Periodic Reporting for period 4 - TECNEC (Preclinical concept validation of tumor endothelial cell metabolism for novel anti-angiogenic therapy)
Berichtszeitraum: 2022-05-01 bis 2022-10-31
This project aimed to characterize the metabolic changes fueling growth of tumor endothelial cells (TECs) versus normal healthy endothelial cells (NECs), and to identify & validate metabolic genes, deregulated in TECs, as targets for anti-angiogenic cancer therapy.
Blood vessel formation (angiogenesis) promotes tumor growth/metastasis. Anti-angiogenesis therapy (AAT) via blockade of the key angiogenesis factor Vascular Endothelial Growth Factor (VEGF) is clinically approved for certain cancers but suffers resistance, toxicity and insufficient efficacy. AAT requires a “face-lift” to better understand the mechanisms of VEGF’s limitations and to develop improved AATs (no longer based on targeting VEGF, but on novel angiogenic molecules). Among the reasons of VEGF’s limitations, several mechanisms have been overlooked, on which we have focused in this ERC project.
OVERALL OBJECTIVES AND KEY RESULTS
• Heterogeneity of TECs at the single cell level. Several scRNA-sequencing analyses revealed that TECs from lung cancer exhibit a previously unrecognized phenotypic heterogeneity and that VEGF has different efficacy in targeting different TEC phenotypes. Also, TECs have roles in tumor immune surveillance (more important than previously recognized) (Cell Metab 2020; Cancer Cell 2020; Cell 2020; Nat Commun. 2022; Comput Struct Biotechnol J., 2022).
• ECs in tumors are immunosuppressive, which impairs infiltration of immune cells (ICs) in the tumor, a primary reason of why so many cancer patients are showing resistance against traditional immunotherapy, targeting cancer and ICs. The lab focused on discovering and silencing immunosuppressive genes expressed in ECs in tumours (IMECs) to render the immunosuppressive endothelium immunostimulatory. Such alternative immunotherapy has been overlooked, yet promises to increase the efficiency of / overcome the resistance to current traditional immunotherapy (Nat Rev Immunol 2022).
• Structural and functional abnormalities of TECs and tumor vessels promote metastasis and impair the delivery and efficacy of chemo- and immunotherapy. Inducing tumor vessel normalization is emerging as a new AAT; computational approaches are used to identify targets contributing to tumor vessel disorganization.
• We performed a study on tumor-vessel co-option as a resistance mechanism against VEGF-targeted AAT. Tumor vessel co-option is poorly understood, yet it is a resistance mechanism against AAT. The heterogeneity of co-opted ECs and pericytes, co-opting cancer and myeloid cells in tumors growing via vessel co-option, has been investigated at the single-cell level (Cell Reports 2021; Cardiovasc Res 2022; Front Oncol 2022 and STAR Protoc. 2022).
Blood vessel formation (angiogenesis) promotes tumor growth/metastasis. Anti-angiogenesis therapy (AAT) via blockade of the key angiogenesis factor Vascular Endothelial Growth Factor (VEGF) is clinically approved for certain cancers but suffers resistance, toxicity and insufficient efficacy. AAT requires a “face-lift” to better understand the mechanisms of VEGF’s limitations and to develop improved AATs (no longer based on targeting VEGF, but on novel angiogenic molecules). Among the reasons of VEGF’s limitations, several mechanisms have been overlooked, on which we have focused in this ERC project.
OVERALL OBJECTIVES AND KEY RESULTS
• Heterogeneity of TECs at the single cell level. Several scRNA-sequencing analyses revealed that TECs from lung cancer exhibit a previously unrecognized phenotypic heterogeneity and that VEGF has different efficacy in targeting different TEC phenotypes. Also, TECs have roles in tumor immune surveillance (more important than previously recognized) (Cell Metab 2020; Cancer Cell 2020; Cell 2020; Nat Commun. 2022; Comput Struct Biotechnol J., 2022).
• ECs in tumors are immunosuppressive, which impairs infiltration of immune cells (ICs) in the tumor, a primary reason of why so many cancer patients are showing resistance against traditional immunotherapy, targeting cancer and ICs. The lab focused on discovering and silencing immunosuppressive genes expressed in ECs in tumours (IMECs) to render the immunosuppressive endothelium immunostimulatory. Such alternative immunotherapy has been overlooked, yet promises to increase the efficiency of / overcome the resistance to current traditional immunotherapy (Nat Rev Immunol 2022).
