Decoding gene regulation in the fly brain.

The Drosophila brain is a frequently used model in neuroscience. Single-cell transcriptome analysis1-6, three-dimensional morphological classification7 and electron microscopy mapping of the connectome8,9 have revealed an immense diversity of neuronal and glial cell types that underlie an array of functional and behavioural traits in the fly. The identities of these cell types are controlled by gene regulatory networks (GRNs), involving combinations of transcription factors that bind to genomic enhancers to regulate their target genes. Here, to characterize GRNs at the cell-type level in the fly brain, we profiled the chromatin accessibility of 240,919 single cells spanning 9 developmental timepoints and integrated these data with single-cell transcriptomes. We identify more than 95,000 regulatory regions that are used in different neuronal cell types, of which 70,000 are linked to developmental trajectories involving neurogenesis, reprogramming and maturation. For 40 cell types, uniquely accessible regions were associated with their expressed transcription factors and downstream target genes through a combination of motif discovery, network inference and deep learning, creating enhancer GRNs. The enhancer architectures revealed by DeepFlyBrain lead to a better understanding of neuronal regulatory diversity and can be used to design genetic driver lines for cell types at specific timepoints, facilitating their characterization and manipulation.

Interpretation of allele-specific chromatin accessibility using cell state-aware deep learning.

Genomic sequence variation within enhancers and promoters can have a significant impact on the cellular state and phenotype. However, sifting through the millions of candidate variants in a personal genome or a cancer genome, to identify those that impact cis-regulatory function, remains a major challenge. Interpretation of noncoding genome variation benefits from explainable artificial intelligence to predict and interpret the impact of a mutation on gene regulation. Here we generate phased whole genomes with matched chromatin accessibility, histone modifications, and gene expression for 10 melanoma cell lines. We find that training a specialized deep learning model, called DeepMEL2, on melanoma chromatin accessibility data can capture the various regulatory programs of the melanocytic and mesenchymal-like melanoma cell states. This model outperforms motif-based variant scoring, as well as more generic deep learning models. We detect hundreds to thousands of allele-specific chromatin accessibility variants (ASCAVs) in each melanoma genome, of which 15%-20% can be explained by gains or losses of transcription factor binding sites. A considerable fraction of ASCAVs are caused by changes in AP-1 binding, as confirmed by matched ChIP-seq data to identify allele-specific binding of JUN and FOSL1. Finally, by augmenting the DeepMEL2 model with ChIP-seq data for GABPA, the TERT promoter mutation, as well as additional ETS motif gains, can be identified with high confidence. In conclusion, we present a new integrative genomics approach and a deep learning model to identify and interpret functional enhancer mutations with allelic imbalance of chromatin accessibility and gene expression.

Cross-species analysis of enhancer logic using deep learning.

Deciphering the genomic regulatory code of enhancers is a key challenge in biology because this code underlies cellular identity. A better understanding of how enhancers work will improve the interpretation of noncoding genome variation and empower the generation of cell type-specific drivers for gene therapy. Here, we explore the combination of deep learning and cross-species chromatin accessibility profiling to build explainable enhancer models. We apply this strategy to decipher the enhancer code in melanoma, a relevant case study owing to the presence of distinct melanoma cell states. We trained and validated a deep learning model, called DeepMEL, using chromatin accessibility data of 26 melanoma samples across six different species. We show the accuracy of DeepMEL predictions on the CAGI5 challenge, where it significantly outperforms existing models on the melanoma enhancer of IRF4 Next, we exploit DeepMEL to analyze enhancer architectures and identify accurate transcription factor binding sites for the core regulatory complexes in the two different melanoma states, with distinct roles for each transcription factor, in terms of nucleosome displacement or enhancer activation. Finally, DeepMEL identifies orthologous enhancers across distantly related species, where sequence alignment fails, and the model highlights specific nucleotide substitutions that underlie enhancer turnover. DeepMEL can be used from the Kipoi database to predict and optimize candidate enhancers and to prioritize enhancer mutations. In addition, our computational strategy can be applied to other cancer or normal cell types.

Identification of genomic enhancers through spatial integration of single-cell transcriptomics and epigenomics.

Single-cell technologies allow measuring chromatin accessibility and gene expression in each cell, but jointly utilizing both layers to map bona fide gene regulatory networks and enhancers remains challenging. Here, we generate independent single-cell RNA-seq and single-cell ATAC-seq atlases of the Drosophila eye-antennal disc and spatially integrate the data into a virtual latent space that mimics the organization of the 2D tissue using ScoMAP (Single-Cell Omics Mapping into spatial Axes using Pseudotime ordering). To validate spatially predicted enhancers, we use a large collection of enhancer-reporter lines and identify ~ 85% of enhancers in which chromatin accessibility and enhancer activity are coupled. Next, we infer enhancer-to-gene relationships in the virtual space, finding that genes are mostly regulated by multiple, often redundant, enhancers. Exploiting cell type-specific enhancers, we deconvolute cell type-specific effects of bulk-derived chromatin accessibility QTLs. Finally, we discover that Prospero drives neuronal differentiation through the binding of a GGG motif. In summary, we provide a comprehensive spatial characterization of gene regulation in a 2D tissue.

