Discovery and Functional Interrogation of Biomarkers Related to Therapeutic Response in Chronic Lymphocytic Leukemia
DISSERTATION
Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University
By
Cecelia R. Miller
Graduate Program in Integrated Biomedical Science Program
The Ohio State University
2016
Dissertation Committee:
John C. Byrd, MD, Advisor
Erin Hertlein, PhD, Co-advisor
Nyla Heerema, PhD
Michael Freitas, PhD
Sameek Roychowdhury, MD/PhD
Copyright by
Cecelia R. Miller
2016
Abstract
Chronic lymphocytic leukemia (CLL) is the most common type of adult leukemia in the United States. CLL is a highly heterogeneous disease; patients may progress within months or live out their lives without requiring intervention. As most patients are asymptomatic and early stage at diagnosis, biomarkers are key to predicting outcomes for these patients. Biomarkers are cellular, biochemical or molecular alterations that are measureable indicators of disease processes or pharmacological responses.
The work presented here focuses on identification of novel biomarkers for CLL.
Chapter one provides background on CLL and describes biomarkers as well as therapeutics currently in use for this disease. Chapter two describes a novel chromosomal abnormality in CLL, the jumping translocation. We found that this abnormality contributed to patients acquiring the aggressive disease markers, complex karyotype and deletion of 17p. Chapter three interrogates the ability of two chromosomal markers, gain of 2p and tetraploidy, to predict for progression on the Bruton’s tyrosine kinase inhibitor, ibrutinib. While gain of 2p showed a trend of increased occurrence in patients who later progressed, it was not a statistically significant predictor for progression. On the other hand, tetraploidy was highly enriched in patients who progressed on ibrutinib through
Richter’s transformation, a transformation to an aggressive lymphoma. Being positive for tetraploidy prior to receiving ibrutinib was significantly associated with progression
ii through Richter’s transformation. Chapter four is focused on the discovery of aberrantly expressed long noncoding RNA (lncRNA) in CLL. LncRNAs are non-protein coding transcripts greater than 200 nucleotides in length. Recent studies have elucidated transcriptional and translational regulatory roles for some of these molecules. We identified a lncRNA, named treRNA, which was associated with aggressive disease markers and was an independent predictor of progression free response on the chemotherapeutic combination of fludarabine and cyclophosphamide. We found enforced expression of treRNA in a CLL cell line resulted in decreased DNA damage when exposed to these agents. Chapter five discusses conclusions and future directions for this work.
Together, this work describes several novel biomarkers for CLL. This work adds to our ability to predict for disease aggressiveness and therapeutic response in CLL.
Furthermore, by identifying these markers we have uncovered novel avenues for interrogation of the biology of this disease.
iii
Dedication
To David…. you have my heart.
iv
Acknowledgments
First, I would like to thank my advisor, Dr. John Byrd. Dr. Byrd, thank you for opportunity to be part of your amazing team. I remember when I joined the lab you told me you would do everything in your power to see me reach my goals. You truly went above and beyond. Thank you so much for believing in me. It has been inspiring to work for someone so dedicated and positive, thank you for teaching me how to make lemonade. To my co-advisor, Dr. Erin Hertlein, thank you for all your guidance in the lab. Your feedback has been critical to my improvement as a scientist. I have learned so much working with you. Dr. Freitas and Dr. Roychowdhury, thank you for being part of my committee. You have both provided invaluable feedback, encouragement, and support. Dr. Heerema, I am so excited for what the future has in store. None of this would be possible without you. I am so grateful and honored by all the support you have given me.
To everyone in the Byrd lab, thank you for becoming like a second family to me, it has been a pleasure working with you all. To my parents and brothers, thank you for all the love and support you have given over the years. To my husband, you are the best thing to have ever happened to me. You have filled my life with joy and love. In everything I do, I know you have my back, and that provides me comfort and strength.
Thank you, my love.
v
Vita
February 22, 1985 ...... Born, Logan, West Virginia, USA
2006 ...... BS, Microbiology, The Ohio State
University
2012 ...... MS, Allied Health, The Ohio State
University
May 2012 to present ...... PhD Candidate, Biomedical Sciences
Graduate Program, The Ohio State
University
Publications
Miller C, Stephens D, Ruppert AS, Racke F, McFaddin A, Breindenbach H, Lin HJ, Waller K, Bannerman T, Jones JA, Woyach JA, Andritsos LA, Maddocks K, Zhao W, Lozanski G, Flynn JM, Grever M, Byrd JC, and Heerema NA. 2015. Jumping translocations, a novel finding in chronic lymphocytic leukaemia. British Journal of Haematology. 2015;170(2):200-7.
Balatti V, Rizzotto L, Miller C, Palamarchuk A, Fadda P, Pandolfo R, Rassenti LZ, Hertlein E, Ruppert AS, Lozanski L, Lozanski G, Kipps TJ, Byrd JC, Croce CM & Pekarsky Y. 2015. TCL1 targeting miR-3676 is codeleted with tumor protein p53 in chronic lymphocytic leukemia. Proceedings of the National Academy of Sciences. 2015;112(7):2169-74.
Miller C, Muthusamy N, Breidenbach H, Puski A, Byrd JC, Heerema N. Metaphase cytogenetics in chronic lymphocytic leukemia. Current Genetic Medicine Reports, In Review vi
Miller C, Ruppert AS, Lehman A, Blachly JS, Lucas DM, Grever M, Tallman MS, Flinn IW, Rassenti LZ, Kipps TJ, Sampath D, Byrd JC, Coombes KR & Hertlein EK. The aberrantly expressed long noncoding RNA, treRNA, predicts for aggressive disease in chronic lymphocytic leukemia. In preparation.
Fields of Study
Major Field: Biomedical Science Graduate Program
Area of Research Emphasis: Genetics
vii
Table of Contents Abstract ...... ii
Dedication ...... iv
Acknowledgments ...... v
Vita ...... vi
Publications ...... vi
Fields of Study ...... vii
Table of Contents ...... viii
List of Tables ...... xi
List of Figures ...... xiii
CHAPTER 1: Introduction ...... 1
1.1 Chronic Lymphocytic Leukemia ...... 1
1.2 Therapy in CLL ...... 2
1.3 Prognostication in CLL ...... 7
1.4 Cytogenetics in CLL ...... 10
1.5 Long noncoding RNA ...... 14
1.6 Significance and Summary of Dissertation ...... 16
viii
CHAPTER 2: Jumping translocations, a novel finding in chronic lymphocytic leukemia 22
2.1 Introduction ...... 22
2.2 Materials and Methods ...... 23
2.3 Results ...... 24
2.4 Discussion ...... 29
CHAPTER 3: Evaluating 2p gain and near-tetraploidy as biomarkers for progression on ibrutinib ...... 45
3.1 Introduction ...... 45
3.2 Materials and Methods ...... 47
3.3 Results ...... 48
3.4 Discussion ...... 50
CHAPTER 4: The long noncoding RNA, treRNA, decreases DNA damage and is associated with poor response to chemotherapy ...... 60
4.1 Introduction ...... 60
4.2 Materials and Methods ...... 62
4.3 Results ...... 68
4.4 Discussion ...... 72
CHAPTER 5: Discussion ...... 92
5.1 Summary of work ...... 92
ix
5.2 Future directions ...... 95
Appendix A: The effects of D-2-hydroxyglutarate on cells in the acute myeloid leukemia microenvironment...... 102
References ...... 133
x
List of Tables
Table 1. Rai and Binet staging systems for CLL...... 19
Table 2. Complete karyotypes of patients with jumping translocations...... 34
Table 3. Jumping translocation breakpoints for the 17p11.2 donor chromosome group. 39
Table 4. Jumping translocation breakpoints for the miscellaneous donor chromosome group...... 41
Table 5. Associations between 2p gain and demographic and molecular variables...... 52
Table 6. In a cumulative incidence analysis 2p gain was not a significant predictor for progression on ibrutinib...... 53
Table 7. Associations between tetraploidy and demographic and molecular variables. .. 54
Table 8. In a cumulative incidence analysis tetraploidy was a significant predictor for progression on ibrutinib...... 55
Table 9. Primers used for gene expression studies for lncRNA in CLL...... 75
Table 10. Associations in the CRC patient set for treRNA...... 76
Table 11. Associations in the ECOG 2997 patient set for treRNA...... 77
Table 12. Associations in the ECOG 2997 patient set for ri-treRNA...... 78
Table 13. Measured values of synthetic D2HG culture spike-ins...... 117
Table 14. D2HG measurements in cultured primary AML samples...... 118
Table 15. D2HG accumulates in AML cell lines expressing IDH mutant constructs. .. 119
xi
Table 16. D2HG is detectable in cell pellets following culture of activated T cells in supernatant from primary AML samples...... 120
Table 17. Synthetic D2HG is able to enter normal T cells when 100uM is present...... 121
Table 18. HS5 stromal cells may take up small amounts of D2HG...... 122
Table 19. Wildtype AML sample did not accumulate D2HG...... 123
xii
List of Figures
Figure 1. The B cell receptor signaling pathway...... 20
Figure 2. LncRNA mechanisms of action...... 21
Figure 3. Time to treatment and survival probabilities for CLL patients with jumping translocations...... 43
Figure 4. Partial karyotypes showing JTs in CLL...... 44
Figure 5. Cumulative incidence curves for progression with and without 2p gain...... 56
Figure 6. Cumulative incidence curves for progression with and without tetraploidy. ... 57
Figure 7. Kaplan-Meier curve of overall survival for patients with or without tetraploidy.
...... 59
Figure 8. Heatmap of CLL vs NB lncRNA microarray...... 79
Figure 9. LncRNAs are aberrantly expressed in CLL...... 80
Figure 10. Kaplan-Meier curves by treRNA expression in the CRC training set...... 81
Figure 11. CLL cells express retained intron treRNA...... 82
Figure 12. Kaplan-Meier curves by treRNA expression in the ECOG 2997 validation set.
...... 83
Figure 13. Ri-treRNA expression correlates with treRNA expression...... 84
Figure 14. TreRNA expression in OSUCLL does not alter viability, proliferation, or migration...... 85
xiii
Figure 15. TreRNA does not alter viability following treatment with fludarabine and mafosfamide...... 86
Figure 16. TreRNA induces less γH2AX in the presence of fludarabine...... 87
Figure 17. γH2AX immunoblot quantitation...... 88
Figure 18. OSUCLL expressing treRNA has less DNA damage induced with chemotherapy...... 89
Figure 19. TreRNA expression correlates with miR-155 and TCL1A...... 90
Figure 20. TreRNA is upregulated by primary CLL cells in culture and this upregulation can be abrogated by stimulation...... 91
Figure 21. IDH family members’ roles in the TCA cycle...... 124
Figure 22. D2HG is a competitive inhibitor of α-ketoglutarate dependent dioxygenases.
...... 125
Figure 23. D2HG remains stable in culture for 7 days...... 126
Figure 24. HS5 stromal cells incubated with octyl-D2HG for 7 days did not alter (a) mitochondrial activity, (b) proliferation, or (c) viability...... 127
Figure 25. D2HG does not increase ER stress in HS5 cells...... 128
Figure 26. Synthetic D2HG does not alter T cell proliferation or viability...... 129
Figure 27. T cell activation is enhanced with supernatant from AML patients...... 130
Figure 28. Synthetic D2HG does not alter T cell activation...... 131
Figure 29. PD-1 expression does not differ between IDH wildtype and mutant AML. 132
xiv
CHAPTER 1: Introduction
1.1 Chronic Lymphocytic Leukemia
Chronic lymphocytic leukemia (CLL) is the most common leukemia in the western world, accounting for 30% of all adult leukemia. In the United States, 14,620 cases were predicted to occur in 2015, with 4,650 deaths occurring due to the disease1.
The disease occurs primarily in the elderly, with a median age at diagnosis of 71 years1.
Although the median survival for CLL is 9 years, and 81.7% of patients survive 5 years1, the course of the disease is extremely variable. Some patients progress within months of diagnosis requiring treatment very early, while others live out their lives without requiring intervention.
CLL is diagnosed by the presence of morphologically mature clonal B lymphocytes accumulated to 5,000 cells per microliter of blood and persisting for more than 3 months2. CLL is characterized by the expression of B cell surface markers CD19,
CD23, and weak CD20, along with CD5 and dim surface immunoglobulin with kappa or lambda light chain restriction3. CLL cells accumulate in the blood, bone marrow, lymph nodes and spleen. These cells accumulate by resisting apoptosis through expression of anti-apoptotic BCL2 protein family members and through protection provided from the microenvironment4-7. CLL cells typically have a low proliferative index, with the
1 proliferative cells predominately localized in the lymphatic tissues, while the majority of circulating cells are arrested in G0/G18-10.
The standard systems of staging the disease are the Rai system and Binet system
(Table 1). Based on the clinical assessment of the patient they are categorized in either early (Rai 0, Binet A), intermediate (Rai I/II, Binet B) or advanced (Rai III/IV, Binet C) stage disease with median estimated survival times of greater than 10 years, 5-7 years, and 1-3 years11,12. However, clinical staging of a patient does not determine if and at what rate the disease will progress. This is particularly important as CLL is considered a
‘watch and wait’ disease; there has been no clinical benefit reported for initiating therapy prior to the onset of symptoms, therefore therapy is not given until a patient becomes symptomatic2. Symptoms seen in CLL that may indicate the need for therapeutic intervention include anemia, thrombocytopenia, massive splenomegaly or lymphadenopathy, weight loss, fatigue, fever and night sweats2. The benefits of early intervention versus the traditional ‘watch and wait’ strategy, particularly with novel kinase inhibitors in patients with poor prognostic markers, are currently being investigated in clinical trials. Due to the heterogeneous nature of the disease and the
‘watch and wait’ approach to treatment, biomarkers are an important tool to help predict when a patient may progress as well as determine the appropriate therapeutic strategy for that patient.
1.2 Therapy in CLL
2
This section presents an abbreviated summary of some of the most commonly used therapeutic options for CLL patients. Patients who are refectory to these therapies may also receive novel agents or combinations in clinical trials. An additional therapeutic approach, and the only potentially curative approach for CLL, is an allogeneic stem cell transplant.
Chemotherapy
Historically, when CLL patients became symptomatic, oral alkylating agents such as cholorambucil or cyclophosphamide were administered; these agents have continued to be utilized in CLL for many decades. Current alkylators in use today are the nitrogen mustards chlorambucil and cyclophosphamide and the bifunctional agent bendamustine.
Alkylating agents are a class of compounds that react with nucleophilic sites in DNA through one or more electron-deficient alkyl groups resulting in irreversible DNA-DNA crosslinks13. DNA crosslinking from alkylating agents results in DNA damage and breaks that can induce cell cycle arrest and apoptosis through the TP53 pathway or cause necrotic cell death14,15. DNA alkylation can be repaired through alkyltransferases or
AlkB dioxygenases and DNA damage repaired via base excision, nucleotide excision, mismatch repair, and recombinational repair16,17. As a single agent, the alkylating agent chlorambucil produced complete responses in 0-5% of CLL, with overall response rates of 30-50%18,19. Median progression free survival (PFS) with chlorambucil was around
14-17.8 months with median overall survival of 56-63.6 months18,19.
The next chemotherapeutic to influence CLL is nucleoside analogs. Fludarabine is the predominately used nucleoside analog in CLL; others that have been utilized are
3 pentostatin and cladribine20,21. Fludarabine is a purine analog, it must be metabolized and converted to its active form, F-ara-ATP, to have cytotoxic activity22. The mechanisms of action for F-ara-ATP are many. It suppresses DNA synthesis by inhibiting DNA polymerase or DNA primase22. F-ara-ATP also inhibits ribonucleotide reductase, which converts dADP to dATP, thus diminishing the pool of nucleotide available22. F-ara-ATP can be incorporated into DNA where it inhibits DNA ligase 1 from ligating single strands22. Outside of interfering with DNA synthesis and repair, F-ara-ATP can also induce the apoptotic cascade23. As a single agent, fludarabine achieved CR in 3-38% of patients with an OR between 32-80%24-27. Fludarabine alone did not improve long term outcome for patients compared to alkylating agents. Median PFS with fludarabine alone was between 18.7-20 months with median overall survival of 45.9-88 months18,19,28.
Combination chemotherapy significantly improved response and PFS. The combination of alkylating agents with nucleoside analogs inhibits DNA repair of alkylator damage through incorporation of the nucleoside analogs in the repair patch17.
Combining cyclophosphamide with fludarabine resulted in CR and OR between 22-46% and 74.3-94%, respectively25,29,30. Median PFS was between 27-48.1 months26,28-30.
However, this combination did not result in improved OS; median OS was between 55-
79.1 months 26,28-30. Most patients ultimately relapse on chemotherapy.
Monoclonal antibodies
The next major advance for CLL therapeutics was the introduction of monoclonal antibodies, predominately targeted against CD20, a surface marker expressed on normal
B as well as malignant B cells31. Binding of rituximab, a chimeric anti-CD20 antibody,
4 can recruit effector cells as well as complement to result in killing of tumor cells31.
Rituximab as a single agent did not improve response rate over chemotherapy32, but resulted in striking improvements when combined with fludarabine and cyclophosphamide (FCR). FCR produced CR in 44-72% of patients and an OR of 90-
95%29,33. Median PFS was between 51.8-80 months29,33. Six year OS was 77%, with a
33% reduced risk of death compared to FC29,33. Today, the chemo-immunotherapy FCR is considered the standard of care for fit younger patients34,35. Second generation monoclonal anti-CD20 antibodies used in CLL include obinutuzumab and ofatumumab36.
Alemtuzumab, an anti-CD52 antibody, is also used in CLL, primarily in patients with aggressive disease characteristics37.
Tyrosine kinase inhibitors
In the last several years therapeutic strategies for CLL have begun a paradigm shift. Small molecule inhibitors targeting the B cell receptor (BCR) pathway (Figure 1) have shown significant response rates as single agents. The BCR is a multiprotein subunit consisting of surface immunoglobulin and a signaling subunit of CD79a and CD79b38,39.
