Human telomeres maintain their overhang length at senescence

Abstract

Normal human cells in culture enter replicative senescence after a finite number of population doublings. The exact molecular mechanisms triggering the growth arrest are poorly understood. A recent report on the disappearance of the G-rich 3' telomeric overhang in senescent cells led to the hypothesis that loss of the 3' G-rich overhang is the molecular signal that triggers senescence. Here, we describe a quantitative assay to measure the length of the G-rich 3' telomeric overhangs from cultured cells. Using both this assay and the conventional nondenaturing hybridization assay for measuring G-rich overhangs, we show that normal human fibroblasts can maintain their overhangs at senescence. Furthermore, cells do not lose their overhangs when they bypass senescence after the inactivation of p53 and Rb. We thus conclude that a global reduction in overhang length is not the molecular signal that triggers replicative senescence.

FIG. 1.

Outline for measuring the size of telomeric 3′ G-rich overhang using the overhang protection assay. Total genomic DNA is incubated with T4 gp32 and GST-UP1, which bind to single-stranded 3′ G-rich overhangs. After the proteins are cross-linked to DNA, unprotected double-stranded DNA is removed by DNase I digestion. After heat inactivation of the DNase I, cross-linking is reversed and proteins are digested to make the protected overhangs available for hybridization to ³²P-labeled C-rich probe. Overhangs are finally size fractionated on a native polyacrylamide gel.

FIG. 2.

Validation of the overhang protection assay. (A) The overhang protection assay was performed on single-stranded oligonucleotides containing (TTAGGG)_n repeats (n = 6, 9, 16, 32, 48, and 64). Oligonucleotides were either annealed directly to the C-rich probe (annealed) or underwent the overhang protection assay (protected). Weighted mean sizes were calculated after quantitating the signals in each lane with ImageQuant software between molecular size (M) marker positions of 36 to 384 bp. (B) Results of the assay in panel A are plotted. (C) The overhang protection assay was performed on the model template pBSK-Rep4, which contained an ∼450-nt 3′ (TTAGGG)_n overhang after 3 kb of double-stranded pBSK vector sequences. Prior to the assay, a different amount of pBSK-Rep4 was mixed with or without HeLa genomic DNA. (D) Whole genomic DNA from HeLa cells was digested with T7 gene 6 exonuclease (a 5′-to-3′ exonuclease) for the indicated times and then treated with or without Exo I before being analyzed in the overhang protection assay. Black lines indicate positions of the bulk of the overhang DNA.

FIG. 3.

Quantitation of mean overhang sizes from the overhang protection assay. The gel (A) was first overlaid with a grid of boxes (B) and then quantitated (C). The values after subtracting the background (D) correspond to the term OD_i, while the values after dividing the signal by the size in nucleotides (E) correspond to the signal OD_i/L_i in the formula Σ(OD_i)/Σ(OD_i/L_i) used to calculate the final mean sizes. See text for details.

FIG. 4.

BJ cells maintain their telomeric 3′ overhangs at senescence. (A) Overhang protection assay of DNA from BJ fibroblasts at different PD up to senescence (sen). DNA was treated with or without Exo I before being analyzed in the overhang protection assay. To verify approximately equal amounts of input DNA, 1/30 of each sample was run on a 0.7% agarose gel and visualized with ethidium bromide (input). (B) Weighted mean sizes of the overhangs were quantified and are plotted. Results shown are representative of at least three independent experiments. Error bars represent one standard deviation. Cells were cultured in three different growth series until senescence, and DNA from these cells was used for the overhang protection assay. (C) Senescence-associated β-galactosidase staining of BJ fibroblasts at PD 32 and at senescence. (D) Nondenaturing hybridization analysis of telomeric restriction fragments from BJ fibroblasts at PD 40 and 90 (senescence). Undigested genomic DNA was hybridized to ³²P-labeled C-rich high-specificity probe and then gel fractionated in 0.5× TBE. After the gel was denatured, overhang signals were transferred to a nylon membrane and exposed to a phosphorimager screen (left panel). The denatured gel was neutralized and rehybridized to the Alu repeat probe (right panel). (E) Relative amounts of overhangs were calculated by normalizing signals from the membrane (overhang signals) by the Alu repeat signal (representing total genomic DNA) and plotted. Results shown are representative of three independent experiments. Error bars represent one standard deviation.

