Biology subjects

Steinmetz, L. M.

Publications and source records attributed to Steinmetz, L. M..

At least 19 recordsLinked to original sources

bioRxiv2026

PHAROS: turning single-cell perturbation models into target-directed drug-combination screens

Combination therapies are central to cancer treatment, but exhaustive screening is impractical. We introduce PHAROS, a framework that turns a pretrained single-cell perturbation model into a target-directed search engine for drug combinations. PHAROS predicts how a cell population changes under a drug, one drug at a time, then chains these predictions together to simulate drug combinations. It scores each simulated outcome against the desired target state and uses a search algorithm to find the most promising combinations, all without retraining the underlying model. Across two independent combinatorial perturbation datasets, PHAROS recovered exact or mechanism-matched two-drug responses in cell lines, both seen and unseen during model training. Its rankings were specific to the requested conversion and were not explained by single-drug effects, additive effects, or shared mechanism of action. In exploratory analyses of patient-derived metastatic HR+/HER2$-$ breast tumors and basal cell carcinoma (BCC), PHAROS prioritized FDA-approved regimens, distinguished combinations by their predicted tumor-versus-immune objective profiles, and nominated pathway-level hypotheses, while explicitly identifying both tumor cohorts as outside the model's supported distribution. PHAROS provides a modular route from pretrained virtual-cell models to inverse, single-cell combination-screening platforms.

bioinformatics↗

bioRxiv2026

Nuclear size is genetically controlled and influences cell fate

The proportional scaling between nuclear and cell size was first described more than 150 years ago and is among the most conserved features of cellular organization. Yet the mechanisms that establish this scaling and its physiological significance have remained unresolved. Here, we address both questions by combining image-enabled cell sorting with genome-wide CRISPR screening, transcriptomics and functional analyses. We identify more than 180 regulators of the nuclear-to-cell ratio and show that distinct classes of genes independently control nuclear and cell size. RNA metabolism predominantly regulates nuclear size, whereas protein synthesis and degradation primarily regulate cell size. This supports a model in which differences in macromolecular partitioning between the nucleus and cytoplasm contribute to osmotic regulation of nuclear size together with mechanical constraints imposed by chromatin, the cytoskeleton, and the nuclear envelope. Changes in nuclear size cause widespread transcriptional remodeling that is independent of changes in cell size. Cells with smaller nuclei exhibit reduced PRC2-dependent H3K27 trimethylation, activation of developmental gene-expression programmes and repression of cell-cycle genes. Consistent with these changes, mouse embryonic stem cells with smaller nuclei show an increased propensity to exit pluripotency and initiate differentiation in response to retinoic acid. Together, our findings provide a mechanistic framework for nuclear size scaling and establish nuclear size as a physical regulator of gene expression and cell-state transitions, linking cellular architecture to cell fate.

cell biology↗

bioRxiv2026

Cell organization and disturbance-mapping using image activated cell-profiling (CODIAC)

Protein localization and abundance are tightly regulated to maintain cellular homeostasis, and their dysregulation is a hallmark of disease. However, methods that monitor the spatial localization changes of many proteins at once, and at scale, remain limited. Here we introduce CODIAC (Cell Organization and Disturbance-mapping using Image-Activated Cell-profiling), which couples image-enabled cell sorting (ICS) with improved Cas12a-assisted endogenous PCR tagging and machine-learning-guided gating to profile pools of fluorescently tagged proteins. Rather than isolating discrete cell populations, CODIAC reads the "phenotypic fingerprint" of each fluorescently tagged protein, defined by its distribution across a fixed set of image-defined sort bins. The method resolves localization-associated phenotypes, reproducibly profiles complex pools, generalizes to proteins absent from the training set, and detects shifts in protein localization and abundance following chemical perturbation. CODIAC thus extends ICS from outlier screening toward systematic profiling of the spatial proteome and its remodeling under perturbation.

cell biology↗

bioRxiv2026

Image-based transposon screening reveals a flavin reductase that restrains intracellular Salmonella replication in macrophages

