Abstract
Sampling of bedding debris within the exhaust systems of ventilated racks may be a mechanism for detecting murine pathogens in colony animals. This study examined the effectiveness of detecting pathogens by PCR analysis of exhaust debris samples collected from ventilated racks of 2 different rack designs, one with unfiltered air flow from within the cage to the air-exhaust pathway, and the other had a filter between the cage and the air-exhaust pathway. For 12 wk, racks were populated with either 1 or 5 cages of mice (3 mice per cage) infected with one of the following pathogens: mouse norovirus (MNV), mouse parvovirus (MPV), mouse hepatitis virus (MHV), Helicobacter spp., Pasteurella pneumotropica, pinworms, Entamoeba muris, Tritrichomonas muris, and fur mites. Pathogen shedding by infected mice was monitored throughout the study. In the filter-containing rack, PCR testing of exhaust plenums yielded negative results for all pathogens at all time points of the study. In the rack with open air flow, pathogens detected by PCR analysis of exhaust debris included MHV, Helicobacter spp., P. pneumotropica, pinworms, enteric protozoa, and fur mites; these pathogens were detected in racks housing either 1 or 5 cages of infected mice. Neither MPV nor MNV was detected in exhaust debris, even though prolonged viral shedding was confirmed. These results demonstrate that testing rack exhaust debris from racks with unfiltered air flow detected MHV, enteric bacteria and parasites, and fur mites. However, this method failed to reliably detect MNV or MPV infection of colony animals.
Abbreviations: MHV, mouse hepatitis virus; MNV, mouse norovirus; MPV, mouse parvovirus
The use of sentinel mice, housed on dirty bedding from colony mice, has been a cost-effective way of monitoring for the presence of adventitious rodent pathogens in large mouse colonies. Historically, sentinels were housed in open-top caging on the bottom row of racks, and dirty bedding was added to facilitate the detection of infectious agents transmissible by airborne and fecal–oral routes.14 With the advent of microisolation caging and IVC, infectious disease containment has become possible at the cage level. This practice has the benefit of reducing transmission of infectious agents and decreasing the overall prevalence of an infectious agent within a colony. However, with improved disease containment, several infectious agents were recognized to be inefficiently transmitted to sentinel animals through contact with soiled bedding. These agents include Helicobacter spp., Pasteurella pneumotropica, and fur mites, which lose viability outside of the host, and respiratory pathogens, such as Sendai virus, which is not transmitted by the fecal–oral route.1,3,4,8,9,12,15 In addition, the efficiency of disease transmission to the sentinel mice depends on many factors, including dose and stability of the pathogen in the dirty bedding, amount of dirty bedding transferred, immune competence, and other physiologic factors of the sentinel animals. Some agents that historically have been readily detected using dirty bedding sentinels may go undetected given the right circumstances. For example, mouse hepatitis virus (MHV) was readily detected by using dirty bedding transfer in a number of studies using IVC systems and dirty-bedding sentinels.2,3,5 However, more detailed experiments conducted to look at variables affecting MHV transmission to sentinel mice revealed that several factors, including duration of virus shedding and timing of bedding transfer, affected transmission and could lead to suboptimal detection.13 To compensate for the weaknesses of dirty-bedding sentinel monitoring, samples collected directly from colony mice and from environmental sites have been used.
