Spatial Proteomics Research Articles

The production of CXCL13 in CD4+ T cells is influenced by adjacent Tregs in the tertiary lymphoid structures of chronic pemphigus blisters

Abstract Pemphigus is a rare and life-threatening autoimmune blistering disease mediated by autoreactive B cells targeting desmogleins (DSGs) in keratinocytes. Anti-DSG autoantibodies disrupt cell-cell adhesion, leading to suprabasal acantholytic blisters. Our investigation reveals the presence of tertiary lymphoid structures (TLSs) harboring DSG-specific B cells and plasma cells in chronic skin blisters of pemphigus. TLSs, resembling B-cell follicles of secondary lymphoid organs, are detected in various inflammatory diseases. CXCL13 is an important chemokine for recruiting CXCR5-expressing naïve B and T follicular helper cells. Through immunofluorescence studies, we found that CXCL13+ cells were mainly produced by CD4+ T cells. Single-cell RNA sequencing revealed clonally expanded CXCL13+CD4+ T cells containing DSG3-specific memory T cells with activated cytotoxic features. Paradoxically, these cells exhibit an attenuation of T-cell receptor (TCR) signaling when expressing high levels of CXCL13. Spatial proteomics reveals that Tregs are spatially adjacent to CXCL13+CD4+ T cells within the skin TLSs. In vitro depletion of Tregs leads to decreased CXCL13 production by CD4+ T cells. These data suggest that adjacent Tregs play a role in regulating CXCL13+CD4+ T cells. In conclusion, skin TLSs are associated with the maintenance of chronic blisters, and the microenvironmental network involving CXCL13+CD4+ T cells and Tregs plays a crucial role in CXCL13 production.

Molecular Pixelation - dissecting the spatial organization of 80 surface proteins in single immune cells

Abstract Immune cell function is dependent on the activity of membrane proteins. Most widely-used cell analysis methods, flow cytometry and scRNAseq, directly or indirectly assess membrane protein abundance, yet provide no information about their spatial distribution. The gold standard technology for spatial protein analysis, fluorescence microscopy, has revealed that the spatial organization of membrane proteins is in fact of importance for their function, as exemplified by the immune synapse and uropods in migrating cells. Limited multiplexing and throughput scalability of microscopy, reduces its potential for forefront spatial proteomics discoveries, highlighting the need for new innovative methods. Here, we present Molecular Pixelation (MPX), an advanced DNA barcode-based method for high-multiplex spatial analysis of proteins in single cells. Combining an NGS-based readout with our dedicated software, Pixelator, MPX enables the reconstruction of 3D cell-surface maps of 80 proteins on over 1000 individual cells, simultaneously. We applied MPX to compare the surface proteome of single resting and in-vitro activated T cells, highlighting differences in protein abundance, as well as revealing dramatic spatial protein re-organization, both in terms of supramolecular clustering and protein-protein co-localization. These results demonstrate that MPX has the potential to pave the way for expanding our understanding of cell surface protein architecture and its role in T cell activation.

Spatio-temporal Phenotyping of Murine Embryos with Ultrahigh-plex Antibody Panels for Characterization of Early Development

Abstract The murine embryo model is one of the most common animal models for studying developmental biology. Embryonic development is a highly regulated process with spatial and temporal expression of proteins for regional specialization and tissue-specific functions. Application of spatial proteomics to developing embryos enables us to understand cellular organization and interactions within tissues along with regulatory mechanisms at the single-cell level. In this study, we applied a 36-plex panel, specifically designed for analysis of mouse embryos at different stages of development using the PhenoCycler™-Fusion (PCF) 2.0 Platform. This panel was developed by selecting biomarkers for immune cell lineages, immune activation, proliferation, structural and neuronal components. Purified antibodies were tagged to PhenoCycler barcodes and verified for specificity and reproducibility. After successful validation, the panel was deployed on the FFPE mouse embryos and ultrahigh-plex whole-slide imaging was performed using the PCF 2.0 platform. Unsupervised clustering analysis and deep phenotyping were performed to generate a high-resolution spatial atlas of mouse embryos at different stages of development. The tools developed as part of this study can be used to conduct in-depth research on various aspects of organogenesis and to study the spatio-temporal dynamics of cellular and molecular expression in developing embryos.

Spatial Proteomics Reveals Alcohol-Induced Damages to the Crypts and Villi of the Mouse Small Intestine.

Alcohol consumption perturbs the gut immune barrier and ultimately results in alcoholic liver diseases, but little is known about how immune-related cells in the gut are perturbed in this process. In this study, we employed laser capture microdissection and a label-free proteomics approach to investigate the consequences of alcohol exposure to the proteomes of crypts and villi in the proximal small intestine. Intestinal tissues from alcohol-fed and pair-fed mice were microdissected to selectively capture cells in the crypts and villi regions, followed by one-pot protein digestion and data-independent LC-MS/MS analysis. We successfully identified over 3000 proteins from each of the crypt or villi regions equivalent to ∼3000 cells. Analysis of alcohol-treated tissues indicated an enhanced alcohol metabolism and reduced levels of α-defensins in crypts, alongside increased lipid metabolism and apoptosis in villi. Immunofluorescence imaging further corroborated the proteomic findings. Our work provides a detailed profiling of the proteomic changes in the compartments of the mouse small intestine and aids in molecular-level understanding of alcohol-induced tissue damage.

