SLAS Discovery

CD28 is a pivotal costimulatory receptor involved in T cell activation and immune regulation, positioning it as a key therapeutic target for inflammatory diseases, including inflammatory bowel disease (IBD). Despite its potential, small molecules targeting CD28 are still limited. To fill this gap, we developed a high-throughput screening (HTS) platform based on Temperature-Related Intensity Change (TRIC) technology, enabling rapid, immobilization-free screening of chemical libraries of small molecules. Using the Dianthus instrument, we applied our optimized TRIC assay for CD28 (signal-to-noise ratio of 21.99) to screen two MedChemExpress libraries: Small Molecule Immuno-Oncology Compounds (SMIOC) and Protein-Protein Interaction Inhibitors (PPII), identifying 50 initial hits. Following exclusion of compounds with dye interference or aggregation artifacts, 12 candidates were prioritized for further validation. Microscale thermophoresis (MST) confirmed dose-dependent binding of seven compounds to CD28, with affinities in the micromolar range. Surface plasmon resonance (SPR) further validated two compounds, EABP 02303 and CTEP, as CD28 binders. These results demonstrate that our TRIC-based HTS platform is robust, scalable, and effective for identifying small molecule CD28 binders. The incorporation of orthogonal validation supports the reliability of our findings and highlights the feasibility of small-molecule discovery targeting CD28.

Phenotypic profiling by high throughput microscopy has become one of the leading tools for screening large sets of perturbations in cellular models. Of the numerous methods used over the years, the flexible and economical Cell Painting (CP) assay has been central in the field, allowing for large screening campaigns leading to a vast number of data-rich images. Currently, to analyze data of this scale, available open-source software (i.e., CellProfiler) requires computational resources that are not available to most laboratories worldwide. In addition, the image-embedded cell-to-cell variation of responses within a population, while collected and analyzed, is usually averaged and unused. Here we introduce SPACe (Swift Phenotypic Analysis of Cells), an open source, Python-based platform for the analysis of single cell image-based morphological profiles produced by CP experiments. SPACe can process a typical dataset approximately ten times faster than CellProfiler on common desktop computers without loss in mechanism of action (MOA) recognition accuracy. It also computes directional distribution-based distances (Earth Movers Distance - EMD) of morphological features for quality control and hit calling. We highlight several advantages of SPACe analysis on CP assays, including reproducibility across multiple biological replicates, easy applicability to multiple ([~]20) cell lines, sensitivity to variable cell-to-cell responses, and biological interpretability to explain image-based features. We ultimately illustrate the advantages of SPACe in a screening campaign of cell metabolism small molecule inhibitors which we performed in seven cell lines to highlight the importance of testing perturbations across models.

Efficiently identifying bioactive compounds towards a target of interest remains a time- and resource-intensive task in early drug discovery. The ability to accurately predict bioactivity using morphological profiles has the potential to rationalize the process, enabling smaller screens of focused compound sets. Towards this goal, we explored the application of deep learning with Cell Painting, a high-content image-based assay, for compound bioactivity prediction in early drug screening. Combining Cell Painting data and unrefined single-concentration activity readouts from high-throughput screening (HTS) assays, we investigated to what degree morphological profiles could predict compound activity across a set of 140 unique assays. We evaluated the performance of our models across different target classes, assay technologies, and disease areas. The predictive performance of the models was high, with a tendency for better predictions on cell-based assays and kinase targets. The average ROC-AUC was 0.744 with 62% of assays reaching [≥]0.7, 30% reaching [≥]0.8 and 7% reaching [≥]0.9 average ROC-AUC, outperforming commonly used structure-based predictions in terms of predictive performance and compound structure diversity. In many cases, bioactivity prediction from Cell Painting data could be matched using brightfield images rather than multichannel fluorescence images. Experimental validation of our predictions in follow-up assays confirmed enrichment of active compounds. Our results suggest that models trained on Cell Painting data can predict compound activity in a range of high-throughput screening assays robustly, even with relatively noisy HTS assay data. With our approach, enriched screening sets with higher hit rates and higher hit diversity can be selected, which could reduce the size of HTS campaigns and enable primary screening with more complex assays.

