Scorpion venom peptide HsTx2 suppressed PTZ-induced seizures in mice via the circ_0001293/miR-8114/TGF-β2 axis

The peptide HsTx2 (purity > 95%) and the random scrambled peptide (AGERSRKILKREHGHICMVTRGAKK) (purity > 95%) were commercially synthesized and provided by Wuhan Bioyeargene Biotechnology Co., Ltd. (China).

BALB/c mice (weight 20–22 g; 8 weeks) were purchased from Hunan SJA Laboratory Animal Co., Ltd. (Certificate no. 43004700043639, China) and housed under laboratory conditions (relative humidity of 45–55%, 12-h light/dark cycle, freely available food and water) at room temperature of 20–23 °C. All experimental protocols were approved by the Animal Experimental Ethical Inspection of Kunming Medical University (KMMU2020076). The epilepsy mouse model was induced by administering PTZ (Sigma–Aldrich, St. Louis, MO, USA) [32]. Briefly, mice were intraperitoneally injected with PTZ (35 mg/kg) once every other day for a total of 10 injections (from day 1 to day 20). After each injection, all animals were immediately observed for 30 min. Mice presenting at least three consecutive seizures with a score of 4 or 5 were considered fully kindled. Dosing was stopped for 1 week, and the model mice were used for experiments. The mice were randomly divided into the following groups. Each group was first given an intraperitoneal injection of pretreatment. PTZ was intraperitoneally injected after 30 min of pretreatment, and behavior observation and/or EEG monitoring were started immediately after PTZ was administered. The mice were randomly divided into the following groups: (1) NC group: mice receiving 0.9% (W/V) NaCl (intraperitoneally, 0.4 mL); (2) PTZ group: mice receiving PTZ (as described above); (3) ethosuximide group (positive control): mice receiving ethosuximide and PTZ; (4) PTZ + HsTx2 group: mice receiving HsTx2 (intraperitoneally, 0.5 and 1 nmol/kg) and PTZ; (5) PTZ + random scrambled peptide group: mice receiving random scrambled peptide (intraperitoneally, 1 nmol/kg) and PTZ; (6) PTZ + HsTx2 + si-circ group: mice receiving HsTx2, si_0001293 (knockdown of circ_0001293 induced by the intrahippocampal injection of a specific siRNA) and PTZ; and (7) PTZ + HsTx2 + si-circ + miR mimic group: mice receiving HsTx2, si_0001293, agomiR-8114 mimic (overexpression of miR-8114 induced by an intrahippocampal injection of agomiR-8114 mimic) and PTZ.

Thirty minutes after each behavioral observation experiment, the mice were anesthetized with 4% chloral hydrate solution and then intracardially perfused with phosphate-buffered saline for 3 min, and the hippocampal tissue was collected. Protein samples were used immediately for biochemical assays or stored in a − 80 °C freezer for total RNA extraction and western blot testing.

Behavioral seizure scoring and electrophysiological (EEG) evaluations were performed at the last PTZ administration. Two unipolar scalp electrodes were placed on the bilateral temporal skin of the mice. Then, the mice were allowed to move freely in transparent cages. The baseline EEG was recorded for approximately 15 min before the injection of PTZ or saline, and then the EEG was recorded for at least 30 min using the numerical acquisition system (BIOPAC System Inc., Goleta, CA, USA, Model MP150). An electrophysiological seizure was defined as a seizure with a high frequency (> 5 Hz) and high amplitude (> 2 times the baseline) that lasted for more than 5 s. The seizure intensity was assessed based on the Racine scale: stage 0, no response; stage 1, mouth and facial movements; stage 2, head nodding; stage 3, forelimb clonus; stage 4, rearing; stage 5, rearing and falling; stage 6, death [33]. The behavioral data captured by the synchronized video recording system were used to confirm EEG seizure activity.

Primary astrocyte cultures were prepared from the hippocampal tissue of postnatal BALB/c mice on day 1 as previously described [34]. Cell culture reagents for astrocytes were purchased from Thermo Fisher Scientific (Waltham, MA). Hippocampal tissues were isolated and trypsinized. Cells were centrifuged, and the supernatants were removed. The pellets were suspended and cultured. The cultures were further purified by shaking after reaching 90% confluence. The culture medium was changed twice a week. Each time, the culture was pipetted up and down gently to remove loosely attached oligodendrocytes, microglia, and neurons. Purified ACSC2 + cells were then obtained by staining cultures with an anti-ACSC2-APC antibody (Miltenyi Biotec, Germany) and subjecting them to flow cytometry analysis with a FACS-Canto flow cytometer (BD Bioscience, USA).

