ABSTRACT
The World Health Organization recently highlighted the serious worldwide problem of the emergence of antibiotic-resistant or antibiotic multidrug-resistant bacteria. Carbapenem-resistant Enterobacterales, including carbapenemase-producing Enterobacterales (CPE), are major antibiotic-resistant bacteria that can be identified by various methods, including antibiotic susceptibility testing, PCR, and immunologic assays. However, there is a need for a faster, more accurate, low-cost, and easy method to detect CPE strains. We previously developed an osmotic shock matrix-assisted laser desorption/ionization mass spectrometry (OS-MALDI MS) method for directly detecting intact Klebsiella pneumoniae carbapenemase (KPC) using osmotic shock cell lysis. In this study, we evaluated the OS-MALDI MS method and compared it with two other methods (octyl-glucoside-aided direct KPC detection method [OG-MALDI MS] and Bruker's MBT subtyping module indirect method [MBT-SM MALDI MS]). We first completed an analytical performance evaluation of the OS-MALDI MS method according to Clinical and Laboratory Standards Institute guidelines. Clinical testing was performed with 437 clinical isolates, including 292 KPC-producing bacteria and 145 non-KPC-producing bacteria. The OS-MALDI MS method exhibited 95.9% sensitivity, 100.0% specificity, and 100.0% precision for detecting KPC. Accuracy of the OS-MALDI MS, OG-MALDI MS, and MBT-SM MALDI MS methods was 97.3%, 55.9%, and 50.2%, respectively. In conclusion, the OS-MALDI MS method clearly outperformed the other methods, exhibiting the highest accuracy and sensitivity of the three methods. We propose the OS-MALDI MS method as a practical, useful method for clinic environments, which may help guide appropriate antibiotic treatment and contribute to the prevention of the spread of CPE.
KEYWORDS: osmotic lysis, MALDI MS, carbapenemase, KPC, Klebsiella pneumoniae
INTRODUCTION
As the carbapenem class of antibiotics has the widest activity spectrum among β-lactam antibiotics and the greatest efficacy against bacteria, they are considered a last resort for bacterial infections (1, 2). However, because of the misuse of antibiotics, carbapenemase-producing organisms (CPOs) have been increasing, posing a serious threat to public health according to various organizations, including the World Health Organization, European Centers for Disease Control and Prevention, and U.S. Centers for Disease Control and Prevention (3–5). As CPOs are resistant to carbapenem antibiotics, they lead to various infections (e.g., gastroenteritis, pneumoniae, sepsis) that are difficult to treat with common antibiotics (6). With the very high mortality rate of CPO infections, worldwide efforts to prevent the emergence and spread of antibiotic-resistant bacteria are crucial, along with comprehensive management of antibiotic supervision systems, such as antimicrobial stewardship programs (6, 7).
Accurate and rapid identification of CPOs is essential for preventing and treating infections, and various diagnostic methods for identifying carbapenemase are constantly being developed (8–10). Currently, phenotypic assays and molecular-based techniques are the two main methods for identifying carbapenemase in CPOs. The phenotypic assays include four types of assays: modified Hodge test (MHT) (11), colorimetric assay (12), modified carbapenem inactivation method (mCIM) (13), and spectrophotometric method (14). The MHT method is simple and inexpensive, but it is time-consuming and has high false-positive and false-negative rates because of poor metallo-β-lactamase (MBL) detection. Colorimetric assay is also a simple, inexpensive method that can detect KPC and most MBLs, but it is inadequate for OXA-48 and has limitations for various infecting factors. mCIM can detect all types of carbapenemases in a simple, cost-effective manner, but it is time-consuming. Furthermore, MHT, colorimetric assay, and mCIM methods are all indirect assays and are unable to detect various carbapenemases. However, molecular-based techniques, such as PCR and real-time PCR, are direct assays that can accurately detect the genes encoding all carbapenemases. However, they have high technical requirements and measurement costs (8).
Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS)-based bacterial identification systems are emerging worldwide. They have several advantages, including high-throughput, ease of use, low-cost per sample, and the possibility of multiplexing (15–17). A MALDI-TOF MS-based ertapenem hydrolysis assay has been developed for CPO detection, which detects the carbapenemase activity of Enterobacteriaceae strains in positive blood cultures in several hours (18). Recently, a direct and rapid method of detecting Klebsiella pneumoniae carbapenemase (KPC) protein has been reported using MS-compatible nonionic detergent and MALDI-TOF MS (19). More recently, we reported a rapid, more sensitive, cost-effective, and detergent-free method of detecting KPC from Gram-negative bacteria using osmotic shock (OS) lysis and MALDI-TOF MS (the OS-MALDI MS method) (20).
In this study, we compared the accuracy, precision, specificity, and sensitivity of the OS-MALDI MS method to that of other types of KPC detection methods using MALDI-TOF MS in clinical isolates. The OS-MALDI MS method is expected to be clinically applicable as an easy, rapid, inexpensive, and reliable assay for detecting KPC. It is also anticipated that this method can be further expanded to detect other types of β-lactamases conferring antibiotic resistance, such as CTX, VIM, and OXA.