Simply put, upon BCR engagement CD79a and CD79b immunoreceptor tyrosine-based activation motifs (ITAMs) are phosphorylated by src-family kinases, such as the tyrosine- protein kinase Lyn (LYN)38,39. Doubly phosphorylated ITAMs recruit and activate the tyrosine-protein kinase Syk (SYK) 38,39. SYK is then able to phosphorylate Bruton’s tyrosine kinase (BTK), then it and SYK phosphorylates phospholipase C gamma 2
(PLCG2). BCR stimulation also activates phosphatidylinositide 3-kinase (PI3K) which
38,40 phosphorylates PI(4,5)P2 to PI(3,4,5)P3 . This step is critical for the recruitment of
5
BTK to the signalosome39. The signaling chain continues to cascade resulting in activation of AKT and ERK1/238,41. These pathways drive transcription factors such as
NFKB, NFAT, AP1, and MYC resulting in the upregulation of genes involved in cell survival and proliferation42,43.
BCR signaling is dysregulated in CLL. CLL cells, particularly those from lymph nodes and bone marrow, have elevated expression of BCR genes, including BTK, LYN and SYK8,44. LYN, PI3K, AKT, and ERK are often constitutively active in CLL cells45.
Recent therapeutics exploits the dependence of CLL cells on this signaling. Ibrutinib is a first in class BTK inhibitor. It irreversibly targets the cysteine 481 residue in the ATP binding pocket of BTK, thus abrogating BCR signaling for the cell46. In treatment-naïve and previously treated patients, ibrutinib has achieved CR in 23 and 7% and OR in 84 and 90%, respectively47. Importantly, with almost three year follow-up, median PFS has not been reached for either group47. Ibrutinib is now the recommended therapy for relapsed/refractory and high risk deletion 17p CLL patients34,35. A second generation
BTK inhibitor currently in clinical trials, acalabrutinib, reduces off target effects and early results show similar response rates as ibrutinib48.
The second kinase inhibitor affecting the BCR pathway that is approved for use in
CLL is idelalisib. This molecule reversibly inhibits the PI3K delta isoform, a predominately B cell isoform49. Idelalisib combined with rituximab in untreated elderly patients had 19% CR and 97% OR rates. With this combination median PFS was not reached with three year follow-up50. Idelalisib as a single agent in relapsed/refractory
CLL had an OR rate of 33% with media PFS of 16 months51.
6
1.3 Prognostication in CLL
In addition to the staging systems discussed previously, there are a number of additional patient demographics and clinical factors that predict prognosis in CLL including age, gender, and performance status. Older age and male gender have been associated with poorer outcome52. In addition, older age is also an indicator for therapeutic response; older individuals had decreased OS with FCR53. Performance status, which assesses if a patient is able to care for his or herself or if they require various levels of assistance, has been reported to be an independent predictor of PFS54.
There are several laboratory diagnostics that have shown associations with disease progression. Lymphocyte doubling time (LDT) is determined by calculating the number of months it takes for a patient’s absolute lymphocyte count to double. A LDT of 12 months or less has been associated with poor prognosis55. LDT has confounding issues; additional factors other than CLL may affect the absolute lymphocyte count56. B-2-
Microglobulin (B2M) is a component of the major histocompatibility complex 1 which is present on all nucleated cells. Serum levels of this molecule correlate with tumor burden in CLL57. Elevated B2M has been shown to be an independent predictor of poor OS in retrospective studies58,59 as well as a prospective study60. However, another prospective evaluation of this marker did not find significance61, thus the usefulness of B2M as a prognostic marker is unclear.
The mutational status of the immunoglobulin variable heavy chain (IGHV) is a stable prognostic marker in CLL; a patient’s IGHV mutational status does not change over time or with treatment56,62. Patients are classified as IGHV mutated if the IGHV gene
7 within their CLL cells has greater than 2% difference in nucleotide sequence when aligned to germline IGHV. Patients are classified as IGHV unmutated with a 2% or less sequence difference63,64. Unmutated IGHV occurs in approximately 30-40% of patients and is associated with shorter mean survival times64-67. Unmutated IGHV was confirmed by several groups as independently associated with poor survival66-68. Interestingly, this biomarker is suggestive to two biologically distinct forms of CLL. Mutated IGHV suggests that the CLL clone may have arisen from a post-germinal antigen experienced memory B cell, while unmutated IGHV may arise from a pre-germinal center naïve B cell or a germinal center-independent memory B cell69,70. However, the cell of origin for CLL remains an issue of debate70. Measuring IGHV mutational status has not been widely adopted clinically due to the need to sequence patients’ samples, which is costly and technically challenging56. Surrogate markers for IGHV have been pursued.
CD38 is a type II transmembrane molecule normally expressed by B lineage progenitors, B cells in the germinal center, and terminally differentiated plasma cells62.
CD38 expression is used to delineate two subgroups of CLL patients based on their percentage of CD38 positive leukemic cells. Surface expression is generally measured using flow cytometry. Typically the threshold is set at ≥30% CD38+ clonal members to be considered positive, however, this threshold is not universally agreed upon56,62. CD38 expression in CLL is associated with an increased response to BCR stimulation and enhanced migration towards the chemokine CXCL1271,72. CD38 positivity has been associated with disease progression, poor response to therapy, and decreased OS56,73,74.
CD38 expression is associated with unmutated IGHV, but is not a robust surrogate for
8
IGHV mutational status56,62. An additional consideration for using CD38 as a prognostic marker is that its expression may vary over time75.
70-kD zeta-associated protein (ZAP70) was identified from a gene expression screen as the most specific gene which separated IGHV mutated and unmutated patients; the majority of unmutated patients were ZAP70 positive and mutated patients ZAP70 negative76. ZAP70 expression is low to absent in normal B cells, instead being expressed in T cells and NK cells77. ZAP70 is a protein tyrosine kinase homologous to SYK in B cells78. CLL patients expressing ZAP70 have enhanced BCR signaling79. Stimulation with the chemokine CXCL12 in ZAP70 positive CLL results in enhanced migration, greater intracellular calcium flux, and prolonged stimulation of ERK80,81. Analysis of
ZAP70 has been performed using multiple methods including flow cytometry, immunohistochemistry, quantitative reverse transcription polymerase chain reaction
(qRT-PCR), immunoblotting, and promoter methylation. Accurate quantification using qRT-PCR and immunoblotting can be hindered by contaminating T cells which express high levels of ZAP70. Immunohistochemistry is unable to provide precise quantification78. ZAP70 has also been assessed using flow cytometry, this method requires the cells to be fixed and permeabilized to allow for intracellular staining. Using additional markers, ZAP70 can be analyzed within the CD19/CD5+ population of B cells; however, ZAP70 is often dimly expressed thus complicating this approach78. Using flow cytometry patients are typically classified as positive when ≥20% of the leukemia cells express ZAP7082. DNA methylation in ZAP70 regulatory regions has been shown to associate with ZAP70 expression83. Additionally, measuring methylation at a specific
9 single CpG dinucleotide in the 5’ regulatory region appears sufficient to stratify patients; patients methylated in this region express low ZAP70 while patients who are unmethylated have high expression84. High ZAP70 expression has been associated with poor PFS and OS, however, there is a typically a strong correlation between ZAP70 and
IGHV mutational status, thus the marker is often not an independent prognostic factor78.
In addition to the prognostic markers described above, cytogenetics studies in
CLL have identified important biomarkers predictive of progressive disease as well as therapeutic response. These markers will be discussed in detail in the following subsection.
1.4 Cytogenetics in CLL
Fluorescence in situ hybridization (FISH)
Fluorescence in situ hybridization is predominately used to detect cytogenetic abnormalities in CLL because it does not rely upon dividing cells. FISH uses DNA probes with incorporated fluorophores with sequences complementary to the DNA region of interest. The labeled probe and the target DNA are denatured and combined allowing the single stranded probe to hybridize to the target region of single stranded DNA typically in either interphase nuclei or on metaphase chromosomes. Signals produced from the fluorophores can be visualized with fluorescence microscopy. Single colored
FISH probes can be used to count signals for gains or losses. There are also break-apart as well as dual color fusion probes which allow you to detect specific chromosomal rearrangements.
10
Approximately 80% of CLL patients have a genomic aberration which can be detected by FISH on interphase cell nuclei with the disease specific probe set for target regions on 13q, 11q, 12 centromere and 17p85. Using interphase cytogenetics, Döhner et al. found an occurrence of 55% for deletion 13q, 18% for deletion 11q, 16% for trisomy
12, and 7% for deletion 17p85. The 17p deletion, 11q deletion, trisomy 12, normal FISH and 13q deletion had respective median survival times of 32, 79, 114, 111, and 133 months85. The Döhner hierarchy was recently validated as well as extended by the CLL
Research Consortium (CRC). They found that patients with deletions of 13q or 17p could be further stratified based of the percentage of cells positive for either abnormality.
Patients with >85% of cells with a deletion of 13q had shorter time to first treatment
(TTFT) compared to patients with lower frequencies of 13q deletion, and patients with
<20% of cells with a deletion of 17p had longer TTFT compared to patients with a higher percentage of cells positive for del(17p)86.
These cytogenetic abnormalities have important biological implications for the cells. Loss of 17p results in the loss of the tumor suppressor gene TP53, which is a transcription factor activated upon cellular stress such as DNA damage leading to cell cycle arrest or apoptosis87. Loss of TP53 allows damaged cells to survive, thus contributing to resistance to chemotherapies87. Deletion of 11q results in the loss of
ATM Serine/Threonine Kinase (ATM). This kinase is involved in recognizing DNA damage and phosphorylating downstream mediators of DNA damage response and cell cycle control, including TP5388. Deletion of 13q results in the loss of the microRNA
(miR) tumor suppressors miR-15a and miR-16. MicroRNAs are small noncoding RNA
11 that can bind to protein coding transcripts to initiate degradation or prevent translation89. miR-15a and miR-16 target the anti-apoptotic protein BCL2, which is overexpressed in
CLL90.
Metaphase karyotyping
Interphase FISH provides valuable prognostic information, however, this technique is limited to the regions targeted by the probes. To evaluate genetic changes across all chromosomes metaphase karyotyping is performed. Metaphase karyotyping involving culturing cells and arresting them during metaphase, when the chromatin is most compacted. This allows the staining of the chromosomes to produce sequential light and dark banding patterns that can be used to identify and differentiate the chromosomes with microscopy. Metaphase karyotyping is labor intensive; typically only 20 metaphases are examined for a patient.
In the past, metaphase karyotyping had limited utility in CLL due to the lack of dividing cells in culture. CLL cells respond weakly to traditionally used B-cell mitogens
(pokeweed mitogen (PWM) and phorbol 12-myristate 13-acetate (PMA)) and the detection of abnormal clones occurred in approximately 30-50% of cases91,92. More recently, cytosine-phosphate-guanosine oligodeoxynucleotides (CpG ODNs) have been shown to dramatically increase the detection of abnormalities in CLL cells as well as the percentages of abnormal cells93-95. CpG ODNs are short (usually 19-25 bases in length) strands of DNA containing unmethylated CpG dinucleotides in particular sequence motifs. These CpG ODNs can be natural (bacterial DNA) or synthetic and are recognized by immune cells as pathogen-associated molecular patterns via Toll-like receptor 9 96,97.
12
Stimulation of normal and CLL B-cells with CpG ODNs results cell proliferation, cytokine production, and regulation of surface molecules96. The CpG ODN used to stimulate primary CLL cells at our institution is ODN-685. This is a 21-mer fully phosphorothioated unmethylated CpG ODN with a sequence of: 5’-TCG TCG ACG TCG
TTC GTT CTC 3’ 98. The CpG ODN DSP30 in combination with the cytokine Il-2 is another commonly used stimulus. CpG ODNs have been shown not to cause the increased abnormalities observed with their use94,99,100.
Metaphase karyotyping is necessary to identify an important biomarker in CLL, the complex karyotype (≥ 3 chromosome abnormalities). Complex karyotype is an independent predictor of adverse prognosis101-103, and is an independent predictor of shorter progression-free survival (PFS) and overall survival (OS) for chemo- immunotherapies104,105. In CLL patients treated with flavopiridol (a cyclin-dependent kinase inhibitor) complex karyotype moderately impacted OS106. Additionally, complex karyotype predicted an inferior outcome following reduced-intensity conditioning allogeneic stem cell transplant for CLL107. Importantly, in CLL patients treated with ibrutinib, complex karyotype was independently associated with progression108, and was confirmed a stronger predictor than deletion (del) 17p for an inferior outcome in relapsed or refractory CLL patients treated with ibrutinib‐based regimens109.
Metaphase karyotyping also offers the opportunity to discover additional novel recurrent abnormalities in CLL. Interrogation of the prognostic significance of novel abnormalities as well as their biologic implications can further our understanding of this disease.
13
1.5 Long noncoding RNA
A significant portion of the genome is transcribed, however, only approximately
2% of this transcription represents protein coding genes110. One major subset of non- protein coding RNA being transcribed is long noncoding RNA (lncRNA) 111.
Traditionally lncRNAs have been viewed as transcriptional noise; however, recent discoveries have begun to change this perception. LncRNA is defined as a non-protein coding RNA greater than 200 nucleotides in length112. LncRNAs are classified by their location relative to nearby coding genes112. For example, lncRNAs can be defined as intergenic, antisense, or intronic lncRNAs. Many lncRNAs are processed like mRNA; transcribed by RNA polymerase II, capped, spliced, and polyadenylated113. LncRNAs show clear evolutionary conservation particularly in exonic and promoter regions; however, they are considerably less conserved than protein coding genes and conservation is variable amongst lncRNA114,115. LncRNA have been identified by the use of chromatin signatures. Genomic regions that are actively transcribed by RNA polymerase II have trimethylation of lysine 4 on histone H3 at their promoter and trimethylation of lysine 36 on histone H3 along the length of the transcribed region, this signature is called the K4-K36 domain114. Using ChIP-seq, Guttman et al. sequenced regions bearing this signature then filtered out all the known protein coding genes to identify putative lncRNA114. A second methodology used to identify lncRNA is RNA- sequencing (RNA-seq). RNA-seq is the high throughput sequencing of cDNA. Using this approach Cabili et al. was able to identify 8,000 putative lncRNA transcripts by filtering out reads mapping to known protein coding regions116.
14
There are primarily two in silico methods used to assess coding potential of a transcript. LncRNA coding potential can be queried by translating the transcript in each possible frame and then comparing these results to known protein families112. The second method used is the codon substitution frequency metric. This approach compares genomic sequences across species to assess the evolutionary conservation of sequences.
Since protein coding genes are highly conserved, and lncRNA are less so, the two can be differentiated117. The non-coding potential of the transcript can be confirmed experimentally through in vitro translation118,119.
Thus far, there are four major functional categories described for lncRNA (Figure
2). LncRNA can act as decoys by binding miRNA, mRNA, or proteins to inhibit their normal function120-122. LncRNA can act as guides associating with DNA and proteins to bring these molecules in close proximity. Importantly, many lncRNAs have been shown to associate with chromatin modifying proteins, suggesting they may be involved in targeting these complexes to particular regions of DNA123-125. LncRNA may serve as scaffolds bringing multiple protein complexes in close proximity to each other126.
LncRNA have also been reported to act as enhancers, increasing the transcription of near- by coding regions119,127. A single lncRNA may perform multiple functions112,127. This wide range of functions was identified across only a small fraction lncRNAs. Overall, this class of RNA is presently functionally poorly understood, with less than 1% well characterized128.
Dysregulated expression of lncRNAs has been documented in a number of cancers112. Many aberrantly expressed lncRNA have recently been discovered as
15 potential biomarkers in various cancers and has stimulated the functional investigation of their role in these diseases. Elevated expression of the lncRNA HOTAIR can predict for poor PFS, OS, and metastasis in several solid tumors independent of other variables129,130.The dysregulated expression of HOTAIR has been shown to have functional implications in several types of cancer129,131,132. In vitro overexpression of
HOTAIR in breast cancer cell lines increased migration and invasion through an extracellular matrix129. HOTAIR overexpression in epithelial cancer cells drove their gene signature towards a phenotype resembling embryonic fibroblasts129. These changes were driven by HOTAIR’s ability to alter the localization of the chromatin modifying complex, Polycomb repressive complex 2133,134. A second example is the lncRNA
SChLAP1. SChLAP1 was identified as an independent predictor of recurrence, progression, and mortality in prostate cancer118. In vitro studies modulating SChLAP1 expression in prostate cell lines implicated the lncRNA in an invasion phenotype; furthermore, SChLAP1 was demonstrated to alter gene expression through antagonizing the SWI/SNF chromatin remodeling complex118. These studies support that lncRNA can serve as robust biomarkers in cancer and interrogating their function sheds light on biological processes driving cancer aggressiveness. Importantly, little is currently known about the role of lncRNA in CLL.
1.6 Significance and Summary of Dissertation
Due to the vast disease heterogeneity seen across CLL patients, biomarkers play an important role in this disease. In particular, identification of markers that can predict
16 therapeutic response remains of great interest. FCR is the gold standard for therapy in the young fit CLL patient population, however, we know patients with deletion of 17p typically respond poorly to this regimen and relapse quickly34. Detection of this marker can inform clinicians that these patients may benefit from alternative therapeutic approaches, such as ibrutinib. While deletion 17p is a strong predictor for FCR response, many patients without this marker do not achieve long-term PFS 33. Identification of additional markers which predict for poor response would be beneficial.
Ibrutinib has shown remarkable efficacy in CLL and is currently the recommended therapy for unfit older patients or those with del(17p)34,35. While the majority of patients have durable responses to ibrutinib, there remains of subset of patients who are progressing on this therapy, either with CLL progression or Richter’s transformation108. Markers that can identify patients at risk for progression are needed.
In response to these unmet needs, the work described in this dissertation focuses on identification of biomarkers related to therapeutic response as well as interrogation of the biological implications for several of the aberrant markers identified in CLL. Chapter
2 describes a novel chromosomal abnormality in CLL, the jumping translocation, which can contribute to the development of the poor prognostic markers del(17p) and complex karyotype. In chapter 3, two chromosomal markers are evaluated for their ability to predict progression on ibrutinib. Chapter 4 describes a lncRNA predictive of poor response to fludarabine plus cyclophosphamide and provides evidence that the lncRNA may have a role in reducing the DNA damage induced by chemotherapeutic agents.
Finally, chapter 5 presents the conclusions of this work and discusses future directions for
17 further interrogation of these markers. The significance of this work is to identify biomarkers which can improve the risk stratification of CLL patients as well as uncover novel biological processes in CLL for further investigation.
18
Rai staging Binet staging Stage 0 I II III IV A B C Lymphocytosis + + + + + Lymphadenopathy + +/- +/- +/- Enlarged spleen + +/- +/- and/or liver Anemia + +/- +* Thrombocytopenia + +* <3 lymph node + involvement ≥3 lymph node + involvement
Table 1. Rai and Binet staging systems for CLL. Clinical parameters that define the stages of Rai and Binet staging11,12. + indicates presence, +/- indicates that the parameter may be present or absent, +* indicates either one or both parameters are present.
