Recapitulating the tumor microenvironment in a dish, one cell type at a time.

The tumor microenvironment harbors a variety of different cell types that differentially impact tumor biology. In this issue of Cell Reports Methods, Raffo-Romero et al. standardized and optimized 3D tumor organoids to model the interactions between tumor-associated macrophages and tumor cells in vitro.

Decomprolute is a benchmarking platform designed for multiomics-based tumor deconvolution.

Tumor deconvolution enables the identification of diverse cell types that comprise solid tumors. To date, however, both the algorithms developed to deconvolve tumor samples, and the gold-standard datasets used to assess the algorithms are geared toward the analysis of gene expression (e.g., RNA sequencing) rather than protein levels. Despite the popularity of gene expression datasets, protein levels often provide a more accurate view of rare cell types. To facilitate the use, development, and reproducibility of multiomic deconvolution algorithms, we introduce Decomprolute, a Common Workflow Language framework that leverages containerization to compare tumor deconvolution algorithms across multiomic datasets. Decomprolute incorporates the large-scale multiomic datasets produced by the Clinical Proteomic Tumor Analysis Consortium (CPTAC), which include matched mRNA expression and proteomic data from thousands of tumors across multiple cancer types to build a fully open-source, containerized proteogenomic tumor deconvolution benchmarking platform. http://pnnl-compbio.github.io/decomprolute.

A co-culture system of macrophages with breast cancer tumoroids to study cell interactions and therapeutic responses.

3D tumoroids have revolutionized in vitro/ex vivo cancer biology by recapitulating the complex diversity of tumors. While tumoroids provide new insights into cancer development and treatment response, several limitations remain. As the tumor microenvironment, especially the immune system, strongly influences tumor development, the absence of immune cells in tumoroids may lead to inappropriate conclusions. Macrophages, key players in tumor progression, are particularly challenging to integrate into the tumoroids. In this study, we established three optimized and standardized methods for co-culturing human macrophages with breast cancer tumoroids: a semi-liquid model and two matrix-embedded models tailored for specific applications. We then tracked interactions and macrophage infiltration in these systems using flow cytometry and light sheet microscopy and showed that macrophages influenced not only tumoroid molecular profiles but also chemotherapy response. This underscores the importance of increasing the complexity of 3D models to more accurately reflect in vivo conditions.

Tongue orthotopic xenografts to study fusion-negative rhabdomyosarcoma invasion and metastasis in live animals.

PAX3/7 fusion-negative rhabdomyosarcoma (FN-RMS) is a childhood mesodermal lineage malignancy with a poor prognosis for metastatic or relapsed cases. Limited understanding of advanced FN-RMS is partially attributed to the absence of sequential invasion and dissemination events and the challenge in studying cell behavior, using, for example, non-invasive intravital microscopy (IVM), in currently used xenograft models. Here, we developed an orthotopic tongue xenograft model of FN-RMS to study cell behavior and the molecular basis of invasion and metastasis using IVM. FN-RMS cells are retained in the tongue and invade locally into muscle mysial spaces and vascular lumen, with evidence of hematogenous dissemination to the lungs and lymphatic dissemination to lymph nodes. Using IVM of tongue xenografts reveals shifts in cellular phenotype, migration to blood and lymphatic vessels, and lymphatic intravasation. Insight from this model into tumor invasion and metastasis at the tissue, cellular, and subcellular level can guide new therapeutic avenues for advanced FN-RMS.

Systematic analysis of proteome turnover in an organoid model of pancreatic cancer by dSILO.

