Cell Types In Tumor Microenvironment Research Articles

Abstract B023: Regulatory T (Treg) cell contributes to tumor cell dissemination via extracellular matrix (ECM) remodeling, which is driven by TAMs in an IFNg-dependent manner

Abstract Background: The extracellular matrix (ECM) is a highly organized non-cellular network consisting of structural proteins, growth factors, cytokines and other secreted molecules. The ECM is highly dynamic and is a critical player in the regulation of local invasion and metastasis. The ECM network is ever-changing and an important immunosuppressive component of the tumor microenvironment (TME). However, the reciprocal effects of the immune system on the tumor-associated ECM are elusive and largely unexplored. Regulatory T (Treg) cells function to enforce peripheral tolerance and are potent suppressors of tumor immunity. We have previously shown that Treg cells promote tumor growth in murine breast cancer models by favoring alternative activation of tumor-associated macrophages (TAMs) via suppression of IFN-g. Aims: In this work, we sought to assess whether Treg cell-mediated immune suppression is associated with changes in the ECM, and whether those ECM changes functionally impact metastatic dissemination. Further, we assessed the requirement of TAMs and IFN-g in the ECM remodeling. Design Methods: In this study, we used spontaneous and transplantable models of breast cancer susceptible to genetic ablation of Treg cells. We used a combination of histopathological analysis, tumor decellularization, bioengineering 3D-chip modeling, in vivo manipulations and bioinformatic analysis, to evaluate remodeling of the ECM, cancer cell invasive behavior and metastatic dissemination. Results Summary: Treg cell ablation resulted in significant ECM remodeling, with reduced amounts and organization of collagen fibers, and increased amounts of fibronectin and laminin. Tumor cells seeding on Treg cell ablated tumor- derived decellularized ECM matrices results in reduced epithelial-mesenchymal transition (EMT) transcription factors, and profound impairment of collective migration. In vivo, Treg cell ablation in a neo-adjuvant setting followed by primary tumor resection resulted in significant reduction of circulating tumor cells (CTC) and impairment of lung metastatic disease. Importantly, we demonstrated that Treg cell ablation-driven changes in the matrisome correlate with long-term improved survival in a cohort of breast cancer patient samples. Through genetic knock-out models and in-vivo neutralization, we observed that Treg cell-dependent changes in ECM-related phenotypes are dependent on IFN-g-signaling. Single-cell RNA-sequencing (scRNA-seq) of murine breast tumors revealed that the matrisome signature is highly upregulated among TAMs, compared to other TME cell types. Lastly, using conditional genetic knock-out models we show these IFN-g-dependent changes are mostly driven by TAMs. Conclusions: Altogether, we identified a novel metastasis-promoting effect of Treg cells in the breast cancer microenvironment through regulation of ECM dynamics beyond their described effects on primary tumor growth. Citation Format: Ailen D. Garcia-Santillan, Jessanne Y. Lichtenberg, He Shen, Jasmine M Rodriguez, Jonathan Barra, Nicholas M Clark, Wei Du, Leandro M Martinez, Ashley D Hadjis, Taylor Calicchia, Mikhail G Dozmorov, Jose J Bravo-Cordero, Amy L Olex, Priscilla Y Hwang, Paula D Bos. Regulatory T (Treg) cell contributes to tumor cell dissemination via extracellular matrix (ECM) remodeling, which is driven by TAMs in an IFNg-dependent manner [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Tumor-body Interactions: The Roles of Micro- and Macroenvironment in Cancer; 2024 Nov 17-20; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2024;84(22_Suppl):Abstract nr B023.

ANGI-05. UNSUPERVISED DECONVOLUTION OF SPATIOTEMPORAL HETEROGENEITY REVEALS HIGH-RESOLUTION INVASION STATES IN GLIOBLASTOMA

Abstract Diffuse invasion of glioblastoma cells through normal brain tissue is a key contributor to tumor aggressiveness, resistance to conventional therapies, and dismal prognosis in patients. A deeper understanding of how components of the tumor microenvironment (TME) contribute to overall tumor organization and to programs of invasion may reveal opportunities for improved therapeutic strategies. Towards this goal, we applied a novel computational workflow to a spatiotemporally profiled GBM xenograft cohort, leveraging the ability to distinguish human tumor from mouse TME to overcome previous limitations in analysis of diffuse invasion. Our analytic approach, based on unsupervised deconvolution, performs reference-free discovery of cell types and cell activities within the complete GBM ecosystem. We present a comprehensive catalogue of 15 tumor cell programs set within the spatiotemporal context of 90 mouse brain and TME cell types, cell activities, and anatomic structures. Distinct tumor programs related to invasion were aligned with routes of perivascular, white matter, and parenchymal invasion. Furthermore, sub-modules of genes serving as program network hubs were highly prognostic in GBM patients. The compendium of programs presented here provides a basis for rational targeting of tumor and/or TME components. We anticipate that our approach will facilitate an ecosystem-level understanding of immediate and long-term consequences of such perturbations, including identification of compensatory programs that will inform improved combinatorial therapies.

