Intracage Ammonia Levels in Static and Individually Ventilated Cages Housing C57BL/6 Mice on 4 Bedding Substrates

Abstract

The relationship among ammonia levels, cage-changing frequency, and bedding types is an important and potentially controversial topic in the laboratory animal science community. Some bedding options may not provide sufficient urine absorption and bacterial regulation to minimize ammonia production during the interval between cage changes. High intracage ammonia levels can cause subclinical degeneration and inflammation of nasal passages, rhinitis and olfactory epithelial necrosis in exposed mice. Here we sought to compare the effects of 4 commonly used bedding substrates (1/4-in. irradiated corncob, reclaimed wood pulp, aspen wood chips, and recycled newspaper) on ammonia generation when housing female C57BL/6 mice in static and individually ventilated caging. Intracage ammonia levels were measured daily for 1 wk (static cage experiment) or 2 wk (IVC experiment). The results of this study suggest that the corncob, aspen wood chip, and recycled newspaper beddings that we tested are suitable for once-weekly cage changing for static cages and for changing every 2 wk for IVC. However, ammonia levels were not controlled appropriately in cages containing reclaimed wood pulp bedding, and pathologic changes occurred within 1 wk in the nares of mice housed on this bedding in static cages.

Minimizing environmental causes of physiologic alterations within the research population has long been a focus of laboratory animal medicine. Past studies have investigated the effects of various bedding substrates on ammonia levels in static cages^25,35 or evaluated ammonia levels in individually ventilated cages (IVC) using a single type of bedding substrate.^{12,19,24,26,30,36} However, there are no published studies that assess the effects of various bedding substrates on ammonia levels in modern IVC systems, making it difficult for facility managers and veterinary staff to determine appropriate cage-changing intervals. In addition, the 8th edition of the Guide for the Care and Use of Laboratory Animals¹³ requires that institutions focus on performance data and microenvironmental conditions when determining acceptable cage sanitation intervals. Numerous studies have described detrimental health effects of increased ammonia levels on laboratory rodents.^2,4,10,35,36 Most of these effects are subclinical but often include histologic and immunologic changes²⁸ that potentially represent unacceptable environmental variables affecting research results.

Although numerous laboratory bedding substrates are available, each has advantages and disadvantages, which are based in both science and the personal opinion of the user. The purpose of the bedding substrate in a rodent housing cage is multifaceted. According to the Guide, “Bedding is used to absorb moisture, minimize the growth of microorganisms, and dilute and limit animals’ contact with excreta, and specific bedding materials have been shown to reduce the accumulation of intracage ammonia.”¹³ Dust content, enrichment value, animal comfort, and cost are also factors to consider when selecting a bedding substrate.^7,15,21

Because of product availability, our institution had switched from aspen woodchip bedding to reclaimed wood pulp (RWP) bedding several years ago. Numerous investigators preferred RWP, believing it provided greater comfort for mice. The institution continued the once-weekly cage changes used with aspen bedding. Simultaneously, the institution was gradually changing to ventilated cages, in part because of the labor savings associated with the 14-d cage-change intervals. After the bedding switch, facilities developed a strong ammonia odor. Given these concerns, the institution switched from RWP to 1/4-in. corncob bedding. This switch was made for several reasons, including to minimize animal and personnel exposure to ammonia and to decrease exposure to potential dust-borne allergens, both of which were perceived to be higher in RWP cages than in corncob cages. Several investigators believed their mice failed to breed on corncob bedding and questioned whether the ammonia levels in the RWP cages were significant.

This current study was performed to determine whether the ammonia odor associated with RWP represented deleterious ammonia levels necessitating more frequent cage changing. In this study, we evaluated 4 similarly priced, commonly available bedding options to determine which provided the best balance between optimizing animal health and minimizing ergonomic stress for animal care staff.

Materials and Methods

Mice and husbandry.

Four bedding substrates were used: corncob (Harlan 1/4 inch irradiated, Harlan Laboratories, Madison, WI), RWP (Harlan Tek-Fresh, Harlan Laboratories), aspen woodchips (Harlan Irradiated Teklad Sani Chips, Harlan Laboratories), and recycled newspaper (Paperchip, Shepherd Specialty Papers, Watertown, TN). Each cage received 1 L of one type of bedding substrate by using a plastic 1000-mL beaker for the purpose of consistency, given that husbandry technicians generally fill cages to a certain depth rather than a certain mass.³ Female C57BL/6 mice (age, 7 wk) were obtained from The Jackson Laboratory (Sacramento, CA) and housed at 5 animals per cage. Five cages of mice were housed on each of the 4 bedding substrates, for a total of 100 mice in each of the 2 study experiments (Figure 1). A control cage was included in each bedding group and included bedding, food, and water but no mice.

