BCR-ABL1 compound mutations combining key kinase domain positions confer clinical resistance to ponatinib in Philadelphia chromosome-positive leukemia

CML 1s caused by a random reciprocal translocation that joins the ABL1 gene on chromosome 9, with the BCR gene on chromosome 22. The result is the formation of the oncogenic BCR-ABL1 gene. This derivative chromosome is widely known as the Philadelphia Chromosome (Ph+), and it encodes a deregulated, constitutively activated tyrosine kinase called BCR-ABL1. Ponatinib is the only approved tyrosine kinase inhibitor (TKI) that suppresses all BCR-ABL1 point mutant-based resistance in Philadelphia chromosome-positive (Ph+) leukemia, including the highly resistant BCRABL1T3151 mutant. However, the development of compound mutants (2 mutations in the same BCR-ABL1 molecule) may inhibit ponatinib binding, and lead to treatment failure. We found that clinically reported BCR-ABL1 compound mutants center on 12 key positions and confer varying resistance to imatinib, nilotinib, dasatinib, ponatinib, rebastinib and bosutinib. T315I-inclusive compound mutants confer high-level TKI resistance, including to ponatinib. Structural explanations for compound mutation-based resistance were obtained through molecular dynamics simulations. Our findings demonstrate that BCR-ABL1 compound mutants present different levels of TKI resistance, requiring rational treatment selection to optimize clinical outcome.

BCR-ABL1 COMPOUND MUTATIONS COMBINING KEY KINASE DOMAIN POSITIONS CONFER CLINICAL RESISTANCE TO PONATINIB IN PHILADELPHIA CHROMOSOME-POSITIVE LEUKEMIA By Johanna Estrada A Senior Honors Thesis Submitted to the Faculty of The University of Utah In Partial Fulfillment of the Requirements for the Honors Degree in Bachelor of Science In Biology Approved: Dr. Thomas O'Hare Co-Director of Research Acivities Dr. Neil J. Vicker, Professor and Chairman Department of Biology Dr. Michael Deininger Professor and Chief, Division of Hematology & Hematological Malignancies Supervisors Dr. Darryl Kropf Professor of Biology Biology Honors Advisor Dr. Sylvia D. Torti Dean, Honors College April 2014 ABSTRACT CML 1s caused by a random reciprocal translocation that joins the ABL1 gene on chromosome 9, with the BCR gene on chromosome 22. The result is the formation of the oncogenic BCR-ABL1 gene. This derivative chromosome is widely known as the Philadelphia Chromosome (Ph+), and it encodes a deregulated, constitutively activated tyrosine kinase called BCR-ABL1. Ponatinib is the only approved tyrosine kinase inhibitor (TKI) that suppresses all BCR-ABL1 point mutant-based resistance in Philadelphia chromosome-positive (Ph+) leukemia, including the highly resistant BCRABL1T3151 mutant. However, the development of compound mutants (2 mutations in the same BCR-ABL1 molecule) may inhibit ponatinib binding, and lead to treatment failure. We found that clinically reported BCR-ABL1 compound mutants center on 12 key positions and confer varying resistance to imatinib, nilotinib, dasatinib, ponatinib, rebastinib and bosutinib. T315I-inclusive compound mutants confer high-level TKI resistance, including to ponatinib. Structural explanations for compound mutation-based resistance were obtained through molecular dynamics simulations. Our findings demonstrate that BCR-ABL1 compound mutants present different levels of TKI resistance, requiring rational treatment selection to optimize clinical outcome. ii TABLE OF CONTENTS ABSTRACT ii INTRODUCTION 1 METHODS 10 RESULTS 14 DISCUSSION 30 ACKNOWLEDGMENTS 35 REFERENCES 36 1 INTRODUCTION ABOUT LEUKEMIA Chronic myeloid leukemia (CML) is characterized as a disease of the blood and bone marrow. More specifically, CML is a cancer of white blood cells, defined by unregulated growth of myeloid cells in the bone marrow, which results in the pathogenic accumulation of these cells in the blood. Hematopoietic stem cells can differentiate into myeloid or lymphoid stem cells. A lymphoid stem cell can become further distinguished as it specializes into lymphoblasts, and eventually into highly specific B-lymphocytes, T lymphocytes and natural killer cells. Myeloid stem cells are the precursors for red blood cells (cells purposed to carry oxygen to all tissues of the body), platelets (fragments of megakaryocytes involved in hemostasis) and myeloblasts (antecedents of granulocytes, eosinophils basophils and neutrophils)(4)(Fig. 1). Granulocytes and lymphocytes are white blood cells that defend the Myeloid stem cell \ and infection. t m <* body against disease Lymphoid stem cell Myeloblast Lymphoblast \ In Granulocytes Basophil Eosinophil CML, an Red blood cells excessive number stem B lymphocyte Neutrophil I of Platelets T lymphocyte Natura| killer cell ______________ I --------- 1--------White blood cells Fig. 1. Blood cell development. A blood stem cell goes through several stages to cells become a red blood cell, platelet, or white blood cell.“ General Information about Chronic Myelogenous Leukemia.” National Cancer Institute. n.p. 2 January 2014. differentiate Web. 5 March 2014 2 into atypical granulocytes. These cells, now considered leukemic, can build up in the blood and bone marrow, and cause a harmful condition in which they outcompete platelets, red blood cells and healthy white blood cells. This interference, an encroachment on space and nutrients caused by the extreme sum ofleukemic cells, results in a decline in the population ofthe healthy, highly specialized cells needed by the body to maintain homeostasis (4). SIGNS AND SYMPTOMS The clinical signs of CML are gradual, shifting as the disease progresses through its three phases of chronic, accelerated and blast. Often, the disease is characterized by an anemia related fatigue, a result of the decline in the population of healthy red blood cells; and a lowFgrade fever accompanied by extreme sweating, a consequence of hypermetabolism. A key physiological effect of CML, is the manifestation of an enlarged spleen, as it becomes a storage site of the extra white blood cells being uncontrollably produced. This enlargem ent causes abdominal pain, prem ature satiety and as a result, a reduced food intake. These symptoms are all caused by the infringem ent of the spleen on the stomach and left upper quadrant of the abdomen (5). CLASSIFICATION There are three phases of CML: chronic, accelerated and blast, with nearly all patients being diagnosed during the chronic phase. Disease phase is defined and characterized by the number of blast cells (immature white blood cells) found in the blood and bone marrow, and the severity of signs and symptoms that are exhibited by the patient. 3 In chronic phase, <10% of the cells in the blood and bone marrow are blast cells; in accelerated phase, blast cells comprise 10% to 19% of the cells; and in blastic phase, >20% of the cells in the blood or bone marrow are considered unspecialized myeloid cells (4). When fatigue, fever and an enlarged spleen occur during blastic phase, the patient is said to be in blast crisis. This phase of CML is typically fatal within 6 months. DIAGNOSIS A variety of tests can be conducted in order to detect and diagnose CML. A general physical exam of the body is performed to detect signs of disease. Splenomegaly is a common physical finding in patients with CML, as the size of the spleen is directly related to the peripheral blood granulocyte counts, with the largest spleens correlating to patients having been observed to exhibit the highest white blood cells counts. An extremely large spleen is often an indication of the transformation into the acute blast crisis form of the disease. A complete blood count (CBC) with differential and a peripheral smear are also standard steps in diagnosis. This blood count includes a summary of the numbers and types of white blood cells, the number and size variation of red blood cells, hematocrit, hemoglobin value, platelet count, and mean corpuscular hemoglobin, measuring how much hemoglobin the patient’s red blood cells contain (6). The values derived from a CBC with differential are compared to normal ranges characterizing a healthy individual for the different parts of the CBC. CML is suspected of being present when the CBC and peripheral smear findings indicate a total WBC count of 20,000-60,000 cells/^L (5), a slight increase in basophils and eosinophils, minor to moderate anemia and abnormal platelet counts. 