Beta1-6 branching of cell surface glycoproteins may contribute to uveal melanoma progression by up-regulating cell motility

Abstract

Purpose: This study investigated the influence of integrin expression as well as the oligosaccharide structure of surface N-glycoproteins on cell behavior of two primary uveal (92-1 and Mel202) and two primary cutaneous (FM55P and IGR-39) melanoma cell lines.

Methods: Cell adhesion to fibronectin and cell migration on fibronectin (wound healing) were selected as the studied cell behavior parameters. The percentage of cells positive for expression of selected integrins was estimated by flow cytometric analysis. The influence of beta1-6 branched complex-type N-oligosaccharides on wound healing on fibronectin was investigated. Cell surface beta1-6 branched N-oligosaccharides were measured by their specific binding to PHA-L followed by flow cytometry, and the fibronectin receptors bearing beta1-6 GlcNAc branched N-linked glycans were identified. In addition, the transcript of GnT-V (the enzyme that catalyzes the addition of N-acetylglucosamine to the core mannose of di- and tri-antennary N-glycans through a beta1-6 linkage) was analyzed by semiquantitative RT-PCR.

Results: Unlike the two examined cutaneous melanoma cell lines, neither of the uveal melanoma cells adhered to fibronectin. The adhesion efficiency of IGR-39 cells was twice that of FM55P cells. In contrast, uveal melanoma cells repaired scratch wounds on fibronectin-coated surfaces twice as fast as cutaneous melanoma cells did. The expression of alpha(3)beta(1), alpha(4)beta(1), alpha(5)beta(1), and alpha(v)beta(3) integrins, acting as fibronectin receptors, differed between the tested cell lines, and no distinct pattern distinguished uveal melanoma from cutaneous melanoma except for high expression of alpha(4)beta(1) integrin on both FM55P and IGR-39 cells. The results also demonstrated that the high levels of alpha(3)beta(1), alpha(4)beta(1), and alpha(5)beta(1) integrin expression on IGR-39 cells promoted their strong attachment to fibronectin-coated surfaces. In addition, 92-1, Mel202, and FM55P cells showed no or low adhesion to fibronectin, perhaps the result of low expression of fibronectin receptors excluding high expression of alpha(4)beta(1) integrin in FM55P cells. Cell migration was significantly decreased in three out of four PHA-L-treated cell lines, suggesting that beta1-6 branched complex type N-oligosaccharides are critical for 92-1, Mel202, and FM55P cell motility. Semiquantitative RT-PCR analysis showed that the tested cells did not differ in mRNA levels of beta1-6 -N-acetylglucosaminyltransferase V. However, FACS analysis showed that 92-1, Mel202 and IGR-39 cells expressed significantly higher amounts of beta1-6 branched N-oligosaccharides on the cell surface than FM55P cells did. All examined alpha(3), alpha(5), alpha(v), and beta(1) integrin subunits were shown to bear beta1-6 branched N-linked glycans.

Conclusions: The role of integrins and their N-glycosylation in the regulation of uveal melanoma growth and progression is largely unknown. These results reveal that cell surface complex-type N-glycans with GlcNAc beta1-6 branches are important factors determining the migration of primary uveal melanoma cells on fibronectin.

Figure 1

Adhesion of uveal (92–1, Mel202) and cutaneous melanoma (FM55P, IGR-39) cells to fibronectin. Each result is the average of three independent experiments done in triplicate. All data are given as percentage of adhesion relative to adhesion on poly-L-lysine (taken as 100%). Error bars indicate standard deviations. Asterisk (*) indicates p<0.05.

Figure 2

Effect of phaseolus vulgaris agglutinin on repair of wounds in monolayers of 92–1, Mel202, FM55P, and IGR-39 cells. A line was scratched with a plastic pipette tip through the confluent monolayer of cells maintained in serum-containing RPMI 1640 on a fibronectin-coated surface. The wounded cultures were allowed to heal for 24 h at 37 °C in the presence or absence of 25 μg/ml phaseolus vulgaris agglutinin (PHA-L) in serum-containing RPMI 1640. A: Panels show migration of cells in the presence or absence of PHAL after 24 h. B: The extent of wound closure was quantified by measurements of the width of the wound space for each case. For this value, the width was measured at twenty different locations in the wound and the mean value was compared to the width of the original closure (0 h). Values are means ± standard deviation of three separate experiments. Asterisk (*) indicates p<0.05.

Figure 3

Expression of integrins on human uveal (92–1, Mel202) and cutaneous melanoma (FM55P, IGR-39) cells. Melanoma cells were examined by flow cytometry for the expression of α₃β₁, α₄β₁, α₅β₁, and α_vβ₃ integrins, and data were compared to cells incubated with normal mouse IgG. Fluorescein isothiocyanate (FITC)-conjugated rabbit anti-mouse IgG (Fab’)₂ fragments were used for detection. Fluorescence signals of 10,000 cells were counted for each integrin subunit tested. Histograms of cells versus log fluorescence were generated. A: Panels show FACS profile for integrin-positive cell lines. Colored areas indicate the fluorescence profile of cells after indirect fluorescence staining with anti-integrin monoclonal antibodies. Open histograms represent background fluorescence. Relative fluorescence is shown as a logarithmic scale of 4 log cycles on the x-axis, and cell number as a linear scale on the y-axis. Data from one of three similar experiments are presented. The negative control for each line is different in some experiments because the experiments were not run on the same occasion. B: Diagram shows percentage of melanoma cells expressing α₃β₁, α₄β₁, α₅β_1, and α_vβ₃ integrins. C: Diagram shows quantitation of data from flow cytometric analyses. Values are means ± standard deviation of three separate experiments. Asterisk (*) indicates p<0.05.

Figure 4

Flow cytometric analysis of phaseolus vulgaris agglutinin binding on the surface of human uveal (92–1, Mel202) and cutaneous melanoma (FM55P, IGR-39) cells. A: Histogram of fluorescence intensity with or without fluorescein isothiocyanate (FITC)-conjugated phaseolus vulgaris agglutinin. B: Quantification of data from flow cytometric analyses. Fluorescence intensity relative to negative control, representing means from three pooled experiments. Asterisk (*) indicates p<0.05.

Figure 5

Immunodetection of α₃, α₅, α_v, and β₁ in materials obtained after precipitation of 92–1, Mel202, FM55P, and IGR-39 cell extracts with phaseolus vulgaris agglutinin bound to agarose. One mg of the cell extracts were incubated overnight with phaseolus vulgaris agglutinin (PHA-L) immobilized on cross-linked 4% beaded agarose. Glycoproteins were released from the complexes by boiling in electrophoresis sample buffer before being subjected to 10% SDS–PAGE. Following separation, the proteins were blotted onto PVDF membrane. After being blocked the blots were incubated with one of the following antibodies specific for different integrin subunits: α₃, α₅, α_v, and β₁. Next, the membranes were incubated with the secondary antibodies either alkaline phosphatase conjugated goat anti-rabbit IgG (for α₃, α₅, α_v, integrin subunits) or alkaline phosphatase coupled goat anti-mouse IgG (for β₁ integrin subunit). Visualization of immunoreactive proteins was achieved with the use of 4-nitroblue-tetrazolium salt/5-bromo-4-chloro-3-indolylophosphate solution. Lane S shows position of molecular weight markers.