Abstract
In this study, matrix-assisted laser desorption ionisation time of flight mass spectrometry (MALDI-TOF MS) was used to identify bacteria from environmental matrices. The aim of this work was to determine the efficacy of this rapid technique and the bacterial community of agricultural samples. Environmental samples included the collection of irrigation waters and manures, and bacteria from the surface of vegetables were also investigated. From food safety point of view, the investigation of these microbial communities is inevitable considering their potential hazardous impact on the food production chain. Altogether 235 bacterial isolates were identified with the most frequent genera being Pseudomonas, Bacillus, Acinetobacter and Aeromonas. Our results indicated that MALDI-TOF MS can be used to identify causative agents of foodborne illnesses, food spoilage and common plant pathogens. However, limitations of the rapid identification technique were also encountered as we obtained correct identification at species level for 30.2% and at genus level for 69.8% of the isolates.
1 Introduction
Consumption of raw vegetables and fruits has been responsible for various foodborne illnesses worldwide in the past decades (Kampmeier et al., 2018; Turner et al., 2019). Since many pathogenic bacteria (Listeria monocytogenes, verotoxigenic Escherichia coli, Salmonella spp., E. coli O157:H7) are able to survive and grow in low nutrient environment (Falardeau et al., 2017), the sources of the microbial hazards could be traced back to irrigation water as well. Therefore, to ensure food safety from farm to fork, the food production chain involving fresh produce should be monitored. Accurate and fast identification of the most common foodborne bacterial pathogens (Campylobacter spp., L. monocytogenes, Salmonella spp., Staphylococcus aureus, Clostridium perfringens, E. coli and other shiga toxin-producing E. coli strains (STEC) (Scallan et al., 2011)) could prevent the growing number of foodborne outbreaks.
Traditional microbiological methods (morphological, physiological, and biochemical tests) and/or staining (e.g. Gram-staining), or molecular methods (e.g. 16S rRNA gene sequence analysis) with the need of trained laboratory personnel, high costs, and being time-consuming, are inadequate for rapid bacterial identification (Böhme et al., 2011; Alnakip et al., 2020). In recent years, MALDI-TOF MS (matrix assisted laser desorption ionisation-time of flight mass spectrometry) has become a popular technique in microbiological identification. The application of MALDI-TOF MS in food microbiology offers a faster, less expensive, and labour-saving technique compared to the above mentioned methods (Böhme et al., 2011; Strejcek et al., 2018). Identification via MALDI-TOF MS is carried out by comparing the PMF (Protein Mass Fingerprint) of the tested microbe to the data of PMF databases, or by pairing the masses of the identified biomarkers of unknown organisms using proteomic databases. Two of the largest systems based on mass spectral microbial identification are BioTyper® (Bruker Daltonics GmbH & Co, Bremen, Germany) and VITEK® MS Plus (bioMérieux, Marcy l’Etoile, France) (Singhal et al., 2015).
MALDI-TOF MS is mostly used for the identification of clinically relevant pathogens. It has been reported that this system can effectively identify bacteria even at species level from clinical samples (van Veen et al., 2010; Ponderand et al., 2020; Chung et al., 2021). However, it is increasingly utilised to detect foodborne pathogens in the complex food chain as well (Böhme et al., 2011; de Koster and Brul, 2016; Horváth et al., 2020). Still, the application of MALDI-TOF MS to characterise the microbial composition of environmental samples such as from waste disposal sites (Kopcakova et al., 2014), soil samples (Strejcek et al., 2018), or from mining samples (Avanzi et al., 2017), seemed to be less competent compared to the results of clinical studies.
Therefore, the aims of our study were to describe the performance and limitation of MALDI-TOF MS using different types of agricultural samples and to characterise the bacterial communities of the environment.
