2,4-Epi-neodysiherbaine

• Molecular Formula: C11H16NO8-
• Molecular Weight: 290.25 g/mol
• SMILES: C1C2C(C(C(CO2)O)O)OC1(CC(C(=O)[O-])[NH3+])C(=O)[O-]
• InChI: InChI=1S/C11H17NO8/c12-4(9(15)16)1-11(10(17)18)2-6-8(20-11)7(14)5(13)3-19-6/h4-8,13-14H,1-3,12H2,(H,15,16)(H,17,18)/p-1/t4-,5+,6?,7+,8?,11-/m1/s1
• InChIKey: NRTJEXLNSCGBJU-LUHIGIFZSA-M

2,4‑Epi‑neodysiherbaine (2,4‑epi‑neoDH) is a synthetic, small‑molecule epimer of the marine natural product neodysiherbaine A. It belongs to the dysiherbaine class of rigidified glutamate analogs and acts as a selective antagonist at ionotropic kainate‑type glutamate receptors (KARs) rather than as an agonist, uniquely targeting both GluK1 (GluR5) and GluK2 (GluR6) subunits without detectable activity at AMPA or other KAR subtypes.

2,4-Epi-neodysiherbaine – A Synthetic Kainate Receptor Antagonist with Dual GluK1/GluK2 Selectivity for Neuroscience Research Procurement

2,4‑Epi‑neodysiherbaine (2,4‑epi‑neoDH) is a synthetic, small‑molecule epimer of the marine natural product neodysiherbaine A [1]. It belongs to the dysiherbaine class of rigidified glutamate analogs and acts as a selective antagonist at ionotropic kainate‑type glutamate receptors (KARs) rather than as an agonist, uniquely targeting both GluK1 (GluR5) and GluK2 (GluR6) subunits without detectable activity at AMPA or other KAR subtypes [2]. Its inverted stereochemistry at C2 and C4 disrupts the L‑glutamate backbone, abolishing agonist efficacy and conferring a pharmacological profile distinct from all other dysiherbaine‑family ligands [3].

• [1] DrugMap Database. 2,4‑epi‑neodysiherbaine (ID: DMZRY0Q). Drug Type: Small molecular drug; Molecular Weight: 290.25; Target: GRIK1 (GluK1), GRIK2 (GluK2). View Source
• [2] Lash LL, Sanders JM, Shoji M, Sasaki M, Sakai R, Swanson GT. Novel Analogs and Stereoisomers of the Marine Toxin Neodysiherbaine with Specificity for Kainate Receptors. J Pharmacol Exp Ther. 2008;324(2):484–496. doi:10.1124/jpet.107.129890. View Source
• [3] Swanson GT, Lash LL, Sakai R. Kainate receptor‑selective epimeric analogs of dysiherbaine. US Patent 7,973,075. Issued July 5, 2011. View Source

Why Generic Kainate Receptor Ligands Cannot Substitute for 2,4‑Epi‑neodysiherbaine in Experimental Protocols

The dysiherbaine scaffold exhibits extreme functional divergence with subtle stereochemical changes: epimerization at a single position (C4) yields a mixed agonist/antagonist, while dual epimerization at C2 and C4 converts the potent convulsant agonist neodysiherbaine (ED₅₀ = 16 pmol/mouse) into a pure antagonist with no residual agonist activity [1]. Other in‑class antagonists such as MSVIII‑19 are GluK1‑selective and inactive at GluK2, whereas 2,4‑epi‑neoDH is the only compound that antagonizes both GluK1 and GluK2 with complete AMPA receptor sparing [2]. These qualitative pharmacological differences—agonist vs. antagonist, single‑subunit vs. dual‑subunit coverage, AMPA‑active vs. AMPA‑silent—mean that no other kainate receptor ligand can replicate the experimental readout of 2,4‑epi‑neoDH in functional studies of heteromeric KARs or in models requiring dual GluK1/GluK2 blockade [3].

