Strain-level Staphylococcus differentiation by CeO2-metal oxide laser ionization mass spectrometry fatty acid profiling
© Saichek et al. 2016
Received: 1 June 2015
Accepted: 2 March 2016
Published: 23 April 2016
The Staphylococcus genus is composed of 44 species, with S. aureus being the most pathogenic. Isolates of S. aureus are generally susceptible to β-lactam antibiotics, but extensive use of this class of drugs has led to increasing emergence of resistant strains. Increased occurrence of coagulase-negative staphylococci as well as S. aureus infections, some with resistance to multiple classes of antibiotics, has driven the necessity for innovative options for treatment and infection control. Despite these increasing needs, current methods still only possess species-level capabilities and require secondary testing to determine antibiotic resistance. This study describes the use of metal oxide laser ionization mass spectrometry fatty acid (FA) profiling as a rapid, simultaneous Staphylococcus identification and antibiotic resistance determination method.
Principal component analysis was used to classify 50 Staphyloccocus isolates. Leave-one-spectrum-out cross-validation indicated 100 % correct assignment at the species and strain level. Fuzzy rule building expert system classification and self-optimizing partial least squares discriminant analysis, with more rigorous evaluations, also consistently achieved greater than 94 and 84 % accuracy, respectively. Preliminary analysis differentiating MRSA from MSSA demonstrated the feasibility of simultaneous determination of strain identification and antibiotic resistance.
The utility of CeO2-MOLI MS FA profiling coupled with multivariate statistical analysis for performing strain-level differentiation of various Staphylococcus species proved to be a fast and reliable tool for identification. The simultaneous strain-level detection and antibiotic resistance determination achieved with this method should greatly improve outcomes and reduce clinical costs for therapeutic management and infection control.
Staphylococci are Gram-positive facultative anaerobes comprising 44 species commonly found in soil or on the skin of animals . S. aureus is the most pathogenic of the genus and is commonly associated with septicemia, osteomyelitis, endocarditis, and skin infection . Isolates of S. aureus are generally susceptible to β-lactam antibiotics, but extensive use of this class of drugs has led to the emergence of resistant strains . In 2011 the Centers for Disease Control and Prevention (CDC) reported 80,461 methicillin-resistant S. aureus (MRSA) infections in the U.S. leading to 11,285 deaths. While improved infection control policies decreased clinical MRSA infections by 52 % between 2005 and 2011, there remains a need to rapidly screen patients for S. aureus and determine antibiotic resistance.
Culture-, biochemical-, and molecular-based methods are the current standard for clinical MRSA detection. Culture methods offer high specificity, but relatively lengthy turnaround times (TAT) of 24–72 h and the requirement for secondary resistance testing contribute significantly to delays in onset of treatment. A retrospective cohort study of bloodstream infections found that mortality rates rose 7.6 % per hour for every hour of delay in the initiation of effective antimicrobial therapy . Chromogenic agars have been used to slightly decrease TAT to 18–24 h, while also improving specificity, but secondary resistance testing is still required .
Some of the most common approaches for analysis of the specific biochemical characteristics of staphylococci include: coagulase and phosphatase activity, hemolysis, nitrate reduction, and aerobic acid production from carbohydrate metabolism . Kloos and coworkers reported a simplified scheme for analyzing the extensive data produced by biochemical results to characterize staphylococci. The commercially available BioMérieux API STAPH-IDENT and American Hospital Supply Corporation MicroScan Systems are based on this approach. The API Staph-IDENT utilizes a battery of 10 microscale biochemical tests, whereas the MicroScan System consists of 27 tests . These systems were reported to have accuracies of 88 and 86.4 %, respectively, but also showed inherent limitations [8–10].
