Occurrence of Campylobacter in Faeces, Livers and Carcasses of Wild Boars Hunted in Tuscany (Italy) and Evaluation of MALDI-TOF MS for the Identification of Campylobacter Species

A total of 193 wild boars hunted in Tuscany, an Italian region with a high presence of wild ungulates, were examined to assess the occurrence of Campylobacter species in faeces, bile, liver and carcasses, with the aim of clarifying their contribution to human infection through the food chain. Campylobacter spp. were found in 44.56% of the animals, 42.62% of the faecal samples, 18.18% of the carcass samples, 4.81% of the liver tissues and 1.97% of the bile samples. The Campylobacter species genotypically identified were C. coli, C. lanienae, C. jejuni and C. hyointestinalis. The prevalent species transpired to be C. coli and C. lanienae, which were isolated from all the matrices; C. jejuni was present in faeces and liver, while C. hyointestinalis only in faeces. Identification was carried out by matrix-assisted laser desorption/ionisation–time-of-flight mass spectrometry (MALDI-TOF MS) on 66 out of 100 isolates identified genotypically, and the technique yielded unsatisfactory results in the case of C. lanienae, which is responsible for sporadic human disease cases. The level of Campylobacter spp. contamination of meat and liver underlines the need to provide appropriate food safety information to hunters and consumers.


Introduction
Campylobacteriosis is a gastro-intestinal infection and the most reported zoonosis in the EU, with Campylobacter jejuni and C. coli as the main causative agents, and it represents a threat to public health worldwide [1][2][3]. It is usually self-limiting; however, in rare cases, and especially in immunocompromised individuals, it can lead to severe post-infection complications: arthropathies, immune-mediated neuropathy such as Guillain-Barré Syndrome, irritable bowel syndrome and septicaemia [4]. Although campylobacteriosis is mostly associated with the consumption of poultry meat and raw milk, Campylobacter spp. is also commonly found in the meat of other animal species [5].
The wild boar is currently the second most abundant ungulate in Europe, and it is considered a pest in many areas mainly because it damages crops, has a negative A total of 193 wild boars (111 females and 82 males; 87 adult, 30 sub-adult and 76 young animals) were sampled as part of a project focused on the isolation of pathogenic micro-organisms in hunted Tuscan wild boar, to study their role as a reservoir for human and animal diseases. The animals' ages were determined based on tooth eruption and lower-jaw tooth wear [25]. Sampling took place during the autumn-winter hunting season from 2018 to 2019, following the regional hunting legislation [26] in 4 Tuscan provinces (Pisa, Livorno, Siena and Grosseto). Sampling was integrated with the routine procedure of delivery of shot animals (by the evening of the hunting day at the latest) to the central collection point, where evisceration and skinning operations were performed.

Sample Collection
The samples were taken by expert microbiology laboratory personnel during the standard course of action when hunted prey was dressed. Faecal swabs were obtained directly from the rectum using an Amies charcoal swab (Copan Italia S.p.A., Brescia, Italy). During the evisceration phase, bile was withdrawn from the gall bladder with a sterile syringe, and a sample of liver (about 50 g) was also collected. After the skinning, a Whirl-Pak Speci-Sponge (Nasco, Madison, WI, USA) (pre-moistened on site with 10 mL of sterile buffered peptone water) was used to sample the carcass tissue; a total area of approximately 300 cm 2 for each half carcass from the belly/flank, medial thigh, and shoulder regions was sampled. Since the sampling was performed in collaboration with the hunters during regular hunting sessions, it was not always possible to collect all four types of samples from each animal examined. The total numbers of analysed samples were as follows: 183 faecal swabs, 187 liver tissue samples, 152 bile samples and 55 carcass sponges. All samples were transported to the laboratory under refrigerated conditions and analysed before 24 h had elapsed from the sampling time.