4 The last factor in the workup may be a bone marrow analysis that can be used to confirm the diagnosis of CML. The most important of these bone marrow findings, and the molecular signature of CML, is the presence of the Philadelphia chromosome that leads to expression of the BCR-ABL1 fusion protein. Additional factors include: hypercellularity, showing the increase of neutrophils, eosinophils and basophils in the myeloid cell line, as well as its progenitor cells; a prominence and possible increase in megakaryocytes; and finally, reticulin fibrosis, indicating the inappropriate maturation of red blood cells. PATHOPHYSIOLOGY CML is caused by a random reciprocal translocation that j oins the ABL 1 gene on chromosome 9, with the BCR gene on chromosome 22 (Fig. 2). The result is the formation Changed chromosome 9 of the oncogenic BCR-ABL1 gene. Normal chromosome 9 Chromosomes break This Changed chromosome 22 (Philadelphia chromosome) Normal chromosome 22 8 i bcr ’ derivative chromosome is widely known as ™ bcr-abl II the Philadelphia Chromosome abl U Fig. 2. Philadelphia chromosome, the result of a reciprocal translocation between chromosome 9 and 22. “ General Information about Chronic Myelogenous Leukemia.” National Cancer Institute. n.p. 2 January 2014. Web. 5 March 2014 (Ph+), in reference to the which city it in was discovered, and it encodes a deregulated, constitutively activated tyrosine kinase called BCR-ABL1. Tyrosine kinases serve as molecular switches in many cellular functions, 5 and BCR-ABL1, when locked in the “on” position, causes uncontrolled proliferation of white blood cells, and the development ofcancer. TARGETED THERAPY Imatinib, a first generation BCR-ABL1 tyrosine kinase inhibitor (TKI), was introduced in 1999 as the first clinically effective suppressor of BCR-ABL1 enzymatic activity, and has fundamentally improved CML outcomes. The success of imatinib is due to the ability of the inhibitor to occupy the ATP-binding pocket of the BCR-ABL1 kinase domain, a modification that prevents the protein from changing its conformation to the enzymatically active form (13). In the International Randomized Study of Interferon versus STI571 (IRIS) clinical trial, patients with CML received imatinib or interferon plus cytarabine as initial therapy. The results of the five-year follow-up showed a drastic contrast between the successes of the treatments, as 69% of patients at 12 months, and 87% of patients at 60 months in the imatinib group, exhibited a complete cytogenetic response (CCR)(1). A complete cytogenetic response to treatment designates that no Ph+ cells can be measured in the bone marrow by conventional or fluorescence in situ hybridization (FISH) cytogenetic testing (2). Responses were much poorer among the patients given interferon plus cytarabine, and 65% had crossed over to imatinib. Since few patients were still receiving interferon plus cytarabine at 60 months, the remainder of the report focused on the long-term follow-up of patients who received imatinib as the initial therapy for CML. This was a dramatic improvement from the previous standard therapies and the IRIS study, alongwithfurtherimatinib-based clinical trials, showedthatpatientswhohad exhibited a CCR, or whose BCR-ABL1 levels had fallen by at least 3 orders of magnitude 6 (1), were dramatically less likely to experience disease progression. Thus, CCR can be used as apowerful and predictive marker of outcome. Patients maintaining a CCR for 2 years have a projected life expectancy comparable to that of the general population (3). RESISTANCE Imatinib is, however, incapable of inhibiting BCR-ABL1 harboring certain point mutations within the ABL1 kinase domain. With eight years of follow-up, it was observed that 16% of IRIS patients discontinued treatment for insufficient efficacy and 6% for adverse events (3); the most frequently reported events being nausea, edema, vomiting, diarrhea and muscle cramps (21).BCR-ABL1 kinase domain mutations are the most pervasive resistance mechanism to TKIs, as they modify imatinib binding, or favor kinase conformations inaccessible to imatinib without substantially altering kinase activity. Consequently, allowing the reactivation of BCR-ABL1. The second generation of BCR-ABL1 TKIs, dasatinib and nilotinib, were shown through randomized trials comparing imatinib to dasatinib (Dasatinib versus imatinib study in treatment native CML patients (DASISION) study) or nilotinib (Evaluating nilotinib efficacy and safety in clinical trials of newly diagnosed Ph-positive CML patients (ENESTnd) study), to be superior with regards to complete cytogenetic response (CCR), where no Ph+ cells were measured by either conventional or FISH cytogenetic testing; major molecular response (MMR), in which the amount of BCR-ABL1 protein in the bone marrow is very low; and complete molecular response (CMR), in which no BCR-ABL1 protein is detectable in the bone marrow through the use of PCR (3). As a result of these favorable responses, both TKIs were approved for first-line therapy of CML in 2010. 7 Although nilotinib and dasatinib are effective inhibitors of most point mutants, there is the exception ofBCR-ABL1T315I, in which the native threonine residue is replaced by isoleucine. Structural studies have shown that threonine is necessary in the binding of imatinib to the kinase domain, as it stabilizes binding through hydrogen-bond interactions, as well as regulates access to a deep hydrophobic pocket in the kinase domain (7). The substitution of isoleucine physically intrudes on the inhibitor-binding site, and promotes the kinase domain to adapt the active, imatinib and nilotinib insensitive conformation (3). Bosutinib, another BCR-ABL1 TKI with an additional inhibitory effect on SRC family kinases, has a unique binding mode that is able to accommodate several kinase domain mutations conferring resistance to nilotinib and dasatinib, but is similarly susceptible to BCR-ABL1T315:, the “gatekeeper” mutation. Ponatinib, a third generation BCR-ABL1 TKI, is invulnerable to all known single kinase domain mutations, including BCR-ABL1T315:. Ponatinib and imatinib have similar binding modes, except that the carbon-carbon triple bond of ponatinib is necessary to enforce compatibility with the T315I residue (3). However, some patients with CML develop resistance to ponatinib and encounter treatment failure. Our pre-clinical and clinical evidence supports the hypothesis that resistance to ponatinib can be divided into two main categories: 1) BCR-ABL1 compound mutation-mediated resistance (two mutations in the same BCR-ABL1 molecule(Fig. 3)); and 2) resistance despite inhibition of BCR-ABL1, necessitating simultaneous inhibition of BCR-ABL1 and newly identified co-critical targets (8). 