Cell Line Screening Research Articles

Large-scale Pan-cancer Cell Line Screening Identifies Actionable and Effective Drug Combinations.

We present the largest cancer drug combination screen published to date with 7 × 7 concentration response matrices for 109 combinations in more than 750 cell lines, complemented by multi-omics predictors of response and identification of "emergent" combination biomarkers. We prioritize hits to optimize clinical translatability, and experimentally validate novel combination hypotheses. This article is featured in Selected Articles from This Issue, p. 695.

CD155/PVR determines acute myeloid leukemia targeting by Delta One T cells

AbstractRelapsed or refractory acute myeloid leukemia (AML) remains a major therapeutic challenge. We have recently developed a Vδ1+ γδ T cell–based product for adoptive immunotherapy, named Delta One T (DOT) cells, and demonstrated their cytolytic capacity to eliminate AML cell lines and primary blasts in vitro and in vivo. However, the molecular mechanisms responsible for the broad DOT-cell recognition of AML cells remain poorly understood. Here, we dissected the role of natural killer (NK) cell receptor ligands in AML cell recognition by DOT cells. Screening of multiple AML cell lines highlighted a strong upregulation of the DNAM-1 ligands, CD155/pulmonary vascular resistance (PVR), CD112/nectin-2, as well as the NKp30 ligand, B7-H6, in contrast with NKG2D ligands. CRISPR-mediated ablation revealed key nonredundant and synergistic contributions of PVR and B7-H6 but not nectin-2 to DOT-cell targeting of AML cells. We further demonstrate that PVR and B7-H6 are critical for the formation of robust immunological synapses between AML and DOT cells. Importantly, PVR but not B7-H6 expression in primary AML samples predicted their elimination by DOT cells. These data provide new mechanistic insight into tumor targeting by DOT cells and suggest that assessing PVR expression levels may be highly relevant to DOT cell–based clinical trials.

Sesquiterpene Lactones and Flavonoid from the Leaves of Basin Big Sagebrush (Artemisia tridentata subsp. tridentata): Isolation, Characterization and Biological Activities.

This research is an exploratory study on the sesquiterpenes and flavonoid present in the leaves of Artemisia tridentata subsp. tridentata. The leaf foliage was extracted with 100% chloroform. Thin-layer chromatography (TLC) analysis of the crude extract showed four bands. Each band was purified by column chromatography followed by recrystallization. Three sesquiterpene lactones (SLs) were isolated-leucodin, matricarin and desacetylmatricarin. Of these, desacetylmatricarin was the major component. In addition, a highly bio-active flavonoid, quercetagetin 3,6,4'-trimethyl ether (QTE), was also isolated. This is the first report on the isolation of this component from the leaves of Artemisia tridentata subsp. tridentata. All the components were identified and isolated by TLC, high-performance liquid chromatography (HPLC) and mass spectrometry (MS) techniques. Likewise, the structure and stereochemistry of the purified components were characterized by extensive spectroscopic analysis, including 1D and 2D nuclear magnetic resonance (NMR) and Fourier transform infrared spectroscopy (FTIR) studies. The antioxidant activities of crude extract were analyzed, and their radical-scavenging ability was determined by Ferric reducing antioxidant power (FRAP) and 2,2-diphenyl-1-picrylhydrazyl (DPPH) assay. The crude extract showed antioxidant activity of 18.99 ± 0.51 and 11.59 ± 0.38 µmol TEg-1 FW for FRAP and DPPH assay, respectively, whereas the activities of matricarin, leucodin, desacetylmatricarin and QTE were 13.22, 13.03, 14.90 and 15.02 µmol TEg-1 FW, respectively, for the FRAP assay. The antitumor properties were probed by submitting the four isolated compounds to the National Cancer Institute (NCI) for NCI-60 cancer cell line screening. Overall, the results of the one-dose assay for each SL were unremarkable. However, the flavonoid's one-dose mean graph demonstrated significant growth inhibition and lethality, which prompted an evaluation of this compound against the 60-cell panel at a five-dose assay. Tests from two separate dates indicate a lethality of approximately 75% and 98% at the log-4 concentration when tested against the melanoma cancer line SK-Mel 5. This warrants further testing and derivatization of the bioactive components from sagebrush as a potential source for anticancer properties.

Spheroid Drug Sensitivity Screening in Glioma Stem Cell Lines.

In vitro drug sensitivity screens are important tools in the discovery of anti-cancer drug combination therapies. Typically, these in vitro drug screens are performed on cells grown in a monolayer. However, these two-dimensional (2D) models are considered less accurate compared to three-dimensional (3D) spheroid cell models; this is especially true for glioma stem cell lines. Cells grown in spheres activate different signaling pathways and are considered more representative of in vivo models than monolayer cell lines. This protocol describes a method for in vitro drug screening of spheroid lines; mouse and human glioma stem cell lines are used as an example. This protocol describes a 3D spheroid drug sensitivity and synergy assay that can be used to determine if a drug or drug combination induces cell death and if two drugs synergize. Glioma stem cell lines aremodified to express RFP. Cells are plated in low attachment round well bottom 96 plates, and spheres are allowed to form overnight. Drugs are added, and the growth is monitored by measuring the RFP signal over time using the Incucyte live imaging system,a fluorescence microscope embedded in the tissue culture incubator. Half maximal inhibitory concentration (IC50), median lethal dose (LD50), and synergy score are subsequently calculated to evaluate sensitivities to drugs alone or in combination. The three-dimensional nature of this assayprovides a more accurate reflection of tumor growth, behavior, and drug sensitivities in vivo, thus forming the basis for further preclinical investigation.

