Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

ABC transporters as predictive factors for chemotherapeutic response in acute myeloid leukemia Ho, Maria Ming Chee 2007-12-31

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
ubc_2007-267264.pdf [ 15.88MB ]
[if-you-see-this-DO-NOT-CLICK]
Metadata
JSON: 1.0100341.json
JSON-LD: 1.0100341+ld.json
RDF/XML (Pretty): 1.0100341.xml
RDF/JSON: 1.0100341+rdf.json
Turtle: 1.0100341+rdf-turtle.txt
N-Triples: 1.0100341+rdf-ntriples.txt
Original Record: 1.0100341 +original-record.json
Full Text
1.0100341.txt
Citation
1.0100341.ris

Full Text

ABC TRANSPORTERS AS PREDICTIVE FACTORS FOR CHEMOTHERAPEUTIC RESPONSE IN ACUTE MYELOID LEUKEMIA by  Maria Ming Chee Ho B.Sc., The University of British Columbia, 2001  A THESIS S U B M I T T E D IN P A R T I A L F U L F I L L M E N T OF T H E R E Q U I R E M E N T S F O R T H E D E G R E E OF  D O C T O R OF P H I L O S O P H Y in T H E F A C U L T Y OF G R A D U A T E STUDIES (Biochemistry and Molecular Biology)  The University of British Columbia A p r i l 2007  Maria M i n g Chee H o , 2007  A B S T R A C T  Multidrug Resistance ( M D R ) , resistance to multiple chemotherapeutic drugs, is a major problem in the treatment o f acute myeloid leukemia ( A M L ) . Overexpression o f members o f the A T P Binding-Cassette ( A B C ) transporter superfamily has been associated with clinical M D R and failure o f conventional chemotherapy. The work in this thesis was the first in investigating expression o f A B C transporters and functional effects o f their modulation in A M L subpopulations along the leukemic stem cell hierarchy: C D 3 4 + C D 3 8 - (primitive and disease maintaining), CD34+CD38+ (differentiating progenitors), and C D 3 4 - (depleted o f progenitors). A n initial profiling o f m R N A expression o f the 47 human A B C transporters in total de novo blasts by R T Real-Time P C R showed no consistent differences between patients who subsequently achieved complete remission following conventional remission induction chemotherapy (responders) and patients who remained refractory (non-responders). Subsequent profiling o f isolated subpopulations, however, revealed elevated expression o f MDR1 and/or BCRP1,  two main drug-resistance A B C transporters, in the primitive C D 3 4 + C D 3 8 - fraction o f  7/10 non-responders compared to 0/7 responders. To test their functional activity ex vivo, daunorubicin sensitivity with or without A B C modulators was determined in A M L subpopulations by the apoptotic assay. I found high ABC-dependent drug resistance, correlated to high MDR1IBCRP1  expression level and reversible by A B C inhibition, in the C D 3 4 + C D 3 8 -  fraction o f non-responders compared to responders. This suggests an active functional role o f A B C transporters in the primitive, disease-maintaining fraction. Taken as a whole, my studies suggest a prognostic significance o f A B C transporters in the primitive C D 3 4 + C D 3 8 - leukemic subpopulation, and support a modified approach in investigating the value o f A B C modulating agents in A M L . It may be possible to pre-screen and identify patients for whom A B C transporters is a major factor for M D R before initial treatment, who are most likely to benefit from the combination o f conventional chemotherapy and A B C inhibitors. This w i l l be invaluable especially to patients with a normal karyotype (50% o f patients), since cytogenetic aberrations currently remain the most useful prognostic marker for A M L .  TABLE OF CONTENTS Abstract.... Table of contents List of tables List of figures List of abbreviations Acknowledgements Dedication  I  ••••••  Introduction... 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9  II  ••  -  •  ii .....iii v i ....viii x xi v  1  Challenges in cancer therapy ...1 Acute Myeloid Leukemia: an overview 3 Current prognostic factors for predicting chemotherapeutic response in A M L 5 ABC transporters: an overview 7 ABC transporters and multidrug resistance in cancer 9 ABC transporters in A M L . 12 The leukemic stem cell model and its implications in drug response 13 ABC transporters in normal and leukemic stem cells 15 Thesis objectives 18  mRNA expression profiling of the ABC transporter superfamily in unfractionated A M L patient samples. ..35 2.1 - Introduction 2.2 - Materials and Methods 2.2.1 - Patient samples, cell lines and culture 2.2.2 - R T - R e a l Time P C R assay.  35 37 37 38  2.2.2.1. Overview 38 2.2.2.2. RNA isolation, DNase treatment and Reverse Transcription 39 2.2.2.3. Primer design and optimization 39 2.2.2.4. Real-Time PCR 39 2.2.2.5. Generation of standard curve 40 2.2.2.6. Data analysis 40 2.2.2.7. Statistical analysis 40 2.3-Results 41 2.3.1 - Profiling o f A B C transporters in the drug-sensitive leukemic cell line C E M and its vinblastine-selected, drug-resistant subline C V T . 0 41 2.3.2 - Lack o f consistent differences was observed i n A B C transporter expression between responsive and non-responsive patients ....42  2.4 - Discussion  III  44  Expression profiling of drug resistance-related transporters in FACSsorted A M L subpopulations 53 3.1 - Introduction 3.2 - Materials and Methods  53 55  in  3.2.1 - F l o w cytometric sorting o f A M L subpopulations ..55 3.2.2 - R N A isolation, DNase treatment and R T - R e a l T i m e - P C R 55 3.3-Results 57 3.3.1 - Profiling o f selected drug resistance-related transporters in FACS-sorted subpopulations o f A M L patient samples 57 3.3.2 - Higher expression oiMDRl and BCRP1 in the CD34+CD38-cells from nonresponders ..57 3.4 - Discussion ......59  IV  Ex vivo drug sensitivity of primitive and mature subpopulations of Acute Myeloid Leukemia and effects of ABC transporter modulation.68 4.1 - Introduction 4.2 - Materials and Methods... 4.2.1 - Exposure o f A M L cells to drugs 4.2.2 - Annexin V-Propidium Iodide assay 4.2.3 - M . T S assay..... 4.2.4 - Analysis  68 70 70 .70 71 71  4.3-Results..... 73 4.3.1 - Comparison o f the Annexin V - P I apoptotic assay to the M T S proliferation assay on C E M and C V 1 . 0 cell lines 73 4.3.2 - Adaptation o f the apoptotic assay to A M L patient cells 74 4.3.3 - Subpopulation size and patient material availability as a source o f limitation 75 4.3.4 - Higher daunorubicin resistance and larger effect o f A B C modulation in the C D 3 4 + C D 3 8 - fraction o f non-responders 75 4.4 - Discussion 78  V  Conclusion and future prospects 5.1 - Overall discussion and conclusion 5.2 - Future Prospects....  93 93 98  Bibliography  105  Appendix: Drug sensitivity curves of CR and NR patients..  115  IV  LIST OF TABLES Table 1.1: FAB classification of A M L  .  19  Table 1.2: WHO classification of A M L  19  Table 1.3: Frequencies of common recurring cytogenetic aberrations in adult A M L  20  Table 1.4: Prognostic significance of frequent chromosomal abnormalities in A M L  21  Table 1.5: List of human ABC transporters, location and physiological function  22  Table 1.6: Substrate specificity of PGP, MRP1 and BCRP1  23  Table 2.1: Patient characteristics  46  Table 4.1: % of CD34+CD38-, CD34+CD38+ & CD34- fractions in A M L patients.  81  v  LIST OF FIGURES Figure 1.1: Morphological differences between different A M L subtypes..........  24  Figure 1.2: A M L translocation (8;21) (left) compared to normal chromosomes 8 and 21 (right) 25 Figure 1.3: A M L translocation (15;17) detected in APL  ........25  Figure 1.4: Overall survival of A M L patients with favorable (A) or adverse (B) cytogenetic abnormalities compared to the group with normal karyotype.... 26 Figure 1.5: Schematic diagram of a typical ABC transporter  27  Figure 1.6: X-ray crystallography structure of MsbA  28  Figure 1.7: Proposed model for lipid A transport by MsbA  29  Figure 1.8: PGP as a classic drug pump in cancer cells  30  Figure 1.9: Prognostic significance of PGP in A M L  31  Figure 1.10: Normal human hematopoiesis  32  Figure 1.11: A M L forms a stem cell hierarchy  33  Figure 1.12: Targeting leukemic stem cells as a curative therapy....  34  Figure 2.1: Flow-chart of RT-Real Time-PCR  47  Figure 2.2: Typical dissociation curve and amplification plot of a Real Time PCR product  48  Figure 2.3: Standard curves for selected genes  49  Figure 2.4: Profiling of the ABC transporter superfamily in C E M and CV1.0  50  Figure 2.5: Profiling of the ABC transporter superfamily in patients CR#7 and NR#9  51  Figure 2.6: mRNA levels of selected ABCs in unfractionated A M L patient samples  52  Figure 3.1: FACS analysis of A M L patient sample NR#9  62  Figure 3.2: mRNA expression levels of MDR1 in FACS-sorted A M L subpopulations  63  Figure 3.3: mRNA expression levels oiMRPl in FACS-sorted A M L subpopulations  64  vi  Figure 3.4: mRNA expression levels of BCRP1 in FACS-sorted A M L subpopulations  65  Figure 3.5: Comparison of expression oiMDRl, BCRP1 and MRP1 between the CR and NR groups in the CD34+CD38- fraction 66 Figure 3.6: CD34+CD38- fraction size in CR and NR A M L patients  .....67  Figure 4.1: Apoptosis as detected by the Annexin V-PI assay  82  Figure 4.2: Schematic representation of conversion of MTS to formazan  83  Figure 4.3: Effects of PSC-833 and verapamil on daunorubicin sensitivity in C E M and CV1.0 as measured by the Annexin V-PI assay  84  Figure 4.4: Effects of PSC-833 and verapamil on daunorubicin sensitivity in C E M and CV1.0 as measured by the MTS assay  85  Figure 4.5: Toxicity assay of PSC-833 and verapamil on A M L patient cells  86  Figure 4.6: Effect of Pgp inhibition on daunorubicin sensitivity of A M L patient #18  87  Figure 4.7: Effect of Pgp inhibition on daunorubicin sensitivity of A M L patient #25  .88  Figure 4.8: Drug sensitivity and effects of ABC modulation on CR patient #1 subpopulations  89  Figure 4.9: Drug sensitivity and effects of ABC modulation on NR patient #9 subpopulations  90  Figure 4.10: Daunorubicin sensitivity of different A M L subpopulations in CR and NR patients .  91  Figure 4.11: Effects of ABC modulation on drug sensitivity in different A M L subpopulations  ..92  Figure 5.1: Models of tumor drug resistance  101  Figure 5.2: Predicting response and overcoming MDR  102  Figure 5.3. Principles of array comparative genomic hybridization..  103  Figure 5.4. C G H karyogram pf patient NR#3  104  vii  LIST OF ABBREVIATIONS ABC ABCR AE AML APL Ara-C BAALC BAC  BCRP1 bp BSEP CBF CD CEBPA CFTR CGH CR CSC Ct Cy DMSO DNA FAB FACS FBS FCS FISH  Flt3 GAPDH G-CSF GSH HSC IMDM ITD LSC MDR  MDR1 MEM MLL  MRP1 MTS NBD NOD-SCID NR  ATP-binding cassette ATP-binding cassette transporter-retina amplification efficiency acute myeloid leukemia acute promyelocyte leukemia arabinoside Brain and Acute Leukemia Cytoplasmic bacterial artificial chromosome Breast cancer resistance protein 1 base pair(s) Bile-salt export pump also known as S P G P Core-binding Factor Cluster o f Differentiation molecule C C A A T enhancer binding protein Cystic Fibrosis Transmembrane conductance Regulator comparative genomic hybridization complete remission cancer stem cells threshold cycle cyanine dimethylsulfoxide deoxy ribonucleic acid French-American-British fluorescence-activated cell sorting fetal bovine serum fetal calf serum fluorescence in situ hybridization F M S - l i k e tyrosine 3 Glyseraldehyde-3-phosphate dehydrogenase granulocytic-colony stimulating factor glutathione hematopoietic stem cell Iscove's modified Dulbecco's medium internal tandem duplication leukemic stem cell multidrug resistance Multidrug Resistance 1 minimal essential medium M i x e d Lineage Leukemia Multidrug Resistance protein 1 tetrazolium salt nucleotide-binding domain non-obese diabetic severe-combined immunodeficiency non-responsive  PB PCR PgP PI PML PS RARa RNA RT Shh SKY SL-IC SP SPGP VLB  peripheral blood polymerase chain reaction Permeability-glycoprotein propidium iodide promyelocytic leukemia phosphatidylserine Retinoic A c i d Receptor a ribonucleic acid reverse transcription Sonic Hedgehog spectral karyotyping severe-combined immunodeficiency leukemia initiating cell side population Sister Permeability-glycoprotein also known as B S E P vinblastine  ix  ACKNOWLEDGEMENTS I thank my supervisor, Dr. Victor L i n g , for giving me the opportunity to work in his laboratory. The guidance, challenge, resources and autonomy he provides are all pivotal in shaping me into an independent thinker. To Dr. Donna Hogge, my collaborator: her support and expert advice are critical to the successful completion of this project. Special thanks to Dr. Jaclyn Hung for constant encouragement and fruitful intellectual discussions. To my fellow lab members and friends, it was a pleasure and privilege working with them. I shall remember this special period o f time in my life. I thank my committee members, D r . Connie Eaves, D r . Donna Hogge, and Dr. Ross M a c G i l l i v r a y , for their help and suggestions on my thesis work. I also thank Gitte Gerhard and Leman Yalcintepe for their technical assistance. Finally, I thank my parents, my sister and brother for their unfailing love and support throughout the years.  x  To my father, who spurred me on the pursuit of knowledge. To my Father from Above: without Him I am nothing.  I  Introduction  1.1 - Challenges in cancer therapy Cancer has become the current leading cause o f death for people under age 85 in North o  America . It is a genetic disease, arising from the transformation o f a normal cell, with a derangement in normal regulation o f cell proliferation, differentiation and death. One major difficulty in the treatment of cancer stems from its heterogeneity: individual cancer patients typically show different combinations o f genetic and/or epigenetic aberrations in specific cellular pathways. This is thought to explain, at least in part, why there is no form o f therapy that is equally successful in all patients presented with the same type o f cancer. Recognition o f this heterogeneity has resulted in increasing support for the concept o f "personalized therapy" or "individual therapy" - tailoring the treatment according to individual patient condition, in particular, therapeutically targeting specific genetic abnormalities for the patient in question - as emphasized in the 2006 annual report from American Society o f Clinical Oncology . 9  There are three broadly used types o f treatment for cancer: surgery, radiation therapy and chemotherapy. While the former two are very effective on local tumors, chemotherapy remains the main form o f systemic treatment for inoperable, metastasized or more advanced cancer. Adverse side effects often ensue, however, due to drug toxicity on normal cells in the body, particularly those that must divide to maintain organ integrity. It would therefore be invaluable i f an individual patient's response to a given regimen could be predicted before deciding on the best form o f therapy available. Recent years has seen the development o f hypothesis-driven, mechanism-based drug discovery in support o f personalized therapy - designing drugs specifically targeting the molecular pathology underlying individual cancers (reviewed by  1  Collins and Workman, 2006 ). This type o f molecular-targeted therapy is best combined with 10  characterization and pre-screening o f patients with the relevant genetic alteration. One prominent example is the use o f Gleevac (imatinib, STI571), a specific inhibitor o f B C R - A B L tyrosine kinase, in chronic myeloid leukemia ( C M L ) " ' . B C R - A B L is the product o f the aberrant 1 2  Philadelphia chromosome and is present in virtually all cases o f the disease. This prototypical specific targeting o f a molecular product in cancer achieves high success in newly diagnosed C M L patients, although cases o f resistance undermines its effectiveness in more advanced cases (reviewed by Kantarjian, 2006 ). In Acute M y e l o i d Leukemia ( A M L ) , the internal tandem 13  duplication (ITD) o f the FLT3 gene is the most common mutation and hence an attractive target for therapy. Flt3 is a tyrosine kinase receptor important in regulation o f proliferation, differentiation and apoptosis o f hematopoietic progenitors . Several studies reported that 14  karyotypically normal patients with F L T 3 activation have a poorer prognosis " , and a number 15  17  of agents against the FLT3 mutation were being tested on patients bearing the mutation ' . 18  19  This thesis sought to evaluate the possible prognostic value o f A B C transporter expression in anticipating the chemotherapeutic response o f the malignant population in patients with A M L . This type o f cancer offers a number o f unique features relevant to such a study. Technically, it is possible to select patients with over 90% leukemic cell counts in their blood system, so that results are not skewed by a substantial or undefined normal cell contamination, a common problem in solid tumors. Since chemotherapy is the main form o f treatment for leukemia, almost all patients w i l l undergo a standard initial drug regimen, thereby allowing direct comparisons with treatment outcome. Furthermore, critical cancer biology models are well established for A M L , notably the origin and perpetuation o f the leukemia by a rare subset o f "leukemic stem cells" o f known phenotype '  2 20  (Discussed in Section 1.7). This forms the basis o f  2  my main hypothesis that it is the rare leukemic stem cells, not the mature or differentiated cells comprising the majority o f leukemia, which dictate clinical outcome. This is supported by results of this study, which highlights the value o f determining A B C transporter levels in a leukemic stem cell-enriched population to predict treatment response. I then went on to illustrate the effects o f A B C inhibitors on this small cell fraction, proposing their use as drug-sensitizing agents in combination with chemotherapy to improve initial treatment outcome for otherwise non-responding patients. This is especially crucial for A M L , in which its fast progression augments the importance of choosing the best treatment at diagnosis.  1.2 - Acute Myeloid Leukemia: an overview Leukemias make up ~ 2 % o f adult cancers  2 1  but comprise a heterogeneous group o f  diseases. The lymphoid leukemias affect the lymphoid lineages (notably B - and T-cells) and the myeloid leukemias affect primarily the myeloid lineages including the granulocytes and monocytes/macrophages, red blood cells, and megakaryocytes. Leukemia can be classified as either acute or chronic. A M L is a malignancy o f the myeloid elements, the hallmark being a block in normal differentiation, leading to the massive accumulation o f immature leukemic "blast cells". This usually results in rapid and severe disruption o f normal bone marrow function, which can take the clinical form o f anemia (decrease in hemoglobin), fever and infection (decrease in white cells), and bleeding and bruising (decrease in platelets). Accumulation o f leukemic cells in other tissues is also common, such as the lymph nodes, spleen and skin. A M L requires urgent diagnosis and treatment. If left untreated, the disease results in death within weeks or days. In contrast, the chronic leukemias are characterized by unregulated proliferation  3  and overexpansion o f a range o f differentiated cells, are slow-growing and usually progress over 21'  a period o f several years . The incidence o f A M L increases sharply with age, from less than 1/100,000 under age 35 to 15/100,000 over 75. Recent Canadian statistics report an incidence o f 816 new cases in 2001 and a mortality rate of 690 deaths in 2003 . While the etiology o f A M L remains largely 8  unknown, a number of risk factors have been identified, including radiation and/or chemical exposure, tobacco use, prior cancer-related chemotherapy or radiation therapy, old age, genetic syndromes, and a history o f prior blood disorders. Certain recurring cytogenetic abnormalities have been found to be closely associated with A M L and form a major area o f investigation (see Section 1.3). Diagnosis is based on morphological data which further classify A M L into eight subgroups (MO-7) under the French-American-British ( F A B ) scheme established in 1976 (Table 1.1). A n example o f distinct morphological and histochemical differences between A M L subtypes is shown in Figure 1.1. M o r e recently the W o r l d Health Organization ( W H O ) has proposed a new classification which includes immunophenotyping, cytogenetics and clinical features to allow a more prognostically relevant description o f A M L (Table 1.2). Given that A M L is a rapidly progressing disease with a fatal outcome i f not adequately controlled, initial treatment is often targeted at eradicating leukemic blasts and re-establishing normal bone marrow function. This is usually achieved through high-dose chemotherapy with general supportive measures such as blood cell transfusions, antibiotics administration, and leukapheresis to temporarily clear patient blood o f blasts. The standard remission-induction chemotherapeutic regimen for A M L is a combination o f cytosine arabinoside (Ara-C) and an anthracycline such as daunorubicin. Clinical M D R (Multidrug Resistance), cross resistance arising in cancer cells to a wide range o f chemically unrelated drugs, is commonly observed and  4  presents a major problem in A M L therapy. While - 7 0 % o f patients achieve remission with initial therapy, approximately 75% o f these w i l l relapse within 2 years o f diagnosis in spite o f additional consolidation chemotherapy . In addition, 20-30% o f patients are unresponsive even 21  to initial chemotherapy. Overall, long-term remissions are obtained in only 2 5 % o f patients . A 21  small number o f patients under 60 years o f age and who have a suitable histocompatible donor are eligible for curative treatment supported by an allogeneic bone marrow or mobilized peripheral blood transplant. However, chemotherapy necessarily remains the main form o f postremission treatment. Given that chemotherapy is a highly invasive treatment with both long term and short term side effects, it would be valuable to be able to predict which patients are more likely to benefit from current chemotherapy treatments. In addition, the poor prognostic rate raises the need for more effective curative therapies.  