Kv12-encoded K+ channels drive the day-night switch in the repetitive firing rates of SCN neurons

Abstract

Considerable evidence suggests that day-night rhythms in the functional expression of subthreshold potassium (K+) channels regulate daily oscillations in the spontaneous firing rates of neurons in the suprachiasmatic nucleus (SCN), the master circadian pacemaker in mammals. The K+ conductance(s) driving these daily rhythms in the repetitive firing rates of SCN neurons, however, have not been identified. To test the hypothesis that subthreshold Kv12.1/Kv12.2-encoded K+ channels play a role, we obtained current-clamp recordings from SCN neurons in slices prepared from adult mice harboring targeted disruptions in the Kcnh8 (Kv12.1-/-) or Kcnh3 (Kv12.2-/-) locus. We found that mean nighttime repetitive firing rates were higher in Kv12.1-/- and Kv12.2-/- than in wild type (WT), SCN neurons. In marked contrast, mean daytime repetitive firing rates were similar in Kv12.1-/-, Kv12.2-/-, and WT SCN neurons, and the day-night difference in mean repetitive firing rates, a hallmark feature of WT SCN neurons, was eliminated in Kv12.1-/- and Kv12.2-/- SCN neurons. Similar results were obtained with in vivo shRNA-mediated acute knockdown of Kv12.1 or Kv12.2 in adult SCN neurons. Voltage-clamp experiments revealed that Kv12-encoded current densities in WT SCN neurons are higher at night than during the day. In addition, the pharmacological block of Kv12-encoded currents increased the mean repetitive firing rate of nighttime, but not daytime, in WT SCN neurons. Dynamic clamp-mediated subtraction of modeled Kv12-encoded currents also selectively increased the mean repetitive firing rates of nighttime WT SCN neurons. Despite the elimination of the nighttime decrease in the mean repetitive firing rates of SCN neurons, however, locomotor (wheel-running) activity remained rhythmic in Kv12.1-/-, Kv12.2-/-, and Kv12.1-targeted shRNA-expressing, and Kv12.2-targeted shRNA-expressing animals.

Figure S1.

Generation of (Kv12.1^−/−) mice harboring a targeted disruption of the Kcnh3 locus. (A) Schematic of the targeted exon 1 Kcnh8 (Ex1) locus, the linearized targeting construct, the initial targeted allele, and the null (knockout) allele generated by Cre-loxP recombination. Targeting of the mouse Kcnh8 (Kv12.1) locus involved homologous recombination (dashed lines) in mouse embryonic stem cells between the native Kcnh8 locus and the targeting vector, and insertion of a loxP site and myristoyl-EGFP into Ex1 immediately upstream of the translation start. A viral 2A sequence joins the myristoyl-EGFP open reading frame to the Kv12.1 open reading frame to potentially allow EGFP-labeling of Kv12.1-expressing cells carrying the targeted allele. Downstream of Ex1 in the first intron, insertions include an frt-bracketed neomycin resistance cassette (Neo) driven by the PGK promoter for positive selection of targeted ES cells on G418, a second loxP site, and dTomato-2A-rtTA (reverse tetracycline trans-activator) cassette (dTOM-rtTA) which includes an SV40 polyadenylation sequence to terminate transcription and block expression of downstream exons. TK (thymidine kinase) and DT (diptheria toxin) expression cassettes flank the left and right, respectively, arms in the targeting construct. These negative selection cassettes are eliminated by homologous recombination but were included to suppress the random insertion of the targeting construct into the ES cell genome. Note the native allele is bracketed by Nhe I restriction sites ∼15 Kb apart. In the targeted allele, the Nhe I sites are preserved, but the distance between them is increased to ∼18 Kb, and a unique Bgl II site is introduced upstream of the 5′ loxP site. Hybridization probes located between either (left or right) arm and the neighboring Nhe I site will label a ∼15 Kb band in a Southern blot of Nhe I/Bgl II digested genomic DNA for the WT allele. The same probes will label bands of 8 Kb (left arm) and 10 Kb (right arm) for the targeted allele. (B) Southern blot analysis of nine G418-resistant ES cell clones following Nhe I/Bgl II digestion of genomic DNA with a probe upstream of the left arm (left) or downstream of the right arm (right). The WT allele is identified by the 15-kb band with both probes, whereas the targeted allele is identified by 8-kb (left arm) and 10-kb (right arm) bands. Arrows indicate the ES cell clone as positive for carrying the targeted allele used to generate the Kv12.1^−/− mouse line. Note that DNA isolation failed for two clones (clones #1 and 3). The complete gels from which the “cut outs” in B were derived are provided in the source data. (C) ES clone #2 was karyotyped to confirm chromosome number and morphology and used for injection into C57BL/6J blastocysts. Three male chimeric mice were obtained, two of which transmitted the targeted allele through the germline. Mice carrying the targeted allele were bred with C57BL/6J-TgN(Zp3-Cre)93Knw females, which express Cre-recombinase in the germline (Kemler et al., 2004), to generate heterozygous Kv12.1^+/− animals, which were then bred to generate Kv12.1^−/− mice. For genotyping, a two-step PCR reaction was used with primers specific to the WT (sense: 5′-TGG​TCA​CAG​TGC​AGC​GGC​CAG​GGA​GTA-3′ and antisense: 5′-AAA​TTA​TTG​CGC​GGA​TGG​AAA​CAG​AGG​A-3′) and targeted (sense: 5′-GTC​ACA​GTG​CAG​CGG​CCA​GGG​AGT​AGC-3′ and antisense: 5′-CTT​GGC​GGT​CTG​GGT​GCC​CTC​GTA​GG-3′) alleles for the Kcnh8 gene. Bands at both 366 (WT) and 434 bp (targeted) identified heterozygous Kv12.1^+/− mice; bands at only 366 or 434 bp identified WT or homozygous Kv12.1^−/− mice, respectively. Source data are available for this figure: SourceData FS1.

