Specific Transcription Factor Binding Research Articles

DeepCTF: transcription factor binding specificity prediction using DNA sequence plus shape in an attention-based deep learning model

DeepCTF: transcription factor binding specificity prediction using DNA sequence plus shape in an attention-based deep learning model

MulTFBS: A Spatial-Temporal Network with Multichannels for Predicting Transcription Factor Binding Sites.

Revealing the mechanisms that influence transcription factor binding specificity is the key to understanding gene regulation. In previous studies, DNA double helix structure and one-hot embedding have been used successfully to design computational methods for predicting transcription factor binding sites (TFBSs). However, DNA sequence as a kind of biological language, the method of word embedding representation in natural language processing, has not been considered properly in TFBS prediction models. In our work, we integrate different types of features of DNA sequence to design a multichanneled deep learning framework, namely MulTFBS, in which independent one-hot encoding, word embedding encoding, which can incorporate contextual information and extract the global features of the sequences, and double helix three-dimensional structural features have been trained in different channels. To extract sequence high-level information effectively, in our deep learning framework, we select the spatial-temporal network by combining convolutional neural networks and bidirectional long short-term memory networks with attention mechanism. Compared with six state-of-the-art methods on 66 universal protein-binding microarray data sets of different transcription factors, MulTFBS performs best on all data sets in the regression tasks, with the average R2 of 0.698 and the average PCC of 0.833, which are 5.4% and 3.2% higher, respectively, than the suboptimal method CRPTS. In addition, we evaluate the classification performance of MulTFBS for distinguishing bound or unbound regions on TF ChIP-seq data. The results show that our framework also performs well in the TFBS classification tasks.

Deciphering the spatiotemporal transcriptional and chromatin accessibility of human retinal organoid development at the single-cell level

Molecular information on the early stages of human retinal development remains scarce due to limitations in obtaining early human eye samples. Pluripotent stem cell-derived retinal organoids (ROs) provide an unprecedented opportunity for studying early retinogenesis. Using a combination of single cell RNA-seq and spatial transcriptomics we present for the first-time a single cell spatiotemporal transcriptome of RO development. Our data demonstrate that ROs recapitulate key events of retinogenesis including optic vesicle/cup formation, presence of a putative ciliary margin zone, emergence of retinal progenitor cells and their orderly differentiation to retinal neurons. Combining the scRNA- with scATAC-seq data, we were able to reveal cell-type specific transcription factor binding motifs on accessible chromatin at each stage of organoid development, and to show that chromatin accessibility is highly correlated to the developing human retina, but with some differences in the temporal emergence and abundance of some of the retinal neurons.

Locus-level L1 DNA methylation profiling reveals the epigenetic and transcriptional interplay between L1s and their integration sites

Long interspersed element 1 (L1) retrotransposons are implicated in human disease and evolution. Their global activity is repressed by DNA methylation, but deciphering the regulation of individual copies has been challenging. Here, we combine short- and long-read sequencing to unveil L1 methylation heterogeneity across cell types, families, and individual loci and elucidate key principles involved. We find that the youngest primate L1 families are specifically hypomethylated in pluripotent stem cells and the placenta but not in most tumors. Locally, intronic L1 methylation is intimately associated with gene transcription. Conversely, the L1 methylation state can propagate to the proximal region up to 300bp. This phenomenon is accompanied by the binding of specific transcription factors, which drive the expression of L1 and chimeric transcripts. Finally, L1 hypomethylation alone is typically insufficient to trigger L1 expression due to redundant silencing pathways. Our results illuminate the epigenetic and transcriptional interplay between retrotransposons and their host genome.

ChEC-seq2: an improved chromatin endogenous cleavage sequencing method and bioinformatic analysis pipeline for mapping in vivo protein-DNA interactions.