19
Figure 1. The B cell receptor signaling pathway. This figure shows key components of BCR signaling pathway. The BCR signaling cascade activates downstream effectors of survival and proliferation. Figure taken from Dal Porto, 200438.
20
Figure 2. LncRNA mechanisms of action. LncRNAs have been described to have several functional mechanisms. These functions can be broadly grouped into four categories; decoys, scaffolding, guides, and enhancers. Figure taken from Rinn & Chang, 2012112.
21
CHAPTER 2: Jumping translocations, a novel finding in chronic lymphocytic leukemia
2.1 Introduction
A jumping translocation (JT) is a rare cytogenetic aberration defined by translocations involving the same fragment of chromosome, called the donor chromosome, and two or more different recipient chromosomes in different cells in the same patient 135. In previous reports, donor chromosome breakpoints predominately occur in centromeric and heterochromatic regions, while recipient chromosome breakpoints frequently occur in telomeric and subtelomeric regions. The most commonly detected donor chromosome segment is 1q 136. Other donor chromosomes have been described, including chromosomes 3, 7, 9, 11, and 17 137-141. The recipient chromosome appears to be random136.
JTs have been associated with aggressive disease and poor prognosis in various hematologic malignancies 137,142,143. They occur most frequently in multiple myeloma followed by acute lymphoblastic leukemia and acute myeloid leukemia 136. To our knowledge, only a single mention of this phenomenon occurring in a CLL patient has been reported 144.
In this study, we used metaphase chromosomal analysis to establish JTs as a type of recurrent chromosomal abnormality occurring in CLL. Here we present 26 CLL patients with karyotypes containing this abnormality.
22
2.2 Materials and Methods
Patient sample and data collection
Records of 878 patients seen at The Ohio State University (OSU) for a diagnosis of CLL by International Workshop on CLL (IWCLL) criteria2, who had submitted cytogenetic material between October 2007 and July 2011, were reviewed for karyotypes containing JTs. Each confirmed case with a JT during this time period was then retrospectively reviewed for clinicopathological data from the time of diagnosis through
May 2013. This study was conducted under an OSU Institutional Review-board approved protocol in accordance with the Declaration of Helsinki.
Conventional cytogenetic analyses
Cytogenetics was performed as previously described145. Briefly, cells from bone marrow or peripheral blood (2.0 X106 cells/mL) were incubated in RPMI 1640 medium
(Fischer Scientific, Houston, TX) with 2% L-Glutamine (Gibco Invitrogen, Carlsbad,
CA), supplemented with 20% fetal bovine serum (Hyclone Laboratories, Logan, TX), and 2% penicillin and streptomycin (Gibco Invitrogen). The mitogens used included: pokeweed mitogen (PWM, 10μg/mL, Sigma Aldrich, St. Louis, MO), phorbol 12- myristate 13-acetate (PMA, 40ng/mL, Sigma Aldrich) and CpG ODN 685 (20 μg/mL, synthesized by Sigma Aldrich). The mitogens were added to the cultures, and the cells were incubated for 72 hours under standard laboratory conditions. All samples were harvested, fixed, G-banded using trypsin, and stained with Wright stain according to standard laboratory procedures. Twenty metaphases were completely analyzed whenever
23 possible. Karyotypes were described following the ISCN 2009 standard146 except single cell abnormalities were reported when they demonstrated a JT.
Fluorescence in situ hybridization (FISH)
FISH analysis was also performed on the mitogen-stimulated cultures. A FISH panel of probes for CLL (data not shown) was analyzed which included TP53 (17p13.1)
(Abbott Molecular, Des Plains, IL). Hybridization was according to the manufacturer’s directions. Two hundred cells per probe were analyzed, 100 by each of two independent observers. Each case was compared for consistency of FISH results with conventional karyotyping results. Additionally, FISH using centromere probes (Abbott Molecular) was performed on metaphases in a subset of patients to confirm dicentric chromosomes.
Statistical Methods
Characteristics of patients who had been identified with JT are described with frequencies and proportions for categorical variables and with medians and ranges for continuous variables. Time to treatment (TTT) was calculated from the date of diagnosis until the date of first treatment, censoring one patient who had not yet started treatment at last follow-up. Overall survival (OS) was calculated from the time of diagnosis until the date of death, censoring patients who were alive at last follow-up. TTT and OS estimates were calculated by the Kaplan-Meier method147.
2.3 Results
Clinical characteristics
24
A total of 26 patients with JTs were identified among 878 CLL patients (3.0%,
95% CI: 1.9% to 4.3%) seen over a period of 3.75 years. Clinically, this group of 26 patients (58% male) exhibited an aggressive disease course, with high-risk cytogenetics developing in all patients.
At diagnosis of CLL in this patient cohort, the median age was 54 years (range:
37-77), with approximately 90% of the patients with Rai Stage 2 or less148. Of the 16 patients with IGHV data available, 11 (69%) had IGHV unmutated disease. The estimated median TTT was 11.5 months (95% CI: 3.0 to 25.6 months) (Figure 3), and all but one patient with 28 months of follow-up had begun treatment. The one patient without treatment is male, diagnosed at age 60 with a Rai Stage of 0, but has since been re-staged to Rai stage 2 and has been identified as also harboring in addition to a JT, del(17p) with a complex karyotype and unmutated IGHV disease. With a median follow-up of 83 months for the JT patients, there have been 13 deaths; the estimated OS at 6 years from
CLL diagnosis is 66% (95% CI: 0.42 to 0.82) (Figure 3).
At the time of JT detection, the median age was 61 years (range: 39-78) and 73% had a Rai stage of 3 or 4. In patients where CD38 and beta-2 microglobulin (B2M) data were available (n=23 and n=20, respectively), 39% of patients were positive for CD38 expression (defined as >20%) and 35% of patients had elevated B2M (≥4 mcg/mL).
Cytogenetically, 23 of the 26 patients (88%) had del(17p). Del(17p) was previously seen in 6 of 26 patients prior to detection of JT; for these patients loss of 17p occurred from several years before the JT to less than a year before the JT developed. All patients eventually developed complex karyotype, defined as at least 3 independent aberrations,
25 with 24 of 26 (92%) having at least 5 independent aberrations (Table 2). In our institution’s overall CLL population at the time of this study del(17p) occurred in 21%
(181/878) of patients. JT patients comprised 13% (23/181) of the del(17p) population.
Complex karyotype was found in 33% (289/878) of patients; JT patients were 9%
(26/289) of this population.
Data regarding JTs at the time of diagnosis was limited for these cases.
Karyotypes were available for only 4 patients within 3 months of diagnosis; of these, JTs were detected in 3 patients. One of these patients had not received any treatment for CLL and did not begin treatment for more than a year following identification of JT. The second patient had received one cycle of fludarabine prior to the identification of JT. The third patient previously received bortezomib and rituximab for an incorrect diagnosis of
Waldenstrom’s disease at the time of CLL diagnosis and prior to JT identification. Not including the patient treated for Waldenstrom’s disease, there were a total of 5 patients
(19%) with a JT in samples obtained prior to any CLL therapy. In the remaining 20 patients who had been treated prior to discovery of JT, the median number of therapies was 3, with a range of 1 to 7.
Following detection of JT, all but two patients have received some form of treatment. Response to first treatment following identification of JT included 5 patients with complete remissions; one with allogenic stem cell transplant, one with rituximab, doxorubicin, etoposide, vincristine, cyclophosphamide and prednisone (REPOCH), one with fludarabine, cyclophosphamide and rituximab (FCR), one on clinical trial with cyclophosphamide, flavopiridol and rituximab, and one on clinical trial with flavopiridol.
26
Of the remaining 19 patients who received subsequent treatment; 9 had partial response,
3 had stable disease, and 7 had progressive disease. Four patients progressed to Richter’s disease, all on or after the sample date when the JT was identified.
Jumping translocation characteristics
Among the 26 cases, there were a total of 97 JTs. Examples of JTs in CLL patients are shown in Figure 4, complete karyotypes are provided in Table 2. Twenty- three patients had unbalanced (loss of chromosomal material) JTs, and two patients (Case
6 and 16) had both balanced (no loss of chromosomal material) and unbalanced JTs. One patient (Case 2) exhibited a balanced JT as the sole abnormality, which is considered a very rare occurrence140,149. Serial cytogenetic samples were available for 22 of the 26 patients. In these serial samples JTs were detected in the first cytogenetic analysis performed at our institution in 12 patients. JTs were absent in the initial cytogenetic analysis in 10 patients; 6 of these patients initially had normal karyotypes, one patient had a single cytogenetic abnormality, and 3 patients had complex karyotypes.
Among the 26 patient samples, a total of 33 donor chromosomes were identified.
For analysis purposes we separated the JTs based on the localization of the donor chromosome breakpoints. Strikingly, 16 (48%) of the donor breakpoints occurred in the centromeric region 17p11.2 (Table 3), whereas the other 17 occurred in at least 11 different chromosomes and have been classified as miscellaneous donor breakpoints
(Table 4). In this miscellaneous group, two repeat donor breakpoints were identified, one involving 4q12 and the other involving 18p11.2. Additional donor breakpoints were found at 1p32, 8p21, 9q12, 11q21, 12p11.2, 13p11.2, 13q14, 13q21, 14p11.2, 15p11.2,
27
19p13.3; two breakpoints were of unidentified chromatin. Thus, of the 17 miscellaneous donor chromosomes, 9 breakpoints (53%) occurred in centromeric regions, 6 (35%) occurred between the pericentromeric and subtelomeric regions, and 2 (12%) were not identified. Across both donor breakpoint groups, seven patients, (Cases 3, 7, 8, 17, 18,
23, and 25) had two unique donor chromosomes involved in JTs at the same time (Case 3 shown in Figure 4). Four of these seven patients with two unique donor chromosomes had a chromosome acting as both a donor and recipient in the JTs.
Among the group of 16 donor chromosomes with breakpoints at 17p11.2, there were a total of 51 recipient chromosomes (Table 3). The majority of recipient breakpoint locations occurred in centromeric regions (n=35, 69%), with fewer occurring, respectively, between the pericentromeric and subtelomeric regions (n=8, 16%) and within the subtelomeric region (n=2, 4%); six (12%) of the recipient breakpoints were unknown. These JTs resulted in the formation of 36 dicentric chromosomes and 2 pseudo-dicentric chromosomes, most with recipient breakpoints near centromeres.
Dicentric chromosomes were identified by the presence of two distinct constrictions
(centromeres); while pseudo-dicentric chromosomes had one primary constriction and the other centromere was inactive. Multiple repeat recipient breakpoint locations, all from centromeric regions, were identified in this group and included 18p11.2 (7), 8p11.2 (3),
14p11.2 (3), 21p11.2 (3), 17p11.2 (2), 6p11.2 (2), 8q11.2 (2), 12p11.2 (2), and 20p11.2
(2).
Among the group of 17 miscellaneous donor chromosomes, a total of 50 recipient chromosomes were identified (Table 4). In contrast with what was found when 17p11.2
28 was the donor breakpoint, the recipient breakpoint locations associated with non-17p11.2 donors were more often found between the pericentromeric and subtelomeric regions or within the subtelomeric region, but less often in centromeric regions (p<0.0001, Fisher’s exact test). Specifically, only 7 breakpoints (14%) involved centromeric regions, whereas
19 (38%) and 18 (36%) were involved between the pericentromeric and subtelomeric regions or within the subtelomeric region, respectively; six (12%) of the recipient breakpoints were unknown. There were only 8 unique dicentric or pseudodicentric chromosomes. Repeat recipient breakpoint locations in the miscellaneous group were found at chromosome bands 5q35 (2), 12q24.1 (2), 17p11.2 (2), and 18q23 (2).
2.4 Discussion
Here we describe the first series of JTs occurring in CLL and provide evidence that this is a recurring event in this disease. We present 26 cases from a population of
878 patients seen at our institution. This subset represents approximately 3% of our patient population which is comparable to the frequency of this abnormality seen in multiple myeloma, the most commonly reported hematologic malignancy with JTs 136,150.
However, the prevalence of this abnormality in our patient population may be slightly underestimated due to the difficulties in detecting JTs that appear as rare nonclonal abnormalities. Our patient population is indicative of a large regional quaternary care center, and patients tend to be at a later stage of their disease at presentation to our institution than the average CLL patient. The frequency of del(17p) in our patient population during the course of this study was 21% and complex karyotype was 33%. As
29 most patients present to outside facilities at their initial CLL diagnosis, we have limited data on the patient’s initial karyotype and frequently only FISH analysis was previously performed. Due to this, we are unable to determine the timing of JT development in 13 patients. For the other 13 patients, JTs were detected in 3 of the 4 patients with karyotypes performed within 3 months of diagnosis. In 10 patients for whom serial cytogenetic samples were available, including the one patient without a JT at diagnosis,
JTs did not occur until later in the course of their disease. Our ability to identify multiple
JTs is likely due to enhanced detection of chromosomal abnormalities by conventional chromosome analysis with the use of CpG ODN to increase mitotic activity of CLL cells in culture 99,145. Using only FISH analysis, in place of metaphase cytogenetics in CLL, these abnormalities and their associated karyotypic complexity cannot be detected.
The majority of patients with CLL who were known to have a JT at some point during the course of their disease appeared to have an aggressive disease course. Twenty- five of the 26 patients had begun treatment, with an estimated TTT of 11.5 months.
Twenty-three patients (88%) showed loss of TP53 by either FISH or karyotype analysis during the course of their disease. This loss occurred prior to or concurrently with the JT.
TP53 is involved in arresting the cell cycle or inducing apoptosis in damaged cells151.
Loss of TP53 in CLL patients is associated with increased drug resistance and shortened
PFS and OS 152-155. This loss may also lead to genomic instability causing patients to develop a greater number of chromosome abnormalities compared to cases with normal
17p 156. JTs occurred in 3 patients with no loss of TP53; however, the TP53 mutational status of these patients is not known. Loss of TP53 may provide a permissive
30 environment for JTs to develop. Alternatively, additional factors not yet understood may also contribute.
While data regarding JT breakpoints across all hematologic malignancies are highly heterogeneous, certain trends have emerged. Breakpoints of donor chromosomes most often occur in centromeric regions and heterochromatic regions, with a preference towards 1q. JTs involving 1q often result in partial trisomy. No recipient chromosome has been reported to have preferential involvement in these translocations. However, JT recipient breakpoints have been found most often in telomeric regions 136. JTs in CLL contrasted with these trends in a number of ways. While the majority of donor breakpoints were centromeric (76%), heterochromatic breakpoints were rare. We found no donor breakpoints at 1q; instead repeat donor breakpoints were seen most commonly at 17p11.2 followed by 18p11.2 and 4q12. These translocations frequently resulted in partial monosomy. Across all donor and recipient chromosomes in CLL JTs, 16 repeat breakpoints were seen. Multiple repeat recipient breakpoints particularly within the
17p11.2 donor group suggest that, contrary to other hematologic malignancies136, recipient breakpoint locations in CLL JTs may not be entirely random nor do they favor telomeric breakpoints. Twelve of the repeat breakpoint locations occurred in chromatin directly adjacent to centromeres. The most frequent recipient breakpoint, 18p11.2, occurred 8 times. Recipient breakpoint locations in CLL were significantly different
(p<0.0001) when comparing the 17p11.2 donor group (Table 3) to the miscellaneous donor group (Table 4). When associating with the 17p11.2 donor breakpoint, recipient breakpoints most often occurred in centromeric regions and very rarely in telomeric
31 regions. Recipient breakpoints associated with the other donor chromosomes occurred between the pericentromeric and subtelomeric regions and in subtelomeric regions at similar rates, while centromeric breakpoints were less frequent. It is tempting to speculate the mechanism behind JTs with 17p11.2 donor breakpoints may differ from other donor breakpoints.
A total of 16 of the 33 donor chromosome breakpoints occurred at 17p11.2
(48%), indicating a preferential involvement of this band as a donor breakpoint for JTs in
CLL. Multiple breakpoint cluster regions have been identified in the centromeric region of 17p 157,158. The 17p11.2 region is characterized by multiple large palindromic low copy repeats (LCR) which may favor its increased involvement in rearrangements 159. LCRs are heavily represented in pericentromeric and subtelomeric regions and may result in unstable genomic regions that are prone to nonallelic homologous recombination 160.
LCRs have been proposed as target regions for JTs in a constitutional case 161. These characteristics of the 17p11.2 region as well as its prevalence as a recurrent rearrangement site in CLL provide a basis for its increased involvement in JTs in CLL.
JTs involving 17p11.2 as a donor breakpoint resulted in the formation of dicentric or pseudodicentric chromosomes 75% of the time. In addition, in 3 of the 10 cases with only a miscellaneous donor chromosome, dicentric chromosomes with 17p11.2 breakpoints occurred without being involved in a JT (Table 2). These translocations resulted in loss of TP53 for these patients. Sawyer et al. 162 described loss of TP53 via
JTs in multiple myeloma with 17p acting as a recipient chromosome. In contrast, in CLL we identified 17p as a frequent donor chromosome. Loss of TP53 in CLL is commonly
32 due to rearrangements involving 17p11.2 158. Dicentric chromosomes involving 17p11, most frequently dicentric (17;18), have been previously reported as a recurrent abnormalities in CLL144,163. JTs appear to be another recurring type of rearrangement in
CLL involving this region. Chromosome instability due to the presence of two centromeres in dicentric chromosomes may promote the development of JTs.
In summary, our findings suggest that JTs are a type of recurrent abnormality occurring in CLL, often in conjunction with dicentric chromosomes involving 17p. These JTs appear to behave non-randomly; sequence homology between the repetitive elements in centromeric and telomeric regions as well as other regions containing LCR may mediate these translocations. Jumping translocations in CLL commonly occur in a complex karyotype, are frequently seen in patients with loss of
TP53 and in many cases may directly contribute to this loss. Both complex karyotype and loss of TP53 are poor prognostic indicators in CLL. Patients who developed JTs developed disease at an earlier age than the average CLL population1 and required treatment relatively quickly. Of the 26 patients with JTs four (15%) progressed to
Richter’s transformation, which is slightly higher than the 5-10% seen in the average
CLL population 164. In total, these patients with JTs in CLL had poor clinical outcomes; though whether this was due to the complex karyotype, loss of TP53, the JT, or a combination of these factors is not clear. Therefore, further prospective investigation into the mechanisms behind JT formation and their clinical consequences is warranted.
