Ponatinib (Pon; lot no. #13771), aspirin (lot no. #323026) and edaravone (Eda; lot no. #35508) were purchased from MedChemExpress. O-dianisidine (lot no. #SHBP1768) was obtained from Sigma-Aldrich (Merck KGaA). Water extracts of the aerial parts of CA (lot no. #24ENLC02), MO leaves (lot no. #24DDLC01), CL rhizomes (lot no. #24DQLC03), PN seeds (lot no. #24DTLC01) and CS fillets (lot no. #24BWLC01) were purchased from PT. Sari Alam Sukabumi, which complies with good manufacturing processes for producing extracts. Stock solutions of the drugs were dissolved in either 100% dimethyl sulfoxide (DMSO) or 0.9% sodium chloride. Stock solutions of extracts were prepared in either deionized water or 1% DMSO. All stock solutions were diluted with E3 medium (Sigma-Aldrich; Merck KGaA) before each experiment.

High-resolution mass spectrometry (MS) analysis was performed using a Waters ACQUITY UPLC^® H-Class system combined with an Xevo G2-S QT of MS (Waters Corporation). The analysis used a C18 column (1.8 µm, 2.1x100 mm; ACQUITY UPLC^® HSS, Waters Corporation) maintained at 50˚C, while the room temperature was maintained at 25˚C. Liquid chromatography was performed using a step gradient with a mobile phase consisting of water containing 5 mM ammonium formate (A) and acetonitrile containing 0.05% formic acid (B), at a flow rate of 0.2 ml/min over a 23-min run time. The injection volume was 5 µl and samples were prefiltered using a 0.2-µm syringe filter. Electrospray ionization in the positive mode was used for MS, scanning a mass range of 50-1,200 m/z. The source temperature was 100˚C and the desolvation temperature was maintained at 325˚C. The cone gas flow rate was 0 l/h, while the desolvation gas flow rate was 794 l/h. A collision energy gradient of 4-60 eV was applied. Data acquisition, analysis and instrument control were performed using MassLynx software v4.1(20).

Adult wild-type zebrafish (54 male and 36 female; age, 4-5 months, 0.4-0.6 g) were purchased from a local breeder from Bogor, Indonesia. The zebrafish species (Danio rerio) was confirmed at the Museum Zoology, School of Life Sciences and Technology, Institut Teknologi Bandung, Bandung, Indonesia (specimen no. 3939/IT1.C11.2/TU/2024). Before the experiment, adult zebrafish were acclimatized for ≥14 days. The zebrafish were maintained under standard laboratory conditions, including a temperature of 26-28˚C, 14/10-h light/dark cycle, continuous water filtration and aeration, pH maintained at 6.9±0.2 and conductivity within a standard range (21). They were fed three times daily with commercial pellets. Any dead animals were removed immediately to maintain water quality. The fish were handled carefully to minimize stress. Animal health and behavior were monitored at least twice daily. Observations included swimming behavior (hypoactivity and erratic movement), feeding response, morphological abnormality and signs of distress (surface gasping and loss of righting reflex). In total, nine adult zebrafish that exhibited severe distress or abnormal behavior were promptly separated and closely monitored throughout the experimental period.

Zebrafish eggs were obtained from breeding adult fish in a 6:4 male: female ratio. The fertilized eggs were collected and transferred to E3 medium. For thrombosis analysis, 10 h after embryo collection, 3% propylthiouracil (PTU; v/v) was added to inhibit melanin formation in the zebrafish embryos. The experimental procedures were conducted over 4 days for the toxicity study and 5 days for the anti-ischemic stroke study, including breeding, treatment exposure and endpoint assessment. At the end of the experiments, euthanasia of zebrafish larvae was performed using 0.4% tricaine solution. Death was confirmed based on established criteria for zebrafish, including absence of opercular (gill) movement, lack of heartbeat (observed under stereomicroscope) and no response to gentle tactile stimulation (22). To ensure accuracy, observations were conducted for ≥5 min before confirming mortality. All procedures were approved by the Animal Research Ethics Committee at the Institut Teknologi Bandung (Bandung, Indonesia; approval nos. KEP/I/2024/II/H211223ND/TAAZ for the toxicity test and KEP/I/2024/VI/H110624NM/ANSZ for the anti-ischemic stroke test).

