bims-gerecp 2024-08-04 papers

bims-gerecp

Biomed News

on Gene regulatory networks of epithelial cell plasticity

Issue of 2024‒08‒04
seventeen papers selected by
Xiao Qin, University of Oxford

Multiome Perturb-seq unlocks scalable discovery of integrated perturbation effects on the transcriptome and epigenome.
Approximate Bayesian computation for inferring Waddington landscapes from single-cell data.
Concepts and new developments in droplet-based single cell multi-omics.
Localized in vivo gene editing of murine cancer-associated fibroblasts.
Enhancer-driven gene regulatory networks inference from single-cell RNA-seq and ATAC-seq data.
scCross: a deep generative model for unifying single-cell multi-omics with seamless integration, cross-modal generation, and in silico exploration.
Drug resistant pancreatic cancer cells exhibit altered biophysical interactions with stromal fibroblasts in imaging studies of 3D co-culture models.
Epigenome editing technologies for discovery and medicine.
Reconstruction of single-cell lineage trajectories and identification of diversity in fates during the epithelial-to-mesenchymal transition.
ERK signalling eliminates Nanog and maintains Oct4 to drive the formative pluripotency transition.
Exploring the Function of Gene Promoter Regulatory Elements Using CRISPR Tools.
Optimized reporters for multiplexed detection of transcription factor activity.
Protocol for the application of single-cell damage in murine intestinal organoid models.
Quantifying genome-wide transcription factor binding affinities for chromatin using BANC-seq.
Cancer, metastasis, and the epigenome.
CELLNET technology: Spatially organized, functional 3D networks at single cell resolution.
Chromosomal instability as a driver of cancer progression.

bioRxiv. 2024 Jul 27. pii: 2024.07.26.605307. [Epub ahead of print]

Multiome Perturb-seq unlocks scalable discovery of integrated perturbation effects on the transcriptome and epigenome.

Eli Metzner, Kaden M Southard, Thomas M Norman.

Single-cell CRISPR screens link genetic perturbations to transcriptional states, but high-throughput methods connecting these induced changes to their regulatory foundations are limited. Here we introduce Multiome Perturb-seq, extending single-cell CRISPR screens to simultaneously measure perturbation-induced changes in gene expression and chromatin accessibility. We apply Multiome Perturb-seq in a CRISPRi screen of 13 chromatin remodelers in human RPE-1 cells, achieving efficient assignment of sgRNA identities to single nuclei via an improved method for capturing barcode transcripts from nuclear RNA. We organize expression and accessibility measurements into coherent programs describing the integrated effects of perturbations on cell state, finding that ARID1A and SUZ12 knockdowns induce programs enriched for developmental features. Pseudotime analysis of perturbations connects accessibility changes to changes in gene expression, highlighting the value of multimodal profiling. Overall, our method provides a scalable and simply implemented system to dissect the regulatory logic underpinning cell state.

DOI: https://doi.org/10.1101/2024.07.26.605307
R Soc Open Sci. 2024 Jul;11(7): 231697

Approximate Bayesian computation for inferring Waddington landscapes from single-cell data.

Yujing Liu, Stephen Y Zhang, Istvan T Kleijn, Michael P H Stumpf.

  Single-cell technologies allow us to gain insights into cellular processes at unprecedented resolution. In stem cell and developmental biology snapshot data allow us to characterize how the transcriptional states of cells change between successive cell types. Here, we show how approximate Bayesian computation (ABC) can be employed to calibrate mathematical models against single-cell data. In our simulation study, we demonstrate the pivotal role of the adequate choice of distance measures appropriate for single-cell data. We show that for good distance measures, notably optimal transport with the Sinkhorn divergence, we can infer parameters for mathematical models from simulated single-cell data. We show that the ABC posteriors can be used (i) to characterize parameter sensitivity and identify dependencies between different parameters and (ii) to construct representations of the Waddington or epigenetic landscape, which forms a popular and interpretable representation of the developmental dynamics. In summary, these results pave the way for fitting mechanistic models of stem cell differentiation to single-cell data.

Keywords:  epigenetic landscape; likelihood free inference; quasi-potential

DOI:  https://doi.org/10.1098/rsos.231697
Trends Biotechnol. 2024 Aug 01. pii: S0167-7799(24)00184-7. [Epub ahead of print]

Concepts and new developments in droplet-based single cell multi-omics.

