The pattern of chromosome-specific variations in telomere length in humans is determined by inherited, telomere-near factors and is maintained throughout life

Abstract

In this study the telomere length distribution on individual chromosome arms in humans has been characterized. Using fluorescent in situ hybridisation (FISH) followed by computer-assisted analysis of digital images, we show that the distribution of telomere length on individual chromosome arms is not random, but that humans have a common telomere profile. This profile exists in both lymphocytes, amniocytes and fibroblasts, and is conserved during life until about the age of 100. We find that the length of the telomeres generally follows the length of the chromosomes and that the chromosome specific differences in telomere length are determined by factors located very distally on the chromosome arms. In addition to the common profile, we also find that each individual has specific characteristics. Based on analysis of both monozygotic and dizygotic twins, we find that these characteristics are partly inherited. For each chromosome, age-related chromosome loss correlates negatively with telomere length. This suggests that decrease in telomere length may be an element in age-related genome instability.

Introduction

Telomeres are GC-rich repetitive structures located at the very end of chromosomes in higher organisms (Blackburn, 1994, Zakian, 1995). The monomeric unit of the telomeres in humans is TTAGGG (Moyzis et al., 1988, Lejnine et al., 1995), and tandem repeats of this unit results in a stretch of repetitions that varies from a few thousands to about 15.000 bp in size (Harley et al., 1990, Lindsey et al., 1991, Frenck et al., 1998). This DNA sequence is surrounded by a number of specific proteins (Chong et al., 1995, Bilaud et al., 1997, Broccoli et al., 1997, Smith et al., 1998, Kim et al., 1999), and the whole structure serves as a cap to protect the chromosome ends (Shay, 1999), possibly by forming a telomere loop structure (Greider, 1999, Griffith et al., 1999). This function is clearly demonstrated by the fact that chromosomes without telomeres are highly unstable and are frequently lost during cell growth (Li et al., 1998).

The telomere repeat sequences also represent a buffer of DNA sequence that appears to be dispensable without adverse effects. The cell needs such a buffer because a short stretch of DNA is lost during every replication cycle due to the well known ‘end replication problem’ (Olovnikov, 1973). In most somatic cells there is no system for compensating this telomere loss. However, in a few cell types, most notably in germ line cells, lymphocytes, skin stem cells and intestinal mucosa, this ability exists (Broccoli et al., 1995, Counter et al., 1995, Harle-Bachor and Boukamp, 1996, Wright et al., 1996). The ability to elongate telomeres resides primarily in the enzyme telomerase, which is a reverse transcriptase that can perform this telomere elongation (Blackburn et al., 1989). However, telomerase independent elongation of telomeres has also been observed (the ALT-pathway) (Bryan et al., 1995, Bryan et al., 1997).

Given the size of the telomere loss at each replication and the average telomere length at conception, telomerase negative cells have the capacity for about 100 cell divisions before telomeres get critically short (Harley, 1995). As telomeres shorten, the cell will at some point go into replicative arrest. It has been suggested that signalling from the shortest telomere activates the DNA-quality control pathway of the cell (Vaziri, 1997, Vaziri et al., 1997, Chin et al., 1999), or alternatively, that it is the mean telomere length that determines the onset of replicative senescence (Martens et al., 2000), possibly through a process with a pronounced stochastic element (Blackburn, 2000).

Immediately before senescence, chromosomes with short telomeres show a tendency to form dicentric chromosomes, probably as a mechanism for end protection (DeLange et al., 1995). This phenomenon may under certain circumstances be problematic since end-to-end fusion of chromosomes can lead to aneuploidy, one of the well-established steps in carcinogenesis. It is therefore important to obtain knowledge concerning maintenance and regulation of telomere length. An impressive body of information on molecular components regulating this system has emerged in recent years, and information about the variations of mean telomere length in relation to various conditions has been obtained (reviewed in: Holt and Shay (1999)). Since it is still not established whether it is the shortest telomeres or the mean telomere length that is the determining factor with regards to consequences of telomere loss—be it in connection with normal aging or in connection with disease—it is surprising that so little research have looked into the length of the individual telomere. Only a few groups have done detailed studies on the chromosome specific pattern of telomere length in man (Lansdorp et al., 1996, Martens et al., 1998, Martens et al., 2000). Using quantitative fluorescence in situ hybridization, Martens et al. (1998) found a sixfold variation in telomere length within the same metaphase and suggested a certain pattern of telomere lengths within individuals. With regards to person-to-person similarities the same group found signs of this, and suggested that chromosome 17p had a tendency always to be one of the shortest (Martens et al., 1998).

In the present communication we have performed a more extensive study on the chromosome specific pattern of telomere length (a telomere profile) in humans, mainly of advanced age. We have used FISH analysis of metaphase spreads and FITC-labelled PNA probes specific for the telomere repeat sequence. We have found evidence for a common pattern shared by all individuals in the study, but we have also found that there are individual specific characteristics in this pattern. By investigating chromosomes with translocations we have found that the chromosome specific pattern is determined by factors located very distally on the chromosome arms, and by studying telomere lengths in twins we have found that the factors that determine individual differences in the telomere profile appear to be inherited. Finally, we have found a correlation between telomere length and the frequency of age related aneuploidy.