Defining the in vivo DNA binding specificity of transcription factors (TFs) has relied nearly exclusively on chromatin immunoprecipitation (ChIP). While ChIP reveals TF binding patterns, its resolution is low. Higher resolution methods employing nucleases such as ChIP-exo, chromatin endogenous cleavage (ChEC-seq) and CUT&RUN resolve both TF occupancy and binding site protection. ChEC-seq, in which an endogenous TF is fused to micrococcal nuclease, requires neither fixation nor antibodies. However, the specificity of DNA cleavage during ChEC has been suggested to be lower than the specificity of the peaks identified by ChIP or ChIP-exo, perhaps reflecting non-specific binding of transcription factors to DNA. We have simplified the ChEC-seq protocol to minimize nuclease digestion while increasing the yield of cleaved DNA. ChEC-seq2 cleavage patterns were highly reproducible between replicates and with published ChEC-seq data. Combined with DoubleChEC, a new bioinformatic pipeline that removes non-specific cleavage sites, ChEC-seq2 identified high-confidence cleavage sites for three different yeast TFs that are strongly enriched for their known binding sites and adjacent to known target genes.

Epigenomic profiling nominates master transcription factors (TFs) driving sarcomatoid differentiation (SD) of renal cell carcinoma (RCC)

Abstract Background Sarcomatoid differentiation of RCC (sRCC) is associated with poor survival. Recent studies showed marked response of sRCC to immune checkpoint blockade (ICB). While distinctive patterns of gene expression in sRCC have been identified, the gene regulatory programs and TFs that drive SD remain unknown. The aim of this study is to nominate TFs responsible for SD and to investigate their association with the clinical outcomes of patients with RCC. Methods Chromatin immunoprecipitation and sequencing (ChIP-seq) for H3K27ac – a histone modification associated with active regulatory elements – was performed on pathologically reviewed sRCC and non-sRCC samples collected at the Dana-Farber Cancer Institute. Regulatory elements that were differentially active between the two groups were identified based on levels of H3K27ac (Benjamini-Hochberg q<0.01, log-fold change (LFC) threshold=1). Enrichment of specific TF binding motifs at activated regulatory elements in sRCC was assessed using HOMER. Differential gene expression analysis of TFs was performed using DESeq2 on RNA-seq data from TCGA. A Mann-Whitney U test was performed on RNA-seq data from the IMmotion151 and Javelin Renal 101 clinical trials to compare mean expression level of TFs in transcript per million (TPM) in these trials. Patients with non-sRCC enrolled in the IMmotion151 trials were divided into quartiles based on gene expression levels of candidate TFs. Progression-free survival (PFS) was compared between non-sRCC patients stratified by expression quartiles as well as patients with sRCC using a multivariable Cox proportional-hazards model accounting for age and IMDC risk score. To validate these findings, a similar analysis was performed in the Javelin Renal 101 trial. Results We obtained high-quality H3K27ac ChIP-seq profiles for 9 sRCC and 17 non-sRCC samples. We identified 278 candidate regulatory elements with increased H3K27ac levels in sRCC vs. non-sRCC. These regulatory elements were enriched for nucleotide motifs bound by the TFs FOSL1 and E2F7. Differential expression analysis between 48 sRCC vs. 493 non-sRCC samples showed that FOSL1 and E2F7 were significantly upregulated in sRCC vs. non-sRCC (LFC=1.7, q=5e-11; LFC=1.8, q=1.3e-20; resp.). Mean TPMs of FOSL1 and E2F7 were significantly increased in sRCC vs. non-sRCC in IMmotion151 cohort and Javelin Renal 101 (all p<0.001). Among patients who received sunitinib, those with the highest quartile of FOSL1 and E2F7 expression showed significantly shorter PFS in IMmotion151 patients compared to patients with the lowest quartile of expression (HR=1.6, 95%CI=1.3-2.2, p=0.008 & HR=2.6, 95%CI=1.8-3.7, p<0.001; resp.). Furthermore, patients with highest quartile of expression showed similar PFS compared to patients with sRCC (p=0.56 and p=0.64; resp.). These results were validated in the sunitinib arm of the Javelin Renal 101 cohort (Figure). Figure: Kaplan-Meier curves of progression-free survival (PFS) in patients with in the sunitinib arm of Javelin Renal 101 by sarcomatoid differentiation and FOSL1 expression levels. Conclusions This is the first study to characterize the epigenomic landscape of sRCC by integrating ChIP-seq and RNA-seq data. Our findings implicated FOSL1 and E2F7 as transcriptional regulators of SD with prognostic relevance. These TFs seems to be associated with aggressive behavior in non-sRCC as well. Further studies are underway to functionally validate these results.

