Evaluation of a Novel Battery-Operated Tumbler Device for Use in the Detection of Mouse Pathogens for Rodent Health Monitoring
The search for alternatives to live animal sentinels in rodent health monitoring programs is fundamental to the 3Rs (Reduction, Replacement, and Refinement) of animal research. We evaluated the efficacy of a novel battery-operated tumbler device that rotates soiled bedding in direct contact with sample media against the use of exhaust sample media and soiled bedding sentinel (SBS) mice. Four rodent racks were used, each with 3 test cages: a cage with a tumbler device that rotated for 10 min twice a week (TUM10), a cage with a tumbler device that rotated for 60 min twice a week (TUM60), and a cage housing 2 female Crl:CD1(ICR) mice. Every 2 wk, each test cage received soiled bedding collected from all cages on each respective rack. In addition to soiled bedding, the tumbler device contained various sample collection media: a contact Reemay filter (3 mo-cRF) that remained in the tumbler for the duration of the study, a contact Reemay filter (1 mo-cRF) that was replaced monthly, adhesive swabs (AS) that were added at every biweekly cage change, and an exhaust Reemay filter located at the exhaust outlet of the cage. All analyses were performed by direct PCR for both sample media in the animal-free methods, and fecal pellet, body swab, and oral swabs were collected from sentinel mice. Out of 16 total pathogens detected, assessment of 1 mo-cRF from both TUM10 and TUM60 cages detected 84% and 79% of pathogens, respectively, while SBS samples detected only 47% of pathogens. AS in TUM60 and TUM10 cages detected the fewest pathogens (24% and 13%, respectively). These results indicate that the novel tumbler device is an effective and reliable tool for rodent health monitoring programs and a suitable replacement for live animal sentinels. In this study, 1 mo-cRF in TUM10 cages detected the highest number of pathogens.Abstract
Introduction
According to the Guide for the Care of and Use of Laboratory Animals, “appropriate procedures should be in place for disease surveillance and diagnosis” and “procedures for disease prevention, diagnosis, and therapy should be those currently accepted in veterinary and laboratory animal practice.”9 As such, health monitoring programs are a critical component of any research facility housing animals. Although rodents obtained from vendors are screened against a list of potential pathogens, situations such as failed quarantine, wild rodent incursion, untreated feed and bedding, and research biologics increase the potential for introducing adventitious agents. Therefore, animal colonies must be continually monitored for infectious diseases that can pose a risk to research, and a reliable method of pathogen detection is essential. Consistency with the 3Rs (Reduction, Replacement, and Refinement) of animal research has led to increasing acceptance of using alternatives to live animals in health monitoring programs, and many animal care and use programs are transitioning their sentinel detection program to nonanimal alternatives.15,28 Although some agents, such as mouse parvovirus (MPV), Helicobacter spp., mouse hepatitis virus (MHV), and murine norovirus (MNV), can be detected successfully in live soiled bedding sentinels (SBS),2,6,13,14,24,27 SBS may inefficiently or never detect some pathogens.3,10-13,17,22 Over the past decade, the transition from static to IVC systems for rodent housing has resulted in numerous studies demonstrating the superiority of environmental health monitoring over the use of SBS mice for the detection of infectious agents.2,3,7,8,13,20,23,29-31 However, depending on the caging type and level of filtration, testing of exhaust air at the rack level may not be an effective option for cage-level filtration.1 The success of detecting pathogens via SBS mice depends on many factors, such as the prevalence of infectious agents, the number of organisms or particles shed, the ability of the organism to transmit to and infect the sentinel, caging type, cage change frequency, immune status of mice, and the number of mice in the colony, all of which affect the likelihood of pathogen detection via SBS.26 Several studies have shown that adding sample media, such as filter paper, flocked swabs, or adhesive swabs (AS), to pooled bedding from colony mice in a mouse-free cage detected more pathogens than did SBS.3,5,7,8,20,29,30 These previous studies showed that stirring and shaking the soiled bedding in mouse-free cages that contain sampling media aid in the detection of infectious agents by mimicking the actions of SBS in agitating the dust particles that bind to the sample media.3,7,8,20,29,30
In this proof-of-concept study, we evaluated a novel battery-operated tumbler that slowly mixes soiled bedding in contact with diagnostic sample media and compared its operation to SBS to determine if this device was an effective nonanimal method of health monitoring with cage-level filtration. Pushing a button on this device triggers agitation of the soiled bedding, thus eliminating the need for husbandry staff to shake or stir the soiled bedding. Agitation of the soiled bedding is important because it promotes the binding of infectious agents to dust particles in the pooled soiled bedding so that they are distributed and can be more readily captured by the media.3,7,29,30 Use of the tumbler device can increase the uniformity of soiled bedding agitation while reducing the risk of injury and fatigue to personnel.