Adjunct testing methods exploiting the common exhaust airflow of the IVC system have been investigated and continue to develop as adjunct testing for infectious disease monitoring. Two different approaches have been examined. One approach has been to alter the exhaust airflow in the IVC system design such that the exhaust air from the entire rack is provided to dedicated air-exhaust sentinel cages within the IVC system. One group found that a combination of the air-exhaust sentinel cage and the transfer of dirty bedding proved successful for detection of MHV, pinworms, and intestinal flagellates but failed to detect mouse parvovirus (MPV).2 Interestingly, the authors compared dirty-bedding sentinels alone with the combination of a dedicated air-exhaust sentinel cages and dirty-bedding transfer for MHV and Enteromonas spp. detection. The detection of these 2 agents did not differ, so it remains unclear whether the exhaust-air sentinel cage provided any additional detection capability over dirty-bedding sentinels in that study.2 Another study compared exhaust airflow sentinel cages with inline exhaust-air filter detection as well as contact and dirty-bedding sentinels for the detection of MHV, Sendai virus, murine rotavirus, MPV, and Helicobacter spp.3 Of the enteric pathogens, MHV was effectively transferred to sentinels by all methods; Helicobacter spp. were detected by using contact and dirty-bedding sentinels and inline exhaust filters but not in air-exhaust sentinel cages. MPV and murine rotavirus were the most difficult pathogens to detect and were only reliably detected by the use of contact sentinels. Sendai virus was detected by using the air-exhaust sentinel cage and inline filters but not by dirty-bedding sentinels. Testing inline exhaust filters for pathogens appeared to be a more versatile adjunct testing method than was screening sentinels from the air-exhaust sentinel cage.3
Subsequent studies have explored the use of PCR-based methods to detect fur mites, Corynebacterium bovis, and pinworms in bedding debris that collected on the surfaces of horizontal air-exhaust manifolds of racks with unfiltered exhaust air. In the fur mite study, a single cage of mite-infested mice was placed at the end of a row farthest from the manifold sampling site on each of 17 racks.6 Mite DNA was detected by PCR analysis of exhaust debris from more than 50% of racks within the first week of cage placement and in debris from 94% of all racks within 4 wk; debris from all racks was positive by 9 wk. Samples collected from the end of each horizontal manifold provided more reliable detection of mite DNA by PCR amplification than did a single sample collected from the top of the vertical plenum. In addition, Corynebacterium bovis was detected by using samples collected at the horizontal exhaust manifold, and cage positioning relative to the exhaust manifold was an important variable.10 The pinworm study determined that screening of exhaust debris did not confirm the extent of a recently identified infection of mice with Aspiculuris tetraptera.7 Taken together; these results highlight the sensitivity of PCR methodology in the detection of target DNA in environmental samples, even when disease prevalence is low.
These studies support the concept that exhaust debris that accumulates on the inner surfaces of the exhaust plenums may provide ideal samples for the detection of adventitious agents in rodent colonies. The published studies just discussed were all performed by using IVC racks with a similar airflow design, in which the air enters the individual cages through an open port or diffuser and exits through an unfiltered exhaust-air manifold.2,3,6,7,10 However, in other IVC rack configurations, air is exchanged in and out of cages through a filtered lid, and whether bedding debris accumulates in the exhaust manifold of these racks is unknown. The current study was designed to better understand the effectiveness of exhaust-air debris samples for microbiologic monitoring of mouse colonies and to determine how well this screening tool works on IVC racks in which exhaust air was or was not filtered at the cage level prior to entry into the exhaust-air manifold.
Materials and Methods
Racks, caging, rack sanitation and mice.
The vivarium at IDEXX BioResearch is part of an AAALAC-accredited animal care and use program at the University of Missouri. The animal studies were approved by the University of Missouri's Animal Care and Use Committee. Two single-sided rack systems with nondisposable caging (IVC Green Line System, Tecniplast, West Chester, PA), designated as racks 1 and 2, and one single-sided rack with disposable caging (Innovive, San Diego, CA), designated as rack 3, were used in the study. These IVC racks were sanitized according to routine manufacturer recommended acid and alkaline detergent wash cycles in a rack washer (900RW, Tecniplast), evaluated by PCR assay, and found to be free of all infectious agents of interest prior to the beginning of the study. Racks were set to 70 air changes hourly, and rack airflow was set such that pressure within individual cages was negative relative to the room. The room was maintained with constant environmental conditions (14:10-h light:dark cycle, an average daily relative humidity of 35% to 50%, and a temperature range of 69 to 71 °F [20.6 to 21.7 °C]). Irradiated commercial diet (Teklad 2918 diet, Envigo, Indianapolis, IN) and autoclaved acidified water were provided free choice.
Nondisposable cages were bedded with autoclaved soft-texture bedding (Paperchip, Shepherd Specialty Papers, Watertown, TN), and disposable cages were purchased prebedded (Alpha-dri, Shepherd Specialty Papers) and sterilized. Every 2 wk, cages were changed in a HEPA-filtered biosafety cabinet. The cabinet was disinfected with 10% bleach prior to and after changing groups of cages; in addition, nitrile gloves and disposable polyester gowns were changed between groups of cages. Nondisposable cages were autoclaved prior to dumping of soiled bedding and washing. Clean cages were bedded and autoclaved prior to use. The mice (Mus musculus) used in this study were either Hsd:ICR females (age, 5 to 7 wk at the time of inoculation or cohousing [Envigo, Dublin, VA]) or were inhouse-bred naturally infected Sencar females (age, 5 to 7 wk at study initiation).
Murine infections.
