Mitochondria underlie different metabolism of hematopoietic stem and progenitor cells

It is becoming increasingly clear that hematopoietic stem cells (HSC) are a heterogeneous set of cells with a diversity of functional states despite (so far seemingly) immunophenotypically identical cells. Long-term (LT-) HSC, those endowed with long-lasting self-renewal, have been shown to differ from more differentiated cells in energy metabolism. LT-HSC reside in a hypoxic microenvironment1 in the bone marrow and preferentially use glycolysis to obtain energy.2 Adjusting oxygen bioavailability and transitions between aerobic and anaerobic metabolism might be one of the ways by which HSC niche cells, such as sub-endothelial nestin mesenchymal stem cells,3 regulate HSC behavior. Intriguingly, nestin mesenchymal stem cells are directly regulated by sympathetic nerve fibers, which control blood flow, oxygen availability and also circadian HSC traffic.4 The role of the microenvironment in regulating HSC metabolism is, therefore, an exciting area of research.

During glycolysis, glucose is converted into pyruvate to generate ATP. Pyruvate can be transformed into acetyl-CoA - later utilized in the mitochondria - by pyruvate dehydrogenase, which is inhibited by pyruvate dehydrogenase kinases (Pdk). The characteristic hypoxia of LT-HSC stabilizes the hypoxia-inducible factor HIF-1α, promoting glycolysis and leading to Pdk activation, pyruvate dehydrogenase inhibition and a reduced supply of acetyl-CoA to mitochondria. These kinases have therefore emerged as critical metabolic regulators of HSC function.5

In the mouse, LT-HSC contain fewer mitochondria than do more differentiated hematopoietic progenitors.2,6 In this issue of Haematologica, Romero-Moya et al. report a similar finding in human HSC-enriched cord blood CD34 cells.7 In their study, human cord blood CD34 cells were sorted according to mitochondrial content. Concordantly with the results previously reported in the mouse,2,6 CD34 cells with fewer mitochondria contained more immunophenotypically-defined (CD34 CD38) HSC.7 Since the authors used immunomagnetically-enriched CD34 cells, it would be interesting to know whether more purified human HSC follow the same trend. Human HSC can be virtually isolated at single-cell resolution as Lin CD34 CD38 CD45RA Thy1 Rho CD49f cells.8 It would also be interesting to know whether those HSC are the ones containing fewer mitochondria and favoring glycolysis. In the study by Romero-Moya et al., cord blood CD34 cells with fewer mitochondria generated slightly fewer colony-forming units in culture (a readout of hematopoietic progenitor activity), mostly of erythroid lineage, and reduced secondary colonies (reflecting more primitive hematopoietic progenitors). As expected, most of the differences between CD34 cells with more or fewer mitochondria were attenuated or disappeared when cells were cultured, presumably under 20% O₂7.

The gold standard assay to test HSC activity is long-term transplantation into mice. The kinetics of hematopoietic reconstitution in conditioned recipients is hierarchically organized in mice9 and humans,10 with more committed progenitors giving rise to blood cells soon after transplantation, while only LT-HSC can contribute to hematopoiesis in the long term. Romero-Moya et al. analyzed the bone marrow of immunodeficient mice 7 days after transplantation and surprisingly found a slight increase in human hematopoietic chimerism (mostly consisting of B cells) from CD34 cells with fewer mitochondria,7 which would suggest that this population might not only contain HSC but also committed precursor cells. Further studies, including purification of HSC and long-term transplantations, could determine whether human LT-HSC and/or other hematopoietic progenitors have relatively fewer mitochondria and a distinct metabolic profile. Controlling HSC metabolism indirectly, e.g. through supporting mesenchymal stem cells, might facilitate expansion of human cord blood HSC for transplantation.11

Mitochondria are the major source of reactive oxygen species (ROS), which are generated as a result of regular bioenergetic metabolism. ROS levels must be tightly regulated in the cells as they play a key role as second messengers, but are detrimental when their levels are too high. In fact, HSC are more sensitive to ROS exposure than are committed progenitors, and small increases in ROS can compromise HSC ‘stemness’ whereas high levels induce HSC apoptosis12 (Figure 1). Thus, the comparatively lesser mitochondrial content in murine2,6 and human7 HSC might contribute to protect the cells from mitochondrial damage and subsequent apoptosis driven by ROS overproduction. Indeed, different studies in the mouse have shown that HSC have lower ROS levels than more differentiated progenitors.13

In addition, other protective mechanisms have been proposed to actively reduce ROS levels in HSC, mainly through the stimulation of antioxidant defenses. The serine/threonine protein kinase Ataxia telangiectasia mutated (ATM) is essential to maintain low ROS levels in HSC.14 Following DNA damage, ATM is activated and phosphorylates several proteins responsible for DNA repair, cell cycle arrest or programmed cell death. HSC deficient in ATM exhibit increased ROS levels and reduced self-renewal.14 In addition, the Foxhead O (FoxO) subfamily of transcription factors regulates ATM expression and ROS levels in HSC, since deletion of FoxO1,3 and 4 resulted in increased ROS production and decreased HSC self-renewal.15

