The activity of human enhancers is modulated by the splicing of their associated lncRNAs

Pervasive enhancer transcription is at the origin of more than half of all long noncoding RNAs in humans. Transcription of enhancer-associated long noncoding RNAs (elncRNA) contribute to their cognate enhancer activity and gene expression regulation in cis. Recently, splicing of elncRNAs was shown to be associated with elevated enhancer activity. However, whether splicing of elncRNA transcripts is a mere consequence of accessibility at highly active enhancers or if elncRNA splicing directly impacts enhancer function, remains unanswered. We analysed genetically driven changes in elncRNA splicing, in humans, to address this outstanding question. We showed that splicing related motifs within multi-exonic elncRNAs evolved under selective constraints during human evolution, suggesting the processing of these transcripts is unlikely to have resulted from transcription across spurious splice sites. Using a genome-wide and unbiased approach, we used nucleotide variants as independent genetic factors to directly assess the causal relationship that underpin elncRNA splicing and their cognate enhancer activity. We found that the splicing of most elncRNAs is associated with changes in chromatin signatures at cognate enhancers and target mRNA expression. We provide evidence that efficient and conserved processing of enhancer-associated elncRNAs contributes to enhancer activity.

Characterization of sequence determinants of enhancer function using natural genetic variation

Sequence variation in enhancers, a class of cis-regulatory elements that control cell type-specific gene transcription, contributes significantly to phenotypic variation within human populations. Enhancers are short DNA sequences (~200 bp) composed of multiple binding sites (4-10 bp) for transcription factors (TFs). The transcriptional regulatory activity of an enhancer is encoded by the type, number, and distribution of TF binding sites that it contains. However, the sequence determinants of TF binding to enhancers and the relationship between TF binding and enhancer activity are complex, and thus it remains difficult to predict the effect of any given sequence variant on enhancer function. Here, we generate allele-specific maps of TF binding and enhancer activity in fibroblasts from a panel of F1 hybrid mice that have a high frequency of sequence variants. We identified thousands of enhancers that exhibit differences in TF binding and/or activity between alleles and use these data to define features of sequence variants that are most likely to impact enhancer function. Our data demonstrate a critical role for AP-1 TFs at many fibroblast enhancers, reveal a hierarchical relationship between AP-1 and TEAD TF binding at enhancers, and delineate the nature of sequence variants that contribute to AP-1 TF binding. These data represent one of the most comprehensive assessments to date of the impact of sequence variation on enhancer function in chromatin, with implications for identifying functional cis-regulatory variation in human populations.

Analysis of long and short enhancers in melanoma cell states

Understanding how enhancers drive cell type specificity and efficiently identifying them is essential for the development of innovative therapeutic strategies. In melanoma, the melanocytic (MEL) and the mesenchymal-like (MES) states present themselves with different responses to therapy, making the identification of specific enhancers highly relevant. Using massively parallel reporter assays (MPRA) in a panel of patient-derived melanoma lines (MM lines), we set to identify and decipher melanoma enhancers by first focusing on regions with state specific H3K27 acetylation close to differentially expressed genes. An in-depth evaluation of those regions was then pursued by investigating the activity of overlapping ATAC-seq peaks along with a full tiling of the acetylated regions with 190 bp sequences. Activity was observed in more than 60% of the selected regions and we were able to precisely locate the active enhancers within ATAC-seq peaks. Comparison of sequence content with activity, using the deep learning model DeepMEL2, revealed that AP-1 alone is responsible for the MES enhancer activity. In contrast, SOX10 and MITF both influence MEL enhancer function with SOX10 being required to achieve high levels of activity. Overall, our MPRAs shed light on the relationship between long and short sequences in terms of their sequence content, enhancer activity, and specificity across melanoma cell states.

Enhancer-associated H3K4 methylation safeguards in vitro germline competence

Germline specification in mammals occurs through an inductive process whereby competent cells in the post-implantation epiblast differentiate into primordial germ cells (PGC). The intrinsic factors that endow epiblast cells with the competence to respond to germline inductive signals remain unknown. Single-cell RNA sequencing across multiple stages of an in vitro PGC-like cells (PGCLC) differentiation system shows that PGCLC genes initially expressed in the naïve pluripotent stage become homogeneously dismantled in germline competent epiblast like-cells (EpiLC). In contrast, the decommissioning of enhancers associated with these germline genes is incomplete. Namely, a subset of these enhancers partly retain H3K4me1, accumulate less heterochromatic marks and remain accessible and responsive to transcriptional activators. Subsequently, as in vitro germline competence is lost, these enhancers get further decommissioned and lose their responsiveness to transcriptional activators. Importantly, using H3K4me1-deficient cells, we show that the loss of this histone modification reduces the germline competence of EpiLC and decreases PGCLC differentiation efficiency. Our work suggests that, although H3K4me1 might not be essential for enhancer function, it can facilitate the (re)activation of enhancers and the establishment of gene expression programs during specific developmental transitions.

Deciphering enhancer sequence using thermodynamics-based models and convolutional neural networks

Deciphering the sequence-function relationship encoded in enhancers holds the key to interpreting non-coding variants and understanding mechanisms of transcriptomic variation. Several quantitative models exist for predicting enhancer function and underlying mechanisms; however, there has been no systematic comparison of these models characterizing their relative strengths and shortcomings. Here, we interrogated a rich data set of neuroectodermal enhancers in Drosophila, representing cis- and trans- sources of expression variation, with a suite of biophysical and machine learning models. We performed rigorous comparisons of thermodynamics-based models implementing different mechanisms of activation, repression and cooperativity. Moreover, we developed a convolutional neural network (CNN) model, called CoNSEPT, that learns enhancer ‘grammar’ in an unbiased manner. CoNSEPT is the first general-purpose CNN tool for predicting enhancer function in varying conditions, such as different cell types and experimental conditions, and we show that such complex models can suggest interpretable mechanisms. We found model-based evidence for mechanisms previously established for the studied system, including cooperative activation and short-range repression. The data also favored one hypothesized activation mechanism over another and suggested an intriguing role for a direct, distance-independent repression mechanism. Our modeling shows that while fundamentally different models can yield similar fits to data, they vary in their utility for mechanistic inference. CoNSEPT is freely available at: https://github.com/PayamDiba/CoNSEPT.