Figure S2.

CX4-sensitive currents in WT SCN neurons and Cybercyte modeled I_Kv12. (A) Representative whole-cell Kv current recordings obtained from an SCN neuron in an acute slice prepared from a night-phased (ZT19–ZT24) WT SCN are shown. Whole-cell Kv currents, evoked during (2 s) voltage steps to potentials ranging from −100 to +50 mV (in 10 mV increments) from a holding potential of −70 mV, were first recorded in ACSF bath solution with 10 mM TEA and 10 mM 4-AP added (A1) and again following superfusion of the 10 mM TEA- and 10 mM 4-AP-containing ACSF with 20 μM CX4 added (A2). The voltage-clamp paradigm (in gray) is illustrated below the current records. Offline digital subtraction of the records obtained in the presence (A2), from the currents recorded in the absence (A1), of 20 μM CX4 provided the CX4-sensitive currents (A3). I_CX4 conductances at each test potential were calculated and normalized to the maximal conductance (G_max), determined in the same cell. (B) The mean ± SEM normalized conductances of activation of the CX4-sensitive currents are plotted as a function of the test potential and fitted with single Boltzmanns. The V_1/2 and the k values derived from these fits for current activation were V_1/2 = −4.9 ± 1.0 mV; k = 8.7 ± 1.9 (n = 21). (C) These parameters were used to tune the I_K12 model to fit the CX4-sensitive currents that were recorded during action potential–clamp experiments (see Fig. 4). The properties of currents produced by the Cybercyte I_Kv12 model (lower panel, red) reliably reproduce the CX4-sensitive currents measured in action potential–clamp recordings (middle panel, black) from WT SCN neurons.

Scheme 1

Scheme 2

Figure S3.

Cybercyte modeled I_A. A Markov model describing the gating of the K⁺ channels that generate I_A in WT SCN neurons was developed based on a previously described model of the rapidly activating and inactivating, I_A-like, K⁺ current in (ferret) ventricular myocytes (Campbell et al., 1993), and was populated using previously acquired voltage-clamp data detailing the time- and voltage-dependent properties of I_A in mouse SCN neurons (Hermanstyne et al., 2017). (A) Representative I_A waveforms recorded from a WT SCN neuron (Hermanstyne et al., 2017) in response to voltage steps to test potentials ranging from −40 to +40 mV (in 5 mV increments) from a HP of −70 mV are shown; the voltage-clamp paradigm is illustrated above the current records. (B) The voltage-dependences of activation and inactivation for I_A in WT SCN neurons (Hermanstyne et al., 2017) were determined using protocols identical to those described above for the CX4-senstive currents. The V_1/2 and the k values derived from these fits for current activation (open symbols) and inactivation (closed symbols) were V_1/2 = −9.3 ± 1.3 mV; k = 12.7 ± 0.8 (n = 12) and V_1/2 = −59.4 ± 2.1 mV, k = 7.9 ± 0.7 (n = 12), respectively. (C) These parameters were used to tune the model. (C) The waveforms of the currents produced by the Cybercyte I_A model reliably reproduced I_A recorded from WT SCN neurons (A).

