Recombination occurs during telomere formation in yeast

Short stretches of cloned telomeric sequences are necessary and sufficient for telomere formation in yeast as long as the sequences are present in the same orientation as they are found in vivo. During telomere formation, DNA termini usually undergo RAD52-independent recombination with other DNA termini as would be predicted by models of recombination-mediated telomere replication.     

DNA sequences of telomeres maintained in yeast

Telomeres, the ends of eukaryotic chromosomes, have long been recognized as specialized structures. Their stability compared with broken ends of chromosomes suggested that they have properties which protect them from fusion, degradation or recombination. Furthermore, a linear DNA molecule such as that of a eukaryotic chromosome must have a structure at its ends which allows its complete replication, as no known DNA polymerase can initiate synthesis without a primer. At the ends of the relatively short, multi-copy linear DNA molecules found naturally in the nuclei of several lower eukaryotes, there are simple tandemly repeated sequences with, in the cases analysed, a specific array of single-strand breaks, on both DNA strands, in the distal portion of the block of repeats. In general, however, direct analysis of chromosomal termini presents problems because of their very low abundance in nuclei. To circumvent this problem, we have previously cloned a chromosomal telomere of the yeast Saccharomyces cerevisiae on a linear DNA vector molecule. Here we show that yeast chromosomal telomeres terminate in a DNA sequence consisting of tandem irregular repeats of the general form C1-3A. The same repeat units are added to the ends of Tetrahymena telomeres, in an apparently non-template-directed manner, during their replication on linear plasmids in yeast. Such DNA addition may have a fundamental role in telomere replication.     

Recognition of a chromosome truncation site associated with alpha-thalassaemia by human telomerase

Telomeres define the ends of chromosomes; they consist of short tandemly repeated DNA sequences loosely conserved in eukaryotes (G1-8(T/A)1-4). Telomerase is a ribonucleoprotein which, in vitro, recognizes a single-stranded G-rich telomere primer and adds multiple telomeric repeats to its 3' end by using a template in the RNA moiety. In conjunction with other components, telomerase may balance the loss of telomeric repeats due to DNA replication. Another role of telomerase may be the de novo formation of telomeres. In eukaryotes like Tetrahymena, this process is an integral part of the formation of macronuclear chromosomes. In other eukaryotes this process stabilizes broken chromosomes. A case of human alpha-thalassaemia is caused by a truncation of chromosome 16 that has been healed by the addition of telomeric repeats (TTAGGG)n. Using an in vitro assay, I show here that human telomerase correctly recognizes the chromosome 16 breakpoint sequence and adds (TTAGGG)n repeats. The DNA sequence requirements are minimal and seem to define two modes of DNA recognition by telomerase.     

Inhibition of telomerase by G-quartet DNA structures

The ends or telomeres of the linear chromosomes of eukaryotes are composed of tandem repeats of short DNA sequences, one strand being rich in guanine (G strand) and the complementary strand in cytosine. Telomere synthesis involves the addition of telomeric repeats to the G strand by telomere terminal transferase (telomerase). Telomeric G-strand DNAs from a variety of organisms adopt compact structures, the most stable of which is explained by the formation of G-quartets. Here we investigate the capacity of the different folded forms of telomeric DNA to serve as primers for the Oxytricha nova telomerase in vitro. Formation of the K(+)-stabilized G-quartet structure in a primer inhibits its use by telomerase. Furthermore, the octanucleotide T4G4, which does not fold, is a better primer than (T4G4)2, which can form a foldback structure. We conclude that telomerase does not require any folding of its DNA primer. Folding of telomeric DNA into G-quartet structures seems to influence the extent of telomere elongation in vitro and might therefore act as a negative regulator of elongation in vivo.     

Telomeres shorten during ageing of human fibroblasts

The terminus of a DNA helix has been called its Achilles' heel. Thus to prevent possible incomplete replication and instability of the termini of linear DNA, eukaryotic chromosomes end in characteristic repetitive DNA sequences within specialized structures called telomeres. In immortal cells, loss of telomeric DNA due to degradation or incomplete replication is apparently balanced by telomere elongation, which may involve de novo synthesis of additional repeats by novel DNA polymerase called telomerase. Such a polymerase has been recently detected in HeLa cells. It has been proposed that the finite doubling capacity of normal mammalian cells is due to a loss of telomeric DNA and eventual deletion of essential sequences. In yeast, the est1 mutation causes gradual loss of telomeric DNA and eventual cell death mimicking senescence in higher eukaryotic cells. Here, we show that the amount and length of telomeric DNA in human fibroblasts does in fact decrease as a function of serial passage during ageing in vitro and possibly in vivo. It is not known whether this loss of DNA has a causal role in senescence.     

