Rapid turnover and impaired cell-surface expression of the human folate receptor in mouse L(tk-) fibroblasts, a cell line defective in glycosylphosphatidylinositol tail synthesis.

The human folate receptor (hFR) is a plasma membrane protein that is anchored to the membrane via a glycosylphosphatidylinositol (GPI) tail in some cell types. The KB hFR cDNA sequence predicts a hydrophobic, alpha-helical 31-residue carboxyl terminus that is thought to be the signal for cleavage and attachment of the GPI tail. Alternatively, this region may serve as a transmembrane domain if GPI attachment is not efficient. In this study, we investigated the latter possibility by expressing the hFR in L(tk-) cells, cells that are unable to synthesize GPI tails for attachment to membrane proteins. We also transfected the same hFR cDNA into Chinese hamster ovary (CHO) cells, cells that can anchor proteins by either a GPI tail or a transmembrane domain. Neither parental cell line expresses detectable levels of folate receptor as determined by folic acid binding assays, Western analysis, or Northern analysis. In L(tk-) cells, we found that the recombinant hFR is not expressed on the cell surface, but is rapidly degraded (t1/2 < or = 4 h). Most (> 95%) of the recombinant hFR remains Endo H sensitive, suggesting retention in the endoplasmic reticulum. In contrast, transfected CHO cells express functional hFR protein at the cell surface, the half-life of the hFR is long (t1/2 > or = 24 h), and the Endo H glycosylation pattern of the recombinant hFR is consistent with normal processing through the Golgi apparatus. Therefore, in the absence of a GPI tail, the hFR is not sorted to the cell surface and the incompletely processed hFR protein is unstable.