Aging is an inevitable process if one lives long enough. There is an ongoing shift in demographics such that older persons will out number young persons in the coming years, and with it age-associated tissue attrition and increased diseases and disorders. There has been an increased influx in literature on the association of the aging process with dysregulation of hematopoietic stem (HSC) and progenitor (HPC) cells, and hematopoiesis. Most such hematopoietic aging studies have been carried out in mice, where it has been reported that the aging process (in this case mice in the range of 2 years old) compared to that of younger mice is associated with increased absolute numbers of phenotypically defined HSCs identified by cell surface antigens in the bone marrow (BM). Yet, the functional capacities of these increased numbers of HSCs are grossly deficient in their engrafting capability in competitive and non-competitive HSC transplants in lethally irradiated mice. Moreover, the differentiation capacity of the engrafted donor BM cells from the old mice is different from that of young donor BM cells; there is a shift in the lymphoid/myeloid ratio of engrafting cells such that the older donor BM cells manifest greater numbers of myeloid to lymphoid cell output. This is the opposite of that of young engrafting mouse BM HSCs. How informative this and other information on aged hematopoiesis is remains to be determined by further investigation. Recent work from our laboratory [1] has observed that at least some of the abnormalities of HSCs from old mice may be more of an artifact of the collection and processing of mouse BM cells, rather than how they manifest their numbers and functional capacities in vivo.

This review provides an extensive, although not necessarily complete, summary on the literature of aged hematopoiesis, HSCs and HPCs. When placed in context of potential artifacts of the collection and processing procedure, that may not be totally representative of the status of HSCs in the in vivo microenvironment of BM, the site in which HSCs, HPCs, and hematopoiesis are nurtured for self-renewal, proliferation, survival, and differentiation some of what we know may have to be re-evaluated. This review encompasses the following sections: A) Aging and Stem Cells in General, B) Age-Related Changes in HSCs/HPCs and Hematopoiesis, C) Age-Related Clonal Hematopoiesis of Indeterminant Potential (CHIP) and Inflammation, D) DNA Damage, Transcriptional and Epigenetic Changes During Aging, E) Metabolic Processes, Mitochondria and Reactive Oxygen Species (ROS) During Aging, F) Apoptosis, Autophagy, Radiation and a Role for the Sirtuin Family of Proteins During Aging, G) The Microbiome, Hematopoiesis, and Aging, H) Additional Age-Related Information, The Microenvironment, Exosomes, Leptin (Lep) and Leptin Receptors (R), and Means to Better Evaluate and Understand Hematopoiesis During Aging in part in context of our recent studies [1], I) COVID-19, SARS-CoV-2, Aging and Hematopoiesis, and J) Conclusions in Context of Potential Future Interventions for Better Health of the Hematopoietic System During Aging. One of the authors (HEB) of this review had an interest in Gerontology, the study of aging, over 50 years ago, but it is only most recently, that he, his lab members, and collaborators have been involved in actual experiments in this area, having previously focused on the regulation of hematopoiesis in the young.Footnote 1

A) Aging and Stem Cells in General

It has been suggested that aging is not caused by active gene programming, but that it rather evolved through limitations in maintenance of somatic cells in which there was a build up of damage [2], which in fact is associated with gene mutations that affect endocrine signaling, stress responses, metabolism and telomere length [3]. Thus, aging is believed to entail “damage” due to multiple mechanisms, some information of which may possibly be used to slow some of the “damage” during aging for healthier outcome. Covered in these papers [2, 3] are the areas of: why aging occurs, is it programmed, how does evolutionary genetics and physiology fit into these processes, how aging is “caused” in terms of molecular mechanisms, mitochondria, and network themes. Whether we yet know enough about the aging process of cells, their organelles, and organisms is still open for debate, although more insight into problems and causes associated with aging could provide the means to potentially intervene at least partially in the future.

Human aging is associated with a number of diseases and defects including the heart, muscle wasting, osteoporosis, and in some cases mental deterioration [4]. Senescent cells and their accumulating damage can contribute to aging, through a number of intracellular signaling pathways including the p53 and RB tumor suppressors, and the influences of neighboring cells in the environment [5]. These, and a number of other genetic pathways have been implicated in aging [6], including nutrient sensing pathways. Over thirty genetic mutations have been reported to extend the lifespan of mice, and a number of genes have been associated in genome wide association studies with longevity of humans [6]. Ames Dwarf mice which harbor a spontaneous mutation in the Prop1df gene resulting in the lack of growth hormone, prolactin, and thyroid-stimulating hormone are known to live close to two times the lifespan of other mouse strains [7,8,9,10], a situation mimicking certain conditions in humans. Why we age has been commented on from an evolutionary point of view [11]. While somatic cells have a limited lifespan, the lifespan of stem cells, which have the property of making more of themselves (self-renewal) and being able under the appropriate stimuli for differentiation to more mature cell types has not yet been conclusively defined, although transcriptional fingerprinting and other pathway analyses suggest that stem cells do themselves age as one gets older [12, 13].

B) Age-Related Changes in HSCs/HPCs and Hematopoiesis

A number of articles and reviews appeared in this area of research since 2005 [14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37] which will be mainly described in chronological order so that the reader can see what research was reported first and then subsequently, as not all research findings may agree. It was reported that the aging of long-term (LT) HSCs was associated with autonomous changes that increased the self-renewal of these cells, but that these HSCs manifested decreased potential for lymphoid cell differentiation and production [14]. This was associated with down-modulation of genes that mediate lymphoid specification, and up-modulation of myeloid fate decisions and functions [14]. Competitive transplants were done using the congenic CD45.1/CD45.2 mouse system with relatively purified populations of donor HSCs to assess engraftment, and self-renewal was estimated by secondary transplants. This paper [14] did not note the decreased engrafting capacity of HSCs from BM of older mice that most of the other numerous publications in this area have reported, and lacked detailed month by month chimerism data comparing engraftment of old vs. young mice; nor did it quantitate numbers of functional HSCs using limiting dilution analysis to calculate competitive repopulating units (CRUs, a measure of the numbers of functional HSCs [38]. Age-related defects in lymphoid-bias [15] and B-lymphopoiesis [16] have been reported by others, and have been suggested to underlie the dominance of myeloid cells in adult leukemia [17]. In order for HSCs to engraft, they must first home to the BM after IV injection. This process of homing for BM cells of old mouse CRUs was about 3-fold lower than the homing efficiency of CRUs from young mice [18], hence one potential reason for decreased engrafting capability noted by others. Of some interest, although not completely understood, the ability to mobilize HSCs from old mice with G-CSF to the blood was increased compared to that of G-CSF mobilization of HSCs from younger mice [19]. This correlated with a reduced adhesion capacity of an immature cell population (not a purified HSC population) to stromal cells, and with increased activation of Cdc42, a small RhoGTPase. This work has not yet been reproduced to the knowledge of the authors of this review, and more rigorous analysis is needed to fully understand this interesting phenomenon. What has not been defined yet is the mobilization of lymphoid vs. myeloid-biased HSCs in old vs. young mice. It will also be of interest to assess the mobilizing capacity of bonified HSCs to the combination of G-CSF plus AMD3100 (Plerixafor), as G-CSF and AMD3100 synergize to mobilize HSCs and HPCs from young mice [39].

Other reviews have noted age-related changes in hematopoiesis of old vs. young mice [20, 21], with one short report [22] not seeing differences in engraftment of sorted populations of HPCs from elderly (>70 years old vs. young) human BM in immune deficient NSG mice. This clearly needs more rigorous investigation in terms of numbers and engrafting capability of rigorously purified populations of functional human BM HSCs (not HPCs) from old and young donors.

While an earlier report [14] suggested increased self-renewal of HSCs from BM of old mice, a later report by others with more in-depth analysis demonstrated that HSCs from the BM of old mice manifested significantly reduced self-renewal in secondary transplants using highly purified populations of LT-HSCs [23] for both primary and secondary engraftment. They [23] as did others [18] showed decreased homing efficiency of HSCs from the BM of old mice. Moreover, they [23] showed significantly delayed proliferative responses of old vs. young BM HSCs.

What is clear is that all studies thus far that have assessed old vs. young BM engrafting HSCs have shown a bias of the myeloid vs. lymphoid production capability of HSCs from old mice [14, 16,17,18, 20,21,22,23,24]. Whether this apparent bias of donor HSCs from old mice might be due to potential artifacts in how donor cells were collected, processed, and injected into recipient mice [1] will be discussed in Section H.

