Identification of 1600 replication origins in S. cerevisiae.

There are approximately 500 known origins of replication in the yeast genome, and the process by which DNA replication initiates at these locations is well understood. In particular, these sites are made competent to initiate replication by loading of the Mcm replicative helicase prior to the start of S phase; thus, 'a site that binds Mcm in G1' might be considered to provide an operational definition of a replication origin. By fusing a subunit of Mcm to micrococcal nuclease, we previously showed that known origins are typically bound by a single Mcm double hexamer, loaded adjacent to the ARS consensus sequence (ACS). Here, we extend this analysis from known origins to the entire genome, identifying candidate Mcm binding sites whose signal intensity varies over at least three orders of magnitude. Published data quantifying single-stranded DNA (ssDNA) during S phase revealed replication initiation among the most abundant 1600 of these sites, with replication activity decreasing with Mcm abundance and disappearing at the limit of detection of ssDNA. Three other hallmarks of replication origins were apparent among the most abundant 5500 sites. Specifically, these sites: (1) appeared in intergenic nucleosome-free regions flanked on one or both sides by well-positioned nucleosomes; (2) were flanked by ACSs; and (3) exhibited a pattern of GC skew characteristic of replication initiation. We conclude that, if sites at which Mcm double hexamers are loaded can function as replication origins, then DNA replication origins are at least threefold more abundant than previously assumed, and we suggest that replication may occasionally initiate in essentially every intergenic region. These results shed light on recent reports that as many as 15% of replication events initiate outside of known origins, and this broader distribution of replication origins suggest that S phase in yeast may be less distinct from that in humans than widely assumed.

The Density of Regulatory Information Is a Major Determinant of Evolutionary Constraint on Noncoding DNA in Drosophila.

Evolutionary analyses have estimated that ∼60% of nucleotides in intergenic regions of the Drosophila melanogaster genome are functionally relevant, suggesting that regulatory information may be encoded more densely in intergenic regions than has been revealed by most functional dissections of regulatory DNA. Here, we approached this issue through a functional dissection of the regulatory region of the gene shavenbaby (svb). Most of the ∼90 kb of this large regulatory region is highly conserved in the genus Drosophila, though characterized enhancers occupy a small fraction of this region. By analyzing the regulation of svb in different contexts of Drosophila development, we found that the regulatory information that drives svb expression in the abdominal pupal epidermis is organized in a different way than the elements that drive svb expression in the embryonic epidermis. While in the embryonic epidermis svb is activated by compact enhancers separated by large inactive DNA regions, svb expression in the pupal epidermis is driven by regulatory information distributed over broader regions of svb cis-regulatory DNA. In the same vein, we observed that other developmental genes also display a dense distribution of putative regulatory elements in their regulatory regions. Furthermore, we found that a large percentage of conserved noncoding DNA of the Drosophila genome is contained within regions of open chromatin. These results suggest that part of the evolutionary constraint on noncoding DNA of Drosophila is explained by the density of regulatory information, which may be greater than previously appreciated.

The structure and mechanism of action of a distinct class of dicistrovirus intergenic region IRESs.

Internal ribosomal entry sites (IRESs) engage with the eukaryotic translation apparatus to promote end-independent initiation. We identified a conserved class of ∼150 nt long intergenic region (IGR) IRESs in dicistrovirus genomes derived from members of the phyla Arthropoda, Bryozoa, Cnidaria, Echinodermata, Entoprocta, Mollusca and Porifera. These IRESs, exemplified by Wenling picorna-like virus 2, resemble the canonical cricket paralysis virus (CrPV) IGR IRES in comprising two nested pseudoknots (PKII/PKIII) and a 3'-terminal pseudoknot (PKI) that mimics a tRNA anticodon stem-loop base-paired to mRNA. However, they are ∼50 nt shorter than CrPV-like IRESs, and PKIII is an H-type pseudoknot that lacks the SLIV and SLV stem-loops that are primarily responsible for the affinity of CrPV-like IRESs for the 40S ribosomal subunit and that restrict initial binding of PKI to its aminoacyl (A) site. Wenling-class IRESs bound strongly to 80S ribosomes but only weakly to 40S subunits. Whereas CrPV-like IRESs must be translocated from the A site to the peptidyl (P) site by elongation factor 2 for elongation to commence, Wenling-class IRESs bound directly to the P site of 80S ribosomes, and decoding begins without a prior translocation step. A chimeric CrPV clone containing a Wenling-class IRES was infectious, confirming that the IRES functioned in cells.

