Affiliation:
1. Huazhong University of Science and Technology
2. Southern University of Science and Technology
Abstract
Abstract
Background
This study aimed to identify new protein markers that can evaluate the metastatic ability of Pancreatic neuroendocrine tumors (PanNETs) and predict prognosis together with Ki-67.
Methods
Multicentric series analysis was performed to evaluate the effect of metastasis on overall survival (N = 1,109). Single-cell RNA sequencing, proteomics (32 PanNETs and 10 normal pancreas tissues), and immunohistochemical staining (152 PanNETs) were conducted to screen proteins associated with PanNET metastatic capability. Associations between expression of screened proteins and survival were examined using Cox regression. In vitro experiments were performed to verify the function of candidate protein. A survival signature was constructed using recursive partitioning and survival forest analyses.
Results
The combination of G grade and metastasis predicted long-term survival more accurately though retrospective analysis. Single-cell RNA-seq highlighted the PanNET characteristics of metastasis that do not correspond with G grade and identified high metastatic capability subgroups in PanNET. Five proteins (PIN1, POSTN, SEMA4F, ASPN, and KCDT12), which may be related to the metastatic capability of PanNETs, were identified by proteomics. Random survival forest analysis showed that PIN1 could be a marker of PanNET metastatic ability. PIN1 overexpression promoted metastasis through LAMIN in PanNET cells. Finally, a survival signature with a consistency index of 0.921 and strong calibration was established, with good ability to predict patient survival.
Conclusion
In summary, we found that PIN1 can precisely evaluate the metastatic potential of PanNETs and, together with the proliferation marker Ki-67, can accurately predict the survival of PanNET patients.
Publisher
Research Square Platform LLC