Comprehensive characterization of the neurogenic and neuroprotective action of a novel TrkB agonist using mouse and human stem cell models of Alzheimer's disease

Abstract

Background: Neural stem cell (NSC) proliferation and differentiation in the mammalian brain decreases to minimal levels postnatally. Nevertheless, neurogenic niches persist in the adult cortex and hippocampus in rodents, primates and humans, with adult NSC differentiation sharing key regulatory mechanisms with development. Adult neurogenesis impairments have been linked to Alzheimer's disease (AD) pathology. Addressing these impairments by using neurotrophic factors is a promising new avenue for therapeutic intervention based on neurogenesis. However, this possibility has been hindered by technical difficulties of using in-vivo models to conduct screens, including working with scarce NSCs in the adult brain and differences between human and mouse models or ethical limitations.

Methods: Here, we use a combination of mouse and human stem cell models for comprehensive in-vitro characterization of a novel neurogenic compound, focusing on the brain-derived neurotrophic factor (BDNF) pathway. The ability of ENT-A011, a steroidal dehydroepiandrosterone derivative, to activate the tyrosine receptor kinase B (TrkB) receptor was tested through western blotting in NIH-3T3 cells and its neurogenic and neuroprotective action were assessed through proliferation, cell death and Amyloid-β (Aβ) toxicity assays in mouse primary adult hippocampal NSCs, mouse embryonic cortical NSCs and neural progenitor cells (NPCs) differentiated from three human induced pluripotent stem cell lines from healthy and AD donors. RNA-seq profiling was used to assess if the compound acts through the same gene network as BDNF in human NPCs.

Results: ENT-A011 was able to increase proliferation of mouse primary adult hippocampal NSCs and embryonic cortical NSCs, in the absence of EGF/FGF, while reducing Aβ-induced cell death, acting selectively through TrkB activation. The compound was able to increase astrocytic gene markers involved in NSC maintenance, protect hippocampal neurons from Αβ toxicity and prevent synapse loss after Aβ treatment. ENT-A011 successfully induces proliferation and prevents cell death after Aβ toxicity in human NPCs, acting through a core gene network shared with BDNF as shown through RNA-seq.

Conclusions: Our work characterizes a novel BDNF mimetic with preferable pharmacological properties and neurogenic and neuroprotective actions in Alzheimer's disease via stem cell-based screening, demonstrating the promise of stem cell systems for short-listing competitive candidates for further testing.

Keywords: Alzheimer’s disease; BDNF; In-vitro drug testing; Neurogenesis; Stem-cell based screening; TrkB; iPSC.

Fig. 1

Screening of ENT-A011. ENT-A011 induces TrkB and its downstream target Akt phosphorylation after 20 min treatment in NIH-3T3 TrkB stable expressed cells. Representative blots (A) and quantification A’ of 8 (TrkB) and 6 (Akt) independent experiments are shown. Error bars represent S.E.M., Student’s t-test against Control; *p < 0.05; ***p < 0.001. B, B′ ENT-A011 protects NIH-3T3 TrkB cells from cell death caused by serum deprivation. Representative images (B) and quantification of Toxicity Assay (B′) on NIH-3T3 TrkB cells after treatment with BDNF or compound ENT-A011 for 24 h with or without TrkB inhibitor, ANA-12. (without ANA-12: BDNF N = 10/ENT-A011 N = 10, with ANA-12: BDNF N = 7/ENT-A011 N = 9), error bars represent S.E.M., Student’s t-test against Control; *p < 0.05, **p < 0.01, ***p < 0.001. C, C′ Nomad biosensor activation via TrkB phosphorylation in HEK 293 TrkB transiently transfected cells after ENT-A011 treatment. Receptor activation leads to increase in fluorescence levels. Representative images (C) and dose response curve for different compound concentrations (C′) are shown. Scalebars = 100 μm