Sample Preparation and Campylobacter Enrichment
For all types of samples, enrichment was carried out in Bolton selective enrichment broth with the addition of the corresponding supplement (cefoperazone, vancomycin or trimethoprim, 20 mg/L in each case, and cycloheximide, 50 mg/L) and laked horse blood (5% v/v). For faecal samples, each swab was put in a tube with 9 mL of Bolton enrichment broth. For liver tissue, a swab was taken from a freshly cut surface exposed using a sterile scalpel and put in a tube with 9 mL of Bolton enrichment broth. Before the incision, the liver surface was sterilised by touching it with the red-hot blade of a scalpel to avoid external contamination [27]. For bile samples, 1 mL of sample was added to 9 mL of enrichment broth. Finally, for carcass samples, 70 mL of enrichment broth was added to each sponge bag, and the contents were mixed by squeezing the sponge repeatedly. All enrichment broths were incubated for 48 h at 42 ± 0.5 • C under microaerobic conditions using CampyGen sachets (5% O 2 , 10% CO 2 , 85% N 2 ). The same incubation conditions were applied for all subsequent phases. All culture media, supplements and microaerobic sachets in all phases were acquired from Oxoid (Basingstoke, UK).

Campylobacter Isolation and Pre-Identification
Campylobacter isolation and preliminary identification at the genus level was carried out following the ISO 10272-1:2017 method modified by adding a supplemental filtration phase of the enrichment broth after the incubation, as described by Pedonese et al. [28,29]. Briefly, 0.3 mL of the enrichment broth was seeded onto a sterile cellulose membrane filter (47 mm diameter, 0.45 µm pore size, Sigma-Aldrich, Milan, Italy) laid on the surface of a modified charcoal cefoperazone deoxycholate agar (mCCDA) plate, and the filtrate was spread onto the agar surface. After incubation, 1 or 2 characteristic colonies per sample were picked from positive plates. Isolates were preliminarily identified at the genus level by using several phenotypic tests: microscope observation of cell morphology and  [30], an oxidase test (oxidase strips, Oxoid), a catalase test (hydrogen peroxide 3% solution, Merck Life Sciences S.r.l., Milan, Italy), growth trials (at 42 • C in aerobiosis and at 25 • C in microaerobic conditions) and with the Oxoid Biochemical Identification System Campy (Oxoid), a rapid identification system based on the detection of L-alanyl aminopeptidase production. Isolates phenotypically confirmed as Campylobacter spp. and fully viable were stored at −80 • C in brain-heart infusion broth with 5% laked horse blood and 15% glycerol (Sigma-Aldrich) for further analyses.

Molecular Identification
The isolates from positive samples were confirmed using a multiplex PCR as described by Wang et al. [31] and a simplex PCR according to Di Giannatale et al. [32], as previously reported [33]. For these analyses, DNA was extracted using the Maxwell 16-tissue DNA purification kit (Promega Corporation, Madison, WI, USA) according to the manufacturer's instructions. In addition, isolates negative in PCR identification were subjected to 16S rRNA gene sequencing. Typing was performed by amplifying the 16S rRNA gene using the primers shown in Table 1 (Eurofins Genomics, Ebersberg, Germany), followed by sequencing of the PCR product. Amplification products were verified by gel electrophoresis. The products of the PCR were purified using ExoSAP-IT reagent (Applied Biosystems, GE Healthcare, Piscataway, NJ, USA) and sequenced using the BigDye Terminator v.3.1 Cycle sequencing kit (Applied Biosystems) according to the recommendations of the manufacturer. After sequencing, DNA was purified with ethanol precipitation using the Agencourt CleanSEQ ® kit (Beckman Coulter, Brea, CA, USA). Sequencing products were analysed with a Genetic Analyzer 3500 ® (Life Technologies, Renfrewshire, UK). Basic local alignment search tool analysis of the partial 16S rRNA gene sequence obtained was performed using the taxonomy browser of the National Center for Biotechnology Information (www.ncbi.nlm.nih.gov, accessed on 6 December 2019).

