Transcriptosome and serum cytokine profiling of ... - Wiley Online Library

American Journal of Hematology 81:779–786 (2006)

Transcriptosome and Serum Cytokine Profiling of an Atypical Case of Myelodysplastic Syndrome with Progression to Acute Myelogenous Leukemia Daruka Mahadevan,1* Johanna DiMento,1 Kimiko Della Croce,1 Christopher Riley,1 Benjamin George,1 Deborah Fuchs,2 Timothy Mathews,3 Charlton Wilson,3 and Michael Lobell1 2

1 University of Arizona Cancer Center, Tucson, Arizona Department of Pathology, University of Arizona, Tucson, Arizona 3 Phoenix Indian Medical Center, Phoenix, Arizona

A Native American-Indian female presenting with anemia and thrombocytosis was diagnosed with myelodysplastic syndrome (MDS, refractory anemia). Over the course of 5 years she developed cytopenias and periods of leukocytosis with normal bone marrow (BM) blast counts, features of an unclassifiable MDS/MPS syndrome. The patient ultimately progressed to acute myelogenous leukemia (AML, FAB M2) and had a normal karyotype throughout her course. The episodes of leukocytosis were associated with infectious complications. Transformation to AML was characterized by a BM blast percentage of 49%. Peripheral blood and BM samples were obtained for serum protein analysis and gene expression profiling (GEP) to elucidate her disease process. An ELISA assay of the serum analyzed ~80 cytokines, which demonstrated that hepatocyte growth factor/scatter factor and insulin-like growth factor binding protein 1 were markedly elevated compared to normal. GEP demonstrated a unique ‘‘tumor molecular profile,’’ which included overexpression of oncogenes (HOXA9, N-MYC, KOC1), proliferative genes (PAWR, DLG5, AKR1C3), invasion/metastatic genes (FN1, N-CAM-1, ITGB5), pro-angiogenesis genes (c-Kit), and down regulation of tumor suppressor genes (SUI1, BARD1) and anti-apoptotic genes (PGLYRP, SERPINB2, MPO). Hence, a biomics approach has provided insight into elucidating disease mechanisms, molecular prognostic factors, and discovery of novel targets for therapeutic intervention. Am. J. Hematol. 81:779– C 2006 Wiley-Liss, Inc. 786, 2006. V Key words: anemia; thrombocytosis; refractory anemia

INTRODUCTION

Myelodysplastic syndrome (MDS) is a clonal disorder of pluripotent stem cells of the bone marrow. These disorders are characterized by ineffective hematopoiesis, with abnormalities in proliferation, differentiation, and apoptosis. The World Health Organization classifies MDS into five subtypes based on blast percentage in the blood and bone marrow, degree of dysplasia (i.e., number of lineages affected), number of ringed sideroblasts, and chromosomal abnormalities [1]. An international prognostic scoring system (IPSS) of MDS predicts for survival and progression to acute myelogenous leukemia (AML) and is based on blast percentage, karyotype, and number of cytopenias; low-, intermediate-, and high-risk groups are identified [2]. In MDS, clonal karyotypic abnormalities are frequently C 2006 Wiley-Liss, Inc. V

observed, occurring in 40–50% of primary MDS and 90% in therapy-related MDS. The bone marrow (BM) karyotypes of MDS are characterized by the loss of genetic material; in contrast to AML, balanced translocations are rarely found. The del(5q) or 5, del(7q) or 7, del(20q), del(11q), and Y are among the most frequently reported cytogenetic abnormalities in MDS [3]. *Correspondence to: Daruka Mahadevan, University of Arizona Cancer Center, Tucson, AZ 85724. E-mail: [email protected] Received for publication 10 July 2005; Accepted 9 March 2006 Published online 12 July 2006 in Wiley InterScience (www.interscience. wiley.com). DOI: 10.1002/ajh.20690

780

Case Report: Mahadevan et al.

