Gonadal hormone‐independent sex differences in GABAA receptor activation in rat embryonic hypothalamic neurons

GABAA receptor functions are dependent on subunit composition, and, through their activation, GABA can exert trophic actions in immature neurons. Although several sex differences in GABA‐mediated responses are known to be dependent on gonadal hormones, few studies have dealt with sex differences detected before the critical period of brain masculinisation. In this study, we assessed GABAA receptor functionality in sexually segregated neurons before brain hormonal masculinisation.


| INTRODUCTION
Sexual differentiation of the brain is mediated by gonadal hormones during embryonic development, within a timeframe called the "critical period," established between embryonic day 18 (E18) and post-natal day 10 (PN10) (McCarthy, Wright, & Schwarz, 2009). In this regard, the hypothalamus is the most sexually dimorphic region of the brain.
Regarding GABA, in cultured neurons (E16) after 9 days in vitro (DIV), we have detected a larger population of male than female neurons depolarising after GABA A receptor stimulation. Male neurons also displayed larger and longer lasting responses than females after GABA A receptor activation, even in the absence of hormone exposure (Mir, Carrer, & Cambiassso, 2017).
In the present work, we explored sex-dependent differences in GABA response in immature hypothalamic neurons. Using cell Ca 2+ imaging recordings and patch-clamp measurements, we found differential responses to GABA between immature male and female hypothalamic neurons (2 DIV) that were cultured before hormone exposure and brain masculinisation. Our data suggest that these differences probably rely on sexual genetic backgrounds, independent of the sexual hormone environment. Hypothalamic cultures of rat embryonic (E16) neurons were prepared, as previously described (Cambiasso et al., 2000). Briefly, the presence of a vaginal plug was designated as gestational day 1; then, on the morning of gestational day 16, pregnant Wistar rats (RRID:RGD_2312511, n = 23) were anaesthetised with CO 2 and killed by cervical dislocation. Fetuses were quickly removed under sterile conditions and separated by sex through identification of the spermatic artery on developing testes. After removing the brain, three to five ventromedial hypothalami of each sex were dissected out, pooled, and then incubated in trypsin-Hank's solution at 37 C for 15 min. The digested tissue was re-suspended in DMEM containing 10% fetal calf serum and mechanically dissociated by gentle aspiration with fire-polished Pasteur pipettes. The dissociated cell suspension was plated at high density (>60,000 cellsÁcm −2 ) on pre-coated poly-L-lysine (1 mgÁml −1 ) 12-mm glass coverslips. Cultures were maintained 2 DIV in an incubator with (1:1) DMEM:Ham's F12 nutrient mixture-astrocyte conditioned media.
Astroglial cultures were prepared as above. Briefly, mesencephalic tissue was chemically and mechanically digested and re-suspended in DMEM-10% fetal calf serum. After that, cells were plated in 25-cm 2 sterile flasks at high density and maintained in an incubator until a confluent monolayer was established around 11-15 DIV. The DMEM: Ham's F12 media conditioned for 48 hr by the astrocyte cultures was used to feed hypothalamic neurons. All cultures were maintained under phenol red-free conditions to avoid "oestrogen-like effects" (Berthois, Katzenellenbogen, & Katzenellenbogen, 1986). In some experiments, male and female hypothalamic cultures were treated with 10-nM testosterone immediately after plating until completing 2 DIV (see figure legends for details). Testosterone was diluted in culture media to an appropriate concentration with 10-mM ethanol stock. The final concentration of ethanol in the culture media never exceeded 0.001%.

