Biology subjects

Yi, R.

Publications and source records attributed to Yi, R..

At least 19 recordsLinked to original sources

bioRxiv2026

Desmoglein-1 coordinates cell proliferation and actin dynamics to promote epidermal morphogenesis.

Skin development depends on coordinated biochemical and mechanical signaling to form a multilayered epidermal barrier. Our previous work demonstrated that the stratified tissue-specific desmosomal cadherin Desmoglein-1 (Dsg1) promotes epidermal differentiation in human in vitro models. Dsg1 first appears in the developing mouse epidermis around E14, coinciding with a developmental transition in stratification and the onset of terminal differentiation, raising the question of how Dsg1 contributes to epidermal morphogenesis in vivo. We generated a wild-type (WT) single-cell RNA sequencing (scRNA-seq) atlas of murine skin and integrated scRNA-seq data from Dsg1 knockout (Dsg1 KO) embryos spanning E14.5, E15.5, E16.5, and E18.5. Six keratinocyte subpopulations were annotated, including a Transitional subpopulation co-expressing basal and suprabasal markers, positioned between basal and differentiated states. Dsg1 KO epidermis exhibited reduced thickness at E14.5-E15.5, but recovered by E16.5-E18.5, suggesting compensatory adaptation. Differential gene expression analysis revealed enrichment of cell-cycle and actin-associated programs in Dsg1 KO during late development. EdU incorporation and PCNA staining confirmed increased proliferation in superficial layers accompanied by altered actin organization and changed cell morphology. Notably, actin and the Arp2/3 actin nucleation-promoting factor N-WASP exhibited marked changes in distribution during early and late development, respectively. Mechanistically, Dsg1 associates with Nck1, an adaptor linking N-WASP to Arp2/3-dependent actin polymerization. Silencing Nck1 or N-WASP reduced Dsg1-mediated stratification significantly, while a Nck1-binding-deficient Dsg1 mutant failed to rescue stratification in Dsg1-deficient keratinocytes in vitro. Together, these findings identify a Dsg1-Nck1-N-WASP axis linking desmosomal adhesion to cytoskeletal remodeling and support a model in which Dsg1 coordinates cell-cycle, cell morphology, and actin dynamics to drive epidermal morphogenesis.

cell biology↗

bioRxiv2026

Multimodal Human Scalp Atlas Defines Cell Landscape and Lineage Architecture In Situ

Human scalp hair has an extraordinary ability to grow continuously for years while maintaining structural and functional integrity. However, the cell states and lineage organization that enable this capacity and how they are disrupted in inflammatory hair loss disorders remain poorly defined in humans. Here we establish a high-resolution, multimodal atlas of human scalp by integrating deep-coverage spatial transcriptomics with single-cell RNA-seq and multiomics data. This reference resolves spatially organized epithelial and mesenchymal states and links in situ transcriptional programs to chromatin accessibility dynamics and lineage trajectories at single-cell resolution, revealing human-specific principles of tissue organization and previously unrecognized features of hair follicle architecture and lineage progression. We validate key aspects of matrix cell organization and cell activities using live imaging, connecting molecularly defined cell states to dynamic cell behaviors and lineage progression in the matrix. Leveraging the atlas as a spatial reference, we project patient scRNA-seq profiles from alopecia areata and lichen planopilaris onto defined cell compartments, resolving disease-specific perturbations in fibroblasts, epithelial and immune populations. This comparison delineates distinct cellular programs associated with non-scarring versus scarring hair loss and highlights compartment- and state-specific pathways with diagnostic and therapeutic potential. Together, this work provides a foundational resource for human hair biology and establishes a generalizable framework for spatially resolved, multimodal interrogation of tissue organization and disease in complex human tissues.

Developmental Biology↗

bioRxiv2026

Coordinated inhibition of SOX9 and cell cycle progression by microRNA-200 restricts sebaceous gland fate specification

