Experimental paradigm in mice. a Swiss Albino mice or C57BL/6/Tar × CBA/Tar GPR39 knockout and wild-type mice were subjected to acute models of electrically or chemically induced seizures or a chronic model of epileptogenesis [chemical kindling induced by pentylenetetrazole (PTZ)], followed by observation of seizure behavior. Animals were pre-treated with agonists of GPR39: the small molecule TC-G 1008 or ZnCl₂ or a standard anti-seizure drug, valproic acid (VPA), i.p. b Detailed mouse allocation in each of the experimental procedures

Experimental paradigm in mice. a Swiss Albino mice or C57BL/6/Tar × CBA/Tar GPR39 knockout and wild-type mice were subjected to acute models of electrically or chemically induced seizures or a chronic model of epileptogenesis [chemical kindling induced by pentylenetetrazole (PTZ)], followed by observation of seizure behavior. Animals were pre-treated with agonists of GPR39: the small molecule TC-G 1008 or ZnCl₂ or a standard anti-seizure drug, valproic acid (VPA), i.p. b Detailed mouse allocation in each of the experimental procedures

The concentrations of TC-G 1008 in the serum (a) and brain (b). Single dose of TC-G 1008 (20 mg/kg, i.p.) was administered in Swiss Albino mice. Data are expressed as means ± SEM of concentrations over time. n = 4 mice per time point

The effects of single doses of VPA (b, c), TC-G 1008 (d, e), or ZnCl₂ (f, g) on the seizure threshold in the maximal electroshock seizure threshold (MEST) test and the seizure threshold in the 6-Hz seizure threshold test. a Drugs or VEH were administered i.p. in Swiss Albino mice. 30 min later, the MEST or 6-Hz seizure threshold test was performed. b–g The doses of drugs are shown on the x-axes in mg/kg. Data are presented as CS₅₀ (in mA) values with upper 95% confidence limits. n = 20 in each group. p values were determined by the Student’s t-test or one-way ANOVA and Dunnett’s multiple comparison test. *p < 0.05, **p < 0.01, ***p < 0.001

The effects of single doses of TC-G 1008 (b–d) or ZnCl₂ (e–g) on acute seizures induced by kainic acid (KA). a Drugs or VEH were administered i.p. in Swiss Albino mice. 30 min later, KA (40 mg/kg, i.p.) was administered. Immediately after KA administration, the mice were subjected to an evaluation of behavioral seizures for 2 h. b–g The doses of drugs are shown on x-axes in mg/kg. Data are expressed as means ± SEM of seizure severity or latency to scratching and the percentage of animals displaying stage 7 characteristics (SE/death). Latency to scratching was impossible to measure in mice that received 8 mg Zn/kg because they proceeded immediately to advanced seizure stages. b–d n = 10 VEH, n = 10 TC-G 1008 10, n = 9 TC-G 1008 20, n = 8 TC-G 1008 40. e–g n = 10 VEH, n = 10 Zn 2, n = 10 Zn 4, n = 12 Zn 8. p values were determined by one-way ANOVA and Bonferroni’s multiple comparison test or the Chi-square test. *p < 0.05, **p < 0.01

Local field potential (LFP) recordings from the optic tectum of zebrafish larvae exposed to TC-G 1008 (b–d) or ZnCl₂ (e–g) and pentylenetetrazole (PTZ). a Zebrafish larvae were incubated with TC-G 1008 (70 µM), ZnCl₂ (65 µM) (at maximally tolerated concentrations, MTCs) or VEH for 20 h and subsequently exposed to PTZ (20 mM) or VEH for 5 min. The LFP recordings began 5 min after removing the larva from VEH/PTZ solution and lasted 20 min. b–g Results are expressed as means ± SEM of the number of epileptiform-like events, the cumulative duration of epileptiform-like events, and the mean duration of epileptiform-like events. b–d n = 10 VEH VEH, n = 8 VEH PTZ, n = 5 TC-G 1008 VEH, n = 14 TC-G 1008 PTZ. e–g n = 10 VEH VEH, n = 8 VEH PTZ, n = 7 Zn VEH, n = 17 Zn PTZ. p values were determined by one-way ANOVA and Bonferroni’s multiple comparison test. *p < 0.05, **p < 0.01, ***p < 0.001

The effects of chronic treatment with VPA, TC-G 1008, or ZnCl₂ on PTZ-induced epileptogenesis in Swiss Albino mice. a VPA (150 mg/kg), TC-G 1008 (10 mg/kg), ZnCl₂ (8 mg Zn/kg), or VEH were injected i.p. once daily on alternate days during weekdays. 30 min later, PTZ (40 mg/kg) was injected i.p. Immediately after each PTZ injection, the mice were subjected to an evaluation of behavioral seizures, which lasted 30 min. The total number of PTZ injections was 19. Data are expressed as means ± SEM of seizure severity after each injection (b) and the percentage of mice displaying consecutive stage 5 seizures (fully kindled mice) after the last PTZ injection (c). n = 15 in each group. p values were determined by two-way repeated measures ANOVA and Dunnett’s multiple comparison test or Fisher’s exact test. *p < 0.05, **p < 0.01

