Fate through Fat: Lipid Metabolism Determines Stem Cell Division Outcome

Abstract

Distinctive metabolism associated with particular cell states is increasingly being defined in normal and malignant cells. Ito et al. (2012) now show that fatty acid oxidation is associated with hematopoietic stem cells and determines whether they undergo symmetric or asymmetric cell division, driving a fundamental property of the stem cell state.

Increased attention to the metabolism of malignant cells has raised the issue of just how different they are from their normal counterparts. This is particularly of interest in the context of cell functions such as self-renewal, shared by both malignant cells and normal stem cells. Whether the process of self-renewal has distinct metabolic requirements in cancer and normal cells is unclear, but is of potential consequence as therapies targeting cancer cell metabolism move toward the clinic. Work by the Pandolfi laboratory now adds a new topic for consideration in this rapidly evolving field. Ito et al. show that fatty acid metabolism is central to a signature feature of hematopoietic and other tissue stem cells, the “decision” between self-renewal and differentiation (Ito et al., 2012).

The metabolism of hematopoietic stem and progenitor cells (HSPCs) has generally been studied with a focus on the relative importance of glycolysis versus oxidative phosphorylation. HSPCs are known to depend upon HIF-1α, for example, and while this gene has many effects, its role in enforcing glycolysis is associated with stem cell maintenance in vivo. Disrupting HIF-1α reduces persistence of the cells upon serial transplantation, a process that depends on self-renewal (Takubo et al., 2010). HSPCs have a relatively low mitochondrial mass per cell, which is consistent with a preferential use of glycolysis over oxidative phosphorylation (Simsek et al., 2010). These characteristics may reflect a hypoxic environment in which HSPCs reside, but this remains controversial, and determinants other than oxygen availability can certainly influence the relative expression of glycolytic genes. For example, there are data indicating that expression of Cripto, a member of the EGF-CPC family, induces glycolytic enzyme expression (Miharada et al., 2011), and other studies suggest that Meis-1, necessary for HSPC generation, is responsible for HIF-1α expression in HSPCs (Simsek et al., 2010). Therefore, indirect evidence points to HSPCs preferentially undergoing glycolysis through enhanced expression of glycolytic pathway genes. It may be logical to consider glycolysis central to self-renewal. However, it has yet to be formally proven. Indeed, it is not clear that any particular metabolic pathway is central to stemness in hematopoiesis, although deleting metabolism-related genes such as LKB1, TSC, and FOXOs results in profound HSPC phenotypes (Gan et al., 2008, 2010; Gurumurthy et al., 2010; Nakada et al., 2010; Tothova et al., 2007). The work of Ito et al. now provides the missing direct evidence and does so by studying a pathway not previously associated with HSPC function.

Members of the peroxisome proliferator-activated receptor (PPAR) family of nuclear receptors are known to be nutrient sensors and potent transcriptional regulators of enzymes important in fatty acid transport and fatty acid oxidation (FAO). Manipulating the levels or activity of PPARδ, Ito et al. found that HSPC function was altered. The deletion of PPARδ led to poor HSPC self renewal in serial transplantation assays and, strikingly, a decrease in the ratio of asymmetric to symmetric divisions of hematopoietic stem cells. Conversely, pharmacologic activation of PPARδ had the opposite effect, promoting self-renewal and increasing asymmetric divisions. These changes in HSPC function were due to PPARδ modification of long-chain FAO, as the effects were recapitulated by pharmacologic inhibition of long-chain FAO (Figure 1). Of note, the effect of FAO inhibition on HSPC was persistent even with secondary transplantation, suggesting that more than just metabolite levels were involved.

Figure 1. Fatty Acid Oxidation Maintains Asymmetric HSC Division.

(A) The promyelocytic (PML) gene regulates PPARδ to increase fatty acid oxidation (FAO) in hematopoietic stem cells (HSCs) and biases them toward the fate of asymmetric cell division, leading to continued maintenance of the HSC pool.

(B) Potential mechanisms by which FAO maintains asymmetric self renewal include (1) shunting of long chain fatty acids away from lipid and cell membrane synthesis, which is not a major requirement for the slowly self renewing HSCs; (2) generation of ATP for HSC energy needs; (3) generation of reducing equivalents for quenching reactive oxygen species (ROS); and (4) generation of acetyl-CoA, chromatin modification, and preservation of the stem cell epigenome.

The hematopoietic stem cell population, like many tissue stem cells, is hypothesized to depend upon the ability of its daughter cells to undergo either self-renewal or differentiation with cell division. Symmetric division, where both daughter cells either differentiate or self-renew, is an inexorable path to stem cell exhaustion or tissue failure from lack of mature cell replenishment, respectively. Asymmetric stem cell division, where one daughter cell differentiates and the other self-renews, is therefore thought to preserve both the stem cell and mature cell pools. Shifting between asymmetric and symmetric division may enable stem cells to dynamically respond to changing physiologic needs. Ito et al. provide evidence that FAO is at the core of this fundamental hematopoietic property.

Tracking cell surface marker distribution in daughter cells of dividing HSPCs, the authors showed that inhibition of FAO enhanced symmetric, differentiating cell division, whereas pharmacologic activation of FAO increased asymmetric division. Genetic deletion of PML, an upstream inducer of PPARδ (and therefore FAO), increased the proportion of cells expressing a differentiation marker, suggesting an increase in symmetric, differentiating divisions in vivo. These data led to the conclusion that FAO is critical for stem cells to divide and have at least one daughter cell self-renew.

Why FAO is critical for self-renewal remains to be answered, but several possibilities can be considered. First, cells exiting the stem cell pool generally proliferate at a higher rate than stem cells. Perhaps reduced FAO positions daughter cells to be able to more readily generate the necessary biomass for rapid cell proliferation. Long-chain fatty acid metabolism first involves formation of fatty-acyl CoA in the outer mitochondrial membrane. This fatty-acyl CoA molecule can either contribute to the formation of lipids or can be transferred to the mitochondrial matrix to undergo β-oxidation, a process that generates acetyl-CoA en route to ATP. Decreased FAO may be permissive of greater lipid membrane generation. Why might stem cells be advantaged by increased FAO? In addition to the acetyl-CoA/ATP that is generated when FAO proceeds, there is also generation of NADH and FADH. NADH can serve as a proton donor for quenching ROS, which are maintained at particularly low levels in HSCs, presumably as a means of protecting against DNA damage (Jang and Sharkis, 2007). Finally, the acetyl-CoA molecules generated by FAO can be used as substrate for altering histone acetylation patterns responsible for epigenetic control of cell fate. Whether increased acetyl-CoA facilitates preservation of the stem cell epigenome is unclear, but may be one of the elements of the persistent change in stem cell function seen in the experiments of Ito et al. Unraveling whether and how metabolic parameters determine cell state is now open for interrogation in multiple pathways in both malignant and normal cells. The next generation of studies should tell us just how sugar, fat, and fate intermingle in the stem cell recipe.