DNA damage in aging, the stem cell perspective

Abstract

DNA damage is one of the most consistent cellular process proposed to contribute to aging. The maintenance of genomic and epigenomic integrity is critical for proper function of cells and tissues throughout life, and this homeostasis is under constant strain from both extrinsic and intrinsic insults. Considering the relationship between lifespan and genotoxic burden, it’s plausible the longest-lived cellular populations would face an accumulation of DNA damage over time. Tissue-specific stem cells are multipotent populations residing in localized niches and are responsible for maintaining all lineages of their resident tissue/system throughout life. However, many of these stem cells are impacted by genotoxic stress. Several factors may dictate the specific stem cell population response to DNA damage, including the niche location, life history, and fate decisions after damage accrual. This leads to differential handling of DNA damage in different stem cell compartments. Given the importance of adult stem cells in preserving normal tissue function during an individual’s lifetime, DNA damage sensitivity and accumulation in these compartments could have crucial implications for aging. Despite this, more support for direct functional effects driven by accumulated DNA damage in adult stem cell compartments is needed. This review will present current evidence for the accumulation and potential influence of DNA damage in adult tissue-specific stem cells and propose inquiry directions that could benefit individual healthspan.

Introduction

The field of aging research has been rapidly advancing in recent decades and with it our understanding of its complexity. As a result, our knowledge on the physiological and molecular hallmarks of aging has greatly expanded (Lopez-Otin et al. 2013). Concomitantly, many theories have been proposed to explain what drives these changes (Vina, Borras, and Miquel 2007; Finch and Kirkwood 2000), including free radicals (Harman 1956), semi-stochastic damage accumulation (Gladyshev 2013), immuno-senescence (Franceschi, Bonafe, and Valensin 2000), metabolism and growth (McCay, Crowell, and Maynard 1989), antagonistic pleiotropy (Williams 1957) and others. Despite the increasing data describing aging phenotypes, our understanding of what drives aging remains unclear. Among the many theories, somatic DNA damage has long been a suspected driver of aging (Kirkwood 1977; Szilard 1959) and is extensively investigated (reviewed in Moskalev et al. 2013). Within the soma, adult stem cells are especially relevant to lifelong tissue maintenance as they have the capacity to generate somatic cell lineages. This functional capacity of the tissue specific stem cells is altered with age (Table 1), via both extrinsic and intrinsic factors. One potential driver of this change could be derived from the inability to maintain genomic integrity. This is particularly relevant to the stem cell compartment as their role in the organism is to maintain and restore the function of the tissue or system they belong to. To maintain homeostasis and to repair injured systems, these cells differentiate into all cells composing the tissue / system (Wagers and Weissman 2004); thus, any change to the genome would be inherited by a multitude of daughter cells. Damage passed to the differentiated progeny could exponentially magnify the impact of the genotoxic damage by amplification in differentiated cells. Second, stem cells possess the unique ability to self-renew, i.e., to proliferate and generate more of themselves (Wagers and Weissman 2004). As such, a mutation could propagate within the stem cell compartment itself. In some cases, these mutations may either lead to loss of the already rare stem cells or worse, confer a selective advantage to stem cells that harbor them, leading to domination of mutated clones (Jaiswal et al. 2014). Moreover, the inherent self-renewal capability of stem cells combined with their undifferentiated state places them in functional proximity to cancer cells (Malta et al. 2018). Stem cells are therefore more susceptible to transformation resulting from the accumulation of mutations (Tomasetti and Vogelstein 2015; Jaiswal et al. 2014; Salk, Fox, and Loeb 2010). Finally, the long-term maintenance of tissue homeostasis depends on the ability of somatic stem cells to replace the effector cells within the system in a consistent manner. In that regard, the attrition of stem cell genomic integrity would inevitably result in a loss of tissue homeostasis either in the direction of hyperproliferation e.g. cancer and fibrosis, or hypo-proliferation.

Table 1.

Evidence for age-related accumulation of DNA damage, or reduced DNA-damage response in stem cells

