Introduction

The use of IVC rack systems to house rodents in biomedical research has become the standard of care in recent decades. In these systems, air is filtered and then supplied to each cage such that cages do not share supply air; exhaust air is filtered as it exits the plenum. Therefore, each cage is a distinct microbiologic unit¹⁰ that prevents transmission of infectious agents between cages and subsequently results in a lower prevalence and spread of adventitious pathogens in rodent colonies.³ Historically, the standard indirect method of detection for rodent pathogens was the use of live soiled-bedded sentinel (SBS) mice that are exposed to dirty bedding for a specific period and then submitted for serology, culture, and visual detection methods to determine the presence of adventitious agents in the sentinels, and, by inference, in other mice in their proximity. Successful detection of infectious agents by the SBS method relied on the sentinel animal becoming infected with pathogens present in the soiled bedding. This method has several major disadvantages that include the use of live mice, dependency on infectious or viable pathogen shedding by infected mice during the interval of exposure, and transfer of infection through soiled bedding.²¹ Several rodent pathogens are known to remain undetected when using SBS method of colony health surveillance, including lymphocytic choriomeningitis virus, Sendai virus, adenovirus, Rodentibacter spp. (formerly Pastuerella spp.), and others.^{2,4,7,14,17,25}

In the past 20 y, the use of molecular methods such as the PCR for the detection of rodent pathogens has increased in animal research facilities, often replacing serological or microscopic and culture-based methods.^6,8 PCR relies on detection and amplification of nucleic acids and therefore can detect with high sensitivity and specificity the presence of an adventitious agent even with very low shedding.^{4,6,12,20,21,27,28} Concurrently, IVC rack systems have evolved to include both reusable and disposable caging. Reusable caging is typically made of polycarbonate or polysulfone material that is strong, optically transparent, and resistant to heat and chemicals, making them attractive for long-term use. Alternatively, disposable caging is made from more flexible plastic such as polypropylene or polyethylene, which is lighter and can be recycled yet is also optically transparent.^1,15 A common feature of disposable caging includes a cage-level disposable Reemay filter at the exhaust air exit before reaching the exhaust plenum; this feature not used for reusable cages (Figure 1). For the purposes of this article, we have defined this filter as SBS cage–level exhaust filter, or SCEF.

View Figure

• Download as Powerpoint
• Download Figure

Because of the differing cage types, various sampling procedures have been used as PCR has gained popularity. These include direct testing of research animals by ante mortem sampling including blood or fecal pellets, body swabs, and oral swabs; exhaust air dust from IVC rack prefilters; exhaust air dust swabbed from rack plenums; exhaust air dust from in-line plenum media; exhaust air dust from in-cage media; and exhaust air dust from in-cage filters.^{9,20,21,22,23,27,28}

Our institution, like many academic programs, uses a variety of caging systems including both reusable and disposable cages, IVC and static systems, metabolic chambers, environmental chambers, and others. While a variety of systems allows our institution to maximally serve our researcher population, it also poses challenges for establishing a rodent colony surveillance program that is reliable, efficient, and cost effective. Therefore, we sought to compare SBS methods (sample collection for serology and PCR) with PCR of exhaust air dust EDT from a filter in the downstream vertical plenum, and PCR of SCEF in disposable cages, for mouse health surveillance. Our hypothesis was both PCR methods would outperform SBS for detection of mouse pathogens.

Materials and Methods

All housing and experimental use of mice were carried out in AAALAC-accredited facilities in accordance with US federal, state, local, and institutional regulations and guidelines governing the use of animals²⁴ and were approved by the University of Oklahoma Health Sciences Center IACUC.

This study was conducted in the rodent barrier facility, which is dedicated to breeding rodent colonies for maintenance and expansion to support biomedical research. Dedicated personnel are assigned to each room to prevent cross-contamination. All mice used in this study were negative for agents on our facility exclusion list before the start of study based on quarterly SBS testing and vendor-provided health reports (Table 1).

