Cell Painting, a high-content image-based assay for morphological profiling using multiplexed fluorescent dyes

Matchmaking at the cell surface using bispecifics to put cells on their best behavior

Intermolecular relationships at the cell surface dictate the behavior and regulatory network of cells. Such interactions often require precise spatial control for optimal response. By binding simultaneously to two different target sites, bispecific binders can bridge molecules of interest. Despite decades of bispecific development, only recently have bispecifics been engineered with programmable, tuneable geometries to replicate natural interaction geometries or achieve new responses from unnatural arrangements. This review highlights emerging methods of protein engineering and modular bioconjugation to control pairing and orientation of binders in bispecific scaffolds. We also describe novel biophysical and phenotypic assays, which reveal how bispecific geometries change cell fate. These approaches are informing design of next-generation precision therapeutics, as well as uncovering fundamental features of signal integration.

Exploring NLRP3-related phenotypic fingerprints in human macrophages using Cell Painting assay

Existing research has proven difficult to understand the interplay between upstream signaling events during NLRP3 inflammasome activation. Additionally, events downstream of inflammasome complex formation such as cytokine release and pyroptosis can exhibit variation, further complicating matters. Cell Painting has emerged as a prominent tool for unbiased evaluation of the effect of perturbations on cell morphological phenotypes. Using this technique, phenotypic fingerprints can be generated that reveal connections between phenotypes and possible modes of action. To the best of our knowledge, this was the first study that utilized Cell Painting on human THP-1 macrophages to generate phenotypic fingerprints in response to different endogenous and exogenous NLRP3 inflammasome triggers and to identify phenotypic features specific to NLRP3 inflammasome complex formation. Our results demonstrated that not only can Cell Painting generate morphological fingerprints that are NLRP3 trigger-specific but it can also identify cellular fingerprints associated with NLRP3 inflammasome activation.

Deciphering the phenotypic, inflammatory, and endocrine disrupting impacts of e-waste plastic-associated chemicals

As the volume of plastic waste from electrical and electronic equipment (WEEE) continues to rise, a significant portion is disposed of in the environment, with only a small fraction being recycled. Both disposal and recycling pose unknown health risks that require immediate attention. Existing knowledge of WEEE plastic toxicity is limited and mostly relies on epidemiological data and association studies, with few insights into the underlying toxicity mechanisms. Therefore, this study aimed to perform comprehensive chemical screening and mechanistic toxicological assessment of WEEE plastic-associated chemicals. Chemical analysis, utilizing suspect screening based on high-resolution mass spectrometry, along with quantitative target chemical analysis, unveiled numerous hazardous compounds including polyaromatic compounds, organophosphate flame retardants, phthalates, benzotriazoles, etc. Toxicity endpoints included perturbation of morphological phenotypes using the Cell Painting assay, inflammatory response, oxidative stress, and endocrine disruption. Results demonstrated that WEEE plastic chemicals altered the phenotypes of the cytoskeleton, endoplasmic reticulum, and mitochondria in a dose-dependent manner. In addition, WEEE chemicals induced inflammatory responses in resting macrophages and altered inflammatory responses in lipopolysaccharide-primed macrophages. Furthermore, WEEE chemicals activated the nuclear factor erythroid 2-related factor 2 (Nrf2) pathway, indicating oxidative stress, and the aryl hydrocarbon receptor (AhR). Endocrine disruption was also observed through the activation of estrogenic receptor-α (ER-α) and the induction of anti-androgenic activity. The findings show that WEEE plastic-associated chemicals exert effects in multiple subcellular sites, via different receptors and mechanisms. Thus, an integrated approach employing both chemical and toxicological methods is essential for comprehensive assessment of the toxicity mechanisms and cumulative chemical burden of WEEE plastic-associated chemicals.

Sea urchin immune cells and associated microbiota co-exposed to iron oxide nanoparticles activate cellular and molecular reprogramming that promotes physiological adaptation

The innate immune system is the first player involved in the recognition/interaction with nanomaterials. Still, it is not the only system involved. The co-evolution of the microbiota with the innate immune system built an interdependence regulating immune homeostasis that is poorly studied. Herein, the simultaneous interaction of iron-oxide nanoparticles (Fe-oxide NPs), immune cells, and the microbiota associated with the blood of the sea urchin Paracentrotus lividus was explored by using a microbiota/immune cell model in vitro-ex vivo and a battery of complementary tools, including Raman spectroscopy, 16S Next-Generation Sequencing, high-content imaging, NanoString nCounter. Our findings highlight the P. lividus immune cells and microbiota dynamics in response to Fe-oxide NPs, including i) morphological rearrangement and immune cell health status maintenance (intracellular trafficking increasing, no phenotypic alterations or caspase 3/7 activation), ii) transcriptomic reprogramming in immune cells (Smad6, Lmo2, Univin, suPaxB, Frizzled-7, Fgfr2, Gp96 upregulation), iii) immune signaling unchanged (e.g., P-p38 MAPK, P-ERK, TLR4, IL-6 protein level unchanged), iv) enrichment in extracellular vesicle released in the co-culture medium, and v) a shift in the composition of microbial groups mainly in favor of Gram-positive bacteria (e.g., Firmicutes, Actinobacteria),. Our findings suggest that Fe-oxide NPs induce a multi-level immune cell-microbiota response restoring homeostasis.

Nellie: automated organelle segmentation, tracking and hierarchical feature extraction in 2D/3D live-cell microscopy

Cellular organelles undergo constant morphological changes and dynamic interactions that are fundamental to cell homeostasis, stress responses and disease progression. Despite their importance, quantifying organelle morphology and motility remains challenging due to their complex architectures, rapid movements and the technical limitations of existing analysis tools. Here we introduce Nellie, an automated and unbiased pipeline for segmentation, tracking and feature extraction of diverse intracellular structures. Nellie adapts to image metadata and employs hierarchical segmentation to resolve sub-organellar regions, while its radius-adaptive pattern matching enables precise motion tracking. Through a user-friendly Napari-based interface, Nellie enables comprehensive organelle analysis without coding expertise. We demonstrate Nellie's versatility by unmixing multiple organelles from single-channel data, quantifying mitochondrial responses to ionomycin via graph autoencoders and characterizing endoplasmic reticulum networks across cell types and time points. This tool addresses a critical need in cell biology by providing accessible, automated analysis of organelle dynamics.

Total syntheses of cyclohelminthol I-IV reveal a new cysteine-selective covalent reactive group

Biocompatible covalent reactive groups (CRGs) play pivotal roles in several areas of chemical biology and the life sciences, including targeted covalent inhibitor design and preparation of advanced biologic drugs, such as antibody-drug conjugates. In this study, we present the discovery that the small, chlorinated polyketide natural product cyclohelminthiol II (CHM-II) acts as a new type of cysteine/thiol-targeting CRG incorporating both reversible and irreversible reactivity. We devise the first syntheses of four simple cyclohelminthols, (±)-cyclohelminthol I-IV, with selective chlorinations (at C2 and C5) and a Ni-catalyzed reductive cross coupling between an enone, a vinyl bromide and triethylsilyl chloride as the key steps. Unbiased biological profiling (cell painting) was used to discover a putative covalent mechanism for CHM-II in cells with subsequent validation experiments demonstrating mechanistic similarity to dimethyl fumarate (DMF) - a known (covalent) drug used in the treatment of multiple sclerosis. Focused biochemical experiments revealed divergent thiol-reactivity inherent to the CHM-II scaffold and through further chemical derivatization of CHM-II we applied activity-based protein profiling (ABPP)-workflows to show exclusive cysteine-labelling in cell lysate. Overall, this study provides both efficient synthetic access to the CHM-II chemotype - and neighboring chemical space - and proof-of-concept for several potential applications of this new privileged CRG-class within covalent chemical biology.