The impact of hematopoiesis in aging primates was investigated by clonal tracking in which clonal output of thousands of genetically barcoded HSCs and HPCs was determined in old vs. young macaques after autologous transplantation [25, 26]. Delayed output from multipotent clones was observed in old macaques with persistence of lineage biased clones noted; in contrast to aging studies in mice which showed persistence of myeloid-biased clones with old age, there was persistent output from both B-lymphoid- and myeloid-biased clones. Whether or not macaque vs. mouse differences were due to aging differences between species requires further investigation as these studies [25] were based on only two old macaques 18 and 25 years of age, which were considered “aged” on their lifespans of captivity of 20-30 years.

The multipotential progenitor (MPP) cell compartment is a composite of 4 different cell types, with the MPP4 compartment being considered to be lymphoid-primed [27]. A yet to be understood observation in context of lymphoid-biased aging studies is the progressive loss and increased cycling of the MPP4 population with aging; other cells and factors may be involved in lymphoid-biased output from engrafted aged HSCs.

Two intriguing reviews on HSC aging are entitled: “The slippery slope of hematopoietic stem cell aging” [30], and “Age-related clonal hematopoiesis: Stem cells tempting the devil” [29]. The latter review touches on clonal hematopoiesis of indeterminate potential (CHIP), an area that will be covered in detail in Section C, and is associated with increased risk of hematological cancers, as well as that of the mortality associated with cardiovascular problems. A number of other more recent reviews are worth noting including: “Aging of hematopoietic stem cells” [31], “Anemia at older age: etiologies, clinical implications and management” [32], “Aged murine stem cells drive aging-associated immune remodeling” [33], “The global complexity of the murine blood system declines throughout life and after serial transplantation” [34], “Hematopoietic stem cells aging, life span and transplantation” [35], “The ageing hematopoietic stem cell compartment” [36], and “Relationships between aging and hematopoietic cell transplantation” [37]. All these reviews suggest that intervention in age-related dysfunction of HSCs may be possible, in part by targeting selected intracellular regulatory pathways. We suggest in Section H, the potential use based on studies in mice of HSCs and HPCs from older individuals for efficient hematopoietic cell transplantation (HCT), if the cells are more appropriately collected and processed under conditions that maintain their in vivo numbers and functional characteristics [40,41,42]. How these cells are collected may be crucial, as cells are currently collected in almost all mouse studies, except those noted by us [40,41,42], and in all human studies, in ambient air (~21% O2 tension). Collection of cells in ambient air subjects them to a phenomenon which we termed Extra Physiological Oxygen Shock Stress (EPHOSS) [40, 41]. Ambient air causes the very rapid loss of HSCs and a concomitant increase in numbers of HPCs due to EPHOSS-induced differentiation of HSCs [40]. This differentiation process during collection of cells in air occurs within minutes, and may likely be needed to be considered in interpretation of at least some of the published information presented. This may require some re-evaluation of past studies to ensure that studies accurately describe the situation of numbers and functions of these cells as when they are present in their BM microenvironment, before removal for collection and analysis, as we reported in [1] and discuss in Section H.

C) Age-Related Clonal Hematopoiesis of Indeterminant Potential (CHIP) and Inflammation

Age-Related Clonal Hematopoiesis

CHIP, also known as age-related clonal hematopoiesis (ARCH), is characterized by expansion of somatic mutations in various hematopoietic lineages of older persons and is associated with risks of developing leukemia [43], as well as other age-associated disorders including cardiovascular disease [44]. Human aging is associated with an exponential increase in the occurrence of CHIP in aged individuals. It is an emerging public health issue that affects at least 15-20% of individuals aged 70 or above [43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59]. A number of reviews and reports on clonal hematopoiesis have been published [45,46,47,48,49,50,51,52,53,54,55,56,57,58,59]. This is currently a heavily researched area of investigation, with the causes still relatively unknown. Clarity is needed on why some cells with mutations, likely involving and caused by several factors [57] noted in the below Sections, persist and/or expand with resultant disorders such as leukemias, myelodysplasias, and cancers associated with aging individuals.

The vast majority of the mutations identified in CHIP are dispersed across the genome. However, five genes, including DNMT3A, TET2, ASXL1, JAK2, and TP53, have high numbers of somatic mutations [54,55,56,57,58,59]. The most common base-pair change in the somatic variants identified in CHIP was a cytosine to thymine (C to T) transition, a somatic mutational signature of aging [54,55,56]. CHIP is an age-dependent risk factor for both hematological malignancies and cardiovascular disease [53,54,55,56,57,58,59]. Thus, preventing CHIP progression may prove to be beneficial for human health. However, mechanisms by which somatic mutations in HSCs and other blood cells contribute to the pathogenesis of age-related diseases are largely unknown.

Clinical studies revealed that hematopoietic clones harboring specific mutations in individuals with CHIP may expand over time [54,55,56,57,58]. However, how different cellular stressors affect clonal expansion is largely unknown. Recently, three different stressors, including hematopoietic transplantation, cytotoxic therapy and inflammation, have been shown to expand hematopoietic clones. TP53 mutations identified in CHIP confer a competitive advantage to HSCs and HPCs following transplantation through modulating epigenetic pathways [52]. Considering that common mutations identified in CHIP affect epigenetic modulators, including DNMT3A, ASXL1, and TET2, these findings underscore the importance of dysregulated epigenetic control in CHIP development.

PPM1D is a phosphatase that negatively regulates p53 and several proteins involved in the DNA damage response (DDR) pathway [60]. Recently, PPM1D mutations were found in CHIP [54,55,56,57,58]. PPM1D mutations result in the expansion of PPM1D-mutant hematopoietic cells following chemotherapy treatment. However, they do not confer competitive advantage to HSCs and HPCs following bone marrow transplantation [61, 62]. TP53 mutations are associated with prior exposure to chemotherapy [63]. Genotoxic stresses selectively expand TP53-mutant HSPCs [50, 64]. While both p53 and PPMID are involved in the DDR pathway, they appear to play distinct roles in promoting of HSCs and HPCs expansion.

The Effects of Chronic, Low-Grade Inflammation Associated with Aging

During aging, chronic and low-grade inflammation - inflammaging - develops, which contributes to the pathogenesis of age-related diseases [65, 66]. Aberrant innate immune activation and pro-inflammatory signaling within the malignant clone and the BM microenvironment have been identified as key pathogenic drivers of myelodysplastic syndrome (MDS), an age-related disease [67]. Mutations identified in CHIP may utilize cell extrinsic mechanisms to promote clonal hematopoiesis. For example, TET2-deficient macrophages exhibit an increased in NLRP3 inflammasome-mediated interleukin-1β secretion [68]. Inflammasomes are multiprotein complexes that activate Caspase-1 and increase the release of pro-inflammatory cytokines such as IL-1β, leading to caspase-1-dependent death, known as pyroptosis [69]. HSCs and HPCs from low to high-risk human patients with MDS manifest activated NLRP3 inflammasome [70]. NLRP1 inflammasome activation increases IL-1β secretion that inhibits wild-type HSPC function through inducing pyroptosis [71]. The NLRP1 inflammasome, but not the NLRP3 inflammasome, is specifically activated in p53 mutant HSPCs, leading to increased secretion of IL-1β, which induces pyroptosis of wild-type HSPCs in a paracrine fashion (YL and HEB, unpublished data). Tet2-deficient hematopoietic stem and progenitor cells manifest a hyperactive IL-6 pathway, which promotes cell survival under basal conditions and in response to inflammatory stress. Inhibiting inflammatory signaling in Tet2 mutant preleukemic cells mitigates stress-induced abnormalities and clonal hematopoiesis [72].

Splicing of pre-mRNAs by the spliceosome plays a key role in tissue development [73, 74]. Genome wide splicing analysis revealed an increased number of spliced genes during aging [75, 76]. Changes in spliceosome gene expression and alterations in pre-mRNA splicing are associated with lifespan in mice and humans [77]. Notably, both human and mouse HSPCs display dysregulated pre-mRNA splicing with age [78, 79]. Further, spliceosome gene mutations, including SF3B1, SRSF2 and U2AF1, were frequently found in CHIP and MDS [54,55,56,57, 80,81,82], implicating that aberrant splicing in hematopoietic cells may contribute to CHIP and pathogenesis of MDS. Although both SRSF2 and SF3B1 mutations alter mRNA splicing, these mutations functionally converge with hyperactivation of NF-κB, a key mediator of the inflammatory response [83]. These findings underscore the importance of chronic inflammation in promoting CHIP development during aging.