EPCO-26. LONGITUDINAL SINGLE-CELL ANALYSES IDENTIFY DRIVERS OF GENETIC, EPIGENOMIC, AND CELLULAR EVOLUTION IN IDH-MUTANT GLIOMA

Abstract Despite standard treatment, IDH-mutant gliomas inevitably recur. Previous studies suggest that therapeutic resistance may result from a combination of intratumoral cellular heterogeneity, epigenetic evolution, and acquired genetic alterations. However, how these multilayered molecular features interact to influence the evolutionary paths of IDH-mutant gliomas remains incompletely understood. To chart this evolution, 75 glioma samples were longitudinally collected from 35 patients (n = 13 oligodendroglioma, n = 22 astrocytoma) and profiled using single nucleus RNA sequencing, ATAC sequencing, and bulk DNA sequencing. Analysis of 331,016 nuclei (snRNA) revealed eight tumor microenvironment cell types and five pan-IDH-mutant malignant cellular states. The malignant states were distributed along a cellular hierarchy of stem-like populations (neural progenitor cell-like, oligodendrocyte progenitor cell-like, and undifferentiated) and more differentiated populations (mesenchymal-like and astrocyte-like) with cycling cells enriched in the stem-like populations. Joint open chromatin accessibility data (snATAC) was available for a subset of malignant cells (71,088 nuclei) and state-specific differentially accessible peaks supported that these states are epigenetically encoded. Across both subtypes, higher tumor grade was associated with reduced malignant cell differentiation and increased cycling populations. Between the two time points, we identified that there was a longitudinal increase in the cycling, undifferentiated, and mesenchymal-like populations with a corresponding decrease in the astrocyte-like population. Longitudinal genetic analysis revealed that 19 of 35 tumors acquired at least one of the following genetic alterations: treatment-associated hypermutation, cell cycle alteration, or large changes in copy number alterations. Importantly, tumors that acquired these key genetic alterations demonstrated significant shifts towards reduced differentiation and increased cycling populations while those tumors that did not had more stable malignant profiles. Collectively, our results suggest a common cellular hierarchy across IDH-mutant gliomas with a shift towards reduced differentiation, increased cycling populations during disease progression that is driven by acquired genetic alterations.

Cancer-associated Fibroblasts (CAFs) Regulate Lung Cancer Malignant Progression by Transferring SERPINE2 (PN1) Via Exosomes.

Cancer-associated fibroblasts (CAFs), one of the most abundant stromal cell types in tumor microenvironment (TME), have been a potential target for cancer treatment such as lung cancer. However, the underlying mechanism by which CAFs promote lung cancer progression remains elusive. We obtained primary CAFs, normal fibroblasts (NFs) and their exosomes, constructed protease nexin-1 (PN1) stably silenced or over-expressed CAFs cells using lentivirus. Bioinformatics was used to obtain the expression of PN1 in lung cancer and normal tissues, the relationship with overall survival, and the enriched pathways. The MTT assays and Transwell assays were performed to detect the proliferation, migration, and invasion abilities of lung cancer cells after treatments. Western blotting, qRT-PCR, immunohistochemistry, and xenograft models were used to illustrate how CAFs functions in lung cancer progression via exosomes. CAFs-derived exosomes, in which PN1 was higher expressed compared with NFs-derived ones, promoted effectively the proliferation, migration, and invasion of lung cancer cells A549 and H1975. Meanwhile, the expression of PN1 expressed higher in lung cancer tissues compared with normal ones, and was negatively associated with the overall survival rate of lung cancer patients. More importantly, over-expressing or silencing PN1 in A549 and H1975 could also promote or inhibit cell proliferation, migration, and invasion correspondingly. Furthermore, treated with PN1 over-expressed CAFs-derived exosomes, the lung cancer cells proliferation, migration, and invasion varied positively, and accompanied by activation of Toll-like and NF-κB signaling pathways. However, this phenomenon can be reversed by AN-3485, an antagonist of Toll-like pathway. Finally, over-expressing PN1 leads to an accelerated tumor growth by increasing the expression of proliferation biomarker Ki67 and activation of NF-κB signaling pathway in vivo. CAFs promoted lung cancer progression by transferring PN1 and activating Toll-like/NF-κB signaling pathway via exosomes.

Spatiotemporal modeling reveals high-resolution invasion states in glioblastoma

BackgroundDiffuse invasion of glioblastoma cells through normal brain tissue is a key contributor to tumor aggressiveness, resistance to conventional therapies, and dismal prognosis in patients. A deeper understanding of how components of the tumor microenvironment (TME) contribute to overall tumor organization and to programs of invasion may reveal opportunities for improved therapeutic strategies.ResultsTowards this goal, we apply a novel computational workflow to a spatiotemporally profiled GBM xenograft cohort, leveraging the ability to distinguish human tumor from mouse TME to overcome previous limitations in the analysis of diffuse invasion. Our analytic approach, based on unsupervised deconvolution, performs reference-free discovery of cell types and cell activities within the complete GBM ecosystem. We present a comprehensive catalogue of 15 tumor cell programs set within the spatiotemporal context of 90 mouse brain and TME cell types, cell activities, and anatomic structures. Distinct tumor programs related to invasion align with routes of perivascular, white matter, and parenchymal invasion. Furthermore, sub-modules of genes serving as program network hubs are highly prognostic in GBM patients.ConclusionThe compendium of programs presented here provides a basis for rational targeting of tumor and/or TME components. We anticipate that our approach will facilitate an ecosystem-level understanding of the immediate and long-term consequences of such perturbations, including the identification of compensatory programs that will inform improved combinatorial therapies.

Integrated analyses reveal IDO1 as a prognostic biomarker coexpressed with PD-1 on tumor-associated macrophages in esophageal squamous cell carcinoma.

Inhibition of indolamine-2,3-dioxygenase 1 (IDO1) has been proposed as a promising strategy for cancer immunotherapy; however, it has failed in clinical trials. Macrophages in the tumor microenvironment (TME) contribute to immune escape and serve as potential therapeutic targets. This study investigated the expression pattern of IDO1 in TME and its impact on prognosis and therapeutic response of patients with esophageal squamous cell carcinoma (ESCC). RNA sequencing data from 95 patients with ESCC from The Cancer Genome Atlas (TCGA) database were used to explore the prognostic value of IDO1. Bioinformatics tools were used to estimate scores for stromal and immune cells in tumour tissues, abundance of eight immune cell types in TME, and sensitivity of chemotherapeutic drugs and immune checkpoint (IC) blockage. The results were validated using digitalized immunohistochemistry and multiplexed immunofluorescence in ESCC tissue samples obtained from our clinical center. TCGA and validation data suggested that high expression of IDO1 was associated with poor patient survival, and IDO1 was an independent prognostic factor. IDO1 expression positively correlated with macrophages in TME and PDCD1 within diverse IC genes. Single-cell RNA sequencing data analysis and multiplexed immunofluorescence verified the coexpression of IDO1 and PD-1 in tumor-associated macrophages (TAMs). Patients with high IDO1 expression showed increased sensitivity to various chemotherapeutic drugs, while were more likely to resist IC blockage. This study identifies IDO1 as an independent prognostic indicator of OS in patients with ESCC, reveals a compelling connection of IDO1, PD-1, and TAMs, and explores the sensitivity of patients with high IDO1 expression to chemotherapeutic drugs and their resistance to IC blockade. These findings open new avenues for potential targets in ESCC immunotherapy.

