Anal. Chem. 2008, 80, 2026-2034
MALDI-TOF Mass Spectrometry Compatible Inactivation Method for Highly Pathogenic Microbial Cells and Spores Peter Lasch,*,† Herbert Nattermann,‡ Marcel Erhard,⊥ Maren Sta 1 mmler,† Roland Grunow,‡ § ‡,# † Norbert Bannert, Bernd Appel, and Dieter Naumann
P25, ZBS2, and ZBS4, Robert Koch-Institut, Nordufer 20, 13353 Berlin, Germany, and AnagnosTec GmbH, Gesellschaft fu¨r Analytische Biochemie und Diagnostik mbH, Am Mu¨hlenberg 11, D-14476 Potsdam-Golm, Germany
Identification of microorganisms, specifically of vegetative cells and spores, by intact cell mass spectrometry (ICMS) is an emerging new technology. The technique provides specific biomarker profiles which can be employed for bacterial identification at the genus, species, or even at the subspecies level holding the potential to serve as a rapid and sensitive identification technique in clinical or food microbiology and also for sensitive detection of biosafety level (BSL) 3 microorganisms. However, the development of ICMS as an identification technique for BSL-3 level microorganisms is hampered by the fact that no MALDI-TOF (matrix-assisted laser desorption/ionization time-of-flight) compatible inactivation procedure for microorganisms, and particularly for bacterial endospores, has been evaluated so far. In this report we describe a new methodology for effective inactivation of microorganisms which is compatible with the analysis of microbial protein patterns by MALDI-TOF mass spectrometry. The main challenge of this work was to define the conditions that ensure microbial inactivation and permit at the same time comprehensive analysis of microbial protein patterns. Among several physical, chemical, and mechanical inactivation procedures, inactivation by trifluoroacetic acid (TFA) proved to be the best method in terms of bactericidal capacity and information content of the mass spectra. Treatment of vegetative cells by 80% TFA alone for 30 min assured complete inactivation of microbial cells under all conditions tested. For spore inactivation, the “TFA inactivation protocol” was developed which is a combination of TFA treatment with basic laboratory routines such as centrifugation and filtering. This MALDITOF/ICMS compatible sample preparation protocol is simple and rapid (30 min) and assures reliable inactiva* Corresponding author. E-mail:
[email protected]. Phone: +49-30-45472405. Fax: +49-30-45472606. † P25, Robert Koch-Institut. ‡ ZBS2, Robert Koch-Institut. § ZBS4, Robert Koch-Institut. ⊥ AnagnosTec GmbH. # Present address: Bundesinstitut fu ¨r Risikobewertung, Thielallee 88-92, 14195 Berlin, Germany.
2026 Analytical Chemistry, Vol. 80, No. 6, March 15, 2008
tion of vegetative cells and spores of highly pathogenic (BSL-3) microorganisms. Direct profiling of microorganisms by intact cell mass spectrometry (ICMS) has emerged as a valuable research tool to obtain specific information on the cell’s total molecular composition.1-7 In ICMS it is not necessary to employ a separation stage (i.e., liquid chromatography (LC) or electrophoresis), and the sample can be analyzed “as a whole” with only minimal sample pretreatment.8 Upon ICMS analysis of microbial cells, or cell extracts, a multiplicity of peptides and proteins is simultaneously detected. The mass spectra of whole cells provide a snapshot of the different protein compositions of individual microbial strains and thus constitute strain-specific suites of biomarkers. Although a molecular identification of individual mass peaks by MS/MS techniques is principally possible, ICMS spectra are frequently considered as spectral fingerprints which do not require complete understanding of the samples’ biochemistry. According to the fingerprint approach, ICMS mass patterns of unknown bacteria can be identified by pattern matching against a validated database of microbial reference spectra.9,10 Alternative concepts of microbial identification and detection rely on proteomic information by searching protein databases5 or on a comparison of the experimental mass spectra with protein masses predicted from microbial genomes.11,12 All these techniques have in common that a multitude of bacterial mass peaks is analyzed which allows (1) Anhalt, J. P.; Fenselau, C. Anal. Chem. 1975, 47, 219-225. (2) Mass Spectrometry for the Characterization of Microorganisms; Fenselau, C., Ed.; ACS Symposium Series 541; American Chemical Society: Washington, DC, 1994. (3) Krishnamurthy, T.; Ross, P. L. Rapid Commun. Mass Spectrom. 1996, 10 (15), 1992-1996. (4) Claydon, M. A.; Davey, S. N.; Edwards-Jones, V.; Gordon, D. B. Nat. Biotechnol. 1996, 14 (11), 1584-1586. (5) Demirev, P. A.; Ho, Y. P.; Ryzhov, V.; Fenselau, C. Anal. Chem. 1999, 71 (14), 2732-2738. (6) Wilkins, C. L.; Lay, J. O. Identification of Microorganisms by Mass Spectrometry; John Wiley and Sons: Hoboken, NJ, 2005. (7) Hathout, Y.; Demirev, P. A.; Ho, Y. P.; Bundy, J. L.; Ryzhov, V.; Sapp, L.; Suttler, J.; Jackman, J.; Fenselau, C. Appl. Environ. Microbiol. 1999, 65 (10), 4313-4319. (8) Fox, A. J. Clin. Microbiol. 2006, 44 (8), 2677-2680. (9) Holland, R. D.; Wilkes, J. G.; Rafii, F.; Sutherland, J. B.; Persons, C. C.; Voorhees, K. J.; Lay, J. O., Jr. Rapid Commun. Mass Spectrom. 1996, 10 (19), 1227-1232. (10) Arnold, R. J.; Reilly, J. P. Rapid Commun. Mass Spectrom. 1998, 12 (10), 630-636. 10.1021/ac701822j CCC: $40.75
© 2008 American Chemical Society Published on Web 02/22/2008