Stem cell	Species	Type of Damage/DDR	Observation	References
Hematopoietic (HSC)	Mouse	γH2AX foci	Accumulate in HSCs	(Rossi et al. 2007; Flach et al. 2014)
		Strand breaks	Accumulate in HSCs	(Beerman et al. 2014)
		53BP1 and p-CHK1	No accumulation	(Flach et al. 2014)
		γH2AX foci and Strand breaks	Accumulate in HSCs (compensated by repair upon stress)	(Moehrle et al. 2015)
		Strand breaks. ROS	Accumulate in HSPCs	(Porto et al. 2015)
		pKap1	Increased in HSC	(Anjos-Afonso et al. 2016)
		γH2AX and 53BP1 foci, Strand breaks Various DDR pathways	Accumulate in HSCs Diminished ATM activity resulting in reduced apoptotic priming	(Gutierrez-Martinez et al. 2018)
	Human	mtDNA mutations	Accumulate in bone marrow CD34+ cells	(Shin et al. 2004)
		mtDNA mutations	Accumulate in mobilized CD34+	(Ogasawara et al. 2005)
		γH2AX foci DSB repair capacity	Accumulate in mobilized CD34+ and bone marrow CD34+CD38- Reduced with age in mobilized CD34+ but not in bone marrow HSPCs	(Rube, Fricke, et al. 2011)
		Microsatellite instability MMR component MLH1	Increases in bone marrow CD34+ Reduced expression in a subset of CD34+ HSPCs	(Kenyon et al. 2012)
		mtDNA mutations	Accumulate with age in mobilized CD34+	(Yao et al. 2013)
		Somatic mutations	Accumulate in circulating Lin-CD34+CD38−	(Welch et al. 2012)
		Somatic mutations	Accumulate in HSCs	(Lee-Six et al. 2018)
		Somatic mutations	Accumulate in bone marrow CD34+CD38−CD45RA−CD90+CD49f+ (HSC)	(Osorio et al. 2018)
Muscle (MuSC)	Mouse	Strand breaks and yH2AX foci	Accumulates with age	(Sinha et al. 2014)
		yH2AX foci	No significant DSB accumulation with age	(Cousin et al. 2013)
		yH2AX foci, Strand breaks	Age-related decline in mitotic catastrophe damage response associated with accumulation of yH2AX foci	(Liu et al. 2018)
		yH2AX foci, Senescence	Aged MuSC are prone to enter cellular senescence with accumulation of yH2AX foci	(Sousa-Victor et al. 2014)
	Human	Mutations	Accumulate with age	(Franco et al. 2018)
Intestinal (ISC)	Drosophila	γH2AX foci, 8-oxo-dG	Accumulates with age	(Park et al. 2012; Na et al. 2013)
		Centrosome amplification	Accumulates with age	(Park et al. 2014)
		ATR dependent DDR	Increased basal and inducible activity	(Park et al. 2015)
		Genomic modifications, mutations, deletions	Accumulate in ISCs, leading to spontaneous neoplasia	(Siudeja et al. 2015)
		Various DDR pathways γH2Av foci, transposable element expression (Gypsy-TRAP)	Age-related decline DDR (correlated with Piwi regulation) Increased with age	(Sousa-Victor et al. 2017)
	Mouse	IR DDR response	Reduced with age	(Martin et al. 1998; Martin, Kirkwood, and Potten 1998)
		DDR gene expression	increased with age	(Moorefield et al. 2017)
	Human	mtDNA mutations	Accumulate with age	(Taylor et al. 2003)
		mtDNA mutations, clonal expansion	Accumulate with age	(Greaves et al. 2014; Greaves et al. 2006)
		Mutations	Accumulate with age	(Blokzijl et al. 2016)
Mesenchymal (MSC) **	Mouse	γH2AX foci	Increased in osteoprogenitors	(Kim, Chang, et al. 2017)
	Human	yH2AX foci, oxidative damage (DCFDA)	Increased with age in dental pulp MSCs	(Feng et al. 2014)
		Strand breaks	Increased in old stromal cells	(Gnani et al. 2019)
		WRN regulated DDR	Reduced WRN expression in dental pulp MSCs from old	(Zhang et al. 2015)
Neural (NSC)	Mouse	Mutation/deletions	Loss of heterozygosity	(Bailey, Maslov, and Pruitt 2004)
		Strand breaks	No accumulation of DNA damage	(Kalamakis et al. 2019)
	Mouse and Rat	Regulation of oxidative stress by Nrf2	Age-related alterations in levels or function of Nrf2	(Ray et al. 2018; Corenblum et al. 2016)
Skin (SSC)	Mouse	Strand breaks and γH2AX foci	Accumulate in HFSCs	(Matsumura et al. 2016)
		yH2AX foci, 8-oxo-dG	Accumulate in epSCs in a circadian dependent manner	(Solanas et al. 2017)
Germ (GSC)	C. elegans	DDR gene expression	Altered with age in oocytes	(Luo et al. 2010)
	Mouse	DDR gene expression	Altered with age in male and female GSCs	(Zhang et al. 2006; Hamatani et al. 2004)
	Rat	DDR gene expression	Altered with age in sperm	(Paul, Nagano, and Robaire 2013, 2011)
	Human	Mutations in progeny	Increase with both parental and maternal age	(Risch et al. 1987; Kong et al. 2012)
		Chromosomal structural aberrations	Increase in sperm with age	(Sloter et al. 2007; Wyrobek et al. 2006; Schmid et al. 2013; Templado et al. 2011)
		Chromosomal trisomy	Increase in oocytes with age	(Oliver et al. 2008; Hassold et al. 1995) (Bugge et al. 1998; Fisher et al. 1995) (Lamb et al. 1997; Lamb et al. 1996)
		Strand breaks	Increase in sperm with age	(Schmid et al. 2007; Alshahrani et al. 2014)
		DDR gene expression	Altered in oocytes	(Steuerwald et al. 2007; Titus et al. 2013)
		γH2AX foci	Accumulates in oocytes with age	(Titus et al. 2013)

^*DDR – DNA damage Response

^**Mesenchymal stromal cells are a heterogeneous cell population, and more refined populations of skeletal stem cells have been recently defined (see text for details)

Understanding the contribution of DNA damage to the aging of tissue-specific stem cells could have crucial implications for our understanding of the aging process and development of aging interventions. In this review, we present an examination of current findings on DNA damage in multiple adult stem cell compartments.

Hematopoietic stem cells (HSCs)

Hematopoietic stem cells are responsible for the production of all blood cells throughout life (Weissman and Shizuru 2008). While remarkably potent (Harrison 1972), HSCs undergo marked age-related changes, including increased numbers, reduced functional potential, and a shift towards intrinsic myeloid bias in both human and mouse models (Beerman et al. 2010; Pang et al. 2011). These age-related changes are also associated with altered epigenetic profiles (Beerman et al. 2013; Sun et al. 2014) and changes in metabolic potential (Ho et al. 2017), and the human HSC frequently presents with an age-associated mutation-driven clonal expansion, often preceding cancer (Shlush et al. 2014; Jaiswal et al. 2014; Welch et al. 2012).

HSCs predominantly reside in the bone marrow, and despite the hematopoietic system’s demanding cell-production burden, due to the high turnover rate of blood cells, the limited HSC population is largely quiescent with few divisions throughout an organism’s lifetime (Wilson et al. 2008; Bernitz et al. 2016). This prolonged dormancy could theoretically protect HSCs from DNA damage by reducing the incidence of replication-based errors, minimizing telomere attrition, and curtailing oxidative damage by lowering metabolic activity (reviewed in Beerman 2017). HSCs are further protected by their hypoxic niche (Suda, Takubo, and Semenza 2011; Parmar et al. 2007; Nombela-Arrieta et al. 2013), hypoxic intracellular state (Takubo et al. 2010), and reliance on glycolytic metabolism (Simsek et al. 2010). Furthermore, the unique ABC transporter profile of HSCs implies greater protection from cytotoxic insults (Tang et al. 2010; Zhou et al. 2002). Despite these buffers, it seems multiple forms of DNA damage accrue with age in HSCs of both mice and humans (Table 2). HSCs are one of the most rigorously studied stem cell compartments, but the assessment of DNA damage accumulation and how it effects the function of aging HSCs is still incomplete. Early studies examining DNA damage accumulation in mice relied on γH2AX foci detection, a proxy of DNA double-strand breaks measuring DNA damage response (DDR) (Table 2). Inclusion of additional assays such as single-cell gel electrophoresis (comet assay), and staining with pKAP1, 53BP1, and p-CHK1 have led to a general consensus that DNA damage accumulates in the aged HSC compartment. Fewer studies have been performed on human hematopoietic stem (and progenitor) cells, but regardless of collection technique (GCSF-mobilized peripheral blood, non-mobilized blood, or bone marrow aspirates), human progenitor cells also consistently display an elevated level of DNA damage in aged individuals (Table 2).

Table 2.

