Surveillance of Shiga toxin-producing Escherichia coli and Campylobacter spp. in wild Japanese deer (Cervus nippon) and boar (Sus scrofa).

Increasing game meat consumption in Japan requires the dissemination of safety information regarding the presence of human pathogens in game animals. Health information regarding the suitability of these animals as a meat source is not widely available. In this study, we aimed to evaluate the safety of game meat and detect potential human pathogens in wild deer (Cervus nippon) and boar (Sus scrofa) in Japan. Fecal samples from 305 wild deer and 248 boars of Yamaguchi, Kagoshima, and Tochigi prefectures collected monthly for 2 years were examined for the prevalence of Shiga toxin-producing Escherichia coli (STEC) and Campylobacter spp. STEC was isolated from 51 deer consistently throughout the year and from three boars; O-antigen genotype O146, the expression of stx2b, and eaeA absence (n=33) were the major characteristics of our STEC isolates. Other serotypes included the medically important O157, stx2b or stx2c, and eaeA-positive (n=4) and O26, stx1a, and eaeA-positive strains (n=1). Campylobacter spp. were isolated from 17 deer and 31 boars. Campylobacter hyointestinalis was the most common species isolated from 17 deer and 25 boars, whereas Campylobacter lanienae and Campylobacter coli were isolated from three and two boars, respectively. Seasonal trends for the isolation of these bacteria were not significant. This study demonstrates that wild game animals carry human pathogens; therefore, detailed knowledge of the safe handling of game meat is needed to prevent foodborne infections.

The marketing of game meat as a food source is a practical method to utilize wild animals as a natural resource. In addition, increasing populations and densities of wild animals, such as deer (Cervus nippon) and boar (Sus scrofa) [16], have accelerated hunting to counter mounting economic damage to agriculture and forestry from the overpopulation of these animals [17]. Thus, the Japanese Ministry of the Environment has proposed addressing the overpopulation issue through nationwide population control measures. The Ministry began by revising the Protection and Control of Wild Birds and Mammals and Hunting Management Law in 2015, with the goal of halving the populations by 2023 [20]. The Ministry of Agriculture, Forestry, and Fisheries also promotes the use of wild animals as a food source as a way to control agricultural economic losses caused by high game animal populations [18]. As a result, the game meat market is rapidly growing [17], but appropriate hygiene for processing game meat has not been fully established in Japan. Therefore, the risk of foodborne infections from game meat is unknown.

There are national guidelines for processing wild deer and boar published by the Ministry of Health, Labor, and Welfare (MHLW) [22]; however, these comprise a modified version of the Slaughterhouse Act, addressing commercial processing, which is not easily followed by laypersons, since additional expertise is required. Both national and local governments have made efforts to improve the food safety of game meat by inspecting slaughter sites and providing advice on the best hygienic practices. However, to accomplish this task, basic information on which pathogens are carried by game animals is necessary. Previous research has reported the prevalence of foodborne bacteria in fecal samples from a number of areas in Japan between July and December [38], specifically Campylobacter spp. in boar (43.8%) and Shiga toxin-producing Escherichia coli (STEC) O157 (2.3%) and O26 (0.8%) in deer. Another research group isolated seven STEC strains (16.3%) from deer in Hokkaido, Japan in June and July [3]. These studies suggest that there is a real possibility of a health hazard in game meat. In domestic animals, foodborne bacteria surveys are performed regularly throughout the year on farms or in specific areas [14, 15, 25]. These surveillance studies show seasonal diversity and area specificity that are not available for wild game animals.

There have been several wild game meat-related cases of pathogenic bacterial infections reported in Japan; for example, three distinct cases of enterohemorrhagic E. coli infection in 1997, 2001, and 2009 were linked to deer meat consumption [21, 34, 40]. Additionally, while human campylobacteriosis related to game meat has not been reported in Japan, Campylobacter is a common agent of bacterial foodborne illness in this country [23]. Moreover, it has, as noted previously herein, been found in wild boar, indicating that establishing the infectious potential of game meat is necessary. Recently, surveillance of bacterial contamination of commercial game meat in Japan revealed STEC contamination in deer meat (0.8% of samples) [1], indicating the potential for game meat-related foodborne infection.

