Materials and methods
D. flexuosum specimens in the flowering phase were harvested from Tsikhijvari, Georgia. A voucher specimen labeled TBPH-21405 is preserved at the Herbarium of the TSMU I. Kutateladze Institute of Pharmacochemistry.

Obtaining of alkaloids

The aerial and underground parts of D. flexuosum were air-dried and powder-grinded, treated with a 5% sodium carbonate solution to remove non-alkaloid impurities, and extracted with chloroform. The chloroform extracts were concentrated to one-fifth of their original volume, and alkaloids were extracted using a 5% aqueous solution of sulfuric acid. After cooling, the acidic extract was rinsed with diethyl ether, alkalinized with sodium carbonate to a pH of 9, and the alkaloids were then extracted using chloroform. After dehydration with anhydrous sodium sulfate and vacuum concentration, crude alkaloids were fractionated according to basicity using a modified polybuffer separation method.8 Fractionation with citrate-phosphate buffers (рН 8.0 → 2,0) yielding eight fractions (4 each for aerial and underground parts of D. flexuosum).

​

Qualitative analysis of alkaloids in comparison with reference samples was conducted using TLC Silica gel 60 F₂₅₄ plates (Merck, Germany) in the following systems: I – chloroform: methanol (6:1), II – chloroform: benzene: 95% ethanol: 25% ammonia (40:40:10:0.2). Detection of alkaloids was done using Dragendorff's reagent.9,10

​

United fractions with a pH of 2-5 and a pH of 6-8 were separated on a silica gel column (100/400), eluted with chloroform, and six mixtures of chloroform-methanol (100:1; 90:1; 50:1; 25:1; 5:1, 1:1). Alkaloids were identified using the 7890B GC System and the 5977B Single Quadrupole GC/MSD System (Agilent Technologies, USA). Alkaloids' melting point was determined by the IA 9100 melting point apparatus (Wenk Lab Tec, Germany).

Experimental animals
Inbred male albino mice (n=72, body weight: 26±2 g) were utilized and randomly divided into six groups. Animals were housed under standard conditions: temperature of 20±2°C, humidity at 40%, with a 12/12 hr. light/dark cycle, and provided with a standard rodent diet and water ad libitum. All animal maintenance and experimental procedures adhered to internationally accepted ethical guidelines.10,11 Statistical analysis of experimental data was conducted using Student's t-test. All animal experiments were approved by the Tbilisi State Medical University Ethical Committee on Animal Experiments (approval #AP-61-2023).

PTZ-induced seizures assay

D.flexuosum fractions were administered intraperitoneally at doses of 10 mg/kg. Thirty minutes after administration, each animal was injected intraperitoneally with 80 mg/kg pentylenetetrazol (PTZ, Sigma, USA). After PTZ injection, mice were placed in a plexiglass acrylic cage, and seizures were recorded until convulsions stopped and the animals were fully recovered. Occurrence, severity, latency, and duration of seizures were analyzed, as well as the number of animals displaying seizures and death due to seizures.12,13

​

The severity of seizures was estimated by the following scale: 0 - no changes in behavior; 1 - isolated myoclonic jerks; 2 - atypical minimal seizures (unilateral, incomplete); 3 - total clonic seizure; 4 - a pattern of tonic-clonic seizures with a suppression of tonic phase; 5 - generalized tonic-clonic seizure and status epilepticus; Mortality percentage was recorded over 60 minutes, with animals surviving beyond one hour deemed protected

“Hot plate” assay
The analgesic activity of alkaloids was evaluated using Eddy's "hot plate" test.14-15 In brief, animals were placed into a 12 cm wide and 25 cm high transparent polycarbonate cylinder on a hot plate maintained at 55±0.2°C. Experimental mice were injected with Delphinium extract 5 and 15 mg/kg intraperitoneally 15 min before the experiment. Intraperitoneally normal saline 0.2 ml/animal was used as the control. The latency period (in seconds) between placing on the plate and the occurrence of response reactions (withdrawal of hind paws, jumping) was recorded. Animals without a response within 30 seconds were excluded from the experiment. Latencies were recorded before (base level) and 15-, 30-, 45-, 60-, and 120-min post-injection of test and reference compounds.