Several studies have evaluated the gene expression profiling (GEP) in MDS in the context of IPSS and each has utilized different control RNA to categorize sets of up and down regulated genes. The first study analyzed CD34(þ) BM cells from 7 patients with lowrisk and 4 patients with high-risk MDS in comparison to CD34(þ) BM cells from 4 controls. Eleven prognostic genes were identified based on a class membership prediction analysis. In low-risk MDS, the retinoicacid-induced gene, the radiation-inducible, immediateearly response gene, and the stress-induced phosphoprotein 1 were down regulated, suggesting that low-risk MDS lacked defensive proteins, resulting in susceptibility to cell damage [4]. A second study utilized BM stem cells with marker AC133 from healthy volunteers and 30 patients at various stages of MDS and identified sets of genes with expression specific to either low- or high-risk MDS. For low-risk MDS the gene PIASy, which catalyzes protein modification with the ubiquitin-like molecule SUMO, was identified. Loss of PIASy expression may contribute to growth of MDS blasts and stage progression [5]. A third study utilized peripheral blood neutrophils of 21 MDS patients; 7 had 5q syndrome and 2 had AML. The most up regulated genes included RAB20, ARG1, ZNF183, and ACPL. The RAB20 gene is a member of the Ras gene superfamily and ARG1 promotes cellular proliferation. The most down regulated genes include COX2, CD18, FOS, and IL7R. COX2 is anti-apoptotic and promotes cell survival. Many genes were identified that are differentially expressed in the different MDS subtypes and AML. A subset of genes was able to discriminate patients with the 5q syndrome from patients with RA and a normal karyotype [6]. Two recent studies of the GEP in large numbers of AML patients identified several signature profiles that correlated with prognosis and overall survival [7,8]. The current study evaluated the GEP and serum markers of prognosis of a 70-year-old Native American-Indian female with an initial diagnosis of refractory anemia (MDS) and a marked thrombocytosis (MPD) with normal cytogenetics. This GEP study was performed after the patient had transformed to AML and therefore these finding may not be generalizable to early-stage MDS. At the time of transformation to AML (FAB M2, normal cytogenetics), blood and BM were studied for prognostic markers and underlying disease mechanisms in the hope of gleaning an insight into the disease mechanisms and to identify potential therapeutic targets. METHODS Clinical Samples and Flow Cytometry

The patient is a 70-year-old female who consented to providing part of her blood (5 mL) and bone marAmerican Journal of Hematology DOI 10.1002/ajh

row (5 mL) for research while samples were taken for routine diagnostic purposes. The study was approved by the Indian Health Services Internal Review Board. The blood was spun, serum collected, and stored at 808C. BM was spun and white blood cells were harvested and stored at 808C. Consent was obtained from a normal healthy 30-year-old female for obtaining 5 mL of blood. This sample was spun, serum collected, and stored at 808C. Flow cytometry on the BM was utilized to identify blast cells positive for CD13, CD33, CD34, and CD117. The flow data were correlated with gene expression of the above markers as validation tool. ELISA Analysis of Serum Proteins

Control and patient sera were hybridized onto preprinted arrays purchased from RayBiotech (No. H0108005, GA) according to the manufacturer’s protocol. The study was repeated once. Analysis was performed via exposure to X-ray film. Quantitation was achieved via analysis of the strength of the black spectrum of color in a black and white scanned image of the film in Adobe Photoshop. Selection of the area for analysis was done with the eyedropper tool, averaging color intensity over an area of 5 pixels. Results are represented as up or down changes in intensity of the patient sample compared to the control. Gene Expression Profiling by Oligonucleotide Microarray

White blood cells were centrifuged, pelletted, and resuspended in cold PBS and counted using a hemacytometer. An aliquot of cells (2 107) was lysed to isolate RNA utilizing the RNeasy Mini kit according to the manufacturer’s protocol (Qiagen No. 74104, CA). The resulting RNA was eluted with RNase-free water (40 mL), quantitated via spectrophotometry, and stored at 808C. Human normal BM myeloid cells (CD33þ) total RNA was purchased from AllCells (No. RNA-PB003, CA) and stored at 808C until use. A total of 5 mg of mRNA was used to generate first-strand cDNA by using a T7-linked oligo(dT) primer. After secondstrand synthesis, in vitro transcription (Ambion) was performed with biotinylated UTP and CTP (Enzo Diagnostics), resulting in 40- to 80-fold linear amplification of RNA. A total of 40 mg of biotinylated RNA was fragmented to 50- to 150-nt size before overnight hybridization to HG- U133A 2.0 Affymetrix array comprising 18,400 transcripts and 22,000 probe sets (Santa Clara, CA). After washing, arrays were stained with streptavidin-phy-