8 BCR-ABL1 polyclonal mutations C M L cc lls BCR-ABL1 compound mutations CML cells Fig. 3.Polyclonal versus compound mutations. In polyclonal mutations, BCR-ABL1 mutations (green and red stars; upper panel) exist separately in different clones. BCR-ABL1 compound mutants exhibit two mutations within the same BCR-ABL1 molecule (green and red stars; lower panel).Khorashad, J. “BCR-ABL compound mutations in tyrosine kinase inhibitor- resistant CML: frequency and clonal relationships.” Blood 121(3), 489-98 (2013) SIGNIFICANCE Sequential therapy with different BCR-ABL1 TKIs may unavoidably promote the development or selection of BCR-ABL1 compound mutants. As a result, the ability to distinguish between polyclonal and compound mutations is clinically important and can be used to determine which TKI is most suitable to overcome resistance (9). As a consequence of several compound mutations having been shown to result in resistance to ponatinib, most notably T315I-inclusive BCR-ABL1 compound mutations, this project was constructed to expound the role of BCR-ABL1 compound mutations in ponatinib resistance, and to identify the nature and presence of compound mutations in CML patients at the start of ponatinib therapy, compared to at treatment failure. This 9 study’s findings have important implications for optimizing therapy selection in patients harboring compound mutations, and have the possibility to facilitate maximum disease control in a large proportion of patients with Ph+leukemia (10). 10 METHODS A Note on Methods: My involvement in the project centered on Methods #3-13, in collaboration with Srinivas Tantravahi, MD. In order to aid the reader in understanding how this work fits into the broader scope of the project, I have included additional procedures and acknowledged the people who did the work. Methods #1 and 2 were the responsibility of Matthew Zabriskie and Kimberly Reynolds, respectively, research scientists in the Deininger/O’Hare laboratory. Method #14 was carried out by Christopher Eide, a research scientist in the Druker laboratory at Oregon Health & Science University. Methods#15 and 16 were the work of postdoctoral associateNadeem Vellore and Professor Riccardo Baron, University of Utah Department of Medicinal Chemistry. 1. Cellular Proliferation Assays Ba/F3 BCR-ABL1-expressing cells were plated in 96-well plates (2x103 cells/well) and incubated in 2-fold escalating concentrations of dasatinib, ponatinib (0-768 nM), imatinib, nilotinib, rebastinib, or bosutinib (0-10240 nM) for 72 h. Proliferation was assessed by methanethiosulfonate (MTS)-based viability assay (CellTiter 96 AQueous One; Promega). IC50 values are reported as the mean of three independent experiments performed in quadruplicate. 2. Isolation of Primary Ph+ Leukemia Cells from Blood or Bone Marrow Patients were consented in accordance with the Declaration of Helsinki and the Belmont Report, and University of Utah Institutional Review Board approved all studies with human specimens. Mononuclear cells (MNCs) were isolated from primary patient peripheral blood or bone marrow specimens by Ficoll-separation. CD34+ cells were enriched by magnetic column separation using a CD34 human microbead kit and the 11 POSSELDS program (AutoMACS; Miltenyi). Purity of the CD34+ fraction was determined to be >90% by fluorescence-activated cell sorting. If MNC yield was limiting (<2e7 cells), the RNA isolation described below was done with an aliquot of MNCs. 3. RNA isolation from primary CML cell lysates at time of TE and TF RNA obtained from primary Ph+ leukemia cell lysates was extracted using the QIAGEN RNeasy Mini Kit, and was used to serve as a template for the synthesis of complementary DNA. 4. RNA quantitation RNA was quantified using the BioTek Epoch instrument equipped with aTake3 Volume plate and Gen5 software. 5. cDNA Synthesis Complementary DNA was synthesized using the BioRad iScript cDNA Synthesis Kit 6. Amplification of BCR-ABL1 kinase domain using two-step nested PCR Amplification of the BCR-ABL1 kinase domain was done by two-step PCR to exclude amplification of normal ABL1. Amplification was achieved through the use of Denville Scientific Choice-Taq DNA Polymerase, a master mix containing DNA polymerase and nucleotides (dNTPs). The primers used were B2A and NTPB+ for forward primers and JamR and NTPE- as reverse primers. The synthesized cDNA was used as the DNA template for the reaction. The PCR reactions were run on the Bio Rad C1000 thermal cycler. 7. Analytical gel electrophoresis The amplified BCR-ABL1 kinase domain from primary Ph+ cells and bacterial colonies, was electrophoresed on 2% Agarose gel to confirm the amplification of a single fragment. 12 8. Purification The amplified fragment from the primary Ph+ cells was purified (QIAquick PCR Purification KIT) 9. Conventional sequencing of the BCR-ABL1 Kinase Domain The now purified and amplified fragment of the BCR-ABL1 kinase domain was subjected to conventional sequencing in both directions using BigDye terminator chemistry on ABI3730 instrument and compared to the ABL1 gene. 10. Cloning ofamplified fragments and introduction into E. coli TOP10 cells The amplified BCR-ABL1 kinase domain was cloned (InvitrogenTOPO cloning system) and introduced into E. coli TOP10 cells (Invitrogen) 11. PCR amplification of 100 bacterial colonies each containing an individual BCRABL1 kinase domain amplicon For amplicon sequencing, individual bacterial colonies (average: 85/specimen; range: 23-100), each carrying a recombinant plasmid with a single BCR-ABL1 kinase domain amplicon inserted, were subjected to BCR-ABL1 kinase domain amplification. (BioRad C1000 thermal cycler) 12. Analytical gel electrophoresis The amplified BCR-ABL1 kinase domain from primary Ph+ cells and bacterial colonies was electrophoresed on 2% Agarose gel to confirm amplification of a single fragment. 13. Sequencing of 100 BCR-ABL 1 kinase domain amplicons obtained from 100 bacterial colonies Individual amplicons from the bacterial colonies were sequenced in both directions (Beckman Coulter Genomics) 14. DNA Sequence Analysis DNA sequence analysis encompassing missense and silent BCR-ABL1 kinase domain mutations was accomplished with Mutation Surveyor software. 15. Molecular Dynamics Simulations Mutant conformations of the ABL1 kinase were prepared using standard methods to generate ABL 1Y253H/E255V, ABL 1E255V/T315I, and ABL 1I315M.For each mutant, both the active and inactive conformations of ABL1 kinase were created. The NAMD simulation package was used for molecular dynamics simulation, and the Amber ff12SB force field was employed for standard protein parameters. 