Prologue
Brain work has enjoyed wide-spread scientific interest since Roy and Sherrington (1890) observed that sensory nerve stimulation expanded cerebral cortex as a result of an increase in blood flow. Increased brain function seemed to produce metabolites that caused the dilation of cerebral blood vessels presumably to fulfill the increased demand for nutrients. These findings sparked intense research to identify the cellular metabolic pathways involved and to uncover the mechanisms that couple blood flow to brain function.
Indeed, brain cells need plenty calories. The brain accounts for an estimated 20 percent of our body's energy consumption. The aerobic metabolism of glucose, also known as aerobic glycolysis, has been shown to constitute the most prominent source for the brain's energy needs.
Aerobic glycolysis entails the oxidation of glucose into water and carbon dioxide to produce energy-rich phosphorous compounds, mainly adenosine triphosphate (ATP). The compounds are used to fuel active transmembrane transport mechanisms of ions and molecules that are needed for nerve cell communication and information processing. However, only insufficient amounts of glucose and no oxygen can be stored in the brain. Both have to be delivered to the brain tissue on demand.
In the late 1960s, Lou Sokoloff and his colleagues at the National Institute of Mental Health succeeded in developing a procedure with which the local cerebral metabolic rates of glucose could be determined in vivo. They named it the autoradiographic deoxyglucose method (Sokoloff and others, 1977). Autoradiography involves the administration of a radioactively labeled tracer, the distribution of which in the tissue can be visualized postmortem by apposing X-ray films to tissue sections. The grain density in the autoradiograms on the developed films corresponds to the tracer concentration in the sections. The grain density can be measured with an optical densitometer and calibrated to represent tracer concentration in the tissue.
Glucose is rapidly metabolized and the metabolites are quickly removed from the brain. Therefore, Sokoloff and colleagues could not use radioactively labeled glucose as tracer for their measurements. They used deoxyglucose instead. Deoxyglucose is a competitive analogue of glucose. It is phosphorylated to deoxyglucose-6-phosphate by the enzyme hexokinase in the first step of glycolysis, but is not metabolized further, and accumulates in the brain tissue. The tracer accumulation is proportionate to the rate of glucose metabolism.
Using this method, numerous studies have shown that glucose metabolism in the central nervous system is associated with electrical nerve cell spiking activity (e.g. Kadekaro and others, 1985). That is, nerve cell stimulation evokes electrical millivolt spikes in the nerve cell's axon by rapid fluxes of sodium ions into the cell and potassium ions out of the cell through ion-specific membrane channels. The spikes are known as action potentials. The axon is the long process that connects the nerve cells. Trains of action potentials traveling along axons essentially constitute packets of data transmitted to other nerve cells.
In order to prepare the cell for the next spike, the ion concentrations must be restored to their pre-action potential level, known as resting potential. The sodium ions must be pumped out of the cells with an energy-consuming pump. In accord, Mata and others, (1980) observed that glucose metabolism diminishes significantly, when the sodium pump is blocked and the production of action potentials ceases, supporting the idea that cerebral energy consumption is related to the spiking activity of nerve cells.
Sokoloff encountered criticism. One critic was the Dutch biochemist Cees Van den Berg, who claimed that the deoxyglucose method did not measure glucose metabolism accurately. He had measured concentrations of pyruvate, a glycolytic key intermediate, and his results suggested a much greater glucose consumption (Van den Berg and Bruntink, 1983). I met him after he had applied for a grant to fund experiments designed to support his claim. He was invited to present a lecture on what he thought needed to be done. Many years later, I asked Lou Sokoloff about Van den Berg's ideas. Lou swiftly and resolutely retorted, "Nonsense!" In as much as the size of a lake does not tell us how fast water is running through it, the concentration of a metabolite is no indication of its metabolic rate.
However, one of Van den Berg's claims still resonates in my mind. He noted that the rate of glucose metabolism measured with the deoxyglucose method would increase no more than two-fold from baseline in normal physiological circumstances. I have carried out hundreds of experiments with this method in rodents. On this point, Van den Berg was correct. Physiological whisker stimulation increases glucose metabolic rates in primary somatic sensory cortex less than two-fold. Indeed, the average is about 1.3-fold.
Nerve cell activity-related increases in glucose metabolism should have been greater, if they had been tightly correlated with the energy demands of axonal sodium pump activity estimated at the time. In their pioneering work, the Noble Prize-laureates Hodgkin and Huxley were the first to describe the ion currents that produced action potentials in squid giant axons (Hodgkin and Huxley, 1952a). In a subsequent paper, the authors suggested a model quantifying these currents (Hodgkin and Huxley,1952b). According to the model, sodium influx was anticipated to rise 4-fold or greater. Consequentially, stimulus-evoked nerve cell spiking activity should have resulted in much greater increases in glucose use to pump out the ions than found with the deoxyglucose method.
