Social Housing of Male CD-1 Mice (Mus musculus) in the Toxicological Setting: A 28-Day Social Compatibility Study of Male Mice following Oral Gavage of Theophylline
Group or pair housing of social animals in laboratory settings is beneficial to animal welfare. Social housing of male mice can be difficult due to possible aggressive behaviors, leading to injury and resulting in animals not being able to continue in the study. Techniques used to decrease male mouse aggression and territorial behaviors have been described in recent literature, including pairing mice before sexual maturity, moving used nesting material over during cage changes, removing high-value items that can encourage territorial behaviors, and using low-stress handling techniques. Using these suggested tactics, we conducted a 28-day study determining the social compatibility of male CD-1 mice in a standard toxicological study design. Forty-eight mice, aged 5 weeks old, were equally divided into single and pair housing and were administered theophylline daily. Animals were exposed to common toxicology study procedures known to cause additional stress including repeated blood collections, daily dosing, weekly clinical observations and body weights, and terminal urine collections. Behavioral assays, including nest scores and time-to-integrate-nest-material testing, were performed weekly, and pelt scores were collected postmortem. Fecal samples were collected intermittently for fecal corticosterone metabolite analyses. No mouse pairs required separation throughout the dosing phase of study, and no significant differences were observed that would affect toxicology studies. This suggests that by using techniques to decrease agonistic behaviors, male mice can be successfully socially housed on acute and subacute toxicology studies.
Introduction
The cohabitation of social species is considered to be an essential feature of animal husbandry and is expected by most regulatory agencies.1 However, social housing of male mice in laboratory settings continues to be a debated practice despite female mice being commonly and successfully housed together. Male mice display territorial behaviors toward conspecifics as well as engaging in hierarchical disputes. These behaviors include chasing, aggression, and wounding to the point of severe injury or death, which can result in not only poor animal welfare but also the loss of valuable study data.2–7 Due to these substantial risks, male mice are commonly singly housed in laboratory settings, or the use of males is avoided altogether in study designs.6,8
While permanently singly housing male mice removes the risks described above, it has a significant impact on the physical and mental health of the animals, potentially affecting study results. Male mice exposed to social isolation have been observed to exhibit anxious and depressive-like behaviors, reduced body weight gain, increased food consumption, increased heart rate, and reduced body temperature.9–14 Chronic stress through social isolation can permanently affect the development and neuroplasticity of the brain and profoundly affect learning and memory in mice.14,15 Leaving male mice singly housed removes the known opportunities and benefits gained from social housing, including expression of natural species-specific behaviors such as grooming, building nests together, and sleeping within close proximity to one another, which aid in thermoregulation.1,6,8,16 Social housing also facilitates social buffering, enabling animals to recover more effectively from stress.17 This strategy is particularly beneficial in laboratory settings, where stress-inducing situations may arise.
The Guide for Care and Use of Laboratory Animals emphasizes the importance of social compatibility, stating, “risks of social incompatibility are greatly reduced if animals to be grouped are raised together from a young age, if group composition remains stable, and if the design of the animals’ enclosure and their environmental enrichment facilitates the avoidance of social conflicts.”1 Because social housing of male mice can be challenging, a variety of strategies have been explored in the literature to reduce aggression between male mice when social housing. Current research recommends that male mice be placed into socially compatible groups at a young age or within sibling groups to reduce aggression by improving familiarity within the pair.1,4,8,18,19 Age recommendations for groups range between 3 and 5 weeks of age, recognizing that certain study types and designs may require different ages. Habituating mice to handling at this young age and using low-stress handling techniques can reduce stress and decrease aggressive behaviors toward handlers and conspecifics.10,19–21 Likewise, social compatibility improves with the addition of environmental enrichment, particularly nesting material in the case of male mice.6,22 Transfer of used nesting material at cage cleanings is recommended to allow odor familiarity in the new cage, decreasing the need to reestablish social hierarchies as mice use their sense of smell to recognize each other and determine their relative social statuses.2,8,19,23,24 In addition, the structural integrity of a nest can be an effective indicator of social compatibility in mice, allowing for cage-side evaluation of compatibility with minimal disturbance.23,25 Care should be taken when adding other enrichment types; however, as high-value enrichment, such as rigid structures, has been found to increase aggression between conspecifics.6,26,27 Since these structures can be easily monopolized by one individual, the need to compete for limited resources may increase, consequently damaging the stabilization and compatibility of a cohort.
Despite known strategies for facilitating the social housing of male mice, social housing continues to be avoided in many toxicological settings within the United States due to historic study design and concerns for interference with study data. These studies typically require frequent handling of animals and performance of study functions (the dosing of test material, blood collections, behavioral assays, and obtaining regular body weights), which causes additional stress for the animals. This stress potentially promotes aggressive behaviors toward handlers and conspecifics.6,28–30 In addition, there is concern about the impact of test articles affecting social compatibility. Annas et al23 anecdotally demonstrated successful social housing of male mice in toxicological studies but did not evaluate if treatment with the test material had a statistically relevant impact on stress and social hierarchy.
We hypothesized that male mice could be successfully socially housed and complete a 28-day toxicological study even when administered a test article with known neurologic side effects (theophylline). The goal of the current study was to demonstrate that social housing of male mice does not statistically impact toxicological study results and may be a viable option for improving animal welfare in laboratory settings.
Materials and Methods
Study design.
The current study was designed to assess the impact of pair housing on standard biologic variables routinely assessed in toxicological studies when mice were administered a test article with known toxicological effects (theophylline). Mice were assigned to single or pair housing and then into one of 3 dose groups: a control group (receiving 0.9% NaCl), a low-dose theophylline group (receiving 50 mg/kg theophylline), or a high-dose theophylline group (receiving 200 mg/kg theophylline) according to the study design shown in Table 1. There were a total of 6 groups (single-housed compared with pair housed) of 8 animals for a total of 48 animals. The experimental unit was determined to be each individual animal, as this is consistent with current statistical practices for toxicological studies and allows for the individual assessment of key biologic variables. For variables in which individual data could not be obtained (eg, nest score, time to integrate nest material testing [TINT]), the result from the cage was assigned to each individual animal.
| Group no. | Test material | Dose level, mg/kg/d | Dose volume, mL/kg | Dose concentration, mg/mL | Housing | No. of animals |
|---|---|---|---|---|---|---|
| 1 | 0.9% NaCl | 0 | 10 | 0 | Single | 8 |
| 2 | Theophylline | 50 | 5 | 8 | ||
| 3 | 200 | 20 | 8 | |||
| 4 | 0.9% NaCl | 0 | 0 | Pair | 8 | |
| 5 | Theophylline | 50 | 5 | 8 | ||
| 6 | 200 | 20 | 8 |
Test material.
Theophylline (100%; CAS no. 58-55-9; C7H8N4O2; molecular weight: 180.16 g/mol) was obtained from Sigma Aldrich (St. Louis, MO). The anhydrous 99% powder form was initially mixed with 3 potential vehicles (0.9% NaCl, corn oil, and sterile water) at 20 mg/mL to determine the most suitable vehicle for dosing, of which 0.9% NaCl was determined to give the best results for homogeneity and ease of dosing. For dosing, theophylline was mixed with 0.9% NaCl at 5 and 20 mg/kg using a 5-minute mechanical stir via stir bar and magnetic stir plate, followed by 5 minutes of homogenization with a Silverson Vortex Mixer (East Longmeadow, MA), and finally stirred again mechanically with a stir bar and stir plate. All dosing formulations were stocked in amber glass jars with lids and released to room temperature storage for dosing.
Experimental animals.
Fifty-four male Crl:CD1(ICR) [CD-1] mice were obtained from Charles River Laboratories (Raleigh, NC). Animals were 5 weeks of age upon arrival. General weight range, commensurate with age, was expected to be between 26 and 41 g and confirmed within 3 days of arrival. Forty-eight animals were assigned to study and were either single or pair housed depending on group assignment; the additional 8 animals that were maintained as replacement animals were all pair housed. Animals were housed in open-topped, transparent polycarbonate cages (406.5 cm, 2 × 12.7 cm in height; Allentown, Allentown, NJ). Bedding was aspen woodchips (Northeastern Products, Warrensburg, NY). Food (Block Lab Diet Certified Rodent Diet No. 5 CR4; PMI Nutrition International, Richmond, IN) and municipal tap water were both available ad libitum. Drinking water was treated with 0.5-2 ppm chlorine dioxide and filtered to 5 μm. Quarterly testing for coliform bacteria and heterotrophic plate counts were performed on the water, as well as annual testing for all primary contaminants included in the US Environmental Protection Agency National Primary Drinking Water Regulations.31 Transfers to clean caging were performed weekly.
