Perturbation of hyaluronan interactions inhibits malignant properties of glioma cells

Abstract

Malignant progression of gliomas is characterized by acquisition of inappropriate growth and invasive properties. In vitro, these malignant properties are reflected in, and measured by, the ability to grow in an anchorage-independent manner and to invade artificial extracellular matrices. The results of numerous studies have suggested that the extracellular and pericellular matrix polysaccharide, hyaluronan, plays an important role in these attributes of malignant cancer cells. However, with respect to glioma cells, most studies have addressed the effect of exogenously added hyaluronan rather than the function of endogenous tumor cell-associated hyaluronan. In this study we manipulate hyaluronan-glioma cell interactions by two methods. The first is administration of small hyaluronan oligosaccharides that compete for endogenous hyaluronan polymer interactions, resulting in attenuation of hyaluronan-induced signaling. The second is overexpression of soluble hyaluronan-binding proteins that act as a competitive sink for interaction with endogenous hyaluronan, again leading to attenuated signaling. We find that both treatments inhibit anchorage-independent growth, as measured by colony formation in soft agar, and invasiveness, as measured by penetration of reconstituted basement membrane matrices. Based on our findings, we conclude that endogenous hyaluronan interactions are essential for these two fundamental malignant properties of glioma cells.

Figure 1.

Inhibition of glioma cell invasion by hyaluronan oligomers. A: C6 glioma cells (2 × 10⁴ per well) were incubated in the presence or absence of 100 μg/ml of hyaluronan (HA) oligomers for 72 hours and the number of cells that invaded the reconstituted basement membrane matrix was measured at each 24-hour interval, as described in Materials and Methods. B: C6 rat glioma cells (2 × 10⁴ per well), A172 human glioma cells (1 × 10⁴ per well), and U87 human glioma cells (1 × 10⁴ per well) were incubated in the presence and absence of 1 to 100 μg/ml of hyaluronan oligomers, 100 μg/ml of chitin oligomers, or 4 μg/ml of antibody against CD44 (Zymed), and the number of cells that invaded was measured after 24 hours. The results in B are presented as means of three experiments ± SD (control versus hyaluronan oligomers; P < 0.05 for each cell type).

Figure 2.

Inhibition of C6 glioma cell invasion by soluble HABPs. C6 cells were infected with recombinant adenoviruses driving expression of β-galactosidase (β-gal), soluble CD44 (solCD44), or brevican link module (BLM), then assayed for invasion as in Figure 1B ▶ . The insets show expression of the HABPs, as measured by Western blotting. Top inset: lanes 1 and 2, soluble CD44 adenovirus; lanes 3 and 4, β-galactosidase adenovirus; lanes 1 and 3, cell extract; lanes 2 and 4, medium. Western blotting was performed with antibody to CD44. Bottom inset: lanes 1 and 2, brevican link module adenovirus; lanes 3 and 4, β-galactosidase adenovirus; lanes 1 and 3, cell extract; lanes 2 and 4, medium. Western blotting was performed with antibody to the hemoagglutinin tag attached to BLM.

Figure 3.

Inhibition of anchorage-independent growth of C6 glioma cells by hyaluronan oligomers or by overexpression of soluble HABPs. A: C6 cells (2 × 10³ per well) were incubated for 1 week in soft agar in the presence or absence of 1, 10, or 100 μg/ml of hyaluronan oligomers, or 50 μg/ml of N-acetylglucosamine plus 50 μg/ml of glucuronic acid monomers, as described in Materials and Methods. The numbers of colonies >0.25 mm in size were counted. The results are presented as means of triplicates ± SD (control versus 10 or 100 μg/ml hyaluronan oligomers; P < 0.05). B: Western blot with antibody against phosphorylated Akt (p-Akt) or Akt for extracts of C6 cells grown in suspension. Lane 1, C6 cells alone; lane 2, C6 cells incubated with 100 μg/ml of chitin oligomers; lane 3, with 100 μg/ml of hyaluronan polymer (∼80 kD); lane 4, with 100 μg/ml of hyaluronan oligomers. C: C6 cells (2 × 10³ per well) were infected with recombinant adenoviruses driving expression of β-galactosidase (β-gal), soluble CD44 (solCD44), or brevican link module (BLM), then assayed for growth of colonies in soft agar. The results are presented as means of three separate experiments, each performed in triplicate, ± SD (control versus soluble CD44 or brevican link module; P < 0.05).