Establishment of in vivo brain imaging method in conscious mice

In vivo imaging, such as PET, requires restriction of body movements and is generally conducted under sedation by anesthetic agents in studies using laboratory animals. Because anesthetics reduce neural activity and metabolism, physiologic neural functions are difficult to assess in animal PET studies. Therefore, use of an appropriate method in conscious animals is important and is a practical requirement for physiologic in vivo brain imaging studies. Here, we established an in vivo imaging system for conscious mice to reveal the physiologic regional cerebral glucose metabolic rate (rCMRglu) with (18)F-FDG PET.

We first developed a head holder to enable brain PET of a conscious mouse. To obtain optimal rCMRglu, we examined the effects of physical and psychologic stresses caused by ambient temperature, intravenous injection, and acclimation to the apparatus and immobile state. Finally, quantitative kinetic analysis was performed for rCMRglu based on a 2-tissue-compartment model with an input function of arterial blood sampling under both conscious and anesthetized (1.5% isoflurane) conditions.

Increasing the ambient temperature increased uptake of (18)F-FDG in the brain significantly while reducing the uptake in skeletal muscle and brown adipose tissue that was caused by shivering. The reduction of brain (18)F-FDG uptake caused by tail holding and manual injection was significantly ameliorated by the use of an automated slow injection. Although brain uptake of (18)F-FDG varied at the first session of PET, uptake at the second and subsequent sessions was stable, even after long-term acclimation. After these beneficial changes, brain uptake of (18)F-FDG improved significantly, to approximately 260% above the preconditioned state, which is comparable with that obtained in mice that have been allowed to move freely about their home cages. Quantitative kinetic analyses revealed that isoflurane anesthesia lowered rCMRglu in the cerebral cortex, striatum, thalamus, and cerebellum by 66%, 59%, 62%, and 22%, respectively, mainly by reducing the k(3) value, a rate constant for phosphorylation by hexokinase.

To our knowledge, this is the first study to report quantitative kinetic analysis of rCMRglu in mice that have been conscious throughout PET. This investigation will open avenues for research into in vivo functional brain molecular imaging in both normal and genetically manipulated mice.