1.3 - Current prognostic factors for predicting chemotherapeutic response in AML Older A M L patients (over 65 years) have a poor prognosis compared to young patients due to their lower drug tolerance and higher toxicity during high-dose chemotherapy . They are 21  also considered clinically ineligible for allogeneic transplant therapy. Other than age, the current most important prognostic variable for patients with A M L is the detection o f cytogenetic 99  94  abnormalities in a diagnostic bone marrow sample ~ . Almost 200 different recurring acquired cytogenetic aberrations have been identified in - 5 0 % o f A M L patients . The most common ones 24  are listed in Table 1.3. These abnormalities, in the forms o f translocations, inversions, deletions monosomies and trisomies, play an important role in determining the biological basis o f A M L . Intense molecular studies o f specific genes at the sites o f aberration revealed that they are usually  5  involved in normal blood cell development and homeostasis. Most inversions and translocations in A M L result in gene fusion products that can dysregulate proliferation, differentiation or apoptosis o f blood cell precursors  . One o f the best known translocations (8;21) (Figure 1.2) is  frequently associated with A M L subtype M 2 . It places the gene AML1 on 21q22 beside the gene E T O (a transcription factor) on 8q22. A M L 1 is the a subunit o f the heterodimeric transcription factor Core Binding Factor ( C B F ) critical for normal hematopoiesis  . The fusion protein retains  the ability to bind to A M L 1 consensus regions and acts as a competitive inhibitor for the normal AML1 product, resulting in defective myeloid differentiation. Another frequent abnormality, inversion (16), is molecularly related to t(8;21) because it disrupts the gene encoding the P subunit o f C B F . A M L subtype M 3 , also known as acute promyelocytic leukemia ( A P L ) , is characterized by translocation (15; 17) (Figure 1.3). This fuses the retinoic acid receptor a gene ( R A R a ) on 17q 12-21 with the promyelocytic leukemia gene ( P M L ) on 15q22, which has a dominant effect on normal R A R a , antagonizing its differentiation function . Large cooperative studies have documented a significant relationship between detection of non-random chromosomal abnormalities and disease outcomes, including complete remission, disease-free survival and overall survival ' " .. The prognostic significance o f common abnormalities is summarized in Table 1.4. The presence o f certain aberrations including inversion (16) and translocations (8;21) and (15; 17) is associated with significantly better overall survival compared to normal cytogenetics (Figure 1.4A). While use o f the novel targeted 6  therapeutic drugs all-trans retinoic acid ( A T R A ) ' 3 1  3 2  and arsenic trioxide " have drastically 33  36  improved outcome for patients with t(15;17), the molecular basis o f higher sensitivity to chemotherapy for t(8;21) or inv(16) A M L remains to be elucidated. O n the other hand, other chromosomal changes such as -5, -7 and del(5) are frequently related to poor prognosis  6  compared to patients whose blasts are cytogenetically normal (Figure 1.4B), although the 6  molecular mechanisms responsible remain unclear. Despite the prognostic value o f the many cytogenetic abnormalities identified almost 50% o f patients present with an apparent normal karyotype. A patient's karyotype is routinely determined by traditional chromosome-banding and less frequently by newer methodologies such as spectral karyotyping ( S K Y ) or fluorescence in situ hybridization (FISH). Studies have confirmed the validity and consistency o f these karyotyping techniques in detecting translocations and large-scale gains or losses o f genomic D N A " . It w i l l be useful, as suggested 3 7  4 0  by large-scale mieroarray gene expression studies ' , to try to subcategorize patients with a 41  42  normal karyotype into groups with varying prognosis. Indeed, ongoing studies have suggested a few genes that may act as possible prognostic markers for A M L patients with a normal karyotype , including F M S - l i k e tyrosine 3 (FLT3) (see Section 1.1), mixed lineage leukemia 43  (MLL) , 44  C C A A T enhancer binding protein (CEBPA) , 45  acute leukemia cytoplasmic {BAALC) . 41  nucleophosmin (NPM1)  46  and brain and  Studies on more biomarkers should facilitate predicting  treatment outcomes within the cytogenetically normal group. The objective o f this thesis was to investigate the prognostic value o f expression o f the multidrug resistance-related ATP-binding cassette transporters ( A B C transporters).  1.4 - ABC transporters: an Overview ATP-binding cassette ( A B C ) transporters represent the largest transmembrane protein superfamily in eukaryotes and prokaryotes and are important factors in drug resistance. They are ATP-dependent protein transporters that actively pump a wide range of substrates across biological membranes. To date, 48 A B C transporters have been identified in humans ' ' . Figure 1.5 shows 1  48  49  7  the structural organization of the A B C transporter protein and gene. A B C proteins share a highly conserved ATP-binding cassette ( A B C ) , also known as nucleotide-binding domain ( N B D ) , which consists of the characteristic motifs Walker A (G-X2-G-X-G-K-S/T-T/S-X4-hydrophobic) and Walker B(R-X-hydrophobic2-X -P/T/S/A-X-hydrophobic4-D-E-A/P/C-T-S/T/A-AG-hydrophobic2  D ) . The A B C gene also contains motif C , or the "signature m o t i f bearing the sequence 5 0  (hydrophobic-S-X-G-Q-R/K-Q-R-hydrophobic-X-hydrophobic-A) organized between motifs A and B . A fully functional unit of A B C transporters consists of two similar halves, each containing a 5 1  transmembrane domain and an ATP-binding domain. Members of the superfamily can either be "full transporters" with both halves present or "half transporters" requiring dimerization for function . 1  The crystal structure of M s b A , an A B C transporter homolog in Escherichia coli, has been described by Chang et al . This provides insights into the molecular structure and transport 5  mechanism of A B C transporters. A s seen in Figure 1.6, M s b A consists o f two similar halves embedded in the lipid bilayer, organized much like a clamp hinging on extracellular connecting loops. The twelve transmembrane domains collectively serve as the substrate binding site. Based on this structure, the same group also proposed a general model for lipid transport by A B C transporters (Figure 1.7). In this "flippase" model, the transmembrane chamber first interacts with and intercepts the substrate, causing conformational changes that result in A T P hydrolysis by the N B D . This in turn causes a conformational shift that brings the two N B D s together. The resulting change in the chamber now produces an energetically unfavorable environment for the hydrophobic substrate and it "flips" from the inner to the outer membrane layer. After the flip, the N B D s separate leading to the expulsion o f the substrate to the extracellular environment. Recently, Dawson and Locher proposed an alternate "access and release" model based on their crystal structure on the bacterial  8  A B C transporter Savl866, that postulates conformational changes to reflect the hydrolysis state o f ATP . 5 2  The human A B C transporters are further categorized into subfamilies A to G based on sequence homology and structural similarity. Table 1.5 lists the current subfamily members, nomenclature and function. Human A B C s perform a variety of physiological functions by facilitating unidirectional shuttling o f compounds within the cell as part of a metabolic process or outside the cell to other organs ' . Unlike other types o f transport proteins, one unique 1 49  characteristic of the members of the A B C superfamily are their wide substrate specificity. The first  53  and best characterized A B C transporter, MDR1 (encoding the P-glycoprotein, PGP), is a promiscuous transporter of hydrophobic substrates. Physiologically, P G P is important in removing toxic metabolites from cells, especially in the brain ' . Other A B C s perform functions ranging 54  55  from liver bile salt excretion (SPGP) to vitamin A transport in photoreceptors ( A B C R ) to cholesterol transport ( A B C A subfamily). A number of A B C transporters have been linked to genetic disorders, for example C F T R (mutations o f which causes cystic fibrosis) and S P G P (mutations o f which result in progressive familial intrahepatic cholestasis 2 ) . The promiscuity of A B C 56  transporters, however, becomes a major clinical problem when cancer occurs: some A B C s are able to transport multiple chemotherapeutic drugs out of cancer cells, resulting in Multidrug Resistance.  1.5 - ABC transporters and multidrug resistance in cancer Overexpression o f members o f the ATP-Binding-Cassette ( A B C ) transporter family is found to be a main factor for M D R in cancer. There are three main A B C transporters involved in multidrug resistance: the classical FGP/MDR1, the multidrug resistance associated protein (MRP1), and the breast cancer resistance protein (BCRP1, ABCG2). MDR1, the prototype A B C  9  transporter, was first cloned from a drug-selected cell line displaying multidrug resistance . Knockout studies showed that mdrla/mdrlb(-l-) (homologs o f human MDR1) mice had a high accumulation o f drug levels in many tissues, especially the brain, confirming that P G P confers a general detoxification function against xenotoxins . The observation that not all M D R cell lines 57  overexpress MDR1 led to the discovery o f another multidrug resistance transporter, multidrug resistance protein 1 (MRP1). Cloned from a small-cell lung cancer cell line, MRP I transports drugs that are conjugated to glutathione ( G S H ) via the G S H reductase pathway . Knockout 58  studies showed that MRP I is important in inflammation as well as for detoxification in the brain . The third multidrug resistance transporter, BCRP1, was identified in cell lines selected 59  for mitoxantrone resistance ' . Unlike P G P and M R P 1 which are full A B C transporters, 60 61  B C R P 1 is a half A B C and is thought to function as a homodimer. Knockout studies have suggested an in vivo role for this transporter in tissue detoxification and heme metabolite transport under hypoxic conditions ' . Hence all three MDR-related A B C transporters are 62  63  physiologically important in protecting normal cells from a broad array o f xenobiotics. This versatile defense mechanism, however, can be utilized by cancer cells for protection against a wide variety o f chemotherapeutic drugs. Early clinical studies found high expression o f P G P in many types o f cancer, such as the leukemias, colon, kidney, adrenocortical and hepatocellular cancers ' . Amplification o f MDR1 is frequently reported in many drug-selected M D R cell 64  65  lines, although this is not a commonly observed mechanism for overexpression in clinical cases . MRP1 is overexpressed in leukemia, esophageal cancer and non-small cell lung 66  cancers . Expression ofBCRPl  has been detected in the leukemias , gastric cancer,  hepatocellular carcinoma, endometrial cancer, colon cancer, melanoma and lung cancer ' . 69  70  Figure 1.8 illustrates the mechanism o f drug resistance via these transporters: the incoming  10  chemotherapeutic drug is intercepted by the transporter in the plasma membrane o f the cancer cell and expelled, thereby failing to reach the intracellular site o f action, resulting in drug resistance and failure o f therapy. Other A B C transporters have been implicated in drug resistance. For instance, A B C A 2 71  can confer resistance to estramustine, a nitrogen mustard derivative o f oestradiol . S P G P , an 79  A B C that shares high homology to P G P , is reported to confer resistance to paclitaxel . Members o f the A B C C subfamily, commonly known as M R P 2 to M R P 9 , bear similarities to M R P 1 and have the potential to confer drug resistance. Given that many members o f the A B C superfamily have not been well characterized, they may also have the capacity to efflux substrates o f clinical interest that has not yet been identified. Table 1.6 lists the common chemotherapeutic drugs that are known substrates o f these transporters. P G P and B C R P 1 preferentially extrude large hydrophobic, positively charged molecules ' 60  7375  , while M R P 1 can extrude both hydrophobic uncharged molecules and water-  soluble anionic compounds ' " . A l l three display capacity to transport daunorubicin , the 58  76  79  80  common drug used for A M L , although daunorubicin efflux by M R P 1 is strictly dependent on G S H levels . M a n y agents have been investigated in an effort to reverse PGP-mediated clinical 81  M D R . These are generally competitive inhibitors that are substrates for the A B C transporters. Verapamil is a commonly used inhibitor that can modulate activity o f a number o f A B C transporters (most effective against PGP). More specific inhibitors have been developed for each transporter, for example PSC-833 for P G P " . Clinical trials using A B C inhibitors yielded 8 2  8 4  mixed results in A M L , with complications including high bone marrow and neurological toxicity, confounding interpretations on the usefulness o f these agents. Nonetheless, successful therapy has been reported for other types o f cancer. In highly drug-resistant retinoblastoma for  11  example, combination o f P G P inhibitor cyclosporin A with chemotherapy resulted in marked increase in relapse-free rate, demonstrating the value o f A B C modulation in improving clinical outcome . 85  1.6 - ABC transporters in A M L Expression o f MDR1 in leukemic cells likely contributes to chemotherapy resistance in A M L patients. Treatment failure can be observed either as intrinsic chemotherapy resistance at diagnosis or at relapse. Overexpression o f P G P is the most extensively studied mechanism o f M D R in A M L . The presence o f P G P detected by antibody staining as measured by flow cytometry has been demonstrated in 20% to 75% o f de novo A M L patients according to different on  OQ  studies " . Expression o f P G P in patient peripheral blood or bone marrow is associated with lower remission rates ' , shorter overall survival and lower disease-free survival (Figure 87  88  1.9) ' ' ' . In addition, a number o f studies have also reported an increase in P G P during 4  7  90  91  relapsed disease " . It is suggested that PGP-positive cells might escape initial chemotherapy 92  94  and remerge at relapse, or alternately, MDR1 expression may be induced during chemotherapy. Ex vivo studies have shown that P G P expression decreases intracellular accumulation o f daunorubicin in leukemic cells, and is reversed by P G P inhibitors ' ' . A n additional antiapoptotic role has also been attributed to P G P in that it appears to protect leukemic cells from caspase-dependent programmed cell death " . 97  99  Several investigations on the M R P 1 transporter suggested a role in A M L drug resistance " , but larger studies failed to identify a consistent relationship between M R P 1 and 100  prognosis '  102  . Studies on the expression o f B C R P 1 in A M L have also yielded mixed results.  While some showed that up-regulation o f BCRP1 is associated with a poor p r o g n o s i s  104  '  105  and is  12  common during relapse  , others reported a lack o f consistent overexpression in A M L  patients " . Several studies have reported the combination o f different drug-resistant 107  109  transporters, or their co-expression with cell survival factors, important in drug resistance ' ' . Almost all past studies, however, have been limited to the three MDR-related 89  100  109  transporters. Another important feature oiMDRl  and BCRP1 are their changes in expression during the early  differentiation o f very primitive hematopoietic cells. A s described in the following sections, this feature has important clinical implications in the treatment o f A M L .  1.7 - The leukemic stem cell model and its implications in drug response The concept o f "cancer stem cells" has emerged as an important theme for cancer research in the past few years. Normal stem cells are defined by their dual capacity to regenerate themselves through self-renewal mechanism and to produce mature cells through differentiation . The "cancer stem cell model" proposes that a tumor is similarly sustained by a 110  biologically distinct subpopulation o f "cancer stem cells" (CSC), with the same ability to perpetuate the production o f progeny with limited proliferative ability. Three observations support the existence o f C S C s in human cancers. First, only a small fraction o f a tumor has the capacity to regenerate a new tumor, operationally demonstrable upon transplantation into immunodeficient mice. Second, these tumor-initiating cancer cells can be identified and prospectively isolated by a distinct phenotype, usually based on flow cytometric or immunomagnetic detection o f differentially expressed surface antigens. Third, secondary tumors regenerated by these cells contain an array o f mixed tumorigenic and non-tumorigenic cells, recapitulating the heterogeneity o f the original tumor . Hence irrespective o f the origin o f C S C , 111  13  this population bears the hallmarks o f stem cells - self-renewal and "differentiation" into a functional hierarchy o f "primitive" (tumorigenic) and "mature" (non-tumorigenic) cells. The C S C model was first and best developed in l e u k e m i a ' . In 1979, Minden and 2  112  colleagues first described colony forming cells, termed "blast progenitor cells", from A M L patient peripheral b l o o d  113  , suggesting that a subpopulation o f leukemic cells was clonogenic. It  was, however, John D i c k and colleagues who provided direct evidence for the leukemic stem cell ( L S C ) model . They described a primitive leukemic cell they termed S L - I C (SCID leukemia initiating cell) that can initiate A M L in mice. The SL-ICs from most patients tested are CD34++/CD38-, a surface phenotype similar to that of primitive normal cells that can regenerate normal hematopoiesis in immunodeficient mice (Figure 1.10). In addition to its tumorigenic potential, the S L - I C also shows the capacity to differentiate into the non-dividing leukemic cells similar to those which constitute the majority population in the malignant clone in the A M L patient from which it was isolated. The authors conclude that it is possible to isolate a small fraction enriched with L S C activity (0.2-100 stem cells in 10 blast cells), and the common 6  surface properties and hierarchical organization support the hypothesis that the L S C originates by transformation o f an initially normal hematopoietic stem cell (Figure 1.11). Nevertheless, it has also been shown that leukemic stem cells can be generated from more mature progenitors that reacquire stem cell properties " . 114  116  C S C s appear to share many common properties with normal stem cells . For example, 3  overlap in their regulation by the Wnt, Notch and Sonic hedgehog (Shh) pathways, are associated both with oncogenesis and normal stem cell renewal . A s well, C S C can share normal stem cell 117  properties that allow a long life-span, such as protection against cytotoxins via expression o f A B C transporters (see section 1.8 below).  14  Validation and adoption o f the leukemic stem cell model calls for a paradigm shift in the treatment o f the disease. Figure 1.12 gives a schematic representation o f new treatment strategies that are likely more effective. Because the leukemia is sustained by the rare L S C s , this small population must be included as a target for effective therapies rather than just the majority o f blast cells that have very limited proliferative ability. Existing drug therapies, however, are commonly targeted against the bulk leukemic population or their immediate precursors. Although a dramatic initial response can often be achieved, i f the L S C s are not also effectively eliminated they can eventually regenerate the disease. Hence to achieve more durable responses or even cure, there is a need for novel treatment methods more specifically directed against this primitive subpopulation.  1.8 - ABC transporters in normal and leukemic stem cells In recent years, MDR-related A B C transporters, in particular MDR1 and BCRP1, have been associated with stem cells o f the hematopoietic system . Early investigation by Chaudhary 118  and Roninson first showed elevated expression oiMDRl  in primitive (CD34+) normal  hematopoietic c e l l s . A s discussed above, C D 3 4 , a cell surface phosphoprotein, is a marker for 119  both normal hematopoietic stem cells ( H S C ) and L S C s . Engraftment studies in humans baboons  121  and m i c e  122  ,  demonstrated the C D 3 4 + population to be highly enriched with stem cell  repopulation activity. It has also been suggested that C D 3 4 expression may be in part regulated by the activation state o f stem c e l l s . Further studies demonstrated a marked decline in MDR1 123  expression during differentiation ' . L i k e MDR1, expression ofBCRPl 119  124  is high in normal  human H S C , drops dramatically in more committed progenitors, and remains low in mature hematopoietic c e l l s . In addition, BCRP1 is reported to be a molecular determinant o f the side 125  15  population (SP) in hematopoietic cells, a small distinct cell fraction with enriched H S C repopulating activity in adult mouse bone m a r r o w  126  or human fetal l i v e r  127  but not vice versa.  The SP, defined by its ability to efflux the fluorescent dye Hoechst 33342, was first identified in 128  murine bone marrow cells  126 129 131  and later also found in many different human tissues  '  "  , in  spite o f its failure to be detected on normal adult human H S C . One important property o f the SP is its ability to exclude a number o f drugs, reflective o f the transport activity o f A B C transporters. Interestingly, Uchida and colleagues reported that the dye (Rhodamine 123 and 132  Hoechst 33342) efflux ability o f P G P and B c r p l are unstable in murine H S C s  . This fluctuation  appears to parallel changes in other H S C markers such as C D 3 4 and C D 3 8 and relate to the activation state o f H S C s , suggesting a common control mechanism operated by a cell cycle checkpoint. One likely function o f A B C transporters in stem cells is to protect this population from toxic substances over their long life-span. Another possible function is that these transporters can efflux regulatory molecules that can alter stem cell fate. For example, studies conducted by our lab revealed that P G P is a functional bile salt transporter in S P G P knockout mice, compensating for the lack o f S P G P . The recently elucidated connection between A B C transporters and normal hematopoietic stem cells sheds a new light on the drug resistance property o f C S C s . It has been suggested that C S C s can employ the same protective mechanisms operating in normal stem cells to defend themselves from chemotherapeutic drugs. Indeed, an increasing number o f studies are associating A B C transporters to C S C s . For example, P G P expression has been correlated to C D 3 4 positivity in A M L  1 3 3 , 1 3 4  . A s well, the SP has been identified in neuroblastoma, breast 135 139  cancer, ovarian cancer, glioblastoma and gastrointestinal cancer cell lines  "  . These studies  demonstrated that this small subpopulation has both enriched tumorigenicity and high drug  16  extrusion capacity. W u l f and colleagues found the SP to be detectable in most A M L patients, displays significantly increased drug efflux ability, and is able to regenerate the disease in N O D SCID mice  140  . Studies by Feuring-Buske and Hogge confirmed the prevalence o f SP in A M L ,  although SP+CD34+CD38- cells appeared to represent normal rather than leukemic primitive cells in A M L patients . 