Cellular Heterogeneity Research Articles

HyGAnno: hybrid graph neural network-based cell type annotation for single-cell ATAC sequencing data.

Reliable cell type annotations are crucial for investigating cellular heterogeneity in single-cell omics data. Although various computational approaches have been proposed for single-cell RNA sequencing (scRNA-seq) annotation, high-quality cell labels are still lacking in single-cell sequencing assay for transposase-accessible chromatin (scATAC-seq) data, because of extreme sparsity and inconsistent chromatin accessibility between datasets. Here, we present a novel automated cell annotation method that transfers cell type information from a well-labeled scRNA-seq reference to an unlabeled scATAC-seq target, via a parallel graph neural network, in a semi-supervised manner. Unlike existing methods that utilize only gene expression or gene activity features, HyGAnno leverages genome-wide accessibility peak features to facilitate the training process. In addition, HyGAnno reconstructs a reference-target cell graph to detect cells with low prediction reliability, according to their specific graph connectivity patterns. HyGAnno was assessed across various datasets, showcasing its strengths in precise cell annotation, generating interpretable cell embeddings, robustness to noisy reference data and adaptability to tumor tissues.

Topological and geometric analysis of cell states in single-cell transcriptomic data.

Single-cell RNA sequencing (scRNA-seq) enables dissecting cellular heterogeneity in tissues, resulting in numerous biological discoveries. Various computational methods have been devised to delineate cell types by clustering scRNA-seq data, where clusters are often annotated using prior knowledge of marker genes. In addition to identifying pure cell types, several methods have been developed to identify cells undergoing state transitions, which often rely on prior clustering results. The present computational approaches predominantly investigate the local and first-order structures of scRNA-seq data using graph representations, while scRNA-seq data frequently display complex high-dimensional structures. Here, we introduce scGeom, a tool that exploits the multiscale and multidimensional structures in scRNA-seq data by analyzing the geometry and topology through curvature and persistent homology of both cell and gene networks. We demonstrate the utility of these structural features to reflect biological properties and functions in several applications, where we show that curvatures and topological signatures of cell and gene networks can help indicate transition cells and the differentiation potential of cells. We also illustrate that structural characteristics can improve the classification of cell types.

Single-cell RNA sequencing data imputation using bi-level feature propagation.

Single-cell RNA sequencing (scRNA-seq) enables the exploration of cellular heterogeneity by analyzing gene expression profiles in complex tissues. However, scRNA-seq data often suffer from technical noise, dropout events and sparsity, hindering downstream analyses. Although existing works attempt to mitigate these issues by utilizing graph structures for data denoising, they involve the risk of propagating noise and fall short of fully leveraging the inherent data relationships, relying mainly on one of cell-cell or gene-gene associations and graphs constructed by initial noisy data. To this end, this study presents single-cell bilevel feature propagation (scBFP), two-step graph-based feature propagation method. It initially imputes zero values using non-zero values, ensuring that the imputation process does not affect the non-zero values due to dropout. Subsequently, it denoises the entire dataset by leveraging gene-gene and cell-cell relationships in the respective steps. Extensive experimental results on scRNA-seq data demonstrate the effectiveness of scBFP in various downstream tasks, uncovering valuable biological insights.

ScEWE: high-order element-wise weighted ensemble clustering for heterogeneity analysis of single-cell RNA-sequencing data.

With the emergence of large amount of single-cell RNA sequencing (scRNA-seq) data, the exploration of computational methods has become critical in revealing biological mechanisms. Clustering is a representative for deciphering cellular heterogeneity embedded in scRNA-seq data. However, due to the diversity of datasets, none of the existing single-cell clustering methods shows overwhelming performance on all datasets. Weighted ensemble methods are proposed to integrate multiple results to improve heterogeneity analysis performance. These methods are usually weighted by considering the reliability of the base clustering results, ignoring the performance difference of the same base clustering on different cells. In this paper, we propose a high-order element-wise weighting strategy based self-representative ensemble learning framework: scEWE. By assigning different base clustering weights to individual cells, we construct and optimize the consensus matrix in a careful and exquisite way. In addition, we extracted the high-order information between cells, which enhanced the ability to represent the similarity relationship between cells. scEWE is experimentally shown to significantly outperform the state-of-the-art methods, which strongly demonstrates the effectiveness of the method and supports the potential applications in complex single-cell data analytical problems.

