Thursday, November 19, 2026
Science
No Result
View All Result
  • Login
  • HOME
  • SCIENCE NEWS
  • CONTACT US
  • HOME
  • SCIENCE NEWS
  • CONTACT US
No Result
View All Result
Scienmag
No Result
View All Result
Home Science News

Cross-Species Analysis Uncovers Conserved Cellular Programs Across Diverse Solid Tumours

November 19, 2026
in Science News
Nathaniel Bowman
By Nathaniel Bowman Scienmag Editorial Profile - Precision Oncology
Reading Time: 6 mins read
0
Cross-Species Analysis Uncovers Conserved Cellular Programs Across Diverse Solid Tumours

Cross-Species Analysis Uncovers Conserved Cellular Programs Across Diverse Solid Tumours

65
SHARES
587
VIEWS
Share on FacebookShare on Twitter
ADVERTISEMENT

A comprehensive meta-analysis of single-cell transcriptomic datasets spanning more than 15 million cells has demonstrated that host cell transcriptional programming in solid tumours follows highly conserved patterns across species and anatomical sites. Published in Nature, the investigation systematically examined stromal and immune compartments across 42 solid cancer types in more than 2,000 human and mouse tumour samples. By algorithmically screening thousands of publicly available single-cell RNA-sequencing studies, the research team identified reproducible cellular architectures that bridge tumour biology across disparate tissues, providing unified reference classifications and a practical foundation for systematic cross-cancer investigation.

Investigations into host tissue responses to solid malignancies have historically evaluated individual cancer types in isolation, limiting comparative understanding of stromal and immune evolution across diverse pathologies. While malignant cells display substantial genetic heterogeneity across organs, non-malignant host components frequently exhibit convergent phenotypic responses to oncogenic stress. To determine the extent of these shared states, researchers curated droplet-based single-cell RNA-sequencing datasets deposited between January 2014 and July 2025 across two primary public repositories: the Gene Expression Omnibus and ArrayExpress. Solid-organ tumours outside the central nervous system and non-haematological cancers were systematically assembled, requiring samples with intact stromal representation to prevent cellular dropouts during pan-cancer integration.

The study protocol implemented strict inclusion criteria, manually screening 2,168 human and 1,807 mouse datasets. Human collections were required to contain at least three primary tumour or non-malignant tissue samples profiled via droplet platforms. Mouse datasets were limited to untreated animals carrying genetically engineered or orthotopic cancer cell line models. Following rigorous quality filtering to eliminate low-quality cells, transcript duplicates, and computational doublets, the team compiled a human atlas comprising 2,855 samples from 190 datasets and a mouse atlas of 526 samples from 68 datasets. In total, 210 human and 79 mouse datasets satisfied these baseline requirements before undergoing downstream multi-step computational reconciliation.

To mitigate the technical challenges of handling tens of millions of cells, the investigators implemented bit-packing compression through the BPCells package to minimize disk read and write demands during matrix processing. Dataset integration utilized sketch-based leverage-score sampling, taking representative subsets of 1,000 cells per dataset to compute variable features, scale expressions, and perform principal component analysis across 30 dimensions. To prevent classification bias stemming from any single mathematical model, the initial cell classifications deployed three independent integration algorithms in parallel: Harmony, FastMNN, and scVI. Cells identified as candidate lineages were subsequently pooled across methods, permitting overlapping initial assignments before undergoing iterative rounds of supervised cleaning and re-clustering.

Supervised cell annotation classified the broad cellular ecosystem into seven lineages: epithelial, fibroblast, endothelial, mural, nerve, lymphoid, and myeloid cells. Marker gene identification utilized canonical transcripts including EPCAM for epithelial cells, PDGFRA for fibroblasts, PECAM1 for endothelial cells, RGS5 for mural cells, CD19 and CD3E for lymphoid populations, CD14 for myeloid cells, and PLP1 for nerve cells. Because epithelial cells in solid carcinomas pose major computational challenges when separating malignant from non-malignant lineages, detailed downstream subtype analyses were directed toward the remaining six compartments. Across multiple iterative rounds of batch correction and sub-clustering using Harmony, investigators generated purified atlases culminating in 45 defined subpopulations in humans and 44 in mice.

