{"paper_id":"49564df1-63e6-4b83-9cd8-2c51b56d9f9d","body_text":"1 \nA Nonsteroidal Reversal Agent Inhibits Allopregnanolone \nModulation of α1β3δ GABAA Receptors \n \nXiaojuan Zhou, Youssef Jounaidi and *Keith W. Miller \n \nAuthor information \nKeith W. Miller, — Department of Anesthesia, Critical Care and Pain Medicine, Harvard \nMedical School, Massachusetts General Hospital, 32 Fruit Street, Boston, Massachusetts 02114, \nUnited States. ORCID # 0000-0002-4105-6639. \nYoussef Jounaidi — Department of Anesthesia, Critical Care and Pain Medicine, Harvard \nMedical School, Massachusetts General Hospital. \nXiaojuan Zhou — Department of Anesthesia, Critical Care and Pain Medicine, Massachusetts \nGeneral Hospital. \n* Corresponding author \nkwmiller@mgh.harvard.edu \n \nKEYWORDS: GABAAR; Allopregnanolone; Reversal Agent; Spiro-hydantoin; Allosterism. \n  \n105 and is also made available for use under a CC0 license. \n(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC \nThe copyright holder for this preprintthis version posted April 17, 2026. ; https://doi.org/10.64898/2026.04.14.718525doi: bioRxiv preprint \n\n2 \nABSTRACT \nThe neurosteroid allopregnanolone is a positive allosteric modulator of GABA(A) receptors, \nwhich has proved beneficial in the treatment of major depressive disorder and epilepsies. It also \nhas a role in treating the mood swings that are associated with fluctuations in its level during the \nmenstrual cycle. Nonetheless, a subset of women do not tolerate high levels of allopregnanolone. \nIso-allopregnanolone, a negative allosteric modulator, as well as synthetic steroid antagonists are \nused to treat such conditions. However, steroid-based medications are difficult to deliver and \ntheir specificity of action can be unclear. Recently introduced novel nonsteroidal agents that, like \niso-allopregnanolone, can reverse the action of positive allosteric modulators without changing \nthe positive action of GABA, might provide an alternative. We surveyed a number of them on \nhuman α1β3δ GABAARs using a [3H]muscimol binding assay. A 6-membered ring spiro-\nhydantoin, DKD99, allosterically reversed the positive allosteric action of allopregnanolone over \na wide concentration range (6 to 1,000 nM). DKD99 shifted allopregnanolone’s modulation \ncurve 10-fold to the right. Furthermore, it has a much lower affinity when exerting similar \nactions on α1β3γ2 receptors. Agents such as this have utility for elucidating underlying \nmechanisms and may offer an alternative pathway for the development of nonsteroidal therapies \nagainst the positive allosteric modulatory actions of neurosteroids. \n  \n105 and is also made available for use under a CC0 license. \n(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC \nThe copyright holder for this preprintthis version posted April 17, 2026. ; https://doi.org/10.64898/2026.04.14.718525doi: bioRxiv preprint \n\n3 \nINTRODUCTION \nAllopregnanolone ( 5α-Pregnan-3α-ol-20-one) is an endogenous neurosteroid that regulates \naffective disorders via its positive allosteric modulator actions on GABAARs.1 It has attracted the \nattention of pharmacologists because of its therapeutic potential in treating depression and \nepilepsy.2-5 This has further lead to the development of Brexanalone, a formulation of \nallopregnanolone for intravenous delivery and Zuranalone, a steroid that can be taken orally.6, 7  \nHowever, the physiological concentration of allopregnanolone fluctuates in response to \nvarious inputs and this can affect mood. Nowhere is this more apparent than in the menstrual \ncycle. In post-menstrual dysphoric disorder (PMDD) the level of allopregnanolone fluctuates \nthroughout the cycle and is associated with mood swings, irritability, anxiety and depression. 2, 8-\n10 In PMDD the role of fluctuating allopregnanolone levels and their link to behavior is complex, \nwith levels rising by an order of magnitude between the follicular and luteal phases. It is then \nparadoxical that in a subset of women PMDD is experienced during this phase. However, the \nneurosteroid iso-allopregnanolone (5β-Pregnan-3α-ol-20-one ), an endogenous  negative \nallosteric modulator, also fluctuates during the cycle.11, 12 Although the mechanism remains \nunclear and may involve fluctuations in the isoforms of the GABAAR involving the δ-subunit, \nthis has led to the introduction of negative allosteric modulator therapeutics such as Sepranolone \nand Golexanalone (GR-3027).1, 5, 7, 12-14 Research with such GABAA receptor modulating steroid \nantagonists (GAMSAs) has revealed that they may also have a therapeutic role to play in diverse \netiologies, often related to neuro-inflammation, such as cognitive function, hepatic \nencephalopathy and motor incoordination.15-17 \nNeurosteroids act on most isoforms of the GABAAR but are particularly effective on δ-\nsubunit containing receptors partly because GABA is a partial agonist and the potential for \nenhancing action is corresponding high.18, 19 Studies of δ-subunit containing extrasynaptic \nreceptors in heterologous systems have been bedeviled by promiscuous assembly in α4/6βxδ \nreceptors. 