Assessing the Biosecurity Risk of Footwear as a Fomite for Transmission of Adventitious Infectious Agents to Mice

Introduction

A 2020 survey revealed that more than 2 million mice and rats were reported outside of New York City (NYC) buildings over a 90-d period.⁴ While the precise number of wild rodents in NYC is unknown, statistical models estimate that a population of approximately 2 million rats cohabitate with over 8 million city residents.⁴ Wild mice in NYC have been found to carry many infectious agents in their feces including Shigella spp.; enteroinvasive, enteropathogenic, and Shiga toxin-producing Escherichia coli; Clostridioides difficile; Clostridium perfringens; Salmonella enterica; Leptospira spp.; murine adenovirus 2; murine parvovirus 2; murine polyomavirus; lactate dehydrogenase–elevating virus; murine hepatitis virus; and murine rotavirus, among other agents.^56,57 Our feral mouse monitoring program has identified many of these agents in mice captured in proximity to our research buildings with vivaria. In fact, we suspected a wild rodent incursion to have been the likely cause of an outbreak of a novel virus, murine astrovirus 2, in one of our vivaria.⁴⁶ A recent New York Times article discussed how the number of wild rodents and the incidence of zoonotic diseases, for which these animals serve as reservoirs, increased in NYC during the COVID-19 pandemic.⁶ This urban wild rodent population is not only a threat to human health, but also poses a significant biosecurity threat to research rodent colonies, as many of the infectious agents they carry are well documented to confound experimental outcomes in research rodents.^5,7,22

To mitigate this risk, animal care programs strive to exclude specific agents from their rodent colonies by implementing stringent biosecurity practices and colony health monitoring programs.^{5,7,9,13,19,33} Given the seemingly ubiquitous presence of wild rodents on NYC streets and in subway stations, it is highly probable that research and animal care staff unwittingly accumulate debris shed by wild rodents as well as their excrement on the soles of their shoes during their travels to and from work. Consequently, street shoes are a plausible source of fomite-mediated transmission of excluded infectious agents to laboratory rodents.

At our institutions, shoe cleaners are used to mechanically remove particulate debris from research and animal care staffs’ shoes prior to vivarium entry. Shoe cleaners were adopted for vivarium use from the electronics manufacturing industry, where cleanrooms are used to ensure that particulates are excluded from specific manufacturing areas, as they can interfere with the function of electronic components. Shoe cleaners consist of motor-driven brushes that remove large and medium-sized particles from the top, bottom, and sides of shoes to reduce the potential that particulates carrying infectious agents are introduced into the vivarium. As these particles are removed, they are collected through a vacuum system into a collection unit. Although uncommon in research vivaria, shoe cleaners are used in lieu of shoe covers in our vivaria because donning shoe covers may lead to the contamination of personnel’s hands as they enter the vivarium, increasing the chance of the inadvertent introduction of adventitious agents into the vivarium.^{2,17,27,28,34,54} Despite significant concern among human healthcare experts that shoes can serve as fomites, there is no conclusive evidence as to whether footwear poses a substantial transmission risk to animals or humans.^32,44,58 We limit shoe cover use to situations when gross contamination of footwear is likely to occur (for example, large animal holding rooms and necropsy) or within areas where hazardous agents are used to reduce the likelihood of hazardous agent contamination outside of the containment area. For similar reasons, human hospitals no longer recommend the use of shoe covers as a form of personal protective equipment to reduce the introduction of infectious agents and only use them to protect from blood and bodily fluid contamination.^28,34,36,54

At our institutions, shoe cleaners are used upon entry to 5 vivaria by hundreds of staff daily. Their use presented a unique opportunity to evaluate whether contaminated shoes pose a substantive biosecurity risk to rodent colonies. In this study we aimed to survey shoe debris collected by shoe cleaners or directly from shoe soles for the presence of adventitious agents using multiplex PCR and confirm whether the material collected can transmit infectious agents to mice. Results from this study provide valuable insights as to whether shoes pose a substantive biosecurity risk associated with staff entering a vivarium.

Materials and Methods

Experimental design.

Vivarium shoe cleaner study.