The animals were anesthetized and transcardially perfused with 4% paraformaldehyde in phosphate buffer. Brains were removed and embedded in paraffin. Brain blocks were then sectioned into serial arrays of 3-μm-thick sections. The tissue sections were deparaffinized, rehydrated and subjected to antigen retrieval. Then, sections and cultured astrocytes were permeabilized with 0.4% Triton X-100 for 10 min and blocked with goat serum (cat. no. ab7481; Abcam, UK) for 1 h to eliminate nonspecific staining. Sections and cultured astrocytes were incubated with a mixture of a rabbit anti-TGF-β2 antibody (dilution 1:250, cat. no. ab113670; Abcam, UK) and rabbit anti-GFAP antibody (dilution 1:200, cat. no. ab7260; Abcam, UK) overnight at 4 °C. Alexa Fluor-conjugated secondary antibodies (cat. no. ab150077; Abcam, UK) were then incubated with the samples for 1 h at room temperature. The cell nucleus was stained with 0.1% Hoechst 33342 (Sigma–Aldrich, USA) for 5 min at room temperature. TGF-β2 and GFAP staining was observed using a Nikon Eclipse 80i microscope (Nikon Corporation).

For this experiment, miR-8114 mimic, miR-NC mimic, miR-8114 inhibitor, miR-NC inhibitor, specific small interfering RNAs (siRNAs) targeting circ_0001293 (si-Circ) and TGF-β2 (si-TGF-β2), and siRNA-negative control (si-NC) were purchased from Shanghai GenePharma (China). The sequences of circ_0001293 were inserted into a pcDNA3.1 plasmid to obtain the circ_0001293 overexpression plasmid pcDNA3.1-circ_0001293 (oe-circ), and an empty pcDNA3.1 plasmid was used as the negative control (oe-NC). Plasmid DNA, siRNA, miR-mimic or miR-inhibitor was transfected into astrocytes (1 × 10⁵), which were subcultured at a density of 80%, with Lipofectamine 2000 reagent (Invitrogen; Thermo Fisher Scientific, Inc., USA) at 37 °C. Forty-eight hours after transfection, the transfection efficiency was detected using RT‒qPCR and Western blotting, and then subsequent experiments were performed.

RNA22 (https://​cm.​jefferson.​edu/​rna22/​Interactive/​) online software was used to predict target binding sites of miRNAs in circRNAs and mRNAs. PmirGLO-circ_0001293-wild (circ-WT)/-mutant (circ-MUT) type and PmirGLO-TGF-β2-WT/-MUT type reporter plasmids were provided by Shanghai GenePharma Co., Ltd. HEK293 cells (2 × 10⁵/well) were cotransfected with the circ-WT/-MUT or TGF-β2-WT/-MUT plasmid and miR-NC mimic or miR-8114 mimic using Lipofectamine 2000 reagent (Invitrogen, USA) at 37 °C. At 48 h post-transfection, luciferase activity was determined using the dual-luciferase reporter assay system (Promega Corporation, USA). Firefly luciferase activities were normalized to Renilla luciferase activities.

The levels of the inflammatory cytokines TNF-α (Abcam, UK), IFN-γ (Abcam, UK), IL-6 (Abcam, UK), IL-1β (Abcam, UK), IL-10 (Abcam, UK), and TGF-β2 (Solarbio, China) in the hippocampus were measured using corresponding ELISA kits according to the manufacturer’s instructions. Absorbance was determined using a microplate spectrophotometer (BioTeke, China).