Case Report: Transcriptosome and Serum Cytokine Profiling in MDS

coerythrin (Molecular Probes) and scanned on a Hewlett-Packard scanner. Intensity values were scaled such that overall intensity for each chip of the same type was equivalent. Intensity for each feature of the array was captured by using GENECHIP software (Affymetrix, CA), and a single raw expression level for each gene was derived from the 20 probe pairs representing each gene by using a trimmed mean algorithm. Affymetrix CEL files were imported into R (Bioconductor version 1.7) using the ‘‘read.affy’’ function provided in the SimpleAffy package (version 2.11) available from Bioconductor (http://bioconductor.org). Normalized log2 expression values and absent/present calls were generated with the ‘‘call.exprs’’ function using the MAS5 algorithm. The TGF (target scaling factor) was 100 and scale factors of 2.088243 and 1.426016 were applied to the Patient and Control chips, respectively. Affymetrix quality controls were investigated using the ‘‘qc’’ function provided in the Simpleaffy package. Average background, hybridization controls, and internal controls performed within standards as presented (in Chapter 4: Guidelines for Assessing Data Quality of the Affymetrix-GeneChip Expression Analysis: Data Analysis Fundamentals manual: Affymetrix, Inc., Santa Clara, CA). Genes called absent on both chips were removed from further analysis. Fold changes were derived from the log2 ratios of the normalized expression values. We converted log2 values to log10 to provide raw expression intensities for each probe set. Patient’s relative gene expression levels was compared to the control sample and lists of ‘‘robust increasers’’ and ‘‘robust decreasers’’ were generated utilizing the Affymetrix Data Analysis Program (Affymetrix MAS 5.0). Gene expression profile was further classified according to the Hallmarks of Cancer [9] to determine an MDS/AML specific, ‘‘tumor profile signature’’ for this patient. These are genes involved in self-sufficiency in growth signals including oncogenes, insensitivities to growth inhibitory signals including tumor suppressors, evasion of apoptosis, limitless replicative potential, sustained angiogenesis, and invasion and metastasis.

RESULTS A Diagnostic Dilemma

The patient was a 70-year-old female whose pertinent past medical history began in 1998, when she was 65 years of age. The diagnosis and clinical course of her disease was a dilemma to the treating hematologist for the following reasons. At that time, she presented with symptoms of dizziness, malaise,

781

and fatigue. There was no family history of hematologic malignancy. She had a macrocytic anemia with an elevated B12 level, accompanied by an elevated platelet count of 873,000/mL. BM biopsy revealed a diagnosis of refractory anemia (MDS) with reactive thrombocytosis. Over the course of 3 years, she was treated with multiple regimens, i.e., anagrelide, hydroxyurea, topotecan with high-dose Ara-C, thalidomide, and prinomastat. Despite these interventions, she remained transfusion dependent with progressive pancytopenia. Repeat BM biopsy (8/2001) showed refractory cytopenia with multilineage dysplasia with a normal karyotype. By 2002, transient elevations in the white blood cell (WBC) count occurred that usually coincided with a systemic infection. In February of 2003, during the course of a lower extremity cellulitis, the WBC rose to 91,000/ml with 5% circulating blasts. BM biopsy revealed 6.8% myeloblasts with a diagnosis of a combined MDS/MPD, again with a normal karyotype. Approximately 4 weeks later, during the course of an Escherichia coli urinary tract infection, the patient’s WBC rose to 164,000/ml with 6% circulating blasts. This hyperleukocytosis promptly responded to leukapheresis and hydroxyurea. Over the next 2 months, the white count increased despite continued hydroxyurea. While experiencing symptoms consistent with an upper respiratory infection, the WBC was 184,000/ml with 24% circulating blasts. Repeat marrow examination at this time revealed 49% myeloblasts with transformation to AML (FAB M2) with myeloid maturation. With the patient’s consent, blood and bone marrow samples were obtained. Figure 1 shows the patients BM aspirate stained with Wright stain to demonstrate the increase in myeloblasts and dysplastic granulocytes. She required transfusions of blood and platelets over the course of the next month prior to her death.

