Determining if Telomeres Matter in Colon Cancer Initiation or Progression

The nucleus of each human cell contains 46 linear chromosomes (92 ends) that are capped by thousands of repetitive TTAGGG DNA sequences (similar to the plastic tips on shoelaces). These ends, called “telomeres” have complex roles in aging and cancer and have proven to be intricately involved in such pivotal processes as the protection of genetic material, the completion of chromosome replication, the regulation of cellular aging due to progressively shortening telomeres throughout life, the initial protection against unlimited cellular growth, and, in combination with other alterations, the promotion of cancer progression. A collection of six proteins that either bind or associate with telomeres (the shelterin complex) begin to resolve a central question of how cells distinguish telomere ends from typical genomic DNA double-strand breaks (1). The consequences of telomere dysfunction are becoming more apparent by examining a growing list of human genetic diseases called telomeropathies. In certain inherited and familial cases of idiopathic pulmonary fibrosis, dyskeratosis congenita, and sporadic bone marrow failure (aplastic anemia), inheritance of short telomeres due to mutations in genes involved in the ribonucleoprotein telomerase holoenzyme lead to premature aging phenotypes (2). Telomerase is the cellular reverse transcriptase that can add TTAGGG repeats onto the telomeres and is active during early human development (3) but is silent in most adult tissues except proliferating stem cells, unless it is upregulated as part of cancer progression (4).

In normal cells (in the absence of other alterations), progressive telomere shortening leads to replicative senescence (not cancer) that is associated with many of the hallmarks of aging. It is believed that cellular senescence may initially be a potent suppressor of cancer by arresting cell-cycle progression, either in response to short telomeres or to oncogenic- or chemotherapy-induced DNA-damaging stimuli. In this issue of the Journal, Roger et al. (5) reports that extensive telomere erosion precedes initiation of colorectal cancer (CRC) in polyps obtained from familial adenomatous polyposis patients. They provide evidence that the combination of short telomeres together with adenomatous polyposis coli (APC) gene alterations may lead to chromosomal instability potentially driving clonal evolution and CRC progression. Roger et al. (5) also observed that the progressive telomere shortening in combination with APC mutations led to large scale genomic rearrangements that were independent of polyp size.

Although, at first glance, telomere shortening contributing to both aging and cancer may appear to be contradictory, there may be a logical explanation. For example, there is a long history of studies showing that viral oncoproteins such as simian virus 40 large T-antigen and human papillomavirus 16 E6/E7 enable bypass of senescence, providing normal human cells with an extended lifespan and eventually leading to a phenomenon known as crisis (where the telomeres are so short that there is ongoing chromosomal end-fusion and bridge-fusion-breakage cycles) (Figure 1). One way to think about telomere shortening is that when a few telomeres are short this leads to a DNA damage signal resulting in cellular growth arrest (called M1 or mortality stage 1) (6). In the absence of other alterations, this would, at least for a period of time, prevent additional cell divisions. However, when certain driver oncogenic changes occur, cells ignore the DNA damage signal arising from a few upcapped telomeres and continue to divide. Eventually the vast majority of telomeres become so short that cells form end-end associations and then end-fusions (7) until they cannot continue to divide (telomere catastrophe). However, these cells in crisis are also being driven forward by oncogenic changes, which results in a tenuous equilibrium between cell growth and apoptosis. In very rare instances, a mechanism is engaged to permit cells to overcome telomere catastrophe. In 85% to 90% of carcinomas, telomerase is greatly upregulated or reactivated (8,9) to stabilize the chromosome ends, and then the cells escape crisis (or M2, for mortality stage 2). There is another much less common pathway involving DNA recombination at telomeres that can also lead to the escape from crisis (10).

Figure 1.

Telomere shortening in colon cancer progression. In human cells, telomeres gradually shorten over a lifetime, even in cells that have some detectable telomerase activity, such as proliferating colonic epithelial cells. In the absence of other genetic or epigenetic alterations, eventually a few telomeres become so short that they can no longer cap and protect the linear telomere ends and are then recognized by the DNA damage-sensing machinery leading to replicative senescence (or M1, mortality stage 1). Such cells do not necessarily undergo apoptosis but can remain viable for years. Cells can bypass M1 if they obtain an appropriate oncogenic gain of function or a loss of tumor suppressor function. This leads to an extended lifespan period, potentially resulting in preneoplastic lesions, such as aberrant crypt foci or small colonic adenomas. Eventually the cells reach crisis, a point in which there are so many shortened telomeres that chromosome end-end fusions occur, leading to cycles of bridge-fusion-breakage until most cells undergo apoptosis. Rarely, a cell escapes crisis, and in colon cancer this is almost universally the transition between adenomas and adenocarcinoma, where the level of telomerase activity is greatly increased. Thus telomerase may slow down genomic instability by maintaining telomeres and reducing the cycles of bridge-fusion-breakage. M2 = mortality stage 2.