The high morbidity rate of ovarian cancer has remained unchanged during the past four decades, partly due to lack of understanding of disease mechanisms and difficulties in developing new targeted therapies. Defective DNA damage detection and repair is one of the hallmarks of cancer cells and is a defining characteristic of ovarian cancer. Most in vitro studies to date, involve viability measurements at scale using relevant cancer cell lines, however, the translation to clinic is often lacking. The use of patient derived organoids is closing that translational gap yet the 3D nature of organoid cultures present challenges for assay measurements beyond viability measurements. In particular, high-content imaging has the potential for screening at scale providing a better understanding of mechanism of action of drugs or genetic perturbagens. In this study we report a semi-automated and scalable immunofluorescence imaging assay utilising the development of a 384-well plate based subnuclear staining and clearing protocol and optimisation of 3D confocal image analysis for studying DNA damage dose response in human ovarian cancer organoids. The assay was validated in four organoid models and demonstrated a predictable response to Etoposide drug treatment with lowest efficacy observed in the clinically most resistant model. This imaging and analysis method can be applied to other 3D organoid and spheroid models for use in high content screening.

Rett syndrome (OMIM 312750) is a rare neurodevelopmental disorder caused by de novo mutations in the Methyl-CpG Binding Protein 2 (MeCP2) gene located on the X-Chromosome, typically affecting girls. Currently, available therapy for Rett Syndrome is only symptomatic. Rett syndrome symptoms first appear between 6 to 18 months of age, characterized by microcephaly and lack of motor coordination being the most prevalent. The disease continues to progress until adulthood when it reaches a stationary phase. More than 800 different mutations causing Rett syndrome have been described, yet the most common is T158M (9% prevalence), located in the Methyl-Binding domain (MBD) of MeCP2. Due to its importance for DNA binding through recognition of methylated CpG, mutations in the MBD have a significant impact on the stability and function of MeCP2. MeCP2 is a nuclear protein and accumulates in liquid-liquid phase condensates visualized as speckles in NIH3T3 by microscopy. We developed a high content phenotypic assay, detecting fluorescent MeCP2 speckles in NIH3T3 cells. The assay allows to identify small molecules that stabilize MeCP2-T158M and phenotypically rescue speckle formation. To validate the assay, a collection of 3572 drugs was screened, including FDA-approved drugs, compounds in clinical trials and biologically annotated tool compounds. 18 hits were identified showing at least 25% of rescue of speckles in the mutant cell line while not affecting wild-type MeCP2 speckles. Primary hits were confirmed in a dose response assay and in a thermal shift assay with recombinant MeCP2. One class of identified hits represents histone deacetylase inhibitors (HDACis) showing 25% speckle rescue of mutant MeCP2 without toxicity. This screening strategycan be expanded to additional compound libraries and support novel drug discovery.

Despite recent advances in melanoma drug discovery, the average overall survival of patients with late stage metastatic melanoma is approximately 3 years, suggesting a need for new approaches and melanoma therapeutic targets. Previously we identified heterogeneous nuclear ribonucleoprotein H2 as a potential target of anti-melanoma compound 2155-14 (Palrasu et al, Cell Physiol Biochem 2019;53:656-86). In the present study, we endeavored to develop an assay to enable a high throughput screening campaign to identify drug-like molecules acting via down regulation of heterogeneous nuclear ribonucleoprotein H that can be used for melanoma therapy and research. ResultsWe established a cell-based platform using metastatic melanoma cell line WM266-4 expressing hnRNPH2 conjugated with green fluorescent protein to enable assay development and screening. High Content Screening assay was developed and validated in 384 well plate format, followed by miniaturization to 1,536 well plate format. All plate-based QC parameters were acceptable: %CV = 6.7{+/-}0.3, S/B = 21{+/-}2.1, Z = 0.75{+/-}0.04. Pilot screen of FDA-approved drug library (n=1,400 compounds) demonstrated hit rate of 0.5%. Two compounds demonstrated pharmacological response and were authenticated by western blot analysis. ConclusionsWe developed a highly robust HTS-amenable high content screening assay capable of monitoring down regulation of hnRNPH2. This assay is thus capable of identifying authentic down regulators of hnRNPH1 and 2 in a large compound collection and, therefore, is amenable to a large-scale screening effort.