The infectious agents selected for this study included MHV1, mouse norovirus (MNV) 4, MPV1e, P. pneumotropica, Helicobacter spp., fur mites (Myocoptes musculinus and Radfordia affinis), pinworms (A. tetraptera and Syphacia obvelata), and large intestinal protozoa (Tritrichomonas muris and Entamoeba muris). MHV, MNV, and MPV were selected because they are among the most prevalent agents and, therefore, present the greatest biosecurity risk for mouse colonies.11 In addition, MNV is persistently shed in the feces, whereas MHV and MPV are shed transiently. P. pneumotropica, Helicobacter spp., and parasites were selected as these pathogens are relatively prevalent pathogens that are not always detected in sentinels exposed to infected bedding from colony mice.8,12
Virus-infected mice.
Groups of 10 ICR mice were inoculated by oral gavage with 0.1 mL DMEM containing MHV (2 × 105 pfu), MNV (9 × 106 pfu), or MPV (1 × 106 pfu). To mimic a natural infection, each experimentally infected mouse was then cohoused with 3 or 4 naïve ICR mice for a period of 1 wk after inoculation. The experimentally inoculated mouse was then removed from the study, and the contact-exposed ICR mice were moved to clean cages (3 mice per cage) and placed on 1 of the 3 IVC racks for the remainder of the study.
Bacteria- and parasite-infected mice.
Sencar mice naturally colonized by P. pneumotropica, Helicobacter ganmani, H. typhlonius, Myobia musculi, Myocoptes musculinus, S. obvelata, A. tetraptera, E. muris, and T. muris were housed 3 per cage in clean cages and transferred at the same time as the ICR mice to the new IVC racks.
Assignment of mice to IVC racks.
Four different groups of mice (ICR mice exposed to MHV, MNV, or MPV and one group of naturally infected Sencar mice) were housed on each IVC rack. Five cages per group (MHV, MNV, MPV, or naturally exposed group) were housed on either a rack containing nondisposable caging (that is, rack 1) or a rack containing disposable caging (that is, rack 3; Figure 1). A single cage of mice per agent (MHV, MNV, MPV, or naturally exposed group) was housed on the second rack containing nondisposable caging (rack 2). The first 2 rows of each rack contained either 5 cages of uninfected mice (racks 1 and 3) or one cage of uninfected mice (rack 2). These uninfected mice were included in the study to mimic animal housing in which a percentage of colony mice housed in a rack may be uninfected yet contribute bedding dust and debris that may serve to dilute dust and debris coming from infected cages. Mice were placed on the rack in order of lowest to highest disease transmission risk with uninfected animals in top rows and MHV and mice containing multiple infectious agents at the lower rows.
Figure 1.
Location of mouse cages on ventilated racks containing 5 cages of mice per agent tested. The image on the left is rack 1 and blower. The image on the right is rack 3. Mice were housed with one agent per row, and the numbers 1–5 indicate the location of the 5 cages per agent. Rows: a, uninfected mice; b, uninfected mice; c, MNV; d, MPV; e, MHV; and f, Sencar mice with Pasteurella pneumotropica, Helicobacter ganmani, Helicobacter typhlonius, Myobia musculi, Myocoptes musculinus, Syphacia obvelata, Aspiculuris tetraptera, Entamoeba muris, and Tritrichomonas muris. For rack 2 containing 1 cage of mice per agent, the mice were housed in column 1, with the same rows containing the same agents as rack 1 containing 5 cages.
The remainder of each rack was filled with empty cages with bedding material. Rack 1 held 70 cages, and rack 3 held 63 cages. Therefore, the percentage of infected cages for each agent on each rack was between 7% to 8% for the racks containing 5 cages per group or 1% for rack 2, containing a single cage per group.
Diagnostic testing.
Mouse testing.
All collected samples were submitted to the IDEXX BioResearch testing laboratory (Columbia, MO) for real-time PCR testing or multiplex fluorescent immunoassay serology. All IDEXX BioResearch real-time PCR assays have been validated to detect 10 or fewer copies of target DNA or RNA and all multiplex fluorescent immunoassays are validated to perform with 98% or greater specificity and sensitivity. To confirm shedding of the various agents, fecal samples were obtained from each cage of mice at the initiation of the study (collected the day after mice were transferred to new cages and placed onto the IVC rack), weekly from each cage for the first 4 wk, and at the end of the 12 wk study. Fecal samples from the virus-exposed Hsd:ICR mice were tested by real-time PCR analysis to confirm shedding of MHV, MNV or, MPV. Feces from the naturally infected Sencar mice were tested by PCR for P. pneumotropica, Helicobacter spp., pinworms, E. muris, and T. muris. Fur swabs (FLOQSwabs, Copan, Murrieta, CA) were collected on the same schedule as feces from the group of naturally infected Sencar mice and were tested by real-time PCR assay for fur mites. At the end of the 12-wk study, blood was collected from the Hsd:ICR mice to screen for antibodies to MHV, MNV, and MPV.
