Age-dependent H3K9 trimethylation by dSetdb1 impairs mitochondrial UPR leading to degeneration of olfactory neurons and loss of olfactory function in Drosophila
- Center for Integrative Biology, Faculty of Sciences, Universidad Mayor, Chile
- Center for Aging Research and Healthy Longevity, Faculty of Sciences, Universidad Mayor, Chile
- Geroscience Center for Brain Health and Metabolism (GERO), Chile
- Center for Resilience, Adaptation and Mitigation, Universidad Mayor, Chile
- Buck Institute for Research on Aging, United States
- Centro Científico y Tecnológico de Excelencia Ciencia and Vida, Fundación Ciencia and Vida, Chile
Figures
Figure 1 with 3 supplements
UPRMT-dependent activation of the Hsp60::dsRed reporter in the antennal lobe (AL) of Drosophila.
(A) Representative confocal images of whole Drosophila brains flies expressing the UPRMT reporter Hsp60::dsRed (magenta). Flies were treated for 48 hr with vehicle, paraquat (paraquat PQ, 10 µM), or doxycycline (Doxy, 100 µM). The white box indicates the quantified region of the AL. Scale bar, 20 µm. (B) Confocal images of the whole Drosophila brains expressing the UPRMT reporter Hsc70-5::dsRed (magenta) under the same treatment conditions as in (A). (C) Quantification of Hsp60::dsRed integrated density in the AL. (n=6, Veh vs. PQ: p=0.0003 and Veh vs. Doxy: p=0.0024).(D) Quantification of Hsc70-5::dsRed integrated density in the AL. (n=5, Veh vs. PQ: p=0.0198 and Veh vs. Doxy: p=0.0184). (E) Hsp60::dsRed in AL with pan-neuronal RNAi knockdown of dve, ubl, crc, post treatment with PQ or vehicle; magenta shows the Hsp60A-dsRed signal, cyan for DAPI. Scale bar, 20 µm. (F) Integrated density of Hsp60::dsRed, Two-way ANOVA Dunnett test between veh vs PQ treated flies: Control (p=0.0038, n=8), dve (p>0.9999, n=8), crc (p>0.9999, n=8), ubl (p>0.9999, n=8). (G) Representative images Hsp60::dsRed reporter in GFP-labeled neurons of the AL of flies treated with 10 mM PQ or vehicle for 48 hr. Hsp60::dsRed in magenta, Pan-neuronal GFP in green, and DAPI in cyan. Scale bar, 20 µm. (H) Integrated density of Hsp60::dsRed in GFP-labeled neurons from 0 (white bars) to 45 dpe (red bars). Two-way ANOVA with Dunnett’s multiple comparisons between ages for vehicle (p=0.2319, n=7) and for PQ treatment (p=0.0017, n=7). For the difference between treatments: vehicle and PQ at 0 dpe (p=0.0404, n=7), and at 45 dpe (p=0.9581, n=7). (I) Neuronal volume (µm 3) in AL from 0 (white bars) to 45 dpe (red bars). Two-way ANOVA with Dunnett’s multiple comparisons between 0 and 45 dpe flies for vehicle (p=0.4255, n=7) and PQ (p=0.5076, n=7). Comparing vehicle to PQ treatment showed no significant difference at 0 dpe (p=0.1133, n=7) and 45 dpe (p=0.0857, n=7). (J) Dot plot visualizing vAChT (blue), vGlut (green), and Gad1 (red) for UPRMT activation analysis via single-cell RNA seq data. (E) AUC scores of Hsp60 expression: vAChT (0 vs 50 dpe, p=0.0049, n=576), vGlut (0 vs 50 dpe, p=0.9998, n=168), and Gad1 (0 vs 50 dpe, p=0.0191, n=168). (F) AUC scores of Hsc70-5 expression: vAChT (0 vs 50 dpe, p<0.0001, n=576), vGlut (0 vs 50 dpe, p<0.0001, n=168), and Gad1 (0 vs 50 dpe, p=0.9914, n=168). For panels C–I, each n represents one brain from an individual animal. For panels J–L, each n represents a single cell. Bars represent mean ± SEM. Statistical significance is denoted as ****p<0.0001; ***p<0.001; **p<0.01; *p<0.05; ns>0.05.
Figure 1—figure supplement 1
Quantification of the fluorescent signal of Hsp60::dsRed reporter using 3D reconstructed surface in the antennal lobe of Drosophila brain.
