Explore the vast range of titles available.

Screening drugs on patient biopsies from solid tumours has immense potential, but is challenging due to the small amount of available material. To address this, we present here a plug-based microfluidics platform for functional screening of drug combinations. Integrated Braille valves allow changing the plug composition on demand and enable collecting >1200 data points (56 different conditions with at least 20 replicates each) per biopsy. After deriving and validating efficient and specific drug combinations for two genetically different pancreatic cancer cell lines and xenograft mouse models, we additionally screen live cells from human solid tumours with no need for ex vivo culturing steps, and obtain highly specific sensitivity profiles. The entire workflow can be completed within 48 h at assay costs of less than US$ 150 per patient. We believe this can pave the way for rapid determination of optimal personalized cancer therapies. Cancer patients exhibit specific sensitivities toward drug combinations that cannot be easily predicted. Here the authors setup a microfluidic platform that allows testing of multiple drug combinations correctly predicting sensitivity in vivo and they use it on patients biopsies to define effective drugs.

High throughput screening (HTS) is routinely used to identify bioactive small molecules. This requires physical compounds, which limits coverage of accessible chemical space. Computational approaches combined with vast on-demand chemical libraries can access far greater chemical space, provided that the predictive accuracy is sufficient to identify useful molecules. Through the largest and most diverse virtual HTS campaign reported to date, comprising 318 individual projects, we demonstrate that our AtomNet® convolutional neural network successfully finds novel hits across every major therapeutic area and protein class. We address historical limitations of computational screening by demonstrating success for target proteins without known binders, high-quality X-ray crystal structures, or manual cherry-picking of compounds. We show that the molecules selected by the AtomNet® model are novel drug-like scaffolds rather than minor modifications to known bioactive compounds. Our empirical results suggest that computational methods can substantially replace HTS as the first step of small-molecule drug discovery.

Key Points Drug discovery approaches for cancer, as for other therapeutic areas, have typically been divided into two classes: target-based drug discovery (TDD) and phenotypic drug discovery (PDD). Cancer drug discovery poses substantial challenges for both targeted and 'classical' phenotypic drug discovery owing to the number, diversity and plasticity of molecular mechanisms and phenotypes underlying tumour initiation and growth. Discovery origins for all 48 small-molecule cancer drugs approved by the US Food and Drug Administration between 1999 and 2013, and for agents in Phase II and II clinical trials at the end of 2013, were analysed and classified. Although a significant number of approved and investigational cancer drugs could be easily classified as targeted, the majority of which (21 out of 29) are kinase inhibitors, we concluded that very few drugs (four out of 48) were discovered entirely by 'classical' PDD. The remainder were discovered by, or developed from chemical lead matter discovered by a combination of phenotypic and target-based assays. Drug discovery using cytoxicity assays and cancer cell lines, although yielding many of the current standard-of-care chemotherapies, is unlikely to result in further drugs with novel mechanisms of action. Knowledge of the molecular pathways and targets required for specific disease-associated phenotypes, along with the ability to use more disease-relevant cell models, improves the probability of discovering drugs with novel mechanisms of action and clinical efficacy in molecularly defined patient populations. We introduce the concept of 'mechanism-informed phenotypic drug discovery' (MIPDD) to include phenotypic assays for specific molecular pathways and targets. Determining the causal relationships between target inhibition and phenotypic effects may well open up new and unexpected avenues of cancer biology. Such an approach presents the best means of discovering drugs that have both an optimal molecular mechanism of action and a diagnostic hypothesis to enable patient selection leading to clinical responses. There has been a resurgence of interest in the use of phenotypic screens in drug discovery as an alternative to target-focused approaches. Moffat and colleagues investigated the contribution of phenotypic assays in oncology by analysing the origins of the new small-molecule cancer drugs approved by the US Food and Drug Administration over the past 15 years. They also discuss technical and biological advances that could empower phenotypic drug discovery in oncology by enabling the development of mechanistically informed phenotypic screens. There has been a resurgence of interest in the use of phenotypic screens in drug discovery as an alternative to target-focused approaches. Given that oncology is currently the most active therapeutic area, and also one in which target-focused approaches have been particularly prominent in the past two decades, we investigated the contribution of phenotypic assays to oncology drug discovery by analysing the origins of all new small-molecule cancer drugs approved by the US Food and Drug Administration (FDA) over the past 15 years and those currently in clinical development. Although the majority of these drugs originated from target-based discovery, we identified a significant number whose discovery depended on phenotypic screening approaches. We postulate that the contribution of phenotypic screening to cancer drug discovery has been hampered by a reliance on 'classical' nonspecific drug effects such as cytotoxicity and mitotic arrest, exacerbated by a paucity of mechanistically defined cellular models for therapeutically translatable cancer phenotypes. However, technical and biological advances that enable such mechanistically informed phenotypic models have the potential to empower phenotypic drug discovery in oncology.

