ABSTRACT: Living organisms detect seasonal changes in day length (photoperiod), and alter their physiological functions accordingly, to fit seasonal environmental changes. This photoperiodic system is implicated in seasonal affective disorders and the season-associated symptoms observed in bipolar disease and schizophrenia. Thyroid-stimulating hormone beta subunit (Tshb), induced in the pars tuberalis (PT), plays a key role in the pathway that regulates animal photoperiodism. However, the upstream inducers of Tshb expression remain unknown. Here we show that late-night light stimulation acutely triggers the Eya3-Six1 pathway, which directly induces Tshb expression. Using melatonin-proficient CBA/N mice, which preserve the photoperiodic Tshb-expression response, we performed a genome-wide expression analysis of the PT under chronic short-day and long-day conditions. These data comprehensively identified long-day and short-day genes, and indicated that late-night light stimulation induces long-day genes. We verified this by advancing and extending the light period by 8 hours, which acutely induced Tshb expression, within one day. In a genome-wide expression analysis under this condition, we searched for candidate upstream genes by looking for expression that preceded Tshb’s, and identified Eya3 gene. These results elucidate the comprehensive transcriptional photoperiodic response in the PT, revealing the complex regulation of Tshb expression and unexpectedly rapid response to light changes in the mammalian photoperiodic system. Mice were separated into 2 groups. One group was maintained under the short-day conditions (light: dark = 8 h:16 h, ZT0 = lights on, ZT8 = lights off, 400 lux) and the other was housed under long-day conditions (light:dark = 16 h:8 h, ZT0 = lights on, ZT16 = lights off, 400 lux) for 2 weeks. The PTs of both groups were retrieved every 4 h for 1 day (6 time points for each group), starting at ZT0. For the experiments performed during the first day of the long-day conditions, we applied two different conditions, following 3 weeks under short-day conditions. In one, the light-onset was advanced by 8 hours (advance condition), and in the other, the dark period was delayed by 8 hours (delay condition). PTs from both groups were obtained every 4 h for 1 day, starting at the lights-on time. (Lights on for the advance condition was ZT16 as defined by the short-day condition. Lights on for the delay condition was ZT0). We sampled 25 mice at each time point. This whole procedure was repeated twice (n = 2) to obtain experimental replicates.
