Species-specific DNA probes for Campylobacter species isolated from pigs with proliferative enteritis

Cloned, chromosomal DNA probes from porcine isolates of Campylobacter hyointestinalis and C. mucosalis were developed for the detection and identification of these putative swine enteric pathogens. High molecular weight chromosomal DNA from each species was used to construct genomic libraries in plasmids. Recombinants were selected which hybridized strongly to the homologous organism, but not to any other species of Campylobacter. Species-specific recombinants were labeled with phosphorus-32 and tested for sensitivity by dot blot hybridization to various dilutions of DNA and bacteria from each swine species, including C. hyointestinalis, C. mucosalis, C. coli and C. jejuni. Specificity was tested by hybridizing these probes against various strains of C. hyointestinalis or C. mucosalis, and against reference strains of all other described Campylobacter species. A C. hyointestinalis-specific probe and a C. mucosalis-specific probe were identified which were capable of detecting 1 ng of DNA or 10(4) cfu by bacterial spot blotting on nylon membranes. These probes hybridized to intestinal mucosal scrapings containing C. hyointestinalis and C. mucosalis obtained from pigs with proliferative enteritis, but not to material from normal pigs. Thus, cloned, chromosomal DNA probes may be useful in the detection and identification of bacteria involved in swine proliferative enteritis.     

Porcine proliferative enteritis: serological, microbiological and pathological studies from three field epizootics.

Three outbreaks of porcine proliferative enteritis were evaluated clinically, pathologically, microbiologically and serologically. The disease was characterized by a chronic intermittent diarrhea. Pathological lesions included a thickened, turbid ileum with the microscopic appearance of proliferating ileal crypt epithelial cells. Comma shaped intracytoplasmic organisms were observed in the apical portions of the proliferating crypt epithelial cells with a Warthin-Starry silver stain. Microbiologically, both Campylobacter sputorum subspecies mucosalis and Campylobacter hyointestinalis, were cultured from ileal specimens of seven pigs with lesions of porcine proliferative enteritis. Microagglutination antibody titers were determined on sera from 12 of 14 pigs with porcine proliferative enteritis and on sera from 91 clinically normal swine. Pigs with porcine proliferative enteritis had a low antibody titer to subspecies mucosalis that ranged from 1-3 with a mean of 2.17. A varied C. hyointestinalis titer from 3-7 with mean of 4.83 was determined. Titers to either subspecies mucosalis and C. hyointestinalis were higher in non-porcine proliferative enteritis pigs. The results indicate that the presence of a positive titer to either C. hyointestinalis or subspecies mucosalis in swine is not indicative of clinical disease. The isolation of C. hyointestinalis from diseased ileal specimens (porcine proliferative enteritis) confirms previous reports implicating this agent in the disease.     

Campylobacter sputorum subsp mucosalis and Campylobacter hyointestinalis infections in the intestine of gnotobiotic pigs

At 4 days of age, 7 gnotobiotic pigs were orally inoculated with broth cultures of both Campylobacter sputorum subsp mucosalis (CSM) and Campylobacter hyointestinalis (CH). One pig was killed and evaluated each week for 7 weeks. Forty-eight hours after inoculation, CH and CSM were recovered from the feces of the pigs; thereafter, only CH was recovered. Organisms morphologically typical of Campylobacter sp were observed on the mucosal surface and on the crypt epithelial cells of the ileum, cecum, and colon from post-inoculation weeks (PIW) 2 through 7. Bacteria were clustered around the surface opening of goblet cells in pigs at PIW 6 and 7. Crypt epithelial cell proliferation and intracellular bacteria were not seen, except in 1 pig (killed at PIW 7) in which intracellular bacteria were seen only in the cecum. Therefore, CSM and CH did not induce porcine proliferative enteritis in gnotobiotic pigs.     