Husbandry staff performed daily animal health checks and environmental monitoring (mouse activity, food, water, temperature, and humidity), and veterinary staff performed health checks at least weekly. Routine rodent health monitoring is performed quarterly (basic pathogen testing) and annually (comprehensive pathogen testing) based on the facility exclusion list (Table 1). To capture as much information regarding rodent pathogen status as possible for purposes of this study, we tested for several agents that are not currently excluded from our rodent colonies (Table 2).

Ninety female CD-1 mice (Mus musculus) aged 4 to 6 wk were purchased from Charles River Laboratories (Wilmington, MA) for use as sentinel mice. Among the agents tested for this study, sentinel mice were positive only for Staphylococcus aureus upon receipt. CD-1 outbred stocks have been used historically as SPF sentinel rodents at our institution because they are generally good serologic responders.²⁶ SBS mice were cohoused in disposable IVC (Allentown EasyCage, Allentown, PA). All cages contained irradiated 1/8” pelleted cellulose bedding (BioFresh, Scott Pharma Solutions, Marlborough, MA). Mice were fed irradiated mouse diet (Purina LabDiet 5053, St. Louis, MO) ad libitum and received reverse osmosis–purified water via automatic watering system (Edstrom Industries, Waterford, WI). Environmental enrichment included paper nesting material (Bed-r’Nest puck, The Andersons, Delphi, IN) and a red plastic shelter (BioServ, Minneapolis, MN). Cages, racks, enrichment items, and supplies were either irradiated or autoclaved before use. All mice were handled in either animal transfer stations (Allentown Phantom2, Allentown, PA) or class 2 biosafety cabinets (NuAire LabGard ES AIR, Class II, Type A2 Biosafety Cabinet, Plymouth, MN).

All animal holding and procedure rooms were maintained on an automated building system monitor (Edstrom Avidity, Edstrom Industries, Waterford, WI) that automatically monitors room temperature, humidity, light cycle, and water flow and sends alarm updates for parameters outside of established acceptable ranges. Room conditions were a 14-h light and 10-h dark cycle, temperature 72 ± 2 °F (22.2 ± −16.7 °C), 40 to 70% relative humidity, 120 to 300 l× light intensity, and 15 air changes per hour under positive pressure (15 to 22 Pa).

Twenty-five individually ventilated mouse racks (NexGen, Allentown, PA) housed breeding mice in disposable cages (EasyCage, Allentown, PA). Racks were either double sided (n = 17) or single sided (n = 8). Before racks were used to house mice, rack plenum swabs were collected with flocked swabs provided by Charles River Laboratories (Wilmington, MA), stored in sterile 50 mL conical tubes, and then shipped to Charles River Laboratories (Wilmington, MA) for PCR testing for excluded rodent pathogens. At the beginning of the 3-mo monitoring period, EDT filters were placed on each IVC rack (Sentinel EAD, Allentown, PA) in the exhaust plenum (Figure 2). SBS cages were placed on each side of the rack on the bottom right cage slot and left. At each 3-wk change of cage bottoms, approximately 1 tablespoon of bedding from each cage on the rack was added to the sentinel cage. At each 6-wk cage lid change, the manufacturer-placed filter paper covering the exhaust air port in the disposable cage lid of SBS cages (approximate 2 × 2 cm size) was aseptically removed using a sterile scalpel blade (10 blade; Covetrus North America, Portland, ME) and placed in a sterile 50-mL conical tube (Thermo Fisher Scientific, Dallas, TX). At our institution, cage bottoms are changed every 3 wk and cage lids are changed every 6 wk, which resulted in 4 filters submitted as pooled samples for double-sided racks and 2 filters submitted as pooled samples for single-sided racks. The total numbers of test samples submitted for SCEF, EDT and SBS were 25, 25, and 42, respectively. After 3 mo, blood was collected from mice via cheek bleed using a lancet, pooled on the dried blood spot card (Charles River Laboratories, Wilmington, MA) for each side of the rack, and submitted for antibody detection by serology for bacterial and viral agents. Fecal samples were collected from bedding of each SBS cage for PCR detection of helicobacter species and pinworms, and fur swabs were collected from each SBS cage for PCR detection of fur mites. The EDT filters and SCEF were collected; SCEF were pooled for each rack prior for submission to the diagnostic laboratory (Charles River Research Animal Diagnostic Services, Wilmington, MA).