MATERIALS AND METHODS
Materials.
Sinapic acid (SA; Sigma-Aldrich Chemical Co., St. Louis, MO, USA), α-cyano-4-hydroxycinnamic acid (HCCA; Bruker Daltonik GmbH, Bremen, Germany), trifluoroacetic acid (TFA; Sigma-Aldrich Chemical Co.), and acetonitrile (ACN; TEDIA Company, Inc., Fairfield, OH, USA) were used for MALDI MS analysis. N-octyl-β-d-glucoside (OG; Avanti Polar Lipids, Inc., USA) and formic acid (FA; FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) were used for cell lysis. Blood agar plates (BAPs) for cell culture were obtained from BANDIO Bioscience (Gyeonggi-do, South Korea).
Expression and purification of KPC-2 standard protein.
The KPC-2 standard protein was prepared as previously reported (20). Briefly, a recombinant KPC gene was transformed into Escherichia coli TOP10 cells and cultured in Luria Bertani liquid media (Becton, Dickinson and Company, Sparks, MD, USA) with 50 mg/L ampicillin at 37°C for 16 h. The cells were harvested by centrifugation (4,000 rpm, 15 min). Then, the harvested cells were lysed and partially purified using osmotic lysis method. The purified KPC-2 protein fraction was confirmed by SDS-PAGE gel and MALDI-TOF MS analysis, prior to use.
Collection of clinical isolates.
Clinical isolates were collected by the microbiology team at Seegene Medical Foundation in South Korea. Carbapenem-resistant Enterobacteriaceae (CRE)-positive isolates or CRE-negative isolates were primarily selected by the selective agar containing antibiotics. All clinical isolates were subcultured twice on BAPs, and single isolates were stored using Microbank (Pro Lab Diagnostics Inc., USA) at −80°C in a deep freezer. The protocol for this study was approved by the Institutional Review Board of the Seegene Medical Foundation (IRB No. SMF-IRB-2021-012).
Identification of clinical isolates.
Clinical isolates from the Microbank stock were subcultured on BAPs at 37°C for 20 h to 30 h. The direct colony on plate extraction method was used as previously described for bacteria identification (21). Briefly, single colonies were randomly selected from the BAPs and then transferred to an MSP 96 target polished steel BC plate (Bruker Daltonics Inc., Bremen, Germany) at the inoculated spot. The sample was then overlaid with 1 μL 70% FA and air-dried. Finally, 1 μL matrix solution (10 mg/mL HCCA in 2.5% TFA/50% ACN) was dropped onto the inoculated spot and air-dried. MALDI-TOF MS analysis was performed using the Microflex LT/SH smart MALDI-TOF MS system (Bruker Daltonics Inc., Bremen, Germany). In all analyses, automatic calibration and quality check were confirmed through bacterial test standard analysis using MALDI BioTyper Compass Explorer version 4.1. A reference library including 3,893 species from 664 genera was used to identify each microorganism by comparing peak profiles of the clinical isolates. For accurate identification, the cutoff score of the program was set at 2.0 or higher, and clinical isolates with a score of less than 2.0 were reconfirmed.
The target gene (blaKPC gene) was amplified by PCR using these universal primers: blaKPC forward, 5′-AACTGCAGGATGTCACTGTATCGCCGTCTA-3′; and blaKPC reverse, 5′-GGAATTCTTACTGCCCGTTGACGCC-3′. The PCR products were confirmed by their appropriate size on 0.8% agarose gel electrophoresis. The KPC gene subtype was confirmed by DNA sequencing analysis of the PCR products (Cosmo Genetech, Seoul, South Korea).
Lysis methods for MALDI-TOF MS analysis.
The optimized OS cell lysis method was used as previously described to prepare samples for MS identification of carbapenemase in clinical isolates of CRE (20). Briefly, cultured cells (colonies collected from BAPs) were resuspended in 100 μL hypertonic solution containing 20 mM Tris-HCl (pH 8.0) and 500 mM NaCl in distilled water. After incubation for 10 min at room temperature, suspended cells were centrifuged at 14,000 × g at 4°C for 10 min. The supernatant was then discarded, and the harvested cells were resuspended in 100 μL hypotonic solution (distilled water). The suspended cells were incubated for 10 min at room temperature and centrifuged at 14,000 × g for 10 min at 4°C, the supernatant was recovered. For sample preparation using the OG lysis method, colonies of clinical isolates cultured on BAPs were lysed in 100 μL 1.5% OG solution (250 μM Tris-HCl, 1.5% OG) at room temperature for 10 min (19). The lysate was then centrifuged at 14,000 × g for 10 min, the supernatant was recovered. For MBT-SM MALDI MS method, single colonies of cultured cells were applied to a 96-well steel target plate and overlaid with 1 μL 70% FA. After drying the FA, an equal volume of HCCA matrix solution was added onto the 96-well steel target plate and dried prior to analysis (22).
MALDI-TOF MS analysis for KPC detection.
