Caloric restriction shortens lifespan through an increase in lipid peroxidation, inflammation and apoptosis in the G93A mouse, an animal model of ALS

Abstract

Caloric restriction (CR) extends lifespan through a reduction in oxidative stress, delays the onset of morbidity and prolongs lifespan. We previously reported that long-term CR hastened clinical onset, disease progression and shortened lifespan, while transiently improving motor performance in G93A mice, a model of amyotrophic lateral sclerosis (ALS) that shows increased free radical production. To investigate the long-term CR-induced pathology in G93A mice, we assessed the mitochondrial bioenergetic efficiency and oxidative capacity (CS--citrate synthase content and activity, cytochrome c oxidase--COX activity and protein content of COX subunit-I and IV and UCP3-uncoupling protein 3), oxidative damage (MDA--malondialdehyde and PC--protein carbonyls), antioxidant enzyme capacity (Mn-SOD, Cu/Zn-SOD and catalase), inflammation (TNF-alpha), stress response (Hsp70) and markers of apoptosis (Bax, Bcl-2, caspase 9, cleaved caspase 9) in their skeletal muscle. At age 40 days, G93A mice were divided into two groups: Ad libitum (AL; n = 14; 7 females) or CR (n = 13; 6 females), with a diet equal to 60% of AL. COX/CS enzyme activity was lower in CR vs. AL male quadriceps (35%), despite a 2.3-fold higher COX-IV/CS protein content. UCP3 was higher in CR vs. AL females only. MnSOD and Cu/Zn-SOD were higher in CR vs. AL mice and CR vs. AL females. MDA was higher (83%) in CR vs. AL red gastrocnemius. Conversely, PC was lower in CR vs. AL red (62%) and white (30%) gastrocnemius. TNF-alpha was higher (52%) in CR vs. AL mice and Hsp70 was lower (62%) in CR vs. AL quadriceps. Bax was higher in CR vs. AL mice (41%) and CR vs. AL females (52%). Catalase, Bcl-2 and caspases did not differ. We conclude that CR increases lipid peroxidation, inflammation and apoptosis, while decreasing mitochondrial bioenergetic efficiency, protein oxidation and stress response in G93A mice.

Figure 1. Anthropometric measures in CR vs. AL G93A mice.

(A) Food intake (g), (B) body weight (g) and (C) body condition of 23 ad libitum (AL: •, 16 females; ▪, 7 males) and 31 calorie restricted (CR, 60% of ad libitum: ○, 21 females; □, 10 males) G93A mice. Data are means ± SEM.

Figure 2. Functional measures in CR vs. AL G93A mice.

(A) Ability to move and (B) paw grip endurance (s) of 23 ad libitum (AL: •, 16 females; ▪, 7 males) and 31 calorie restricted (CR, 60% of ad libitum: ○, 21 females; □, 10 males) G93A mice. Data are means ± SEM.

Figure 3. Probablility of clinical onset in CR vs. AL G93A mice.

Probability of clinical onset in 23 ad libitum (AL: •, 16 females; ▪, 7 males) and 31 calorie restricted (CR, 60% of ad libitum: ○, 21 females; □, 10 males) G93A mice. The rate of attaining clinical onset (i.e., the hazard ratio) in the CR mice was 2-fold higher (95% CI: 1.4, 4.7) than the AL mice. CR females had a 2-fold (95% CI: 1.2, 5.3) higher rate of attaining clinical onset as compared with AL females (P = 0.016). There was no sex difference in the rate of attaining clinical onset.

Figure 4. Correlation between clinical score and ability to move and between clinical score and paw grip endurance in CR vs. AL G93A mice.

(A) Relation between clinical score (from clinical onset to euthanasia) and ability to move and (B) between clinical score and paw grip endurance (s) in 23 ad libitum (AL: •, 16 females; ▪, 7 males) and 31 calorie restricted (CR, 60% of ad libitum: ○, 21 females; □, 10 males) G93A mice. For ability to move, slopes and intercepts were not different with the following pooled equation: ability to move  =  (−1.4) x clinical score + (7.1). For paw grip endurance, data for the females followed a sigmoidal relationship (AL females, r² = 0.990; CR females, r² = 0.995; curves were significantly different, P<0.0001) whereas data for the males followed a linear relationship (slopes were significantly different, P = 0.0011). For AL males (r = −0.984, P<0.0001), paw grip endurance  =  (−27.8±1.1) x clinical score + (133.5±3.8); for CR males (r = −0.995, P<0.0001), paw grip endurance  =  (−31.8±0.7) x clinical score + (159.4±2.9), with the data presented as means ± SD. Data are means of each group on the same day.

Figure 5. Probability of survival in CR vs. AL G93A mice.

Probability of survival using as endpoint a clinical score of (A) 4 and (B) 5 in 23 ad libitum (AL: •, 16 females; ▪, 7 males) and 31 calorie restricted (CR, 60% of ad libitum: ○, 21 females; □, 10 males) G93A mice. The rate of reaching endpoint in the CR mice (i.e., the hazard ratio) was 3.1-fold higher (95% CI: 2.9, 10.7) than the AL mice. The rate of reaching endpoint was 3.1-fold (95% CI: 2.6, 13.5) higher in the CR vs. AL females (P<0.0001) and 3.4-fold (95% CI: 1.9, 20.7) higher in the CR vs. AL males (P = 0.003).

