RICERCA ITALIANA     

Myelodysplastic syndromes: pathogenetic models and promise of new therapies

Università Cattolica del Sacro Cuore

Abstract

BACKGROUND
Myelodysplastic syndrome (MDS) are a heterogeneous group of clonal haematopoietic disorders, which exhibit ineffective haematopoiesis leading to pancytopenia and with an increased risk (30%) of transformation to acute myeloid leukaemia (AML). Expansion of the abnormal clone is characterized by morphological dysplasia, impaired differentiation, defective cellular functions, and genetic instability. Clonal cytogenetic abnormalities are demonstrable in 40%-50% of patients with primary MDS and in up to 80% of patients with secondary or treatment-induced MDS. The specific chromosome affected and the number of chromosomal abnormalities (i.e., complexity) offer powerful prognostic information. Evolution to AML, however, may be influenced by epigenetic events. For example, silencing of the proto-oncogene p15INK4b by methylation is detected in >70% of patients experiencing AML evolution.
Preclinical investigations indicate that reciprocal interactions between the malignant clone and the microenvironment serve to create a hostile milieu that reinforces ineffective blood cell production. Ineffective hematopoiesis, the hallmark of MDS, arises from impaired progenitor responsiveness to normal trophic signals and excess local generation of inhibitory cytokines, which promote accelerated apoptosis of progenitors and their progeny. Evidence supporting this model derives from cytokine neutralization studies and the direct relationship between plasma tumor necrosis factor concentration and apoptosis via intracellulary oxigen free radical production and DNA oxidation of malignant CD34+ progenitors. Recent investigations indicate that angiogenic molecules generated by malignant myelomonocytic precursors represent integral diffusible signals that reinforce leukemia progenitor self-renewal, while promoting the generation of proapoptotic cytokines and medullary angiogenic response. Delineation of such biologic features that are central for the pathobiology of MDS provides a reliable framework for the development of novel therapeutics, including demethylating agents, anti-angiogenesis and anti-TNF drugs.

OBJECTIVES
1) GENETIC STUDIES
a. Identification of new genes and molecular mechanisms underlying myelodysplastic syndromes and other related myeloid neoplasm, using conventional cytogenetics, interphase and metaphase FISH, comparative genomic ibridization and microarray technology;
b. Cytogenetic and molecular characterization of 3q21 and 3q26 break point regions with analysis of EVI1, MEL1, RPN1 genes and of the fusion genes RPN1-EVI 1 and MEL1-RPN1;
c.Molecular analysis of tyrosine-kinase mutations (FLT3, PDGFR, VEGFR e AKT).
2) EPIGENETIC STUDIES
a. methylation pattern of the promoters of some DNA-repair, apoptosis and angiogenesis regulatory genes at basal condition and after therapy with hypomethilating drugs;
b.DNA histone acetylation pattern of CD34+ cells collected from MDS patients;
c. modulation of the transcriptional activity induced by ATRA and deacetylase inhibitors on cell lines and CD34+ cells isolated from patients with MDS;
d. Relationship between genetic and epigenetic alterations (EVI1 expression and deacetylase activity.
3. Immunophenotyping of MDS to evaluate:
a.the pattern of expression of adhesion molecules (Lecam1, ICAM1, VLA-4, VLA-5, Thy1-CD90) and differentiation antigens(CD45 CD34, HLA-DR, CD13, CD33, CD15, CD11b, CD16) in bone marrow samples of patients with MDS, compared to that of patients with acute myeloid leukemia or normal donors;
b. abnormalities of apoptosis, by testing the expression of apo/fas (CD95), bax, bcl-2, annexin V and caspase-3;
4. Monitoring the biologic profile of the disease during therapy with hypomethylating agents and acetylating agents;
5. Specific associations between genomic rearrangements and morphology and/or immunology, according to WHO criteria.
6. Therapeutic role of inhibitors of tyrosin-kinase,TNF alfa and angiogenesis . <<<

Principal Investigator

Giuseppe LEONE Università Cattolica del Sacro Cuore

Research Objectives

Aim of the project is to identify specific biological profiles of MDS patients, classified according to WHO criteria, at diagnosis and after treatment with drugs able to revert epigenetic changes. MDS subtypes will be identified and characterized by performing genetic and epigenetic studies, which will be correlated to immunophenotype and morphology.

a) TASKS OF GENETIC STUDIES
1) BIOLOGICAL DEFINITION OF MYELODYSPLASTIC SYNDROMES WITH MISSING GENETIC LABELS IN THE WHO CLASSIFICATION.
This part of the project has been conceived to discover new cytogenetic/molecular markers to define new entities within the WHO classification and to identify new and/or cryptic genomic lesions either in cases with karyotypic anomalies or in normal and failed cases at conventional cytogenetics.