Rack testing.
Airflow patterns differ in each rack manufacturer and sampling locations need to be selected based on the airflow pattern of the ventilated rack system. The design of racks 1 and 2 has an open air flow, with no filter barrier between the cage and the exhaust manifolds (Figure 2 A). Air from the room enters the intake area at the top of the blower and passes through a prefilter and a HEPA filter, exits the blower into a horizontal manifold at the top of the rack, and travels down the vertical plenums to enter into the cage through an open port in the cage lid (Figure 2 A, white arrows). Exhaust air then leaves through a second open port in the cage lid and travels down a vertical exhaust plenum (Figure 2 A, black arrows). The exhaust air travels to the bottom horizontal manifold and enters into the lower exhaust portion of the blower passing through a prefilter and then a HEPA filter before being exhausted to the room. Because IVC racks 1 and 2 have the exhaust connections at the bottom of the rack and debris accumulates along the entire exhaust system, filter material (Filtrete 1500 HVAC filter, 3M, Maplewood, MN) was fitted over the rack exhaust prefilter to allow for easy access and sequential collection of exhaust debris samples (Figure 3 A). A 2-cm square piece of the filter was removed from the bottom edge of the filter at each sampling time point.
Figure 2.
(A) Racks 1 and 2 with blower airflow pattern. The white arrows show the direction and pathway of the supply air, and the black arrows show the direction and pathway of the exhaust air. The white arrowhead indicates the prefilter access door on the blower unit. The inset shows the ports through which air enters and leaves through an individual cage. (B) Rack 3 with blower airflow pattern. The white arrows show the direction and pathway of the supply air, and the black arrows show the direction and pathway of the exhaust air. The inset shows the ports through which air enters and leaves through an individual cage. Air enters through an open port on the cage lid but exits through a filtered port on the cage lid.
Figure 3.
(A) An image of the prefilter utilized in racks 1 and 2 with the attached high-efficiency filter. The black arrow indicates the 2-cm sections that were removed for testing. (B) Rack 3 customized with sampling ports at the level of each horizontal plenum. The black arrow indicates an open sampling port. For sampling, the port was opened and a swab inserted into the port; the sample was collected by rubbing the swab on the horizontal plenum surface at the junction of the horizontal and vertical plenum.
In contrast, IVC rack 3 has the exhaust blower mounted on the top of the rack rather than the bottom of the rack (Figure 2 B). In addition, the design of rack 3 has a closed air flow, with a filter barrier between the cage and the exhaust plenum (Figure 2 B, inset). Air from the room enters the intake blower and passes through a prefilter and a HEPA filter, exiting into a vertical manifold at the top of the rack. The supply air travels down the vertical manifold into horizontal plenums before entering into the cage through an open port in the cage lid (Figure 2B, white arrows). Exhaust air then leaves the cage by passing through a filter in the cage lid and passes into the horizontal plenum (Figure 2 B, black arrows). The exhaust air then travels to the vertical manifold and enters in the exhaust blower at the top of the rack passing through a prefilter and then a HEPA filter before being exhausted to the room. Due to the top mounted blower and a previous finding that sampling at the terminal vertical manifold was not as reliable as sampling at the junction of the vertical and horizontal manifolds, rack 3 was not sampled at the exhaust-blower level.6 Instead, the rack was ordered with optional access ports at the level of each horizontal plenum. Sampling of the horizontal air plenums through access ports located in the vertical exhaust air duct was performed by using flocked swabs (Figure 3 B). Each port was opened on the vertical exhaust-air duct, and a swab was inserted and rubbed along the horizontal air plenum.
Exhaust samples were obtained at the initiation of the study (time point 0), weekly intervals for 4 wk after study initiation, and then monthly until study end (12 wk). Samples were tested by real-time PCR analysis for the presence of MHV, MNV, MPV, P. pneumotropica, Helicobacter spp., fur mites, pinworms, E. muris, and T. muris. In addition, debris that had accumulated for 12 wk in the tubing connecting the blower of rack 1 was collected at the end of the study and stored at room temperature for 12 mo. Aliquots of dust (200 μL each) were tested at 3, 9, and 12 mo by real-time PCR analysis for the presence of MHV, MNV, MPV, P. pneumotropica, Helicobacter spp., fur mites, pinworms, E. muris, and T. muris.
