Passive protection of segmented swine colonic loops against swine dysentery

Swine-ligated loops were used to demonstrate passive protection against swine dysentery. Loops inoculated with immune sera containing complement and with homologous Treponema hyodysenteriae were normal at necropsy. Loops inoculated with heat-inactivated immune sera and heterologous T hyodysenteriae were not protected. Loops inoculated with heat-inactivated immune sera and homologous T hyodysenteriae were partially protected. Positive control loops inoculated with isolate B204 (88%) or B234 (44%) T hyodysenteriae and normal sera developed lesions typical of swine dysentery, whereas negative control loops inoculated with nonexposed sera only were normal.     

The distribution of bmpB, a gene encoding a 29.7 kDa lipoprotein with homology to MetQ, in Brachyspira hyodysenteriae and related species

The distribution of the bmpB gene encoding BmpB, a 29.7 kDa outer membrane lipoprotein of the intestinal spirochaete Brachyspira hyodysenteriae, was investigated. Using PCR, the gene was detected in all the 48 strains of B. hyodysenteriae examined and in Brachyspira innocens strain B256T, but not in 11 other strains of B. innocens nor in 42 strains of other Brachyspira spp. The gene was sequenced from B. innocens strain B256T and from 11 strains of B. hyodysenteriae. The B. hyodysenteriae genes shared 97.9-100% nucleotide sequence similarity and had 97.5-99.5% similarity with the gene of B. innocens strain B256T. Southern hybridisation indicated that bmpB was present on a 1.9 kb HindIII fragment of the B. hyodysenteriae genome and on a 3.1 kb fragment of the B. innocens B256T genome. The B. innocens lipoprotein did not react in Western blots with monoclonal antibody BJL/SH1 that reacts with the B. hyodysenteriae lipoprotein. The difference in binding with the monoclonal antibody may reside in the replacement of a serine residue with a tyrosine residue at base position 210 in the lipoprotein from B. innocens B256T. Comparison of the BmpB amino acid sequence with sequences in the SWISS-PROT protein database indicated that it has 33.9-39.9% similarity with the d-methionine binding proteins (MetQ) of a number of pathogenic bacterial species. The bmpB gene was confirmed to be the same as a gene of B. hyodysenteriae that was recently designated "blpA".     

Induction of swine dysentery in swine by the intravenous injection of filtered Treponema hyodysenteriae.

Swine dysentery was induced in 18 swine exposed by intravenous injection of a filtrate which contained Treponema hyodysenteriae and was obtained from macerated colonic scrapings of swine dysentery. However, swine dysentery did not develop in swine injected intravenously with a pure culture of T. hyodysenteriae or when combined with a colonic filtrate from normal swine. Diarrheal feces from the swine injected intravenously with the filtered T. hyodysenteriae contained more mucus, and fecal smears contained more T. hyodysenteriae and fewer other bacteria than did swine exposed orally to colon infected with swine dysentery or filtered T. hyodysenteriae. In the colons of the 12 swine injected intravenously with filtered T. hyodysenteriae that died, there was a minimum amount of croupous membrane and, microscopically, the T. hyodysenteriae were located deep in the colonic crypts. Five of the six surviving swine injected intravenously with filtered T. hyodysenteriae developed serum anti-T. hyodysenteriae antibodies using the indirect fluorescent antibody test and four of these swine developed diarrhea when reexposed with swine dysentery infected colon six weeks after initial exposure. None of the swine injected intravenously with cultured T. hyodysenteriae developed serum anti-T. hyodysenteriae antibodies and all were highly susceptible to swine dysentery.     

Growth of Serpulina (Treponema) hyodysenteriae under iron-restricted conditions.

Reference strains of Serpulina hyodysenteriae expressed at least three iron-regulated proteins with apparent molecular masses of > 200, 134, and 109 kDa when grown under iron-restricted conditions. Cells of S. hyodysenteriae grown under these conditions also showed increased outer membrane bleb formation when examined by electron microscopy after negative staining. S. hyodysenteriae did not use the 2 most common types of siderophore, namely catechol and hydroxamate. Western blotting with serum from a pig experimentally infected with S. hyodysenteriae B204 indicated that the 109-kDa major iron-regulated protein was expressed in vivo and was conserved among all strains tested.     

Sequence characterization of two new members of a multi-gene family in Serpulina hyodysenteriae (B204) with homology to a 39 kDa surface exposed protein: vspC and D.