Potential molecular metabolic mechanisms underlying the effects of cimifugin in gastric cancer through single-cell and bulk RNA sequencing combined with network pharmacology.

Gastric cancer (GC) is a leading cause of cancer-related mortality worldwide, posing a significant clinical challenge due to its complex tumor microenvironment (TME) and metabolic heterogeneity. Despite continuous improvements in treatment strategies including surgery, chemotherapy, and targeted therapies, the metabolic reprogramming in GC continues to impede treatment efficacy, highlighting an urgent need for the development of novel therapeutic strategies. This persistent issue underscores the urgent need for novel therapeutic approaches that can effectively address the diverse and dynamic characteristics of GC. Cimifugin, a traditional Chinese medicine (TCM), has garnered attention for its potential role in alleviating inflammation, neurological disorders, pain, and metabolic disorders. Its multi-targeting properties and minimal side effects suggest a broad potential for cancer management, which is currently being explored. This study aims to delineate the molecular mechanisms that cimifugin may impact within the TME and metabolic pathways of GC, with the expectation of contributing to a deeper understanding of GC and the development of innovative treatment strategies. We identified the GC-related TME cell types and metabolic profiles and pathways by using relevant data from the single-cell RNA sequencing (scRNA-seq) database GSE134520 and the stomach adenocarcinoma (STAD) data set from The Cancer Genome Atlas (TCGA). We also assessed the effects of cimifugin on MKN28 cell proliferation, invasion, and migration. By using six public platforms, we comprehensively predicted the potential biological targets of cimifugin. Clinical prognosis and immunohistochemistry (IHC), molecular docking, and dynamics simulations were used to confirm the clinical relevance and stability of the aforementioned targets. Cimifugin inhibited MKN28 cell proliferation, migration, and invasion. Cimifugin may potentially act on various metabolic pathways in GC, including folate biosynthesis, xenobiotic metabolism via cytochrome P450 (CYP), glutathione metabolism, steroid hormone biosynthesis, and tryptophan metabolism. Cimifugin was noted to stably bind to three significant core targets associated with metabolic reprogramming in GC: AKR1C2, MAOB, and PDE2A; all three targets were strongly expressed in endocrince cells, pit mucous cells (PMCs), and common myeloid progenitors (CMPs). We verified the pharmacological effects of cimifugin on GC cell proliferation, invasion, and migration. AKR1C2, MAOB, and PDE2A were identified as the key targets of cimifugin in GC-related metabolic reprogramming and pathogenesis. Our research provides preliminary insights into the potential therapeutic effects of cimifugin, which could be considered for future exploration in the context of GC treatment.

Abstract LB065: Tumor-infiltrating lymphocytes (TIL) engineered with regulatable membrane-bound IL15 (mbIL15) and LIGHT (TNFSF14) show enhanced efficacy in fibroblast-containing cold tumors

Abstract Introduction: Adoptive cellular therapies (ACT) have encountered challenges in solid tumors due in part to the immunosuppressive tumor microenvironment (TME). We have developed OBX-115, TIL engineered to express mbIL15 regulatable using the cytoDRiVE® platform, which allows for TIL expansion, persistence, and anti-tumor efficacy under control of the FDA-approved small-molecule ligand, acetazolamide (ACZ), eliminating the need for co-administration of IL2 (NCT05470283). LIGHT, a tumor necrosis factor family member, interacts with lymphotoxin beta receptor (LTbR) and herpes virus entry mediator (HVEM) found on various TME cell types, including stromal cells such as cancer associated fibroblasts (CAF). In preclinical studies, LIGHT expression within a tumor has been linked to the formation of tertiary lymphoid structures and vascular normalization (Ramachandran Cancer Cell 2023), both associated with better clinical outcomes (Sautès-Fridman Nat Rev Cancer 2019). We hypothesized that engineering TIL with regulatable mbIL15 and LIGHT expression could enhance their efficacy by modifying the TME. Methods: TIL from colorectal (CRC) and head and neck squamous cell carcinoma (HNSCC) were transduced with retroviral vectors to express regulatable mbIL15 and LIGHT. ACZ-induced surface expression of mbIL15 and LIGHT in expanded TIL was examined using flow cytometry. Functional signaling of LIGHT was assessed through co-culture with Jurkat-HVEM-NF-kappaB reporter cells and LTbR+ HUVEC cells. In vitro, engineered TIL were tested in stromal-rich tumor models (CRC and HNSCC) by co-culturing with autologous patient-derived tumor/CAF hybrid spheroids. In vivo, antigen-independent TIL persistence was assessed in NSG mice without exogenous IL2. Syngeneic studies were performed to assess the efficacy of adoptively transferred mbIL15 and LIGHT-engineered Pmel cells (CD8+ T cells transgenic for a gp100-specific T cell receptor) in a subcutaneous cold tumor model (B16-F10). Results: Engineered TIL were successfully expanded without exogenous IL2. ACZ-dependent mbIL15 and LIGHT expression were confirmed, validating co-regulation and functionality in vitro. TIL engineered with mbIL15 and LIGHT displayed significantly increased cytotoxicity against autologous tumor/CAF spheroids compared with TIL expressing mbIL15 alone (p<0.005) in CRC and HNSCC tumor/CAF hybrid models. TIL with mbIL15 and LIGHT expanded in vivo and persisted for ≥42 days without exogenous IL2 support. Moreover, Pmel cells engineered with mbIL15 and LIGHT demonstrated durable anti-tumor efficacy in B16-F10 tumor-bearing mice, which was greater than Pmel cells engineered with mbIL15 alone (p<0.01). Conclusions: These preclinical results suggest that TIL engineered with regulatable mbIL15 and LIGHT using the cytoDRiVE platform have the potential to address the high unmet clinical need in cold tumors with suppressive TME, which are currently not amenable to ACT. Citation Format: Balazs Koscso, Zheng Ao, Carmela Passaro, Nirzari Shah, Ngoc Ly, Patricia Timpug, Bulent A. Aksoy, Dexue Sun, Dan Jun Li, Kerri-Lynn Sheahan, Violet Young, Theresa Ross, Benjamin Primack, Meghan Langley, Jeremy Tchaicha, Dhruv K. Sethi, Jan ter Meulen, Michelle Ols. Tumor-infiltrating lymphocytes (TIL) engineered with regulatable membrane-bound IL15 (mbIL15) and LIGHT (TNFSF14) show enhanced efficacy in fibroblast-containing cold tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr LB065.