Serum HGF/SF and IGFBP 1 as Markers of Progression and Poor Prognosis

Plasma cytokine, chemokine, and growth factor profile for the patient and control is shown in Figure 2. The control sample had four proteins that were higher than the patient sample (leptin, NAP-2, PDGF-b, and RANTES). RANTES was expressed at 85% intensity compared to the average of the positive controls in the control sample, but was undetectable in the patient. NAP-2 was expressed at 66% intensity in the control but was also not detectable above background in the patient. PDGF-b was detected at 98% intensity, and leptin was detected at 97% intensity; neither was found to be above background in the patient sample. HGF/SF proAmerican Journal of Hematology DOI 10.1002/ajh

782


Fig. 1. Wright stain of bone marrow aspirate demonstrates numerous myeloblasts (thick arrows) and hypogranular neutrophil precursors (thin arrows) with nuclear hypolobulation (pseudo Pelger-Huet nuclei). [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

duced by mesenchymal cells, including BM stromal cells has potent mitogenic and motogenic effects on a variety of cell types [10]. Plasma levels of HGF correlate with shorter survival of AML but not MDS patients [11,12]. The patient had a HGF level 35-fold higher than control but there are no data prior to her progression. HGF is an angiogeneic factor and like VEGF and bFGF is secreted by neoplastic hematopoietic cells to promote growth and proliferation of the leukemic cells in an autocrine fashion [13]. Moreover, soluble HGF is more potent than VEGF in controlling the mobilization of blast cells from bone marrow to peripheral blood and tissues and is postulated to have played a role in this patient. Insulin-like growth factors (IGFs) and IGFbinding proteins (IGFBPs) may play a critical role in tumor proliferation. The index case had an IGFBP-1 level 25-fold higher than control. However, the other IGF binding proteins 2, 3, and 4 were not elevated. IGFBPs interact with cell or matrix components and may concentrate IGFs near their receptor, enhancing IGF activity. IGFBP-1 interacts with a5b1 integrin, influencing cell adhesion and migration [14,15], a potential mechanism of blast mobilization. Taken together, elevated levels of HGF and IGFBP-1 appear to signal the transition from MDS to AML, promotes tumor proliferation and mobilization, and are markers of poor prognosis for disease progression and survival. This is a novel finding and warrants further investigation in a large cohort of similar patients to validate prognostic significance and method of targeting by validating in a mouse model of MDS. American Journal of Hematology DOI 10.1002/ajh

Fig. 2. Serum cytokine profile showing the up regulated HGF/SF (A) and IGFBP-1 (B). Top left and bottom right are positive and negative controls. [Color figure can be viewed in the online issue, which is available at www.interscience. wiley.com.]

The Transcriptosome Profile Identifies a Unique Tumor Profile

The GEP compared normal BM CD33þ cells to the patient CD33þ BM blast cells. The mRNA level of CD33 was equivalent in both samples and correlated with flow cytometry for CD33. The robust up and down regulated genes, fold change, and their known functions are shown in Table I. The highly up regulated genes appear to be intimately linked to cellular survival and proliferation while the highly down regulated genes most likely from gene silencing appear to have many distinct functions that if normally expressed or up regulated may provide a survival benefit. Of the highly up regulated genes, HLA DRB4 was 150-fold increased while of the highly down regulated genes, HLA DQB1 was 114-fold decreased. It is hypothesized that the latter may be due to a small deletion leading to gene silencing. Genes encoding the six hallmarks of cancer [9] are shown in Table II and can be utilized to ascertain the unique nature of each patient’s disease as well as common relative gene expression changes in patients with similar diagnoses and provide biological insights to disease mechanisms and potential targets for therapeutic intervention.