(A) Localization of the region of interest (ROI) in the Drosophila brain. Left: Schematic diagram of the brain highlighting the antennal lobe (AL, green box). Right: Representative confocal image of a whole brain expressing the Hsp60::dsRed reporter (magenta). The white box delineates the AL, which corresponds to the ROI used for subsequent image analyses. Scale bar, 50 µm. (B) Quantification of Hsp60::dsRed and Xbp1::GFP signals alone: For experiments shown in Figures 1A, B, E, 2B, D, a 100 μm2 ROI was selected around the AL. 3D reconstruction and surface rendering of the fluorescent signal were performed using Imaris software. The integrated density, representing the cumulative fluorescence signal within the ROI, was calculated by multiplying the average fluorescence intensity by the volume of the signal-bearing domain. (C) Quantification of Hsp60::dsRed signal in GFP-labeled neurons: For experiments shown in Figures 1G and 4D, the GFP signal was used to define the volume of labeled neurons in the AL. 3D reconstruction and surface rendering of the GFP signal were performed, and the integrated density of the Hsp60::dsRed signal within the GFP-labeled volume was quantified. This approach normalizes the Hsp60::dsRed signal to the GFP signal, accounting for variations in cell number and volume. (D) Quantification of H3K9me3 signal in GFP- and ToPro3-labeled OPNs: For experiments shown in Figure 4D. To quantify the specific signal in olfactory projection neurons (OPNs), a surface of GFP-labeled OPNs was rendered. Then, the To-Pro3-labeled nuclei signal was masked and its surface rendered. Finally, the H3K9me3 signal within the OPN nuclei was quantified using integrated density as described previously. Scale bar: 10 µm.
Figure 1—figure supplement 2
The Hsp60A regulatory region::dsRed transcriptional reporter is induced by mitochondrial stress in vivo.
(A) Schematic of the endogenous Hsp60A locus (RefSeq NM_078560.3). Orange arrows indicate the annotated Hsp60A exons in the reference gene model. (B) Diagram of the transgenic construct used to generate the UPRMT reporter. The magenta arrow represents the full Hsp60A::dsRed reporter cassette (Hsp60A 5′ regulatory fragment driving the dsRed reporter sequence; see Methods). The red arrow denotes the dsRed-Express coding sequence within the cassette. (C) Confocal images of adult brain tissue expressing Mito::GFP (green), Hsp60A::dsRed (red), and DAPI (cyan), under vehicle or paraquat (PQ, 10 mM) treatment. Merged images and colocalization panels (Coloc). Coloc panels show the colocalized signal as a mask of overlapping voxels between Mito::GFP and Hsp60A::dsRed and are displayed as a single channel. Scale bar: 10 µm. (D) Confocal images of adult gut expressing Mito::GFP (green), Hsp60A::dsRed (red), and DAPI (cyan), under vehicle or PQ (10 mM) treatment. Merged images and Coloc panels. Scale bar: 10 µm. (E) Quantification of Hsp60A::dsRed integrated fluorescence intensity in the brain, normalized to vehicle-treated controls (PQ vs vehicle, p<0.05). (F) Manders’ coefficient of colocalization between Mito::GFP and Hsp60A::dsRed in the brain (PQ vs vehicle, p<0.01). (G) Quantification of Hsp60A::dsRed integrated fluorescence intensity in the gut, normalized to vehicle-treated controls (PQ vs vehicle, p<0.05). (H) Manders’ coefficient of colocalization between Mito::GFP and Hsp60A::dsRed in the gut (PQ vs vehicle, p<0.05). Bars represent mean ± SEM, n=5–6 brains or gut of individual animals per group. Unpaired t-test: *p<0.05, **p<0.01.
Figure 1—figure supplement 3
The Hsp60::dsRed reporter is specifically activated by mitochondrial stress, not by endoplasmic reticulum (ER) stress.
(A) Representative confocal images of the antennal lobe (AL) from 0-day-old flies expressing the UPRMT reporter Hsp60::dsRed (magenta). Flies were treated for 48 hr with vehicle, paraquat (PQ, 10 µM), or the ER stressor tunicamycin (Tm, 0.25 µg/µl). Nuclei are stained with DAPI (cyan). Scale bar, 20 µm. (B) Quantification of Hsp60::dsRed integrated density for the conditions shown in (A). (One-way ANOVA with Bonferroni’s multiple comparisons test; vehicle vs. PQ, p=0.0061; PQ vs. Tm, p=0.0119; vehicle vs. Tm, p=0.9863. n=9 for vehicle, n=11 for PQ, n=8 for Tm). (C) Representative confocal images of the AL from flies expressing the UPRER reporter Xbp1::GFP (green), treated with vehicle or Tm. Nuclei are stained with DAPI (cyan). Scale bar, 20 µm. (D) Quantification of Xbp1::GFP integrated density for the conditions shown in (C). (Unpaired t-test; p=0.0231. n=8 for vehicle, n=7 for Tm). Data are presented as mean ± SEM. Each dot represents one biological replicate (a single antennal lobe). Statistical significance is denoted as ***p<0.001; *p<0.05; ns, not significant.