Evidence for DNA damage interventions in stem cells that affect the aging phenotype

Stem cell	Species	Intervention on	Observation	Reference
Hematopoietic (HSC)	Mouse	NER (Ercc1−/−)	Multilineage cytopenia, accelerated fatty replacement of bone marrow	(Prasher et al. 2005)
		Telomere maintenance (mTR−/−) NER (XPDTTD) NHEJ (Ku80−/−)	Accelerated age-related functional decline of HSCs	(Rossi et al. 2007)
		NHEJ (Lig4Y288C/Y288C)	Accelerated loss of erythropoiesis and bone marrow cellularity	(Nijnik et al. 2007)
		DDR checkpoint regulation (ATRs/s)	Pancytopenia, old-like bone marrow cellularity and profile	(Murga et al. 2009)
		NER (Ercc1−/Δ)	Progressive loss of HSPC number and function	(Cho et al. 2013)
		NHEJ (XLF−/−)	Accelerated age-related functional decline of HSCs	(Avagyan et al. 2014)
		Telomere maintenance (mTerc−/− 3rd gen)	Accelerated aging accompanied by Increased DNA damage and γH2AX foci in HSCs	(Wang et al. 2014)
		Telomere maintenance (TERTER/ER)	Various MDS phenotypes	(Colla et al. 2015)
		Replication damage tolerance (PcnaK164R)	Progressive skewing of bone marrow progenitor populations and reduction of HSC function	(Pilzecker et al. 2017)
		MMR (MSH2−/−)	Formed bone marrow with preleukemic HSCs	(Qing and Gerson 2017)
	Human	Irradiation	Reduced reconstitution potential and myeloid bias of HPCs in NOG mice	(Wang et al. 2015)
Muscle (MuSC)	Mouse	Spindle assembly checkpoint regulation (BubR1 hypomorphic mutation)	Accelerated sarcopenia and senescence.	(Baker et al. 2004)
		NHEJ (Xrcc5−/− or Xrcc5+/−)	Accelerated skeletal muscle aging phenotypes (reduced MuSC function and accumulation of γH2AX foci)	(Didier et al. 2012)
		NER (Ercc1−/−)	Accelerated functional decline of progenitor cells	(Lavasani et al. 2012)
		GDF11 supplementation	Rejuvenated functional potential to old murine MuSCs, and reduced DNA damage	(Sinha et al. 2014)
		Bmi1 upregulation	Improved muscle regeneration via oxidative stress resistance with reduction of 8-oxoG	(Di Foggia et al. 2014)
		Telomere maintenance (Tertmdx)	Duchenne muscular dystrophy, activation of DDR leading to impaired differentiation	(Latella et al. 2017)
		NER (Ercc1−/Δ)	Impaired function	(Takayama et al. 2017)
		ROS (Pitx2/3 single and double mutant mice)	Over generation of ROS resulted in DNA damage and premature differentiation or senescence	(L’Honore et al. 2018)
Intestinal (ISC)	Drosophila	Oxidative stress (Paraquat administration)	Mimics several age-related phenomena	(Biteau, Hochmuth, and Jasper 2008)
		CncC or Keap RNAi (NRF2 Oxidative damage response)	Accelerated age-related intestinal degeneration	(Hochmuth et al. 2011)
		DDR checkpoint regulation (ATM/ATR knockdown)	Reduced ISC maintenance	(Park et al. 2015)
		DDR checkpoint regulation (Chk2 KD)	Increased age-related hyperproliferation	(Nagy, Sandor, and Juhasz 2018)
		Multiple DDR pathways (enterocyte specific KO)	ISC hyperproliferation, increased centrosome amplification and DNA damage accumulation	(Park et al. 2018)
	Mouse	Accelerated mtDNA mutations (PolgD257A)	Accelerated ISC aging phenotype	(Fox et al. 2012)
		Irradiation	Reduced ISC regenerative potential but increased IR-resistance	(Booth et al. 2015)
		Telomerase deficient accelerated aging (G3Terc−/−)	Impaired Gadd45a regulated DDR	(Diao et al. 2018)
Mesenchymal Stromal Cells (MSC)	Mouse	DDR checkpoint regulation (ATM−/−)	Accelerated bone aging phenotype due to altered MSC potential	(Hishiya et al. 2005)
		DDR checkpoint regulation (p53 KO, Mdm2 conditional KO)	Altered osteogenic and osteoblast differentiation phenotype	(Wang et al. 2006; Lengner et al. 2006)
		DDR checkpoint regulation (ATM−/−)	Osteoporosis due to osteoblast differentiation deficit but normal progenitor numbers	(Rasheed et al. 2006)
		NER (Ercc1−/−/Ercc1−/Δ)	Accelerated osteoporosis accompanied by progenitor senescence and accumulation of DNA damage	(Chen et al. 2013; Niedernhofer et al. 2006)
		DDR checkpoint regulation (ATRflox/−)	Osteoporosis, possibly due to MSC depletion	(Ruzankina et al. 2007)
		Telomere maintenance (Wrn KO, Terc KO)	Impaired osteoblast differentiation and accelerated loss of bone mass	(Wang, Chen, et al. 2012)
		NER (Ercc2R722W TTD model)	Accelerated age-related decline in MSC osteogenic potential and bone strength	(Nicolaije et al. 2012; Diderich et al. 2012)
		Various DDR pathways (WRN KO)	Recapitulated aging phenotypes upon in vitro expansion	(Zhang et al. 2015)
	Rat	Wnt/β-catenin ROS induction	Induced several aging phenotypes accompanied by γH2AX foci accumulation	(Zhang et al. 2013)
		Sirt1 KO and OE	Sirt1 protected telomere DNA damage and affected the aging phenotype	(Chen et al. 2014)
	Human	Irradiation	Reduced stemness accompanied by ROS (DCFDA) and γH2AX foci accumulation in vitro	(Hou et al. 2013)
		Oxidative stress (H2O2)	Induced cellular senescence in MSCs in vitro	(Borodkina et al. 2014)
		Antioxidants induced DNA damage	Induced cellular senescence in MSCs in vitro	(Kornienko et al. 2019)
		MMR (miR675)	Accelerated MSC transformation in vitro	(Lu et al. 2019)
Neural (NSC)	Mouse	DDR checkpoint regulation (ATM−/−)	Reduced proliferation/survival of NSCs	(Allen et al. 2001; Wong et al. 2003)
		NHEJ (XRCC4 KO, p53 KO)	Predisposed NSPCs to medulloblastoma	(Yan et al. 2006)
		p16INKArf modulation	Altered proliferative profile of NSCs	(Nishino et al. 2008)
		mtDNA mutatagenesis (Polg mutation)	Disrupted NSC quiescence and accelerated aging phenotypes	(Ahlqvist et al. 2012)
		Irradiation	Induced proliferative arrest (senescence) and astrocytic differentiation	(Schneider et al. 2013)
		Hydroxyurea treatment	Induced proliferative arrest (senescence) with inhibited differentiation	(Dong et al. 2014)
		Irradiation (neonatal)	Induced early-onset adult neurodegeneration	(Kempf et al. 2014)
		Irradiation	Induces long-term impairment of NSC proliferation	(Mineyeva et al. 2018)
		Irradiation (in combination with SOD mimetics and p16 KD)	Reduced neurogenesis via expression of p16INK4A, independent of ROS	(Le et al. 2018)
	Mouse and Rat	Various Nrf2 manipulations	Several aging or rejuvenation phenotypes	(Ray et al. 2018; Corenblum et al. 2016)
	Human	Low-dose irradiation	Increased sensitivity of NSCs, elevated apoptosis and cell cycle arrest	(Acharya et al. 2010)
Skin (SSC)	Mouse	DDR checkpoint (ATRflox/−)	Led to hair loss, hair graying, and progressive loss of follicle stem cells	(Ruzankina et al. 2007)
		Irradiation DDR checkpoint regulation (ATM−/−)	Activated DDR that pushed melanocyte stem cells to differentiate Increased sensitization of melanocyte stem cells to differentiate after irradiation stress	(Inomata et al. 2009)
		DDR checkpoint regulation (BRCA1−/−)	Hair follicle specific degeneration with increased γH2AX foci	(Sotiropoulou et al. 2013)
		Mitochondrial DSBs (SystemicIndmito-PstI mice) Various	Hair loss, hair graying, progressive loss of follicle stem cells Various	(Pinto et al. 2017), (Luo and Murphy 2011)
Germs (GSC)	C. elegans Drosophila	I-CreI expression and irradiation	Lok kinase dependent reduced survival and loss of differentiation potential	(Ma et al. 2016)
	Mouse	DSB repair (Brca1, MRE11, Rad51, and ATM oocyte specific KD)	Reduced fertility and oocyte survival	(Titus et al. 2013)
		DSB repair and signaling (ATLD1H/H)	Accelerated oocyte attrition	(Inagaki, Roset, and Petrini 2016)
	Human	Recombination associated DSB (Spo11−/−, Dmc1−/− and Msh5−/−) and DDR checkpoint regulation (ATM−/−)	Reduced oocyte survival	(Di Giacomo et al. 2005)
		DSB repair (Brca1, MRE11, Rad51, and ATM oocyte specific KD)	Reduced fertility and oocyte survival	(Titus et al. 2013)