Duplication of a chromosome 2 segment in production CHO cell lines correlates with age-related growth improvement.

Monitoring the stability of recombinant Chinese Hamster Ovary (CHO) cell lines is essential to ensure the selection of production cell lines suitable for biomanufacturing. It has been frequently observed that recombinant CHO cell lines develop phenotypic changes upon aging, such as accelerated cell growth in late generation cultures. However, the mechanism responsible for age-correlated changes is poorly understood. In this study, we investigated the molecular mechanisms underlying the age-correlated cell growth improvement in Pfizer's platform fed-batch production process, by examining multiple cell lines derived from different CHO expression systems, expressing a variety of monoclonal antibodies (mAbs). Comprehensive whole-genome resequencing analysis revealed duplication of a continuous 50.2 Mbp segment in chromosome 2 (Chr2) specific to clones that showed age-correlated growth change as compared to clones that did not exhibit age-correlated growth change. Moreover, such age- and growth-related Chr2 duplication was independent of the presence or type of recombinant monoclonal antibody expression. When we compared transcriptome profiles from low-growth and high-growth cell lines, we found that>95% of the genes overexpressed in high-growth cell lines were in the duplicated Chr2 segment. To the best of our knowledge, this is the first report of large genomic duplication, specific to Chr2, being associated with age-correlated growth change. Investigation of the cause-and-effect relationship between the genes identified in the duplicated regions and age-correlated growth change is underway. We are confident that this effort will lead to improved cell line screening and targeted rational cell line engineering efforts to develop cell lines with improved stability performance.

Repurposing proteasome inhibitors for improved treatment of triple-negative breast cancer

Triple-negative breast cancer (TNBC) is associated with poor prognosis and limited treatment options due to the lack of important receptors (estrogen receptor [ER], progesterone receptor [PR], and human epidermal growth factor receptor 2 [HER2]) used for targeted therapy. However, high-throughput in vitro drug screening of cell lines is a powerful tool for identifying effective drugs for a disease. Here, we determine the intrinsic chemosensitivity of TNBC cell lines to proteasome inhibitors (PIs), thereby identifying potentially potent 2-drug combinations for TNBC. Eight TNBC cell lines (BT-549, CAL-148, HCC1806, HCC38, HCC70, MDA-MB-436, MDA-MB-453, and MDA-MB-468) and two controls (MCF-10A and MCF-7) were first exposed to 18 drugs (11 PIs and 7 clinically relevant chemotherapeutic agents) as monotherapy, followed by prediction of potent 2-drug combinations using the IDACombo pipeline. The synergistic effects of the 2-drug combinations were evaluated with SynergyFinder in four TNBC cell lines (CAL-148, HCC1806, HCC38, and MDA-MB-468) and three controls (BT-474, MCF-7, and T47D) in vitro, followed by further evaluation of tumor regression in zebrafish tumor models established using HCC1806 and MCF-7 cells. Monotherapy identified nine effective drugs (bortezomib, carfilzomib, cisplatin, delanzomib, docetaxel, epoxomicin, MLN-2238, MLN-9708, and nedaplatin) across all cell lines. PIs (e.g., bortezomib, delanzomib, and epoxomicin) were highly potent drugs in TNBC cells, of which bortezomib and delanzomib inhibited the chymotrypsin-like activity of the 20 S proteasome by 100% at 10 µM. Moreover, several potent 2-drug combinations (e.g., bortezomib+nedaplatin and epoxomicin+epirubicin) that killed virtually 100% of cells were also identified. Although HCC1806- and MCF-7-derived xenografts treated with bortezomib+nedaplatin and carboplatin+paclitaxel were smaller, HCC1806 cells frequently metastasized to the trunk region. Taken together, we show that PIs used in combination with platinum agents or topoisomerase inhibitors exhibit increased efficiency with almost 100% inhibition in TNBC cell lines, indicating that PIs are therefore promising compounds to use as combination therapy for TNBC.

A comprehensive clinically informed map of dependencies in cancer cells and framework for target prioritization

Genetic screens in cancer cell lines inform gene function and drug discovery. More comprehensive screen datasets with multi-omics data are needed to enhance opportunities to functionally map genetic vulnerabilities. Here, we construct a second-generation map of cancer dependencies by annotating 930 cancer cell lines with multi-omic data and analyze relationships between molecular markers and cancer dependencies derived from CRISPR-Cas9 screens. We identify dependency-associated gene expression markers beyond driver genes, and observe many gene addiction relationships driven by gain of function rather than synthetic lethal effects. By combining clinically informed dependency-marker associations with protein-protein interaction networks, we identify 370 anti-cancer priority targets for 27 cancer types, many of which have network-based evidence of a functional link with a marker in a cancer type. Mapping these targets to sequenced tumor cohorts identifies tractable targets in different cancer types. This target prioritization map enhances understanding of gene dependencies and identifies candidate anti-cancer targets for drug development.

