Telomeres, aging, and cancer: the big picture

Abstract

The role of telomeres in human health and disease is yet to be fully understood. The limitations of mouse models for the study of human telomere biology and difficulties in accurately measuring the length of telomere repeats in chromosomes and cells have diverted attention from many important and relevant observations. The goal of this perspective is to summarize some of these observations and to discuss the antagonistic role of telomere loss in aging and cancer in the context of developmental biology, cell turnover, and evolution. It is proposed that both damage to DNA and replicative loss of telomeric DNA contribute to aging in humans, with the differences in leukocyte telomere length between humans being linked to the risk of developing specific diseases. These ideas are captured in the Telomere Erosion in Disposable Soma theory of aging proposed herein.

Figure 1.

Q-FISH measurements of metaphase chromosomes from human fibroblasts at different population doublings (PDs). The telomere length values are increasingly skewed toward short telomeres as the cultured cells approach senescence. (A) Actual data (bars; adapted from Martens et al, with permission²⁸) correspond to predictions based on a mathematical model (lines; adapted from Rodriguez-Brenes and Peskin with permission⁹⁹) but show more pronounced skewing. (B) Snapshots of Q-FISH images of chromosomes X and 17 (also hybridized with a Chr17-specific probe) from individual metaphase cells of different diploid fibroblasts clones. For details see Martens et al, Asterisks indicate sporadic loss of telomere repeats on specific chromosome arms. Q-FISH, quantitative fluorescence in situ hybridization.

Figure 2.

Loss of telomere repeats with age in leukocytes. Adapted, with permission, from Baerlocher et al and Aubert et al. (A-B) Telomeric DNA in human granulocytes (Gr, n = 808) and naive T cells (nT, n = 832) from normal human donors is lost most rapidly in the first year of life, slows down after puberty, and is more pronounced in T cells than in granulocytes. The range in distribution of telomere length at a given age is expressed as a percentile based on best fit regression lines: 99th (red), 90th (green), 50th (black), 10th (green) and 1st (blue). The shaded blue area reflects ∼3 kb of subtelomeric DNA included in calculated values to allow comparisons with telomere length estimates by terminal restriction fragment (TRF) analysis. (C) The median telomere length in human granulocytes and naive T cells in female (pink; n = 29) vs male (blue; n = 29) cord blood samples. (D) The decline in telomere length in nucleated blood cells from a baboon was nonlinear and showed a pronounced drop after 1 year. (E) In humans, naive T cells from females (n = 414) have, on average, longer telomeres compared with those in males (n = 418). (F) Patients heterozygous for a mutation in either TERT or TERC (n = 58) have very short telomeres compared with siblings (n = 37) without the mutation.

Figure 3.

Normalized telomere length distribution in lymphocytes and granulocytes from normal, healthy individuals at the indicated ages. Data for schematic distributions are shown in Figure 2. Calculated length of telomere repeats rather than data compatible with terminal restriction fragment (TRF) results are shown by subtraction of 3 kb of subtelomeric DNA (included in TRF values and data shown in Figure 2). Note that the age-related decline in telomere length is not linear over time and is more pronounced in lymphocytes than in granulocytes.

Figure 4.

Short telomeres reduce the risk of cancer early in life at the expense of impaired regeneration late in life. Genetic variation in telomere length between humans is shown in blue. The TEDS theory of aging proposes that telomere erosion allowed lifespan to increase by suppressing the growth of malignant tumors before reproduction. Telomere loss has pleiotropic detrimental effects late in life by limiting cell renewal in the immune system and other tissues. Long and short telomeres increase the risk of cancer late in life via distinct, but partially overlapping, mechanisms (see text for details). Note that the rate of telomere loss depends, not only on cell divisions, but also on damage to telomeric DNA and variable levels of telomerase. The replication potential or maximum number of cell divisions (n) in stem cells is not known but is predicted to be less than 100 by TEDS: 50 to 60 divisions predicted by Schrödinger plus 0 to 40 additional cell divisions to account for stem cell renewal over a lifetime.