Detonation Nanodiamonds for Rapid Detection of Clinical Isolates of

Aug 20, 2012 - With on diamond digestion, we confirm the DND captured antigen is cell filtrate protein 10 (CFP-10) because its Mascot analysis shows a...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/ac

Detonation Nanodiamonds for Rapid Detection of Clinical Isolates of Mycobacterium tuberculosis Complex in Broth Culture Media Po-Chi Soo,‡ Ching-Jen Kung,† Yu-Tze Horng,‡ Kai-Chih Chang,‡ Jen-Jyh Lee,§ and Wen-Ping Peng*,† †

Department of Physics, National Dong Hwa University, Shoufeng, Hualien, Taiwan 97401, R.O.C. Department of Laboratory Medicine and Biotechnology, Tzu Chi University, Hualien, Taiwan, R.O.C. § Department of Internal Medicine, Buddhist Tzu Chi General Hospital, Tzu Chi University, Hualien, Taiwan, R.O.C. ‡

S Supporting Information *

ABSTRACT: Routinely used molecular diagnostic methods for mycobacterium identification are expensive and time-consuming. To tackle this problem, we develop a method to streamline identification of Mycobacterium tuberculosis complex (MTBC) in broth culture media by using detonation nanodiamonds (DNDs) as a platform to effectively capture the antigen secreted by MTBC which is cultured in BACTEC MGIT 960, followed by the analysis of matrixassisted laser desorption/ionization mass spectrometry (MALDI-TOF MS). The 5 nm DNDs can capture the MTBC secretory antigen without albumin interference. With on diamond digestion, we confirm the DND captured antigen is cell filtrate protein 10 (CFP-10) because its Mascot analysis shows a score of 68. The dot blotting method further verifies a positive reaction with anti-CFP10, indicating that CFP-10 is secreted in the medium of mycobacterium growth indicator tube (MGIT) and captured by DNDs. The minimal CFP-10 protein detection limit was 0.09 μg/mL. Furthermore, our approach can avoid the falsepositive identification of MTBC by immunological methods due to cross-reactivity. Five hundred consecutive clinical specimens subjected to routine mycobacteria identification in hospital were used in this study, and the sensitivity of our method is 100% and the specificity is 98%. The analysis of each MTBC sample from culture solution can be finished within 1 h and thus shortens the turnaround time of MTBC identification of gold standard culture methods. In sum, DND MALDI-TOF MS for the detection of MTBC is rapid, specific, safe, reliable, and inexpensive.

T

quickly identify MTBC (e.g., AccuProbe took 3.5 h, ProbeTec 3.5−4 h, AMTD2 2.5 h, and AMPLICOR 6−7 h4,5). However, they have the following shortcomings. First, long turnaround times lead to delay in reporting results. Second, it is expensive to require reagents due to cold storage and shipping. Last but not least, it is labor-intensive and expensive to extract nucleic acid, so samples are usually processed in batches in some hospitals to lower the cost.6 The above-mentioned methods are suitable for both solid and liquid media, but the use of liquid cultures is recommended by the World Health Organization (WHO) for mycobacterial species identification and drug susceptibility tests in high TB burdened countries due to their rapid detection and increasing yield.7 Nevertheless, liquid cultures also grow nontuberculous mycobacteria (NTM), which sometime exists in the upper respiratory track and causes opportunistic infections in the immunocompromised individuals,1and therefore, rapid and reliable differentiation of MTBC and NTM is important for the treatment and effective control of TB.

he genus Mycobacterium encompasses approximately 140 heterogeneous species of rapid- and slow-growing bacilli.1 Most species of this genus live in the environment, and some of these species cause diseases in humans and animals. The most important pathogenic species are Mycobacterium tuberculosis complex (MTBC) which can cause tuberculosis (TB). TB is a major public health problem that resulted in 8.8 million incident cases and 1.1 million deaths in 2010 and is the second leading cause of death from an infectious disease worldwide.2 MTBC has become a significant cause of death in many developing countries and continues to be a public health problem globally. Therefore, the development of a rapid, reliable, and early diagnosis method for MTBC is crucial to prevent further spread. Conventional methods of identifying MTBC are colony morphology, pigmentation and biochemical tests,3 which require several weeks for adequate MTBC growth and sometimes cannot make an accurate identification of MTBC; therefore, molecular biology tools have been developed to diagnose MTBC, such as DNA hybridization (AccuProbe, GenProbe, San Diego, CA) and nucleic acid amplification (BD ProbeTec ET Direct TB system, Becton Dickinson Sparks, MD; AMTD2 Gen-Probe, Inc., San Diego, CA; and COBAS AMPLICOR MTB assay, Roche, Basel, Switzerland). These can © 2012 American Chemical Society