A study was conducted to assess the potential transmission of murine adventitious agents to mice exposed to debris collected from shoe cleaners. Shoe cleaner debris (SCD; Figure 1) was collected from 5 NYC vivaria (designated A to E). We surveyed for 84 infectious agents present in each SCD sample by analyzing 25 mL of unaltered SCD, as well as contact media (CM; PathogenBinder™, Charles River Laboratories, Wilmington, MA) exposed to the same 25 mL of SCD, by multiplex PCR (Table 1).

View Figure

• Download as Powerpoint
• Download Figure

Table 1.List of viral, bacterial, and protozoal agents tested by serology and/or PCR

Pathogen	Shoe cleaner Study	Street shoes direct testing study
	SCD or CM	CM exposed to shoe soles	Mice exposed to CM exposed to shoe soles
Boone cardiovirus-1	✓	✓	NP
Ectromelia virus	✓	✓	✓
Hantaan virus	✓	✓	✓
Lactate dehydrogenase–elevating virus (LDV)	✓	✓	✓
Lymphocytic choriomeningitis virus (LCMV)	✓	✓	✓
Minute virus of mice (MVM)	✓	NP	NP
Mouse hepatitis virus (MHV)	✓	✓	✓
Murine norovirus (MNV)a	✓	✓	✓
Generic MVM/mouse parvovirus (MPV)	✓	✓	✓
MPV-1	✓	NP	NP
MPV-2	✓	NP	NP
MPV-3	✓	NP	NP
MPV-4	✓	NP	NP
Mouse thymic virus	✓	✓	✓
Epizootic diarrhea of infant mice (EDIM)	✓	✓	✓
Murine adenovirus 1 and 2 (MAV 1 and 2)	✓	✓	✓
Murine alphacoronavirus	✓	✓	NP
Murine astrovirus 1 (MuAstV1)a	✓	✓	✓
Murine astrovirus 2 (MuAstV2)	✓	✓	✓
Murine chaphamaparvovirus 1 (MuCPV)a	✓	✓	✓
Murine cytomegalovirus (MCMV)	✓	✓	✓
Murine kobavirus-1	✓	✓	NP
Murine kobavirus-2	✓	✓	NP
Murine picornavirus	✓	✓	NP
Murine sapovirus	✓	✓	NP
New World hantaviruses	✓	✓	✓
Old World hantaviruses	✓	✓	✓
Papillomavirus	✓	✓	✓
Pneumonia virus of mice (PVM)	✓	✓	✓
Murine polyoma virus (MuPyV)	✓	✓	✓
Reovirus-3	✓	✓	✓
Sarbecovirus	✓	✓	✓
Sendai virus	✓	✓	✓
Theiler encephalomyelitis virus (TMEV)	✓	✓	✓
Group A Streptococcus	✓	✓	✓
Group B Streptococcus	✓	✓	✓
Group C Streptococcus	✓	✓	✓
Group G Streptococcus	✓	✓	✓
Bordetella bronchiseptica	✓	✓	✓
Bordetella pseudohinzii	✓	✓	✓
Candidatus Savagella (SFB)a	✓	✓	✓
Corynebacterium bovisa	✓	✓	✓
Campylobacter spp.	✓	✓	✓
Campylobacter coli	✓	✓	✓
Campylobacter jejuni	✓	✓	✓
Chlamydia muridaruma	✓	✓	✓
Francisella tularensis	✓	✓	✓
Helicobacter spp.a	✓	✓	✓
Helicobacter bilis	✓	✓	✓
Helicobacter ganmani	✓	✓	✓
Helicobacter hepaticus	✓	✓	✓
Helicobacter mastromyrinus	✓	✓	✓
Helicobacter rodentium	✓	✓	✓
Helicobacter typhlonius	✓	✓	✓
Klebsiella oxytocaa	✓	✓	✓
Klebsiella pneumoniaea	✓	✓	✓
Leptospira spp.	✓	✓	NP
Mycoplasma pulmonis	✓	✓	NP
Pasteurella multocida	✓	✓	✓
Proteus mirabilisa	✓	✓	✓
Pseudomonas aeruginosaa	✓	✓	✓
Rodentibacter heyliia	✓	✓	✓
Rodentibacter pneumotropicusa	✓	✓	✓
Salmonella spp.	✓	✓	✓
Staphylococcus aureusa	✓	✓	✓
Staphylococcus xylosusa	✓	✓	✓
Streptobacillus moniliformis	✓	NP	NP
Streptococcus pneumoniae	✓	✓	✓
Chilomastix muris	✓	✓	✓
Cryptosporidium spp.	✓	✓	✓
Demodex spp.	✓	✓	✓
Demodex musculia	✓	NP	NP
Encephalitozoon cuniculi	NP	✓	NP
Entamoeba spp.a	✓	✓	✓
Giardia spp.	✓	✓	✓
Hexamastix muris	✓	✓	✓
Mite (Myobia musculi, Myocoptes musculinus, and Radfordia affinis)	✓	✓	✓
Mouse Eimeria spp.	✓	✓	✓
Ornithonyssus bacoti	✓	✓	✓
Pinworm (Syphacia spp. and Aspiculuris spp.)	✓	✓	✓
Pneumocystis spp.	✓	✓	NP
Spironucleus murisa	✓	✓	✓
Tapeworm	✓	✓	NP
Tritrichomonas spp.a	✓	✓	✓