20-23 However, the α1βxδ subunits are thought to assemble homogeneously as in α1βγ2 \nreceptors (β–α1–δ–β–α1) and were therefore chosen for this study. 24-26 \n105 and is also made available for use under a CC0 license. \n(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC \nThe copyright holder for this preprintthis version posted April 17, 2026. ; https://doi.org/10.64898/2026.04.14.718525doi: bioRxiv preprint \n\n4 \n Recently, a series of spiro-barbiturates and \nspiro-hydantoins have been assayed for their \nability to reverse the action of positive \nallosteric modulators on synaptic (α1β3γ2) and \nextra-synaptic (α1β3) GABAARs. Some of \nthese agents acted negative allosteric \nmodulators of positive allosteric modulators \nincluding the steroid general anesthetic \nalphaxalone but. as null allosteric ligands of \northosteric agonist binding. They have been \ntermed reversal agents and some progress has \nbeen made towards discovering their \nmechanism of action.27-30 Their actions \nresemble those reported for iso-\nallopregnanolone 12, 31. In this report, we have \nevaluated a subset of these compounds on \nheterologously expressed human full length \nα1β3δ GABAARs (Fig. 1). Although, the \nstructure activity relationships differed from \ntheir action on α1β3γ2 receptors, two of them \nselectively reversed allopregnanolone’s \npositive allosteric modulation at low \nmicromolar concentrations that had no action \non α1β3γ2 receptors. \n  \n \nFigure 1. The chemical structures of the ligands \nstudied. \nTop row: Allopregnanolone or 3α-Hydroxy-5α-\npregnan-20-one. \nMiddle row: Spiro-barbiturate reversal agents. \nBottom row: Spiro-hydantoin reversal agents. \n105 and is also made available for use under a CC0 license. \n(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC \nThe copyright holder for this preprintthis version posted April 17, 2026. ; https://doi.org/10.64898/2026.04.14.718525doi: bioRxiv preprint \n\n5 \nRESULTS \nPrinciple of the assay. We assayed for reversal activity at \nequilibrium using an [3H]muscimol binding assay based on the \nobservation that the GABAAR exists in a dynamic equilibrium \nbetween a low affinity resting state and a smaller fraction of \nreceptors in a high affinity desensitized state.32 By using a low \nconcentration of [3H]muscimol that mainly binds to the small \nfraction of high affinity receptors, the action of positive \nallosteric modulators (PAMs), such as neurosteroids and \ngeneral anesthetics, to stabilize the desensitized state can be \nobserved. The assay is a convenient way to test for agents that \nreverse positive allosteric modulator action. It has been \nemployed in αβγ and αβ receptors with success, but this is the \nfirst time it has been employed in α1β3δ receptors. \nDo reversal agents interact with the orthosteric \nagonist site? It is important to ensure that reversal agents \ndo not displace [3H]muscimol from its binding site. Each \nreversal agent in Fig. 1 was titrated between 0.1 and 100 \nμM against 3 nM [3H]muscimol binding in α1β3δ \nGABAARs. None decreased [3H]muscimol binding and \none, BWC10, was a weak PAM, enhancing \n[3H]muscimol from 100 to 125 ± 4.5%. This is of little \nfunctional significance compared to the maximum \nenhancement caused by 100 μM etomidate of 441 ± 19% \n(n = 22).  \nThe \nstructural \ndependence \nof reversal \naction. We \ntested the \nability of the reversal agents in Fig. 1, to reverse the \npositive allosteric action of 100 nM \nallopregnanolone on α1β3δ GABAARs with the \nspecific goal of discovering a reversal agent with \ngood efficacy, and an IC50 in the low micromolar \nrange. We chose 100 nM allopregnanolone for this \nsurvey because it enhanced [3H]binding by 288 ± \n13% (n = 6) making it easier to detect reversal \nactivity.  \nNone of the three spiro barbiturates reversed \nallopregnanolone’s enhancing action with the \nexception of DKB21, which was inactive at 30 μM \nbut at 100 μM modestly reduced allopregnanolone’s \n \nFigure 3. DKD99 reverses \nallopregnanolone’s enhancement of \n[3H]muscimol binding over a wide \nconcentration range. Each point determined in \ntriplicate. \n \nFigre 1, & more 20260318.cvd \n \nFigure 2. DKD29 partially reverses \nallopregnanolone’s PAM action on α1β3δ \nGABAARs without displacing \n[3H]muscimol binding. N =3 at each \nconcentration. Standard deviations are \nshown when larger than symbols. \n \n400\n350\n300\n250\n200\n150\n100\n0.01 0.1 1 10 100\nDKD29, µM\nPlus 100 nM 3α5αP\nControl\nTable 1. Reversal agents have \nlittle action on [3H]muscimol \nbinding. \nAgent Average ± SD N \nDKD21 100 ± 0.02 15 \nBWC10 118 ± -\n12.7 17 \nDKD29 99 ± 5.1 23 \nDKD99 103 ± 4.9 27 \nControl = 100 \n \n105 and is also made available for use under a CC0 license. \n(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC \nThe copyright holder for this preprintthis version posted April 17, 2026. ; https://doi.org/10.64898/2026.04.14.718525doi: bioRxiv preprint \n\n6 \nenhancing action by 10 % (n = 5, p= 0.02).  \nIn contrast, both of the spiro-hydantoins tested reversed allopregnanolone’s action. The 5-\nmembered ring hydantoin, DKD29, modestly reversed allopregnanolone-enhanced \n[3H]muscimol binding from 325 to 253% without displacing [3H]muscimol binding itself (Fig. \n2), whereas the 6-membered ring hydantoin, DKD99, was twice as efficacious. Although \nefficacy was dependent on ring size, potency was not, and both agents had similar IC50s of 5 μM \n(Table 2). Encouragingly, this compares to IC50s of 40 μM in synaptic α1β3γ2 GABAARs \nindicating that this reversal action is quite selective for extrasynaptic α1β3δ receptors. \nBecause of its higher efficacy we studied DKD99 in more detail. Allopregnanolone’s \nenhancement of [3H]muscimol binding was reversed by DKD99 at 6, 100 and 1,000 nM. At both \n6 and 100 nM allopregnanolone reversal was close to complete, but at 1 μM allopregnanolone \nreversal was far from complete perhaps indicating a ceiling effect in the allosteric interaction \nbetween the allopregnanolone sites and the reversal site(s) (Table 2, Fig. 3). \nAllopregnanolone enhances \n[3H]muscimol binding over a wide range \nof concentrations. Allopregnanolone \nenhanced [3H]muscimol binding in a \nconcentration-dependent manner with \nenhancement reaching 20% between 0.1 \nand 0.3 nM and plateauing at 370% at ≥10 \nμM, which is 86% of that for etomidate \n(Fig. 4). Etomidate and steroids both bind \nin the same β+/α– interfaces in the \ntransmembrane domain but etomidate is \nsituated closer to the orthosteric agonist \nsite than steroids. This may offer an \nexplanation for the difference in \nenhancing efficacy. The enhancement \ncurve had a midpoint of 14 nM and a Hill \ncoefficient of 0.5 (Table 3).  \nDKD99 shifts allopregnanolone’s \nenhancement curve to the right. Reversal \nagents that act allosterically are expected \nTable 2. Reversal curves for allopregnanolone’s enhancement of [3H]muscimol binding \nReversal \nagent \n3α5αP \nnM \n1C50 \nμM ± SD  Max ± SD  Min ± SD  Normalized \nEfficacy ± SD \nDKD99 6 2.2 ± 0.3  219 ± 2.7  92 ± 3.1  0.58 ± 0.01 \nDKD99 100 5.0 ± 0.6  264 ± 2.3  115 ± 3.6  0.56 ± 0.01 \nDKD99 1,000 5.2 ± 1.7  331 ± 5.3  214 ± 8.6  0.35 ± 0.01 \nDKD29 100 4.5 ± 1.1  327 ± 2.6  253 ± 4.0  0.23 ± 0.01 \nNormalized efficacy is the difference between the maximum and minimum enhancement \nnormalized to the maximum enhancement. \n \nOriginal in α1β3δ manuscript 20260112.key, slide 27 on 4/2/26. \n \nFigure 4. In α1β3δ GABAARs, the concentration-\ndependence allopregnanolone’s enhancement of \n[3H]muscimol binding shifted to the right and reduced in \namplitude in the presence of the reversal agent DKD99 (50 \nμM). The curves are fits to a two site Adair equation. \nNumber of data points: Control, 89; +DKD99, 53. \n105 and is also made available for use under a CC0 license. \n(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC \nThe copyright holder for this preprintthis version posted April 17, 2026. ; https://doi.org/10.64898/2026.04.14.718525doi: bioRxiv preprint \n\n7 \nto shift this curve to the right. 27 To test this we repeated the above titration in the presence of a \nfixed concentration (50 μM) of the spiro-hydantoin DKD99. DKD99 reversed \nallopregnanolone’s action over the whole concentration range (6 to 1,000 nM). Compared to the \ncontrol curve, DKD99 shifted allopregnanolone’s enhancement curve 10-fold to the right and \nlowered its maximum some 30% without changing its Hill coefficient. The low Hill coefficient \nsuggests that there is more than one site or process underlying the enhancement. \nInterpretation of the low Hill coefficient. The goal of this study was to test whether reversal \nagents can reverse the action of allopregnanolone on α1β3δ receptors so we did not seek the \ncause of the low Hill coefficient. It has been claimed that in α1β2γ2 receptors there are three \nnoninteracting steroid sites that act independently.33However, a single particle Cryo–EM \nstructure of α1β2γ2 receptors in the presence of GABA and allopregnanolone shows it bound to \nthe two classic sites in the β+/α– interface in the transmembrane domain. 