Proximity dependent labeling using engineered enzymes has been used extensively to identify protein-protein interactions, and map protein complexes in-vitro and in-vivo. Here, we extend the use of engineered promiscuous biotin ligases to the identification of small molecule protein targets. Chimeric bi-functional chemical probes ("recruiters") are used to effectively recruit tagged biotin ligases for proximity dependent labeling of target and target interactors. The broad applicability of this approach is demonstrated with probes developed from a multi-kinase inhibitor, a bromodomain targeting moiety, and an FKBP targeting molecule. While complementary to traditional chemo-proteomic strategies such as photo-affinity labeling (PAL), and activity-based protein profiling (ABPP), this approach is a useful addition to the target ID toolbox with opportunities for tunability based on the inherent labeling efficiencies of different engineered enzymes and control over the enzyme cellular localization.

Viral exoribonucleases are uncommon in the world of RNA viruses. To date, this activity has been identified only in the Arenaviridae and the Coronaviridae families. These exoribonucleases play important but different roles in both families: for mammarenaviruses the exoribonuclease is involved in the suppression of the host immune response whereas for coronaviruses, exoribonuclease is both involved in a proofreading mechanism ensuring the genetic stability of viral genomes and participating to evasion of the host innate immunity. Because of their key roles, they constitute attractive targets for drug development. Here we present a high-throughput assay using fluorescence polarization to assess the viral exoribonuclease activity and its inhibition. We validate the assay using three different viral enzymes from SARS-CoV-2, lymphocytic choriomeningitis and Machupo viruses. The method is sensitive, robust, amenable to miniaturization (384 well plates) and allowed us to validate the proof-of-concept of the assay by screening a small focused compounds library (23 metal chelators). We also determined the IC50 of one inhibitor common to the three viruses. HighlightsO_LIArenaviridae and Coronaviridae viral families share an exoribonuclease activity of common evolutionary origin C_LIO_LIArenaviridae and Coronaviridae exoribonuclease is an attractive target for drug development C_LIO_LIWe present a high-throughput assay in 384 well-plates for the screening of inhibitors using fluorescence polarization C_LIO_LIWe validated the assay by screening of a focused library of 23 metal chelators against SARS-CoV-2, Lymphocytic Choriomeningitis virus and Machupo virus exoribonucleases C_LIO_LIWe determined the IC50 by fluorescence polarization of one inhibitor common to the three viruses. C_LI

Morphological profiling with the Cell Painting assay has emerged as a promising method in drug discovery research. The assay captures morphological changes across various cellular compartments enabling the rapid identification of the effect of compounds. We present a comprehensive morphological profiling dataset using the carefully curated and well-annotated EU-OPENSCREEN Bioactive Compound Set. Our profiling dataset was generated across multiple imaging sites with high-throughput confocal microscopes using the Hep G2 as well as the U2 OS cell line. We employed an extensive assay optimization process to achieve high data quality across the different imaging sites. An analysis of the four replicates validates the robustness of the generated data. We compare morphological features of the different cell lines and map the profiles to activity, toxicity, and basic compound targets to further describe the dataset as well as to demonstrate the potential of this dataset to be used for mechanism of action exploration.

