A core inflammatory gene network associated with poor prognosis serves chemokine production in cancer associated fibroblasts in pancreatic ductal adenocarcinoma

Fangfei Li, Liu Yang, Zheng Chen, Shuangying Qiao, Yalan Sheng, Debajyoti Chowdhury, Hiu Fung Yip, Meiheng Sun, Aiping Lu

Research output: Contribution to journalConference articlepeer-review


Introduction: Inflammation plays an important role on the tumor microenvironment (TME) of pancreatic ductal adenocarcinoma (PDAC). Nevertheless, due to the variable inflammatory characteristics and TME profiles among PDAC patients, it is still unclear which inflammatory factors are crucially associated with PDAC prognosis and how the TME is influenced. Previously, we found a core inflammatory gene network (CIGN) by analyzing bulk RNA seq data of 183 PDAC patients from the Cancer Genome Altas (TCGA) based on 104 inflammatory gene sets from the Molecular Signatures Database. The CIGN is defined by two markers (DCBLD2 and PLAU) and is associated with poor prognosis. Single-cell RNA (scRNA) seq data provides valuable information on multiple types of cells, and it is necessary to utilise scRNA data to analyze the impact of our CIGN on PDAC TME.

Method: To investigate the tumor microenvironment (TME) associated with CIGN, we employed both bulk and scRNA seq data of tumor tissue from PDAC patients from TCGA and Gene Expression Omnibus. Firstly, the Enrichplot package and online Metascape were used to perform functional enrichment analysis of genes increased in association with CIGN from bulk RNA seq data. Secondly, CIGN identified prognostic criteria from bulk RNA seq was applied to scRNA data. Then, the Seurat package was used to analyse three scRNA seq series: GSE212966 (6 PDAC patients), GSE155698 (15 PDAC patients), and GSE214295 (3 PDAC patients). Finally, the effects of CIGN on cancer-associated fibroblasts (CAFs) (proliferation, migration and chemokine expression profile)were investigated ex vivo in mouse CAFs isolated from KPC mice.

Results: Genes associated with CIGN were enriched in processes related to the extracellular matrix, endoderm formation, collagen binding, response to wounding and receptor-ligand activity. Then, we performed CIGN grouping on three scRNA seq series and found a higher accumulation of CAFs in CIGN group compared to non-CIGN group, while pancreatic progenitor cell infiltration in CIGN group was much lower, implying a more immune suppressive, desmoplastic, and hypoxic TME in CIGN patients. Furthermore, we found two marker genes of CIGN (DCBLD2 and PLAU) in both expressed mostly in fibroblast cells. Moreover, in CAFs, DCBLD2 and PLAU expression is significantly higher in the CIGN group than in the non-CIGN group. Finally, the proliferation and migration rates of si-DCBLD2 and si-PLAU groups were significantly decreased compared with the control group ex vivo. What is more, the secretion of inflammatory chemokine (CXCL7 and CXCL12) increased in supernatants of CAFs.

Conclusion: A core inflammatory gene network was found specifically functions in CAF and induced chemokine production in pancreatic ductal adenocarcinoma.
Original languageEnglish
Article numberB043
JournalCancer Immunology Research
Issue number12, Supplement
Publication statusPublished - 1 Dec 2023
EventAACR Special Conference in Cancer Research: Tumor Immunology and Immunotherapy - Toronto, Canada
Duration: 1 Oct 20234 Oct 2023


Dive into the research topics of 'A core inflammatory gene network associated with poor prognosis serves chemokine production in cancer associated fibroblasts in pancreatic ductal adenocarcinoma'. Together they form a unique fingerprint.

Cite this