Novel roles of reactive oxygen species in the pathogenesis of acute myeloid leukemia

Review of ROS in the pathogenesis of acute myeloid leukemia (AML), and the rationale for redox therapies.

Abstract

It has become apparent that regulation of ROS is important in cell signaling and homeostasis. Accumulation of ROS triggers oxidative stress in various cell types and contributes to the development, progression, and persistence of cancer. Recent research has demonstrated that redox dysregulation caused by ROS promotes proliferation, differentiation, genomic, and epigenetic alterations; immune evasion; and survival in leukemic cells. ROS act as signaling molecules to regulate redox-sensitive transcriptional factors, enzymes, oncogenes, and other downstream effectors. Thus, a thorough understanding the role of ROS as key mediators in leukemogenesis is likely to provide opportunities for improved pharmacological intervention. In this review, we summarize the recent findings that support a role for ROS in the pathogenesis of AML and outline innovative approaches in the implementation of redox therapies for myeloid malignancies.

Introduction

ROS are a family of chemically active molecules that contain free radicals and are involved in modulating biological cell functions, cell signaling, and homeostasis [1–3]. Cells' fate depends on the levels of ROS in the redox microenvironment. Low levels provide a beneficial effect, supporting cell proliferation and survival pathways. Sublethal levels of ROS are generated during environmental stress, resulting in DNA damage in normal or precancerous cells that promote malignant transformation. However, excessive ROS levels can cause detrimental oxidative stress that results in cell death [4, 5]. Frequently, an adequate balance between producing and eliminating ROS is maintained in cells via antioxidants or specific enzymatic pathways.

ROS have been termed mediators in intracellular signaling. Key to the renewal of this field was the finding that endogenous ROS levels are elevated in tumor cells [6]. Increased levels of steady-state ROS have been linked to numerous cellular processes that are associated with tumor development, such as malignant transformation, prosurvival pathway up-regulation, and DNA damage-induced mutations. ROS also modulate the expression of critical kinases, phosphatases, redox-sensitive transcription factors, and associated genes, suggesting that they promote malignant transformation. The role of ROS in leukemia was supported by the results of several studies [7–9].

AML is an aggressive hematopoietic malignancy that is characterized by highly proliferative blast cells and increasing incidence with age. Clinically, the disease is associated with hyperleukocytosis, extramedullary disease, and abnormal coagulation. Moreover, AML patients inevitably experience relapse, even after standard chemotherapy and stem cell transplantation. Most relapses occur in patients with a poor prognosis index [10, 11]. Current therapeutic strategies appear to have reached the limit of their effectiveness; thus, novel strategies are urgently needed to target leukemic cells exclusively. Several recent studies have demonstrated that altered antioxidant status is a distinct feature of AML, and in blasts isolated from patients with various types of leukemia, oxygen radical levels were significantly higher in malignant cells than in normal leukocytes [12–14]. We hypothesized that altered intracellular and extracellular ROS trigger acute myeloid leukemogenesis.

In this paper, we review the recent medical literature and address the role of intracellular and extracellular ROS in acute myeloid leukemogenesis. Opportunities for pharmacological intervention are identified to demonstrate how recent findings may facilitate the implementation of redox therapies for AML.

ROS REGULATION OF HSCs

ROS are produced as normal products during cellular metabolism and include superoxide anion, H₂O₂, hydroxyl radical, and NO. They are signal messengers that are generated from multiple sources, including the mitochondrial electron transport system, the NOX complex, and other cellular oxidant-generating systems [15].