2) IDENTIFICATION OF NEW GENES AND MOLECULAR MECHANISMS UNDERLYING MYELODYSPLASTIC SYNDROMES AND OTHER RELATED MYELOID NEOPLASMS.
Comparative Genomic Hybridization (CGH) and metaphase FISH in MDS with complex karyotype will be performed. Molecular investigations will include the characterization of genes DUP16 (a new oncosuppressor gene at 12p13, encoding for a MAPK phosphatase)and BUP98.

3)MATRIX-CGH AND GENE EXPRESSION ANALYSIS ON CHROMOSOMES AND ON MICROARRAYS. Matrix-CGH is helpful to increase the resolution power of CGH in metaphase.

4)CYTOGENETIC AND MOLECULAR CHARACTERIZATION OF MDS WITH INVOLVEMENT OF 3q21 AND 3q26 BREAKPOINT REGIONS.
The relationship between the MEL1-RPN1 fusion gene and other genes, EAP, GR6 and GATA2, on band 3q21, will be studied.
The rearrangement and the expression of RPN1-Evi1 (and of the others genes mentioned above) will be studied:
a) by conventional cytogenetic analysis and fluorescence in situ hybridisation (FISH)
b) by qualitative and quantitative Real Time RT-PCR analysis.

5) ROLE OF EVI1 AND MDS-EVI1 IN PROMOTING THE LEUKEMIC PHENOTYPE AND ITS ROLE IN THE PATHOGENESIS OF MDS.
It is known that Evi1 and Mds1-Evi1 are induced by retinoic acid in some human cell lines. The expression profile of involved genes of blast cells before and after in vitro and in vivo ATRA exposition will be analyzed by microarrays.

b) TASKS OF EPIGENETIC STUDIES
1)Determine the activity of 4 different HDAC inhibitors on histone H2A and B, H3, H4 of MDS CD34-positive primary cells, compared to normal progenitor cells.

2)Explore the significance of histone methylation in the regulation of gene expression.

3)Gene profiling after histone methylation/acetylation by HDACi. The project is to elucidate the pattern of possible re-expression of silenced genes after HDACi treatment, but also verify the downregulation of genes promoting apoptosis, proliferation and drug resistance.

4)Analysis of signal transduction patterns and crosstalks in MDS cells treated with HDACi.

5)Proteomic analysis of CD34+ myelodysplastic cells compared to normal CD34+ cells, focusing on post-translational modifications of the identified proteins, in particular the degree of acetylation before and after the addition of HDACi and DNMTi, both for histone and non-histone proteins (about 200 proteins).

6)The role of Evi1 associated with its histone deacetylases (HDACs) activity will be also studied. A potent inhibitor of histone deacetylase, such as Tricostatin A/Valproic Acid, will be used in vitro to inhibit Evi1 activity, with the derepression of target genes.

7) METHYLATION STATUS OF THE PROMOTER REGIONS OF DNA-REPAIR GENES.
The methylation status of BRCA-1 and MGMT will be investigated. Mutations of the BRCA1 gene are known to confer susceptibility to breast and ovarian cancer in high-risk families. We will study the methylation status of BRCA-1 in MDS, to evaluate its role as tumor suppressor gene in this disease. MSP for MGMT and BRCA1 will be carried out after bisulfite treatment of DNA, according to Esteller et al (2002).

8) METHYLATION STATUS OF ANGIOGENIC FACTORS : COX-2 AND THROMBOSPONDIN
It has been shown that angiogenesis plays a pathogenetic role in MDS. Among genes important for angiogenesis, we will study COX-2, which has a pro-angiogenic and anti-apoptotic role, and thrombospondin, which has an anti-angiogenic function.