Previous cloning and sequencing of clones from a genomic library constructed from Serpulina hyodysenteriae B204 had identified a tandem pair of open reading frames, identified as vspA and vspB (variable surface protein) expected to encode proteins with homology to ( but not identical with) a 39 kDa surface exposed membrane protein from this animal pathogen. Additional screening of the genomic library was performed to retrieve the remainder of the vspB gene using new oligonucleotide probes based upon the cloned gene sequences. Not only was this goal met but we also discovered two more adjacent and related vsp genes (vspC and vspD) and have completely sequenced them. They are all in a parallel orientation and appear to have a set of similar but distinct regulatory elements that may control separate expression of their open reading frames (ORFs). Thus, there are four contiguous vsp genes which are predicted to encode a family of structurally conserved proteins. The four adjacent open reading frames (ORFs) are of similar size (384-389 codons) and share from 83% to 90% identity in their amino acid sequence. Preliminary data suggests there may be yet another homologous gene copy in a distal location of S. hyodysenteriae that faithfully encodes the 39 kDa surface protein. The organization and homologies of these highly conserved multiple gene copies are discussed.     

Evidence that the 36 kb plasmid of Brachyspira hyodysenteriae contributes to virulence

Swine dysentery (SD) results from infection of the porcine large intestine with the anaerobic intestinal spirochaete Brachyspira hyodysenteriae. Recently the genome of virulent Australian B. hyodysenteriae strain WA1 was sequenced, and a 36 kilobase (kb) circular plasmid was identified. The plasmid contained 31 genes including six rfb genes that were predicted to be involved with rhamnose biosynthesis, and others associated with glycosylation. In the current study a set of PCRs was developed to amplify portions of nine of the plasmid genes. When used with DNA extracted from virulent strain B204, PCR products were generated, but no products were generated with DNA from avirulent strain A1. Analysis of the DNA using pulsed field gel electrophoresis (PFGE) identified a plasmid band in strains WA1 and B204, but not in strain A1. These results demonstrate that strain A1 does not contain the plasmid, and suggests that lack of the plasmid may explain why this strain is avirulent. To determine how commonly strains lacking plasmids occur, DNA was extracted from 264 Australian field isolates of B. hyodysenteriae and subjected to PCRs for three of the plasmid genes. Only one isolate (WA400) that lacked the plasmid was identified, and this absence was confirmed by PFGE analysis of DNA from the isolate and further PCR testing. To assess its virulence, 24 pigs were experimentally challenged with cultures of WA400, and 12 control pigs were challenged with virulent strain WA1 under the same conditions. Significantly fewer (P=0.03) of the pigs challenged with WA400 became colonised and developed SD (13/24; 54%) compared to the pigs infected with WA1 (11/12; 92%). Gross lesions in the pigs colonised with WA400 tended to be less extensive than those in pigs colonised with WA1, although there were no obvious differences at the microscopic level. The results support the likelihood that plasmid-encoded genes of B. hyodysenteriae are involved in colonisation and/or disease expression.     

3-Acetyl-4'-isovaleryl tylosin for prevention of swine dysentery

The 21 field isolates of Treponema hyodysenteriae which were tested were sensitive to 3-acetyl-4'-isovaleryl tylosin (AIV); the minimal inhibitory concentration was 0.25 to 16 micrograms/ml. 3-Acetyl-4'-isovaleryl tylosin administered prophylactically to pigs at concentrations of 5 to 100 mg/kg of feed and tylosin at 110 mg/kg of feed for 28 or 31 days prevented swine dysentery induced by tylosin-sensitive T hyodysenteriae strain SQ2; 15 nonmedicated, inoculated control pigs had bloody diarrhea, and 9 pigs died. In 2 additional trials, AIV administered prophylactically for 28 days at 55 or 110 mg/kg of feed prevented swine dysentery induced by tylosin-insensitive T hyodysenteriae strain B204. All of the inoculated principal pigs medicated with AIV at 55 or 110 mg/kg of feed or carbadox at 55 mg/kg of feed and the noninoculated sentinel pigs for each group had solid feces throughout the 56-day trial. In the nonmedicated, inoculated control groups, bloody diarrhea began at 4 to 5 days after inoculation was done, and 9 of 10 principal pigs and 6 of 9 sentinel pigs had dysentery; 2 pigs died. In the groups medicated with AIV at 27.5 or 5.5 mg/kg of feed, all 5 principal pigs and 3 or 4 sentinel pigs in each group had dysentery; 3 or 4 pigs in each group died. In the group medicated with tylosin at 110 mg/kg of feed, 7 of 10 principal pigs and all 9 sentinel pigs had dysentery; 1 pig died.(ABSTRACT TRUNCATED AT 250 WORDS)     