33
Time since Case first JT Karyotype Number sample (months)
45,XX,- 4,del(6)(q13q21),add(16)(p13.1),der(17)t(4;17)(q21;p11.2)[cp8]/ - 45,sl,add(7)(q11.2)[cp2]/45,XX,t(5;19)(q31;p11.2),del(6)(q13q21),di c(16;17)(p11.2;p11.2)[4]/45,XX,del(6)(q13q21),- 8,der(17)t(8;17)(q11.2;p11.2) [2]/ 46,XX[4] Case 1 >1 45,XX,t(5;19)(q13;p11.2),del(6)(q13q21),dic(16;17)(p11.2;p11.2)[cp 5]/nonclonal[1] 46,XX,del(6)(q13q21)[cp3]/45,sl,-4,add(16)(p13.1), 4 der(17)t(4;17)(q21;p11.2)[cp8]/43-44,sdl1,dic(8;22)(p11.2;p11.2) [cp2]/45,sl,t(5;19)(q13;p11.2),dic(16;17)(p11.2;p11.2)[cp7]
Case 2 - 46,XY,t(1;3)(p32;q21)[1]/46,XY,t(1;4)(p32;q28)[1]/46,XY,t(1;14)( p32;q24)[1]/46,XY,t(1;22)(p32;q13)[1]/46,XY[16] 45,XX,dic(13;17)(p11.2;p11.2)[cp6]/45,XX,dic(17;20) (p11.2;q13.3)[4]/45,XX,dic(17;18)(p11.2;p11.2)[4]/46,XX,- - 17,+mar[3]/ Case 3 45,XX,dic(14;17)(p11.2;p11.2)[2]/45,XX,dic(14;22)(p11.2;p11.2)[1 ]/46,XX[10]/nonclonal[1] 8 45,XX,dic(13;17)(p11.2;p11.2)[1]/45,XX,dic(14;17)(p11.2;p11.2) [1]/45,XX,dic(17;18)(p11.2;p11.2)[1]/46,XX[11]/nonclonal[6]
45,XX,dic(8;17)(p11.2;p11.2)[cp8]/45,sl,-4[cp2]/45,sdl1,+r[1]/ - 44,sdl1,der(17)t(4;17)(q12;q25)[1]/44,sdl1,der(18)t(4;18)(q12;q23) Case 4 [1]/44,sdl1,der(21)t(4;21)(q12;q22)[1]//46,XX[6] 3 45,XX,dic(8;17)(p11.2;p11.2)[cp8]/45,sl,del(14)(q22q24)[5]/44,sl,- X,+11,der(11)t(4;11)(q12;q13)[1]/46,XX[3]/nonclonal[3]
Case 5 - 45,XY,dic(4;17)(p12;p11.2),add(8)(q24.3)[7]/44,sl,+8,-add(8)(q24), add(11)(q25),-15[cp7]/44,sdl1,add(18)(q23)[4]/46,XY[2]
46,XX,del(17)(p13)[cp3]/46,sl,t(12;13)(q24.1;q14)[4]/46,sl,t(1;13)(q Case 6 - 42;q14),del(4)(p15)[4]/46,sl,der(4)del(4)(p15)t(4;13)(q31;q14),- 13,add(15)(q26)[3]/ 46,XX[5]/nonclonal[1] 44,XX,der(3)t(3;?)(p25;?)t(?;13)(?;q21)add(3)(q27),del(4)(p16),-9,- 10,del(13)(q14),add(15)(p13),add(17)(p11.2)[cp13]/44,sl,der(8)t(8;9 Case 7 - )(p23;q12)[2]/44,sl,der(3)t(3;9)(q27;q12)[1]/44,XX,add(1)(p13) ,-9,add(9)(p23),del(12)(q15q24.3),- 13,add(17)(p11.2)[1]/nonclonal[1]
Table 2. Complete karyotypes of patients with jumping translocations. Karyotypes are described using ISCN 2009 nomenclature(ref). Jumping translocations are bolded.
Continued
34
Table 2 continued
44,XX,der(3)t(3;?)(p25;?)t(?;13)(?;q21)add(3)(q27),add(4)(p16),- 9,add(10)(q22),del(13)(q12q22),add(15)(p13),-17[5] /44,sl,der(8)t(8;9)(p23;q12)[cp5]/44,sl,+add(3)(q21),-der(3),+10,- add(10)[4]/44,sl,del(6)(q22)[3]/44,sl,add(13)(q34)[2]/44,sl,der(3)t(3 ;9)(q27;q12)[2]/44,sl,der(6)t(6;9)(q22;q12)[2]/44,sl,dic(7;9)(q36;p 1 12)[2]/44,sl,der(12)t(9;12)(q12;q24.1)[2]/44,sl,der(1)t(1;9)(q44;q1 2)[1]/44,sl,der(6)t(6;9)(p26;q12)[1]/44,sl,der(12)t(9;12)(q12;p13)[ 1]/44,sl,der(14)t(9;14)(q12;q32)[1]/45,XX,add(3)(q27),add(8)(p11. 2),der(9)t(9;12)(q32;q14),del(13)(q12q22),add(17)(p11.2),- 19[2]/45,XX,add(1)(p13),add(9)(p24),dic(9;17) (q12;p11.2),del(12)(q15q24.3),- 13,+mar[cp6]/46,XX[1]/nonclonal[2] 44,XX,der(3)t(3;13)(p25;q21)ins(3);?)(p25;?)add(3)(q27),add(4)(p1 6),-9,add(10)(q22),del(13)(q12q22),add(15)(p13),-17[cp12]/ 2 44,sl,der(8) t(8;9)(p23;q12)[cp3]/45,XX,add(1)(p13), add(9)(p24),dic(9;17)(q12;p11.2),del(12)(q15q24.3),-13, +mar[3]/44,XX,add(1)(p13),-9,add(9)(p23), del(12)(q15q24.3),- 13,add(17)(p11.2)[1]/46,XX[1] 44,XX,der(3)t(3;13)(p25;q21)ins(3;?)(p25;?)add(3)(q27),add(4)(p16 7 ),-9,add(10)(q22),del(13)(q12q22),add(15)(p13),-17[10]/44,sl,der(8) t(8;9)(p23;q12)[3]/44,sl,der(6)t(6;9)(q22;q12)[1]/44,sl,dic(7;9)(q36 ;p12)[1]/44,sl,add(13)(q34)[1]/46,XX[4] - 43,XY,add(5)(q35),-8,-13,-15,add(17)(p11.2),add(21)(p11.2)[1]/ 46,XY[18]/nonclonal[1] 43,XY,add(5)(q35),add(7)(q36),-8,-13,-15,add(17)(p11.2),add(21) (p11.2)[cp5]/43-44,X,-Y,-3,-10,del(13)(q12q14),der(17)t(?3;17) 12 (p13;q25),+der(?)t(?;3)(?;q?21)[cp2]/46,XY,add(1)(p31),add(8)(q24 ),-15,-16,-17,add(19)(q13.1), der(19)t(1;19)(q21;q13.3), +mar2,+mar3,+mar4 [cp2]/44,XY,-10,add(13)(q12), add(15)(p11.2),dic(17;18)(p11.2;p11.2)[1] /46,XY[9]/nonclonal[1] 43,XY,add(5)(q35),add(7)(q36),-8,-13,-15,add(17)(p11.2), Case 8 13 add(21)(p11.2)[cp5]/43,sl,+5,-add(5)(q35),add(11)(q24)[2] /43,sl,add(18)(p11.2)[cp3]/46,XY[8]/nonclonal[2] 40,XY,add(5)(q35),add(7)(q36),-8,-13,-15,add(21) 15 (p11.2)[cp1]/40,add(X)(q26),-Y,add(7)(q36),-8,-13,- 15,add(17)(p11.2),add(21)(p11.2)[cp1]//46,XX[18]donor 43,XY,add(5)(q35),add(7)(q36),-8,-13,-15,add(17)(p11.2) ,add(21)(p11.2)[cp7]/43,sl,add(18)(p11.2)[1]/44,X,-Y,add(1) 17 (p31),add(8)(q24),-15,-16,-17,add(18)(q22),add(19)(q13.1), der(19)t(1;19)(q21;q13.3),+mar2,+mar3,+mar4[cp2]/45,XY,der(10)t (10;21)(q25;q11.2),?del(13)(q12q14),-21[2]/46,XY[8] - 45,XX,dic(17;18)(p11.2;p11.2)[2]/46,XX[20] 4 45,XX,dic(17;18)(p11.2;p11.2)[5]/47,XX,+12[2]/46,XX[22]/nonclo nal[1] Case 9 8 45,XX,dic(17;18)(p11.2;p11.2)[2]/46,XX[17]/nonclonal[1]
45,XX,dic(17;18)(p11.2;p11.2)[cp3]/42,XX,dic(17;21)(p11.2;p11.2 17 )[1]/46,XX[9]/nonclonal[5]
Continued
35
Table 2 continued
45,XX,dic(17;18)(p11.2;p11.2)[cp8]/45,XX,- 20 8,der(17)t(8;17)(q21;p11.2) [4]/45,XX,- 9,add(17)(p11.2)[cp4]/46,XX[3]/nonclonal[1] 45,XX,dic(17;18)(p11.2;p11.2)[cp8]/45,XX,- 23 8,der(17)t(8;17)(q21;p11.2) [cp1]/45,XX,- 9,add(17)(p11.2)[cp3]/nonclonal[1]//46,XY[cp7]donor 45,XX,dic(17;18)(p11.2;p11.2)[cp7]/45,XX,-9 25 ,add(17)(p11.2)[cp4]/45,XX,-8,der(17)t(8;17)(q21;p11.2)[1]/ 47,XX,dic(17;21)(p11.2;p11.2)[1]/nonclonal[2]//46,XY[5]donor - 46,XX,del(13)(q14q22)[2]/46,sl,dic(17;18)(p11.2;p11.2)[2]/46,XX,[1 5]/ nonclonal[1] 2 45,XX,del(13)(q14q22),dic(17;18)(p11.2;p11.2)[4]/46,XX,idic(17)(p 11.2) [3]/46,XX[14]/nonclonal[1] Case 10 8 45,XX,del(13)(q14q22),dic(17;18)(p11.2;p11.2)[9]/46,XX,idic(17)(p 11.2) [cp9]/46,XX[12] 10 46,XX,idic(17)(p11.2)[cp10]/45,XX,del(13)(q14q22),dic(17;18)(p11. 2;p11.2)[3] 46,XX[6]//nonclonal[1] - 44,XY,add(3)(p26),der(11)t(11;15)(q21;q13),del(12)(q22),- 15,dic(17;18)(p11.2;p11.2)[cp10]/46,XY[9]/nonclonal[1] 44,XY,add(3)(p26),der(11)t(11;15)(q21;q13),del(12)(q22),ider(13)(q1 2 0)del(13)(q14.1q22),-15,dic(17;18)(p11.2;p11.2)[cp12]/43,sl,+13,- ider(13),+psu dic(17;8)(p11.2;p21),-dic(17;18)[cp7]/46,XY[1]
44,XY,add(3)(p26),der(11)t(11;15)(q21;q13),del(12)(q22),ider(13)(q1 8 0)del(13)(q14.1q22),-15,dic(17;18)(p11.2;p11.2)[cp6] Case 11 /44,sl,+add(13)(q14),-ider(13)[8]/42,sl,-9,+13,-ider(13),- dic(17;18),+psu dic(17;8)(p11.2;p21)[cp4 one is 4n]/46,XY[2] 44,XY,add(3)(p26),der(11)t(11;15)(q21;q13),del(12)(q22),- 15,dic(17;18) (p11.2;p11.2)[cp4]/43-45,sl,dic(13;18) 9 (p11.2;q11.2)[cp7]/42,sl,-8,-9,+mar1,+mar2[cp5,one is 4n]/ 43,sl,+mar3[cp2]/44,sl,ider (13)(q10)del(13)(q14.1q22) [1]/44,sl,add(13)(q14)[1] 43,Y,t(X;9)(q26;q13),-1,add(2)(q31),t(2;14)(p23;q32), del(4)(p14),der(8)add(8)(p11.2)t(8;15)(q24.3;q15),der(9)add(9)(p22)a dd(9)(q22),derdic(12;19)(p11.2;p13.1)ins(12;?)(p11.2;?),der(14)t(14 Case 12 - ;15)(q32;q15),-15,-15,-17,der(20)t(15;20)(q15;q13.3) ins(20;8)(q13.3;q24.1q24.3),der(22)add(22)(p11.2)t(14;22)(q32;q13), +mar1,+mar2[cp23 one is 4n]/43,sl,add(8)(p21)[cp9]/43-45,sl,der dic(11;12)(q25;p11.2) ins(11;?)(q25;?),+12,-der dic(12;19)[cp3] 44,XX,der(10)t(10;15)(q23;q13),dic(15;18)(p11.2;p11.2),del(17)(p11. 2),-18[cp5]/44,sl,+der(15) t(15;18)inv(15;18)(18qter- - >18q21.1::15q13->15p11.2::18q21.1->18q11.2::15q13->15qter),- dic(15;18)(p11.2;p11.2)[2]/43,sl,-3,der(12)t(3;12) Case 13 (q12;q24)[11]/42,sl,dic(6;17) (p11.2;p11.2)[1]/46,XX[1] 44,XX,der(10)t(10;15)(q23;q13),dic(15;18)(p11.2;p11.2),del(17)(p11. 2),-18[3]/43,sl,-3,der(12)t(3;12)(q12;q24)[10]/82,sdl1x2,dic(14;17) 13 (p11.2;p11.2)[1]/85,sdl1x2,idic(17)(p11.2)[1]/84,sdl1x2,dic(7;17)(q3 6;p11.2)dup(7)(q11.2q32)[1]/82-85,sdl1x2[cp2]/42,sld1,dic(4;17) (p12;p11.2)[1]/46,XX[1] Continued
36
Table 2 continued
44,XX,der(10)t(10;15)(q23;q13),dic(15;18)(p11.2;p11.2),del(17)(p11 19 .2),-18[cp4]/43,sl,-3,der(12)t(3;12) (q12;q24)[cp9]/42,sdl1,- del(17),dic(17;20)(p11.2;p11.2)[2]/42,sdl1,-del(17),dic(4;17) (p12;p11.2)[2]/46,XX[3]
46,XY,add(5)(q35),add(13)(q12)[cp3]/46,sl,add(17)(p11.2)[cp2]/47,s Case 14 - l,add(3)(q29),+9,del(9)(q22)[cp3]/45,sl,psu dic(17;3)(p11.2;p13) [cp7]/43,sl, add(1)(q32),add(3)(p21),dic(12;17)(p11.2;p11.2),-22,- 22[cp4]/ 46,XY[1]/nonclonal[2] 44-46,XX,der(4)t(4;17)(p14;q11.2),-13,-17[1]/42,sl,-X,del(2) (p22),del(6)(q13q21),-9,add(14)(q32)[1]/42,sl,-X,del(6)(q13q21),- 8,del(11)(q21q23),dic(14;15)(p11.2;p11.2),dup(14)(q24.1q13), Case 15 - ,+der(?)(8qterf8q24.1::?->8cen->?::8q11.2->8qter)[cp10]/ 42,sdl2,add(4)(q35)[cp2]/41,sdl3,+4,-add(4),dic(10;18)(q26;p11.2) [3]/41,sdl4,t(5;10)(p13.1;p13),+10,-dic(10;18),+der(14) dic(14;18)(p11.2;p11.2)dup(14)(q24.1q13),-dup(14)[cp5] 45,XY,dic(17;18)(p11.2;p11.2)[cp12]/45,sl,-13,+mar[2]/46,XY, - add(4)(p15),t(8;17)(q11.2;p11.2)[cp4]/45,XY,- Case 16 15,der(17)t(15;17)(q15;p11.2)[1]/nonclonal[2]
>1 45,XY,dic(17;18)(p11.2;p11.2)[14]/46,XY,del(6)(q13)[1]/46,XY,add (4)(p15),t(8;17)(q11.2;p11.2)[1]/46,XY[1]/nonclonal[3]
44,XY,-13,dic(14;17)(p11.2;p11.2),der(19)t(13;19)(q14;p13.3) Case 17 - ins(19;?) (p13.3;?)[5]/44,sl,add(Y)(p11.3)[5]/46,XY,add(19) (p13.3)[2]/46,XY[16]/nonclonal[2] 45,XY,dic(17;20)(p11.2;p11.2)[6]/45,X,-Y[3]/43,XY,add(4)(q23), - psu dic(5;13)(p13;p11.2),der(9)t(4;9)(q23;p13),- 12,dic(17;20)(p11.2;p11.2), add(18)(p11.2)[1]/46,XY[10] 45,XY,dic(17;20)(p11.2;p11.2)[cp6]/44,sl,add(4)(q23),der(9)t(4;9)(q Case 18 4 23;p13),-12[2]/43,sdl1,psu dic(5;13)(p13;p11.2),add(18)(p11.2) [2]/44,sl,add(2) (q37),add(13)(q34)[cp4]/44,sl,-13,add(18) (q23)[2]/46,XY[6]/nonclonal[3] 43,XY,add(4)(q23),psu dic(5;13)(p13;p11.2),der(9)t(4;9)(q23;p13),- 9 12,dic(17;20)(p11.2;p11.2),add(18)(p11.2)[3]/45,XY,t(1;10)(p32;q21 ),-3,dic(13;18)(p11.2;p11.2),+mar[1]/46,XY[15]/nonclonal[1] 45,XY,add(3)(q11.2),psu dic(8;3)(p21;p21),-14[4]/45, Case 19 - XY,add(8)(p21),-14[3]/45,XY,dic(8;13) (p11.2;p11.2), del(17) (p11.2)[cp3]/46,XY,add(1)(p13),t(1;8)(p21;p21)[1]/46,XY[13] 46,XX,der(3)t(3;3)(p13;q23),dic(8;17)(p11.2;p11.2),+12[cp4]/ - 46,sl,add(1)(q42),del(9)(q13q22),+add(17)(p11.2),-dic(8;17) [2]/46,XX,t(1;9)(q42;p13),+12,dic(17;20)(p11.2;p11.2)[cp3] /46,XX[10]/nonclonal[4] Case 20 3 46,XX,+12,dic(17;20)(p11.2;p11.2)[8]/46,sl,t(1;9)(q42;p13)[1]/46,X X[8]/nonclonal[3] 46,XX,+12,dic(17;20)(p11.2;p11.2)[cp5]/46,sl,t(1;9)(q42;p13)[1]/45 5 -46,XX,-5,-6,dic(8;17)(p11.2;p11.2),-9,del(13)(q14q22), add(15)(p11.2),+mar1,+mar2[cp2]/46,XX[9]/nonclonal[3] Continued
37
Table 2 continued
46,XX,+12,dic(17;20)(p11.2;p11.2)[cp12]/46,sl,del(X)(q26),t(1;9)(q42; 8 p13)[cp3]/45,XX,-5,-6,dic(8;17)(p11.2;p11.2),add(13)(q14), +mar1,+mar2[4]/46,XX[9]/nonclonal[2] - 45,XY,dic(8;17)(p11.2;p11.2),del(14)(q24q32)[cp20]
46,XY,del(14)(q24q32)[2]/46,sl,del(17)(p13)[2]/45,sl,dic(8;17)(p11.2;p 7 11.2)[cp12]/45,sl,dic(12;17)(p11.2;p11.2),der(13)t(12;13)(p11.2;p11.2) [3]/46,XY[1] 46,XY,del(14)(q24q32)[cp2]/45,sl,dic(8;17)(p11.2;p11.2)[cp12]/45,sl,d 9 ic(12;17)(p11.2;p11.2),der(13)t(12;13)(p11.2;p11.2)[3]/44,sdl1,add(2)( q37),-13,-15,+mar[cp2]/46,XY[1]
46,XY,del(14)(q24q32)[cp1]/45,sl,dic(8;17)(p11.2;p11.2)[cp12]/45,sl,d 11 Case 21 ic(12;17)(p11.2;p11.2),der(13)t(12;13)(p11.2;p11.2)[2]/46,sl,del(17)(p1 2)[cp1]/46,XY[4] 13 45,XY,dic(8;17)(p11.2;p11.2),del(14)(q24q32)[cp12]/46,XY[2]//46,X X[6]donor 46,XY,del(14)(q24q32)[1]/45,sl,dic(8;17)(p11.2;p11.2)[cp15]/45,sdl1,a 15 dd(1)(q21),der(2)t(2;15)(q23;q15)ins(2;?)(q37;?),add(9)(p21),del(13)(q 14q32),del(15)(q15)[2]//46,XX[3]donor 18 46,XY,del(14)(q24q32)[1]/45,sl,dic(8;17)(p11.2;p11.2)[cp11]//46,XX[ 8]donor 47,XY,+12,del(13)(q14q21.2)[cp3]/47,sl,der(8)t(8;8)(p21;q13)[10]/47,sl Case 22 - ,der(4)t(4;13)(q33;q21),-del(13),+der(13)t(4;13)(q33;q14)[8]/47,sl, der(12)t(12;13)(q15;q21),der(13)t(12;13)(q15;q14)[5]/46,XY[1] 39-46,XX,add(11)(q13),-16,der(17)t(11;17)(q21;p11.2),add(18)(p11.2) - [cp10]/41-43,sl,der(2)t(2;3)(p13;q25)t(3;11)(q29;q21)t(2;8) (q33;q22),add(3)(q25),-4,t(4;9)(q25;q21),-8,-der(17)t(11;17) Case 23 ,+add(17)(p11.2)[cp8]/43,sdl1,del(17)(p11.2)[2] 45,XX,add(11)(q13),-16,der(17)t(11;17)(q21;p11.2),add(18)(p11.2) 1 [cp7]/ 43,sl,der(2)t(2;3)(p13;q25)t(3;11)(q29;q21)t(2;8)(q33;q22) ,add(3)(q25),-4,t(4;9)(q23;q21),-8,-der(17)t(11;17),+add(17)(p11.2) [cp9]/ 45,sl,add(4)(p16),del(5)(q13),add(6)(p25)[cp3]/46,XX[1] 45,XY,t(1;1)(p36.3;q42),add(4)(q31),add(8)(p21),der(11)t(11;11) - (p11.2;q13),dic(11;17)(q13;p11.2)[cp9]/45,sl,del(3)(p13)[1]/45- 46,XY,del(3)(p13),add(8)(p21),der(17)t(4;17)(q25;p11.2)[cp2]/46,XY[ Case 24 3]/nonclonal(XY)[3]//46,XX[2]donor 46,XY,t(1;1)(p36.3;q42)[1]/45,sl,add(4)(q31),add(8)(p21),der(11)t(11;1 >1 1)(p11.2;q13),dic(11;17)(q13;p11.2)[cp5]/45,XY,dic(17;19)(p11.2;q12 )[1]/44,XY,dic(6;17)(q12;p11.2),-19[1]/46,XY[4]/nonclonal(XY)[4] //46,XX[4]donor 42-46,XY,add(10)(q22),idic(17)(p11.2)[cp6]/43,sl,-2,dic(8;17)(p11.2; Case 25 - p11.2),-13,der(15)t(2;15)(q21;p11.2),-idic(17)[cp2]/43,sdl1,add(19) (p13)[cp6]/45,sl,i(8)(q10)[2]/45,sl,dic(15;18)(p11.2;p11.3),add(20)(q13 )[cp2]/46,XY[2] 47,XY,der(5)t(4;5)(q12;q35),+19[cp2]/47,sl,+add(5)(q11.2),-der(5)t(4; Case 26 - 5),der(9)t(4;9)(q12;q34),add(11)(p15),der(15)t(5;15)(q22;q22),add(16) (p13.3)[cp16]/46,XY[2]
38
Case Donor breakpoint Recipient breakpoint
4q21 1 17p11.2 8q11.2 16p11.2* 13p11.2* 20q13.3* 3 17p11.2 14p11.2* 18p11.2* 9q12* 7 17p11.2 Unknown Unknown 8 17p11.2 18p11.2* 18p11.2* 8q21 9 17p11.2 Unknown 21p11.2* 18p11.2* 10 17p11.2 17p11.2* 18p11.2* 11 17p11.2 8p21^ 20p11.2* 4p12* 14p11.2* 13 17p11.2 17p11.2* 7q36* 6p11.2*
Table 3. Jumping translocation breakpoints for the 17p11.2 donor chromosome group. Bold, repeat breakpoint location across all JTs identified in CLL; *, dicentric chromosome; ^, pseudodicentric chromosome. Green background, centromeric breakpoint; pink background, telomeric breakpoint.
