A mutation in the dynein heavy chain gene compensates for energy deficit of mutant SOD1 mice and increases potentially neuroprotective IGF-1

Abstract

Background: Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease characterized by a progressive loss of motor neurons. ALS patients, as well as animal models such as mice overexpressing mutant SOD1s, are characterized by increased energy expenditure. In mice, this hypermetabolism leads to energy deficit and precipitates motor neuron degeneration. Recent studies have shown that mutations in the gene encoding the dynein heavy chain protein are able to extend lifespan of mutant SOD1 mice. It remains unknown whether the protection offered by these dynein mutations relies on a compensation of energy metabolism defects.

Results: SOD1(G93A) mice were crossbred with mice harboring the dynein mutant Cramping allele (Cra/+ mice). Dynein mutation increased adipose stores in compound transgenic mice through increasing carbohydrate oxidation and sparing lipids. Metabolic changes that occurred in double transgenic mice were accompanied by the normalization of the expression of key mRNAs in the white adipose tissue and liver. Furthermore, Dynein Cra mutation rescued decreased post-prandial plasma triglycerides and decreased non esterified fatty acids upon fasting. In SOD1(G93A) mice, the dynein Cra mutation led to increased expression of IGF-1 in the liver, increased systemic IGF-1 and, most importantly, to increased spinal IGF-1 levels that are potentially neuroprotective.

Conclusions: These findings suggest that the protection against SOD1(G93A) offered by the Cramping mutation in the dynein gene is, at least partially, mediated by a reversal in energy deficit and increased IGF-1 availability to motor neurons.

Figure 1

Dynein mutation increased adipose stores in early symptomatic SOD1(G93A) mice. A- mRNA levels of alpha subunit of the nicotinic acetylcholine receptor (AchRα) in gastrocnemius muscles of wild type (+/+) and dynein mutant mice (Cra/+) bearing SOD1(G93A) transgene (SOD1m, black columns) or not (Wt, empty columns) *P < 0.05 versus Wt. mRNA levels were standardized using 18S ribosomal RNA as a control. N = 9 mice per group. B-C- Relative weight of epididimary (EPI) and retroperitoneal (RP) white adipose tissue fat pad with regard to body weight (B). The panel C shows absolute weight (in mg) of EPI and RP in the same mice than in A. *P < 0.05 versus Wt; #, p < 0.05 as indicated. N = 9 mice per group.

Figure 2

Dynein mutation does not decrease global hypermetabolism but shifts energy metabolism of SOD1(G93A) mice towards carbohydrate use. A- Total (left panel) and resting (right panel) energy expenditure of wild type (+/+) and dynein mutant mice (Cra/+) bearing SOD1(G93A) transgene (SOD1m, black columns) or not (Wt, empty columns) *P < 0.05 versus Wt. Note that SOD1(G93A) mice are hypermetabolic and that the dynein mutant genotype has no effect on this hypermetabolism. N = 9 mice per group. B- Energy expenditure of the same mice than in A as a function of time. Mice bearing a Cra dynein mutation are denoted by empty symbols, and their corresponding controls by a filled symbol. Mice bearing a SOD1(G93A) (SOD1m) transgene are labeled by a squared symbol and their controls by a circle. Note that in the diurnal period, there are no differences between +/+ and Cra/+ mice but that the presence of a SOD1(G93A) transgene increases energy expenditure. On the contrary, during nocturnal activity period, the presence of either Cra/+ mutation or SOD1(G93A) transgene increases energy expenditure. N = 9 mice per group. C-D- Respiratory quotient of the same mice than in A. The four groups of mice are shown in two graphs for clarity reasons. Mice non transgenic for SOD1(G93A) (SOD1m) are shown in panel C (filled symbols, +/+; empty symbols, Cra/+). Mice transgenic for SOD1(G93A) are shown in panel D (filled symbols, +/+; empty symbols, Cra/+). Note that the presence of a dynein mutation increases respiratory quotient in otherwise wild type mice, and even more potently in SOD1(G93A) mice.

