Cellular Readouts Research Articles

Training deep learning models on personalized genomic sequences improves variant effect prediction.

Sequence-to-function models have broad applications in interpreting the molecular impact of genetic variation, yet have been criticized for poor performance in this task. Here we show that training models on functional genomic data with matched personal genomes improves their performance at variant effect prediction. Variant effect representations are retained even when transfer learning models to unseen cellular contexts and experimental readouts. Our results have implications for interpreting trait-associated genetic variation.

Simple methodology to score micropattern quality and effectiveness.

Micropatterns (MPs) are widely used as a powerful tool to control cell morphology and phenotype. However, methods for determining the effectiveness of how well cells are controlled by the shape of MPs have been inconsistently used and studies rarely report on this topic, indicating lack of standardization. We introduce an evaluation score that quantitatively assesses the MP fabrication quality and effectiveness, which can be broadly used in conjunction with all currently available MP design types. This score uses four simple and quick steps: (i) scoring MP and (ii) background fabrication quality, (iii) defining the type(s) of MP of interest, and (iv) assigning so-called efficiency descriptors describing cell behavior. These steps are based on visual inspection and quick categorization of various aspects of MP fabrication quality and cell behaviour, presented in illustrations and microscopy image examples intended to serve as a reference "atlas". To illustrate the advantage of using this score, we determined differences in cell morphology and F-actin intensity between scored vs. non-scored cells. These measurements, which could be different in other studies, were chosen because both are understood as markers of cell phenotype and function. We combined intensity-calibrated immunofluorescence microscopy and image-based single cell protein analysis. Importantly, significant differences in cell morphology and cytoskeletal protein content between scored vs. non-scored cells were noted: the unconditional inclusion of all experimental read-outs (i.e., all MP data regardless of MP quality and effectiveness) into the final results significantly misjudged the experimental readouts vs. only including experimental read-outs of quality-controlled and effective MPs, identified by scoring. Specifically, non-scoring underestimated the F-actin intensity per cell and quantitative cellular morphometric descriptors circularity and solidity and overestimated aspect ratio. Scoring improved the precision of cellular readouts, advocating the use of a MP quality and efficiency score as a quantitative decision-supporting tool in deciding whether or not particular MPs should be used for experiments, saving time and money. This simple scoring methodology can be used for improving MP fabrication, comparing results across studies, benefiting basic science studies and potential future clinical use of MPs by introducing standardization.

Photodynamic Therapy of Head and Neck Cancers with Ruthenium-based Photosensitizers

SignificanceThe management of patients with head and neck cancers relies mainly on platinum-based chemotherapies. However, severe side effects and tumor resistance impair treatment efficacy, notably due to frequent mutations of the TP53 tumor suppressor. Here we propose a ruthenium-based photodynamic therapy as a p53-independent therapeutic strategy, and show that its phototoxicity relies on the induction of endoplasmic reticulum stress and immunogenic cell death. ApproachSQ20B cells were treated with the Ru I ruthenium-based photosensitizer prior to illumination. Ru I cytotoxicity and mode of action were studied using molecular and cellular read-outs. ResultsRu I showed an interesting phototoxicity and triggered the expression of key markers of endoplasmic reticulum stress and early stages of immunogenic cell death. ConclusionsRu I-PDT appears to be efficient in vitro on head and neck cancer models. Its phototoxicity seems to be associated to two p53-independent pathways, whose functional links remain to be clarified.

FRESH™ 3D bioprinted cardiac tissue, a bioengineered platform for in vitro pharmacology.

There is critical need for a predictive model of human cardiac physiology in drug development to assess compound effects on human tissues. In vitro two-dimensional monolayer cultures of cardiomyocytes provide biochemical and cellular readouts, and in vivo animal models provide information on systemic cardiovascular response. However, there remains a significant gap in these models due to their incomplete recapitulation of adult human cardiovascular physiology. Recent efforts in developing in vitro models from engineered heart tissues have demonstrated potential for bridging this gap using human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) in three-dimensional tissue structure. Here, we advance this paradigm by implementing FRESH™ 3D bioprinting to build human cardiac tissues in a medium throughput, well-plate format with controlled tissue architecture, tailored cellular composition, and native-like physiological function, specifically in its drug response. We combined hiPSC-CMs, endothelial cells, and fibroblasts in a cellular bioink and FRESH™ 3D bioprinted this mixture in the format of a thin tissue strip stabilized on a tissue fixture. We show that cardiac tissues could be fabricated directly in a 24-well plate format were composed of dense and highly aligned hiPSC-CMs at >600 million cells/mL and, within 14 days, demonstrated reproducible calcium transients and a fast conduction velocity of ∼16 cm/s. Interrogation of these cardiac tissues with the β-adrenergic receptor agonist isoproterenol showed responses consistent with positive chronotropy and inotropy. Treatment with calcium channel blocker verapamil demonstrated responses expected of hiPSC-CM derived cardiac tissues. These results confirm that FRESH™ 3D bioprinted cardiac tissues represent an in vitro platform that provides data on human physiological response.