Sodium fluorescein uptake by the tumor microenvironment in human gliomas and brain metastases.

Intravenous sodium fluorescein (SF) is increasingly used during surgery of gliomas and brain metastases to improve tumor resection. Currently, SF is believed to permeate the brain regions where the blood-brain barrier (BBB) is damaged and to accumulate in the extracellular space but not in tumor or healthy cells, making it possible to demarcate tumor margins to guide resection. By evaluating the immune contexture of a number of freshly resected gliomas and brain metastases from patients undergoing SF-guided surgery, the authors recurrently observed fluorescence-positive cells. Therefore, the aim of this study was to determine if SF accumulates inside the cells of the tumor microenvironment (TME), and if so, in which type of cells, and whether incorporation can also be observed in the leukocytes of peripheral blood. Freshly resected tumor specimens were dissociated to single cells and analyzed by multiparametric flow cytometry. Peripheral blood leukocytes, macrophages, and a glioma cell line were treated with SF in vitro, and their cell uptake was assessed by multiparametric and imaging flow cytometry and by confocal microscopy. The ex vivo and in vitro analyses revealed that SF accumulates intracellularly in leukocytes as well as in tumor cells, but with a high variability of incorporation in the different cell subsets analyzed. Myeloid cells showed the highest level of fluorescence. In vitro uptake experiments showed that SF accumulation increases over time. The imaging analyses confirmed the internalization of the compound inside the cells. SF is not just a marker of BBB damage, but its intracellular detection suggests that it selectively accumulates intracellularly. Future efforts should target the mechanisms of its differential uptake by the different TME cell types in depth.

Identification of a Novel Circadian Rhythm-Related Signature for Predicting Prognosis and Therapies in Hepatocellular Carcinoma Based on Bulk and Single-Cell RNA Sequencing

Background. Circadian rhythm disruption involves tumorigenesis and tumor progression. However, the influences of circadian rhythm on the tumor microenvironment (TME) and the prognosis of hepatocellular carcinoma (HCC) are unknown. Methods. Bulk RNA-seq and single-cell RNA-seq from TCGA, ICGC, and GEO were used to comprehensively identify prognostic circadian control cells and circadian rhythm associated genes (CRRGs) using R and Python packages. Besides, the circadian rhythm-related prognostic signature was identified and validated. The biological function, immune infiltration, and therapeutic response associated with circadian rhythm-related (CR) risk were detected. Results. A total of 252 differentially expressed CRRGs in HCC were identified, and HCC with a high CR score revealed poor survival. We annotated 11 major cell types in TME; immune cells (B cells, myeloid, CD4+ cells, CD8+ cells, NK cells, Tregs) with high CR score, and hepatocyte, bio-potent cells, fibroblasts, and endothelial cells with low CR score were identified. Moreover, five CRRGs (RPL29, PFKFB3, RPS7, SLC6A6, and RPLP2) were selected and validated as the prognostic signature in HCC. The risk score was calculated based on the prognostic signature, and patients then were divided into high-risk and low-risk groups according to the median value of the risk score. High risk is linked to several metabolism-related pathways and canonical cancer-related pathways and is negatively associated with immunotherapeutic responses and positively associated with some chemotherapeutic drugs. Conclusion. Our finding provides the novel circadian rhythm-related prognostic signature and represents a novel viable “time-dependent” therapeutic option for HCC treatment.

Abstract 3387: Spatial analysis revealed distinct molecular and immune landscape in genetic versus carcinogen-induced lung adenocarcinoma precancer mouse models