Figure 2
dSetdb1 negatively regulates UPRMT in aging through increasing H3K9me3 levels in antennal lobe (AL) of Drosophila.
(A) Western blot for H3K9me3 levels with pan-neuronal downregulation of dSetdb1. Two-way ANOVA Bonferroni’s multiple comparisons test results: Control at 0 vs 45 days post eclosion (dpe) (p=0.0255, n=3), dSetdb1 RNAi at 0 vs 45 dpe (p=0.4951, n=3), Control vs. dSetdb1 RNAi at 0 dpe (p>0.9999, n=3), and at 45 dpe (p=0.4556, n=3). n=20 fly heads. (B) Representative confocal images of AL from UPRMT reporter flies with dSetdb1 loss of function (lof), displaying Hsp60A::dsRed in magenta. Scale bar, 20 µm. (C) Normalized integrated density of Hsp60A::dsRed. Two-way ANOVA Bonferroni’s multiple comparisons test between 0 vs 45 dpe within dSetb1 l of genotype (p<0.0001, n=10) and control (p>0.9999, n=10). (D) Confocal images of Hsc70-5::dsRed in AL from flies at 0 and 45 dpe with dSetdb1 loss of function. Scale bar, 20 µm. (E) Integrated density of Hsc70-5::dsRed normalized to control values. Two-way ANOVA Bonferroni’s multiple comparisons between 0 vs 45 dpe within dSetdb1 Lof genotype (p<0.0001, n=8) and control (p=0.8633, n=8). (F) Dot plot from Scope single-cell RNA-seq analysis depicting neuronal types: vAChT (blue), vGlut (green), and Gad1 (red). (G) AUCell scores for dSetdb1 expression in single neurons at 0, 15, 30, and 50 dpe. One-way ANOVA with Dunnett’s multiple comparison against 0 dpe. vAChT neurons: 50 dpe (p=0.2906, n1=2932, n2=656). vGlut neurons: 50 dpe (p=0.7386, n1=712, n2=172). Gad1 neurons: 50 dpe (p=0.8916, n1=576, n2=197). (H) AUCell scores for utx expression in single neurons at 0, 15, 30, and 50 dpe. One-way ANOVA with Dunnett’s multiple comparison against 0 dpe. vAChT neurons: 50 dpe (p=0.0001, n1=2932,, n2=656). vGlut neurons: 50 dpe (p=0.321, n1=712, n2=172). Gad1 neurons: 50 dpe (p=0.0688, n1=576, n2=197). (I) AUCell scores for kdm2 expression in single neurons at 0, 15, 30, and 50 dpe. One-way ANOVA with Dunnett’s multiple comparison against 0 dpe. vAChT neurons: 50 dpe (p<0.0001, n1=2932, n2=656). For vGlut neurons: 50 dpe (p=0.321, n1=712, n2=172). For Gad1 neurons: 50 dpe (p=0.667, n1=576, n2=197). For panels C and E, each n represents one brain from an individual animal. For panels G–I, n1 and n2 represent single cells from each age group. All error bars represent mean ± SEM. P-value: ****p<0.0001; ***p<0.001; **p<0.01, *p<0.05 and ns>0.05.
-
Figure 2—source data 1
PDF file containing original western blots for Figure 2A, indicating the relevant bands and treatments.
- https://cdn.elifesciences.org/articles/103118/elife-103118-fig2-data1-v2.zip
-
Figure 2—source data 2
Original files for western blot analysis displayed in Figure 2A.
- https://cdn.elifesciences.org/articles/103118/elife-103118-fig2-data2-v2.zip
Figure 3 with 1 supplement
dSetdb1 pan-neuronal downregulation preserves olfactory function in aging.