| Experimental protocol
All the studies were designed to generate groups of equal size, using randomization and blinded analysis.
Inhibitors (bicuculline or nifedipine) were added during this time in some experiments.
Coverslips containing cells were then mounted in a recording chamber with 500-μl aCSF for live cell imaging using spinning disc microscopy (Olympus optical, Tokyo, Japan) with epifluorescence illumination (150 W Xenon lamp), a microprocessor, a Hamamatsu CCD video camera, and image intensifier at standard parameters (λexcitation = 492 nm, λemission = 514 nm; imaging frequency: 0.5 Hz; exposure time: 50 ms); 20-30 neurons of each sex were morphologically identified using a 63X objective. After 1 min of recording resting Ca +2 (baseline), 500 μl of aCSF containing either GABA (10 μM), muscimol (10 μM), or GABA + inhibitors (50-μM bicuculline or 20-μM nifedipine) at a 2× concentration was manually added and signals recorded for 4 min. In some experiments, after 3 min of GABA stimulation, the aCSF was removed, and GABA + 5-μM propofol was added, and Ca 2+ signals were obtained over a period of 3 min. During the last minute of all recordings and after aCSF removal, cultures were stimulated with 90-mM KCl. Only neurons that responded to KCl stimulation were included for analysis.
After background subtraction, the fluorescence was quantified with the Fiji/ImageJ Time Series Analyzer plug-in (NIH, Bethesda, MD, USA). Fluorescence intensities (F) were normalised by the average baseline (F 0 ). The ratio F/F 0 represents the cytoplasmic Ca 2+ signal as a function of time. We documented the number of neurons with response (F/F 0 > 20% of baseline) and without response (F/F 0 < 20% of baseline); peak (maximum fluorescence intensity after drug application); rise time (time [seconds] to reach the peak); and decay time (time [seconds] needed to return to F/F 0 = 20%).
Electrical measurements were carried out at room temperature with an Axopatch-200A amplifier (Axon Instruments, Foster City, CA, USA). Data were sampled at 10 kHz and pass filtered at 2 kHz, digitalised with an A/D Digidata 1000 using pClamp software (pClamp, RRID:SCR_011323, Molecular Devices, Union City, CA, USA). Pipettes were visualised under a microscope (Olympus optical, Tokyo, Japan) and positioned over the cells by micromanipulators. Whole-cell voltage clamp recordings were performed after formation of a GΩ resistance seal and break-in, while perforated patch-clamp recordings were started at least 20 min after cell-attached formation and transient capacitive peak apparition. Pipette and whole-cell capacitance and series resistance were compensated using amplifier circuitry. Cells were clamped at −50 mV, and only those that presented inward currents in response to a voltage ramp (−80 to +30 mV) and less than 100 pA of leak current were included for analysis.
Drugs were diluted in aCSF to an appropriate concentration with 10-mM DMSO or distilled water stocks. Drug delivery was performed using a gravity-driven system connected to a capillary HPLC that was positioned 50 μm from the recorded neuron. Neuronal responses to 1-s exposure to the drugs were assessed every minute. The rationale behind the modulators used in this study was based on their ability to identify GABA A receptor subtypes by characterising their responses to each drug. Diazepam potentiation depends on γ2 and either α1, α2, α3, or α5 subunits in GABA A receptors. Furosemide is a diuretic and a strong inhibitor of GABA A receptors formed by α4β2/3γ2 and α6β2/3γ2, whereas α1/2/3/5β2/3γ2 conformations are practically insensitive. THIP works as a partial agonist of α4βγ2 receptors; a total agonist effect is observed in α4βδ receptors. Ro 15-4513, another synthetic agonist, enhances currents in α4βγ2/ α6βγ2 and α6βδ/α4βδ receptors. Similarly, La +3 blocks only the GABA A receptors made up of α6β3γ2L/α6β3δ and α4β3γ2/α4β3δ subunits. On the other hand, the efficacy of the synthetic neurosteroid alfaxalone is higher in δ-GABA A receptors than in γ2-GABA A receptors. Ethanol produces a mild potentiation in γ2-GABA A receptors and a high potentiation in δ-GABA A receptors . Moreover, ethanol has no effect on GABA A receptors lacking δ or γ subunits. Zn 2+ is a non-competitive antagonist for receptors expressing αβε but lacking γ or δ subunits. Propofol potentiates the majority of GABA A receptor subtypes by a β subunit-dependent mechanism. Particularly, ε-GABA A receptors show some resistance to propofol with effects greater in α3βθε than in α3βε conformations (Johnston, 1996;Korpi et al., 2002;Olsen & Sieghart, 2008).
Current density in GABA A receptors was determined dividing the current amplitude of a saturating GABA dose (500 μM) by the capacitance of the same neuron. EC 50 and the Hill coefficient were obtained by fitting the Hill equation to dose-response current amplitudes for each neuron using ORIGIN ® software (Northampton, Massachusetts, USA). All measurements were stored in a PC, and offline analyses were performed with Clampfit (Molecular Devices, Union City, CA, USA).

