Cell Type-specific Binding Research Articles

Design of Recombinant Spider Silk Proteins for Cell Type Specific Binding (Adv. Healthcare Mater. 9/2023)

Cell-Selective Spider Silk Proteins In article number 2202660, Vanessa Tanja Trossmann and Thomas Scheibel design recombinant spider silk proteins to guide cell-specific materials interactions. The fusion of spider silk proteins with short peptides from extracellular matrix proteins such as KGD yield cell type selective binding, which enables the development of hierarchically organized tissues.

Fission yeast Swi2 designates cell-type specific donor and stimulates Rad51-driven strand exchange for mating-type switching.

A haploid of the fission yeast Schizosaccharomyces pombe expresses either the P or M mating-type, determined by the active, euchromatic, mat1 cassette. Mating-type is switched by Rad51-driven gene conversion of mat1 using a heterochromatic donor cassette, mat2-P or mat3-M. The Swi2-Swi5 complex, a mating-type switching factor, is central to this process by designating a preferred donor in a cell-type-specific manner. Swi2-Swi5 selectively enables one of two cis-acting recombination enhancers, SRE2 adjacent to mat2-P or SRE3 adjacent to mat3-M. Here, we identified two functionally important motifs in Swi2, a Swi6 (HP1 homolog)-binding site and two DNA-binding AT-hooks. Genetic analysis demonstrated that the AT-hooks were required for Swi2 localization at SRE3 to select the mat3-M donor in P cells, while the Swi6-binding site was required for Swi2 localization at SRE2 to select mat2-P in M cells. In addition, the Swi2-Swi5 complex promoted Rad51-driven strand exchange in vitro. Taken together, our results show how the Swi2-Swi5 complex would localize to recombination enhancers through a cell-type specific binding mechanism and stimulate Rad51-driven gene conversion at the localization site.

Design of Recombinant Spider Silk Proteins for Cell Type Specific Binding.

Cytophilic (cell-adhesive) materials are very important for tissue engineering and regenerative medicine. However, for engineering hierarchically organized tissue structures comprising different cell types, cell-specific attachment and guidance are decisive. In this context, materials made of recombinant spider silk proteins are promising scaffolds, since they exhibit high biocompatibility, biodegradability, and the underlying proteins can be genetically functionalized. Here, previously established spider silk variants based on the engineered Araneus diadematus fibroin 4 (eADF4(C16)) are genetically modified with cell adhesive peptide sequences from extracellular matrix proteins, including IKVAV, YIGSR,QHREDGS, and KGD. Interestingly, eADF4(C16)-KGD as one of 18tested variants is cell-selective for C2C12mouse myoblasts, one out of 11 tested cell lines. Co-culturing with B50rat neuronal cells confirms the cell-specificity of eADF4(C16)-KGD material surfaces for C2C12mouse myoblast adhesion.

Deep learning approaches for noncoding variant prioritization in neurodegenerative diseases.

Determining how noncoding genetic variants contribute to neurodegenerative dementias is fundamental to understanding disease pathogenesis, improving patient prognostication, and developing new clinical treatments. Next generation sequencing technologies have produced vast amounts of genomic data on cell type-specific transcription factor binding, gene expression, and three-dimensional chromatin interactions, with the promise of providing key insights into the biological mechanisms underlying disease. However, this data is highly complex, making it challenging for researchers to interpret, assimilate, and dissect. To this end, deep learning has emerged as a powerful tool for genome analysis that can capture the intricate patterns and dependencies within these large datasets. In this review, we organize and discuss the many unique model architectures, development philosophies, and interpretation methods that have emerged in the last few years with a focus on using deep learning to predict the impact of genetic variants on disease pathogenesis. We highlight both broadly-applicable genomic deep learning methods that can be fine-tuned to disease-specific contexts as well as existing neurodegenerative disease research, with an emphasis on Alzheimer's-specific literature. We conclude with an overview of the future of the field at the intersection of neurodegeneration, genomics, and deep learning.