PCR analysis was performed on the contact filter paper and AS placed in the tumbler and on filter paper collected from the cage exhaust port. In addition, body/fur swabs, fecal pellets, and oral swabs were collected from SBS and submitted for PCR testing. This experimental design was created to closely mimic our institution’s current sentinel health monitoring program and to simulate a real-life scenario for the use of these tumblers. Time intervals used to run the tumbler devices were determined based on a previous in-house study evaluating the transmission of murine pathogens in mouse-free cages that contained pooled soiled bedding mixed with adhesive and flocked swabs; in that study, at least 20 manual cage rotations were required to achieve pathogen transmission to the diagnostic sample media.8 However, one full rotation of the tumbler device required 30 s, so 10 min of tumbling was equivalent to 20 manual shakes. We also evaluated a 60-min rotation time because preliminary work with the tumbler prototypes showed that 60 min of rotation provided adequate accumulation of dust and debris on the exhaust filter paper. We hypothesized these tumblers would detect more pathogens than would SBS. We also compared pathogen detection after 10 and 60 min of tumbling time.
Materials and Methods
Mice.
All use of mice was approved by the University of Tennessee Health Science Center’s IACUC and conducted in an AAALAC-accredited facility. Female Crl:CD1(ICR) (n = 8, age 6 to 8 wk) mice were obtained from Charles River Laboratories (Wilmington, MA) and housed 2 per cage. Sentinel mice used in this study were specified by the vendor to be free of Sendai virus, pneumonia virus of mice (PVM), MHV, minute virus of mice (MVM), MPV, MNV, Theiler murine encephalomyelitis virus (TMEV), reovirus, lymphocytic choriomeningitis virus (LCMV), ectromelia, mouse adenovirus, mouse cytomegalovirus (MCMV), K virus, polyomavirus, hantavirus, lactate dehydrogenase virus elevating virus, murine chapparvovirus (MuCPV), Bordetella bronchiseptica, Citrobacter rodentium, Filobacterium rodentium (formerly Cilia-Associated Respiratory Bacillus), Corynebacterium kutscheri, Helicobacter spp., Klebsiella oxytoca, Klebsiella pneumoniae, Mycoplasma pulmonis, Pasteurella multocida, Rodentibacter heylii (formerly Pasteurella pneumotropica Heyl), Rodentibacter pneumotropicus (formerly Pasteurella pneumotropica Jawetz), Pseudomonas aeruginosa, Salmonella spp., Streptobacillus moniliformis, Streptococcus pneumoniae, β-Streptococcus spp., Clostridium piliforme, ectoparasites (Myocoptes musculinus, Myobia musculi, and Radfordia affinis), Aspiculuris tetraptera, Syphacia muris, Syphacia obvelata, Spironucleus muris, Chilomastix bettencourti, Cryptosporidium muris, Cryptosporidium parvum, Eimeria spp., Entamoeba muris, Giardia muris, Hexamastix muris, Spironucleus muris, Trichomonas muris, and Tritrichomonas muris. Other mice housed on racks where soiled bedding was collected were not specified to be free from all the pathogens mentioned. Historically, these mice have tested positive for Helicobacter spp. and MNV as these pathogens are not excluded from the colony. A previous in-house study conducted on the same racks revealed the presence of nonexcluded pathogens such as Astrovirus-1, Proteus mirabilis, Staphylococcus aureus, Rodentibacter heylii, Rodentibacter pneumotropicus, Chilomastix muris, and Entamoeba muris.
Mice were housed in an ABSL-1 facility in Optimice IVC cages, specifically with autowater cage bottoms for rack-mounted valves and standard cage tops (Animal Care Systems, Centennial, CO) on a 12:12-h light:dark cycle, 68 to 79 °F (20 to 26 °C) temperature range, 30% to 70% humidity, and 10 to 15 air changes per hour. Mice were fed a standard irradiated rodent feed (Teklad 7912; Inotiv, Indianapolis, IN) and received hyperchlorinated water ad libitum. All mice were housed in autoclaved cages on corncob rodent bedding (The Andersons, Maumee, OH). During biweekly cage changes, one teaspoon of soiled bedding was collected from each cage on the rack, pooled into a single cage, mixed thoroughly, and transferred to an autoclaved cage until bedding was 1/4-in. deep.
Equipment.