All cages contained one Bed-r’Nest 8g nesting puck (The Andersons, Maumee, OH) per animal and 1 pinch of ALPHA-twist (Shepherd Specialty Papers, Watertown, TN). Each cage was allotted one red translucent Mouse Tunnel (Bio-Serv, Flemington, NJ); singly housed animals received the tunnel immediately, while socially housed animals received the tunnels after 2 weeks to minimize territoriality surrounding the durable item while dominance hierarchies were being established. Bed r-Nest nesting pucks were transferred at each cage cleaning for socially housed animals. Low-stress handling practices21 (using either the cupping method or the tunnel) were used exclusively when removing mice from their home cages. Temperature and relative humidity were monitored and recorded daily and maintained between 20 to 26 °C and 30% to 70%, respectively. A 12-hour:12-hour light:dark cycle was maintained using fluorescent lights, with all functions performed during the light cycle. Mice were procured free of ectromelia virus, Hantaan virus, K virus, lactic dehydrogenase elevating virus, lymphocytic choriomeningitis virus, minute virus of mice, mouse adenovirus type 1 and 2, mouse cytomegalovirus, mouse hepatitis virus, mouse parvovirus, mouse polyoma virus, mouse rotavirus, mouse thymic virus, murine norovirus, pneumonia virus of mice, respiratory enteric virus III, Sendai virus, Theiler murine encephalomyelitis virus, Bordetella bronchiseptica, Filobacterium rodentium, Corynebacterium bovis, Corynebacterium kutscheri, Mycoplasma pulmonis, Salmonella spp., Streptobacillus moniliformis, Clostridium piliforme, Helicobacter spp., Klebsiella pneumoniae, Staphylococcus aureus, Rodentibacter pneumotropicus, endoparasites, and ectoparasites. Facility standard operating procedures dictated that sentinel testing be performed quarterly; due to the short duration of the study and the specific period of time during which the animals were on site, no additional health monitoring was performed on these animals. Colonization by commensal organisms from colony animals was anticipated due to the open-topped housing status of the animals.
Sample size.
Group size and number of groups were based on Organisation for Economic Co-operation and Development (OECD) guidelines for 28-day repeat dosing,32 with modifications based on a single sex being used for the current study and the use of microsampling for toxicokinetic (TK) sample collections. Each group was comprised of 8 animals per group with a control group, a low-dose theophylline group, and a high-dose theophylline for both single- and pair-housed animals (48 animals total). As theophylline has a well-known TK profile, in accordance with 3Rs practices, it was deemed unnecessary to include a mid-dose group.
Inclusion and exclusion criteria.
The first 48 mice removed from the shipping boxes were assigned to the study. Remaining animals not assigned to the study were retained as spares for replacement if necessary. It was predetermined that animals would be separated and removed from the study if pairs required separation for any reason (eg, incompatibility, loss of an individual, etc) or if any clinical abnormality could not be treated with veterinary support. Animals were deemed incompatible if aggression escalated to the point of obvious physical injury or guarding of resources or continuous harassment progressed to the point that there was a significant clinical impact on an animal (significant weight loss or dehydration). One cage of pair-housed animals allocated to the control group was found to be incompatible during the first week of the stabilization period, where aggression resulted in wounding and therefore the pair was excluded from the study and replaced by one pair from the spare replacement animals. An additional pair of animals in the low-dose theophylline group was removed and replaced on the first day of dosing due to premature euthanasia of one of the pair as a result of blood collection complications.
Subject allocation.
Mice were assigned to groups upon receipt at the facility based on the order in which they were removed from the shipping crate (ie, the first animal removed from the crate became the first animal in group 1, the second animal removed became the first animal in group 2, etc). This “randomization upon receipt” is the primary practice used for rodents received at all Charles River Laboratory facilities.
Blinding.
Individuals performing the dosing, observations, and body weight evaluations were not blinded. Those performing nest scoring, TINT testing, fecal corticosterone metabolite analysis, clinical pathology analysis, and pelt aggression lesion scores were blinded.
Experimental procedures.
Animals were given 2 weeks to acclimate to the facility before per os dosing began (aka the stabilization period). Starting on study day 1, groups 1 and 4 animals received 0.9% NaCl, while groups 2 and 5 received 50 mg/kg/d theophylline, and groups 3 and 6 received 200 mg/kg theophylline via oral gavage. Dosing continued daily 5 days per week (Monday to Friday) for a total of 20 doses at approximately 9:00 am each day. Animals were then evaluated for body weight, clinical observations, nest scores, TINT test scores, fecal corticosterone metabolite values, clinical pathology values, and pelt aggression lesions scores (PALS), as described below. Figure 1 depicts the study timeline.
Citation: Journal of the American Association for Laboratory Animal Science 2025; 10.30802/AALAS-JAALAS-25-068
Outcome measures.
Mortality/cageside observations.
All animals were observed at least twice daily (morning and afternoon) from the time of arrival through study termination. Animals were observed cage side for morbidity, mortality, injury, and availability of food and water. If animals were noted to have abnormal clinical signs, they were submitted for veterinary consultation and examination by a member of the veterinary services staff. Any observations noted were recorded in Provanits 10 (Instem, West Conshohocken, PA).
Detailed clinical observations.
Detailed clinical observations were performed on the day after arrival and then once weekly throughout the study. Individual body weights were obtained on the day after animal arrival, on study day −3, and once weekly for the duration of the study. Animals were initially observed cage side for approximately 10 seconds and then removed from the cage for approximately 30 seconds per animal. Observations were performed by visualization and palpation and included evaluation of the skin, fur, eyes, ears, nose, oral cavity, thorax, abdomen, external genitalia, limbs and feet, respiratory and circulatory effects, autonomic effects such as salivation, and nervous system effects including tremors, convulsions, reactivity to handling (if animals attempted to escape being gently restrained within the handler’s hand or bite the handler), and unusual behaviors such as increased or decreased activity or hypersensitivity compared with conspecifics and animals’ own baseline observations. These were recorded in Provanits 10 (Instem, West Conshohocken, PA). All staff received training on observations through a formal site-wide training program, which included a glossary of available descriptive terms within the Provantis lexicon. Staff members performing detailed clinical observations regularly worked with a variety of mice within the vivarium and remained consistent throughout the study, and were therefore familiar with the predose presentation of the animals and were able to accurately describe behavioral changes from each animal’s baseline.
Nest scoring.
Approximately 7-9 hours following the start of the light cycle, nest scoring was conducted on the day following transfer to clean caging once during the prestudy period and then weekly throughout the dosing period. Nest scores were obtained by trained and blinded observers. Nest scoring was based on the article by Gaskill et al25 and is described in Table 2. Individuals performing the scoring were blinded to the drug treatment group by removing any external cage identifiers and physically relocating cages on the holding rack approximately 20 minutes in advance of scoring as cages were normally kept in dose group order. It was observed that by week 3 of the study, the nesting material that had been transferred over during each cage cleaning in the pair-housed animals had visibly declined in quality, losing its rigidity and potentially impacting nest quality. To counteract this, an additional 8 g of BedRNes nesting puck was offered to all pair-housed animals (for a total of 16 g of nesting material), incorporating it into the existing nesting material after nest scoring was performed on week 3. Cages were returned to their original location on the rack immediately following completion of scoring.
| Score | Description |
|---|---|
| 1 | No evidence of a nest |
| 2 | Flat nest with no walls |
| 3 | Nest slightly cup-shaped where the wall is less than one-half of the height of the dome |
| 4 | Nest is cup-shaped with a wall that is one-half of the height of the dome |
| 5 | Nest is cup-shaped with walls that are taller than one-half of the height of the dome; the walls may or may not fully enclose to make a dome |
TINT.
TINT was performed once during the prestudy period and then weekly throughout the dosing period, 3 days after transfer to clean caging within 3 hours of the start of the light cycle based on the method described in Rock et al.33 A 2.5 cm2 square of cotton nesting material (Cotton square; Ancare, Bellmore, NY) was placed in the front left corner of each cage. The TINT was scored as positive if, after 10-minute of placing the material in the cage, the material had been moved from its original location and/or showed evidence of having been actively interacted with (shredding). If the material had not been moved or interacted with, it was scored as negative. One observation noted at the time of performing the TINT was that some pair-housed mice did seem to become territorial with respect to the added nesting square. This territoriality was resolved when an additional nesting square was added after the initial TINT observation was completed so that each animal had access to their own nesting square. Individuals performing the scoring were blinded to the drug treatment group by removing any external cage identifiers and physically relocating cages on the holding rack approximately 20 minutes before the start of the test, with cages being returned to their original locations immediately following test conclusion.
Fecal corticosterone metabolites.
Fecal samples (6-8 fecal boli per animal) were collected from each cage pretest and following transfer to clean caging in weeks 1 and 4 for fecal corticosterone metabolite testing. Initially, samples were collected 3 hours postrotation; however, it was determined that inadequate volumes of fecal material were available at this time, so in weeks 1 and 4, samples were collected approximately 24 hours after cage rotation. For pair-housed animals, a random selection of fecal boli was alternately placed in each animal’s collection tube. Samples were frozen at −51 to −68 °C within 30 minutes of collection. Sample analysis was conducted using the Corticosterone ELISA kit (cat. no. ADI-900-097; ENZO Life Sciences, Farmingdale, NY) following the product manual instructions.
