Featured Article - May 2011
Short description: Proteoglycans can exert opposing effects on neuronal growth by competing to control the oligomerization of a common cell surface receptor.
Type IIa receptor protein tyrosine phosphatases (RPTPs), such as RPTPσ, LAR and RPTPδ, are cell surface receptors with important functions in neuronal development, function and repair. The extracellular regions, or ectodomains, of RPTPs interact with heparan sulfate proteoglycans (HSPGs) and chondroitin sulfate proteoglycans (CSPGs) with typically opposing effects on cell function, but how these opposing effects are mediated at the molecular level has been unknown.
Reporting in Science, Coles and colleagues show that neurocan, a CSPG, reduces outgrowth of dorsal root ganglion neurons, whereas in RPTPσ−/− neurons this inhibitory effect is decreased. Conversely, glypican-2, a HSPG, strongly promotes outgrowth of wild-type, but not RPTPσ−/−, neurons. They further show that the glycosaminoglycan (GAG) chains of neurocan and glypican-2 must be involved, and that their opposing effects are mediated through a common receptor, RPTPσ. Previous mutagenesis studies suggested a shared GAG-binding site in the N-terminal Ig domain of RPTPσ, so the authors analyzed the structural basis of proteoglycan recognition.
The crystal structures of the two N-terminal Ig domains (Ig1-2) of various members of the RPTP family across different species reveal a V-shaped arrangement of Ig1 and Ig2, which is stabilized by conserved interactions. Residues of RPTPσ previously shown to mediate GAG binding lie on loops between Ig1 β-strands C-D and E-F, forming an extended positively charged surface. This region is highly conserved across family members and species, suggesting a common GAG-binding mode.
The crystal structure of human LAR Ig1-2 in complex with a synthetic heparin mimic confirms the GAG-binding site location and reveals a conformational plasticity of the C-D loop, as ligand binding triggers an outward movement of residues in the C-D loop following rupture of a salt bridge. The modified topology of the GAG-binding site maintains an overall positive charge, suggesting that the combination of basic side chains used by the GAG-binding site may vary to accommodate chemically diverse GAGs.
The dimensions of the proteoglycan-binding surface from the chicken RPTPσ Ig1-2 crystal structure suggest that GAG chains may assemble RPTPσ oligomers. Indeed, heparin fragments comprising eight or more saccharide units and heparan sulfate (which consists of 30–150 saccharide units) induce RPTPσ oligomerization. In contrast, comparable chondroitin sulfate quantities did not induce clustering of any RPTP construct tested. In addition, excess chondroitin sulfate inhibits heparan sulfate–induced RPTPσ clustering, suggesting that the HSPG:CSPG ratio and its effect on receptor clustering may influence neuronal function.
Interestingly, immunofluorescence studies show that HSPGs colocalize with RPTPs in the puncta on sensory neurons in culture, whereas CSPGs localize to the extracellular matrix. This suggests that HSPGs might act in cis on cell surface receptors, whereas CSPGs function in trans. Exogenous addition of HSPG or CSPG shifts the HSPG:CSPG ratio, thereby switching the cellular response. The authors propose a model in which high levels of HSPG promote clustering of RPTPs molecules, causing an uneven distribution of phosphatase activity on the cell surface and the formation of microdomains with high phosphotyrosine levels that would support neuronal extension. Conversely, increasing the CSPG:HSPG ratio shifts the balance away from growth-promoting RPTPσ clusters, resulting in stalled axon growth. According to this scenario, molecules able to promote RPTPσ clustering may prove useful in therapeutic strategies following neuronal injury.
C.H. Coles et al.
Science (31 March 2011). doi:10.1126/science.1200840