Structural basis for cell type specific DNA binding of C/EBPβ: The case of cell cycle inhibitor p15INK4b promoter

C/EBPβ is a key regulator of numerous cellular processes, but it can also contribute to tumorigenesis and viral diseases. It binds to specific DNA sequences (C/EBP sites) and interacts with other transcription factors to control expression of multiple eukaryotic genes in a tissue and cell-type dependent manner. A body of evidence has established that cell-type-specific regulatory information is contained in the local DNA sequence of the binding motif. In human epithelial cells, C/EBPβ is an essential cofactor for TGFβ signaling in the case of Smad2/3/4 and FoxO-dependent induction of the cell cycle inhibitor, p15INK4b. In the TGFβ-responsive region 2 of the p15INK4b promoter, the Smad binding site is flanked by a C/EBP site, CTTAA•GAAAG, which differs from the canonical, palindromic ATTGC•GCAAT motif. The X-ray crystal structure of C/EBPβ bound to the p15INK4b promoter fragment shows how GCGC-to-AAGA substitution generates changes in the intermolecular interactions in the protein—DNA interface that enhances C/EBPβ binding specificity, limits possible epigenetic regulation of the promoter, and generates a DNA element with a unique pattern of methyl groups in the major groove. Significantly, CT/GA dinucleotides located at the 5′ends of the double stranded element maintain local narrowing of the DNA minor groove width that is necessary for DNA recognition. Our results suggest that C/EBPβ would accept all forms of modified cytosine in the context of the CpT site. This contrasts with the effect on the consensus motif, where C/EBPβ binding is modestly increased by cytosine methylation, but substantially decreased by hydroxymethylation.

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.

Toward a base-resolution panorama of the in vivo impact of cytosine methylation on transcription factor binding

BackgroundWhile methylation of CpG dinucleotides is traditionally considered antagonistic to the DNA-binding activity of most transcription factors (TFs), recent in vitro studies have revealed a more complex picture, suggesting that over a third of TFs may preferentially bind to methylated sequences. Expanding these in vitro observations to in vivo TF binding preferences is challenging since the effect of methylation of individual CpG sites cannot be easily isolated from the confounding effects of DNA accessibility and regional DNA methylation. Thus, in vivo methylation preferences of most TFs remain uncharacterized.ResultsWe introduce joint accessibility-methylation-sequence (JAMS) models, which connect the strength of the binding signal observed in ChIP-seq to the DNA accessibility of the binding site, regional methylation level, DNA sequence, and base-resolution cytosine methylation. We show that JAMS models quantitatively explain TF occupancy, recapitulate cell type-specific TF binding, and have high positive predictive value for identification of TFs affected by intra-motif methylation. Analysis of 2209 ChIP-seq experiments results in high-confidence JAMS models for 260 TFs, revealing a negative association between in vivo TF occupancy and intra-motif methylation for 45% of studied TFs, as well as 16 TFs that are predicted to bind to methylated sites, including 11 novel methyl-binding TFs mostly from the multi-zinc finger family.ConclusionsOur study substantially expands the repertoire of in vivo methyl-binding TFs, but also suggests that most TFs that prefer methylated CpGs in vitro present themselves as methylation agnostic in vivo, potentially due to the balancing effect of competition with other methyl-binding proteins.

From genetic variants to diseases: the role of LACC1 in T cells

Abstract Expression quantitative trait loci (eQTLs) studies provide associations of genetic variants with the expression of particular genes. But how these eQTLs directly affect gene expression and in what cell types of the immune system they act remain largely unknown. One hypothesis is that functional eQTLs in the cis-regulatory elements may affect cell type-specific transcription factor binding, thus reducing gene expression and even contributing to the cause of autoimmune and inflammatory diseases. Previously, we identified T cell-specific eQTLs of a gene known as laccase domain containing 1 (LACC1), which are GWAS hits for Crohn’s disease. Four of them are located in the LACC1 promoter region in the same haploblock. Direct association of disease-risk variants with lower LACC1 expression was confirmed by comparing pre-mRNA quantity of the different alleles in LACC1 heterozygous human CD4 T cells. We further validated the functional role of the eQTLs using CRISPR Cas9-mediated base-pair editing. LACC1 function has been analyzed primarily in macrophages and shown to affect their function, but the GWAS hits do not affect expression in macrophages. Moreover, in our in vitro studies, human CD4 T cells with LACC1 knockdown showed altered metabolic programming and proliferation, which may contribute to inflammatory diseases. Overall, our study provides an excellent model for analyzing and explaining how genetic variants contribute to diseases by influencing gene expression in specific cell types and changing cell functions.