Clinical pathology.
Blood samples were collected via submental vein on study days 1 and 26 at 0.5 and 4 hours postdose to mimic TK sample collection. Collections were scheduled to be 32 μL per collection to allow for repeated collections in the same animal. Due to an error in protocol evaluation, the first collection on day 1 was inadvertently collected as 250 μL; the second collection that day was reduced to a single droplet (10 μL) to prevent excessive blood loss. As the TK profile of theophylline is already well established, these samples were not analyzed and were discarded postcollection.
Hematologic and clinical chemistry evaluations were conducted on blood collected via vena cava following euthanasia via CO2 inhalation. The order of bleeding was alternated such that 1 animal from each dose group was bled before returning to the first group to reduce handling and timing biases. The following hematologic values were assessed using Sysmex XN-1000V (Sysmex Corporation, Kobe, Japan): red blood cell count, hemoglobin concentration, hematocrit, mean corpuscular volumes (MCV), red blood cell distribution width, mean corpuscular hemoglobin concentration (MCHC), mean corpuscular hemoglobin (MCH), reticulocyte count, platelet count, white blood cell, neutrophil count, lymphocyte count, monocyte count, eosinophil count, basophil count, large unstained cell count, and other cell count; a blood smear was also preserved and stained for manual evaluation, if needed. Clinical chemistry values were assessed using Beckman Coulter AU5800 (Brea, CA) and included alkaline phosphatase, total bilirubin, aspartate aminotransferase, alanine aminotransferase, urea nitrogen, creatinine, total protein, albumin, globulin, albumin/globulin ratio, glucose, total cholesterol, triglycerides, sodium, potassium, chloride, calcium, and phosphorus.
Urinalysis was performed on samples collected via free-catch during approximately 16 hours of single housing in 251.8 cm2 × 12.7-cm wire-bottom cages containing perforated enrichment tunnels (Mariplast North America, Greenville, SC) before euthanasia. Single housing was pursued for this time point to mirror toxicological study designs that frequently separate pair-housed animals for final urine collections. During this time, one unit of DietGel 76A (ClearH2O, Westbrook, ME) and one unit of HydroGel (ClearH2O, Westbrook, ME) were provided to prevent contamination of urine with pelleted diet. Urinalysis was performed using Sysmex Clinitek Novus (Sysmex Corporation, Kobe, Japan) and included the following measurements: volume, specific gravity, and pH.
Terminal procedures.
All animals were euthanized using CO2 inhalation and confirmed by laceration of the diaphragm following terminal blood collection. Animals were euthanized in alternating order with one animal from each dose group euthanized before returning to the first group to reduce handling and timing biases. A gross necropsy, performed by a veterinarian trained in pathology, was only performed on the group 5 animal that was euthanized and replaced on day 1 to determine the most likely cause of death.
Pelt aggression lesion scores were performed on all animals that survived until the main study necropsy. All carcasses were inspected for signs of injury and disease. Pelts were removed from the animals by 2 veterinarians via midline incisions made from the pubis to the mandible, and then detaching the skin from the carcass. The pelt was then placed on a dissection surface with the subcutis facing upward so that photographs could be obtained. One veterinarian coded the obtained images so that a second blinded veterinarian could score the pelts in accordance with the paper by Gaskill et al,34 consistent with the methodology used for albino animals.10
Statistical methods.
All statistics were completed in SAS 9.4 with SAS/STAT 15.2 by the facility biostatistician and data analyst (third and fourth authors, respectively) using statistical methods routinely implemented in toxicological studies of similar design. The Shapiro-Wilk test for normality of the residuals and the Levene test for homogeneity of variances were conducted to determine if continuous variable data (body weight, body weight change, hematology, urinalysis, clinical chemistry, and fecal corticosterone metabolite) were normally distributed35,36; if either test was significant (P ≤ 0.01), the data were subjected to log transformation. If both tests were not significant on the log-transformed data, then a natural log (loge[x]) normal distribution was used; if either was significant, indicating a nonparametric distribution, then a rank transformation (PROC RANK) was used. An F statistic in a 2-way ANOVA was used to test each endpoint for the effects of treatment, housing (single compared with pair), and the interaction between treatment and housing.37 Linear contrasts were constructed for 2-tailed, pairwise comparisons of each treatment group within each housing type and between housing types within each treatment if the interaction effect was found to be significant (P < 0.05). Otherwise, linear contrasts were constructed among treatment groups overall if the main effect of treatment was significant (P < 0.05) and among housing types overall if the main effect of housing type was significant (P < 0.05). Results of all pairwise comparisons were adjusted for multiple comparisons using the methods of Edwards and Berry.38
Overall differences between response and treatment groups for clinical observations and nest scores were assessed using an exact Mantel-Haenszel for each housing type if there was sufficient variability among the groups.39 A Fisher exact test was performed for TINT results, where response values corresponded to the presence or absence of an observation, for each type of housing. If the results were significant (P < 0.05), follow-up pairwise comparisons were performed using the Fisher exact test. If the resulting P value was significant (P < 0.05), follow-up pairwise comparisons were performed if groups had sufficient sample size (n ≥ 3). Adjustments were made using a step-down Sidak method40 with significance reported at P ≤ 0.05.
Ethical review.
All treatment, toxicity assessments, and data acquisition were performed at Charles River Laboratories, a facility that is accredited by AAALAC, International. The study protocol was reviewed and approved by the IACUC prior to initiation.
Results
Detailed clinical observations.
Overall, pair-housed mice had significantly fewer animals with observations of “activity increased” than single-housed mice (χ2 = 16.07; P < 0.01). Per the Provantis glossary, “activity increased” is defined as “the animal’s activity level being increased compared with the normal state.” Single-housed high-dose theophylline mice had significantly more animals with observations of “activity increased” than pair-housed high-dose theophylline mice (χ2 = 19.5; P < 0.01) and than single-housed control mice (χ2 = 36.0; P < 0.01). The pair-housed high-dose theophylline mice also had significantly more animals with “activity increased” recorded than pair-housed control mice (χ2 = 6.5; P < 0.05). Similarly, pair-housed mice had significantly fewer animals with observations of “hypersensitive, moderate” than single-housed animals (χ2 = 10.7; P < 0.01). Per the Provantis lexicon, “hypersensitivity” is defined as: “the animal has an exaggerated reaction to skin touch, toe pinch, or tail pinch” with the modifier of “moderate” being defined as the animal being “moderately reactive to physical touch during study procedures and procedures may be impacted.” In this case, single-housed high-dose theophylline mice had a significantly higher number of animals with “hypersensitive, moderate” observations than single-housed control animals (χ2 = 8.0; P < 0.01), as well as pair-housed high-dose theophylline animals (χ2 = 5.5; P < 0.05). Pair-housed low-dose theophylline animals had significantly fewer “hypersensitive, moderate” animals than did single-housed low-dose theophylline animals (χ2 = 5.8; P < 0.05).
Pair-housed high-dose theophylline mice had significantly more animals with thin fur cover over the muzzle than pair-housed control animals (χ2 = 5.4; P < 0.05). The Provantis glossary definition of thin fur cover refers to fur cover being reduced or thinner than normal in a specific area.
Body weight and body weight change.
Mean body weights and SD for the various study time points are presented in Table 3. Single-housed low-dose theophylline mice were significantly heavier than single-housed control mice prestudy (t = 3.36; P < 0.05); however, this difference resolved once dosing started. Pair-housed control animals were significantly heavier than single-housed control animals prestudy (t = 3.53; P < 0.01), and remained significantly heavier through study weeks 1 (t = 3.09; P < 0.05), 2 (t = 2.96; P < 0.05), and 4 (P < 0.01). During week 3, pair-housed animals were significantly heavier than single-housed animals overall (t = 2.20; P < 0.05); however, this effect did not transition to significant changes between specific dose levels between housing conditions.
