Telomeres, Atherosclerosis, and Human Longevity

Human telomere(definition) length, as expressed in leukocytes, is strongly fashioned by heritability.1–5 A number of genes that explain some of the inter-individual variation in leukocyte telomere length have already been identified.6–9 Leukocyte telomere length is longer in women than in men3,10,11 and in African Americans than in whites of European descent.12,13 In addition, offspring of older fathers display a longer leukocyte telomere length, a finding that may reflect longer telomeres in sperm of older men.14–17 In contrast to the male germline, telomeres in replicative somatic tissues, including the hematopoietic system, display age-dependent shortening. A shorter leukocyte telomere length is associated with aging-related diseases, principally atherosclerosis18,19 and reduced longevity.20,21 The prevailing interpretation is that a shorter leukocyte telomere length is the outcome of processes, mainly the age-dependent accruing burden of oxidative stress and inflammation22 that ultimately shorten the human lifespan. As such, leukocyte telomere length is considered to be a non-causal biomarker, ie, a telomeric “clock,” of human aging. We propose that telomere length is not only associated with atherosclerosis and longevity but is also a causal determinant of both. This hypothesis is based on a model of leukocyte telomere length dynamics (leukocyte telomere length and its attrition rate) that incorporates two key features. First, the model differentiates the findings of studies of telomere length dynamics in vitro from those in vivo. Second, the model distinguishes two phases of human leukocyte telomere length dynamics in vivo: growth and adulthood. THE TELOMERIC “CLOCK” BASED ON EXTRAPOLATIONS FROM IN VITRO STUDIES The metaphor of telomeres as a biological clock is based on research showing that somatic cell replication in culture causes telomere attrition, ultimately leading to replicative senescence(definition) (or apoptosis).23 Although senescence might be triggered by mechanisms other than critically short telomeres,24–26 the evidence is strong that the telomeric clock is a major factor in replicative senescence in vitro. Studies in cultured cells also revealed that oxidative stress heightens the loss of telomere repeats per replication, presumably due to the sensitivity of telomeres to oxidative stress.27 Leukocytes are the most frequently used somatic cells to study telomere biology in epidemiological settings. The traditional model of leukocyte telomere length dynamics, extrapolated from cultured cells to the in vivo state, is based on the premise that oxidative stress augments the loss of telomere repeats per replication of all somatic cells. However, in hematopoietic stem cells (HSCs) not only oxidative stress but also inflammation increases telomere shortening, because HSCs undergo more replication to replace leukocytes consumed by the inflammatory response. Shorter leukocyte telomere length, which reflects HSC telomere length,28,29 has therefore been attributed to a faster HSC telomere-length attrition due to a higher cumulative burden of oxidative stress and inflammation—two biological processes that are considered the hallmarks of atherosclerosis and aging.22 The concept of leukocyte telomere length as a biomarker of aging entails the premise that at birth, if not at conception, all individuals have a biological age—a telomeric clock time—of zero. In newborns, however, we see great variation in leukocyte telomere length, with values ranging from 8 to 11 kb or more.30,31 This variation among newborns exceeds the total average leukocyte telomere length shortening throughout the adult life course (20–90 years) estimated at <2.0 kb, assuming an average leukocyte telomere length attrition rate of 25–30 bp/year.32 Thus, the use of leukocyte telomere length as a biomarker of aging in vivo should account for not only leukocyte telomere length attrition but also leukocyte telomere length at birth for each individual, an important point often overlooked. TELOMERE DYNAMICS IN VIVO IN THE HEMATOPOIETIC SYSTEM Cultured cells typically display exponential proliferation, whereby one cell divides into two daughter cells, ie, symmetric replication. However, in vivo, HSCs experience both symmetric replication to two daughter HSCs and asymmetric replication to a daughter hematopoietic progenitor cell (HPC) and a daughter HSC.33 Symmetric replication, as in cultured cells, serves to expand exponentially the HSC reservoir during growth. It takes, for example, four cycles of symmetric HSC replication to generate 16 HSCs [2-4-8-16]. In contrast, asymmetric HSC replication during growth, which linearly expands the HPC pool and ultimately the peripheral blood cell mass, takes 16 cycles to generate 16 HPCs. Therefore, during growth, expansion of the HPC pool in tandem with the growing soma requires far more replication and more telomere attrition than the expansion of the HSC reservoir.29,34 Accordingly, the expansion of the HPC pool is probably the key explanation for the rapid leukocyte telomere length attrition during growth, which amounts to 1.5–2 kb. The first 20 years of life comprise the full period of growth and cross- sectional analysis clearly shows rapid leukocyte telomere length attrition during this period.35 However, modeling suggests that most leukocyte telomere length attrition during growth occurs during early development (by age 5 years).29 Moreover, recent studies suggest that fetal and childhood exposures might influence leukocyte telomere length,36,37 but this effect might be small compared with the joint effect on leukocyte telomere length of heritability, sex, race, and paternal age. In adulthood, there is still an ongoing need for symmetric and asymmetric HSC replication. Symmetric replication occurs to replace some HSCs that died or experienced senescence. Asymmetric replication accommodates the death and senescence of HPCs and needs for replenishment of the blood cell mass. Based on current evid