SINGLE-CELL TRANSCRIPTOMIC ANALYSIS OF SMALL INTESTINAL NEUROENDOCRINE TUMORS REVEALED POTENTIAL MECHANISMS OF MESENTERIC FIBROSIS

Background: Nearly half of patients with small intestinal neuroendocrine tumor (SI-NET) experienced local mesenteric fibrosis, significantly impacting their prognosis. However, the mechanism by which SI-NET causes mesenteric fibrosis remains unclear.
Aims: We aim to elucidate the molecular mechanisms underlying SI-NET fibrosis and explore potential therapeutic targets.
Methods: We conducted single-cell RNA sequencing (scRNA) on 5 primary tumor specimens and their corresponding adjacent non-tumor tissues, integrating one public SI-NET scRNA data (Figure 1A). Samples were grouped into SI-NET with mesenteric fibrosis (SINET_MF) and without mesenteric fibrosis (SINET_NF) groups based on pathological or radiological data. We obtained bulk RNA-seq data from public databases and performed validation using immunohistochemistry staining and ELISA.
Results: After dimensionality reduction and clustering, all samples were categorized into eight major cell lineages (Figure 1B). IGFBP3 expression was notably higher in SINET_MF tumor cells compared to SINET_NF, as validated by immunohistochemistry staining (Figure 1C, D). To identify the cell types primarily affected by the high expression of IGFBP3 in tumor cells, we developed a new algorithm that utilized bulk RNA sequencing data. We stratified the publicly available bulk RNA sequencing data for SINET into two groups based on the median expression of IGFBP3: the high IGFBP3 expression group and the low IGFBP3 expression group (Figure 1E). A total of 752 differentially expressed genes that were significantly upregulated in the high IGFBP3 expression group were designated as the IGFBP3-associated gene signature (P.adj < 0.05 & Foldchange > 1). The IGFBP3-related gene signature was associated with fibroblasts, endothelial cells, and perivascular-like cells (Figure 2A). LUM+ fibroblasts, responsible for extracellular matrix synthesis and collagen production, were significantly enriched in SINET_MF (Figure 2B). Trajectory and cell-cell interaction analysis revealed that IGFBP3 secreted by tumor cells binds to the TGFβ2/3 receptors of LUM+ fibroblasts, activating the TGFβ signaling pathway, thereby promoting mesenteric fibrosis (Figure 2C, D). ESM1+ endothelial cells, responsible for vascular budding, were enriched in SINET_MF (Figure 2E). IGFBP3 might promote the proliferation of ESM1+ endothelial cells by activating the PI3K-AKT signaling pathway (Figure 2F). Immature perivascular-like cells might cause endothelial destabilization and vascular leakage through ANGPT2-TIE2 interaction with endothelial cells in tumors.
Conclusions: We revealed a potential mechanism underlying mesenteric fibrosis in SINET, suggesting IGFBP3 as a potential drug target and biomarker for SI-NET patients with mesenteric fibrosis.
Keywords: small intestinal neuroendocrine tumors, single-cell RNA sequencing, mesenteric fibrosis, fibroblasts