Abstract B005: Tolerance to colibactin correlates with response to chemotherapeutic agents in colorectal cancer

Abstract Introduction. The bacterial genotoxin colibactin is enriched in colorectal cancer (CRC) and promotes the accumulation of mutations that drive tumorigenesis. However, systematic assessment of its impact on DNA damage response is lacking and the effect of colibactin exposure on response to other genotoxic agents (such as chemotherapy) is missing. Materials and methods: We implemented an in vitro bacteria-coculture system to assess the effect of colibactin on a representative subset of 40 molecularly and pharmacologically annotated CRC cell lines and in a panel of isogenic DDR KO cell lines we generated. We further validated our results in patient-derived organoids. Finally, we recapitulated prolonged exposure to colibactin occurring during tumorigenesis by chronically infecting sensitive cells until the emergence of a tolerant phenotype. Results: We found that different cell lines display specific sensitivity to colibactin’s genotoxic stress: while colibactin-tolerant cells are capable of quickly and efficiently repairing colibactin-induced DNA damage, sensitive cells lack this ability. Moreover, we found that homologous recombination (HR) proficiency discriminates colibactin-tolerant cells, which display higher levels of RAD51 foci (as marker of activation of HR) compared to sensitive cells upon infection with colibactin. Screening of isogenic DDR KO cell lines revealed that genetic inactivation of the intertwined pathways of HR (through KO of ATM) and replication stress (RS) response (through KO of ATRIP) significantly sensitized cells to colibactin. In addition, we found that restoration of HR activity was sufficient to induce a colibactin-tolerant phenotype in initially sensitive cell lines. Notably, thanks to a previous effort of pharmacological characterization of CRC cell lines in our lab, we found a significant correlation between sensitivity to colibactin and irinotecan active metabolite SN38, but not oxaliplatin. We validated the same correlation in patient-derived organoids annotated for response to SN38. While colibactin, SN38 and oxaliplatin all induced RS in treated cells, we found that colibactin and SN38 showed a similar DNA damage response which involved activation of ATM. Finally, chronic re-infection of sensitive, HR-deficient CRC cells with colibactin selected a tolerant phenotype characterized by restoration of HR activity. Of translational relevance, colibactin-tolerant derivative cells acquired cross-resistance to SN38 and PARP inhibitor olaparib but not to oxaliplatin. Conclusion: Our results shed novel insight into colibactin’s genotoxic mechanism and support a model in which colibactin both promotes tumorigenesis and acts as an evolutionary bottleneck which selects HR proficient CRC cells. Furthermore, our study provides preclinical evidence on colibactin’s role in promoting chemoresistance in colorectal cancer. Citation Format: Alberto Sogari, Emanuele Rovera, Nicole Megan Reilly, Simona Lamba, Erika Durinikova, Annalisa Lorenzato, Marco Avolio, Eleonora Piumatti, Mariangela Russo, Sabrina Arena, Livio Trusolino, Manuela Donalisio, Federica Di Nicolantonio, David Lembo, Alberto Bardelli. Tolerance to colibactin correlates with response to chemotherapeutic agents in colorectal cancer [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: DNA Damage Repair: From Basic Science to Future Clinical Application; 2024 Jan 9-11; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2024;84(1 Suppl):Abstract nr B005.

Abstract B023: SMARCAL1 is a novel candidate therapeutic target for ALT-positive osteosarcoma

Abstract The DNA translocase SMARCAL1 has been characterized by our laboratory and others as a DNA replication-associated factor that promotes the remodeling of stalled replication forks to facilitate the repair of DNA damage and the restart of DNA synthesis. Recent genome-wide analyses of CRISPR-Cas9 loss-of-function screens in pediatric cancer cell lines have suggested an essential function for SMARCAL1 in a subset of pediatric tumors (Dharia NV et al., Nature Genetics, 2021). In agreement with these observations, pan-cancer analysis of dependency scores in over 1,000 cell lines from the DepMap project revealed an important role for SMARCAL1 in tumors of mesenchymal origin and especially in osteosarcoma. We found that SMARCAL1 dependency in osteosarcoma is selectively associated with lack of telomerase expression. In the absence of telomerase, cancer cells mostly rely on a recombination-based mechanism, known as ALT (Alternative Lengthening of Telomeres), for the maintenance of their telomere length. Surprisingly, by comparing genome-wide CRISPR screening data of known ALT-positive vs ALT-negative osteosarcoma cell lines, we found that SMARCAL1 is among the genes that are most selectively essential in ALT-positive tumors. We validated in multiple cellular models a strict correlation and functional relationship between ALT status and SMARCAL1 dependency. In line with previous findings (Cox KE et al., Cell Reports, 2016), we observed that SMARCAL1 localizes to a subset of telomeres in ALT-positive cells and loss of SMARCAL1 induces accumulation of telomere damage. SMARCAL1 enzymatic activity and its RPA-binding function contribute to suppress DNA damage in ALT-positive cells. Furthermore, our studies suggest that SMARCAL1 protects telomere stability in ALT-positive cancer cells by limiting PRIMPOL-mediated restart of stalled replication forks. In addition, SMARCAL1 deficiency enhances cell-intrinsic immune response and anti-tumor immunity. Collectively, our studies have uncovered SMARCAL1 as a novel candidate therapeutic target for the treatment of ALT-positive cancer patients. In these patients, SMARCAL1 inhibition may have a dual beneficial effect by promoting selective cancer cell toxicity combined with stimulation of anti-tumor immunity and enhancement of immunotherapy response. Citation Format: Angelo Taglialatela, Xiao Chen, Zahra Fatema Khan, Filemon Dela Cruz, Dua Shivi, Dillon Oswald, Azeroglu Benura, Eros Lazzerini Denchi, Min Jaewon, Andrew L. Kung, Alberto Ciccia. SMARCAL1 is a novel candidate therapeutic target for ALT-positive osteosarcoma [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: DNA Damage Repair: From Basic Science to Future Clinical Application; 2024 Jan 9-11; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2024;84(1 Suppl):Abstract nr B023.