16. Docking Simulations The Schrodinger suite of programs (Suite 2012: Maestro, version 9.3) was used for docking studies. In the final 50 ns of the simulation, 50 conformations were extracted as docking receptors. Selected conformations were prepared using Protein Preparation Wizard. Ligands (ponatinib and dasatinib) were prepared (Suite 2012: LigPrep, version 2.5) and initial docking simulation was performed using the GlideXP module (version 5.7) of the Schrodinger program. To enhance binding conformations and allow receptor flexibility, docked conformations were subjected to induced fit simulations. Docking scores were computed using the GlideXP module. RESULTS A RESTRICTED SET OF KEY BCR-ABL1 KINASE DOMAIN RESIDUES ARE HEAVILY REPRESENTED IN CLINICALLY REPORTED BCR-ABL1 COMPOUND MUTANTS Although there are many distinct single BCR-ABL 1 mutants, a careful inventory of BCR-ABL1 compound mutations associated with TKI resistance, have confirmed the involvement of 12 key residues in the resistance of one or more TKIs: dasatinib, imatinib, nilotinib, bosutinib, and ponatinib. These residues have been identified as: M244, G250, Q252, Y253, E255, V299, F311, T315, F317, M351, F359, and H396 (Fig.4). All clinically reported compound mutations (100%) involve at least one of the Fig.4. Crystal structure of the ABL1 kinase domain in key mutated residues, and the complex with imatinib (green). Twelve key positions majority accounting for most clinical BCR-ABL1 TKI resistance, (65%) involve two. including compound mutation-based resistance, are highlighted (orange; T315 is in red). Clinical examples of the T315I mutation paired with all other key residues except 317 and 299 have been documented, and among the 66 possible pairings of the 12 key resistance positions, 33 (50%)havebeen reported to date. Although variations at the specific substitution level have been reported for several pairings, such as T315I/F359C and T315I/F359V in the T3151-inclusive category, as well as E255K/F317L and E255V/F317I in the non-T315I category, data confirms that compound mutations are largely derived from a limited set of mutated residues within the kinase domain of BCR-ABL1. The involvement of these 12 key residues in TKI resistance, is only further supported by their frequency in the assessment of the baseline conventional sequencing traces for 439 patients entering the phase 2 Ponatinib Ph+ALL and CML Evaluation (PACE) trial (14). Enrollment required: 1) resistance to, or unacceptable toxicity from, nilotinib or dasatinib or 2) a documented baseline T315I mutation. This evaluation identified that mutations occurring in more than one patient were confined to 16 positions, including 11/12 key resistance residues (with the exception of position 252). In total, 95.4% of the mutations observed in baseline pre-ponatinib specimens from the PACE trial, occurred at key resistance residues. With respect to EOT (end of treatment) specimens, all but one of the compound mutations inferred from conventional sequencing traces were composed exclusively of key resistance residues (15/16 (93.8%)); the only non-key residue, E275K, has been implicated in imatinib resistance (15). CLINICALLY AVAILABLE TKIS EXHIBIT DIFFERENTIAL ACTIVITY AGAINST BCR-ABL1 COMPOUND MUTANTS Proliferation assays comparing six TKIs were performed with Ba/F3 cells (an increasingly popular cell line used as a model system to produce efficient kinasedependent cellular assay systems (12)) expressing native BCR-ABL1, BCR-ABL1 single mutants at each of the 12 key positions, and clinically reported BCR-ABL1 compound mutants. With the exception of I315M, each single mutant was effectively inhibited by at least one TKI. The six TKIs displayed partially overlapping resitance profiles for BCR-ABL1 single mutants, with T315I inhibited only by ponatinib and rebastinib (DCC-2036)(Fig.5). T315I-inclusive compound mutants were insensitive to all TKIs except ponatinib and rebastinib, which exhibited only marginal efficacy in most cases. The most resistant mutant, E255V/T315I, exhibited 11.9- and 22.7-fold higher ponatinib resistance than E255V or T315I, respectively. The Q252H/T315I, T315I/M351T, T315I/F359V and T315I/H396R mutants displayed minimal ponatinib sensitivity and high-level rebastinib resistance. M244V/T315I was the only T315I-inclusive compound mutant in the panel predicted to be sensitive to ponatinib and rebastinib at physiologically achievable levels (Fig5). All non-T315I compound mutants were inhibited by one or more TKI. For some compound mutants (e.g.Y253H/F317L), several TKI options exist. For others, a single TKI stands out as the leading choice, most notably dasatinib for Y253H/E255V (Fig. 5). In summary, non-T315I BCR-ABL1 compound mutants exhibited a spectrum of TKI sensitivities, suggesting in vitro resistance profiles may serve as a guide for clinical TKI selection. HIGH RESOLUTION DYNAMIC COMPUTATIONAL MODELING OF THE Y253H/E255V COMPOUND MUTANT RATIONALIZES DIFFERENTIAL SENSITIVITY TO TKIS The X-ray crystallographic structure of the ABL1 kinase in complex with ponatinib and imatinib has shown that both TKIs have similar binding modes. A short motif near the N-terminal region of the activation loop assumes a particular conformation denoted to as “DFG-out.” Ponatinib binds to the mode and hinders the conformationally active form of the kinase (11). DFG-out 17 Fig. 5. Heat map of TKI IC50s for single and tandem mutants. I315M and T315I/E453K are included for comparison. A color gradient from green (sensitive) to yellow (moderately resistant) to red (highly resistant) denotes the IC50 sensitivity to each TKI. Ponatinib binding at the BCR-ABL1 kinase domain is centered on an adenine pocket extending from the phosphate-binding loop (P-loop) to the C-helix region in which the trifluoromethyl element ofponatinib tightly binds the pocket produced by the DFG-out conformation of the protein. In contrast, dasatinib binding is characterized by less conformational constraints, and is less reliant on direct P-loop and C-helix interactions (16). 18 Since ponatinib compared favorably with dasatinib against all non-T315I compound mutants except Y253H/E255V, we investigated structural features that account for the striking difference in the case of this mutant. These features were examined by molecular dynamics (MD) simulations carried out for a prolonged interval (100 ns), and docking simulation was completed by the GlideXP method on a collection of 50 conformations of the Y253H/E255V mutant, extracted at regular intervals. It was observed that the introduction of these two mutations markedly shifts the positions of both residues, and consequently, collapses the P-loop partially into the ponatinib-binding site (Fig. 6). In particular the replacement of the □Y253H/E255V □ Native negatively charged and highly hydrophilic C-helix glutamate residue, with the hydrophobic valine residue at position 255, ponatinib results in the positioning of the valine side-chain Fig. 6. Structural alignment of native ABL1 kinase with the ABL1Y253H/E255V mutant kinase (blue). Ponatinib is shown in inside of the ABL1 kinase wireframe. Native positions Y253 and E255 (green) and mutated domain, in order positions H253 and V255 (red) are shown as spheres. The region of the to backbone (P- loop, C-helix and DFG loop) that showed major conformational change is highlighted in either orange (native) or blue (Y253H/E255V). placate its hydrophobic nature. The structural changes within the P-loop also trigger distortions within the Nterminal portion of the C-helix at the distal end of the ponatinib binding site. Specifically, loss of Y253-F382 aromatic pi stacking pushes F382 into the ponatinib-binding site. The structural changes of Y253H/E255V mutant have further negative results with respect to 19 ponatinib binding, as N-terminal C-helix distortion results in the loss of the critical K271E286 salt bridge,whose absence shifts the positions of resides L248, K271, E286, and R362 and unfavorably affects ponatinib binding. In contrast, modeling predicts that in dasatinib, the formation of a favorable hydrogen bond between the side chain of H253 and the amide nitrogen, is assisted in the Y253H/E255V mutant (Fig. 7), and the structural changes experienced by the mutant residues do not impose on the dasatinib-binding site, as compared to the native kinase □Y253H/E255V domain. M290 Dasatinib binding was also shown to K271 dasatinib be more effective than binding H253 Fig. 7. Relative positions of selected residues in the dasatinib binding site of ABL1 Y253H/E255V (cyan). The Y253H substitution establishes a new hydrogen bond between this residue and dasatinib, in addition to the existing T315-dasatinib hydrogen bond (red dashed lines). The chemical structure of