Figure 1.

The repetitive firing rates of Kv12.1^−/− and Kv12.2^−/− SCN neurons are similar during the day and at night. (A and B) Representative daytime (A) and nighttime (B) whole-cell current-clamp recordings from WT (black squares), Kv12.1^−/− (red squares), and Kv12.2^−/− (blue squares) SCN neurons are shown. (C) Daytime and nighttime repetitive firing rates, measured in individual WT (black squares), Kv12.1^−/− (red squares), and Kv12.2^−/− (blue squares) SCN neurons are plotted; mean ± SEM repetitive firing rates are indicated; n = numbers of cells. As reported previously (Brown and Piggins, 2007; Allen et al., 2017; Belle and Allen, 2018; Harvey et al., 2020), the mean ± SEM repetitive firing rate in WT SCN neurons is higher (P = 0.0001, one-way ANOVA) during the day than at night (Table 1). In contrast, there are no day–night differences (one-way ANOVA) in the mean ± SEM repetitive firing rates of Kv12.1^−/− (red squares; P = 0.99) or Kv12.2^−/− (blue squares; P = 0.79) SCN neurons (Table 1). (D and E) Input resistances (D) and membrane potentials (E), measured in individual WT (black squares), Kv12.1^−/− (red squares), and Kv12.2^−/− (blue squares) SCN neurons during the day and at night, are plotted; mean ± SEM values and P values are indicated. In contrast to WT SCN neurons, there are no day–night differences (one-way ANOVA) in the mean ± SEM input resistances (D) or membrane potentials (E) of Kv12.1^−/− (red squares) or Kv12.2^−/− (blue squares) SCN neurons (Table 1).

Figure S4.

Repetitive firing rates in individual Kv12.1^−/−, Kv12.2^−/−, Kv12.1-targeted shRNA-expressing, and Kv12.2-targeted shRNA-expressing animals/slices. Comparisons of repetitive firing rates and membrane properties of WT (black squares), Kv12.1^−/− (red squares), Kv12.2^−/− (blue squares), NT shRNA-expressing (black circles), Kv12.1-targeted shRNA-expressing (red circles), and Kv12.2-targeted shRNA-expressing (blue circles) daytime (ZT 7–ZT 12; open symbols) and nighttime (ZT 18–ZT 24; closed symbols) SCN neurons; the individual data points represent the average values determined in slices from individual animals; N = number of animals. (A–C) Repetitive firing rates (A), input resistances (B), and membrane potentials (C) were measured in WT, Kv12.1^−/−, and Kv12.2^−/− SCN neurons in slices prepared during the day (open symbols: N = 9 for WT [black square]; N = 6 for Kv12.1^−/− [red square]; N = 4 for Kv12.2^−/− [blue square]) or at night (closed symbols: N = 5 for WT [black square]; N = 3 for Kv12.1^−/− [red square]; N = 4 for Kv12.2^−/− [blue square]) and average values (in individual slices/animals) are plotted. (D–F) Average daytime and nighttime repetitive firing rates (D), input resistances (E), and membrane potentials (F), determined in daytime (open symbols) NT shRNA- (black circles; N = 4), Kv12.1-targeted shRNA- (red circles; N = 3), Kv12.2-targeted shRNA- (blue circles; N = 3), and nighttime (closed symbols) NT shRNA- (black circles; N = 4), Kv12.1-targeted shRNA- (red circles; N = 3), and Kv12.2-targeted shRNA- (blue circles; N = 3) expressing SCN neurons are presented. In A–F, mean ± SEM values are plotted and P values are indicated. All data are also tabulated in Table S1.

Figure S5.