The microRNA-200 family (miR-200s) is widely recognized for their potent role in inhibiting epithelial-to-mesenchymal transition and cell cycle progression in cancer. However, their functional specificity in normal epithelial development remains poorly understood. Here we show that miR-200s are highly enriched in hair matrix progenitors but conspicuously absent from the upper hair follicle (HF), the anatomical location where sebaceous gland (SG) and HF stem cells are specified. We demonstrate that elevated miR-200 expression in this region abolishes SG fate specification while permitting hair morphogenesis. Genome-wide identification of miR-200 targets reveals that miR-200s regulate multiple negative regulators of WNT signaling, in addition to cell cycle regulators. Single-cell and spatial transcriptomic analyses uncover a mutually exclusive expression patterns between WNT activity and SOX9 in the upper HF, which is disrupted by miR-200 induction, resulting in compromised SOX9 function. Mechanistically, genome-wide identification of SOX9 targets uncovers a broad network of lipid and fatty acid metabolism genes critical for the transition of upper HF progenitors to the SG fate. The coordinated inhibition of the SOX9-dependent lipogenic program and the potent restriction of cell cycle progression, both mediated by miR-200s, collectively blocks SG specification. Taken together, this work reveals an unexpected specificity of miR-200 in restricting epithelial plasticity and elucidates a spatially defined SOX9 regulatory network essential for SG development.

developmental biology↗

bioRxiv2026

Structural insights into the recognition and catalysis of tRNA by human NSUN2

The human RNA m5C methyltransferase NSUN2 catalyzes site-specific cytosine methylation across diverse RNA substrates and thereby regulates a wide range of biological and physiological processes. However, the molecular basis by which NSUN2 achieves broad substrate recognition while maintaining catalytic specificity has remained unclear. Here we determine structures of human NSUN2 in both substrate-free and substrate-bound states using X-ray crystallography and cryo-electron microscopy. Structures of NSUN2 in complex with multiple tRNA substrates reveal a structure-first, sequence-tolerant strategy in which NSUN2 actively remodels tRNA architecture, exposing the buried target cytosine and positioning it within the catalytic pocket for methyl transfer. This recognition strategy enables NSUN2 to accommodate diverse tRNA substrates through a largely conserved interaction interface. Guided by these structural insights, we identify a small-molecule inhibitor that suppresses NSUN2 activity and cancer cell proliferation. Together, our findings define the molecular principles underlying NSUN2-mediated RNA m5C modification and provide a structural foundation for targeting NSUN2 in cancers.

molecular biology↗

bioRxiv2025

Compartmentalization of mRNA and Translation by Desmosomes

Subcellular compartmentalization enables cells to concentrate and locally regulate functions. Here we find that translational machinery is compartmentalized to the cell cortex in epidermal cells. Furthermore, we observed broad enrichment of transcripts at the cortex, establishing a novel axis of mRNA organization in these cells. Mechanistically, we identified the desmosomal protein, desmoplakin, to be essential for cortical recruitment of both ribosomes and mRNAs, through distinct mechanisms. While mRNA localization is usually proposed to promote local translation, we found that most cortical transcripts were translationally repressed. These findings reveal spatial and transcript-dependent features of the cortical translatome. One mediator of this regulation is the RNA-induced Silencing Complex (RISC) which is also cortically enriched in a desmoplakin-dependent manner. Under homeostatic conditions, cortical RISC associates with transcripts for cell adhesion and the cytoskeleton. Upon wounding RISC is delocalized from the cortex and its associated transcripts become translationally upregulated. Together our data demonstrate a desmosome-dependent cortical compartmentalization of translation that dynamically responds to barrier perturbations, including wounding.

cell biology↗

bioRxiv2023

MIIP downregulation promotes colorectal cancer progression via inducing adjacent adipocytes browning

The enrichment of peri-cancerous adipocytes is one of the distinctive features of colorectal cancer (CRC), which accelerates the disease progression and worsens the prognosis. The communication between tumor cells and adjacent adipocytes plays an important role in this process. However, the detailed mechanisms remain largely unknown. Here we demonstrated MIIP is downregulated in high-grade CRC, and revealed its role and mechanism in tumor cell-adipocyte communication. By detecting MIIP expression in CRC tissues and adjacent normal tissues, we found MIIP was significantly decreased in CRC, and was closely related to adjacent adipocytes browning. In detail, MIIP reduction altered the N-linked glycosylation modification of AZGP1 and thus alleviated the inhibition of secretion. AZGP1, a critical lipid mobilization factor, led to the intensification of adipocytes browning and the release of free fatty acids (FFAs), which in turn fueled for CRC progression. Our data demonstrate that MIIP plays a key regulatory role in the communication between CRC and neighboring adipocytes by mediating AZGP1 secretion, and MIIP reduction leads to adipose browning-induced cancer rapid progression and poor prognosis.

cancer biology↗

bioRxiv2023

Reconstruction of 3-dimensional tissue organization at the single-cell resolution

