The animal care and experimental protocols were reviewed and approved by the IACUC of the University of Tennessee Health Science Center, which is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care.

C57BL/6J mice of both sexes, all 8–12 weeks old, were anesthetized with a mixture of xylazine/ketamine (12/100 mg/kg of weight) and kept anesthetized for the duration of the experiment with subsequent ketamine doses (50 mg/kg of weight) every 15 min or as needed. A catheter was inserted into the internal carotid artery so that experimental drug infusions were directed toward the brain rather than the thoracic cavity. An area of the skull was cleared of tissue and thinned in order to visualize the branching arteries originating from the middle cerebral artery (MCA) on the brain side where the catheter was inserted, above the zygomatic arch, between the ear and eye [27, 40]. The arteries branching out from the MCA were monitored using a Leica MC170 HD microscope with a mounted camera (Leica M125 C) connected to a computer monitor. Drugs were diluted to their final concentration in 0.9% NaCl and administered via catheter at 0.1 mL/25 g of mouse weight. Cranial window images before and after drug administration were acquired every 60 s for later analysis; a sample of n = 5–6 was acquired for each group (with n representing the number of separate animals).

Male and female C57BL/6J mice, all 8–12 weeks old, were deeply anesthetized with isoflurane via inhalation using an open-drop method in a bell jar. Upon losing their response to toe pinch, animals were quickly decapitated with sharp scissors. Resistance-size MCAs (∼100 μm in outer diameter) were dissected from the mouse brains. Endothelium was removed by passing an air bubble through the vessel lumen for 90 s [29]. Arterial segments (0.5 cm long) were cannulated at each end, and the artery exterior was continuously perfused with physiologic sodium saline (PSS) of the following composition (mM): 119 NaCl, 4.7 KCl, 1.2 KH₂PO₄, 1.6 CaCl₂, 1.2 MgSO₄, 0.023 EDTA, 11 glucose, and 24 NaHCO₃; pH = 7.4, at 35°C–37°C. PSS was continuously bubbled with O₂/CO₂/N₂ at 21/5/74%. Vehicle control (dimethyl sulfoxide; DMSO), PREG, EtOH, or the PREG + EtOH combination were diluted into PSS and perfused over the arterial segment. The artery external wall diameter was measured using the automatic edge-detection function of the IonWizard software package (IonOptix) via a Leica MC170 HD microscope with a mounted camera (Leica M125 C) connected to a computer monitor.

For all electrophysiological recordings, whether in mouse cerebral artery myocytes or following heterologous expression of recombinant BK channels in Xenopus laevis oocytes, ionic currents were recorded from excised membrane patches in the inside-out (I/O) patch-clamp configuration. Patch-recording electrodes were pulled from glass capillaries and treated as described previously [41]. When filled with high K+ solution (see below for composition of electrode solutions), the vast majority of tip resistances were ∼2 MΩ, with a few reaching 5 MΩ. Series resistance was electronically compensated up to 80% by the EPC8 amplifier. In all experiments, whether on myocytes or oocytes, the nominal free [Ca²⁺] in experimental solutions was calculated with MaxChelator Sliders (Stanford University) and validated experimentally using Ca²⁺-selective and reference electrodes [42]. Solutions were applied to the cytosolic side of the patch using an automated, pressurized Octaflow system (ALA Scientific) through a micropipette tip with an internal diameter of 100 μm. Experiments were carried out at room temperature (20°C–22°C). Ionic currents at single-channel resolution were recorded using an EPC8 amplifier (HEKA) at 1 kHz. Data were digitized at 5 kHz using a Digidata 1320A A/D converter and pCLAMP 8.0 (Molecular Devices).

For ionic current recordings from MCA smooth muscle BK channels, cerebral artery myocytes were isolated from adult mouse MCA as described in detail elsewhere [43]. Bath and electrode solutions contained (mM): 130 KCl, 5 EGTA, 1.6 HEDTA, 2.28 MgCl₂ ([Mg²⁺]_free = 1 mM), 15 HEPES; pH 7.4. Free [Ca²⁺] in the solution (30 µM) was adjusted to the desired value by adding CaCl₂. An agar bridge with Cl⁻ as the main anion was used as a ground electrode.

