specifically expressed in terminally exhausted cells. These data suggest the balance of Batf and Nr4a2 may modulate the enhancer landscape to promote terminal exhaustion, while hypoxia simultaneously promotes hypermethylation and gene repression. Conclusions Our study defines for the first time the features of epigenetic dysfunction in tumor-mediated T cell exhaustion and deepens our understanding of the epigenetic regulation of gene expression. These observations are the bases for future work that will elucidate that factors that drive progression towards terminal T cell exhaustion at the epigenetic level and identify novel therapeutic targets to restore effector function of tumor T cells and mediate tumor clearance. http://dx.doi.org/10.1136/jitc-2020-SITC2020.0518 519 DIACYLGLYCEROL KINASE z LIMITS IL-2-DEPENDENT CONTROL OF PD-1 EXPRESSION IN TUMOR- INFILTRATING T LYMPHOCYTES 1 Javier Arranz-Nicolas*, 1 Cristina Rodríguez-Rodríguez, 1 Rosa Liébana, 2 Judith Leitner, 1 Antonia Ávila-Flores, 2 Peter Steinberger, 1 Isabel Mérida. 1 Spanish National Centre for Biotechnology (CNB-CSIC), Madrid, Spain; 2 Medical University of Vienna, Vienna, Austria Background Tumors evade T cell responses targeting them through the upregulation of tolerance-inducing mechanisms. One of the best characterized is that of PD-1/PD-1L engage- ment, that in healthy CD8+ T cells limits cytotoxic responses against self-antigens and that tumors employ to neutralize T cell attack. Antibody-based therapies aimed to block the PD-1/ PD-1L axis have rendered notable results, but most patients eventually develop resistance. This failure is attributed to CD8 + T cells achieving an exhausted phenotype where recovery is hardly feasible. The dysfunctional phenotype of tumor-infiltrat- ing T cells is largely triggered by the unbalance of diacylgly- cerol (DAG)- and Ca2+-regulated signals that results in alteration of the transcriptional T cell program. DAG kinase (DGK) z-dependent DAG consumption contributes to hypo- functional T cell states while DGKz deficiency facilitates tumor rejection in mice without apparent adverse autoimmune effects. In spite of its therapeutic potential, little is known about DGKz function in human T cells and there are not iso- form-specific inhibitors targeting this DGK isoform. Methods Here we used of a human triple parameter reporter (TPR) cell line to examine the consequences of DGKz deple- tion in the transcriptional restriction imposed by PD-1 liga- tion. We also investigated the effect of DGKz deficiency in the expression dynamics of PD-1, as well as the impact of the absence of this DGK isoform in the in vivo growth of a MC38 adenocarcinoma cell line. Results We demonstrate that DGKz depletion enhances DAG- regulated transcriptional programs, favoring IL-2 production and limiting PD-1 expression. Diminished PD-1 expression and enhanced expansion of cytotoxic CD8+ T cell popula- tions is also observed even in the context of immunosuppres- sive milieus and correlates with the failure of MC38 adenocarcinoma cells to form tumors in DGKz-deficient mice. Conclusions Our results suggest the relevance of DGKz as a therapeutic target on its own as well as a biomarker of CD8 + T cell dysfunctional states. http://dx.doi.org/10.1136/jitc-2020-SITC2020.0519 520 THE IMMUNE LANDSCAPE OF PEDIATRIC TUMORS 1 Shimaa Sherif*, 2 Jessica Roelands, 2 William Mifsud, 2 Blessing Dason, 2 Darawan Rinchai, 2 Adrian Charles, 2 Ayman Saleh, 2 Chiara Cugno, 2 Khalid Fakhro, 1 Borbala Mifsud, 2 Davide Bedognetti, 2 Wouter Hendrickx. 1 Hamad Bin Khalifa University, Doha, Qatar; 2 Sidra Medicine, Doha, Qatar Background It is now well established that the immune system has a substantial role in controlling cancer growth and pro- gression. Immunotherapy is quickly coming to the forefront of cancer treatment,however the implementation of immunother- apy in pediatric solid cancers, which classically display a low mutational load, is hindered by insufficient understanding of the determinants of cancer immune responsiveness in children. In order to better understand tumor-host interplay, we sought to characterize solid pediatric cancers based on immunological parameters 1 using analytes extracted from gene expression data. Methods We performed single sample GeneSet Enrichment Analysis for 105 immune signatures previously described on 5pediatric tumors (410patients) from TARGET dataset 1 to identify coherent signature modules. Then