99mTc-labeled chemotactic peptides: influence of coligand on distribution of molecular species and infection imaging properties. Synthesis and structural characterization of model complexes with the {Re(η 2-HNNC 5H 4N)(η 1-NNC 5H 4N)} core

Abstract

99mTc-labeled hydrazino nicotinamide (HYNIC) derivatized chemotactic peptides are useful infection imaging agents. However, the reagent that is used for radiolabeling (‘coligand’) can have profound effects on biodistribution. In this study, the distribution of molecular species formed with a variety of coligands was correlated with infection localization and biodistribution in E. coli infected rabbits. Five 99mTc-coligand complexes (mannitol, glucamine, glucarate, tricine and glucoheptonate) were used to label f-MLFK-HYNIC and radiolabeled species were characterized by reverse phase HPLC. One mCi of each 99mTc-coligand–peptide complex was injected into rabbits 24 h after infection and imaging was performed 3–4 and 16–17 h later. After acquiring the final images, the animals were euthanized and biodistribution was measured. With all five coligands, 99mTc-labeled peptide was obtained in >90% radiochemical purity. Multiple radiolabeled species were obtained with each reagent; however, mannitol yielded the most homogeneous product. Model studies on the robust {Re(η 2-HNNC 5H 4N)(η 1-NNC 5H 4)} core demonstrate that ligands such as mannitol, tricine, tris(hydroxymethyl)propane, glucarate-a, glucarate-b, and glucoheptonate yield intractable mixtures of products as a consequence of the ambidentate nature of these ligands and their facile displacement by other donor groups, including solvent molecules. In contrast, dihydroxylic ligands with an additional imine functionality and a meridional geometric preference yield monophasic materials upon reaction with [ReCl 3(η 1-NNC 5H 4NH)(η 2-HNNC 5H 4N)]. Both [Re{η 3-C 5H 3N-2,6-(CH 2O) 2}(η 1-NNC 5H 4N)(η 2-HNNC 5H 4N)] ( 1) and [Re{η 3-OC 6H 4C(H)NC 6H 4O}(η 1-NNC 5H 4N)(η 2-HNNC 5H 4N)] ( 2) exhibit distorted octahedral geometries with a meridional disposition of the bidentate pyridinediazene and pyridinediazenido(1−) ligands and the three remaining meridional positions occupied by the two oxygen and the nitrogen donors of the tridentate coligand. Image analysis demonstrated that 99mTc-mannitol-f-MLFK-HYNIC produced the highest target to background ratios (T/B). The T/B values were: 3.91±0.99, 6.15±1.07, 2.8±1.07, 2.57±1.07, 3.41±1.07 and 1.8±1.07 at 3–4 h and 11.9±0.99, 3.94±1.07, 3.79±1.07, 4.81±1.17, 7.0±1.31 and 5.32±1.07 at 16–17 h for mannitol, tricine, glucamine, glucarate-a, glucarate-b, and glucoheptonate, respectively. At both imaging times, 99mTc-mannitol-f-MLFK-HYNIC had the lowest relative levels of accumulation in bowel. Biodistribution measurements demonstrated that the mannitol preparation had the highest level of accumulation (%ID g −1) in infected tissue; mannitol>glucoheptonate=glucarate-b>glucarate-a>glucamine>tricine; P<0.01. The infected to normal muscle ratio for 99mTc-mannitol-f-MLFK-HYNIC was ∼50:1. These results support our previous findings in rats that coligands can markedly effect the biodistribution and infection localization of 99mTc-labeled HYNIC chemotactic peptides. For infection imaging, the mannitol preparation had the most favorable combination of accumulation in infected tissue, T/B ratio and biodistribution in uninfected organs.