Combined Single Cell Flow Cytometry and Imaging Analyses Reveal Immunomodulatory Effects Exerted By Targeted Phospho-SYK Inhibitors with Elevated Sensitivity in NPM1 Mutated AML

Acute myeloid leukemia (AML) is driven by genetic lesions that can result in addiction to signaling pathways that lead to aberrant proliferation and survival. Several new targeted drugs are being assessed for activity in AML with some recently approved for the treatment of subgroups of patients with specific genetic features, such as mutations in FLT3, IDH1 or IDH2. Entospletinib (ENTO) and lanraplenib (LANRA) are inhibitors of spleen tyrosine kinase (SYK); the latter is a next-generation SYK inhibitor and is currently being evaluated in combination with gilteritinib (GILT) in patients with relapsed or refractory FLT3-mutated AML (NCT05028751). Although SYK is rarely mutated, it has been found to cooperate with FLT3, and is highly active especially in AML patients with mutated FLT3. In addition, SYK is important for immune cell signaling through antibody constant region Fc and B cell receptors. To examine the multi-modulatory effect of pSYK inhibitors in AML, we used combined high parameter flow cytometry (HP-FC) and single cell miniaturized imaging (sc-MI) approaches to profile the immunologic, mutation, and functional protein pathway targets of ENTO and LANRA. Two different AML patient cohorts were assessed by ex vivo treatment with ENTO, LANRA and combinations of LANRA (trametinib, gilteritinib (GILT), aPD-1) over a range of time points (2h, 4h, 3d and 9d) and readouts for cellular phenotype (e.g., differentiation state, immune cell populations) and functional readouts (pSYK, viability) for each cohort were analyzed by HP-FC and sc-MI and associated with available mutation information (bulk gene and single cell RNA sequencing). Differential analyses of data from flow cytometry (AML-HP-FC Cohort) and single cell imaging (AML-sc-MI Cohort) demonstrated: 1) sensitivity to ENTO and LANRA (³ 1uM) was higher in NPM1 and FLT3 mutated AML samples, as measured via cell viability and pSYK target inhibition; sensitivity was not drastically altered in the presence of other mutations (e.g. IDH1/2, MLL); 2) lower doses of ENTO and LANRA (<1uM) induced proliferation in specific cell types, including progenitor myeloid cells, and T and B lymphocytes, but induced more T cell proliferation in AML samples that have low baseline proliferation; analysis with next generation bulk gene and single cell RNA sequencing suggested downregulation of STAT1expression, downregulation of MYC- and mTOR-dependent pathways, and upregulation of IL-1 pathway genes; and 3) LANRA was more broadly effective at targeting pSYK inhibition than ENTO; however, both were effective in combinations (ENTO with trametinib and GILT, and LANRA with aPD-1). In sum, analysis using collective flow and single cell imaging provides corroboratory and complementary information revealing new roles for ENTO and LANRA as immune modulating agents supporting B and T cell proliferation at lower dosages, and with demonstrated greater effect in individuals with NPM1 and FLT3 mutations. These support studies investigating the use of pSYK inhibitors in combination with other therapies to reduce tumor burden, and as an improved line of treatment for AML. Acknowledgements: We are appreciative of the patients and families for their time and participation.

Disruption of tonic endocannabinoid signalling triggers cellular, behavioural and neuroendocrine responses consistent with a stress response.

Endocannabinoid (eCB) signalling gates many aspects of the stress response, including the hypothalamic-pituitary-adrenal (HPA) axis. The HPA axis is controlled by corticotropin releasing hormone (CRH) producing neurons in the paraventricular nucleus of the hypothalamus (PVN). Disruption of eCB signalling increases drive to the HPA axis, but the mechanisms subserving this process are poorly understood. Using an array of cellular, endocrine and behavioural readouts associated with activation of CRH neurons in the PVN, we evaluated the contributions of tonic eCB signalling to the generation of a stress response. The CB1 receptor antagonist/inverse agonist AM251, neutral antagonist NESS243 and NAPE PLD inhibitor LEI401 all uniformly increased Fos in the PVN, unmasked stress-linked behaviours, such as grooming, and increased circulating CORT, recapitulating the effects of stress. Similar effects were also seen after direct administration of AM251 into the PVN, while optogenetic inhibition of PVN CRH neurons ameliorated stress-like behavioural changes produced by disruption of eCB signalling. These data indicate that under resting conditions, constitutive eCB signalling restricts activation of the HPA axis through local regulation of CRH neurons in the PVN.

