Circadian Disruptions in the Myshkin Mouse Model of Mania Are Independent of Deficits in Suprachiasmatic Molecular Clock Function

Background Alterations in environmental light and intrinsic circadian function have strong associations with mood disorders. The neural origins underpinning these changes remain unclear, although genetic deficits in the molecular clock regularly render mice with altered mood-associated phenotypes. Methods A detailed circadian and light-associated behavioral characterization of the Na+/K+-ATPase α3 Myshkin (Myk/+) mouse model of mania was performed. Na+/K+-ATPase α3 does not reside within the core circadian molecular clockwork, but Myk/+ mice exhibit concomitant disruption in circadian rhythms and mood. The neural basis of this phenotype was investigated through molecular and electrophysiological dissection of the master circadian pacemaker, the suprachiasmatic nuclei (SCN). Light input and glutamatergic signaling to the SCN were concomitantly assessed through behavioral assays and calcium imaging. Results In vivo assays revealed several circadian abnormalities including lengthened period and instability of behavioral rhythms, and elevated metabolic rate. Grossly aberrant responses to light included accentuated resetting, accelerated re-entrainment, and an absence of locomotor suppression. Bioluminescent recording of circadian clock protein (PERIOD2) output from ex vivo SCN revealed no deficits in Myk/+ molecular clock function. Optic nerve crush rescued the circadian period of Myk/+ behavior, highlighting that afferent inputs are critical upstream mediators. Electrophysiological and calcium imaging SCN recordings demonstrated changes in the response to glutamatergic stimulation as well as the electrical output indicative of altered retinal input processing. Conclusions The Myshkin model demonstrates profound circadian and light-responsive behavioral alterations independent of molecular clock disruption. Afferent light signaling drives behavioral changes and raises new mechanistic implications for circadian disruption in affective disorders.

2 provided ad libitum. Animals were maintained in light-tight cabinets and exposed to a 12h:12h LD cycle (~56 μW/cm 2 from a broad spectrum fluorescent light source) for a minimum of 14 days before release into DD for a minimum of 14 days to assess free-running rhythms. The wheel-running rhythms of a separate cohort of female mice (+/+ n=8, Myk/+ n=7) were assessed as described above in LD, DD, and an 8h advanced LD (see below) as well as in constant light (LL; for up to 15 days; light intensity 56 μW/cm 2 ).
Behavioral assessment without running-wheels: To monitor behavior in the absence of wheel-running activity, mice (+/+ n=7, Myk/+ n=9) were singly-housed in cages measuring 425 x 265 x 150 mm that were equipped with a running wheel that was permanently disabled (unable to rotate) for the duration of the experiment (TSE Systems, Bad Homburg, Germany). Cages were fitted with an infrared activity monitoring system (Inframot, TSE Systems) to record locomotor activity in the homecage. Activity was recorded every 10min using PhenoMaster software (TSE Systems).
Mice were maintained under a 12h light:12h dark cycle (LD; light intensity 56μW/cm 2 ) for a minimum of 15 days before transfer to constant darkness (DD) for the remainder of the experiment. After an initial acclimatization period of ~5 days, behavioral activity was monitored for the last 10 days of LD and the first 10 days of DD, during which we assessed the period of activity and the duration of the active phase (alpha) under both LD and DD.
Alpha was assessed from actograms using eyefit regression lines through the daily onsets and offsets of activity. Period was assessed using Clocklab software (Actimetrics, Evanston, IL, USA).
Phase shifting protocol: Animals were exposed to either an Aschoff type I (+/+ n=18; Myk/+ n=15) or a type II light phase shifting protocol (+/+ n=11; Myk/+ n=11) (3). Under type I conditions, animals were allowed to free-run in DD for 14 days prior to receiving a light pulse. On the day of the pulse, CT12 for each animal was predicted by fitting a regression line to the wheel-running actogram. At the predicted circadian time (CT; CT14, CT20, CT23), 3 animals were carefully transferred to a light-tight cabinet in an adjacent room with the lights on (~56 μW/cm 2 ) for one hour before return to home cages. For type II pulses, animals were released into DD for 24h and the cabinet light turned on for one hour (~56 μW/cm 2 ) at the appropriate CT, calculated relative to ZT0 under 12h:12h conditions. For this protocol, animals were not moved to another cabinet for light pulsing.