dasatinib is shown inside of a space-filling wireframe. ponatinib to the Y253H/E255V mutant in docking the score histograms, which identified several dasatinib-binding conformations as being energetically favored over the lowest ponatinib binding conformation, with a gap of nearly 2 kcal/mol between the lowest energy conformers for the respective TKIs. Experimental results only further support the predictions based on computational modeling, by demonstrating that dasatinib is a more potent inhibitor of Ba/F3 cells expressing the BCR-ABL1Y253H/E255Vcompound mutant, than ponatinib. Dasatinib, as indicated by these results, is the best TKI, in terms of potency, for controlling the Y253H/ E255V mutant. CONVENTIONAL AND AMPLICON DEEP SEQUENCING ESTABLISH CORRELATIONS BETWEEN BASELINE MUTATIONS AND RESPONSE TO PONATINIB To elucidate the role of compound mutations in resistance to ponatinib, we analyzed 100 specimens from 64 patients treated on the PACE trial (N=50) or other approved protocols, including the ponatinib expanded access program (N=14). The BCR-ABL1 kinase domain of the specimens was analyzed using conventional Sanger sequencing, as well as the sequencing of ~85 individual BCR-ABL1 kinase domain amplicons per specimen. With the amplicon sequencing being at least an order of magnitude more sensitive than conventional sequencing, and permitting dependable identification of mutations expressed in as little as ~2% of the total population. A second critical advantage of amplicon sequencing is the capacity to decisively distinguish between compound mutations and unlinked, point mutations. Baseline samples were evaluated for all patients; for 30 patients, longitudinal and/or EOT (End of treatment) specimens were also analyzed. Patients were grouped according to their baseline mutation status as determined by conventional sequencing into (i) patients with T315I, (ii) patients with mutations other than T315I and (iii) patients with no detectable BCRABL1 kinase domain mutations. Patients with a baseline T315I mutation. T3151 mutations were detectable in 22 of 64 patients (34.4%) at baseline by conventional sequencing, including: 8 CML-CP (Chronic phase), 6 CML-AP (Acute phase), 5 CML-BP (Blast phase) and 3 Ph+ ALL (Acute Lymphoblastic Leukemia) patients. The T315I mutation was the only baseline mutation detected in 18/22 specimens; three specimens carried a second detectable baseline mutation: K285E (Patient #2), F317L (Patient #10) and H396R (Patient #18). Patients with mutations other than T315I at baseline. Among the remaining patients surveyed, 17/64 (26.6%) had one or more nonT315I baseline mutations detected by conventional sequencing: 9CML-CP, 3 CML-AP, 2 CML-BP and 3 Ph+ ALL. Baseline mutations representing 9 of 11 non-T315I key resistance residues were observed among these examples (all except positions 244 and 311). The majority of the baseline samples (11/17; 64.7%) harbored a baseline mutation at a single position; the remainder (6/17; 35.3%) carried two baseline mutations. These included: F317L;E450G (#23), F317L;E459K (#27), E255V;F317L (#32),F317I;F359V (#34), Y253H; E255V (#35) and G250E; F317L (#39). Patients with no detectable mutations at baseline. Twenty-five pre-ponatinib specimens (25/64; 39.1%) showed no evidence of a baseline mutation by conventional sequencing: 12 CML-CP, 5 CML-AP, and4 CML-BP and 2 Ph+ ALL. Notably, none of the patients lacking a detectable mutation at baseline, including those who subsequently discontinued ponatinib therapy, were found to harbor a detectable compound mutation in their baseline or EOT amplicon sequencing profile. These results may reflect a degree of BCR-ABL1 independent resistance, potentially through activation of alternative oncogenic pathways, in this heavily pre-treated group. For patient populations carrying a baseline T315I or non-T315I mutation, we next evaluated outcomes on ponatinib therapy. T3151-INCLUSIVE COMPOUND MUTATIONS ARE ASSOCIATED WITH PONATINIB FAILURE Supporting our in vitro profiling of compound mutations including the T315I mutant residue, which concluded that most pairings of a key resistance residue with 22 T315I, would result in moderate- to high-level resistance to ponatinib; we observed three patients who exhibited clinical failure of ponatinib with a noticeable development of a T315I-inclusive compound mutant. Patient #38 began ponatinib treatment with Ph+ALL, and had previously shown resistance to imatinib and dasatinib. By conventional sequencing, there was detection of a baseline E255V point mutation, and a low-level (~15%) T315I mutation was also observed in the conventional sequencing trace. Baseline amplicon sequencing confirmed the existence of the E255V mutation, present in 85% of amplicons, and additionally revealed an E255V/T315I compound mutant, present in 15% of amplicons. The patient had a temporary response to ponatinib, followed by rapid relapse at 7 weeks, at which time the E255V/T315Imutant (the most ponatinib-resistant compound mutant identified in pre-clinical studies) was highly predominant, detected in 69% of amplicons (17) (Fig.8). In silico (through computer modeling) MD simulations were performed in order to better understand E255V/T315I resistance to ponatinib. It was revealed that the large A Fig.8. A)Summary of BCR-ABL1 amplicon sequencing for Patient #38. B) Molecular dynamics comparison of E255V/T315I and T315I mutants C) Structural realignments in E255V/T315I reduction in ponatinib affinity toward the E255V/T315I mutant, as compared to its affinity to either just a single E255V or T315I mutation, is a result of the re-orientation of the P-loop and the N-terminal region of the C-helix. Both of these structural features are needed to accommodate the hydrophobic valine side-chain of V255, and these adjustments force the L248 and I315 side chains into the ponatinib-binding site. The Chelix distortion also results in steric hindrance in the ponatinib-binding site, particularly due to repositioning of residues M290, F359, and D381. Finally, a reduction in the distance between F382 and I315 unfavorably narrows the channel into which ponatinib normally binds. Two additional clinical examples, in which baseline point mutations experienced disease progression and resulted in the emergence of a T315I compound mutation, are discussed below. In the first case, a CML-BP (Patient #36), with a baseline F359C mutation was found to have a T315I/F359C compound mutation at EOT that was not detectable in the baseline amplicon sequencing profile (Fig. 9A). In vitro profiling of the related T315I/F359V mutant predicts that compound mutations with a pairing of these key residues, are expected to be moderately resistant to ponatinib at clinically achievable concentrations. In the next example, a CML-AP patient (Patient #12), with a baseline T315I mutation, experienced disease progression with emergence of a predominant T315I/E453K compound mutation (90% of amplicons) that was not detectable in the baseline amplicon sequencing profile (Fig. 9C). To date, point mutations at position 453 (E453G/K) have been associated with imatinib resistance (18) but not to compound mutation-based resistance. 24 Patient 36 Prior TKIs: D, N Patient 12 Prior TKIs: I, D, N A m plicon S equencing Pre-ponatinib: 100-1 Pre-ponatinib: sample 12-360A May 2011 --------------------- sample 12-342A Pre-ponatinib: CML-BP CML-AP Baseline conventional seq: Baseline conventional seq: F359C (100%) T315I (90%) co 0 40­ ° § K ey None Mutation: G250R/ V299M/ F311S/ T315I/ T315I/ F359C T315I/ F359C F359C F359C F359C/ F359C/ H396Y H396Y May 2012 ---------------------► Longitudinal: EOT: T315I; E453K (100%; 100%) T315I; F359C (100%; 100%) Key Mutation- EO T due to progression None B L Key + Additional Ponatinib "o -—. 