This study investigated the prevalence of STEC and Campylobacter spp. in wild deer and boar prior to meat processing. Fecal samples were regularly collected to obtain year-round data in a fixed area by sampling every month for 2 years in Yamaguchi and Kagoshima prefectures and for several months in Tochigi prefecture. We sought to determine whether isolates were identical clones or varied between animal species and sites. Variation in STEC was evaluated by O-antigen genotype (Og), stx gene type, and the presence of eaeA (intimin). Stx gene typing and eaeA detection were conducted to provide epidemiological information of STEC in wild animals to consider the possible threat to humans. Campylobacter species were identified using matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) and confirmed through an assessment of 16S rRNA gene sequences. The findings of this study, the largest survey of its kind in Japan, demonstrated differences in the rates of bacterial presence, seasonal trends, and geographical locations of STEC and Campylobacter spp. in wild deer and boars.

MATERIALS AND METHODS

Sample collection

Fresh fecal samples were collected from freshly killed wild deer and boar hunted in Japan in Yamaguchi and Kagoshima prefectures from June 2014 to May 2016 and in Tochigi prefecture from November 2015 to February 2016. Local hunters acted as volunteer sample collectors under instructions provided by the members of this study group. Hunting procedures complied with the Protection and Control of Wild Birds and Mammals and Hunting Management Law [19]. Thirty-seven boar samples were collected in Tochigi, 90 in Yamaguchi, and 121 in Kagoshima prefectures. A total of 142 deer samples were collected in Yamaguchi and 163 in Kagoshima. In total, samples from 248 boars and 305 deer were tested. In Yamaguchi and Kagoshima prefectures, it was possible to obtain samples on a regular basis. The number of animals that were sampled every month at each area ranged from 1 to 32, based on the monthly average over a 2-year period. Outside the normal hunting season (April to October), animals were hunted for wildlife control. Rectal fecal samples were collected in sterilized 10- or 30-ml tubes, and then immediately placed in a cold box or refrigerator until they were sent to Kagoshima University using refrigerated transportation and stored at 4°C until testing. Samples were processed for cultivation within 3 days of collection, except for samples from Tochigi, which were processed 5 days after collection due to a longer transport time.

STEC isolation and analysis

Using sterilized cotton swabs, fecal samples were streaked directly onto CHROMagar STEC medium (Kanto Chemical Co., Inc., Tokyo, Japan) and incubated at 37°C for 18 hr. Purple colonies were selected and inoculated onto DHL agar plates (Eiken Chemical Co., Inc., Tokyo, Japan) to confirm lactose degradation and then inoculated onto a nutrient agar plate (Nissui Pharmaceutical Co., Inc., Tokyo, Japan) for further characterization. If purple colonies were numerous, multiple colonies were selected for further evaluation. For species identification of isolated bacteria, MALDI-TOF MS was performed using an autoflex speed TOF/TOF-KG (Bruker Daltonics Inc., Bremen, Germany) based on the software MBT_FC1.par method of flexControl (Burker Daltonics) and real-time classification with the Autoxecute method of the MALDI Biotyper (Burker Daltonics). Bacterial samples were prepared with the formic acid-extraction method, according to the manufacturer’s instructions, and spotted onto an MTP 384 target plate polished steel BC (Burker Daltonics). A Bacterial Test Standard (Burker Daltonics) was used for the calibration of each test run. A matching score over 1.7 was considered successful identification.

Shiga toxin type 1 (stx1) and 2 (stx2) genes were detected using PCR and sequenced for subtyping as previously reported [39]. E. coli isolates were considered STEC when they carried either stx1 or stx2. The intimin-coding gene eaeA was detected by PCR as previously reported [5].