Recent advances in spatial transcriptomics (ST) have allowed for the mapping of tissue heterogeneity, but this technique lacks the resolution to investigate gene expression patterns, cell-cell communications and tissue organization at the single-cell resolution. ST data contains a mixed transcriptome from multiple heterogeneous cells, and current methods predict two-dimensional (2D) coordinates for individual cells within a predetermined space, making it difficult to reconstruct and study three-dimensional (3D) tissue organization. Here we present a new computational method called scHolography that uses deep learning to map single-cell transcriptome data to 3D space. Unlike existing methods, which generate a projection between transcriptome data and 2D spatial coordinates, scHolography uses neural networks to create a high-dimensional transcriptome-to-space map that preserves the distance information between cells, allowing for the construction of a cell-cell proximity matrix beyond the 2D ST scaffold. Furthermore, the neighboring cell profile of a given cell type can be extracted to study spatial cell heterogeneity. We apply scHolography to human skin, human skin cancer and mouse brain datasets, providing new insights into gene expression patterns, cell-cell interactions and spatial microenvironment. Together, scHolography offers a computational solution for digitizing transcriptome and spatial information into high-dimensional data for neural network-based mapping and the reconstruction of 3D tissue organization at the single-cell resolution.

bioinformatics↗

bioRxiv2022

MicroRNA-205 promotes hair regeneration by modulating mechanical properties of hair follicle stem cells

Stiffness and actomyosin contractility are intrinsic mechanical properties of animal cells required for the shaping of tissues. However, whether tissue stem cells (SCs) and progenitors located within SC niche have different mechanical properties that govern their size and functions remains unclear. We show that hair follicle SCs in the bulge are stiff with high actomyosin contractility and resistant to size change, whereas hair germ (HG) progenitors are soft and periodically enlarge and contract during quiescence. During activation, HGs reduce contraction and more frequently enlarge, a process that is associated with weakening of the actomyosin network, nuclear YAP accumulation and cell cycle re-entry. Induction of miR-205, a novel regulator of the actomyosin cytoskeleton, reduces actin contractility, decreases the stiffness of the bulge and HG, and activates hair regeneration in young and old mice. This study reveals the control of tissue SC size and activities by spatiotemporally compartmentalized mechanical properties and demonstrates the possibility to stimulate tissue regeneration by fine-tuning cell mechanics.

genetics↗

bioRxiv2022

MOF-mediated Histone H4 Lysine 16 Acetylation Governs Mitochondrial and Ciliary Functions by Controlling Gene Promoters

Histone H4 lysine 16 acetylation (H4K16ac), governed by the histone acetyltransferase (HAT) MOF, orchestrates critical functions in gene expression regulation and chromatin interaction. However, how does MOF and H4K16ac control cellular function and regulate mammalian tissue development remains unclear. Furthermore, whether the function of MOF is mediated by MSL or NSL, two distinct MOF-containing HAT complexes, have not been determined during mammalian development. Here we show that conditional deletion of Mof but not Kansl1, an essential component of the NSL complex, causes severe defects during murine skin development. In the absence of Mof and H4K16ac, basal epithelial progenitors of mammalian skin fail to establish the basement membrane and cell polarity, causing the failure of self-renewal. Furthermore, epidermal differentiation and hair growth are severely compromised, leading to barrier defects and perinatal lethality. Single-cell and bulk RNA-seq, in combination with MOF ChIP-seq, reveal that MOF regulated genes are highly enriched in mitochondria and cilia. Mechanistically, MOF coordinates with RFX2 transcription factor, which preferentially binds to gene promoters, to regulate ciliary and mitochondrial genes. Importantly, genetic deletion of Uqcrq, a nuclear-encoded, essential subunit for electron transport chain (ETC) Complex III, recapitulates the defects of epidermal differentiation and hair follicle growth observed in MOF cKO. Together, this study reveals the requirement of MOF-mediated epigenetic mechanism for mitochondria and cilia, and demonstrates the important function of the MOF/ETC axis for mammalian skin development.

developmental biology↗

bioRxiv2022

FOXP2 confers oncogenic effects in prostate cancer through activating MET signalling