Reproduction of proliferative enteritis in gnotobiotic pigs

Gnotobiotic pigs dosed orally with filtrates (0.8 and 0.65 micron) of intestinal mucosa from a pig affected by proliferative haemorrhagic enteropathy developed lesions of proliferative enteritis, affecting mainly the ilea. Other piglets dosed with filtrates of affected mucosa from the same source and from other proliferative haemorrhagic enteropathy or intestinal adenomatosis mucosae, did not develop lesions. All inocula contained numerous campylobacter-like organisms evident in stained smears, Campylobacter coli and C mucosalis. C coli colonised the intestines of all the pigs, C hyointestinalis (which was not detected in the inocula) did so in some affected and unaffected pigs while C mucosalis was not recovered from any of the intestines. Although other explanations are possible the number and viability of the intracellular campylobacter-like forms is likely to be the critical factor in infectivity. In affected intestines the crypts were colonised by campylobacter-like organisms, and their attachment and entry into enterocytes was associated with cellular proliferation. Immunofluorescence reactions suggested that the intracellular campylobacter-like organisms were antigenically distinct from the known Campylobacter species. It is possible, therefore, that porcine proliferative enteritis is caused by a further unidentified Campylobacter species, or that there is a marked antigenic change of C hyointestinalis or C coli on entry into porcine enterocytes.     

Behaviour of Campylobacter sputorum subspecies mucosalis in gnotobiotic pigs

Gnotobiotic pigs were dosed orally with Campylobacter sputorum subspecies mucosalis, either alone, or combined with rotavirus or non-pathogenic Escherichia coli and Streptococcus bovis to study the behaviour of C s mucosalis in defined conditions, to assess intracellular parasitism of enterocytes by C s mucosalis, and if possible to establish an experimental model of porcine intestinal adenomatosis. C s mucosalis colonised the gut of gnotobiotic pigs, persisting for up to 47 days after infection, but did not induce adenomatosis. Despite evidence of limited penetration of the mucosa up to two days after infection, the majority of C s mucosalis remained in the gut lumen. Rotavirus did not enhance invasion of enterocytes by C s mucosalis. The presence of E coli and S bovis caused an increase in the total numbers of C s mucosalis in the gut, but did not affect their distribution. Thus C s mucosalis was largely non-pathogenic in gnotobiotic pigs.     

Isolation, characterization, and molecular cloning of cryptic plasmids isolated from Edwardsiella ictaluri

Fifty-five isolates of Edwardsiella ictaluri were examined for the presence of plasmid DNA by a rapid alkaline extraction procedure. All 49 isolates from channel catfish and a single isolate from Bengal danio carried 2 plasmids with molecular masses of approximately 3.2 and 3.7 megadaltons (Mdal). Five E ictaluri isolates from other fish contained 1 to 3 plasmids, which had molecular masses ranging from 2.5 to 45 Mdal. The 2 plasmids (3.2 and 3.7 Mdal) from the type strain of E ictaluri (ATCC 33202) were ligated into pUC19 cloning vectors, and restriction endonuclease maps of each insert were prepared.     

Linkage of genes for heat-stable enterotoxin, drug resistance, K99 antigen, and colicin in bovine and porcine strains of enterotoxigenic Escherichia coli

Linkages among genes for production of Escherichia coli heat-stable enterotoxin (ST), drug resistance, K99 antigen, and colicin were investigated in 2 bovine and 3 porcine strains of enterotoxigenic E coli. In conjugation experiments, all 5 isolates transferred enterotoxigenicity and the ability to produce K99; 4 of the 5 isolates also transferred antibiotic resistance markers, and 3 colicinogenic strains transmitted the ability to produce colicin. In 2 of the 3 colicin-producing strains, the genes for colicin were located along with those for K99 and ST on a single plasmid. One of these 2 strains transferred the genes for tetracycline resistance and production of both mouse-active ST (STa) and mouse-inactive ST, whereas the other transmitted the gene(s) for STa only. Transformation studies with the 3rd strain revealed that the K99 determinant resided on a 22-megadalton (Mdal)R plasmid, and that STa and colicin production were on a 65-Mdal plasmid. Analysis of the plasmids from the transformation experiments revealed that the larger plasmid was conjugative and the smaller plasmid was nonconjugative; stable cointegrate formation occurred between these 2 plasmids. Genetic information coding for the production of STa and K99 were also present on a single plasmid of approximately 80 Mdal in both noncolicinogenic strains.     

