INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Cats are recognized reservoirs for enteric Campylobacter spp., including Campylobacter jejuni, Campylobacter coli, Campylobacter upsaliensis, and Campylobacter helveticus (5, 13, 20, 37, 58). C. jejuni and C. coli are among the most frequently encountered human enteric pathogens worldwide (2, 3). C. upsaliensis, a catalase-negative or weakly positive Campylobacter sp. initially isolated from diarrheic or nondiarrheic domestic dogs and cats (20, 48), has also been associated with enteritis (25, 44, 55) and bacteremia (31, 39, 44) in humans. More-serious illnesses, including spontaneous abortion and hemolytic-uremic syndrome, have also been reported for a human infected with C. upsaliensis (27). Campylobacter infections, particularly C. jejuni infections, are zoonotic and are a particular problem among puppies and kittens from shelters (13, 18, 45). C. helveticus, which is closely related to C. upsaliensis, has also been isolated from domestic cats and dogs but has not been linked with human disease (52).

Although Helicobacter spp. are better known as gastric pathogens (7, 10, 18, 19, 28, 41, 43), there has been an increasing interest in enterohepatic Helicobacter spp. isolated from humans and animals. Helicobacter canis has been isolated from normal and diarrheic dogs, cats, and diarrheic humans as well as from the liver of a dog with hepatitis (6, 12, 17, 54); "Flexispira rappini " strains, which represent at least 10 Helicobacter taxa, have been isolated from feces of mice, sheep, dogs, and humans and have been associated with abortion in sheep (9). Helicobacter pullorum, first isolated from the feces and liver of chickens, also has been cultured from diarrheic humans (53). Helicobacter canadensis, originally misdiagnosed as H. pullorum, has been isolated from Canadian patients with diarrhea (14). Helicobacter cinaedi, isolated from the feces of healthy hamsters, also has been recovered from the inflamed lower bowel and blood of immunocompromised adults and children with diarrhea (11, 24, 57). Recently, mixed infections of Helicobacter spp. and Campylobacter spp. have been noted in diarrheic children residing in developing countries (33). The purpose of this study is to describe for the first time coinfection of enteric Helicobacter spp. and Campylobacter spp. in cats and to provide molecular characterization of novel Helicobacter species in these same animals.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Animals. During a 6-year period, 227 purpose-bred cats were obtained from three commercial sources. The cats purchased were from specific-pathogen-free colonies certified to be negative for feline leukemia virus, feline immunodeficiency virus, and feline coronavirus. All cats were evaluated for body condition, appetite, and episodic diarrhea while in quarantine. At the time of fecal culture all cats were clinically healthy.

Bacterial culture. Rectal swabs from the cats were streaked onto cefoperazone-vancomycin-amphotericin B antibiotic-impregnated media (Remel Laboratories, Lenexa, Kans.) and grown under microaerobic conditions in vented jars containing N₂, H₂, and CO₂ (80:10:10) at 37 and 42°C. The primary isolates were Gram stained and tested for urease, oxidase (Bactidrop; Remel Laboratories), and catalase (3% H₂O₂). Isolates were also assayed for their ability to hydrolyze hippurate (30). The sensitivity of the isolates to nalidixic acid and cephalothin was tested with antibiotic-impregnated disks. Bacteria identified as Campylobacter-like organisms (CLOs) were frozen at 20°C in 20% glycerol and brucella broth. Bacteria were subsequently reisolated on primary media for molecular characterization.

DNA extraction. DNA was extracted from individual colonies grown on cefoperazone-vancomycin-amphotericin B plates by using InstaGene matrix (Bio-Rad Laboratories, Hercules, Calif.). Bacteria were resuspended in 1 ml of double-distilled water in a microfuge tube. After centrifugation and removal of the supernatant, 200 µl of InstaGene matrix was added to the pellet and incubated at 56°C for 30 min. The samples were then boiled for 10 min and centrifuged for 5 min at high speed, and then 10 µl of the supernatant was used for the PCR.