Evaluating feature extraction in ovarian cancer cell line co-cultures using deep neural networks

Single-cell image analysis is crucial for studying drug effects on cellular morphology and phenotypic changes. Most studies focus on single cell types, overlooking the complexity of cellular interactions. Here, we establish an analysis pipeline to extract phenotypic features of cancer cells cultured with fibroblasts. Using high-content imaging, we analyze an oncology drug library across five cancer and fibroblast cell line co-culture combinations, generating 61,440 images and ∼170 million single-cell objects. Traditional phenotyping with CellProfiler achieves an average enrichment score of 62.6% for mechanisms of action, while pre-trained neural networks (EfficientNetB0 and MobileNetV2) reach 61.0% and 62.0%, respectively. Variability in enrichment scores may reflect the use of multiple drug concentrations since not all induce significant morphological changes, as well as the cellular and genetic context of the treatment. Our study highlights nuanced drug-induced phenotypic variations and underscores the morphological heterogeneity of ovarian cancer cell lines and their response to complex co-culture environments.

Cellular Activity of CQWW Nullomer-Derived Peptides

Analysis of observed protein sequences across all species within the UniProtKB/Swiss-Prot data set reveals CQWW as the shortest absent stretch of amino acids. While DNA can be found encoding the CQWW sequence, it has never been observed to be translated or included in manually curated sets of proteins, existing only in predicted, tentative sequences and in a single mature antibody sequence. We have synthesized this "nullomer" peptide, along with 13 derivatives, reversed, truncated, stereoisomers, and alanine-scanning peptides, conjugated to polyarginine stretches to increase cellular uptake. We observed their impact against a healthy neuronal line and six patient-derived glioblastoma cell lines spanning three clinical subtypes. Results reveal IC50 values averaging 4.9 μM for inhibition of cell survival across tested oncogenic cell lines. High-content phenotypic analysis of cellular features and reverse-phase protein arrays failed to discern a clear mode of action for the nullomer peptide but suggests mitochondrial impairment through the inhibition of GSK3 and isoforms, supported by observations of reduced mitochondrial stain intensities. With a recent increase in interest in nullomer peptides, we see the results in this study as a starting point for further investigation into this potentially therapeutic peptide class.

CellFIE: Integrating Pathway Discovery With Pooled Profiling of Perturbations Uncovers Pathways of Huntington's Disease, Including Genetic Modifiers of Neuronal Development and Morphology

Genomic screens and GWAS are powerful tools for identifying disease-modifying genes, but it is often challenging to understand the pathways by which these genes function. Here, we take an integrated approach that combines network analysis and an imaging-based pooled genetic perturbation study to examine modifiers of Huntington's disease (HD). The computational analysis highlighted several genes in a subnetwork enriched for modifiers of neuronal development and morphology. To test the functional roles of these genes, we developed an experimental pipeline that allows pooled CRISPRi KD of 21 genes in human iPSC-derived neurons followed by optical analysis of genotypes, neuronal arborization, multiplexed pathway activity and morphological fingerprint readout. This approach recovered known genes involved in morphology and confirmed unexpected links from the network between several genetic modifiers of HD and morphology. Our approach overcomes challenges in pooled measurement of neuronal function and health and could be adapted for other phenotypes in HD and other neurological diseases.

Single-cell morphological tracking of cell states to identify small-molecule modulators of liver differentiation

We have developed a single-cell assay that combines Cell Painting-a morphological profiling assay-with trajectory inference analysis. We have applied this morphological trajectory inference to the bi-potent HepaRG liver progenitor cell line allowing us to track liver cell fate and map small-molecule-induced changes using a morphological atlas of liver cell differentiation. Our overarching goal is to demonstrate the potential of Cell Painting to study biological processes as continuous trajectories at the single-cell level, enhancing resolution and biological understanding. This work has identified small-molecule Src family kinase inhibitors that promote the differentiation of HepaRG cells toward a hepatocyte-like lineage as well as primary human hepatic progenitor cells toward a hepatocyte-like phenotype in vitro. These findings could significantly advance research on liver cell regeneration mechanisms and facilitate the development of cell-based and small-molecule therapies.

Rare missense variants of the leukocyte common antigen related receptor (LAR) display reduced activity in transcellular adhesion and synapse formation

The leukocyte common antigen related receptor (LAR) is a member of the LAR receptor protein tyrosine phosphatase (RPTP) family of synaptic adhesion molecules that contribute to the proper alignment and specialization of synaptic connections in the mammalian brain. LAR-RPTP members have been genetically associated with neuropsychiatric disorders, but the molecular consequences of genetic perturbations of LAR remain unstudied. Using exome sequencing data from psychiatric patients and controls, we identify rare missense variants of LAR that render the extracellular domain (ECD) unstable and susceptible to proteolytic cleavage. Using recombinant and cellular systems, we describe three variants that cause disruption of the LAR:NGL-3 interaction, which results in loss of transcellular adhesion and synaptogenic effects. Furthermore, we show that overexpression of two of these variants elicit altered morphological phenotypes in an imaging-based morphological profiling assay compared to wild type LAR, suggesting that destabilization of the LAR ECD has broad effects on LAR function. In conclusion, our study identifies three rare, missense variants in LAR that could provide insights into LAR involvement with psychiatric pathobiology.

Automated detection of c-Fos-expressing neurons using inhomogeneous background subtraction in fluorescent images

Although many methods for automated fluorescent-labeled cell detection have been proposed, not all of them assume a highly inhomogeneous background arising from complex biological structures. Here, we propose an automated cell detection algorithm that accounts for and subtracts the inhomogeneous background by avoiding high-intensity pixels in the blur filtering calculation. Cells were detected by intensity thresholding in the background-subtracted image, and the algorithm's performance was tested on NeuN- and c-Fos-stained images in the mouse prefrontal cortex and hippocampal dentate gyrus. In addition, applications in c-Fos positive cell counting and the quantification for the expression level in double-labeled cells were demonstrated. Our method of automated detection after background assumption (ADABA) offers the advantage of high-throughput and unbiased analysis in regions with complex biological structures that produce inhomogeneous background.