Integrative spatial analysis reveals a multi-layered organization of glioblastoma

Glioma contains malignant cells in diverse states. Here, we combine spatial transcriptomics, spatial proteomics, and computational approaches to define glioma cellular states and uncover their organization. We find three prominent modes of organization. First, gliomas are composed of small local environments, each typically enriched with one major cellular state. Second, specific pairs of states preferentially reside in proximity across multiple scales. This pairing of states is consistent across tumors. Third, these pairwise interactions collectively define a global architecture composed of five layers. Hypoxia appears to drive the layers, as it is associated with a long-range organization that includes all cancer cell states. Accordingly, tumor regions distant from any hypoxic/necrotic foci and tumors that lack hypoxia such as low-grade IDH-mutant glioma are less organized. In summary, we provide a conceptual framework for the organization of cellular states in glioma, highlighting hypoxia as a long-range tissue organizer.

Cross-link assisted spatial proteomics to map sub-organelle proteomes and membrane protein topologies

The functions of cellular organelles and sub-compartments depend on their protein content, which can be characterized by spatial proteomics approaches. However, many spatial proteomics methods are limited in their ability to resolve organellar sub-compartments, profile multiple sub-compartments in parallel, and/or characterize membrane-associated proteomes. Here, we develop a cross-link assisted spatial proteomics (CLASP) strategy that addresses these shortcomings. Using human mitochondria as a model system, we show that CLASP can elucidate spatial proteomes of all mitochondrial sub-compartments and provide topological insight into the mitochondrial membrane proteome. Biochemical and imaging-based follow-up studies confirm that CLASP allows discovering mitochondria-associated proteins and revising previous protein sub-compartment localization and membrane topology data. We also validate the CLASP concept in synaptic vesicles, demonstrating its applicability to different sub-cellular compartments. This study extends the scope of cross-linking mass spectrometry beyond protein structure and interaction analysis towards spatial proteomics, and establishes a method for concomitant profiling of sub-organelle and membrane proteomes.

Abstract PL03-02: From microniches to single cells: Understanding the impact of intratumoral microbes on spatial and cellular heterogeneity in cancer

Abstract The tumor microenvironment (TME) is a dynamic network that includes a variety of non-malignant cells such as endothelial cells, fibroblasts, and immune cells, which play a crucial role in tumor behavior. Advances in molecular analysis have recently identified the presence of intratumoral microbiota in several solid cancer types. However, the specific ways these microorganisms interact with human cells within the TME and contribute to disease progression are not yet fully understood. Our recent research provides new insights into how bacterial components of the intratumoral microbiota contribute to tumor heterogeneity, particularly in head and neck and colorectal cancers. To investigate these interactions, we have adapted and applied emerging technologies in spatial transcriptomics, spatial proteomics, and single-cell sequencing (INVADEseq). Our spatial analyses show that the intratumoral microbiota are not distributed randomly within human tumors; instead, they localize to distinct microniches that promote immune and epithelial cell functions supportive of cancer progression. These bacterial-colonized microniches are characterized by myeloid cell infiltration, upregulation of MAPKs and immune checkpoint proteins, and reduced T-cell levels. Furthermore, INVADEseq single-cell RNA sequencing has enabled us to identify the dominant host cell-associated bacteria, the human cell types they infect, and their impact on the transcriptional programs of infected cells from patient tumor specimens. In oral squamous cell carcinoma, Fusobacterium and Treponema species have been identified as the dominant cell-associated bacterial taxa, primarily infecting subsets of macrophages and aneuploid cancer epithelial cells and upregulating pathways associated with cancer progression. Our findings offer insights into the complex interplay between intratumoral microbiota and human components of the TME, emphasizing their role in supporting intratumoral heterogeneity and influencing cancer progression. Understanding these dynamics holds significant promise for the development of targeted therapeutic strategies aimed at disrupting bacterial-human spatial and cellular interactions within the TME. Citation Format: Susan Bullman. From microniches to single cells: Understanding the impact of intratumoral microbes on spatial and cellular heterogeneity in cancer [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 PL03-02.

Comparative spatial proteomics between SARS-CoV-2 infected lungs of Roborovski dwarf hamsters and Syrian hamsters

Comparative spatial proteomics between SARS-CoV-2 infected lungs of Roborovski dwarf hamsters and Syrian hamsters

SpiDe-Sr: blind super-resolution network for precise cell segmentation and clustering in spatial proteomics imaging

Spatial proteomics elucidates cellular biochemical changes with unprecedented topological level. Imaging mass cytometry (IMC) is a high-dimensional single-cell resolution platform for targeted spatial proteomics. However, the precision of subsequent clinical analysis is constrained by imaging noise and resolution. Here, we propose SpiDe-Sr, a super-resolution network embedded with a denoising module for IMC spatial resolution enhancement. SpiDe-Sr effectively resists noise and improves resolution by 4 times. We demonstrate SpiDe-Sr respectively with cells, mouse and human tissues, resulting 18.95%/27.27%/21.16% increase in peak signal-to-noise ratio and 15.95%/31.63%/15.52% increase in cell extraction accuracy. We further apply SpiDe-Sr to study the tumor microenvironment of a 20-patient clinical breast cancer cohort with 269,556 single cells, and discover the invasion of Gram-negative bacteria is positively correlated with carcinogenesis markers and negatively correlated with immunological markers. Additionally, SpiDe-Sr is also compatible with fluorescence microscopy imaging, suggesting SpiDe-Sr an alternative tool for microscopy image super-resolution.

