Hazuki Takahashi

Cell differentiation towards neurons is accompanied by widespread changes in three-dimensional (3D) genome organization and gene expression. Chromatin-associated RNAs have been proposed to be important regulators of such changes; however, the type, abundance, and role of these RNAs during neuronal differentiation remain largely unexplored. Here, we integrate multi-omic data generated in the frame of the Functional ANnoTation Of the Mammalian genome (FANTOM6) to chart 3D genome, RNA-DNA contactome, and transcriptome changes occurring during in vitro differentiation of human induced pluripotent stem cells to neural stem cells and neurons. We reveal a previously unreported phenomenon, in which intronic RNAs engage in long-distance contacts with DNA loci distributed all over the genome. These trans-contacting intronic RNAs (TIRs) are produced from exceptionally long (mean length: 750 kilobases, kb) protein-coding genes that carry ultra-long introns and are selectively expressed in neurons. We show that TIRs do not undergo rapid co-transcriptional degradation but rather accumulate in the nucleus of neuronal cells, forming large ‘dot clouds’ around their source loci and spreading across the nucleus, as visualized by single-molecule RNA fluorescence in situ hybridization. TIRs engage in contacts with a set of genomic regions (TIR-contacted regions or TIRCs) that carry much shorter (mean length: ~30 kb) neuronally expressed genes forming high-connectivity hubs. We also show that the expression of genes within TIRCs contacted by the same set of TIRs is highly co-varied, and that TIR source genes, especially their introns, are enriched in genetic risk loci for neurodevelopmental and neuropsychiatric disorders. Our findings point to a functional role of the persistence of long intronic RNAs in the nucleus of neuronal cells and might contribute to explain why neurons uniquely express many ultra-long genes. We propose a model in which TIRs form pan-nuclear scaffolds—which we propose to name introsomes—that constitute a dynamically self-renewing regulatory layer, where transcription itself continuously regenerates the very scaffold that organizes genome function in neurons, and that might be involved in the pathogenesis of neurodevelopmental and neuropsychiatric disorders.

Repeat elements (REs) comprise over half of the human genome and are increasingly recognized as regulatory components, yet their role in therapy resistance remains unclear. Here, we integrate CAGE and RADICL-seq to map regulatory elements and RNA-chromatin interactions in estrogen receptor-positive breast cancer cells (MCF7) and their long-term estrogen-deprived (LTED) derivatives. We identify repeat element-anchored enhancer-promoter interactions (REEPIs), defined by the presence of annotated REs at both enhancer and promoter sites. In LTED cells, REEPIs are enriched for ERV1 and ERVL elements and involve specific repeat pairings associated with transcription factor combinations such as SP1-SMAD3 and SP1-NFYC. At the gene level, these REEPIs are preferentially linked to genes involved in chromatin remodeling, DNA repair, and lineage plasticity. Our findings reveal a repeat-guided regulatory architecture associated with transcriptional adaptation in endocrine resistance and identify REEPIs as a previously underexplored component of regulatory reprogramming in hormone-resistant breast cancer.

More than twenty years ago, the FANTOM consortium uncovered that mammalian genomes are pervasively transcribed, revealing multitudes of RNAs with unknown functions. A subset of these transcripts has since then been linked to transcriptional control and to chromatin organization via their ability to interact with DNA, suggesting that chromatin-associated RNAs could be key players in genome regulation. Although recent technological advances now enable the mapping of genome-wide RNA-DNA contacts, a lack of analyses integrating these methods with other genomic features and across multiple cellular contexts hinders our comprehensive understanding of the principles underlying RNA-DNA interactions and of their biological importance. As part of the FANTOM6 project, we thus generated RNA-DNA interaction maps in 16 different human cell types, then combined these contacts with multiple layers of other genomic data to investigate how patterns of interaction between RNA and DNA relate to chromatin organization and function. We show that the RNA-DNA interactome is highly dynamic yet reproducibly organized in cell-type specific networks, constituted of a great diversity of interactions that vary in function of their distance, the nature of their sources and the chromatin state of their targets. In particular, we detected numerous regulatory elements that exhibit marked changes in activity when differentially bound by transcripts, implying that thousands of RNA-DNA interactions can play a mechanistic role in gene expression. This regulatory function correlates with RNA-protein interactions and significantly associates with cell type-relevant and disorder-related traits. In addition to providing essential resources for future research in RNA-mediated chromatin regulation, cellular biology and human diseases, our study thus establishes the RNA-DNA interactome as a new genome regulatory layer that defines and maintains cellular identity and behavior.

Noncoding RNAs (ncRNAs) regulate gene expression through RNA–chromatin interactions, yet their role in human brain evolution remains unclear. To map these interactions, we integrated FANTOM6 neurogenic data with a comprehensive atlas of regulatory regions active during early cortical development. This integration effort linked ncRNAs and transcription factors (TFs) to their genomic targets, revealing functional interactions captured by RNA-DNA contact mapping, yet largely missed by chromatin conformation data. Overlaying these interactions with nearly fixed variants in Homo sapiens relative to extinct hominins uncovered widespread rewiring of TF binding, affecting genes implicated in progenitor proliferation and neuronal differentiation. Gene regulatory network reconstruction identified TEAD2 and ONECUT2 as key regulators of intermediate progenitors and migrating neurons. Functional perturbation followed by Cut&Tag and ATAC-seq linked the two TFs to disease- and evolutionarily- relevant pathways. Together, these results show how ncRNA–TF cooperation and selective cis -regulatory rewiring contributed to sapiens-specific features of cortical development.

