Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Mar 1.
Published in final edited form as: J Appl Microbiol. 2019 Dec 9;128(3):697–709. doi: 10.1111/jam.14507

Detection and Identification of a Protein Biomarker in Antibiotic Resistant E. coli using intact protein LC Offline MALDI-MS and MS/MS

Anthony Maus a, Bledar Bisha b, Clifton Fagerquist c, Franco Basile a,*
PMCID: PMC7444928  NIHMSID: NIHMS1059436  PMID: 31715076

Abstract

Aims:

The identification and differentiation of antibiotic resistant bacteria by MALDI-TOF-MS profiling remains a challenge due to the difficulty in detecting unique protein biomarkers associated with this trait. To expand the detectable proteome in antibiotic-resistant bacteria, we describe a method implementing offline LC protein separation/fractionation prior to MALDI-ToF-MS and top-down MALDI-ToF/ToF-MS (tandem MS or MS/MS) for the analysis of several antibiotic-resistant Escherichia coli isolates.

Methods and Results:

Coupling offline LC with MALDI-ToF-MS increased the number of detected protein signals in the typically analyzed mass regions (m/z 3,000–20,000) by a factor of 13. Using the developed LC-MALDI-ToF-MS protocol in conjunction with supervised principal components analysis (sup-PCA), we detected a protein biomarker at m/z 9355 which correlated to β-lactam resistance among the E. coli bacteria tested. Implementing a top-down MALDI-ToF/ToF-MS approach, the pre-fractionated protein biomarker was inferred as a DNA-binding HU protein, likely translated from the blaCMY-2 gene (encoding AmpC-type β-lactamase) in the incompatibility plasmid complex A/C (IncA/C).

Conclusions:

Our results demonstrate the utility of LC-MALDI-MS and MS/MS to extend the number of proteins detected and perform MALDI-accessible protein biomarker discovery in microorganisms.

Significance and Impact of Study:

This outcome is significant since it expands the detectable bacterial proteome via MALDI-ToF-MS.

Keywords: Proteomics, Resistance, Detection, E. coli (all potentially pathogenic types), Agriculture

Introduction

Intact protein analysis via Matrix Assisted Laser Desorption/Ionization (MALDI)-Time of Flight (TOF)-Mass Spectrometry (MS) is now an established and viable method for the routine identification of microorganisms in the clinical setting.(Anon, 2011; Charles L. Wilkins & Jackson O. Lay, Jr., 2006) In this approach, intact bacteria cells are deposited directly onto the MALDI plate (or their proteins quickly extracted) and analyzed by MALDI-MS to obtain a protein profile characteristic of the bacteria being analyzed. However, the MALDI-MS profiling approach is limited in its ability to reliably and consistently identify antibiotic resistant bacteria.(Alekshun & Levy, 2007; Schaumann, Knoop, Genzel, et al., 2012) Established MALDI-MS profiling techniques primarily detect abundant ribosomal proteins under 15 kDa. Because these proteins have highly conserved sequences, they provide little differentiation power at the strain level.(Pineda, Antoine, Demirev, et al., 2003; Stelzl, Connell, Nierhaus, et al., 2001; Sandrin, Goldstein & Schumaker, 2013) Detection in profiling mode is typically limited to only about 30–50 protein signals per mass spectrum, limiting the detection of potential biomarkers for differentiation/identification.(Freiwald & Sauer, 2009) For example, there are about 1600 detectable proteins in E. coli (based on 2-dimmensional (2D) gel experiments), thus profile-based MALDI-MS measurements are only accessing about 3% of the detectable proteome.(Basile & Mignon, Rudolph K., 2016)

Due to widespread use of β-lactam antibiotics, identification of resistance to this class of antibiotics in bacteria is of particular interest. Although other mechanisms of antibiotic resistance exist, the primary mechanism of resistance to β-lactam antibiotics is through the production of the enzyme β-lactamase, which catalyzes the hydrolysis of the amide bond in the β-lactam ring.(Lima, Pinto, Ribeiro, et al., 2013) β-Lactamases form a family of heterogeneous enzymes with molar masses (MW) ranging from 29 to 39 kDa.(Bush & Jacoby, 2010) Since MALDI-MS profiling protocols detect proteins in the mass range of 5–20 kDa, this approach has had mixed success in the detection of antibiotic resistant bacteria.(Schaumann, Knoop, Genzel, et al., 2012; Egli, Tschudin-Sutter, Oberle, et al., 2015) In a notable example of direct detection of intact β-lactamase, Camara and Hays(Camara & Hays, 2007a) analyzed bacterial extracts with MALDI-MS in the m/z 3,000–40,000 range. Using their protocol, they were able to detect a peak near m/z 29,000 for ampicillin-resistant E. coli, which was later confirmed by SDS-PAGE and peptide mass fingerprinting (PMF). In another example of direct detection of β-lactamase, Papagiannitsis and coworkers(Papagiannitsis, Kotsakis, Tuma, et al., 2014) prepared extracts of periplasmic proteins from Enterobacteriaceae clinical isolates using a modified sucrose method and the extracts were analyzed by MALDI- MS. Using this technique, they detected a signal at m/z 39,850, which was identified as Citrobacter freundii-derived CMY-2-like cephalosporinases (using SDS-PAGE and PMF).

New strategies have been developed to detect β-lactam resistance employing MALDI-MS detection. In a targeted approach, MALDI-MS was used to determine β-lactamase activity by detecting the hydrolysis products from enzyme activity.(Hrabák, Chudáčková & Walková, 2013; Kostrzewa, Sparbier, Maier, et al., 2013) Recently, stable isotope-labeled media with antibiotic drugs have been used for detection of various antibiotic resistances in microorganisms.(Demirev, Hagan, Antoine, et al., 2013; Jung, Eberl, Sparbier, et al., 2014; Sparbier, Lange, Jung, et al., 2013)

Other approaches incorporate a Liquid Chromatography (LC) step prior to MS analysis in order to increase the selectivity and dynamic range of the number of proteins detected (i.e., increased proteome depth).(Aguilar, 2004) Although LC-tandem mass spectrometry (MS/MS) instrumentation has primarily been used to analyze peptides in a bottom-up proteomic approach, the analysis of intact proteins via a top-down method has several advantages. For example, Everly et al demonstrated the use of LC-electrospray ionization (ESI)-Quadrupole (Q)-TOF-MS to differentiate strains of E. coli.(Everley, Mott, Wyatt, et al., 2011) When compared to MALDI-MS profiling, the LC-MS technique detected more high MW proteins that can be used as biomarkers for differentiation. In a follow-up study, the same laboratory detected proteins up to 50,000 Da and identified a number of specific biomarkers.(Mott, Everley, Wyatt, et al., 2010) Williams and coworkers analyzed protein extracts from V. parahaemolyticus using LC-ESI-Q-TOF-MS and found that pandemic strains expressed a histone-like DNA-binding protein, HU-α, with a C-terminal amino acid sequence different from those of other strains, which could be used as a biomarker. In a later study, the same group differentiated thermal resistant strains of bacteria using biomarkers discovered via LC-Q-TOF-MS.(Williams, Monday, Edelson-Mammel, et al., 2005) McFarland and coworkers used LC-Q-TOF-MS/MS to obtain intact protein profiles and differentiated several Salmonella serovars.(McFarland, Andrzejewski, Musser, et al., 2014) Investigators used MS/MS analysis of intact proteins to identify several of the most intense protein signals and identify single-nucleotide polymorphisms.