Inflammation

Inflammation is a double-edged sword in hematopoiesis and disease. The hematopoietic system gives rise to the immune cells of the body and is, therefore, closely linked to inflammation. Even at early stages of hematopoietic development, inflammatory cytokines, such as interleukin-1 (IL-1), interferon-γ (IFN-γ), tumor necrosis factor (TNF), and granulocyte colony stimulating factor (G-CSF), play critical roles in the specification of hematopoietic stem cells (HSCs) [84]. Remarkably, different levels and combinations of inflammatory signaling molecules can elicit opposing responses, suggesting that context may be significant. For example, while IFN-γ and TNF are linked to bone marrow failure and decreased self-renewal capacity in adults, they enable hematopoiesis during development [84]. Under normal conditions in the mature hematopoietic system, cytokines influence HSC proliferation, differentiation, and self-renewal [84]. IFN-γ, IL-3, and IL-1 can influence the differentiation of HSCs toward myeloid lineages by activating myeloid transcription factors [84]. These same inflammatory factors stimulate hematopoiesis to support the immune system during infection and injury in a process called emergency granulopoiesis [85]. This process leads to the expansion of myeloid cells, which serve as the first line of defense against foreign pathogens [85]. These mechanisms rely on inflammatory mediators and are essential for maintaining homeostasis.

While inflammation enables hematopoietic development and stimulation of HSCs during illness and injury, it also contributes to pathogenesis and disease progression. During infection, inflammatory signals, such as IFN-γ and IL-27, trigger the proliferation and differentiation of HSCs to bolster the immune response either by acting directly on HSCs or indirectly via mature hematopoietic cells, endothelial cells, or the bone marrow microenvironment [86,87,88,89,90]. However, prolonged bacterial or viral infection can hinder self-renewal and competitive repopulation capacity, leading to HSC depletion [91, 92]. This exhaustion of HSCs during chronic inflammation may be attributed to increased myeloid differentiation [93]. Notably, dysregulation of the myeloid cell compartment also occurs in patients with severe COVID-19 [94]. This inflammation-mediated HSC dysfunction may occur via the TLR4-TRIF-ROS-p38 signaling pathway rather than Myd88 signaling, suggesting that the mechanism underlying chronic inflammation may be distinct from that of emergency granulopoiesis [91]. In the context of sepsis, activation of Myd88 caused myelosuppression without significant effects on HSCs, whereas activation of TRIF strongly inhibited HSC self-renewal without direct effects on myeloid cells, inferring cell type-specific effects of these inflammatory mechanisms [95]. Targeting these two pathways may have therapeutic value. While foreign infections can alter hematopoiesis by triggering inflammation, the normal microbiome can also influence the hematopoietic system. See more on this in Section G. Antibiotic-treated mice exhibit depletion of HSCs and progenitor cells as well as anemia, thrombocytosis, pan-lymphopenia, and leukopenia [96]. The complexity of the intestinal microbiome regulates the size of the myeloid cell population in the bone marrow via Myd88 signaling [97]. The disruption of these interactions may have implications for the potential contribution of infection to the progression of preleukemic conditions to hematological disease. For example, disruption of the intestinal barrier promotes myeloproliferation in mice lacking the preleukemic gene Tet2, whereas germ-free Tet2-deficient mice do not exhibit myeloproliferation [98]. Germline Tet2 loss of function is associated with immunodeficiency and lymphoma in children [99, 100]. Importantly, myeloproliferation was alleviated in Tet2-deficient mice with loss of intestinal integrity by treatment with antibiotics [98, 101]. Similarly, bacterial signals cause the expansion of HSCs lacking Tet2 and induce the production of IL-6 from HSCs, bolstering the role of infection and inflammation in the pathogenesis of hematological malignancy [102]. A link to these effects in aged animals remains to be better elucidated, with effects of bacteria on tumor progression and metastasis covered in Section I.

Remarkably, many of the same inflammatory pathways that guide hematopoietic development and strengthen the immune system under normal conditions can also drive leukemia in the context of infection or other inflammatory conditions. For example, IL-6 facilitates the development of chronic myelogenous leukemia in mice, and IL-33 contributes to myeloproliferative neoplasms by altering myelopoiesis [103, 104]. In addition, inflammation can cause genotoxic stress, the accumulation of mutations, and the progression of preleukemic conditions to leukemia [105, 106]. An unexplored area is the elucidation of the factors and events responsible for the transition from the normal inflammatory response to the deleterious inflammation that can promote hematological disease. As during hematopoietic development, the context of the inflammatory response may be important in shaping disease outcomes. For example, chronic IL-1 signaling can reduce HSC self-renewal, limit hematopoietic lineages, and impair the response of HSCs to replicative challenges [107]. A recent study of MDS indicates that inflammation acts as a selective pressure that specifically fosters expansion of preleukemic or malignant HSCs compared to normal HSCs [108]. Distinct hematopoietic cell subtypes may exhibit differential responses to specific inflammatory signaling molecules, further supporting heterogeneous HSC populations that can drive leukomogenesis [84, 109].

The effect of inflammation on HSCs has valuable implications for age-associated diseases, as older individuals exhibit elevated levels of inflammation [110]. Inflammation plays a significant role in expansion of HSC clones carrying preleukemic mutations in CHIP. Characterized by acquisition of somatic mutations in hematopoietic lineages with age, CHIP is associated with both hematological malignancy and cardiovascular disease (CVD), broadening the role of preleukemic mutations in disease states [111, 112]. The pro-inflammatory response observed in HSCs carrying preleukemic mutations indicates the existence of intrinsic mechanisms of inflammation for HSCs [102, 113, 114]. The addition of a pro-inflammatory preleukemic mutation can alter the presentation of hematological malignancy [115]. It has been proposed that HSCs and mature hematopoietic lineages propagate the inflammatory response via a feedforward mechanism, in which inflammatory signals from one cell type amplifies the other [116]. Pro-inflammatory macrophages secreting IL-1α enable CHIP-associated CVD, underscoring that some of the same pathways involved in hematopoietic homeostasis and leukemogenesis also contribute to the non-hematopoietic manifestations of CHIP [68]. In addition, as in emergency granulopoiesis, aging promotes myeloid expansion, further emphasizing potential mechanistic overlap [109]. While intrinsic factors appear to be important to the inflammatory response involved in the pathogenesis of CHIP-associated diseases, extrinsic mechanisms may also influence this process as the conditioned media from aged mesenchymal stromal cells impairs the function of young HSCs [117]. In addition to acting directly on HSCs, inflammation can also remodel the bone marrow microenvironment, which regulates hematopoiesis [118]. Additional studies are needed to understand the ways in which these pathways are altered during aging and pathogenesis, and how they may be modified for health benefit.

Elderly populations may be uniquely susceptible to the effects of inflammation and CHIP as comorbidities and CHIP are more common in aging individuals. Several comorbidities with inflammatory components, such as ulcerative colitis, rheumatoid arthritis, and systemic sclerosis, have been linked to increased clonal hematopoiesis [119,120,121]. Pro-inflammatory features of CHIP-associated HSCs/HPCs may influence outcomes of hematopoietic stem cell transplants, leading to cytopenias, chronic graft vs. host disease, and/or donor-derived leukemia (DDL) [122]. These patient populations may especially be vulnerable to development of hematological malignancies and CHIP-associated diseases following infection, as infection may exacerbate the existing inflammatory response in these patients. Inflammation may be a key factor contributing to the heterogeneity observed in hematological malignancies and CHIP-associated diseases and should be considered in the clinical management of patients with these conditions. In particular, differences in the clinical presentation between donors and recipients that develop DDL highlight the potential role of inflammation and comorbidities in promoting leukemogenesis. It is yet not known if these different sources of inflammation influence hematological malignancies and CHIP-associated diseases via common mechanisms.