| Study period/social housing status | Dose group | Overall | ||
|---|---|---|---|---|
| 0.9% NaCl (0 mg/kg/d) | Theophylline (50 mg/kg/d) | Theophylline (200 mg/kg/d) | ||
| Predose | ||||
| Single | ||||
| Mean | 31.51 | 34.98 b | 33.86 | 33.45 |
| SD | 1.95 | 1.68 | 1.32 | 2.17 |
| n a | 8 | 8 | 8 | 24 |
| Paired | ||||
| Mean | 35.16 d | 34.38 | 32.91 | 34.16 |
| SD | 1.01 | 2.70 | 2.85 | 2.45 |
| n | 8 | 10 | 8 | 26 |
| Overall | ||||
| Mean | 33.33 | 34.65 | 33.38 | |
| SD | 2.41 | 2.26 | 2.20 | |
| n | 16 | 18 | 16 | |
| Week 1 | ||||
| Single | ||||
| Mean | 33.12 | 35.50 | 35.28 | 34.63 |
| SD | 2.43 | 2.31 | 1.04 | 2.23 |
| n | 8 | 8 | 8 | 24 |
| Paired | ||||
| Mean | 36.66 c | 36.80 | 33.96 | 35.80 |
| SD | 1.29 | 3.10 | 2.80 | 2.75 |
| n | 8 | 8 | 8 | 24 |
| Overall | ||||
| Mean | 34.89 | 36.15 | 34.62 | |
| SD | 2.62 | 2.72 | 2.15 | |
| n | 16 | 16 | 16 | |
| Week 2 | ||||
| Single | ||||
| Mean | 33.91 | 37.05 | 36.60 | 35.85 |
| SD | 2.69 | 2.29 | 1.11 | 2.48 |
| n | 8 | 8 | 8 | 24 |
| Paired | ||||
| Mean | 37.52 c | 38.25 | 35.52 | 37.10 |
| SD | 1.23 | 3.40 | 2.97 | 2.83 |
| n | 8 | 8 | 8 | 24 |
| Overall | ||||
| Mean | 35.71 | 37.65 | 36.06 | |
| SD | 2.75 | 2.87 | 2.23 | |
| n | 16 | 16 | 16 | |
| Week 3 | ||||
| Single | ||||
| Mean | 34.92 | 37.81 | 37.41 | 36.68 |
| SD | 2.95 | 2.26 | 1.72 | 2.64 |
| n | 8 | 8 | 7 | 23 |
| Paired | ||||
| Mean | 38.67 | 39.76 | 36.80 | 38.41 c |
| SD | 1.56 | 3.57 | 2.99 | 2.98 |
| n | 8 | 8 | 8 | 24 |
| Overall | ||||
| Mean | 36.80 | 38.78 | 37.08 | |
| SD | 2.99 | 3.06 | 2.42 | |
| n | 16 | 16 | 15 | |
| Week 4 | ||||
| Single | ||||
| Mean | 35.52 | 38.48 | 37.90 | 37.27 |
| SD | 2.90 | 2.31 | 1.43 | 2.59 |
| n | 8 | 8 | 7 | 23 |
| Paired | ||||
| Mean | 39.91 d | 40.47 | 37.16 | 39.18 |
| SD | 1.20 | 3.94 | 2.43 | 3.02 |
| n | 8 | 8 | 8 | 24 |
| Overall | ||||
| Mean | 37.71 | 39.48 | 37.50 | |
| SD | 3.12 | 3.28 | 1.99 | |
| n | 16 | 16 | 15 | |
Sample size.
Significantly different from 0.9% NaCl (P < 0.05).
Significantly different from single-housed animals (P < 0.05).
Significantly different from single-housed animals (P < 0.01).
The change in body weight from prestudy to week 1 of study was significantly greater in pair-housed low-dose theophylline mice than in single-housed low-dose theophylline mice (t = 3.34; P < 0.05). Between weeks 1 and 2, low and high-dose theophylline animals gained significantly more weight than control animals overall (t = 2.74 and t = 2.49, respectively; P < 0.05). Between weeks 2 and 3, pair-housed mice gained significantly more weight than single-housed mice overall (t = 2.18, P < 0.05).
Nest scoring.
Mean nest scores across the study are pictured in Figure 2. In study week 1, nest scores of pair-housed low- and high-dose theophylline animals were significantly lower than single-housed animals at the same dose levels (t = 7.50 and t = 7.35, respectively; P < 0.01) and significantly lower than pair-housed control animals (t = 15.20 and t = 12.10, respectively; P < 0.01). In weeks 3 and 4, both pair-housed low-dose (t = 6.75 and t = 10.0, respectively; P < 0.05) and high-dose (t = 8.45 and t = 5.87, respectively; P < 0.05) theophylline animals had significantly lower nest scores when compared with pair-housed control animals. Pair-housed high-dose theophylline animals also had significantly lower nest scores compared with single-housed high-dose theophylline animals in week 3 (t = 5.82; P < 0.05). In addition, single-housed high-dose theophylline animals had significantly lower scores than single-housed control animals in weeks 2 and 4 (t = 7.14 and t = 5.52, respectively; P < 0.05), with single-housed low-dose theophylline animals also having significantly lower scores than single-housed control animals in week 4 (t = 6.51; P < 0.05).
Citation: Journal of the American Association for Laboratory Animal Science 2025; 10.30802/AALAS-JAALAS-25-068
TINT.
The number of positive TINT scores observed throughout the study is pictured in Figure 3. Overall, pair-housed animals were noted with significantly fewer animals with positive TINT scores than single-housed animals prestudy (χ2 = 9.55; P < 0.01). Pair-housed high-dose theophylline mice had significantly fewer animals with positive TINT scores prestudy compared with single-housed animals of the same dose level (χ2 = 9.6, P < 0.01). This observation was reversed in week 4 of the study, where significantly more pair-housed high-dose theophylline mice had positive TINT scores than single-housed high-dose theophylline mice (χ2 = 6.23; P < 0.05).
Citation: Journal of the American Association for Laboratory Animal Science 2025; 10.30802/AALAS-JAALAS-25-068
Fecal corticosterone metabolites.
Mean fecal corticosterone metabolite values and SD for the various study time-points are presented in Table 4. During week 4, pair-housed animals had significantly higher fecal corticosterone metabolite values than single-housed animals (t = 2.62; P < 0.05), when not accounting for dose group.
| Study period/social housing status | Dose group | Overall | ||
|---|---|---|---|---|
| 0.9% NaCl (0 mg/kg/d) | Theophylline (50 mg/kg/d) | Theophylline (200 mg/kg/d) | ||
| Prestudy | ||||
| Single | ||||
| Mean | 258.6 | 199.3 | 229.6 | 229.2 |
| SD | 65.44 | 57.94 | 58.83 | 63.18 |
| n a | 8 | 8 | 8 | 24 |
| Paired | ||||
| Mean | 216.1 | 218.8 | 231.0 | 222.0 |
| SD | 57.72 | 44.44 | 52.36 | 49.93 |
| n | 8 | 8 | 8 | 24 |
| Overall | ||||
| Mean | 237.4 | 209.0 | 230.3 | |
| SD | 63.52 | 50.89 | 53.81 | |
| n | 16 | 16 | 16 | |
| Week 1 | ||||
| Single | ||||
| Mean | 202.8 | 285.3 | 366.3 | 284.8 |
| SD | 56.89 | 97.56 | 131.36 | 117.40 |
| n | 8 | 8 | 8 | 24 |
| Paired | ||||
| Mean | 262.8 | 255.0 | 303.5 | 273.8 |
| SD | 46.11 | 92.20 | 34.23 | 63.74 |
| n | 8 | 8 | 8 | 24 |
| Overall | ||||
| Mean | 232.8 | 270.1 | 334.9 b | |
| SD | 58.84 | 93.02 | 98.23 | |
| n | 16 | 16 | 16 | |
| Week 4 | ||||
| Single | ||||
| Mean | 202.1 | 237.1 | 251.3 | 229.3 |
| SD | 49.46 | 68.04 | 138.50 | 89.04 |
| n | 8 | 8 | 7 | 23 |
| Paired | ||||
| Mean | 240.9 | 237.4 | 380.9 | 286.4 c |
| SD | 40.48 | 63.30 | 147.02 | 113.83 |
| n | 8 | 8 | 8 | 24 |
| Overall | ||||
| Mean | 221.5 | 237.3 | 320.4 | |
| SD | 48.03 | 63.48 | 153.32 | |
| n | 16 | 16 | 15 | |
Sample size.
Significantly different from 0.9% NaCl (P < 0.01).
Significantly different from single-housed animals (P < 0.05).
Clinical pathology.
Hematology.
