Oncogene-induced telomere dysfunction enforces cellular senescence in human cancer precursor lesions

Abstract

In normal human somatic cells, telomere dysfunction causes cellular senescence, a stable proliferative arrest with tumour suppressing properties. Whether telomere dysfunction-induced senescence (TDIS) suppresses cancer growth in humans, however, is unknown. Here, we demonstrate that multiple and distinct human cancer precursor lesions, but not corresponding malignant cancers, are comprised of cells that display hallmarks of TDIS. Furthermore, we demonstrate that oncogenic signalling, frequently associated with initiating cancer growth in humans, dramatically affected telomere structure and function by causing telomeric replication stress, rapid and stochastic telomere attrition, and consequently telomere dysfunction in cells that lack hTERT activity. DNA replication stress induced by drugs also resulted in telomere dysfunction and cellular senescence in normal human cells, demonstrating that telomeric repeats indeed are hypersensitive to DNA replication stress. Our data reveal that TDIS, accelerated by oncogene-induced DNA replication stress, is a biological response of cells in human cancer precursor lesions and provide strong evidence that TDIS is a critical tumour suppressing mechanism in humans.

Figure 1

Melanocytic cells of benign- and dysplastic nevi, but not cells of malignant melanoma, display hallmarks of telomere dysfunction-induced cellular senescence. (A) Tissue sections from indicated lesions were immunostained with antibodies against melanA (red) and 53BP1 (green). Insets display an enlarged section of the indicated area. (B) Quantitation of 53BP1-positive melanocytic cells in benign nevi (Bng; n=13, 3529 cells), dysplastic nevi (Dsp; n=13, 2300 cells), melanoma (Mel; n=18, 5401 cells), and in normal epidermal melanocytes adjacent to the lesion (Norm; n=12, 485 cells). Values are shown as mean±s.d.; *P<0.001. (C) Dysfunctional telomeres in nevi. Tissue sections from benign nevi were processed by immunoFISH to simultaneously detect 53BP1 (green) and telomeres (red). Enlarged versions of the numbered DNA damage foci showing colocalization with telomeres are shown in the right micrographs. Note that only one optical slice is displayed. (D) Quantitation of TIF positive cells in indicated lesions (mean±s.d.). A total of 13 benign nevi (1355 53BP1 foci), 13 dysplastic nevi (2968 53BP1 foci), and 7 melanoma (891 53BP1 foci) were counted; *P<0.001. (E) Distribution of telomere lengths based on their signal intensities (x-axis; arbitrary units). Top histogram: all telomeric signals in cells of nevi (average signal intensity 268±46). Bottom histogram: single (red bars; average signal intensity 281±46) and multiple/diffuse (blue bars; average signal intensity 275±56) telomeric signals associated with 53BP1 foci. n: number of telomeric signals analysed (F) Tissue sections from a dysplastic naevus (top), and invasive melanoma (bottom) were immunostained using antibodies against 53PB1 (red) and macroH2A (green). Arrows point to stromal cells and basal layer epidermal keratinocytes that did not display elevated macroH2A levels. Dashed line separates epidermis (bottom left) from naevus (top right). Scale bars: 25 μm. Statistical significance was calculated by one-way ANOVA followed by Tukey’s post hoc test.

Figure 2

Epithelial cells of precursor lesions to colon- and breast-cancers display hallmarks of telomere dysfunction-induced cellular senescence. (A) Tissue sections from colonic adenomas (colon) and DH of the breast (breast; left column) and from invasive colon- and breast-carcinomas (right column) were immunostained with antibodies against 53BP1 (red, top row; green, bottom row). (B) Quantitation of 53BP1-positive epithelial cells (mean±s.d). Top graph: colonic adenomas (CA, n=10, 5588 cells), colonic adenocarcinomas (Carc, n=14, 4898 cells), epithelial cells from adjacent normal colonic mucosa (Norm, n=9, 2474 cells). Bottom graph: usual- and atypical-DH (Hyp, n=14, 4359 cells), invasive ductal carcinomas (Carc, n=15, 2905 cells), and luminal epithelial cells from adjacent normal ducts (Norm, n=9, 3153 cells); *P<0.001 by one-way ANOVA followed by Tukey’s post hoc test. (C) Quantitation of TIF-positive cells (mean±s.d.). CA: n=10; 3309 foci; Carc: n=14; 205 foci; Hyp: n=14, 1510 cells; Carc: n=15, 661 cells; *P<0.001 by unpaired t-test. (D) Tissue sections from indicated lesions were immunostained using antibodies against 53PB1 (red) and macroH2A (green). Scale bar: 50 μm.

