Targeted DNA demethylation and activation of endogenous genes using programmable TALE-TET1 fusion proteins

Abstract

Genome-wide studies have defined cell type-specific patterns of DNA methylation that are important for regulating gene expression in both normal development and disease. However, determining the functional significance of specific methylation events remains challenging, owing to the lack of methods for removing such modifications in a targeted manner. Here we describe an approach for efficient targeted demethylation of specific CpGs in human cells using fusions of engineered transcription activator-like effector (TALE) repeat arrays and the TET1 hydroxylase catalytic domain. Using these TALE-TET1 fusions, we demonstrate that modification of critical methylated promoter CpG positions can lead to substantial increases in the expression of endogenous human genes. Our results delineate a strategy for understanding the functional significance of specific CpG methylation marks in the context of endogenous gene loci and validate programmable DNA demethylation reagents with potential utility for research and therapeutic applications.

Figure 1. TALE-TET1 Fusion Proteins Induce Targeted Demethylation in Human Cells

(A) Schematic illustrating the predicted domain architecture of the full-length TET1 protein (TET1 FL) and the catalytic domain of the TET1 protein (TET1 CD). CXXC = CXXC-type zinc-binding domain, NLS = nuclear localization signal, Cys-rich = cysteine-rich region, DSBH = double-stranded β helix domain (DSBH). (B) Schematic illustrating general structure of fusions between a TALE repeat array DNA-binding domain and either TET1 FL or TET1 CD. (C) Schematic illustrating the human KLF4 locus with CpGs indicated with black arrows. Red arrows indicate the location and direction (5’ to 3’) of TALE-TET1 fusion protein binding sites. Numbering on the bottom line indicates position on the DNA relative to the start site of transcription (right-angle arrow) and numbering on the top line indicates position relative to the beginning of intron 2/3. (D) Demethylation activities of KL-1 TALE-TET1 CD (top) and KL-1 TALE-TET1 FL (bottom) fusion proteins in human K562 cells. Graphs show the fraction of CpGs methylated (y-axis) for different positions along the length of the second KLF4 intron (x-axis, numbered relative to the start of the intron). Blue arrow indicates the location and direction (5’ to 3’) of TALE target binding site. Each point represents the mean of three independent transfection experiments with bars indicating standard errors of the mean. Methylation status for each experiment was assessed using high-throughput bi-sulfite sequencing. Off-target TALE fusions bind to an unrelated sequence in the EGFP reporter gene. Note that the GFP control datapoints shown are the same for both graphs and are depicted twice for ease of comparison. (E) Demethylation activities of KL-2 TALE-TET1 CD (top) and KL-2 TALE-TET1 FL (bottom) fusion proteins in human K562 cells. Same as in (C) but using a TALE repeat array targeted to a second site in the KLF4 intron.

Figure 2. Targeted demethylation and activation of the human RHOXF2 gene by TALE-TET1 Fusion Proteins

(A) Schematic illustrating the human RHOXF2 locus with CpGs indicated with black arrows. Red arrows indicate the location and direction (5’ to 3’) of 11 TALE-TET1 fusion protein binding sites. Numbering indicates position on the DNA relative to the start site of transcription (right-angle arrow). (B) Expression levels of RHOXF2 mRNA (normalized to β-actin mRNA levels) in HeLa cells transfected with plasmids expressing RH-3 or RH-4 TALE-TET1 proteins, RH-3 or RH-4 TALE-TET1 proteins bearing mutations that inactivate TET1 catalytic function, or an off-target TALE-TET1 protein (targeted to KLF4) or a control GFP expression plasmid. Means of three independent samples each assayed three times by quantitative RT-PCR are shown with bars representing standard errors of the mean. Asterisks indicate samples with values significantly greater than those obtained with the off-target control as determined by a one-sided Welch’s t-test (n=3, p<0.05). (C) Expression levels of RHOXF2 mRNA (normalized to β-actin mRNA levels) in 293 cells transfected with plasmids expressing the indicated TALE-TET1, catalytically inactivated TALE-TET1 or off-target TALE-TET1 fusion protein or a control GFP expression plasmid. Data presented as in (B). (D) Demethylation activities of RH-3 and RH-4 TALE-TET1 fusion proteins in HeLa (top) and 293 (bottom) cells. Graphs show the fraction of CpGs methylated (y-axis) for different positions along the length of the RHOXF2 promoter (x-axis, numbered relative to the transcription start site) in cells transfected with plasmids expressing the RH-3, RH-4 (blue), or an off-target (to a site in the KLF4 gene) TALE-TET1 protein (orange) or GFP (green). Each data point represents the mean of three independent transfection experiments with bars indicating standard errors of the mean. Methylation status for each experiment was assessed using high-throughput bi-sulfite sequencing. Note that the GFP and off-target TALE-TET1 control data points shown are the same in both panels for each cell type and are depicted multiple times only for ease of comparison with experimental samples. Blue arrow indicates the location and direction (5’ to 3’) of the RH3 or RH4 TALE-TET1 binding sites. (E) Demethylation activities of catalytically inactive RH-3 and RH-4 TALE-TET1 fusion proteins in HeLa (top) and 293 (bottom) cells. Data represented as in (D) but here showing fraction of CpGs methylated in cells transfected with plasmids expressing RH-3 or RH-4 TALE-TET1 fusions bearing mutations that inactivate TET1 function (red). Note that RH-3, RH-4 and GFP data are the same as in (D) and are depicted again here for ease of comparison with catalytically inactive controls.