Unveiling the functional heterogeneity of cytokine-primed human umbilical cord mesenchymal stem cells through single-cell RNA sequencing

BackgroundMesenchymal stem cells (MSCs) hold immense promise for use in immunomodulation and regenerative medicine. However, their inherent heterogeneity makes it difficult to achieve optimal therapeutic outcomes for a specific clinical disease. Primed MSCs containing a certain cytokine can enhance their particular functions, thereby increasing their therapeutic potential for related diseases. Therefore, understanding the characteristic changes and underlying mechanisms of MSCs primed by various cytokines is highly important.ResultsIn this study, we aimed to reveal the cellular heterogeneity, functional subpopulations, and molecular mechanisms of MSCs primed with IFN-γ, TNF-α, IL-4, IL-6, IL-15, and IL-17 using single-cell RNA sequencing (scRNA-seq). Our results demonstrated that cytokine priming minimized the heterogeneity of the MSC transcriptome, while the expression of MSC surface markers exhibited only slight changes. Notably, compared to IL-6, IL-15, and IL-17; IFN-γ, TNF-α, and IL-4 priming, which stimulated a significantly greater number of differentially expressed genes (DEGs). Functional analysis, which included Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses, indicated that IFN-γ, TNF-α, and IL-4-primed hUC-MSCs are involved in interferon-mediated immune-related processes, leukocyte migration, chemotaxis potential, and extracellular matrix and cell adhesion, respectively. Moreover, an investigation of various biological function scores demonstrated that IFN-γ-primed hUC-MSCs exhibit strong immunomodulatory ability, TNF-α-primed hUC-MSCs exhibit high chemotaxis potential, and IL-4-primed hUC-MSCs express elevated amounts of collagen. Finally, we observed that cytokine priming alters the distribution of functional subpopulations of MSCs, and these subpopulations exhibit various potential biological functions. Taken together, our study revealed the distinct regulatory effects of cytokine priming on MSC heterogeneity, biological function, and functional subpopulations at the single-cell level.ConclusionsThese findings contribute to a comprehensive understanding of the inflammatory priming of MSCs, paving the way for their precise treatment in clinical applications.

Image-based profiling and deep learning reveal morphological heterogeneity of colorectal cancer organoids

Patient-derived organoids have proven to be a highly relevant model for evaluating of disease mechanisms and drug efficacies, as they closely recapitulate in vivo physiology. Colorectal cancer organoids, specifically, exhibit a diverse range of morphologies, which have been analyzed with image-based profiling. However, the relationship between morphological subtypes and functional parameters of the organoids remains underexplored. Here, we identified two distinct morphological subtypes (“cystic” and “solid”) across 31360 bright field images using image-based profiling, which correlated differently with viability and apoptosis level of colorectal cancer organoids. Leveraging object detection neural networks, we were able to categorize single organoids achieving higher viability scores as “cystic” than “solid” subtype. Furthermore, a deep generative model was proposed to predict apoptosis intensity based on a apoptosis-featured dataset encompassing over 17000 bright field and matched fluorescent images. Notably, a significant correlation of 0.91 between the predicted value and ground truth was achived, underscoring the feasibility of this generative model as a potential means for assessing organoid functional parameters. The underlying cellular heterogeneity of the organoids, i.e., conserved colonic cell types and rare immune components, was also verified with scRNA sequencing, implying a compromised tumor microenvironment. Additionally, the “cystic” subtype was identified as a relapse phenotype featuring intestinal stem cell signatures, suggesting that this visually discernible relapse phenotype shows potential as a novel biomarker for colorectal cancer diagnosis and prognosis. In summary, our findings demonstrate that the morphological heterogeneity of colorectal cancer organoids explicitly recapitulate the association of phenotypic features and exogenous perturbations through the image-based profiling, providing new insights into disease mechanisms.

Single-cell profiling transcriptomic reveals cellular heterogeneity and cellular crosstalk in choroidal neovascularization model

Choroidal neovascularization (CNV) is a hallmark of neovascular age-related macular degeneration (nAMD) and a major contributor to vision loss in nAMD cases. However, the identification of specific cell types associated with nAMD remains challenging. Herein, we performed single-cell sequencing to comprehensively explore the cellular diversity and understand the foundational components of the retinal pigment epithelium (RPE)/choroid complex. We unveiled 10 distinct cell types within the RPE/choroid complex. Notably, we observed significant heterogeneity within endothelial cells (ECs), fibroblasts, and macrophages, underscoring the intricate nature of the cellular composition in the RPE/choroid complex. Within the EC category, four distinct clusters were identified and EC cluster 0 was tightly associated with choroidal neovascularization. We identified five clusters of fibroblasts actively involved in the pathogenesis of nAMD, influencing fibrotic responses, angiogenic effects, and photoreceptor function. Additionally, three clusters of macrophages were identified, suggesting their potential roles in regulating the progression of nAMD through immunomodulation and inflammation regulation. Through CellChat analysis, we constructed a complex cell-cell communication network, revealing the role of EC clusters in interacting with fibroblasts and macrophages in the context of nAMD. These interactions were found to govern angiogenic effects, fibrotic responses, and inflammatory processes. In summary, this study reveals noteworthy cellular heterogeneity in the RPE/choroid complex and provides valuable insights into the pathogenesis of CNV. These findings will open up potential avenues for deep understanding and targeted therapeutic interventions in nAMD.

