Brain Glycogen: An Angel or a Devil for Ischemic Stroke?

Brain Glycogen Works as an Angel Under Physiological Condition

Glycogen Metabolism in the Brain

One of the unique characteristics of the brain is its diversity in neural cell types, and neurons and astrocytes exhibit different metabolic patterns [1]. Glycogen mainly exists in astrocytes, not in neurons [2]. Glycogen metabolism is conservative throughout the body. Similar to the liver and muscle, the astrocytic glycogen level is controlled by glycogenesis and glycogenolysis [3]. Glycogen synthase (GS) and glycogen branching enzyme (GBE) are the rate-limiting enzymes that regulate glycogenesis, and glycogen phosphorylase (GP) and glycogen debranching enzyme (AGL) are the key enzymes for glycogenolysis [3]. Phosphorylated GP is the active form and unphosphorylated GP is the inactive form. Conversely, phosphorylated GS is the inactive form, and unphosphorylated GS is the active form. GS has two isoforms: glycogen synthase 1 (GYS1) and glycogen synthase 2 (GYS2). GP has three isoforms: a muscle isoform, a brain isoform (PYGB), and a liver isoform [3]. The brain mainly has GYS1 and seldom has GYS2. As for GP, PYGB is the predominant isoform in the brain. GS activity is regulated by protein kinase A (PKA) and GS kinase-3 (GSK-3). GP activity is modulated by GP kinase (PhK), which is further controlled by PKA. PhK is also activated by phosphorylation and inactivated by dephosphorylation [3].

Energy Reserve Role of Brain Glycogen

Brain glycogen is an energy reserve and can be rapidly broken down to meet adenosine triphosphate (ATP) deficiency during an energy crisis. Glucose-6-phosphate, the catabolic product of glycogen, can enter into glycolysis and generate pyruvate, which can transfer into mitochondria for aerobic oxidation and generate 30-32 ATPs [4]. Notably, the concentration of brain glycogen is ~10% of muscle glycogen and 1% of liver glycogen [5], so cannot support brain energy consumption for a long period when the cerebrovascular glucose supply is blocked, since the brain consumes ~20% of the body’s energy [6]. Glycogenolysis not only helps to commonplace the energy supply but also satisfies special requirements such as higher local energy demand induced by regional stimulation, stability maintenance during hypoglycemia, responding to rapid and high-demand needs signaled by neuromodulator factors such as norepinephrine, drug addiction, memory formation, and consolidation, as well as sleep and development [7]. Consequently, dysfunction of brain glycogenolysis has been reported to impair memory formation, as well as to increase cortical spreading depression and susceptibility to epileptic seizures [5].

Epigenetic Regulatory Role of Brain Glycogen

Glycosylation is one of the most abundant post-translational modifications regulating multiple processes, such as enzymatic activity, protein folding, subcellular localization, and protein-protein interaction [8]. The biological relevance of glycosylation is evidenced by the fact that 2% of human genes take part in glycan metabolism and >125 human diseases are associated with mutations in those genes [8]. Glycosylation can be divided into several types, including O-, S-, C-, N-, and P-glycosylation [9]. Glucosamine is the substrate for these glycosylations [9]. A recent study found that brain glycogen, which contains 25% glucosamine, is different from liver and muscle glycogen [8]. In addition, GS and GBE can incorporate glucosamine-6-phosphate into brain glycogen, and glycogen can also be broken down into glucosamine-6-phosphate by GP and AGL (Fig. 1) [8]. Not only nuclear histone but also several proteins in the cytoplasm, including some enzymes, can be glycosylated by glucosamine [10]. Glycosylation disorders are a hallmark of many brain diseases. In ischemic stroke, serum IgG N-glycosylation is a biomarker for the severity of inflammation [11]. Previous studies have suggested that the B2R glycosylation level is increased during thrombolysis with recombinant tissue plasminogen activator (rt-PA) at 72 h after reoxygenation in the brain endothelial cell; this accelerates the hemorrhagic transformation in thrombolysis [12]. In addition, the highly glycosylated CD147 in the endothelium caused by diabetes, also increases the risk of hemorrhagic transformation during thrombolysis by rt-PA after ischemic stroke [13]. From this perspective, glycogen has evolved into an epigenetic regulator by glycosylation modification.

Fig. 1.