Received: June 25, 2012 Accepted: August 20, 2012 Published: August 20, 2012 7972

dx.doi.org/10.1021/ac301767z | Anal. Chem. 2012, 84, 7972−7978

Analytical Chemistry

Article

assays, the turnaround time of DND MALDI-TOF MS is greatly reduced from 5 to 1 h and the sensitivity increases by about a factor of 2. Compared with the MALDI-TOF MS method, the DND MALDI-TOF MS analysis shows 40-fold increase in detection sensitivity of CFP-10. In summary, our method is simple and can minimize the operator’s exposure to pathogens in the clinical laboratory. Besides, the cost of DNDs in diagnosis is very low, approximately $10−20 per gram, and we only need 50 μg for one sample analysis. DNDs can be directly used without surface modifications and operated in a wide pH range.21,26 Our results showed 100% sensitivity and 98% specificity in identifying 500 clinical MTBC isolates.

Developing immunological methods to detect TB with antibody and antigen reaction may provide an alternative for hospitals with limited resources. The proteins secreted into the extracellular environment by MTBC are known to elicit MTBC-specific immune responses and are of diagnostic value. To date, secreted proteins including MPT45, MPT63, MPB64, MPT70, antigen 85 complex, and ESAT-6/CFP-108 are the known major candidates used for MTBC identification. The MPB64 antigen has been used in the immunochromatographic assay (ICA; the Capilia TB assay and TBc ID test) for rapid differentiation of MTBC in liquid cultures.9 However, it is found that false-negative reports are observed in ICA because there are few bacterial numbers in the culture or bacteria with mpb64 mutation.10 MTBC has cross-reactivity with certain clinical isolates, such as M. chelonae and M. intracellulare by detecting Mpb64.11,12 ESAT-6/CFP-10, an alternative of Mpb64, were reported to identify MTBC in cultures by enzyme-linked immunosorbent assay (ELISA) and ICA.13 The limitation of the ESAT-6/CFP-10 ICA is its cross-reactivity with some NTM-containing cultures. Genes coding for ESAT6-like proteins and CFP-10-like proteins were demonstrated in M. kansasii, M. marinum, M. szulgai, M. flavescens, M. gastri, and M. smegmatis14,15 and found to produce the false-positive results. Thus, a desirable new method to overcome these defects has to be developed. Here we present detonation nanodiamond matrix-assisted laser desorption ionization time-of-flight mass spectrometry (DND MALDI-TOF MS) to rapidly and correctly identify MTBC. Hettick et al. and others have used MALDI-TOF MS for identification of Mycobacterium species.16−18 However, comprehensive bacterial profiling databases of specific biomarkers are still not well-developed and the global experimental data are still quite limited for a rapid and reliable identification of Mycobacterium species. Besides, the whole cell MALDI-TOF MS experiments expose operators to pathogenic MTBC, and identifications of strains obtained from mycobacterium growth indicator tube (MGIT) media become worse as compared with those of strains from solid media.19 Nanodiamonds (NDs), due to their chemical inertness, hardness, optical transparency, and biocompatibility, have been applied in analytical chemistry, biology, catalysis, spectroscopy, materials, electronic applications, and quantum computing.20 A particular type of ultrananocrystalline material called detonation nanodiamonds (DNDs) with the characteristic size of 4−5 nm were produced by detonation of carbon-containing explosives in the former USSR in the 1960s and show good extraction performance in mass analysis.19,21 The combination of highpressure high-temperature (HPHT) NDs with TOF mass spectrometry can also effectively detect proteins,22 biopolymers,19 DNA oligonucleotides,23 and multiphosphorylated peptides24 from complex biosamples feasible.25 Therefore, ND MALDI-TOF MS is a good platform to detect potential biomarkers under critical environments (e.g., salt and detergent interferences, low protein concentrations) in detecting MTBC. We first showed that DND MALDI-TOF MS can enhance the sensitivity and credibility of detecting specific biomarker, CFP10, which can identify MTBC from MGIT culture media. On diamond digestion further confirms that the antigen extracted by 5 nm DNDs is CFP-10 proteins. The dot blotting method shows a positive reaction with anti-CFP-10 of DND captured proteins, which confirms that CFP-10 is presented in the medium of MGIT with MTBC growing but not with NTM growing. Compared with the commercial molecular diagnostic