34 In the absence of a \nstructure for the α1β3δ receptor, the low Hill coefficient could be interpreted in several ways. If \nthere are two or more different allopregnanolone sites, they could either act independently but \nhave different affinities or they could have similar affinities but interact with negative \nallosterism. Alternatively, there could be two different states or conformations that have high \naffinity for [3H]muscimol but different affinities for allopregnanolone. We will call the two \nactions “components” to avoid implying a mechanism.  \nWe chose simply to fit the data to a two independent binding site model to provide a robust \ndescription that aims to deconvolute the two phases of action. This model is sometimes referred \nto as a two site Adair equation.  \nf(A) = Min+(Max–Min)*( (fract1*(A/(A+K1))+(1–fract1)*(A/(A+K2)))) \nwhere Max and Min are the respective amplitudes at zero and the plateau allopregnanolone \nconcentration, and A is the concentration of allopregnanolone. K1 and K2 are the dissociation \nconstants of the two sites or conformations and fract1 is the fractional population of site or \nconformation 1.  \nTable 3. The concentration-dependence of allopregnanolone’s enhancement of [3H]muscimol binding in \nα1β3δ GABAARs in the absence and presence of 50 μM of the reversal agent DKD99. \nEquation Parameter α1β3δ ± SD  N  \nα1β3δ \n+ \n50  μM \nDKB99 \n± SD  N \nHill EC50 14 ± 1.9 nM 78  141 ± 51 nM 52 \n nH 0.51 ± 0.03    0.46 ± 0.05   \n Max 378 ± 5.1    287 ± 10 %  \n             \nAdair K1 1.3 ± 0.23 nM 78  0.85 ± 0.29 nM 52 \n K2 148 ± 32 nM   629 ± 106 nM  \n Fraction \nSite 1 0.52 ± 0.03    0.28 ± 0.02   \n Max 371 ± 3.3 %   285 ± 3.8 %  \nThe concentration of [3H]muscimol was 3 nM. Each curve is the combined data from four experiments. \n \n \nSlide 9, α1β3δ manuscript 20260112.key \n \n105 and is also made available for use under a CC0 license. \n(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC \nThe copyright holder for this preprintthis version posted April 17, 2026. ; https://doi.org/10.64898/2026.04.14.718525doi: bioRxiv preprint \n\n8 \nThis analysis deconvoluted allopregnanolone’s concentration-response curve into a high \naffinity component with a dissociation constant of 1 nM and a low affinity component whose \ndissociation constant was some 100-fold higher. The population of these two components was \ndistributed equally (Table 2).  \nFitting the data to a two site Adair equation revealed that the high and low affinity sites \nreacted to DKD99 differently, which supports the idea that they represent different states. \nDKD99 acted on allopregnanolone’s high affinity site to decrease its fractional contribution \nwithout changing its EC50. In contrast, DKD99 shifted allopregnanolone’s low affinity site’s \nEC50 4-fold to higher concentrations without changing its overall contribution to enhancement. \nThat is, the decrease in overall enhancement originates entirely from DKD99’s action on \nallopregnanolone’s high affinity site, and the right shift from its low affinity site. This \nconclusion, unlike the Hill equation’s description, is model-dependent and determining the \nmechanism of these unexpected functional actions of DKD99 will require more detailed work \nincluding structural studies. \nCONCLUSION \nThe spiro-hydantoin reversal agent, DKD99, reverses with micromolar potency and good \nefficacy allopregnanolone’s positive allosteric action on extrasynaptic α1β3δ GABAARs over a \nwide concentration range. Previously, reversal of allopregnanolone’s positive allosteric actions \nhas only been accomplished with steroid antagonists that can be challenging to formulate for \nclinical use and whose metabolism may have downstream effects. Nonsteroidal agents such as \nDKD99 may both prove useful in dissecting mechanisms in more complex systems and provide a \nnovel starting point for therapeutic development. \nEXPERIMENTAL METHODS \nCell line creation. A new α1β3δ inducible HEK293 cell line was created using constructs for \nhuman full length β3- and δ-subunits that have been previously described 35-37. A new full length \nhuman α1-subunit (Uniprot P14867) construct was synthesized (Synbio technologies, Inc) \nbearing an N-terminal twin strep tag \n(YPYDVPDYAGGSWSHPQFEKGGGSGGGSGGSAWSHPQFEK) inserted five amino acid \nafter the α1-subunit signal peptide cleavage site and followed by a flexible linker GGS. The α1-\ntwin strep tagged subunit was cloned into Doxycycline inducible expression plasmid \npCDNA4TO (EcoRI-XhoI), with selection marker Zeocin (See Supplemental Materials).  \nThe α1-, β3- and δ- constructs were stably transfected into a HEK293T TetR cell line with \nLipofectamine 2000 reagent as directed by the manufacturer’s protocol. The transfection ratio of \nα1- and δ-subunits being 2 : 0.25 and using ratios for the β3-subunit of either 0.3 or 1. At 48 \nhours post-transfection cells were trypsinized and plated in the presence of Zeocin 250 μg/ml, \nHygromycin 50 μg/ml and G418 100u μg/ml with 5 μg/ml Blasticidin for one week. The selected \npools were then sorted twice by flow cytometry for GABA receptor expression using APC \nconjugated anti-1D4 antibodies against the 1D4-tagged delta subunit as a surrogate for a full \npentameric expression at the cell surface. The twice enriched pool was then used to isolate clones \nby single cell plating in a 96 well plate and antibiotic selection for another week. The 18 best \n105 and is also made available for use under a CC0 license. \n(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC \nThe copyright holder for this preprintthis version posted April 17, 2026. ; https://doi.org/10.64898/2026.04.14.718525doi: bioRxiv preprint \n\n9 \ngrowing clones were then analyzed by flow cytometry using the same methodology and the best \nclones were further characterized by [3H]muscimol binding assay. \nSelection and pharmacological characterization of clones. Judged by FACS the lower ratio of \nβ3-subunit gave better yields overall, but individual clones from both transfection ratios gave \ngood specific activities. The clone selected was from the 2:0.3:0.25 (α1:β3:δ) group. It had a \ngood growth rate and a specific activity of 17 ± 2 pmol of muscimol binding sites/mg of \nmembrane protein, comparable to a similar cell line expressing α1β3γ2 receptors.36  DS2 is a δ-\nsubunit selective ligand.38 It strongly enhanced [3H]muscimol binding  by 315 ± 25% in α1β3δ \nreceptors compared to 8 ± 3% in α1β3γ2 receptors. We estimated the dissociation constant of \nGABA by titrating it against 3 nM [3H]muscimol, a concentration chosen as a compromise \nbetween low occupancy and having sufficient cpm. The titration from 1 nM to 1 mM yielded an \nIC50 of 150 ± 10 nM (n = 32; 3 separate experiments). A similar experiment with α1β3γ2 \nreceptors yielded an IC50 of 75 ± 9 μM (n = 34; 2 separate experiments). The difference in \nIC50s is consistent with the higher agonist affinity of α1β3δ over α1β3γ2receptors. \nThe [3H]muscimol binding assay. The methods have been recently described.27 Briefly, \nbinding assays were performed on cell membranes harvested from HEK cells. [3H]muscimol \n(PerkinElmer) was used at 3 nM final concentration and nonspecific binding was corrected for \nby displacement of specific binding by 1 mM GABA. Samples were made in 7 mL glass sample \nvials. Allopregnanolone was added to glass sample vials after serial dilution from a 20 mM stock \nin DMSO. Samples were equilibrated for 30 min and filtered on GFB filters, which were dried \nunder a lamp before being added to 5 mL of scintillation cocktail and counted. Analysis was \ncarried out using Wavemetrics’s Igor software. \nAcknowledgements  \nThis research was supported by a grant from the National Institutes of Health, USA, R01 \nGM135550, and with support from the Department of Anesthesia, Critical Care and Pain \nMedicine, Massachusetts General Hospital, Boston.  \nReferences \n1. MacKenzie G, Maguire J. The role of ovarian hormone-derived neurosteroids on the \nregulation of GABAA receptors in affective disorders. Psychopharmacology (Berl). \n2014;231(17):3333-42. Epub 2014/01/10. doi: 10.1007/s00213-013-3423-z. PMID: 24402140. \n2. Maguire JL, Stell BM, Rafizadeh M, Mody I. Ovarian cycle-linked changes in GABA(A) \nreceptors mediating tonic inhibition alter seizure susceptibility and anxiety. Nat Neurosci. \n2005;8(6):797-804. Epub 20050515. doi: 10.1038/nn1469. PMID: 15895085. \n3. Rogawski MA, Loya CM, Reddy K, Zolkowska D, Lossin C. Neuroactive steroids for the \ntreatment of status epilepticus. Epilepsia. 2013;54 Suppl 6(0 6):93-8. doi: 10.1111/epi.12289. \nPMID: 24001085. \n105 and is also made available for use under a CC0 license. \n(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC \nThe copyright holder for this preprintthis version posted April 17, 2026. ; https://doi.org/10.64898/2026.04.14.718525doi: bioRxiv preprint \n\n10 \n4. Blanco MJ, La D, Coughlin Q, Newman CA, Griffin AM, Harrison BL, Salituro FG. \nBreakthroughs in neuroactive steroid drug discovery. Bioorg Med Chem Lett. 2018;28(2):61-70. \nEpub 20171202. doi: 10.1016/j.bmcl.2017.11.043. PMID: 29223589. \n5. Maguire JL, Mennerick S. Neurosteroids: mechanistic considerations and clinical \nprospects. Neuropsychopharmacology. 