Reducing mouse anxiety during handling: effect of experience with handling tunnels

Abstract

Handling stress is a well-recognised source of variation in animal studies that can also compromise the welfare of research animals. To reduce background variation and maximise welfare, methods that minimise handling stress should be developed and used wherever possible. Recent evidence has shown that handling mice by a familiar tunnel that is present in their home cage can minimise anxiety compared with standard tail handling. As yet, it is unclear whether a tunnel is required in each home cage to improve response to handling. We investigated the influence of prior experience with home tunnels among two common strains of laboratory mice: ICR(CD-1) and C57BL/6. We compared willingness to approach the handler and anxiety in an elevated plus maze test among mice picked up by the tail, by a home cage tunnel or by an external tunnel shared between cages. Willingness to interact with the handler was much greater for mice handled by a tunnel, even when this was unfamiliar, compared to mice picked up by the tail. Once habituated to handling, C57BL/6 mice were most interactive towards a familiar home tunnel, whereas the ICR strain showed strong interaction with all tunnel handling regardless of any experience of a home cage tunnel. Mice handled by a home cage or external tunnel showed less anxiety in an elevated plus maze than those picked up by the tail. This study shows that using a tunnel for routine handling reduces anxiety among mice compared to tail handling regardless of prior familiarity with tunnels. However, as home cage tunnels can further improve response to handling in some mice, we recommend that mice are handled with a tunnel provided in their home cage where possible as a simple practical method to minimise handling stress.

Figure 1. Handling methods used to pick up mice.

(A) Tail handling: the most widespread method currently used for handling mice. The animal is lifted by the base of the tail between thumb and forefingers and supported on the handler’s arm or hand. (B) Tunnel handling. The animal is guided into a plastic tunnel and held inside the tunnel. (C) Tunnels may be provided in the home cage as source of enrichment for mice as well as a tool for handling.

Figure 2. Voluntary interaction with the handler immediately before and after mice were picked up by a shared tunnel or tail.

Percentage of test time interacting with the handler immediately before (b, solid bars) and after (a, hatched bars) the first, fifth and ninth handling session for C57 and ICR mice picked up by one of two different handling methods (mean ± s.e.m., n = 8 cages for each handling group and strain). Shared Tunnel Only and Tail mice were kept under the same housing conditions and neither had any prior experience of tunnels. P values indicate the effect of handling method (repeated measures ANOVAs, full analyses given in Table S1). There was a significant interaction between handling method and strain in session 1 (F_1,26 = 22.2, P<0.001) and session 5 (F_1,26 = 19.5, P<0.001) but not in session 9 (F_1,24 = 0.1, P = 0.74).

Figure 3. Effect of tunnel experience on voluntary interaction before and after handling by tunnel.

Percentage of test time interacting with the handler immediately before (b, solid bars) and after (a, hatched bars) handling sessions one, five and nine for C57 and ICR mice handled either using a home tunnel or a shared tunnel (mean ± s.e.m., n = 8 cages per handling group for each strain). Mice handled with a shared tunnel either had prior experience with a home tunnel (Shared Tunnel Exp.) or no prior tunnel experience (Shared Tunnel Only). P values indicate the effect of tunnel experience for each strain (repeated measures ANOVAs, full analyses are given in Table S1). Asterisks indicate a significant planned contrast between the home tunnel method and shared tunnel groups (* P<0.05, ** P<0.01, ***P<0.001).

Figure 4. Effect of handling method and tunnel experience on anxiety measures.

Frequency of open arm entries (A), time on open arms (B) and protected stretch attend postures (C) in an elevated plus maze test for ICR (open bars) and C57 (hatched bars) mice after nine sessions of handling by home tunnel, shared tunnel or tail (mean ± s.e.m., n = 16 mice per handling group). Mice handled with a shared tunnel either had prior experience with a home tunnel (Shared Tunnel Exp.) or no prior tunnel experience (Shared Tunnel Only). P values indicate the effect of tunnel experience between tunnel groups, or compared shared tunnel versus tail handling (one way ANOVAs, full analyses are given in Table S2). There were no significant interactions between method and strain, except for a difference in protected stretch attend between shared tunnel only and tail handled mice (F_1,57 = 5.6, P = 0.02). Protected stretch attend was significantly higher among ICR mice handled by tail (P = 0.01) but not among C57 (P = 0.78).