Identification oncogenes is fundamental to revealing the molecular basis of cancer. Here, we found that FOXP2 is overexpressed in human prostate cancer cells and prostate tumors, but its expression is absent in normal prostate epithelial cells and low in benign prostatic hyperplasia. To date, little is known regarding the link of FOXP2 to prostate cancer. We observed that high FOXP2 expression and frequent amplification are significantly associated with high Gleason score. Ectopic expression of FOXP2 induces malignant transformation of mouse NIH3T3 fibroblasts and human prostate epithelial cell RWPE-1. Conversely, FOXP2 knockdown suppresses the proliferation of prostate cancer cells. Transgenic overexpression of FOXP2 in the mouse prostate causes prostatic intraepithelial neoplasia. Overexpression of FOXP2 aberrantly activates oncogenic MET signalling and inhibitors targeting MET signalling effectively reverts the FOXP2-induced oncogenic phenotype. Additionally, the novel recurrent FOXP2-CPED1 fusion identified in prostate tumors results in high expression of truncated FOXP2, which exhibit a similar capacity for malignant transformation. Together, our data demonstrate for the first time that FOXP2 is an oncogene involved in tumorigenicity of prostate.

cancer biology↗

bioRxiv2022

Biological responses to terahertz radiation with different power density in primary hippocampal neurons

BackgroundTerahertz (THz) radiation is a valuable imaging and sensing tool which is widely used in industry and medicine. However, it biological effects including genotoxicity and cytotoxicity are not clear, particularly on the nervous system. In this study, we investigated how THz radiation with different frequency and intensity would affect the morphology, cell growth and function of rat hippocampal neurons in vitro. MethodsPrimary hippocampal neuronal cells were isolated from newborn Wistar rat pups and were exposed to THz radiation with the frequencies of 0.12 THz and 0.141 THz, at an power intensities of 10 mW/cm2 and 30 mW/cm2 respectively. The cellular activities were evaluated by Cell Counting Kit-8 after exposure for 10 min and 30 min. The cellular apoptosis was examined by Annexin V staining using flow cytometry and the ultrastructure was detected by electron microscopy. Moreover, the amount of amino acid in the cultured neurons was measured by high performance liquid chromatography (HPLC). ResultsNo obvious morphological changes, such as axon length and axon number, could be detected in primary neurons after exposure to 10mW/cm2 or 50 mW/cm2 THz for 30 min. However, the overall cellular activities were differentially affected by 10mW/cm2 and 50 mW/cm2 THz. Obvious cellular apoptosis was induced by both 10mW/cm2 and 50 mW/cm2 exposure. We also found that the amounts of most amino acids did not change significantly after exposure to these two types of THz radiation. ConclusionsThe effects of THz radiation on hippocampal neurons depend on frequency, power, exposure time as well as the measurement time after the exposure. These results will provide important information for further investigation of the effect of THz radiation on the nervous system.

neuroscience↗

bioRxiv2021

Impact of Protein Identity on Tumor-Associated Antigen Uptake into Infiltrating Immune Cells: A Comparison of Different Fluorescent Proteins as Model Antigens

Effective immune responses depend on efficient antigen uptake in the periphery, transport of those antigens to, and presentation in draining lymph nodes (LNs). These processes have been studied intensively using stable fluorescent proteins (FPs) as model antigens. To date, ZsGreen is the only FP that can be tracked efficiently towards LNs, hence, it is difficult to compare studies using alternated tracking proteins. Here, we systematically compared six different FPs. We included ZsGreen, ZsYellow, DsRed, AsRed, mCherry, and mRFP based on sequence homology and/or origin species, and generated FP-expressing tumor cell lines. Stability of fluorescent signal was assessed in vitro over time, across different pH environments, and in vivo through FP antigen uptake and transfer to immune cells isolated from tumors and tumor-draining LNs. ZsGreen could be detected in high percentages of all analyzed tumor-infiltrating immune cells, with highest amounts in tumor-associated macrophages (TAMs) and type 2 conventional dendritic cells (cDC2s). ZsYellow, AsRed, and DsRed followed a similar pattern, but percentages of FP-containing immune cells in the tumor were lower than for ZsGreen. Strikingly, mRFP and mCherry demonstrated a non-canonical antigen uptake pattern where percentages of FP-positive tumor-infiltrating immune cells were highest for cDC1s not TAMs and cDC2s despite comparable stabilities and localization of all FPs. Analysis of antigen-containing cells in the LN was hindered by intracellular degradation of FPs. Only ZsGreen could be efficiently tracked to the LN, though some signal was measurable for ZsYellow and DsRed. In summary, we find that detection of antigen uptake and distribution is subject to variabilities related to fluorophore nature. Future experiments need to consider that these processes might be impacted by protein expression, stability, or other unknown factors. Thus, our data sheds light on potential under-appreciated mechanisms regulating antigen transfer and highlights potential uses and necessary caveats to interpretation based on FP use.

immunology↗

bioRxiv2021

NetTIME: improving multitask transcription factor binding site prediction with base-pair resolution

