Developmental mitochondrial complex I activity determines lifespan

Abstract

Aberrant mitochondrial function has been associated with an increasingly large number of human disease states. Observations from in vivo models where mitochondrial function is altered suggest that maladaptations to mitochondrial dysfunction may underpin disease pathology. We hypothesized that the severity of this maladaptation could be shaped by the plasticity of the system when mitochondrial dysfunction manifests. To investigate this, we have used inducible fly models of mitochondrial complex I (CI) dysfunction to reduce mitochondrial function at two stages of the fly lifecycle, from early development and adult eclosion. Here, we show that in early life (developmental) mitochondrial dysfunction results in severe reductions in survival and stress resistance in adulthood, while flies where mitochondrial function is perturbed from adulthood, are long-lived and stress resistant despite having up to a 75% reduction in CI activity. After excluding developmental defects as a cause, we went on to molecularly characterize these two populations of mitochondrially compromised flies, short- and long-lived. We find that our short-lived flies have unique transcriptomic, proteomic and metabolomic responses, which overlap significantly in discrete models of CI dysfunction. Our data demonstrate that early mitochondrial dysfunction via CI depletion elicits a maladaptive response, which severely reduces survival, while CI depletion from adulthood is insufficient to reduce survival and stress resistance.

Keywords: Ageing; Complex I; Drosophila; Mitochondria; Mitochondrial Disease.

Figure 1. Depletion of respiratory complex I from development but not from adulthood shortens lifespan.

(A) A schematic illustrating the periods during which the GS inducible expression system was used to reduce CI function, from early development into adulthood (D + A) or from adulthood only (A-only). (B) Survival of male flies where CI subunit, ND-18, has been depleted from early development (D + A) or from adulthood (A-only). Log-rank test, P < 0.0001, n = 262-266 flies per experimental group. (C) Survival of male flies where CI subunit, ND-75, has been depleted from early development (D + A) or from adulthood (A-only). Log-rank test, P < 0.0001, n = 151–220 flies per experimental group. (D) Normalized levels of CI-linked respiration in ND-18 KD (left) and ND-75 KD (right) 5–7 day adult males. Kolmogorov–Smirnov test, P = 0.0460 (ND-18 KD); T test, P = 0.7131 (ND-75 KD), n = 4–10 independent samples per group. (E) Normalized levels of CI-linked respiration in ND-18 KD (left) or ND-75 KD (right) male 3rd instar larvae in the presence (drug) or absence (vehicle) of the GS inducer, RU-486. T test, P = 0.0016 (ND-18 KD) & P = 0.0429 (ND-75 KD), n = 3–6 independent samples per group. (F) A schematic illustrating the developmental stages where the UAS/GAL4 expression system, coupled with the temperature-sensitive transcriptional repressor GAL80ts, was employed to reduce CI function. (G) % eclosion of flies where depletion of CI subunit, ND-18, has been induced at distinct developmental stages. n = 3–5 independent vias per group. (H) Survival of male flies where depletion of CI subunit, ND-18, has been induced at distinct developmental stages. Log-rank test, P < 0.0001, n = 44–71 flies per group. Please note that in the wL3 > 29 ND-18-KD group, only four flies survived after eclosion (flies got stuck in the fly food); therefore, this group was excluded from the statistical analysis. In bar graphs, means are presented with ± SEM. Significance is indicated as follows: *P < 0.05, **P < 0.01, ***P < 0.001; ns not significant. Source data are available online for this figure.

Figure 2. Developmental but not adulthood CI depletion increases stress sensitivity and alters energy homeostasis.

(A) Survival under oxidative stress conditions of male flies where either ND-18 (left) or ND-75 (right) has been depleted from early development (D + A) or from adulthood (A-only). Log-rank test, P < 0.0001, n = 80–140 flies per experimental group. (B) Survival under thermal stress conditions of male flies where either ND-18 (upper) or ND-75 (lower) has been depleted from early development (D + A) or from adulthood (A-only). Log-rank test, P < 0.0001, n = 85–304 flies per experimental group. (C) Survival under starvation conditions of male flies where either ND-18 (left) or ND-75 (right) has been depleted from early development (D + A) or from adulthood (A-only). Log-rank test, P < 0.0001, n = 51–254 flies per experimental group. (D) Quantification of triacylglyceride levels in male flies where either ND-18 (left) or ND-75 (right) has been depleted from early development (D + A) or from adulthood (A-only). T test, P = 0.0090 (ND-18 KD) & P = 0.0006 (ND-75 KD), n = 3–5 samples per experimental group. (E) Confocal imaging of fat bodies from male flies where ND-75 left have been depleted from early development (D + A) or from adulthood (A-only) stained with LipidTOX Red and DAPI, scale bar (90 μM), inset scale bar (10 μM). Quantification is displayed on the right. T test, P = 0.0045, n = 4–5 independent samples per experimental group. In bar graphs, means are presented with ± SEM. Significance is indicated as follows: *P < 0.05, **P < 0.01, ***P < 0.001. Source data are available online for this figure.

