Increased mitochondrial fusion allows the survival of older animals in diverse C. elegans longevity pathways

Abstract

Mitochondria are dynamic organelles that undergo fusion and fission events. Mitochondrial dynamics are required for mitochondrial viability and for responses to changes in bioenergetic status. Here we describe an insulin-signaling and SCF^LIN-23-regulated pathway that controls mitochondrial fusion in Caenorhabditis elegans by repressing the expression of the mitochondrial proteases SPG-7 and PPGN-1. This pathway is required for mitochondrial fusion in response to physical exertion, and for the associated extension in lifespan. We show that diverse longevity pathways exhibit increased levels of elongated mitochondria. The increased mitochondrial fusion is essential for longevity in the diverse longevity pathways, as inhibiting mitochondrial fusion reduces their lifespans to wild-type levels. Our results suggest that increased mitochondrial fusion is not a major driver of longevity, but rather is essential to allow the survival of older animals beyond their normal lifespan in diverse longevity pathways.Mitochondria can undergo shape changes as a result of fusion and fission events. Here the authors describe how insulin signalling regulates mitochondrial fusion in C. elegans, and show that mitochondrial fusion is necessary, but not sufficient, for longevity of worms with mutations that increase lifespan.

Fig. 1

CAND-1/SCF^LIN-23 regulates mitochondrial shape via spg-7 and ppgn-1. a Real-time PCR quantification of spg-7 mRNA levels in adults of the indicated genotypes, presented in arbitrary units (a.u.) normalized to control rpl-19 (ribosomal protein L19) mRNA. Error bars denote s.e.m. from four biological replicates, each with at least two technical replicates. P values were determined by Student’s t-test. b Images of tubular, elongated, and fragmented mitochondria visualized by mitochondria-targeted GFP expressed in body-wall muscle cells. Scale bar, 10 µm. c The percentages of muscle cells with tubular, elongated, and fragmented mitochondria for the indicated genotypes and RNAi treatments. All animals were fed either gene-specific RNAi or control RNAi so that the feeding conditions were matched. P values were determined by χ ²-test. Sample size (n) of muscle cells from left to right are: 729; 126; 210; 124, 210; 201; 234; 274; 209; 297; 318; 457. Mitochondrial morphology was scored blinded. d Real-time PCR quantification of ppgn-1 mRNA levels normalized to control rpl-19 mRNA in adults of the indicated genotypes. Error bars denote s.e.m. from three biological replicates, each with at least two technical replicates. P values determined by Student’s t-test. For all panels, asterisks above bars denote P value comparisons to wild-type/controls; asterisks above lines denote comparisons under the lines: *P < 0.05; **P < 0.01; ***P < 0.001; ns = not significant

Fig. 2

CAND-1/SCF^LIN-23 activates AKT-1 to inhibit DAF-16 and reduce EAT-3 levels. a DAF-16::GFP nuclear localization in body-wall muscle cells for animals with the indicated RNAi treatments. White arrows indicate nuclei in the control RNAi image. Scale bar, 10 µm. b Quantification of the mean level of nuclear and cytoplasmic DAF-16::GFP intensity in body-wall muscle cells. Error bars denote s.e.m. P values were determined by Student’s t-test. Sample size (n) of body-wall muscle cells from left to right are: 44; 61; 102; 87. c Western blot with anti-GFP of AKT-1::GFP from whole-animal lysate of L4/young-adult-stage animals treated with the indicated RNAi showing the altered mobility of AKT-1::GFP on SDS-PAGE; anti-histone H4 staining is used as a loading control. Similar results were obtained in two to five biological replicates. d Western blot showing staining for an antibody that recognizes AKT phospho-substrates in animals expressing AKT-1::GFP and subjected to the indicated RNAi treatments. Significant decreases in AKT phospho-substrates were observed for cand-1, cul-1, and lin-23 RNAi conditions in three to five biological replicates, with pdk-1 RNAi shown as a control that is known to reduce AKT-1 activity. The identities of the AKT-1 phospho-substrate proteins observed in the western blot are not known. e Graph showing the levels of AKT phospho-substrate signal relative to control protein bands (anti-tubulin or anti-histone H4) for three to five biological replicates. The control RNAi is set to 100 in arbitrary units. Error bars denote s.e.m. P values were determined by Student’s t-test. f Western blot showing EAT-3 protein levels for the indicated genotypes and RNAi treatments. The expected molecular weights of the L-isoform and S-isoform of EAT-3 are 110.1 and 85.5–86.4 kDa, respectively. g Quantification of EAT-3 levels normalized to α-tubulin. Error bars denote s.e.m. from two to four biological replicates. For all panels, asterisks above bars denote P value comparisons to wild type/controls; asterisks above lines denote comparisons under the lines: *P < 0.05; **P < 0.01; ***P < 0.001; ns = not significant