Abstract 113: Cell-specific Chromatin Landscape Of Human Coronary Artery Resolves Mechanisms Of Disease Risk

Coronary artery disease (CAD) is a complex inflammatory disease involving genetic influences across several cell types. Genome-wide association studies (GWAS) have identified over 170 loci associated with CAD, where the majority of risk variants reside in noncoding DNA sequences impacting cis -regulatory elements (CREs). Here, we applied single-cell ATAC-seq to profile 28,316 cells across coronary artery segments from 41 patients with varying stages of CAD, which revealed 14 distinct cellular clusters. We mapped over 320,000 accessible sites across all cells, identified cell type-specific elements, transcription factors, and prioritized functional CAD risk variants via quantitative trait locus and sequence-based predictive modeling. Using differential peak analyses we identified a number of candidate mechanisms for smooth muscle cell transition states (e.g. fibromyocytes). By integrating these profiles with GWAS meta-analysis summary data we resolved cell type-specific putative binding sites for the majority of CAD risk variants. In particular, we prioritized functional variants predicted to alter MEF2 binding in smooth muscle cells at the MRAS locus. We also identify variants predicted to alter macrophage-specific regulation of LIPA. We further employed DNA to gene linkage to nominate disease-associated key driver transcription factors such as PRDM16 and TBX2. Together, this single cell atlas provides a critical step towards interpreting cis -regulatory mechanisms in the vessel wall across the continuum of CAD risk.

Dynamic CTCF binding directly mediates interactions among cis-regulatory elements essential for hematopoiesis

While constitutive CCCTC-binding factor (CTCF)–binding sites are needed to maintain relatively invariant chromatin structures, such as topologically associating domains, the precise roles of CTCF to control cell-type–specific transcriptional regulation remain poorly explored. We examined CTCF occupancy in different types of primary blood cells derived from the same donor to elucidate a new role for CTCF in gene regulation during blood cell development. We identified dynamic, cell-type–specific binding sites for CTCF that colocalize with lineage-specific transcription factors. These dynamic sites are enriched for single-nucleotide polymorphisms that are associated with blood cell traits in different linages, and they coincide with the key regulatory elements governing hematopoiesis. CRISPR-Cas9–based perturbation experiments demonstrated that these dynamic CTCF-binding sites play a critical role in red blood cell development. Furthermore, precise deletion of CTCF-binding motifs in dynamic sites abolished interactions of erythroid genes, such as RBM38, with their associated enhancers and led to abnormal erythropoiesis. These results suggest a novel, cell-type–specific function for CTCF in which it may serve to facilitate interaction of distal regulatory emblements with target promoters. Our study of the dynamic, cell-type–specific binding and function of CTCF provides new insights into transcriptional regulation during hematopoiesis.

Ledidi: Designing genomic edits that induce functional activity

Abstract The development of modern genome editing tools has enabled researchers to make edits with high precision, but has left unsolved the problem of designing these edits. We propose Ledidi, an approach that treats the design of genomic edits as an optimization problem where the goal is to produce the desired output from a predictive model. The discrete nature of biological sequences makes direct optimization challenging, but we overcome this by using the Gumbel-Softmax reparameterization trick. We validate Ledidi by pairing it with the Basenji model, which makes predictions for thousands of functional profiles, and designing edits that affect CTCF binding and induce cell type-specific binding of JUND.

A Comparative Study of Supervised Machine Learning Algorithms for the Prediction of Long-Range Chromatin Interactions

The role of three-dimensional genome organization as a critical regulator of gene expression has become increasingly clear over the last decade. Most of our understanding of this association comes from the study of long range chromatin interaction maps provided by Chromatin Conformation Capture-based techniques, which have greatly improved in recent years. Since these procedures are experimentally laborious and expensive, in silico prediction has emerged as an alternative strategy to generate virtual maps in cell types and conditions for which experimental data of chromatin interactions is not available. Several methods have been based on predictive models trained on one-dimensional (1D) sequencing features, yielding promising results. However, different approaches vary both in the way they model chromatin interactions and in the machine learning-based strategy they rely on, making it challenging to carry out performance comparison of existing methods. In this study, we use publicly available 1D sequencing signals to model cohesin-mediated chromatin interactions in two human cell lines and evaluate the prediction performance of six popular machine learning algorithms: decision trees, random forests, gradient boosting, support vector machines, multi-layer perceptron and deep learning. Our approach accurately predicts long-range interactions and reveals that gradient boosting significantly outperforms the other five methods, yielding accuracies of about 95%. We show that chromatin features in close genomic proximity to the anchors cover most of the predictive information, as has been previously reported. Moreover, we demonstrate that gradient boosting models trained with different subsets of chromatin features, unlike the other methods tested, are able to produce accurate predictions. In this regard, and besides architectural proteins, transcription factors are shown to be highly informative. Our study provides a framework for the systematic prediction of long-range chromatin interactions, identifies gradient boosting as the best suited algorithm for this task and highlights cell-type specific binding of transcription factors at the anchors as important determinants of chromatin wiring mediated by cohesin.