Within the fibroblast compartment, the analysis reconciled fragmented historical nomenclatures into four overarching superclusters: antigen-presenting cancer-associated fibroblasts (apCAFs), inflammatory CAFs (iCAFs), myofibroblastic CAFs (mCAFs), and steady-state CAFs (ssCAFs). The scale of the dataset allowed classical LRRC15-positive myofibroblasts and mechanosensitive COL4A1/SPP1-positive states to be integrated into a unified mCAF compartment reflecting shared extracellular matrix remodelling programs. Furthermore, the analysis demonstrated that traditional CXCL12-positive iCAF states transcriptionally align with tissue-resident ssCAF populations expressing CXCL14 and CXCL12, separating them from true inflammatory phenotypes that respond to microenvironmental cytokine cues.

Refined resolution within the inflammatory fibroblast compartment revealed a specialized subpopulation characterized by high ISG15 expression. These ISG15-positive iCAFs proved strongly enriched across primary solid tumours compared to non-malignant tissue and increased across advanced neoplastic stages. Downstream signalling analyses utilizing CellChat and NicheNet indicated that these cells display transcriptional profiles consistent with receiving interferon-gamma signalling from cytotoxic CD8-positive granzyme-B-positive T lymphocytes. Correlative analyses incorporating bulk RNA-sequencing data from 9,880 primary tumour samples in The Cancer Genome Atlas across 31 solid tumour types revealed that high proportions of ISG15-positive iCAFs were linked to significantly decreased patient overall survival.

Cross-species integration established that these stromal and immune phenotypes are deeply conserved between human and murine solid malignancies. Transforming mouse genes into human orthologues using NicheNet enabled combined dimensionality reduction and anchor-based label transfer between the two species. Conserved cellular identities spanned lymphoid, myeloid, endothelial, and mural populations, with coincidence matrices demonstrating robust alignments between equivalent human and mouse clusters. This conservation indicates that preclinical murine models reliably recapitulate core human microenvironmental programming, while simultaneously providing a mechanistic baseline to guide translational experimental interventions across solid tumours.

To validate the computational cell–cell interaction predictions experimentally, the researchers cultured primary human pancreatic ductal adenocarcinoma cancer-associated fibroblasts derived from a surgical resection. Cells were stimulated for 24 hours with recombinant human interferon-gamma, lipopolysaccharide, or vehicle control, followed by bulk RNA-sequencing across dedicated sequencing lanes. Transcriptional profiling confirmed that primary fibroblasts exposed to interferon-gamma specifically induced the signature marker panel characterizing ISG15-positive iCAFs, reproducing the computational model predictions. Spatial transcriptomic and multiplex immunofluorescence analyses on histological specimens further localized these specialized stromal phenotypes in situ alongside interacting immune populations.

To facilitate broad scientific use, the authors codified these classifications into scTME, an open-source R package utilizing SingleR classification frameworks to annotate candidate single-cell datasets against the 45 curated pan-cancer subpopulations. Alongside the software, an interactive web atlas (tmeatlas.stanford.edu) provides public access to the underlying cell embeddings, raw matrices, and clinical association data. By establishing a unified structural baseline for solid tumour microenvironments, the study bridges long-standing divides between isolated organ-specific investigations, offering standardized reference tools for dissecting host-tumour ecosystems across human oncology and translational models.

Beyond broad cellular cataloguing, the computational framework incorporated rigorous statistical standards to measure subpopulation enrichment across clinical subsets. Subpopulation frequencies were assessed across three clinical axes: non-malignant tissue versus primary tumours, early-stage (stage I and II) versus late-stage (stage III and IV) disease, and treatment-responsive versus non-responsive tumours. To prevent mathematical instability from sparsely sampled biopsies, investigators excluded any sample contributing 30 or fewer cells to a parent lineage. Proportional comparisons deployed a two-sided weighted Welch t-test with sample weights scaled to single-cell capture counts, deriving error boundaries via Kish effective sample size calculations and controlling false discovery rates through lineage-stratified Benjamini–Hochberg adjustments. For bulk clinical deconvolution, researchers downsampled the single-cell atlas to 1,000 cells per subpopulation to inform CIBERSORTx running in S-mode across 9,880 primary tumour and 740 tumour-adjacent profiles from The Cancer Genome Atlas. Hazard ratios evaluating patient mortality across 31 solid cancer types were estimated with Cox proportional hazards regression in survcomp, synthesizing a pan-cancer metric using inverse-variance weighting.