Bolded agents were also tested by serology. ✓, performed in the study; N/A, not applicable; NP, not performed.

^aEnzootic in sampled vivaria. Other agents are either excluded or not monitored.

View Fullscreen

The SCD from each of the 5 vivaria (5 mice/vivarium; n = 25) was also used to expose cohoused, naive, immunodeficient NOD.Cg-Prkdc^scid Il2rg^tm1Wjl/SzJ (NSG; n = 3) and immunocompetent Swiss Webster (SW; n = 2) mice. Control mice (n = 3 NSG and n = 2 SW cohoused in a single cage) were exposed to autoclaved SCD pooled from all 5 vivaria. Mice were inoculated with a SCD suspension orally and intranasally as well as by direct contact (DC) for 7 d by adding 25 mL of unaltered SCD to the cage bedding (Figure 2). Following a 7-d exposure period, mice were transferred into a clean autoclaved cage and monitored for up to 63 d postinoculation (DPI). Detailed procedures for collecting and processing of the SCD samples are described in subsequent sections.

View Figure

• Download as Powerpoint
• Download Figure

Fecal pellets as well as fur and oral swabs were collected from each mouse on 0, 3, 7, 14, and 63 DPI. All samples from 0, 14, and 63 DPI were pooled by day and cage, while samples collected on 3 and 7 DPI were combined into a single sample and pooled by cage for analysis. Samples were stored at −80 °C (−112 °F) until they were tested by PCR for 84 murine and zoonotic viruses, bacteria, and parasites. One NSG mouse per cage was euthanized on 7 and 14 DPI while the remaining NSG and 2 SW mice were euthanized on 63 DPI. Terminal blood collection (SW mice only) was performed immediately after euthanasia and sera were stored at −80 °C (−112 °F) until tested for 20 agents by multiplexed fluorometric immunoassay (MFIA; Charles River Laboratories). Following euthanasia, deep skin scrapes of the cervical, thoracic, and sacral skin and a complete necropsy were performed.

To confirm results from the initial shoe debris exposure, an equal volume of SCD (6 mL/vivarium) was pooled and used to expose 2 cages, each with 5 NSG female mice as described previously. The mice were monitored for clinical signs and euthanized on 63 DPI. A gross necropsy was performed on each mouse. No further analyses were performed, as mice were healthy and lesion free.

Street shoes direct testing study.

As SCD accumulated over periods ranging from 30 up to 335 d, depending on reservoir volume, and was subject to recognized and unrecognized environmental stressors during collection and storage, an additional study was conducted to detect the presence of infectious agents collected directly from the soles of 27 animal care staff’s shoes on building entry for their assigned shift. The samples were collected on 26 April 2024. The temperature for the preceding 24 h was 46 to 60 °F (4.4 to 15.5 °C), and there had been a light drizzle (less than 0.1 in. of precipitation).⁴⁰ Detailed procedures for collecting and processing of the street shoe samples are described in subsequent sections.