An unusual DNA structure detected in a telomeric sequence under superhelical stress and at low pH

Telomeric sequences of DNA, which are found at the ends of linear chromosomes, have been attracting attention as potential sites for the formation of unusual DNA structures. They consist of (GnTm) or (GnATm) motifs (n greater than or equal to m) and, in the single-stranded state, form hairpins stabilized by non-canonical G.G pairs. In the duplex state and under superhelical stress they exhibit hypersensitivity to SI nuclease which by analogy with homopurine-homopyrimidine sequences may reflect the formation of an unusual structure. To determine whether this is the case we have inserted into a plasmid the Tetrahymena telomeric motif (G4T2).(A2C4) and probed it by two-dimensional gel electrophoresis, chemical modification and oligonucleotide binding. Our data demonstrate that, under superhelical stress and at low pH, the insert does indeed adopt a novel DNA conformation. We have concluded that in this structure the C-rich strand forms a hairpin stabilized by non-Watson-Crick base pairs C.C+ and A.A+, whereas the G-rich strand remains unstructured. We term this new DNA structure the (C,A)-hairpin.     

A truncated human chromosome 16 associated with alpha thalassaemia is stabilized by addition of telomeric repeat (TTAGGG)n

The instability of chromosomes with breaks induced by X-irradiation led to the proposal that the natural ends of chromosomes are capped by a specialized structure, the telomere. Telomeres prevent end-to-end fusions and exonucleolytic degradation, enable the end of the linear DNA molecule to replicate, and function in cell division. Human telomeric DNA comprises approximately 2-20 kilobases (kb) of the tandemly repeated sequence (TTAGGG)n oriented 5'----3' in towards the end of the chromosome, interspersed with variant repeats in the proximal region. Immediately subtelomeric lie families of unrelated repeat motifs (telomere-associated sequences) whose function, if any, is unknown. In lower eukaryotes the formation and maintenance of telomeres may be mediated enzymatically (by telomerase) or by recombination; in man the mechanisms are poorly understood, although telomerase has been identified in HeLa cells. Here we describe an alpha thalassaemia mutation associated with terminal truncation of the short arm of chromosome 16 (within band 16p13-3) to a site 50 kb distal to the alpha globin genes, and show that (TTAGGG)n has been added directly to the site of the break. The mutation is stably inherited, proving that telomeric DNA alone is sufficient to stabilize the broken chromosome end. This mechanism may occur in any genetic disease associated with chromosome truncation.     

Telomeric DNA dimerizes by formation of guanine tetrads between hairpin loops

The telomeric ends of eukaryotic chromosomes are composed of simple repeating sequences in which one DNA strand contains short tracts of guanine residues alternating with short tracts of A/T-rich sequences. The guanine-rich strand is always oriented in a 5'-3' direction towards the end of the chromosome and is extended to produce a 3' overhang of about two repeating units in species where the telomeric terminus is known. This overhang has been implicated in the formation of several unusual intra-and intermolecular DNA structures, although none of these structures has been characterized fully. We now report that oligonucleotides encoding Tetrahymena telomeres dimerize to form stable complexes in solution. This salt-dependent dimerization is mediated entirely by the 3'-terminal telomeric overhang (TT-GGGGTTGGGG) and produces complexes in which the N7 position of every guanine in the overhangs is chemically inaccessible. We therefore propose that telomeric DNA dimerizes by hydrogen bonding between two intramolecular hairpin loops, to form antiparallel quadruplexes containing cyclic guanine base tetrads. These novel hairpin dimers may be important in telomere association and recombination and could also provide a general mechanism for pairing two double helices in other recombinational processes.     

Cloning of human telomeres by complementation in yeast

Telomeres confer stability on chromosomes by protecting them from degradation and recombination and by allowing complete replication of the end. They are genetically important as they define the ends of the linkage map. Telomeres of lower eukaryotes contain short repeats consisting of a G-rich and a C-rich strand, the G-rich strand running 5'-3' towards the telomere and extending at the end. Telomeres of human chromosomes share characteristics with those of lower eukaryotes including sequence similarity as detected by cross-hybridization. Telomeric repeats from many organisms can provide telomere function in yeast. Here we describe a modified yeast artificial chromosome (YAC) vector with only one telomere which we used to clone human telomeres by complementation in yeast. YACs containing human telomeres were identified by hydridization to an oligonucleotide of the trypanosome telomeric repeat. A subcloned human fragment from one such YAC is immediately subtelomeric on at least one human chromosome.     