■

EXPERIMENTAL SECTION Chemical and Materials. Sinapic acid (SA), alpha-cyano-4hydroxy-cinnamic acid (α-CHCA), bovine trypsin, Tween 20, iodoacetamide, and β-mercaptoethanol were purchased from Sigma-Aldrich (St. Louis, MO, USA). HPLC grade trifluoroacetic acid (TFA) was obtained from Alfa Aesar. HPLC grade acetonitrile, methanol, and acetone were obtained from J.T. Baker. Nonfat dry milk was from Anchor. Phosphate-buffered saline (PBS) powder was obtained from BIO BASIC. Antimouse IgG was from Jackson ImmunoResearch. CFP-10 monoclonal antibody was bought from Novus Biologicals. Chemiluminescence reagent was from PerkinElmer. BBL MGIT, BACTEC MGIT Growth Supplement, and BBL MGIT PANTA were obtained from Becton Dickinson (Cockeysville, MD). Nitrocellulose membrane (NC membrane) was from Amersham Biosciences. PBST is a mixture of 1× phosphate-buffered saline and 0.1% Tween 20. Specimen Collection and Processing. The specimens were digested and decontaminated by the N-acetyl-L-cysteine− 4% sodium hydroxide method and concentrated by centrifugation.27 Cultures were performed by inoculating 0.5 mL of sediment into liquid culture BACTEC MGIT 960 system (Becton Dickinson, Cockeysville, MD). The BACTEC MGIT 960 system was automatically monitored until a positive signal presented. A 1 ml portion of the MGIT cultivated medium was filtered, followed by sample preparation for MALDI-TOF MS, DND MALDI-TOF MS, or dot blotting analysis. Dot Blotting. A 1 μL portion of MGIT medium containing native or recombinant CFP-10 was loaded on nitrocellulose membrane (NC membrane). After air drying, the NC membrane was incubated in the blocking solution (5% nonfat dry milk in PBST) for 1 h. Subsequently, the membrane was incubated with monoclonal anti-CFP10 for 1 h, followed by wash with PBST for 15 min three times. The membrane was incubated with antimouse IgG conjugated horseradish peroxidase (HRP) for 1 h. After it was washed with PBST, the enhanced chemiluminescence (ECL) reagent was put in and incubated for 2 min. The intensity of the dot was analyzed by the gel catcher 2850 chemiluminescence camera system (CLUBIO, Taipei, Taiwan). Sample Preparation for DND MALDI-TOF MS Analysis. The DNDs were suspended in DI water at a concentration of 1 mg/mL. The DND solution was sonicated for 5 min before use. A 50 μL portion of DND solution (1 mg/mL) was put to 1 ml of filtered MGIT medium. After weakly vortexing for 30 min at room temperature, the protein-loaded DNDs were centrifuged at 13 000 rpm for 3 min.28 The supernatant was removed, and the DNDs were additionally washed with DI water to remove residual contaminants. A 1.5 μL portion of saturated SA solvent was mixed with DNDs followed by depositing 0.8 μL of 7973

dx.doi.org/10.1021/ac301767z | Anal. Chem. 2012, 84, 7972−7978

Analytical Chemistry

Article

Figure 1. Mass spectra of cell-free filtrate from MTBC and NTM cultivated in the MGIT. The filtrate was analyzed by MALDI-TOF MS without DND pretreatment (A) or with either 5 (B) or 50 nm DND pretreatment (C) The filtered medium MGIT background was a control. The potential biomarker for differentiating MTBC from NTM is indicated by a bracket.