Top-down proteomic analyses have also been conducted with MALDI and tandem ToF mass analyzers (ToF/ToF) to successfully fragment and analyze moderately sized proteins of up to 12 kDa in size. This capability was first demonstrated by Lin et al in 2003 where they showed fragmentation of proteins at aspartyl and prolyl residues using high energy CID (2 keV collision energy). Subsequently, Fagerquist et al(Fagerquist, Garbus, Williams, et al., 2009) and Demirev and coworkers(Demirev, Feldman, Kowalski, et al., 2005) each independently developed protocols and software capable of identifying proteins based on their MALDI-tandem mass spectra. Both software programs work by performing in-silico digestions of proteins to produce a list of potential fragments. The tandem mass spectra are then compared with the theoretical fragments and scored, thus assigning a confidence in the inference (identification) of the protein and its microbial source. An added dimension of selectivity can be introduced by performing off-line LC coupled with MALDI-ToF/ToF. For example, Maltman et al applied LC-MALDI-ToF/ToF to analyze the low mass proteome (i.e. peptidome) of neural progenitor cells,(Maltman, Brand, Belau, et al., 2011) while Quinton and coworkers used LC-MALDI-ToF/ToF to identify disulfide bridges and proteins in crude venom samples.(Quinton, Demeure, Dobson, et al., 2007)

The work presented herein uses a two-prong approach implementing offline LC-MALDI-MS and top-down MS/MS to greatly improve protein biomarker detection and identification in microorganisms. We demonstrate this approach with several E. coli isolates with varying degrees of antibiotic resistance and discover a protein-biomarker related to the resistance phenotype. The approach uses supervised Principal Component Analysis (sup-PCA) of the LC-MALDI-MS data to initially recognize potential biomarkers. In subsequent experiments, isolated LC fractions containing such potential biomarkers are then re-analyzed via a top-down proteomic strategy using a MALDI-ToF/ToF instrument. With this approach, we identify a DNA-binding HU protein exhibiting exceptional correlation to β-lactam resistance and trace its source to the Inc(A/C) plasmid.

Materials and Methods

Chemicals:

Deionized water (18 MΩ) was used (Sartorius, Bohemia, NY) and acetonitrile (ACN; HPLC grade), agar, Tris base, isopropyl alcohol (HPLC grade), and sequence grade formic acid (all from Fisher Scientific, Fair Lawn, NJ) were used for sample preparation and solutions. Optima grade (Fisher Scientific) water and ACN were used for LC mobile phases. Hydrochloric acid was obtained from EMD Millipore (Billerica, MA). Triflouroacetic acid (TFA), iodoacetamide, dithiothreitol (DTT), insulin (Bovine), cytochrome c (horse heart), myoglobin (horse skeletal muscle), ethanol, sinapinic acid, and α-cyano-4-hydroxycinnamic acid (CHCA) were all purchased from Sigma-Aldrich (St. Louis, MO). Brain Heart Infusion broth (BHI) and agar (BHIA) were from BD, Franklin Lakes, NJ.

Microbiological analyses:

Escherichia coli isolates were environmental isolates from the collection of Dr. Bisha in the food microbiology laboratory, which were collected in previous studies from whole gastrointestinal (GIT) tracts of European starlings (Sturnus vulgaris) associated with beef cattle concentrated animal feeding operations (CAFOs). No animal work was performed in the current study. For confirmation of isolates, MALDI Biotyping was performed using a previously described formic acid-acetonitrile extraction procedure (Bizzini, Durussel, Bille, et al., 2010). One loopful (approximately 1 μL) of bacterial colony material was suspended in a 1:3 solution HPLC grade water and absolute ethanol. Bacteria were collected by centrifugation at 17,000 × g, pellets allowed to air dry, and then suspended 1:1 in acetonitrile and 70% formic acid. Insoluble material was pelleted by centrifugation as above and 1 μL of the supernatant was applied polished steel target plate (Bruker, Billerica, MA). Samples were then air-dried and overlaid with 1 μl of freshly prepared α-cyano-4-hydroxycinnamic acid matrix (Bruker, Billerica, MA). MALDI Biotyping was accomplished using a Bruker Ultraflex II ToF/ToF or Bruker Microflex LRF (Billerica, MA) operating with Bruker Biotyper RTC software (Version 3.1) and pre-calibrated with Bruker Bacterial Test Standard. Species level identification of the isolates was accepted if a score of ≥ 2.0 was assigned by the MALDI Biotyper algorithm.

Antimicrobial Susceptibility Testing:

Antimicrobial susceptibility testing of E. coli isolates was performed in accordance with Clinical and Laboratory Standards Institute’s (CLSI’s) M02-A11 and M100-S24 protocols. Sensi-Discs impregnated with 18 different antibiotics were used for susceptibility testing, including 9 β-lactam antibiotics representative of several groups as described below. Penicillins: ampicillin (AMP, 10 μg), and piperacillin (PIP, 100 μg); β-lactam/β-lactamase inhibitor combinations: amoxicillin-clavulanate (AMC, 20/10 μg); Carbapenems: imipenem (IPM, 10 μg); Cephems: cefazolin (CFZ, 30 μg), cefotaxime (CTX, 30 μg), cefoxitin (FOX, 30 μg), ceftazidime (CAZ, 30 μg); and Monobactams: aztreonam (AZA, 30 μg). Following standards defined by CLSI, isolates were grouped as sensitive (S), intermediate (I), or resistant (R) to the respective antibiotics.

