Rapid MRSA detection via tandem mass spectrometry of the intact 80 kDa PBP2a resistance protein

Abstract

Treatment of antibiotic-resistant infections is dependent on the detection of specific bacterial genes or proteins in clinical assays. Identification of methicillin-resistant Staphylococcus aureus (MRSA) is often accomplished through the detection of penicillin-binding protein 2a (PBP2a). With greater dependence on mass spectrometry (MS)-based bacterial identification, complementary efforts to detect resistance have been hindered by the complexity of those proteins responsible. Initial characterization of PBP2a indicates the presence of glycan modifications. To simplify detection, we demonstrate a proof-of-concept tandem MS approach involving the generation of N-terminal PBP2a peptide-like fragments and detection of unique product ions during top-down proteomic sample analyses. This approach was implemented for two PBP2a variants, PBP2a_mecA and PBP2a_mecC, and was accurate across a representative panel of MRSA strains with different genetic backgrounds. Additionally, PBP2a_mecA was successfully detected from clinical isolates using a five-minute liquid chromatographic separation and implementation of this MS detection strategy. Our results highlight the capability of direct MS-based resistance marker detection and potential advantages for implementing these approaches in clinical diagnostics.

Figure 1

Identification of PBP2a_mecA protein from MRSA cell extract by LC–MS/MS. (a) Intact wild-type PBP2a_mecA MS spectrum from a representative MRSA strain (ATCC 33,591) separated by LC. Intact protein precursor ion charge states are labelled from 80 to 110. Data are representative for multiple MRSA strains (n ≥ 3). (b) Recombinant His₆-PBP2a_mecA MS spectrum acquired using direct infusion. Identical precursor ion charges states are labelled as wild-type PBP2a_mecA spectrum. (c) Intact wild-type PBP2a_mecA MS spectrum produced during LC separation of cell extract and PTCR-mediated separation of superposed protein ion populations. Isolation window was centered at m/z 777 with a width of 5 m/z. Precursor ion charge states are labelled from 80 to 95. Data are representative of multiple technical replicates from the same MRSA cell extract (ATCC 33,591). (d) MS/MS spectrum of intact wild-type PBP2a_mecA precursor ion at m/z 793 (charge state = 102), using 1.5 m/z isolation window and fragmented with an HCD collision energy of 10 eV. Abundant N-terminal (b-ions) and C-terminal (y-ions) fragment ions are labelled. Fragmentation data are representative for multiple charge states and different MRSA strains (n ≥ 100). (e) Associated b- and y-ion fragment location for MS/MS of intact wild-type PBP2a_mecA protein indicated by purple vertical lines. Red colored amino acids indicate protein coverage (72.2%) for complementary bottom-up peptide analysis and protein characterization. Protein coverage associated with peptide data are from multiple analyses (n = 12).

Figure 2

PBP2a_mecA protein detection using in-source generated peptide-like fragments and MS/MS fragmentation. (a, b) Source-induced dissociated N-terminal peptide-like fragments from PBP2a_mecA (ATCC MRSA isolate BAA-44) at precursor m/z 596.8961 (a) and m/z 653.4382 (b) for selected ion monitoring (insets) and associated representative targeted MS/MS spectra (n ≥ 100). (c) Synthetic peptide confirmation of source-induced dissociated peptide-like precursor m/z 596.8961 and associated MS/MS spectrum (inset spectrum series displays peptide spectrum without in-source energy, with in-source dissociation, and magnified spectrum of target precursor following in-source dissociation). (d) Synthetic peptide confirmation of source-induced dissociated peptide-like precursor m/z 653.4382 and associated MS/MS spectrum (inset spectrum series displays peptide spectrum without in-source energy, with in-source dissociation, and magnified spectrum of target precursor following in-source dissociation). Data collection for synthetic peptides was repeated multiple times on different MS instruments (n = 3).

Figure 3

PBP2a_mecC protein detection using in-source generated peptide-like fragments and MS/MS fragmentation. (a, b) Source-induced dissociated N-terminal peptide-like fragments from PBP2a_mecC (ATCC MRSA isolate BAA-2312) at precursor m/z 658.9042 (a) and m/z 715.4462 (b) for selected ion monitoring (insets) and associated representative targeted MS/MS spectra (n ≥ 100). (c) Synthetic peptide confirmation of source-induced dissociated peptide-like precursor m/z 658.9042 and associated MS/MS spectrum (inset spectrum series displays peptide spectrum without in-source energy, with in-source dissociation, and magnified spectrum of target precursor following in-source dissociation). (d) Synthetic peptide confirmation of source-induced dissociated peptide-like precursor m/z 715.4462 and associated MS/MS spectrum (inset spectrum series displays peptide spectrum without in-source energy, and magnified spectrum of target precursor following in-source dissociation). Data collection for synthetic peptides was repeated multiple times on different MS instruments (n = 3).

Figure 4

Evaluation of source-induced peptide detection of PBP2a across representative MRSA panel and clinical isolates. (a) Performance of source-induced detection method for PBP2a_mecA (top panel) and PBP2a_mecC (bottom panel) N-terminal peptide-like fragments for MRSA strains exhibiting different SCCmec and PFGE genetic backgrounds, as well as negative MSSA isolates (listed in Supplemental Table S6). Strains were cultured on TSA plates and harvest after overnight growth. Colored bars represent the mean maximum intensity for each product ion over a 60-min LC protein separation. Results are from distinct biological replicates (n = 3). (b) Performance of MSSA and MRSA quality control strains along with clinical isolates over five-minute PBP2a_mecA detection method using custom SPE trap and source-induced dissociation method. Strains were cultured on Blood Agar plates and harvested after overnight growth. Data are the mean maximum ion intensity for each product ion and calculated from distinct biological replicates (n = 2 or 3).