Acetylshikonin extract was obtained from Huakang Pharmaceutical (China) and was confirmed by high-performance liquid chromatography (HPLC). 3-(4,5-dimethylthiazol- 2-yl)-2,5-diphenyltetrazolium bromide (MTT) was obtained from Sigma Chemical (USA). Cyclophosphamide was obtained from Hengrui Pharmaceutical (China).

Malignant cell lines in this study include human lung adenocarcinoma epithelial cell line A549 (ATCC CCL-185), human breast adenocarcinoma cell line MCF-7 (ATCC HTB-22TM) and mouse Lewis lung carcinoma (LLC) (ATCC CRL-1642) were obtained from American Type Culture Collection (USA). Human hepatocellular carcinoma cell line Bel-7402 was obtained from the Cell Bank of the Chinese Academy of Sciences. The cells were cultured in RPMI 1640 (Gibco, USA) supplemented with 10% fetal bovine serum (Minhai Bio-engineering, China) and maintained at 37°C with 4% CO₂ in a humidified atmosphere. Cell viability was determined with 0.1% trypan blue.

MTT assay 2 was performed to measure the anti-proliferation effects of acetylshikonin on the cell lines of A549, Bel-7402, MCF-7 and LLC. Acetylshikonin was diluted and added to target cells in triplicates with final concentrations at 25.6, 12.8, 6.4, 3.2, 1.6, 0.8, 0.4 μg/ml. The cells were incubated for 48 hours and 20 μl of 5 mg/ml solution of MTT in phosphate-buffered saline (PBS) was added to triplicate samples and the plates were incubated for additional 4 hours. The plates were then centrifuged and the medium was removed. Two hundred microliters (200 μl) of DMSO was added to each well to dissolve the purple blue sediment, the absorbance was determined at 590 nm on a microplate reader (Model 550, Bio-Rad, USA). The inhibition rate was calculated as follows:

The 50% inhibitory concentrations (IC₅₀) of the 48 hours were calculated with Bliss assay.

Similarly, A549 cell was used to observe the effects of acetylshikonin on growth curve at various time points. Acetylshikonin was added to A549 cell with various final concentrations (3.2, 1.6, 0.8 μg/ml). MTT assay was performed on the cells in triplicates for each concentration after the cells were incubated for 12, 24, 48, 72 hours respectively. Adriamycin (0.1 μg/ml) served as a positive control.

The C₅₇BL/6 mice (Experimental Animal Center, West China Center for Medical Sciences, Sichuan University, China) were transplanted with LLC according to protocols of transplanted tumor research. At 24 hours after tumor transplantation, the mice were divided into five groups (12 mice per group) randomly: (1) control group (0.9% normal saline), (2) cyclophosphamide group (60 mg per kg of body weight), (3) acetylshikonin group (0.5 mg per kg of body weight), (4) acetylshikonin group (1 mg per kg of body weight) and (5) acetylshikonin group (2 mg per kg of body weight). Mice in cyclophosphamide group received injections only on Day 1, while mice in all other groups received injections once every two days (six times in total). All injections were carried out intraperitoneally. Data were collected and calculated as follows:

(1) Tumor volume: The length (A) and width (B) of the tumor were measured of the tumor issue of each mouse once every two days since Day 5. Tumor volume (V) = AB²/2

(2) Inhibitory rate: The animals were sacrificed on Day 13 and tumors were exercised and weighed. Tumor inhibition rate (%) = (1- tumor tissue weight of treatment group/tumor tissue weight of control group) ×100%

The animal handlings and experimental procedures were approved by the Animal Ethics Committee of Sichuan University.

Tumor tissues were fixed with 4% formaldehyde solution at 4°C for 24 hours, dehydrated in graded concentrations of ethanol embedded in paraffin and sliced.

Streptavidin/biotin-peroxidase (SP) method was used for immunohistochemical staining. The primary antibodies, namely bcl-2, bax and caspase-3 (Wuhan Boster Biological Technology, China), were diluted at 1:100. PBS was used as control. Each slice was photographed and the integrated optical density (IOD) was measured with Image pro plus 5.02 (Media Cybernetics, USA).

Data were expressed as mean ± standard deviation (SD) unless otherwise indicated. Statistical differences between the treatment and control groups were determined by Mann-Whitney test with SPSS 12.0 (SPSS, USA). P < 0.05 was considered statistically significant.

After treatment with acetylshikonin for 48 hours, the untreated tumor cells grew and the cytoskeletons were clearly visible under inverted light microscope. Cells treated with 25.6 μg/ml acetylshikonin became round in shape and condensed nuclei were seen, many of which lost intact membranes, leading to necrosis (Figure 2).

Cell inhibition rates were determined by MTT assay. The results revealed that acetylshikonin inhibited the growth of A549, Bel-7402, MCF-7 and LLC in a dose-dependent manner. IC₅₀ for 48 hours were 5.6 ± 0.86 μg/ml, 6.82 ± 1.5 μg/ml, 3.04 ± 0.44 μg/ml and 2.72 ± 0.38 μg/ml respectively (Figure 3). Acetylshikonin inhibited the growth of A549 in a time-dependent manner as indicated in the growth curve (Figure 4).

Since Day 5, significant differences were observed between the control and cyclophosphamide and acetylshikonin groups. Tumor volume of the mice in the cyclophosphamide group increased slowly, suggesting that the growth of LLC was suppressed by cyclophosphamide. Tumor volume of the mice in the acetylshikonin group (2 mg/kg) was markedly smaller than that in the control group, suggesting that the growth of LLC was suppressed by acetylshikonin (Figure 5).

Cyclophosphamide and acetylshikonin (1, 2 mg/kg) had significant inhibitory effects on the growth of LLC in mice. Inhibitory rate of the cyclophosphamide group was 67.81%; inhibition rates of the three acetylshikonin groups were 42.85%, 21.86% and 11.11% respectively, all in a dose-dependent manner (Table 1, Figure 6).

The photomicrograph of immunohistochemistry staining is shown in Figure 7. The positive reaction located in cytosol was stained in brown. The color of the stain is positively correlated to the protein expression.

The IOD of each group indicates that the expression of bax and caspase-3 in the acetylshikonin groups increased, whereas bcl-2 decreased in the same groups, resulting in higher bax/bcl-2 ratios, all in a dose-dependent manner (Table 2).