Abstract A029: Epigenomic profiling nominates master transcription factors (TFs) driving sarcomatoid differentiation of renal cell carcinoma (RCC)

Abstract Introduction Sarcomatoid RCC (sRCC) is associated with poor prognosis. Recent studies showed promising response rates of sRCC to immune checkpoint blockade (ICB). Although distinct patterns of gene expression in sRCC have been reported, the gene regulatory programs that drive SD remain unknown. This study aims to nominate TFs driving SD and to evaluate their impact on the clinical outcomes of patients (pts) with RCC. Methods Chromatin immunoprecipitation and sequencing (ChIP-seq) for H3K27ac was performed on histologically annotated areas of sRCC and non-sRCC fresh frozen tissue samples collected at Dana-Farber Cancer Institute. Regulatory elements that were differentially active between the two groups were identified (cutoff p<0.01, log-fold change (LFC) threshold=1). Enrichment of specific TF binding motifs at activated regulatory elements in sRCC was assessed using HOMER. Differential gene expression analysis of TFs was performed using DESeq2 on RNA-seq data from TCGA (Benjamini-Hochberg q <0.01, LFC threshold=1). An unpaired student’s t-test was performed to compare mean gene expression levels in transcript per million (TPM) on baseline sRCC and non-sRCC tissue samples from the IMmotion151 trial (sunitinib vs. atezolizumab-bevacizumab in metastatic RCC). Pts enrolled in the IMmotion151 trial were divided into quartiles based on gene expression levels of candidate TFs. Progression-free survival (PFS) was compared between the quartiles using a multivariable Cox proportional-hazards model accounting for age and IMDC risk score. The association of TF expression with the objective radiographic response (ORR) was analyzed using a logistic regression model. Results ChIP-seq was performed on 9 sRCC & 17 non-sRCC samples. We identified 278 candidate regulatory elements with increased H3K27ac levels in sRCC vs. non-sRCC that were enriched for nucleotide motifs bound by the TFs FOSL1 and E2F7. Differential expression analysis between 48 sRCC vs. 493 non-sRCC samples showed that FOSL1 & E2F7 were significantly upregulated in sRCC vs. non-sRCC (LFC=1.7, q=5e-11; LFC=1.8, q=1.3e-20; resp.). Mean TPMs of FOSL1 & E2F7 were significantly increased in sRCC vs. non-sRCC in IMmotion151 cohort (2.5 vs. 1.5 TPM for FOSL1, and 2.9 vs. 1.9 TPM for E2F7, all p<0.001). Pts in the highest quartile of FOSL1 and E2F7 expression showed significantly shorter PFS in IMmotion151 pts receiving sunitinib alone (HR=1.6, 95%CI=1.3-2.2, p=0.008 & HR=2.6, 95%CI=1.8-3.7, p<0.001; resp.). We observed similar results even after controlling for the presence of SD. Increased expression of E2F7 was associated with worse ORR in IMmotion151 pts who received sunitinib alone, even after controlling for the presence of SD (OR=0.24±0.4, p<0.001). Conclusion This is the first study to analyze the epigenomic landscape of sRCC by integrating ChIP-seq and RNA-seq data. Our findings implicated FOSL1 & E2F7 as transcriptional regulators of SD with prognostic relevance. Further studies are underway to validate these results. Citation Format: Karl Semaan, Talal El Zarif, Marc Eid, Valisha Shah, Brad Fortunato, Renee Maria Saliby, Amin H. Nassar, Sarah Abou Alaiwi, Ji-Heui Seo, Chris Labaki, Ziad Bakouny, Sayed Matar, Nourhan El Ahmar, Yasmin Nabil Laimon, Gwo-Shu Mary Lee, Mark Pomerantz, Jacob Berchuck, Sabina Signoretti, Eliezer Van Allen, Toni Choueiri, David Braun, Matthew Freedman, Sylvan Baca. Epigenomic profiling nominates master transcription factors (TFs) driving sarcomatoid differentiation of renal cell carcinoma (RCC) [abstract]. In: Proceedings of the AACR Special Conference: Advances in Kidney Cancer Research; 2023 Jun 24-27; Austin, Texas. Philadelphia (PA): AACR; Cancer Res 2023;83(16 Suppl):Abstract nr A029.