Intracellular bacterial pathogens can survive and replicate within host cells, yet an isogenic population follows divergent fates: some bacteria are killed, some arrest growth, and others replicate to high numbers. Identifying the bacterial functions behind each fate requires recovering mutants from within the host cells displaying it. But systematic approaches, such as transposon insertion analysis, can only estimate fitness of mutants from the whole infected population. Here, we developed an approach that couples a genome-wide mutagenesis library to image-enabled cell sorting (ICS), sorting infected cells by the number of bacteria they contain and assigning mutants to defined replication outcomes. We applied this approach to Salmonella enterica serovar Typhimurium (S.Tm) in macrophages, and revealed genes required for replication from genes that restrain it, whose disruption increased replication. Among the latter we identified the cytosolic flavin reductase Fre, which supplies reduced flavins to a broad range of bacterial processes. We uncovered a mechanism whereby loss of fre protected S.Tm from oxidative and nitrosative damage and increased bacterial numbers. Inside macrophages this advantage was mediated by the upregulation of the iron-sulfur-independent cytochrome bd-I oxidase. By resolving a mutant library into phenotypically defined subpopulations, this framework can be applied to characterize bacterial or host genes that drive infection phenotypes in any infection model.

microbiology↗

bioRxiv2026

Ubiquitin selective ribosome profiling reveals systematic principles of co-translational quality control

Protein biogenesis is a stress- and error-sensitive process that can lead to nascent protein misfolding and aggregation, challenging cellular proteostasis.1 Co-translational ubiquitination (CTU) is a critical surveillance mechanism,2,3 yet its regulatory principles remain unclear due to limited known substrates and the lack of translatome-wide methods to query CTU. Here, we introduce UbSeRP, an approach that enables ubiquitin linkage-specific, and translatome-wide mapping of CTU. Focusing on ribosome-associated quality control (RQC), we expand the known endogenous RQC substrates in S. cerevisiae from few4-7 to thousands, reveal that only a subset of all identified disomes undergoes RQC, and uncover biophysical features of nascent chains that predict RQC engagement. We also identify widespread RQC-independent ubiquitination of nascent proteins, implicating broader roles of CTU in protein complex assembly. In a chronological aging model, we show that aging remodels translation and diminishes RQC engagement, favoring RQC-independent proteasomal pathways. Our findings provide systematic insight into the determinants and adaptability of CTU in maintaining proteostasis under changing physiological conditions.

biochemistry↗

bioRxiv2026

Multimodal phenotyping defines variant-to-function maps for RBM20 in dilated cardiomyopathy

Multiplex assays of variant effects have linked thousands of genotypes to fitness effects, yet we lack profound understanding of how variants impact molecular phenotypes. Here, we introduce a deep mutational scanning framework that quantifies disease-determining molecular phenotypes in human cells, allowing readouts of protein localization and splicing regulatory function at scale. Applied to the dilated cardiomyopathy (DCM)-associated protein RBM20, we profiled [~]4,300 amino acid substitutions across disease-linked protein domains. Complemented by structure-function investigations of RBM20 bound to its nuclear import receptor TNPO3, we discover new variant hotspots affecting protein function. Finally, we systematically probed nuclear relocalization to identify variants that may be amenable to this therapeutic strategy. Together, we create comprehensive variant-to-function maps that predict variant impact, enhance clinical interpretation, and stratify RBM20-mediated DCM into mechanistically distinct therapeutic classes.

genomics↗

bioRxiv2026

PerturbPlan: An analytical framework for designing Perturb-seq experiments

CRISPR screens with single-cell RNA-seq readouts provide a powerful tool for characterizing the functions of noncoding elements and genes. However, designing these experiments to balance statistical power and cost is challenging, given the large number of design parameters. The only available tool for this purpose is a simulation-based power calculator, but it is computationally costly and requires high-performance computing to run. We derive a novel analytical formula for the power to detect perturbation-expression associations, recapitulating power estimates from the simulation-based tool while reducing runtime by up to seven orders of magnitude. This acceleration unlocks the possibility of interactive single-cell CRISPR screen design. Accordingly, we develop PerturbPlan, an interactive web application built on the analytical power formula. PerturbPlan helps users address 11 design questions for two types of single-cell CRISPR screens, Perturb-seq and targeted Perturb-seq (TAP-seq). We apply PerturbPlan to carry out a comparative analysis of three recent Perturb-seq designs, demonstrating how optimal design varies across experiments of different scales. We also use PerturbPlan to quantify the cost savings of a recent TAP-seq study relative to a hypothetical Perturb-seq study assaying the same perturbations, illustrating how the tool can inform decisions about targeted versus whole-transcriptome readouts. In sum, PerturbPlan is the first tool to facilitate flexible and interactive design of well-powered single-cell CRISPR screen experiments.

genomics↗

bioRxiv2026

Genome-scale mapping of variant, enhancer and gene function in primary human CD4+ T cells

