Luteolin [(2-(3,4-dihydroxyphenyl)-5,7-dihydroxychromen-4-one), 98%, Kingherbs, China] was used in its raw and micronized forms. Acetone (99.5%, Vetec, Brazil) and CO₂ (99.9% in a liquid phase, White Martins S.A., Brazil) were used to micronize luteolin. Diazepam (Fagron Pharmaceuticals, Brazil) was used as the positive control. Pentylenetetrazole (PTZ, Sigma-Aldrich, Germany) was used to induce seizures. Commercial kits for RNA extraction (PureLink RNA Mini Kit), RNA and cDNA quantification (Qubit RNA HS Assay Kit and Qubit dsDNA HS Assay Kit), and cDNA synthesis (High-Capacity cDNA Reverse Transcription Kit) were supplied by Thermo Fisher Scientific Inc, United States. PowerUp SYBR Green Master Mix (Invitrogen, United States) was used for quantitative polymerase chain reactions (qPCR). Primers were synthesized by Invitrogen (Brazil).

Micronized luteolin was obtained by the gas antisolvent technique (Aguiar et al., 2018; Garbinato et al., 2021; dos Santos et al., 2022; de Oliveira et al., 2023). First, luteolin was briefly mixed in acetone and while being stirred into complete dissolution. The obtained solution was placed in a chamber, and CO₂ was added to increase the ideal pressure to complete the process. When the pressure was reached, the antisolvent flow was stopped, and the chamber continued to agitate the solution for 10 min at 300 rpm (Aguiar et al., 2018; Garbinato et al., 2021; de Oliveira et al., 2023). After this, the washing stage began with the antisolvent flow rate of 10 mL‧min⁻¹, and the pressure was isobaric for 60 min for the complete removal of acetone from the system. The operating conditions were pressure 80 bar and temperature 35 °C. Finally, the obtained material was stored at 4 °C.

Luteolin (raw) and micronized luteolin were characterized by scanning electron microscopy. The particle size was determined using the software Meter Size (version 1.1) (Aguiar et al., 2018; Garbinato et al., 2021; dos Santos et al., 2022; de Oliveira et al., 2023). The melting points of the luteolin and micronized luteolin were verified by a differential scanning calorimeter (Jade-DSC, Perkin Elmer, United States). Samples (5–10 mg) measurements were performed by heating the compounds from 30 °C to 350 °C, at a heating speed of 20 °C‧min^-1 in an inert atmosphere (N₂ flow: 20 mL‧min^-1) (Aguiar et al., 2018; Garbinato et al., 2021; de Oliveira et al., 2023).

The zebrafish larvae used in this study were obtained from adult zebrafish (Danio rerio, AB strain), maintained in our institutional vivarium. Zebrafish were mated, and eggs were obtained following previous protocol (Bertoncello et al., 2018; Decui et al., 2020). Fertilized eggs were collected, freed of debris, and disposed of in sterile cell culture plates containing embryo medium (osmosis water equilibrated with Instant Ocean salts; pH 7.0–7.4). Plates were placed in an incubator at 27 °C, under a 14–10-h light/dark cycle photoperiod (100 lux) until animals reached 5 days post-fertilization (dpf). All experimental practices were approved by the Institutional Ethics Committee for Animal Use (CEUA Unochapecó; Protocol #009/2019) and respected the European Community instructions.

The protocol, illustrated in Figure 1, consisted of four experimental rounds conducted on different days, totaling 20 animals per experimental group (5 larvae from each group per experimental round) (Bertoncello et al., 2018; Decui et al., 2020). Larvae were randomly allocated into the experimental groups using a computerized random number generator. To improve research reporting, animal housing and experimental procedures followed the ARRIVE guidelines (Percie du Sert et al., 2020). The concentrations of each treatment, the time elapsed between treatment and PTZ exposure, and the PTZ concentration were defined based on previous studies (Bertoncello et al., 2018; Decui et al., 2020).

