Transcription Factor Binding Sites Research Articles

Deep learning model for predicting genetic diseases using DNA sequence data

The emergence of machine learning in the recent decade has excelled in determining new potential features and nonlinear relationships existing between the data derived from the DNA sequences of genetic diseases. Machine learning also enhances the process of handling data with maximum predicted variables compared to observations during the data mining process of prediction. In this context, our study presents a deep learning model for predicting Transcription Factor Binding Sites (TFBS) in DNA sequences, with a focus on features within genetic data associated with diseases. Transcription Factors (TFs) play a crucial role in modulating gene expression by binding to TFBS. The accurate prediction of TFBS is essential for understanding genome function and evolution. Thus, we develop an efficient deep learning model that considers TFBS prediction as a nucleotide-level binary classification task. In our proposed model, first we create an input matrix using the original DNA sequences. Next, we encode these DNA sequences using one-hot encoding, representing them as a sequence of numerical values. We then employ three convolutional layers, allowing our model to capture intricate patterns and motif features over a larger spatial range. To capture important features within the DNA sequence and to focus on them, we incorporate an attention layer. Finally, a dense layer, consisting of two fully connected layers and a dropout layer, calculates the probability of TF binding site occurrence based on the features learned by the proposed model. Our experimental results, using in-vivo datasets obtained from Chip-seq, demonstrate the superior performance of our proposed deep learning model in TFBS prediction compared to other existing state-of-the-art methods. The improvement in accuracy is due to additional layers of CNN and then an attention layer in the model. Thus, this result in a better performance of our approach in predicting the transcription factor binding sites and enhancing our understanding of gene regulation and genome function.

Reconnaissance of Allostery via the Restoration of Native p53 DNA-Binding Domain Dynamics in Y220C Mutant p53 Tumor Suppressor Protein.

Allosteric regulation of protein dynamics infers a long-range deliberate propagation of information via micro- and macroscale interactions. The Y220C structural mutant is one of the most frequent cancerous p53 mutants. The mutation is distally located from the DNA-binding site of the p53 DNA-binding domain yet causes changes in DNA recognition. This system presents a unique opportunity to examine the allosteric control of mutated proteins under a drug design paradigm. We focus on the key case study of p53 Y220C mutation restoration by a series of new compounds suggested to have Y220C reactivation properties in comparison to our previous findings on the restorative potential of PK11000, a compound studied extensively for reactivation in vitro and in vivo. Previously, we implemented all-atom molecular dynamics (MD) simulations and our lab's techniques of MD-Sectors and MD-Markov state models on the wild type, the Y220C mutant, and Y220C with PK11000 to characterize the effector's restorative properties in terms of conformational dynamics and hydrogen bonding. In this study, we turn to probing the effects made by docking the battery of a new but less well-tested set of aminobenzothiazole derivative compounds reported by Baud et al., which show promise of Y220C rescue. We find that while complete and precise reconstitution of p53 WT molecular dynamics may not be observed as was the case with PK11000, dispersed local reconstitution of loop dynamics provides evidence of rescuing effects by aminobenzothiazole derivative N,2-dihydroxy-3,5-diiodo-4-(1H-pyrrol-1-yl)benzamide, Effector 22, like what we observed for PK11000. Generalizable insights into the mutation and allosteric reactivation of p53 by various effectors by reconstitution of WT dynamics observed in statistical conformational ensemble analysis and network inference are discussed, considering the development of allosteric drug design rooted in first principles.

DeepReg: a deep learning hybrid model for predicting transcription factors in eukaryotic and prokaryotic genomes

Deep learning models (DLMs) have gained importance in predicting, detecting, translating, and classifying a diversity of inputs. In bioinformatics, DLMs have been used to predict protein structures, transcription factor-binding sites, and promoters. In this work, we propose a hybrid model to identify transcription factors (TFs) among prokaryotic and eukaryotic protein sequences, named Deep Regulation (DeepReg) model. Two architectures were used in the DL model: a convolutional neural network (CNN), and a bidirectional long-short-term memory (BiLSTM). DeepReg reached a precision of 0.99, a recall of 0.97, and an F1-score of 0.98. The quality of our predictions, the bias-variance trade-off approach, and the characterization of new TF predictions were evaluated and compared against those produced by DeepTFactor, as well as against experimental data from three model organisms. Predictions based on our DLM tended to exhibit less variance and bias than those from DeepTFactor, thus increasing reliability and decreasing overfitting.