Abstract Introduction: Our understanding of early carcinogenesis of lung adenocarcinoma (LUAD) is still rudimentary. Single-cell technologies have unveiled the intricate interaction between tumor cells and tumor microenvironment (TME) with unprecedented precision. However, the spatial context of single-cell phenotypes within TME during LUAD early carcinogenesis remain poorly understood. Here, we sought to characterize human-relevant murine LUAD precancer models to elucidate the molecular and immune changes in the spatial context during early LUAD carcinogenesis to inform lung cancer interception strategies. Methods: We have developed genetically engineered mouse models (GEMMs) (129S4 KrasG12D) and carcinogen-induced LUAD precancer model (CITM) (129S4 Urethane). Lung tissues were collected at various time points after induction and analyzed using whole exome sequencing (WES), imaging mass cytometry (IMC), and spatial transcriptomics (10x Visium). Results: Pathological image analysis demonstrated a significantly higher tumor burden (p<0.01) in the 129S4 KrasG12D model compared to the 129S4 Urethane model across different time points. WES revealed a progressive increase in tumor mutation burden (TMB) from precancers to invasive LUADs in both models, with the CITM model having significantly higher TMB (p<0.05) at matched timepoints than GEMM model. IMC assessed 39 biomarkers in 284 regions of interest (ROIs) and identified 1.4 million individual cells with 17 major cell types. Macrophages were the most prevalent immune cell type in TME, representing 9.00% of total cells in GEMM and 8.29% in CITM. The percent immune cells of the total cells progressively decreased from early-stage hyperplasia to late-stage adenocarcinoma in both models. The CITM exhibited a significantly (p<0.05) higher immune proportion than the GEMM model at hyperplasia and adenoma stages, with no significant difference at the invasive adenocarcinoma tumor cores between these 2 models. On the other hand, CITM had a significantly higher overall immune proportion (p<0.05) and the CD8 T cell and B cell proportions than GEMM tumor adjacent lung tissues at the adenocarcinoma stage leading to a significantly higher immune cell gradient between tumor adjacent lung tissues and tumor cores in CITM models. Spatial transcriptomics (ST) data demonstrated an increased presence of immune response signaling at the tumor adjacent lung tissues compared to the tumor core in the latest stage. Conclusion: Spatial analysis revealed different molecular and immune landscapes in the genetic and carcinogen-induced LUAD precancer models, leading to distinct cellular spatial arrangements within the tumor microenvironment in each model. Overall, innate immunity, particularly macrophage, probably plays critical roles during early carcinogenesis of LUAD regardless of carcinogen exposure. Citation Format: Bo Zhu, Pingjun Chen, Jian-Rong Li, Muhammad Aminu, Chenyang Li, Hong Chen, Yanhua Tian, Frank R. Rojas, Parra Cuentas Edwin Roger, Ou Shi, Monique B. Nilsson, Alissa Poteete, Shawna M. Hubert, Khaja B. Khan, Wei Lu, Xingzhi Song, Jianhua Zhang, Don L. Gibbons, Luisa M. Solis Soto, Ignacio I. Wistuba, John V. Heymach, Chao Cheng, Jia Wu, Jianjun Zhang. Spatial analysis revealed distinct molecular and immune landscape in genetic versus carcinogen-induced lung adenocarcinoma precancer mouse models [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3387.

Abstract 899: Development of a deep learning model for cell type mapping in colorectal cancer using H&E images leveraging image-based spatial transcriptomics data

Abstract Purpose The tumor microenvironment (TME) is crucial in colorectal cancer as it influences disease progression, treatment response, and patient outcomes, providing valuable insights for personalized therapies and prognostic assessments. Here, we have developed a deep learning model by integrating hematoxylin and eosin (H&E) stained images of colorectal cancer and image-based spatial transcriptomics (Xenium) to infer spatial mapping of cell types in TME only using H&E images. Methods A total of 30 H&E images of colorectal cancer obtained by tissue microarray were registered with image-based spatial transcriptomics data (Xenium). Utilizing a Variational Autoencoder (VAE) based model and leveraging reference single-cell data enables the acquisition of cell types for individual cells in image-based spatial transcriptomics. A convolutional neural network (CNN) model was developed using H&E image as inputs to predict cell types in H&E-stained tissue image patches of colorectal cancer collected from various patients. The model also estimated the cell types from H&E-stained whole slide tissue image of colorectal cancer of The Cancer Genome Atlas (TCGA-COAD). Results The accuracy of the model's predictions for cell types using H&E image patches was notably high and exhibited a significant concordance with the results obtained through the validation. The Intersection over Union (IoU) metric for image segmentation indicated a value of 0.66 for epithelial cells and 0.44 for TNK cells. The output of deep learning model for epithelial cells and T/NK cells from TCGA-COAD tissue images showed a correlation with human-labeled regions of cancer epithelium and immune cells. Conclusions Leveraging image-based spatial transcriptomics, we developed a deep learning model capable of discerning various cell types within the tumor microenvironment solely from H&E images. This clinically translatable approach is valuable for investigating tumor microenvironment to develop biomarkers associated with various cancer therapeutics particularly immuno-oncology drugs. This approach can yield objective deep learning-based models without human labels for characterizing the tumor microenvironment in single-cell resolution, particularly regarding spatial immune distribution. Citation Format: Seungho Cook, Dongjoo Lee, Myunghyun Lim, Jae Eun Lee, Daeseung Lee, Hyung-Jun Im, Jung-Soo Pyo, Kwon Joong Na, Hongyoon Choi. Development of a deep learning model for cell type mapping in colorectal cancer using H&E images leveraging image-based spatial transcriptomics data [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 899.

Abstract 6539: RNAi mediated silencing of STAT3 in tumor-associated immune cells induces robust anti-tumor effects in immunotherapy resistant tumors

Abstract Malignant tumors are often associated with an immunosuppressive tumor microenvironment (TME), rendering most of them resistant to standard-of-care immune checkpoint inhibitors (CPIs). Signal transducer and activator of transcription 3 (STAT3), a ubiquitously expressed transcription factor, has well-defined immunosuppressive functions in several leukocyte populations within the TME. Since the STAT3 protein has been challenging to target using conventional pharmaceutical modalities, we investigated the feasibility of applying systemically delivered RNA interference (RNAi) agents to silence its mRNA directly in tumor-associated immune cells. In preclinical murine tumor models, chemically stabilized acetylated small-interfering RNAs (siRNAs) selectively silenced Stat3 mRNA in multiple relevant cell types in TME and tumor-draining lymph nodes (TDLN), reduced STAT3 protein levels, and increased cytotoxic T-cell infiltration. In murine models of CPI-resistant tumors, RNAi-mediated Stat3 silencing resulted in tumor growth inhibition, which was further enhanced and resulted in tumor regressions in combination with CPIs. Our findings suggest that targeting STAT3 directly and reducing STAT3 protein in the suppressive immune cell populations in the TME and TDLN may overcome some of the current limitations of immunotherapy and be of therapeutic benefit especially for cancer patients who failed to respond to standard immunotherapy treatments alone. A human active STAT3 conjugate is currently in Phase 1 clinical trial for treatment of immunotherapy refractory cancer types (NCT06098651). Citation Format: Shanthi Ganesh, Minju Kim, Jenny Lee, Xudong Feng, Harini Sivagurunatha Krishnan, Rene Rijnbrand, Amber Beaudry, Jennifer A. Lockridge, Vinita Uttamsingh, John Hanrahan. RNAi mediated silencing of STAT3 in tumor-associated immune cells induces robust anti-tumor effects in immunotherapy resistant tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6539.

