Enhancer-associated H3K4 methylation safeguards in vitro germline competence

Tracking-seq reveals the heterogeneity of off-target effects in CRISPR-Cas9-mediated genome editing

The continued development of novel genome editors calls for a universal method to analyze their off-target effects. Here we describe a versatile method, called Tracking-seq, for in situ identification of off-target effects that is broadly applicable to common genome-editing tools, including Cas9, base editors and prime editors. Through tracking replication protein A (RPA)-bound single-stranded DNA followed by strand-specific library construction, Tracking-seq requires a low cell input and is suitable for in vitro, ex vivo and in vivo genome editing, providing a sensitive and practical genome-wide approach for off-target detection in various scenarios. We show, using the same guide RNA, that Tracking-seq detects heterogeneity in off-target effects between different editor modalities and between different cell types, underscoring the necessity of direct measurement in the original system.

Quantifying 3'UTR length from scRNA-seq data reveals changes independent of gene expression

Although more than half of all genes generate transcripts that differ in 3'UTR length, current analysis pipelines only quantify the amount but not the length of mRNA transcripts. 3'UTR length is determined by 3' end cleavage sites (CS). We map CS in more than 200 primary human and mouse cell types and increase CS annotations relative to the GENCODE database by 40%. Approximately half of all CS are used in few cell types, revealing that most genes only have one or two major 3' ends. We incorporate the CS annotations into a computational pipeline, called scUTRquant, for rapid, accurate, and simultaneous quantification of gene and 3'UTR isoform expression from single-cell RNA sequencing (scRNA-seq) data. When applying scUTRquant to data from 474 cell types and 2134 perturbations, we discover extensive 3'UTR length changes across cell types that are as widespread and coordinately regulated as gene expression changes but affect mostly different genes. Our data indicate that mRNA abundance and mRNA length are two largely independent axes of gene regulation that together determine the amount and spatial organization of protein synthesis.

The transcription factor OCT6 promotes the dissolution of the naïve pluripotent state by repressing Nanog and activating a formative state gene regulatory network

In the mouse embryo, the transition from the preimplantation to the postimplantation epiblast is governed by changes in the gene regulatory network (GRN) that lead to transcriptional, epigenetic, and functional changes. This transition can be faithfully recapitulated in vitro by the differentiation of mouse embryonic stem cells (mESCs) to epiblast-like cells (EpiLCs), that reside in naïve and formative states of pluripotency, respectively. However, the GRN that drives this conversion is not fully elucidated. Here we demonstrate that the transcription factor OCT6 is a key driver of this process. Firstly, we show that Oct6 is not expressed in mESCs but is rapidly induced as cells exit the naïve pluripotent state. By deleting Oct6 in mESCs, we find that knockout cells fail to acquire the typical morphological changes associated with the formative state when induced to differentiate. Additionally, the key naïve pluripotency TFs Nanog, Klf2, Nr5a2, Prdm14, and Esrrb were expressed at higher levels than in wild-type cells, indicating an incomplete dismantling of the naïve pluripotency GRN. Conversely, premature expression of Oct6 in naïve cells triggered a rapid morphological transformation mirroring differentiation, that was accompanied by the upregulation of the endogenous Oct6 as well as the formative genes Sox3, Zic2/3, Foxp1, Dnmt3A and FGF5. Strikingly, we found that OCT6 represses Nanog in a bistable manner and that this regulation is at the transcriptional level. Moreover, our findings also reveal that Oct6 is repressed by NANOG. Collectively, our results establish OCT6 as a key TF in the dissolution of the naïve pluripotent state and support a model where Oct6 and Nanog form a double negative feedback loop which could act as an important toggle mediating the transition to the formative state.

Decoding chromatin states by proteomic profiling of nucleosome readers

DNA and histone modifications combine into characteristic patterns that demarcate functional regions of the genome1,2. While many 'readers' of individual modifications have been described3-5, how chromatin states comprising composite modification signatures, histone variants and internucleosomal linker DNA are interpreted is a major open question. Here we use a multidimensional proteomics strategy to systematically examine the interaction of around 2,000 nuclear proteins with over 80 modified dinucleosomes representing promoter, enhancer and heterochromatin states. By deconvoluting complex nucleosome-binding profiles into networks of co-regulated proteins and distinct nucleosomal features driving protein recruitment or exclusion, we show comprehensively how chromatin states are decoded by chromatin readers. We find highly distinctive binding responses to different features, many factors that recognize multiple features, and that nucleosomal modifications and linker DNA operate largely independently in regulating protein binding to chromatin. Our online resource, the Modification Atlas of Regulation by Chromatin States (MARCS), provides in-depth analysis tools to engage with our results and advance the discovery of fundamental principles of genome regulation by chromatin states.

