Notes
High-Throughput Identification of Coliforms in Processed Soybean Products Using Matrix-Assisted Laser Desorption Ionization Time-of-Flight Mass Spectrometry
2014 Volume 20 Issue 5 Pages 1093-1097
Details
2014 Volume 20 Issue 5 Pages 1093-1097
Standard methods for bacterial identification include culturing and biochemical identification. These methods are time-consuming because they require culturing in different media for at least a few days. Matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) has recently been demonstrated as a rapid, reliable method for bacterial identification. In this study, we developed an improved sample preparation method to accurately identify coliforms in processed soybean products using MALDI-TOF MS. Our method included harvesting bacterial cells using isoelectric precipitation and anionic surfactant treatment to eliminate food matrices such as proteins and fats. At least 107 colony-forming units (CFU) / mL of typical test strains of Escherichia coli, Klebsiella pneumoniae, and Citrobacter freundii were successfully identified in soy milk samples. After sample incubation steps at 37°C for 18 h, low levels of coliforms (101 CFU/mL) could be identified. Our proposed procedure was verified by accurately identifying coliforms in commercial processed soybean products.
Analytical methods for bacterial identification based on Gram staining, culturing, and biochemical assays for phenotypic characterization have remained unchanged for many years. Several rapid detection methods are currently used in the food industry (Gracias and Mckillip, 2004). PCR is one of the most useful of these methods because of its high sensitivity. However, PCR requires extraction of nucleic acids from food samples, and is occasionally inhibited by food matrices such as fats and proteins (Schrader et al., 2012; Gadkar and Filion, 2013).
Based on the ribosomal protein profiles of different bacteria, matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) was recently shown to be a rapid, reliable method for bacterial identification (Lay, 2001; Mazzeo et al., 2006). Considering its speed and accuracy, MALDI-TOF MS may become a popular and routinely used method for bacterial identification in food microbiology (Angelakis et al., 2011; Böhme et al., 2011; Hochel et al., 2012; Kuda et al., 2014). Therefore, identifying bacteria, such as coliforms, using MALDI-TOF MS obviates the requirements for classical culture methods, considerably shortening the time required for bacterial identification.
The increased consumption of processed soybean products worldwide is generating demand for high quality and safely processed soybean products. Food-borne pathogenic and spoilage bacteria can be present as residual microbiota or can be introduced as contaminants during processing. Furthermore, bacterial species play an important role in soybean food-borne illness and spoilage. Thus, rapid identification of bacteria is as important as enumeration of bacteria (Katase and Tsumura, 2011; Tsumura and Tsuboi, 2012; Katase et al., 2013).
Bacterial separation methods to harvest bacterial cells from complex food matrices interfering with their accurate detection have been investigated (Benoit and Donahue, 2003; Stevens and Jaykus, 2004; Fukushima et al., 2007). Moreover, it is difficult to detect and identify bacteria in foods that contain fat matrices (Schrader et al., 2012; Gadkar and Filion, 2013). In particular, there is little information regarding non-culture-based identification of bacteria in food samples using MALDI-TOF MS (Ferreira et al., 2010; Barreiro et al., 2012; Furukawa et al., 2013).
In this study, we evaluated an improved sample preparation method for use in the accurate, high-throughput identification of coliforms in processed soybean products using MALDI-TOF MS.
Bacterial strains and culture conditions Typical test strains, including Escherichia coli NCTC9001, Klebsiella pneumoniae NBRC14940, and Citrobacter freundii NBRC12681, were used as artificial contaminants for coliform testing. These strains were cultured in lactose broth media (Nissui Seiyaku, Tokyo, Japan) and incubated at 37°C for 18 h. After incubation, bacterial cells were harvested by centrifugation at 11,000 × g for 10 min (Himac CR22; Hitachi Koki, Tokyo, Japan). When enrichment culture was used, 15 mL of soy milk with 15 mL of 2× brilliant-green bile lactose broth (BRILA broth; Merck, Darmstadt, Germany) was prepared and incubated at 37°C for 18 h.
