Search

Fibrotic diseases affect multiple organs and are associated with morbidity and mortality. To examine organ-specific and shared biologic mechanisms that underlie fibrosis in different organs, we developed machine learning models to quantify T1 time, a marker of interstitial fibrosis, in the liver, pancreas, heart and kidney among 43,881 UK Biobank participants who underwent magnetic resonance imaging. In phenome-wide association analyses, we demonstrate the association of increased organ-specific T1 time, reflecting increased interstitial fibrosis, with prevalent diseases across multiple organ systems. In genome-wide association analyses, we identified 27, 18, 11 and 10 independent genetic loci associated with liver, pancreas, myocardial and renal cortex T1 time, respectively. There was a modest genetic correlation between the examined organs. Several loci overlapped across the examined organs implicating genes involved in a myriad of biologic pathways including metal ion transport (SLC39A8, HFEand TMPRSS6), glucose metabolism (PCK2), blood group antigens (ABOand FUT2), immune function (BANK1and PPP3CA), inflammation (NFKB1) and mitosis (CENPE). Finally, we found that an increasing number of organs with T1 time falling in the top quintile was associated with increased mortality in the population. Individuals with a high burden of fibrosis in ≥3 organs had a 3-fold increase in mortality compared to those with a low burden of fibrosis across all examined organs in multivariable-adjusted analysis (hazard ratio = 3.31, 95% confidence interval 1.77–6.19; P= 1.78 × 10−4). By leveraging machine learning to quantify T1 time across multiple organs at scale, we uncovered new organ-specific and shared biologic pathways underlying fibrosis that may provide therapeutic targets.

Mapping gene networks requires large amounts of transcriptomic data to learn the connections between genes, which impedes discoveries in settings with limited data, including rare diseases and diseases affecting clinically inaccessible tissues. Recently, transfer learning has revolutionized fields such as natural language understanding1,2and computer vision3by leveraging deep learning models pretrained on large-scale general datasets that can then be fine-tuned towards a vast array of downstream tasks with limited task-specific data. Here, we developed a context-aware, attention-based deep learning model, Geneformer, pretrained on a large-scale corpus of about 30 million single-cell transcriptomes to enable context-specific predictions in settings with limited data in network biology. During pretraining, Geneformer gained a fundamental understanding of network dynamics, encoding network hierarchy in the attention weights of the model in a completely self-supervised manner. Fine-tuning towards a diverse panel of downstream tasks relevant to chromatin and network dynamics using limited task-specific data demonstrated that Geneformer consistently boosted predictive accuracy. Applied to disease modelling with limited patient data, Geneformer identified candidate therapeutic targets for cardiomyopathy. Overall, Geneformer represents a pretrained deep learning model from which fine-tuning towards a broad range of downstream applications can be pursued to accelerate discovery of key network regulators and candidate therapeutic targets.

Myocardial interstitial fibrosis is associated with cardiovascular disease and adverse prognosis. Here, to investigate the biological pathways that underlie fibrosis in the human heart, we developed a machine learning model to measure native myocardial T1 time, a marker of myocardial fibrosis, in 41,505 UK Biobank participants who underwent cardiac magnetic resonance imaging. Greater T1 time was associated with diabetes mellitus, renal disease, aortic stenosis, cardiomyopathy, heart failure, atrial fibrillation, conduction disease and rheumatoid arthritis. Genome-wide association analysis identified 11 independent loci associated with T1 time. The identified loci implicated genes involved in glucose transport (SLC2A12), iron homeostasis (HFE, TMPRSS6), tissue repair (ADAMTSL1, VEGFC), oxidative stress (SOD2), cardiac hypertrophy (MYH7B) and calcium signaling (CAMK2D). Using a transforming growth factor β1-mediated cardiac fibroblast activation assay, we found that 9 of the 11 loci consisted of genes that exhibited temporal changes in expression or open chromatin conformation supporting their biological relevance to myofibroblast cell state acquisition. By harnessing machine learning to perform large-scale quantification of myocardial interstitial fibrosis using cardiac imaging, we validate associations between cardiac fibrosis and disease, and identify new biologically relevant pathways underlying fibrosis.

