LH pulse and surge profiles in mice with associated firing activity of Kiss1ARH neurons.

A, Representative example of an LH pulse profile in an ovariectomized (OVX) mouse in the absence of gonadal steroid feedback. Pulses detected by the DynPeak algorithm are indicated with an asterisk (Lin et al., 2021). B, LH surge profile (closed symbol) in mice undergoing the OVX+E2+E2 surge inducing protocol (6-10 days post OVX and implantation of 17β-estradiol capsule delivering diestrous levels of steroid, mice received subcutaneous injections of estradiol benzoate (1µg/20g body weight) on 2 consecutive days at 08:30 h with blood samples collected every 30 min from 14:30 - 20:30 h for LH measurement (Lin et al., 2021); mean ± SEM; (n=9). Open symbol representing control mice not receiving the second dose of estradiol benzoate (OVX+E2+oil). The expected LH surge occurred approximately 1 h before lights off (19:00 h). Values significantly different from basal LH concentrations are indicated by # (#p<0.05, ##p<0.01, repeated measures ANOVA). C, illustration of synchronized firing of a Kiss1ARH neuron induced by TACR3 agonist senktide in an OVX female. D, demonstration of burst firing of a Kiss1ARH neuron induced by glutamate in an E2-treated, OVX female. The spike activities in C and D have been expanded to emphasize the notable effects of senktide or glutamate on the firing activity of Kiss1ARH neurons.

Relative contribution of voltage-gated calcium currents in Kiss1ARH neurons from OVX mice.

A-E, representative current –voltage relationships showing that Cd2+ (non-selective blocker of calcium channels) sensitive peak currents were inhibited by different calcium channel blockers: A, nifedipine; B, ω-conotoxin GIVA; C, ω-agatoxin IVA; D, SNX-482; E, TTA-P2. F, the maximum peak currents were measured at -10 mV. The proportions of Ca2+ currents inhibited by nifidipine (L type), ω-conotoxin GVIA (N type), ω-agatoxin IVA (P/Q), SNX-482 (R type) and TTA-P2 (T type). Data are expressed as mean ± SEM, n = cell numbers.

Blockade of HVA Ca2+ channels decreases the slow EPSP in Kiss1ARH neurons.

A-C, representative traces showing that the slow EPSPs were abolished by perfusing the blocker of the L-type calcium channel, nifedipine (A) or N- and P/Q -type calcium channels, ω-conotoxin MVIIC (B), or the R-type calcium channel, SNX 482 (C), respectively. The arrows indicate the measurements of slow EPSP amplitude, denoted as R1 and R2, after low-pass filtering. D. Bar graphs summarizing the effects of drugs on the R2/R1 ratios. The slow EPSP was generated in OVX Kiss1-Cre::Ai32 mice. Comparisons between different treatments were performed using a one-way ANOVA analysis (F (3, 44) = 19.72, p<0.0001) with the Bonferroni’s post hoc test. **, **** indicates p<0.01, 0.001, respectively vs. control.

Primer Table

A, E2 increases the expression of low and high voltage-activated calcium channels in Kiss1ARH neurons.

Kiss1ARH neurons (three to four 10-cell pools) were harvested from each of 5 vehicle- and 5 E2-treated, OVX females to quantify ion channel mRNA expression of low and high voltage activated calcium channels as described in the Methods. The analysis included: T-type (Cav3.1) low voltage-activated, as well as the following high-voltage activated channels: R-type (Cav 2.3), L-type (Cav 1.2), N-type (Cav 2.2) and P/Q-type (Cav 2.1) calcium channels. Interestingly, all of these channels were upregulated with E2 treatment, which significantly increased the whole-cell calcium current (see Figure 5). B, E2 also increased the expression of hyperpolarization-activated, cyclic-nucleotide gated HCN1 and HCN2 channels in Kiss1ARH neurons. The same Kiss1ARH neuronal pools were also analyzed for mRNA expression of HCN1 and HCN2 ion channels. HCN1 channel mRNA expression was the most highly upregulated by E2 treatment in Kiss1ARH neurons; although, the mRNA expression of HCN2 was also significantly increased. The expression values were calculated via the ΔΔCT method, normalized to GAPDH and relative to the oil control values. Bar graphs represent the mean ± SEM. * p < 0.05, **p < 0.01, ***p < 0.005, ****p < 0.001, oil versus E2. C, in addition, we have shown previously that E2-treatment also increases the associated T- and h-currents (not shown), as well as the neuronal excitability (measured as rebound excitation) in Kiss1ARH neurons. The left panel illustrates an example of rebound burst firing in Kiss1ARH neurons, while the right panel displays bar graphs representing the mean ± SEM. ***p < 0.005 (Qiu J. et al., 2018).

