Impaired SNF2L Chromatin Remodeling Prolongs Accessibility at Promoters Enriched for Fos/Jun Binding Sites and Delays Granule Neuron Differentiation

Abstract

Chromatin remodeling proteins utilize the energy from ATP hydrolysis to mobilize nucleosomes often creating accessibility for transcription factors within gene regulatory elements. Aberrant chromatin remodeling has diverse effects on neuroprogenitor homeostasis altering progenitor competence, proliferation, survival, or cell fate. Previous work has shown that inactivation of the ISWI genes, Smarca5 (encoding Snf2h) and Smarca1 (encoding Snf2l) have dramatic effects on brain development. Smarca5 conditional knockout mice have reduced progenitor expansion and severe forebrain hypoplasia, with a similar effect on the postnatal growth of the cerebellum. In contrast, Smarca1 mutants exhibited enlarged forebrains with delayed progenitor differentiation and increased neuronal output. Here, we utilized cerebellar granule neuron precursor (GNP) cultures from Smarca1 mutant mice (Ex6DEL) to explore the requirement for Snf2l on progenitor homeostasis. The Ex6DEL GNPs showed delayed differentiation upon plating that was not attributed to changes in the Sonic Hedgehog pathway but was associated with overexpression of numerous positive effectors of proliferation, including targets of Wnt activation. Transcriptome analysis identified increased expression of Fosb and Fosl2 while ATACseq experiments identified a large increase in chromatin accessibility at promoters many enriched for Fos/Jun binding sites. Nonetheless, the elevated proliferation index was transient and the Ex6DEL cultures initiated differentiation with a high concordance in gene expression changes to the wild type cultures. Genes specific to Ex6DEL differentiation were associated with an increased activation of the ERK signaling pathway. Taken together, this data provides the first indication of how Smarca1 mutations alter progenitor cell homeostasis and contribute to changes in brain size.

Keywords: ATAC-seq; ISWI chromatin remodeler; SMARCA1; Snf2L; cerebellar granule neuron progenitors; chromatin accessibility.

FIGURE 1

Generation of GNP cultures from the Ex6DEL mice. (A) Schematic diagram of the procedure for generating GNP and granule neuron cultures. (B) Representative images of WT cultures 1 DIV (top image) and 3 DIV (bottom image) stained with neurofilament-200 (red). Scale bar, 25 μM. (C) Protein extracts of primary GNP cultures at I and 3 DIV were immunoblotted for the ISWI proteins (Snf2h, Snf2l), neuronal (Tuj, NeuN) and proliferation (Ki67) markers. Vinculin was used as a loading control. (D) Immunoblot (top panel), Smarca1 genotyping (middle panel), and RNAseq analysis (bottom panel) from WT and Ex6DEL GNP cultures confirmed the loss of exon 6 in the GNP cultures isolated from Ex6DEL mice. The red rectangle outlines the absence of RNAseq reads corresponding to the position of exon 6. (E) Phase contrast images of GNP cultures from WT and Ex6DEL mice. Scale bar, 50 μM.

FIGURE 2

Ex6DEL cultures show delayed differentiation. (A) Graph depicting the fraction of BrdU + cells in WT (white bars) and Ex6DEL (gray bars) GNP cultures at the specified times. Cultures were pulsed with BrdU for 2 h prior to harvesting cells for analysis. (n = 4 independent experiments; five images per coverslip used for quantification). ***p < 0.001. (B) P10 cerebellar sections from WT and Ex6DEL mice stained for committed granule neurons (Pax6, magenta), Purkinje neurons (Calbindin, green) or all cell nuclei (DAPI, blue). Arrowheads indicate the migrating postmitotic granule neurons. EGL, external granule layer; ML, molecular layer; PCL, Purkinje cell layer; IGL, inner granule layer. Scale bar, 50 μM. (C) Plot of the percentage of Pax6⁺ cells within the ML. ^∗p < 0.05. (D) Immunoblot analysis of cerebellar extracts from WT and Ex6DEL mice co-immunoprecipitated for Snf2l, Cecr2, or a control rabbit antibody (Rb IgG).