141  The emerging paradigm on cancer biology posits functional heterogeneity within a cancer and the existence o f a small distinct group o f C S C reminiscent o f normal stem cells. It remains to be investigated whether higher expression and activity o f MDR-related A B C transporters contributes to higher drug tolerance in L S C , the presumptive subpopulation responsible for perpetuation o f A M L .  17  1.9 - Thesis objectives The overall objective o f this thesis was to investigate A B C transporter expression as a possible predictive factor for initial drug response in A M L patients. M y original hypothesis was that upregulation o f A B C transporters is responsible for the lack o f response to initial chemotherapy in A M L patients. In that context, I investigated the expression level o f A B C transporters in total A M L blast cells and in different subpopulations along the leukemic hierarchy - C D 3 4 C D 3 8 " (most primitive), C D 3 4 C D 3 8 , and CD34" (most mature). Hence the +  +  +  three goals o f this thesis were: 1. To profile m R N A expression patterns o f the A B C transporter superfamily in bulk A M L patient materials, and compare profiles between patients that responded or failed to respond to initial chemotherapy. 2. To examine expression patterns o f key MDR-related A B C transporters in primitive and mature A M L subpopulations o f responders and non-responders. 3. To determine and compare the ex vivo drug sensitivity o f CD34+CD38-, CD34+CD38+, and C D 3 4 - A M L subpopulations o f responders and non-responders. The first and second goals are addressed in Chapters 2 and 3 where the RT-Real-TimeP C R technique was utilized to profile m R N A levels o f A B C transporters in unsorted and sorted populations o f A M L patient samples. In Chapter 4,1 investigated the ex vivo drug sensitivity o f sorted leukemic subpopulations using the fluorescence-based Annexin V - P I assay for apoptosis.  18  Table 1.1. FAB classification of A M L . FAB MO Ml  Morphology M i n i m a l l y differentiated M y e l o b l a s t s leukemia without maturation  M2  M y e l o b l a s t s leukemia with maturation  M3  Hypergranular promyelocytic leukemia  M4  Myelomonocytic blasts  M4Eo  Variant, increase i n marrow eosinophils  M5  Monocytic leukemia  M6  Erythroleukemia  M7  Megakaryoblastic leukemia  Table 1.2. WHO classification of A M L . A M L with recurrent cytogenetic translocations A M L with t(8;21)(q22;q22) A M L l / C B F a l p h a / E T O Acute promyelocytic leukemia: A M L with t( 15; 17)(q22;ql2) and variants P M L / R A R a l p h a A M L with abnormal bone marrow eosinophils inv(16)(pl3;q22) vagy t(16;16)(pl3;q22) CBFbeta/MYHl A M L with 1 lq23 M L L abnormalities A M L with multilineage dysplasia With prior M D S Without prior M D S A M L with myelodysplastic syndrome, therapy related Alkylating agent related Epipodophyllotoxin related Other types A M L not otherwise categorized A M L minimally differentiated A M L without maturation A M L with maturation Acute myelomonocytic leukemia Acute monocytic leukemia Acute erythroid leukemia Acute megakaryocyte leukemia Acute basophilic leukemia Acute panmyelosis with myelofibrosis M y e l o i d sarcoma Acute Leukemias of ambiguous lineage  Table 1.3. Frequencies of common recurring cytogenetic aberrations in adult A M L . Cytogenetic  Cooperative Group Study (No. of patients)  abnormality  Adults total (n - 4257)  CALGB  MRC  SWOG/ECOG  (n = 1311)  (n= 2337)  (n = 609)  No.(%)  No. (%)  No. (%)  582 (44)  1096 (47)  244 (40)  1922 (45)  +8  123 (9)  211 (9)  53(9)  387 (9)  -7/7q-  95(7)  209 (9)  52 (9)  356 (8)  t(15;17)(q22;q21)  88 (7)  210(9)  27 (4)  325 (8)  -5/5q-  86 (7)  183 (8)  36(6)  305 (7)  t(8;21)(q22;q22)  81(6)  104 (4)  50(8)  235 (6)  inv(16)  96 (7)  53 (2)  53 (9)  202 (5)  -Y  58(4)  NA  20 (3)  78 (4)  t/inv(llq23)  54 (4)  45 (2)  42 (7)  141 (3)  +21  28 (2)  51(2)  NA  79 (2)  abn(17p)  30 (3)  NA.  12(2)  42(2)  del(9q)  33 (3)  37(2)  17(3)  87 (2)  inv(3)  12(1)  61(3)  12(2)  85 (2)  Complex with  135 (10)  NA  71 (12)  206(11)  99 (8)  222(9)  53 (9)  374 (9)  None (normal  No. (%)  karyotype)  > 3 abn Complex with > 5 abn  Research C o u n c i l ' ; S W O G / E C O G , Southwest Oncology Group/Eastern Cooperation 6  29  Oncology Group ; abn, abnormality; N A , not available. Modified from Mrozek et al, Blood  Reviews, 2004 . 24  Table 1.4. Prognostic significance of frequent chromosomal abnormalities in A M L .  Good  Standard  Poor  inv(16)  Normal  -5, del(5q)  t(8:21)  +8  -7  t(15;17)  +21  Abnormal 3q  +22  Complex  del(7q) del(9q) Abnormal 1 lq23 A l l other structural abnormalities Modified from Grimwade et al, Blood, 1998 .  21  Table 1.5. List of human ABC transporters, location and physiological function. Gene (subfamily) ABCAI ABCA2 ABCA3 ABCA4 ABCA5 ABCA6 ABCA7 ABCA8 ABCA9 A B C A 10 A B C A 12 A B C A 13 ABCB1 ABCB2 ABCB3 ABCB4 ABCB5 ABCB6 ABCB7 ABCB8 ABCB9 ABCB10 ABCB11 ABCC1 ABCC2 ABCC3 ABCC4 ABCC5 ABCC6 ABCC7 ABCC8 ABCC9 ABCC10 ABCC11 ABCC12 ABCD1 ABCD2 ABCD3 ABCD4 ABCE1 ABCF1 ABCF2 ABCF3 ABCGI ABCG2 ABCG4 ABCG5 ABCG8  Common name ABC1 ABC2 ABC3, A B C C ABCR  M D R 1/PGP TAP1 TAP2 PGY3, M D R 3  TAPL SPGP, BSEP MRP1 MRP2 MRP3 MRP4 MRP5 MRP6 CFTR SUR1 SUR2 MRP7 MRP8 MRP9 ALDP ALDR PMP70 PMP69 OABP, RNASELI ABC50  WHITE 1 BCRP1,ABCP WH1TE2 WH1TE3 WHITE4  Location 9q31.1 9q34 16pl3.3 Ip22.1-p21 17q24 17q24 19pl3.3 17q24 17q24 17q24 2q34 7pll-qll 7p21 6p21 6p21 7q21.1 7pl4 2q36 Xql2-ql3 7q36 12q24 lq42 2q24 16pl3.1 10q24 17q21.3 13q32 3q27 16pl3.1 7q31.2 llpl5.1 12pl2.1 6p21 16qll-ql2 16qll-ql2 Xq28 12qll-ql2 Ip22-p21 14q24.3 4q31 6p2133 7q36 3q25 21q22.3 4q22 llq23 2p21 2p21  Expression Ubiquitous Brain Lung Photoreceptors Muscle, heart, testes Liver Spleen, thymus Ovary Heart Muscle, heart Stomach Low in all tissues Adrenal, kidney, brain A l l cells A l l cells Liver Ubiquitous Mitochondria Mitochondria Mitochondria Brain, testis, spinal cord Mitochondria Liver Lung, testes, P B M C Liver Lung, intestines, liver Prostate Ubiquitous Kidney, liver Exocrine tissue Pancreas Heart, muscle Low in all tissues Low in all tissues Low in all tissues Peroxisomes Peroxisomes Peroxisomes Peroxisomes Ovary, testes, spleen Ubiquitous Ubiquitous Ubiquitous Ubiquitous Placenta, intestines Liver Liver, intestines Liver, intestines  Function Cholesterol efflux  N-retinylidene-PE efflux  Multidrug resistance Peptide transport Peptide transport PC transport Iron transport Fe/S cluster transport Peptide transport Bile salt transport Drug resistance Organic anion efflux Nucleoside transport Nucleoside transport Chloride ion channel Sulfonylurea receptor  V L C F A transport regulation  Oligoadenylate binding  Cholesterol transport Toxin efflux, drug resistance Sterol transport Sterol transport  Modified from Dean et al, Genome Research, 2001 .  22  Table 1.6. Substrate specificity of PGP, MRP1 and BCRP1. Gene ABCB1  Protein PGP  Non-chemotherapy substrates  Chemotherapy substrates  Neutral and cationic organic  Doxorubicin, daunorubicin,  compounds, many commonly used  vincristine, vinblastine,  drugs  actinomycin-D, paclitaxel, docetaxel, etoposide, teniposide, bisantrene, STI571  ABCC1  MRP1  86  Glutathione and other conjugates,  Doxorubicin, daunorubicin,  organic anions, leukotriene C 4  epirubicin, etoposide, vincristine, methotrexate  ABCG2  BCRP1, ABCP  Prazosin  CO H(L TQ  '  Doxorubicin, daunorubicin, mitoxantrone, topotecan, S N 2g60,73-75  23  F i g u r e 1.1. M o r p h o l o g i c a l differences between different A M L subtypes. A M L subtypes M l ( A & B ) and M 6 (C & D) are shown as examples. A , M l myeloblasts with eccentric nuclei. B , Sudan Black stain o f M l blasts. C , M 6 erythroblasts. D , coarse P A S stain o f M 6 blasts. Courtesy o f Haematological Malignancy Diagnostic Service http://www.hmds.org.uk.  24  8 F i g u r e 1.2. A M L translocation (8;21) (left) compared to n o r m a l chromosomes 8 and 21 (right). Courtesy o f Guide for Detection o f M R D in A M L www.meds.com/leukemia/guide.  detfl 7)  •d«tl5)  'lb: mm  dft *  "^1  ail  '  \  ^  •  PAC 833D9  9  P A C 933118  F i g u r e 1.3. A M L translocation (15;17) detected i n A P L . T w o probes (red: chromosome 15, green: chromosome 17) are utilized to visualize the translocation (yellow arrows). Courtesy o f Atlas o f Genetics and Cytogenetics in Oncology and Hematology.  25  25  -I  2  3 Years from entry  normal <n=680) -5{rt=26) -7 <n=01) del(5q) (rt=28) abn(3q)(n=40) complex (n=95)  100  Years from entry  Figure 1.4. Overall survival of A M L patients with favorable (A) or adverse (B) cytogenetic abnormalities compared to the group with normal karyotype. Reproduced from Grimwade et al, Blood, 1998 .  26  B  Figure 1.5. Schematic diagram of a typical A B C transporter. A , an A B C protein is embedded in the lipid bilayer o f a cellular membrane (yellow). The transmembrane domain is depicted as blue rectangles and the N B D as red circles. B , Common sequence organization o f the ATP-binding cassette o f an A B C gene, with Walker motifs in the order A - C - B . Reproduced from Dean et al, Genome Research, 2001 \  27  Figure 1.6. X-ray crystallography structure of MsbA. A , V i e w o f dimer looking into the chamber opening. The transmembrane domain, NBD, intracellular domain, and connecting loops are in red, cyan, dark blue and green, respectively. A potential substrate lipid A is shown to the right o f the structure. Solid and dotted green lines represent the boundaries o f the membrane bilayer leaflets. B, V i e w o f M s b A from the extracellular side, perpendicular to the membrane with lipid A . Reproduced from Chang et al, Science,  2001 . 5  28  Substrate bindhg and recruitment  Chamber opening and substrate expulsion  Chamber closure ant! substrate flip-flop  Figure 1.7. Proposed model for lipid A transport by MsbA. Stages 1 to 3 begin at top and proceeds clockwise. See text for details. (1) L i p i d A binding, triggering o f A T P hydrolysis, and recruitment o f substrate to chamber. (2) Closure o f the chamber and translocation o f lipid A . Interaction between the two N B D s is possible. (3) Opening o f the chamber, movement o f T M 2 / 5 , release o f lipid A to the outer bilayer leaflet, and nucleotide exchange. A small yellow rectangle and a green circle denote the hydrophobic tails and sugar head groups o f lipid A , respectively. The transmembrane domain ( T M ) , intracellular domain (ICD), and nucleotidebinding domain ( N B D ) are labeled. Blue regions indicate positive charge lining the chamber, and purple regions represent the intracellular domain. The gray region on the outer membrane side o f the chamber is hydrophobic. Red and black arrows show the movement o f substrate and structural changes o f M s b A , respectively. Reproduced from Chang et al, Science, 2001 . 5  29  Anticancer drug Outside  Pgp  Anticancer drug #  Reversal agent  Outside  I  -mar  Inside  Antimitotic activity  MembraneInside Antimitotic activity  Figure 1.8. PGP as a classic drug pump in cancer cells. Left, P G P is located in the plasma membrane. A drug molecule going from the outside to the inside o f the cell is intercepted by P G P in the membrane and subsequently transported out o f the cell, resulting in drug resistance. Right, Addition o f a reversal agent such as PSC-833 inhibits P G P transport activity by binding to its substrate binding site. Reproduced from Chemtech 1998, 28 (6), 31-36.  30  .1" 'o  "j Rhl23-efflux positive 10  20  30  AO 50  60  overall survival <in months)  70  80  9C  Log Rank = 0.0013  Figure 1.9. Prognostic significance of PGP in A M L . A , P G P protein positivity as measured by U I C 2 antibody staining is associated with reduced complete remission duration in the poor risk cytogenetic group. B , PGP-related functional activity as measured by dye Rhodamine 123 efflux correlates with lower survival in de novo A M L . Reproduced from D e l Poeta et al, Leukemia Research,  1999 (A) and Wuchter et al, Haematologica, 4  2000 (B). 7  31  STEM C E L L S  C O M MITED PROGENITORS  MATURE C E L L S T-lymphocyte  fM.ymphoeyte /Plasmacelt  Megakaryocyte /Platelets  Basophil /Mast cell  Eosinophil  Neutrophil Monocyte/ Macrophage/ Kupffer cell Langerhans cell Dendritic ceil Osteoclast  F i g u r e 1.10. N o r m a l h u m a n hematopoiesis. Hematopoietic stem cells give rise to all the types of blood cells of the lymphoid lineage (B lymphocytes, T lymphocytes, natural killer cells) and the myeloid lineage (red blood cells neutrophils, basophils, eosinophils, monocytes, macrophages, and platelets) via more commited progenitors. Reproduced from Metcalf D , Blood Lines, 2005. www.hloodlines.stemcells.com.  32  Leukaemic  Normal  F i g u r e 1.11. A M L forms a stem cell hierarchy. Leukemia cells are believed to be mainly derived from transformed CD34++/CD38- hematopoietic stem cells ( H S C ) and share common surface markers with the H S C . The H S C is capable o f self-renewal and production of the normal myeloid and lymphoid lineages through a series o f progenitors (right). Similarly, the leukemic stem cell ( L S C ) is responsible for producing the leukemic progenitors and nonclonogenic blast cells which form the bulk o f the leukemia (left). Reproduced from Huntly & Gilliland et al., Nature Reviews/Cancer, 2005 . 2  33  Recurrence of disease  F i g u r e 1.12. Targeting leukemic stem cells as a curative therapy, a, Current treatment focuses on the eradication o f all leukemia cells (grey) and alleviation o f symptoms, but is ineffective for long-term remission since remaining leukemic stem cells ( L S C s , green) are capable o f repopulating the leukemia, b, Specific targeting o f these L S C s , with or without combination o f conventional chemotherapy, can allow effective cure o f the disease. Reproduced from Huntly & Gilliland, Nature Reviews/Cancer, 2005 . 2  34  II  mRNA expression profiling of the ABC transporter superfamily in unfractionated AML patient samples  2.1 - Introduction  A B C transporters are an important factor in cancer M D R (see Section 1.5), a major problem in A M L treatment. Because neither radiation therapy nor surgery is applicable to leukemia, and few patients are eligible for allogeneic transplants, most A M L patients rely on high-dose chemotherapy to overcome their disease. Due to presentation o f M D R during initial treatment or at relapse; however, only 20-30% patients w i l l ultimately achieve long-term remission. There is an urgent need to investigate the role o f drug resistance factors in A M L in order to circumvent M D R . Previous studies o f A B C transporters on A M L have largely focused on the three known MDR-related transporters - P G P , M R P 1 , and B C R P 1 , with mixed results. With the identification o f more novel, uncharacterized A B C transporters (up to 48), it is possible that some o f these promiscuous transporters also play a role in drug resistance. A systematic study o f the expression o f the whole A B C superfamily is required to determine how many o f these transporters might contribute to M D R in A M L patients. Classic detection o f MDR-related A B C s has mainly relied on use o f monoclonal antibodies to measure protein levels by immunocytochemistry or flow cytometry, typically defining A B C "positivity" by arbitrary thresholds. But because antibodies are not available for many o f the more recently identified A B C transporters, protein measurement was not suitable for a systematic study. A s well, an international, multi-centered workshop organized by W i l l i a m Beck's group in 1 9 9 6  142  identified the variability in measurements by these methods as a major  35  impediment to reaching consensual conclusions on the role o f P G P in A M L . This inconsistency was most apparent when measuring low levels o f A B C transporters in clinical samples, calling for an improvement in the methods used for their detection. Since then, advances in technology have allowed simultaneous, sensitive detection o f many genes with known sequences at the m R N A level. Data generated by this R T - R e a l Time P C R assay gives a profile that is semi-quantitative and represents expression as a continuous variable. In this chapter, I tailored and utilized the R T - R e a l Time P C R assay for semiquantification o f m R N A expression of the A B C superfamily in A M L samples. Because A B C transporters are expressed at relatively low levels, past results by old methods were likely obscured by the limitation o f low sensitivity. Since 20%-30% patients fail to respond to initial therapy, intrinsic resistance mechanisms such as A B C transporter overexpression may already be in place at diagnosis in these patients. I therefore hypothesized that intrinsic m R N A expression o f A B C transporters might be predictive of chemotherapeutic response. To test this, expression profiles were generated on samples o f leukemic cells taken from A M L patients at diagnosis and the results between patients that responded or not to treatment were then compared retrospectively.  36  2.2 - Materials and Methods  2.2.1 - Patient samples, cell lines and culture C C R F - C E M and C E M / V L B human acute lymphoblastic leukemia cell lines established in our laboratory ' 143  144  were grown in A l p h a M E M with 10% F B S ( G I B C O Invitrogen, Burlington,  Ontario, Canada). A n additional 1.0 ug/ml vinblastine (Sigma-Aldrich, Oakville, Ontario, Canada) was supplemented to C E M / V L B to maintain drug resistance. Peripheral blood (PB) cells were obtained from 31 patients with newly diagnosed A M L , after an informed consent and with the approval o f the Clinical Research Ethics Board o f the University o f British Columbia. Diagnosis and classification, o f A M L were based on the criteria of the F A B group. Cytogenetic analysis was performed on the bone marrow at initial diagnosis. Mononuclear cells from A M L blood samples were isolated by F i c o l l Hypaque density gradient centrifugation (Pharmacia, Uppsala Sweden) and cryopreserved in Iscove's modified Dulbecco's medium ( I M D M ) with 50% F B S (both from StemCell Technologies, Vancouver, Canada) and 10% dimethylsulfoxide. More than 90% o f the cells in A M L samples were leukemic blasts. Thawed cells were washed twice in I M D M containing 10% F B S before use in the experiments described below. R N A was extracted immediately from steady-state P B samples. A l l patients selected for this study received remission induction therapy consisting o f daunorubicin (45 mg/m daily for 3 days) and cytarabine (100 mg/m I V q l 2 h x 7 days for patients > 60 years old or 1.5 g/m I V q l 2 h x 6 days for patients < 60 years old). Patients who entered a complete morphological remission with this therapy constitute the complete remission group (CR; responders) in this study. Patients who failed to achieve remission with this initial chemotherapy plus one additional cycle o f treatment (either a second course of the same regimen  37  or an alternate regimen usually containing cyclophosphamide and etoposide  ) constitute the  non-responding group ( N R ; non-responders). Complete remission ( C R ) was defined as less than 5% blasts in a normocellular bone marrow with a neutrophil count >1.0 x 10 /L, and an 9  unsupported hemoglobin o f > 100 g m / L and platelet count >100 x 10 /L. C R patients received 9  consolidation therapy consisting o f either two cycles o f additional chemotherapy the same as induction treatment or allogeneic transplantation (for patients <50-60 years o f age with a suitable sibling donor and intermediate risk or a suitable sibling or unrelated donor and high risk cytogenetics as defined by the M R C ( U K ) criteria ). 6  2.2.2 - RT-Real Time P C R assay 2.2.2.1. Overview. A reverse transcription (RT) -Real Time polymerase chain reaction ( P C R ) assay was developed and utilized to detect the relative expression levels o f A B C transporters in A M L patient samples. This methodology was preferred over expression array chips for its high sensitivity. This proved crucial since many A B C transporters were expressed at low levels. A flow-chart o f the assay was shown in Figure 2.1. Total R N A was first extracted from frozen patient samples and DNase was used to eliminate contaminating genomic D N A . R N A was then reverse-transcribed into c D N A and an aliquot o f the same R T reaction was used for individual Real Time P C R reactions for each gene. The resulting Ct value (see below) from each P C R was used to calculate the relative expression o f each gene o f interest. The amount o f P C R product was detected and measured by the fluorescence intensity caused by the binding o f the fluorescent dye S Y B R Green to D N A . The threshold cycle, Ct, was defined as the cycle when fluorescence first becomes detectable above the threshold value. The Ct is inversely proportional to the amount o f starting R N A transcript and is used to calculate the  38  relative expression level o f the gene. To correct for differences in amplification efficiency ( A E ) between primer pairs, a standard curve was constructed for each gene. A housekeeping gene (GAPDH)  was amplified from the same sample as an endogenous control to account for  variability in concentration and quality o f total R N A , and in the R T reaction efficiency.  2.2.2.2. - RNA isolation, DNase treatment and Reverse Transcription. Total R N A was isolated using Trizol for A M L P B samples (Invitrogen) and quantified by measuring its absorbance at 260/280 nm on a U-2000 spectrophotometer (Hitachi, Tokyo, Japan). Total R N A was treated with DNase I (Invitrogen) following manufacturer's instructions and subsequently reverse-transcribed using random hexamers and the Superscript II R T enzyme (Invitrogen) at a concentration o f 1 u.g total R N A per 20 ul reaction. 