The experimental groups were: “sham” (animals did not experience any drug treatment and were not exposed to PTZ), “medium + PTZ” (animals were treated with embryo medium and exposed to PTZ), “diazepam + PTZ” (positive control group; animals were treated with diazepam at 75 μM to corroborate the zebrafish response to a clinically used antiseizure drug in the PTZ-induced seizure model), “luteolin + PTZ” (animals were treated with luteolin at 1 μM and exposed to PTZ), and “micronized luteolin + PTZ” (animals were treated with micronized luteolin at 1 μM and exposed to PTZ) (Bertoncello et al., 2018; Decui et al., 2020).

Considering the occurrence of each seizure stage and the latency to reach each seizure stage (Figures 2, 3), the exposed data corresponds to the groups “medium + PTZ”, “diazepam + PTZ”, “luteolin + PTZ”, and “micronized luteolin + PTZ” since the “sham” group did not experience seizures. Considering the animals’ motor function and genetic markers (Figures 4, 5), the exposed data corresponds to “sham” and “medium + PTZ” groups. We intended to verify the neuroprotective potential of luteolin and micronized luteolin through these markers. However, as the seizures did not alter the motor function of the animals as well as the genetic parameters, it would not make sense to investigate the effects of luteolin and micronized luteolin on these biomarkers.

Each zebrafish larva was exposed to its respective treatment for 30 min prior to PTZ exposure (Bertoncello et al., 2018; Decui et al., 2020). A total of 20 larvae per group were individually placed in 6-well plates (1 larva per well, in 5,000 μL of the corresponding solution) containing embryo medium, diazepam (75 μM), luteolin (1 μM), or micronized luteolin (1 μM) (Bertoncello et al., 2018).

Immediately after the treatment, each zebrafish larvae were individually exposed to PTZ solution (3 mM) in 6-well plates (1 larva/well with 5,000 μL of solution) for 10 min. Under these conditions, animals consistently exhibit seizure-like behavior and respond to the positive control drug (Bertoncello et al., 2018; Decui et al., 2020). The animals’ exposure to PTZ was recorded for further analysis. Each video was analyzed by trained and experienced observers blindly. The occurrence of each seizure stage and the latency for the first signal of each seizure stage were evaluated (Bertoncello et al., 2018; Decui et al., 2020). Three convulsive stages were considered: stage I - a dramatic increase in swimming activity; stage II - swimming behavior in whirlpools, and stage III–seizures similar to the clonus state, followed by loss of posture when the animal falls to the side and remains immobile for 1–3 s (Baraban et al., 2005; Bertoncello et al., 2018). After exposure to PTZ, the animals were placed in housing plates (containing embryo medium) according to their respective experimental group. Zebrafish larvae remained in housing plates for 24 h until being tested in the novel tank. The experimental number in the PTZ-induced seizure protocol is 20 (n = 20).

The animals’ motor function 24 h after PTZ exposure was analyzed by the larvae' novel tank paradigm, adapted from previous studies (Bertoncello et al., 2018; Pedroso et al., 2022). Zebrafish larvae were individually tested in a 24 wells-plate (1 larva/well with 120 μL of embryo medium). The animal locomotion was recorded for 6 min and further analyzed using the ANY-Maze^® software (Stoelting Co., Wood Dale, IL, United States).

The analyzed parameters were: total distance traveled, number of line crossings (transitions between the different zones of the plate), mean speed, immobility time, distance traveled in the center zone, and time spent in the center zone. Firstly, we compared larvae that did not experience any drug or PTZ (sham group) with larvae treated with embryo medium and submitted to PTZ-induced seizures (medium + PTZ group). As the seizures did not alter the motor function of the animals, the effects of luteolin and micronized luteolin on biomarkers of motor function were not investigated.

To assess whether luteolin or micronized luteolin could prevent seizure-induced motor changes, we initially compared the profiles of the “Sham” group and the “Medium + PTZ” group. Since no significant differences were observed between these groups, further analyses to investigate the potential effects of luteolin or micronized luteolin on motor function were not performed.