A single-step high-throughput bioassay for quantifying Fc-containing recombinant proteins based on non-classical calculation of fluorescence polarization.

The titer of recombinant proteins is one of the key parameters in biopharmaceutical manufacturing processes. The fluorescence polarization (FP)-based assay, a homogeneous, high-throughput and real-time analytical method, had emerged as a powerful tool for biochemical analysis and environmental monitoring. In this study, an FP-based bioassay was utilized to quantify antibody fragment crystallizable (Fc)-containing proteins, such as recombinant monoclonal antibodies (mAbs) and mAb derivatives, in the cell culture supernatant, and the impacts of tracer molecular weight and FITC-coupling conditions on fluorescence polarization were methodically examined. Distinct from the fluorescence polarization potency calculated by classical formula, we for the first time proposed a new concept and calculation of fluorescence polarization intensity, based on which an analytical method with broader detection range and analysis window was established for quantifying Fc-containing proteins. This provided new ideas for the practical application of fluorescence polarization theory. The established method could detect 96 samples within 30 minutes, with dynamic titer range of 2.5-400 mg L-1, and a linear fitting R2 between the measured and actual concentration reaching 0.99. The method had great application prospects in determining the titer of recombinant proteins with Fc fragments, especially when applied to large-scale screening of high-yield and stable expression CHO cell lines commonly used in biopharmaceutical industry.

In vitro and in vivo evaluation of novel chromeno[2,3-d]pyrimidinones as therapeutic agents for triple negative breast cancer.

Triple-negative breast cancer (TNBC) is the most aggressive subtype of breast cancer, and the limited therapeutic options show poor efficacy in patients, associated to severe side effects and development of resistance. Considering that chromene-based scaffolds proved to be attractive candidates for cancer therapy, herein we prepared new chromeno[2,3-d]pyrimidinone derivatives by a simple two step procedure, starting from the reaction of cyanoacetamide and a salicylaldehyde. A cell viability screening in several breast cancer cell lines allowed to identify two promising compounds with IC50 values in the low micromolar range for TNBC cells. These chromenes inhibited cell proliferation, induced cell cycle arrest and triggered cell death through apoptosis. In vivo studies revealed a safe profile in invertebrate and vertebrate animal models and confirmed their capacity to inhibit tumor growth in the CAM model, inducing significant tumor regression after 4 days of treatment. The two compounds identified in this study are promising drug candidates for TNBC treatment and valuable hits for future optimization, using the versatile synthetic platform that was developed.

JAK2 Loss Arising From Tumor-Spread-Through-Air-Spaces (STAS) Promotes Tumor Progression by Suppressing CD8+ T Cells in Lung Adenocarcinoma: A Machine Learning Approach.

Tumor spread through air spaces (STAS) is a recently discovered risk factor for lung adenocarcinoma (LUAD). The aim of this study was to investigate specific genetic alterations and anticancer immune responses related to STAS. By using a machine learning algorithm and drug screening in lung cancer cell lines, we analyzed the effect of Janus kinase 2 (JAK2) on the survival of patients with LUAD and possible drug candidates. This study included 566 patients with LUAD corresponding to clinicopathological and genetic data. For analyses of LUAD, we applied gene set enrichment analysis (GSEA), in silico cytometry, pathway network analysis, in vitro drug screening, and gradient boosting machine (GBM) analysis. The patients with STAS had a shorter survival time than those without STAS (P < 0.001). We detected gene set-related downregulation of JAK2 associated with STAS using GSEA. Low JAK2 expression was related to poor prognosis and a low CD8+ T-cell fraction. In GBM, JAK2 showed improved survival prediction performance when it was added to other parameters (T stage, N stage, lymphovascular invasion, pleural invasion, tumor size). In drug screening, mirin, CCT007093, dihydroretenone, and ABT737 suppressed the growth of lung cancer cell lines with low JAK2 expression. In LUAD, low JAK2 expression linked to the presence of STAS might serve as an unfavorable prognostic factor. A relationship between JAK2 and CD8+ T cells suggests that STAS is indirectly related to the anticancer immune response. These results may contribute to the design of future experimental research and drug development programs for LUAD with STAS.

