Using Reduced Personal Protective Equipment in an Endemically Infected Mouse Colony

Abstract

Personal protective equipment (PPE) frequently is used to reduce the risk of spreading adventitial diseases in rodent colonies. The PPE worn often reflects the historic practices of the research institution rather than published performance data. Standard PPE for a rodent facility typically consists of a disposable hair bonnet, gown, face mask, shoe covers, and gloves, which are donned on facility entry and removed on exiting. This study evaluated the effect of a reduced PPE protocol on disease spread within an endemically infected mouse colony. In the reduced protocol, only the parts of the wearer that came in direct contact with the mice or their environment were covered with PPE. To test the reduced PPE protocol, proven naïve mice were housed in a facility endemically infected with murine norovirus and mouse hepatitis virus for 12 wk. During that time, routine husbandry operations were conducted by using either the standard or reduced PPE protocols. All study mice remained free of virus antibody when reduced PPE was implemented. These results indicate that reduced PPE is adequate for disease containment when correct techniques for handling microisolation caging are used. Reducing the amount of PPE used in an animal facility affords considerable cost savings yet limits the risk of disease spread.

The introduction of harmful microbial agents is potentially devastating to research and can be time-consuming and expensive to eradicate. Animal research facilities typically implement a variety of practices designed to prevent the introduction and spread of adventitial agents. Personal protective equipment (PPE) is a mainstay of colony biosecurity and personnel protection. However, there is no accepted standard for the use of PPE in rodent rooms and therefore great variability in PPE use exists among research institutions.^1,2,14

PPE was a topic of discussion at the 2010 American College of Laboratory Animal Medicine forum, where the results of a questionnaire on PPE usage at academic and commercial institutions highlighted this variation.¹ Of the respondents, 49% required head covering, 70% required shoe covering or dedicated shoes, 23% required disposable sleeves, and 97% required a set of gloves to be worn when manipulating immunocompetent rodents. At our institution, a disposable hair bonnet, gown, shoe covers, and gloves are required to enter the SPF rodent colonies, with the wearing of face masks being optional. According to responses to the forum survey, this PPE requirement is comparable to those of other institutions, but it is based on historic practices rather than performance data.

Together, PPE and barrier housing are considered to be the primary facility and management procedures designed to isolate animals from infectious agents¹³ and to maintain SPF rodents. Barriers can be established at the cage, room, suite, or facility level.²⁰ Establishing the barrier at the cage level requires the following:²⁰ isolator caging systems,^19,24 sanitized cages and water bottles, uncontaminated food and bedding, and the completion of all manipulations in change stations providing laminar-flow HEPA-filtered air.⁴ This process is termed microisolation technique and is implemented at our institution, where the barrier is maintained at the cage level.

Despite these practices, breaks in the microbiologic status still occur. All factors being equal, prevalent pathogens pose a higher risk than do rare ones, and 2 of the 3 most prevalent murine viral pathogens in research facilities are murine norovirus (MNV) and mouse hepatitis virus (MHV).^6,26 MNV is a nonenveloped single-stranded RNA virus that was first described in a colony of RAG2/STAT1^−/− mice,¹⁸ and MHV is an enveloped single-stranded RNA virus belonging to the family Coronaviridae.⁷ Recent surveys found MNV to be the most prevalent viral pathogen in North America and Europe, with MHV being the third most common.^6,26 Rodent colonies at our institution are endemically infected with MNV, and one facility is endemically infected with MHV also.