Real-Time PCR Detection of CIT-type AmpCs:

The presence or absence CIT-type AmpCs was detected by employing a multiplex real-time PCR designed to detect the predominant class A beta-lactamase as adapted from Roschanski et al.(Roschanski, Fischer, Guerra, et al., 2014) In addition, the protocol allowed for detection of blaCTX-M, blaSHV, and blaTEM in the same one-step reaction. A CFX96 Touch Real-Time PCR Detection System (Bio-Rad laboratories, Hercules, CA, USA) was used to perform all PCR reactions in 25 μL reaction volumes. Real-time PCR conditions were as follows: 95°C for 15 min, then 30 cycles at 95°C for 15 s, 50°C for 15 s, and 70°C for 20 s.

Biological Samples and Protein Extraction for Proteomic Analysis:

A total of 10 E. coli isolates samples for proteomic analyses were selected based on a susceptibility/resistance test against β-lactam antibiotics. All isolates were grown on BHIA for 24 hours at 37° C. Proteins were extracted using the method from Camara and Hays(Camara & Hays, 2007b) with some modifications. Briefly, 10 mg of microbial material (approximately 3×109 cells) was placed in a microcentrifuge tube with 1 mL of 0.1% TFA and washed. A 50 μL aliquot of extraction solution (17:33:50, formic acid: isopropyl alcohol: water) was added to the cells and vortexed for 1 minute and centrifuged for 5 minutes at 15,000xg. The supernatant containing the suspended proteins was then transferred to a new microcentrifuge tube. The solvent was then removed and the proteins resuspended in 75 μL of 20% ACN in water and frozen at this stage until HPLC analysis. For comparison, MALDI-MS profiling mass spectra were obtained using the method published by Freiwald & Sauer.(Freiwald & Sauer, 2009)

Liquid Chromatography:

Proteins were separated using a reversed-phase liquid chromatography column (15 cm long, 100 μm inner diameter (ID) column; packed in-house with C-18 stationary phase 3.6 μm diameter and 200 Å pores; Phenomenex, Torrance, CA) using a HPLC unit (NLC 400x autosampler coupled to an ekspert™ nanoLC 425; Sciex) with a 2 μL sample injection (detailed HPLC parameters can be found in the Supplemental Information section).

During separation, an ekspot™ spotter (Sciex) was used to spot eluate pre-mixed with MALDI matrix directly onto the MALDI plate. The MALDI matrix was 5 mg/mL sinapinic acid in 80% ACN, 20% water, and 1% TFA. The MALDI matrix flow rate was set at 0.63 μL/min and mixed online with the LC eluate. Each fraction (i.e., MALDI spot) was collected for 24 s onto a 384 well plate (Sciex), and fractions were collected between 30 and 120 min of the LC gradient.

Mass Spectrometry:

After protein LC fractionation, the MALDI plate was analyzed by MALDI-ToF/ToF (5800™, Sciex). Two mass ranges were analyzed designated ‘mid-mass’ and ‘high-mass’. The mid mass TOF range for detection was m/z 3,000–20,000 with a focus mass of 10,000. A fixed laser intensity of 6,000 arbitrary units (au) was used and 500 shots were averaged for each MALDI spot (i.e., a 24 s LC fraction).

Data Processing:

Following MS analysis of the entire MALDI plate (~225 fractions or spots), a peak list was generated using the Peak Explorer™ software (Sciex). The generated peak lists were transferred to the MarkerView™ software (Sciex) for Principle Component Analysis (PCA). Peak alignment and filtering (removing redundant signals in the final peak list) was first performed on all data sets. A conservative mass tolerance of 1,000 ppm was used, and all signals within the 120 min separation time were considered (see Supplemental Information for additional PCA parameters). Supervised PCA was performed on all data sets, with log weighting and autoscaling. Supervised PCA groups were based on the antimicrobial resistance phenotypes established by the disk diffusion testing: highly susceptible, intermediate (susceptibility) and highly resistant. Supervised PCA plots were constructed with discriminant components D1 and D2, each accounting for 50% of the total variance amongst all the samples. The top 10 loadings for each antibiotic resistance were then manually verified in all data files in an effort to confirm distinct and unique biomarkers. Mirror-image plots were generated using the mMass software program.(Strohalm, Hassman, Košata, et al., 2008)

Collection, Digestion, and Bottom-up Proteomic Analysis of Selected LC Fractions:

Based on supervised PCA results, fractions containing potential biomarkers were collected into microcentrifuge tubes in subsequent LC fractionations. The fractions were dried by vacuum centrifuge and resuspended in 50 μL of 100 mM Tris-HCl at pH 8 for subsequent digestion (see details in Supplemental Information).

Digested samples were then analyzed by offline LC MALDI-ToF-ToF-MS/MS using a MALDI plate pre-coated with graphite powder to increase the peptide signal intensity and thus the number of protein identifications.(Maus, Mignon & Basile, 2018).

Peptides fractionated and collected onto the MALDI plate were immediately analyzed by MALDI-ToF/ToF-MS/MS. The reflector-mode MS analysis mass window was m/z 900–4,000 with a focus mass set at 2,000. A fixed laser intensity of 2,700 au was used and each mass spectrum resulted from the average of 1,000 laser shots.

After reflector-mode MS analysis was completed on all fractions collected, an interpretation method was used to select a maximum of 15 precursor ions per fraction (spot) for MS/MS analysis. Each MS/MS mass spectrum was an average of 1,000 laser shots.

Tandem mass spectrometry measurements were processed using the Peak Explorer™ and ProteinPilot™ software packages (Sciex). The criteria for peptide identification was set at a minimum of 95% confidence, and the protein inference in the Paragon™ algorithm was set to a minimum of 99% confidence score. These settings allowed for protein inference (at 99% confidence) with a minimum of one peptide. ProteinPilot™ was also used to export MS/MS results in Mascot Generic Format (.mgf) for data analysis using the Mascot™ database search tool.

Top-Down Proteomic Analysis of Biomarker:

For top-down LC-MALDI-ToF/ToF-MS analysis, the LC parameters were unchanged from those discussed above except a 4 μL injection volume was used. The MALDI plate was also coated with graphite powder using the method described above. Based on supervised PCA results, LC fractions containing potential biomarker candidates were subjected to MS/MS analyses. The ToF/ToF was set to 1 kV without Collision Induced Dissociation (1 kV-no CID) and used for all top-down MS/MS measurements. The laser was operated at 7,900 au and the detector voltage was set to −2.400 kV. Other important parameters for the ToF/ToF top-down measurements are listed in the Supplemental Information section. Selected tandem mass spectra were exported (in .t2d format; Sciex) and analyzed via a web-based software for top-down proteomics.(Fagerquist, Garbus, Williams, et al., 2009) For the potential biomarker detected, the experimental precursor protein mass of 9356 ± 15 Da was used, which yielded 934 E. coli proteins within the specified parameters. The spectra were searched using both residue specific (aspartic acid and glutamic acid only) and non-residue specific in silico fragment ion comparisons with a fragment ion tolerance of ±1.0 u.