The battery-operated spinning tumbler prototypes (patent pending) were provided by Animal Care Systems and can hold up to 100 g of soiled bedding (Figure 1A). Before the study, the tumbler prototypes and rod cylinder base were disinfected thoroughly using Peroxigard (Virox Technologies, Ontario, Canada). The tumbler cylinder, measuring 20.32 cm × 11.43 cm × 11.43 cm, opens into halves that are secured together by placing plastic rubber bands on both ends (Figure 1A). The base of the tumbler device runs on spinning rods (measuring 24.13 cm × 10.16 cm × 1.27 cm) with 2 attached D batteries that rotate at a speed of 1 full rotation every 30 s (Figure 1B). The current study used 4 racks; each rack contained 2 tumbler devices that were placed in separate Optimice cages. Each tumbler device was placed inside a cage with the lid on and docked onto the rack, making sure to remove the water automatic watering system before docking. The duration of tumbler rotations was controlled by an external timer that is connected to the rod base, where the flexible wire sits under the lid and hooks outside on the front of the cage (Figure 1D). Sample media that were used for pathogen detection in the tumblers were autoclaved 2024 Reemay Spunbonded Polyester Nonwoven Fabric filter paper that measured approximately 5.4 cm × 3.6 cm (Animal Care Systems) and individually wrapped AS that were approximately 8 cm in length (Charles River Laboratories) (Figure 1C). To ensure functionality of the tumbler device, an imprinted arrow on the surface of the tumbler cylinder is visible from the outside of the cage.
Citation: Journal of the American Association for Laboratory Animal Science 63, 3; 10.30802/AALAS-JAALAS-23-000053
Collection of soiled bedding.
At the start of the study, all cages were autoclaved and an autoclaved Reemay filter was placed and secured at the cage exhaust port as per the cage manufacturer’s instructions. Sentinel mice were placed in clean cages 1 wk before the start of the study. Soiled bedding for both the SBS and tumbler cages was collected in a biosafety cabinet at regularly scheduled cage changes every 2 wk. One teaspoon of soiled bedding from each cage on the respective racks was collected into either an autoclaved cage or autoclaved plastic container (ULINE, Pleasant Prairie, WI) for distribution into tumblers and SBS cages. The number of cages on each rack varied depending on animal census. During the 3-mo study duration, each rack had 10 to 89 occupied cages with an average of 3 mice per cage. Changing the bedding in the tumbler devices involved cleaning and disinfecting the biosafety cabinet with Peroxigard before use. The tumbler cylinder consisted of 2 halves secured by rubber bands at both ends (Figure 1A); these were easily loosened, and then the device was opened for replacement of soiled bedding. Half of the precollected soiled bedding was poured from the autoclaved collection containers into one-half of the tumbler, with the amount varying depending on the number of cages present at the time of cage changing (Figure 1C). Sample media were transferred to the new soiled bedding using an aseptic technique with sterile forceps. Gloves were changed between cages. The tumbler devices remained in the same cage for the study duration.
Study design.
The study used 4 different racks in a single mouse housing room. This study began when both the SBS and tumbler devices were exposed to pooled soiled bedding from each of the 4 racks after an exposure period of 2 wk. There was also a placement of a new exhaust Reemay filter (exRF) on each cage housing a tumbler device. The tumbling was evaluated for durations of 10 and 60 min. Each rack contained 3 test cages: 1) a cage containing 2 SBS mice, 2) a cage assigned to 10 min of tumbling (TUM10), and 3) a cage assigned to 60 min of tumbling (TUM60). Each tumbler contained 3 different sample media that rotated in contact with the dirty bedding: 1) a Reemay filter paper that remained in the device for the entire 3-mo duration of the study (3 mo-cRF), 2) a Reemay filter paper that was removed and replaced monthly (1 mo-cRF), and 3) an AS that was added at every cage change. The AS was added at every cage change as a follow-up to a prior in-house study in which cages were shaken manually, an AS was added at each cage change, and swabs were then pooled for analysis at the end of the study.8 A 1-cm horizontal cut was made on the monthly Reemay filters to distinguish them from the Reemay filters that remained in the tumbler device for the 3-mo study duration. The Reemay filter located at the cage exhaust port (exRF) was submitted for analysis at the end of the 3-mo study.
Sample collection and testing from tumblers.
1 mo-cRF were stored together in a sterile 50-mL screwcap centrifuge tube (Globe Scientific, Mahwah, NJ) and were submitted as one sample; AS were pooled in a separate sterile 50-mL screwcap centrifuge tube. The 2 tubes were submitted for analysis as 2 samples. The filter paper that had been placed at the cage exhaust outlet (exRF) was collected at the end of the 3-mo study using sterile forceps and was placed in a sterile 50-mL screwcap centrifuge tube. All samples were stored at room temperature between 68 and 77 °F (20 to 25 °C).
Sample collection and testing from SBS.
Samples collected from each sentinel mouse consisted of fur/body AS and oral swabs. These samples were pooled by cage and stored separately in a sterile 50-mL screwcap centrifuge tube and a 5-mL round-bottom polystyrene test tube (Corning Falcon, Fischer Scientific, Pittsburgh, PA), respectively. Up to 10 fecal pellets were collected from the soiled bedding and directly from the anus of the mice with sterile forceps and pooled together in a sterile 5.0-mL microfuge tube (Eppendorf North America) as one sample per cage.
Sample submission and processing.