Construction and application of a multifunctional CHO cell platform utilizing Cre/lox and Dre/rox site-specific recombination systems.

During the development of traditional Chinese hamster ovary (CHO) cell lines, target genes randomly integrate into the genome upon entering the nucleus, resulting in unpredictable productivity of cell clones. The characterization and screening of high-yielding cell lines is a time-consuming and expensive process. Site-specific integration is recognized as an effective approach for overcoming random integration and improving production stability. We have designed a multifunctional expression cassette, called CDbox, which can be manipulated by the site-specific recombination systems Cre/lox and Dre/rox. The CDbox expression cassette was inserted at the Hipp11(H11) locus hotspot in the CHO-K1 genome using CRISPR/Cas9 technology, and a compliant CHO-CDbox cell platform was screened and obtained. The CHO-CDbox cell platform was transformed into a pool of EGFP-expressing cells using Cre/lox recombinase-mediated cassette exchange (RMCE) in only 2weeks, and this expression remained stable for at least 75 generations without the need for drug stress. Subsequently, we used the Dre/rox system to directly eliminate the EGFP gene. In addition, two practical applications of the CHO-CDbox cell platform were presented. The first was the quick construction of the Pembrolizumab antibody stable expression strain, while the second was a protocol for the integration of surface-displayed and secreted antibodies on CHO cells. The previous research on site-specific integration of CHO cells has always focused on the single functionality of insertion of target genes. This newly developed CHO cell platform is expected to offer expanded applicability for protein production and gene function studies.

Abstract B041: Defining the Determinants of Immune Response in DNA Homologous Recombination Deficient Tumors

Abstract Immune checkpoint blockade (ICB) has shown unprecedented success in improving clinical outcomes for numerous cancer patients, but many patients still fail to respond to treatment. Studies have demonstrated that the tumor microenvironment, tumor mutational burden, and patient DNA Damage Repair (DDR) deficiency may play a key role in determining response to ICB. To better characterize this and define key determinants of ICB response, we created an isogenic knockout BRCA2 in a murine 4T1 metastatic TNBC background to model these differential clinical responses and to assess how these tumor-intrinsic programs influence the tumor immune microenvironment to poise tumors for immunotherapy response. Differential immune landscapes were found via scRNAseq at baseline and with ICB driven by tumor DDR status with notable differences in the myeloid compartment and CXCR3-expressing T cell, NK, pDC, DCs, and plasma cells. Differential expression and interferon stimulated gene (ISG) metagene analysis showed notable differences in ISG expression in DDR-deficient samples at baseline, particularly in clusters containing monocytes and macrophages. Additionally, differential expression of CXCL10 in monocytes/macrophages and assessment of ligand-receptor interactions by CellChat similarly showed enhanced contribution CXCL10-CXCR3. Parallel bulk RNAseq analysis showed increased production of T cell trafficking chemokines that would poise the tumor for response to ICB. Our lab established a genetic interferon reporter system to measure both tumor intrinsic interferon-driven inflammation in cis and in trans with tumor associated macrophages, which enables multidimensional tracking of bidirectional signaling between tumor cells and macrophages by flow. This system enabled screening of inhibitors and tumor knockout cell lines to identify that tumor-intrinsic cGAS/STING and IFNb1 reinforce production of tumor and myeloid CXCL10/11 to poise the tumor for response to ICB. Further assessment of tumor cell lines via cellular fractionation experiments identified the presence of both gDNA and R-loops in the cytoplasm of BRCA2-mutant mammary cancer cell lines that may be detected as damage associated molecular patterns by tumor cells or myeloid cells in the tumor bed. Future studies aim at validating this DDR-driven production of IFNb1 as a driver of T cell chemokines in vivo; tumor-intrinsic knockout of IFNb1 and other T cell promoting chemokines (CXCL10/11/CCL5) to assess tumor or myeloid driven nature of this response to ICB and depletion of myeloid cells are being performed in parallel to assess contribution of tumor and myeloid cells, respectively. Taken together, our results have important implications for understanding the key drivers of ICB response in the tumor microenvironment and how patient-intrinsic mutations differentially poise the microenvironment for immunotherapy response. Citation Format: Natalie Vaninov, Robert Samstein. Defining the Determinants of Immune Response in DNA Homologous Recombination Deficient Tumors [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Tumor Immunology and Immunotherapy; 2023 Oct 1-4; Toronto, Ontario, Canada. Philadelphia (PA): AACR; Cancer Immunol Res 2023;11(12 Suppl):Abstract nr B041.