The heart, the first organ to develop in the embryo, undergoes complex morphogenesis that when defective results in congenital heart disease (CHD). With current therapies, more than 90% of patients with CHD survive into adulthood, but many suffer premature death from heart failure and non-cardiac causes1. Here, to gain insight into this disease progression, we performed single-nucleus RNA sequencing on 157,273 nuclei from control hearts and hearts from patients with CHD, including those with hypoplastic left heart syndrome (HLHS) and tetralogy of Fallot, two common forms of cyanotic CHD lesions, as well as dilated and hypertrophic cardiomyopathies. We observed CHD-specific cell states in cardiomyocytes, which showed evidence of insulin resistance and increased expression of genes associated with FOXO signalling and CRIM1. Cardiac fibroblasts in HLHS were enriched in a low-Hippo and high-YAP cell state characteristic of activated cardiac fibroblasts. Imaging mass cytometry uncovered a spatially resolved perivascular microenvironment consistent with an immunodeficient state in CHD. Peripheral immune cell profiling suggested deficient monocytic immunity in CHD, in agreement with the predilection in CHD to infection and cancer2. Our comprehensive phenotyping of CHD provides a roadmap towards future personalized treatments for CHD.

Heart failure encompasses a heterogeneous set of clinical features that converge on impaired cardiac contractile function1,2and presents a growing public health concern. Previous work has highlighted changes in both transcription and protein expression in failing hearts3,4, but may overlook molecular changes in less prevalent cell types. Here we identify extensive molecular alterations in failing hearts at single-cell resolution by performing single-nucleus RNA sequencing of nearly 600,000 nuclei in left ventricle samples from 11 hearts with dilated cardiomyopathy and 15 hearts with hypertrophic cardiomyopathy as well as 16 non-failing hearts. The transcriptional profiles of dilated or hypertrophic cardiomyopathy hearts broadly converged at the tissue and cell-type level. Further, a subset of hearts from patients with cardiomyopathy harbour a unique population of activated fibroblasts that is almost entirely absent from non-failing samples. We performed a CRISPR-knockout screen in primary human cardiac fibroblasts to evaluate this fibrotic cell state transition; knockout of genes associated with fibroblast transition resulted in a reduction of myofibroblast cell-state transition upon TGFβ1 stimulation for a subset of genes. Our results provide insights into the transcriptional diversity of the human heart in health and disease as well as new potential therapeutic targets and biomarkers for heart failure.

In this study, Pekarek et al. demonstrate a novel oxytocin-dependent mechanism of adult-born neuron synaptic maturation and circuit integration. They demonstrate that oxytocin receptor signaling promotes synaptic maturation of newly integrating adult-born neurons by regulating their morphological development and expression of mature synaptic AMPARs and other structural proteins.

The exploration of S- or Se-for-Oexchange with (Z)-1-R-2-(4′,4′-dimethyl-2′-oxazolin-2′-yl)-eth-1-en-1-ols (1: R = Ph, Me, CF3, Ph-p-OMe) is described. The use of P4S10, intended to facilitate O↔Sexchange, leads to the isolation of rare examples of air-stable thioester analogues of dithiooxophosphoranes. Clean heteroatom exchange can be realized using Lawesson's Reagentin good to moderate yields. Woollins'selenation reagent leads to the isolation of an unusual example of a selenazolidine product (10) in trace amounts.The exploration of S- or Se-for-Oexchange with (Z)-1-R-2-(4′,4′-dimethyl-2′-oxazolin-2′-yl)-eth-1-en-1-ols is described. The use of P4S10leads to the isolation of rare examples of air-stable thioester analogues of dithiooxophosphoranes. Clean heteroatom exchange can be realized using Lawesson's Reagentin good to moderate yields. Woollins'reagent leads to the isolation of an example of a selenazolidine product (10) in trace amounts.

Transcriptional enhancers commonly work over long genomic distances to precisely regulate spatiotemporal gene expression patterns. Dissecting the promoters physically contacted by these distal regulatory elements is essential for understanding developmental processes as well as the role of disease-associated risk variants. Modern proximity-ligation assays, like HiChIP and ChIA-PET, facilitate the accurate identification of long-range contacts between enhancers and promoters. However, these assays are technically challenging, expensive, and time-consuming, making it difficult to investigate enhancer topologies, especially in uncharacterized cell types. To overcome these shortcomings, we therefore designed LoopPredictor, an ensemble machine learning model, to predict genome topology for cell types which lack long-range contact maps. To enrich for functional enhancer-promoter loops over common structural genomic contacts, we trained LoopPredictor with both H3K27ac and YY1 HiChIP data. Moreover, the integration of several related multi-omics features facilitated identifying and annotating the predicted loops. LoopPredictor is able to efficiently identify cell type–specific enhancer-mediated loops, and promoter–promoter interactions, with a modest feature input requirement. Comparable to experimentally generated H3K27ac HiChIP data, we found that LoopPredictor was able to identify functional enhancer loops. Furthermore, to explore the cross-species prediction capability of LoopPredictor, we fed mouse multi-omics features into a model trained on human data and found that the predicted enhancer loops outputs were highly conserved. LoopPredictor enables the dissection of cell type–specific long-range gene regulation and can accelerate the identification of distal disease-associated risk variants.