In order to improve the specificity and selectivity of Staphylococcus detection, molecular methods for analyzing specific genetic markers have been explored. In an attempt to identify S. aureus and assay for methicillin resistance, multiplexed PCR has been used to simultaneously target the staphylococcal nuc gene, encoding a thermostable nuclease (TNase), and the mecA gene, encoding a penicillin binding protein . PCR results agreed with coagulase production and agar screening tests for single-step identification of MRSA. In an attempt to identify coagulase-negative staphylococcal strains (CoNS), one study targeted a 429-bp amplicon of the sodA gene encoding the manganese-dependent superoxide dismutase . Clinical isolates and ATCC reference strains were identified with 83 % accuracy in about 8 h. While culturing and biochemical assays offer comparable specificity to results obtained by hsp60  and 16S rRNA sequencing ; TAT is still typically greater than 24 h.
Turnaround time was significantly reduced using phage amplification-based lateral flow immunochromatography (LFI) . This work led to the FDA-approved MicroPhage KeyPath MRSA/MSSA blood culture test . Exploitation of S. aureus-specific phage amplification targeting clinical blood isolates allowed for simultaneous identification and methicillin resistance determination with a TAT of 5 h and 98.3 % accuracy .
Published reports suggest the rise of non-S. aureus infections in clinical studies, some with resistance to multiple classes of antibiotics [17–19]. CoNS are among the most commonly reported bloodstream isolates (37.3 % compared to 12.6 % for S. aureus) . These reports place emphasis on the importance of S. epidermidis, S. saprophyticus, S. lugdunensis, and S. schleiferi infection and further demonstrate the need for more rapid techniques for simultaneous species-level Staphylococcus identification and antibiotic resistance determination. Bacterial protein-profiling by matrix assisted laser desorption/ionization-time of flight mass spectrometry (MALDI-TOF MS) has been used to identify S. aureus and CoNS in prosthetic joint infections . Although this method was relatively rapid, only 52 % highly probable species-level identification was obtained.
A report by Dubois and coworkers using the Bruker Biotyper MALDI-TOF MS protein analysis of 152 staphylococcal isolates correctly identify 151 samples at the species level. These results confirmed their earlier findings using a PCR-based sodA gene array . Rajakurna et al. correctly identified a different set of Staphylococcus isolates with 97 % accuracy using the MicrobeLynx macromolecule profiling database, developed by Waters Corporation .
A MALDI mass spectral-bacterial profiling approach using fatty acids as diagnostic biomarkers rather than proteins was recently reported [24–26]. Employing MALDI with CeO2 (metal oxide laser ionization [MOLI] MS) as an in situ saponification catalyst and matrix replacement, bacterial samples were identified to the species level with 97 % accuracy . In a follow up study, suites of Enterobacteriaceae, Listeria, and Acinetobacter were analyzed in parallel by MOLI MS fatty acid profiling and the Bruker Biotyper protein profiling . The results from this study clearly established fatty acid MOLI MS profiling for strain-level differentiation of closely-related phylotypes with 98–100 % accuracy. In comparison, protein profiling of the same samples correctly identified Enterobacteriaceae with 30 %, Listeria with 64 % and Acinetobacter with 66 % accuracy at the species level.
The present study describes MOLI MS CeO2 fatty acid profiling of 31 non-aureus Staphylococcus strains and 19 S. aureus strains (nine MRSA and ten MSSA). A fuzzy rule building expert system (FuRES)  and a self-optimizing partial least squares discriminant analysis (PLS-DA)  were used for classification.
Results and discussion
FuRES and PLS-DA calculations correctly classified the data into 27 strains. These results were obtained with 100 bootstraps and three Latin partitions. FuRES and PLS-DA had 93.9 ± 0.4 % and 84.1 ± 0.4 % prediction rates, respectively. From the PCA scores, it was shown that strains of the same species exhibit profiles that were highly similar.