• Structural and functional abnormalities of TECs and tumor vessels promote metastasis and impair the delivery and efficacy of chemo- and immunotherapy. Inducing tumor vessel normalization is emerging as a new AAT; computational approaches are used to identify targets contributing to tumor vessel disorganization.
• We performed a study on tumor-vessel co-option as a resistance mechanism against VEGF-targeted AAT. Tumor vessel co-option is poorly understood, yet it is a resistance mechanism against AAT. The heterogeneity of co-opted ECs and pericytes, co-opting cancer and myeloid cells in tumors growing via vessel co-option, has been investigated at the single-cell level (Cell Reports 2021; Cardiovasc Res 2022; Front Oncol 2022 and STAR Protoc. 2022).
LONG-TERM VISION OF THE PROJECT
Discovering new targets driving patho-biological processes in human ECs and functionally validate them in mouse models, in order to develop new therapeutic strategies (“from the bench to the bedside”).
• TARGET DISCOVERY by studying TEC heterogeneity. We utilized (amongst other techniques) unbiased multi-omics, including single cell analyses, in combination with state-of-the-art (meta-analysis) bioinformatics methods (BIOMEX, EC-GEM, EndoDB, etc.) as described in our studies: Rohlenova et al. Cell Metab 2020; Goveia et al. Cancer Cell 2020; Kalucka et al. Cell 2020. We have already identified multiple targets in TECs from lung cancer using this approach, which now need functional validation before proceeding to selection for drug development. By focusing on genes that, by analysing scRNA-seq datasets were found to be upregulated in proliferating ECs and were conserved in both patients and mouse models from different disorders that are characterized by excessive angiogenesis, and by combining the scRNAseq analyses with other complementary bioinformatics meta-analyses, we were able to identify previously unrecognized angiogenic targets (Cell Metab 2020; Cancer Cell 2020). As a next step, we will use similar approaches to identify targets, involved in EC heterogeneity in other human cancer types.
• “SMARTER” DISCOVERY & SELECTION OF TARGETS: An estimated third of the human coding genome contains (up to 6,000) “mystery” genes, lacking functional annotation and publications. We developed a novel Artificial Intelligence (AI)-based prototype tool (SCMYSTERYDENTIFIER) that uses classical classifier algorithms and single cell transcriptomics data to successfully predict immunosuppressive functions for “mystery” genes in ECs, never associated with alternative immunotherapy. Building further on this, we are developing an improved AI-based BRAIN-FOR-BIOTECH (BFBIO) tool that integrates multiple data modalities and graph neural networks to discover mystery immunosuppressive targets in ECs more accurately (BFBIO-V1.0).
• TARGET VALIDATION is a critical step in the process of drug development starting from target discovery to functionally characterizing a large number of candidates. Validation occurs at several levels. We developed a novel revolutionizing (“REVOLT”) strategy to generate EC-selective knockout mice. REVOLT is based on the combined use of: a) lipid nanoparticles with improved EC-selective targeting containing sgRNA; and b) endothelial Cas9 expressing mice. We will validate the role of target genes in vivo, by generating EC gene manipulated mice (using REVOLT) and phenotyping their tumor growth & response to traditional immunotherapy.
Discovering new targets driving patho-biological processes in human ECs and functionally validate them in mouse models, in order to develop new therapeutic strategies (“from the bench to the bedside”).
• TARGET DISCOVERY by studying TEC heterogeneity. We utilized (amongst other techniques) unbiased multi-omics, including single cell analyses, in combination with state-of-the-art (meta-analysis) bioinformatics methods (BIOMEX, EC-GEM, EndoDB, etc.) as described in our studies: Rohlenova et al. Cell Metab 2020; Goveia et al. Cancer Cell 2020; Kalucka et al. Cell 2020. We have already identified multiple targets in TECs from lung cancer using this approach, which now need functional validation before proceeding to selection for drug development. By focusing on genes that, by analysing scRNA-seq datasets were found to be upregulated in proliferating ECs and were conserved in both patients and mouse models from different disorders that are characterized by excessive angiogenesis, and by combining the scRNAseq analyses with other complementary bioinformatics meta-analyses, we were able to identify previously unrecognized angiogenic targets (Cell Metab 2020; Cancer Cell 2020). As a next step, we will use similar approaches to identify targets, involved in EC heterogeneity in other human cancer types.
• “SMARTER” DISCOVERY & SELECTION OF TARGETS: An estimated third of the human coding genome contains (up to 6,000) “mystery” genes, lacking functional annotation and publications. We developed a novel Artificial Intelligence (AI)-based prototype tool (SCMYSTERYDENTIFIER) that uses classical classifier algorithms and single cell transcriptomics data to successfully predict immunosuppressive functions for “mystery” genes in ECs, never associated with alternative immunotherapy. Building further on this, we are developing an improved AI-based BRAIN-FOR-BIOTECH (BFBIO) tool that integrates multiple data modalities and graph neural networks to discover mystery immunosuppressive targets in ECs more accurately (BFBIO-V1.0).