Intrinsic and extrinsic sources of inflammation may represent potential therapeutic targets for hematological malignancies and CHIP-associated diseases. Inhibition of inflammation can impede clonal expansion in response to inflammatory stimuli [102]. Blocking inflammation may also be a valuable therapeutic approach in CVD, as inhibiting IL-1 receptor signaling from pro-inflammatory macrophages can prevent CHIP-associated CVD in mice [68]. It is currently unclear whether suppressing inflammatory HSCs or mature hematopoietic lineages is more effective. Inflammatory signals from the bone marrow microenvironment may also be viable therapeutic targets but have not yet been investigated. A more thorough understanding of the timing of the inflammatory response and the cell types involved will facilitate the effective inhibition of inflammation in hematological malignancies and CHIP-associated diseases. Strategies to both prevent pathogenesis and to treat existing disease will be valuable. A critical aspect of targeting inflammation in these disease contexts will be to maintain the normal inflammatory response necessary for responses to infection and injury while targeting aberrant inflammatory pathways that promote disease; however, additional studies are needed to elucidate factors governing these processes. Thus, Inflammation acts as a dysregulator of tissue maintenance and regeneration during aging as evidenced by the fact that HSC do not regenerate well after inflammatory challenge [123].

D) DNA Damage, Transcriptional, and Epigenetic Changes During Aging

DNA damage accumulates with age, and defects in DNA repair can cause cellular changes that resemble a premature aging phenomenon [124,125,126,127]. Tables on selected models of premature aging in mice and their common features have been summarized in a review [126] and p53 implicated in DNA damage [87]. While DNA damage to HSCs and HPCs during aging is clearly impacted, such damage to the microenvironmental niche cannot be overlooked [124]. Transcriptional changes in stem cell populations have been profiled for HSCs and other stem cell types, but it is not clear yet if a common age-related signature has been identified [12]. A role for epigenetics in the aging process is also considered [128]. Epigenetic hallmarks of aging and senescence have been diagrammed, as have been the pros and cons of using model systems to study aging and senescence in a variety of species, along with a short listing of repositories and tools for evaluating a role for research in the aging process [128].

Repair of damage has been shown to offset deficient HSC function during aging [129]. This was especially apparent under stress conditions, in which DNA damage led to loss of the potential of HSC reconstitution, proliferative capacity, self-renewal activity, enhanced apoptosis, and then exhaustion of function [129]. It was suggested that the accrual of DNA damage may be a means contributing to HSC functional defects of these cells to respond to acute stress or injury [129].

A shift from canonical to non-canonical signaling by Wnt, in response to elevated expression of Wnt5a was associated with the process of HSC aging [130]. Treatment of cells from young mice with Wnt5a induced aging associated HSC apolarity, reduced their capacity for regeneration, and resulted in a age-related shift in myeloid/lymphoid differentiation, that was associated with activation of the GTPase Cdc42 [130]. Moreover, haploinsufficiency of Wnt5a resulted in the attenuation of the aging phenotype of HSCs [130]. Other studies defined replication stress as a driver of functional declines in HSCs during aging [131]. This was associated with decreased expression of mini-chromosomal maintenance helicase components and altered DNA replication fork dynamics [132].

There are reports on a role for epigenetics in abnormalities associated with HSCs in old mice [132,133,134,135]. While the decline of HSC function seemed to be dependent on their proliferative history, it was noted to be independent of the length of their telomeres [133]. HSCs from old mice manifested reduced signaling of transforming growth factor-beta with changes in genes involved in proliferation and differentiation of HSCs [135]. HSCs from old mice had broader peaks of H3K4me3 with increased methylation of DNA at the transcription factor binding sites that were associated with genes involved in promotion of differentiation, and a reduction of genes associated with maintenance of HSCs [135]. Ribosomal biogenesis was found to be a particular target of this age-related HSC phenotype; there was increased transcription of ribosomal protein and RNA genes, and the hypomethylation of genes for ribosomal RNA [135].

Proteosome analysis [136] and single-cell RNA sequencing [137] have been performed on HSCs from old mice. How much these analyses really inform us about the prime drivers in HSC dysfunction remains to be determined, especially since the cells were collected in ambient air prior to analysis, which may not be optimal for assessing physioxia associated effects [1]. Of some interest, deletion of inhibition of DNA binding1 (Id1), a helix-loop-helix transcription faction protected HSCs from both the effects of stress-induced exhaustion and that of aging [138].

E) Metabolic Processes, Mitochondria and Reactive Oxygen Species (ROS) During Aging

Metabolism, mitochondria and ROS are important aspects of HSC function [40, 41], as well as for other stem and progenitor cell types [139,140,141]. Aging is associated with extensive changes in metabolism [75,76,77]. A short report questioned whether or not metabolic mechanisms of stem cell maintenance might explain aging and its associated impact on stem cells [142]. Another review concentrated on mitochondrial contributions to dysfunction of somatic stem cells in general and in context of aging [143] and a review on mitochondrial metabolic checkpoints and aging of HSCs implicated mitochondrial maintenance mechanisms including mitophagy and asymmetric segregation of “aged” mitochondria [144]. This is an area that clearly requires more detailed investigation, although it has been suggested that mutations in mitochondrial DNA are not a primary driver of stem cell aging [145].

ROS has been implicated in various stem cell functions [40, 41, 146,147,148], and STAT3, mitochondrial dysfunction, and overproduction of ROS has been associated with a rapid aging-like phenotype [149]. Symmetric divisions of stem cells, including HSCs, results in increased stem cell numbers with maintenance of stem cell characteristics of the original “mother” cell. However, asymmetric division of stem cells can result in one daughter cell maintaining the original stem cell characteristics of the “mother” stem cell, while the other cell can be a more differentiated progenitor cell. To assess selective apportioning of subcellular contents between “daughter” cells using mammary stem like cells, it was found that “daughter” cells that received fewer “old” mitochondria were associated with maintenance of stem cell traits; inhibition of mitochondrial fission disrupted age-dependent subcellular localization and segregation of mitochondria with resultant loss of stem cell properties in the progeny [150]. It is not clear if such studies with mitochondria apportuning between HSCs undergoing symmetric or asymmetric divisions have yet been done, but it is certainly an area of interest if done in context of HSC from young and old bone marrow HSCs, and their collection and processing under physioxia conditions as noted [1].

There is still much to be learned regarding how stem cells maintain metabolic homeostasis. The unfolded protein response has been implicated as modulating the HSC pool during stress [151], but has apparently not yet been evaluated in HSCs from aged mice. However, a regulatory branch of the mitochondrial unfolded protein response, mediated by the interplay of the sirtuin, SIRT7 (more on sirtuins in Section F), and nuclear respiratory factor 1 (NRF1) which is a master regulator of mitochondria, was interrogated in HSCs [152]. It was noted that inactivation of SIRT7 resulted in reduced quiescence, increased mitochondrial protein folding stress, and decreased regenerative capacity of HSCs. Moreover, expression of SIRT7 was decreased in HSCs from old mice, and up-regulation of SIRT7 in the aged HSCs improved their regenerative capacity. This implicated the mitochondrial unfolded protein response-mediated metabolic checkpoint as a contributor to HSCs in old mice [152]. In addition, mitochondrial DNA polymerase, when defective, has been associated with premature aging in mice [153], but how, if at all, this relates to HSC function from old mice remains to be determined.

Thioredoxin-interacting protein (TXNIP) is a 397 amino acid residue, belonging to the arrestin family of proteins. It has been reported to regulate HSC quiescence and mobilization after stress [154,155,156], and is likely to be involved in HSC function, but has not to our knowledge been extensively investigated. Reasons to evaluate this during aging is that Txnip-/- mice have decreased HSC reconstitution resulting in HSC exhaustion, effects associated with hyperactive signaling of Wnt, an active cell cycle, and reduced expression of p21cip1. These stresses also affect the BM microenvironment resulting in decreased expression of CXCL12 (a chemotactic and homing chemokine)- and osteopontin-mediated interactions between HSCs and the BM [154]. TXNIP helps to maintain the pool of HSCs by functional switching of p53 after oxidative stress, effects that have been reviewed [155].

There is much to be learned regarding metabolic influences in aging, and molecular mechanisms underlying aging effects on HSCs still remain unclear. Elevated activity of the small RhoGTPase cdc42, previously noted by the investigators in another paper was linked casually to effects on HSCs in old mice [152], with a correlation of the loss of polarity in these cells. Moreover, by inhibiting cdc42 activity by pharmacological means, it “rejuvenated” the aged populations of HSCs by increasing the percent of polarized cells and restoring the level and spatial distribution of histone H4 lysine16 acetylation such that it was similar to that in HSCs isolated from young mice [152, 157]. This information further identified epigenetic regulatory changes in functional effects of HSCs from old mice, and may relate to metabolic changes.