accurate typing of microorganisms at the genus, species, and sometimes even at the subspecies level. In microbiology, the matrix-assisted laser desorption/ionization time-of-flight ICMS (MALDI-TOF/ICMS) technique has a number of potential advantages over alternative typing methods. First, sample preparation for ICMS is comparably simple and can be carried out within minutes. Second, the technique does not rely on taxon-specific or cost-intensive consumables such as antibodies. The ICMS workflow is simple, fast, and can be standardized for most of the bacterial species. Furthermore, many of the procedures for sample preparation, data acquisition, and evaluation can be automated. The information content of the ICMS spectra was thus employed in numerous studies in clinical, food, and environmental microbiology, taxonomical research, and tested also as a rapid detection technique for highly pathogenic bacteria as potential agents in bioterrorist attacks.13-15 One of the disadvantages of current ICMS approaches that hampered broad application lies in the fact that no MALDI-TOF/ ICMS compatible inactivation procedure for pathogenic microorganisms, and particularly for bacterial endospores, is available so far. Yet, inactivation is considered to be crucial for utilization of an ICMS-based methodology to rapidly detect naturally or intentionally released biosafety level 3 (BSL-3) microorganisms. Within the scope of the present study we have therefore systematically investigated a number of well-established techniques for the inactivation of microorganisms and tested the compatibility of the methods with MALDI-TOF mass spectrometry. Among these, inactivation by trifluoroacetic acid (TFA) proved to be the best technique in terms of bactericidal capacity and information content of the mass spectra. In this study we present quantitative data on the TFA inactivation tests for vegetative cells and bacterial endospores and give a comparison of the ICM spectra from untreated and inactivated microorganisms. Furthermore, the spectral reproducibility was investigated and an example of the discriminative potential of the new MALDITOF sample processing procedure is given. EXPERIMENTAL SECTION Microbial Strains and Isolates. Most Yersinia strains investigated in this study originated from the strain collection at the Robert Koch-Institute (RKI) and from a clinical study performed by AnagnosTec GmbH, Potsdam. The collection of Yersinia strains was further complemented by reference strains from the German Collection of Microorganisms and Cell Cultures GmbH (DSMZ, Braunschweig, Germany). With the exception of two strains of Bacillus anthracis, which were a kind gift of W. Beyer (Universita¨t Hohenheim, Germany), all Bacillus strains originated from the strain collection at RKI. An overview of the strains and isolates used is given in Supporting Information Table S-1. Bacterial cultures were prepared by growing each strain for two passages under aerobic conditions on Caso agar (Merck KGaA, Darmstadt, (11) Demirev, P. A.; Lin, Y. S.; Pineda, F. J.; Fenselau, C. Anal. Chem. 2001, 73, 4566-4573. (12) Warscheid, B.; Fenselau, C. Anal. Chem. 2003, 75, 5618-5627. (13) Ryzhov, V.; Hathout, Y.; Fenselau, C. Appl. Environ. Microbiol. 2000, 66 (9), 3828-3834. (14) Elhanany, E.; Barak, R.; Fisher, M.; Kobiler, D.; Altboum, Z. Rapid Commun. Mass Spectrom. 2001, 15 (22), 2110-2116. (15) Castanha, E. R.; Fox, A.; Fox, K. F. J. Microbiol. Methods 2006, 67 (2), 230-240.
Germany) for 48 h at 28 °C (Yersinia) or on LB (Luria-Bertani) agar (RKI, Berlin, Germany) for 24 h at 37 °C (Bacillus, Burkholderia, Escherichia), respectively. Cells were harvested by transferring an equivalent of three wire loops from each agar plate to 20 µL of water. If not otherwise stated, the concentration of colony forming units (CFU/mL) given throughout the paper refers to these suspensions. Bacilli spores were prepared as detailed in ref 16. The purity of spore preparations was determined as outlined in the same reference.16 The determination of cell counts (enumeration) of vegetative cells and spores was carried out as described in the European norms EN 13727 and EN 14347, respectively.16,17 Electron Microscopy. All steps in the preparation of the samples were performed at room temperature unless indicated otherwise. For transmission electron microscopy (TEM), spores were either treated for 30 min in 80% TFA, or fixed for 2 h in a solution of 10% p-formaldehyde (FA), 0.05% glutaraldehyde (GA) in 0.05 M HEPES buffer pH 7.2 (control treatment). Spores of B. atrophaeus DSM 2277 were spun down and embedded in 3% lowmelting-point agarose. The polymerized agarose containing the spores was cut into small blocks for staining. After 1 h of fixation in a solution of 2.5% GA, 0.05 M HEPES, pH 7.2, the blocks were rinsed in HEPES buffer and stained with 1% osmium tetroxide for 1 h. After washing with distilled water, the spores were further stained with 2% uranyl acetate for 1 h. The samples were then dehydrated in a graded ethanol series and embedded in LR white (London Resin Company, U.K.) starting the infiltration with a 1:1 mixture of LR white and ethanol (for 20 min) followed by three 20 min steps with pure LR white. After transfer into new LR white, the overnight polymerization at 60 °C was started. Ultrathin sections of the embedded samples were prepared and placed on 400 mesh copper grids. Sections were subsequently stained with 2% uranyl acetate (20 min) and lead citrate (3 min) and stabilized with carbon by carbon evaporation (BAE 250, Bal Tec, Liechtenstein). Sections were examined under a Zeiss EM 902 electron microscope operated at 80 kV. Images were taken with a slow scan CCD camera (Proscan, Scheuering, Germany). Inactivation with Trifluoroacetic Acid. For TFA inactivation, a volume of 20 µL of the aforementioned microbial suspension was carefully mixed with 80 µL of pure TFA (Uvasol, Merck KGaA, Darmstadt, Germany). Mixtures were prepared in screw-capped Eppendorf tubes. It was important to gently shake the inoculates to ensure adequate mixing (at room temperature). After 30 min of inoculation the solutions were diluted 10-fold by HPLC grade water (Mallinckrodt Baker B.V., Deventer, Netherlands). For MALDI-TOF/ICMS, 2 µL of the diluted sample solution were mixed with 2 µL of a 12 mg/mL R-cyano-4-hydroxycinnamic acid (HCCA, Bruker Daltonics, Bremen, Germany) solution. Preparation of the HCCA solution was achieved by dissolving HCCA in TA2, a 2:1 (v/v) mixture of 100% acetonitrile and 0.3% TFA. One microliter of the sample/HCCA mixture was spotted onto ground steel sample targets from Bruker Daltonics. (16) Chemical Disinfectants and AntisepticssBasic Sporicidal ActivitysTest Method and Requirement (phase 1), German version EN14347:2005; DIN Deutsches Institut fu ¨ r Normierung e.V. Beuth Verlag GmbH: Berlin. (17) Chemical Disinfectants and AntisepticssQuantitative Suspension Test for the Evaluation of Bacteriacidal Activity of Chemical Disinfectants for Instruments Used in the Medical AreasTest Method and Requirements (Phase 2, Step 1), German version EN 13727:2003; DIN Deutsches Institut fu ¨ r Normierung e.V. Beuth Verlag GmbH: Berlin.