FIG. 5.

IMR90 cells maintain their telomeric 3′ overhangs at senescence. (A) Overhang protection analysis of DNA from IMR90 fetal lung fibroblasts at different PD up to senescence (sen) and from IMR90 cells expressing SV40 L-Tg or HPV E6 and E7 (E6/E7). Cells that are in the extended life span through expression of L-Tg (PD 74) or E6/E7 (PD 87) are indicated as M1-M2. DNA was treated with or without Exo I before being analyzed in the overhang protection assay. To verify approximately equal amounts of input DNA, 1/30 of each sample was run on a 0.7% agarose gel and visualized with ethidium bromide (input). (B) Weighted mean sizes of overhangs were quantified and plotted. Results shown are representative of at least three independent experiments. Error bars represent one standard deviation. (C) Nondenaturing hybridization analysis of telomeric restriction fragments from young (PD 30) and senescent (PD 58) IMR90 fetal lung fibroblasts. Undigested genomic DNA was hybridized to ³²P-labeled C-rich high-specificity probe and then gel fractionated in 0.5× TBE. After the gel was denatured, overhang signals were transferred to a nylon membrane and exposed to a phosphorimager screen (left panel). The denatured gel was then neutralized and rehybridized to the Alu repeat probe (right panel). (D) Relative amounts of overhangs were calculated by normalizing signals from the membrane (overhang signals) to the Alu repeat signal (representing total genomic DNA) and plotted. Results shown are representative of three independent experiments. Error bars represent one standard deviation.

FIG. 6.

Cells maintain their 3′ overhangs during their extended life span. (A) Growth curve of BJ fibroblasts ectopically expressing HPV E6 only, E7 only, or pLXSN (vector-only control). BJs were infected at PD 78 with retrovirus containing E6 (filled squares), E7 (filled triangles), or vector only (diamonds). After selection, cells were cultured until no cell number increase in 4 weeks was observed. (B) TRAP analysis of BJ cells ectopically expressing E6, E7, E6 and E7, L-Tg, or hTERT. Nonradioactive TRAP assays were performed by using whole-cell lysates from 2,500 cells. ITAS represents the 36-bp internal TRAP assay standards. (C) Weighted mean sizes of overhangs from the overhang protection assay from BJ fibroblasts ectopically expressing E6 or E6 and E7 (E6/E7) at different PD. Results are representative of at least three independent experiments. Error bars represent one standard deviation.

FIG. 7.

Loss of telomeric overhangs in cells expressing SV40 L-Tg, as shown in the overhang protection analysis of DNA from BJ fibroblasts infected with retrovirus containing SV40 L-Tg at different PD. Cells bypassed senescence at approximately PD 90. DNA was treated with or without Exo I before being analyzed in the overhang protection assay. To verify approximately equal amounts of input DNA, 1/30 of each sample was run on a 0.7% agarose gel and visualized with ethidium bromide (input). (B) Weighted mean sizes of overhangs were quantified and are plotted. Results shown are representative of at least three independent experiments. Error bars represent one standard deviation. (C) Relative amount of overhangs of BJ/L-Tg cells measured by nondenaturing hybridization assay. Undigested whole genomic DNA from BJ at PD 40 and at senescence was first hybridized to the C-rich telomeric probe and then gel fractionated. After the gel was denatured, overhang signals were transferred to a membrane and the gel was rehybridized to the Alu repeat probe. Relative amounts of overhangs were calculated by normalizing signals from the membrane (overhang signals) by the Alu repeat signal (representing total genomic DNA). Results shown are representative of three independent experiments. Error bars represent one standard deviation.