Total RNA was extracted from cells using TRIzol® reagent (Invitrogen, USA) according to the manufacturer's protocol. First-strand cDNAs were synthesized from total RNA using the PrimeScript™ RT reagent kit (Takara Biotechnology Co., Ltd., Japan) according to the manufacturer’s instructions. RT–qPCR was subsequently performed using the SYBR-Green qPCR kit (Thermo Fisher Scientific, Inc., USA) according to the manufacturer’s protocols. The following primer sequences were used for RT‒qPCR: miR-8114, forward 5′-TCACCCATCTCCTCTCC-3′ and reverse 5′-TGTCGTGGAGTCGGC-3′; TGF-β2, forward 5′-CCAGGGGGAAGGAGGTCATA-3′ and reverse 5′-CTTCGGCAGACACGTGTTTG-3′; circ_0001293, forward 5′-AAATCCTCTGCAGCCCTTCC-3′ and reverse 5′-TTCATCTCTTGGTCGAGGCG-3′; U6, forward 5′-CTCGCTTCGGCAGCACA-3′ and reverse 5′-AACGCTTCACGAATTTGCGT-3′; and β-actin, forward 5′-CACTGTGCCCATCTACGAGG-3′ and reverse 5′-TAATGTCACGCACGATTTCC-3′. RT‒qPCR experiments were performed using an Applied Biosystems 7900HT Fast Real-time PCR system (Applied Biosystems; Thermo Fisher Scientific, Inc., USA). The relative expression levels were calculated using the 2-ΔΔCq method [35] and normalized to those of the internal reference genes β-actin (mRNA) and U6 (miRNA).

Tissues and cells were isolated, total proteins were extracted using RIPA buffer, and protein concentrations were measured using a BCA Protein Determination kit (Pierce Biotechnology, USA). Equal quantities of protein samples were separated on 10% SDS–PAGE gels and transferred to polyvinylidene difluoride membranes, which were then washed and blocked. Specific primary antibodies against TGF-β2 (1:1000; cat. no. ab205150; Abcam, UK), p-ERK (1:500; cat. no. ab47310; Abcam, UK), p-p38 (1:1000; cat. no. ab240335; Abcam, UK), p-JNK (1:1000; cat. no. ab124956; Abcam, UK), p–p65 (1:1000; cat. no. ab76302; Abcam, UK), and β-actin (1:5000; cat. no. ab8226; Abcam, UK) were incubated with membranes, followed by an incubation with an HRP-conjugated secondary antibody (1:5000; cat. no. ab20272; Abcam, UK). Antibody binding was detected using enhanced chemiluminescence reagent (Thermo Fisher Scientific, Inc., USA). ImageJ (Version 1.49; NIH, USA) was used to analyze the gray value of each band on the membrane.

Forty-eight hours after cells were seeded in 96-well plates at a density of 2500 cells per well, cell viability was assessed using a CCK-8 assay (Solarbio, China) according to the manufacturer’s instructions. Absorbance was measured using a microplate spectrophotometer (BioTeke, China).

All experiments were performed in triplicate. Statistical analyses were performed using GraphPad Prism 7 (GraphPad Software Inc., San Diego, CA, USA). The results are presented as the means ± standard deviations (SD). Two groups were compared with Student’s t test, and three or more treatments or groups were compared with one-way ANOVA followed by the Tukey–Kramer post hoc test. A P value < 0.05 was defined as statistically significant.

As shown in our previous study, HsTx2 ameliorates cerebral ischemic injury in rats by exerting neuroprotective effects [15], suggesting that it may be a new potential ASD. The latency to tonic‒clonic seizures, seizure stage and survival of each group of mice are shown in Fig. 1a–c. Ethosuximide and HsTx2 increased the latency to tonic‒clonic seizures compared to the negative control, and 1 nmol/kg HsTx2 increased the latency to tonic‒clonic seizures compared to the positive control (Fig. 1a). Our spontaneous seizure analysis showed that both ethosuximide and HsTx2 decreased spontaneous seizure severity compared with the negative control, but 1 nmol/kg HsTx2 did not change spontaneous seizure severity compared with the positive control (Fig. 1b). The survival ratio was increased by both ethosuximide and HsTx2 (Fig. 1c). Compared with 0.5 nmol/kg HsTx2, greater antiseizure activity was observed for 1 nmol/kg HsTx2. Thus, 1 nmol/kg HsTx2 was used in subsequent assays. As shown in Fig. 1d, the amplitude of seizure spike waves was significantly increased in the PTZ group compared to the NC group. Interestingly, HsTx2 reduced the amplitude of seizure spike waves in PTZ-induced mice. HsTx2 decreased the levels of TNF-α, IFN-γ and IL-6 but increased the levels of IL-10 and TGF-β2 (Fig. 1e). A randomly scrambled peptide was used as a control. In Fig. 1f, the treatment of randomly scrambled peptide could not reverse the decrease of TNF-α, IFN-γ and IL-6, and the increased of IL-10 and TGF-β2 levels by PTZ stimulation. RNA-depleted total RNA sequencing was performed to profile circRNA expression and screen potential driver circRNAs involved in the effects of HsTx2 on epilepsy. The heatmap of identified differentially expressed circRNAs among hippocampal tissues from the control group, PTZ group and HsTx2 group is shown in Fig. 1g, and some of those differential circRNAs were validated to be related to the therapeutic effects of HsTx2 on epilepsy. We found that circ_0001293 expression was reduced by PTZ administration, but this change was reversed by HsTx2 treatment (Figs. 1g and 8a). Therefore, we further explored the underlying molecular mechanism responsible for the effects of circ_0001293 on astrocytes. CircRNAs have been shown to function as miRNA sponges in the cytoplasm [28]. We first predicted the potential target miRNAs in the CircNet database using bioinformatics to understand the actions of circ_0001293. As shown in Fig. 1h, circ_0001293 (red) was predicted to interact with mmu-miR-8114 and rno-miR-666-3p (blue), but rno-miR-666-3p is a circRNA in rats. Hence, miR-8114 was considered the miRNA target of circ_0001293. Subsequently, we predicted the potential target mRNAs (blue) of miR-8114 (yellow) to design the miRNA‒mRNA regulatory network using TargetScan and miRanda software. Interestingly, we identified TGF-β2 as a target gene of miR-8114. Ultimately, the circRNA–miRNA‒mRNA network of the circ_0001293/miR-8114/TGF-β2 axis was assessed in subsequent studies.