For ionic current recordings in Xenopus laevis oocytes, isolated oocytes (stages V and VI) were purchased from Xenopus 1. Oocytes were defolliculated with forceps under a microscope and stored at 18°C until injection with cbv1-coding cRNA injection. Each oocyte was injected with 23 nL of 40 ng/μL cbv1 cRNA, with patch-clamp recordings being conducted 36–72 h after injection. Immediately before patch recordings, each oocyte was manually freed from its vitelline layer as described [41]. Both bath and electrode solutions contained (mM) 135 K⁺ gluconate, 5 EGTA, 2.28 MgCl₂, 15 HEPES, and 1.6 HEDTA, pH 7.4. As for solutions used with myocyte experiments, free [Ca²⁺] in the solution (30 µM) was adjusted to the desired value by adding CaCl₂. An agar bridge with K⁺ gluconate as the main anion was used as a ground electrode [41]. Two major Ca²⁺-dependent gating parameters, i.e., the Ca²⁺ dissociation constant (K_d) and the allosteric factor (C) that couples Ca²⁺-binding to channel close-open transitions in absence of stimuli, were estimated from the Ca²⁺ dependence of Po at very negative voltages; such estimates have been used previously to study the effect of EtOH on the Ca²⁺ gating behavior of cbv1 channels [44]. To do this, we obtained R0, i.e., the NPo ratio in the presence of Ca²⁺ (0.1–100 µM) over the NPo ratio in absence of Ca²⁺ (determined +30 mV) in the absence and presence of 10 µM PREG; data were then fitted using the following equation: R 0   C a 2 + = N P 0 V , C a 2 + N P 0 V , 0 = 1 + K C 1 + K 4 = 1 + C C a 2 + / K d 1 + C a 2 + / K d 4

Pregnenolone was purchased from Abcam. Ethanol (200 proof; E7023) and all other chemicals were purchased from Sigma Aldrich. PREG stock solution was prepared in DMSO and diluted into saline, PSS, or patch-clamp bath solution immediately before application to the animal, artery, or membrane patch, respectively. Each animal or pressurized artery was exposed to vehicle, PREG, EtOH, or the PREG + EtOH combination only once in order to avoid any possible receptor (i.e., BK channel) desensitization [45]. Membrane patches were perfused with increasing concentrations of Ca²⁺, first in the absence and then in the presence of PREG.

Analysis was performed by investigators who were blind to experimental group identity. Changes in artery diameter obtained from cranial window experiments were determined using the ImageJ software package (ImageJ 1.52a, downloaded from https://imagej.nih.gov/ij/download.html). Changes in artery diameter in vitro were determined using the IonWizard software package (IonOptix). The product of the number of channels in the membrane patch (N) and the individual open probability (Po) was used as an index of channel steady-state activity. NPo was obtained using a built-in function in pCLAMP 8.0 (Molecular Devices).

Statistical analysis was performed using the InStat3.05 software package (GraphPad). Data distributions were checked using a Kolmogorov–Smirnov approach in cases where the number of observations ≥10. For normally distributed data (Gaussian type), the t-test was used to test for statistically significant differences between two groups. For data following a non-Gaussian distribution or whose mode of distribution could not be established with certainty (number of observations <10), the statistical methods employed included the Mann–Whitney test for comparisons between two experimental groups, and the Kruskal–Wallis test followed by Dunn’s post-test for comparisons of three or more experimental groups. The threshold for significance was set at p < 0.05; group sizes were determined to achieve greater than or equal to 80% power at this significance threshold. Data are reported in the form mean ± SEM. Final data plotting and fitting processes were conducted using the Origin 2020 software package (OriginLab).