We demonstrated the utility of CeO2-MOLI MS FA profiling coupled with multivariate statistical analysis for performing strain-level differentiation of various Staphylococcus species. The emergence of MRSA and CoNS clinical isolates and the need for rapid clinical intervention has made it increasingly important to differentiate Staphylococcus isolates at the species and strain level. LOSOCVs yielded 100 % correct classification at the species and strain level. FuRES classification, with a more rigorous evaluation, also consistently achieved 94 % accuracy. Preliminary analysis differentiating MRSA from MSSA demonstrated the feasibility of simultaneously determining strain identification and antibiotic resistance, which is increasingly important for therapeutic management and infection control. By eliminating the need for secondary testing, this could decrease the delay of drug administration by up to 54 h over conventional diagnostic techniques. Ultimately, as is also the case in protein profiling, construction of a comprehensive database will be necessary for identification of unknown isolates.
Staphylococcus species used
A. Species-level study
B. Strain-level study
C. Resistance study
Fatty acids used in Principal Component Analysis
Lipids were extracted as previously described [24, 27]. Briefly, individual colonies were suspended in 50 μL of a 1:2 v/v methanol/chloroform (Pharmco-AAPER, Shelbyville KY and Fischer, Pittsburgh PA, respectively) and vortexed for 120 s. to allow for cell disruption. An equal volume of phosphate buffer saline (PBS) at a pH of 7.4 was added prior to additional vortexing to facilitate phase separation. Extracts were centrifuged prior to MALDI sample preparation.
Sample preparation for MOLI MS analysis was carried out as previously described . Briefly, 100 mg of CeO2 (Cermac Inc., Milwaukee, WI) was suspended in 1 mL of n-hexane (Sigma Aldrich) prior to spotting 1 μL of the resulting slurry on a standard Bruker stainless steel MALDI plate. Two μL of each lipid extract was deposited directly on a CeO2 spot and allowed to air dry prior to analysis. MOLI-MS measurements were performed with a Bruker Ultraflextreme MALDI-TOF MS (Bruker Daltronics, Billerica, MA) in negative-ion reflectron mode with a grid voltage of 50.3 %, a delayed extraction time of 120 ns, and a sampling frequency of 1 kHz on a 355 nm Nd:YAG laser. Five replicates of each isolate were analyzed as 500 shot composites using automated laser rastering to ensure instrument stability.
Mass spectra were exported as ASCII files and processed using a Python algorithm to select and centroid 29 specific fatty acid peaks (Table 2), and scale each peak to total ion intensity. Processed data were written as.xls files for import into R (Ver. 3.0.2, R Foundation, Vienna, Austria) as a data frame. The prcomp()function mean centered and calculated PCA scores before plotting with the built-in plot()function. Leave-one-spectrum-out cross-validation (LOSOCV) was performed using linear discriminant analysis to validate the classification rate.
Processed fatty acid profiles were further analyzed with MATLAB 2014a (Mathworks, Natick, MA). Generalized prediction rates were measured using three Latin partitions and 100 bootstraps . Two classifiers were evaluated: a fuzzy rule-building expert system (FuRES)  and partial least squares discriminant analysis (PLS-DA) . The PLS-DA algorithm used two Latin partitions and ten bootstraps to calculate average pooled prediction errors . The number of components (i.e., latent variables) that minimized error was selected and used to build a model from the set of training data, which was then used as a prediction set. Training data consisted of a set of profiles used to build the classifiers; the test data was the set of profiles used to evaluate the performance of these classifiers. Hierarchical cluster analysis was used to generate dendrograms and graphically illustrate linkage distances (Euclidean distances) obtained from an agglomerative algorithm. The distances were between pairs of profiles or between the averages of profiles from subclusters.