Elucidating the Immunogenic Mechanisms of Merck’s Lipid Nanoparticle (LNP) Adjuvants

Abstract Lipid nanoparticles (LNPs) have been shown to be a safe and potent platform for vaccine delivery. When added to subunit proteins, they have the potential to induce strong immune responses capable of generating long-lasting immunity. Merck has developed a variety of LNPs utilizing different cationic lipid components including LNP A, which has a long half-life in vivo, and LNP B, which has a shorter half-life in vivo and is more biodegradable. Protein subunit vaccines containing LNP A have been shown to generate robust antigen-specific humoral and cellular responses with some dose dependent reactogenicity while vaccines formulated with LNP B elicit lower immunogenicity and increased tolerability. To gain insight into the innate signaling pathways utilized by these LNPs, we performed an immunogenicity study in MyD88 knockout mice utilizing RSV F subunit protein vaccine plus LNP A or B. Wild type mice immunized with LNP A vaccines produced significantly higher levels of pro-inflammatory and neutrophil recruitment serum cytokines than the MyD88 knockout mice. However, with LNP B, MyD88 deficiency had no effect on cytokine expression. Humoral and cellular immune readouts are ongoing. Next, we stimulated whole human blood with the LNPs to compare at cytokine production. Stimulation with LNP A produced cytokine significant upregulation of monocyte and neutrophil-specific cytokines at 24 hours while stimulation with LNP B had no effect on cytokine production. Finally, we are evaluating differential gene expression to gain insight into the cellular responses of leukocytes to stimulation with the LNPs. Together, these experiments will help mechanistically determine how these fundamentally different LNPs can activate immune responses.

Selection of Optimal Cell Lines for High-Content Phenotypic Screening

High-content microscopy offers a scalable approach to screen against multiple targets in a single pass. Prior work has focused on methods to select "optimal" cellular readouts in microscopy screens. However, methods to select optimal cell line models have garnered much less attention. Here, we provide a roadmap for how to select the cell line or lines that are best suited to identify bioactive compounds and their mechanism of action (MOA). We test our approach on compounds targeting cancer-relevant pathways, ranking cell lines in two tasks: detecting compound activity ("phenoactivity") and grouping compounds with similar MOA by similar phenotype ("phenosimilarity"). Evaluating six cell lines across 3214 well-annotated compounds, we show that optimal cell line selection depends on both the task of interest (e.g., detecting phenoactivity vs inferring phenosimilarity) and distribution of MOAs within the compound library. Given a task of interest and a set of compounds, we provide a systematic framework for choosing optimal cell line(s). Our framework can be used to reduce the number of cell lines required to identify hits within a compound library and help accelerate the pace of early drug discovery.

The role of EphA2 in ADAM17- and ionizing radiation-enhanced lung cancer cell migration.

Ionizing radiation (IR) enhances the migratory capacity of cancer cells. Here we investigate in non-small-cell-lung-cancer (NSCLC) cells a novel link between IR-enhanced ADAM17 activity and the non-canonical pathway of EphA2 in the cellular stress response to irradiation. Cancer cell migration in dependence of IR, EphA2, and paracrine signaling mediated by ADAM17 was determined using transwell migration assays. Changes of EphA2 pS897 and mRNA expression levels upon different ADAM17-directed treatment strategies, including the small molecular inhibitor TMI-005, the monoclonal antibody MEDI3622, and shRNAs, were mechanistically investigated. ADAM17-mediated release and cleavage of the EphA2 ligand ephrin-A1 was measured using ELISA and an acellular cleavage assay. Irradiation with 5 Gy enhanced tumor cell migration of NSCLC NCI-H358 cells in dependence of EphA2. At the same time, IR increased growth factor-induced EphA2 S897 phosphorylation via auto- and paracrine signaling. Genetic and pharmaceutical downregulation of ADAM17 activity abrogated growth factor (e.g. amphiregulin) release, which reduced MAPK pathway-mediated EphA2 S897 phosphorylation in an auto- and paracrine way (non-canonical EphA2-pathway) in NCI-H358 and A549 cells. These signaling processes were associated with reduced cell migration towards conditioned media derived from ADAM17-deficient cells. Interestingly, ADAM17 inhibition with the small molecular inhibitor TMI-005 led to the internalization and proteasomal degradation of EphA2, which was rescued by amphiregulin or MG-132 treatment. In addition, ADAM17 inhibition also abrogated ephrin-A1 cleavage and thereby interfered with the canonical EphA2-pathway. We identified ADAM17 and the receptor tyrosine kinase EphA2 as two important drivers for (IR-) induced NSCLC cell migration and described a unique interrelation between ADAM17 and EphA2. We demonstrated that ADAM17 influences both, EphA2 (pS897) and its GPI-anchored ligand ephrin-A1. Using different cellular and molecular readouts, we generated a comprehensive picture of how ADAM17 and IR influence the EphA2 canonical and non-canonical pathway in NSCLC cells.

Predicting progression to Alzheimer's disease with human hippocampal progenitors exposed to serum.

Adult hippocampal neurogenesis is important for learning and memory and is altered early in Alzheimer's disease. As hippocampal neurogenesis is modulated by the circulatory systemic environment, evaluating a proxy of how hippocampal neurogenesis is affected by the systemic milieu could serve as an early biomarker for Alzheimer's disease progression. Here, we used an in vitro assay to model the impact of systemic environment on hippocampal neurogenesis. A human hippocampal progenitor cell line was treated with longitudinal serum samples from individuals with mild cognitive impairment, who either progressed to Alzheimer's disease or remained cognitively stable. Mild cognitive impairment to Alzheimer's disease progression was characterized most prominently with decreased proliferation, increased cell death and increased neurogenesis. A subset of 'baseline' cellular readouts together with education level were able to predict Alzheimer's disease progression. The assay could provide a powerful platform for early prognosis, monitoring disease progression and further mechanistic studies.