Optimization of mouse kidney digestion protocols for single-cell applications.

Single-cell technologies such as flow cytometry and single-cell RNA sequencing have allowed for comprehensive characterization of the kidney cellulome. However, there is a disparity in the various protocols for preparing kidney single-cell suspensions. We aimed to address this limitation by characterizing kidney cellular heterogeneity using three previously published single-cell preparation protocols. Single-cell suspensions were prepared from male and female C57BL/6 kidneys using the following kidney tissue dissociation protocols: a scRNAseq protocol (P1), a multi-tissue digestion kit from Miltenyi Biotec (P2), and a protocol established in our laboratory (P3). Following dissociation, flow cytometry was used to identify known major cell types including leukocytes (myeloid and lymphoid), vascular cells (smooth muscle and endothelial), nephron epithelial cells (intercalating, principal, proximal, and distal tubule cells), podocytes, and fibroblasts. Of the protocols tested, P2 yielded significantly less leukocytes and type B intercalating cells compared with the other techniques. P1 and P3 produced similar yields for most cell types; however, endothelial and myeloid-derived cells were significantly enriched using P1. Significant sex differences were detected in only two cell types: granulocytes (increased in males) and smooth muscle cells (increased in females). Future single-cell studies that aim to enrich specific kidney cell types may benefit from this comparative analysis.NEW & NOTEWORTHY This study is the first to evaluate published single-cell suspension preparation protocols and their ability to produce high-quality cellular yields from the mouse kidney. Three single-cell digestion protocols were compared and each produced significant differences in kidney cellular heterogeneity. These findings highlight the importance of the digestion protocol when using single-cell technologies. This study may help future single-cell science research by guiding researchers to choose protocols that enrich certain cell types of interest.

Preclinical Advances in LNP-CRISPR Therapeutics for Solid Tumor Treatment.

Solid tumors, with their intricate cellular architecture and genetic heterogeneity, have long posed therapeutic challenges. The advent of the CRISPR genome editing system offers a promising, precise genetic intervention. However, the journey from bench to bedside is fraught with hurdles, chief among them being the efficient delivery of CRISPR components to tumor cells. Lipid nanoparticles (LNPs) have emerged as a potential solution. This biocompatible nanomaterial can encapsulate the CRISPR/Cas9 system, ensuring targeted delivery while mitigating off-target effects. Pre-clinical investigations underscore the efficacy of LNP-mediated CRISPR delivery, with marked disruption of oncogenic pathways and subsequent tumor regression. Overall, CRISPR/Cas9 technology, when combined with LNPs, presents a groundbreaking approach to cancer therapy, offering precision, efficacy, and potential solutions to current limitations. While further research and clinical testing are required, the future of personalized cancer treatment based on CRISPR/Cas9 holds immense promise.

Single cell transcriptome profiling of infrapatellar fat pad highlights the role of interstitial inflammatory fibroblasts in osteoarthritis

ObjectivesOsteoarthritis (OA) is a whole-joint disease in which the role of the infrapatellar fat pad (IFP) in its pathogenesis is unclear. Our study explored the cellular heterogeneity of IFP to understand OA and identify therapeutic targets. MethodsSingle-cell and single-nuclei RNA sequencing were used to analyze 10 IFP samples, comprising 5 from OA patients and 5 from healthy controls. Analyses included differential gene expression, enrichment, pseudotime trajectory, and cellular communication, along with comparative studies with visceral and subcutaneous fats. Key subcluster and pathways were validated using multiplex immunohistochemistry. ResultsThe scRNA-seq performed on the IFPs of the OA and control group profiled the gene expressions of over 49,674 cells belonging to 11 major cell types. We discovered that adipose stem and progenitor cells (ASPCs), contributing to the formation of both adipocytes and synovial-lining fibroblasts (SLF). Interstitial inflammatory fibroblasts (iiFBs) were a subcluster of ASPCs that exhibit notable pro-inflammatory and proliferative characteristics. We identified four adipocyte subtypes, with one subtype showing a reduced lipid synthesis ability. Furthermore, iiFBs modulated the activities of macrophages and T cells in the IFP. Compared to subcutaneous and visceral adipose tissues, iiFBs represented a distinctive subpopulation of ASPCs in IFP that regulated cartilage proliferation through the MK pathway. ConclusionThis study presents a comprehensive single-cell transcriptomic atlas of IFP, uncovering its complex cellular landscape and potential impact on OA progression. Our findings highlight the role of iiFBs in OA, especially through MK pathway, opening new avenues for understanding OA pathogenesis and developing novel targeted therapies.