2.2.2.3. Primer design and optimization. Primers for Real-Time P C R for all 47 A B C transporter genes were designed using PrimerExpress software, Version 2.0 (Applied Biosystems, Streetsville, Ontario, Canada). A l l primers were designed to yield a unique gene-specific product that does not overlap with consensus A B C walker sequences. Parameters used for design included 100 bp amplicon size, 19-22 bp primer size, 40-60 % G C content and 80-90 °C melting temperature. The validity o f the primers was tested by conventional P C R and products were analyzed by agarose gel electrophoresis to ensure they gave a single product o f the correct size. 2.2.2.4. Real-Time PCR. Real-Time P C R was performed with S Y B R Green Real-Time Core Reagents (Applied Biosystems) according to manufacturer's instructions on the A B I Prism 7900 Sequence Detection System (Applied Biosystems). Each 15 ul P C R reaction contained 1.5 ul diluted c D N A (24 ng starting total R N A ) . Thermal cycling conditions were 50°C for 2 min and 95°C for 5 min, followed by 40 cycles o f 15 sec at 95°C, 30 sec at 58°C and 30 sec at 72°C. A n  39  additional cycle - 15 sec at 95°C, 15 sec at 60°C and 15 sec at 95°C - was performed at the end of the reaction to generate the dissociation curve o f the amplicon to ensure a single, specific product with the corresponding melting temperature was produced (Figure 2.2). A negative R T control without the R T enzyme was included for each sample total R N A to ensure no reminiscent genomic D N A was amplified in the P C R reaction, and a negative P C R water control without c D N A was included per P C R reaction plate to check for reagent contamination. 2.2.2.5. Generation of standard curves. To determine the A E , a standard curve was constructed for each gene on a 2x c D N A dilution series equivalent from 30 ng to 0.47 ng o f starting total R N A (Figure 2.3). The A E was calculated from the formula: 1 0  I / M  - 1 , where M = the slope o f  the standard curve. 2.2.2.6. Data analysis. The Sequence Detector Software S D S 2.0 (Applied Biosystems) was used for data analysis. The threshold cycle value (Ct), defined as the cycle at which a statistically significant increase in S Y B R fluorescence (normalized to a passive reference Rox) is first detected, is automatically calculated and reported by S D S 2.0 for each reaction. Ct is inversely proportional to the log o f c D N A . T o determine the fold expression o f a gene relative to the housekeeping gene, we used the formula: (1 + A E ) "  d C t  where A E is the amplification efficiency  of the specific gene and dCt = (Ct o f the gene) - (Ct o f the housekeeping gene). 2.2.2.7. Statistical analysis. T w o statistical tests, the student's t test and the permutation test, were performed to assess expression differences between the C R and N R patient groups. A two tail-distribution, homoscedastic (assume two sample groups with equal variance) t test was utilized to compare the means of the patient groups. The permutation test is a randomization test which requires no assumptions about statistical distributions (random-sampling, equal variance). Statistical significance was set at p<0.05 for the t test and Z>2 for the permutation test.  40  2.3 - Results  2.3.1 - Profiling of ABC transporters in the drug-sensitive leukemic cell line C E M and its vinblastine-selected, drug-resistant subline CV1.0 I first profiled all 47 human A B C transporters in the lymphoblastic leukemic cell lines C C R F - C E M and C E M / V L B (although 48 human A B C s were predicted, sequences were available for only 47). These were chosen for comparison since the C C R F - C E M parental line is drug-sensitive while the vinblastine-selected C E M / V L B is multidrug resistant (at least 500-fold more resistant to vinblastine and 150-fold more resistant to doxorubicin than C C R F - C E M )  1 4 3  '  1 4 4  .  A s described in the Materials and Methods Section, Ct values generated from the amplification plots were used to calculate the level o f expression o f each test gene relative to that o f  GAPDH  (set at 10 ). Bustin and colleagues reported the transcript copy number per cell o f GAPDH 6  in the order o f 2 x 10 to 5 x 10 in P B 3  3  1 4 6 , 1 4 7  to be  . In addition, I found Ct values above 35 to be  generally unreliable because it approaches the machine detection limit (40 = undetectable). Under the conditions used, this translates to approximately 5 x 1 0 -fold fewer transcripts than GAPDH.  I therefore set the tentative "biologically relevant" expression level at 1/10 o f  GAPDH  (discussed in Section 2.4), which I estimated to correspond to ~2-5 transcripts per cell. Similar to previous reports, MDR1 m R N A was upregulated by at least 2 x 10 -fold i n C E M / V L B 3  cells due to gene amplification as compared to C E M (Figure 2 . 4 ) . The drug-sensitive parental 148  C C R F - C E M cells expressed very low level o f MDR1 (almost 10-fold lower than the ~2 copy per cell reference line). There is no significant difference in MRP1 expression and BCRP1 levels are below the level o f detection. A number o f genes in the ABCA subfamily, notably ABCA5, 6, 9, JO, appear to be also upregulated in the drug-selected cell line. A s these are all clustered on  41  17q24, this chromosomal segment is likely to be amplified independently o f MDR1 located on 7q21 during drug selection. Amplification o f this cluster o f genes in drug resistance warrants further investigation as the function o f these A B C transporters are not well known.  2.3.2 - Lack of consistent differences was observed in ABC transporter expression between responsive and non-responsive patients Table 2.1 lists the characteristics o f 31 A M L patients selected for expression profiling. Samples were chosen to represent patients differing in clinical response to induction chemotherapy with the combination o f cytosine arabinoside and daunorubicin. O f the 31 samples, 18 were from patients who achieved C R after initial therapy and 13 were from the N R group (patients who did not achieve remission). 25 o f the 31 patients had only normal karyotypes seen in the diagnostic bone marrow sample. In an initial set o f experiments, amplification profiles were generated for all 47 A B C transporters from P B samples o f 12 o f the C R patients, and 6 N R patients. Figure 2.5 shows representative expression profiles (CR, Patient #7 and N R , Patient #9). Transcripts for over 40 A B C transporters were detectable in these A M L samples. These included the MDR-related transporters MDR1, MRP1 and BCRP1,  as well as A B C transporters that have been previously  reported to be restricted to other tissues, such as MDR3 (ABCB4) that transports phosphatidylcholine in the liver, SPGP(ABCB11)  that exports bile salts in the liver, CFTR  (ABCC7) the chloride ion channel in the lung, WHITES (ABCG5) and WHITE4 (ABCG8) which transport sterols in the liver and the intestine. However, expression levels o f A B C transporters were generally low (at least 10-fold lower than GAPDH),  with many below the reference line (2  copies per cell). This may represent differential expression among subpopulations o f cells. A s  42  well, there was a significant variation (10 to 100-fold) among these 18 A M L samples in the m R N A levels o f each A B C transporter detected. MDR1 expression in patients generally fell between that o f C C R F - C E M and C E M / V L B . To evaluate the difference in expression between the C R and N R patients for every gene, the data set was tested independently using both a t test and a permutation test. Both statistical tests indicated no significant difference (p>0.05 for the t test and Z<2 for the permutation test) in any A B C transporter between the C R and N R groups, implying that m R N A levels o f A B C transporters in the bulk population prior to treatment is not predictive o f drug response. W e also compared expression levels between A M L samples from the N R group and a C R subgroup o f 6 patients (#24, #30 to #34) who achieved long-term remission for over 3 years - C R ( L ) . But, again, no statistically significant difference (p>0.05 for the t test and Z<2 for the permutation test) was found for any o f the genes examined. Expression o f a selected subset o f 9 A B C transporters including the MDi?-related A B C transporters MDR], MRP1 and BCRP1,  as well as ABCA2, ABCA3, ABCB9, SPGP, MRP4, and  WHITE 1was evaluated in an expanded population o f 31 A M L samples which included 18 C R and 13 N R patients. These genes were selected because they were either MDR-related or closely related to the MDR-transporter genes. Expression profiles o f these genes are shown in Figure 2.6. Again, for each A B C transporter, both groups covered a wide expression range that overlapped with each other. Statistical analysis also confirmed that there was no consistent difference in m R N A expression between the two groups for these A B C transporters. A s apparent in the overlapping expression ranges, some C R patients actually show the highest expression o f MDR-related transporters among all patients, while some N R patients have very low expression of these transporters.  43  2.4 - Discussion  In this section, I used the sensitive RT-Real Time PCR technique to quantify and compare expression of the A B C transporter superfamily in various leukemic cell populations of defined drug responsiveness. Testing on the parental C E M cell line and its drug-selected sub-line validated the ability of this approach to detect differences in A B C expression levels. Consistent with prior reports, dramatic upregulation of the MDR] gene was observed in the drug-resistant CV1.0 cell line. Interestingly, a number of other A B C transporters were also elevated (the ABCA subfamily), most of which are poorly characterized. Whether or not these are inducible by drugs or related to drug resistance are not known and warrant further investigation. In this initial study I constructed the mRNA expression profiles of the A B C transporter superfamily in patient blast cells of 90% purity. Because A B C transporters are a major contributor to M D R in other cancers, I asked whether this would also be the case in A M L . However, I observed no consistent, statistically significant difference between the C R and N R patients' cells in all A B C transporters examined. Hence my results suggest that expression levels of A B C transporters in the total blast population are not a useful predictor of response to initial chemotherapy. There are several interpretations for this apparent discordance with older studies relating A B C expression to a poor prognosis) ' ' ' ' " ' ' . The first may be purely technical since 4  7  90  91  100  102  104  105  differences in laboratory techniques and experimental conditions are known to produce variable results that confounding comparisons. In particular, old studies commonly measured the protein levels of A B C transporters, setting somewhat arbitrary levels of "positivity". Levels of A B C s , however, are likely a continuous variable. Thus my data likely give a more accurate depiction of  44  the expression profiles. O n the other hand, m R N A expression levels may not translate to corresponding protein levels and drug efflux activity o f the A B C transporters. Third, A B C transporters may simply not contribute significantly towards resistance to initial chemotherapy via high intrinsic levels, as suggested by the very low levels o f expression detected. Instead, A B C transporters may play a role in A M L - i n d u c e d resistance via rapid and dramatic upregulation o f expression under the stress o f drug exposure. In support o f this, acute induction of MDR1 expression after exposure to d o x o r u b i c i n  149  and carcinogens  150  has been reported.  Another plausible explanation, in line with the existing paradigm o f cancer heterogeneity, is that the difference in expression lies not in the heterogeneous bulk sample, but in subpopulations o f cells. This is supported by the very low expression levels (below reference line - less than one copy per cell) observed in a significant number o f transporters in many patients. Possibly, a small subpopulation expressed a much higher, biologically-relevant level that is "diluted" in the average profile for the total population. This could also explain why some C R patients express relatively high levels o f known MDR-related transporters and some N R patients express relatively low levels. Perhaps it is expression in the "relevant" cell fraction that accounts for differences in drug response. In order to obtain a more comprehensive picture on the subject, an examination o f sorted subpopulations could be useful.  45  Table 2.1. Patient characteristics. Age at Patient no. Gender diagnosis  FAB  Cytogenetics  1 2 4 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34  M4eo M5B M5A M5A M2 M5A M4 Ml Ml Ml M4 M2 M2 M5 M2 M4 Ml M5B M4 M2 M6 Ml M3 M4 M4 M4 M4 M2 M2 M2 M4eo  Inv(16)(pl3;q22),+22 Normal +8,del(5)(q31;q33) Normal Inv(3)(q21;q26) Normal Normal  M M F M M M F M F M F F F F M F M M M M M F M F M F M M  M F M  36 39 17 54 36 18 63 68 25 69 68 46 25 49 57 28 51 41 46 36 69 39 69 66 . 27 47 22 25 64 62 50  +11 Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal +4, t(8;21)(q22;q22) Normal Normal inv(16)(pl3;q22), +22  Treatment Response CR CR NR CR NR NR NR NR NR NR NR NR NR NR NR NR CR CR CR CR CR(L) CR CR CR CR CR CR(L) CR(L) CR(L) CR(L) CR(L)  C R indicates complete remission achieved after induction chemotherapy; C R ( L ) , continuous remission >3years; N R , no response to induction therapy.  Total R N A  >cDNA DNase Reverse transcription  11111  Individual Realtime P C R  Uniplex Real-time PCR: Triplicate for each gene  Run PCR reactions in ABI Prism Sequence Detection System  Software analysis: Obtain Ct value and dissociation curve for each reaction  Calculate relative expression from Ct: (Gene A Ct) - (Housekeeping gene Ct) = dCt Expression fold = (1 + A E )  _dCt  F i g u r e 2.1. F l o w - c h a r t of R T - R e a l T i m e - P C R . Total R N A isolated from patient peripheral blood was first DNased and reverse-transcribed. The c D N A was subsequently used for uniplex Real Time P C R for each gene. P C R reactions were prepared and aliquoted into a 3 84-well plate and run in the A B I Prism detection system, generating a Ct value per reaction. Expression o f gene A was normalized to the housekeeping gene (GAPDH)  by subtracting the housekeeping  Ct value from Ct o f gene A . The resulting dCt value and the amplification efficiency ( A E , see below) were used to calculate the relative expression, expressed as a fold difference to the housekeeping gene.  47  . Dissorcintiort Plot  // 1  J ••/  •  1 1\  jJ  3 .OOO E -  f i •  /  /'  4(•  V „- j1 s7  \  :  V,/  V  •vs.  '•;>*  :  -J  1-  J  \  \\  SO.O T o n m p e r a t u r e :  i.1 . O O O E+1 -  I.OOO  E-1-  "'1^pDO;E-2-  1 :ooo E - 3 -  '•20* Cycle  Figure 2.2. Typical dissociation curve and amplification plot of a Real Time PCR product. A dissociation curve (top) was generated for the amplified product of the housekeeping gene GapDHin  an A M L sample at the end of a Real Time P C R cycle. The melting temperature  (Tm), 83 °C, corresponds to that of the expected GapDH amplicon from the primer design. The threshold cycle (Ct) of the product is 24 as observed in the amplification plot (bottom), with the red horizontal line indicating the threshold intensity.  48  Standard curves for selected genes  Figure 2.3. Standard curves for selected genes. Ct values of real time PCR reactions were plotted against the amount of starting total R N A . For details, see above.  49  1000000  g 100000  a.  a I Q  Q. a O c o (0 0) 0) 1. a. X <u  10000  1000'  100  £  a) _> -*-* w  10-1  a> LY.  Figure 2.4. Profiling of the ABC transporter superfamily in C E M and CV1.0. m R N A expression levels normalized to GapDH (set at 10 ) o f 47 known A B C transporters in cell lines 6  C E M (black bar) and C V 1 . 0 (white bar). Solid horizontal line indicates the biologicallyrelevant reference line.  50  100000  10000-  1000-  100-  HI 10-1  §  111,  ? g § g 55  Figure 2.5. Profiling of the ABC transporter superfamily in patients CR#7 and NR#9. m R N A expression levels normalized to GapDH (set at 10 ) o f 47 known A B C transporters in 6  cell lines CR#7 (black bar) and NR#9 (white bar). Solid horizontal line indicates the biologically-relevant reference line.  51  1000000-1  II  Q 100000 Q. re o D.  o c o  f  10000  (A (A 0  1000  >  100  _re  CR NR g  ......  -  f  •  :  0) re u  i 1 f  (A  o> _lo ABCA2  ABCA3  MDR1  ABCB9  BSEP  MRP1  MRP4  WHITE 1  I •  BCRP  F i g u r e 2.6. m R N A levels of selected A B C s i n unfractionated A M L patient samples. A B C transporters were profiled in cells from 18 C R patients, and 13 N R patients. For each column (gene), each point represents a single patient from 3 groups (left: remission, right: refractory). Dotted lines indicate expression levels in cell lines (Blue: C E M , Red: C V 1 . 0 )  52  Ill  Expression profiling of drug resistance-related transporters in FACS-  sorted A M L subpopulations  3.1 - Introduction Results from Chapter 2 raised the question o f whether A B C transporters are differentially expressed in different subpopulations o f leukemic cells. Based on the L S C model, high A B C expression in the fraction enriched in leukemia-initiating C D 3 4 C D 3 8 " cells, would be expected +  to be the most critical to explain initial A M L treatment failure. A s discussed in Chapters 1.6 and 1.7, this fraction is chiefly responsible for initiation and maintenance o f the whole leukemic population, and its frequency has been implicated to be an independent prognostic marker in A M L . M a n y studies have been conducted to examine the prognostic value o f C D 3 4 expression in A M L (typically categorizing patients into CD34-positive and CD34-negative for comparison of clinical outcome), albeit with mixed results (reviewed by Kanda et al in 2 0 0 0 ) . Although 151  older studies frequently report an association between C D 3 4 "positivity" and lower remission rates " , more recent studies found no such correlation " . 152  157  158  161  C D 3 4 + C D 3 8 - leukemic cells display self-renewal and differentiating properties that are reminiscent o f the function o f normal H S C . It is, therefore, also likely that these cells would show higher drug tolerance, another characteristic o f normal H S C . Recent studies by de Grouw et a l  1 6 2  and Peeters et a l  1 6 3  have demonstrated preferential expression o f A B C transporters in  both normal and leukemic C D 3 4 + C D 3 8 - cells. Since C D 3 4 + C D 3 8 - cells that express MDRrelated A B C s w i l l have a survival advantage under cytotoxic stress, the presence o f these transporters in this fraction o f cells may predict chemotherapeutic failure. This may in part explain the mixed results on the prognostic value o f C D 3 4 expression  151  - it may be A B C  53  "positivity" o f the C D 3 4 positive leukemic cells, rather than C D 3 4 "positivity" among total A M L blasts, that is prognostically significant. In this section, I profiled the key MDR-related A B C transporters - MDR1, MRP1, and BCRP1 - in different A M L subpopulations. M y working hypothesis was that intrinsic A B C expression in the "relevant" subpopulation, specifically the C D 3 4 + C D 3 8 - fraction, might be more predictive o f initial treatment response than the bulk cells, and might possibly be an even more useful prognostic factor than the number o f these primitive leukemic cells present.  54  3.2 - Materials and Methods  3.2.1 - Flow cytometric sorting of A M L subpopulations Frozen A M L patient cells were thawed at 37 °C and counted in 1:1 trypan blue for viability. A small aliquot was reserved as the unsorted population and the remaining was centrifuged for 9 min at 1000 rpm in 10 m l Iscove's medium (Invitrogen) + 20 % F B S (Invitrogen). Cells were resuspended in H F N (Hank's Balanced Salt Solution + 2% F B S + 0.05% N a l S y + 5% human serum at a concentration o f 7 x 10 cells per u l . For each 2 x 10 4  7  cells, 12 ul o f each o f the following antibodies were added: C D 3 - F I T C , C D 1 9 - F I T C , C D 3 8 - P E , C D 3 4 - C y 5 ( A P C - A ) . After incubation in the dark for 30 m i n on ice, 1 m l H F N was added to each tube and centrifuged for 7 min 1000 rpm. Cells were resuspended in 1 m l H F N with 2ug/ml PI, re-centrifuged and resuspended in 1.5 m l H F N . FACS-sorting was performed on a F A C S A r i a flow cytometer (Becton Dickinson). PI negative cells were first gated as the viable fraction. CD34+CD38-, CD34+CD38+, and C D 3 4 - cells were gated within the viable C D 3 - , C D 1 9 fraction. The following controls were included: I G G 1 - F I T C only, I G G 1 - P E only (non-specific staining), C D 3 - F I T C only, C D 1 9 - F I T C only, C D 3 8 - P E only, C D 3 4 - C y 5 only (for compensation). A sample F A C S sort is shown in Figure 3.1.  3.2.2 - RNA isolation, DNase treatment and RT-Real Time-PCR Sorted A M L subpopulation cells were centrifuged at 1200 rpm for 5 min. Lysis buffer was added to the cell pellet immediately after centrifugation and stored at -80 °C until R N A isolation. Total R N A was isolated using the RNeasy M i c r o spin column kit (Invitrogen) according to manufacturer's instructions. N o quantification was performed due to limitation o f  55  material, and all total R N A isolated was directly treated with DNase I (Invitrogen), reversetranscribed and subsequently used for Real Time P C R using a similar protocol as in Chapter 2.2.2.4 scaled down for smaller amounts o f R N A . A two-tailed, homoscedastic student's t test was utilized to evaluate statistical difference between the N R and C R groups. Statistical significance was set at p < 0.05.  56  3.3 - Results 3.3.1 - Profiling of selected drug resistance-related transporters in FACS-sorted subpopulations of A M L patient samples Three fractions from each o f 7 C R and 10 N R patients were sorted by F A C S : C D 3 4 + C D 3 8 - (most primitive), CD34+CD38+ (differentiating progenitors), C D 3 4 - (depleted o f progenitors). C D 3 - C D 1 9 - cells were first gated to exclude contaminating normal T and B lymphocytes. The three major MDR-related A B C transporters were profiled for each subpopulation: MDR] (Figure 3.2), MRP1 (Figure 3.3), BCRP1 (Figure 3.4). Variation in levels of each gene among subpopulations of the same patient is apparent, indicating heterogeneity within the cancer. Expression levels o f MDR1 and MRP1 were frequently detected above the reference line. Detectable BCRP1 levels were less frequent and fall below the reference. Significant expression for MDR1 and BCRP1 appeared to be restricted to the C D 3 4 - fraction for C R patients, while N R patients showed high levels also in the primitive subpopulations (see below). MRP1 expression, on the other hand, was observed across all fractions.  