MMGAT: a graph attention network framework for ATAC-seq motifs finding.

Motif finding in Assay for Transposase-Accessible Chromatin using sequencing (ATAC-seq) data is essential to reveal the intricacies of transcription factor binding sites (TFBSs) and their pivotal roles in gene regulation. Deep learning technologies including convolutional neural networks (CNNs) and graph neural networks (GNNs), have achieved success in finding ATAC-seq motifs. However, CNN-based methods are limited by the fixed width of the convolutional kernel, which makes it difficult to find multiple transcription factor binding sites with different lengths. GNN-based methods has the limitation of using the edge weight information directly, makes it difficult to aggregate the neighboring nodes' information more efficiently when representing node embedding. To address this challenge, we developed a novel graph attention network framework named MMGAT, which employs an attention mechanism to adjust the attention coefficients among different nodes. And then MMGAT finds multiple ATAC-seq motifs based on the attention coefficients of sequence nodes and k-mer nodes as well as the coexisting probability of k-mers. Our approach achieved better performance on the human ATAC-seq datasets compared to existing tools, as evidenced the highest scores on the precision, recall, F1_score, ACC, AUC, and PRC metrics, as well as finding 389 higher quality motifs. To validate the performance of MMGAT in predicting TFBSs and finding motifs on more datasets, we enlarged the number of the human ATAC-seq datasets to 180 and newly integrated 80 mouse ATAC-seq datasets for multi-species experimental validation. Specifically on the mouse ATAC-seq dataset, MMGAT also achieved the highest scores on six metrics and found 356 higher-quality motifs. To facilitate researchers in utilizing MMGAT, we have also developed a user-friendly web server named MMGAT-S that hosts the MMGAT method and ATAC-seq motif finding results. The advanced methodology MMGAT provides a robust tool for finding ATAC-seq motifs, and the comprehensive server MMGAT-S makes a significant contribution to genomics research. The open-source code of MMGAT can be found at https://github.com/xiaotianr/MMGAT , and MMGAT-S is freely available at https://www.mmgraphws.com/MMGAT-S/ .

Defensins of the stable fly (Stomoxys Calcitrans) have developmental-specific regulation and evolve at different rates.

Organisms produce antimicrobial peptides (AMPs) either in response to infection (induced) or continuously (constitutively) to combat microbes encountered during normal trophic activities and/or through pathogenic infections. The expression of AMPs is tightly regulated often with specificity to particular tissues or developmental stages. As a result, AMPs face varying selective pressures based on the microbes the organism's tissue or developmental stage encounters. Here, we analyzed the evolution and developmental-specific expression of Defensins, which are ancient AMPs in insects, in the stable fly (Stomoxys calcitrans). Stable fly larvae inhibit microbe-rich environments, whereas adult flies, as blood-feeders, experience comparatively fewer encounters with diverse microbial communities. Using existing RNA-seq datasets, we identified six Defensins that were only expressed in larvae (larval Defensins) and five that were not expressed in larvae (non-larval Defensins). Each of the non-larval Defensins was expressed in at least one adult tissue sample. Half of the larval Defensins were induced by mating or feeding in adults, and all three of the induced Defensins were located downstream of canonical binding sites for an Imd transcription factor involved in the highly conserved NF-κB signaling that regulates induction of AMPs. The larval and non-larval Defensins were located in distinct genomic regions, and the amino acid sequences of the larval Defensins formed a monophyletic clade. There were more amino acid substitutions across non-larval Defensins, with multiple genes losing a highly conserved furin cleavage site thought to be required for the removal of the amino terminus from the mature Defensin domain. However, larval Defensins had a higher proportion of radical amino acid substitutions, altering amino acid size and polarity. Our results reveal insights into the developmental stage-specific regulation of AMPs, and they suggest different regulatory regimes impose unique selection pressures on AMPs, possibly as a result of variation in exposure to microbial communities across development.

