Amyloid Precursor Protein (APP) Regulates Gliogenesis and Neurogenesis of Human Neural Stem Cells by Several Signaling Pathways

Abstract

Numerous studies have focused on the pathophysiological role of amyloid precursor protein (APP) because the proteolytic processing of APP to β-amyloid (Aβ) peptide is a central event in Alzheimer's disease (AD). However, many authors consider that alterations in the physiological functions of APP are likely to play a key role in AD. Previous studies in our laboratory revealed that APP plays an important role in the differentiation of human neural stem cells (hNSCs), favoring glial differentiation (gliogenesis) and preventing their differentiation toward a neuronal phenotype (neurogenesis). In the present study, we have evaluated the effects of APP overexpression in hNSCs at a global gene level by a transcriptomic analysis using the massive RNA sequencing (RNA-seq) technology. Specifically, we have focused on differentially expressed genes that are related to neuronal and glial differentiation processes, as well as on groups of differentially expressed genes associated with different signaling pathways, in order to find a possible interaction between them and APP. Our data indicate a differential expression in genes related to Notch, Wnt, PI3K-AKT, and JAK-STAT signaling, among others. Knowledge of APP biological functions, as well as the possible signaling pathways that could be related to this protein, are essential to advance our understanding of AD.

Keywords: RNA sequencing; amyloid precursor protein; gliogenesis; neural stem cells; neurogenesis; signaling pathways.

Figure 1

Confirmation of APP overexpression in hNS1 cells after transient nucleofection. (A) Schematic representation of the protocol used for proliferation (division) and differentiation (4 days after nucleofection) of hNS1 cells. (B) Representative images in phase contrast and the corresponding field with GFP+ cells in control-hNS1 cells and APP-hNS1 cells on different days after transfection (division and day 4 of differentiation). Scale bar = 50 μm. (C) Percentage of GFP+ cells with respect to total cells for both experimental groups (control and APP) in cell division. (D) Relative expression levels of APP695 mRNA by quantitative real-time PCR (RT-qPCR) in APP-hNS1 cells and control-hNS1 cells on different days after transfection (division and day 4 of differentiation). (E) Western blot (WB) analysis of APP (using 22C11 antibody) in cell extracts of APP-hNS1 cells and control-hNS1 cells on different days after transfection (division and day 4 of differentiation). Actin was used as a loading control. (F) Relative protein levels of APP by densitometry analysis in both experimental groups (control and APP) on different days after transfection (division and day 4 of differentiation). Data represent mean ± SD (n = 3 for three independent samples). Statistical analysis was performed using a t-test between APP and control groups; * p < 0.05; *** p < 0.001.

Figure 2

Representation of Differentially Expressed Genes (DEGs) found in APP-hNS1 cells versus control-hNS1 cells. (A) Heatmap showing differential RNA-seq counts of the genes with FDR ≤ 0.05 and sorted by logFold Change (logFC). There were three replicates of control samples (derived from control-hNS1 cells) and three replicates of APP samples (derived from APP-hNS1 cells). This heatmap was built using DESeq 2 on normalized gene read counts. (B) Heatmap of the three samples for each condition (APP-hNS1 cells and control-hNS1 cells) and gene expression levels of selected genes using average linkage clustering with Pearson correlation as the default distance metric. (C) Volcano plot from RNA-seq data of APP-hNS1 cells and control-hNS1 cells. The log2FC is plotted on the x-axis, and the negative log10(FDR) is plotted on the y-axis. Some relevant genes increased in APP-hNS1 cells (red dots) are indicated inside a box. (D) Replot of the volcano excluding the highest negative log10(FDR) genes. Red dots show the distribution of increased genes, and blue dots show the distribution of decreased genes in APP-hNS1 cells versus control-hNS1 cells. Some genes related to gliogenesis, neurogenesis, and different pathways mentioned in this study are indicated inside a box.

Figure 3

Functional enrichment of the genes differentially expressed (DEGs) in RNA-seq assays due to APP overexpression. (A) Gene Ontology (GO) enrichment analysis for the DEGs altered by APP overexpression versus control in hNS1 cells, obtained with Enrichr. Representation of the GO biological function terms is indicated with the bar length by the number of genes and the p-value by the color blue gradient. (B) Diagram with genes of RNA-seq assays dysregulated due to APP overexpression that are associated with gliogenesis and neurogenesis. Represented genes are based in terms of the GO, CellMarker, and Panglao databases. (C) Chord plot representing via colored ribbons the expression of genes associated with gliogenesis and neurogenesis. Genes are ordered according to log2FC in APP-hNS1 cells versus control-hNS1 cells, which are displayed in the intensity of red and blue squares next to the selected genes.

Figure 4

Signaling pathways enrichment of the genes differentially expressed (DEGs) in RNA-seq assays due to APP overexpression. (A) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways enrichment for all the DEGs altered by APP overexpression versus control in hNS1 cells, obtained with Enrichr. Representation of the signaling pathways are indicated with the bar length by the number of genes and the p-value by the color blue gradient. (B) Dotplot showing the results of KEGG pathways enrichment for the genes upregulated in APP versus control hNS1 cells. The y-axis represents the ratio between the number of genes participating in the analysis and the total annotated participants of the KEGG pathway. The dot color indicates the p-value, and the dot size indicates the number of genes. (C) Results of the RNA-seq analysis of APP-hNS1 cells versus control-hNS1 cells, representing the expression of genes as log2FC, included in KEGG signaling pathways of Notch signaling (i), Wnt signaling (ii), PI3K-AKT signaling (iii), MAPK signaling (iv), JAK-STAT signaling (v), and BMP signaling (vi). Significant differentially expressed genes are marked with an asterisk (*).

Figure 5

Validation of RNA-seq results on gliogenesis and neurogenesis processes in APP overexpression hNS1 cells. (A) STRING network of protein–protein interactions between APP, neurogenesis-related genes, and gliogenesis-related genes according to our transcriptomic analysis. (B) STRING network of protein–protein interactions between APP and signaling pathways-related genes according to our transcriptomic analysis. Thicker network edges show greater evidence of a biological relationship between the nodes (genes). (C) Representative images at day 4 of differentiation by immunocytochemistry (ICC), showing immunoreactivity for GFAP (red) and β-III-tubulin (BIIITub; red) in control-hNS1 cells and APP-hNS1 cells. Scale bar = 50 μm. (D) Percentage of positively stained cells for GFAP and β-III-tubulin markers relative to the total cells (Hoechst; blue) at day 4 of differentiation after nucleofection. (E) Relative expression levels of GFAP, S100B, and TUBB3 mRNA by quantitative real-time PCR (RT-qPCR) in APP-hNS1 cells and control-hNS1 cells at day 4 of differentiation after nucleofection. Data represent mean ± SD (n = 3 for three independent samples). Statistical analysis was performed using T-test between APP and control groups; * p < 0.05; ** p < 0.01.

Figure 6

Schematic illustration of the suggested effects of APP overexpression on hNSCs. The positive and negative effects of APP on neurogenesis could be mediated, at a global level, by Notch signaling, Wnt signaling, and/or PI3K-AKT signaling. The positive effects of APP on gliogenesis could be mediated, at a global level, by Notch signaling, JAK-STAT signaling, and/or BMP signaling.