Identifying non-toxic doses of manganese for manganese-enhanced magnetic resonance imaging to map brain areas activated by operant behavior in trained rats

Manganese-enhanced magnetic resonance imaging (MEMRI) offers unique advantages such as studying brain activation in freely moving rats, but its usefulness has not been previously evaluated during operant behavior training. Manganese in a form of MnCl2, at a dose of 20 mg/kg, was intraperitoneally infused. The administration was repeated and separated by 24 h to reach the dose of 40 mg/kg or 60 mg/kg, respectively. Hepatotoxicity of the MnCl2 was evaluated by determining serum aspartate aminotransferase, alanine aminotransferase, total bilirubin, albumin and protein levels. Neurological examination was also carried out. The animals were tested in visual cue discriminated operant task. Imaging was performed using a 3T clinical MR scanner. T1 values were determined before and after MnCl2 administrations. Manganese-enhanced images of each animal were subtracted from their baseline images to calculate decrease in the T1 value (ΔT1) voxel by voxel. The subtracted T1 maps of trained animals performing visual cue discriminated operant task, and those of naive rats were compared. The dose of 60 mg/kg MnCl2 showed hepatotoxic effect, but even these animals did not exhibit neurological symptoms. The dose of 20 and 40 mg/kg MnCl2 increased the number of omissions and did not affect the accuracy of performing the visual cue discriminated operant task. Using the accumulated dose of 40 mg/kg, voxels with a significant enhanced ΔT1 value were detected in the following brain areas of the visual cue discriminated operant behavior performed animals compared to those in the controls: the visual, somatosensory, motor and premotor cortices, the insula, cingulate, ectorhinal, entorhinal, perirhinal and piriform cortices, hippocampus, amygdala with amygdalohippocampal areas, dorsal striatum, nucleus accumbens core, substantia nigra, and retrorubral field. In conclusion, the MEMRI proved to be a reliable method to accomplish brain activity mapping in correlation with the operant behavior of freely moving rodents.