Epidemiological study on feline gastric Helicobacter spp. in Japan.

Epidemiological and pathological studies on Helicobacter spp. in feline stomachs in Japan were conducted using genus- and species-specific (H. felis, H. bizzozeronii, H. heilmannii sensu stricto [s.s.] and H. pylori) polymerase chain reactions (PCRs), ureAB gene sequencing and histopathology. PCR results showed that 28 of 56 cats were infected with Helicobacter spp., and H. heilmannii s.s. was the most prevalent species by both PCR (28/28) and ureAB gene sequencing (26/28). Some of the sequences showed high similarities with those from human patients with gastric diseases (99%). There were no significant differences between Helicobacter spp.-positive and -negative cats in the severity of chronic gastritis (P=0.69). This is the first extensive epidemiological study on feline gastric Helicobacter spp. in Japan.

The genus Helicobacter is gram-negative, microaerophilic spiral bacteria and contains at least 40 species [1]. Helicobacter pylori (H. pylori) is the predominant species in humans and associated with various human gastric diseases [12, 17, 21]; however, it has been reported only three times in cats [2, 9, 18].

Non-H. pylori helicobacters (NHPH) have also been detected in the stomachs of humans and several animal species. According to previous reports, the prevalence of NHPH in cats is 41–100% [8]. The predominant NHPH in cats were H. heilmannii sensu stricto (s.s.), H. felis and H. bizzozeronii, whereas H. salomonis was less often detected. H. baculiformis was recently isolated from a feline stomach, and its prevalence has not yet been determined [8]. Most of these NHPH are thought to be associated with human gastric diseases including active chronic gastritis, acute gastritis, gastric ulcer and gastric low-grade MALT lymphoma [5, 8, 15, 22] and are responsible for causing a growing concern as zoonosis. In contrast, there are many conflicting reports regarding the pathogenicity of each NHPH in cats, and therefore, the pathogenic significance of these gastric NHPH in cats is controversial at present [8]. The conflicting results of the pathogenicity of each NHPH in cats may occur due to the differences in virulence between different isolates, as has also been described for H. pylori [6, 12]. Despite their potential importance, there are insufficient epidemiological data to estimate the prevalence of Helicobacter spp. in feline stomachs in Japan.

The purposes of this study were to determine the prevalence of different Helicobacter spp. in feline stomachs in Japan and to evaluate the associations of these species with histopathological changes.

Fifty-six cats that underwent upper gastrointestinal endoscopy for various reasons at the Veterinary Medical Center of the University of Tokyo (VMC-UT) from April 2013 to June 2016 and Japan Small Animal Medical Center from June 2015 to June 2016, without a prior prescription of the effective eradication protocols for gastric Helicobacter spp. for cats using the combination of proton pump inhibitor (omeprazole or lansoprazole) or famotidine and two of the following: amoxicillin, metronidazole or clarithromycin [3, 7], were enrolled in this study (Table 1). Informed consent was obtained from all cat owners, and the study protocol was approved by the animal care committee of VMC-UT. Gastric biopsy samples were obtained under anesthesia from the antrum, corpus and fundus, in this order, using an endoscope designed for animals (the Olympus VQ-8143B; Olympus Medical Systems Corp., Tokyo, Japan) and biopsy forceps (FB-54Q-1; Olympus Medical Systems Corp.). The biopsy forceps were washed intensively with water and 70% ethanol in between each collection from the targeted gastric regions. The samples obtained were subjected to polymerase chain reaction (PCR) and histopathology.

Table 1.

Characteristics of cats included in this study (n=56)

Gender
Female	24 (21 spayed)
Male	32 (30 castrated)
Ages (years)	1.6 to 15.0 (median: 9.3)
Body weight (kg)	1.8 to 8.8 (median: 4.1)
Breeds	Mixed breed (n=36), Russian Blue (n=5), American Shorthair (n=4), Maine Coon and Abyssinian (n=2 each) and others (n=7)
Chief complaints	Vomiting (n=37), anorexia (n=16), diarrhea or bloody stool (n=15), weight loss (n=10) and others (n=6)

DNA was extracted from specimens using a QIAamp DNA Mini Kit (Qiagen, Santa Clarita, CA, U.S.A.). Genus-specific PCR was performed, followed by species-specific (H. felis, H. bizzozeronii, H. heilmannii s.s. and H. pylori) PCR and a sequencing analysis of the partial ureAB gene to investigate Helicobacter at the strain level. Primers and amplification parameters for each specific reaction are shown in Tables S1 and S2, respectively.