Records presenting interferences (like shadows) that affected the quality of analyses by the ANY-Maze^® recording software were discarded. This procedure resulted in two animals per group being removed from the novel tank analyses. The experimental number in the novel tank test is 18 (n = 18).

To investigate if the seizure occurrence had any effects on genetic markers of the inflammatory response (IL-1β, IL-6, and TNF-α), mTOR signaling pathway (p70S6Ka and p70S6Kb), neurogenesis (BDNF), and apoptosis (caspase-3), we performed quantitative polymerase chain reactions (qPCR) analyses. Firstly, we compared larvae that did not experience any drug or PTZ (sham group) with larvae treated with embryo medium and submitted to PTZ-induced seizures (medium + PTZ group). As the seizures did not alter the analysed transcript levels, the effects of luteolin and micronized luteolin on these biomarkers were not investigated.

For molecular analyses, a second set of 60 animals per experimental group was submitted to the protocol. Zebrafish larvae were cryoanesthetized and euthanized by decapitation after being submitted to the protocol (Gawel et al., 2024). Molecular experiments were performed using n = 3 (Ghaddar et al., 2020; Mazzolini et al., 2020; Pedroso et al., 2022). Each sample consisted of a pool of 20 whole zebrafish larvae (Gawel et al., 2024).

Total RNA was isolated from samples using the PureLink RNA Mini Kit following the manufacturer’s recommendations. The total RNA was quantified by using the Qubit RNA HS Assay Kit. The cDNA was synthesized using the High-Capacity cDNA Reverse Transcription Kit. An average of 0.5 μg of extracted RNA in a reaction with a final volume of 20 μL was used to synthesize cDNA. cDNA quantification was performed by using the Qubit dsDNA HS Assay Kit, and the samples were subsequently diluted to a final concentration of 5 ng‧μL^-1.

Following the manufacturer’s recommendations, qPCR was performed by using the PowerUp SYBR Green Master Mix. β-actin was used as an internal control to normalize the expression of genes of interest (Almeida et al., 2021; Garbinato et al., 2021). Based on our data, the expression of β-actin was not altered by treatment, validating its use as an appropriate housekeeping gene for normalization in this study (Tang et al., 2007). All primer sequences are pointed out in Table 1 (Tang et al., 2007; Mei et al., 2008; Frank et al., 2017; van der Vaart et al., 2017; Kirsten et al., 2018; Carty et al., 2019).

Each reaction contained 10 ng of cDNA and 0.5 mM of each primer in a final volume of 10 μL. Each sample was analyzed in triplicate. The PCR cycles had the following conditions: 50 °C for 2 min, 95 °C for 10 min, 40 cycles at 95 °C for 15 s, 60 °C for 1 min. The dissociation occurred at 95 °C (1.6 °C‧s^-1) for 15 s, then 60 °C (1.6 °C‧s^-1) for 1 min, and, finally, 95 °C (1.6 °C‧s^-1) for 15 s. The equipment used to perform qPCRs was QuantStudio 3 (Thermo Fisher Scientific). Relative gene expression levels were determined by applying the RQ = 2 ^^ΔΔCt method (Livak and Schmittgen, 2001).

To assess whether luteolin or micronized luteolin could prevent seizure-induced gene expression changes, we initially compared the profiles of the “Sham” group and the “Medium + PTZ” group. Since no significant differences were observed between these groups, further analyses to investigate the potential modulatory effects of luteolin or micronized luteolin on gene expression were not performed.