Continued
39
Table 3 continued
Unknown 14 17p11.2 3p13^ 12p11.2* 18p11.2* 16 17p11.2 8q11.2 15q15 6p11.2* 17 17p11.2 21p11.2* 14p11.2* 20p11.2* 20 17p11.2 8p11.2* Unknown 8p11.2* 12p11.2* 22p11.2* 21 17p11.2 18p11.2* 15p11.2* 21p11.2* 11q21 23 17p11.2 Unknown 4q25 11q13* 24 17p11.2 6q12* 19p12* 17p11.2* 25 17p11.2 8p11.2*
40
Case Donor breakpoint Recipient breakpoint
3q21 4q28 2 1p32 14q24 22q13 17p11.2* 3 14p11.2 22p11.2* 11q13 17q25 4 4q12 18q23 21q22 8q24.3 5 Unknown 11q25 18q23 12q24.1 6 13q14 1q42 4q31 8p23 12q24.1 6q22 3q27 7 9q12 17p11.2* 1q44 6p26 12p13 14q32 Table 4. Jumping translocation breakpoints for the miscellaneous donor chromosome group. Bold, repeat breakpoint location across all JTs identified in CLL; *, dicentric chromosome; ^, pseudodicentric chromosome. Green background, centromeric breakpoint; pink background, telomeric breakpoint. Continued
41
Table 4 continued
5q35 7q36 8 Unknown 11q24 Xq26 Unknown* 12 12p11.2 Unknown* 10q26* 15 18p11.2 14p11.2* Unknown 17 19p13.3 Unknown 5p13^ 13p11.2 18p11.2* 18 13p11.2* 18p11.2 Unknown 3p21^ 19 8p21 1p21 Unknown 4q33 22 13q21 12q15 17p11.2 23 11q21 3q29 2q21 25 15p11.2 18p11.3* 5q35 26 4q12 9q34
42
Figure 3. Time to treatment and survival probabilities for CLL patients with jumping translocations. Probabilities of time to treatment and overall survival from diagnosis for CLL patients who developed jumping translocations. The estimated median TTT was 11.5 months and the estimated OS at 6 years from diagnosis was 66%.
43
Figure 4. Partial karyotypes showing JTs in CLL. Arrows indicate breakpoint locations. (A) JTs from Case 2 with balanced JTs showing t(1;3)(p32;q21), t(1;4)(p32;q28), t(1;14)(p32;q24), and t(1;22)(p32;q13). (B) JTs from Case 3 with both 17p11.2 and 14p11.2 acting as both donor and recipient breakpoints. The first JT has 17p11.2 as a donor breakpoint (top) translocating to form the dicentric chromosomes dic(13;17)(p11.2;p11.2), dic(14;17)(p11.2;p11.2), dic(17;18)(p11.2;p11.2), and dic(17;20)(p11.2;q13.3). The second JT in this patient has the 14p11.2 recipient breakpoint from the first JT now acting as a donor breakpoint (bottom) translocating to form the dicentric chromosomes dic(14;17)(p11.2;p11.2) and dic(14;22)(p11.2;p11.2). (C) JT from Case 10 showing dicentric chromosomes idic(17)(p11.2) and dic(17;18)(p11.2;p11.2).
44
CHAPTER 3: Evaluating 2p gain and near-tetraploidy as biomarkers for progression on ibrutinib
3.1 Introduction
Ibrutinib is a first-in-class oral covalent inhibitor of Burton’s tyrosine kinase
(BTK)165 recently FDA approved for use in CLL. Ibrutinib is profoundly changing the landscape of CLL treatment, with substantial response rates in relapsed/refractory CLL as well as previously untreated patients 166-168. Despite these promising results, there remains a subset of patients who progress on ibrutinib. Patients who relapse on ibrutinib do so with progressive CLL or through Richter’s transformation, a transformation of their
CLL into aggressive lymphoma, predominately diffuse large B cell lymphoma
(DLBCL)108. Patients who progress with CLL typically acquire mutations in BTK, in the
C481 binding pocket for ibrutinib, which confers resistance to the drug 169. Additionally, mutations in downstream PLCG2 drive resistance by bypassing the need for signaling through BTK 170. Mutations in these genes are found in the majority of patients who progress with CLL; however, they are considerably less frequent in patients who progress with Richter’s transformation108. Identifying biomarkers that are predictive for progression on ibrutinib, particular with Richter’s transformation, is critical at this time as an increasing number of patients begin to receive ibrutinib and those who progress often have very aggressive disease and poor prognosis108,171. Cytogenetic markers have been
45 identified as predictive for ibrutinib progression, including BCL6 abnormalities and complex karyotype (3 or more chromosomal abnormalities)108,172. In addition, various prognostic markers in CLL have been associated with increased risk of Richter’s transformation, including inactivation TP53 or CDKN2A, CMYC abnormalities, NOTCH1 mutation, and CD38 expression173-177. However, there are currently no specific risk factors associated with developing Richter’s transformation.
Gain or amplification of the 2p14-16 region occurs in approximately 3-10% of
CLL178-180. The chromosomal region of 2p14-16 gained in CLL contains several potential genes of interest; including the NFKB subunit REL, nuclear transporter XPO1, and proto- oncogene BCL11A 180,181. 2p gain has previously been reported as associated with poor
OS and increased risk of Richter’s transformation182,183. Gain of this region has been detected in Richter transformed cells and also occurs in de novo DLBCL174,183,184. 2p gain was recently described as a late stage driver in CLL185. In this study we sought to determine if gain of 2p would be predictive for progression on ibrutinib, for either
Richter’s transformation or CLL progression.
While performing fluorescence in situ hybridization for 2p we observed a number of cases had signals indicative of a near-tetraploid cell, a cell with four copies of most chromosomes. Near-tetraploidy has been described in various lymphomas186,187, however, to our knowledge the incidence of near-tetraploidy in CLL has not been examined 188.
Tetraploidy in cancer cells can promote chromosome instability and may provide a fitness advantage to the cells by buffering them against deleterious effects caused by
46 gross aneuploidy, rearrangements, and mutations189. Therefore, we evaluated tetraploidy, in addition to 2p gain, in these patients.
3.2 Materials and Methods
All patients were consented to participate in the study in accordance with the declaration of Helsinki and from protocols approved by the Ohio State University IRB.
FISH analysis was performed on the mitogen-stimulated cultures from patients peripheral blood or bone marrow samples collected prior to receiving ibrutinib. FISH probes for
REL (2p15) and DIRC1 (2q32.1) were used (Empire). Hybridization was according to the manufacturer’s directions. Two hundred cells per probe were analyzed, 100 by each of two independent observers. To confirm near-tetraploidy, previous FISH results from the standard CLL panel performed on these samples were analyzed for consistence with a near-tetraploid clone. From here forward we will refer to near-tetraploidy as well as tetraploidy as simply tetraploidy, however, using FISH alone we cannot determine definitively how many chromosomes are present.
A total of 300 patients were included in the tetraploid analysis and 296 in the gain of 2p analysis. The characteristics of this patient population from four ibrutinib clinical trials have been previously described by Maddocks et al 108. Time to discontinuation of treatment was measured from the first date of treatment with ibrutinib until the off-study date, censoring patients who had not discontinued ibrutinib therapy at the date of last contact and patients who went off study for transplant or continued treatment at another institution. Median follow-up was calculated among all patients (n = 300) censored for
47 time to discontinuation of treatment. Gray models of cumulative incidence were fit to identify variables associated with a particular progression type and in the presence of competing risks190,191. All models were adjusted for monotherapy with ibrutinib vs combination of ibrutinib with ofatumumab. Survival following discontinuation was calculated from the off-study date until the date of death from any cause, censoring patients at last contact. Survival estimates were calculated by the Kaplan-Meier method147 and differences between curves were tested with the log-rank test. All tests were 2-sided, and statistical significance was declared at α=.05.
3.3 Results
2p gain results
2p gain was observed in 44 (14.9%) of 296 patients analyzed from four ibrutinib trials. We examined if gain of 2p was associated with any other baseline characteristics for these patients and found gain of 2p was strongly associated with BCL6 and MYC abnormalities, and less strongly associated with deletion 17p and complex karyotype
(Table 5). Next we looked at 2p gain in relation to outcome. Of the patients with gain of
2p, 20% progressed with CLL on ibrutinib, 11% progressed with Richter’s transformation, 14% discontinued therapy for reasons unrelated to progression, and 50% were still on treatment. In comparison, for patients with no gain of 2p, 9% progressed with CLL on ibrutinib, 8% progressed with Richter’s transformation, 22% discontinued therapy for reasons unrelated to progression, and 57% were still on treatment (Table 6).
However, while the proportion of patients who progressed with CLL or Richter’s
48 transformation was higher in the gain of 2p group, in a time to event analysis 2p gain was not a significant predictor for these events (Table 6). The cumulative incidence for progression + transformation and for transformation alone showed a trend of an increased incidence for the 2p positive patients but this did not reach statistical significance (Figure
5).
Tetraploidy results
Tetraploidy was observed in 9 (3%) of 300 patients analyzed. Tetraploidy was associated with the baseline characteristics Rai stage III/IV, trisomy 12, deletion 17p, and complex karyotype (Table 7). In terms of outcome, of the patients positive for tetraploidy, 11% progressed with CLL on ibrutinib, 67% progressed with Richter’s transformation, and 22% were still on treatment. In comparison, for patients with no tetraploidy, 11% progressed with CLL on ibrutinib, 7% progressed with Richter’s transformation, 21% discontinued therapy for reasons unrelated to progression, and 56% were still on treatment (Table 8). In a cumulative incidence analysis tetraploidy was significantly associated with progression + transformation and for transformation alone
(Figure 6). It was not significantly associated with progression with CLL alone (Figure
6), indicating tetraploidy primarily occurs in patients who transform. All patients with tetraploidy also had complex karyotype; however, this group of patients had an increased incidence for transformation compared to patients with complex karyotype without tetraploidy (Figure 6). In a multivariable model previously described108 adjusting for lactate dehydrogenase, MYC abnormalities, complex karyotype, and prior therapies, tetraploidy was strongly associated with a higher incidence of discontinuing ibrutinib due
49 to Richter’s transformation (HR 7.62, 95% CI 3.20-18.13, p < 0.0001). Tetraploidy also showed a trend towards decreased OS that was not statistically significant (Figure 7).
3.4 Discussion
We aimed to identify novel prognostic markers in patient samples obtained prior to starting ibrutinib therapy associated with progression. As a retrospective study we have several limitations. Patient samples varied in the site they were collected from, peripheral blood or bone marrow. Time of sampling also varied from days to months prior to beginning treatment. Serial samples were unavailable for most patients, thus we are unable to determine if the frequency of the 2p or tetraploid clones altered during treatment, nor do we know if gain of these abnormalities after starting treatment will predict for progression. It would be useful to monitor patients serially, prior to and throughout treatment, for these abnormalities to address these questions.
2p gain has previously been associated with Richter’s transformation182; however, in this study 2p gain was not significantly associated with progression on ibrutinib through transformation. Our studies differ in the time of sampling; Rinaldi et al. examined 2p gain in patient samples at diagnosis182 while our study was enriched for patients later in the course of their disease. We may also have not reached significance due to the small number of CLL patients who have progressed on ibrutinib and short follow-up time. There was a trend for increased incidence of progression in patients with gain of 2p after 24 months, therefore, analysis after a longer follow-up period would be useful to determine if 2p gain may predict for a late progression event.
50
In this study we identified tetraploidy as a prognostic marker for progression on ibrutinib through Richter’s transformation. While this marker co-occurs with complex karyotype, patients with both tetraploidy and complex karyotype had an increased incidence for transformation compared to complex karyotype alone. Additionally, CLL cytogenetics laboratories often do not perform stimulated karyotype and are unable to identify patients with complexity. FISH for tetraploidy is a viable alternative in these settings to increase the ability to identify patients at risk for transformation on ibrutinib.
An additional benefit for analyzing tetraploidy as a prognostic marker is that it can be detected using the standard FISH panel and no additional probes may be necessary. It is critical to validate our finding in an independent data set. It would also be interesting to determine if tetraploidy predicts for transformation in settings outside of ibrutinib therapy.
