Supplementary Figures 1-11 Supplementary Figure 1: Kv7 currents in neonatal CA1 neurons measured with the classic M- current voltage-clamp protocol. (a), Voltage-clamp recordings from CA1 pyramidal neurons of P7 control and mutant mice under control conditions (black) and after application of 10 μm XE991 (XE) (gray). Neurons were held at 25 mv and hyperpolarized to 45 mv for 1 s. The red current trace shows the resulting difference current that represents the Kv7 current (controls and mutants n = 10 cells from 3 mice). (b), Quantification of voltage-clamp experiments shown in (a) (*** P < 0.001; unpaired t-test).
Supplementary Figure 2: Analysis of spontaneous, giant depolarizing potentials (GDPs) recorded from CA3 neurons in acute slices from P7 mice. (a), Example traces of spontaneous activity and (b), quantification of GDP occurrence for baseline (ACSF) recordings from control or mutant CA3 neurons. (c), Example traces and (d), quantification of relative GDP occurrence in control or mutant CA3 neurons in ACSF or ACSF+XE991 (XE), (controls and mutants, n = 7 animals each for both conditions; * P < 0.05, Newman-Keuls post-hoc after 2-way ANOVA with genotype and treatment as between group factors.)
Supplementary Figure 3: Spontaneous network activity in the isolated hippocampus in vitro. (a), Spontaneous oscillations in hippocampi from P13 mutants and controls. Examples of the local field potential recorded in hippocampi isolated from a control (top trace) or mutant (bottom trace) mouse. (b), After 20 minutes baseline recording, instantaneous frequency was estimated for 40 oscillation cycles using ClampFit (n = 18 hippocampi (9 mice)) and mutants (n = 8 hippocampi (4 mice), values shown as mean ± standard deviation; P = 0.35, unpaired t-test.)
Supplementary Figure 4: Visualization of depth-electrode recording sites. (a), Coronal, NeuroTrace-stained and corresponding atlas sections 63 through primary visual (V1) cortex and (b), hippocampus of a P7 neonate following acute in vivo silicone probe depth-recordings (Figs. 2,4; Supplementary Figs. S5, S7). The electrode tracks were visualized by DiI staining (red).
Supplementary Figure 5: Spontaneous in vivo hippocampal LFPs in P5-7 mutants. (a), Example LFP traces, CSD, and pseudocolor CSD from an acute silicon probe depth-profile recording (50 µm inter-site spacing) in the hippocampus of a P7 control during sharp waves (SPWs). (b), Extended example of a 1.2 s LFP epoch overlaid on its CSD pseudocolor representation during SPWs and hippocampal beta oscillations
(HBOs). (c), Spectrogram of LFP channel 9 (highlighted in (b)) from same epoch as (b). (d), Example LFP traces, CSD, and pseudocolor CSD from a recording in the hippocampus of a P7 mutant during SPWs. (e), Extended example of a 1.4 s LFP epoch overlaid on its CSD pseudocolor representation, with same time axis as in (c). (f), Spectrogram of LFP channel 13 as highlighted in (e) from same epoch as (e). (g), Top, cross-correlogram (CCG) of SPW with HBO events, left; auto-correlogram (ACG) of HBO events, right (ACGs and CCGs were pooled over all (n = 13 control and n = 15 mutant) recordings because there were no differences between genotypes.); Middle, CCGs of movement with SPW and HBO events indicate that movement preceded both event types, and that HBOs were mildly inhibited 2-3 s after twitches; Bottom, ACGs of twitches and hippocampal events show SPWs often occur in pairs or bursts. (h), Quantification of SPW and HBO properties indicate no differences between genotypes (n = 13 controls and n = 15 mutants; unpaired t-tests). Supplementary Figure 6: Spontaneous in vivo mutant P10-11 hippocampal spiking exhibits burstiness. Histograms and boxplots appear as described in Figure 2c. (a), Firing rate distributions for mutant and control neurons (P = 0.55). (b), Fraction of inter-spike-intervals (ISIs) between 2 and 15 ms in mutants vs. controls. (c), Distribution of the percentage of mutant or control hippocampal neurons spikes which occurred in groups of 4 within 60 ms windows, termed quadruplets, or as 5 within 75 ms, quintuplets. (n = 151 units from 8 control mice; n = 117 units from 5 mutant mice; *** P < 0.001; n.s., not significant; Mann-Whitney U test).