A high throughput CRISPR perturbation screen identifies epigenetic regulators impacting primordial germ cell development

Certain environmental factors can impact fertility and reproductive parameters such as the number and quality of sperm and eggs. One possible mechanism is the perturbation of epigenetic landscapes in the germline. To explore this possibility, we conducted a CRISPRi screen of epigenetic-related genes to identify those that specifically perturb the differentiation of embryonic stem cells (ESCs) into primordial germ cell-like cells (PGCLCs), exploiting a highly scalable cytokine-free platform. Of the 701 genes screened, inhibition of 53 decreased the efficiency of PGCLC formation. NCOR2, a transcriptional repressor that acts via recruitment of Class I and Class IIa histone deacetylases (HDACs) to gene targets, was particularly potent in suppressing PGCLC differentiation. Consistent with evidence that histone deacetylation is crucial for germline differentiation, we found that the HDAC inhibitors (HDACi) valproic acid (VPA; an anti-convulsant) and sodium butyrate (SB; a widely-used dietary supplement) also suppressed ESC>PGCLC differentiation. Furthermore, exposure of developing mouse embryos to SB or VPA caused hypospermatogenesis. Transcriptome analyses of HDACi-treated, differentiating ESC>PGCLC cultures revealed suppression of germline-associated pathways and enhancement of somatic pathways. This work demonstrates the feasibility of conducting large-scale functional screens of genes, chemicals, or other agents that may impact germline development.

METTL14 regulates chromatin bivalent domains in mouse embryonic stem cells

METTL14 (methyltransferase-like 14) is an RNA-binding protein that partners with METTL3 to mediate N6-methyladenosine (m6A) methylation. Recent studies identified a function for METTL3 in heterochromatin in mouse embryonic stem cells (mESCs), but the molecular function of METTL14 on chromatin in mESCs remains unclear. Here, we show that METTL14 specifically binds and regulates bivalent domains, which are marked by trimethylation of histone H3 lysine 27 (H3K27me3) and lysine 4 (H3K4me3). Knockout of Mettl14 results in decreased H3K27me3 but increased H3K4me3 levels, leading to increased transcription. We find that bivalent domain regulation by METTL14 is independent of METTL3 or m6A modification. METTL14 enhances H3K27me3 and reduces H3K4me3 by interacting with and probably recruiting the H3K27 methyltransferase polycomb repressive complex 2 (PRC2) and H3K4 demethylase KDM5B to chromatin. Our findings identify an METTL3-independent role of METTL14 in maintaining the integrity of bivalent domains in mESCs, thus indicating a mechanism of bivalent domain regulation in mammals.

Epigenetic Control of Cell Potency and Fate Determination during Mammalian Gastrulation

Pluripotent embryonic stem cells have a unique and characteristic epigenetic profile, which is critical for differentiation to all embryonic germ lineages. When stem cells exit the pluripotent state and commit to lineage-specific identities during the process of gastrulation in early embryogenesis, extensive epigenetic remodelling mediates both the switch in cellular programme and the loss of potential to adopt alternative lineage programmes. However, it remains to be understood how the stem cell epigenetic profile encodes pluripotency, or how dynamic epigenetic regulation helps to direct cell fate specification. Recent advances in stem cell culture techniques, cellular reprogramming, and single-cell technologies that can quantitatively profile epigenetic marks have led to significant insights into these questions, which are important for understanding both embryonic development and cell fate engineering. This review provides an overview of key concepts and highlights exciting new advances in the field.

