A prevalent variant in PPP1R3A impairs glycogen synthesis and reduces muscle glycogen content in humans and mice

Abstract

Background: Stored glycogen is an important source of energy for skeletal muscle. Human genetic disorders primarily affecting skeletal muscle glycogen turnover are well-recognised, but rare. We previously reported that a frameshift/premature stop mutation in PPP1R3A, the gene encoding RGL, a key regulator of muscle glycogen metabolism, was present in 1.36% of participants from a population of white individuals in the UK. However, the functional implications of the mutation were not known. The objective of this study was to characterise the molecular and physiological consequences of this genetic variant.

Methods and findings: In this study we found a similar prevalence of the variant in an independent UK white population of 744 participants (1.46%) and, using in vivo (13)C magnetic resonance spectroscopy studies, demonstrate that human carriers (n = 6) of the variant have low basal (65% lower, p = 0.002) and postprandial muscle glycogen levels. Mice engineered to express the equivalent mutation had similarly decreased muscle glycogen levels (40% lower in heterozygous knock-in mice, p < 0.05). In muscle tissue from these mice, failure of the truncated mutant to bind glycogen and colocalize with glycogen synthase (GS) decreased GS and increased glycogen phosphorylase activity states, which account for the decreased glycogen content.

Conclusions: Thus, PPP1R3A C1984DeltaAG (stop codon 668) is, to our knowledge, the first prevalent mutation described that directly impairs glycogen synthesis and decreases glycogen levels in human skeletal muscle. The fact that it is present in approximately 1 in 70 UK whites increases the potential biomedical relevance of these observations.

Figure 1. Human PPP1R3A FS Carrier Phenotype

(A) Muscle glycogen, (B) blood glucose, and (C) insulin levels in humans with the PPP1R3A FS mutation before and after two standardised meals (given at 60 and 240 min). Data from PPP1R3A FS carriers (n = 4; open circles) are compared with those from volunteers without diabetes (n = 9; solid squares) and from participants with type 2 diabetes (T2DM; n = 9; filled triangles). Mean data from two digenic participants (double heterozygotes for the PPP1R3A FS and PPARG FS) are included in graphs (A) and (B). Mean fasting insulin levels from the two digenic individuals were 199, 5,744, 9,534, 1,303, 13,689, and 10,291 pmol/l, too high to include in (C), at 0, 60, 120, 240, 360, and 480 minutes, respectively.

Figure 2. Muscle Glycogen Metabolism in R_GL Knock-in and Wild-Type Mice

(A) R_GL Western blots of skeletal muscle extracts from WT, R_GL knock-in heterozygous (R_GL kin het), and R_GL knock-in homozygous (R_GL kin hom) mice. The predicted molecular weight (MW) of the truncated R_GL (R_GL trunc, 643 amino acids) is 72,000 but it migrates on SDS polyacrylamide gel electrophoresis with an apparent MW of 83,000. (B) GS activity was assayed in the absence (−) or presence (+) of G6P in extracts of skeletal muscle from WT, R_GL knock-in heterozygous (R_GL kin het), and R_GL knock-in homozygous (R_GL kin hom) mice not injected or injected intraperitoneally with 5 mU/g insulin for 10 min (plain and chequered bars, respectively). (C) Total GS activity (mU/mg) measured in the presence of 7.2 mM G6P. (D) Glycogen phosphorylase (Ph) activity was assayed in the absence (−) or presence (+) of 2 mM AMP. (E) Total glycogen phosphorylase (Ph) activity (U/mg) measured in the presence of 2 mM AMP. (F) Glycogen content in muscle. n = 4–9 per group; * p < 0.05 versus WT basal; # p < 0.05 insulin versus basal.

Figure 3. Interactions between WT R_GL, R_GL Knock-In Truncated Mutant (R_GL trunc), and GS in Skeletal Muscle Extracts from WT, R_GL Knock-In Heterozygous (R_GL kin het) and R_GL Knock-In Homozygous (R_GL kin hom) Mice

(A) Western blots of protein phosphatase-1 catalytic subunit delta (PP1cδ) and GS in muscle extracts. (B) Western blots for R_GL and GS following R_GL immunoprecipitation with a R_GL N-terminal antibody (Ab). (C) Western blots for R_GL and GS following R_GL immunoprecipitation with a R_GL C-terminal antibody (Ab).

Figure 4. Glycogen Synthase Pull Down

GS was pulled down with GST-glycogenin—GST-GN(297–333)—fusion protein before and after α-amylase digestion of glycogen in skeletal muscle extracts from wild type (R_GL WT) and R_GL knock-in homozygous (R_GL kin hom) mice. (A) GST-GN(297–333) pulls down almost all GS, but (B) neither the WT nor the truncated R_GL (R_GL trunc) was pulled down by GST-GN(297–333).

Figure 5. R_GL Western Blot Analysis of Skeletal Muscle Extracts

Western blotting of high-speed supernatants and pellets of skeletal muscle extracts from WT, R_GL knock-in heterozygous (het), and R_GL knock-in homozygous (hom) mice. High speed (HS) ultracentrifugation at 100,0000g for 90 min was used to pellet glycogen. Western blots for R_GL, GS, and GP were then performed on the supernatant and pellet fractions.

Figure 6. Glucose Tolerance and Insulin Sensitivity in R_GL Knock-in Mice

(A) Glucose tolerance following intraperitoneal glucose (2 mg/g body weight) administration in WT, R_GL knock-in heterozygous (R_GL kin het), and R_GL kin homozygous (R_GL kin hom) mice. (B) Plasma glucose response to intraperitoneal insulin (0.75 mU/g body weight) in WT-, R_GL knock-in heterozygous (R_GL kin het)-, and R_GL knock-in homozygous (R_GL kin hom) mice. Peripheral and hepatic insulin sensitivity were assessed by means of hyperinsulinemic-euglycemic clamps in WT and R_GL kin het mice (C–G). Glucose infusion rates (C); peripheral glucose turnover (D); and basal (E) and suppressed (F) endogenous glucose production (EGP) during hyperinsulinemic-euglycemic clamps. (G) Whole body (WB) glycolysis and glycogen synthesis were measured during the clamps. Data are expressed as mean values ± SEM for 6–9 mice per treatment group.