51
Characteristic Total No Gain of 2p Gain of 2p P-value (n = 296) (n = 253) (n = 43) Study OSU-10032 47 (16) 34 (13) 13 (30) OSU-10053 67 (23) 55 (22) 12 (27) 0.03 OSU-11133 145 (49) 131 (52) 14 (32) OSU-12024 37 (13) 32 (13) 5 (11) Monotherapy with IB No 67 (23) 55 (22) 12 (27) 0.44 Yes 229 (77) 197 (78) 32 (73) Age Median 65 65 62 0.21 Range 26-91 26-91 50-79 Sex, number (%) Male 209 (71) 183 (73) 26 (59) 0.08 Female 87 (29) 69 (27) 18 (41) Rai Stage, number (%) 0 11 (4) 10 (4) 13 (30) 0.86 I 67 (23) 56 (22) 12 (27) II 22 (7) 20 (8) 14 (32) (1.00 for Rai III 43 (15) 38 (15) 5 (11) 3/4 vs lower) IV 153 (52) 129 (51) 24 (55) Elevated LDH No 110 (38) 94 (38) 16 (37) 1.00 Yes 178 (62) 151 (62) 27 (63) LDH Median 219 217 221 0.92 Range 96-1485 96-1485 100-495 Number of prior therapies Median 3 3 3.5 0.74 Range 0-16 0-16 0-12 BCL6 abnormality No 266 (91) 232 (93) 34 (79) 0.009 Yes 27 (9) 18 (7) 9 (21) MYC abnormality . No 230 (79) 204 (82) 26 (60) 0.004 Yes 63 (22) 46 (18) 17 (40) Trisomy 12 No 241 (82) 207 (83) 34 (79) 0.52 Yes 52 (18) 43 (17) 9 (21) Del(13q) No 141 (48) 121 (48) 20 (47) 0.87 Yes 152 (52) 129 (52) 23 (53) Del(11q) No 210 (72) 175 (70) 35 (81) 0.15 Yes 83 (28) 75 (30) 8 (19) Del(17p) No 176 (60) 156 (62) 20 (47) 0.06 Yes 117 (40) 94 (38) 23 (53) Del(11q)/Del(17p) No 109 (37) 94 (38) 15 (35) 0.86 Yes 184 (63) 156 (62) 28 (65) Complex karyotype No 119 (41) 108 (44) 11 (26) 0.03 Yes 170 (59) 138 (56) 32 (74) IGVH Mutated 53 (20) 48 (22) 5 (12) 0.21 Unmutated 211 (80) 174 (78) 37 (88)
Table 5. Associations between 2p gain and demographic and molecular variables. Associations were tested using the Wilcoxon rank sum and Fisher’s exact tests. Abbreviations, IB, ibrutinib, LDH, lactate dehydrogenase, del, deletion.
52
Total No Gain of 2p Gain of 2p P- Outcome (n = 296) (n = 252) (n = 44) value Reason Discontinued Progressive CLL 32 (11) 23 (9) 9 (20) Transformation 26 (9) 21 (8) 5 (11) Other Event 61 (21) 55 (22) 6 (14) Tx Elsewhere 12 (4) 10 (4) 2 (5) N/A – Still on Treatment 165 (56) 143 (57) 22 (50) Cuminc: Progression (CLL+T) Number of Events 58 44 14 % at 12m (95% CI) 5.4 (2.8-8.0) 5.6 (2.7-8.5) 4.5 (0-10.8) 0.12 % at 24m (95% CI) 12.7 (8.8-16.6) 11.9 (7.8-16..) 16.8 (5.2-28.3) % at 36m (95% CI) 20.5 (15.2-25.8) 18.1 (12.6-23.6) 31.3 (16.1-46.6) Cuminc: CLL Progression Number of Events 32 23 9 % at 12m (95% CI) 0.7 (0-1.6) 0.8 (0-1.9) 0 0.12 % at 24m (95% CI) 5.2 (2.5-7.8) 4.2 (1.6-6.8) 9.9 (0.5-19.3) % at 36m (95% CI) 10.8 (6.6-15.1) 9.1 (4.8-13.5) 18.9 (5.9-31.8) Cuminc: Transformation Number of Events 26 21 5 % at 12m (95% CI) 4.8 (2.3-7.2) 4.8 (2.1-7.4) 4.5 (0-10.8) 0.61 % at 24m (95% CI) 7.5 (4.5-10.6) 7.7 (4.3-11.0) 6.9 (0-14.5) % at 36m (95% CI) 9.6 (6.0-13.3) 9.0 (5.2-12.8) 12.5 (2.0-22.9) Cuminc: Other Event Number of Events 61 55 6 % at 12m (95% CI) 12.9 (9.1-16.7) 14.7 (10.3-19.1) 2.3 (0-6.9) 0.14 % at 24m (95% CI) 17.5 (13.1-21.9) 18.9 (14.0-23.7) 9.6 (0.5-18.6) % at 36m (95% CI) 22.2 (17.1-27.3) 23.2 (17.6-28.8) 15.6 (3.8-27.5)
Table 6. In a cumulative incidence analysis 2p gain was not a significant predictor for progression on ibrutinib. Associations were tested using the Gray’s test. Abbreviations: Tx, treatment, Cuminc, cumulative incidence, m, months
53
Characteristic Total Not Tetraploid Tetraploid P-value (n = 300) (n = 291) (n = 9) Study OSU-10032 49 (16) 47 (16) 2 (22) OSU-10053 69 (23) 66 (23) 3 (33) 0.25 OSU-11133 146 (49) 144 (49) 2 (22) OSU-12024 36 (12) 34 (12) 2 (22) Monotherapy with IB No 69 (23) 66 (23) 3 (33) 0.43 Yes 231 (77) 225 (77) 6 (67) Age Median 65 65 62 0.83 Range 26-91 26-91 52-79 Sex, number (%) Male 213 (71) 208 (71) 5 (56) 0.29 Female 87 (29) 83 (29) 4 (44) Rai Stage, number (%) 0 11 (4) 11 (4) 0 (0) 0.28 I 69 (23) 69 (24) 0 (0) II 22 (7) 22 (8) 0 (0) (0.03 for Rai III 44 (15) 43 (15) 1 (11) 3/4 vs lower) IV 154 (51) 146 (50) 8 (89) Elevated LDH No 113 (39) 112 (40) 1 (11) 0.16 Yes 179 (61) 171 (60) 8 (89) LDH Median 219 218 227 0.18 Range 96-1485 96-1485 142-765 Number of prior therapies Median 3 3 5 0.13 Range 0-16 0-16 2-8 BCL6 abnormality No 270 (91) 267 (93) 7 (78) 0.15 Yes 27 (9) 21 (7) 2 (22) MYC abnormality . No 232 (78) 230 (80) 5 (56) 0.09 Yes 65 (22) 58 (20) 4 (44) Trisomy 12 No 245 (82) 240 (83) 5 (56) 0.05 Yes 52 (18) 48 (17) 4 (44) Del(13q) No 145 (49) 140 (49) 5 (56) 0.75 Yes 152 (51) 148 (51) 4 (44) Del(11q) No 214 (72) 207 (72) 7 (78) 1.00 Yes 83 (30) 81 (28) 2 (22) Del(17p) No 177 (60) 175 (61) 2 (22) 0.03 Yes 120 (40) 113 (39) 7 (78) Del(11q)/Del(17p) No 110 (37) 109 (39) 1 (11) 0.16 Yes 187 (63) 179 (62) 8 (89) Complex karyotype No 119 (41) 119 (42) 0 (0) 0.01 Yes 172 (59) 163 (58) 9 (100) IGVH Mutated 53 (20) 53 (20) 0 (0) 0.35 Unmutated 215 (80) 208 (80) 7 (100)
Table 7. Associations between tetraploidy and demographic and molecular variables. Associations were tested using the Wilcoxon rank sum and Fisher’s exact tests. Abbreviations, IB, ibrutinib, LDH, lactate dehydrogenase, del, deletion.
54
Outcome Total Not Tetraploid Tetraploid P-value (n = 300) (n = 291) (n = 9) Reason Discontinued Progressive CLL 33 (11) 32 (11) 1 (11) Transformation 27 (9) 21 (7) 6 (67) Other Event 62 (21) 62 (21) 0 (0) Tx/Rx Elsewhere 12 (4) 12 (4) 0 (0) N/A – Still on Treatment 166 (55) 164 (56) 2 (22) Cuminc: Progression (CLL+T) <0.0001 Number of Events 60 53 7 % at 12m (95% CI) 5.4 (2.8-7.9) 4.5 (2.1-6.9) 33.3 (0.3-66.3) % at 24m (95% CI) 12.5 (8.7-16.4) 11.5 (7.8-15.3) 44.4 (9.4-79.5) % at 36m (95% CI) 20.5 (15.3-25.8) 18.4 (13.2-23.5) 77.8 (46.1-100) Cuminc: CLL Progression 0.53 Number of Events 33 32 1 % at 12m (95% CI) 0.7 (0-1.6) 0.7 (0-1.7) 0 % at 24m (95% CI) 5.1 (2.5-7.7) 5.3 (2.6-8.0) 0 % at 36m (95% CI) 10.6 (6.4-14.7) 10.4 (6.3-14.6) 11.1 (0-35.7) Cuminc: Transformation <0.0001 Number of Events 27 21 6 % at 12m (95% CI) 4.7 (2.3-7.1) 3.8 (1.6-6.0) 33.3 (0.3-66.3) % at 24m (95% CI) 7.4 (4.4-10.4) 6.3 (3.5-9.1) 44.4 (9.4-79.5) % at 36m (95% CI) 10.0 (6.3-13.6) 7.9 (4.6-11.3) 66.7 (32.3-100) Cuminc: Other Event 0.11 Number of Events 62 62 0 % at 12m (95% CI) 13.0 (9.2-16.9) 13.5 (9.5-17.4) 0 % at 24m (95% CI) 17.6 (13.2-22.0) 18.2 (13.7-22.6) 0 % at 36m (95% CI) 22.1 (17.1-27.2) 22.9 (17.7-28.1) 0 Overall Survival 0.06 Number of Events 82 77 5 Median, months (95% CI) NR (56.5 to NR) NR (56.5 to NR) 29.7 (8.8-NR) % at 12m (95% CI) 87.3 (83.0-90.6) 87.6 (83.3-90.9) 77.8 (36.5-93.9) % at 24m (95% CI) 80.8 (75.9-84.9) 81.3 (76.3-85.3) 66.7 (28.1-87.8) % at 36m (95% CI) 72.2 (66.1-77.3) 73.3 (67.2-78.4) 41.7 (10.9-70.8)-
Table 8. In a cumulative incidence analysis tetraploidy was a significant predictor for progression on ibrutinib. Associations were tested using the Gray’s test. Abbreviations: Tx, treatment, Cuminc, cumulative incidence, m, months
55
(a)
(b)
(c)
Figure 5. Cumulative incidence curves for progression with and without 2p gain. (a) Incidence for CLL progression and Richter’s transformation combined. (b) Incidence for CLL progression. (c) Incidence for Richter’s transformation
56
(a)
(b)
(c)
Figure 6. Cumulative incidence curves for progression with and without tetraploidy. (a) Incidence for CLL progression and Richter’s transformation combined (b) incidence for CLL progression (c) incidence for Richter’s transformation (d) incidence for transformation with complexity with or without tetraploidy. Continued
57
Figure 6 continued
(d)
P<0.0001
58
Figure 7. Kaplan-Meier curve of overall survival for patients with or without tetraploidy.
59
CHAPTER 4: The long noncoding RNA, treRNA, decreases DNA damage and is associated with poor response to chemotherapy
4.1 Introduction
As described in the introduction, CLL has a highly heterogeneous clinical course; while some patients have indolent disease that might not require therapy for years, others progress quite rapidly. Biomarkers such as unmutated IGHV63,64, unmethylated ZAP70 promoter84, and the cytogenetic abnormalities deletion 17p and deletion 11q153 have been identified as independent markers for aggressive disease. However, these markers do not demarcate all aggressive cases and many of the techniques used to measure these biomarkers tend to be variable across institutions. Identifying additional molecular markers that predict for aggressive disease or that defines a potentially different biology remains of considerable interest, particularly those which can predict response to treatment.
Gene expression can be a useful prognostic indicator in CLL, particularly the expression of noncoding RNAs such as microRNAs (miRs). Calin et. al has described a 9 miR signature associated with decreased time to treatment in CLL that includes miR-
181a (high), miR-155 (high), miR-146 (high), and miR-29c (low)192. High expression of miR-155 also associates with shorter overall survival193 and progression free survival with chemo-immunotherapy194. Beyond their prognostic significance, several miRs have
60 been shown to play critical biological roles in CLL. miR-155 expression enhances sensitivity to B-cell receptor (BCR) ligation193. Alternatively, miR-150 expression decreases the intensity of BCR signaling195. Deletion or downregulation of miR-15 and miR-16, negative regulators of the anti-apoptotic protein BCL2, occurs in the majority
(~68%) of CLL196,197. While the expression of miRs has been clearly demonstrated as important in CLL, the contribution of other noncoding RNAs has not been well established. Long noncoding RNAs (lncRNAs) are noncoding transcripts greater than
200 nucleotides in length, and recently, it has been shown that lncRNAs are able to regulate gene expression and are associated with diverse biological processes112,198. The deregulated expressions of lncRNAs, such as HOTAIR, MALAT1, and SChLAP1, have been associated with disease progression in multiple types of cancer118,129,131,199-201. In
CLL, lncRNAs, including linc-p21 and NEAT1, have been shown to be induced by DNA damage in a TP53 dependent mechanism202. In addition, circulating levels of linc-p21 in plasma from CLL patients is lower than healthy donor controls203.
In the current study, we explored the differential expression of lncRNA in CLL and identified treRNA (TRERNA1), as a lncRNA that is overexpressed in CLL and predicts for poor prognosis. TreRNA has been described to have enhancer-like function119 as well as translational regulatory functions204, and is overexpressed in breast cancer lymph-node metastases and colon cancer204. In CLL, we found high expression of treRNA was associated with poor response to chemotherapy independent of other variables, and suggest that treRNA may serve as a valuable prognostic factor in this disease. Finally, using a CLL cell line model we show that over-expression of treRNA
61 results in decreased DNA damage caused by exposure to chemotherapeutic agents, likely contributing to the impaired response to these agents in patients.
4.2 Materials and Methods
Patient samples and cell culture conditions
Blood was obtained from CLL patients with written informed consent in accordance with the Declaration of Helsinki and under a protocol approved by the
Institutional Review Board of the Ohio State University. CLL cells and normal B cells were isolated using ficoll density gradient centrifugation (Ficoll-Plaque Plus; Amersham
Biosciences) and enriched for B cells using the Rosette-Sep negative selection kit
(StemCell Technologies) according to manufacturer protocol. Cryopreserved cells utilized in the prognostic training and validation sets were obtained from the CLL
Research Consortium (CRC) tissue bank or from the ECOG-2997 clinical trial respectively. Cells were thawed in RPMI 1640 media then washed in PBS to obtain cell pellets.
Primary CLL cells and the OSUCLL cell line were grown in RPMI 1640 media supplemented with 10% fetal bovine serum, 2mM L-glutamine (Invitrogen), 100U/mL penicillin, and 100ug/mL streptomycin (Sigma). For stimulating primary CLL samples 6 well plates were coated with anti-IgM (MPbio) at a concentration of 1ug/mL. Anti-IgM was adhered overnight at 4oC then washed twice with PBS prior to adding cells at a concentration of 1x107cells/mL. Cells were stimulated with 1mg/mL CD40 ligand
(PeproTech) and 800U/mL recombinant human IL4 (PeproTech). For drugging OSUCLL
62 cells, 2x106 cells from OSUCLL empty vector or OSUCLL treRNA were incubated with vehicle (DMSO), 10uM fludarabine (Sigma), 2.36uM mafosfamide (Santa Cruz), or
10uM fludarabine plus 2.36uM mafosfamide. All conditions were given equivalent volumes of DMSO.
Microarray
Using ArrayStar Human LncRNA Array v2.0, two pools of RNA from CLL patient samples (n=6 per pool) with mixed clinical and molecular histories, both pools unstimulated and one pool also stimulated with CD40 ligand (PeproTech), were compared to a pool of healthy donor B cells (n=5) isolated from Red Cross leukopaks then cultured with or without CD40 ligand. As an exploratory analysis to identify potential differentially expressed genes for further validation, the pooled CLL samples and pooled normal B cells samples with and without CD40L stimulation were treated as independent samples in the analysis; however, these conditions were performed on the same set of pooled samples. All analyses were performed in version 3.1.0 of the R
Statistical Programming Environment. Data processing was performed with version
3.20.1 of the limma package. Background correction was performed using the “normal- exponential” model, and then quantile normalization applied. Probe-by-probe smoothed t-tests were used to identify genes differentially expressed between CLL cells and normal
B-cells. In order to account for multiple testing, a beta-uniform-mixture (BUM) model was used on the p-values to estimate the false discovery rate (FDR).
Quantitative reverse transcription PCR (qRT-PCR)
63
RNA was extracted by phenol chloroform isolation using TRIzol reagent
(Invitrogen) then purified using the MirVANA kit total RNA isolation procedure
(Ambion). Following isolation RNA was treated with Turbo DNase (Ambion). cDNA was prepared with SuperScript First-Strand Synthesis System (Invitrogen). Quantitative reverse transcription PCR (qRT-PCR) to validate the microarray was performed using
SybrGreen master mix (Applied Biosystems). Prognostic training and validation sets were run using TaqMan master mix (Applied Biosystems). Detection was performed using an ABI Prism 7700 detection system (Applied Biosystems). LncRNA expression was normalized to internal control genes TATA-box binding protein (TBP) or
Glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Custom-designed primers were used for all SybrGreen reactions. Custom TaqMan primers were used for treRNA and
ENST00000413901 TaqMan reactions, with TaqMan TBP primer # 4326322E
(ThermoFisher) used as the internal control. Custom designed primer sequences are provided in Table 9. ). DeltaCT was determined by subtracting the CT value of the housekeeping gene from the CT value of the lncRNA.
Nuclear and cytoplasmic RNA extraction
Nuclear and cytoplasmic lysate fractions were isolated from CLL patients’ cell pellets using the NE-PER kit (Thermo Scientific). RNA was then collected from the lysates using the miRVANA kit (Ambion) using the manufacturer protocol for cultured cells followed by total RNA extraction. Following isolation, RNA was treated with Turbo
DNase (Ambion). Standard PCR was performed for treRNA, U2, and S14 as previously described by Gummireddy et al204.