^*Abbreviations: DDR – DNA damage response; DSB – Double strand break; KO – Knockout; MDS – Myelodysplastic syndrome; MMR – Mismatch repair; NER – Nucleotide excision repair; NHEJ – Non- homologous end joining; OE – Overexpression; ROS – Reactive oxygen species

Early studies examining the role of DNA damage accrual in HSCs used genetic models to ablate key DNA damage repair pathways. Mice with deficient nucleotide excision repair, homologous recombination, and non-homologous recombination pathways showed various aging HSC phenotypes under transplant paradigms (Nijnik et al. 2007; Prasher et al. 2005; Rossi et al. 2007). More recently, these observations have been expanded using other models (Table 3) and other DDR pathways such as replication damage response (Pilzecker et al. 2017) and mismatch repair (Qing and Gerson 2017). Together, the studies demostrate that varied genetic disruptions of DDR result in HSC age-related phenotypes under stress conditions and support the role of DNA damage in HSC aging phenotypes. While genetic manipulations were perfomed in mice, studies of Fanconi anemia disease indicate an analogous relationship in humans between DDR and long-term HSC maintenance (Brosh et al. 2017). Though there is strong support for the accumulation of DNA damage in the aged HSC compartment and evidence that aberrant DDR function leads to decreased functional potential, not all forms of DNA damage can be directly correlated to aging phenotypes. It is important to note that several studies have shown that neither DNA damage nor defects in repair are sufficient to drive aging of this compartment (Pugh et al. 2016; Kaschutnig et al. 2015) and that changes in DDR can be independent of the actual damage accrual (Flach et al. 2014). Several papers have also demonstrated loss of telomere maintenance and mitochondrial DNA damage are not physiological drivers of aging phenotypes (Chen et al. 2015; Samper et al. 2002; Norddahl et al. 2011; Ameur et al. 2011; Allsopp et al. 2003). Furthermore, the aged HSC compartment appears to have the capacity to recognize, respond to, and repair strand breaks once driven into cell cycle (Beerman et al. 2014; Gutierrez-Martinez et al. 2018). Taken together, this suggests that DNA damage is not the sole culprit in HSC aging. That said, forms of DNA damage, such as base mutations, have been directly associated with aging and disease progression (Table 4). By mid-age, humans acquire mutations in HSCs that are detectable at low allele frequencies in the peripheral blood. Some of these mutations confer selective advantages to the stem cells and can lead to clonal expansion defined as age-related clonal hematopoiesis (ARCH) or clonal hematopoiesis of indeterminate potential CHIP (reviewed in Busque et al. 2018). Further, these mutations have been shown to play a role in the age-related acute myeloid leukemia (Reviewed in Short, Rytting, and Cortes 2018), myelodysplastic syndromes (Reviewed in Zhou et al. 2015; Zhou et al. 2013), and atherosclerosis (Fuster et al. 2017).

Table 3.

Disease models associated with both aging and DNA damage in stem cells

Stem cell	Species	Aging/longevity phenotype	Observation	References
Hematopoietic (HSC)	Human	Fanconi anemia	Progressive and accelerated bone marrow failure and AML onset	(Reviewed in Brosh et al. 2017)
Muscle (MuSC)	Mouse	Duchenne muscular dystrophy (Mdx)	Accelerated muscle wasting accompanied by increased 8-oxoG in MuSCs	(Di Foggia et al. 2014)
Intestinal (ISC)	Mouse	ATM and telomerase deficient accelerated aging	Increased apoptosis, p53 activation and anaphase bridges	(Wong et al. 2003)
Mesenchymal (MSC)	Mouse	Werner syndrome and HGPS (Lmna mutant combined with Wrn KO)	Accelerated MSC aging accompanied by accumulation of γH2AX/53BP1 foci	(Wu et al. 2018)
		Trichothiodystrophy (TTD) (Ercc2R722W)	Defective NER	(Diderich et al. 2012)
	Human and mouse	Werner syndrome	WRN regulates MSC heterochromatin stability	(Zhang et al. 2015)
	Human	Werner syndrome	Accelerated in vitro MSC cellular senescence caused by telomerase deficiency	(Cheung et al. 2014)
		Hutchinson-Gilford progeria syndrome (HGPS)	MSCs differentiated in vitro from HGPS derived iPSCs display increased γH2AX foci and reduced stress fidelity	(Zhang et al. 2011)
Skin (SSC)	Mouse	Accelerated skin aging by Tap63 KO	Accelerated accumulation of γH2AX foci in the hair follicle sheath	(Su et al. 2009)
	Human	Photoaging (UV damage)	Reduction in keratinocyte progenitors in photoaged skin	(Kwon et al. 2008)
		Xeroderma Pigmentosum	Genetic correction of epidermal stem cells is beneficial	(Rouanet et al. 2013)

Table 4.