Trial in Progress: A Phase I/II Study to Assess Safety, Tolerability and Preliminary Efficacy of Bexmarilimab in Combination with Standard of Care in Patients with Hematological Malignancies (BEXMAB)

Background: Emerging data on the role of leukemic cells on shaping the tumor immune environment places macrophages as target cells and immune checkpoints in acute myeloid leukemia (AML) (Miari et. al. Front. Cell Dev. Biol.2021; Sallman et. al. Blood 2019). CLEVER-1 (common lymphatic endothelial and vascular endothelial receptor-1) constitutes such macrophage checkpoint. The receptor contributes to tumor immune tolerance by antigen clearing and T cell inhibition. In multiple models, tumor growth is significantly reduced in CLEVER-1 absence (Viitala et. al. Clin Cancer Res 2019). CLEVER-1 expression level is a prognostic factor for poor outcomes in AML (Pölönen et. al. Cancer Research 2019) and its suppression inhibits leukemia cell growth. Analyses suggests lower expression is associated with venetoclax sensitivity (Lin et al. Mol Ther 2019). Bexmarilimab (FP-1305; BEX), an IgG4 monoclonal antibody, binds CLEVER-1 and alters the function of macrophages and inactive T cells. BEX was shown to increase antigen presentation, induce proinflammatory cytokines and infiltration of activated T cells (Viitala et. al. Clin Cancer Res 2019). The therapeutic potential of BEX is supported with data from the MATINS study (NCT03733990) evaluating single agent in advanced solid tumors. Up to 40% of patients in selected indications experience disease control (stable disease/partial response) associated with a 65% chance of 1-year overall survival (OS) compared to 11% of patients without disease control. BEX treatment is well tolerated without any dose limiting toxicity (DLT) reported. CLEVER-1 is highly expression on malignant blasts and monocytes of bone marrow from AML and myelodysplastic syndrome (MDS) patients. Treatment of these BM cells with BEX alone or in combination with azacitidine or venetoclax results in enhanced antigen presentation capacity and increased activation markers on effector T cells with synergistic effects (EHA 2022 P380). Study Design and Methods: The PhI/II study investigates BEX treatment plus standard of care in MDS, chronic myelomonocytic leukemia (CMML) frontline or hypomethylating agent (HMA)-failed patients or newly diagnosed AML or relapsed/refractory (r/r) AML patients (NCT05428969). Enrollment is ongoing. Ph I is a dose escalation using a Bayesian optimal interval (BOIN) design to explore 4 dose levels of BEX in a cohort size of 5 with 1.0 mg/kg as starting dose. BEX plus azacitidine (doublet) is assessed in a mixed cohort of MDS, CMML frontline and hypomethylating agent (HMA)-failure patients or r/r AML patients. Dose determination may then be performed in separate indications. BEX plus azacitidine + venetocolax (triplet) will follow a similar BOIN design for dose determination in patients with newly diagnosed AML unfit for induction chemotherapy. BEX is dosed in 28-day cycles Q1W during cycles 1-3 followed by Q2W dosing, until disease progression or intolerable toxicity. Standard of care therapy is administered as per label. During Ph II, selected indications are explored for efficacy when BEX is administered at recommended phase 2 dose (RP2D) determined in the doublet and triplet. Dose expansion follows a 2-Stage design with indication-specific target efficacy rates, including complete remission +/- incomplete hematologic recovery, overall response rate or minimal residual disease (MRD)-negativity. Primary endpoints are safety and tolerability, RP2D of BEX and preliminary efficacy. Secondary endpoints include extended efficacy for duration of response, event free survival and OS. Exploratory endpoints are focused on pharmacodynamic parameters for immunological and cellular read-outs and the association of predictive biomarker of BEX clinical activity in different patient sub-groups. Key eligibility includes patient ≥ 18 years of age presenting with morphologically confirmed diagnosis of either MDS (intermediate-very high rIPSS risk), CMML-2 with indication for azacitidine treatment; CMML and MDS patients with failure to HMA therapy, confirmed diagnosis of r/r AML following at least 1 prior treatment line or diagnosis of AML in patients unfit for induction therapy with indication for azacitidine + venetoclax. Patients with APL or myeloproliferative CMML, active auto-immune disorder or requiring systemic corticosteroid or other immunosuppressive treatment are excluded. Approximately 181 patients at 10 US and EU sites will be enrolled.