PCR amplification. The nucleotide sequences and the sources of all primers used to amplify the cat fecal isolates are listed in Table 1. PCR amplifications were performed with a Thermal Cycler and an Expand high-fidelity PCR system (Roche Molecular Biochemical, Indianapolis, Ind.). Each reaction mixture (100 µl) contained 1× polymerase buffer, a 0.5 µM concentration of each of two primers, a 200 µM concentration of each deoxyribonucleotide triphosphate, and bovine serum albumin (200 µg/ml). The samples were heated at 94°C for 4 min, briefly centrifuged, and cooled to 58°C. Polymerase (2.5 U) was then added. Amplification was achieved by denaturation at 94°C for 1 min, annealing at 58°C for 2 min, and elongation at 72°C for 2 min. For primers specific for C. helveticus, C. upsaliensis, C. jejuni, and C. coli, the annealing was at 55°C. A 15-µl portion of the sample was then electrophoresed through a 1% agarose gel and followed by ethidium bromide staining and UV illumination.

Restriction fragment length polymorphism of Helicobacter 16S rRNA gene. Primers C97 and C05 (Table 1) were used to amplify the 1.2-kb PCR fragments from all the Helicobacter species isolates identified in this study. Amplified DNA (20 µl) was digested with 10 U of AluI in the buffer recommended by the enzyme manufacturer at 37°C for 3 h. Restriction patterns were compared after the digested PCR products were separated on a 6% Visigel separation matrix (Stratagene, LaJolla, Calif.).

Cloning and sequencing 16S ribosomal DNA PCR products. A pGEM-T vector (Promega, Madison, Wis.) was used for cloning the PCR products. The PCR products were purified from a low-melting-point agarose gel with the QIAquick PCR purification kit (Qiagen, Valencia, Calif.). Fifty nanograms of purified PCR product was ligated with 50 ng of pGEM-T vector at 4°C overnight and used to transfer into competent JM109 cells. Ampicillin plates with 5-bromo-4-chloro-3-indolyl--D-galactopyranoside (X-Gal) and isopropyl--D-thiogalactopyranoside (IPTG) were used to select positive clones. Plasmid DNA was isolated from Escherichia coli using a Qiaprep mini spin kit (Qiagen). The 1,600-bp DNA sequences of the 16S rRNA cistrons of two pure isolates and the 1,200-bp sequences of eight PCR products obtained with the use of Helicobacter genus-specific primers were obtained by cycle sequencing. Purified DNA from the PCR and plasmid DNA were sequenced using an ABI prism cycle sequencing kit (BigDye Terminator cycle sequencing kit with AmpliTaq DNA polymerase FS; Perkin-Elmer). The primers listed in Table 1 were used for sequencing. Quarter dye chemistry was used with 80 µM primer and 1.5 µl of DNA in a final volume of 20 µl. Cycle sequencing was performed using an ABI 9700 thermal cycler, with 25 cycles of denaturation at 96°C for 10 s and annealing and extension at 60°C for 4 m. Sequencing reactions were run on an ABI 377 DNA sequencer.

16S rRNA data analysis. Sequences were first screened by a BLAST analysis comparing them to all entries in GenBank (1). Sequence data were then entered into RNA, a program set for data entry, editing, sequence alignment, secondary structure comparison, similarity matrix generation, and dendrogram construction for 16S rRNA in Microsoft QuickBasic for use with personal computers, and were aligned as previously described (42). Our database contains over 1,000 sequences obtained in our laboratory and over 500 obtained from GenBank. Dendrograms were constructed by the neighbor-joining method (47).

Nucleotide sequence accession numbers. The GenBank accession numbers for the strains MIT 98-1705-1 and MIT 95-234-6 are AF336947 and AF336948, respectively. Sequences for the 1,200-bp partial sequences are available from the corresponding author.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Campylobacter spp. and Helicobacter spp. commonly colonize the intestines of cats used for biomedical research. During the 6 years of the survey, 227 fecal samples were collected for culture of CLOs. Sixty-four cats were initially diagnosed as positive for CLOs, based on primary isolation and characterization of bacteria by colony morphology and biochemical tests. The bacteria were slightly curved or spiral-shaped, gram-negative organisms and grew under microaerobic conditions at 37°C. Selected isolates grew at 42°C. Isolates typically were urease negative, oxidase positive, weakly catalase positive or negative, and sensitive to disks containing 30 µg of nalidixic acid and 30 µg of cephalothin.