Analytical Chemistry, Vol. 80, No. 6, March 15, 2008
2027
Direct ICMS Analysis. The direct analysis method included the deposition of a small amount of the microbial sample directly from the culture plates onto MALDI ground steel targets (Bruker Daltonics), drying at ambient temperature, and subsequent covering of the sample material by 1 µL of a saturated HCCA matrix solution MALDI-TOF Mass Spectrometry. ICMS profiles of microbial extracts from vegetative cells or spores were acquired using a Bruker Daltonics Autoflex mass spectrometer (Bruker Daltonics). The instrument was equipped with a nitrogen UV laser (337 nm), which operates at pulse rates up to 20 Hz. In order to optimize peak intensities the laser was slightly defocused. Pulse ion extraction time was set to 200 ns. Measurements were carried out in linear mode using an acceleration voltage of 20.00 (ion source 1) or 18.45 (ion source 2) kV. Lens voltage was 6.70 kV. Mass spectra were stored in the m/z range between 2000 and 20 000 Da. Insulin (5734.52 Da), ubiquitin I (8565.89 Da), cytochrome c (12 360.97 Da), and myoglobin (16 952.31 Da) were used as external calibrants enabling a mass accuracy of about 300 ppm. At least 600 individual laser shots were coadded for each spectrum. Data Analysis. Data acquisition and data analysis was carried out employing Bruker’s Compass 1.1 software package (Bruker Daltonics) and Matlab-based software developed in-house (The Mathworks Inc., Natick, MA). First steps of spectral analysis consisted in spectral preprocessing routines such as smoothing, baseline correction, and intensity normalization. For advanced visualization of the microbial biomarkers, the Matlab routines allowed also the generation of a simulated gel view from sets of preprocessed mass spectra. RESULTS In the following we will present the results of inactivation tests carried out under consideration of the following criteria: (1) The sample preparation technique should reliably inactivate vegetative cells and spores even at very high concentrations. To give an example, the preparation in the Daschle letter in October 2001 has been reported to contain extraordinarily high numbers of B. anthracis endospores with an estimated concentration of 10111012 CFU/g.18,19 Understandably, the analysis and handling of such intentionally released BSL-3 level agents, or their products, require either BSL-3 facilities or complete inactivation if samples from contaminated areas are prepared for dispatch to standard laboratories. (2) In order to be of practical use for MALDI-TOF/ICMS, the inactivation method should preserve the structural integrity of the biomolecules on which the ICMS technique is based. Thus, the new inactivation procedure should completely prevent germination and at the same time avoid destruction of the structural integrity of microbial proteins. Within the scope of the present work we have tested a number of inactivation methods: heat inactivation (autoclavation), radiation (UV and γ-irradiation), inactivation by FA, peracetic acid (PAA), or TFA. Among these methods, inactivation by concentrated TFA turned out to be a simple, efficient, and MALDI-TOF/ICMS compatible inactivation procedure. A systematic comparison with an interpretation of the MALDI-TOF spectra obtained by the other inactivation methods will be given in a separate publication. (18) Kennedy, H. New York Daily News, Oct 31, 2001, p 5. (19) Bartlett, J. G.; Inglesby, T. V., Jr.; Borio, L. Clin. Infect. Dis. 2002, 35, 851858.