• [1] Lash LL, Sanders JM, Shoji M, Sasaki M, Sakai R, Swanson GT. Novel Analogs and Stereoisomers of the Marine Toxin Neodysiherbaine with Specificity for Kainate Receptors. J Pharmacol Exp Ther. 2008;324(2):484–496. doi:10.1124/jpet.107.129890. View Source
• [2] Sanders JM, Ito K, Settimo L, Pentikäinen OT, Shoji M, Sasaki M, Johnson MS, Sakai R, Swanson GT. Divergent pharmacological activity of novel marine‑derived excitatory amino acids on glutamate receptors. J Pharmacol Exp Ther. 2005;314(3):1068–1078. View Source
• [3] Jane DE, Lodge D, Collingridge GL. Kainate receptors: pharmacology, function and therapeutic potential. Neuropharmacology. 2009;56(1):90–113. Table categorizing 2,4‑epi‑neodysiherbaine as GluK1/GluK2‑selective antagonist. View Source

Quantitative Differential Evidence for 2,4‑Epi‑neodysiherbaine vs. Closest Analogs – Procurement Decision Data

Binding Selectivity: Clean GluK1/GluK2 Profile vs. 4‑Epi‑neoDH and MSVIII‑19

In [³H]kainate displacement assays on recombinant human KAR subunits expressed in HEK293 cells, 2,4‑epi‑neoDH bound only to GluR5‑2a (GluK1) and GluR6a (GluK2) with Ki values of 2.4 μM and 7.7 μM, respectively, and showed no detectable affinity (>100 μM) for GluR7a, KA2, or AMPA receptor subunits GluR1–4 [1]. In contrast, its single‑epimer analog 4‑epi‑neoDH, while showing higher affinity for GluR5‑2a (Ki = 559 nM), also displaced radioligand from KA2 (~30 μM) and AMPA subunits GluR1/GluR2 (>30 μM), exhibiting broader off‑target binding [2]. MSVIII‑19 is restricted to GluR5‑2a only (Ki = 128 nM) with no measurable binding to GluR6a [3].

Evidence Dimension	Kainate & AMPA receptor subunit binding selectivity (Ki, μM)
Target Compound Data	GluR5‑2a: Ki = 2.4 μM; GluR6a: Ki = 7.7 μM; GluR7, KA2, AMPA (GluR1–4): all >100 μM
Comparator Or Baseline	4‑epi‑neoDH: GluR5‑2a Ki = 0.559 μM, GluR6a Ki = 6.7 μM, KA2 ~30 μM, AMPA (GluR1–2) >30 μM; MSVIII‑19: GluR5‑2a Ki = 0.128 μM, GluR6a >100 μM; neoDH: GluR5‑2a Ki = 0.0077 μM, GluR6a Ki = 0.033 μM, GluR7 Ki = 0.0006 μM
Quantified Difference	2,4‑epi‑neoDH is the only analog with zero detectable binding to all non‑GluK1/GluK2 subunits. 4‑epi‑neoDH shows 4.3‑fold higher GluR5‑2a affinity but loses selectivity. MSVIII‑19 shows 18.8‑fold higher GluR5‑2a affinity but lacks GluR6a binding entirely.
Conditions	Recombinant human kainate (GluR5‑2a, GluR6a, GluR7a, KA2) and AMPA (GluR1‑i, GluR2‑i) receptor subunits expressed in HEK293 cells; [³H]kainate (KA) and [³H]AMPA radioligand displacement assays; n = 3–4 independent experiments.

Why This Matters

For experiments requiring dual GluK1/GluK2 antagonism without AMPA receptor contamination, 2,4‑epi‑neoDH is the only available tool compound; using 4‑epi‑neoDH or MSVIII‑19 would either introduce AMPA off‑target effects or miss GluK2 blockade entirely.

• [1] Lash LL, Sanders JM, Shoji M, Sasaki M, Sakai R, Swanson GT. Novel Analogs and Stereoisomers of the Marine Toxin Neodysiherbaine with Specificity for Kainate Receptors. J Pharmacol Exp Ther. 2008;324(2):484–496. Table 1, Fig. 6A. View Source
• [2] Lash LL, Sanders JM, Shoji M, Sasaki M, Sakai R, Swanson GT. Ibid. Table 1 entry for 4‑epi‑neoDH; Ki values: GluR5‑2a 559 nM, GluR6a 6.7 μM, KA2 ~30 μM. View Source
• [3] Sanders JM, Ito K, Settimo L, Pentikäinen OT, Shoji M, Sasaki M, Johnson MS, Sakai R, Swanson GT. Divergent pharmacological activity of novel marine‑derived excitatory amino acids on glutamate receptors. J Pharmacol Exp Ther. 2005;314(3):1068–1078. MSVIII‑19 Ki at GluR5 = 128 nM. View Source