The authors acknowledge the Colorado School of Mines Department of Chemistry for support of NRSaichek. Acquisition of the MALDI–TOF mass spectrometer was supported in part by National Science Foundation MRI grant CHE-1229156.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Lowy FD. Staphylococcus aureus infections. N Engl J Med. 1998;339(8):520–32.View ArticlePubMedGoogle Scholar
- Centres for Disease Control and Prevention (US). Antibiotic resistance threats in the United States, 2013. Centres for Disease Control and Prevention, US Department of Health and Human Services; 2013.Google Scholar
- Centers for Disease Control and Prevention. Active Bacterial Core Surveillance Report, Emerging Infections Program Network, Methicillin-Resistant Staphylococcus aureus; 2011.Google Scholar
- Kumar A, et al. Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock*. Crit Care Med. 2006;34(6):1589–96.View ArticlePubMedGoogle Scholar
- Flayhart D, et al. Multicenter evaluation of BBL CHROMagar MRSA medium for direct detection of methicillin-resistant Staphylococcus aureus from surveillance cultures of the anterior nares. J Clin Microbiol. 2005;43(11):5536–40.View ArticlePubMedPubMed CentralGoogle Scholar
- Kloos WE, Schleifer KH. Simplified scheme for routine identification of human Staphylococcus species. J Clin Microbiol. 1975;1(1):82–8.PubMedPubMed CentralGoogle Scholar
- Kloos WE, Wolfshohl JF. Identification of Staphylococcus species with the API STAPH-IDENT system. J Clin Microbiol. 1982;16(3):509–16.PubMedPubMed CentralGoogle Scholar
- Hussain ZAFAR, et al. Comparison of the MicroScan system with the API Staph-Ident system for species identification of coagulase-negative staphylococci. J Clin Microbiol. 1986;23(1):126–8.PubMedPubMed CentralGoogle Scholar
- Sasaki T, et al. Reclassification of phenotypically identified Staphylococcus intermedius strains. J Clin Microbiol. 2007;45(9):2770–8.View ArticlePubMedPubMed CentralGoogle Scholar
- Pottumarthy S, et al. Clinical isolates of Staphylococcus intermedius masquerading as methicillin-resistant Staphylococcus aureus. J Clin Microbiol. 2004;42(12):5881–4.View ArticlePubMedPubMed CentralGoogle Scholar
- Brakstad OG, Maeland JA, Tveten Y. Multiplex polymerase chain reaction for detection of genes for Staphylococcus aureus thermonuclease and methicillin resistance and correlation with oxacillin resistance. Apmis. 1993;101(7–12):681–8.View ArticlePubMedGoogle Scholar
- Martineau F, et al. Development of a PCR assay for identification of staphylococci at genus and species levels. J Clin Microbiol. 2001;39(7):2541–7.View ArticlePubMedPubMed CentralGoogle Scholar
- Bukau B, Horwich AL. The Hsp70 and Hsp60 chaperone machines. Cell. 1998;92(3):351–66.View ArticlePubMedGoogle Scholar
- Edwards U, Rogall T, Blöcker H, Emde M, Böttger EC. Isolation and direct complete nucleotide determination of entire genes. Characterization of a gene coding for 16S ribosomal RNA. Nucleic Acids Res. 1989;17(19):7843–53.View ArticlePubMedPubMed CentralGoogle Scholar
- Bhowmick T, et al. Controlled multicenter evaluation of a bacteriophage-based method for rapid detection of Staphylococcus aureus in positive blood cultures. J Clin Microbiol. 2013;51(4):1226–30.View ArticlePubMedPubMed CentralGoogle Scholar
- Smith D. Accurate Detection of Nasal MRSA Carriage by the Bacteriophage Amplification Test. 46th Annual Meeting. Idsa; 2008.Google Scholar
- David MZ, Daum RS. Community-associated methicillin-resistant Staphylococcus aureus: epidemiology and clinical consequences of an emerging epidemic. Clin Microbiol Rev. 2010;23(3):616–87.View ArticlePubMedPubMed CentralGoogle Scholar