Arthur Chow, Caleb A Lareau.

  Single cell sequencing technologies have become a fixture in the molecular profiling of cells due to their ease, flexibility, and commercial availability. In particular, partitioning individual cells inside oil droplets via microfluidic reactions enables transcriptomic or multi-omic measurements for thousands of cells in parallel. Complementing the multitude of biological discoveries from genomics analyses, the past decade has brought new capabilities from assay baselines to enable a deeper understanding of the complex data from single cell multi-omics. Here, we highlight four innovations that have improved the reliability and understanding of droplet microfluidic assays. We emphasize new developments that further orient principles of technology development and guidelines for the design, benchmarking, and implementation of new droplet-based methodologies.

Keywords:  bioinformatics; microfluidics; multi-omics; sequencing; single cell

DOI:  https://doi.org/10.1016/j.tibtech.2024.07.006
bioRxiv. 2024 Jul 16. pii: 2024.07.11.603114. [Epub ahead of print]

Localized in vivo gene editing of murine cancer-associated fibroblasts.

Nicholas F Kuhn, Itzia Zaleta-Linares, William A Nyberg, Justin Eyquem, Matthew F Krummel.

Discovering the role of fibroblasts residing in the tumor microenvironment (TME) requires controlled, localized perturbations because fibroblasts play critical roles in regulating immunity and tumor biology at multiple sites. Systemic perturbations can lead to unintended, confounding secondary effects, and methods to locally genetically engineer fibroblasts are lacking. To specifically investigate murine stromal cell perturbations restricted to the TME, we developed an adeno-associated virus (AAV)-based method to target any gene-of-interest in fibroblasts at high efficiency (>80%). As proof of concept, we generated single (sKO) and double gene KOs (dKO) of Osmr , Tgfbr2 , and Il1r1 in cancer-associated fibroblasts (CAFs) and investigated how their cell states and those of other cells of the TME subsequently change in mouse models of melanoma and pancreatic ductal adenocarcinoma (PDAC). Furthermore, we developed an in vivo knockin-knockout (KIKO) strategy to achieve long-term tracking of CAFs with target gene KO via knocked-in reporter gene expression. This validated in vivo gene editing toolbox is fast, affordable, and modular, and thus holds great potential for further exploration of gene function in stromal cells residing in tumors and beyond.

DOI: https://doi.org/10.1101/2024.07.11.603114
Brief Bioinform. 2024 Jul 25. pii: bbae369. [Epub ahead of print]25(5):

Enhancer-driven gene regulatory networks inference from single-cell RNA-seq and ATAC-seq data.

Yang Li, Anjun Ma, Yizhong Wang, Qi Guo, Cankun Wang, Hongjun Fu, Bingqiang Liu, Qin Ma.

  Deciphering the intricate relationships between transcription factors (TFs), enhancers, and genes through the inference of enhancer-driven gene regulatory networks (eGRNs) is crucial in understanding gene regulatory programs in a complex biological system. This study introduces STREAM, a novel method that leverages a Steiner forest problem model, a hybrid biclustering pipeline, and submodular optimization to infer eGRNs from jointly profiled single-cell transcriptome and chromatin accessibility data. Compared to existing methods, STREAM demonstrates enhanced performance in terms of TF recovery, TF-enhancer linkage prediction, and enhancer-gene relation discovery. Application of STREAM to an Alzheimer's disease dataset and a diffuse small lymphocytic lymphoma dataset reveals its ability to identify TF-enhancer-gene relations associated with pseudotime, as well as key TF-enhancer-gene relations and TF cooperation underlying tumor cells.

Keywords:  Steiner forest problem model; biological network; data integration; scATAC-seq; scRNA-seq; submodular optimization

DOI:  https://doi.org/10.1093/bib/bbae369
Genome Biol. 2024 Jul 29. 25(1): 198

scCross: a deep generative model for unifying single-cell multi-omics with seamless integration, cross-modal generation, and in silico exploration.

Xiuhui Yang, Koren K Mann, Hao Wu, Jun Ding.