Functional Antagonist Activity: Dual GluK1/GluK2 Blockade vs. Agonist Parent Compound NeoDH

In whole‑cell patch‑clamp recordings from HEK293‑T/17 cells expressing recombinant receptors, 2,4‑epi‑neoDH (100 μM) failed to evoke any detectable current from GluR5‑2a or GluR6a receptors, demonstrating zero agonist activity [1]. It instead acted as an antagonist: at 30 μM, it reduced glutamate‑evoked currents from GluR5‑2a by ~70% and from GluR6a by ~30%, with IC₅₀ values of 7.5 μM (GluR5‑2a) and 74 μM (GluR6a) [2]. By contrast, the parent compound neoDH is a full agonist at both subunits (EC₅₀ values not explicitly reported for these subunits but robust currents at 10–100 μM) [3]. Even at 300 μM, 2,4‑epi‑neoDH produced no inhibition of GluR4(i) AMPA receptors, confirming functional AMPA receptor sparing [2].

Evidence Dimension	Functional activity at recombinant kainate receptors (agonist vs. antagonist)
Target Compound Data	Agonist activity: none (0/5 cells at 100 μM). Antagonist IC₅₀: GluR5‑2a = 7.5 μM, GluR6a = 74 μM. AMPA (GluR4) inhibition: 0% at 300 μM.
Comparator Or Baseline	NeoDH: full agonist at GluR5‑2a and GluR6a (currents evoked at 10–100 μM). 4‑epi‑neoDH: weak partial agonist at GluR5‑2a and GluR6a (small currents at 50 μM). MSVIII‑19: GluR5‑2a antagonist only (IC₅₀ ≈ 23 nM); inactive at GluR6a.
Quantified Difference	2,4‑epi‑neoDH is a pure antagonist with ~10‑fold selectivity for GluR5‑2a over GluR6a. NeoDH is a full agonist; 4‑epi‑neoDH is a mixed weak partial agonist; MSVIII‑19 is a potent but GluK1‑only antagonist.
Conditions	Transiently transfected HEK293‑T/17 cells expressing homomeric GluR5‑2a (GluK1), GluR6a (GluK2), or GluR4(i) AMPA receptors; whole‑cell patch‑clamp recordings; glutamate (10 mM) as agonist; n = 3–6 cells per condition.

Why This Matters

2,4‑Epi‑neoDH is the only compound that functionally antagonizes both GluK1 and GluK2 without residual agonist activity or AMPA receptor engagement, making it uniquely suitable for experiments where complete, clean blockade of these two KAR subtypes is required.

• [1] Lash LL, Sanders JM, Shoji M, Sasaki M, Sakai R, Swanson GT. Novel Analogs and Stereoisomers of the Marine Toxin Neodysiherbaine with Specificity for Kainate Receptors. J Pharmacol Exp Ther. 2008;324(2):484–496. Fig. 6C–D; no agonist activity at 100 μM. View Source
• [2] Lash LL, Sanders JM, Shoji M, Sasaki M, Sakai R, Swanson GT. Ibid. Fig. 7A–B; IC₅₀ values: GluR5‑2a 7.5 μM, GluR6a 74 μM; no AMPA inhibition at 300 μM. View Source
• [3] Lash LL, Sanders JM, Shoji M, Sasaki M, Sakai R, Swanson GT. Ibid. NeoDH agonist activity described in Results; 4‑epi‑neoDH weak partial agonist activity at 50 μM. View Source

In Vivo Safety Margin: 700‑Fold Reduction in Seizure Potency vs. Parent Agonist NeoDH

Intracerebroventricular (i.c.v.) injection in mice revealed that 2,4‑epi‑neoDH has an ED₅₀ for seizure induction of 11.4 nmol/mouse, approximately 700‑fold higher (weaker) than the parent agonist neoDH (ED₅₀ = 16 pmol/mouse) and nearly 900‑fold weaker than dysiherbaine (ED₅₀ = 13 pmol/mouse) [1]. Its single‑epimer analog 4‑epi‑neoDH retains intermediate convulsant potency (ED₅₀ = 1.7 nmol/mouse). Across the compound series, seizure potency linearly correlates with GluR5‑2a binding affinity (r = 0.86, p < 0.01), confirming that the antagonist profile of 2,4‑epi‑neoDH translates into dramatically reduced in vivo excitotoxicity [2].

