Volume 14, Issue 4, Pages 501-513 (May 2004) Telomere Shortening Triggers Senescence of Human Cells through a Pathway Involving ATM, p53, and p21CIP1, but Not p16INK4a  Utz Herbig, Wendy A Jobling, Benjamin P.C Chen, David J Chen, John M Sedivy  Molecular Cell  Volume 14, Issue 4, Pages 501-513 (May 2004) DOI: 10.1016/S1097-2765(04)00256-4

Figure 1 Upregulation of p16 and p21 in Single Cells (A) Visualization of p21 and p16 expression. Random fields of live p21+/GFP cells were marked and imaged for EYFP (left). Cells were considered p21-positive if their fluorescence exceeded basal levels in early passage cells. Samples were fixed and processed for p16 IHC, and the marked fields were imaged under phase (right). All four expression patterns were seen: arrow #1, p16−/p21+; arrow #2, p16−/p21−; arrow #3, p16+/p21+; arrow #4, p16+/p21−. (B) Visualization of p21, p16, and BrdU. Cultures were labeled with BrdU for 48 hr, and random fields were imaged for EYFP. BrdU was detected by immunofluorescence, the same fields were imaged, and the images were merged (left). Finally, p16 was detected by IHC, and the same fields were imaged under phase (right). Arrow #1, p21+/p16−/BrdU−; arrow #2, p21−/p16+BrdU−; arrow #3, p21−/p16−/BrdU+; arrow #4, p21+/p16+/BrdU−. (C) Left panel: Single parameter analysis of p21, p16, SA-β-gal expression and BrdU incorporation. RPD, remaining population doublings. Upper right panel: Two parameter analysis of p21 expression in p16+ and p16− cells. p21+ cells are plotted as percentages of p16+ and p16− cells. Lower right panel: Analysis of p16 expression in p21+ and p21− cells. Data were reanalyzed, and p16+ cells are shown as percentages of p21+ and p21− cells. >100 cells were analyzed at each time point. (D) Quantification of three parameter data from (B), at RPD = 10 (left) and RPD = 2 (right). >250 cells were analyzed at each time point. Molecular Cell 2004 14, 501-513DOI: (10.1016/S1097-2765(04)00256-4)

Figure 2 Accumulation of Telomeric γ-H2AX Foci in Senescent HDF (A and B) Cells were immunostained with γ-H2AX antibody. At the indicated RPD, γ-H2AX foci were quantified in LF1 cells (A) or p21−/− LF1 cells (B). The bars are divided into the fraction of cells containing 1 (blue), 2 (red), 3 (green), 4–5 (yellow), and over 5 (black) foci. TERT, cells immortalized with telomerase. >200 cells were analyzed at each time point. (C and D) Cells were processed by immunoFISH to visualize γ-H2AX foci and telomeres. At the indicated RPD TIF-positive cells were quantified in LF1 cells (A) or p21−/− LF1 cells (B). A cell was considered TIF-positive if ≥50% of its γ-H2AX foci colocalized with telomeres (PNA). >80 γ-H2AX-positive cells were analyzed at each time point. Molecular Cell 2004 14, 501-513DOI: (10.1016/S1097-2765(04)00256-4)

Figure 3 Localization of TRF1 and TRF2 to Telomeres in Senescent Cells (A) Near-senescent LF1 cells (RPD = 12) were immunostained with antibodies against TRF1 or TRF2 (red) and γ-H2AX (green). Arrows point to sites of colocalization. Enlarged views of deconvoluted images are shown in the right panels. (B) Quantification of two parameter analysis (A) of γ-H2AX/TRF1 and γ-H2AX/TRF2 TIF-positive cells. A cell was considered positive if ≥50% of its γ-H2AX foci colocalized with TRF1 or TRF2. >80 γ-H2AX-positive cells were analyzed in each case. (C) Three parameter immunofluorescent staining of TRF1 (red), TRF2 (green), and γ-H2AX foci (blue). Merge, merged images. Arrows, colocalization of γ-H2AX foci with TRF1 and TRF2. Enlarged views of deconvoluted images are shown in the bottom right panels. (D) Quantification of three parameter analysis (C) of colocalization between γ-H2AX foci and TRF1 and/or TRF2. TRF1, TRF1 only; TRF2, TRF2 only; TRF1+2, both TRF1 and TRF2; none, absence of both TRF1 and TRF2. >80 γ-H2AX-positive cells were analyzed. Molecular Cell 2004 14, 501-513DOI: (10.1016/S1097-2765(04)00256-4)

