Dinuclear DOTA-Based Gd-III Chelates - Revisiting a Straightforward Strategy for Relaxivity Improvement

The need for magnetic resonance imaging (MRI) contrast agents with improved relaxivity maintains the development of new Gd-III chelates as an intensive and demanding field of research. In this work, we introduce the new dimeric chelators bis{1,4,7,10-tetraazacyclododecane-1-[(6-amino)hexanoic]- 4,7,10-triacetic acid} adipate [L2, bis(DOTA-AHA)-adipate] and bis{1,4,7,10-tetraazacyclododecane-1-[(6amino) hexanoic]-4,7,10-triacetic acid} 1,3-phenyldiacetate [L3, bis(DOTA-AHA)1,3-phenyldiacetate], which are based on the bifunctional ligand 1,4,7,10-tetraazacyclododecane-1-[(6-amino) hexanoic]-4,7,10-triacetic acid (L1, DOTA-AHA). Their Gd-III chelates were studied by variable-temperature H-1 nuclear magnetic relaxation dispersion (NMRD) and O-17 NMR spectroscopy to measure their relaxivities and the parameters that govern them. The rates of exchange of innersphere water from the monomer GdL1 and from the two dinuclear chelates Gd(2)L2 and Gd(2)L3 are very similar ((298)k(ex) approximate to 6.5 x 10(6) s(-1)) and slightly faster than that for [Gd(DOTA)-H2O](-) ((298)k(ex) = 4.1 x 10(6) s(-1)). All three compounds form weakly bound aggregates with equilibrium constants K-298 of 2.9, 15.6, and 14.6 for GdL1, Gd(2)L2, and Gd(2)L3, respectively. Even though the aggregates contain only 10 to 15% of the total amount of Gd-III ions, a marked increase in relaxivity between 30 and 100 MHz is observed. Furthermore, the distance between the two Gd-III centers in the dinuclear compounds has been determined by double electron-electron resonance (DEER) spectroscopy experiments and by molecular modeling studies, which afforded comparable distances. The linkers between the chelating moieties allow Gd-III-Gd-III distances of ca. 3.0 nm for the completely stretched linker conformation and less than 1.9 nm for the conformation with the metal centers at a closer distance. These metal-to-metal distances by themselves cannot explain the considerably long tumbling times of the chelates in solution. Only a model consistent with some level of aggregation for the dinuclear chelates in aqueous solution could satisfactorily explain our results