9) FUNCTIONAL STUDIES ON DAP-KINASE HYPERMETHYLATION IN MDS TREATED WITH DEMETHYLATING AGENTS IN-VITRO AND IN VIVO.
Dap-kinase is a serin/threonin kinase, regulated by calmodulin, which participates in a wide array of apoptotic systems started by gamma-interferon, TNF-alpha, Fas, and by detachment from the cellular matrix. We have shown that DAP-kinase is hypermethylated in almost 50% of MDS and AML- We will study the DAP-kinase methylation status and its functionl effects in primary cells of MDS and s-AML patients, treated by demethylating agents and hystone acetilase inhibitors
To study the effect of DAP-kinase reinduction on the apoptotic machinery, the percentage of apoptotic cells will be evaluated by Annexin-V-FLUOS Staining Kit, using flow cytometry.

10)TRANSCRIPTIONAL ACTIVITY INDUCED BY METHYLTRANSFERASE INHIBITORS ON CELL LINES AND PRIMARY TISSUE EXPRESSING EVI1.
The expression profile of MDS cells after exposition to drugs able to modify the methyltransferase activity (MTA) such as 5-Azacytidine (Pharmion) or Decytabine (SuperGen) will be studied. To study epigenetic modifications induced by Evi1 and HDAC activity, the gene silencing method will be used. Double-strand RNA (dsRNA)-based post-transcriptional gene silencing, also recognized as RNA interference (RNAi), is a powerful tool to suppress specific gene expression and allows gene function studies both in cell colture and in vivo. RNAi will be used to induce the silencing of enzymes involved in epigenetics.

c) IMMUNOPHENOTYPING AND OTHER STUDIES
- Aim of the project is to investigate the phenotypic profile of myelodysplastic syndromes (MDS) with a particular focus on expression of adhesion molecule, differentiation antigens and apoptosis.
Specifically, the main objectives of the study are:
1.To determine the pattern of expression of adhesion molecules(Lecam1, ICAM1, VLA-4, VLA-5, Thy1-CD90) and differentiation antigens(CD45, CD34, HLA-DR, CD13, CD33, CD15, CD11b, CD16) in bone marrow samples of patients affected by MDS, and compare it with that of samples collected from patients with acute myeloid leukemia or normal donors. This step will also investigate whether or not MDS express these markers abnormally and, if this is the case, will test whether these aberrant phenotypes may be useful for minimal residual disease detection;
2.To analyze abnormalities of apoptosis by testing the expression of apo/fas (CD95), bax, bcl-2, annexin V and caspase-3;
3.To determine whether altered antigen expression and/or abnormalities of apoptosis have a prognostic impact on disease outcome;
4.To correlate the altered antigen expression and apoptosis with other variables of prognostic relevance such as cytogenetics and
FAB/WHO categories;
5.To monitor the possible modulations of the patient biologic profile during the therapy with Infliximab (low-risk MDS) or 5-Aza/DAC (high-risk MDS);
6.To analyze possible correlations between biologic changes induced by the drugs and clinical outcome.

- EVALUATION OF CLONAL EVOLUTION OF DIFFERENT TYPES OF MYELODYSPLASIA AND CORRELATION TO THE METHYLATION PATTERN AND TREATMENT RESPONSE
Hemopoiesis clonality in patients who will obtain hematological complete remission after chemotherapy will be studied. X-CIP of granulocytes from peripheral blood and bone marrow mononuclear cells will be evaluated using a technique, which studies the polymorphism of the first exon of human androgen receptor gene (HUMARA). <<<

