Targeted ultrasound contrast agent for molecular imaging of inflammation in high-shear flow

Targeted ultrasound contrast materials (gas-filled microbubbles carrying ligands to endothelial selectins or integrins) have been investigated as potential molecular imaging agents. Such microbubbles normally exhibit good targeting capability at the slower flow conditions. However, in the conditions of vigorous flow, binding may be limited. Here, we describe a microbubble capable of efficient binding to targets both in slow and fast flow (exceeding 4 dyne/cm(2) wall shear stress) using a clustered polymeric form of the fast-binding selectin ligand sialyl Lewis(X). Microbubbles were prepared from decafluorobutane gas and stabilized with a monolayer of phosphatidylcholine, PEG stearate and biotin-PEG-lipid. Biotinylated PSLe(x) (sialyl Lewis(X) polyacrylamide) or biotinylated anti-P-selectin antibody (RB40.34) was attached to microbubbles via a streptavidin bridge. In a parallel plate flow chamber targeted adhesion model, PSLe(x) bubbles demonstrated specific adhesion, retention and slow rolling on P-selectin-coated plates. Efficiency of firm targeted adhesion to a P-selectin surface (140 molecules/microm(2)) was comparable for antibody-carrying bubbles and PSLe(x)-targeted bubbles at 0.68 dyne/cm(2) shear stress. At fast flow (4.45 dyne/cm(2)), PSLe(x)-targeted bubbles maintained their ability to bind, while antibody-mediated targeting dropped more than 20-fold. At lower surface density of P-selectin (7 molecules/microm(2)), targeting via PSLe(x) was more efficient than via antibody under all the flow conditions tested. Negative control casein-coated plates did not retain bubbles in the range of flow conditions studied. To confirm echogenicity, targeted PSLe(x)-bubbles were visualized on P-selectin-coated polystyrene plates by ultrasound imaging with a clinical scanner operated in pulse inversion mode; control plates lacking targeted bubbles did not show significant acoustic backscatter. In vivo, in a murine model of inflammation in the femoral vein setting, targeting efficacy of intravenously administered PSLe(x)-microbubbles was comparable with targeting mediated by anti-P-selectin antibody, and significantly exceeded the accumulation of non-targeted control bubbles. In the inflamed femoral artery setting, PSLe(x)-mediated microbubble targeting was superior to antibody-mediated targeting.