Abstract
Recent literature has indicated that animals seropositive for Clostridium piliforme may have negative findings on polymerase chain reaction (PCR) testing. This study examines and reports on the results of serology, PCR and histopathology tests performed on a group of 20 laboratory rabbits seropositive for C. piliforme using the latest available diagnostic methodologies. The presence of the organism was not confirmed by either PCR or histopathology in this group of 20 for C. piliforme seropositive rabbits. This presents challenges for laboratory animal veterinarians and scientists wishing to establish the presence of the organism with commonly available diagnostic methods and means that clinical signs of disease in addition to diagnostic results must be interpreted together.
Clostridium piliforme, an anaerobic, Gram-negative, sporeforming bacterium, is the causative agent of Tyzzer's disease. 1 The clinical disease, first described by Ernst Tyzzer in 1917, is characterized in rabbits by voluminous watery diarrhoea, anorexia, dehydration and death primarily in rabbits around the age of weaning.1,2 The bacterium, which had been called Bacillus piliforme until it was reclassified in the mid 1993,3,4 infects a wide variety of hosts including mice, rats, hamsters, guineapigs, rabbits, dogs, cats, birds, pandas, deer, horses, cattle, and non-human primates, both old and new world species.5–15 One human has been diagnosed with the bacterium. In that case, a man positive for HIV-1 infection had the bacterium associated with several chronic skin lesions that were described as crusted and verrucous. 16
During an outbreak of diarrhoea in a colony of rabbits at the age of weaning, Tyzzer's disease is considered a major differential diagnosis.1,2,17 Many commercial vendors and institutions include C. piliforme in their list of organisms to exclude for a colony. Colonies are routinely screened for evidence of infection by serology using enzyme-linked immunosorbent assay or other methods. 2 Positive serological findings are not definitive as they may be due to crossreacting antibodies elicited by non-pathogenic bacteria. 4 A definitive diagnosis of Tyzzer's disease is based on the observation of histopathology and intracellular bacteria in specially (e.g. Warthin-Starry) stained sections of ileum, heart or liver.5,18,19 Infection can also be substantiated by polymerase chain reaction (PCR) targeting 16S ribosomal ribonucleic acid (rRNA) genomic sequences.3,4,6,16,20,21 The incidence of Tyzzer's disease in a colony has been shown to vary according to the host species and strain of C. piliforme, but in rodent colonies it is often quite low. To facilitate a diagnosis, Tyzzer's disease can be provoked in asymptomatically infected hosts by immunosuppression with cyclophosphamide 22 or cortisone acetate. 23
Most studies of the pathobiology and diagnosis of Tyzzer's disease have been on mice and rats; they generally support the conclusion that C. piliforme infection does not persist in immunocompetent hosts following seroconversion. In this study, we investigated C. piliforme infection by histopathology and PCR in serologically positive rabbits.
Materials and Methods
Animals
Twenty 13-14-month-old, female, non-actively breeding, New Zealand White [Hra: (NZW) SPF] rabbits, from an active breeding facility (Covance Research Products, Denver, PA, USA) accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care International, were selected based on predetermined criteria and submitted to Charles River Research Animal Diagnostic Services (CR-RADS, Wilmington, MA, USA) for the panel of C. piliforme tests described below. This represented a collaborative research project with Institutional Animal Care and Use Committee approval of procedures. These rabbits were selected at random from a pool of 150 rabbits from a colony with animals that had previously tested positive for C. piliforme by an immunofluorescence assay (IFA) and a multiplexed fluorometric immunoassay (MFIA) performed at CR-RADS. More specific tests, such as PCR, had not been used to further confirm a diagnosis in the pool of 150 rabbits. From that pool of animals, 52 (34.7%) were positive for C. piliforme on MFIA. Confirmatory testing on those positive on MFIA was conducted using IFA techniques and 73% of those (38) were positive. Approximately, three to four weeks after the initial testing of the pool of 150 rabbits, the 20 rabbits used in this study were randomly selected from the 38 animals that were positive on both median fluorescence index (MFI) and IFA.
Clinical signs of Tyzzer's disease were not observed in any of the C. piliforme seropositive rabbits or in the colonies from which they originated. According to the results of monthly health monitoring, the colony from which the 20 rabbits originated were free from Pasteurella multocida, Pasteurella pneumotropica, Treponema cuniculi, Encephalitoon cuniculi, Salmonella sp., Eimeria stidedae, Toxoplasma gondii, Passalurus ambiguous, ectoparasites and dermatophytes. Rabbits for this study originated from one building, were not actively breeding, and were either group or singly housed in three-tiered, fixed, stainless-steel caging with slatted floors and no bedding; environmental parameters included an ambient temperature range of 18.3-23.8°F, a relative humidity range of 30-70%, a light cycle of 14 h light:10 h dark with artificial lighting, ventilation with 100% fresh outside area with a minimum of 10 air changes per hour; Purina 5321 (Purina, Richmond, IN, USA) feed provided in measured daily amounts, and water treated by particulate filtration (5 μm), chlorination and ultraviolet irradiation provided ad libitum by an automatic watering system.