Abstract LB034: Evaluation of a novel single-cell RNA sequencing methodology for use in clinical trial settings

Abstract Bulk RNA sequencing (RNA-seq) is regularly employed for transcriptional profiling but provides only a snapshot of the average population. Particularly when studying disease, understanding differences in diverse cells is essential thus single-cell resolution is a necessity due to significant cellular heterogeneity. High-throughput droplet- and microfluidic-based approaches for performing single-cell RNA-seq (scRNA-seq) are invaluable techniques for uncovering changes in cells such as differential gene expression and epigenetic states in cancer research. A challenge for scRNA-seq, however, is sample preparation as typically tissues must be processed immediately after collection. We and others have reported alternative approaches to stabilize samples prior to analysis such as cryopreservation of cells and tissues and methanol-based fixation of cells. However, the inclusion of these approaches is still limited in many settings due to the challenging logistics involved in sample collection. To further broaden the ability of researchers to deploy scRNA-seq, we report our experiences from testing a pre-commercial workflow developed by 10x Genomics to resolve these sample management limitations. The use of paraformaldehyde (PFA) to fix samples at the collection site allows samples to be transported to remote laboratories for downstream processing without sacrificing integrity or data quality. This advance has enabled new possibilities for sample accessibility, throughput, and batched analysis in basic and translational research settings. Furthermore, this approach also allows for analysis of transcriptome-wide gene expression at high sensitivity with simultaneous cellular readouts. Here we report a case study comparing the PFA method to standard approaches, and highlight advantages and potential applications previously not possible with reported sample preparation strategies. Citation Format: David Corney, Yang Han, Yu Qiu, Yongjun Fan, Michael Stephens, Andrea O'Hara, Laure Turner, Christopher Mozdzierz, Haythem Latif, Ginger Zhou. Evaluation of a novel single-cell RNA sequencing methodology for use in clinical trial settings [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr LB034.

Transverse endoplasmic reticulum expansion in hereditary spastic paraplegia corticospinal axons.

Hereditary spastic paraplegias (HSPs) comprise a large group of inherited neurologic disorders affecting the longest corticospinal axons (SPG1-86 plus others), with shared manifestations of lower extremity spasticity and gait impairment. Common autosomal dominant HSPs are caused by mutations in genes encoding the microtubule-severing ATPase spastin (SPAST; SPG4), the membrane-bound GTPase atlastin-1 (ATL1; SPG3A) and the reticulon-like, microtubule-binding protein REEP1 (REEP1; SPG31). These proteins bind one another and function in shaping the tubular endoplasmic reticulum (ER) network. Typically, mouse models of HSPs have mild, later onset phenotypes, possibly reflecting far shorter lengths of their corticospinal axons relative to humans. Here, we have generated a robust, double mutant mouse model of HSP in which atlastin-1 is genetically modified with a K80A knock-in (KI) missense change that abolishes its GTPase activity, whereas its binding partner Reep1 is knocked out. Atl1KI/KI/Reep1-/- mice exhibit early onset and rapidly progressive declines in several motor function tests. Also, ER in mutant corticospinal axons dramatically expands transversely and periodically in a mutation dosage-dependent manner to create a ladder-like appearance, on the basis of reconstructions of focused ion beam-scanning electron microscopy datasets using machine learning-based auto-segmentation. In lockstep with changes in ER morphology, axonal mitochondria are fragmented and proportions of hypophosphorylated neurofilament H and M subunits are dramatically increased in Atl1KI/KI/Reep1-/- spinal cord. Co-occurrence of these findings links ER morphology changes to alterations in mitochondrial morphology and cytoskeletal organization. Atl1KI/KI/Reep1-/- mice represent an early onset rodent HSP model with robust behavioral and cellular readouts for testing novel therapies.

Microplastics and Their Impact on Reproduction-Can we Learn From the C. elegans Model?

Biologically active environmental pollutants have significant impact on ecosystems, wildlife, and human health. Microplastic (MP) and nanoplastic (NP) particles are pollutants that are present in the terrestrial and aquatic ecosystems at virtually every level of the food chain. Moreover, recently, airborne microplastic particles have been shown to reach and potentially damage respiratory systems. Microplastics and nanoplastics have been shown to cause increased oxidative stress, inflammation, altered metabolism leading to cellular damage, which ultimately affects tissue and organismal homeostasis in numerous animal species and human cells. However, the full impact of these plastic particles on living organisms is not completely understood. The ability of MPs/NPs to carry contaminants, toxic chemicals, pesticides, and bioactive compounds, such as endocrine disrupting chemicals, present an additional risk to animal and human health. This review will discusses the current knowledge on pathways by which microplastic and nanoplastic particles impact reproduction and reproductive behaviors from the level of the whole organism down to plastics-induced cellular defects, while also identifying gaps in current knowledge regarding mechanisms of action. Furthermore, we suggest that the nematode Caenorhabditis elegans provides an advantageous high-throughput model system for determining the effect of plastic particles on animal reproduction, using reproductive behavioral end points and cellular readouts.

Pharmacological tools to investigate inositol polyphosphate kinases - Enzymes of increasing therapeutic relevance.