Samples from 2 single-housed low-dose theophylline, one group single-housed high-dose theophylline, and 2 pair-housed low-dose theophylline mice were omitted from analysis due to clotting. Mean hematologic values and standard deviations are presented in Table 5. Overall, white blood cell counts were significantly lower in pair-housed low and high-dose theophylline animals compared with pair-housed controls (t = −2.82 and t = −3.53, respectively; P < 0.05). Pair-housed low-dose theophylline animals also had significantly lower white blood cell counts than single-housed low-dose theophylline animals (t = −3.02; P < 0.05). Both the percentage and absolute count of lymphocytes were significantly lower in mice receiving the low dose of theophylline (t = −2.72 and t = −2.68, respectively; P < 0.05) than control mice; absolute lymphocyte counts were also significantly lower in high-dose theophylline mice (t = −5.81; P < 0.01). The percentage of reticulocytes and the absolute value of reticulocytes were significantly lower in low-dose theophylline animals compared with control animals (t = −2.79 and t = −2.90, respectively; P < 0.05). High-dose theophylline mice had significantly higher mean corpuscular hemoglobin (t = 2.83; P < 0.05) and mean corpuscular volume (t = 4.31; P < 0.05) than control mice. Pair-housed mice had significantly higher platelet counts than single-housed mice (t = 3.24; P < 0.01).
| Study period/social housing status | Dose group | Overall | ||
|---|---|---|---|---|
| 0.9% NaCl (0 mg/kg/d) | Theophylline (50 mg/kg/d) | Theophylline (200 mg/kg/d) | ||
| Basophils, 103/µL | ||||
| Single | ||||
| Mean | 0.01 | 0.00 | 0.00 | 0.01 |
| SD | 0.01 | 0.00 | 0.00 | 0.00 |
| n a | 8 | 6 | 6 | 20 |
| Paired | ||||
| Mean | 0.00 | 0.01 | 0.01 | 0.01 |
| SD | 0.00 | 0.00 | 0.00 | 0.00 |
| n | 8 | 6 | 8 | 22 |
| Overall | ||||
| Mean | 0.01 | 0.00 | 0.01 | |
| SD | 0.00 | 0.00 | 0.00 | |
| n | 16 | 12 | 14 | |
| Eosinophils, 103/µL | ||||
| Single | ||||
| Mean | 0.12 | 0.13 | 0.06 | 0.11 |
| SD | 0.11 | 0.12 | 0.06 | 0.10 |
| n | 8 | 6 | 6 | 20 |
| Paired | ||||
| Mean | 0.11 | 0.08 | 0.05 | 0.08 |
| SD | 0.10 | 0.06 | 0.03 | 0.07 |
| n | 8 | 6 | 8 | 22 |
| Overall | ||||
| Mean | 0.12 | 0.10 | 0.05 | |
| SD | 0.10 | 0.09 | 0.05 | |
| n | 16 | 12 | 14 | |
| Lymphocytes, 103/µL | ||||
| Single | ||||
| Mean | 4.36 | 3.81 | 2.81 | 3.73 |
| SD | 1.73 | 0.87 | 0.91 | 1.39 |
| n | 8 | 6 | 6 | 20 |
| Paired | ||||
| Mean | 4.08 | 2.95 | 3.05 | 3.40 |
| SD | 0.76 | 0.59 | 0.71 | 0.85 |
| n | 8 | 6 | 8 | 22 |
| Overall | ||||
| Mean | 4.22 | 3.38 b | 2.95 c | |
| SD | 1.30 | 0.84 | 0.78 | |
| n | 16 | 12 | 14 | |
| Monocytes, 103/µL | ||||
| Single | ||||
| Mean | 0.08 | 0.08 | 0.07 | 0.07 |
| SD | 0.08 | 0.03 | 0.05 | 0.06 |
| n | 8 | 6 | 6 | 20 |
| Paired | ||||
| Mean | 0.09 | 0.11 | 0.07 | 0.09 |
| SD | 0.04 | 0.05 | 0.06 | 0.05 |
| n | 8 | 6 | 8 | 22 |
| Overall | ||||
| Mean | 0.08 | 0.09 | 0.07 | |
| SD | 0.06 | 0.04 | 0.06 | |
| n | 16 | 12 | 14 | |
| Neutrophils, 103/µL | ||||
| Single | ||||
| Mean | 0.46 | 0.93 | 0.44 | 0.60 |
| SD | 0.22 | 1.10 | 0.16 | 0.63 |
| n | 8 | 6 | 8 | 20 |
| Paired | ||||
| Mean | 0.68 | 0.56 | 0.49 | 0.58 |
| SD | 0.50 | 0.13 | 0.15 | 0.32 |
| n | 8 | 6 | 8 | 22 |
| Overall | ||||
| Mean | 0.57 | 0.75 | 0.47 | |
| SD | 0.39 | 0.77 | 0.15 | |
| n | 16 | 12 | 14 | |
| Hemoglobin, g/dL | ||||
| Single | ||||
| Mean | 17.11 | 17.58 | 17.81 | 17.46 |
| SD | 1.56 | 1.19 | 1.59 | 1.43 |
| n | 8 | 6 | 6 | 20 |
| Paired | ||||
| Mean | 17.16 | 17.93 | 16.18 | 17.01 |
| SD | 1.42 | 1.54 | 2.01 | 1.76 |
| n | 8 | 6 | 8 | 22 |
| Overall | ||||
| Mean | 17.13 | 17.75 | 16.88 | |
| SD | 1.44 | 1.33 | 1.96 | |
| n | 16 | 12 | 14 | |
| Mean corpuscular hemoglobin concentration, g/dL | ||||
| Single | ||||
| Mean | 29.42 | 29.08 | 29.03 | 29.20 |
| SD | 0.44 | 0.50 | 0.80 | 0.58 |
| n | 8 | 6 | 6 | 20 |
| Paired | ||||
| Mean | 29.28 | 29.58 | 29.16 | 29.32 |
| SD | 0.86 | 0.57 | 0.50 | 0.66 |
| n | 8 | 6 | 8 | 22 |
| Overall | ||||
| Mean | 29.35 | 29.33 | 29.10 | |
| SD | 0.66 | 0.57 | 0.62 | |
| n | 16 | 12 | 14 | |
| Mean corpuscular hemoglobin, pg | ||||
| Single | ||||
| Mean | 15.46 | 15.58 | 15.90 | 15.63 |
| SD | 0.57 | 0.35 | 0.34 | 0.46 |
| n | 8 | 6 | 6 | 20 |
| Paired | ||||
| Mean | 15.45 | 15.73 | 15.97 | 15.71 |
| SD | 0.37 | 0.54 | 0.53 | 0.51 |
| n | 8 | 6 | 8 | 22 |
| Overall | ||||
| Mean | 15.45 | 15.65 | 15.94 b | |
| SD | 0.46 | 0.44 | 0.44 | |
| n | 16 | 12 | 14 | |
| Mean corpuscular volume, fL | ||||
| Single | ||||
| Mean | 52.61 | 53.60 | 54.78 | 53.56 |
| SD | 1.63 | 0.74 | 1.38 | 1.57 |
| n | 8 | 6 | 6 | 20 |
| Paired | ||||
| Mean | 52.76 | 53.20 | 54.72 | 53.59 |
| SD | 1.29 | 1.57 | 1.33 | 1.59 |
| n | 8 | 6 | 8 | 22 |
| Overall | ||||
| Mean | 52.68 | 53.40 | 54.75 b | |
| SD | 1.42 | 1.19 | 1.30 | |
| n | 16 | 12 | 14 | |
| Hematocrit, % | ||||
| Single | ||||
| Mean | 58.18 | 60.50 | 61.30 | 59.81 |
| SD | 5.27 | 4.15 | 4.01 | 4.58 |
| n | 8 | 6 | 6 | 20 |
| Paired | ||||
| Mean | 58.55 | 60.60 | 55.45 | 57.98 |
| SD | 4.17 | 4.44 | 6.48 | 5.38 |
| n | 8 | 6 | 8 | 22 |
| Overall | ||||
| Mean | 58.36 | 60.55 | 57.95 | |
| SD | 4.59 | 4.09 | 6.15 | |
| n | 16 | 12 | 14 | |
| Platelet count, 103/µL | ||||
| Single | ||||
| Mean | 1,022.40 | 1,031.80 | 914.33 | 984.81 |
| SD | 411.62 | 335.75 | 271.08 | 320.83 |
| n | 5 | 5 | 6 | 16 |
| Paired | ||||
| Mean | 1,471.14 | 1,262.40 | 1,493.37 | 1,427.85 e |
| SD | 301.40 | 496.61 | 473.74 | 415.93 |
| n | 7 | 5 | 8 | 20 |
| Overall | ||||
| Mean | 1,284.16 | 1,147.10 | 1,245.21 | |
| SD | 405.65 | 417.71 | 487.38 | |
| n | 12 | 10 | 14 | |
| Red blood cell count, 106/µL | ||||
| Single | ||||
| Mean | 11.07 | 11.29 | 11.20 | 11.18 |
| SD | 1.13 | 0.85 | 0.92 | 0.94 |
| n | 8 | 6 | 8 | 20 |
| Paired | ||||
| Mean | 11.10 | 11.40 | 10.14 | 10.83 |
| SD | 0.84 | 0.87 | 1.26 | 1.12 |
| n | 8 | 6 | 8 | 22 |
| Overall | ||||
| Mean | 11.09 | 11.34 | 10.59 | |
| SD | 0.96 | 0.82 | 1.22 | |
| n | 16 | 12 | 14 | |
| Red blood cell distribution width, % | ||||
| Single | ||||
| Mean | 17.61 | 18.26 | 19.30 | 18.31 |
| SD | 1.54 | 0.67 | 0.81 | 1.30 |
| n | 8 | 6 | 6 | 20 |
| Paired | ||||
| Mean | 18.16 | 18.31 | 18.37 | 18.28 |
| SD | 0.75 | 1.03 | 1.05 | 0.90 |
| n | 8 | 6 | 8 | 22 |
| Overall | ||||
| Mean | 17.88 | 18.29 | 18.77 | |
| SD | 1.21 | 0.83 | 1.04 | |
| n | 16 | 12 | 14 | |
| Reticulocytes, 109/L | ||||
| Single | ||||
| Mean | 593.05 | 476.48 | 567.90 | 550.53 |
| SD | 199.49 | 127.31 | 196.80 | 178.07 |
| n | 8 | 6 | 6 | 20 |
| Paired | ||||
| Mean | 564.40 | 533.58 | 631.57 | 580.42 |
| SD | 103.43 | 128.08 | 155.07 | 131.18 |
| n | 8 | 6 | 8 | 22 |
| Overall | ||||
| Mean | 578.72 | 505.03 b | 604.28 | |
| SD | 154.22 | 125.35 | 170.04 | |
| n | 16 | 12 | 14 | |
| White blood cell count, 103/µL | ||||
| Single | ||||
| Mean | 5.04 | 4.97 | 3.40 c | 4.53 |
| SD | 1.86 | 1.70 | 0.93 | 1.68 |
| n | 8 | 6 | 6 | 20 |
| Paired | ||||
| Mean | 4.98 | 3.72 b , d | 3.68 c | 4.16 |
| SD | 1.04 | 0.70 | 0.85 | 1.05 |
| n | 8 | 6 | 8 | 22 |
| Overall | ||||
| Mean | 5.01 | 4.34 | 3.56 | |
| SD | 1.45 | 1.40 | 0.86 | |
| n | 16 | 12 | 14 | |
Sample size.