Results
Infected mice.
Fecal samples collected from mouse cages demonstrated that naïve ICR mice exposed to MNV, MPV, and MHV were infected, except for one cage of MPV-exposed mice in rack 1 (Table 1). Cages housing mice infected with MPV remained positive by fecal PCR analysis throughout the study, indicating continual shedding, and MPV-infected mice were seropositive for MPV at 84 d (the end of the study). MHV was transferred to all cohoused mouse cages, but shedding decreased or was absent starting at day 21, and all cages were negative for MHV by fecal PCR analysis by day 84. All MHV-exposed mice were seropositive for MHV at 84 d. In contrast, several MNV cages were negative by fecal PCR analysis at the day 0 time point after being cohoused with inoculated mice for 1 wk. However, by 14 d, all MNV cages were positive for MNV fecal PCR analysis and remained positive through day 28 of the study. At the end of the study (day 84), several MNV cages were negative by fecal PCR analysis, but all mice had seroconverted to MNV.
Table 1.
Rack 1 (5 cages per group)
| Mouse PCR results on daya |
Serology on day 84b | Prefilter test results on day |
||||||||||||
| 0 | 7 | 14 | 21 | 28 | 84 | 0 | 7 | 14 | 21 | 28 | 56 | 84 | ||
| MNV | 3 | 5 | 5 | 5 | 5 | 1 | 15 | - | - | - | - | - | - | - |
| MPV | 4 | 4 | 4 | 4 | 4 | 4 | 12 | - | - | + | - | - | - | - |
| MHV | 5 | 5 | 5 | 5 | 2 | 0 | 15 | - | + | + | + | + | + | + |
| Helicobacter spp. | 5 | 5 | 5 | 5 | 5 | 5 | not done | - | + | + | + | + | + | + |
| P. pneumotropica | 5 | 5 | 5 | 5 | 5 | 5 | not done | - | + | + | + | + | + | + |
| E. muris | 5 | 5 | 5 | 5 | 5 | 5 | not done | - | + | + | + | + | + | + |
| T. muris | 5 | 5 | 5 | 5 | 5 | 5 | not done | - | - | + | + | + | + | - |
| Pinworms | 5 | 5 | 5 | 4 | 4 | 4 | not done | - | + | + | + | + | + | + |
| Fur mites | 5 | 5 | 5 | 5 | 5 | 5 | not done | - | + | + | + | + | + | + |
Number of positive cages among 5 cages tested
Number of positive mice among 15 mice tested
All cages of Sencar mice were positive for Helicobacter spp., P. pneumotropica, and fur mites at every time point tested (Tables 1 through 3). In addition, these cages were positive for pinworms at every time point tested, except for one cage that tested negative on day 21 and continued to be negative throughout the study including day 84 (Table 1). Most cages tested positive for E. muris and T. muris at every time point tested, but single cages tested sporadically negative; this pattern was seen more often with T. muris (Tables 2 and 3). Overall, these results demonstrate that mice were infected with the agents of interest and continued to shed pathogens for varying periods of time throughout the 3-mo study.
Table 2.
Rack 3 (5 cages per group)
| Mouse PCR results on daya |
Serology on day 84b | Prefilter test results on day |
||||||||||||
| 0 | 7 | 14 | 21 | 28 | 84 | 0 | 7 | 14 | 21 | 28 | 56 | 84 | ||
| MNV | 5 | 5 | 5 | 5 | 5 | 2 | 15 | - | - | - | - | - | - | - |
| MPV | 5 | 5 | 5 | 5 | 5 | 5 | 15 | - | - | - | - | - | - | - |
| MHV | 5 | 5 | 5 | 4 | 3 | 0 | 15 | - | - | - | - | - | - | - |
| Helicobacter spp. | 5 | 5 | 5 | 5 | 5 | 5 | not done | - | - | - | - | - | - | - |
| P. pneumotropica | 5 | 5 | 5 | 5 | 5 | 5 | not done | - | - | - | - | - | - | - |
| E. muris | 4 | 5 | 5 | 5 | 5 | 5 | not done | - | - | - | - | - | - | - |
| T. muris | 4 | 5 | 5 | 4 | 5 | 5 | not done | - | - | - | - | - | - | - |
| Pinworms | 5 | 5 | 5 | 5 | 5 | 5 | not done | - | - | - | - | - | - | - |
| Fur mites | 5 | 5 | 5 | 5 | 5 | 5 | not done | - | - | - | - | - | - | - |
Number of positive cages among 5 cages tested
Number of positive mice among 15 mice tested
Table 3.
