Use of Matrix-Assisted and Laser Desorption/Ionization Time-of-Flight Technology in the Identification of Aeromonas Strains Isolated from Retail Sushi and Sashimi

The genus Aeromonas includes well-known pathogenic species for fishes and humans that are widely distributed in the aquatic environment and foods. Nowadays, one of the main issues related to wild Aeromonas isolates is their identification at the species level, which is challenging using classical microbiological and biomolecular methods. This study aims to test MALDI-TOF MS technology in the identification of Aeromonas strains isolated from n. 60 retail sushi and sashimi boxes using an implemented version of the SARAMIS software V4.12. A total of 43 certified Aeromonas strains were used to implement the SARAMIS database by importing the spectra obtained from their identification. The original SARAMIS version (V4.12) failed to recognize 62.79% of the certified strains, while the herein-implemented version (V4.12plus) allowed the identification of all the certified strains at least to the genus level with a match of no less than 85%. Regarding the sushi and sashimi samples, Aeromonas spp. was detected in n. 18 (30%) boxes. A total of 127 colonies were identified at the species level, with A. salmonicida detected as the most prevalent species, followed by A. bestiarum and A. caviae. Based on the results of the present study, we could speculate that MALDI-TOF technology could be a useful tool both for the food industry to monitor product contamination and for clinical purposes to make diagnoses effectively and quickly.


Introduction
The genus Aeromonas includes a group of 36 Gram-negative bacteria species widely distributed in the aquatic environment, particularly in rivers, lakes, and sewage [1].Epidemiological evidence shows how water acts as a vehicle for Aeromonas spread and is responsible for the wide range of organisms and foods in which it can be isolated [2].
Aeromonas is, in fact, well known as a pathogen of several freshwater and marine animals, causing even severe diseases, such as the "red leg" (septicemic disease) caused by A. hydrophila in frogs or "furunculosis" caused by A. salmonicida in salmonids, responsible for major economic losses to the aquaculture sector [3].
Due to its wide distribution, humans can be easily exposed to Aeromonas through various routes of infection, resulting in even severe cases of gastroenteritis, primary and secondary septicemia in immunocompromised people, and severe wound infections in healthy people [4,[11][12][13].Epidemiological evidence suggests that children are the most Pathogens 2024, 13, 432 2 of 15 susceptible population in which co-infections by Salmonella and Campylobacter are frequently observed (polymicrobial infection) [14].Out of 36 Aeromonas species, 19 are considered pathogenic to humans; however, most of the infections are attributed to only 4 species: A. caviae, A. dhakensis, A. veronii biovar sobria, and A. hydrophila [15,16].
The foodborne outbreaks related to this bacterium are mainly related to the consumption of contaminated water and food, such as freshwater fish and shellfish, meats, and fresh vegetables [6][7][8][9][10][11][12].Among these, ready-to-eat food such as sushi and sashimi could represent a relevant source of infection considering the use of raw fish (in sushi and sashimi) and often also raw vegetables (in sushi) in their preparation.Sushi and sashimi are traditional Japanese dishes whose consumption has increased annually worldwide, posing concerns about the magnitude of the global exposure.Even though Aeromonas represents a relevant risk for these types of products, only a little evidence of its prevalence in sushi and sashimi is available, and further studies are certainly desirable [6].