| Data and statistical analysis
The data and statistical analysis comply with the recommendations of the British Journal of Pharmacology on experimental design and analysis in pharmacology. All treatments were assigned randomly. The data are presented as mean ± SEM or percentage change over control.
The Student's t test was used to compare the response of male and female neurons to individual treatments. The χ 2 test was used to compare the proportion of cells with or without response to a specific drug in each sex. Power analysis was set at P < .05 value. The statistical analysis for each experiment was performed with data obtained from at least three independent cultures and was undertaken only when each group size was at least n = 5 neurons. All statistical analysis was performed with STATISTICA software (STATISTICA, RRID: SCR_014213, StatSoft Inc., Tulsa, OK, USA).

| Materials
The following compounds were obtained from Tocris Bioscience

| GABA elevates intracellular Ca 2+ levels differentially in male and female neurons
To evaluate sex-associated responses of immature neurons after GABA stimulation, hypothalamic neurons were cultured for 2 DIV, a period in which most neurons display a symmetric arrangement of short neurites (minor processes) and one single axon (two to three times longer than minor processes ;Cambiasso et al., 1995Cambiasso et al., , 2000. Accordingly, we measured GABA-mediated

| GABA A receptor-induced [Ca 2+ ] i increases are mediated by L-type VDCC
To explore whether [Ca 2+ ] i increases were mediated by GABA A receptors, we used muscimol, a potent selective agonist for these receptors, to stimulate hypothalamic neurons. Neurons were treated acutely with 10-μM muscimol, which raised [Ca 2+ ] i similarly in males and females (Figure 2a,b). Moreover, co-application of 10-μM GABA and 50-μM for male and 30.2 ± 2.5 mM (n = 6) for female neurons. Of note, none of these parameters differed between males and females, suggesting that the sex differences observed in depolarising responses by Ca 2+ imaging cannot be attributed to variations in Cl − electrochemical gradient force between sexes (Figure 3b). vs. females: 155 ± 31 pA, n = 9). Moreover, muscimol (10 μM) did not enhance GABA A receptor currents differently between male and female neurons (300 ± 47 pA, n = 7 vs. 218 ± 32 pA, n = 9, respectively).

| Characterisation of GABA
We also analysed several intrinsic membrane properties such as RMP (which ranged from −30 to −57 mV in both sexes), capacitance and membrane resistance, to compare male and female electrical responses; however, no differences were detected between sexes (Table 1). Therefore, neuronal membrane voltage was set at −50 mV to simplify further comparative analysis. We also estimated GABA A receptor current density, establishing a ratio between the current amplitude evoked by a saturating dose of GABA (500 μM) and the capacitance of the same neuron; this ratio is proportional to GABA A receptor levels. Average current amplitudes recorded were 607 ± 104 pA and 530 ± 175 pA for males and females respectively.
Moreover, mean GABA A receptor current density values were 55.8 ± 7.6 pA/pF for males (n = 10) and 52.5 ± 15.5 pA/pF for females (n = 10). Together, these data suggest that male and female neurons share similar levels of active GABA A receptors at this developmental stage.

| Pharmacological characterisation of GABA A receptors in male and female hypothalamic neurons
The physiological properties of GABA A receptors vary depending on its subunit composition. Therefore, we performed an exhaustive functional characterisation of male and female GABA A receptors to explore sex-dependent composition.

