Improved cage cleaning regimes as a husbandry refinement for laboratory mice

Lead Research Organisation: University of Oxford
Department Name: Zoology


We are assessing if we can improve the welfare of laboratory mice by cleaning out their cages less frequently.
Cage-cleaning is a particularly disruptive aspect of life for laboratory mice. Mice naturally avoid disturbance and rely heavily on smells for communication. So it is not surprising that cage-cleaning, which involves handling and the removal of home-cage odours, elicits many responses that indicate that they find it unpleasant. However, regular cage-cleaning is necessary to prevent high ammonia levels, which can cause health problems.
We are investigating whether reducing cage-cleaning frequency (fortnightly vs. weekly) and providing ‘low-ammonia’ (vs. standard) bedding helps to improve welfare without adversely affecting health. We are assessing the effects of these factors in two commonly used mouse strains – BALB/c and C57BL/6 – via non-invasive physiological and behavioural welfare indicators and analysis of ammonia-related nasal and lung tissue damage. We are also collaborating with a commercial breeding facility to assess whether cage-cleaning frequency affects the reproductive output of breeding females.
The results of this research will enable us to make recommendations about cage-cleaning practices most benefit mice, by optimising their well-being and health; scientists (and ultimately the general public), by improving the validity of mice as research models; and commercial facilities, by improving their productivity.

Technical Summary

The aim of this project is to assess the impact of cage cleaning procedures on the welfare, reproductive output and health of laboratory mice, and by doing so to inform best practice for husbandry procedures in research and commercial mouse units. The project is designed, firstly, to test the hypothesis that disruptive cage cleaning procedures are a source of chronic stress, reflected by increased HPA activity, abnormal behaviours (such as stereotypies and barbering), infant mortality and ?emotionality? in behavioural tests. This project will also assess the relationship between cage-cleaning frequency and the effect of ammonia levels on irritation of the eyes and pathology of the respiratory system. Finally this project will assess whether differences in stress, caused by differences in cage-cleaning procedures (particularly inter-laboratory differences), help shed light on currently unexplained sources of variation between research facilities and studies.
This research will investigate the effects of two aspects of cage cleaning that currently frequently differ: frequency of cleaning (twice-weekly and fortnightly) and bedding type (standard woodchips and low ammonia bedding). The treatments will be assessed in a 2x2 cross-factored design. Each treatment will be replicated in each of four animal houses, and in two commonly used inbred strains of mice, BALB/c and C57BL/6J. The subjects will be: 1) adult males, differentially cage-cleaned since weaning age; 2) breeding females, differentially cage-cleaned throughout their economic breeding life; and 3) male offspring from differentially cage-cleaned breeding females, to test for lasting effects of pre- and early post-natal experience.
Welfare will be assessed using conventional, non-invasive welfare indicators, including urinary corticosterone, prevalence and severity of stereotypy, and infant mortality. Reproductive output will be assessed via litter size and weight at weaning, inter-birth interval and total lifetime output. Health will be assessed through behavioural signs of mucosal irritation (e.g. blink rates) and by post-mortem health screening and histological analysis of respiratory tract tissues.
This will be the first detailed study of the effects of cleaning regimes on the welfare, reproductive output and health of laboratory mice. The results will enable us to make practical recommendations to improve the cage-cleaning regimes used in both research and commercial mouse units.


10 25 50