Specific Biomarker Selection for MTBC by DND MALDI-TOF MS Analysis. After the pretreatment by 5 nm DNDs and MALDI-TOF MS analysis, the characteristic mass spectrum from MTBC consisted of several peaks which were absent from NTM (Figure 1). Further examination of these peaks in the samples from 12 NTM isolates and 12 MTBC isolates were done. Table 1 showed that four peaks were absent in the samples from NTM isolates and presented in most or all of these MTBC isolates. Three peaks showed frequency up to 80%, but only the peak at 10 675 m/z showed the highest intensity and the highest frequency (100%) in the cell-free filtrate from the liquid media growing MTBC (Table 1).

mixture solution on a plate, and then, the sample was analyzed by MALDI-TOF MS.

■

RESULTS AND DISCUSSION Signal Enhancement by DND MALDI-TOF MS Analysis. In order to screen the potential biomarkers of MTBC from MGIT liquid-culture media, one MTBC and one NTM were cultivated in the MGIT separately. After the culture is positive, the filtered liquid media of MGIT were directly analyzed by MALDI-TOF MS. The results showed several lowintensity peaks in mass spectrum (Figure 1A). It was speculated that the low concentrations of these potential biomarkers in the cultured media led to the weak intensity of the peaks. Many findings report the NDs could be used to concentrate proteins from liquid media and elevate the total numbers of proteins for MALDI-TOF MS analysis.25 In this study, 5 and 50 nm of DNDs were used to concentrate the proteins from MGIT broth. The intensities of the peaks were enhanced about a factor of 50 as compared with those of the peaks without 5 nm DNDs treatment. The intensities of peaks by 5 nm DNDs pretreatment were even greater than 50 nm DNDs pretreatment (Figure 1B). These results demonstrate that 5 nm DNDs is the best platform to concentrate the biomarkers in liquid media of MGIT growing mycobacterium species followed by DND MALDI-TOF MS analysis.

Table 1. Selected Peaks from Spectra Obtained from MTBC Cultured in MGIT Broth DND MALDI-TOF analysis m/z [Da]

intensity (%)a

frequency (%)b

10114.31 10128.12 10675.79 10690.44

37.04 15.46 77.12 41.93

88 80 100 60

a

Averaged peak intensity from 12 MTBC isolates analyzed by DND MALDI-TOF MS. bThe positive rate of the peak present in mass spectra obtained from 12 MTBC isolates. 7974

dx.doi.org/10.1021/ac301767z | Anal. Chem. 2012, 84, 7972−7978

Analytical Chemistry

Article

Figure 2. Comparison of the mass spectra of cell-free filtrate from MTBC cultivated in the MGIT without (A-i) or with (A-ii) trypsin digestion on nanodiamonds. The peak of the biomarker at 10 675 m/z absent after trypsin digestion was marked with solid-line arrow (A-i). The tryptic peptides were from supernatant solution and the CFP-10 related peptides were marked with dash-line arrow (B). 7975

dx.doi.org/10.1021/ac301767z | Anal. Chem. 2012, 84, 7972−7978

Analytical Chemistry

Article

Therefore, the peak at 10 675 m/z was chosen as specific biomarker to distinguish MTBC from NTM in clinical specimens. In order to characterize this specific biomarker, the cell-free filtrate from the media growing MTBC was concentrated by 5 nm DNDs, followed by trypsin digestion and MALDI-TOF MS analysis. The peak at 10 675 m/z was absent after trypsin digestion (Figure 2A), indicating this biomarker was a protein, not a chemical compound, polysaccharide, or lipid. Because the molecular weight, Mw = 10 675, is very close to theoretic molecular weight of CFP-10 protein (Mw 10794)29 and we notice the lost of N-terminal Met is shown in the literature,29,30 the theoretic Mw is turned to 10 663. This value is very close to 10 675 in DND MALDI-TOF MS analysis. The tryptic digest peptide masses (m/z) were marked as 477.02, 906.39, 1138.34, 2004.9, and 2613.48 (Figure 2B). Compared with the peaks of theoretical trypsin digestion of every peptide in the SwissProt database through the Mascot software, those marked five peaks (shown in Figure 2B) result in a significant probability-based MOWSE score of 68 which indicates the peak at m/z 10 675 is a CFP-10 antigen (Figure 3A). This is consistent with the literature reports.29,30 The CFP-10 protein is secreted from MTBC and involved in the pathogenicity of MTBC by