2024;49(1):73-82. Epub 20230627. doi: 10.1038/s41386-\n023-01626-z. PMID: 37369775. \n6. Clayton AH, Lasser R, Parikh SV , Iosifescu DV , Jung J, Kotecha M, Forrestal F, Jonas J, \nKanes SJ, Doherty J. Zuranolone for the Treatment of Adults With Major Depressive Disorder: A \nRandomized, Placebo-Controlled Phase 3 Trial. Am J Psychiatry. 2023;180(9):676-84. Epub \n20230503. doi: 10.1176/appi.ajp.20220459. PMID: 37132201. \n7. Singhal M, Modi N, Bansal L, Abraham J, Mehta I, Ravi A. The Emerging Role of \nNeurosteroids: Novel Drugs Brexanalone, Sepranolone, Zuranolone, and Ganaxolone in Mood \nand Neurological Disorders. Cureus. 2024;16(7):e65866. Epub 20240731. doi: \n10.7759/cureus.65866. PMID: 39219949. \n8. Longone P, Rupprecht R, Manieri GA, Bernardi G, Romeo E, Pasini A. The complex \nroles of neurosteroids in depression and anxiety disorders. Neurochem Int. 2008;52(4-5):596-\n601. Epub 20071006. doi: 10.1016/j.neuint.2007.10.001. PMID: 17996986. \n9. Mishra S, Elliott H, Marwaha R. Premenstrual Dysphoric Disorder.  StatPearls. Treasure \nIsland (FL)2025. \n10. Backstrom T, Bixo M, Johansson M, Nyberg S, Ossewaarde L, Ragagnin G, Savic I, \nStromberg J, Timby E, van Broekhoven F, van Wingen G. Allopregnanolone and mood disorders. \nProg Neurobiol. 2014;113:88-94. Epub 20130823. doi: 10.1016/j.pneurobio.2013.07.005. PMID: \n23978486. \n11. Hantsoo L, Epperson CN. Allopregnanolone in premenstrual dysphoric disorder \n(PMDD): Evidence for dysregulated sensitivity to GABA-A receptor modulating neuroactive \nsteroids across the menstrual cycle. Neurobiol Stress. 2020;12:100213. Epub 20200204. doi: \n10.1016/j.ynstr.2020.100213. PMID: 32435664. \n12. Backstrom T, Das R, Bixo M. Positive GABA(A) receptor modulating steroids and their \nantagonists: Implications for clinical treatments. J Neuroendocrinol. 2022;34(2):e13013. Epub \n20210801. doi: 10.1111/jne.13013. PMID: 34337790. \n13. Bixo M, Ekberg K, Poromaa IS, Hirschberg AL, Jonasson AF, Andreen L, Timby E, \nWulff M, Ehrenborg A, Backstrom T. Treatment of premenstrual dysphoric disorder with the \nGABA(A) receptor modulating steroid antagonist Sepranolone (UC1010)-A randomized \ncontrolled trial. Psychoneuroendocrinology. 2017;80:46-55. Epub 20170301. doi: \n10.1016/j.psyneuen.2017.02.031. PMID: 28319848. \n14. Thompson SM. Modulators of GABA(A) receptor-mediated inhibition in the treatment of \nneuropsychiatric disorders: past, present, and future. Neuropsychopharmacology. 2024;49(1):83-\n95. Epub 20230914. doi: 10.1038/s41386-023-01728-8. PMID: 37709943. \n15. Backstrom T, Doverskog M, Blackburn TP, Scharschmidt BF, Felipo V . Allopregnanolone \nand its antagonist modulate neuroinflammation and neurological impairment. Neurosci Biobehav \nRev. 2024;161:105668. Epub 20240410. doi: 10.1016/j.neubiorev.2024.105668. PMID: \n38608826. \n16. Llansola M, Mincheva G, Arenas YM, Izquierdo-Altarejos P, Pedrosa MA, Blackburn TP, \nBackstrom T, Scharschmidt BF, Doverskog M, Felipo V . Golexanolone Attenuates \nNeuroinflammation, Fatigue, and Cognitive and Motor Impairment in Diverse \n105 and is also made available for use under a CC0 license. \n(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC \nThe copyright holder for this preprintthis version posted April 17, 2026. ; https://doi.org/10.64898/2026.04.14.718525doi: bioRxiv preprint \n\n11 \nNeuroinflammatory Disorders. Pharmaceuticals (Basel). 2025;18(11). Epub 20251118. doi: \n10.3390/ph18111757. PMID: 41304999. \n17. Yilmaz C, Karali K, Fodelianaki G, Gravanis A, Chavakis T, Charalampopoulos I, \nAlexaki VI. Neurosteroids as regulators of neuroinflammation. Front Neuroendocrinol. \n2019;55:100788. Epub 20190909. doi: 10.1016/j.yfrne.2019.100788. PMID: 31513776. \n18. Stell BM, Brickley SG, Tang CY , Farrant M, Mody I. Neuroactive steroids reduce \nneuronal excitability by selectively enhancing tonic inhibition mediated by delta subunit-\ncontaining GABAA receptors. Proc Natl Acad Sci U S A. 2003;100(24):14439-44. Epub \n2003/11/19. doi: 10.1073/pnas.2435457100. PMID: 14623958. \n19. Wohlfarth KM, Bianchi MT, Macdonald RL. Enhanced neurosteroid potentiation of \nternary GABA(A) receptors containing the delta subunit. J Neurosci. 2002;22(5):1541-9. doi: \n10.1523/JNEUROSCI.22-05-01541.2002. PMID: 11880484. \n20. Kaur KH, Baur R, Sigel E. Unanticipated structural and functional properties of delta-\nsubunit-containing GABAA receptors. J Biol Chem. 2009;284(12):7889-96. Epub 20090113. \ndoi: 10.1074/jbc.M806484200. PMID: 19141615. \n21. Eaton MM, Bracamontes J, Shu HJ, Li P, Mennerick S, Steinbach JH, Akk G. gamma-\naminobutyric acid type A alpha4, beta2, and delta subunits assemble to produce more than one \nfunctionally distinct receptor type. Mol Pharmacol. 