2 Materials and methods
2.1 Bacterial isolation
In this study, irrigation water samples were collected and examined from different regions of Hungary. Sampling sites were chosen due to their utilisation as irrigation water. Sampling was also done in different regions, where irrigation water contamination (e.g. by manure) could have occurred or from irrigated crops (e.g. corn, lettuce, onion, sorrel, spinach, and tomato), where irrigation water might have transmitted the microbes to them. Two groups were formed from these crops in accordance with their origins as Vegetables1 and Vegetables2. Group of Vegetables1 includes onion, corn, lettuce, spinach, while group of Vegetables2 contains tomato, spinach, and sorrel. Manure samples originated from different swine farms located in Hungary. Manure2 and Manure3 were from the same sample spot (Bátya, Southern Hungary), however, Manure2 was liquid sample. Three samples from wells in different towns are marked as Irrigation water samples. Details of samples and names of them as referred later are shown in Table 1.
Origin of samples and bacterial isolates. All analysed samples were collected from the area of Hungary (HU)
| Types | Sample name | Origin of samples* | Location (city, region) | Isolates | |
| Gram-positive | Gram-negative | ||||
| Still water | Lake1 | Kavicsos-tó | Szigetszentmiklós (Central HU) | 1 | 5 |
| Lake2 | Szelidi-tó | Dunapataj (Central HU) | 1 | 10 | |
| Running water | River1 | Tisza | Tiszakécske (Eastern HU) | 5 | 14 |
| River2 | Tisza | Szolnok (Eastern HU) | 0 | 13 | |
| River3 | Danube | Csepel (Central HU) | 8 | 23 | |
| River4 | Danube | Kalocsa (Southern HU) | 1 | 7 | |
| River5 | Vajas | Bátya (Southern HU) | 0 | 14 | |
| Well | Irrigation water1 | Soroksár | Soroksár (Central HU) | 6 | 8 |
| Irrigation water2 | Debrecen | Debrecen (Eastern HU) | 3 | 4 | |
| Irrigation water3 | Nagykunsági-főcsatorna | Abádszalók (Eastern HU) | 2 | 5 | |
| Liquid manure | Manure1 | Békéscsaba | Békéscsaba (Eastern HU) | 4 | 4 |
| Manure2 | Bátya | Cegléd (Eastern HU) | 7 | 28 | |
| Manure | Manure3 | Bátya | Bátya (Southern HU) | 7 | 13 |
| Manure4 | Cegléd | Cegléd (Eastern HU) | 8 | 0 | |
| Vegetables | Vegetables1 | Soroksár | Soroksár (Central HU) | 6 | 7 |
| Vegetables2 | Debrecen | Debrecen (Eastern HU) | 14 | 7 | |
*: Two samples per sample spots were taken regarding every sample type.
Bacterial isolation was performed after preparing a ten-fold serial dilution in buffered peptone water (BPW) (Thermo Fisher Scientific Inc., Oxoid Ltd., Basingstoke, UK) up to dilution 10−3. The isolates were plated on Trypticase Soy Agar (TSA, Biokar Diagnostics, Allonne, France) plates. Agar plates were incubated at 30 °C to grow overnight cultures. Extended direct transfer procedure was used to identify the isolates as outlined in the User Manual provided by the manufacturer (Bruker Daltonics GmbH & Co, Bremen, Germany). The identification process was done using MALDI Biotyper (Bruker Daltonics GmbH & Co, Bremen, Germany). Furthermore, beside MALDI-TOF MS identification, catalase and oxidase activities of the isolates were in concordance with the MALDI-TOF MS results.
2.2 MALDI-TOF MS data acquisition and processing
MALDI-TOF MS spectra of the samples were collected using a Microflex LT/SH (Bruker Daltonics GmbH & Co, Bremen, Germany) mass spectrometer equipped with a nitrogen laser (lambda = 337 nm) at a laser frequency of 60 Hz operating in linear positive ion detection mode under MALDI Biotyper Compass 3.0 and FlexControl 3.4 (Bruker Daltonics GmbH & Co, Bremen, Germany). Mass spectra were acquired in the range of 2,000–21,000 Da for each sample analysed for species level microbial identification. The spectra were generated from 240 single spectra that were created in 40-laser-shot steps from random positions of each isolate. The system was calibrated using E. coli ribosomal protein standard derived from the manufacturer (Bruker IVD Bacterial Test Standard). FlexControl 3.4 and FlexAnalysis 3.4 (Bruker Daltonics GmbH & Co, Bremen, Germany) were used for data acquisition and data processing, respectively.