Significantly different from 0.9% NaCl (P < 0.05).
Significantly different from 0.9% NaCl (P < 0.01).
Significantly different from single-housed animals (P < 0.05).
Significantly different from single-housed animals (P < 0.01).
Chemistry.
Samples from 2 single-housed control, 2 single-housed low-dose theophylline, 2 single-housed high-dose theophylline, and one pair-housed low-dose theophylline mice were omitted from analysis due to hemolysis. Samples from 7 single-housed control, 7 single-housed low-dose theophylline, 6 single-housed low-dose theophylline, 6 pair-housed control, 8 pair-housed low-dose theophylline, and 5 pair-housed high-dose theophylline mice had inadequate sample volume for all parameters to be run. Mean biochemistry values and SD are presented in Table 6. The albumin:globulin ratio was significantly higher in high-dose theophylline animals compared with vehicle animals (t = 3.23; P < 0.01), although no significant differences were noted in total albumin or globulin levels.
| Study period/social housing status | Dose group | Overall | ||
|---|---|---|---|---|
| 0.9% NaCl (0 mg/kg/d) | Theophylline (50 mg/kg/d) | Theophylline (200 mg/kg/d) | ||
| Alanine aminotransferase, U/L | ||||
| Single | ||||
| Mean | 79.71 | 23.33 | 39.83 | 49.31 |
| SD | 110.18 | 17.52 | 24.26 | 70.07 |
| n a | 7 | 6 | 6 | 19 |
| Paired | ||||
| Mean | 27.50 | 43.28 | 32.00 | 33.87 |
| SD | 5.87 | 44.92 | 34.59 | 31.40 |
| n | 8 | 7 | 8 | 23 |
| Overall | ||||
| Mean | 51.86 | 34.07 | 35.35 | |
| SD | 77.11 | 35.27 | 29.78 | |
| n | 15 | 13 | 14 | |
| Albumin, g/dL | ||||
| Single | ||||
| Mean | 3.60 | 3.37 | 3.80 | 3.60 |
| SD | 0.30 | 0.18 | 0.28 | 0.30 |
| n | 3 | 4 | 5 | 12 |
| Paired | ||||
| Mean | 3.36 | 3.52 | 3.58 | 3.49 |
| SD | 0.25 | 0.32 | 0.13 | 0.23 |
| n | 5 | 4 | 6 | 15 |
| Overall | ||||
| Mean | 3.45 | 3.45 | 3.68 | |
| SD | 0.27 | 0.25 | 0.23 | |
| n | 8 | 8 | 8 | |
| Albumin/globulin ratio, ratio | ||||
| Single | ||||
| Mean | 1.16 | 1.07 | 1.28 | 1.18 |
| SD | 0.05 | 0.09 | 0.08 | 0.11 |
| n | 3 | 4 | 5 | 12 |
| Paired | ||||
| Mean | 1.12 | 1.12 | 1.26 | 1.18 |
| SD | 0.08 | 0.09 | 0.08 | 0.10 |
| n | 5 | 4 | 6 | 15 |
| Overall | ||||
| Mean | 1.13 | 1.10 | 1.27 c | |
| SD | 0.07 | 0.09 | 0.07 | |
| n | 8 | 8 | 11 | |
| Alkaline phosphatase, U/L | ||||
| Single | ||||
| Mean | 62.37 | 57.85 | 62.66 | 60.95 |
| SD | 15.69 | 29.89 | 16.60 | 20.69 |
| n | 8 | 7 | 6 | 21 |
| Paired | ||||
| Mean | 64.12 | 63.85 | 55.75 | 61.13 |
| SD | 18.90 | 14.53 | 13.23 | 15.59 |
| n | 8 | 7 | 8 | 23 |
| Overall | ||||
| Mean | 63.25 | 60.85 | 58.71 | |
| SD | 16.81 | 22.79 | 14.59 | |
| n | 16 | 14 | 14 | |
| Aspartate aminotransferase, U/L | ||||
| Single | ||||
| Mean | 120.28 | 79.66 | 141.66 | 114.21 |
| SD | 65.71 | 19.17 | 73.24 | 60.78 |
| n | 7 | 6 | 6 | 19 |
| Paired | ||||
| Mean | 112.87 | 96.85 | 91.75 | 100.65 |
| SD | 53.24 | 36.40 | 66.25 | 52.42 |
| n | 8 | 7 | 8 | 23 |
| Overall | ||||
| Mean | 116.33 | 88.92 | 113.14 | |
| SD | 57.29 | 29.92 | 71.30 | |
| n | 15 | 13 | 14 | |
| Calcium, mg/dL | ||||
| Single | ||||
| Mean | b | 11.30 | 11.60 | 11.45 |
| SD | b | b | b | 0.21 |
| n | 0 | 1 | 1 | 2 |
| Paired | ||||
| Mean | 11.95 | b | 11.10 | 11.44 |
| SD | 0.77 | b | 0.43 | 0.68 |
| n | 2 | 0 | 3 | 5 |
| Overall | ||||
| Mean | 11.95 | 11.30 | 11.22 | |
| SD | 0.77 | b | 0.43 | |
| n | 2 | 1 | 4 | |
| Chloride, mEq/L | ||||
| Single | ||||
| Mean | b | 123.50 | 123.50 | 123.50 |
| SD | b | 2.12 | 2.12 | 1.73 |
| n | 0 | 2 | 2 | 4 |
| Paired | ||||
| Mean | 120.50 | 122.00 | 124.20 | 122.88 |
| SD | 2.12 | 1.41 | 3.96 | 3.37 |
| n | 2 | 2 | 5 | 9 |
| Overall | ||||
| Mean | 120.50 | 122.75 | 124.00 | |
| SD | 2.12 | 1.70 | 3.36 | |
| n | 2 | 4 | 7 | |
| Cholesterol, mg/dL | ||||
| Single | ||||
| Mean | 189.00 | 250.00 | 283.00 | 261.55 |
| SD | b | 15.52 | 56.68 | 51.59 |
| n | 1 | 3 | 5 | 9 |
| Paired | ||||
| Mean | 278.75 | 250.00 | 254.50 | 260.14 |
| SD | 81.34 | 55.63 | 55.94 | 59.98 |
| n | 4 | 4 | 6 | 14 |
| Overall | ||||
| Mean | 260.80 | 250.00 | 267.45 | |
| SD | 81.07 | 40.34 | 55.42 | |
| n | 5 | 7 | 11 | |
| Creatinine, mg/dL | ||||
| Single | ||||
| Mean | 0.20 | 0.22 | 0.23 | 0.22 |
| SD | 0.00 | 0.04 | 0.05 | 0.04 |
| n | 4 | 5 | 6 | 15 |
| Paired | ||||
| Mean | 0.20 | 0.22 | 0.21 | 0.21 |
| SD | 0.00 | 0.04 | 0.04 | 0.03 |
| n | 6 | 5 | 6 | 17 |
| Overall | ||||
| Mean | 0.20 | 0.22 | 0.22 | |
| SD | 0.00 | 0.04 | 0.04 | |
| n | 10 | 10 | 12 | |
| Globulin, g/dL | ||||
| Single | ||||
| Mean | 3.03 | 3.17 | 2.96 | 3.05 |
| SD | 0.28 | 0.26 | 0.25 | 0.25 |
| n | 3 | 4 | 5 | 12 |
| Paired | ||||
| Mean | 2.98 | 3.12 | 2.86 | 2.97 |
| SD | 0.25 | 0.27 | 0.31 | 0.28 |
| n | 5 | 4 | 6 | 15 |
| Overall | ||||
| Mean | 3.00 | 3.15 | 2.90 | |
| SD | 0.25 | 0.25 | 0.27 | |
| n | 8 | 8 | 11 | |
| Glucose, mg/dL | ||||
| Single | ||||
| Mean | 210.50 | 290.00 | 298.00 | 279.18 |
| SD | 20.50 | 30.99 | 96.22 | 72.11 |
| n | 2 | 4 | 5 | 11 |
| Paired | ||||
| Mean | 216.00 | 229.75 | 297.16 | 254.71 |
| SD | 70.06 | 30.68 | 70.84 | 69.03 |
| n | 4 | 4 | 6 | 14 |
| Overall | ||||
| Mean | 214.16 | 259.87 | 297.54 | |
| SD | 55.11 | 43.03 | 78.82 | |
| n | 6 | 8 | 11 | |
| Phosphorus, mg/dL | ||||
| Single | ||||
| Mean | b | 14.20 | 15.70 | 14.95 |
| SD | b | b | b | 1.06 |
| n | 0 | 1 | 1 | 2 |
| Paired | ||||