One microliter of supernatant obtained by OS lysis was loaded on the inoculated spot on an MSP 96 target polished steel BC plate (Bruker Daltonics, Inc.) and air-dried. Next, 1 μL of matrix solution (20 mg/mL SA in 0.1% TFA/50% ACN) was dropped onto the spot and air-dried. MALDI-TOF MS analysis was performed using the Microflex LT/SH smart MALDI-TOF MS system (Bruker Daltonics, Inc.). In all analyses, calibration and quality checks were confirmed with the KPC-2 protein purified from OS lysis of KPC-producing E. coli standard cells (20). MALDI MS spectrometry was performed in linear and positive mode (ion source voltage 1, 20 kv; ion source voltage 2, 18 kv; detector gain voltage, 2.5 kv; pulsed ion extraction, 450 ns; and detection quality range, 12,000 to 32,000 Da). The parameters for data collection were absolute laser power, 82.5%; detector gain, 7.9×; accumulate shots, 2,000; and frequency, 100 kHz. MALDI-TOF MS spectra for the sample were acquired as previously described (23).
Analytical performance evaluation.
Analyte (i.e., purified KPC-2 protein) concentrations near the cutoff were determined in the detectable range of purified KPC-2 standard protein (0 to 12 ng/μL). The concentration of purified KPC-2 standard protein was previously determined by bicinchoninic acid assay. The experiments were repeated 25 times at each concentration (1, 2, 3, 4, 5, 6, 8, 9, 11, and 12 ng/μL) for bias and imprecision studies. To estimate the accuracy of the method, analytical performance evaluation was conducted with 24 clinical strains, including 12 KPC-positive and 12 KPC-negative bacteria. The number of samples required for evaluation was calculated using the G-Power program. Bias and imprecision studies and comparisons of methods were evaluated according to the Clinical and Laboratory Standards Institute (CLSI) guideline EP12-A2 (24). To determine the intensity threshold for KPC identification in spectra, probability density function was calculated with signal or noise intensity distribution (see Fig. S1 for details).
Data analysis for KPC detection.
Spectral data from OS-MALDI MS and OG-MALDI MS were analyzed using flexAnalysis software version 4.1 (Bruker Daltonik GmbH, Bremen, Germany). The spectra were normalized and smoothed using these settings: (i) for the peak detection algorithm: S/N threshold of 3, peak width of 10 m/z, height of 90%, maximum number of peaks of 300, and minimum intensity threshold of 10; and (ii) for the centroid and smoothing select algorithm: Savizky-Golay method, with a width of 30 m/z and cycle number of 1. MBT-SM MALDI MS samples were analyzed with Bruker’s MBT subtyping module (MBT-SM) software. Pairwise compassion studies between two methods (OS versus OG, OS versus MBT-SM, and OG versus MBT-SM) were performed by Mann-Whitney test (independent samples) using the MedCalc program (version 20.106).
RESULTS
Collection and distribution of clinical isolates.
In total, 437 clinical isolates were used to evaluate the three MALDI-TOF MS-based KPC detection methods (Table 1). The isolates were identified using the Microflex LT/SH smart MALDI-TOF MS system and classified into eight species: K. pneumoniae, 287 isolates; Klebsiella aerogenes, four isolates; Klebsiella variicola, seven isolates; Klebsiella oxytoca, one isolate; E. coli, 111 isolates; Citrobacter koseri, 24 isolates; Enterobacter cloacae, two isolates; and Acinetobacter baumannii, one isolate. Of the 437 isolates, 292 (66.74%) were KPC-positive strains and 145 (33.26%) were KPC-negative strains. The 292 blaKPC genes were identified by PCR, with each subtype confirmed by DNA sequencing: KPC-2, 288; KPC-4, two; KPC-19, one; and KPC-70, one. The blaKPC-2 gene was identified in all seven species, whereas the blaKPC-4 (n = 2) and blaKPC-70 (n = 1) genes were identified in only K. pneumoniae, and the blaKPC-19 gene (n = 1) was identified in C. koseri.
TABLE 1.
Clinical test results for the three MALDI-TOF MS methods in 437 clinical isolates
| ID (MALDI BioTyper) |
No. of KPC-pos/no. of KPC-neg | KPC gene IDc |
OS | OG | MBT-SM | ||
|---|---|---|---|---|---|---|---|
| Genus | Species | 292/145 | Subtype | n | 280 | 99 | 74 |
| Klebsiella | K. pneumoniae a | 235/52 | bla KPC-2 | 232 | 222 | 72 | 69 |
| bla KPC-4 b | 2 | 2 | 0 | 0 | |||
| bla KPC-70 | 1 | 1 | 1 | 1 | |||
| K. aerogenes | 4/0 | bla KPC-2 | 4 | 4 | 2 | 0 | |
| K. variicola a | 5/2 | bla KPC-2 b | 5 | 4 | 0 | 0 | |
| K. oxytoca | 1/0 | bla KPC-2 | 1 | 1 | 1 | 0 | |
| Escherichia | E. coli a | 21/90 | bla KPC-2 | 21 | 20 | 6 | 4 |
| Citrobacter | C. koseri | 24/0 | bla KPC-2 | 23 | 23 | 15 | 0 |
| bla KPC-19 | 1 | 1 | 1 | 0 | |||
| Enterobacter | E. cloacae a | 1/1 | bla KPC-2 | 1 | 1 | 1 | 0 |
| Acinetobacter | A. baumannii | 1/0 | bla KPC-2 b | 1 | 1 | 0 | 0 |
Detected in both KPC-positive and KPC-negative isolates.