Macrophages are innate immune cells present in most tissues of the body, whose molecular programs are determined by their ontogeny and environment. From the earliest stages of embryonic development, macrophages are recruited into developing tissues where they support organogenesis with trophic factors such as WNT, VEGF and PDGF. While macrophage subsets have been described in different tissues at single cell resolution, little is known about transcript isoforms and proteoforms that underpin their differentiation and function. Here we assessed enhancer, promoter, transcript and proteomic variation as pluripotent stem cells differentiate to macrophages, identifying over 200 previously uncharacterised genes and over 20,000 new mRNA isoforms, updating our current understanding of the human genome, its regulation and potential output. Newly discovered myeloid-expressed transcripts and proteins were enriched for motifs associated with secreted proteins, and these included previously uncharacterised isoforms of growth factors, in which we predict N-terminal changes impact on their location and function. Activation of primary adult monocytes and monocyte-derived macrophages was also characterised by the expression of diverse transcript isoforms, largely arising from alternate transcription initiation sites and predicted to impact on the acute response to bacterial or fungal stimuli. Understanding the full spectrum of gene products expressed by these cells further extends our understanding of the phenotypic plasticity and trophic potential of macrophages in human development and may lead to the discovery of new clinical targets for tissue engineering or immune-related studies. Graphical In this manuscript, Berrocal-Rubio and colleagues examined the differentiation of human macrophages using an iPSC model of tissue macrophage biology. Combining long-read sequencing technology with promoter profiling identified over 17, 700 genes implicated in pluripotency-myeloid specification. 7% of transcripts profiled from previously characterised genes were predicted to encode new proteins, and a further 3% of transcripts were derived from genes newly discovered in this project. They confirmed that a high proportion of these alternate transcripts were detectable in primary monocytes but also discovered that activation of primary monocytes led to further alternate promoter usage, with the potential to further diversify the innate immune responses to a broad set of pathogens. The newly described macrophage genes and transcripts encoded proteins enriched for motifs associated with secreted peptides. These data suggest that alternate transcription of macrophage genes leads to new effectors of innate immune function, that include a substantially expanded number of growth factors or secreted proteins. Created in BioRender. Berrocal, M. (2025) https://BioRender.com/d0n47le

Cis-regulatory networks are essential for determining gene transcriptional states, yet the mechanisms that mediate cellular signaling and enhancer-promoter communications are not yet well understood. Here, we integrate high-resolution enhancer-promoter interactions obtained using targeted chromosome conformation capture and RNA-DNA contacts from RADICL-seq to uncover the role of RNA in modulating enhancer-promoter interactions. Both datasets were generated from human induced pluripotent cells at different stages across neural differentiation. As expected, most enhancer-promoter interactions were dynamic with only 38% shared across all cell states and 68% of promoter interacting regions overlapped at least one annotated neural enhancer. Integration with RADICL-seq data revealed a 9.3-fold enrichment for cis-interacting DNA regions to associate with at least one RNA. Among 18,346 cis-interacting promoters, only 1,170 (6.3%) lacked any RNA association, whereas 3,702 (28%) showed stable RNA association and 13,579 (74%) displayed dynamic RNA association across differentiation. Promoters associated with RNA engaged with higher number of enhancers and exhibited higher expression levels, while enhancers with RNA associations showed higher levels of chromatin activation marks. Promoters with dynamic RNA association were also more likely to be differentially expressed and enriched for relevant biological processes, phenotypes and diseases. Importantly, RNA-DNA association dynamics correlated strongly with DNA-DNA interaction dynamics; gain of RNA association on enhancers was typically accompanied by enhancer-promoter interaction gain, whereas concurrent RNA association gain on both promoter and enhancers frequently led to interaction loss. The correlation of RNA association on cis-interaction dynamics across differentiation were also captured in the community structure of the DNA-RNA network. Together, our results reveal extensive coupling between RNA-DNA and DNA-DNA networks supporting a coordinating role of RNA in gene regulation.