Abstract A136: Inferring therapeutic vulnerability within tumors through integration of pan-cancer cell line and single-cell transcriptomic profiles

Abstract Single-cell sequencing techniques have greatly advanced our current understanding of intratumoral heterogeneity through identifying tumor subpopulations with distinct biologies and therapeutic responses. However, translating biological differences into treatment strategies is challenging, as we still lack tools to facilitate efficient drug discovery that tackles heterogeneous tumors. One key step in development of such approaches centers around accurate prediction of drug response at the single-cell level to offer therapeutic options to specific cell subpopulations. Here, we present a transparent computational framework (nicknamed scIDUC) to predict therapeutic efficacies on an individual-cell basis by integrating single-cell transcriptomic profiles with large, data-rich pan-cancer cell line screening datasets. Our method detects shared expression patterns between the two data sources and utilizes such information to project cellular drug response. This method achieves high accuracy, with predicted sensitivities easily able to separate cells into their true cellular drug resistance status as measured by effect size (Cohen's d &amp;gt; 1.0); this holds when using single-cell RNA-seq from both cell line and in vivo models. More importantly, we examine our method’s utility with three distinct prospective tests, and in each our predicted results are accurate and mirrored biological expectations. In the first two tests, we investigated predicting drugs for cell subpopulations that are resistant to standard-of-care (SOC) therapies due to intrinsic resistance or effects of tumor microenvironments. In both, our results showed high consistency with results from the original studies. In the third test, we generated SOC therapy resistant cell lines, used scIDUC to identify drugs predicted effective on the resistant line, and validated the predictions with in vitro experiments. Together, scIDUC quickly and directly translates scRNA-seq data into meaningful cellular drug response for individual cells, displaying the potential to be used by researchers as a first-line tool for nuanced and heterogeneity-aware drug discovery. Citation Format: Weijie Zhang, Danielle Maeser, Adam Lee, Yingbo Huang, Robert F Gruener, Sampreeti Jena, Anand G Patel, R. Stephanie Huang. Inferring therapeutic vulnerability within tumors through integration of pan-cancer cell line and single-cell transcriptomic profiles [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2023 Oct 11-15; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2023;22(12 Suppl):Abstract nr A136.

Abstract LB_A21: NCI60 HTS384: The next phase of the NCI60 screen

Abstract The NCI60 human tumor cell line screen, which has operated as a free service to the cancer research community for over 20 years, has proven to be an important drug discovery tool. The screen has operated using cutting-edge technology of the 1990s including a 96-well microplate format, a 2-day drug exposure period and following cell fixation, and a visible absorbance endpoint by the protein-staining dye sulforhodamine B (SRB). The SRB endpoint determines cell density based on protein content with a lower resolution limit of 1,000 – 2,000 cells. This report describes the next phase of this important cancer research tool, the NCI60 HTS384 screen format. While the cell lines remain the same, the screen is performed using 384-well microplates with a 3-day compound exposure period and a CellTiter-Glo luminescence endpoint. The CellTiter-Glo endpoint determines cell number based upon ATP content with a lower resolution limit of &amp;lt;10 cells. To develop a baseline comparison, data were generated from a library of 1,000 FDA-approved and investigational small molecule anticancer agents by the two assay formats. Although the mean-graph patterns and COMPARE analyses are very similar for many compounds, there is a divergence between the two screens for some molecular targets. Within subsets of approved and investigational agents directed toward the same target (e.g., EGFR, BRAF/MEK/ERK, and PI3K inhibitors), the clustering of compound data from the NCI60 HTS384 screen is qualitatively similar to the clustering of compound data from the classic NCI60 SRB screen and there is a strong COMPARE correlation between the two screens. However, for some drug targets there is very little or no clustering between data from the two screens, although the clustering within each screen is moderate-to-good. Examples include mTOR rapalogs, BET bromodomain and NAMPRTase inhibitors. Differences in data between the two screens might be partially attributed to increases in the assay endpoint sensitivity and compound exposure time of the NCI60 HTS384 assay. Compounds submitted for NCI60 screening after September 30, 2023, will be tested in the modernized NCI60 HTS384 format. This project was funded in part with federal funds from the NCI, NIH, under contract no. HHSN261201500003I and by the Intramural Research Program of the NIH, National Cancer Institute. Citation Format: Beverly A Teicher, Mark W Kunkel, Joel Morris, Ronald C Taylor, Thomas S Dexheimer, Nathan P Coussens, James H Doroshow. NCI60 HTS384: The next phase of the NCI60 screen [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2023 Oct 11-15; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2023;22(12 Suppl):Abstract nr LB_A21.

2‐Aryl Imidazolium Salts as Potential Bladder Cancer Chemotherapeutic Candidates

AbstractAs potential intravesical exfoliants for the treatment of bladder cancer, four new imdiazolium salts have been prepared. These compounds are modified on the nitrogen positions of the 4,5‐diphenyl‐1H‐imidazole or benzimidazole rings with naphthalen‐2‐ylmethyl groups, and the 2 positions on the azolium cations are occupied by arene rings. The 4,5‐diphenyl‐1H‐imidazole and benzimidazole precursors are produced from 1,2‐phenylenediamine and benzil respectively, and the imidazolium cations are generated from sequential alkylation of the nitrogen positions using 2‐bromomethyl naphthalene. The four imidazolium salts were screened for their chemotherapuetic activity using the NCI‐60 Human Tumour Cell Line Screen, and we observed that the cytotoxicity was more dependent on the identity of the imidazolium group (benzimidazolium versus diphenylimidazolium) rather than on the 2‐aryl position structure.

