Proliferation: Driver of HSC aging phenotypes?

Abstract

The decline of stem cell performance with age is a potential paramount mechanism of aging. Hematopoietic stem cells (HSCs) are perhaps the most studied and best characterized tissue-specific somatic stem cells. As such, HSCs offer an excellent research model of how aging affects stem cell performance, and vice versa. Studies from recent years have elucidated major aging phenotypes of HSCs including a decline in reconstitution potential, altered differentiation predisposition, an increase in number, accumulation of DNA damage/mutations and several others. However, what drives these changes, and exactly how they translate to pathology is poorly understood. Recent studies point to proliferative stress of HSCs as a potential driver of their aging and the resulting pathologies. Here we discuss the recent discoveries and suggest the context in which aging phenotypes could be driven, and the relevant mechanisms by which HSCs could be affected.

1. Hematopoiesis demands and stem cell potential

Blood is a tissue with high cellular turnover, with estimates of ~5 × 10¹¹ new cells created per day in an adult human, necessitating immense production during homeostasis and even more so during times of hematologic or immunologic stress (Alberts et al., 2015). This necessitates the bone marrow, as the main hematopoietic tissue, to possess massive cellular production potential (Scheiermann et al., 2015). Within the bone marrow, hematopoietic stem cells (HSCs) are largely responsible for lifelong maintenance of balanced hematopoiesis (Weissman and Shizuru, 2008). However, the proliferation and differentiation of HSCs into mature blood cells is a multi-step process involving intermediaries of increasingly restricted differentiation potential such as multipotent progenitors, oligopotent progenitors (e.g. common lymphoid or myeloid progenitors), and lineage restricted cells (e.g. Pro-B, Pro-T, MkP etc.) (Sawai et al., 2016). Recent studies reveal that these intermediaries bear the brunt of the proliferative load of native hematopoiesis (Busch et al., 2015; Gentek et al., 2018; Gomez Perdiguero et al., 2015; Rodriguez--Fraticelli et al., 2018) and in transplant settings (Boyer et al., 2019). As a result, HSCs under native conditions proliferate rarely (Bernitz et al., 2016; Foudi et al., 2009; Qiu et al., 2014; Rodriguez-Fraticelli et al., 2018; Wilson et al., 2008). For example, recent estimates suggest human HSCs divide every 2–20 months (Lee-Six et al., 2018). This tight regulation of a predominantly quiescent state indicates that excessive proliferations could have negative ramifications, should they occur. While HSCs rarely proliferate, they are the only cells capable of full, long-term restoration of the hematopoietic system after severe myeloablative stress, such as lethal total body irradiation. This regenerative process requires a multitude of proliferations and can be derived from even a single HSC, indicating that these stem cells possess immense proliferative potential under stress conditions (Alberts et al., 2015; Benveniste et al., 2003; Harrison, 1972; Krause et al., 2001; Rodriguez-Fraticelli et al., 2020; Yamamoto et al., 2013). However, that potential does seem to be finite (Harrison, 1979). Importantly, a similar replicative stress, such as one inflicted on HSCs during a stem cell transplant, could also be relevant outside of clinical or experimental settings. Animals in an unprotected environment undergo blood and immune system stresses of varying degrees throughout their life that could push HSCs to proliferate, even beyond what would be ideal during “normal” homeostasis (Batsivari et al., 2020). As such, the question of whether and if so, how, replicative stress affects hematopoietic stem cells and their aging phenotype is relevant both to our understanding of aging and to the implications of stem cell transplants in the clinic. Here we briefly review the potential consequences of excessive HSC proliferations on aging phenotypes.

2. DNA damage

The relationship between DNA damage and proliferation can be regarded as a double-edged sword in the stem cell compartment. On one side, the cell cycle can act as a negative influence on genomic integrity, where proliferation can induce various forms of DNA damage (Flach et al., 2014; Halazonetis et al., 2008; Walter et al., 2015); on the other side, the bulk of DNA response and repair is directly associated with the process of DNA replication (Barr et al., 2017; Beerman et al., 2014). Importantly, proliferation of a stem cell with accumulated damage or mutations will result in exponential propagation of the accrued damage: perpetuated by maintenance of the damage/mutations during self-renewal and also through inheritance of these aberrations in progeny after HSC differentiation.

