Structure and ligand binding properties of human serum albumin

1.

Serum albumin possesses a unique capability to bind, covalently or reversibly, a great number of various endogenous and exogenous compounds. Several different transport proteins exist in blood plasma but albumin only is able to bind a wide diversity of ligands reversibly with high affinity. The subject of the present thesis is these binding properties. In 1981 the author proposed in a review a model for binding of ligands to serum albumin. In the model, binding of ligands to at least 6 distinct regions was considered. The purpose of the experimental work described here was to test the validity of the model. This was done by performing a large number of competition experiments. From these new data a revised model for ligand binding is presented. 2.

Human serum albumin consists of 585 amino acids forming a single polypeptide of known sequence. A number of well characterized genetic variants have been reported. The physico-chemical characteristics of the protein are well-established. By contrast, the complete secondary and tertiary structures are not known; information about major structural features only has been obtained. The albumin molecule seems to have an overall ellipsoidal shape (about 140 x 40 A) and to be composed of domains. On the basis of the amino acid sequence, Brown (1977a) proposed a 3-domain model for the protein. Each domain is believed to consist of 6 helices forming a hydrophobic channel with basic and hydrophobic amino acid residues placed at the ends. Experimental data, however, indicate that the domains cannot be identical. Long-chain fatty acid ions are proposed to bind with high affinity within the channels. The ability to fluctuate between isomeric forms in aqueous solution could assist in adapting the albumin molecule to bind ligands with a diverse nature with high affinity. This possibility is discussed on the basis of several physico-chemical techniques including hydrogen-deuterium exchanges. Also the importance of the N-B transition for ligand binding is considered. 3.

Single binding of ligands to serum albumin is usually described as high-affinity binding to one or two sites and weaker binding to a larger number of sites. In this chapter, the original binding model for high-affinity binding is elaborated. Region 1 seems to be specific for binding of one, or possibly two, ions of long-chain fatty acids. Region 2 is somewhat less specific and includes binding of octanoate, tryptophan, chlorazepate, thyroxine, p-iodobenzoate and possibly also chloride. Region 3 accommodates bilirubin, Phenol Red, Bromophenol Blue and iopanoate. Region 4 is a special site for strong binding of metal ions such as Cu++ and Ni++. Evidence is presented for placing the primary haemin site in a separate region (no. 5). The existence of additional binding regions, well-suited for high-affinity binding of drugs, is discussed...