Genome editing demonstrates that the -5 kb Nanog enhancer regulates Nanog expression by modulating RNAPII initiation and/or recruitment

Abstract

Transcriptional enhancers have been defined by their ability to operate independent of distance and orientation in plasmid-based reporter assays of gene expression. At present, histone marks are used to identify and define enhancers but do not consider the endogenous role of an enhancer in the context of native chromatin. We employed a combination of genomic editing, single cell analyses, and sequencing approaches to investigate a Nanog-associated cis-regulatory element, which has been reported by others to be either an alternative promoter or a super-enhancer. We first demonstrate both distance and orientation independence in native chromatin, eliminating the issues raised with plasmid-based approaches. We next demonstrate that the dominant super-enhancer modulates Nanog globally and operates by recruiting and/or initiating RNA Polymerase II. Our studies have important implications to how transcriptional enhancers are defined and how they regulate gene expression.

Keywords: Nanog; embryonic stem cells; enhancers; gene expression; gene regulation; pluripotency; super-enhancers; transcription regulation.

Figure 1

Deletion of the −5 CRE.A, biallelic deletion of −5 CRE was achieved by inserting two loxP sites around the enhancer and cells were treated with tamoxifen (4OHT) for 6 days. Left panel, schematic; right panel, Nanog mRNA levels in bulk vehicle or 4OHT-treated cells. n = 3. B, day 6 mRNA expression of mesodermal, ectodermal, and trophectodermal differentiation markers. n = 3. C, day 6 mRNA expression of endodermal promoting transcription factors. n = 3. D, stable biallelic deletion of the −5 CRE was achieved by rescuing with a mouse Nanog cDNA. Endogenous Nanog expression is measured via the Nanog 3’UTR, while Total Nanog measured endogenous and the exogenous expression. Left panel, schematic; right panel, mRNA levels in bulk treated cells. n = 3. E, pluripotency markers in bulk treated cells in −5 CRE-deleted cells expressing Nanog in trans shown in (D). n = 3. All mRNA levels measured by RT-qPCR and shown as 2^∧ΔΔCT compared with wildtype or vehicle treated. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001 Student’s two sample t test. CRE, cis-regulatory element.

Figure 2

The −5 CRE regulates Nanog exclusively.A, differentially expressed genes, defined to be at least 2-fold, statistically significant (adj p < 0.05, n = 3) change, on chromosome 5,6, and 7 via RNA-Seq. Nanog RNAi microarray data are shown for comparison (20). B, differentially expressed genes, irrespective of significance or fold-change, on chromosome 6 within the Nanog TAD, one TAD upstream and one downstream. The −5 SE-deleted cells express Nanog in trans to prevent a loss of pluripotency. CRE, cis-regulatory element; SE, super-enhancer; TAD, topologically associated domain.

Figure 3

Manipulation of the −5 CRE.A, the −5 CRE was flipped by inserting opposing loxP sites followed by tamoxifen (4OHT) treatment. Left panel, schematic; right panel, mRNA levels of Nanog and relevant pluripotency genes in individual isolated clones. n = 3. B, an approximately 2-kb region between the −5 CRE and Nanog TSS was deleted. Left panel, schematic; right panel, Nanog mRNA levels. n = 5. C, two copies of the −5 CRE were inserted downstream of Nanog in cells where the endogenous enhancer is floxed. Endogenous enhancers were deleted via treatment with tamoxifen. Left panel, schematic; right panel, mRNA levels of Nanog in bulk cells treated with vehicle or 4OHT. n = 3. All mRNA levels measured by RT-qPCR and shown as 2^∧ΔΔCT compared with wildtype. ∗p < 0.05, ∗∗p < 0.01 Student’s two sample t test. None of the cell lines shown express Nanog in trans. CRE, cis-regulatory element.

Figure 4

−5 SE constituent enhancers are both required.A, Integrated Genome Viewer snapshot of the −5 SE showing two constituent enhancers and locations of guide RNAs. The x-axis is genomic position; y-axis is normalized tag count. B, schematic. C, mRNA levels of Nanog and relevant pluripotency genes when two constituent enhancers within the −5 SE were deleted. n = 3. All mRNA levels measured by RT-qPCR and shown as 2^∧ΔΔCT compared with wildtype. ∗∗p < 0.01, ∗∗∗p < 0.001 Student’s two sample t test. SE, super-enhancer.

Figure 5

−5 SE operates in all cells. Single-cell RT-qPCR for endogenous Nanog via Nanog 3’UTR, Total Nanog, Oct4, and ERCC3 in −5 SE-deleted cells. Expression is depicted as Log2exp relative to the Limit of Detection, as described by Fluidigm. n = 51 to 77. ∗∗∗p < 0.001 by Mann–Whitney test. The −5 SE-deleted cells express Nanog in trans to prevent a loss of pluripotency. SE, super-enhancer.

Figure 6

−5 SE operates prior to RNAPII pause release.A, Integrated Genome Viewer snapshot of CUT&Tag for Total RNAPII (n = 1) and RNAPII-Ser5P (n = 3) in wildtype and 0 copy of −5 SE cells. The left panel shows the −5 SE (y-axis = 0–233), and the right panel shows Nanog (y-axis = 0–51). Note that the exonic regions of the 0 copy cell line are confounded by the exogenous Nanog cDNA. B, Integrated Genome Viewer snapshot of ATAC-Seq data, separated by each sample of each wildtype cells where the −5 SE is floxed and cells with the −5 SE completely deleted. Differential peaks identified using DiffBind are shown in the last track. The left panel shows the −5 SE (y-axis = 0–495) and the right panel shows Nanog (y-axis = 0–292). Genes and SEs are shown below. The x-axis is genomic position; y-axis is normalized read count. The −5 SE-deleted cells express Nanog in trans (mNanogV5) to maintain pluripotency, and overlapping regions are shown in grayscale and are not included in the analysis. SE, super-enhancer.