Fetal liver macrophages contribute to the hematopoietic stem cell niche by controlling granulopoiesis.

During embryogenesis, the fetal liver becomes the main hematopoietic organ, where stem and progenitor cells as well as immature and mature immune cells form an intricate cellular network. Hematopoietic stem cells (HSCs) reside in a specialized niche, which is essential for their proliferation and differentiation. However, the cellular and molecular determinants contributing to this fetal HSC niche remain largely unknown. Macrophages are the first differentiated hematopoietic cells found in the developing liver, where they are important for fetal erythropoiesis by promoting erythrocyte maturation and phagocytosing expelled nuclei. Yet, whether macrophages play a role in fetal hematopoiesis beyond serving as a niche for maturing erythroblasts remains elusive. Here, we investigate the heterogeneity of macrophage populations in the murine fetal liver to define their specific roles during hematopoiesis. Using a single-cell omics approach combined with spatial proteomics and genetic fate-mapping models, we found that fetal liver macrophages cluster into distinct yolk sac-derived subpopulations and that long-term HSCs are interacting preferentially with one of the macrophage subpopulations. Fetal livers lacking macrophages show a delay in erythropoiesis and have an increased number of granulocytes, which can be attributed to transcriptional reprogramming and altered differentiation potential of long-term HSCs. Together, our data provide a detailed map of fetal liver macrophage subpopulations and implicate macrophages as part of the fetal HSC niche.

Abstract 3772: Tissue-niche-based and cell-type-selective in-depth proteomics

Abstract Introduction: The development, physiology function, and pathogenesis of the tissues are intricately orchestrated by the spatial organization of diverse cell populations. Proteins are the fundamental components within individual cells that control every function. Therefore, by integrating both information on cellular location and molecular profiling, spatial proteomics has emerged as an indispensable field for understanding mechanisms governing tissue homeostasis, disease progression, and therapeutic responses. Here, I introduce a spatial proteomics method designed to target selective cell phenotypes within definite tissue niches for in-depth spatial biology analysis. Method: In this approach, we developed a sequential pipeline highlighted with immunofluorescence staining, microscopic photobleaching, fluorescence cell sorting, and liquid chromatography-mass spectrometry (LC-MS) analysis. The process begins with staining 400-μm-thick tissue macrosections with fluorescence-conjugated antibodies to visualize specific cell types of interest. We employed two different fluorescence-conjugated antibodies of the same clone (e.g., Alexa488-anti-CD11c and Alexa647-anti-CD11c antibodies) for the staining on every targeted cell (e.g., CD11c+ immune cells). We then introduced “photobleaching barcodes” using a confocal microscope to photobleach/exhaust either one of the fluorescence signals stained on cells by exposing it to the matched laser. This allows us to create differentiable fluorescence barcodes that label cells at distinct regions, incorporating cell spatial information into our fluorescence detection. After tissue dissociation, barcoded cells were sorted into different groups which were also classified by their original locations in the macrosection. Finally, protein extraction and LC-MS analysis were conducted on the collected cells to enable comprehensive spatial proteomic analysis. Results: The initial application in investigating dendritic cell (DC) subsets in mouse spleen during lipopolysaccharide (LPS)-induced inflammation reveals significant proteome differences among three splenic DC subsets, categorized by their locations in or outside spleen T-cell zones as well as control DCs. Our result aligns with previously published proteome data in splenic DCs, underscoring the feasibility and reliability of our technology. Currently, we are in the process of evaluating breast cancer cell heterogeneity in human biopsy specimens using our approach to profile the tumor-immune microenvironment. Conclusion: Spatial proteomic findings from both applications are poised to discover potential drug targets that benefit treatment efficacy for inflammatory diseases and breast cancer. This technology features deep protein profiling for cell-type-specific spatial proteomics and is designed for broad applications across diverse tissue specimens in various diseases. Citation Format: Yi-Chien Wu, Elie Abi Khalil, Samuel Weng, Steve Lee. Tissue-niche-based and cell-type-selective in-depth proteomics [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 3772.

