Reduced activity of AMP-activated protein kinase protects against genetic models of motor neuron disease

Abstract

A growing body of research indicates that amyotrophic lateral sclerosis (ALS) patients and mouse models of ALS exhibit metabolic dysfunction. A subpopulation of ALS patients possesses higher levels of resting energy expenditure and lower fat-free mass compared to healthy controls. Similarly, two mutant copper zinc superoxide dismutase 1 (mSOD1) mouse models of familial ALS possess a hypermetabolic phenotype. The pathophysiological relevance of the bioenergetic defects observed in ALS remains largely elusive. AMP-activated protein kinase (AMPK) is a key sensor of cellular energy status and thus might be activated in various models of ALS. Here, we report that AMPK activity is increased in spinal cord cultures expressing mSOD1, as well as in spinal cord lysates from mSOD1 mice. Reducing AMPK activity either pharmacologically or genetically prevents mSOD1-induced motor neuron death in vitro. To investigate the role of AMPK in vivo, we used Caenorhabditis elegans models of motor neuron disease. C. elegans engineered to express human mSOD1 (G85R) in neurons develops locomotor dysfunction and severe fecundity defects when compared to transgenic worms expressing human wild-type SOD1. Genetic reduction of aak-2, the ortholog of the AMPK α2 catalytic subunit in nematodes, improved locomotor behavior and fecundity in G85R animals. Similar observations were made with nematodes engineered to express mutant tat-activating regulatory (TAR) DNA-binding protein of 43 kDa molecular weight. Altogether, these data suggest that bioenergetic abnormalities are likely to be pathophysiologically relevant to motor neuron disease.

Figure 1.

Bioenergetic abnormalities are present in mSOD1 models of motor neuron disease. A, ATP turnover (respiration supporting ATP synthesis) in whole-cell extracts from mixed spinal cord cultures infected for 48 or 72 h with either HSV-WT SOD1 or G37R mSOD1 virus. ^#p ≤ 0.01, significantly different from 48 h time point; *p < 0.01, significantly different from WT SOD1; n = 4–9 dishes per group. B, Representative polarographs of whole-cell extracts from HSV-WT SOD1 and HSV-G37R mSOD1 cultures infected for 48 and 72 h. Numbers under the curves represent the slope of each segment of the trace in nanoatoms Oxygen per minute. C, RCR of spinal cord homogenates from P40 and P90 mice. ^#p < 0.0001, significantly different from P40 time point; *p < 0.01, significantly different from non-Tg controls; n = 4–8 animals per group. D, Representative polarographs in the presence of glutamate (10 mm) and malate (2 mm) in mouse spinal cord homogenates from P40 and P90 G93A mSOD1 mice and non-Tg control counterparts. Numbers under the curves represent the slope of each segment of the trace in nanoatoms Oxygen per minute. Error bars represent ±SEM.

Figure 2.

Increased AMPK activity in mSOD1 models of motor neuron disease. A, Mixed spinal cord cultures infected with either HSV-WT SOD1 or HSV-G37R mSOD1 for 48 h. pAMPK is significantly increased in mSOD1-infected compared to WT SOD1-infected cultures, with no change in overall AMPK levels. The expression of the AMPK downstream target pACC is also increased, with no change in overall ACC levels. Expression of p4EBP1 and pp70 S6 Kinase, downstream targets of AMPK and the mTOR pathway, are decreased with no change in overall protein levels. n = 3–5 dishes per group. B, Quantification of pAMPK/AMPK (*p ≤ 0.003, difference between WT SOD1 and G37R mSOD1, four different experiments) and pACC/ACC levels (*p ≤ 0.02, difference between WT SOD1 and G37R mSOD1). C, No change in pAMPK and pACC levels in P40 G93A mSOD1 mouse spinal cords compared to non-Tg controls. D, However, pAMPK and pACC levels are significantly increased in P90 G93A mSOD1 mouse spinal cords compared to non-Tg controls. The effect was specific to spinal cord (SC) and was found neither in cerebellum (Cer) nor liver. E, Quantification of pAMPK/AMPK (*p < 0.001, difference between WT SOD1 and G93A mSOD1) and pACC/ACC levels (*p < 0.0002, difference between WT SOD1 and G37R mSOD1) from spinal cord of P40 and P90 mice. n = ∼5 animals per group. For B and E, data are presented as percentages compared to controls (either HSV-WT SOD1 or non-Tg controls). Error bars indicate ±SEM.

Figure 3.