Processed soybean product samples Soy milk was obtained from Fuji Oil Co., Ltd., (Osaka, Japan). Soy milk samples with different fat contents were prepared by mixing soy milk and soybean oil (Fuji Oil Co., Ltd.) using an IKA T-18 basic Ultra Turrax homogenizer (IKA Works, Inc., Wilmington, NC, USA). The fat content of soy milk was expressed as percent solid content. Twenty processed soy products were purchased from different retail outlets in the Osaka region of Japan and stored at 4°C until analysis. All samples were examined within 2 days of purchase.
Isoelectric precipitation and surfactant treatment To separate coliforms in processed soybean products, conventional isoelectric precipitation was performed. To 30 mL of a soy milk sample, 75 µL of concentrated hydrochloric acid was added (Wako Pure Chemical Industries, Ltd., Osaka, Japan), after which the sample was incubated at 45°C for 5 min. Consequently, soy proteins in the sample precipitated, and a crude supernatant was collected. This supernatant (1 mL) was centrifuged at 13,000 ×g for 2 min (Model 3615; Kubota Corporation, Tokyo, Japan), after which the remaining supernatant was carefully discarded using a 1-mL pipette. To remove the remaining soy protein and fat matrices, bacterial cells in the pellet were suspended in 1 mL of 0.5% sodium dodecyl sulfate (SDS; Wako Pure Chemical Industries) solution. Bacterial cells were harvested by a second centrifugation step at 13,000×g for 2 min, and the supernatant was carefully discarded. Bacterial cells in the pellet were suspended in 1 mL of sterile water and centrifuged at 13,000×g for 2 min. This supernatant was discarded, and the microbial cells in the pellet were suspended in 300 µL of sterile water. This cell suspension was subsequently used for the direct identification of coliforms using MALDI-TOF MS.
Sample preparation from commercial processed soybean products Before bacterial cell recovery, a culture enrichment step was conducted because processed foods generally contain few viable coliforms cells. Twenty samples of commercial soybean products were evaluated by culture enrichment prior to MALDI-TOF MS. A processed soybean product (10 g) was mixed with 90 mL of BRILA broth and incubated at 37°C for 18 h. The increased number of coliform cells in samples was then subjected to MALDI-TOF MS analysis and 16S rRNA gene sequencing. Next, 1 mL of this suspension was filtered through a stomacher bag (Tempo bag, bioMérieux, Lyon, France) and centrifuged at 13,000 × g for 2 min. The harvested coliforms cells were washed twice with SDS solution to ensure complete removal of fat matrices. Coliform pellets were suspended in 0.3 mL of sterile water and used for MALDI-TOF MS and 16S rRNA gene sequencing; 16S rRNA gene sequencing was performed as described previously (Furukawa et al., 2013).
MALDI-TOF MS MALDI-TOF MS was performed according to standard procedures (La Scola and Raoult, 2009). In brief, 300 µL of bacterial cell suspension was mixed with 900 µL of absolute ethanol (99.5%; Kanto Kagaku, Tokyo, Japan) and centrifuged at 13,000 ×g for 2 min. The supernatant was discarded and the residual ethanol was removed by repeated centrifugation. Then, 10 µL of formic acid (70%; Kanto Kagaku) was added to the pellet and thoroughly mixed by pipetting, followed by the addition of 10 µL of acetonitrile (98%; Kanto Kagaku). This mixture was centrifuged at 13,000 ×g for 2 min. The supernatant (1 µL) was spotted on a steel target and air-dried at room temperature. The sample spots were overlaid with 1 µL of matrix solution [saturated solution of α-cyano-4-hydroxycinnamic acid (HCCA) in an organic solvent (50% acetonitrile and 2.5% trifluoroacetic acid)] and air-dried at room temperature. Measurements were performed using an Autoflex II MALDI-TOF/TOF mass spectrometer (Bruker Daltonics, Leipzig, Germany) equipped with a 20-Hz nitrogen laser. Spectra were recorded in the linear positive ion mode at a laser frequency of 20 Hz within a mass range of 2 – 20 kDa. The IS1 voltage was 20 kV, the IS2 voltage was maintained at 18.5 kV, the lens voltage was 6 kV, and the extraction delay time was 250 ns. For each spectrum, 300 laser shots were acquired and analyzed. The spectra were externally calibrated using a standard calibration mixture (Bruker Daltonics). For automated data analysis, raw spectra were processed using MALDI Biotyper 3.0 software (Bruker Daltonics) with default settings. To identify unknown bacteria, each peak was directly matched against reference libraries (3,995 strains) using the integrated pattern-matching algorithm incorporated in the Biotyper 3.0 software. Identifications obtained using MALDI-TOF MS were evaluated according to modified scores (ranging from 0 to 3) proposed by the manufacturer. A score of more than 1.7 indicated probable bacterial identification, whereas a score of less than 1.7 indicated no reliable bacterial identification.