BESTDR: Bayesian quantification of mechanism-specific drug response in cell culture

Understanding drug responses at the cellular level is essential for elucidating mechanisms of action and advancing preclinical drug development. Traditional dose-response models rely on simplified metrics, limiting their ability to quantify parameters like cell division, death, and transition rates between cell states. To address these limitations, we developed Bayesian Estimation of STochastic processes for Dose-Response (BESTDR), a novel framework modeling cell growth and treatment response dynamics to estimate concentration-response relationships using longitudinal cell count data. BESTDR quantifies rates in multi-state systems across multiple cell lines using hierarchical modeling to support high-throughput screening. We validated BESTDR with synthetic and experimental datasets, demonstrating its robustness and accuracy in estimating drug response. By integrating mechanistic modeling of cytotoxic, cytostatic and other effects, BESTDR enhances dose-response studies, facilitating robust drug comparisons and mechanism-specific analyses. BESTDR offers a versatile tool for early-stage preclinical research, paving the way for drug discovery and informed experimental design.

Orpinolide disrupts a leukemic dependency on cholesterol transport by inhibiting OSBP

Metabolic alterations in cancer precipitate in associated dependencies that can be therapeutically exploited. To meet this goal, natural product-inspired small molecules can provide a resource of invaluable chemotypes. Here, we identify orpinolide, a synthetic withanolide analog with pronounced antileukemic properties, via orthogonal chemical screening. Through multiomics profiling and genome-scale CRISPR-Cas9 screens, we identify that orpinolide disrupts Golgi homeostasis via a mechanism that requires active phosphatidylinositol 4-phosphate signaling at the endoplasmic reticulum-Golgi membrane interface. Thermal proteome profiling and genetic validation studies reveal the oxysterol-binding protein OSBP as the direct and phenotypically relevant target of orpinolide. Collectively, these data reaffirm sterol transport as a therapeutically actionable dependency in leukemia and motivate ensuing translational investigation via the probe-like compound orpinolide.

Cell Painting: a decade of discovery and innovation in cellular imaging

Modern quantitative image analysis techniques have enabled high-throughput, high-content imaging experiments. Image-based profiling leverages the rich information in images to identify similarities or differences among biological samples, rather than measuring a few features, as in high-content screening. Here, we review a decade of advancements and applications of Cell Painting, a microscopy-based cell-labeling assay aiming to capture a cell's state, introduced in 2013 to optimize and standardize image-based profiling. Cell Painting's ability to capture cellular responses to various perturbations has expanded owing to improvements in the protocol, adaptations for different perturbations, and enhanced methodologies for feature extraction, quality control, and batch-effect correction. Cell Painting is a versatile tool that has been used in various applications, alone or with other -omics data, to decipher the mechanism of action of a compound, its toxicity profile, and other biological effects. Future advances will likely involve computational and experimental techniques, new publicly available datasets, and integration with other high-content data types.

Progress in toxicogenomics to protect human health

Toxicogenomics measures molecular features, such as transcripts, proteins, metabolites and epigenomic modifications, to understand and predict the toxicological effects of environmental and pharmaceutical exposures. Transcriptomics has become an integral tool in contemporary toxicology research owing to innovations in gene expression profiling that can provide mechanistic and quantitative information at scale. These data can be used to predict toxicological hazards through the use of transcriptomic biomarkers, network inference analyses, pattern-matching approaches and artificial intelligence. Furthermore, emerging approaches, such as high-throughput dose-response modelling, can leverage toxicogenomic data for human health protection even in the absence of predicting specific hazards. Finally, single-cell transcriptomics and multi-omics provide detailed insights into toxicological mechanisms. Here, we review the progress since the inception of toxicogenomics in applying transcriptomics towards toxicology testing and highlight advances that are transforming risk assessment.

Identification of 5-amino-1,3,4-thiadiazole appended isatins as bioactive small molecules with polypharmacological activities

The identification of heterocyclic small molecules that cover unexplored chemical space is of great importance for the development of new small-molecule therapeutics. In this study, we synthesized a series of 5-amino-1,3,4-thiadiazoles appended isatins (UZ-1-20) that exhibited polypharmacological properties, as evaluated in a cell-painting assay assessing induced cellular morphological changes. A high hit rate ranging from 55% to 80% was observed for the tested compounds at varied concentrations. The most active compounds showed significant activity in inducing cellular morphological changes with a measured induction value of more than 30% and shared a high biological profiling similarity with an antifungal agent itraconazole and a chemokine receptor inhibitor. The synthesized compounds exhibited moderate to good antiproliferative activity against tested cancer cell lines in the MTT assay. Molecular docking studies were performed to theoretically probe and compare the binding modes between the most active UZ compounds and ITZ or BI-6901, respectively. Additionally, ADMET analysis indicated favorable pharmacokinetic parameters including good oral bioavailability, balanced hydrophilicity, and minimal toxicity. Overall, the findings in this study highlight the potential of developing the aminothiadiazole appended isatins as bioactive agents.

Semisupervised Contrastive Learning for Bioactivity Prediction Using Cell Painting Image Data

Morphological profiling has recently demonstrated remarkable potential for identifying the biological activities of small molecules. Alongside the fully supervised and self-supervised machine learning methods recently proposed for bioactivity prediction from Cell Painting image data, we introduce here a semisupervised contrastive (SemiSupCon) learning approach. This approach combines the strengths of using biological annotations in supervised contrastive learning and leveraging large unannotated image data sets with self-supervised contrastive learning. SemiSupCon enhances downstream prediction performance of classifying MeSH pharmacological classifications from PubChem, as well as mode of action and biological target annotations from the Drug Repurposing Hub across two publicly available Cell Painting data sets. Notably, our approach has effectively predicted the biological activities of several unannotated compounds, and these findings were validated through literature searches. This demonstrates that our approach can potentially expedite the exploration of biological activity based on Cell Painting image data with minimal human intervention.

A genome-wide atlas of human cell morphology

A key challenge of the modern genomics era is developing empirical data-driven representations of gene function. Here we present the first unbiased morphology-based genome-wide perturbation atlas in human cells, containing three genome-wide genotype-phenotype maps comprising CRISPR-Cas9-based knockouts of >20,000 genes in >30 million cells. Our optical pooled cell profiling platform (PERISCOPE) combines a destainable high-dimensional phenotyping panel (based on Cell Painting) with optical sequencing of molecular barcodes and a scalable open-source analysis pipeline to facilitate massively parallel screening of pooled perturbation libraries. This perturbation atlas comprises high-dimensional phenotypic profiles of individual cells with sufficient resolution to cluster thousands of human genes, reconstruct known pathways and protein-protein interaction networks, interrogate subcellular processes and identify culture media-specific responses. Using this atlas, we identify the poorly characterized disease-associated TMEM251/LYSET as a Golgi-resident transmembrane protein essential for mannose-6-phosphate-dependent trafficking of lysosomal enzymes. In sum, this perturbation atlas and screening platform represents a rich and accessible resource for connecting genes to cellular functions at scale.