Other studies linked the interaction of ROS dependent DNA damage, mitochondria, and p38 MAPK with senescence of adult mesenchymal stem/stroma cells (MSCs) from humans, with pharmacological inhibition of p38 MAPK partially recovered the senescence phenotype by partial prevention of hydrogen peroxide-induced senescence [158]. How linked senescence phenotypes are to the function of HSCs in aged persons remains to be determined. Somatic cell mitochondrial DNA (mtDNA) mutations contribute to such age-related disorders as those associated with myelodysplasia (MDS), and it was noted that the mito-protective effect of autophagy was impaired in erythroid cells of old mice [159]. mtDNA-mutated mice had somatic mtDNA mutations that were a targeted defect in the function of proofreading mtDNA polymerase, PolgA, and developed macrocytic anemia similar that seen in MDS patients. Mechanistic insight into these processes was reported [159], but whether or not these processes reflected changes in HSCs from old mice was not explored.

F) Apoptosis, Autophagy, Radiation, and a Role for the Sirtuin Family of Proteins During Aging

Aging-Related Apoptosis and Autophagy

Apoptosis, the phenomenon of programmed cell death, and autophagy, a self-degrative process responsible for eliminating cytosolic constituents such as long-lived proteins, aggregated proteins, and damaged organelles (mitochondria, ribosomes, peroxysomes) [160] have been linked to functional changes noted during aging [161, 162] and HCT [163]. Autophagy is associated with repair pathways that can protect hematopoiesis from injury due to nuclear radiation [164, 165]. Inhibition of autophagy by genetic manipulation was associated with normal and pathological aging, with its inhibition compromising the “longevity-promoting” effects of restriction of calories, the activation of SIRT1, inhibition of insulin and insulin growth factor signaling, and the administration of rapamycin, resveratrol, or spermine [161]. Autophagy was shown to maintain the metabolism of HSCs from both young and old mice [164]. These influences were not noted in all HSCs from old mice, with about a third of HSCs from aged mice demonstrating high autophagy levels being associated with a low metabolic state and high potential for regeneration [165]. This suggests that not all HSCs in aged mice are functionally compromised, an important point in aging HSC research that can be overlooked when studying HSCs from old mice at a total HSC population level. It is known that there are subsets of rigorously purified HSC populations that differ in mitotic history [166], and intracellular characteristics [1]. FOXO4 was suggested as a pivotal agent in the area of cellular senescence [167]. Using a FOXO4 peptide that disrupted the FOXO4 interactions with p53 in vivo where it was tolerated, restored certain functions in naturally aged and in fast aging XpdTTD/TTD mice. How this relates to HSCs in old mice remains to be evaluated.

Mitophagy is a process that is evolutionary conserved involving autophagic targeting and clearance of mitochondria that are destined for removal [168]. It is induced by short ubiquitin chains on the mitochondria [169]. Reviews on this process have been reported [168, 169] and discuss how metformin, an oral diabetes medication, both enhances and normalizes mitochondrial function that leads to alleviation of inflammation associated with aging [170]. What remains to be determined is if there is a role for mitophagy in HSC and HPC during aging, and if this can be modulated for health benefit.

Radiation Effects and Aging

Like aging, exposure to radiation is an additional stressor to the hematopoietic (H) system, the most sensitive tissue in the body to radiation damage. Therapeutic radiation, nuclear accidents, and malicious exposure from radiologic-warfare put mankind at risk for life-threatening acute radiation syndromes (ARS) and the delayed effects of acute radiation exposure (DEARE) in those fortunate to survive ARS. H-ARS, due to direct and indirect effects of radiation exposure on all classes of hematopoietic cells, leads to death within weeks if untreated [171, 172].

Hematopoietic DEARE, also known as residual bone marrow damage (RBMD), is characterized by diminished immunity and decreased production of blood cells persisting for years after radiation exposure [173,174,175,176,177,178,179]. Survivors of H-ARS exhibit severe lifelong damage to HSC, characterized by significantly decreased complete blood count, loss of HSC repopulating potential, loss of HSC quiescence, decreased numbers of HPC, and dramatic myeloid skewing, all most evident under stress [173,174,175,176,177,178,179,180,181,182,183]. An increased incidence of lymphoid malignancies, shortened life span, decreased mesenchymal stem/progenitor cell (MSC) number, and aberrant levels of endothelial cell-derived HSC niche proteins in aged H-ARS survivors have also been documented (Orschell, unpublished data). Long-term damage to the HSC-supportive niche also likely contributes to HSC dysfunction and RBMD. As enhanced cycling of HSC is believed to lead to loss of self-renewal potential and is detrimental to engraftment potential [184], it seems likely that the enhanced cycling of HSC from H-ARS survivors is a major contributor to RBMD. These data illustrate an unrecoverable loss of HSC self-renewal and differentiation potential, the two hallmarks of HSC [185,186,187,188,189,190,191], in survivors of H-ARS and suggest that compensatory mechanisms of hematopoietic support cannot overcome the “second hit” imparted by aging [14, 124, 129, 192, 193].

The DEARE are generally thought to result from persistent inflammation and chronic oxidative stress [194,195,196,197,198,199,200,201], leading to fibrosis [202] and loss of stem cell self-renewal functions. Indeed, elevated levels of pro-inflammatory cytokines associated with oxidative stress [203] have been reported in Japanese atomic bomb survivors [198, 202]. Other studies in atomic bomb survivors have shown possible reductions in self renewal capability of HSC secondary to dose-dependent DNA damage [204], as well as detriments in immune function [205], corroborating mouse H-ARS data. NAD(P)H oxidase, xanthine oxidase, and mitochondria have all been implicated as primary oxidant sources in various models and conditions [203, 206,207,208], and ROS has been documented in HSC post-irradiation as well [181]. NF-kB, one of several transcription factors activated by ionizing radiation [209], plays a central role in inflammation [210], is activated by oxidative stress and induces oxidative stress through interactions with cytokines [211], creating a potential feed-forward mechanisms to maintain chronic inflammation and oxidative stress.

Cellular senescence, and its associated oxidative/pro-inflammatory phenotype, has recently emerged as a causative mechanism of DEARE [211,212,213,214], making senescent cells a new therapeutic target for RBMD and other DEARE. Importantly, senolytic drugs have the potential to be used as an effective treatment for DEARE even after DEARE becomes a progressive disease [212], but it is not yet clear how this might be used in and for elderly exposed individuals.

As mice age, like humans, health issues and phenotypic changes begin to manifest and variability in experimental endpoints increases, necessitating the need for larger group sizes for sufficient statistical power. For example, mice of similar strains have been shown to exhibit significantly different life spans [215, 216] and radiation sensitivities when aged (Orschell, unpublished data), as well as differing susceptibilities to radiation-induced swollen muzzle syndrome [217], all depending on the vendor from which they were sourced. Mice from different vendors have also been shown to possess different fecal microbiota [218], which may contribute to vendor-specific phenotypic differences. For these reasons, investigators should consider stringent control of the vendor and barrier of their mice for aging studies to ensure optimal stability of their experimental models.

It is noted that more profound effects of aging may be produced not by life-threatening ARS (where the majority of those are exposed to high dose radiation), but rather by moderate or even low dose exposure.

The Role of Sirtuins in Regulating Aged HSC Function

Sirtuins are part of a large family of molecules, some of which have been linked in longevity/aging studies. The role of the sirtuin SIRT1 in stem cell biology, the aging process and in HSC function in old mice had been reviewed [219]. In this review it was noted that although the role of SIRT1 in teleomere maintenance was not resolved, its role in mitochondria and generation of ROS was highly implicated. It was observed that the genetic, hormonal, or drug manipulation of stem cell mitochondria may be useful as an intervening tool for manipulating HSCs from old mice. It was later reported that deficiency of SIRT1 compromised mouse embryonic stem cell hematopoietic cell differentiation in addition to embryonic and adult mouse hematopoiesis [220]. SIRT1 was reported to be required for maintenance of HSCs and lineage specification, in part by the transcription factor FOXO3 [221]. These investigators also suggested that SIRT1 may be involved in HSC function during aging by “delaying” HSC functional abnormalities [221], but this has not been rigorously studied. Although the role of SIRT1 and other sirtuins in the caloric restriction of modifying the aging process have been extensively reviewed [222,223,224], such studies do not always take into account the sex and mouse strains utilized [225] which could influence the reported results. SIRT3, while found to be dispensable for maintenance of HSCs and homeostasis of tissues during young age, was reported to be essential following stress and with the aging process [226]. SIRT3 expression was decreased with aging and upregulation of SIRT3 expression in HSCs of old mice improved their regenerative capacity, effects involving a role for mitochondrial homeostasis [226]. As noted above in Section E, SIRT7 in the mitochondrial unfolded protein response and aging-associated changes in HSCs in the old mice were linked [152]. This was discussed more thoroughly in a short commentary [227].