we clustered sam- ples according to representative signatures 1 and compared sur- vival across clusters. We completed the analysis by analyzing the enrichment of immune subpopulations and the expression of the immune checkpoints. The degree of dysregulation of Abstract 520 Figure 1 Immune subtypes of pediatric solid tumors A. Spearman Correlation matrix of 105 cancer immune signatures showing 5 main modules. B. Spearman’s correlation of the 105 cancer immune signatures, identifies separation of the 5 immune signatures in different clusters C. Distribution of cancer types within immune subtypes. The percentage of samples belonging to each tumor is shown in colors. D. Distribution of immune subtypes within TARGET pediatric tumors. The percentage of samples belonging to each immune subtype is shown in colors. E. Distributions of signature scores within the six immune subtypes (rows), with dashed line indicating the median. F. Kaplan-Meier OS curve for the 6 immune subtypes (S1-S6) showing different outcomes. Abstracts A318 J Immunother Cancer 2020;8(Suppl 3):A1–A559 on April 17, 2022 by guest. Protected by copyright. http://jitc.bmj.com/ J Immunother Cancer: first published as 10.1136/jitc-2020-SITC2020.0520 on 10 December 2020. Downloaded from
Transcript
specifically expressed in terminally exhausted cells. These datasuggest the balance of Batf and Nr4a2 may modulate theenhancer landscape to promote terminal exhaustion, whilehypoxia simultaneously promotes hypermethylation and generepression.Conclusions Our study defines for the first time the featuresof epigenetic dysfunction in tumor-mediated T cell exhaustionand deepens our understanding of the epigenetic regulation ofgene expression. These observations are the bases for futurework that will elucidate that factors that drive progressiontowards terminal T cell exhaustion at the epigenetic level andidentify novel therapeutic targets to restore effector functionof tumor T cells and mediate tumor clearance.
http://dx.doi.org/10.1136/jitc-2020-SITC2020.0518
519 DIACYLGLYCEROL KINASE z LIMITS IL-2-DEPENDENTCONTROL OF PD-1 EXPRESSION IN TUMOR-INFILTRATING T LYMPHOCYTES
1Javier Arranz-Nicolas*, 1Cristina Rodríguez-Rodríguez, 1Rosa Liébana, 2Judith Leitner,1Antonia Ávila-Flores, 2Peter Steinberger, 1Isabel Mérida. 1Spanish National Centre forBiotechnology (CNB-CSIC), Madrid, Spain; 2Medical University of Vienna, Vienna, Austria
Background Tumors evade T cell responses targeting themthrough the upregulation of tolerance-inducing mechanisms.One of the best characterized is that of PD-1/PD-1L engage-ment, that in healthy CD8+ T cells limits cytotoxic responsesagainst self-antigens and that tumors employ to neutralize Tcell attack. Antibody-based therapies aimed to block the PD-1/PD-1L axis have rendered notable results, but most patientseventually develop resistance. This failure is attributed to CD8+ T cells achieving an exhausted phenotype where recovery ishardly feasible. The dysfunctional phenotype of tumor-infiltrat-ing T cells is largely triggered by the unbalance of diacylgly-cerol (DAG)- and Ca2+-regulated signals that results inalteration of the transcriptional T cell program. DAG kinase(DGK) z-dependent DAG consumption contributes to hypo-functional T cell states while DGKz deficiency facilitates tumorrejection in mice without apparent adverse autoimmuneeffects. In spite of its therapeutic potential, little is knownabout DGKz function in human T cells and there are not iso-form-specific inhibitors targeting this DGK isoform.Methods Here we used of a human triple parameter reporter(TPR) cell line to examine the consequences of DGKz deple-tion in the transcriptional restriction imposed by PD-1 liga-tion. We also investigated the effect of DGKz deficiency inthe expression dynamics of PD-1, as well as the impact of theabsence of this DGK isoform in the in vivo growth of aMC38 adenocarcinoma cell line.Results We demonstrate that DGKz depletion enhances DAG-regulated transcriptional programs, favoring IL-2 productionand limiting PD-1 expression. Diminished PD-1 expressionand enhanced expansion of cytotoxic CD8+ T cell popula-tions is also observed even in the context of immunosuppres-sive milieus and correlates with the failure of MC38adenocarcinoma cells to form tumors in DGKz-deficient mice.Conclusions Our results suggest the relevance of DGKz as atherapeutic target on its own as well as a biomarker of CD8+ T cell dysfunctional states.
