Fig 1.
Status epilepticus (SE)-induced Wnt signaling and associated signaling networks.
(A) Model of Wnt signaling and mTOR activation (via TSC2 and AMPK). TSC2 requires the phosphorylation of both AMPK and GSK3βS9 to fully inhibit the downstream activity of mTOR pathway, including reduced phosphorylation of P70S6K. The figure is based on the literature [21, 24–26] and was created with BioRender.com. (B) Wnt signaling is activated in the early epileptogenic period (pilocarpine induction). Status epilepticus was induced in male 7-week FVBN mice with pilocarpine (220 mg/kg). Representative western blots and quantified data of hippocampal extracts from 3-, 5- or 7-day post-pilocarpine (Pilo; n = 4) or control saline (Sal; n = 4) are shown for pGSK3βS9, GSK3β, β-catenin, and actin (loading control). (C). Expression of selected Wnt pathway targets and components reflects increases in Wnt signaling. Hippocampal mRNA levels for representative Wnt pathway target gene (Axin 2), a Wnt ligand (Wnt8B), and Wnt pathway inhibitors sFRP1 [27, 28] and HBP1 [29, 30] were measured by quantitative RT-PCR at day 5 post-SE (pilocarpine) compared to control mice. (n = 4–6 mice for each group). The data were represented as mean ± SEM using unpaired student’s t-test. (Prism 9.0, Graphpad. *P <0.05, **P < 0.01). (D) Wnt target genes are activated in the early epileptogenic period. A qRT- PCR based array of representative mouse Wnt pathway gene targets (Qiagen) was used to analyze RNA from mouse control and 5-day post-kainate hippocampi. The results were expressed as a volcano plot (left). Genes to the right or left of the vertical grey lines indicate >2-fold (+/-) changes in gene expression. Genes above the horizontal grey line indicate significant changes in the cohort. A list of Wnt-target genes that were significantly induced >2-fold is shown (right). (E, F) Wnt signaling occurs in neurons. Representative confocal images (x63) show the neuronal localization of pGSK3βs9 (green), and (F) β-catenin (green) in hippocampus at day 5 post-SE with pilocarpine, when compared to control mice. [DAPI (blue; all nuclei), GFAP (red; astrocytes) in (E) and NeuN (red, neurons) in (F)]. The percentage of positive pGSK3βS9 in (E), and of β-catenin in (F) neuronal cells of the dentate gyrus (DG) region at 5-days post-SE with pilocarpine and in control mice data are compiled from multiple sections (see Methods) and represented as mean ± SE. *P <0.05, ** P< 0.01, *** P< 0.001 by unpaired student’s t-test, using Prism 9.0 (Graphpad) or MS-Excel. Refer to supplemental materials for more details on the number of replicates and statistical analyses for each experiment.
Fig 2.
Altered pyruvate metabolism at day 5 post-SE.
(A) Pyruvate dehydrogenase 4 (PDK4) mRNA and protein are increased at day 5 post-SE. Quantitative RT-PCR of PDK4 expression (left) and representative western blots with quantified data comparing the expression of PDK4 protein in hippocampus at day 5 following SE with pilocarpine (Pilo) or control (saline)–Right. (n = 4 mice/group). (B) PDK4 is localized primarily in neurons post-SE day 5. Representative confocal images (x63) and percentage of positive PDK4 (green) neuronal cells are higher in the DG region of 5 days post-SE with pilocarpine hippocampi when compared to control mice. [NeuN (red) and Dapi (blue)]. (C) Ser 293 phosphorylation of pyruvate dehydrogenase (PDH) is increased at day 5, post SE. Quantitative and representative western blot of pPDHS293, total PDH, and actin (loading control) in hippocampi of day 5 induced SE with pilocarpine (Pilo) and control (Sal). (n = 4 for each group). (D) Confocal images (x63) and percentage of positive pPDHS293 (red) neuronal cells (see supplemental methods) are higher in the DG region of 5 days post-SE with pilocarpine hippocampi when compared to control mice. [GFAP (green) and Dapi (blue)]. Data are represented as mean ± SEM. *P <0.05, **P < 0.01 by unpaired student’s t-test using Prism 9.0 (Graphpad). Refer to supplemental materials for more details on the number of replicates and statistical analyses for each experiment.
Fig 3.
Status-induced Warburg metabolic enzymes.