Detected by only OS-MALDI MS.
ID, identification; KPC, Klebsiella pneumoniae carbapenemase; MALDI-TOF MS, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry; MBT-SM, Bruker’s MBT subtyping module; neg, negative; OG, n-octyl-β-d-glucoside; OS, osmotic shock; pos, positive.
Analytical performance testing of OS-MALDI MS method.
We previously showed the superiority of the OS-MALDI MS method in terms of sensitivity (20), but the method was not sufficiently validated in clinical isolates. Before initiating large-scale clinical testing, we selected 12 KPC-positive and 12 KPC-negative clinical isolates and performed qualitative performance testing according to CLSI guideline (EP12-A2), as shown in Fig. 1 and Fig. S1. KPC peak intensities were extracted from the MALDI-MS spectra of the 12 KPC-positive isolates (n = 360: 12 isolates × 3 persons × 10 days), and noises were recalculated from peak signal-to-noise ratio values, obtained using the flexAnalysis program. The intensity cutoff value of the KPC peak in the MALDI-TOF MS spectra was determined using the probability density function for the signal and noise frequency distributions (Fig. S1). The intensity threshold was chosen to be 1.67 logarithmic scale (arbitrary unit, 46.84) based on the three-sigma value (Fig. S1), the value of maximal differentiation between signal and noise. When we obtained a normal distribution for the KPC mass (bin size, 10), most KPC peaks were observed in the range of ± 100 m/z, with a 99.7% positive rate. Thus, the above criteria were applied for KPC identification. The 12 KPC-producing and 12 non-KPC-producing bacteria were confirmed as positive and negative, respectively, with 100% sensitivity and 100% specificity. The median ± standard deviation (SD) m/z of the 12 KPC peaks repeatedly confirmed 360 times was 28,726.20 ± 39 (95% confidence interval [CI] of the median = 28,722.77 to 28,729.12 m/z).
FIG 1.
Summary of three methods used for KPC detection by MALDI-TOF MS. The entire process of analytical performance test for OS-MALDI MS is described in the left panel. Schematic diagram of clinical tests for three methods (OS, OG, MBT-SM) is shown in the right panel. An MALDI MS method for analytical performance test and clinical tests (OS and OG) was performed as marked by asterisks. MBT-SM MALDI MS method was followed as manufacturer’s protocol.
Clinical tests for KPC-producing bacteria using various MALDI-TOF MS methods.
The performance results of the two direct KPC detection methods (OS-MALDI MS and OG-MALDI MS) and one indirect KPC detection method (MBT-SM MALDI MS) using clinical isolates are summarized in Table 1. Representative spectra for the three methods are shown in Fig. S2. Overall, the OS-MALDI MS correctly identified 95.9% (280 of 292) KPC-positive isolates, while OG and MBT-SM MALDI MS correctly identified 33.9% (99 of 292) and 25.3% (74 of 292) of KPC-positive isolates, respectively. MBT-SM MALDI MS detected KPC-related protein (11,109 Da) in only K. pneumoniae and E. coli, whereas OS-MALDI MS directly detected KPC in all species and OG-MALDI MS directly detected KPC in all species except K. variicola. KPC protein was not detected by any method in any of the KPC-negative isolates.
All clinical isolates were identified by applying the criteria determined in the analytical performance evaluation. The diagnostic values of all three methods are presented in Table 2. The sensitivity and specificity of OS-MALDI MS method for detecting KPC protein were 95.9% and 100%, respectively. The accuracy and precision of the OS-MALDI MS method were 97.3% and 100%, respectively. Based on a disease prevalence of 66.74% (292/437) in the evaluated samples, the OS-MALDI MS method had a positive predictive value of 99.29% (95% score confidence limit = 97.24% to 99.82%) and a negative predictive value of 91.82% (95% score confidence limit: 86.84% to 95.03%). However, the OG-MALDI MS method had lower sensitivity (33.9%) and accuracy (55.9%) than OS-MALDI MS. The MBT-SM MALDI MS method uses an integrated special algorithm for indirectly detecting KPC using the blaKPC-related peak (at 11,109 m/z) of the pKpQIL (Fig. S2). The presence of KPC was determined using the subtyping module included in the BioTyper system. The MBT-SM MALDI MS method had the lowest sensitivity (25.3%) and accuracy (50.2%) among all methods.
TABLE 2.
Diagnostic values of the three MALDI-TOF MS methods for detecting KPC protein in 437 clinical isolatesa
| Methods | No. of results |
Sensitivity (%) |
Specificity (%) |
Accuracy (%) |
Precision (%) |
AUC | |||
|---|---|---|---|---|---|---|---|---|---|
| TP | TN | FP | FN | ||||||
| OS | 280 | 145 | 0 | 12 | 95.9 | 100.0 | 97.3 | 100.0 | 0.979 |
| OG | 99 | 145 | 0 | 193 | 33.9 | 100.0 | 55.9 | 100.0 | 0.670 |
| MBT-SM | 74 | 145 | 0 | 218 | 25.3 | 100.0 | 50.2 | 100.0 | 0.627 |
AUC, area under the receiver operating characteristics curve; FN, false negative; FP, false positive; MALDI-TOF MS, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry; MBT-SM, Bruker’s MBT subtyping module; OG, n-octyl-β-d-glucoside; OS, osmotic shock; TN, true negative; TP, true positive.