Repetitive firing rates of Kv12.1^−/− and Kv12.2^−/− SCN neurons are higher than in WT SCN neurons throughout the night. Cell-attached voltage recordings were obtained for WT, Kv12.1^−/−, and Kv12.2^−/− SCN neurons in acute slices prepared at various times throughout the circadian cycle. As is evident, mean ± SEM peak firing rates were high throughout the day (yellow shaded region) in WT (black circles) SCN neurons (n = 132), subsequently decreased during the transition from day to night, and remained low throughout the night. Conversely, the mean ± SEM peak firing rates of Kv12.1^−/− (red circles; n = 121) and Kv12.2^−/− (blue circles; n = 125) SCN neurons did not vary measurably over time and were consistently high (similar to the daytime repetitive firing rates of WT SCN neurons) throughout the day and night.

Figure 2.

Validation of the Kv12.1-targeted and Kv12.2-targeted shRNAs. To examine the specificity of the selected Kv12.1-targeted and Kv12.2-targeted shRNAs, each was expressed in tsA-201 cells with a cDNA construct encoding Kv12.1-eYFP, Kv12.2-eYFP, or Kv4.1-eYFP; parallel experiments were completed with the NT shRNA. Approximately 24 h following transfections, cell lysates were prepared, fractionated by SDS-PAGE, transferred to PVDF membranes, and probed with an anti-GFP antibody. All blots were also subsequently probed with an anti-α-tubulin antibody to verify equal protein loading. (A) Co-expression with the Kv12.1-targeted shRNA markedly reduced Kv12.1-eYFP protein expression, whereas the Kv12.2-targeted shRNA and the NT shRNA were without effects on Kv12.1-eYFP. (B) Similarly, co-expression with the Kv12.2-targeted shRNA, but not the Kv12.1-targeted or the NT shRNA, reduced Kv12.2-eYFP protein expression. (C) In contrast, neither the Kv12.1-targeted nor the Kv12.2-targeted shRNA measurably affected expression of Kv4.1-eYFP. The complete blots from which the “cut outs” shown in A–C were derived are provided in the source data. (D) eGFP-expressing neurons were readily identified in acute SCN slices prepared 2 wk following bilateral injections of the Kv12.1-targeted (or the Kv12.2-targeted) shRNA-expressing AAV8 into the SCN; scale bar = 250 μm. Source data are available for this figure: SourceData F2.

Figure 3.

Acute in vivo shRNA-mediated knockdown of Kv12.1 or Kv12.2 increases the nighttime, but not the daytime, repetitive firing rates of SCN neurons. (A and B) Representative daytime (A) and nighttime (B) whole-cell current-clamp recordings obtained from NT shRNA- (black circles), Kv12.1-targeted shRNA- (red circles), and Kv12.2-targeted shRNA- (blue circles) expressing SCN neurons during the day (ZT7–ZT12) and at night (ZT18–ZT24) are shown. (C) Daytime and nighttime repetitive firing rates measured in individual SCN neurons expressing the NT (black circles), Kv12.1-targeted (red circles), or Kv12.2-targeted shRNA (blue circles) are plotted; mean ± SEM repetitive firing rates are indicated. Similar to WT SCN^[7–10] neurons (Fig. 1), the mean ± SEM repetitive firing rate of NT shRNA-expressing (black circles) SCN neurons was lower (P = 0.0001, one-way ANOVA) at night than during the day (Table 2). In marked contrast, the mean ± SEM repetitive firing rates of Kv12.1- (red circles) and Kv12.2- (blue circles) targeted shRNA-expressing SCN neurons at night are not (one-way ANOVA) different from daytime firing rates (Table 2). (D and E) The input resistances (D) and membrane potentials (E) of NT shRNA- (black circles), Kv12.1-targeted shRNA- (red circles), and Kv12.2-targeted shRNA-expressing (blue circles) SCN neurons during the day and at night were also determined. Similar to WT SCN neurons^[7–10], the mean ± SEM input resistance of NT shRNA-expressing (black circles) SCN neurons was higher (P = 0.0001, one-way ANOVA) during the day than at night, and the mean ± SEM membrane potential was more depolarized (P = 0.004, one-way ANOVA), during the day than at night (Table 2). In contrast, there were no day–night differences (one-way ANOVA) in the mean ± SEM input resistances (D) or membrane potentials (E) of Kv12.1-targeted shRNA- (red circles) and Kv12.2-targeted shRNA-expressing (blue circles) SCN neurons (Table 2).