MALDI-TOF MS Identification
Sixty-six isolates were subjected to MALDI-TOF MS identification to verify the effectiveness of this technique in identifying Campylobacter species. The identification of bacterial isolates was preceded by preliminary extraction of proteins with ethanol and formic acid. For this purpose, a single colony of each bacterial culture was suspended in 150 µL of sterile deionized water, after which 450 µL of pure ethanol (Merck, Darmstadt, Germany) was added. Then, each sample was mixed thoroughly by vortexing. The resulting sample was then centrifuged for 5 min at 13,000 rpm. After that, the supernatant was discarded, 40 µL of 70% aqueous formic acid and then 40 µL of acetonitrile (Merck) were added to the precipitate, and the sample was thoroughly mixed by vortexing. After centrifugation (12,470× g), 1 µL of the obtained supernatant was pipetted onto a metal plate (Anchorchip 800 384, Bruker Daltonik, Bremen, Germany) and allowed to dry at room temperature. Then, 1 µL of cyano-4-hydroxycinnamic acid was added as the matrix solution (Bruker Daltonik), and the sample was left to dry at room temperature. The metal plate with the samples was subsequently placed in a MALDI chamber for analysis. An automatic measurement of the spectrum and a comparative analysis with reference spectra of bacteria were performed using the UltrafleXtreme mass spectrometer and MALDI-Biotyper 3.0 software (Bruker Daltonik). The analysis was performed in triplicate for each isolate. Mass spectra were recorded in active positive reflector mode within the 2000-20,000 mass-to-charge ratio (m/z) range using "complete sample" laser walk and 2000 shots per sample. The mass spectrometer automatically generated spectra of peaks corresponding to ions with different m/z, providing their number, intensity and peak correlation. Characteristic spectra for Campylobacter species generated by the Bruker UltrafleXtreme MALDI-TOF system were presented in Figures 1 and S1-S3. Before identification, raw spectra were smoothed, and baseline was corrected and then compared with reference microbial spectra using MALDI-Biotyper 3.0 software with a database containing microbial profiles. The probability of correct identification in MALDI Biotyper 3.0 is expressed as a score index, and the following values indicate the level of identification: 2.300-3.000 is highly probable identification of the micro-organism to species level; 2.000-2.299 is highly probable identification of the micro-organism to genus level and likely identification to species level; 1.700-1.999 is likely identification to genus level; and 0-1.699 is unreliable identification. Using molecular diagnosis as the gold standard, the sensitivity and specificity of MALDI-TOF MS species identification was determined. The Campylobacter species identification was considered valid if the cut-off score reached or exceeded 2.0, referring to the database provided by Bruker Daltonics. The data were processed with Excel (Microsoft Corporation, Redmond, WA, USA). automatic measurement of the spectrum and a comparative analysis with reference spectra of bacteria were performed using the UltrafleXtreme mass spectrometer and MALDI-Biotyper 3.0 software (Bruker Daltonik). The analysis was performed in triplicate for each isolate. Mass spectra were recorded in active positive reflector mode within the 2000-20,000 mass-to-charge ratio (m/z) range using "complete sample" laser walk and 2 000 shots per sample. The mass spectrometer automatically generated spectra of peaks corresponding to ions with different m/z, providing their number, intensity and peak correlation. Characteristic spectra for Campylobacter species generated by the Bruker UltrafleXtreme MALDI-TOF system were presented in Figures 1 and S1-S3. Before identification, raw spectra were smoothed, and baseline was corrected and then compared with reference microbial spectra using MALDI-Biotyper 3.0 software with a database containing microbial profiles. The probability of correct identification in MALDI Biotyper 3.0 is expressed as a score index, and the following values indicate the level of identification: 2.300-3.000 is highly probable identification of the micro-organism to species level; 2.000-2.299 is highly probable identification of the micro-organism to genus level and likely identification to species level; 1.700-1.999 is likely identification to genus level; and 0-1.699 is unreliable identification. Using molecular diagnosis as the gold standard, the sensitivity and specificity of MALDI-TOF MS species identification was determined. The Campylobacter species identification was considered valid if the cut-off score reached or exceeded 2.0, referring to the database provided by Bruker Daltonics. The data were processed with Excel (Microsoft Corporation, Redmond, WA, USA).

Results and Discussion
Wild animals may represent a reservoir and source of various foodborne infections for humans, including campylobacteriosis. This disease can be transmitted indirectly by contaminated water, vegetables or domestic animals, or directly through contact with contaminated carcasses or by consuming raw or undercooked meat and meat products [34]. Previous studies have shown a significantly higher occurrence of Campylobacter spp. in wild boars as compared to other wild animals [15]. In this study, 44.56% (86/193) of the tested animals were found to be positive for Campylobacter spp. (Table 2).