Comparative analysis of the performance of the three MALDI MS methods for major Enterobacteriales.
The OS-MALDI MS and comparator methods (OG-MALDI MS and MBT-SM MALDI MS) were analyzed as a three-way comparison between the candidate method, the comparison method, and diagnostic accuracy criteria according to the CLSI EP12-A2 guideline. The difference in sensitivity between the OS-MALDI MS and OG-MALDI MS methods was 62.0% (95% CI for the difference = 54.56% to 66.51%). There was no significant difference in specificity between the two methods (95% CI for the difference = −3.73% to 1.91%). The difference in sensitivity between the OS-MALDI MS and MBT-SM MALDI MS methods was 70.6% (95% CI for the difference = 64.41% to 75.57%). Statistical analyses for nonparametric data from the methods (OS versus OG, OS versus MBT-SM, and OG versus MBT-SM) were performed using the Mann-Whitney test as shown in Fig. 2. OS-MALDI MS showed significantly higher sensitivity than OG (P-value < 0.05) and MBT-SM MADLI MS (P-value < 0.05). OG-MALDI MS was not significantly higher than MBT-SM MALDI MS (P-value > 0.05). There was no significant difference in specificity between the methods (95% CI for the difference = −3.73% to 2.59%).
FIG 2.
Box-and-whisker plot for comparing KPC protein peak intensities of the three methods. In each box plot, the bottom and top line of the boxes indicate the 25th and 75th percentiles, respectively. The horizontal line within each box represents the median. The red dotted line is the cutoff value of intensity for detecting KPC protein (square root transformation value is 6.84, as the value of peak intensity is 46.86). OS, OG, and MBT-SM MALDI MS are represented by red circles, blue squares, and black triangles, respectively.
The KPC protein was identified by all three methods in only 24 of 292 KPC-positive isolates (Fig. 3). In the largest proportion of isolates (135), the protein was identified by only the OS-MALDI MS method. Interestingly, the KPC protein in the K. variicola pneumoniae-related species was only confirmed by OS-MALDI MS (Table 1). Only three isolates (1.0%) were identified by just the MBT-SM MALDI MS method (Fig. 3). The remaining eight isolates (0.03% of 292) were not detected by any method.
FIG 3.
Venn diagram for the detection of KPC protein using the three methods. OS, OG, and MBT-SM methods are shown in red, blue, and black, respectively. OS and OG; OS and MBT-SM; and OS, OG, and MBT-SM are shown in light blue, green, and purple, respectively. The number of clinical isolates is shown for each condition.
Identification of KPC subtypes using the three MALDI-TOF MS methods.
The OS-MALDI MS method identified all KPC subtypes (KPC-2, KPC-4, KPC-19, and KPC-70) in various bacterial strains (Table 1). All subtypes were identified with high detection rates (>95%) by the OS-MALDI MS method. The detection rate was 95.9% for the KPC-2 subtype and 100.0% for all other subtypes: KPC-2 (276/288), KPC-4 (2/2), KPC-19 (1/1), and KPC-70 (1/1). Using the OG-MALDI MS method, the detection rate was 33.7% (97/288) for KPC-2, 100.0% (1/1) for KPC-19, and 100.0% (1/1) for KPC-70. Using the MBT-SM MALDI MS method, the detection rate was 25.3% (73/288) for KPC-2, 0% (0/0) for KPC-19, and 100.0% (1/1) for KPC-70. Both the OG-MALDI MS and MBT-SM MALDI MS methods were unable to detect any KPC-4 subtype.
DISCUSSION
The OS-MALDI MS method is an effective method for identifying periplasmic proteins (e.g., KPC) derived from Gram-negative bacteria and is expected to be used clinically for diagnosing antibiotic resistance. In this study, we demonstrated that KPC protein can be detected more accurately using osmotic lysis and MALDI MS in hundreds of clinical isolates. Additionally, compared with two other methods, OG-MALDI MS (direct KPC detection) and MBT-SM MALDI MS (indirect KPC detection), the OS-MALDI MS method yielded a much higher rate of correct identification.
Optimization of MS acquisition parameters in OS-MALDI MS method.
The Biotyper application uses typical mass range from 2,000 to 20,000 m/z, where most of ribosomal proteins can be detected. Because KPC protein has a molecular weight of 28.7 kDa, it is reasonable to use a higher mass range to detect it. Therefore, we selected a 12,000 to 32,000 m/z range to detect all peaks of single- and double-charged KPC protein. Furthermore, to increase detection sensitivity of the KPC target protein with a high molecular weight, we have tested the higher level of laser power (~60% → 80%), detector gain (~3 → 7.9), and accumulation shot (500 → 2,000) than that of Biotyper’s parameters, which targets mainly smaller size of ~10 kDa ribosomal proteins. Our optimal parameters allowed us a successful KPC detection with higher sensitivity and reproducibility on MALDI-TOF MS.