Swine dysentery: studies of gnotobiotic pigs inoculated with Treponema hyodysenteriae, Bacteroides vulgatus, and Fusobacterium necrophorum

Transmission experiments were carried out in gnotobiotic pigs to determine whether lesions typical of swine dysentery could be produced by oral inoculation of Treponema hyodysenteriae in combination with Bacteroides vulgatus or Fusobacterium necrophorum, or both. Each of the organisms had been isolated from swine with early lesions of the disease. Lesions were not found in 6 pigs inoculated with T hyodysenteriae alone, in 4 pigs given F necrophorum and T hyodysenteriae, or in 4 pigs given B vulgatus and F necrophorum. Lesions typical of swine dysentery developed in 8 pigs given B vulgatus, F necrophorum, and T hyodysenteriae as well as in 3 of 4 pigs given B vulgatus and T hyodysenteriae. In both of these groups, the inoculated bacteria were recovered from the colon, and T hyodysenteriae was demonstrated in the colonic crypts, epithelium, and lamina propria. The pathogenicity of the T hyodysenteriae was shown by the development of characteristic signs and lesions of swine dysentery in 12 of 14 naturally farrowed pigs inoculated with T hyodysenteriae alone.     

Antimicrobial susceptibility of Brachyspira hyodysenteriae isolated from 21 Polish farms

Swine dysentery (SD) is a common disease among pigs worldwide, which contributes to major production losses. Antimicrobial susceptibility testing of B. hyodysenteriae, the etiological agent of SD, is mainly performed by the agar dilution method. This method has certain limitations due to difficulties in interpretation of results. The aim of this study was the analysis of antimicrobial susceptibility of Brachyspira hyodysenteriae (B. hyodysenteriae) Polish field isolates by broth microdilution procedure. The study was performed on 21 isolates of B. hyodysenteriae, collected between January 2006 to December 2010 from cases of swine dysentery. VetMIC Brachyspira panels with antimicrobial agents (tiamulin, valnemulin, doxycycline, lincomycin, tylosin and ampicillin) were used for susceptibility testing of B. hyodysenteriae. The minimal inhibitory concentration (MIC) was determined by the broth dilution procedure. The lowest antimicrobial activity was demonstrated for tylosin and lincomycin, with inhibition of bacterial growth using concentrations > 128 microg/ml and 32 microg/ml, respectively. In the case of doxycycline, the MIC values were < or = 2.0 microg/ml. No decreased susceptibility to tiamulin was found among the Polish isolates and MIC values for this antibiotic did not exceed 1.0 microg/ml. The results of the present study confirmed that Polish B. hyodysenteriae isolates were susceptible to the main antibiotics (tiamulin and valnemulin) used in treatment of swine dysentery. Further studies are necessary to evaluate a possible slow decrease in susceptibility to tiamulin and valnemulin of B. hyodysenteriae strains in Poland.     

Identification of genes associated with prophage-like gene transfer agents in the pathogenic intestinal spirochaetes Brachyspira hyodysenteriae, Brachyspira pilosicoli and Brachyspira intermedia

VSH-1 is an unusual prophage-like gene transfer agent (GTA) that has been described in the intestinal spirochaete Brachyspira hyodysenteriae. The GTA does not self-propagate, but it assembles into a virus-like particle and transfers random 7.5kb fragments of host DNA to other B. hyodysenteriae cells. To date the GTA VSH-1 has only been analysed in B. hyodysenteriae strain B204, in which 11 late function genes encoding prophage capsid, tail and lysis elements have been described. The aim of the current study was to look for these 11 genes in the near-complete genomes of B. hyodysenteriae WA1, B. pilosicoli 95/1000 and B. intermedia HB60. All 11 genes were found in the three new strains. The GTA genes in WA1 and 95/1000 were contiguous, whilst some of those in HB60 were not-although in all three strains some gene rearrangements were present. A new predicted open reading frame with potential functional importance was found in a consistent position associated with all four GTAs, located between the genes for head protein Hvp24 and tail protein Hvp53, overlapping with the hvp24 sequence. Differences in the nucleotide and predicted amino acid sequences of the GTA genes in the spirochaete strains were consistent with the overall genetic distances between the strains. Hence the GTAs in the two B. hyodysenteriae strains were considered to be strain specific variants, and were designated GTA/Bh-B204 and GTA/Bh-WA1 respectively. The GTAs in the strains of B. intermedia and B. pilosicoli were designated GTA/Bint-HB60 and GTA/Bp-95/1000 respectively. Further work is required to determine the extent to which these GTAs can transfer host genes between different Brachyspira species and strains.     

