1
Telomeres: All’s Well That Ends Well

Patrizia ALBERTI

Structure and Instability of Genomes, National Museum of Natural History, CNRS, INSERM, Paris, France

1.1. Introduction

Eukaryotes have linear chromosomes, which means that each chromosome has two extremities. Two crucial questions underlie the studies that have led, over the years, to the understanding of the peculiar DNA–protein structures that constitute the extremities of eukaryotic linear chromosomes. The first question concerns the so-called “end-protection problem”, raised by the observations of Müller and McClintock in the 1930s–1940s (Müller 1938; McClintock 1941). When a chromosome undergoes a double-strand break, specialized cellular components try to repair it. How do the repair machineries distinguish between the extremities of a double-strand break and the extremities of a linear chromosome? What makes the extremities of a chromosome different from the extremities of a double-strand break? The second question concerns the so-called “end-replication problem”, raised by Olovnikov and Watson in the 1970s (Watson 1972; Olovnikov 1973). DNA synthesis by polymerases proceeds from 5′ to 3′ and requires an RNA primer for replication initiation. As a consequence of this directional mechanism, the very 3′ ends of a chromosome cannot be fully replicated, since, once the RNA primer near the 3′ end is removed, it leaves a 3′ single-stranded tail that cannot be replicated. Hence, in the absence of counterbalancing mechanisms, chromosome ends shorten at each round of DNA duplication. How do eukaryotes resolve these problems? In the late 1970s, researchers began to discover the features that make extremities of chromosomes in eukaryotes so special and understand how these features ensure the proper replication and protection of chromosome ends, preventing inappropriate repair (that would lead to chromosome fusions) and loss of genetic information. Basically, the ends of eukaryotic chromosomes, named telomeres, feature a peculiar DNA composition, a peculiar protein composition and a mechanism that is able to elongate them. In a large variety of eukaryotes, telomeres exhibit a further characteristic: owing to their particular DNA composition, one of the two telomeric strands can fold into four-stranded structures, named “quadruplexes of guanines” or “G-quadruplexes” or simply “G4s”.

In this chapter, I will provide an overview of the history and structures of telomeres, from ciliates (where the main features of telomeres were discovered) to other eukaryotes. Section 1.2 retraces the pathways of two independent discoveries: the discovery of G-quadruplexes and the discovery of the peculiar properties of the ends of chromosomes. We will see when and where G-quadruplexes and telomeres finally met. Section 1.3 provides an overview of DNA sequences, proteins and structures of telomeres in model eukaryotes. My aim is to show how telomeric architectures in eukaryotes present common aspects modulated by species-specific aspects and constitute a fascinating example of the unity and diversity of life. Section 1.4 focuses on G-quadruplexes at telomeres: their structure and what they represent for proteins that deal with them. G-quadruplexes will be the starting and end points of this overview.

1.2. The beginning of the end

In this section, I will retrace the first steps of two independent research topics: one that led to the discovery of G-quadruplexes and the other to the discovery of the peculiar properties of the ends of chromosomes. It will be a journey in the past, from chemistry and biophysics to biology, an example of how unrelated research pathways (G-quadruplexes and telomeres) can at some point cross each other, paving the way for a deeper understanding.

1.2.1. Chemists and biophysicists at work: G-quartets and G-quadruplexes disclose themselves

“DNA comes in many forms” (Rich 1993). Although DNA is a linear polymer basically composed of only four types of nucleotides, it exhibits extraordinary properties of self-assembly and can fold into many conformations. The structure of the DNA double-helix was published in 1953 and immediately opened the way to the understanding of how genetic material is duplicated (Watson and Crick 1953). Just a few years after the elucidation of the structure of DNA in its double-helix form, a new class of DNA structures, G-quadruplexes, began to emerge in the studies of chemists and biophysicists. Nevertheless, as we will see, it took many years before biologists become interested in these structures.

