Figure 1.
Human melanoma cell lines and melanoma patient specimens showed high levels of basal autophagy.
(A) Melanoma cell lines were transiently transfected with an EGFP-LC3 expressing plasmid and grown in normal medium. 24 hours after transfection, the cells were fixed and the autophagosomes in the cells were visualized by the presence of LC3 puncta under fluorescence microscopy, with the percentages of cells showing LC3 puncta (mean ± SD) indicated. All tested cell lines showed positive punctation indicating high basal autophagy. (B) Western blot shows that melanoma cell lines have high basal LC3-II and p62 protein levels. Upon HBSS induced starvation, increased cleavage of LC3-I to LC3-II indicative of autophagy induction and decreased or unchanged p62 levels were seen. Extracts from autophagy competent and autophagy deficient iBMK cells served as positive and negative controls. (C) Immunofluorescence staining for endogenous LC3 on a human melanoma tissue microarray. Punctate LC3 localization for autophagosomes is indicated by red arrows (top panel). The percentages of specimens with punctate LC3 staining in malignant, metastatic and benign nevus groups are indicated in the table (bottom panel). (D) Western blot showed decreased expression levels of Atg7 (top panel) and impaired clonogenic survival (bottom panel) in response to lentiviral shRNA knockdown of the essential autophagy regulator Atg7.
Figure 2.
Autophagic flux is induced by CCI-779 and blocked by HCQ.
(A) Melanoma cell lines transiently transfected with an EGFP-LC3-expressing plasmid were seeded on coverslips and treated with 10 µM of CCI-779 and 20 µM of HCQ, alone and in combination for 24 hours. Autophagosomes were visualized by the presence of LC3 puncta. The drug combination treatment showed more autophagosome accumulation than either single agent alone. The percentages of cells showing LC3 puncta (mean ± SD) are indicated. (B) Western blot showed that the combination treatment resulted in considerable increases in LC3-II and increases in p62 protein levels compared to treatment with single agent HCQ alone, indicating the blockade of autophagic flux. (C) Quantitation of actin-normalized changes in p62 (upper panel) and the ratio of LC3-II/LC3-I (lower panel) after single and combination treatment in comparison to the untreated control.
Figure 3.
HCQ and CCI-779 synergize to induce melanoma cell death.
(A) Melanoma cells (2×104) were plated in 12 well plates, and treated with CCI-779 and HCQ alone and in combination at the concentrations indicated. The cell viability was examined at day 4 by Trypan blue stain based cell counting and normalized to untreated cells at the time zero. Relative viability showed that CCI-779 and HCQ cooperate to promote cell death. (B) Clonogenic survival assay showed that melanoma cells treated with CCI-779 and HCQ in combination at the concentration as circled in (A) impaired colony formation. (C) Synergistic effects were analyzed by CalcuSyn software, with a fixed ratio of CCI- 779 and HCQ concentration (1∶3) as indicated. Experimental points fall mostly below the borderline (CI = 1) indicating strong synergism. (D) Inhibition of melanoma cell growth in three-dimensional culture. Melanoma cell spheroids were grown as described in the Materials and Methods. After 72 hours of incubation with CCI-779 and HCQ alone and in combination, the cells were treated with cell LIVE/DEAD Viability kit wherein living cells stain green and dead cells stain red. Representative 3-D spheroid pictures showed synergism between CCI-779 and HCQ (scale bar = 500 µM) (left), with the graph at right displaying the percentages of spheroid diameters (mean ± SD) normalized to the untreated spheroids. Spheroids of UACC903 disintegrated after the treatment. (E) Combination HCQ and CCI-779 induced cell death via apoptosis. Western blot of melanoma cells treated with HCQ and CCI-779 at indicated concentrations demonstrated increased active caspase-3 indicating apoptotic cell death.
Figure 4.
CCI-779 and HCQ in combination inhibit UACC903 tumor xenograft growth.
(A) Nude mice bearing 100 mm3 tumors were given daily s.c. injection of 65 mg/kg HCQ and 0.5 mg/kg CCI-779 at the days indicated. Tumor size was measured once every three days. Each data point represents the mean ± SE tumor volume from 5 mice, 10 tumors, with tumor suppression greatest in mice treated with both CCI-779 and HCQ (*P = 0.001, **P = 0.014). (B) Representative pictures of immunohistochemical analysis showed that Ki67 was reduced in tumors treated with CCI-779, HCQ either alone or in combination. The proliferation index was calculated as the proportion of Ki67 positive cells among 200 cells in 4 randomly selected fields. Error bars represent standard errors. (*) P = 0.004; (**) P = 0.0013 (t-test) compared with the vehicle control. Active-caspase-3 staining was increased in tumors treated with HCQ, and was further increased in tumors treated with CCI-779 and HCQ. The apoptotic index was calculated as the proportion of active caspase 3 positive cells among 200 cells in 4 randomly selected fields. (*) P = 0.007; (**) P = 0.0014 (t-test) compared with vehicle control.