ADNP is essential for sex-dependent hippocampal neurogenesis, through male unfolded protein response and female mitochondrial gene regulation

Abstract

Essential for brain formation and protective against tauopathy, activity-dependent neuroprotective protein (ADNP) is critical for neurogenesis and cognitive functions, while regulating steroid hormone biogenesis. As such, de novo mutations in ADNP lead to syndromic autism and somatic ADNP mutations parallel Alzheimer's disease progression. Furthermore, clinical trials with the ADNP fragment NAP (the investigational drug davunetide) showed efficacy in women suffering from the tauopathy progressive supranuclear palsy and differentially boosted memory in men (spatial) and women (verbal), exhibiting prodromal Alzheimer's disease. While autism is more prevalent in boys and Alzheimer's disease in women, both involve impaired neurogenesis. Here, we asked whether ADNP sex-dependently regulates neurogenesis. Using bromodeoxyuridine (BrdU) as a marker of neurogenesis, we identified two-fold higher labeling in the hippocampal sub-ventricular zone of ADNP-intact male versus female mice. Adnp haplo-insufficient (Adnp^+/-) mice or mice CRSIPR/Cas9-edited to present the most prevalent neurodevelopmental ADNP syndrome mutation, p.Tyr718* (Tyr) showed dramatic reductions in male BrdU incorporation, resulting in mutated females presenting higher labeling than males. Treatment with NAP compensated for the male reduction of BrdU labeling. Mechanistically, hippocampal RNAseq revealed male-specific Tyr down-regulation of endoplasmic reticulum unfolded protein response genes critical for sex-dependent organogenesis. Newly discovered mitochondrial accessibility of ADNP was inhibited by the Tyr718* mutation further revealing female-specific Tyr downregulation of mitochondrial ATP6. NAP moderated much of the differential expression caused by p.Tyr718*, accompanied by the down-regulation of neurotoxic, pro-inflammatory and pro-apoptotic genes. Thus, ADNP is a key regulator of sex-dependent neurogenesis that acts by controlling canonical pathways, with NAP compensating for fundamental ADNP deficiencies, striding toward clinical development targeting the ADNP syndrome and related neurodevelopmental/neurodegenerative diseases.

Fig. 1. Comparison of BrdU-labeled cell concentrations in the SVZ of two ADNP mouse models, Adnp^+/− on ICR background and Tyr mice on C57BL6/NJ background.

A–C Adnp^+/− mice, (D–F) Tyr mice, (G) wild type, ICR and C57BL6/NJ comparisons. (A, B, D, E) Representative images (scale bar = 100 µm). C, F, G Group differences in BrdU-positive cells/mm2 (mean ± SEM) were compared using a one-way analysis of variance with Tukey’s post-hoc test. Technical replicates (six per animal) were used for statistical analysis. Outliers were excluded based on Grubb’s test. C For Adnp^+/− males, a statistically significant difference for BrdU-positive cells was discovered (F_2,77 = 33.030, p < 0.001), with Tukey’s post-hoc test revealing a significant reduction in BrdU-positive cells in Adnp^+/− DD-treated males (N = 4), as compared to WT (N = 5) (***P < 0.001), which was significantly corrected upon NAP treatment (***P < 0.001, N = 5). No such effect was found in females (N = 5 per group). Significant sex differences were discovered in the WT and the NAP-treated Adnp^+/− group (***P < 0.001 for both comparisons). F For Tyr females, a statistically significant difference for BrdU-positive cells was discovered (F_2,77 = 11.272, ***P < 0.001), with Tukey’s post-hoc test revealing a significant reduction in BrdU-positive cells in NAP-treated Tyr females (N = 5), as compared to WT (N = 4) (***P < 0.001), with no difference being seen when compared to DD-treated Tyr females (N = 5). For Tyr males, a statistically significant difference for BrdU-positive cells was discovered (F_2,89 = 12.777, ***P < 0.001), with Tukey’s post-hoc test revealing a significant reduction in BrdU-positive cells in DD- and NAP-treated Tyr males (N = 5 for both groups), as compared to WT (N = 5) (***P < 0.001). Significant sex differences were discovered in the WT and the DD-treated Tyr groups (P < 0.05 for both comparisons). G Significant differences were discovered among males of the different tested strains (P < 0.01).