3.3.2 - Higher expression of MDR1 and BCRP1 in the CD34+CD38- cells from nonresponders  Figure 3.5 shows a scatter plot o f expression levels in the C D 3 4 + C D 3 8 - fraction. For MRP1, there was significant overlap between the C R and N R groups and no significant difference was observed. MDR1 and BCRP1, however, showed interesting patterns in the primitive fraction. A l l 7 C R patients expressed uniformly low levels o f MDR1 and BCRP1 in the C D 3 4 + C D 3 8 - cells. In contrast, 5/10 N R patients (NR6, N R 8 , N R 1 0 , N R 1 4 , N R 1 5 , 50%) show significantly higher expression o f MDR] (all o f which are above the median), and 5/10 N R patients (NR6, N R 8 , N R 9 , N R 1 5 , N R 1 8 , 50%) have high expression ofBCRPl  (most o f which are significantly  57  above the median). Differences for both genes reached statistical significance (p<0.05). In combination, 7/10 N R patients (70%) show high expression o f MDR1 and/or Interestingly, MDR1/BCRP1  BCRP1.  expression profiles o f the N R group do not resemble a continuum.  Rather, there seem to be a distinct segregation within the group, with patients' cells expressing high MDR1/BCRP1  being well separated from the rest o f the N R patients clustered with the C R  group (especially for BCRP1).  O f note, MDR1/BCRP1  levels in C D 3 4 + C D 3 8 - cells are very  low, especially BCRP1 (>10 -fold lower than the housekeeping gene GAPDH). 3  To test i f the  proportion o f C D 3 4 + C D 3 8 - cells among A M L blasts could predict treatment response, the C D 3 4 + C D 3 8 - fraction size was also compared between N R and C R patients (Figure 3.6). Contrary to old studies reporting the prognostic significance o f C D 3 4 expression " , however, 152  157  comparison o f the two groups revealed no statistically significant difference (means o f % C D 3 4 + C D 3 8 - f o r C R and N R = 2.9 and 12 respectively, p>0,05).  58  3.4 - Discussion  In this chapter, I profiled the MDR-related transporters in FACS-sorted subpopulations o f C R and N R patients. Results demonstrate that, consistent with the concept o f cancer heterogeneity, A B C transporters were differentially expressed across subpopulations along the leukemic hierarchy. This argues against the concept that expression levels are homogeneous across all cells in the leukemic clone. Indeed, profiles on subpopulations proved far more revealing on the potential underlying mechanisms o f A M L M D R . N o difference in MRP1 expression between the two patient groups was observed, showing that MRP1 is not a useful predictive factor o f initial treatment outcome and likely not an important contributor to drug resistance in A M L . O n the other hand, expression o f MDR1 and/or BCRP1 was elevated in C D 3 4 + C D 3 8 - cells o f 7/10 N R patients compared to uniformly low levels in the C R group. Hence intrinsic high expression o f either transporter in this primitive fraction enriched for leukemia-initiating cells appears predictive o f poor response. Furthermore, MDR1/BCRP1  m R N A levels in C D 3 4 + C D 3 8 - cells were more predictive o f response than the  size o f the CD34+CD38- fraction itself, which showed no significant difference between the C R and N R groups. Both MDR1 and BCRP1 can efflux daunorubicin, an important chemotherapeutic drug used for most A M L patients. According to the L S C model, C D 3 4 + C D 3 8 - cells are responsible for maintenance o f the whole leukemic population, thus high expression o f these transporters w i l l give them a critical survival advantage under drug exposure, allowing them to regenerate a leukemic population resulting in refractory disease. Although high levels o f A B C transporters were also observed in the C D 3 4 - fraction o f some responders, this mature subpopulation is  59  incapable o f propagating the disease on its own and thus w i l l not contribute to treatment outcome. Comparing differences between the C R and N R groups, both MDR1 and BCRP1 reached statistical significance. More notably however, it is observed that within the N R group, expression levels are either significantly elevated (above the mean) or very low like the C R group, making these "outliers" clearly identifiable. Hence it may be possible to distinguish the potential ABC-dependent (high expression) responders within the N R group for more specifically targeted therapy (see Chapters 4 and 5). Higher relative expression o f A B C transporters in the primitive leukemic fraction is not an entirely surprising finding. In fact, based on the current existing paradigm that L S C originate from transformed H S C , the leukemic C D 3 4 + C D 3 8 - cells would be expected to be found to express the highest A B C levels among the three subpopulations examined, as previously reported for their normal counterparts ' . Indeed, profiling o f a limited number o f CD34+ 119 125  normal P B samples (with H S C mobilization by G - C S F injection) showed MDR1 transcript levels to be even higher than those o f the high-expressing N R patients (data not shown). Hence these patients appear to be retaining existing normal stem cell protection mechanisms, while the lowexpression patients may have actually "lost" this defense system during leukemogenesis. Even more intriguing is the observation that some C D 3 4 - fractions displayed high A B C expression. Since normal C D 3 4 - cells do not express high levels o f these transporters ' ' , their leukemic 119  124  125  counterpart may have abnormally turned on the expression o f these genes. Whether this is a random product o f aberrant epigenetics in A M L or part o f a more systematic mechanism remains to be elucidated. Results o f this study suggest the utility o f evaluating the leukemic C D 3 4 + C D 3 8 subpopulation for both MDR1/BCRP1  expression in predicting a N R outcome. Despite the  60  apparent difference between the C R and N R groups; however, the low MDR1/BCRP1 expression levels detected in N R patients raises the possibility of their biologically relevance. In particular, BCRP1 levels were so low that they fell below the biological reference line. Further studies are required to discern what expression level translates to functionally significant activity.  61  July  July 06-July24 - NR9  I|I*MM| I II IIMj * I I 11 III! I I I till -86  0  10  10  2  3  10  FL1 FITC-A  4  1  -139  06-Julv24 - NR9  0 10  PE-A  Tube; Jufy24 - NR9 Population O Ail Events P1 -~Wi P2  #Events %Pafent %Total  L  P4 PS  mips  61.200 44.374 41.521 38.745 6,350 19,824 7,073  72.5 93.6 93,3 16.4 51.2 18.3  100.0 72.5 67.8 63.3 10.4 32.4 11.6  Figure 3.1. F A C S analysis of A M L patient sample N R # 9 . Left: Staining o f cells with C D 3 F I T C and C D 1 9 - F I T C to gate the C D 3 - C D 1 9 - (P3) fraction. Right: Staining and sorting o f C D 3 4 + C D 3 8 - (P4), CD34+CD38+ (P5), and C D 3 4 - (P6) cells from the P3 gate. Bottom: Size of each gate is shown as a percentage o f the parental gate or percentage o f total population.  62  20000  MICD34+38ES5CD34+38+ ^ CD34-  17500  TT i  Q a  Complete Remission  Non-responsive  15000  ro  0  -2  12500  C o  10000-\  Q. V) X  '»  7500 ^  E  a> > £  5000  —  a>  or  2500-I  £  rd  ^  *  *  f  ^  ^  *«ict / •  Figure 3.2. mRNA expression levels of MDR1 in FACS-sorted A M L subpopulations. CD34+CD38- (red), CD34+CD38+ (green) and CD34- (blue) cell fractions of AML patient samples were sorted by FACS analysis for Real Time-PCR. Expression levels were expressed relative to GAPDH (set at 10 ). Solid horizontal line indicates the biologically-relevant reference line. 6  90000-,  _  JL  MCD34+38tSS3CD34+38+ ^ CD34-  80000-|  Complete Remission  70000  Non-responsive  D Q. «  CD 60000-\ o c o S 50000 4>  Q. X  £  40000  30000  a> a>  20000  10000-1  III cf  0  #  J  J  £  ^_ ^  $  c  #  /  /  #  *  Figure 3.3. mRNA expression levels of MRP1 in FACS-sorted A M L subpopulations. CD34+CD38- (red), CD34+CD38+ (green) and CD34- (blue) cell fractions of AML patient samples were sorted by FACS analysis for Real Time-PCR. Expression levels were expressed relative to GAPDH'(set at 10 ). Solid horizontal line indicates the biologically-relevant reference line. 6  800  ••CD34+38ES3CD34+38+ ^ CD34-  700-  JL  Complete Remissi on  Q 600-  Non-responsive  Q.  <0  o  o  *->  c o  500-  'vt  a> 400M i_ Q.  X  < Z  a:  300-  E a>  > 200-  100  jaJ  J.*  B.  # #  ts  c  ^^^ #  N  ^/ ^  ^  Figure 3.4. mRNA expression levels of BCRP1 in FACS-sorted A M L subpopulations. C D 3 4 + C D 3 8 - (red), CD34+CD38+ (green) and C D 3 4 - (blue) cell fractions of A M L patient samples were sorted by F A C S analysis for Real T i m e - P C R . Expression levels were expressed relative to GAPDH reference line at 1000 (not shown).  (set at 10 ). Expression levels fall significantly below the biologically-relevant 6  17500  250n  NR10  A  B  70000  NR6  225H "to  15000  '60000 200  T—  JL  i  '  £ 12500 (0 O J  50000  175 150  10000  NR9,18  40000  '</>  <n a>  NR14  i_  a.  * <  A  7500  125  NR15  A  30000'  100  or E 5  NR6, 8 *  5000-  75  ro  a:  NR15  2500-  20000H  A  50  A A  NR 8  IOOOOH  25 A . ,  complete remission  non-responsive  Figure 3.5. Comparison of expression oiMDRl,  complete remission  non-responsive  complete remission  non-responsive  BCRP1 and MRP1 between the CR and NR groups in the CD34+CD38-  fraction. Expression levels o f C R patients (square) and N R patients (triangle) were plotted for MDR1 ( A ) , BCRP1 (B) and MRP1 (C). R e d and blue horizontal lines represent median levels of C R and N R , respectively. Patients with high MDR1/BCRP1 levels are identified beside their symbols. Expression levels relative to GAPDH (set at 10 ). Asterisk indicates statistical significance. 6  70-  60H c  m CD  o. 50c c 40o  o oo 30-  m a o +  8 : 20  10  complete remission  non-responsive  Figure 3.6. CD34+CD38- fraction size in CR and NR A M L patients. Percentages o f the FACS-sorted CD34+CD38- subpopulation is plotted for C R (square) and N R (triangle) patients. Horizontal lines represent the median o f each group.  67  IV  Ex vivo drug sensitivity of primitive and mature subpopulations of Acute Myeloid Leukemia and effects of ABC transporter modulation  4.1 - Introduction  Expression profiling as described in the previous chapter allowed rapid examination o f a high number o f genes in patient samples. However, high gene expression does not necessarily correlate to high protein expression or functional activity. To evaluate the biological relevance o f expression data in drug response, a functional assay is necessary. In this chapter I investigated the ex vivo drug sensitivity o f A M L patient cells to determine the functional relevance o f MDR1/BCRP1  gene expression, particularly in the primitive C D 3 4 + C D 3 8 - subpopulation. I first  adapted and validated two functional assays, the Annexin V - P I apoptotic assay and the M T S proliferation assay, on the C E M and C V 1 . 0 leukemic cell lines (see section 2.3.1), These two assays were used to test the sensitivity o f C E M and C V 1 . 0 to exposure o f daunorubicin, a major chemotherapeutic drug used for A M L treatment and a known substrate for both P G P and B C R P 1 . The C E M cell line with low A B C transporter expression should display higher sensitivity to the drug than the C V 1 . 0 cell line with MDR1 amplification. Addition o f A B C modulators PSC-833 (highly specific inhibitor o f P G P ) and verapamil (inhibitor o f A B C transporters including P G P and B C R P 1 ) is expected to result in re-sensitization o f C V 1 . 0 to daunorubicin. The objectives o f this chapter were two-fold. The first was to examine possible differences in drug sensitivity among FACS-sorted subpopulations from A M L patients. A s discussed earlier, the L S C fraction exhibits properties reminiscent o f the normal H S C . Hence I  68  hypothesized that the primitive C D 3 4 + C D 3 8 - fraction would be associated with drug resistance, mirroring protection o f H S C from cytotoxins. Specifically, I hypothesized that the C D 3 4 + C D 3 8 fraction of N R patients would exhibit the highest tolerance to daunorubicin. M y second objective was to investigate the functional significance o f A B C transporters in A M L subpopulations. Based on my observations o f higher A B C gene expression in C D 3 4 + C D 3 8 - cells from A M L N R as compared to C R patients (Chapter 3), I hypothesized that ex vivo drug resistance would be associated with high expression levels o f MDR1 and/or BCRP1,  and that addition o f A B C  inhibitors would have the largest drug re-sensitization effects on the primitive fraction o f N R patients, while the low expresser C R fractions would remain unaffected.  69  4.2 - Materials and Methods  4.2.1 - Exposure of A M L cells to drugs Unsorted or FACS-sorted A M L patient cells were centrifuged and re-suspended in Iscove's medium (Invitrogen) + 10% F B S at a concentration o f 250 cells per u l . To each well, 100 ul cells, 50 ul daunorubicin at varying dilutions (0.001 to 0.5 jig/ml) and 50 u l PSC-833 (1 u M or 3 u M ) or verapamil (5 ug/ml or 20 ug/ml) were added. Cells were incubated at 37 °C for 24 hours before subjected to apoptosis or proliferation assay (see below).  4.2.2 - Annexin V-Propidium Iodide assay The Annexin V - P I assay utilizes two cell-death markers Annexin V (conjugated to F I T C fluorochrome) and propidium iodide (PI) to stain for cells which are undergoing or have undergone apoptosis, respectively. In the early stages o f apoptosis, membrane rearrangement causes the translocation o f phosphatidylserine (PS) from the inner to the outer leaflet o f the plasma membrane. Thus the binding o f PS by Annexin V - F I T C allows detection o f cells in the early stages o f apoptosis. This is coupled with the vital dye PI that stains for later-stage apoptotic cells with a loss o f membrane integrity. Cells that are both Annexin V - F I T C and PI negative are defined as viable. Figure 4.1 shows the detection o f viable cells as double negatives under exposure o f a low and high concentration o f daunorubicin. Apoptosis was measured using the Annexin V - F I T C + PI detection kit I ( B D Biosciences) with a modified manufacturer's protocol. After incubation with daunorubicin for 24 hours, A M L patient cells were centrifuged for 5 min at 1200 rpm. M e d i u m (170 ul) was removed from the top o f each well with caution before addition o f 100 u l o f l x Binding Buffer  70  and 20 u l o f 1/8 diluted Annexin-V and PI mix. I G G 1 - F I T C Antibody ( B D Biosciences) was added to the mock-treatment well as negative control. Additional staining controls were carried out with addition o f Annexin V - F I T C only and PI only. Cells were incubated in the dark at room temperature for 15 min. Reactions were terminated by addition o f 100 u l l x Binding Buffer. Fluorescence was measured using a F A C S C a l i b u r flow cytometer (High Throughput Sampler, B D ) and analyzed by F l o w Jo v.2.0 software.  4.2.3 - M T S assay The M T S assay is a colorimetric, proliferation assay that utilizes the soluble chemical tetrazolium salt ( M T S ) . M T S is reduced to formazan by metabolically active cells (Figure 4.2). Production o f formazan can be detected by development o f a brown colour and is proportional to the number o f viable cells. Proliferation was measured using the CellTiter 96 Aqueous N o n Radioactive C e l l proliferation Assay (Promega). After 24-hour incubation with daunorubicin, 20 ul o f Aqueous Solution 1 was added to cells and incubated for 2 hours at 37 °C. Color development was quantified as absorbance at 490 nm by the M R X Microplate Reader (Dynex Technology).  4.2.4 - A n a l y s i s For the M T S assay, absorbance was plotted against daunorubicin concentration to generate a drug sensitivity curve for each concentration o f A B C inhibition (no inhibition, l u M PSC-833, 3 u M PSC-833, 5 ug/ml verapamil, 20 ug/ml verapamil). For the Annexin V - P I assay, % viability was defined as % o f cells in the Annexin V - F I T C and PI double-negative quadrant (in reference to the non-specific I G G 1 control). Using this value, % k i l l was calculated  71  for each daunorubicin concentration with respect to the viability control (no daunorubicin) using the formula (% viable control-% viable sample)/% viable control x 100%. The drug sensitivity curve was then generated by plotting % k i l l against daunorubicin concentration for each concentration o f A B C inhibition. IC50 for each curve was obtained as the daunorubicin concentration at 50% k i l l . Fold change in IC50 by A B C modulation was calculated using the formula IC50 unmodulated/ IC50 highest inhibition dose. A two-tailed, heteroscedastic student's t test was utilized to evaluate statistical difference between the N R and C R groups. Statistical significance was set at p <0.05.  72  4.3 - Results  4.3.1 - Comparison of the Annexin V-PI apoptotic assay to the MTS proliferation assay on C E M and CV1.0 cell lines. The Annexin V - P I assay and the M T S assay were performed to measure and compare drug tolerance o f C E M and C V 1 . 0 cells. Daunorubicin sensitivity was first determined by the apoptosis assay. Cells were stained with Annexin V - F I T C and PI after exposure to different concentrations o f daunorubicin with or without PSC-833 or verapamil. A s expected, Figure 4.3 shows that C V 1 . 0 cells without A B C modulation is much more resistant to daunorubicin (only 20% k i l l at highest dose o f 0.5 ug/ml) than parental C E M cells ( I C  50  = 0.025 ug/mi). Neither  PSC-833 nor verapamil had a significant effect on C E M while both had a dose-dependent resensitizing effect on C V 1 . 0 cells, which is consistent with MDR1 amplification being a main cause o f the drug-resistance mechanism in this cell line. Results from the M T S assay on a similar drug study were represented in Figure 4.4. Again, C V 1 . 0 cells were more resistant, as indicated by their higher proliferation activity (proportional to higher absorbance) after daunorubicin exposure, and A B C modulation had a dose-dependent effect in decreasing their proliferation/survival. Similar results from the two assays gave results consistent with expectations showing that both assays provided valid measures o f ex vivo drug responses o f leukemic cells. Because o f its single-cell analysis, the apoptotic assay was more sensitive and reliable compared to the colorimetric M T S assay, and was thus used for subsequent studies on A M L patient cells.  73  4.3.2. Adaptation of the apoptotic assay to A M L patient cells.  Due to the limitation o f sorted cells, I next tested assay conditions on unsorted A M L patient cells to adapt the apoptotic assay further to clinical material. To determine the maximum dose o f A B C modulators PSC-833 (dissolved in D M S O ) and verapamil that can be used, bulk cells from A M L patient #18 were subjected to different concentrations o f the A B C modulators in culture and % k i l l was determined using the Annexin V - P I apoptotic assay. Because the concentrated PSC-833 stock solution was dissolved in D M S O , the matching volume equivalent of D M S O was also tested for each PSC-833 concentration. A s seen in Figure 4.5, the maximum dose o f PSC-833 and verapamil without significant toxicity on A M L primary cells was 3 u M and 20 ug/ml, respectively. D M S O had no independent toxic effect in the assay conditions used. Next, a range o f daunorubicin concentrations with or without addition o f the modulator PSC-833 was tested on two unsorted patient samples. A s expected, A B C modulation has no significant effect on patient #18 (Figure 4.6), whose expression o f MDR1 was among the lowest o f all patients (Figure 4.6, small panel). O n the other hand, Figure 4.7 shows that patient #25, who had the highest MDR1 expression (Figure 4.7, small panel), displayed a more resistant profile (% k i l l plateaued at about 60%) than patient #18 (% k i l l reached about 100% at 0.1 ug/ml). Moreover, PSC-833 had a dose-dependent effect, increasing the sensitivity o f patient #25 cells to the P G P substrate daunorubicin. This validated the application o f the apoptotic assay on clinical samples and demonstrated A B C inhibition as a useful measure o f A B C transporter activity.  74  4.3.3 - Subpopulation size and patient material availability as a source of limitation.  For study coherence and to allow direct comparison between expression data and functional results, I performed the ex vivo drug sensitivity assay on the same patient samples used in Chapters 2 and 3. Unfortunately, there were limitations in the selection o f patient samples for functional studies. The main limitation was the size o f subpopulations obtainable. Taking into account the length o f sort time and number o f patient cells per frozen aliquot, a size o f at least 2% o f each subpopulation was typically needed from a single patient. Table 4.1 lists the fraction sizes o f the same patients that have been profiled for A B C expression in their subpopulations, with an asterisk marking those with high expression o f MDR1 and/or BCRP1 in the C D 3 4 + C D 3 8 - cells (Chapter 2). Another limitation was the availability o f patient samples (for example, no more cells were available for patient CR#30). Taken together, two C R patients (CR#1 and #34) and six N R patients (NR#8, #9, #14, #15, #16, #19) matched the criteria of availability and ample fraction size. Four o f the six N R patients ( N R #8, #9, #14, #15) had high MDR1/BCRP1  expression in C D 3 4 + C D 3 8 - cells. These eight patients were used for the  functional assay.  4.3.4. Higher daunorubicin resistance and larger effect of ABC modulation in the CD34+CD38- fraction of non-responders.  CD34+CD38-, CD34+CD38+, and C D 3 4 - fractions were sorted by F A C S from the eight selected A M L patients. These as well as unsorted cells were subjected to 24-hour daunorubicin exposure in the presence or absence o f PSC-833 or verapamil followed by the apoptotic assay.  75  A s illustrated by examples o f drug sensitivity curves from the C R and N R groups in Figures 4.8 and 4.9 (other profiles were included in the Appendix), striking differences were observed between the two. For C R patient #1, unsorted (total), CD34+CD38-, CD34+CD38+ and C D 3 4 fractions showed high sensitivity to daunorubicin, reaching over 85% k i l l at 0.1 (a.g/ml (IC50 = 0.03 to 0.04 ug/ml) with or without A B C modulation, and addition o f either PSC-833 or verapamil only had a slight effect o f increasing its sensitivity. In contrast, N R patient #9 exhibited significant differences among subpopulations. A s shown in Figure 4.9, unmodulated C D 3 4 + C D 3 8 - and CD34+CD38+ (35% - 40% k i l l at 0.1 ug/ml, I C  5 0  > 0.1 ug/ml) o f NR#9 were  significantly more resistant than C D 3 4 - and unsorted (55% and 70% at 0.1 ug/ml respectively). A B C modulation had a dose-dependent re-sensitizing effect on all subpopulations, with a magnitude following the order C D 3 4 + C D 3 8 - > CD34+CD38+ > C D 3 4 - , suggesting that A B C transporter activity was a significant contributor to drug resistance in the non-responder, especially in the primitive subpopulation. Figure 4.10 shows the IC50 o f daunorubicin for each subpopulation o f C R and N R patients. Although only two C R patients were tested, they gave similar results. Both showed very high drug sensitivity (IC50 O . 