Isolate O-antigens were determined using O-genotyping multiplex PCR, which can comprehensively identify 147 individual O serogroups [12]. When the O-genotype (Og) of an isolate was identified as one of the major serotypes for foodborne STEC infection in Japan (O26, O103, O111, O121, O145, O157, or O165) [32], a seroagglutination test was performed to confirm the O-antigen type using commercial antiserum according to the manufacturer’s protocol (Denka Seiken Co., Ltd., Tokyo, Japan).

Campylobacter isolation and analysis

One gram of a fecal sample was added to 9 ml of Preston enrichment broth (Oxoid Ltd., Hampshire, UK) and incubated under microaerobic conditions (7.5% CO2, 5% O2, 7.5% H2, and 80% N2) at 42°C for 48 hr. The enriched cultures were then streaked on Butzler agar plates (Oxoid Ltd.) and again incubated under microaerobic conditions at 42°C for 48 hr. Colonies were observed for the typical screw movement indicative of Campylobacter using phase-contrast microscopy (Nikon Corp., Tokyo, Japan) and then tested for catalase and oxidase as previously described [14]. Bacterial species of suspected colonies were identified by MALDI-TOF MS as described previously herein. For species confirmation, 16S rRNA gene sequences of the isolates were analyzed using primers CG12F (TTGATCCTGGCTCAGAGT) and CG1507R (TTCACCCCAGTCGCTGAT) [27, 30] to amplify a 1,495-bp product and the additional primer C412F (GGATGACACTTTTCGGAGC) for sequence analyses as reported elsewhere [33, 37]. The resulting sequences were searched for homologous sequences using BLASTn (https://blast.ncbi.nlm.nih.gov/Blast.cgi). Multiplex PCR specific for the Campylobacter jejuni hipO gene and Campylobacter coli ceuE gene was conducted to confirm these two species as previously reported [42].

Statistical analyses

Statistics were performed using the statistical program file ystat2013 (Igaku Tosho Shuppan, Tokyo, Japan). Isolation rate comparisons between animal species, prefectures, and seasons were calculated by the 2 × 2 or m × n χ2 test or 2 × 2 or m × n Yate’s χ2 test when the dataset had fewer than 10 entries. P<0.05 was considered significant. Seasonal differences were compared over four seasons as defined by the Japan Meteorological Agency as follows: winter (December to February), spring (March to May), summer (June to August), and autumn (September to November).

RESULTS

STEC prevalence

STEC was isolated from wild deer (n=51, 16.7%) and boar (n=3, 1.2%) in Yamaguchi and Kagoshima prefectures. STEC presence in deer was significantly higher than that in boar (P<0.05, Yate’s χ2 test; Table 1), and multiple STEC isolates were obtained from some individual animals; thus, the number of isolates (n=66) exceeded the number of positive animals (Table 2). Total numbers of STEC-carrying animals were not significantly different between prefectures. STEC was regularly isolated from wild deer throughout the year in both prefectures and no significant difference was observed.

Table 1.

Prevalence of Shiga toxin-producing Escherichia coli (STEC) and Campylobacter species in wild deer and boars and their prefecture of origin