Intercellular crosstalk analyses broadened the scope of microenvironmental dynamics beyond interferon-gamma signalling. Evaluating 100,000 downsampled cells through CellChat and NicheNet regulatory potential modeling highlighted multiple coordinated axes, including macrophage-monocyte and mural cell transmission of vascular endothelial growth factor signals to ESM1-positive endothelial tip cells, as well as macrophage and dendritic cell communication through migration inhibitory factor to ISG15-positive fibroblasts. Spatial transcriptomic and proteomic validations situated these cell types within intact tissue microenvironments. Investigators coupled the NiCo pipeline with public 10X Genomics Xenium assays and performed multiplexed CODEX immunofluorescence on formalin-fixed pancreatic ductal adenocarcinoma sections from the Stanford Department of Pathology. These spatial surveys aligned with automated fiber architecture algorithms trained on 354,838 hematoxylin and eosin image tiles from 85 pancreatic specimens, directly confirming that spindle-shaped HLA-DR-positive fibroblasts localized within dense extracellular matrix regions associated with elevated histological pseudotime. Complementary immunofluorescence assays confirmed comparable PDGFR-alpha and HLA-DR co-localization across independent pancreatic, colorectal, and papillary thyroid carcinoma specimens.

Controlled in vitro culture experiments provided granular transcriptional measurements of stromal plasticity. Investigators isolated primary fibroblasts from a surgical resection of a single male patient with pancreatic ductal adenocarcinoma under institutional review board approval. Following four passages, cells were plated across nine independent wells and exposed for 24 hours to vehicle control, recombinant human interferon-gamma at 20 nanograms per milliliter, or lipopolysaccharide at 100 nanograms per milliliter. Extracted RNA was prepared with rRNA depletion chemistry and sequenced on a Singular Genomics G4 instrument across nine dedicated lanes, yielding paired-end reads of 75 base pairs without multiplexing index cycles. Reads were mapped to the human GRCh37-hg19 reference genome using TopHat, establishing 26,363 filtered genes that were evaluated through DESeq2 Wald testing to quantify signature enrichment scores.

The authors also highlighted distinct experimental and computational boundaries inherent to large-scale data aggregation. Because historical public single-cell repositories rarely deposit unfiltered raw count matrices, the authors could not run droplet de-noising packages such as CellBender or SoupX to subtract ambient RNA. In vitro validation relied on a primary fibroblast line derived from one male donor without mycoplasma screening or post-isolation marker re-verification. Furthermore, while the cross-species integration harmonized mouse and human cellular states, no new animal experiments were performed; all murine single-cell data stemmed entirely from previously published datasets.

Subject of Research: Pan-cancer cross-species meta-analysis of the solid tumour microenvironment

Article Title: Conservation of the tumour microenvironment across organs and species

Article References: Januszyk, M., Lu, J. M., Korah, M., Jing, S. L., Xie, P. Y., Fallah, M., Pham, B., Guo, J. L., Berry, C. E., Wang, H. H., Talbott, H. E., Nee, K., Le, T., Garcia, C. A., Gonçalves, A., Nosrati, F., Bauer-Rowe, K. E., Guo, C., Agolia, J. P., … Longaker, M. T. (2026). Conservation of the tumour microenvironment across organs and species. Nature. https://doi.org/10.1038/s41586-026-11270-4

Image Credits: AI Generated

DOI: 10.1038/s41586-026-11270-4

Keywords: tumour microenvironment, single-cell RNA-seq, cancer stroma, cancer immunology, cancer-associated fibroblasts, cross-species genomics, pan-cancer meta-analysis, transcriptomics, Harmony integration, scTME, ISG15 iCAFs, endothelial cells, myeloid cells, cellular heterogeneity, computational biology, intercellular communication

Cite Scienmag News

Nathaniel Bowman. (November 19, 2026). Cross-Species Analysis Uncovers Conserved Cellular Programs Across Diverse Solid Tumours. Scienmag. https://scienmag.com/cross-species-analysis-uncovers-conserved-cellular-programs-across-diverse-solid-tumours/

Nathaniel Bowman. "Cross-Species Analysis Uncovers Conserved Cellular Programs Across Diverse Solid Tumours." Scienmag, 19 November 2026, https://scienmag.com/cross-species-analysis-uncovers-conserved-cellular-programs-across-diverse-solid-tumours/. Accessed 19 November 2026.