Figure 3

Dynein mutation reverts the systemic and molecular alterations associated with energy deficit in SOD1(G93A) mice. A- mRNA levels of lipoprotein lipase (LPL) in the epididimary white fat pad of wild type (+/+) and dynein mutant mice (Cra/+) bearing SOD1(G93A) transgene (SOD1m, black columns) or not (Wt, empty columns). *P < 0.05 versus Wt; #, p < 0.05 as indicated. Note that the SOD1(G93A)-associated decreased expression of LPL in the EPI is rescued by the dynein mutation in compound SOD1(G93A)/Cra mice. N = 9 mice per group. B- mRNA levels of fatty acid synthase (FAS), carnithine palmitoyl transferase 1A (liver form, CPT1A), Medium chain acyl CoA dehydrogenase (MCAD), peroxisome-proliferator activated receptor alpha (PPARα), peroxisome-proliferator activated receptor gamma co-activator 1 alpha (PGC1α), phosphoenolpyruvate carboxykinase (PEPCK) and peroxisome-proliferator activated receptor gamma (PPARγ) in the liver of the same mice than in A. *P < 0.05 versus Wt; #, p < 0.05 as indicated. Note that the SOD1(G93A)-associated decreased expression of FAS in the liver is rescued by the dynein mutation in compound SOD1(G93A)/Cra mice. N = 9 mice per group. C- Circulating triglycerides levels in the same mice than in A either in fed (right) or fasted conditions (left). Note that the SOD1(G93A) transgene leads to decreased fed triglycerides levels and that this is partially reverted in compound SOD1(G93A)/Cra mice. N = 9 mice per group. D- Circulating non-esterified fatty acids (NEFAs) levels in the same mice than in A in fasted conditions. Note that the SOD1(G93A) transgene leads to decreased fasted NEFAs levels and that this is fully reverted in compound SOD1(G93A)/Cra mice. N = 9 mice per group. E- mRNA levels of carnithine palmitoyl transferase 1B (muscle form, CPT1A), Medium chain acyl CoA dehydrogenase (MCAD), peroxisome-proliferator activated receptor gamma (PPARγ) and peroxisome-proliferator activated receptor gamma co-activator 1 alpha (PGC1α) in the gastrocnemius muscle of the same mice than in A. *P < 0.05 versus Wt. N = 9 mice per group.

Figure 4

Dynein mutation increases liver IGF-1 expression. A- mRNA levels of Insulin-like growth factor (IGF-1) in the liver of wild type (+/+) and dynein mutant mice (Cra/+) bearing SOD1(G93A) transgene (SOD1m, black columns) or not (Wt, empty columns) *P < 0.05 versus Wt; #, p < 0.05 as indicated. N = 9 mice per group. B- mRNA levels of Insulin-like growth factor (IGF-1) and its muscle specific splice variant mechano-growth factor (MGF) in the gastrocnemius muscle of wild type (+/+) and dynein mutant mice (Cra/+) bearing SOD1(G93A) transgene (SOD1m, black columns) or not (Wt, empty columns). N = 9 mice per group. C- Serum IGF-1 levels in wild type (+/+) and dynein mutant mice (Cra/+) bearing SOD1(G93A) transgene (SOD1m, black columns) or not (Wt, empty columns) #, p < 0.05 as indicated. Note that circulating IGF-1 levels are increased in dynein mutant mice and that this increase is abolished in compound SOD1(G93A)/Cra mice. N = 9 mice per group.

Figure 5

Dynein mutation increases potentially neuroprotective IGF-1. A- Spinal cord IGF-1 levels in wild type (+/+) and dynein mutant mice (Cra/+) bearing SOD1(G93A) transgene (SOD1m, black columns) or not (Wt, empty columns) #, p < 0.05 as indicated. Note that spinal cord IGF-1 levels are increased in dynein mutant mice and that this increase is stronger in compound SOD1(G93A)/Cra mice. N = 9 mice per group. B- mRNA levels of Insulin-like growth factor (IGF-1) in the spinal cord of wild type (+/+) and dynein mutant mice (Cra/+) bearing SOD1(G93A) transgene (SOD1m, black columns) or not (Wt, empty columns). N = 9 mice per group. C- mRNA levels of matrix metallo proteinase 9 (MMP9) in the spinal cord of wild type (+/+) and dynein mutant mice (Cra/+) bearing SOD1(G93A) transgene (SOD1m, black columns) or not (Wt, empty columns). *P < 0.05 versus Wt; #, p < 0.05 as indicated. N = 9 mice per group.

Figure 6

a working model for dynein mutant protection against SOD1-ALS. We propose that dynein mutations provide neuroprotection by converging paths. First, degeneration of proprioceptive neurons (PN) modify network activity (1) by decreasing direct glutamatergic input to motor neurons (MN), but also to inhibitory interneurons (IN). Second, CNS-linked hyperactivity (2) might on its own lead to increased motoneuronal activity, and MMP9 activation. Last, by modifying energy metabolism and favoring carbohydrate oxidation over lipid oxidation (3), dynein mutation reverts the energy deficit of SOD1(G93A) mice, thereby increasing IGF-1 production in the liver. This increased IGF-1 production is able to enter the CNS due to MMP9 activation. Other circulating factors might also increase IGF-1 supply to motor neurons.