Abstract 1143: Dissection of melanoma and cutaneous lymphoma spatial molecular architecture by multimodal single-cell and spatial transcriptomics with generative AI

Abstract Advances in single-cell multiomic technologies have revolutionized our understanding of how heterogeneous cell types and cell states shape the tumor microenvironment (TME). However, dissociated single cell profiling lacks spatial context. To resolve the organization, cell-cell communication, and granular structures in the TME, new single-cell spatial transcriptomic (ST) characterizations are needed. Here, we dissected the spatial molecular architecture of the melanoma (30 tumors, 30 patients) and transformed cutaneous T-cell lymphoma (tCTCL, 12 tumors, 6 patients) TMEs by Multiplexed Error-Robust FISH (305 gene MERFISH), with benchmarking scRNAseq data from matched tissues (6 melanoma scFFPEseq, 6 tCTCL scVDJ/RNAseq) for cross-platform validation. We profiled 565,122 cells in melanoma and 92584 cells in tCTCL by MERFISH. We interrogated both datasets using a novel computational framework that includes cell typing by Leiden clustering and marker genes, spatial neighborhood and receptor-ligand (R-L) analyses, and spatial/scRNA imputation via deep learning. In the melanoma dataset, we identified tumor cells, major TME cell types, and rare immune cell types (TCF7+ stem-like T-cells and CD3+TCRα+PAX5+CD79a+ B/T dual-expressor lymphocytes enriched in tertiary lymphoid structures). All cell types were validated in scFFPEseq. As immune R-L interactions are critical therapeutic targets, we developed a novel computation method to quantify spatial R-L interactions by accounting for cell-cell distance, identifying spatially informed R-L pairs that differ from dissociation-based inferences. CNV inference from melanoma scFFPEseq revealed intratumoral heterogeneity (ITH). We therefore adapted the conditional variational autoencoder ENVI to simultaneously incorporate scRNA and MERFISH spatial data into a unified latent embedding. ENVI-ITH successfully learned ITH subgroup information, enabling imputation of phylogenetic lineages on spatial MERFISH. CTCL MERFISH data revealed major TME cell types, yet separation of malignant vs benign T-cells by clustering and marker approaches is hindered by their overlapping gene expression. To overcome this barrier, we deployed ENVI to impute patient-specific malignant and benign TCR clonotypes by training on the matched scVDJ/RNA data, yielding an ENVI-TCR pipeline that can delineate malignant vs benign T-cells in situ. We validated the spatial pattern of malignant T-cells by patient-specific in situ TCR probes. In sum, we have interrogated the melanoma and CTCL TME by MERFISH and presented a novel computational framework for robustly co-embedding scRNA data with imaging-based ST to spatially resolve melanoma ITH and malignant/benign T-cell populations. We believe this approach will be highly impactful for melanoma and CTCL and can be broadly applied to other solid tumors and lymphomas. Citation Format: Zichao Liu, Xiaofei Song, Jiang He, Justin He, Jodi Balasi, Jeffrey H. Chuang, Pei-Ling Chen. Dissection of melanoma and cutaneous lymphoma spatial molecular architecture by multimodal single-cell and spatial transcriptomics with generative AI [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 1143.

Abstract 5480: Single cell and spatial transcriptomic profile of appendiceal tumour peritoneal disease

Abstract Introduction: Appendiceal tumors (AT) are rare and often develop peritoneal disease (PD), with 5 year survival of 15-63% for high grade adenocarcinomas (HG) and 46-96% for low grade mucinous neoplasms (LG). While the mainstay of treatment is cytoreductive surgery and heated intraperitoneal chemotherapy (CRS/HIPEC), the role of systemic chemotherapy is less clear with limited biological characterization of AT and its tumor microenvironment (TME). The purpose of this study was to evaluate the TME of AT PD. Here we present an atlas of AT PD cells alongside a spatially resolved transcriptomic profile comparing HG and LG. Methods: We collected tissue from 29 patients with AT PD during CRS prior to HIPEC (18 LG, nine HG and two no tumor control). We performed comprehensive single-cell RNA sequencing (38224 cells) on fresh tissue specimens from nine patients (four HG, five LG), and whole transcriptome spatial analysis (>18000 genes) using the GeoMx Digital Spatial Profiler on formalin-fixed paraffin-embedded tissue from 22 patients (five HG, 13 LG). In the spatial profiling, we used anti-CK20, anti-CD45 and SYTO 83 as morphology markers to select areas of interest (AOIs): tumor (CK20), immune (CD45) and stroma (non-CK20/CD45). Analyses were performed with R and a false discovery rate threshold of 0.05. Results: Single-cell RNA sequencing showed that both LG and HG have a preponderance of T-lineage cells including CD8, CD4 and NK T-cells. Both LG and HG tumors display association of mesothelial cells with cancer-associated fibroblasts (CAF), and CAF subsets expressing genes of epithelial-mesenchymal transition (EMT). LG and HG tumors display different patterns of mucin gene expression - epithelial cells of HG express the secreted mucin MUC20 in addition to MUC2 and MUC5B also expressed in LG; a subset of mesothelial cells in LG but not HG express the transmembrane mucin genes MUC1, MUC3A, MUC12 and MUC16. Spatial analysis of tumor AOIs revealed HG compared to LG had increased expression of genes associated with MYC targets, oxidative phosphorylation, anti-tumor immune pathways (IFNα/γ response, intestinal immune network for IGA production), cell cycle pathways (E2F targets, G2M checkpoint) and EMT. LG compared to HG had increased expression of genes associated with inflammation-mediated disease (TNFα signaling via NFKB, inflammatory response), regulation of tumor growth (P53), hypoxia and early cell cycle (estrogen response) and glycosylation pathways. Immune and stromal AOIs did not have significant transcriptomic differences between LG and HG. Conclusion: This study provides novel insights into the biological profile of AT PD by describing TME cell types, immune function, tumor growth and cell cycle pathways. These findings may be used to identify clinically relevant biomarkers and new therapeutic targets. Citation Format: Madeleine C. Strach, Nicole Yeung, Eva Apostolov, Tony Wang, Hui-Ming Lin, Nabila Ansari, Cherry Koh, Joo-Shik Shin, James Kench, Alexander Swarbrick, Lisa Horvath, Kate L. Mahon. Single cell and spatial transcriptomic profile of appendiceal tumour peritoneal disease [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 5480.