Cancer drug development has been riddled with high attrition rates, in part, due to poor reproducibility of preclinical models for drug discovery. Poor experimental design and lack of scientific transparency may cause experimental biases that in turn affect data quality, robustness and reproducibility. Here, we pinpoint sources of experimental variability in conventional 2D cell-based cancer drug screens to determine the effect of confounders on cell viability for MCF7 and HCC38 breast cancer cell lines treated with platinum agents (cisplatin and carboplatin) and a proteasome inhibitor (bortezomib). Variance component analysis demonstrated that variations in cell viability were primarily associated with the choice of pharmaceutical drug and cell line, and less likely to be due to the type of growth medium or assay incubation time. Furthermore, careful consideration should be given to different methods of storing diluted pharmaceutical drugs and use of DMSO controls due to the potential risk of evaporation and the subsequent effect on dose-response curves. Optimization of experimental parameters not only improved data quality substantially but also resulted in reproducible results for bortezomib- and cisplatin-treated HCC38, MCF7, MCF-10A, and MDA-MB-436 cells. Taken together, these findings indicate that replicability (the same analyst re-performs the same experiment multiple times) and reproducibility (different analysts perform the same experiment using different experimental conditions) for cell-based drug screens can be improved by identifying potential confounders and subsequent optimization of experimental parameters for each cell line.

Three-dimensional (3D) cell culture technologies, such as organoids, are physiologically relevant models for basic and clinical applications. Automated microfluidics offers advantages in high-throughput and precision analysis of cells but is not yet compatible with organoids. Here, we present an automated, high-throughput, microfluidic 3D organoid culture and analysis system to facilitate preclinical research and personalized therapies. Our system provides combinatorial and dynamic drug treatments to hundreds of cultures and enables real-time analysis of organoids. We validate our system by performing individual, combinatorial, and sequential drug screens on human-derived pancreatic tumor organoids. We observe significant differences in the response of individual patient-based organoids to drug treatments and find that temporally-modified drug treatments can be more effective than constant-dose monotherapy or combination therapy in vitro. This integrated platform advances organoids models to screen and mirror real patient treatment courses with potential to facilitate treatment decisions for personalized therapy. The use of organoids in personalized medicine is promising but high throughput platforms are needed. Here the authors develop an automated, high-throughput, microfluidic 3D organoid culture system that allows combinatorial and dynamic drug treatments and real-time analysis of organoids.

Transfer learning, which transfers patterns learned on a source dataset to a related target dataset for constructing prediction models, has been shown effective in many applications. In this paper, we investigate whether transfer learning can be used to improve the performance of anti-cancer drug response prediction models. Previous transfer learning studies for drug response prediction focused on building models to predict the response of tumor cells to a specific drug treatment. We target the more challenging task of building general prediction models that can make predictions for both new tumor cells and new drugs. Uniquely, we investigate the power of transfer learning for three drug response prediction applications including drug repurposing, precision oncology, and new drug development, through different data partition schemes in cross-validation. We extend the classic transfer learning framework through ensemble and demonstrate its general utility with three representative prediction algorithms including a gradient boosting model and two deep neural networks. The ensemble transfer learning framework is tested on benchmark in vitro drug screening datasets. The results demonstrate that our framework broadly improves the prediction performance in all three drug response prediction applications with all three prediction algorithms.

The challenges associated with traditional drug screening, such as high costs and long screening times, have led to an increase in the use of single-cell isolation technologies. Small sample volumes are required for high-throughput, cell-based assays to reduce assay costs and enable rapid sample processing. Using microfluidic chips, single-cell analysis can be conducted more effectively, requiring fewer reagents and maintaining biocompatibility. Due to the chip's ability to manipulate small volumes of fluid, high-throughput screening assays can be developed that are both miniaturized and automated. In the present study, we employ microfluidic chips for drug screening in chronic myeloid leukemia. This study aimed to establish a robust methodology integrating diverse assays, providing a holistic understanding of drug response. Herein, we have used a chronic myeloid leukemia derived cell line (K562) for drug screening with an innovative microfluidic-based drug screening approach to investigate the efficacy of imatinib in K562 cells. Cell viability was assessed using MTT assay. Apoptosis was measured using Annexin/PI staining by flow cytometry. Significant increased apoptosis was seen in K562 cells treated with imatinib in the microfluidic device compared to cells treated with imatinib in 24- and 96-well plates. Moreover, in the microfluidic chip, drug screening time was reduced from 48 hours to 24 hours. Compared to traditional approaches, microfluidic-based drug screening efficiently evaluates the efficacy of imatinib in K562 cells. This approach is promising for drug discovery and treatment optimization, as it increases sensitivity and streamlines the screening process.