Chance promoter activities illuminate the origins of eukaryotic intergenic transcriptions.

It is debated whether the pervasive intergenic transcription from eukaryotic genomes has functional significance or simply reflects the promiscuity of RNA polymerases. We approach this question by comparing chance promoter activities with the expression levels of intergenic regions in the model eukaryote Saccharomyces cerevisiae. We build a library of over 105 strains, each carrying a 120-nucleotide, chromosomally integrated, completely random sequence driving the potential transcription of a barcode. Quantifying the RNA concentration of each barcode in two environments reveals that 41-63% of random sequences have significant, albeit usually low, promoter activities. Therefore, even in eukaryotes, where the presence of chromatin is thought to repress transcription, chance transcription is prevalent. We find that only 1-5% of yeast intergenic transcriptions are unattributable to chance promoter activities or neighboring gene expressions, and these transcriptions exhibit higher-than-expected environment-specificity. These findings suggest that only a minute fraction of intergenic transcription is functional in yeast.

Initiation of translation on nedicistrovirus and related intergenic region IRESs by their factor-independent binding to the P site of 80S ribosomes.

Initiation of translation on many viral mRNAs occurs by noncanonical mechanisms that involve 5' end-independent binding of ribosomes to an internal ribosome entry site (IRES). The ∼190-nt-long intergenic region (IGR) IRES of dicistroviruses such as cricket paralysis virus (CrPV) initiates translation without Met-tRNAi Met or initiation factors. Advances in metagenomics have revealed numerous dicistrovirus-like genomes with shorter, structurally distinct IGRs, such as nedicistrovirus (NediV) and Antarctic picorna-like virus 1 (APLV1). Like canonical IGR IRESs, the ∼165-nt-long NediV-like IGRs comprise three domains, but they lack key canonical motifs, including L1.1a/L1.1b loops (which bind to the L1 stalk of the ribosomal 60S subunit) and the apex of stem-loop V (SLV) (which binds to the head of the 40S subunit). Domain 2 consists of a compact, highly conserved pseudoknot (PKIII) that contains a UACUA loop motif and a protruding CrPV-like stem--loop SLIV. In vitro reconstitution experiments showed that NediV-like IRESs initiate translation from a non-AUG codon and form elongation-competent 80S ribosomal complexes in the absence of initiation factors and Met-tRNAi Met Unlike canonical IGR IRESs, NediV-like IRESs bind directly to the peptidyl (P) site of ribosomes leaving the aminoacyl (A) site accessible for decoding. The related structures of NediV-like IRESs and their common mechanism of action indicate that they exemplify a distinct class of IGR IRES.

Hematopoietic/erythroid enhancers activate nearby target genes by extending histone H3K27ac and transcribing intergenic RNA.

Enhancers activate gene transcription remotely, which requires tissue specific transcription factors binding to them. GATA1 and TAL1 are hematopoietic/erythroid-specific factors and often bind together to enhancers, activating target genes. Interestingly, we found that some hematopoietic/erythroid genes are transcribed in a GATA1-dependent but TAL1-independnet manner. They appear to have enhancers within a relatively short distance. In this study, we paired highly transcribed hematopoietic/erythroid genes with the nearest GATA1/TAL1-binding enhancers and analyzed these putative enhancer-gene pairs depending on distance between them. Enhancers located at various distances from genes in the pairs, which was not related to transcription level of the genes. However, genes with enhancers at short distances away tended to be transcriptionally unaffected by TAL1 depletion. Histone H3K27ac extended from the enhancers to target genes. The H3K27ac extension was maintained without TAL1, even though it disappeared owing to the loss of GATA1. Intergenic RNA was highly transcribed from the enhancers to nearby target genes, independent of TAL1. Taken together, TAL1-independent transcription of hematopoietic/erythroid genes appears to be promoted by enhancers present in a short distance. These enhancers are likely to activate nearby target genes by tracking the intervening regions.