Within 2 h of swabbing the soles with CM, each of the 27 CM samples was divided in half. One half of each sample was added directly to a cage containing 3 NSG and 2 SW female mice, while the other half was pooled and tested by PCR for 75 infectious agents (Table 1). Following a 7-d exposure period, mice were transferred into a new autoclaved cage and monitored until 63 DPI, at which point they were euthanized and deep skin scrapes of the cervical, thoracic, and sacral skin, as well as a complete necropsy, were performed.

Sample collection and processing.

Vivarium shoe cleaner study.

SCD was collected from the shoe cleaner collection reservoirs from each of 5 vivaria (A through E). Details regarding the numbers of laboratory staff that have access to each vivaria and its average daily cage census are detailed in Table 2 to provide context as to the number of people (animal care staff excluded) using the shoe cleaners. Samples were pooled and thoroughly mixed in vivaria with multiple shoe cleaners (A and C). The shoe cleaning system in vivaria A and C consists of 5 individual shoe cleaners (shoe brush 2001-TB, Liberty Industries, East Berlin, CT) with each connected through a centralized vacuum system to a reservoir into which the SCD from the associated cleaner is collected and stored. Approximately 100 mL of debris from each of the aforementioned shoe cleaner reservoirs from the same building was pooled and processed. For vivaria B, D, and E, each has a single shoe cleaner (shoe brush 2010SC, Liberty Industries) that collects debris directly into a collection bag located within the shoe cleaner. Debris (500 mL) from each collection bag was collected and processed. Samples from each vivaria were aliquoted into five 50-mL sterile conical tubes (Falcon 50-mL high clarity PP conical bottom sterile centrifuge tube, 352070, Corning, Corning, NY) containing approximately 25 mL of SCD. Excess SCD was stored for use as needed in confirmatory studies. Aliquots from each vivaria were allocated for direct PCR testing, exposure of CM (which was subsequently tested by PCR), preparing a suspension for intranasal and oral inoculation of mice, exposing mice by DC, and autoclaving for exposure of mice serving as controls.

Table 2.Cage and user census per vivaria

Vivarium	Number of laboratory staff with access to vivarium	Average daily cage census
A	306	11,285
B	34	1,529
C	537	25,552
D	41	1,660
E	100	6,222

View Fullscreen

To prepare the CM for PCR testing, a 25-mL SCD aliquot from each vivaria and the autoclaved control SCD were each transferred into individual autoclaved plastic boxes (PathogenBinder™ kit box, Charles River Laboratories) containing a new CM. The sample was processed according to the manufacturer’s instructions, which included shaking the box 20 times horizontally and 20 times vertically within a BSC (LabGard S602-500, NuAire) to ensure full exposure of the CM to the SCD over a 15- to 20-s period.¹¹ Precautions were taken to prevent cross-contamination of CM from different vivaria during sample processing, including changing gloves and disinfecting all interior surfaces of the BSC between samples with a 0.5% accelerated hydrogen peroxide disinfectant (Peroxigard RTU, Virox Technologies, Oakville, ON, Canada), allowing for a minimum of 5-min contact time before wiping up the solution with a cloth (WypAll economizer L30 ¼ fold wipers, Kimberly-Clark, Roswell, GA).

To prepare the intranasal and oral inoculum, a 15-mL aliquot from each SCD was transferred to a 50-mL sterile conical tube, sterile PBS (15 mL) was added, and the tube was vortexed for 60 s to ensure homogeneity followed by centrifugation at 10,000 × g for 20 min. The supernatant was collected and administered by the intranasal and oral routes within 1 h of preparation.

Control SCD was prepared by pooling 25-mL aliquots of SCD from each vivarium in a microisolation cage (no. 19, Thoren Caging Systems) with a filter top secured using heat-sensitive autoclave tape (Medline, Mundelein, IL). The cage was then placed within an autoclavable biohazard bag (part no. 848, autoclavable biohazard waste bags with polypropylene w/indicator, Medegen Medical Products, Hauppauge, NY) and taped shut with heat-sensitive autoclave tape. The bagged cage was then autoclaved using the cycle parameters described above. Sterilization was confirmed by autoclave tape color change and post hoc verification of cycle chamber operating conditions. Following autoclaving, control debris was prepared and used as described.