(A) Olfactory T-maze was used to perform the olfactory behavioral test. Flies are presented to an experimental odor or vehicle. Flies have 60 s to discriminate between odors and go to an arm of the T-maze. At the end of the time, an image is acquired, and the preference index is calculated for every trial; every dot corresponds to 10 trials of 15 flies each. (B) Olfactory preference index shows the aging-associated functional decline in the olfactory system. (C) Olfactory preference index in flies with pan-neuronal downregulation of dve, ubl, and crc. n=3 populations of 10 flies each. (D) Western blot analysis of H3K9me3 levels in young flies with pan-neuronal downregulation of dSetdb1 (green), utx (gray), and kdm2 (calypso). Each n represents homogenized pools of 20 fly heads. One-way ANOVA Bonferroni’s multiple comparisons test results: Control vs. dSetdb1 RNAi (p=0.9738, n=3), Control vs. utx RNAi (p=0.0391, n=3), Control vs. kdm2 RNAi (p=0.1951, n=3). (E) Olfactory preference indices for 0 and 45 days post eclosion (dpe) flies with downregulation of dSetdb1, utx, and kdm2, exposed to odors 3-octanol (-) and 2,3-butanedione (+). Two-way ANOVA Bonferroni’s test results for 3-octanol (-), control at 0 vs 45 dpe (p<0.0001, n=3), dSetdb1 (p=0.0624, n=3), utx (p=0.1356, n=3), kdm2 (p=0.298, n=3); for 2,3-butanedione (+), control (p<0.0001, n=3), dSetdb1 (p=0.1356, n=3), utx (p=0.1374, n=3), kdm2 (p>0.9999, n=3). All error bars represent mean ± SEM.
-
Figure 3—source data 1
PDF file containing original western blots for Figure 3D, indicating the relevant bands and treatments.
- https://cdn.elifesciences.org/articles/103118/elife-103118-fig3-data1-v2.zip
-
Figure 3—source data 2
Original files for western blot analysis displayed in Figure 3D.
- https://cdn.elifesciences.org/articles/103118/elife-103118-fig3-data2-v2.zip
Figure 3—figure supplement 1
dSetdb1 pan-neuronal downregulation does not increase max survival, but increases healthspan (A) Survival curve of flies bearing pan-neuronal downregulation of dSetdb1 (green), utx (gray), and kdm2 (calypso) compared to control (black).
Statistical analysis: Gehan-Breslow-Wilcoxon test for survival curve comparison, followed by one-way ANOVA with Dunnett’s multiple comparisons test against the control genotype. P-value: ****p<0.0001; ***p<0.001; **p<0.01, *p<0.05, ns >0.05. Pairwise comparisons using the Gehan-Breslow-Wilcoxon test revealed a significant difference in survival between control and dSetdb1 RNAi flies (p=0.0002), but no significant differences between control and utx RNAi (p=0.1803) or control and kdm2 RNAi (p=0.1677). Each curve represents a cohort of 100 individual flies.
Figure 4 with 2 supplements
dSetdb1 downregulation preserves olfactory projection neurons (OPNs) function in aging by reducing H3K9me3 and enabling UPRMT.
(A) Representative images of GFP-tagged OPNs in the antennal lobe (AL) of 0 and 45 days post eclosion (dpe) flies bearing the downregulation of dSetdb1 under the control of the GH146 driver. Nuclei are in cyan (ToPro3), H3K9me3 is in yellow, and the right panel is a merge of three channels with OPNs in green. Scale bar, 20 µm. (B) Analysis of H3K9me3 integrated density in the nuclei of GFP-tagged OPNs across aging in Drosophila with dSetdb1 knockdown. Using two-way ANOVA with Bonferroni’s correction, control flies showed a significant reduction in density from 0 to 45 dpe (p=0.0128, n=7), a difference lost at dSetdb1 RNAi flies (p=0.5022, n=7). (C) Quantification of H3K9me3 volume, specifically in nuclei of GFP signal, normalized to control of 0 dpe untreated flies. White bars represent 0 dpe flies and red bars 45 dpe flies, n=7. Two-way ANOVA Bonferroni’s multiple comparisons showed for control flies of 0 vs 45 dpe p=0.0057; or dSetdb1 RNAi group of 0 vs 45 dpe (p>0.9999), Control vs dSetdb1 RNAi at 0 dpe p=0.0604, for 45 dpe, Control vs dSetdb1 RNAi p=0.0004 (D) Representative images of OPNs labeled with GFP (green) bearing the UPRMT reporter Hsp60::dsRed (magenta) in the AL of 0 and 45 dpe flies