Background It is now well established that the immune systemhas a substantial role in controlling cancer growth and pro-gression. Immunotherapy is quickly coming to the forefront ofcancer treatment,however the implementation of immunother-apy in pediatric solid cancers, which classically display a lowmutational load, is hindered by insufficient understanding ofthe determinants of cancer immune responsiveness in children.In order to better understand tumor-host interplay, we soughtto characterize solid pediatric cancers based on immunologicalparameters1 using analytes extracted from gene expressiondata.Methods We performed single sample GeneSet EnrichmentAnalysis for 105 immune signatures previously described on5pediatric tumors (410patients) from TARGET dataset1 toidentify coherent signature modules. Then we clustered sam-ples according to representative signatures1 and compared sur-vival across clusters. We completed the analysis by analyzingthe enrichment of immune subpopulations and the expressionof the immune checkpoints. The degree of dysregulation of
Abstract 520 Figure 1 Immune subtypes of pediatric solid tumorsA. Spearman Correlation matrix of 105 cancer immune signaturesshowing 5 main modules.B. Spearman’s correlation of the 105 cancer immune signatures,identifies separation of the 5 immune signatures in different clustersC. Distribution of cancer types within immune subtypes. The percentageof samples belonging to each tumor is shown in colors.D. Distribution of immune subtypes within TARGET pediatric tumors.The percentage of samples belonging to each immune subtype is shownin colors.E. Distributions of signature scores within the six immune subtypes(rows), with dashed line indicating the median.F. Kaplan-Meier OS curve for the 6 immune subtypes (S1-S6) showingdifferent outcomes.
Abstracts
A318 J Immunother Cancer 2020;8(Suppl 3):A1–A559
on April 17, 2022 by guest. P
rotected by copyright.http://jitc.bm
j.com/
J Imm
unother Cancer: first published as 10.1136/jitc-2020-S
oncogenic pathways was also assessed. The performance ofpreviously identified immune signatures as the ImmunologicConstant of Rejection(2,3),which captures an active Th1/cyto-toxic response associated with favorable prognosis and respon-siveness to immunotherapy, was also checked within eachtumor subtype.Results We found 5main modules, in agreement with resultsobtained in adult solid tumors:Wound Healing,TGF-B signal-ing,IFN-G signaling, Macrophages and Lymphocytes (figure1). These 5 modules clustered pediatric patients into 6immune subtypes S1-S6 with distinct survival (S2vsS4,p=0.0044, adjusted for cancer type),S2 cluster has the bestoverall survival and characterized by low enrichment ofwound healing signature, high Th1, low Th2 and highexpression of HLA 1 and HLA2, while the opposite holdstrue for cluster S4 with the worst survival and highestenrichment of wound healing signature, high Th2, and lowTh1.The S6 cluster is characterized by highest enrichment oflymphocyte signature, the highest expression of immunecheckpoints accompanied by elevated expression of exhaus-tion markers, and an unpolarized immune response with highabundance of macrophages. Additionally, pan-cancer, theupregulation of WNT-Beta catenin pathway is associated withadverse outcome and lack of T-cell infiltration. In the per-cancer analysis, ICR is associated with better survival inosteosarcoma and with worse survival in Wilms’ tumors,simi-larly with what observed in adult kidney’s cancer despite thedifferent embryological origin.Conclusions We demonstrated that pediatric solid cancers canbe classified according to their immune disposition, unveilingunexpected similarity with adults’ tumors.Immunologicalparameters might be explored to refine diagnostic and prog-nostic biomarkers and to identify potential immune responsivetumors. This is the first pan-cancer immunogenomic analysisin children.