E2 treatment (positive-feedback regimen) increases the Ca2+ currents in Kiss1ARH neurons.

A-B, Ca2+ currents in Kiss1ARH neurons with the same membrane capacitance from oil-treated (A) or E2-treated (B) animals. C, the maximum peak currents were measured at -10 mV. The current amplitudes were normalized to the cell capacitance in all cases to calculate current density. The bar graphs summarized the density of Ca2+ current in Kiss1ARH neurons from oil-treated and E2-treated animals. The mean density was significantly greater in E2-treated (13.4 ± 0.9 pA/pF, n = 11) than in oil-treated OVX females (7.2 ± 0.5 pA/pF, n = 40) (unpaired two-tailed t-test, t(49) = 5.75, ****p < 0.0001). D. The modeling predicts that E2-treated, OVX females exhibit a significantly greater inward Ca2+ current (red trace) than the vehicle-treated females (black trace). The green arrow in the red trace indicates the T-channel “inflection.” E. Relative contribution of voltage-gated calcium currents in Kiss1ARH neurons from OVX, E2-treated mice. The maximum peak currents were measured at -10 mV. The proportions of Ca2+ currents inhibited by nifidipine (L type), ω-Conotoxin GVIA (N type), ω-agatoxin IVA (P/Q), SNX-482 (R type) and TTA-P2 (T type). Data are expressed as mean ± SEM, n = cell numbers.

Voltage dependence of ICa in Kiss1ARH neurons from OVX and OVX+E2 mice.

A-B, top panels: activation and inactivation protocol. Bottom: representative traces. C-D, the mean V1/2 values for calcium channel activation were not significantly different for cells from controls versus cells from estrogen-treated females. Similarly, the V1/2 values for channel steady-state inactivation were similar for both groups.

Small conductance, calcium-activated K+ (SK) channel is involved in repolarization of burst firing Kiss1ARH neurons in OVX and E2-treated, OVX mice.

