Comparative studies of antagonist-based metabotropic glutamate-receptor targeted MRI contrast agents

Targeting receptors is a widely used concept for competitive binding in brain imaging techniques such as positron emission tomography and optical imaging and an ongoing debate exists whether antagonist- or agonist-based approaches are more beneficial. So far, applying competitive binding for brain functional magnetic resonance imaging (fMRI) has not been demonstrated. Our intention is to develop antagonist-based glutamate (Glu) "responsive" MRI contrast agents (CAs) that would allow imaging of Glu fluctuations in specific brain regions associated with neural activity. We have therefore chosen CAs that bind to the metabotropic Glu-receptor subtype 5 (mGluR5).[1] Such molecules will bind both to neuronal postsynaptic receptors as well as to those expressed on astrocytes. Hence, these CAs can act both as "markers" of receptor density and as indicator of neuronal activation. Ideally, for the latter, after Glu binding to the receptor (i.e. upon Glu release at the synapse) the CA will be released, in turn leading to a reduction in image contrast, followed by a restoration of equilibrium and re-binding of the CA to the receptor. These events are believed to occur over a period of a few seconds allowing data acquisition using modern fast MR-techniques.[2] We have synthesized different prospective CAs derived from potent mGluR5-antagonists (alkynes (e.g. MPEP or MTEP) and dipyridyl/heterobiaryl amides [1]) coupled to DOTA-derived gadolinium chelates. These CAs were evaluated for cellular labeling, cytotoxicity, antagonistic activity, and receptor binding: (A) Binding affinity studies in cultured primary rat astrocytes (expressing mGluR5 as shown by immunofluorescence) showed that three of the compounds increased the cellular relaxation rate R1,cell up to 135% of control (200M, 45 min., 37°C). These studies were also done at 0°C (to prevent endocytosis). Under these conditions R1,cell was increased only to ~110% of control. MRI-measurements were done with a 3T human whole body scanner. (B) None of the CAs had cytotoxic effects at the concentrations used (24h incubation). (C) Antagonistic activity of the CAs was calculated as changes in EC50 of Glu using a calcium fluorescence assay on primary rat astrocytes. Two of the compounds with increased R1,cell also retained significant antagonistic activity, one in each structural class.[1] (D) Receptor binding and reversibility was demonstrated for one of the dipyridyl derivatives (which have an inherent fluorescence that changes upon binding) on commercially available receptor preparations containing recombinant human mGluR5.[1] By using primary rat astrocytes as cellular model to investigate our newly developed mGluR5-targeted MRI contrast agents, we were able to identify two promising candidates. The CAs are based on the structures of antagonists to mGluR5 and our studies demonstrate that it might be possible to use these targeted CAs to image receptor density in the brain using MRI. Furthermore, our results point towards the possibility of using MRI for brain functional measurements employing competitive binding approaches