Lipid metabolism characterization in gastric cancer identifies signatures to predict prognostic and therapeutic responses

Jiawei Zeng; Honglin Tan; Bin Huang; Qian Zhou; Qi Ke; Yan Dai; Jie Tang; Bei Xu; Jiafu Feng; Lin Yu

doi:10.3389/fgene.2022.959170

Lipid metabolism characterization in gastric cancer identifies signatures to predict prognostic and therapeutic responses

Front Genet. 2022 Nov 3:13:959170. doi: 10.3389/fgene.2022.959170. eCollection 2022.

Authors

Jiawei Zeng¹, Honglin Tan², Bin Huang³, Qian Zhou¹, Qi Ke⁴, Yan Dai⁵, Jie Tang¹, Bei Xu¹, Jiafu Feng¹, Lin Yu^{1

6}

Affiliations

¹ Department of Clinical Laboratory, Mianyang Central Hospital, School of Medicine, University of Electronic Science and Technology of China, Mianyang, China.
² Development and Regeneration Key Lab of Sichuan Province, Department of Histology and Embryology, Chengdu Medical College, Chengdu, China.
³ Emergency Department, Mianyang Central Hospital, School of Medicine, University of Electronic Science and Technology of China, Mianyang, China.
⁴ Department of Pathology, Mianyang Central Hospital, School of Medicine, University of Electronic Science and Technology of China, Mianyang, China.
⁵ Department of Ophthalmology, Mianyang Central Hospital, School of Medicine, University of Electronic Science and Technology of China, Mianyang, China.
⁶ NHC Key Laboratory of Nuclear Technology Medical Transformation, (Mianyang Central Hospital), Mianyang, China.

Abstract

Purpose: Increasing evidence has elucidated the significance of lipid metabolism in predicting therapeutic efficacy. Obviously, a systematic analysis of lipid metabolism characterizations of gastric cancer (GC) needs to be reported. Experimental design: Based on two proposed computational algorithms (TCGA-STAD and GSE84437), the lipid metabolism characterization of 367 GC patients and its systematic relationship with genomic characteristics, clinicopathologic features, and clinical outcomes of GC were analyzed in our study. Differentially expressed genes (DEGs) were identified based on the lipid metabolism cluster. At the same time, we applied single-factor Cox regression and random forest to screen signature genes to construct a prognostic model, namely, the lipid metabolism score (LMscore). Next, we deeply explored the predictive value of the LMscore for GC. To verify the specific changes in lipid metabolism, a total of 90 serum, 30 tumor, and non-tumor adjacent tissues from GC patients, were included for pseudotargeted metabolomics analysis via SCIEX triple quad 5500 LC-MS/MS system. Results: Five lipid metabolism signature genes were identified from a total of 3,104 DEGs. The LMscore could be a prognosticator for survival in different clinicopathological GC cohorts. As well, the LMscore was identified as a predictive biomarker for responses to immunotherapy and chemotherapeutic drugs. Additionally, significant changes in sphingolipid metabolism and sphingolipid molecules were discovered in cancer tissue from GC patients by pseudotargeted metabolomics. Conclusion: In conclusion, multivariate analysis revealed that the LMscore was an independent prognostic biomarker of patient survival and therapeutic responses in GC. Depicting a comprehensive landscape of the characteristics of lipid metabolism may help to provide insights into the pathogenesis of GC, interpret the responses of gastric tumors to therapies, and achieve a better outcome in the treatment of GC. In addition, significant alterations of sphingolipid metabolism and increased levels of sphingolipids, in particular, sphingosine (d16:1) and ceramide, were discovered in GC tissue by lipidome pseudotargeted metabolomics, and most of the sphingolipid molecules have the potential to be diagnostic biomarkers for GC.

Keywords: LMscore; gastric cancer; lipidome pseudotargeted metabolomics; sphingolipid metabolism; therapeutic response.