Transcriptomic Plasticity of the Circadian Clock in Response to Photoperiod: A Study in Male Melatonin-Competent Mice

Seasonal daylength, or circadian photoperiod, is a pervasive environmental signal that profoundly influences physiology and behavior. In mammals, the central circadian clock resides in the suprachiasmatic nuclei (SCN) of the hypothalamus where it receives retinal input and synchronizes, or entrains, organismal physiology and behavior to the prevailing light cycle. The process of entrainment induces sustained plasticity in the SCN, but the molecular mechanisms underlying SCN plasticity are incompletely understood. Entrainment to different photoperiods persistently alters the timing, waveform, period, and light resetting properties of the SCN clock and its driven rhythms. To elucidate novel molecular mechanisms of photoperiod plasticity, we performed RNAseq on whole SCN dissected from mice raised in Long (LD 16:8) and Short (LD 8:16) photoperiods. Fewer rhythmic genes were detected in Long photoperiod and in general the timing of gene expression rhythms was advanced 4-6 hours. However, a few genes showed significant delays, including Gem. There were significant changes in the expression clock-associated gene Timeless and in SCN genes related to light responses, neuropeptides, GABA, ion channels, and serotonin. Particularly striking were differences in the expression of the neuropeptide signaling genes Prokr2 and Cck, as well as convergent regulation of the expression of three SCN light response genes, Dusp4, Rasd1, and Gem. Transcriptional modulation of Dusp4 and Rasd1, and phase regulation of Gem, are compelling candidate molecular mechanisms for plasticity in the SCN light response through their modulation of the critical NMDAR-MAPK/ERK-CREB/CRE light signaling pathway in SCN neurons. Modulation of Prokr2 and Cck may critically support SCN neural network reconfiguration during photoperiodic entrainment. Our findings identify the SCN light response and neuropeptide signaling gene sets as rich substrates for elucidating novel mechanisms of photoperiod plasticity. Data is also available on our corresponding website, where users can search and view the expression and rhythmic properties of genes across these photoperiod conditions. Overall design: RNAseq was conducted on whole SCNs dissected from mice with intact melatonin signalling (C3A.BLiA-Pde6b+/J strain; JAX stock #001912) raised in Long-day (hours Light: hours Dark 16:8) and Short-day (LD 8:16) photoperiods from embryonic day zero (E0). To accomplish this, breeder pairs were placed under Short or Long photoperiods (ca. 100 lux) and litters were born and maintained under these conditions. As rodent circadian rhythms develop in utero, the first litters born under the experimental photoperiods were not used in this experiment in order to avoid any potential after-effects of parental entrainment to previous photoperiod (standard institutional housing (LD 12:12)). Male mice from second and subsequent litters were used at age postnatal day 50 (P50; early adolescence). At postnatal day 50, the lighting regime was switched to constant darkness (DD) at the respective time of lights off (ZT 12) for each photoperiod. Mice spent a minimum of 36h in DD before sample collection began, such that gene expression in the samples likely represents the state of the endogenous clock rather than acute responses to the previously light cycle (Hughes et al., 2017). After this interval, SCN dissections took place at intervals of 4h over 2 circadian cycles (Hughes et al., 2017), until a minimum n of 4 was achieved per timepoint per cycle. Short and Long photoperiod samples were collected at 36, 40, 44, 48, 52, 56, 60, 64, 68, 72, 76, and 80 hours after lights off. A dim red headlamp was worn by the experimenter during the first steps of dissection. Mice were euthanized by cervical dislocation and the eyes removed to prevent potential light signaling to the SCN. The subsequent steps took place under light: the optic nerve was cut and the brain was removed such that it was situated ventral side up in a petri dish. The SCN was dissected from the brain under a microscope. Each SCN sample was put in a 1.5mL sterile DNase- RNase-free tube, and tubes were immediately placed in powered dry ice. RNA and DNA were extracted from each SCN sample using the AllPrep DNA/RNA Micro Kit (Qiagen) according to the instructions. (DNA was stored for later use and is not used in this study).