Abstract 7398: Unveiling the tumor microenvironment of hepatocellular carcinoma using AI trained by spatial transcriptomics: A preliminary study to predict response to immunotherapy

Abstract Background. Immune checkpoint inhibitors (ICIs), particularly targeting the PD-1 pathway, are promising in treating hepatocellular carcinoma (HCC). However, their variable effectiveness among individuals calls for a better understanding of the tumor microenvironment (TME) and reliable predictive biomarkers. Here, we employ spatial transcriptomics (ST) to develop a deep-learning model that aims to investigate the TME in HCC using Hematoxylin and eosin (H&E) staining images. Our focus is on identifying cell types within the TME that are linked to the response to ICI in a cohort of patients treated with nivolumab. Methods. Our study leveraged 21 Visium ST datasets from HCC to develop a model aimed at inferring the cellular composition of TME from H&E images that matched with ST data. The comprehensive TME cell types enrichment score, encompassing Tumor Endothelial Cells (TECs), Tumor-Associated Macrophages (TAMs), B cells, T cells, and Cancer-Associated Fibroblasts (CAFs) in each spot, was calculated by the CellDART algorithm, a method that maps cell type by integrating with a reference scRNA-seq data. The model was trained in a patch-wise manner, utilizing a pre-trained ResNetRS50 as the backbone model. For the internal validation, Spearman correlation coefficients were evaluated for predicted cell types using H&E, and the trained model was correlated with cell type enrichment scores estimated by ST datasets of 4 samples independent of the training set. The model was applied to H&E images of 16 patients with HCC who underwent nivolumab treatment. The model predicted the 5 cell types enrichment score in the tumor and tumor-adjacent normal liver tissues. Results. The internal validation set results demonstrate robust Spearman correlations between our model predictions and actual cellular compositions in the TME. Specifically, we observe strong correlations for T cells (rho = 0.66), TAMs (rho=0.46), CAFs (rho=0.32), B cells (rho=0.30), and TECs (rho=0.24). When we applied the model to the H&E images of the cohort of patients who underwent nivolumab, immunotherapy non-responders showed a significantly higher predicted enrichment score of CAFs in tumor-adjacent regions (Mann-Whitney test, p<.01). Upon stratifying patients based on stromal CAF prediction, those with low stromal CAF levels exhibited better overall survival (Log-rank test, p<.05). Conclusions. We present a deep learning model to analyze TME in HCC solely on H&E images trained by ST data. Our findings showed a potential relationship between CAFs in tumor-adjacent regions and non-responders to nivolumab treatment in HCC. These insights underscore the potential to predict nivolumab treatment responders using H&E images combined with the deep learning model. This approach could provide a significant advancement in personalized treatment strategies for HCC patients. Citation Format: Dongjoo Lee, Haenara Shin, Seungho Cook, Daeseung Lee, Hongyoon Choi, Won-Mook Choi, Changhoon Yoo, Kwon Joong Na. Unveiling the tumor microenvironment of hepatocellular carcinoma using AI trained by spatial transcriptomics: A preliminary study to predict response to immunotherapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 7398.

Agent-based modeling of the prostate tumor microenvironment uncovers spatial tumor growth constraints and immunomodulatory properties

Inhibiting androgen receptor (AR) signaling through androgen deprivation therapy (ADT) reduces prostate cancer (PCa) growth in virtually all patients, but response may be temporary, in which case resistance develops, ultimately leading to lethal castration-resistant prostate cancer (CRPC). The tumor microenvironment (TME) plays an important role in the development and progression of PCa. In addition to tumor cells, TME-resident macrophages and fibroblasts express AR and are therefore also affected by ADT. However, the interplay of different TME cell types in the development of CRPC remains largely unexplored. To understand the complex stochastic nature of cell-cell interactions, we created a PCa-specific agent-based model (PCABM) based on in vitro cell proliferation data. PCa cells, fibroblasts, “pro-inflammatory” M1-like and “pro-tumor” M2-like polarized macrophages are modeled as agents from a simple set of validated base assumptions. PCABM allows us to simulate the effect of ADT on the interplay between various prostate TME cell types. The resulting in vitro growth patterns mimic human PCa. Our PCABM can effectively model hormonal perturbations by ADT, in which PCABM suggests that CRPC arises in clusters of resistant cells, as is observed in multifocal PCa. In addition, fibroblasts compete for cellular space in the TME while simultaneously creating niches for tumor cells to proliferate in. Finally, PCABM predicts that ADT has immunomodulatory effects on macrophages that may enhance tumor survival. Taken together, these results suggest that AR plays a critical role in the cellular interplay and stochastic interactions in the TME that influence tumor cell behavior and CRPC development.

