Ganglioside GD2 Enhances the Malignant Phenotypes of Melanoma Cells by Cooperating with Integrins

Abstract

Gangliosides have been considered to modulate cell signals in the microdomain of the cell membrane, lipid/rafts, or glycolipid-enriched microdomain/rafts (GEM/rafts). In particular, cancer-associated gangliosides were reported to enhance the malignant properties of cancer cells. In fact, GD2-positive (GD2+) cells showed increased proliferation, invasion, and adhesion, compared with GD2-negative (GD2-) cells. However, the precise mechanisms by which gangliosides regulate cell signaling in GEM/rafts are not well understood. In order to analyze the roles of ganglioside GD2 in the malignant properties of melanoma cells, we searched for GD2-associating molecules on the cell membrane using the enzyme-mediated activation of radical sources combined with mass spectrometry, and integrin β1 was identified as a representative GD2-associating molecule. Then, we showed the physical association of GD2 and integrin β1 by immunoprecipitation/immunoblotting. Close localization was also shown by immuno-cytostaining and the proximity ligation assay. During cell adhesion, GD2+ cells showed multiple phospho-tyrosine bands, i.e., the epithelial growth factor receptor and focal adhesion kinase. The knockdown of integrin β1 revealed that the increased malignant phenotypes in GD2+ cells were clearly cancelled. Furthermore, the phosphor-tyrosine bands detected during the adhesion of GD2+ cells almost completely disappeared after the knockdown of integrin β1. Finally, immunoblotting to examine the intracellular distribution of integrins during cell adhesion revealed that large amounts of integrin β1 were localized in GEM/raft fractions in GD2+ cells before and just after cell adhesion, with the majority being localized in the non-raft fractions in GD2- cells. All these results suggest that GD2 and integrin β1 cooperate in GEM/rafts, leading to enhanced malignant phenotypes of melanomas.

Keywords: GEM/rafts; cancer-associated antigen; ganglioside; integrin; melanoma.

Figure 1

Establishment of cell lines expressing ganglioside GD2. (A) Biosynthetic pathway of gangliosides and their synthetic enzymes. (B) Expressions of gangliosides GD2 and GD3 in GD2+ (S1 and S6) and GD2− (V4 and V9) cells were analyzed by flow cytometry, using anti-GD2 and anti-GD3 mAbs. (C) Immunofluorescence images of the expression of GD2 in S1 and S6 cells using purified anti-GD2 mAb 220-51 at 4.5 and 36 μg/mL, and an FITC anti-mouse IgG 2nd antibody. Scale bar: 50 μm.

Figure 2

GD2+ cells showed increased proliferation, invasion, and cell adhesion. (A) Proliferation was analyzed using GD2+ and GD2− cells by the MTT assay. Cells were seeded and MTT solution was added and incubated on each day, as indicated. Absorbance was measured at 590/620 nm and relative absorbance was plotted. MTT assay was performed in triplicates. The means ± SD are presented, and analyzed using the two-way ANOVA with a Tukey post-hoc test between V-series vs. S-series cell lines. * p < 0.05. (B) Invasion activity was analyzed using GD2+ cells and GD2− cells with cell culture inserts. After 24 h of incubation, cells that invaded the lower chamber were stained with Giemsa, and the number was counted under a microscope. The invasion assay was performed in triplicates. The means ± SD are presented and analyzed using Student’s t-test. * p < 0.05 and ** p < 0.01. (C) Microscopic images of invaded cells are shown. Scale bar: 50 μm. (D), cell adhesion was analyzed by the RT-CES system using GD2+ and GD2− cells in collagen I-precoated plates.

Figure 3

Effects of anti-GD2 mAbs on cell proliferation and adhesion of GD2+ cells. (A) Proliferation was analyzed by the MTT assay using GD2+ S1 and GD2− V4 cells. Cells were seeded in each well, and purified anti-GD2 mAbs 220-51 were added at 72, 18, and 4.5 μg/mL. The MTT solution was added and incubated on each day as indicated. Absorbance was measured at 590/620 nm. The experiment was performed in triplicates (and the mean ± SD are presented) and analyzed by two-way ANOVA with Tukey post-hoc test between control (S1-C or V4-C) vs. anti-GD2 treated cell lines (S1+mAbs or V4+mAbs). *** p < 0.005 and ** p < 0.01. (B,C) Effects of treatment with anti-GD2 mAb 220-51 on cell adhesion were analyzed by the RT-CES system, using GD2+ and GD2− cells in collagen I-precoated microplates. Cells were seeded in the plates containing 200 μL of culture medium. Purified anti-GD2 antibody (220-51), 72 μg/200 μL, was added to the wells at 0.5 h (B) and 3.0 h (C), after starting. Antibody-treated cells are indicated as ‘T’.

Figure 4

Identification of glycolipid-associated molecules by EMARS/MS. (A) A scheme of the EMARS/MS approach. EMARS was performed using GD2+ S1 cells, anti-GD2 mAb 220-51, and HRP-conjugated anti-mouse IgG. (B) Western blotting of EMARS reaction products. FITC-labeled molecules were immunoprecipitated and subsequently immunoblotted with anti-FITC antibodies. Bands were visualized using ImmunoStar^TM LD detection kits and analyzed by Imager LI-COR^TM.