		Origin prefecture		Number of positive animals (%)
				All season	Winter	Spring	Summer	Autumn
STEC				54 / 553 (9.8)	10 / 179 (5.6)	18 / 132 (13.6)	13 / 111 (11.7)	13 / 131 (9.9)
	Deer	Total		51 / 305 (16.7)a)	10 / 82 (12.2)	17 / 83 (20.5)	12 / 77 (15.6)	12 / 63 (19.0)
			Yamaguchi	22 / 142 (15.5)	0 / 19 (0)	9 / 43 (20.9)	9 / 56 (16.1)	4 / 24 (16.7)
			Kagoshima	29 / 163 (17.8)	10 / 63 (15.9)	8 / 40 (20.0)	3 / 21 (14.3)	8 / 39 (20.5)
	Boar	Total		3 / 248 (1.2)a)	0 / 97 (0)	1 / 49 (2.0)	1 / 34 (2.9)	1 / 68 (1.5)
			Yamaguchi	2 / 90 (2.2)	0 / 23 (0)	0 / 28 (0)	1 / 16 (6.3)	1 / 23 (4.3)
			Kagoshima	1 / 121 (0.8)	0 / 63 (0)	1 / 16 (6.3)	0 / 8 (0)	0 / 24 (0)
			Tochigi	0 / 37 (0.0)	0 / 11 (0)	0 / 5 (0)	NT	0 / 21 (0)
Campylobacter spp.				48 / 553 (8.7)	21 / 179 (11.7)	9 / 132 (6.8)	10 / 111 (9.0)	8 / 131 (6.1)
	Deer	Total		17 / 305 (5.6)b)	7 / 82 (8.5)	4 / 83 (4.8)	2 / 77 (2.6)	4 / 63 (6.3)
			Yamaguchi	1 / 142 (0.7)c)	0 / 19 (0)	0 / 43 (0)	1 / 56 (1.8)	0 / 24 (0)
			Kagoshima	16 / 163 (9.8)c)	7 / 63 (11.1)	4 / 40 (10.0)	1 / 21 (4.8)	4 / 39 (10.3)
	Boar	Total		31 / 248 (12.5)b)	14 / 97 (14.4)	5 / 49 (10.2)	8 / 34 (23.5)	4 / 68 (5.9)
			Yamaguchi	12 / 90 (13.3)	7 / 23 (30.4)	1 / 28 (3.6)	2 / 16 (12.5)	2 / 23 (8.7)
			Kagoshima	18 / 121 (14.9)	7 / 63 (11.1)	4 / 16 (25.0)	6 / 18 (33.3)	1 / 24 (4.2)
			Tochigi	1 / 37 (2.7)	0 / 11 (0)	0 / 5 (0)	NT	1 / 21 (4.8)§

NT, Not tested. § Only October and November. a), b), c): Significant difference between same alphabet (P<0.05).

Table 2.

Characteristics of Shiga toxin-producing Escherichia coli isolates

O antigen genotype	Stx gene	eaeA gene	Animal	Originb)	Number of isolates
Og 11	stx2b	-	Deer	K	11
Og 26	stx2a	+	Deer	Y	1
Og 54	stx2a	-	Deer	K	1
	stx2b	-	Deer	K	8
Og 84	stx1a	+	Deer	K	3
Og 113	stx1c + stx2d	-	Deer	K	1
Og 128	stx2b	-	Deer	Y	1
Og 131	stx2b	-	Deer	Y	1
Og 146	stx1c + stx2b	-	Deer, Boar	Y	6, 1
	stx2b	-	Deer	K, Y	12, 14
	stx2c	-	Deer	Y	1
Og 156	stx1 + stx2a)	+	Deer	K	1
Og 157	stx2b	+	Boar	Y	1
	stx2c	+	Deer	K	1
		+	Boar	Y, K	1, 1

a) Subtype undetermined. b) K, Kagoshima; Y, Yamaguchi.

STEC isolate characteristics

O-antigen genotyping, stx gene identification, and eaeA presence revealed the diversity of STEC in wild deer and boar. Ten O antigens were identified from wild deer and two from boar (Table 2). The stx gene identification was determined by sequence analyses and sequences were deposited in GenBank as follows: stx1, LC388483–LC388499; stx2, LC388500–LC388568. By combining the three identifiers (Og, stx, and eaeA), we were able to characterize our STEC isolates. The major subtype of STEC was Og146, stx2b, and eaeA-negative (n=33) from deer and boar, followed by Og11, stx2b, and eaeA-negative (n=11) and Og54, stx2a or stx2b, and eaeA-negative (n=9) from deer.

Two major serotypes from severe enterohemorrhagic E. coli infection in humans [32] were found in deer and confirmed serologically, specifically O26 (n=1) and O157 (n=4). The O157 isolates, one with stx2b and one with stx2c, were obtained from the same animal, demonstrating that a single animal can carry different subtypes of STEC. Of the stx gene subtypes tested, stx2b was dominant (n=55) in both deer and boar, but stx1a, stx1c, stx2a, stx2b, stx2c, and stx2d were also found (Table 2). Furthermore, eaeA was present in certain Og types, though not in the major ones noted previously herein.