Fig. 2. Heterozygous Adnp expression in Tyr mice.

A Bar plot of hippocampal Adnp expression in WT and Tyr mice, treated by vehicle or NAP as before [22] by allele (read count) and (B) boxplot of total Adnp gene expression (normalized expression). Note that allelic expression is based solely on reads overlapping the Tyr718* mutation locus, while overall gene expression is based on all reads overlapping the gene, regardless of genotype.

Fig. 3. Adnp-specific regulated expression in Tyr males.

A Heatmap of differential expression colored according to log2-fold change for each gene (row) at each comparison (column). In cases of significant differential expression (FDR < 0.05), the cell is marked with a star. If differential expression was only significant at the transcript level, the gene name is marked with a circled suffix. B A graph of relationships between differentially expressed genes (small nodes) and terms with which they are associated. Gene nodes are marked red (up-regulation) or blue (down-regulation) according to their significant differential expression trend. C GSEA plot of genes up-regulated by the HSP-90 inhibitor geldanamycin according to their differential expression in the Tyr male comparison. The position of the enriched genes is marked in a ranked list at the bottom, starting from the most downregulated (blue) to the most up-regulated (red).

Fig. 4. Adnp-specific regulated expression in Tyr females.

A Heatmap of differential expression colored according to log2-fold change for each gene (row) at each comparison (column). In cases of significant differential expression (FDR < 0.05), the cell is marked with a star. If differential expression was only significant at the transcript level, the gene name is marked with a circled suffix. B A graph of relationships between differentially expressed genes (small nodes) and terms with which they are associated. Gene nodes are marked red (up-regulation) or blue (down-regulation) according to their significant differential expression trend. C1 Mouse neuroblastoma N1E-115 cell clones expressing CRISPR/Cas9-edited full-length ADNP or heterozygous ADNP p.Tyr718* [25, 29] were endogenously stained with MitoTracker (red)-. Co-localization of ADNP and MitoTracker is represented by white dots. Quantitative analysis of ADNP/MitoTracker merged staining, reflecting co-localization, is presented in the graph; images were viewed using a x63 oil immersion lens. C2 Statistical analysis of the co-localization rate calculated by addressing merged staining in a Leica sp8 fluorescent microscope. A two-tailed t-test confidence level of 95% was determined using PRISM Statistics software, version 24 (IBM, Armonk, NY), *P < 0.05. D ADNP contains a mitochondrial targeting sequence [58, 59] ADNP is represented in dark green, while the internal mitochondrial targeting sequence is colored red (D1). The sequence is shortened in ADNP p.Tyr719*, with further structural differences as previously highlighted [8] (D2). ADNP and ADNP p.Tyr719* structures were retrieved using the I-TASSER server. The figures were created using PyMOL software as before [8, 14].

Fig. 5. Differential expression across multiple RNA-seq experiments revealing variations according to Adnp mutation and mouse model.

A The number of genes with significant differential expression across comparisons, either at the whole gene (DGE) or the single transcript level (DTE). B Upset plot of significant differentially expressed genes (FDR < 0.05), where each column is representative of a subset of genes from one or more comparisons (represented by rows). The size of each gene subset is visualized in the bar-plot in the top panel, with both subsets (columns) and comparisons (rows) sorted from the most to the fewest genes. C Heatmap of genes significantly differentially expressed (FDR < 0.05) in five or more comparisons. Significant results (FDR < 0.05) are marked with a star, while near-significant results (0.05 < FDR < 0.1) are marked with a dot.

Fig. 6. ADNP is essential for sex-dependent hippocampal neurogenesis, through male unfolded protein response and female mitochondrial gene regulation, schematic representation.

The scheme shows the discovery of ADNP’s association with increased neurogenesis in males that is reduced in the face of ADNP deficiencies and corrected in part by NAP (davunetide) treatment. It further depicts the discovery of the differential ADNP/NAP regulation of the unfolded protein response in males versus the regulation of the essential MT-ATP6 in females, corrected by NAP (davunetide) treatment. Lastly, the newly revealed involvement of protocadherin in ADNP function is highlighted.