Results and Discussion
Wild animals may represent a reservoir and source of various foodborne infections for humans, including campylobacteriosis. This disease can be transmitted indirectly by contaminated water, vegetables or domestic animals, or directly through contact with contaminated carcasses or by consuming raw or undercooked meat and meat products [34]. Previous studies have shown a significantly higher occurrence of Campylobacter spp. in wild boars as compared to other wild animals [15]. In this study, 44.56% (86/193) of the tested animals were found to be positive for Campylobacter spp. (Table 2). A total of 100 isolates were genotypically identified at the species level. The presence of four Campylobacter species (C. coli, C. lanienae, C. jejuni and C. hyointestinalis) was registered by analysing wild boar faecal swabs, liver tissue, bile, and carcass samples (Table 3). Campylobacter spp. were found in 42.62% (78/183) of the faecal samples tested, of which 50% (39/78) were identified as C. coli, 41.03% (32/78) as C. lanienae, 5.13% (4/78) as C. hyointestinalis and 3.85% (3/78) as C. jejuni. Of the liver samples, 4.81% (9/187) were positive for Campylobacter spp.; C. coli, C. jejuni and C. lanienae were found evenly at 33.33% (3/9) of each species. In the tested bile samples, a prevalence of only 1.97% (3/152) was observed for Campylobacter, and the following species were confirmed: C. coli in 33.33% (1/3) and C. lanienae in 66.66% (2/3) of specimens. Regarding carcasses, 18.18% (10/55) of samples were positive for Campylobacter; C. coli and C. lanienae were found in 50% (5/10) of these contaminated carcasses.   Reviewing global studies, the Campylobacter rate of positivity and species diversity in wild boar varied significantly depending on the geographical areas where the animals came from and the type of sample tested. In Swedish wild boar, Wahlström et al. reported 10.6% Campylobacter positivity in faecal samples. In Switzerland, Campylobacter was not found in the faecal samples or the tonsils of 153 wild boar [11,12]. More recently, Tomino et al. reported 12.5% positive faecal samples among 248 wild boars hunted in Japan, with C. hyointestinalis as the most represented species; however, C. lanienae and C. coli were also confirmed in this study [35]. Generally, a high rate of Campylobacter prevalence in wild boar faecal samples was reported in Spain: 66% of 287 samples in southern central Spain, with prevalence ranging from 33% to 100% depending on the hunting region, and 38.9% of 126 samples in southern Spain, with most isolates being C. lanienae (n = 34) [13]. However, C. coli (n = 8) and C. jejuni (n = 2) were also confirmed [11]. In the same period, Navarro-Gonzalez et al. reported 12% of samples positive for Campylobacter bacteria presence out of 150 tested in north-eastern Spain, with a higher prevalence (24.6%) when sample enrichment was performed [12]. In this case, C. lanienae was also found as the most prevalent Campylobacter species. A very high prevalence (54.6%) of Campylobacter was reported in the faeces of wild boar in the Czech Republic, with C. coli as the predominant species (46.9%), followed by C. jejuni (13.4%) [36]. It should also be stated that studies regarding contamination of wild boar carcasses with Campylobacter are limited. A low carcass contamination level was reported in Germany, with only two positive samples for C. coli (1.57%) and one for C. jejuni (0.79%) out of 127 carcasses examined [37].
Similarly, data on the occurrence of Campylobacter spp. in Italian wild boar are scant. Ercolini et al. tested 50 muscle tissue samples from wild boars that were collected in Liguria and reported that only 2% were Campylobacter-positive (for C. lari) [38]. More recently, Stella et al. studied the presence of Campylobacter in wild boar hunted in northern Italy, and found 51.8% of faecal samples positive out of 56 tested and 16.7% of carcasses positive out of 30, with absolute values similar to those obtained in our study, in which 42.62% of faecal samples and 18.18% of carcasses samples were positive [39].
Our studies also showed the presence of Campylobacter in the liver and gallbladder of examined wild boar. Hepatic contamination with Campylobacter was proven in poultry, where it was a not-uncommon cause of human food-borne disease strongly linked to the ingestion of undercooked pâté [40]. Previous research also investigated these bacteria in suids: The prevalence of Campylobacter in pig livers was studied in Germany by von Altrock et al., and 10% of the 1500 pig livers sampled from 50 fattening herds were found to be positive, with C. coli as the predominant species [41]. Also in the UK, Campylobacter was isolated from 18.3% of pig offal samples, including liver, drawn from retailed food [42]. It should be stated, however, that new data regarding the occurrence of Campylobacter in wild boar liver are only gathered sporadically [43], and the first confirmation that raw liver may represent an important source of infection for the consumer or the operators who eviscerate wild boar after hunting is provided in our results. It should also be emphasised that even though C. lanienae has often been isolated from wild boar [44], this is the first time that this species has been isolated from wild boar liver and bile.
