Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 7 (4), 1020-9

CTCF Haploinsufficiency Destabilizes DNA Methylation and Predisposes to Cancer

Affiliations

CTCF Haploinsufficiency Destabilizes DNA Methylation and Predisposes to Cancer

Christopher J Kemp et al. Cell Rep.

Abstract

Epigenetic alterations, particularly in DNA methylation, are ubiquitous in cancer, yet the molecular origins and the consequences of these alterations are poorly understood. CTCF, a DNA-binding protein that regulates higher-order chromatin organization, is frequently altered by hemizygous deletion or mutation in human cancer. To date, a causal role for CTCF in cancer has not been established. Here, we show that Ctcf hemizygous knockout mice are markedly susceptible to spontaneous, radiation-, and chemically induced cancer in a broad range of tissues. Ctcf(+/-) tumors are characterized by increased aggressiveness, including invasion, metastatic dissemination, and mixed epithelial/mesenchymal differentiation. Molecular analysis of Ctcf(+/-) tumors indicates that Ctcf is haploinsufficient for tumor suppression. Tissues with hemizygous loss of CTCF exhibit increased variability in CpG methylation genome wide. These findings establish CTCF as a prominent tumor-suppressor gene and point to CTCF-mediated epigenetic stability as a major barrier to neoplastic progression.

