Circadian rhythms in multiple behaviors depend on sex, neuropeptide signaling, and ambient light
Institutional Repository Document
Overview
Research
Identity
Other
View All
Overview
abstract
AbstractOrganisms have evolved circadian (near-24 h) rhythms in behavior to anticipate daily opportunities and challenges such as mating and predation. However, the ethological investigation of circadian behavioral rhythms has been traditionally limited to studying easy-to-measure behaviors at higher temporal resolution or difficult-to-measure behaviors with limited temporal resolution. Our ability to simultaneously record circadian rhythms in multiple behaviors has also been limited by currently available technology. We thus sought to examine eight overt, ethologically-relevant behaviors never before studied simultaneously as a function of time of day: eating, drinking, grooming, rearing, nesting, digging, exploring, and resting. To address the hypothesis that the daily patterning of these behaviors depends on neuropeptide signaling, sex, and ambient light, we used high-throughput machine learning to automatically score millions of video frames of freely-behaving male and female wild-type and vasoactive intestinal peptide (Vip)-deficient mice. Automated frame-by-frame predictions of the eight behaviors correlated highly with consensus labels by trained human classifiers. We discovered reliable daily rhythms in many previously unreported behaviors that peaked at stereotyped times of day and persisted in constant darkness. Surprisingly, nesting and digging rhythms differed dramatically in both phase and amplitude between male and female mice. InVip-deficient mice, daily rhythms in most behaviors were low amplitude and peaked earlier in the day in a light:dark cycle, while rhythms in all behaviors peaked randomly throughout the day in constant darkness. We also found that for most behaviors, time budgets predominantly differed by light cycle, but transition probabilities predominantly differed with VIP signaling and by sex. We conclude that machine learning can be used to reveal novel sex, neuropeptide, and light-dependent behaviors at time scales from seconds to days.