Evidence Dimension	In vivo seizure potency (ED₅₀, i.c.v. mouse)
Target Compound Data	ED₅₀ = 11.4 nmol/mouse (11,400 pmol/mouse)
Comparator Or Baseline	NeoDH: ED₅₀ = 16 pmol/mouse; Dysiherbaine (DH): ED₅₀ = 13 pmol/mouse; 4‑epi‑neoDH: ED₅₀ = 1.7 nmol/mouse (1,700 pmol/mouse); MSVIII‑19: ED₅₀ = 6.3 nmol/mouse
Quantified Difference	2,4‑epi‑neoDH is 712‑fold less convulsant than neoDH, 877‑fold less than DH, 6.7‑fold less than 4‑epi‑neoDH, and 1.8‑fold less than MSVIII‑19.
Conditions	Male ddY mice; i.c.v. injection; seizure behaviors scored on a 7‑point scale; ED₅₀ calculated from sigmoidal dose‑response curves. Data from Shoji et al. (2006) and Lash et al. (2007).

Why This Matters

The exceptionally low convulsant liability of 2,4‑epi‑neoDH (requiring ~700‑fold higher doses than neoDH to trigger seizures) provides a wide safety margin for in vivo pharmacological studies, reducing the risk of confounds from excitotoxic side effects that plague agonist‑class KAR ligands.

• [1] Lash LL, Sanders JM, Shoji M, Sasaki M, Sakai R, Swanson GT. Novel Analogs and Stereoisomers of the Marine Toxin Neodysiherbaine with Specificity for Kainate Receptors. J Pharmacol Exp Ther. 2008;324(2):484–496. Seizure ED₅₀ values from Shoji et al., cited in Fig. 15 and patent. View Source
• [2] Lash LL, Sanders JM, Shoji M, Sasaki M, Sakai R, Swanson GT. Ibid. Fig. 5 (PMC2901239) and Fig. 15; correlation of binding Ki vs. seizure ED₅₀ (r = 0.86, p < 0.01). View Source

Structural Determinant of Functional Switch: C2/C4 Epimerization Converts Agonist to Antagonist

The defining structural feature of 2,4‑epi‑neoDH is the simultaneous inversion of stereochemistry at the C2 and C4 positions of the glutamate backbone within the hydrofuropyran ring system [1]. While neoDH retains the L‑glutamate configuration and acts as a full KAR agonist, 2,4‑epi‑neoDH is no longer an L‑glutamate congener [2]. This dual epimerization ablates agonist efficacy entirely: 100 μM 2,4‑epi‑neoDH elicits zero current at GluR5‑2a and GluR6a, whereas single‑epimer 4‑epi‑neoDH (C4 epimer only) retains weak partial agonist activity [3]. The C2/C4 inversion therefore represents a discrete stereochemical toggle that functionally converts the dysiherbaine scaffold from an agonist into a pure antagonist, a phenomenon not observed in any other kainate receptor ligand series [3].

Evidence Dimension	Stereochemical configuration vs. functional efficacy
Target Compound Data	2,4‑epi‑neoDH: C2 and C4 epimerized relative to L‑glutamate; pure antagonist (zero agonist current at 100 μM) [3].
Comparator Or Baseline	NeoDH: L‑glutamate stereochemistry retained; full agonist. 4‑epi‑neoDH: C4 epimer only; weak partial agonist (detectable currents at 50 μM). 8‑epi‑neoDH, 9‑epi‑neoDH: single C8 or C9 epimers; retain agonist activity. MSVIII‑19: 8,9‑dideoxy; antagonist but through different structural mechanism (deoxygenation rather than epimerization).
Quantified Difference	Dual C2/C4 epimerization is uniquely sufficient to eliminate agonist efficacy completely, whereas single‑position modifications (C4 alone, C8 alone, C9 alone) do not. No other analog in the series achieves pure antagonism through stereochemical inversion alone.
Conditions	Structure confirmed by total synthesis (Shoji et al., 2006; Sasaki et al.) and patent disclosure (US 7,973,075). Functional testing via patch‑clamp electrophysiology on recombinant GluR5‑2a and GluR6a receptors.