Evidence for the involvement of stem cell DNA damage in age-related diseases

Stem cell	Species	Age-related disease	Observation	References
Hematopoietic (HSC)	Mouse	Atherosclerosis	Lipoprotein receptor–deficient (Ldlr−/−) mice chimeric with Tet2 mutant HSCs displayed clonal expansion and increased atherosclerosis	(Fuster et al. 2017)
	Human and mouse	MDS	Multiple	(Reviewed in Zhou et al. 2015; Zhou et al. 2013)
		AML	Multiple	(Reviewed in Short, Rytting, and Cortes 2018)
Intestinal (ISC)	Drosophila	Cancer	Age-related somatic genomic modifications in ISCs lead to spontaneous neoplasia	(Siudeja et al. 2015)
	Mouse	Colorectal Cancer	ISCs are sensitive to alkylation-induced DNA damage, leading to hyperproliferation and transformation	(Kim et al. 2016)
			Bcl-2 protects ISCs from IR-induced apoptosis and is also involved in adenoma development	(van der Heijden et al. 2016)
			Loss of tumor-suppressor in ISCs leads to cancer via DNA damage accumulation	(Myant et al. 2017)
	Various	Colorectal cancer	Various	Reviewed in (Testa, Pelosi, and Castelli 2018)
Skin (SSC)	Mouse	Basal cell carcinoma	Hedgehog signaling induced carcinoma only acted via stem cells	(Sanchez-Danes et al. 2016)
		Squamous and basal cell carcinoma	Various	Reviewed in (Sotiropoulou and Blanpain 2012)
Neuronal (NSC)	Mouse	Medulloblastoma	NHEJ deficiency in NSPCs predisposes to cancer	(Yan et al. 2006)
		Various brain cancer types	Combination of genomic deletions caused brain tumors from NSCs but not astrocytes	(Jacques et al. 2010)
		Brain cancer	Recurrent DSB sites in NSCs promote oncogenic translocations	(Schwer et al. 2016)
		Neuroblastoma	Recurrent DSB sites in NSCs occur near tumor-suppressor genes	(Wei et al. 2016), (Wei et al. 2018)
	Human	Glioblastoma	Constitutive DDR allows glioblastoma stem cells to resist IR	(Carruthers et al. 2018)
		Medulloblastoma-conditioned media	Increases sensitivity to DNA damage and oxidative stress in NSCs	(Sharma et al. 2016)
Mesenchymal (MSC)	Human	Malignant transformation	Accelerated by miR675 regulated MMR dysfunction	(Lu et al. 2019)
Germ (GSC)	Human	Testicular spermatocytic seminoma	Multiple	(Romano et al. 2016)

^*AML – Acute myeloid leukemia; MDS – Myelodysplastic syndrome

Satellite Cells / Muscle stem cells

Muscle stem cells, most commonly referred to as satellite cells, reside in the skeletal muscle tissue below the lamina surrounding the muscle fibers (Sacco et al. 2008). These largely quiescent adult stem cells facilitate the regeneration of skeletal muscle through differentiation into multinucleated myofibers. Loss of muscle mass with age correlates with an overall decline in function, potential, and numbers of satellite cells (Table 1). Epigenetic alterations, including histone modifications and DNA methylation, have been associated with satellite cell aging (Bigot et al. 2015; Liu et al. 2013; Schworer et al. 2016) and may also contribute to their age-associated dysfunction.

One of the first reported examinations of age-related decline of satellite cells did not see an accumulation of strand breaks (Cousin et al. 2013). To the contrary, more recent evidence has shown increased levels of DNA damage in satellite cells isolated from aged mice (Sinha et al. 2014; Liu et al. 2018) (Table 2). Similarly, in humans, functional impairment of old satellite cells was associated with loss of genomic stability (Franco et al. 2018). Additionally, genetic manipulation of key DDR genes leads to various muscle stem cell defects that mirror aging phenotypes (Table 3). For example, compromising nucleotide excision repair by manipulation of ERCC1 resulted in accelerated satellite cell functional decline (Lavasani et al. 2012; Takayama et al. 2017). Similarly, interference with the non-homologous end-joining process by XRCC manipulation resulted in acceleration of muscle aging and reduced satellite cell function (Didier et al. 2012).

Interestingly, geriatric mice, over 28 months old, exhibit senescent satellite cells (Sousa-Victor et al. 2014); as senescence is often coupled with DNA damage, this further provides indirect support for DNA damage accrual in this compartment. This transition to senescence (presumably after accumulation of DNA damage) is in contrast to HSCs which have been shown to differentiate (Wang, Sun, et al. 2012), attempt to repair the accrued damage (Beerman et al. 2014), or undergo apoptosis (Gutierrez-Martinez et al. 2018) but interestingly provided a therapeutic target. After clearance of senescent cells, through induced p16^INK4a-dependent apoptosis (Baker et al. 2011) or restored autophagy (Garcia-Prat et al. 2016) the aged satellite cell population had significantly improved function. In addition, treatment with the antioxidant Trolox, which again may act by mitigating DNA damage, was also demonstrated to improve aged satellite cell function by attenuating a transition from quiescence to senescence (Sousa-Victor et al. 2014). Heterochronic parabiosis studies, where mice of different ages share blood circulation, also demonstrated improved function of the aged muscle (Conboy et al. 2005; Sinha et al. 2014) and one study examined DNA damage after heterochronic parabiosis and saw decreased pH2Ax-foci in satellite cells from aged mice exposed to the young systemic environment (Sinha et al. 2014). Thus, muscle stem cells appear to be amenable to interventions, and indeed many aging phenotypes have been reversed or mitigated in experimental settings; highlighting the potential restoration of function in aged satellite cells, associated with the clearance of DNA damage.

Intestinal stem cells (ISCs)

Intestinal stem cells give rise to every cell in the intestinal epithelium, a tissue with an exceptionally high turn-over (Hu et al. 2013), similar to the hematopoietic system. ISCs reside within intestinal crypts (Gervais and Bardin 2017) and their location and cell-cycle status appear directly correlated with their function. The active population of ISCs is typically referred to as crypt base columnar (CBC) cells reside at the base of the crypt and replicate to maintain homeostasis of the rapidly replaced cells of the intestinal epithelium. A second population of ISCs, located at the +4 crypt position, remain largely quiescent (Tian et al. 2011; Yan et al. 2012) and become active under stress conditions (Takeda et al. 2011; Tian et al. 2011). A large fraction of the published research on ISCs is generated using Drosophila melanogaster (Tables 2–5); however, recent studies in the murine system have found species-specific aging phenotypes. For instance, studies in Drosophila show an increase in ISC numbers with age (Martin et al. 1998; Biteau, Hochmuth, and Jasper 2008), but aging in mouse models is associated with maintained (Nalapareddy et al. 2017) or decreased ISC numbers (Mihaylova et al. 2018; Igarashi et al. 2019). However, in both aging is associated with a functional decline (Biteau, Hochmuth, and Jasper 2008; Choi et al. 2018; Cui et al. 2019; Igarashi et al. 2019; Mihaylova et al. 2018; Nalapareddy et al. 2017).

Table 5.