Fig. 2

Effect of ENT-A011 on mouse astrocytes. The compound induces TrkB and its downstream target Akt phosphorylation after 20 min treatment in astrocytes. Representative blots (A) and quantification (A′) of 4 independent experiments are shown. Error bars represent S.E.M., Student’s t-test against Control; *p < 0.05; ***p < 0.001. B ENT-A011 increases Bdnf, Wnt, Creb and TrkB expression (detected by RT-qPCR) after LPS treatment in astrocytes. N = 3–4, error bars represent S.E.M., Student’s t-test against Control; *p < 0.05, **p < 0.01

Fig. 3

ENT-A011 induces neurogenesis on mouse adult hippocampal NSCs. A, A′ ENT-A011 increases proliferation in adult hippocampal NSCs. Representative fluorescence microscopy images of Immunostaining for BRDU and Nestin in adult hippocampal NSCs treated with BDNF or ENT-A011 for 24 h with or without Aβ42 (A). Quantification of BRDU fluorescence change with or without Aβ42 (A′). Quantification represents BRDU positive cells normalized against Nestin positive cells. (without Aβ42: N = 7–8, with Aβ42: N = 3), error bars represent S.E.M., Student’s t-test against Control; *p < 0.05, **p < 0.01, ***p < 0.001. B, B′. The compound rescues adult hippocampal NSCs from Aβ-induced cell death. Representative fluorescence microscopy images of Celltox cytotoxicity assay on hippocampal NSCs in the presence of Aβ42, treated with BDNF or ENT-A011, with or without the TrkB inhibitor ANA-12 for 24 h (B). Quantification of fluorescence change and MTT levels after BDNF or compound treatment, with or without ANA-12 N = 3–5. (B′). Quantification represents Celltox positive cells normalized against Hoechst positive cells (without ANA-12: N = 5–8, with ANA-12: N = 3)., error bars represent S.E.M., Student’s t-test against Control; *p < 0.05, **p < 0.01, ***p < 0.001. C, C′ ENT-A011 promotes differentiation of adult hippocampal NSCs. Representative fluorescence microscopy images of Immunostaining for Tuj1 and GFAP in adult hippocampal neural stem cells treated with BDNF or ENT-A011 for 10 days (C). Fold change in expression of TUJ1 (N = 6 for untreated and BDNF, N = 5 for ENT-A011), MAP2 (N = 5) and GFAP (N = 4) assessed by RT-qPCR in BDNF or ENT-A011 treated adult hippocampal NSCs relative to untreated control cells. (C′) Scalebars = 20 μm

Fig. 4

ENT-A011 has positive effect on mouse cortical NSCs. A, A′ ENT-A011 increases proliferation on cortical NSCs, Representative fluorescence microscopy images of Immunostaining for BRDU and Nestin in adult hippocampal NSCs treated with BDNF or ENT-A011 for 24 h with or without Aβ42 (A). Quantification of BRDU fluorescence change with or without Aβ42 (A′). Quantification represents BRDU positive cells normalized against Nestin positive cells. (without Aβ42: N = 6, with Aβ42: N = 3), error bars represent S.E.M., Student’s t-test against Control; *p < 0.05, **p < 0.01, ***p < 0.001. B, B′ The compound rescues cortical NSCs from Aβ-induced cell death. Representative fluorescence microscopy images of Celltox cytotoxicity assay on hippocampal NSCs in the presence of Aβ42, treated with BDNF or ENT-A011, with or without the TrkB inhibitor ANA-12 for 24 h (B′ right). Quantification of fluorescence change and MTT levels after BDNF or compound treatment, with or without ANA-12 (without ANA-12: N = 3–5, with ANA-12: N = 5) (B′ left). Quantification represents Celltox positive cells normalized against Hoechst positive cells. (without ANA-12: N = 5–7, with ANA-12: N = 3–5), error bars represent S.E.M., Student’s t-test against Control; *p < 0.05, **p < 0.01, ***p < 0.001. Scalebars = 20 μm