Characterization of Campylobacter lanienae from pig feces

To isolate Campylobacter spp., the feces of healthy cattle, pigs, and broilers were examined between June 1999 and January 2000. Campylobacter lanienae strains were isolated from the feces of healthy pigs, but not from the feces of cattle or broilers. In six C. lanienae isolates, there was only 21-38% DNA-DNA homology to Campylobacter hyointestinalis subsp. lawsonii strain NCTC 12901. Thus, the primary host of C. lanienae is likely to be the pig and C. lanienae appears to be a species distinct from C. hyointestinalis subsp. lawsonii. In addition, an intervening sequence of 226 bp in the 16S rRNA gene was found in four isolates.     

Southern blot analysis of strain variation in Campylobacter mucosalis.

A panel of three DNA probes were derived at random from a genomic DNA library of Campylobacter mucosalis strain E8384-4. Each probe hybridized specifically to C. mucosalis DNA from bacteria fixed to nylon membranes. The probes did not hybridize to DNA from other Campylobacter species or to other bacteria even at 100-fold higher amounts. Each probe hybridized to all of 24 isolates of C. mucosalis which had been collected over time from different geographic locations. Southern blot analysis of selected C. mucosalis isolates was carried out to determine if the probes would be useful for differentiating among various isolates. It indicated that restriction fragment length polymorphisms (RFLPs) exist at the loci identified by our probes. These differences were used to characterize seven C. mucosalis isolates recovered from pigs in Minnesota. The results suggest that RFLP analysis may be a useful tool for epidemiological studies of C. mucosalis.     

Pathogenicity of Campylobacter jejuni and Campylobacter coli strains in the pregnant guinea pig model

Pathogenicity of 17 Campylobacter isolates for pregnant guinea pigs was investigated. Of 14 isolates, 12 (86%) produced rates of abortion ranging from 13% to 87%. Two isolates did not produce abortion. Reference strains of C fetus subsp venerealis produced abortion in 60% to 87% and C fetus subsp fetus produced abortion in 60% of the guinea pigs. Inoculated organisms were recovered from uterus, blood, liver, kidney, spleen, and gallbladder of the guinea pigs at rates as high as 83% for 2 ovine isolates and as low as 13% for 2 bovine and 1 human isolates. Most isolations were from the uterus. Two avian isolates were not recovered. Within the C jejuni and C coli group, the ovine and the human isolates appear to be more pathogenic. Swine, bovine, and avian isolates were less pathogenic. Seemingly, the pregnant guinea pig was a suitable and practical model for evaluating the pathogenicity of Campylobacter organisms, regardless of their host of origin.     

Use of embryonating eggs for isolation of Campylobacter species from intestines of swine with proliferative enteritis

Intestinal tissues from swine affected with proliferative enteritis were ground, filtered through a 0.65-micron pore membrane filter, diluted, and injected into 7-day-old embryonated hens' eggs via the yolk sac. At 2, 4, and 7 days later, yolk sac swab specimens taken from live embryos were cultured for Campylobacter species. Campylobacter hyointestinalis was recovered from eggs injected with tissues of swine with acute hemorrhagic proliferative enteritis at dilutions up to 10(-4). Campylobacter mucosalis was recovered from eggs injected with tissues of swine with chronic proliferative enteritis at dilutions up to 10(-6). Campylobacter coli was recovered from several specimens without lesions of proliferative enteritis and also from some specimens with lesions of proliferative enteritis. Two previously undescribed hemolytic Campylobacter species designed as hemolytic number 1 and hemolytic number 2 were recovered from normal and experimentally inoculated swine tissues. Few contaminating organisms grow in eggs and these were usually recovered at dilutions of 10(-2) or less. Recovery of Campylobacter species by use of these techniques was seldom successful in tissues stored at -70 C for more than 6 months.     