A statistical approach for identifying single nucleotide variants that affect transcription factor binding

Non-coding variants located within regulatory elements may alter gene expression by modifying transcription factor (TF) binding sites, thereby leading to functional consequences. Different TF models are being used to assess the effect of DNA sequence variants, such as single nucleotide variants (SNVs). Often existing methods are slow and do not assess statistical significance of results. We investigated the distribution of absolute maximal differential TF binding scores for general computational models that affect TF binding. We find that a modified Laplace distribution can adequately approximate the empirical distributions. A benchmark on invitro and invivo datasets showed that our approach improves upon an existing method in terms of performance and speed. Applications on eQTLs and on a genome-wide association study illustrate the usefulness of our statistics by highlighting cell type-specific regulators and target genes. An implementation of our approach is freely available on GitHub and as bioconda package.

Tissue-specific atlas of trans-models for gene regulation elucidates complex regulation patterns

BackgroundDeciphering gene regulation is essential for understanding the underlying mechanisms of healthy and disease states. While the regulatory networks formed by transcription factors (TFs) and their target genes has been mostly studied with relation to cis effects such as in TF binding sites, we focused on trans effects of TFs on the expression of their transcribed genes and their potential mechanisms.ResultsWe provide a comprehensive tissue-specific atlas, spanning 49 tissues of TF variations affecting gene expression through computational models considering two potential mechanisms, including combinatorial regulation by the expression of the TFs, and by genetic variants within the TF.We demonstrate that similarity between tissues based on our discovered genes corresponds to other types of tissue similarity. The genes affected by complex TF regulation, and their modelled TFs, were highly enriched for pharmacogenomic functions, while the TFs themselves were also enriched in several cancer and metabolic pathways. Additionally, genes that appear in multiple clusters are enriched for regulation of immune system while tissue clusters include cluster-specific genes that are enriched for biological functions and diseases previously associated with the tissues forming the cluster. Finally, our atlas exposes multilevel regulation across multiple tissues, where TFs regulate other TFs through the two tested mechanisms.ConclusionsOur tissue-specific atlas provides hierarchical tissue-specific trans genetic regulations that can be further studied for association with human phenotypes.

Benzo[a]pyrene exposure causes exonal switch resulting in reduced surface CD5 expression in an AHR-dependent manner

The function of CD5 protein in T cells is well documented, but regulation of its surface-level expression has yet to be fully understood. However, variation in its surface expression is associated with various immunopathological conditions and haematological malignancies. Briefly, expression of an alternate exon E1B of a human endogenous retroviruses (HERV) origin directly downregulates the conventional transcript variant (E1A), as its expression leads to the retention of the resultant protein at the intracellular level (cCD5). A separate promoter governs the expression of E1B and may be influenced by different transcription factors. Hence, we performed in silico transcription factor binding site (TFBS) analysis of the 3 kb upstream region from TSS of exon E1B and found five putative DREs (Dioxin Response elements) with good similarity scores. Further, we observed the upregulation in E1B expression after the exposure of BaP (a dioxin) and the reduction of E1A expression and their respective protein, i.e. sCD5 and cCD5. The binding of AHR at the predicted DRE sites was confirmed by ChIP qPCR and AHR specific inhibitor and gene silencing studies suggested the involvement of AHR in exonal switch. This study indicates that the polycyclic aromatic hydrocarbon decreases the sCD5 expression by upregulating alternative exon expression, which may adversely affect the overall T cell functions.

Transcription Factor Binding Site in Promoter Determines the Pattern of Plasmid-Based Transgene Expression In Vivo.