(A) The mRNA expression of Warburg metabolic enzyme isoforms is elevated after pilocarpine-induced SE. Quantitative RT-PCR was used to measure the mRNA for hexokinase 2 (HK II), pyruvate kinase M2 isoform (PKM2) and lactate dehydrogenase A-isoform (LDHA) from the hippocampi of the 3-, 5- and 7-day post-status epilepticus mice in (pilocarpine-treated) compared to control (saline) mice. (B) Protein expression of Warburg metabolic enzyme isoforms are elevated after pilocarpine-induced SE. Representative and quantified western blots of hippocampal extracts from 3-, 5- or 7-day pilocarpine (Pilo), or control saline (Sal) are shown for PKM2, LDHA, and actin (loading control, n = 3–5 for each group). (C) Differential localization of PKM2 after pilocarpine-induced SE. Confocal images (x63) and percentage of positive PKM2 (red) astrocytes (see supplemental methods) are higher in the DG region of 5 days post-SE with pilocarpine hippocampus when compared to control mice. [GFAP (green) and Dapi (blue)]. Data are represented as mean ± SEM. *P <0.05, **P< 0.01 by unpaired student’s t-test using Prism 9.0 (Graphpad). Refer to supplemental materials for more details on the number of replicates and statistical analyses for each experiment.
Fig 4.
Metabolite analysis reveals elevated lactate and glucose metabolism in 5-day post-SE hippocampal metabolism—Signatures of Warburg metabolism.
4–5 different hippocampi per group were independently processed and analyzed as described in methods. (A) Hierarchical cluster analysis of metabolites consistent with Warburg metabolism (Metaboanalyst 4.0, [62, 63]). The top 15 metabolites by significance are displayed. Metabolites were segregated into two clusters with clear up- and down-changes in metabolites between control and pilocarpine-treated mice. (B) Warburg effect identified by metabolite set enrichment analysis (MSEA, Metaboanalyst 4.0, [62, 63]). The top 20 metabolic pathways enriched in 5-day post-SE hippocampus relative to control were identified by analysis of metabolites from S1 Table using MSEA. P-value and fold enrichment are displayed. (C) Principal component analysis (Metaboanalyst 4.0, [62, 63]) using the metabolites from S1 Table demonstrates a significant difference between control and 5-day post-SE metabolism. The metabolite principal components showed significant separation (>95% confidence) between control mice (red) and 5-day post pilocarpine-treated mice (green) in 2D plot. (D) Hierarchical cluster analysis of 13C metabolite carbons demonstrates glucose flux into lactate (Metaboanalyst 4.0, [62, 63]). The top 15 13C-labeled metabolites by significance are displayed. Unbiased analysis of 13C-labeled metabolites segregated into two clusters with clear up- and down-changes in metabolites between control and 5-day post-pilocarpine-treated mice. (E) Schematic representation of possible isoptomers arising from [U-13C] glucose in glycolysis in the first turn of the TCA cycle. Each glucose carbon is color-coded for its fate in glycolysis, with the glucose split at the aldolase step, eventually ending at pyruvate. Astrocyte pyruvate is converted to lactate, transported to neurons and subsequently re-converted to pyruvate, where it enters the TCA cycle via pyruvate dehydrogenase (PDH). This gives labeling citrate, 2-oxoglutarate, glutamate and GABA on carbons 4 and 5. Pyruvate remaining in astrocytes can similarly be converted via PDH, but can also undergo the anaplerotic pyruvate carboxylase step, converting to oxaloacetate (OAA) with label at carbons 1, 2 and 3. Subsequent condensation of OAA with acetyl CoA yields some of the citrate pool labeled at carbons 1–5, readily detectable by NMR via the interaction of carbon 3 with carbon 4. GAD67, Glutamate decarboxylase; GS, Glutamine synthetase; GLS, Glutaminase. (F) GAD67, GS, and GLS mRNA Levels at day 5 post-SE. GAD67 and Glutaminase mRNA levels decreased while glutamine synthetase mRNA increased in hippocampus at 5 days post-SE with pilocarpine. (n = 4–6 mice per group). Data are represented as mean ± SEM. *P <0.05, **P < 0.01 by unpaired student’s t-test using Prism 9.0 (Graphpad). Refer to supplemental materials for more details on the number of replicates and statistical analyses for each experiment.
Fig 5.
Induced Wnt signaling post-SE coordinates AMPK and mTOR signaling pathways.