At the end of the 3-mo study, samples from all the tumblers, exRF, and SBS were submitted to Charles River Laboratories for PCR analysis under the Mouse Surveillance PRIA Plus panel, which tests for mouse parvoviruses (MVM/MPV 1 to 5); MNV; MHV; murine rotavirus (MRV/EDIM); mouse theilovirus (TMEV and GDVII); adenovirus type 1 and 2 (MAV1 and MAV2); reovirus type 1, 2, 3, and 4; murine orthopneumovirus (PVM); Sendai virus; ectromelia; LCMV; new world hantaviruses; old world hantaviruses; hantavirus; lactate dehydrogenase-elevating virus (LDV); Astrovirus-1; Astrovirus-2; murine chapparvovirus (MuCPV, MKPV, and RoChPV-1); mouse thymic virus (MTLV); MCMV; mouse polyomavirus; Helicobacter; Citrobacter rodentium; Mycoplasma pulmonis; Streptobacillus moniliformis; Rodentibacter pneumotropicus; Rodentibacter heylii; Clostridium piliforme; Pseudomonas aeruginosa; Salmonella; Campylobacter; Bordetella bronchiseptica; Bordetella hinzii/B. pseudohinzii; Corynebacterium kutscheri; Corynebacterium bovis; S. aureus; Streptococcus pneumoniae; Klebsiella pneumoniae; Klebsiella oxytoca; Beta-hemolytic Streptococcus group A (S. pyogenes); Beta-hemolytic Streptococcus group B (S. agalactiae and S. dysgalactiae subsp. equisimilis); Beta-hemolytic Streptococcus group C (S. equi subsp. equi and zooepidemicus); Beta-hemolytic Streptococcus group G (S. dysgalactiae subsp. dysgalactiae and subsp. equisimilis); Chlamydia muridarum; Pasteurella multocida; Staphylococcus xylosus; fur mites (Myobia, Myocoptes, and Radfordia); pinworms (Aspiculuris and Syphacia); Demodex; tropical rat mite (Ornithonyssus bacoti); Giardia; Spironucleus muris; Cryptosporidium; Entamoeba; Eimeria (Coccidia, Cyclospora, and Isospora); Hexamastix; Chilomastix; Tritrichomonas; and Pneumocystis.
Biostatistical analysis.
The percentage of pathogens detected was computed by dividing the number of pathogens detected by a specific type of collection media by the total number of pathogen types that were detected by all collection media. The distribution of the residuals was verified to ensure that normality assumptions were met. Two-way crossed-classified ANOVA with method of pathogen detection and location of rack as fixed main effects was performed. The mean percentage of detected pathogen types was estimated for each combination of the type of collection media and rack. The interaction of method of pathogen detection and the rack location was not significant; therefore, the sample media were compared after averaging results from all 4 racks and using unpaired t test in the context of ANOVA. Statistical significance was determined at a type I error rate of 0.05. Biostatistical data analysis was performed using SAS Software Version 9.4 of the SAS System for Windows 10 Pro (2016, SAS Institute).
Results
Pathogens present on rack.
Racks were considered positive if any of the environmental health monitoring methods or SBS on the rack tested positive. Rack 1 was positive for 8 agents: 2 viruses (Astrovirus-1 and MNV) and 6 bacteria (Helicobacter genus, H. ganmani, H. mastomyrinus, Proteus mirabilis, S. aureus, and Rodentibacter heylii) (Table 1). Rack 2 was positive for 11 agents: 2 viruses (Astrovirus-1 and MNV), 7 bacteria (Campylobacter genus, Helicobacter genus, H. bilis, H. rodentium, H. typhlonius, Proteus mirabilis, and S. aureus), and 2 parasites (Chilomastix muris and Entamoeba) (Table 1). Rack 3 was positive for 13 agents: 2 viruses (Astrovirus-1 and MNV), 8 bacteria (Campylobacter genus, Helicobacter genus, H. ganmani, H. hepaticus, H. mastomyrinus, H. typhlonius, Rodentibacter heylii, and Rodentibacter pneumotropicus), and 3 parasites (Chilomastix muris, Demodex, and Entamoeba) (Table 1). Rack 4 was positive for 7 agents: 2 viruses (Astrovirus-1 and MNV), 8 bacteria (Helicobacter genus, H. ganmani, S. aureus, and Rodentibacter heylii), and one parasite (Entamoeba) (Table 1).
| Pathogen | Rack 1 | Rack 2 | Rack 3 | Rack 4 | Total positive racks | |
|---|---|---|---|---|---|---|
| Viruses | Astrovirus-1 | + | + | + | + | 4 |
| MNV | + | + | + | + | 4 | |
| Bacteria | Campylobacter genus | – | + | + | – | 2 |
| Helicobacter genus | + | + | + | + | 4 | |
| H. bilis | – | + | – | – | 1 | |
| H. ganmani | + | – | + | + | 3 | |
| H. hepaticus | – | – | + | – | 1 | |
| H. mastomyrinus | + | – | + | – | 2 | |
| H. rodentium | – | + | – | – | 1 | |
| H. typhlonius | – | + | + | – | 2 | |
| P. mirabilis | + | + | – | – | 2 | |
| S. aureus | + | + | – | + | 3 | |
| R. heylii | + | – | + | + | 3 | |
| R. pneumotropicus | – | – | + | – | 1 | |
| Parasites | C. muris | – | + | + | – | 2 |
| Demodex | – | – | + | – | 1 | |
| Entamoeba | – | + | + | + | 3 |
Racks were considered positive (+) or negative (–) for an agent if any of the testing media (Reemay filter papers, exRF, and SBS) detected a particular pathogen.