Gene expressions associated with longer lifespan and aging exhibit similarity in mammals.

Although molecular features underlying aging and species maximum lifespan (MLS) have been comprehensively studied by transcriptome analyses, the actual impact of transcriptome on aging and MLS remains elusive. Here, we found that transcriptional signatures that are associated with mammalian MLS exhibited significant similarity to those of aging. Moreover, transcriptional signatures of longer MLS and aging both exhibited significant similarity to that of longer-lived mouse strains, suggesting that gene expression patterns associated with species MLS contribute to extended lifespan even within a species and that aging-related gene expression changes overall represent adaptations that extend lifespan rather than deterioration. Finally, we found evidence of co-evolution of MLS and promoter sequences of MLS-associated genes, highlighting the evolutionary contribution of specific transcription factor binding motifs such as that of E2F1 in shaping MLS-associated gene expression signature. Our results highlight the importance of focusing on adaptive aspects of aging transcriptome and demonstrate that cross-species genomics can be a powerful approach for understanding adaptive aging transcriptome.

Comparative analysis reveals epigenomic evolution related to species traits and genomic imprinting in mammals

DNA methylation is an epigenetic modification that plays a crucial role in various regulatory processes, including gene expression regulation, transposable element repression, and genomic imprinting. However, most studies on DNA methylation have been conducted in humans and other model species, whereas the dynamics of DNA methylation across mammals remain poorly explored, limiting our understanding of epigenomic evolution in mammals and the evolutionary impacts of conserved and lineage-specific DNA methylation. Here, we generated and gathered comparative epigenomic data from 13 mammalian species, including two marsupial species, to demonstrate that DNA methylation plays critical roles in several aspects of gene evolution and species trait evolution. We found that the species-specific DNA methylation of promoters and noncoding elements correlates with species-specific traits such as body patterning, indicating that DNA methylation might help establish or maintain interspecies differences in gene regulation that shape phenotypes. For a broader view, we investigated the evolutionary histories of 88 known imprinting control regions across mammals to identify their evolutionary origins. By analyzing the features of known and newly identified potential imprints in all studied mammals, we found that genomic imprinting may function in embryonic development through the binding of specific transcription factors. Our findings show that DNA methylation and the complex interaction between the genome and epigenome have a significant impact on mammalian evolution, suggesting that evolutionary epigenomics should be incorporated to develop a unified evolutionary theory.

Complementary strategies for directing in vivo transcription factor binding through DNA binding domains and intrinsically disordered regions

DNA binding domains (DBDs) of transcription factors (TFs) recognize DNA sequence motifs that are highly abundant in genomes. Within cells, TFs bind a subset of motif-containing sites as directed by either their DBDs or DBD-external (nonDBD) sequences. To define the relative roles of DBDs and nonDBDs in directing binding preferences, we compared the genome-wide binding of 48 (∼30%) budding yeast TFs with their DBD-only, nonDBD-truncated, and nonDBD-only mutants. With a few exceptions, binding locations differed between DBDs and TFs, resulting from the cumulative action of multiple determinants mapped mostly to disordered nonDBD regions. Furthermore, TFs' preferences for promoters of the fuzzy nucleosome architecture were lost in DBD-only mutants, whose binding spread across promoters, implicating nonDBDs' preferences in this hallmark of budding yeast regulatory design. We conclude that DBDs and nonDBDs employ complementary DNA-targeting strategies, whose balance defines TF binding specificity along genomes.

