Effects of housing density on weight gain, immune function, behavior, and plasma corticosterone concentrations in BALB/c and C57BL/6 mice

Abstract

The Guide for the Care and Use of Laboratory Animals contains recommended housing densities for rodent species that are commonly used by the scientific community. However, at the time of the Guide's publication, housing density recommendations were based heavily on the professional judgment of qualified scientists. Some scientists therefore question whether rodents can be housed at greater densities, whereas others wonder whether the space currently provided for rodents is sufficient. The present study was designed to determine the effect of housing adult female BALB/c- and C57BL/6-mice in standard 75-in(2) (484-cm(2)) ventilated cages at various housing densities (n = 2, 5, and 10 mice/cage). Measures of weight gain, plasma corticosterone, behavior, and immune parameters were evaluated at 7, 28, and 70 d after housing allocation. Housing BALB/c mice at 10/cage had negative effects on weight gain, corticosterone, behavior, and immune parameters. Housing C57BL/6 mice at 10/cage did not affect immune function or weight gain, although behavior and corticosterone showed statistical trends implying a negative effect Differences associated with housing densities of 2 and 5 mice/cage were less robust for all variables measured. We conclude that housing female BALB/c mice at 10 mice/cage (that is, at twice the Guide-recommended density) affects their physiology. We also conclude that mice vary in their responses in the parameters measured. These observations support the conclusion that it will be extremely challenging to scientifically determine an optimal cage density standard that can be uniformly applied across all mouse strains.

Figure 1.

Experimental design. Each time point represents a single mouse that was used for terminal collection of blood (immune and stress assays) or behavioral testing.

Figure 2.

Weight change (grams of weight gained from time 0) in BALB/c and C57BL/6 mice housed at 2, 5, or 10 mice/cage for (A) 7 d, (B) 28 d, or (C) 70 d. Throughout the experiment, BALB/c mice housed at 2/cage gained significantly more weight than did BALB/c mice housed at 10/cage. This effect was not seen in C57BL/6 mice. *, P < 0.05 (3-way ANOVA).

Figure 3.

The number of entries into a central area of the open field by (A) BALB/c and (B) C57BL/6 mice housed at 2, 5 or 10 mice/cage for 7, 28, or 70 d. In both strains, mice housed at lower densities generally had more open field entries than did those housed at higher densities. *, P < 0.05 (3-way ANOVA).

Figure 4.

The percentage of time spent in the outer wall area by BALB/c and C57BL/6 mice housed at 2, 5, or 10 mice/cage for (A) 7 d, (B) 28 d, or (C) 70 d. After 70 d, BALB/c mice housed at 10/cage spent more time in the outer wall area than did BALB/c housed at either 2/cage or 5/cage. Housing density did not affect this measure in C57BL/6 mice. *, P < 0.05 (3-way ANOVA).

Figure 5.

The plasma corticosterone levels in BALB/c and C57BL/6 mice housed at 2, 5, or 10 mice/cage for (A) 7 d, (B) 28 d, or (C) 70 d. BALB/c mice housed at high density (10/cage) had higher corticosterone levels than did those housed at low density (2/cage). Mice housed at medium density (5/cage) initially mirrored the low-density group. However, by the 70-d time point, corticosterone levels for the mice housed at medium density became more similar to those of mice housed at high density. This effect was not seen in C57BL/6 mice. *, P < 0.05 (3-way ANOVA).

Figure 6.

Percentage of splenic (A) total, (B) helper, and (C) cytotoxic T cells in BALB/c and C57BL/6 mice housed at 2, 5 or 10 mice/cage at the 70-d time point. Percentage of total T cells was unaffected by housing density in both strains (A). However, BALB/c mice housed at 10/cage had significantly lower levels of helper T cells than did those housed at the lower densities (B). Housing density had no effect on cytotoxic T cell levels in BALB/c mice (C) or on any T cell population in C57BL/6 mice. *, P < 0.05.

Figure 7.

Correlation of number of helper T cells to corticosterone levels in BALB/c mice housed at 2, 5 or 10 mice/cage for 70 d. An inverse relationship was present, with low helper T-cell levels correlating with high corticosterone levels. Furthermore, a clustering of housing densities was present. *, P < 0.05).