In HSCs, it seems that DNA damage is an aging mechanism (McNeely et al., 2019). This statement is largely derived from: (i) age-related diseases that originate from DNA damage such as acute myeloid leukemia (AML); (ii) evidence that disruption of DNA damage response results in aging HSC phenotypes; and (iii) DNA damage and mutations accumulate with age in HSCs (McNeely et al., 2019). However, the relative contribution of DNA damage to HSC aging is not fully clear (Ameur et al., 2011; Norddahl et al., 2011). Interestingly, strand breaks primarily accumulate in quiescent HSCs, which are capable of repairing this damage when they enter the cell cycle (Beerman et al., 2014; Gutierrez-Martinez et al., 2018). Conversely, excessive stress induced proliferations have been shown to drive DNA damage accumulation in HSCs (Walter et al., 2015). The latter is of special relevance to clinical HSC and bone marrow transplants, where the cells are forced to undergo numerous proliferations under a “stress setting” which could result in the loss of fidelity of the reconstituted hematopoietic system.

Overall, it seems that HSCs hold a fine balance of tight cell cycle regulation to maintain their genome integrity, whereby either extreme periods of dormancy (e.g. during “native” hematopoiesis) or excessive proliferations (e.g. during low cell number transplants) could result in accumulation of DNA damage. A better understanding of the balance between these extremes would be useful to understand the pathogenesis of age-related diseases such as AML and to improve our ability to provide long-term stable stem cell grafts.

3. Population dynamics and clonal expansion

One of the most fascinating and perhaps impactful consequences of DNA damage in HSCs directly relates to their regulation of proliferation. When specific mutations occur, they can provide a distinct selective advantage to the HSCs that harbor them, resulting in expansion of specific clones (Fig. 1). This clonal expansion can occur independently of clinical blood disorders (Cooper and Young, 2017) but is considered a predisposition for them, including associations to hematologic cancers (Genovese et al., 2014; Jaiswal et al., 2014; Shlush et al., 2014; Welch et al., 2012), atherosclerosis (Jaiswal et al., 2017), and all-cause mortality (Jaiswal et al., 2014). Clonal expansion is surprisingly common during hematopoietic aging, and can in fact be defined as a blood aging phenotype (Jaiswal et al., 2014; McKerrell et al., 2015; Verovskaya et al., 2013), indicating that selective advantage among HSCs is an inevitable process of aging. Since this selection must operate via self-renewal of specific clones, and since HSCs natively prefer dormancy (see above), it stands to reason that ascendant clones have endured some measure of proliferative stress in order to achieve dominance (Kirschner et al., 2017). This stress could in turn be compounded by loss of cell-cycle checkpoint fidelity, granting HSCs another advantage, but also resulting in expansion of cells that are less equipped to handle the very proliferative stress that gave rise to them (Fig. 1).

Fig. 1.

Hematopoietic stem cells’ proliferation potential and numbers with aging. While hematopoietic stem cells (HSCs) increase in number with age (lower panel), they concomitantly display a reduced potential (upper panel) and accumulate adverse effects that can be driven by over-proliferation. The process is accompanied by an expansion of mutated clones with increased proliferative potential (in red).

Besides the implications for aging, this is also extremely relevant to stem cell transplants. Especially in transplant situations starting with low donor stem cell numbers, as the transplanted stem cells would need many cellular proliferations to effectively reconstitute the hematopoietic system. As such, our understanding of the clonal selection processes during a transplant, and of the differential proliferative stress on different clones, could greatly impact our clinical capacities. The same would also be true for any future potential HSC rejuvenation treatment: these treatment paradigms would need to consider the ramifications of inadvertently selecting for undesirable clones.

4. Potential for Perpetuation of Proliferative Stress

For proliferative stress to influence HSC aging phenotypes, its impact must be perpetuated over time and carried over to the progeny. In theory, the transfer of “proliferative scars” is less likely to be mediated by cellular protein or membrane composition, as both of these components would ultimately get diluted during multiple proliferations (unless the dilution itself encompasses the change; however, this has not been shown). Importantly, this dilution of cellular components is not necessarily symmetrical as young HSCs predominantly divide in a polar fashion whereas old HSCs shift towards symmetrical divisions (Florian et al., 2018). The selective distribution of components to the daughter cells can include organelles (Loeffler et al., 2019) and could also affect the epigenetic landscape (Florian et al., 2018) contributing to altered potential and function.