Figure 4.

CX4-sensitive K⁺ current densities in SCN neurons are higher at night than during the day. (A) Whole-cell outward K⁺ currents, evoked by an action potential voltage command recorded from a nighttime WT SCN neuron, were recorded in WT and DKO (Kv12.1^−/−/Kv12.2^−/−) SCN neurons before and after local application 20 μM CX4. Subtraction of the currents recorded before/after application of CX4 in each cell provided the CX4-sensitive K⁺ currents. (B and C) Representative CX4-sensitive outward K⁺ currents, obtained by offline digital subtraction of the currents recorded in the presence of CX4 from the currents recorded (in the same cell) in the control bath solution during the day and at night in WT (B) and DKO (C) SCN neurons, are shown. The CX4-sensitive outward K⁺ currents recorded in WT daytime and nighttime SCN neurons are also displayed on an expanded time scale in the insets in B. (D) Peak CX4-sensitive K⁺ current densities measured in WT (black squares) and DKO (magenta squares) SCN neurons during the day and at night are plotted. The mean ± SEM CX4-sensitive K⁺ current density in WT (black squares) SCN neurons is much higher (P = 0.002, one-way ANOVA) at night (n = 13) than during the day (n = 17). In contrast, CX4-sensitive K⁺ current densities in DKO (magenta squares) SCN neurons are low during the day (n = 8) and at night (n = 8), and there is no day–night difference (P = 0.85, one-way ANOVA) in mean ± SEM CX4-sensitive K⁺ current densities in DKO (magenta squares) SCN neurons.

Figure 5.

Pairwise comparisons of repetitive firing rates in WT SCN neurons in the absence and presence of CX4. (A and B) Representative daytime (A) and nighttime (B) whole-cell current-clamp recordings obtained from WT SCN neurons before and after application of 20 μM CX4 are shown. (C and D) The repetitive firing rates measured in individual daytime (C) and nighttime (D) WT SCN neurons before and after application of 20 μM CX4 are plotted. In the vast majority (15 of 17; ∼90%) of WT SCN neurons, daytime repetitive firing rates were not affected (P = 0.6, repeated measures t test) by the application of 20 μM CX4 (C). In contrast, nighttime repetitive firing rates were increased (P = 0.001, repeated-measures Student’s t test) in most (11 of 15; ∼70%) WT SCN neurons upon application of 20 μM CX4 (D). Approximately 30% (4 of 15) of nighttime WT SCN neurons, however, were not measurably affected by 20 μM CX4.

Figure 6.

Dynamic clamp-mediated subtraction of I_Kv12 increases the repetitive firing rates of nighttime WT SCN neurons. (A–C, and E–G) Representative whole-cell current-clamp recordings from a WT nighttime (A–C) and a WT daytime (E–G) SCN neuron at baseline (A and E) and with different magnitudes of modeled I_Kv12 subtracted (B, C, F, and G) via dynamic clamp are shown. In the insets below C and G, the waveforms of individual action potentials (black) recorded in the representative WT nighttime (C) and daytime (G) SCN neurons are displayed and plotted on an expanded timescale. Below these voltage records, the modeled (−3x) I_Kv12 waveforms (purple) are shown; the zero current levels are indicated by the dotted lines. As is evident, the effect of subtracting increasing I_Kv12 is greater in the nighttime (B and C), than in the daytime (F and G), WT SCN neuron. Similar results were obtained in current-clamp recordings from additional WT nighttime and daytime SCN neurons in which varying amplitudes (−x, −2x, −3x, −4x, and −5x) of the minimal modeled I_Kv12 (0.5 pA) were subtracted via dynamic clamp. (D and H) The mean ± SEM percent changes in the spontaneous repetitive firing rates of WT daytime (empty square; n = 10) and nighttime (filled square; n = 7) SCN neurons are plotted as a function of the magnitude of modeled I_Kv12 subtracted.

Figure 7.