Brain glycogen serves as a glycosylation regulator and glycophagy substrate. On the one hand, brain glycogen can directly degrade into glucosamine-6-phosphate (glucosamine-6-P). Glucosamine-6-P first converts into acetylglucosamine-6-phosphate (acetylglucosamine-6-P) via glucosamine-6-phosphate acetyltransferase 1 (GNPNAT1), then transfers into acetylglucosamine-1-phosphate (acetylglucosamine-1-P) via phosphoglucomutase 3 (PGM3), and changes into UDP-acetylglucosamine (UDP-GlcNAc) via UDP-acetylglucosamine pyrophosphorylase 1 (UAP1), which is the substrate for glycosylated protein via O-GlcNAc transferase (OGT). The glycosylated protein can also transform into a non-glycosylated protein via O-GlcNAcase (OGA). On the other hand, glycogen can move into the phagophore under the guidance of starch-binding domain-containing protein 1 (STBD1) and γ-aminobutyric acid receptor-associated protein-like 1 (GABARAPL1). After the combination of STBD1 and GABARAPL1, glycophagosomes form and move into lysosomes to degrade glycogen into glucose by acid α-glucosidase (GAA).

Glycophagy Substrate Role of Brain Glycogen

The glycogen particle is not only a significant glucose store but also a cellular organelle. Autophagy is the basic physiological process by which cells eliminate waste organelles, particularly during stress and starvation [14]. It occurs mainly in three modes: microautophagy, macroautophagy, and chaperone-mediated autophagy [14]. Glycophagy is a form of chaperone-mediated autophagy that degrades glycogen by lysosomes [14]. Briefly, starch-binding domain-containing protein 1 allows glycogen to combine with γ-aminobutyric acid receptor-associated protein-like 1, the marker protein in the phagophore [14]. After the combination, the glycogen is packaged by the phagophore and sent into the lysosome, where it is then degraded by acid α-glucosidase [14]. Thus, glycophagy is accomplished using three organelles, the glycogen particle, the phagophore, and the lysosome (Fig. 1). Different from glycogenolysis, the product of glycophagy is glucose, not glucose-6-phosphate [15]. Glycophagy works as a complementary glycogen degradation pathway, and its dysfunction induces many diseases, such as infantile Pompe disease and diabetic cardiomyopathy [15]. A common feature of these diseases is the accumulation of glycogen in the cytoplasm [15]. Furthermore, dysfunctional glycophagy-related glycogen accumulation is the cause of neurodegeneration in Lafora disease [15]. Their findings indicate that glycophagy is necessary for maintaining cellular survival and organismal health.

Brain Glycogen Turns into a Devil During Reperfusion After Ischemic Stroke

Astrocytic Glycogen is Accumulated During Ischemia/Reperfusion (I/R) Injury

Astrocytic glycogen changes significantly in the different phases during ischemic stroke. In the reperfusion phase, astrocytic glycogen is excessively accumulated in the penumbra [16, 17]. Brain glycogen level begins to increase 6 h after cerebrovascular recanalization; this lasts for at least 5 days and disappears at day 7 post-reperfusion [18]. Further studies revealed that glycogen accumulates mainly in astrocytes, not in neurons [18]. Based on the in vivo mouse and in vitro cultured astrocyte models, glycogenolysis dysfunction, not glycogenesis dysfunction, is responsible for the astrocytic glycogen accumulation in the reperfusion phase post-stroke [17, 18]. Decreased GP expression and activity are the main reasons for glycogenolysis dysfunction [17, 18]. In addition, the activities of PKA and PhK, the upstream regulators of GP, are reduced after reperfusion, which leads to GP inactivation [18]. Surprisingly, the GS activity is not affected because of the neutralizing effect of PKA suppression and GSK-3β activation [18].

It is noteworthy that glycogen can also be degraded into glucose by glycophagy [15]. However, the alteration of astrocytic glycophagy and its relationship with glycogen accumulation in cerebral I/R has not been reported. It is possible that astrocytic glycophagy is impaired during I/R, which aggravates glycogen accumulation. Another possibility is that astrocytic glycophagy is enhanced, which can alleviate glycogen accumulation to a certain extent during I/R. Besides, astrocytic glycophagy might also be unaffected during I/R. Accordingly, this issue deserves to be further investigated in the future.