Understanding the regulation of transgene expression is critical for the success of plasmid-based gene therapy and vaccine development. In this study, we used two sets of plasmid vectors containing secreted embryonic alkaline phosphatase or the mouse IL-10 gene as a reporter and investigated the role of promoter elements in regulating transgene expression in vivo. We demonstrated in mice that hydrodynamic transfer of plasmids with the CMV promoter resulted in a high level of reporter gene expression that declined rapidly over time. In contrast, when plasmids with albumin promoters were used, a lower but sustained gene expression pattern was observed. We also found that plasmids containing a shorter CMV promoter sequence with fewer transcription factor binding sites showed a decrease in the peak level of gene expression without changing the overall pattern of reporter gene expression. The replacement of regulatory elements in the CMV promoter with a single regulatory element of the albumin promoter changed the pattern of transient gene expression seen in the CMV promoter to a pattern of sustained gene expression identical to that of a full albumin promoter. ChIP analyses demonstrated an elevated binding of acetylated histones and TATA box-binding protein to the promoter carrying regulatory elements of the albumin promoter. These results suggest that the strength of a promoter is determined by the number of appropriate transcription factor binding sites, while gene expression persistence is determined by the presence of regulatory elements capable of recruiting epigenetic modifying complexes that make the promoter accessible for transcription. This study provides important insights into the mechanisms underlying gene expression regulation in vivo, which can be used to improve plasmid-based gene therapy and vaccine development.

Concordant B and T Cell Heterogeneity Inferred from the Multiomic Landscape of Peripheral Blood Mononuclear Cells in a Crohn's Disease Cohort.

Crohn's disease is characterized by inflammation in the gastrointestinal tract due to a combination of genetic, immune, and environmental factors. Transcriptomic and epigenomic profiling of intestinal tissue of Crohn's disease patients have revealed valuable insights into pathology, however have not been conducted jointly on less invasive peripheral blood mononuclear cells (PBMCs). Furthermore, the heterogeneous responses to treatments among individuals with Crohn's disease imply hidden diversity of pathological mechanisms. We employed single nucleus multiomic analysis, integrating both snRNA-seq and snATAC-seq of PBMCs with a variety of open source bioinformatics applications. Our findings reveal a diverse range of transcriptional signatures among individuals, highlighting the heterogeneity in PBMC profiles. Nevertheless, striking concordance between three heterogeneous groups was observed across B cells and T cells. Differential gene regulatory mechanisms partially explain these profiles, notably including a signature involving TGFß signaling in two individuals with Crohn's disease. A mutation mapped to a transcription factor binding site within a differentially accessible peak associated with the expression of this pathway, with implications for a personalized approach to understanding disease pathology. This study highlights how multiomic analysis can reveal common regulatory mechanisms that underlie heterogeneity of PBMC profiles, one of which may be specific to inflammatory disease.

Molecular phenotyping of small cell lung cancer using targeted cfDNA profiling of transcriptional regulatory regions.

We report an approach for cancer phenotyping based on targeted sequencing of cell-free DNA (cfDNA) for small cell lung cancer (SCLC). In SCLC, differential activation of transcription factors (TFs), such as ASCL1, NEUROD1, POU2F3, and REST defines molecular subtypes. We designed a targeted capture panel that identifies chromatin organization signatures at 1535 TF binding sites and 13,240 gene transcription start sites and detects exonic mutations in 842 genes. Sequencing of cfDNA from SCLC patient-derived xenograft models captured TF activity and gene expression and revealed individual highly informative loci. Prediction models of ASCL1 and NEUROD1 activity using informative loci achieved areas under the receiver operating characteristic curve (AUCs) from 0.84 to 0.88 in patients with SCLC. As non-SCLC (NSCLC) often transforms to SCLC following targeted therapy, we applied our framework to distinguish NSCLC from SCLC and achieved an AUC of 0.99. Our approach shows promising utility for SCLC subtyping and transformation monitoring, with potential applicability to diverse tumor types.

Synthesis, Anti-Inflammatory Activities, and Molecular Docking Study of Novel Pyxinol Derivatives as Inhibitors of NF-κB Activation.

