G6PD protects from oxidative damage and improves healthspan in mice

Abstract

Reactive oxygen species (ROS) are constantly generated by cells and ROS-derived damage contributes to ageing. Protection against oxidative damage largely relies on the reductive power of NAPDH, whose levels are mostly determined by the enzyme glucose-6-phosphate dehydrogenase (G6PD). Here, we report a transgenic mouse model with moderate overexpression of human G6PD under its endogenous promoter. Importantly, G6PD-Tg mice have higher levels of NADPH, lower levels of ROS-derived damage, and better protection from ageing-associated functional decline, including extended median lifespan in females. The G6PD transgene has no effect on tumour development, even after combining with various tumour-prone genetic alterations. We conclude that a modest increase in G6PD activity is beneficial for healthspan through increased NADPH levels and protection from the deleterious effects of ROS.

Figure 1. Characterization of the G6PD-Tg mice.

(a) Relative G6PD mRNA of the indicated tissues from mice 10–12 weeks old, one male and two females for the WT group, one male (M) and five females (F) in the Tg group. (b) Levels of G6PD protein in the indicated tissues from 3 to 4-months-old mice and mouse embryo fibroblasts (MEFs) where each lane corresponds to one individual animal for the tissues (WAT: white adipose tissue; RBC: red blood cells), or an independent MEF preparation for the MEFs. The anti-G6PD antibody reacts with both the mouse and human G6PD protein. In some tissues (liver, heart), a second upper band was also detected. ACTIN or GAPDH were used as loading controls. (c) G6PD enzymatic activity in the indicated samples. (d) Levels of NADPH in the indicated tissues. (e) Uric acid concentration in plasma. (f) Lactate production in isolated erythrocytes (left panel) and total levels of lactate in plasma (right panel). (g,h) Percentage of cell survival in G6PD-Tg and WT primary MEFs upon treatment with the oxidative stress inducers diamide (g) and paraquat (h). (i) Survival curve of mice treated with a lethal dose of paraquat. Bars represent the mean±s.e.m. Statistic significance was assessed using the two-tailed Student's t-test for all experiments, with the exception of i, where the logrank test was applied: *P<0.05; **P<0.01; ***P<0.001.

Figure 2. G6PD-Tg mice display improved healthspan.

(a,b) Survival curves of female (a) and male (b) mice of the indicated genotypes. (c) Weight of males during the indicated time. (d,e) GTT (d) and ITT (e) on 1–year-old male mice. Inset, area under the curve (AUC). (f) Rotarod test on 1.5–2-year-old mice. Bars and dots represent average±s.e.m. Statistical significance was assessed using two-tailed Student's t-test with the exception of panels a and b, where logrank test was used: *P<0.05.

Figure 3. Cancer susceptibility is not affected in G6PD-Tg mice.

(a) Growth curves of primary MEFs at the indicated serum concentrations. (b,c) Number of neoplastic foci in MEFs transduced with E1a/HRas^GV12 (b) or with E6/HRas^G12V (c) after 2 weeks. (d) Fold change of reprogramming efficiency of primary MEFs after transduction with Oct4, Sox2 and Klf4. (e–i) Survival curves of the indicated mouse lines and experiments. (e) p53-KO;G6PD-WT, n=21, 7 females (F); 14 males (M). p53-KO;G6PD-Tg, n=32, 12F, 20M. (f) ATM-KO;G6PD-WT, n=7, 3F, 4M; ATM-KO;G6PD-Tg, n=16, 6F, 10M. (g) MMTV-PyMT-Tg;G6PD-WT, n=7F; MMTV-PyMT-Tg;G6PD-Tg, n=10F. (h) Eμ-Myc-Tg;G6PD-WT, n=6, 3F, 3M; Eμ-Myc-Tg;G6PD-Tg, n=9, 4F, 5M. (i) Survival curve of mice injected intramuscularly with 3MC, G6PD-WT n=9, 1F, 8M; G6PD-Tg n=8M. Bars represent mean±s.e.m. Statistical significance was determined by the two-tailed Student's t-test for a–d, and by the logrank test for e–i. None of the differences was statistically significant.

Figure 4. Overexpression of G6PD in vivo protects from oxidative damage.

(a,b) G6PD mRNA expression levels (a) and G6PD activity (b) in liver (left panels) and brain (right panels) of neonates, young (10–35 weeks) and old (125–143 weeks) female or male mice. For RNA measures: neonates, n=5 for all groups. Liver: young WT females, n=11; young Tg females, n=13; old WT females, n=22; old Tg females, n=22; young WT males, n=12; young Tg males, n=13; old WT males, n=13; old Tg males, n=11. Brain: young WT females, n=3; young Tg females, n=6; old WT females, n=15; old Tg females, n=15; young WT males, n=4; young Tg males, n=6; old WT males, n=6; old Tg males, n=5. For G6PD activity: neonates, n=4 for all groups. Liver: young WT females, n=8; young Tg females, n=8; old WT females, n=4; old Tg females, n=5; young WT males, n=4; young Tg males, n=4; old WT males, n=5; old Tg males, n=10. Brain: young WT females, n=3; young Tg females, n=4; old WT females, n=4; old Tg females, n=5; young WT males, n=4; young Tg males, n=4; old WT males, n=5; old Tg males, n=4. (c,d) DNA oxidation (8-OHdG adducts) in liver (c) or brain (d). (e) Lipid oxidation (MDA adducts) in liver. (f–h) GSH:GSSG ratio (f) and GSH (g) or GSSG (h) concentrations were assayed in liver samples from 2-year-old female mice. Bars represent mean±s.e.m. Statistical significance was assayed using the two-tailed Student's t-test: *P<0.05; **P<0.01; ***P<0.001.