Fig. 3

Mitochondrial morphology in C. elegans body-wall muscles. a, b The percentages of muscle cells with predominantly tubular, elongated, or fragmented mitochondria in adult hermaphrodites of the indicated RNAi treatments visualized by mitochondria-targeted GFP expressed in body-wall muscle cells. P values were determined by χ ²-test. Sample size (n) of muscle cells from left to right are: a 119, 128, 152, 198, 173; and b 151, 111, 173, 125, 136. Mitochondrial morphology was scored blinded. For all panels, asterisks above bars denote P value comparisons to wild type/controls; asterisks above lines denote comparisons under the lines: ***P < 0.001; ns = not significant

Fig. 4

Model for regulation of mitochondrial fusion. a Proposed linear pathway for CAND-1 and SCF^LIN-23 regulation of mitochondrial fusion; see text for description. b, c Schematic of the proposed intracellular pathway regulating mitochondrial fusion and lifespan extension in the presence b or absence c of CAND-1

Fig. 5

DAF-16 is required for physical exertion-induced mitochondrial fusion. a ATP levels decrease during swimming. Graph of ATP levels (normalized to whole-animal protein levels) at the indicated times of continuous swimming. Error bars denote s.e.m. from two biological replicates, each with three technical replicates. P values were determined by Student’s t-test. b Average swim strokes per minute of 12 animals each for the indicated genotypes upon induction of swimming behavior. Error bars denote s.e.m. P values determined by Student’s t-test. c The percentages of muscle cells with elongated mitochondria in wild type, daf-16(mu86), and cand-1(tm1683); spg-7(ek25) animals for the indicated times post induction of swimming behavior. Full distributions of mitochondrial morphology and sample size (n) are shown in Supplementary Fig. 9. d DAF-16::GFP nuclear localization in body-wall muscle cells for animals at 0 and 300 min post induction of swimming behavior. White arrows indicate nuclei in the 0 min image. Scale bar, 10 µm. For all panels, asterisks above bars denote P value comparisons to wild type/controls; asterisks above lines denote comparisons under the lines: *P < 0.05; **P < 0.01; ***P < 0.001; ns = not significant

Fig. 6

Increased mitochondrial elongation extends lifespans. a–e Survival curves for adults of the indicated RNAi treatments. The wild-type survival curves for a–c were analyzed at the same time and are shown in each panel for comparison. RNAi depletions of lin-23 a, cand-1 a, ppgn-1 c, and spg-7 + ppgn-1 d significantly increased mean lifespan. b cand-1 mutants and spg-7(tm2312) mutants d had increased the mean lifespan relative to wild type, while cand-1; spg-7(ek25) c animals had lifespan comparable to wild type. e, f eat-3 RNAi e and fzo-1 RNAi f depletions significantly decreased the mean lifespan of daf-2(e1370) mutants. All lifespan experiments were performed with four biological replicates. See Supplementary Table 2 for statistics. g The percentages of muscle cells with predominantly tubular, elongated, or fragmented mitochondria in adult hermaphrodites of the indicated genotypes/RNAi treatments visualized by mitochondria-targeted GFP expression in body-wall muscle cells. P values were determined by χ ²-test. Sample size (n) of muscle cells from left to right are: 278; 102; 179; 207; 136; 218. Mitochondrial morphology was scored blinded. The wild-type control from Fig. 1c was analyzed at the same time and is shown here for comparison. Asterisks above bars denote P value comparisons to wild type/control; asterisks above lines denote comparisons under the lines: *P < 0.05; **P < 0.01; ***P < 0.001; ns = not significant