Pyxinol, an active metabolite of ginsenosides in human hepatocytes, exhibits various pharmacological activities. Here, a series of C-3 modified pyxinol derivatives was designed and virtually screened by molecular docking with the key inflammation-related proteins of the nuclear factor kappa B (NF-κB) pathway. Some of the novel derivatives were synthesized to assess their effects in inhibiting the production of nitric oxide (NO) and mitochondrial reactive oxygen species (MtROS) in lipopolysaccharide-triggered RAW264.7 cells. Derivative 2c exhibited the highest NO and MtROS inhibitory activities with low cytotoxicity. Furthermore, 2c decreased the protein levels of interleukin 1β, tumor necrosis factor α, inducible nitric oxide synthase, and cyclooxygenase 2 and suppressed the activation of NF-κB signaling. Cellular thermal shift assays indicated that 2c could directly bind with p65 and p50 in situ. Molecular docking revealed that 2c's binding to the p65-p50 heterodimer and p50 homodimer was close to their DNA binding sites. In summary, pyxinol derivatives possess potential for development as NF-κB inhibitors.

Physalin H, physalin B, and isophysalin B suppress the quorum-sensing function of Staphylococcus aureus by binding to AgrA.

The virulence of Staphylococcus aureus, including methicillin-resistant S. aureus (MRSA), depends on the expression of toxins and virulence factors controlled by the quorum-sensing (QS) system, encoded on the virulence accessory gene regulator (agr) locus. The aim of this study was to identify a phytochemical that inhibits Agr-QS function and to elucidate its mechanism. We screened 577 compounds and identified physalin H, physalin B, and isophysalin B--phytochemicals belonging to physalins found in plants of the Solanaceae family--as novel Agr-QS modulators. Biological analyses and in vitro protein-DNA binding assays suggested that these physalins suppress gene expression related to the Agr-QS system by inhibiting binding of the key response regulator AgrA to the agr promoters, reducing the function of hemolytic toxins downstream of these genes in MRSA. Furthermore, although physalin F suppressed gene expression in the Agr-QS system, its anti-hemolytic activity was lower than that of physalins H, B, and isophysalin B. Conversely, five physalins isolated from the same plant with the ability to suppress Agr-QS did not reduce bacterial Agr-QS activity but inhibited AgrA binding to DNA in vitro. A docking simulation revealed that physalin interacts with the DNA-binding site of AgrA in three docking states. The carbonyl oxygens at C-1 and C-18 of physalins, which can suppress Agr-QS, were directed to residues N201 and R198 of AgrA, respectively, whereas these carbonyl oxygens of physalins, without Agr-QS suppression activity, were oriented in different directions. Next, 100-ns molecular dynamics simulations revealed that the hydrogen bond formed between the carbonyl oxygen at C-15 of physalins and L186 of AgrA functions as an anchor, sustaining the interaction between the carbonyl oxygen at C-1 of physalins and N201 of AgrA. Thus, these results suggest that physalin H, physalin B, and isophysalin B inhibit the interaction of AgrA with the agr promoters by binding to the DNA-binding site of AgrA, suppressing the Agr-QS function of S. aureus. Physalins that suppress the Agr-QS function are proposed as potential lead compounds in the anti-virulence strategy for MRSA infections.

Myxococcus xanthus translesion DNA synthesis protein ImuA is an ATPase enhanced by DNA.

Translesion DNA synthesis pathways are necessary to ensure bacterial replication in the presence of DNA damage. Translesion DNA synthesis carried out by the PolV mutasome is well-studied in Escherichia coli, but ~one third of bacteria use a functionally homologous protein complex, consisting of ImuA, ImuB, and ImuC (also called DnaE2). Numerous in vivo studies have shown that all three proteins are required for translesion DNA synthesis and that ImuC is the error-prone polymerase, but the roles of ImuA and ImuB are unclear. Here we carry out biochemical characterization of ImuA and a truncation of ImuB from Myxococcus xanthus. We find that ImuA is an ATPase, with ATPase activity enhanced in the presence of DNA. The ATPase activity is likely regulated by the C-terminus, as loss of the ImuA C-terminus results in DNA-independent ATP hydrolysis. We also find that ImuA binds a variety of DNA substrates, with DNA binding affinity affected by the addition of ADP or adenylyl-imidodiphosphate. An ImuB truncation also binds DNA, with lower affinity than ImuA. In the absence of DNA, ImuA directly binds ImuB with moderate affinity. Finally, we show that ImuA and ImuB self-interact, but that ImuA is predominantly a monomer, while truncated ImuB is a trimer in vitro. Together, with our findings and the current literature in the field, we suggest a model for translesion DNA synthesis, where a trimeric ImuB would provide sufficient binding sites for DNA, the β-clamp, ImuC, and ImuA, and where ImuA ATPase activity may regulate assembly and disassembly of the translesion DNA synthesis complex.