MotivationMachine learning models for predicting cell-type-specific transcription factor (TF) binding sites have become increasingly more accurate thanks to the increased availability of next-generation sequencing data and more standardized model evaluation criteria. However, knowledge transfer from data-rich to data-limited TFs and cell types remains crucial for improving TF binding prediction models because available binding labels are highly skewed towards a small collection of TFs and cell types. Transfer prediction of TF binding sites can potentially benefit from a multitask learning approach; however, existing methods typically use shallow single-task models to generate low-resolution predictions. Here we propose NetTIME, a multitask learning framework for predicting cell-type-specific transcription factor binding sites with base-pair resolution. ResultsWe show that the multitask learning strategy for TF binding prediction is more efficient than the single-task approach due to the increased data availability. NetTIME trains high-dimensional embedding vectors to distinguish TF and cell-type identities. We show that this approach is critical for the success of the multitask learning strategy and allows our model to make accurate transfer predictions within and beyond the training panels of TFs and cell types. We additionally train a linear-chain conditional random field (CRF) to classify binding predictions and show that this CRF eliminates the need for setting a probability threshold and reduces classification noise. We compare our methods predictive performance with two state-of-the-art methods, Catchitt and Leopard, and show that our method outperforms previous methods under both supervised and transfer learning settings. AvailabilityNetTIME is freely available at https://github.com/ryi06/NetTIME and the code is also archived at https://doi.org/10.5281/zenodo.6994897 Contactrb133@nyu.edu

systems biology↗

bioRxiv2021

Single-cell gene regulatory network inference atscale: The Inferelator 3.0

MotivationGene regulatory networks define regulatory relationships between transcription factors and target genes within a biological system, and reconstructing them is essential for understanding cellular growth and function. Methods for inferring and reconstructing networks from genomics data have evolved rapidly over the last decade in response to advances in sequencing technology and machine learning. The scale of data collection has increased dramatically; the largest genome-wide gene expression datasets have grown from thousands of measurements to millions of single cells, and new technologies are on the horizon to increase to tens of millions of cells and above. ResultsIn this work, we present the Inferelator 3.0, which has been significantly updated to integrate data from distinct cell types to learn context-specific regulatory networks and aggregate them into a shared regulatory network, while retaining the functionality of the previous versions. The Inferelator is able to integrate the largest single-cell datasets and learn cell-type specific gene regulatory networks. Compared to other network inference methods, the Inferelator learns new and informative Saccharomyces cerevisiae networks from single-cell gene expression data, measured by recovery of a known gold standard. We demonstrate its scaling capabilities by learning networks for multiple distinct neuronal and glial cell types in the developing Mus musculus brain at E18 from a large (1.3 million) single-cell gene expression dataset with paired single-cell chromatin accessibility data. AvailabilityThe inferelator software is available on GitHub (https://github.com/flatironinstitute/inferelator) under the MIT license and has been released as python packages with associated documentation (https://inferelator.readthedocs.io/).

systems biology↗

bioRxiv2021

Epigenetic Programming during thymic development sets the stage for optimal function in effector T cells via DNA demethylation

The repressive effect of DNA methylation at promoters is well-known. However, its role within conserved sequences in intragenic and intergenic regions is less clear. Using Cd4 as a model gene, here we show that DNA methylation regulates the function of stimulus-responsive regulatory elements in effector T cells. Two cis-elements orchestrate intra-and intergenic DNA demethylation of the Cd4 gene during thymic development, which in turn licenses a stimulus-responsive element, E4a, for its later function in effector cells. Deficiency in DNA demethylation leads to impaired E4a function, reduced H3K4me3 promoter levels and an inability to repel de novo DNA methylation during replication, ultimately leading to gene silencing. This physiological reduction in CD4 expression leads to a defect in Th1 polarization during cutaneous Leishmaniasis. Similar patterns of regulation were observed in a broad number of genes, highlighting an essential role for DNA demethylation during thymic development in modulating the function of stimulus-responsive elements.

immunology↗

bioRxiv2019

Health behaviors among male and female university students in Cambodia: a cross-sectional survey

