Folate Receptor

Receptor-Mediated Folate Transport: An Example of Potocytosis

Studying cells grown in low levels of folate (<1 nM), several laboratories identified a folate receptor (FR) on the membrane of KB (human nasopharyngeal carcinoma) cells, CACO2 (colon carcinoma) cells, IGROV-1 (ovarian carcinoma), MA104 (monkey kidney epithelial) cells, and, more recently, in both murine and human leukemia cells (35-39: reviewed in refs. 19 and 40). A soluble form of the protein was initially found in milk 30 yr ago and in some chronic myelogenous leukemia cell extracts more than 25 yr ago (in early history is reviewed in ref. 40).

We proposed that receptor-coupled folate transport occurs in discrete areas of the cell membrane called caveolae (20). Although the anatomic clustering is the subject of much debate (41-43), the kinetics of 5-methyl[3H] tetrahydrofolate in MA104 cells are suggestive of a coupled or channeled process (Fig. 1). First, the ligand binds to the surface via an "externalized" FR. This folate is removable with an acid/saline wash; second, the receptor/ ligand complex is sequestered into an acid-resistant but still membrane-bound compartment; third, the compartment is acidified and folate dissociates from the receptor and is transferred to the cytoplasm; fourth, the receptor returns to the cell surface (acid labile fraction) to begin another round of folate binding and internalization.

Molecular Biology/Biochemistry of FR

Several laboratories have reported the isolation of cDNA clones for FR (44-47). The cDNA from CACO2, KB, and human placenta are identical and there is only one difference near the amino terminus in the ovarian carcinoma cDNA. The predicted amino acid sequence of the membrane form is >99% identical to the actual analysis of the soluble receptor from milk. This correlation at the nucleic acid level and the relationship of the membrane and soluble form assessed by [35S]-labeling KB cells in vitro (48) supports


Compounds under each step effect FR movement, folate binding or release

Fig. 1. Receptor-coupled transport of folate (potocytosis).

Compounds under each step effect FR movement, folate binding or release

Fig. 1. Receptor-coupled transport of folate (potocytosis).

the hypothesis that the soluble form of the receptor is a processed product of the membrane form, specifically that the source of the soluble folate-bind-ing protein is FR that was released from the cell membrane. In this regard, we showed that the cDNA sequence correctly predicted that the receptor would be coupled to the membrane via a glycosyl-phosphatidylinositol linkage (GPI) by releasing the receptor with phospholipase C specific for a GPI linkage (45). Others have confirmed this through direct analysis and enzyme release in both KB cells and fresh tissue samples (49,50). There is also some evidence that the receptor may be released through the action of a Mg2+-dependent protease. In addition, Ratnam and colleagues found a second clone in a human placental library (51) and a third species predominating in marrow elements (52). The first species, initially referred to as MFP2, is now termed FR-a; the second, also from placenta and originally named MFP1, is FR-P; the third, recently described by molecular techniques, is FR-y. FR-P is about 70% identical to FR-a and maintains a GPI anchor. FRY is also about 70% identical to FR-a and is GPI anchored; however FR-y' is truncated at the carboxy terminus and is a secreted protein. The function of FR-P and FR-y are not known. Preliminary studies in receptor-negative CHO cells transfected with FR-P allowed us to show that membrane binding had no influence on the accumulation of 5-methyl[3H]tetrahydrofolate. Moreover, despite the molecular detection of FR-P using polymerase chain reaction (PCR) technology in a number of different tissues, the total amount of message is very small compared to FR-a and there are no known cell lines that express FR-P naturally.

Recently, it has been suggested that the nomenclature for the whole family may be somewhat misleading. Da Costa and Rothenberg, noting that the molecules are so similar, suggested that tissue identification may be more useful (53). Based on studies of soluble as well as membrane-bound protein in selected overexpressing cell lines as well as transfected cells, we believe that there is some merit to this classification. Moreover, it is our opinion, especially because plasma folate across species lines vary by more than 10fold (e.g., man is 5-15 nM and mouse is 150-250 nM), that even the function of the receptors, which have differing affinities for 5-methyl-tetrahydrofolate, may be different. Regardless, the role and regulatory pathways for the receptor(s) need to be determined. For convenience, we will continue to use the currently accepted terminology as outlined in recent reviews (19,40) and, moreover, unless otherwise noted, the studies presented here are derived exclusively from studies of FR-a unless otherwise indicated. In addition, there are very little data on the fate/function of the soluble protein and our early studies showed that, at least in vitro, it did not deliver bound folate to some cells in a specific manner; therefore, it will not be reviewed here.

0 0

Post a comment