Figure 1.

Bedding groups and cage density.

Rodent laboratory chow (Harlan Teklad F6 Rodent Diet, Harlan Laboratories) and water were provided ad libitum. Housing rooms were maintained at an ambient temperature of 65 to 74 °F and 70 to 74 °F, relative humidity of 35% to 65%, and a 12:12-h light:dark cycle. Environmental enrichment was provided with 1 piece of shredable nesting material per cage (Nestlet, Ancare, Bellmore, NY). During the static-cage experiment of the study, mice were housed in 75-in.² static polysulfone cages with microisolation tops (Max 75, Alternative Design Manufacturing and Design, Siloam Springs, AR). During the IVC experiment of the study, mice were housed in 82-in.² ventilated polysulfone cages (Blue Line, Tecniplast Buguggiate, Italy) with 60 air changes hourly. Air flow was confirmed by means of a smoke test prior to placing mice in the cages. Mice were acclimated for 1 wk on their respective bedding substrates, and a new cage base with new bedding was provided prior to beginning measurements.

Soiled-bedding sentinel mice are tested quarterly for and remain free of mouse hepatitis virus, Sendai virus, pneumonia virus of mice, Mycoplasma pulmonis, Theiler mouse encephalomyelitis virus, reovirus 3, mouse oarvovirus, epizootic diarrhea of infant mice virus, lymphocytic choriomeningitis virus, ectromelia virus, pinworms, and ectoparasites. This study was conducted between May 2012 and September 2012 and was approved by the University of California Berkeley's IACUC. The University of California Berkeley is accredited by AAALAC.

Ammonia measuring procedure.

Ammonia measurements were performed daily between 1330 and 1530 by using a chip measurement system (model 6405300, Dräger Safety, Pittsburgh, PA), consisting of gas-detecting chips and an analyzer fitted with a remote system (model 6405060, Dräger Safety). Ammonia chips with a detection range of 10 to 150 ppm or 100 to 2000 ppm were used; the higher range chip was used when a value greater than 100 ppm was detected initially by using the lower-range chip. Chips reportedly are accurate to 8% to 10% of the measured value and reproducible at 10% (SD).^8,9 Flushing of the sampling hose was performed according to the manufacturer's instructions, to minimize or eliminate effects associated with the use of the sampling hose, such as memory effect or dead volume. Room air measurements were performed on the first day, the last day, and on a day halfway through each experiment in the study. Cages were placed on the housing racks in random order, to alleviate possible variations in light intensity, room air circulation, and temperature. Cages were numbered and measured in random order. Ammonia levels were measured in every cage daily. Average ammonia level (ppm) for each bedding group was calculated from the 5 test cages in each group for each day. During the static-cage experiment of the study, a housing rack was designated for the purpose of this study in an otherwise full housing room. During the IVC experiment of the study, cages assigned to the study were housed on one side of an IVC rack; the other side was full of cages owned by another investigator, again in an otherwise full housing room.

For measurements performed during the static caging experiment of the study, the tip of the remote system hose was inserted at the front of the cage to 2 cm above the bedding surface, which is the approximate height of a mouse's nose. To insert the hose, a small perforation was made in the filter top of each cage in the same location, just wide enough for the hose to fit through (Figure 2). Between measurements, this hole was covered with a small sticker to decrease the opportunity for ammonia to escape. For measurements performed during the ventilated caging experiment of the study, the tip of the remote system hose was inserted through the sipper tube opening at the front of the cage and advanced to the right rear corner of the cage under the exhaust port (Figure 3). Measurements were performed with the tip of the hose 1 to 2 cm above the bedding surface, as was done in the static caging experiment. Cage tops were not removed for the duration of the study (7 d for static cages and 14 d for individually ventilated cages).

Figure 2.

(A) Small perforation in the filter top of each static cage, made in the same location and just wide enough for the hose to fit through. (B) Between measurements, the hole was covered with a small sticker to decrease the opportunity for ammonia to escape. (C) Mouse in static cage demonstrating position of hose during measurements.

Figure 3.

(A) Insertion of hose through water-bottle port for measurements in IVC. (B) Hose position 1 to 2 cm above bedding surface in IVC.

Nasal histopathology.