BackgroundStudents go through a transition when they enter university, which involves major individual and contextual changes in every domain of life that may lead to several behavioral and health problems. This paper describes a wide range of health behaviors among male and female university students in Cambodia.\n\nMethodsA cross-sectional survey was conducted in 2015 among students randomly selected from the Royal University of Phnom Penh and University of Battambang. Health-related behaviors in different domains were collected using a structured questionnaire. Chi-square test, Fishers exact test, or independent Students t-test was used as appropriate to describe and compare the variables among male and female students.\n\nResultsThis study included 1359 students, of whom 50.8% were male, and the mean age was 21.3 years (SD= 2.3). Of the total, 79.5% reported not having any vigorous-intensity activities, 25.9% not having moderate-intensity activities, and 33.5% not having walked continuously for 10 minutes during the past seven days. The prevalence of substance use was low with 38.3% currently drinking alcohol, 1.1% smoking tobacco, and 0.4% using an illicit drug during the past 12 months. About one in ten (10.6%) reported having sexual intercourse, with a mean number of partners of 2.1 (SD= 2.4) during the past 12 months, and 42.4% not using a condom in the last intercourse. Only 7.1% reported having been diagnosed with a sexually transmitted infection in the past 12 months; of whom, 60% sought for treatment. About one-third (33.6%) reported eating fast food once or twice, and 5.3% having it three times or more over the last week. More than half (55.6%) had one to two servings of fruits and vegetables daily, and 9.9% did not eat any fruits and vegetables over the last week.\n\nConclusionsWe found that the prevalence of sexual risk behaviors and substance use was plausibly low among university students in this study. However, the rates of inactive lifestyle and unhealthy food consumption were concerning. Public policy and universities should promote healthy behaviors among the students. The interventions may take advantage of and expend upon the good health behaviors and consider gender differences.

scientific communication and education↗

bioRxiv2019

Integrated analysis of directly captured microRNA targets reveals the impact of microRNAs on mammalian transcriptome

MicroRNA (miRNA)-mediated regulation is widespread, relatively mild but functionally important. Despite extensive efforts to identify miRNA targets, it remains unclear how miRNAs bind to mRNA targets globally and how changes in miRNA levels affects the transcriptome. Here we apply an optimized method for simultaneously capturing miRNA and targeted RNA sites to wildtype, miRNA knockout and induced epithelial cells. We find that abundantly expressed miRNAs can bind to thousands of different transcripts and many different miRNAs can regulate the same gene. Although mRNA sites that are bound by miRNAs and also contain matches to seed sequences confer the strongest regulation, [~]50% of miRNAs bind to RNA regions without seed matches. In general, these bindings have little impact on mRNA levels and reflect a scanning activity of miRNAs. In addition, different miRNAs have different preferences to seed matches and 3end base-pairing. For a single miRNA, the effectiveness of mRNA regulation is highly correlated with the number of captured miRNA:RNA fragments. Notably, elevated miRNA expression effectively represses existing targets with little impact on newly recognized targets. Global analysis of directly captured mRNA targets reveals pathways that are involved in cancer, cell adhesion and signaling pathways are highly regulated by many different miRNAs in epithelial cells. Comparison between experimentally captured and TargetScan predicted targets indicates that our approach is more effective to identify bona fide targets by reducing false positive and negative predictions. This study reveals the global binding landscape and impact of miRNAs on mammalian transcriptome.

molecular biology↗

bioRxiv2018

Transcriptional regulatory networks that promote and restrict identities and functions of intestinal innate lymphoid cells

Innate lymphoid cells (ILCs) can be subdivided into several distinct cytokine-secreting lineages that promote tissue homeostasis and immune defense but also contribute to inflammatory diseases. Accumulating evidence suggests that ILCs, similarly to other immune populations, are capable of phenotypic and functional plasticity in response to infectious or environmental stimuli. Yet the transcriptional circuits that control ILC identity and function are largely unknown. Here we integrate gene expression and chromatin accessibility data to infer transcriptional regulatory networks within intestinal type 1, 2, and 3 ILCs. We predict the \"core\" sets of transcription-factor (TF) regulators driving each ILC subset identity, among which only a few TFs were previously known. To assist in the interpretation of these networks, TFs were organized into cooperative clusters, or modules that control gene programs with distinct functions. The ILC network reveals extensive alternative-lineage-gene repression, whose regulation may explain reported plasticity between ILC subsets. We validate new roles for c-MAF and BCL6 as regulators affecting the type 1 and type 3 ILC lineages. Manipulation of TF pathways identified here might provide a novel means to selectively regulate ILC effector functions to alleviate inflammatory disease or enhance host tolerance to pathogenic microbes or noxious stimuli. Our results will enable further exploration of ILC biology, while our network approach will be broadly applicable to identifying key cell state regulators in other in vivo cell populations.

immunology↗