The role of protein turnover in pancreatic ductal adenocarcinoma (PDA) metastasis has not been previously investigated. We introduce dynamic stable-isotope labeling of organoids (dSILO): a dynamic SILAC derivative that combines a pulse of isotopically labeled amino acids with isobaric tandem mass-tag (TMT) labeling to measure proteome-wide protein turnover rates in organoids. We applied it to a PDA model and discovered that metastatic organoids exhibit an accelerated global proteome turnover compared to primary tumor organoids. Globally, most turnover changes are not reflected at the level of protein abundance. Interestingly, the group of proteins that show the highest turnover increase in metastatic PDA compared to tumor is involved in mitochondrial respiration. This indicates that metastatic PDA may adopt alternative respiratory chain functionality that is controlled by the rate at which proteins are turned over. Collectively, our analysis of proteome turnover in PDA organoids offers insights into the mechanisms underlying PDA metastasis.

Cell-free DNA end characteristics enable accurate and sensitive cancer diagnosis.

The fragmentation patterns of cell-free DNA (cfDNA) in plasma can potentially be utilized as diagnostic biomarkers in liquid biopsy. However, our knowledge of this biological process and the information encoded in fragmentation patterns remains preliminary. Here, we investigated the cfDNA fragmentomic characteristics against nucleosome positioning patterns in hematopoietic cells. cfDNA molecules with ends located within nucleosomes were relatively shorter with altered end motif patterns, demonstrating the feasibility of enriching tumor-derived cfDNA in patients with cancer through the selection of molecules possessing such ends. We then developed three cfDNA fragmentomic metrics after end selection, which showed significant alterations in patients with cancer and enabled cancer diagnosis. By incorporating machine learning, we further built high-performance diagnostic models, which achieved an overall area under the curve of 0.95 and 85.1% sensitivity at 95% specificity. Hence, our investigations explored the end characteristics of cfDNA fragmentomics and their merits in building accurate and sensitive cancer diagnostic models.

Mimicking and analyzing the tumor microenvironment.

The tumor microenvironment (TME) is increasingly appreciated to play a decisive role in cancer development and response to therapy in all solid tumors. Hypoxia, acidosis, high interstitial pressure, nutrient-poor conditions, and high cellular heterogeneity of the TME arise from interactions between cancer cells and their environment. These properties, in turn, play key roles in the aggressiveness and therapy resistance of the disease, through complex reciprocal interactions between the cancer cell genotype and phenotype, and the physicochemical and cellular environment. Understanding this complexity requires the combination of sophisticated cancer models and high-resolution analysis tools. Models must allow both control and analysis of cellular and acellular TME properties, and analyses must be able to capture the complexity at high depth and spatial resolution. Here, we review the advantages and limitations of key models and methods in order to guide further TME research and outline future challenges.

Multiomics2Targets identifies targets from cancer cohorts profiled with transcriptomics, proteomics, and phosphoproteomics.

The availability of data from profiling of cancer patients with multiomics is rapidly increasing. However, integrative analysis of such data for personalized target identification is not trivial. Multiomics2Targets is a platform that enables users to upload transcriptomics, proteomics, and phosphoproteomics data matrices collected from the same cohort of cancer patients. After uploading the data, Multiomics2Targets produces a report that resembles a research publication. The uploaded matrices are processed, analyzed, and visualized using the tools Enrichr, KEA3, ChEA3, Expression2Kinases, and TargetRanger to identify and prioritize proteins, genes, and transcripts as potential targets. Figures and tables, as well as descriptions of the methods and results, are automatically generated. Reports include an abstract, introduction, methods, results, discussion, conclusions, and references and are exportable as citable PDFs and Jupyter Notebooks. Multiomics2Targets is applied to analyze version 3 of the Clinical Proteomic Tumor Analysis Consortium (CPTAC3) pan-cancer cohort, identifying potential targets for each CPTAC3 cancer subtype. Multiomics2Targets is available from https://multiomics2targets.maayanlab.cloud/.

iSubGen generates integrative disease subtypes by pairwise similarity assessment.