Timescale

24 months

National and international background

Myelodysplastic syndromes(MDS) are a heterogeneous group of clonal haematopoietic disorders, which exhibit ineffective haematopoiesis leading to pancytopenia and with an increased risk (30%) of transformation into acute myeloid leukaemia (AML). Expansion of the abnormal clone is characterized by morphological dysplasia, impaired differentiation, defective cellular functions, and genetic instability. Clonal cytogenetic abnormalities are demonstrable in 40%-50% of patients with primary MDS and in up to 80% of patients with secondary or treatment-induced MDS. The specific chromosome affected and the number of chromosomal abnormalities (i.e., complexity) offer powerful prognostic information. Results from cytogenetic studies applied to hematological malignancies provided new insights for new classifications. First, the so-called MIC (Morphology, Immunology, Cytogenetics) classification of acute and chronic leukaemias, as well as of MDS, was based on specific associations between morphological, immunological and cytogenetic features. More recently, the World Health Organisation (WHO) proposed a wide ranging classification of hematological malignancies and established the state of the art in genetic diagnosis of lympho- and myelo-proliferative diseases. Cytogenetics and molecular biology of several diseases are well known, however the most still remain unknown.
Among acute myeloid leukemias and myelodysplastic syndromes around 60% of cases with both normal karyotype or unusual changes remain to be characterized at molecular level.
Knowledge of the genetic basis of hematological malignancies expanded greatly when conventional cytogenetics was flanked from molecular cytogenetics i.e. in situ hybridization with fluorescent genomic probes (FISH) and comparative genomic hybridization (CGH). By applying these techniques important advances were obtained over the past ten years: number of diagnoses of known cytogenetic sub-groups are increased; new genes have been cloned; new fusion proteins with transforming effects have been detected; cases where conventional cytogenetics failed have been re-analyzed with success; monitoring of residual disease after chemotherapy and/or bone marrow transplant has improved; new cytogenetic markers have been identified for distinct sub-groups of hematological malignancies sharing clinical and hematological characteristics.
Many of these changes are clonal and recurrent and might lead to genetic alterations that contribute to oncogenesis . Among the most frequent cytogenetic alterations of MDS, deletions of a whole chromosome or of part of it, such as the 5q-, the monosomy of chromosome 7 and the 20q-, are the most frequent. Also relevant appear chromosomal abnormalities involving 3q defining a specific subset of high risk (HR) MDS patients. In particular 3q21 and 3q26 are involved simultaneously in inv(3)(q21q26)(3) and in t(3;3)(q21;q26). These abnormalities are observed in patients with acute myeloid leukemia (AML), MDS and chronic myelocytic leukemia (CML) in blastic phase. Inv(3) and t(3;3) are associated to a very aggressive disease with poor response to chemotherapy, failure to achieve complete remission and short survival. Patients with 3q aberrations frequently have normal or elevated platelet counts and an expanded pool of partially dysplastic megakaryocytes with abnormal nuclear and cytoplasmic maturation. Two genes located in 3q26 have been implicated in the development or progression of myeloid leukemia: Evi1 and Mds1. Evi encodes a nuclear DNA binding protein. Evi1 has been shown to interfere with erythroid and granulocytic differentiation in certain model systems and it has suggested also a possible association of dysmegakaryocytopoiesis with overexpression of Evi1. Evi1 contains a SET domain. The proteins containing this domain (like Evi1, G9a, SUV39H1, etc.) are associated with histone methyltransferasi activity (MTA), with characteristic regulation of gene expression (epigenetic). A second function of these type of transcription factor is associated with a poli ADP-ribosylation. Several epigenetic modifications work together in the correct regulation of gene expression: histone acetylation and DNA methylation are the two best characterized epigenetic modifications. Post-synthetic modifications of DNA and of chromatin proteins are of extreme importance as by interfering with chromatin structure they determine its re-modeling, which is necessary to modulate the accessibility to information that is present on DNA.
Methylation is able to modify DNA and consists in the introduction of methyl groups on cytosines mainly present in the CpG dinucleotides. This epigenetic modification introduces as fifth base on DNA, the 5mC. As for the distribution of 5mCs it is well-known that they distribute themselves in a non-random fashion in genomic DNA so that methylated cytosines are localized on the bulk of DNA, while the unmethylated ones are mainly located within particular regions termed CpG islands.