Supplemental Digital Content is available in the text.

Oncogenic mutations confer on cells the ability to propagate indefinitely, but whether oncogenes alter the cell fate of these cells is unknown. Here, we show that the transcriptional regulator PRDM16s causes oncogenic fate conversion by transforming cells fated to form platelets and erythrocytes into myeloid leukemia stem cells (LSCs). Prdm16s expression in megakaryocyte-erythroid progenitors (MEPs), which normally lack the potential to generate granulomonocytic cells, caused AML by converting MEPs into LSCs. Prdm16s blocked megakaryocytic/erythroid potential by interacting with super enhancers and activating myeloid master regulators, including PU.1. A CRISPR dropout screen confirmed that PU.1 is required for Prdm16s-induced leukemia. Ablating PU.1 attenuated leukemogenesis and reinstated the megakaryocytic/erythroid potential of leukemic MEPs in mouse models and human AML with PRDM16 rearrangement. Thus, oncogenic PRDM16s expression gives MEPs an LSC fate by activating myeloid gene regulatory networks.

In this study, Xiao et al. set out to identify molecular mechanisms and genetic pathways that regulate the cellular makeup of a tissue. Using mammalian cardiac fibroblast as a model and unbiased genomic approaches, including high throughput single-cell transcriptomics and enhancer connectome profiling, the authors show that the Hippo signaling pathway is essential in maintaining cardiac fibroblasts in a homeostatic condition.

Mammalian organs vary widely in regenerative capacity. Poorly regenerative organs, such as the heart are particularly vulnerable to organ failure. Once established, heart failure commonly results in mortality. The Hippo pathway, a kinase cascade that prevents adult cardiomyocyte proliferation and regeneration, is upregulated in human heart failure. Here we show that deletion of the Hippo pathway component Salvador (Salv) in mouse hearts with established ischaemic heart failure after myocardial infarction induces a reparative genetic program with increased scar border vascularity, reduced fibrosis, and recovery of pumping function compared with controls. Using translating ribosomal affinity purification, we isolate cardiomyocyte-specific translating messenger RNA. Hippo-deficient cardiomyocytes have increased expression of proliferative genes and stress response genes, such as the mitochondrial quality control gene, Park2. Genetic studies indicate that Park2 is essential for heart repair, suggesting a requirement for mitochondrial quality control in regenerating myocardium. Gene therapy with a virus encoding Salv short hairpin RNA improves heart function when delivered at the time of infarct or after ischaemic heart failure following myocardial infarction was established. Our findings indicate that the failing heart has a previously unrecognized reparative capacity involving more than cardiomyocyte renewal.

Ischemic cardiomyopathy (ICM) is the leading cause of heart failure worldwide, yet the cellular and molecular signature of this disease is largely unclear. Using single-nucleus RNA sequencing (snRNA-seq) and integrated computational analyses, we profile the transcriptomes of over 99,000 human cardiac nuclei from the non-infarct region of the left ventricle of 7 ICM transplant recipients and 8 non-failing (NF) controls. We find the cellular composition of the ischemic heart is significantly altered, with decreased cardiomyocytes and increased proportions of lymphatic, angiogenic, and arterial endothelial cells in patients with ICM. We show that there is increased LAMININ signaling from endothelial cells to other cell types in ICM compared with NF. Finally, we find that the transcriptional changes that occur in ICM are similar to those in hypertrophic and dilated cardiomyopathies and that the mining of these combined datasets can identify druggable genes that could be used to target end-stage heart failure.

