SummaryStre pto b a c illus m o nil ifo rm is is a Gram-negat ive bac terium found in various laboratory anim al species and is the cause of rat bit e fever and Haverhill fever in man. In order to evaluate a polymerase chain reaction (PCR ) for the detection of this zoonotic bact erium in anim al tissues a set of primers was designed based on the DNA base sequence of part of the 16S rRNA gene from 11 S. m o nili fo rm is strains. T he PCR detected as few as 2±6 copies of S. m o nilifo rm is DNA. A 296 bp DNA fragm ent was ampli®ed from S. m o nilifo rm is strains from rodents, humans and turkeys. Amplicons of about the same size were obtained from T he PCR detected S. m o ni lifo rm isinfection in all four orally-and four intravenously-infected C57BL/6 mice and the bac terium was cultured from all but one mouse. T he PCR detected S. m o nilifo rm is infection in all 12 orally-infected WU rats, and in ®ve of eight rats exposed to natural infection. Enzyme linked immunosorbent assay (ELISA) and PCR were equally successful in detecting infection in rats but S. m o ni lifo rm is was not detected by using culture. Keywords S. m o nilifo rm is; PCR; mouse; rat; health monitoring Stre pto b a c illus m o nil ifo rm isis a Gramnegative bac terium that can occur in various laboratory animal species including gerbils, guineapigs, mice and rats. Humans may become infected through ingestion of contam inated water or through a bit e or a scratch of an animal, leading to Haverhill fever (HF) and rat bite fever (RBF), respectively (Wullenweber 1995). It is commonly believed that with the use of modern maintenance system s streptobac illosis has been eradicated from laboratory animal colonies. T here are however recent reports on the isolation of S. m o nilifo rm is from laboratory rodents (Wullenweber e t a l. 1990, Koopman e t a l. 1991, Kirchner e t a l. 1992, Wullenweber e t a l. 1992, Glastonbury e t a l. 1996 ), and serological observations using an ELISA (Boot e t a l. 1993 ) suggest that latent infection may be common in rats (unpublished observations). As the culture of S. m o ni lifo rm is from healthy animals is dif®cult, we developed and evaluated a PCR for the detection of the bacterium in tissues of mice and rats. Materials and methods Mic ro o rga nism sSe ve nte e n S. m o nilifo rm is strains were obtai ned from mice (3 ), rats (3 ), a spinifex hopping mouse (1 ), turkeys (4 ) and from cases of HF (1 ) and RBF (5 ) in humans (Table 1).

SummaryAn enzyme-linked immunosorbent assay (ELISA) to measure Streptobacillus moniliformis antibodies in mice and rats was developed. Twelve S. moniliformis strains originating from cases of rat-bite fever and Haverhill fever in man and from various rodent species, showed considerable serological relationship.The ELISA appeared specific since antibodies to S. moniliformis were absorbed by autologous and homologous antigen, but not by heterologous bacterial antigens.Acholeplasma laidlaw;; showed partial serological relationship with S. moniliformis.The ELISA was validated using experimental infections in mice and rats. These studies and observations in naturally infected feral rats, confirmed that S. moniliformis is difficult to grow on primary isolation, and that the ELISA for S. moniliformis antibodies revealed more contaminated animals than culture. Keywords: Streptobacillus moniliformis;Serology; ELISA; Mouse; Rat Streptobacillus monilijormis has been associated with rat-bite fever and Haverhill fever in humans, following a bite and contamination of food by rat urine respectively. The rat has been identified as the main reservoir of S. monilijormis and the bacterium is considered a commensal inhabitant of the nasopharynx in this species of animal. S. moni/iformis has been reported to