• TARGET VALIDATION is a critical step in the process of drug development starting from target discovery to functionally characterizing a large number of candidates. Validation occurs at several levels. We developed a novel revolutionizing (“REVOLT”) strategy to generate EC-selective knockout mice. REVOLT is based on the combined use of: a) lipid nanoparticles with improved EC-selective targeting containing sgRNA; and b) endothelial Cas9 expressing mice. We will validate the role of target genes in vivo, by generating EC gene manipulated mice (using REVOLT) and phenotyping their tumor growth & response to traditional immunotherapy.
PROGRESS & EXPECTED RESULTS
* We developed an EC-tailored genome scale computational metabolic model (EC-GEM) to in silico predict metabolic genes essential for EC growth. To lift the power of EC GEM modeling to a next level, we will construct GEMs specific for: (a) tumor ECs (TECs) from different tumor types to predict tumor-tailored TEC targets; (b) ECs subjected to classical anti-angiogenic therapy (AAT), which may induce metabolic adaptation and subsequent treatment escape, to predict so-called synthetic lethality targets that can be inhibited to overcome metabolic adaptation to AATs; and (c) specific TEC subtypes which initiate angiogenesis and possess a distinct metabolic profile.
* scRNA-seq analysis of freshly isolated TECs/NECs from lung cancer patients and mouse models identified distinct TEC clusters. We will further analyse whether TEC populations escape anti-angiogenic therapy due to an adapted metabolism. Hence, targeting the “escape” metabolic rewiring in resistant TEC subpopulations would make them again more vulnerable and therapy responsive.
* Immunotherapy (IT) is a highly promising anti-cancer therapy but is plagued by resistance caused by insufficient recruitment of ICs to the tumor stroma and by tolerance. To improve IT, we explore the immunosuppressive behaviour of the tumor endothelium and tune TECs to become phenotypically immunocompetent and to allow for improved T cell recruitment into the tumor stroma. To gain knowledge on the (patho)-biological role of immunomodulatory ECs (IMECs), we use state-of-the-art single cell omics and construct a Mega-Atlas of all publicly available single-cell transcriptome data of ECs (single-cell vasculome atlas).
* For a selected number of validated targets (including mystery genes), pharmacological tool compounds (monoclonal antibodies or nanobodies) might be generated (outsourced) and tested in mouse tumor models to increase the value for licensing to big pharma. Discovering and validating mystery targets offers a goldmine for academia & pharma.
* We developed an EC-tailored genome scale computational metabolic model (EC-GEM) to in silico predict metabolic genes essential for EC growth. To lift the power of EC GEM modeling to a next level, we will construct GEMs specific for: (a) tumor ECs (TECs) from different tumor types to predict tumor-tailored TEC targets; (b) ECs subjected to classical anti-angiogenic therapy (AAT), which may induce metabolic adaptation and subsequent treatment escape, to predict so-called synthetic lethality targets that can be inhibited to overcome metabolic adaptation to AATs; and (c) specific TEC subtypes which initiate angiogenesis and possess a distinct metabolic profile.
* scRNA-seq analysis of freshly isolated TECs/NECs from lung cancer patients and mouse models identified distinct TEC clusters. We will further analyse whether TEC populations escape anti-angiogenic therapy due to an adapted metabolism. Hence, targeting the “escape” metabolic rewiring in resistant TEC subpopulations would make them again more vulnerable and therapy responsive.
* Immunotherapy (IT) is a highly promising anti-cancer therapy but is plagued by resistance caused by insufficient recruitment of ICs to the tumor stroma and by tolerance. To improve IT, we explore the immunosuppressive behaviour of the tumor endothelium and tune TECs to become phenotypically immunocompetent and to allow for improved T cell recruitment into the tumor stroma. To gain knowledge on the (patho)-biological role of immunomodulatory ECs (IMECs), we use state-of-the-art single cell omics and construct a Mega-Atlas of all publicly available single-cell transcriptome data of ECs (single-cell vasculome atlas).
* For a selected number of validated targets (including mystery genes), pharmacological tool compounds (monoclonal antibodies or nanobodies) might be generated (outsourced) and tested in mouse tumor models to increase the value for licensing to big pharma. Discovering and validating mystery targets offers a goldmine for academia & pharma.