Superiority of OS-MALDI MS method.
OS-MALDI MS method showed great performance in sensitivity, mass accuracy, and the ability of KPC subtyping. As reported previously (20), MALDI-TOF MS analysis of KPC protein was most effective when using distilled water rather than detergent, salts, or solvents. Here, we confirmed that, as a salt- and detergent-free method, the OS-MALDI MS method is useful for identifying clinical isolates with KPC proteins. Even the analysis of the 98 samples simultaneously identified in both OS- and OG-MALDI MS methods (both direct methods) revealed that the KPC peak intensity in OS-MALDI MS was 1.2- up to 120-fold higher than OG-MALDI MS method, indicating that OS-MALDI MS was significantly more sensitive the OG-MALDI MS (Fig. 2; Fig. S3). When we compared OS- and OG-MALDI MS for the average mass difference between theoretical m/z and observed m/z of KPC peaks (ΔMtheo-obs), OS-MALDI MS showed ~10-fold higher mass accuracy than OG-MALDI MS: the mean value of ΔMtheo-obs of OS-MALDI MS was only −4.17 Da (n = 292) and that of OG-MALDI MS was −47.8 Da (n = 99). We presume that a salt or detergent-free condition in OS-MALDI MS allows substantial increase of mass accuracy in KPC detection, indicating that OS-MALDI MS is more applicable to KPC-subtyping. Interestingly, the OS-MALDI MS method certainly exhibited very good performance for identifying KPC subtypes (KPC-2, KPC-4, KPC-19, and KPC-70) in this study. Although the mass difference between KPC-2 and KPC-4 (or KPC-19 or KPC-70) was +19 Da (or +18 or +46 Da, respectively), all these KPC subtypes were sufficiently observed in the mass range. The result implies that the OS-MALDI MS method has a discriminant ability in CPE (or extended-spectrum β-lactamase) gene’s subtyping if the mass difference of subtypes or mass resolution is high enough. We suggest that simultaneous subtyping performance of OS-MALDI MS could be one of main advantages that differs prominently from other major methods for CPE identification without additional primer/substrate design or antibody in commercial RT-PCR or carbapenemase’s activity-based assays or antibody-based assays.
Likelihood of KPC detection in all bacterial species by OS-MALDI MS.
Comparative analysis of KPC detection performance of all three methods showed that only the OS-MALDI MS method could detect the KPC protein in all the species (eight species, K. pneumoniae; K. aerogenes; K. variicola; K. oxytoca; E. coli; C. koseri; E. cloacae; and A. baumannii). The MBT-SM MALDI MS method detected KPC only in two of eight species (K. pneumoniae and E. coli), as reported previously (23, 25) and OG-MALDI MS method detected KPC in six of eight species (except K. variicola and A. baumannii). In this study, we classified a total of seven K. variicola isolates and one A. baumannii isolate. PCR and DNA sequencing methods confirmed the presence of the blaKPC gene in five of the K. variicola isolates, as well as the A. baumannii isolate. PCR and DNA sequencing methods confirmed the presence of the blaKPC gene in five of the K. variicola isolates, as well as the A. baumannii isolate. Acinetobacter species have been known as one of the most important nosocomial opportunistic pathogens causing pneumoniae and sepsis (26, 27). In particular, carbapenem-resistant Acinetobacter are often multidrug-resistant bacteria that are simultaneously resistant to other antibiotics and are associated with significantly increased mortality and hospital length of stay (28, 29). K. variicola has also been recently reported as one of the most important nosocomial opportunistic pathogens (30, 31). Our OS-MALDI MS method identified KPC protein in four of the five K. variicola isolates (80.0%) and the A. baumannii isolate (100.0%), whereas the other two methods failed to detect KPC in isolates. In OS-MALDI MS, one of five K. variicola isolates was determined as KPC negative in our clinical test, according to an ambiguity and the peak shift being accidentally out of mass range in the spectrum. However, re-analysis of the K. variicola isolate confirmed a clear KPC peak in the mass range, correctly identified as a KPC positive isolate. This indicates that the introduction of internal mass calibration and improvement of peak resolution is to be considered a future project worth exploring to improve the accuracy of clinical assay. We are currently developing specific subjects which can increase mass accuracy and peak resolution using an internal calibrator. We expect that it could ultimately provide subtyping performance for all kinds of carbapenemases.
Proceeding of an easy, rapid, and affordable OS-MALDI MS method.
Sample preparation of OS-MALDI MS includes a three-step procedure: colony incubations, centrifugations, and spotting on MALDI plate. The turnaround time (TAT) by OS-MALDI MS analysis is short and estimated at <45 min (if optimized, the TAT can be shortened by <25 min). Furthermore, because OS-MALDI MS uses osmotic shock with distilled water and centrifugation, the cost of the method is affordable, similar to conventional bacterial identification by Biotyper. However, if bacterial identification is required, it takes more time due to the need for additional Biotyper analysis.