Cell Painting for cytotoxicity and mode-of-action analysis in primary human hepatocytes

High-throughput, human-relevant approaches for predicting chemical toxicity are urgently needed for better decision-making in human health. Here, we apply image-based profiling (the Cell Painting assay) and two cytotoxicity assays (metabolic and membrane damage readouts) to primary human hepatocytes after exposure to eight concentrations of 1085 compounds that include pharmaceuticals, pesticides, and industrial chemicals with known liver toxicity-related outcomes. Three computational methods (CellProfiler, a Cell Painting-specific convolutional neural network, and a pretrained vision transformer) were compared to extract morphology features from single cells or entire images. We used these morphology features to predict activity in the measured cytotoxicity assays, as well as in 412 curated ToxCast assays that span cytotoxicity, cell-based, and cell-free categories. We found that the morphological profiles detect compound bioactivity at lower concentrations than standard cytotoxicity assays. In supervised analyses, they predict cytotoxicity and targeted cell-based assay readouts, but not cell-free assay readouts. We also found that the various feature extraction methods performed relatively similarly and that filtering out non-bioactive or cytotoxic concentrations did not boost supervised assay prediction performance for any assay endpoint category, although it did have a large influence on unsupervised cluster analysis. We envision that image-based profiling could serve as a key component of modern safety assessment.

Unbiased identification of cell identity in dense mixed neural cultures

Induced pluripotent stem cell (iPSC) technology is revolutionizing cell biology. However, the variability between individual iPSC lines and the lack of efficient technology to comprehensively characterize iPSC-derived cell types hinder its adoption in routine preclinical screening settings. To facilitate the validation of iPSC-derived cell culture composition, we have implemented an imaging assay based on cell painting and convolutional neural networks to recognize cell types in dense and mixed cultures with high fidelity. We have benchmarked our approach using pure and mixed cultures of neuroblastoma and astrocytoma cell lines and attained a classification accuracy above 96%. Through iterative data erosion, we found that inputs containing the nuclear region of interest and its close environment, allow achieving equally high classification accuracy as inputs containing the whole cell for semi-confluent cultures and preserved prediction accuracy even in very dense cultures. We then applied this regionally restricted cell profiling approach to evaluate the differentiation status of iPSC-derived neural cultures, by determining the ratio of postmitotic neurons and neural progenitors. We found that the cell-based prediction significantly outperformed an approach in which the population-level time in culture was used as a classification criterion (96% vs 86%, respectively). In mixed iPSC-derived neuronal cultures, microglia could be unequivocally discriminated from neurons, regardless of their reactivity state, and a tiered strategy allowed for further distinguishing activated from non-activated cell states, albeit with lower accuracy. Thus, morphological single-cell profiling provides a means to quantify cell composition in complex mixed neural cultures and holds promise for use in the quality control of iPSC-derived cell culture models.

Super-resolution microscopy as a drug discovery tool

At the turn of the century a fundamental resolution barrier in fluorescence microscopy known as the diffraction limit was broken, giving rise to the field of super-resolution microscopy. Subsequent nanoscopic investigation with visible light revolutionised our understanding of how previously unknown molecular features give rise to the emergent behaviour of cells. It transpires that the devil is in these fine molecular details, and essential nanoscale processes were found everywhere researchers chose to look. Now, after nearly two decades, super-resolution microscopy has begun to address previously unmet challenges in the study of human disease and is poised to become a pivotal tool in drug discovery.

High-throughput hit identification with acoustic ejection mass spectrometry

This mini-review provides an overview of recent developments in AEMS supporting hit identification in drug discovery, emphasizing its potential to enhance the quality and efficiency of label-free HTS. Future advancements that may further expand the role of AEMS in the drug discovery process will also be discussed.

Self-Driving Microscopes: AI Meets Super-Resolution Microscopy

The integration of Machine Learning (ML) with super-resolution microscopy represents a transformative advancement in biomedical research. Recent advances in ML, particularly deep learning (DL), have significantly enhanced image processing tasks, such as denoising and reconstruction. This review explores the growing potential of automation in super-resolution microscopy, focusing on how DL can enable autonomous imaging tasks. Overcoming the challenges of automation, particularly in adapting to dynamic biological processes and minimizing manual intervention, is crucial for the future of microscopy. Whilst still in its infancy, automation in super-resolution can revolutionize drug discovery and disease phenotyping leading to similar breakthroughs as have been recognized in this year's Nobel Prizes for Physics and Chemistry.

Harnessing the potential of human induced pluripotent stem cells, functional assays and machine learning for neurodevelopmental disorders

Neurodevelopmental disorders (NDDs) affect 4.7% of the global population and are associated with delays in brain development and a spectrum of impairments that can lead to lifelong disability and even mortality. Identification of biomarkers for accurate diagnosis and medications for effective treatment are lacking, in part due to the historical use of preclinical model systems that do not translate well to the clinic for neurological disorders, such as rodents and heterologous cell lines. Human-induced pluripotent stem cells (hiPSCs) are a promising in vitro system for modeling NDDs, providing opportunities to understand mechanisms driving NDDs in human neurons. Functional assays, including patch clamping, multielectrode array, and imaging-based assays, are popular tools employed with hiPSC disease models for disease investigation. Recent progress in machine learning (ML) algorithms also presents unprecedented opportunities to advance the NDD research process. In this review, we compare two-dimensional and three-dimensional hiPSC formats for disease modeling, discuss the applications of functional assays, and offer insights on incorporating ML into hiPSC-based NDD research and drug screening.

Predicting cell morphological responses to perturbations using generative modeling

Advancements in high-throughput screenings enable the exploration of rich phenotypic readouts through high-content microscopy, expediting the development of phenotype-based drug discovery. However, analyzing large and complex high-content imaging screenings remains challenging due to incomplete sampling of perturbations and the presence of technical variations between experiments. To tackle these shortcomings, we present IMage Perturbation Autoencoder (IMPA), a generative style-transfer model predicting morphological changes of perturbations across genetic and chemical interventions. We show that IMPA accurately captures morphological and population-level changes of both seen and unseen perturbations on breast cancer and osteosarcoma cells. Additionally, IMPA accounts for batch effects and can model perturbations across various sources of technical variation, further enhancing its robustness in diverse experimental conditions. With the increasing availability of large-scale high-content imaging screens generated by academic and industrial consortia, we envision that IMPA will facilitate the analysis of microscopy data and enable efficient experimental design via in-silico perturbation prediction.

A combination of high-throughput in vitro and in silico new approach methods for ecotoxicology hazard assessment for fish

Fish acute toxicity testing is used to inform environmental hazard assessment of chemicals. In silico and in vitro approaches have the potential to reduce the number of fish used in testing and increase the efficiency of generating data for assessing ecological hazards. Here, two in vitro bioactivity assays were adapted for use in high-throughput chemical screening. First, a miniaturized version of the Organisation for Economic Co-operation and Development (OECD) test guideline 249 plate reader-based acute toxicity assay in RTgill-W1 cells was developed. Second, the Cell Painting (CP) assay was adapted for use in RTgill-W1 cells along with an imaging-based cell viability assay. Then, 225 chemicals were tested in each assay. Potencies and bioactivity calls from the plate reader and imaging-based cell viability assays were comparable. The CP assay was more sensitive than either cell viability assay in that it detected a larger number of chemicals as bioactive, and phenotype altering concentrations (PACs) were lower than concentrations that decreased cell viability. An in vitro disposition (IVD) model that accounted for sorption of chemicals to plastic and cells over time was applied to predict freely dissolved PACs and compared with in vivo fish toxicity data. Adjustment of PACs using IVD modeling improved concordance of in vitro bioactivity and in vivo toxicity data. For the 65 chemicals where comparison of in vitro and in vivo values was possible, 59% of adjusted in vitro PACs were within one order of magnitude of in vivo toxicity lethal concentrations for 50% of test organisms. In vitro PACs were protective for 73% of chemicals. This combination of in vitro and in silico approaches has the potential to reduce or replace the use of fish for in vivo toxicity testing.