Fig. 5

ENT-A011 protects mouse hippocampal neurons and rat cortical neurons against Aβ42 and glutamate toxicity respectively. A, A′, ENT-A011 reduces Aβ42 induced hippocampal neuron cell death. Primary hippocampal neurons were treated with Aβ42 for 48 h in presence of ENT-A011 and cell death was quantified using TUNEL assay. Representative images from 4 independent experiments. Data are shown as SEM. One-way ANOVA, Tukey’s Test correction: **p < 0.01; ***p < 0.001. B, B′ ENT-A011 protects against Aβ42-induced hippocampal synapse loss. Primary hippocampal neurons were treated with Aβ42 for 4 h in presence of ENT-A011, followed by staining against synaptophysin to assess synapse number. Total Synaptophysin area was normalized to total Tuj1 area in each image. Representative images from 4 independent experiments. Data are shown as SEM. One-way ANOVA, Tukey’s Test correction: ** p < 0.01. ENT-A011 reverts the decrease of neurite branches caused by glutamate on rat Cortical neurons. Representative fluorescence microscopy images of Immunostaining for Tuj1 in cortical neurons treated with BDNF or ENT-A011 for 1 h before glutamate treatment (C). Quantification of neurite branches (C′). N = 3, error bars represent S.E.M., Student’s t-test against Control; *p < 0.05, **p < 0.01, ***p < 0.001. White Scalebars = 20 μm, Yellow Scalebars = 5 μm

Fig. 6

ENT-A011 effect on human cortical Neural Progenitor Cells (NPCs). Change in BRDU + cells (calculated as percentage of total Hoechst + cells) after BDNF or ENT-A011 treatment for 48 h in NPC derived from 3 human induced pluripotent stem cell (iPSC) lines. Representative fluorescence microscopy images of Immunostaining for BRDU and Hoechst (A) and quantification of BRDU fluorescence change (A′). B, B′ NPC derived from 3 human iPSC lines were treated with Aβ_1-42 oligomers and BDNF or ENT-A011 for 48 h and compound effect on reducing toxicity was assessed via the Celltox cytotoxicity assay. Representative fluorescence microscopy images of Celltox cytotoxicity assay in the presence of Aβ42, treated with BDNF or ENT-A011 for 24 h (B). Quantification of fluorescence change and MTT levels after BDNF or compound treatment, with or without ANA-12 (B′). Quantification represents Celltox positive cells normalized against Hoechst positive cells. Three different lines were differentiated, N = 3–4, error bars represent S.E.M., Student’s t-test against Control; *p < 0.05, **p < 0.01, ***p < 0.001. Scalebars = 20 μm

Fig. 7

Differential expression analysis of ENT-A011 and BDNF treated APOE4 human Cortical Neural Progenitor Cells (NSCs) compared to untreated control cells. a Volcano plot of differentially regulated genes after BDNF treatment compared to untreated control cells. Negative log10 adjusted p values (y axis) are plotted against log2 fold change in expression for each gene. Genes with Padj lower or equal to 0.1 (blue horizontal line) are highlighted in bold blue colour and genes that are also differentially regulated by ENT-A011 (p adj ≤ 0.1) are highlighted with red circles. b Volcano plot of differentially regulated genes after ENT-A011 treatment compared to untreated control cells. Negative log10 adjusted p values (y axis) are plotted against log2 fold change in expression for each gene. Genes with Padj lower or equal to 0.1 (blue horizontal line) are highlighted in bold blue colour and genes that are also differentially regulated by ENT-A011 (p adj ≤ 0.1) are highlighted with red circles. c Heatmap of log2 Fold Change in expression for genes differentially regulated by both BDNF and ENT-A011, ranked by decreasing log2FC values in BDNF treated cells. d Scatter plots of log2 fold change in gene expression after BDNF (x axis) or ENT-A011 (y axis) treatment. Spearman correlation coefficient (0.9) and p value (2.2e⁻¹⁶) are displayed next to line of best fit. e Heatmap of negative log10 p values for gene ontology enrichment analysis of upregulated genes (p adj ≤ 0.1) after BDNF (left column) or ENT-A011 treatment. Unique terms to either treatment group are plotted in pale orange in the other group