Results

The benefits of LC separation prior to MALDI-MS analysis can be seen in Figure 1, wherein the average number of total proteins detected using an established MALDI-MS profiling protocol (Freiwald & Sauer, 2009) is compared to the offline LC-MALDI-MS technique (all 10 E. coli isolates tested; both in the range m/z 3,000–20,000). The MALDI-MS profiling protocol detected an average of about 30 unique protein signals, while the offline LC MALDI-MS methodology detected an average of 300 unique protein signals. Representative mass spectra for the E. coli isolate 1413 from different LC fractions are compared with a bacterial MALDI-MS profile mass spectrum in Figure 2. This comparison clearly points out the benefits of LC separation prior to MALDI-MS analysis over MALDI-MS profiling of bacteria. These mass spectra at different fraction numbers show a much greater variety of protein signals than would be otherwise observed in a single MALDI-MS profile mass spectrum.

Figure 1.

Figure 1.

Open in a new tab

Average number of total protein signals detected from E. coli isolates when using established MALDI-MS methods and the LC-MALDI-MS method presented herein (error bars=±1 std. dev.). LC separation greatly improved protein detection capabilities. M/z range used in both analyses: 2,000 – 20,000.

Figure 2.

Figure 2.

Open in a new tab

Comparison of bacteria MALDI-MS profile mass spectrum with LC-MALDI-MS fraction mass spectra of the same E. coli isolate (strain 1413).

Supervised PCA (sup-PCA) was applied to the pre-processed data and the resulting plots are shown in Figure 3 (scatter and loading plots). The sup-PCA groups were based on the number of β-lactam antibiotics each isolate was resistant out of 9 antibiotics tested. From the loadings plot, it can be concluded that the biomarker detected at m/z 9355 is highly correlated with isolates displaying resistance to a larger number of β-lactam antibiotics used in testing. Overall, the sup-PCA biomarker discovery process yielded 4 potential biomarker signals at m/z’s 9355, 7625, 7135, and 17582 (Table 1). To avoid artifacts and possible false positives, these signals were manually verified to be only present in mass spectra of the isolates resistant to a larger number of antimicrobials tested, and not in the mass spectra of susceptible strains.

Figure 3.

Figure 3.

Open in a new tab

Supervised Principal Component Analysis of LC-MALDI-MS analyses of E. coli strains: a) scatter PCA plot and b) Loadings plot. Supervised PCA groups were based on the antimicrobial resistance phenotypes established by the disk diffusion testing: highly susceptible, intermediate (susceptibility) and highly resistant. The supervised PCA plots was constructed with discriminant components D1 and D2, each accounting for 50% of the total variance amongst all the samples

Table 1.

Potential biomarkers derived from supervised-PCA and β-lactam resistance characteristics. The numbers in the chart are number of antibiotics tested (out of 9 total; #R= number of highly resistant, #S= number of highly susceptible, #int= intermediate resistant). The check marks indicate the frequency that a biomarker was detected during the analysis of that isolate (with two total analyses performed on each isolate).

graphic file with name nihms-1059436-t0007.jpg
Open in a new tab

The potential biomarker detected at m/z 9355 was of particular interest for further characterization since it was present in 4 of the 5 isolates resistant to most antimicrobial tested, with a consistent and reproducible signal at a signal-to-noise ratio (S/N) greater than 5. Moreover, this signal was not detected in any of the MALDI-MS profiling experiments performed on the same sample set, making it a unique candidate for a novel biomarker. The top mass spectrum shown in Figure 4 was obtained from an isolate displaying resistance to a broad spectrum of β-lactams and it is compared with the mass spectrum of an isolate with higher susceptibility to β-lactams (bottom mass spectrum). These mass spectra were obtained from the same LC fraction in each analysis. In the case of the highly resistant isolate, the biomarker peak at m/z 9355 is clearly detected, while it is undetected in the highly susceptible isolate.

Figure 4.

Figure 4.

Open in a new tab

Mirror-image plot of MALDI-mass spectra resulting from the same LC fraction of a highly resistant isolate (1383) and highly susceptible isolate (1612). In the highly resistant spectra, the biomarker peak is present with high intensity. In contrast, this peak is absent in this and all fractions of the highly susceptible isolate. The signal is also not detected in any spectra using established profiling techniques.

To identify the biomarker at m/z 9355, the appropriate LC-fraction (or MALDI spot) was subjected to MALDI-ToF/ToF-MS analysis and the resulting tandem mass spectrum is shown in Figure 5. Database search was performed using aspartic acid (D) and glutamic acid (E) residue-specific fragmentation during MS/MS. The top protein sequence match was a DNA-binding HU protein with a theoretical m/z 9354.5 for the [M+H]+ ion (hereafter referred to as HU-biomarker) and the top 10 sequence matches are listed in Table 1S (in the Supplemental Information section). An average m/z 9356.6 was measured for the [M+H]+ ion of the HU-biomarker, which corresponds to a 0.02% error (versus theoretical), well within the experimental accuracy for this instrument when operated in the linear mode and with external mass calibration. The P-value assigned to this sequence match was 4.6×10−10, with the nearest alternative sequence P-value being 9.3×10-3.

Figure 5.

Figure 5.

Open in a new tab

Annotated tandem spectrum (MS/MS) of the biomarker at m/z 9,355 used for database search and protein inference. The assigned sequence and mass are shown. The spectrum was searched using a residue specific (D and E) database search. Fragments corresponding to all but 1 D or E residue were detected.

To determine the genetic origin of the HU-biomarker, the experimentally derived amino acid sequence was searched against the genome database using the NCBI tblastn tool.(NCBI, NIH) A total of 16 high confidence matches were obtained and they are listed in Table 2. Interestingly, all these matches correspond to plasmids, many of which are responsible for resistance to β-lactam antibiotics.

Table 2.

Matches between the experimentally derived amino acid sequence of HU-biomarker and the genomic database search. All the top hits are plasmids. Entries in bold represent plasmid sequences reported by Fernández-Alarcón et al.