Re-entrainment protocol: To assess the rate of re-entrainment to a new light-dark cycle and hence the animal's response to 'jet-lag', male and female (+/+ n=10, Myk/+ n=10) animals were individually housed in a 12h:12h LD cycle for 14 days and then the LD cycle was advanced 8 hours, with lights-on now occurring some 8h earlier relative to the previous LD cycle. Animals were maintained on this advanced 12h:12h LD cycle for a further 14 days before the LD cycle was delayed by 8 hours. To delay the LD cycle, lights were kept on for 8 hours into the dark phase and the new 12h:12h LD cycle was maintained for a further 14 days. For the transient phase advance protocol (+/+ n=7, Myk/+ n=5), the LD cycle was advanced by 7h as described above, but this new lighting schedule was maintained for only 48h and then the animals were released into DD. Masking protocol: To assess the effects of "negative" masking on behavior, after 14 days under a 12h:12h LD cycle animals were given a 1-hour light-pulse (~56 μW/cm 2 ) without physical disturbance at either ZT14 or ZT20 (+/+ n=18; Myk/+ n=23). In addition, a cohort of +/+ and Myk/+ animals were exposed to an 8-hour light pulse (~56 μW/cm 2 ) in which lights were turned on between ZT16-ZT24.
Open-field assessment: All animals (+/+ n=8; Myk/+ n=7) were group housed and maintained under a 12h:12h LD cycle. Assessments were performed in a 300 x 300 x 240 mm polyethylene arena ( Figure S2) and data acquired through a video camera mounted above the recording area. Animals were habituated to the arena as a group then individually over a period of 5 days prior to testing. During data collection, the experimental room was illuminated in dim light (~1.5 μW/cm 2 ) and each animal placed in the arena between ZT15-18 and left for 15 minutes with minimal disturbance for video acquisition (24 frames s -1 ). Videos were analysed using EthoVision XT (Noldus, Netherlands).
Optic Nerve Crush (ONC) protocol: Mice were anaesthetised with a mixture of 13 mg/kg of Rompum™ (Bayer Inc., Mississauga, Ont., Canada) and 87 mg/kg of Ketalar® (ERFA Canada 2012 Inc., Montreal, QC, Canada). Under a dissecting microscope, a small incision was made with spring scissors in the conjunctiva, beginning inferior to the globe and around the eye temporally. To expose the posterior aspect of the globe, allowing visualization of the optic nerve, the edge of the conjunctiva next to the globe was retracted with micro-forceps, rotating the globe nasally. Using Dumont cross-action forceps, the exposed optic nerve was clamped approximately 1-3 mm from the globe for 10 s, after which the optic nerve was Whole-cell current-clamp recordings were made using a npi BA-01X bridge amplifier (npi electronics, Tann, Germany). Recording electrodes were fashioned from borosilicate glass capillaries pulled on a two-stage pipette puller (PC-10, Narishige, Tokyo, Japan). Pipettes (7-10MΩ) were half-filled with 0.22µm filtered intracellular solution; K-gluconate 130mM; KCl 7 10mM; MgCl 2 2mM; K 2 -ATP 2mM; Na-GTP 0.5mM; HEPES 20mM; EGTA 0.5mM; pH 7.28 with KOH; Osmolarity 295-300mOsmol kg -1 ; stored on ice to prevent ATP and GTP degradation. Cells were targeted using infrared video-enhanced differential interface contrast microscopy using a 40x water-immersion lens. Signals were sampled at 30 KHz and stored on a personal computer running Spike2 software for analysis (Cambridge Electronic Design, Cambridge, UK).