Inositol poly- and pyrophosphates (InsPs and PP-InsPs) are a group of central eukaryotic metabolites and signaling molecules. Due to the diverse cellular functions and widespread diseases InsPs and PP-InsPs are associated with, pharmacological targeting of the kinases involved in their biosynthesis has become a significant research interest in the last decade. In particular, the development of inhibitors for inositol hexakisphosphate kinases (IP6Ks) has leaped forward, while other inositol phosphate kinases have received scant attention. This review summarizes the efforts undertaken so far for discovering potent and selective inhibitors for this diverse group of small molecule kinases. The benefits of pharmacological inhibition are highlighted, given the multiple kinase-independent functions of inositol phosphate kinases. The distinct structural families of InsP and PP-InsP kinases are presented, and we discuss how compound availability for different inositol phosphate kinase families varies drastically. Lead compound discovery and optimization for the inositol kinases would benefit from detailed structural information on the ATP-binding sites of these kinases, as well as reliable biochemical and cellular read-outs to monitor inositol phosphate kinase activity in complex settings. Efforts to further tune well-established inhibitors, while simultaneously reviving tool compound development for the more neglected kinases from this family are indisputably worthwhile, considering the large potential therapeutic benefits.

Abstract P247: Evaluating TRKB activity of novel preclinical brain-penetrant ROS1 and ALK inhibitors

Abstract Oncogenic ROS1 and ALK rearrangements are detected in up to 3% and up to 5% of advanced non-small cell lung cancer (NSCLC), respectively, with up to 40% of patients presenting with brain metastases at diagnosis. Although there are FDA-approved ROS1 and ALK inhibitors, there is a need for next-generation brain-penetrant inhibitors that have activity against drug-resistance mutations while sparing tropomyosin kinase B (TRKB) function in the central nervous system (CNS). TRKB inhibition in the CNS has been implicated in adverse events such as cognitive impairment, mood disorders, sleep effects or sleep disturbances, dizziness, ataxia or gait/motor disturbances, and weight gain, which have been observed with FDA-approved dual TRK/ROS1 inhibitor entrectinib and FDA-approved ALK inhibitor lorlatinib. TRKB-related CNS adverse events present a key challenge for the development of next-generation ROS1 and ALK therapies. We previously reported the discovery of NVL-520 and NVL-655, preclinical inhibitors of ROS1 and ALK, respectively, which have demonstrated activity against common drug-resistance mutations, selectivity versus TRKB, and brain penetrance in preclinical experiments. An important part of the discovery process was generating reliable and relevant TRKB potency measurements to enable structure-activity relationship (SAR) efforts. Here we describe four in vitro TRKB potency assays and explain their utility in guiding selectivity optimization for new inhibitors. These assays include (1) biochemical, (2) cell viability, (3) cellular coupled-enzyme, and (4) direct cellular phosphorylation readouts. Although biological contexts vary, each assay exhibits good cross-correlations, and we recommend multi-assay testing to improve confidence in evaluating TRKB inhibition. Using these assays, we present a multi-assay analysis of the preclinical TRKB potency and selectivity of NVL-520 and NVL-655 alongside FDA-approved or experimental inhibitors of ROS1 (crizotinib, entrectinib, repotrectinib, and taletrectinib) and ALK (crizotinib, ceritinib, alectinib, brigatinib, lorlatinib, and ensartinib). In these preclinical assays, we observed entrectinib and lorlatinib to exhibit poor ROS1-vs-TRKB and ALK G1202R-vs-TRKB selectivity, respectively, which we believe is consistent with their CNS adverse events. In the same experiment, we observed NVL-520 and NVL-655 to exhibit selectivity across all assays, including for ROS1 G2032R and ALK G1202R/L1196M mutations, respectively. In conclusion, we have devised a multi-assay approach to evaluate TRKB inhibition and guide discovery of selective ROS1 and ALK inhibitors, with the goal of minimizing adverse events and driving durable responses for patients. Citation Format: Anupong Tangpeerachaikul, Joshua C. Horan, Henry E. Pelish. Evaluating TRKB activity of novel preclinical brain-penetrant ROS1 and ALK inhibitors [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2021 Oct 7-10. Philadelphia (PA): AACR; Mol Cancer Ther 2021;20(12 Suppl):Abstract nr P247.

Abstract P084: DCC-3116, a first-in-class selective inhibitor of ULK1/2 kinases and autophagy, synergizes with EGFR inhibitors osimertinib and afatinib in NSCLC preclinical models

