Background and Purpose: Episodic ataxia type 1 (EA1) is an autosomal dominant neurological disorder caused primarily by loss-of-function mutations in the voltage-gated potassium channel Kv1.1 (KCNA1). Small molecules that restore Kv1.1 activity hold promise as targeted therapies for EA1, yet current pharmacological strategies remain limited.

Experimental Approach: Two electrode voltage-clamp electrophysiology and the Xenopus laevis oocyte expression system were used to study effects of carnosic acid, a phenolic diterpene from Salvia rosmarinus (rosemary) leaf extract on Kv1.1-linked EA1 mutant channels. We assessed ataxia therapeutic efficacy by comparing performance of Kcna1^+/+ and Kcna1^E283K/+ mice, a model of human EA1, on a balance beam under isoproterenol challenge in the absence or presence of 0.3 mg/kg carnosic acid.

Key Results: Rosemary leaf extract and rosemary metabolite carnosic acid are efficacious Kv1.1 openers. Carnosic acid fully or partially corrects heterozygous, heteromeric Kv1.1 EA1 mutant channel activity in vitro and restores normal function in Kv1.1^E283K/+ mice at 0.3 mg/kg. Experimental validation of unbiased in silico docking reveals carnosic acid occupies a binding pocket between the S1 and S4 helices of the voltage-sensing domain (VSD) of Kv1.1. Finally, we determined the chemical properties that endow carnosic acid with the ability to open Kv1.1 channels.

Conclusions and Implications: Our findings uncover a Kv1.1 opener with the potential to reverse EA1 and other Kv1.1-linked disorders.

Collection and preparation of plant samples

Salvia rosmarinus leaves were collected from the corresponding author’s garden. We homogenized the sample using a bead mill with porcelain beads in a 50 ml tube (Omni International, Kennesaw, GA, United States). We resuspended the homogenate in 80% methanol/20% water (100 ml per 5 g solid) and then incubated for 48 hours at room temperature, occasionally inverting the bottles to resuspend the extracts. We next removed the methanol using evaporation in a fume hood for 24 hours at room temperature. We centrifuged extracts for 10 minutes at 21 °C, 4000 RCF to remove the remaining particulate matter, followed by storage at -20 °C. On the day of electrophysiological recording, we thawed the extracts and diluted them 1:50 in bath solution immediately before use.

Channel subunit cRNA preparation and Xenopus laevis oocyte injection.

All potassium channel constructs used as part of this study were human. cRNA transcripts encoding Kv1.1-1.5 channel subunits were generated using the T7 mMESSAGE mMACHINE kit (Ambion, Inc., UK) from cDNA sub-cloned into pMAX plasmids. cRNA transcripts for Kv7.1-7.5, hERG, Kv2.1, Kv4.2, Kv4.3, and Trek-1 were generated using the SP6 mMESSAGE mMACHINE kit (Ambion, Inc., UK) from cDNA sub-cloned into pTLNx plasmids. The human Kv1.1-linked EA1 mutant constructs were generated by Genscript (Piscataway, NJ, USA). Defolliculated stage V and VI Xenopus laevis oocytes (Xenoocyte, Dexter, MI, USA) were injected with 0.5-10 ng of cRNA and incubated at 16 ^oC in ND96 containing penicillin and streptomycin. Oocytes were washed daily, for 1-3 days prior to two-electrode voltage-clamp (TEVC) recording.

Two-electrode voltage clamp (TEVC)

TEVC was performed as described previously (Manville & Abbott, 2025b). Briefly, experiments were performed at room temperature using an OC-725C amplifier (Warner Instruments, Hamden, CT, USA) and pClamp 8 software (Molecular Devices, Sunnyvale, CA, USA) 1-3 days after cRNA injection. Chemicals were either sourced from Combi-Blocks (San Diego, CA, USA) or synthesized in-house. Stock solutions of chemicals were made in DMSO and added to the bath solution (in mM): 96 NaCl, 4 KCl, 1 MgCl₂, 1 CaCl_2, 10 HEPES (pH 7.6). Compounds and rosemary leaf extract were introduced into the oocyte recording bath by gravity perfusion at a constant flow of 1 ml per minute for 3 minutes prior to recording. Pipettes were of 1-2 MΩ resistance when filled with 3 M KCl. Currents were recorded in response to voltage pulses between -80 mV and +40 mV at 10 mV intervals, followed by a tail pulse to -40 mV, from a holding potential of -80 mV, to yield current-voltage relationships. We analyzed data using Clampfit (Molecular Devices) and GraphPad Prism software (GraphPad, San Diego, CA, USA), stating values as mean ± SEM. We plotted raw or normalized tail currents versus prepulse voltage and fitted with a single Boltzmann function.