CRISPR-Cas9-mediated functional dissection of the foxc1 genomic region in zebrafish identifies critical conserved cis-regulatory elements.

FOXC1 encodes a forkhead-domain transcription factor associated with several ocular disorders. Correct FOXC1 dosage is critical to normal development, yet the mechanisms controlling its expression remain unknown. Together with FOXQ1 and FOXF2, FOXC1 is part of a cluster of FOX genes conserved in vertebrates. CRISPR-Cas9-mediated dissection of genomic sequences surrounding two zebrafish orthologs of FOXC1 was performed. This included five zebrafish-human conserved regions, three downstream of foxc1a and two remotely upstream of foxf2a/foxc1a or foxf2b/foxc1b clusters, as well as two intergenic regions between foxc1a/b and foxf2a/b lacking sequence conservation but positionally corresponding to the area encompassing a previously reported glaucoma-associated SNP in humans. Removal of downstream sequences altered foxc1a expression; moreover, zebrafish carrying deletions of two or three downstream elements demonstrated abnormal phenotypes including enlargement of the anterior chamber of the eye reminiscent of human congenital glaucoma. Deletions of distant upstream conserved elements influenced the expression of foxf2a/b or foxq1a/b but not foxc1a/b within each cluster. Removal of either intergenic sequence reduced foxc1a or foxc1b expression during late development, suggesting a role in transcriptional regulation despite the lack of conservation at the nucleotide level. Further studies of the identified regions in human patients may explain additional individuals with developmental ocular disorders.

Stroke-associated intergenic variants modulate a human FOXF2 transcriptional enhancer.

SNPs associated with human stroke risk have been identified in the intergenic region between Forkhead family transcription factors FOXF2 and FOXQ1, but we lack a mechanism for the association. FoxF2 is expressed in vascular mural pericytes and is important for maintaining pericyte number and stabilizing small vessels in zebrafish. The stroke-associated SNPs are located in a previously unknown transcriptional enhancer for FOXF2, functional in human cells and zebrafish. We identify critical enhancer regions for FOXF2 gene expression, including binding sites occupied by transcription factors ETS1, RBPJ, and CTCF. rs74564934, a stroke-associated SNP adjacent to the ETS1 binding site, decreases enhancer function, as does mutation of RPBJ sites. rs74564934 is significantly associated with the increased risk of any stroke, ischemic stroke, small vessel stroke, and elevated white matter hyperintensity burden in humans. Foxf2 has a conserved function cross-species and is expressed in vascular mural pericytes of the vessel wall. Thus, stroke-associated SNPs modulate enhancer activity and expression of a regulator of vascular stabilization, FOXF2, thereby modulating stroke risk.

LRF Promotes Indirectly Advantageous Chromatin Conformation via BGLT3-lncRNA Expression and Switch from Fetal to Adult Hemoglobin.

The hemoglobin switch from fetal (HbF) to adult (HbA) has been studied intensively as an essential model for gene expression regulation, but also as a beneficial therapeutic approach for ß-hemoglobinopathies, towards the objective of reactivating HbF. The transcription factor LRF (Leukemia/lymphoma-related), encoded from the ZBTB7A gene has been implicated in fetal hemoglobin silencing, though has a wide range of functions that have not been fully clarified. We thus established the LRF/ZBTB7A-overexpressing and ZBTB7A-knockdown K562 (human erythroleukemia cell line) clones to assess fetal vs. adult hemoglobin production pre- and post-induction. Transgenic K562 clones were further developed and studied under the influence of epigenetic chromatin regulators, such as DNA methyl transferase 3 (DNMT3) and Histone Deacetylase 1 (HDAC1), to evaluate LRF's potential disturbance upon the aberrant epigenetic background and provide valuable information of the preferable epigenetic frame, in which LRF unfolds its action on the ß-type globin's expression. The ChIP-seq analysis demonstrated that LRF binds to γ-globin genes (HBG2/1) and apparently associates BCL11A for their silencing, but also during erythropoiesis induction, LRF binds the BGLT3 gene, promoting BGLT3-lncRNA production through the γ-δ intergenic region of ß-type globin's locus, triggering the transcriptional events from γ- to ß-globin switch. Our findings are supported by an up-to-date looping model, which highlights chromatin alterations during erythropoiesis at late stages of gestation, to establish an "open" chromatin conformation across the γ-δ intergenic region and accomplish ß-globin expression and hemoglobin switch.