Results of the identifications were categorised following the standard identification scores provided by the manufacturer. High-confidence identification indicates a score in the range of 2.00–3.00, which means reliable identification at species level. Low-confidence identification is accepted at genus level, with the score of 1.7–1.99. Scores below 1.7 are considered as not reliable identifications.
3 Results and discussion
Bacterial isolates in this study included 235 isolates from different water, vegetables, and manure samples. The samples contained more Gram-negative isolates than Gram-positive ones. In the case of Gram-negative bacteria, the software produced better results compared to Gram-positive bacteria, as a bigger proportion of those were identified both at species and genus levels. In this study, Biotyper could not identify, even at genus level, a significant part of total isolates (Table 2).
Identification results of MALDI-TOF MS of bacteria isolated from environmental samples
| Organisms | MALDI-TOF MS identification scores | |||
| Isolates | Species identification (2–3) | Genus identification (>1.7) | Not reliable identification (0–1.69) | |
| Gram-positive bacteria | 73 | 18 (24.7%) | 45 (61.6%) | 28 (38.4%) |
| Gram-negative bacteria | 162 | 53 (32.7%) | 119 (73.5%) | 43 (26.5%) |
| Total | 235 | 71 (30.2%) | 164 (69.8%) | 71 (30.2%) |
Kopcakova et al. (2014) used MALDI-TOF MS to identify the microbiota from waste disposal sites having an identification rate lower than 20% at species level, and Strejcek et al. (2018) reported 35% correct identifications at species level from soil samples. Similarly, we achieved 30.2% correct species identification, however, our study involved more isolates (235) than the aforementioned two (22 and 49, respectively).
Among isolated bacteria, Pseudomonas was the most abundantly occurring genus as 45 isolates belonged to that group. It has been widely represented in the samples of River (1–5), Lake (1–2), Irrigation water1, and Manures (1–2) (Table 3). Plant pathogens belonging to Pseudomonas fluorescens group were isolated and identified such as Pseudomonas marginalis, which causes rots of plant tissues, and Pseudomonas azotoformans. Pseudomonas alcaligenes and Pseudomonas veronii, both used for bioremediation purposes, and Pseudomonas extremorientalis, usually found in drinking water, were also isolated from water samples (River1-5).
Bacterial isolates identified by MALDI-TOF MS from environmental samples. Coloured cells indicate bacterial genera, within which more species were identified from the samples. Identified species are not listed in the table as those are presented in the text
 |
As a ubiquitous genus in nature, species of Bacillus were also frequently found in most samples (Table 3). At least one isolate of genus Bacillus was found in every type of sample (irrigation water, vegetable, manure). However, most of the Bacillus isolates were found on vegetables. Bacillus cereus and Bacillus licheniformis, causative agents of foodborne illnesses or food spoilage, were found in both groups of Vegetables1 and 2.
The genus of Acinetobacter, a common one in soil and water, was identified from different types of samples (water, manure, vegetable). Acinetobacter lwoffii, an opportunistic human pathogen, and Acinetobacter johnsonii, part of the human skin flora, as well as Acinetobacter pittii and Acinetobacter calcoaceticus both belonging to Acinetobacter baummannii complex were also identified.
Members of the genus Aeromonas were also recurring with isolates identified from different water samples (Lake1-2, River1, River3, River5). Aeromonas hydrophila, a human and fish pathogen, and Aeromonas salmonicida, which infects salmon, were also identified.
In addition, MALDI-TOF MS could identify several Gram-negative ubiquitous bacteria from the environmental samples at species level. Pantoea agglomerans, a common plant and opportunistic human pathogen, was identified from several samples including Irrigation water2, Vegetables1, and Manure1. The genus Brevundimonas, commonly found in the environment, was also identified from different types of samples. One isolate was identified as Brevundimonas diminuta from sample Manure1, another isolate was identified as Brevundimonas vesicularis from sample Irrigation water3. Identified genera regarding each sample are shown in Table 3.
