An approach to molecular imaging of atherosclerosis, thrombosis, and vascular inflammation using microparticles of iron oxide

Abstract

The rapidly evolving field of molecular imaging promises important advances in the diagnosis, characterization and pharmacological treatment of vascular disease. Magnetic resonance imaging (MRI) provides a modality that is well suited to vascular imaging as it can provide anatomical, structural and functional data on the arterial wall. Its capabilities are further enhanced by the use of a range of increasingly sophisticated contrast agents that target specific molecules, cells and biological processes. This article will discuss one such approach, using microparticles of iron oxide (MPIO). MPIO have been shown to create highly conspicuous contrast effects on T(2)(*)-weighted MR images. We have developed a range of novel ligand-conjugated MPIO for molecular MRI of endothelial adhesion molecules, such as vascular cell adhesion molecule-1 (VCAM-1) and P-selectin expressed in vascular inflammation, as well as activated platelet thrombosis. This review discusses the application of ligand-targeted MPIO for in vivo molecular MRI in a diverse range of vascular disease models including acute vascular inflammation, atherosclerosis, thrombosis, ischemia-reperfusion injury and ischemic stroke. The exceptionally conspicuous contrast effects of ligand-conjugated MPIO provide a versatile and sensitive tool for quantitative vascular molecular imaging that could refine diagnosis and measure response to treatment. The potential for clinical translation of this new class of molecular contrast agent for clinical imaging of vascular syndromes is discussed.

Fig. 1

MRI relies on the delivery of relatively high payloads of either (A) gadolinium chelates or (B) iron oxide. Gd chelates decorate the surface of the carriage vehicle since Gd requires interaction with local water molecules to produce contrast effects. Iron oxide particles (size range, ∼10 nm to 5 μm) are typically contained within polymer shells. USPIO: ultrasmall particles of iron oxide; CLIO: cross-linked iron oxide nanoparticles; MPIO: microparticles of iron oxide.

Fig. 2

Imaging of VCAM-1 in acute inflammation. (A) Confocal microscopy of TNF-α stimulated sEND-1 cells. Green fluorescence reflects VCAM-1 expression on the cell surface. Prior incubation of VCAM-MPIO with Fc-ICAM-1 had no effect on VCAM-MPIO binding (autofluorescent green spheres), whereas pre-incubation with Fc-VCAM-1 abolished VCAM-MPIO retention, despite demonstrable VCAM-1 surface expression. Graph depicts retained VCAM-MPIO (mean ± S.D.) with and without pre-incubation with soluble Fc-VCAM-1 or Fc-ICAM-1 (^*P < 0.0001). (B) In vivo T₂^*-weighted MR coronal images (4 images per brain; resolution ∼90 μm³). Intense low signal areas (highlighted with red box) on the left side of the brain reflect specific MPIO retention (VCAM-MPIO (row 1) VCAM + P-selectin MPIO (row 2)) on acutely activated vascular endothelium with almost absent contrast effect in the contralateral hemisphere (green box). No contrast effects were observed with IgG-MPIO control (row 3) or pre-treatment with VCAM-1 antibody prior to VCAM-MPIO administration, which effectively blocked VCAM-MPIO binding (row 4). Scale bar, 5 mm. (C) Three-dimensional volumetric maps of VCAM-MPIO contrast effects (red) delineate the architecture of cerebral vasculature in the IL-1β-stimulated hemisphere (left half of top image) with almost total absence of binding on the contralateral, non-activated side. The midlines are indicated by vertical sections. Pre-administration of VCAM-1 antibody abolished VCAM-MPIO retention (lower image). Quantitative analyses of MPIO contrast effects found that specific VCAM-MPIO contrast was increased >100-fold, compared with brains without IL-1β injection. Dual-targeted VCAM + P-selectin MPIO also bound specifically but did not further enhance contrast effects. Substitution of IgG-MPIO (IgG/IL-1β⁺), sham intracerebral injection (VCAM/NaCl), no intracerebral injection (VCAM/IL-1β⁻) and pre-blocking (VCAM/IL-1β⁺ with block) were not associated with specific contrast effects. Bars indicate mean values for each group (^*P = 0.02) .

Fig. 3

Dual-targeted MPIO binding in mouse atherosclerosis. (A) Dual-ligand MPIO recognizing VCAM-1 and P-selectin showed 7-fold enhanced binding to aortic root plaque endothelium compared to single-ligand MPIO targeting either VCAM-1 or P-selectin, following left ventricular injection. **P < 0.01; *P < 0.05. (B) Dense dual-targeted MPIO binding to endothelium overlying atherosclerotic plaque. Scale bar, 20 μm. (C) Ex vivo MRI of aortic roots 30 min after i.v. injection of MPIO. Dual-targeted MPIO binding appeared as distinct circular low signal areas adherent to endothelium overlying atherosclerotic plaque. Minimal contrast effects were observed with negative isotype IgG-MPIO. Scale bar, 500 μm. (D) 3D reconstruction of dual-targeted MPIO contrast effects through the aortic root .

Fig. 4

MRI of activated platelets using MPIO conjugated to single-chain antibodies directed against ligand-induced binding sites (LIBS). (A) (i) Ex vivo MRI of a wire-injured femoral artery exposed to LIBS-MPIO shows multiple, intensely low signal, lobulated areas at the interface between vessel wall and lumen (arrows). (ii) Three-dimensional reconstruction shows diffuse and relatively even LIBS-MPIO binding along the luminal surface of the injured femoral artery. (iii) Co-localization of LIBS-MPIO and platelets was confirmed using immunohistochemistry for CD62 . (B) Confocal microscopy of human platelets immobilized on fibrinogen and detected by immunofluorescence using CD62 antibody (green) . (i) LIBS-MPIO (red) show specific binding to platelets. (ii) No binding was observed with control-MPIO. (iii) 3D rendering shows multiple LIBS-MPIO binding to clusters of activated platelets, via GPIIb/IIIa. (C) In vivo T₂^*-weighted MRI after carotid artery injury . Transverse sections demonstrate the injured right carotid artery (red circle), and the non-injured left carotid artery (green circle). Following LIBS-MPIO injection, there is increasing signal drop at 12, 24, and 72 min compared with preinjection and the non-injured left carotid artery. For control-MPIO, signal intensity is similar at 12, 48, and 72 min in both vessels. (D) Immunohistochemistry of wall-adherent thrombus in a LIBS-MPIO-injected animal. In the inset, arrows depict bound MPIO on the thrombus surface (thrombus area itself appears red). Quantification of MPIO bound to wall-adherent thrombosis shows significantly higher LIBS-MPIO binding compared to control-MPIO or to mice treated with human urokinase or mouse recombinant urokinase to induce thrombolysis, prior to LIBS-MPIO administration .

Fig. 5

(A) Histological section depicting the considerably smaller size of MPIO (1 μm diameter) (black arrow) bound to endothelium overlying aortic root plaque, compared to adjacent red blood cells (red arrows). (B) In vivo biodistribution studies show that MPIO retention by the lungs is minimal, while MPIO uptake by the spleen and liver is rapid . (C) Schematic representation of the biological handling properties of biodegradable MPIO. Efficient removal of contrast agent from the circulation via the reticulo-endothelial system is advantageous for in vivo imaging, where background blood MPIO may otherwise obscure specifically bound contrast. Disintegration and dispersal of the dextran coat and iron via normal iron handling pathways is required.