Mitochondrial chaperone TRAP1 activates the mitochondrial UPR and extends healthspan in Drosophila

Abstract

The molecular mechanisms influencing healthspan are unclear but mitochondrial function, resistance to oxidative stress and proteostasis are recurring themes. Tumor necrosis factor Receptor Associated Protein 1 (TRAP1), the mitochondrial analog of Hsp75, regulates levels of reactive oxygen species in vitro and is found expressed at higher levels in tumor cells where it is thought to play a pro-survival role. While TRAP1-directed compartmentalized protein folding is a promising target for cancer therapy, its role at the organismal level is unclear. Here we report that overexpression of TRAP1 in Drosophila extends healthspan by enhancing stress resistance, locomotor activity and fertility while depletion of TRAP1 has the opposite effect, with little effect on lifespan under both conditions. In addition, modulating TRAP1 expression promotes the nuclear translocation of homeobox protein Dve and increases expression of genes associated with the mitochondrial unfolded protein response (UPR(mt)), indicating an activation of this proteostasis pathway. Notably, independent genetic knockdown of components of the UPR(mt) pathway dampen the enhanced stress resistance observed in TRAP1 overexpression flies. Together these studies suggest that TRAP1 regulates healthspan, potentially through activation of the UPR(mt).

Keywords: Ageing; Drosophila; Healthspan; Mitochondria; TRAP1; UPR.

Figure 1. TRAP1 regulates ROS levels in vivo

(A–C) Optic lobes of young adult Drosophila brains stained with Mitosox, and (D–F) larval segmental nerves of corresponding genotypes stained with antibody against mitochondrial Complex V. (A and D)w¹¹¹⁸; +/+; +/+(B and E)w; TRAPΔ4/TRAPΔ4; +/+, (C and F) w; +/+; UAS-TRAP1^7M/ActinGal4. (G) MitoSox staining is significantly increased in w; TRAPΔ4/TRAPΔ4; +/+ mutants. While staining is marginally reduced in overexpression flies w; +/+; UAS-TRAP1^7M/ActinGal4, it is not significantly different from wildtype. Error bars denote standard deviation of means. N = 3–5. (**) indicate p < 0.001 from Student’s t-test. Mitochondrial distribution appears normal in mutant and overexpression animals in segmental nerves of Drosophila larvae. In all cases, scale bars equal 10 µm.

Figure 2. TRAP1 regulates oxidative stress resistance in young and old Drosophila males

(A–E) Survival curves of young flies (5 days) of indicated genotypes exposed to 20 mM paraquat. (A) Median survival of male w; TRAP1Δ4/ TRAP1Δ4; +/+ (18.5 hrs ± 0.9) is significantly less than that of w¹¹¹⁸; +/+; +/+ (28.4 hrs ± 1.6; p<0.0001). (B) Median survival of male w; +/+; UAS-TRAP1^7M/ActinGal4 (63.2 hrs ± 4.8) is significantly more than that of w; +/+; UAS-TRAP1^7M/ + (43.9 hrs ± 3.2; p=0.0002), and that of w; +/+; ActinGal4 (36.2 ± 2.8; p<0.0001). (C) Median survival of female w; TRAP 1A4/TRΔP 1Δ4; +/+ (20.3 hrs ± 1.4) is comparable to that of w¹¹¹⁸; +/+; +/+ (22.4 hrs ± 1.4; p=0.4). (D) Median survival of female w; +/+; UAS-TRAP1^7M/ActinGal4 (43.9 hrs ± 3.3) is significantly higher than that of w; +/+; ActinGal4/+ (32.4 hrs ± 2.2; p=0.004) but comparable to that of w; +/+; UAS-TRAP1^7M/ + (36.9 hrs ± 2.7; p=0.1). (E) Median survival of young w; TRAP 1A4/TRAP 1A4; UAS-TRAP1^7M/ActinGal4 males (25.26 hrs ± 1.88) is significantly higher than w; TRAP 1Δ4/TRAP 1Δ4; +/+ (15 hrs ± 0.4; p<0.0001). (F) Median survival is increased in old (40 days) w; +/+; UAS-TRAP1^7M/ GS-tub5Gal4 males maintained on 5 mM RU-486 (41.47 hrs ± 2.99), relative to control flies of the same genotype, sex and age maintained on vehicle (32.88 ± 1.78; p = 0.03). In all cases, errors denote standard deviation. Statistical significance was determined using the standard chi-squared based log-rank test.