| GABA sensitivity
First, we studied GABA sensitivity in male and female hypothalamic neurons at 2 DIV, through a concentration-response analysis; Figure 4a shows that the higher the concentration of GABA, the higher the current registered, revealing more recruitment of GABA A receptors . A weak response was elicited by the lowest concentration (1 μM) as well as fast desensitisation after 500-μM GABA stim- 2.94 ± 1.98, n = 11) differed between sexes (Figure 4b). Propofol is an anaesthetic that potentiates the action of GABA at most GABA A receptor subtypes, although some resistance is observed in GABA A receptors containing ε subunits (Davies, Hanna, Hales, & Kirkness, 1997;Davies, Kirkness, & Hales, 2001). In our model, while 5-μM propofol enhanced GABA currents in most male or female hypothalamic neurons, 30% remained insensitive or were slightly inhibited. Moreover, the potentiation achieved by propofol was greater in female than in male neurons ( Figure 5), suggesting differences in GABA A receptor subunits between sexes. Table S1 summarises our main conclusions regarding the GABA A receptor subtypes detected in male and female neurons based on this screening.

| Effect of propofol on GABA A receptormediated [Ca 2+ ] i increases
As propofol (5 μM) showed a selective enhancement in female neurons, we explored whether co-stimulation with GABA (10 μM) would increase Ca 2+ influx. However, this strategy blunted a GABA A receptor-mediated [Ca 2+ ] i increase , a phenomenon observed only in female neurons (Figure 7a,b).
In this work, we showed that sex differences in GABA responses of hypothalamic neurons are manifested before the critical period of brain masculinisation, mostly independent of hormonal treatment.
Our data suggest that a greater number of male than female neurons were depolarised by GABA (as shown by Ca 2+ influx), exhibiting depolarising responses lasting longer periods of time ( Figure 1). These results are consistent with previous evidence obtained in 9 DIV neurons (Mir et al., 2017), suggesting that sex differences are established early in neuronal development, even before the peak of testosterone levels at E18. We also found that GABA A receptors mediate Ca 2+ influx, membrane depolarisation, and L-type VDCC opening (Figure 2), the canonical pathway by which GABA excites immature neurons (Ben-Ari, 2002). Surprisingly, the inhibition of Ca 2+ influx by nifedipine was stronger in male than in female neurons (Figure 2f), suggesting differences in regulatory mechanisms for L-type VDCC. In this regard, Considering the responses recorded using allosteric modulators, as well as by expression profiles previously published (Laurie et al., 1992;Pape et al., 2009), we conclude that both male and female hypothalamic neurons possess a large variety of functional GABA A receptors (Figures 5 and 6;Johnston, 1996;Korpi et al., 2002;Olsen & Sieghart, 2008). From our data, we would hypothesise the coexistence of several populations of hypothalamic neurons in culture, an observation also supported by the high variation of GABA-mediated responses (even within each sex). Accordingly, we infer that almost all 2 DIV neurons display α2β2/3γ2, α3β2/3γ2, and α5β2/3γ2 conformations of GABA A receptors. Nevertheless, a subpopulation (around 50%) could also display functional GABA A receptor containing α4, α3, θ, and/or ε subunits. It should be noted that we also found that propofol-dependent current potentiation was higher in females than in males ( Figure 5) but that propofol acted as a potent blocker of Ca 2+ influx mediated by GABA in females but not in males (Figure 7), suggesting sex differences in GABA A receptor subunit composition. In other words, this propofol-dependent effect may reveal differences between sexes in β, θ, and/or ε GABA A receptor composition (Table S1). Thus, it is important to highlight that administration of testosterone did not erase sex differences, either in GABA A receptordependent depolarisation or in propofol response (Figure 8), suggesting hormone-independent effects. In fact, testosterone treatment increased the sex differences reported in this study.