TABLE I. The Top 20 Robust Up and Down Regulated Genes

Gene HLA-DRB4

Fold increase 177

PXDN HOXA9

83 72

FN1

63

MYCN LOC51334 HOXB6 ITM2C

48 44 38 38

ZBTB1

37

FLJ12787 PPFIBP1; LOC440091

34 33

SPATA7 SPINK2 IMP-3/KOC1

30 30 28

TNNT1

28

XIST CDKN1C

26 26

MEG3

23

TRBV21-1; TRBV19; TRBV5-4; TRBV3-1; TRBC1 PLOD2

22

Gene

22 Fold decrease

MMP9

58

OLR1

28

LCN2

25

OLFM4 CYP4F3 MMP8 cAMP TCN1 IL8 CLC ANXA3 ALOX5AP LTF S100A12 ARG1 YWHAB

22 21 20 18 18 14 13 9 8 8 8 7 7

CCL3; CCL3L1; CCL3L3

7

SCAP2 STOM MPO

7 7 7

Function Presents extracellular derived proteins, belongs to the HLA class II beta chain paralogues Peroxidasin homolog (Drosophila) Associated w/ the nuclear pore complex for signal mediated import and export Cell adhesion and migration along protein fibers within the ECM DEAD box protein is a putative RNA helicase Mesenchymal stem cell protein DSC54, similar to rubredoxins Homeobox protein Hox-B6, transcript variant 2 Integral membrane protein 2C, found in proteins implicated in dementia, respiratory distress, and cancer Stratagene lung carcinoma, pyruvate kinase M2 isozyme, involved in transcriptional regulation UTP15, U3 small nucleolar ribonucleoprotein, homolog (yeast) PTRF interacting protein, binding protein 1 (liprin beta 1); similar to PTPRF interacting protein binding protein 1 isoform 1; liprin-beta 1; liprin related protein; proteintyrosine phosphatase receptor-type f polypeptide-interacting protein-binding protein 1 Hypothetical protein HSD-3.1, associated w/ spermatogenesis Serine protease inhibitor, Kazal type (acrosin-trypsin inhibitor) Found in nucleolus interacting w/ U3 snoRNA or complexing w/ IGF-II mRNA binding protein Human slow skeletal troponin T, shows expression in cardiac muscle X (inactive)-specific transcript Tight-binding inhibitor of several G1 cyclin/Cdk complexes and a negative regulator of cell proliferation Very hypothetical protein from MEG3 locus (Protein PRO0518) T-cell receptor beta variable 21-1; T cell receptor variable 19; T cell receptor beta variable 5-4, T cell receptor beta variable 3-1; T cell receptor beta constant 1 Procollagen-lysine, 2-oxoglutarate 5-dioxygenase 2

Function Multifunctional proinflammatory cytokine belonging to the TNF superfamily, mainly secreted by macrophages Encodes receptor protein in C-type lectin superfamily; binds, internalizes and degrades oxidized LDL Marker for dysregulated keratinocyte differentiation in human skin Olfactomedin 4 Cytochrome P450, subfamily IVF Matrix metalloproteinase 8 (neutrophil collagenase) Cellular signaling cascade molecule, diverse function Transcobalamin I (vitamin B12 binding protein) Interleukin 8 Charcot–Leyden crystal protein Annexin A3 Arachidonate 5-lipoxygenase-activating protein Lactotransferrin S100 calcium binding A12 (calgranulin C) Arginase, liver Tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein, beta polypeptide Chemokine (C-C motif) ligand 3; chemokine (C-C motif) ligand 3-like 1; chemokine (C-C motif) ligand 3-like 3 Src family associated phosphoprotein 2 Stomatin myeloperoxidase

Probe intensity (log 10) patient

Probe intensity (log 10) control

1308.22 155.33

7.39 1.88

123.46

1.7

178.14 74.34 19.51 104.67

2.83 1.55 0.44 2.76

272.14

7.26

13.89 29.93

0.38 0.88

32.42 46.29 217.31

0.99 1.56 7.32

71.93

2.56

145.97 685.61

5.23 26.43

141.31

5.47

336.71

14.46

74.43 12.64

3.34 0.58



63.74

3673.52

16.28

459.56

191.78 61.06 48.34 124.37 357.35 82.99 171.36 163.45 72.82 352.79 698.92 496.49 120.15

4723.05 1348.93 997.54 2446.06 6263.15 1453.84 2421.65 2114.18 645.1 2902.35 5450.09 3845.85 888.82

81.96

600.13

102.6 42.8 122.16 270.14

715.56 296.77 842.45 1841.94

784


TABLE II. Hallmarks of Cancer

Gene symbol Oncogenes

Fold change

HOXA9

72

MYCN IMP-3/KOC1

48 28

Proliferation

BMII/PCGF4 R-RAS IMP-3/KOC1

10 4 28

Anti-apoptosis

PAWR DLG5 AKR1C3 PA26/SESN1 BAG3

13 11 7 3 3

Tumor suppressors

SOCS2 CDKN1C

3 26

Angiogenesis Replication

KIT LTBP3

Invasion and metastasis

FN1

63

NCAM1

21

ITGB5 COL6A2 C1QA

12 11 9

3 4

Function(s) Associated w/the nuclear pore complex for signal mediated import/export DEAD box protein is a putative RNA helicase Found in nucleolus interacting w/U3 snoRNA or complexing w/ IGF-II mRNA binding protein Regulation of self-renewal of leukemic stem cells Immortalization and transformation Found in nucleolus interacting w/U3 snoRNA or complexing w/ IGF-II mRNA binding protein Binds WT tumor suppressor Membrane-associated guanylate kinase Aldo/keto reductase PA26, a target of the p53 and a GADD family member Bind to the ATPase domain of Hsc70 and inhibit its chaperone activity in a Hip-repressible manner Suppressor of cytokine signaling 2 Tight-binding inhibitor of several G1 cyclin/Cdk complexes and a negative regulator of cell proliferation Receptor tyrosine kinase Assembly, secretion and targeting of TGFb1 to sites at which it is stored and/or activated Cell adhesion and migration along protein fibers within the ECM Cell adhesion molecule involved in neuron-neuron adhesion, neurite fasciculation, outgrowth of neuritis, etc. Integrin b5 Collagen 6A2 Complement 1QA