The present study confirmed multiple infections involving two species of Campylobacter in four tested wild boar. The possibility of a co-infection with different strains and species of Campylobacter was previously highlighted both in animals and humans [45][46][47]. The simultaneous occurrence of several Campylobacter species was confirmed in dog faeces samples where carriages of C. coli/C. lari, C. jejuni/C. coli, C. jejuni/C. upsaliensis and C. upsaliensis/C. lari were observed [48]. In humans, co-infection with multiple Campylobacter was described in infants co-infected with C. jejuni/C. infans/C. curvus/C. concisus/C. helveticus/C. upsaliensis, C. jejuni/C. infans/C. upsaliensis and C. concisus/C. infans species [49].
Our study also indicates a higher prevalence of Campylobacter in males (25.39%) than in females (19.17%). This phenomenon is difficult to explain; however, in this context, Zeng et al. demonstrated a more efficient immune response to Gram-negative bacteria in female than in male mice [43]. It was proven that innate antibodies against Gramnegative bacteria, such as enteropathogenic Escherichia coli, were present only in female mice, developed as a response to oestrogen hormones, and were not dependent on previous exposure to the antigen [50]. Thus, differences in sex hormone levels could play a significant role in the immune response to bacterial infection and partially explain the results obtained in this study. This correlation was also supported by human Campylobacter incidence studies where, in meta-analysis, the data confirmed higher incidence rates of campylobacteriosis in men than in women [51]. Our study also proved that young animals can be an important source of Campylobacter, hosting larger bacteria populations than sub-adult and adult individuals. In fact, there is little information in the global literature about age-based changes in gut microbiota composition in wild boar. It is known, however, that the process of Campylobacter gastrointestinal tract colonisation in chickens may be affected by the maturation stage of the immune system and the developmental stage of the gut microbiota composition [52,53]. In studies conducted on pigs, it has also been indicated that Proteobacteria (to which Campylobacter belongs) are relatively more abundant in the guts of piglets, but their relative abundance considerably decreases in adulthood [54].
The successful identification of micro-organisms using MALDI-TOF MS mainly relies on the database containing the appropriate spectra of known organisms and on the obtainment of a sufficient protein signal [20]. The level (score) of identification is determined by existing mass-spectral signature patterns in the database and factors which influence spectra quality, especially culture conditions such as the type of media used and the atmo-sphere in which bacterial isolates are grown, technical issues such as the age and intensity of the laser and the number of laser shots applied, as well as the type of MALDI-TOF MS target plates and matrix used [57]. It has been proven that for infrequently encountered microorganisms, the results of MALDI-TOF MS identification are reliable if the protein spectra of such micro-organisms are covered by the database; otherwise, the generated results may only give a hint by finding a related species or, in some cases, may misidentify the micro-organism as a very closely related species [58,59]. Campylobacter lanienae, for which the lowest scores were given, is not often identified to the species level using the MALDI-TOF technique, and this species is also often overlooked in food safety and animal health diagnostic protocols. Therefore, the database used may contain a lower number of reference spectra for C. lanienae, and the paucity of data could explain the lower score values for this species that were obtained in this study. There is a need to extend the databases with the spectra from different C. lanienae strains, and particularly those isolated from various game-origin food matrices. Recently, the reliability of MALDI-TOF MS for Campylobacter identification was tested on 27 strains of C. coli and C. jejuni isolated from birds and gave 100% correct results [56]. On a tenfold higher number of Campylobacter isolates from poultry and turkey products, MALDI-TOF MS identified C. jejuni, C. coli and C. lari 96% correctly overall [59].
As for other Campylobacter species commonly reported in wild boar, such as the C. lanienae evident in this study, few data are available on the usefulness of MALDI-TOF MS for their identification. Kérouanton et al. used MALDI-TOF MS to identify isolates of C. lanienae, although they did not report data on the reliability and quality of the identification or whether some strains were not identified [60]. Identifying the species of the strain involved in the first case in Europe of C. lanienae-related human gastroenteritis, Fornefett et al. reported a low MALDI-TOF MS identification score and only confirmed the isolate identification by 16S rRNA gene partial sequencing and species-specific PCR [61]. It was shown for the first time using MALDI-TOF MS that it is possible to obtain secure species identification for some C. lanienae isolates. This study shows reliable identification scores for 34% of the tested C. lanienae. Making a correct identification of C. lanienae should not be overlooked, as its actual pathogenic potential is still not fully known. In fact, besides the first European case of gastroenteritis with this aetiological agent in a butcher from Germany investigated by Fornefett et al. (which was mentioned previously), there is also on record the report of the first human case of enteritis caused by C. lanienae in a 39-year-old Canadian woman living on a pig farm [24,61]. The microorganism has also been isolated in slaughterhouse workers in Switzerland [62].