G) The Microbiome, Hematopoiesis, and Aging

The microbiome describes microorganisms such as bacteria, viruses, and fungi that colonize the human and animal body and influence various biological processes. Most studies that explored microbiome–hematopoiesis interactions are based on characterizing HSC and HPC populations in germ-free (GF) or broad-spectrum antibiotic-treated mice and in human subjects under prolonged antibiotic regimens or diagnosed with gut dysbiosis such as inflammatory bowel syndrome [228,229,230]. GF mice demonstrated myelosuppression, smaller HSC, MPP, and common lymphoid progenitor (CLP) populations, and impaired neutrophils, monocytes, and T-cell functions. Recolonization of GF mice restored immune response to infection [96, 228, 231]. However, a closer evaluation of HSCs and HPCs in oral antibiotic-treated mice revealed normal HSC and HPC populations but reduced mature T cell, B cell, and granulocyte populations, suggesting impaired differentiation of mature immune cells in microbiota-depleted mice after oral antibiotics treatment and introduced some discrepancies between animal models used to study the microbiome [232]. Following HCT, microbiota-depleted recipient mice immune reconstitution was significantly lower than their control counterparts [232] supporting the conclusion that the microbiome plays a role in regulating mature immune cell development. Several studies have linked the human gut microbiome imbalance or dysbiosis in conditions such as inflammatory bowel syndrome, malnutrition, and obesity to altered hematopoiesis [229].

The microbiome and hematopoiesis have been intimately linked [233,234,235,236,237,238]. Gut microbiota are known to sustain hematopoiesis [233], microbiota can regulate HSC differentiation by altering the BM niche [234], and CX3CR1+ mononuclear cells influence HPCs [235]. Reduced mPB is noted in mice receiving antibiotics [236] and microbiota modification has been discussed in context of hematology [237]. Moreover, gut microbiota are known to control bacterial infection by promoting hematopoiesis [238], but definitive and rigorous comparative studies on a role of the microbiome on hematopoiesis in the young and old are yet to be done.

Bacteria and the microbiome present a not uncomplicated scenario that has not been adequately addressed in context of aged hematopoiesis and this needs adequate attention. Certain bacteria, using that of Fusobacterium nucleatum as an example, have been implicated in enhancing metastasis of cancer cells [239,240,241,242,243,244,245,246,247,248]. If such bacteria have this capacity for cancer cell metastasis, then why not for HSC and HPC, migration and/or homing an area worthy of investigation.

Up to 15% of patients with a history of prolonged antibiotic use have suffered hematological adverse effects in the form of neutropenia, anemia, and pancytopenia [249, 250]. Although associations between the hematopoietic system and microbiota imbalance is apparent in mice and humans, mechanistic understanding of this interaction is limited. In the signal transducer and activator of transcription protein 1 (STAT1) knockout mouse, the antibiotic effects on HSC and HPC numbers were abrogated [96]. In another report, administrating a ligand of the pathogen recognition receptor 1 (NOD1) restored HSC and HPC numbers in GF mice [251]. To the authors' knowledge, aged-HSC and HPC function and phenotyping in relation to the microbiome have not yet been reported in mice or humans. Aging is associated with perturbation of intestinal epithelial integrity and upregulation of permeability, allowing microbiota entrance to the circulation and induces a chronic inflammatory state in the aged subject. Aging results in microbiota-associated increases in pro-inflammatory cytokine levels (e.g., TNF-α, TGF-β, IL-6, etc), changes in T-cell numbers (e.g., Treg, Th1, and Th2 T-cell subsets), and activation of TLR2, NF-κB and mTOR [252]. Considering how the well-characterized low-grade, chronic inflammation associated with aging affects the hematopoietic system, a role for microbiota promoting such inflammation is a strong possibility. How this might regulate hematopoiesis in the young and old remains an unexplored area to be better studied in steady- and stressed-states.

H) Additional Age-Related Information and Means to Better Evaluate/Understand Hematopoiesis During Aging

Role of Collection/Processing of Cells

Many studies have acknowledged the probable effects of oxidative stress on functional changes in stem cells during aging [253]. This oxidative stress is associated with damage to macromolecules including that of nucleic acids, proteins, lipids, and carbohydrates that could contribute to changes in HSC function. This however, did not consider how even the mere removing of HSC-containing populations of cells from mouse BM [40], mouse mobilized peripheral blood (mPB) [42], or human cord blood [40] could so quickly change HSC numbers and impinge on the function of the removed HSCs. As previously mentioned, collecting and processing cells under ambient air conditions for as little as 15 minutes exposes the cells to extra physiological oxygen stress/shock or EPHOSS. EPHOSS is associated with increased differentiation of HSCs to HPCs through a sequence of events involving p53, the opening of the mitochondrial permeability transition pore, cyclophilin D, hypoxia inducing factor (HIF)1-alpha and the hypoxamir, miR210 [40, 41]. By collecting and processing BM and mPB from mice and cord blood from humans in a hypoxic chamber set at 3% O2 and taking care to make sure that the cells are never exposed to ambient air conditions, it is possible to obtain many more phenotyped and functional HSCs [40,41,42]. Increased numbers of collected HSCs under hypoxia has also been reported for BM cells from Fanca-/- and Fancc-/- mice [254]. It is possible that some of the EPHOSS-related effects on HSCs exposed to ambient air can be compensated for by collection and processing of these cells at ~4°C [255]. Such cold collections and processing of human cord blood and mouse BM cells mimic at least some of the effects seen during hypoxia collection/processing of cells including increased numbers of collected HSCs. However, the mechanisms involved with preserving HSC numbers/function following cold collection/processing of cord blood cells have not yet been worked out and may differ somewhat from that of physioxia/hypoxia collected/processed cell populations.

Collecting/processing BM cells from old vs. young mice under different oxygen tensions [1], allowed us to demonstrate that functional engrafting HSCs from old mouse BM collected/processed at 3% O2 were equal in number to that of ambient air (~21% O2) collected/processed young mouse BM HSCs. Perhaps more importantly, the abnormal myeloid to lymphoid ratio seen when aged BM cells were engrafted into lethally irradiated recipient mice in a competitive transplant setting was not noticeable and completely resembled that seen after engraftment of young mouse BM. This was consistent with increased CLP numbers and decreased CMP and GMP numbers in the donor BM from old mice following their collection in hypoxia. These phenomena seen with old BM HSCs collected/processed in hypoxia, were associated with decreased total and mitochondria ROS, and decreased expression of stress-induced proteins [1]. Hence, aged mouse BM HSC function may not be as dysregulated as many others have reported, with differences perhaps due to their increased response to EPHOSS following ex vivo analysis in ambient air. A corollary of this may suggest that with the more physiological collection and processing of HSCs (a.k.a. in hypoxia conditions, or by means of a physiological strategy (e.g. collecting cells in the cold [255]) from BM of older individuals may be more functional in context of clinical HCT than if they were collected/processed as usual in ambient air O2. It is known that there are distinct populations of even rigorously purified populations of HSCs [1, 256]. Which of these HSC populations survive EPHOSS effects when cells are collected/processed in ambient air from young and old mouse BM remains to be better elucidated. While much more obviously needs to be done to more mechanistically and physiologically understand the true state of HSCs isolated from older human individuals or mice, it may be that much of the literature on aged HSCs needs to be re-evaluated for their functional status as they are within their in vivo physioxia microenvironment. Studies to be done with hypoxia vs. ambient collected HSCs include more in-depth intracellular events and signaling pathways that include gene and protein expression profiles, as well as epigenetic changes, work currently ongoing.

Molecular chaperones and heat shock responses have been postulated to play a role in longevity and aging [257], but rigorous studies in the area of aged HSCs are lacking. Age and organ specific differences in bioenergistics in Brown Norway rats have been noted [258], and p53 deficiency induces diverse dysregulated processes under physiological oxygen/physioxia [259], but we know that such events are influenced by modes of cell collection and processing [40, 41].