Rack 2 (1 cage per agent)
| Mouse PCR results on daya |
Serology on day 84b | Prefilter test results on day |
||||||||||||||
| 0 | 7 | 14 | 21 | 28 | 84 | 0 | 7 | 14 | 21 | 28 | 56 | 84 | ||||
| MNV | 0 | 0 | 1 | 1 | 1 | 0 | 3 | - | - | - | - | - | - | - | ||
| MPV | 1 | 1 | 1 | 1 | 1 | 1 | 3 | - | - | - | - | - | + | - | ||
| MHV | 0 | 1 | 1 | 1 | 1 | 0 | 3 | - | + | + | + | + | + | - | ||
| Helicobacter spp. | 1 | 1 | 1 | 1 | 1 | 1 | not done | - | - | + | + | + | + | + | ||
| P. pneumotropica | 1 | 1 | 1 | 1 | 1 | 1 | not done | - | - | - | - | - | + | + | ||
| E. muris | 1 | 1 | 1 | 1 | 1 | 1 | not done | - | - | - | - | + | + | + | ||
| T. muris | 1 | 0 | 0 | 1 | 1 | 1 | not done | - | - | - | - | - | + | + | ||
| Pinworms | 1 | 1 | 1 | 1 | 1 | 1 | not done | - | + | + | + | + | + | + | ||
| Fur mites | 1 | 1 | 1 | 1 | 1 | 1 | not done | - | + | + | + | + | + | + | ||
Number of positive cages (1 tested per time point)
Number of positive mice (3 tested)
Rack monitoring.
At day 0, samples from all 3 racks tested negative by PCR analysis for all agents tested, demonstrating the racks were not contaminated prior to the start of testing (Tables 1 through 3). Despite housing mice shedding a number of viral, bacterial, and parasitic agents, samples from rack 3 tested negative for all tests at every time point (Table 2). In contrast, samples from racks 1 and 2 tested positive for multiple agents (Tables 1 and 3). Rack 1, which contained 5 cages per agent, was positive from day 7 through day 84 for MHV, Helicobacter spp., P. pneumotropica, E. muris, pinworms, and fur mites (Table 1). This rack was also positive for T. muris from day 14 through day 56. In contrast, MNV was not detected by rack testing, and MPV was positive at only one time point (day 14). Rack 2, which contained one cage per agent, had a generally lower rate of positive rack samples (Table 3). Samples for rack 2 were positive for pinworms and fur mites on days 7 through 84 and for MHV on days 7 through 56. Samples tested for P. pneumotropica, E. muris, and T. muris were found positive between days 28 and 84 of the study. MNV was not detected in rack 2 samples, and MPV was detected at only a single time point (day 56). These results are similar to the findings for rack 1 (Table 3).
Pathogen stability in exhaust debris.
At 3 mo after the end of the study, exhaust debris collected from the tubing connecting the rack to the blower was negative for MNV and MPV, and positive for MHV, Helicobacter spp., P. pneumotropica, E. muris, pinworms, and fur mites (Table 4). The sample remained positive for Helicobacter spp., P. pneumotropica, E. muris, pinworms, and fur mites for 1 y. These results confirm the stability of the nucleic acids and the ability to detect the presence of an agent by PCR long after the source mice are removed from the ventilated rack.
Table 4.
Stability of agent in exhaust debris at room temperature according to PCR analysis
| 3 mo | 9 mo | 12 mo | |
| MNV | - | - | - |
| MPV | - | - | - |
| MHV | + | + | - |
| Helicobacter spp. | + | + | + |
| P. pneumotropica | + | + | + |
| E. muris | + | + | + |
| T. muris | + | + | + |
| Pinworms | + | + | + |
| Fur mites | + | + | + |
Discussion
Sentinel mice have been the cornerstone of rodent health monitoring for the past 5 decades, but increasing biosecurity using microisolation and ventilated caging has resulted in sequestration of disease to the cage level such that each cage has its own microbial status. The deficits in pathogen detection using dirty-bedding sentinel monitoring for agents such as fur mites, P. pneumotropica, and Helicobacter spp. have been recognized, and testing of environmental samples has improved disease detection. The detection of fur mite DNA in the bedding debris in the exhaust air manifolds of a ventilated rack offered the possibility of a cost-effective method for monitoring large populations of rodents for infectious pathogens.6 When we explored the use of exhaust debris to monitor for murine pathogens, we learned that the design of the ventilated rack must be considered if environmental rack monitoring is to be incorporated into an animal health monitoring program. In rack 3, bedding debris in the exhaust air is trapped by the filter cage lids such that debris cannot enter the exhaust manifold, and our study demonstrated that no pathogens were detected on samples collected from these exhaust manifolds. In fact, debris was not found on the inner surfaces of the manifolds, and the prefilter at the exhaust level remained free of visible debris.
