Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Feb 9;10(6):eadk3384.
doi: 10.1126/sciadv.adk3384. Epub 2024 Feb 9.

CDCA7-associated global aberrant DNA hypomethylation translates to localized, tissue-specific transcriptional responses

Affiliations

CDCA7-associated global aberrant DNA hypomethylation translates to localized, tissue-specific transcriptional responses

Maja Vukic et al. Sci Adv. .

Abstract

Disruption of cell division cycle associated 7 (CDCA7) has been linked to aberrant DNA hypomethylation, but the impact of DNA methylation loss on transcription has not been investigated. Here, we show that CDCA7 is critical for maintaining global DNA methylation levels across multiple tissues in vivo. A pathogenic Cdca7 missense variant leads to the formation of large, aberrantly hypomethylated domains overlapping with the B genomic compartment but without affecting the deposition of H3K9 trimethylation (H3K9me3). CDCA7-associated aberrant DNA hypomethylation translated to localized, tissue-specific transcriptional dysregulation that affected large gene clusters. In the brain, we identify CDCA7 as a transcriptional repressor and epigenetic regulator of clustered protocadherin isoform choice. Increased protocadherin isoform expression frequency is accompanied by DNA methylation loss, gain of H3K4 trimethylation (H3K4me3), and increased binding of the transcriptional regulator CCCTC-binding factor (CTCF). Overall, our in vivo work identifies a key role for CDCA7 in safeguarding tissue-specific expression of gene clusters via the DNA methylation pathway.