In addition to being prevalent, these viruses affect animal health and research. MNV initially was shown to cause lethality in STAT1^−/− mice¹⁸ and chronic infection of RAG2^−/− mice¹⁸ and subsequently has been shown to cause clinical and subclinical disease in other immunodeficient and immunocompetent mice.^25,28 In addition, MNV can modulate the disease course of other murine viruses when coinfected.¹⁰ Enteric MHV, now having largely superseded polytropic MHV strains in contemporary mouse colonies,¹⁶ causes subclinical infections in immunocompetent mice,⁷ but leads to disseminated multiorgan infection in immunodeficient mice.⁷ Transmission of both viruses is via the fecal–oral route. MHV is highly transmissible via direct contact, soiled bedding, and air.^9,27 However MHV is labile in the environment.^9,27 MNV is readily transmissible via contact and soiled bedding²¹ but is highly stable in the environment.^4,12

The aim of the current study was to evaluate whether the type of PPE worn affected viral spread to naïve animals within a dynamic mouse colony. To our knowledge, this study is the first to evaluate disease spread in a facility in the context of different levels of PPE. We hypothesized that the level of PPE worn when microisolation technique was used would not affect the exposure of naïve mice to MNV and MHV in a colony that was endemically infected with these viruses. The rationale was that only the parts of the body in direct contact with the mouse and its environment need to be covered with PPE and sanitized. In this way, focused PPE use could prevent disease spread.

Materials and Methods

Animals.

Female (n = 106) immunocompetent Crl:CD1(ICR) mice (age, 4 to 6 wk; Charles River, Wilmington, MA) specified by the supplier to be free of murine viruses, pathogenic bacteria, and endo- and ectoparisites were used for these studies. All work was conducted in accordance with the Guide for the Care and Use of Laboratory Animals¹⁷ in an institution accredited by AAALAC. All procedures outlined in the study were approved by the Columbia University IACUC.

To ensure the integrity of the shipping and unpacking process, 2 mice from each shipping crate were transferred to a flexible-film isolator (ParkBio, Groveland, MA) and kept there for testing. For each of the groups, 4 mice were separated and tested to ensure that incoming mice were virus-naïve. After 6 d, a fresh fecal pellet collected from each mouse tested for MHV and MNV by RT-PCR, and after 14 d blood collected from the submandibular vein was tested for MNV and MHV antibodies (Charles River Diagnostics Labs, Wilmington, MA).

Mice were housed in individually ventilated cages (JAG75, Allentown Caging System, Allentown, NJ) or in static (7115, Allentown Caging System) microisolation cages. Cage components were sanitized in a mechanical washer (model 7 tunnel washer, Buxton, Lindenhurst, NY) by using a detergent and reaching a temperature of 180 °F before being autoclaved (model 8502, Buxton). The ventilated racks distribute room air through a HEPA filter to the cages, and exhausted air was delivered directly into the building's HVAC system. HEPA filters were certified annually and replaced as needed. All individually ventilated cages were kept at positive pressure relative to the room. Mice were housed on autoclaved 1/8-in. corncob bedding (Cob Products, The Andersons, Maumee, OH) and were provided with γ-irradiated commercial diet (Picolab 5053, PMI, St Louis, MO) and acidified water (pH 2.5 to 2.8) supplied in water bottles ad libitum. Mice were singly housed and provided with enrichment in the form of autoclaved nesting pads (Fisher Feed, Somerville, NJ). The macroenvironment was controlled at 68 to 79 °F, at 30% to 60% relative humidity, and on a 12:12-h light:dark cycle.

Mouse facility.

The multi-room facility has a capacity of 7500 cages and houses an approximately 50:50 mix of ventilated and static double-sided racks. The facility contains 19 animal housing rooms and 4 procedure rooms, all of which were endemically infected with MNV and 10 of which were endemically infected with MHV, according to sentinel results from the preceding 3 y. There are 2 anterooms for entry into the facility that allow access to all rooms. During the last 3 y, no new rooms became infected with MHV, and rooms that were infected routinely continued to test positive on quarterly sentinel results. Traffic flow was designed so that MHV-positive rooms were entered after MHV-negative rooms, thereby minimizing the risk of interroom transmission. Separate procedure space was available for mice from MHV positive and negative rooms. Rooms that were positive for MHV were kept at a negative pressure relative to the corridor, whereas MHV-negative rooms were kept under positive pressure relative to the corridor. For this study, 5 rooms positive for MHV and MNV were selected, all of which belonged to the same investigator. The room capacity ranged from 480 to 840 cages, with each rack approximately 80% occupied during the study. Of these 5 rooms, 3 used ventilated caging, and the remaining 2 used static caging systems.

