Serum Eicosanoids Metabolomics Profile in a Mouse Model of Renal Cell Carcinoma: Predicting the Antitumor Efficacy of Anlotinib

Abstract

Anlotinib (ANL) shows promising efficacy in patients with renal cell cancer (RCC). Here, for the first time, a serum eicosanoid metabolomics profile and pharmacodynamics in Renca syngeneic mice treated with ANL was performed and integrated using our previous HPLC-MS/MS method and multivariate statistical analysis. The tumor growth inhibition rates of ANL were 39% and 52% at low (3 mg/kg) and high (6 mg/kg) dose levels, without obvious toxicity. A total of 15 disturbed metabolites were observed between the normal group and the model group, and the intrinsic metabolic phenotype alterations had occurred due to the treatment of ANL. A total of eight potential metabolites from the refined partial least squares (PLS) model were considered as potential predictive biomarkers for the efficacy of ANL, and the DHA held the most outstanding sensitivity and specificity with an area under the receiver operating characteristic curve of 0.88. Collectively, the results of this exploratory study not only provide a powerful reference for understanding eicosanoid metabolic reprogramming of ANL but also offer an innovative perspective for the development of therapeutic targets and strategies, the discovery of predictive biomarkers, and the determination of effective tumor monitoring approaches.

Keywords: anlotinib; eicosanoids; metabolomics; pharmacodynamics; renal cell carcinoma.

Figure 1

In vivo antitumor activities of anlotinib in a Renca cell syngeneic model in mice. (A) The timeline of the animal experiment. (B) Bodyweight. (C) Bodyweight variance. (D) Relative tumor volume at indicated time points (n=5). Data are shown as means ± SEM. *p < 0.05, **p < 0.01 vs saline.

Figure 2

The typical tumors harvested from mice sacrificed on the last day of experiment. (A) The photographs of tumors on the final day. (B) Tumor weight of different groups. *p < 0.05.

Figure 3

The eicosanoids metabolic profiles among different treatments. (A) OPLS-DA score plots of three groups. (B) Random permutation test with 200 iterations. (C) The correlation bioplot of OPLS-DA model.

Figure 4

The eicosanoid metabolomic profiles of anlotinib. (A) DModX curve. (B) PCA. (C) OPLS-DA. (D) Permutation test with 200 iterations. (E) Differential metabolite correlation heatmaps. (F) HCA of eicosanoid metabolomic profiling.

Figure 5

The three significant changed metabolites in Renca cell syngeneic mice treated by anlotinib (6 mg/kg) when the cutoffs were set as VIP>1 and p<0.05. (A) 5-HE-TE. (B) DHA. (C) 19,20-EpDPA. (D) The heatmap of the above-mentioned metabolites. *p < 0.05.

Figure 6

The supervised PLS model for correlating metabolic profiles with the efficacy of anlotinib. (A, D) are initial PLS for the first latent variables (ratio of peak area of metabolites, X block) of tumor volume (ANL, 3 mg/kg and 6 mg/kg, Y block) prediction model, respectively. (B, E) are loading plots for tumor volume, respectively. The square (blue) means the response variable; each triangle represents a metabolite, and the triangles (red) show the metabolites with VIP > 1.0. (C, F) are refined models to predict tumor volume based on the screened biomarkers.

Figure 7

The OPLS-DA to differentiate the efficacy groups based on the screened biomarkers. Green shape marks indicate mice treated by saline (A) or (B), blue stars mean mice treated by ANL [3 mg/kg, (A)], and red circle represents mice treated by ANL [6 mg/kg, (B)]. (C) The ROC curve of DHA for the ANL (6 mg/kg) group.