There are myriad types of biomedical data-molecular, clinical images, and others. When a group of patients with the same underlying disease exhibits similarities across multiple types of data, this is called a subtype. Existing subtyping approaches struggle to handle diverse data types with missing information. To improve subtype discovery, we exploited changes in the correlation-structure between different data types to create iSubGen, an algorithm for integrative subtype generation. iSubGen can accommodate any feature that can be compared with a similarity metric to create subtypes versatilely. It can combine arbitrary data types for subtype discovery, such as merging genetic, transcriptomic, proteomic, and pathway data. iSubGen recapitulates known subtypes across multiple cancers even with substantial missing data and identifies subtypes with distinct clinical behaviors. It performs equally with or superior to other subtyping methods, offering greater stability and robustness to missing data and flexibility to new data types. It is available at https://cran.r-project.org/web/packages/iSubGen.

Nextflow pipeline for Visium and H&E data from patient-derived xenograft samples.

We designed a Nextflow DSL2-based pipeline, Spatial Transcriptomics Quantification (STQ), for simultaneous processing of 10x Genomics Visium spatial transcriptomics data and a matched hematoxylin and eosin (H&E)-stained whole-slide image (WSI), optimized for patient-derived xenograft (PDX) cancer specimens. Our pipeline enables the classification of sequenced transcripts for deconvolving the mouse and human species and mapping the transcripts to reference transcriptomes. We align the H&E WSI with the spatial layout of the Visium slide and generate imaging and quantitative morphology features for each Visium spot. The pipeline design enables multiple analysis workflows, including single or dual reference genome input and stand-alone image analysis. We show the utility of our pipeline on a dataset from Visium profiling of four melanoma PDX samples. The clustering of Visium spots and clustering of H&E imaging features reveal similar patterns arising from the two data modalities.

A platform for rapid patient-derived cutaneous neurofibroma organoid establishment and screening.

Localized cutaneous neurofibromas (cNFs) are benign tumors that arise in the dermis of patients affected by neurofibromatosis type 1 syndrome. cNFs are benign lesions: they do not undergo malignant transformation or metastasize. Nevertheless, they can cover a significant proportion of the body, with some individuals developing hundreds to thousands of lesions. cNFs can cause pain, itching, and disfigurement resulting in substantial socio-emotional repercussions. Currently, surgery and laser desiccation are the sole treatment options but may result in scarring and potential regrowth from incomplete removal. To identify effective systemic therapies, we introduce an approach to establish and screen cNF organoids. We optimized conditions to support the ex vivo growth of genomically diverse cNFs. Patient-derived cNF organoids closely recapitulate cellular and molecular features of parental tumors as measured by immunohistopathology, methylation, RNA sequencing, and flow cytometry. Our cNF organoid platform enables rapid screening of hundreds of compounds in a patient- and tumor-specific manner.

Subtype-WGME enables whole-genome-wide multi-omics cancer subtyping.

We present an innovative strategy for integrating whole-genome-wide multi-omics data, which facilitates adaptive amalgamation by leveraging hidden layer features derived from high-dimensional omics data through a multi-task encoder. Empirical evaluations on eight benchmark cancer datasets substantiated that our proposed framework outstripped the comparative algorithms in cancer subtyping, delivering superior subtyping outcomes. Building upon these subtyping results, we establish a robust pipeline for identifying whole-genome-wide biomarkers, unearthing 195 significant biomarkers. Furthermore, we conduct an exhaustive analysis to assess the importance of each omic and non-coding region features at the whole-genome-wide level during cancer subtyping. Our investigation shows that both omics and non-coding region features substantially impact cancer development and survival prognosis. This study emphasizes the potential and practical implications of integrating genome-wide data in cancer research, demonstrating the potency of comprehensive genomic characterization. Additionally, our findings offer insightful perspectives for multi-omics analysis employing deep learning methodologies.

Tracing unknown tumor origins with a biological-pathway-based transformer model.