Epigenetic changes, such as alterations of methylation and acetylation with the consecutive DNA remodeling, have acquired increased recognition as mechanisms actively involved in the pathogenesis of MDS. Aberrant DNA hypermethylation is thought to be relevant for leukemic transformation. Several groups have shown that in MDS bone marrow cells present several methylation lesions. Some of the most frequently examined genes in MDS are the cyclin-dependent kinase inhibitors, p15 and p16, and more recently p21 and p57.The p15INK4B gene, and its functional homologue, the p16INK4A gene, encode proteins that negatively regulate the cell cycle by inhibiting the cyclin-dependent kinases 4 and 6, and control the progression of cell cycle. Aberrant methylation of CpG islands in the p15 promoter region commonly occurs in MDS, such as refractory anemia with excess of blasts (RAEB) or RAEB in transformation, and is associated with loss of p15 expression. Using a methylation-sensitive PCR, p15 methylation was not detectable in "low-risk" MDS, whereas in "high-risk" MDS it ranged from 23% at diagnosis to 30% as detected on disease progression. Patients with p15INK4B gene methylation at diagnosis or at any time-points during follow-up had a significantly higher chance of disease progression to acute leukemia than those without gene methylation. Methylation may also play a role in the pathogenesis of therapy-related MDS. Exposure to chemo-radiotherapy induces changes at molecular level in hematopoietic precursors. This includes methylation of the promoter-associated CpG-rich regions, of several genes which are important for DNA-repair process, progression of cell-cycle and cell survival. Death-associated protein kinase (DAP-kinase) is among genes identified as deregulated, through hypermethylation, and that could be of functional significance since they work like tumor suppressor genes. Voso et al found aberrant DAP-kinase hypermethylation in 27.5% of the samples from patients with MDS. Alteration in the apoptotic response, due to the loss of DAP-kinase function, may be an early event in the dysplastic process and subsequent leukemic transformation. Hypermethylation of DNA repair genes such as MGMT and BRCA-1, represents another relevant issue in MDS pathogenesis. The DNA repair protein MGMT removes mutagenic adducts from the O6 position of guanine, thereby protecting the genome against G to A transition mutations. MGMT is hypermethylated in many human cancers and has been associated with G to A mutations of K-ras in colorectal cancer. MGMT expression level varies in different tissue and tumor types. Methylation of CpG islands in the MGMT promoter region is associated with silencing of the gene, resulting in reduced detoxification of alkylating agents. Low levels of MGMT in hematopoietic progenitors may predispose to an accumulation of mutagenic events leading to MDS.
BRCA1 germline mutations are known to confer susceptibility to breast and ovarian cancer in high-risk families. Several studies suggest that promoter hypermethylation may be common to a significant proportion of sporadic breast and ovarian cancers. This suggests that BRCA1 is a tumor suppressor gene which plays an important role in maintaining genomic stability of hematopoetic precursors by repairing DNA.
Chromatin remodeling at the site of promoter genes enables transcriptional initiation and histone acetylation is one of the best characterized epigenetic modification triggering transcription. Histone acetylation occurs on lysine residues at the N-terminal tail of H2B, H3 and H4 histones and this process facilitate the access to DNA by destabilizing histone-DNA interaction. Histone acetylation is regulated by the equilibrium of two enzymes histone acetyltransferase (HAT) and histone deacetylase (HDAC) which are recruited locally by sequence specific DNA binding proteins.The histone octamere, with the tight DNA wrapping constitutes the nucleosome. In leukemic cells, DNA methyltransferase methylates CpG islands and this recruits specific proteins such as MeCP2 which binds transcription repressor complexes including HDACs. This event leads to loss of the acetyl tails of histones and to their strict packaging to DNA, which becomes inaccessible for transcription.
Post transcriptional modifications of histones combinatorial "histone code".
Complex interactions between methyltransferases (DNMTs), methyl-CpG binding proteins (MBDs), demethylase enzyme, and histone acetylases and deacetylases (HDAC), transcription factors, and chomatin structure have been demonstrated. The association of the methylation machinery, DNMTs and MBDs, with HDACs, provides a co-operative linkage in transcriptional silencing between DNA methylation and histone deacetylation. Numerous examples of promoter hypermethylation of tumor suppressor genes, resulting in gene silencing and presumably conferring a growth advantage to involved cells, have been described. A strong correlation between DNA hypermethylation, transcriptional silencing and tightly compacted chromatin has been established in many different systems. Work over the past 5 years has led to a remarkable increase in the knowledge of the mechanisms of chromatin structure modulation and has revealed that chromatin is a dynamic structure which plays an important role in