Abstract 5674: Characterizing Cellular Heterogeneity and Identifying Potential Biomarkers in Esophageal Squamous Cell Carcinoma (ESCC) Using Single-Cell RNA Sequencing

Abstract Esophageal squamous cell carcinoma (ESCC) exhibits significant cellular heterogeneity, rendering it susceptible to developing radioresistance and recurring tumors. Single-cell RNA sequencing (scRNA-seq) is an advanced technique that allows for the exploration of distinct gene expression profiles in individual cells. To investigate cellular heterogeneity in ESCC, scRNA-seq analysis was performed on tumor tissues from ten ESCC patients, as well as adjacent non-malignant esophageal tissues. The cellular composition revealed 19 distinct cell clusters, further categorized into seven cell populations: 69.31% Fibroblast, 15.36% T cell, 5.30% Monocytes, 5.26% Endothelial cells, 2.68% plasma cell, 1.91% Epithelial cells, and 0.18% ambiguous. Notably, Fibroblast and Endothelial cells were more prevalent in non-malignant tissues, while the remaining cell types were enriched in tumor tissues. Differential gene expression analysis identified 49 genes with higher expression in tumor cells and 37 genes with lower expression. Gene Set Enrichment Analysis (GSEA) revealed the top three most relevant GO Biological Processes pathways: positive regulation of immune response, regulation of T cell activation, and positive regulation of cell migration. Furthermore, in post-treatment follow-up, four patients were observed to develop metastasis within one year. Comparative analysis between metastatic and non-metastatic patients unveiled 9 Differentially Expressed Genes (DEGs), including KRT6A and AKR1B10, which exhibited specific expression in Epithelial cells and were highly expressed in tumor tissues. Additionally, IGLC2, IGHG4, and IGHG1 demonstrated specific expression in plasma cells, with elevated expression in tumor tissues. These findings suggest that KRT6A, AKR1B10, IGLC2, IGHG4, and IGHG1 may serve as potential biomarkers for predicting post-chemotherapy metastasis in ESCC patients. Citation Format: Tzu-Hung Hsiao, Chih-Hung Lin, Ting-Shuan Wu, Li-Wen Lee, Chung-Ping Hsu. Characterizing Cellular Heterogeneity and Identifying Potential Biomarkers in Esophageal Squamous Cell Carcinoma (ESCC) Using Single-Cell RNA Sequencing [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 5674.

Abstract 6853: Spatial proteomic and transcriptomic characterization of the inflammatory landscape of formalin fixed paraffin embedded Barrett’s esophagus tissues

Abstract Esophageal adenocarcinoma (EAC), a significant cause of morbidity and mortality, develops from its metaplastic precursor lesion termed Barrett’s esophagus (BE). BE progresses to EAC within an inflammation rich environment, however a detailed understanding of the types of inflammatory cells present, how that population changes over time or diagnosis, and the related immunomodulatory pathways that may mediate BEs progression is still understudied. Traditional immunohistochemistry is often inadequate to provide detailed information about a broad number of cells, local cellular proportions, and cellular heterogeneity; while bulk or single cell RNA sequencing lose the spatial context and relationships between the stroma and epithelium. Utilizing modern multiplex profiling technologies such as Nanostring’s GeoMx digital Profiler (DSP) aid in identifying both cellular composition and distribution of cells within the epithelium and stroma. Therefore, our aim of this study was to determine the suitability of the DSP platform for analyzing formalin fixed paraffin embedded (FFPE) endoscopic samples of BE and determine any differences between non-dysplastic BE (NDBE) samples from patients who eventually progressed (progressors) compared to those who did not (non-progressors), with the goal to gain deeper insights about cellular composition and spatial distribution of immune cells that may dictate progression of BE. Our secondary goal was to investigate the relationship of these cells to the BE epithelial cells through understanding the immune signaling and regulatory pathways. Our pilot study utilizing a 49-plex antibody panel on 7 NDBE samples from progressors and 8 NDBE samples from non-progressors showed good quality metrics and deciphered an immune rich microenvironment in BE with several immune markers including CD66b (neutrophils), CD14 (monocytes), CD68 (macrophages) and CD56 (NK cells) being overexpressed in the stroma of progressors compared with non-progressors, corrected p value less than 0.05. The myeloid activation marker CD80 was overexpressed in the BE epithelial cells of progressors. To determine how spatial transcriptomic profiling performed in FFPE endoscopic samples, 1,812 transcripts within 48 regions of interest of four BE, HGD, and EAC samples was interrogated. Only 36 (2%) probes fell below the limit of detection quality threshold. Thus, our pilot results show that robust proteomic and transcriptomic profiling can be obtained from endoscopic BE biopsies using this system. Based on these results, we systematically interrogated the immune cell makeup and distribution as well as their related inflammatory and immunoregulatory pathways in 40 NDBE biopsies (20 progressors and 20 with stable non-progressing disease) within discrete compartments (stroma and epithelium) through spatial whole transcriptome profiling. Citation Format: Qurat ul Ain, Nicola Frei, Amir M. Khoshiwal, Pim Stougie, Robert Odze, Sophie Camilleri-Broet, Lorenzo Ferri, Lucas C. Duits, Jacques Bergman, Matthew D. Stachler. Spatial proteomic and transcriptomic characterization of the inflammatory landscape of formalin fixed paraffin embedded Barrett’s esophagus tissues [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 6853.

