The main application of MALDI-TOF mass spectrometry in clinical microbiology is the identification of microorganisms by analysis of their total proteins (ribosomal proteins and proteins associated with membranes).
This technique can identify most bacteria in just a few minutes. The method is fast, precise, reliable and cost-effective compared to conventional phenotypic methods. Other applications of MALDI-TOF mass spectrometry are under development, such as the detection of bacterial toxins or antibiotic resistance mechanisms.
General principles to remember
The management of a patient suspected of bacterial infection is conventionally based on the identification of the pathogen at the site of infection and on the choice of the best antibiotic treatment, based on the antibiogram. The identification of a pathogen is therefore crucial, both to diagnose a bacterial infection but also to guide antibiotic therapy.
The laboratories develop and use reliable, ever faster methods for bacterial identification. The usual strategy for these identifications consists of several stages: after rapid simple orientation tests such as Gram staining or catalase and oxidase tests, phenotypic tests, based on the biochemical characteristics of bacteria, complete the ‘identification. For decades, this strategy was the only one applicable in a routine bacteriological laboratory.
The only notable evolution of these techniques over the years has been their miniaturization and automation. These techniques are expensive and always require several hours of incubation before obtaining a species identification. The methods used in routine laboratories guarantee reliable identification only for the species most frequently encountered in the clinic.
In most cases, bacterial identification takes place 48 hours after receipt of the sample; this period can be further extended in the case of bacteria or fungi with slow and / or difficult growth. For several years, molecular biology tests have enabled rapid bacterial identification, applicable to all microorganisms, including non-cultivable infectious agents.
Molecular biology tests using PCR techniques (Polymerase Chain Reaction ) unfortunately only identify the infectious agent that we suspect and for which we have a technique. In addition, these tests do not always make it possible to obtain a discriminating identification down to the species. Furthermore, the high cost and the high level of technical expertise required make molecular biology a technique currently unsuitable for routine identification.
Only a few tests are marketed for the rapid identification of specific pathogens, alone or in panels, and other applications using microarray technology have improved identification strategies in molecular biology. However, the cost of these tests and, sometimes, the workload remain high and represent limiting factors for their use in clinical microbiology laboratories. An innovative technique has recently appeared on the bacteriology market, allowing a routine laboratory not only to identify a large number of bacteria and fungi, more or less frequently encountered in the clinic, but also to carry out this identification in a way fast, reliable and inexpensive: mass spectrometry!
Mass spectrometry: MALDI-TOF
Mass spectrometry is a highly sensitive physical analysis technique, which has existed for almost a century, making it possible to detect and identify molecules of interest. We can diagram a mass spectrometer in 4 parts: the sample introduction system, the ionization chamber, producing ions in the gas phase, the analyzer, separating the ions according to their mass-to-charge ratio ( m / z) and the detector, transforming the ion current into electric current. Ionization is the most important step in identifying molecules.
MALDI-TOF mass spectrometry is based on an ionization technique, developed in the 80s, and leading to the identification of biomarkers of high molecular weight: it is a laser-assisted desorption-ionization by matrix (or MALDI: Matrix- Assisted Laser Desorption Ionization ). The sample to be analyzed is deposited on a target and is processed by an appropriate matrix. After introduction of the target into the system, it is bombarded by a laser.
The ions thus generated in the ionization chamber are accelerated in an electric field which directs them in a flight tube towards the analyzer. The latter makes it possible to separate and classify the accelerated ions according to their time of flight (TOF: Time-Of-Flight) and to produce a mass spectrum. The mass spectrum obtained is a kind of specific and unique fingerprint of the protein composition of the microorganism analyzed, which can be compared to a database of spectra.
Note that it was only recently that MALDI-TOF mass spectrometry was adapted as a fast, precise and inexpensive technique for the routine of microbiology laboratories. Currently, two manufacturers offer systems allowing the identification of microorganisms by MALDI-TOF mass spectrometry, with a fairly comparable mode of operation and quality: the B iotyper MALDI-TOF MS from Bruker and the AXIMA system from Shimadzu using the SARAMIS database of Anagnostec .
Importance of the database
Identification by MALDI-TOF is based on the following discoveries:
- the spectral imprint varies between microorganisms,
- among the compounds detected in the mass spectrum, certain peaks are specific to the genus, to the species and even in certain cases of the subspecies ,
- the spectra obtained are reproducible provided that the growth of the bacteria takes place under the same conditions. However, even if the culture conditions vary, many peaks are preserved: these are those which are potentially the most suitable for being used as specific biomarkers, allowing bacterial identification.
This identification is based on the comparison of the position of the peaks of the unknown mass spectrum with all the typical spectra recorded in the spectra database. A matching score ranks the spectra and specifies the most plausible bacterial identification (s), in order of probability. Spectrum libraries are supplied, validated and updated by the firms marketing the MALDI-TOF systems.
Currently, the Bruker’s spectra bank allows the identification of 3,476 cellular organisms: 3,216 bacteria (enterobacteria, non- fermenting Gram negative bacilli , staphylococci, streptococci, mycobacteria, anaerobic bacteria, etc.) and 260 fungi (Candida, filamentous fungi, …). This database, allowing the identification of pathogens of clinical interest and of microorganisms of the environment, can be enriched by the user: it is possible to introduce new spectra which will be added to those already configured by the constructor. It is important to know that bacterial species whose protein profile is comparable, even identical, cannot be discriminated by the MALDI-TOF technique.