DISCUSSION

The diagnostic dilemma posed by this patient’s unclassifiable MDS/MPS prompted a molecular investigation that searched for biological clues in her bone marrow and serum proteome. Although CD33 was utilized as a selection marker, the cellular composition of the control sample by morphology was unknown. Therefore, differences in cellular composition between the control and patient samples may exist and contribute to the GEP identified. However, cytospins and flow histograms of the patient sample were gated on dim CD45þ cells with low side scatter, which were blasts, and the majority were CD33þ. The results obtained demonstrate dysregulated gene expression despite normal cytogenetics and increased levels of HGF and IGFBP-1 in disease progression. The tumor profile signature based on the six hallmarks of cancer clearly demonstrates the complex nature of this patient’s disease. Several overexpressed oncogenes appear to collaborate in MDS initiation and progression in this patient. With regard to dysregulated tumor growth the patient had up regulation of Hoxa9, a homeobox protein transcription American Journal of Hematology DOI 10.1002/ajh



Affy_ID

123.46 74.34

1.7 1.55

209905_at 209757_s_at

71.93 579.31 202.12

2.56 56.5 47.08

203819_s_at 202265_at 212647_at

71.93 20.74 38.78 53.51 315.16

2.56 1.64 3.62 7.97 90.03

203819_s_at 204005_s_at 201681_s_at 209160_at 218346_s_at

125.21 47.29

39.96 16.63

217911_s_at 203372_s_at

141.31 215.4

5.47 61.95

213182_x_at 205051_s_at

77.46

20.04

219922_s_at

178.14

2.83

211719_x_at

108.28 17.97 95.57 411.44

5.16 1.51 8.63 44.1

212843_at 214020_x_at 209156_s_at 218232_at

factor that is known to be overexpressed in AML, especially in relation to the expression of the mixedlineage leukemia oncogene [16] Overexpression of the Hoxa9 gene in BM cells induces stem cell expansion and is correlated with a poor prognosis [17]. The MYC oncogene family is known to be deregulated and overexpressed in many malignancies including leukemia and the different isoforms can substitute for each other with the patient having overexpression of N-MYC. The KOC1 gene encodes KOC protein, which contains an RNA-binding motif known as the KH domain. It was isolated as a gene overexpressed in human pancreatic cancer and is thought to play a role in tumor cell proliferation [18]. Bmi-1 is a member of the Polycomb group (PcG) of genes, which have an essential role in embryogenesis, regulation of the cell cycle, and lymphopoiesis. PcG genes are responsible for preservation of gene silencing and are therefore essential for upholding cell identity. Bmi-1 is indispensable for regulation of self-renewal by normal and leukemic stem cells [19]. Finally, R-Ras is a ras oncogene member that is implicated in immortalization and transformation of normal cells to malignancy. Additionally, two tumor