The fish embryo acute toxicity test, based on the Organization for Economic Co-operation and Development protocol no. 236, was conducted to evaluate toxicity (23). E3 medium and 4 µg/ml 3,4-dichloroaniline solution were used as negative and positive controls, respectively. Toxicity testing was performed on zebrafish embryos <6 h post-fertilization (hpf). The embryos were placed in a 24-well plate, with one embryo/well. Subsequently, the embryos were incubated with CA (156.28-5,000.00 µg/ml), MO (31.25-1,000.00 µg/ml), CL (156.28-5,000.00 µg/ml), PN (31.25-1,000.00 µg/ml) and CS (31.25-1,000.00 µg/ml) for 24-96 h at room temperature. Observations of embryonic abnormality, including embryo coagulation, imperfect somite formation and tail bud release and absence of heartbeat, were carried out every 24 h (24). The 10% (LC₁₀) and 50% lethal concentration (LC₅₀) were calculated by probit analysis using a Microsoft Excel 2021-based template (25). The experiments were performed in triplicate.

Fig. 1 presents a schematic diagram of the experimental design. The method for producing ischemic stroke in zebrafish larvae was adapted from a previous study (19). Zebrafish larvae at 4 days post fertilization (dpf) were placed into 24-well plates, with each well containing 5 larvae. Larvae were incubated for 24 h (room temperature) in a mixture of 3 µg/ml Pon solution and sample solutions (¼ LC₁₀, ½ LC₁₀ and LC₁₀ of each extract) or reference drug solutions (12.5 µg/ml Asp and Eda). Larvae incubated in E3 or E3 + PTU were used as the control group.

Following ischemic stroke induction, the larvae were euthanized with 0.4% tricaine solution, and whole larvae were stained in room temperature with 5.85 mmol/l o-dianisidine solution, sodium acetate buffer (0.1 M, pH 4.5), deionized water and 30% (v/v) hydrogen peroxide (20:5:20:1) for 15 min in the dark. Larvae were washed three times with 100% DMSO and mounted on glass slides with low-melting point agarose. The zebrafish larvae were observed under a light microscope. The area of cerebral thrombosis was characterized by a dark brown color and was quantified using ImageJ software (version 1.54g) (26).

Following induction, zebrafish larvae were placed in a Petri dish containing 3 ml incubation medium and recorded individually. Zebrafish larvae were subjected to light/dark stimulation, comprising a 1 min light period followed by a 1 min dark period. Larval locomotion was recorded for 3 min using a camera. If the larvae did not exhibit movement at the beginning of the recording period, tactile stimulation by touching the head or tail was performed to trigger movement (27). The video recordings were analyzed with AnimalTA software v3.2.2(28). Locomotor parameters used to assess the anti-ischemic stroke effect included total distance traveled, average swimming speed and swimming path (29).

Data were analyzed with GraphPad Prism 8.0 (Dotmatics) and are presented as the mean ± SEM of ≥3 independent experimental repeats. Data were analyzed by one-way ANOVA or Kruskal-Wallis test followed by Dunnett's test. P<0.05 was considered to indicate a statistically significant difference.

Fig. 2 shows the chromatogram and predicted metabolites of CA, MO, CL, PN and CS. CA comprised various metabolites classified into alkaloid, phenolic and terpenoid groups. MO contained chemical components belonging to the phenolic, flavonoid and glycoside classes. Additionally, phenolic and terpenoid groups were detected in CL. PN exhibited compounds classified into alkaloid and phenolic classes. Amino acids, fatty acids and peptides were detected in CS.

To evaluate the toxicity of the extracts on zebrafish embryonic development, survival rate was measured 24-96 hpf following exposure. No surviving embryos were found at 72 hpf following exposure to 5,000 µg/ml CA or CL, 2,500 µg/ml CL or 1,000 µg/ml CS (Fig. 3). PN at 1,000 µg/ml exhibited a strong lethal effect, with all embryos dying by 24 hpf. Additionally, more than half of the embryos remained viable at 96 hpf following exposure to ≤1,250 µg/ml CA, ≤500 µg/ml MO, ≤1,250 µg/ml CL and ≤250 µg/ml PN and CS. Compared with the control, embryos exposed to 156.25-625 µg/ml CA, 31.25-250 µg/ml MO, 156.25-625 µg/ml CL and 31.25-125 µg/ml PN and CS did not show a significant difference in survival rate. These results indicate that lower extract concentrations were associated with higher embryo survival.