It should be noted that the exhaust port in the horizontal manifold above each cage in both rack designs tested is an open port. If cages are not present at each location in the rack, room air and debris can be drawn through these openings into the exhaust plenum. This feature may be an important consideration if partially filled ventilated racks in either of the tested designs are placed into an animal room with higher burdens of aerosolized bedding debris. Increased bedding debris aerosolization can occur with cages that do not have filtered tops and when cages are changed on open benches and not in a vertical laminar flow or biosafety cabinet. Bedding debris that may be aspirated through these exhaust ports into the exhaust manifold may not reflect the health of animals housed within cages on the rack but rather may represent other animals housed within the room. For this reason, when evaluating results of ventilated rack studies, it is critical to know how all animals are housed and handled within the room in which the study is being performed.
Racks 1 and 2 accumulated debris at all points along the exhaust airflow pathway. The rack design has the particulate prefilter at the blower, and samples of filter material could be easily removed for sample collection. In addition, bedding debris may have concentrated on the filter and thus may have improved pathogen detection in the debris by PCR analysis. For example, fur mites were readily detected in filter samples from rack 2, which housed only one cage of fur-mite–infested mice. This finding is unlike a fur mite study in which sampling at a single point in the terminal region of the vertical exhaust plenum provided false-negative PCR results when mite-infested mice were housed on the rack.6
Overall, results of testing exhaust debris by PCR analysis corroborated the MHV, bacterial, and parasite infection status of mice housed on the rack. These results are consistent with the findings of others who detected fur mites,6 Helicobacter spp., MHV, and pinworms3 in exhaust debris in racks with a similar open airflow design. Our study confirmed that enteric protozoa and P. pneumotropica can also be detected by PCR analysis of exhaust debris. MPV and MNV were not reliably detected in exhaust debris, even though mice were documented to shed virus throughout the study. MNV was not detected at any time point on either rack 1 or 2, and MPV was only sporadically detected. Sporadic detection of MPV in exhaust debris from cages with mice confirmed to shed virus was found in a previous study as well.3 Further study is needed to determine whether a higher prevalence of infected mice or longer duration of infection would improve the detection of these viruses in bedding dust.
Furthermore, our study revealed that when low numbers of positive animals are housed on a rack, the detection of pathogens in exhaust debris may occur weeks after the onset of pathogen shedding, and suggesting it takes longer to accumulate enough contaminated debris of some pathogens to obtain positive results. The detection of a limited number of infected mice, as can occur during a disease outbreak, might be delayed until the prevalence of infection increases or until the continued accumulation of contaminated debris results in positive PCR results. However, by the end of study, the results of pathogen screening in debris from rack 2 (1 cage per agent) was the same as was found in debris from rack 1 (5 cages per agent).
Debris collected from the tubing connecting the rack to the blower of rack 1 yielded basically the same results as those from the prefilter samples. This material was collected 3 mo after the end of the study, and the rack had remained empty with the blowers running since the end of the study. The material remained PCR-positive for 1 y, demonstrating the stability of the contamination once it is introduced into a ventilated rack and emphasizes the need to thoroughly decontaminate racks and verify that they test PCR-negative prior to housing animals on the rack when using exhaust-debris monitoring.
Overall, testing of exhaust debris in racks with unfiltered exhaust air flow reliably detected the agents that are difficult to detect by dirty-bedding sentinel monitoring (fur mites, Helicobacter spp., and P. pneumotropica) as well as several agents that are reliably detected using dirty-bedding sentinels (MHV, E. muris, T. muris). If the ventilated rack system includes an inline filter between the cage and the exhaust system, rack testing was not a reliable method for detection of these agents. Testing exhaust samples was not a reliable method for the detection of MPV and MNV regardless of the rack type used. This study demonstrates that, depending on the ventilated rack design, rack testing can be a reliable method for detection of some—but not all—agents found in contemporary mouse colonies. In addition, the type of ventilated rack and the specific design of the rack is an important consideration when adding rack monitoring to an animal health monitoring program.
