Heterogenous cancer-associated fibroblasts related tumor microenvironment marked by CD10/KLF4/TIAM1 were identified in pancreatic adenocarcinoma by integrated transcriptomics

Abstract

Pancreatic ductal adenocarcinoma (PDAC) is a highly aggressive malignancy characterized by a dense and heterogeneous tumor microenvironment (TME) composed of various cancer-associated fibroblasts (CAFs). In this study, we explored the composition and proportions of CAF subtypes within the PDAC TME and identified three distinct CAF-related TME subtypes: iCAF-rich, myCAF-rich, and PSC-rich. We observed significant heterogeneity in CAF populations across different patients, which correlated with patient prognosis and the mechanical and fibrotic properties of the TME. Our analysis revealed that these CAF subtypes exhibit distinct gene expression profiles, with the myCAF-rich subtype showing upregulation of hypoxia- and glycolysis-related genes, such as LDHA. Furthermore, gene set and survival analyses demonstrated that specific CAF subtypes harbor unique protective and risk factors, which were non-overlapping between the subtypes. These findings suggest that the heterogeneity of CAF subtypes plays a critical role in PDAC progression and therapeutic response. By utilizing multiplex immunohistochemistry and spatial transcriptomics, we also identified key CAF subpopulations, such as iCAF_17, iCAF_19, and myCAF_12, which were found to interact closely with tumor cells and macrophages. In chemotherapy-treated patients, myCAFs were positioned at the tumor boundary, potentially acting as a barrier to tumor invasion. This study provides novel insights into CAF-related TME subtypes, offering a foundation for future therapeutic strategies targeting CAFs in PDAC.

Keywords: ScRNA-seq; cancer-associated fibroblasts; pancreatic adenocarcinoma; spatial transcriptomics; tumor microenvironment.

Figure 1

Overall design of experiment and alignment of scRNA-seq and stRNA-seq. (A) Schematic showing workflow, and sampling methods of PDAC samples. (B) UMAP visualization of the cell types in PDAC TME. (C) Proportion of different cell types in PDAC samples. (D) UMAP visualization of CAFs and PSCs subset from integrated dataset. (E, F) Proportion of different subtypes of PSCs and CAFs in each patient, showed in bar (E) and clustered heatmap (F). (G) Four CAF-related TME subtypes were identified, and the proportions of CAFs and PSCs in each subtype were calculated.

Figure 2

MyCAF-rich TME is associated with typical niche and unfavorable clinical outcome. (A–C) UMAP showed cell density in each TME subtypes. The red arrows indicated representative cell types. (D) The Ro/e value (log scaled) of each cell type in different TME subtypes. (E) Geneset enrichment analysis suggested the myCAF-rich TME were shown regulated EMT, hypoxia, and glycolysis activity. (F) Forest plot showed the unfavorable clinical outcome of myCAF-rich TME signatures, while the other iCAF-rich and PSC-rich were not significant.

Figure 3

Different CAFs-related TME subtypes have heterogenous clinical outcomes and distinct feature genes. (A–C) Survival analysis indicated that the myCAF-rich, iCAF-rich, and PSC-rich TME subtypes have distinct clinical outcomes. The myCAF-rich TME was considered as the most significant classifier for overall survival (D, E) Forest plot showed the top 5 hazard (D) and protect (F) genes and their effect for predicting clinical outcomes.

Figure 4

Different CAFs-related subtypes TME cross talked with tumor cells and immune cells. (A) Dotplot exhibited the expression of clinical prediction signatures in distinct CAFs-related subtypes patients. (B) The CAFs and PSC subtypes infiltration ratio in four subtypes TME. (C) Multi-color IHC staining showed the tight co-localization of CAF, Macrophage(Mφ), and tumor cells(Tm). (D) Ductal cells and Macrophage interacted with iCAF_17 and iCAF_20 by IGF, EGF and MIF, CXCL signaling.

Figure 5

Existence and spatial arrangement of three key subtypes CAFs. (A, B) Multi-color IHC staining exhibited the existence of each CAFs subtypes. CD10(MME) for myCAF_12, KALF4 for iCAF_17, and TIAM1 for iCAF_19 (C) The spatial arrangement of major cell types and the dynamic change before and after chemotherapy. The grey dashed lines showed the boundary of tumor.