Can Metabolic Mechanisms of Stem Cell Maintenance Explain Aging and the Immortal Germline?

Abstract

The mechanisms underlying the aging process are not understood. Even tissues endowed with somatic stem cells age while the germline appears immortal. I propose that this paradox may be explained by the pervasive use of glycolysis by somatic stem cells as opposed to the predominance of mitochondrial respiration in gametes.

The question ‘Why do we age’ is not solved. Understanding the mechanisms driving aging may lead to innovative strategies to increase health span, an effort that would carry enormous human and economic benefit. The fact that many species (typically, though not exclusively, more slowly developing, longer-lived and larger species) possess somatic stem cells capable of self-renewal and tissue regeneration calls into question why these organisms and their somatic stem cells do age whereas the germline apparently does not. It is also unclear how evolutionary theories of aging that are currently accepted as at least plausible can be reconciled with the biological properties of somatic stem cells. It is proposed here that somatic stem cell maintenance mechanisms lead to preferential accumulation, rather than disposal, of damaged stem cells. On the other hand stringent selection in the germ line renders this lineage seemingly immortal. Furthermore, use of glycolysis for ATP production in somatic stem cells as opposed to mitochondrial respiration in the germ line suggests that mitochondria play a critical role in stem cell maintenance and gamete selection. This hypothesis is consistent with prevailing evolutionary theories of aging, which I will introduce below, and with a critical role for mitochondria in aging.

Evolutionary theories of aging

The most efficacious medical interventions harness naturally evolved processes such as wound repair (surgery), immunological memory (vaccination), endocrine hormones (hormone replacement therapies), and antimicrobials. Understanding the evolutionary rationale of the aging process is therefore critical if we are to prolong life/health span. Several evolutionary theories attempt to explain aging, though none satisfactorily (Kirkwood, 2005). Currently, the most widely accepted explanations are the mutation accumulation and the disposable soma theories (Figure 1). The basis of the mutation accumulation theory is the fact that individuals in the wild typically do not reach old age because of starvation, predation and exposure (external mortality). This external mortality allows deleterious alleles whose effects occur late in life, and therefore are less subject to purifying selection, to accumulate in a population. The disposable soma theory posits that aging is the consequence of the separation of soma and germline (perhaps driven by external mortality), and is caused by preferential allocation of resources to the germline at the expense of the soma.

Figure 1.

Schematic representation of the disposable soma and the mutation accumulation theories of aging.

Both theories suffer from a lack of defined underlying mechanisms, and from observations that have highlighted major exceptions. For example, under the assumptions of the mutation accumulation theory, increased predation should shorten longevity. However, guppies exposed to increased predation did not evolve earlier senescence (Kirkwood, 2005). The disposable soma theory, which focuses mainly on allocation of metabolic resources, may be valid in hermaphroditic species such as C. elegans, where the germline makes up large fraction of the organism. It is unclear however which resources are differentially allocated to the germline at the expense of the soma in sexually dimorphic species, where the metabolic investment in reproduction is lower, in particular in males, which often even have shorter life spans than females.

Maintenance mechanisms in somatic stem cells

The function of somatic stem cells declines with age (Kirkwood, 2005; Snoeck, 2013), and this decline is at least in part explained by cell-intrinsic mechanisms (Snoeck, 2013). While often viewed as a degenerative condition, aging of somatic stem cells may in fact be a reflection of the pervasive action of protective stem cell maintenance mechanisms that confer differential susceptibility to stress and injury compared to mature cells. Maintenance mechanisms in hematopoietic stem cells (HSCs), the best-characterized postnatal stem cells, include quiescence (associated with the use of the error-prone non-homologous end joining (NEHJ) DNA repair pathway), more active autophagy, and higher resistance to starvation and radiation-induced apoptosis compared to their progeny (Blanpain et al., 2011; Ito and Suda, 2014). Though not as thoroughly investigated, similar maintenance mechanisms are operative in other stem cells. Skin stem cells in the bulge of the hair follicle use NHEJ for DSB repair, have a more transient DNA damage response, and are less radiosensitive than progenitors. Mammary gland stem cells also display radioresistance (Blanpain et al., 2011). There is little or no evidence that stem cell-protective mechanisms fail with age. Radioresistance, starvation resistance, increased autophagy and the quiescence-associated use of the NHEJ DNA repair pathway suggest that somatic stem cells favor repair, even though incomplete, over disposal by damage-induced apoptosis. This preference would lead to stem cell compartments that expand with age, as has been shown for HSCs (reviewed in (Snoeck, 2013), and in doing so temporarily maintain their overall function. However, these mechanisms ultimately lead to accumulation of damaged stem cells. Indeed, the repopulation capacity of individual HSCs becomes severely compromised with age (Snoeck, 2013). In contrast, the germline, immortal in a transgenerational sense, uses exactly the inverse mechanism, selection of the fittest gametes (Wallace, 2013). Similarly, asexual reproduction may involve stringent selection of somatic cells.