Telomere crisis generally occurs in the transition period between benign lesions and carcinomas in most epithelial solid tumors (8). In the case of CRC, APC alterations are believed to occur early, perhaps before or during the early adenoma stage, and this may result in bypass of replicative senescence, leading to crisis when global genomic alterations occur and very rarely leading to cell immortalization (Figure 1) (11). It has been established that telomere length abnormalities occur early in the initiation of most, if not all, human epithelial cancers (12). In addition, colonocytes of ulcerative colitis patients with progressive chronic inflammatory disease show premature shortening of telomeres, which might in part explain the increased predisposition and earlier risk of CRC (13,14). These investigators (13,14) also observed that telomere shortening is associated with chromosomal instability and anaphase bridges (a result of end-to-end chromosomal fusions), providing a mechanistic connection between telomere shortening, chromosomal damage, and cancer.

It is important to remember that human CRC almost always emerges over years, if not decades (15). Thus, it is not surprising that premalignant cells may have some regulated telomerase activity (8,16). This has been noted for many tissues for which there is a high rate of cellular turnover, such as in the bone marrow, the skin, and the gastrointestinal tract (4,8). Thus, the low levels of telomerase activity in polyps in the Roger et al. study (5) may extend the proliferative lifespan of colonic stem cells but are insufficient to immortalize them. It is only in the context of crisis and critically short telomeres that telomerase becomes expressed at much higher levels, most likely to stabilize the complex genomic chaos caused by continuous bridge-fusion-breakage cycles. Thus, telomerase re-expression or upregulation may actually provide a mechanism to stabilize the genome and suggests that inhibiting telomerase in patients with premalignant lesions may actually prevent CRC progression.

With increased age not only does the emerging colon polyp contain shorter telomeres but so also do the cells of the microenvironment, such as fibroblasts around the colonic crypts, the endothelial cells of the vasculature, and the cells of the immune system. There are emerging reports demonstrating that short telomeres in stromal cells may contribute to cancer by a phenomenon known as the senescence-associated secretory pathway, which may fuel chronic inflammation (17). In premalignant epithelial cells, senescence-associated secretory pathways can induce an epithelial–mesenchyme transition and invasiveness, hallmarks of malignancy by autocrine or paracrine mechanisms. Thus in the prepolyp stage, stimuli that engage cellular senescence can be both beneficial, in initially preventing damaged cells from dividing, and deleterious, by having effects on the precancerous microenvironment. Thus short telomeres in cells of the colon microenvironment could result in bypass of senescence, extended cell divisions, initiating specific alterations such as in the APC gene (as well as in other oncogenes and tumor suppressor genes), leading to crisis and telomere end-end fusions. Although the Roger et al. (5) studies were done on familial adenomatous polyposis patients, it is likely that these same changes would occur in patients with sporadic CRC because almost all have mutations in the APC gene.

It has been suggested that modulation of telomere biology and/or inhibition of telomerase may be excellent targets for cancer prevention and treatment (18–21). For example, development of predictive biomarkers directed toward specific subsets of cancers has ushered in a new era of personalized therapeutics. Will the use of telomere length as an enrichment biomarker for clinical trials be valuable because short telomeres correlate with poor outcomes (20,21)? The challenge for the present is to understand telomere-associated senescence more fully to harness its benefits (e.g., tumor suppression) while preventing its drawbacks (e.g., genomic instability and cancer progression).

Funding

The JWS laboratory is supported in part by the Southland Financial Corporation Distinguished Chair in Geriatric Research, Simmons Cancer Center Support Grant (5P30 CA 142543-03); a grant from National Institutes of Health (C06 RR30414); grants from the National Aeronautics and Space Administration (NNX11AC15G, NNJ05HD36G, NNX09AU95G); and the National Cancer Institute SPORE Grant (P50CA70907).