Acute myeloid leukemia (AML) is used to describe a cluster of neoplastic disorders characterized by the clonal expansion of immature blood cells of the myeloid lineage in the bone marrow (BM), blood or in other tissue , due to increased cell proliferation, survival and/or a block in the ability of the hematopoietic progenitor cells to differentiate3. These progenitor cells include cells of the granulocytic, monocyte/macrophage, erythroid and megakaryocytic lineages.
Diagnosis and classification of AML is primarily made on the basis of morphology and cyotchemical analysis. Using the widely adopted French-American-British (FAB) classification system4, AML can be classified into 8 subtypes as determined based on morphology, cellularity, blast percentage and cytochemistry. These subtypes are distinguished based on both the degree of differentiation and cell lineage. Cytochemical stains, including myeloperoxidase, nonspecific esterase and sudan black B are used in conjunction with morphology in the identification of the subtypes. Table 1.1 depicts the various AML subytpes as classified under the FAB system and their morphological and cytochemical presentations.
More recently, in a new classification system from the World Health Organization (WHO), other recurring molecular parameters associated to the various AML as diagnosed by cytogenetics, molecular genetics and immnophenotyping are also considered in the classification of the different AML types.
Although each subtype of AML may differ vastly in their genetic backgrounds, a hallmark of all the AML types is the severe block of myeloid differentiation. Thus it is thought that aberrations involving key transcription factors and its associated co-activators and co-repressors which are essential for the differentiation process is a major driving force for AML pathogenesis.
Table 1: The French-American-British classification of Acute Myeloid Leukemia (BOB LÃ-WENBERG, M.D., JAMES R. DOWNING. M.D., AND ALAN BURNETT, M.D.: Acute Myeloid Leukemia. NEJM 341(14): 1051-1062, 1999)
1.2 Role of the Transcription Machinery in AML.
The transcription machinery plays a crucial role in the control of normal hematopoiesis by having a major influence in the differentiation of the various hematopoietic lineages7-10. In normal hematopoiesis, the hematopoietic stem cell (HSC) matures into more committed multi-potential progenitors and finally to specific cell types of the different lineages. This process of growth and maturation of hematopoietic cells is regulated during normal hematopoiesis through a balance between its capacity to self renew and proliferate versus lineage commitment and differentiation.
Regulation is achieved via the controlled expression or repression of certain genes that are involved in self renewal, proliferation and differentiation as well as cell survival. This control is exerted via the cell's transcription machinery and its associated cofactors which include the various co-activator and co-repressor proteins.
A number of members of the transcription machinery have been identified to be crucial in the commitment of HSCs to the lymphoid and erythroid lineages11-15 and in the recent years, those which are critical in the control of development of HSCs towards the myeloid lineage have also been identified and studied. Particularly, the recent studies on normal hematopoiesis and leukemogenesis have suggested that factors involved in transcription have a major influenence on these 2 processes16-18 ; especially when mutated or dysregulated, these factors become the driving force of AML pathogenesis.
The nuclear receptor co-repressor N-CoR is a 270 kDa protein which is a key component of multi-protein co-repressor complex involved in transcriptional control mediated by various transcriptional factors. It mediates gene repression by binding to unliganded nuclear receptors such as the retinoic acid and thyroid hormone receptors19 and consists of both the nuclear binding domains as well as multiple repressor domains (Figure 1.1).
N-CoR recruits histone deacetylase 3 (HDAC3) to the promoter regions of genes and promotes the deacetylation of histones at those regions, thus changing the conformation of the chromatin making them less accessible to transcription activators resulting in gene silencing20-23. Other than HDAC3, N-CoR also recruits other factors such as Transducin B-Like 1 (TBL1), the TBL1-related protein (TBLR1)24 and G Protein Pathway Suppressor 2 (GPS-2)25 which together mediates the repression by multiple nuclear receptors such as the unliganded thyroid hormone receptor(TR)26 as well as the retinoic acid receptor (RAR), the peroxisome-proliferator-activated receptors (PPARs) PPARÎ±, PPARÎ² (also known as PPARÎ´) and PPARÎ³, and the liver X receptors (LXRs) LXRÎ± and LXRÎ²2. Figure 1.2 depicts the two-interaction feed forward model for recruiting N-CoR complexes to chromatin by the unliganded TR via histone deacetylation to mediate gene repression proposed by Yoon et al.
Other than nuclear receptors, N-CoR is also found to interact with the mammalian switch-independent 3 protein (mSin3)29-31. It is thought that this interaction is involved in the repression of several non-receptor transcription factors such as Mad/Max31.
