PerspectiveCELL PROLIFERATION

Unlimited Mileage from Telomerase?

See allHide authors and affiliations

Science  12 Feb 1999:
Vol. 283, Issue 5404, pp. 947-949
DOI: 10.1126/science.283.5404.947

If oncogene activation is akin to a jammed gas pedal and tumor suppressor inactivation to loss of the brakes, what then is telomerase activation in this hackneyed cancer-car analogy? Telomerase, the reverse transcriptase that maintains the ends of eukaryotic chromosomes, has long been stigmatized for its association with human cancer. This judgment has grown from the observations that, although most normal cells are devoid of telomerase's enzymatic activity and lack its main protein component (human telomerase reverse transcriptase, or hTERT), nearly all human tumors express hTERT and have active telomerase (1, 2). New data confirm that telomerase is neither tumor suppressor nor oncogene, underscoring the unique role of telomeres in tumorigenesis.

In normal cells, insufficient telomerase activity and a finite store of telomeric DNA limit the number of divisions a cell can undergo before critical telomere shortening signals entry into replicative senescence, defined by a finite capacity for cell division (3, 4). As such, replicative senescence is believed to represent a prominent genetic roadblock on the way to cancer, one that can be avoided by activation of telomerase (4, 5) but that can also be detoured by oncogene activation (for example, Myc) or loss of tumor suppressor function (such as RB/p53) (68). However, cells that use Myc or loss of RB/p53 to circumvent senescence will eventually experience rampant genome instability due to the loss of their telomeres and require a mechanism to maintain these essential elements (9). Thus, no matter which road the aspiring cancer cell travels, counteracting telomere shortening appears to be a key step. On the basis of the frequent activation of telomerase in human cancer, up-regulation of this enzyme is apparently the simplest way toward this end.

Thus, telomere erosion and the associated limitations in replicative life-span have been proposed as a potent tumor suppression mechanism, and telomerase was censured as a “bad” enzyme whose activity subverts our normally constrained somatic cells by providing the opportunity for boundless growth. Initial efforts to understand the mechanics underlying the telomerase-tumorigenesis connection seemed to yield more questions than answers. Is its activation essential for cancers to develop? Is telomerase really all that is needed for cellular immortalization, and will enforced somatic expression of telomerase lead to a cancer-prone condition?

Definitive answers to these questions have yet to emerge. However, the first major advance was provided a year ago with the finding that ectopic expression of hTERT in primary human cells could confer endless growth in culture (4, 5). Although the cells in question, human foreskin fibroblasts and retinal pigment epithelial cells (RPEs), normally ceased dividing after 40 to 80 population doublings, telomerase-positive derivatives able to maintain their telomeres progressed unimpeded beyond that usual life-span and have now been maintained in continuous growth for more than a year (10, 11). For practical purposes, these cells can be viewed as immortal—a characteristic illegitimately appropriated by many human cancers but normally preserved for the few cells that make up our germ line.

These studies received much attention as a potential cellular fountain of youth, with visions of an immediate impact on normal tissue and transplant repositories, while the popular press was distracted with speculations that telomerase could attenuate organismal aging and promote longevity in humans. A more guarded view (12) raised concerns that unscheduled telomerase expression in vivo may lead to an increase in cancer incidence by eliminating replicative senescence, hence obviating a potential tumor suppression mechanism.

The simple interpretation that hTERT expression alone can endow all cell types with unlimited growth potential has given way to a more complex story since the finding that immortalization of mammary epithelial cells and keratinocytes required not only hTERT expression, but also compromise in the RB pathway (13). Moreover, the observation that activated RAS and RAF signals can induce cellular senescence in pre-senescent primary fibroblast cultures well before telomeres have reached a critical length suggests that some physiological stimuli may be capable of acting dominantly to subvert the actions of hTERT (14, 15). Together, these findings raised concerns as to whether the life-extended hTERT-expressing fibroblasts and RPEs also had sustained additional genetic lesions. Two recent reports have gone a long way in addressing these important issues (10, 11). An extensive molecular and biological characterization of hTERT-immortalized fibroblasts and RPEs now indicates that they behave like their normal pre-senescent counterparts, harboring an intact RB pathway, functional DNA damage checkpoints, and normal karyotypes while lacking well-established hallmarks of neoplasia such as reduced serum requirements, anchorage-independent growth, and tumor formation in nude mice. The non-oncogenic nature of hTERT is in accord with the inability of this gene to substitute for the immortalizing oncoprotein Myc in the classical Myc/RAS cotransformation assay (16).