| Mean | 17.00 | b | 14.96 | 15.78 |
| SD | 1.41 | b | 0.05 | 1.31 |
| n | 2 | 0 | 3 | 5 |
| Overall | ||||
| Mean | 17.00 | 14.20 | 15.15 | |
| SD | 1.41 | b | 0.36 | |
| n | 2 | 1 | 4 | |
| Potassium, mEq/L | ||||
| Single | ||||
| Mean | b | 11.15 | 10.10 | 10.62 |
| SD | b | 1.48 | 0.00 | 1.05 |
| n | 0 | 2 | 2 | 4 |
| Paired | ||||
| Mean | 10.10 | 10.10 | 10.44 | 10.28 |
| SD | 0.00 | 0.00 | 0.76 | 0.56 |
| n | 2 | 2 | 5 | 9 |
| Overall | ||||
| Mean | 10.10 | 10.62 | 10.34 | |
| SD | 0.00 | 1.05 | 0.64 | |
| n | 2 | 4 | 7 | |
| Sodium, mEq/L | ||||
| Single | ||||
| Mean | b | 159.00 | 159.00 | 159.00 |
| SD | b | 1.41 | 1.41 | 1.15 |
| n | 0 | 2 | 2 | 4 |
| Paired | ||||
| Mean | 163.00 | 160.50 | 162.80 | 162.33 |
| SD | 4.24 | 2.12 | 1.78 | 2.34 |
| n | 2 | 2 | 5 | 9 |
| Overall | ||||
| Mean | 163.00 | 159.75 | 161.71 | |
| SD | 4.24 | 1.70 | 2.43 | |
| n | 2 | 4 | 7 | |
| Total bilirubin, mg/dL | ||||
| Single | ||||
| Mean | 0.31 | 0.33 | 0.37 | 0.34 |
| SD | 0.12 | 0.16 | 0.14 | 0.14 |
| n | 8 | 7 | 6 | 21 |
| Paired | ||||
| Mean | 0.27 | 0.33 | 0.29 | 0.30 |
| SD | 0.09 | 0.15 | 0.11 | 0.11 |
| n | 8 | 7 | 8 | 23 |
| Overall | ||||
| Mean | 0.29 | 0.33 | 0.33 | |
| SD | 0.11 | 0.15 | 0.12 | |
| n | 16 | 14 | 14 | |
| Total protein, g/dL | ||||
| Single | ||||
| Mean | 6.63 | 6.56 | 6.76 | 6.66 |
| SD | 0.56 | 0.30 | 0.43 | 0.39 |
| n | 3 | 5 | 6 | 14 |
| Paired | ||||
| Mean | 6.34 | 6.65 | 6.45 | 6.46 |
| SD | 0.47 | 0.50 | 0.44 | 0.45 |
| n | 5 | 4 | 6 | 15 |
| Overall | ||||
| Mean | 6.45 | 6.60 | 6.60 | |
| SD | 0.49 | 0.38 | 0.45 | |
| n | 8 | 9 | 12 | |
| Triglycerides, mg/dL | ||||
| Single | ||||
| Mean | 108.00 | 115.66 | 124.00 | 119.44 |
| SD | b | 31.34 | 17.13 | 20.66 |
| n | 1 | 3 | 5 | 9 |
| Paired | ||||
| Mean | 200.25 | 145.66 | 121.66 | 151.38 |
| SD | 110.50 | 49.80 | 52.21 | 76.47 |
| n | 4 | 3 | 6 | 13 |
| Overall | ||||
| Mean | 181.80 | 130.66 | 122.72 | |
| SD | 104.21 | 40.68 | 38.49 | |
| n | 5 | 6 | 11 | |
Sample size.
Unable to calculate due to insufficient sample numbers.
Significantly different from 0.9% NaCl (P < 0.01).
Urinalysis.
Samples from 3 single-housed control, 2 single-housed low-dose theophylline, one single-housed high-dose theophylline, 2 pair-housed control, and 4 pair-housed high-dose theophylline mice had inadequate sample volume for all parameters to be run. Samples from one pair-housed control and one pair-housed low-dose theophylline animal were contaminated by feces. Mean urinalysis values and SD are presented in Table 7. No significant differences were noted between any groups for any urinalysis parameters.
| Study period/social housing status | Dose group | Overall | ||
|---|---|---|---|---|
| 0.9% NaCl (0 mg/kg/d) | Theophylline (50 mg/kg/d) | Theophylline (200 mg/kg/d) | ||
| Urine volume, mL | ||||
| Single | ||||
| Mean | 1.50 | 2.36 | 2.77 | 2.20 |
| SD | 1.16 | 1.04 | 1.87 | 1.45 |
| n a | 7 | 6 | 7 | 20 |
| Paired | ||||
| Mean | 1.68 | 1.38 | 5.25 | 2.71 |
| SD | 0.81 | 0.76 | 6.01 | 3.72 |
| n | 7 | 8 | 7 | 22 |
| Overall | ||||
| Mean | 1.59 | 1.80 | 4.01 | |
| SD | 0.96 | 0.99 | 4.47 | |
| n | 14 | 14 | 14 | |
| Urine specific gravity | ||||
| Single | ||||
| Mean | 1.04 | 1.03 | 1.04 | 1.03 |
| SD | 0.01 | 0.01 | 0.00 | 0.01 |
| n | 7 | 6 | 7 | 20 |
| Paired | ||||
| Mean | 1.03 | 1.03 | 1.02 | 1.03 |
| SD | 0.00 | 0.01 | 0.01 | 0.01 |
| n | 7 | 8 | 7 | 22 |
| Overall | ||||
| Mean | 1.03 | 1.03 | 1.03 | |
| SD | 0.00 | 0.01 | 0.01 | |
| n | 14 | 14 | 14 | |
| Urine pH | ||||
| Single | ||||
| Mean | 7.20 | 6.75 | 6.83 | 6.91 |
| SD | 0.57 | 0.41 | 1.08 | 0.73 |
| n | 5 | 6 | 6 | 17 |
| Paired | ||||
| Mean | 7.00 | 6.92 | 7.87 | 7.18 |
| SD | 0.61 | 0.73 | 0.47 | 0.72 |
| n | 5 | 7 | 4 | 16 |
| Overall | ||||
| Mean | 7.10 | 6.84 | 7.25 | |
| SD | 0.56 | 0.59 | 1.00 | |
| n | 10 | 13 | 10 | |
Sample size.
Pelt aggression lesion scores.
No significant differences were present between any groups for average pelt aggression lesion scores. Only one animal had a score greater than 0; this was a group 5 animal with a score of 0.11.
Discussion
For the duration of the 28-day study, all animals placed on study remained successfully pair housed from study day 1 until study completion regardless of required study functions and dosage of theophylline. The majority of pairings were deemed socially compatible before study start using recommended strategies to facilitate social compatibility, including pairing at 5 weeks of age, implementing low stress handling, removing high value structures during initial acclimation, and moving nesting over during cage changes. A single pair of mice was replaced before the study start due to social incompatibility during the acclimation period, which demonstrates the importance of providing adequate habituation time before the study as well as frequently monitoring and properly identifying incompatible groups before study start dates. By administering varying doses of theophylline, we were able to assess the impact of test articles with known clinical effects on social compatibility in male mice.
Despite animals in the high-dose group experiencing notable theophylline-related side effects, these side effects appeared to not affect social compatibility, as all social pairs remained compatible from study start to study end.