CD4+ T cells harbor a disproportionate enrichment of immune disease risk loci and represent the primary cellular context for immune disease biology, yet the genes and regulatory programs these variants affect remain largely unknown. We combined targeted Perturb-seq of 1,032 cis-regulatory elements (CREs) overlapping 4,724 variants across 14 immune diseases with genome-wide Perturb-seq of all expressed genes in primary human CD4+ T cells, spanning 4.1 million cells. We identified 626 CRE-gene pairs, and connected CRE targets to downstream regulatory cascades. At the TYK2 and DEXI/CLEC16A loci, we resolved target genes and linked noncoding variants to inflammatory and metabolic programs. Across diseases, we revealed that dispersed variants converged on shared and disease-specific programs. Our work provides a framework for tracing variant-to-CRE-to-gene-to-network in disease-relevant primary cells.

genetics↗

bioRxiv2025

Improved cloning-free one-step CRISPR-Cas12a-assisted tagging of mammalian genes using PCR generated reagents (PCR tagging)

Precise DNA integration in mammalian genomes using CRISPR/Cas based strategies remains challenging due to competing DNA repair pathways. We present an update for our CRISPR/Cas12a-assisted PCR-tagging method using linear donor cassettes and self-targeting gRNAs for gene tagging. By incorporating 2A-linked selection markers and chemically modifying PCR cassettes with phosphorothioate and biotin, along with the inhibition of NHEJ, we improve precise insertions while reducing unwanted integration products and we devise a variant of the method for seamless gene tagging. To profile integration outcomes, we further developed our integration site sequencing protocol (Tn5-Anchor-Seq, Meurer et al., 2018) to obtain a protocol using long-read (ONT) sequencing coupled with a custom computational pipeline. This enables high-resolution quantification of all integration events with respect to integration by HDR or NHEJ, as well as the presence of concatemers and off-target integrations. Our data reveal gene-specific editing profiles and demonstrate that combining cassette design with DNA repair modulation yields up to 5-fold increases in HDR and suppression of concatemers and indels. Together, our improvements outline an robust strategy for efficient and high-fidelity gene tagging in mammalian cells, facilitating functional genomics and cell engineering applications.

molecular biology↗

bioRxiv2025

An unbiased survey of distal element-gene regulatory interactions with direct-capture targeted Perturb-seq

Identifying the impact of distal regulatory elements on gene expression is a core challenge in human genetics. Large-scale CRISPR screens have not captured lower effect size element-gene interactions due to selection bias and limited statistical power. We developed a framework for highly powered CRISPR screens, consisting of Direct-Capture Targeted Perturb-seq (DC-TAP-seq), unbiased target selection, and a pipeline accounting for statistical power. Surveying 10,000 random distal element-gene pairs revealed most element-gene interactions have effect sizes <10%, which were virtually undetectable in prior studies. Most interactions occur within 100kb, many elements bind CTCF without classical enhancer chromatin, and housekeeping genes have similar frequencies of distal regulatory elements but with weaker effects. We also highlight limitations of predictive models and suggest that new models consider elements with smaller effect sizes. Our study provides an expanded view of distal regulatory elements and a framework for building more comprehensive maps of distal regulation.

genomics↗

bioRxiv2025

Mutational scanning by multiplexed genome editing of the essential transcription termination factor Nrd1.

Proteins operate through a few critical residues, yet most proteins remain uncharacterized at the deep molecular resolution, particularly within essential genes, where functional dissection is obstructed by lethality. Here, we establish a platform for mutational scanning of essential genes at their endogenous locus, combining a repressible complementation system with multiplexed CRISPR-based genome editing in budding yeast. Our approach provides a generalizable framework for dissecting essential protein function in vivo, expanding the capacity to map critical residues underlying essential cellular processes. We applied this strategy to NRD1, encoding an essential RNA Polymerase II (RNAPII) termination factor and performed a systematic alanine scanning with near-saturation coverage. We discovered novel and unexpected lethal mutations in the CTD-interacting domain (CID), thus revealing an unanticipated importance for this domain. Overall, our results demonstrate the power of our mutation scanning platform to map critical residues underlying essential cellular processes.