The effects of GPR39 KO in C57BL/6/Tar × CBA/Tar mice on seizure threshold in the MEST test (a) and on PTZ-induced epileptogenesis (b–d). a WT and GPR39 KO mice were subjected to the MEST test. Data are expressed as CS₅₀ (in mA) with upper 95% confidence limits. n = 20 in each group. P value was determined by the Student’s t-test (b) WT and GPR39 KO mice were injected i.p. once daily with TC-G 1008 (10 mg/kg) or VEH on every alternate day during weekdays. 30 min later, the mice were injected i.p. with PTZ (25 mg/kg). Immediately after each PTZ injection, the mice were subjected to an evaluation of behavioral seizures, which lasted 30 min. The total number of PTZ injections was 14. Data are expressed as means ± SEM of seizure severity after each injection (c) and the percentage of mice displaying consecutive stage 5 seizures after the last PTZ injection (d). One outlier was excluded in the WT TC-G 1008 group leading to n = 8 WT VEH, n = 6 WT TC-G 1008, n = 8 KO VEH, n = 7 KO TC-G 1008. p values were determined by three-way repeated measures ANOVA and Bonferroni’s multiple comparison test or Fisher’s exact test. *p < 0.05, **p < 0.01, ***p < 0.001

The effects of PTZ-kindling in GPR39 KO or WT mice and chronic treatment with TC-G 1008 (10 mg/kg) on zinc levels in serum and hippocampus. a The PTZ kindling in GPR39 KO or WT (C57BL/6/Tar × CBA/Tar) mice consisted of 14 injections of PTZ (25 mg/kg). The biochemical analyses were performed 24 h after the last PTZ injection. b Total zinc (protein bound and [Zn²⁺]_I) was measured in serum by ICP-OES. Data are expressed as means ± SEM. n = 8 WT VEH, n = 6 WT TC-G 1008, n = 7 KO VEH, n = 7 KO TC-G 1008. c Total zinc was analyzed semi-quantitatively in coronal hippocampal sections by LA-ICP-MS. Data are expressed as means ± SEM of counts per second (CPS). n = 4 in each group. d The protein expression level of the putative [Zn²⁺]_I sensor, metal regulatory transcription factor 1 (MTF1), was analyzed in the hippocampus. The results (means ± SEM) are presented as the MTF1/β-actin ratio. n = 7 in each group. e Representative blots of MTF1. The observed band (~ 130 kDa). f–h [Zn²⁺]_I was examined in coronal hippocampal sections by a cell-membrane permeable probe, Zinpyr-1 (ZP-1). The mean ZP-1 grey values ratio between mouse sections in the CA3, CA1, or dentate gyrus (DG) regions of the hippocampus is shown. n = 6 WT VEH, n = 6 WT TC-G 1008, n = 5 KO VEH, n = 6 KO TC-G 1008. i A greyscale image of the hippocampal section after staining with ZP-1. p values were determined by two-way ANOVA and Bonferroni’s multiple comparison test. *p < 0.05

The effects of PTZ-kindling in GPR39 KO or WT (C57BL/6/Tar × CBA/Tar) mice and chronic treatment with TC-G 1008 (10 mg/kg) on the expression of proteins in the GPR39 signaling pathway in the hippocampus. a The PTZ kindling in GPR39 KO or WT (C57BL/6/Tar × CBA/Tar) mice consisted of 14 injections of PTZ (25 mg/kg). The biochemical analyses were performed 24 h after the last PTZ injection. b–d The expression of p-CREB, CREB, BDNF, p-TrkB, and TrkB was normalized to β-actin. The results (means ± SEM) are shown as the p-CREB/CREB, BDNF/β-actin, or p-TrkB/TrkB. b n = 7 in each group. c–d n = 6 in each group. e Representative blots of p-CREB, CREB (~ 46 kDa), BDNF (~ 14 kDa), p-TrkB, TrkB (~ 140 kDa), and β-actin (~ 42 kDa). ^#p-CREB, BDNF, and p-TrkB come from the same blot; ^$CREB and TrkB come from the same blot, thus sharing the corresponding β-actin band. f, h The expression p-ERK1/2, ERK1/2, and SIRT1 was normalized to the total protein concentration. f The results (means ± SEM) are shown as p-ERK1/2/ERK1/2. n = 6 WT VEH, n = 6 WT TC-G 1008, n = 7 KO VEH, n = 7 KO TC-G 1008. h n = 7 WT VEH, n = 6 WT TC-G 1008, n = 5 KO VEH, n = 6 KO TC-G 100. g Representative blots of p-ERK1/2, ERK1/2 (~ 44, 42 kDa) and SIRT1 (the observed band ~ 130 kDa). p values were determined by two-way ANOVA and Bonferroni’s multiple comparison test. *p < 0.05, **p < 0.01

Summary of findings and the proposed mechanism. TC-G 1008 is a small molecule agonist at GPR39 [28]. GPR39 interacts physically with the serotonin 5-HT_1A receptor [78]. TC-G 1008 also binds to the 5-HT_1A receptor [29]. a We found that TC-G 1008 facilitated PTZ-epileptogenesis in vivo by acting selectively at GPR39. However, it increased the activation of CREB in the hippocampus of kindled GPR39 KO mice. b Our data showed that TC-G 1008 acts at GPR39 and other targets, presumably the 5-HT_1A receptor. (+) enhancement