View larger version (26K): [in this window] [in a new window]   FIG. 1.   Cocolonization of cats by Campylobacter spp. and Helicobacter spp. (A) Campylobacter genus-specific primers were used to amplify DNA extracted from cat fecal isolates. Lane MW, 100-bp DNA ladder; lane 1, reagent control; lane 2, Campylobacter-positive control; lanes 3 to 19, isolates from cat feces. (B) Helicobacter genus-specific primers were used to amplify DNA extracted from cat fecal isolates. Lane MW, 1-kb DNA ladder; lane 2, Helicobacter-positive control; lane 3, reagent control; lanes 3 to 19, isolates from cat feces.

Differentiation of Campylobacter spp. isolated from cat feces. PCR assays specific for C. coli, C. helveticus, C. upsaliensis, and C. jejuni were used to screen the 47 fecal isolates which were positive for Campylobacter spp. by analysis with genus-specific primers. Of the bacterial cultures, 34% (16 of 47 cultures) were found to be mixed cultures of more than one Campylobacter sp. (Fig. 2). Eighty-three percent (39 of 47 cultures) were positive for C. helveticus; 47% (22 of 47 cultures) were positive for C. upsaliensis; and 6% (3 of 47 cultures) were positive for C. jejuni. In tests using C. coli 16S rRNA gene-specific primers, positive results were not obtained for any of the cultures.

View larger version (83K): [in this window] [in a new window]   FIG. 2.   Results of genus-specific PCR. (A) Primers specific for C. helveticus amplified 16S rRNA PCR products from 12 of 17 cat fecal isolates. Lane MW, 1-kb DNA ladder; lane 1, reagent control; lane 2, positive control; lanes 3 to 19, isolates from cat feces. (B) Primers specific for C. upsaliensis amplified 16S rRNA products from 8 of 17 isolates from cat feces. Lane MW, 1-kb DNA ladder; lane 1, reagent control; lane 2, positive control; lanes 3 to 19, isolates from cat feces. (C) Primers specific for C. jejuni amplified hippuricase gene PCR products from 3 of 17 cat isolates. Lane MW, 1-kb DNA ladder; lane 1, reagent control; lane 2, positive control; lanes 3 to 19, isolates from cat feces.

Restriction fragment length polymorphism (RFLP) analysis of Helicobacter species PCR products from cat isolates. The 1.2-kb products from Helicobacter genus-specific PCR analysis of 19 cat isolates were digested by AluI. Five patterns were observed. Three patterns grouped taxonomically with H. bilis or Helicobacter ("Flexispira") taxon 8; one was similar to H. canis, and the remaining pattern matched that of a novel Helicobacter sp. isolated from a woodchuck (21) (Fig. 3 and 4).

View larger version (82K): [in this window] [in a new window]   FIG. 3.   Products (1.2 kb) of PCR using Helicobacter genus-specific primers were digested by AluI and analyzed by electrophoresis on 6% Visigel matrix. Five patterns were observed. Lane MW, 100-bp DNA ladder; lane 1, Helicobacter ("Flexispira") taxon 8 (ATCC 49317); lane 2, MIT 98-90; lane 3, MIT 95-1850-65; lane 4, MIT 95-513-27; lane 5, H. bilis ATCC 51630; lane 6, MIT 95-234-6; lane 7, H. canis cat isolate (11); lane 8, MIT 95-1114-42; lane 9, MIT 95-1114-46; lane 10, MIT 94-55-4; lane 11, Helicobacter sp. woodchuck isolate (19); lane 12, MIT 95-513-29; lane 13, MIT 94-2635-37; lane 14, MIT 98-1705-4.

View larger version (55K): [in this window] [in a new window]   FIG. 4.   Neighbor-joining phylogenetic tree for Helicobacter spp. isolated from feces of cats and reference Campylobacter and Helicobacter species. The scale bar represents phylogenetic distance as estimated using the Jukes Cantor correction. Distances can be determined by adding the lengths of all of the horizontal lines connecting any two species. GenBank accession numbers appear in brackets. The 1,200-bp partial sequences (marked as such) for which accession numbers are not provided were not deposited in GenBank and are available from the corresponding author.