In order to evaluate the combinatory actions of PREG + EtOH on male and female cerebral arteries at the organismal level, we used the cranial window technique. This technique allows for the continuous monitoring of resistance-size pial arteries that branch out of the MCA, and has previously been used to evaluate the pharmacological effects of each drug on these vessels [19, 27]. Intra-carotid infusion of either volume control (0.9% NaCl) or vehicle control (DMSO) failed to evoke significant changes in MCA diameter when compared to averaged pre-infusion values (i.e., the baseline in Figure 1), which were obtained via continuous artery diameter monitoring for no less than 3 min. For the testing of PREG, a concentration of 10 μM was chosen because this constitutes the lowest PREG concentration that is able to evoke maximal constriction of the mouse MCA in vitro and in vivo (EC_max) [19]. For EtOH, 50 mM was chosen because this concentration is reached in blood circulation after a moderate-to-heavy alcohol consumption episode and is close to EC_max for ethanol for constriction of mouse cerebral arteries that branch out of Willis’ circle, including the MCA [28]. In contrast to the controls, bolus injections of either PREG or EtOH administered to male animals resulted in 28% ± 3.8% and 31.2% ± 2.1% reductions in artery diameter, respectively (Figure 1). For both agents, maximal constriction was detected around 3 min after bolus injection. The effect of each agent differed significantly from the time-matched effects of either saline or DMSO (p = 0.0079–0.0043). In female animals, PREG and EtOH also evoked a peak constriction around 3 min after bolus injection, with arterial diameter decreasing by 21.3% ± 2.2% and 19.5% ± 4.1%, respectively. These vasoconstrictive responses did not differ statistically from those observed in males (Figures 1A, C, E vs. Figures 1B, D, F).

In males and females, concomitant application of PREG + EtOH caused MCA constriction of magnitude 33.6% ± 2.8% and 19.3% ± 3.2%, respectively. Within each sex, the response to the combination PREG + EtOH did not differ statistically from the responses evoked by the individual agents (Figures 1C–F). Since the two agents were applied locally in bolus with the injectate directed toward their site of action (the MCA pial artery branch under recording), the lack of synergism between the PREG and EtOH vasoconstrictive actions is unlikely to have been due to modification of the pharmacokinetic properties (i.e., absorption, distribution, metabolism, and/or elimination) of one drug caused by the simultaneous presence of the other. Rather, the lack of synergism can be explained by (i) the system reaching its maximal level of constriction under each agent and under their combination (a “ceiling effect”), or (ii) convergence of the constrictions elicited by either PREG or EtOH on a given organ/tissue pathway.

To investigate the possibility that EtOH and PREG constrict MCA through a common pathway, we used a wide range of PREG concentrations (10 nM–100 µM [19]) in the presence and absence of 50 mM EtOH. If the two cerebrovascular constrictors were acting through a common pathway, then submaximal concentrations of PREG in combination with EtOH at 50 mM would show additivity in constricting the MCA [46]. Since we have previously documented that MCA constriction by either agent does not require the endothelium, but is mediated by targets and mechanisms located in the vascular SM [19, 29], MCA segments were de-endothelialized before pressurization, as described in the Materials and Methods section.

Remarkably, submaximal concentrations (<EC_max, 0.001–0.1 μM) of PREG that evoke constriction displayed synergism with 50 mM EtOH (Figures 2C, D); in males, the degrees of constriction induced by 10 nM and 100 nM PREG in the presence of co-administrated EtOH (9.32% ± 2.16% and 5.88% ± 0.83% constriction, respectively) were significantly greater than the degrees of constriction produced solely by PREG (2.96% ± 0.38% and 3.31% ± 0.38%); p = 0.0159 and p = 0.026, respectively. Importantly, this synergism was lost at maximally effective and supramaximal concentrations of PREG (i.e., 1, 10, and 100 μM [19]) (Figure 2C), which is to be expected in a case of two agents acting via a common pathway or common target(s).

In females, the concentration–response curve of MCA constriction in response to PREG was restricted to evaluate higher concentrations, shown to be effective in our previous publication [19]. Records from these animals also showed synergism between PREG and EtOH when PREG was probed at submaximal concentrations (1 µM): 2.46% ± 0.91% constriction vs. 0.66% ± 0.38% constriction with PREG alone (p = 0.04206; Figure 2D). As found with males, MCA constriction in females under exposure to EtOH and PREG was characterized by loss of synergism when PREG was probed at maximally effective concentrations (10 and 100 μM; [19]) (Figure 2D). Since it has been documented by us and others, both in this system and under identical conditions, that depolarizing 60 mM KCl constricts MCA by >20% in both males and females (see [28] and references cited therein), the lack of synergism between PREG and EtOH in evoking MCA constriction cannot be explained by a “ceiling effect” (i.e., by MCA segments reaching their maximal possible degree of constriction). Therefore, the synergism between EtOH and PREG at submaximal concentrations and the loss of synergism when either EC_Max is reached on isolated, de-endothelialized MCA segments indicate that the two drugs converge on a common pathway or target(s), likely located in the vascular SM itself.