Polycomb proteins, and more specifically Bmi-1, are master epigenetic regulators of HSC self-renewal and fate. This gave rise to the idea of a stem cell signature that defines the identity of the HSC. The ability of Bmi-1 to maintain the ‘stemness’ of HSC relies in part on the silencing of one of its major targets, the locus encoding the p16 and p19 tumor suppressors, which promote cell senescence.16 Notably, Bmi-1 also regulates mitochondrial ROS production independently of that pathway, further linking ROS and HSC fate.17

Another potential protective mechanism against oxidative stress that has been less explored is represented by mitochondrial DNA (mtDNA) variants. In mice, frequent and non-pathological mtDNA variants determine differences in the performance of mitochondrial oxidative phosphorylation and ROS production.18 Some inherited mtDNA variants have been related to healthy aging in human centenarians,19 whereas a certain somatic mtDNA mutation accumulation in leukocytes was suggested to enhance immunity and promote longevity.20 Recently, CD34 cells from members of the same family were shown to share several unique mtDNA variants, but the overall age-related accumulation of mtDNA mutations in CD34 cells varied in different families. Future study of mtDNA variants could pave the way to discerning susceptibility to age-related HSC failure/dysfunction mediated by ROS.21

Figure 1.ROS produced during mitochondrial metabolism determine the fate of HSC. Quiescent LT-HSC possess protective mechanisms to prevent abnormally high ROS levels, including low mitochondrial content, preferential use of glycolysis and reinforced antioxidant defenses. Small changes in ROS content may affect HSC fate, stimulating these cells’ proliferation and differentiation, and reducing their self-renewal capacity. High levels of ROS may lead to either HSC apoptosis or transformation.

Abnormally high ROS levels can decrease HSC self-renewal through activation of the p38 mitogen-activated protein kinase (MAPK) pathway.22 In fact, treatment with a p38 inhibitor has been shown to restore normal function in HSC with a high ROS content.13,22 High ROS levels have also been associated with increased expression of the mammalian target of rapamycin (mTOR).13 In hypoxic HSC, mTOR is inhibited by its major negative regulator Tuberous sclerosis complex (TSC1). TSC1 mutations have been shown to increase mitochondrial biogenesis and ROS accumulation, while reducing HSC self-renewal.23 In addition, mTOR is negatively regulated by the AMP-activated protein kinase (AMPK). When nutrients are low, AMPK is activated by the tumour suppressor Lkb1, leading to mTOR repression and reduced cell proliferation. Therefore, multiple pathways regulating ROS production and/or directly affected by ROS can control HSC energy metabolism and function.

Recent studies have shown that Lkb1 regulates quiescence and energy metabolism in HSC but, unexpectedly, not through the AMPK-mTOR pathway but, instead, by largely unknown mechanisms.24–26 Deletion of Lkb124–26 or AMPK26 decreases mitochondrial biogenesis and energy production in HSC. This seems to be mediated, at least partially, through the master regulator of mitochondrial biogenesis peroxisome proliferator-activated receptor gamma (PPARγ) coactivator protein 1α (PGC-1α).25 During stress hematopoiesis, PGC-1α is required to induce mitochondrial biogenesis and stimulate glucose uptake and rapid proliferation of HSC and hematopoietic progenitors, despite the relatively low oxygen availability.27 Altogether, these studies have coupled ROS levels and master regulators of energy metabolism to HSC function. Controlling HSC metabolism has, therefore, emerged as a promising avenue to stimulate engraftment28 and also to sensitize malignant cells.29 Nevertheless, important metabolic differences have been noted between normal and malignant HSC. Unlike normal HSC, leukemic stem cells and their derivatives have been proposed to contain a greater mitochondrial mass, have a higher O₂ consumption and be particularly sensitive to inhibition of mitochondrial translation.30 Future studies will determine whether these promising avenues for obtaining metabolic control of normal and malignant HSC are therapeutically valuable.

Footnotes

• Dr. Lorena Arranz is a Postdoctoral Associate in the Stem Cell Niche Pathophysiology Group led by Dr. Simón Méndez-Ferrer at CNIC (Madrid, Spain). Dr. Álvaro Urbano-Ispizúa is Professor of Medicine at the Department of Hematology, University of Barcelona, IDIBAPS and Institute of Research Josep Carreras; Director of the Institute of Hematology and Oncology at Clínic Hospital (Barcelona, Spain). His main field of interest is hematopoietic progenitor cell transplantation. Dr. Simón Méndez-Ferrer is Assistant Professor at the Department of Cardiovascular Development and Repair, Centro Nacional de Investigaciones Cardiovasculares Carlos III (CNIC, Madrid, Spain) and Adjunct Assistant Professor at the Department of Medicine, Icahn School of Medicine at Mount Sinai (New York, USA). His main field of interest is stem cell niche regulation.
• Financial and other disclosures provided by the author using the ICMJE (www.icmje.org) Uniform Format for Disclosure of Competing Interests are available with the full text of this paper at www.haematologica.org.