Imaging of brain tissue pH and metabolites. A new approach for the validation of volume-selective NMR spectroscopy.

For the validation of volume-selective 1H and 31P NMR spectroscopy of the brain methods are required that allow high resolution quantitative mapping of tissue pH and metabolites on intact brain slices. The following techniques are proposed for this purpose. Tissue pH is imaged on cryostat sections of in situ frozen brains, using umbelliferone as a fluorescent pH indicator (Csiba et al, Brain Res 289, 334-337 (1983]. Regional tissue ATP content is measured in adjacent cryostat sections, using the luciferine/luciferase system of fireflies for evoking substrate-specific bioluminescence (Kogure and Furones Alonso, Brain Res. 154, 273-284 (1978]. Lactate content is imaged in a similar way by inducing substrate-specific bioluminescence with lactate dehydrogenase and luciferase from vibrio Fischeri (Paschen, J. Cereb. Blood Flow Metab. 5, 609-612 (1985]. The spatial resolution of these techniques is better than 100 mu, as exemplified in experimental brain tumors and brain infarct of cats. The applicability of biochemical mapping for the validation of NMR spectroscopy was tested in a global brain ischemia model of cat by correlating surface coil 31P and 1H spectra with the corresponding regional biochemical data, measured in the sensitive volume of the coil. Correlation coefficients were r = 0.907, 0.852 and 0.924 for pH, lactate and ATP, respectively. These results demonstrate that the biochemical measurements obtained by bioluminescence and fluoroscopic imaging correlate closely with the NMR data and, therefore, are appropriate for the validation of more complex applications, such as volume-selective spectroscopy of brain infarcts or tumors.