FIGURE 3

RNAseq analysis of WT and Ex6DEL cultures. (A) Schematic diagram showing the numbers of differentially expressed genes (DEGs) between different samples (WT and Ex6DEL) and days in culture. (B) GO term analysis of the DEGs observed at 1 DIV. (C) DEG lists for some of the top GO terms. (D) RT-qPCR validation of Fos/Jun genes from RNA isolated from P10 cerebella. ^∗p < 0.05.

FIGURE 4

ATACseq analysis of WT and Ex6DEL cultures. (A) Compilation of all ATACseq peaks showing enrichment at the TSS under all culture conditions. Mutant cultures at 1 DIV (Ex6DEL GNP) showed the greatest enrichment of peaks at the TSS, followed by Ex6DEL GN (3 DIV), WT GNPs and WT GNs. (B) Peaks aligned to a normalized genes showed that Ex6DEL chromatin (light blue) was more accessible than WT chromatin (dark blue) and more similar to the profile of ESCs (yellow). TSS, transcriptional start site; TES, transcriptional end site. (C) Schematic diagram showing the numbers of differentially accessible regions (DARs) between WT and Ex6DEL samples at 1 and 3 days in culture. (D) MA plot of the 3231 differentially accessible regions. Pink dots represent significant changes in accessibility, with dots above the line representing increased accessibility and below the line decreased accessible regions. (E) Plot showing the frequency of DARs at different genomic positions. Promoter/TSS, −1KB to +100 bp; 5′UTR, >100 bp from TSS; TTS, transcriptional termination sites.

FIGURE 5

Accessible regions are enriched for Fos/Jun binding sites. ATACseq peak alignment with binding sites for CTCF (A) or Fos (B) showed enrichment in the Ex6DEL samples for Fos binding sites but not CTCF. (C) ATACseq peaks at 1 DIV were assigned to the nearest gene and then cross referenced to the up- and down-regulated genes to generate the bar graphs. Open (green) closed (pink) or chromatin regions showing no change (blue) are shown. (D,E) ATACseq and RNAseq reads shown in IGV browser format for the Arc (D) and Fosb (E) genes. The GNP ATACseq tracks are shown in the blue box. The GNP RNAseq is shown in the black box. The unboxed tracks correspond to ATACseq and RNAseq from GNs.

FIGURE 6

Ex6DEL cultures are delayed in differentiation. (A) Schematic diagram showing the numbers of differentially expressed genes (DEGs) between different samples (WT and Ex6DEL) and days in culture (same as Figure 3A). (B) Venn diagram showing overlapping differentially expressed genes in WT or Ex6DEL cultures during differentiation (gene expression comparison between I and 3 DIV). (C) Venn diagram highlighting that only 1/3 of the uniquely expressed genes from (B) (Ex6DEL specific: 1532; WT-specific: 2228) remain differentially expressed at 3 DIV. Green circle comprises the 2967 down- and 551 up-regulated DEGs between WT and Ex6DEL cultures. (D,E) GO term analysis of the DEGs at # DIV that were specific to the Ex6DEL (D) or WT (E) cultures.

FIGURE 7

Granule neuron precursor differentiation is associated with ERK signaling activation. (A) Immunoblots from WT and Ex6DEL GNP cultures (1 and 3 DIV) for Erk and phospho-Erk (pErk). Vinculin serves as a protein loading control. (B) Quantification of the immunoblots in A showing the normalized change in Erk activation (pErk1 + pErk2 level) to total Erk protein level. ^∗p < 0.05. (C) RT-qPCR validation of the expression of downstream target genes of ERK/Fos/Jun signaling. ^∗p < 0.05. (D) Genome browser views of the Sox2 (D) and Bdnf (E) genes. The blue boxes highlight the ATACseq peaks and the black boxes mark the RNAseq reads at 3 DIV. All unboxed tracks correspond to the I DIV time.