Evidence for serum-deprivation-induced co-release of FGF-1 and S100A13 from astrocytes

Since fibroblast growth factor (FGF)-1 lacks conventional amino-terminal signal peptide essential for endoplasmic reticulum (ER)-Golgi pathway, the mode of release of this polypeptide remains to be fully understood. We attempted to characterize the non-classical (non-vesicular) mode of FGF-1 release in the analyses using immunocytochemistry and immunoblot of conditioned medium (CM) from astrocytes. FGF-1 was completely released from astrocytes upon serum-deprivation stress in a Brefeldin A-insensitive manner. In the immunoprecipitation study using anti-FGF-1 IgG, S100A13 was identified to be the major protein co-eluted with FGF-1. The interaction between GST-FGF-1 and Strep-tag II-S100A13 was found to be Ca2+-sensitive, and to require the C-terminal 11 amino acid peptide sequence of S100A13. The overexpression of Δ88–98 mutant of S100A13 selectively inhibited the serum-deprivation stress-induced release of FGF-1, but not the release of S100A13 mutant from C6 glioma cells. However, amlexanox, anti-allergic drug whose target is S100A13, completely inhibited the stress-induced release of FGF-1 as well as S100A13. The stress-induced release of both proteins was also abolished by BAPTA-AM, an intracellular Ca2+ chelating agent. The serum-deprivation caused Ca2+ spikes in ω-conotoxin GVIA and thapsigargin-sensitive manner. All these results suggest that S100A13 is a cargo molecule for the serum-deprivation stress-induced non-classical release of FGF-1, and that its driving force of protein–protein interaction and release is possibly mediated by Ca2+-induced Ca2+ release (CICR) coupled to N-type Ca2+ channel activity.