Atherosclerosis is caused by chronic inflammation of the arterial wall leading to serious complications such as myocardial infarction and stroke. Inflammation is driven by recruitment of mononuclear cells to the vascular wall, a process that is exacerbated by platelet-dependent deposition of chemokines CCL5 (RANTES) and CXCL4 (platelet factor 4: PF4) on the inflamed endothelium. Recently it was found that heterodimer formation of RANTES and PF4 promotes monocyte recruitment to the vascular wall and might thus be a very suitable clinical target.
In order to study the mode of action of the RANTES-PF4 heterodimer bridging the monocyte-endothelium interactions, synthetic access to a homogenous covalently linked RANTES-PF4 molecule was needed. Molecular dynamics simulations and HSQC chemical shift mapping show that non covalent RANTES-PF4 interactions are occurring in the N-termini of the chemokines1. However, as RANTES interacts with its receptor via the N-terminus it was hypothesized that a coupling of the two termini would not lead to an active dimer. Based on the molecular dynamics, positions for covalent bond formation between the two proteins were identified that will not influence chemokine activity. The covalent linkage will be made with an oxime bond using novel oxime formation chemistry. The usual introduction of a ketone functionality through levulinic acid led to low yields in a model setup. It was found that cyclization leads to a side product that is unable to react with the aminooxy modified protein. A mechanism for cyclization of levulinic acid was proposed based on NMR measurements. The oxime formation was optimized using alternative keto-acids; kinetics of the oxime formation were characterized. Both PF4 and RANTES were synthesized in three parts and ligated using native chemical ligation. PF4 and RANTES were modified with ketone and aminooxy moieties, respectively. Results of the heterodimer synthesis with optimized oxime chemistry will be presented.