Gene expression is controlled by proximal and distal cis-regulatory elements (CREs), containing DNA motifs bound by various transcription factors (TFs). Other sequence features, such as specific k-mers or low complexity regions, have also been implicated [1–3]. However, in a dynamic biological process such as cell differentiation, we lack an understanding of how the transcriptional activity of CREs progressively change and what sequence features underlie these transitions, which may reflect common and/or coordinated regulatory processes. Here, we use single-cell ATAC-seq and RNA-seq to follow, at a genome scale, CREs along differentiation of induced pluripotent stem cells into cortical neurons and develop a method to automatically identify the diversity of CRE profiles and their underlying sequence features. We propose a machine-learning guided clustering algorithm, STOIC (Statistical learning TO Inform Clustering), that jointly learns an unsupervised clustering of the CREs in the space of the activity profiles and a supervised predictor associated with each cluster in the DNA-sequence space. STOIC is specifically designed to provide readily interpretable results. We show that the method identifies CRE profiles associated with highly predictive sequence features and outperforms methods solely concerned with co-activity clustering on this task. Orthogonal data collected in the same settings link the inferred CRE clusters to specific enhancer or promoter signatures. Furthermore, we show that the DNA features unveiled by STOIC reflect biologically relevant regulators and offer a valuable basis to dissect elements of the cisregulatory grammar. Finally, we demonstrate the general applicability of STOIC by analyzing five bulk CAGE datasets of human cells responding to various treatments.

Cell differentiation towards neurons is accompanied by widespread changes in three-dimensional (3D) genome organization and gene expression. Chromatin-associated RNAs have been proposed to be important regulators of such changes; however, the type, abundance, and role of these RNAs during neuronal differentiation remain largely unexplored. Here, we integrate multi-omic data generated in the frame of the Functional ANnoTation Of the Mammalian genome (FANTOM6) to chart 3D genome, RNA-DNA contactome, and transcriptome changes occurring during in vitro differentiation of human induced pluripotent stem cells to neural stem cells and neurons. We reveal a previously unreported phenomenon, in which intronic RNAs engage in long-distance contacts with DNA loci distributed all over the genome. These trans-contacting intronic RNAs (TIRs) are produced from exceptionally long (mean length: 750 kilobases, kb) protein-coding genes that carry ultra-long introns and are selectively expressed in neurons. We show that TIRs do not undergo rapid co-transcriptional degradation but rather accumulate in the nucleus of neuronal cells, forming large ‘dot clouds’ around their source loci and spreading across the nucleus, as visualized by single-molecule RNA fluorescence in situ hybridization. TIRs engage in contacts with a set of genomic regions (TIR-contacted regions or TIRCs) that carry much shorter (mean length: ∼30 kb) neuronally expressed genes forming high-connectivity hubs. We also show that the expression of genes within TIRCs contacted by the same set of TIRs is highly co-varied, and that TIR source genes, especially their introns, are enriched in genetic risk loci for neurodevelopmental and neuropsychiatric disorders. Our findings point to a functional role of the persistence of long intronic RNAs in the nucleus of neuronal cells and might contribute to explain why neurons uniquely express many ultra-long genes. We propose a model in which TIRs form pan-nuclear scaffolds—which we propose to name introsomes—that constitute a dynamically self-renewing regulatory layer, where transcription itself continuously regenerates the very scaffold that organizes genome function in neurons, and that might be involved in the pathogenesis of neurodevelopmental and neuropsychiatric disorders.

Long-read sequencing has transformed transcriptome profiling, yet capturing full-length, non-polyadenylated transcripts like enhancer RNAs (eRNAs) remains challenging. Here, we introduce CFC-seq, combining cap-trapping and in vitro poly(A)-tailing to sequence poly(A) and non-poly(A) RNAs with precise transcription start site. Paired with our assembler, SALA, we identified 39,425 novel transcriptional units, including ∼24,000 eRNAs. Our data reveal a distinct genomic code governing eRNA fate dictated by core promoter architecture. CpG-island enhancers show high chromatin connectivity but yield short, exosome-sensitive RNAs. Conversely, TATA-box enhancers systematically co-opt LTR retrotransposons to inherit structural motifs that produce long, stable, and spliced RNAs. Mechanistically, the pioneer factor NF-Y activates these viral elements to license transcription, balanced by TEAD4 activity across a dual-gear regulatory axis. Finally, non-poly(A) eRNAs terminate via exosome-associated processing at structural-depleted cleavage zones. This comprehensive annotation links enhancer sequence architecture to RNA fate, providing a new transformative framework for decoding the functional human genome. Highlights: Expanded genomic architecture: CFC-seq unmasks a hidden layer of human transcriptome, identifying 39,425 novel transcriptional units with high-confidence TSS support, including ∼24,000 eRNAs. TSS-first assembler: We introduce SALA, a specialized long-read assembler that prioritizes authentic 5’ Cap-trapped ends to accurately reconstruct the TSS-resolved transcript models. Genomic code of eRNA fate: CGI enhancers drive short and exosome-sensitive transcripts associated with repressive H3K27me3 mark and high chromatin connectivity. TATA-box enhancers produce cell-type-specific, long, stable, and frequently spliced eRNAs. Evolutionary co-option of retrotransposons: A major fraction of TATA-box eRNAs originate from LTR retrotransposons, providing a direct mechanism for integration of viral elements into the human regulatory landscape. A dual-gear pioneering axis: The pioneer factor NF-Y activates unprimed LTR-TATA enhancers to license transcription independent of histone acetylation cascades, operating in parallel with TEAD4-mediated activation. Structural determinants of eRNA termination: Non-poly(A) eRNA TES features a secondary structure depletion zone that coordinates pol II termination and calibrates exosome-mediated turnover.