Abstract Background: Activating mutations in EGFR have been reported in ~30% of patients with non-small cell lung cancer (NSCLC). Three generations of small molecule EGFR kinase inhibitors have been approved by the FDA to treat these patients, however, multiple mechanisms of resistance cause cancer progression. In addition to drug-resistant mutations that arise and re-activate EGFR, other signaling pathways can be activated to cause resistance. Although EGFR inhibitors such as osimertinib and afatinib (also a pan-ErbB inhibitor) have demonstrated clear clinical benefits, patients inevitably develop resistance. Herein we demonstrate mutant EGFR NSCLCs activate autophagy upon treatment with EGFR inhibitors as a drug resistance mechanism. Therefore, a combination of an EGFR inhibitor with an autophagy inhibitor has the potential to deepen and prolong responses and improve patient outcomes. Materials and Methods: Human cancer cells with EGFR mutations were cultured using recommended complete medium. Inhibition of ULK1/2 was measured through standard biochemical assays and cellular readouts including NanoBRET and ELISA-based ATG13 phosphorylation assays. Autophagosome formation was measured using the CytoID assay (Enzo Life Sciences). For in vivo studies, NCI-H1975 cells that harbor an EGFR T790M resistance mutation were inoculated into BALB/c nude mice. Statistical analyses for the in vivo studies were performed using the Student’s t-test. Results: Erlotinib, gefitinib, osimertinib, and afatinib activated autophagy 3–4-fold over basal levels in the HCC827 cell line (EGFR exon 19 deletion) as measured by increases in phosphorylated ATG13, a cellular substrate of the autophagy-initiating kinases ULK1/2. DCC-3116, an investigational potent and selective dual inhibitor of ULK1 (IC50 6 nM) and ULK2 (9 nM) in cellular assays, inhibited both EGFR-induced and basal phosphorylation of ATG13 with IC50 values of 61–66 nM. Treatment of the NCI-H1975 EGFR mutated (L858R/T790M) NSCLC cell line with osimertinib or afatinib induced autophagy 3-fold over basal levels. DCC-3116 potently inhibited osimertinib and afatinib induced phosphorylation of ATG13 with IC50 values of 91 nM and 71 nM, respectively, and inhibited the increase in autophagosomes induced by these agents. Importantly, these in vitro effects translated to in vivo efficacy. The combination of DCC-3116 with osimertinib or afatinib resulted in significantly greater tumor responses than single agent treatments in the NCI-H1975 NSCLC xenograft model (combination vs. osimertinib p = 0.0005; combination vs. afatinib p = 0.0001; osimertinib combination vs vehicle p < 0.0001; afatinib combination vs. vehicle p < 0.0001). These data provide a strong rationale to study the combination of the ULK inhibitor DCC-3116 with EGFR inhibitors such as osimertinib and afatinib in cancer patients. DCC-3116 is currently in a Phase 1 clinical trial in patients with advanced solid tumors with a documented RAS or RAF mutation (NCT04892017). Citation Format: Madhumita Bogdan, Mary J. Timson, Hikmat Al-Hashimi, Yu Zhan, Bryan D. Smith, Daniel L. Flynn. DCC-3116, a first-in-class selective inhibitor of ULK1/2 kinases and autophagy, synergizes with EGFR inhibitors osimertinib and afatinib in NSCLC preclinical models [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2021 Oct 7-10. Philadelphia (PA): AACR; Mol Cancer Ther 2021;20(12 Suppl):Abstract nr P084.

Abstract 1021: Investigating the mechanism of PARP7 inhibition in type I interferon signaling by arrayed CRISPR screening

Abstract Genomic instability in cancer cells leads to cellular stress through the accumulation of aberrant nucleic acid species in the cytosol. We have shown that PARP7, a monoPARP, is a negative regulator of cytosolic nucleic acid sensing in cancer cells. RBN-2397 is a potent and selective PARP7 inhibitor that induces antitumor immunity in preclinical models and is currently being evaluated in a Phase I clinical trial. In our preclinical investigations, we found that in a subset of cancer cell lines, such as NCI-H1373, inhibition of PARP7 triggers Type I IFN release, STAT1 phosphorylation, and growth arrest. In contrast, other cell lines, for example, HARA, do not mount an IFN-response upon PARP7 inhibition, even though they are responsive to transfection of exogenous nucleic acids and PARP7 is expressed and enzymatically active. To investigate the underlying mechanism of PARP7 inhibition and to determine the drivers of the differential sensitivity across cell lines we performed arrayed CRISPR knockout screens, targeting approximately 240 genes in the nucleic acid sensing and IFN signaling pathways, in the presence and absence of PARP7 inhibition. Our arrayed screens confirmed multiple hits from a previous genome-wide pooled synthetic/lethal CRISPR dropout screen. For example, targeting genes in the cGAS/STING pathway conferred resistance to PARP7 inhibition in the NCI-H1373 responder cells, suggesting a critical dependence on this sensing pathway. In the PARP7 inhibitor-resistant HARA cells, deletion of components of innate immune-signaling (such as AIM2 and ADAR1), the NF-κB pathway, and genes involved in autophagy sensitized the cells to PARP7 inhibition. We further delineated the function of PARP7 by comparing the effects of the CRISPR perturbation across different cellular readouts such as STAT1 phosphorylation, IFN release, and proliferation. With our work, we shed light on the mechanism by which PARP7 acts as a critical suppressor of the innate immune response. Our findings demonstrate both redundancy and crosstalk between different nucleic acid-sensing pathways and may explain why some cell lines are resistant to PARP7 inhibition. Citation Format: Bin Gui, Ryan Abo, Patrick Flynn, Alvin Z. Lu, Jan-Rung Mo, Joseph M. Gozgit, Melissa M. Vasbinder, Zacharenia A. Varsamis, Andrew Santospago, Victoria M. Richon, Kevin W. Kuntz, Heike Keilhack, Timothy J. Mitchison, Mario Niepel. Investigating the mechanism of PARP7 inhibition in type I interferon signaling by arrayed CRISPR screening [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 1021.