Caenorhabditis elegans telomere-binding proteins TEBP-1 and TEBP-2 adapt the Myb module to dimerize and bind telomeric DNA

Protecting chromosome ends from misrecognition as double-stranded (ds) DNA breaks is fundamental to eukaryotic viability. The protein complex shelterin prevents a DNA damage response at mammalian telomeres. Mammalian shelterin proteins TRF1 and TRF2 and their homologs in yeast and protozoa protect telomeric dsDNA. N-terminal homodimerization and C-terminal Myb-domain-mediated dsDNA binding are two structural hallmarks of end protection by TRF homologs. Yet our understanding of how Caenorhabditis elegans protects its telomeric dsDNA is limited. Recently identified C. elegans proteins TEBP-1 (also called DTN-1) and TEBP-2 (also called DTN-2) are functional homologs of TRF proteins, but how they bind DNA and whether or how they dimerize is not known. TEBP-1 and TEBP-2 harbor three Myb-containing domains (MCDs) and no obvious dimerization domain. We demonstrate biochemically that only the third MCD binds DNA. We solve the X-ray crystal structure of TEBP-2 MCD3 with telomeric dsDNA to reveal the structural mechanism of telomeric dsDNA protection in C. elegans. Mutagenesis of the DNA-binding site of TEBP-1 and TEBP-2 compromises DNA binding in vitro, and increases DNA damage signaling, lengthens telomeres, and decreases brood size in vivo. Via an X-ray crystal structure, biochemical validation of the dimerization interface, and SEC-MALS analysis, we demonstrate that MCD1 and MCD2 form a composite dimerization module that facilitates not only TEBP-1 and TEBP-2 homodimerization but also heterodimerization. These findings provide fundamental insights into C. elegans telomeric dsDNA protection and highlight how different eukaryotes have evolved distinct strategies to solve the chromosome end protection problem.

A PadR family transcriptional repressor regulates the transcription of chromate efflux transporter in Enterobacter sp. Z1.

Chromium is a prevalent toxic heavy metal, and chromate [Cr(VI)] exhibits high mutagenicity and carcinogenicity. The presence of the Cr(VI) efflux protein ChrA has been identified in strains exhibiting resistance to Cr(VI). Nevertheless, certain strains of bacteria that are resistant to Cr(VI) lack the presence of ChrB, a known regulatory factor. Here, a PadR family transcriptional repressor, ChrN, has been identified as a regulator in the response of Enterobacter sp. Z1(CCTCC NO: M 2019147) to Cr(VI). The chrN gene is cotranscribed with the chrA gene, and the transcriptional expression of this operon is induced by Cr(VI). The binding capacity of the ChrN protein to Cr(VI) was demonstrated by both the tryptophan fluorescence assay and Ni-NTA purification assay. The interaction between ChrN and the chrAN operon promoter was validated by reporter gene assay and electrophoretic mobility shift assay. Mutation of the conserved histidine residues His14 and His50 resulted in loss of ChrN binding with the promoter of the chrAN operon. This observation implies that these residues are crucial for establishing a DNA-binding site. These findings demonstrate that ChrN functions as a transcriptional repressor, modulating the cellular response of strain Z1 to Cr(VI) exposure.