10-min duration tumbling.
1 mo-cRF.
The 1 mo-cRF detected the highest mean percentage of overall pathogens (84 ± 9%; 95% CI = 64% to 103%, P < 0.001) (Table 2 and Figure 2) and viral pathogens (88%) (Table 3). A mean percentage of bacterial and parasitic pathogens were detected at 82% and 67% respectively (Table 3). Among the 4 racks sampled, pathogens detected were MNV, Campylobacter spp., Helicobacter genus, H. bilis, H. ganmani, H. hepaticus, H. mastomyrinus, H. typhlonius, R. heylii, R. pneumotropicus, Entamoeba spp., P. mirabilis, Astrovirus-1, and C. muris (Table 4).
| Diagnostic | Rack 1 | Rack 2 | Rack 3 | Rack 4 | Overall | Standard error | P Value | 95% Confidence limits |
|---|---|---|---|---|---|---|---|---|
| Monthly ReemayTUM10 | 75% | 82% | 92% | 86% | 84% | 9% | <0.0001* | 64% to 103% |
| Monthly ReemayTUM60 | 75% | 91% | 92% | 57% | 79% | 9% | <0.0001* | 59% to 98% |
| 3-Month ReemayTUM10 | 25% | 82% | 85% | 71% | 66% | 9% | <0.0001* | 46% to 85% |
| 3-Month ReemayTUM60 | 25% | 73% | 85% | 71% | 63% | 9% | <0.0001* | 44% to 83% |
| Exhaust ReemayTUM10 | 62% | 82% | 77% | 71% | 73% | 9% | <0.0001* | 54% to 93% |
| Exhaust ReemayTUM60 | 38% | 82% | 77% | 57% | 63% | 9% | <0.0001* | 44% to 83% |
| Adhesive swabsTUM10 | 25% | 0% | 0% | 29% | 13% | 9% | 0.1690 | −6% to 33% |
| Adhesive swabsTUM60 | 25% | 0% | 15% | 57% | 24% | 9% | 0.0164* | 5% to 44% |
| SBS | 25% | 45% | 31% | 71% | 47% | 9% | <0.0001* | 27% to 66% |
Citation: Journal of the American Association for Laboratory Animal Science 63, 3; 10.30802/AALAS-JAALAS-23-000053
| Testing media | Viral pathogen detection | Bacterial pathogen detection | Parasitic pathogen detection |
|---|---|---|---|
| Monthly ReemayTUM10 | 88%* | 82%* | 67%* |
| Monthly ReemayTUM60 | 63%* | 83%* | 75%* |
| 3-Month ReemayTUM10 | 63%* | 66%* | 67%* |
| 3-Month ReemayTUM60 | 75%* | 63%* | 54%* |
| Exhaust ReemayTUM10 | 75%* | 75%* | 42%* |
| Exhaust ReemayTUM60 | 63%* | 67%* | 42%* |
| Adhesive swabsTUM10 | 38% | 4% | 0% |
| Adhesive swabsTUM60 | 50%* | 23%* | 25% |
| SBS | 50%* | 46%* | 38%* |
| Pathogen | 10-min Tumbler | 60-min Tumbler | Racks positive | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| 1 mo-cRF | 3 mo-cRF | exRF | AS | 1 mo-cRF | 3 mo-cRF | exRF | AS | SBS | ||
| MNV | + | + | + | + | + | + | + | + | + | 4 |
| Campylobacter | + | + | + | – | + | + | + | – | – | 2 |
| Genus | ||||||||||
| Helicobacter | + | + | + | – | + | + | + | + | + | 4 |
| Genus | ||||||||||
| H. bilis | + | + | + | – | + | + | + | – | – | 1 |
| H. ganmani | + | + | + | – | + | + | + | + | + | 3 |
| H. hepaticus | + | + | + | – | + | + | + | – | + | 1 |
| H. mastomyrinus | + | + | + | – | + | + | + | – | + | 2 |
| H. rodentium | – | – | – | – | + | – | – | – | – | 1 |
| H. typhlonius | + | + | + | – | + | + | + | – | + | 2 |
| S. aureus | – | – | – | – | – | – | – | – | + | 3 |
| R. heylii | + | + | + | – | + | + | + | + | – | 3 |
| R. pneumotropicus | + | + | + | – | + | + | + | – | – | 1 |
| Demodex | – | – | – | – | + | – | – | – | – | 1 |
| Entamoeba | + | + | + | – | + | + | + | – | + | 3 |
| P. mirabilis | + | + | + | + | + | + | + | + | + | 2 |
| Astrovirus-1 | + | + | + | + | + | + | + | + | + | 4 |
| C. muris | + | + | + | – | + | + | + | – | + | 2 |
Racks are considered positive (+) for an agent if any of the methods of detection detected a particular pathogen. 1 mo-cRFL filter: Paper replaced monthly and pooled together and submitted as one sample. 3 mo-cRF: Single filter paper that remained the same throughout the 3-mo study. exRF: Exhaust filter that remained the same throughout the 3-mo study. AS: Biweekly added sticky swabs that were pooled together and submitted as one sample. SBS: Fur/body swab, oral swab, and fecal pellet PCR submitted from live animal SBS.