In conclusion, this study demonstrated the effectiveness of three different methods of MALDI-TOF MS for detecting KPC protein and showed that OS-MALDI MS was the most effective sample preparation method for detecting intact KPC in clinical isolates. We used a sample preparation method suitable for MALDI-based direct KPC protein detection methods and confirmed the high sensitivity and accuracy of OS-MALDI MS. Our results also suggest that the OS-MALDI MS method may be clinically applied for the detection of other disease-related proteins, including their subtypes in Gram-negative bacteria (e.g., KPC-4, OXA-232, CTX-M-15).
Footnotes
Supplemental material is available online only.
Contributor Information
Je-Hyun Baek, Email: jhbaek@mf.seegene.com.
Alexander Mellmann, University Hospital Munster.
REFERENCES
- 1.Nicolau DP. 2008. Carbapenems: a potent class of antibiotics. Expert Opin Pharmacother 9:23–37. 10.1517/14656566.9.1.23. [DOI] [PubMed] [Google Scholar]
- 2.Codjoe FS, Donkor ES. 2017. Carbapenem resistance: a review. Med Sci (Basel) 6:1. 10.3390/medsci6010001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kadri SS. 2020. Key takeaways from the U.S. CDC's 2019 antibiotic resistance threats report for frontline providers. Crit Care Med 48:939–945. 10.1097/CCM.0000000000004371. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.WHO . 2015. Global action plan on antimicrobial resistance. [Google Scholar]
- 5.ECDC. 2019. Antimicrobial resistance in the EU/EEA (EARS-Net) - Annual epidemiological report for 2019. https://wwwecdceuropaeu/sites/default/files/documents/surveillance-antimicrobial-resistance-Europe-2019pdf.
- 6.Bonomo RA, Burd EM, Conly J, Limbago BM, Poirel L, Segre JA, Westblade LF. 2018. Carbapenemase-producing organisms: a global scourge. Clin Infect Dis 66:1290–1297. 10.1093/cid/cix893. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Rodriguez-Bano J, Gutierrez-Gutierrez B, Machuca I, Pascual A. 2018. Treatment of infections caused by extended-spectrum-beta-lactamase-, AmpC-, and carbapenemase-producing Enterobacteriaceae. Clin Microbiol Rev 31. 10.1128/CMR.00079-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Cui X, Zhang H, Du H. 2019. Carbapenemases in Enterobacteriaceae: detection and antimicrobial therapy. Front Microbiol 10:1823. 10.3389/fmicb.2019.01823. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Osei Sekyere J, Govinden U, Essack SY. 2015. Review of established and innovative detection methods for carbapenemase-producing Gram-negative bacteria. J Appl Microbiol 119:1219–1233. 10.1111/jam.12918. [DOI] [PubMed] [Google Scholar]
- 10.Baeza LL, Pfennigwerth N, Greissl C, Gottig S, Saleh A, Stelzer Y, Gatermann SG, Hamprecht A. 2019. Comparison of five methods for detection of carbapenemases in Enterobacterales with proposal of a new algorithm. Clin Microbiol Infect 25:1286e9–1286e15. 10.1016/j.cmi.2019.03.003. [DOI] [PubMed] [Google Scholar]
- 11.Girlich D, Poirel L, Nordmann P. 2012. Value of the modified Hodge test for detection of emerging carbapenemases in Enterobacteriaceae. J Clin Microbiol 50:477–479. 10.1128/JCM.05247-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Nordmann P, Poirel L, Dortet L. 2012. Rapid detection of carbapenemase-producing Enterobacteriaceae. Emerg Infect Dis 18:1503–1507. 10.3201/eid1809.120355. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kuchibiro T, Komatsu M, Yamasaki K, Nakamura T, Nishio H, Nishi I, Kimura K, Niki M, Ono T, Sueyoshi N, Kita M, Kida K, Ohama M, Satoh K, Toda H, Mizutani T, Fukuda N, Sawa K, Nakai I, Kofuku T, Orita T, Watari H, Shimura S, Fukuda S, Nakamura A, Wada Y. 2018. Evaluation of the modified carbapenem inactivation method for the detection of carbapenemase-producing Enterobacteriaceae. J Infect Chemother 24:262–266. 10.1016/j.jiac.2017.11.010. [DOI] [PubMed] [Google Scholar]
- 14.Bernabeu S, Poirel L, Nordmann P. 2012. Spectrophotometry-based detection of carbapenemase producers among Enterobacteriaceae. Diagn Microbiol Infect Dis 74:88–90. 10.1016/j.diagmicrobio.2012.05.021. [DOI] [PubMed] [Google Scholar]
- 15.Liu H, Du Z, Wang J, Yang R. 2007. Universal sample preparation method for characterization of bacteria by matrix-assisted laser desorption ionization-time of flight mass spectrometry. Appl Environ Microbiol 73:1899–1907. 10.1128/AEM.02391-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Ruelle V, El Moualij B, Zorzi W, Ledent P, Pauw ED. 2004. Rapid identification of environmental bacterial strains by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Rapid Commun Mass Spectrom 18:2013–2019. 10.1002/rcm.1584. [DOI] [PubMed] [Google Scholar]