Mice were euthanized via carbon dioxide inhalation on day 7 (static-cage experiment) or day 14 (IVC experiment) immediately after measurement of intracage ammonia levels. Gross necropsies were performed, and tissues were collected from all study animals. Histopathologic analysis was performed on 1 mouse from each cage, to examine for potential effects of increased ammonia levels. Subsequently, the remaining 4 mice in the RWP cages were analyzed due to the presence of nasal pathology in all of the mice examined initially. For these analyses, the mouse's head was fixed in 10% neutral buffered formalin, decalcified, and embedded in paraffin. Coronal sections (thickness, 5 µm) were made at 1 mm from the tip of the nose, 4 mm from the tip of the nose, and the medial canthus of the eye. These 3 sections were mounted on glass slides, stained with hematoxylin and eosin, and examined for lesions by a blinded evaluator (RVA).

Statistics.

Statistical significance was preset at a P value of less than 0.05. Differences in measured ammonia levels between groups were analyzed by using a Kruskall–Wallis test. Analyses were generated by using VassarStats.¹⁸ The cage served as the experimental unit. Ammonia measurements of less than 10 ppm were assigned a value of zero for the purposes of data analysis.

Results

There were no clinical health issues reported during daily health checks for any of the mice assigned to this project for the duration of the study. During the IVC experiment of the study, one cage with corncob bedding flooded and was excluded from data analysis. All other cages were included in analysis, regardless of ammonia level.

Ammonia measurements.

Results are reported as the daily average ammonia concentration (ppm) for each bedding type. During the static-cage experiment, ammonia levels in all RWP static cages exceeded the commonly used human exposure limit of 25 ppm, with these cages rising to an average concentration of 264 ppm by day 7 (Figure 4). These ammonia levels were significantly (P = 0.0016) higher than those of the corncob, aspen woodchip, and recycled newspaper groups (Kruskall–Wallis H = 15.21) over a 7-d period. During the first week of the IVC experiment, ammonia levels were undetectable in all cages. However during the second week of this experiment, ammonia levels were significantly (P = 0.0016) higher for the RWP group than for the corncob, aspen woodchip, and recycled newspaper groups (Kruskall–Wallis H = 15.21; Figure 5).

Figure 4.

Average ammonia concentrations in static cages. Each time point represents 5 cages of 5 mice. Bars, standard error. Corn cob (CC), aspen wood chip (AWC), reclaimed wood pulp (RWP), and recycled newspaper (RN).

Figure 5.

Average ammonia concentrations in IVC. Each time point represents 5 cages of 5 mice. Bars, standard error. Corn cob (CC), aspen wood chip (AWC), reclaimed wood pulp (RWP), and recycled newspaper (RN).

Nasal histopathology.

No lesions were detected in any of the mice analyzed that were housed on corncob, paper chip, or aspen chip bedding in either static or ventilated cages. Mice housed on RWP in static cages demonstrated marked nasal pathology on day 7. All static-cage mice housed on RWP demonstrated generalized epithelial necrosis of the turbinate and septal surfaces, generalized inflammatory cell infiltrates with a dominance of neutrophils, multifocal depletion of cilia, and variable multifocal submucosal edema, congestion, and hemorrhage (Figure 6). Interestingly, although ammonia levels did not exceed 25 ppm in the standardized measuring location, early changes suggestive of mild airway insult, including turbinate edema and vasodilation, were present in RWP IVC mice.

Figure 6.

Representative histologic sections from mice in static cages on (A and B) aspen chip and (C and D) RWP bedding. Generalized epithelial necrosis and inflammatory cell infiltrates (for example open arrows) are evident in mice housed on RWP, as are focal congestion (asterisk), edema (closed arrows), and hemorrhage. All sections are 4 mm from the nares. Magnification, 10× (A and C), 40× (B and D).

Discussion

In this study, we compared 4 bedding substrates in static and IVC for their capacity to minimize ammonia levels over a 7- or 14-d cage-change interval. These results suggest that in static cages with 5 mice, nonRWP bedding substrates offer the potential for once-weekly cage changes whereas RWP requires more frequent changes. The data from IVC were less definitive, given that ammonia levels were relatively low at the standardized sites, regardless of bedding type. This result may be expected, given the nature of airflow dynamics within IVC. Intriguingly, nasal edema was evident in several mice bedded on RWP in IVC despite their relatively low ammonia measurements. This finding may be the result of higher ammonia levels in the latrine sites in these cages. In fact, ammonia levels as high as 63 ppm (data not shown) were measured in RWP latrine sites, whereas ammonia was undetectable in latrine sites from the other 3 bedding substrates evaluated. Latrine sites were determined by noting the area of the cage containing wet discolored bedding. This difference in ammonia levels in 2 different locations within the ventilated cages further demonstrates the potential complexity in evaluating cage type, bedding substrates, and husbandry practices.