While the potential of patient-derived organoids (PDOs) to predict patients’ responses to anti-cancer treatments has been well recognized, the lengthy time and the low efficiency in establishing PDOs hamper the implementation of PDO-based drug sensitivity tests in clinics. We first adapt a mechanical sample processing method to generate lung cancer organoids (LCOs) from surgically resected and biopsy tumor tissues. The LCOs recapitulate the histological and genetic features of the parental tumors and have the potential to expand indefinitely. By employing an integrated superhydrophobic microwell array chip (InSMAR-chip), we demonstrate hundreds of LCOs, a number that can be generated from most of the samples at passage 0, are sufficient to produce clinically meaningful drug responses within a week. The results prove our one-week drug tests are in good agreement with patient-derived xenografts, genetic mutations of tumors, and clinical outcomes. The LCO model coupled with the microwell device provides a technically feasible means for predicting patient-specific drug responses in clinical settings. The lengthy time in establishing patient-derived organoids(PDOs) hampers the implementation of PDO-based drug sensitivity tests in clinics. Here, the authors show a microwell array-based method to predict patient’s responses to anti-cancer therapies in a week using lung cancer organoids.

Cancer drug screening in patient‐derived cells holds great promise for personalized oncology and drug discovery but lacks standardization. Whether cells are cultured as conventional monolayer or advanced, matrix‐dependent organoid cultures influences drug effects and thereby drug selection and clinical success. To precisely compare drug profiles in differently cultured primary cells, we developed DeathPro , an automated microscopy‐based assay to resolve drug‐induced cell death and proliferation inhibition. Using DeathPro , we screened cells from ovarian cancer patients in monolayer or organoid culture with clinically relevant drugs. Drug‐induced growth arrest and efficacy of cytostatic drugs differed between the two culture systems. Interestingly, drug effects in organoids were more diverse and had lower therapeutic potential. Genomic analysis revealed novel links between drug sensitivity and DNA repair deficiency in organoids that were undetectable in monolayers. Thus, our results highlight the dependency of cytostatic drugs and pharmacogenomic associations on culture systems, and guide culture selection for drug tests. Synopsis DeathPro , an automated microscopy‐based assay resolves cell death and proliferation inhibition in 2D and 3D cultures. Drug screens using DeathPro provide insights into the impact of culture systems on drug effects and their links to genomic features. DeathPro resolves cytotoxic and cytostatic effects in drug screens with patient‐derived ovarian and lung cancer cells, organoids and co‐cultures with fibroblasts. Drug responses in cancer organoids are more diverse than in 2D cultured cells. Cytostatic drugs depend on culture systems, cytotoxic effects are independent of the culture format. Genomic analysis of cancer patients links DNA repair deficiency to drug sensitivity in organoids. Graphical Abstract DeathPro , an automated microscopy‐based assay resolves cell death and proliferation inhibition in 2D and 3D cultures. Drug screens using DeathPro provide insights into the impact of culture systems on drug effects and their links to genomic features.

Epithelial-mesenchymal transition (EMT) is a crucial pathological event in cancer, particularly in tumor cell budding and metastasis. Therefore, control of EMT can represent a novel therapeutic strategy in cancer. Here, we introduce an innovative three-dimensional (3D) high-throughput screening (HTS) system that leads to an identification of EMT inhibitors. For the establishment of the novel 3D-HTS system, we chose NanoCulture Plates (NCP) that provided a gel-free micro-patterned scaffold for cells and were independent of other spheroid formation systems using soft-agar. In the NCP-based 3D cell culture system, A549 lung cancer cells migrated, gathered, and then formed multiple spheroids within 7 days. Live cell imaging experiments showed that an established EMT-inducer TGF-β promoted peripheral cells around the core of spheroids to acquire mesenchymal spindle shapes, loss of intercellular adhesion, and migration from the spheroids. Along with such morphological change, EMT-related gene expression signatures were altered, particularly alteration of mRNA levels of ECAD/CDH1, NCAD/CDH2, VIM and ZEB1/TCF8. These EMT-related phenotypic changes were blocked by SB431542, a TGF-βreceptor I (TGFβR1) inhibitor. Inside of the spheroids were highly hypoxic; in contrast, spheroid-derived peripheral migrating cells were normoxic, revealed by visualization and quantification using Hypoxia Probe. Thus, TGF-β-triggered EMT caused spheroid hypoplasia and loss of hypoxia. Spheroid EMT inhibitory (SEMTIN) activity of SB431542 was calculated from fluorescence intensities of the Hypoxia Probe, and then was utilized in a drug screening of EMT-inhibitory small molecule compounds. In a pilot screening, 9 of 1,330 compounds were above the thresholds of the SEMTIN activity and cell viability. Finally, two compounds SB-525334 and SU9516 showed SEMTIN activities in a dose dependent manner. SB-525334 was a known TGFβR1 inhibitor. SU9516 was a cyclin-dependent kinase 2 (CDK2) inhibitor, which we showed also had an EMT-inhibitory activity. The half maximal inhibitory concentration (IC50) of SB-525334 and SU9516 were 0.31 μM and 1.21 μM, respectively, while IC50 of SB431542 was 2.38 μM. Taken together, it was shown that this 3D NCP-based HTS system was useful for screening of EMT-regulatory drugs.