ROS have been implicated as the chemical mediators in normal hematopoiesis. HSCs have a self-replicating potential that allows them to constantly replenish blood cells. Fluctuations of relatively low ROS levels control their self-renewal and proliferative capacity [16]. HSCs are found primarily in a quiescent state in the bone marrow stem cell niche [17]. However, HSCs with increased ROS levels undergo proliferation and differentiation after mobilizing to the more oxygen-rich bloodstream. Elevated ROS levels lead to HSCs changing their phenotype [16]. Numerous genes involved in DNA-damage responses or longevity-related signaling contribute to maintaining the self-renewal capacity of HSCs. Repression of intracellular ROS levels in HSCs, after treating with antioxidant ROS scavengers, rescues HSC functions, suggesting that excessive ROS contributes to the phenotypic change in HSCs [18].

Recent studies demonstrated that FoxOs are required for the long-term regenerative potential of the HSC compartment, as FoxOs regulate the response of HSCs to physiologic oxidative stress, quiescence, and survival [19]. Moreover, FoxOs affect HSC integrity by regulating ROS. In particular, a recent finding indicates that the FoxO-deficient HSC phenotype causes elevated ROS concentrations that are correlated with aberrant expression of ROS-regulating genes. Furthermore, daily treatment with the antioxidant N-acetyl-l-cysteine for 5 weeks reverses the FoxO-deficient HSC phenotype [20]. Similarly, germline loss of FoxO3 results in increased ROS levels and increased activation of p38 MAPK in the HSC compartment [21]. In brief, the forkhead transcription factors of the FoxO family participate in diverse physiologic processes by regulating ROS, including cell-cycle arrest, stress resistance, differentiation, apoptosis, and metabolism.

HSCs undergo self-renewing cell divisions and maintain blood production for life. Appropriate control of the self-renewal capacity of HSCs is critical for maintaining hematopoietic homeostasis. Hypoxia helps preserve the progenitor characteristics of HSCs and hematopoietic progenitor cells in the bone marrow. It effectively preserves the biological characteristics of CD³⁴⁺ cells by regulating NOX, thereby maintaining lower intracellular ROS levels [22]. Clearly, a low-oxygenic niche in the bone marrow limits ROS production to provide long-term protection for HSCs from ROS stress. For example, Mdm2 is an E3 ubiquitin ligase that targets p53 for degradation and is required to control ROS-induced p53 levels for sustainable hematopoiesis [23]. However, excessive oxygen is toxic to HSCs because of a state of unstable, dynamic ROS balance in the surrounding niche. In fact, elevated ROS levels also limit the lifespan of HSCs [24].

BMI1 has been reported as an oncogene in human carcinogenesis and is necessary for efficient self-renewing cell divisions in adult HSCs. Rizo et al. [25] demonstrated that BMI1 down-regulation impairs the long-term expansion and progenitor-forming capacity of cord blood CD³⁴⁺ cells in cytokine-driven liquid cultures and in bone marrow stromal cocultures. The Tsc1 acts as a tumor suppressor and is a key regulator of cellular metabolism. The studies by Chen et al. [26] also found that Tsc1 deletion in HSCs drives them from quiescence to rapid cycling, with increased mitochondrial biogenesis and elevated ROS levels, whereas in vivo treatment with a ROS antagonist restores HSC numbers and functions. The ATM gene modulates the cell cycle and is involved in oxidative defense. In ATM-deficient mice, increased levels of ROS induced HSC-specific phosphorylation of p38 MAPK, followed by a defect in maintaining HSC quiescence. p38 MAPK inhibition restored the ROS-defective HSC population capacity and HSC quiescence retention, suggesting that the ROS-p38 MAPK pathway contributes to HSC depletion [24]. Recent findings by Kocabas et al. [27] demonstrated that inducible Meis1 deletion in adult mice with increased ROS levels resulted in apoptosis, loss of HSC quiescence, and failure of bone marrow repopulation after transplantation. ROS act as signal molecules that can control HSC functions via gene regulation. Indeed, this deletion of genes reduced hematopoiesis and HSC self-renewal dramatically (Fig. 1).

Figure 1. ROS regulation of hematopoiesis via stem cell self-renewal.