Additional studies are necessary to confirm that the pathologic changes we noted in this study are due strictly to the presence of increased levels of ammonia in the microenvironment. Other possible contributors to nasal pathology include—but are not limited to—endotoxin, dust, and coliform levels in bedding^34,37 as well as increased intracage temperature, humidity, and carbon dioxide.^5,17,27,29 However, previously published data suggest that endotoxin levels are lowest in paper beddings, dust levels are highest in hardwood beddings, and coliform counts are highest in corncob beddings.³⁷ We believe that these data support our conclusion that ammonia levels contributed to nasal pathology in the paper bedding group. Definitive proof of this hypothesis requires precise studies using nose-only or whole-body exposure inhalation systems.

To date, ammonia standards have not been established for rodent caging. The National Institute for Occupational Safety and Health standard for humans (25 ppm) is set as an 8 h time-weighted exposure limit.²⁰ In the past, this standard of 25 ppm has been used as a guideline for the maximal level in rodent cages,^1,7,11,14,23 but conflicting reports on the effects of increased ammonia on the rodent respiratory tract have hindered the development of a laboratory-animal–specific exposure limit.⁶ Many authors disagree regarding whether 25 ppm is too high or too low for optimal rodent health and welfare. Those studies suggesting that this level is too high note that rodents are constantly exposed to ammonia,^16,22 whereas the 25-ppm standard is for a daily 8 h working day exposure.²⁶ Those suggesting that the limits should be higher note that wild rodents often live in crowded underground burrows with little airflow.^11,26 To our knowledge, no studies have measured ammonia levels in natural burrows, but most rodent species, including mice, are thought to ‘latrine,’ meaning that urine is frequently removed from nesting sites.³³ Although additional studies are necessary to determine an upper ammonia limit for mice, the results of our current study indicate that lesions due to uncontrolled ammonia develop as early as 7 d, and RWP was unable to maintain ammonia levels at or below 25 ppm as early as one day after cage change in static caging.

We examined husbandry practices, bedding substrates, and consequences of uncontrolled ammonia exposure in mice and demonstrated a challenge in creating a one-size-fits-all approach to cage management. Intracage ammonia levels are influenced by several factors, including the sex, strain, and age of the animals housed; number of animals in the cage; temperature; humidity; number of air changes hourly; and amount and type of bedding substrate used.^{7,25,26,30,32,36} We chose small female mice to minimize the risk of losing mice to fight wounds and the subsequent need to remove cage tops throughout the study. In addition, we chose C57BL/6 mice because this strain is a common background for transgenic mice. Although these choices standardized the production of ammonia during our study, ammonia levels would likely be even higher in cages housing larger mice, male mice, and diabetic strains.^23,25,36 Cages with fewer than 5 mice may have lower intracage ammonia levels, whereas cages with nursing pups will certainly report much higher levels.^6,24 Therefore, we believe the conditions that we chose represent midlevel ammonia exposure observed under standard husbandry conditions. However, cages in this study were not opened during the 7 or 14 d between cage changes, to minimize air turbulence variables in ammonia production. As a result, the ammonia levels we noted may not be representative of levels in cages that are opened at least once—if not several times—during the week.³¹

Historically, many husbandry standards have developed from accepted practices rather than systematic data collection. Federal policies and accrediting agencies are increasingly requiring data to support performance-based standards. Therefore, there is a justifiable need to provide such data. As shown in the current study, ammonia and potentially other microenvironmental contaminants are associated with laboratory rodent health. Although some historic studies have evaluated the health effects of intracage ammonia levels, our current study is the first to systematically compare several bedding substrates and cage types for SPF mice under modern husbandry practices. Although ammonia exposure limits have not been established for laboratory rodents, the presence of clear pathology after 7 d in a static cage demonstrates the need for additional systematic evaluation to set such a standard.

In conclusion, this information will enable facility managers and veterinary staff to evaluate the acceptability of these bedding substrates with regard to cage-change interval, laboratory animal health, and ability to meet the needs of investigators. These results are immediately applicable in laboratory animal facilities using these bedding types and document the importance of controlling ammonia levels in rodent health and management.