Article Figures & Data
Figures
- Figure 1.
Opa1+/− mice show adult hippocampal neurogenesis impairments. A, Schematic of the proliferation and maturation steps of adult-born neurons (in days) within the dentate gyrus of the hippocampus in mouse, and specifically their critical period (21–28 d). B, Labeling against Ki67 and (C) DCX allowed to evaluate the number of proliferating cells and immature neurons, respectively, for each group of mice (mean values per DG, Ki67+ cells in Opa1+/+: 534.8 ± 53.53; in Opa1+/−: 484.7 ± 43.57; DCX+ cells in Opa1+/+: 6575 ± 686.4; in Opa1+/−: 6488 ± 558.4; 7–9 mice/group). D, Numbers of 28-d-old BrdU+ cells (mean values per DG, Opa1+/+: 221.3 ± 15.09; Opa1+/−: 169.5 ± 15.37; 7–9 mice/group, *p = 0.0329; unpaired t test). E, Representative 3D reconstruction images showing GFP+ (green) dendritic spines (arrows) and mitoDsRed+ (red) mitochondria. F, Density of GFP+ dendritic spines (Opa1+/+: 8.34 ± 0.46/10 μm dendrite; Opa1+/−: 5.22 ± 0.51/10 μm dendrite) and (G) spine subtypes (stubby: Opa1+/+: 0.93 ± 0.10 vs Opa1+/−: 0.47 ± 0.07 spines/10 μm; thin: Opa1+/+: 5.27 ± 0.3 vs Opa1+/−: 3.25 ± 0.3 spines/10 μm; mushroom: Opa1+/+: 1.78 ± 0.16 vs Opa1+/−: 1.29 ± 0.2 spines/10 μm; *p < 0.05, **p < 0.01, ***p < 0.001; two-tailed Mann–Whitney test). H, M, Mitochondrial parameters in the somatic and dendritic compartments of 21-d-old neurons. H, K, Total mitochondrial biomass per somatic or dendritic GFP+ volume. Somas: G, Opa1+/+: 4.98 ± 0.27; Opa1+/−: 3.32 ± 0.33; Dendrites: J, Opa1+/+: 15.49 ± 1.23; Opa1+/−: 10.93 ± 1.08. I, L, Mean volume of individual somatic or dendritic mitochondria. Somas: H, Opa1+/+: 0.39 ± 0.02 μm3; Opa1+/−: 0.32 ± 0.02 μm3; Dendrites: K, Opa1+/+: 0.58 ± 0.05 μm3; Opa1+/−: 0.41 ± 0.02 μm3. K, M, Number of mitochondria per somatic or dendritic GFP+ volume. Somas: I, Opa1+/+: 13.41 ± 0.85 particles/100 μm3; Opa1+/−: 11.57 ± 1.01 particles/100 μm3; Dendrites: L, Opa1+/+: 32.47 ± 3.28 particles/100 μm3; Opa1+/−: 27.35 ± 2.64 particles/100 μm3. Mean ± SEM; five mice/group, five to seven somas/mouse, six to eight dendritic segments/mouse (*p < 0.05, **p < 0.01, ***p < 0.001; two-tailed Mann–Whitney test). See Extended Data Figure 1-1, mitochondria in GFP+ adult born neurons from mouse dentate gyrus.
- Figure 2.
Opa1+/− mice have impaired spatial memory but intact nonspatial memory. A, Schematic setup of the spatial version of the Barnes maze. B, Number of errors before entering into the escape hole across training sessions, Opa1+/+ (n = 10) and Opa1+/− (n = 11), genotype, p = 0.88; two-way ANOVA with repeated measures). C, Two days after the last training session (day 8), memory was evaluated by a spatial memory probe test. Mice from both genotypes visited significantly more often the holes located in the target quadrant than in the other quadrants (target quadrant: ***p < 0.0001; two-way ANOVA). D, Opa1+/− mice visited the target hole significantly less often than the Opa1+/+ mice (Opa1+/+: 10.80; Opa1+/−: 7.27; *p < 0.05; unpaired t test). E, Schematic of the object location task. F, Preference for the displaced object compared with chance level: Opa1+/+: 62.98 ± 1.24%; n = 13, ###p < 0.0001; Opa1+/−: 53.40 ± 1.77%; n = 12, p>0.05; index versus 50%, one-sample t test. Compared with Opa1+/+ mice, mutant mice exhibited a spatial memory deficit (***p < 0.001; unpaired t test). G, Schematic of the object recognition task. H, During memory testing, both genotypes explored preferentially the novel object than the familiar one (Opa1+/+, 61.14 ± 1.51%, n = 14; Opa1+/−, 57.60 ± 1.57%, n = 12, ###p < 0.001, respectively; index vs 50%, one-sample t test). In F and H, dotted lines indicate equal exploration of the two objects. See Extended Data Figure 2-1, intact anxiety-related and locomotor behavior in Opa1+/− mice. See Extended Data Figure 2-2, Opa1+/− mice display intact performances in the nonspatial Barnes maze. See Extended Data Figure 2-3, long-term spatial memory is less accurate in Opa1+/− mice than in Opa1+/+ mice; Opa1+/+ and Opa1+/− mice express similar interest for all objects.
