Telomere biology and age-related diseases

Introduction

Biological age describes the functional status of the body relative to its chronological age. For over half a century researchers have tried to estimate the biological age of individuals using test batteries [1], [2], [3], [4]. These test batteries assess a range of physiological, psychological and cognitive functions. Today, blood biomarkers are increasingly used as they provide complementary information that help to predict an individual’s biological age better.

Deterioration of genomic integrity and genomic instability are key aspects of aging [5], [6]. Telomeres represent the protective end caps of telomeres that are of critical importance for genomic integrity and stability. The continuous shortening of telomeres with increasing age has attracted a lot of interest in recent years [7]. Telomere length has been considered a ‘biological clock’ that is useful for the establishment of an individual’s biological age [8]. In addition, the measurement of telomere length may help to estimate the risk for age-related disease. The present review provides an overview about the structure and function of telomeres, the relationship between telomere length and age-related diseases, and analytical aspects.

Telomeres – structure and function

Telomeres, from Greek telos ‘end’ and meros ‘part’, are nucleoprotein structures located at both ends of a chromosome. They are composed by a non-coding, repetitive DNA sequence and associated proteins that form the shelterin complex. Telomeric DNA is double-stranded for most of its length with a short, G-rich, single-stranded overhang at the 3′-OH end. With the help of the shelterin proteins telomeric DNA folds into a complex three-dimensional structure, which is important for telomere function [7], [8].

Telomeres are evolutionary conserved and were first identified in 1938 by Hermann Müller. Subsequently, the Nobel-prize winners Müller and McClintock identified the protective function of telomeres [9], [10]. The first human telomeres were isolated in 1988 by Moyzis et al. [11]. Milestones in the discovery of telomere biology are shown in Figure 1.

Figure 1:

Milestones in telomere biology discovery.

In humans, telomeres span several kilobase pairs (kbp) and are composed by thousands of repeats of the hexanucleotide TTAGGG [7], [8] (Figure 2). They have two pivotal functions. On one side telomeres are involved in the protection of chromosomal ends, as they prevent unwanted recombination and degradation. On the other side, telomeres play an important role during DNA replication as they prevent the loss of coding DNA. The three-dimensional folding of telomeres protects the free 3′-OH end of each DNA strand from being recognized as a double strand break (DSB). Dysfunctional telomeres activate the DNA damage response (DDR) cascade and drive cells into senescence [8], [13], [14]. It has been shown that the removal of telomeres from yeast chromosomes causes a dramatic loss of DNA [15].

Figure 2:

Telomere structure.

(A) Telomeres are composed by a double strand region of –TTAGGG– repetitions and by a single strand region called G-strand overhang. Two protein complexes are bound to telomeres, the telomere repeat binding factor 1 (TRF1) complex and the telomere repeat binding factor 2 (TRF2) complex. (B) The G-strand overhand can fold back and invades the double strand region leading to the formation of T-loop and D-loop structures. The resulting 3D conformation protects the 3′OH end of the chromosome. (C) Composition of the two main telomere-associated protein complexes. The TRF1 complex is involved in telomere length control, whereas the TRF2 complex functions as protective end cap of telomeres. Modified from Pusceddu et al. [12].

The proteins of the shelterin-complex are involved in the control of telomere length by regulating the access of telomerase to the G-strand overhang and by protecting it from degradation [13], [14]. In addition, with the help of shelterin proteins end-to-end fusions of chromosomes are prevented. The shelterin complex is composed of six telomere-specific proteins: TRF1 (Telomeric Repeat Binding Factor 1), TRF2, RAP1 (Repressor/Activator Protein 1), TIN2 (TRF1 Interacting Nuclear Factor 2), POT1 (Protector of Telomeres 1) und TPP1 (POT1-TIN2 organizing protein) [16]. With the help of the shelterin complex the 3′G rich single-stranded DNA overhang at the very end of all telomeres can fold back and anneal with double-stranded telomeric DNA forming a loop structure known as T-loop (Figure 2). The closed configuration of the T-loop hides the free 3′-OH end at the very end of each DNA-strand thus avoiding it from being recognized as a DSB [7], [8]. Because of its shape and function the T-loop can be seen as a protective cap that defines the natural end of a chromosome. The thymidine- and guanine-rich hexanucleotide of human telomeres is believed to form a quadruple helix [7], [8]. First, the 3′-OH strand overhang pairs with itself by forming abnormal GG-double bonds. The resulting double strand pairs again with itself forming a quadruple helix in which the guanine residues participate in Hoogsten bonds [7], [8].