MLL3/MLL4 methyltransferase activities control early embryonic development and embryonic stem cell differentiation in a lineage-selective manner

H3K4me1 methyltransferases MLL3 (KMT2C) and MLL4 (KMT2D) are critical for enhancer activation, cell differentiation and development. However, roles of MLL3/4 enzymatic activities and MLL3/4-mediated enhancer H3K4me1 in these processes remain unclear. Here we report that constitutive elimination of both MLL3 and MLL4 enzymatic activities prevents initiation of gastrulation and leads to early embryonic lethality in mice. However, selective elimination of MLL3/4 enzymatic activities in embryonic, but not extraembryonic, lineages leaves gastrulation largely intact. Consistent with this, embryonic stem cells (ESCs) lacking MLL3/4 enzymatic activities can differentiate toward the three embryonic germ layers but show aberrant differentiation to extraembryonic endoderm (ExEn) and trophectoderm. The failure in ExEn differentiation can be attributed to markedly reduced enhancer-binding of the lineage-determining transcription factor GATA6. Furthermore, we show that MLL3/4-catalyzed H3K4me1 is largely dispensable for enhancer activation during ESC differentiation. Together, our findings suggest a lineage-selective, but enhancer activation-independent, role of MLL3/4 methyltransferase activities in early embryonic development and ESC differentiation.

Loss of MLL3/4 decouples enhancer H3K4 monomethylation, H3K27 acetylation, and gene activation during embryonic stem cell differentiation

BACKGROUND

Enhancers are essential in defining cell fates through the control of cell-type-specific gene expression. Enhancer activation is a multi-step process involving chromatin remodelers and histone modifiers including the monomethylation of H3K4 (H3K4me1) by MLL3 (KMT2C) and MLL4 (KMT2D). MLL3/4 are thought to be critical for enhancer activation and cognate gene expression including through the recruitment of acetyltransferases for H3K27.

RESULTS

Here we test this model by evaluating the impact of MLL3/4 loss on chromatin and transcription during early differentiation of mouse embryonic stem cells. We find that MLL3/4 activity is required at most if not all sites that gain or lose H3K4me1 but is largely dispensable at sites that remain stably methylated during this transition. This requirement extends to H3K27 acetylation (H3K27ac) at most transitional sites. However, many sites gain H3K27ac independent of MLL3/4 or H3K4me1 including enhancers regulating key factors in early differentiation. Furthermore, despite the failure to gain active histone marks at thousands of enhancers, transcriptional activation of nearby genes is largely unaffected, thus uncoupling the regulation of these chromatin events from transcriptional changes during this transition. These data challenge current models of enhancer activation and imply distinct mechanisms between stable and dynamically changing enhancers.

CONCLUSIONS

Collectively, our study highlights gaps in knowledge about the steps and epistatic relationships of enzymes necessary for enhancer activation and cognate gene transcription.

Could senescence phenotypes strike the balance to promote tumor dormancy?

After treatment and surgery, patient tumors can initially respond followed by a rapid relapse, or respond well and seemingly be cured, but then recur years or decades later. The state of surviving cancer cells during the long, undetected period is termed dormancy. By definition, the dormant tumor cells do not proliferate to create a mass that is detectable or symptomatic, but also never die. An intrinsic state and microenvironment that are inhospitable to the tumor would bias toward cell death and complete eradication, while conditions that favor the tumor would enable growth and relapse. In neither case would clinical dormancy be observed. Normal cells and tumor cells can enter a state of cellular senescence after stress such as that caused by cancer therapy. Senescence is characterized by a stable cell cycle arrest mediated by chromatin modifications that cause gene expression changes and a secretory phenotype involving many cytokines and chemokines. Senescent cell phenotypes have been shown to be both tumor promoting and tumor suppressive. The balance of these opposing forces presents an attractive model to explain tumor dormancy: phenotypes of stable arrest and immune suppression could promote survival, while reversible epigenetic programs combined with cytokines and growth factors that promote angiogenesis, survival, and proliferation could initiate the emergence from dormancy. In this review, we examine the phenotypes that have been characterized in different normal and cancer cells made senescent by various stresses and how these might explain the characteristics of tumor dormancy.