A Generative Neural Network for Maximizing Fitness and Diversity of Synthetic DNA and Protein Sequences.

SUMMARYEngineering gene and protein sequences with defined functional properties is a major goal of synthetic biology. Deep neural network models, together with gradient ascent-style optimization, show promise for sequence design. The generated sequences can however get stuck in local minima and often have low diversity. Here, we develop deep exploration networks (DENs), a class of activation-maximizing generative models, which minimize the cost of a neural network fitness predictor by gradient descent. By penalizing any two generated patterns on the basis of a similarity metric, DENs explicitly maximize sequence diversity. To avoid drifting into low-confidence regions of the predictor, we incorporate variational autoencoders to maintain the likelihood ratio of generated sequences. Using DENs, we engineered polyadenylation signals with more than 10-fold higher selection odds than the best gradient ascent-generated patterns, identified splice regulatory sequences predicted to result in highly differential splicing between cell lines, and improved on state-of-the-art results for protein design tasks.

Developmental regulation of cell type-specific transcription by novel promoter-proximal sequence elements.

Cell type-specific transcriptional programs that drive differentiation of specialized cell types are key players in development and tissue regeneration. One of the most dramatic changes in the transcription program in Drosophila occurs with the transition from proliferating spermatogonia to differentiating spermatocytes, with >3000 genes either newly expressed or expressed from new alternative promoters in spermatocytes. Here we show that opening of these promoters from their closed state in precursor cells requires function of the spermatocyte-specific tMAC complex, localized at the promoters. The spermatocyte-specific promoters lack the previously identified canonical core promoter elements except for the Inr. Instead, these promoters are enriched for the binding site for the TALE-class homeodomain transcription factors Achi/Vis and for a motif originally identified under tMAC ChIP-seq peaks. The tMAC motif resembles part of the previously identified 14-bp β2UE1 element critical for spermatocyte-specific expression. Analysis of downstream sequences relative to transcription start site usage suggested that ACA and CNAAATT motifs at specific positions can help promote efficient transcription initiation. Our results reveal how promoter-proximal sequence elements that recruit and are acted upon by cell type-specific chromatin binding complexes help establish a robust, cell type-specific transcription program for terminal differentiation.

Annotations capturing cell type-specific TF binding explain a large fraction of disease heritability.

Regulatory variation plays a major role in complex disease and that cell type-specific binding of transcription factors (TF) is critical to gene regulation. However, assessing the contribution of genetic variation in TF-binding sites to disease heritability is challenging, as binding is often cell type-specific and annotations from directly measured TF binding are not currently available for most cell type-TF pairs. We investigate approaches to annotate TF binding, including directly measured chromatin data and sequence-based predictions. We find that TF-binding annotations constructed by intersecting sequence-based TF-binding predictions with cell type-specific chromatin data explain a large fraction of heritability across a broad set of diseases and corresponding cell types; this strategy of constructing annotations addresses both the limitation that identical sequences may be bound or unbound depending on surrounding chromatin context and the limitation that sequence-based predictions are generally not cell type-specific. We partitioned the heritability of 49 diseases and complex traits using stratified linkage disequilibrium (LD) score regression with the baseline-LD model (which is not cell type-specific) plus the new annotations. We determined that 100bp windows around MotifMap sequenced-based TF-binding predictions intersected with a union of six cell type-specific chromatin marks (imputed using ChromImpute) performed best, with an 58% increase in heritability enrichment compared to the chromatin marks alone (11.6× vs. 7.3×, P = 9 × 10-14 for difference) and a 20% increase in cell type-specific signal conditional on annotations from the baseline-LD model (P = 8 × 10-11 for difference). Our results show that TF-binding annotations explain substantial disease heritability and can help refine genome-wide association signals.

Predicting transcription factor binding using ensemble random forest models.

Background: Understanding the location and cell-type specific binding of Transcription Factors (TFs) is important in the study of gene regulation. Computational prediction of TF binding sites is challenging, because TFs often bind only to short DNA motifs and cell-type specific co-factors may work together with the same TF to determine binding. Here, we consider the problem of learning a general model for the prediction of TF binding using DNase1-seq data and TF motif description in form of position specific energy matrices (PSEMs). Methods: We use TF ChIP-seq data as a gold-standard for model training and evaluation. Our contribution is a novel ensemble learning approach using random forest classifiers. In the context of the ENCODE-DREAM in vivo TF binding site prediction challenge we consider different learning setups. Results: Our results indicate that the ensemble learning approach is able to better generalize across tissues and cell-types compared to individual tissue-specific classifiers or a classifier built based upon data aggregated across tissues. Furthermore, we show that incorporating DNase1-seq peaks is essential to reduce the false positive rate of TF binding predictions compared to considering the raw DNase1 signal. Conclusions: Analysis of important features reveals that the models preferentially select motifs of other TFs that are close interaction partners in existing protein protein-interaction networks. Code generated in the scope of this project is available on GitHub: https://github.com/SchulzLab/TFAnalysis (DOI: 10.5281/zenodo.1409697).