3 mo-cRF.
The 3 mo-cRF detected an overall average of 66 ± 9% of pathogens (95% CI = 46% to 85%; P < 0.001) (Table 2). The 3 mo-cRF detected 63% of viral pathogens, 66% of bacterial pathogens, and 67% of parasitic pathogens (Table 3). Among the 4 racks sampled, the 3-mo filter papers detected MNV, Campylobacter spp., Helicobacter genus, H. bilis, H. ganmani, H. hepaticus, H. mastomyrinus, H. typhlonius, R. heylii, R. pneumotropicus, Entamoeba spp., P. mirabilis, Astrovirus-1, and C. muris (Table 4).
ExRF.
The exRF detected an overall average of 73 ± 9% (95% CI = 54% to 93%, P < 0.001) (Table 2). The exRF papers detected 75% of viral pathogens, 75% of bacterial pathogens, and 42% of parasitic pathogens (Table 3). Among the 4 racks sampled, the exhaust filter papers detected MNV, Campylobacter spp., Helicobacter genus, H. bilis, H. ganmani, H. hepaticus, H. mastomyrinus, H. typhlonius, R. heylii, R. pneumotropicus, Entamoeba spp., P. mirabilis, Astrovirus-1, and C. muris (Table 4).
AS.
The AS detected an overall average of 13 ± 9% of pathogens (95% CI = –6% to 33%, P = 0.1690) (Table 2 and Figure 2). AS detected the lowest mean percentage of viral (38%), bacterial (4%), and parasitic (0%) pathogens (Table 3). Among the 4 racks sampled, the biweekly added AS detected MNV, Helicobacter genus, P. mirabilis, and Astrovirus-1 (Table 4).
60-min duration tumbling.
1 mo-cRF.
The 1 mo-cRF detected an overall average 79 ± 9% of pathogens (95% CI = 59% to 98%, P < 0.001) (Table 2). 1 mo-cRF detected 83% of bacterial pathogens and 75% of parasitic pathogens (Table 3). Among the 4 racks sampled, the pathogens detected were MNV, Campylobacter spp., Helicobacter genus, H. bilis, H. ganmani, H. hepaticus, H. mastomyrinus, H. rodentium, H. typhlonius, R. heylii, R. pneumotropicus, Demodex sp., Entamoeba spp., P. mirabilis, Astrovirus-1, and C. muris (Table 4).
3 mo-cRF.
The 3 mo-cRF detected an overall average of 63 ± 9% of pathogens (95% CI = 44% to 83%, P < 0.001) (Table 2). 3 mo-cRF detected 75% of viral, 63% of bacterial, and 54% of parasitic pathogens (Table 3). Among the 4 racks sampled, the detected pathogens were MNV, Campylobacter spp., Helicobacter genus, H. bilis, H. ganmani, H. hepaticus, H. mastomyrinus, H. typhlonius, R. heylii, R. pneumotropicus, Entamoeba sp., P. mirabilis, Astrovirus-1, and C. muris (Table 4).
ExRF.
The exRF detected an overall average of 63 ± 9% of pathogens (95% CI = 44% to 83%, P < 0.001) (Table 2). ExRF detected 63% of viral, 67% of bacterial, and 42% of parasitic pathogens (Table 3). Among the 4 racks sampled, detected pathogens were MNV, Campylobacter spp., Helicobacter genus, H. bilis, H. ganmani, H. hepaticus, H. mastomyrinus, H. typhlonius, R. heylii, R. pneumotropicus, Entamoeba sp., P. mirabilis, Astrovirus-1, and C. muris (Table 2).
AS.
The AS detected an overall average of 24 ± 9% of pathogens (95% CI = 5% to 23%, P = 0.0164) (Table 2). AS detected 23% of bacterial pathogens and 25% parasitic pathogens (Table 3). Among the 4 racks sampled, detected pathogens were MNV, Helicobacter genus, H. ganmani, R. heylii, P. mirabilis, and Astrovirus-1 (Table 4).