Every division of a somatic cell is accompanied by a loss of 50–200 bp of telomeric DNA resulting in a continuous shortening of telomeres. The incomplete DNA replication at the 5′-end of all newly synthesized daughter strands is due to a phenomenon called end-replication problem [7], [8], [17]. DNA replication always occurs in 5′→3′ direction and is catalyzed by DNA polymerase. However, the enzyme can only add nucleotides to a free 3′-OH end of an existing nucleotide. At the very beginning of a DNA strand there is no such 3′-OH group to which DNA polymerase can add nucleotides. Therefore, RNA primers that provide a 3′-OH group are needed to start DNA synthesis. Once replication is complete, primers are removed and the resulting gaps are filled with DNA. However, at the 5′-end of each daughter strand, the space formerly occupied by the primer cannot be filled with DNA, as a free 3′-OH group is lacking [17], [18].

In the absence of telomeres, there would be inevitable loss of genetic information from the leading strand with every mitosis. As telomeres are a non-coding sequence, but are composed by repetitive elements, telomere shortening avoids the loss of genetic information. The consequence of telomere shortening is that somatic cells can undergo only a defined number of divisions before telomeres become critically short and lose their protective properties [7], [13], [14]. When a critical telomere length of approximately 4 kbp is reached, cells cannot divide anymore and undergo senescence or apoptosis. Consequently, telomere shortening limits the life span of cells and is an effective tumor suppressor mechanism. When cells with critically short telomeres continue to divide, such as in some cancer cells, chromosomes become instable. Because of the physiologic telomere shortening in somatic cells telomere length can be considered as a mitotic or molecular clock [19].

Telomerase

In contrast to most somatic cells, hematopoietic stem cells, keratinocytes in the basal layer of the epidermis, uterine endometrial cells, germ cells and various tumors avoid telomere shortening by activation of telomerase [20], [21]. Telomerase is a ribonucleoprotein with reverse transcription activity, which adds de novo telomere hexanucleotide repeats to the chromosomal ends [7], [13], [14]. Telomerase contains a highly conserved reverse transcriptase (human telomerase reverse transcriptase, hTERT), an associated template RNA (telomerase RNA component, TERC) and a key auxiliary protein known as Dyskerin. Inducing telomerase activity in primary human fibroblasts by retroviral gene transfer is sufficient to counteract telomere erosion and to prevent cells from entering senescence [7], [13], [14]. The resulting maintenance of telomere length immortalizes most human cell types. In some mammalian cancer cells and immortalized cell lines, telomeres are extended in a telomerase-independent manner, called ALT (Alternative Lengthening of Telomeres). It has been suggested that ALT mechanisms rely on the homologous DNA recombination between telomeric sequences [7], [13], [14].

For many years telomeres were viewed as transcriptionally inert. However, transcription of the C-strand of telomeres by RNA polymerase II produces long UUAGGG-containing transcripts, known as TERRA (telomeric repeat-containing RNA) that seems to have structural functions [7], [8].