Using the bioinformatics database starBase to search for potential targets, a putative interaction between miR-8114 and circ_0001293/TGF-β2 was identified, and the target binding sequence is shown in Fig. 2a and b. A dual-luciferase reporter assay revealed that circ-WT/TGF-β2-WT and miR-8114 mimic cotransfection significantly decreased the luciferase activities in HEK293 cells (Fig. 2c, e), while circ-MUT/TGF-β2-MUT and miR-8114 mimic cotransfection failed to alter the luciferase activity in HEK293 cells. Transfection of oe-circ into HEK293 cells significantly decreased miR-8114 expression (Fig. 2d), and transfection of miR-8114 mimic into HEK293 cells significantly decreased TGF-β2 expression (Fig. 2f). Subsequently, the circRNA–miRNA‒mRNA network of the circ_0001293/miR-8114/TGF-β2 axis was explored. RT–qPCR was conducted to examine the expression of circ_0001293, miR-8114 and TGF-β2, and the overexpression of circ_0001293 increased the expression of circ_0001293 and TGF-β2 and decreased miR-8114 expression, but these effects were reversed by the overexpression of miR-8114 (Fig. 2g). Knockdown of circ_0001293 reduced the expression of circ_0001293 and TGF-β2 and increased miR-8114 expression, but miR-8114 upregulation and TGF-β2 downregulation were reversed upon the downregulation of miR-8114. Western blot results showed increased and decreased levels of the TGF-β2 protein following circ_0001293 upregulation and downregulation, respectively, but these changes were reversed by miR-8114 mimic and miR-8114 inhibitor treatment, respectively (Fig. 2h). Based on these observations, circ_0001293 sponges miR-8114, and miR-8114 targets TGF-β2.

Astrocytes were purified from hippocampal tissues to confirm the molecular mechanisms underlying the antiseizure effect of HsTx2. As shown in Fig. 3a, 98.2% of the microbead-labeled cells were ACSA2+. Immunofluorescence staining confirmed that over 90% of bead-bound cells expressed GFAP (Fig. 3b). These astrocytes were treated with different concentrations of HsTx2 (0.1, 0.5, 1, 5 or 10 nM) for 48 h to examine whether HsTx2 regulated the expression of circ_0001293, miR-8114 and TGF-β2 in IL-1β-stimulated astrocytes. As shown in Fig. 3c and d, HsTx2 increased the expression of circ_0001293 and TGF-β2 and reduced miR-8114 expression. Next, the viability and inflammation of astrocytes were determined. As shown in Fig. 3e, cell viability was reduced after IL-1β induction but increased by HsTx2 treatment. ELISA results showed that IL-1β not only increased the levels of TNF-α, IFN-γ and IL-6 but also decreased the levels of IL-10 and TGF-β2, which were reversed upon HsTx2 treatment (Fig. 3f). Various studies have shown that the role of nuclear factor kappa B (NF-κB) in different neurodegenerative diseases is related to inflammation [36]. Elevated levels of the transcription factor NF-κB in the epileptic brain induce the transcription of genes encoding proinflammatory mediators, such as IL-1β, IL-6 and TNF-α [37]. As shown in our previous study, HsTx2 also regulates the MAPK signaling pathway [15]. Thus, we investigated whether HsTx2 regulated the NF-κB and MAPK pathways in IL-1β-treated astrocytes. IL-1β increased the levels of p-ERK, p-p38, p-JNK and p-p65. The levels of p-ERK and p-p38 increased continuously after HsTx2 treatment, but the p-JNK and p-p65 levels were reduced (Fig. 3g). Compared with 0.1, 0.5, 5 and 10 nM HsTx2, the effects of 1 nM HsTx2 on TGF-β2 expression, cell viability, NF-κB pathway activation and inflammatory cytokine levels were significant. Thus, cells treated with 1 nM HsTx2 for 48 h were used in subsequent experiments.