Cancer of unknown primary (CUP) represents metastatic cancer where the primary site remains unidentified despite standard diagnostic procedures. To determine the tumor origin in such cases, we developed BPformer, a deep learning method integrating the transformer model with prior knowledge of biological pathways. Trained on transcriptomes from 10,410 primary tumors across 32 cancer types, BPformer achieved remarkable accuracy rates of 94%, 92%, and 89% in primary tumors and primary and metastatic sites of metastatic tumors, respectively, surpassing existing methods. Additionally, BPformer was validated in a retrospective study, demonstrating consistency with tumor sites diagnosed through immunohistochemistry and histopathology. Furthermore, BPformer was able to rank pathways based on their contribution to tumor origin identification, which helped to classify oncogenic signaling pathways into those that are highly conservative among different cancers versus those that are highly variable depending on their origins.

Cross-attention enables deep learning on limited omics-imaging-clinical data of 130 lung cancer patients.

Deep-learning tools that extract prognostic factors derived from multi-omics data have recently contributed to individualized predictions of survival outcomes. However, the limited size of integrated omics-imaging-clinical datasets poses challenges. Here, we propose two biologically interpretable and robust deep-learning architectures for survival prediction of non-small cell lung cancer (NSCLC) patients, learning simultaneously from computed tomography (CT) scan images, gene expression data, and clinical information. The proposed models integrate patient-specific clinical, transcriptomic, and imaging data and incorporate Kyoto Encyclopedia of Genes and Genomes (KEGG) and Reactome pathway information, adding biological knowledge within the learning process to extract prognostic gene biomarkers and molecular pathways. While both models accurately stratify patients in high- and low-risk groups when trained on a dataset of only 130 patients, introducing a cross-attention mechanism in a sparse autoencoder significantly improves the performance, highlighting tumor regions and NSCLC-related genes as potential biomarkers and thus offering a significant methodological advancement when learning from small imaging-omics-clinical samples.

An autologous ex vivo model for exploring patient-specific responses to viro-immunotherapy in glioblastoma.

Oncolytic virus (OV) clinical trials have demonstrated remarkable efficacy in subsets of patients with glioblastoma (GBM). However, the lack of tools to predict this response hinders the advancement of a more personalized application of OV therapy. In this study, we characterize an ex vivo co-culture system designed to examine the immune response to OV infection of patient-derived GBM neurospheres in the presence of autologous peripheral blood mononuclear cells (PBMCs). Co-culture conditions were optimized to retain viability and functionality of both tumor cells and PBMCs, effectively recapitulating the well-recognized immunosuppressive effects of GBM. Following OV infection, we observed elevated secretion of pro-inflammatory cytokines and chemokines, including interferon γ, tumor necrosis factor α, CXCL9, and CXCL10, and marked changes in immune cell activation markers. Importantly, OV treatment induced unique patient-specific immune responses. In summary, our co-culture platform presents an avenue for personalized screening of viro-immunotherapies in GBM, offering promise as a potential tool for future patient stratification in OV therapy.

A 3D lung lesion variational autoencoder.

In this study, we develop a 3D beta variational autoencoder (beta-VAE) to advance lung cancer imaging analysis, countering the constraints of conventional radiomics methods. The autoencoder extracts information from public lung computed tomography (CT) datasets without additional labels. It reconstructs 3D lung nodule images with high quality (structural similarity: 0.774, peak signal-to-noise ratio: 26.1, and mean-squared error: 0.0008). The model effectively encodes lesion sizes in its latent embeddings, with a significant correlation with lesion size found after applying uniform manifold approximation and projection (UMAP) for dimensionality reduction. Additionally, the beta-VAE can synthesize new lesions of varying sizes by manipulating the latent features. The model can predict multiple clinical endpoints, including pathological N stage or KRAS mutation status, on the Stanford radiogenomics lung cancer dataset. Comparisons with other methods show that the beta-VAE performs equally well in these tasks, suggesting its potential as a pretrained model for predicting patient outcomes in medical imaging.

An improved approach to generate IL-15+/+/TGFßR2-/- iPSC-derived natural killer cells using TALEN.

