Fig. 7

Diagnostic value of serum RRM2 in predicting liver cancer. a ROC curves for serum RRM2, AFP, and the combination of serum RRM2 and AFP for the discriminating patients with liver cancer from healthy individuals. The AUC value represents the combined effects of the sensitivity and specificity of single or combined biomarkers in the diagnosis of patients with liver cancer. All the data were obtained from three biological replicates. b Serum RRM2 is positively associated with tumor stage in liver cancer patients. The first quartile of serum RRM2 levels in liver cancer patients was no more than 300 pg/ml, the second quartile was from 300 to 1200 pg/ml, and the bottom quartile was more than 1200 pg/ml. Liver cancer patients were divided into three groups according to these quartiles. All the data were from three biological replicates. The associations between tumor stage and serum RRM2 in liver cancer patients were analyzed using the χ² test. c Correlation between intracellular RRM2 and RRM2 in culture medium from liver cancer cell lines as analyzed by Spearman rank-correlation analysis. d RRM2 expression across multiple cancer types from the UALCAN database. TPM, transcriptions per million. e The model of the study. Under homeostasis, RRM2 is phosphorylated at the T33 site (can be dephosphorylated by NU6102) to sustain GSS and GSH levels and suppress potential ferroptosis. In the ferroptotic state, RRM2 is dephosphorylated and has increased affinity GSS, leading to the subsequent proteasomal degradation of both proteins. With such an effect, GSH eventually declines to facilitate ferroptosis. The data from a-c are shown from three biological replicates. Analysis of the receiver operator characteristics (ROC) and calculation of the area under the curve (AUC) was performed to find the diagnostic values of RRM2, AFP and the combination of RRM2 and AFP for the prediction of liver cancer. **P < 0.01 indicates statistical significance. Data from b were analyzed using the χ² test. Data from c were analyzed by Spearman rank-correlation analysis. Data in d were analyzed using Student’s t test