Fig. 7

Exercise extends lifespan that is dependent on the mitochondrial fusion pathway. a–d Survival curves for adults that were kept continuously on agar plates or removed from the plates for brief periods for the described swim regimens. a Comparison of wild-type kept on agar plates continuously or subject to swim regimen A (30 min of swimming per day). b Test of swim regimens B and C. Swim regimen B was the following swimming times per day on the listed days: 30 min on days 1–5; 25 min on days 6 and 7; 20 min on days 8 and 9; 15 min on day 10; 10 min on day 11; and 5 min on day 12. Swim regimen C was 30 min of swimming on day 1, followed by a reduction in swim time of 2 min per day on subsequent days. c, d Test of swim regimen C with wild type, cand-1(tm1683); spg-7(ek25) c, or daf-16(mu86) d. The wild-type control and wild-type swim regimen C survival curves were analyzed at the same time and are shown in c, d for comparison. All swimming regimens significantly increased the mean lifespan of wild type a–d. Swim regimen C did not increase the mean lifespan of cand-1(tm1683); spg-7(ek25) c or daf-16(mu86) mutants d relative to the control, non-swimming condition. All lifespan experiments were performed with four biological replicates. See Supplementary Table 3 for statistics

Fig. 8

Diverse life extension pathways have increased levels of elongated mitochondria. a, b The percentages of body-wall muscle cells with tubular, elongated, and fragmented mitochondria in the indicated genotypes, overexpression (oe), and RNAi treatments. P values were determined by χ ²-test. Sample size (n) of muscle cells from left to right are: a 254, 239, 301, 331, 103, 110, 303, 270, 141, 148; and b 148, 181, 120, 121, 98, 114, 153, 122, 260, 264, 143. For all panels, asterisks above bars denote P value comparisons to wild-type/control; asterisks above lines denote comparisons under the lines: *P < 0.05; **P < 0.01; ***P < 0.001; ns = not significant

Fig. 9

Mitochondrial elongation is required for longevity in diverse mutants. a–i Survival curves for adults of the indicated genotypes, overexpressions (oe), and RNAi treatments. The wild-type and eat-3(RNAi) survival curves are shown in graphs for comparison. eat-3 RNAi significantly reduced the extended lifespans relative of all tested genotypes, overexpression, and RNAi treatments except for vhl-1 mutants and wild-type control e. All lifespan experiments were performed with four biological replicates. See Supplementary Table 2 for statistics

Fig. 10

The UPR^mt is not involved in the mitochondrial fusion pathway. a Representative images of the UPR^mt marker Phsp-6::GFP in L4/young-adult-stage animals for the indicated RNAi treatments. Scale bar, 200 µm. b Quantification of Phsp-6::GFP signal for the RNAi treatments shown in a for 20 animals. Error bars denote s.e.m. P values were determined by Student’s t-test. c, d Survival curves for atfs-1(et17) gain-of-function mutants with the indicated RNAi treatments; the wild-type survival curve is shown in both graphs for comparison. e The percentages of body-wall muscle cells with tubular, elongated, and fragmented mitochondria in the indicated genotypes and RNAi treatments. P values were determined by χ ²-test. Sample size (n) of muscle cells from left to right are: 427; 234; 333. f Quantification of Phsp-6::GFP signal for 20 animals swimming for the indicated times. Error bars denote s.e.m. P values determined by Student’s t-test. The quantification was performed as for b and the two graphs can be compared directly. No significant differences in Phsp-6::GFP were observed relative to 0 min control animals. For all panels, asterisks above bars denote P value comparisons to wild type/controls; asterisks above lines denote comparisons under the lines: *P < 0.05; **P < 0.01; ***P < 0.001; ns = not significant