Chromatin conformation capture (3C)-based technologies have enabled the accurate detection of topological genomic interactions, and the adoption of ChIP techniques to 3C-based protocols makes it possible to identify long-range interactions. To analyze these large and complex datasets, computational methods are undergoing rapid and expansive evolution. Thus, a thorough evaluation of these analytical pipelines is necessary to identify which commonly used algorithms and processing pipelines need to be improved. Here we present a comprehensive benchmark framework, Bacon, to evaluate the performance of several computational methods. Finally, we provide practical recommendations for users working with HiChIP and/or ChIA-PET analyses.

Introduction:Atrial fibrillation (AF) is the most common sustained arrhythmia and is associated with substantial morbidity and mortality. AF is known to have a heritable component, with >100 associated common variant loci. Rare variant studies have yielded limited robust associations for AF. We aimed to utilize large genome and exome sequencing data to discover rare genetic variants conferring large effects on AF risk.Methods:We meta-analyzed genome and exome sequencing data from 36 studies, including TOPMed, CCDG, UK Biobank and FOURIER. We performed exome-wide gene burden testing of rare (MAF<0.1%) loss-of-function and deleterious missense variants, and single variant testing of low-frequency and rare (MAF<1%) coding variants. Within genome sequenced samples, we performed gene burden testing of rare structural variants. Novel signals were replicated in MyCode. Finally, we functionally validated a novel gene by siRNA knockdown in pluripotent-induced atrial cardiomyocytes.Results:We included 52,416 AF cases and 277,762 controls, of which 49.6% were female, 83.4% were of European ancestry, and the mean baseline age was 56 years. In analysis of rare coding variation, we identified 4 novel genes associated with AF, including MYBPC3(OR 3.5, P=2.1x10-15), LMNA(OR 5.7, P=8.8x10-11), PKP2(OR 1.9, P=5.2x10-8) and KDM5B(OR 2.3, P=3.0x10-6). These signals were robust to removal of heart failure and cardiomyopathy cases and were replicated in independent datasets. Single variant analysis identified 2 novel signals in FAM189A2(OR 3.9, P=7.86x10-8) and ZFC3H1(OR 5.9, P=9.7x10-8). Rare deletions in CTNNA3(OR 4.5, P=7.0x10-9) were associated with increased AF risk and were supported by independent coding variant analyses, while duplications of GATA4(OR 0.24, P=2.1x10-5) were associated with reduced AF risk. In functional studies, knockdown of KDM5Bresulted in shortening of the atrial action potential duration.Conclusions:Our analyses show the contribution of rare coding and structural variants to AF risk, highlight the shared genetic pathways underlying cardiomyopathy and AF, and implicate the histone demethylase gene KDM5Bin AF susceptibility. In sum, we expanded our understanding of the rare variant architecture of this common arrhythmia.

In response to retinal damage, the Müller glial cells (MGs) of the zebrafish retina have the ability to undergo a cellular reprogramming event in which they enter the cell cycle and divide asymmetrically, thereby producing multipotent retinal progenitors capable of regenerating lost retinal neurons. However, mammalian MGs do not exhibit such a proliferative and regenerative ability. Here, we identify Hippo pathway-mediated repression of the transcription cofactor YAP as a core regulatory mechanism that normally blocks mammalian MG proliferation and cellular reprogramming. MG-specific deletion of Hippo pathway components Lats1and Lats2, as well as transgenic expression of a Hippo non-responsive form of YAP (YAP5SA), resulted in dramatic Cyclin D1upregulation, loss of adult MG identity, and attainment of a highly proliferative, progenitor-like cellular state. Our results reveal that mammalian MGs may have latent regenerative capacity that can be stimulated by repressing Hippo signaling.

Cellular heterogeneity within the mammalian brain poses a challenge toward understanding its complex functions. Within the olfactory bulb, odor information is processed by subtypes of inhibitory interneurons whose heterogeneity and functionality are influenced by ongoing adult neurogenesis. To investigate this cellular heterogeneity and better understand adult-born neuron development, we utilized single-cell RNA sequencing and computational modeling to reveal diverse and transcriptionally distinct neuronal and nonneuronal cell types. We also analyzed molecular changes during adult-born interneuron maturation and uncovered developmental programs within their gene expression profiles. Finally, we identified that distinct neuronal subtypes are differentially affected by sensory experience. Together, these data provide a transcriptome-based foundation for investigating subtype-specific neuronal function in the olfactory bulb (OB), charting the molecular profiles that arise during the maturation and integration of adult-born neurons and how they dynamically change in an activity-dependent manner.