Analysis of 16S rRNA sequences. Full 16S rRNA sequences were determined for isolates MIT 95-234-6 and MIT 98-1705-1. Isolate MIT 95-234-6 was identical to the sequence of the type strain of H. bilis (GenBank accession no. U18766) except that it did not contain a 187-base intervening sequence (22) in the 198-219 helix, but rather the sequence GGUUUUUC. Isolate MIT 98-1705-1 was identical to a Helicobacter sp. previously isolated from a woodchuck, MIT 98-6070 (GenBank accession no. AF333341). The eight 1,200-base partial sequences clustered into three groups. Clones MIT 95-513-29 and MIT 94-2635-37 differed by 1 base from MIT 98-1705-1. Clones MIT 95-54-4, MIT 95-1114-42, and MIT 95-1114-46 differed from the sequence of the type strain of H. canis (GenBank accession no. L13464) by the following three base changes: G592A, A616G, and A645G (numbering according to the sequence of E. coli). Clones MIT 95-513-27, MIT 95-1850-65 and MIT 98-90 differed from the sequence of Helicobacter ("Flexispira") taxon 8 (GenBank accession no. M88138) by 0 to 3 bases. A neighbor-joining phylogenetic tree was constructed and is shown in Fig. 4.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

This study for the first time demonstrated a high prevalence of mixed infections of Campylobacter and Helicobacter species in a large number of clinically healthy cats obtained from three commercial sources located in different geographic regions. CLOs were isolated from 28% of 227 cats, and from 33% of these, mixed cultures of Campylobacter organisms and Helicobacter organisms were obtained. The Campylobacter spp. most frequently isolated from the feces of these cats were C. upsaliensis and C. helveticus; 86% of the cultures contained C. helveticus, while 47% of the cultures contained C. upsaliensis.

The high prevalence of Campylobacter spp. in laboratory-reared cats is consistent with widespread Campylobacter infection observed for domestic animals. C. upsaliensis was first isolated from the feces of healthy and diarrheic dogs in Sweden (63 of 98 [64%] of the Campylobacter strains isolated from the feces of these dogs over 2 years were C. upsaliensis) (48). We have previously documented the presence of C. upsaliensis in cat feces using biochemical characterization and DNA hybridization assays (20). In Switzerland, Campylobacter spp. were isolated from 31% of a group of diarrheic and healthy pet animals; 50% of the isolates from cats were C. upsaliensis. For cats, there was no association between Campylobacter carriage and disease, irrespective of the animals' age. For dogs older than 12 months, there was also no difference in Campylobacter carriage rate between diarrheic and healthy animals. However, 44% of the younger dogs with diarrhea shed Campylobacter species organisms in their feces, more than twice the rate observed for clinically healthy dogs (5). In the United Kingdom, 50% of 156 healthy domestic pets and laboratory animals were positive for Campylobacter spp., with 60% of the cats shedding C. upsaliensis in their feces (38). However, none of the authors of the above-mentioned four studies isolated Helicobacter spp. from the feces of these animals.

C. upsaliensis has also been isolated from the feces of children and adults with diarrhea (25, 44, 55), as well as from the blood of pediatric patients and adults with septicemia (31, 39, 44). Other extraintestinal sites from which this organism has been cultured include a breast abscess (23) and the fetoplacental tissue of an 18-week-pregnant woman who had contact with a household cat. Both isolates had similar sodium dodecyl sulfate gel protein patterns (27). There is other epidemiological evidence that suggests that C. upsaliensis may have zoonotic potential. One study of C. upsaliensis infection reported that four of seven humans infected had animal contact (44). C. upsaliensis was also isolated from a diarrheic patient and his clinically healthy dog (26).

C. helveticus was isolated in a high percentage of cats in this study, which confirmed the results of earlier studies in England, where it was cultured from the feces of healthy cats (52). However, the organism's clinical relevance for pets, if any, has not been reported.

The natural habitat of most Campylobacter spp., including C. jejuni, is the intestinal tract of warm-blooded animals, including birds (40). Campylobacter infection is transmitted to humans from animals either by fecal-oral contact or indirectly by food, milk, or water. Campylobacterosis in humans is largely a result of food-borne infection in which foods of animal origin, particularly poultry, play an important role (8). Domestic animals are common reservoir hosts for C. jejuni, and zoonotic infections have been acquired from pets, including cats with or without diarrhea (4, 8, 13; M. B. Skirrow, G. L. Turnbull, R. E. Walker, and S. E. J. Young, Letter, Lancet i:1188, 1980; A. Svedhem and G. Norkrans, Letter, Lancet i:713-714, 1980). C. jejuni has been isolated from dogs and cats housed in animal shelters in addition to being isolated from dogs and cats used in biomedical research (13, 15, 45). In our study, 4% of cats had C. jejuni in their feces. This prevalence was lower than the 10.7% previously reported for research cats, but it was higher than the 1% C. jejuni isolation rate recorded for pet cats and cats sampled at a humane society shelter (13, 29).