Gross necropsy and collection of specimens for C. piliforme testing
After rabbits were euthanized with carbon dioxide, blood was collected for C. piliforme serology and gross necropsies were performed. Samples of organs that are potential targets of C. piliforme infection, including ileum, caecum, colon, heart and liver, were collected for histopathology. Caecum samples were collected for PCR.
Histopathology
Tissue specimens for histopathology were fixed in 10% neutral-buffered formalin and embedded in paraffin blocks, from which 4 μm tissue sections were prepared. Stained sections were microscopically examined by an American College of Veterinary Pathologists board-certified veterinary pathologist. Haematoxylin and eosin-stained sections were examined for lesions indicative of Tyzzer's disease. Warthin-Starry silver-stained sections were examined for the ‘pick-up-sticks’ arrays of bacilli observed in C. piliforme-infected cells.
C. piliforme serology
C. piliforme antigen was prepared from cultures of the BRL-3A rat liver cell-line (American Type Cell Culture, Rockville, MD, USA) infected with the R1 rat strain of C. piliforme that has been shown to be cross reactive with a variety of species.24,25 C. piliforme organisms harvested from infected cultures were depleted of host cells by lowspeed centrifugation (at 400 × g) for 10 min and pelleted by high-speed centrifugation at 12,000 × g for 20 min. The bacterial pellet was resuspended in phosphate-buffered saline (PBS), pH 7.4, pelleted by centrifugation and resuspended in PBS. For IFA, suspended C. piliforme were dispensed into glass-slide wells; slides were dried at room temperature and then immersed in cold acetone to fix the C. piliforme. To prepare antigen for MFIA, suspended C. piliforme were incubated in PBS with 1% (w/v) n-Octyl-β-D-glucopyranoside (Calbiochem, San Diego, CA, USA) for several hours at room temperature; intact bacteria were removed by high-speed centrifugation and the clarified detergent extract was dialysed in PBS and covalently coupled to colour-coded microbeads according to the manufacturer's protocol (Luminex Corp, Austin, TX, USA).
The indirect IFA was performed by incubating 10 μL of serum, diluted 10-fold, in a C. piliforme-coated slide well followed by a wash step, incubation with FITC-labelled goat IgG anti-rabbit IgG and a final wash. Both incubations were for 30 min at 39 ± 2°C in humidified air. IFA slides were washed by immersion in several PBS baths. Following the final wash, slides were covered with 0.1 mmol/L Tris-HCl-buffered glycerol mounting medium, pH 8.7, and a coverslip and were examined at 400× magnification with an epi-illumination fluorescence microscope (Carl Zeiss AG, Oberkochen, Germany). IFA reactions were classified as positive, negative or non-specific according to the pattern and intensity of fluorescence. 26
The MFIA, which utilizes Luminex's xMAP® suspension microarray technology (Luminex, Austin, TX, USA), was performed in 96-well filter-bottom plate wells (1.2 μm poresize Multiscreen, Millipore Corporation, Bedford, MA, USA) as follows. Fifty microlitres of a pooled suspension of assay and control microbead sets (Table 1), containing 2500 microbeads per set, and an equal volume of serum diluted 50-fold were dispensed per assay well and incubated together for 60 min. This and subsequent incubations were carried out at 27 ± 2°C in darkness, with orbital shaking at approximately 500 rpm. After being incubated with serum, microbeads were washed two times by the addition and evacuation (through the well filter bottom) of wash solution. Rabbit antibodies bound to the microbeads were demonstrated by 30 min incubations with 100 μL per well of biotinylated goat IgG anti-rabbit IgG (Jackson ImmunoResearch Laboratories, West Grove, PA, USA), followed by streptavidin-phycoerythrin conjugate (Invitrogen, Carlsbad, CA, USA). Microbeads were washed after each incubation as described. The MFI for each microbead set was determined with a BioPlex Suspension Array Reader System (Bio-Rad Laboratories, Hercules, CA, USA). MFI were used to calculate scores, which were classified as positive, negative or equivocal if they were ≥2.5, <1.5, or ≥1.5 and <2.5, respectively, provided that the tissue control score was <2.0. 28
Rabbit serology multiplexed fluorometric immunoassay (MFIA) microbead panel
