STRUCTURE AND FUNCTION OF MAMMALIAN TELOMERES

Telomeres are structural elements required for the stability and complete replication of chromosome termini. Originally, telomeres were defined based on the observation that naturally occurring chromosome ends behave differently than induced double-stranded DNA breaks. Such breaks are highly reactive entities that are susceptible to nucleolytic attack and give rise to chromosome rearrangements. In contrast, telomeres confer stability upon naturally occurring ends of DNA molecules, thereby allowing the accurate transmission of chromosomes. Human telomeres are composed of tandem repeats of the sequence 5' TTAGGG 3' complexed with sequence-specific DNA binding proteins. The resulting nucleoprotein structure is responsible for mediating telomere function.

In somatic cells, telomeres are dynamic structures and lose DNA sequences with each round of cell division. In most human tumors, however, telomere length is stabilized due to the activation of telomerase, a specialized reverse transcriptase necessary for telomere replication. These data have led to the hypothesis that telomere length acts as a type of clock to limit the number of divisions a cell will undergo, i.e., as a tumor suppressor mechanism. Expression of the catalytic subunit of telomerase is sufficient to immortalize human cells. The data suggest that when the telomere becomes critically shortened a signal is generated that prevents further cell division. The molecular mechanism(s) responsible for telomere length-dependent limitation of proliferative potential remain obscure and several candidate hypotheses exist. First, critically shortened telomeres might be unable to recruit sufficient telomeric proteins for a functional protective complex to form, which in turn would result in an exposed end that resembles a double-stranded DNA break. Alternatively, the shortened telomere repeat itself, or an increase in the concentration of free telomeric proteins arising due to lack of substrate binding sites, may generate a signal to the cell (Figure 1).

Two proteins that bind to the double-stranded region of mammalian telomeres have been identified, TTAGGG repeat binding factor 1 (TRF1) and TRF2; these proteins are related and have a similar domain organization. Both proteins are associated with telomeres throughout the cell cycle and bind to the cognate telomeric sequence as homodimers using a carboxy-terminal myb-type DNA binding domain. The identification of TRF1 and TRF2, as well as the cloning of several components of human telomerase, has made it possible to directly address questions regarding the molecular mechanisms regulating telomere replication and telomere-mediated chromosome stability. Our laboratory focuses on dissecting two aspects of mammalian telomere dynamics: 1) the molecular basis of cellular responses to telomere malfunction and, 2) the role of altered telomere dynamics in BRCA1 mediated transformation.

The formation of a nucleoprotein complex that sequesters the end of the DNA molecule and prevents it from activating DNA damage pathways is thought to be responsible for the protective function of telomeres. TRF2 has been shown to play a role in chromosome stability and is essential for the protective function of the telomere. Removing TRF2 from the telomere results in end-to-end fusions, which are manifested as anaphase bridges.

We used adenoviruses expressing a dominant negative allele of TRF2 to begin dissection of the cellular response to loss of telomere function (Figure 2A). Adenovirus infection is a powerful system making it possible to efficiently introduce alleles of interest into a wide variety of cell types. Dominant negative TRF2 efficiently disrupted the association of the endogenous protein within 24 hours of infection and anaphase bridges, representing loss of telomere protective function, were present within 36 hours of infection (Figure 2B). The response to disruption of the telomeric complex is dependent upon the cellular background. We have found that a variety of cell lines such as HeLa cells, mouse embryo fibroblasts (MEFs), and primary human B cells will undergo apoptosis in response to loss of telomere protective function.

The apoptotic response to loss of telomere function is dependent upon both the p53 and ataxia telangiectasia (ATM) gene products. This was demonstrated using human cell lines derived from individuals carrying ATM mutations and in MEFs derived from embryos that are homozygous null for p53. Infection of either of these cell types with the dominant negative TRF2 allele did not result in apoptosis. However, MEFs that were homozygous null for p21, pRB or INK4a/ARF exhibited a wild type apoptotic response indicating that these proteins are not involved in this response to telomere malfunction.

In contrast to the pathway described above, primary human fibroblasts do not undergo apoptosis in response to telomere disruption. The ability to define the cellular phenotype resulting from telomere malfunction in this cellular background is hampered by the detrimental effects of adenovirus infection, which are manifested 96 hours following infection. In the future, we plan to concentrate our efforts on dissecting the response of primary human fibroblasts to loss of telomere protective function using a retroviral-based gene delivery system.

BRCA1 mutations are responsible for the majority of hereditary cases of ovarian cancer. Evidence has been accumulating indicating that BRCA1 plays a role in genome maintenance. However, the molecular mechanisms underlying neoplastic predisposition remain ambiguous. The study of the genetic alterations leading to the development of ovarian tumors has been hampered by the paucity of early diagnostic markers for this disease. We have evidence that telomeres in cells derived from individuals carrying a mutated BRCA1 allele are significantly shorter than those in age-matched control individuals. This is the first observation linking an easily assessable phenotype with an increased risk for future development of breast and/or ovarian cancer.

The current working hypothesis regarding the link between telomere length and tumor formation predicts that premature senescence, rather than a predisposition for tumor development, would be manifested in individuals with shorter telomeres. Other hereditary syndromes associated with perturbed DNA metabolism such as Werner syndrome and ataxia telangiectasia also have an accelerated rate of telomere attrition coupled with a predisposition to neoplasias. Given the central role of telomeres in human oncogenesis, determining the basis for the observed difference in telomere structure and establishing the role of abnormal telomere dynamics in contributing to neoplastic transformation is of singular importance. In addition, this BRCA1 system represents a unique opportunity to investigate the molecular mechanisms by which alterations in telomere metabolism contribute to the development of a specific (i.e., breast and ovarian) tumor.

BROCCOLI, D., GODWIN, A.K. Telomere length changes in human cancer. In Methods in Molecular Biology: The Molecular Analysis of Cancer, edited by J. Boultwood, C. Fidler. The Humana Press (in press).

KARLSEDER, J., BROCCOLI, D., DAI, Y., HARDY, S., DE LANGE, T. Unmasked telomeres induce ATM/p53-dependent apoptosis. Science (in press).

^§   Fox Chase researcher

^a   J. Karlseder, T. de Lange: The Rockefeller University, New York, NY 10021

^b   S. Hardy: Chiron Corporation, Emeryville, CA 94608

Illustrations or unpublished data in these reports should not be used without permission of the author.

Fox Chase Cancer Center

Scientific Report 1998