Mitochondrial GTP metabolism controls reproductive aging in C. elegans

Abstract

Healthy mitochondria are critical for reproduction. During aging, both reproductive fitness and mitochondrial homeostasis decline. Mitochondrial metabolism and dynamics are key factors in supporting mitochondrial homeostasis. However, how they are coupled to control reproductive health remains unclear. We report that mitochondrial GTP (mtGTP) metabolism acts through mitochondrial dynamics factors to regulate reproductive aging. We discovered that germline-only inactivation of GTP- but not ATP-specific succinyl-CoA synthetase (SCS) promotes reproductive longevity in Caenorhabditis elegans. We further identified an age-associated increase in mitochondrial clustering surrounding oocyte nuclei, which is attenuated by GTP-specific SCS inactivation. Germline-only induction of mitochondrial fission factors sufficiently promotes mitochondrial dispersion and reproductive longevity. Moreover, we discovered that bacterial inputs affect mtGTP levels and dynamics factors to modulate reproductive aging. These results demonstrate the significance of mtGTP metabolism in regulating oocyte mitochondrial homeostasis and reproductive longevity and identify mitochondrial fission induction as an effective strategy to improve reproductive health.

Keywords: GTP metabolism; bacteria-host interaction; gene-environment interaction; mitochondrial distribution; mitochondrial dynamics; oocyte quality control; reproductive aging; succinyl-CoA synthetase; vitamin B12.

Figure 1.. GTP-specific Succinyl-CoA Synthetase (SCS) regulates reproductive aging

(A) A diagram of SCS enzymatic function and its ATP or GTP specificity. (B) Wild-type (WT) worms subjected to sucg-1 RNA interference (RNAi) have a significantly longer reproductive lifespan (RLS) than those subjected to the empty vector (EV) control. (C) WT worms subjected to sucl-2 RNAi have a longer RLS than those subjected to the EV control. (D) Day 7 and 9 WT hermaphrodites subjected to sucg-1 or sucl-2 but not suca-1 RNAi show higher rates of reproduction than those subjected to the EV control, when mated with day-2-old males. (E) WT worms subjected to suca-1 RNAi show no significant differences in RLS compared to those subjected to the EV control. (B, C, E) n.s. p > 0.05, *** p < 0.001 by log-rank test; n = 3 biological independent replicates, ~20 worms per replicate, see Supplementary Table 1 for full RLS Data. (D) Error bars represent mean ± s.e.m., n = 3 biologically independent samples, n.s. p > 0.05, * p < 0.05, ** p < 0.01, *** p < 0.001 by Fisher’s exact test adjusted with the Holm–Bonferroni method for multiple comparisons, ~20 worms per replicate.

Figure 2.. GTP-specific SCS functions in the germline to regulate oocyte mitochondria during reproductive aging

(A) Confocal imaging of the SUCG-1::eGFP knock-in line, in which the endogenous sucg-1 is tagged with egfp, shows its predominant expression in the germline but weak expression in the intestine, pharynx, muscle, hypodermis and neurons (Scale bar: 100μm; Dashed white line: germline). (B) The SUCG-1::eGFP level in the germline is increased at day 5 compared to day 1 (Scale bar: 100μm). (C) Quantification of SUCG-1::eGFP level in the germline at day 1 and day 5. (AU: arbitrary unit) (D) Germline-specific RNAi inactivation of sucg-1 extends RLS. (E) Germline-specific RNAi inactivation of sucl-2 extends RLS. (F) SUCG-1::eGFP colocalizes with mitochondrial TOMM-20::mKate2 in the germline (Scale bar: 30μm for the images with lower magnification; 5μm for the images with higher magnification). (G) Overexpression of mitochondria-targeted ndk-1(mito::ndk-1) in the germline suppresses the RLS extension caused by sucg-1 RNAi knockdown. (H) Representative images show that oocyte mitochondria are largely dispersed at day 1 while experiencing increasing perinuclear distribution at day 5 (Scale bar: 5μm; Dashed white line: oocyte outline; N: nucleus). (I) Perinuclear clustering of oocyte mitochondria is increased from day 1 to day 5. (J) The increase in perinuclear distribution of oocyte mitochondria at day 5 is suppressed by sucg-1 or sucl-2, but not suca-1 RNAi knockdown. (C) *** p < 0.001 by Student’s t-test; n = 31 (day 1), n = 32 (day 5). (D, E, G) ** p < 0.01, *** p < 0.001 by log-rank test; n = 4 (D) or 3 (E, G) biological independent replicates, ~20 worms per replicate, see Supplementary Table 1 for full RLS Data. (I) n = 43 (day 1), n = 40 (day 5); *** p < 0.001 by Chi-squared test. (J) n= 45 (EV, D1), n = 45 (sucg-1, D1), n = 45 (sucl-2, D1), n = 45 (suca-1, D1), n = 42 (EV, D5), n = 45 (sucg-1, D5), n = 44 (sucl-2, D5), n = 48 (suca-1, D5); RNAi vs. EV, n.s. p > 0.05, *** p < 0.001 by Chi-squared test adjusted with the Holm–Bonferroni method for multiple comparisons.