Protein name/description Identity Accession
Escherichia coli plasmid pV139-a DNA, contig: V139-a_scaffold_6, strain: V139 100% LC056335.1
Escherichia coli O145:H28 str. RM12581 plasmid pRM12581, complete sequence 100% CP007137.1
Escherichia coli O145:H28 str. RM13514 plasmid pRM13514, complete sequence 100% CP006029.1
Escherichia coli strain SCEC2 plasmid pSCEC2, complete sequence 100% KF152885.1
Escherichia coli strain PG010208 plasmid pPG010208, complete sequence 100% HQ023861.1
Escherichia coli strain Y5 plasmid pECY53, complete sequence 100% KT997783.1
Escherichia coli strain H4H plasmid peH4H, complete sequence 100% FJ621586.1
Escherichia coli strain 39R861 plasmid p39R861–4, complete sequence 100% KP276584.1
Escherichia coli UMNK88 plasmid pUMNK88, complete sequence 100% HQ023862.1
Escherichia coli strain APEC1990_61 plasmid pAPEC1990_61, complete sequence 100% HQ023863.1
Escherichia coli strain AR060302 plasmid pAR060302, complete sequence 100% FJ621588.1
Escherichia coli strain 6409 plasmid p6409–202.186kb, complete sequence 100% CP010373.2
Escherichia coli strain YDC637 plasmid pYDC637, complete sequence 100% KP056256.1
Escherichia coli strain Ecol_316 plasmid pEC316_KPC, complete sequence 100% CP018956.1
Escherichia coli strain BK32533 plasmid pBK32533, complete sequence 100% KP345882.1
Escherichia coli strain Ecol_AZ155 plasmid pECAZ155_KPC, complete sequence 100% CP019001.1
Open in a new tab

Examination of the bottom-up proteomics data revealed the detection/identification of the DNA-binding HU-α and HU-β proteins from E. coli K12. With knowledge of the molecular masses of the HU proteins inferred by bottom-up proteomics, two signals were tentatively assigned to correspond to the HU-α and Hu-β proteins (at m/z’s 9537 and 9226, respectively) in the LC-MALDI-MS data of intact proteins. The signals for the HU-α and HU-β proteins were detected in all isolates, regardless of resistance characteristics. An extracted ion chromatogram (XIC; shown in Figure 6) was constructed with data from a highly resistant isolate. This XIC includes traces with the signals from the HU-α, HU-β, and the HU-biomarker proteins, the later unique to the resistant strains tested. The close LC retention times among these 3 proteins also indicates a shared sequence homology. It should be noted that in several LC-MALDI-MS/MS bottom-up proteomic analyses, the HU-biomarker protein was not detected/identified, while the AmpC β-lactamase (blaCMY-2), HU-α and HU-β were all confidently detected (data not shown).

Figure 6.

Figure 6.

Open in a new tab

Superimposed extracted ion chromatograms of the HU-α (m/z 9537.7±2.9, green trace), HU-β (9226.8±2.9, blue trace) and HU-Biomarker proteins (9356.6±2.8, red trace). The theoretical and average experimental [M+H]+ are shown with a ±1 std. dev.

Discussion

Many of the limitations of the MALDI-MS profiling technique for the analysis of bacteria can be attributed to the limited number of proteins detected relative to the available proteome. This shortcoming is mostly a consequence of ionization suppression effects known to take place during the MALDI process when a complex mixture is analyzed. This is the case in bacteria MALDI-MS profiling, where abundant ribosomal proteins are preferentially ionized. The implementation of an LC separation step prior to MALDI-MS analysis alleviates these ionization suppression effects, resulting in a significant increase in the number of detected protein signals (Figure 1 and 2).

Based on the sup-PCA results (Figure 3), the biomarker at m/z 9355 was deemed to have the strongest correlation with antibiotic resistance within the bacteria tested. In addition, genome database searching provided strong evidence that the identified HU-biomarker was translated by the plasmid IncA/C responsible for antibiotic resistance based on the production of β-lactamase. Moreover, the presence of the plasmid-mediated AmpC gene in these isolates was confirmed via real-time PCR, even in an isolate which did not express typical AmpC phenotype. Interestingly, this isolate yielded a mass spectrum containing the biomarker at m/z 9355, thus leading us to conclude that the HU-biomarker is most likely plasmid-mediated.

Further evidence is presented linking the HU-biomarker to the IncA/C plasmid. In a study published by Fernández-Alarcón et al(Fernández-Alarcón, Singer & Johnson, 2011) four IncA/C plasmids isolated from E. coli (from livestock in the United States and Chile) were sequenced and analyzed. In particular, two plasmids studied by this group are among the high confident matches listed in Table 2, the pUMNK88 and pAR060302 plasmids. These plasmids contain a single insertion of the blaCMY-2 gene, which encodes for extended-spectrum AmpC-type β-lactamases.(Fernández-Alarcón, Singer & Johnson, 2011; Call, Singer, Meng, et al., 2010) In addition, based on the genetic sequence of these Inc(A/C) plasmids, the authors inferred that these proteins share similarities to the HU-beta family of proteins, among others,(Fernández-Alarcón, Singer & Johnson, 2011) and supports our hypothesis that the discovered HU-biomarker in the bacteria tested is plasmid mediated. Therefore, it is highly likely that the HU-biomarker protein detected in this study is the predicted HU-like protein translated from the same plasmid responsible for β-lactam resistance. As a result, the identified biomarker is indirectly linked to the mechanism of resistance.

Extended-spectrum and AmpC-type β-lactamases are acknowledged as very important in Enterobacteriaceae.(Pitout & Laupland, 2008; Liebana, Carattoli, Coque, et al., 2013) In particular, therapeutics for gram-negative bacteria producing plasmid-mediated β-lactamases remains difficult, because these bacteria will express resistance to all β-lactam antibiotics except cefepime, cefpirome, and carbapenems.(Pérez-Pérez & Hanson, 2002) The problem is compounded by the inability of phenotypic tests to distinguish between chromosomal and plasmid-mediated β-lactamases (especially in hyperproducing E. coli strains and bacteria with inducible chromosomal AmpC enzymes) as well as between the six known families of plasmid-mediated AmpC β –lactamases, requiring utilization of genotypic characterization by PCR to determine such differences.(Pérez-Pérez & Hanson, 2002; Thomson, 2010) However, the ability to apply this differentiation remains paramount for surveillance, epidemiology studies, and hospital infection mitigation, due to the fact that plasmid-mediated genes such as those encoding extended spectrum β-lactamases (ESBL) or AmpC can disseminate to other bacteria within the hospital settings, food animals and the environment.(Pérez-Pérez & Hanson, 2002; Gazouli, Tzouvelekis, Prinarakis, et al., 1996; Blanc, Mesa, Saco, et al., 2006) Among plasmid-mediated AmpC cephalosporinases, the CIT-type AmpC β-lactamases (e.g. CMY-2,−3,−4, LAT-1, LAT-2, BIL-1) are often detected, with the CMY-2 type being the most frequent AmpC β-lactamase from patients in hospitals, community, livestock and ground meat.(Roschanski, Fischer, Guerra, et al., 2014; Polsfuss, Bloemberg, Giger, et al., 2011; Woodford, Reddy, Fagan, et al., 2007; Ewers, Bethe, Semmler, et al., 2012) As a consequence, the CMY-2 type AmpC is considered to be the most common plasmid-mediated AmpC β-lactamase worldwide.(Ewers, Bethe, Semmler, et al., 2012) In the sample set studied, the identified HU-biomarker is a unique protein that shows correlation with AmpC resistance phenotype, and as such, it could be used as another biomarker for plasmid-mediated AmpC β-lactam resistance.