We constructed stable circ_0001293 knockdown cell lines using si-circ_0001293, and si-circ_0001293 significantly downregulated the expression of circ_0001293 (Fig. 4a). The results of the RT–qPCR analysis indicated that IL-1β reduced circ_0001293 levels, which was reversed by HsTx2 treatment, but circ_0001293 expression was finally repressed by si-circ_0001293 (Fig. 4b). Furthermore, cell viability and inflammatory cytokine levels were detected using the CCK-8 assay and ELISA, respectively. As shown in Fig. 4c, cell viability was reduced after IL-1β treatment but was alleviated by HsTx2 treatment, which was terminally inhibited by si-circ_0001293. The levels of the inflammatory cytokines TNF-α, IFN-γ and IL-6 were increased after IL-1β treatment but reduced by HsTx2 treatment, which was terminally increased by si-circ_0001293 (Fig. 4d). In contrast, IL-1β reduced the levels of IL-10 and TGF-β2, which were reversed by HsTx2 treatment, but the levels of IL-10 and TGF-β2 were ultimately repressed by si-circ_0001293. Based on these findings, HsTx2 alleviates IL-1β-induced astrocyte inflammation by upregulating circ_0001293.

We next overexpressed miR-8114 in astrocytes and established cell lines stably overexpressing miR-8114 (Fig. 5a). The RT–qPCR results showed that circ_0001293 expression was significantly inhibited by IL-1β but increased after circ_0001293 overexpression, which was terminally repressed by miR-8114 mimic transfection (Fig. 5b). In contrast, miR-8114 expression was increased after IL-1β treatment but reduced by circ_0001293 overexpression, which was ultimately increased by the miR-8114 mimic. In addition, cell viability was reduced by IL-1β treatment but increased after circ_0001293 overexpression and was ultimately reduced by miR-8114 mimic transfection (Fig. 5c). Furthermore, ELISAs showed that the levels of the inflammatory cytokines TNF-α, IFN-γ and IL-6 were increased after IL-1β treatment but reduced upon the overexpression of circ_0001293 and increased after transfection of the miR-8114 mimic (Fig. 5d). In contrast, IL-1β reduced the levels of IL-10 and TGF-β2, changes that were reversed by oe_0001293, but the levels of IL-10 and TGF-β2 were reduced by miR-8114 overexpression. Therefore, circ_0001293 alleviates IL-1β-induced astrocyte inflammation by targeting miR-8114.

We constructed stable TGF-β2 knockdown cell lines using si-TGF-β2, and si-TGF-β2 significantly downregulated TGF-β2 expression (Fig. 6a). As shown in Fig. 6b, miR-8114 expression was increased and TGF-β2 expression was reduced in the IL-1β group compared to the NC group. Inhibition of miR-8114 decreased miR-8114 expression and increased TGF-β2 expression, whereas TGF-β2 expression was reversed upon TGF-β2 knockdown. The CCK-8 results showed that IL-1β reduced cell viability, which was reversed upon miR-8114 inhibition. Cell viability was reduced by TGF-β2 knockdown (Fig. 6c). ELISAs showed that IL-1β increased the levels of TNF-α, IFN-γ and IL-6, but the levels of these cytokines were reduced by the miR-8114 inhibitor and ultimately increased by downregulating TGF-β2 (Fig. 6d). In contrast, IL-1β reduced the levels of IL-10 and TGF-β2, changes that were reversed by the miR-8114 inhibitor, but the levels of IL-10 and TGF-β2 were finally reduced by downregulating TGF-β2. Collectively, these results suggest that miR-8114 inhibition alleviates IL-1β-induced astrocyte inflammation by targeting TGF-β2.