Figure 3. TRAP1 confers resistance against heat stress

Fly survival at 6 days and 19 days of age, 24 hrs after each heat stress is delivered. Data is presented as the percentage of flies subjected to the stress that survives. (A) On the 6^th day, survival of male w; TRAP 1Δ4/TRAP 1Δ4; +/+ (75 ± 6%) trends towards being lower than w¹¹¹⁸; +/+; +/+ (81 ± 8 %; p=0.05), but is significantly decreased (7.14 ± 0.66%) at 20 days of age relative to control (37.41 ± 0.77%; p<0.001). (B) Survival of female w; TRAP 1Δ4/TRAP 1Δ4; +/+ (93 ± 1%) is comparable to that of w¹¹¹⁸; +/+; +/+ (97 ± 1%; p=0.1) on the 6^th day, but significantly decreased at 20 days (32.05 ± 7%) compared to control (85.62 ± 3.5%; p<0.001). (C) On 6^th day, survival of male w; +/+; UAS-TRAP1^4M/ActinGal4 (96.5 ± 1.5%) is significantly higher than w; +/+; ActinGal4/+ (81 ± 6%; p=0.01), but comparable to w; +/+; UAS-TRAP1^4M/ + (93.5 ± 0.5%; p=0.1). However, on the 20^th day, survival of male w; +/+; UAS-TRAP1^4M/ActinGal4 (90.64 ± 1%) is remarkably higher than w; +/+; ActinGal4/+ (29.71 ± 2.16; p<0.0001), and w; +/+; UAS-TRAP1^4M/ + (57.14 ± 0.75%, p<0.001). (D) Survival of female w; +/+; UAS-TRAP1^4M/ActinGal4 (97.5 ± 1.5%) is not significantly different from w; +/+; ActinGal4/+ (94 ± 1%; p=0.1), and w; +/+; UAS-TRAP1^4M/ + (97 ± 1%; p=0.2) at 6 days. However at 20 days, survival of w; +/+; UAS-TRAP1^4M/ActinGal4 (94.62 ± 2%) is significantly more than w; +/+; ActinGal4/+ (87.18 ± 2.5%; p<0.001), and w; +/+; UAS-TRAP1^4M/ + (89.18 ± 2.5%; p=0.01). In all cases, errors denote standard errors of mean. Statistical significance was determined using one-way ANOVA.

Figure 4. TRAP1 has a marginal influence on lifespan

Lifespan curves of indicated genotypes. (A) Median lifespan of male w¹¹¹⁸; +/+; +/+ (60.4 ± 1.4 days) is comparable to that of w; TRAP 1Δ4/TRAP 1Δ4; +/+ (64.9 ± 1.1 days; p=0.48). (B) Median lifespan of female w¹¹¹⁸; +/+; +/+ (64.6 ± 1.44 days) is comparable to w; TRAP 1Δ4/TRAP 1Δ4; +/+ (64.32 ± 0.98 days; p=0.08). (C) Median lifespan of male w; +/+; UAS-TRAP1^7M/ActinGal4 (58.6 ± 1.3 days) is significantly more than w; +/+; ActinGal4/+ (40.7 ± 1.4 days; p<0.001), and marginally more than that of w; +/+; UAS-TRAP1^7M/ + (56.5 ± 1.1 days; p=0.05). (D) Median lifespan of female w; +/+; UAS-TRAP1^7M/ActinGal4 (75.5 ± 1.4 days) is significantly more than that of w; +/+; ActinGal4/+ (53.5 ± 1.2 days; p<0.001), and marginally more than that of w; +/+; UAS-TRAP1^7M/ + (66.5 ± 1.9; p=0.044). (E) Median lifespan of w; +/+; UAS-TRAP1^4M/ GS-tub5Gal4 males maintained on RU-486 (43.66 ± 0.88 days) is similar to control flies of the same genotype, sex and age maintained on vehicle (44.86 ± 0.64 days; p = 0.15). (F) Median lifespan of w; +/+; UAS-TRAP1^4M/ GS-tub5Gal4 females on RU-486 (47.24 ± 0.97 days) is also comparable to controls (44.84 ± 0.81 days; p=0.04). In all cases, errors denote standard deviation. Statistical significance was determined using the standard chi-squared based log-rank test.

Figure 5. TRAP1 modulates healthspan

(A and B) TRAP1 mutant flies tend to climb more poorly than controls in the negative geotaxis assay and the difference becomes significant at older ages. There is a significant drop in performance of mutant females at 30 days. By 60 days, fewer male and female w; TRAP1A4/TRAP1A4; +/+ flies climbed successfully as compared to w¹¹¹⁸; +/+; +/+(C and D) Locomotor performance after 40 days of age is well maintained in w; +/+; UAS-TRAP1^7M/ActinGal4 males and females on ubiquitous overexpression of TRAP1. (E)w; TRAP1Δ4/TRAP1Δ4; +/+ females produce fewer progeny than controls at all stages of their life. (F)w; +/+; UAS-TRAP1^7M/ActinGal4 females produce more progeny than controls at all stages of their life. (G and H) Male and female flies overexpressing TRAP1 exclusively in the nervous system displayed significant improvement in locomotor ability at older age-points in w; +/+; UAS-TRAP1^4M/ elavGal4. In all cases error bars denote standard error of means; (*) indicates p < 0.05, (**) indicates p < 0.001. Statistical significance was determined by one-way ANOVA with Dunnett’s post hoc comparison. (See also Figure S 5)