Conclusions
The present study shows that wild boar may be a conspicuous source of Campylobacter spp., because 44.56% of the tested animals were found to be positive. It should be emphasised that all identified species of Campylobacter, i.e., C. coli, C. lanienae, C. jejuni and C. hyointestinalis, are a public health concern. The Campylobacter spp. on carcasses and offal may be a direct source of human infection, especially for hunters, game dressers, butchers and game consumers, while Campylobacter in faecal samples may implicate faecal contamination of the environment in indirect zoonotic transmission. It was shown that MALDI-TOF MS combined with a reliable database could be a powerful method for the identification of C. coli and C. jejuni species (the species of most significant epidemiological importance and those frequently identified). Taking into account the achieved C. lanienae MALDI-TOF MS identification parameters, it should be stated that there is a need to constantly extend spectrum databases in order to increase the sensitivity of the method.
There are some limitations in our research. Firstly, it was not always possible to collect all four types of samples for each wild boar individual. In addition, not all collected isolates were confirmed using MALDI-TOF because of the loss of their vitality. Despite these problems, we conducted an in-depth analysis of Campylobacter infections in wild boar, taking into account the sampling site, the animal age and its sex. We contended that MALDI-TOF MS is a useful tool for major Campylobacter species identification and showed the limitations of this method in identifying a minor species such as C. lanienae. The investigation draws attention to this species, which is frequently found in wild boar and has neglected but not totally negligible pathogenic potential. The occurrence of Campylobacter in wild boar faeces, meat and offal requires communication measures targeted to hunters and consumers and intended to minimize the possibility of food-borne disease outbreaks linked to the improper handling of hunted animals and their meat.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/foods12040778/s1, Figure S1: Characteristic spectra of Campylobacter coli isolates (n = 35) generated by the Bruker Ultraflextreme MALDI TOF system. The intensities and m/z of the ions are shown on the Y-and X-axes, respectively. Sample codes (in accordance with those in Table 3) are displayed on the right at the end of each spectrum; Figure S2: Characteristic spectra of Campylobacter lanienae isolates (n = 29), generated by the Bruker Ultraflextreme MALDI TOF system. The intensities and m/z of the ions are shown on the Y-and X-axes, respectively. Sample codes (in accordance with those in Table 3) are displayed on the right at the end of each spectrum; Figure S3: Characteristic spectra of Campylobacter jejuni isolates (n = 2) generated by the Bruker Ultraflextreme MALDI TOF system. The intensities and m/z of the ions are shown on the Y-and X-axes, respectively. Sample codes (in accordance with those in Table 3) are displayed on the right at the end of each spectrum.  Institutional Review Board Statement: Ethical review and approval were waived for this study since animals were sampled during hunting activities regulated by national law.

Informed Consent Statement: Not applicable.
Data Availability Statement: The data supporting reported results are available from the corresponding authors upon request.