Acknowledgments
We thank Alaina Armentrout, Corri Phillips, Shannon Primm, Greg Purdy, Amberly Schulz, and Laurie Wisdom for providing technical support during this study. We also thank Tecniplast for providing racks 1 and 2 with caging for use in this study. This study was funded by and the authors are employed by IDEXX BioResearch.
References
- 1.Artwohl JE, Cera LM, Wright MF, Medina LV, Kim LJ. 1994. The efficacy of a dirty-bedding sentinel system for detecting Sendai virus infection in mice: a comparison of clinical signs and seroconversion. Lab Anim Sci 44:73–75. [PubMed] [Google Scholar]
- 2.Brielmeier M, Mahabir E, Needham JR, Lengger C, Wilhelm P, Schmidt J. 2006. Microbiologic monitoring of laboratory mice and biocontainment in individually ventilated cages: a field study. Lab Anim 40:247–260. [DOI] [PubMed] [Google Scholar]
- 3.Compton SR, Homberger FR, Paturzo FX, Clark JM. 2004. Efficacy of 3 microbiologic monitoring methods in a ventilated cage rack. Comp Med 54:382–392. [PubMed] [Google Scholar]
- 4.Dillehay DL, Lehner ND, Huerkamp MJ. 1990. The effectiveness of a microisolator cage system and sentinel mice for controlling and detecting MHV and Sendai virus infections. Lab Anim Sci 40:367–370. [PubMed] [Google Scholar]
- 5.Grove KA, Smith PC, Booth CJ, Compton SR. 2012. Age- associated variability in susceptibility of Swiss Webster mice to MPV and other excluded murine pathogens. J Am Assoc Lab Anim Sci 51:789–796. [PMC free article] [PubMed] [Google Scholar]
- 6.Jensen ES, Allen KP, Henderson KS, Szabo A, Thulin JD. 2013. PCR testing of a ventilated caging system to detect murine fur mites. J Am Assoc Lab Anim Sci 52:28–33. [PMC free article] [PubMed] [Google Scholar]
- 7.Leblanc M, Berry K, Graciano S, Becker B, Reuter JD. 2014. False-positive results after environmental pinworm PCR testing due to rhabditid nematodes in corncob bedding. J Am Assoc Lab Anim Sci 53:717–724. [PMC free article] [PubMed] [Google Scholar]
- 8.Lindstrom KE, Carbone LG, Kellar DE, Mayorga MS, Wilkerson JD. 2011. Soiled-bedding sentinels for the detection of fur mites in mice. J Am Assoc Lab Anim Sci 50:54–60. [PMC free article] [PubMed] [Google Scholar]
- 9.Livingston RS, Riley LK, Besch-Williford CL, Hook RR, Jr, Franklin CL. 1998. Transmission of Helicobacter hepaticus infection to sentinel mice by contaminated bedding. Lab Anim Sci 48:291–293. [PubMed] [Google Scholar]
- 10.Manuel CA, Pugazhenthi U, Leszczynski JK. 2016. Surveillance of a ventilated rack system for Corynebacterium bovis by sampling exhaust-air manifolds. J Am Assoc Lab Anim Sci 55:58–65. [PMC free article] [PubMed] [Google Scholar]
- 11.Pritchett-Corning KR, Cosentino J, Clifford CB. 2009. Contemporary prevalence of infectious agents in laboratory mice and rats. Lab Anim 43:165–173. [DOI] [PubMed] [Google Scholar]
- 12.Scharmann W, Heller A. 2001. Survival and transmissibility of Pasteurella pneumotropica. Lab Anim 35:163–166. [DOI] [PubMed] [Google Scholar]
- 13.Smith PC, Nucifora M, Reuter JD, Compton SR. 2007. Reliability of soiled bedding transfer for detection of mouse parvovirus and mouse hepatitis virus. Comp Med 57:90–96. [PubMed] [Google Scholar]
- 14.Thigpen JE, Lebetkin EH, Dawes ML, Amyx HL, Caviness GF, Sawyer BA, Blackmore DE. 1989. The use of dirty bedding for detection of murine pathogens in sentinel mice. Lab Anim Sci 39:324–327. [PubMed] [Google Scholar]
- 15.Whary MT, Cline JH, King AE, Hewes KM, Chojnacky D, Salvarrey A, Fox JG. 2000. Monitoring sentinel mice for Helicobacter hepaticus, H rodentium, and H bilis infection by use of polymerase chain reaction analysis and serologic testing. Comp Med 50:436–443. [PubMed] [Google Scholar]