0 4 ug/ml) in the primitive C D 3 4 + C D 3 8 - fraction, while the mature C D 3 4 - fractions were slightly more resistant to daunorubicin. O n the other hand, the N R patients showed a range o f resistance in terms o f higher IC50 values across all fractions. A s a group, N R patients were much more resistant than the C R patients, particularly in the C D 3 4 + C D 3 8 - fraction in which there was no overlap between the two groups and difference reached statistical significance (p <0.05). Drug sensitivity corresponded with A B C expression, in that the four patients ( N R #8, #9, #14, #15) with high MDR1/BCRP1  expression displayed the highest  resistance in this primitive fraction.  76  Figure 4.11 shows the effect o f A B C modulation on drug sensitivity for each subpopulation. Both C R patients showed little change in the presence o f A B C inhibitors, especially in the C D 3 4 + C D 3 8 - fraction, and only a minor increase in sensitivity in the C D 3 4 - fraction. Contrastingly, A B C inhibition markedly decreased IC50 in the primitive C D 3 4 + C D 3 8 - and CD34+CD38+ fractions o f N R patients. There is a statistically significant difference between the C R and N R groups in CD34+CD38- (p <0.05, student's t test) but not in CD34+CD38+, C D 3 4 and total fractions. A s with drug resistance, observed effects o f A B C modulation is consistent with expression data, with the MDRl/BCKPl-expressing  N R patients being most affected by  A B C inhibition, demonstrating the M D R activity o f these transporters.  77  4.4 - Discussion  In this section, I followed up on the expression results from Chapter 3 with ex vivo functional studies to test the drug sensitivity o f sorted patient subpopulations and the activity o f A B C transporters in these fractions. A s with levels o f A B C expression, differences in functional characteristics were apparent among subpopulations along the leukemic hierarchy. Sensitivity curves o f unsorted cells, as well as the effects o f A B C modulation on these, were generally similar to that o f the predominant fraction o f C D 3 4 - cells rather than the small primitive C D 3 4 + C D 3 8 - fraction, again showing that studies on total population may not be representative o f the tumorigenic cells o f importance for prognosis. Although the population size is small, results o f this study revealed that ABC-dependent daunorubicin resistance in the C D 3 4 + C D 3 8 fraction was common among N R patients who are mostly karyotypically normal (not observed in the C R group) and might thus be predictive o f poor clinical response to initial therapy. This difference between the C R and N R patients was not seen in either the mature C D 3 4 - fraction or the bulk population. For C R patients, the C D 3 4 - fraction displayed slightly higher resistance than the primitive cells which can be overcome by A B C modulation, showing some level o f A B C transporter activity in this fraction. Nevertheless, this did not correlate with clinical outcome ( C R patients), consistent with the hypothesis that the features o f the non-tumorigenic fraction may not be prognostic. Modulation o f A B C activity increased daunorubicin sensitivity in a dose-dependent fashion in the C D 3 4 + C D 3 8 - cells o f most N R patients studied. Such an effect was most dramatic for NR#8, #9 and #15, patients with high MDR1/BCRP1  expression (3/6 tested). A B C  transporter activity was consistent with the expression data in that high  MDR1/BCRP1  78  expression corresponded with high drug resistance, reversible by ABC-specific inhibition. This can have important clinical implications. Drug studies, especially on subpopulations, is timeconsuming and laborious, hence may not be feasible for routine testing in a clinical laboratory, especially for an acute disease such as A M L that requires rapid intervention. O n the other hand, because expression studies are faster, require less clinical material, and may be adapted to high through-put protocols, it is potentially more valuable for identification o f the N R cases where MDR1/BCRP1  is a significant contributor to drug resistance.  Interestingly, although C D 3 4 - cells o f N R patients exhibited a range o f drug resistance like that o f C D 3 4 + C D 3 8 - cells, effects o f A B C modulation on C D 3 4 - is much smaller (except N R 1 6 ) . Therefore, while MDR1/BCRP1  expression may significantly contribute to drug  resistance in primitive leukemic cells, it is not the dominant resistance mechanism in mature cells. Based on the existing paradigm that the L S C originate from H S C , the L S C w i l l likely retain many o f the normal stem cell properties, including its protective mechanism against cytotoxins. A s evident from my studies, L S C from a number o f patients did appear to have high A B C expression like H S C s , which became advantageous under chemotherapy resulting in treatment failure. What additional mechanisms drive drug resistance in the C D 3 4 - fraction remains to be elucidated. The L S C model is in congruence with older principles o f chemotherapy, in particular the classification o f cell cycle-specific and non-cell cycle-specific drugs. Cycle-specific drugs, such as A r a - C , have a dramatic but exclusive effect on proliferating cells in the S-phase o f the cell cycle, while non-cycle-specific drugs such as daunorubicin are active throughout the cell cycle, killing both proliferating cells and quiescent (Go) cells. The latter, however, tends to associate with much higher tissue toxicity compared to cycle-specific drugs. In light o f the L S C model, an  79  updated interpretation o f this is that cycle-specific drugs are effective in killing the more differentiated, actively proliferating cancer cells that usually comprise o f the majority o f the tumor. O n the other hand, non-cycle-specific drugs are needed to eradicate the rare C S C that may be in Go. Hence in A M L , although A r a - C is useful in tumor de-bulking and alleviating symptoms, it is daunorubicin (the substrate for MDR1/BCRP1)  that targets the L S C fraction and  is necessary for long-term remission. Following this premise, it makes sense how patients with high MDR1/BCRP1  in C D 3 4 + C D 3 8 - should fail to achieve complete remission in my study,  since daunorubicin is ineffective against the disease-maintaining compartment.  80  Table 4.1. % of CD34+CD38-, CD34+CD38+ & CD34- fractions in A M L patients. Patient  % CD34+38-  % CD34+38+  % CD34-  CR1  12.3  41.2  27.9  CR20  0.1  63.5  11.8  CR21  0.1  0.8  78  CR25  0.1  10.4  60.6  CR30  3.3  39.9  38  CR31  0.2  75.3  1.2  CR34  4.2  21.6  36.5  NR4  0.6  *NR6  0.05  2.7  89.3  *NR8  69.8  13.8  12.1  *NR9  16.4  51.2  18.3  *NR10  5.2  1.1  80.7  *NR14  6.1  15.2  70.7  *NR15  4.8  63.9  17.3  NR16  11.9  18.6  i i i i i i i i  *NR18  0.1  69.8  11.7  NR19  7.1  48.7  36.4  79.3  * - N R patients with high expression of M A K 7 / B C R P 1 in CD34+CD38-. Highlighted - patient samples suitable A N D available for functional studies.  Ungated " 280706CR1^1 A01 Event Count: 7000 F L 3  H :  P l  Ungated FL3-H: PI 280706CR1 -1A06 Event Count: 7000  F i g u r e 4.1. Apoptosis as detected by the A n n e x i n V - P I assay. A M L patient cells were exposed to 0.001 ug/ml (left) and 0.5 ug/ml (right) daunorubicin for 24 hours before stained by Annexin V and PI and analyzed. The blue gate indicates the quadrant containing viable cells. Cells were significantly more viable when exposed to low drug concentrations (80% viable) than high drug concentrations (1.8% viable).  82  dehydrogenase  Product  Substrate  NAD  m  0^%fX*X'  -  KADH  ETR-reduced  ETR  ^ ^s ^__^y^ Ss  m  J O V * X »  S  CX^^^^C^  F i g u r e 4.2. Schematic representation of conversion of M T S to formazan. Production o f N A D H or N A D P H by metabolically-active cells is the main cause for conversion o f M T S to formazan. Electron transfer occurs from N A D H to an electron transfer reagent ( E T R ) such as P M S , which subsequently causes the production o f brown formazan that can be detected by a colorimetric method. Reproduced from www.ebiotrade.com.  83  — • — C E M DNR only —*—  CEMPSCIuM C E M PSC 3uM  — * — C E M Verapamil 5ug/ml — • — CEM Verapamil 20ug/ml CV1.0 DNR only CV1.0 PSC 1uM CV1.0 P S C 3 u M ------- CV1.0 Verapamil 5ug/ml ------- CV1.0 Verapamil 20 ug/ml 0.001  0.01  0.1  1  DNR ug/ml  Figure 4.3. Effects of PSC-833 and verapamil on daunorubicin sensitivity in C E M and CV1.0 as measured by the Annexin V-PI assay. C E M and C V 1 . 0 cells were exposed to daunorubicin at concentrations ranging from 0.001 to 0.5 ug/ml for 24 hours, with or without PSC-833 (1 u M , 3 u M ) or verapamil (5 ug/ml, 20 ug/ml). Annexin V and PI were subsequently added to the cells and incubated at room temperature for 15 min in the dark. F A C S analysis was performed to determine the viability (defined as both Annexin V and PI negative).  84  — • — C E M DNR only — - A — - C E M P S C 1uM  • — C E M P S C 3uM — * — C E M Verapamil 5 ug/ml — • — C E M Verapamil 20 ug/ml CV1.0 DNR only CV1.0 P S C 1uM CV1.0 P S C 3uM CV1.0 Verapamil 5 ug/ml CV1.0 Verapamil 20ug/ml 0.001  0.01  0.1  1  DNR ug/ml  Figure 4.4. Effects of PSC-833 and verapamil on daunorubicin sensitivity in C E M and CV1.0 as measured by the MTS assay. C E M and C V 1 . 0 cells were exposed to daunorubicin at concentrations ranging from 0.001 to 0.5 ug/ml for 24 hours, with or without PSC-833 (1 u M , 3 u M ) or verapamil (5 ug/ml, 20 ug/ml). M T S was subsequently added to the cells and incubated for 2 hours at 37 °C. Color development was measured as absorbance at 490 nm.  85  PSC-833 in medium  0.01  0.1  1  10  PSC-833 uM 100 90 80 70 60  1  50 40 30 20 10 0 0.001  0.01  0.1  1  10  100  1000  Verapamil pg/ml  Figure 4.5. Toxicity assay of PSC-833 and verapamil on A M L patient cells. Primary A M L patient cells were exposed to PSC-833 at concentrations ranging from 0.03 u M to 8 u M and the D M S O volume equivalent ( A ) or verapamil at concentrations ranging from 0.002 to 200 ug/ml (B) for 24 hours. Viability and % k i l l were determined by the Annexin V - P I assay.  86  DNR ug/ml  Figure 4.6. Effect of P G P inhibition on daunorubicin sensitivity of A M L patient #18. Unsorted cells from patient #18, was subjected to daunorubicin exposure with or without P S C 833 for 24 hours before the Annexin V - P I assay. Small panel: MDR1 expression in total population o f #18 (arrow) i n comparison to other patients.  87  DNR ug/ml  Figure 4.7. Effect of PGP inhibition on daunorubicin sensitivity of A M L patient #25. Unsorted cells from patient #25, was subjected to daunorubicin exposure with or without P S C 833 for 24 hours before the Annexin V - P I assay. Small panel: MDR1 expression in total population o f #25 (arrow) in comparison to other patients.  88  A10  B  /ft/ /•'ft/  /, ft / f,' ft/ t' ft/ ft/- 1  -DNR  only  -m— P S C 0 . 3 u M -•—  PSC  1 uM  <>- - P S C 3uM — V e r a p a m i l 5 ug/ml -A- - Verapamil 20 ug/ml  DNR  ug/ml  DNR  ug/ml  D  -  D N R only  -  P S C 0.3uM  - o —w  P S C 1 uM - P S C 3uM Verapamil 5 ug/ml  - -A- - Verapamil 20 ug/ml  0.001 DNR  ug/ml  DNR  ug/ml  Figure 4.8. Drug sensitivity and effects of ABC modulation on CR patient #1 subpopulations. FACS-sorted CD34+38- ( A ) , CD34+38+ (B) and C D 3 4 - (C) and unsorted cells (D) were exposed to daunorubicin at a range o f concentrations +/- A B C inhibitors verapamil and PSC-833 for 24 hours. C e l l viability was measured by the A V - P I assay and expressed as % k i l l .  OO  Figure 4.9. Drug sensitivity and effects of ABC modulation on NR patient #9 subpopulations. FACS-sorted CD34+38- ( A ) , CD34+38+ (B) and C D 3 4 - (C) and unsorted cells (D) were exposed to daunorubicin at a range o f concentrations +/- A B C inhibitors verapamil and PSC-833 for 24 hours. C e l l viability was measured by the A V - P I assay and expressed as % k i l l .  0.275  n  0.250 0.225 D> 0 200  NR9  A  NR8  *  c o  S 0.175 RS 3 T3 O 0.150  E  o §  0.125-1  3  O JC  0.100  NR15  I J  A A  0.075 NR14  0.050  AA  A A  0.025'J 0.000  CD34+CD38*  CD34+CD38+  CD34-  Unsorted  Figure 4.10. Daunorubicin sensitivity of different A M L subpopulations in CR and NR patients. IC o o f daunorubicin on A M L cells without A B C modulation was plotted for unsorted 5  cells and each subpopulation. Square, C R patients. Triangle, N R patients. Patients with high MDR1/BCRP1  expression are identified beside their symbols. Horizontal line represents  median o f each group. Statistical significance in difference between C R and N R indicated by an asterisk *.  91  10T CO  C  TO  en  JC! O  o  NR14  c:  A  II  c o  If 3  NR9 *  O  E u  CD  < JC  1  4-  o  c .  3-  o  c: CO  o  NR8  A A  NR15  A A  - W -  A A A  2n  2 • :  "tTA-  o  li-  CD34+CD38*  C034+CD38+  CD34-  Unsorted  Figure 4.11. Effects of ABC modulation on drug sensitivity in different A M L subpopulations. Fold-change in daunorubicin IC50 by the highest dose o f A B C inhibitor (IC50 unmodulated / IC50 inhibited) was plotted for unsorted cells and each subpopulation. Square, C R patients. Triangle, N R patients. Patients with high MDR1/BCRP1  expression are identified  beside their symbols. Horizontal line represents median o f each group. Statistical significance in difference between C R and N R indicated by an asterisk *.  92  V  Conclusion and future prospects  5.1 - Overall discussion and conclusion Drug resistance has been a major obstacle in cancer treatment. Some forms o f resistance, which result from alteration o f a specific drug target or loss o f the surface receptor for a given drug, are specific to a small number o f related drugs and can be overcome by combination therapy . The emergence o f M D R , however, which involves resistance to multiple unrelated 164  drugs, poses a far more difficult problem for which combination therapy is not a solution. To effectively reverse M D R , an understanding o f the underlying mechanisms is necessary. This thesis sought to address the problem o f M D R in A M L by investigating the role o f A B C transporters, a classic contributor to M D R in cell lines. In Chapter 2,1 applied the R T - R e a l Time P C R assay to profile and compare the expression levels o f the 47 known human A B C transporters between the A M L responders and non-responders to initial chemotherapy. This is the first systematic study on the prognostic significance o f the full A B C transporter superfamily in A M L . I first asked whether some o f these transporters, especially the MDR-related transporters, might be present in high levels in the bulk leukemic cells from N R patients. However, I found no consistent difference in the expression o f any A B C gene between the bulk samples o f the two patient groups. Based on the L S C model, I then hypothesized that expression differences might be hidden within the most primitive subpopulations. This was addressed in Chapter 3, where subpopulations along the leukemic functional hierarchy were isolated and profiled for expression o f MDR-related A B C transporters. H i g h MDR1 and/or BCRP1 expression in the primitive C D 3 4 + C D 3 8 - fraction was found to be consistently associated with N R outcomes. Neither the more committed CD34+CD38+  93  progenitors nor the mature C D 3 4 - cells provided such an association. This is the first indication o f a possible prognostic value of A B C transporter expression in the C D 3 4 + C D 3 8 - cells, the small subpopulation that is thought to be responsible for maintenance o f the leukemia in patients and for relapses. In Chapter 4,1 further investigated the ex vivo drug sensitivity o f patient subpopulations and functional relevance of A B C transporters. Using the Annexin V - P I apoptotic assay, I confirmed that ABC-dependent resistance, corresponding to high MDR1IBCRP1  expression and  reversibility by ABC-specific inhibitors, is common among non-responsive patients, particularly those with a normal karyotype. This suggested that the expressed P G P / B C R P 1 are actively extruding drugs from the L S C and thus may make a significant contributor to intrinsic drug resistance in vivo. M y study demonstrated that the properties o f primitive subpopulations may facilitate better understanding o f how a cancer operates than examination o f the properties o f the bulk cells. In particular, the drug response o f the primitive C D 3 4 + C D 3 8 - A M L subpopulation seems a more accurate predictor o f treatment outcome than the bulk leukemic population. This is in line with the L S C model and calls for a continued research focus on this small fraction o f cells. O f note, the very low m R N A levels observed (especially for BCRP1 which was consistently below the biological reference line) even in the relative "high" expressers hints towards a further subset within the CD34+CD38- fraction that expresses much higher levels. Hence although C D 3 4 + C D 3 8 - marks a subpopulation enriched in tumorigenicity that is important for prognostic and therapeutic purposes, it is likely a heterogeneous group o f cells in itself. Figure 5.1 shows the different proposed models o f drug resistance in cancer . The 3  conventional model o f cancer drug resistance (Figure 5.1 A ) conceives that a number o f cells  94  acquired mutations that confer drug resistance. These cells outgrow the others to form a new resistant tumor population following chemotherapy. In the C S C model (Figure 5.IB), the C S C (CD34+CD38- in A M L ) is inherently drug-resistant. A t least some o f these survive chemotherapy to regenerate a tumor similar to the original disease. A variation o f the C S C model, the acquired-resistance C S C model (Figure 5.1C), posits that additional mutations in the surviving C S C generate a drug-resistant tumor. In the last intrinsic resistance model (Figure 5.ID), both the C S C and its descendants are intrinsically drug-resistant. Therapies are ineffective, resulting in uncontrolled tumor growth. The results o f my study are consistent with the C S C model and the acquired-resistance model in that C D 3 4 + C D 3 8 - cells with intrinsic high MDR1/BCRP1  expression and activity can survive chemotherapy. These can drive tumor  regeneration that is rapid enough to result in persistent disease, hence failure to achieve complete remission (no detectable leukemic cells). In this respect, non-responsiveness may be viewed as a very fast relapse within the chemotherapeutic regimen. During or following initial therapy, the primitive fraction may acquire further mutations that confer M D R to its descendants as well (acquired resistance model). A B C transporter expression may also be induced in the general population upon drug exposure, producing a drug resistant disease often seen in A M L relapse. The presence o f drug resistance mechanisms such as A B C transporters in the L S C is a more accurate predictor o f response than the de novo size o f this fraction. Thus not all L S C are resistant enough to withstand the initial high-dose chemotherapy given to A M L patients. Those patients without high MDR1IBCRP1  expression in C D 3 4 + C D 3 8 - cells responded well to initial  therapy, while those that apparently retained this normal H S C defense system did not respond. Clearly, A M L is a heterogeneous disease and the L S C that originates it possesses varying properties in each patient.  95  It is interesting that non-responders with high A B C expression (MDR1IBCRP1)  appear  well separated from the other non-responders. Practically, this may represent a convenient means to distinguish the ABC-dependent (high expression) from the ABC-independent (low expression) N R patients. Conceptually, this segregation also demonstrates the value in identifying outliers within a pre-defined group, an approach that has recently been successfully applied by Tomlins and colleagues to discover the T M P R S S 2 - E T S fusion in prostate cancer . Cancer as a dynamic 165  disease can display heterogeneity within both itself (as exemplified by expression differences observed among subpopulations) and among its assigned "type" and "group", such as responders and non-responders in this study. In fact, it is important to bear in mind that these "subcategories" merely serve as working definitions and hence is subjected to redefinition or further division as more information became known. In this case, patients within the N R group appear to fall into the ABC-dependent and ABC-independent subcategories that may require different strategies in treatment (Section 5.2). Overall, my work fulfilled all three objectives: to profile expression levels o f the A B C transporter superfamily in total A M L patient samples, to compare expression o f MDR-related A B C transporters in subpopulations along the leukemic hierarchy, and to investigate the drug resistance o f subpopulations and functional activity o f A B C transporters. Based on my studies, the following conclusions are reached. First, expression o f A B C transporters in de novo unfractionated patient samples is not predictive o f response. Second, high intrinsic levels o f MDR1 and/or BCRP1 in the primitive C D 3 4 + C D 3 8 - fraction are associated with poor response to initial chemotherapy. Third, high ex vivo functional activity correlates with high expression levels o f MDR11BCRP1  in CD34+CD38- cells. Fourth, ABC-dependent drug resistance in  C D 3 4 + C D 3 8 - is common among non-responders, especially those with normal cytogenetics, and  96  is reversible via A B C - i n h i b i t i o n . Taken as a whole, my studies suggest a prognostic significance of A B C transporters in the primitive C D 3 4 + C D 3 8 - leukemic subpopulation, and support a modified approach in investigating the value o f A B C modulating agents in A M L , as discussed in the next section.  