Association of mutation signature effectuating processes with mutation hotspots in driver genes and non-coding regions.

Cancer driving mutations are difficult to identify especially in the non-coding part of the genome. Here, we present sigDriver, an algorithm dedicated to call driver mutations. Using 3813 whole-genome sequenced tumors from International Cancer Genome Consortium, The Cancer Genome Atlas Program, and a childhood pan-cancer cohort, we employ mutational signatures based on single-base substitution in the context of tri- and penta-nucleotide motifs for hotspot discovery. Knowledge-based annotations on mutational hotspots reveal enrichment in coding regions and regulatory elements for 6 mutational signatures, including APOBEC and somatic hypermutation signatures. APOBEC activity is associated with 32 hotspots of which 11 are known and 11 are putative regulatory drivers. Somatic single nucleotide variants clusters detected at hypermutation-associated hotspots are distinct from translocation or gene amplifications. Patients carrying APOBEC induced PIK3CA driver mutations show lower occurrence of signature SBS39. In summary, sigDriver uncovers mutational processes associated with known and putative tumor drivers and hotspots particularly in the non-coding regions of the genome.

Nearest-Neighbor Effects Modulate loxP Spacer DNA Chemical Shifts and Guide Oligonucleotide Design for Nuclear Magnetic Resonance Studies.

The Cre-loxP gene editing tool enables site-specific editing of DNA without leaving lesions that must be repaired by error-prone cellular processes. Cre recombines two 34-bp loxP DNA sites that feature a pair of palindromic recombinase-binding elements flanking an asymmetric 8-bp spacer region, via assembly of a tetrameric intasome complex and formation of a Holliday junction intermediate. Recombination proceeds by coordinated nucleophilic attack by pairs of catalytic tyrosine residues on specific phosphodiester bonds in the spacer regions of opposing strands. Despite not making base-specific contacts with the asymmetric spacer region of the DNA, Cre exhibits a preference for initial cleavage on one of the strands, suggesting that intrinsic properties of the uncontacted 8-bp spacer region give rise to this preference. Furthermore, little is known about the structural and dynamic features of the loxP spacer that make it a suitable target for Cre. To enable NMR spectroscopic studies of the spacer, we have aimed to identify a fragment of the 34-bp loxP site that retains the structural features of the spacer while minimizing the spectral crowding and line-broadening seen in longer oligonucleotides. Sequence-specific chemical shift differences between spacer oligos of different lengths, and of a mutant that inverts strand cleavage order, reveal how both nearest-neighbor and next-nearest-neighbor effects dominate the chemical environment experienced by the spacer. We have identified a 16-bp oligonucleotide that preserves the structural environment of the spacer, setting the stage for NMR-based structure determination and dynamics investigations.

Regulation of closely juxtaposed proto-oncogene c-fms and HMGXB3 gene expression by mRNA 3' end polymorphism in breast cancer cells.

Sense-antisense mRNA pairs generated by convergent transcription is a way of gene regulation. c-fms gene is closely juxtaposed to the HMGXB3 gene in the opposite orientation, in chromosome 5. The intergenic region (IR) between c-fms and HMGXB3 genes is 162 bp. We found that a small portion (∼4.18%) of HMGXB3 mRNA is transcribed further downstream, including the end of the c-fms gene generating antisense mRNA against c-fms mRNA. Similarly, a small portion (∼1.1%) of c-fms mRNA is transcribed further downstream, including the end of the HMGXB3 gene generating antisense mRNA against the HMGXB3 mRNA. Insertion of the strong poly(A) signal sequence in the IR results in decreased c-fms and HMGXB3 antisense mRNAs, resulting in up-regulation of both c-fms and HMGXB3 mRNA expression. miR-324-5p targets HMGXB3 mRNA 3' UTR, and as a result, regulates c-fms mRNA expression. HuR stabilizes c-fms mRNA, and as a result, down-regulates HMGXB3 mRNA expression. UALCAN analysis indicates that the expression pattern between c-fms and HMGXB3 proteins are opposite in vivo in breast cancer tissues. Together, our results indicate that the mRNA encoded by the HMGXB3 gene can influence the expression of adjacent c-fms mRNA, or vice versa.