Relation of plasmids to virulence and other properties of salmonellae from avian sources

A collection of 185 isolates of 34 serovars of Salmonella from avian sources was examined for plasmids, drug resistance, biochemical properties, serum resistance, and virulence. No serovars other than S. enteritidis, S. typhimurium, and S. heidelberg showed evidence of serovar-associated plasmids. All S. enteritidis isolates carried a single plasmid of 36 Mdal and were resistant to guinea pig serum; one strain that was tested was virulent. Of 27 isolates of S. typhimurium, 11 possessed a 60-Mdal plasmid and 17 harbored a 2.3-Mdal plasmid. Among isolates of S. heidelberg, 21 of 24 carried a 2.2-Mdal plasmid. The only biochemical property that varied was fermentation of inositol, which tended to be related to serovar. Of 172 isolates, 54 were resistant to at least one drug. Multiple drug resistance was usually associated with R plasmids, and transmissible plasmids that encoded resistance to chloramphenicol and gentamicin were demonstrated. Of 117 isolates tested, 43 were resistant to guinea pig serum. Resistance appeared to be a characteristic of isolates rather than serovar and could not be related to plasmids. Twenty-five isolates highly resistant to guinea pig serum were all susceptible to the bactericidal action of chicken serum. In tests for virulence using intraperitoneally (i.p.) and orally inoculated Balb/c mice and day-old chicks, only i.p.-inoculated chicks proved useful in demonstrating large differences among isolates: LD50's ranged from 10(0) to 10(8).     

Serological diagnosis of the porcine proliferative enteropathies: implications for aetiology and epidemiology

Campylobacter mucosalis and C hyointestinalis have been associated with the proliferative enteropathies of pigs. An examination of the antibody response to these organisms and to the intracellular campylobacter-like organism was undertaken. Antibody to the campylobacter-like organism was predominantly IgM, short lived, and could be detected by an immunofluorescence test using bacteria released from lesions as antigen. The majority (75 per cent) of pigs with proliferative enteropathy at necropsy were antibody positive and a small number (4 per cent) of pigs in which lesions were not observed were found to have antibody. Antibody appeared to be correlated with the presence of lesions rather than with exposure to infection and was independent of the presence of antibody to C mucosalis or C hyointestinalis. In natural outbreaks of the disease antibody to the campylobacter-like organism was more prevalent than clinical signs in the affected animals.     

Immuno-histochemical and -cytochemical evidence suggesting the presence of Campylobacter jejuni and Campylobacter coli in cases of porcine intestinal adenomatosis

Antisera against a number of Campylobacter species were used in immuno-histochemical and -cytochemical studies on cases of porcine intestinal adenomatosis. Avidin-biotin-complex (ABC) and streptavidin immunoperoxidase methods were used on formalin-fixed, paraffin-embedded and frozen sections. Protein A gold method was used on formaldehyde fixed and frozen sections for immuno-cytochemistry. The antisera used were raised in rabbits by subcutaneous or intravenous injection of living or formalin treated organisms. Anti-sera against different serotypes of the thermotolerant, catalase positive campylobacters, Campylobacter jejuni and Campylobacter coli, gave positive reactions in the immuno-histochemical studies. The staining was found in intestinal epithelial cells both in the ileum and in the colon and was restricted to the apical cytoplasm of adenomatous epithelial cells. The staining had a granular pattern, the positive structures sometimes having the shape of Campylobacter. Epithelial cells in areas with normal differentiation of goblet cells did not stain. In contrast, no staining resulted with antisera against Campylobacter sputorum subsp. mucosalis and Campylobacter hyointestinalis. Immuno-cytochemistry, using antisera against Campylobacter jejuni, showed that the positive staining in altered epithelial cells were restricted to intracellular organisms having a structure resembling Campylobacter spp.     