Dynamic clamp-mediated addition of modeled Kv12-encoded (I_Kv12) currents in WT SCN neurons. (A–C, and E–G) Representative whole-cell current-clamp recordings from a WT nighttime (A–C) and a WT daytime (E−G) SCN neuron at baseline (A and E) and with different magnitudes of modeled I_Kv12 added (B, C, F, and G) via dynamic clamp are shown. In the insets below C and G, the waveforms of individual action potentials (black) recorded in the representative WT nighttime (C) and daytime (G) SCN neurons are displayed and plotted on an expanded timescale. Below these voltage records, the modeled (+16x) I_Kv12 waveforms (purple) are shown; the dotted lines indicate the zero current levels. As is evident, increasing I_Kv12 reduced the rate of repetitive firing in both the WT nighttime (B and C) and the WT daytime (F and G) SCN neuron, although the impact is greater on nighttime (B and C) than on daytime (F and G) firing. Similar results were obtained in recordings from additional WT nighttime and daytime SCN neurons in which varying amplitudes (+x, +2x, +4x, +8x, and +16x) of the minimal modeled I_Kv12 (2 pA) were added during current-clamp recordings. (D and H) The mean ± SEM percent changes in the spontaneous repetitive firing rates of WT daytime (empty circles, n = 7) and nighttime (filled circles, n = 10) SCN neurons are plotted as a function of the magnitude of modeled I_Kv12 added.

Figure S6.

Dynamic clamp-mediated addition of I_Kv12 decreases the rate of repetitive firing in nighttime Kv12.1^−/− SCN neurons. (A–C) Representative whole-cell current-clamp recordings obtained from a nighttime (ZT18–ZT20) Kv12.1^−/− SCN neuron under control conditions (A) and with dynamic clamp-mediated addition of modeled I_Kv12 (B and C). Similar to WT SCN neurons (see Fig. 7), the addition of modeled I_Kv12 (B and C) reduced the repetitive firing rates of Kv12.1^−/− SCN neurons in direct proportion to the amplitude of the injected current. (D) The mean ± SEM (n = 11) percentage changes in the repetitive firing rates of nighttime Kv12.1^−/− SCN neurons (red circles) are plotted as a function of the amplitude of I_Kv12 added. The percent changes in the repetitive firing rates of WT neurons with the addition of I_Kv12 (from Fig. 7 D) are replotted here (black circles) to facilitate direct comparison of the results in WT and Kv12.1^−/− SCN neurons.

Figure S7.

Kinetic differences between the Cybercyte modeled I_A and I_Kv12. Comparison of the normalized currents produced by the Cybercyte I_Kv12 (lower panel, purple) and I_A (lower panel, blue) models generated by action potential waveforms (upper panel, black) from a representative WT SCN neuron.

Figure 8.

Dynamic clamp-mediated subtraction of modeled I_A in WT SCN neurons. (A–D) Representative whole-cell current-clamp recordings from WT nighttime (A and C) and daytime (B and D) SCN neurons with different magnitudes of modeled I_A subtracted are shown. In the insets below C and D, the waveforms of individual action potentials (black), recorded in the representative WT nighttime (C) and daytime (D) SCN neurons, are plotted on an expanded timescale. The modeled (−3x) I_Kv12 waveforms (blue) generated in these cells are shown; the zero current levels are indicated by the dotted lines. As is evident, the rates of repetitive firing of the representative nighttime (A and C) and daytime (B and D) SCN neurons are increased with the subtraction of modeled I_A. Similar results were obtained in recordings from additional nighttime and daytime WT SCN in which varying amplitudes of the minimal modeled I_A (20 pA) were subtracted (−x, −2x, −3x, −4x, and −5x) during current-clamp recordings. (E and F) The mean ± SEM percent changes in the spontaneous repetitive firing rates of WT SCN neurons in response to subtracting (open and closed squares; n = 16) modeled I_A are plotted as a function of the magnitude of the modeled I_A subtracted.

Figure 9.

Dynamic clamp-mediated addition of modeled I_A in WT SCN neurons. (A–D) Representative whole-cell current-clamp recordings from WT nighttime (A and C) and daytime (B and D) SCN neurons with different magnitudes of modeled I_A added are shown. In the insets below C and D, the waveforms of individual action potentials (black), recorded in the representative WT nighttime (C) and daytime (D) SCN neurons, are plotted on an expanded timescale. The modeled (+3x) I_Kv12 waveforms (blue) generated in these cells are shown; the zero current levels are indicated by the dotted lines. The rates of repetitive firing of both the representative nighttime (A and C) and the representative daytime (B and D) SCN neurons are decreased with the addition of modeled I_A. Recordings from additional nighttime and daytime WT SCN neurons in which varying amplitudes (x, 2x, 3x, 4x, 5x) of the minimal modeled I_A (20 pA) were added also revealed a decrease in repetitive firing rates. (E–H) The mean ± SEM percent changes in the spontaneous repetitive firing rates of WT SCN neurons in response to adding (open and closed circles; n = 14) modeled I_A are plotted as a function of the magnitude of the modeled I_A added.