We present a TALEN-based workflow to generate and maintain dual-edited (IL-15+/+/TGFßR2-/-) iPSCs that produce enhanced iPSC-derived natural killer (iNK) cells for cancer immunotherapy. It involves using a cell lineage promoter for knocking in (KI) gene(s) to minimize the potential effects of expression of any exogenous genes on iPSCs. As a proof-of-principle, we KI IL-15 under the endogenous B2M promoter and show that it results in high expression of the sIL-15 in iNK cells but minimal expression in iPSCs. Furthermore, given that it is known that knockout (KO) of TGFßR2 in immune cells can enhance resistance to the suppressive TGF-ß signaling in the tumor microenvironment, we develop a customized medium containing Nodal that can maintain the pluripotency of iPSCs with TGFßR2 KO, enabling banking of these iPSC clones. Ultimately, we show that the dual-edited IL-15+/+/TGFßR2-/- iPSCs can be efficiently differentiated into NK cells that show enhanced autonomous growth and are resistant to the suppressive TGF-ß signaling.

An engineering strategy to target activated EGFR with CAR T cells.

Chimeric antigen receptor (CAR) T cells have shown remarkable response rates in hematological malignancies. In contrast, CAR T cell treatment of solid tumors is associated with several challenges, in particular the expression of most tumor-associated antigens at lower levels in vital organs, resulting in on-target/off-tumor toxicities. Thus, innovative approaches to improve the tumor specificity of CAR T cells are urgently needed. Based on the observation that many human solid tumors activate epidermal growth factor receptor (EGFR) on their surface through secretion of EGFR ligands, we developed an engineering strategy for CAR-binding domains specifically directed against the ligand-activated conformation of EGFR. We show, in several experimental systems, that the generated binding domains indeed enable CAR T cells to distinguish between active and inactive EGFR. We anticipate that this engineering concept will be an important step forward to improve the tumor specificity of CAR T cells directed against EGFR-positive solid cancers.

Single-cell signatures identify microenvironment factors in tumors associated with patient outcomes.

The cellular components of tumors and their microenvironment play pivotal roles in tumor progression, patient survival, and the response to cancer treatments. Unveiling a comprehensive cellular profile within bulk tumors via single-cell RNA sequencing (scRNA-seq) data is crucial, as it unveils intrinsic tumor cellular traits that elude identification through conventional cancer subtyping methods. Our contribution, scBeacon, is a tool that derives cell-type signatures by integrating and clustering multiple scRNA-seq datasets to extract signatures for deconvolving unrelated tumor datasets on bulk samples. Through the employment of scBeacon on the The Cancer Genome Atlas (TCGA) cohort, we find cellular and molecular attributes within specific tumor categories, many with patient outcome relevance. We developed a tumor cell-type map to visually depict the relationships among TCGA samples based on the cell-type inferences.

Identification of activity-based biomarkers for early-stage pancreatic tumors in blood using single-molecule enzyme activity screening.

Single-molecule enzyme activity-based enzyme profiling (SEAP) is a methodology to globally analyze protein functions in living samples at the single-molecule level. It has been previously applied to detect functional alterations in phosphatases and glycosidases. Here, we expand the potential for activity-based biomarker discovery by developing a semi-automated synthesis platform for fluorogenic probes that can detect various peptidases and protease activities at the single-molecule level. The peptidase/protease probes were prepared on the basis of a 7-amino-4-methylcoumarin fluorophore. The introduction of a phosphonic acid to the core scaffold made the probe suitable for use in a microdevice-based assay, while phosphonic acid served as the handle for the affinity separation of the probe using Phos-tag. Using this semi-automated scheme, 48 fluorogenic probes for the single-molecule peptidase/protease activity analysis were prepared. Activity-based screening using blood samples revealed altered single-molecule activity profiles of CD13 and DPP4 in blood samples of patients with early-stage pancreatic tumors. The study shows the power of single-molecule enzyme activity screening to discover biomarkers on the basis of the functional alterations of proteins.