A. Representative traces of the inhibition of outward currents before (left, control) and after the specific SK blocker Apamin (500 nM, middle). Apamin sensitive currents were calculated from the subtraction of control and apamin at depolarized potentials (right). Cells were clamped at -70 mV and given 500 ms voltage pulses from - 60 mV to +40 mV in 10 mV steps at 0.2 Hz, as shown in A at the bottom. B. Mean current density-voltage relationships measured at the end of the 500 ms voltage step ranging from -60 mV to +40 mV were obtained in the absence and presence of apamin (two-way ANOVA: main effect of treatment (F(1, 4) = 7.697, p = 0.0501), main effect of time (F(10, 40) = 99.3, p < 0.0001) and interaction (F(10, 40) = 7.645, p < 0.0001); mean ± SEM; n = 3; post hoc Bonferroni test, **p < 0.01, ***p < 0.005, ****p < 0.001). C. Apamin sensitive current densities were obtained from C (mean ± SEM, n = 3). D. Representative traces of the inhibition of outward currents before (left, control) and after the specific SK blocker Apamin (500 nM, middle). Apamin sensitive currents were resulted from the subtraction of control and apamin at depolarized potentials (right). E. Mean current density-voltage relationships measured at the end of the 500 ms voltage step ranging from -60 mV to +40 mV were obtained in the absence and presence of apamin (two-way ANOVA: main effect of treatment (F(1, 8) = 9.433, p = 0.0153), main effect of time (F(10, 80) = 184.9, p < 0.0001) and interaction (F(10, 80) = 8.791, p < 0.0001); mean ± SEM, n = 4; post hoc Bonferroni test, *p < 0.05, ***p < 0.005, ****p < 0.001). F. Apamin sensitive current densities were obtained from C and E (ns; Two-way ANOVA followed by Bonferroni post hoc test; mean ± SEM; OVX, n = 3; OVX+E2, n = 4). G. Kiss1ARH neurons (three to four 10-cell pools) were harvested from each of 5 vehicle- and 5 E2-treated, OVX females to quantify the mRNA expression of SK3 ion channel. E2 did not increase the mRNA expression small conductance calcium-activated K+ (SK3) channels in Kiss1ARH. The expression values were calculated via the ΔΔCT method, normalized to GAPDH and relative to the oil control values. Bar graphs represent the mean ± SEM (unpaired two-tailed t-test for SK3, ns). H. The mathematical model was calibrated on the electrophysiology data from Kiss1ARH neurons for E2-treated females before and after treatment with the specific SK blocker apamin, left panel versus middle panel respectively (see Table S1 for gSK). The modeled apamin-sensitive current (right panel) matches the electrophysiological data. For the calibration it was assumed that the applied concentration of apamin (500nM) completely blocked the SK current.

Large conductance, calcium-activated K+ (BK) channels contributes to the repolarization of Kiss1ARH neurons in OVX and E2-treated, OVX mice.

A. Representative traces of the inhibition of outward currents before (left, control) and after the specific BK blocker iberiotoxin (IbTx; 200 nM, middle). IbTx sensitive currents were calculated from the subtraction of control and IbTx at depolarized potentials (right). Cells were clamped at -70 mV and given 500 ms voltage pulses from -60 mV to +40 mV in 10 mV steps at 0.2 Hz, as shown in A at the bottom. B. Mean current density-voltage relationships measured at the end of the 500 ms voltage step ranging from -60 mV to +40 mV were obtained in the absence and presence of IbTx (two-way ANOVA: main effect of treatment (F(1, 8) = 0.8841, p = 0.3746), main effect of time (F(10, 80) = 71.56), p < 0.0001) and interaction (F(10, 80) = 1.127, p = 0.3528); mean ± SEM, n = 5; post hoc Bonferroni test, p > 0.05). C. IbTX sensitive current densities were obtained from B (mean ± SEM, n = 5). D. Representative traces of the inhibition of outward currents before (left,control) and after the specific BK blocker iberiotoxin (IbTx; 200 nM, middle). IbTx sensitive currents were resulted from the subtraction of control and IbTx at depolarized potentials (right). E. Mean current density-voltage relationships measured at the end of the 500 ms voltage step ranging from -60 mV to +40 mV were obtained in the absence and presence of IbTX (two-way ANOVA: main effect of treatment (F(1, 6) = 3.181, p = 0.1248), main effect of time (F(10, 60) = 52.90, p < 0.0001) and interaction (F(10, 60) = 3.667, p = 0.0007); mean ± SEM, n = 4; post hoc Bonferroni test, *p < 0.05, **p < 0.01). F. IbTx sensitive current densities were obtained from C and E (two-way ANOVA: main effect of treatment (F(1, 7) = 31.63, p = 0.0008), main effect of time (F(10, 70) = 80.41, p < 0.0001) and interaction (F(10, 70) = 21.54, p <0.0001); mean ± SEM, OVX, n = 5; OVX+E2, n = 4; Bonferroni post hoc test, **p < 0.01, ****p < 0.001). G. Kiss1ARH neurons (three to four 10-cell pools) were harvested from each of 5 vehicle- and 5 E2-treated, OVX females to quantify the mRNA expression of BKα channel. E2-treatment increased the mRNA expression of BKα. The expression values were calculated via the ΔΔCT method, normalized to GAPDH and relative to the oil control values. Bar graphs represent the mean ± SEM (unpaired two-tailed t-test for BK, t(6) = 3.479, **p < 0.01). H. The mathematical model was calibrated to reproduce the current voltage relationship observed in Kiss1ARH neurons from E2-treated animals (see Table S1 for gBK) before and after treatment with IbTx. The modeled IbTx -sensitive current (right panel) matches the electrophysiological data. For the calibration it was assumed that the applied concentration of IbTx (200nM) completely blocked the BK current.