Figure S8.

Direct comparison of the effects of dynamic clamp-mediated subtraction of I_Kv12 versus I_A in daytime WT SCN neurons. (A, B, D, and E) Representative whole-cell current-clamp recordings were obtained from a day-phased (ZT7–ZT12) WT SCN neuron with dynamic clamp-mediated subtraction of modeled I_A (A and B) or I_Kv12 (D and E). Subtracting modeled 1x or 3x I_A increased the rate of repetitive firing (A and B), whereas subtracting modeled 1x or 3x I_Kv12 (D and E) in the same cell did not measurably alter the firing rate. In the insets below B and E, the waveforms of individual action potentials (black), recorded in a representative WT daytime SCN neuron with subtracted modeled (−3x) I_A (B) or (−3x) I_Kv12 (E), are plotted on an expanded timescale. The modeled (−3x) I_A (blue, B) and (−3x) I_Kv12 (purple, E) waveforms in these cells are shown; the zero current levels are indicated by the dotted lines. Similar results were obtained in six additional cells. (C and F) Plotting the mean ± SEM (n = 7) percentage changes in the repetitive firing rates of daytime WT SCN neurons reveals the mark difference in the effects of subtracting I_A (C) versus I_Kv12 (F).

Figure S9.

Expression of Kcnh8 and Kcnh3 in the SCN. The expression levels of the transcripts encoding Per2, Bmal, Kcnh8, and Kcnh3 were determined by quantitative RT-PCR analyses of SCN tissues samples, obtained every 4 h over a 48 h time-period, from animals maintained in the standard and reversed 12:12 h LD conditions, as described in Materials and methods. The expression of each transcript was normalized to the expression of Hprt in the same sample. Mean ± SEM (n = 7–8) values for each transcript are plotted. Fitting the mean data for each transcript, using JTK cycle analysis (Hughes et al., 2010) with the period set to 24 h, reveals that, as expected (Kuhlman and McMahon, 2006; Hastings et al., 2019), the Per2 (P = 0.0001) and Bmal (P = 0.0001) transcripts display 24 rhythms in expression. In contrast, the expression levels of the Kcnh8 (P = 1.00) and Kcnh3 (P = 0.102) transcripts do not display 24 h rhythms. The dashed lines on days 1 and 2 are at 7 pm.

Figure 10.

Rhythmic wheel-running activity is maintained in Kv12.1^−/− and Kv12.2^−/− mice and with acute in vivo shRNA-mediated knockdown of Kv12.1 or Kv12.2 in adult animals. (A and B) Representative recordings of wheel-running activity of WT (top), Kv12.1^−/− (middle), and Kv12.2^−/− (bottom) mice (A) and WT mice that received bilateral SCN injections of the non-targeted (top), Kv12.1-targeted (middle), or Kv12.2-targeted (bottom) shRNA-expressing AAV8 (B). Continuous recordings were obtained for ∼10 d in 12:12 h LD (indicated by the grey and white backgrounds, respectively) conditions, followed by at least 20 d in DD (constant darkness, indicated by the white background). The circadian periods, measured in each WT (black squares), Kv12.1^−/− (red squares), and Kv12.2^−/− (blue squares) mouse and in each animal expressing the NT (black circles), Kv12.1-targeted (red circles) and Kv12.2-targeted (blue circles) shRNA in DD, are plotted; P values (one-way ANOVA are indicated). (C) The mean ± SEM circadian periods of locomotor activity determined in WT (black squares; n = 9), Kv12.1^−/− (red squares; n = 15) and Kv12.2^−/− (blue squares; n = 14) are similar. (D) The mean ± SEM circadian periods (D) of locomotor activity measured in animals expressing the Kv12.1-targeted shRNA (red circles; n = 7), the Kv12.2-targeted shRNA (blue circles; n = 7) and the NT shRNA (black circles; n = 11) are also similar.