Soiled-bedding sentinels.
The SBS detected an overall average 47 ± 9% of pathogens (95% CI = 27% to 66%, P < 0.0001) (Table 2). SBS detected 50% of viral, 46% of bacterial, and 38% of parasitic pathogens (Table 3). Among the 4 racks sampled, detected pathogens were MNV, Helicobacter genus, H. ganmani, H. hepaticus, H. mastomyrinus, H. typhlonius, S. aureus, Entamoeba sp., P. mirabilis, Astrovirus-1, and C. muris (Table 4).
Evaluation of tumbler device.
The tumbler devices were easy to use and performed automated agitation of pooled samples of soiled bedding. The D-sized batteries used in the device did not require changing during the 3-mo study. The timers for the tumbler devices were located in front of the cage (Figure 1D), allowing easy access to turn on the tumblers without having to open the cage. While the tumbler devices overall functioned well, a few technical issues occurred during the 3-mo study. For the tumbler devices placed on racks that were close to full cage capacity, soiled bedding would spill outside of the tumbler device during rotation (Figure 3). This was remedied by decreasing the amount of soiled bedding transferred, not to exceed 75 teaspoons (1.5 cups). The other technical issue encountered was that sticky swabs fell into the periphery of the tumbler device (Figure 4).
Citation: Journal of the American Association for Laboratory Animal Science 63, 3; 10.30802/AALAS-JAALAS-23-000053
Citation: Journal of the American Association for Laboratory Animal Science 63, 3; 10.30802/AALAS-JAALAS-23-000053
Discussion
This study examined the efficacy of a novel tumbler device that contained both contact (1 mo-cRF, 3 mo-cRF, and AS) and exhaust air media (exRF) as compared with SBS for rodent health monitoring programs.
Among the 4 racks tested, contact media (1 mo-cRF and 3 mo-cRF) in both TUM10 and TUM60 (with the exception of AS in TUM10) were successful in detecting MNV, Campylobacter spp., Helicobacter genus, H. bilis, H. ganmani, H. hepaticus, H. mastomyrinus, H. typhlonius, R. heylii, R. pneumotropicus, Entamoeba spp., P. mirabilis, Astrovirus-1, and C. muris.
1 mo-cRF that were pooled and submitted as one sample at the end of the 3-mo study detected the highest overall percentage of pathogens identified on the 4 racks in both TUM10 and TUM60 (Table 2). 1 mo-cRF in TUM60 detected H. rodentium and Demodex, which were not detected by any other method.
Overall, AS were the least effective method of pathogen detection of all the methods tested, including SBS. AS in TUM60 detected more pathogens, MNV, H. ganmani, and R. heylii than did AS in TUM10, which is to be expected given that AS were in contact with agitated soiled bedding longer after 60 rather than 10 min of exposure. However, some AS fell outside of the tumbler during rotation, thus limiting their contact with the soiled bedding (Figure 4). The devices have walls that help to contain the soiled bedding within the tumbler (Figure 1C), but devices that were loaded with close to the maximum of 100 g of soiled bedding had more AS loss than did devices on racks that held only 25% of the rack cage capacity. We addressed the problem of AS falling into the periphery by limiting the maximum amount of soiled bedding to 75 g. A recent publication evaluated detection of an opportunistic pathogen in manually shaken cages that contained both filter paper and AS and found that AS was not effective in detecting pathogens, possibly due to decreased adhesive quality over time.30 In our study, even with the addition of a new AS at biweekly cage changes, AS was still the worst method overall for pathogen detection, even as compared with SBS. This finding is inconsistent with the previous study, indicating that the results obtained using AS added at every cage change and then pooled together at the end of the 3-mo study period were deemed a viable alternative to SBS based on the comparable to superior pathogen detection.8
Helicobacter spp. (including H. bilis, H. ganmani, H. hepaticus, H. mastomyrinus, and H. typhlonius) were detected on both contact and exhaust Reemay filters in both TUM10 and TUM60. 1 mo-cRF in TUM60 was the only media that detected H. rodentium. SBS reliably detected Helicobacter spp., H. ganmani, H. hepaticus, and H. typhlonius but did not reliably detect as many Helicobacter species as the contact and exhaust Reemay filter paper in both the TUM10 and TUM60. A previous study demonstrated H. hepaticus was reliably detected on exhaust air particles at a low prevalence compared with SBS.16 Although not always excluded from standard vivaria, detection of Helicobacter spp.22 is important due to the potential of inciting inflammatory bowel disease, gastritis, and gastrointestinal disorders, particularly in immunodeficient mice.4
1 mo-cRF in TUM60 was the only media that detected Demodex, which are small, cigar-shaped mites that principally feed and reproduce in hair follicles and adnexal glands of mammals. Although Demodex are not often assessed or excluded with conventional colony animals, infestations with this parasite can lead to ocular lesions and ulcerative dermatitis in immunocompromised animals18 A low copy number of 12 (data not shown) could result in a false positive result because Demodex mites are traditionally diagnosed via deep skin scrape or fur plucking.19 All positive samples were confirmed via the diagnostic laboratory by repeating the real-time PCR amplification to confirm the initial result, but the low copy number and negative findings for other sample media or racks prevented a conclusion concerning which sample medium was most effective at detecting this particular parasite. A future study could test soiled bedding from mice known to be positive for Demodex to study this issue.