For the confirmatory study, 3 mL of SCD from each vivarium stored at room temperature for 3 mo was pooled with approximately 3 mL of freshly collected debris from the same vivaria to inoculate NSG mice with pooled SCD. NSG mice were inoculated through oral, intranasal, and DC exposure as described above. In total, these cages were inoculated with 30 mL of pooled SCD representing all vivaria.

Street shoes direct testing study.

Samples were collected by firmly pressing a new CM onto the heel of the sole, moving it from the heel to the toe to the heel, and then repeating the process on the other sole. All exposed CMs (n = 27) were pooled into a single sterile bottle (500-mL sterile Corning™ octagonal PET storage bottle with 31.7-mm screw caps, Corning) until processing within an hour after collection. No identifying information was collected from volunteers.

Each CM was cut in half on the diagonal using autoclaved stainless-steel iris scissors. Half of each CM (n = 27) was combined and introduced into a single cage housing 3 NSG and 2 SW mice within 3 h of collection (Figure 3); the mice were exposed to the CM for 7 d until the next cage change, at which point they were housed in autoclaved cages for the remainder of the study. The remaining CM halves were returned to the 500-mL sterile bottle for PCR testing for 75 murine and zoonotic agents (Table 1).

View Figure

• Download as Powerpoint
• Download Figure

Results

Discussion

Animal care programs strive to exclude specific infectious agents from rodent colonies by implementing stringent biosecurity practices and verify their absence by implementing colony health monitoring programs.^5,33,45 At our institutions, we use mechanical shoe cleaners to remove debris from shoes prior to entering vivaria. Shoe cleaners replaced the use of shoe covers, as the latter have questionable effectiveness at preventing entry and spread of potential contaminants and may, in fact, increase the likelihood for contamination of personnel hands and gowns during their donning.²⁷ Despite widespread use of shoe covers, and in our case mechanical shoe cleaners, to prevent entry of adventitious agents to vivaria, the risk of shoes serving as fomites for the spread of and/or introduction of murine infectious agents has never been directly assessed.^17,27,44

Our widespread programmatic use of shoe cleaners in an urban environment known to host large numbers of wild mice and rats provided a unique opportunity to assess the biosecurity risk posed by footwear. These wild rodents have been shown to carry MHV, MAV 1 and 2, MNV, MuAstV2, MuCPV, sarbecovirus, Klebsiella pneumoniae, Campylobacter jejuni, Streptobacillus moniliformis, Cryptosporidium parvum, and Ornithonyssus bacoti among a host of other infectious agents.^{18,20,53,56,57} While we do not exclude all of these agents from our colonies, many are bona fide biosecurity risks. In addition to assessing shoe debris collected by shoe cleaners by PCR for many murine infectious agents, we also sampled the soles of street shoes immediately on entry to the facility and we exposed both immunocompromised and immunocompetent mice to samples to ascertain whether any of these materials could transmit infectious agents to mice.

The study also provided the opportunity to compare the effectiveness of directly testing SCD with the use of CM to indirectly assess the presence of nucleic acid by PCR. While results from SCD and CM were not always congruent, overall, both samples identified a similar number of infectious agents. Discrepancies between both positive and negative results for the same sample and agent occurred 68 times. Since each CM was exposed to the same aliquot of SCD submitted for PCR, it can be assumed that these discrepancies arose from the ability of a specific agent or its nucleic acid to adhere to the CM or not, as well as the sensitivity of the PCR assay for the specific agent. The most likely reason for the failure to detect these agents likely reflected poor adherence of the specific agent or nucleic acid to the media. We presume that CM’s enhanced effectiveness reflected the specific agent’s or nucleic acid’s ability to adhere to the media, concentrating it above the detection limit of the assay. While individual assay failures could also account for false-negative results, a control was used for each reaction.