SLIT2, a secreted glycoprotein involved in axon guidance, immune modulation, and tumor progression, remains largely unexplored as a pharmacological target due to the absence of small-molecule modulators. Here, we present a proof-of-concept high-throughput screening platform that integrates Temperature-Related Intensity Change (TRIC) technology with time-resolved Forster resonance energy transfer (TR-FRET) to identify small molecules capable of disrupting the SLIT2/ROBO1 interaction. Screening a lipid metabolism-focused compound library (653 molecules) yielded bexarotene, as the most potent small molecule SLIT2 binder reported to date, with a dissociation constant (KD) of 2.62 {micro}M. Follow-up TR-FRET assays demonstrated dose-dependent inhibition of SLIT2/ROBO1 interaction, with an IC50 value of [~]22.8 {micro}M and maximal inhibition of [~]15- 25%. These findings suggest a novel extracellular activity of bexarotene and validate the combined use of TRIC and TR-FRET as a scalable screening strategy for SLIT2-targeted small molecules. This platform lays the groundwork for future high-throughput discovery efforts against SLIT2 and its signaling axis. Graphical abstractTRIC-based small molecule screening platform protocol steps with implementation of TR-FRET for the identification of SLIT2 inhibitors. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=122 SRC="FIGDIR/small/663693v1_ufig1.gif" ALT="Figure 1"> View larger version (26K): org.highwire.dtl.DTLVardef@e585fborg.highwire.dtl.DTLVardef@bf3de5org.highwire.dtl.DTLVardef@178910dorg.highwire.dtl.DTLVardef@7b2086_HPS_FORMAT_FIGEXP M_FIG C_FIG

KRAS is an important oncogenic driver which is mutated in numerous cancers. Recent advances in the selective targeting of KRAS mutants via small molecule inhibitors and targeted protein degraders have generated an increase in research activity in this area in recent years. As such, there is a need for new assay platforms to profile next generation inhibitors which improve on the potency and selectivity of existing drug candidates, while evading the emergence of resistance. Here, we describe the development of a new panel of biochemical and cell-based assays to evaluate the binding and function of known chemical entities targeting mutant KRAS. Our assay panels generated selectivity profiles and quantitative binding interaction dissociation constants for small molecules and degraders against wild type, G12C, G12D, and G12V KRAS, which were congruent with published data. These assays can be leveraged for additional mutants of interest beyond those described in this study, using both overexpressed cell-free systems and cell-based systems with endogenous protein levels. TABLE OF CONTENTS/ABSTRACT GRAPHIC O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=111 SRC="FIGDIR/small/604418v1_ufig1.gif" ALT="Figure 1"> View larger version (27K): org.highwire.dtl.DTLVardef@b5bb4aorg.highwire.dtl.DTLVardef@11b1c91org.highwire.dtl.DTLVardef@f07d63org.highwire.dtl.DTLVardef@b81c33_HPS_FORMAT_FIGEXP M_FIG C_FIG

Phenotypic profiling methods, such as Cell Painting and gene expression, have been widely used to predict compound bioactivity, often showing improvement over predictive models based on chemical structures alone. We discovered that a large subset of assays in widely-used benchmark datasets either directly relate to cell health and cytotoxicity or are assays intending to capture a more specific phenotype but whose active compounds impact cell count, while inactives do not. As a result, counting cells can achieve similar predictive performance as Cell Painting or gene expression data. Filtering benchmarks to include only assays relating to protein targets reveals that Cell Painting can capture information that cannot be predicted by mere cell counting. We re-evaluated three benchmark datasets used with Cell Painting data and observed that, in many cases, cell count models produced an AUC comparable to models using the full Cell Painting profiles. However, in protein-target-specific benchmarks across 17 distinct protein targets, Cell Painting features demonstrated unique predictive power, outperforming mean balanced accuracy from cell count models with a relative improvement of 19.6%. We propose five practical recommendations for benchmarking machine learning models for predicting bioactivity, including using cell count as a baseline feature. Although multi-class classification applications (such as matching samples based on their morphological profile) are less likely to be predictable by cell count than bioactivity benchmarks, these recommendations are broadly applicable to machine learning for drug discovery.