Chromatin accessibility landscapes revealed the subgenome-divergent regulation networks during wheat grain development

Development of wheat (Triticum aestivum L.) grain mainly depends on the processes of starch synthesis and storage protein accumulation, which are critical for grain yield and quality. However, the regulatory network underlying the transcriptional and physiological changes of grain development is still not clear. Here, we combined ATAC-seq and RNA-seq to discover the chromatin accessibility and gene expression dynamics during these processes. We found that the chromatin accessibility changes are tightly associated with differential transcriptomic expressions, and the proportion of distal ACRs was increased gradually during grain development. Specific transcription factor (TF) binding sites were enriched at different stages and were diversified among the 3 subgenomes. We further predicted the potential interactions between key TFs and genes related with starch and storage protein biosynthesis and found different copies of some key TFs played diversified roles. Overall, our findings have provided numerous resources and illustrated the regulatory network during wheat grain development, which would shed light on the improvement of wheat yields and qualities.

Target search dynamics of Sox transcription factors.

Transcription factors (TFs) are central players in regulating gene expression through their ability to bind sequence-specific sites. Nonetheless, how TFs search the genome to locate their binding sites is unclear. The facilitated diffusion theory proposes that TFs combine 1D-sliding on DNA and 3D-diffusion in the nucleus, and that local nonspecific interactions mediated by electrostatic forces drive the search efficiency beyond the one that would be reached by free diffusion alone. While such models are descriptive of the situation in prokaryotes, how eukaryotic TFs search the chromatinized genome is poorly understood. Furthermore, the relevance of regions outside DNA-binding domains (DBDs) for the search process remains unknown. Herein, we hypothesize that positive charges in DBD flanking regions accelerate TF target search through nonspecific interactions with DNA. A comparative study between Sox2 and Sox17 can demonstrate how electric charges within intrinsically disordered regions (IDRs) affect the target search process. These Sox-TF recognize the same motifs but have opposite charge distributions at their DBD C-terminus. Single-molecule imaging techniques allow for investigating the search efficiency. our results show that Sox2 exhibited a two-fold higher pseudo-on-rate than Sox17 and a Sox2 variant having an acidic stretch of Sox17. This is consistent with in vitro measurements, directly reporting a two-fold higher on-rate for wild-type Sox2 compared to the negatively-charged Sox2 variant. Moreover, the Sox2 variant showed a decreased ability to bind target sites within compact chromatin. Since Sox2 is known as a pioneer factor, quantifying their interactions with synthetically-chromatinized DNA will elucidate search mechanisms of pioneer TFs. In this regard, we foresee that more positively-charged IDRs stabilize the specific binding of TFs to nucleosomal DNA and help them to invade chromatin. Altogether, this work will dissect how the charge distribution of DBD flanking regions modulates TF target search on chromatin.

GNet: An integrated context-aware neural framework for transcription factor binding signal at single nucleotide resolution prediction.

Transcription factors (TFs) are important factors that regulate gene expression. Revealing the mechanism affecting the binding specificity of TFs is the key to understanding gene regulation. Most of the previous studies focus on TF-DNA binding sites at the sequence level, and they seldom utilize the contextual features of DNA sequences. In this paper, we develop an integrated spatiotemporal context-aware neural network framework, named GNet, for predicting TF-DNA binding signal at single nucleotide resolution by achieving three tasks: single nucleotide resolution signal prediction, identification of binding regions at the sequence level, and TF-DNA binding motif prediction. GNet extracts implicit spatial contextual information with a gated highway neural mechanism, which captures large context multi-level patterns using linear shortcut connections, and the idea of it permeates the encoder and decoder parts of GNet. The improved dual external attention mechanism, which learns implicit relationships both within and among samples, and improves the performance of the model. Experimental results on 53 human TF ChIP-seq datasets and 6 chromatin accessibility ATAC-seq datasets shows that GNet outperforms the state-of-the-art methods in the three tasks, and the results of cross-species studies on 15 human and 18 mouse TF datasets of the corresponding TF families indicate that GNet also shows the best performance in cross-species prediction over the competitive methods.