Figure 1.1 Schematic of the N-CoR corepressor. The primary structure of the human N-CoR from N- to C-terminus [Horlein et al., 1995; Ordentlich et al., 1995]. Codon numbering is indicated on top. The locations of the repression domains (RD1 to RD3), the deacetylase activating domain (DAD), and the conserved SANT motifs that include sites of histone interaction and of the CoRNR box /nuclear receptor interaction sites (N1, N2, and N3 in N-CoR) are indicated. Interaction sites for transcription factors that utilize N-CoR for repression are indicated in yellow, whereas interaction sites for additional components of the corepressor complex or the general transcriptional machinery are shown in red. The actual sites of contact may be smaller than the experimentally mapped domains shown, and not all interacting proteins have been proven to interact with both N-CoR and SMRT.
(Nuclear Receptor Signaling (2005) 3, Michael Goodson, Brian A. Jonas and Martin A. Privalsky et al, Figure 1)
Figure 1.2 A two-interaction, feed-forward model for targeting SMRT/N-CoR complexes to chromatin by unliganded TR. As TR/RXR heterodimers, unliganded TR binds constitutively to its target genes in chromatin and unstably recruits SMRT/N-CoR. This unstable recruitment initiates limited histone deacetylation through associated HDAC3. Histone deacetylation generates limited hypoacetylated histone tails (H2B and H4) and allows TBL1/TBLR1 to bind. The binding of TBL1/TBLR1 to hypoacetylated histones H2B and H4 stabilizes the recruitment of SMRT/N-CoR complexes by unliganded TR, and the stable recruitment of SMRT/N-CoR complexes in turn leads to further deacetylation and finally transcriptional repression26.
(Molecular and Cellular Biology (2005) 25, Ho-Geun Yoon, Youngsok Choi, Philip A. Cole, and Jiemin Wong, Figure 9)
126.96.36.199 N-CoR in Normal Development.
Physiologically, N-CoR is important in many developmental processes such as proliferation, differentiation and apoptosis and its importance is underscored by the fact that N-CoR knockout mouse models produced an embryonically lethal phenotype with severe anemia due to defects in definitive erythropioesis as well as defects in thymocyte and neural development20.
N-CoR has been reported to be important in neural stem cell differentiation to astrocytes and this was accompanied by the cytosolic export of N-CoR via phosphorylation due to PI3K/Akt1 kinase activity resulting in the concomitant loss of N-CoR nuclear function32. Recently, N-CoR's role in erythroid differentiation was also established, with it being important in the regulation of the heme biosynthesis enzyme 5-aminolevulinate synthase (ALA-S2) in K562 cells33. A role for N-CoR in the differentiation of pituitary cells34 as well as in myogenesis35 had also been cited, highlighting the critical role of N-CoR in regulating differentiation.
Other than the regulation of differentiation, a novel role for N-CoR in the regulation of circadian metabolic physiology was recently published. Genetic disruption of the N-CoR-HDAC3 interaction in mice resulted in the aberrant regulation of clock genes and caused abnormal circadian behavior. Loss of functional N-CoR-HDAC3 complex altered the oscillatory patterns of several metabolic genes, implying that the activation of HDAC3 by N-CoR is a key event in the epigenetic regulation of circadian and metabolic physiology36.
188.8.131.52 N-CoR in Carcinogenesis.
N-CoR being a key component of the transcriptional repression machinery; its expression and function is tightly regulated in normal development. Deregulations of N-CoR function due to changes in expression or aberrant nuclear export have been reported to contribute to carcinogenesis.
N-CoR has been shown to be important in the repression of the PI3K/Akt1 signaling pathway in thyrocytes and its observable loss of expression in thyroid cancer cells was thought to enhance the survival of these cells by activating the PI3K/Akt 1kinase signaling pathway37. Its role in the transcriptional control of POZ/zinc finger transcription factor BCL-6 was also reported to be essential for the survival of tumour cells in diffuse large B-cell lymphomas38. In glioblastoma multiforme (GBM), tumor cells with nuclear localization of Nâ€‘CoR demonstrates an undifÂferentiated phenotype, but are subject to astroglial differentiation upon exposure to agents promoting phosphorylation of Nâ€‘CoR and its subsequent translocation to the cytoplasm39. It was also observed that in colorectal cancer primary patient samples, N-CoR displayed aberrant cytosolic localization. This was due to its phosphorylation by IKKÎ± and this was thought to result in the derepression of genes which promote the proliferation and survival of these cancer cells40.