The Microenvironment

Not to be ignored in studies of aged HSCs is the role that their microenvironment in vivo plays in the functional cellular and intracellular abnormalities/defects associated with their engrafting deficiencies and biased differentiation patterns. The microenvironment niche for HSCs has been studied for young mice [260], but more in-depth studies of the aged BM microenvironment is warranted, especially in context of oxygen content [260], as we noted [1]. There is a report of the rejuvenation of progenitors from old mice when placed into and exposed to a young BM environment [261], and a more recent report demonstrated that degeneration of adrenergic nerves in BM affects aging of the HSC niche [262], and that aging in humans alters the special organization between populations of CD34+ cells (contains mainly HPCs, but also a small percentage of HSCs) and adipocytes in the BM [263]; this is of potential relevance as increased adiposity is associated with the aged BM microenvironment and can alter the functionality of surrounding cells. MSCs showed aging related expression of cxcr4 [264], a “homing” HSC/HPC receptor, but this has not yet been evaluated under conditions of hypoxia collection and processing such as in [40, 41].

Leptin (Lep) and Lep Receptors (R)

Metabolic activities of cells are adaptively regulated by systemic signals that reflect the nutritional status of an organism throughout its life [265, 266]. This is particularly crucial in the case of HSCs as they are rare, and they maintain the integrity of the entire hematopoietic organ system. One way that the body communicates nutritional cues to HSCs is via systemic hormonal regulations. Various metabolic hormones have been documented to influence hematopoiesis, and aging can significantly alter these metabolic messengers, hence indirectly affecting HSC functional behavior [267,268,269].

Among them, leptin controls the body energy expenditure and storage through both central and peripheral mechanisms [270]. As an adipokine, it is recognized to have broad spectrum effects on numbers and functions of different immune cells under homeostasis [271]. Aging is known to be associated with multiple dysregulations of the immune system, including a declined adaptive immune response [272]. Lep induces gene expression of p16, a marker of cellular immunosenescence in human B cells from young lean adults. These cells also exhibited lowered class switching activity and ability to produce influenza-specific IgG [273, 274]. Unfortunately, the study was limited to in vitro treatment only and did not provide full mechanistic insights. In line with this finding, it was demonstrated that lep induced significantly higher levels of IL-10, TNF-α, and IL-6 from aged human B cells as compared to young controls, and this effect was mediated through the STAT3 signaling pathway downstream of lepR activation [275]. Another group reported that sustainably higher lep levels were found in LPS-treated older (24 month) compared to young (2 month) rats, and the old rats showed delayed but longer febrile response. Elevated lep concentrations were accompanied by increased levels of pro-inflammatory cytokines including IL-6 and IL-1Rα; however, it was not determined whether the increase in lep level was causative and how [276]. Although lep signaling was consistently reported to be altered in aged animals, more rigorously designed studies are needed to help us understand how this well-known proinflammatory neuroendocrine adipokine may play any roles in aging-associated immunological changes [277, 278].

Beyond its role in immunity, lepR-expressing MSCs in murine hematopoiesis have been well-characterized as an indispensable source of stem cell factor for maintenance of both HSCs and more differentiated progenitor cells [279,280,281,282]. In addition, it has been demonstrated that BM adipose tissue possessed brown fat properties that declined in old or diabetic mice [283]. It was demonstrated that during the process of aging or in obesity, MSCs preferentially differentiated into adipocytes, which impaired hematopoietic recovery [284]. Since BM adipocytes could be a potential source of lep, it is important to know how BM adipocyte-derived lep can directly or indirectly alter hematopoiesis as the animal ages. In the context of HSC biology, we recently discovered that Lepr marks a small subset of robustly repopulating and self-renewing long-term HSCs in adult murine BM; Lepr+ HSCs (defined as LSKCD150+CD48-) were found to generate equal myeloid-lymphoid outputs as compared to Lepr-HSCs [285]. Given that lepr (OB-R) was also reported to be expressed by different types of both myeloid and lymphoid leukemic cells [286,287,288,289], it will be intriguing and important to determine whether aging has a differential selection pressure on Lepr+HSCs as compared to the rest of total BM HSCs, particularly in the context of clonal hematopoiesis. Although RNA-seq data suggested that Lepr+HSCs and Lepr+MPPs (defined as LSKCD150-CD48-) predominantly expressed the truncated short isoforms OB-Ra and OB-Rc, it remains to be determined whether aged HSCs or leukemic cells can express the long functional isoform OB-Rb (LepR). Future studies with more mechanistic insight should be able to address these questions, perhaps providing potential meaningful clinical implications.

More on Inflammation

Aging-related inflammation promoted aging characteristics of HSCs through a tumor necrosis (TNF)-alpha, ERK, ETS1, IL-27 receptor (R) pathway [290]. TNF-alpha increases during aging and induced expression of HSC IL-27R-alpha by ERK-ETS-1 intracelluar signaling, with deletion of IL-27R-alpha rescuing functional decline and myeloid bias of HSCs. Old IL-27R-alpha knockout mice had reduced proportions of myeloid-biased HSCs. Thus, this is another report implicating factors external to HSCs that effect the functional capacity of HSCs from old mice. Somatotrophic/Insulin-Insulin-Like Growth Factor has also been implicated in aging effects on stem cells as well, in these cases on a population of very small embryonic like stem cells [291, 292].

Exosomes

Exosomes are a subset of small extracellular vesicles that range in size from 30-150nm, that are produced by normal and malignant cells [293, 294], originate from the endocytic compartment of producer cells [295], and have emerged as a universal intercellular communication system [296, 297]. Exosomes are in a protective protein/lipid bilayer, are delivered to recipient cells without degradation, and freely cross biological barriers [296, 297]. Exosomes [298] influence proliferation of HPCs [299], but there is no information on effects of exosomes on HSCs and HPCs during aging.

Biological Time Keeping and Circadian Rhythms

Biological time keeping [300] and circadian variations are known to influence numbers of HSCs and HPCs in BM and blood, but how this might occur in old mice has not yet been explored. Such studies may be of interest, based on when it is best to collect HSCs and HPCs from older individuals from BM and mPB for use in HCT. Studies of circadian deep sequencing revealed stress-response genes that adapted to high rhythmic expression during aging [301], and daily onset of light and darkness have been reported to manifest control over HSC maintenance and differentiation [302]. Moreover, circadian host and microbiome interactions have been suggested to play relevant roles [303]. There is a noted sexual dimorphism in body clock regulation [304] which has not been adequately addressed, especially in context of aging, and aged HSCs and HPCs.

HSC homing to the BM plays an important role in the engrafting capability of the HSCs [305], and more consideration needs to be given to this capacity of aged HSCs. There are a number of means to enhance the homing of HSCs for enhanced engraftment [306,307,308,309], but these have not yet been evaluated in context of HSCs from old mice, or when cells are collected/processed in hypoxia vs. ambient air. Notably, there are a number of animal models to study aging, with an example being the Ames hypopituitary dwarf mice [7,8,9,10], that live approximately two times longer than that of most other mice. Imbalanced myelopoiesis was noted in myelopoiesis between BM and spleen in Ames dwarf mice [310].

I) COVID-19/SARS-CoV-2, Aging, and Hematopoiesis

Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) is the pathogen responsible for causing Coronavirus Disease 2019 (COVID-19), a disease which has spread worldwide, infected over 32 million people and claimed the lives of nearly 1 million people as of September 2020 [311]. SARS-CoV-2 is a virus which infects cells by binding to cell surface proteins via its Spike protein that extends out from the viral envelope, facilitating entry to the host cell and allowing for viral replication within the cell [312,313,314]. The most well-studied presentation of COVID-19 is an infection of the lungs, with symptoms ranging from a mild cough and fever to a severe pneumonia [315, 316].

The prognosis and severity of COVID-19 in patients appears strongly tied to the age of the patient [317]. The risk for SARS-CoV-2 infection leading to symptomatic disease rises dramatically with age. According to the Center for Disease Control (CDC), as age increases the likelihood of being hospitalized for COVID-19 increases, with adults over 65 having 5-13x higher hospitalization rate [318]. Older people are, as presently known, also more likely to die from the disease, with a 90-630x higher death rate in patients over 65 [318], although it is likely that the full story on this is not yet known. This is not likely due to older people being more susceptible to infection, as RT-PCR tests for the presence of SARS-CoV-2 in mild to moderate cases of COVID-19 demonstrate that the host cells of younger patients contain more viral RNA than older patients [319]. A potential interpretation of this is that a higher viral load is necessary for younger people to display symptoms compared to older people. This suggests that children are equally likely to be infected, but may be less likely to display symptoms of the disease, likely due to a general lack of additional contributing factors, possibly including a predisposition to hematologic disorders.