Importantly, many organisms identified by PCR in SCD and/or the CM samples, including MuAstV1, MNV, MuCPV, SFB, Chlamydia muridarum, Staphylococcus xylosus, Corynebacterium bovis, Helicobacter spp., Klebsiella oxytoca, Spironucleus muris, and Tritrichomonas spp., are enzootic in most colonies in our vivaria. The study could not ascertain whether the source of nucleic acid for these agents was from NYC streets or vivaria floors, as most users enter and exit the facilities multiple times daily using the shoe cleaner on each entry. Nucleic acid for 22 tested agents could only have originated from shoes contaminated with outdoor debris, as these agents are excluded from our colonies. Most importantly, regardless of the nucleic acid source, none was infectious, as none of the mice exposed to SCD and/or CM, or inoculated with extracts from these materials, became infected with any agent.

We routinely test wild mice captured in or near our vivaria to identify agents of concern to our research mouse populations. These mice frequently are positive for LDV, MCMV, MPV-1, MPV-2, MPV-5, MNV, MTV, MuAstV1, MuAstV2, MuCPV, murine picornavirus, and EDIM. Therefore, it was not surprising to find that the SCD and CM were positive for many of these agents. Interestingly, while MCMV, MTV, EDIM, LDV, and murine picornavirus are commonly detected in wild mice, they were never detected in any SCD or CM sample. While endoparasites and ectoparasites are among the most common contaminants of contemporary mouse colonies, it would appear unlikely that their introduction or spread is the result of contaminated footwear.^15,43,55 While none of the SCD or CM was shown to contain infectious material, the detection of ectromelia, MAV 1 and 2, MVM, and MPV is of significant concern, as these agents are commonly excluded from research mouse colonies and both MAV 2 and MPV appear to be circulating within colonies of wild NYC mice.⁵⁶

Some of the bacterial or protozoal agents assessed in this study for which nucleic acid was detected in the SCD and/or CM samples are excluded from all vivaria tested in this study. Therefore, as with some of the viruses, detection of these agents cannot be attributed with certainty to shoes contaminated with outdoor debris. Although we do not test and exclude sarbecovirus from our colonies, we assume it initially originated from outdoor debris, given that it was nonexistent in the United States before 2020. SCD and/or CM samples from 3 vivaria tested positive for Ornithonyssus bacoti, an ectoparasite excluded from all vivaria. Surprisingly we did not detect other mites or nematodes in SCD or CM even though we frequently find murine pinworms (for example, Syphacia obvelata) and fur mites (for example, Radfordia affinis) in wild mice captured in or near our vivaria.

In addition, although not specifically reported in NYC wild mouse or rat populations, rodents are known to carry other pathogens such as Bordetella bronchiseptica, Campylobacter coli, Filobacterium rodentium, Francisella tularensis, Leptospira spp., Salmonella spp., LCMV, Eimeria spp., and Hymenolepis spp.^{1,18,20,23,24,39,48,53,55–57,59} PCR results from SCD and CM were negative for all these agents, suggesting a low biosecurity risk.

Interestingly, the control CM and SCD samples tested positive for several agents despite being autoclaved. Previous studies have shown that 15 min of autoclaving at 121 °C (250 °F) and 15 psi is insufficient to degrade DNA and prevent nucleic acid reamplification during PCR.¹⁶ Using these temperature and pressure parameters, it can take up to 2 h of autoclaving to eliminate all nucleic acids.²¹ Our autoclave cycle consisted of a pulsed vacuum cycle with 4 pulses, a maximum pressure of 12 psi, and a temperature of 121 °C (250 °F) for 20 min. Therefore, it is not surprising that we detected nucleic acids in our samples. However, this is not consistent with previous work assessing infectious agent transmission from autoclaved cages.^10,42,47

The most surprising and important aspect of this study was that SW and NSG mice directly inoculated with and housed in direct contact with SCD and/or CM exposed to SCD did not become colonized or infected with any of the agents found in SCD and/or CM, with perhaps Pseudomonas aeruginosa being the sole exception. As mice used in this study were infected with MuAstV1, SFB, and Staphylococcus xylosus, as these agents are enzootic in their colonies of origin, we could not assess whether SCD/CM could transmit these agents; however, based on the findings with numerous other agents detected in this study, we think their transmission would be highly unlikely.