Identification and functional modelling of plausibly causative cis-regulatory variants in a highly-selected cohort with X-linked intellectual disability

Identifying causative variants in cis-regulatory elements (CRE) in neurodevelopmental disorders has proven challenging. We have used in vivo functional analyses to categorize rigorously filtered CRE variants in a clinical cohort that is plausibly enriched for causative CRE mutations: 48 unrelated males with a family history consistent with X-linked intellectual disability (XLID) in whom no detectable cause could be identified in the coding regions of the X chromosome (chrX). Targeted sequencing of all chrX CRE identified six rare variants in five affected individuals that altered conserved bases in CRE targeting known XLID genes and segregated appropriately in families. Two of these variants, FMR1CRE and TENM1CRE, showed consistent site- and stage-specific differences of enhancer function in the developing zebrafish brain using dual-color fluorescent reporter assay. Mouse models were created for both variants. In male mice Fmr1CRE induced alterations in neurodevelopmental Fmr1 expression, olfactory behavior and neurophysiological indicators of FMRP function. The absence of another likely causative variant on whole genome sequencing further supported FMR1CRE as the likely basis of the XLID in this family. Tenm1CRE mice showed no phenotypic anomalies. Following the release of gnomAD 2.1, reanalysis showed that TENM1CRE exceeded the maximum plausible population frequency of a XLID causative allele. Assigning causative status to any ultra-rare CRE variant remains problematic and requires disease-relevant in vivo functional data from multiple sources. The sequential and bespoke nature of such analyses renders them time-consuming and challenging to scale for routine clinical use.

The MLL3/4 H3K4 methyltransferase complex in establishing an active enhancer landscape

Enhancers are cis-regulatory elements that play essential roles in tissue-specific gene expression during development. Enhancer function in the expression of developmental genes requires precise regulation, while deregulation of enhancer function could be the main cause of tissue-specific cancer development. MLL3/KMT2C and MLL4/KMT2D are two paralogous histone modifiers that belong to the SET1/MLL (also named COMPASS) family of lysine methyltransferases and play critical roles in enhancer-regulated gene activation. Importantly, large-scale DNA sequencing studies have revealed that they are amongst the most frequently mutated genes associated with human cancers. MLL3 and MLL4 form identical multi-protein complexes for modifying mono-methylation of histone H3 lysine 4 (H3K4) at enhancers, which together with the p300/CBP-mediated H3K27 acetylation can generate an active enhancer landscape for long-range target gene activation. Recent studies have provided a better understanding of the possible mechanisms underlying the roles of MLL3/MLL4 complexes in enhancer regulation. Moreover, accumulating studies offer new insights into our knowledge of the potential role of MLL3/MLL4 in cancer development. In this review, we summarize recent evidence on the molecular mechanisms of MLL3/MLL4 in the regulation of active enhancer landscape and long-range gene expression, and discuss their clinical implications in human cancers.

Epigenetic landscapes of intracranial aneurysm risk haplotypes implicate enhancer function of endothelial cells and fibroblasts in dysregulated gene expression

Background Genome-wide association studies have identified many single nucleotide polymorphisms (SNPs) associated with increased risk for intracranial aneurysm (IA). However, how such variants affect gene expression within IA is poorly understood. We used publicly-available ChIP-Seq data to study chromatin landscapes surrounding risk loci to determine whether IA-associated SNPs affect functional elements that regulate gene expression in cell types comprising IA tissue. Methods We mapped 16 significant IA-associated SNPs to linkage disequilibrium (LD) blocks within human genome. Using ChIP-Seq data, we examined these regions for presence of H3K4me1, H3K27ac, and H3K9ac histone marks (typically associated with latent/active enhancers). This analysis was conducted in several cell types that are present in IA tissue (endothelial cells, smooth muscle cells, fibroblasts, macrophages, monocytes, neutrophils, T cells, B cells, NK cells). In cell types with significant histone enrichment, we used HiC data to investigate topologically associated domains (TADs) encompassing the LD blocks to identify genes that may be affected by IA-associated variants. Bioinformatics were performed to determine the biological significance of these genes. Genes within HiC-defined TADs were also compared to differentially expressed genes from RNA-seq/microarray studies of IA tissues. Results We found that endothelial cells and fibroblasts, rather than smooth muscle or immune cells, have significant enrichment for enhancer marks on IA risk haplotypes (p < 0.05). Bioinformatics demonstrated that genes within TADs subsuming these regions are associated with structural extracellular matrix components and enzymatic activity. The majority of histone marked TADs (83% fibroblasts [IMR90], 77% HUVEC) encompassed at least one differentially expressed gene from IA tissue studies. Conclusions These findings provide evidence that genetic variants associated with IA risk act on endothelial cells and fibroblasts. There is strong circumstantial evidence that this may be mediated through altered enhancer function, as genes in TADs encompassing enhancer marks have also been shown to be differentially expressed in IA tissue. These genes are largely related to organization and regulation of the extracellular matrix. This study builds upon our previous (Poppenberg et al., BMC Med Genomics, 2019) by including a more diverse set of data from additional cell types and by identifying potential affected genes (i.e. those in TADs).