Refinement of the kinetic model of the 2-[14C]deoxyglucose method to incorporate effects of intracellular compartmentation in brain.

A translocase to transport hexose phosphate formed in the cytosol into the cisterns of the endoplasmic reticulum, where the phosphatase resides, is absent in brain (Fishman and Karnovsky, 1986). 2-Deoxyglucose-6-phosphate (DG-6-P) may therefore have limited access to glucose-6-phosphatase (G-6-Pase), and transport of the DG-6-P across the endoplasmic reticular membrane may be rate limiting to its dephosphorylation. To take this compartmentation into account, a five-rate constant (5K) model was developed to describe the kinetic behavior of 2-deoxyglucose (DG) and its phosphorylated product in brain. Loss of DG-6-P was modeled as a two-step process: (a) transfer of DG-6-P from the cytosol into the cisterns of the endoplasmic reticulum; (b) hydrolysis of DG-6-P by G-6-Pase and subsequent return of the free DG to the precursor pool. Local CMRglc (LCMRglc) was calculated in the rat on the basis of this model and compared with values calculated on the basis of the three-rate constant (3K) and the four-rate constant (4K) models of the DG method. The results show that under normal physiological conditions all three models yield values of LCMRglc that are essentially equivalent for experimental periods between 25 and 45 min. Therefore, the simplest model, the 3K model, is sufficient. For experimental periods from 60 to 120 min, the 4K and 5K models do not correct completely for loss of product, but the 5K model does yield estimates of LCMRglc that are closer to the values at 45 min than those obtained with the 3K and 4K models.