2028
Analytical Chemistry, Vol. 80, No. 6, March 15, 2008
Neutralization of TFA. According to regulations of the European Committee for Standardization for “Chemical Disinfectants and AntisepticssBasic Sporicidal Activity” (EN 14347)16 and a “Quantitative Suspension Test for the Evaluation of Bactericidal Activity of Chemical Disinfectants for Instruments Used in the Medical Area” (EN 13727),17 quantitative microbiological tests for potential survival require complete neutralization of the disinfectant. The goal of neutralization is to stop antimicrobial activity of the disinfectant and thus to obtain precise numbers of surviving bacteria. Neutralization is usually achieved at the end of a specific contact time between the microorganisms and the disinfectant by transferring aliquots of the suspensions into neutralization tubes. The tubes contain the neutralizer and a cultivation medium and should serve for complete neutralization of the bactericidal and/or bacteriostatic activity of the disinfectant. Basic principles of the main experiment used here to quantitatively test the survival rates after treatment by the disinfectant are given in Supporting Information Figure S-1 (see column “main experiment” to the left). The tests included incubation of vegetative cell, or spore suspensions, at room temperature in TFA by mixing the microbial suspension with TFA. After a defined time of incubation the suspensions were diluted 10-fold, and a volume of 100 µL was transferred into a neutralization tube containing 9.9 mL of tryptic soy broth (TSB) and the neutralizer. The composition of the neutralizer was established in a separate series of experiments. We found that neutralization media containing 9% Tween 80, 0.9% lecithin, 3% histidine, and TSB allowed complete neutralization of the antimicrobial activity of TFA. The experimental protocol for testing the effectiveness of disinfectants also included the determination of spore or cell numbers in absence of the disinfectant (see central column of Supporting Information Figure S-1) and a toxicity control of the neutralizer itself (see right column of Supporting Information Figure S-1). These supplementary experiments permitted the calculation of disinfectant-specific (logarithmic) reduction factors (RFs) and demonstrated that the neutralizer alone did not inhibit bacterial growth (data not shown). Time and Concentration Dependence of TFA Inactivation of Spores. The main goal of these tests was to precisely define experimental conditions at which microorganisms are effectively inactivated by TFA. Because endospores are extremely resistant to a variety of harsh treatments,20-23 the optimization of TFA inactivation parameters were carried out on spores. Among the tested spore preparations, B. cereus ATCC 10987 turned out to be the most resistant strain, and thus it has been used for optimization. In these experiments, TFA concentrations of 25%, 50%, and 80%, respectively, were adjusted in spore suspensions of an original concentration of approximately 3.8 × 109 CFU/mL. The inactivation process was stopped after 5, 15, or 30 min of incubation, respectively, by mixing 100 µL of the TFA/spore suspension with 9.9 mL of the neutralization medium. Recultivation was carried out by using 0.2 mL aliquots according to the “spattle method” on tryptic soy agar (TSA, 48 h growth time). After (20) Setlow, P. Environ. Mol. Mutagen. 2001, 38 (2-3), 97-104. (21) Nicholson, W. L.; Munakata, N.; Horneck, G.; Melosh, H. J.; Setlow, P. Microbiol. Mol. Biol. Rev. 2000, 64 (3), 548-572. (22) Setlow, B.; Loshon, C. A.; Genest, P. C.; Cowan, A. E.; Setlow, C.; Setlow, P. J. Appl. Microbiol. 2002, 92 (2), 362-375. (23) Setlow, P. J. Appl. Microbiol. 2006, 101 (3), 514-525.
Table 1. Inactivation of Vegetative Cells by 80% TFA (30 min Contact Time, No Filtration)a species
no. of strains
no. of inactivation tests
cases with survival
Y. enterocolitica Y. pseudotuberculosis Y. pestis Burkholderia spp. Bacillus spp. (16) B. anthracis
56 14 2 12 49 7
156 42 2 14 209 7
0 0 0 0 0 0
∑:
140
430
0
a In all experiments the concentration of microorganisms was above 109 CFU/mL. Survival was tested by recultivation in enrichment media (14 days) according to the European norm EN 13727 (ref 17).
Table 2. Inactivation of Bacillus Spores by TFA (80%, 30 Min) without Filtrationa Figure 1. Logarithmic reduction factors (RFs) obtained from inactivation experiments for spores from B. cereus ATCC 10987 by trifluoroacetic acid (TFA) as a function of time and concentration. Initial spore concentration was 3.8 × 109 CFU/mL. Survival rates were obtained by transferring 0.1 mL of a spore/TFA mixture into 9.9 mL of a neutralizer and recultivation of 0.2 mL aliquots for 48 h. Each spore killing curve is an interpolation based on the means of three independent inactivation experiments. Bars represent the original data value range (see also text for details).
determination of CFU numbers, logarithmic RFs were determined for each time point and TFA concentration. The results of the TFA inactivation tests are given in Figure 1. In this figure, each of the survival curves depicts the mean, minimum, and maximum values of the logarithmic RFs of three independent inactivation tests. According to these data, the RF never dropped below a value of 5 indicating that TFA is a highly efficient chemical for spore inactivation. The curves also demonstrate that the efficiency of inactivation correlates with the TFA concentration, with the maximal spore killing effect in the first 5 min. With an average of 6.5, the highest RF was found after 30 min of incubation with TFA 80%. The curves also suggest that further increase of TFA concentration, or incubation time, would not result in a significant increase of the TFA killing rate. Thus, a TFA concentration of 80% and an incubation time of 30 min are considered a reasonable tradeoff between the spore killing rate and experimental efforts that would be still acceptable in the laboratory routine, or more importantly, in an emergency situation. TFA Inactivation of Vegetative Cells. Inactivation tests for vegetative cells of the genera Bacillus, Burkholderia, Escherichia, and Yersinia followed in all details the principles already described for spores (vide supra). In 430 separate experiments on the example of 140 different bacterial strains we could show that treatment with 80% TFA for 30 min resulted in complete inactivation of all strains tested (see Table 1). In all of the conducted tests the cell concentration exceeded the value of 109 CFU/mL. TFA Inactivation of Bacilli Spores. Spore suspensions were also deactivated by 80% TFA at a contact time of 30 min. Survival was tested after neutralization of the TFA by recultivation in enrichment media at day 1 and after 14 days of growth (EN 14347). At spore concentrations of 109 CFU/mL the inactivation tests yielded complete absence of bacterial growth of the following species: B. licheniformis, B. thuringiensis, and B. anthracis (see
species
concn [CFU/mL]
no. of strains
no. of inactivation tests
cases with survivals
B. anthracis B. cereus B. cereus B. licheniformis B. subtilis B. thuringiensis
109 107 109 109 109 109
2 2 2 1 1 1
8 6 13 5 7 6
0 0 10 0 1 0
9
45
11
∑:
a Survival was tested by recultivation from enrichment media at day 1 and after 14 days according to the European norm EN 14347 (ref 16). Survival of spores was found only for the strains B. cereus ATCC 10987 and B. subtilis DSM 347 in suspensions of 109 CFU/mL.