PPE.

PPE is required to be worn by everyone entering the mouse facility. The standard PPE protocol (sPPE) at our institution comprised a hair bonnet, disposable gown, shoe covers, and gloves (all from Physician Sales and Services, Jacksonville, FL). These items comprised the control condition for comparison of MNV and MHV spread with the reduced PPE. Reduced PPE (rPPE) comprised water-resistant sleeves (Physician Sales and Services) and gloves. For animal care staff, PPE is worn over dedicated scrubs and shoes. PPE was donned in an anteroom prior to entering the facility and then removed in the same room on the way out. PPE was not changed between rooms. During the time that the first group of study mice was housed, all routine manipulations conducted in the housing room to both the study and colony mice took place while caretakers wore sPPE. During the second part of the study, these same manipulations were conducted by using rPPE.

Participants.

Study participants were animal care technicians assigned to the rooms by the facility supervisor. No change in personnel was made prior to initiation of the study or during the study except for vacation and illness. A total of 6 animal care technicians took part in the study, and each was blinded to the study, unaware of the presence of the study mice on the rack. To reduce bias by the caretakers due to any perceived difference in PPE, the sPPE study was evaluated before the rPPE study began.

Microisolator technique.

All manipulations that involved opening the microisolation cage lid were conducted in a laminar-flow cage-change station (Lab Products, Seaford, DE). For this study, manipulations were restricted to routine husbandry only, with cage bottoms and water bottles changed weekly, and wire-bar lids and microisolation cage lids changed monthly. This requirement was the same for both the static and the ventilated cages. The monthly frequency of cage-lid changing was based on performance data and approved by the Columbia University IACUC as an exemption to the Guide.¹⁷ The laminar-flow cage-change station was sanitized at the beginning and end of the work period. Mice were transferred by using forceps that were soaked with disinfectant between animals. Gloves and sleeves, if worn, were sanitized between cages. All sanitization was done by using the broad-spectrum disinfectant Virkon S (DuPont, Physician Sales and Services).

Serologic testing.

In each of the 5 rooms enrolled in the study, 1 female CD1 experimental mouse was placed in a cage and housed on each side of each rack. The position on the rack was assigned randomly to each experimental cage. The mice were housed for 12 wk, undergoing routine husbandry throughout that time. At the end of the 12 wk, blood collected from submandibular vein by using manual restraint was submitted to a commercial laboratory (Charles River Diagnostic Labs) for MNV and MHV serologic testing. Once the results were received, the mice were euthanized. The study protocol was the same for both the sPPE and rPPE groups, with a new set of mice being used for the rPPE study.

Colony health monitoring.

Dirty-bedding sentinels were used as part of the health monitoring program to detect the presence of excluded murine pathogens in the rooms enrolled in this study. Outbred Crl:CD1(ICR) mice were used as dirty-bedding sentinels. During the weekly cage change, dirty bedding was collected from every cage in the rack by using a disposable scoop that collected, on average, 7 g of dirty bedding. The sentinels were housed on this dirty bedding to allow for exposure, and seroconversion to pathogens present in the colony cages. The dirty-bedding sentinels were housed for 12 wk; at the end of this period, blood samples were collected and analyzed to determine seroconversion. The results from the dirty-bedding sentinels were used to determine whether MNV and MHV were present during the experimental period, with seroconversion of the sentinel indicating exposure to the virus.

Statistics.

All data manipulations and statistics were done by using Excel 2007 (Microsoft, Redmond, WA) with the StatPlus plugin. Results were determined to be statistically significant when the P value was less than 0.05. Data were analyzed by using paired t tests and one-way ANOVA.