treated with paraquat (PQ) or vehicle for 48 hr. Scale Bar, 20 µm. (E) Quantification of Integrated density of Hsp60A::dsRed specifically in the GFP-labeled neurons in the AL of 0 and 45 dpe flies treated with vehicle or PQ 10 mM for 48 hr. Two-way ANOVA Bonferroni’s test result at 0 dpe between Control Vehicle vs. Control PQ p=0.0027, n=9; Control Vehicle vs. dSetdb1 Veh p=0.0291, n=9; Control Vehicle vs. dSetdb1 PQ p=0.0003, n=9 and at 45 dpe between Control Vehicle vs. Control PQ p=0.9872, n=9; Control Vehicle vs. dSetdb1 Veh p=0.0154, n=9; Control Vehicle vs. dSetdb1 PQ p<0.0001, n=9. (F) Dot plot of expression cluster showing specifically OPN cluster in red. One-way ANOVA with Dunnett’s multiple comparisons test against 0 dpe was performed for (G) Hsp60A Expression: At 0 vs 50 dpe p=0.0102, n1=60, n2=38, and (I) Hsc70-5 expression 0 vs 50 dpe p<0.0001, n1=60, n2=38. (I–K) Expression levels of dSetdb1, kdm2, and utx in OPNs through aging. One-way ANOVA with Dunnett’s multiple comparisons test against 0 dpe was performed, (I): dSetdb1 expression 0 vs 50 dpe (p=0.0102, n1=84, n2=56). (J) utx expression 0 vs 50 dpe (p=0.9996, n1=84, n2=56). (K) kdm2 0 vs 50 dpe (p=0.0425, n1=84, n2=56). Each n represents the AUC values for expression in a single cell. (L) Olfactory preference index of flies bearing the GH146 Gal4 driven knockdown of dSetdb1, dve, crc, and the double knockdown of dSetdb1/dve and dSetdb1/crc, respectively. White bars are 0 dpe flies and red bars are 45 dpe flies. n=3. Results from a Two-way ANOVA Bonferroni’s multiple comparisons test are as follows: Control at 0 vs 45 dpe: p<0.0001, n=3, dSetdb1 at 0 vs 45 dpe: p>0.9999, n=3; dve at 0 vs 45 dpe: p>0.9999, n=3; dSetdb1/dve at 0 vs 45 dpe: p>0.9999, n=3; crc at 0 vs 45 dpe: p>0.9999, n=3; dsetdb1/crc at 0 vs 45 dpe: p=0.2606, n=3. P-value: ****p<0.0001; ***p<0.001; **p<0.01, *p<0.05 and ns >0.05. For panels B, C, and E, each n represents one brain from an individual animal. For panels G–K, n1 and n2 represent single cells from each age group. In panel L, each n represents one population of 10 flies. All error bars represent mean ± SEM.
Figure 4—figure supplement 1
Validation of the dSetdb1 function in aged flies using independent genetic tools.
Olfactory preference index (PI) for an aversive odorant (3-octanol, left) and an attractive odorant (2,3-butanedione, right) in 45-day-old flies. Genotypes tested were control, dSetdb1 RNAi (TRiP.JF01310), an independent short-hairpin RNAi line (dSetdb1 RNAishort, TRiP. HMS00112), and a loss-of-function allele (dSetdb1 lof, egg²³⁵). Comparisons to control for 3-octanol were: dSetdb1 RNAi (p<0.0001), dSetdb1 RNAishort (p<0.0001), and dSetdb1 lof (p=0.0002). Comparisons to control for 2,3-butanedione were: dSetdb1 RNAi (p=0.0006), dSetdb1 RNAishort (p=0.0003), and dSetdb1 lof (p=0.0005). Data are presented as mean ± SEM; n=3 biological replicates of ten flies per group. Statistical significance was determined by one-way ANOVA with Tukey’s multiple comparisons test. ****p<0.0001; ***p<0.001.
Figure 4—figure supplement 2
Olfactory projection neuron (OPN)-specific dSetdb1 knockdown restores age-dependent UPRMT activation.
(A) Representative confocal images of olfactory projection neurons (OPNs; GH146>GFP, green) expressing the mitochondrial stress reporter Hsc70-5::dsRed (magenta). Images compare young (0 days post-eclosion, dpe) and aged (45 dpe) flies under vehicle or paraquat (PQ, 10 µM) treatment. Scale bar, 10 µm. (B) Quantification of Hsc70-5::dsRed integrated density in GFP-positive OPNs. (n=7 brains per condition; one-way ANOVA with Tukey’s multiple comparisons test). Comparisons shown are Control Veh vs. Control PQ at 0 dpe (p=0.0416); Control PQ at 0 dpe vs. 45 dpe (p=0.0004); Control vs. dSetdb1 RNAi under PQ at 45 dpe (p=0.0243); and Control vs. dSetdb1 RNAi under vehicle at 45 dpe (p=0.0003). Data are presented as mean ± SEM. Each n represents a brain from an individual animal . Statistical significance is denoted as ***p<0.001; *p<0.05; ns, not significant.