- Figure 3.
Opa1+/− mice show deficits in metric spatial pattern separation. A, Schematic of the pattern separation procedure. B, Exploration time during exposition phase (session: ***p < 0.0001; two-way ANOVA with repeated measures). No genotype effect was observed. C, Exploration ratio during memory test in Opa1+/+ mice (61.4 ± 1.3%, n = 12, t = 8.914, df = 11, ###p < 0.0001, one-sample t test) and Opa1+/− mice (52.8 ± 1.9%, n = 12, t = 1.430, df = 11, p = 0.181, one-sample t test). Performances in pattern separation were significantly different between genotypes (**p < 0.01; unpaired t test). Dotted line indicates equal exploration of the two objects.
- Figure 4.
Voluntary running corrects hippocampal-dependent memory deficits in Opa1+/− mice. A, Timeline of the protocol used to test the effect of running. B, Preference indexes for the displaced object for both genotypes in standard and running conditions (Opa1+/+ standard, 61.50 ± 1.30%, n = 8, ###p < 0.0001; Opa1+/+ running, 58.54 ± 1.91%, n = 7, ##p < 0.01; Opa1+/− standard, 51.85 ± 1.56%; n = 8, p > 0.05; Opa1+/− run, 62.24 ± 2.99%, n = 7, ##p < 0.01: index vs 50%, one-sample t test). Dotted line indicates equal exploration of the two objects. See Extended Data Figure 4-1, voluntary running increased the number of DCX+ cells in Opa1+/+ and Opa1+/− mice and rescued mitochondrial content in Opa1+/−.
- Figure 5.
Mdivi-1 treatment corrects hippocampal-dependent memory deficits in Opa1+/− mice. A, Timeline of the protocol used to test the effect of Mdivi-1 treatment. B, Preference indexes for the displaced object for both genotypes in standard and running conditions Opa1+/+ before treatment: 57.94 ± 2.41%, n = 9, #p < 0.05, index versus 50%, one-sample t test; after treatment: 62.37 ± 2.41%, n = 9, ###p < 0.0001, index versus 50%, one-sample t test; Opa1+/− before treatment: 54.67 ± 3.01%, n = 9, p = 0.158; index versus 50%, one-sample t test; after treatment: 64.67 ± 1.35%, n = 9, ###p < 0.0001, one-sample t test. Dotted line indicates equal exploration of the two objects.
Extended Data
Extended Data Figure 1-1
Mitochondria in GFP+ adult born neurons from mouse dentate gyrus. Images of MitoDsRed+ mitochondria inside GFP+ adult-born neurons in the dentate gyri from Opa1+/+ and Opa1+/− were captured from mouse brain sections, using a confocal microscope with a 63x oil lens and digital zoom of 6. A representative image of a 3D-projection from z-series acquired from each genotype, is shown, with inverted colors, after deconvolution using the Huygens Essential deconvolution software (SVI) (A) and after segmentation using the Imaris surface tool (B). Download Figure 1-1, EPS file.
Extended Data Figure 2-1
Intact anxiety-related and locomotor behavior in Opa1+/− mice. OPA1 deficiency has no impact on anxiety-like behavior measured by (A) the time spent in the open arms of the elevated plus maze (in %, Opa1+/+: 29.01 ± 4.27; Opa1+/−: 30.72 ± 2.28, t = 0.34, df = 25, p = 0.734; t test). Mice from both genotypes show similar locomotor activity in the open-field, evaluated by (B) the time spent in the center area (in %, Opa1+/+: 31.07 ± 2.03; Opa1+/−: 30.73 ± 2.46, t = 0.11, df = 25, p = 0.915; t test), and (C) the total distance travelled (in cm, Opa1+/+: 3246 ± 437.7; Opa1+/−: 3423 ± 145.1, p = 0.2388; U = 66, two-tailed Mann–Whitney test). Opa1+/+, n = 14; Opa1+/−, n = 13. Download Figure 2-1, EPS file.