Resurrection of a Viral Internal Ribosome Entry Site from a 700 Year Old Ancient Northwest Territories Cripavirus.

The dicistrovirus intergenic region internal ribosome entry site (IGR IRES) uses an unprecedented, streamlined mechanism whereby the IRES adopts a triple-pseudoknot (PK) structure to directly bind to the conserved core of the ribosome and drive translation from a non-AUG codon. The origin of this IRES mechanism is not known. Previously, a partial fragment of a divergent dicistrovirus RNA genome, named ancient Northwest territories cripavirus (aNCV), was extracted from 700-year-old caribou feces trapped in a subarctic ice patch. The aNCV IGR sequence adopts a secondary structure similar to contemporary IGR IRES structures, however, there are subtle differences including 105 nucleotides upstream of the IRES of unknown function. Using filter binding assays, we showed that the aNCV IRES could bind to purified ribosomes, and toeprinting analysis pinpointed the start site at a GCU alanine codon adjacent to PKI. Using a bicistronic reporter RNA, the aNCV IGR can direct translation in vitro in a PKI-dependent manner. Lastly, a chimeric infectious clone swapping in the aNCV IRES supported translation and virus infection. The characterization and resurrection of a functional IGR IRES from a divergent 700-year-old virus provides a historical framework for the importance of this viral translational mechanism.

The Halastavi árva Virus Intergenic Region IRES Promotes Translation by the Simplest Possible Initiation Mechanism.

Dicistrovirus intergenic region internal ribosomal entry sites (IGR IRESs) do not require initiator tRNA, an AUG codon, or initiation factors and jumpstart translation from the middle of the elongation cycle via formation of IRES/80S complexes resembling the pre-translocation state. eEF2 then translocates the [codon-anticodon]-mimicking pseudoknot I (PKI) from ribosomal A sites to P sites, bringing the first sense codon into the decoding center. Halastavi árva virus (HalV) contains an IGR that is related to previously described IGR IRESs but lacks domain 2, which enables these IRESs to bind to individual 40S ribosomal subunits. By using in vitro reconstitution and cryoelectron microscopy (cryo-EM), we now report that the HalV IGR IRES functions by the simplest initiation mechanism that involves binding to 80S ribosomes such that PKI is placed in the P site, so that the A site contains the first codon that is directly accessible for decoding without prior eEF2-mediated translocation of PKI.

In silico characterisation of stand-alone response regulators of Streptococcus pyogenes.

Bacterial "stand-alone" response regulators (RRs) are pivotal to the control of gene transcription in response to changing cytosolic and extracellular microenvironments during infection. The genome of group A Streptococcus (GAS) encodes more than 30 stand-alone RRs that orchestrate the expression of virulence factors involved in infecting multiple tissues, so causing an array of potentially lethal human diseases. Here, we analysed the molecular epidemiology and biological associations in the coding sequences (CDSs) and upstream intergenic regions (IGRs) of 35 stand-alone RRs from a collection of global GAS genomes. Of the 944 genomes analysed, 97% encoded 32 or more of the 35 tested RRs. The length of RR CDSs ranged from 297 to 1587 nucleotides with an average nucleotide diversity (π) of 0.012, while the IGRs ranged from 51 to 666 nucleotides with average π of 0.017. We present new evidence of recombination in multiple RRs including mga, leading to mga-2 switching, emm-switching and emm-like gene chimerization, and the first instance of an isolate that encodes both mga-1 and mga-2. Recombination was also evident in rofA/nra and msmR loci with 15 emm-types represented in multiple FCT (fibronectin-binding, collagen-binding, T-antigen)-types, including novel emm-type/FCT-type pairings. Strong associations were observed between concatenated RR allele types, and emm-type, MLST-type, core genome phylogroup, and country of sampling. No strong associations were observed between individual loci and disease outcome. We propose that 11 RRs may form part of future refinement of GAS typing systems that reflect core genome evolutionary associations. This subgenomic analysis revealed allelic traits that were informative to the biological function, GAS strain definition, and regional outbreak detection.