(A) Coordinated inhibition of AMPK and activation of P70S6 Kinase. (Left) Representative western blots with quantification (Right) comparing the expression of pAMPKαT172 and pP70S6KT389 in hippocampi at day 5 following SE with pilocarpine (Pilo) or control (saline). (n = 3–4 mice/group). Data are represented as mean ± SEM. *P <0.05, **P < 0.01 by unpaired student’s t-test using Prism 9.0 (Graphpad). (B and C) Localization of AMPK and mTOR activation by confocal microscopy. Representative confocal sections with quantified analysis of positive pAMPKαT172 (green; B) and pP70S6KT389 (red; C) neuronal cells (see supplemental methods) are increased in the DG region of 5-day post-pilocarpine induced-SE when compared to control saline. [GFAP (red in B; green in C) and Dapi (blue)]. Data are represented as mean ± SEM. *P <0.05, ***P < 0.001 by unpaired student’s t-test using Prism 9.0 (Graphpad). (D) pP70S6KT389 and pGSK3βS9 co-localize to neurons. Immunostaining with pP70S6KT389 (red) and pGSK3βS9 (green) show both co-localize to neuronal cells and are higher in percentage (see supplemental methods) at day 5 post-SE with pilocarpine when compared to control. [Dapi (blue)]. (E) TSC2 phosphorylation inhibited at day 5 post-SE. Representative and quantified western blots of hippocampal extracts from pilocarpine- treated (Pilo), or control saline (Sal) are shown for pTSC2S1387, and total TSC2 (loading control). (n = 3–5 for each group). Data are represented as mean ± SEM. **P < 0.01 by unpaired student’s t-test using Prism 9.0 (Graphpad). Refer to supplemental materials for more details on the number of replicates and statistical analyses for each experiment. (F) Summary and mechanistic model of Wnt-AMPK-mTOR signaling in epileptogenesis. Activation of canonical Wnt signaling results in inhibition of GSK3β, increases β-catenin protein levels, resulting metabolic gene expression, including increases HKII, PKM2 and LDHA consistent with activation of aerobic glucose metabolism. This “Warburg” metabolism results in decreased AMP and inhibition of AMP-dependent protein kinase (AMPK). The coordinated inhibition of GSK3β and AMPK, TSC2 from its inhibitory role in the mTOR pathway, resulting in mTOR activation and a change in seizure susceptibility. The figure was created with BioRender.com.
Fig 6.
Glycolytic inhibitor 2-DG alters downstream AMPK and mTOR signaling, and neuronal gene expression, but does not affect Wnt signaling.
(A) SE-induced Wnt signaling and Warburg metabolic enzyme PKM2 are unaffected by 2DG. The protein expression of pGSK3βS9, pPDHS293, and PKM2 in the epileptogenic period was measured by western blot. Representative blots are on the left and quantified data (n = 4–6 mice per group) on the right. Data are represented as mean ± SEM. *P <0.05, ***P <0.001 by unpaired student’s t-test using Prism 9.0 (Graphpad). (B) The Wnt target gene PDK4 expression was unaffected by 2DG. Quantitative RT-PCR (top panel) shows PDK4 mRNA levels induced at 5-day post-SE relative to control. Representative western blots of PDK4 protein (center panel) and quantified data (bottom panel) are shown. There was no change in the expression levels of PDK4 in hippocampus at day 5 following SE with pilocarpine (Pilo) and treated with 2DG–right panel. (n = 4–6 mice per group). Data are represented as mean ± SEM. *P <0.05, by unpaired student’s t-test using Prism 9.0 (Graphpad). (C) SE-induced activation of AMPK (pAMPKαT172) and of mTOR (pP70S6KT389) is reversed with 2DG treatment. The levels of activated AMPK and activated P70S6K were detected by quantitative western blot from extract of the hippocampus from 5 days post SE mice, in the presences and absence of 2DG (see Methods). Data are represented as mean ± SEM. *P <0.05 by unpaired student’s t-test; (n = 4–6 mice for each group). Representative western blots with quantified data showed 2-DG treatment at day 5 post-SE with pilocarpine inhibit mTOR signaling downstream activation by decreased pP70S6KT389 and activating pAMPKαT172. (n = 4–6 mice for each group) Data are represented as mean ± SEM. *P <0.05 by unpaired student’s t-test using Prism 9.0 (Graphpad). (D and E) SE-induced localization of activated AMPK and mTOR activation is restored with 2DG treatment. Sections from 5-day post-SE hippocampus in the absence or presence of 2DG treatment (see Methods). Confocal images (x63) from treated and control hippocampi were analyzed for positive pAMPKαT172 (D, red), and pP70S6KT389 (E, red) in the DG region. [GFAP (green) and Dapi (blue)]. Refer to supplemental materials for more details on the number of replicates and statistical analyses for each experiment.
Fig 7.
The role of HBP1 and Wnt signaling in acquired seizure susceptibility.
