S.L. Oxberry

The prevalence of Serpulina pilosicoli in humans and domestic animals in the Eastern Highlands of Papua New Guinea

In a survey of five villages in the Eastern Highlands of Papua New Guinea, Serpulina pilosicoli was isolated from rectal swabs from 113 of 496 individuals (22.8%). Colonization rates ranged from 22.6-30.1% in four of the villages but was only 8.6% in the other village. In comparison colonization was demonstrated in only 5 of 54 indigenous people (9.3%) and none of 76 non-indigenous people living in an urban environment in the same region. Colonization did not relate to reported occurrence of diarrhoea, age, sex, or length of time resident in a village. A second set of 94 faecal specimens was collected from 1 village 6 weeks after the first set. S. pilosicoli was isolated from 27 of 29 individuals (93.1%) who were positive on the first sampling and from 7 of 65 individuals (10.8%) who previously were negative. In this case, isolates were significantly more common in watery stools than in normal stools. The annual incidence of infection in the village was calculated as 93.6%, with an average duration of infection of 117 days. S. pilosicoli could not be isolated from any village pig (n = 126) despite its confirmed presence in 17 of 50 commercial pigs (34.0%) sampled at a local piggery. Four of 76 village dogs (5.3%) and 1 of 2 village ducks were colonized with S. pilosicoli, suggesting the possibility of cross transmission between humans and animals.

Serpulina pilosicoli, waterbirds and water: potential sources of infection for humans and other animals

Serpulina pilosicoli was isolated from 8 of 43 (19%) faecal specimens obtained from feral waterbirds sampled around a small lake at Perth Zoological Gardens, Western Australia, and from 3 of 7 (43%) samples of the lake water. The organism was only isolated from 1 of 204 (0.5%) samples from captive birds and animals in the zoological collection. Multilocus enzyme electrophoresis analysis of the isolates showed that they were genetically diverse, and none had identical electrophoretic profiles as those previously obtained from human beings, dogs, pigs and other avian species. To determine the survival time of S. pilosicoli in water, cells of strain 1648 were seeded into lake and tap water, and incubated at 4, 25 and 37 degrees C. The organism could be recultured from lake water for up to 66 days at 4 degrees C, and for 4 days at 25 degrees C. A healthy human volunteer who drank water seeded with S. pilosicoli strain Wes B became colonized, and developed abdominal discomfort and headaches. Contamination of water by faeces may represent a source of S. pilosicoli infection for both humans and animals.

Population genetic analysis of Serpulina pilosicoli and its molecular epidemiology in villages in the Eastern Highlands of Papua New Guinea

The population genetics of Serpulina pilosicoli and its molecular epidemiology in villages in the Eastern Highlands province of Papua New Guinea were investigated. Multilocus enzyme electrophoresis (MLEE) was used to analyse 164 isolates from humans and animals. These were divided into 33 electrophoretic types (ETs), four of which contained 65% of the isolates. The mean genetic diversity (n = number of ETs) for 145 human isolates was 0.18, and the mean number of alleles at five polymorphic loci was 2.6. The species appeared to be recombinant, as there was a lack of linkage disequilibrium, and 25% of all the possible combinations of alleles was present in the population. PFGE analysis using the enzymes M/ul and Sa/l divided 157 of the isolates into 99 PFGE types, demonstrating the existence of considerable strain diversity in a geographically restricted area. The two techniques were in excellent agreement; however, PFGE was more discriminatory for strain typing than was MLEE. Nine out of 19 (47.4%) culture-positive individuals were colonized by the same PFGE type of S. pilosicoli when retested after 6 weeks. For three individuals, the PFGE profiles of the second isolate differed from the first in only one or two DNA bands, while the other seven individuals were colonized with distinct PFGE types on each occasion. In two cases, strains with the same PFGE pattern were isolated from humans and dogs, suggesting that cross-species transmission of S. pilosicoli may occur naturally and that the infection can be zoonotic.

Influences of diet and vaccination on colonisation of pigs by the intestinal spirochaete Brachyspira (Serpulina) pilosicoli.