Case Report: Transcriptosome and Serum Cytokine Profiling in MDS

suppressor genes, BARD1 and SUI 1, are down regulated, of which BARD 1 is implicated in BRCA1mediated tumor suppression [20], while SUI 1 encodes a translation initiation factor that is suppressed during tumorigenesis [21]. This unique combination of up regulation of five potent oncogenes (Hoxa9/ N-MYC/Koc1/BMI1/R-Ras) and down regulation of two tumor suppressors (BARD1/SUI1) appear to be the molecular basis of rapid progression from MDS to AML in this patient. The overexpression of CD117 (c-Kit) on the patient’s blast cells most likely provides a potent stimulus for both proliferation and angiogenesis. A recent phase 2 trial evaluated the efficacy of imatinib mesylate in 21 patients with c-kit positive AML; all received 600 mg imatinib orally once daily. Five responses were observed in patients starting with relatively low blast count in BM and PB. Two patients, considered refractory to chemotherapy on the basis of persistence of blasts in PB and BM, achieved a complete hematologic remission. One patient had no evidence of leukemia, while two patients achieved a minor response. DNA-sequencing indicated wildtype c-KIT with no mutant forms identified [22]. These five cases suggest that imatinib has modest clinical activity in a subset of patients with c-kitpositive AML. A trial of imatinib may have benefited this patient. To make matters more complex several proteins involved in invasion and metastasis are also overexpressed. These are fibronectin (Fn1), neural cell adhesion molecule 1 (NCAM1), integrin B5 (ITGB5), complement C1QA, and collagen COL6A2, forming a complex network of cell surface adhesion molecules and extracellular matrix molecules that appear to facilitate blast mobilization. In order to enhance leukemia cell survival several apoptotic genes are up (PAWR, BAG3, SOCS2, TNFSF9, or CD137L) and down regulated (MPO, SERPINB2, PGLYRP). Of these only Serpinb2 (plasminogen activator inhibitor-2, PAI-2) and SOCS-2 (suppressor of cytokine signaling-2) are anti-apoptotic genes that are down regulated. The rest of the apoptotic genes fit into the category of promoting cell survival. PAWR (Wilm’s tumor 1 interacting protein par-4) binds to WT1 tumor suppressor protein and promotes cell survival by inhibiting apoptosis [23]. BAG3 is a co-chaperone proteins that share the Bcl-2-associated athanogene (BAG) domain, interacts with heat shock proteins (Hsp), steroid hormone receptors, Bcl-2, and Raf-1. It is involved in regulating protein folding, proliferation, and apoptosis by interfering with cytochrome c release, apoptosome assembly, and other events in the death process [24]. TNFSF9 (tumor necrosis factor ligand superfamily 9), also known as CD137L

785

(CD137 ligand), was recently shown to correlate with rapid progression of MDS to AML [25] and is a potential marker of prognosis. The role of CD137/CD137L system in leukemia is unknown; these co-stimulatory molecules may confer an advantage to hematologic tumors promoting survival, sustaining cellular growth, and contributing to drug resistance. One of the critical observations with this patient was her predisposition to infectious complications. Myeloperoxidase (MPO), a known cytochemical marker in the evaluation of AML, plays a role in host defense in polymorphonuclear leukocytes and is responsible for microbicidal activity against a wide range of organisms. It is eightfold down regulated and may correlate with hyperleukocytosisassociated infectious episodes observed in this patient. The BM from June 2004 (Figure 1) shows hypogranular myeloblasts and neutrophil consistent with this observation. Peptidoglycan recognition proteins (PGLYRP) are pattern recognition receptors of the innate immune system that bind and in some cases hydrolyze peptidoglycans on bacterial cell walls. These molecules participate in host defense against both gram-positive and gram-negative bacteria. This molecule is down regulated in the patient, thus enhancing her susceptibility to bacterial infection [26].

CONCLUSIONS

MDS is an excellent model of AML development with a progressive increase of blasts in the BM, but the genetic events that lead to this evolution remain poorly understood. The diagnostic dilemma presented by this patient with MDS/MPD with normal cytogenetics, periods of hyperleukocytosis with infectious complications, and progression to AML can be rationalized by a functional genomics and proteomics approach. Moreover, this study provides markers of prognosis, disease progression, and several therapeutic targets for intervention.

ACKNOWLEDGMENTS

We thank the Phoenix Indian Medical Center (PIMC) and their IRB for approving the manuscript. They also provided the necessary education to allow research to be successful in Indian country. The views are those of the authors and do not necessarily represent the view of the Indian Health Services. Dr. Lobell thanks the NCI for providing funding (Special Population Network) to allow PIMC and the Arizona Cancer Center to develop a partnerAmerican Journal of Hematology DOI 10.1002/ajh