Nonetheless, as it can be seen in our results, the identification score of environmental isolates is lower compared to studies involving clinical isolates. Low identification rate can be explained by several factors. As Bruker's database is mostly made for clinically relevant microbes, environmental isolates regarding food safety and quality are underrepresented in it. This finding is similar to previously reported ones (De Koster and Brul, 2016; Strejcek et al., 2018). Moreover, several species of genus Bacillus, the most abundantly occurring Gram-positive genus in our study, such as Bacillus drentensi, Bacillus pumilus, and Bacillus thuringiensis have either been reported as missing from the database, misidentified, or identified with low confidence by other authors (Ashfaq et al., 2022). Another factor that contributes to poor identification result is that the database contains an inadequate number of reference spectra for a given species. As remarked by Edouard et al. (2012) and Kopcakova et al. (2014), who used MALDI-TOF MS to identify environmental isolates of Propionibacterium spp., a database expansion with environmental isolates would be vital.
4 Conclusions
The application of MALDI-TOF MS is gaining space for identification of food- and waterborne pathogens in the complex food production chain due to its faster and inexpensive identification process compared to traditional methods. Our analysis showed that Bruker Biotyper is able to identify causative agents of foodborne illnesses and food spoilage as well as common plant pathogens, however, its library still has room for improvement. Our research also revealed that the applicability of MALD-TOF MS in environmental bacteriology is limited due to the fact that the software failed to identify one third of the isolates even at genus level. However, according to identifications based on the library of Biotyper, Hungarian irrigation waters can be considered as microbiologically safe as no serious human pathogens were detected.
Funding
This work was supported by the EFOP - The Project is supported by the European Union and co-financed by the European Social Fund (grant agreement no. EFOP-3.6.3-VEKOP-16-2017-00005).
Acknowledgements
The authors acknowledge the Doctoral School of Food Science of the Hungarian University of Agriculture and Life Sciences for supporting this research.
References
Alnakip, M.E.A., Rhouma, N.R., Abd-Elfatah, E.N., Quintela-Baluja, M., Böhme, K., Fernández-No, I., Bayoumi, M.A., Abdelhafez, M.M., Taboada-Rodríguez, A., Calo-Mata, P., and Barros-Velázquez, J. (2020). Discrimination of major and minor streptococci incriminated in bovine mastitis by MALDI-TOF MS fingerprinting and 16S rRNA gene sequencing. Research in Veterinary Science, 132: 426–438. https://doi.org/10.1016/j.rvsc.2020.07.027.
Ashfaq, M.Y., Da'na, D.A., and Al-Ghouti, M.A. (2022). Application of MALDI-TOF MS for identification of environmental bacteria: a review. Journal of Environmental Management, 305: 114359. https://doi.org/10.1016/j.jenvman.2021.114359.
Avanzi, I.R., Gracioso, L.H., dos Passos Galluzzi Baltazar, M., Karolski, B., Perpetuo, E.A., and Nascimento, C.A.O. (2017). Rapid bacteria identification from environmental mining samples using MALDI-TOF MS analysis. Environmental Science and Pollution Research, 24: 3717–3726. https://doi.org/10.1007/s11356-016-8125-8.
Böhme, K., Fernández‐No, I.C., Barros‐Velázquez, J., Gallardo, J.M., Cañas, B., and Calo‐Mata, P. (2011). Rapid species identification of seafood spoilage and pathogenic Gram‐positive bacteria by MALDI‐TOF mass fingerprinting. Electrophoresis, 32(21): 2951–2965. https://doi.org/10.1002/elps.201100217.
Chung, Y., Han, M., and Kim, J.S. (2021). Comparative evaluation of Bruker Biotyper and ASTA MicroIDSys for species identification in a clinical microbiology laboratory. Diagnostics, 11(9): 1683. https://doi.org/10.3390/diagnostics11091683.
