Berbamine inhibits the growth of liver cancer cells and cancer-initiating cells by targeting Ca²⁺/calmodulin-dependent protein kinase II

Abstract

Liver cancer is the third leading cause of cancer deaths worldwide but no effective treatment toward liver cancer is available so far. Therefore, there is an unmet medical need to identify novel therapies to efficiently treat liver cancer and improve the prognosis of this disease. Here, we report that berbamine and one of its derivatives, bbd24, potently suppressed liver cancer cell proliferation and induced cancer cell death by targeting Ca(2+)/calmodulin-dependent protein kinase II (CAMKII). Furthermore, berbamine inhibited the in vivo tumorigenicity of liver cancer cells in NOD/SCID mice and downregulated the self-renewal abilities of liver cancer-initiating cells. Chemical inhibition or short hairpin RNA-mediated knockdown of CAMKII recapitulated the effects of berbamine, whereas overexpression of CAMKII promoted cancer cell proliferation and increased the resistance of liver cancer cells to berbamine treatments. Western blot analyses of human liver cancer specimens showed that CAMKII was hyperphosphorylated in liver tumors compared with the paired peritumor tissues, which supports a role of CAMKII in promoting human liver cancer progression and the potential clinical use of berbamine for liver cancer therapies. Our data suggest that berbamine and its derivatives are promising agents to suppress liver cancer growth by targeting CAMKII. Mol Cancer Ther; 12(10); 2067-77. ©2013 AACR.

Figure 1

Structures of BBM and its derivative 2-methylbenzoyl berbamine (bbd24).

Figure 2. Berbamine (BBM) suppresses liver cancer cell growth and inhibited CAMKII phosphorylation

A&B. Cell proliferation was assayed 72 hours after BBM treatment for (A) liver cancer cells with epithelial morphology and (B) two mesenchymal–like cell lines and a normal embryonic liver cell line, CL48. C. Huh7 and SK-Hep-1 cells were used to generate xenografts. On the indicated days after starting BBM treatment, the sizes of Huh7 xenograft were measured. D. On Day 26, the mice were euthanized and the tumors were weighed. E. On the indicated days, the SK-Hep-1 xenograft tumors were measured. F. On Day 30, the tumors were weighed. *, p < 0.05.

Figure 3. BBM and its derivative bbd24 induce liver cancer deaths and inhibit the phosphorylation of CAMKII

A. Huh7 cells were incubated with 16 μg/ml BBM for 72 hours. The cells were then stained with PI and Annexin V-FITC. B. Huh7 cells, SK-Hep-1, and MHCC97H were treated with BBM for 48 hours, and then their lysates were analyzed by Western blot. HSP70 was used as a loading control. C. A derivate of BBM, bbd24, was applied to Huh7 cells with a concentration of 2 ug/ml. The photos were taken 6 hour after the treatment. D. The same cells were analyzed by flow cytometry at the different time points. E. Huh7 cells were treated with bbd24 for 48 hours, and then their lysates were analyzed by Western blot.

Figure 4. BBM preferentially targets liver cancer initiating cells and suppresses the phosphorylation of CAMKII

A. A series of doses of BBM were applied to the sorted CD133+ Huh7 cells, the death of which was analyzed by flow cytometry. B. Huh7 cells were treated with bbd24 for 24 hours, and then labeled with CD133-PE antibody and analyzed. C. Huh7, MHCC97H, and SNU398 cells were treated with BBM for 48 hours, and then trysinized and re-suspended with sphere formation conditioning medium, and seeded in low attachment plates. The spheres were photographed 1 week after seeding cells. D. The sorted CD133+ Huh7 cells were treated with BBM for 48 hours and then analyzed with Western blot.

Figure 5. Responses of liver cancer cell lines to a chemical inhibitor of CAMKII, KN93

A. A series of doses of KN93 and its non-targeting analogue KN92 were applied to the liver cancer cell lines. Cell proliferation assay with MTS was performed 72 hours after the treatment. *, p < 0.05. B. The Huh7 and MHCC97H cells under 5 μM KN93 were labeled with CD133-PE and CD90-FITC antibodies, respectively, and then analyzed by flow cytometry. C. Huh7, MHCC97H, and SNU398 cells were treated with KN93 or KN92 for 48 hours, and then trysinized and resuspended with sphere formation conditioning medium, and seeded in low attachment plates.

Figure 6. Knockdown of CAMKIIγ suppresses liver cancer cell growth, and overexpression of CAMKIIγ antagonizes the effect of BBM

A. shRNAs were expressed by a lentiviral vector pLKO.1-puro. An MOI of 3 was applied for infection of Huh7 cells. Each of the 4 pairs of shRNAs targeting CAMKIIγ and a scramble control was tested in Huh7 cells by Western blot. B. Cell proliferation assay with MTS was performed 96 hours after infection of Huh7, SK-hep-1, and MHCC97H by the pool of the 4 lentiviral vectors. C. The transduced Huh7 cells were selected by puromycin. The stable cells were used for the xenograft on NOD/SCID mice. The tumor sizes were measured on Day 7, Day 15, and Day 28 after the s.c. injection. The tumors were weighed on Day 28. *, p < 0.05. D. Overexpression of CAMKIIγ by infecting HepG2 and Huh7 cells with a retroviral vector pMSCV. E. The enhanced growth of the CAMKIIγ overexpression HepG2 and Huh7 cells. F. Reduced cytotoxicity of 5-FU by the CAMKIIγ overexpression in HepG2 cells. G. Tetracycline-derivative doxycycline (DOX) induction of CAMKIIγ by infecting HepG2 and Huh7 cells with a retroviral vector pRetroX-Tight-puro. H. DOX-induced CAMKIIγ expression antagonized the cytotoxicity of 72 hours BBM treatment. *, p < 0.05. I. The stable HepG2 and Huh7 cells with CAMKIIγ knockdown were treated with BBM at indicated concentrations. The cell survival was measured by MTS assay.

Figure 7. CAMKII is hyperphosphorylated in human HCC compared with the paired peri-tumor tissues

A. Western blot analyses of oncogenic signaling in liver tumors from patients with different tumor stages. HSP70 was used as a loading control. B. Tumor stages in patients with different pCAMKII between tumors and non-tumor tissues. ‘Down’ means the patients have down-regulated CAMKII phosphorylation in tumors compared with that in non-tumor liver tissues, while ‘Up’ means the patients have higher CAMKII phosphorylation in tumors than that in non-tumor tissues. *, p < 0.05. C. The quantification of CAMKII phosphorylation in patients of different tumor stages.