Nathaniel Bowman. "Cross-Species Analysis Uncovers Conserved Cellular Programs Across Diverse Solid Tumours." Scienmag. November 19, 2026. https://scienmag.com/cross-species-analysis-uncovers-conserved-cellular-programs-across-diverse-solid-tumours/

Tags: Cancer immunologycancer stromacancer-associated fibroblastscellular heterogeneitycomputational biologycross-species genomicsendothelial cellsHarmony integrationintercellular communicationISG15 iCAFsmyeloid cellspan-cancer meta-analysisscTMEsingle-cell RNA-seqTranscriptomicstumour microenvironment
Share26Tweet16
Previous Post

South Asian Avian Influenza Research Surged as Clade 2.3.4.4b Spread Across Borders

Next Post

G20 Headache Study Reveals Population Growth and Saturation Effects Shaping Future Burdens

Related Posts

Qualitative Study Reveals How COVID-19 Disrupted Key Healthcare Services in Malawi
Medicine

Qualitative Study Reveals How COVID-19 Disrupted Key Healthcare Services in Malawi

November 19, 2026
Rule-Based Transcriptomic Modeling Identifies Early Markers of Primate SIV Vaccine Protection
Science News

Rule-Based Transcriptomic Modeling Identifies Early Markers of Primate SIV Vaccine Protection

November 19, 2026
Lysine-Derived Carbon Catalyst Converts Carbon Dioxide Into Cyclic Carbonates Without Metals
Technology and Engineering

Lysine-Derived Carbon Catalyst Converts Carbon Dioxide Into Cyclic Carbonates Without Metals

November 19, 2026
Salvage Macrolide Treats Resistant Mycoplasma Pneumonia and Plastic Bronchitis in Child
Medicine

Salvage Macrolide Treats Resistant Mycoplasma Pneumonia and Plastic Bronchitis in Child

November 19, 2026
Cattail Roots Release Oxygen into Wetland Sediments During Active Growth Pulses
Chemistry

Cattail Roots Release Oxygen into Wetland Sediments During Active Growth Pulses

November 19, 2026
Free Farrowing Systems Improve Sow Welfare but Require Rigorous Parturition Monitoring
Medicine

Free Farrowing Systems Improve Sow Welfare but Require Rigorous Parturition Monitoring

November 19, 2026
Next Post
G20 Headache Study Reveals Population Growth and Saturation Effects Shaping Future Burdens

G20 Headache Study Reveals Population Growth and Saturation Effects Shaping Future Burdens

  • Mothers who receive childcare support from maternal grandparents show more optimized

    Mothers who receive childcare support from maternal grandparents show more parental warmth, finds NTU Singapore study

    27656 shares
    Share 11059 Tweet 6912
  • University of Seville Breaks 120-Year-Old Mystery, Revises a Key Einstein Concept

    1061 shares
    Share 424 Tweet 265
  • Bee body mass, pathogens and local climate influence heat tolerance

    682 shares
    Share 273 Tweet 171
  • Researchers record first-ever images and data of a shark experiencing a boat strike

    546 shares
    Share 218 Tweet 137
  • Groundbreaking Clinical Trial Reveals Lubiprostone Enhances Kidney Function

    531 shares
    Share 212 Tweet 133
Science

Embark on a thrilling journey of discovery with Scienmag.com—your ultimate source for cutting-edge breakthroughs. Immerse yourself in a world where curiosity knows no limits and tomorrow’s possibilities become today’s reality!

RECENT NEWS

  • Qualitative Study Reveals How COVID-19 Disrupted Key Healthcare Services in Malawi
  • Rule-Based Transcriptomic Modeling Identifies Early Markers of Primate SIV Vaccine Protection
  • Lysine-Derived Carbon Catalyst Converts Carbon Dioxide Into Cyclic Carbonates Without Metals
  • Salvage Macrolide Treats Resistant Mycoplasma Pneumonia and Plastic Bronchitis in Child

Categories

  • Agriculture
  • Anthropology
  • Archaeology
  • Athmospheric
  • Biology
  • Biotechnology
  • Blog
  • Bussines
  • Cancer
  • Chemistry
  • Climate
  • Earth Science
  • Editorial Policy
  • Marine
  • Mathematics
  • Medicine
  • Pediatry
  • Policy
  • Psychology & Psychiatry
  • Science Education
  • Science News
  • Social Science
  • Space
  • Technology and Engineering

Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 5,151 other subscribers

© 2026 Scienmag - Science Magazine

Welcome Back!

Login to your account below

Forgotten Password?

Retrieve your password

Please enter your username or email address to reset your password.

Log In
No Result
View All Result
  • HOME
  • SCIENCE NEWS
  • CONTACT US

© 2026 Scienmag - Science Magazine

Discover more from Science

Subscribe now to keep reading and get access to the full archive.

Continue reading