Table 2). However, occasional survival of spores of B. cereus strain ATCC 10987 and B. subtilis DSM 347 in suspensions of 109 CFU/ mL was observed. At lower concentrations (107 CFU/mL) the TFA treatment resulted in complete spore killing (see Table 2). In conclusion, we found that treatment of microorganisms with 80% TFA reliably inactivated vegetative cells andsat concentrations of 107 CFU/mL and smallersalso spores. This is quite efficient, but it must be noted that at concentrations larger than 107 CFU/mL TFA treatment alone does not always ensure complete elimination of spore viability. We have therefore evaluated supplementary methods that can be easily combined with the TFA treatment. Reduction of the Cell Concentration in the Supernatant by Centrifugation. Decanting the supernatant after centrifugation is known as a simple and effective method to reduce the cell count of cell suspensions. We have thus determined spore reduction factors for a selection of bacterial species in 25 individual centrifugation experiments (see Table 3). For practical reasons, the determination of precise reduction factors was carried out in water. Centrifugation time was 20 min applying a g-force of 16.000g. The logarithmic RFs of spores in the supernatant are given in Table 3 and varied between 2.3 (spores of B. anthracis) and 5.2 (B. subtilis). These results indicated that original spore concentrations of 108-1010 CFU/mL are reduced by a factor of at least 102.3 ∼ 200. Consequently, the resulting cumulative RF of the combination of TFA treatment (>6.5) and centrifugation (>2.3) is larger than 8.8. It can be estimated that this high level of sporicidal capacity will be adequate in any real-world scenario when, for Analytical Chemistry, Vol. 80, No. 6, March 15, 2008
2029
Table 3. Reduction of Viable Spore Counts in the Supernatant by Centrifugation (16.000g) for 20 min for Various Species and Initial Spore Concentrationsa species
concn [CFU/mL]
no. of strains
no. of expts
reduction factor [∆log CFU]
B. anthracis B. cereus B. cereus B. licheniformis B. licheniformis B. pseudomycoides B. subtilis B. subtilis B. subtilis B. thuringiensis
108 108 109 109 1010 109 108 109 1010 108
2 4 3 1 1 1 2 2 2 1
2 8 3 2 2 1 2 2 2 1
2.4 (2.3-2.6) 3.4 (2.7-4.3) 4.3 (3.6-4.9) 4.0 3.3 (2.7-3.9) 4.0 3.9 (3.7-4.0) 3.9 (3.8-4.0) 5 (4.8-5.2) 2.5
19
25
∑:
a The reduction factors are given in a logarithmic scale (∆log CFU) as average numbers with min/max values.
Table 4. Spore Inactivation by the TFA Inactivation Protocola species/strain B. anthracis 527 B. anthracis B 22/39 B. anthracis B 11/38 B. atrophaeus DSM 675 B. atrophaeus DSM 2277 B. cereus ATCC 14579 B. cereus ATCC 10987 B. licheniformis DSM 13 B. licheniformis DSM 13 B. licheniformis ATCC 12759 B. megaterium DSM 90 B. pseudomycoides DSM 1244 B. subtilis ATCC 6633 B. subtilis DSM 347 B. thuringiensis DSM 350 ∑:
concn no. of cases with [CFU/mL] inactivation tests survival 1.0 × 109 4.0 × 108 3.0 × 108 1.4 × 109 7.5 × 109 7.0 × 109 1.0 × 109 1.0 × 1010 2.3 × 1010 2.2 × 109 1.5 × 109 1.0 × 109 2.0 × 1010 2.9 × 1010 3.0 × 109
1 1 1 5 7 4 6 4 2 6 6 6 6 6 6
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
67
0
a Spore suspensions of 14 strains were reliably sterilized in 67 separate inactivation tests by the TFA inactivation protocol which involves initial treatment of the microorganisms by TFA, 80% for 5 min, centrifugation of the TFA suspension for 20 min at 16.000g, and sterile filtration of the supernatant. Suspensions were recultivated for 14 days after neutralization of the TFA in enrichment media.
example, the infectious dose for the inhalational form of anthrax is considered (8000-50 000 spores).24-26 Although the factual RF will be in most of the cases much larger than 8.8, occasional survival of a few spores can still not completely ruled out. It is, however, anticipated that survival of a few spores that germinate only after neutralization of TFA in enrichment media does not represent an imminent risk of infection. On the other hand, since German legislation prohibits any release of potentially infectious agents from BSL-3 laboratories, we suggest here a third supplemental procedure which should finally rule out all cases of residual spore survival. Sterile Filtration. Subsequent to centrifugation, the supernatant can be carefully aspirated and transferred for filtering into Millipore’s Ultrafree MC filter tubes of a pore size of 0.22 µm (24) Brachman, P. Ann. N.Y. Acad. Sci. 1980, 353, 83-93. (25) Jemski, J. V. DTIC Recovery No. AD 1963, 498-288. (26) Albrink, W. S.; Goodlow, R. J. Am. J. Pathol. 1959, 35, 1055-1065.
