Brain glycogen serves as a critical glucosamine cache required for protein glycosylation.

Glycosylation defects are a hallmark of many nervous system diseases. However, the molecular and metabolic basis for this pathology is not fully understood. In this study, we found that N-linked protein glycosylation in the brain is metabolically channeled to glucosamine metabolism through glycogenolysis. We discovered that glucosamine is an abundant constituent of brain glycogen, which functions as a glucosamine reservoir for multiple glycoconjugates. We demonstrated the enzymatic incorporation of glucosamine into glycogen by glycogen synthase, and the release by glycogen phosphorylase by biochemical and structural methodologies, in primary astrocytes, and in vivo by isotopic tracing and mass spectrometry. Using two mouse models of glycogen storage diseases, we showed that disruption of brain glycogen metabolism causes global decreases in free pools of UDP-N-acetylglucosamine and N-linked protein glycosylation. These findings revealed fundamental biological roles of brain glycogen in protein glycosylation with direct relevance to multiple human diseases of the central nervous system.

Discovery and Development of Small-Molecule Inhibitors of Glycogen Synthase.

The overaccumulation of glycogen appears as a hallmark in various glycogen storage diseases (GSDs), including Pompe, Cori, Andersen, and Lafora disease. Accumulating evidence suggests that suppression of glycogen accumulation represents a potential therapeutic approach for treating these GSDs. Using a fluorescence polarization assay designed to screen for inhibitors of the key glycogen synthetic enzyme, glycogen synthase (GS), we identified a substituted imidazole, (rac)-2-methoxy-4-(1-(2-(1-methylpyrrolidin-2-yl)ethyl)-4-phenyl-1H-imidazol-5-yl)phenol (H23), as a first-in-class inhibitor for yeast GS 2 (yGsy2p). Data from X-ray crystallography at 2.85 Å, as well as kinetic data, revealed that H23 bound within the uridine diphosphate glucose binding pocket of yGsy2p. The high conservation of residues between human and yeast GS in direct contact with H23 informed the development of around 500 H23 analogs. These analogs produced a structure-activity relationship profile that led to the identification of a substituted pyrazole, 4-(4-(4-hydroxyphenyl)-3-(trifluoromethyl)-1H-pyrazol-5-yl)pyrogallol, with a 300-fold improved potency against human GS. These substituted pyrazoles possess a promising scaffold for drug development efforts targeting GS activity in GSDs associated with excess glycogen accumulation.

Redox Switch for the Inhibited State of Yeast Glycogen Synthase Mimics Regulation by Phosphorylation.

Glycogen synthase (GS) is the rate limiting enzyme in the synthesis of glycogen. Eukaryotic GS is negatively regulated by covalent phosphorylation and allosterically activated by glucose-6-phosphate (G-6-P). To gain structural insights into the inhibited state of the enzyme, we solved the crystal structure of yGsy2-R589A/R592A to a resolution of 3.3 Å. The double mutant has an activity ratio similar to the phosphorylated enzyme and also retains the ability to be activated by G-6-P. When compared to the 2.88 Å structure of the wild-type G-6-P activated enzyme, the crystal structure of the low-activity mutant showed that the N-terminal domain of the inhibited state is tightly held against the dimer-related interface thereby hindering acceptor access to the catalytic cleft. On the basis of these two structural observations, we developed a reversible redox regulatory feature in yeast GS by substituting cysteine residues for two highly conserved arginine residues. When oxidized, the cysteine mutant enzyme exhibits activity levels similar to the phosphorylated enzyme but cannot be activated by G-6-P. Upon reduction, the cysteine mutant enzyme regains normal activity levels and regulatory response to G-6-P activation.

Incorporation of phosphate into glycogen by glycogen synthase.

The storage polymer glycogen normally contains small amounts of covalently attached phosphate as phosphomonoesters at C2, C3 and C6 atoms of glucose residues. In the absence of the laforin phosphatase, as in the rare childhood epilepsy Lafora disease, the phosphorylation level is elevated and is associated with abnormal glycogen structure that contributes to the pathology. Laforin therefore likely functions in vivo as a glycogen phosphatase. The mechanism of glycogen phosphorylation is less well-understood. We have reported that glycogen synthase incorporates phosphate into glycogen via a rare side reaction in which glucose-phosphate rather than glucose is transferred to a growing polyglucose chain (Tagliabracci et al. (2011) Cell Metab13, 274-282). We proposed a mechanism to account for phosphorylation at C2 and possibly at C3. Our results have since been challenged (Nitschke et al. (2013) Cell Metab17, 756-767). Here we extend the evidence supporting our conclusion, validating the assay used for the detection of glycogen phosphorylation, measurement of the transfer of (32)P from [ß-(32)P]UDP-glucose to glycogen by glycogen synthase. The (32)P associated with the glycogen fraction was stable to ethanol precipitation, SDS-PAGE and gel filtration on Sephadex G50. The (32)P-signal was not affected by inclusion of excess unlabeled UDP before analysis or by treatment with a UDPase, arguing against the signal being due to contaminating [ß-(32)P]UDP generated in the reaction. Furthermore, [(32)P]UDP did not bind non-covalently to glycogen. The (32)P associated with glycogen was released by laforin treatment, suggesting that it was present as a phosphomonoester. The conclusion is that glycogen synthase can mediate the introduction of phosphate into glycogen, thereby providing a possible mechanism for C2, and perhaps C3, phosphorylation.

Discovery of a series of aromatic lactones as ALDH1/2-directed inhibitors.

In humans, the aldehyde dehydrogenase superfamily consists of 19 isoenzymes which mostly catalyze the NAD(P)(+)-dependent oxidation of aldehydes. Many of these isoenzymes have overlapping substrate specificities and therefore their potential physiological functions may overlap. Thus the development of new isoenzyme-selective probes would be able to better delineate the function of a single isoenzyme and its individual contribution to the metabolism of a particular substrate. This specific study was designed to find a novel modulator of ALDH2, a mitochondrial ALDH isoenzyme most well-known for its role in acetaldehyde oxidation. 53 compounds were initially identified to modulate the activity of ALDH2 by a high-throughput esterase screen from a library of 63,000 compounds. Of these initial 53 compounds, 12 were found to also modulate the oxidation of propionaldehyde by ALDH2. Single concentration measurements at 10µM compound were performed using ALDH1A1, ALDH1A2, ALDH1A3, ALDH2, ALDH1B1, ALDH3A1, ALDH4A1, and/or ALDH5A1 to determine the selectivity of these 12 compounds toward ALDH2. Four of the twelve compounds shared an aromatic lactone structure and were found to be potent inhibitors of the ALDH1/2 isoenzymes, but have no inhibitory effect on ALDH3A1, ALDH4A1 or ALDH5A1. Two of the aromatic lactones show selectivity within the ALDH1/2 class, and one appears to be selective for ALDH2 compared to all other isoenzymes tested.