Abstract C176: Cancer-specific AI identifies multi-modal biomarkers of therapeutic response for 1,951 drugs including TNG348, a highly selective USP1 inhibitor

Abstract Understanding drug mechanism-of-action and improving clinical response rates are critical challenges to cancer therapeutic development. We designed a deep learning framework that learns molecular mechanisms of drug sensitivity and resistance from preclinical and/or clinical drug response data. The trained model can then generate biomarker strategies, suggest companion diagnostics, and propose rational drug-drug combinations that increase or restore drug sensitivity. The key innovation is the model’s architecture; the framework learns an embedding that reflects an interpretable relationship between a cellular process and drug response by emulating a tumor cell’s proteomic architecture. We applied the framework to learn genetic determinants of response for 1,951 drugs and 17,386 CRISPR gene knockouts screened in human cancer cell lines. The learned embeddings reflected known and novel mechanisms. RAF inhibitors, MEK inhibitors, and BRAF CRISPR knockout colocalized (P=1.7×10-11) due to their shared dependency on hyperactivation of RAF signaling. CDK4/6 inhibitor efficacy was strongly associated with a set of physically interacting proteins that control G1-to-S transition (P=5.6×10-12). Validating the clinical applicability of the finding, the trained model stratified patient response in a retrospective analysis of 70 ER+ metastatic breast cancer patients treated with the CDK4/6 inhibitor palbociclib (P=0.05, log-rank predicted responsive vs. non-responsive). As a case study, we examined mechanisms of sensitivity and resistance to TNG348 – a potent and highly selective USP1 inhibitor – in non-small cell lung cancer (NSCLC). By integrating genetics, gene expression, and CRISPR screens, the model identified seven genes whose expression explained &amp;gt;50% of the variation in response to TNG348 across 32 held-out NSCLC cell lines (Spearman R=0.74; P=8.7×10-7). The findings extend TNG348’s known relationship with homologous repair to additional protein complexes involved in cell cycle and DNA repair. The model leveraged the new associations to generate a six-biomarker inclusion-exclusion logic that captures both a large population (31.4% of TCGA lung adenocarcinoma patients meet inclusion criteria) and substantially enriches for strong response amongst preclinical models (OR=14.8, P=3.0×10-3). Finally, we asked the model to design rational combinations to overcome acquired resistance mechanisms. Proposed combinations included clinically successful combinations such as RAF+MEK inhibitors and were enriched for synergistic matches in a cell line screen of 986 drug pairings (Spearman rho P=6.6×10-4). These results highlight the promise of explainable AI to learn complex cellular mechanisms-of-action and generate non-obvious biomarker hypothesis and rational drug combination strategies for clinical development. Citation Format: Adam Yaari, Lee McDaniel, Antoine Simoneau, Samuel Meier, Oliver Priebe, Eduardo Farias, Alan Huang, Jannik Andersen, Yi Yu, Maxwell Sherman. Cancer-specific AI identifies multi-modal biomarkers of therapeutic response for 1,951 drugs including TNG348, a highly selective USP1 inhibitor [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2023 Oct 11-15; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2023;22(12 Suppl):Abstract nr C176.

Abstract PR004: PRISM high-throughput screening of antibody-drug conjugates uncovers clinically relevant targets

Abstract Antibody-drug conjugates (ADCs) are a rapidly emerging class of cancer therapeutics composed of an antibody scaffold, a cytotoxic payload, and a joining linker. They exhibit properties of both biologics and small molecule inhibitors, can exert activity through multiple mechanisms, and some are capable of inducing target-independent cell killing via a phenomenon known as the bystander effect. Due to this inherent structural and biological complexity, cell-based functional screening is an essential component of ADC development. However, the majority of preclinical in vitro studies for ADCs are limited to a small number of cell lines and thus provide minimal insights into their activity. High-throughput drug-screening platforms, such as PRISM (Profiling Relative Inhibition Simultaneously in Mixtures), that have been highly effective in characterizing small molecule inhibitors could facilitate characterization of ADCs at an unprecedented scale and provide more comprehensive insights into ADC mechanisms, target specificity, and potency. To test the suitability of the PRISM platform for ADC screening, we investigated several ADCs conjugated to the HER2-targeting antibody trastuzumab that differed in terms of linker type (cleavable and non-cleavable), payload class, and drug-to-antibody ratio. ADCs were screened in parallel with controls including unconjugated antibodies and free payloads across a panel of nearly 500 cell lines consisting of diverse cell lineages. We explored whether ADC selectivity was associated with target antigen expression and found a clear association between cell line sensitivity and ERBB2 (HER2) gene expression, with the strongest effects on viability observed in a subset of cell lines overexpressing ERBB2. Univariate biomarker analysis identified ERBB2 gene expression as a significantly correlated feature of sensitivity to all HER2-targeting ADCs. Notably, this correlation was retained across ADCs with payloads capable of exerting bystander killing of cells with low ERBB2 expression. Upon further comparison of biomarker profiles across different ADCs, established relationships between payload resistance and ABCB1 efflux pumps emerged for a subset of ADC payloads. These results indicate that PRISM pooled cell line screening is able to distinguish ADCs with different payload characteristics while maintaining selectivity of the primary target. Overall, our study highlights how PRISM can be leveraged to analyze ADC activity and uncover clinically relevant targets. This approach can aid in evaluating potential ADC candidates by providing a better understanding of ADC mechanisms of action and how their unique structural components drive cytotoxic activity. Citation Format: Jillian N Eskra, Ellen Nguyen, Aydin Golabi, Shiker Nair, Antonella Masciotti, Anthony Fazio, Mustafa Kocak, Melissa Ronan, Matthew G Rees, Jennifer A Roth. PRISM high-throughput screening of antibody-drug conjugates uncovers clinically relevant targets [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2023 Oct 11-15; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2023;22(12 Suppl):Abstract nr PR004.

