KCNMA1 encodes the voltage- and calcium-activated K⁺ (BK) channel, which regulates suprachiasmatic nucleus (SCN) neuronal firing and circadian behavioral rhythms. Gain-of-function (GOF) and loss-of-function (LOF) alterations in BK channel activity disrupt circadian behavior, but the effect of human disease-associated KCNMA1 channelopathy variants has not been studied on clock function. Here, we assess circadian behavior in two GOF and one LOF mouse lines. Heterozygous Kcnma1^N999S/WT and homozygous Kcnma1^D434G/D434G mice are validated as GOF models of paroxysmal dyskinesia (PNKD3), but whether circadian rhythm is affected in this hypokinetic locomotor disorder is unknown. Conversely, homozygous LOF Kcnma1^H444Q/H444Q mice do not demonstrate PNKD3. We assessed circadian behavior by locomotor wheel running activity. All three mouse models were rhythmic, but Kcnma1^N999S/WT and Kcnma1^D434G/D434G showed reduced circadian amplitude and decreased wheel activity, corroborating prior studies focused on acute motor coordination. In addition, Kcnma1^D434G/D434G mice had a small decrease in period. However, the phase-shifting sensitivity for both GOF mouse lines was abnormal. Both Kcnma1^N999S/WT and Kcnma1^D434G/D434G mice displayed increased responses to light pulses and took fewer days to re-entrain to a new light:dark cycle. In contrast, the LOF Kcnma1^H444Q/H444Q mice showed no difference in any of the circadian parameters tested. The enhanced sensitivity to phase-shifting stimuli in Kcnma1^N999S/WT and Kcnma1^D434G/D434G mice was similar to other Kcnma1 GOF mice. Together with previous studies, these results suggest that increasing BK channel activity decreases circadian clock robustness, without rhythm ablation.

Wheel activity was sampled every 1 min in ClockLab software (Actimetrics). Actograms were constructed by double-plotting consecutive days of activity over the recording period. Circadian period (τ), χ2 amplitude and Fourier transform relative power (FFT rPSD for 0.04–0.042 cycles/hr) were determined from the last 10 days in LD and the first 15 days of wheel running activity in DD in ClockLab software. Activity period (α) was determined as the length of time an animal had consolidated activity using default settings with manual adjustment, with ρ defined as the portion of the cycle outside of α. For re-entrainment experiments, after 7 days of stable entrainment, the LD cycle was phase advanced by 6 h. The response was calculated as the number of days to stable re-entrainment of at least 3 days. After keeping mice in DD for 7 days, phase shifts in response to a light pulse were calculated as the number of hours between the activity onset regression fits the day before and after a 30-min light pulse delivered in early subjective night (CT16). All bout parameters (bout duration, bouts/day, counts/bout) were calculated from the first 15 days in DD, where a bout of activity is defined as periods during which the activity does not fall below 30 counts/min for more than 10 min. Data was excluded from mice that failed to record activity counts for two consecutive days. All animal experiments were conducted blinded to genotype during data collection and analysis from at least three independent cohorts for each transgenic line. Data was obtained from males and females, and no randomization was used to allocate animals into cohorts. WT controls were compared to transgenic littermates within each individual transgenic line. 

Current-clamp recordings were made with a Multiclamp 700B amplifier and pCLAMP 11.2 software (Molecular Devices, Sunnyvale, CA). Data were acquired at a 50-kHz sampling rate. Recording window was at the peak (ZT4–8) of the circadian rhythm. All recordings were made with synaptic transmission intact to more closely approximate in vivo activity. Electrodes (4–8 MΩ resistance) were filled internal solution (in mM: 123 K-methane-sulfonate, 9 NaCl, 0.9 EGTA, 9 HEPES, 14 Tris-phosphocreatine, 2 Mg-ATP, 0.3 Tris-GTP, and 2 Na2-ATP, pH adjusted to 7.3 with KOH). After gigaseal formation and whole cell break-in, membrane properties were elicited from a +50-mV voltage step from a holding potential (V_h) of -70 mV. Access resistance was verified to be <30 MΩ with less than 20% change at the end of the recording. Series resistance was compensated at 60–70%. For action potentials, data were acquired in 10 s sweeps. Silent cells were identified by injecting a 20-pA current to elicit an action potential. Frequency was calculated as the average of each sweep. Baseline potential was determined as the average inter-spike potential in the presence of spontaneous activity. Input resistance was calculated as the linear slope of the voltage from hyperpolarizing current injections (‒10 to ‒25 pA in 5 pA increments). Instantaneous frequency histograms were constructed by normalizing the number of events in 0.5-Hz bins to the total number of events from all neurons (GraphPad Prism v9.3.0 (San Diego, CA)). All data were corrected for liquid junctional potential (10 mV).

Statistics and graphs:  Graphpad Prism v9.3

Wheel actogram data:  Actimetrics Clocklab running in MatLab (available from the authors upon request)

Action potential recordings:  pClamp v11.2