The PCR products of the partial ureAB gene from Helicobacter spp.-positive samples were purified and cloned using the Wizard PCR Preps DNA Purification System (Promega Corp., Madison, WI, U.S.A.), 10× A-attachment mix (Toyobo Co., Ltd., Osaka, Japan) and pGEM-T Easy vector (Promega), followed by sequencing of plasmid DNA from eight clones per sample using BigDye Terminator v3.1 (Applied Biosystems, Foster City, CA, U.S.A.). Homology sequences matches were searched in the NCBI/GenBank database using the BLAST search tool.

A phylogenetic tree was constructed using the neighbor-joining method from sequences of clinical strains obtained in this study, the sequences of Helicobacter spp. in NCBI/GenBank most closely related to each clinical strain and the partial ureAB genes from urease-positive reference Helicobacter spp. [19, 20].

Histopathological evaluations were performed by two pathologists (K. U. and J. C.). The gastric biopsies were evaluated for evidence of gastritis according to the World Small Animal Veterinary Association criteria, and changes were graded as normal, mild, moderate or severe [4].

Fisher’s exact test was used to determine the association between the Helicobacter status (positive or negative) and category of gastritis (normal to mild or moderate to severe). Statistical analyses were performed using JMP Pro 11 (SAS Institute, Cary, NC, U.S.A.), with the level of significance set at P=0.05.

PCR results showed that 28 of 56 cats were infected with Helicobacter spp. (50%): H. heilmannii s.s. (25/28), H. bizzozeronii (7/28), H. felis (7/28) and H. pylori (0/28). Three of the genus-positive samples could not be amplified by any of the species-specific primers. The prevalence of Helicobacter spp. in this study falls within the range previously reported in cats outside of Japan (41–100%) [8]. The predominant Helicobacter species in the present study, as determined by PCR, was H. heilmannii s.s. (89%), a finding similar to those of epidemiological studies in Switzerland (86%) and Korea (86%) [10, 16], whereas the infection rate of H. felis in the present study (25%) was higher than those reported in Switzerland (0.0%) and Korea (9.5%). The lack of H. pylori infections found in the present study was the same as that found in these other countries. The prevalence of H. bizzozeronii was not determined in the previous studies. Correlations between the Helicobacter spp. infection rate and age and sex were not statistically significant. The relationship between the Helicobacter spp. infection rate and type of breed could not be investigated owing to the small number of each pure breed included in the present study. The detection rate of Helicobacter spp. was the highest in the gastric fundus (50%), followed by the corpus (43%) and antrum (21%), which is identical to the tendency observed in dogs in Japan [11]. There were no differences between each of the Helicobacter species regarding the tendency of the detection rate in the three gastric regions.

To investigate Helicobacter at the strain level, partial ureAB genes from 27 of the 28 genus-positive samples were sequenced, and 24 unique sequences were obtained (GenBank/EMBL/DDBJ accession numbers are shown in Fig. 1). No amplification was observed in one case. Homology matches identified using the BLAST program revealed that 18 strains from 26 cases were most closely related to H. heilmannii s.s. (AB778507, L25079, AB462258 or HM625826), with sequence similarities of 91–99%; two strains from two cases were most closely related to H. bizzozeronii (FR871757, 96–98%); and four strains from five cases were most closely related to H. felis (FQ670179, 98–99%) (Table 2). Of these sequences, nine strains showed high similarities (99%) with those of H. heilmannii s.s. strains (L25079, AB778507 and AB462258) that were obtained from human patients with gastric diseases, suggesting that these strains may contribute to zoonosis [13, 14, 20]. Another nine strains were most closely related (about 90% similarity) to known H. heilmannii s.s. strains, suggesting that there are novel subspecies of H. heilmannii s.s. in Japan.

Fig. 1.