Evolved mechanisms must affect reproductive fitness to be subject to selective pressure. Somatic stem cell maintenance mechanisms are likely geared to maximize the probability of reaching reproductive age. Selection, as observed in the germline, would in somatic stem cells lead to rapid attrition and therefore to insufficient tissue maintenance and replacement to be compatible with life up to reproductive age. A similar mechanism of differential maintenance and selection may play a role in the well-established skewing in the differentiation potential of HSCs with age as well. Individual HSCs are heterogeneous with respect to differentiation potential, ranging from predominantly myeloid to predominantly lymphoid. Myeloid-biased HSCs become predominant with age (Snoeck, 2013). While mostly described as a consequence of aging, a shift towards myeloid-biased HSC clones is already observed before birth, suggesting a developmentally regulated mechanism subject to natural selection (Snoeck, 2013). Interpreted in the context of evolved mechanisms of maintenance and selection, stem cell-specific maintenance mechanisms may be more potent in HSCs with predominant myeloid potential, while lymphoid HSCs may be more prone to selection and early attrition. As a consequence, early in life the hematopoietic system generates myeloid cells and long-lived lymphocytes capable of vigorous memory responses, required to protect individuals during reproductive age and to provide passive immunity to their offspring. Later in life long-lived memory lymphocytes persist and myeloid-biased HSCs provide a continuous supply of shorter-lived myeloid cells. What is interpreted as aging may therefore to some extent reflect developmental programming of stem cell maintenance mechanisms aimed at maximizing reproductive fitness.

Role of mitochondria in germline selection and somatic stem cell maintenance

Mechanisms underlying maintenance and accumulation of damaged somatic stem cells should act in a diametrically opposite directions in gametes, where they mediate selection of the fittest to effect reproduction. The differential reliance of somatic stem cells and gametes on mitochondrial respiration for ATP production may be at least one such mechanism (Figure 2).

Figure 2.

Schematic illustration of how the differential use of the mitochondrial respiration in gametes and somatic stem cells leads to selection in the germline and maintenance in somatic stem cells, and contributes to tissue dysfunction and aging.

Oocytes use mitochondrial respiration from pyruvate provided by follicle cells to generate ATP. Strong purifying selection against non-synonymous mutations in mtDNA in genes encoding components of the electron transport chain (ETC) occurs in the female germline, where in humans, only 400 oocytes are ovulated from >10⁶ proto-oocytes (Wallace, 2013). Sperm cells are continuously produced from spermatogonial stem cells (SSCs). Although SSCs undergo age-related functional decline (Oatley and Brinster, 2012), they produce millions of sperm cells, of which the fittest in terms of motility, capacitation and zone pellucida binding will fertilize an oocyte. Although sperm motility also requires ATP generated from glycolysis, sperm’s success at fertilizing the oocyte is profoundly affected by mitochondrial dysfunction. In fact, mitochondrial dysfunction and mtDNA mutations, even those that have no discernible consequences in females, are strongly associated with male infertility (Beekman et al., 2014). Stringent selection of sperm during fertilization based on mitochondrial function is therefore likely. Although male mitochondria are lost after fertilization, such selection mechanism may reduce the transmission of genomic DNA mutations as mitochondrial dysfunction is typically associated with increased oxidative stress.

In contrast to gametes, all somatic stem cells where metabolism has been studied rely predominantly on glycolytic ATP production, while most mature cells use mitochondrial respiration, which is more efficient (Ito and Suda, 2014). Glycolysis in HSCs is typically viewed as a HIF1α-mediated response to the hypoxic BM environment (Ito and Suda, 2014). However, recent evidence suggests that HIF1α is in fact not directly stabilized by hypoxia (except in very severe hypoxia), but by hypoxia-induced reactive oxygen species (ROS) (Sena and Chandel, 2012). These observations run counter to the finding that HSCs are exquisitely sensitive to ROS and exhibit very low ROS levels (Ito and Suda, 2014). Glycolytic ATP production in HSCs may therefore be at least in part hardwired. If gametes are selected based on mitochondrial function, then the pervasive reliance on glycolysis in somatic stem cells, whether hardwired or in response to a hypoxic niche, may achieve the opposite: prevention of selection and maintenance of the stem cell pool, even at the expense of accumulation of subfunctional cells.