Uncovering the gene regulatory network of type 2 diabetes through multi-omic data integration

BackgroundType 2 diabetes (T2D) onset is a complex, organized biological process with multilevel regulation, and its physiopathological mechanisms are yet to be elucidated. This study aims to find out the key drivers and pathways involved in the pathogenesis of T2D through multi-omics analysis.MethodsThe datasets used in the experiments comprise three groups: (1) genomic (2) transcriptomic, and (3) epigenomic categories. Then, a series of bioinformatics technologies including Marker set enrichment analysis (MSEA), weighted key driver analysis (wKDA) was performed to identify key drivers. The hub genes were further verified by the Receiver Operator Characteristic (ROC) Curve analysis, proteomic analysis, and Real-time quantitative polymerase chain reaction (RT-qPCR). The multi-omics network was applied to the Pharmomics pipeline in Mergeomics to identify drug candidates for T2D treatment. Then, we used the drug-gene interaction network to conduct network pharmacological analysis. Besides, molecular docking was performed using AutoDock/Vina, a computational docking program.ResultsModule-gene interaction network was constructed using MSEA, which revealed a significant enrichment of immune-related activities and glucose metabolism. Top 10 key drivers (PSMB9, COL1A1, COL4A1, HLA-DQB1, COL3A1, IRF7, COL5A1, CD74, HLA-DQA1, and HLA-DRB1) were selected by wKDA analysis. Among these, COL5A1, IRF7, CD74, and HLA-DRB1 were verified to have the capability to diagnose T2D, and expression levels of PSMB9 and CD74 had significantly higher in T2D patients. We further predict the co-expression network and transcription factor (TF) binding specificity of the key driver. Besides, based on module interaction networks and key driver networks, 17 compounds are considered to possess T2D-control potential, such as sunitinib.ConclusionsWe identified signature genes, biomolecular processes, and pathways using multi-omics networks. Moreover, our computational network analysis revealed potential novel strategies for pharmacologic interventions of T2D.

A variable gene family encoding nodule-specific cysteine-rich peptides in pea (Pisum sativum L.)

Various legume plants form root nodules in which symbiotic bacteria (rhizobia) fix atmospheric nitrogen after differentiation into a symbiotic form named bacteroids. In some legume species, bacteroid differentiation is promoted by defensin-like nodule-specific cysteine-rich (NCR) peptides. NCR peptides have best been studied in the model legume Medicago truncatula Gaertn., while in many other legumes relevant information is still fragmentary. Here, we characterize the NCR gene family in pea (Pisum sativum L.) using genomic and transcriptomic data. We found 360 genes encoding NCR peptides that are expressed in nodules. The sequences of pea NCR genes and putative peptides are highly variable and differ significantly from NCR sequences of M. truncatula. Indeed, only one pair of orthologs (PsNCR47–MtNCR312) has been identified. The NCR genes in the pea genome are located in clusters, and the expression patterns of NCR genes from one cluster tend to be similar. These data support the idea of independent evolution of NCR genes by duplication and diversification in related legume species. We also described spatiotemporal expression profiles of NCRs and identified specific transcription factor (TF) binding sites in promoters of “early” and “late” NCR genes. Further, we studied the expression of NCR genes in nodules of Fix– mutants and predicted potential regulators of NCR gene expression, one among them being the TF ERN1 involved in the early steps of nodule organogenesis. In general, this study contributes to understanding the functions of NCRs in legume nodules and contributes to understanding the diversity and potential antibiotic properties of pea nodule-specific antimicrobial molecules.

