Analysis of KLF4 regulated genes in cancer cells reveals a role of DNA methylation in promoter- enhancer interactions

ABSTRACT

Recent studies have revealed an unexpected role of DNA methylation at promoter regions in transcription activation. However, whether DNA methylation at enhancer regions activates gene expression and influences cellular functions remains to be determined. In this study, by employing the transcription factor krÜppel-like factor 4 (KLF4) that binds to methylated CpGs (mCpGs), we investigated the molecular outcomes of the recruitment of KLF4 to mCpGs at enhancer regions in human glioblastoma cells. First, by integrating KLF4 ChIP-seq, whole-genome bisulfite sequence, and H3K27ac ChIP-seq datasets, we found 1,299 highly methylated (β >0.5) KLF4 binding sites, three-quarters of which were located at putative enhancer regions, including gene bodies and intergenic regions. In the meantime, by proteomics, we identified 16 proteins as putative targets upregulated by KLF4-mCpG binding at enhancer regions. By chromosome conformation capture (3C) analysis, we demonstrated that KLF4 bound to methylated CpGs at the enhancer regions of the B-cell lymphocyte kinase (BLK) and Lim domain only protein 7 (LMO7) genes, and activated their expression via 3D chromatin loop formation with their promoter regions. Expression of mutant KLF4, which lacks KLF4 ability to bind methylated DNA, or removal of DNA methylation in enhancer regions by a DNA methyltransferase inhibitor abolished chromatin loop formation and gene expression, suggesting the essential role of DNA methylation in enhancer-promoter interactions. Finally, we performed functional assays and showed that BLK was involved in glioblastoma cell migration. Together, our study established the concept that DNA methylation at enhancer regions interacts with transcription factors to activate gene expression and influence cellular functions.

KEYWORDS:

• Chromosome conformation capture
• methylated DNA
• enhancer
• KrÜppel-like factor 4 (KLF4)
• B-cell lymphocyte kinase

Acknowledgments

The bioinformatics analyses were performed by Johns Hopkins University School of Medicine and the Collaborative Core for Cancer Bioinformatics (C3B) shared by Indiana University Simon Cancer Center and Purdue University Center for Cancer Research supported by Walther Cancer Foundation. This work was supported by grants from NIH R01NS091165 (S.X.), NIH EY024580 (J.Q.), NIH EY023188 (J.Q.), NIH GM111514 (H.Z. and J.Q.), NIH R33CA186790 (H.Z.), NIH U54 HG006434 (H.Z.), NIH U24 CA160036 (H.Z.), Ford Foundation pre-doctoral fellowship program (O.O) and NIH T32 GM007445 (O.O).

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the National Institute of Neurological Disorders and Stroke [NS091165].