Throughout the study, significantly more animals within the high-dose groups were recorded with “activity increased” observations. These results are consistent with reported clinical observations associated with high doses of theophylline, where neurologic effects up to and including death have been observed.41–44 However, it was noted that the pair-housed high-dose theophylline animals had fewer observations of “activity increased” than the single-housed high-dose mice. This may be due to social buffering, which suggests that social animals recover more effectively from stress when socially housed17; unfortunately, this has been minimally studied in mice, but extrapolation from other social species suggests that this could be a reasonable explanation for the observed results. It is possible that mice housed in compatible pairs were better able to cope with the side effects of theophylline. Social buffering may also explain the results that pair-housed low-dose theophylline groups were recorded with significantly fewer “hypersensitive, moderate” observations than single-housed low-dose theophylline pairs. Further research into social buffering in mice needs to be completed before this can be confirmed. In addition, significantly more pair-housed high-dose theophylline mice were observed with thin fur cover over the muzzle. Social and mutual grooming is observed as an affiliative behavior between social animals.6 However, when combined with symptoms of increased activity, it is suspected that thin fur cover occurred due to overgrooming of social partners.
Due to the randomization procedure used before study start, significant differences in body weight existed between groups before study start. Despite this, it was observed that pair-housed mice gained more weight than single-housed mice, especially in control animals. Previous studies have demonstrated that social housing in mice improves thermoregulatory abilities and therefore reduces energy expenditure.16 This reduction in energy expenditure is thought to contribute to increased body weight gains in socially housed mice,11,12 as was reflected in the results of this study. Despite these changes in body weight gain, all values were within the historical control ranges for Crl:CD1 mice of this age at the Charles River Mattawan site, suggesting that the changes, while statistically significant, are likely not significant for toxicological studies.
Behavioral assays were suggestive of differences between singly housed and pair-housed animals. Both pair-housed and single-housed high-dose theophylline groups had lower nest scores for the duration of the study compared with both pair-housed and single-housed control groups. This is suspected to be due to the neurologic impacts of high-dose theophylline, with the increased activity resulting in compression of the nests. The pair-housed animals dosed with theophylline, however, showed significantly lower nest scores than single-housed theophylline animals during the first week of the study. This is hypothesized to be due to the impact of social instability during that time as scores improved following the initial stabilization period. As this effect was not seen in control animals, the administration of theophylline may have exacerbated this effect. After the new nesting material was added to pair-housed cages following nesting scoring in week 3, all pair-housed animals had higher nesting scores on average during week 4, with the high-dose theophylline pairs scoring an entire point higher than the previous week. This suggests that the quality of nesting material offered was affecting nest scoring, and therefore, the fact that we were transferring nesting material for the pair-housed but not the single-housed animals was a confounding variable in nest quality evaluation. It is therefore recommended that in future studies where nest scores are to be evaluated, existing nesting material is transferred for both single- and pair-housed animals with the potential for adding additional nesting material to all cages to help maintain nest quality. An additional observation made during the nest scoring procedure was that approximately 10% of the pair-housed animals were observed to be showing signs of conflict when their cages were relocated on the rack to allow for blinding. This was resolved by adding small amounts of foraging materials (eg, seed mixes; Alpha dri) to affected animals to distract the animals at the time the conflict was noted or to self-resolve once cages were returned to their previously allotted space on the rack after the evaluation was completed.
The TINT, in which mice are offered a small amount of new nesting material and then observed to see if they integrate the nesting material into the main nesting site, has been routinely used to measure well-being in mice.33,45 A positive score indicates integration of the new material and is thought to reflect a more positive animal welfare status.33,45 During the prestudy period, pair-housed animals overall were noted to have significantly fewer positive TINT scores than single-housed animals. This was unexpected as previous studies33,45 have demonstrated either no difference in TINT score between individual and group-housed mice, or an increased likelihood for single-housed mice to have a negative TINT score. It has been previously reported46,47 that new social pairs of male mice will experience higher levels of instability during initial hierarchy determination, but soon stabilize over time. It is suspected that this initial instability likely contributed to the lack of positive TINT scores in pair-housed animals in the prestudy period. The statistically significant improvement in the number of positive TINT scores for pair-housed high-dose theophylline animals compared with single-housed high-dose theophylline animals at week 4 highlights that this instability is temporary and that social housing may overall reduce stress, allowing the animals to more routinely participate in species-typical behaviors.
Pelt scoring has been used in previous studies34 evaluating social housing of male mice to help identify the presence of traumatic lesions due to fighting that are not readily visible on the surface of the skin, to minimize interrater variance in external wound scoring, and to help differentiate between ulcerative dermatitis lesions and wounding lesions. The current study revealed that all animals but one showed no evidence of dermal or subcutaneous injury; and there were no significant differences between groups in terms of injuries. The animal with a positive pelt score had a score of 0.11, suggestive of minimal injuries at most. This further supports the conclusion that animals can be socially housed without causing significant injury to one another, and that the measures used for observing animal compatibility in real time were accurate for predicting social compatibility.
Corticosterone levels can be measured in a variety of bodily substances and are routinely used as a measure of stress in rodents, with increased corticosterone levels suggestive of increased stress and anxiety experienced by the animals.48–50 In the current study, fecal corticosterone metabolite levels were observed to be significantly higher in high-dose theophylline groups during week 1, suggesting that the initial clinical effects of theophylline were stressful, regardless of housing condition. However, these effects were observed to have resolved by week 4, indicating that the stress experienced was likely temporary. Previous studies12,51,52 have examined the impact of social housing status on corticosterone in both urine and blood in mice. Those studies demonstrated either no impact of social housing status on corticosterone levels or overall increases in corticosterone levels in group-housed males. As male mice appear to excrete up to 73% of corticosterone metabolites via the feces,53 it was thought that fecal corticosterone metabolites may be more reflective of stress than urine values and would therefore be the most appropriate noninvasive method to assess corticosterone metabolite levels in the current study. During the last week of the study (week 4), pair-housed animals had significantly higher fecal corticosterone metabolite values than single-housed animals, when not accounting for dose group. Although there were no observed aggressive behaviors between paired animals, and pelt scores indicated no wounding had occurred between paired animals, higher fecal corticosterone metabolite levels could indicate increased unseen stress. Since week 4 is the last measured fecal corticosterone metabolite data point, it remains unclear whether the observed increase would continue over time or if it represents a temporary change. At week 4 of the study, the animals were approximately 11 weeks of age. As it is known that male mouse aggression can increase with age6 and, therefore, potentially result in increased stress levels, it is recommended that further research be conducted to define the correlation between fecal corticosterone metabolite levels and social compatibility as male mice age.
Clinical pathology was assessed to mirror similar requirements in toxicological studies. While statistically significant differences were noted in several hematologic parameters, all values were within the historical control range for CD-1 mice of this age at the Charles River Mattawan site and within currently published ranges.54 This suggests that the observed hematologic differences, while statistically significant, are not significant for toxicological studies. High doses of theophylline have also been reported to cause increases in MCV and MCHC and decreases in lymphocytes in mice,41–44 which was reflected in the high-dose theophylline animals in the current study. In addition, while reductions in lymphocyte counts alone were not statistically significant, for the low- and high-dose pair-housed theophylline animals compared with the control pair-housed animals, these reduced values appeared to be the major contributors to the statistically significant reductions in total white blood cell counts.
We do recognize that several limitations exist for the current study. The study duration was limited to 28-days following a 2-week-long stabilization period. As mentioned above, given that aggression has been reported to increase with age in male mice, studies of increased duration need to be performed to see if the steps followed in the current study will allow continued social housing of male mice. In addition, only one strain of mice was examined, so how well these data can be extrapolated to other strains is currently uncertain. While CD-1 mice were specifically selected for this study due to their frequency of use in toxicological research, they have also been reported as one of the more highly aggressive mouse strains.55 Therefore, we do anticipate that males of those strains known to be equally or less aggressive will have similar success rates when socially housed; however, further research is needed to confirm this. Finally, as previously mentioned, this study was designed based on OECD guidelines and the requirements of regulatory agencies such as the FDA, with statistical analyses performed using their standards. For some of the variables measured, such as nest and TINT scoring, these types of analyses may be contrary to methods routinely used by those exclusively studying behavior. We opted to include these measures despite this, as we felt they gave us further insight into the animals’ experience beyond just clinical observations and commonly reported toxicological variables; however, we recognize that it would be beneficial to perform future studies focusing more intensely on the behavioral components included here.
In summary, we successfully pair-housed CD-1 male mice on a 28-day orally dosed toxicology study. Based on parameters assayed, there were no effects on parameters typically measured for similarly designed types of toxicology studies, and behavioral assessments suggest that there may be benefits to pair housing once hierarchies have been established. Further studies should be conducted on other mouse strains/stocks, other types of dosing, and for longer duration studies, with additional focus on the impacts on animal welfare and stress. These results are encouraging, however, in advancing the goal of social housing for all mice used in toxicology studies.
Study Timeline. BW/CO, body weight/clinical observations. *Animals placed into cages individually or pair housed at the time of receipt.
Nest Scores over Time. Significant differences between pair-housed (P) animals from control (*), between single-housed (S) animals from control (&), and between single- and pair-housed animals at the same dose level (#) are highlighted. Arrow indicates time at which additional nesting material was provided to the pair-housed animals.
Number of Positive TINT Scores over Time. Significant differences between single- and pair-housed animals at the same dose level (#) are highlighted. TINT, time-to-integrate-nest-material testing.
Contributor Notes