SummaryTwo indirect immunofluorescence (IIF)assays, two enzyme-linked immunosorbent assays (ELISAs)and the carbon immunoassay (CIA)for determination of antibodies to Encephalitozoon cuniculi were compared using 210 sera of rabbits, 135 of which originated from seven infected colonies, while 75 originated from four uninfected colonies. There was no evidence of a difference between the different assays with respect to the number of positive sera. There was a clear correlation between the quantitative response measured by IIF and CIA and the other assays, and between both IIF tests, while no such correlation was found in the quantitative response measured by ELISAs,which might be explained by the less quantitative nature of the ELISA.Therefore quantitative determination of antibodies to E. cuniculi should be performed by IIF and not by ELISA.The nosographic sensitivities Nt and specificities N2 of the assays were~0.94 and~0.97 respectively. Small differences in Nt and N2 between the assays, although not statistically significant, were responsible for differences in the calculated predictive values of a positive test and of a negative test. As expected, the magnitude of these differences depended on the fraction of positive sera sampled from a given colony. There was strong evidence of such a difference between the fraction of positive sera found in different colonies, but the sample size from some colonies was too small to allow any conclusion, whether this was due to differences in the prevalences of the infection in the colonies or something else. We conclude that any of the assays will be suitable for the routine health monitoring of laboratory rabbit colonies for E. cuniculi infection, as recommended by the Federation of European Laboratory Animal Science Associations.

No abstract

An enzyme-linked immunosorbent assay (ELISA) was developed for the detection of Pasteurella pneumotropica antibodies in the sera of rats, mice, hamsters and Mastomys. P. pneumotropica from mice and rats showed cross-reactivity. The ELISA using P. pneumotropica NCTC 8284 detected more infected animals than selective culture in groups of rodents from which P. pneumotropica, Haemophilus sp and/or Actinobacillus sp were cultured. Cross reactivity between P. pneumotropica NCTC 8284 and haemophilus and actinobacillus isolates were not studied.

Culture of guinea pig and rabbit respiratory tracts for bacteria using X- (haemin) and V- (NAD) factor in agar media detected infection by V-factor dependent Pasteurellaceae (Haemophilus sp.) in three colonies of guinea pigs and a group of rabbits. The 12 Haemophilus strains comprised three API NH codes classed as Haemophilus parainfluenzae and two codes classed as Haemophilus aphrophilus/paraphrophilus. Six cell wall lipid profiles were detected, but these were not related to API NH codes. Both bacteriological properties were used to select strains for serological studies but any relationship between bacteriological and serological properties of the Haemophilus strains was not evident. Varying serological relationships occurred between the newly isolated Haemophilus strains, [Pasteurella] pneumotropica NCTC 8284 and Haemophilus strains previously isolated from rats.

SummaryMice and rats were experimentally infected with Pasteurellaceae isolated from mice, rats, hamsters and gerbils. Mice and rats were most heavily colonized by strains originally isolated from mice and rats respectively, and to a lesser extent by Pasteurellaceae from hamsters and gerbils. Colonization was generally accompanied by seroconversion. Gross pathology of the lungs was not observed. We conclude that Pasteurellaceae-free SPF mice and rats can be colonized by members of this bacterial family present in other rodent species. KeywordsPasteurellaceaej rodents; experimental infection; species-specificity Pohl 1979(Mannheim 1984 show a remarkable degree of species-specificity in colonizing hosts (Kilian & Frederiksen 1981). Although Pasteurellaceae have been isolated frequently from both healthy and diseased laboratory rodents, notably from mice and rats, host-specific parasitism of P. pneumotropica and related bacteria has hardly been studied. Morphological and biochemical differences between P. pneumotropica strains according to rodent species of origin, have not been reported. However, antigenic differences between mouse and rat strains of P. pneumotropica have been suggested (Nakagawa & Saito 1984) and we found differences in the haemagglutinating activity between mouse and rat strains of P. pneumotropica (Boot et a1. 1993a bacteriological, serological and pathological observations in SPF mice and rats which were experimentally infected with Pasteurellaceae isolated from mice, rats, hamsters and gerbils. Our study was aimed at increasing insight into the epidemiology of pasteurellosis in mouse and rat colonies. Members of the family Pasteurellaceae

Cross-reactivity studies with Pasteurellaceae from guineapigs revealed 5 serologically distinct groups, comprising Pasteurella multocida, Sp group bacteria, SP-like bacteria, Pasteurella pneumotropica and an actinobacillus-like bacterium. Guineapig Pasteurellaceae differed serologically from mouse-derived P. pneumotropica NCTC 8284. An enzyme-linked immunosorbent assay (ELISA) using SP group antigen, developed to monitor 'natural' infections by SP group Pasteurellaceae in guineapigs, detected significantly more infection than did cultivation, and was found superior to an ELISA performed with P. pneumotropica NCTC 8284.