Mice exposed to SCD from vivarium B were PCR positive for beta Streptococcus group B and Hexamastix muris in the pooled day 3 and 7 samples; however, the copy numbers were extremely low, and the negative results from samples collected at later time points indicate that these results were likely false positives. Similarly, at 63 DPI, mice in the control group exposed to the autoclaved SCD inoculum tested positive for Demodex spp., Demodex musculi, and Pseudomonas aeruginosa by PCR. The copy numbers for Demodex spp. and D. musculi were low at 100 and 12, respectively. Although we have seen Demodex musculi transmission to soiled bedding sentinels in our health monitoring program, colonization typically occurs from dam to offspring, and mite numbers in immunocompromised mice are usually high enough to consistently identify positive results by skin scrapes and PCR.^38,50 Considering the inoculum was autoclaved, the low copy number, and negative confirmatory tests on NSG mice (deep skin scrapes and skin biopsies), we consider the PCR result to most likely reflect a false positive result.

Pseudomonas aeruginosa was also detected in the autoclaved control SCD and vivarium D samples on day 63; however, the copy number for the control sample was very low (25), the mice had not tested positive at earlier time points, and the NSG mice were normal clinically at necropsy, suggesting that these results also likely reflect false positive results. Highly immunocompromised mice such as NSG, SCID, and NOG are known to be highly susceptible to Pseudomonas aeruginosa–induced disease.^14,41 The copy number for vivarium D was 1622, suggesting the bacterium may have been present in the sample. However, none of the earlier samples was positive. This timing raises doubts as to whether infection resulted from exposure to SCD or, more likely, another source. Pseudomonas aeruginosa is a common environmental opportunist found ubiquitously in soil and water, with horizontal transmission occurring within 5 d in mice.³⁰

SW mice exhibited lymphoid and gut-associated lymphoid tissue hyperplasia at the final timepoint. Gut-associated lymphoid tissue hyperplasia has been associated with Chlamydia muridarum infection.³⁷ However, MOMP immunohistochemistry staining of large intestinal tissue and testing of pooled lymph node biopsies by PCR were negative for the bacterium. We could not identify an infectious etiology. We speculate that these lesions likely reflected antigenic stimulation following exposure to the SCD. All other microscopic findings were considered incidental, spontaneous, strain, or age related, and were not associated with SCD exposure.^8,35,49

As the vivarium shoe cleaner study results were unexpected, we inoculated additional NSG mice with pooled freshly collected and original sample SCD from all vivaria. These mice remained clinically normal and lesion free at 63 DPI. As the SCD used in the initial experiments had been collected over a period of up to 335 d, we also assessed footwear directly by collecting samples from the soles of animal care staff’s street shoes immediately on facility entry using CM and exposed mice directly to these CM samples. While nucleic acid from 11 agents was detected in an extract from the pooled CM sample, all NSG and SW mice exposed to contaminated CM samples were PCR negative at 63 DPI. However as discussed, the mice used in this study were MuAstV1 and SFB positive, and therefore we could not assess whether these agents could be transmitted via contaminated footwear.

We encountered 6 PCR assays yielding inconclusive results, preventing determination of the full range of infectious agents that may have been present in the samples. These agents of interest included MPV-4, SFB, Campylobacter coli, Campylobacter jejuni, Helicobacter rodentium, and Streptobacillus moniliformis. Inconclusive results were obtained more frequently in SCD samples as compared with CM, and when testing for bacteria as compared with viruses or parasites. However, if these agents were present and infectious, we would have detected them in the samples collected from the exposed and inoculated mice.

In conclusion and importantly, these results demonstrate that while nucleic acid from many rodent pathogens can be found on footwear, it is unlikely that footwear poses a biosecurity risk to research mouse colonies, questioning the necessity of not only the use of shoe cleaners, but also the use of shoe covers as a component of a rodent biosecurity program.