Epiblast-like stem cells established by Wnt/β-catenin signaling manifest distinct features of formative pluripotency and germline competence

Different formative pluripotent stem cells harboring similar functional properties have been recently established to be lineage neutral and germline competent yet have distinct molecular identities. Here, we show that WNT/β-catenin signaling activation sustains transient mouse epiblast-like cells as epiblast-like stem cells (EpiLSCs). EpiLSCs display metastable formative pluripotency with bivalent cellular energy metabolism and unique transcriptomic features and chromatin accessibility. We develop single-cell stage label transfer (scSTALT) to study the formative pluripotency continuum and reveal that EpiLSCs recapitulate a unique developmental period in vivo, filling the gap of the formative pluripotency continuum between other published formative stem cells. WNT/β-catenin signaling activation counteracts differentiation effects of activin A and bFGF by preventing complete dissolution of naive pluripotency regulatory network. Moreover, EpiLSCs have direct competence toward germline specification, which is further matured by an FGF receptor inhibitor. Our EpiLSCs can serve as an in vitro model for mimicking and studying early post-implantation development and pluripotency transition.

Developmental phenomics suggests that H3K4 monomethylation confers multi-level phenotypic robustness

How histone modifications affect animal development remains difficult to ascertain. Despite the prevalence of histone 3 lysine 4 monomethylation (H3K4me1) on enhancers, hypomethylation appears to have minor effects on phenotype and viability. Here, we genetically reduce H3K4me1 deposition in Drosophila melanogaster and find that hypomethylation reduces transcription factor enrichment in nuclear microenvironments, disrupts gene expression, and reduces phenotypic robustness. Using a developmental phenomics approach, we find changes in morphology, metabolism, behavior, and offspring production. However, many phenotypic changes are only detected when hypomethylated flies develop outside of standard laboratory environments or with specific genetic backgrounds. Therefore, quantitative phenomics measurements can unravel how pleiotropic modulators of gene expression affect developmental robustness under conditions resembling the natural environments of a species.

Mouse primordial germ-cell-like cells lack piRNAs

PIWI-interacting RNAs (piRNAs) are small RNAs bound by PIWI-clade Argonaute proteins that function to silence transposable elements (TEs). Following mouse primordial germ cell (mPGC) specification around E6.25, fetal piRNAs emerge in male gonocytes from E13.5 onward. The in vitro differentiation of mPGC-like cells (mPGCLCs) has raised the possibility of studying the fetal piRNA pathway in greater depth. However, using single-cell RNA-seq and RT-qPCR along mPGCLC differentiation, we find that piRNA pathway factors are not fully expressed in Day 6 mPGCLCs. Moreover, we do not detect piRNAs across a panel of Day 6 mPGCLC lines using small RNA-seq. Our combined efforts highlight that in vitro differentiated Day 6 mPGCLCs do not yet resemble E13.5 or later mouse gonocytes where the piRNA pathway is active. This Matters Arising paper is in response to von Meyenn et al. (2016), published in Developmental Cell. See also the correction by von Meyenn et al. published in this issue.

Histone post-translational modifications - cause and consequence of genome function

Much has been learned since the early 1960s about histone post-translational modifications (PTMs) and how they affect DNA-templated processes at the molecular level. This understanding has been bolstered in the past decade by the identification of new types of histone PTM, the advent of new genome-wide mapping approaches and methods to deposit or remove PTMs in a locally and temporally controlled manner. Now, with the availability of vast amounts of data across various biological systems, the functional role of PTMs in important processes (such as transcription, recombination, replication, DNA repair and the modulation of genomic architecture) is slowly emerging. This Review explores the contribution of histone PTMs to the regulation of genome function by discussing when these modifications play a causative (or instructive) role in DNA-templated processes and when they are deposited as a consequence of such processes, to reinforce and record the event. Important advances in the field showing that histone PTMs can exert both direct and indirect effects on genome function are also presented.

UvKmt2-Mediated H3K4 Trimethylation Is Required for Pathogenicity and Stress Response in Ustilaginoidea virens

