We hypothesize that apoA-I is the major HDL platform and functions as a conformationally dynamic scaffold that facilitates the interaction of a host of other apolipoproteins and lipid remodeling factors. It follows from this hypothesis that a detailed understanding of HDL assembly is not possible through traditional unidisciplinary approaches but requires an integrated multidisciplinary team approach. Our objective in this project is to use that approach to understand, in unprecedented detail, the structural basis for HDL assembly. We will combine in silico approaches with more traditional in-solution experimental approaches: i) high resolution mass spectrometry combined with thiol-cleavable cross-linking chemistry (Core D), ii) site-directed mutagenesis (Core D), and iii) functional studies (Core C). To achieve this objective, we propose three specific aims:
Aim 1 : Three early steps in HDL assembly: i) ApoA-I monomer structure and dynamics. We will experimentally test our models using cross-linking and site-directed mutagenesis (Core D) and assays of function (Core C). ii) ABCA1 structure and dynamics. We will use computational methods to develop the extracellular- and intracellular-facing conformations and the molecular basis for this conformational cycle of ABCA1 and how this cycle produces a phospholipid pump (Core B). Using purified and mammalian cell-derived ABCA1, we will experimentally test the models by, respectively, mass spectroscopy-chemical cross-linking (Cores C and D) and site-directed mutagenesis (Core D) using cell-based assays of HDL function and quantification of HDL particles (Core C). iii) Physical basis for assembly of nascent HDL. We will test the monolayer pleating hypothesis by MD simulations, examining, for example, the effects of lysoPC and FFA on disc-membrane fission and mechanisms of lipidation of dimeric apoA-I (Core B). Mutagenesis will target key Lys residues in the extracellular domain of ABCA1 where pleating and apoA-I association is propose to occur. High resolution EM tomography will be used to examine 3D images of budding discs under various conditions of lipid composition.
Aim 2 : Two middle steps in HDL assembly: i) Simulations of the interactions of LCAT with apoA-I and dHDL (Core B). The models will be experimentally tested using cross-linking and mutagenesis (Heinecke). ii) Simulations of the interactions of apoA-II with apoA-I and HDL. The models will be experimentally tested using cross-linking (Davidson).
Aim 3 : Three later steps in HDL assembly. i) ABCG1 structure and dynamics. The models will be developed using homology modeling with MODELLER, double state normal mode analysis and MD simulations (Core B). ii) Detailed molecular models of PLTP and its interactions with HDL (Davidson). The models will be experimentally tested using homology modeling with MODELLER (Core B), assays of function (Core C), cross- linking (Cores C and D) and high resolution EM tomography. iii) Detailed molecular models of the interactions of PON-1 with HDL. ApoA-I mutations in helical repeat 4 in apoA-I (or other domains identified by Davidson) will be reconstituted as dHDL and PON-1 activity determined.
HDL, the good cholesterol, is an important target for future drugs to prevent heart attacks. Unfortunately, the relationship of HDL to heart disease is extremely complex. The combination of computer and test tube studies of HDL complexity that we propose, a combination unique to this program, provides a molecular blueprint for future drug development aimed at HDL.
