The use of sterile male mice to induce pseudopregnancy in female mice assigned for the implantation of embryos is a vital component in the production of Genetically Altered Animals (GAA). Usually the male is vasectomized, which is a surgical procedure under general anesthesia, with time required for the male to recover and the spermatozoa count to drop to zero. One refinement to GAA production has been the development of genetically sterile mice, in particular the double transgenic Protomine1 (Prm1) strain of mouse. The male mice born with the dominant Prm1-EGFP transgene were found to be naturally sterile and can be used as a replacement for vasectomized mice (Haueter et al., 2010). These GAA mice have successfully replaced the need to use vasectomized male mice for generations of pseudo pregnant females in our facility. As the male mice utilized to induce pseudopregnancy cannot subsequently be re-housed with other males due to fighting, these mice are individually housed for up to 9 months depending on their breeding capability. Although male mice are not generally considered to be social animals there is evidence that shows they prefer to be housed with conspecifics (Goldsmith et al., 1978; Van Loo et al., 2002; Kalliokoski et al., 2014). In a breeding colony, fighting between a male: female pair or trio made up of one male and up to two female mice is less common, indicating that the sterile males may benefit from having a female companion over a male companion. Following on from work carried out by the originator of this model, each of our cohorts of sterile male mice have a companion female housed with them to reduce anxiety associated with long term single housing (Kalliokoski et al., 2014). When the male is required for the breeding program the companion female is replaced with a new female in the morning. The new female mouse is left in the cage for at least 24 h, or until a copulatory plug is observed in the vagina, indicating that mating has occurred.

We cannot be certain by visual observations alone what the duration and amount of disruption this intervention will cause in a mouse cage. Some of the reasons for this are: if there is an observer in the room it can lead mice to become more active than if they were undisturbed and if direct observations are used (where a person is in the room monitoring cages) there will undoubtedly be some subjectivity and or human error. The solution of a video recording can reduce this; however, it is very time consuming to review footage on a cage by cage basis.

To investigate the disruption caused to mice during the day of the replacement and how long it would take for the cage activity to return to baseline in an objective way, we housed mice in the Digital Ventilated Cage (DVC^®) system, Tecniplast S.p.A, which is a non-intrusive mouse activity tracking system. This enables us to monitor mice without the need to use a video camera, subcutaneous microchip or have a telemetry device surgically implanted. Each cage position on the rack has a sensor board underneath it, containing 12 electromagnetic sensors to continuously track and monitor spontaneous mouse activity (Iannello, 2019) i.e., mouse activity without human intervention. The cage sits on the runners in the same way as any standard Individually Ventilated Cage (IVC) would, so there is no disturbance to the animals. The individual sensor board sends data generated by the cage occupants to a Master computer, which has a temporary data storage area and sends the data directly into a cloud storage platform from where it can be downloaded. Each time a mouse moves over an electrode, the mouse activates a signal which is sent to the Master computer, these activations form the “Animal Activity Index” which is a normalized measurement assuming values between 0% (no sensors activated) and 100% (all electrodes activated simultaneously). The summary measure over the cage is the Average Activity Index (AAI); calculated as the average of all twelve electrode activations per minute. When we discuss cage activity in this paper we are referring only to the AAI.

Other studies have shown that procedures involving cage changing handling or the transfer of the animals disrupts their basic activity pattern and induces changes that appear to impact them over an extended period of time (Gray and Hurst, 1995; Lim et al., 2019; Pernold et al., 2019). We wanted to investigate the disruption to cage activity during the day of the female exchange and how long it would take for the activity to return to baseline. We initially carried out a Pilot study with a small cohort of mice to investigate whether timed mating activity is masked by cage changing, followed by a subsequent (Main study) investigation of female replacement as part of timed mating. We hypothesized that the impact of mouse activity in the cage is significantly increased when a companion female is replaced with a new female.

The exploratory data analyses from these studies highlighted that there was a lot of variability between cages (including within groups) across each 24 h time frame (see Figure 1). In order to understand any changes in cage activity we decided to focus on specific 2-h time intervals. The 2-h time intervals we discuss here are when the mice are active after a cage is moved or opened [08:00–10:00 h (Main study only) and 10:00–12:00 h], when the mice were inactive (12:00–14:00 h) and, when the mice are active directly after the lights went out (20:00–22:00 h).