The purpose of this study was to determine whether methods used to control swine dysentery (SD), caused by the intestinal spirochaete Brachyspira (Serpulina) hyodysenteriae, would also be effective in controlling porcine intestinal spirochaetosis (PIS) caused by the related spirochaete Brachyspira (Serpulina) pilosicoli. Weaner pigs in Groups I (n=8) and II (n=6) received a standard weaner pig diet based on wheat and lupins, whilst Group III (n=6) received an experimental diet based on cooked white rice and animal protein. Pigs in Group II were vaccinated intramuscularly twice at a 3-week-interval with a formalinised bacterin made from B. pilosicoli porcine strain 95/1000 resuspended in Freunds incomplete adjuvant. Eleven days later pigs in all groups were infected orally with 10(10) cells of strain 95/1000 on three successive days. One control pig in Group I developed acute diarrhoea, and at post-mortem had a severe erosive colitis with end-on attachment of spirochaetes to the colonic epithelium. All other pigs developed transient mild diarrhoea and had moderate patchy colitis at post-mortem 3 weeks later. B. pilosicoli was isolated from the faeces of all pigs, except for one fed rice, and was isolated from the mesenteric nodes of three pigs from Group I and from one vaccinated pig in Group II. Consumption of the rice-based diet, but not vaccination, delayed and significantly (p<0.001) reduced the onset of faecal excretion of B. pilosicoli after experimental challenge. Vaccination induced a primary and secondary serological response to B. pilosicoli, as measured using sonicated whole cells of strain 95/1000 as an ELISA plate coating antigen. Antibody titres in the vaccinated pigs then declined, despite intestinal colonisation by B. pilosicoli. Both groups of unvaccinated animals also failed to develop a post-infection increase in circulating antibody titres.

Potential for zoonotic transmission of Brachyspira pilosicoli.

To the Editor: Anaerobic intestinal spirochetes of the genus Brachyspira colonize the large intestine (1). Most Brachyspira species have a restricted host range, whereas Brachyspira (formerly Serpulina) pilosicoli colonizes a variety of animal and bird species and humans. B. pilosicoli is an important colonic pathogen of pigs and chickens (2). It occurs at high prevalence rates in humans in developing countries and in male homosexuals and HIV-positive persons in industrialized countries (3). Its potential as a human pathogen was emphasized after its identification in the bloodstream of a series of debilitated persons (4).

Antimicrobial susceptibility testing of Australian isolates of Brachyspira hyodysenteriae using a new broth dilution method

The antimicrobial susceptibilities of 76 field isolates of Brachyspira hyodysenteriae from different states of Australia were tested in a newly developed broth dilution procedure. The antimicrobial agents used were tiamulin, valnemulin, tylosin, erythromycin, lincomycin and clindamycin. The results from the broth dilution susceptibility testing of 39 of the isolates were compared with results obtained for the same isolates using the agar dilution method. Amongst the isolates tested by broth dilution, 17 were from three farms and had been collected over a number of years. Their pulsed field gel electrophoresis pattern previously had been determined. The broth dilution technique was simple to use, less labor intensive than agar dilution, and gave clear end points. The results obtained using the two methods generally corresponded well, although in a few cases the MIC obtained by broth dilution were lower than those with agar dilution. For the 76 isolates tested by broth dilution, the MIC(90) (mg/l) was: tiamulin, 1; valnemulin, 0.5; tylosin>256; erythromycin>256; lincomycin, 64 and clindamycin, 16. Only minor differences in susceptibility patterns were found amongst isolates from different Australian states. Over all the isolates, and also amongst the isolates obtained from different years on the three farms, there was no trend for the susceptibility of the isolates to alter with time.