TPP1 is a homologue of ciliate TEBP-beta and interacts with POT1 to recruit telomerase

Telomere dysfunction may result in chromosomal abnormalities, DNA damage responses, and even cancer. Early studies in lower organisms have helped to establish the crucial role of telomerase and telomeric proteins in maintaining telomere length and protecting telomere ends. In Oxytricha nova, telomere G-overhangs are protected by the TEBP-alpha/beta heterodimer. Human telomeres contain duplex telomeric repeats with 3' single-stranded G-overhangs, and may fold into a t-loop structure that helps to shield them from being recognized as DNA breaks. Additionally, the TEBP-alpha homologue, POT1, which binds telomeric single-stranded DNA (ssDNA), associates with multiple telomeric proteins (for example, TPP1, TIN2, TRF1, TRF2 and RAP1) to form the six-protein telosome/shelterin and other subcomplexes. These telomeric protein complexes in turn interact with diverse pathways to form the telomere interactome for telomere maintenance. However, the mechanisms by which the POT1-containing telosome communicates with telomerase to regulate telomeres remain to be elucidated. Here we demonstrate that TPP1 is a putative mammalian homologue of TEBP-beta and contains a predicted amino-terminal oligonucleotide/oligosaccharide binding (OB) fold. TPP1-POT1 association enhanced POT1 affinity for telomeric ssDNA. In addition, the TPP1 OB fold, as well as POT1-TPP1 binding, seemed critical for POT1-mediated telomere-length control and telomere-end protection in human cells. Disruption of POT1-TPP1 interaction by dominant negative TPP1 expression or RNA interference (RNAi) resulted in telomere-length alteration and DNA damage responses. Furthermore, we offer evidence that TPP1 associates with the telomerase in a TPP1-OB-fold-dependent manner, providing a physical link between telomerase and the telosome/shelterin complex. Our findings highlight the critical role of TPP1 in telomere maintenance, and support a yin-yang model in which TPP1 and POT1 function as a unit to protect human telomeres, by both positively and negatively regulating telomerase access to telomere DNA.     

Crystal structure of parallel quadruplexes from human telomeric DNA

Telomeric ends of chromosomes, which comprise noncoding repeat sequences of guanine-rich DNA, are fundamental in protecting the cell from recombination and degradation. Disruption of telomere maintenance leads to eventual cell death, which can be exploited for therapeutic intervention in cancer. Telomeric DNA sequences can form four-stranded (quadruplex) structures, which may be involved in the structure of telomere ends. Here we describe the crystal structure of a quadruplex formed from four consecutive human telomeric DNA repeats and grown at a K(+) concentration that approximates its intracellular concentration. K(+) ions are observed in the structure. The folding and appearance of the DNA in this intramolecular quadruplex is fundamentally different from the published Na(+)-containing quadruplex structures. All four DNA strands are parallel, with the three linking trinucleotide loops positioned on the exterior of the quadruplex core, in a propeller-like arrangement. The adenine in each TTA linking trinucleotide loop is swung back so that it intercalates between the two thymines. This DNA structure suggests a straightforward path for telomere folding and unfolding, as well as ways in which it can recognize telomere-associated proteins.     

Defects in mismatch repair promote telomerase-independent proliferation

Mismatch repair has a central role in maintaining genomic stability by repairing DNA replication errors and inhibiting recombination between non-identical (homeologous) sequences. Defects in mismatch repair have been linked to certain human cancers, including hereditary non-polyposis colorectal cancer (HNPCC) and sporadic tumours. A crucial requirement for tumour cell proliferation is the maintenance of telomere length, and most tumours achieve this by reactivating telomerase. In both yeast and human cells, however, telomerase-independent telomere maintenance can occur as a result of recombination-dependent exchanges between often imperfectly matched telomeric sequences. Here we show that loss of mismatch-repair function promotes cellular proliferation in the absence of telomerase. Defects in mismatch repair, including mutations that correspond to the same amino-acid changes recovered from HNPCC tumours, enhance telomerase-independent survival in both Saccharomyces cerevisiae and a related budding yeast with a degree of telomere sequence homology that is similar to human telomeres. These results indicate that enhanced telomeric recombination in human cells with mismatch-repair defects may contribute to cell immortalization and hence tumorigenesis.     