2014;86(6):647-56. Epub 2014/09/23. doi: \n10.1124/mol.114.094813. PMID: 25238745. \n22. Wongsamitkul N, Baur R, Sigel E. Toward Understanding Functional Properties and \nSubunit Arrangement of alpha4beta2delta gamma-Aminobutyric Acid, Type A (GABAA) \nReceptors. J Biol Chem. 2016;291(35):18474-83. Epub 2016/07/07. doi: \n10.1074/jbc.M116.738906. PMID: 27382064. \n23. Sente A, Desai R, Naydenova K, Malinauskas T, Jounaidi Y , Miehling J, Zhou X, \nMasiulis S, Hardwick SW, Chirgadze DY , Miller KW, Aricescu AR. Differential assembly \ndiversifies GABAA receptor structures and signalling. Nature. 2022;604(7904):190-4. Epub \n2022/04/01. doi: 10.1038/s41586-022-04517-3. PMID: 35355020. \n24. Botzolakis EJ, Gurba KN, Lagrange AH, Feng HJ, Stanic AK, Hu N, Macdonald RL. \nComparison of gamma-Aminobutyric Acid, Type A (GABAA), Receptor alphabetagamma and \nalphabetadelta Expression Using Flow Cytometry and Electrophysiology: Evidence for \nalternative subunit stoichiometries and arrangements. J Biol Chem. 2016;291(39):20440-61. \nEpub 2016/08/06. doi: 10.1074/jbc.M115.698860. PMID: 27493204. \n25. Feng HJ, Forman SA. Comparison of alphabetadelta and alphabetagamma GABAA \nreceptors: Allosteric modulation and identification of subunit arrangement by site-selective \ngeneral anesthetics. Pharmacol Res. 2018;133:289-300. Epub 2018/01/03. doi: \n10.1016/j.phrs.2017.12.031. PMID: 29294355. \n26. Liao VWY , Chebib M, Ahring PK. Efficient expression of concatenated alpha1beta2delta \nand alpha1beta3delta GABAA receptors, their pharmacology and stoichiometry. Br J Pharmacol. \n2021;178(7):1556-73. Epub 2021/01/26. doi: 10.1111/bph.15380. PMID: 33491192. \n27. Koinas D, Zhou X, Wu B, Miller KW, Bruzik KS. Novel Spiro-Barbiturates Can Reverse \nthe Action of General Anesthetics on the GABA(A)R. J Med Chem. 2025;68(8):8025-45. Epub \n20250407. doi: 10.1021/acs.jmedchem.4c02514. PMID: 40193703. \n28. Koinas D, Zhou X, Wu B, Bruzik KS, Miller KW. Spiro Hydantoins Can Reverse the \nAction of Positive Allosteric Modulators on GABAARs. ACS Med Chem Lett. 2025;16:2078–\n83. Epub 06 May 2025. doi: https://doi.org/10.1021/acsmedchemlett.5c00499. \n105 and is also made available for use under a CC0 license. \n(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC \nThe copyright holder for this preprintthis version posted April 17, 2026. ; https://doi.org/10.64898/2026.04.14.718525doi: bioRxiv preprint \n\n12 \n29. Miehling J. Mechanism of anaesthetic activation, combination and antagonism [Thesis]. \nApollo–University of Cambridge: University of Cambridge; 2022. \nhttps://doi.org/10.17863/CAM.87926 \n30. Zuo Y , Zhao Y , Liu G, Sun Q. Recent advances in GABA(A) receptor targeting ligands. \nEur J Med Chem. 2026;308:118651. Epub 20260205. doi: 10.1016/j.ejmech.2026.118651. \nPMID: 41719803. \n31. Stromberg J, Lundgren P, Taube M, Backstrom T, Wang M, Haage D. The effect of the \nneuroactive steroid 5beta-pregnane-3beta, 20(R)-diol on the time course of GABA evoked \ncurrents is different to that of pregnenolone sulphate. Eur J Pharmacol. 2009;605(1-3):78-86. \nEpub 20090110. doi: 10.1016/j.ejphar.2008.12.038. PMID: 19168059. \n32. Chang Y , Ghansah E, Chen Y , Ye J, Weiss DS. Desensitization mechanism of GABA \nreceptors revealed by single oocyte binding and receptor function.[erratum appears in J Neurosci \n2002 Oct 15;22(20):1b Note: Chang YongChang [corrected to Chang Yongchang]]. Journal of \nNeuroscience. 2002;22(18):7982-90. \n33. Germann AL, Pierce SR, Tateiwa H, Sugasawa Y , Reichert DE, Evers AS, Steinbach JH, \nAkk G. Intrasubunit and Intersubunit Steroid Binding Sites Independently and Additively \nMediate alpha1beta2gamma2L GABA(A) Receptor Potentiation by the Endogenous \nNeurosteroid Allopregnanolone. Mol Pharmacol. 2021;100(1):19-31. Epub 20210506. doi: \n10.1124/molpharm.121.000268. PMID: 33958479. \n34. Legesse DH, Fan C, Teng J, Zhuang Y , Howard RJ, Noviello CM, Lindahl E, Hibbs RE. \nStructural insights into opposing actions of neurosteroids on GABA(A) receptors. Nat Commun. \n2023;14(1):5091. Epub 20230822. doi: 10.1038/s41467-023-40800-1. PMID: 37607940. \n35. Dostalova Z, Liu A, Zhou X, Farmer SL, Krenzel ES, Arevalo E, Desai R, Feinberg-\nZadek PL, Davies PA, Yamodo IH, Forman SA, Miller KW. High-level expression and \npurification of Cys-loop ligand-gated ion channels in a tetracycline-inducible stable mammalian \ncell line: GABAA and serotonin receptors. Protein Sci. 2010;19(9):1728-38. Epub 2010/07/28. \ndoi: 10.1002/pro.456. PMID: 20662008. \n36. Dostalova Z, Zhou X, Liu A, Zhang X, Zhang Y , Desai R, Forman SA, Miller KW. \nHuman alpha1beta3gamma2L gamma-aminobutyric acid type A receptors: High-level production \nand purification in a functional state. Protein Sci. 2014;23(2):157-66. Epub 2013/11/30. doi: \n10.1002/pro.2401. PMID: 24288268. \n37. Zhou X, Desai R, Zhang Y , Stec WJ, Miller KW, Jounaidi Y . High-level production and \npurification in a functional state of an extrasynaptic gamma-aminobutyric acid type A receptor \ncontaining alpha4beta3delta subunits. PLoS One. 2018;13(1):e0191583. Epub 2018/01/21. doi: \n10.1371/journal.pone.0191583. PMID: 29352320. \n38. Jensen ML, Wafford KA, Brown AR, Belelli D, Lambert JJ, Mirza NR. A study of \nsubunit selectivity, mechanism and site of action of the delta selective compound 2 (DS2) at \nhuman recombinant and rodent native GABA(A) receptors. Br J Pharmacol. 