Comparison of 2 blood sampling methods in mice to increase animal welfare and the reliability of experimental results

Abstract

There is a strong need to gain systematic knowledge about the potential impact of routine procedures on laboratory animals to ensure animal welfare and reliability of experimental results. Blood sampling is a frequently used procedure within laboratory animal research. Two of the most common anesthesia-free techniques to obtain blood in mice are the saphenous venipuncture and submandibular bleeding. It is common practice that these methods are applied repeatedly to 1 individual. The impact of both techniques, either performed singly or repeatedly, on the animal’s welfare is not fully known yet. We directly compared the effects of both sampling methods on physiological stress responses, behavior, tissue damage, and procedure duration in 2 frequently used mouse strains: C57BL/6 and BALB/c mice (n = 45/strain). The mice were randomly assigned to 3 treatment groups (saphenous venipuncture, submandibular bleeding, control) of 15 mice/strain and were sampled weekly (80 μl/ blood sample) during 4 wk for the experimental groups and 5 wk for the control group. To evaluate the acute physiological stress response, plasma corticosterone levels were determined. As an indicator of chronic stress, body weights were recorded once a week and the thymus was weighed after necropsy. Behavior was recorded using video cameras in front of the cages and afterwards analyzed by a well-trained observer using event logging software. Tissue damage was evaluated by scoring pre-and postmortem hemorrhages and hematomas at the injection site and by measuring alanine transaminase, aspartate aminotransferase, and creatine kinase plasma levels. Submandibular bleeding appeared to be less time consuming (P<0.05) while providing higher sample volumes. However, this method led to increased acute stress responses in C57BL/6 mice (P = 0.03). All animals of both strains undergoing this procedure expressed more pain-related behavior (P < 0.05) and the extent of tissue damage at the injection site was significantly higher when compared to the saphenous venipuncture (P < 0.05). Thus, in order to avoid the undesirable effects of stress on experimental outcomes and to reduce the extent of tissue damage, saphenous venipuncture should be preferred above submandibular bleeding whenever possible.