genetics↗

bioRxiv2025

CAMK2D causes heart failure in RBM20 cardiomyopathy

Although heart disease can arise from different etiologies, current treatment is not tailored to the different underlying causes but is rather a one-size-fits-all approach. Importantly, not all patients benefit from this treatment regimen, which means the number needed to treat is very high. Moreover, this makes clinical trials large and costly, limiting clinical translation. Thus, there is a high medical need to develop a first etiology-specific therapy. Mutations in RBM20, a splicing factor that targets multiple pivotal cardiac genes including TTN and CAMK2D, cause a clinically aggressive form of dilated cardiomyopathy (DCM) with a high risk of malignant ventricular arrhythmias. Here, we hypothesized that CAMK2D is the heart disease-causing target of RBM20. We crossed Camk2d- to Rbm20-deficient mice and found that double knockout (DKO) mice were protected from heart failure and sudden cardiac death. Phosphorylation of multiple CAMK2D targets was increased in Rbm20-deficient mouse hearts, which was reverted in DKO hearts, confirming that CAMK2D is not only misspliced but also overactivated. AAV9-mediated re-expression of single CAMK2D splice variants in DKO mice reintroduced cardiac dysfunction irrespective of the splice variant, unmasking that overactivation rather than missplicing underlies the detrimental phenotype. To test whether heart failure could pharmacologically be reversed, we treated heterozygous Rbm20-R636Q knock-in (KI) mice with hesperadin, a potent CAMK2 inhibitor, which rescued both cardiac function and ventricular geometry. These data demonstrate that overactivation of CAMK2D underlies heart failure in RBM20 cardiomyopathy. Pharmacological inhibition of CAMK2D could therefore become the first cause-directed DCM therapy.

molecular biology↗

bioRxiv2025

Sarcomeric Remodelling in Human Heart Failure unraveled by single molecule long read sequencing

Dysregulation of alternative splicing - mediated by factors such as RBM20 or SLM2 - can affect proper gene isoform control disrupts gene isoform homeostasis and underpins severe cardiomyopathy in both animal models and patients. Although innovative therapies target various sarcomeric components, the impact of isoform switching in cardiac disease remains poorly understood. Here, we applied nanopore long-read sequencing to map the full-length transcriptome of left ventricular tissue from thirteen nonfailing controls, ten patients with dilated cardiomyopathy (DCM), and ten with ischemic cardiomyopathy (ICM). Our analysis identified 78,520 transcripts, 31% of which represent novel isoforms of known genes. Notably, the transcriptomes of DCM and ICM were largely indistinguishable, indicating that end-stage heart failure is characterized by a convergent isoform landscape, irrespective of disease etiology. Among 11 prototypical sarcomere genes, 10 displayed highly significant isoform shifts (p= 5.23x10-45 - 2.89x10-200). Focusing on tropomyosin, we observed that while the predominant cardiac gene TPM1 showed moderate up-regulation of its transcript isoforms, transcripts derived from TPM3--typically expressed at lower levels in the healthy heart--were markedly increased in heart failure.

genetics↗

bioRxiv2024

Mapping enhancer-gene regulatory interactions from single-cell data

Mapping enhancers and their target genes in specific cell types is crucial for understanding gene regulation and human disease genetics. However, accurately predicting enhancer-gene regulatory interactions from single-cell datasets has been challenging. Here, we introduce a new family of classification models, scE2G, to predict enhancer-gene regulation. These models use features from single-cell ATAC-seq or multiomic RNA and ATAC-seq data and are trained on a CRISPR perturbation dataset including >10,000 evaluated element-gene pairs. We benchmark scE2G models against CRISPR perturbations, fine-mapped eQTLs, and GWAS variant-gene associations and demonstrate state-of-the-art performance at prediction tasks across multiple cell types and categories of perturbations. We apply scE2G to build maps of enhancer-gene regulatory interactions in heterogeneous tissues and interpret noncoding variants associated with complex traits, nominating regulatory interactions linking INPP4B and IL15 to lymphocyte counts. The scE2G models will enable accurate mapping of enhancer-gene regulatory interactions across thousands of diverse human cell types.

genetics↗

bioRxiv2024

Pathogenic mechanisms of Plasmodium falciparum egress unveiled by a microvascular 3D blood-brain barrier model