PGC-1α and downstream target transcript levels in in vitro and in vivo mSOD1 models of motor neuron disease. A, Mixed spinal cord cultures infected with HSV-G37R show no difference in PGC-1α transcript levels and the levels of the antioxidant enzyme GPx1 compared to controls, HSV-WT SOD1 (n = 4–5 dishes per group, three independent experiments). B, Spinal cords from P120 G93A mice also reveal no differences in the expression levels of PGC-1α and the antioxidant enzymes GPx1 and Sod2 compared to non-Tg controls. C, Hindlimb muscles from P120 G93A mSOD1 animals do not show a significant increase in mRNA levels of PGC-1α and its downstream targets GPx1 and Sod2. For B and C, n = 5–7 animals per group. For A–C, all data are normalized to GAPDH and are presented as percentages compared to controls (either HSV-WT SOD1 or non-Tg controls). Error bars indicate ±SEM.

Figure 4.

Inhibiting AMPK activity in vitro protects against mSOD1-induced toxicity. A, Representative immunoblots of mixed spinal cord cultures treated with 5 μm CC for 48 h. B, Quantification of immunoblots from two independent experiments. The ratios of pAMPK/AMPK and pACC/ACC are significantly reduced in CC-treated cultures compared to vehicle-treated cultures (*p ≤ 0.003 and *p ≤ 0.0001, respectively). C, mSOD1-induced toxicity assays were conducted in 14 DIV mixed spinal cord cultures. Cultures infected with G37R mSOD1 and treated with DMSO show ∼50% survival in SMI-32(+) neurons compared to WT SOD1-infected cultures (p ≤ 0.002). However, mSOD1-infected cultures treated with CC demonstrate 100% survival (*p ≤ 0.01, mSOD1 plus CC compared with mSOD1 plus DMSO, three independent experiments). Furthermore, treatment with AICAR (10 mm) has no effect and also shows decreased survival compared to mSOD1 plus CC (*p ≤ 0.001). n = 6 slides per group, 4–5 fields each. D, Representative immunoblots of mixed spinal cord cultures infected with either an HSV-WT AMPK or dnAMPK construct. E, Quantification of immunoblots from two independent experiments. In dnAMPK-infected compared to WT AMPK-infected cultures, there is a significant decrease in the ratio of pAMPK/AMPK (*p < 0.03) and a decrease in pACC/ACC (*p ≤ 0.05). F, In an mSOD1-induced toxicity assay, cultures infected with mSOD1 and coinfected with Lac-Z show expected decrease in survival rate after infection compared to WT SOD1 plus Lac-Z (p < 0.002). However, mSOD1-infected cultures treated with dnAMPK show close to 100% survival, while cultures infected with WT AMPK show no change in survival (*p ≤ 0.04 mSOD1 plus dnAMPK compared with mSOD1 plus Lac-Z, three independent experiments). Furthermore, cultures infected with WT SOD1 plus Lac-Z, WT AMPK, or dnAPMK show no effect among groups (data not shown). n = 4 slides per group, 4–5 fields each. For B, C, E, and F, data are presented as percentages relative to WT SOD1-infected cultures. Asterisks indicate significance in an ANOVA, followed by post hoc tests when necessary. Error bars indicate ±SEM.

Figure 5.