Regardless, we propose that proliferative stress is most likely to be perpetuated by the following mechanisms: The first would be DNA damage/mutations as discussed above. One such form of DNA damage which seems to be specifically relevant towards that purpose is telomere shortening. Indeed, telomere dysfunction in HSCs affects the intrinsic checkpoint regulation and differentiation via p21 (Choudhury et al., 2007) and BATF (Wang et al., 2012) resulting in myeloid biased hematopoiesis (Chen et al., 2015; Wang et al., 2012), predisposing to myelodysplastic syndrome (MDS) (Colla et al., 2015), and negatively impacting HSC reconstitution potential (Samper et al., 2002). However, the proliferation capacity of murine HSCs seems to be telomerase independent (Allsopp et al., 2003; Samper et al., 2002) though it is important to note the caveat that murine telomeres are far more extensive than human telomeres. Indeed, telomere-related human pathologies are often associated with bone marrow failure (Fiorini et al., 2018), emphasizing the importance of telomere maintenance for human HSCs. In mice, the progeny of HSCs seems to be more heavily impacted by shortened telomeres, but the resulting damage can feedback to the HSCs and induce skewed hematopoiesis in a cell non-autonomous fashion (Ju et al., 2007; Song et al., 2010).

Another potential compartment to harbor long-term proliferative stress effects would be the mitochondrial content or activity, which has previously been implicated in HSC aging (Hinge et al., 2020; Ho et al., 2017). Until recently, mitochondrion were thought to be sparse in HSCs due to experimental assays measuring mitochondrial dye incorporation (together with reported low respiratory activity). However, HSCs have since been reported to efficiently efflux mitochondrial dyes, and it is now clear that mitochondria are abundant in HSCs (de Almeida et al., 2017). Future studies on how mitochondrial content is generated and distributed during HSC proliferation and aging would improve our currently limited understanding.

While this review focuses on processes that are intrinsic to HSCs, it is important to note that HSC proliferation is also linked to their surroundings. This environment, the bone marrow niche, is instrumental to both HSC biology and aging (Guidi et al., 2017; Ho et al., 2019; Li et al., 2018; Morrison and Scadden, 2014; Pinho and Frenette, 2019). Recently, it has also been shown that quiescent HSCs reside mostly in protected perisinusoidal niches (Sacma et al., 2019) and that this niche plays a dominant role in activating HSC proliferation following myeloablation. While this is a still developing, complex field of HSC research (Kokkaliaris, 2020), the fact that the niche can both regulate HSC biology and differentially harbor HSCs based on their proliferative history, mandates serious consideration for any potential rejuvenation strategy. This is particularly true when considering the phenomenon of “inflammaging” (Franceschi and Campisi, 2014). As HSCs are the precursors of the immune response, they would naturally react to immune signals and would be strongly affected by inflammatory signals. Thus, it will also be very interesting to understand how local sterile inflammatory signaling will affect HSC proliferative history.

Perhaps the most appealing candidate compartment to carry the scars of proliferative stress are epigenetic regulators such as DNA methylation or chromosomal modifications. For one, the epigenetic landscape is altered with aging in HSCs (Beerman et al., 2013; Beerman and Rossi, 2015; Sun et al., 2014). More specifically, the overall methylation profile of HSCs that are forced to proliferate has many commonalities with aged HSCs (Beerman et al., 2013). In addition, mutations that drive clonal expansion often directly influence epigenetic programs. The most prominent of these are mutations in DNMT3A (Shlush et al., 2014), TET2, and ASXL1 (Jaiswal et al., 2014), all three of which directly regulate the chromatin landscape. However, the bright side of considering epigenetic modifications as major players driving HSC aging is that they are potentially reversible, as exemplified by Wahlestedt et al. (2013) who demonstrated that an epigenetic reset could rejuvenate HSC functions, independent of their proliferative history. Another example of the importance of epigenetic modifications is seen in the regulation of chromatin architecture by Cdc42 modulation (Grigoryan et al., 2018). Thus, in order to improve upon such rejuvenation strategies (Spehar et al., 2020), future studies should first (i) develop improved epigenetic engineering techniques that could reliably edit HSCs and (ii) better understand the modifications that occur as a result of proliferative stress in HSCs to know where to specifically target.