Abstract 2762: Spatial proteomics of lung adenocarcinoma mouse models with divergent propensities for metastasis

Abstract Introduction: Lung cancer is the leading cause of cancer deaths worldwide with most deaths attributed to metastasis. Metastases in existing mouse models of lung adenocarcinoma (LUAD) is sporadic and often requires several months of aging. These factors limit these models for basic and pre-clinical research aimed at identifying mechanisms of metastatic disease. We have recently developed two novel mouse models of LUAD that feature constitutively active Kras, loss of Trp53 function and truncated Dicer1 in different cell types. The observed survival and metastatic phenotypes of the models are dependent on the cell type in which the genetic alterations are made. We applied spatial proteomics and metabolomics to investigate molecular drivers underlying the phenotypic differences in our models. Methods and Results: Mouse models were generated by adding mutations in Dicer1, an RNAse III enzyme within the microRNA (miRNA) biosynthesis pathway, to a mouse model of Kras-driven LUAD. For both models, tumorigenesis was induced by conditional expression of an oncogenic allele of Kras, deletion of both alleles of Trp53 and deletion of one allele of Dicer1 in one cell type and the expression of a truncated Dicer1 allele in a different cell type. Mice expressing KrasG12D and deleted Trp53 in club cells without Dicer1 truncation have a reported median survival of 28.6 weeks after tumor induction. We observed accelerated development of LUAD and lymph node metastasis (12.1 weeks) only when we induced tumorigenesis in club cells and truncate Dicer1 in alveolar type II (ATII) cells. Induction of tumorigenesis in ATII cells and truncation of Dicer1 in club cells did not accelerate tumorigenesis or lead to detectable metastasis. Prelminary spatial proteomic analysis revealed significant upregulation of proteins associated with mitochondrial respiratory chain complex I in the tumors of both models which is consistent with processes of bioenergetic adaptation of tumor cells to hypoxic conditions. Gene Ontology annotation term enrichment revealed that the tumors of the two models have distinctive molecular function profiles. Both models show significant disruption of amino acid metabolism but only the model with increased metastatic potential showed apparent disruption in metabolites associated with steroidogenesis. Conclusions: Through cell type specific truncation/deletion of Dicer1 we have generated a mouse model that rapidly develops Kras/Trp53 driven LUAD and metastatic disease. Our findings suggest that tumorigenesis and metastasis are influenced by miRNA-mediated communication between different cell types. Additional studies are underway to link miRNA expression in these models to the results reported here for spatial proteomics and metabolomics. We will continue to develop this model to investigate the molecular mechanisms underlying metastasis. Citation Format: Paige Ramkissoon, Naomi Mitchell-Hutchinson, Julie Wells, Rick Maser, Tim Stodola, Brian Hoffmann, Anne Marchini, Elaine Bechtel, Rosalinda Doty, Carol J. Bult. Spatial proteomics of lung adenocarcinoma mouse models with divergent propensities for metastasis [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 2762.

Abstract 3802: Integrating multiplex-staining and multiplex-DIA: Profiling the tumor microenvironment spatial proteome for precision cancer research

Abstract Our innovative dual-dimensional multiplexing approach ushers in a new era of spatial proteomic exploration. Multiplex-staining deepens understanding of tumor-stroma interactions. Our robust dimethyl labeling-based multiplex-DIA workflow enhances throughput and proteome depth in the Deep Visual Proteomics (DVP) technology. Now, our audacious integration of multiplex-staining with multiplex-DIA holds the promise of pushing the boundaries of spatial proteome profiling. This innovation combines the power of mass spectrometry (MS) with clinical viability in 1) requiring just one FFPE section for spatial proteome analysis at the resolution of individual cell types; 2) ingeniously transmuting MS revelations into a clinically feasible staining strategy. A force in personalized oncology, our innovation unravels the intricate dynamics of the cancer ecosystem while ensuring direct clinical relevance. We conducted multiplex immunofluorescent staining on human cancers using the MACSima platform, followed by cellular phenotyping and automated single-cell laser microdissection. Subsequently, we loaded dimethyl-labelled 3-plex tryptic peptides onto Evotips and performed LC-MS analysis using the Evosep One coupled to timsTOF SCP. Pre- and post-22-plex staining proteome analysis confirmed no significant protein loss, validating microdissection compatibility and excellent recovery. The key findings include: 1) profiling of 21 distinct cell types/subtypes in human tonsil cancer tissues, resulting in the identification of 4941 protein groups from a cellular volume of 110,859 µm3, equivalent to approximately 55 cells; 2) observation of a loss of CD45RA/RO expression in tumor-infiltrating cytotoxic T cells, contributing to a reduction in MET and PDGF signaling activity within tumor cells; 3) the presence of abundant cytotoxic T cells, B cells, and plasma cells, coupled with a reduction in tumor cell proliferation, indicative of a hot tumor microenvironment. Proteome analysis revealed enhanced immune cell mobility and interactions with tumor cells. Moreover, our workflow facilitated the identification of potential diagnostic markers for mycosis fungoides, a marker-lacking skin lymphoma, where we achieved the identification of 2964 significant proteins out of a total of 6009. In addition, a rare lymphoma case study not only guided ABVD chemotherapy and suggested IL4 inhibition to address chemoresistance but also identified CDK, MCM, and PSMA/B/C/D/E inhibitors as promising candidates for targeted therapy. The integration of multiplex-staining and multiplex-DIA represents a significant stride in advancing personalized diagnostics and therapeutic strategies. Citation Format: Xiang Zheng, Andreas Mund, Matthias Mann. Integrating multiplex-staining and multiplex-DIA: Profiling the tumor microenvironment spatial proteome for precision cancer research [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 3802.

