Dihydroartemisinin inhibits TCTP-dependent metastasis in gallbladder cancer

Abstract

Background: Patients with metastatic or relapsed gallbladder cancer generally have a poor prognosis. Therefore, targeting metastasis is one arm of therapeutic strategies to treat gallbladder cancer.

Methods: Levels of translationally controlled tumor protein (TCTP) were measured in samples of gallbladder cancer by immunohistochemical staining. Wound healing, migration and invasion assays were used to investigate the motility of cells. Western blot assay was used to investigate the levels of TCTP and other proteins. Liver metastasis models and lung metastasis models were established to investigate the inhibitory effect of Dihydroartemisinin on gallbladder cancer metastasis.

Results: TCTP is aberrantly expressed in gallbladder cancer patients and associated with metastasis and a poor prognosis. Depleting TCTP significantly inhibited gallbladder cancer cell migration and invasion. We found that Dihydroartemisinin as a potent inhibitor of TCTP inhibited TCTP-dependent cell migration and invasion by reducing cell division control protein 42 homolog (Cdc42) activation. In addition, in mice with xenografted tumors, treatment with Dihydroartemisinin decreased gallbladder cancer cell metastases and improved survival.

Conclusions: These findings provide new insights into the therapeutic activity of Dihydroartemisinin as a treatment for gallbladder cancer metastasis.

Keywords: Dihydroartemisinin; Gallbladder cancer; Invasion; Metastasis; TCTP.

Fig. 1

TCTP is associated with gallbladder cancer metastasis. a The expression levels of TCTP were detected in 73 gallbladder cancer (GBC) specimens and 103 cholecystitis tissue using IHC staining. Representative IHC images of TCTP expression are shown. b The average staining scores for TCTP expression in GBC and cholecystitis tissues were measured using IHC. ***, p < 0.001, Mann–Whitney U test. c TCTP IHC staining scores for non-metastatic and metastatic GBC tissues obtained from patients. ***, p < 0.001, Mann–Whitney U test. d IHC staining scores for TCTP expression in microvascular and neural invasive and non-invasive tissue samples obtained from GBC patients. ***, p < 0.001, Mann–Whitney U test. e TCTP mRNA levels were detected using qPCR in 5 primary tumor and metastatic lymph node samples. f Kaplan–Meier plots of the overall survival of GBC patients based on TCTP-high (n = 54) or low (n = 19) level expression

Fig. 2

TCTP promotes GBC cell migration and invasion. a Western blot analysis of endogenous TCTP expression in six GBC cell lines. β-Actin was used as a loading control. b Analysis of TCTP expression in control and TCTP knock-down GBC cells. c-e Wound-healing, migration and invasion assays were used to compare GBC cells transfected with either scramble or TCTP shRNA. Cell motility, migration and invasion were quantified by setting the observations in the scramble controls as 100%. Each experiment was independently performed three times, and the graphed data represent the mean ± SD. **, p < 0.01

Fig. 3

DHA inhibits TCTP-dependent cell migration and invasion. a In the migration assays, the TCTP-positive cell lines NOZ, GBC-SD, and OCUG-1, and the TCTP-negative cell lines EH-GB-2 and SGC-996 were pre-treated with either vehicle or DHA (40 μM) for 2 days and then seeded in transwell plates for 24 h. b The relative migration rates are shown in a bar graph. c, d SGC-996 cells were transfected with an empty or TCTP expression vector (c), treated with vehicle or DHA for 2 days, and then seeded in transwell plates for migration assays. The relative migration rates are shown in a bar graph (d). e, f TCTP was depleted in NOZ cells using shRNA (e). The cells were then treated with DHA for 2 days and seeded in transwell plates for migration assays. The relative migration rates are shown in a bar graph (f). The percentage of cells that migrated was scored and normalized to the percentage of migrated vehicle-treated cells. The graphed data represent the mean ± SD of 3 independent experiments. *, p < 0.05; **, p < 0.01, ns: no significant difference (p > 0.05)

Fig. 4

DHA altered cell attachment, spreading, and focal adhesions. a The morphologies of GBC-SD and OCUG-1 cells that were treated with DHA were examined under a microscope. b GBC-SD and OCUG-1 cells were treated with DHA for 2 days. They were then re-plated on fibronectin-coated coverslips for 6 h before being washed, fixed, and stained with rhodamine-conjugated phalloidin. The attached cells were counted using light microscopy. c GBC-SD and OCUG-1 cells were processed and stained as described in (b). Images of the cells were taken using a confocal microscope, and cell sizes were measured using Volocity software. At least 50 cells were measured for each time group. d Representative images of focal adhesions in GBC-SD and OCUG-1 cells that were treated with DHA. Cells were re-plated on fibronectin-coated coverslips, fixed, and stained with anti-vinculin antibodies and rhodamine-conjugated phalloidin. The cells were visualized using a confocal microscope. e Total cellular protein levels were detected using western blot analysis with antibodies against vinculin and E-cadherin. β-actin was used as an internal control

Fig. 5

DHA reduces Cdc42 activation in TCTP-expressing cells. a, b NOZ and GBC-SD cells, which express TCTP, and SGC-996 cells, which do not express TCTP, were treated with DHA for 2 days. Cdc42 activation was analyzed using a configuration-specific monoclonal antibody that specifically recognizes Cdc42-GTP. The level of active Cdc42 was then quantified and normalized to the level of total Cdc42 in each sample, and all results were normalized to the level in the vehicle-treated control cells in each experiment. c, d NOZ cells were transfected with a negative scramble or a siRNA targeting Cdc42. The effect of knockdown was verified using immunoblotting. The NOZ cells were then treated with DHA for 2 days before they were seeded in transwell plates for migration assays. The relative migration rates are shown in a bar graph. e, f NOZ cells were transfected with an empty or myc-tagged active Cdc42 Q61L vector. Cdc42 Q61L overexpression was verified using immunoblotting. The NOZ cells were then treated with DHA for 2 days and plated in transwell plates for migration assays. The relative migration rates are shown in bar graph. The graphed data represent the mean ± SD of 3 independent experiments. *, p < 0.05; **, p < 0.01, ns: no significant difference (p > 0.05)

Fig. 6

DHA reduces TCTP-dependent metastasis in vivo. a GBC cells were injected into the spleens of immunodeficient mice to establish a spleen-to-liver metastasis model, and the mice were then treated with DHA or vehicle control (PBS) via an IP injection every day. Representative photos of histological H&E-stained liver metastasis tissues are shown for each group. b-c A bar graph summarizing the number of liver metastases in the DHA-treated and control NOZ (b) and EH-GB-2 (c) cells. d-e A bar graph summarizing the incidence of liver metastasis in the DHA-treated and control NOZ (d) and EH-GB-2 (e) cells. f To establish a lung metastasis model, mice were intravenously injected into the tail with NOZ cells that expressed luciferase and then treated with DHA or vehicle control (PBS) via an IP injection every day. The bioluminescence of the cells was monitored every 2 weeks. Proton flux was evaluated using Xenogen IVIS LuminaXR software. The data shown represent the mean ± SD. **p < 0.001. NS: no significant difference. g Representative photos of histological lung metastasis tissues are shown for each group. A bar graph is used to summarize the number of lung metastases in the DHA-treated and control groups. h Kaplan–Meier plots of survival in the mice in the DHA-treated and control groups. Each group contained 10 mice