64
Retrovirus vectors and generation of treRNA expressing cell lines
A 589bp PCR product encoding the spliced transcript of treRNA was cloned into the EcoRI/Not I sites of pRetro-tight-puro (Clontech). pRetro-tight-puro vector without insert was used as a control. Retrovirus was produced by co-transfecting the plasmid
DNA of interest and ecotropic-helper plasmids (pVSV and pGPZ) into the 293T cell line using calcium phosphate precipitation. Inducible Tet activator OSUCLL cell line
(pRetrox-Tet-on) was established according to the manufacturer protocol (Clontech).
OSUCLL-pTet-on cells were then infected with retrovirus by culturing for 10 hours in the conditioned media with 8 μg/mL of polybrene. Cells were then washed to eliminate virus and polybrene and incubated in complete media for 48 h before selection with
1μg/ml of puromycin and 500ug/mL G418.
Viability
Viability was assessed by staining with Annexin V-FITC and propidium iodide
(PI). Data was collected on a Beckman Coulter FC500 flow cytometer then analyzed using Kaluza software.
Migration
Cells were suspended in RPMI at 5x106 cells/mL, and 100 µl was placed in the upper chamber of a 24-transwell plate with a 5µm filter. Chambers were placed into wells containing media containing no chemokine (control), recombinant human CXCL12
(200ng/mL, Millipore) or CXCL13 (1000ng/mL, R&D). Migration was permitted for 3 hours, and cells in the lower chamber were collected and counted for 20 seconds on high
65 speed on a Beckman Coulter FC500 flow cytometer. A 1/20 dilution of input cells was also determined.
Proliferation
Proliferation was measured using the Click-iT Plus EdU Alexa Fluor 488 Flow
Cytometry Assay Kit (Invitrogen) following the manufacturer protocol. Cells were incubated with 10uM EdU for 2 hours before staining.
Immunoblot
Whole cell lysates were prepared by lysing PBS-washed OSUCLL cell pellets in cold lysis buffer containing phosphatase inhibitor cocktail 1 and 2, protease inhibitor cocktail P8340 and 1mM phenylmethylsulfonyl fluoride (all from Sigma). Protein was quantified by the BCA method (Pierce). Protein (25ug/lane) was separated on 12% polyacrylamide gels and transferred onto nitrocellulose. After antibody incubations, proteins were detected with chemiluminescent substrate (Advansta) and quantified using a ChemiDoc system with Quantity One software (Bio-Rad Laboratories). The following antibodies were used for detection, anti-TP53, anti-ACTB (both Santa Cruz
Biotechnologies), and anti-γH2AX (Abcam).
Single Cell Gel Electrophoresis assay (SCGE, also known as comet assay)
Comet assay was performed using reagents from the Trevigen COMET assay kit.
Following a one hour drugging of OSUCLL pRetro and treRNA, samples were washed with ice cold PBS then combined with comet agarose at a 1:10 ratio. Samples were adhered for 30 minutes at 4oC then immersed in lysis buffer for 60 minutes at 4oC.
Samples were then submerged in alkaline unwinding buffer for 30 minutes at 4oC.
66
Samples were electrophoresed in alkaline running buffer, fixed with 70% ethanol, and then stained with SYBR gold (Fisher).
Statistical analysis
Linear mixed effects models were applied to the deltaCT values (relative to control: GAPDH or TBP), allowing for dependencies among observations from the same sample. From the models, which contained the interaction between target and cell type, differences in deltaCT values between CLL and normal B cells for each target were estimated, with 95% CI, and then converted to fold changes. To control the overall Type
I error at 0.05, Holm’s step-down procedure was used to adjust the p-values from each of the individual target comparisons. Analyses were performed using SAS/STAT software,
Version 9.4 of the SAS System for Windows (SAS Institute Inc.).
Linear mixed effects models were used for statistical analysis to compare the olive tail moment between cells with treRNA and empty vector under each treatment condition. Analyses were performed using SAS/STAT software, Version 9.4 of the SAS
System for Windows (SAS Institute Inc.).
Clinical endpoints were defined as follows: Time to treatment (TTT) was measured from the date of diagnosis until the date of first treatment, censoring patients who had not started treatment at last follow-up. PFS was defined as the time from randomization until documented disease progression or death without progression, censoring patients alive and progression-free at the date of last reported contact, and OS was defined as the time from randomization until date of death, censoring patients alive at last contact date. Associations between lncRNAs and TTT, PFS, or OS were initially
67 explored using Kaplan–Meier plots and differences between low and high expression groups were evaluated with the log-rank test. Multivariable models were fit using Cox proportional hazards models. Associations between treRNA with other clinical and molecular features were tested using the Wilcoxon rank sum or Fisher’s exact tests for continuous and categoric variables, respectively. All tests were two-sided and statistical significance was declared for p<0.05.
4.3 Results lncRNAs are differentially expressed in CLL
We first sought to explore if lncRNAs are aberrantly expressed in CLL by comparing pooled CLL patients’ RNA to pooled RNA from healthy donor B cells, with and without stimulation with CD40 ligand, using the ArrayStar human lncRNA microarray, a platform that analyzes over 30,000 lncRNA transcripts in addition to
30,000 coding transcripts. We found gene expression was highly dysregulated in CLL compared to normal B cells (Figure 8). 1051 probes, including both mRNA and lncRNA, were significantly different between CLL and normal B cells at a FDR=7%, of these 311 probes were annotated lncRNA. We selected 14 highly differentially expressed lncRNA to validate using qRT-PCR in an additional 14 CLL samples and 5 normal B cell samples
(independent of those used in the microarray). We were able to validate 8 lncRNAs as having significantly different expression when comparing CLL samples to normal B cell samples (Figure 9). We noticed that two of the validated lncRNAs, treRNA and
ENST00000413901, clustered based on the IGHV mutational status in the patient
68 samples; unmutated IGHV samples had higher expression of these lncRNAs (Figure 9).
This led us to investigate further the prognostic significance of these two lncRNAs in a large well-characterized cohort of patient samples.
High expression of treRNA is associated with shorter time to treatment
The initial set of CLL samples used to assess prognostic significance of treRNA and ENST00000413901 were obtained from 144 previously untreated asymptomatic CLL patients enrolled in the CRC registry. In this particular patient set, ZAP70 protein expression and methylation status had been previously described84. Expression of treRNA and ENST00000413901 were measured using qRT-PCR and normalized to the housekeeping gene TBP. We separated patients into high or low expression of treRNA and ENST00000413901 based on the median expression value. We found that high expression of treRNA associated with poor prognostic indicators: low (<20%) ZAP70 methylation, high (>20%) ZAP70 protein expression, and unmutated IGHV (Table 10).
High expression of treRNA also associated with shorter TTT (Figure 10). Interestingly, high expression of treRNA identified patients with shorter TTT in the favorable prognostic subgroups, low CD38 expression and low ZAP70 protein expression (Figure
10). Although LncRNA ENST00000413901 was also associated with poor prognostic indicators low ZAP70 methylation, high ZAP70 protein expression, and unmutated
IGHV, it was not as strongly associated with TTT and we therefore focused our remaining studies on treRNA (data not shown).
High expression of treRNA is an independent predictor of shorter progression free survival in patients receiving chemotherapy
69
While interrogating the clinical significance of treRNA we discovered that CLL cells contain a transcript in the nuclear RNA fraction that retains the intron between the two coding exons, likely due to insufficient splicing (Figure 11). Therefore we investigated the prognostic significance of both spliced and retained intron treRNA (ri- treRNA) in our validation set. To validate the associations we identified in the CRC dataset we used 147 pretreatment samples from symptomatic previously untreated patients enrolled in the Eastern Cooperative Oncology Group E2997 (E2997) phase III clinical trial comparing the nucleoside analog fludarabine to the combination of fludarabine plus the alkylating agent cyclophosphamide 205. In this setting, we confirmed the association between high expressers of treRNA and IGHV mutational status and identified an additional association with cytogenetic markers; high treRNA expression occurred more frequently in patients with deletion 17p, and less frequently with trisomy
12 (Table 11). High treRNA was also associated with shorter PFS and OS (Figure 12).
Even though ri-treRNA expression correlates with treRNA expression (Figure 13), these associations were not significant for the unspliced transcript (Table 12). This could be related to the fact that patients express considerably lower levels of ri-treRNA compared to spliced treRNA and have a smaller dynamic range of expression. We next included treRNA expression in a multivariable model previously described155. TreRNA expression did not add prognostic information for OS in the multivariable model once accounting for treatment arm, age, and molecular group. However, the impact of treRNA on PFS was significantly different depending on whether patients received fludarabine plus cyclophosphamide or fludarabine alone. In a model accounting for the effect modification
70 of treatment arm on the relationship between treRNA and PFS and adjusting for age, sex,
Rai stage, and molecular group, high treRNA expression was strongly associated with a higher risk of progression or death for those who received fludarabine plus cyclophosphamide (HR 3.14, 95% CI 1.61-6.14, p = 0.0008); for those who received fludarabine alone, PFS was short irrespective of treRNA expression (HR 1.12, 95% CI
0.62-2.02, p = 0.70) (Figure 12).
The presence of treRNA results in less DNA damage with fludarabine and mafosfamide
The inferior response to fludarabine plus cyclophosphamide in patients with high treRNA expression suggests a role for treRNA in mediating DNA damage response; therefore we established a stable retroviral system to further study this observation in vitro. We used the CLL cell line established in our lab (OSU-CLL)206 to exogenously express treRNA, as well as a control line transduced with the empty viral vector.
Expression of treRNA did not alter viability, proliferation, or migration (Figure 14). The treRNA cell line drugged with fludarabine or fludarabine combined with mafosfamide
(an active metabolite of cyclophosphamide) did not show significantly different viability compared to empty vector (Figure 15). However, we observed a trend towards less induction of the DNA damage indicator, γH2AX, when exposed to fludarabine or fludarabine plus mafosfamide, but not mafosfamide alone in OSU-CLL expressing treRNA, although this did not reach statistical significance (Figure 16) (quantitation
Figure 17).
Due to the inherent issues with quantitation of immunoblots207, we verified the differences in DNA damage using the more quantifiable comet assay. Following one hour
71 incubation with fludarabine, mafosfamide, or the combination, DNA damage was markedly less induced in all drugging conditions in OSUCLL treRNA compared to empty vector (Figure 18). These results indicate that the presence of treRNA results in decreased DNA damage in cells exposed to chemotherapeutic agents.
4.4 Discussion
Here we characterized the differential expression of lncRNA in CLL cells compared to normal B cells. While validating differentially expressed lncRNA identified by microarray, we narrowed our interest to two lncRNAs, treRNA and
ENST00000413901, due to their variable expression in CLL patients and association with
IGHV mutational status. Of these two, only treRNA showed clinical significance in the initial prognostic dataset. While we did not pursue further studies with
ENST00000413901, the clear separation of patients who highly express this lncRNA in contrast to patients with low to undetectable levels suggests that this lncRNA may reflect biologically distinct subgroups.
Our initial dataset indicated that patients with high expression of treRNA are enriched for aggressive disease markers and required treatment earlier than those with lower expression. These findings were supported in a second, independent dataset that confirmed the association between treRNA and the aggressive IGHV disease, and also identified that treRNA expression is associated with shorter PFS and OS. Interestingly, we found low expression of treRNA expression was an independent prognostic factor for improved PFS in patients receiving fludarabine plus cyclophosphamide. TreRNA
72 expression was not associated with PFS in patients receiving fludarabine alone, however, this arm of the study overall did very poorly. It would it be interesting to see if the prognostic significance of treRNA is specifically associated with therapy involving these
DNA damaging agents and if this association with PFS will persist for patients receiving fludarabine plus cyclophosphamide in combination with rituximab, as this is a standard treatment strategy in CLL34,35. Given that treRNA has been reported as expressed in breast cancer lymph-node metastases as well as colorectal cancer204, and that chemotherapeutic agents are also used for treatment of these types of cancer, it would also be of interest to explore if treRNA expression is predictive of chemotherapeutic response in these solid tumors.
The association with PFS of the combination treatment of fludarabine + cyclophosphamide suggests that treRNA may play a role in response to DNA damaging agents prompting further exploration of this. In vitro studies using the OSU-CLL cell line expressing treRNA showed a slight trend towards having less induction of the DNA damage marker γH2AX. We validated this trend using the comet assay, which more readily quantifies DNA repair, and found that OSU-CLL expressing treRNA had markedly reduced induction of DNA damage after one hour exposure to fludarabine, mafosfamide, and the combination compared to OSU-CLL empty vector. These results suggest that treRNA may be directly involved in decreasing DNA damage after exposure to chemotherapeutic agents and provides a basis for why this marker is independently associated with PFS in patients who receive these agents. However, the cell line model may be unable to fully recapitulate what is happening in the patient cells due to inherent
73 differences between cell lines and primary samples, the lack of microenvironment interactions in the in vitro setting, and this model does not address the role of treRNA with prolonged exposure to chemotherapeutic agents.
TreRNA has been reported to affect both transcription119 and translation204, however, the targets of treRNA may be cell type specific204. In normal B cells treRNA was low to undetectable; however, in CLL patients we found a wide range of expression.
This expression positively correlated with aggressive disease markers, as well as TCL1A and miR-155 (Figure 19) which are known oncogenic factors in B cells208,209, suggesting that treRNA expression is part of an aggressive B cell leukemia phenotype. However, the factors that drive the expression of treRNA, and functional effects of treRNA outside of
DNA damage response in CLL, remain to be elucidated. In primary patient samples we observed an upregulation of treRNA as CLL cells sat in culture, this upregulation could be abrogated by providing stimulation (Figure 20), suggesting treRNA may be regulated by a stress response.
LncRNAs have been reported to function as regulators of gene expression and can directly contribute to aggressive cancer phenotypes. Our results suggest lncRNAs are significantly deregulated in CLL compared to normal B cells, and can be used as prognostic indicators in this disease. Investigation into the function of these aberrantly expressed lncRNA may help us further understand CLL pathogenesis, and provide important insight into response to therapy.
74
Sybr green primer sequences
lncRNA Forward Reverse complement
ENST00000503154 CCTGGGACTTCACAACACCAGCCG CCACACTCTCTTTGGCCACTCCAGG
ENST00000456588 GGCACAGGCTGTACAGGAAGCG GTCATGTAAGACGTGCCTGCTCTCC
ENST00000413901 GCAGCGTCCAGAGGCTGG GGCAGCGAGGATGCTGAACC
ENST00000423967 CCACCCACCGCGTATACCTCTG CTGGCACATCTTACCGGTTTCTGCG treRNA CGTGGCCGATTTGAGAGAGTGAGAC CCAGGTCTGGGCAAAGAGAGGC
AK126772 GCTGGAGTGGCCTTGTCCTCATTC GCCAGTCTAGGTCTACCTGAGCCTC
AK000998 GGGTTCACGCCATTCTCCTGGATGAC GGGTTCAGGCACTGACCATGTGTTCC lncREL1.1 TTGTAATCCCAGCACTTCAGG AAAGAACTACGGCCAGACAC
Taqman primer sequences
lncRNA Forward Reverse Probe treRNA-CRC CGTGGCCGATTTGAGAG CCAGGTCTGGGCAAA GCTGTAGCCCTGGCAACCTCC
AGTGAGACC GAGAGGC ACTCCGCCTG treRNA-ECOG GGTGGTTTTACGTGGCC CCAGGTCTGGGCAAA GCTGTAGCCCTGGCAACCTCC
GATTTGAG GAGAGGC ACTCCGCCTG ri-treRNA GTGGCCAAAAGGGGACC CCAGGTCTGGGCAAA GCTGTAGCCCTGGCAACCTCC
GAAG GAGAGGC ACTCCGCCTG
ENST0000041 GCAGCGTCCAGAGGCTG GGCAGCGAGGATGC GGATGCTGCTGGCAGGTGCAC
3901 G TGAACC CACACTGAGTTGC
Table 9. Primers used for gene expression studies for lncRNA in CLL.
75
Low treRNA High treRNA Variable P-value (n=72) (n=72)
0.33 Median Age, yrs. (Range) 53 (32-78) 52 (26-82)
Female, Num. (%) 18 (25) 26 (36) 0.21 ZAP70 Methylation Low, Num. (%) 44 (61) 63 (88) <0.001 High (> 20%) 28 (39) 9 (13) ZAP70, Num. (%) Negative 45 (63) 26 (36) 0.003 Positive (>20%) 27 (38) 46 (64) CD38, Num. (%) Negative 39 (54) 31 (43) 0.24 Positive (>20%) 33 (46) 41 (57) IGHV, Num. (%) Mutated 30 (42) 8 (11) <0.0001 Unmutated (>98%) 42 (58) 64 (89) Time to Treatment N 64 65 0.04 Median, yrs. (95% CI) 3.9 (2.9-4.9) 2.5 (2.0-3.5) At 5 Yrs, (95% CI) 0.36 (0.25-0.48) 0.22 (0.12-0.33) Overall Survival N 72 72 0.17 Median, yrs. (95% CI) 12.1 (9.1-12.6) 7.9 (6.7-13.5) At 5 Yrs, (95% CI) 0.83 (0.72-0.90) 0.80 (0.67-0.89)
Table 10. Associations in the CRC patient set for treRNA. Associations were tested using the Wilcoxon rank sum and Fisher’s exact tests. Abbreviations: yrs, years; num, number; del, deletion.
76
Low High Variable treRNA treRNA P-value (n = 74) (n = 73)
Treatment Arm, Num. (%) Fludarabine (F) 37 (50) 38 (52) 0.87 F+Cyclophosphamide 37 (50) 35 (48)
Median Age, yrs. (Range) 62 (33-83) 60 (42-78) 0.34
Female, Num. (%) 20 (27) 22 (30) 0.72
0.35 Rai Stage II/III/IV, Num. (%) 51 (69) 56 (77)
IGHV Num. (%) Mutated 29 (49) 16 (23) 0.0029 Unmutated 30 (51) 53 (77) Unknown 15 4
ZAP70 Methylation Low, Num. (%) 32 (70) 45 (79) 0.36 High (> 20%) 14 (30) 12 (21) Unknown 28 16
Dohner Classification del(17p), Num. (%) 2 (3) 11 (15) del(11q) 8 (11) 11 (15) +12 25 (34) 12 (17) 0.018 del(6q) 1 (2) 4 (6) Normal 13 (18) 11 (15) del(13q) 24 (33) 22 (31) Unknown 1 2
Table 11. Associations in the ECOG 2997 patient set for treRNA. Associations were tested using the Wilcoxon rank sum and Fisher’s exact tests. Abbreviations: yrs, years; num, number; del, deletion.