All of the mice were bred at the establishment and housed in GM500 mouse cages (Tecniplast S.p.A) with a conventional wire lid. We included all available pairs of Prm1 males with their companion females and the CD1 females used as part of colony management. We had 14 Prm1 males and 14 CD1 female companions (bred at the establishment) and 10 CD1 females (Charles River Laboratories, United Kingdom) used as replacement females in the Pilot study and 20 Prm1 males, 20 CD1 companion females and 10 CD1 (Charles River Laboratories, United Kingdom) replacement females in the Main study. The male weights ranged between 33–56 gm and the female weights between 28–37 gm at the start of the Pilot study (see Supplementary Table 1).

Mice were transferred from the breeding room to the animal room. And on arrival the mice were health checked carefully and acclimatized for 7 days prior to the start of each study in XT GM500 Digitally Ventilated Cage (DVC^®) (Tecniplast S.p.A, Italy) made from a clear Polypropylene plastic base with a stainless-steel bar lid, the internal measurements of the cage are 35.5 cm × 17.5 cm × 12 cm. The DVC^® was placed along the side wall near the rear of the room. A health monitoring program for mice was in place and excluded infectious agents described in the FELASA guidelines (Berard et al., 2014). For identification and randomization purposes all the females in these studies were ear marked prior to the start of the study. All new female mice were group housed in groups of four on an IVC rack in the room (prior to being introduced to the male) and established male/female pairs were housed on the DVC^® rack. When the companion female mice were removed from the cage, they were group housed with three other companion females on the IVC rack in the same animal holding room. The IVC rack was located along the side wall next to the DVC^® both were connected to a SmartFlow^® Air Handling Unit (Tecniplast S.p.A) located centrally between them.

All cages had 10 mm depth of Lignocel BK8/15 (IPS) bedding, a mouse igloo measuring 10 cm diameter × 6 cm high, and an aspen wood chew block 50 mm × 10 mm × 10 mm long (Datesand, United Kingdom) with a Lignocel Large Wood Wool disk (IPS, United Kingdom) and a mouse Mini Fun Tunnel 76 mm × 38 mm circumference (LBS, United Kingdom). Cages were changed in line with study protocol; where the mice are removed from the home cage and placed in a fresh cage with fresh identical bedding and fresh enrichment items.

Mice were fed ad libitum with 5LF2 irradiated rodent diet (IPS) via food hoppers at the rear of the cage, and ad libitum animal grade reverse osmosis filtered, and UV treated drinking water via bottles attached to the front of the cages. Food and water levels were monitored by the staff and the DVC^® system. After cage changing all enrichment remained with the animals for the duration of each study. Throughout the facility there was piped radio tuned into a local commercial station during the light phase. The light cycle was on a 14:10 light: dark with light phase from 06.00–20.00 h and a gradual increase or decrease of lighting over a 10-min dawn: dusk period. The room temperature was 21–24°C, humidity was 55% ± 10%, with 20 air changes per hour in the room and 75 air changes per hour in the DVC^® and IVC.

In the Pilot study females were replaced between 09:30 and 10:00 h, the comparison of Day 3 vs. Day 2 (pre-female replacement) showed a significant increase in activity for the 10:00–12:00 h time frame, due to both Female Replacement and Cage Change (see Table 2). There was no significant difference in activity of female replacement or cage activity for the timeframes 12:00–14:00 h and 20:00–22:00 h, see Supplementary Table 2 for a complete set of results.

In the Main study females were replaced between 07:00–08:00 h and the T-tests showed, in the cages where females were replaced (Grp_1.m), there was a significant difference in activity in the time frame comparison for 10:00–12:00 h when compared to the average activity of previous days, and for the comparisons of the time frame for 10:00–12:00 h and 12:00–14:00 h when compared to subsequent days (see Table 2).

Function on scalar regression tests showed that the cage activity in Grp_1.m (where females were replaced) had higher activity peaks and returned to baseline ∼ 4 h post intervention (Figure 8) compared to the sham replacement in Grp_2.m where there was no significant increase in activity compared to pre-replacement.

Initially we investigated whether the impact of replacing a female would be masked during the cage change process as studies have shown that there is a significant increase in activity post cage change (Pernold et al., 2019). We found the increased spike in activity in the 2 h after the female replacement was conducted was much more significant than the actual disruption caused by cage changing (Table 2). However, cage change was still a significant effect within this time frame and therefore, we removed cage changing from the Main study where we wanted to investigate whether replacing females 6 days after cage changing would show higher activity than not replacing the females and how long it would take for the disruption to return closer to the pre-replacement baseline.

