Those proteins that bind folate with high affinity are referred to as folate binding proteins (FBPs) and are displayed on cell surfaces where they function as a receptor for transport of folate across cell membranes. They are also found in a free, soluble form in a number of biological fluids, including milk.
Unusual physical characteristics are described for bovine milk FBP; namely (i) its affinity for ligand appears to decrease with increased protein concentration, without a change in ligand capacity, and (ii) the binding of ligand appears to be co-operative when tested by Scatchard analysis (e.g. Salter et al 1981; Hansen et al 1983). Since bovine milk FBP is a single subunit glycoprotein of 222 amino acid residues and MW 27,500 for the peptide alone, there is no clear physical basis for these characteristics. However, they are most apparent at pH values above 6, conditions that allow polymerisation of FBP; hence these characteristics have been attributed to FBP polymerisation. This thesis re-examines the polymerisation and binding properties of bovine milk FBP to test this putative relationship.
FBP was purified from whey protein concentrate by affinity chromatography. Analysis of the preparation by SDS-PAGE revealed a mixture of proteins, of which two displayed folate-binding activity, comprised 65 and 25% of the Coomassie blue dye-binding activity of the preparation and separated, respectively, at 30kDa and 33kDa. The remaining 10% of dye-binding activity was comprised of three contaminating proteins that lacked folate-binding activity and had subunit MW of 80kDa, 55kDa and 23kDa.
Polymerisation was investigated at 22°C by analytical molecular sieve chromatography at three pH values, 5.0, 6.8 and 9.0, and FBP concentrations ranging from 0.5mg/mL to 50mg/mL. Analytical sedimentation equilibrium ultracentrifugation at 20°C was carried out at pH 5.0 and 6.8 and FBP concentrations in the range 0.2 to 7.5 mg/mL. Visible aggregation was compared at 20°C and at 12°C. At pH 6.8 and 9.0 FBP readily polymerised. The tendency to polymerization was greater with greater FBP concentration, higher temperature and the presence of ligand and was greater at pH 9.0 than at pH 6.8. At pH 5.0 FBP remained monomeric at concentrations ranging from 0.5 to 50 mg/.mL, both in the absence and presence of ligand, and at temperatures of 20ºC and 72ºC.
Ligand binding to FBP was measured by incubation of tritiated folic acid with FBP followed by separation of free folic acid from bound, by centrifugal ultrafiltration. FBP concentrations were 0.66 and 77 µM and binding was measured at pH 5.0, 7.3 and 9.0. Binding functions were fitted to untransformed data sets without weighting and the goodness of each fit was compared by the F test.
At each examined pH value and FBP concentration, the cooperative binding model fitted the data significantly better than the one-site saturation model. While the cooperative model appeared to provide an adequate description of the data, a significantly better fit was provided by the stoichiometric binding model, in which there is quantitative formation of the folate-FBP complex at each point along the titration curve due to the experimental use of concentrations of folic acid and FBP orders of magnitude above the K^ value. To test binding to lower FBP concentrations, data published originally as Scatchard plots was re-analyzed by the methods adopted here, namely binding data for 100 µM FBP at pH 5.1 and 7.2 (Salter et al, 1981) and for 12 µM FBP at pH 7.4 (Hansen et al., 1983). Again each data set was described adequately by the cooperative binding model but significantly better by the stoichiometric binding model.
Adoption of either the one-site saturation model or the cooperative binding model allowed the derivation of apparent dissociation constants for the FBP-ligand complex, and these "constants" indicated the apparent weakening of affinity as the FBP concentration increased, as previously concluded. Whereas changing ligand affinity with changing FBP concentration is possible at pH values consistent with FBP polymerisation, such behaviour is clearly impossible at pH 5 at which FBP remains monomeric. Co-operative binding is similarly impossible at pH 5. The more appropriate interpretation of the binding data as examples of stoichiometric binding removes these apparent difficulties, along with the notion that binding is cooperative, at all tested pH values.