2030
Analytical Chemistry, Vol. 80, No. 6, March 15, 2008
(Millipore, Carrigtwohill, Ireland). Typical spin time for filtering was 3 min at 12 000g. Ultrafree filter tubes contain a Durapore PVDF (poly(vinylidene fluoride)) membrane which exhibits a low protein binding capacity. For us it was important that the filter material turned out to be resistant to highly concentrated TFA. Tests of several batches of TFA pretreated Ultrafree MC filters demonstrated structurally intact filter membranes with a preserved ability to retain spores. Furthermore, MALDI-TOF control measurements of filtrates of a 80% TFA/20% water mixture did not provide spectral contaminations from the filter tubes or the PVDF filter material (data not shown). The TFA Protocol: A Combination of TFA Treatment, Centrifugation, and Filtration. In the previous sections we have presented three MALDI-TOF/ICMS compatible techniques that each served the purpose to reduce the probability of release of microorganisms: TFA treatment (i), centrifugation (ii), and filtration of the supernatant (iii). The combination of these techniques is supposed to completely inactivate any type of bacteria, including bacterial endospores, even at concentrations which are expected in preparations of weaponized anthrax. In the following the combination of these basic laboratory procedures will be referred to as the “TFA inactivation protocol”: (i) An initial treatment of cells, or spores, in TFA, 80% for 5 min. Adequate mixing and shaking is required. (ii) Centrifugation of the microbial suspensions for 20 min at 16.000g. (iii) Sterile filtration of the supernatant by TFA resistant spore filters. Although the cumulative reduction factors of these three procedures strongly suggest that the TFA inactivation protocol is applicable for reliable and ICMS compatible inactivation, experimental data are required that unequivocally demonstrate this conclusion. The results of these inactivation experiments are summarized in Table 4. In 67 independent tests, spore suspensions of 14 bacterial strains could be completely sterilized by applying the TFA inactivation protocol. None of the cultivation trials showed signs of spore survival after neutralization of the TFA and recultivation for 14 days in enrichment media. Moreover, it could be shown that the inactivation technique is effective also for very high spore concentrations of up to 2.9 × 1010 CFU/mL (see Table 4). Hence, the combination of TFA treatment, centrifugation, and filtration of the supernatant represents an exceptionally efficient inactivation technique which assures complete inactivation of vegetative cells and spores even at very high concentrations. A schematic workflow overview of the TFA protocol is given in Figure 2. Once highly pathogenic microorganisms are sterilized in biosafety laboratories, or in contaminated areas, the sterile TFA extracts can be further processed for MALDI-TOF/ICMS analysis using standard laboratory equipment and facilities. Alternatively, the inactivation protocol would be also appropriate for preparing field samples in contaminated environments for dispatch to specialized laboratories. In this context it is furthermore important to note that TFA-treated microbial suspensions did not exhibit time-dependent spectral changes over several weeks. Comparison of ICM Spectra from Viable Microorganisms and TFA Extracts. MALDI-TOF mass spectra of Figure 3 are given for comparative purposes and were obtained by applying either the so-called direct ICMS measurement technique (see the
Figure 2. Schematic overview on the procedures for MALDI-TOF/ ICMS compatible inactivation of vegetative cells and spores.
“Experimental Section” or ref 27) or sample preparation by the TFA inactivation protocol. In this figure mass spectra originate from viable cells of Escherichia coli RKI A139 (traces A and C) and from spores of B. subtilis DSM 347 (traces B and D). The systematic comparison of the spectra of Figure 3 was carried out on the basis of peak lists compiled by an objective Matlab-based routine. This routine was used to determine the 30 most intense mass peaks in the m/z range of 2-20 kDa. The
comparison of the peak tables, but also the visual inspection of the spectra, demonstrated a high degree of consistency between the respective peak patterns. We found 77% (23/30) of the E. coli mass peaks and 57% (17/30) of the B. subtilis mass signals independent of the sample preparation technique (data not shown). Description of Electron Micrographs of Sectioned Spores of B. atrophaeus. In order to demonstrate the TFA-induced degradation of spores and to get more insight into the molecular mechanism of TFA inactivation we have investigated the effects of spore treatment by TFA also by means of electron microscopy (EM). At the Robert Koch-Institute, diagnostic EM is routinely applied in the Centre for Biological Safety as a rapid identification method for infectious agents (spores, viruses). This includes also EM analyses of “white powders” which are oftentimes contained in hoax anthrax letters. Mainly for this purpose an EM compatible inactivation method was developed which is supposed to conserve the structural features, i.e., the morphology of spores and viruses.28 The method is based on a 2 h treatment of the substrate with a 10% FA, 0.05% GA buffer and will be referred to in the following as the EM control method. The comparison of spore appearance after FA/GA control treatment and TFA inactivation is given in the electron micrographs of Figure 4. The micrographs in Figure 4, parts B and D, display sectioned spores of B. atrophaeus DSM 2277 after control treatment; the TFA-induced morphology changes from the same spore batch are depicted in panels A and C. By comparison of the TFA-treated and the control spores it may be inferred that TFA causes the loss of material of the spore’s core, presumably due to extraction of core materials. Furthermore, substantial changes of the morphology of cortex and inner coat are observed which are accompanied by a remarkable increase of the diameter of TFA-treated spores. The outer coat seems to be less heavily damaged.
Figure 3. MALDI-TOF/ICM spectra of E. coli A139 (vegetative cells, left column) and B. subtilis DSM 347 (spores, right column). (A and B) No prior inactivation. Sample preparation was carried out by the direct analysis method in which a bottom layer of a microbial sample is covered by the matrix (saturated HCCA dissolved in 2.5% trifluoroacetic acid (TFA) and 50% acetonitrile. (C and D) Inactivation by TFA 80% (30 min).
Analytical Chemistry, Vol. 80, No. 6, March 15, 2008
2031
Figure 4. Electron micrographs of sectioned spores of B. atrophaeus DSM 2277. Right: spore morphology after control treatment by a formaldehyde/glutaraldehyde mixture (FA/GA). Left: spores after 30 min of trifluoroacetic acid (TFA) treatment. Note the extraction of the spore’s core by TFA. Labels: ic, inner coat; oc, outer coat; cx, cortex; co, core; m, core membrane.