Evidence for longevity phenotypes or interventions that influence DDR or DNA integrity in stem cells

Stem cell	Species	Aging/longevity phenotype	Observation	References
Intestinal (ISC)	Drosophila	Metformin longevity model	Improved aged ISC function with reduced DNA damage	(Na et al. 2013)
	Mouse	Caloric Restriction	Increased DNA damage protection in ISCs	(Yousefi et al. 2018)
Mesenchymal (MSC)	Rat	In vitro treatment with old rat serum	Accelerated in vitro accumulation of γH2AX foci	(Zhang, Wang, and Tan 2011)
Neuronal (NSC)	Mouse	CR	Reduces inflammation in SVZ and neurogenesis decline	(Apple et al. 2019)
	Rat	High-intensity exercise	Improves adult neurogenesis	(Nokia et al. 2016)
Germ (GSC)	Mouse	CR	Prevention of age-related aneuploidy	(Selesniemi et al. 2011)

ISCs reside within a tissue that is continuously exposed to chemical threats which naturally puts them at risk for DNA insults. In addition, the constitutively mitotically active subpopulation of ISCs is under constant replicative stress which could be aggravated by putative exposure to external stressors. Examination of both the mitochondrial and nuclear genome demonstrate evidence of age-associated DNA damage accumulation (Table 2). The DDR of ISCs may also undergo a decrease in potency (Sousa-Victor et al. 2017; Martin et al. 1998; Martin, Kirkwood, and Potten 1998), while displaying increased basal activity (Moorefield et al. 2017). These observations are likely most relevant to the quiescent ISC subpopulation (Yan et al. 2012), as active ISCs respond to most non-physiological levels of genotoxicity with cell death (Otsuka et al. 2018). This is similar to the increased propensity of more committed hematopoietic progenitor cells to undergo apoptosis as compared to HSCs (Guiterrez-Martinez et al 2018). This suggests maintenance of the quiescent stem cell population is vital for the long-term maintenance of the gut epithelium. Models inducing DNA damage in ISCs has been shown to mimic several aging phenotypes (Table 3), including the known age-related colorectal cancer (Table 5). Interestingly, ISCs during the aging process, like HSCs, seem to be at a risk to undergo clonal expansion (Lopez-Garcia et al. 2010) which has been demonstrated to be possible following K-Ras mutagenesis and could predispose for cancer (Snippert et al. 2014; Gierut et al. 2015).

Mesenchymal stromal cells (MSCs)

Mesenchymal stromal cells are a heterogenous population of stem, and likely, progenitor cells which can give rise to bones, muscle, fat, cartilage, fibroblasts, endothelial cells, and stroma (Chan et al. 2018) often referred to as mesenchymal stem cells. Recently, a more defined population of skeletal stem cells (SkSC) that can generate bone, cartilage and stromal tissue have been characterized in both mice (Chan et al. 2015) and humans (Chan et al. 2018); however, limited information is available about the DNA damage on this refined population, so we will discuss data generated on the more heterogeneous MSC compartment.

MSCs undergo reported phenotypical changes with age but some of these may be complicated given the heterogeneity of the populations, leading to alternative conclusions on aging phenotypes (Table 1) (Sethe, Scutt, and Stolzing 2006), and most functional analysis is performed ex vivo. Examination of DDR after exogenous damage induction was more robust in early-passage MSCs compared to late-passage MSCs (Wu et al. 2017) suggesting an “aging” defect in damage response. Increased passaging of MSCs leads to accumulation of DNA damage and mutation accrual and ultimately senescence – suggesting replication stress significantly contributes to these defects (Kim, Rhee, et al. 2017; Alves et al. 2010; Hare et al. 2016; Yu et al. 2018; Galderisi et al. 2009). Recently, direct evidence of DNA damage accumulation has also been shown in aged human MSCs, where strand break accrual was accompanied by elevated ROS levels - but not increased 8- oxo- dG staining (Gnani et al. 2019). This increased damage was also associated with a decline of MSC potential (measured by colony forming assays) and decreased osteogenic lineage differentiation (Gnani et al. 2019).

Manipulation of DNA damage repair or response pathways in MSCs has frequently resulted in accelerated aging phenotypes (Table 3), mostly through demonstration of osteoporosis and bone density loss in mice. However, a direct link between DNA damage in MSCs and any human age-related disease has yet to be reported (Table 5). This is surprising considering that intrinsic MSC dysfunction has been linked to osteoporosis (Pino, Rosen, and Rodriguez 2012), chronic inflammation (Miranda et al. 2019), arthritis (Barcia et al. 2015), and osteosarcoma (Walkley et al. 2008). Still, these data may be confounded by the heterogeneous nature of these analyzed populations, and novel insight into DNA damage accrual of SkSC and evaluation of their contribution to these aging phenotypes will likely be forthcoming.

Neuronal stem cells (NSCs)

Neuronal stem cells give rise to all major cell types in the adult brain including neurons, astrocytes, oligodendrocytes, and glial cells (Gage and Temple 2013). NSCs reside in two distinct niches: the sub-ventricular zone of the lateral ventricles and the sub-granular zone of the hippocampus (Gage and Temple 2013). They contribute to the maintenance of neural tissue throughout adulthood and also provide plasticity for the generation of new structures required for learning and memory. In rodents, aging is accompanied by decreased neurogenesis by resident stem and progenitor cells (Kuhn, Dickinson-Anson, and Gage 1996). In mice, there is also a sharp age-related decrease in the total number of NSCs in the ventricular-subventricular zone (Kalamakis et al. 2019; Shook et al. 2012; Ahlenius et al. 2009). This significant decline is already apparent in mid-aged brains 7–12 months with only a modest further reduction occurring over the following year (Shook et al. 2012; Kalamakis et al. 2019). A significant loss of quiescent hippocampal NSCs is also seen between one- and 24-month-old mice (Encinas et al. 2011).

Data on age-associated DNA damage accumulation in NSCs is sparse and at times contradictory. A recent study has found no significant age-associated accumulation of DNA damage in either active or quiescent NSCs, as detected by the comet assay. Further, similar functional potency was observed in NSCs isolated from young and aged mice (Kalamakis et al. 2019). However, loss of chromosomal heterozygosity (suggestive of DNA damage) was observed in almost all neurospheres derived from aged neural stem and progenitor cells (NSPCs) generated from C57Bl/6 × DBA/2 hybrid mice (Bailey, Maslov, and Pruitt 2004). Further, seminal work assaying chromosomal architecture in cycling NSPCs, demonstrated that replication stress introduces recurrent DSB clusters in key genes regulating brain functions (Wei et al. 2016). These fragile sites are hot spots for chromosomal rearrangement events and may serve as a mechanism contributing to the somatic diversity of brain as described by McConnell et al. (2013). Thus, NSCs must balance the generation of genetic diversity contributing to short-term memory and learning (Wei et al. 2016) against maintaining the genomic fidelity of the NSC pool. It has been suggested that a subset of normally quiescent NSCs permanently exit the stem cell pool once they become activated (Encinas et al. 2011) in order to protect against a long term accumulation of DNA damage in this compartment.

