Integration of Different "-omics" Technologies Identifies Inhibition of the IGF1R-Akt-mTOR Signaling Cascade Involved in the Cytotoxic Effect of Shikonin against Leukemia Cells

Abstract

Hematological malignancies frequently have a poor prognosis and often remain incurable. Drug resistance, severe side effects, and relapse are major problems of currently used drugs, and new candidate compounds are required for improvement of therapy success. The naphthoquinone shikonin derived from the Chinese medicinal herb, Lithospermum erythrorhizon, is a promising candidate for the next generation of chemotherapy. The basal cellular mechanism of shikonin is the direct targeting of mitochondria. Cytotoxicity screenings showed that the compound is particularly effective against leukemia cells suggesting an additional cellular mechanism. mRNA and miRNA microarrays were used to analyze changes in gene expression in leukemia cells after shikonin treatment and combined with stable-isotope dimethyl labeling for quantitative proteomics. The integration of bioinformatics and the three "-omics" assays showed that the PI3K-Akt-mTOR pathway was affected by shikonin. Deregulations of this pathway are frequently associated with cancerogenesis, especially in a wide range of hematological malignancies. The effect on the PI3K-Akt-mTOR axis was validated by demonstrating a decreased phosphorylation of Akt and a direct inhibition of the IGF1R kinase activity after shikonin treatment. Our results indicate that inhibiting the IGF1R-Akt-mTOR signaling cascade is a new cellular mechanism of shikonin strengthening its potential for the treatment of hematological malignancies.

Figure 1

Cytotoxic effect of shikonin against U937 leukemia cells. (a) Chemical structure of shikonin. (b) Dose response curves of U937 cells after treatment with different concentrations of shikonin. Resazurin reduction assays were performed to determine dose response curve. Viability of U937 cells is represented by mean ± SEM of three independent experiments, and it is expressed as percentage survival of control.

Figure 2

Results and relations of mRNA microarray, miRNA microarray, and stable-isotope dimethyl labeling for quantitative proteomics. Only molecules from the omics datasets that met the expression fold change cutoff of ≥±1.2 are shown in this diagram. The intersections of the Venn diagram indicate the number of corresponding molecules deregulated in different assays independently from the direction of their deregulated expression.

Figure 3

Complementarity of various “-omics” technologies provides a system-level understanding of shikonin's effects in U937 cells. (a) Comparison analysis of molecular and cellular functions affected by shikonin in three “-omics” assays. The bar graph displays only functions disturbed in all three assays. Right-tailed Fisher's exact test was used to calculate a P value determining the probability that each biological function assigned to the datasets is due to chance alone. (b) Overlapping signaling pathways deregulated after shikonin treatment. Numbers of identical molecules deregulated in overlapping pathways are shown.

Figure 4

Effect of shikonin on the PI3K-mTOR signaling cascade. (a) U937 cells were treated with 0.3 μM shikonin for 24 h and subsequently stained with phospho-specific antibodies against p-Akt and p-ribosomal protein S6. Shikonin treatment significantly decreased the amount of phosphorylated Akt (left shift), but no effect on the phosphorylation status of ribosomal protein S6 was detected. (b) Statistical quantification of p-Akt or p-ribosomal protein S6 negative cells after shikonin treatment. Data points represent mean ± SD of at least three independent experiments. (*Significantly different according to Student's t-test, P < 0.05).

Figure 5

Dose response curve of shikonin in an IGF1-R kinase assay. A radiometric protein kinase assay (33PanQinase Activity Assay) was used for measuring the kinase activity of IGF1R. IC50 values were calculated by nonlinear regression using Prism 5.04 (Graphpad, CA, USA).