Conserved Landscape of Chemokine Receptor Co-expression Defines the Functional States of CD8+ T Cells in Melanoma
Macedo, R.; Harle, D. W.; Hoffer-Hawlik, K.; Wang, X. K.; McMahon-Skates, T.; Matschiner, A.; Belay, K.; Saenger, Y. M.; Izar, B.; Azizi, E.; Reshef, R.
Show abstract
Cancer immunotherapies, from checkpoint blockade to adoptive cell therapies like tumor-infiltrating lymphocytes (TILs), have revolutionized cancer treatment but are limited by variable efficacy and significant toxicities. A central challenge is identifying ideal T-cell populations that effectively eliminate tumors without causing off-target damage, a distinction not captured by existing biomarkers. We show that co-expression patterns of chemokine receptors (CRs) CXCR3, CCR5, and CXCR6 on CD8+ T cells provide a functional "code" defining subsets with divergent roles in on-target immunity versus off-target inflammation. In mouse and human melanoma, a triple-positive (CXCR3+CCR5+CXCR6+) T-cell subset is essential for tumor control, and its genetic signature correlates with positive clinical response, while a distinct CCR5+CXCR6+ subset drives liver immune-related adverse events (IRAEs). Crucially, this CR code reveals that immunotherapy actively reshapes T-cell trafficking patterns, uncovering profound heterogeneity within conventional populations and distinguishing potent anti-tumor progenitors from cells predisposed to exhaustion or off-target migration. This work establishes CR co-expression as a practical tool, providing a surface marker-based strategy to identify and enrich optimized T cells for adoptive therapies, thereby offering a framework to uncouple efficacy from toxicity. One Sentence SummaryCo-expression of CCR5, CXCR6, and CXCR3 provides a functional code that separates T-cell-mediated anti-tumor efficacy from off-target toxicity, enabling the selection of superior cells for safer and more effective cancer immunotherapies. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=120 SRC="FIGDIR/small/693486v2_ufig1.gif" ALT="Figure 1"> View larger version (29K): org.highwire.dtl.DTLVardef@19b0f44org.highwire.dtl.DTLVardef@1076129org.highwire.dtl.DTLVardef@17c0be6org.highwire.dtl.DTLVardef@f14893_HPS_FORMAT_FIGEXP M_FIG C_FIG HighlightsO_LI{middle dot} CXCR6, CXCR3 and CCR5 co-expression signature stratifies patient survival in human melanoma C_LIO_LI{middle dot} CD8+ T cells co-expressing CXCR6, CXCR3 and CCR5 are critical effectors with high proliferative, cytotoxic, and activation profile in human and mice melanoma C_LIO_LI{middle dot} CD8+ T cells co-expressing CXCR6, CXCR3 and CCR5 drive anti-tumoral responses during checkpoint blockade in both human and mice C_LI
Matching journals
The top 4 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- LAG-3 blockade reactivates the CD8+ T cell expansion program to re-expand contracted clones in the tumor 98%
- Immunity in Omicron SARS-CoV-2 breakthrough COVID-19 in vaccinated adults 97%
- Agonistic CD40 antibody therapy induces tertiary lymphoid structures but impairs the response to immune checkpoint blockade in glioma 97%
Similar papers in this journal
- Soluble CTLA-4 mainly produced by Treg cells inhibits type 1 inflammation without hindering type 2 immunity to allow for inflammation resolution 96%
- Opposing effects of pre-existing antibody and memory T cell help on the dynamics of recall germinal centers 96%
- Permanent lymphocyte subset elimination upon a single dose of AAV-delivered depletion antibody dissects immune control of chronic viral infection 94%
Similar papers in this journal
- Distinct Subsets of Multi-Lymphoid Progenitors Support Ontogeny-Related Changes in Human Lymphopoiesis 97%
- Extremely potent monoclonal antibodies neutralize Omicron and other SARS-CoV-2 variants 97%
- HIV-1 Vpr drives a tissue residency-like phenotype during selective infection of resting memory T cells 96%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.