Why, then, are fibroblasts and RPEs different from mammary epithelial cells and keratinocytes? Part of the answer may lie in the simple fact that amounts of p16INK4a [a critical inhibitor of the RB pathway and key mortality gene (17, 18)] are low in fibroblasts, thus perhaps making it easier for telomerase alone to bypass senescence in those cells. A more likely explanation, however, could relate to cell type-specific differences in the signaling responses activated upon adaptation to culture and how those responses ultimately affect mortality pathways, particularly those governed by p16INK4a and its surrogate pRB. These cell culture-based studies underscore the need to frame these questions in a more physiological context, in which the long-term consequences of broad somatic TERT transgenic expression can be monitored.

It is nevertheless reassuring to know that the hTERT gene is not behaving as a conventional oncogene and that its effects are restricted to telomere metabolism. But do these findings exonerate telomerase as a culprit in cancer? Should we view it now as a “good” enzyme that can be used ad libitum to manipulate the life-span of human cells? Can we look forward to the repair of human tissues and rejuvenation of stem cell populations based on telomerase therapy? If telomerase does not conspire in the tortuous pathway of human tumorigenesis, why then is the enzyme activated in so many cancers? Although telomerase activity is associated with high proliferative rates in some cell types, such regulation fails to explain the appearance of hTERT in most cancers (19). After all, many normal human cell types do not express telomerase while they proliferate in vitro, but tumors derived from such cells do.

Could telomerase simply be a harmless by-product of one of the oncogenes causing malignant transformation? In this regard, it has recently emerged that the transcription of the hTERT gene is regulated directly by the immortalizing oncoprotein Myc (16, 20), whose up-regulation is an obligate feature of virtually all human cancers. Is telomerase just an innocent passenger driven by c-Myc but not lending any growth advantage to tumor cells? This view is made unlikely by the finding that inhibition of telomerase or experimental interference with telomere function arrests and often kills cells even if they are transformed (2123). Thus, telomerase activity would appear to make an important contribution to the viability of transformed cells, but its action does not fit the usual roles ascribed to oncogenes and tumor suppressors.

Instead of gas pedal or brake, telomerase and more specifically telomeres may be best viewed as the gasoline tank. Gasoline is not sufficient to drive or accelerate the car, nor does it affect the brakes, but when the gas is used up the car stops regardless of the status of its brakes or how hard one steps on the gas pedal. At times, other braking systems may operate early; such is the case with RAS-induced activation of the p16/Rb pathway or other oncogenic signals (14, 15). In those cases, telomerase introduction alone does not immortalize the cells, because they also must overcome the cell cycle arrest. That is, telomerase is not sufficient for transformation, but cells will have indefinite replicative capacity upon telomerase activation if there is a drive for proliferation and if nothing else arrests the cells.

The major remaining challenges are to determine whether shortening of somatic telomeres really constitutes a tumor suppressor mechanism in vivo and to assess the actual contribution of telomerase to cancer. Answers to these questions could emerge from several mouse models: the telomerase knockout mice (24, 25) and mice transgenic for telomerase in experimental settings where telomeres are limiting, in which it can be determined whether there is increased incidence of spontaneous or carcinogen-induced cancer. Ultimately, the definitive answer may have to come from the use of telomerase inhibitors in cancer patients. Although a complete understanding of the role of telomeres in cancer seems far off, the continued interest in telomerase should provide sufficient fuel to carry us to the end of this journey.

References and Notes

View Abstract

Stay Connected to Science

Navigate This Article