184.108.40.206.1 N-CoR in AML Pathogenesis.
Although N-CoR has been reported to contribute to the pathogenesis of various types of cancers, it is most widely implicated in the pathogenesis of AML, especially in AMLs involving the AML1-ETO fusion protein and in Acute Promyelocytic Leukemia (APL/ AML-M3). In AML1-ETO positive myeloid leukemias, it was observed that aberrant recruitment of the N-CoR-HDAC3 complex to ETO in the fusion oncogene repressed gene transcription and inhibited differentiation in hematopoietic precursors41.
In APL, N-CoR's role in the pathogenesis of the disease has been extensively studied. It is believed that in APL, N-CoR recruited by PML-RARÎ± acted as a repressor for the genes that respond to Retinoic Acid (RA) and these are also the genes that were essential for myeloid cell maturation. By dissociating N-CoR from PML-RARÎ± in RA treatment, the repression for these genes is lifted which leads to myeloid cell maturation and differentiation42-44. Recently a different role for N-CoR in the pathogenesis of APL was elucidated. It was shown that PML-RARÎ± mediated the accumulation of insoluble N-CoR in the endoplasmic reticulum (ER). This led to the induction of ER stress and subsequent activation of the unfolded protein response (UPR) resulting in differentiation arrest in APL cells; RA inhibited this ER stress by enhancing the solubility of N-CoR45. A similar mechanism for Genistein mediated differentiation and maturation of APL cells was also published46. Genistein was shown to improve the solubility of N-CoR in APL cells resulting in the differentiation and maturation of the APL cells. In another report, N-CoR misfolding and its subsequent cleavage by a protease was also shown to be a contributing factor to the pathogenesis of APL47. Here it was reported that N-CoR cleavage in APL cells reduced ER stress to levels below the threshold needed to evoke ER stress mediated apoptosis. This enabled APL cells to escape cell death due to the accumulation of the misfolded N-CoR.
1.3 Protein Misfolding and Disease.
The normal function of proteins is determined by its three dimensional structure which is acquired through the folding of the polypeptide chain encoded by the genome. Any changes in the polypeptide chain either via abnormal amino acid modifications or aberrant post-translational modifications may result in changes of the folding process causing protein misfolding48.
Traditionally, protein misfolding has been linked to the pathogenesis of neurological diseases such as Alzheimer's disease and Parkinson's disease49-51, however recent studies are beginning to reveal new evidence of the involvement of protein misfolding in a multitude of other disease such as lung diseases, eye tissue diseases54-59, skeletal muscle diseases60-64, epithelial tissue diseases65-67, Atherosclerosis68-70 as well as in cancers.
The basis of pathology of most of these conformational diseases is a cellular inability to degrade these misfolded proteins resulting in the formation of cytotoxic aggregates. In these diseases, pathology is predominantly determined by cell damage associated with the aggregation thus exhibiting what is considered a 'gain-of-function' pathology71. This group of disease includes the neurological disorders Alzheimer's, Parkinson's and Huntington's disease72-74.
In another group of disease, which includes cystic fibrosis75, phenylketonuria76 and fatty acid oxidation defects77, the misfolded proteins are degraded rapidly resulting in the 'loss-of-function' pathology related to the decrease in the steady amount of the protein.
1.3.1 Protein Misfolding in Carcinogenesis
In normal conditions, tumor suppressors and oncogenes play critical roles in the regulation of cell division and survival. Tumorigenesis often arises due to inhibition of the proper function of these proteins as a result of genetic or post-translational aberrations. In some cases, these misfolded tumor suppressors are inactivated resulting in a subsequent loss of function, while in certain cases, these mutated proteins may adopt an aberrant conformation which is regulated differently from its wild-type counterparts.
One of the most well studied misfolded protein dependent conformational loss of function of a tumor suppressor in carcinogenesis is the study involving p53. p53 is a tumor suppressor protein which is commonly found to be deregulated in multiple cancers due to genetic aberrations. It is known that mutations involving the core domain (p53C) are found in more than 50% of all cancers. The misfolded variant of p53 is inactive and is found to have a dominant negative effect on normal wild type p5381. The loss of functional p53 results in the accumulation of mutations in the genome due to the inability of the cell to effectively repair DNA lesions. This inactive conformational variant of p53 has been described to be aggregated in the nucleus and cytoplasm of multiple cancers such as neuroblastomas, retinoblastomas, breast cancers and colon cancers.