PubMed Disclaimer

Figures

Fig. 1.
Fig. 1.. An ICF3-relevant Cdca7 missense mutation disrupts CDCA7 association with chromatin in vivo.
(A) Top: Representative Sanger sequencing traces from wild type (WT) (Cdca7+/+), heterozygotes (Cdca7G305V/+), and homozygotes (Cdca7G305V/G305V) depicting the c.914G > T mutation, and reference sequence for comparison. Bottom: Schematic of CDCA7 protein and the corresponding ICF3 patient mutation. (B) Bar chart showing the percentage of animals and their genotypes at mid-gestation (E14.5) and at weaning (P21). N = number of animals, collected from 10 (E14.5) and 14 litters (P21), respectively. (C) Western blot showing CDCA7 protein levels in different tissues at P21. TUBULIN was used as a loading control; two biological replicates per genotype. (D) Left: Quantitative reverse-transcriptase polymerase chain reaction (RT-qPCR) of relative Cdca7 mRNA levels in P21 spleen (normalized to β-actin; n = 4 biological replicates; data are presented as mean with black dots as individual values; error bar = SEM). Right: Western blot showing CDCA7 protein in P21 spleen; two biological replicates per genotype; TUBULIN was used as a loading control. (E) Western blot showing CDCA7 protein levels in cytoplasmic extract (CE), membrane extract (ME), soluble nuclear extract (SNE), and chromatin-bound fraction (CB) isolated from E10.5 (4 pooled embryos per genotype). TUBULIN and H3 were used as loading controls. (F) Left: Percentage green fluorescent protein (GFP)–positive erythrocytes (TER-119–positive) in WT (n = 7) and Cdca7G305V homozygous (n = 6) embryonic spleens. ****P < 0.0001 (two-sided unpaired t test). Right: Transgene methylation levels in WT and Cdca7G305V homozygous embryonic spleen measured by bisulfite sequencing (filled circles – methylated cytosines; empty circle – unmethylated cytosines). (G) Left: Gfp mRNA levels determined by RNA sequencing (RNA-seq) and represented in rpm (reads per million) [data are presented as mean with black dots as individual values; error bars = SEM; ***P < 0.0001 (two-sided unpaired t test)]. Right: Schematic of the GFP transgene and methylation levels of WT and Cdca7G305V homozygous P21 spleen measured by whole-genome bisulfite sequencing (WGBS) (average of two biological replicates per genotype).
Fig. 2.
Fig. 2.. CDCA7 is required to maintain global DNA methylation levels in multiple tissues.
(A) Bar chart showing DNA methylation levels in different tissues of P21 animals measured by LUMA. Three biological replicates per genotype [error bars = SEM; *P < 0.05, **P < 0.005, and ***P < 0.0005 (unpaired t test)]. (B) Southern blots showing minor satellite, intracisternal A particle (IAP), and major satellite repeat methylation levels in the P21 spleen. Genomic DNA was digested with HpaII (H; methylation sensitive) or MspI (M; methylation insensitive). Two or three biological replicates per genotype. (C) Western blot showing DNMT1, UHRF1, and HELLS protein levels in P21 spleen; two biological replicates per genotype. TUBULIN or VINCULIN were used as loading controls. (D) Left: Western blot showing DNMT1, HELLS, and UHRF1 protein levels in CE, ME, SNE, and CB isolated from E10.5 (four pooled embryos per genotype). TUBULIN and H3 were used as loading controls. Right: Bar chart showing DNA methylation levels in E10.5 embryos measured by LUMA; two biological replicates per genotype.
Fig. 3.
Fig. 3.. CDCA7 primarily mediates DNA methylation of the B genomic compartment in the spleen.
(A) Boxplots slowing global methylation levels measured by WGBS. Each boxplot indicates one biological replicate [middle line indicates median, box limits indicate upper and lower quartiles, and whiskers extend to 1.5× interquartile range (IQR) from quartiles]. (B) Middle: Stacked bar chart showing proportions of differentially methylated or unchanged CpGs in Cdca7G305V homozygotes. Proportions of hypomethylated (left) or hypermethylated (right) CpG sites with respect to their genomic annotation. (C) Violin plots showing WGBS methylation levels at different genomic features (average methylation levels over 1-kb tiles). Average of two biological replicates per genotype. (D) Genome browser view depicting methylation profiles over part of chromosome 2; two replicates per genotype. Replication ChIP-seq data (ENCFF001JUT) and BED tracks for previously defined common partially methylated domains (PMDs) and highly methylated domains (HMDs) (17) and cLADs (GSE17051) and RefSeq annotations are shown. Yellow shading, representative hypomethylated domains; gray rectangle, representative gene poor hypomethylated region (zoom-in in fig. S4D, top); dark orange rectangle, representative hypomethylated gene cluster (zoom-in in fig. S4D, bottom). (E) Top: Active A (green) and inactive B (blue) genomic compartments of the nucleus are illustrated. Profile plots showing methylation levels ±3 kb over cLADs, (GSE17051). Bottom: Common HMDs and PMDs (17). Methylation levels were calculated over 50 bp for cLADs, and 10-bp bins for common PMDs and HMDs. Numbers indicate a number of regions over which average methylation was calculated. (F) Violin plots showing methylation at previously defined solo-WCGWs (17) located in common HMDs or PMDs; two biological replicates per genotype. The number indicates the number of CpGs plotted (fig. S5B). (G) Profile plots showing methylation levels ±3 kb over (top) L1 and B1 elements (calculated over 10-bp bins) and (bottom) over previously defined L1- and SINE-enriched genes (21) (calculated over 10-bp bins).
Fig. 4.
Fig. 4.. CDCA7 is dispensable for H3K9me3 deposition, while aberrant DNA hypomethylation is accompanied by increased H3K27me3 at gene clusters in the spleen.
