Clustering of GPI-anchored folate receptor independent of both cross-linking and association with caveolin.

Abstract

The distribution of the glycosyl-phosphatidylinositol (GPI)-anchored folate receptor (FR) in a diffuse pattern vs. functional clusters associated with caveolae has been debated. The equivocal nature of direct localization studies is due to possible experimental artifacts such as cross-linking of the protein by the antibody probes prior to fixation and alternatively the use of a disruptive fixation method. Such studies have also been complicated by the use of cells that vastly overexpress FR. In this study a monovalent probe, i.e., a biotinylated folate affinity analogue was used to covalently label FR. Cells expressing moderate levels of FR, i.e., JAR epithelial cells expressing FR-alpha and recombinant CHO fibroblasts expressing FR-beta, were used. The affinity label and either caveolin or antigenic sites on FR were localized by electron microscopy using colloidal gold conjugated antibody probes post-embedding in the relatively permeable LR White resin. The method avoided both receptor cross-linking and early fixation steps and also enabled the use of transport permissive conditions while labeling FR at the cell surface. The results indicate that in steady-state FR is not significantly colocalized with caveolin. However, the receptor molecules occur predominantly in clusters, independent of cross-linking, providing a physical basis for the observed kinetics of receptor internalization and recycling during folate transport. Evidence is also presented to suggest that early mild fixation will disrupt the clustering of FR.