Abstract B025: MHC class II and CD1d expressing conventional dendritic cells with immunoglobulin therapy drives anti-tumor abscopal effect by recruiting Natural Killer T cells to the tumor microenvironment

Abstract Background Dendritic cells (DC) are the primary antigen presenting cells in the body responsible for processing and presenting MHC Class I and II antigens to initiate adaptive T cell response against cancer. In addition to MHC, DC also express non-classical receptors such as CD1d to activate invariant TCR α chain natural killer T (iNKT) cells. The conventional type-I dendritic cells (cDC1) presenting MHC class-II tumor associated antigens within tumor microenvironment (TME) have shown robust efficacy in mouse models of breast cancer (BC). Here, we have extended prior knowledge to deliver DC-based therapeutics, co-presenting class II antigens and CD1d directly to TME to show the potential of co-stimulation of multiple effector T cells in combination of tumor specific antibody (IgG1) mediating antibody dependent cellular cytotoxicity (ADCC). We hypothesize that merger of adaptive and innate responses will attain greater pathologic response in pre-clinical and clinical settings to design futuristic therapeutics against BC. Methods Present study uses variety of pre-clinical mouse models of BC and other types of cancer to test the therapeutic efficacy of intratumoral delivery of mouse bone marrow-derived conventional type I (CD103+ CD1d+) polarized DC pulsed with MHC class II HER2-derived peptides (HER2-DC1) in combination with systemic delivery of tumor specific IgG1 (aSEMA4D) to evoke ADCC. Cellular mechanism of response was evaluated using downstream methods including multiplex immunofluorescence, flow cytometry, and transcriptomics assays. Results We exhibit that specifically intratumoral combination therapy of HER2-DC1 expressing CD1d and aSEMA4D primes both innate and adaptive immune responses to completely regress tumor in 75-80% of mice with significant survival advantage in comparison to mice that received no therapy or monotherapies. The early CD4 Th1 response through interferon gamma combined with ADCC played a critical role in the decrease of pro-tumorigenic myeloid cells and increase in anti-tumor effector cells with a major contribution from classical CD4+ NKT cells within the TME. Interestingly, intratumoral combination therapy intensified DC1 migration from primary tumor to secondary tumor sites along with dynamic changes in subpopulations of DC with an associated increase in NKT cells at untreated tumor sites to display abscopal effect. Likewise, early data from phase I/II clinical trial illustrated great potential of cDC1 and IgG1 combination therapy to recruit anti-tumor effector cells such CD8 and NKT cells to mediate cancer eradication in HER2+ BC setting. Conclusions Specifically, intratumoral delivery of HER2-DC1 in combination with anti-tumor antibody mediating ADCC shows great promise for future BC clinical trials and improved patient outcomes. Furthermore, present study highlighted the role of NKT cells in favorable anti-tumor response. Our future studies are focused on understanding the crosstalk between DC and NKT to modulate the function of various cell types in TME to design improved cancer immunotherapies. Citation Format: Saurabh K Garg, Colin Snyder, Ganesan Ramamoorthi, Elizabeth E Evans, Amy L Aldrich, Namrata Gautam, Doris Wiener, Marie C Lee, Hyo S Han, Koski K Gary, Brian J Czerniecki. MHC class II and CD1d expressing conventional dendritic cells with immunoglobulin therapy drives anti-tumor abscopal effect by recruiting Natural Killer T cells to the tumor microenvironment [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Breast Cancer Research; 2023 Oct 19-22; San Diego, California. Philadelphia (PA): AACR; Cancer Res 2024;84(3 Suppl_1):Abstract nr B025.

Forces at play: exploring factors affecting the cancer metastasis.

Metastatic disease, a leading and lethal indication of deaths associated with tumors, results from the dissemination of metastatic tumor cells from the site of primary origin to a distant organ. Dispersion of metastatic cells during the development of tumors at distant organs leads to failure to comply with conventional treatments, ultimately instigating abrupt tissue homeostasis and organ failure. Increasing evidence indicates that the tumor microenvironment (TME) is a crucial factor in cancer progression and the process of metastatic tumor development at secondary sites. TME comprises several factors contributing to the initiation and progression of the metastatic cascade. Among these, various cell types in TME, such as mesenchymal stem cells (MSCs), lymphatic endothelial cells (LECs), cancer-associated fibroblasts (CAFs), myeloid-derived suppressor cells (MDSCs), T cells, and tumor-associated macrophages (TAMs), are significant players participating in cancer metastasis. Besides, various other factors, such as extracellular matrix (ECM), gut microbiota, circadian rhythm, and hypoxia, also shape the TME and impact the metastatic cascade. A thorough understanding of the functions of TME components in tumor progression and metastasis is necessary to discover new therapeutic strategies targeting the metastatic tumor cells and TME. Therefore, we reviewed these pivotal TME components and highlighted the background knowledge on how these cell types and disrupted components of TME influence the metastatic cascade and establish the premetastatic niche. This review will help researchers identify these altered components' molecular patterns and design an optimized, targeted therapy to treat solid tumors and restrict metastatic cascade.

Diagnostic and Therapeutic Application of Fibroblast Activation Protein Inhibitors in Oncologic and Nononcologic Diseases.

Fibroblast activation protein inhibitor positron emission tomography (PET) has gained interest for its ability to demonstrate uptake in a diverse range of tumors. Its molecular target, fibroblast activation protein, is expressed in cancer-associated fibroblasts, a major cell type in tumor microenvironment that surrounds various types of cancers. Although existing literature on FAPI PET is largely from single-center studies and case reports, initial findings show promise for some cancer types demonstrating improved imaging when compared with the widely used 18F-fludeoxyglucose PET for oncologic imaging. As we expand our knowledge of the utility of FAPI PET, accurate understanding of noncancerous uptake seen on FAPI PET is crucial for accurate evaluation. In this review, we summarize potential diagnostic and therapeutic applications of radiolabeled FAP inhibitors in oncological and nononcological disease processes.