Figure 1.
Mitochondrial Dysfunction and Retrograde Response in Senescent Cells
(A) and (B) Confocal images of actively proliferating (PD 34) and senescent (PD 47) MRC-5 fibroblasts. (A) Red MitoSOX fluorescence indicates mitochondrial superoxide. Blue: DAPI. Note lower average red fluorescence and cell-to-cell heterogeneity in proliferating cultures. (B) JC-1 ratio images. A shift from red to green fluorescence indicates low MMP.
(C) MitoSOX, DHR, and JC-1 ratio intensities in senescent MRC-5 fibroblasts as measured by FACS (percentage of young fibroblast control, set to 100%). Data are mean ± s.e.m., n = 3. An asterisk (*) indicates a significant difference to young controls with p < 0.001.
(D) Integrity of mtDNA as measured by the relative amplification efficiency for an 11-kb mtDNA fragment in young and senescent MRC5 cells. Data are mean ± s.e.m. An asterisk (*) indicates a significant difference to young controls with p < 0.001.
(E) Single cell traces of cytoplasmic calcium levels [Ca2+]i following a challenge with 100 mM CaCl2 as measured by Fluo3 fluorescence confocal imaging.
(F) Basal cytoplasmic [Ca2+]i levels and recovery times in young and senescent cells. Data are mean ± s.e.m. from 13 young and 8 senescent cells. An asterisk (*) indicates a significant difference between young and senescent cells with p < 0.001.
(G–J) Relative mRNA expression levels in young (YOU), senescent (SEN), and hyperoxia-treated (HYP) MRC5 (averages of two [HYP] to four [YOU] independent experiments per condition). Fold-change colour code is indicated on the bottom. For some genes, the arrays contained probe sets resulting in multiple occurrences. All genes showing at least a 2-fold expression difference between YOU and SEN were selected if the expression difference was confirmed in HYP and if they fell into one of the following categories: (G) Ca-dependent signalling; (H) glycolysis and Krebs cycle; (I) mitochondrial function and biogenesis; and (J) stress response and apoptosis.
Figure 2.
Cellular ROS Levels Control MMP via UCP-2 Expression
(A) Interventions that modify ROS result in an inverse change in MMP. Fibroblasts were grown under 5% (squares) or 40% (triangles) oxygen. Cells were also treated with the ROS scavenger PBN (cross-haired square and inverted triangle) or transfected with an SOD3 transgene [34] (diamond). Open symbols: MitoSOX, filled symbols: DHR. Controls (circles) were set as 100% for each experiment. Data are mean ± s.e.m from three to five experiments per treatment. Solid line: linear regression, dotted lines: 95% confidence intervals. Red symbols: senescent and SES MRC-5 fibroblasts.
(B) UCP-2 expression in MRC-5. Top: Duplex-RT-PCR of UCP-2 versus GAPDH. Bottom: quantification of results, n = 2, mean + difference between means (Δm). NORM indicates young controls proliferating under normoxia, HYPER indicates young cells under 40% hyperoxia, and SEN indicates senescent cells under normoxia.
(C) UCP-2 versus β-tubulin (β-TUB) protein content. Top: Immunoblots; bottom: quantification of results. Data are mean + Δm, n = 2.
(D) Suppression of UCP-2 expression in MRC5 fibroblasts by treatment with siRNA (si1-UCP), but not with a scrambled control siRNA (si-C) as measured by RT-PCR. N indicates untreated control.
(E) si1-UCP expression results in increased mitochondrial superoxide levels (SOX) and increased MMP (JC-1 ratio). Data are mean + Δm, n = 2. Similar results were obtained with si2-UCP and si3-UCP.
Figure 3.
Mild Uncoupling by DNP Prolongs Replicative Lifespan of Human Fibroblasts
MRC5 fibroblasts were treated with 250 μM DNP from PD 35 till senescence.
(A) MitoSOX fluorescence. Data are mean + Δm, n = 2. Fluorescence intensity at t = 0 was set to 100%. Solid line: linear regression, dotted lines: 95% confidence intervals.
(B) Replicative lifespan. Data are from three (DNP) respective four (control) independent experiments. Hyperbolic best fits and 95% confidence intervals are shown.
(C) γ-H2A.X staining. The bar denotes 20 μm. Insert: higher magnification showing focal staining.
(D) Frequency of γ-H2A.X-positive fibroblasts. Data are mean ± s.d., n = 3. Differences are significant (p < 0.05) between day 8 and 46.
Figure 4.
Delayed Senescence in DNP-Treated Fibroblasts Is Telomere Dependent
(A) Telomere in-gel hybridisation of untreated (control) and DNP-treated MRC5. The size-weighted average of each peak is indicated by a white line. A λHindIII DNA digest is used as size marker (M).