Activation of GABA A receptors is the main excitatory signal for embryonic developing circuits, modulating Ca 2+ -mediated processes F I G U R E 7 Propofol-dependent modulation of GABA A receptor induces intracellular calcium increases. (a) Cal-520 fluorescence (F/F 0 ) measured in single cells after 10-μM GABA, 10-μM GABA + 5-μM propofol (PROP), and 90-mM KCl stimulation in both male and female hypothalamic neurons at 2 DIV. Each trace represents a time-dependent fluorescence signal measured in a single neuron soma. Bars represent drug exposure time. (b) Maximum amplitudes (peak) of Ca 2+ signals after 10-μM GABA and 10-μM GABA + 5-μM propofol treatments in male and female cultured hypothalamic neurons. Values represent the means ± SEM. *P < .05, significantly different as indicated; Student's t test such as neuronal differentiation, neurite outgrowth, and survival (Represa & Ben-Ari, 2005;Sernagor, Chabrol, Bony, & Cancedda, 2010). Earleir work had shown that propofol, through GABA A receptors and L-type VDCC activation, modified the axonal and dendritic morphology of cortical neurons (Briner et al., 2011;Mintz, Barrett, Smith, Benson, & Harrison, 2013) and also produces cell death of hippocampal neurons (Kahraman, Zup, McCarthy, & Fiskum, 2008). Moreover, male hippocampal neurons were more vulnerable than those of females to GABA A receptor-dependent excitotoxicity, apparently due to failures in switching-off Ca 2+ transients elicited by GABA A receptor over-activation. This effect has been mainly attributed to a hormone-dependent sex difference of GABA A receptor subunits (Nuñez & McCarthy, 2008). Nevertheless, differences detected in our study are independent of testosterone treatment and most probably dependent on sex chromosome complement.
Considering our results blocking L-type VDCC with nifedipine in male and female neurons, we do not discount sex differences in the composition and expression of these channels (Figure 2f). In fact, propofol inhibits L-type VDCC by a voltage-dependent inactivation mechanism (Fassl, High, Stephenson, Yarotskyy, & Elmslie, 2011;Martella et al., 2005), which could explain the selective inhibition by propofol in females (Figure 7). Several reports support the notion that sex chromosomes encode many transcription factors regulating both autosomal and sexual genes, leading to imbalances in gene expression between XX (female) and XY (male) cells (Carrel & Willard, 2005;Lee & Bartolomei, 2013;Wijchers & Festenstein, 2011). In fact, the cluster of genes encoding α3/θ/ε GABA A receptor subunits (Simon  et al., 2004) and the Ca v 1.4, L-type VDCC-subunit gene (Catterall, Perez-Reyes, Snutch, & Striessnig, 2005) are located on the X chromosome, and their expression could be different in XX and XY hypothalamic neurons.
The hypothalamus is one of the most sexually dimorphic regions in the brain, controlling important sexually dimorphic behaviours (Flanagan-Cato, 2011;Griffin & Flanagan-Cato, 2009;Yang et al., 2013). Hypothalamic sex differences are largely connected to early events in development such as proliferation and apoptosis, lineage commitment, neuronal migration, and connectivity. Accordingly, GABAergic signalling, the main excitatory input at embryonic developmental stages (Ben-Ari, 2002), is critical to sustain developmental and physiological aspects of developing neurons before the establishment of synapses. Several trophic and paracrine roles have been described for GABA, ranging from cell proliferation control, migration, neurite outgrowth, and synapse formation (Cancedda et al., 2007;Chudotvorova et al., 2005;Represa & Ben-Ari, 2005;Reynolds et al., 2008;Sernagor et al., 2010). Therefore, the sexually dimorphic depolarising effects of GABA reported in this work could differentially influence the morphology, physiology, and connectivity of male and female hypothalamic neurons, even before exposure to gonadal hormones. Moreover, our pharmacological screening reinforces the importance of considering sex as a key variable for pharmacological studies. Of note, our results suggest sex differences in GABAergic signalling after treatment with propofol, which is regularly used as an anaesthetic. Such an observation has clinical relevance for men's and women's health (Briner et al., 2011;Kahraman et al., 2008;Mintz et al., 2013).
In summary, our work shows that male and female hypothalamic neurons differ in their GABAergic physiology, independent of gonadal hormones. Hormone administration did not erase differences, suggesting that the sexual genetic backgrounds of sexes are the most probable basis for these findings. To our knowledge, this is the first study reporting GABAergic differences between male and female neurons before brain sexual differentiation, and the consequent importance of considering this issue in the biology and physiology of hypothalamic neurons.