Prioritizing sequence variants in conserved non-coding elements in the chicken genome using chCADD.

The availability of genomes for many species has advanced our understanding of the non-protein-coding fraction of the genome. Comparative genomics has proven itself to be an invaluable approach for the systematic, genome-wide identification of conserved non-protein-coding elements (CNEs). However, for many non-mammalian model species, including chicken, our capability to interpret the functional importance of variants overlapping CNEs has been limited by current genomic annotations, which rely on a single information type (e.g. conservation). We here studied CNEs in chicken using a combination of population genomics and comparative genomics. To investigate the functional importance of variants found in CNEs we develop a ch(icken) Combined Annotation-Dependent Depletion (chCADD) model, a variant effect prediction tool first introduced for humans and later on for mouse and pig. We show that 73 Mb of the chicken genome has been conserved across more than 280 million years of vertebrate evolution. The vast majority of the conserved elements are in non-protein-coding regions, which display SNP densities and allele frequency distributions characteristic of genomic regions constrained by purifying selection. By annotating SNPs with the chCADD score we are able to pinpoint specific subregions of the CNEs to be of higher functional importance, as supported by SNPs found in these subregions are associated with known disease genes in humans, mice, and rats. Taken together, our findings indicate that CNEs harbor variants of functional significance that should be object of further investigation along with protein-coding mutations. We therefore anticipate chCADD to be of great use to the scientific community and breeding companies in future functional studies in chicken.

Structural Determinants within the Adenovirus Early Region 1A Protein Spacer Region Necessary for Tumorigenesis.

It has long been established that group A human adenoviruses (HAdV-A12, -A18, and -A31) can cause tumors in newborn rodents, with tumorigenicity related to the presence of a unique spacer region located between conserved regions 2 and 3 within the HAdV-A12 early region 1A (E1A) protein. Group B adenoviruses are weakly oncogenic, whereas most of the remaining human adenoviruses are nononcogenic. In an attempt to understand better the relationship between the structure of the AdE1A spacer region and oncogenicity of HAdVs, the structures of synthetic peptides identical or very similar to the adenovirus 12 E1A spacer region were determined and found to be α-helical using nuclear magnetic resonance (NMR) spectroscopy. This contrasts significantly with some previous suggestions that this region is unstructured. Using available predictive algorithms, the structures of spacer regions from other E1As were also examined, and the extent of the predicted α-helix was found to correlate reasonably well with the tumorigenicity of the respective virus. We suggest that this may represent an as-yet-unknown binding site for a partner protein required for tumorigenesis.IMPORTANCE This research analyzed small peptides equivalent to a region within the human adenovirus early region 1A protein that confers, in part, tumor-inducing properties to various degrees on several viral strains in rats and mice. The oncogenic spacer region is α-helical, which contrasts with previous suggestions that this region is unstructured. The helix is characterized by a stretch of amino acids rich in alanine residues that are organized into a hydrophobic, or "water-hating," surface that is considered to form a major site of interaction with cellular protein targets that mediate tumor formation. The extent of α-helix in E1A from other adenovirus species can be correlated to a limited degree to the tumorigenicity of that virus. Some serotypes show significant differences in predicted structural propensity, suggesting that the amino acid type and physicochemical properties are also of importance.

Binding of a viral IRES to the 40S subunit occurs in two successive steps mediated by eS25.