Figure 4 Colocalization of DNA Damage Response Factors with γ-H2AX in Senescent Cells (A) Near-senescent LF1 cells (RPD = 8) were immunostained with antibodies against γ-H2AX (middle panels, green) and phospho-ATM(S1981), 53BP1, MRE11, phospho-Chk2(T68), phospho-Chk1(S317), phospho-Rad17(S645), and BRCA1 (top panels, red). DNA was counterstained with DAPI (blue). Merged images are shown in the bottom panels. (B) LF1 cells at the same passage were processed by immunoFISH for the indicated DNA damage response factors (green) and telomeres (PNA probe, red). DNA was counterstained with DAPI (blue). Enlarged views of deconvoluted images are shown in the bottom panels. Molecular Cell 2004 14, 501-513DOI: (10.1016/S1097-2765(04)00256-4)

Figure 5 DNA Damage Foci in Senescent Cells Correlate with Growth Arrest and Upregulation of p21 but Not of p16 (A) LF1 cells were immunostained with antibodies against γ-H2AX (green; red in top right panel) and phospho-p53(S15) (red), p21 (green), BrdU (red), or p16 (red). BrdU labeling was for 48 hr. DNA was counterstained with DAPI (blue). Arrows point to γ-H2AX foci. (B) Three parameter immunoFISH to visualize γ-H2AX foci (green), telomeres (red), and BrdU incorporation (blue). Near-senescent LF1 cells (RPD = 6) were labeled with BrdU for 48 hr. Arrows point to γ-H2AX foci. Enlarged views of deconvoluted images are shown in the bottom panels (numbers correspond to arrows in top panel). Note that the two BrdU-positive cells contain one γ-H2AX focus each (arrows 1 and 2), neither of which colocalizes with PNA; the single BrdU-negative cell (bottom left) contains two γ-H2AX foci (arrows 3 and 4), both of which colocalize with PNA. (C) Three parameter immunofluorescent staining of near-senescent LF1 cells (RPD = 10) to visualize p16 (red) and p21 (blue) expression, and γ-H2AX foci (green). Arrows point to γ-H2AX foci. Note that the p16-positive cell in the top panel is p21-negative and the p16-positive cell in the bottom panel is p21-positive. (D) Quantification of two parameter analysis (A) of γ-H2AX foci versus phospho-p53(S15), p21, or p16 expression, or BrdU incorporation. Replicatively young (RPD = 38) and near-senescent (RPD = 10) cells were analyzed. The bar graphs enumerate γ-H2AX foci, the bars being divided into the fraction of cells containing 0 (white), 1 (blue), 2 (red), 3 (green), 4–5 (yellow), and over 5 (black) γ-H2AX foci. The second parameter (p53(S15), p21, p16, BrdU) is indicated on the x axis, with both positive and negative cells being scored independently. p53*, phospho-p53(S15). >150 cells were scored at each passage. (E) Quantification of three parameter analysis (B) of γ-H2AX/PNA TIF versus BrdU incorporation. >75 cells were analyzed. (F) Quantification of three parameter analysis (C) of p16, p21, and γ-H2AX. p16-positive cells (only) were scored for p21 and γ-H2AX staining. >60 p16-positive cells were analyzed. Molecular Cell 2004 14, 501-513DOI: (10.1016/S1097-2765(04)00256-4)