De Koster, C.G. and Brul, S. (2016). MALDI-TOF MS identification and tracking of food spoilers and food-borne pathogens. Current Opinion in Food Science, 10: 76–84. https://doi.org/10.1016/j.cofs.2016.11.004.
Edouard, S., Couderc, C., Raoult, D., and Fournier, P.E. (2012). Mass spectrometric identification of Propionibacterium isolates requires database enrichment. Advances in Applied Microbiology, 2(4): 497–504. http://dx.doi.org/10.4236/aim.2012.24063.
Falardeau, J., Johnson, R.P., Pagotto, F., and Wang, S. (2017). Occurrence, characterization, and potential predictors of verotoxigenic Escherichia coli, Listeria monocytogenes, and Salmonella in surface water used for produce irrigation in the Lower Mainland of British Columbia, Canada. PloS One, 12(9): e0185437. https://doi.org/10.1371/journal.pone.0185437.
Horváth, B., Peles, F., Szél, A., Sipos, R., Erős, Á., Albert, E., and Micsinai, A. (2020). Molecular typing of foodborne coagulase-positive Staphylococcus isolates identified by MALDI-TOF MS. Acta Alimentaria, 49(3): 307–313. https://doi.org/10.1556/066.2020.49.3.9.
Kampmeier, S., Berger, M., Mellmann, A., Karch, H., and Berger, P. (2018). The 2011 German enterohemorrhagic Escherichia coli O104:H4 outbreak — the danger is still out there. In: Frankel, G. and Ron, E. (Eds.), Escherichia coli, a versatile pathogen. Current topics in microbiology and immunology, Vol. 416, pp. 117–148. https://doi.org/10.1007/82_2018_107.
Kopcakova, A., Stramova, Z., Kvasnova, S., Godany, A., Perhacova, Z., and Pristas, P. (2014). Need for database extension for reliable identification of bacteria from extreme environments using MALDI TOF mass spectrometry. Chemical Papers ,68(11): 1435–1442. https://doi.org/10.2478/s11696-014-0612-0.
Ponderand, L., Pavese, P., Maubon, D., Giraudon, E., Girard, T., Landelle, C., Maurin, M., and Caspar, Y. (2020). Evaluation of Rapid Sepsityper® protocol and specific MBT-Sepsityper module (Bruker Daltonics) for the rapid diagnosis of bacteremia and fungemia by MALDI-TOF-MS. Annals of Clinical Microbiology and Antimicrobials, 19: 60. https://doi.org/10.1186/s12941-020-00403-w.
Scallan, E., Hoekstra, R.M., Angulo, F.J., Tauxe, R.V., Widdowson, M.A., Roy, S.L., Jones, J.L., and Griffin, P.M. (2011). Foodborne illness acquired in the United States—major pathogens. Emerging infectious diseases, 17(1): 7–15. https://doi.org/10.3201/eid1701.P11101.
Singhal, N., Kumar, M., Kanaujia, P.K., and Virdi, J.S. (2015). MALDI-TOF mass spectrometry: an emerging technology for microbial identification and diagnosis. Frontiers in Microbiology, 6: 791. https://doi.org/10.3389/fmicb.2015.00791.
Strejcek, M., Smrhova, T., Junkova, P., and Uhlik, O. (2018). Whole-cell MALDI-TOF MS versus 16S rRNA gene analysis for identification and dereplication of recurrent bacterial isolates. Frontiers in Microbiology. 9: 1294. https://doi.org/10.3389/fmicb.2018.01294.
Turner, K., Moua, C.N., Hajmeer, M., Barnes, A., and Needham, M. (2019). Overview of leafy greens–related food safety incidents with a California link: 1996 to 2016. Journal of Food Protection, 82(3): 405–414. https://doi.org/10.4315/0362-028X.JFP-18-316.
van Veen, S.Q., Claas, E.C., and Kuijper, E.J. (2010). High-throughput identification of bacteria and yeast by matrix-assisted laser desorption ionization-time of flight mass spectrometry in conventional medical microbiology laboratories. Journal of Clinical Microbiology, 48(3): 900–907. https://doi.org/10.1128/JCM.02071-09.