80n _ II §I 60^ ^ 1 | 40-1 °§ m 'H K pv 0 Key Mutation: G250R/V299M/ F311S/ T315I/ T315I/ F359C t ^ si/ F359C F359C F359C F359C/ F359C/ H396Y H396Y I Key Mutation sample 12-34 100-1 Longitudinal conventional seq: EOT conventional seq: 1 None p i ^ T315I ^ T315I/ 1 E453K EOT conventional si T315I; E453K (100%; 85%) T3 80- sa5z0II560­ jj 8 g 40o o 20- EO T due to progression Key | Key Mutation | Key + Additional • Prominent Non-Key , • Prominent Non-Key Mutation + Additional Fig.9.A)Summary of BCR-ABL1 amplicon sequencing for Patient #36 B)Fold-change in ponatinib Ba/ F3 cellular IC50 values for BCR-ABL1F359V and BCR-ABL 1T315I/F359V compared to native BCR-ABL1. C)ummary of BCR-ABL1 amplicon sequencing for Patient #12 D) Fold-change in ponatinib Ba/F3 cellular IC50 values for BCR-ABL1T315I and BCR-ABL 1T315I/E453K compared to native BCR-ABL1. Ba/F3 BCR-ABL 1T315I/E453Kcells showed a substantially higher level of ponatinib resistance relative to those expressing the T315I mutant and were insensitive to all other TKIs tested except rebastinib. Both ponatinib and rebastinib were effective only at clinically unachievable concentration. In each of these three examples, the development of a T315I-inclusive compound mutation resulted in TKI resistance. Interestingly, however, in only one of the three cases (E255V/T315I) was the EOT compound mutant detected at baseline and only by amplicon sequencing, suggesting the treatment failurecausing mutation was acquired on therapy or was below the detection limit of amplicon sequencing in the other two cases. By contrast, we identified two additional patients (Patients 17 and 18) in which principal compound mutations existent at baseline and predicted to confer ponatinib resistance (Y253H/T315I and T315I/H396R, respectively) also predominated at EOT. These outcomes advocate that although not all compound mutants detected at high frequency at relapse may be easily detected at baseline, T315I-inclusive compound mutants can result in substantial inhibition of ponatinib binding and lead to high-level clinical resistance. THE I315M MUTATION ARISES FROM T315I AND CONFERS HIGH-LEVEL PONATINIB RESISTANCE In nearly all instances in which there was detected a noticeable shift from the baseline mutational status to a predominant BCR-ABL1 mutant clone at the time of treatment failure, it was in direct consequence of the presence of a compound mutation involving two of the 12 key residues. However, a patient was identified in which a previously unrecognized point mutation-based escape route, resulted in a secondary point mutation of the same residue, 315. Patient #22 (Ph+ ALL with a T315I mutation at baseline) achieved a complete cytogenetic response on ponatinib but progressed after 7 months. Longitudinal and EOT amplicon sequencing profiles revealed a newly emergent and increasingly dominant secondary point mutation in the codon for position 315, causing a change of the baseline I315 to methionine (I315M) through a single nucleotide change nucleotide change (ATT to ATG) (Fig. 10A). We previously recovered the ponatinib resistant I315M mutant in Ba/F3 BCRABL1T31SI cell-based resistance screens, and in vitro profiling of Ba/F3 BCR-ABL1I315M cells confirmed pan-TKI resistance. The level of ponatinib resistance conferred by the I315M mutant exceeded all tested single and compound mutants except E255V/T315I. To understand the structural basis for I315M-based ponatinib resistance, we modeled this mutant using in silico MD simulations and found that the methionine residue directly encroaches on the ponatinib binding site. Resulting structural adjustments at residues 269,290,317,359 and 381 also disfavor ponatinib binding. The interaction of M315 with residue M290 of the C-helix, in particular, leads to a disruption in the hydrophobic spine architecture (19). Also, the shift of M315 towards M290 in turn causes F317 to occupy the region of the adenine pocket that overlaps with the ponatinib-binding site. These findings illustrate that I315M as a single point mutation can lead to ponatinib treatment failure. /\ B Patient 22 Prior TKIs: I C Ponatinib Amplicon Sequencing Apr 2011 - sample 11-0 Ph+ALL I Is .' »8 E255K E255K/ E255K/ E255V/ E255V/ T315I T315I/ 1315M L II60 Si E255K E255K/ E255K/ E255V/ E255V/ T315I T315I/ 1315M EOT due to progression Ponatinib' Mutation: ■ Key Mutation ■ F359 Key + Additional Fig. 10-A) Summary of BCR-ABL1 amplicon sequencing for Patient #22 .B) Fold-change in ponatinib Ba/F3 cellular IC50 values for BCR-ABL1 I315M and BCR-ABL1 T315I compared to native BCRABL1. C) Ponatinib in complex with ABL1T315I; ABL1I315M is superimposed.D) M315 penetrates deeply into the ponatinib site, and structural adjustments at positions 269, 290, 317, 359 and 381 also disfavor ponatinib binding. 27 NON-T3151 COMPOUND MUTATIONS IMPART DIFFERENTIAL LEVELS OF TKI RESISTANCE In vitro profiling of non-T315I compound mutations showed varying levels of sensitivity to tested TKIs, with 7/8 lines demonstrating sensitivity to ponatinib. Among patients for whom EOT samples were available, only one (Patient #37) demonstrated clear evidence of a non-T315I compound mutation at failure. Following treatment with imatinib and dasatinib, this patient exhibited a baseline F317I mutation. The patient experienced disease progression and discontinued ponatinib, with EOT sequencing revealing an E255V/F317I mutation (100% of amplicons). A second patient (Patient #34) diagnosed with CML-AP, was treated with imatinib for several years, but was subsequently switched to dasatinib in response to intolerance and limited response. After a temporary response to dasatinib she experienced hematologic relapse and was switched to ponatinib. Conventional sequencing identified two dominant mutations (F317I and F359V) in the baseline sample preceding ponatinib therapy, which were confirmed as 100% F317I/F359V compound mutation by amplicon sequencing. A complete hematological response was achieved within one month of ponatinib therapy, and at 7 months a major molecular response was reached. Amplicon sequencing done at this time confirmed that the maj ority of the clones were F317I/F359V (87%). MMR is sustained as of last follow-up (January 2014), and these findings are also consistent with our in vitro profiling data signifying that compound mutants pairing these two residues remain sensitive to ponatinib. A CML-CP patient (Patient #23) who had previously been treated with four successive TKIs (imatinib, dasatinib, nilotinib, bosutinib) exhibited an F317L/E450G compound mutation (83% of amplicons) in the baseline amplicon sequencing profile. The E450G substitution is an uncommon mutation associated with weak imatinib resistance (20) and has not been previously reported in the setting of a compound mutation. The patient achieved a CHR on ponatinib and remained in stable chronic phase over a period of greater than two years. Ponatinib treatment was recently discontinued, for undisclosed reasons, while the patient was still in CHR. EOT sequencing showed no F317L/E450G, suggesting this mutant is resistant to previous TKIs but sensitive to ponatinib. B Patient 37 Prior TKIs: I, D Patient 34 Prior TKIs: I, D A m plicon Sequencing Pre-ponatinib: Ph+ ALL Baseline conventional seq: Sep 2012 ---------------► sample 11-003 100n Pre-ponatinib: CML-AP Baseline conventional seq: 80' F317I (100%) K ey A m plicon Sequencing Pre-ponatinib: sample 12-303 Pre-ponatinib: F317I; F359V (100%; 100%) E255V/ E255V/ V299L V299L/ F317I F317I F317I None