A first implication of this idea is that through mechanisms that are as yet unclear, mitochondria may play a key role in somatic stem cell maintenance that is not dependent on ATP production. Indeed, loss of quiescence is typically associated with stem cell depletion and a shift towards OXPHOS (Ito and Suda, 2014), and may be mediated at least to some extent by increased selection and attrition.

A second implication is that mitochondria may be both the driving force and one of the executioners of the aging process. Aging is associated with mitochondrial dysfunction and bioenergetic failure (Wallace, 2013). mtDNA is highly susceptible to mutation and deletion (Wallace, 2013). Absence of mitochondrial selection in somatic stem cells implies imperviousness to accumulation of mitochondrial damage and dysfunction. Supporting this idea are the findings that PolgA^mut mice, which accumulate mtDNA mutations and display features resembling premature aging, show defects in lymphoid and erythroid lineages, but not in HSCs (Norddahl et al., 2011). The permissiveness of stem cells to mitochondrial damage will contribute to age-associated tissue heteroplasmy, which is defined as the presence of cells with increased mtDNA mutation load. This will affect the function of mature cells and tissues, as these predominantly rely on mitochondrial respiration. Furthermore, defects in the ETC associated with mtDNA damage will increase ROS production. ROS recruit at least some stem cells into cell cycle and differentiation (Ito and Suda, 2014). In this way heteroplasmic stem cells could progressively, and even preferentially participate in tissue remodeling. Preferential recruitment of stem cells that have accumulated mtDNA damage may also contribute to the mechanistically ill-understood phenomenon of the apparent selection for heteroplastic cells in tissues of patients with maternally transmitted mtDNA mutations (Wallace, 2013).

Tissues with low turnover and interventions that increase life span

In some tissues, such as the central nervous system, cellular replacement is low to regionally absent. Neurons are extremely reliant on mitochondrial respiration. Interestingly, defects in mitochondrial quality control particularly affect the nervous system (Rugarli and Langer, 2012), suggesting that, in contrast to somatic stem cells, the brain evolved and requires potent mitochondrial quality control mechanisms, which we predict to be less active in somatic stem cells. Such tissues may nevertheless be profoundly affected by age-associated changes in other tissues. For example, aging is characterized by a state of inflammation that may set in motion or perpetuate aging in other tissues. The aging hematopoietic system may itself be a source of inflammatory mediators, produced by senescent memory lymphoid cells, which predominate in the immune system of aged individuals (Snoeck, 2013). Thus, age-associated tissue dysfunction caused by changes in stem cell compartments driven by pervasive stem cell maintenance mechanisms can have systemic consequences, which could in turn also affect stem cell function and the stem cell niche in various tissues.

Several interventions, including caloric restriction, genetic inhibition of nutrient signaling and sirtuin activation extend life span, in particular in shorter-lived organisms, and may extend health span in long-lived organisms (Kirkwood, 2005). These interventions metabolically rewire cells towards more efficient nutrient utilization, and optimize cellular quality control and stress resistance, often at the expense of reproductive capacity. None of these interventions abolishes aging however. The underlying mechanisms likely evolved to allow organismal maintenance during temporary metabolic stress caused primarily by starvation (Kirkwood, 2005), potentially explaining why the largest effects on lifespan are observed in the shortest living organisms.

Stem Cell Maintenance Mechanisms Integrates Theories of Aging

The hypothesis that somatic stem cell maintenance as opposed to germline selection contributes to aging is consistent with the disposable soma theory, which fundamentally posits that the distinction between soma and germline constitutes the origin of the aging process. It would not support the notion that differential allocation of resources between soma and germline mechanistically underlies aging however. It is also stands to reason that this division of labor arose to maintain the species in the face of external mortality, the driving force of aging according to the mutation accumulation theory. Finally, somatic stem maintenance as a mechanism underlying organismal aging is remarkably consistent with a variant of the mutation accumulation theory, the antagonistic pleiotropy theory, which proposes that mechanisms that provide reproductive or survival benefit early in life are detrimental late in life (Kirkwood, 2005). The idea is testable in a variety of model organisms and stem cells, and raises a number of intriguing questions, such as the role of mitochondria in stem cell maintenance and the role and activity of mitochondrial quality control mechanisms in somatic stem cells.