Figure 3.. Mitochondrial fission-fusion factors regulate reproductive longevity

(A) A diagram showing regulation of mitochondrial fission-fusion by GTPase DRP-1, FZO-1 and EAT-3 (IMM: Inner mitochondrial membrane; OMM: Outer mitochondrial membrane). (B) Germline-specific RNAi inactivation of eat-3 extends RLS. (C) Day 7 and 9 aged hermaphrodites subjected to germline-specific eat-3 RNAi have a higher rate of reproduction than those subjected to the EV control when mated with day-2-old young males, while germline-specific RNAi inactivation of fzo-1 or drp-1 RNAi does not affect the rate of reproduction at all ages. (D) Germline-specific RNAi inactivation of fzo-1 does not affect RLS. (E) Germline-specific RNAi inactivation of drp-1 does not affect RLS. (F) Germline-specific overexpression of drp-1 prolongs RLS. (G) The perinuclear clustering of oocyte mitochondria at day 5 is decreased in the transgenic strain with germline-specific drp-1 overexpression. (H) The increase in the perinuclear distribution of oocyte mitochondria at day 5 is decreased upon eat-3 but not fzo-1 RNAi knockdown. The distribution of oocyte mitochondria is not scorable in day 5 aged worms subjected to drp-1 RNAi knockdown due to distorted germline. (B, D, E, F) n.s. p > 0.05, *** p < 0.001 by log-rank test; n = 3 biological independent replicates, ~20 worms per replicate, see Supplementary Table 1 for full RLS Data. (C) Error bars represent mean ± s.e.m., n = 4 biologically independent samples, n.s. p > 0.05, * p < 0.05 by Fisher’s exact test adjusted with the Holm–Bonferroni method for multiple comparisons, ~15 worms per replicate. (G) n= 46 (WT, D1), n = 42 (drp-1 OE, D1), n = 40 (WT, D5), n = 46 (drp-1 OE, D5); WT vs. drp-1 OE, n.s. p > 0.05, *** p < 0.001 by Chi-squared test. (H) n= 43 (EV, D1), n = 38 (eat-3, D1), n = 40 (fzo-1, D1), n = 46 (drp-1, D1), n = 41 (EV, D5), n = 41 (eat-3, D5), n = 42 (fzo-1, D5); RNAi vs. EV, n.s. p > 0.05, *** p < 0.001 by Chi-squared test adjusted with the Holm–Bonferroni method for multiple comparisons.

Figure 4.. GTP-specific SCS regulates reproductive aging through mitochondrial fission factor

(A) A diagram demonstrating auxin-induced degradation of endogenous DRP-1 tagged with GFP and Degron. (B) Confocal imaging of GFP shows that the endogenous DRP-1 protein is specifically depleted in the germline upon the auxin treatment (Scale bar: 100μm for the images with lower magnification; 30μm for the images with higher magnification). (C) Auxin-induced germline-specific depletion of DRP-1 does not affect RLS. (D) Auxin-induced germline-specific depletion of DRP-1 abrogates the RLS extension caused by sucg-1 RNAi. (E) The drp-1 loss-of-function mutant increases the perinuclear clustering of oocyte mitochondria at day 1, which is not suppressed by sucg-1 RNAi knockdown. (C, D) n.s. p > 0.05, *** p < 0.001 by log-rank test; n = 3 biological independent replicates, ~20 worms per replicate, see Supplementary Table 1 for full RLS Data. (E) n= 38 (WT, EV RNAi, D1), n = 41 (drp-1(tm1108), EV RNAi, D1), n = 41 (WT, sucg-1 RNAi, D1), n = 46 (drp-1(tm1108), sucg-1 RNAi, D1); RNAi vs EV and WT vs drp-1 mutant, n.s. p > 0.05, * p < 0.05, *** p < 0.001 by Chi-squared test adjusted with the Holm–Bonferroni method for multiple comparisons.