In Vivo Modeling of T-Cell Acute Lymphoblastic Leukemia Reveals Synergistic Oncogenic Pathways and Bcl11b Haploinsufficiency As a Potential Therapeutic Vulnerability

For decades, T-ALL models were based on retroviral overexpression of NOTCH1 in murine hematopoietic stem cells (HSCs). This induces aggressive T-ALL but does not reflect the genetic and clinical heterogeneity of the disease. Large-scale genomic analyses have mapped the detailed landscape of T-ALL driver events over recent years, but how these diverse oncogenic events interact in T-ALL pathogenesis and which vulnerabilities this creates remains incompletely explored. This is in part due to the lack of appropriate mouse models that recapitulate the complex disease biology and are suitable to identify and study potential therapeutic targets and interventions in a more realistic context. We modeled the genetic heterogeneity of T-ALL through in vivo multiplexing of gain-of-function and loss-of-function oncogenic events known to drive the disease. We developed a novel mouse strain that expresses Cre-recombinase and an inducible Cas9 only in the thymus ( LSL.Cas9 x Lck-cre) along with a lentiviral vector system into which T-ALL specific oncogenes are cloned in antisense flanked by LOX66/71 sites. Gain-of-function events, such as the overexpression of transcriptional regulators ( Tlx1, Tal1, or Lmo2) and signal transducers ( NRAS G12D, PIK3CD E1021K) were induced in developing thymocytes by Cre-mediated inversion together CRISPR-Cas9 editing of recurrent T-ALL loss-of-function drivers. This combinatorial approach allowed to probe more than 2000 possible oncogenic mutational combinations in vivo. Transplantation of HSCs from LSL.Cas9 x Lck-cre mice transduced with our vector system into sublethally irradiated mice led to T-ALL development with different morphology and disease phenotypes. Developing leukemia harbored reproducible phenotypes, covering very immature CD4-CD8-CD25+CD44+ T-ALL in the Tlx1_NRAS subgroup, in contrast to the classical cortical CD4+CD8+CD3- phenotype (Tlx1_PIK3CD subgroup), as well as mature single-positive CD4+CD3+ or CD8+CD3+ blasts (Tal1_Lmo2 subgroup). We identified Tlx1, Tal1, Lmo2, NRAS G12Dand PIK3CD E1021Kas the main driver of disease phenotype and gene expression profile, whereas the loss-of-function mutations contributed to and accelerated the onset of the disease. Tumor-derived cell lines could be easily established and kept their phenotypic characteristics and mutation patterns. The resulting T-ALL tumors contained up to eight different disease-relevant genetic alterations, thus recapitulating the genetic complexity in humans and overcoming the long latency for spontaneous mutations in less complex mouse models. Among these, co-evolution of insertions and deletions in Notch1, Cdkn2a, Bcl11b, and Pten were among the most frequently observed mutations in definitive T-ALL, whereas others, e.g., in Dnm2, Phf6, Etv6, and Lef1, occurred more randomly. While gene editing by CRISPR-Cas9 mostly resulted in frame-shift mutations in both alleles, Bcl11b exhibited a strong haploinsufficient phenotype with exactly one knock-out and one intact allele in each case examined. Following this observation, further suppression of Bcl11b by shRNAs in Bcl11b +/- cell lines derived from our tumor models demonstrated rapid induction of cell death, which was more intense in T-ALL cells compared to non-malignant thymocytes. In addition, re-analysis of CRISPR-based loss-of-function screens of human hematopoietic cell lines retrieved from the Cancer Dependency Map confirmed BCLL11B as a strong selective dependency in human T-ALL. In summary, we model subgroup-specific T-ALL transformation through conditional and multiplexed in vivo gene editing and oncogene overexpression, thus creating a resource of novel mouse models that recapitulate the genetic and biological heterogeneity of the disease. Using our resource, we identified Bcl11b as a strong and selective context-specific dependency in T-ALL. Although Bcl11b is a transcription factor, our results could serve as an starting point for novel therapies that interfere with Bcl11b function, such as molecular glue degraders or protein interaction inhibitors.