Figure 3

Oncogenic signals cause stalling of telomeric DNA replication forks. (A) Schematic illustrating a replication fork at the telomeric locus. Following pulse labelling of mouse fibroblasts with halogenated nucleotides, the DNA was extracted, combed, and processed by immunoFISH as indicated in Supplementary Figure S7. As an example, a newly fired DNA replication origin (ORI) that incorporates IdU during the first pulse (blue) is shown. CldU (green) is then incorporated during the second pulse. Telomeric repeats were detected using a Cy3-conjugated telomeric PNA (red). We interpret symmetric replication signals as normal replication fork progression (#3). Asymmetric DNA replication bubbles bordering, or extending into-telomeric repeats, as inferred by asymmetric replication signals were considered as telomeric fork stalling events (#1 and 2). These events were quantified in Figure 3D. (B) Representative images of analysed DNA replication patterns. Rows 1: merged image; 2: IdU (blue); 3: CldU (green); 4: telomere (red). Scale bar: 40 kb. (C) Quantitation of the percentage of DNA combing signals revealing DNA replication forks arrest at the start of a telomeric tract (within 2 kb), partial and complete telomere replication in empty vector- (EV; black bars; n=140) and H-RasV12-expressing (Ras; red bars; n=171) cells. *P<0.001, **P=0.02. (D) Quantitation of fork stalling events at telomeres in empty vector- (EV; n=140) and H-RasV12-expressing (Ras; n=171) cells. *P=0.03.

Figure 4

Oncogenic signals cause fragile telomeres and stochastic telomere attrition. (A) Metaphases from BJ cells that were contact inhibited, transduced with lentivirus encoding Ras, and subsequently released into colcemid containing medium. Chromosome ends, labelled using a telomeric Cy3-conjugated-PNA (red), either displayed telomeric doublets (1), different signal intensities on sister chromatids (2), absence of signal (3), or diffuse telomeric staining patterns (4). Bar graph: percentage of aberrant telomeric structures (± s.d.) in GFP (69 metaphases) and Ras (98 metaphases) expressing cells (n=3); *P<0.001. (B) Contact inhibited BJ cells were transduced with lentivirus encoding GFP (control) or Ras followed by release into colcemid-containing medium. Telomere fluorescence (arbitrary units, mean±s.d.) as measured by flow-FISH or q-FISH (C) in Ras- and GFP-transduced cells; *P<0.001. Statistical significance was calculated by unpaired t-test.

Figure 5

Oncogenic Ras causes transient non-telomeric- and persistent-telomeric DDR foci in non-telomerized human cells. (A) Growth curve of BJ cells transduced with indicated retroviral combinations. Experiments were performed a total of five times (three times in BJ cells and two times in unrelated normal human foreskin fibroblasts) with similar results. p: encoding puromycin resistance; n: encoding neomycin resistance. Note that cells were selected using puromycin only. (B) Percentage of BrdU-positive cells 14 days following transduction (or 12 days following removal of the selection drug puromycin; BrdU added for 48 h). (C) Immunoblot showing Ras expression levels in BJ cells at indicated days following retroviral transduction. γ-Tubulin served as a loading control. Numbers represent relative telomerase activity following retroviral transduction, as measured by TRAPeze assay. (D) BJ cells were transduced with indicated combinations of retroviruses and analysed for the percentages of 53BP1-positive cells (±s.d.) at indicated times after transduction. Colour bars indicate the frequency of 53BP1 foci per cell nucleus. At least 100 cells were analysed for each group and time point; *P=0.001; **P=0.004 by unpaired t-test. (E) Percentage of 53BP1 (±s.d.) foci colocalizing with telomeric DNA at indicated times after transduction of indicated retroviral combinations. Only DDR-positive cells were analysed. A minimum of 30 cells were analysed for each group and time point (average of 165 53BP1 foci/group and time point); *P<0.001 by repeated measures one-way ANOVA.

Figure 6

DNA replication stress causes TDIS in normal human cells. (A) BJ cells and hTERT-expressing BJ cells were incubated with low concentrations of hydroxyurea (HU; 50 μM) and aphidicolin (Aph: 0.3 μM) for 4 days followed by a release from the block for 48 h. Bar graph: percentage of 53BP1-positive cells (± s.d.). The frequency of DNA damage foci per cell nucleus is indicated with coloured bars. A minimum of 400 cells per group were analysed (n=3); *P=0.002, **P<0.001. (B) Percentage of TIF-positive cells (±s.d.). A minimum of 100 cells per group were analysed (n=3); *P<0.001. (C) Percentage of cells that did not incorporate BrdU over a 48 h labelling period and were therefore considered senescent (±s.d.). BrdU was added to the culture medium following drug removal. A minimum of 400 cells per group were analysed (n=3); *P=0.007, **P=0.012. (D) Colocalization between 53BP1 (red) and ATR(Ser428; P-ATR) foci in nevi (top row), ADH (middle row), and colonic adenomas (bottom row). White arrows indicate colocalizations. Statistical significance was calculated by paired t-test.

Figure 7

Model illustrating the role of TDIS in suppressing malignant cancer progression in humans. Cells encountering oncogenic signals are exposed to telomeric replication stress, which leads to stochastic telomere attrition and telomere dysfunction in cells that lack telomerase activity. In normal somatic cells, telomere dysfunction results in cellular senescence, thereby preventing malignant cancer progression. In contrast, telomere dysfunction is suppressed in cells with high telomerase activity allowing these cells to continue proliferating. Similarly, in cells with compromised senescence responses, telomere dysfunction generates chromosomal instability, an event that is associated with reactivation of telomerase and malignant cancer progression.