We found that there was an increase in activity across the cages where the female was replaced compared to when the companion female was briefly removed. From the T-Test analyses of the Main study, we inferred that across the study period it took 4 h for the activity levels to be similar to pre-replacement baseline activity levels across all the timeframes, we also tested the return to baseline more precisely using Functional regression and found the activity levels returned to baseline 4 h directly post intervention. Home Cage Monitoring gave us an insight into the impact of timed mating on the expected activity pattern of the study mice. We found a notable increase in activity during the light phase and an unexpected decrease in activity during the dark phase (see Figures 6, 7) as a response to female replacement, while not statistically significant it is an interesting finding that we would not have known if it were not for the DVC^®, which is another benefit to having 24 h cage monitoring.

Understanding how our routines and procedures impact animal activity is important, especially if it is likely to have a negative effect on the animals and likely to be repeated with the same animal over a number of months such as timed mating for stud male mice. Interventions that are considered “non-invasive” such as cage changing (Rosenbaum et al., 2009; Pernold et al., 2019), or “sub threshold” such as timed mating, can cause significant spikes in activity which are not fully understood using a traditional scoring system based on cage side observation. For example, changes in activity are usually included in welfare score sheets for moderate to severe protocols. These score systems (Ullman-Cullere and Foltz, 1999) have been used as an objective assessment of changes in behavior and body condition which can indicate pain and distress in mice. This is an “in the moment” assessment which usually involves having to remove a cage from the rack. This disturbance would not give any meaningful observation data when trying to observe the response of mice to timed mating. Therefore, significant disruptions in the activity patterns of mice (Pernold et al., 2019), which could lead to increased levels of stress may be missed.

The importance of sleep in mammals has been demonstrated (Horne and McGrath, 1984; Tang et al., 2005; Palchykova et al., 2006; Colavito et al., 2013; Febinger et al., 2014) and the resulting stress associated with sleep disturbance could cause a confounding factor for experiments (Van Loo et al., 2002). Our study has indicated that timed mating may have an effect on the wellbeing of mice due to the potential disturbance in their activity pattern. The disruption in the activity pattern could lead to increase stress levels in male mice if they are used for repeated timed mating. While we can be confident the activity is as a result of replacing the companion female with a new female, we cannot be certain whether the activity is just the male, female or more likely it is a combination of both mice.

Our initial program of work was designed with a focus on the potential impact of timed mating on a male mouse, however, in doing so we missed out on observing the female’s response to their own different scenarios and whether there is any impact on their welfare. It would be interesting to follow the companion females and see how they react to being put in a new cage with other companion females and what happens when they are returned to the male. We postulated activity may be lower when she is returned to a familiar cage compared to if she were put in with an unfamiliar male in a new cage, especially if there are still smells and cues that are familiar. If, in addition, we included the impact on the new female before and after she has been with the male, this could give insight of how timed mating impacts all the mice involved.

Synchronizing our interventions to the activity pattern of mice may lead to improving their wellbeing (Febinger et al., 2014; Pernold et al., 2019). The impact of timed mating could be reduced if we carried out this procedure later in the afternoon as this is closer to the time that mice are more active. In the Pilot study the new female was left with the male for 5 days however, it should be noted that routinely the new female is removed after 24 h, but there are occasions when the initial mating is not successful, and females are left with the male for an extended period. This could lead to the male becoming accustomed to the new female, and when she is replaced with the companion female it becomes another disturbance in his natural activity pattern. Anecdotal observations from the research team were that mating was observed immediately after female replacement and continued for the duration of time that they were present in the room and was the most likely reason for the peak in activity, the use of a video camera would enable us to be certain that this observation was correct.

In conclusion this study has shown there is a potential impact on the welfare of male mice during timed mating as a result of the changes in their activity pattern. The impact of this intervention could be reduced if the timed mating occurred in the late afternoon when the mice are in the active part of their day.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The animal study was reviewed and approved by GlaxoSmithKline, United Kingdom. All studies were ethically reviewed and carried out in accordance with Animals (Scientific Procedures) Act 1986 and in the GSK Policy on the Care, Welfare and Treatment of Animals.

JM devised the initial study, worked with the staff in the room carrying out the practical work, and wrote the first draft of the manuscript’s introduction and methods. EB helped refine the study design and method for the second study, oversaw the statistical analysis on the data, helped produce the results, and wrote the results. Both authors contributed to the discussion and editing and refining the whole manuscript to prepare it for submission.

JM and EB are employed by the company GlaxoSmithKline, Brentford, United Kingdom.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.