Figure 5 with 2 supplements
dSetdb1 downregulation mitigates age-related mitochondrial oxidation and preserves mitochondrial morphology in OPNs.
(A) Scheme of GH146-driven GFP expression in olfactory projection neurons (OPNs), and mitochondrial GFP reporter expressed in those neurons. Neurons possess their soma in the antennal lobe (AL) and project their axons to the synapsis-enriched zone, the lateral horn (LH). Scale bar, 100 µm. (B) Mitochondria labeled with GFP in the AL of OPNs of 0 and 45 days post eclosion (dpe) flies bearing the dSetdb1 RNAi; mitochondria: green, nuclei: cyan. Scale bar, 5 µm. (C) Mitochondrial integrated density in the AL of 0 and 45 dpe flies with dSetdb1 knockdown analyzed via Two-way ANOVA with Bonferroni’s multiple comparisons test. Control flies showed a significant difference at 0 vs 45 dpe (p=0.0018, n=9), while dSetdb1 RNAi flies showed no significant change (p=0.0855, n=9). (D) Analysis of total mitochondrial volume in the AL of OPNs with dSetdb1 knockdown compared to controls at 0 and 45 dpe. Two-way ANOVA with Bonferroni’s multiple comparisons test indicates a decrease in mitochondrial volume in control flies from 0 to 45 dpe (p<0.0001, n=9). dSetdb1 RNAi of 0 vs 45 dpe flies did not show a change in volume (p=0.2387, n=9). (E) AL mitochondrial fragmentation index of images shown in B. Two-way ANOVA Bonferroni’s multiple comparisons test for mitochondrial fragmentation index in control flies from 0 to 45 dpe (p=0.0145, n=9) and dSetdb1 RNAi flies showed no significant change (p>0.9999, n=9). Control vs. dSetdb1 RNAi at 0 dpe (p=0.0101, n=9); no significant change at 45 dpe (p>0.9999, n=9). (F) AL sphericity index of mitochondria from images shown in B. Graph shows results of Two-way ANOVA with Bonferroni’s multiple comparisons test. For 0 vs 45 dpe, control flies (p=0.0019, n=9) and dSetdb1 RNAi flies (p=0.0257, n=9). At 0 dpe, control vs dSetdb1 RNAi flies (p=0.012, n=9), and at 45 dpe (p=0.0008, n=9). (G) Representative images of axonal mitochondria in the green of 0 and 45 dpe flies bearing the knockdown of dSetdb1. Scale bar, 5 µm. (H) Axonal MitoGFP integrated density; control increase (p=0.0005, n=9), dSetdb1 RNAi (p=0.754, n=9). (I) LH MitoGFP confocal images. Scale bar, 10 µm. (J) LH MitoGFP integrated density; control (p=0.1197, n=9), dSetdb1 RNAi (p=0.0751, n=9). (K) Representative images of 0 and 45 dpe Drosophila’s AL showing GH146-GAL4;UAS-MitoTimer, an in vivo mitochondrial oxidation reporter. Reduced (green) and oxidized (red) MitoTimer signals are shown, and merge (yellow). Scale bar, 5 µm. (L) AL Red/Green integrated density ratio; 0 vs 45 comparison reports a significant oxidation increase in control flies (p<0.0001; n=8), while dSetdb1 RNAi flies show a non-significant change (p=0.4359; n=8). Control vs dSetdb1 RNAi at 0 dpe (p=0.2618; n=8) and at 45 dpe (p=0.0143; n=8). (M) Representative images of 0 and 45 dpe Drosophila’s axons showing GH146-GAL4;UAS-MitoTimer. Scale bar, 5 µm. (N) Axonal Red/Green integrated density ratio; 0 vs 45 comparison reports an increase in oxidation in control flies (p=0.0447; n=8), with no change in dSetdb1 RNAi flies (p>0.9999; n=8). Control vs dSetdb1 RNAi at 0 dpe (p=0.1266; n=8) and at 45 dpe (p>0.9999; n=8). (O) Representative images of 0 and 45 dpe Drosophila’s lateral horn (LH) showing GH146-GAL4;UAS-MitoTimer. Scale bar, 5 µm. (P) LH Red/Green integrated density ratio; 0 vs 45 comparison reports a significant oxidation increase in control flies (p<0.0001; n=8), while dSetdb1 RNAi flies do not display change (p=0.1263; n=8). Control vs dSetdb1 RNAi at 0 dpe (p=0.4114; n=8) and at 45 dpe (p=0.0001; n=8). P-value: ***p<0.0001; ***p<0.001; **p<0.01; *p<0.05; ns >0.05. For all quantified panels, each n represents one brain from an individual animal. All error bars represent mean ± SEM.
