisms. , A major computational difficulty encountered when using TMT labeling is caused by co-fragmentation of several precursor ions, resulting in interference associated with the measured intensities of TMT reporter ions. − To evaluate how well the EM algorithm implemented in the proposed approach handles interference, we used 94 MS/MS data files having a variety of instrumental parameters, a wide range of protein concentration ratios across TMT-labeled channels, and containing two organisms per channel for 90 of the MS/MS data files. The samples containing two organisms can be viewed as samples of two organisms per TMT-labeled channel or as samples of a single microorganism per TMT-labeled channel embedded in a host background.…”

“…31,35 MiCId is a workflow, with accurate statistical significance assignment, designed for the rapid identification of microorganisms, antibiotic resistance proteins, and estimation of biomass, i.e., the relative amount of protein in the sample for each identified microorganism. 21,35,57,62 The E value, the statistic employed by MiCId to compute statistical significance for identified microorganisms, utilizes all confidently identified taxon-specific peptides and taxon-nonspecific peptides mapped to a given microorganism. 31,35 MiCId's workflow has been extensively tested and can be used to identify microorganisms from samples composed of single and multiple microorganisms.…”

Metataxonomic and metaproteomic profiling of the oral microbiome in oral lichen planus - a pilot study

“…Samples collected were thawed and the MolYsis™ kit (MolYsis Basic5 kit, Molzym GmbH & Co., Bremen, Germany) was used to remove human biomass, according to the supplier’s protocol, with minor modifications [ 36 ]. The resulting bacterial pellets were resuspended in 150 µl ammonium bicarbonate (20 mM) buffer containing 1% sodium deoxycholate (SDC) final concentration (from 5% stock solution in 20 mM ammonium bicarbonate).…”