(A) Heatmap showing expression levels of the 10 differentially expressed genes. Light blue denotes low, and dark blue indicates high expression. (B) Profile plots depicting average methylation levels over protein-coding expressed and nonexpressed genes and 3-kb flanking regions in WT and Cdca7G305V homozygous spleen. (C) Profile plots and heatmaps showing CpG methylation over the center of H3K9me3 consensus peaks and flanking 10-kb regions (methylation levels calculated over 1-kb bins). (D) Profile plots showing (left) CpG methylation or (right) H3K9me3 levels at ERVK, ERVL, and ERV1 and 3-kb flanking regions in two WT and two Cdca7G305V homozygous P21 spleens (TE, transposable element). (E) Genome browser view depicting H3K9me3 and methylation over the cytochrome P450 gene cluster. Black rectangles, H3K9me3 peaks; RefSeq and Repeat masker annotations are shown; Yellow shading, H3K9me3-covered IAP. (F) Left: Heatmap showing average H3K27me3 levels at 310 differentially enriched peaks in two WT and two Cdca7G305V homozygotes [DiffBind, default settings - DEseq2, false discovery rate (FDR) ≤ 0.05]. Right: Stacked bar chart showing the proportion of H3K27me3 differential peaks (gained or lost in Cdca7G305V homozygotes) with respect to genomic annotation. (G) DNA methylation levels over promoters (TSS ± 5 kb) that gained or lost H3K27me3 in Cdca7G305V homozygotes represented by profile plots and heatmaps (methylation levels calculated over 100-bp bins). (H) Genome browser screenshot of H3K27me3 and methylation levels over clustered protocadherin alpha genes. (I) Box plot showing median expression of all genes, and genes with promoter H3K27me3 gain or loss in Cdca7G305V homozygotes (DEseq2 median of ratios for each gene was used as a measure of expression; middle line indicates median, box limits indicate upper and lower quartiles, and whiskers extend to 1.5× IQR from quartiles).
Fig. 5.
Fig. 5.. Dysregulation of clustered protocadherin gene expression and chromatin state in Cdca7G305V homozygous cerebrum.
(A) Uniform Manifold Approximation and Projection (UMAP) representations of E14.5 cerebrum integrated snRNA-seq data (n = 25,923 nuclei). Left UMAP: The colors represent four different samples grouped by genotype. Right: Integrated dataset colored according to the four annotated cell types. A small number of nuclei (n = 201 in one WT and n = 104 in one Cdca7G305V/G305V biological replicate) that could not be assigned to a specific cell type were designated “unknown” and were not included in the subsequent analyses. (B) MA plots showing differential RNA levels between WT and Cdca7G305V homozygous E14.5 cerebrum for the different cell types. Red, significantly up-regulated; blue, significantly down-regulated genes (Padj ≤ 0.01 and −1.5 ≤ log2FC ≤ 1.5). Horizontal dashed lines represent log2FC thresholds of 1.5 and −1.5; n = 2 biological replicates per genotype. (C) Heatmap showing average H3K4me3 levels at 21 differentially enriched peaks in WT and Cdca7G305V homozygous E14.5 cerebrum (DiffBind, default settings - DEseq2, FDR ≤ 0.05). (D) Miami plot visualization of H3K27me3 ChIP-seq differential analysis using sicer_df showing (top) regions that gain H3K27me3 (Cdca7G305V/WT) and (bottom) regions that lose H3K27me3 (WT/Cdca7G305V). All top-scoring dots on chromosome 18 correspond to the clustered Pcdh locus (highlighted in yellow). All four Hox clusters lose H3K27me3 as indicated (triangles indicate out-of-the-range values). (E) Bar chart showing DNA methylation levels in E14.5 cerebrum measured by LUMA. Three biological replicates per genotype [error bar = SEM; ***P < 0.0005 (unpaired t test)]. (F) Genome browser screenshot of the clustered Pcdha locus. Representative tracks for snRNA-seq (EN, excitatory neurons; IN, inhibitory neurons) and ChIP-seq (H3K4me3 and H3K27me3) from E14.5 cerebrum are shown (rectangles below H3K4me3 and H3K27me3 tracks, called peaks; RefSeq annotation is shown below the tracks).
Fig. 6.
Fig. 6.. CDCA7 is a modifier of protocadherin alpha stochastic promoter choice.
(A) Schematic of the Pcdha locus composed of variable and constant exons and experimental design. Each variable exon is spliced to the three constant exons. In a single neuron, one to five variable isoforms are randomly chosen and expressed (green rectangles). (B) Methylation levels of Pcdha1, Pcdha5, Pcdha7, and Pcdha8 promoters were measured by bisulfite sequencing in WT and Cdca7G305V homozygous E14.5 cerebrum (filled circles, methylated cytosines; empty circles, unmethylated cytosines; black box, first exon; black arrows, primers). (C) Top UMAP of WT and Cdca7G305V homozygous E14.5 cerebrum (n = 2 per genotype). Single nuclei are colored by cell type. Bottom: Feature plots showing nuclei expressing Pcdha isoforms in the annotated cell types. (D) Quantification of Pcdha isoform expression frequencies in WT and Cdca7G305V homozygous excitatory and inhibitory neurons. Frequency (in percentage %) shown on the line graph is the mean ± standard error of the two biological replicates; normalized by the total number of nuclei for each cell type (fig. S11C). Pcdha10/11 were combined because of difficulties with Cell Ranger annotation. Chi-square test *P < 0.05, **P < 0.01, and ***P < 0.001. (E) Quantification of the number of Pcdha isoforms expressed in WT and Cdca7G305V homozygous single nuclei in excitatory and inhibitory neurons. Frequency (in percentage %) shown is an average from two biological replicates per genotype; normalized by the total number of nuclei for each cell type (fig. S11C). Chi-square test *P < 0.05, **P < 0.01, and ***P < 0.001. (F) ChIP-qPCR showing CTCF occupancy at promoters of different variable Pcdha isoforms and the HS5-1 enhancer; two biological replicates per genotype. One dot indicates primers spanning conserved sequence element (CSE) before exon 1; two dots indicate primers spanning exonic CSE (eCSE). **P < 0.005 (two-tailed unpaired t test; error bars = SEM).
Fig. 7.
Fig. 7.. CDCA7 is required for Pcdh cluster methylation in the embryo.
(A) Western blot showing DNMT3B and CDCA7 protein levels in E7.5, E8.5, and E10.5 whole WT embryos; VINCULIN and TUBULIN were used as loading controls. (B) DNA methylation levels of Pcdha1 and Pcdha8 promoters measured by Sanger bisulfite sequencing in WT and Cdca7G305V homozygous E7.5 whole embryos (filled circles, methylated cytosines; empty circle, unmethylated cytosines). (C) Schematic representation of pipeline for adaptive sampling with Oxford Nanopore Technology, where only regions of interest, provided within a BED file, are sequenced. (D) Methylation levels over the Pcdh cluster, (E) the Mael and IIdr2 promoter regions, and (F) the Kcnq1 imprinted region obtained from ONT long-read sequencing. From top to bottom, the figure panels show (i) the genomic position of interest, (ii) a diagram displaying the correspondence between genome space and CpG space, and (iii) the smoothed methylated fraction plot.
Fig. 8.
Fig. 8.. Model of CDCA7-mediated DNA methylation utilization at two stochastically expressed loci.
The multicopy GFP transgene and the variable Pcdha promoters are characterized by mosaic DNA methylation, which is associated with variegated transgene expression in erythrocytes and influences stochastic Pcdha isoform choice in neuronal cells, respectively. CDCA7 disruption leads to hypomethylation and an increase in the probability of a cell expressing GFP in erythrocytes and Pcdha isoforms in neurons.