View Figure

• Download as Powerpoint
• Download Figure

All samples were stored at 4 °C before shipment to the testing laboratory, where they were evaluated for multiple murine pathogens (Table 2). Samples for PCR and serology were submitted to a commercial diagnostic laboratory (Charles River Research Animal Diagnostic Services, Wilmington, MA). Briefly, for PCR testing, total nucleic acid was isolated from both rodent-derived and filter samples via an established automatic magnetic isolation protocol and PCR analysis was performed by using qualified, proprietary TaqMan PCR assays. Internal sample suitability controls were used to monitor all samples for PCR inhibition.¹³ PCR inhibition was not detected in any of the samples. All positive results were repeated to confirm reproducibility.

Serologic testing was performed to detect the presence of pathogen antibodies in blood samples from SBS by using established and validated Multi-Fluorometric Immunoassays.²⁹ All positive findings were repeated and confirmed by repeat assays.

Statistical analysis was performed using GraphPad Prism version 10.1.0 for Windows (GraphPad Software, Boston, MA). A one-tailed t test was used to compare the number of positive detections by agent and the percentage of positive racks by agent between SCEF and SBS and between EDT and SBS. A confidence interval of 95% and a P value < 0.05 were used to assess significance.

Results

Table 3 summarizes agents detected, positive racks detected, and relevant detection methods. Nine different infectious agents were detected in total. Five different infectious agents (1 virus, 3 bacteria, 1 protozoan) were found on SCEF. Eight different infectious agents (2 viruses, 4 bacteria, 2 protozoa) were found in EDT filter media. Three different agents (1 virus, 1 bacteria, 1 protozoan) were detected among SBS by serology and PCR. PCR on SCEF and EDT consistently detected agents more often than did SBS (P = 0.03 and P = 0.04, respectively). PCR on SCEF and EDT also consistently detected agents on more racks than did SBS (P = 0.003 and P = 0.001, respectively). PCR on EDT detected a greater variety of agents (n = 8) than either SCEF or SBS (n = 5 and n = 1, respectively).

Among racks tested, no infectious agents were detected by any one method on 3 racks, and at least one infectious agent was detected on 22 racks. Among positive cage racks from which at least one agent was detected, the percentage of positive detection by SCEF, EDT, and SBS was 91%, 16%, and 9%, respectively (Table 3). Because 2 SBS cages were tested on double-sided racks compared with only one pooled SCEF and one EDT, SBS was scored as a positive if either SBS was positive on either side of the rack. Although this provides a calculation bias in favor of SBS, it removes calculation bias favoring single positives for SCEF and EDT if only 1 of 2 SBS was positive.