Development and evaluation of polymerase chain reaction tests as an aid to diagnosis of swine dysentery and intestinal spirochaetosis

Polymerase chain reaction (PCR) tests were established for detection of Serpulina hyodysenteriae, the agent of swine dysentery, and S. pilosicoli, the agent of intestinal spirochaetosis. Both reactions were specific when tested with DNA from 107 strains of various intestinal spirochaetes. For diagnostic use, faeces were plated to selective medium, and diatomaceous earth extraction used to obtain DNA prior to PCR. This procedure detected 103–104 cells of either organism seeded into 0·2 g of faeces. When applied to 63 samples from pigs of eight piggeries naturally infected with either S. hyodysenteriae or S. pilosicoli, both PCRs were specific, more rapid, and detected more positive samples than did routine faecal culture and isolation.

Evidence for Serpulina hyodysenteriae being recombinant, with an epidemic population structure

The population structure of Serpulina hyodysenteriae was investigated using multilocus enzyme electrophoresis. A total of 231 isolates were divided into 50 electrophoretic types (ETs), with a mean genetic diversity of 0.29 for the number of ETs and 0.23 for the number of isolates. Subsets of isolates from two Australian states (71 isolates from Victoria and 68 isolates from Queensland) exhibited as much genetic variation as the entire collection. The calculated index of association (IA) for the number of ETs (0.29 +/- 0.17) was not significantly different from zero, and hence provided evidence for the occurrence of significant genetic recombination accounting for the observed variation between strains. In contrast, the IA for the number of isolates (3.93 +/- 0.03) was significantly different from zero, with seven of the 50 ETs (ETs 4, 6, 13, 14, 20, 33 and 35) containing 51% of all the isolates. Even when multiple isolates from the same farm were removed from the analysis, the IA value for the number of isolates remained significantly greater than zero (IA 9.87 +/- 0.04), indicating that it was not biased by their inclusion. The results suggest that S. hyodysenteriae has an epidemic population structure.

Analysis of Serpulina hyodysenteriae strain variation and its molecular epidemiology using pulsed-field gel electrophoresis

Pulsed-field gel electrophoresis (PFGE) was applied as a molecular typing tool for the spirochaete Serpulina hyodysenteriae, the agent of swine dysentery. Analysis of a collection of 40 mainly Australian isolates, previously characterized by other methods, divided these into 23 PFGE types. This confirmed that there are many strains of the spirochaete in Australia. PFGE was more discriminatory for strain typing than both multilocus enzyme electrophoresis and serotyping. It had similar discriminatory power to restriction endonuclease analysis, but the results of PFGE were easier to interpret. When applied to 29 isolates collected from 4 farms over periods of up to 8 years, 2 PFGE patterns were found on 3 farms, and a single pattern on the other. In each case a new strain had apparently emerged as a variant of an original parent strain. PFGE was found to be a powerful technique for investigating the molecular epidemiology of swine dysentery outbreaks.

Epidemiological studies of Brachyspira pilosicoli in two Australian piggeries

The epidemiology of infection with the intestinal spirochaete Brachyspira pilosicoli within pig herds is incompletely understood. To investigate this further, cross-sectional and cohort studies were undertaken on two piggeries. Faeces were subjected to selective culture, and DNA was extracted from growth on the primary media and amplified by polymerase chain reaction (PCR). On one farm, samples from other animal species and the environment were also examined. Isolates were subjected to multilocus enzyme electrophoresis (MLEE) and pulsed field gel electrophoresis (PFGE). The prevalence on farm A (>2000 sows) was 2.4% (95% confidence interval (CI): 0.3, 4.4%). Infection was largely confined to grower/finisher pigs. The six isolates of B. pilosicoli recovered belonged to a single MLEE electrophoretic type (ET) and a single PFGE type. On piggery B, an 80-sow unit located on a research farm, the prevalence amongst growers and finishers was 12.2% (95% CI: 4.7, 19.6%). There was also evidence that weaners were being infected. Ten isolates obtained were genetically heterogeneous, being divided into six ETs and seven PFGE types. One of four isolates in one ET had an identical PFGE type to those on piggery A, and may have been introduced to piggery B in stock from piggery A. On farm B, B. pilosicoli was also detected by PCR in chickens, effluent pond water and wild ducks on the pond. An isolate from the pond belonged to the same ET as one from a pig, whereas the duck isolates were distinct. This study demonstrates the complex epidemiology of B. pilosicoli infections in piggeries.