KCNQ channels (M-current) contribute to the slow AHP in Kiss1ARH neurons.

A-B. Representative current traces of the M-current inhibition caused by 40 µM XE-991 perfused for 10 min in (A) OVX-oil and (B) OVX+E2-treated female mice. Inset: M-current deactivation protocol. C-D. Current density-voltage plots from –75 to –30 mV of vehicle and XE-991 perfusion in (C) OVX-oil and (D) OVX+E2-treated mice. Two-way ANOVA for C: main effect of treatment (F(1, 17) = 1.908, p = 0.1851), main effect of time (F(9, 153) = 187.1, p < 0.0001), and interaction (F(9, 153) = 3.901, p = 0.0002); Veh, n = 11; XE-991, n = 8; Bonferroni post hoc test, p > 0.05. For D: main effect of Veh and XE-991 (F(1, 24) = 24.92, p < 0.0001), main effect of time (F(9, 216) = 174.5, p < 0.0001), and interaction (F(9, 216) = 52.75, p < 0.0001); Veh, n = 13; XE-991, n = 13; Bonferroni post hoc test, a = p < 0.05, b = p < 0.001. E. Treatment with E2 elevated, while XE-991 diminished the maximum peak current density elicited by a -30 mV step in OVX- and OVX+E2-treated mice. Two-way ANOVA: main effect of Veh and XE-991 (F(1, 41) = 47.59, p < 0.0001), main effect of OVX and OVX+E2 (F(1, 41) = 15.76, p = 0.0003), and interaction (F(1, 41) = 18.2, p = 0.0001; Veh, n = 11; XE-991, n = 8; Bonferroni post hoc test, Veh: OVX vs. OVX+E2, a = p < 0.001. XE-991: OVX vs. OVX+E2, p > 0.05. F. Kiss1ARH neurons (three to four 10-cell pools) were harvested from each of 5 vehicle- and 5 E2-treated, OVX females to quantify the mRNA expression of Kcnq2. E2 treatment increased the mRNA expression of Kcnq2. Unpaired t-test, t(8) = 4.850, **p = 0.0013. G. Percent contribution of the different K+ currents to the repolarization current during burst-type firing activity in the OVX+E2 state. At each time point, the length of each color bar denotes the percent contribution of the corresponding current to the total outward current.

Estradiol decreases Tac 2, Trpc5 and Kcnj6 but increases Vglut 2 mRNA expression in Kiss1ARH neurons.

A. qPCR amplification curves illustrating the cycle threshold (CT) for Tac2, Gapdh, Kiss1, Trpc5, Slc17a6 (Vglut2) and Kcnj6 (GIRK2) in Kiss1ARH five cell (C,D,E) or ten cell (F) neuronal pools (three to six pools from each animal) in OVX Oil-treated, and B, in OVX E2-treated females. C. Quantitative real-time PCR analysis of Tac2 mRNA (n=5 animals), D, Vglut2 (n=7 animals), E, Trpc5 (n=5 animals), F, Kcnj6 (n=5 animals). Comparisons were made between Oil-treated and E2-treated, OVX females using the comparative 2-ΔΔCT method. Bar graphs represent the mean ± SEM (Unpaired t test for Tac2, t(6)= 6.350, p<0.001; Unpaired t test for Vglut2, t(8)= 4.522, p<0.001; Unpaired t test for Trpc5, t(6)= 4.818, p<0.01; Unpaired t test for Kcnj6, t(6)= 3.457, p<0.01). The data for C and D (Tac2 and Slc17a6) has been published previously (Qiu J. et al., 2018).