Group sizes were selected based on prior studies using comparable electrophysiological and behavioral paradigms in Xenopus oocytes and mouse models of episodic ataxia, where these sample sizes consistently detected biologically meaningful effects with acceptable variance and reproducibility.

Cell Culture

Human embryonic kidney 293T (HEK293T) cells were obtained from American Type Culture Collection (ATCC) and maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum and 1% penicillin–streptomycin. Cells were transiently transfected with the human Kv1.1 plasmid using Lipofectamine 2000 (Thermo Fisher Scientific) according to the manufacturer’s instructions. Whole-cell patch clamp experiments were performed 24 h post-transfection.

Whole-cell patch clamp

Kv1.1-mediated K⁺ currents were recorded in the whole-cell configuration using an Axopatch 200B amplifier (Molecular Devices, Sunnyvale, CA, USA). Data acquisition and stimulation were controlled using a Digidata 1322A interface and pCLAMP 10 software (Molecular Devices, Sunnyvale, CA, USA). Cells were continuously perfused with an external solution containing 4 mM KCl, 120 mM NaCl, 5 mM glucose, 10 mM HEPES, 0.5 mM CaCl₂, and 1.2 mM MgCl₂ (pH 7.4). Patch pipettes (2-3 MΩ resistance) were filled with an internal solution composed of 136 mM KCl, 2 mM K₂ATP, 5 mM EGTA, 10 mM HEPES, and 1 mM MgCl₂ (pH 7.2). Capacitive transients were compensated online. Currents were sampled at 10 kHz and low pass filtered at 1 kHz. Carnosic acid was applied via a micro perfusion system after current stability was confirmed using test pulses to -50 mV (1 s duration) from a holding potential of -80 mV, delivered at 10 s intervals. Current-voltage (I-V) relationships were elicited by depolarizing steps from a holding potential of -80 mV to test potentials ranging from -80 mV to +40 mV (1.5 s duration, 10 mV increments), followed by a step to -40 mV for 1 s. Interpulse intervals were 10 s, consistent with those used in TEVC recordings.

Chemical sourcing and synthesis

Carnosic acid (Cat#. QC-4383; purity 97%) and carnosol (Cat#. HA-3496; purity 97%) were obtained from Combi-Blocks (San Diego, CA, USA). Rosmarinic acid (Cat#. R4033; purity 98%) was obtained from Sigma-Aldrich (St. Louis, MO). Methyl carnosate (purity >95%), carnosic acid γ-lactone (purity >95%), carnosic acid diacetate (purity >95%) and carnosic acid dimethyl ether (purity >95%) were synthesized as described previously (Manville et al., 2023a). The ketal of carnosic acid (purity >95%) and acetone (purity >95%) was prepared from carnosic acid as in the literature with minor modifications (Li et al., 2025; Huang et al., 2025). Purity of synthesized compounds was assessed following chromatography or crystallization using thin-layer chromatography.

In silico docking

For in silico ligand docking predictions of compounds binding to Kv1.1 we performed unguided docking to predict potential binding sites, using DockThor (Guedes et al., 2024). We used AlphaFold2 to generate a predicted structure for Kv1.1, as described previously (Manville et al., 2025a). We prepared the channel structure for docking using DockPrep in UCSF Chimera https://www.rbvi.ucsf.edu/chimera), with which we also generated docking figures.