Biallelic missense variants in ZBTB11 can cause intellectual disability in humans.

Exploring genes and pathways underlying intellectual disability (ID) provides insight into brain development and function, clarifying the complex puzzle of how cognition develops. As part of ongoing systematic studies to identify candidate ID genes, linkage analysis and next-generation sequencing revealed Zinc Finger and BTB Domain Containing 11 (ZBTB11) as a novel candidate ID gene. ZBTB11 encodes a little-studied transcription regulator, and the two identified missense variants in this study are predicted to disrupt canonical Zn2+-binding residues of its C2H2 zinc finger domain, leading to possible altered DNA binding. Using HEK293T cells transfected with wild-type and mutant GFP-ZBTB11 constructs, we found the ZBTB11 mutants being excluded from the nucleolus, where the wild-type recombinant protein is predominantly localized. Pathway analysis applied to ChIP-seq data deposited in the ENCODE database supports the localization of ZBTB11 in nucleoli, highlighting associated pathways such as ribosomal RNA synthesis, ribosomal assembly, RNA modification and stress sensing, and provides a direct link between subcellular ZBTB11 location and its function. Furthermore, given the report of prominent brain and spinal cord degeneration in a zebrafish Zbtb11 mutant, we investigated ZBTB11-ortholog knockdown in Drosophila melanogaster brain by targeting RNAi using the UAS/Gal4 system. The observed approximate reduction to a third of the mushroom body size-possibly through neuronal reduction or degeneration-may affect neuronal circuits in the brain that are required for adaptive behavior, specifying the role of this gene in the nervous system. In conclusion, we report two ID families segregating ZBTB11 biallelic mutations disrupting Zn2+-binding motifs and provide functional evidence linking ZBTB11 dysfunction to this phenotype.

CNKSR1 gene defect can cause syndromic autosomal recessive intellectual disability.

The advent of high-throughput sequencing technologies has led to an exponential increase in the identification of novel disease-causing genes in highly heterogeneous diseases. A novel frameshift mutation in CNKSR1 gene was detected by Next-Generation Sequencing (NGS) in an Iranian family with syndromic autosomal recessive intellectual disability (ARID). CNKSR1 encodes a connector enhancer of kinase suppressor of Ras 1, which acts as a scaffold component for receptor tyrosine kinase in mitogen-activated protein kinase (MAPK) cascades. CNKSR1 interacts with proteins which have already been shown to be associated with intellectual disability (ID) in the MAPK signaling pathway and promotes cell migration through RhoA-mediated c-Jun N-terminal kinase (JNK) activation. Lack of CNKSR1 transcripts and protein was observed in lymphoblastoid cells derived from affected patients using qRT-PCR and western blot analysis, respectively. Furthermore, RNAi-mediated knockdown of cnk, the CNKSR1 orthologue in Drosophila melanogaster brain, led to defects in eye and mushroom body (MB) structures. In conclusion, our findings support the possible role of CNKSR1 in brain development which can lead to cognitive impairment.

Single-cell spatial multi-omics and deep learning dissect enhancer-driven gene regulatory networks in liver zonation.

In the mammalian liver, hepatocytes exhibit diverse metabolic and functional profiles based on their location within the liver lobule. However, it is unclear whether this spatial variation, called zonation, is governed by a well-defined gene regulatory code. Here, using a combination of single-cell multiomics, spatial omics, massively parallel reporter assays and deep learning, we mapped enhancer-gene regulatory networks across mouse liver cell types. We found that zonation affects gene expression and chromatin accessibility in hepatocytes, among other cell types. These states are driven by the repressors TCF7L1 and TBX3, alongside other core hepatocyte transcription factors, such as HNF4A, CEBPA, FOXA1 and ONECUT1. To examine the architecture of the enhancers driving these cell states, we trained a hierarchical deep learning model called DeepLiver. Our study provides a multimodal understanding of the regulatory code underlying hepatocyte identity and their zonation state that can be used to engineer enhancers with specific activity levels and zonation patterns.

Analysis of long and short enhancers in melanoma cell states.

Understanding how enhancers drive cell-type specificity and efficiently identifying them is essential for the development of innovative therapeutic strategies. In melanoma, the melanocytic (MEL) and the mesenchymal-like (MES) states present themselves with different responses to therapy, making the identification of specific enhancers highly relevant. Using massively parallel reporter assays (MPRAs) in a panel of patient-derived melanoma lines (MM lines), we set to identify and decipher melanoma enhancers by first focusing on regions with state-specific H3K27 acetylation close to differentially expressed genes. An in-depth evaluation of those regions was then pursued by investigating the activity of overlapping ATAC-seq peaks along with a full tiling of the acetylated regions with 190 bp sequences. Activity was observed in more than 60% of the selected regions, and we were able to precisely locate the active enhancers within ATAC-seq peaks. Comparison of sequence content with activity, using the deep learning model DeepMEL2, revealed that AP-1 alone is responsible for the MES enhancer activity. In contrast, SOX10 and MITF both influence MEL enhancer function with SOX10 being required to achieve high levels of activity. Overall, our MPRAs shed light on the relationship between long and short sequences in terms of their sequence content, enhancer activity, and specificity across melanoma cell states.