  Single-cell multi-omics data reveal complex cellular states, providing significant insights into cellular dynamics and disease. Yet, integration of multi-omics data presents challenges. Some modalities have not reached the robustness or clarity of established transcriptomics. Coupled with data scarcity for less established modalities and integration intricacies, these challenges limit our ability to maximize single-cell omics benefits. We introduce scCross, a tool leveraging variational autoencoders, generative adversarial networks, and the mutual nearest neighbors (MNN) technique for modality alignment. By enabling single-cell cross-modal data generation, multi-omics data simulation, and in silico cellular perturbations, scCross enhances the utility of single-cell multi-omics studies.

Keywords:  Autoencoder; Cross-modal generation; Generative adversarial network; In silico perturbations; Multimodal integration; Muti-omics; Single cell

DOI:  https://doi.org/10.1186/s13059-024-03338-z
bioRxiv. 2024 Jul 17. pii: 2024.07.14.602133. [Epub ahead of print]

Drug resistant pancreatic cancer cells exhibit altered biophysical interactions with stromal fibroblasts in imaging studies of 3D co-culture models.

Eric Struth, Maryam Labaf, Vida Karimnia, Yiran Liu, Gwendolyn Cramer, Joanna B Dahl, Frank J Slack, Kourosh Zarringhalam, Jonathan P Celli.

Interactions between tumor and stromal cells are well known to play a prominent roles in progression of pancreatic ductal adenocarcinoma (PDAC). As knowledge of stromal crosstalk in PDAC has evolved, it has become clear that cancer associated fibroblasts can play both tumor promoting and tumor suppressive roles through a combination of paracrine crosstalk and juxtacrine interactions involving direct physical contact. Another major contributor to dismal survival statistics for PDAC is development of resistance to chemotherapy drugs. Though less is known about how the acquisition of chemoresistance impacts upon tumor-stromal crosstalk. Here, we use 3D co-culture geometries to recapitulate juxtacrine interactions between epithelial and stromal cells. In particular, extracellular matrix (ECM) overlay cultures in which stromal cells (pancreatic stellate cells, or normal human fibroblasts) are placed adjacent to PDAC cells (PANC1), result in direct heterotypic cell adhesions accompanied by dramatic fibroblast contractility which leads to highly condensed macroscopic multicellular aggregates as detected using particle image velocimetry (PIV) analysis to quantify cell velocities over the course of time lapse movie sequences. To investigate how drug resistance impacts these juxtacrine interactions we contrast cultures in which PANC1 are substituted with a drug resistant subline (PANC1-OR) previously established in our lab. We find that heterotypic cell-cell interactions are highly suppressed in drug-resistant cells relative to the parental PANC1 cells. To investigate further we conduct RNA-seq and bioinformatics analysis to identify differential gene expression in PANC1 and PANC1-OR, which shows that negative regulation of cell adhesion molecules, consistent with increased epithelial mesenchymal transition (EMT), is also consistent with loss of hetrotypic cell-cell contact necessary for the contractile behavior observed in drug naïve cultures. Overall these findings elucidate the role of drug-resistance in inhibiting an avenue of stromal crosstalk which is associated with tumor suppression and also help to establish cell culture conditions useful for further mechanistic investigation.

DOI: https://doi.org/10.1101/2024.07.14.602133
Nat Biotechnol. 2024 Jul 29.

Epigenome editing technologies for discovery and medicine.

Sean R McCutcheon, Dahlia Rohm, Nahid Iglesias, Charles A Gersbach.

Epigenome editing has rapidly evolved in recent years, with diverse applications that include elucidating gene regulation mechanisms, annotating coding and noncoding genome functions and programming cell state and lineage specification. Importantly, given the ubiquitous role of epigenetics in complex phenotypes, epigenome editing has unique potential to impact a broad spectrum of diseases. By leveraging powerful DNA-targeting technologies, such as CRISPR, epigenome editing exploits the heritable and reversible mechanisms of epigenetics to alter gene expression without introducing DNA breaks, inducing DNA damage or relying on DNA repair pathways.

DOI: https://doi.org/10.1038/s41587-024-02320-1
Proc Natl Acad Sci U S A. 2024 Aug 06. 121(32): e2406842121

Reconstruction of single-cell lineage trajectories and identification of diversity in fates during the epithelial-to-mesenchymal transition.

Yu-Chen Cheng, Yun Zhang, Shubham Tripathi, B V Harshavardhan, Mohit Kumar Jolly, Geoffrey Schiebinger, Herbert Levine, Thomas O McDonald, Franziska Michor.

