Laboratory Animal Science Vol 48, No 5 Copyright 1998 October 1998 by the American Association for Laboratory Animal Science Detection of Pasteurellaceae in Rodents by Polymerase Chain Reaction Analysis Frank Bootz,1* Susanne Kirschnek,2 Werner Nicklas,2 Stefanie K. Wyss,1 and Felix R. Homberger1 The family Pasteurellaceae presently consists of the gen- sidering P. pneumotropica the only relevant species of era Pasteurella, Actinobacillus, and Haemophilus. Infection Pasteurellaceae is, therefore, short sighted at best. Although with Pasteurellaceae is found in all mammalian and avian there is no universal agreement, we believe that other species. These bacteria are the most common organisms members of the Pasteurellaceae may also be of importance isolated from contemporary conventional and barrier-main- for laboratory animal medicine. If not principally as agents tained rodent colonies (1, 2). A number of them have been causing overt disease, they may influence biomedical re- documented to be opportunistic pathogens and have been search by causing subclinical infection. This is based on implicated as causative agents of various diseases. Infec- our interpretation of the information available, either pub- tions are usually asymptomatic and persistent. Clinical lished or from personal experiences. In addition, we believe manifestation of infection may include pneumonia; conjunc- that the name P. pneumotropica, due to a lack of sufficiently tivitis; metritis; cystitis; pleural, peritoneal, and orbital sensitive differentiation methods, has been used ambigu- abscesses; dermatitis; and panopthalmitis (3). ously in the past. This should be investigated further. Newer Pasteurellaceae also are known to affect the immune sys- methods, such as polymerase chain reaction (PCR) assays tem and alter the immune response (4). Owing to the bio- (12, 13), will help to answer this question by differentiat- chemical diversity of the Pasteurellaceae, identification of ing Pasteurella strains on a genetic level. Therefore, to not isolates often has been challenging (5). Animals infected miss an organism that might be potentially detrimental to with these organisms are usually asymptomatic carriers biomedical research, a state-of-the-art health-monitoring harboring few bacteria in their respiratory or genital tract. program in laboratory rodents needs to include all mem- The organisms sometimes are difficult to culture because bers of the Pasteurellaceae in its screening spectrum (14). they are present in small numbers or may be overgrown Monitoring for bacterial agents is usually performed by by other bacteria; some are even growth factor dependent culture methods. Another method used to detect infection (6). Actually, this phenomenon has historically been used with Pasteurellaceae in rodents is serologic testing, which for taxonomic purposes. All factor-dependent Pasteurellaceae may be unreliable due to lack of cross-reactivity between have been assigned to the Haemophilus genus. However, different isolates or may yield false-positive results due to recent studies involving comparison of the 16S rRNA gene unpredictable cross-reactivities. Owing to the diversity of have suggested that classical taxonomy done on the basis of the Pasteurellaceae, a number of antigens are required to factor dependency and biochemical profiling may not be cover all species in the family. In our hands at least, we adequate. Factor-dependent bacteria have been found in all have found a high number of false-positive reactions, most three genera of Pasteurellaceae, and the biochemical profile often due to the age of the animals from which the sera is not directly linked with phylogenetic relationship (6). were obtained. To avoid false-positive reactions, it has been Pasteurella pneumotropica is considered the most (often recommended that animals older than 12 weeks (15) not only) pathogenic member of the Pasteurellaceae in rodents. be tested. It was specifically looked for in routine health moni- The most promising approach for detecting infection with toring; all other members of this family were disre- Pasteurellaceae might be PCR. This method, as well as se- garded (7). However, other members of this family, such rologic testing, was documented to detect P. pneumotropica as H. influenzaemurium (8, 9), P. ureae (later classified as in culture-negative animals (16–18). This recently described Actinobacillus muris [10]), and a not yet definitely described PCR is based on the 16S rRNA sequence of the type strain Pasteurella species found in rats (11), have been described and is therefore unlikely to detect all biotypes that are ex- as causing clinical disease. Also, different isolates from the pected to infect laboratory rodents. On the basis of this same species