Abstract 1271: PMEPA1 acts as a switch to modulate cooperative cellular invasion and drive NSCLC tumor progression

Abstract Metastatic disease drives cancer patient mortality. One primary mode of metastasis is collective invasion, whereby cohesive groups of cells invade into the adjacent stroma while maintaining cell-cell contacts. Importantly, these cellular packs harbor genetically and phenotypically heterogeneous subpopulations that cooperate to drive invasion. To deconstruct how phenotypic cellular heterogeneity facilitates distinct molecular profiles within a single tumor, we established the technique Spatiotemporal Cellular and Genomic Analysis (SaGA). SaGA is an image-guided approach that optically highlights phenotypically defined cell(s) for live fluorescence-activated cell sorting and analysis. To elucidate the transcriptional heterogeneity across invasive packs within a 3D physiologically-relevant microenvironment, we combined SaGA with single cell RNA sequencing for isolation and analysis of actively invading cells. This novel single cell transcriptomics approach reveals that - during cellular invasion into the microenvironment - the follower subpopulation supplies TGFβ to heterogeneously stimulate both autocrine canonical and paracrine noncanonical signaling pathways. This cooperative exchange facilitates a symbiotic relationship between tumor subpopulations in non-small cell lung carcinoma both in vitro and in vivo. In vitro, we determine that PMEPA1 protein levels mitigate TGFβ downstream effector signaling within heterogeneous tumor cells to modulate JAG1 expression and activity. In vivo, we show that a reduction in JAG1 levels decreases primary tumor cell invasion and progression. We mechanistically define a TGFβ-PMEPA1-JAG1 signaling axis across leader and follower subpopulations critical for tumor cell invasion and cancer progression. Together, these data highlight the importance of cellular heterogeneity and subpopulation communication required for tumor progression. Citation Format: Tala O. Khatib, Brian A. Pedro, Christina M. Knippler, Veronika Y. Matsuk, Sarah F. Webster, Sergei Bombin, Bhakti Dwivedi, Jeanne Kowalski, Henry R. Johnston, Janna K. Mouw, Adam I. Marcus. PMEPA1 acts as a switch to modulate cooperative cellular invasion and drive NSCLC tumor progression [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 1271.

Abstract 4927: Combining single-cell ATAC and RNA sequencing for supervised cell annotation

Abstract Background: Analysis of samples at the single-cell level offers insights into cellular heterogeneity and cell function. Cell type annotation is the first critical step for performing such an analysis. While current methods primarily utilize single-cell RNA sequencing (scRNA-seq) for annotation, several studies have demonstrated improved classification accuracy by combining scRNA-seq with transposase-accessible chromatin sequencing (ATAC-seq) using unsupervised methods. However, the utility of ATAC-seq features for supervised cell-type annotation has not been explored. Aims/Objectives: The objective of this study was to evaluate the relative performance of supervised cell-type classification using scRNA-seq alone vs. in multimodal combination with ATAC-seq; and how these data interplay with choice of classification and dimensionality reduction methods. Methods: A peripheral-blood mononuclear cell multi-omic dataset from a single, healthy female donor wasanalysed in this study. Ground truth annotations were generated using unsupervised annotation with the weighted nearest neighbour clustering method. Two dimensionality reduction methods (principal component analysis (PCA), single-cell Variational Inference (scVI) autoencoder) and four classification models (logistic regression, random forest, support vector machine (SVM)) were implemented and performance metrics (F1 score, precision, and recall) were compared over 10 bootstrap samples. Results: ATAC-seq features improved annotation quality and prediction confidence when using scVI embeddings, independent of the classifier. The best-performing model (SVM with scVI embeddings) showed an increase from a median macro F1 score of 0.907 (IQR = [0.902, 0.910]) using scRNA-seq alone to 0.946 (IQR = [0.940, 0.949], p <0.05) with ATAC-seq added. For PCA embeddings, improvements in macro F1 score were insignificant. All cell types (B, T, monocytes, natural killer and dendritic cells) showed significant improvements when using ATAC-seq with scVI embeddings. CD4 T effector memory cells showed the largest gain in F1 score (0.112, p <0.01), whilst type-2 conventional dendritic cells showed the smallest improvement (0.006, p <0.05). Prediction confidence was improved in B cells, monocytes, natural killer cells, CD4 and CD8 naïve cells, CD4 T central memory cells and CD8 T effector memory cells. Improvements in F1 scores were lost when only classifying major cell types rather than subtypes. Conclusions: Employing ATAC-seq embeddings with scVI autoencoder enhances supervised annotation quality over scRNA-only methods. Further studies should explore the use of ATAC to improve the annotation of highly heterogeneous tissues such as tumours. Citation Format: Jaidip Gill, Abhijit Dasgupta, Brychan Manry, Natasha Markuzon. Combining single-cell ATAC and RNA sequencing for supervised cell annotation [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 4927.