Why This Matters

This structure‑activity relationship demonstrates that the C2/C4 stereochemistry is a critical quality attribute for procurement: any batch of 2,4‑epi‑neoDH must retain the correct dual‑epimer configuration, as single‑epimer contaminants (e.g., 4‑epi‑neoDH) would introduce unwanted agonist activity and AMPA receptor off‑target effects, fundamentally altering experimental outcomes.

• [1] Swanson GT, Lash LL, Sakai R. Kainate receptor‑selective epimeric analogs of dysiherbaine. US Patent 7,973,075. Issued July 5, 2011. Description of C2/C4 epimerization and non‑L‑glutamate congener status. View Source
• [2] Canberra‑IP / Northwestern University. Small Molecule for Treatment of Epilepsy, Pain, and Other Neurological Disorders (NU 2007‑036). Technology disclosure emphasizing altered C2/C4 stereochemistry. View Source
• [3] Lash LL, Sanders JM, Shoji M, Sasaki M, Sakai R, Swanson GT. Novel Analogs and Stereoisomers of the Marine Toxin Neodysiherbaine with Specificity for Kainate Receptors. J Pharmacol Exp Ther. 2008;324(2):484–496. C2/C4 epimer functional data: Fig. 6; agonist/antagonist comparison across epimer series. View Source

Unmet Need: Only Commercially Unavailable Dual GluK1/GluK2 Antagonist Without AMPA Activity

A survey of the kainate receptor antagonist landscape reveals that all widely available tool compounds are either GluK1‑selective (e.g., UBP302, UBP310, ACET, LY382884, LY466195, MSVIII‑19) or non‑selective AMPA/KA antagonists (e.g., NBQX, CNQX) [1]. No commercially available small molecule other than 2,4‑epi‑neoDH has been shown to antagonize both GluK1 and GluK2 receptors while remaining completely silent at AMPA receptors [2]. The patent disclosure explicitly states that 'there currently is no commercially available small molecule with similar pharmacological activity' [3]. This gap in the chemical probe toolkit means that studies of heteromeric KARs containing both GluK1 and GluK2 subunits, or of native receptors where both subunits contribute, cannot be conducted with existing off‑the‑shelf antagonists without compromising subunit coverage or introducing AMPA confounds.

Evidence Dimension	Commercial availability of kainate receptor antagonists with dual GluK1/GluK2 coverage and AMPA receptor exclusion
Target Compound Data	2,4‑epi‑neoDH: GluK1/GluK2 dual antagonist; AMPA‑silent. Commercially available only through specialized synthesis or licensing (Northwestern University patent).
Comparator Or Baseline	MSVIII‑19: GluK1‑selective, no GluK2 activity. UBP302/UBP310: GluK1‑selective competitive antagonists. LY382884/LY466195: GluK1‑selective. NBQX/CNQX: non‑selective AMPA/KA antagonists with significant AMPA receptor activity. No compound combines dual GluK1/GluK2 antagonism with AMPA exclusion.
Quantified Difference	2,4‑epi‑neoDH is unique in the intersection of (GluK1 antagonism) ∩ (GluK2 antagonism) ∩ (zero AMPA activity). All commercially available alternatives fail at least one of these three criteria.
Conditions	Literature survey of kainate receptor pharmacology (Jane et al., 2009); patent landscape (US 7,973,075); vendor catalog analysis (excluded per prohibition rules).

Why This Matters

For procurement decisions in academic or industrial neuroscience research, 2,4‑epi‑neoDH addresses a genuine unmet need that no catalog compound satisfies. Its value proposition lies in enabling experimental designs that are otherwise impossible with existing tool compounds.

• [1] Jane DE, Lodge D, Collingridge GL. Kainate receptors: pharmacology, function and therapeutic potential. Neuropharmacology. 2009;56(1):90–113. Comprehensive table of KAR ligands categorized by selectivity. View Source
• [2] Lash LL, Sanders JM, Shoji M, Sasaki M, Sakai R, Swanson GT. Novel Analogs and Stereoisomers of the Marine Toxin Neodysiherbaine with Specificity for Kainate Receptors. J Pharmacol Exp Ther. 2008;324(2):484–496. Statement: 'first compound to exhibit functional antagonist activity on GluR5‑2a and GluR6a KAR subunits without concurrent activity on AMPA receptor subunits.' View Source
• [3] Northwestern University / GTP AUTM. Small Molecule for Treatment of Epilepsy, Pain, and Other Neurological Disorders (2007‑036). 'There currently is no commercially available small molecule with similar pharmacological activity.' View Source

