The significance of galectin-3 as a new basal cell marker in prostate cancer

Abstract

Prostate cancer may originate from distinct cell types, resulting in the heterogeneity of this disease. Galectin-3 (Gal-3) and androgen receptor (AR) have been reported to play important roles in the progression of prostate cancer, and their heterogeneous expressions might be associated with different cancer subtypes. Our study found that in various prostate cancer cell lines Gal-3 expression was always opposite to AR expression and other luminal cell markers but consistent with basal cell markers including glutathione S-transferase-π and Bcl-2. This expression pattern was confirmed in human prostate cancer tissues. Our results also showed that prostate cancer cells positive with basal cell markers were more aggressive. Downregulation of Gal-3 expression resulted in increased apoptotic potential and decreased metastasis potential of prostate cancer cells. Our findings demonstrate for the first time that Gal-3 may serve as a new marker for basal characteristics of prostate cancer epithelium. This study helps us to better understand the heterogeneity of prostate cancer. The clinical significance of this study lies in the application of Gal-3 to distinguish prostate cancer subtypes and improve treatment efficacy with designed personalized therapy.

Figure 1

The expression pattern of Gal-3 and AR in human prostate cancer cells. The expression levels of related proteins were analyzed by western blot. (a) The expression of Gal-3 and AR in LNCaP, DU145, and PC3 cells. (b) Gal-3 modulation had no effect on endogenous AR level. VC, cells transfected with a non-target control vector; no. 29-11 and no. 29-23, LNCaP clones with Gal-3 overexpression; siGal3-11 and siGal3-19, PC3 clones with Gal-3 knockdown. (c) AR overexpression had no effect on Gal-3 level in PC3 cells. PC3-AR, PC3 cells transfected with AR. β-Actin was used as the loading control. Data are representative of three independent experiments

Figure 2

Gal-3 expression may serve as a new basal cell marker for human prostate cancer cells. The expression profile of basal and luminal markers in LNCaP, DU145, and PC3 cells (Aa); LNCaP, C4-2B, and VUI3 cells (Ab); and normal prostate epithelial cells PZ-HPV-7 (B). The expression pattern of Gal-3 is in accordance with markers for basal phenotype such as GST-π and Bcl-2 but opposite with makers for luminal phenotype such as CK-18 and AR. β-Actin was used as the loading control. Data are representative of three independent experiments. (C) The pattern of opposite expressions of Gal-3 and AR in human prostate cancer tissues. Double staining of Gal-3 and AR in the same section was performed using immunohistochemistry. Gal-3-positive cells were negative for AR (arrows point to Gal-3, which is represented with bluish-gray staining in nucleus and cytoplasm), whereas AR-positive cells (arrows point to brownish-red color in nucleus) were negative for Gal-3. Line boxed pictures localized in the upper right corner of each picture represent × 400 magnification of corresponding areas in × 200 magnification pictures. (a) Gleason 2; (b) Gleason 5

Figure 3

More aggressive phenotype of prostate cancer cells expressing basal cell markers. (A) VUI3 cells had a higher growth rate. The alive cell number was counted using automated cell counter. (B) VUI3 cells were resistant to the chemotherapeutic drug cisplatin, and the resistance was reversed by Gal-3 knockdown. (Ba) Cells were treated with 50 μM cisplatin for 16 h and then subjected to western blot analysis of active caspase-3. β-Actin was used as the loading control. 1 and 4, LNCaP; 2 and 5, C4-2B; 3 and 6, VUI3. (Bb) FITC annexin V apoptosis detection. Cells were treated with 100 μM cisplatin for 4 h and then stained with FITC annexin V/7-AAD and analyzed by flow cytometry. Early-apoptotic cells were shown in the lower right corner of the image. (Bc) Gal-3 knockdown increased the number of apoptotic VUI3 cells in response to cisplatin. VUI3 cells were transfected with Gal-3 siRNA or control siRNA as described in the Materials and methods section. Thirty-eight hours later, cells were treated with 100 μM cisplatin for 4 h, stained with FITC annexin V/7-AAD, and analyzed by flow cytometry. VUI3 cells showed higher colony formation rate in six-well plates (C) and soft agar (D). Upper panel, picture of colonies; lower panel, the graph of colony-forming efficiency. (E) Gelatin zymography assay showed more active MMP-2 and MMP-9 in VUI3 cells. (F) Chemoinvasion assay showed more VUI3 cells invaded through Matrigel. (a) C4-2B or VUI3 versus LNCaP; (b) VUI3 versus C4-2B. Error bars represent S.D.; *P<0.05, **P<0.01, ***P<0.001. Data are representative of three independent experiments

Figure 4

Confirmation of Gal-3 as a new basal marker in human prostate cancer tissue assays. Double staining of Gal-3/AR (left) and single staining of p63 (right) were performed by immunohistochemistry. Gal-3 positive, bluish gray in nucleus and cytoplasm; AR positive, brownish red in nucleus; p63 positive, red in nucleus. Gal-3-positive cells were negative for AR staining but were in accordance with p63-positive areas. Line boxed pictures localized in the lower right corner of each picture represent × 400 magnification of corresponding areas in × 200 magnification pictures. Arrows indicate Gal-3⁻/AR⁺ areas, which are also p63⁻. (a) Normal; (b) Gleason 2; (c) Gleason 3; (d) Gleason 4

Figure 5

Downregulation of Gal-3 increased apoptotic potential and decreased metastasis potential of prostate cancer cells in a xenograft model. Appropriate cancer cells were injected into nude mice, and the tumor model was established as previously described. Tumors were sectioned and subjected to immunohistochemistry as described in the Materials and methods section. The positive staining for all related proteins was shown in brown color. VC, PC3 cells transfected with a non-target control vector; siGal3-11, PC3 clone with Gal-3 knockdown. Magnification, × 200