Understandably, many studies of SARS-CoV-2 have focused on its infection of and impact on the lungs. However, it has become increasingly evident that COVID-19 is systemic in nature [320], affecting many different systems including primitive and mature hematopoietic cells [317, 321,322,323,324]. The impact that the disease has on the hematopoietic system is evident in the hematological manifestations of COVID-19. A review looking at hematological factors in COVID-19 patients found that both lymphocytopenia and thrombocytopenia are common symptoms with hospitalized patients [322]. It is also evident that more severe cases of COVID-19 were more frequently associated with these hematological factors [322]. Further, one of the more devastating effects of SARS-CoV-2 infection is the induction of a "cytokine storm" [325, 326]. Cytokine storm refers to a toxic excessive release of immune cytokines leading to an autoimmune response. Thus, there is a strong need to address the mechanisms and effects of SARS-CoV-2 exposure on primitive and mature hematopoietic cells and the impact it may have for COVID-19 patients, especially in older individuals.

Recently it was demonstrated that ACE2 is expressed on the cell surface of small numbers of HPCs and mature immune cells [324] and moderate to large numbers of HSCs [323, 324]. Importantly, exposure of human HSCs to SARS-CoV-2 Spike protein induces increases in expression of inflammatory molecules such as NLRP3 and IL-1beta [323], indicating that the cytokine storm may be mediated in part through primitive hematopoietic cells. Further, human HSCs and HPCs exposed to Spike protein ex vivo have decreased capacity for functional HPC colony formation, exhibit decreased cell growth, and decreased expansion of HSCs, HPCs, and functional HPC colony forming units compared to cells that were unexposed to the viral protein [324]. These effects can be neutralized by co-treatment with Angiotensin1-7 [323, 324], a peptide linked to ACE2 regulation of hypertension [327]. The effects of the SARS-CoV-2 Spike protein on colony forming capacity and expansion can also be neutralized by treatment with an antibody targeting SARS-CoV-2 Spike protein or by treatment with soluble human ACE2 [324]. Human peripheral blood mature immune cells also exhibit a response to exposure to Spike protein ex vivo, with monocytes upregulating CD14 and undergoing aberrant changes in morphology [324]. It is clear that SARS-CoV-2 does not have to infect HSCs and HPCs to cause some of the above noted effects of the SARS-CoV-2 Spike protein [323, 324]. These, and that yet to be reported, data are important because they may help to explain the origin of hematological manifestations of COVID-19 such as lymphocytopenia and thrombocytopenia and provide insight into neutralizing these effects on the hematopoietic system.

It is very possible that one of the contributing factors to COVID-19 disease severity in the aged population is due to the impact of the disease on a hematopoietic system that has already been accumulating alterations and damage for many years. Additionally, effects of cytokine storm associated inflammation may further damage aged hematopoietic cells, possibly making them even more vulnerable to the development of hematological disorders even after recovery from COVID-19. The relationships between hematologic manifestations of COVID-19 and age should be further studied, as should the effects of SARS-CoV-2 exposure on HSC/HPC and mature immune cells from the aged versus young. It will also be important to determine whether the hematological manifestations of COVID-19 may be neutralized by specifically targeting the effects on the hematopoietic system, thus potentially relieving some of the disease burden on more severely affected patients, including older patients.

J) Conclusions in Context of Potential Future Intervention for Better Health of the Hematopoietic System During Aging

A number of studies have reviewed the aging process in general including protein sequestration at the nuclear periphery, and pathways of cellular proteostasis, the effects of aging on stem cell populations, and potential therapeutic interventions including that for aged HSCs [328,329,330,331,332,333,334]. Organoids have been suggested as experimental means to study the process of cellular aging [335], but much more rigorous work is needed in this area, especially with analysis ex vivo of HSCs in a relevant physioxia microenvironment BM niche model.

It has been noted that there are molecules in aged blood that promote the spread of cancer [336] and the accumulation of methylmalonic acid promotes tumor progression in the aged [337]. How these phenomenon relate to HSC, HPC, and hematopoiesis and pre-leukemia and leukemia, and to effects on cells collected in hypoxia/physioxia remains to be determined. A recent book [338] has described the aging process from the perspective of a long-time investigator in this field and noted the mTOR, AMPK and sirtuin pathways as main longevity signaling pathways.

Other considerations in context of aging and hematopoiesis to be elucidated are: how mitochondrial ROS acts as a double-edged conundrum for that of host defense in contrast to infection associated pathological inflammation [339] and its effects on HSC and HPC [340]. More in-depth insights can be gained from approaches incorporating single cell multiomics [341] and what role the mechanoregulation [342] of hematopoiesis might play during aging and disorders associated with aging. There is also the question of a role for mitochondrial transfer from Cx43-expressing HPC to stroma [343] during aging of the hematopoietic system and its interaction with the bone marrow microenvironment during regeneration. As noted in more recent reviews on CHIP [344,345,346], there is still much we do not understand about this phenomenon and its relationship to aging and aged hematopoiesis.

Cytokines, Chemokines and Intracellular Signaling

HSCs, HPCs, and hematopoiesis are regulated by numerous interacting cytokines and chemokines [347], effects mediated by receptor-induced intracellular signaling (348; Broxmeyer; submitted solicited review on Cytokines/Chemokines/Other Growth Regulators and their Receptors, 8th Edition, Hematology: Basic Principles and Practice, Eds. R. Hoffman, et. al., 2020). Some recent intracellular players involved in HSC and HPC function have been reported [307,308,309, 349, 350]. However, all such intracellular signaling events have been carried out with HSCs, HPCs, and immune cells of young mice, or human CB or human BM or mPB from younger individuals. Whether or not such regulatory intracellular signaling is similar in cells from old vs. the young remains to be determined, and such studies need to be assessed with purified populations of HSCs and HPCs, and in context of such cells isolated and processed under physioxia conditions, so that they are not exposed to ambient air oxygen which will likely modify the signaling events.

Dipeptidylpeptidase (DPP)4

We had noted that it may be feasible to use HSCs from older individuals for hematopoietic cell transplantation (HCT) if the cells are collected so that they are not induced by EPHOSS to decrease HSC numbers [1] when stressed by ambient air collection [40,41,42], but are there means to decrease the acute GVHD associated with allogeneic HCT, which perhaps has the potential to be more aggressive when donor cells are from more aged individuals? Inhibition of the enzyme dipeptidylpeptidase (DPP)4/CD26 has been shown to enhance mouse BM HCT and to accelerate recovery after radiation and selected chemotherapeutic drugs [351, 352] and to enhance time to engraftment of cord blood (CB) HCT [353,354,355]. This enhancement in time to neutrophil engraftment was also associated with a decrease in the already low acute GVHD noted for CB HCT [356]. It is now clear that the orally active DPP4 inhibitor, sitagliptin used in the CB Trials also greatly dampens acute GVHD in the setting of clinical mPB HCT [357]. Hence, there may be a role for DPP4 inhibitors such as sitagliptin in context of aged hematopoiesis and HCT. DPP4 has also been implicated in exosomes from patients with acute myeloid leukemia [299]. More in depth information on DPP4 during the aging process is clearly warranted. There are many proteins that have purported DPP4-truncation sites [358, 359]. This is of relevance as DPP4-trucated proteins can have decreased activity and block the effects of the full length molecules [352, 360]. Hence, a better undertaking of DPP4 on hematopoiesis in the old, as well as young, could uncover additional means to enhance hematopoiesis during health, aging and disease.

Concluding Thoughts

There are still many unknowns, and still to be elucidated mechanistic insights to understand changes in hematopoiesis during aging. Enhanced genomics may provide additional clues to age-related hematopoiesis [361]. Some of these areas for further investigation are noted in Table 1. It will likely take a multidisciplinary approach to fully understand the causes of abnormalities associated with the aging process, not only of the hematopoietic system. This information might at least partially control what might not be the inevitable consequences of the time- and disease-induced aging process. Future studies of hematopoiesis and HSCs during the aging process that investigate these processes in context of their status in an in vivo/physioxia environment, and ex-vivo under conditions that more closely mimic their in vivo condition of O2 tension and with microenvironmental niche cell interactions [362] will bring us closer to better understanding aged HSCs and what their true functional capacities and abnormalities are. Only with this enhanced understanding can we truly know how this information can best be used and better modified, if necessary and possible, for health benefit.

Table 1 Suggested areas of investigation requiring more thorough and rigorous analysis to better understand HSC and HPC function and hematopoiesis during the aging process
Full size table