In recent years, more proteins whose loss of function due to a misfolded conformation resulted in the cancer phenotype have been identified. These include the WT1 zinc-finger transcription factor where the tumorigenic WT1 inactivation mutation (WT/AR) which is a result of improper splicing is thought to have a role in the development of Wilms' Tumor a pediatric cancer of the kidney82-84. The von Hippel Lindau (VHL) tumor suppressor is a 213 amino acid protein that assembles with elongins B and C to form the VBC complex85. When it fails to correctly fold as observed in multiple tumorigenic mutants of this protein86, its ability to form the VBC complex is compromised resulting in the development of tumors associated with VHL syndrome. Merlin is another protein which misfolding has been reported to result in a loss-of-function phenotype that leads to the development of tumors associated with Neurofibromatosis type II. A missense point mutation in the N-terminus of this 65 kDa protein disrupts the folded conformation thus affecting its ability to dimerize for proper function87.
1.3.2 Protein Misfolding in AML.
In the AML model, pathogenesis of the disease due to the misfolded conformation dependent loss-of-function of a tumor suppressor is most studied in Acute Promyelocytic Leukemia (APL) or AML of the M3 subtype under the FAB classification.
PML-RARÎ± the fusion oncogene associated with more than 95% of the known cases of APL. In the traditional proposed model of APL pathogenesis, it was thought that the fusion oncogene PML-RARÎ± brings about differentiation arrest via the induction of transcription inhibition of RA target genes. However recent studies have revealed a novel mechanism of APL pathogenesis brought about by the fusion oncogene PML-RARÎ±. New studies have demonstrated the direct role of PML-RARÎ± in the misfolding of the generic nuclear receptor co-repressor N-CoR. In these studies it was shown that when expressed with PML-RARÎ±, N-CoR displayed several characteristics of misfolding, including its accumulation in the ER as insoluble protein aggregates, aberrant post-translational modifications and destabilization. PML-RARÎ± which is a de novo misfolded protein by itself, apparently acted as a nucleating event to trigger the conformational re-arrangement of N-CoR resulting in its misfolding and destabilization90. This results in the loss of normal N-CoR function, contributing to the pathogenesis of APL.
1.4 The Cancer Stem Cell Hypothesis
In recent years there has been a growing body of evidence supporting the hypothesis that carcinogenesis is initiated and maintained by a population of cancer cells which share similar biological properties to normal adult stem cells. These cells are known as the Cancer Stem Cells (CSCs). There are currently 2 schools of thought as to where these CSCs have its origin. One theory proposes that cancers arise from the stem cells directly while the other proposes that progenitors of stem cells which normally undergo limited numbers of cell divisions aberrantly acquire the ability to self renew. In either case both theories propose that these cells have the capacity to survive long enough to become long lived targets for additional genetic lesions to occur, inducing the tumorigenic phenotype91.
However despite the differences in opinion on the origin of CSCs, these cells are defined by its stem cell-like properties of the ability to proliferate indefinitely and give rise to both more CSCs and progeny that ultimately differentiate into the different cell types in a tumor92. However, the main difference between normal and tumor cells is the loss of the ability and mechanisms to maintain normal cell growth checks usually occurring at the stem cell levels resulting in the loss of the ability to maintain normal cell numbers.
1.4.1 The Leukemic Stem Cell.
With the advent of technological advancements, especially with the invention of multi-parametric Flow Cytometry, the Weissman laboratory had successfully identified an ordered sequence of phenotypically distinct hematopoietic stem cell and intermediate-precursor populations leading to the isolation and characterization of these populations in the hematopoietic system, using the unique combination of expression of certain cell surface markers for each of these populations93. This led to the subsequent elucidation of the leukemic stem cell through the ground breaking work of Bonnet and Dick94.
Leukemias can be viewed as aberrant hematopoietic processes initiated by the rare Leukemic Stem Cell (LSCs) which maintained or reacquired the capacity for indefinite proliferation. It has been proposed that LSCs may either be initiated by the transforming events which took place in the hematopoietic stem cell (HSC) compartment or it may also arise from the more committed progenitor cells owing to genetic or post-translational aberrations resulting in the enhancement of their otherwise limited self-renewal properties. In fact recent studies have demonstrated that both hypotheses were plausible and similar end stage leukemias were observed in mouse models regardless of the origin of the LSCs91.
220.127.116.11 Demonstration of the LSC.
It was previously demonstrated by Phillip Fialkow and colleagues using elegant experiments involving patterns of inactivation in X-linked genes that leukemias such as chronic myelogeneous leukemia (CML) and AML are clonal in origin. However it was not until the recent advances in the identification and separation of discrete tumor-cell populations and the availability of appropriate assays that the LSC was first demonstrated in 1997 by investigators based in the University of Toronto.