Despite its wide environmental distribution, the consequent diffusion in the food, and the pathogenicity in humans, Aeromonas is still recognized as a potential or emerging foodborne pathogen since only a few cases of human infection are documented.Several authors have hypothesized that the low number of infection reports could be related to the lack of routine tests, the difficulty in strain identification, and the frequent occurrence of polymicrobial infection [14].
In this regard, rapid and efficient identification of the Aeromonas isolates at the species level would allow better risk management under different perspectives.The limited accuracy of the available methods in the identification of the Aeromonas strains is related to the high phenotypic and genotypic similarity that some species show between them [17].Traditional techniques used for its identification are based both on the use of culture media that allow a morphological evaluation of the isolates as well as biochemical tests used for their physiological characterization.These last methods have the disadvantage of generally being time-consuming and not very accurate [18].Over the years, various molecular methods have been developed, such as DNA-DNA hybridization and housekeeping genes; although these methods are known to be highly accurate in the identification of Aeromonas, they are rarely used in routine analysis due to the high cost, long time required for analysis, and the need for highly trained staff [1].Furthermore, although 16S rRNA gene sequencing is considered one of the most popular tools for bacterial species identification, due to the high sequence conservation of the 16S rRNA gene (97.8-100%) in the Aeromonas genus and the existence of several copies of the gene with intragenomic heterogeneity in some strains, the usefulness of the 16S rRNA gene for taxonomic analysis at the species level is limited [19].
A valid alternative to the conventional identification methods just described is the mass spectrometry (MS) technique based on MALDI-TOF (matrix-assisted and laser desorption/ionization time-of-flight) technology [20].MALDI-TOF MS started to be routinely utilized as a first-line identification method in microbiology laboratories in the last 12-15 years [21,22].This technology offers many advantages over conventional microbiological and molecular techniques, which include reliability and rapidness, as it takes only a few minutes to identify microbes; simplicity; cost-effectiveness; and no highly trained staff are needed [23,24].MALDI-TOF technology is based on the detection of the mass-to-charge ratio of specific targets, the ribosomal proteins of the bacteria, which provide a unique mass spectrum of the microorganism within a short time [23].To compare the mass spectra of unknown bacteria with the reference mass spectra, there are several commercial databases available, such as SARAMIS software that was used in the present study.The identification accuracy via MALDI-TOF depends upon the database, which, in general, at the species level, is above 90% [25].Therefore, an effective way to improve the accuracy of MALDI-TOF instrument identification is to expand and update the database [21].As reported by several studies on the identification of bacteria of the Aeromonas genus, MALDI-TOF technology correctly identifies isolates at the genus level but shows variable reliability in the identification at the species level depending on the species examined [3,22,26].
Against this background, the present study aims to test MALDI-TOF MS technology in the identification of Aeromonas strains isolated from retail sushi and sashimi boxes using an implemented version of the SARAMIS software.