V299A/ F317I/ F317I/ F359V F359V F359V None V299A/ F317I/ F317I/ F359V F359V F359V Mutation: V299M/ F317I Longitudinal conventional seq: EOT conventional seq: F317I; F359V (100%; 100%) E255V;F317I ( 100 %; 100 %) EOT due to progression — i----------- 1----------- 1----------- 1----------- 1— F317I Key 0 E255V/ E255V/ V299L V299U Mutation: V299M/ F317I F317I MMR F317I Key + Additional Prior TKIs: I, D, N, B Oct 2010 -------------Pre-ponatinib: D A m plicon Sequencing Pre-ponatinib: 100 t tion P atient 27 Prior TKIs: I, D Key + Additional A m plicon Sequencing Pre-ponatinib: sample 12-356B Pre-ponatinib: CML-CP CML-CP Baseline conventional seq: F317L; E450G (90%, 90%) Key Mutation None F311S/ F317L/ E450G F317L F317S L F317L/ F317L/ F359V E450G F359V Baseline conventional seq: F317L; E459K (100%; 100%) Key u Mutation: E450G sample 12-356C Apr 2013 EOT conventional seq: Longitudinal conventional seq: No mutations F317L (40%) Key Mutation U ndisclosed reasons ■ None F311S/ F317L/ E450G F317L F317S F317L/ F317L/ F359V E450G | Key + Additional • Prominent Non-Key , Mutation • Prominent Non-Key Mutation + Additional F359V E450G | Key Mutation | Key + Additional • Prominent Non-Key • Prominent Non-Key Mutation + Additional Fig. 11. A) Summary of BCR-ABL1 amplicon sequencing for Patient #37 B) Summary of BCR-ABL1 amplicon sequencing for Patient #34 C) Summary of BCR-ABL1 amplicon sequencing for Patient #23 D) Summary of BCR-ABL1 amplicon sequencing for Patient #27 Finally, A CML-CP patient (Patient #27) who had previously been treated with imatinib and dasatinib, exhibited an F317L/E459K compound mutation (98% of amplicons) in the baseline amplicon sequencing profile. After a year on ponatinib therapy and achievement of CHR but neither CCyR nor MMR, F317L/E459K was reduced to a minor component (6% ofamplicons), suggesting sensitivity to ponatinib. 30 DISCUSSION Drug resistant compound mutations within the BCR-ABL1 kinase domain are an emerging clinical problem for patients receiving sequential TKI therapy. As we predicted for ponatinib and rebastinib, some of these mutations confer resistance that is several-fold higher than that of either contributing mutation in isolation. We prospectively investigated the role of BCR-ABL1 compound mutations in TKI resistance, focusing on ponatinib due to its unique effectiveness against the T315I single mutant and clinical availability. In the U.S., ponatinib is approved for patients with refractory CML or Ph+ ALL harboring a T315I mutation, or for whom no other TKI is indicated, based on results of the PACE trial demonstrating significant activity at a median follow-up of 15 months. Despite the impressive efficacy of ponatinib, ponatinib does experience resistance. Our analysis has led us to believe that there are two fundamental escape routes involved in ponatinib resistance. The first route implicates compound mutations as being the mechanisms of BCR-1 dependent resistance. The second escape route centers on BCRABL1 independent resistance, resistance that occurs despite TKI mediated inhibition of the BCR-ABL1 kinase domain. Although acknowledging the existence and importance of both systems conferring resistance, our work focuses primarily on BCR-ABL1 dependent resistance, more specifically as it relates to compound mutations. Our in vitro studies predict that at a 30 mg/day dose, compared to the starting dose of 45 mg/day, ponatinib is efficient against 7/8 non-T315I compound mutants tested in our panel. At a daily dose of 15mg, ponatinib is predicted to prevent outgrowth of 5/8 non-T315I compound mutants in our panel, with the Y253H/E255V, E255V/V299L and F317L/F359V mutants remaining potentially problematic. In contrast, therapeutic utility is less promising with respect to T315I-inclusive compound mutants, where 9/10 in our panel showed little or no sensitivity to ponatinib or any o f the other TKIs tested. We provide examples o f clinical ponatinib failure attributable to E255V/T315I, T315I/F359C, Y253H/T315I, T315I/H396R and T315I/ E453K. Given ponatinib’s unique efficacy against the T315I single mutant and its current revised U.S. clinical indication, a significant fraction o f future patients treated with ponatinib will be expected to harbor a T315I mutation at baseline. More sensitive, routine screening ofbaseline samples from these patients may be warranted to determine whether problematic T315I-inclusive compound mutants are present. Detection o f two mutations by conventional sequencing does not provide sufficient information to identify the best treatment option since this may represent two clones, each with a single mutation. In contrast, amplicon sequencing readily discerns compound from polyclonal mutations. For example, while Patients #10, #26, and #35 each had two baseline mutations by conventional sequencing, amplicon sequencing demonstrated mutual exclusivity of these mutations at the clonal level. Verification that Y253H and E255V exist in different clones as opposed to as a highly ponatinib resistant Y253H/E255V compound mutant (Patient #35) is of importance for clinical decision­ making. In our study, no patient beginning ponatinib with native BCR-ABL1 by conventional sequencing exhibited a contributing compound mutation at EOT, similar to reports on BCR-ABL1 single mutants. These results suggest that patients who fail multiple TKIs without a BCR-ABL1 mutation are unlikely to experience ponatinib failure due to emergence o f a resistance-conferring compound mutation. Effective therapy for these patients may require an approach involving the blockade of a second pathway in addition to BCR-ABL1. Additionally, consistent with inferred compound mutational status data reported in the PACE trial, we found detection of compound mutations at EOT to be more frequent among CML-BP and Ph+ ALL patients than those with CML-CP, suggesting an increased risk of compound mutation- based ponatinib resistance in advanced disease. Although we identified a number of resistance-conferring compound mutants at EOT, longer follow-up and a larger number of specimens will be required to make definitive prognostic use of baseline sequencing profiles of patients beginning a new TKI. Among the 12 key positions identified, there appear to be pairing constraints for generation of TKI resistant compound mutants. For example, the current snapshot of reported compound mutations includes position 315 in tandem with 9/11 other key positions, whereas position 252 has only been reported in tandem with positions 255 and 315. While additional resistant pairings will undoubtedly be observed in the future, findings to date may suggest that only a limited number of compound mutation possibilities avoid damaging, non-tolerated effects on kinase function or fitness of the mutated clone. The broad potency of ponatinib against BCR-ABL1 point mutants can be traced to the extensive network of contacts that stabilize its binding to the kinase domain. However, certain pairings of mutations, each of which is susceptible to a given TKI in isolation, confer increased resistance when present as a compound mutation. Molecular dynamics-guided modeling performed for Y253H/E255V, E255V/T315I and I315M reveals commonalities that could aid in designing TKIs to treat compound mutants. For instance, several compound mutations involving the P-loop result in significant distortion of this region, suggesting it may prove advantageous to minimize direct TKI/P-loop interactions. Also intriguing isour characterization of an I315M pointmutation in clinical resistance to ponatinib due to direct encroachment of the mutant side chain on drug binding (Patient # 22). Notably, ponatinib is a poor inhibitor of kinases in which methionineisthenative ‘gatekeeper’ positionanalogoustoBCR-ABL1 position 315. For example, the insulin receptor is ~500 fold less sensitive to ponatinib than ABL1T315:. By contrast, the T315A mutant is uniquely resistant to dasatinib and inhibited by each of the other five TKIs including ponatinib. These findings argue that efforts to develop new BCR-ABL1 TKIs should also consider the capacity to accommodate multiple different specific substitutions at the gatekeeper position. Computational methods have been applied to BCR-ABL1 single and compound mutants to predict TKI binding, including ponatinib. The use of different computational methods and the inherent limitations of modeling mandate experimental validation of predictions. For example, one computational approach identified T315I/F359V as ~4fold more resistant to ponatinib than Y253H/T315I. In contrast, our comprehensive, direct experimental comparison o f compound mutants and findings in cell-based resistance analyses identify Y253H/T315I as 3.5-fold more ponatinib resistant than T315I/F359V. We performed a comprehensive investigation of BCR-ABL1 compound mutations as major mechanisms for ponatinib resistance. Specifically we: 1) determined that resistance to ponatinib is attributed to pairwise combinations of twelve key residues; 2) we profiled and directly compare six (5 FDA approved, one investigational) BCRABL1 TKIs against a cell line panel of clinically-reported BCR-ABL1 compound mutants. This has important ramifications for clinical decision-making, by allowing physicians to better match existing TKIs to the compound mutants they inhibit most effectively; 3) We also demonstrated that while most T315I-inclusive compound mutations frustrate all TKIs including ponatinib, most non-T315I compound mutations are sensitive to at least one TKI; 4) Utilizing in silico molecular dynamics simulations we provided structural explanations and unifying concepts for compound mutation-based resistance that can help to guide the development of new TKIs; 5) by screening for BCRABL1 compound mutations in baseline, longitudinal and/or end-of-therapy specimens from more than 60 Ph+ leukemia patients treated with ponatinib, we hope to provide a framework for individualized TKI selection based on mutation profile. We believe this work has important implications for clinical decision-making in Ph+ leukemia. 35 ACKNOWLEDGMENTS Foremost, I would like to express my sincere gratitude to my advisor, Dr. Thomas O’Hare, Co-Director of Research activities, whose continuous support and assistance has guided me in all stages of my research, and in the completion of this thesis. I could not have been paired with a better mentor, and have benefited from our interactions immensely. I am extremely grateful for his influence thus far, and look forward to our future collaborations. I would like to thank all the members of the Deininger/O’Hare Lab who have trained and assisted me in my work as an undergraduate research assistant, and provided me with constant access to professional input and advice, in all aspects of this project. A special thanks to Dr. Srinivas Tantravahi, Hematology and Oncology Fellow, who I had the privilege to work with regularly, for the maj ority of this proj ect. Thank you for encouraging me to understand my lab work on a deeper level, and for providing me with persistent encouragement and assistance. Deininger/O’Hare Lab Michael Deininger Thomas O’Hare Kimberly Reynolds Matt Zabriskie Johanna Estrada Qian Yu Jamshid Khorashad Anna Eiring Tony Pomicter Ira Kraft William Heaton Clint Mason Tian Zhang Anna Senina Candice Ott Baron Lab Ricardo Baron Nadeem Vellore Druker Lab Chris Eide Jade Bryant Marc Loriaux Jeffrey Tyner 36 REFERENCES 1. Druker, B. J. et al. “Five-year follow-up of patients receiving imatinib for chronic myeloid leukemia.” N. Engl. J. Med. 355, 2408-2417 (2006). 2. “Response.” The National CML Society. n.p, n.d. Web. 5 March 2014 http:// www.nationalcmlsociety.org/living-cml/response 3. O’Hare, T. “Pushing the limits of targeted therapy in chronic myeloid leukemia.” Nature Rev. Cancer 12, 513-526 (2012). 4. “ General Information about Chronic Myelogenous Leukemia.” National Cancer Institute. n.p. 2 January 2014. Web. 5 March 2014 http://www.cancer.gov/cancertopics/ pdq/treatment/CML/Patient/page1 5. Besa, Emmanuel C “ Chronic Myelogenous Leukemia.” Medscape. n.p. 6 January 2014. Web. 5 march 2014 http://emedicine.medscape.com/article/199425-overview 6. “Complete Blood Count with Differential.” University of Rochester Medical Center. n.p, n.d. Web. 5 March 2014 http://www.urmc.rochester.edu/encyclopedia/ content. aspxContentTypeID= 167&ContentID=complete_blood_count_w_differential 7. Azam, M., Seeliger, M. A., Gray, N. S., Kuriyan, J. & Daley, G. Q. “Activation of tyrosine kinases by mutation of the gatekeeper threonine.” Nature Struct. Mol. Biol. 15, 1109-1118 (2008). 8. Packer LM, Rana S, Hayward R, O'Hare T, Eide CA, Rebocho A, et al. “Nilotinib and MEK inhibitors induce synthetic lethality through paradoxical activation of RAF in drugresistant chronic myeloid leukemia.” Cancer cell. 20,715-27 (2011) 9.Khorashad, J. “BCR-ABL compound mutations in tyrosine kinase inhibitor- resistant CML: frequency and clonal relationships.” Blood 121(3), 489-98 (2013) 10. Lin et al., Y.L. Lin, Y. Meng, W. Jiang, B. Roux “Explaining why Gleevec is a specific and potent inhibitor of Abl Kinase.” PNAS 110, 1664-1669 (2013) 11. WarmuthM, SungjoonK, GuX, XiaG, Adrian F. “Ba/F3 cells and their use in kinase drug discovery.” Curr Opin Oncol 19(1), 55-60 (2007) 12.DeiningerMW, Goldman JM, Melo JV. “ The molecular Biology of Chronic Myeloid Leukemia.” Blood 96, 3343-3356 (2000) 13. Apperley JF. “Part I: mechanisms of resistance to imatinib in chronic myeloid leukaemia.” Lancet Oncol 8:1018-29 (2007) 14. Cortes JE, Kim DW, Pinilla-Ibarz J, le Coutre P, Paquette R, Chuah C, et al. “A phase 2 trial of ponatinib in Philadelphia chromosome-positive leukemias.” N Engl J Me 369:1783-96 (2013) 15. Sherbenou DW, Hantschel O, Kaupe I, Willis S, Bumm T, Turaga LP, et al. “BCRABL SH3-SH2 domain mutations in chronic myeloid leukemia patients on imatinib.” Blood 116:3278-85 (2010) 16. Tokarski JS, Newitt JA, Chang CY, Cheng JD, Wittekind M, Kiefer SE, et al. “The structure of Dasatinib (BMS-354825) bound to activated ABL kinase domain elucidates its inhibitory activity against imatinib-resistant ABL mutants.” Cancer Res 66:5790-7 (2006) 17. O'Hare T, Shakespeare WC, Zhu X, Eide CA, Rivera VM, Wang F, et al. “AP24534, a pan-BCR-ABL inhibitor for chronic myeloid leukemia, potently inhibits the T315I mutant and overcomes mutation-based resistance.” Cancer Cell 16:401-12. (2009) 18. Soverini S, De Benedittis C, Machova PolakovaK, Brouckova A, Horner D, Iacono M, etal. “Unraveling the complexity oftyrosine kinase inhibitor-resistant populations by ultra-deep sequencing of the BCR-ABL kinase domain.” Blood (2013) 19. Azam M, Seeliger MA, Gray NS, Kuriyan J, Daley GQ. “Activation of tyrosine kinases by mutation of the gatekeeper threonine.” Nat Struct Mol Biol 15:1109-18 (2008) 20. Jones D, Chen SS, Jabbour E, Rios MB, Kantarjian H, Cortes J. “Uncommon BCRABL kinase domain mutations in kinase inhibitor-resistant chronic myelogenous leukemia and Ph+ acute lymphoblastic leukemia show high rates of regression, suggesting weak selective effects.” Blood 115:5428-9 (2010) 21. Talpaz M. “Imatinib induces durable hematologic and cytogenic responses in patients with accelerated phase chronic myeloid leukemia: results of phase 2 study.” Blood 99:1928-1937(2002)