The tissue control microbead set measures non-specific binding of serum immunoglobulin. The reporter dye signal, or median fluorescence index (MFI), is subtracted from the microbial antibody assay MFI to determine net MFI, which is the specific serological response. A sample that reacts strongly with the tissue control bead set is not suitable for MFIA. The rabbit IgG control bead set is an internal system-suitability control that evaluates assay conditions, reagents and reader performance for each test well. The serum IgG control shows whether the concentration, species and type of immunoglobulin in a serum sample are in general suitable for an MFIA. For the results of microbial antibody assays to be valid, the MFI for the anti-rabbit IgG and serum IgG control tests must be ≥ the cut-off
Cut-off MFI were assigned to each assay by inspecting the results for known-negative and known-positive sera. The microbial antibodyassay-cut-off values are net MFI calculated by subtracting the tissue MFI. The validity of the net MFI cut-off values was verified by receiver operating characteristic (ROC) analysis, which employs a variety of graphs and statistics to summarize the specificity and sensitivity of a diagnostic test for a range of different cut-offs 27
SF: Spodoptera frugiperda; NA: not available
Fluorogenic endonuclease PCR (fnPCR)
Previous work has demonstrated that C. piliforme infection persists longer in the intestines than in the liver. 29 Thus, the decision was made to perform PCR analysis on intestinal tissues (caecal luminal contents and wall tissue). DNA was extracted from 6-8 mg of faeces or 15-20 mg of unwashed caecum by column isolation (DNeasy Mini Kit, Qiagen, Valencia, CA, USA) and eluted in 100 μL. PCR primers and probes were designed to target ribosomal (r) RNA genomic sequences that are unique to C. piliforme. Primer and probe nomenclature is based on their location within the sequence of Genbank accession# L07416. Five microlitres of eluted DNA were added to 25 μL final reaction volume containing 1× TaqMan Universal PCR Master mix (Applied Biosystems, Foster City, CA, USA), 0.01% tween-20 (Sigma-Aldrich, St Louis, MO, USA), 0.05% gelatin (Sigma-Aldrich), and the following oligos: CPIF924 (AACCTTACC TAAACTTGACATACCAT) at 275 nM, CPIR1004 (GCACCA CCTGTATCCAT) at 725 nM, CPIP950A (FAM-TGACAGA CTACGTAAAGTAGTTTTCCTTCGGGA-BHQ1) at 125 nM and CPIP950B (FAM-TGACAGGCTACGTAAAGTAGCTTT CCTTCG-BHQ1) at 125 nM. To detect PCR inhibition, 5 μL eluted sample was added to a 25 μL final reaction volume, which included the same reagents as above except that it included 100 copies of the plasmid pGEM-luc (Promega, Madison, WI, USA) and different oligos that targeted the luciferase gene in the plasmid: LUCF227 (TCGCGGTTGTTACTT GACTGG) 650 nM, LUCR348 (GGCAGGTCTTCCCGACG) 275 nM and LUCP309 (FAM-ACGCCGGTGAACTTCCCG CC-BHQ1) 200 nM. Reactions were cycled on a 9700 thermalcycler (Applied Biosystems) using the following parameters: one cycle 50°C for 2 min and 95°C for 12 min, five cycles of 95°C for 15 s and 64°C for one minute, and 55 cycles of 95°C for 15 s and 58°C for one minute. Amplification was detected by measuring FAM and ROX fluorescent signals on an Ascent-Fluoroskan fluorometer (Thermo Fisher Scientific Inc, Waltham, MA, USA) electron and calculating a normalized reporter value (Rn = FAM/ROX). Samples were diluted 1:4 and retested if the PCR inhibition control failed (reduced or absent amplification).
Results
Clinical findings
The original pool of 150, including the 20 selected for this study, did not exhibit clinical signs of active Tyzzer's disease (e.g. no diarrhoea, morbidity or mortality were present).
Serology and PCR
The 20 rabbits in this study were randomly chosen from a group of 150 rabbits, 38 of which were shown to be strongly seropositive for C. piliforme antibodies by IFA and MFIA. The average MFIA net score for the initial testing of the 20 seropositive animals was 24, which is well above the cut-off of 3 and near the maximum possible score of approximately 30. A serum sample from one animal was unavailable for repeat serology. Of the 19 animals that were retested, 26% (6) were no longer MFIA positive and about half (10) were IFA negative at retest; the average MFIA net score was 14 (Table 2). The caecal specimens from all animals were C. piliforme PCR negative.