Experimental Plan
The experimental plan of the present study consisted of three different steps: 1. Evaluation of SARAMIS software-V4.12 in Aeromonas identification; 2. Implementation of SARAMIS Database for Aeromonas identification and evaluation of its efficiency; 3. Evaluation of Aeromonas spp. in retail sushi and sashimi boxes.
The steps are detailed below.

Preparation of the Bacterial Colonies for Step 1 and Step 2
Table 1 reports the n.43 certified strains of different species belonging to the genus Aeromonas used in steps 1 and 2, as described below.In detail, n. 29 (67.44%)certified strains were isolated from human clinical specimens, while n. 14 (32.56%)strains were isolated from environmental, food, or animal sources.The strains were kept frozen in Brain Heart Infusion Broth (Biolife, Milan, Italy) + 15% glycerol (Sigma-Aldrich, St. Louis, MO, USA) at − 80 • C at the microbial collection of the "Food Microbiology Laboratory" of the Department of Veterinary Sciences, University of Messina (Messina, Italy).

Evaluation of SARAMIS Software-V4.12 in Aeromonas Identification
The available version of the SARAMIS Knowledge Base software, V4.12, was used.n. 15 "Reference-Spectra" and n. 18 "Super-Spectra" of Aeromonas were present in the database (Table 2).According to the SARAMIS manual [27], the following explanation of "Reference-Spectra" and "Super-Spectrum" are given: in Reference Spectra (or Consensus Spectra), only those mass signals with high frequency in a batch of spectra are recorded; thus, if the frequency threshold is set to 70%, only those mass signals recorded in at least 70% of the compared spectra are selected.The differences among spectra are due to the variability between isolates, the variability of mass spectra of single isolates, and analytical deviations.In Reference Spectra, these differences are eliminated, and only a reduced number of conserved mass signals is retained.The remaining masses are thus typically recorded in mass spectra of the particular taxon, irrespective of the isolates' origin and cultivation conditions.In Reference Spectra, mass signals' relative intensities are averaged.The Super Spectrum is created from a Reference Spectrum by assigning peak weights to each mass signal.Peak weights are generally higher for species-specific mass signals and lower for mass signals that are specific only at higher taxonomic levels, such as genera or families.Since the latter mass signals have no pertinence for species identification, the peak weights are set low, or the mass signals are even ignored by setting the peak weight to zero.Super Spectra are therefore highly specific artificial mass spectra that allow the unambiguous identification of an unknown isolate when its mass spectrum shows the specific mass pattern of a particular Super Spectrum.
To carry out a preliminary evaluation of the robustness and efficiency of SARAMIS software V4.12 in the identification of strains belonging to the Aeromonas genus, all the n.43 certified strains reported in Table 1 were tested following the procedure reported below.
Once thawed, a total of 20 µL of each strain from the frozen stock was inoculated into 10 mL of Tryptone Soy Broth (TSB; Biolife, Milan, Italy) and incubated overnight at 30 • C. The broth cultures were then plated using a 10 µL loop onto Tryptone Soy Agar (TSA) (Biolife, Milan, Italy), TSA + 5% defibrinated mutton blood (TSAS; Biolife, Milan, Italy) (medium reported as optimal for identification), and incubated at 30 • C for 24 h to be tested with VITEK MS (bioMérieux Italia, Florence, Italy), automated device at our disposal for analyses based on MALDI-TOF MS technology.
At the same time, Escherichia coli ATCC 8739 (reference strain for calibration of the VITEK MS device) was prepared on TSAS and incubated at 37 • C for 24 h.The identification analyses were performed using disposable plates VITEK MS-DS with 48 spots (bioMérieux Italia, Florence, Italy) using the G3 and G4 as calibration spots for E. coli ATCC 8739.
Isolated colonies from each certified strain were picked using a sterile 1 µL loop from each medium plate, taking care not to unintentionally withdraw agar.Each picked colony was smeared in the center of a spot, and then 1 µL of alpha-cyano-4-hydroxycinnamic acid (CHCA) MALDI matrices (bioMérieux Italia, Florence, Italy) was added.The matrix/microorganism suspension was allowed to dry and crystallize completely and then placed in the acquisition station and loaded into the VITEK-MS.Crystallization was determined by visually checking the formation of crystals as a yellowish film on the spot surface.The VITEK-MS was used with the following settings: positive linear mode, laser frequency of 50 Hz, acceleration voltage of 20 kV, and extraction delay time of 200 ns.The mass spectra range was set to detect from 2000 to 20,000 Da.A unique mass spectrum was generated for each tested colony, which was transferred into SARAMIS software and compared to the database of Reference Spectra and Super Spectra.For each strain, the spectra of 3 colonies grown in each culture medium were acquired.
For the interpretation of the results returned by the SARAMIS software, confidence levels were established based on the match percentage between the spectra of the colonies tested and the spectra in the database: a match ≥99.9% was considered as an "excellent identification", a match ranging between 60% and 99.8% was considered as a "good identification", while a match <60% was interpreted as "no identification" [24].