Figure 5—figure supplement 1
Quantification of fluorescent signal within GFP-labeled mitochondria expressed in olfactory projection neuron (OPN) in the antennal lobe, axons, and lateral horn.
(A) Representative 20 X confocal image of a Drosophila brain expressing mitoGFP (green) in OPNs (GH146-Gal4 driver). White squares indicate regions of interest: Antennal Lobe (AL), axon, and Lateral Horn (LH). (B) Representative 63 X confocal images of mitoGFP-labeled mitochondria in the AL, distal axon, and LH at different ages. Left panels: raw GFP signal. Right panels: 3D surface reconstruction of mitoGFP signal for quantification. Scale bars: 50 µm (A), 5 µm (B). For mitochondrial morphology analysis, a 3D surface rendering of mitoGFP was performed using Imaris software. The following parameters were quantified to assess age-related changes: total mitoGFP volume (µm3), puncta number, fragmentation index (volume/area, µm), sphericity index, average size (µm3), and integrated density.
Figure 5—figure supplement 2
Analysis of mitochondrial morphology in olfactory projection neurons (OPNs) unaffected by dSetdb1 downregulation with aging in Drosophila.
(A) Quantification of mitochondrial count in the antennal lobe (AL). (B) Average size of mitochondria in the AL. (C) Total mitochondrial volume in the AL. (D) Mitochondrial count in axonal regions. (E) Fragmentation index of axonal mitochondria. (F) Sphericity index of mitochondria in the AL. (G) Average size of mitochondria in axonal regions. (H) Mitochondrial volume in axonal regions. (I) Mitochondrial count in the lateral horn (LH). (J) Fragmentation index of LH mitochondria. (K) Sphericity index of LH mitochondria. (L) Average size of mitochondria in the LH. In panels A, D, and I, mitochondrial count remains unchanged by dSetdb1 downregulation across the age spectrum. In panels B and G, average mitochondrial size decreases with age in both control and dSetdb1 RNAi flies. In panel C, total mitochondrial volume is reduced in control flies with age but not significantly affected by dSetdb1 knockdown. In panels E, F, J, and K, the fragmentation and sphericity indices in axonal and LH regions show no differences between control and dSetdb1 RNAi flies, indicating that dSetdb1 does not affect mitochondrial morphology in these regions. Similarly, mitochondrial volume in axonal regions (H) and average mitochondrial size in the LH (L) are not affected by dSetdb1 knockdown. Data are presented as mean ± SEM; n=6. Scale bars: 5 µm for axonal regions, 10 µm for AL and LH images. P-value significance: ns = not significant, *p<0.05, **p<0.01, ***p<0.001. Statistical significance was determined using two-way ANOVA with Bonferroni’s multiple comparisons test. For all quantified panels, each n represents one brain from an individual animal. All error bars represent mean ± SEM.
Figure 6
dSetdb1 knockdown preserves neuronal integrity and synaptic density in the aging Drosophila OPNs.
