Neuronal glycogen synthesis contributes to physiological aging

Abstract

Glycogen is a branched polymer of glucose and the carbohydrate energy store for animal cells. In the brain, it is essentially found in glial cells, although it is also present in minute amounts in neurons. In humans, loss-of-function mutations in laforin and malin, proteins involved in suppressing glycogen synthesis, induce the presence of high numbers of insoluble polyglucosan bodies in neuronal cells. Known as Lafora bodies (LBs), these deposits result in the aggressive neurodegeneration seen in Lafora's disease. Polysaccharide-based aggregates, called corpora amylacea (CA), are also present in the neurons of aged human brains. Despite the similarity of CA to LBs, the mechanisms and functional consequences of CA formation are yet unknown. Here, we show that wild-type laboratory mice also accumulate glycogen-based aggregates in the brain as they age. These structures are immunopositive for an array of metabolic and stress-response proteins, some of which were previously shown to aggregate in correlation with age in the human brain and are also present in LBs. Remarkably, these structures and their associated protein aggregates are not present in the aged mouse brain upon genetic ablation of glycogen synthase. Similar genetic intervention in Drosophila prevents the accumulation of glycogen clusters in the neuronal processes of aged flies. Most interestingly, targeted reduction of Drosophila glycogen synthase in neurons improves neurological function with age and extends lifespan. These results demonstrate that neuronal glycogen accumulation contributes to physiological aging and may therefore constitute a key factor regulating age-related neurological decline in humans.

Keywords: Drosophila; aging; corpora amylacea; glycogen; protein aggregation; stress response.

Figure 1

Age-dependent PG accumulation is GS-dependent. (A) 30-μm-thick (see Experimental procedures for details) brain (hippocampus) sections of 1-, 3-, 6-, 16- and 24-month-old wild-type mice (n = 3) were immunostained for MGS (brown). Scale bars = 100 μm. Arrows point to MGS accumulations. (B, C) Brain (hippocampus) sections of 16-month-old wild-type (A) and MGS^KO (B) mice. Examples of 4-μm-thick (see Experimental procedures for details) consecutive brain sections stained with periodic acid–Schiff (PAS) and immunostained for MGS (brown) as indicated (n = 3). All the sections showed in (B, C) were counterstained with hematoxylin (blue). Arrows point to PGBs, and arrowhead to an astrocyte showing normal MGS expression. Scale bars = 200 μm. Second and fourth row panels show magnifications of the squared fields.

Figure 2

Consequences of MGS knockout on PGB accumulation during normal aging. (A) CA markers accumulated with polyglucosan bodies (PGBs) in aged (16-month-old) wild-type and malin^KO mice. Aged (16-month-old) MGS^KO mice did not show these accumulations. Hippocampus sections from wild-type, malin^KO and MGS^KO mice are shown (n = 3). 4-μm-thick sections (see Experimental procedures for details) consecutive to those stained with periodic acid–Schiff (PAS, not shown for clarity) were stained with iodine (purple) or immunostained with antibodies (brown) against laforin, ubiquitin, 70-kDa heat-shock protein (HSP70), advanced glycation end products (AGEP), alpha-synuclein, or parvalbumin. Immunostained sections were counterstained with hematoxylin (light blue). Scale bar = 25 μm. Laforin cellular localization appeared to be mainly nuclear in the absence of glucose polymers in MGS^KO brains. (B, C) PGBs in aged (16-month-old) wild-type and malin^KO mice accumulated glycogen phosphorylase (BGP) and were associated with astrocytes. Confocal images are shown for brains of 16-month-old wild-type and malin^KO mice. Antibodies were used against polyglucosan (red, B), brain glycogen phosphorylase (BGP, magenta, C), and glial fibrillary acidic protein (GFAP, green). Hoechst (blue) was used for nuclear staining. Scale bar = 10 μm. Hippocampus sections are shown in A–C.

Figure 3

A role for GS in age-related PG accumulation in the Drosophila central nervous system.(A–H) Transmission electron microscopy (EM) analysis (×26 500) of optic lobe neuropils of young (5 days, A, D, E, H) and old (31 days, B, C, F, G) flies belonging to the following genotypes: elav-Gal4/w¹¹¹⁸ (Control, A, B, E, F), elav-Gal4/+; UAS-GS-RNAi-III/+ (e>GS-RNAi, C, G), and elav-Gal4/+; UAS-MGS/+ (e>MGS, D, H). Red arrows point to glycogen granules and blue arrowheads to glycogen clusters. CP (capitate projection), M (mitochondrion near to glycogen), S (postsynaptic density), R (ribosomes in glial cell processes). Scale bar = 500 nm, inset 200 = nm. A–D, males, and E-H, females.