Implementation of SARAMIS Database for Aeromonas Identification and Evaluation of Its Efficiency
For creation of new Super Spectra or implementation of existing ones, the overnight TSB culture of each of the n.43 certified strain was plated using a 10 µL loop in four different growth media, including two selective media for growth of Aeromonas spp.: (i) Aeromonas starch DNA agar base (Biolife, Milan, Italy) (AEStarch); (ii) Pseudomonas Aeromonas Selective Agar Base acc. to KIELWEIN (GSP) (Merck, Darmstadt, Germany) supplemented with sodium penicillin G (10 mg/L; IGN Biomedicals, OH, USA); and two nonselective media: (iii) TSA; (iv) TSAS.All plates were incubated at 30 • C for 24 h.At the same time, the calibration strain E. coli ATCC 8739 was prepared on TSAS and incubated at 37 • C for 24 h.
For each Aeromonas certified strain, n. 4 colonies from each of the four media (a total of 16 colonies) were collected and processed for spectra acquisition, as described above (see Section 2.3) in order to increase the variability and the number of spectra to compare (16 for each strain) so that all possible mass variants resulting from both intra-species genetic differences and different culture conditions could be included in the spectra.
Through the data loading system "Target manager" in SARAMIS software, for each processed colony, the "genus", "species", and "type" corresponding to the identification number of the certified strain, "medium of origin", and "incubation temperature" were reported.These inserted data automatically determined the creation within SARAMIS software of the respective "species" and "type" folders within the "Aeromonas" genus folder.The acquired spectra were subjected to an initial selection maintaining only those that fell within a range of 100 to 200 peaks, eliminating those with a number of peaks less than 100 and greater than 200 according to SARAMIS manual.Then, a dendrogram of the selected spectra was created using a specific tool in SARAMIS software.Only spectra with a similarity of 70% between duplicates and 65% between species were selected, while those with lower similarity were eliminated.Selected spectra were further compared to each other to determine the most frequently occurring mass signals.Only spectra with 60% of the masses in common were maintained, while all the other strains below the threshold were eliminated.The remaining spectra with a total of 40 mass signals in common represented the new Reference Spectra that were loaded into the SARAMIS database.To create Super Spectra from Reference Spectra, it was necessary to determine mass signals specific to the selected species by distinguishing conserved masses that were not at a higher taxonomic level (e.g., genus).In this regard, a comparison of the mass signals between the implemented Reference Spectra and the spectra of Aeromonas genus was performed, and the masses in common were excluded in order to create new Super Spectra of the considered species maintaining only characteristic mass signals.Finally, the specificity of the mass signals was evaluated by running a comparison of the newly created Super Spectra against the entire database since individual mass signals can also occur in spectra of different taxa.The new Super Spectra met two basic conditions: (i) the sum of peak weights was not more than 1400 points; (ii) the sum of peak weights of the mass signals in common with other taxa did not exceed 600 points.
The 43 certified strains already evaluated with the original version of SARAMIS software (see Section 2.3) were analyzed once again using the newly implemented version (V4.12 plus ).
2.5.Evaluation of Aeromonas spp. in Retail Sushi and Sashimi Boxes 2.5.1.Sampling of Retail Sushi and Sashimi Boxes A total of 60 retail sushi and sashimi boxes, including 5 different formulations, were purchased from different retailers in Sicily (southern Italy).Each box represented a sample and, in detail, the following were collected: 10 sushi-nigiri boxes (ingredients: raw salmon, rice, rice vinegar, and soy sauce), 10 sushi-hosomaki boxes (ingredients: nori seaweed, rice, rice vinegar, soy sauce, and salmon), 10 sushi-uramaki boxes (ingredients: sesame seeds, rice, rice vinegar, nori seaweed, soy sauce, salmon, and avocado), 15 salmon sashimi boxes (ingredients: raw salmon and soy sauce), and 15 tuna sashimi boxes (ingredients: raw tuna and soy sauce).The boxes contained 6 pieces each of the corresponding formulation.
Half of the nigiri, hosomaki, and uramaki boxes were prepared at the time of sale directly in-store, while the other half were sold already packaged and prepared at an industrial level.
At the time of sampling, boxes were stored in refrigeration regime and, once purchased, were transported inside coolers to the "Food Microbiology Laboratory" of the Department of Veterinary Sciences, University of Messina (Messina, Italy), and immediately analyzed as follows.