Beyond phenotypic and genotypic methods, many studies have been used to determine resistance characteristics by MALDI-MS detection, either by profiling methods or measuring the β-lactamase enzyme activity. Our results demonstrate the ability of detecting blaCMY-2 plasmid-mediated proteins that may not be directly responsible for resistance. In our case, this protein was of a mass and expression level amenable to MS analysis. In addition, our approach could be applied to detect/discover other plasmid-mediated biomarkers for antibiotic resistance in different microorganisms and different types of antibiotic resistances. Although many similar studies have been conducted using LC-ESI-MS/MS instruments, to the best of our knowledge this is the first study successfully using LC-MALDI-MS for microorganism biomarker discovery. It is expected that MALDI-MS techniques will continue to grow in use, especially in the clinical setting. Our investigation demonstrates the utility of combining LC separation prior to MALDI-MS for protein biomarker discovery. However, because the LC separation adds a significant amount of time and complexity to the MALDI-MS analysis, this approach is far from becoming routine analysis in many labs. In addition, reproducible detection of proteins > 20 kDa remains a challenge with this method, excluding the possibility of detecting potential biomarkers known to exist in this mass range. Even though a variety of established techniques such as polymerase chain reaction (PCR) and phenotypic methods are being used to detect antibiotic resistance in microorganisms, our work demonstrated that incorporating offline LC to MALDI-MS has the potential to discover potential biomarkers and determine microbial resistance characteristics, especially in laboratories that already have access to MALDI-MS and/or MS/MS instrumentation.

Supplementary Material

Supp info

Acknowledgements

We thank Jennifer Anders and Leslie Day for their assistance in preparing the microbial samples and providing disk measurements. Funding from NSF (NSF-MRI 1429615) for the acquisition of the MALDI-MS/MS instrument is gratefully acknowledged. B. Bisha acknowledges funding from NIFA WYO-511-14. A. Maus acknowledges the financial support from NIH (NIH-INBRE) to the University of Wyoming; The NIH-INBRE project at the U. of Wyoming was supported by a grant from the National Institute of General Medical Sciences (P20GM103432) from the National Institutes of Health. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Footnotes

