MALDI-TOF Mass Spectrometry: Transformative Proteomics for Clinical Microbiology

Clinical microbiology laboratories today operate with the use of a blend of high- and low-technology procedures. Despite huge steps forward in diagnostic methods based on nucleic acid amplification, clinical microbiology laboratories still rely heavily on culturing bacteria and fungi for diagnostic purposes because of the plethora and diversity of potentially isolated organisms. This process is typically done on solid (e.g., blood agar) and occasionally in liquid (i.e., broth) media. Once plate growth is observed (the so-called “colony”), the challenge is accurate identification of that growth. Historically, this identification has been considered to be a complex task, and numerous textbooks have been written about it. Traditional (phenotypic) identification of bacteria and fungi involves examination of morphologic characteristics, staining, and conducting biochemical tests (either manually or as an automated panel). Manual biochemical assays (e.g., catalase, oxidase) may provide a rapid turnaround time, but they identify only a limited number of organism types. Automated systems (e.g., VITEK®, bioMérieux; BD Phoenix™ Automated Microbiology System; MicroScan®, Siemens Healthcare Diagnostics) identify a broader range of organisms, but suffer from longer turnaround time and expensive consumables. Furthermore, the user often must have a priori knowledge of the organism type (e.g., gram-negative bacillus) being tested. Introduction of MALDI-TOF mass spectrometry into the clinical microbiology laboratory has markedly altered workflow, allowing bacterial and fungal colonies to be accurately, rapidly, and inexpensively identified. Contemporary applications in clinical microbiology combine the powerful capabilities of mass spectrometry and bioinformatics to yield a novel methodology for bacterial and fungal identification. Utilization of MALDI-TOF mass spectrometry in the clinical microbiology laboratory is explained in its simplest form by discussing the routine workflow of a microbiology technologist. Starting from a single isolated colony on a culture plate, a colony is “picked” to a spot on a MALDI-TOF mass spectrometry plate, which is a solid, typically stainless steel plate that has a defined number of test spots (e.g., 24, 48, 96, 384). The organism on the plate is overlain with a small amount of matrix (e.g., 1–2 μL of α-cyano-4-hydroxycinnamic acid dissolved in 50% acetonitrile and 2.5% trifluoroacetic acid) and the plate is dried (typically for 5 min or less). Identification and reagent contamination controls and calibration standard(s) are tested on separate spots. The plate is then placed into the chamber of the MALDI-TOF mass spectrometer. The process is automated (following instrument calibration), with the mass spectrum of the test isolate generated and compared against a database of reference spectra. Software scores the relatedness of the test isolate's spectrum to the known spectra stored in the system library, resulting in a ranked (e.g., percentage or score) list of the possible organisms and a level of confidence in identification. There is no formal mass spectral interpretation required by the technologist because this task is performed by the software in an analogous fashion to running a Google search. Single or multiple isolates may be tested at a time. The analytical turnaround time is ≤3 min per isolate. Few disposables and reagents are required and the plates are reusable, which makes this a fairly “green” technology compared to more traditional biochemical-based automated instruments. Disposable MALDI-TOF mass spectrometry plates are available if desired by individual laboratories. How does MALDI-TOF mass spectrometry identify bacteria and fungi? The intact organism is embedded in a matrix, which assists in subsequent desorption and ionization of microbial proteins. The dried spot containing matrix and microorganism is hit by multiple laser shots. The matrix protects the microbial proteins, converts laser energy to excitation energy, and aids in desorbing (vaporizing) and ionizing microbial proteins, which are primarily highly abundant ribosomal proteins, although the analytes are not specifically identified. The ionized analytes enter a flight tube and are separated by their individual mass-to-charge ratios, although because the analytes are typically singly charged with MALDI mass spectrometry, separation is effectively by mass alone. The smallest analytes travel the fastest and hit the detector first, followed by progressively larger analytes. A mass spectrum is generated and represents the numbers of ions hitting the detector at the end of the flight tube at a specific time (time of flight). MALDI-TOF mass spectrometry is made clinically useful in microbiology by coupling the powerful analytical features with well-curated databases of spectra that represent a comprehensive collection of clinically relevant bacteria and fungi. Two commercial systems (which include commercial databases) are available, the Bruker Biotyper (Bruker Daltonics) and the VITEK® MS (bioMérieux). As with any identification system, the limitations of each database must be clear, especially in regard to organisms that were not captured or may have been incorrectly identified. In general, the databases available in the commercial systems are best for common aerobic gram-negative bacilli and gram-positive cocci, but may be less comprehensive for esoteric organisms (e.g., anaerobic bacteria). Given the rapidly changing nomenclature and descriptions of new species, consistent and routine updates to the commercial databases used in clinical laboratories will be