Microbiological Analysis: Aeromonas Detection and Enumeration
The protocol adopted for the detection and enumeration of presumptive Aeromonas spp. was inspired by Lee et al. [7].In detail, the sushi and sashimi pieces of each box were homogenized, and a representative sample of 10 g was aseptically put into a stomacher bag, diluted in a ratio of 1:9 w/v with sterile peptone water (Biolife, Milan, Italy), and homogenized through a stomacher (400 Circulator; International PBI s.p.a., Milano, Italy) for 60 s at 230 rpm.The homogenate was decimal diluted and plated onto GSP agar and incubated at 30 • C for 24-48 h.The typical yellow colonies of Aeromonas spp.were visually enumerated.
A maximum of 10 colonies per sample were collected using a sterile loop, streaked onto TSA, and incubated at 30 • C for 24-48 h.The isolations obtained were then identified by MALDI-TOF MS using the same settings reported in Section 2.4.through our implemented version of SARAMIS software (V4.12 plus ).

Data Analysis
The data acquired were presented as mean ± standard deviation or parts of the whole as percentages.
The normal distribution of the data of the bacterial loads detected in the sushi and sashimi samples was tested using the D'Agostino-Pearson omnibus test, and any significant differences between the different sushi samples were tested using ordinary one-way analysis of variance (ANOVA) or Welch's t-test.
The critical level of significance (p) was set at 5% (0.05), and the test was performed two-tailed using Graph Pad Prism 9.1.1software (Graph Pad Prism, San Diego, CA, USA).

Preliminary Assessment of SARAMIS V4.12 Performance in Aeromonas Identification
The results of the identifications obtained by SARAMIS software V 4.12 are shown in Table 3.Out of n. 43 strains tested, n. 27 (62.79%)were not even identified at the genus level, whereas the remaining n. 16 strains were identified as follows: n. 13 (30.23%)as Aeromonas sp., with an identification percentage ranging from 71.7% to 92.5%; n. 2 (4.65%) strains of A. sobria were identified, with an identification percentage ranging between 77.1% and 89.3%; while n. 1 strain of A. veronii (CECT4258) was misidentified both as A. sobria and Aeromonas sp.n.i.= no identification; TSA = Tryptic Soy Agar; TSAS = TSA + 5% defibrinated mutton blood.

Evaluation of the SARAMIS Version Implemented in Aeromonas Identification
The identification results of the Aeromonas strains processed using the implemented version of SARAMIS software (V4.12 plus ) proposed herein are shown in Table 4.

Table 4.
Comparison between the identification of different Aeromonas strains using SARAMIS software V 4.12 and an implemented version of the software (V4.12 plus ) proposed in the present study.The best percentages of strain identifications performed with both versions of the software are reported regardless of the type of growth medium used.