Extended Data Figure 2-2
Opa1+/− mice display intact performances in the non-spatial Barnes maze. A, Schematic setup of the non-spatial (cued) Barnes maze. B, During the five days of training, the number of errors before entering the escape hole decreased across training sessions for mice of both genotypes, indicating they learned to locate the escape hole (session/day: F(4,88) = 17.82, ###p < 0.0001; two-way ANOVA with repeated measures). No difference was found between genotypes (F(1,22) = 0.002, p = 0.96; two-way ANOVA with repeated measures). C, During the non-spatial probe test held 24 h after training (Day 6), both groups of mice visited more often the holes located in the target quadrant than in each of the other quadrants (F(3,88) = 51.35, ***p < 0.0001; two-way ANOVA) and no difference was observed between genotypes (F(1,88) = 2.355, p = 0.128; two-way ANOVA). D, Opa1+/+ and Opa1+/− mice exhibited a similar preference for the target hole (t = 0.4089, df = 38, p = 0.6849; unpaired t test). Opa1+/+, n = 12; Opa1+/−, n = 12. Download Figure 2-2, EPS file.
Extended Data Figure 2-3
Long-term spatial memory is less accurate in Opa1+/− mice than in Opa1+/+ mice. A, During the spatial memory probe test held 7 d after training (Day 13), mice from both genotypes visited significantly more often the holes located in the target quadrant than those located in the other quadrants (F(3,76) = 56.56, ***p < 0.0001; two-way ANOVA). However, Opa1+/− visited significantly less often the holes of the target quadrant (Opa1+/+: 22.40, Opa1+/−, 16.45; #p = 0.0482; unpaired t test). B, This effect was strictly attributable to a difference in the number of visits to the target hole between Opa1+/− and Opa1+/+ mice (Opa1+/+: 9.90, Opa1+/−, 6.27; t = 3.433, df = 19, **p = 0.0028; unpaired t test), revealing a less precise spatial memory in mutant mice. Opa1+/+, n = 10; Opa1+/−, n = 11. Opa1+/+ and Opa1+/− mice express similar interest for all objects. Mice from both genotypes spent the same amount of time exploring the left and right objects during the acquisition phase of the object location (C) and object recognition (D) tasks. Mean exploration time was not different between genotypes, indicating that both groups of mice had a similar interest for each object. Object location: Opa1+/+, n = 13, left object: 15.08 ± 1.5, right object: 15.46 ± 1.33; Opa1+/−, n = 12, left object: 14.92 ± 1.44, right object: 14.75 ± 2.09. Object recognition: Opa1+/+, n = 14, left object: 16.79 ± 1.80, right object: 16.79 ± 1.89; Opa1+/−, n = 12, left object: 16.67 ± 1.50, right object: 16.83 ± 1.40. Download Figure 2-3, EPS file.
Extended Data Table 1
Data structure, test and power are shown for each figure. Download Table 1-1, XLS file.
Extended Data Figure 4-1
Voluntary running increased the number of DCX+ cells in Opa1+/+ and Opa1+/− mice and rescued mitochondrial content in Opa1+/−. Mice of eight/nine months of age were housed in standard conditions (standard) or submitted to a three-week period of voluntary running (running). A, Number of DCX+ cells; Opa1+/+: standard, n = 5, 554.4 ± 54.10; running, n = 7, 1187 ± 194.4; Opa1+/−: standard, n = 5, 572.4 ± 91.66; running, n = 7, 1134 ± 138.6, *p < 0.05, unpaired t test. B, Running increases the mean volume of somatic mitochondria per 100 μm3 of GFP+ cell volume by 33%: Opa1+/− mice (5.072 ± 0.475; n = 5) compared to Opa1+/− mice housed under standard (3.76 ± 0.328; n = 7) conditions (*p < 0.05, U = 158, one-tailed Mann–Whitney test). Download Figure 4-1, EPS file. Voluntary running increased the number of DCX+ cells in Opa1+/+ and Opa1+/− mice and rescued mitochondrial content in Opa1+/−. Mice of eight to nine months of age were housed in standard conditions (standard) or submitted to a three-week period of voluntary running (running) (A) Number of DCX+ cells ; Opa1+/+: standard, n = 5, 554.4 ± 54.10; running, n = 7, 1187 ± 194.4, **p = 0.0065, unpaired t test; Opa1+/−: standard, n = 5, 572.4 ± 91.66; running, n = 7, 1134 ± 138.6, *p < 0.05, unpaired t test. (B) Running increases the mean volume of somatic mitochondria per 100μm3 of GFP+ cell volume by 33%: Opa1+/− mice (5.072 ± 0.475; n = 5) compared to Opa1+/− mice housed under standard (3.76 ± 0.328; n = 7) conditions; *p < 0.05, U = 158, one-tailed Mann-Whitney test).
In this issue