Mouse Studies

Experimental animals

The KCNA1^E283K/+ transgenic mouse line was created from C57BL/6 mice by CRISPR knock-in at the Chao Family Comprehensive Cancer Center Transgenic Mouse Facility at UCI as described previously (Manville et al., 2025a). Mice were group-housed on a 12-hour light/dark cycle with ad libitum access to food and water. Adult, male mice (8-12 weeks of age) were selected for this initial proof-of-concept therapeutic study. The behavioral experiments were performed under protocol AUP-24-095, approved by the Institutional Animal Care and Use Committee at the University of California, Irvine. Behavioral cohort sizes were consistent with previous studies using the Kcna1^E283K/+ model.

KCNA1^E283K/+ mouse balance beam experiments

Carnosic acid was made up at 0.03 mg/mL in sterile PBS containing 1% DMSO. Isoproterenol hydrochloride (Sigma-Aldrich, St. Louis, MO, USA) was made up at 0.75 mg/mL free base in sterile PBS. All injections were performed intraperitoneally at 10 mL/kg. The balance beam apparatus consisted of a hardwood dowel (1 cm diameter, 100 cm length) that was suspended 30 cm above the ground. On day one (training), each mouse was trained to cross the balance beam once in two separate sessions. On day two (experiment), each mouse was initially recorded crossing the balance beam for a baseline evaluation. Thirty minutes after the baseline trial, each mouse received an intraperitoneal injection of either vehicle or carnosic acid. Next, 25 minutes after the first injection, animals received an intraperitoneal injection of isoproterenol; 25 minutes after the isoproterenol treatment, each mouse performed two recorded trials on the balance beam separated by 10 minutes. The dosing paradigm and timing of carnosic acid administration were selected to permit systemic absorption and CNS exposure prior to isoproterenol challenge, based on prior in vivo studies of carnosic acid and related diterpenes (Manville et al., 2025a; Zhang et al., 2015; Farr et al., 2016; Vaka et al., 2011; Vaquero et al., 2013; Satoh et al., 2007). The balance beam was wiped down with 70% ethanol and allowed to dry between each trial and mouse. The balance beam videos were scored by a blinded observer who counted the number of hindfoot missteps. Statistics were determined by two-way ANOVA.

Data and statistical analysis

Experiments were designed to generate groups of approximately equal size. Animals were randomly assigned to treatment groups prior to experimentation. Behavioral scoring was conducted by investigators blinded to genotype and treatment condition. Where indicated, electrophysiological data were normalized to baseline or matched control values to reduce inter-oocyte variability between batches while preserving relative drug effects. No transformations were applied to artificially generate Gaussian distributions. Normalized electrophysiological responses are presented as fold change relative to matched control values. Accordingly, normalized datasets and associated figure axes are reported as “fold matched control values” or “fold mean of matched controls,” reflecting normalization relative to corresponding vehicle-treated or baseline control recordings. All values are expressed as mean ± SEM. Statistical analysis was conducted only for datasets in which each experimental group contained at least n = 5 independent biological replicates. Multiple comparison statistics were conducted using a one-way ANOVA with a post-hoc Dunnett’s test or by two-way ANOVA. For multigroup comparisons using parametric statistics, post hoc analyses were performed only following significant omnibus ANOVA results. Data distributions and variance homogeneity were assessed prior to application of parametric tests. Comparison of two groups was conducted using a t-test; all p values were two-sided. Statistical significance was defined as P < 0.05. Electrophysiological data were confirmed in at least two batches of oocytes. The declared n values represent independent biological replicates (individual oocytes or individual animals), and technical replicates were not treated as independent observations for statistical analysis. No predetermined exclusion criteria were applied. Where indicated, lower-throughput mechanistic experiments are presented to illustrate qualitative consistency across experimental systems and were interpreted conservatively. In limited cases where experimental throughput constrained sample size, data were included to illustrate mechanistic observations or consistency of electrophysiological effects across experimental systems but were not subjected to formal statistical analysis. This manuscript complies with BJP’s recommendations and requirements on experimental design and analysis (Curtis et al., 2018).

Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in https://www.guidetopharmacology.org, and are permanently archived in the Concise Guide to PHARMACOLOGY 2023/24 (Alexander et al., 2023).