Abstract 3649: A novel spatial multi-omic approach for biological discoveries in colonic diseased tissues using a comprehensive Immuno-Oncology Proteome Atlas and Whole Transcriptome Atlas

Abstract The advancement of spatially resolved, multiplex proteomic and transcriptomic technologies has revolutionized and redefined the approaches to complex biological questions pertaining to tissue heterogeneity, tumor microenvironments, cellular interactions, cellular diversity, and therapeutic response. While spatial transcriptomics has traditionally led the way in plex, multiple studies have demonstrated a poor correlation between RNA expression and protein abundance, owing to transcriptional and translational regulation, target turnover, and most critically, post-translational protein modifications. Therefore, a more holistic, ultra-high-plex, and high-throughput proteomic atlas approach becomes critical for the next phase of discovery biology. Here, we present a barrier-breaking, spatial proteomics panel that was designed to accelerate scientific discoveries. A Digital Spatial Profiler platform is uniquely suited to support high-plex proteomics, allowing for the simultaneous analyses of proteins from discrete regions of interest (ROIs) in FFPE tissue sections while preserving spatial context. The assay relies upon abcam antibodies coupled to photocleavable DNA barcodes readout with NGS sequencing, allowing for theoretically unlimited plex. Here we introduce the Human Immuno-Oncology Proteome Atlas (IPA), a 570+ antibody-based proteomic discovery panel, compatible with immunohistochemistry on FFPE tissues with NGS readout. IPA is the highest-plex, most comprehensive, antibody-based multi-omic panel to date focusing on key areas of immuno-oncology, oncology, immunology, epigenetics, metabolism, cell death, and signaling pathway regulation. Here we demonstrate the performance of IPA on various cell lines and tissue. Additionally, we show the power of IPA, using the spatial multi-omic assay along with the GeoMx® Whole Transcriptome Atlas (> 18,000 transcripts), a 40-plex custom antibody panel and microbiome-curated RNA custom spike-in (~220 transcripts) to evaluate 70 different colon disease samples across 5 pathologies including adenocarcinoma, hyperplasia, and chronic inflammation. This is the highest-plex multi-omic (~610-plex proteins and >18,220 genes) study ever implemented for spatial biology. When we compared the diseased tissue to normal tissue, we observed an upregulation of specific pathways associated with tumorigenesis and inflammation. Furthermore, we observed distinct differences in proteomic and transcriptomic landscape between pathologies. The cutting-edge, data-driven, expert-curated IPA panel is at the forefront of spatial proteomics, empowering the researcher for the acceleration of biological discoveries. FOR RESEARCH USE ONLY. Not for use in diagnostic procedures. Citation Format: Alyssa Rosenbloom, Shilah Bonnett, Mark Conner, Christine Kang, Erin Piazza, Brian Filanoski, Rhonda Meredith, Hye Son Yi, Lori Hamanishi, Eduardo Ignacio, Nadine Nelson, Michael Prater, Vik Devgan, Rudy Van Eijsden, Melanie Moon, Lesley Isgur, Terence Theisen, Margaret Hoang, Gary Geiss, Joseph M. Beechem. A novel spatial multi-omic approach for biological discoveries in colonic diseased tissues using a comprehensive Immuno-Oncology Proteome Atlas and Whole Transcriptome Atlas [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 3649.

Abstract 4407: Simultaneous spatial epigenomic and transcriptomic analysis of gastric adenocarcinoma reveals regulatory patterns governing tumor and microenvironment architecture at the cellular level

Abstract Recent advances in spatial transcriptomics and spatial proteomics have enabled increasingly complex questions on the nature of gene regulation and expression in cellular subtypes in tumor tissue and the tumor microenvironment. However, most spatial omics techniques do not profile the epigenomic landscape responsible for downstream gene expression. Furthermore, current spatial technologies have yet to profile the epigenome and transcriptome simultaneously, and thus it remains a challenge to correlate multi-omics data across sections of extremely heterogenous tumor tissue. Recently, co-profiling of spatial epigenomics and transcriptomics using principles of Deterministic Barcoding in Tissue for spatial omics sequencing (DBiT-seq) has been demonstrated on normal brain tissue. Joint spatial profiling of chromatin states and whole transcriptome in tissue allows for parallel characterization of gene regulation programs across all cell types, while preserving the tissue architecture for greater understanding of the cellular environment. Here we present the first application of spatial ATAC-seq and spatial transcriptomics on the same tissue section to characterize the tumor microenvironment of an invasive gastric adenocarcinoma (GAC) and adjacent normal tissue. GAC is the fifth most common cancer and commonly exhibits mutations in epigenetic modifiers, including ARID1A and MLL1-4. Distinct spatial clusters representing different cell subtypes were identified via both spatial chromatin accessibility and spatial transcriptomics. Spatial ATAC-seq profiling of accessible regulatory elements correlated well with RNA expression of target genes. Spatial patterns of transcription factor motif accessibility also correlated well with the observed transcriptional program of tumor tissue. When compared to adjacent normal tissue, spatial co-profiling of chromatin accessibility and the transcriptome revealed that the epigenetic landscape is significantly altered in tumorigenesis of GAC. Future work will focus on development of co-profiling of histone modifications and the transcriptome to enable the study of another layer of the epigenomic landscape, especially as targeting epigenetic modifiers such as EZH2 has been identified as a potential therapeutic strategy in GACs. Overall, we present a solution to profile multiple layers of gene regulation and expression with spatial context, which can be applied to most tumor types for better understanding of tumorigenesis and the consequences of new targeted therapies. Citation Format: Katelyn J. Noronha, Jennifer M. Garbarino, Daniel Massucci, Abigail R. Tyree, Colin Ng. Simultaneous spatial epigenomic and transcriptomic analysis of gastric adenocarcinoma reveals regulatory patterns governing tumor and microenvironment architecture at the cellular level [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 4407.