Discussion

As an alternative to SBS for infectious agent detection, we investigated PCR on SCEF and EDT for environmental health monitoring in disposable IVC mouse cages. As compared with SBS, we found that both SCEF and EDT sampling for PCR are superior to SBS evaluated by PCR and serology, as both testing methods outperformed SBS in both frequency of detection and number of agents detected. This finding is consistent with the growing body of literature describing nonanimal PCR-based methods of colony health surveillance.^{3,4,7,9,5,11,16,18,19,21,22,28} One group recently evaluated filter media surveillance as a replacement for SBS methods. Our findings corroborate theirs, suggesting that PCR of filter media is superior to SBS for detection of adventitious rodent pathogens in a large health-monitoring program.²⁸ Our data indicate that cage-level filter testing (SCEF) may be preferable to plenum filter testing (EDT) for viral agents, as SCEF detected Astrovirus-1 more often (7 of 25 samples) than did EDT filter media (1 of 25 samples). Although both the EDT filter media and SBS methods detected mouse Norovirus, only one positive sample was found and the copy number was low (a single copy), suggesting contamination rather than viral shedding (data not shown). Astrovirus was used as a surrogate marker for viruses because it is persistently shed at high copy numbers. SCEF detected Astrovirus in 88% of the cages (7 of 8), whereas EDT detected Astrovirus in only 13% of the cages (1 of 8). Bacteria were detected well by both SCEF and EDT. Either method of environmental monitoring may be acceptable for routine surveillance of colonies housed in disposable cages. However, if concern arises for low prevalence of colonization by opportunistic bacteria, direct sampling of mice may provide better detection. SCEF and EDT detected protozoa at the same frequency, but EDT detected 2 types of protozoa where SCEF detected only one. Tritrichomonas was detected in high copy numbers (50 to 1,630) in EDT and theoretically should have been detected in the SCEF as well. In addition, SCEF had positive results more often than did EDT; the cage-level filter may remove some agents before they reach the rack plenum. Some agents may not be able to exit the cage because of their size and the cage-level exhaust filter. More investigation could help identify an explanation for the observed discrepancy.

Because we performed this study in an active research environment, some limitations were inherent in the study design and some variables were unknown or uncontrolled. Because of a screening and exclusion biosecurity program for infectious agents, the racks used in the study contained a limited number of agents. Testing mice with a greater array of known microbiologic agents could provide better insight into the advantages of the 2 nonanimal screening methods for routine colony surveillance. The prevalence for each agent on racks used in this study were unknown. Therefore, we could not evaluate the influence of rack prevalence on detection by any method. Furthermore, our facility does not routinely test for several of the agents that were detected by either SCEF or EDT; therefore, we cannot assess how well these methods detect adventitious agents over a longer sampling period. Another consideration is that the agents we detected are shed persistently; agents that are cleared by the immune system over time would not be represented in this study.

Future studies focused on agents that are frequently implicated as contributing to aberrant research results (e.g., Helicobacter spp.) would be beneficial. Our facility operates a Helicobacter spp.–negative breeding facility; however, the agent is not excluded from conventional spaces. This means that any given testing period may have positive Helicobacter spp. results from pooled fecal samples. Because the majority of conventional mice are transferred out of the rodent barrier facility, these positive results are uncommon and intermittent. Establishing the frequency of positive tests of known positive samples would help us to identify our true prevalence rate and better guide selection of the type of housing needed.

Choosing a method for colony health surveillance should include consideration of the ease of testing, technical compliance, and staff time needed perform sampling. We found that SCEF required more labor than simply retrieving EDT from the rack plenum because the former required sterile instruments, sterile conical tubes, and careful aseptic handling of the cage tops in a biosafety cabinet. SCEF also required placement of a sentinel cage and labor to transfer dirty bedding at each cage change. Exhaust air dust filter media that can be placed directly in the exhaust plenum has the advantages of easy placement with no additional labor needed until retrieval. This method is also free of variation in the volume of dirty bedding collected from individual cages, mixing technique, and cage change interval. Both SCEF and EDT require less labor than does the SBS methods. Labor required for SBS monitoring includes placing dirty bedding the sentinel cages at each cage change and the need to collect blood, feces, and pelt samples quarterly from each sentinel mouse. We estimate that the time required for SBS sample collection for this study, which included only 45 sentinel cages and 90 mice, was approximately 30 h when scaled for a large facility with a high rodent census; the time needed to use the SBS method can be a large component of normal husbandry operations.

In conclusion, our data indicate both methods of environmental health monitoring are superior to SBS in disposable cages, which supports ethical animal use in accordance with the 3R’s. While more information is needed regarding detection of unsuspected organisms and monitoring in various caging types, our institution has moved toward environmental monitoring in lieu of live animal testing for rodent colony health surveillance based on the data in this study.