Across many solid tumor types, cancer-associated fibroblasts (CAFs) are abundant and highly heterogeneous, with distinct subpopulations exerting immunomodulatory functions. Though they have been well characterized in malignancies considered to be highly desmoplastic, how distinct CAF populations in lung adenocarcinoma (LUAD) influence the tumor–immune microenvironment is not well understood. Here, we identify a novel population of immunomodulatory CAFs (imCAFs) in the LUAD tumor microenvironment, characterized by CHL1 expression and enriched in transcriptional programs involving immune regulation and chemokine signaling. Through single-cell and spatial transcriptomics, we demonstrate that imCAFs are spatially colocalized with CXCR3+ regulatory T (Treg) cells, a hyper-suppressive T cell subset that accumulates at the tumor border. Mechanistically, imCAFs produce CXCL9, a key chemoattractant that mediates the recruitment of CXCR3+ Tregs to the tumor, promoting an immunosuppressive microenvironment. Functional profiling reveals that tumor-resident CXCR3+ Tregs possess enhanced proliferative and suppressive capacity and are transcriptionally distinct from CXCR3- counterparts. Genetic ablation of Cxcr3 in Treg cells or Cxcl9 in stromal cells significantly reduces Treg cell accumulation, resulting in enhanced CD8+ T cell activation and decreased tumor burden. Analogous CHL1+ imCAF-like fibroblasts are also observed in human NSCLC samples and were found to colocalize with Treg cells, with increased CHL1 expression in lung tumors being associated with reduced cytotoxicity scores and decreased progression free survival. These findings uncover a critical stromal–immune axis wherein imCAF-derived CXCL9 orchestrates the localization and function of hyper-suppressive CXCR3+ Tregs in LUAD and highlight the imCAF–CXCL9–CXCR3+ Treg cell recruitment circuit as a promising therapeutic target to restore anti-tumor immunity.
from sddb import Catalog
cohort = Catalog().collection("geo-gse303162")
cohort.members # the 2 datasets as a table
print(cohort.licenses()) # check the terms before you download
sdatas = cohort.to_dict() # {uid: SpatialData}, cached locally
# or keep the .zarr files in a folder of your choice:
# sdatas = cohort.to_dict(dest="data/")Needs pip install lamindb and network access to the Lamin database, which is limited during the alpha. If it fails, use the Client tab.
import lamindb as ln
db = ln.DB("scverse/spatialdata-db")
coll = db.Collection.get(key="geo-gse303162")
coll.ordered_artifacts.to_dataframe() # the 2 members, ordered
# download every member .zarr to the local cache:
paths = [a.cache() for a in coll.ordered_artifacts.all()]# sddb client: download every member .zarr into data/
sddb-cli download --collection geo-gse303162 --dest data/Licenses: 2 NCBI GEO policy
@article{Ringham_2026,
title = {A novel CAF population coordinates hyper-suppressive regulatory T cell recruitment and localization in lung cancer},
volume = {27},
ISSN = {1529-2916},
url = {http://dx.doi.org/10.1038/s41590-026-02607-2},
DOI = {10.1038/s41590-026-02607-2},
number = {9},
journal = {Nature Immunology},
publisher = {Springer Science and Business Media LLC},
author = {Ringham, Olivia R. and Rivera, Monica and Loffredo, Lucas F. and Ozsoy, Melih Arda and Healy, Christina M. and Cheng, Maye F. and Jin, Yinuo and Chen, Noah and de los Santos-Alexis, Kenia and Azizi, Elham and Saqi, Anjali and Buechler, Matthew B. and Concepcion-Crisol, Carla P. and Arpaia, Nicholas},
year = {2026},
month = Aug,
pages = {1899–1912}
}