Figure 3. Targeted demethylation and activation of the human HBB gene by TALE-TET1 Fusion Proteins

(A) Schematic illustrating the human HBB locus with CpGs indicated with black arrows. Numbering indicates position on the DNA relative to the start site of transcription (right-angle arrow). Colored arrows indicate the location and direction (5’ to 3’) of 10 TALE-TET1 fusion protein binding sites. (B) Demethylation activities of HB-1, -2, -3, -4, -5, and -6 TALE-TET1 fusion proteins in K562 cells. Graphs show the fraction of CpGs methylated (y-axis) for different positions along the length of the HBB promoter (x-axis, numbered relative to the transcription start site) in K562 cells transfected with plasmids expressing one of the six HB TALE-TET1 fusion proteins (colored circles), an off-target TALE-TET1 fusion to a site in the KLF4 gene (black triangles), or GFP (green triangles). Each data point represents the mean of three independent transfection experiments with bars indicating standard errors of the mean. Methylation status for each experiment was assessed using high-throughput bisulfite sequencing. Note that the GFP and off-target TALE-TET1 control data points shown are the same in both panels and are depicted twice for ease of comparison with experimental samples. Colored arrows indicate the location and direction (5’ to 3’) of the various HB TALE-TET1 binding sites. (C) Demethylation activities of HB-7, -8, -9, and -10 TALE-TET1 fusion proteins in K562 cells. Data represented as in (B) (D) Expression levels of HBB mRNA (normalized to β-actin mRNA levels) in K562 cells transfected with indicated TALE-TET1 fusion protein expression plasmids or a control GFP expression plasmid. Means of three independent samples assayed by quantitative RT-PCR are shown with bars representing standard errors of the mean. Asterisks indicate samples with values significantly greater than those obtained with an off-target control TALE-TET1 fusion protein (targeted to a site in the human KLF4 gene) as determined by a one-sided Welch’s t-test (n=3, p<0.05). (E) Graphs showing the fraction of CpGs methylated (y-axis) for different positions along the length of the HBB promoter (x-axis, numbered relative to the transcription start site) in cells transfected with plasmids expressing the HB-4, HB-5 or HB-6 TALE-TET1 fusion, HB-4, HB-5 or HB-6 TALE-TET1 bearing mutations that inactivate TET1 catalytic function, or GFP. Each data point represents the mean of three independent transfection experiments with bars indicating standard errors of the mean. Methylation status for each experiment was assessed using high-throughput bi-sulfite sequencing. Note that the control GFP data shown is the same in all three panels and is depicted multiple times for ease of comparison with experimental samples. Blue arrow indicates the location and direction (5’ to 3’) of the TALE-TET1 fusion protein binding sites. (F) Expression levels of HBB mRNA (normalized to β-actin mRNA levels) in K562 cells transfected with plasmids expressing HB-4, HB-5 or HB-6 TALE-TET1 fusions, HB-4, HB-5 or HB-6 TALE-TET1 fusions bearing mutations that inactivate TET1 catalytic function, an off-target TALE-TET1 fusion targeted to an unrelated site in the human KLF4 gene, or GFP. Means of three independent samples assayed by quantitative RT-PCR are shown with bars representing standard errors of the mean. Asterisks indicate samples with values significantly greater than those obtained with the off-target control TALE-TET1 fusion protein as determined by a one-sided, two-sample equal variance t-test (n=3, p<0.05).