Kainic Acid 2-Carboxy-3-Carboxymethyl-4-Isopropenylpyrrolidine

The compound (2S,3S,4S)-2-carboxy-4-(prop-1-en-2-yl)pyrrolidine-3-acetic acid, designated kainic acid and supplied as the monohydrate under CAS 58002-62-3 (anhydrous 487-79-6), is a rigidified glutamic acid analog isolated from the marine rhodophyte *Digenea simplex*. Structurally, the pyrrolidine ring constrains the glutamate backbone into a fixed conformation that confers selective agonist activity at the kainate subclass of ionotropic glutamate receptors (iGluRs). Commercial research-grade lots carry a minimum purity specification of ≥98.0% by HPLC at 210 nm, with the balance accounted for by the monohydrate water of crystallization (theoretical 7.8% w/w) and a trace enantiomeric impurity, typically the (2R,3R,4R)-isomer, limited to ≤0.5% area normalization. The product is provided as a white to off-white lyophilized powder in septum-sealed amber vials under argon overlay, with a recommended pre-opening equilibration to 20–25 °C to avoid condensation. Identity is confirmed by ¹H‑NMR (400 MHz, D₂O) matching the characteristic vinyl proton doublet at δ 5.25–5.30 and the C‑3 methine multiplet at δ 2.85–2.95, and by high-resolution mass spectrometry (ESI‑TOF, negative mode) yielding [M−H]⁻ at *m/z* 212.0928 ± 3 ppm. Residual inorganic content (sulfated ash) is controlled to ≤0.1% per Ph. Eur. 2.4.14.

Why Does Purity Determination Require Orthogonal Chromatographic Methods?

Kainic acid’s weak UV chromophore (λ_max ≈ 205 nm) renders conventional HPLC-UV quantification susceptible to baseline interference from dissolved atmospheric CO₂ and trace mobile phase contaminants. A validated ion-pairing reverse‑phase method on a C18 column (150 × 4.6 mm, 5 µm particle, 30 °C) with tetrabutylammonium hydrogen sulfate (5 mM, pH 2.5) and acetonitrile (3% v/v) isocratic elution achieves a retention factor *k’* of 4.2 for kainic acid, separating the main peak from the des‑isopropenyl degradation product (formed by acid-catalyzed hydration of the exocyclic double bond) and the *trans*-configured diastereomer. For lots intended for *in vivo* neurotoxicity modeling, a secondary HILIC method with charged aerosol detection (CAD) is employed to quantify non‑UV‑active process impurities, notably residual pyrrolidine‑2,3-dicarboxylic acid intermediates, which can act as weak NMDA receptor ligands and confound behavioral endpoints. Acceptable lots show CAD impurities totaling ≤1.5% area. Each batch data sheet includes both HPLC-UV (210 nm) and HILIC‑CAD chromatograms, along with the 98.0% purity acceptance limit anchored to Ph. Eur. monograph 2535 for related substances.

Excitotoxicity Induction: Injection Parameters and Regional Vulnerability Gradients

Kainic acid is used primarily to produce region‑specific neuronal loss and acute seizure phenotypes in rodents, replicating aspects of temporal lobe epilepsy. Stereotaxic microinfusion into the dorsal hippocampus (coordinates from Bregma: AP −3.6 mm, ML ±2.4 mm, DV −3.0 mm, flat‑skull) at a dose of 0.5 µg in 0.5 µL phosphate‑buffered saline (pH 7.4) over 10 minutes yields a CA3>CA1 pyramidal cell loss exceeding 70% at 7 days post‑injection, as quantified by NeuN immunohistochemistry. Infusion rates must be controlled by a microprocessor‑driven syringe pump (≤0.05 µL/min); higher rates cause mechanical shear‑induced tissue cavitation unrelated to receptor activation. The pH of the reconstituted solution requires careful adjustment—the free acid form yields a pH of 3.2–3.5 at 1 mg/mL in normal saline, which is pro‑convulsant independent of kainate receptor agonism; neutralization with 0.1 N NaOH to pH 7.2–7.4 is mandatory and must be performed immediately prior to injection because alkaline solutions promote oxidation of the isopropenyl side‑chain, with a measured half‑life of 45 minutes at 25 °C under ambient O₂. Doses and injection coordinates for amygdala‑kindling models, intraperitoneal systemic administration (10–30 mg/kg in C57BL/6J mice producing Racine stage 4–5 seizures within 30–45 minutes), and intra‑accumbens microinjection for psychosis‑relevant circuit disruption are governed by the same pH and rate constraints.