Epigenetics – DNA and histone methylation

In addition to the 3D conformational status and the shelterin complex, a third mechanism known as epigenetics, regulates the function of telomeres. Epigenetics include DNA methylation, predominantly at CpG islands and histone modifications, such as methylation and acetylation. In humans, three main DNA methyltransferases (DNMTs) are responsible of the methylation status of DNA. DNMT1 functions as a maintenance DNMT that copies parental strand DNA methylation onto the daughter strand after replication. DNMT3a and DNMT3b function as de novo DNMTs. Human telomere repeats cannot be methylated because they lack CpG sequences, which are the substrates for DNMTs. However, the subtelomeric region contains a high proportion of CpG dinucleotides, which are heavily methylated. For example, DNMT knockdown mice models are characterized by hypomethylation of subtelomeres and abnormal elongation of telomeres. Reintroduction of DNMT3a and DNMT3b in these cells results in the re-methylation of subtelomeric domains and decreased telomere recombination [22], [23], [24]. In addition, methylation is not equally distributed throughout the genome indicating that DNA methylation regulates distinct biologic functions. Furthermore, telomeres and subtelomeres are bound by nucleosomes that are enriched in trimethylated histone 3 Lysine 9 (H3K9). With the help of Histone-Methyltransferases (HMT) and Histone-Demethylases (HDM) histones can be methylated and demethylated at lysine and arginine residues. Cells that lack the HMT are characterized by decreased levels of histone trimethylation and by aberrant telomere elongation. Moreover, it has been shown that progressive telomere shortening leads to epigenetic changes, both in telomeric and subtelomeric regions [22], [23], [24]. However, most of these studies were performed in mice models and the real effect in humans needs to be further elucidated.

Telomere length – influencing factors

It is well established that telomeres shorten with age. This phenomenon is the result of the end-replication problem [25]. Mean LTL at birth is 11 kbp and declines to less then 4 kbp in elderly individuals. A lot of factors have been related to LTL, such as gender, physical activity, smoking, body mass index (BMI), alcohol consumption, hormone replacement therapy, dietary antioxidants, vitamins, trace elements, chronic inflammation, socioeconomic status, perceived stress levels, and paternal age [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [40], [41]. For example, telomeres are longer in women, as the consequence of higher estrogen levels, that increase telomerase activity and have antioxidant effects. Similarly, psychological stress affects cellular aging through oxidative stress and telomerase activity. Highly stressed women are characterized by lower telomerase activity and higher oxidative stress compared to women with a low stress level [39], [40], [41]. In addition, regular physical activity has also been associated with decreased levels of oxidative stress and inflammation, which affect telomere shortening. For example, moderate and vigorous-intensity activity were associated with increased LTL, and telomerase activity was higher among trained athletes and after 3-month of lifestyle intervention, which included the association of moderate aerobic exercise [41].

Finally, several nutritional factors like vitamins (including folate, nicotinamide, vitamin A, B12, D, C and E), minerals (including magnesium, zinc and iron) and other bioactive dietary components (like omega-3 fatty acid, polyphenols and curcumin), are able to directly or indirectly influence LTL trough several mechanisms. Recent studies have shown consistent associations between LTL and the availability of B and D vitamins [12], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38]. Preliminary results from own studies demonstrate that serum folate and its metabolites correlate with LTL (Figure 3) [29]. In agreement with this observation also HCY, a functional marker of folate and vitamin B12 availability, is consistently correlated with LTL (Figure 3). Anti-oxidant activity, DNA methylation and prevention of DNA damage are the most important mechanisms trough which these nutritional factors slow down telomere attrition [12], [27]. In summary, a healthy lifestyle with a diet rich in fruits and vegetables combined with exercise, lower BMI and no smoking is associated with longer telomeres.

Figure 3:

Relative telomere length and markers of B vitamin status.

(A) Correlation between relative telomere length and total folate in 53 elderly subjects of the KNOVIB study [29]. (B) Correlation between relative telomere length and homocysteine in 336 subjects of the STVS study (not published).