CRISPR mutagenesis of Trpc5 channels in Kiss1ARH neurons.

A, structure of AAV1-FLEX-SaCas9-U6sgTrpc5-exon2. Exon 2 of Trpc5 is denoted with guide sequence highlighted in red, the PAM is underlined. B. Structure of AAV1-FLEX-SaCas9-U6sgTrpc5-exon7. Exon 7 of Trpc5 is denoted with guide sequence highlighted in red, the PAM is underlined. C1, image of coronal section through the ARH from Kiss1-Cre::Ai32 mouse with dual co-injections of AAV-DIO-mCherry and AAV1-FLEX-SaCas9-U6-sgTrpc5. Scale = 200 µm. C2, C3, higher power overlays of epifluorescence (EYFP & mCherry) images with recording pipette patched onto Kiss1ARH-Cre:mCherry cell (C3). Scale = 40 µm. D, quantitative PCR measurements of Trpc5 transcripts in double sgRNA mutagenesis of Trpc5 (second sgRNA against pore forming region) in Kiss1ARH neurons. Primers were targeted to 1st or 2nd guide, respectively.

Double CRISPR mutagenesis of Trpc5 attenuates slow EPSP, increases rheobase and shifts the F-I curve.

A, high-frequency photo-stimulation (20 Hz) generated slow EPSP in Kiss1ARH neuron from ovariectomized, control mouse. Red trace is slow EPSP after low-pass filtering. B, slow EPSP in Kiss1ARH neuron from OVX, double sgTrpc5 -targeted mouse. C, summary of the effects of Trpc5 mutagenesis on slow EPSP amplitude in female mice (**** p < 0.0001). D, double sgRNA mutagenesis of Trpc5 channels in Kiss1ARH neurons significantly increased the RMP (control: -64.5±1.4 mV versus double sgTrpc5 1, 2: -71.1±1.2 mV, *p = 0.0007). E, current ramp showing the increased rheobase in sgTrpc5 double mutagenesis (control: 31.1±1.2 pA, n=31, versus sgTrpc5 1&2, 35.3±1.0 pA, n=33, p = 0.0073). F, firing frequency vs. current (F-I) curves for control versus sgTrpc5 double mutagenesis (*<0.05; **<0.01 and ***p < 0.005, respectively). G, model simulations of the effects of a current ramp (50 pA/sec) for OVX (left panel) and OVX female with reduced (muted) TRPC5 conductance (right panel), and H, the associated firing frequency vs. current curves. In the latter case the TRCP5 conductance was halved, which is a conservative estimation of the CRISPR state in which the Trpc5 is much more mutated in Kiss1ARH neurons (Figure 11).

Computational modeling of a Kiss1ARH neuron in the OVX and OVX + E2 state demonstrates its distinct dynamic responses.

A model of the Kiss1ARH neuron was developed and calibrated using molecular data and electrophysiological recordings of Kiss1ARH neurons from OVX and OVX+E2 mice. A. Simulations of the OVX-parameterized model demonstrating high frequency activity in response to NKB stimulation. The balance between GIRK and TRCP5 conductance controls the response of the neuron to NKB stimulation, with neuronal response eliminated when TRPC5 conductance is low (red triangle) relative to the GIRK conductance. B. The OVX+E2 parameterized models demonstrate sustained burst firing activity. The bursting activity that is supported by elevated h- and Ca 2+-currents (red square). C. In the OVX+E2 state, burst firing activity is also supported by high conductance of HVA Ca2+ channels relative to the conductance of TRPC5 channels. Representative points in the parameter space giving rise to burst firing activity are marked with red squares, whereas red triangles are used for points resulting in regular spiking. The black line separates these two regions of activity.

Schematic diagram of the conductance based mathematical model of Arcuate nucleus Kiss1 neurons.

Table of model parameters.