Role of portal region lysine residues in electrostatic interactions between heart fatty acid binding protein and phospholipid membranes.

The structure of heart fatty acid binding protein (HFABP) is a flattened beta-barrel comprising 10 antiparallel beta-sheets capped by two alpha-helical segments. The helical cap region is hypothesized to behave as a portal "lid" for the entry and release of ligand from the binding pocket. The transfer of fatty acid from HFABP is thought to occur via effective collisional interactions with membranes, and these interactions are enhanced when transfer is to membranes of net negative charge, thus implying that specific basic residues on the surface of HFABP may govern the transfer process [Wootan, M. G., & Storch, J. (1994) J. Biol. Chem. 269, 10517-10523]. To directly examine the role of charged lysine residues on the HFABP surface in specific interactions with membranes, chemical modification and selective mutagenesis of HFABP were used. All surface lysine residues were neutralized by acetylation of recombinant HFABP with acetic anhydride. In addition, seven mutant HFABPs were generated that resulted in charge alterations in five distinct sites of HFABP. Modification of the protein did not significantly alter the structural or ligand binding properties of HFABP, as assessed by circular dichroism, fluorescence quantum yield, and ligand binding analyses. By using a resonance energy transfer assay, transfer of 2-(9-anthroyloxy)palmitate (2AP) from acetylated HFABP to membranes was significantly slower than transfer from native HFABP. In addition, in distinct contrast to transfer from native protein, the 2AP transfer rate from acetylated HFABP was not increased to acceptor membranes of increased negative charge. Transfer of 2AP from HFABP mutants involving K22, located on alpha-helix I (alpha-I) of the helical cap region, was 3-fold slower than transfer from wild-type protein, whereas rates from a mutant involving the K59 residue, located on the beta 2-turn of the barrel near the helical cap, were 2-fold faster than those of wild type. A double mutant involving K22 and K59 resulted in transfer rates identical to those of wild type, indicating that at least two domains are involved in determining the overall rate of ligand transfer. In addition, 2AP transfer rates from HFABP mutated at position 22 were totally unaffected by the charge characteristics of acceptor membranes, in marked contrast to wild type and other members of the mutant series. Further, by introducing a positive charge to alpha-helix II (alpha-II) of the helical cap region, 2AP transfer rates increased by 4-fold and properties of HFABP transfer began to approach those seen for AFABP, another member of the FABP family thought to transfer ligand via collisional interactions with membranes, which has a lysine residue in the alpha-II helix. These studies demonstrate that the helical cap region of HFABP may play an important role in governing ionic interactions between binding protein and membranes.