Results

Validation of SPF status of test animals.

To ensure that mice were free of MNV and MHV upon arrival from the vendor, we tested 4 mice by RT-PCR at 6 and 14 d after they were received at our facility. Results at both time points for both groups were negative, ensuring that any MNV and MHV infections detected could be attributed to conditions at our facility and not to those at the vendor facility or during the transport process.

Effect of PPE levels on seroconversion.

To determine the effect of PPE on viral transmission, MNV and MHV seroconversion of study animals was compared between the sPPE and rPPE groups. After 12 wk, none of the 49 mice in either the sPPE or rPPE groups had seroconverted to MNV or MHV Within the rPPE group, no difference was seen between static (n = 20) and ventilated (n = 29) caging systems.

Seroconversion rates of dirty bedding sentinels.

To determine the presence of virus during both the study periods, data from dirty-bedding sentinels were collected and analyzed. The seroconversion rate was calculated by dividing the number of sentinels that seroconverted by the total number of sentinels placed in that room. Dirty-bedding sentinels from all 5 rooms across both study periods showed high MNV seroconversion, with the average seroconversion rate of all rooms being 0.89. By contrast for MHV, only 1 of 5 rooms showed high seroconversion during both study periods (average seroconversion rate, 0.79); 2 rooms showed low seroconversion during the rPPE period (average seroconversion rate, 0.125); and 2 rooms showed no sentinel conversion during either study period. A paired t test showed no significant difference between the seroconversion of dirty-bedding sentinels during the sPPE and rPPE periods (MNV, P = 0.82; MHV, P = 0.77; Figure 1).

Figure 1.

Seroconversion rates (mean ± SEM) of dirty-bedding sentinels during the study period, according to virus and level of PPE. There was no significant difference between the sPPE and rPPE groups for either MNV or MHV.

Historic sentinel data for the 5 rooms in the study was collated from the preceding 2 y and analyzed (Figure 2). One-way ANOVA revealed no difference in the sentinel seroconversion rates over this period for either MNV or MHV (MNV: F_{8, 36} =1.40, P = 0.23; MHV: F_{8, 36} = 0.29, P = 0.97), suggesting that both viruses were indeed endemic in the infected rooms. The seroconversion rate for MNV was consistently high in 4 of 5 rooms; however one room had less seroconversion than did the other rooms. The MHV seroconversion rate was low but consistent, although one of the rooms had a high rate of seroconversion compared with that of MNV. Historic sentinel seroconversion rates then were compared with seroconversion rates during the 2 test periods (Figure 2). According to one-way ANOVA, no significant difference was detected between the 2 study groups and the historic sentinel data for either MNV or MHV (MNV: F_{10, 44} = 0.93, P = 0.51; MHV: F_{10, 44} = 0.23, P = 0.99).

Figure 2.

Seroconversion rates (mean ± SEM) for dirty-bedding sentinels to MNV (red) and MHV (blue) over preceding years. 2013 quarter 1 represents the time when the sPPE study animals were housed, and 2013 quarter 2 represents rPPE. No significant difference was seen in average MNV or MHV seroconversion rates for the 2-y period leading up to the study or between the study period and the preceding 2 y.

Discussion

In this study, we demonstrated that a reduced level of PPE combined with microisolator technique can be used to maintain naïve mice virus-free in an endemically infected colony independent of caging system used. During the 12-wk period, no study animals seroconverted with either sPPE or rPPE, consistent with our hypothesis, even though dirty-bedding sentinel seroconversion during the same time period demonstrated the presence of virus. We therefore conclude that wearing PPE only on parts of the body that enter the animal's microenvironment (that is, rPPE) is just as effective in preventing contamination as is wearing is our facility's sPPE. This study suggests that focusing on microisolator technique and maintaining the barrier at the cage level enables the use of decreased PPE regardless of the housing type.