Figure 6. Loss and overexpression of TRAP1 activates the UPR^mt pathway

(A – C) Anti-Dve antibody (green) reveals tight cytosolic localization of Dve in neuronal cell bodies in brains of adult w¹¹¹⁸; +/+; +/+ males. Nuclear dye DAPI (red) stains the entire nucleus although it appears brighter in the nucleolus. The merged image of anti-Dve and DAPI stain reveals crisp cytosolic sub-cellular localization of Dve. (D – F) Dve staining appears diffuse in neuronal cell bodies of male w; TRAP1A4/TRAP1A4, +/+ flies and exhibits considerable overlap with the nucleus. (G – I) Dve staining appears similarly diffused and overlaps with nuclear stain in neuronal cell bodies of male w; +/+; UAS-TRAP1^4M/ActinGal4 flies. Scale bar equals 3.5 µm. (J) Ratio of staining intensity of Dve in the cytoplasm relative to the nucleus in indicated genotypes. (K) mRNA expression of Dve in w¹¹¹⁸; +/+; +/+ females is not significantly different from mutant (p=0.2) and overexpression (p=0.9) strains. (L, M and N) Male w; TRAP1Δ4/TRAP1Δ4, +/+ and w; +/+; UAS-TRAP1^7M/ActinGal4 flies exhibit significantly increased mRNA expression of Hsp60, mtHsp70 and CG5045 as compared to w¹¹¹⁸; +/+; +/+. (O, P and Q) Female w; TRAP1A4/TRAP1A4; +/+ exhibit significantly increased mRNA expression of Hsp60 as compared to w¹¹¹⁸; +/+; +/+, but not that of mtHsp70 and CG5045. In all cases, data are presented as fold change from control. Error bars denote standard error of means; statistical significance was determined using one-way ANOVA. (*) indicates p < 0.05, (**) indicates p < 0.001.

Figure 7. Activation of the UPR^mt is required for TRAP1 mediated heat stress resistance in older males

Fly survival at 6 days and 21 days of age, 24 hrs after each heat stress is delivered. (A) In 6-day old male flies with conditional Gal4, survival of w; +/+; UAS TRAP1^4M/ GS-Tub5Gal4 +/+ in control group (97 ± 4.2%) is comparable to identical flies maintained on RU-486 (99 +/− 1.4%; p=0.6), w; UAS CG5045 RNAi/+; UAS TRAP1^4M/ GS-Tub5Gal4 in control group (100%) are the same as identical flies maintained on RU-486 (100%), and w; UAS DVE RNAi/+; UAS TRAP1^4M/ GS-Tub5Gal4 in control group (100%) are the same as identical flies maintained on RU-486 (100%). (B) In 6-day old female flies with conditional Gal4, survival of w; +/+; UAS TRAP1^4M/ GS-Tub5Gal4 +/+ in control group (100%) is the same as identical flies maintained on RU-486 (100%), w; UAS CG5045 RNAi/+; UAS TRAP1^4M/ GS-Tub5Gal4 in control group (100%) is the same as identical flies maintained on RU-486 (100%), and w; UAS DVE RNAi/+; UAS TRAP1^4M/ GS-Tub5Gal4 in control group (100%) are the same as identical flies maintained on RU-486 (100%). (C) In 21-day old male flies with conditional Gal4, survival of w; +/+; UAS TRAP1^4M/ GS-Tub5Gal4 +/+ of flies maintained on RU-486 (85 +/− 4.2%) is significantly higher than controls (70 ± 2.8%; p = 0.05). However, w; UAS CG5045 RNAi/+; UAS TRAP1^4M/ GS-Tub5Gal4 in control group (73 +/− 9.8%) is comparable as identical flies maintained on RU-486 (76 +/− 5.6%; p = 0.7), and w; UAS DVE RNAi/+; UAS TRAP1^4M/ GS-Tub5Gal4 in control group (69 +/−1.4%) are similar to identical flies maintained on RU-486 (71 +/− 2.8%; p = 0.8). (D) In 21-day old female flies with conditional Gal4, survival of w; +/+; UAS TRAP1^4M/ GS-Tub5Gal4 +/+ in control group (95 +/− 1.4%) is comparable to identical flies maintained on RU-486 (92 +/− 5.6%; p = 0.5), w; UAS CG5045 RNAi/+; UAS TRAP1^4M/ GS-Tub5Gal4 in control group (95 +/− 1.5%) is similar to identical flies maintained on RU-486 (96 +/− 5.6%; p = 0.8), and w; UAS DVE RNAi/+; UAS TRAP1^4M/ GS-Tub5Gal4 in control group (91 +/−1.4%) are the comparable to identical flies maintained on RU-486 (88 +/− 2.8%; p = 0.3). In all cases, errors denote standard deviation. Statistical significance was determined using one-way ANOVA.