Figure 6 ATR/ATRIP Recruitment to Telomeres Is Increased in ATM−/− Cells (A) Top rows: Near-senescent LF1 and ATM−/− cells were immunostained with antibodies against ATRIP (left panels, red) and γ-H2AX (middle panels, green). Bottom row: Near-senescent ATM−/− cells were processed for immunoFISH using antibodies against ATRIP (left panel, green) and PNA (middle panel, red). In all cases DNA was counterstained with DAPI (blue). Merged images are shown in the right panels. Enlarged views of deconvoluted images are shown in the right-most small panels. (B) Quantification of γ-H2AX foci in proliferating ATM+/− and near-senescent ATM−/− cells. Senescent LF1 cells (Sen) are included for comparison. The bars are divided into the fraction of cells containing 1 (blue), 2 (red), 3 (green), 4–5 (yellow), and over 5 (black) γ-H2AX foci. >130 cells per cell line were analyzed. (C) Quantification of γ-H2AX/PNA TIF in proliferating ATM+/− and near-senescent ATM−/− cells. >60 γ-H2AX-positive cells were analyzed for each cell line. (D) Quantification of ATRIP foci in proliferating ATM+/− and near-senescent ATM−/− cells. >80 cells were analyzed for each cell line. (E) Quantification of colocalization between ATRIP and telomeres labeled with a PNA probe (ATRIP/PNA TIF) in ATM−/− cells. A cell was considered TIF-positive if ≥50% of its ATRIP foci colocalized with telomeres. >50 ATRIP positive cells were analyzed. (F) Quantification of colocalization between γ-H2AX foci and ATRIP, Chk1(S317), or Chk2(T68) foci in near-senescent ATM+/− and ATM−/− cells. >50 γ-H2AX foci were analyzed for each cell line. (G) ChIP was performed with antibodies against the indicated proteins. Data were normalized to the no antibody control for each cell line and are expressed as fold-change relative to early passage LF1 cells. ChIP was performed on two separate occasions with equivalent results. Molecular Cell 2004 14, 501-513DOI: (10.1016/S1097-2765(04)00256-4)

Figure 7 Cell Cycle Checkpoints in Normal and ATM−/− Cells (A) Quantification of three parameter immunoFISH analysis of ATRIP/PNA TIF and BrdU incorporation in near-senescent ATM−/− cells. TIF were scored as indicated in Figure 6E; BrdU labeling was for 48 hr. >60 ATRIP-positive cells were scored. (B) Near-senescent ATM−/− cells were immunostained for γ-H2AX (red) and p21 (green). DNA was counterstained with DAPI (blue). Arrows point to γ-H2AX foci. Note that the top cell is γ-H2AX-positive and p21-negative, the middle cell is γ-H2AX-positive and p21-positive, and the bottom cell is γ-H2AX-negative and p21-positive. (C) Quantification of two parameter immunofluorescent analysis of γ-H2AX foci and p21 expression (B). The bars are divided into the fraction of cells containing 0 (white), 1 (blue), 2 (red), 3 (green), 4–5 (yellow), and over 5 (black) γ-H2AX foci. >60 γ-H2AX-positive cells were scored. (D) Effect of 2-AP treatment on BrdU incorporation in LF1 cells. Cells in their subterminal passage (RPD = 2) were incubated with BrdU for 48 hr and with 2-AP (5 mM) only for the initial 2 hr. Pilot experiments showed the 2 hr 2-AP treatment to be optimal for stimulating BrdU incorporation. BrdU incorporation was visualized by immunofluorescent staining. >600 cells were scored. (E) Effect of 2-AP treatment on phospho-ATM(1981) foci in LF1 cells. Cells were treated as in (D), except that cells were processed immediately after the 2 hr 2-AP treatment for immunofluorescent detection of phospho-ATM(1981) (green). DNA was counterstained with DAPI (blue). (F) Quantification of the effect of 2-AP treatment on phospho-ATM(1981) foci (E). >175 cells were scored. (G) Effect of siRNA treatment on BrdU incorporation in LF1 cells. Cells in their terminal passage were incubated with BrdU for 48 hr and with the indicated siRNA for the initial 8 hr. CP, cyclophilin siRNA; ATM, ATM siRNA only; ATM+ATR, combined incubation with ATM and ATR siRNAs. BrdU incorporation and p16 expression were visualized by immunofluorescent staining. Data is presented for the p16-negative fraction of cells. p16-positive cells were >95% BrdU-negative. >80 cells were scored for each siRNA construct. (H) Quantification of total nuclear DNA content in near-senescent LF1 and ATM−/− cells with and without γ-H2AX/TRF1 TIF. After the completion of immunofluorescent staining for γ-H2AX/TRF1 TIF, samples were incubated for 1 hr in 10 mM Tris-HCl (pH 7.5), 1 M NaCl, 2 μM Hoechst 33258. Images were captured at 40× magnification, and Hoechst 33258 nuclear signals were quantified using Metamorph software. TIF-positive cells were subsequently identified at 160× magnification. A total of 503 LF1 cells (78 TIF-positive) and 829 ATM−/− cells (98 TIF-positive) were scored. Molecular Cell 2004 14, 501-513DOI: (10.1016/S1097-2765(04)00256-4)