A phylogenetic tree constructed based on genetic distances and from the partial ureAB gene sequences amplified from 27 Helicobacter-positive cats and other urease-positive Helicobacter species. All strain names are accompanied by GenBank/EMBL/DDBJ accession numbers. Strains with a black circle (●) at the beginning of their names are reference strains, whereas strains without a black circle are clinical strains obtained in this study. Strains with “T” at the end of their names and in front of accession numbers are type strains for their species. Clinical strain names include each cat’s individual identification number, and the numbers after hyphens indicate clones. The 24 strains in the box with a dotted line are thought to constitute H. heilmannii s.s. clusters. The clinical strains with underlines in the box with a dotted line indicate strains that were obtained from samples found positive for H. heilmannii s.s. by species-specific PCR. Bootstrap values (for branches present in more than 40% of 1,000 resamplings of the data) are indicated. H. pylori strains DA, 26695 and OB were used as an outgroup. Evolutionary analyses were conducted in MEGA 6.0.6.

Table 2.

Sequence similarities of the partial ureAB gene from 24 strains detected in the stomachs of 27 Helicobacter-positive cats

Number of strains (cases detected)	Most closely related species		Similarity
	Species	Gene accession no.	Within the specie (%)	Among the clinical strains (%)
18 (26)	H. heilmannii s.s.	AB778507	91–99	90–99
		L25079
		AB462258
		HM625826
2 (2)	H. bizzozeronii	FR871757	96–98	97
4 (5)	H. felis	FQ670179	98–99	98–99

Mild to severe chronic gastritis were the most frequently observed conditions in both Helicobacter spp.-positive and -negative cases (Table 3). The proportion of cats with moderate to severe chronic gastritis did not differ between the groups (P=0.69). H. heilmannii s.s. strains which were most closely related to those isolated from human patients with gastric diseases with sequence similarities of 99% were detected in 22 cats of the 28 Helicobacter spp.-positive cases, but there was no significant difference in the severity of gastritis between these strains-positive and Helicobacter spp.-negative groups (P=0.67). A comparison of gastric histopathological diagnoses between cases infected with different NHPH could not be performed, because H. heilmannii s.s. was detected in almost all the Helicobacter-positive cases (26/28), as determined by ureAB gene sequencing. Gastric lymphomas (low-grade or high-grade) were seen in both Helicobacter spp.-positive cats (25%) and -negative (11%) cats, while gastric carcinoma and amyloidosis were seen only in infected cats, and atrophic gastritis and large granular lymphocytic lymphoma were seen only in uninfected cats. Gastric adenocarcinoma was seen in a cat infected with Helicobacter species which was most closely related to H. heilmannii s.s. SH6 (AB462258, with similarity of 99%), which have been detected in a human patient with gastritis, and amyloidosis was observed in a cat infected with one which was similar to H. felis (FQ670179; 99%) by the sequence analysis of the partial ureAB gene.

Table 3.

Histopathological diagnoses of feline stomachs (n=56)

Histopathological diagnosis		Not infected with Helicobacter spp. (n=28) / (%)	Infected with Helicobacter spp. (n=28) / (%)
Normal		1 (3.6)	0 (0.0)
Chronic gastritis	Mild	19 (68)	15 (54)
	Moderate	2 (7.1)	2 (7.1)
	Severe	0 (0.0)	2 (7.1)
Atrophic gastritis		1 (3.6)	0 (0.0)
Gastric adenocarcinoma		0 (0.0)	1 (3.6)
Lymphoma	Low-grade	1 (3.6)	2 (7.1)
	High-grade	2 (7.1)	5 (18)
	LGLa)	2 (7.1)	0 (0.0)
Amyloidosis		0 (0.0)	1 (3.6)

a) Large granular lymphocytic.

There were some limitations to our study: all cases included in this study showed some signs of gastrointestinal disease, we utilized only two referral veterinary centers in Tokyo and Saitama, and some cases had been administered antibiotics prior to this study. These factors might have affected our determinations of the Helicobacter spp. infection rate and the proportion of each NHPH found in this study. In addition, the investigation of each NHPH pathogenicity was insufficient, because a comparison of gastric histopathological diagnoses between cases infected with different NHPH could not be performed in this study because H. heilmannii s.s. was detected in almost all (26/28) of the Helicobacter-positive cases, as determined by partial ureAB sequencing. To properly investigate the pathogenicity of each NHPH, in vitro cultivation of each NHPH and experimental infection of feline gastric cell lines or cats will be necessary.

In summary, this is the first extensive epidemiological study on feline Helicobacter spp. in Japan. H. heilmannii s.s. was the predominant Helicobacter species in feline stomachs, and some of the strains represented a possible zoonosis. Further studies, including in vitro cultivation of each NHPH, will be necessary to properly investigate the pathogenicity of each NHPH.