inhibiting lipopolysaccharide (LPS)-induced production of reactive oxidative species (ROS).31 In order to confirm the CFP-10 were present in the samples containing MTBC not NTM, the cell-free filtrates with 5 nm DNDs processing were analyzed by dot blotting using anti-CFP-10 monoclonal antibody. The dot blotting result showed a positive reaction with anti-CFP-10 antibody, indicating that CFP-10 was secreted in the medium of MGIT growing MTBC not NTM (Figure 3B). Together, the specific biomarker CFP-10 could be used to identify MTBC from clinical specimens by DND pretreatment followed by MALDI-TOF MS analysis. Detection Limit of Specific Biomarker CFP-10 from MGIT Broth. The detection limit of CFP-10 in MGIT medium by DND MALDI-TOF MS was measured using serial 2-fold dilution of recombinant CFP-10 (rCFP-10) proteins in MGIT medium. The rCFP10 proteins were diluted from 14.4 to 0.04 μg/mL. The 100 μL of MGIT medium and rCFP10 proteins were mixed with 50 μL of 5 nm DNDs and incubated for 30 min. Subsequently, the DNDs absorbing proteins were analyzed by MALDI-TOF MS and confirmed by dot blotting using anti-CFP-10 monoclonal antibody. We found the optimal detection limit by dot blotting method is 0.18 μg/mL, whereas the detection limit of MALDI-TOF MS (without DND pretreatment) is 3.6 μg/mL. The minimal concentration of rCFP-10 protein in MGIT medium for identification by DND MALDI-TOF MS is 0.09 μg/mL (Figure 4). The above results indicate that the DND MALDI-TOF MS analysis shows the best detection sensitivity as compared to that of dot blotting method and conventional MALDI-TOF MS analysis. Feng et al. showed the detection limit for ESAT-6/CFP-10 by ELISA was approximately 60 pg/mL,13 but their method employed

Figure 3. Biomarker in the cell-free filtrate from MTBC treated with trypsin identified using Mascot search algorithms with peptide fragment mass spectra generated by MALDI-TOF mass analysis (A) and confirmed by dot blotting using anti-CFP-10 antibody (B). (A-i) Identified peptide sequences of CFP-10 by MALDI-TOF MS shown as bold and underlined. (A-ii) Mascot analysis result showing positions of start and end amino acid residue of tryptic peptides sequences (Start-End), the observed peak of tryptic peptides identified by MALDI-TOF MS analysis (Observed), the expected molecular weight of predicted tryptic peptides (Mr.), and the site cleaved by trypsin marked with dot. NC and PC are negative control and positive control, respectively.

Figure 4. Detection limit of rCFP-10 by DND MALDI-TOF MS analysis. (A) The serial dilution of rCFP-10 was analyzed by MALDITOF MS (ii) or DND MALDI-TOF MS (iii) and confirmed by dot blotting using anti-CFP-10 antibody (i). The peak of rCFP-10 at 14 410 m/z shown in mass spectra was marked with +; − indicated no peak. (B) Mass spectra of various concentrations of rCFP-10. 7976

dx.doi.org/10.1021/ac301767z | Anal. Chem. 2012, 84, 7972−7978

Analytical Chemistry

Article

long incubation time of antigen−antibody reaction; the turnaround time was estimated to be 3 h. Besides, the preparation of the ELISA plate is also time-consuming, including coating for a night and blocking for 2 h. Furthermore, the cost for cold storage of antibody and peroxidase-conjugated streptavidine is high. Here, we show that DND MALDI-TOF MS is a fast and easy method by greatly shortening the turnaround time to 1 h. Protein Capture Ability with Different Sized NDs and Surface Morphology of Diamond Nanocrystalline. The capture ability of 5 nm DNDs is better than 50 nm DNDs in detecting CFP-10 proteins in MGIT media as shown in Figure 5. The 50 nm detonation NDs show singly and doubly charged