POT1 as a terminal transducer of TRF1 telomere length control

Human telomere maintenance is essential for the protection of chromosome ends, and changes in telomere length have been implicated in ageing and cancer. Human telomere length is regulated by the TTAGGG-repeat-binding protein TRF1 and its interacting partners tankyrase 1, TIN2 and PINX1 (refs 5-9). As the TRF1 complex binds to the duplex DNA of the telomere, it is unclear how it can affect telomerase, which acts on the single-stranded 3' telomeric overhang. Here we show that the TRF1 complex interacts with a single-stranded telomeric DNA-binding protein--protection of telomeres 1 (POT1)--and that human POT1 controls telomerase-mediated telomere elongation. The presence of POT1 on telomeres was diminished when the amount of single-stranded DNA was reduced. Furthermore, POT1 binding was regulated by the TRF1 complex in response to telomere length. A mutant form of POT1 lacking the DNA-binding domain abrogated TRF1-mediated control of telomere length, and induced rapid and extensive telomere elongation. We propose that the interaction between the TRF1 complex and POT1 affects the loading of POT1 on the single-stranded telomeric DNA, thus transmitting information about telomere length to the telomere terminus, where telomerase is regulated.     

Insertion of the eukaryotic transposable element Ty1 creates a 5-base pair duplication

The his4-912 mutation results from insertion of a transposable element into the 5'-non-coding region of the his4 gene of yeast. A duplication of 5 base pairs of wild-type his4 DNA flanks the inserted element. The major class of His+ revertants result from excision of most of the transposable element, leaving an inserted segment of 334 base pairs flanked by the 5-base pair repeats.     

Telomerase primer specificity and chromosome healing

Chromosome healing by de novo telomere addition at nontelomeric sites has been well characterized in several organisms. The Tetrahymena telomerase ribonucleoprotein uses an internal RNA template to catalyse d(TTGGGG)n telomere addition to the 3' end of telomeric sequence in vitro and in vivo. Studies of telomerase RNA indicated that hybridization of the RNA template region, 5'-CAACCCCAA-3', to the 3' end of single-stranded telomeric oligonucleotides might be important for primer recognition and utilization. The apparent requirement of telomerase for pre-existing telomeric sequence has raised questions regarding its role in chromosome healing. We report here that Tetrahymena telomerase can specifically elongate single-stranded DNA oligonucleotides whose termini are not complementary to the RNA template sequence 5'-CAACCCCAA-3'. These data suggest that telomerase may be able to heal chromosomes directly in vivo.     

Covalently closed circles of adenovirus 5 DNA

The genome of adenoviruses is a double-stranded linear DNA molecule with inverted terminal repeats about 100 base pairs (bp) in length and a terminal protein covalently linked to the 5' nucleotide of each strand. Both of these features permit the formation of DNA circles, the inverted repeats allowing the circularization of single-stranded DNA and the terminal protein the joining of one or more molecules to yield double-stranded circles or concatemers. However, although the existence of covalently closed circles has been postulated, double-stranded viral DNA purified from virions or infected cells by conventional methods (that is, using proteases and phenol or chloroform) has always been obtained in a linear form. Here, we present evidence for the existence in adenovirus 5 (Ad5) infected cells of novel structures resulting from covalent head-to-tail joining of viral DNA molecules and show that these structures are due at least in part to the formation of covalently closed circles.     

DNA self-recognition in the structure of Pot1 bound to telomeric single-stranded DNA

Telomeres, specialized protein-DNA complexes that cap the ends of linear chromosomes, are essential for protecting chromosomes from degradation and end-to-end fusions. The Pot1 (protection of telomeres 1) protein is a widely distributed eukaryotic end-capping protein, having been identified in fission yeast, microsporidia, plants and animals. Schizosaccharomyces pombe Pot1p is essential for telomere maintenance, and human POT1 has been implicated in telomerase regulation. Pot1 binds telomeric single-stranded DNA (ssDNA) with exceptionally high sequence specificity, the molecular basis of which has been unknown. Here we describe the 1.9-A-resolution crystal structure of the amino-terminal DNA-binding domain of S. pombe Pot1p complexed with ssDNA. The protein adopts an oligonucleotide/oligosaccharide-binding (OB) fold with two loops that protrude to form a clamp for ssDNA binding. The structure explains the sequence specificity of binding: in the context of the Pot1 protein, DNA self-recognition involving base-stacking and unusual G-T base pairs compacts the DNA. Any sequence change disrupts the ability of the DNA to form this structure, preventing it from contacting the array of protein hydrogen-bonding groups. The structure also explains how Pot1p avoids binding the vast excess of RNA in the nucleus.