Figures

Figure 1
Figure 1. Ctcf+/− mice are susceptible to tumor development
A, Kaplan-Meier analysis of tumor-free survival of Ctcf+/− (n = 42) and Ctcf+/+ (n = 29) mice, P < 0.0001. B, H&E staining of spontaneous primary histiocytic sarcoma and corresponding lung metastasis from a Ctcf+/− mouse. Scale bars, 100 µm. C, Hematopoietic neoplasms by Ctcf genotype, **** P < 0.0001. D, Kaplan-Meier analysis of tumor-free survival of irradiated Ctcf+/− (n = 52) and Ctcf+/+ (n = 24) mice, P < 0.0001. E, H&E staining of DMBA-induced tumors from Ctcf+/− mice. Left, uterine endometrial adenocarcinoma with epithelial and mesenchymal components. Right, mammary gland adenocarcinoma with areas of spindle cell differentiation (t) and invasion of adjacent skeletal muscle (m); magnification of boxed region (bottom) reveals nuclear atypia (arrowhead). Scale bars, 100 µm. F, Frequency of endometrial lesions (left) and mammary gland histopathology (right) from DMBA-treated mice. Spindle indicates prominent neoplastic spindle cells intermixed within the tumor, ** P < 0.01, * P < 0.05. G, Kaplan-Meier analysis of tumor-free survival in urethane-treated Ctcf+/− (n = 39) and Ctcf+/+ (n = 40) mice, P < 0.0001. H, H&E staining of urethane-induced lung tumors. Scale bars, 1 mm (top) and 50 µm (bottom). I, Distribution of adenomas/adenocarcinomas in each genotype, ** P = 0.002. See also Figure S1.
Figure 2
Figure 2. Ctcf is haploinsufficient for tumor suppression
A, Southern blot analysis of lung tumors from Ctcf+/+ (lanes 1–3) and Ctcf+/− (lanes 4–7) mice. B, qPCR analysis of genomic DNA from normal lung (N) and lung tumors (T) from each genotype. C, Immunoblot analysis of CTCF protein in lung tumors from Ctcf+/+ (lanes 1–3) and Ctcf+/− (lanes 4–7) mice. α-tubulin served as loading control. D, Gel shift analysis of nuclear extracts from Ctcf+/− lung tumors show CTCF binding at both the H19/Igf2 ICR and β-globin insulator FII loci. Positions of protein-DNA complexes with 11ZF CTCF DNA binding domain (11ZF) or full length CTCF protein (CTCF) are indicated. α-CTCF antibody (α) was used to super-shift CTCF-DNA complexes. E, qRT-PCR analysis of Ctcf mRNA in Ctcf+/+ (n=3) and Ctcf+/− (n=4) MEFs, mean +/− s.e.m. F, Proliferation of Ctcf+/+ and Ctcf+/− MEFS. Assays performed in triplicate for two clones, each genotype; mean +/− s.d. G, Foci formation in MEFS cultured from Ctcf+/− compared to Ctcf+/+ mice. H, Increased foci formation in Ctcf+/− MEFs; mean +/− s.e.m., *P < 0.05, **P < 0.01.
Figure 3
Figure 3. Aberrant DNA methylation in non-cancerous lung tissue from Ctcf+/− mice
A, Examples of non-divergent and divergent CpGs within 15 lung samples. B, Hierarchical clustering of divergent autosomal CpG methylation values. C, Principle component analysis (PCA) plot for divergent autosomal CpG methylation values. D, Genomic distribution of divergent autosomal CpGs vs. all represented CpGs. E, Density plots of median CpG methylation values according to Ctcf genotype; non-divergent autosomal CpGs (top) and divergent autosomal CpGs (middle), P<7.3e-63). Divergent CpGs tend to have intermediate methylation values (bottom). F, Density plots of CpG methylation variance according to Ctcf genotype; non-divergent autosomal CpGs (left) and divergent autosomal CpGs (right), P<2.3e-05). G, CpG methylation values within 1kb intervals up- and downstream of divergent CpGs. The differences of median CpG methylation values for each sample relative to the median of all seven wild type CpG methylation values and associated P values are plotted. Shaded area indicates region with significant difference between genotypes, P<1e-25, ANOVA.
Figure 4
Figure 4. CTCF hemizygous deletion and mutation in human breast and endometrial cancer
A,B, Size plots for breast invasive carcinoma (A) and uterine corpus endometrioid carcinoma (B) indicating the total number of samples with either copy number change or mutation (yellow) or samples diploid for CTCF (blue). Copy number values are presented in discrete increments of 0.5. C,D Relative CTCF mRNA levels in breast (n=856, P < 10e-16) (C) and endometrial (n=362, P < 7.15e-12) (D) tumors are plotted according to CTCF DNA copy number. E,F, Somatic mutations within the CTCF protein coding sequence are plotted based on amino acid position (Uniprot Identifier P49711) for breast carcinoma (E) and endometrioid carcinoma (F). Synonymous (green), missense (blue), frameshift InDels (gold), and nonsense (red) mutations are shown. Blue rectangles indicate 11 ZN finger domains. G, Rendition of a typical C2-H2 type zinc finger (ZF) showing composite of missense mutations from endometrial and breast cancers. Amino acids at positions −1, 2, 3, and 6 that contact DNA directly and histidine (H) and cysteine (C) residues that coordinate Zn are indicated.
Figure 5
Figure 5. CTCF status correlates with genome wide DNA methylation patterns and patient survival in endometrial cancer
A, Chromosomal plot of significantly differentially methylated CpGs between CTCF CNA (n = 45) vs. CTCF diploid (n = 114) endometrial tumors, P < 0.001. Positive and negative differences indicate methylation probes with increased or decreased methylation in the CTCF CNA tumors, respectively. The H19 locus is indicated with a red arrow. Right, volcano plot of permutation-based significance (blue, P < 0.001) as a function of differences in average DNA methylation between CTCF CNA vs. wild type endometrial tumors (beta-value differentials). B, Chromosomal plot of significantly differentially methylated CpGs (P < 0.018) between CTCF mutant (n = 45) vs. CTCF diploid (n = 114) endometrial tumors. Note, COL14A1 at chr8 with a differential methylation value of 0.25 was deleted for clarity (see Table S8). Right, volcano plot of permutation-based significance (blue, P < 0.018) as a function of differences in average DNA methylation between CTCF mutant vs. wild type tumors (beta-value differentials). C, Kaplan Meier survival plot of patients with endometrial cancer (n=492) stratified by CTCF copy number (P < 0.05, log-rank test).

Similar articles

See all similar articles

Cited by 53 PubMed Central articles

See all "Cited by" articles

Publication types

Associated data

Feedback