of Pasteurellaceae may have different bio- background, we decided to develop a PCR assay that is quick logical traits such as pathogenicity. Pathogenicity of an iso- and easy to perform, and is able to detect all strains of late does not correlate with its biochemical profile. Con- Pasteurellaceae that might be infective for rodents. The investigation was carried out using animals from Institute of Laboratory Animal Science, University of Zurich, Switzerland1 our routine health-monitoring program. A total of 35 mice, and Central Animal Laboratories, German Cancer Research Center, Heidel- 23 rats, 2 gerbils, 10 Syrian hamsters, and 1 rabbit of vari- berg, Germany2 *Address correspondence to Dr. Frank Bootz, Institute of Laboratory Ani- ous ages and either sex were examined. The animals were mal Science, University of Zurich, Winterthurerstrasse 190, CH-8057 derived from different experimental colonies of the Uni- Zurich, Switzerland. 542 Note Table 1. Phenotypic properties of Pasteurellaceae isolates that tested positive by polymerase chain reaction (PCR) Pasteurella pneumotropica isolates ATCC NCTC 12555a I198012 E409011 8141b H199011 K272011 E323021 G019021 J426011 F443021 F065011 K205031 Heyl Heyl Heyl Jawetz Jawetz Jawetz 1a 1a 1a 20 5 6 NAD dependency - ---- - -- - --- Mannitol - ---- - -- - --- Urease + +++++ ++ ++++ Phosphatase + +++++ ++ ++++ Indole - +++++ - - -+++ Arabinos + + + - - - - - - - - + Xylose + +++++ ++ + -- + Trehalose + +++++ ++ ++- - Melibiose + + + - - - + + + - - - Raffinose + + + - - - + + + - - - Ribose + +++++ ++ ++++ Esculin - - ---- -- - --- Origin of isolate Mouse Rat Mouse Mouse Mouse Mouse Rat Mouse Rat Mouse Rat Rat NCTC NCTC NCTC HIM K227011 K048012 11051c 11146d K063022 K091017 K026021 J384012 G106012 I159022 7857e 931-8f Pp21 Pp22 P. sp.(rat) H. infl. m. H. infl. m. P. sp. HW Act. murisActino 11 Actino 12 Hae. para. Hae. para. Hae. sp. NAD dependency - ----- -- --++ Mannitol - ----- -+ ++-- Urease + + ---- -+ ++-+ Phosphatase - ----- -- --++ Indole - + ---- -- ++-+ Arabinose - ----- -- ---- Xylose - ----- -- ---+ Trehalose + +++++ ++ ++ -- Melibiose + + ---- +- ++ -- Raffinose + + ---- ++ ++ +- Ribose + + - + + + + + + + - + Esculin - + ---- ++ +- -- Origin of isolate Mouse Mouse Rat Mouse Mouse Mouse Mouse Mouse Mouse Mouse Human Rat aPasteurella pneumotropica biotype Heyl bP. pneumotropica biotype Jawetz through this incision into the nasopharynx from which cPasteurella sp. (rat) specimens were collected for culture. Pharynx, trachea, and dHaemophilus influenzaemurium eH. parainfluenzae lung specimens were collected and stored at -708C before fHaemophilus sp. taxon B Kilian HK 447 being submitted to DNA extraction. The bacterial strains tested were chosen to include a Table 2. Bacterial strains that tested negative by PCR member of each genus and subgroup or phenotype of all Enterobacteriaceae subgroup 1 hitherto detected rodent biotypes and to cover the broad Citrobacter freundii Enterobacter cloacae spectrum of the family. We also tested a number of bacte- E. sakazakii ria not of the Pasteurellaceae family, either closely related Escherichia coli Klebsiella pneumoniae A 189/2/83 to the target organism or often present in diagnostic Proteus vulgaris samples, such as Enterobacteriaceae (average phylogenetic Salmonella spp. difference of 12%) (19), Vibrionaceae, Pseudomonadaceae, Yersinia enterocolitica A 329/2/95 Vibrionaceae subgroup 2 and Mycoplasmataceae. A complete list of bacterial strains Aeromonas sobria B 6/2/95 used in this study is given in Tables 1 and 2. Pseudomonaceae Twenty-four strains of Pasteurellaceae isolated or main- Bordetella bronchiseptica rat 437/95 Pseudomonas aeruginosa tained at the German Cancer Research Center (DKFZ) (5) Mycoplasmataceae were grown on sheep blood agar. Three strains of Mycoplasma pneumoniae Micrococcaceae Pasteurellaceae (phenotypic properties not shown) and Staphylococcus aureus those not Pasteurellaceae were maintained at the Insti- tute of Laboratory Animal Science of the University of versity of Zurich animal facility. Cages, feeders, and water Zurich as reference strains on various media. One-day-old bottles were changed each week. All animals had access to colonies were picked and diluted in 100 ml of phosphate- feed (no. 890 mouse and rat maintenance diet; NAFAG, buffered saline for DNA extraction. Gossau, Switzerland) and water ad libitum. Animal rooms Swab specimens were plated directly onto 5% sheep blood were kept at 22 6 28C with a 12/12-h light/dark cycle and and chocolate agars, and incubated in an oxygen-reduced 55 6 10% relative humidity. Quarterly health monitoring atmosphere in a candle jar at 378C. After 24 and 48 h, the of rodents included comprehensive ante- and postmortem plates were examined for colonies