(A) HBP1 gene is expressed in the CA1 and DG regions. The expression pattern of the HBP1 gene was determined in the HBP1-/- mouse through the inserted β-galactosidase gene. Note that the HBP1-/- mouse was generated by insertion of the β-galactosidase gene into intron 1 of the HBP1 genomic loci [72]. Thus, β-galactosidase activity is driven by the HBP1 promoter and is a reporter of the HBP1 promoter activity. Representative slides of HBP1-/- adult mouse hippocampus with β-galactosidase gene inserted into intron 1 of HBP1 gene. (B) Activation of Wnt and mTOR signaling in HBP1-/- and HBP1+/- mice. Representative western blot with quantified data of baseline levels of Wnt and mTOR signaling markers in FVBN mouse strain were increased in HBP1-/- knockout (KO) and HBP1+/- heterozygous (HET) when compared to control (WT). (C-E) Localization of activated Wnt and mTOR signaling in HBP-/- mice by confocal microscopy. Representative confocal sections with quantified analysis of positive pGSK3βS9 (green; x63 in C), β-catenin (green; x20 in D), and pP70S6KT389 (red; x20 in E) neuronal cells are increased in the DG region of KO when compared to WT. [GFAP (red in C and D; green in E) and Dapi (blue)]. (n = 3–5 mice for each group). Data are represented as mean ± SEM. *P < 0.05, **P < 0.01, *** P < 0.001 by unpaired student’s t-test using Prism 9.0 (Graphpad). (F) Wnt targets PDK4 and phosphorylation of pyruvate dehydrogenase (PDH) at Ser293 are increased in HBP-/- mice. Quantitative RT-PCR of PDK4 gene expression (left) and representative western blot with quantified data (right) of pPDHS293, total PDH, and actin (loading control) in hippocampi of HBP1-/- Knockout (KO), HBP1+/- heterozygous (HET), and control (WT) (n = 4 for each group). Data are represented as mean ± SEM. *P <0.05 unpaired student’s t-test using Prism 9.0 (Graphpad). (G) Phosphorylated PDHS293 localized in neuronal cells of HBP-/- mice. Confocal images (x63) and percentage of positive pPDHS293 (green) neuronal cells are higher in the DG region of KO when compared to WT mice. [GFAP (red) and Dapi (blue)]. Data are represented as mean ± SEM. *P <0.05, **P < 0.01 by unpaired student’s t-test using Prism 9.0 (Graphpad). (H) Elevation of Warburg metabolic enzyme isoforms in HBP-/- mice. Quantitative and representative western blots of PKM2 protein levels were increased in KO group. (n = 3–5 mice for each group). Data are represented as mean ± SEM. *P <0.05 by unpaired student’s t-test. (I) Localization of Warburg metabolic enzyme isoform (PKM2) in HBP-/- mice. Confocal images (x63) and percentage of positive PKM2 (red) astrocytes (see supplemental methods) are higher in the DG region of 5 days post-SE with pilocarpine hippocampus when compared to control mice. [GFAP (green) and Dapi (blue)]. Data are represented as mean ± SEM. ****P <0.0001 by unpaired student’s t-test using Prism 9.0 (Graphpad). (J) Decreased survival after seizure induction in HBP1-/-KO mice. Survival curves were made for the first 5 days after seizure induction. (n = 5 mice for each group). Data are represented as mean ± SEM. *P <0.05 by unpaired student’s t-test using Prism 9.0 (Graphpad). (K) Increased seizure sensitivity in HBP1-/- KO mice for FVBN strain. Results based on the number and severity of seizures on a 5-point modified Racine scale [73]. (n = 11–17 mice for each group). Data are represented as mean ± SEM. *P <0.05, **P < 0.01 by unpaired student’s t-test using Prism 9.0 (Graphpad). Refer to supplemental materials for more details on the number of replicates and statistical analyses for each experiment.
Fig 8.
A summary scheme for potential drugs to target the Wnt-AMPK-mTOR pathway during epileptogenesis.
Our epileptogenic mechanism model shows three target points for potential anti-epileptogenic drugs. First, a Wnt inhibitor may target the upstream of this pathway, but current treatments do not cross the BBB. Secondly, altered metabolism can be targeted by ketogenic diets and Stiripentol. Thirdly, mTOR activation can be targeted with mTOR inhibitors (e.g., Sirolimus and rapamycin). Collectively, the ideal therapy is a combination and nontoxic therapy aimed at multiple targets to alter remodeling and neuronal differentiation during the epileptogenic period to forestall the progression of chronic seizures.