transcriptional regulation. At least a major portion of chromatin remodeling appears to be accomplished through acetylation and deacetylation of the histone octamer tails. Many of the acetylation and deacetylation enzymes turn out to be known transcriptional enhancer and repressor proteins, respectively. Heavily methylated, inactive regions of DNA (bulk chromatin) were also found to be enriched in hypoacetylated histones. The process of DNA methylation, mediated by Dnmt1, may depend on or generate an altered chromatin state via histone deacetylase activity. A major advance in understanding the link between DNA methylation, histone hypoacetylation, and gene silencing came from the recent studies of two laboratories (Fucks et al, Nan et al.). Both groups demonstrated that the methyl-binding protein MeCP2, known to be involved in transcriptional repression of methylated DNA, recruits a histone deacetylase, HDAC1, via the bridging protein Sin3a. Recently, in vitro studies have also shown that DNMT1 physically interacts with either HDAC1 or HDAC2. The connection between DNA methylation and histone deacetylation could have therapeutic implications. DNA methylation and histone deacetylation appear to act as synergistic layers, although dense CpG island methylation is dominant for the maintenance of a silent state of genes during oncogenesis. Inhibitors of histone deacetylases, alone or in combination with DNA hypomethylating agents, may be useful in reactivating tumor suppressor genes in cancer cells. A study using 5-azacytidine followed by trichostatin A (TSA), a potent and reversible HDAC inhibitor of human colon cancer and leukemia cells, robustly reactivated multiple hypermethylated genes such as MLH1, TIMP3, p15INK4B and p16INK.
Beside the notion that MDS arises from sequential genetic damages occurring within the stem cell compartment, followed by epigenetic modifications, a number of evidences suggest that reciprocal interactions between the malignant clone and the microenvironment serve to create a hostile milieu that reinforces ineffective blood cell production. Ineffective hematopoiesis, the hallmark of MDS, arises from impaired progenitor responsiveness to normal trophic signals and excess local generation of inhibitory cytokines, which promote accelerated apoptosis of progenitors and their progeny. Evidence supporting this model derives from cytokine neutralization studies and the direct relationship between plasma tumor necrosis factor concentration and apoptosis. Recent investigations indicate that angiogenic molecules generated by malignant myelomonocytic precursors represent integral diffusible signals that reinforce leukemia progenitor self-renewal.
In this context, the role of TNFalfa has been extensively studied, and it became clear that it might control dysplastic growth either directly, through the generation of free radicals which in turn cause molecular and genetic breakages, or indirectly by inducing hematopoietic stroma malfunctioning.
On the other hand, a great deal of attention has been dedicated to the abnormalities of adhesion properties of stem cells and hematopoietic microenviroment. Defective expression of adhesion molecules may lead to misprogramming of signal transduction thus affecting survival, differentiation, and apoptosis of hematopoietic cells. In this view, some critical molecules such as integrins and selectins have been identified: a defective expression of VLA-4 may alter the normal function of fibronectine which in turn results in an excess of apoptosis of hematopoietic cells. Previous investigations by the Amadori group have shown that in the stem cell compartment of patients with MDS, Lecam1 is generally defective whereas Icam1 is up-regulated. Based on this observation, a Lecam-1/ICAM-1 ratio was derived, expecting low values for MDS CD34+ cells, and high values for normal ones. In line with our working hypothesis, MDS CD34+ cells showed significantly lower values of the proposed ratio as compared to CD34+ progenitors from normal controls. Moreover, the value of Lecam-1/ICAM-1 ratio correlated inversely with bone marrow blast infiltration.
Interestingly, the up-regulation of Icam1 parallels the over-expression of anti-apoptotic proteins and, in some cell lines, is associated with the amplification of the anti-apoptotic homeobox gene DLX-7. Gene derangement may also explain the down-modulation of other adhesion molecules such as CD11a, CD11b, CD11c, CD49d, L-Selectin and of differentiation antigens (CD45, CD34, HLA-DR, CD13, CD33, CD15, CD16). Furthermore, the Amadori group has recently observed an over-expression of Thy1 in the blast compartment of MDS and sAML as compared to de novo AML. The role of Thy1 is not yet fully elucidated, however there are indications that it belongs to the adhesion molecule family with a regulatory function of signal transduction. Based on this background, the use of biologic modifiers hypomethylating agents, inhibitors of HDAC and tyrosine kinases and anti-cytokine drugs, is very attractive, not only in terms of therapeutic approch but even to investigate the the way these drugs interfere with the cell epigenetic modifications occuring in MDS. <<<