SEMplMe: a tool for integrating DNA methylation effects in transcription factor binding affinity predictions

MotivationAberrant DNA methylation in transcription factor binding sites has been shown to lead to anomalous gene regulation that is strongly associated with human disease. However, the majority of methylation-sensitive positions within transcription factor binding sites remain unknown. Here we introduce SEMplMe, a computational tool to generate predictions of the effect of methylation on transcription factor binding strength in every position within a transcription factor’s motif.ResultsSEMplMe uses ChIP-seq and whole genome bisulfite sequencing to predict effects of methylation within binding sites. SEMplMe validates known methylation sensitive and insensitive positions within a binding motif, identifies cell type specific transcription factor binding driven by methylation, and outperforms SELEX-based predictions for CTCF. These predictions can be used to identify aberrant sites of DNA methylation contributing to human disease.Availability and ImplementationSEMplMe is available from https://github.com/Boyle-Lab/SEMplMe.

Target Search Dynamics of Sox Transcription Factors

Here we investigate the impact of regions outside of the DNA‐binding domain (DBD) on target search efficiency of transcription factors (TFs). In living organisms, TFs are key for gene regulation by binding sequence DNA‐binding sites. The facilitated diffusion theory proposes a search process combining TF 1D sliding on DNA and 3D diffusion in the nucleus. Moreover, the local non‐specific TF‐DNA interactions mediated by electrostatic forces enhance the target search efficiency to reach on‐rates faster than allowed by free diffusion. While such theoretical models fit TF behaviors in prokaryotes, how TFs search the chromatinized eukaryotic genome is poorly understood. More importantly, the relevance of regions outside of the DNA‐binding domain for the search process remains to be clarified.We hypothesize that positive charges flanking the DBD accelerate TF target search by interacting non‐specifically with negatively charged DNA. We tackle this hypothesis with a comparative study between Sox2 and Sox17. While recognizing the same DNA motif, Sox2 displays higher non‐specific binding affinity, which we think is caused by a higher content in positive charges close to the DBD.To probe the kinetics of in vivo Sox‐DNA interactions in living mouse embryonic stem cells, we used highly inclined and laminated optical sheet (HiLo) microscopy. We determined that Sox2 and Sox17 have similar profiles of residence time with specific binding of ~ 12 s and non‐specific binding of ~ 0.7 s. We then quantified the search efficiency of Sox2 and Sox17 by determining their pseudo on‐rate, which we define as the time‐averaged number of specifically bound particles per time interval and per TF concentration. Since Sox2 and Sox17 bind to the same DNA motifs, they thus have the same number of targets in the mouse genome. Therefore, the pseudo on‐rate scales linearly with association rate at a given TF concentration. We found that Sox2 has a two‐fold higher pseudo on‐rate than Sox17. We also generated a Sox2 mutant in which its DBD flanking region was swapped with the one from Sox17. This led to a two‐fold decrease in pseudo on‐rate as compared to the wild‐type Sox2. These results thus directly support the hypothesis that non‐specific electrostatic interactions of positively charged residues in regions flanking the DBD enhance their DNA association rates.One limitation of in vivo experiments is that the chromatin context in which TF‐DNA interactions are observed is unknown. To dissect the differential interactions of Sox2 and Sox17 with naked versus nucleosomal DNA, we are using chemically‐synthesized in vitro chromatin combined with single‐molecule tracking via total internal reflection (TIRF) microscopy. Since Sox2 but not Sox17 is known as a pioneer factor, quantifying their interactions with chromatinized DNA will shed light on search mechanisms of pioneer TFs in compact chromatin. In that sense, we foresee that more positively‐charged DBD flanking regions may stabilize specific binding of TF to nucleosomal DNA and help TFs to invade and remodel chromatin.Taken together, our combination of in vivo and in vitro single‐molecule imaging techniques allows dissecting how the charge distribution of DBD flanking regions modulates TF target search on chromatin.