Models that introduce DNA damage in the brain reproduce several aging phenotypes (Table 3), largely recapitulating aberrant differentiation potential. In young NSCs, irradiation leads to robust induction of DDR, correlated with loss of stem cell markers and differentiation towards astrocytes (Schneider et al. 2013). Another model of damage induction utilizing Polg disruption, leads to mtDNA mutagenesis and also loss of NSC self-renewal (Ahlqvist et al. 2012). While these studies again support the role of differentiation to eliminate the potentially damaged NSCs from the stem cell pool (similar to HSCs and ISCs), the majority of these studies were only performed on young NSCs and are analyses of significant extrinsic perturbations of the DNA integrity. It is yet to be fully established how more endogenous stimuli and physiologic aging may affect the fidelity of DNA and the functional potential of NSCs.

While reduced functional potential of aged NSCs could be a direct driver of dementia and Parkinson’s, there is a little evidence linking DNA damage in NSCs to common age-related diseases (Table 4). For instance, low-dose irradiation induction of an Alzheimer’s-like phenotype in mice did not act through NSCs (Kempf et al. 2016) and NSPCs are less prone to apoptosis than their progeny (Kashiwagi et al. 2018), similar to aforementioned stem cells. However, murine models may not fully recapitulate Alzheimer’s phenotypes, and the possible connection between damage and disease remain to be comprehensively examined.

Skin stem cells (SSCs)

Skin stem cells also often refer to a heterogenous population of cells residing in various locations in the skin including the hair follicles, interfollicular space, and sebaceous glands (Blanpain and Fuchs 2006). SSCs generate multiple lineages of skin cells during both regular maintenance and regeneration following an insult (such as a skin wound) (Fuchs 2008; Mascre et al. 2012). These cells are typically classified as either epidermal stem cells (EpSC) hair follicle stem cells (HFSC), or melanocyte stem cells (MeSC). During aging, SSCs undergo functional decline that varies based on subtype and exposure to external factors (Table 1). In the absence of environmental insults, EpSC numbers are retained with age, although a functional decline still occurs (Giangreco et al. 2008; Keyes et al. 2013). On the other hand, examination of HFSC antibodies in C57BL/6 mice showed an age-associated reduction in signal, with dysregulation of the stem cell signatures at regions of hair thinning and loss (Matsumura et al. 2016).

SSCs are responsible for the maintenance of a high turnover tissue but are mostly dormant (Tumbar et al. 2004), and potentially derive similar protections from DNA damage during quiescence as HSCs. However, the proximity of the skin to the atmosphere may subject SSCs to external insults that threaten DNA integrity, including high oxygen intake (Stucker et al. 2002), air pollutants, and UV radiation (Panich et al. 2016). The last has been shown to directly damage EpSCs in vivo at physiological doses (Ruetze et al. 2011). As an additional layer of protection against this exposure, EpSCs have a circadian-regulated timing of DNA replication, with proliferation occurring during the night (Janich et al. 2011). With aging, this tight regulation of S-phase entry is lost, and the EpSCs continue to cycle during the day when DNA is potentially exposed to UV and oxidative damage. Accordingly, aged EpSCs display high levels of 8-OHdG and γH2AX staining throughout the day, indicating elevated DNA damage (Solanas et al. 2017).

Perhaps the most intuitive link between SSC aging and DNA damage is hair graying and hair loss, both known aging phenotypes. HFSCs in the aged skin display elevated levels of damage, indicated by 53BP1, γH2AX-foci, and the comet assay; this damage accumulation is more pronounced in regions of hair graying and loss (Matsumura et al. 2016). Additionally, MeSC depletion, due to excessive differentiation driven by DNA damage, induces hair graying (Inomata et al. 2009). Hair thinning is also a common feature of progeroid syndromes such as Werner’s syndrome and Ataxia-telangiectasia (Hasty et al. 2003), both of which are driven by genomic instability again drawing connections between DNA damage and aging phenotypes. Similar loss of homeostatic function was observed after induction of DNA damage by irradiation or when using the XPD^TTD/TTD progeroid mouse model (Matsumura et al. 2016). Notably, the age-associated accumulation of strand breaks in HFSCs was reported to not exclusively colocalize with telomeric DNA, suggesting the accumulation of damage foci is independent of telomere shortening (Schuler and Rube 2013).

Overall, these lines of evidence suggest DNA damage accumulation in aged skin stem cells and raises the question of how these stem cells address this burden. HFSCs, as many of the tissue specific stem cells, are more resistant to DNA damage induced apoptosis compared to more differentiated cells (Solanas et al. 2017; Gutierrez-Martinez et al. 2018). Further, both MeSCs and HFSCs commit to differentiation after significant genotoxic stress, (Inomata et al. 2009; Matsumura et al. 2016), a strategy also used by aged HSCs (Wang, Sun, et al. 2012; Wingert et al. 2016). These mechanisms could serve as a fail-safe to eliminate damaged stem cells from the niche to minimize cancer potential.

Germline stem cells (GSCs)

Germline stem cells are, by definition, not somatic or responsible for adult tissue homeostasis. However, they reside within aging organisms and may be affected by aging processes, such as DNA damage accumulation. The major downstream impact of genomic damage and mutagenesis in GSCs would manifest in the progeny or as infertility with direct consequences to the germline, but there are also instances in which accumulation of damage could affect the host soma. Unlike other stem cells, GSCs differ noticeably between the sexes. In humans, oocyte generation ceases after birth while spermatogenesis involves constant replication throughout life, greatly increasing the risk of mutagenesis due to replication errors and necessitating powerful genome maintenance strategies (Goriely and Wilkie 2012). Accordingly, spermatogenial stem cells also utilize unique DNA damage responses (Rube, Zhang, et al. 2011; Ishii et al. 2014; Ahmed et al. 2007), including constitutively active telomerase (Pech et al. 2015), similar to somatic tissue specific stem cells. In fact, telomerase is so highly active that their telomeres have been reported to elongate with age (Antunes et al. 2015). Despite such protective mechanisms, male reproductive potential declines with age (Paul, Nagano, and Robaire 2013) correlating with a decline in GSC number and function (Paul, Nagano, and Robaire 2013, 2011; Zhang et al. 2006). Notably, a study demonstrated that in mice the intrinsic spermatogenesis cellular potential exceeds the natural mouse lifespan (Ryu et al. 2006), indicating a strong influence of the niche in the aging phenotype. Despite this potential, there is evidence spermatogenial stem cells also accumulate DNA damage with age (Sloter et al. 2007; Schmid et al. 2013; Templado et al. 2011) and display altered DDR gene expression profiles (Paul, Nagano, and Robaire 2013, 2011; Zhang et al. 2006). These aberrations have been linked to both infertility (Alshahrani et al. 2014; Wyrobek et al. 2006) and mutagenesis that predisposes progeny to various types of illness (Kong et al. 2012; Goriely and Wilkie 2012). These mutations could also lead to a selective advantage of specific clones, leading to clonal expansion of spermatogenial cells (Goriely et al. 2005; Goriely et al. 2003). Interestingly, testicular germ cell tumors originate from germ cells and in the case of spermatocytic seminoma is a disease of late life (Bahrami, Ro, and Ayala 2007). This cancer, which is generally associated with deficiency in the DDR (Romano et al. 2016), exemplifies how a non-somatic stem cell can drive an aging phenotype in the adult.