ECBD: European chemical biology database

The European Chemical Biology Database (ECBD, https://ecbd.eu) serves as the central repository for data generated by the EU-OPENSCREEN research infrastructure consortium. It is developed according to FAIR principles, which emphasize findability, accessibility, interoperability and reusability of data. This data is made available to the scientific community following open access principles. The ECBD stores both positive and negative results from the entire chemical biology project pipeline, including data from primary or counter-screening assays. The assays utilize a defined and diverse library of over 107 000 compounds, the annotations of which are continuously enriched by external user supported screening projects and by internal EU-OPENSCREEN bioprofiling efforts. These compounds were screened in 89 currently deposited datasets (assays), with 48 already being publicly accessible, while the remaining will be published after a publication embargo period of up to 3 years. Together these datasets encompass ∼4.3 million experimental data points. All public data within ECBD can be accessed through its user interface, API or by database dump under the CC-BY 4.0 license.

A new class of natural anthelmintics targeting lipid metabolism

Parasitic helminths are a major global health threat, infecting nearly one-fifth of the human population and causing significant losses in livestock and crops. Resistance to the few anthelmintic drugs is increasing. Here, we report a set of avocado fatty alcohols/acetates (AFAs) that exhibit nematocidal activity against four veterinary parasitic nematode species: Brugia pahangi, Teladorsagia circumcincta and Heligmosomoides polygyrus, as well as a multidrug resistant strain (UGA) of Haemonchus contortus. AFA shows significant efficacy in H. polygyrus infected mice. In C. elegans, AFA exposure affects all developmental stages, causing paralysis, impaired mitochondrial respiration, increased reactive oxygen species production and mitochondrial damage. In embryos, AFAs penetrate the eggshell and induce rapid developmental arrest. Genetic and biochemical tests reveal that AFAs inhibit POD-2, encoding an acetyl CoA carboxylase, the rate-limiting enzyme in lipid biosynthesis. These results uncover a new anthelmintic class affecting lipid metabolism.

LIN28-Targeting Chromenopyrazoles and Tetrahydroquinolines Induced Cellular Morphological Changes and Showed High Biosimilarity with BRD PROTACs

The probing of small molecules with heterocyclic scaffolds covering unexplored chemical space and the evaluation of their biological relevance are essential parts of forward chemical genetics approaches and for the development of potential small-molecule therapeutics. In this study, we profiled sets of chromenopyrazoles (CMPs) and tetrahydroquinolines (THQs), originally developed to target the protein-RNA interaction of LIN28-let-7, in a cell painting assay (CPA) measuring cellular morphological changes. Selected LIN28-inactive CMPs and THQs induced cellular morphological changes to different extents. The most CPA-active CMPs 2 and 3 exhibited high bio-similarity with the LCH and BET clusters, while the most CPA-active THQs 13 and 20 indicated a mechanism of action beyond the currently established biosimilarity clusters. Overall, this work demonstrated that CPA is useful in revealing "hidden" biological targets and mechanisms of action for biologically inactive small molecules, which are CMPs and THQs targeting the RNA-binding protein LIN28 in this case, evaluated in target-based strategies. When compared with annotated reference compounds, CMP 3 exhibited a high biosimilarity with the dual BRD7/9 degrading PROTAC VZ185, suggesting that CPA could potentially function as a new phenotypic approach to identify degrader molecules.

The present and future of the Cancer Dependency Map

Despite tremendous progress in the past decade, the complex and heterogeneous nature of cancer complicates efforts to identify new therapies and therapeutic combinations that achieve durable responses in most patients. Further advances in cancer therapy will rely, in part, on the development of targeted therapeutics matched with the genetic and molecular characteristics of cancer. The Cancer Dependency Map (DepMap) is a large-scale data repository and research platform, aiming to systematically reveal the landscape of cancer vulnerabilities in thousands of genetically and molecularly annotated cancer models. DepMap is used routinely by cancer researchers and translational scientists and has facilitated the identification of several novel and selective therapeutic strategies for multiple cancer types that are being tested in the clinic. However, it is also clear that the current version of DepMap is not yet comprehensive. In this Perspective, we review (1) the impact and current uses of DepMap, (2) the opportunities to enhance DepMap to overcome its current limitations, and (3) the ongoing efforts to further improve and expand DepMap.

Imaging phenotype reveals that disulfirams induce protein insolubility in the mitochondrial matrix

The cell painting assay is useful for understanding cellular phenotypic changes and drug effects. To identify other aspects of well-known chemicals, we screened 258 compounds with the cell painting assay and focused on a mitochondrial punctate phenotype seen with disulfiram. To elucidate the reason for this punctate phenotype, we looked for clues by examining staining steps and gene knockdown as well as examining protein solubility and comparing cell lines. From these results, we found that the punctate phenotype was caused by protein insolubility in the mitochondrial matrix. Interestingly, the punctate phenotype of disulfiram was sensitive to the relative expression of LonP1, a protease in the mitochondrial matrix that regulates proteostasis, suggesting that the punctate phenotype manifests when the protein quality control capacity in the mitochondrial matrix is exceeded. Moreover, we discovered that disulfiram and its derivatives, which have all been reported to increase acetaldehyde in the blood after the in vivo intake of alcohol, induced a punctate phenotype as well. The investigated punctate phenotype not only provides a novel clue for elucidating the common mechanism of action among disulfiram derivatives but is also a novel example of chemical perturbation of proteostasis in the mitochondrial matrix.

MorphoDiff: Cellular Morphology Painting with Diffusion Models

Understanding cellular responses to external stimuli is critical for parsing biological mechanisms and advancing therapeutic development. High-content image-based assays provide a cost-effective approach to examine cellular phenotypes induced by diverse interventions, which offers valuable insights into biological processes and cellular states. In this paper, we introduce MorphoDiff, a generative pipeline to predict high-resolution cell morphological responses under different conditions based on perturbation encoding. To the best of our knowledge, MorphoDiff is the first framework capable of producing guided, high-resolution predictions of cell morphology that generalize across both chemical and genetic interventions. The model integrates perturbation embeddings as guiding signals within a 2D latent diffusion model. The comprehensive computational, biological, and visual validations across three open-source Cell Painting datasets show that MorphoDiff can generate high-fidelity images and produce meaningful biology signals under various interventions. We envision the model will facilitate efficient in silico exploration of perturbational landscapes towards more effective drug discovery studies.