The effect of fermentable carbohydrates on experimental swine dysentery and whip worm infections in pigs

An experiment was conducted to study the effect of diets with contrasting fermentability in the large intestine on experimental infections with Brachyspira hyodysenteriae, the causative agent of swine dysentery, and the whip worm, Trichuris suis, in pigs. Two diets with organically grown ingredients were composed. Both diets were based on triticale and barley and supplemented with either rape seed cake (Diet 1) or dried chicory root and sweet lupins (Diet 2). The study had a three-factorial design, with eight groups of pigs receiving Diet 1 or Diet 2, +/-B. hyodysenteriae, and +/-T. suis. Pigs fed Diet 2 and challenged with B. hyodysenteriae did not develop swine dysentery and B. hyodysenteriae was not demonstrated in any of the pigs during the study. In contrast, 94% of the B. hyodysenteriae challenged pigs fed Diet 1 showed clinical symptoms of swine dysentery and all the pigs were shedding B. hyodysenteriae in faeces at some points in time during the experiment. The number of T. suis was lower in pigs fed Diet 2 compared to pigs fed Diet 1, but the differences were not significant. Pigs on Diet 1 and challenged with both pathogens showed clinical symptoms of SD for a longer period than pigs inoculated with B. hyodysenteriae only. The study showed that diets supplemented with highly fermentable carbohydrates from dried chicory roots and sweet lupins can protect pigs against developing swine dysentery, but do not have any significant influence on T. suis.     

Effect of highly fermentable dietary fiber on the development of swine dysentery and on pig performance in a "Pure--Culture Challenge Model"

This study tried to evaluate the effect of highly fermentable fiber on the incidence and severity of swine dysentery (SD) after experimental oral infection with pure cultures of Brachyspira (B.) hyodysenteriae. Forty eight growing pigs were allocated to two groups and treated until slaughter as follows: Group 1 (n = 24): infected with B. hyodysenteriae and fed with a food containing 9.6% highly fermentable neutral detergent fiber. Group 2 (n = 24): infected with B. hyodysenteriae and fed with a food containing 6.1% low fermentable neutral detergent fiber. Pigs of each group were intragastrically inoculated on each of three consecutive days with pure culture of 1.8 x 10(10) B. hyodysenteriae. All pigs were monitored daily until slaughter. Faecal shedding of B. hyodysenteriae by polymerase chain reaction, antibody response by IFA, clinical signs, growth performance and extents of gross and microscopical lesions specific for swine dysentery were determined. Faecal shedding of B. hyodysenteriae and antibodies specific for B. hyodysenteriae were detected at day 30 post infectionem. Significant (p < 0.05) milder clinical signs typical for swine dysentery were detected in group 1, fed with 9.6% high fermentable fiber compared to group two fed with a food containing 6.1% low fermentable neutral detergent fiber. Daily weight gain differed significantly (p < 0.05) between the groups (group one 780 g vs. group two 760 g). Food conversion efficiency showed in group one a significant (p < 0.05) better (3.28) result than in group two (3.38). Feed consumption presented significantly (p < 0.001) better results in group one compared to group two (2.38 kg vs. 2.25 kg). From our experimental findings we conclude that in production units suffering of swine dysentery high levels of highly fermentable fiber in diet may increase health and performance.     

