Targeted transcriptional activation of silent oct4 pluripotency gene by combining designer TALEs and inhibition of epigenetic modifiers

Abstract

Specific control of gene activity is a valuable tool to study and engineer cellular functions. Recent studies uncovered the potential of transcription activator-like effector (TALE) proteins that can be tailored to activate user-defined target genes. It remains however unclear whether and how epigenetic modifications interfere with TALE-mediated transcriptional activation. We studied the activity of five designer TALEs (dTALEs) targeting the oct4 pluripotency gene. In vitro assays showed that the five dTALEs that target distinct sites in the oct4 promoter had the expected DNA specificity and comparable affinities to their corresponding DNA targets. In contrast to their similar in vitro properties, transcriptional activation of oct4 by these distinct dTALEs varied up to 25-fold. While dTALEs efficiently upregulated transcription of the active oct4 promoter in embryonic stem cells (ESCs) they failed to activate the silenced oct4 promoter in ESC-derived neural stem cells (NSCs), indicating that as for endogenous transcription factors also dTALE activity is limited by repressive epigenetic mechanisms. We therefore targeted the activity of epigenetic modulators and found that chemical inhibition of histone deacetylases by valproic acid or DNA methyltransferases by 5-aza-2'-deoxycytidine facilitated dTALE-mediated activation of the epigenetically silenced oct4 promoter in NSCs. Notably, demethylation of the oct4 promoter occurred only if chemical inhibitors and dTALEs were applied together but not upon treatment with inhibitors or dTALEs only. These results show that dTALEs in combination with chemical manipulation of epigenetic modifiers facilitate targeted transcriptional activation of epigenetically silenced target genes.

Figure 1.

Activation of a transgenic oct4 reporter construct by dTALEs in HEK293T cells. (A) Schematic representation of the 102-bp fragment upstream of the transcriptional start site (TSS) of the oct4 promoter, including the binding site of the Sp1/Sp3 transcription factors, the hormone responsive element (HRE) and two CpG sites (open circles). oct4-specific dTALEs are depicted in correspondence of the location of their target sequence and designated according to the distance between the 5′ end of their target sequence and the TSS. (B) Fluorescence microscopy images of HEK293T cells co-transfected with the poct4- GFP reporter construct and the T-83 dTALE constructs. Left panel shows cells transfected with the T-83 dTALE fused to the wild-type AD (wt AD). Right panel shows cells transfected with the T-83 dTALE fused to the VP16 AD. Scale bar = 200 µm. (C) Transcriptional activation of the unmethylated poct4-GFP reporter construct by oct4-specific dTALEs. eGFP expression was normalized to cells co-transfected with a control plasmid encoding the fluorescent protein mCherry (mCh) and poct4-GFP reporter construct. (D) Transcriptional activation of the in vitro methylated poct4-GFP reporter construct by oct4-specific dTALEs. eGFP expression was normalized to cells co-transfected with a control plasmid (mCh) and poct4-GFP reporter construct. To allow for a direct comparison of expression levels in (C) and (D) the data observed on the methylated promoter were normalized to the mCherry values observed with the unmethylated promoter (C). Error bars in (C) and (D) represent standard deviation from three independent experiments.

Figure 2.

The location of a dTALE target sequence within the oct4 promoters can affect its functionality. (A) Schematic representation of an oct4 promoter deletion construct in which base pairs −31 to −102 relative to the TSS were deleted and the target sequences of the four dTALEs were inserted yielding the reporter constructs TB31, TB60, TB68 and TB83. (B) Transcriptional activation of the reporter constructs TB31, TB60, TB68 and TB83 by corresponding dTALEs. (C) Background activity of the mutated reporter constructs in cells co-transfected with respective reporter and mCherry control.

Figure 3.

(A)DNA-binding properties of oct4 eGFP-dTALE fusion proteins in vitro. Schematic representation of the 102-bp upstream of the transcriptional start site (TSS) of the oct4 promoter, including the binding site of the Sp1/Sp3 transcription factors, the hormone responsive element (HRE) and two CpG sites (open circles). oct4 eGFP–dTALE fusion proteins are depicted at the position of their target sequence and numbered according to the distance between the 5′ end of their target sequence and the TSS of the oct4 gene. Binding assays were performed using fluorescently labeled double-stranded DNA substrates corresponding to position −39 to +18 [substrate A (SA)] and −88 to −31 [substrate B (SB)] relative to the TSS of the oct4 gene. Note that substrate A includes the targeting sequences of dTALEs T-11 and T-31 and substrate B includes the targeting sequences of dTALEs T-60, T-68 and T-83. (B) DNA binding of eGFP-dTALE fusions to the specific substrate in competition with the respective unspecific substrate. Shown are fluorescent intensity ratios of bound labeled DNA substrate/eGFP-dTALE fusions. eGFP was used as negative control. Values represent means and ±SEM from three independent experiments. (C) DNA affinity measurements of the five dTALEs as measured by fluorescence polarization. Upper panel shows the data points acquired for each dTALE and the corresponding fitted curves. The table contains the K_d values for each dTALE calculated from the fittings using gnuplot and the function .

Figure 4.

Activation of the endogenous oct4 gene in NSCs requires inhibition of repressive epigenetic mechanisms. (A) Relative eGFP intensities as measured by flow cytometry of mCherry-positive ogNSCs transfected with the VP16 T-83 dTALE construct (T-83). Cells transfected with control plasmid (blue) or T-83 (red) were untreated or treated with TSA (30 nM), VPA (620 µM), 5azadC (10 nM) or a combination of VPA (310 µM) and five azadC (5 nM). (B) Relative levels of endogenous oct4 mRNA measured by quantitative real-time PCR of transfected, mCherry-positive ogNSCs from (A) as well as untransfected ogNSCs and ogESCs as a reference. (C) DNA methylation levels of the oct4 promoter in samples from (A) and of ogESCs as well as ogNSCs as reference. Percentage of methylation represents the average of five CpG sites in the proximal part of the oct4 promoter. (D, E) Relative mRNA levels of tet1 and nanog as determined by quantitative real-time PCR of samples from (A) and of ogESCs as well as ogNSCs as reference. (F) Fluorescence microscopy images of ogNSCs transfected with the T-83 construct in combination with 5azadC treatment (10 nM) or no drug. Samples were stained for Oct4 protein (A647) and counterstained with DAPI. mCherry channel shows cells transfected with T-83. eGFP channel shows expression of the oct4 reporter transgene. Scale bar represents 25 µm. For images of samples treated with the other inhibitors, see Supplementary Figure S4. Error bars represent standard deviation from two to three independent experiments. Asterisks indicate samples where no mRNA was detectable by quantitative real-time PCR.