AI-based classification of anticancer drugs reveals nucleolar condensation as a predictor of immunogenicity

BACKGROUND

Immunogenic cell death (ICD) inducers are often identified in phenotypic screening campaigns by the release or surface exposure of various danger-associated molecular patterns (DAMPs) from malignant cells. This study aimed to streamline the identification of ICD inducers by leveraging cellular morphological correlates of ICD, specifically the condensation of nucleoli (CON).

METHODS

We applied artificial intelligence (AI)-based imaging analyses to Cell Paint-stained cells exposed to drug libraries, identifying CON as a marker for ICD. CON was characterized using SYTO 14 fluorescent staining and holotomographic microscopy, and visualized by AI-deconvoluted transmitted light microscopy. A neural network-based quantitative structure-activity relationship (QSAR) model was trained to link molecular descriptors of compounds to the CON phenotype, and the classifier was validated using an independent dataset from the NCI-curated mechanistic collection of anticancer agents.

RESULTS

CON strongly correlated with the inhibition of DNA-to-RNA transcription. Cytotoxic drugs that inhibit RNA synthesis without causing DNA damage were as effective as conventional cytotoxicants in inducing ICD, as demonstrated by DAMPs release/exposure and vaccination efficacy in mice. The QSAR classifier successfully predicted drugs with a high likelihood of inducing CON.

CONCLUSIONS

We developed AI-based algorithms for predicting CON-inducing drugs based on molecular descriptors and their validation using automated micrographs analysis, offering a new approach for screening ICD inducers with minimized adverse effects in cancer therapy.

A combined AI and cell biology approach surfaces targets and mechanistically distinct Inflammasome inhibitors

Inflammasomes are protein complexes that mediate innate immune responses whose dysregulation has been linked to a spectrum of acute and chronic human conditions, which dictates therapeutic development that is aligned with disease variability. We designed a scalable, physiologic high-content imaging assay in human PBMCs that we analyzed using a combination of machine-learning and cell biology methods. This resulted in a set of biologically interpretable readouts that can resolve a spectrum of cellular states associated with inflammasome activation and inhibition. These methods were applied to a phenotypic screen that surfaced mechanistically distinct inflammasome inhibitors from an annotated 12,000 compound library. A set of over 100 inhibitors, including an array of Raf-pathway inhibitors, were validated in downstream functional assays. This approach demonstrates how complementary machine learning-based methods can be used to generate profiles of cellular states associated with different stages of complex biological pathways and yield compound and target discovery.

A highly efficient, scalable pipeline for fixed feature extraction from large-scale high-content imaging screens

Applying artificial intelligence (AI) to image-based morphological profiling cells offers significant potential for identifying disease states and drug responses in high-content imaging (HCI) screens. When differences between populations (e.g., healthy vs. diseased) are unknown or imperceptible to the human eye, large-scale HCI screens are essential, providing numerous replicates to build reliable models and accounting for confounding factors like donor and intra-experimental variations. As screen sizes grow, so does the challenge of analyzing high-dimensional datasets in an efficient way while preserving interpretable features and predictive power. Here, we introduce ScaleFEx℠, a memory-efficient, open-source Python pipeline that extracts biologically meaningful features from HCI datasets using minimal computational resources or scalable cloud infrastructure. ScaleFEx can be used together with AI models to successfully identify phenotypic shifts in drug-treated cells and rank interpretable features, and is applicable to public datasets, highlighting its potential to accelerate the discovery of disease-associated phenotypes and new therapeutics.

Molecular and functional profiling of primary normal ovarian cells defines insights into cancer development and drug responses

Patients with ovarian cancer, especially the high-grade serous ovarian cancer (HGSOC) subtype, face poor prognosis due to late diagnosis and treatment resistance. Owing to the high heterogeneity of HGSOC, identifying the origin of the disease and optimal treatments is difficult. Here, we characterized two primary immortalized human ovarian cell lines, human ovarian surface epithelium (HOSE)1C and HOSE2C, comparing their molecular profiling with representative HGSOC cells. We identified molecular features associated with normal and malignant phenotype of ovarian cells by applying single-cell transcriptomics, high-content image-based cell painting, and high-throughput drug testing. Our findings reveal distinct transcriptomic and morphological profiles for the two HOSEs, with a stromal phenotype. Moreover, their responses to the tumor microenvironment differ, exemplified by STAT1 and GREM1 upregulation in HOSE1C and HOSE2C, respectively. We identified selective activation of ERK/MEK targeted inhibitors in cancer cells compared to HOSEs. This study offers insights into the normal and malignant ovarian cells, shedding light on cancer development and drug responses.

Classification and functional characterization of regulators of intracellular STING trafficking identified by genome-wide optical pooled screening

Stimulator of interferon genes (STING) traffics across intracellular compartments to trigger innate responses. Mutations in factors regulating this process lead to inflammatory disorders. To systematically identify factors involved in STING trafficking, we performed a genome-wide optical pooled screen (OPS). Based on the subcellular localization of STING in 45 million cells, we defined 464 clusters of gene perturbations based on their cellular phenotypes. A secondary, higher-dimensional OPS identified 73 finer clusters. We show that the loss of the gene of unknown function C19orf25, which clustered with USE1, a protein involved in Golgi-to-endoplasmic reticulum (ER) transport, enhances STING signaling. Additionally, HOPS deficiency delayed STING degradation and consequently increased signaling. Similarly, GARP/RIC1-RGP1 loss increased STING signaling by delaying STING Golgi exit. Our findings demonstrate that genome-wide genotype-phenotype maps based on high-content cell imaging outperform other screening approaches and provide a community resource for mining factors that impact STING trafficking and other cellular processes.

Semi-quantitative scoring criteria based on multiple staining methods combined with machine learning to evaluate residual nuclei in decellularized matrix

The detection of residual nuclei in decellularized extracellular matrix (dECM) biomaterials is critical for ensuring their quality and biocompatibility. However, current evaluation methods have limitations in addressing impurity interference and providing intelligent analysis. In this study, we utilized four staining techniques-hematoxylin-eosin staining, acetocarmine staining, the Feulgen reaction and 4',6-diamidino-2-phenylindole staining-to detect residual nuclei in dECM biomaterials. Each staining method was quantitatively evaluated across multiple parameters, including area, perimeter and grayscale values, to establish a semi-quantitative scoring system for residual nuclei. These quantitative data were further employed as learning indicators in machine learning models designed to automatically identify residual nuclei. The experimental results demonstrated that no single staining method alone could accurately differentiate between nuclei and impurities. In this study, a semi-quantitative scoring table was developed. With this table, the accuracy of determining whether a single suspicious point is a cell nucleus has reached over 98%. By combining four staining methods, false positives caused by impurity contamination were eliminated. The automatic recognition model trained based on nuclear parameter features reached the optimal index of the model after several iterations of training in 172 epochs. The trained artificial intelligence model achieved a recognition accuracy of over 90% for detecting residual nuclei. The use of multidimensional parameters, integrated with machine learning, significantly improved the accuracy of identifying nuclear residues in dECM slices. This approach provides a more reliable and objective method for evaluating dECM biomaterials, while also increasing detection efficiency.