Isolation of Treponema hyodysenteriae from sources other than swine

Fecal samples were collected from animals and environments on 3 swine farms and cultured for Treponema hyodysenteriae. Each farm was a farrow-to-finish operation and, at the time of sampling, swine dysentery was enzootic among 8- to 22-week-old pigs. Pathogenic T hyodysenteriae was isolated from pigs on all 3 farms. On farm A, nonpathogenic T hyodysenteriae was isolated from a sample of lagoon water. On farm B, pathogenic T hyodysenteriae was isolated from a waste-holding pit. On farm C, a dog was observed to be eating feces of pigs that had swine dysentery. The dog was diarrheic and a fecal sample yielded a pathogenic isolant of T hyodysenteriae. Further isolation attempts were unsuccessful after the dog was removed from the infected premises. Isolation of pathogenic and nonpathogenic organisms from waste-holding systems emphasizes the need for cultural techniques in detecting pathogenic T hyodysenteriae.     

Demonstration of genes encoding virulence and virulence life-style factors in Brachyspira spp. isolates from pigs

The distribution of many genes encoding virulence and virulence life-style (VL-S) factors in Brachyspira (B.) hyodysenteriae and other Brachyspira species are largely unknown. Their knowledge is essential e.g. for the improvement of diagnostic methods targeting the detection and differentiation of the species. Thus 121 German Brachyspira field isolates from diarrhoeic pigs were characterized down to the species level by restriction fragment length polymorphism analysis of the nox gene and subsequently subjected to polymerase chain reaction detecting VL-S genes for inner (clpX) and outer membrane proteins (OMPs: bhlp16, bhlp17.6, bhlp29.7, bhmp39f, bhmp39h), hemolysins (hlyA/ACP, tlyA), iron metabolism (ftnA, bitC), and aerotolerance (nox). For comparison, B. hyodysenteriae reference strains from the USA (n=7) and Australia (2) were used. Of all genes tested only nox was detected in all isolates. The simultaneous presence of both the tlyA and hlyA/ACP was restricted to the species B. hyodysenteriae. The hlyA infrequently occurred also in weakly hemolytic Brachyspira. Similarly to tlyA and hlyA all B. hyodysenteriae strains contained the ferritin gene ftnA which was also found in two Brachyspira intermedia isolates. OMP encoding genes were present in B. hyodysenteriae field isolates in rates of 0% (bhlp17.6, bhmp39h), 58.1% (bhlp29.7), and 97.3% (bhmp39f). Since the study revealed a high genetic heterogeneity among German B. hyodysenteriae field isolates differentiating them from USA as well as Australian strains, targets for diagnostic PCR were limited to the nox gene (genus specific PCR) as well as to the species specific nox(hyo) gene and the combination of hlyA and tlyA which allow to specifically detect B. hyodysenteriae.     

Characterisation of multiresistant Brachyspira hyodysenteriae isolates from Czech pig farms

The aim of the present study was to evaluate the importance of clonal spread of Brachyspira hyodysenteriae resistant to pleuromutilins (tiamulin, valnemulin) on farms in the Czech Republic. Agar dilution method and macrorestriction fragment profile analysis by pulsed field gel electrophoresis were used to characterise 35 B hyodysenteriae isolates that were obtained from clinical cases of swine dysentery on 32 farms between 2000 and 2005. Most isolates showed multiple resistances to tiamulin, valnemulin, tylosin and lincomycin. A total of six pulsotypes were detected in these multiresistant isolates. An analysis of epidemiological data showed that multiresistant B hyodysenteriae isolates were more often detected on fattening farms (59 per cent), compared with farms with other types of production. Furthermore, it was found that multiresistant B hyodysenteriae clones were most frequently selected on farms with endemic incidence of swine dysentery. This finding was confirmed by the characterisation of 21 B hyodysenteriae isolates obtained from three large-scale operations in seven consecutive years.     

Development of an experimental model allowing discrimination between virulent and avirulent isolates of Serpulina (Treponema) hyodysenteriae.

Variation in virulence among different strains of Serpulina hyodysenteriae was studied by oral inoculation of specific pathogen free piglets and CD-1 mice. Piglets infected with serotype 2 reference strain B204 and an untypable field strain LHV-90-9-I had severe diarrhea tainted intermittently with mucus and fresh blood. The piglets inoculated with B169, B8044, B6933, and ACK300-8 reference strains representing serotypes 3, 5, 6, and 7 respectively developed moderate diarrhea. However, reference strains B234 and A-1 of serotypes 1 and 4, respectively, failed to cause any diarrhea. None of the S. hyodysenteriae strains caused diarrhea in mice. The results indicate a great variation in virulence among strains of different serotypes of S. hyodysenteriae. Mice were less susceptible to infection with S. hyodysenteriae.     

