Fig 2: MiDAC controls a neurodevelopmental gene expression program.(A) Subunits with domains of the human histone deacetylase complex MiDAC. (B) WB for the indicated MiDAC components from total cell lysates of WT, Dnttip1 KO1 and Dnttip1 KO2 mESCs. Actin is the loading control. (C) IPs were carried out with IgG and ELMSAN1 antibodies from nuclear extracts of WT and Dnttip1 KO1 mESCs followed by WB for the indicated MiDAC components. The asterisk marks the IgG heavy chain. (D) Scatter plot comparing all DEGs in Dnttip1 KO (KO1 and KO2) versus WT mESCs from Figure 1—figure supplement 2A (x-axis) with DEGs in Elmsan1 KO (KO1 and KO2) versus WT mESCs from Figure 1—figure supplement 2B (y-axis). Both axes depict normalized gene expression (log2 FC of CPM). (E) RNA-seq heatmap depicting DNTTIP1 and ELMSAN1 co-regulated (MiDAC-regulated) genes in mESCs (fold change (FC) >1.5 or <-1.5, p<0.01). The color scale depicts normalized gene expression (log2 CPM). (F) Reactome analysis showing the most highly enriched gene categories of genes that are positively regulated by MiDAC (both down in Dnttip1 KO (KO1 and KO2) and Elmsan1 KO (KO1 and KO2) versus WT mESCs, FC <-1.5, p<0.01). Pathways associated with neural differentiation and function are highlighted in red. (G) RNA-seq heatmaps depicting down- and upregulated genes from a gene set of neurodevelopmental genes that is mutually regulated by DNTTIP1 and ELMSAN1 (FC <-1.5 or >1.5, p<0.05). The color scale depicts the z-score of normalized gene expression (log2 CPM). (D) DEGs, (E) MiDAC-regulated genes, (F) Reactome gene categories and (G) differentially expressed neurodevelopmental genes were determined based on two biological replicates each from WT mESCs and two Dnttip1 KO and two Elmsan1 KO clones, respectively.

Fig 3: MiDAC regulates neurite outgrowth.(A) Schematic outline of neuro-ectoderm (NE) differentiation protocol. (B) MAP2 immunofluorescence (IF) staining of WT, Dnttip1 KO1 and Elmsan1 KO1 NE after 12 days of differentiation. Nuclei were stained with DAPI. For analysis the neuronal cell body (blue) and its neurites were manually traced with ImageJ software. The white scale bar represents 20 µm. (C, D) qRT-PCR for (C) Tubb3 and (D) Map2 mRNA in WT, Dnttip1 KO1 and Elmsan1 KO1 mESCs (day 0) and NE after 12 and 21 days of differentiation. Expression was normalized to Gapdh. (E–G) Quantification of (E) neurite length, (F) the total number of neurites per neuron and (G) the percentage of neurons within the total cell population from traced neurites in WT, Dnttip1 KO1 and Elmsan1 KO1 neurons after 12 days of differentiation as determined by MAP2 IF. (E, F) The neurites of 200 neurons within the total cell population were assessed per genotype. (H) Gene expression levels of the indicated neurodevelopmental genes in WT, Dnttip1 KO1 and Elmsan1 KO1 NE after 12 days of differentiation as analyzed by qRT-PCR. Expression was normalized to Gapdh. Unpaired t-test was performed throughout where ***, p=0.001; **, p=0.01; *, p=0.05; and ns, p>0.05 is not significant.

Fig 4: MiDAC binds to and modulates the expression of genes that regulate neural differentiation and neurite outgrowth.(A) Pie charts displaying the genome-wide distribution (left) and promoter distal distribution (right) of DNTTIP1 in WT mESCs. DNTTIP1 peaks within 1 kb of the transcription start site (TSS) were assigned to TSS. (B) Venn Diagram depicting the overlap between DNTTIP1, H3K4me1 and H3K27ac peaks in WT mESCs. (C) Venn diagram showing the number of downregulated and upregulated genes in Dnttip1 KO (KO1 and KO2) versus WT mESCs that are bound by DNTTIP1 in WT mESCs. (D) Heatmaps displaying the genome-wide distribution of all DNTTIP1 binding sites in WT, Dnttip1 KO1 and Elmsan1 KO1 mESCs sorted by enrichment in descending order in WT mESCs and compared to HDAC1 occupancy in WT mESCs. The color scale depicts the normalized ChIP-seq signal intensity (log2 CPM per 20 bp bin). (E) Venn diagram showing the co-occupancy between DNTTIP1 and HDAC1 peaks in WT mESCs. (A–E) DNTTIP1, H3K4me1 and H3K27ac ChIP-seq peaks were determined based on the average of two replicates each from WT mESCs and where applicable based on the average of two replicates each from Dnttip1 KO1 and Elmsan1 KO1 mESCs, respectively. HDAC1 ChIP-seq data were obtained from GSE55437. Downregulated and upregulated genes were determined based on two biological replicates each from WT mESCs and two Dnttip1 KO (KO1 and KO2) clones. (F, G) ChIP-seq profiles of the (F) Slit3 and (G) Spry4 loci for DNTTIP1 in WT, Dnttip1 KO1 and Elmsan1 KO1 mESCs and for HDAC1 in WT mESCs. Promoter and putative enhancer regions used for manual ChIP experiments in Figure 4 are highlighted by orange boxes. The track files depict the average of two ChIP-seq replicates each from WT mESCs and the average of two replicates each from the Dnttip1 KO1 and Elmsan1 KO1 clone, respectively. (H) WB for the specified histone acetylation marks from total cell lysates of WT, Dnttip1 KO1 and Elmsan1 KO1 mESCs. H3 and H4 are loading controls.

Fig 5: MiDAC directly targets positive and negative regulators of neurite outgrowth during neural differentiation.(A–J) qPCR from manual ChIP experiments against (A, F) DNTTIP1, (B, G) ELMSAN1, (C, H) HDAC1, (D, I) H3K27ac and (E, J) H4K20ac from WT, Dnttip1 KO1 and Elmsan1 KO1 NE targeting select promoter, putative enhancer and intragenic control regions of (A–E) Slit3 or (F-J) Spry4 loci as highlighted in Figure 3F and G. IgG was used as a control antibody. Unpaired t-test was performed throughout where **, p=0.01; and ns, p>0.05 is not significant.