John Dick and colleagues first demonstrated that normal human stem cells as well as unfractionated AML, CML and Acute Lymphoblastic Leukemia (ALL) isolated from patients could engraft and proliferate in SCID mice and the progeny of these cells can be detected by Flow Cytometry. The leukemic cells recapitulated the human leukemias in the recipient mice while recipients of the normal stem cells were able to regenerate hematopoietic cells of all matured lineages with the exception of the T cells97-99. These data indicated that tumor cells which fulfilled the criteria to be called "stem cells" existed.
With the development of the NOD/SCID leukemia model which successfully eliminated limitations of the SCID mouse model and allowed for the separation of leukemic cells into subpopulations which could be evaluated for engraftment and serial-transplantation capacities, it was discovered that the ability to transfer human AML to recipient mice resided exclusively in the CD34+CD38- fraction. Furthermore, with the exception of APL/AML-M3, all subtypes of AML could be successfully induced. The secondary recipients developed AML which closely resembled that of the human donor demonstrating the long term self renewal capabilities of the LSCs. It also showed that the LSCs determined the stage of differentiation block during leukemogenesis100. Due to the phenotypic and functional similarities of the LSCs to normal HSCs, it was proposed that the HSC was most likely the target for transformation into the LSC.
18.104.22.168 Characterization of the LSC.
Phenotypic evidence points to the somatic stem cells as target cells for malignant transformation. Both the normal and tumor stem cells express identical levels of telomerase which is required to maintain telomere length and prevent replicative senescence101. As stem cells persist throughout adulthood and undergo increased numbers of cell divisions, they have a greater opportunity than that of the short-lived cells to acquire multiple mutations required for malignant transformation102-104.
The LSC population which generated the leukemic phenotype in the NOD/SCID mouse model had the CD34+CD38- phenotype similar to that of the somatic HSCs which induced multi lineage hematopoiesis when engrafted into the same mouse model99. Cytogenetic anomalies which are consistently associated with certain leukemias have also been detected in the HSC compartment in patients with AML,CML and ALL105-108. Furthermore, the BCR-ABL gene rearrangement most commonly associated with CML can be detected in cells of multiple lineages indicating that the initial transforming event occurred in a cell with the ability to differentiate into cell types of the different lineages such as the HSC.
However, committed progenitors cells which lack the capacity for self-renewal might also be transformed through the activation of certain oncogenic pathways which re-confer the properties of self renewal to such cells. Proof of this concept have recently been established in various mouse models for leukemia. In these models, purified populations of progenitor cells were transduced with either the MLL-ENL or MOZ-TIF2 oncogenes; both were originally cloned from human leukemia and generated leukemia on retroviral expression in mouse models. Transduction with either of the oncogenes resulted in the restoration of the self-renewal properties in the committed progenitors in in vitro assays. These transduced progenitors were also able to generate leukemia when transplanted into mice with the resulting leukemias being serially transplantable into secondary recipients. This transformation capacity was conferred solely by the oncogenes as cells transduced with the defective mutants of these oncogenes did not bring about the same phenomenon, plus self-renewal was not conferred by all leukemia associated oncogenes as transduction with BCR-ABL failed to bring about similar observations of reacquisition of self-renewal properties of the committed progenitors.
Evidence for progenitor cells as targets for malignant transformation has also been reported in human leukemias. Recent reports indicate that the organization of the LSC compartment following the progression of CML to blast crisis phase suggests a committed progenitor population might be the target for further oncogenic assaults that re-confer the properties for self-renewal. GMPs for example, purified from CML patients in blast crisis have self-renewal properties in vitro112, raising the possibility that ongoing acquisition of secondary mutations conferring self-renewal potential to progenitor cells could also be a contributing factor to the hierarchy of LSCs with have long-term self-renewal potential113.
22.214.171.124 AML : A Stem Cell Disorder.
Although AML is increasingly being recognized as a stem cell disorder, the true origin of LSCs in AML is still unclear. Whether LSCs in AML are initiated in the primitive stem cell compartment, or if the transforming event is a re-acquisition of stem cell like characteristics of committed cells is still a matter of debate. In the recent years, several elegant studies involving mouse models have suggested that APL/AML-M3 could arise in the committed progenitor cells114-116. Likewise, it was recently shown that chromosomal translocations involving the MLL1 gene locus especially when partnered with the transcription factor AF9 resulting in the fusion gene product MLL1-AF9 most commonly found in AML-M5, was able to impart stem cell like properties to committed progenitor cells and that leukemic cells in AML-M5 are initiated in the matured cells when these matured cells ectopically regain stem-cells like properties.