Table 1. Comparative Receptor Binding Affinity Profiles (K_i, nM, [³H] ligand displacement, rat cortex membranes)

Compound	GluK1 (iGluR5)	GluK2 (iGluR6)	GluA2 (AMPA)	GluN1/N2A (NMDA)
Kainic Acid	4.8	5.6	18.2	>10,000
Domoic Acid	1.3	1.9	32.0	>10,000
(S)-AMPA	1200	>10,000	8.7	>10,000
L-Glutamate	420	380	190	65

Handling Incompatibilities That Elevate Background Excitotoxicity

Dissolved kainic acid must be shielded from polycarbonate and polystyrene labware. The compound adsorbs to untreated polymer surfaces via ionic interaction of its two carboxylate moieties, reducing the free concentration in solution by up to 30% within 2 hours at 4 °C as measured by LC‑MS/MS. Low‑protein‑binding polypropylene tubes (siliconized) or borosilicate glass vials are required for all dilutions. Contact with phosphate buffers of concentration ≥50 mM precipitates calcium‑sensitive kainic acid‑phosphate complexes when Ca²⁺ concentration exceeds 1.8 mM, which is relevant because artificial cerebrospinal fluid routinely contains Ca²⁺ at 1.2–2.4 mM. Precipitation yields a white flocculent material that clogs microdialysis probes (CMA 12 Elite, 4 mm polyarylethersulfone membrane) and unpredictably lowers the delivered dose. When calcium‑containing vehicles are unavoidable, a chelator‑free strategy is to maintain a kainic acid concentration below 0.1 mM and verify solubility by dynamic light scattering (Z‑average ≤ 10 nm). For systemic intraperitoneal injections, the solution should be passed through a 0.22 µm PVDF syringe filter immediately before dosing to remove any microparticulates formed during pH adjustment.

The product is stable for 36 months from the date of manufacture when stored unopened at −20°C ± 5°C, with water content determined by Karl Fischer titration (USP <921> Method Ia) maintained at 3.8–8.5% (theoretical monohydrate 7.8%). After reconstitution in sterile water (resistivity 18.2 MΩ·cm) at 1 mg/mL and neutralization, the solution retains ≥95% of the initial concentration for 24 hours at 2–8 °C protected from light. Repeated freeze‑thaw cycles cause a cumulative loss of 4–7% per cycle due to surface denaturation at the ice‑liquid interface; single‑use aliquots are recommended.

When Structural Isomerism Dictates Pharmacophore Recognition at the Ligand Binding Domain

Kainic acid differs from the structurally related marine toxin domoic acid (isolated from *Pseudo‑nitzschia* diatoms) by the absence of the glutamic acid side‑chain extension found in domoic acid’s C7′–C8′ region. In domoic acid, the addition of a C7′ carboxyl‑containing substituent forms a third electrostatic contact with Arg523 in the GluK1 subunit, lowering the K_d to 1.3 nM and dramatically slowing dissociation (t_{1/2 off} > 20 minutes versus 0.8 seconds for kainic acid at GluK1). This kinetic trapping underlies domoic acid’s 30‑fold greater toxic potency in mollusks and vertebrates and its designation as an amnesic shellfish poison. Synthetically, the C4 isopropenyl group of kainic acid is a minimal hydrophobic anchor occupying a sub‑pocket lined by Tyr140 and Met191 in the GluK2 ligand binding domain; hydrogenation to the isopropyl analog reduces affinity by 120‑fold, confirming the π‑interaction of the exocyclic double bond with the aromatic cage. A racemic mixture of (2S,3S,4S)‑ and (2R,3R,4R)‑kainic acid shows half the potency of the enantiopure natural product in electrophysiological recordings from Xenopus oocytes expressing homomeric GluK2 receptors (EC₅₀ shifts from 5.1 µM to 9.8 µM), consistent with the inactive enantiomer acting as a silent competitor. Thus, specifications strictly enforce enantiomeric excess (>99.0% ee by chiral HPLC on Chiralpak® ZWIX(+) column).