References

    1. Greenberg M. V. C., Bourc'his D., The diverse roles of DNA methylation in mammalian development and disease. Nat. Rev. Mol. Cell Biol. 20, 590–607 (2019). - PubMed
    1. Janssen S. M., Lorincz M. C., Interplay between chromatin marks in development and disease. Nat. Rev. Genet. 23, 137–153 (2022). - PubMed
    1. Thijssen P. E., Ito Y., Grillo G., Wang J., Velasco G., Nitta H., Unoki M., Yoshihara M., Suyama M., Sun Y., Lemmers R. J. L. F., de Greef J. C., Gennery A., Picco P., Kloeckener-Gruissem B., Güngör T., Reisli I., Picard C., Kebaili K., Roquelaure B., Iwai T., Kondo I., Kubota T., van Ostaijen-ten Dam M. M., van Tol M. J. D., Weemaes C., Francastel C., van der Maarel S. M., Sasaki H., Mutations in CDCA7 and HELLS cause immunodeficiency-centromeric instability-facial anomalies syndrome. Nat. Commun. 6, 7870 (2015). - PMC - PubMed
    1. Vukic M., Daxinger L., DNA methylation in disease: Immunodeficiency, centromeric instability, facial anomalies syndrome. Essays Biochem. 63, 773–783 (2019). - PMC - PubMed
    1. Velasco G., Grillo G., Touleimat N., Ferry L., Ivkovic I., Ribierre F., Deleuze J. F., Chantalat S., Picard C., Francastel C., Comparative methylome analysis of ICF patients identifies heterochromatin loci that require ZBTB24, CDCA7 and HELLS for their methylated state. Hum. Mol. Genet. 27, 2409–2424 (2018). - PubMed

Publication types

LinkOut - more resources