Discussions
Preliminary assessment of the SARAMIS V4.12 showed the poor efficiency of the software in the identification of different strains of the genus Aeromonas, possibly related to the few number of Reference Spectra and Super Spectra available in the database.Out of 43 strains tested, only the 2 strains (4.65%) of A. sobria were correctly identified at the species level (with a match ranging between 77.1% and 89.3%), n. 1 was misidentified as A. sobria instead of A. veronii, n. 13 (30.23%)were identified only at the genus level, and the remaining n. 27 (62.79%)were not identified.
In this regard, consulting the data in the literature, SARAMIS V4.12 showed less accuracy compared to other more updated MALDI-TOF MS databases.A comparison to identify clinical isolates of Aeromonas spp.documented by Kitagawa et al. between MALDI-Biotyper (Brucker Daltonics, Bremen, Germany) Flex Control software ver.3.4 and VITEK MS and its analysis software VITEK MS ver.KB3.2 showed, for MALDI-Biotyper, correct identification at the genus level for all isolates.In contrast, the identification at the species level presented an important species-dependent variability, with an identification match of 0%, 0%, 72.3%, and 78.9% for A. jandei, A. dhakensis, A. veronii, and A. caviae, respectively, with the best result for A. hydrophila (93.3%).Regarding VITEK MS, similarly, all isolates were identified correctly at the genus level, but only A. hydrophila and A. caviae were accurately identified at the species level.Isolates of all other species (A.jandei, A. dhakensis, and A. veronii) were detected as more than one species [1].
Porte et al. evaluated the use of two MALDI-TOF MS platforms (Microflex LT, from Bruker Daltonics; and Vitek MS, from BioMerieux) for the microbiological diagnosis of several microorganisms in a routine laboratory in Chile; the Aeromonas genus identification match at species level proved to be low, concluding that this technique has good precision in genus accuracy but with limited species differentiation [28].
In another clinical study, MALDI-TOF MS with BIOTYPER 2.0 (Bruker Daltonics, Bremen, Germany) correctly identified a high number of reference strains at the species level: A. hydrophila (35/35) Based on the identification results obtained using the implemented version of SARAMIS software (V4.12 plus ) proposed herein, we could speculate that, with an appropriate upgrade, MALDI-TOF technology based on VITEK MS and SARAMIS software can represent a rapid and relatively les-expensive method for the identification of bacterial colonies morphologically ascribable to Aeromonas spp.Although 16S rRNA gene sequencing and DNA-DNA hybridization, which is considered the gold standard method, allow for the identification of a greater number of species, the results of this study suggest that MALDI-TOF MS may be a valid, first-line identification system, reserving the use of molecular methods for those cases where the instrument fails to recognize isolates at the species level [30].This would undoubtedly be a benefit in those routine situations where rapid identification is required.In this regard, there may be many fields of application for the instrument.
Contaminated foods represent one of the main sources of Aeromonas infection for humans, and several studies in the literature document an increasing occurrence of the bacterium in ready-to-eat foods, particularly in seafood, such as sushi and sashimi [4,[6][7][8]10].From this perspective, we investigated the occurrence of Aeromonas in retail sushi and sashimi boxes, detecting a prevalence of 30%.There is a wide variability in the prevalence of Aeromonas in retail sushi and sashimi reported in the literature.
The prevalence observed herein was relatively lower than that reported by Lee et al. in a similar study conducted on 30 retail sushi boxes purchased in Norwegian markets in which 18 Aeromonas strains, all identified as A. media, were isolated in 17% of the samples analyzed, with average loads of 1.5 Log CFU/g [7].A higher prevalence was instead detected in another study conducted on 21 retail sushi products purchased in Italy, which isolated Aeromonas spp. in 90.5% of the samples [31].Hoel et al. also reported a higher prevalence (71%) of Aeromonas spp. in nigiri and maki sushi compared to that observed herein, detecting an average load of 3 ± 0.3 Log CFU/g [32].The similar load of 2.93 ± 0.89 Log CFU/g detected in the present study poses concerns considering the A. hydrophila concentration of 2.3-3 Log CFU/g estimated in a cold dish identified as the source of a Chinese outbreak affecting 349 people in 2012 [33].However, using SARAMIS version V4.12 plus , MALDI-TOF technology did not identify any A. hydrophila among the isolates, which instead were A. salmonicida (67.57%), A. bestiarum (16.22%), and A. caviae (2.03%).A. salmonicida is the only psychrophilic species among those isolated, and this could explain the higher prevalence since the retail sushi was kept at a refrigeration temperature at the time of sampling.Interestingly, there is evidence that 25 • C represents the temperature limit beyond which A. salmonicida cannot survive [14].This contrasts with the results obtained in the present study, as the protocols adopted herein for the isolation of Aeromonas spp.involved an incubation temperature of the growth media of 30 • C.However, recent studies have proposed a new classification for A. salmonicida distinguishing between the subspecies that do not grow at temperatures above 37 • C (A. salmonicida subsp.salmonicida, smithia, achromogenes and masoucida) and the A. salmonicda subsp.pectynolityca that grows well even at 37 • C [14].Against this background, we could speculate that A. salmonicda subsp.pectynolityca could be the one isolated in the present study.
