Nesting behavior is associated with body weight and grip strength loss in mice suffering from experimental arthritis

Abstract

Objective animal health evaluation is essential to determine welfare and discomfort in preclinical in vivo research. Body condition scores, body weight, and grimace scales are commonly used to evaluate well-being in murine rheumatoid arthritis (RA) and osteoarthritis experiments. However, nest-building, a natural behavior in mice, has not yet been evaluated in wild type (WT) or genetically modified rodents suffering from collagen antibody-induced arthritis (CAIA). To address this, we analyzed nesting behavior in WT mice, calcitonin gene-related peptide alpha-deficient (αCGRP^-/-) mice, and calcitonin receptor-deficient (Calcr^-/-) mice suffering from experimental RA compared to healthy control (CTRL) groups of the same genotypes. CAIA was induced in 10-12-week-old male mice, and clinical parameters (body weight, grip strength, clinical arthritis score, ankle size) as well as nesting behavior were assessed over 10 or 48 days. A slight positive association between the nest score and body weight and grip strength was found for animals suffering from CAIA. For the clinical arthritis score and ankle size, no significant associations were observed. Mixed model analyses confirmed these associations. This study demonstrates that clinical effects of RA, such as loss of body weight and grip strength, might negatively affect nesting behavior in mice. Assessing nesting behavior in mice with arthritis could be an additional, non-invasive and thus valuable health parameter in future experiments to monitor welfare and discomfort in mice. During severe disease stages, pre-formed nest-building material may be provided to animals suffering from arthritis.

Figure 1

Clinical course of CAIA and CTRL animals. Longitudinal development of (a) body weight, (b) grip strength, (c) semi-quantitative clinical arthritis score, and (d) ankle size for CAIA and CTRL animals (WT, αCGRP^-/-, Calcr^-/-) from day 315 following arthritis induction. Displayed are mean values and standard error of the mean (SEM).

Figure 2

Relative distribution of nest scores (1–5) for CAIA and CTRL animals, based on observations per mouse and day for different disease stages of experimental RA.

Figure 3

OR estimates derived from mixed logistic regression models, along with respective scatterplots, including repeated-measure Spearman correlation coefficients. Forest plots show unadjusted and adjusted OR estimates (along with 95% CI) derived from mixed logistic regression models for the dependent variable nest score ≥ 3 and independent variables (a) body weight [per 1 g], (b) grip strength [per 10 g], (c) clinical arthritis score [per 1 unit] and (d) ankle size [per 0.1 mm]; separately for all mice, CAIA, and CTRL animals. Scatterplots include repeated-measure Spearman correlation coefficients (along with 95% CI) for mouse-day observations for (e) body weight and nest score, (f) grip strength and nest score, (g) clinical arthritis score and nest score, and (h) ankle size and nest score; separately for CAIA and CTRL mice.

Figure 4

Schematic and real-life representation of naturalistic nest score. (a) Schematic and (b)-(f) real-life images of the naturalistic nest score based on Gaskill et al.. White dotted circles indicate the nest center, except in (f), where the entrance to the nest is shown.