In 1972, Futterman and Heller, using fluorescence measurements, first reported that bovine b-lactoglobulin, like RBP, formed water-soluble complexes with retinol. The degree of fluorescence enhancement of retinol bound to b-lactoglobulin was about half that of retinol bound to RBP, indicating that retinol was bound with greater affinity to RBP than to bovine b-lactoglobulin. Using fluorescence titration and circular dichroism, b-lactoglobulin has been found to display two high affinity binding sites for retinol per protein dimer, each with an association constant of 2 x 108 M-1. This value is in good agreement with that determined by continuous elution analytical affinity chromatography. However, using equilibrium dialysis and radiolabel retinol, the number of binding sites is in accordance with that previously determined, but the association constant appears to be four orders of magnitude less (about 1.5 x 104 M-1). Gel filtration or reversed-phase chromatography of retinol incubated with b-lactoglobulin has shown that the b-lactoglobulin peak contains only a small amount of ligand but most of the retinol is eluted as free retinol, indicating that b-lactoglobulin binds retinol with a relatively low affinity.

The binding of retinol to b-lactoglobulin mainly involves hydrophobic interactions. Furthermore, an electrostatic contribution to the interaction has also been suggested because of the dependence of the affinity on pH and ionic strength.

Chemical modification of b-lactoglobulin by methods, such us methylation, ethylation, esterification, or alkylation enhances the binding affinity for retinol by opening up a second binding site and increasing the overall hydrophobicity of the protein.

The existence of a specific binding site for retinol in b-lactoglobulin has been proposed, but the location and nature of this site are still unclear. Two independently performed crystallographic analysis of b-lactoglobulin have led to different hypothesis about the location of the putative retinol-binding pocket. Model building studies have proposed that retinol, as in RBP, is bound to b-lactoglobulin in a deep pocket within the central calyx. However, an alternative proposition support the idea that retinol would be bound in a hydrophobic pocket located on the surface of b-lactoglobulin.

The binding of retinol by b-lactoglobulin appears different from RBP on the basis of three types of experimental results. First, retinol bound to b-lactoglobulin is displaced from the complex by another retinoid whereas nothing analogous to this has been described for RBP. Second, a higher rate of oxidation of retinol by liver alcohol dehydrogenase was observed when it was presented bound to b-lactoglobulin compared to RBP. Third, rotational relaxation measurements suggested that retinol binds in a region of the b-lactoglobulin molecule that is flexible with respect to the rest of the protein.

Monaco et al. (1987) interpreted these data to razionalizate their observed surface binding pocket. It is possible that this model represents an artefactual model of binding due to the retinol-b-lactoglobulin complex shows a rather weak electron density assigned to the retinol molecules, possibly as a result of low occupancy.

The binding of retinol to b-lactoglobulin in vitro, as well as its similarity to RBP, has led to speculation that b-lactoglobulin may participate in the transport of retinol in the intestine of the newborn. However, despite the indirect experimental support for in vitro interaction, the presence of retinol bound to b-lactoglobulin in milk has not so far been detected. Furthermore, the presence of a receptor for b-lactoglobulin-like proteins at the brush border membrane of the rat enterocyte has been suggested. However, the presence of a receptor for this protein would be unusual, because rat milk does not contain b-lactoglobulin.

References .

Return to first page of review

Return to Milk Science page