Figure 6. Mitochondrial energetics in CR vs. AL G93A mice.

(A) COX to citrate synthase activity ratio was lower in the quadriceps of CR vs. AL G93A male mice (34%, P = 0.034). (B) Cytochrome c oxidase subunit-IV to citrate synthase protein content ratio was higher in CR vs. AL G93A male mice (2.3-fold, P = 0.059). (C) UCP3 was higher in CR vs. AL female quadriceps (1.5-fold, P = 0.042). Data are presented as means ± SEM. n = 27; AL, 7 males and 7 females; CR, 7 males and 6 females. Asterisks denote significant changes (P≤0.05 vs. AL); dagger denotes strong trend (0.05<P≤0.1 vs. AL male).

Figure 7. Lipid peroxidation in CR vs. AL G93A mice.

(A) MDA was higher in CR vs. AL (83%, P<0.001; main effect of diet), and in female vs. male (32%, P = 0.045; main effect of sex) red gastrocnemius. Within each sex, MDA was higher in CR vs. AL females (2-fold, P = 0.003), and in CR vs. AL males (56%, P = 0.005). (B) MDA was higher in CR vs. AL white gastrocnemius (14%, P = 0.095; main effect of diet), with no sex differences. Data are presented as means ± SEM. Asterisks denote significant changes (P≤0.05 vs. AL); dagger denotes strong trend (0.05<P≤0.1 vs. AL).

Figure 8. Potein oxidation in CR vs. AL G93A mice.

(A) PC was lower in CR vs. AL red gastrocnemius (62%, P<0.001; main effect of diet). Within each sex, PC was lower in CR vs. AL females (50%, P = 0.012), and in CR vs. AL males (74%, P<0.001). (B) PC was lower in CR vs. AL white gastrocnemius (30%, P = 0.003; main effect of diet). PC was lower in CR vs. AL females (43%, P<0.001). Data are presented as means ± SEM. n = 27; AL, 7 males and 7 females; CR, 7 males and 6 females. Asterisks denote significant changes (P≤0.05 vs. AL).

Figure 9. SOD content in CR vs. AL G93A mice.

(A) MnSOD was higher in CR vs. AL (3-fold, P = 0.031; main effect of diet), and in female vs. male (2.5-fold, P = 0.026; main effect of sex) red gastrocnemius. MnSOD was significantly elevated in CR vs. AL females (4-fold, P = 0.015), but not in CR vs. AL males. (B) MnSOD was higher in CR vs. AL white gastrocnemius (78%, P = 0.062; main effect of diet). MnSOD was higher in CR vs. AL females (2.2-fold, P = 0.038). (C) Cu/Zn-SOD was higher in CR vs. AL red gastrocnemius (67%, P = 0.096; main effect of diet). Cu/Zn-SOD was significantly higher in CR vs. AL females (2.6-fold, P = 0.020). (D) There was no change in Cu/Zn-SOD protein content in white gastrocnemius. Data are presented as means ± SEM. n = 27; AL, 7 males and 7 females; CR, 7 males and 6 females. Asterisks denote significant changes (P≤0.05 vs. AL); dagger denotes strong trend (0.05<P≤0.1 vs. AL); double-dagger denotes significant changes (P≤0.05 vs. AL females).

Figure 10. Catalase activity in CR vs. AL G93A mice.

Catalase enzyme activity remained unchanged in the white gastrocnemius of G93A mice. Data are presented as means ± SEM. n = 27; AL, 7 males and 7 females; CR, 7 males and 6 females.

Figure 11. TNF-α content in CR vs. AL G93A mice.

TNF-α was higher in the quadriceps of CR vs. AL mice (52%, P = 0.030; main effect of diet). Data are presented as means ± SEM. n = 27; AL, 7 males and 7 females; CR, 7 males and 6 females. Asterisks denote significant changes (P≤0.05 vs. AL).

Figure 12. Hsp70 content in CR vs. AL G93A mice.

Hsp70 was lower in the quadriceps of CR vs. AL mice (62%, P = 0.002; main effect of diet), and in males vs. females (37%, P = 0.030; main effect of sex). Within each sex, Hsp70 was significantly decreased in CR vs. AL females (64%, P<0.001) and in CR vs. AL males (59%, P = 0.039). Data are presented as means ± SEM. n = 27; AL, 7 males and 7 females; CR, 7 males and 6 females. Asterisks denote significant changes (P≤0.05 vs. AL).

Figure 13. Apoptosis regulatory proteins content in the quadriceps of CR vs. AL G93A mice.

(A) Bax was significantly higher in CR vs. AL mice (41%, P = 0.027; main effect of diet). Bax was increased in CR vs. AL females (52%, P = 0.048). (B) There was no significant difference in Bcl-2 protein content. (C) Bax/Bcl-2 was significantly higher in CR vs. AL mice (68%, P = 0.040; main effect of diet). Bax/Bcl-2 was increased in CR vs. AL females (2.3-fold, P = 0.029). No significant changes were observed in the protein content of: (D) caspase 9, (E) cleaved caspase 9, and (F) the ratio of cleaved caspase 9/caspase 9 for diet or sex. Data are presented as means ± SEM. n = 27; AL, 7 males and 7 females; CR, 7 males and 6 females. Asterisks denote significant changes (P≤0.05 vs. AL).