Campylobacter prevalence

Campylobacter species were also isolated from both animals in all prefectures (Table 1). Wild boar (n=31, 12.5%) had a significantly higher Campylobacter spp. isolation rate than deer (n=17, 5.6%; P<0.05, χ2 test). Deer Campylobacter spp. isolation rates were higher in Kagoshima prefecture than in Yamaguchi (P<0.05, Yate’s χ2 test), but boar Campylobacter spp. isolation rates did not differ significantly between prefectures. Seasonal trends were not significant for any animal or prefecture. Isolation rates peaked in winter for deer and in summer for boar; however, in boar, the season with the highest isolation rate differed by prefecture (Table 1). The dominant species for all conditions was Campylobacter hyointestinalis, and results for this species showed statistical significance for the same factors as those for all Campylobacter spp.; specifically, boar had higher isolation rates than deer (P<0.05, χ2 test) and deer in Kagoshima had higher isolation rates than those in Yamaguchi (P<0.05, Yate’s χ2 test; Table 3).

Table 3.

Number of animals carrying Campylobacter species and the species of isolated Campylobacter

Origin prefecture		Number of animals	Number of positive animals (%)
			Total	C. hyointestinalis	C. lanienae	C. coli	Campylobacter sp.
Deer	Total	305	17 (5.6)a)	17 (5.6)b)	0 (0)	0 (0)	0 (0)
	Yamaguchi	142	1 (0.7)c)	1 (0.7)d)	0 (0)	0 (0)	0 (0)
	Kagoshima	163	16 (9.8)c)	16 (9.8)d)	0 (0)	0 (0)	0 (0)
Boar	Total	248	31 (12.5)a)	25 (10.1)b)	3 (1.2)	2 (0.8)	1 (0.4)
	Yamaguchi	90	12 (13.3)	12 (13.3)	0 (0)	0 (0)	0 (0)
	Kagoshima	121	18 (14.9)	12 (9.9)	3 (2.5)	2 (1.7)	1* (0.8)
	Tochigi	37	1 (2.7)	1 (2.7)	0 (0)	0 (0)	0 (0)
Total		553	48 (8.7)	42 (7.6)	3 (0.5)	2 (0.4)	1 (0.2)

a), b), c), d): Significant difference between same alphabet (P<0.05). *Isolate C147, identified as C. jejuni via time of flight mass spectrometry, C. coli from 16S rRNA sequence, and negative for both C. jejuni and C. coli via multiplex PCR.

Species identification of Campylobacter isolates

A total of 48 isolates of Campylobacter were initially identified by TOF-MS. The major species was C. hyointestinalis (n=42, 87.5% of all Campylobacter isolates), which was found in both deer and boar in all areas studied. C. coli (two isolates in winter), C. jejuni (one isolate in winter), and Campylobacter lanienae (three isolates, in winter, spring, and summer) were isolated only from wild boar in Kagoshima.

The MALDI TOF-MS Biotyper database was constructed primarily for medical use, but the isolates in the current study were field samples. Thus, 16S rRNA gene sequence analysis was conducted to confirm bacterial species identification. The resulting sequences were deposited in GenBank (16S rRNA, LC388439–LC388476). Eight isolates were not sequenced owing to inadequate bacterial storage. Using BLAST analyses, the subspecies of all C. hyointestinalis isolates was identified as C. hyointestinalis subspecies hyointestinalis with 100% identity (1,303 bp; Accession numbers AB301960.1 and AB301956.1). Isolates C139 and C198, however, had only 99.8% and 99.9% identity to the same sequences, respectively. C439, however, had 100% identity to another unspecified C. hyointestinalis subspecies sequence (Accession numbers LR698977.1, NR_118518.1, and NR_037052.1).