Telomeres and age-related diseases

Short telomeres and telomere dysfunction have been linked to numerous disorders. Several genetic disorders with mutations in genes involved in the expression and regulation of shelterins, telomerase or other telomere-related proteins, have been described [26], [42]. All these diseases are characterized by an accelerated rate of telomere shortening. For example, aplastic anemia, dyskeratosis congenital, idiopatic pulmonary fibrosis are all characterized by mutations in TERC, TERT or dyskerin [26], [42]. In epidemiologic studies, individuals with shorter telomeres are characterized by a higher mortality rate, which is nearly twice as high as in those with longer telomeres (Figure 4) [43]. For example, Mons et al. [45] analyzing more than 12,000 subjects of two population-based studies (ESTHER and Nurses’ Health Study), identified that subjects with shorter telomeres (1st quintile) were characterized by higher hazard ratio for all-cause mortality (1.66, 95%CI 1.09–2.53, p=0.018) compared to those with higher telomere length (5th quintile). In addition, a reduced LTL may indicate an existing or an elevated risk for future age-related disease, such as cardiovascular disease (CVD), type 2 diabetes mellitus (T2DM), neurodegenerative diseases, osteoporosis and premature aging syndromes [42], [43], [46], [47], [48], [49].

Figure 4:

Telomere length, mortality and age-related disease.

(A) Mortality by decile of 5-year change in telomere length (p for trend <0.001) [43]. (B) Telomere restriction fragment (TRF) length among healthy women aged 50 and older with clinical osteoporosis at one and two or more sites. Means and standard errors are adjusted for age, smoking status, body mass index, HRT and menopausal status [44]. (C) Survival in 608 patients with coronary heart disease, stratified by change in telomere length during the previous 5 years (overall log rank test p<0.0001) [43].

Leukocyte telomere as a surrogate marker for telomere shortening in the entire body

Measurement of telomere length in tumor tissue harbors substantial practical limitations that prevent its use as a biomarker in clinical practice. The most important issue is that the analysis can only be performed after a tumor has been diagnosed. Furthermore, an invasive biopsy is necessary to obtain a sample. Because of the invasive sample collection, serial measurements are not feasible. In order to overcome these obstacles, the measurement of LTL in blood leukocytes has been proposed as a surrogate marker. In blood leukocytes, telomeres shorten with an average annual rate of 30–35 bp. As this places leukocytes in the middle of all tissue types, it has been proposed that LTL might be a suitable surrogate marker for telomere shortening in the entire body.

LTL is often used in clinical studies. However, doubts remain to what extent this parameter reflects the situation in other organs and tissues. Until today, only few studies have explored the relationship between telomere length in different cell types and results are not so clear [91], [92], [93], [94], [95]. For example, Dlouha et al. [91] measured telomere length in 12 different human tissues (peripheral blood leukocytes, liver, kidney, heart, spleen, brain, skin, triceps, tongue mucosa, intercostal skeletal muscle, subcutaneous fat and abdominal fat) from 12 cadavers. Telomere length was significantly higher in blood compared to liver, brain, muscles, skin, spleen and mucosa but was not different compared to adipose and renal tissue. In addition, the largest telomere length variability was observed in leukocytes and kidney, and the smallest telomere length variability was observed in brain, heart and skin [91]. In another study, Friedrich et al. [93] reported a linear relationships between telomere length measured in blood, skin and synovia of nine elderly patients. In this study, significantly shorter telomeres were found in leukocytes compared to skin and synovia [93]. However, in this study, age of patients ranged from 73 to 95 years. Telomere length from leukocytes and from skin correlated negatively with age [93]. Richard et al. [94], who analyzed telomere length in aortic biopsies and paired blood leukocyte from 20 patients, identified that shorter blood telomeres reflect shorter aortic telomeres, indicating that telomere attrition in blood leukocytes is indicative of similar changes in the vascular well. Altogether these findings could reflect the expected life span of different cell types. One aspect to consider is that senescent cells are eliminated by apoptosis so that only a few or no senescent cells persist in vivo. On the other site, the stem cell compartment of a tissue is enriched in cells with longer telomeres, as they express telomerase and therefore avoid telomere shortening. Indeed, longer telomeres and higher telomerase activity were found in several stem cell compartments, including hair follicle, skin, small intestine, cornea and testis [95]. Finally, it has also been speculated that a common genetically controlled mechanism regulates telomere length differently in various tissues [93].