HSCs have self-replicating potential that allows them to constantly replenish blood cells. Elevated ROS levels, present in normal hematopoiesis, lead to reduced self-renewal and hematopoiesis. Elevated ROS levels reduce the self-renewal of HSCs and inhibit hematopoiesis. Deletion of specific genes affects HSC integrity by modulating ROS levels. ^−/−, Deficient; RNAi, RNA interference.

ROS-RELATED LEUKEMOGENESIS

Acute leukemias are malignant clonal disorders of hematopoietic precursor cells, in which leukemic stem cells acquire mutations that confer self-renewal capabilities, altered hematopoietic differentiation, and increased proliferative capacity. ROS modulate the expression of redox-sensitive transcription factors, enzymes, oncogenes, and other downstream effectors [28]. Recent advances in our understanding of AML biology have contributed to better identify genetic and epigenetic abnormalities in leukemic cells and modulate ROS in leukemic hematopoiesis (Fig. 2).

Figure 2. Altered ROS levels stimulate leukemogenesis.

Oxidative stress has been implicated in the pathogenesis of AML. Aberrant ROS signaling triggers damage to nuclear and mitochondrial DNA and compromises DNA repair via increasing a cell's mutation rate and/or promoting and maintaining the oncogenic phenotype as a secondary messenger in intracellular signaling cascades.

Elevated ROS levels in AML cells

In normal mammalian hematopoietic systems, ROS remain at low levels in HSCs. However, common myeloid progenitors produce significantly higher levels of ROS than do normal leukocytes. Leukemic cells are intrinsically under oxidative stress and are therefore more vulnerable to further stress. Scavenging the ROS from hematopoietic progenitors using in vivo genetic tools impairs their differentiation into mature blood cells [29]. In contrast, increasing the hematopoietic progenitor ROS to above basal level triggers precocious differentiation of mature blood cell types in Drosophila, through the JNK signal pathway, FoxO activation, and polycomb down-regulation [29]. In the identification of genes associated with leukemic cell differentiation, Yang et al. [30] reported that ectopic overexpression of mda-7/IL-24 and the splice variant of mda-7/IL-24 lacking exon 5 (IL-24 delE5) significantly promotes ROS production beyond the threshold of tolerability in leukemic cells, resulting in differentiation of leukemic cells [30]. Through a ROS-dependent mechanism, oncoprotein mucin 1-C inhibition also induces a terminally differentiated myeloid phenotype in AML cell lines and primary blasts [31]. These new studies have significantly shifted our understanding of how cell differentiation is controlled via ROS signaling.

It is worth mentioning that the leukemic phenotype is correlated with elevated ROS levels. The serine-threonine kinase Akt is constitutively activated in most primary AML cells, followed by FoxO function inhibition and increased ROS levels [19, 32]. Juntilla et al. [33] found that intracellular ROS levels depend on Akt. Moreover, Akt1/2 double-deficient HSCs competed poorly against WT cells in the development of myeloid and lymphoid cells in in vivo reconstitution assays. Furthermore, constitutively activated Akt and FLT3 regulate leukemic cell survival and resistance to chemotherapy [34]. These results suggest that ROS produced in tumor cells alters the balance of genes expression, which may affect AML progression.