Abstract LB257: Possible involvement of PDL1 intrinsic signaling to acquired resistance to immunotherapy for renal cell carcinoma (RCC)

Abstract Immune checkpoint inhibitor (ICI) therapy has shown unprecedented results in RCC treatment, but durable responses have been hampered by acquired resistance (AR, initially respond but eventually develop cancer progression). Mechanisms for AR remain ambiguous. Previously, we reported elevated blood GPNMB to strongly associate with AR development. To study molecular mechanisms for AR-linked GPNMB elevation, we created an ICI-resistant RenCa (R-RenCa) mouse model by repeated selection of largest tumors in mice treated with PD1 Ab ICI. Parental cells (Gpnmblow) respond to ICI (S-RenCa), while R-RenCa is completely unresponsive to ICI and acquired Gpnmb overexpression (but PDL1 remained unchanged), recapitulating strong association of GPNMB with AR development. R-RenCa tumors displayed more immunosuppressive and angiogenic phenotypes and enhanced response to anti-Gpnmb Ab. In probing for genetic alterations unique to R-RenCa, whole exome seq analysis showed R-RenCa to bear a greater tumor mutation burden (184.9 and 174.6 mut/Mb for R- and S-RenCa, respectively) but no somatic mutations in beta-2 microglobulin and IFN-signal molecule genes known to be involved in AR. RNA-seq analysis showed; more transcriptionally active genes in R-RenCa tumors (heatmap analysis); more responsive genes to ICI (Volcano analysis); and dysregulation in microphthalmia (Mitf) family gene expression. Mitf RNA was upregulated 3-fold, but Tfe3 and other members downregulated 2-3-fold. RNA start sites of Mitf transcripts are highly heterogenous in R-RenCa compared to S-RenCa (Sashimi plot), indicating generation of more isoforms. Since Gpnmb promoter has E-boxes (DNA binding sites) for Mitf factor, we examined whether Mitf is responsible for Gpnmb expression using promoter-Luc reporter assays and found 4-fold higher in R-RenCa, which was lost completely in E-box-mutated promoter. In seeking transcriptional activators for Mitf gene, we found Sox10 factor to be upregulated in R-RenCa. The molecular link of Sox10-Mitf-Gpnmb was shown by the data of gene-knock down experiments that Sox10-siRNA diminished Mitf and Gpnmb expression completely, and Mitf-siRNA had no effect on Sox10 expression but diminished Gpnmb expression. We then studied mechanisms for Sox10 activation and found PDL1 signaling (triggered by Ab-crosslinking) to induce Sox10 RNA expression in tumor cells, which was 80% abrogated by deleting the RMLDVEKC cytoplasmic motif known to protect tumor cells from IFN cytotoxicity. For human relevance of Sox10-Mitf axis activation, we employed nested qRT-PCR of cell-free RNA (cfRNA) isolated from plasma of RCC patients (n=3) immediately before and after AR. SOX10 and MITF cfRNAs were increased by 2.3 ± 0.85 and 1.77 ± 0.05-fold, respectively, between these two time points. Thus, PDL1 signaling is involved in AR by dysregulating the SOX10-MITF-GPNMB cascade, indicating the potential of the cascade-targeting to break AR hindrance for ICI therapy. Citation Format: Jin-Sung Chung, Lei Guo, Vinita Popat, Ponciano D. Cruz, Lin Xu, Hans Hammers, Kiyoshi Ariizumi. Possible involvement of PDL1 intrinsic signaling to acquired resistance to immunotherapy for renal cell carcinoma (RCC) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr LB257.

Extensive Loss and Gain of Conserved Noncoding Elements During Early Teleost Evolution.