imperative to ensure clinical utility into the future. Laboratories are able to compile their own mass spectral databases, assuming the requisite isolates are available. How well does microbial identification work with the use of MALDI-TOF mass spectrometry? Many investigators have attempted to answer this question. At the Mayo Clinic we compared an automated biochemical-based identification system (the BD Phoenix automated microbiology system) to the Bruker Biotyper mass spectrometry system for identification of gram-negative bacilli. The BD Phoenix correctly identified 363 (83%) and 330 (75%) of isolates to the genus and species levels, respectively, whereas the mass spectrometry system correctly identified 408 (93%) and 360 (82%) of isolates to the genus and species levels, respectively, statistically outperforming the comparator. The Bruker Biotyper was then evaluated for its ability to identify 298 clinical isolates of staphylococci, streptococci, and related genera (1). The MALDI-TOF mass spectrometry system correctly identified 284 (95%) and 207 (69%) of isolates to the genus and species levels, respectively. Subgroup analysis was performed for staphylococci, streptococci, and enterococci (n = 217) and “related genera” (n = 81). For the former subgroup, 213 (98%) and 171 (79%) of isolates were identified to the genus and species levels, respectively. In contrast, in the subgroup of “related genera,” only 71 (88%) and 36 (44%) of isolates were identified to the genus and species levels, respectively. The Bruker Biotyper was also validated for identification of Corynebacterium species, generally perceived as a more challenging group of organisms to identify to the species level compared to common gram-negative bacilli and gram-positive cocci. Eighty of 92 (87%) yielded a species-level identification, and all were accurate with the exception of Corynebacterium aurimucosum, which was misidentified as the closely related Corynebacterium minutissimum. Studies with the Biotyper system were also conducted for identification of 138 common and 103 archived yeast isolates, and 96% and 85% accurate species-level identifications were determined for common and archived yeasts, respectively. Further evaluation of dermatophytes (n = 171) by use of the Biotyper system identified 93% of dermatophyte isolates to the genus and 60% to the species level. Investigators from around the world have reported genus-level identifications of 96%–99% and species-level identifications of 85%–97% when testing routinely isolated bacteria and yeast using Bruker Biotyper MALDI-TOF mass spectrometry [Table 1 (2–5)]. MALDI-TOF mass spectrometry has been used to identify additional types of bacteria (e.g., anaerobic bacteria, mycobacteria) and fungi, although in some cases special processing and/or database enhancements were applied. Clinical laboratories looking to use this exciting new technology will need to choose between the Bruker Biotyper and VITEK® MS systems, although neither instrument has received clearance from the US Food and Drug Administration for testing patient isolates. The analytical principles underlying these 2 platforms are similar; however, there are fundamental differences in the algorithms used to identify organisms and in the databases. The Bruker MicroFlex LT mass spectrometer is a desktop instrument, whereas the VITEK® MS instrument is a larger floor model. The numeric values generated by each system's software are produced with different algorithms and reported on different scales, and are therefore not directly interchangeable or necessarily comparable. Despite these differences, the system performance characteristics and accuracy of isolate identification have remained similar within the limited studies performed to date. There are several issues which remain to be addressed before routine clinical use, including definition of the exact cutoff scores for reliable organism identification, development of robust QCs and standards, and a need for preparatory extraction (1). One limitation of this technology is that it does not provide antimicrobial susceptibility results. Strategies are in development, but these lack comprehensive susceptibility results and require testing and analysis methods that are different from those described earlier. The technology does not, with rare exceptions, apply to direct testing of clinical specimens—organisms must first be cultivated. Rapid identification of microorganisms growing in blood culture bottles continues to be a promising application, but there currently remains a requirement for preparatory specimen processing. Finally, application of mass spectrometry within the clinical microbiology laboratory requires chemical hygiene practices that are different from those traditionally encountered by microbiologists, although this issue is easily addressed with proper education and training of laboratory personnel. In summary, MALDI-TOF mass spectrometry is a quantum leap forward for identification of cultured bacteria and fungi in the clinical microbiology laboratory. Rapid and facile identification of bacteria and fungi by use of MALDI-TOF mass spectrometry is incontrovertibly beneficial to clinical laboratories, and the decrease in direct and indirect expenses (the reagent cost is approximately $0.25 per isolate tested, which is several-fold lower than traditional automated biochemical-based identification) is advantageous to the laboratory and patients. Whether or not implementation of MALDI-TOF mass spectrometry in clinical microbiology laboratories will result in a substantial improvement in patient outcome remains to be determined. I thank Scott A. Cunningham for his thoughtful review of this Perspective.