Figure 5.. Bacterial inputs regulate germline mitochondrial GTP and reproductive aging

(A) A diagram showing the strategy to obtain sucg-1 homozygous knockout (KO) mutants from heterozygous mutants with sucg-1 KO at one locus and sucg-1::egfp (GFP) at the other locus. (B) sucg-1 KO/KO mutants show a significant increase in RLS compared to sucg-1 GFP/GFP and sucg-1 KO/GFP worms. (C) With OP50 bacteria, sucg-1 KO/KO mutants show no significant differences in RLS compared to sucg-1 GFP/GFP or sucg-1 KO/GFP worms. (D) Germline mitochondrial GTP (mtGTP) level is increased by 9.4-fold in day 5 aged worms compared to day 1 young worms on HT115 bacteria. With OP50 bacteria, the germline mtGTP level increase from day 1 to day 5 is 3-fold. The germline mtGTP level is higher in worms on HT115 bacteria than those on OP50 bacteria at day 5, but not at day 1. (E) Germline mitochondrial ATP (mtATP level) is not significantly different in worms of different ages and on different bacteria. (B, C) n.s. p > 0.05, *** p < 0.001 by log-rank test; n = 3 biological independent replicates, ~80 worms per replicate split into 3 genotypes, see Supplementary Table 1 and Supplementary Table 3 (C) for full RLS Data. (D, E) Error bars represent mean ± s.e.m., n = 4 biologically independent samples, n.s. p > 0.05, * p < 0.05, ** p < 0.01, *** p < 0.001 by Student’s t-test adjusted with the Holm–Bonferroni method for multiple comparisons.

Figure 6.. Mitochondrial fission-fusion factors mediate bacterial effects on reproductive longevity

(A) The perinuclear clustering of oocyte mitochondria is decreased in day 5 worms on OP50 compared to those on HT115 bacteria. (B) Auxin-induced germline-specific depletion of DRP-1 reduces RLS in worms on OP50 bacteria. (C) Adult-only germline-specific depletion of DRP-1 reduces RLS in worms on OP50 bacteria. (D) Germline-specific overexpression of drp-1 prolongs RLS in worms on OP50 bacteria. (E) Germline-specific RNAi inactivation of eat-3 fails to extend RLS in worms on OP50 bacteria. (F) With OP50 bacteria, the distribution of oocyte mitochondria is not significantly different between EV control worms and those subjected to eat-3 or fzo-1 RNAi knockdown at day 5. With drp-1 RNAi knockdown, oocyte mitochondrial distribution becomes unscorable due to distorted germline. (A) n = 48 (HT115, D1), n = 53 (OP50, D1), n = 47 (HT115, D5), and n = 48 (OP50, D5); HT115 vs. OP50, n.s. p > 0.05, *** p < 0.001 by Chi-squared test. (B, C, D, E) n.s. p > 0.05, ** p < 0.01, *** p < 0.001 by log-rank test; n = 3 (B, C, D) or 4 (E) biological independent replicates, ~20 worms per replicate, see Supplementary Table 3 for full RLS Data. (F) n= 44 (EV, D1), n = 43 (eat-3, D1), n = 41 (fzo-1, D1), n = 42 (drp-1, D1), n = 43 (EV, D5), n = 43 (eat-3, D5), n = 43 (fzo-1, D5); OP50 condition; RNAi vs. EV, n.s. p > 0.05 by Chi-squared test adjusted with the Holm–Bonferroni method for multiple comparisons.

Figure 7.. Bacterial VB12 regulates oocyte mitochondria and reproductive aging

(A, B) As a VB12-deficiency reporter, the acdh-1:gfp signal level is higher in day-1 worms on OP50 than those on HT115 bacteria (Scale bar: 100μm in A). GFP signal quantification is shown in B (AU: arbitrary unit). (C) Supplementation of meCbl shortens RLS of WT worms on OP50 bacteria. (D) Supplementation of meCbl increases the perinuclear clustering of oocyte mitochondria in WT worms on OP50 bacteria at day 5. (E) Supplementation of meCbl does not shorten RLS of the sucg-1 knockout worms on OP50 bacteria. (F) Summary model representing mitochondrial GTP metabolism and mitochondrial fission-fusion couple in the oocyte to regulate reproductive longevity, which is modulated by metabolic inputs from bacteria. (B) n = 15 (HT115), n = 15 (OP50); *** p < 0.001 by Student’s t-test. (C) *** p < 0.001 by log-rank test; n = 3 biological independent replicates, ~20 worms per replicate, see Supplementary Table 3 for full RLS Data. (D) n= 40 (EV, D1), n = 45 (128nM meCbl, D1), n = 42 (EV, D5), n = 38 (128nM meCbl, D5); OP50 condition; 128nM meCbl vs EV, n.s. p > 0.05, *** p < 0.001 by Chi-squared test. (E) n.s. p > 0.05 by log-rank test; n = 3 biological independent replicates, ~80 worms per replicate split into 3 genotypes, see Supplementary Table 3 for full RLS Data.