Isolates identified as C. lanienae by TOF-MS corresponded with the C. lanienae strains with identity up to 99.9% (1301 bp; Accession number AF550664.1). However, the identification of C. jejuni and C. coli isolates (C116, C211, and C147) was confusing. Isolate C147 was initially identified as C. jejuni via TOF-MS; however, the 16S rRNA gene sequence (1315 bp) had 100% identity with that of C. coli (Accession number HG326877.1). Conventional duplex PCR was negative for both C. jejuni and C. coli, and thus, we could not confirm species identification for this isolate. The C. coli isolates C116 and C211 were first identified by TOF-MS, and the sequences of the 16S rRNA genes, 1,335 bp of C116 and 1,318 bp of C211, had 100% identity with those of C. coli strains (multiple sequences, including Accession number AF550621.1 for C116 and CP046317.1 for C211). Interestingly, C211 also had 100% identity with many C. jejuni stains (i.e. Accession numbers CP048769.1, CP048767.1, CP048765.1, and others). Moreover, C116 and C211 were positive for C. coli and negative for C. jejuni by duplex PCR. To clarify the identification of these three isolates, we added a biochemical characterization test using API Campy (bioMerieux SA, Marcy l’Etoile, France); C116 and C211 were confirmed as C. coli but C147 was nonspecific (data not shown). From these results, we concluded that Campylobacter species identification by TOF-MS and gene sequences was consistent for all isolates, except for the one isolate identified as C. jejuni. Final species identification results are summarized in Table 3.

DISCUSSION

Foodborne disease-causing bacteria were isolated from wild deer (C. nippon) and boar (S. scrofa), both hunted and processed locally for consumption in Japan. The study was conducted over 2 years and at three locations. Regular, year-round sampling resulted in the largest investigation ever into STEC and Campylobacter occurrence in these two major game animal species. STEC O157, which is well-known to cause severe disease in humans [6], was found in both animals. Previously wild deer in Japan have been reported to carry O157 [38], but it has not been reported in boar. While STEC prevalence was low in this study, particularly in wild boar, both deer and boar meat could still be a source of foodborne infection. Characterization of isolates revealed heterogeneity of the pathogens in both animals.

To date, STEC is one of the most serious bacterial hazards from game meat. In Japan, STEC outbreaks in humans from the consumption of game meat were reported in 1997, 2001, and 2009 [21, 34, 40]. However, the official foodborne STEC-infection record is limited, but STEC from retail deer meat was recently isolated in Japan [1] and has been previously identified in deer game meat in Belgium (16 to 22%) and Spain (5.6%) [7, 36]. Here, we isolated many serotypes of STEC with the ability to infect humans, as shown in the official records in Japan [32]. We were able to identify a variety of STEC O-antigens using genotyping PCR [12], and we recommend this method for serotyping E. coli isolated from miscellaneous sources, since the seroagglutination test using commercial antiserum did not work in our case. In our hands, isolates from our field samples agglutinated with multiple O-antigen antisera, even though control STEC strains reacted as expected (data not shown). Therefore, we suggest PCR for the initial identification of O antigens.

The relationship between O antigens and stx genes and their subtypes is unclear. However, Og146 and stx2b were widely detected in the present study, particularly in deer. This might indicate some unique characteristics of wild deer, but further work would be required. Recent studies have suggested the potential adaptation of specific STEC strains to deer, as they frequently detected stx2d from deer fecal samples from Hokkaido and commercially available deer meat in Japan [1,2,3]. Hokkaido is the northernmost prefecture and is geographically separated from other major islands of Japan. It is possible that this has caused some regional specificity in STEC characteristics.

The actual pathogenicity of the STEC isolates associated with stx genes and subtypes was not examined in our study, since none of our fecal samples appeared to include diarrhea (data not shown), suggesting that these do not function as apparent enteropathogenic factors in deer and boar. Some of our isolates contained stx1a or stx2a, both of which have been shown to correlate with severe infections in humans [4, 28]. We considered the presence of the eaeA gene, an additional STEC virulence criterion that was found in certain O-antigen types, suggesting that game animals in Japan have the potential to carry highly virulent bacteria.