Abstract 1151: Reconstruction of the spatial ecosystem of glioblastoma reveals relationships between tumor cell states and microenvironment

Abstract Glioblastoma multiforme (GBM) is the most aggressive form of primary brain tumor, with no curative treatment options. Multiple studies have characterized at single cell resolution the GBM as being composed of transcriptional cell states interconnected with components in the tumor immune microenvironment (TME). Our group proposed and validated the first single cell guided functional classification of GBM in four tumor-intrinsic cell states which informed clinical outcome and delivered therapeutic options. However, single cell technologies are unable to unravel the spatial relationships among the cell states of GBM and between GBM cell states and TME. Spatially resolved transcriptomic technologies are emerging as powerful tools to reconstruct the spatial architecture of a tissue. We performed spatial transcriptomics of multicellular regions of interest (ROI) in 6 primary IDH wild-type GBM and 2 recurrent GBM with both CosMx Spatial Molecular Imager, which analyzes 1,000 RNA probes and 64 proteins at single cell resolution, and GeoMx Digital Spatial Profiler which profiles the whole transcriptome (~18,000 genes) at ROI resolution. The development of computational tools aimed to integrate spatial proximity and CosMx derived single-cell transcriptomics revealed spatial segregation of the tumor cell clones and cellular states and highlighted recurrent patterns of cell states, distinct TME cell types associated with coherent histopathological features across multiple samples. The development of a spatial informed intercellular communication algorithm and the reconstruction of ligand-receptor-target networks will allow the discovery of tumor cell states-TME cross-talks and the biological signaling regulated by these interactions that are driving the heterogeneity of GBM and therefore potentially therapeutically targetable. Analysis of matched regions of interest profiled by GeoMx and spatial proteomics with CosMx further cross-validated the spatial ecosystem of glioblastoma as reconstructed at single-cell resolution. Our studies established a scalable approach to resolve the transcriptional heterogeneity of GBM and reconstruct the architecture of GBM cell states and tumor microenvironment. Citation Format: Bruno Adabbo, Simona Migliozzi, Luciano Garofano, Fulvio D'Angelo, Pedro Davila, Sakir H. Gultekin, Daniel Bilbao Cortes, Benjamin Currall, Sion L. Williams, Marc Sanson, Franck Bielle, Anna Luisa Di Stefano, Michele Ceccarelli, Anna Lasorella, Antonio Iavarone. Reconstruction of the spatial ecosystem of glioblastoma reveals relationships between tumor cell states and microenvironment [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 1151.

Abstract 6092: Three-dimensional multi-omic analysis of early invasion of human pancreatic ductal adenocarcinoma from IPMN

Abstract Background: Pancreatic ductal adenocarcinoma (PDAC) is a deadly cancer, with only 11% of patients surviving beyond 5 years. Intraductal papillary mucinous neoplasm (IPMN), a precancerous lesion and the most common type of neoplastic pancreatic cyst, presents a critical opportunity for cancer interception, but the drivers of malignant progression in IPMN are still largely unknown. Although the immunosuppressive microenvironment of PDAC has been well characterized, the timing of the development of these immune alterations in IPMNs is not well studied and could provide a rational foundation for cancer immuno-prevention. Moreover, the inception of invasive PDAC and its spread into the microenvironment from IPMN is complex, multifactorial, and, due to its unique anatomy, challenging to analyze with traditional 2-dimensional visualization. We sought to capture the precise point of transition from IPMN to invasive carcinoma utilizing CODA, a supervised deep learning tool for three-dimensional (3D) reconstruction of serially sectioned human tissue, in order to quantitatively assess the molecular and cellular alterations associated with malignant progression. Method: Formalin-fixed paraffin-embedded (FFPE) tissue blocks with PDAC arising from IPMN were serially sectioned, every third slide was stained with H&E, and digitized at 20x magnification. Following pathologist-guided annotations on a subset of H&E slides, CODA generated 3D models of each tissue block, including automated annotation of 9 pancreatic tissue components. Using these annotations, we identified the transition from IPMN to invasive carcinoma in each sample and selected regions for multi-omic profiling based on quantitative features of the cellular microenvironment. Multi-omic profiling of IPMN, transition zone, and PDAC included laser capture microdissection followed by whole exome sequencing to identify somatic DNA alterations, as well as spatial transcriptomics to identify alterations in gene expression in neoplastic cells and spatial proteomics to identify cellular alterations in the surrounding microenvironment. Results: CODA generated accurate annotation on all the H&E slides from each case and robustly identified regions of interest for multi-omic profiling. Spatial proteomic profiling using imaging mass cytometry revealed a decrease in T cell density as IPMNs transition into PDAC. Sub- clustering of the T cell compartment identified decreases in multiple T cell subsets, including activated cytotoxic T cells and helper T cells, in the transition zone compared to IPMN. Conclusion: In this study, we integrated multi-omic profiling with CODA-generated high-resolution 3D tissue maps to identify molecular and cellular drivers of malignant progression in human PDAC. Citation Format: Shalini Datta, Sarah M. Shin, Jessie Kanacharoen, Michael Johannes Pflüger, Katsuya Hirose, André Forjaz, Sarah Graham, Pei-Hsun Wu, Ralph H. Hruban, Denis Wirtz, Won Jin Ho, Ashley L. Kiemen, Laura D. Wood. Three-dimensional multi-omic analysis of early invasion of human pancreatic ductal adenocarcinoma from IPMN [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 6092.

