PuSH - Publikationsserver des Helmholtz Zentrums München

Hains, A.E.* ; Marques, J.G.* ; Chetal, K.* ; Nakatani, T. ; Ettinger, A. ; Biagi, C.A.O.d.* ; Gonzalez‐Sandoval, A.* ; Pillai, R.* ; Gatto, A.* ; Torres-Padilla, M.E. ; Sadreyev, R.I.* ; Filbin, M.G.* ; Rechem, C.V.*

Abstract SY23-01: Diffuse midline gliomas: From cell cycle to therapeutic opportunities.

Cancer Res. 85, SY23 - 01 (2025)
DOI
Lung squamous carcinoma (LUSC) is a lethal cancer with still unknown etiology and limited detection and therapeutic options. One hypothesis links its origin to the transformation of basal epithelium. Notably immunotherapy has changed cancer treatment. We integrated transcriptomics, proteomics, genomics, epigenetics, and single-cell RNA sequencing (scRNAseq) to explore basal cell-dependent signatures as therapy response predictors in LUSC. To identify whether subtypes of LUSC respond differently to immunotherapy, we first used the TCGA-LUSC cohort and performed unsupervised clustering using a set of hub (basal and immune) genes obtained through extensive analysis. We obtained 2 distinct clusters, the "Basalimmune-high" cluster had higher immune scores, immune cell infiltration and a better predictive immunotherapy response. The "Basalimmune-low" showed higher proliferative basal cell score and KRT5/6 and TP63 expression. Interestingly, there was an inverse correlation between KRT5/6 methylation and its RNA expression in "Basalimmune-high", with no difference in TP63 methylation between the groups, when analyzing epigenetic data. The "Basalimmune-high" cluster was associated with poor overall survival, when treated with standard chemotherapy. A prognostic signature of 15 genes was identified. These findings were validated using transcriptomic and proteomic data from other four LUSC cohorts. Using the top upregulated genes between the 2 clusters, we selected and measure their protein serum levels, to identify potential cluster predictors. Changes in serum levels of CXCL5 and CXCL11 in a real-life treatment naïve LUSC cohort (Stage 1-4,n=100) confirmed the clustering and poorer prognosis of the "Basalimmune-high"group (p<0.03). Using MutSig2CV, we found unique mutational landscapes in the two clusters: NOTCH1 in "Basalimmune-high" and ARID1A in "Basalimmune-low". Furthermore, copy number variants showed shared and distinct amplifications (Basalimmune-high: KAT6A; Basalimmune-low: EGFR) and deletions (Basalimmune-high: ROBO1; -low: STK11). GSEA of transcriptomic and proteomic data revealed some biological mechanisms underlying the immune responses of "Basalimmune-high" patients, such as upregulation of signaling pathways for complement, IL2, STAT5, IFN-α/γ, and other inflammatory responses. Given these associations, we performed an intercellular communication analysis on the scRNAseq data to identify novel communications between basal and immune cells that might influence immunotherapy. The pathways identified in silico were validated in vitro in co-cultures of CD8+ T cells, NK cells, and monocytes with primary human bronchial epithelial cells and identified as predictive of basal-immune cell communication. Altogether, this is the first demonstration that a multi-omic integrative approach has successfully identified distinct clusters on LUSC patients and may predict personalized immunotherapy options based on their genetic and molecular profiles.
Altmetric
Weitere Metriken?
[➜Einloggen]
Zusatzinfos bearbeiten [➜Einloggen]
Publikationstyp Artikel: Journalartikel
Dokumenttyp Meeting abstract
Korrespondenzautor
ISSN (print) / ISBN 0008-5472
e-ISSN 1538-7445
Zeitschrift Cancer Research
Quellenangaben Band: 85, Heft: 8_Supplement_2, Seiten: SY23 - 01 Artikelnummer: , Supplement: ,
Verlag American Association for Cancer Research (AACR)
Verlagsort Philadelphia, Pa.
Nichtpatentliteratur Publikationen
Begutachtungsstatus Peer reviewed
Institut(e) Institute of Epigenetics and Stem Cells (IES)