Protein modification by interferon-stimulated gene 15 (ISG15), termed ISGylation, exhibits antiviral properties and influences tumorigenesis, genome stability and metabolic processes. ISGylation is counteracted by the specific protease USP18. Likewise, viral proteases such as the papain-like protease (PLpro) from SARS-CoV-2 cleave ISG15 to undermine the host immune response. Beyond its role as a deISGylating enzyme, USP18 acts as a major negative regulator of the IFN signaling pathway in a STAT2-dependent manner. In humans, unconjugated ISG15 secures USP18 stability and the absence of USP18 or impaired STAT2/USP18 binding cause fatal interferonopathies. Thus, the USP18 signaling hub represents a critical checkpoint for type I IFN signaling and ISGylation, qualifying it as a promising immune and cancer drug target. However, suitable assays to monitor protein-protein interactions (PPIs) within the USP18/ISG15/STAT2 signaling hub and to screen for PPI modulators are missing and no specific inhibitors targeting USP18 interactions are available. To address this gap, we developed a method based on the NanoLuc luciferase (NLuc) Bioluminescence Energy Transfer (NanoBRET) assay system to study PPIs. Firstly, we generated stable cell lines suitable to monitor USP18/ISG15 and USP18/STAT2 interactions, providing a semi high-throughput screening (HTS)-compatible platform. In combination with a virtual pre-screen of 60,000 compounds against USP18 in silico, this assay allowed us to identify a first small molecule (ZHAWOC8655) that compromises cellular USP18/ISG15 binding and inhibits USP18 protease activity in vitro. To further explore the potential of using the NanoBRET system for testing PPI modulators, we evaluated the effect of GRL0617, a compound which was shown to disrupt the interaction between SARS-CoV-2 PLpro/ISG15 as well as SARS-CoV-2 PLpro/ubiquitin. NanoBRET based stable cell lines as presented here will be suitable for monitoring PPIs in other multiprotein complexes after various stimuli, mutations or small molecule administration and can be challenged with siRNA or CRISPR/Cas9 libraries to identify previously unrecognized regulators.

DNA-encoded library screening represents a significant advancement in the field of drug discovery. Its ability to rapidly and cost-effectively identify potential drug candidates from large compound libraries has the potential to revolutionize the way new medicines are discovered and developed. While the strategies for DEL screening and data analysis have improved over the years, data normalization remains an open challenge. Existing normalization methods can yield poor correlation for compounds with high read count, and they do not account for inherent sources of noise. To overcome these drawbacks, we have developed a robust normalization technique using an antibody fragment and a DNA-conjugated peptide as an internal control. This innovative approach allows for normalization between samples of different conditions and accounts for technical challenges that occur during screening. Table of Contents Graphic O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=109 SRC="FIGDIR/small/700605v1_ufig1.gif" ALT="Figure 1"> View larger version (23K): org.highwire.dtl.DTLVardef@1b04b91org.highwire.dtl.DTLVardef@1312295org.highwire.dtl.DTLVardef@d59713org.highwire.dtl.DTLVardef@b1786a_HPS_FORMAT_FIGEXP M_FIG C_FIG SynopsisNormalization of DNA-encoded library selection data reduces bias and noise, enabling accurate identification of true binders and reliable enrichment analysis.

In the recent issue of Nature Communications (2023 Nov 30;14(1):7908), Xue et. al. reported a very interesting and significant discovery of a possible DCAF11 ligand chemotype that could be used as the "warhead" to design bifunctional compounds for targeted degradation via engaging the E3 ligase DCAF11 1 (annotated as ref 1 hereafter). The discovery is of importance to the targeted protein degradation field and was inspired by previous reports suggesting that similar compounds may also engage the autophagosome protein LC3 for degradation and function as autophagy-tethering compounds (ATTECs) 2, 3, 4, 5, 6, 7, which seem to be inconsistent with ref 1. We think that the conclusions based on these data are not necessarily mutually exclusive. After performing additional experiments and analyses, we would like to discuss some possibilities explaining such discrepancies and make a few points of clarification.

Robust, generalizable approaches to identify compounds efficiently with undesirable mechanisms of action in complex cellular assays remain elusive. Such a process would be useful for hit triage during high-throughput screening and, ultimately, predictive toxicology during drug development. We generated cell painting and cellular health profiles for 218 prototypical cytotoxic and nuisance compounds in U-2 OS cells in a concentration-response format. A diversity of compounds causing cellular damage produced bioactive cell painting morphologies, including cytoskeletal poisons, genotoxins, nonspecific electrophiles, and redox-active compounds. Further, we show that lower quality lysine acetyltransferase inhibitors and nonspecific electrophiles can be distinguished from more selective counterparts. We propose that the purposeful inclusion of cytotoxic and nuisance reference compounds such as those profiled in this Resource will help with assay optimization and compound prioritization in complex cellular assays like cell painting.