In summary, available data suggest that telomere length is correlated between different tissues in the same individual and that peripheral blood telomere length could be used as a surrogate marker of telomere length in other organs [91], [92], [93], [94], [95].

Analytical aspects

Telomeres can be measured in all solid organ tissues and blood leukocytes. Although the analysis of telomere length in the tissue of interest is most representative, the invasive sample collection procedure and other practical aspects often render this approach not feasible for routine use. Because of the simple and non-invasive specimen collection, blood leukocytes are the most widely used matrix for clinical and research purposes. Telomere length can be measured with a range of different methods [95], [96], [97], [98], [99], [100], [101], [102], [103]. Generally, methods that use isolated DNA have to be distinguished from those using intact cells. Southern blotting and quantitative-polymerase chain reaction (Q-PCR) are the analytical principals that are applied for the analysis in isolated DNA. The Southern blot technique was the first method that has been developed and is still considered the gold standard for the measurement of telomere length [96]. In the first step, DNA is digested by endonucleases that do not cut within the telomeric TTAGGG sequence. Subsequently, the digested DNA is separated by gel electrophoresis. Telomeres are visualized by labeled DNA probes that are complementary to the telomere sequence. The longer the telomere, the more probes can bind, resulting in a higher signal intensity [96]. The obtained signal represents the average telomere length of all chromosomes from cells in the sample. It provides an absolute value of telomere length expressed in base pair, requires a substantial quantity of DNA (i.e. micrograms), is time consuming and is not suitable for large population-based studies. However, as the analytical principle of this hybridization-based method, it implies a decreasing signal as telomere become shorter. There is a threshold telomere length below which no detectable signal will be generated. The inability to detect very short telomeres limits the utility of this method in studies of cellular aging, where short telomeres are of particular interest [96]. The Q-PCR-based assay requires the amplification of the telomeric region and of a single copy gene (SCG) [97], [98]. Telomere length is quantified as the amount of telomeric sequence compared to SCG and is expressed as a relative telomere length (RTL) or telomere/SCG (T/S) ratio [97], [98]. As the Southern blot method, it is not possible to evaluate telomere length of individual cells or chromosomes. However, the Q-PCR-based technique is fast, highly sensitive and requires low amounts of DNA (i.e. nanograms). Because of these characteristics, it is often chosen for large population-based studies [97], [98]. At the cellular level, telomere length can be evaluated by quantitative fluorescence in situ hybridization (Q-FISH) [100] or by flow cytometry (flow-FISH) [101]. Q-FISH can be combined with a conventional FISH, allowing the evaluation of chromosome-specific telomere length. Similarly to the Southern blot-based method described above, Q-FISH is also a hybridization technique, which implies the existence of a threshold telomere length below which no detectable signal is produced. As Q-FISH is performed on cells in metaphase, the analysis is limited to cells that proliferate in vitro. Consequently, not all cells are suitable to be analyzed with this method. The confocal telomere Q-FISH method (Telomapping) allows the quantification of telomere length in tissue samples or biopsies [95].

More work is needed for the harmonization of telomere length assessment between laboratories, including the development of a common standard reference, of internal and external quality control programs and of reference ranges stratified for age and gender.

Conclusions

Telomeres and telomerase play an important role during the process of aging. Short telomeres in various tissues and in blood leukocytes are associated with age-related diseases, such as CVD, Alzheimer’s disease and cancer. Although a number of studies have proposed the measurement of LTL as a predictor of incident age-related diseases, disease progression and mortality, it is still far from being used in clinical practice. Analytical issues (lack of standardization) and pathophysiological aspects that are insufficiently understood are the main obstacles that impede the widespread use of telomere length measurement. More work is needed for the harmonization of telomere length assessment between laboratories, including the development of a common standard reference, of internal and external quality control programs and of reference ranges stratified for age and gender. In addition, a range of preanalytical aspects need to be clarified. For example, under what circumstances the measurement of LTL should be ordered? When should the blood sample be collected? So far, the evaluation of telomere length will be reasonable only in research programs, clinical trials, longitudinal and cross-sectional clinical studies.