Abstract 1378: A novel, highly potent PI3Kα covalent inhibitor deconvolutes class I PI3K isoforms in cancer cells

Abstract Phosphoinositide 3-kinase (PI3K) imposes central regulatory signals that promote cell growth, proliferation, motility, and suppression of apoptosis. PI3K is dysregulated in a broad spectrum of human tumours.[1] Considering the central role attributed to PI3K signaling in cancer biology, the success of clinical trials in the field was below expectations, mainly due to metabolic on-target adverse effects. Here, we present a novel, highly selective PI3Kα-covalent inhibitor (compound 19), which combines the exceptional affinity of the pan-PI3K inhibitor PQR514[2] with a linker moiety, directing a warhead to selectively react via a Michael addition with isoform-specific Cys862 of PI3Kα. Cellular assay platforms were combined to validate 19, specifically, to assess its target selectivity and occupancy, and to translate its in vitro activity to cellular readouts. Overall, 19 exhibited excellent cellular activity with a superior profile compared to CNX-1351[3], a previously reported covalent PI3Kα inhibitor. A NanoBRET (Bioluminescence Resonance Energy Transfer) drug-target engagement assay allowed for an investigation of inhibitory efficacy in HEK293 cells. NanoBRET experiments exploiting the Cys862Ser mutant PI3Kα protein confirmed selective reactivity with Cys862. Various cell lines either constitutively activated in PI3Kα (MCF7, T47D, SKOV3) or deficient in PTEN (PC3 and A2058) were used to resolve target engagement, which confirmed a persistent and prolonged suppression of protein kinase B (PKB/Akt, Ser473) phosphorylation after washout of 19 over non-covalent structural analogs thereof that lack a functional warhead. Furthermore, nucleo-cytoplasmic translocation probes based on the FOXO1 transcription factor enabled the investigation of signal propagation downstream of PKB/Akt. In PI3Kα-dependent cell lines, FOXO1, maintained a predominant nuclear localization even after removal of 19, confirming extended PI3Kα targeting upon covalent attachment of specifically this drug. Importantly, 19 provided a considerable gain in potency over BYL719, a clinically approved non-covalent PI3Kα-selective inhibitor. Taken together, our results present a major step towards an increased local and temporal control of PI3Kα signaling through isoform-selective covalent anchoring. This strategy is a step towards the development of drug-like, isoform-selective, covalent PI3K inhibitor scaffolds, and 19 provides a template for the development of PI3Kα inhibitors with overall reduced adverse side effects in cancer therapy. Moreover, our covalent inhibition strategy will pave the way for a better molecular understanding of PI3K isoform signaling in general. [1] Wymann M. P. and Schneiter R. Nat. Rev. Mol. Cell Biol. 2008, 9, 162-176. [2] Borsari, C et. al. ACS Med. Chem. Lett. 2019, 10 (10), 1473-1479. [3] Nacht M et. al. J. Med. Chem. 2013, 56 (3), 712-721. Citation Format: Erhan Keles, Chiara Borsari, Rohitha Sriramaratnam, Thorsten Schäfer, Matthias P. Wymann. A novel, highly potent PI3Kα covalent inhibitor deconvolutes class I PI3K isoforms in cancer cells [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 1378.

Partial agonism improves the anti-hyperglycaemic efficacy of an oxyntomodulin-derived GLP-1R/GCGR co-agonist

ObjectiveGlucagon-like peptide-1 and glucagon receptor (GLP-1R/GCGR) co-agonism can maximise weight loss and improve glycaemic control in type 2 diabetes and obesity. In this study, we investigated the cellular and metabolic effects of modulating the balance between G protein and β-arrestin-2 recruitment at GLP-1R and GCGR using oxyntomodulin (OXM)-derived co-agonists. This strategy has been previously shown to improve the duration of action of GLP-1R mono-agonists by reducing target desensitisation and downregulation. MethodsDipeptidyl dipeptidase-4 (DPP-4)-resistant OXM analogues were generated and assessed for a variety of cellular readouts. Molecular dynamic simulations were used to gain insights into the molecular interactions involved. In vivo studies were performed in mice to identify the effects on glucose homeostasis and weight loss. ResultsLigand-specific reductions in β-arrestin-2 recruitment were associated with slower GLP-1R internalisation and prolonged glucose-lowering action in vivo. The putative benefits of GCGR agonism were retained, with equivalent weight loss compared to the GLP-1R mono-agonist liraglutide despite a lesser degree of food intake suppression. The compounds tested showed only a minor degree of biased agonism between G protein and β-arrestin-2 recruitment at both receptors and were best classified as partial agonists for the two pathways measured. ConclusionsDiminishing β-arrestin-2 recruitment may be an effective way to increase the therapeutic efficacy of GLP-1R/GCGR co-agonists. These benefits can be achieved by partial rather than biased agonism.