Astrocytes are ‘good scouts': being prepared also helps neighboring neurons

Abstract

Recognition of glycogen as an active participant in the energetics of brain activation is replacing the long-held concept of glycogen as an emergency energy reserve, but the functional roles of glycogen and the cellular utilization of glycogen carbon are unresolved issues. Metabolic modeling by DiNuzzo et al, in this issue predicts that mobilization of glycogen during brain activation provides fuel for activated astrocytes and increases product inhibition of hexokinase thereby reducing astrocytic utilization of blood-borne glucose and increasing glucose availability for activated neurons. Glucose buffering and glucose channeling (not lactate shuttling to neurons) are proposed to be the consequences of glycogenolysis.

Brain requires a continuous glucose–oxygen supply to satisfy energy demands, but astrocytes also store glucose as glycogen and consume it during brain activation. Contributions of astrocytes and glycogen to overall brain energetics are not well understood, and establishing glycogen's functional roles requires quantitative studies that are hindered by technical difficulties related to its lability, isotopic labeling, and sensitivity to metabolic, physiological, and behavioral states of subjects. Early studies reported low glycogen levels in rodent brain (∼1.5 to 3 μmol/g, ≅brain (glucose)), slow ‘resting' turnover, and rapid consumption during energy failure. These data supported the long-held notion of glycogen as a small emergency depot, a concept inconsistent with its rapid mobilization by many neurotransmitters, elevated [K⁺]_{extracellular}, oxidative stress, and sensory stimulation (Hertz et al, 2007). When stimuli are minimized and glycogen phosphorylase totally inactivated, glycogen level is 10 to 12 μmol/g, equivalent to 30 to 40 μmol/g in astrocytes (assumed to be 30% of brain mass). Glycogen now exceeds astrocytic glucose level by 10- to 25-fold, and it can be a major, dynamic fuel, consistent with the large compensatory rise in blood-borne glucose utilization evoked by sensory stimulation during glycogenolysis blockade (Dienel et al, 2007). It is not known if astrocytes completely metabolize the glycogen in vivo, and its metabolic and cellular fates are important, unresolved issues. Negligible brain glucose-6-phosphatase activity prevents conversion of glycogen to glucose. Cultured astrocytes release glycogen-derived lactate, and neuronal protection and prolongation of neuronal function by glycogen under severe pathophysiological conditions probably involves metabolite trafficking. However, the notion that astrocyte-derived lactate is a major neuronal fuel during physiological activation of normoglycemic conscious subjects is controversial. Lactate cannot fulfill neuronal requirements for glucose involving flux through the glycolytic (e.g., synaptic vesicle loading, ion pumps) and pentose phosphate shunt (reduced nicotinamide adenine dinucleotide phosphate, NADPH, to manage oxidative stress) pathways.

In this Journal issue, DiNuzzo et al address the glucose–glycogen–lactate–fuel issue by evaluating the influence of glycogen turnover on neuroenergetics of cellular activation. They extend their simulation analysis (DiNuzzo et al, 2010) of nutrient transport and metabolite trafficking, and their model-predicted outcomes reveal unanticipated, paradigm-shifting insights into glucose supply demand dynamics and astrocyte–neuron interactions. Rapid generation of glucose-6-phosphate from glycogen compensates for astrocytic glucose transport limitations during high activation. Glucose-6-phosphate not only fuels astrocytes, its higher level increases the extent of product inhibition of hexokinase thereby reducing astrocytic phosphorylation of glucose, particularly at stimulus onset; this makes more blood-borne, extracellular glucose available for working neurons. Avid neuronal glucose uptake and metabolism also creates a concentration gradient that enables unmetabolized, intracellular glucose to diffuse from astrocytes to neurons. Glycogenolysis does not alter predominant neuronal lactate generation and lactate shuttling to astrocytes during the early phase of high astrocytic stimulation. As glycogen is an on-demand, renewable buffer that transiently reduces astrocytic dependence on blood glucose during high activation, the scout motto applies, ‘be prepared.' Sparing of astrocytic glucose utilization channels more glucose to activated neurons from blood and from the astrocytic syncytium. Metabolic modeling provides a framework for experimental design, and these novel concepts present technically difficult challenges to experimentalists.

The author declares no conflict of interest.