Moreover, the SEM image of 100 nm HPHT NDs shows a disperse structure which is very different from that of 5 and 50 nm DNDs, because 5 and 50 nm DNDs have the oxygen containing functional groups and graphitic materials. Therefore, the 5 nm detonation NDs was chosen as a nice platform to concentrate the CFP-10 biomarker without albumin interference. Besides, conventional treatment of DNDs including carboxylation and oxidization in strong acids also showed no obvious difference in mass spectra analysis. The 5 nm DNDs can be directly used without any treatments, and proteins can be absorbed by 5 nm DNDs in a dilute solution within a wide pH range.21 Screen of Clinical Specimens by DND MALDI-TOF MS. A total of 500 consecutive clinical specimens subjected to routine mycobacteria identification in Buddhist Tzu Chi General Hospital (TCH) during the period from February 2009 to February 2011 were analyzed. A total of 42 specimens showed positive signals reported by the BACTEC MGIT 960 system, followed by DND MALDI-TOF MS analysis. Among the specimens identified by culture and biochemical methods, a total of 13 specimens were reported to contain MTBC strains and 29 specimens were reported to contain NTM strains. The results were summarized in Table 2. Compared with the culture Table 2. Comparison of Clinical Diagnosis of MTBC by DND MALDI-TOF MS and Conventional Culture with Biochemical Methods culture (n = 42)a DND MALDI-TOF (no. of specimens)

positive

negative

positive (14) negative (28) overall (42)

13 0 13

1 28 29

a

Culture and biochemical diagnosis results were from mycobacteriology laboratory in TCH. Sensitivity 100%; specificity 98%; positive predictive value 93%; negative predictive value 100%.

and biochemical methods, 13 MTBC-containing MGIT medium and 1 NTM-containing MGIT medium showed the peak at 10 675 m/z by DND MALDI-TOF MS. However, 28 NTM-containing MGIT medium did not show the CFP-10 protein peak. Results from DND MALDI-TOF MS showed that the sensitivity is 100%, the negative predictive value is 100%, the specificity is 98%, and the positive predictive value is 93%. Only one specimen was reported as a false-positive by DND MALDI-TOF MS. It was speculated that there was mixed culture of fast growing NTM with MTBC which was a slow growing strain. The conventional biochemical diagnosis could not distinguish the mixed mycobacterial culture in broth which needed to be confirmed by culture on solid agar media or molecular diagnosis.33 The cross-reactivity has been reported by immunological method to detect ESAT-6/CFP-1014,15 because some NTM clinical isolates produced CFP-10-like proteins which had different molecular weights from CFP-10. It could be differentiated by DND pretreatment coupled with high resolution mass spectrometry, e.g. FTICR34 and Oribtrap35 mass spectrometers.

Figure 5. Mass spectra of cell-free filtrated from MTBC with (A) 5 nm DND, (B) 50 nm DND, and (C) 100 nm HPHT ND pretreatment. The arrow indicates the CFP-10 peak, and the inset photos are the SEM images of NDs.

albumin proteins peaks in Figure 5B. According to the scanning electron microscopy (SEM) images shown in Figure 5, we infer that the gaps between porous structures of aggregated 5 nm DNDs are very small as compared to the size of albumin proteins. The 50 nm DNDs have similar surface structure and morphology of 5 nm DNDs, but the gaps of the aggregated structure of 50 DNDs are bigger to allow the capture of albumin proteins easier. Meanwhile, we notice the CFP-10 peak intensities of mass spectra by 5 nm DNDs and by 50 nm DNDs are almost the same, but more peaks are shown in the range from m/z 5000−15 000 by 5 nm DNDs, which indicates that the protein capture ability is better than 50 DNDs due to less albumin proteins interference. Figure 5C exhibits the use of high-pressure high-temperature (HPHT) 100 nm NDs for mass analysis, and we observed that 100 nm HPHT NDs could not capture the CFP-10 proteins, which is because the functional groups (CH−, OH−, NH−, CO−, COOH−) on 5 nm DND surfaces show high abundance in measured absorption spectra as compared to 100 nm HPHT NDs.32