Epidemiology of ovine Campylobacter infection determined by numerical analysis of electrophoretic protein profiles

The relationship of 50 Campylobacter strains isolated from aborted ovine foetuses, and the faeces of sheep, cattle and chickens were determined by numerical analysis of electrophoretic (SDS-PAGE) protein profiles. Comparison of protein patterns by numerical methods revealed differences between C. fetus ssp. fetus, C. jejuni, and C. coli strains as well as heterogeneity among isolates from different outbreaks. Isolates from each farm produced a distinct cluster and flocks from different locations were found to be infected with relatively different strains. In most cases, protein patterns of ovine foetal isolates were very similar to those of ovine faecal isolates. Ovine isolates of C. fetus ssp. fetus, C. jejuni and C. coli gave similar protein patterns to the corresponding Campylobacter species isolated from cattle or chicken, on the same farm. Thus, it was concluded that certain protein types of ovine Campylobacter strains were more likely associated with local areas, and Campylobacter strains causing ovine abortions are distributed in the environment more widely than assumed to date.     

Production and some properties of cytotoxins produced by Campylobacter species isolated from proliferative enteropathy in swine.

Cytotoxin production by Campylobacter species isolated from proliferative enteropathy in swine was examined. Twenty-one of 29 strains of C. hyointestinalis, 10 of 27 strains of C. mucosalis and 10 of 10 strains of C. coli were cytotoxin positive. By the gel filtration chromatography of C. hyointestinalis culture filtrate, cytotoxin activities were observed in two peaks (fraction I and fraction II). Most of the cytotoxic activities lay in fraction I, which is heat-labile, trypsin-sensitive and the molecular weight was estimated at 40,000. On the other hand, fraction II cytotoxin was heat-stable, trypsin-insusceptible and molecular weight was approximately 1,000.     

Possible relationship of proliferative enteritis in pigs and hamsters

Three- to six-week-old hamsters were orally inoculated with broths containing one of the following cultures: Campylobacter mucosalis; C. hyointestinalis; C. coli; C. jejuni, all of porcine proliferative enteritis origin, or else C. jejuni of hamster origin. Hamsters given the last of those organisms were shown to have colonisation of their intestines by C. jejuni and 36 of 40 developed an acute enteritis. Mild hyperplasia of enterocytes in ileal crypts was evident in one hamster 2 days after it was given C. coli. No other lesions were detected. Further 3-week-old hamsters were orally inoculated with homogenised intestinal mucosa collected from 4 pigs (A-D) affected by proliferative enteritis. Lesions of proliferative enteritis were detected in 7 of 41 hamsters necropsied 10-21 days after being dosed with mucosas B or D. Marked hyperplasia of ileal enterocytes, associated with numerous intracellular Campylobacter-like organisms, were invariably detected in experimentally affected hamsters. No particular Campylobacter sp. was consistently isolated. None of the controls had demonstrable lesions. The results suggested that cross-species transmission of proliferative enteritis was possible from pigs to hamsters. Therefore a common initiating or aetiological agent may be present. No specific organism was identified as filling this role by inoculation of hamsters with pure cultures.     

Isolation and association of Escherichia coli AIDA-I/STb, rather than EAST1 pathotype, with diarrhea in piglets and antibiotic sensitivity of isolates.

