A method has been developed for the simultaneous measurement of the rates of glucose consumption in the various structural and functional components of the brain in vivo. The method can be applied to most laboratory animals in the conscious state. It is based on the use of z -d e o x y -~ ['4C]glucose (['4C]DG) as a tracer for the exchange of glucose between plasma and brain and its phosphorylation by hexokinase in the tissues. [14C]DG is used because the label in its product, [ '4C]deoxyglucose-6-phosphate, is essentially trapped in the tissue over the time course of the measurement. A model has been designed based on the assumptions of a steady state for glucose consumption, a first order equilibration of the free [14C]DG pool in the tissue with the plasma level, and relative rates of phosphorylation of [14C]DG and glucose determined by their relative concentrations in the precursor pools and their respective kinetic constants for the hexokinase reaction. An operational equation based on this model has been derived in terms of determinable variables. A pulse of [14C]DG is administered intravenously and the arterial plasma [ 14C]DG and glucose concentrations monitored for a preset time between 30 and 45min. At the prescribed time, the head is removed and frozen in liquid N,-chilled Freon XII, and the brain sectioned for autoradiography. Local tissue concentrations of ['4C]DG are determined by quantitative autoradiography. Local cerebral glucose consumption is calculated by the equation on the basis of these measured values.The method has been applied to normal albino rats in the conscious state and under thiopental anesthesia. The results demonstrate that the local rates of glucose consumption in the brain fall into two distinct distributions, one for gray matter and the other for white matter. In the conscious rat the values in the gray matter vary widely from structure to structure (54-197pmo1/100g/min) with the highest values in structures related to auditory function, e.g. medial geniculate body, superior olive, inferior colliculus, and auditory cortex. The values in white matter are more uniform (i.e. 33-40 pmo1/100 g/min) at levels approximately one-fourth to one-half those of gray matter. Heterogeneous rates of glucose consumption are frequently seen within specific structures, often revealing a pattern of cytoarchitecture. Thiopental anesthesia markedly depresses the rates of glucose utilization throughout the brain, particularly in gray matter, and metabolic rate throughout gray matter becomes more uniform at a lower level THE MAMMALIAN brain is a complex heterogeneous organ comprising many structural and functional components with different and independently regulated levels of functional activity and energy metabolism. Much of our present knowledge of cerebral .

The autoradiographic diffusible tracer technique for the measurement of local cerebral blood flow was originally designed for use with the radioactive, inert gas 131I-labeled trifluoroiodomethane and is applicable only with tracers that exhibit unrestricted diffusion through the blood-brain barrier. Because of the technical problems associated with the use of gaseous tracers, a suitable nonvolatile tracer has been sought. [14C] Antipyrine has been used previously and found to be unsuitable because of limitations in its diffusion through the blood-brain barrier. An analogue of [14C]antipyrine, iodo [14C]antipyrine, exhibits higher partition coefficients than [14C]antipyrine between nonpolar solvents and water and might, therefore, be expected to diffuse more freely through the barrier. Its use as the tracer in the local blood flow technique leads to values considerably above those obtained with [14C]antipyrine in the rat and cat and essentially the same as those obtained with the gas trifluoro[131I]iodomethane in the cat. Iodo[14C]antipyrine appears, therefore, to be a satisfactory nonvolatile tracer for the measurement of local cerebral blood flow.

The [14C]deoxyglucose method for quantitative determination of local cerebral glucose utilization was extended to the macaque monkey. The necessary constants required by its operational equation were evaluated. The lumped constant, measured in 7 normal conscious monkeys, was found to equal 0.344 (SEM, +/- 0.036). The rate constants were also estimated and found to be very similar to those obtained previously in the rat. With these essential constants evaluated, the method was applied to normal conscious monkeys. Local cerebral glucose utilization was found to vary marked throughout the brain but to fall within two distributions, a higher one in gray matter and lower one in white matter. In general, the values fell in a range to be expected from previous measurements of average energy metabolism in the brain as a whole. The values were considerably lower than those observed previously in the conscious rat. Marked heterogeneity of rates of glucose utilization were observed in a number of anatomical structures. In some cases the patterns of heterogeneity were consistent with known histological cytoarchitecture; in others the heterogeneity did not conform with known cytoarchitectural features but corresponded to patterns previously demonstrated by electrophysiological techniques. Many of the regions of the cerebral cortex showed columnar patterns of distribution of higher and lower rates of glucose utilization. These may be a metabolic reflection of the columnar organization of function within the cerebral cortex.