Given the involvement of BK channels in EtOH- [25, 28, 29] and PREG-induced [19] constriction of de-endothelialized cerebral arteries, we next probed whether these channels were involved in MCA constriction by PREG or EtOH when these agents were applied in combination vs. separately. Synergism in MCA constriction was not detected either in males or in females when EtOH was probed at 75 mM (Figures 2E, F), extending our findings shown in Figures 2C, D. More importantly, in both males and females, paxilline at a concentration that selectively blocks BK channels (1 μM; [47]) completely abolished the constriction evoked by PREG alone and by PREG + EtOH (Figures 2E, F). This outcome indicates that the common pathway implicated in constriction induced by PREG and EtOH involves SM BK channels.

To determine whether SM BK channels were indeed shared targets of EtOH and PREG, we set out to explore whether the lack of synergism in concomitant application of EtOH + PREG could be observed at the level of BK channel activity itself, independently of cell signaling and internal organelles. Thus, we recorded BK channel steady-state activity (NPo) in excised, I/O membrane patches from myocytes freshly isolated from mouse MCA (Figure 3A). We chose 10 µM PREG because this concentration constitutes EC_max for PREG-induced constriction of MCA (Figure 2) and for BK channel inhibition by this neurosteroid [19]. The data showed that the inhibition of channel activity by PREG was indistinguishable from that evoked by PREG + EtOH. Indeed, the application of 10 μM PREG to the patch decreased BK NPo to 0.7 ± 0.04 of pre-drug levels (baseline in Figures 3B, C), while the concomitant application of PREG + EtOH decreased NPo to 0.71 ± 0.04 from the baseline (Figures 3B, C). Moreover, the inhibition of BK channel steady-state activity by either 10 µM PREG or 10 µM PREG+50 mM EtOH reported here is identical to the inhibition evoked by 50 mM EtOH alone [25, 29]. Since cerebrovascular SM BK channel inhibition is a well-known mechanism leading to cerebral artery constriction [30, 32], the lack of synergism between EtOH and PREG actions when probed at maximal concentrations in terms of their impact on SM BK channel activity, de-endothelialized MCA segments, and MCA in vivo supports the idea that inhibition of cerebrovascular SM BK channels is the common mechanism underlying MCA constriction induced by these drugs (see Discussion).

We have previously documented the finding that EtOH inhibition of BK channels at physiological levels of Ca²⁺ found in cerebrovascular myocytes requires the presence of modulatory, smooth muscle-abundant BK β1 subunits [25]. Moreover, the β1 subunit TM2 acts as a specific EtOH sensor [35]. In contrast, PREG-induced inhibition of these channels does not involve β1 regulatory proteins; instead, channel-forming α subunits suffice for steroid action [19]. While each ligand inhibits BK channel activity through recognition by different subunits that form part of the SM BK channel heteromer, the action of the two ligands must converge on a gating mechanism or mechanisms in order to explain their lack of synergism at maximal concentrations (shown in Figure 3). Therefore, we probed the effect of PREG on BK channel currents mediated by recombinant BK channel proteins cloned from cerebrovascular smooth muscle (cbv1 isoform; Material and Methods) and expressed in Xenopus laevis oocytes; this system allows for proper comparison with data previously obtained with EtOH under identical recording conditions [44]. Importantly, PREG has been shown to be ineffective in the absence of activating concentrations of Ca²⁺ [48]. Therefore, we focused on determining the action of PREG on the Ca²⁺-driven gating of cbv1 channels. Specifically, we derived the changes in the channel’s Ca²⁺ dissociation constant (K_d) and the allosteric coupling parameter (i.e., parameter C in the HA model [49]) that links Ca²⁺-binding to the intrinsic channel gating (i.e., closed-to-open transitions) occurring in the absence of Ca²⁺ binding and membrane depolarization. Both parameters were obtained as described in the Materials and methods section. Figure 3D shows that 10 µM PREG, surprisingly, did not increase K_d, but rather decreased it. However, PREG did decrease C, an allosteric decoupling that likely contributes to the inhibitory action of PREG on this channel. Remarkably, these two parameters of cbv1 channel gating are also targeted by EtOH, and the overall effect of this drug on cbv1 channel activity is largely determined by its actions on K_d and C [44]. Whether PREG actions on K_d and C are the primary determinants of overall PREG-induced inhibition of BK channels remains to be confirmed (see Discussion).