Figure 3. Developmental CI depletion results in a specific transcriptomic signature.

(A, B) Volcano plot of genes differentially expressed between flies where ND-18 (A) or ND-75 (B) was depleted from early development (D + A) or from adulthood only (A). Each dot represents a single gene. In pink are those genes that are significantly altered; in red are those which are significantly altered in both ND-18 and ND-75 D + A KD flies. In gray are genes which were not significantly altered. FDR, P < 0.05, n = 4 independent samples per experimental group. (C) Venn diagram illustrating the relationship between the differential expression seen in ND-18 D + A vs A-only and ND-75 D + A vs A-only, n = 4 independent samples per experimental group. (D) GO Enrichment Analysis performed on those genes significantly altered in both ND-18 D + A vs A and ND-75 D + A vs A reveals significant enrichment in the following classifications. FDR, P < 0.05, n = 4 independent samples per experimental group. (E–G) Volcano plot of genes significantly differentially expressed between ND-18 D + A and ND-18 A-only (gray), highlighted in pink are those genes whose expression is known to be highly enriched in the fat body (E), midgut (F) and heart (G). FDR, P < 0.05, n = 4 independent samples per experimental group. Source data are available online for this figure.

Figure 4. Developmental CI depletion results in a specific metabolomic and proteomic signature.

(A, B) Volcano plots of metabolites with significantly different abundance between flies where ND-18 (A) or ND-75 (B) was depleted from early development (D + A) vs from adulthood only (A-only). FDR, P < 0.05, n = 4 independent samples per experimental group. (C) Functional Enrichment Analysis using MetaboAnalyst MS Peaks to Pathways (KEGG) of the differential metabolic profiles of ND-18 and ND-75 KD D + A vs A-only, identified enrichment of the following categories. FDR, P < 0.05, n = 4 independent samples per experimental group. (D) Volcano plot of proteins with significantly different abundance between flies where ND-18 has been depleted from development (D+A) vs adulthood only (A-only). FDR, P < 0.05, n = 5 independent samples per experimental group. (E) Dot plot showing the STRING Functional Enrichment Analysis results of ND-18 KD D + A vs A-only flies. FDR, P < 0.05, n = 5 independent samples per experimental group. (F) Dot plot showing the most significant pathways according to KEGG altered at the protein level in ND-18 KD D + A flies. FDR, P < 0.05, n = 4 independent samples per experimental group. Source data are available online for this figure.

Figure EV1. Related to Figure 1.

(A) mRNA quantification of KD in ND-18 KD (left) and ND-75 KD (right) adult males. Kolmogorov–Smirnov test, P = 0.100 (ND-18 KD) & P = 0.100 (ND-75 KD), n = 3 independent samples per experimental group. (B) mRNA quantification of KD in ND-18 KD (left) and ND-75 KD (right) 3rd instar larvae in the presence (drug) or absence (vehicle) of the GS inducer, RU-486. T test, P = 0.0003 (ND-18 KD) & P = 0.0009 (ND-75 KD), n = 4 independent samples per experimental group. (C) Western blot analysis of CI core subunit, NDUFV2 (ND-24) in control males and males where ND-18 or ND-75 has been depleted from early development (D + A) or from adulthood only (A-only), beta-actin was used as a loading control. ANOVA with Tukey’s multiple comparison test (P = 0.0059 (ND-18 KD) & P = 0.0069 (ND-75 KD)). n = 3–6 independent samples per experimental group. (D) Quantification of ND-18 (upper) and ND-75 (lower) mRNA levels in adult males exposed to the inducer from early development (D + A) or from adulthood (A-only). T test, P = 0.2852 (ND-18 KD) & P = 0.5807 (ND-75 KD), n = 3 independent samples per experimental group. (E) Survival of male control flies exposed to the inducer, RU-486, from early development (D + A) or from adulthood (A-only). Log-rank test, P = 0.1120, n = 223–262 individuals per experimental group. (F) Survival of female flies where CI subunit, ND-18, has been depleted from early development (D + A) or from adulthood (A-only). Log-rank test, P < 0.001, n = 50–58 flies per experimental group. (G) Survival of female flies where CI subunit, ND-75, has been depleted from early development (D + A) or from adulthood (A-only). Log-rank test, P < 0.001, n = 55–57 flies per experimental group. (H) Levels of CI-linked respiration in control adult males exposed to the inducer from early development (D + A) or from adulthood (A-only). T test, P = 0.2045, n = 10–11 independent samples per group. (I) % Eclosion of flies where depletion of CI subunit, ND-75, has been induced at distinct developmental stages. n = 5 independent samples per experimental group. (J) Survival of male flies where depletion of CI subunit, ND-75, has been induced at distinct developmental stages. n = 92–100 flies per experimental group. (K) % Eclosion of control male flies moved from 18 °C to 29 °C at distinct developmental stages. n = 4–5 independent samples per experimental group. (L) Survival of control male flies moved from 18 °C to 29 °C at distinct developmental stages. n = 26–72 flies per experimental group. In bar graphs, means are presented with ± SEM. Significance is indicated as follows: *P < 0.05, **P < 0.01, ***P < 0.001; ns = not significant.