- Sievert DM, et al. Vancomycin-resistant Staphylococcus aureus in the United States, 2002–2006. Clin Infect Dis. 2008;46(5):668–74.View ArticlePubMedGoogle Scholar
- Thati V, Shivannavar CT, Gaddad SM. Vancomycin resistance among methicillin resistant Staphylococcus aureus isolates from intensive care units of tertiary care hospitals in Hyderabad. Indian J Med Res. 2011;134(5):704.View ArticlePubMedPubMed CentralGoogle Scholar
- Rogers KL, Fey PD, Rupp ME. Coagulase-negative staphylococcal infections. Infect Dis Clin N Am. 2009;23(1):73–98.View ArticleGoogle Scholar
- Harris LG, El-Bouri K, Johnston S, et al. Rapid identification of staphylococci from prosthetic joint infections using MALDI-TOF mass-spectrometry. Int J Artif Organs. 2010;33(9):568–74.PubMedGoogle Scholar
- Dubois D, et al. Identification of a variety of Staphylococcus species by matrix-assisted laser desorption ionization-time of flight mass spectrometry. J Clin Microbiol. 2010;48(3):941–5.View ArticlePubMedPubMed CentralGoogle Scholar
- Rajakaruna L, et al. High throughput identification of clinical isolates of Staphylococcus aureus using MALDI-TOF-MS of intact cells. Infect Genet Evol. 2009;9(4):507–13.View ArticlePubMedGoogle Scholar
- Voorhees KJ et al. Modified MALDI MS fatty acid profiling for bacterial identification. J Mass Spectrom. 2013;48(7):850–5.View ArticlePubMedGoogle Scholar
- McAlpin CR, Voorhees KJ, Corpuz AR, Richards RM. Analysis of lipids: metal oxide laser ionization mass spectrometry. Anal Chem. 2012;84(18):7677–83.View ArticlePubMedGoogle Scholar
- Voorhees KJ, McAlpin CR, Cox CR. Lipid profiling using catalytic pyrolysis/metal oxide laser ionization-mass spectrometry. J Anal Appl Pyrolysis. 2012;98:201–6.View ArticleGoogle Scholar
- Voorhees KJ, Saichek NR, Jensen KR, Harrington PB, Cox CR. Comparison of metal oxide catalysts for pyrolytic MOLI–MS bacterial identification. J Anal Appl Pyrolysis. 2014. doi:10.1016/j.jaap.2014.10.016.Google Scholar
- Cox CR, Jensen KR, Saichek NR, Voorhees KJ. Strain-level bacterial identification by CeO2-catalyzed MALDI-TOF MS fatty acid analysis and comparison to commercial protein-based methods. Nat Sci Rep. 2015. doi:10.1038/srep10470.Google Scholar
- Harrington PB. Fuzzy multivariate rule‐building expert systems: minimal neural networks. J Chemom. 1991;5(5):467–86.View ArticleGoogle Scholar
- Harrington PB, Kister J, Artaud J, Dupuy N. Automated principal component-based orthogonal signal correction applied to fused near infrared− mid-infrared spectra of French olive oils. Anal Chem. 2009;81(17):7160–9.View ArticleGoogle Scholar
- Harrington PB. Statistical validation of classification and calibration models using bootstrapped Latin partitions. TrAC Trends Anal Chem. 2006;25(11):1112–24.View ArticleGoogle Scholar
- Edwards-Jones V, et al. Rapid discrimination between methicillin-sensitive and methicillin-resistant Staphylococcus aureus by intact cell mass spectrometry. J Med Microbiol. 2000;49(3):295–300.View ArticlePubMedGoogle Scholar
- Bernardo K, et al. Identification and discrimination of Staphylococcus aureus strains using matrix‐assisted laser desorption/ionization‐time of flight mass spectrometry. Proteomics. 2002;2(6):747–53.View ArticlePubMedGoogle Scholar
- Mishra NN, et al. Daptomycin resistance in enterococci is associated with distinct alterations of cell membrane phospholipid content. PLoS One. 2012;7(8):e43958.View ArticlePubMedPubMed CentralGoogle Scholar