Abstract 4911: PAIRWISE: Deep learning-powered personalized medicine approach identifies actionable drug combinations in B cell lymphoma

Abstract Combination drug therapies offer promise for improving cancer treatment. However, identifying actionable drug combinations for specific cancer subtypes and individual patients remains challenging due to the vast number of possibilities. Existing approaches like high-throughput screening (HTS) and rule-based drug additions have limitations in their ability to predict clinically translatable synergistic combinations. Therefore, advanced machine learning focused on patient-centric modelling have the potential to optimize combination therapies. PAIRWISE is a patient-centric deep learning-powered drug combination recommendation platform. Given gene expression along with chemical structures and targets from a set of two drugs being applied, the output is the binary combination outcome. PAIRWISE introduces a new approach of representing tumor genomics using the unsupervised pretraining design on 11K TCGA tumors. The power of PAIRWISE is derived from knowledge-guided transfer learning by learning from in vitro experiments with drug-drug-cell interactions and cellular heterogeneity. PAIRWISE can accurately predict combination synergy with area under the curve (AUC) > 0.85 across 10 tissue types in a large dataset of 200K in vitro drug combinations and significantly outperform all state-of-the-art methods. Diffuse large B cell lymphoma (DLBCL) is the most common B cell non-Hodgkin lymphoma and remains incurable in around 40% of patients. Finding optimal combination therapy for each DLBCL patient is a critical unmet need. As independent validation, we applied PAIRWISE to an external HTS study of Bruton Tyrosine Kinase inhibitor (BTKi) combinations (Griner et al., 2014), confirming the reproducibility and accuracy of high-confidence predictions (AUC = 0.72). PAIRWISE was also applied to transcriptomic data of 29 DLBCL cell lines (Bal et al., 2022) and predicted BTKi-containing combinations across ABC subtypes with higher synergy scores than GCB (P = 0.02). This is consistent with expected better overall efficacy of BTK inhibitors in ABC subtype (Younes et al., 2019; Davis et al., 2010; Wilson et al., 2015). To predict opportunities in DLBCL, PAIRWISE was used to identify BTKi partners for an additional 211 anti-cancer drugs using 562 published ex vivo DLBCL samples (Schmitz et al., 2018). PAIRWISE effectively identified novel DLBCL sub-population with significantly higher response scores for predicted BTKi combinations and favorable survival. Notably, patients predicted as responsive had significantly better outcomes upon first-line chemotherapy treatments than non-responsive ones (P < 0.0001). Collectively, we demonstrate that PAIRWISE is a highly efficient and generalizable framework that enables confident predictions of pairwise drug combinations at disease subtype level, with the potential of informing treatment options and clinical trial design. Citation Format: Chengqi Xu, Jake Cohen-Setton, Zhouyang Shen, Marta Milo, Ilkay US, Ben Sidders, Jude Fitzgibbon, Heng Pan, Olivier Elemento, Krishna Bulusu. PAIRWISE: Deep learning-powered personalized medicine approach identifies actionable drug combinations in B cell lymphoma [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 4911.

Monkey multi-organ cell atlas exposed to estrogen

Abstract Awareness of estrogen’s effects on health is broadening rapidly. The effects of long-term high levels of estrogen on the body involve multiple organs. Here, we used both single-cell chromatin accessibility and RNA sequencing data to analyze the potential effect of estrogen on major organs. The integrated cell map enabled in-depth dissection and comparison of molecular dynamics, cell-type compositions, and cellular heterogeneity across multiple tissues and organs under estrogen stimulation. We also inferred pseudotime cell trajectories and cell–cell communications to uncover key molecular signatures underlying their cellular processes in major organs in response to estrogen. For example, estrogen could induce the differentiation of IFIT3+ neutrophils into S100A9+ neutrophils involved in the function of endosome-to-lysosome transport and the multivesicular body sorting pathway in liver tissues. Furthermore, through integration with human genome-wide association study data, we further identified a subset of risk genes during disease development that were induced by estrogen, such as AKT1 (related to endometrial cancer), CCND1 (related to breast cancer), HSPH1 (related to colorectal cancer), and COVID-19 and asthma-related risk genes. Our work uncovers the impact of estrogen on the major organs, constitutes a useful resource, and reveals the contribution and mechanism of estrogen to related diseases.