- 17.Allen AD, Velez-Quinones M, Eribo BE, Morris V. 2015. MALDI-TOF MS as a supportive tool for the evaluation of bacterial diversity in soils from Africa and the Americas. Aerobiologia (Bologna) 31:111–126. 10.1007/s10453-014-9351-5. [DOI] [Google Scholar]
- 18.Yu J, Liu J, Li Y, Yu J, Zhu W, Liu Y, Shen L. 2018. Rapid detection of carbapenemase activity of Enterobacteriaceae isolated from positive blood cultures by MALDI-TOF MS. Ann Clin Microbiol Antimicrob 17:22. 10.1186/s12941-018-0274-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Yoon EJ, Lee EH, Hwang DH, Lee H, Baek JH, Jeong SH. 2020. Direct detection of intact Klebsiella pneumoniae carbapenemases produced by Enterobacterales using MALDI-TOF MS. J Antimicrob Chemother 75:1174–1181. 10.1093/jac/dkaa007. [DOI] [PubMed] [Google Scholar]
- 20.Cheon DH, Lee S, Yang WS, Hwang S, Jang H, Kim MJ, Baek JH. 2021. Optimization of a lysis method to isolate periplasmic proteins from Gram-negative bacteria for clinical mass spectrometry. Prot Clin Appl 15:e2100044. 10.1002/prca.202100044. [DOI] [PubMed] [Google Scholar]
- 21.Wang J, Wang H, Cai K, Yu P, Liu Y, Zhao G, Chen R, Xu R, Yu M. 2021. Evaluation of three sample preparation methods for the identification of clinical strains by using two MALDI-TOF MS systems. J Mass Spectrom 56:e4696. 10.1002/jms.4696. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Cordovana M, Kostrzewa M, Glandorf J, Bienia M, Ambretti S, Pranada AB. 2018. A full MALDI-based approach to detect plasmid-encoded KPC-producing Klebsiella pneumoniae. Front Microbiol 9:2854. 10.3389/fmicb.2018.02854. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Lau AF, Wang H, Weingarten RA, Drake SK, Suffredini AF, Garfield MK, Chen Y, Gucek M, Youn JH, Stock F, Tso H, DeLeo J, Cimino JJ, Frank KM, Dekker JP. 2014. A rapid matrix-assisted laser desorption ionization-time of flight mass spectrometry-based method for single-plasmid tracking in an outbreak of carbapenem-resistant Enterobacteriaceae. J Clin Microbiol 52:2804–2812. 10.1128/JCM.00694-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Clinical and Laboratory Standard Institute. 2008. User protocol for evaluation of qualitative test performance; approved guideline. Clinical and Laboratory Standard Institute, Wayne PA. [Google Scholar]
- 25.Youn JH, Drake SK, Weingarten RA, Frank KM, Dekker JP, Lau AF. 2016. Clinical performance of a matrix-assisted laser desorption ionization-time of flight mass spectrometry method for detection of certain blaKPC-containing plasmids. J Clin Microbiol 54:35–42. 10.1128/JCM.01643-15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Fournier PE, Richet H. 2006. The epidemiology and control of Acinetobacter baumannii in health care facilities. Clin Infect Dis 42:692–699. 10.1086/500202. [DOI] [PubMed] [Google Scholar]
- 27.Garza-Gonzalez E, Llaca-Diaz JM, Bosques-Padilla FJ, Gonzalez GM. 2010. Prevalence of multidrug-resistant bacteria at a tertiary-care teaching hospital in Mexico: special focus on Acinetobacter baumannii. Chemotherapy 56:275–279. 10.1159/000319903. [DOI] [PubMed] [Google Scholar]
- 28.Falagas ME, Karveli EA, Kelesidis I, Kelesidis T. 2007. Community-acquired Acinetobacter infections. Eur J Clin Microbiol Infect Dis 26:857–868. 10.1007/s10096-007-0365-6. [DOI] [PubMed] [Google Scholar]
- 29.Siegman-Igra Y, Bar-Yosef S, Gorea A, Avram J. 1993. Nosocomial Acinetobacter meningitis secondary to invasive procedures: report of 25 cases and review. Clin Infect Dis 17:843–849. 10.1093/clinids/17.5.843. [DOI] [PubMed] [Google Scholar]
- 30.Andrade BG, de Veiga Ramos N, Marin MF, Fonseca EL, Vicente AC. 2014. The genome of a clinical Klebsiella variicola strain reveals virulence-associated traits and a pl9-like plasmid. FEMS Microbiol Lett 360:13–16. 10.1111/1574-6968.12583. [DOI] [PubMed] [Google Scholar]
- 31.Garza-Ramos U, Silva-Sanchez J, Barrios H, Rodriguez-Medina N, Martinez-Barnetche J, Andrade V. 2015. Draft genome sequence of the first hypermucoviscous Klebsiella variicola clinical isolate. Genome Announc 3. 10.1128/genomeA.01352-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental material. Download jcm.01062-22-s0001.pdf, PDF file, 0.5 MB (510.8KB, pdf)