Enzyme-free isothermal amplification methods offer a promising alternative to enzymatic assays for nucleic acid detection, particularly in low-resource settings. The nonenzymatic isothermal strand displacement and amplification (NISDA) assay was recently introduced as a highly sensitive, enzyme-free detection strategy. Here, we attempted to replicate its reported performance. Despite extensive testing, we failed to replicate the reported sensitivity and could only produce detectable signals at extremely high target concentrations ([≥]1 x 1011 copies/{micro}L). Our results highlight the critical importance of independent validation in the development of nonenzymatic diagnostic assays.

Non-structural protein 2 (nsP2), which plays an essential role in replication of CHIKV, contains a protease, helicase, and methyltransferase-like domain. We executed a simple a screen using malachite green to detect compounds that decreased ATP hydrolysis and tested a library of diverse compounds to find inhibitors of CHIKV nsP2 helicase.

DNA-encoded library (DEL) technology has become a powerful tool in modern drug discovery. Fully harnessing its potential requires the use of extensive computational methodologies, which are often available only through proprietary software. This restricts accessibility for small teams lacking robust informatics support, hindering the growth of the technology. Here, we present DELi, an open-source DEL informatics platform designed for library design, NGS decoding and calling, and enrichment analysis. DELi supports a simple and easy to understand configuration setup to present a straightforward user interface. To showcase its capabilities, we used DELi to design an in-house custom, benzimidazole-based DEL (UNC DEL006), and performed proof-of-concept selection experiments against Bromodomain-containing Protein 4 (BRD4). The DELi decoding and analysis modules identified top-performing compounds, leading to the off-DNA synthesis of UNC11951, which was confirmed as a nanomolar BRD4 binder via isothermal titration calorimetry (ITC) and differential scanning fluorimetry (DSF). These results demonstrate DELi as an effective tool for DEL design and analysis. Furthermore, its open-source nature will promote ongoing development and contributions from the DEL community to expand its applications and capabilities, making DEL technology more widely accessible. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=94 SRC="FIGDIR/small/640184v3_ufig1.gif" ALT="Figure 1"> View larger version (25K): org.highwire.dtl.DTLVardef@1d4ca88org.highwire.dtl.DTLVardef@13cb4feorg.highwire.dtl.DTLVardef@8ea7ecorg.highwire.dtl.DTLVardef@1b294fc_HPS_FORMAT_FIGEXP M_FIG C_FIG

Determining the mechanism of action (MOA) for natural products remains a significant bottleneck in drug discovery, particularly for researchers with limited computational resources or small compound libraries. Traditional approaches require screening large numbers of annotated compounds alongside unknowns, which is cost-prohibitive, or depend on complex machine learning models that need substantial computational resources and large datasets. Here, we present a dissertation chapter excerpt: MOAST (Mechanism of Action Similarity Tool), a BLAST-inspired computational workflow that addresses these limitations by providing rapid MOA hypotheses for newly screened compounds. This chapter investigates two complementary approaches: a kernel density estimation (KDE) method providing statistical significance measures and E-values for MOA class membership, and a CatBoost machine learning classifier for multi-class prediction with ranked outputs. Using cytological profiling data from HeLa and A549 cell lines, MOAST achieved 22% accuracy for the top 5 predictions among [~] 300 MOA classes, with the CatBoost classifier reaching 10% balanced accuracy--significantly better than the [~] 3% reported in literature. The tool suggests a 0.8 prediction probability threshold for trustworthy results and demonstrates robust performance across multiple feature reduction strategies. MOAST provides a practical, accessible solution that bridges traditional phenotypic screening and modern computational approaches, making MOA determination feasible for researchers with limited resources while maintaining statistical rigor and interpretability.