Batch‑to‑Batch Variability in Murine Seizure Latency: A Process‑Control Analysis

Across 15 consecutive production batches (Lot #K21001–K21015), the median latency to first Racine stage 4 seizure in male C57BL/6J mice (8‑week‑old, 22–25 g) dosed intraperitoneally at 20 mg/kg was 28.5 minutes (interquartile range 24.7–32.1 minutes). Two outlier batches (K21003, K21011) exhibited prolonged latencies (35.1 and 36.8 minutes) that correlated with an elevated CAD impurity peak identified as kainic acid lactam, formed by intramolecular cyclization between the C2 carboxyl and C3 acetic acid side‑chain under acidic lyophilization conditions. Lactam content ≥0.8% w/w was associated with delayed seizure onset; the lactam is a weak competitive antagonist at GluK2 (K_B ≈ 2.3 µM) that reduces the net agonist drive when co‑administered. Implementation of a controlled‑pH lyophilization protocol (final cake pH 4.8–5.2 when reconstituted at 1 mg/mL) reduced lactam to ≤0.15% in subsequent batches, tightening the seizure latency distribution to ±3.5 minutes (SD). These process refinements are documented in the batch‑specific certificate of analysis, which also reports residual ethanol (headspace GC, USP <467>) ≤100 ppm and endotoxin ≤0.01 EU/µg (LAL kinetic chromogenic method, USP <85>) for lots intended for CNS infusion.

Comparative neurotoxicity with the alternative excitotoxin ibotenic acid (a non‑selective NMDA and metabotropic glutamate agonist) reveals that kainic acid‑induced lesions are confined to regions enriched in GluK2/GluK5 heteromers, sparing striatal medium spiny neurons that express high GluN2B‑containing NMDA receptors but negligible kainate receptor density. Thus, for targeted ablation of hippocampal CA3 pyramidal cells without the cholinergic fiber damage seen with ibotenate, kainic acid is the agonist of choice at infusion volumes below 0.3 µL, provided that the vehicle does not contain HEPES buffer at concentrations exceeding 10 mM, which has been shown to potentiate GluK1 currents by 15% in patch‑clamp recordings (pH 7.3, −60 mV holding potential).

Table 2. Key Specifications and Reference Methods

Parameter	Specification Limit	Method Reference
Assay (anhydrous basis)	98.0–102.0%	HPLC‑UV 210 nm, Ph. Eur. 2.2.29
Enantiomeric Purity	≥99.0% ee	Chiral HPLC, USP <621>
Water Content	3.8–8.5%	Karl Fischer, USP <921> Ia
Residual Solvents	Ethanol ≤100 ppm	HS‑GC, USP <467>
Endotoxin	≤0.01 EU/µg	LAL Kinetic, USP <85>
Lactam Impurity	≤0.3%	HILIC‑CAD
Appearance	White to off‑white powder	Visual, Ph. Eur. 2.2.1