The Helicobacter spp. most frequently isolated from cats in this study were H. canis, Helicobacter ("Flexispira") taxon 8, and a novel Helicobacter species previously isolated from woodchucks. The novel species shared at least 96% sequence identity with all Helicobacter spp. in the GenBank database but was essentially identical to an isolate from a woodchuck (MIT 98-6070) (21). This novel species shared 97% 16S rRNA sequence homology with a mouse Helicobacter sp. isolate, MIT 94-022. H. canis has been previously reported to have been found in diarrheic cats (12), in a child with gastroenteritis (6), and in dogs with or without diarrhea (54). The organism was also isolated from the liver of a puppy with necrotizing hepatitis (17). Organisms with "Flexispira rappini " morphology isolated from a number of hosts have been divided into 10 taxa (9). For example, H. bilis was identified, by cloning and sequencing of 16S rRNA, in gall bladders of patients with chronic cholecystitis (16). Although the 16S rRNA sequences of these taxa are very similar, the RFLP patterns may be different (50). To our knowledge, organisms of the Helicobacter ("Flexispira") taxa (including H. bilis) have not been isolated from cats. However, such organisms have been cultured from the feces of three dogs and their owners, diarrheic children, and rodents (22, 33, 46, 49). They are also increasingly isolated from the blood of immunocompromised patients, including two that had a history of contact with puppies (51, 56).

In our experience, for the best recovery of CLOs, fecal samples should be placed in glycerol medium for transportation. Higher H₂ levels (5 to 10%) are required for optimal Helicobacter sp. isolation. Unfortunately, this atmosphere is not available in the commercially available diagnostic kits used for Campylobacter isolation. Identification of multiple species of microaerobic bacteria in the feces of an animal poses a diagnostic challenge, particularly when these microaerobes grow on similar media in comparable atmospheric conditions. Primary isolation of these microaerophilic bacteria may be misleading, because Helicobacter spp. may be present in smaller numbers and grow at a slower rate than Campylobacter spp. The similar phenotypic traits and biochemical profiles of these genera also complicate a diagnosis. Using Campylobacter and Helicobacter genus-specific PCR assays allowed us to distinguish between the two genera. The PCR-RFLP assay was also useful for Helicobacter sp. identification.

Investigators in South Africa have recently published results for a protocol that has been in use in their diagnostic laboratory since 1990 and that allows primary isolation of multiple species of Campylobacter and Helicobacter from the diarrheic specimens of individual children. Filtrates are plated onto antibiotic-free blood agar plates and incubated in an H₂-enriched atmosphere (32, 33). The authors not only documented an increase in the number of CLOs isolated but also were able to culture C. upsaliensis for the first time. The authors have reported a 16.2% prevalence of multiple species of CLOs based on primary isolation, biochemical characterization, and serologic confirmation. They frequently recovered between two and five species of CLOs from one stool sample, with C. jejuni (different serotypes), C. coli, C. upsaliensis, Helicobacter fennelliae, and H. cinaedi being commonly isolated (32). Further analysis using the filtration isolation technique with cat and dog feces may yield prevalence rates for mixed Helicobacter and Campylobacter infections even higher than those reported in the present study.

In summary, cats used for biomedical research were commonly colonized with intestinal Helicobacter spp. and Campylobacter spp. Accurate diagnosis of mixed infections with these bacteria may require diagnostic laboratories to incorporate PCR-based assays using Helicobacter and Campylobacter genus- and species-specific primers. This recommendation is supported by a recent study which reported improved sensitivity for PCR compared to conventional culture techniques in identifying mixed infections of Campylobacter spp in cases of human gastroenteritis (34). Although all the CLOs in this study were isolated from clinically healthy cats, some of these species have been linked with diarrheal diseases in humans and animals. The zoonotic importance of intestinal cocolonization with Campylobacter and Helicobacter, as well as their importance in causing disease in cats, other animals, and humans, requires further studies.