  Exploring the complexity of the epithelial-to-mesenchymal transition (EMT) unveils a diversity of potential cell fates; however, the exact timing and mechanisms by which early cell states diverge into distinct EMT trajectories remain unclear. Studying these EMT trajectories through single-cell RNA sequencing is challenging due to the necessity of sacrificing cells for each measurement. In this study, we employed optimal-transport analysis to reconstruct the past trajectories of different cell fates during TGF-beta-induced EMT in the MCF10A cell line. Our analysis revealed three distinct trajectories leading to low EMT, partial EMT, and high EMT states. Cells along the partial EMT trajectory showed substantial variations in the EMT signature and exhibited pronounced stemness. Throughout this EMT trajectory, we observed a consistent downregulation of the EED and EZH2 genes. This finding was validated by recent inhibitor screens of EMT regulators and CRISPR screen studies. Moreover, we applied our analysis of early-phase differential gene expression to gene sets associated with stemness and proliferation, pinpointing ITGB4, LAMA3, and LAMB3 as genes differentially expressed in the initial stages of the partial versus high EMT trajectories. We also found that CENPF, CKS1B, and MKI67 showed significant upregulation in the high EMT trajectory. While the first group of genes aligns with findings from previous studies, our work uniquely pinpoints the precise timing of these upregulations. Finally, the identification of the latter group of genes sheds light on potential cell cycle targets for modulating EMT trajectories.

Keywords:  EMT; cell fate; scRNA-seq

DOI:  https://doi.org/10.1073/pnas.2406842121
Development. 2024 Jul 15. pii: dev203106. [Epub ahead of print]151(14):

ERK signalling eliminates Nanog and maintains Oct4 to drive the formative pluripotency transition.

Carla Mulas, Melanie Stammers, Siiri I Salomaa, Constanze Heinzen, David M Suter, Austin Smith, Kevin J Chalut.

  Naïve epiblast cells in the embryo and pluripotent stem cells in vitro undergo developmental progression to a formative state competent for lineage specification. During this transition, transcription factors and chromatin are rewired to encode new functional features. Here, we examine the role of mitogen-activated protein kinase (ERK1/2) signalling in pluripotent state transition. We show that a primary consequence of ERK activation in mouse embryonic stem cells is elimination of Nanog, which precipitates breakdown of the naïve state gene regulatory network. Variability in pERK dynamics results in heterogeneous loss of Nanog and metachronous state transition. Knockdown of Nanog allows exit without ERK activation. However, transition to formative pluripotency does not proceed and cells collapse to an indeterminate identity. This outcome is due to failure to maintain expression of the central pluripotency factor Oct4. Thus, during formative transition ERK signalling both dismantles the naïve state and preserves pluripotency. These results illustrate how a single signalling pathway can both initiate and secure transition between cell states.

Keywords:  Differentiation; ERK; Formative; Mouse embryonic stem cells; Naive; Nanog

DOI:  https://doi.org/10.1242/dev.203106
Methods Mol Biol. 2024 ;2844 145-156

Exploring the Function of Gene Promoter Regulatory Elements Using CRISPR Tools.

Sara Yousefi Taemeh, Nima Dehdilani, Lena Goshayeshi, Hesam Dehghani.

  Gene promoters serve as pivotal regulators of transcription, orchestrating the initiation, rate, and specificity of gene expression, resulting in cellular diversity found among distinct cell types within multicellular organisms. Identification of the sequence and function of promoters' regulatory elements and their complex interaction with transcription factors, enhancers, silencers, and insulators is fundamental to coordinated transcriptional processes within cells. Identifying these regulatory elements and scrutinizing their functions and interactions through the use of synthetic promoters can pave the way for researchers in various fields ranging from uncovering the origins of diseases associated with promoter mutations to harnessing these regulatory components in biotechnological applications.In this chapter, we describe the manipulation of regulatory elements within promoters, with a specific focus on the use of CRISPR technology on enhancers and silencer elements of the Ovalbumin gene promoter. We explain and discuss processes for the deletion of/interference with regulatory elements within the promoter, employing CRISPR-based approaches. Furthermore, we demonstrate that a CRISPR/Cas-manipulated promoter can activate gene transcription in cell types where it is normally inactive. This confirms that CRISPR technology can be effectively used to engineer synthetic promoters with desired characteristics, such as inducibility, tissue-specificity, or enhanced transcriptional strength. Such an approach provides valuable insights into the mechanisms and dynamics of gene expression, thereby offering new opportunities in the fields of biotechnology and medicine.