To identify emerging Escherichia coli that have the potential to cause diarrhea in pigs, the prevalence of E. coli pathotypes was determined among 170 and 120 isolates from diarrheic and nondiarrheic piglets, respectively. The isolates were tested for F4, F5, F6, F18, and F41 fimbriae, for E. coli attaching and effacing (EAE), porcine attaching and effacing-associated (Paa), and adhesin involved in diffuse adherence (AIDA-I) factors, for LT, STa, STb, and enteroaggregative heat-stable (EAST1) enterotoxins, and for Shiga toxins (Stxl, Stx2, and Stx2e), using DNA hybridization and polymerase chain reaction. All isolates were O-serotyped and tested for antibiotic resistance against 10 drugs. Seventeen different pathotypes, accounting for 40.0% of the isolates, were recovered from diarrheic piglets. The main pathotypes included EAST1 (13.5%), F4/LT/STb/EAST1 (6.5%), AIDA-I/STb/EAST1 (4.1%), F5/STa (2.9%), EAE/EAST1 (2.9%), and AIDA-I/F18 (2.3%). Only 3 pathotypes, EAE (11.7%), EAST1 (10.8%), and EAE/EAST1 (3.3%), were recovered from nondiarrheic piglets. Paa factor was detected in 8.8% and 7.5% of isolates from diarrheic and nondiarrheic piglets, respectively, and always was associated with other virulence determinants. Overall, 22.9% of isolates from diarrheic piglets appeared to be enteropathogens: enterotoxigenic E. coli (11.7%), enteropathogenic E. coli (3.5%), and E. coli isolates (3.0%) for which none of the above adherence factors was detected. Pathotypes AIDA-I/STb/EAST1 and AIDA-I/STb were isolated only from diarrheic piglets and accounted for 4.7% of isolates. Strains of these pathotypes induced diarrhea when inoculated into newborn colostrum-deprived pigs, in contrast to an isolate positive only for EAST1, which did not induce diarrhea. Antibiotic sensitivity test showed that isolates of the AIDA-I/STb/EAST1 and AIDA-I/STb pathotypes were the only strains sensitive to enrofloxacin, gentamicin, neomycin, and trimethoprim-sulfamethoxazole. This study showed that at least 20.5% of isolates from diarrheic piglets appeared to be associated with AIDA-I/STb pathotype and that EAST1 pathotype is probably not an important marker for diarrhea in piglets.     

Experimentally induced porcine proliferative enteritis in specific-pathogen-free pigs

Thirty-three, 10-week-old, specific-pathogen-free pigs were randomly allotted to 3 treatment groups: group 1--intragastrically given homogenized intestinal mucosa (crude inoculum) from pigs with naturally occurring proliferative enteritis; group 2--given cultures of Campylobacter sputorum subsp mucosalis; and group 3--controls. One pig from each group was killed 4, 7, 10, 14, 18, 21, 24, 28, 31, 36, and 38 days after inoculation. The earliest intestinal lesion observed in groups 1 and 2 was leukocytic exudate within crypt lumina and focal inflammation of the surrounding lamina propria. The lesions occurred primarily over ileal aggregated lymphoid nodules (Peyer's patches). These changes were followed by focal proliferation of immature crypt epithelial cells and infiltration of increasing numbers of macrophages into the lamina propria. Campylobacter sp-like organisms were observed within the cytoplasm of affected epithelial cells by light and electron microscopies. Lesions progressed to diffuse crypt cell proliferation, elongation of crypts, and loss of villi. Mucosal necrosis was not a prominent feature.     

Thermophilic Campylobacter from fecal samples of animals and their behavior in the ganglioside (GM1)-ELISA, Vero cell test and salt aggregation test

From 397 fecal specimens from apparently healthy and from diarrheic pigs, dogs, cats and cattle 59 strains (= 15%) of thermophilic Campylobacter (C.) spp. were isolated by culture. 39 strains were identified as C. coli and 18 as C. jejuni whereas 2 isolates could not be classified. None of the strains was found to be positive for cytotoxic enterotoxin in the GM1-ELISA. In the Vero-cell test 5 isolates showed a cytotoxic effect. The salt aggregation test (SAT) for indicating cell surface hydrophobicity was positive with 24 strains (5 C. jejuni, 19 C. coli). A correlation of isolation results with clinical manifestation could not be observed.