αα-hub coregulator structure and flexibility determine transcription factor binding and selection in regulatory interactomes

Formation of transcription factor (TF)–coregulator complexes is a key step in transcriptional regulation, with coregulators having essential functions as hub nodes in molecular networks. How specificity and selectivity are maintained in these nodes remain open questions. In this work, we addressed specificity in transcriptional networks using complexes formed between TFs and αα-hubs, which are defined by a common αα-hairpin secondary structure motif, as a model. Using NMR spectroscopy and binding thermodynamics, we analyzed the structure, dynamics, stability, and ligand-binding properties of the Arabidopsis thaliana RST domains from TAF4 and known binding partner RCD1, and the TAFH domain from human TAF4, allowing comparison across species, functions, and architectural contexts. While these αα-hubs shared the αα-hairpin motif, they differed in length and orientation of accessory helices as well as in their thermodynamic profiles of ligand binding. Whereas biologically relevant RCD1–ligand pairs displayed high affinity driven by enthalpy, TAF4–ligand interactions were entropy driven and exhibited less binding-induced structuring. We in addition identified a thermal unfolding state with a structured core for all three domains, although the temperature sensitivity differed. Thermal stability studies suggested that initial unfolding of the RCD1–RST domain localized around helix 1, lending this region structural malleability, while effects in TAF4–RST were more stochastic, suggesting variability in structural adaptability upon binding. Collectively, our results support a model in which hub structure, flexibility, and binding thermodynamics contribute to αα-hub–TF binding specificity, a finding of general relevance to the understanding of coregulator–ligand interactions and interactome sizes.

Pioneer transcription factors are associated with the modulation of DNA methylation patterns across cancers

Methylation of cytosines on DNA is a prominent modification associated with gene expression regulation. Aberrant DNA methylation patterns have recurrently been linked to dysregulation of the regulatory program in cancer cells. To shed light on the underlying molecular mechanism driving this process, we hypothesised that aberrant methylation patterns could be controlled by the binding of specific transcription factors (TFs) across cancer types. By combining DNA methylation arrays and gene expression data with TF binding sites (TFBSs), we explored the interplay between TF binding and DNA methylation in 19 cancer types. We performed emQTL (expression–methylation quantitative trait loci) analyses independently in each cancer type and identified 13 TFs whose expression levels are correlated with local DNA methylation patterns around their binding sites in at least 2 cancer types. The 13 TFs are mainly associated with local demethylation and are enriched for pioneer function, suggesting a specific role for these TFs in modulating chromatin structure and transcription in cancer patients. Furthermore, we confirmed that de novo methylation is precluded across cancers at CpGs lying in genomic regions enriched for TF binding signatures associated with SP1, CTCF, NRF1, GABPA, KLF9, and/or YY1. The modulation of DNA methylation associated with TF binding was observed at cis-regulatory regions controlling immune- and cancer-associated pathways, corroborating that the emQTL signals were derived from both cancer and tumor-infiltrating cells. As a case example, we experimentally confirmed that FOXA1 knock-down is associated with higher methylation in regions bound by FOXA1 in breast cancer MCF-7 cells. Finally, we reported physical interactions between FOXA1 with TET1 and TET2 both in an in vitro setup and in vivo at physiological levels in MCF-7 cells, adding further support for FOXA1 attracting TET1 and TET2 to induce local demethylation in cancer cells.

Epigenetic reorganization during early embryonic lineage specification.

Dynamic chromatin reorganization occurs during two waves of cell lineage specification process, blastocyst formation and gastrulation, to generate distinct cell types. Epigenetic defects have been associated with severe developmental defects and diseases. How epigenetic remodeling coordinates the two lineage specification waves is becoming uncovered, benefiting from the development and application of new technologies including low-input or single-cell epigenome analysis approached in the past few years. In this review, we aim to highlight the most recent findings on epigenetic remodeling in cell lineage specification during blastocyst formation and gastrulation. First, we introduce how DNA methylation dynamically changes in blastocyst formation and gastrulation and its function in transcriptional regulation lineage-specific genes. Then, we discuss widespread remodeling of histone modification at promoters and enhancers in orchestrating the trajectory of cell lineage specification. Finally, we review dynamics of chromatin accessibility and 3D structure regulating developmental gene expression and associating with specific transcription factor binding events at stage specific manner. We also highlight the key questions that remain to be answered to fully understand chromatin regulation and reorganization in lineage specification. Here, we summarize the recent advances and discoveries on epigenetic reorganization and its roles in blastocyst formation and gastrulation, and how it cooperates with the lineage specification, painting from global sequencing data from mouse in vivo tissues.