ROS-related genetic instability in AML

Major activated oncogenes, such as Bcr/Abl, Ras, and c-myc, have been found to correlate with enhanced ROS production [35, 36]. For example, Ras-induced ROS promote growth factor-independent proliferation in human CD³⁴⁺ hematopoietic progenitor cells [37]. ROS stress causes DSBs and altered repair, leading to genomic changes and carcinogenesis [38, 39]. Defects in nonhomologous end-joining, a main repair pathway for DSBs, lead to abnormal repair, resulting in chromosome deletions and translocations [40]. Genetic instability as a result of increased DNA damage and altered DNA repair contributes to the initiation and progression of inherited and sporadic human leukemias [41]. Furthermore, increased ROS generation enhances the regulatory effect from oncogenic proteins. Woolley et al. [42] found that FLT3 drives H₂O₂ production in AML cells via the NOX p22^phox. Moreover, knocking down p22^phox dramatically reduced H₂O₂ after 24 h in the endoplasmic reticulum. Activating mutations of the FLT3 receptor, seen in ∼30% of AML patients, are associated with the poor prognosis and aggressive clinical behavior of AML. FLT3 mutations, through ITD, cause increased ROS production in FLT3/ITD patients, leading to increased DNA DSBs and repair errors [7]. Alternatively, ROS molecules act as secondary messengers in intracellular signaling cascades, leading to direct DNA damage by increasing the mutation rate of cells. Rassool et al. [43] found that DNA damage and error-prone repair occur after increased ROS in the myeloid leukemic disease progression model. Furthermore, recent studies showed that endogenous levels of 8-hydroxy-2′-deoxyguanosine, an oxidized nucleoside of DNA, were higher in tumor tissues than in controls, suggesting that oxidative DNA damage contributes to AML development [12].

ROS and mitochondrial damage

Mitochondrial dysfunction and oxidative stress have been implicated in the pathophysiology of many diseases [44]. Variations in the mitochondrial basal capacity constitute an important and highly efficient channel for regulating ROS levels in a cell. Disturbances in and weakening of respiratory function in mitochondria further enhance the initial pro-oxidative state of cells. Mutations in mtDNA occur at a high frequency in human tumors and are associated with ROS overproduction [45]. Enhanced ROS not only cause damage to nuclear and mtDNA but also compromise DNA repair. Mitochondria have a limited DNA repair system, and their genome instability was demonstrated recently in AML [46].

Mitochondrial signaling cascades have been implicated in the regulation of cell proliferation and death via multiple pathways [47]. Sirt3 is a mitochondrial tumor-suppressor gene that connects aberrant cellular ROS and carcinogenesis. Murine tumors that lack Sirt3 exhibit abnormally high levels of ROS that directly induce genomic instability and increase HIF-1α protein levels. The subsequent transcription of HIF-1α-targeted genes mediates cellular metabolic reprogramming and increases cellular glucose consumption [48]. In short, it promotes crucial biologic processes in terms of mitochondrial ROS signaling.

Cytokines contribute to myeloid cell growth via ROS

Various tumor-derived factors alter myeloid cell differentiation and maturation via surface receptors and signaling pathways. Recently, several cytokines and growth factors were found to induce ROS production, which may alter the balance of expression of different genes. G-CSF stimulation increases ROS production in a time- and dose-dependent manner and contributes to myeloid cell growth, correlating with activation of Lyn and Akt. ROS production was enhanced further in bone marrow-derived neutrophils that express G-CSFRδ715, a truncated receptor [49]. ROS production induced by IL-3 can protect leukemic cells from apoptosis, but the effect can be counteracted by antioxidants, such as the combined SOD/catalase mimetic, ebselen (a seleno-organic compound and radical scavenger), and hydroxylamine probe. The function of IL-3 and exogenous H₂O₂ on cell viability seems to be mediated by inhibition of apoptosis [50].

REDOX-DEPENDENT REGULATION OF CELL SIGNALING

Cells maintain a network of sensory mechanisms, signaling pathways, and response systems to cope with environmental challenges, such as hypoxia and inflammation. ROS may regulate signaling pathways specifically, by altering the activity of direct-target molecules. Recent studies have begun to shed light on ROS-dependent regulation of cell signaling and ROS homeostasis in AML.

