Isolation and Characterization of a Chinese Hamster Ovary Heparan Sulfate Cell Mutant Defective in Both Met Receptor Binding and Hepatocyte Growth Factor NK1/Met Signaling.

BACKGROUND/AIMS: The up-regulation of hepatocyte growth factor/receptor, HGF/Met, signal transduction is observed in most of human cancers. Specific heparan sulfate structures enhance the HGF/Met signaling at both cell and animal-based model systems. Biochemical studies indicate that heparan sulfate interacts with HGF and a natural occurring splicing variant NK1 of HGF with similar affinity. However, it is currently unknown if cell surface heparan sulfate binds to Met at physiological conditions and if specific cell surface heparan sulfate structures are required for effective HGF/Met or NK1/Met signaling.

METHODS: An established flow sorting strategy was used to isolate a soluble Met recombinant protein-binding positive or negative CHO cell clones different only in specific heparan sulfate structures. The cell surface bindings were imaged by confocal microscopy and flow cytometry analysis. Glucosamine vs. galactosamine contents from media-, cell surface-, and cell association glycosaminoglycans were quantified by HPLC. 35S-sulfate labeled glycosaminoglycans were characterized by anion exchange and size-exclusion HPLC. Heparan sulfate disaccharide compositions were determined by HPLC-MS analysis. Western blot analyses of MAPK-p42/44 were used to monitor HGF- and NK1-facillated Met signaling.

RESULTS: CHO-Positive but not CHO-Negative cell surface heparan sulfate bound to Met recombinant protein and HGF/NK1 further promoted the binding. Overall glycosaminoglycan analysis results indicated that the CHO-Negative cells had reduced amount of heparan sulfate, shorter chain length, and less 6-O-sulfated disaccharides compared to that of CHO-Positive cells. Moreover, CHO-Negative cells were defective in NK1/Met but not HGF/Met signaling.

CONCLUSIONS: This study demonstrated that soluble Met recombinant protein bound to cell surface HS at physiological conditions and a Met /HGF or NK1/HS ternary signaling complex might be involved in Met signaling. Shorter HS chains and reduced 6-O-sulfation might be responsible for reduced Met binding and the diminished NK1-initiated signaling in the CHO-Negative cells. The unique CHO-Positive and CHO-Negative cell clones established in current study should be effective tools for studying the role of specific glycosaminoglycan structures in regulating Met signaling. Such knowledge should be useful in developing glycosaminoglycan-based compounds that target HGF/Met signaling.