Synthesis, Chemical Characterization and Multiscale Biological Evaluation of a Dimeric-cRGD Peptide for Targeted Imaging of αVβ3 Integrin Activity

Cyclic peptides containing the Arg-Gly-Asp (RGD) sequence have been shown to specifically bind the angiogenesis biomarker α V β 3 integrin. We report the synthesis, chemical characterization, and biological evaluation of two novel dimeric cyclic RGD-based molecular probes for the targeted imaging of α V β 3 activity (a radiolabeled version, 64Cu-NOTA-PEG4-cRGD2, for PET imaging, and a fluorescent version, FITC-PEG4-cRGD2, for in vitro work). We investigated the performance of this probe at the receptor, cell, organ, and whole-body levels, including its use to detect diabetes associated impairment of ischemia-induced myocardial angiogenesis. Both versions of the probe were found to be stable, demonstrated fast receptor association constants, and showed high specificity for α V β 3 in HUVECs (K d ~ 35 nM). Dynamic PET-CT imaging indicated rapid blood clearance via kidney filtration, and accumulation within α V β 3-positive infarcted myocardium. 64Cu-NOTA-PEG4-cRGD2 demonstrated a favorable biodistribution, slow washout, and excellent performance with respect to the quality of the PET-CT images obtained. Importantly, the ratio of probe uptake in infarcted heart tissue compared to normal tissue was significantly higher in non-diabetic rats than in diabetic ones. Overall, our probes are promising agents for non-invasive quantitative imaging of α V β 3 expression, both in vitro and in vivo.

Within NOTA fragment all carbon resonances were distinguishable, however direct labeling of few protons and carbon signals by N5, N6, N8, N9, N1a, N7b and N1b, N4b, N7b was not possible. Also, due to the compound symmetry, the two identical cRGD fragments cannot be differentiated; however, their integration in comparison to well separated H-E3 and H-E4 protons could serve as proof of their presence. The amino acid sequence was established based on the NOE's NH/CHα and the H/C NH/C=O/CHα correlation.

Ligand-Receptor Kinetics Studies
All ligand-receptor kinetics studies were performed with surface plasmon resonance (SPR) using the BIAcore 3000 instrument (Biacore International AB, Uppsala, Sweden).

Protein immobilization
Pre-concentration studies were performed to determine optimal pH conditions for protein immobilization. This helped to ensure that the target levels of immobilized protein could be achieved precisely while also enabling us to conserve materials. Receptor solutions were prepared at 20 µg/mL with acetate buffer (10 mmol, pH ranging from 0.5 to 3.5, at -1.0 below the protein's isoelectric point, Table 1). Each receptor solution (20 µL) was injected at a flow rate of 5 µL/min, followed by a 5 µL injection of ethanolamine-HCL (GE Healthcare AB, Uppsala, Sweden) to clear out the binding surface of the receptor. The optimal pH of acetate buffer for each protein was selected according to the maximum level of protein immobilization reached in the pre-concentration study sensograms.

Ligand-receptor kinetics measurements
The kinetic rate constants were determined by performing global kinetic analysis on the binding curves for each ligand-receptor pair with the BIAevaluation software (Version 4.1.1, GE Healthcare). In global kinetic analysis, nonlinear least squares was used to determine the association and dissociation constants that produced the best fit for multiple response curves, simultaneously. Global analysis is considered to produce the most accurate results in comparison to fitting of a single response curve 2 . Raw sensogram were processed for a given ligand-receptor pair. Briefly, each raw response curve was aligned and subtracted from the reference cell signal of each receptor signal to remove the effects of non-specific interactions. Once processed, a 1:1 Langmuir binding model (Equation 1) was fit to the data, and the association (k on ) and dissociation (k o f f ) kinetic constants, as well as the goodness-of-fit parameter (χ 2 ) and the peak magnitude of the signal response (R max ) were determined. This analysis was applied to each set of kinetic studies, except where the subtracted association phase was negative. A negative association curve occurs when the non-specific binding events on the reference cell exceed the binding events observed for the actual ligand-receptor pair; hence, it is indicative of a non-interaction.
Next, the χ 2 -to-R max ratios were calculated which provide a well-established heuristic to determine the validity of obtained parameters for each ligand-receptor kinetic study. Previous studies have suggested that (χ 2 ) value obtained by fitting kinetic constants globally should be at least 10% of, or lower than, the signal's R max (the change in signal from the highest level to the baseline) 3,4 .