Pre-marked chromatin and transcription factor co-binding shape the pioneering activity of Foxa2.

Pioneer transcription factors (PTF) can recognize their binding sites on nucleosomal DNA and trigger chromatin opening for recruitment of other non-pioneer transcription factors. However, critical properties of PTFs are still poorly understood, such as how these transcription factors selectively recognize cell type-specific binding sites and under which conditions they can initiate chromatin remodelling. Here we show that early endoderm binding sites of the paradigm PTF Foxa2 are epigenetically primed by low levels of active chromatin modifications in embryonic stem cells (ESC). Priming of these binding sites is supported by preferential recruitment of Foxa2 to endoderm binding sites compared to lineage-inappropriate binding sites, when ectopically expressed in ESCs. We further show that binding of Foxa2 is required for chromatin opening during endoderm differentiation. However, increased chromatin accessibility was only detected on binding sites which are synergistically bound with other endoderm transcription factors. Thus, our data suggest that binding site selection of PTFs is directed by the chromatin environment and that chromatin opening requires collaboration of PTFs with additional transcription factors.

Predicting transcription factor binding using ensemble random forest models

Background: Understanding the location and cell-type specific binding of Transcription Factors (TFs) is important in the study of gene regulation. Computational prediction of TF binding sites is challenging, because TFs often bind only to short DNA motifs and cell-type specific co-factors may work together with the same TF to determine binding. Here, we consider the problem of learning a general model for the prediction of TF binding using DNase1-seq data and TF motif description in form of position specific energy matrices (PSEMs). Methods: We use TF ChIP-seq data as a gold-standard for model training and evaluation. Our contribution is a novel ensemble learning approach using random forest classifiers. In the context of the ENCODE-DREAM in vivo TF binding site prediction challenge we consider different learning setups. Results: Our results indicate that the ensemble learning approach is able to better generalize across tissues and cell-types compared to individual tissue-specific classifiers or a classifier built based upon data aggregated across tissues. Furthermore, we show that incorporating DNase1-seq peaks is essential to reduce the false positive rate of TF binding predictions compared to considering the raw DNase1 signal. Conclusions: Analysis of important features reveals that the models preferentially select motifs of other TFs that are close interaction partners in existing protein protein-interaction networks. Code generated in the scope of this project is available on GitHub: https://github.com/SchulzLab/TFAnalysis (DOI: 10.5281/zenodo.1409697).

FactorNet: A deep learning framework for predicting cell type specific transcription factor binding from nucleotide-resolution sequential data.

Due to the large numbers of transcription factors (TFs) and cell types, querying binding profiles of all valid TF/cell type pairs is not experimentally feasible. To address this issue, we developed a convolutional-recurrent neural network model, called FactorNet, to computationally impute the missing binding data. FactorNet trains on binding data from reference cell types to make predictions on testing cell types by leveraging a variety of features, including genomic sequences, genome annotations, gene expression, and signal data, such as DNase I cleavage. FactorNet implements several convenient strategies to reduce runtime and memory consumption. By visualizing the neural network models, we can interpret how the model predicts binding. We also investigate the variables that affect cross-cell type accuracy, and offer suggestions to improve upon this field. Our method ranked among the top teams in the ENCODE-DREAM in vivo Transcription Factor Binding Site Prediction Challenge, achieving first place on six of the 13 final round evaluation TF/cell type pairs, the most of any competing team. The FactorNet source code is publicly available, allowing users to reproduce our methodology from the ENCODE-DREAM Challenge.

Accurate prediction of cell type-specific transcription factor binding

Prediction of cell type-specific, in vivo transcription factor binding sites is one of the central challenges in regulatory genomics. Here, we present our approach that earned a shared first rank in the “ENCODE-DREAM in vivo Transcription Factor Binding Site Prediction Challenge” in 2017. In post-challenge analyses, we benchmark the influence of different feature sets and find that chromatin accessibility and binding motifs are sufficient to yield state-of-the-art performance. Finally, we provide 682 lists of predicted peaks for a total of 31 transcription factors in 22 primary cell types and tissues and a user-friendly version of our approach, Catchitt, for download.