97  5.2 - Future Prospects  Previous clinical trials o f P G P inhibition, such as second-generation inhibitor PSC-833, yielded largely negative results in A M L , with high toxicity and lack o f significant improvement in outcome ' . M y findings suggest a functional role o f A B C transporters in the primitive, 83  84  disease-maintaining fraction o f some N R A M L patients. Based on this initial study, disappointing results from clinical A B C inhibition can be viewed in a new light. MDR1IVG? expression and activity was low in all C R samples and half of the N R samples, to whom P G P inhibition is unlikely to have a significant effect. Because past studies did not distinguish these patients from the small subset o f ABC-expressing non-responders, clinical benefits o f A B C modulation was probably diluted and underestimated. M y work raises the importance and feasibility o f pre-screening patients for targeted therapy in A M L . While A B C activity may not be the major mechanism of M D R in all nonresponders, it may be possible to identify those where A B C transporter expression is a major factor for drug resistance and apply appropriate therapeutic intervention. Figure 5.2 outlines a proposed scheme for predicting treatment response and overcoming M D R . Before a patient is subjected to chemotherapy, levels o f MDR1/BCRP1 in the C D 3 4 + C D 3 8 - fraction can be quickly determined. Patients with high expression o f either A B C transporter are predicted to be A B C dependent non-responders (to chemotherapy alone). These are likely to significantly benefit from the combination o f conventional chemotherapy and A B C inhibitors, outweighing the high toxicity effects. The low-expressers are predicted to be either responsive to chemotherapy alone or ABC-independent non-responders and may be spared o f the toxicity o f A B C inhibition. A s  98  more MDR-related factors (factor X ) are characterized, it may be possible to incorporate additional testing and targeted-therapeutic options to the treatment scheme. A s discussed in the introduction, cytogenetic aberrations are the major prognostic factor in A M L . But for patients who are karyotypically normal (half o f all patients), only a few molecular biomarkers have been investigated (discussed in Section 1.3). Most o f the patients (25/31) profiled for A B C transporter expression in my study belonged to this category. In search for other prognostic markers especially for this cytogenetically normal group, I have identified high expression and activity of MDR1/BCRP1  in CD34+CD38- cells as one significant  predictive factor o f poor response. Although this was a common M D R mechanism among the patients tested, however, about half o f the non-responders appear relatively independent o f A B C activity. Evidently, A B C transporters are not the only determinants for treatment outcome. Technical advancements in recent years have opened up new possibilities to study the molecular genetics o f a disease. One such technique, comparative genomic hybridization ( C G H ) , can be used to detect micro amplifications and deletions in D N A . Recent construction by L a m and colleagues o f a high-resolution array with over 30,000 B A C clones that cover the entire human genome allows analysis o f copy-number changes in the global genome, revealing down to gene-size alterations (Figure 5.3) . While half o f the patients are cytogenetically normal, there 166  may be micro-scale alterations in their cancer genome not detected by karyotyping. It is possible that these micro-amplifications and deletions recur at the loci o f novel genes that account for the lack o f response to chemotherapy. A s a further extension o f my thesis, I would seek to explore genomic changes occurring in A M L samples using the novel array-CGH technique. I hypothesize that micro genomic differences can be detected between drug sensitive and resistant patients. The objective o f this  99  study would be two-fold: first, to determine whether high expression o f MDR1 and BCRP1 in N R cells is due to micro-amplifications; second, to see i f additional amplified markers can be identified in a genomic signature predictive o f clinical response. Eight C R patients and nine N R patients (17 in total) from previous studies were selected for arrayCGH, 12 o f which have normal cytogenetics. Microalterations not reported by cytogenetics have been found in each sample. A s shown in Figure 5.4, micro-gains and losses were observed in the different chromosomes o f patient NR#3 (A). For example, there was a fragment loss on the p arm o f chromosome 9 (B) not detected by cytogenetics, and even smaller but clear micro-amplifications and micro-deletions on chromosome 5q. Some o f these gains and losses were recurrent in the samples tested, which may indicate genes critical to leukemogenesis. Furthermore, some differences are seen between the C R and N R samples. Further work on these may provide insight into finding possible prognostic patterns in the de novo leukemic genome to predict treatment response, and identifying novel genes that contribute to the pathogenesis and/or drug response o f A M L .  100  a  MDR eels  I)  Tumou- stem cell  Figure 5.1. Models of tumor drug resistance. A , in the conventional model o f cancer drug resistance, rare cells with genetic alterations that confer M D R form a resistance clone (yellow). These cells survive chemotherapy and proliferate, giving rise to relapsed disease with offspring o f the resistant clone. B , in the C S C model, the C S C (red) that has protective mechanisms survive chemotherapy while the committed cells (blue) are killed. The C S C repopulates a functional tumor hierarchy. C , in the acquired resistance C S C model, C S C surviving chemotherapy accumulates mutations (yellow) conferring a resistant phenotype in also the committed descendants (purple). D , in the intrinsic resistance model, both the stem cells and the committed cells are inherently resistant. Therapy has no effect, resulting in tumor expansion. Reproduced from Dean et al, Nature Reviews/Cancer, 2005 . 3  101  ^ %  D  e  n  0  V  0  A  M  L  RT-PCR H i g h M D R 1 / B C R P 1 in CD34+CD38-  Low MDR1 & BCRP1 in CD34+CD38-  CR or ABCindependent  ABC-dependent NR Chemotherapy + A B C inhibition  Test other Jf factors X  4m " *  Eradication of A M L  X-dependent NR Chemotherapy  Chemotherapy + X inhibition  Eradication of A M L F i g u r e 5.2. P r e d i c t i n g response and overcoming M D R . A t diagnosis, A M L patients can be tested for MDR] and BCRP1 expression in the CD34+CD38- cells. Those with high expression are likely ABC-dependent non-responders ( N R ) to chemotherapy alone and w i l l benefit from the combination o f chemotherapy and A B C inhibitors. Patients with low A B C expression are either responders who w i l l achieve complete remission with chemotherapy alone ( C R ) or A B C independent non-responders who has alternate drug resistance mechanisms. A s more knowledge becomes available, patients may be tested for other MDR-related factors ( X ) and appropriate therapeutic intervention may be applied for eradication o f the disease.  102  "T"iJ ITS out*  F^t &fN&f& 1*^0(61  Figure 5.3. Principles of array comparative genomic hybridization. A , normal and tumor D N A samples are isolated and used to create fluorescently labeled probes, commonly with cyanine 3 (Cy3, green) and cyanine 5 (Cy5, red) dyes. The probes are pooled and competitively cohybridized to a glass slide spotted with a known array o f mapped genomic clones. The arrays are analyzed with a microarray scanner, producing an image that is used to assess the log2 ratios o f the C y 5 to C y 3 intensities for each clone. B , A log2 ratio profile is assembled to determine relative copy number changes between the cancer and the normal samples. Each dot on the graph represents a clone. Values to the left o f " 0 " indicate a loss o f a genomic region, while values to the right indicate a gain or amplification. Values at " 0 " indicate no change. Reproduced from Davies et al, Chromosome Research, 2005.  103  F i g u r e 5.4. C G H k a r y o g r a m p f patient NR#3. D N A from patient NR#3 was isolated and subjected to a r r a y C G H . The raw data was analyzed using the S e e G H program to generate a karygram for every chromosome. Each dot is a B A C clone representing a small section o f D N A . Values to the left o f the centre line ("0") indicate a loss, and values to the right indicate a gain. A , microalterations on the 23 chromosome pairs. B , Fragment loss on chromosome 9p. C , micro-amplifications and micro-deletions on chromosome 5q.  104  Bibliography  1. Dean M , Rzhetsky A , Allikmets R. The human A T P - b i n d i n g cassette ( A B C ) transporter superfamily. Genome Res. 2001;11:1156-1166. 2. Huntly B J , Gilliland D G . Leukaemia stem cells and the evolution o f cancer-stem-cell research. Nat Rev Cancer. 2005;5:311-321. 3. Dean M , Fojo T, Bates S. Tumour stem cells and drug resistance. Nat Rev Cancer. 2005;5:275-284. 4. D e l Poeta G , Venditti A , Stasi R, et al. P-glycoprotein and terminal transferase expression identify prognostic subsets within cytogenetic risk classes in acute myeloid leukemia. Leuk Res. 1999;23:451-465. 5. Chang G , Roth C B . Structure o f M s b A from E . coli: a homolog o f the multidrug resistance A T P binding cassette ( A B C ) transporters. Science. 2001 ;293:1793-1800. 6. Grimwade D , Walker H , Oliver F , et al. The importance o f diagnostic cytogenetics on outcome in A M L : analysis of 1,612 patients entered into the M R C A M L 10 trial. The Medical Research Council Adult and Children's Leukaemia Working Parties. Blood. 1998;92:2322-2333. 7. Wuchter C , Leonid K , Ruppert V , et al. Clinical significance o f P-glycoprotein expression and function for response to induction chemotherapy, relapse rate and overall survival in acute leukemia. Haematologica. 2000;85:711-721. 8. Canadian Cancer Statistics. 2006. 9. Ozols R F , Herbst R S , Colson Y L , et al. Clinical cancer advances 2006: major research advances in cancer treatment, prevention, and screening—a report from the American Society o f Clinical Oncology. J C l i n Oncol. 2007;25:146-162. 10. Collins I, Workman P. N e w approaches to molecular cancer therapeutics. Nat Chem B i o l . 2006;2:689-700. 11. Druker B J , Sawyers C L , Kantarjian H , et al. Activity o f a specific inhibitor o f the B C R A B L tyrosine kinase in the blast crisis o f chronic myeloid leukemia and acute lymphoblastic leukemia with the Philadelphia chromosome. N Engl J M e d . 2001;344:1038-1042. 12. Druker B J , Talpaz M , Resta D J , et al. Efficacy and safety o f a specific inhibitor of the B C R - A B L tyrosine kinase in chronic myeloid leukemia. N Engl J M e d . 2001;344:1031-1037. 13. Kantarjian H M , Talpaz M , Giles F , O'Brien S, Cortes J. N e w insights into the pathophysiology o f chronic myeloid leukemia and imatinib resistance. A n n Intern M e d . 2006;145:913-923. 14. Stirewalt D L , Radich JP. The role o f F L T 3 in haematopoietic malignancies. Nat Rev Cancer. 2003;3:650-665. 15. Whitman SP, Archer K J , Feng L , et al. Absence o f the wild-type allele predicts poor prognosis in adult de novo acute myeloid leukemia with normal cytogenetics and the internal tandem duplication o f F L T 3 : a cancer and leukemia group B study. Cancer Res. 2001 ;61:72337239. 16. Frohling S, Schlenk R F , Breitruck J, et al. Prognostic significance o f activating F L T 3 mutations in younger adults (16 to 60 years) with acute myeloid leukemia and normal cytogenetics: a study o f the A M L Study Group U l m . Blood. 2002;100:4372-4380. 17. K a i n z B , Heintel D , Marculescu R , et al. Variable prognostic value o f F L T 3 internal tandem duplications in patients with de novo A M L and a normal karyotype, t(15;17), t(8;21) or inv(16). Hematol J. 2002;3:283-289.  105  18. Smith B D , Levis M , Beran M , et al. Single-agent C E P - 7 0 1 , a novel F L T 3 inhibitor, shows biologic and clinical activity in patients with relapsed or refractory acute myeloid leukemia. Blood. 2004;103:3669-3676. 19. Stone R M , DeAngelo D J , K l i m e k V , et al. Patients with acute myeloid leukemia and an activating mutation in F L T 3 respond to a small-molecule F L T 3 tyrosine kinase inhibitor, P K C 4 1 2 . B l o o d . 2005;105:54-60. 20. Bonnet D , D i c k JE. Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nat M e d . 1997;3:730-737. 21. Mitus A , Rosenthal D . The Adult Leukemias. Clinical Oncology. 1999. 22. Schoch C , Haferlach T. Cytogenetics in acute myeloid leukemia. Curr Oncol Rep. 2002;4:390-397. 23. Mrozek K , Heinonen K , Bloomfield C D . Prognostic value o f cytogenetic findings in adults with acute myeloid leukemia. Int J Hematol. 2000;72:261-271. 24. Mrozek K , Heerema N A , Bloomfield C D . Cytogenetics in acute leukemia. B l o o d Rev. 2004;18:115-136. 25. D e V i t a J V , Hellman S, Rosenberg S. Cancer Principles and practice o f oncology. 200 l;2nd edition. 26. M i y o s h i H , Shimizu K , K o z u T, Maseki N , Kaneko Y , Ohki M . t(8;21) breakpoints on chromosome 21 in acute myeloid leukemia are clustered within a limited region o f a single gene, A M L 1 . Proc Natl A c a d Sci U S A . 1991;88:10431-10434. 27. Warrell R P , Jr., de The H , Wang Z Y , Degos L . Acute promyelocytic leukemia. N Engl J M e d . 1993;329:177-189. 28. B y r d J C , Mrozek K , Dodge R K , et al. Pretreatment cytogenetic abnormalities are predictive o f induction success, cumulative incidence o f relapse, and overall survival in adult patients with de novo acute myeloid leukemia: results from Cancer and Leukemia Group B ( C A L G B 8461). Blood. 2002;100:4325-4336. 29. Grimwade D , Walker H , Harrison G , et al. The predictive value o f hierarchical cytogenetic classification in older adults with acute myeloid leukemia ( A M L ) : analysis o f 1065 patients entered into the United K i n g d o m Medical Research Council A M L 1 1 trial. Blood. 2001;98:1312-1320. 30. Slovak M L , Kopecky K J , Cassileth P A , et al. Karyotypic analysis predicts outcome o f preremission and postremission therapy in adult acute myeloid leukemia: a Southwest Oncology Group/Eastern Cooperative Oncology Group Study. Blood. 2000;96:4075-4083. 31. Fenaux P, Chevret S, Guerci A , et al. Long-term follow-up confirms the benefit o f alltrans retinoic acid in acute promyelocytic leukemia. European A P L group. Leukemia. 2000;14:1371-1377. 32. Tallman M S , Andersen J W , Schiffer C A , et al. All-trans retinoic acid in acute promyelocytic leukemia: long-term outcome and prognostic factor analysis from the North American Intergroup protocol. Blood. 2002; 100:4298-4302. 33. Shen Z X , Chen G Q , N i J H , et al. Use o f arsenic trioxide ( A s 2 0 3 ) in the treatment o f acute promyelocytic leukemia ( A P L ) : II. Clinical efficacy and pharmacokinetics in relapsed patients. B l o o d . 1997;89:3354-3360. 34. Soignet S L , Frankel SR, Douer D , et al. United States multicenter study o f arsenic trioxide in relapsed acute promyelocytic leukemia. J C l i n Oncol. 2001;19:3852-3860.  106  35. Raffoux E , Rousselot P, Poupon J, et al. Combined treatment with arsenic trioxide and all-trans-retinoic acid in patients with relapsed acute promyelocytic leukemia. J C l i n Oncol. 2003;21:2326-2334. 36. Shen Z X , Shi Z Z , Fang J, et al. All-trans retinoic acid/As203 combination yields a high quality remission and survival in newly diagnosed acute promyelocytic leukemia. Proc Natl A c a d Sci U S A . 2004;101:5328-5335. 37. Garson O M , Hagemeijer A , Sakurai M , et al. Cytogenetic studies o f 103 patients with acute myelogenous leukemia in relapse. Cancer Genet Cytogenet. 1989;40:187-202. 38. Frohling S, Skelin S, Liebisch C , et al. Comparison o f cytogenetic and molecular cytogenetic detection o f chromosome abnormalities in 240 consecutive adult patients with acute myeloid leukemia. J C l i n Oncol. 2002;20:2480-2485. 39. Trost D , Hildebrandt B , M u l l e r N , Germing U , Royer-Pokora B . Hidden chromosomal aberrations are rare in primary myelodysplastic syndromes with evolution to acute myeloid leukaemia and normal cytogenetics. Leuk Res. 2004;28:171-177. 40. Casas S, Aventin A , Fuentes F , et al. Genetic diagnosis by comparative genomic hybridization in adult de novo acute myelocytic leukemia. Cancer Genet Cytogenet. 2004;153:16-25. 41.. Bullinger L , Dohner K , Bair E , et al. Use o f gene-expression profiling to identify prognostic subclasses in adult acute myeloid leukemia. N Engl J M e d . 2004;350:1605-1616. 42. V a l k P J , Verhaak R G , Beijen M A , et al. Prognostically useful gene-expression profiles in acute myeloid leukemia. N Engl J M e d . 2004;350:1617-1628. 43. Marcucci G , Mrozek K , Bloomfield C D . Molecular heterogeneity and prognostic biomarkers in adults with acute myeloid leukemia and normal cytogenetics. Curr Opin Hematol. 2005;12:68-75. 44. Hess J L . M L L : a histone methyltransferase disrupted in leukemia. Trends M o l M e d . 2004;10:500-507. 45. Frohling S, Schlenk R F , Stolze I, et al. C E B P A mutations in younger adults with acute myeloid leukemia and normal cytogenetics: prognostic relevance and analysis o f cooperating mutations. J C l i n Oncol. 2004;22:624-633. 46. Thiede C , K o c h S, Creutzig E , et al. Prevalence and prognostic impact o f N P M 1 mutations in 1485 adult patients with acute myeloid leukemia ( A M L ) . B l o o d . 2006;107:40114020. 47. Baldus C D , Tanner S M , Ruppert A S , et al. B A A L C expression predicts clinical outcome of de novo acute myeloid leukemia patients with normal cytogenetics: a Cancer and Leukemia Group B Study. Blood. 2003;102:1613-1618. 48. Childs S, L i n g V . The M D R superfamily o f genes and its biological implications. Important A d v Oncol. 1994:21-36. 49. Higgins C F . A B C transporters: from microorganisms to man. A n n u Rev Cell B i o l . 1992;8:67-113. 50. Walker J E , Saraste M , Runswick M J , Gay N J . Distantly related sequences in the alphaand beta-subunits o f A T P synthase, myosin, kinases and other ATP-requiring enzymes and a common nucleotide binding fold. Embo J. 1982;1:945-951. 51. Hyde S C , Emsley P, Hartshorn M J , et al. Structural model o f ATP-binding proteins associated with cystic fibrosis, multidrug resistance and bacterial transport. Nature. 1990;346:362-365.  107  52. Dawson R J , Locher K P . Structure o f a bacterial multidrug A B C transporter. Nature. 2006;443:180-185. 53. Juliano R L , L i n g V . A surface glycoprotein modulating drug permeability in Chinese hamster ovary cell mutants. B i o c h i m Biophys Acta. 1976;455:152-162. 54. Rao V V , Dahlheimer J L , Bardgett M E , et al. Choroid plexus epithelial expression o f M D R 1 P glycoprotein and multidrug resistance-associated protein contribute to the bloodcerebrospinal-fluid drug-permeability barrier. Proc Natl A c a d Sci U S A . 1999;96:3900-3905. 55. Borst P, Evers R, K o o l M , Wijnholds J. The multidrug resistance protein family. B i o c h i m Biophys Acta. 1999;1461:347-357. 56. Gottesman M M , Ambudkar S V . Overview: A B C transporters and human disease. J Bioenerg Biomembr. 2001;33:453-458. 57. Schinkel A H , Smit JJ, van Tellingen O, et al. Disruption o f the mouse m d r l a Pglycoprotein gene leads to a deficiency in the blood-brain barrier and to increased sensitivity to drugs. Cell. 1994;77:491-502. 58. Borst P, Evers R, K o o l M , Wijnholds J. A family o f drug transporters: the multidrug resistance-associated proteins. J Natl Cancer Inst. 2000;92:1295-1302. 59. Wijnholds J, deLange E C , Scheffer G L , et al. Multidrug resistance protein 1 protects the choroid plexus epithelium and contributes to the blood-cerebrospinal fluid barrier. J C l i n Invest. 2000;105:279-285. 60. Doyle L A , Y a n g W , Abruzzo L V , et al. A multidrug resistance transporter from human M C F - 7 breast cancer cells. Proc Natl A c a d Sci U S A . 1998;95:15665-15670. 61. Miyake K , M i c k l e y L , Litman T, et al. Molecular cloning o f c D N A s which are highly overexpressed in mitoxantrone-resistant cells: demonstration o f homology to A B C transport genes. Cancer Res. 1999;59:8-13. 62. Jonker J W , Buitelaar M , Wagenaar E , et al. The breast cancer resistance protein protects against a major chlorophyll-derived dietary phototoxin and protoporphyria. Proc Natl A c a d Sci U S A . 2002;99:15649-15654. 63. Krishnamurthy P, Ross D D , Nakanishi T, et al. The stem cell marker B c r p / A B C G 2 enhances hypoxic cell survival through interactions with heme. J B i o l Chem. 2004;279:2421824225. 64. Fojo A T , Ueda K , Slamon D J , Poplack D G , Gottesman M M , Pastan I. Expression o f a multidrug-resistance gene in human tumors and tissues. Proc Natl A c a d Sci U S A . 1987;84:265269. 65. Goldstein L J , Galski H , Fojo A , et al. Expression o f a multidrug resistance gene in human cancers. J Natl Cancer Inst. 1989;81:116-124. 66. Roninson I B . From amplification to function: the case o f the M D R 1 gene. Mutat Res. 1992;276:151-161. 67. Gottesman M M , Fojo T, Bates S E . Multidrug resistance in cancer: role o f A T P dependent transporters. Nat Rev Cancer. 2002;2:48-58. 68. Ross D D , Karp J E , Chen T T , Doyle L A . Expression o f breast cancer resistance protein in blast cells from patients with acute leukemia. Blood. 2000;96:365-368. 69. Diestra J E , Scheffer G L , Catala I, et al. Frequent expression o f the multi-drug resistanceassociated protein B C R P / M X R / A B C P / A B C G 2 in human tumours detected by the B X P - 2 1 monoclonal antibody in paraffin-embedded material. J Pathol. 2002;198:213-219. 70. Kawabata S, Oka M , Soda H , et al. Expression and functional analyses o f breast cancer resistance protein in lung cancer. C l i n Cancer Res. 2003;9:3052-3057.  108  71. M a c k JT, Beljanski V , Tew K D , Townsend D M . The ATP-binding cassette transporter A B C A 2 as a mediator o f intracellular trafficking. Biomed Pharmacother. 2006;60:587-592. 72. Childs S, Y e h R L , H u i D , L i n g V . Taxol resistance mediated by transfection o f the liverspecific sister gene o f P-glycoprotein. Cancer Res. 1998;58:4160-4167. 73. Allikmets R , Schriml L M , Hutchinson A , Romano-Spica V , Dean M . A human placentaspecific A T P - b i n d i n g cassette gene ( A B C P ) on chromosome 4q22 that is involved in multidrug resistance. Cancer Res. 1998;58:5337-5339. 74. Komatani H , Kotani H , Hara Y , et al. Identification o f breast cancer resistant protein/mitoxantrone resistance/placenta-specific, ATP-binding cassette transporter as a transporter o f N B - 5 0 6 and J-107088, topoisomerase I inhibitors with an indolocarbazole structure. Cancer Res. 2001;61:2827-2832. 75. Maliepaard M , Scheffer G L , Faneyte IF, et al. Subcellular localization and distribution o f the breast cancer resistance protein transporter in normal human tissues. Cancer Res. 2001;61:3458-3464. 76. Cole SP, Bhardwaj G , Gerlach J H , et al. Overexpression o f a transporter gene in a multidrug-resistant human lung cancer cell line. Science. 1992;258:1650-1654. 77. L o e D W , Deeley R G , Cole SP. Characterization o f vincristine transport by the M(r) 190,000 multidrug resistance protein ( M R P ) : evidence for cotransport with reduced glutathione. Cancer Res. 1998;58:5130-5136. 78. Jedlitschky G , Leier I, Buchholz U , Barnouin K , K u r z G , Keppler D . Transport o f glutathione, glucuronate, and sulfate conjugates by the M R P gene-encoded conjugate export pump. Cancer Res. 1996;56:988-994. 79. M u l l e r M , Meijer C , Zaman G J , et al. Overexpression o f the gene encoding the multidrug resistance-associated protein results in increased ATP-dependent glutathione S-conjugate transport. Proc Natl A c a d Sci U S A . 