Repeatability of ICM Spectra Obtained by the TFA Method. A high degree of spectral repeatability is essential for a microbiological typing technique. In the following we present the results of our investigations which aimed at probing the repeatability of mass spectra produced on the basis of the TFA inactivation protocol. ICMS measurements were carried out on bacterial samples of two reference strains: E. coli RKI A139 and B. thuringiensis DSM 350. In these experiments cells were grown on LB agar plates at a temperature of 37 °C. After 24 h of growth, vegetative cells were harvested and inactivated according to the TFA inactivation procedure. From four independent cultivations of each of the strains, altogether 40 spectra were obtained. Representative mass spectra from E. coli and B. thuringiensis are given in panels A and C of Figure 5, respectively. Furthermore a contour plot of the gray-scaled ICMS intensities shows the abscissa values as mass/charge ratios (m/z) and spectral numbers as the ordinates (see Figure 5B). Note that in this “gel view” individual spectra appear as rows. Mass spectra from E. coli are given in the upper 20 rows of Figure 5, panel B, whereas the lower lines depicts spectra of B. thuringiensis. The illustration demonstrates a high degree of spectral repeatability and, at the same time, specific protein patterns in the mass spectra of both strains. DISCUSSION Viable microorganisms, including bacterial endospores, can be rapidly and reliably disintegrated and separated by a combination (27) Liu, H.; Du, Z.; Wang, J.; Yang, R. Appl. Environ. Microbiol. 2007, 73 (6), 1899-1907. (28) Gelderblom, H. R.; Bannert, N.; Pauli, G. J. Infect. 2007, 54 (3), 307-308. (29) Hathout, Y.; Setlow, B.; Cabrera-Martinez, R. M.; Fenselau, C.; Setlow, P. Appl. Environ. Microbiol. 2003, 69 (2), 1100-1107. (30) Robinow, C. F. J. Bacteriol. 1953, 66, 300-311.
2032 Analytical Chemistry, Vol. 80, No. 6, March 15, 2008
Figure 5. Repeatability of MALDI-TOF mass spectrometry. Data were obtained from vegetative cells after the TFA inactivation procedure. Panel A shows a representative mass spectrum of E. coli RKI A139, and panel C shows a spectrum of B. thuringiensis DSM 350. To generate the image of panel B, intensities of 20 individual mass spectra obtained from four independent cultivations of E. coli and B. thuringiensis were gray-scaled and plotted as function of the m/z values.
of three routine laboratory procedures: treatment with concentrated TFA, centrifugation, and subsequent filtration of the supernatant through spore-tight filter membranes. The present work demonstrates that a combination of these simple procedures, which we called the TFA inactivation protocol, can be employed as an effective inactivation method that is compatible with the MALDI-TOF typing technique for microorganisms. The development of the TFA inactivation protocol was initially inspired by a work of Hathout et al., which recommended 1 N hydrochloric acid for extraction of small, acid-soluble proteins (SASP) from different Bacillus spores.29 In this work, the SASPs are considered as species-specific biomarkers of Bacillus spores suitable for bacterial typing by MALDI-TOF mass spectrometry. In the context of our work it was particularly interesting that the authors reported that treatment of spores of B. subtilis by 1 N HCl for 1 h did not only extract SASPs but caused also spore killing (>99%, or RF > 2).29 Treatment of microorganisms by strong acids is known for a long time to kill spores. For example, concentrated nitric acid (1 N HNO3) can cause the rupture of the outer spore coat of B. megaterium spores30 which is accompanied by the extrusion of nuclear material. Setlow et al. confirmed these findings and reported also low survival rates (3%) for spores of a wild-type B. subtilis strain incubated with 1 N HCl for 2 h.22 Interestingly, the
resistance of growing cells to acid treatment was found to be much lower since incubation in 300 mmol/L HCl gave >99.99% killing (RF of >4).22 The experimental data of our work with TFA support these data: in comparison with spores, the resistance of vegetative cells to strong acids is significantly lower (cf., Tables 1 and 2). The high killing efficiency of TFA treatment alone (80%, 30 min) to vegetative cells suggests that the centrifugation and filtration steps could be omitted when the presence of spores can be excluded. The mechanisms of spore killing by strong acids apparently involves the breakdown of permeability barriers, presumably of the spore’s core membrane which was found to be irreversibly altered in acid killed spores.22 The coat of the spores does not substantially contribute in the protection against acids since decoating has only a slight effect on the spore’s resistance to acid treatment. This is possibly due to the small molecular size of the penetrating ions.22 The second aspect of TFA activity in microbial suspensions, which is of particular importance for ICMS, is the ability of TFA to act as an organic solvent, i.e., to effectively dissolve proteins. For example, it is well-known that anhydrous TFA readily dissolves a large variety of proteins such as various albumins, globulins, lysozyme, ribonucleases, and other.31 The mechanism of spore inactivation involves therefore not only the destruction of permeability barriers but also the effective extraction of soluble microbial proteins from the core and from other morphological structures of the spores. Obviously, the ability of TFA to effectively penetrate permeability barriers and to extract, or dissolve, significant amounts of microbial proteins are fundamental principles of the TFA inactivation technique of microorganisms. In fact, the results of EM fit nicely with this interpretation. For example, TFA-treated spores of B. atrophaeus appear in our EM micrographs as “ghosts” formed by empty hull structures of the outer coat, whereas inner structures such as the spore’s core, cortex, and inner coat are almost completely extracted (cf., Figure 4, parts A and C). In the light of these considerations it becomes clear that most of the spore’s proteins are effectively dissolved in TFA and are thus detectable by ICMS. The additional two procedures of the TFA inactivation protocol, centrifugation and filtration, are not only helpful in removing potentially surviving spores but improve also the quality of the ICM spectra. For example, centrifugation was found to be effective for separating the supernatant, which contains the major fraction of microbial biomarkers, from cell or spore debris and small insoluble contaminants such as dust particles in environmental samples. Indeed, comparative ICMS measurements of clear supernatant solutions and the pellets have proven a higher signal content in spectra of the supernatant (data not shown). Furthermore, centrifugation is considered as a precondition for filtration since it prevents clogging of the spore filters. Solubilization of microbial proteins in highly concentrated TFA is known to be accompanied by conformational changes (unfolding). Although in a 80% TFA solution the native secondary and tertiary structure of the proteins is lost, there is no experimental evidence for large-scale biochemical modifications of the primary protein structure such as side-chain modifications, cross-links, or hydrolytic cleavages. (31) Katz, J. J. Nature 1954, 174 (4428), 509.