2013;168(5):1118-\n32. Epub 2012/10/16. doi: 10.1111/bph.12001. PMID: 23061935. \n  \n105 and is also made available for use under a CC0 license. \n(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC \nThe copyright holder for this preprintthis version posted April 17, 2026. ; https://doi.org/10.64898/2026.04.14.718525doi: bioRxiv preprint \n\n13 \nSupplemental Materials \nHuman α1 N-Terminal Twin-strep based on Uniprot # P14867 \nRed is the signal peptide. \nGreen is the HA tag 9-amino acid peptide epitope (YPYDVPDYA) derived from the human \ninfluenza hemagglutinin (HA) molecule. \nBlue is the two Twin Strep tags (WSHPQFEK) connected by a flexible linker. \nYellow are flexible linkers. \nProtein sequence: \nMRKSPGLSDCLWAWILLLSTLTGRSYGQPSLQYPYDVPDYAGGSWSHPQFEKGGGSGGGSGGS\nAWSHPQFEKGGSDELKDNTTVFTRILDRLLDGYDNRLRPGLGERVTEVKTDIFVTSFGPVSDHDMEYTIDV\nFFRQSWKDERLKFKGPMTVLRLNNLMASKIWTPDTFFHNGKKSVAHNMTMPNKLLRITEDGTLLYTMRLTV\nRAECPMHLEDFPMDAHACPLKFGSYAYTRAEVVYEWTREPARSVVVAEDGSRLNQYDLLGQTVDSGIVQSST\nGEYVVMTTHFHLKRKIGYFVIQTYLPCIMTVILSQVSFWLNRESVPARTVFGVTTVLTMTTLSISARNSLPKVAY\nATAMDWFIAVCYAFVFSALIEFATVNYFTKRGYAWDGKSVVPEKPKKVKDP \nLIKKNNTYAPTATSYTPNLARGDPGLATIAKSATIEPKEVKPETKPPEPKKTFNSVSKID \nRLSRIAFPLLFGIFNLVYWATYLNREPQLKAPTPHQ*  \nNucleotide sequence: \nATGAGAAAGAGCCCTGGCCTGAGCGATTGTCTGTGGGCCTGGATTCTGCTGCTGA\nGCACCCTGACAGGCAGAAGCTATGGCCAGCCTAGCCTGCAGTACCCCTACGACGTGC\nCAGATTATGCCGGCGGATCTTGGAGCCATCCTCAGTTCGAAAAAGGCGGCGGTTCTG\nGCGGTGGATCTGGCGGATCTGCTTGGTCACACCCACAGTTTGAGAAAGGCGGAAGCG\nACGAGCTGAAGGACAACACCACCGTGTTCACCAGAATCCTGGACAGACTGCTGGAC\nGGCTACGACAACAGACTGAGGCCTGGCCTCGGCGAGAGAGTGACCGAAGTCAAGAC\nCGACATCTTCGTGACCAGCTTCGGCCCCGTGTCCGACCACGATATGGAGTACACCATC\nGACGTGTTCTTCCGGCAGAGCTGGAAGGACGAGCGGCTGAAGTTTAAGGGCCCCAT\nGACCGTGCTGCGGCTGAACAATCTGATGGCCAGCAAGATCTGGACCCCTGACACATT\nCTTCCACAACGGCAAGAAAAGCGTGGCCCACAACATGACCATGCCTAACAAGCTGCT\nGCGGATCACCGAGGATGGCACCCTGCTGTACACCATGAGGCTGACAGTCAGAGCCGA\nGTGTCCCATGCACCTGGAAGATTTCCCTATGGACGCCCACGCCTGTCCTCTGAAGTTT\nGGCAGCTACGCCTACACAAGAGCCGAGGTGGTGTACGAGTGGACCAGAGAACCTGC\nCAGATCTGTGGTGGTGGCCGAGGACGGAAGCAGACTGAACCAGTATGATCTGCTGGG\nCCAGACCGTGGACTCTGGCATTGTGCAAAGCAGCACCGGCGAGTACGTGGTCATGAC\nAACCCACTTCCACCTGAAGCGGAAGATCGGCTACTTCGTGATCCAGACCTACCTGCC\nTTGCATCATGACAGTGATCCTGAGCCAGGTGTCCTTCTGGCTGAACCGGGAATCTGTG\nCCTGCCAGAACAGTGTTCGGCGTGACCACCGTGCTGACCATGACCACACTGAGCATC\nAGCGCCAGAAACAGCCTGCCTAAGGTGGCCTACGCCACCGCTATGGACTGGTTTATC\n105 and is also made available for use under a CC0 license. \n(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC \nThe copyright holder for this preprintthis version posted April 17, 2026. ; https://doi.org/10.64898/2026.04.14.718525doi: bioRxiv preprint \n\n14 \nGCCGTGTGCTACGCCTTCGTGTTCAGCGCCCTGATCGAGTTCGCCACCGTGAACTACT\nTCACCAAGAGAGGCTACGCCTGGGACGGCAAGTCTGTGGTGCCAGAGAAGCCCAAG\nAAAGTGAAGGACCCTCTGATCAAGAAGAACAACACATACGCCCCTACCGCCACCAG\nCTACACCCCTAATCTTGCCAGAGGCGATCCTGGCCTGGCCACAATTGCCAAGTCTGCC\nACCATCGAGCCTAAAGAAGTGAAGCCCGAGACAAAGCCTCCTGAGCCTAAGAAAAC\nCTTCAACAGCGTGTCCAAGATCGACCGGCTGAGCCGGATTGCCTTTCCTCTGCTGTTC\nGGCATCTTCAACCTGGTGTACTGGGCCACCTACCTGAACAGAGAGCCCCAGCTGAAA\nGCCCCTACACCTCACCAGTGA \n \nClone selection \n \n \n \n \n \n \n \nFigure S1. Flow cytometry selection analysis of α1-, β3- and δ- constructs stably transfected into a HEK293T TetR \nclonal cell lines using two transfection ratios α1:β3:δ: A, 2:0.3:0.26 or B, 2:1:0.25. \nThe isolated clones were sorted for GABAAR receptor expression using APC conjugated anti-1D4 antibodies against \nthe 1D4-tagged delta subunit. Cells were analyzed at the MGH Flow Cytometry Core facility using a BD 5 laser \nSORP FACS Vantage SE Diva system or Facsaria (BD Biosciences). FACS data and ∑Median statistics were \nanalyzed using FlowJo 10.8.1 software (Tree Star, Inc.). \n \n \nFACSfinl.pdf \n105 and is also made available for use under a CC0 license. \n(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC \nThe copyright holder for this preprintthis version posted April 17, 2026. ; https://doi.org/10.64898/2026.04.14.718525doi: bioRxiv preprint","source_license":"Public-Domain","license_restricted":false}