Standard plate method The conventional plate method was tested according to “Standard Methods of Analysis in Food Safety Regulation—Biology— ” (Ministry of Health, Labor and Welfare, Japan, 2004) using the pour plate method. Total coliform counts were enumerated after inoculation on desoxycholate agar medium (Merck) and incubation at 37°C for 24 h. The final results are expressed as CFU/mL.
Fat matrices interfere with accurate identification of coliforms in soy milk Soy milk samples containing different fat content (40% – 70% solids content) were prepared and spiked with E. coli at 108 CFU/mL. After anionic surfactant treatment (0.5% SDS), accurate identification scores were obtained using soy milk samples with up to 70% fat content (Table 1). When using a previously described method for removing proteins from soy milk (Furukawa et al., 2013), MALDI-TOF MS scores were less than 2.0 when the fat content of soy milk samples was 50% and greater. This indicated that residual fat matrices may have interfered with the raw mass spectra and affected spectrum matching. Preliminary experiments revealed that 0.5% SDS treatment was sufficient for soy milk samples ranging from 40% to 70% fat content.
| Percent fat (solid content) | MALDI-TOF/MS score | |
|---|---|---|
| Isoelectric precipitation and anionic surfactant treatment | Isoelectric precipitation | |
| 40% | 2.2 | 2.0 |
| 50% | 2.1 | 1.9 |
| 60% | 2.2 | 1.8 |
| 70% | 2.2 | 1.8 |
Soy milk samples were spiked with E. coli at 108 CFU/mL.
In this study, to separate coliforms in processed soybean products, we used an isoelectric precipitation method commonly used to prepare soy protein isolates (Petenate and Glatz, 1983; Tsumura et al., 2004) in combination with SDS treatment. The recovery rates of bacterial cells were estimated to be between 50% and 90% (data not shown). In bacterial identification of blood samples using MALDI-TOF MS, a method employing an anionic surfactant has been known to dissolve blood particles (Weller and Maier, 2011).
Rapid detection methods, including PCR, are occasionally inhibited by fat matrices (Schrader et al., 2012; Gadkar and Filion, 2013). In addition, it is necessary to remove the fatty phase for the bacterial inspection of butter (International Organization for Standardization, 2010). In this manner, fat matrices inhibit amplification of oligonucleotides in DNA-based analysis, and disrupt accurate colony counts in the standard plate method. The remaining fat matrices were removed by SDS treatment after eliminating soy proteins using isoelectric precipitation. Therefore, accurate identification with MALDI-TOF MS was achieved according to the surfactant treatment in this study.