We next examined the expression of circ_0001293 and its downstream genes miR-8114 and TGF-β2 to further validate whether HsTx2 exerts its function via the circ_0001293/miR-8114/TGF-β2 axis. As shown in Fig. 7a and b, IL-1β reduced the expression of circ_0001293 and TGF-β2, and the cell viability. HsTx2 treatment reversed these changes, and the expression of circ_0001293 and TGF-β2 was repressed in cells transfected with si-circ_0001293, miR-8114 mimic and si-TGF-β2. In contrast, miR-8114 expression was increased after IL-1β treatment but reduced by HsTx2 treatment, and its expression was increased in cells transfected with si-circ_0001293, miR-8114 mimic and si-TGF-β2. ELISA results showed that IL-1β not only increased the levels of TNF-α, IFN-γ and IL-6 but also decreased the levels of IL-10 and TGF-β2. HsTx2 treatment reduced the increases in TNF-α, IFN-γ and IL-6 levels and reversed the decrease in IL-10 and TGF-β2 levels after IL-1β stimulation, and the effects of HsTx2 were reversed after the transfection of si-circ_0001293, miR-8114 mimic and si-TGF-β2 (Fig. 7c). Western blot results showed that IL-1β induced increased p-ERK and p-p38 levels, and the increases were continuous with HsTx2 treatment, but si-circ_00001293, miR-8114 mimic and si-TGF-β2 reversed the effects of HsTx2 on p-ERK and p-p38 levels (Fig. 7d). IL-1β increased the levels of p-JNK and p-p65, HsTx2 treatment reversed these changes, and si-circ_00001293, miR-8114 mimic and si-TGF-β2 transfection ultimately increased the levels of these proteins. These data suggest that HsTx2 alleviates IL-1β-induced astrocyte inflammation via the circ_0001293/miR-8114/TGF-β2 axis.

We next assessed and confirmed whether the potential antiseizure activity of HsTx2 was mediated by the circ_0001293/miR-8114/TGF-β2 axis in vivo. Epileptic mice were injected with HsTx2, lenti-si-circ_0001293 or miR-8114 mimic, and the expression levels of these genes were analyzed. As shown in Fig. 8a, the expression of circ_0001293 and TGF-β2 was inhibited by PTZ but increased after HsTx2 treatment; their expression was repressed by si-circ_0001293 and miR-8114 mimic. In contrast, miR-8114 expression was increased after PTZ treatment but reduced by HsTx2 treatment, and its expression was increased by si-circ_0001293 and miR-8114 mimic. Both immunofluorescence staining and Western blotting revealed reduced levels of the TGF-β2 protein in animals receiving PTZ, HsTx2 treatment increased its levels, and si-circ_0001293 and miR-8114 mimic decreased levels of the TGF-β2 protein (Fig. 8b and d). In addition, scalp EEG results showed a significant decrease in the amplitude of seizure spike waves in the HsTx2 group compared to the PTZ group, but these effects of HsTx2 were reversed by si-circ_0001293 and miR-8114 mimic (Fig. 8c). ELISA results showed that PTZ increased the levels of TNF-α, IFN-γ, IL-6 and IL-1β, which were reversed by HsTx2 treatment, but the levels of TNF-α, IFN-γ, IL-6 and IL-1β were increased by si-circ_0001293 and miR-8114 mimic (Fig. 8e). In contrast, PTZ reduced the levels of IL-10 and TGF-β2, HsTx2 treatment reversed these changes, and the levels of IL-10 and TGF-β2 were reduced by si-circ_0001293 and miR-8114 mimic. In addition, Western blot results showed that PTZ induced increased p-ERK and p-p38 levels, and the increase was continuous with HsTx2 treatment, but si-circ_00001293 and miR-8114 mimic reversed the effects of HsTx2 on p-ERK and p–p38 levels (Fig. 8f). PTZ increased the levels of p-JNK and p–p65, HsTx2 treatment reversed these changes, and the levels of these proteins were increased by si-circ_00001293 and miR-8114 mimic. Taken together, HsTx2 alleviates epilepsy-induced inflammation via the circ_0001293/miR-8114/TGF-β2 axis.