Figure 5

Physical association of GD2 and integrin β1. (A) Cell surface expressions of integrin β1 in GD2+ and GD2− cells were analyzed by flow cytometry using anti-integrin β1 mAb. (B) The mRNA expression of integrin β1 was analyzed by qRT-PCR using GD2+ and GD2− cells. The experiment was performed in triplicates and the mean ± SD are presented. (Ca) Expressions of integrin β1 as well as GD2 were analyzed by Western immunoblotting. GD2 and integrin β1 were detected separately with anti-GD2 mAb and anti-integrin β1 mAb. (Cb) The binding of ganglioside GD2 and integrin β1 was analyzed by immunoprecipitation with rabbit anti-integrin β1 antibodies, and subsequent immunoblotting with anti-integrin β1 mAb or anti-GD2 mAb.

Figure 6

Colocalization of GD2 and integrin β1. (A) Cell surface localization of GD2 and integrin β1 was examined by immunocytochemistry using GD2+ cells. After fixation and permeabilization, cells were stained with mouse anti-GD2 mAb and rabbit anti-integrin β1 mAb. Then, cells were incubated with an Alexa 568-conjugated goat anti-mouse IgG antibody and an Alexa 488-conjugated donkey anti-rabbit IgG antibody. Microscopic visualization was performed using a confocal microscope. The green color indicates integrin β1 and the red color indicates GD2. Scale bar = 15 μm. (B) Association between GD2 and integrin β1 was analyzed by PLA. GD2+ and GD2− cells were incubated with anti-GD2 and anti-integrin β1 mAb. Duolink^TM in situ PLA probes anti-mouse PLUS and anti-rabbit MINUS were added, then a ligation–ligase solution was added. Finally, an amplification reaction was carried out. Cells were visualized under a confocal microscope. Scale bar = 20 μm.

Figure 7

Adhesion signals activated during cell adhesion to collagen-coated plates. Tyrosine-phosphorylated proteins were detected during cell adhesion. (A) Diagram for preparing cells to obtain lysates during cell adhesion. GD2+ (S1) and GD2− (V4) cells were detached using 0.02% EDTA/PBS after culturing in serum-free DMEM at 37 °C for 23 h, and rotated at 37 °C for 1 h. Cell suspensions were placed in collagen I pre-coated plates in DMEM, and incubated as indicated at 37 °C. Then, cells were lysed and lysates were used for SDS-PAGE. (B) Immunoblotting was performed with anti-phosphotyrosine PY20 mAb. Anti-β actin Ab was used as a loading control.

Figure 8

Knockdown of integrin β1 and its effects on cell phenotypes. (A,B) Knockdown efficiency of integrin β1 was examined with 4 types of siRNA (37, 75, 74, and ITG1). Using cell lysates and RNAs from GD2+ and GD2− cells, Western immunoblotting (A) and qRT-PCR (B) were performed, respectively. Gene expression levels were analyzed using the Student’s t-test. ** p < 0.01. (C) Cell proliferation was analyzed by the MTT assay, using GD2+ and GD2− cells treated with anti-integrin β1 si-RNA ITG1. Cells (3 × 10³) were seeded in 96-well plates. MTT assay was performed, as described in Figure 2. The analysis was performed in triplicates (and the mean ± SD are presented) and analyzed by two-way ANOVA with the Tukey post-hoc test. * p < 0.05, ** p < 0.01. (D) Cell adhesion was analyzed by the RT-CES system. GD2+ and GD2− cells were transfected with integrin β1 si-RNA, ITG1, and used for RT-CES, as described in Figure 2. (E,F) Invasion activity was analyzed using GD2+ and GD2− cells treated by integrin β1 si-RNA ITG1 with cell culture inserts. (F) A summary of the invasion assay. The invasion assay was performed in triplicates (and the mean ± SD are presented) were analyzed by Student’s t-test. * p < 0.05. Scale bar = 20 μm.

Figure 9

Multiple phospho-tyrosine bands were detected during the cell adhesion of GD2+ cells. (A) Immunoblotting with anti-phosphotyrosine mAb PY20. GD2+ S1 and GD2− V4 cells were transfected with anti-integrin β1 si-RNA ITG1. After 36 h of culture in regular medium, the cells were prepared, as described in Figure 7. Then, the cells were lysed and used for immunoblotting. (B) Band intensities in A were scanned by Images J^TM and plotted for S1 bands (a) and V4 bands (b).

Figure 10

Intracellular distribution of integrin β1 before and during adhesion to CL type I. (A) GD2+ (S1) and control GD2− (V4) cells were detached using 0.5 mM EDTA/PBS. Cell suspension (5 × 10⁵ cells) was placed in collagen I pre-coated plates in DMEM, and incubated for 0~30 min at 37 °C. Then, cell lysates were prepared using 1% Brij 020 in a TNE buffer and separated by Optiprep gradient ultracentrifugation at 42,000 rpm and 4 °C for 5 h, and fractionated (500 μL). Each fraction (13.5 μL) was used for immunoblotting with anti-integrin β1 mAb, anti-GD2 mAb, anti-flotillin, or anti-caveolin-1 antibodies. Flotillin and caveolin-1 were used as GEM/raft markers. (B) Band intensities of integrin β1 were measured by ImageJ^TM software and the relative intensity of bands are presented.