(A) GFP-labeled olfactory projection neurons (OPNs) in the antennal lobe (AL) of 0 and 45 days post eclosion (dpe) control flies and flies with knockdown of dSetdb1 and dve. GFP-positive OPNs are gray, and the right panel shows merged channels with nuclei labeled with ToPro3 in cyan. Scale bar, 20 µm. (B) Nucleus count in GH146-positive OPNs. Graph and two-way ANOVA with Bonferroni’s multiple comparisons between 0 and 45 dpe show that control flies had a significant decrease in nucleus count (p=0.0233, n=10). dSetdb1 RNAi (p>0.9999, n=10); Control vs. dSetdb1 RNAi of 45 dpe, (p=0.0398, n=10). (C) Orthogonal view of the 3D reconstruction of distal axonal tract of OPNs tagged with GFP. Panel shows the combination of axis Y and X, Z and X, and Z and Y for axons from control flies and knockdown flies for dSetdb1. Scale bar, 5 µm. (D) Quantification of axonal integrated density of axons shown in C. Two-way ANOVA multiple comparison between Control flies shows a significant decrease in axonal integrated density from 0 (white bars) to 45 dpe (red bars) (p<0.0001, n=10). dSetdb1 RNAi (p>0.9999, n=10). (E) Representative images of orthogonal view of the 3D reconstruction of GFP-tagged OPNs in the LH. Images show the combination of axis Y and X, Z and X, and Z and Y. Panel shows LH from 0 and 45 dpe control flies, knockdown flies for dSetdb1 and dve. Scale bar, 10 µm. (F) Quantification of GFP integrated density in the LH of images shown in E. Two-way ANOVA with Bonferroni’s multiple comparisons test shows a significant decrease in GFP volume in the LH of control flies from 0 to 45 dpe (p<0.0001, n=10), and no change for dSetdb1 RNAi (p>0.9999, n=10). (G) Orthogonal view of representative 3D reconstruction images of Brp::GFP-labeled presynaptic densities in LH of 0 and 45 dpe flies bearing the dSetdb1 GH146 knockdown. Scale bar, 20 µm (H) Quantification of BrpGFP integrated density of images shown in G. LH Brp::GFP integrated density showed a significant decrease in Brp::GFP integrated density in control flies from 0 to 45 dpe (p=0.0031, n=6), while dSetdb1 RNAi flies did not show a significant change (p=0.0758, n=6). (I) Quantification of the number of presynaptic densities labeled with BrpGFP in the LH of flies bearing the dSetdb1 knockdown shown in G. Control flies showed a reduction in the number of presynaptic densities labeled with Brp::GFP from 0 to 45 dpe (p<0.0001, n=6), and dSetdb1 RNAi showed no significant change (p>0.9999, n=6). White and red bars represent 0 and 45 dpe flies, respectively. n=independent fly brain. P-value: ****p<0.0001; ***p<0.001; **p<0.01, *p<0.05 and ns >0.05. For all quantified panels, each n represents one brain from an individual animal. All error bars represent mean ± SEM.
Figure 7
Schematic representation of age-related changes in H3K9 methylation and its impact on UPRMT and neuronal integrity in Drosophila.
In young organisms, mitochondria are challenged by various insults that lead to the accumulation of mitochondrial dysfunction, causing damage and activating a retrograde response from mitochondria to the nucleus. Ubl, crc, and DVE are translocated to the nucleus, and DVE maintains an open chromatin state, allowing the binding of transcriptional modulators of the UPRMT. This event activates the transcription of chaperones and proteases to recover mitochondrial homeostasis and oxidation. During aging, trimethylation of H3K9 is a mark of heterochromatin associated with the repression of transcription. This trimethylated state of H3K9 does not allow the binding of the transcriptional modulators of UPRMT, dve, crc, and ubl, inhibiting the mitochondrial response to aging-causing damage. Thus, mitochondrial function persists and builds up in a time-dependent manner, increasing mitochondrial oxidation and contributing to aging phenotypes, such as neurodegeneration marked by the reduction of OPNs, axonal volume, and presynaptic connections. Figure created with BioRender.com.
Author response image 1
DVE overexpression triggers cell death and neurodegeneration in Drosophila.
(A, C) Survival curves of flies with ubiquitous (ActG4) or pan-neuronal (ElavG4) knockdown (RNAi) or overexpression (UAS) of DVE. (B, D) Percentage of expected flies expressing UAS-DVE driven by ActG4 or ElavG4 compared to controls. (E, F) Representative eye images of flies with GMR-Gal4 driven DVE knockdown (RNAi), overexpression (UAS), or overexpression with dIAP2 (inhibitor of apoptosis) co-expression. (G) Representative wing arc images of flies with Dpr-Gal4 driven DVE knockdown (RNAi) or overexpression (UAS). White arrowheads indicate CD8::GFP-labeled neuronal cell bodies. (H) Quantification of GFP-labeled cell bodies in the wing arc of flies with DVE knockdown (RNAi) or overexpression (UAS). Statistical analyses: Gehan-Breslow-Wilcoxon test for survival curves, two-way ANOVA with Dunnet's multiple comparisons test for other data. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.
Additional files
-
MDAR checklist
- https://cdn.elifesciences.org/articles/103118/elife-103118-mdarchecklist1-v2.docx
-
Source data 1
Raw data and analysis outputs for all the figures.
- https://cdn.elifesciences.org/articles/103118/elife-103118-data1-v2.xlsx
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Age-dependent H3K9 trimethylation by dSetdb1 impairs mitochondrial UPR leading to degeneration of olfactory neurons and loss of olfactory function in Drosophila
eLife 15:e103118.
https://doi.org/10.7554/eLife.103118