ROS generated by NOX as a redox messenger

ROS appear to be essential for several downstream signaling events. Piccoli et al. [22] described that ROS generated by NOX are redox messengers that control HSC proliferation and differentiation. ROS production appears to contribute to the proliferation and migration of hematopoietic cells that express a variety of oncogenic tyrosine kinases. NOX is a multiprotein complex that generates ROS in response to a wide range of stimuli. Recent studies by Reddy et al. [9] reported that NOX regulate cell growth and migration in myeloid cells transformed by oncogenic tyrosine kinases. With the use of targeted approaches against components of the superoxide-producing NOX, myeloid cells showed slow growth and spontaneous migration. Activated Ras strongly up-regulated ROS production by stimulating NOX activity, and excessive ROS production in the context of Ras activation promoted proliferative responses in normal hematopoietic cells [37]. Thus, the combination of a NOX inhibitor and the T cell-derived cytokine IL-2 is an effective relapse-prevention strategy in AML [51, 52].

ROS scavenger as a redox regulator

Trx is a redox-active protein with a molecular mass of 13 Kda, and it is widely distributed in tissues and organs. Reduced Trx has oxygen radical-scavenging and protein-refolding activities in vitro. Inhibition of the Trx system may induce antileukemic effects [53, 54]. Trx enhances the DNA-binding activity of the AP-1 transcription factor via an activation cascade—Trx-redox factor-1-AP-1—and by direct cysteine-mediated associations and reduction processes [55, 56]. To maintain homeostasis, TXNIP Trx-binding protein 2, the endogenous inhibitor of Trx, binds directly to the protein, resulting in decreased antioxidant activity [57]. Deregulated TXNIP expression contributes to experimental murine leukemia and human AML [58]. Furthermore, 3-deazaneplanocin A, a histone methyltransferase inhibitor, targets leukemic cells in AML by up-regulating TXNIP and increases ROS production [59].

Prx, a family of small nonseleno peroxidases, are capable of protecting cells from ROS insult and regulating AML signal transduction pathways. Godfrey et al. [60] reported recently that cell transformation by FLT3 in AML involves oxidative inactivation of the tumor suppressor PTP, DEP-1/PTPRJ. Decreased ROS result in DEP-1 reactivation in FLT3 ITD–expressing cell lines by inhibiting NOX or overexpressing catalase or Prx-1. Prx2 is an epigenetically silenced tumor-suppressor gene, and low expression is clinically associated with poor prognosis in AML patients. Agrawal-Singh et al. [61] found recently that Prx-2 acted as an inhibitor of myeloid cell growth by reducing levels of ROS generated in response to cytokines. Furthermore, forced Prx-2 expression inhibited c-Myc-induced leukemogenes in vivo in bone marrow-transplanted mice. These studies support the concept that ROS scavenger regulates cell signaling to affect pathological responses in redox microenvironment.

Transcription factors controlling gene expression in AML

Several transcription factors, such as Nrf2, Bach1, NF-κB, AP-1, and HIF-1α, regulate the expression of genes that are critically involved in AML pathogenesis in response to redox alterations. In general, transcription factors activate the transcription of genes involved in diverse aspects of cellular and integrative physiology, including energy metabolism, cell growth, survival, invasion, migration, and angiogenesis.

Nrf2 is a positive regulator of myeloid differentiation in AML cells [62]. Indeed, Nrf2 activity is up-regulated in several leukemia types and contributes to leukemogenesis [63]. An essential role of Nrf2 is transactivating antioxidants, such as HO-1, Trx, and enzymes in response to oxidative stress [64]. Rushworth and MacEwan [65] found that Nrf2 plays a key role in up-regulating HO-1 expression and ultimately protects AML cells from TNF-mediated cell death, resulting in clinically devastating levels of apoptosis resistance. Interplay between Bach1 and Nrf2 is apparent in AML cell lines. ROS cause the oxidation of cysteine residues in the Bach1 transcriptional repressor, thereby allowing the dissociation of Bach1 from the antioxidant response element, its translocation to the cytosol, and its degradation [66]. NF-κB (p50/p65 subunits), a ubiquitous transcription factor, is sequestered in an inactive form in the cytoplasm, as a heterotrimeric complex bound to its inhibitory IκBα protein. Upon phosphorylation of the IκBα, NF-κB dissociated from the complex, and the activated NF-κB is then translocated into the nucleus to activate specific target gens. Interestingly, the transcription factor NF-κB (p50/p65) suppressed HO-1 levels in human AML, and chemical inhibition of NF-κB resulted in elevated HO-1 protein levels in THP-1 and HL-60 cells [67]. Miyazaki et al. [68] also found that Bach1 regulated HO-1 gene expression in AML cells, and its low expression enhanced cell survival by up-regulating HO-1 expression. Mullican et al. [69] reported that neuron-derived orphan receptor 1 and orphan nuclear receptor TR3, critical tumor suppressors of myeloid leukemogenesis, promote the expression of the AP-1 transcription factors JunB and c-Jun and inhibit the development of abnormal HSCs and myeloid progenitors.