Abstract 1041: Spatial concordance between genome and proteome in oncology: A systematic review

Abstract We recently reported that spatial transcriptome data can be used to spatially detect copy number alterations (Erickson et al., Nature, 2022). Given that copy-number alterations may also be reflected by altered protein expression in human cancers, we sought to determine what the empirical concordance of the spatial proteome and genome in human cancer samples. We performed a systematic review of Immunohistochemistry (IHC) & Fluorescence/Chromogenic In-Situ Hybridization (FISH/CISH) across human cancers. We performed a PUBMED query for records up to April 1st, 2023, yielding 9,294 records. Our inclusion criteria included human sample studies that performed both IHC and FISH/CISH against the same target, with a minimum of two samples. We excluded case reports, RNA ISH- and non-human studies, studies without IHC/ISH performed on the same target, and articles not in English or inaccessible. We screened 9,294 PUBMED article abstracts using the AS Review active learning model (van de Schoot et al., 2021), and used a pre-defined threshold of 5 % consecutive irrelevant abstracts to stop screening. A total of 1,309 abstracts were manually excluded using the AS Review, directing the active learning model to automatically exclude 3,413 abstracts out of the original 9,294. Of the 4,572 articles assessed for full-article screening, 1,167 were excluded for being out of scope, 303 for access issues, 227 for language incompatibility, and 236 for imprecise information, primarily due to non-representative samples or incorrect gene/protein target identification. Ultimately, out of 2,413 articles assessed for eligibility, 2,294 did not meet inclusion criteria, leaving 119 studies for our analysis. From these articles, we manually extracted kappa values for analysis and visualization in R. We identified articles from 19 different cancers with the majority studying Breast Cancer, Non-Small Cell Lung Cancer, Gastric Cancer, and Mesothelioma. We observed high Cohen’s Kappa values in Breast Cancer (median = 0.76), Non-Small Cell Lung Cancer (median = 0.91) and Mucinous Ovarian Carcinoma (median = 0.93). We identified studies from 14 different genes with majority being HER2, ALK and MYC. We observed high Cohen’s Kappa values in HER2 (median = 0.80), ALK (median = 0.80), ROS1 (median = 0.95) and MDM2 (median = 0.72). Our preliminary analysis indicates significant heterogeneity between tumor types for protein expression and copy-number alterations. To our knowledge, there are no previously reported systematic reviews that have empirically assessed the spatial concordance of proteome and genome, in human cancer samples. The authors often reported IHC/FISH results at patient, or section level, potentially introducing selection bias by not accounting for intra-tumoral heterogeneity at single-cell resolution. Future efforts are needed to spatially resolve gene-copy number and protein expression at single-cell resolution. Citation Format: Laura K. Sysivirta, Franziska Niemeyer, Veera Kurki, Inkalilja Pirinen, Laura Savolainen, Ceren Seref-Vujaklilja, Andrew Erickson. Spatial concordance between genome and proteome in oncology: A systematic review [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 1041.

Abstract 1495: Lipocalin-2 promotes breast cancer metastasis via tumor cell intrinsic and extrinsic mechanisms