Abstract 4938: Prediction of gene mutation from colorectal adenocarcinoma whole slide images via integrated deep learning pipeline

Abstract Colorectal cancer (CRC) stands as one of the deadliest cancer types worldwide. Identifying specific gene mutations crucial for targeted therapy is pivotal for CRC treatment. However, the current genetic testing method involves liquid biopsy, which is often cumbersome and cost-intensive. Histopathological whole-slide images (WSIs) are routinely generated for diagnostic purposes and can offer valuable insights into cellular heterogeneity. In this study, we aim to develop an integrated deep learning pipeline for predicting TP53 mutations in CRC, one of the most frequently mutated genes in CRC. We commenced by collecting 424 diagnostic WSIs from the TCGA-COAD cohorts. Each WSI was split as small patches at a resolution of 0.5 micrometers per pixel for following analyses. Quality control was implemented using Otsu thresholding and Gaussian blur filtering, both at the slide-level masking and patch-level filtering stages. Additionally, we applied the Macenko algorithm for color normalization to eliminate systematic errors. Following image preprocessing, we constructed a tissue-type classification model based on VGG19, which was trained and tested using the NCT-CRC-HE-100K and CRC-VAL-HE-7K datasets, both containing pixels with tissue types labeled by human experts. The trained classifier was then used to identify adenocarcinoma epithelium patches from TCGA-COAD. Subsequently, we applied multiple pre-trained Convolutional Neural Network (CNN) models, including ResNet-101, DenseNet-201, and Inception-ResNet-v2, to predict TP53 mutations for the identified adenocarcinoma epithelium patches from TCGA-COAD. A novel integration approach was employed to aggregate patch-level predictions into slide-level outcomes. We evaluated the trained models using various metrics, including accuracy, precision, recall, F1 score, AUC (Area Under the Curve), and Cohen's Kappa coefficient. The tissue-type classification model exhibited robust performance, achieving an overall accuracy of 94.7%, a Cohen's Kappa coefficient of 94%, a precision of 95.3%, a recall of 99.76%, and an F1 score of 97.48% for identifying adenocarcinoma epithelium patches. Regarding the TP53 mutation prediction models, ResNet-101, DenseNet-201, and Inception-ResNet-v2 exhibit the predictive ability to both patch level (AUC=65%) and slide level (AUC=70%). In summary, the proposed deep learning pipeline offers an efficient way to acquire gene mutation information from WSIs. Moreover, it demonstrates a capability to discover potential molecular aberration from the clinical images. Citation Format: Geng-Yun Tien, Liang-Chuan Lai, Tzu-Pin Lu, Mong-Hsun Tsai, Hsiang-Han Chen, Eric Y. Chuang. Prediction of gene mutation from colorectal adenocarcinoma whole slide images via integrated deep learning pipeline [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 4938.

Abstract 1179: A single-cell atlas of human gastric malignancy reveals CD8+ T cell rejuvenation as a mediator of immunotherapeutic response

Abstract Immune checkpoint inhibition (ICI) has emerged as a mainstream treatment for patients with advanced stage gastric cancer (GC). Cellular heterogeneity in the tumor microenvironment (TME) has a profound impact on the therapeutic responses to ICI. Here we characterize the kinetics of tumor-infiltrating immune populations in GC patients treated with anti-PD-1 using single-cell RNA and TCR sequencing (scRNA/TCR-seq), including primary tumors and metastases. Integrated with public GC scRNA-seq datasets, we construct the largest multi-omic atlas of GC cell states to date that captures the full trajectory of GC malignancy and comprises >320,000 cells derived from 73 samples. We demonstrate the rejuvenation of exhausted CD8+ T cells into a proliferative state marked by the loss of LAYN expression as a major mediator of ICI response in the TME. In parallel, we observe enhanced immunogenicity of cancer cells in post-ICI tumors, thus highlighting the coordination between tumor-intrinsic and microenvironmental factors elicited by ICI. Our study provides a valuable resource of investigating the tumor response to anti-PD-1 treatment and suggests novel strategies for combination therapy and patient stratification. Citation Format: Han Liang, Jian Chen, Yikai Luo, Muxing Kang, Wei Liu, Lie Wang. A single-cell atlas of human gastric malignancy reveals CD8+ T cell rejuvenation as a mediator of immunotherapeutic response [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 1179.