Previous studies have suggested that high air temperature and humidity (summer) can promote Campylobacter (particularly C. jejuni and C. coli) colonization in broiler chicken flocks [14, 24, 35]. Interestingly, the season associated with the highest isolation rate for Campylobacter spp. in this study was not summer and varied according to animal species and sampling location, which suggests that there are no seasonal trends for Campylobacter contamination of game meat. We speculate that the difference between wild and domestic animals with respect to Campylobacter colonization season was due to the dominant Campylobacter species (C. hyointestinalis in wild animals as opposed to C. jejuni and C. coli, as noted previously herein, in domestic flocks). Nevertheless, C. coli was found in samples in the present study but were obtained during winter, suggesting a difference between domesticated animals and wild game meat animals. Since the usual hunting season in Japan is from November to February, seasonal information for Campylobacter is critical to evaluate human health hazards, especially to hunters, abattoir personnel, and consumers. A previous study on STEC in dairy cattle revealed that bacteria are significantly more likely to be shed at higher temperatures [43]. Again, in our study, there was no seasonal variation in STEC colonization among wild deer and boars. This demonstrates the importance of our year-round survey, which enables a more complete collection of data that accurately reflects possible seasonal variations within each population. Our study also showed that wild animals have a different bacterial presence rate compared to domestic animals and that the distribution of bacteria in wild animals differs by geographical location.

We showed that wild boars carried C. coli, C. lanienae, and C. hyointestinalis, but deer carried only C. hyointestinalis. Campylobacter was more common in boars than deer. The low prevalence that we found for C. jejuni and C. coli, which are the major species associated with worldwide foodborne bacterial contamination [41], is consistent with previous reports for wild deer and boar. Inoue et al. reported C. jejuni in deer (4.5%) and boar (1.9%) and C. coli only in boar (6.5%) in Tokushima prefecture [13], whereas Sasaki et al. reported C. jejuni only in boar (0.8%) and no Campylobacter species in deer in the selected areas in Japan [38]. Different isolation rates could occur due to different populations, sample quantities, and/or culture media and conditions, but worldwide epidemiology tends to have similar issues. Since our study and previous studies [13, 38] used high culture temperature for thermotolerant C. jejuni and C. coli, the prevalence of non-thermotolerant species might be higher in wild deer and boars. C. hyointestinalis was originally found in domestic swine and ruminants [33] and later in humans with gastric diseases, including a case wherein the agent was identified as C. hyointestinalis subsp. hyointestinalis [9, 26]. C. hyointestinalis subsp. hyointestinalis was reported as the most common Campylobacter species in semi-domestic reindeer in Finland [10], which suggests that it might be common to the family Cervidae. C. lanienae was first isolated from a routine hygiene survey in abattoir workers [31] and was later identified in healthy domestic porcine populations [33, 37]. Recently, evidence of C. lanienae as a human pathogen has been reported [29]. Although C. hyointestinalis and C. lanienae might not be foodborne agents for humans, people who have frequent contact with these animals should understand the risk of infection.

There was one Campylobacter isolate from a boar that could not identified to the species level by the methods that we used. It was not a major species in this study. However, this demonstrates the difficulty in identifying microorganisms from wild animals that might carry unidentified or novel species.

The data from this study emphasize the importance of proper handling of game animals after hunting, which is upstream of game meat processing. The guidelines established by MHLW for the hygienic slaughtering process of wildlife, based on the Slaughterhouse Act of 2013 and the HACCP-based hygienic control program of 2018, remind us of the risk of zoonotic infection during wild animal handling and that the risk is not limited to foodborne transmission. Human pathogenic bacteria commonly exist in different areas throughout the year, including during the hunting season, as confirmed in our study. The prevalence of pathogenic bacteria in game animals, mainly deer and boars, has been surveyed in many countries [7, 8, 10, 11, 36]. These studies show that the presence of human pathogenic bacteria is common in game animals. Therefore, epidemiological data of these pathogens is important to evaluate the risk associated with game meat in Japan and to properly convey that risk to consumers and hunters. The necessity of public education based on scientific data to prevent foodborne illness is of particular importance with the recent increase in game meat consumption. Future studies will continue the surveillance for human pathogens, both foodborne and otherwise, in game animals to provide further information for zoonotic disease control.