Abstract While metastasis is the primary factor predicting poor outcomes in solid tumors, survival is lowest in patients initially diagnosed with a localized malignancy that then progresses to metastatic disease. Notably, nearly 25% of early-stage breast cancers (BCs) metastasize during treatment or following an initial positive response to therapeutic intervention. To define the molecular and cellular factors that contribute to metastatic progression of BC, we previously developed the Py230-C57Bl/6 syngeneic experimental metastasis model for interrogating mechanisms by which soluble factors condition premetastatic lung tissue to potentiate BC cell seeding/expansion. Using this model and the TCGA, we identified lipocalin-2 (Lcn2) as a molecular candidate that is upregulated in lung-tropic Py230 cells and other human triple-negative breast cancer (TNBC) cells, and that predicts poor outcomes in human BCs. While Lcn2 has been previously reported to mediate BC progression in genetically engineered mouse models, its mechanism(s) of action remain undefined. Here, we demonstrate that Py230 cell-secreted Lcn2 is necessary for optimal reprogramming of the premetastatic lung during metastatic progression. In agreement with these data, we defined a cell intrinsic role for Lcn2 in BC cell organoid assembly/invasion and expression of proinflammatory factors. To define BC cell extrinsic roles for Lcn2 in remodeling the premetastatic lung, we collected global transcriptome and single-cell spatial proteome data for lung tissue from C57Bl/6 mice systemically educated with mock media or media conditioned with Py8119 (Lcn2low) or Py230 (Lcn2hi) BC cells. Mice receiving Py230 conditioned media treatments were further divided between two groups – one receiving an isotype control IgG and the other receiving an Lcn2 immunoneutralizing antibody. Global transcriptome data was used to determine novel gene set enrichment signatures and clinically relevant interactomes – notable candidates predicted to cooperate with Lcn2 in decreasing patient survival were affiliated with cilia biogenesis and function. Single-cell spatial proteome data were analyzed for 28 different cell populations within the premetastatic lung. Interestingly, Py230 cell-derived Lcn2 did not affect the percentage of ciliated epithelial cells indicating that the global transcriptome data referenced above represent a real increase in expression of cilia-related genes and not an increase in cilia-positive cell types. Notably, we identified that Lcn2 increased the percentage of CD45neg/Despos mechanomodulatory cells and proliferating CD45pos/CD3pos/CD4neg/CD8pos cytotoxic T cells in the premetastatic lung. Taken together, these data define new targets around which new therapeutic strategies can be developed to improve BC patient outcomes. Citation Format: Joshua Gamez, Francesca Sanchez, Gabriela Ortiz-Soto, Jane Cox, Amir Gerami, Jonathan A. Kelber. Lipocalin-2 promotes breast cancer metastasis via tumor cell intrinsic and extrinsic mechanisms [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 1495.

Abstract 3988: Deep spatial immunophenotyping of lymphoid aggregates in pancreatic cancer using multi-omic integration of ultra high-plex proteomics and transcriptomics

Abstract The most well-described mechanism of immune evasion in PDAC is exclusion of immune cells by a desmoplastic stroma, but such immune exclusion is not uniform. Through single-nucleus RNA-seq and whole-transcriptome digital spatial profiling of patient tumors, we previously reported marked heterogeneity in the immune composition of PDAC. Here, we developed a novel multi-omics workflow integrating single cell spatial proteomics (96-plex, Akoya Phenocycler®-Fusion) and transcriptomics (990-plex, Nanostring CosMx) in primary patient PDAC tumors to systematically study spatial relationships and functional states of malignant, stromal, and immune cells. Consecutive sections from four independent tumors enabled whole slide proteomic assessment of >3.5M cells, and transcriptomic assessment of >200K cells. Strikingly, we observed an abundance of immune cell aggregates in all tissue sections, up to 57 within a 1 cm2 tumor section. Aggregates were diverse in composition with representation of myeloid cells, CD8+ T cells, and stromal cells, including fibroblasts and nerves. B/T lymphoid aggregates were well vascularized with co-localized CD31+CD34+ endothelial cells and appeared at various phases of maturation towards putative tertiary lymphoid structures (TLSs) containing CD38+ B cells, CD4+ T cells, and CD21+ follicular DCs. These tissues were enriched for interferon stimulated genes in T cells and type 1 interferon response signatures in malignant cells and cancer associated fibroblasts (CAFs) forming inflammatory multicellular hubs that have been associated with improved response to immune checkpoint blockade. Surprisingly, B cells, in addition to macrophages and CD8+ T cells, were major producers of interferon gamma (B cells: 13.7%, macrophages: 36.6%, CD8+ T cells: 12.1% of IFNG+ cells). Finally, to gain insight into the necessary factors for TLS formation in PDAC, we quantified the expression of chemokines and adhesion molecules associated with TLS initiation. Activated dendritic cells expressed CCL19 and CCL21, endothelial cells expressed CCL21, and in concert with inflammatory CAFs (iCAFs), CXCL12. Malignant cells expressed CXCL13 and CCL20 and displayed high levels of cell adhesion molecules ICAM1, VCAM1, and ICAM2. Together, our data suggests that the functions of lymphoid tissue organizer cells can be substituted by endothelial cells, dendritic cells, iCAFs, and, surprisingly, malignant cells. Our spatial immunophenotyping study of pancreatic cancer thus highlights the marked immune heterogeneity of the disease and suggests a central role for malignant and stromal cells in mediating diverse mechanisms of immune evasion and engagement. Functional studies will be conducted to investigate these mechanisms of tumor-immune crosstalk and identify novel immunotherapeutic vulnerabilities for this deadly disease. Citation Format: Dennis Gong, Jingyi Cao, Peter Wang, Niyati Jhaveri, Yue Hou, Bassem B. Cheikh, HaYeun Ji, Mark Gregory, Jason Reeves, Michael Patrick, Carina Shiau, Jennifer Su, Jimmy A. Guo, Joseph M. Beechem, Martin Hemberg, William L. Hwang, Ryan Park. Deep spatial immunophenotyping of lymphoid aggregates in pancreatic cancer using multi-omic integration of ultra high-plex proteomics and transcriptomics [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 3988.