Abstract:

Objective To screen the differential prognostis ferroptosis genes of thyroid cancer （TC） and construct the prognostic risk model of TC ferroptosis related genes（FRGs）， and to clarify its potential mechanism at the molecular level. Methods The gene expression and clinical data were obtained from The Cancer Genome Atlas （TCGA） Database. The FRGs were obtained from FerrDb and GeneCards Databases， and R software was used to screen the PFRGs of TC；the PFRGs mRNA expressions in TC and thyroid tissues were obtained from TCGA and GTEx Databases， and the immunohistochemical results were obtainted from the Human Protein Atlas （HPA） Database to verify the differences in the expressions of PFRGs mRNA and protein； time-receiver operating characteristic （time-ROC） curve and Kaplan-Meier curve were used to evaluate the relationships between PFRGs and survival and prognosis of the TC patients； univariate and multivariate Cox regression analysis were used to calculate the risk scores of PFRGs expression， the clinical data of the TC patients were included， the independent prognostic analysis was performed， and the Nomogram chart was constructed. Spearman correlation analysis was used to obtain the correlation between the expressions of PFRGs and the expressions of other genes in TCGA Database and expressions of various genes，the correlation coefficient was calculated and the co-expressing genes were screened；the co-expression genes of PFRGs were analyzed by protein-protein interaction（PPI） network diagram，Geno Ontology（GO） and Kyoto Encyclopedia of Genes and Genomes（KEGG） enrichment analysis. Results A total 3 317 up-regulated genes and 3 456 down-regulated genes in TC were screened out by differential analysis；343 differentially expressed genes（DEGs） screened out by univariate Cox regression were associated with the survival and prognosis of the TC patients， including CD44， Annexin A1（ANXA1）， and nuclear receptor subfamily 4 group A member 1（NR4A1）. The Kaplan-Meier and time-ROC curves results showed that the expressions of CD44， ANXA1， and NR4A1 were associated with the survival and prognosis of the TC patients （P=0.048， P=0.005， P=0.036）， and all had good 1-year， 3-year， and 5-year survival prediction effects；the risk scoring system of three genes was constructed to calculate the risk score，and the risk score was a prognostic factor of the TC patients ［hazard ratio（HR）=8.882， 95%CI=1.561－50.547，P=0.014］， and the higher the risk score was，the worse the survival prognosis was［P=0.011，area under curve（AUC）=0.761，0.767，and 0.722）； the Nomogram chart （C-index=0.938） constructed by the risk score combined with the clinical characteristics of the TC patients had a good predictive effect on the survival of the TC patients. The co-expression and enrichment analysis results showed that TC ferroptosis and its co-expressing genes （DUSP1， DUSP5， DUSP6，FOS，IL1RAP， JUN， MET， RASGRF1，TGFA，TGFBR1，andTNFRSF1A）mediated the MAPK signaling pathway and affected the MAP kinase/serine/threonine tyrosine phosphatase activity and MAP kinase phosphatase activity，and regulated the inactivation of the MAPK activity. Conclusion The TC differential PFRGs CD44， ANXA1， and NR4A1 based on bioinformatics analysis are related to the survival and prognosis of the TC patients. The prediction model constructed by three genes has a good predictive ability，and its mechanism may be related to the polygenic network’s regulation of the MAPK signaling pathway.

Key words: Thyroid neoplasm, Ferroptosis, Risk score, Nomogram, Survival prognosis