Figure 4

GS contributes to age-dependent PG accumulation in Drosophila brains. (A, B) Quantification of glycogen granules (A) and clusters (B) from TEM images. Young (5 days) and old (31 days) fly groups are shown in blue and red, respectively. (C) Quantification of maximum glycogen granule diameter as an average of the 10 largest granules identified for each group. Significance is versus aged (31d) group. In A–C, n = X,Y,Z where X = animals, Y = images, Z = neuronal processes. Control (young) n = 470 436; control (old) n = 5116 676; dGS-RNAi-III (old) n = 5144 1006; dGS-RNAi X (old) n = 390 730, MGS (young) n = 344 231. (D) Determination of dGS expression levels in young (5 days) dGS-RNAi animals by Western blot from whole head homogenates. Western blot membranes were exposed to anti-hMGS antibody (3886 clone, Cell Signaling, 1/1000). tub – tubulin loading control. Quantification in A–C is from pooled data of 2 males and 2 females (young, control), 2 male and 3 females (old, control), 4 females and 1 male (old, dGS-RNAi-III), 4 females (old, dGS-RNAi X), and 2 males and 1 female (young, MGS). (E) Alignment of amino acid sequences in human (Hs), mouse (Mm), and Drosophila (Dm) for comparison of the arginine-rich cluster involved in binding of the allosteric activator glucose-6-phosphate (G6P). Amino acid sequence numbers correspond to human MGS. Essential arginine residues for G6P activation are highlighted in gray. (F) GS activity ratio (−G6P/+G6P). Values were calculated from triplicate activity measurements in a single experiment (n = 60 fly heads, 30 each sex, per genotype). For all panels, e> denotes elav-GAL4 pan-neuronal driver, ′control′ refers to elav-Gal4 > w1118 flies, blue relates to young flies (5d old) and red to aged flies (31d old). Genotypes: elav-Gal4/w¹¹¹⁸ (control), elav-Gal4/+; UAS-dGS-RNAi-III/+ (e>dGS-RNAi-III), elav-Gal4/UAS-GS-RNAi-X (e>dGS RNAi X), elav-Gal4/+; UAS-MGS (e>MGS), and elav-Gal4/+; UAS-dGS-RNAi-NIG-III/+ (e>dGS-RNAi-NIII), elav-Gal4/+; UAS-GFP/+ (e>GFP),. *** P < 0.001, ** P < 0.01, * P < 0.05, ns – not significant, and data are expressed ±SEM.

Figure 5

PAS-coupled silver staining contrasts glycogen granules in aged fly neurons. EM images (×15 000–×21 000) of optic lamina neuropil from ultrathin sections of male fly brains stained by the Thiéry method. Flies were young (5 days, A, D) or old (31 days, B, C) and belonged to the following genotypes: elav-Gal4/w¹¹¹⁸ (Control, A, B), elav-Gal4/+; UAS-dGS-RNAi-III/+ (e>GS-RNAi, C), and elav-Gal4/+; UAS-MGS/+ (e>MGS, D). Scale bar = 500 nm, inset 200 = nm. Red arrows point to glycogen granules and blue arrowheads to glycogen clusters.

Figure 6

Functional consequences of reduced GS in the nervous system of aging Drosophila. (A, B) Kaplan–Meier survival curves of dGS RNAi lines driven with elav for male (A) and female (B) flies. Adult-specific RNAi was achieved by co-expressing gal80^ts and raising animals at 18 °C until adulthood. Genotypes: elav-Gal4/w¹¹¹⁸ (control), elav-Gal4/+; UAS-GS-RNAi-NIG-III/tub-gal80ts (e g80ts>dGSi NIII), elav-Gal4/+; UAS-GS-RNAi-III/ tub-gal80ts (e g80ts>dGSi III), and elav-Gal4/ UAS-GS-RNAi-X; tub-gal80ts/+ (e g80ts>dGSi X). Median lifespans and number of animals tested: (A) control 46 d, n = 120; e g80ts>dGSi III 51 d, n = 120; e g80ts>dGSi NIII 51 d, n = 120; (B) control 51 d, n = 97; e g80ts>dGSi III 51 d, n = 120; e g80ts>dGSi NIII 51 d, n = 120; e g80ts>dGSi X 49 d, n = 120. (C–F) Average (C, D) and maximum (E, F) climbing speed of young (9 d, blue) and old (25 d, red) male flies for dGS RNAi NIII (C, E) and dGS RNAi TRiP (D, F) lines driven with elav-Gal4 in neurons. Genotypes: elav-Gal4/+; UAS-GFP/+ (control, C, E), elav-Gal4/+; UAS-GS-RNAi-NIG-III/+ (e>dGSi NIII, C, E), elav-Gal4/+; UAS-GFPval/+ (control, D, F), elav-Gal4/+; UAS-GS-RNAi-TRiP/+ (e> dGSi TRIP, D, F). ** P < 0.01, *** P > 0.001.