Hatching rate was determined by quantifying the number of embryos that emerged from the chorion between 24 and 96 hpf. Embryos showed no evidence of hatching at 24 h, with a significant increase observed at 48 h, particularly following exposure to lower concentrations of extracts (Fig. 4). Control embryos (incubated in E3) hatched normally at 48 hpf. Exposure to 5,000 µg/ml CA, 1,000 µg/ml MO, 1,250-5,000 µg/ml CL and 1,000 µg/ml PN and 1,000 µg/ml CS resulted in delayed hatching of all embryos. Conversely, more than half of the embryos completed hatching by 96 hpf when exposed to ≤625 µg/ml CA, ≤500 µg/ml MO, ≤312.5 µg/ml CL, ≤125 µg/ml PN and ≤250 µg/ml CS. Lower extract concentrations, such as 156.25-625 µg/ml CA, 31.25-250 µg/ml MO, 156.25 µg/ml CL and 31.25-125 µg/ml PN and CS, did not induce any effect on the hatching rate compared with the control group. These findings suggest that embryos exposed to lower extract concentrations exhibited greater hatching success.

To assess the toxic effects of extracts, morphological alterations in embryonic development were observed every 24 h (Figs. S1, S2, S3, S4 and S5). After 24 h extract exposure, high concentrations of the extracts (≥2,500 µg/ml for CA and CL, and 1,000 µg/ml for MO, PN and CS) caused delayed gastrulation, absence of somites and coagulation. At lower concentrations, exposure to certain extracts caused pericardial edema (72 and 96 hpf following exposure to 625 µg/ml CA and CL, respectively), yolk sac edema (96 hpf following exposure to 1,250 µg/ml CA or 500 µg/ml MO or CS) and spine malformation (48 and 72 hpf following exposure to 312.5 µg/ml CA and 125 µg/ml PN, respectively). Additionally, delayed embryonic development was observed, including the lack of eye bud development (24 hpf after exposure to 1,250 µg/ml CA, 500 µg/ml MO and 125 µg/ml CS and 72 hpf after exposure to 2,500 µg/ml CA) and reduced body size (96 hpf after exposure to 2,500 µg/ml CA, 500 µg/ml MO and 625 µg/ml CL). By contrast, embryos exposed to lower extract concentrations (156.25 µg/ml CA, 31.25 µg/ml MO, 156.25-312.5 µg/ml CL and 31.25-62.5 µg/ml PN and CS) exhibited normal developmental progression.

To evaluate the mortality effects of the extracts on embryos, lethal concentrations were calculated using probit analysis (Fig. 5; Table I). The mortality effect order was as follows: CS > PN > MO > CL > CA. Based on the Globally Harmonized System of Classification and Labeling of Chemicals, all extracts were classified as non-toxic based on their LC₅₀ values (LC₅₀>100 µg/ml) (30). Therefore, these extracts were used at concentrations corresponding to ¼ LC₁₀, ½ LC₁₀, and LC₁₀ for subsequent experiments in the zebrafish ischemic stroke model (Table I).

The anti-ischemic stroke effects of the extracts were evaluated by measuring cerebral thrombosis. Following 24 h incubation with Pon, the zebrafish larvae exhibited an increased area of cerebral thrombosis compared with the control group (Fig. 6). By contrast, 114.50 and 229.00 µg/ml CA, 75.00 µg/ml MO, 275.00 µg/ml CL and 14.50-58.00 µg/ml CS demonstrated significant anti-brain thrombosis activity. The effects of CA, MO and CL were dose-dependent, with lower extract concentrations resulting in larger cerebral thrombosis areas than higher concentrations. These three extracts decreased the thrombosis area by 1.85-2.51-fold compared with the Pon group. Conversely, none of the PN concentrations significantly decreased the area of cerebral thrombosis. CS showed a comparable anti-thrombosis effect between concentrations. Moreover, CS resulted in a 1.4-3.2-fold decrease in cerebral thrombosis area relative to the Pon group, with the most pronounced decrease at 29 µg/ml CS, suggesting potent anti-thrombotic activity.

Locomotor function in zebrafish larvae following extract treatment was assessed based on the average swimming speed, total distance traveled and swimming trajectory. After Pon induction, notable locomotor impairment was observed compared with the control group in larvae treated with 114.5 µg/ml CA, 75 µg/ml MO, 26.25 and 105 µg/ml PN and 14.5 and 29 µg/ml CS (Figs. 7, 8 and 9). CL had no effect on the restoration of locomotor damage in an ischemic environment. CA and CS at their highest concentrations demonstrated a reduced effect compared with lower concentrations. Conversely, a moderate PN concentration (52.5 µg/ml) showed a weaker effect than both low (26.25 µg/ml) and high (105 µg/ml) concentrations. These findings indicated that each extract exerts its effect at a specific optimal dose.

MO was the only extract that showed a dose-dependent effect. Furthermore, among all tested concentrations, CS at 29 µg/ml demonstrated the greatest effect, increasing the average speed and total distance traveled by ~3.22 and 2.97-fold, respectively, compared with the Pon group. This indicates its high potency in restoring locomotor damage following ischemia.