Ionic Interactions of the C-Terminal Domain of Apolipoprotein A-I are Responsible for Oligomerization

Abstract

Human apoA-I is an anti-atherogenic protein circulating in plasma. The protein is a single polypeptide of 243 amino acid residues, and is the major protein component of high density lipoprotein. The three-dimensional structure of the protein remains unknown, however, secondary structure analysis revealed that apoA-I is predominately α-helical in nature. The N-terminal (NT) domain is made of a bundle of amphipathic α-helices, while the structure of the C-terminal (CT) domain is less defined. Lipid-free apoA-I exists as a heterogeneous population of oligomers. The CT domain is responsible for oligomerization, and several critical lysine residues that may be responsible for oligomerization have been identified. Site-directed mutagenesis was used to design a series of mutant proteins in which lysine residues of the CT domain were changed into glutamine (K195, K206, K208, K226, K238, and K239). The proteins were over-expressed in Escherichia coli and purified with Ni-affinity and gel filtration chromatography. Circular dichroism analysis showed that the helical content and midpoint of guanidine denaturation was not affected by the amino acid changes. Crosslinking with dimethylsuberimidate and size-exclusion FPLC showed that the extent of oligomerization of apoA-I decreased incrementally with each lysine substitution. The lysines of CT-apoA-I may form ionic bonds with oppositely charged residues in neighboring apoA-I. E234 and E235 are in close proximity to K238 and K239, and may form intermolecular ionic bonds upon self-association. To test this, E234 and E235 were mutated to glutamine. These mutant proteins retained their secondary structure and stability properties were not changed. However, the amount of cross-linking was significantly reduced. This result indicates that both lysine and glutamate residues in CT-apoA-I are critical residues responsible for self-association. Therefore, the glutamine variants will be excellent candidates to obtain the high-resolution structure of the protein.