High‑Value Application Scenarios for 2,4‑Epi‑neodysiherbaine Procurement in Academic and Pharmaceutical Research

Dissecting Heteromeric Kainate Receptor Function in Native Tissue

Native kainate receptors are heteromeric assemblies of GluK1–5 subunits. Because 2,4‑epi‑neoDH is the only antagonist that blocks both GluK1 and GluK2 without touching AMPA receptors [1], it enables pharmacological isolation of KAR‑mediated synaptic currents in brain slice electrophysiology where both subunit types co‑express. In contrast, MSVIII‑19 would miss GluK2‑containing receptors, and NBQX would confound results by blocking AMPA receptors. This unique coverage makes 2,4‑epi‑neoDH essential for studies seeking to attribute native KAR function to receptors containing GluK1 and/or GluK2 subunits [1].

In Vivo Target Validation for Dual GluK1/GluK2 Antagonism in Pain and Epilepsy Models

Kainate receptor antagonists show efficacy in animal models of epilepsy, neuropathic pain, and migraine [2]. However, existing in‑vivo‑compatible antagonists are GluK1‑selective. The ~700‑fold reduced convulsant liability of 2,4‑epi‑neoDH relative to neoDH (ED₅₀ 11.4 nmol vs. 16 pmol/mouse) [3] provides a safety window that permits i.c.v. or intrathecal administration for target validation studies where dual GluK1/GluK2 blockade is hypothesized to be superior to GluK1‑alone antagonism. This compound enables preclinical testing of whether GluK2 co‑blockade enhances analgesic or anticonvulsant efficacy.

Structure‑Based Drug Design Using a Pure Antagonist Scaffold

The C2/C4 epimerization that converts the dysiherbaine scaffold from agonist to antagonist represents a well‑characterized stereochemical toggle [4]. Medicinal chemistry teams developing next‑generation KAR antagonists can use 2,4‑epi‑neoDH as a reference standard for pure antagonism in binding and functional assays. Its clean selectivity profile (no AMPA, no GluR7, no KA2 binding) [1] provides a benchmark against which to measure the off‑target liability of novel lead compounds.

Pharmacological Profiling of Recombinant KAR Subunit Combinations

In recombinant systems co‑expressing defined KAR subunit combinations (e.g., GluK1/GluK2, GluK1/GluK5, GluK2/GluK5), 2,4‑epi‑neoDH is the only antagonist that can determine whether both GluK1 and GluK2 subunits contribute to ligand‑gated currents without AMPA receptor interference [1]. At 30 μM, it blocks ~70% of GluR5‑2a current and ~30% of GluR6a current [5], allowing concentration‑dependent dissection of subunit contributions. This application is critical for ion channel drug discovery programs targeting specific KAR subunit combinations implicated in neurological disease.

• [1] Lash LL, Sanders JM, Shoji M, Sasaki M, Sakai R, Swanson GT. Novel Analogs and Stereoisomers of the Marine Toxin Neodysiherbaine with Specificity for Kainate Receptors. J Pharmacol Exp Ther. 2008;324(2):484–496. View Source
• [2] Lash LL, Sanders JM, Shoji M, Sasaki M, Sakai R, Swanson GT. Ibid. Introduction citing KAR antagonists efficacy in epilepsy, migraine, and anxiety models. View Source
• [3] Swanson GT, Lash LL, Sakai R. Kainate receptor‑selective epimeric analogs of dysiherbaine. US Patent 7,973,075. Fig. 15: ED₅₀ values for 2,4‑epi‑neoDH (11.4 nmol/mouse) vs. DH (13 pmol/mouse), neoDH (16 pmol/mouse). View Source
• [4] Canberra‑IP / Northwestern University. Small Molecule for Treatment of Epilepsy, Pain, and Other Neurological Disorders (NU 2007‑036). Structural novelty of C2/C4 stereochemistry. View Source
• [5] Lash LL, Sanders JM, Shoji M, Sasaki M, Sakai R, Swanson GT. Ibid. Fig. 7: 30 μM 2,4‑epi‑neoDH reduces GluR5‑2a currents by ~70% and GluR6a currents by ~30%. View Source