1994;91:13033-13037. 80. Marbeuf-Gueye C , Broxterman H J , Dubru F , Priebe W , Gamier-Suillerot A . Kinetics o f anthracycline efflux from multidrug resistance protein-expressing cancer cells compared with Pglycoprotein-expressing cancer cells. M o l Pharmacol. 1998;53:141-147. 81. Renes J, de Vries E G , Nienhuis E F , Jansen P L , M u l l e r M . A T P - and glutathionedependent transport o f chemotherapeutic drugs by the multidrug resistance protein M R P 1 . B r J Pharmacol. 1999;126:681-688. 82. Tidefelt U , Liliemark J, Gruber A , et al. P-Glycoprotein inhibitor valspodar ( P S C 833) increases the intracellular concentrations o f daunorubicin in vivo in patients with P-glycoproteinpositive acute myeloid leukemia. J C l i n Oncol. 2000;18:1837-1844. 83. Greenberg P L , Lee SJ, Advani R , et al. Mitoxantrone, etoposide, and cytarabine with or without valspodar in patients with relapsed or refractory acute myeloid leukemia and high-risk myelodysplastic syndrome: a phase III trial (E2995). J C l i n Oncol. 2004;22:1078-1086. 84. Baer M R , George S L , Dodge R K , et al. Phase 3 study o f the multidrug resistance modulator PSC-833 in previously untreated patients 60 years o f age and older with acute myeloid leukemia: Cancer and Leukemia Group B Study 9720. B l o o d . 2002;100:1224-1232. 85. Chan H S , DeBoer G , Thiessen JJ, et al. Combining cyclosporin with chemotherapy controls intraocular retinoblastoma without requiring radiation. C l i n Cancer Res. 1996;2:14991508. 86. Ambudkar S V , Dey S, Hrycyna C A , Ramachandra M , Pastan I, Gottesman M M . Biochemical, cellular, and pharmacological aspects o f the multidrug transporter. A n n u Rev Pharmacol Toxicol. 1999;39:361-398.  109  87. Marie JP, Zittoun R, Sikic B I . Multidrug resistance (mdrl) gene expression in adult acute leukemias: correlations with treatment outcome and in vitro drug sensitivity. Blood. 1991;78:586-592. 88. Pirker R, Wallner J, Geissler K , et al. M D R 1 gene expression and treatment outcome in acute myeloid leukemia. J Natl Cancer Inst. 1991;83:708-712. 89. Leith C P , Kopecky K J , Chen I M , et al. Frequency and clinical significance o f the expression o f the multidrug resistance proteins M D R l / P - g l y c o p r o t e i n , M R P 1 , and L R P in acute myeloid leukemia: a Southwest Oncology Group Study. Blood. 1999;94:1086-1099. 90. Zochbauer S, Gsur A , Brunner R, Kyrle P A , Lechner K , Pirker R. P-glycoprotein expression as unfavorable prognostic factor in acute myeloid leukemia. Leukemia. 1994;8:974977. 91. Borg A G , Burgess R, Green L M , Scheper R J , Y i n J A . Overexpression o f lung-resistance protein and increased P-glycoprotein function in acute myeloid leukaemia cells predict a poor response to chemotherapy and reduced patient survival. B r J Haematol. 1998;103:1083-1091. 92. Lyttelton M P , Hart S, Ganeshaguru K , Hoffbrand A V , Mehta A B . Quantitation o f multidrug resistant M D R 1 transcript in acute myeloid leukaemia by non-isotopic quantitative cDNA-polymerase chain reaction. B r J Haematol. 1994;86:540-546. 93. W o o d P, Burgess R, MacGregor A , Y i n J A . P-glycoprotein expression on acute myeloid leukaemia blast cells at diagnosis predicts response to chemotherapy and survival. B r J Haematol. 1994;87:509-514. 94. Zhou D C , Zittoun R, Marie JP. Expression o f multidrug resistance-associated protein ( M R P ) and multidrug resistance ( M D R 1 ) genes in acute myeloid leukemia. Leukemia. 1995;9:1661-1666. 95. M i c h i e l i M , Damiani D , Ermacora A , et al. P-glycoprotein, lung resistance-related protein and multidrug resistance associated protein in de novo acute non-lymphocytic leukaemias: biological and clinical implications. B r J Haematol. 1999;104:328-335. 96. Broxterman H J , Sonneveld P, Pieters R, et al. D o P-glycoprotein and major vault protein ( M V P / L R P ) expression correlate with in vitro daunorubicin resistance in acute myeloid leukemia? Leukemia. 1999;13:258-265. 97 r Smyth M J , Krasovskis E , Sutton V R , Johnstone R W . The drug efflux protein, Pglycoprotein, additionally protects drug-resistant tumor cells from multiple forms o f caspasedependent apoptosis. Proc Natl A c a d Sci U S A . 1998;95:7024-7029. 98. Johnstone R W , Cretney E , Smyth M J . P-glycoprotein protects leukemia cells against caspase-dependent, but not caspase-independent, cell death. B l o o d . 1999;93:1075-1085. 99. Robinson L J , Roberts W K , L i n g T T , Lamming D , Sternberg SS, Roepe P D . Human M D R 1 protein overexpression delays the apoptotic cascade in Chinese hamster ovary fibroblasts. Biochemistry. 1997;36:11169-11178. 100. van der K o l k D M , de Vries E G , van Putten W J , et al. P-glycoprotein and multidrug resistance protein activities in relation to treatment outcome in acute myeloid leukemia. C l i n Cancer Res. 2000;6:3205-3214. 101. Legrand O, Simonin G , Beauchamp-Nicoud A , Zittoun R, Marie JP. Simultaneous activity o f M R P 1 and Pgp is correlated with in vitro resistance to daunorubicin and with in vivo resistance in adult acute myeloid leukemia. Blood. 1999;94:1046-1056. 102. Legrand O, Simonin G , Perrot J Y , Zittoun R, Marie JP. Pgp and M R P activities using c a l c e i n - A M are prognostic factors in adult acute myeloid leukemia patients. Blood. 1998;91:4480-4488.  110  103. Filipits M , Suchomel R W , Zochbauer S, Brunner R, Lechner K , Pirker R. Multidrug resistance-associated protein in acute myeloid leukemia: N o impact on treatment outcome. C l i n Cancer Res. 1997;3:1419-1425. 104. Damiani D , Tiribelli M , Calistri E , et al. The prognostic value o f P-glycoprotein ( A B C B ) and breast cancer resistance protein ( A B C G 2 ) in adults with de novo acute myeloid leukemia with normal karyotype. Haematologica. 2006;91:825-828. 105. Steinbach D , Sell W , Voigt A , Hermann J, Zintl F , Sauerbrey A . B C R P gene expression is associated with a poor response to remission induction therapy in childhood acute myeloid leukemia. Leukemia. 2002;16:1443-1447. 106. van den Heuvel-Eibrink M M , Wiemer E A , Prins A , et al. Increased expression o f the breast cancer resistance protein ( B C R P ) in relapsed or refractory acute myeloid leukemia ( A M L ) . Leukemia. 2002;16:833-839. 107. Abbott B L , Colapietro A M , Barnes Y , M a r i n i F , Andreeff M , Sorrentino B P . L o w levels of A B C G 2 expression in adult A M L blast samples. Blood. 2002;100:4594-4601. 108. van der K o l k D M , Vellenga E , Scheffer G L , et al. Expression and activity o f breast cancer resistance protein ( B C R P ) in de novo and relapsed acute myeloid leukemia. B l o o d . 2002;99:3763-3770. 109. van der P o l M A , Broxterman H J , Pater J M , et al. Function o f the A B C transporters, Pglycoprotein, multidrug resistance protein and breast cancer resistance protein, in minimal residual disease in acute myeloid leukemia. Haematologica. 2003;88:134-147. 110. Reya T, Morrison SJ, Clarke M F , Weissman I L . Stem cells, cancer, and cancer stem cells. Nature. 2001;414:105-111. 111. Dalerba P, Cho R W , Clarke M F . Cancer Stem Cells: Models and Concepts. A n n u Rev M e d . 2006. 112. Huntly B J , Gilliland D G . Cancer biology: summing up cancer stem cells. Nature. 2005;435:1169-1170. 113. B u i c k R N , Minden M D , M c C u l l o c h E A . Self-renewal in culture o f proliferative blast progenitor cells in acute myeloblastic leukemia. Blood. 1979;54:95-104. 114. Turhan A G , Lemoine F M , Debert C , et al. Highly purified primitive hematopoietic stem cells are P M L - R A R A negative and generate nonclonal progenitors in acute promyelocytic leukemia. Blood. 1995;85:2154-2161. 115. Castor A , Nilsson L , Astrand-Grundstrom I, et al. Distinct patterns o f hematopoietic stem cell involvement in acute lymphoblastic leukemia. Nat M e d . 2005; 11:630-637. 116. Brown D , Kogan S, Lagasse E , et al. A P M L R A R a l p h a transgene initiates murine acute promyelocytic leukemia. Proc Natl A c a d Sci U S A . 1997;94:2551-2556. 117. Taipale J, Beachy P A . The Hedgehog and Wnt signalling pathways in cancer. Nature. 2001;411:349-354. 118. Bunting K D . A B C transporters as phenotypic markers and functional regulators o f stem cells. Stem Cells. 2002;20:11-20. 119. Chaudhary P M , Roninson I B . Expression and activity o f P-glycoprotein, a multidrug efflux pump, in human hematopoietic stem cells. C e l l . 1991;66:85-94. 120. Berenson R J , Bensinger W I , H i l l R S , et al. Engraftment after infusion o f CD34+ marrow cells in patients with breast cancer or neuroblastoma. Blood. 1991;77:1717-1722. 121. Berenson R J , Andrews R G , Bensinger W I , et al. Antigen CD34+ marrow cells engraft lethally irradiated baboons. J C l i n Invest. 1988;81:951-955.  Ill  122. Krause D S , Ito T, Fackler M J , et al. Characterization o f murine C D 3 4 , a marker for hematopoietic progenitor and stem cells. B l o o d . 1994;84:691-701. 123. Sato T, Laver J H , Ogawa M . Reversible expression o f C D 3 4 by murine hematopoietic stem cells. B l o o d . 1999;94:2548-2554. 124. Zinovyeva M V , Zijlmans J M , Fibbe W E , Visser J W , Belyavsky A V . Analysis o f gene expression in subpopulations o f murine hematopoietic stem and progenitor cells. E x p Hematol. 2000;28:318-334. 125. Scharenberg C W , Harkey M A , Torok-Storb B . The A B C G 2 transporter is an efficient Hoechst 33342 efflux pump and is preferentially expressed by immature human hematopoietic progenitors. B l o o d . 2002;99:507-512. 126. Zhou S, Schuetz J D , Bunting K D , et al. The A B C transporter B c r p l / A B C G 2 is expressed in a wide variety o f stem cells and is a molecular determinant o f the side-population phenotype. Nat M e d . 2001;7:1028-1034. 127. Uchida N , Fujisaki T, Eaves A C , Eaves C J . Transplantable hematopoietic stem cells in human fetal liver have a CD34(+) side population (SP)phenotype. J C l i n Invest. 2001; 108:10711077. 128. Goodell M A , Brose K , Paradis G , Conner A S , M u l l i g a n R C . Isolation and functional properties o f murine hematopoietic stem cells that are replicating in vivo. J E x p M e d . 1996;183:1797-1806. 129. A l v i A J , Clayton H , Joshi C , et al. Functional and molecular characterisation o f mammary side population cells. Breast Cancer Res. 2003;5:R1-8. 130. Asakura A , Rudnicki M A . Side population cells from diverse adult tissues are capable o f in vitro hematopoietic differentiation. E x p Hematol. 2002;30:1339-1345. 131. Lechner A , Leech C A , Abraham E J , Nolan A L , Habener JF. Nestin-positive progenitor cells derived from adult human pancreatic islets o f Langerhans contain side population (SP) cells defined by expression o f the A B C G 2 ( B C R P 1 ) ATP-binding cassette transporter. Biochem Biophys Res Commun. 2002;293:670-674. 132. Uchida N , Dykstra B , Lyons K , Leung F , Kristiansen M , Eaves C . A B C transporter activities o f murine hematopoietic stem cells vary according to their developmental and -activation status. B l o o d . 2004;103:4487-4495. 133. te Boekhorst P A , de Leeuw K , Schoester M , et al. Predominance o f functional multidrug resistance ( M D R - 1 ) phenotype in CD34+ acute myeloid leukemia cells. Blood. 1993;82:31573162. 134. Drenou B , Fardel O, A m i o t L , Fauchet R. Detection o f P glycoprotein activity on normal and leukemic C D 3 4 + c e l l s . Leuk Res. 1993;17:1031-1035. 135. Haraguchi N , Utsunomiya T, Inoue H , et al. Characterization o f a side population o f cancer cells from human gastrointestinal system. Stem Cells. 2006;24:506-513. 136. Hirschmann-Jax C , Foster A E , W u l f G G , et al. A distinct "side population" o f cells with high drug efflux capacity in human tumor cells. Proc Natl A c a d Sci U S A . 2004; 101:1422814233. 137. K o n d o T, Setoguchi T, Taga T. Persistence o f a small subpopulation o f cancer stem-like cells in the C 6 glioma cell line. Proc Natl A c a d Sci U S A . 2004;101:781-786. 138. Smalley M J , Clarke R B . The mammary gland "side population": a putative stem/progenitor cell marker? J Mammary Gland B i o l Neoplasia. 2005;10:37-47.  112  139. Szotek P P , Pieretti-Vanmarcke R , Masiakos P T , et al. Ovarian cancer side population defines cells with stem cell-like characteristics and Mullerian Inhibiting Substance responsiveness. Proc Natl A c a d Sci U S A . 2006; 103:11154-11159. 140. W u l f G G , Wang R Y , Kuehnle I, et al. A leukemic stem cell with intrinsic drug efflux capacity in acute myeloid leukemia. Blood. 2001 ;98:1166-1173. 141. Feuring-Buske M , Hogge D E . Hoechst 33342 efflux identifies a subpopulation o f cytogenetically normal CD34(+)CD38(-) progenitor cells from patients with acute myeloid leukemia. B l o o d . 2001;97:3882-3889. 142. Beck W T , Grogan T M , W i l l m a n C L , et al. Methods to detect P-glycoprotein-associated multidrug resistance in patients' tumors: consensus recommendations. Cancer Res. 1996;56:3010-3020. 143. Bradley G , N a i k M , L i n g V . P-glycoprotein expression in multidrug-resistant human ovarian carcinoma cell lines. Cancer Res. 1989;49:2790-2796. 144. Efferth T, Davey M , Olbrich A , Rucker G , Gebhart E , Davey R. Activity o f drugs from traditional Chinese medicine toward sensitive and M D R 1 - or MRPl-overexpressing multidrugresistant human C C R F - C E M leukemia cells. B l o o d Cells M o l Dis. 2002;28:160-168. 145. Johny A , Song K W , Nantel S H , et al. Early stem cell transplantation for refractory acute leukemia after salvage therapy with high-dose etoposide and cyclophosphamide. B i o l B l o o d Marrow Transplant. 2006;12:480-489. 146. Bustin S A , Gyselman V G , Williams N S , Dorudi S. Detection o f cytokeratins 19/20 and guanylyl cyclase C in peripheral blood o f colorectal cancer patients. B r J Cancer. 1999;79:18131820. 147. Bustin S A , Siddiqi S, Ahmed S, Hands R , Dorudi S. Quantification o f cytokeratin 20, carcinoembryonic antigen and guanylyl cyclase C m R N A levels in lymph nodes may not predict treatment failure in colorectal cancer patients. Int J Cancer. 2004;108:412-417. 148. H u X F , Slater A , W a l l D M , et al. Rapid up-regulation o f m d r l expression by anthracyclines in a classical multidrug-resistant cell line. B r J Cancer. 1995;71:931-936. 149. Abolhoda A , Wilson A E , Ross H , Danenberg P V , Burt M , Scotto K W . Rapid activation of M D R 1 gene expression in human metastatic sarcoma after in vivo exposure to doxorubicin. C l i n Cancer Res. 1999;5:3352-3356. 150. Thorgeirsson SS, Huber B E , Sorrell S, Fojo A , Pastan I, Gottesman M M . Expression o f the multidrug-resistant gene in hepatocarcinogenesis and regenerating rat liver. Science. 1987;236:1120-1122. 151. Kanda Y , Hamaki T, Yamamoto R , et al. The clinical significance o f C D 3 4 expression in response to therapy o f patients with acute myeloid leukemia: an overview o f 2483 patients from 22 studies. Cancer. 2000;88:2529-2533. 152. Campos L , Guyotat D , Archimbaud E , et al. Surface marker expression in adult acute myeloid leukaemia: correlations with initial characteristics, morphology and response to therapy. B r J Haematol. 1989;72:161-166. 153. Borowitz M J , Gockerman JP, Moore J O , et al. Clinicopathologic and cytogenic features of C D 3 4 ( M y 10)-positive acute nonlymphocytic leukemia. A m J C l i n Pathol. 1989;91:265-270. 154. Geller R B , Zahurak M , Hurwitz C A , et al. Prognostic importance o f immunophenotyping in adults with acute myelocytic leukaemia: the significance o f the stem-cell glycoprotein C D 3 4 ( M y l O ) . B r J Haematol. 1990;76:340-347.  113  155. Campos L , Guyotat D , Archimbaud E , et al. Clinical significance o f multidrug resistance P-glycoprotein expression on acute nonlymphoblastic leukemia cells at diagnosis. Blood. 1992;79:473-476. 156. Solary E , Casasnovas R O , Campos L , et al. Surface markers in adult acute myeloblastic leukemia: correlation o f C D 19+, C D 3 4 + and C D 14+/DR—phenotypes with shorter survival. Groupe d'Etude Immunologique des Leucemies ( G E I L ) . Leukemia. 1992;6:393-399. 157. Lee E J , Y a n g J, Leavitt R D , et al. The significance o f C D 3 4 and T d T determinations in patients with untreated de novo acute myeloid leukemia. Leukemia. 1992;6:1203-1209. 158. L a m y T, Goasguen J E , Mordelet E , et al. P-glycoprotein (P-170) and C D 3 4 expression in adult acute myeloid leukemia ( A M L ) . Leukemia. 1994;8:1879-1883. 159. Bradstock K , Matthews J, Benson E , Page F , Bishop J. Prognostic value o f immunophenotyping in acute myeloid leukemia. Australian Leukaemia Study Group. Blood. 1994;84:1220-1225. 160. Fruchart C , Lenormand B , Bastard C , et al. Correlation between C D 3 4 expression and chromosomal abnormalities but not clinical outcome in acute myeloid leukemia. A m J Hematol. 1996;53:175-180. 161. Arslan O, A k a n H , Beksac M , et al. Lack o f prognostic value o f C D 3 4 in adult A M L . Leuk Lymphoma. 1996;23:185-186. 162. de Grouw E P , Raaijmakers M H , Boezeman J B , et al. Preferential expression o f a high number o f A T P binding cassette transporters in both normal and leukemic C D 3 4 + C D 3 8 - cells. Leukemia. 2006;20:750-754. 163. Peeters S D , van der K o l k D M , de Haan G , et al. Selective expression o f cholesterol metabolism genes in normal C D 3 4 + C D 3 8 - cells with a heterogeneous expression pattern in A M L cells. E x p Hematol. 2006;34:622-630. 164. Gottesman M M . Mechanisms o f cancer drug resistance. A n n u Rev M e d . 2002;53:615627. 165. Tomlins S A , Rhodes D R , Perner S, et al. Recurrent fusion o f T M P R S S 2 and E T S transcription factor genes in prostate cancer. Science. 2005;310:644-648. 166. Ishkanian A S , M a l l o f f C A , Watson S K , et al. A tiling resolution D N A microarray with complete coverage o f the human genome. Nat Genet. 2004;36:299-303.  114  Appendix: Drug sensitivity curves of CR and NR patients. CR34 Effects of PSC-833 and verapamil on DNR sensitivity in AML patient CR#34 CD34+CD38+  Effects of PSC-833 and verapamil on DNR sensitivity in AML patient CR#34 CD34+CD38-  D N R only P S C 0.3 uM P S C 1 uM P S C 3 uM Verapamil.5 ug/ml • Verapamil 20 ug/ml  DNR ug/ml  DNR ug/ml  Effects of PSC-833 and verapamil on DNR sensitivity in AML patient CR#34 Unsorted  Effects of PSC-833 and verapamil on DNR sensitivity in AML patient CR#34 CD34100  110  90  100  80  90  70  80  ^  ^  ^  I.H»  70  60  //  /  50 40  //  •—~~  30  60 — • — D N R only -m—  P S C 0.3 uM P S C 1 uM  20 10  ^  * P S C 3 uM  •—  0 0 .001  'i^r>  X  ^ ^ ^ ^  Verapamil 5 ug/ml  — • — V e r a p a m i l 20 ug/ml •0.01  0.1 DNR ug/ml  50  — • — D N R only  40  -»-PSC0.3uM  30  P S C 1 uM  20 10  -  >•  '^^j  —*—Verapamil 5 ug/ml  0 0..001  P S C 3 uM  • 0.01  0.1 DNR ug/ml  Verapamil 20 ug/ml  NR8  Effect of PSC-833 and Verapamil on DNR sensitivity of AMLNR8CD34+38+  Effect of PSC-833 and Verapamil on DNR sensitivity of AML NR8 CD34+38110  110  - • — D N R only  100 90  • /r/-"~"~~  P S C 1uM  80  - • — D N R only  100  - • > - P S C 0.3uM  ir  P S C 3uM  /  ///  70  — * - Verapamil 5ug/ml  60  - • — Verapamil 20ug/ml  ///  ///  50  //* / y /•/  40 30 20  - * - P S C 0.3uM  90  /  /?yy-~—\ /ir / j /  P S C 1uM  80  / /  •  P S C 3uM  70  - » — Verapamil 5ug/ml  60  - • — Verapamil 20ug/ml  / /  1 / • /* i /  30  yy^y *sy^$y  10  0  0 0. 001  0.1  0.01  DNR (jg/ml  110  Effect of PSC-833 and Verapamil on DNR sensitivity of AML NR8 Unsorted  - * - D N R only .  120  0.3uM  110  psc  100  P S C 1uM  90 80 70  - * — D N R only -*-PSC0.3uM  100  f  —:  *>• P S C 3uM  0.1  DNR pg/ml  Effect of PSC-833 and Verapamil on DNR sensitivity of AMLNR8CD34120  /  // / /*  20  0.01  /  / / V  40  10  0 001  /  50  90  P S C 1uM "  *  80  —•— Verapamil 20ug/ml  70 " - • — V e r a p a m i l 20ug/ml 60  60 50  &s'  Verapamil 5ug/ml  II/ //'  50  40  ///'  /  40 •  30 20  ^^^^  10 ' •  0 IE0.001  30  yy y?y  20  0 0.01  0.1  DNR pg/ml  .^yy^cy  10  ''  0.001  .  P S C 3uM  —«— Verapamil 5ug/ml  - = = ^ ^ ^ ^ 0.01  0.1  DNR wg/m)  ' **  /  / '  NR14  Effect of PSC-833 and Verapamil on DNR sensitivity of AMLNR14CD34+38+  Effect of PSC-833 and Verapamil on DNR sensitivity of -AMLNR14CD34+38-  110 -j 100  :  0.001'  110 i  0.01  0.1  1  0.001  • 0.01  DNR ug/ml  0.1  1  DNR pg/ml  Effect of PSC-833 and Verapamil on DNR sensitivity of AML NR14 CD34-  Effect of PSC-833 and Verapamil on DNR sensitivity of AML NR14 Unsorted 110  : r  //  //  «// S?  ''/if  50  ///  —•— DNR only ~-«-PSC 0.3uM  ////  • 0.001  /  /  /  /'//  P S C 1uM P S C 3uM -*—Verapamil 5ug/ml -•—Verapamil 20ug/ml  0.01  0.1 DNR ug/ml  1  0.001  0.01  0.1 DNR pg/ml  1  NR15  110 100  Effect of PSC-833 and Verapamil on DNR sensitivity of AMLNR15CD34+38110  « - DNR only »-PSC0.3uM  90  - * - PSC 0.3uM  90  P S C 1uM  y-  70  •—Verapamil 5ug/ml  P S C 3uM  PSCIuM  80 -  /  P S C 3uM  /  70 - -*— Verapamil 5ug/ml  *— Verapamil 20ug/m:  60 . —•— Verapamil 20ug/ml  50  50  40  40  30  30  20  20  10  10  0 0 001  -*— DNR only  100  .  80  60  Effect of PSC-833 and Verapamil on DNR sensitivity of AMLNR15CD34+38+  /  /'  /  /  '  / "  /  / y y y  //  /  /  // )  y  0 0.01  0.1  0. 001  0.01  DNR pg/ml  0.1 DNR pg/ml  Effect of PSC-833 and Verapamil on DNR sensitivity of AML NR15 CD34-  Effect of PSC-833 and Verapamil on DNR sensitivity of AML NR15 Unsorted  120  - DNR only  110  - P S C 0.3uM  100  PSC 1uM  90 80  H  70  PSC 3uM -Verapamil 5ug/ml -Verapamil 20ug/ml  60 50 40 30 20 10 0 0.001 DNR pg/ml  0.01  0.1 DNR pg/ml  NR16  Effect of PSC-833 and Verapamil on DNR sensitivity of  Effect of PSC-833 and Verapamil on DNR sensitivity of AML NR16CD34+38-  AMLNR16CD34+38+ 100 90 80 70 60  /S'y/  50  .  40  : - • — DNR only  .  -»-PSC0.3uM  30 "  "~  ^ y  y y  PSCIuM *  20  P S C 3uM  - * — Verapamil 5ug/ml  10  —•— Verapamil 20ug/ml  0 0.001  0.01  DNR ug/ml  Effect of PSC-833 and Verapamil on DNR sensitivity of  Effect of PSC-833 and Verapamil on DNR sensitivity of AML NR16 CD34-  0.1 DNR ug/ml  AMLNR16Unsorted 100 90 80 70 60 50  /si  40  - * - DNR only P S C 0.3uM P S C 1uM  30  x-  20 10 0  ,  0. 001 DNR ug/ml  MD  / y --^ y 0.01  - • — Verapamil 20ug/ml  0.1 DNR pg/ml  P S C 3uM  —«h- Verapamil 5ug/ml  i  NR19  Effect of PSC-833 and Verapamil on DNR sensitivity of AML patient NR#19 CD34+38-  DNR pg/ml  Effect of PSC-833 and Verapamil on DNR sensitivity of AML patient NR#19CD34-  DNR pg/ml  Effect of PSC-833 and Verapamil on DNR sensitivity of AML patient NR#19 CD34+38+  DNR pg/ml  Effect of PSC-833 and Verapamil on DNR sensitivity of AML patient NR#19 Unsorted  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Country Views Downloads
China 31 54
United States 13 0
Argentina 4 4
Mexico 2 0
France 2 0
United Kingdom 1 0
Canada 1 0
Japan 1 0
City Views Downloads
Shenzhen 22 54
Beijing 8 0
Unknown 6 4
Buenos Aires 4 4
Ashburn 4 0
Denver 2 0
Seattle 2 0
Washington 2 0
Ciudad Juárez 1 0
Changsha 1 0
Redmond 1 0
Sunnyvale 1 0
Cuernavaca 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0100341/manifest

Comment

Related Items