Retesting (after 3–4 weeks) of rabbits found to be C. piliforme seropositive by MFIA and IFA *
In the MFIA, a net score of 3 or higher is classified as positive, unless the sample shows a significant non-specific reaction with the tissue control (i.e. tissue control score >2); none of the sera in the study showed a significant non-specific reaction with the tissue control. A score of 1 or less is negative and 2 is equivocal. A score of 3 has a net MFI (i.e. assay – tissue control) equal to the cut-off, which for the CPIL MFIA was 3000
IFA results were classified according to the intensity and pattern of fluorescence; fluorescence matching that observed with positive control serum is positive (+)
MFIA: multiplexed fluorometric immunoassay; IFA: immunofluorescence assay; NT: not tested
Gross pathology and histopathology
The gross necropsy examination of 19 of the animals noted no abnormalities. One animal was found to have mild, diffuse, multifocal areas of yellow to white discoloration on the left medial lobe of the liver. Of all the organs examined in this study, the liver demonstrated the highest level of microscopic abnormalities (20% of the animals were affected) (Table 3). One animal demonstrated lesions in the heart (Figure 1). No other organs were affected. One of the five hepatic lesions, animal 19 (Figure 2), and the cardiac lesion, from animal 11, had elements consistent with clinical Tyzzer's disease (necrosis of the tissue) (Table 3). Sections of these animal's organs were stained with a Warthin-Starry silver stain to aid in the detection of C. piliforme organisms. No bacteria were observed.
Lesions observed histopathology test results
Necrosis in cardiac muscle tissue
Necrosis in hepatic tissue
Discussion
Serology is the most widely used methodology for C. piliforme surveillance of laboratory animal colonies because it is rapid and inexpensive. Monitoring by cultural isolation is not feasible because C. piliforme is a fastidious, obligate intracellular parasite that cannot be grown in cell-free medium. The applicability of culture, as well as PCR testing, is further limited by the apparently transient (and largely asymptomatic) course of C. piliforme infections in immunocompetent hosts. 29 Provocation of Tyzzer's disease in latently infected animals by immunosuppression, while definitive, is time consuming and unreliable. Although C. piliforme serosurveillance is convenient, unsubstantiated positive serological findings are not sufficient for a definitive diagnosis of C. piliforme infection because serological assays typically utilize complex antigen preparations comprising whole organisms or organism extracts that may cross react with antibodies to other bacterial species.
In this study, we evaluated whether standard, widely available diagnostic techniques, including gross and microscopic pathology and PCR testing, would corroborate C. piliforme infection in rabbits (from a historically seropositive colony) that were identified as C. piliforme seropositive by MFIA, a novel MFIA, and standard IFA. Postmortem gross and histological examination of gastrointestinal tract, heart or liver specimens from the seropositive rabbits did not reveal Tyzzer's disease; moreover, intracellular bacteria, which are pathognomonic for C. piliforme infection, were not observed in Warthin-Starry silver-stained tissue sections. As noted, active Tyzzer's disease had not been observed in the colony from which the study rabbits originated. Lastly, C. piliforme was not detectable in caecal specimens by the PCR assay. Possible explanations for these negative postmortem findings are that the rabbits had cleared the infection, as has been demonstrated for immunocompetent hosts of other species, or that the positive MFIA and IFA results were non-specific.
Although all rabbits were clearly MFIA and IFA seropositive when selected for this study, approximately one-quarter and half of the study animals were MFIA and IFA seronegative, respectively, when retested three to four weeks later at the time of necropsy. In addition, the mean MFIA score dropped from 24 in the first round of serological testing to 13 in the second. Thus, serum antibody levels appeared to drop substantially over a short period. That a higher percentage of animals stayed seropositive by MFIA than by IFA is consistent with the former assay method being generally more sensitive than the latter and may be related to antigen preparation and presentation differences. For instance, acetone fixation of IFA antigen might degrade certain important antigenic determinants.
In summary, we were unable to demonstrate C. piliforme infection of seropositive rabbits by gross and microscopic pathology or PCR testing. Although these negative findings are not surprising, since C. piliforme infections are often subclinical and cleared by immunocompetent hosts following seroconversion, 29 they illustrate the substantial challenges of definitively diagnosing C. piliforme infection and Tyzzer's disease.
Based on the results of this study, the authors would suggest that positive serological findings be consistently confirmed through the use of PCR testing and/or histopathological findings since serological methods are likely to remain the standard for health monitoring in laboratory animal facilities for years to come. In ideal situations, both PCR testing and histopathology results should be obtained to have maximal value. Lack of a positive result through PCR and/or histopathology could indicate false-positive test results through serology, a transient colonization with C. piliforme, or false-negative PCR/histopathology results. In those cases, follow-up serology performed at regular intervals (e.g. 3-4 weeks) could enable further evaluation as to the nature of positive serological response. Diagnostic value can be further enhanced by performing disease provocation studies on weanling or immunosuppressed animals in the colony. These study results may provide definitive evidence of the presence of C. piliforme and will provide additional samples for PCR/histopathology testing.
Footnotes
Acknowledgement
The authors would like to acknowledge Dr Ralph Bunte for his assistance with the manuscript.