Covalent conjugates of the cross-linked iron oxide nanoparticles (CLIO) and high-affinity (K(d)(app) = 8.5 nM) anti-human E-selectin (CD62E) F(ab')(2) fragments were prepared and tested in vitro to establish feasibility of endothelial proinflammatory marker magnetic resonance (MR) imaging. The conjugates were obtained by using thiol-disulfide exchange reaction between 3-(2-pyridyl)propionyl-CLIO and S-acetylthioacetate-modified F(ab')(2) fragments. The purified CLIO-F(ab')(2) conjugates (average hydrodynamic diameter 40.6 nm) were used in experiments with the live human endothelial umbilical vein cells (HUVEC). Cells treated with IL-1 beta expressed E-selectin and showed a 100-200 times higher binding of CLIO particles (83-104 ng iron/million cells) than control cells. The binding resulted in a high superparamagnetism of HUVEC with the transverse water proton relaxation time (T2) decrease to 30-40 ms in cell precipitates. Cells did not bind/internalize CLIO-F(ab')(2) conjugates prepared using a control fragment or nonconjugated iron oxide particles before or after treatment with IL-1 beta. MR imaging of cells showed a highly specific T2-weighted signal darkening associated with cells treated with IL-1 beta and incubated with anti-E selectin. Demonstration of MR imaging of E-selectin expression justifies further development of MR-targeted agents for monitoring tumor vascular endothelial proliferation, angiogenesis, and atherosclerosis.

Measurements of tumoral VVF at high-resolution MR imaging with long-circulating iron oxide are feasible and correlate with angiogenic burden in experimental tumor models.

A rapid and accurate assessment of the antitumor efficacy of new therapeutic drugs could speed up drug discovery and improve clinical decision making. Based on the hypothesis that most effective antitumor agents induce apoptosis, we developed a near-infrared fluorescent (NIRF) annexin V to be used for optical sensing of tumor environments. To demonstrate probe specificity, we developed both an active (i.e., apoptosis-recognizing) and an inactive form of annexin V with very similar properties (to account for nonspecific tumor accumulation), and tested the agents in nude mice each bearing a cyclophosphamide (CPA) chemosensitive (LLC) and a chemoresistant LLC (CR-LLC). After injection with active annexin V, the tumor-annexin V ratio (TAR; tumor NIRF/background NIRF) for untreated mice was 1.22+/-0.34 for LLC and 1.43+/-0.53 for CR-LLC (n=4). The LLC of CPA-treated mice had significant elevations of TAR (2.56+/-0.29, P=.001, n=4), but only a moderate increase was obtained for the CR-LLC (TAR=1.89+/-0.19, P=.183). The in vivo measurements correlated well with terminal deoxyribosyl transferase-mediated dUTP nick end labeling indexes. When inactive Cy-annexin V was used, with or without CPA treatment and in both CCL and CR-CCL tumors, tumor NIRF values ranged from 0.91 to 1.17 (i.e., tumor were equal to background). We conclude that active Cy-annexin V and surface reflectance fluorescence imaging provide a nonradioactive, semiquantitative method of determining chemosensitivity in LLC xenografts. The method maybe used to image pharmacologic responses in other animal models and, potentially, may permit the clinical imaging of apoptosis with noninvasive or minimally invasive instrumentation.