Abstract 5501: Novel whole slide imaging modes for imaging mass cytometry unveil extensive cellular heterogeneity in human glioma

Abstract Gliomas present a complex form of cancer that is challenging to diagnose and treat, with a median survival of just over one year after diagnosis for primary glioblastoma (GBM). GBM can occur as multi-lesion, remote, or diffuse tumors, and often contains a tumor microenvironment (TME) that is devoid of peripheral immune cells. Additional hallmark features of GBM include necrosis, hemorrhage, and pseudo palisades, making it a highly heterogeneous disease requiring further investigation. Identification of the cellular and spatial level composition of the TME is vital for interpretation of GBM disease origin, progression, prediction, and treatment. A 40-plus-marker neuro-oncology Imaging Mass Cytometry™ (IMC™) antibody panel was used to determine the cellular and structural landscape of the brain TME. We applied the panel on a tissue microarray (TMA) containing dozens of human glioma cores and uncovered the spatial distribution of over 40 distinct molecular markers. We performed imaging using two new features of the Hyperion XTi™ Imaging System that provide whole slide scanning capabilities. Ultrafast preview mode was applied to rapidly screen tumor cores for expression signatures associated with tumor immuno-oncology processes. This enabled biomarker-guided selection of areas in tumor tissue that were imaged at higher resolution and analyzed using single-cell analysis. In parallel, a high-throughput tissue mode was applied to perform a detailed scan of the brain tumor TMA followed by pixel-based analysis to unravel the spatial composition of the TME. Using tissue mode imaging, we successfully mapped the spatial location of cell populations making up human gliomas, such as neurons, astrocytes, microglia, and oligodendrocytes, across the entire TMA. Various tumor cell phenotypes, resident and infiltrating cells, and resting and activated microglia were detected across all TMA cores. Subsequent single-cell analysis of select regions of interest provided a quantitative assessment of the cellular composition of the brain TME. We classified the distinct states of neurons and quantified myeloid and lymphoid immune cell infiltration across normal, astrocytoma, and GBM tissues. Striking cellular and protein heterogeneity was observed between acquired cores, indicating the complexity of each case that is impossible to capture with low-plex visualization techniques. Therefore, IMC is a highly relevant tool capable of quantitative spatial evaluation of the high-plex protein composition in the brain TME without the complications of autofluorescence, tissue degradation, and spectral overlap. Empowered by the neuro-oncology panel and new whole slide imaging modes, IMC accelerates neurological research and provides insights into the spatial complexity of gliomas and other tumors. For research purpose only. Not for use in diagnostic procedures. Citation Format: Nick Zabinyakov, Qanber Raza, Thomas D. Pfister, Nikesh Parsotam, David Howell, Liang Lim, Christina Loh. Novel whole slide imaging modes for imaging mass cytometry unveil extensive cellular heterogeneity in human glioma [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 5501.

Abstract 3646: A spatially resolved single cell proteomic atlas of small bowel adenocarcinoma

Abstract Small bowel adenocarcinoma (SBA) is a rare malignancy associated with poor prognosis. The cellular and proteomic heterogeneity within the tumor immune microenvironment (TIME) of SBA is a likely driver of prognosis, disease progression and response to therapy. There is, however, a major gap in our knowledge of tumor and immune interactions in SBA. Cyclic Immunofluorescence (CycIF), an antibody-based, highly multiplexed imaging technology for spatially resolved single cell level proteomic profiling, is well suited to map the proteomic heterogeneity and organization of TIME in SBA. Here, using CycIF, we have generated a spatially resolved single-cell proteomic atlas of TIME for 136 tumor and normal samples from clinically and genomically annotated SBA patients (N=37). The SBA TIME Atlas covers information on expression levels, spatial position, and cell morphology for > 40 proteomic markers of tumor-intrinsic processes and states, diverse immune cell types, immune checkpoints, and tumor vascularization. In total, we collected proteomics data from > 600,000 cells. Using a hierarchical decision tree of cell type markers, we have generated the cell identity annotations for > 350,000 cells, with positional information. A comprehensive computational analysis of the SBA atlas revealed various spatial and heterogeneity features that may impact the TIME organization in fine detail. Using those features, we have compared the proteomic heterogeneity in tumor vs. normal cell populations, identifying spatial and single-cell correlates of key clinical variables (e.g., patient survival, age, sex, disease stage, metastatic status). We have also performed differential analyses of spatial features involving protein markers and cell types with respect to genomic alteration states and have mapped the spatial enrichment of likely actionable immune check proteins to immune and tumor cells. Such mapping may guide future therapeutic decisions. Overall, we expect that the SBA TIME Atlas will provide a useful resource and inform future translational studies, as well as basic research for better mechanistic understanding and treatment of this disease. Citation Format: Zeynep Dereli, Behnaz Bozorgui, Huamin Wang, John Weinstein, Michael Overman, Anil Korkut. A spatially resolved single cell proteomic atlas of small bowel adenocarcinoma [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 3646.