First, the normality of the data was analyzed by the Shapiro-Wilk test. To investigate the influence of the treatments (medium, diazepam, luteolin, micronized luteolin; independent variable) on the occurrence of each seizure stage, Fisher’s exact test (two-tailed) was used to carry out pairwise comparisons of all treatments in PAST v.4.03 (Hammer et al., 2001). Pairwise p-values were adjusted for multiple comparisons using Bonferroni correction with α = 0.05 (Rice, 1989). The Kruskal–Wallis test, followed by Dunn’s post hoc test, was implemented to investigate the influence of the treatments (medium, diazepam, luteolin, micronized luteolin; independent variable) on the latency to reach each seizure stage (dependent variable). To investigate if the seizure occurrence had any effects on motor function and molecular parameters, we performed “sham” and “medium + PTZ” comparisons using Welch’s t-test. Results were considered significant at a p < 0.05 level. GraphPad Prism was used to produce graphs. Seizure occurrence data were expressed as the number of animals that reached each seizure stage. Latency, behavioral, and molecular data were expressed as median with interquartile range.

The scanning electron microscopy analyses (Supplementary Figure S1) showed luteolin presenting irregular particle size and significant agglomeration (Supplementary Figure S1a). Micronized luteolin presented a homogeneous structure (Supplementary Figure S1b). Luteolin has an average particle size of 22.75 µm, and micronized luteolin showed an average size of 2.31 µm.

At the differential scanning calorimeter (Supplementary Figure S2), luteolin showed an endothermic peak at 285.33 °C with ΔH = 38.76 J∙g^-1, which is characteristic of its melting point. A change in the melting point can be seen in micronized luteolin with the appearance of two endothermic peaks, the first at 249.66 °C with ΔH = 3.23 J∙g^-1 and the second peak at 321.94 °C with ΔH = 30.67 J∙g^-1, which indicates a change in the crystalline structure, represented by the change in the melting point of the compound.

The treatments diazepam, luteolin, and micronized luteolin decreased the seizure occurrence (Figure 2). Stage I occurrence was reduced by all treatments compared to medium (negative control) (Bonferroni corrected significance level < 0.05). The occurrence of seizure stage II was reduced by diazepam, luteolin, and micronized luteolin treatments (Bonferroni corrected significance level < 0.001, 0.01, and 0.01, respectively). Finally, the occurrence of the seizure stage III was reduced by diazepam, luteolin, and micronized luteolin treatments (Bonferroni corrected significance level < 0.001, 0.01, and 0.001, respectively).

The treatments diazepam, luteolin, and micronized luteolin slowed the seizure development compared to embryo medium (Figure 3). Diazepam, luteolin, and micronized luteolin treatments increased the latency to stage I (Kruskal–Wallis statistic = 23.74; p < 0.0001), II (Kruskal–Wallis statistic = 37.06; p < 0.0001), and III (Kruskal–Wallis statistic = 41.80; p < 0.0001).

There were no differences among the “sham” and “medium + PTZ” groups (Figure 4) for total traveled distance (Welch corrected t = 0.854; df = 33.55; p = 0.143), line crossing (Welch corrected t = 0.560; df = 32.79; p = 0.197), mean speed (Welch corrected t = 1.400; df = 31.53; p = 0.225), immobility time (Welch corrected t = 1.350; df = 33.63; p = 0.063), distance traveled in center zone (Welch corrected t = 1.213; df = 31.10; p = 0.264), and time spent in center zone (Welch corrected t = 0.132; df = 32.05; p = 0.134).

There were no differences between the “sham” and “medium + PTZ” groups for molecular results (Figure 5). Our results show that there were no changes in the IL-1β (Welch corrected t = 0.990; df = 2.654; p = 0.403), IL-6 (Welch corrected t = 0.487; df = 3.797; p = 0.652), and TNF- α (Welch corrected t = 0.375; df = 2.631; p = 0.735) transcript levels. Also, there were no changes in the p70S6Ka (Welch corrected t = 0.594; df = 2.677; p = 0.598) and p70S6Kb (Welch corrected t = 2.943; df = 2.019; p = 0.097) transcript levels. Finally, there were no changes (Figure 5) in the BDNF (Welch corrected t = 0.191; df = 2.705; p = 0.861) and caspase-3 (Welch corrected t = 0.563; df = 2.266; p = 0.623) transcript levels in zebrafish larvae 24 h after PTZ-induced seizures.