Differential Thermal Stability and Shipping Validation Under ICH Q1A Conditions

Forced degradation studies at 60°C/75% RH for 14 days revealed a primary degradation product with a relative retention time of 1.32 (HPLC‑UV method) identified by LC‑MS as the decarboxylated C2‑COOH derivative, formed at a rate of 0.8% per day. This instability imposes a maximum exposure temperature of +40°C during shipping, verified by a validated cold‑chain packaging system using PCM (phase change material) panels qualified over 96‑hour transit duration. Temperature loggers (Type K thermocouple, ±0.5°C accuracy) are included in each shipment, and data are reviewed upon receipt. Products exposed to a confirmed temperature excursion above +40°C for >4 hours are recalled per the manufacturer’s corrective action procedure CAPA‑SHP‑004. Conversely, storage at −80°C causes no measurable loss of receptor binding affinity after 60 months, as determined by [³H]kainic acid displacement on rat cortical synaptosomes (K_d unchanged at 4.9 ± 0.3 nM). The product monograph therefore permits long‑term archiving at ≤−70°C with desiccant protection, though water re‑equilibration upon rewarming must be controlled by a 2‑hour equilibration period at 25°C before opening to prevent condensation‑induced hydrolysis.

For investigators transitioning from the commercially available kainate receptor agonist ATPA ((RS)‑2‑amino‑3‑(3‑hydroxy‑5‑tert‑butyl‑4‑isoxazolyl)propionic acid), it is critical to note that ATPA exhibits 15‑fold selectivity for GluK1 over GluK2, whereas kainic acid is essentially equipotent at both subunits. This renders kainic acid more suitable for global kainate receptor activation in slice electrophysiology where both synaptic and extrasynaptic GluK2‑containing receptors are recruited. However, ATPA’s lower glutamate site deactivation rate (τ_w 4.2 ms vs 1.1 ms for kainic acid at GluK1) yields a higher steady‑state current that may be preferred for pharmacological silencing experiments. The choice between the two agents thus hinges on the desired subunit‑recruitment profile and the temporal pattern of receptor activation, not simply on nominal EC₅₀ values.

Sterility Assurance Levels for Intracerebroventricular Administration

Aseptic filling of lyophilized kainic acid into pre‑sterilized Type‑I glass vials is performed under ISO 5 laminar airflow with a sterility assurance level (SAL) of 10⁻⁶ per Ph. Eur. 5.1.1. The terminal sterilization of the powder by gamma irradiation is explicitly avoided because irradiation at 25 kGy induces 11% radiolytic degradation of the isopropenyl moiety, producing a mixture of hydroxy‑isopropyl derivatives detectable by ESI‑MS. Instead, the bulk solution is sterilizing‑grade filtered through a PVDF membrane (0.1 µm pore) prior to lyophilization. Post‑lyophilization bioburden monitoring on 1% of vials per sub‑lot (USP <1111>) confirms CFU counts <0.1 per gram. For non‑GMP research applications, the product is shipped with the Endosafe® certificate for endotoxin and sterility certificate for aerobic/anaerobic cultures (USP <71>). Investigators undertaking chronic intraventricular infusions with osmotic minipumps (Alzet® model 1004, 0.11 µL/hr, 28‑day duration) must prepare the filling solution with preservative‑free, endotoxin‑tested saline (USP <797> pharmaceutical compounding standards) and pre‑flushed catheters to minimize glial scar encapsulation, which alters local diffusion kinetics and reduces the effective concentration at the target ventricular wall by an estimated 40% over 14 days (based on microdialysis recovery data in Sprague‑Dawley rats).

When the objective is to produce a focal lesion in the basolateral amygdala complex while preserving the adjacent central nucleus, the inherent diffusivity of kainic acid (aqueous diffusion coefficient D_37°C ≈ 6.8 × 10⁻⁶ cm²/s) imposes a maximum injection volume of 0.1 µL; larger volumes result in retrograde spread along the cannula track and undesirable piriform cortex involvement, producing respiratory arrest that confounds survival studies. This spatial constraint is unique to kainic acid and does not apply to ibotenic acid or NMDA, whose rapid uptake by excitatory amino acid transporters (EAAT1–3) limits the radius of excitotoxic spread to ≤300 µm from the infusion site.