786


ship stressing education, research, and teaching as a means to decrease the burden of cancer in the Arizona Native Americans. The authors thank Dr. Alan List for critical review of the manuscript. Finally, we thank Dr. Thomas P. Miller, Chief of Hematology/ Oncology at the Arizona Cancer Center, and Laurence Cooke for support and encouragement. REFERENCES 1. Jaffe ES, et al. World Health Organization classification of tumors. Pathology and genetics of tumors of hematopoietic and lymphoid tissues. IARC Press: Lyon; 2001. 2. Cortes JE, et al. In: Pazdur R, editor. Myelodysplastic syndromes. Cancer management: a multidisciplinary approach, 6th ed, 2001; Chap 37, p 735–750. 3. Boultwood J, Fidler C. Chromosomal deletions in myelodysplasia. Leukemia Lymphoma 1995;17:71–78. 4. Hofmann WK, et al. Characterization of gene expression of CD34þ cells from normal and myelodysplastic bone marrow. Blood 2002;100(10):3553–3560. 5. Ueda M, et al. DNA microarray analysis of stage progression mechanism in myelodysplastic syndrome. Br J Haematol 2003; 123(2):288–296. 6. Pellagatti A, et al. Gene expression profiling in the myelodysplastic syndromes using cDNA microarray technology. Br J Haematol 2004;125(5):576–583. 7. Bullinger L, et al. Use of gene-expression profiling to identify prognostic subclasses in adult acute myeloid leukemia. N Engl J Med 2004;350(16):1605–1616. 8. Valk PJM, et al. Prognostically useful gene-expression profiles in acute myeloid leukemia. N Engl J Med 2004;350(16):1617– 1628. 9. Hanahan D, Weinberg R. The hallmarks of cancer. Cell 2000; 100(1):57–70. 10. Weimar IS, et al. Hepatocyte growth factor/scatter factor (HGF/SF) affects proliferation and migration of myeloid leukemic cells. Leukemia 1998;12(8):1195–1203. 11. Hjorth-Hansen H, et al. Elevated serum concentrations of hepatocyte growth factor in acute myelocytic leukaemia. Eur J Haematol 1999;62(2):129–134. 12. Verstovsek S, et al. Plasma hepatocyte growth factor is a prognostic factor in patients with acute myeloid leukemia but not in

American Journal of Hematology DOI 10.1002/ajh

13. 14.

15.

16.

17.

18.

19.

20.

21.

22.

23. 24.

25.

26.

patients with myelodysplastic syndrome. Leukemia 2001;15(8): 1165–1170. Albitar M. Angiogenesis in acute myeloid leukemia and myelodysplastic syndrome. Acta Haematol 2001;106(4):170–176. Aref S, et al. Soluble hepatocyte growth factor (sHGF) and vascular endothelial growth factor (sVEGF) in adult acute myeloid leukemia: relationship to disease characteristics. Hematology 2002;7(5):273–279. Firth SM, Baxter RC. Cellular actions of the insulin-like growth factor binding proteins. Endocr Rev 2002;23(6):824– 854. Yokoyama A, et al. Leukemia proto-oncoprotein MLL forms a SET1-like histone methyltransferase complex with menin to regulate Hox gene expression. Mol Cell Biol 2004;24(13):5639– 5649. Thorsteinsdottir U, Mamo A, Kroon E, et al. Overexpression of the myeloid leukemia-associated Hoxa9 gene in bone marrow cells induces stem cell expansion. Blood 2002;99(1):121–129. Kobayashi K, et al. Identification of genes whose expression is upregulated in lung adenocarcinoma cells in comparison with type II alveolar cells and bronchiolar epithelial cells in vivo. Oncogene 2004;23(17):3089–3096. Raaphorst FM. Self-renewal of hematopoietic and leukemic stem cells: a central role for the Polycomb-group gene Bmi-1. Trends Immunol 2003;24(10):522–524. Wu LC, et al. Identification of a RING protein that can interact in vivo with the BRCA1 gene product. Nat Genet 1996;13: 430–440. Lian Z, et al. The translation initiation factor, hu-Sui1 may be a target of hepatitis B X antigen in hepatocarcinogenesis. Oncogene 1999;18(9):1677–1687. Kindler T, et al. Efficacy and safety of imatinib in adult patients with c-kit-positive acute myeloid leukemia. Blood 2004; 103(10):3644–3654. Johnstone RW, et al. Mapping of the human PAWR (par-4) gene to chromosome 12q21. Genomics 1998;53(2):241–243. Bonelli P, et al. BAG3 protein regulates stress-induced apoptosis in normal and neoplastic leukocytes. Leukemia 2004;18(2): 358–360. Salih HR, et al. Serum levels of CD137 ligand and CD178 are prognostic factors for progression of myelodysplastic syndrome. Leukemia Lymphoma 2004;45(2):301–308. Liu C, et al. Peptidoglycan recognition proteins: a novel family of four human innate immunity pattern recognition molecules. J Biol Chem 2001;276(37):34686–34694.