To detect the presence of adventitious diseases in a rodent colony, an appropriate health-monitoring program is essential. Dirty-bedding sentinels often play a key role, given that this method provides an economic way to screen a large, dynamic population.^9,23,26 The use of dirty-bedding sentinels was central to the current study, demonstrating the presence of virus in the room housing study animals as well as the endemic status of MNV and MHV in the colony. However, when isolator caging systems are used, dirty-bedding sentinels cannot be used to determine the true prevalence of virus in the room,⁵ although it is the commonly used metric in rodent colony management.^15,21,23 In the current study, the seroconversion rate was used as a proxy for prevalence and demonstrates the consistent presence of virus in the room during both study periods.

However limits to the efficacy of dirty-bedding sentinels include the volume of bedding transferred, fecal content of bedding, age and strain of sentinel mice, and transmission route of the tested pathogen.^11,21,27 To control for age, mice were replaced after the sPPE arm of the study so that the rPPE arm used purchased mice of identical age, even though one study reports no significant difference in detection by outbred dirty-bedding sentinels ranging from 4 to 44 wk.¹¹ For MNV, outbred immunocompetent mice reliably seroconvert when exposed to an infective dose^21,25 and shed virus persistently.^21,25 MNV has been demonstrated to remain infective in fecal material at ambient temperature.³ Thus dirty-bedding sentinels are a reliable method for detecting MNV when the virus is prevalent.^21,26 One room in our study had a decreased rate of dirty-bedding seroconversion for MNV during the study. This result may reflect a true reduction in prevalence of MNV in that room but more likely reflects one of the limitations of dirty-bedding transfer described earlier. For MHV, many studies have looked at the transmission and efficacy of detection via different sentinel methods.^9,11,22,27 Unlike MNV, MHV is unstable in the fecal environment;²⁷ therefore, virus transmission and dirty-sentinel detection is likely to occur only during active infection.²⁷ This window can vary greatly, given that different strains shed virus for different durations after infection—for example, BALB/c mice shed virus for as long as 2 wk, C57BL/6 mice can shed for 4 wk,⁸ and other strains can shed virus even longer.¹⁵ In the 5 rooms in our study, MHV seroconversion varied from quarter to quarter, although this variation was nonsignificant and may reflect periodic failure of dirty-bedding transfer or detection at lower prevalence.²⁷ Therefore, for both MNV and MHV, the seroconversion of dirty-bedding sentinels can be used to reliably demonstrate viral presence in a colony.

A reduction in PPE usage will yield considerable cost savings for institutions by reducing direct costs which may affect per-diem charges. In the last year, our institution spent approximately $250,000 on PPE across all rodent facilities. Cost estimates for that same time period using rPPE are $100,000, thus saving approximately $150,000 per year. These figures represent not only direct cost savings but also savings through reduced employee time spent on PPE ordering, storage, donning, removing, and disposing. These savings are in line with those from a study conducted at another large institution, which reported a 17% reduction in cost when shoe covers alone were not worn.¹⁴ At neither our facility nor in the cited study¹⁵ was PPE required to be changed between rodent rooms. However, a high percentage of survey respondents¹ reported that disposable PPE items were changed between rooms (for example, 43% of respondents reported changing gowns), therefore cost savings could become even larger with a rPPE strategy. In addition, during discussions at the end of the current study, the caretakers involved expressed a strong preference for wearing the rPPE. Furthermore, a reduction in waste disposal with this approach is evident as well, making the use of rPPE an environmentally sustainable approach.

In summary, a decreased PPE protocol that focuses on covering only those parts of the body that enter the animals’ microenvironment can be used to maintain virus-naïve mice in an endemically infected mouse colony. This study refines PPE use, yet maintains the SPF status of the cage when PPE use is combined with correct microisolator technique. This refinement has implications for PPE practices used in rodent colonies and likely will achieve considerable cost savings without adverse effects on microbiologic status.