Conflict of Interest

No conflict of interest declared

References

  1. Aguilar M-I (2004) HPLC of peptides and proteins: methods and protocols. Totowa, N.J, Humana Press. [Google Scholar]
  2. Alekshun MN & Levy SB (2007) Molecular Mechanisms of Antibacterial Multidrug Resistance. Cell. [Online] 128 (6), 1037–1050. Available from: doi: 10.1016/j.cell.2007.03.004. [DOI] [PubMed] [Google Scholar]
  3. Anon (n.d.) tblastn: search translated nucleotide databases using a protein query. [Online]. Available from: https://blast.ncbi.nlm.nih.gov/Blast.cgi?PROGRAM=tblastn&PAGE_TYPE=BlastSearch&BLAST_SPEC=&LINK_LOC=blasttab&LAST_PAGE=blastn [Accessed: 23 July 2018].
  4. Basile F & Mignon Rudolph K. (2016) Methods and Instrumentation in Mass Spectrometry for the Differentiation of Closely Related Microorganisms In: Applications of Mass Spectrometry in Microbiology: From Strain Characterization to Rapid Screening for Antibiotic Resistance. Plamen Demirev and Todd Sandrin (Eds). Springer; pp. 13–50. [Google Scholar]
  5. Bizzini A, Durussel C, Bille J, Greub G, et al. (2010) Performance of Matrix-Assisted Laser Desorption Ionization-Time of Flight Mass Spectrometry for Identification of Bacterial Strains Routinely Isolated in a Clinical Microbiology Laboratory. Journal of Clinical Microbiology. [Online] 48 (5), 1549–1554. Available from: doi: 10.1128/JCM.01794-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Blanc V, Mesa R, Saco M, Lavilla S, et al. (2006) ESBL- and plasmidic class C β-lactamase-producing E. coli strains isolated from poultry, pig and rabbit farms. Veterinary Microbiology. [Online] 118 (3), 299–304. Available from: doi: 10.1016/j.vetmic.2006.08.002. [DOI] [PubMed] [Google Scholar]
  7. Bush K & Jacoby GA (2010) Updated Functional Classification of β-Lactamases. Antimicrobial Agents and Chemotherapy. [Online] 54 (3), 969–976. Available from: doi: 10.1128/AAC.01009-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Call DR, Singer RS, Meng D, Broschat SL, et al. (2010) blaCMY-2-Positive IncA/C Plasmids from Escherichia coli and Salmonella enterica Are a Distinct Component of a Larger Lineage of Plasmids. Antimicrobial Agents and Chemotherapy. [Online] 54 (2), 590–596. Available from: doi: 10.1128/AAC.00055-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Camara JE & Hays FA (2007a) Discrimination between wild-type and ampicillin-resistant Escherichia coli by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Analytical and Bioanalytical Chemistry. [Online] 389 (Copyright (C) 2014 American Chemical Society (ACS). All Rights Reserved.), 1633–1638. Available from: doi: 10.1007/s00216-007-1558-7. [DOI] [PubMed] [Google Scholar]
  10. Camara JE & Hays FA (2007b) Discrimination between wild-type and ampicillin-resistant Escherichia coli by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Analytical and Bioanalytical Chemistry. [Online] 389 (5), 1633–1638. Available from: doi: 10.1007/s00216-007-1558-7. [DOI] [PubMed] [Google Scholar]
  11. Wilkins Charles L. & Lay Jackson O. Jr. (2006) Identification of Microorganisms by Mass Spectrometry. (1st edition . Hoboken, NJ, John Wiley & Sons, Inc. [Google Scholar]
  12. Demirev PA, Feldman AB, Kowalski P & Lin JS (2005) Top-Down Proteomics for Rapid Identification of Intact Microorganisms. Analytical Chemistry. [Online] 77 (22), 7455–7461. Available from: doi: 10.1021/ac051419g. [DOI] [PubMed] [Google Scholar]
  13. Demirev PA, Hagan NS, Antoine MD, Lin JS, et al. (2013) Establishing Drug Resistance in Microorganisms by Mass Spectrometry. Journal of The American Society for Mass Spectrometry. [Online] 24 (8), 1194–1201. Available from: doi: 10.1007/s13361-013-0609-x. [DOI] [PubMed] [Google Scholar]
  14. Egli A, Tschudin-Sutter S, Oberle M, Goldenberger D, et al. (2015) Matrix-Assisted Laser Desorption/Ionization Time of Flight Mass-Spectrometry (MALDI-TOF MS) Based Typing of Extended-Spectrum β-Lactamase Producing E. coli – A Novel Tool for Real-Time Outbreak Investigation. PLOS ONE. [Online] 10 (4), e0120624. Available from: doi: 10.1371/journal.pone.0120624. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Everley RA, Mott TM, Wyatt SA, Toney DM, et al. (2011) Liquid chromatography/mass spectrometry characterization of Escherichia coli and Shigella species. Journal of the American Society for Mass Spectrometry. [Online] 19 (11), 1621–1628. Available from: doi: 10.1016/j.jasms.2008.07.003. [DOI] [PubMed] [Google Scholar]
  16. Ewers C, Bethe A, Semmler T, Guenther S, et al. (2012) Extended-spectrum β-lactamase-producing and AmpC-producing Escherichia coli from livestock and companion animals, and their putative impact on public health: a global perspective. Clinical Microbiology and Infection. [Online] 18 (7), 646–655. Available from: doi: 10.1111/j.1469-0691.2012.03850.x. [DOI] [PubMed] [Google Scholar]
  17. Fagerquist CK, Garbus BR, Williams KE, Bates AH, et al. (2009) Web-Based Software for Rapid Top-Down Proteomic Identification of Protein Biomarkers, with Implications for Bacterial Identification. Applied and Environmental Microbiology. [Online] 75 (13), 4341–4353. Available from: doi: 10.1128/AEM.00079-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Catherine Fenselau & Plamen Demirev (eds.) (2011) Rapid Characterization of Microorganisms by Mass Spectrometry. ACS Symposium Series 1065 [Online]. American Chemical Society; Available from: http://pubs.acs.org/doi/book/10.1021/bk-2011-1065 [Accessed: 26 June 2015]. [Google Scholar]
  19. Fernández-Alarcón C, Singer RS & Johnson TJ (2011) Comparative Genomics of Multidrug Resistance-Encoding IncA/C Plasmids from Commensal and Pathogenic Escherichia coli from Multiple Animal Sources. PLOS ONE. [Online] 6 (8), e23415. Available from: doi: 10.1371/journal.pone.0023415. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Freiwald A & Sauer S (2009) Phylogenetic classification and identification of bacteria by mass spectrometry. Nat. Protocols. [Online] 4 (5), 732–742. Available from: doi: 10.1038/nprot.2009.37. [DOI] [PubMed] [Google Scholar]
  21. Gazouli M, Tzouvelekis LS, Prinarakis E, Miriagou V, et al. (1996) Transferable cefoxitin resistance in enterobacteria from Greek hospitals and characterization of a plasmid-mediated group 1 beta-lactamase (LAT-2). Antimicrobial Agents and Chemotherapy. 40 (7), 1736–1740. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hrabák J, Chudáčková E & Walková R (2013) Matrix-Assisted Laser Desorption Ionization–Time of Flight (MALDI-TOF) Mass Spectrometry for Detection of Antibiotic Resistance Mechanisms: from Research to Routine Diagnosis. Clinical Microbiology Reviews. [Online] 26 (1), 103–114. Available from: doi: 10.1128/CMR.00058-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Jacoby GA (2009) AmpC β-Lactamases. Clinical Microbiology Reviews. [Online] 22 (1), 161–182. Available from: doi: 10.1128/CMR.00036-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Jung JS, Eberl T, Sparbier K, Lange C, et al. (2014) Rapid detection of antibiotic resistance based on mass spectrometry and stable isotopes. European Journal of Clinical Microbiology & Infectious Diseases. [Online] 33 (6), 949–955. Available from: doi: 10.1128/CMR.00036-08. [DOI] [PubMed] [Google Scholar]