A versatile information retrieval framework for evaluating profile strength and similarity

In profiling assays, thousands of biological properties are measured in a single test, yielding biological discoveries by capturing the state of a cell population, often at the single-cell level. However, for profiling datasets, it has been challenging to evaluate the phenotypic activity of a sample and the phenotypic consistency among samples, due to profiles' high dimensionality, heterogeneous nature, and non-linear properties. Existing methods leave researchers uncertain where to draw boundaries between meaningful biological response and technical noise. Here, we developed a statistical framework that uses the well-established mean average precision (mAP) as a single, data-driven metric to bridge this gap. We validated the mAP framework against established metrics through simulations and real-world data applications, revealing its ability to capture subtle and meaningful biological differences in cell state. Specifically, we used mAP to assess both phenotypic activity for a given perturbation (or a sample) as well as consistency within groups of perturbations (or samples) across diverse high-dimensional datasets. We evaluated the framework on different profile types (image, protein, and mRNA profiles), perturbation types (CRISPR gene editing, gene overexpression, and small molecules), and profile resolutions (single-cell and bulk). Our open-source software allows this framework to be applied to identify interesting biological phenomena and promising therapeutics from large-scale profiling data.

Decoding phenotypic screening: A comparative analysis of image representations

Biomedical imaging techniques such as high content screening (HCS) are valuable for drug discovery, but high costs limit their use to pharmaceutical companies. To address this issue, The JUMP-CP consortium released a massive open image dataset of chemical and genetic perturbations, providing a valuable resource for deep learning research. In this work, we aim to utilize the JUMP-CP dataset to develop a universal representation model for HCS data, mainly data generated using U2OS cells and CellPainting protocol, using supervised and self-supervised learning approaches. We propose an evaluation protocol that assesses their performance on mode of action and property prediction tasks using a popular phenotypic screening dataset. Results show that the self-supervised approach that uses data from multiple consortium partners provides representation that is more robust to batch effects whilst simultaneously achieving performance on par with standard approaches. Together with other conclusions, it provides recommendations on the training strategy of a representation model for HCS images.

ToxProfiler: A novel human-based reporter assay for in vitro chemical safety assessment

In vitro chemical safety assessment often relies on simple and general cytotoxicity endpoint measurements and fails to adequately predict human toxicity. To improve the in vitro chemical safety assessment, it is important to understand the underlying mechanisms of toxicity. Here we introduce ToxProfiler, a novel human-based reporter assay that quantifies the chemical-induced stress responses at a single-cell level and reveals the toxicological mode-of-action (MoA) of novel drugs and chemicals. The assay accurately measures the activation of seven major cellular stress response pathways (oxidative stress, cell cycle stress, endoplasmic reticulum stress, ion stress, protein stress, autophagy and inflammation) that play a role in the adaptive responses prior to cellular toxicity. To assess the applicability of the assay in predicting the toxicity MoA of chemicals, we tested a set of 100 chemicals with well-known in vitro and in vivo toxicological profiles. Concentration response modeling and point-of-departure estimation for each reporter protein allowed for chemical potency ranking and revealed the primary toxicological MoA of chemicals. Furthermore, the assay could effectively group chemicals based on their shared toxicity signatures and link them to specific toxicological targets, e.g. mitochondrial toxicity and genotoxicity, and different human pathologies, including liver toxicity and cardiotoxicity. Overall, ToxProfiler is a quantitative in vitro reporter assay that can accurately provide insight into the toxicological MoA of compounds, thereby assisting in the future mechanism-based safety assessment of chemicals.

Applying cell painting in non-tumorigenic breast cells to understand impacts of common chemical exposures

The general population is exposed to many chemicals which have putative, but incompletely understood, links to breast cancer. Cell Painting is a high-content imaging-based in vitro assay that allows for unbiased measurements of concentration-dependent effects of chemical exposures on cellular morphology. We used Cell Painting to measure effects of 16 human exposure relevant chemicals, along with 21 small molecules with known mechanisms of action, in non-tumorigenic mammary epithelial cells, the MCF10A cell line. Using CellProfiler image analysis software, we quantified 3042 morphological features across approximately 1.2 million cells. We used benchmark concentration modeling to identify features both conserved and different across chemicals. Benchmark concentrations were compared to exposure biomarker concentration measurements from the National Health and Nutrition Examination Survey to assess which chemicals induce morphological alterations at human-relevant concentrations. We found significant feature overlaps between chemicals, including similarities between the organochlorine pesticide DDT metabolite p,p'-DDE and an activator of Wnt signaling CHIR99201. We validated these findings by assaying the activation of Wnt, as reflected by translocation of ꞵ-catenin, following p'-p' DDE exposure. Consistent with Wnt signaling activation, low concentration p',p'-DDE (25 nM) significantly enhanced the nuclear translocation of ꞵ-catenin. Overall, these findings highlight the ability of Cell Painting to enhance mode-of-action studies for toxicants which are common in our environment but incompletely characterized with respect to breast cancer risk.

Towards generative digital twins in biomedical research

Digital twins in biomedical research, i.e. virtual replicas of biological entities such as cells, organs, or entire organisms, hold great potential to advance personalized healthcare. As all biological processes happen in space, there is a growing interest in modeling biological entities within their native context. Leveraging generative artificial intelligence (AI) and high-volume biomedical data profiled with spatial technologies, researchers can recreate spatially-resolved digital representations of a physical entity with high fidelity. In application to biomedical fields such as computational pathology, oncology, and cardiology, these generative digital twins (GDT) thus enable compelling in silico modeling for simulated interventions, facilitating the exploration of 'what if' causal scenarios for clinical diagnostics and treatments tailored to individual patients. Here, we outline recent advancements in this novel field and discuss the challenges and future research directions.

SPACe: an open-source, single-cell analysis of Cell Painting data

Phenotypic profiling by high throughput microscopy, including Cell Painting, has become a leading tool for screening large sets of perturbations in cellular models. To efficiently analyze this big data, available open-source software requires computational resources usually not available to most laboratories. In addition, the cell-to-cell variation of responses within a population, while collected and analyzed, is usually averaged and unused. We introduce SPACe (Swift Phenotypic Analysis of Cells), an open-source platform for analysis of single-cell image-based morphological profiles produced by Cell Painting. We highlight several advantages of SPACe, including processing speed, accuracy in mechanism of action recognition, reproducibility across biological replicates, applicability to multiple models, sensitivity to variable cell-to-cell responses, and biological interpretability to explain image-based features. We illustrate SPACe in a defined screening campaign of cell metabolism small-molecule inhibitors tested in seven cell lines to highlight the importance of analyzing perturbations across models.