Placing G85R mSOD1 worms in an AMPK α2 (aak-2) null background improves locomotor behavior. A, In a crawling assay, L4 WT SOD1 worms exhibit decreased locomotor behavior compared to N2 worms (*p < 0.03), while G85R mSOD1 worms exhibit an even more severe locomotor defect compared to WT SOD1 animals (*p < 0.001). L4 G85R;aak-2(ok524) double-mutant worms demonstrate a significant improvement in crawling behavior compared to G85R counterparts (^#p ≤ 0.01), although still impaired compared to aak-2(ok524) controls (*p < 0.01). B, Adult WT SOD1 worms exhibit a trend toward decreased locomotor behavior compared to N2 worms (p = 0.08), while G85R worms still exhibit severely impaired behavior compared to WT SOD1 animals (*p < 0.001). G85R;aak-2 worms also demonstrate significantly improved crawling behavior during adulthood compared to G85R counterparts (^#p < 0.001), although are still impaired compared to aak-2 controls (*p < 0.003). C, In a swimming assay, L4 WT SOD1 show expected locomotor defect compared to N2 worms (*p < 0.002), and G85R worms show an even more severe locomotor defect compared to WT SOD1 worms (*p < 0.0007). L4 G85R;aak-2 animals display improved swimming behavior compared to G85R counterparts (^#p < 0.001), although still performed worse compared to aak-2 controls (p < 0.0001). There is a trend toward decreased swimming activity in WT SOD1;aak-2(ok524) compared to WT SOD1 animals (p = 0.09). D, Like in B, WT SOD1 worms do not exhibit significant locomotor impairment compared to N2 animals; however, G85R worms show a severe locomotor phenotype compared to WT SOD1 worms (*p < 0.003). The significant improvement in swimming behavior of G85R;aak-2 worms compared to G85R counterparts persists until adulthood (^#p < 0.002), although G85R;aak-2 double mutants still performed worse compared to aak-2 controls (p < 0.0001). In contrast, WT SOD1;aak-2 worms exhibit significantly decreased swimming activity compared to WT SOD1 animals (^#p < 0.04). E, Another G85R mutant line [G85R(18)], which expresses the mSOD1 transgene at a higher copy number compared to the previously used G85R mutants, also shows rescue of locomotor activity when placed in the background of the aak-2(ok524) worm [*p ≤ 0.03 and *p ≤ 0.04, significantly different between G85R(18);aak-2 and G85R(18) worms in the L4 and adult stages, respectively]. F, There is a decreased ratio of soluble (S) to insoluble (I) human SOD1 protein in both G85R and G85R;aak-2 worm lysates compared to WT SOD1 controls. Furthermore, there is no difference in the S/I SOD1 protein ratio between the two mutant groups. G, RNAi knockdown of aak-2 by feeding in G85R worms has no effect, but has a significant effect in the background of a sensitized RNAi mutant strain (*p ≤ 0.04 in G85R;sid-1 worms fed aak-2 RNAi compared to EV RNAi). For A–E and G, asterisks indicate significance between mutant strains and appropriate controls, whereas hash marks in A–D indicate significance between G85R and G85R;aak-2 animals or WT SOD1 and WT SOD1; aak2. A–D are representative figures, and at least two other independent experiments were performed separately. For E and G, each bar graph represents the average data from three independent experiments. For all locomotor assays, each bar graph represents data from 5–12 replicates, containing ∼10 animals per group, for a total of 50–120 tested animals. Error bars indicate ±SEM.

Figure 6.

G85R worms exhibit fecundity defects, which are rescued by ablating aak-2. A, Approximately 50% of G85R worms are sterile compared to 0% in WT SOD1 controls and <10% in G85R;aak-2 worms. B, Number of viable offspring for each genotype, broken down per day. Compared to N2 controls, self-fertilized WT SOD1 hermaphrodites demonstrate decreased numbers of viable progeny laid throughout the reproductive span, especially during Days 2 and 3 (⁺p < 0.0001 for both days). From the subset of worms that were fertile, G85R animals demonstrate a dramatic reduction of viable offspring compared to WT SOD1 controls, especially during Days 1 and 2 of active reproduction (*p < 0.002 and *p ≤ 0.003, respectively). Placing the G85R mSOD1 animals into the aak-2(ok524) null background rescues this fecundity defect, through out the majority of the reproductive span (Days 1 through 7; ^#p ≤ 0.02). There are no differences in brood sizes between aak-2(ok524) animals and N2 controls. C, Sum of viable offspring through out the entire reproductive span. There is a decrease in the total number of viable offspring produced by WT SOD1 compared to N2 worms (*p < 0.001). However, G85R mSOD1 worms show a significantly lower number of total viable offspring compared to WT SOD1 controls (*p < 0.002). This defect is rescued when G85R worms are placed into the aak-2 null background (^#p < 0.0001 G85R;aak-2 vs G85R animals). There is no difference in the total number of viable offspring during the entire reproductive span between aak-2 animals and N2 controls. D, Breakdown of dropped eggs from WT SOD1, G85R, and G85R;aak-2 animals into three categories: (1) live or viable, (2) unhatched, and (3) dead or unfertilized. When counted 2–3 d after egg drop, a significant number of G85R eggs still remained unhatched compared to 0% in WT SOD1 worms and ∼10% in G85R;aak-2 double-mutant worms. Unfertilized embryos are also present a day earlier in G85R worms (Day 6) compared to WT SOD1 animals (Day 7). Although G85R;aak-2 worms also begin producing unfertilized embryos at Day 6, only a little over 40% of total eggs laid were dead compared to close to 100% in G85R worms. For B and C, plus signs indicate significance between WT SOD1 and N2 controls, asterisks indicate significance between WT SOD1 and G85R, and hash marks indicate significance between G85R and G85R;aak-2 animals. For each genotype, n = 9–20 hermaphrodites each. Error bars indicate ±SEM.