Keywords:  CRISPR technology; Gene editing; Promoter; Regulatory sequences

DOI:  https://doi.org/10.1007/978-1-0716-4063-0_10
bioRxiv. 2024 Jul 26. pii: 2024.07.26.605239. [Epub ahead of print]

Optimized reporters for multiplexed detection of transcription factor activity.

Max Trauernicht, Teodora Filipovska, Chaitanya Rastogi, Bas van Steensel.

In any given cell type, dozens of transcription factors (TFs) act in concert to control the activity of the genome by binding to specific DNA sequences in regulatory elements. Despite their considerable importance in determining cell identity and their pivotal role in numerous disorders, we currently lack simple tools to directly measure the activity of many TFs in parallel. Massively parallel reporter assays (MPRAs) allow the detection of TF activities in a multiplexed fashion; however, we lack basic understanding to rationally design sensitive reporters for many TFs. Here, we use an MPRA to systematically optimize transcriptional reporters for 86 TFs and evaluate the specificity of all reporters across a wide array of TF perturbation conditions. We thus identified critical TF reporter design features and obtained highly sensitive and specific reporters for 60 TFs, many of which outperform available reporters. The resulting collection of "prime" TF reporters can be used to uncover TF regulatory networks and to illuminate signaling pathways.HIGHLIGHTS: Systematic design and optimization of transcriptional reporters for 86 TFsCharacterization of TF-specific reporter design optimization rulesEvaluation of reporter TF-specificity across a wide array of TF perturbationsIdentification of a collection of 60 "prime" TF reporters with optimized performance.

DOI: https://doi.org/10.1101/2024.07.26.605239
STAR Protoc. 2024 Jul 30. pii: S2666-1667(24)00318-6. [Epub ahead of print]5(3): 103153

Protocol for the application of single-cell damage in murine intestinal organoid models.

Anna Elisabeth Seidler, Sören Donath, Lara Gentemann, Manuela Buettner, Alexander Heisterkamp, Stefan Kalies.

  Spatially defined organoid damage enables the study of cellular repair processes. However, capturing dynamic events in living tissues is technically challenging. Here, we present a protocol for the application of single-cell damage in intestinal organoid models. We describe steps for isolating and cultivating murine colon organoids, lentivirus generation and transduction of organoids, single-cell ablation by a femtosecond laser, and follow-up imaging analysis. We provide examples for the image acquisition pipeline of dynamic processes in organoids using a confocal microscope. For complete details on the use and execution of this protocol, please refer to Donath et al.1,2.

Keywords:  Biophysics; Microscopy; Molecular Biology; Organoids; Single Cell

DOI:  https://doi.org/10.1016/j.xpro.2024.103153
Nat Protoc. 2024 Jul 30.

Quantifying genome-wide transcription factor binding affinities for chromatin using BANC-seq.

Roelof A Wester, Hannah K Neikes, Rik G H Lindeboom, Michiel Vermeulen.

Transcription factors (TFs) bind specific DNA sequences to regulate transcription. Apart from DNA sequences, local factors such as DNA accessibility and chromatin structure determine the affinity of a TF for any given locus. Including these factors when measuring TF-DNA affinities has proven difficult. To address this challenge, we recently developed a method called binding affinities in native chromatin by sequencing (BANC-seq). In BANC-seq, intact mammalian nuclei are incubated with a concentration range of epitope-tagged TF, followed by either chromatin immunoprecipitation or cleavage under target and release using nuclease with spike-in DNA. This allows determination of apparent dissociation constant (KdApp) values, defined by the concentration of TF at which half-maximum binding occurs, across the genome. Here we present a detailed stepwise protocol for BANC-seq, including downstream data analysis. In principle, any molecular biologist should be able to perform a BANC-seq experiment in as little as 1.5 d (excluding analysis). However, preprocessing and analysis of the sequencing data does require some experience in command-line shell and R programming.

DOI: https://doi.org/10.1038/s41596-024-01026-7
Mol Cancer. 2024 Aug 02. 23(1): 154

Cancer, metastasis, and the epigenome.