The New Find
Yesterday, Science Magazine published a study the results of which may resolve this apparent inconsistency (Alle and others, 2009). The authors used intracellular patch clamp recordings in mice to quantify the action potential-related transmembrane sodium and potassium ion fluxes in mossy fibers of a phylogenetically old part of the cerebral cortex known as hippocampus. They are unmyelinated. That is, like axons of short-distance nerve cell connections in neocortex, mossy fibers lack a fatty sheeth for electrical insulation and, therefore, propagate action potentials similarly as the axons of local intracortical circuits. The authors observed sodium and potassium ion fluxes that were less overlapping and smaller than those found in squid giant axons. They suggest that these differences reduce the action potential-related increase in energy demand to 1.3-fold. This increase is in good agreement with the observed stimulus-related increases in glucose utilization rates discussed above.
In addition, supra-cellular mechanism play a crucial role in the reduction of cortical energy needs. Cortical nerve cells are able to control their gain. That is, the rate of action potentials can be reduced in response to an increase in the strength of stimulation without losing information (Melzer and others, 2006). Van den Berg erred in his critique of deoxyglucose method, because the energy demand of cerebral function is lower than he presumed.
On the other hand, it must be noted that sodium pump activity does not solely account for stimulus-related increases in cerebral glucose metabolism. Nerve cells release glutamate at their synapses, that is the contacts between nerve cells, to excite other nerve cells. Glutamate is the most prevalent neurotransmitter in the cerebral cortex. Astrocytes near the synapses are known to remove the released glutamate with active, that is energy-consuming, transport (Pellerin and Magistretti, 1994; Takahashi and others, 1995). Hence, astrocytic glutamate uptake may contribute considerably to the cerebral rates of glucose metabolism. This contribution remains to be quantified.
References
- Alle H, Roth A, Geiger JRP (2009) Energy-efficient potentials in hippocampal mossy fibers. Science 325:1405-1408.
- Hodgkin AL, Huxley AF (1952a) Propagation of electrical signals along giant nerve fibers. Proc R Soc Lond B Biol Sci 140:177-183.
- Hodgkin AL, Huxley AF (1952b) A quantitative description of membrane current and its application to conduction and excitation in nerve. J Physiol (Ldn) 117:500-544.
- Kadekaro M, Crane AM, Sokoloff L (1985) Differential effects of electrical stimulation of sciatic nerve on metabolic activity in spinal cord and dorsal root ganglion in the rat. Proc Natl Acad Sci U S A 82:6010-6013.
- Mata M, Fink DJ, Gainer H, Smith CB, Davidsen L, Savaki H, Schwartz WJ, Sokoloff L (1980) Activity-dependent energy metabolism in rat posterior pituitary primarily reflects sodium pump activity. J Neurochem 34:213-215.
- Melzer P, Sachdev RN, Jenkinson N, Ebner FF (2006) Stimulus frequency processing in awake rat barrel cortex. J Neurosci 26:12198-2205.
- Pellerin L, Magistretti PJ (1994) Glutamate uptake into astrocytes stimulates aerobic glycolysis: a mechanism coupling neuronal activity to glucose utilization. Proc Natl Acad Sci U S A 91:10625-10629.
- Roy CS, Sherrington CS (1890) On the regulation of the blood-supply of the brain. J Physiol (Ldn) 11:85-108.
- Sokoloff L, Reivich M, Kennedy C, Des Rosiers MH, Patlak CS, Pettigrew KD, Sakurada O, Shinohara M (1977) The [14C]deoxyglucose method for the measurement of local cerebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J Neurochem 28:897-916.
- Takahashi S, Driscoll BF, Law MJ, Sokoloff L (1995) Role of sodium and potassium ions in regulation of glucose metabolism in cultured astroglia. Proc Natl Acad Sci U S A 92:4616-4620.
- Van den Berg CJ, Bruntink R (1983) Glucose oxidation in the brain during seizures: experiments with labeled glucose and deoxyglucose. In Glutamine, Glutamate and GABA in the Central Nervous System (Hertz L, Kvamme E, McCeer EG, Schousboe A, eds), New York, Alan R. Liss:pp 619−624.
can still be found as a used book. He was known as an avid aviator and a superb physician scientist. Elaborating on 