Accumulated Glycogen Is Harmful and Aggravates Cerebral I/R Injury

Does the increased brain glycogen benefit the outcome of cerebral reperfusion injury? Conversely, the more brain glycogen accumulates, the worse the cerebral injury becomes [18]. One study found that the dysfunction of glycogenolysis impairs the degradation of accumulated glycogen into glucose-6-phosphate and accordingly the level of glucose-6-phosphate is decreased during I/R. The glucose-6-phosphate can enter the pentose phosphate pathway and generate nicotinamide adenine dinucleotide phosphate (NADPH) and glutathione, which are important antioxidants in cells. The decrease of glucose-6-phosphate leads to an increase of reactive oxygen species (ROS), which aggravates oxidative damage and affects brain repair during reperfusion after ischemic stroke [19]. Moreover, cultured astrocytes are more resistant to oxygen-glucose deprivation (OGD) when accumulated glycogen is degraded via the overexpression of PYGB, and the survival of neurons is also enhanced when cocultured with astrocytes overexpressing PYGB [18]. In addition, classical neuroprotectants, including insulin and salvianolic acid B, activate the astrocytic glycogenolysis flux, which accelerates the recovery from brain damage after ischemic stroke [18, 19]. These results provide evidence that the accumulated glycogen is harmful and works as an evil in cerebral reperfusion injury.

Mechanism Behind Neuroprotection Mediated by Accumulated Glycogen Degradation in I/R Injury

Glucose-6-phosphate, known as the product of glycogenolysis, has three metabolic directions: glycolysis, the pentose phosphate pathway (PPP), or gluconeogenesis [20]. The glucose-6-phosphate level is increased after glycogenolysis activation following OGD in vitro [21]. Besides, the PPP pathway flux is enhanced and the levels of astrocytic antioxidants are increased following glucose-6-phosphate enhancement after glycogen mobilization, which accordingly causes a decrease of the intracellular ROS level [21]. The decreased astrocytic ROS level leads to the suppression of nuclear transcription factor-κB (NF-κB) and the activation of signal transducer and activator of transcription 3 (STAT3) [21]. A previous study reported that ischemia induces astrocytes to transform into A1 (toxic) or A2 (protective) subtypes [22]. NF-κB and STAT3 are the key modulators for A1 and A2 astrocytes, respectively [21]. Consequently, the restoration of glycogenolysis can change the astrocytic status after OGD and enhance A2 astrocyte formation and decrease A1 astrocyte formation, achieved by ROS-mediated STAT3 activation and NK-κB inhibition.

Notably, PPP enhancement may not be the only pathway for the neuroprotection mediated by the restoration of glycogenolysis in reperfusion injury after ischemic stroke. A previous study has suggested that 25% of brain glycogen is formed from glucosamines [8], and a large amount of glucosamine-6-phosphate can be generated after the breakdown of accumulated glycogen, which might be beneficial to reperfusion injury. In addition, previous studies have pointed out that astrocytes provide neurons with lactate to meet energy deficiency during an energy crisis [1]. The glucose-6-phosphate enhancement induced by the breakdown of accumulated glycogen enter into glycolysis and produce lactate, which is transported into surrounding neurons via the astrocyte–neuron lactate shuttle and generates more energy for neuron survival [1]. Although gene modification or drug interference of the synthesis and breakdown of brain glycogen may help to enhance the resistance of astrocytes and surrounding neurons to OGD damage, brain glycogen is not a appropriate therapeutic target because it affects many physiological processes and may have unexpected side-effects during reperfusion injury after ischemic stroke.

Conclusion

The transformation of brain glycogen from an angel under physiological conditions into a devil under pathological conditions coincides with the philosophy that things will develop in the opposite direction when they become extreme (Fig. 2). Although treatments targeting GP to accelerate accumulated glycogen degradation seem like a promising therapeutic strategy to reverse this angel-to-devil transformation, they might disrupt the whole cellular metabolism and cause unpredictable damage to the brain during reperfusion after ischemic stroke.

Fig. 2.

A diagram to summarize the role of astrocytic glycogen under physiological conditions and in the reperfusion phase after ischemic stroke. Under physiological conditions, glycogen can degrade into glucose-6-phosphate (G6P). The G6P can change into lactate via glycolysis, which works as a metabolic support for cells. The G6P can also enter the pentose phosphate pathway (PPP) to generate NADPH, which decreases cellular ROS and has antioxidative effects. Besides, the glycogen can directly degrade into glucosamine-6-phosphate (glucosamine-6-P) and then convert into the glycosylated substrate UDP-GlcNAc, which regulates the glycosylation of astrocytic protein. In ischemic stroke, the glycogenolysis is dysfunctional due to the impairment of glycogen phosphorylase (GP) and causes a decrease of G6P and glucosamine-6-P, which accordingly aggravates the astrocytic metabolic deficiency, oxidative damage, and epigenetic disorder during the reperfusion phase after ischemic stroke.

Conflict of interest

The author declare that they have no Conflict of interest.