Next, soy milk samples were spiked with different E. coli concentrations to determine the minimum identification limit using this sample treatment. E. coli could be correctly identified when present at a minimum of 107 CFU/mL (Table 2). This result indicates an improvement in the identification limit was obtained compared to previous results (Furukawa et al., 2013). K. pneumoniae and C. freundii could also be identified with adequate identification scores when present at the same concentration (data not shown).
| Microorganism | Spiked coliform count (CFU/mL) | MALDI-TOF MS score |
|---|---|---|
| 3 × 109 | 2.3 | |
| E. coli | 3 × 108 | 2.3 |
| 4 × 107 | 2.0 | |
| 5 × 106 | NI |
NI, no reliable identification.
Soymilk containing 60% fat (solid content) was used in these tests.
Identification of coliforms in soy milk after culture enrichment Considering the low concentrations of coliforms usually found in processed soybean products, culture enrichment was performed for accurate identification. Three coliform test strains in soy milk (101 CFU/mL each) were incubated at 37°C for 18 h in BRILA broth, and the increased coliform numbers (108 CFU/mL) were accurately identified without using colony formation by plate culture (Table 3). When culture enrichment was performed using BRILA broth, poor separation of fat matrices occurred during the isoelectric precipitation step. This was probably due to the presence of bile acids, which are emulsifiers, in BRILA broth. Bile acids are commonly found in selective enrichment broths used for coliforms. However, there was no interference from fat matrices and no problems with bacterial identification when using MALDI-TOF MS because these cells were washed away with the SDS solution.
| CFU/mL | |||
|---|---|---|---|
| Microorganism | Initial coliform count | Coliform count after culture enrichment | MALDI-TOF MS score |
| E. coli | 3 × 101 | 6 × 108 | 2.3 |
| K. pneumoniae | 4 × 101 | 3 × 108 | 2.2 |
| C. freundii | 3 × 101 | 5 × 108 | 2.2 |
Soymilk containing 60% fat (solid content) was used in these tests.
Identification of coliforms in processed soybean products To determine the feasibility of our proposed identification method using MALDI-TOF MS, commercial processed soybean products were evaluated. In four of the 20 collected samples, increased coliform numbers were observed after enrichment culture. Coliforms determined in these samples using 16S rRNA gene sequencing were successfully identified by MALDI-TOF MS with adequate identification scores (Table 4).
| Product | MALDI-TOF MS score1) | 16S rRNA2) | CFU/mL | |
|---|---|---|---|---|
| Coliform count before culture enrichment | Coliform count after culture enrichment | |||
| Fried-Tofu 1 | Enterobacter cloacae (2.3) | Enterobacter sp. (80%) Klebsiella pneumoniae |
1 × 101 | 2 × 108 |
| Fried-Tofu 2 | Klebsiella pneumoniae (2.3) | Klebsiella pneumoniae (80%) Enterobacter cloacae, Enterobacter aerogenes |
4 × 102 | 3 × 108 |
| Fried-Tofu 3 | Escherichia vulneris (2.1) | Escherichia vulneris (90%) Enterobacter sacchari |
4 × 101 | 1 × 108 |
| Soy milk | Enterobacter cancerogenus (1.8) | Enterobacter sp. (100%) | 0 × 101 | 2 × 107 |
Contamination of processed foods is not necessarily due to a single bacterial species. However, in most cases the concentration of coliforms in commercial processed products is comparatively low. The dominant bacterial species can be identified after selective culture enrichment. The isolated coliforms from commercial soybean products belonged to the Enterobacteriaceae family, ubiquitous microbes commonly found in soybean food (Préstamo et al., 2000; Ananchaipattana et al., 2012). Currently, in order to prevent false-positive results, confirmation tests are required after the successful detection of coliforms cultured in BRILA broth. However, because MALDI-TOF MS can directly identify bacterial cells, these tests are not required to confirm contamination by coliforms. Considering the discriminatory capabilities of MALDI-TOF MS, this direct identification method for bacteria in foods could be used for process monitoring, quality control, and confirming the presence of bacteria identified by other bacterial tests.
In conclusion, our proposed method using MALDI-TOF MS demonstrated high-throughput detection of coliforms in fat-containing processed soybean products. It is anticipated that this will be useful for product inspections to ensure the safety of processed soybean foods.