(B) Average telomere (TRF) length in controls and DNP-treated MRC5 fibroblasts. Data are mean ± s.e.m. from quadruplicate gels. Solid line: linear regression, dotted lines: 95% confidence intervals. Differences between slopes are significant (2-tailed t test, p = 0.002).
(C) Telomere Q-FISH histograms of untreated (control) and DNP-treated MRC5 at the indicated PD. Medians are indicated by red bars and are at 6.0 arbitrary units (a.u.) (control PD 38), 1.3 a.u. (control PD 43), and 5.7 a.u. (DNP PD 45). Histograms of control PD 43 and DNP PD 45 are significantly different from each other according to a Mann-Whitney test (p-value < 2.2 × 10−16).
(D) ImmunoFISH of senescent DNP-treated cells at PD 50. γ-H2A.X–containing foci are shown in green and telomeres in red. The boxed area is shown as deconvoluted image at higher magnification at the right. Pixels that show significant co-localisation between foci and telomeres according to a Pearson correlation analysis are shown in white (indicated by arrows).
Figure 5.
Cell-to-Cell Heterogeneity in Replicative Senescence: Involvement of Mitochondrial ROS and Telomeres
(A) The frequency of senescent cells as measured by γ-H2A.X (filled bars) or Sen-β-Gal (open bars) staining is time- and oxidative stress–dependent. Experiments started at PD 38, and cells were grown for the indicated time under either 21% oxygen (CONTROL), 40% normobaric oxygen (HYPEROXIA), normoxia + 250 μM DNP (DNP), 5% normobaric oxygen (HYPOX), or normoxia and 400 μM PBN (PBN). Data are mean ± s.e.m., n ≥ 3.
(B) Cells with high mitochondrial superoxide production are γ-H2AX positive. Cells were stained with MitoSOX, photographed, and then fixed and immunostained for γ-H2A.X; then the same area was photographed again and merged.
(C) ImmunoFISH in an actively growing MRC5 culture shows frequent co-localization (yellow) between γ-H2A.X foci (green) and telomeres (red) in those cells that do contain foci. This co-localisation is significant according to a Pearson correlation analysis (see Figure 6H).
(D) Telomere Q-FISH of metaphases from young (left and middle) and near-senescent (right) MRC5 cultures. A few metaphases from young cultures show weak telomeric signals (middle image) similar to typical metaphases from a near-senescent culture (right image).
(E) Cumulative Q-FISH telomere frequency histograms from a young (PD 13, top left) and a near-senescent (PD 43, top right) population together with histograms from two individual metaphases at PD 13 (bottom). Metaphases of the near-senescent type (bottom right) are found with low frequency (<10%) in young cultures.
Figure 6.
Separation of Cells with a Senescent Phenotype out of Proliferating Cultures Shows Mitochondrial Dysfunction and TIFs
(A) Outline of the sorting experiments. Senescent (SES) and proliferating (PROL) cells were sorted by FACS out of growing MRC5 fibroblast cultures according to size and lipofuscin content as described [51], then replated, and the indicated parameters were measured.
(B) Morphology of sorted cells 1 wk after replating.
(C) Frequency of senescent cells as measured by γ-H2A.X (filled bars) or Sen-β-Gal (open bars) staining in SES- and PROL-sorted MRC5 fibroblasts. Data are mean ± s.e.m., n = 3.
(D) DHR and JC-1 ratio intensities and mtDNA copy numbers in SES cells relative to those in PROL-sorted fibroblasts, mean ± s.e.m., n ≥ 3.
(E) Duplex-RT-PCR of UCP-2 versus GAPDH in SES- and PROL-sorted MRC5 fibroblasts (top) and quantification of results (bottom). Data are mean ± s.e.m., n = 3.
(F) Absence of γ-H2A.X staining in PROL cells and co-localisation of γ-H2A.X foci (green) with telomeres (red) in SES cells. Pixels that show significant co-localisation between foci and telomeres according to a Pearson correlation analysis are shown in white (indicated by arrows) in the magnified image on the right.
(G) Absence of γ-H2A.X staining in cells sorted for low MitoSOX fluorescence intensity (SOX-) and co-localisation of foci (green) with telomeres (red) in cells with high MitoSOX fluorescence (SOX+). Magnified image on the right shows significant co-localisation as white pixels (indicated by arrows).
(H) Average foci-telomere Pearson correlation coefficients for cells in senescent culture (SEN), DNP-treated senescent cells (DNP), cells that contain γ-H2A.X foci in young culture (YOUNG), SES cells, cells sorted for high MitoSOX staining intensity (SOX+), and cells 48 h after 20-Gy γ-irradiation (IR) as negative control. All conditions are significantly different from the negative control, and the asterisk (*) marks a significant difference to SEN (ANOVA). Data are mean ± s.e.m., with the number of cells evaluated given for each condition.