HIF-1α is also a redox-responsive transcription factor that regulates the expression of numerous genes under physiological and pathological conditions [70]. Interestingly, HIF-1α-induced differentiation of myeloid leukemic cells is independent of the transcriptional activity of HIF-1α. Moreover, HIF-1α increased the DNA-binding ability and hematopoietic transcriptional activity of Runt-related transcription factor 1 (also known as AML-1) [71]. The results of these studies strongly suggest that specific transcription factors play a dominant role in AML pathogenesis, making them redox-sensitive targets.

IMMUNE EVASION VIA ROS

NK cells and T cells play a major role in AML immune surveillance [72]. They both possess efficient immune-modulatory and cytotoxic properties that are essential for recognizing and attacking AML cells. However, multiple deficiencies of T and NK cell functions that confer relapse risk and poor prognosis have been observed in patients with AML. Shifts in the redox balance between ROS and antioxidants regulate immunity at various levels. Tumor-associated myeloid cells are the major type of inflammatory cells involved in regulating antitumor immune responses. Although similar ROS-mediated immunosuppression has been reported in nonmalignant myeloid cells, malignant AML cells use ROS to evade antileukemic effector lymphocytes. Aurelius et al. [73] proposed a novel mechanism, whereby mature, malignant cells, in monocytic forms of leukemia, can produce ROS via the NOX and thus, trigger poly(ADP-ribose) polymerase 1-dependent apoptosis in adjacent NK cells, CD4⁺ T cells, and CD8⁺ T cells. In vitro evidence suggests that the generation and release of ROS produced by monocytes and macrophages abrogate cytokine activation of NK and T cells and induce apoptosis of these cells in vitro. Thus, it is probable that inhibition of ROS by histamine will allow cytokines to activate NK and T cells and restore their antineoplastic capabilities [74]. ROS are essential components of the innate immune response [75]. Malignant myeloid cells produce ROS as an immune-evasion strategy to avoid destruction by antileukemic lymphocytes, such as NK cells and CTLs. The proposed model of how the ROS pathway contributes to immune evasion is illustrated in Fig. 3.

Figure 3. Immune evasion via ROS.

(A) NK cells are potential effectors of innate immunity that are regulated by efficient immunomodulatory and cytotoxic mechanisms that are essential for eliminating AML cells. (B) Elevated ROS levels trigger apoptosis in NK cells, CD4⁺ T cells, and CD8⁺ T cells, whereas AML cells use ROS signaling as a strategy to evade cell-mediated immunity.

CONCLUDING REMARKS

Accumulating data suggest that ROS are involved in myeloid leukemogenesis. The action of ROS signaling, including intracellular redox state alterations and oxidative protein modifications, is critical in the AML process. Abnormal ROS levels are characteristic of AML and provide a platform for the development of antileukemic drugs. Key to the development of any successful therapy is an understanding of disease etiology and progression [76]. Thus, targeting ROS as an antileukemic strategy is expected to provide a favorable clinical benefit for AML patients.

AUTHORSHIP

All authors wrote the paper.

DISCLOSURES

The authors declare no conflict of interest.