Conserved noncoding elements in vertebrates are enriched around transcription factor loci associated with development. However, loss and rapid divergence of conserved noncoding elements has been reported in teleost fish, albeit taking only few genomes into consideration. Taking advantage of the recent increase in high-quality teleost genomes, we focus on studying the evolution of teleost conserved noncoding elements, carrying out targeted genomic alignments and comparisons within the teleost phylogeny to detect conserved noncoding elements and reconstruct the ancestral teleost conserved noncoding elements repertoire. This teleost-centric approach confirms previous observations of extensive vertebrate conserved noncoding elements loss early in teleost evolution, but also reveals massive conserved noncoding elements gain in the teleost stem-group over 300 million years ago. Using synteny-based association to link conserved noncoding elements to their putatively regulated target genes, we show the most teleost gained conserved noncoding elements are found in the vicinity of orthologous loci involved in transcriptional regulation and embryonic development that are also associated with conserved noncoding elements in other vertebrates. Moreover, teleost and vertebrate conserved noncoding elements share a highly similar motif and transcription factor binding site vocabulary. We suggest that early teleost conserved noncoding element gains reflect a restructuring of the ancestral conserved noncoding element repertoire through both extreme divergence and de novo emergence. Finally, we support newly identified pan-teleost conserved noncoding elements have potential for accurate resolution of teleost phylogenetic placements in par with coding sequences, unlike ancestral only elements shared with spotted gar. This work provides new insight into conserved noncoding element evolution with great value for follow-up work on phylogenomics, comparative genomics, and the study of gene regulation evolution in teleosts.

TBCA: Prediction of transcription factor binding sites using a deep neural network with lightweight attention mechanism.

The identification of transcription factor binding sites (TFBSs) is crucial for understanding the regulatory mechanisms of gene expression, which contributes to unraveling cellular functions and disease development. Currently, the most common approach involves the use of deep learning techniques to predict TFBSs by combining sequence and shape features. Although significant progress has been made with these methods, the integration of local features extracted from DNA sequences and shapes with global features has not yet reached a sufficient level, and there is still significant room for improvement in the accuracy of prediction results. In this paper, we propose a novel framework based on convolution and attention mechanisms, referred to as TBCA, which combines DNA sequence information and shape information for predicting transcription factor binding sites. In this work, we employ a two-layer convolutional neural network (CNNs) and self-attention mechanism to extract complex sequence features from DNA. What's more, we utilize a Fourier-transform-enhanced multi-head attention along with channel attention to extract high-order shape features of DNA. Finally, these high-order sequence and shape features are integrated into the channel dimension to achieve accurate TFBSs prediction. Our research results demonstrate that TBCA exhibits superior predictive performance in 165 validated ChIP-seq datasets. Furthermore, the employed attention mechanisms can automatically learn important features at different positions and scales, enhancing the accuracy and robustness of feature representation. We also conduct an in-depth analysis of the contributions of five different shapes to site prediction, revealing that shape features can enhance the prediction of transcription factor DNA binding.

The TELOMERE REPEAT BINDING proteins TRB4 and TRB5 function as transcriptional activators of PRC2-controlled genes to regulate plant development

Plant-specific transcriptional regulators called TELOMERE REPEAT BINDING proteins (TRBs) combine two DNA-binding domains, the GH1 domain, which binds to linker DNA and is shared with H1 histones, and the Myb/SANT domain, which specifically recognizes the telobox DNA-binding site motif. TRB1, TRB2, and TRB3 proteins recruit Polycomb group complex 2 (PRC2) to deposit H3K27me3 and JMJ14 to remove H3K4me3 at gene promoters containing telobox motifs to repress transcription. Here, we demonstrate that TRB4 and TRB5, two related paralogs belonging to a separate TRB clade conserved in spermatophytes, regulate the transcription of several hundred genes involved in developmental responses to environmental cues. TRB4 binds to several thousand sites in the genome, mainly at transcription start sites and promoter regions of transcriptionally active and H3K4me3-marked genes, but, unlike TRB1, it is not enriched at H3K27me3-marked gene bodies. However, TRB4 can physically interact with the catalytic components of PRC2, SWINGER, and CURLY LEAF (CLF). Unexpectedly, we show that TRB4 and TRB5 are required for distinctive phenotypic traits observed in clf mutant plants and thus function as transcriptional activators of several hundred CLF-controlled genes, including key flowering genes. We further demonstrate that TRB4 shares multiple target genes with TRB1 and physically and genetically interacts with members of both TRB clades. Collectively, these results reveal that TRB proteins engage in both positive and negative interactions with other members of the family to regulate plant development through both PRC2-dependent and -independent mechanisms.