Cell Painting and Chemical Structure Read-Across Can Complement Each Other for Rat Acute Oral Toxicity Prediction in Chemical Early Derisking

Early derisking decisions in the development of new chemical compounds enable the identification of novel chemical candidates with improved safety profiles. In vivo studies are traditionally conducted in the early assessment of acute oral toxicity of crop protection products to avoid compounds, which are considered "very acutely toxic", with an in vivo lethal dose of 50% (LD50) ≤ 60 mg/kg body weight. Those studies are lengthy and costly and raise ethical concerns, catalyzing the use of nonanimal alternatives. The objective of our analysis was to assess the predictive efficacy of read-across approaches for acute oral toxicity in rats, comparing the use of chemical structure information, in vitro biological data derived from the Cell Painting profiling assay on U2OS cells, or the combination of both. Our findings indicate that the classification of compounds as very acute oral toxic (LD50 ≤ 60 mg/kg) or not is possible using a read-across approach, with chemical structure information, morphological profiles, or a combination of both. When classifying compounds structurally similar to those in the training set, the chemical structure was more predictive (balanced accuracy of 0.82). Conversely, when the compounds to be classified were structurally different from those in the training set, the morphological profiles were more predictive (balanced accuracy of 0.72). Combining the two models allowed for the classification of compounds structurally similar to those in the training set to slightly improve the predictions (balanced accuracy of 0.85).

Combining NeuroPainting with transcriptomics reveals cell-type-specific morphological and molecular signatures of the 22q11.2 deletion

Neuropsychiatric conditions pose substantial challenges for therapeutic development due to their complex and poorly understood underlying mechanisms. High-throughput, unbiased phenotypic assays present a promising path for advancing therapeutic discovery, especially within disease-relevant neural tissues. Here, we introduce NeuroPainting, a novel adaptation of the Cell Painting assay, optimized for high-dimensional morphological phenotyping of neural cell types, including neurons, neuronal progenitor cells, and astrocytes derived from human stem cells. Using NeuroPainting, we quantified cell structure and organelle behavior across various brain cell types, creating a public dataset of over 4,000 cellular traits. This extensive dataset not only sets a new benchmark for phenotypic screening in neuropsychiatric research but also serves as a gold standard for the research community, enabling comparisons and validation of results. We then applied NeuroPainting to identify morphological signatures associated with the 22q11.2 deletion, a major genetic risk factor for schizophrenia. We observed profound cell-type-specific effects of the 22q11.2 deletion, with significant alterations in mitochondrial structure, endoplasmic reticulum organization, and cytoskeletal dynamics, particularly in astrocytes. Transcriptomic analysis revealed reduced expression of cell adhesion genes in 22q11.2 deletion astrocytes, consistent with recent post-mortem findings. Integrating the RNA sequencing data and morphological profiles uncovered a novel biological link between altered expression of specific cell adhesion molecules and observed changes in mitochondrial morphology in 22q11.2 deletion astrocytes. These findings underscore the power of combined phenomic and transcriptomic analyses to reveal mechanistic insights associated with human genetic variants of neuropsychiatric conditions.

Insights into the Identification of iPSC- and Monocyte-Derived Macrophage-Polarizing Compounds by AI-Fueled Cell Painting Analysis Tools

Macrophage polarization critically contributes to a multitude of human pathologies. Hence, modulating macrophage polarization is a promising approach with enormous therapeutic potential. Macrophages are characterized by a remarkable functional and phenotypic plasticity, with pro-inflammatory (M1) and anti-inflammatory (M2) states at the extremes of a multidimensional polarization spectrum. Cell morphology is a major indicator for macrophage activation, describing M1(-like) (rounded) and M2(-like) (elongated) states by different cell shapes. Here, we introduced cell painting of macrophages to better reflect their multifaceted plasticity and associated phenotypes beyond the rigid dichotomous M1/M2 classification. Using high-content imaging, we established deep learning- and feature-based cell painting image analysis tools to elucidate cellular fingerprints that inform about subtle phenotypes of human blood monocyte-derived and iPSC-derived macrophages that are characterized as screening surrogate. Moreover, we show that cell painting feature profiling is suitable for identifying inter-donor variance to describe the relevance of the morphology feature 'cell roundness' and dissect distinct macrophage polarization signatures after stimulation with known biological or small-molecule modulators of macrophage (re-)polarization. Our novel established AI-fueled cell painting analysis tools provide a resource for high-content-based drug screening and candidate profiling, which set the stage for identifying novel modulators for macrophage (re-)polarization in health and disease.

Advanced single-cell and spatial analysis with high-multiplex characterization of circulating tumor cells and tumor tissue in prostate cancer: Unveiling resistance mechanisms with the CoDuCo in situ assay

BACKGROUND

Metastatic prostate cancer is a highly heterogeneous and dynamic disease and practicable tools for patient stratification and resistance monitoring are urgently needed. Liquid biopsy analysis of circulating tumor cells (CTCs) and circulating tumor DNA are promising, however, comprehensive testing is essential due to diverse mechanisms of resistance. Previously, we demonstrated the utility of mRNA-based in situ padlock probe hybridization for characterizing CTCs.

METHODS

We have developed a novel combinatorial dual-color (CoDuCo) assay for in situ mRNA detection, with enhanced multiplexing capacity, enabling the simultaneous analysis of up to 15 distinct markers. This approach was applied to CTCs, corresponding tumor tissue, cancer cell lines, and peripheral blood mononuclear cells for single-cell and spatial gene expression analysis. Using supervised machine learning, we trained a random forest classifier to identify CTCs. Image analysis and visualization of results was performed using open-source Python libraries, CellProfiler, and TissUUmaps.

RESULTS

Our study presents data from multiple prostate cancer patients, demonstrating the CoDuCo assay's ability to visualize diverse resistance mechanisms, such as neuroendocrine differentiation markers (SYP, CHGA, NCAM1) and AR-V7 expression. In addition, druggable targets and predictive markers (PSMA, DLL3, SLFN11) were detected in CTCs and formalin-fixed, paraffin-embedded tissue. The machine learning-based CTC classification achieved high performance, with a recall of 0.76 and a specificity of 0.99.

CONCLUSIONS

The combination of high multiplex capacity and microscopy-based single-cell analysis is a unique and powerful feature of the CoDuCo in situ assay. This synergy enables the simultaneous identification and characterization of CTCs with epithelial, epithelial-mesenchymal, and neuroendocrine phenotypes, the detection of CTC clusters, the visualization of CTC heterogeneity, as well as the spatial investigation of tumor tissue. This assay holds significant potential as a tool for monitoring dynamic molecular changes associated with drug response and resistance in prostate cancer.

Capturing cell heterogeneity in representations of cell populations for image-based profiling using contrastive learning

Image-based cell profiling is a powerful tool that compares perturbed cell populations by measuring thousands of single-cell features and summarizing them into profiles. Typically a sample is represented by averaging across cells, but this fails to capture the heterogeneity within cell populations. We introduce CytoSummaryNet: a Deep Sets-based approach that improves mechanism of action prediction by 30-68% in mean average precision compared to average profiling on a public dataset. CytoSummaryNet uses self-supervised contrastive learning in a multiple-instance learning framework, providing an easier-to-apply method for aggregating single-cell feature data than previously published strategies. Interpretability analysis suggests that the model achieves this improvement by downweighting small mitotic cells or those with debris and prioritizing large uncrowded cells. The approach requires only perturbation labels for training, which are readily available in all cell profiling datasets. CytoSummaryNet offers a straightforward post-processing step for single-cell profiles that can significantly boost retrieval performance on image-based profiling datasets.