Cerebral malaria is a severe neurovascular complication of Plasmodium falciparum infection, with high mortality rates even after treatment with effective antimalarials. Limitations in current experimental models have hindered our knowledge of the disease. We developed a 3D blood-brain barrier (BBB) model with enhanced barrier properties using primary brain endothelial cells, astrocytes and pericytes. Exposure to parasite egress products increased microvascular permeability, likely due to transcriptional downregulation of junctional and vascular development genes in endothelial cells. In addition, it increased the expression of ferroptosis markers, antigen presentation and type I interferon genes and upregulated the JAK-STAT pathway across all BBB cell types. Incubation with cytoadherent schizont-stage P. falciparum-infected erythrocytes induced a similar, but highly localized transcriptional shift, along with inter-endothelial gaps at sites of parasite egress, significantly increasing permeability. The co-administration of egress products with the JAK-STAT inhibitor Ruxolitinib prevented junctional disruption and BBB breakdown. These findings provide key insights into the parasite-mediated mechanisms driving brain microvascular pathogenesis in cerebral malaria and suggest potential avenues for adjunctive therapies.

microbiology↗

bioRxiv2024

Cohesin-mediated 3D contacts tune enhancer-promoter regulation

Enhancers are key drivers of gene regulation thought to act via 3D physical interactions with the promoters of their target genes. However, genome-wide depletions of architectural proteins such as cohesin result in only limited changes in gene expression, despite a loss of contact domains and loops. Consequently, the role of cohesin and 3D contacts in enhancer function remains debated. Here, we developed CRISPRi of regulatory elements upon degron operation (CRUDO), a novel approach to measure how changes in contact frequency impact enhancer effects on target genes by perturbing enhancers with CRISPRi and measuring gene expression in the presence or absence of cohesin. We systematically perturbed all 1,039 candidate enhancers near five cohesin-dependent genes and identified 34 enhancer-gene regulatory interactions. Of 26 regulatory interactions with sufficient statistical power to evaluate cohesin dependence, 18 show cohesin-dependent effects. A decrease in enhancer-promoter contact frequency upon removal of cohesin is frequently accompanied by a decrease in the regulatory effect of the enhancer on gene expression, consistent with a contact-based model for enhancer function. However, changes in contact frequency and regulatory effects on gene expression vary as a function of distance, with distal enhancers (e.g., >50Kb) experiencing much larger changes than proximal ones (e.g., <50Kb). Because most enhancers are located close to their target genes, these observations can explain how only a small subset of genes -- those with strong distal enhancers -- are sensitive to cohesin. Together, our results illuminate how 3D contacts, influenced by both cohesin and genomic distance, tune enhancer effects on gene expression.

molecular biology↗

bioRxiv2024

Functional phenotyping of genomic variants using multiomic scDNA-scRNA-seq

Genomic variation ranging from single nucleotide polymorphisms to structural variants can impact gene function and expression, contributing to disease mechanisms such as cancer progression. The systematic study of this variation is hindered by inefficient precision editing tools making it challenging to confidently link genotype and gene expression in pooled screens. Additionally, assessing heterogenous variants in primary tumor samples at scale is difficult with current single-cell technologies. We developed droplet-based multiomic targeted scDNA-scRNAseq (SDR-seq) to precisely link genotypes with gene expression profiles in high-throughput. SDR-seq simultaneously assesses up to 480 RNA and gDNA targets with high coverage and sensitivity across thousands of cells. Using SDR-seq, we associate coding and non-coding variants with distinct gene expression profiles in human iPSCs. Furthermore, we demonstrate that in primary B-cell lymphoma samples, cells with a higher mutational burden exhibit elevated B-cell receptor signaling and tumorigenic gene expression. SDR-seq has broad potential for gaining functional insights into regulatory mechanisms encoded by genetic variants at diverse loci, advancing our ability to study gene expression regulation and its implications for disease.

molecular biology↗

bioRxiv2024

Dissecting quantitative trait nucleotides by saturation genome editing

Genome editing technologies have the potential to transform our understanding of how genetic variation gives rise to complex traits through the systematic engineering and phenotypic characterization of genetic variants. However, there has yet to be a system with sufficient efficiency, fidelity, and throughput to comprehensively identify causal variants at the genome scale. Here we explored the ability of templated CRISPR editing systems to install natural variants genome-wide in budding yeast. We optimized several approaches to enhance homology-directed repair (HDR) with donor DNA templates, including donor recruitment to target sites, single-stranded donor production by bacterial retrons, and in vivo plasmid assembly. We uncovered unique advantages of each system that we integrated into a single superior system named MAGESTIC 3.0. We used MAGESTIC 3.0 to dissect causal variants residing in 112 quantitative trait loci across 32 environmental conditions, revealing an enrichment for missense variants and loci with multiple causal variants. MAGESTIC 3.0 will facilitate the functional analysis of the genome at single-nucleotide resolution and provides a roadmap for improving template-based genome editing systems in other organisms.

genetics↗