Comparison of methods for antimicrobial susceptibility testing and MIC values for pleuromutilin drugs for Brachyspira hyodysenteriae isolated in Germany

In Germany treatment of swine dysentery is hampered by Brachyspira hyodysenteriae strains showing elevated MIC values to the few antibiotics licensed. Therefore, susceptibility testing of clinical isolates is an important service to the swine practitioner. This study compares the established agar dilution procedure for antimicrobial susceptibility testing of this fastidious anaerobe to the broth microdilution test newly developed [Anim. Health Res. 2 (2001) 59; Vet. Microbiol. 84 (2002) 123; J. Clin. Microbiol. 41 (2003) 2596]. A total of 221 isolates were examined twice with either test procedure using tiamulin and valnemulin as antibiotics. Both methods gave reproducible results, and the MIC values for the reference strains B. hyodysenteriae B204 and Staphylococcus aureus ATCC 29213 corresponded to previously published data. However, the results for individual strains differed significantly for both tests (P < 0.001) with MIC values being on average one dilution step lower in the broth dilution method. The 221 strains used for comparing test procedures were isolated between 1989 and 2001. An additional 102 strains isolated in 2002 were tested only with the broth dilution procedure. A significant rise in the average MIC value for both pleuromutilins could be demonstrated when comparing earlier isolates to those from 2000 to 2001 (P < 0.05), while in 2002 the average MIC significantly decreased when compared to the value in 2000 (P < 0.05). However, strains with MIC values for tiamulin as high as 8 microg/ml (broth dilution) could still be isolated.     

Investigations on Brachyspira--diagnostic and therapeutic strategies in swine dysentery

The infectious agent of swine dysentery, Brachyspira (Br.) hyodysenteriae, seems to be widespread in German pig herds. Due to different reasons the eradication is increasingly difficult. Not only the success of therapeutic procedures but also the possibilities of diagnostics are unsatisfactory. Although only the bacteriological investigation of faeces or intestinal probes by culture techniques allows the typing of Brachyspira strains and the testing of drug resistance, however, the rate of false negative results is relatively high. In comparison with the cultural method an easy, prompt and cheap immunofluorescent test (IFT) resulted in a good sensitivity (90%). The higher rate of negative results by culture techniques can not be attributed to a lower specificity of the IFT, but to an insufficient transport of samples to the laboratory. The IFT therefore has to be considered as a valuable supplement to the cultural diagnostic of Br. hyodysenteriae. It is absolutely necessary to establish strategies in eradication of swine dysentery which result in pig breeding herds free of Br. hyodysenteriae. Only weaner pigs which are reliable free of this germ guarantee a fattening period sufficiently free of swine dysentery. The principles of different measures in effective eradication are described.     

SmpB: a novel outer membrane protein present in some Brachyspira hyodysenteriae strains

A novel outer membrane protein-encoding gene was identified in Brachyspira hyodysenteriae. The predicted protein, SmpB, was encoded by a gene that contains regions of identity with that encoding the previously identified lipoprotein SmpA. However, the majority of the reading frame encoding SmpA and SmpB share no detectable similarity. Analysis of several strains revealed that B. hyodysenteriae harbours either smpA or the newly identified gene smpB, but not both. smpB encodes for a slightly larger protein than smpA, 17.6 and 16.8 kDa, respectively. The predicted proteins share an identical leader sequence and the first 10 amino acids of the mature protein, however, the remainder of the predicted protein sequence shows no similarity. It is hypothesised that smpA and smpB are present on the same area of the chromosome. The proteins are antigenically unique, as antisera raised against a strain of B. hyodysenteriae that expresses SmpA cannot detect SmpB and vice versa. Although the presence of an identical leader peptide suggests identical localisation of SmpA and SmpB, it is not known if the two predicted proteins share similar function.     

Swine dysentery: protection of pigs by oral and parenteral immunisation with attenuated Treponema hyodysenteriae.

An attenuated strain of Treponema hyodysenteriae was used to immunise 18 pigs in three experiments. Live attenuated spirochaetes were dosed orally and injected intra-peritoneally, and killed spirochaetes were injected intramuscularly with adjuvant. The vaccinated pigs, which developed high serum agglutination titres against T hyodysenteriae, and 18 unvaccinated litter-mates were repeatedly challenged with virulent T hyodysenteriae. Nine vaccinated pigs and 16 control pigs developed typical swine dysentery.