REFERENCES1. Thorsson V, Gibbs DL, Brown SD, et al. The immune landscape of cancer. Immun-
ity 2018. 48(4):812–830.2.. Roelands J, Hendrickx W, et al. ‘Oncogenic states dictate the prognostic and pre-
dictive connotations of intratumoral immune response.’ Journal for immunotherapyof cancer 2020;vol. 8:1.
3. Galon J, Angell HK, Bedognetti D, Marincola FM. The continuum of cancer immu-nosurveillance: prognostic, predictive, and mechanistic signatures. Immunity2013;39:11–26.
http://dx.doi.org/10.1136/jitc-2020-SITC2020.0520
521 IMMUNE REGULATORY METABOLITES IN HUMANOVARIAN CANCER
Background Immune regulatory metabolites are key features ofthe tumor microenvironment (TME), yet with a few excep-tions, their identities remain largely unknown. Importantly, lit-tle is known about the heterogeneity of metabolites that arepresent or absent in specimens from human tumors andimmune compartments.Methods Here, we profiled tumor and T cells from tumor andascites of patients with high-grade serous carcinoma (HGSC)
to uncover the metabolomes of these distinct TME compart-ments. We devised a stringent and robust protocol to enrichcell populations from surgically resected samples in patientswith HGSC. We conducted mass spectrometry-based analysisand developed machine learning tools to highlight novelmetabolites that are present in different cellular lineages ofthe tumor.Results Cells within the ascites and tumor had pervasivemetabolite differences, with a striking enrichment in 1-meth-ylnicotinamide (MNA) in T cells infiltrating the tumor com-pared to ascites. Despite the elevated levels of MNA in Tcells, the expression of nicotinamide N-methyltransferase, theenzyme that catalyzes the transfer of a methyl group from S-adenosylmethionine to nicotinamide, was restricted to fibro-blasts and tumor cells. T cells treated with MNA stimulatedsecretion of the tumor promoting cytokine tumor necrosisfactor alpha.Conclusions Our studies provide the first catalogue of metabo-lites in patient-derived tumors and T cells. We found thatTME-derived MNA contributes to the immune modulation ofT cells and represents a potential immunotherapy target totreat human cancer.Ethics Approval This study was approved by the University ofBritish Columbia and BC Cancer Research Ethics Board (H07-00463).Consent Written informed consent was obtained from thepatient to use the results of this study for educational pur-poses including publications. A copy of the written consentis on file and available for review by the Editor of thisjournal.
http://dx.doi.org/10.1136/jitc-2020-SITC2020.0521
522 METABOLIC REQUISITES FOR T CELL PROTEINTRANSLATION IN TUMORS
Katie Hurst, Megan Tennant*, Alex Andrews, Lee Leddy, David Neskey, Lauren Ball,Jessica Thaxton. Medical University of South Carolina, Charleston, SC, USA
Background T cells are a secretory immune subset with thecapacity to control solid tumors. Protein translation is of para-mount importance in CD8 T cells, controlling proliferation,stimulation and lineage fate.Methods Herein, we used both the fluorescent analogue ofmethionine homopropargylglycine (HPG) incorporation assayand O-propargyl-puromycin (OPP) method which enters theA-site of the ribosome and effectively labels and terminatesnascent polypeptide chains to monitor protein synthesis inmouse and human tumors. Moreover, we employed label freequantitative proteomics (LFQ), lipidomics, metabolic analysis,and in vivo animal modeling to elucidate mechanisms of pro-tein translation in antitumor immunity.Results We found that canonical protein synthesis is restrictedin endogenous CD8 tumor infiltrating lymphocytes (TILs) bythe tumor microenvironment (TME). Proteomic analysisrevealed that gluconeogenesis and B-oxidation of fatty acids(FAO) were upregulated in CD8 T cells under tumor stressbut these metabolic sources were unable to support translationin the TME. Further, we discovered that glucose metabolismand mammalian target of rapamycin complex 1 (mTORC1)preferentially hinder protein synthesis in CD8 TILs. Thesedata enabled the discovery that proteasomal protein
Abstracts
J Immunother Cancer 2020;8(Suppl 3):A1–A559 A319
on April 17, 2022 by guest. P
rotected by copyright.http://jitc.bm
j.com/
J Imm
unother Cancer: first published as 10.1136/jitc-2020-S