After inactivation, the preparation of microbial extracts for MALDI-TOF/ICMS included a dilution step, in which the concentration of TFA is reduced from 80% to 8% (1:9 dilution v/v). This step turned out to be mandatory since it avoided spreading of the highly concentrated TFA extracts over the surface of MALDI steel targets. The dilution step is furthermore crucial because it causes nearly complete precipitation of the dissolved microbial proteins. The drawback of the required dilution step lies in a decrease of the analytical sensitivity. A reduced sensitivity could be avoided if the precipitate is spun down by a second short centrifugation step and redissolved in a small volume of a standard acetonitrile/TFA (0.1%) solution. This and alternative concepts of increasing the analytical sensitivity of the TFA sample preparation technique are currently investigated in our laboratory. Although we have presented here an extraordinarily efficient inactivation protocol, it must be, however, mentioned that this technique cannot be evaluated for all conceivable circumstances that may happen in real-life situations. For example, the influence of acid neutralizers which might be present as additives of spore preparations (e.g., BSA) on the killing capacity of TFA was not a subject of our investigations. However, although neutralizers of this type may be able to reduce the killing efficiency, the TFA inactivation protocol provides sufficient capacity to ensure a highly efficient inactivation. Nevertheless, routine tests which intermittently check the success of inactivation are obligatory and should be carried out on a regular basis. Reproducibility and repeatability are both considered as indispensable prerequisites for any bacterial typing technique. Whereas reproducibility refers to the ability to accurately replicate the results of an experiment (possible by someone else working independently), repeatability measures the success rate in successive experiments. Systematic investigations on the repeatability of ICMS have demonstrated that the TFA inactivation/sample preparation technique can be effectively employed to repeatedly obtain bacterial mass spectra. For example, the illustration of Figure 5 illustrates nearly identical mass spectra from independent cultures of the test microorganisms E. coli and B. thuringiensis (low intraspecies variance). Furthermore, the gel view of Figure 5 demonstrates also a high degree of distinctness, or a high interspecies variance, between ICMS patterns of the test strains. It is plausible that a low intraspecies and a high interspecies variances qualify MALDI-TOF/ICMS as an ideal technique for bacterial typing. We are currently investigating TFA extracts from the genus Yersinia. The analysis of these mass spectra demonstrated the suitability of the TFA sample preparation protocol for typing bacteria by MALDI-TOF/ICMS. CONCLUSIONS In this work we have elaborated a protocol for MALDI-TOF/ ICMS compatible inactivation of microorganisms, with particular emphasis on the inactivation of bacterial endospores. We have shown that a combination of treatments with concentrated TFA, centrifugation, and subsequent filtration of the supernatant provides sterile protein extracts of microbial cells and bacterial endospores suitable for MALDI-TOF/ICMS. The new inactivation procedure was validated for a large number of microorganisms and permits comprehensive detection of microbial protein patterns by MALDI-TOF mass spectrometry. When compared with the direct ICMS technique, the TFA inactivation protocol allows us Analytical Chemistry, Vol. 80, No. 6, March 15, 2008
2033
to produce spectra which exhibit similar numbers of mass peaks and show also a high degree of spectral reproducibility. We are confident that the TFA inactivation protocol will be applied in the near future as a simple, reliable, and rapid sample preparation technique for MALDI-TOF/ICMS of highly pathogenic (BSL-3) microorganisms. ACKNOWLEDGMENT The authors thank W. Beyer (Universita¨t Hohenheim, Germany) for providing the strains of B. anthracis, A. Brandt (RKI, Berlin) for performing the reproducibility measurements, and M. Laue, and G. Holland for performing the EM experiments. The authors are grateful to Ralf Dieckmann (AnagnosTec, Potsdam), T. Maier, M. Kostrzewa (Bruker Daltonics, Leipzig), and H. Russmann (Wehrwissenschaftliche Institut fu¨r Schutztechnologien, Munster) for fruitful discussions and support. Furthermore, the excellent technical assistance of P. Lochau, R. Heinrich, and R. Ho¨mke (RKI, Berlin) is acknowledged. This work was sup-
2034
Analytical Chemistry, Vol. 80, No. 6, March 15, 2008
ported by the “Bundesamt fu¨r Bevo¨lkerungsschutz und Katastrophenhilfe” (BBK F2-440-00-185/04) and the “Wehrwissenschaftliche Institut fu¨r Schutztechnologien” (E/E590/4Z015/N5129 and ZVZ-DRMZ-1368-652). The first two authors contributed equally to this work. SUPPORTING INFORMATION AVAILABLE Figure showing the basic scheme of the procedure used to test disinfectants (quantitative suspension assay: EN 14347) and a table giving an overview of all bacterial strains used in this study. This material is available free of charge via the Internet at http:// pubs.acs.org.
Received for review August 29, 2007. Accepted January 11, 2008. AC701822J