Supplementary Figure 7: Spontaneous in vivo V1 cortical network activity at P5-7. (a), Spectrogram of the CSD constructed from channels ~150 um deep in an acute silicone probe depth-profile recording in V1 cortex of a P7 mutant. (b), Spectra of channels approx. 100 µm, 200 µm, and 300 µm deep. (c), LFP traces of the top 8 recording channels (50-400 µm deep) display spontaneous bursts of spindle-like oscillations in the low-beta frequency range highlighted at 6 s in (a), and shown in (b). This event spanned deeper layers. (d), Spindle burst lengths and spectral peak frequency did not differ between genotypes (n = 13 controls and n = 15 mutants.) (e), Example of the two types of slow activity transients (SATs) observed in P5-7 mice. The rate of SAT events, with the fraction of type 1 SATs indicated below the gray divider and type 2 above (n = 12 controls and n = 14 mutants for SAT quantifications). (f), There were no strong cross-correlations between cortical events and twitches and no differences between genotypes (CCGs shown here from pooled data); however, during movement there was a small decrease in the ability to detect SAT events. (g), Quantification of rate and length of twitch episodes in controls (n = 13) and mutants (n = 15) showed no differences. ((d),(e),(g): unpaired t-tests).
Supplementary Figure 8: Bumetanide normalizes mutant hippocampal burstiness without affecting controls. Histograms and boxplots appear as described in Figure 2c. (a), Left, firing rate distribution of neurons from vehicle (1.2 ± 0.16 Hz) vs. bumetanide treated (1.06 ± 0.19 Hz) mutants (P = 0.047); middle controls were not significantly affected by treatment, and bumetanide-treated mutants did not significantly differ from vehicle-treated controls. (b), The distribution of bumetanide-treated mutants hippocampal neurons fraction of short inter-spike-intervals (between 2 and 15 ms) significantly differed from vehicle-treated mutants, and was not significantly different from vehicle-treated controls. The fraction of spikes in quadruplets (4 spikes within a 60 ms window) (c) or quintuplets (5 within 75 ms) (d) in bumetanide-treated mutants was significantly different from vehicle-treated mutants but not significantly different from vehicle-treated controls. (n = 117 units from 5 vehicle-treated mutant mice; n = 145 units from 8 bumetanide-treated mutant mice; n = 151 units from 8 vehicle-treated control mice, n = 186 units from 7 bumetanide-treated control mice; * P < 0.05; *** P < 0.001; n.s., not significant; Kruskal-Wallis multiplecomparisons test after Kruskal-Wallis one-way ANOVA.)
Supplementary Figure 9: LFP genotype differences in twice-daily treated groups. (a), Quantification of cortical spindle burst (SB) length, (b), hippocampal beta oscillation (HBO) occurrence rate, and (c), mean HBO length in in vivo recordings from vehicle- or bumetanide-treated controls or mutants (n = 8 control vehicle, n = 11 control bumetanide, and n = 9 each for mutant vehicle and bumetanide-treated mice). Significant genotype effects were observed (* P < 0.05, ** P < 0.01, by two-way ANOVA having genotype and treatment as between group factors.)
Supplementary Figure 10: Partial behavioral improvement by picrotoxin but not levetiracetam (a), Nissl-stained coronal brain sections from the dorsal hippocampal area of adult vehicle-treated control and mutant mice, as well as picrotoxin (0.1 mg/kg once a day, sub-epileptic dosage) and levetiracetam (titrated up to 100 mg/kg split in 2 doses/day) treated mutants (top to bottom.) Both picrotoxin and levetiracetam administration during the first two postnatal weeks modestly attenuated Kv7-dysfunction-mediated neuronal dispersion in the adult hippocampal CA1 layer. (b), Four example animals 15-minute open-field trials indicate hyperactivity in vehicle and levetiracetam-treated mutants, which was attenuated by once-daily 0.1 mg/kg picrotoxin. (c), Picrotoxin significantly reduced Kv7DN mutant hyperactivity (n = 26 vehicle-treated controls, n = 22 vehicletreated mutants, n = 18 picrotoxin-treated mutants, n = 7 levetiracetam-treated mutants; *** P < 0.001; KruskalWallis multiple comparisons test after one-way ANOVA.)
Supplementary Figure 11: Intermediate improvement by single daily bumetanide treatment. (a to c), The hippocampal sections were stained as in Figs. 2e and 6a, and represent adjacent sections from the same animal. Nissl- (a), GFAP- (b), or isolectin-b4-stained (c) coronal sections through the hippocampal region of a mutant adult mouse following single daily neonatal (P0-P14) bumetanide treatment shows absence of CA1 pyramidal layer dispersion but signs of inflammation, although reduced in intensity (n = 3). Scale bars, 500 µm.