  25. Kostrzewa M, Sparbier K, Maier T & Schubert S (2013) MALDI-TOF MS: an upcoming tool for rapid detection of antibiotic resistance in microorganisms. PROTEOMICS – Clinical Applications. [Online] 7 (11–12), 767–778. Available from: doi: 10.1002/prca.201300042. [DOI] [PubMed] [Google Scholar]
  26. Liebana E, Carattoli A, Coque TM, Hasman H, et al. (2013) Public Health Risks of Enterobacterial Isolates Producing Extended-Spectrum β-Lactamases or AmpC β-Lactamases in Food and Food-Producing Animals: An EU Perspective of Epidemiology, Analytical Methods, Risk Factors, and Control Options. Clinical Infectious Diseases. [Online] 56 (7), 1030–1037. Available from: doi: 10.1093/cid/cis1043. [DOI] [PubMed] [Google Scholar]
  27. Lima TB, Pinto MFS, Ribeiro SM, Lima L.A. de, et al. (2013) Bacterial resistance mechanism: what proteomics can elucidate. The FASEB Journal. [Online] 27 (4), 1291–1303. Available from: doi: 10.1096/fj.12-221127. [DOI] [PubMed] [Google Scholar]
  28. Maltman DJ, Brand S, Belau E, Paape R, et al. (2011) Top-down label-free LC-MALDI analysis of the peptidome during neural progenitor cell differentiation reveals complexity in cytoskeletal protein dynamics and identifies progenitor cell markers. PROTEOMICS. [Online] 11 (20), 3992–4006. Available from: doi: 10.1002/pmic.201100024. [DOI] [PubMed] [Google Scholar]
  29. Maus A, Mignon R & Basile F (2018) Enhanced protein identification using graphite-modified MALDI plates for offline LC-MALDI-MS/MS bottom-up proteomics. Analytical Biochemistry. [Online] 54531–37. Available from: doi: 10.1016/j.ab.2018.01.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. McFarland MA, Andrzejewski D, Musser SM & Callahan JH (2014) Platform for Identification of Salmonella Serovar Differentiating Bacterial Proteins by Top-Down Mass Spectrometry: S. Typhimurium vs S. Heidelberg. Analytical Chemistry. [Online] 86 (14), 6879–6886. Available from: doi: 10.1021/ac500786s. [DOI] [PubMed] [Google Scholar]
  31. Mott TM, Everley RA, Wyatt SA, Toney DM, et al. (2010) Comparison of MALDI-TOF/MS and LC-QTOF/MS methods for the identification of enteric bacteria. International Journal of Mass Spectrometry. [Online] 291 (1–2), 24–32. Available from: doi: 10.1016/j.ijms.2009.12.015. [DOI] [Google Scholar]
  32. Papagiannitsis CC, Kotsakis SD, Tuma Z, Gniadkowski M, et al. (2014) Identification of CMY-2-type cephalosporinases in clinical isolates of Enterobacteriaceae by MALDI-TOF MS. Antimicrobial Agents and Chemotherapy. [Online] AAC.02418–13. Available from: doi: 10.1128/AAC.02418-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Pérez-Pérez FJ & Hanson ND (2002) Detection of Plasmid-Mediated AmpC β-Lactamase Genes in Clinical Isolates by Using Multiplex PCR. Journal of Clinical Microbiology. [Online] 40 (6), 2153–2162. Available from: doi: 10.1128/JCM.40.6.2153-2162.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Pineda FJ, Antoine MD, Demirev PA, Feldman AB, et al. (2003) Microorganism Identification by Matrix-Assisted Laser/Desorption Ionization Mass Spectrometry and Model-Derived Ribosomal Protein Biomarkers. Analytical Chemistry. [Online] 75 (15), 3817–3822. Available from: doi: 10.1021/ac034069b. [DOI] [PubMed] [Google Scholar]
  35. Pitout JD & Laupland KB (2008) Extended-spectrum β-lactamase-producing Enterobacteriaceae: an emerging public-health concern. The Lancet Infectious Diseases. [Online] 8 (3), 159–166. Available from: doi: 10.1016/S1473-3099(08)70041-0. [DOI] [PubMed] [Google Scholar]
  36. Polsfuss S, Bloemberg GV, Giger J, Meyer V, et al. (2011) Practical Approach for Reliable Detection of AmpC Beta-Lactamase-Producing Enterobacteriaceae. Journal of Clinical Microbiology. [Online] 49 (8), 2798–2803. Available from: doi: 10.1128/JCM.00404-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Quinton L, Demeure K, Dobson R, Gilles N, et al. (2007) New Method for Characterizing Highly Disulfide-Bridged Peptides in Complex Mixtures: Application to Toxin Identification from Crude Venoms. Journal of Proteome Research. [Online] 6 (8), 3216–3223. Available from: doi: 10.1021/pr070142t. [DOI] [PubMed] [Google Scholar]
  38. Roschanski N, Fischer J, Guerra B & Roesler U (2014) Development of a Multiplex Real-Time PCR for the Rapid Detection of the Predominant Beta-Lactamase Genes CTX-M, SHV, TEM and CIT-Type AmpCs in Enterobacteriaceae. PLOS ONE. [Online] 9 (7), e100956. Available from: doi: 10.1371/journal.pone.0100956. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Sandrin TR, Goldstein JE & Schumaker S (2013) MALDI TOF MS profiling of bacteria at the strain level: A review. Mass Spectrometry Reviews. [Online] 32 (3), 188–217. Available from: doi: 10.1002/mas.21359. [DOI] [PubMed] [Google Scholar]
  40. Schaumann R, Knoop N, Genzel GH, Losensky K, et al. (2012) A step towards the discrimination of beta-lactamase-producing clinical isolates of Enterobacteriaceae and Pseudomonas aeruginosa by MALDI-TOF mass spectrometry. Medical Science Monitor : International Medical Journal of Experimental and Clinical Research. [Online] 18 (9), MT71–MT77. Available from: doi: 10.12659/MSM.883339. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Shilov IV, Seymour SL, Patel AA, Loboda A, et al. (2007) The Paragon Algorithm, a Next Generation Search Engine That Uses Sequence Temperature Values and Feature Probabilities to Identify Peptides from Tandem Mass Spectra. Molecular & Cellular Proteomics. [Online] 6 (9), 1638–1655. Available from: doi: 10.1074/mcp.T600050-MCP200. [DOI] [PubMed] [Google Scholar]
  42. Sparbier K, Lange C, Jung J, Wieser A, et al. (2013) MALDI Biotyper-Based Rapid Resistance Detection by Stable-Isotope Labeling. Journal of Clinical Microbiology. [Online] 51 (11), 3741–3748. Available from: doi: 10.1128/JCM.01536-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Stelzl U, Connell S, Nierhaus KH & Wittmann-Liebold B (2001) Ribosomal Proteins: Role in Ribosomal Functions In: eLS. [Online]. John Wiley & Sons, Ltd; p. Available from: http://onlinelibrary.wiley.com/doi/10.1038/npg.els.0000687/abstract [Accessed: 8 September 2015]. [Google Scholar]
  44. Strohalm M, Hassman M, Košata B & Kodíček M (2008) mMass data miner: an open source alternative for mass spectrometric data analysis. Rapid Communications in Mass Spectrometry. [Online] 22 (6), 905–908. Available from: doi: 10.1002/rcm.3444. [DOI] [PubMed] [Google Scholar]
  45. Thomson KS (2010) Extended-Spectrum-β-Lactamase, AmpC, and Carbapenemase Issues. Journal of Clinical Microbiology. [Online] 48 (4), 1019–1025. Available from: doi: 10.1128/JCM.00219-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Williams TL, Monday SR, Edelson-Mammel S, Buchanan R, et al. (2005) A top-down proteomics approach for differentiating thermal resistant strains of Enterobacter sakazakii. Proteomics. [Online] 5 (16), 4161–4169. Available from: doi: 10.1002/pmic.200401263. [DOI] [PubMed] [Google Scholar]
  47. Woodford N, Reddy S, Fagan EJ, Hill RLR, et al. (2007) Wide geographic spread of diverse acquired AmpC β-lactamases among Escherichia coli and Klebsiella spp. in the UK and Ireland. Journal of Antimicrobial Chemotherapy. [Online] 59 (1), 102–105. Available from: doi: 10.1093/jac/dkl456. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supp info
NIHMS1059436-supplement-Supp_info.docx (62.8KB, docx)

RESOURCES