Rodentibacter (R. pneumotropicus and R. heylii) was detected on all filter paper media in both TUM10 and TUM60 but was not detected on AS and SBS. Previous studies have shown that Rodentibacter spp. are not reliably detected in SBS due to the failure of the bacteria to survive long enough in soiled bedding to be infectious.25 In addition, exhaust air dust PCR is a more sensitive way to detect this bacteria in IVC caging systems.17 In our study, 1 mo-cRF, 3 mo-cRF, and exRF all could detect both agents, likely due to the presence of bacterial DNA on the filter paper.21
The only pathogen detected by SBS that was not detected with nonanimal media was S. aureus. The sentinel mice were not considered to be specifically free of S. aureus based on health reports from the vendor, and they could have been opportunistically contaminated after receipt; these factors could explain why S. aureus was detected only in the SBS. Pathogens such as Campylobacter spp., H. bilis, H. rodentium, R. heylii, R. pneumotropicus, and Demodex were detected by the nonanimal media but were not detected by SBS (Table 4).
In comparing pathogen detection by nonanimal and SBS methods, the tumbler devices overall detected more pathogens than did SBS. 1 mo-cRF in TUM10 detected the highest percentage of viral and overall pathogens (Table 2 and Figure 2) while 1 mo-cRF in TUM60 detected the highest percentage of bacterial and parasitic pathogens (Table 3). Perhaps replacing the Reemay filter papers monthly avoids saturation of the filters that might occur when they remain in the tumbler for the entire 3-mo duration and subsequently allows more pathogens to bind to it.
Tumbler devices can hold up to 100 g of soiled bedding per device, so when used in routine operation, one tumbler device should be used per 100 cages. In this study, soiled bedding was kept to a maximum of 75 g to prevent spilling of bedding, but future modifications of the device may allow for containment of up to 100 g of bedding. We discarded the tumblers at the completion of the 3-mo study and therefore did not evaluate decontamination methods. Future studies could examine and verify various methods of decontamination to reduce the possibility of detecting residual DNA and obtaining false positive results.15,23 Because the tumbler we tested is a patent-pending prototype, its final cost has not yet been determined because design modifications and the final choice of material may influence the final price. A tumbler cylinder that can be autoclaved may be more expensive than one that can be disposed of and replaced every quarter.
Limitations to the study were the sample size and the limited number of tumblers used. Further studies could test replicates on each rack to aid in differentiating true and false positives. Technical issues such as soiled bedding spilling out the sides of the tumbler (Figure 3) or spent batteries could interfere with detection of a pathogen. However, future modifications could prevent these problems in the future. Future studies could also compare different collection media to determine which would bind infectious agents most effectively. The tumbler device should also be tested with regard to other rodents (for example, rats).
The data we obtained in this study are not sufficient for a recommendation of the optimal interval for reliable pathogen detection. 1 mo-cRF appears to have detected more pathogens because replacing it monthly and submitting multiple swabs as one sample prevent saturation of the Reemay filter papers. The exRF detected more pathogens than did SBS, but Reemay filter paper in contact with the soiled bedding during agitation may detect more pathogens. Our results are informative because quarterly testing of environmental samples or SBS is commonly used in rodent health monitoring programs. However, when we evaluated the efficacy of the tumbler device compared with the use of SBS, we found that these devices are a valid replacement for the use of live animals in mouse health monitoring programs in standard SPF colonies.
Tumbler device situated inside the Optimice IVC cage. The tumbler is positioned and secured with 2 black rubber bands on both sides indicated by the green arrow in front of the exRF at the rear of the cage (A). Tumbler device is supported on rotating rods connected to a D-battery source (B). Reemay filter paper and AS are placed inside the tumbler device along with soiled bedding (C). Exterior front of cage with the on/off and timer switch that fits on the Optimice IVC cage card plastic holder and red arrow pointing to the arrow indicator imprinted on the tumbler cylinder (D).
Percent pathogen detection of viral agents, bacterial agents, parasitic agents, and overall pathogen detection averaged over 4 racks sampled among various methodologies Whiskers indicate the minimum and maximum values whereas boxes represent the upper and lower interquartile range. Circles represent the outliers in the data. Within the box, the X represents the mean average of pathogen detection whereas the horizontal line represents the median.
Tumbler device showing soiled bedding spilling outside into the cage.
Tumbler device pictured shows AS stuck in the periphery during biweekly soiled bedding change.
Contributor Notes