■

CONCLUSIONS We have demonstrated that a combination of 5 nm detonation NDs and MALDI-TOF MS analysis can easily detect secretory CFP-10 protein of MTBC from MGIT liquid culture medium. This early secreted CFP-10 protein can be used as a biomarker 7977

dx.doi.org/10.1021/ac301767z | Anal. Chem. 2012, 84, 7972−7978

Analytical Chemistry

Article

(15) van Ingen, J.; de Zwaan, R.; Dekhuijzen, R.; Boeree, M.; van Soolingen, D. J. Bacteriol. 2009, 191, 5865−5867. (16) Hettick, J. M.; Kashon, M. L.; Slaven, J. E.; Ma, Y.; Simpson, J. P.; Siegel, P. D.; Mazurek, G. N.; Weissman, D. N. Proteomics 2006, 6, 6416−6425. (17) Lefmann, M.; Honisch, C.; Bocker, S.; Storm, N.; von Wintzingerode, F.; Schlotelburg, C.; Moter, A.; van den Boom, D.; Gobel, U. B. J. Clin. Microbiol. 2004, 42, 339−346. (18) Pignone, M.; Greth, K. M.; Cooper, J.; Emerson, D.; Tang, J. J. Clin. Microbiol. 2006, 44, 1963−1970. (19) Lotz, A.; Ferroni, A.; Beretti, J. L.; Dauphin, B.; Carbonnelle, E.; Guet-Revillet, H.; Veziris, N.; Heym, B.; Jarlier, V.; Gaillard, J. L.; Pierre-Audigier, C.; Frapy, E.; Berche, P.; Nassif, X.; Bille, E. J. Clin. Microbiol. 2010, 48, 4481−4486. (20) Holt, K. B. Philos. Trans. R. Soc. A-Math. Phys. Eng. Sci. 2007, 365, 2845−2861. (21) Wei, L. M.; Shen, Q.; Lu, H. J.; Yang, P. Y. J. Chromatogr. B 2009, 877, 3631−3637. (22) Kong, X. L.; Huang, L. C. L.; Hsu, C. M.; Chen, W. H.; Han, C. C.; Chang, H. C. Anal. Chem. 2005, 77, 259−265. (23) Kong, X. L.; Huang, L. C. L.; Liau, S. C. V.; Han, C. C.; Chang, H. C. Anal. Chem. 2005, 77, 4273−4277. (24) Chang, C. K.; Wu, C. C.; Wang, Y. S.; Chang, H. C. Anal. Chem. 2008, 80, 3791−3797. (25) Kong, X. L.; Cheng, P. Materials 2010, 3, 1845−1862. (26) Schrand, A. M.; Hens, S. A. C.; Shenderova, O. A. Crit. Rev. Solid State Mater. Sci. 2009, 34, 18−74. (27) Chou, C. H.; Huang, Y. T.; Hsu, H. L.; Lai, C. C.; Liao, C. H.; Hsueh, P. R. Int. J. Tuberc. Lung Dis. 2009, 13, 996−1001. (28) Chen, W. H.; Lee, S. C.; Sabu, S.; Fang, H. C.; Chung, S. C.; Han, C. C.; Chang, H. C. Anal. Chem. 2006, 78, 4228−4234. (29) Ge, Y.; ElNaggar, M.; Sze, S. K.; Bin Oh, H.; Begley, T. P.; McLafferty, F. W.; Boshoff, H.; Barry, C. E. J. Am. Soc. Mass Spectrom. 2003, 14, 253−261. (30) Rosenkrands, I.; Weldingh, K.; Jacobsen, S.; Hansen, C. V.; Florio, W.; Gianetri, I.; Andersen, P. Electrophoresis 2000, 21, 935− 948. (31) Ganguly, N.; Giang, P. H.; Gupta, C.; Basu, S. K.; Siddiqui, I.; Salunke, D. M.; Sharma, P. Immunol. Cell Biol. 2008, 86, 98−106. (32) Chung, P. H.; Perevedentseva, E.; Tu, J. S.; Chang, C. C.; Cheng, C. L. Diamond Relat. Mater. 2006, 15, 622−625. (33) Middleton, A. M.; Chadwick, M. V.; Gaya, H. Clin. Microbiol Infect. 1997, 3, 668−671. (34) Kelleher, N. L.; Senko, M. W.; Siegel, M. M.; McLafferty, F. W. J. Am. Soc. Mass Spectrom. 1997, 8, 380−383. (35) Makarov, A.; Denisov, E.; Kholomeev, A.; Baischun, W.; Lange, O.; Strupat, K.; Horning, S. Anal. Chem. 2006, 78, 2113−2120. (36) Mahmoudi, M.; Lynch, I.; Ejtehadi, M. R.; Monopoli, M. P.; Bombelli, F. B.; Laurent, S. Chem. Rev. 2011, 111, 5610−5637.

