Circadian rhythm and response to an acute stressor of urinary corticosterone, testosterone, and creatinine in adult male mice

In small laboratory species, steroid measures can be obtained more frequently and less invasively from urine than blood. Insofar as urinary levels reflect systemic levels, they could provide advantages particularly for measurement of glucocorticoids, whose blood levels react rapidly to handling and stress. In Experiment 1, urinary samples were collected from male mice every second hour over a 14:10 h light:dark cycle. Samples were analyzed via enzyme immunoassay for corticosterone, testosterone, and creatinine. Corticosterone had peak concentrations 1 h after light offset and a trough 1 h after light onset. Testosterone showed peak concentrations 5-7 h after light onset and lowest concentrations during the dark phase of the cycle. Creatinine showed some variation over the light-dark cycle, but steroid measures showed similar trends with and without adjustment for creatinine. In Experiment 2, mice were stressed via an injection at times close to the determined peak and trough levels of corticosterone. In urinary samples taken 90 min after injection, corticosterone was significantly higher in injected animals at both times relative to levels in control animals, but testosterone was unaffected by injection stress. In Experiment 3, serum and urine samples were collected from mice every sixth hour across the diurnal cycle. Corticosterone peaked in urine and serum immediately after light offset, and urinary measures predicted those in serum. These data indicate that urinary corticosterone reflects systemic levels in mice, document circadian variation in urinary testosterone, and indicate that circadian variation in creatinine is minimal, but potentially relevant in stressed animals.