Like male GSCs, it seems oocytes of many species also decline in function with age (Ge et al. 2015; Luo and Murphy 2011; Zhao et al. 2008). In humans, oocytes cease to proliferate after birth. On one hand, this protects them from replication errors; on the other, they face a unique challenge to maintain genome integrity and function for a long time in a cell arrested at meiotic prophase I (Luo and Murphy 2011). Possibly to provide the females with the optimal oocytes, a block in DNA damage repair signaling by RNF12 and HORMAD shunt damaged cells towards apoptosis opposed to repair implicating a DNA quality control mechanism that drives the dramatic reduction of total number of oocytes seen in development (Qiao et al. 2018). In the adult, quiescent oocytes display active DNA repair with highly tuned quality control mechanisms to eliminate cells when repair fails (Qiao et al. 2018; Rinaldi et al. 2017; Stringer et al. 2018; Tuppi et al. 2018). Despite this, several genes for DNA maintenance have been shown to be mis-regulated in aging human and mouse oocytes (Hamatani et al. 2004; Steuerwald et al. 2007; Titus et al. 2013). Accordingly, aged oocytes have been shown to accumulate gH2Ax-foci with age, accompanied by a decline in BRCA1, Mre11, Rad51, and ATM (Titus et al. 2013), indicating that DSB repair is especially strained. DNA damage in oocytes has been shown to directly affect their functional decline in flies (Ma et al. 2016) and the link in humans is also quite clear as aneuploidy in the embryo almost always stems from the oocyte (Nagaoka, Hassold, and Hunt 2012); with these errors consistently increasing with age (Nagaoka, Hassold, and Hunt 2012). A fascinating aspect in this field are the dynamics with which different abnormalities accumulate with age: for example, the frequency of trisomy 16 rises sharply immediately following sexual maturity (Hassold et al. 1995) in a fairly linear fashion; whereas the frequency of trisomy 21 increases slowly at first but surges later in life (Nagaoka, Hassold, and Hunt 2012). Taken together, these patterns indicate a combination of specific types of failure- such as loss of chromosome cohesion (Hunt and Hassold 2010) as well as a strong influence of the niche; they also clearly demonstrate that the aging of this system begins long before menopause and is not solely driven by oocyte depletion. This fact also opens the door for the use of model organisms where oocytes are continuously generated to study the aging of this compartment, such as worms and flies (reviewed in Luo and Murphy 2011). In mice, it has been reported that a 40% caloric restriction diet attenuates the age-related oocyte aneuploidy (Selesniemi et al. 2011). Interestingly, reduced oocyte function with age has been proposed to be driven by telomere attrition, despite their absolute lack of replicative stress (Keefe, Marquard, and Liu 2006).

Discussion

Overall, evidence suggests tissue specific stem cells accrue DNA damage during aging despite being considered protected populations (Table 2). This accumulation differs between the various stem cell compartments, and divergence could derive from stem cells’ inherent properties (e.g. DNA methylation profile, potency etc.), their functional dynamics (e.g. quiescent vs. highly proliferative), and the niche in which they reside. The complexity of how these factors interact is well exemplified in quiescent HSCs, in which decreased cycling mitigates replicative associated damage, but consequently allows for accumulation of other types of damage, as DNA damage response pathways are attenuated until cell cycle re-entry (Beerman et al. 2013). A similar concept is exemplified in the sex difference in human germ cells; both have highly tuned defense mechanisms yet accumulate DNA damage with age, but with very distinct types depending on their extremely different proliferative profile.

Despite apparent accumulation of DNA damage with age, the evidence is incomplete to definitely state damage accrual drives the aging of stem cells or the tissues they support. While disturbance of genomic integrity can mimic some stem cell aging phenotypes (Tables 3 and 4), it has yet to be demonstrated that alleviation of the damage burden results in overall extension of health or lifespan. Similarly, there is little evidence showing increased stem cell genomic fidelity in models of longevity, such as centenarians (Table 5). While DNA damage in stem cells has been repeatedly linked to age-related neoplasias (Table 5), no study so far has shown a similar connection to any other age-related diseases with the exception of HSCs in atherosclerosis (Fuster et al. 2017). Of course, some aging processes of the somatic stem cell compartment could be entirely independent of DNA damage; such as alterations in the epigenetic profile (Beerman and Rossi 2015) or proteolytic regulation (Leeman et al. 2018), or potentially combinations of these contributions- such as mutation accrual leading to loss of DNA methylation and vice versa (Xia, Han, and Zhao 2012).

Regardless of whether DNA damage in tissue-specific stem cells is a driver or passenger of aging processes, it appears some damage is inevitable. While we may be able to minimize some insults (e.g. UV exposure), other drivers of damage could be nearly impossible to prevent (e.g. replicative stress). In addition, considering the complex and acutely regulated endogenous DNA damage repair and response network, introducing exogenous improvements to this exquisitely fine-tuned pathway is a daunting task. Thus, any intervention that aims to mitigate age-related diseases driven by DNA damage in stem cells, or to serve as a rejuvenation therapy, would likely need to be applied with the underlying assumption that damage has occurred. It could perhaps be more useful to focus on targeting the downstream effects of this damage on both the stem cell population and its progeny, rather than its prevention. This would require an understanding of the tissue specific implications of said damage. For example, DNA damage in the HSC compartment leads to mutations which can lead to clonal expansion of cells bearing mutations that give them a selective advantage. Identifying and characterizing different clone-types would serve crucial in diagnosis of individuals predisposed to hematologic cancers. Age-related changes in the HSC pool could also alter the trajectory of any potential intervention. For one, undesired clonality should be avoided in a transplantation setting. On the other hand, identification of “bad” clones opens the door to investigate how to eliminate them distinctively while preserving the “good” clones”. However, we have yet to fully understand the classification between “good” and “bad” clones. A similar understanding of parallel phenomenon in other stem cell compartments would be of equal importance for any future medical treatment or intervention aimed to alleviate aging phenotypes or age-related diseases requiring a solid understanding of stem cell type-specific aging phenotypes. Thus, there are still several avenues of research to pursue to fully characterize age-associated loss of stem cell potential; ideally providing more insight into targets for improvement of overall health of each tissue/system through maintenance of stem cell function.

Fig. 1.

Classification and aging phenotypes of adult somatic stem cells.

* negative for mouse

** See text for references