Figure 7.

Decreasing aak-2 activity in the M337V mutant TDP-43 model of motor neuron disease in C. elegans also improves locomotor behavior. A, In the L4 stage, WT TDP-43 worms exhibit modest locomotor defects compared to N2 worms (*p ≤ 0.002); however, M337V mutant TDP-43 worms exhibit a more severe locomotor defect compared to WT TDP-43 worms (*p ≤ 0.002). When M337V worms were placed into the aak-2(ok524) null background, no improvement in swimming behavior was observed in the M337V;aak-2 double mutants compared to M337V controls. No difference was also detected between WT TDP-43 and WT TDP-43;aak-2 animals. B, In adults, although there is no significant difference in swimming behavior between N2 and WT TDP-43 worms, there remains a severe locomotor defect between WT TDP-43 and M337V worms (*p ≤ 0.001). There is a significant improvement in locomotor activity in M337V;aak-2 double mutants compared to M337V worms (^#p ≤ 0.02), which persists until later stages of adulthood (data not shown). C, M337V mutant TDP-43 and M337V; aak-2 worms show significant loss of GABAergic neurons compared to the CZ1200 control reporter strain during L4 and young adulthood (*p < 0.0001). However, double-mutant M337V;aak-2 animals do not show decreased neuronal loss compared to M337V at both time points. For A–C, asterisks indicate significance between mutant strains and appropriate controls (p ≤ 0.05), while hash marks indicate significance between M337V mutant TDP-43 and M337V;aak-2 controls. For A and B, each bar graph represents data from 10 replicates, containing ∼10 animals per group, for a total of 100 tested animals. For C, each graph contains data from 10–20 worms per group. Error bars indicate ±SEM.

Figure 8.

The relationship between TBC1D1/tbc-11 and models of motor neuron disease. A, Mixed spinal cord cultures infected with HSV-G37R mSOD1 for 48 h do not show significant changes in TBC1D1 mRNA levels compared to WT SOD1-infected cultures (n = 3–5 dishes per group, three independent experiments). Furthermore, spinal cord and hindlimb muscle from P120 G93A mSOD1 mice do not show altered levels of TBC1D1 compared to non-Tg controls (n = 5–7 animals per group). All data are normalized to 18s RNA and are presented as percentages compared to controls (either HSV-WT SOD1 or non-Tg controls). B, The expression of tbc-11, the worm paralog of TBC1D1, does not differ in N2, aak-2, and WT SOD1 worms. On the contrary, young adult G85R worms show significantly elevated tbc-11 mRNA levels compared to N2 controls (p = 0.003) and WT SOD1 animals (*p < 0.02). Placing G85R worms into the aak-2 null background normalizes tbc-11 transcript expression back to control levels (^#p = 0.002 compared to G85R worms). Both young adult WT TDP-43 and M337V mutant TDP-43 worms do not show any alterations in tbc-11 mRNA expression compared to N2 controls. Representative figures are normalized to the housekeeping gene cdc-42 and are presented as percentages compared to N2 controls. At least one to two other independent experiments were performed separately. C, In a locomotor assay, RNAi knockdown of tbc-11 by feeding has no statistically significant effect on G85R and G85R;sid-1 worms compared to the EV control. There is a trend toward improvement in the G85R;sid-1 worms fed tbc-11 RNAi bacteria. D, G85R worms were placed into the background of a tbc-11(ok2576) null background to produce G85R;tbc-11 double-mutant worms. No significant improvement in swimming behavior is observed at the L4 time point compared to G85R controls. M337V mutant TDP-43 worms were placed into the background of a tbc-11(ok2576) null background to produce M337V;tbc-11 double-mutant worms. No significant improvement in swimming behavior is observed at the L4 time point compared to M337V controls. E, There is no significant difference in swimming behavior between L4 G85R;aak-2;tbc-11 triple-mutant worms compared to G85R;aak-2. double-mutant worms Furthermore, G85R;aak-2 and G85R;aak-2;tbc-11 worms showed a significant improvement in locomotor activity compared to G85R worms (^#p < 0.03). For C and D, each bar graph represents the average data from three to five independent experiments. For all locomotor assays, each experiment comprises five replicates, containing ∼10 animals per group, for a total of ∼50 animals per experiment. Error bars indicate ±SEM.