Saurav Kiri, Tyrone Ryba.

  Cancer is the second leading cause of death worldwide and disease burden is expected to increase globally throughout the next several decades, with the majority of cancer-related deaths occurring in metastatic disease. Cancers exhibit known hallmarks that endow them with increased survival and proliferative capacities, frequently as a result of de-stabilizing mutations. However, the genomic features that resolve metastatic clones from primary tumors are not yet well-characterized, as no mutational landscape has been identified as predictive of metastasis. Further, many cancers exhibit no known mutation signature. This suggests a larger role for non-mutational genome re-organization in promoting cancer evolution and dissemination. In this review, we highlight current critical needs for understanding cell state transitions and clonal selection advantages for metastatic cancer cells. We examine links between epigenetic states, genome structure, and misregulation of tumor suppressors and oncogenes, and discuss how recent technologies for understanding domain-scale regulation have been leveraged for a more complete picture of oncogenic and metastatic potential.

Keywords:  Chromatin state; EMT; Epigenetic; Metastasis; Topology; Transcription

DOI:  https://doi.org/10.1186/s12943-024-02069-w
bioRxiv. 2024 Jul 16. pii: 2024.07.12.603216. [Epub ahead of print]

CELLNET technology: Spatially organized, functional 3D networks at single cell resolution.

Arun Poudel, Puskal Kunwar, Ujjwal Aryal, Anna-Blessing Merife, Pranav Soman.

Cells possess the remarkable ability to generate tissue-specific 3D interconnected networks and respond to a wide range of stimuli. Understanding the link between the spatial arrangement of individual cells and their networks' emergent properties is necessary for the discovery of both fundamental biology as well as applied therapeutics. However, current methods spanning from lithography to 3D photo-patterning to acoustofluidic devices are unable to generate interconnected and organized single cell 3D networks within native extracellular matrix (ECM). To address this challenge, we report a novel technology coined as CELLNET. This involves the generation of crosslinked collagen within multi-chambered microfluidic devices followed by femtosecond laser ablation of 3D microchannel networks and cell seeding. Using model cells, we show that cell migrate within ablated networks within hours, self-organize and form viable, interconnected, 3D networks in custom architectures such as square grid, concentric circle, parallel lines, and spiral patterns. Heterotypic CELLNETs can also be generated by seeding multiple cell types in side-chambers of the devices. The functionality of cell networks can be studied by monitoring the real-time calcium signaling response of individual cells and signal propagation within CELLNETs when subjected to flow stimulus alone or a sequential combination of flow and biochemical stimuli. Furthermore, user-defined disrupted CELLNETs can be generated by lethally injuring target cells within the 3D network and analyzing the changes in their signaling dynamics. As compared to the current self-assembly based methods that exhibit high variability and poor reproducibility, CELLNETs can generate organized 3D single-cell networks and their real-time signaling responses to a range of stimuli can be accurately captured using simple cell seeding and easy-to-handle microfluidic devices. CELLNET, a new technology agnostic of cell types, ECM formulations, 3D cell-connectivity designs, or location and timing of network disruptions, could pave the way to address a range of fundamental and applied bioscience applications.Teaser: New technology to generate 3D single cell interconnected and disrupted networks within natural extracellular matrix in custom configurations.

DOI: https://doi.org/10.1101/2024.07.12.603216
Nat Rev Genet. 2024 Jul 29.

Chromosomal instability as a driver of cancer progression.

Xuelan Chen, Albert S Agustinus, Jun Li, Melody DiBona, Samuel F Bakhoum.

Chromosomal instability (CIN) refers to an increased propensity of cells to acquire structural and numerical chromosomal abnormalities during cell division, which contributes to tumour genetic heterogeneity. CIN has long been recognized as a hallmark of cancer, and evidence over the past decade has strongly linked CIN to tumour evolution, metastasis, immune evasion and treatment resistance. Until recently, the mechanisms by which CIN propels cancer progression have remained elusive. Beyond the generation of genomic copy number heterogeneity, recent work has unveiled additional tumour-promoting consequences of abnormal chromosome segregation. These mechanisms include complex chromosomal rearrangements, epigenetic reprogramming and the induction of cancer cell-intrinsic inflammation, emphasizing the multifaceted role of CIN in cancer.

DOI: https://doi.org/10.1038/s41576-024-00761-7