The Role of T-loops in Aging of Human Cells

1999 senior Scholar Award in aging

In collaboration with Jack Griffith (UNC) we have found that human telomeres form large duplex loops (t-loops). We propose that t-loops represent the mechanism by which telomere ends are masked from the cellular machinery that detects DNA breaks. Our working hypothesis is that telomere shortening in aging human cells results in chromosome ends that no longer form t-loops. Such t-loop deficient termini are proposed to constitute the main signal leading to cell cycle arrest, senescence, and apoptosis in aging cells.

We have previously identified a key player at mammalian telomeres, the telomeric DNA binding factor, TRF2. Loss of TRF2 from chromosome ends results in immediate deprotection of telomeres. Cells respond to this insult with the activation of a DNA damage pathway that includes the ATM kinase and p53, resulting in cell cycle arrest. Depending on the cellular context, such cells undergo apoptosis or display a phenotype similar to senescence. The unprotected telomeres lose the 3' protrusion of TTAGGG repeats typical of mammalian chromosome ends and eventually become ligated, forming dicentric and multicentric chromosomes. Thus, the removal of TRF2 from chromosome ends mimics events observed in aging human cells in which the telomeres have become critically shortened.

A likely mechanism for TRF2-mediated telomere protection has now been revealed in a study on the structure of the telomeric complex. Using electron microscopy, we have found that telomeres can exist in a specific higher order conformation (t-loops). EM analysis of psoralen crosslinked telomeric DNA demonstrated frequent large t-loops at natural chromosome ends. Molecular studies indicate that t-loops are formed through invasion of the 3' telomeric overhang into the duplex telomeric repeat array. We propose that t-loops are the main mechanism by which cells sequesters telomere ends from DNA damage checkpoints. In agreement with its protective role at telomeres, TRF2 has the ability to promote t-loop formation in vitro, suggesting that its main function in vivo is to facilitate remodeling of telomeric DNA into the t-loop configuration.

The presence of t-loops at human chromosome ends suggests a mechanism by which telomere shortening could induce cellular senescence and apoptosis. We propose that critically shortened telomeres fail to form t-loops, resulting in unfolded, exposed chromosome ends that activate a senescence or apoptotic signalling pathway, possibly involving ATM and p53. This hypothesis will be tested by determining the relationship between t-loop loss, telomere shortening, and cellular aging phenotypes. In addition, we will aim to manipulate the presence of t-loops at chromosome ends in order to test whether loss of t-loops can induce premature senescence and apoptosis. Conversely, we will test whether improved t-loop formation can extend the in vitro life-span of human cells.

Titia de Lange M.D. , Ph.D.
Rockefeller University

Our research focused on the aging of human cells as it is observed in laboratory cultures. It has long been known that normal human cells have a finite life-span. After cells have divided approximately 50 times, they stop. Although the cells stay alive, they have lost the ability to divide again and adopt a specific state, referred to as senescence. Early experiments indicated that cells have an intrinsic knowledge of the number of divisions they have executed and how many divisions in total they are allowed. This remarkable ability of cells to count divisions was later shown to be dependent on their telomeres, the elements at the ends of chromosomes that protect the genetic endowment. Telomeres were shown to shorten a bit with every cell division. Once the telomere reserve has run out, cells stop dividing. Telomeres are functioning like a clock, ticking away with each cell division, keeping track of the number of divisions, and sounding an alarm once 50 divisions have taken place.

In our research, we asked how this telomere clock works. In particular, we wished to understand how telomeres alert cells at the end of 50 divisions. We had previously found that telomeres form little loops at the end of the chromosome, a structure we refer to as a telomere loop, or t-loop. T-loops are formed by inserting the end of the chromosome back into the DNA of the chromosome. We speculated that t-loops protect the end of a chromosome by simply ├Čtucking it in├«. If this is correct, one might imagine that very short telomeres, such as those present in old human cells, can no longer form t-loops. The exposure of these chromosome ends, now no longer tucked in, would signal an alert to the cell.

Several aspects of this hypothesis were addressed in our studies. We studied a protein, called TRF2, that binds to telomeric DNA and has the ability to make t-loops in vitro. When the telomeres shorten, they contain less and less TRF2. So TRF2 is part of the telomere clock. We found that when TRF2 is experimentally removed from the telomeres, the alarm goes off: cells enter senescence as if the telomeres are very short. Furthermore, we found that if we load the clock up with much more TRF2, cells misread time and they continue to divide, even though the telomeres are very short. These and other experiments have informed us about the molecular features of the telomere clock. This knowledge will help us understand how telomeres affect human health and aging.