Epigenetic modification is important for cellular functions. Trimethylation of histone H3 lysine 4 (H3K4me3), which associates with transcriptional activation, is one of the important epigenetic modifications. In this study, the biological functions of UvKmt2-mediated H3K4me3 modification were characterized in Ustilaginoidea virens, which is the causal agent of the false smut disease, one of the most destructive diseases in rice. Phenotypic analyses of the ΔUvkmt2 mutant revealed that UvKMT2 is necessary for growth, conidiation, secondary spore formation, and virulence in U. virens. Immunoblotting and chromatin immunoprecipitation assay followed by sequencing (ChIP-seq) showed that UvKMT2 is required for the establishment of H3K4me3, which covers 1729 genes of the genome in U. virens. Further RNA-seq analysis demonstrated that UvKmt2-mediated H3K4me3 acts as an important role in transcriptional activation. In particular, H3K4me3 modification involves in the transcriptional regulation of conidiation-related and pathogenic genes, including two important mitogen-activated protein kinases UvHOG1 and UvPMK1. The down-regulation of UvHOG1 and UvPMK1 genes may be one of the main reasons for the reduced pathogenicity and stresses adaptability of the ∆Uvkmt2 mutant. Overall, H3K4me3, established by histone methyltransferase UvKMT2, contributes to fungal development, secondary spore formation, virulence, and various stress responses through transcriptional regulation in U. virens.

Cnot8 eliminates naïve regulation networks and is essential for naïve-to-formative pluripotency transition

Mammalian early epiblasts at different phases are characterized by naïve, formative, and primed pluripotency states, involving extensive transcriptome changes. Here, we report that deadenylase Cnot8 of Ccr4-Not complex plays essential roles during the transition from naïve to formative state. Knock out (KO) Cnot8 resulted in early embryonic lethality in mice, but Cnot8 KO embryonic stem cells (ESCs) could be established. Compared with the cells differentiated from normal ESCs, Cnot8 KO cells highly expressed a great many genes during their differentiation into the formative state, including several hundred naïve-like genes enriched in lipid metabolic process and gene expression regulation that may form the naïve regulation networks. Knockdown expression of the selected genes of naïve regulation networks partially rescued the differentiation defects of Cnot8 KO ESCs. Cnot8 depletion led to the deadenylation defects of its targets, increasing their poly(A) tail lengths and half-life, eventually elevating their expression levels. We further found that Cnot8 was involved in the clearance of targets through its deadenylase activity and the binding of Ccr4-Not complex, as well as the interacting with Tob1 and Pabpc1. Our results suggest that Cnot8 eliminates naïve regulation networks through mRNA clearance, and is essential for naïve-to-formative pluripotency transition.

Analysis of sub-kilobase chromatin topology reveals nano-scale regulatory interactions with variable dependence on cohesin and CTCF

Enhancers and promoters predominantly interact within large-scale topologically associating domains (TADs), which are formed by loop extrusion mediated by cohesin and CTCF. However, it is unclear whether complex chromatin structures exist at sub-kilobase-scale and to what extent fine-scale regulatory interactions depend on loop extrusion. To address these questions, we present an MNase-based chromosome conformation capture (3C) approach, which has enabled us to generate the most detailed local interaction data to date (20 bp resolution) and precisely investigate the effects of cohesin and CTCF depletion on chromatin architecture. Our data reveal that cis-regulatory elements have distinct internal nano-scale structures, within which local insulation is dependent on CTCF, but which are independent of cohesin. In contrast, we find that depletion of cohesin causes a subtle reduction in longer-range enhancer-promoter interactions and that CTCF depletion can cause rewiring of regulatory contacts. Together, our data show that loop extrusion is not essential for enhancer-promoter interactions, but contributes to their robustness and specificity and to precise regulation of gene expression.

Distal and proximal cis-regulatory elements sense X chromosome dosage and developmental state at the Xist locus

Developmental genes such as Xist, which initiates X chromosome inactivation, are controlled by complex cis-regulatory landscapes, which decode multiple signals to establish specific spatiotemporal expression patterns. Xist integrates information on X chromosome dosage and developmental stage to trigger X inactivation in the epiblast specifically in female embryos. Through a pooled CRISPR screen in differentiating mouse embryonic stem cells, we identify functional enhancer elements of Xist at the onset of random X inactivation. Chromatin profiling reveals that X-dosage controls the promoter-proximal region, while differentiation cues activate several distal enhancers. The strongest distal element lies in an enhancer cluster associated with a previously unannotated Xist-enhancing regulatory transcript, which we named Xert. Developmental cues and X-dosage are thus decoded by distinct regulatory regions, which cooperate to ensure female-specific Xist upregulation at the correct developmental time. With this study, we start to disentangle how multiple, functionally distinct regulatory elements interact to generate complex expression patterns in mammals.