The mechanism for how internal ribosome entry sites (IRESs) recruit ribosomes to initiate translation of an mRNA is not completely understood. We investigated how a 40S subunit was recruited by the cricket paralysis virus intergenic region (CrPV IGR) IRES to form a stable 40S-IRES complex. Kinetic binding studies revealed that formation of the complex between the CrPV IGR and the 40S subunit consisted of two-steps: an initial fast binding step of the IRES to the 40S ribosomal subunit, followed by a slow unimolecular reaction consistent with a conformational change that stabilized the complex. We further showed that the ribosomal protein S25 (eS25), which is required by functionally and structurally diverse IRESs, impacts both steps of the complex formation. Mutations in eS25 that reduced CrPV IGR IRES activity either decreased 40S-IRES complex formation, or increased the rate of the conformational change that was required to form a stable 40S-IRES complex. Our data are consistent with a model in which eS25 facilitates initial binding of the CrPV IGR IRES to the 40S while ensuring that the conformational change stabilizing the 40S-IRES complex does not occur prematurely.

Analysis of accessible chromatin landscape in the inner cell mass and trophectoderm of human blastocysts.

Early embryonic development is characterized by drastic changes in chromatin structure that affects the accessibility of the chromatin. In human, the chromosome reorganization and its involvement in the first linage segregation are poorly characterized due to the difficulties in obtaining human embryonic material and limitation on low input technologies. In this study, we aimed to explore the chromatin remodeling pattern in human preimplantation embryos and gain insight into the epigenetic regulation of inner cell mass (ICM) and trophectoderm (TE) differentiation. We optimized ATAC-seq (an assay for transposase-accessible chromatin using sequencing) to analyze the chromatin accessibility landscape for low DNA input. Sixteen preimplantation human blastocysts frozen on Day 6 were used. Our data showed that ATAC peak distributions of the promoter regions (<1 kb) and distal regions versus other regions were significantly different between ICM versus TE samples (P < 0.01). We detected that a higher percentage of accessible binding loci were located within 1 kb of the transcription start site in ICM compared to TE (P < 0.01). However, a higher percentage of accessible regions was detected in the distal region of TE compared to ICM (P < 0.01). In addition, eight differential peaks with a false discovery rate <0.05 between ICM and TE were detected. This is the first study to compare the landscape of the accessible chromatin between ICM and TE of human preimplantation embryos, which unveiled chromatin-level epigenetic regulation of cell lineage specification in early embryo development.

The histone mark H3K36me2 recruits DNMT3A and shapes the intergenic DNA methylation landscape.

Enzymes that catalyse CpG methylation in DNA, including the DNA methyltransferases 1 (DNMT1), 3A (DNMT3A) and 3B (DNMT3B), are indispensable for mammalian tissue development and homeostasis1-4. They are also implicated in human developmental disorders and cancers5-8, supporting the critical role of DNA methylation in the specification and maintenance of cell fate. Previous studies have suggested that post-translational modifications of histones are involved in specifying patterns of DNA methyltransferase localization and DNA methylation at promoters and actively transcribed gene bodies9-11. However, the mechanisms that control the establishment and maintenance of intergenic DNA methylation remain poorly understood. Tatton-Brown-Rahman syndrome (TBRS) is a childhood overgrowth disorder that is defined by germline mutations in DNMT3A. TBRS shares clinical features with Sotos syndrome (which is caused by haploinsufficiency of NSD1, a histone methyltransferase that catalyses the dimethylation of histone H3 at K36 (H3K36me2)8,12,13), which suggests that there is a mechanistic link between these two diseases. Here we report that NSD1-mediated H3K36me2 is required for the recruitment of DNMT3A and maintenance of DNA methylation at intergenic regions. Genome-wide analysis shows that the binding and activity of DNMT3A colocalize with H3K36me2 at non-coding regions of euchromatin. Genetic ablation of Nsd1 and its paralogue Nsd2 in mouse cells results in a redistribution of DNMT3A to H3K36me3-modified gene bodies and a reduction in the methylation of intergenic DNA. Blood samples from patients with Sotos syndrome and NSD1-mutant tumours also exhibit hypomethylation of intergenic DNA. The PWWP domain of DNMT3A shows dual recognition of H3K36me2 and H3K36me3 in vitro, with a higher binding affinity towards H3K36me2 that is abrogated by TBRS-derived missense mutations. Together, our study reveals a trans-chromatin regulatory pathway that connects aberrant intergenic CpG methylation to human neoplastic and developmental overgrowth.