The fact that concentrated solutions of guanylic acid (i.e. monomers of guanosine monophosphate (GMP)) form a gel has been known since the early 1900s (it was reported by the biochemist Ivar Bang in 1910). The nature of this gel was unveiled only in the early 1960s, when X-ray diffraction patterns of fibers from dried GMP gels clearly showed “a helical structure with four units per turn of the helix” (Gellert et al. 1962). It was proposed that GMP self-assembled into a cyclic arrangement of four hydrogen-bonded guanines and that these tetrads (later named “G-quartets”) stacked on top of each other, forming a sort of cylindrical structure with a hole (a channel) in the middle (Figure 1.1) (Gellert et al. 1962).

The idea of a structure based on tetrads of guanines had already been proposed a few years earlier by Alexandre Rich to interpret X-ray diffraction pattern of fibers of polyinosinic acid (poly(rI), an RNA polymer made of deaminated guanines) (Rich 1958), proposed and rejected in a single article! Rich proposed and wrongly rejected the hypothesis of a tetrahelical structure on the basis of a right remark: he thought that the large hole in the center of a four-stranded structure, if unfilled, would introduce considerable instability. Subsequent studies on polyguanylic acid (poly(rG), an RNA polymer made of guanines) supported the formation of a four-stranded helical structure based on tetrads of guanines (Zimmermann et al. 1975), and the hole in the middle of the structure turned out to be one of the key factors in the stability of this class of structures. Let us see why.

One of the first studies on a guanosine gel showed that its stability strongly depended on the nature of the cation present in the solution. Intriguingly, when plotting the melting temperature of the gel as a function of the ionic radius of several monocations (radius Li+ < Na+ < K+ < Rb+ < NH4+ < Cs+), the result was a bell-shaped curve with a maximum corresponding to the potassium cation (Chantot and Guschlbauer 1969): the potassium cation (K+) was the one that stabilized the gel most strongly, while the smaller cations lithium (Li+) and sodium (Na+) and the larger cations rubidium (Rb+), ammonium (NH4+) and cesium (Cs+) stabilized the gel to a lesser extent. This puzzling behavior was referred to as “cation anomaly” because it contrasted with what was observed for polyphosphate long-chains and calf thymus DNA, for which the binding constants of alkali metal ions decreased slightly with increasing metal ion radius (Strauss and Ross 1959; Ross and Scruggs 1964). The “cation anomaly” displayed by guanine tetrahelical structures was understood only 10 years later, when researchers grasped that, depending on their dimensions, cations can reside in the electronegative central cavity of the tetrahelical structure and coordinate (more or less tightly, depending on their dimensions) with the carbonyl oxygens of the guanines, thus stabilizing the structures (Figure 1.1) (Miles and Frazier 1978; Pinnavaia et al. 1978). This explained why the stabilizing effect of alkali metal ions on guanine tetrahelical structures (gel and polymers) followed this ranking: K+ > Na+, Rb+ >> Li+, Cs+: in order to be tightly coordinated to tetrads of guanines, a cation must be small enough to fit in the electron-rich cavity and large enough to bridge the guanine carbonyl oxygens. Na+ can fit in a tetrad, the larger K+ can lay between two tetrads, Li+ is too small to coordinate tightly with the carbonyl oxygens, Cs+ is too large to fit in the cavity and to allow the optimal distance between two tetrads. It did not escape the attention of researchers that, curiously, the cation that stabilized G-quadruplexes the most was the physiologically relevant cation potassium (Miles and Frazier 1978).

Figure 1.1. G-quartet: The four guanines auto-assembly through the formation of eight hydrogen bonds between N(1) and N(2) as donors and O(6) and N(7) as acceptors. G-quartets stack on top of each other forming a helical structure. The auto-assembly is driven by the presence of a cation (M+) that fits into the central cavity and coordinates with the carbonyl oxygens of guanines

Twenty years have passed from the first publication of an X-ray diffraction image of a G-quadruplex structure in 1958 (Rich 1958) to the understanding of the “cation anomaly” exhibited by this class of structures. We leave here the pathway of the discovery of G-quadruplexes to follow the pathway of the discovery of the peculiar characteristics of...