The same species were identified through PCR analysis in retail sushi by Hoel et al., who isolated 118 Aeromonas strains using starch ampicillin agar incubated at 37 • C, identifying A. salmonicida as the most prevalent species (74%), followed by A. bestiarum (9%) and A. caviae (5%) [34].Hoel et al. considered all isolates as potentially pathogenic due to the high prevalence of genes encoding different pathogenic factors (hemolysin, aerolysin, cytotoxic enterotoxin, and heat-labile and heat-stable cytotonic enterotoxin) posing concerns for the findings of the present study [34].Among the isolated species, only A. caviae is part of a subset of species most implicated in human infections, together with A. hydrophila, A. dhakensis, and Aeromonas veronii biovar sobria [6].The species A. salmonicida and A. bestiarum are basically considered primary pathogens of fishes, especially in salmonid culture systems and freshwater fishes, respectively [35,36]; however, several recent reports have highlighted the zoonotic potential of A. salmonicida [35].
Despite the significant evidence of the pathogenic effects on aquatic organisms and humans, the role of Aeromonas as a true foodborne pathogen continues to be questioned.Overall, the occurrence of Aeromonas in food has been mostly related to contamination with contaminated water [37].The common physical-chemical treatments used for the purification of drinking water in a food plant are normally capable of devitalizing Aeromonas; however, ineffective water supply management could be responsible for its contamination.Furthermore, the preparation of sushi involves numerous manipulations and, therefore, the routes of exposure to contaminated water can be several and related to improper processing or hygiene practices.A further significant source of Aeromonas for sushi and sashimi is represented by fish, as intra-vitam contamination from the aquatic environment is certainly possible [38].Interestingly, in the present study, Aeromonas was only detected in sashimi (raw fish) samples, while it was not detected in the more elaborate sushi, such as nigiri, hosomaki, and uramaki, in which even other ingredients (e.g., vegetables) could act as sources of contamination.This result could be related to the effect of the acidification of the rice that, in combination with low storage temperatures, supports Aeromonas growth [39].It should also be considered that, although most of the Aeromonas species pathogenic for humans are mesophilic, such as A. caviae herein detected, they still manage to replicate at low temperatures, such as those at which retail sushi is kept on the market.Against this background, the choice of raw materials of high quality and the maintenance of the cold chain, as well as a proper application of good manufacturing and hygiene practices, are crucial for producing safe and quality sushi and sashimi.
As previously stated, there are only a few reports of human Aeromonas outbreaks associated with food consumption, and detailed information regarding infection doses is lacking [6].Data collected during Aeromonas outbreaks that occurred in Norway and Sweden suggested an infection dose ranging between 10 6 and 10 8 cells, which is much higher than the average load (2.93 ± 0.89 Log) detected herein.However, other studies reported lower infective doses of 10 3 to 10 4 cells [40].The current knowledge from clinical studies and recent outbreaks suggests that, depending on host-microbe interactions, even exposure to low and moderate doses of pathogenic Aeromonas can lead to infections, and the prevalence is probably underestimated due to many cases remaining undiagnosed [6].Further concerns arise from the growing detection of multi-resistant Aeromonas strains [41] for which, in addition to finding effective contrast strategies, MALDI-TOF technology could be proposed as a valid and innovative support in the identification of resistant strains through a proteomic approach [42][43][44].
Against this background, the use of MALDI-TOF technology could be useful both for the food industry to monitor product contamination and for clinical purposes to make diagnoses effectively and quickly.

Conclusions
The results of the present study highlighted how MALDI-TOF technology can be a suitable tool for the rapid and effective identification of wild strains of Aeromonas isolated from food matrices.Furthermore, the detection of strains pathogenic to humans in sushi and sashimi samples available on the market stresses the need not to underestimate Aeromonas risk.Although the investigations were conducted only on Aeromonas strains, the results obtained and the simplicity of the analytical approach allow us to speculate on the use of MALDI-TOF for the routine identification of other pathogenic bacteria, not only in the food sector.

Table 1 .
List of the certified Aeromonas strains used in the present study to implement SARAMIS software version V4.12.

Table 2 .
Reference Spectra and Super Spectra list of Aeromonas species included in SARAMIS Database V4.12.

Table 3 .
Identification results of different Aeromonas strains processed using MALDI-TOF technology with identification software SARAMIS, version 4.12.

Table 5 .
MALDI-TOF MS identifications of 148 presumptive Aeromonas colonies isolated from salmon and tuna sashimi sampled in different supermarkets and restaurants in Messina (Southern Italy).