Figure EV2. Related to Figure 2.

Survival under oxidative stress (A), Log-rank test, P = 0.0979, n = 156–159 individuals per group), thermal stress (B), Log-rank test, P = 0.5462, n = 274–281 individuals per group), and starvation (C), Log-rank test, P = 0.8445, n = 250–254 individuals per group) conditions of control male flies exposed to the inducer, RU-486 from early development (D+A) or from adulthood (A-only). (D) Quantification of triacylglyceride levels in control male flies exposed to the inducer, RU-486 from early development (D+A) or from adulthood (A-only). T test, P = 0.7662. n = 5 independent samples per experimental group. (E) Confocal imaging of fat bodies from male flies where ND-18 has been depleted from D + A or A-only stained with LipidTOX Red and DAPI, scale bar (90 μM), inset scale bar (10 μM). n = 2 independent samples per group. (F) Confocal imaging of fat bodies from male flies exposed to the inducer, RU-486 from early development (D+A) or from adulthood (A-only) stained with LipidTOX Red and DAPI, scale bar (90 μM), inset scale bar (10 μM). T test, P = 0.1168, n = 4–6 independent samples per experimental group. (G) Levels of NAD+ (t test, P = 0.2269), NADH (t test, P = 0.4408), and the relative ratio of NAD( + ):NADH (t test, P = 0.6101) in control flies. n = 3 independent samples per group. (H) Levels of NAD+ (t test, P < 0.001), NADH (t test, P = 0.0041), and the relative ratio of NAD( + ):NADH (t test, P = 0.0179) in ND-75 KD flies. n = 3 independent samples per group. (I) Levels of NAD+ (t test, P = 0.5993), NADH (t test, P = 0.1191), and the relative ratio of NAD( + ):NADH (t test, P = 0.2759) in ND-18 KD flies. n = 3 independent samples per group. In bar graphs, means are presented with ± SEM. Significance is indicated as follows: *P < 0.05, **P < 0.01, ***P < 0.001; n.s. = not significant.

Figure EV3. Related to Figure 3.

(A) PCA analysis of controls and ND-18 KD D + A and A-only conditions. n = 4 independent samples per group. (B) PCA analysis of controls and ND-75 KD D + A and A-only conditions. n = 4 independent samples per group. (C) Volcano plot of genes significantly differentially expressed between control D + A and control A. FDR, P < 0.05, n = 4 independent samples per group. (D) KEGG Enrichment Analysis (FlyEnricher) performed on those genes significantly altered in both ND-18 D + A vs A-only and ND-75 D + A vs A-only reveals significant enrichment in the following classifications. FDR, P < 0.05, n = 4 independent samples per group. (E–G) Volcano plots of genes significantly differentially expressed between ND-75 D + A and ND-75 A-only (gray), highlighted in pink, are those genes with highly enriched expression in the fat body (G), midgut (H) and heart (I). FDR, P < 0.05, n = 4 independent samples per group.

Figure EV4. Related to Figure 3.

(A, B) Clustergrams of the indicated GO terms expressing the differential expression (FC) of genes associated with this term. All FCs are from ND-75 KD D + A vs A-only comparison unless otherwise indicated by a dark gray line (ND-18 KD D + A vs A-only) or a purple line (not detected in our study). n = 4 independent samples per group.

Figure EV5. Related to Figure 4.

(A) Volcano plot of metabolites with significantly different abundance between control D + A and control A-only. FDR, P < 0.05, n = 4 independent samples per group. (B, C) PCA analysis of metabolomics data from controls and ND-18 KD (B) and controls and ND-75 KD (C) D + A vs A-only conditions. n = 4 independent samples per group. (D) PCA analysis of proteomics data from control and ND-18 KD D + A vs A-only conditions. n = 5 independent samples per group. (E) Heat maps depicting the expression of OXPHOS subunits (Complexes I–V). Asterisks (*) indicate significant differences in protein expression, FDR, P < 0.05. The comparisons are: (A) ND-18 A versus control A, (D + A) ND-18 D + A versus control D + A, (D + A vs A) ND-18 D + A versus ND-18 A, and (Control) control D + A versus control A. n = 5 independent samples per group. (F) Dot plot showing the most significant pathways according to STRING wikiPathways altered at the protein level in ND-18 KD D + A flies. FDR, P < 0.05. n = 5 independent samples per group.