Thus, the technique is not very effective for the correct identification of streptococcus viridans and for their discrimination with regard to Streptococcus pneumoniae for example. The identification of certain species such as streptococci, anaerobic bacteria or fungi could be improved by updating and enriching the database concerning them. There is theoretically no limit to the capacity for identification by MALDI-TOF mass spectrometry provided that the database contains sufficient adequate reference spectra.
Identification by MALDI-TOF MS
in a routine laboratory of medical microbiology The types of samples that can be analyzed by a MALDI-TOF MS system can be classified into 2 categories: bacterial cultures on agar, which are deposited directly on the target, and yeast cultures, filamentous fungi, mycobacteria, positive blood cultures and urine specimens, which require prior extraction.
Analysis of bacterial culture on agar
The bacterial colony is directly deposited in the form of a thin smear on the surface of a metal plate, the target, then covered with an appropriate matrix. Up to 96 strains per series can be studied for the Bruker system and 384 for the Shimadzu system . The target is then introduced into the MALDI-TOF system and, in less than 2 minutes, the first mass spectrum is produced and analyzed.
By comparison with the database, the computer software offers the most probable identification. Several studies have been published on the performance of MALDI-TOF MS systems for the identification of microorganisms in routine clinical microbiology laboratories. Seng et al. have shown in particular that, among the 1,660 bacterial isolates studied, representing 109 different species, 84.1% were correctly identified up to the level of the species and 11.3% up to the level of the genus.
According to the authors, the absence of identification (46 strains or 2.8%) or incorrect identification (28 strains, or 2.8%) are due to incorrect use of the database. In addition, the authors did not note any discrepancy between Gram staining and MALDI-TOF, suggesting that the latter could be used in the first line, even before Gram staining. All of the publications confirm the role of MALDI-TOF MS systems as a first-line identification tool in a routine laboratory, but also stresses the importance of updating the database.
Analysis of microorganisms after extraction
For the identification of certain microorganisms (yeasts, filamentous fungi, mycobacteria), prior protein extraction is recommended in order to produce an exploitable mass spectrum. The same is true for primary samples such as positive blood cultures and urine. This extraction step extends the handling time from 10 to 15 minutes per extract, compared to the direct deposition technique. In clinical microbiology laboratories, blood cultures always represent the most significant sample for the diagnosis of severe acute bacterial infections.
Automata, such as BactAlert ® ( bioMérieux ), or Bactec ® ( Becton Dickinson), continuously monitor bacterial growth in order to detect it early. Once the blood culture is positive, a rapid presumptive identification based on direct examination after Gram staining already makes it possible to approximately adapt the antibiotic therapy. At this stage, a complete identification, in routine, is generally carried out in 1 to 2 days but can be longer for tedious or atypical microorganisms. Molecular biology techniques (real-time PCR, microarrays , FISH) are being evaluated for the rapid detection of bacteria and their resistance mechanism directly in positive blood cultures .
Currently, these closed systems only allow the detection of a limited number of pathogens, they are expensive and often require special technical skills. MALDI-TOF mass spectrometry allows the identification of microorganisms present in blood cultures, routinely, on the day of their positivity, one to two days earlier than by a conventional phenotypic technique. An important extraction step is necessary: it consists in separating the bacteria from the other cellular components of the sample. Several extraction methods exist such as for example differential centrifugation or the lysis of cell membranes by a detergent .
According to La Scola and Raoult , there are however some limits: (i) for polymicrobial blood cultures (which are infrequent), only one of the species present, in the best of cases, can be identified, and (ii) the problem due to lack of discrimination between certain bacterial species. At the same time, using a concentration step, direct identification in the urine is achievable , without however allowing a quantification of bacteria. Reliable identification by MALDI-TOF MS will depend on the number of microorganisms present in the urine sample (> 105 CFU / ml at least) but also on the bacterial species. Furthermore, this technique does not seem to be practical in a routine laboratory given the large number of urine samples and the existence of diagnostic guidance techniques such as Gram staining or flow cytometry for the analysis of urinary figurative elements.
Other applications and developments
MALDI-TOF MS technology, by producing spectra specific to each bacteria, also makes it possible to envisage the epidemiological study of microorganisms. Far from matching the performance of genotypic typing techniques, it is however possible, with a MALDI-TOF MS system, to compare the protein profiles of different bacterial strains and thus already draw some conclusions as to their possible homology . It is conceivable, for example, to compare multiresistant Acinetobacter baumanii isolated in different hospital units and, if their spectra have exactly the same peaks, to suspect that they belong to the same epidemic clone.
Furthermore, from the similarity scores of the spectra, a dendrogram can be established making it possible to link the closely related species within a given bacterial genus, or to visualize among the same species the most comparable strains.
Detection of markers of antibiotic resistance mechanisms
In addition to its routine use for the identification of microorganisms, the detection of antibiotic resistance mechanisms is of particular interest. Some studies have shown the ability of MALDI-TOF mass spectrometry to distinguish Staphylococcus aureus resistant to methicillin (MRSA) that bear the gene mecA encoding a specific protein, PBP2a, strains methicillin -sensitive (MSSA). The detection of a resistance mechanism involving characteristic proteins (beta- lactamase , methylase , efflux system for example) is therefore possible and under investigation.
Detection of virulence markers
The search and detection of pathogenic strains producing particular virulence factors associated with specific proteins, such as toxins for example, could also help the clinician in the management of infections. Bittar et al. found a peak, a marker of the presence of leukocidin of Panton and Valentine (PVL toxin) which differentiates PVL producing S. aureus strains from those that do not occur. This approach seems promising but, as for the detection of antibiotic resistance, it requires development to allow its clinical application.