Immunoreactive folate-binding proteins from human saliva. Isolation and comparison of two distinct species.

Human saliva contains a single 72,000-M(r) species which specifically reacted with rabbit anti-[human placental folate receptor (PFR)] serum on SDS/PAGE and Western blots. Although a specific radioimmunoassay for human PFR and related folate-binding proteins (FBPs) identified 55 ng of cross-reacting material (CRM) per mg of crude salivary proteins, only a minor fraction (1.6 ng) specifically bound radiolabelled folate. The major fraction of CRM did not contain bound endogenous folate and did not bind radiolabelled folates. On the basis of folate binding, salivary CRM species to PFR were designated as either functional (f-FBP) or non-functional (nf-FBP) species respectively. nf-FBPs and f-FBPs were isolated by different purification schemes. Both purified f-FBPs and nf-FBPs migrated as a single apparent 72,000-M(r) species on SDS/PAGE, but on Sephacryl S-200 gel filtration and sucrose-density-gradient ultracentrifugation they were eluted/sedimented with 40,000-M(r) markers. Each microgram of purified f-FBP and nf-FBP was measured in the radioimmunoassay for PFR as being equivalent to 18 ng and 24 ng of CRM respectively, indicating low epitope-relatedness to PFR. The Kd of f-FBPs was 50 pM and 0.94 mol of folate was bound/mol of protein. f-FBPs exhibited an unusual dependence on Triton X-100 for optimal ligand binding, despite the fact that Triton X-100 micelle binding was not demonstrated. The relative order of affinity of f-FBPs for pteroylglutamate greater than methotrexate greater than 5-formyltetrahydrofolate greater than 5-methyltetrahydrofolate was also distinct from that of purified PFR. Whereas amino acid and carbohydrate analysis revealed that nf-FBP (M(r) 51,400) and f-FBP (M(r) 39,200) were distinct glycoproteins with 8 and 13% carbohydrate respectively, isoelectric focusing and immunological studies suggested some structural identity. The presence of f-FBP and nf-FBP in normal saliva raises new questions about their possible role in vivo.