to differentiate MTBC from NTM. The detection limit of DND MALDI-TOF MS analysis is approximately 90 ng/mL without any amplification procedure. Our approach is more sensitive than the dot blotting method. The 5 nm detonation NDs can be directly used without any preparation procedures. The filtered MGIT liquid media minimizes the operator’s exposure to pathogenic bacteria. This method is safe, reliable, and inexpensive and greatly reduces the turnaround time in detecting TB. The study demonstrates a good application of nanoparticles in the sample enrichment and purification for MS study.36

■

ASSOCIATED CONTENT

S Supporting Information *

Details on sample preparation for MALDI-TOF analysis and on-diamond digestion. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The present study was supported partly by grants (contract number NSC99-2112-M-259-002-MY3 and NSC100-2320-B320-006) from National Science Council and partly by grants TCIRP99002-04 from Tzu Chi University. The authors also thank Dr. Huan-Cheng Chang and Prof. Chia-Liang Cheng of the institute of Atomic and Molecular Sciences of Academia Sinica and Department of Physics of National Dong Hwa University for offering the nanodiamond materials.

■

REFERENCES

(1) Glassroth, J. Chest 2008, 133, 243−251. (2) WHO Report 2011/Global Tuberculosis Control. http://www. who.int. Accessed May 2012. (3) Eing, B. R.; Becker, A.; Sohns, A.; Ringelmann, R. J. Clin. Microbiol. 1998, 36, 2023−2029. (4) GEN-PROBE. http://www.gen-probe.com. Accessed May 2012. (5) Piersimoni, C.; Scarparo, C. J. Clin. Microbiol. 2003, 41, 5355− 5365. (6) Muchwa, C.; Akol, J.; Etwom, A.; Morgan, K.; Orikiriza, P.; Mumbowa, F.; Odong, P. R.; Kateete, D. P.; Eisenach, K. D.; Joloba, M. L. BMC Res. Notes 2012, 5, 44. (7) Nyendak, M. R.; Lewinsohn, D. A.; Lewinsohn, D. M. Curr. Opin. Infect. Dis. 2009, 22, 174−182. (8) Nagai, S.; Wiker, H. G.; Harboe, M.; Kinomoto, M. Infect. Immun. 1991, 59, 372−382. (9) Yu, M. C.; Chen, H. Y.; Wu, M. H.; Huang, W. L.; Kuo, Y. M.; Yu, F. L.; Jou, R. J. Clin. Microbiol. 2011, 49, 802−807. (10) Hirano, K.; Aono, A.; Takahashi, M.; Abe, C. J. Clin. Microbiol. 2004, 42, 390−392. (11) Shen, G. H.; Chen, C. H.; Hung, C. H.; Wu, K. M.; Lin, C. F.; Sun, Y. W.; Chen, J. H. Int. J. Tuberc. Lung Dis. 2009, 13, 371−376. (12) Wang, J. Y.; Lee, L. N.; Lai, H. C.; Hsu, H. L.; Jan, I. S.; Yu, C. J.; Hsueh, P. R.; Yang, P. C. Diagn. Microbiol. Infect. Dis. 2007, 59, 395− 399. (13) Feng, T. T.; Shou, C. M.; Shen, L.; Qian, Y.; Wu, Z. G.; Fan, J.; Zhang, Y. Z.; Tang, Y. W.; Wu, N. P.; Lu, H. Z.; Yao, H. P. Int. J. Tuberc. Lung Dis. 2011, 15, 804−810. (14) Shen, G. H.; Chiou, C. S.; Hu, S. T.; Wu, K. M.; Chen, J. H. J. of clin. Microbiol. 2011, 49, 902−907. 7978

dx.doi.org/10.1021/ac301767z | Anal. Chem. 2012, 84, 7972−7978