77
Low ri- High ri- Variable treRNA treRNA P-value (n = 73) (n = 74) Treatment Arm, Num. (%) Fludarabine (F) 34 (47) 41 (55) 0.32 F+Cyclophosphamide 39 (53) 33 (45) Median Age, yrs. (Range) 60 (33-83) 61 (40-78) 0.22 Female, Num. (%) 18 (25) 24 (32) 0.36 Rai Stage II/III/IV, Num. (%) 55 (75) 52 (70) 0.58 IGHV Mutated, Num. (%) 27 (44) 18 (27) 0.07 Unknown 11 8 ZAP70 Meth > 20%, Num. (%) 14 (30) 12 (21) 0.36 Unknown 27 17 Cyto./Molec.Group, Num. (%) del(17p) 5 (8) 8 (12) del(11q) 9 (15) 10 (15) +12/Notch Mut 8 (13) 7 (11) 0.59 IgHV Unmutated 20 (32) 27 (41) Other 20 (32) 14 (21) Unknown 11 8 Dohner Classification del(17p) 5 (7) 8 (11) del(11q) 9 (12) 10 (14) +12 22 (30) 15 (21) 0.36 del(6q) 1 (1) 4 (6) Normal 15 (21) 9 (13) del(13q) 21 (29) 25 (35) Unknown 0 3
Table 12. Associations in the ECOG 2997 patient set for ri-treRNA. Associations were tested using the Wilcoxon rank sum and Fisher’s exact tests. Abbreviations: yrs, years; num, number; del, deletion, cyto, cytogenetic, molec, molecular.
78
P1
P1/CD40L
P2
NB
NB/CD40L
Figure 8. Heatmap of CLL vs NB lncRNA microarray. Microarray of two pools of CLL (P1 & P2), one stimulated with CD40L, compared to one pool of normal B cells (NB),
7 with and without CD40L stimulation.
9
79
blacnk 79
(a)
(b)
Figure 9. LncRNAs are aberrantly expressed in CLL. (a) Relative expression of lncRNAs in primary CLL cells compared to normal B cells measured by qRT-PCR. P-values were determined by linear mixed effects models. (b) Relative expression of treRNA and ENST00000413901 in IGHV unmutated (black) compared to IGHV mutated (red) primary CLL cells. Gray dots are samples with unknown mutational status. P-values were determined by two-sided t-test.
80
(a)
(b)
(c)
Figure 10. Kaplan-Meier curves by treRNA expression in the CRC training set. Time to treatment in the (a) overall population (b) CD38 negative patients (c) ZAP70 protein expression negative patients. High or low treRNA based on the population median. P-values were determined by log rank test.
81
(a)
(b)
Figure 11. CLL cells express retained intron treRNA. (a) RT-PCR for treRNA and ri-treRNA in normal B samples and primary CLL samples. PCR for the IL-6 promoter was performed as a control to detect genomic DNA contamination. (b) RT-PCR for treRNA, U2 (nuclear enrichment control), and S14 (cytoplasmic enrichment control) as previously described by Gummireddy et al198 in nuclear (NE) and cytoplasmic (CE) fractions in primary CLL patient (Pt) samples (n=3).
82
(a)
(b)
(c)
Figure 12. Kaplan-Meier curves by treRNA expression in the ECOG 2997 validation set. (a) Progression free survival, (b) overall survival, and (c) progression free survival by treatment arm. High or low treRNA based on the population median. P-values were determined by log rank test.
83
Figure 13. Ri-treRNA expression correlates with treRNA expression. Correlation between treRNA and treRNA retained intron expression measured by qRT- PCR in the ECOG 2997 patient samples.
84
Figure 14. TreRNA expression in OSUCLL does not alter viability, proliferation, or migration. (a) Viability of OSUCLL empty vector and OSUCLL treRNA by PI staining. (b) Proliferation of OSUCLL empty vector and OSUCLL treRNA by EdU incorporation. (c) Transwell migration towards media or chemokine (CXCL12 or CXCL13) by OSUCLL empty vector and OSUCLL treRNA.
85
Figure 15. TreRNA does not alter viability following treatment with fludarabine and mafosfamide. OSUCLL empty vector and OSUCLL treRNA drugged with fludarabine (F, 10uM), mafosfamide (M, 2.36uM), or M+F for 24 hours. Viability assessed by Annexin/PI staining.
86
Figure 16. TreRNA induces less γH2AX in the presence of fludarabine. Representative immunoblot of γH2AX induction in OSUCLL-treRNA versus empty vector cell lines following 24-hour treatment with vehicle (V, DMSO), fludarabine (F, 10uM), mafosfamide (M, 2.36uM) or a combination of fludarabine plus mafosfamide (F+M). Actin probed as a loading control.
87
Figure 17. γH2AX immunoblot quantitation. Blots were quantified using Quantity One, data was then log transformed and tested using linear models. Differences were not significant. Abbreviations: vehicle (V), fludarabine (F), mafosfamide (M) or a combination of fludarabine plus mafosfamide (M+F).
88
(a)
(b)
Figure 18. OSUCLL expressing treRNA has less DNA damage induced with chemotherapy. DNA damage was measured using the comet assay in 100 cells per condition following a 1 hour drugging with vehicle (control), fludarabine (F), mafosfamide (M), or the combination (F+M)(n=3). (a) Representative comet tail images. (b) Quantification of comet tails using olive tail moment.
89
Figure 19. TreRNA expression correlates with miR-155 and TCL1A. Expression was quantified from RNA sequencing data in CLL patients.
90
Figure 20. TreRNA is upregulated by primary CLL cells in culture and this upregulation can be abrogated by stimulation. Primary CLL samples (n=11) were cultured for 24 hours in media alone, with plate bound anti-IgM, or with CD40 ligand plus IL-4. TreRNA expression was measured using qRT-PCR. Significance was tested using mixed effects models.
91
CHAPTER 5: Discussion
5.1 Summary of work
The work presented here focuses on identifying novel biomarkers in chronic lymphocytic leukemia. In particular, we evaluated markers associated with therapeutic response. These markers are critical as they may identify patients who benefit from alternative therapeutic approaches or identify ‘at risk’ patients who may require more frequent monitoring for signs of progression. Chapter one provides background on chronic lymphocytic leukemia, highlighting the heterogeneous nature of this disease. We discuss the major therapeutic agents used in CLL and their basic mechanisms of action.
Biomarkers currently used in CLL are described along with their potential pitfalls. We provide in depth discussion on cytogenetics in CLL and cover the two major cytogenetic techniques utilized in CLL. Particularly important is that two cytogenetic markers, deletion 17p and complex karyotype, have been shown to be associated with poor response to therapeutics currently used in CLL. The clinical feasibly for analyzing cytogenetic markers provides rationale for further evaluation of novel cytogenetic biomarkers that can identify patients at risk for progression with specific therapies in
CLL. Finally, we discuss a new category of RNA, lncRNA, being evaluated as biomarkers in various types of cancer. In these cancers, identification of lncRNA biomarkers has led to important discoveries regarding the functions of lncRNA as well as
92 the biology of these cancers. Importantly, the role of lncRNA in CLL is presently poorly understood. From the knowledge summarized in chapter one, we provide a rationale for investigating cytogenetic and lncRNA biomarkers in CLL.
In chapter 2 we describe a novel chromosomal abnormality in CLL, the jumping translocation. Jumping translocations are when the same piece of chromosome, the donor, is found translocating to different partner chromosomes in different metaphases analyzed for a patient. This type of abnormality is only identifiable using metaphase karyotyping. Jumping translocations were identified in 3% of the patient population. Of particular interest, donor chromosome breakpoints frequently occurred at 17p11.2. This is in contrast to what has been described for other hematologic malignancies136. Also in contrast, JT recipient breakpoints in CLL commonly occurred in centromeric regions rather than telomeric regions. Jumping translocations contribute to complexity in CLL; all patients with these abnormalities had complex karyotypes. Additionally, JTs with breakpoints at 17p11.2 resulted in loss of 17p, which includes the tumor suppressor gene
TP53. These results suggest that JTs can be a mechanism of TP53 loss in CLL.
Chapter 3 focuses on interrogating cytogenetic abnormalities detectable by FISH that could predict for progression on ibrutinib. Patients progress on ibrutinib either with
CLL progression through the acquisition of a mutation in BTK or PLCG2 or through
Richter’s transformation, a transformation to an aggressive lymphoma. Richter’s transformation has been previously associated with a chromosomal gain of 2p and 2p gain has been classified as a late stage driver event in CLL 182,185. We interrogated if gain of 2p prior to starting ibrutinib would be predictive for progression. While there were an
93 increased proportion of patients who progressed with the 2p abnormality, it was not a statistically significant predictor for transformation or CLL progression. Interestingly, while examining patient samples for gain of 2p we noticed a number of patients had near- tetraploid cells. Tetraploidy is when a cell has four copies of each chromosome. This phenomenon is rarely reported in CLL188; however, it has been reported for various lymphomas186,187. Due to this, we decided to also analyze tetraploidy for associations with Richter’s transformation. In a univariate analysis tetraploidy was significantly associated with transformation, but not progression. All patients with tetraploidy had complex karyotype; however, this group had an increased cumulative incidence for transformation compared to complex karyotype alone.
Chapter four focused on lncRNA as a prognostic factor in CLL. Since little is known about lncRNA in CLL, we first identified eight lncRNA with aberrant expression in CLL compared to normal B cells. From there we narrowed down to two lncRNA,
ENST00000413901 and treRNA, to analyze further for prognostic significance. In the training set, treRNA was significantly associated with poor prognostic markers as well as a short TTT, while ENST00000413901 did not have statistically significant associations.
In the validation set we confirmed treRNA association with poor prognostic markers. In addition, we found high treRNA expression was an independent predictor for poor PFS in patients receiving F+C. We interrogated the effects of treRNA expression in a CLL cell line model by enforcing expression of treRNA using a retroviral plasmid. TreRNA expression resulted in decreased DNA damage, measured by γH2AX immunoblot and comet, following exposure to fludarabine and mafosfamide. This data supports that not
94 only is treRNA a biomarker for chemotherapy response, but may play a biological role in responding to DNA damage caused by these agents.
5.2 Future directions
Jumping translocations
In this work we found JTs were associated with the poor prognostic markers complex karyotype and deletion 17p. Due to the small number of patients, the variability in patient clinical history, and the variability in when during the course of the disease the abnormalities were detected, we were unable to assess if JTs are associated with a more aggressive disease in comparison to patients with deletion 17p and complex karyotypes without JTs. To do this analysis would require a large prospective study where metaphase karyotyping is performed at uniform time points.
JTs in CLL had unique biology in comparison to other hematologic malignancies.
In addition, we found significant differences within the JTs in CLL. Forty-eight percent of JTs in CLL had a donor breakpoint at 17p11.2. These translocations had recipient breakpoint locations predominately in centromeric regions (n=35, 69%), and frequently
(75%) resulted in the formation of dicentric or pseudodicentric chromosomes. This subgroup also had a frequent recurrent translocation, dic(17;18)(p11.2;p11.2), which occurred seven times. In contrast, the remainder of the donor chromosomes, grouped as miscellaneous, had recipient breakpoints predominately along the chromosome arm
(38%) and in the subtelomeric region (36%). In other hematologic malignancies the most frequently reported donor chromosome is 1q, recipient chromosome breakpoints are
95 predominately subtelomeric, and recipient breakpoints do not favor a particular chromosome136. These findings suggest that JT formation in CLL may differ in mechanism compared to other diseases.
Jumping translocations, as well as dicentric chromosomes, have been proposed to be associated with shortened telomeres210-213. Telomeres are DNA-protein complexes which contain short repetitive DNA sequences (TTAGGG) that protect and prevent the chromosome from being recognized as a double stranded break214. Telomere erosion occurs in CLL and has been associated with genetic complexity and high-risk genomic rearrangements215,216. In addition to shorten telomeres, a subset of CLL patients has mutations in the POT1 gene, a component of the shelterin complex, which provides protection to telomere ends217. These mutations result in loss of function and increased susceptibility to genomic rearrangements217. While telomeric fusions can result in breakpoints far from the telomeres, it seems unlikely, particularly for the 17p donor group, that this would be the only mechanism for JT formation. However, telomere instability may play a significant role with miscellaneous donor JTs, where recipient breakpoints favored telomeric and chromosome arm breakpoints. Telomere length as well as POT1 mutational status could be assessed in these patients to determine the likelihood of this mechanism contributing to these rearrangements.
Interesting studies have been performed in acute myeloid leukemia (AML) to understand dicentric chromosome formation which may be relevant to understanding JTs in CLL. Breakpoint mapping for some translocations in AML indicate a typical translocation with loss of the acentric chromosome product, other translocations had
96 interstitial deletions with preservation of the more telomeric sequences suggestive of telomeric fusions either preceding or followed by the interstitial deletions218. Further characterization of CLL JTs by breakpoint mapping with bacterial artificial chromosome probes as well as subtelomeric and telomeric probes would begin to uncover evidence for mechanistically how these aberrations form.
2p gain and tetraploidy
There was a trend for increased incidence of progression in patients with gain of
2p; however, this did not reach statistical significance. Re-analysis after a longer follow- up period would be useful to determine if 2p gain may predict for a late progression event. Our FISH probe targeted the gene REL; however, some groups have suggested that gain of the MYCN, telomeric to the REL locus, is prognostically relevant178,182. Further characterization of the 2p positive patients in our data set could be performed using a
FISH probe targeted to MYCN, or for a more precise delineation of the region gained single nucleotide polymorphism-arrays could be used. The consequences of 2p gain are not fully understood. Transcriptional profiling of 12 cases using microarray reported no changes in gene expression for many notable genes in the region, including REL and
BCL11A180. Our data set may be useful to characterize the transcriptional profile of patients with gain of 2p; we have a greater number of 2p positive patients and could utilize RNA-seq to avoid biases associated with microarray.
We identified tetraploidy as significantly associated with progression through transformation on ibrutinib. As a biomarker it will be important to confirm this association in a second independent data set. Little is known about tetraploidy in CLL. It
97 would be useful to examine the frequency of this abnormality as well as its clinical associations with Richter’s transformation in other settings, such as with other therapies, at the time of first treatment, and at diagnosis.
Tetraploidy in cancer cells can promote chromosome instability and may provide a fitness advantage to the cells 189. Tetraploidy can occur by several mechanisms. These cells can arise through cell fusion, through mitotic slippage, or through failure to undergo cytokinesis219. Mitotic slippage is when cells are arrested at the spindle assembly checkpoint (SAC) and then exit mitosis without going through anaphase. Mitotic slippage can occur in cells due to altered expression of proteins regulating the SAC219.
Defects in the retinoblastoma pathway (RB) can result in aberrant expression of SAC and cytokinesis proteins189,220. A member of the RB family, RB1, is encompassed in the deleted chromosomal material for some CLL patients with deletion 13q221. In CLL and lymphoma cell lines overexpression of the chromatin modifier PRMT5 silences expression of the RB family222. Alternatively, overexpression of MYC combined with loss of TP53 in a human myeloid cell line results in increased tetraploidization which may be due to mitotic slippage223. Tetraploidy in our patient set was associated with deletion 17p and four patients had MYC abnormalities.
Tetraploidy has also been attributed to telomere dysfunction. Telomere dysfunction induced by knockout of Pot1a/b in p53-deficeint mouse embryonic fibroblasts resulted in increased level of tetraploidy224. TP53 plus RB deficient human fibroblasts result in telomere crisis and increased tetraploidization due to both mitotic
98 slippage and cytokinesis failure225. As discussed in the previous section, telomere crisis and POT1 mutations have been described in CLL216,217,226.
To investigate the possible contribution of these pathways to tetraploidy in our patient population it would be useful to perform RNA as well as whole-exome sequencing on the tetraploid Richter’s transformation patient samples in comparison to patients who transform without signs of tetraploidy as well as patients who have remained on treatment. Gene set enrichment analysis can then be performed for these comparisons. We can also investigate telomere length in these samples using telomere
FISH flow cytometry227,228. treRNA
TreRNA expression was found to independently predict for decreased PFS with the combination of F+C. As a biomarker for CLL it would be important to assess the predictive power of treRNA in more relevant therapies, such as F+C in combination with rituximab or with chlorambucil or bendamustine in combination with anti-CD20 antibodies. In our study, treRNA expression was only measured in previously untreated patients, it is not known if exposure to chemotherapy could select for a clone with high expression of treRNA. TreRNA as a prognostic marker in previously treated patients would be interesting to explore. Caution must be used when measuring and comparing treRNA expression between patients. We found that treRNA expression is upregulated in patient samples over time in culture. Additionally, this upregulation could be abrogated if the cells were provided stimulation. These results suggest that treRNA expression may change if samples are not processed similarly and that expression of treRNA may vary in
99
B cells from different cellular compartments. We could evaluate this by measuring treRNA expression in paired samples collected from bone marrow and peripheral blood.
In addition to looking at treRNA in CLL, it would be interesting to evaluate the prognostic potential of this RNA in other cancers. TreRNA has been reported as expressed in breast cancer metastases as well as colorectal cancer204. As these two cancers are also treated with chemotherapeutic agents, treRNA could be evaluated for predicting response the DNA damaging agents in these tissue types. However, treRNA expression may be technically challenging to assess routinely. As an RNA molecule, expression is typically quantitatively measured using qRT-PCR. Accurate assessment of treRNA expression may be susceptible to contamination from other cell types. Ensuring pure populations of CLL cells, or other tumor tissue types, will increase cost and be more labor intensive.
Expression of treRNA in OSUCLL resulted in decreased γH2AX, a marker for
DNA breaks229, as well as decreased accumulation of DNA damage measured via the comet assay. It remains to be elucidated mechanistically how treRNA is reducing DNA damage. Several resistance mechanisms to chemotherapeutic agents have been described including decreased uptake or increased cellular export, increased inactivation of the compound in the cell, enhanced DNA repair, or a decreased activity of mechanisms that drive cell death230. TreRNA was found to decrease DNA damage for two types of chemotherapeutic agents, a nucleoside analog and alkylating agent, indicating transport and drug inactivation are unlikely mechanisms for decreased DNA damage since the two compounds do not share common mechanisms22,230,231. However, these mechanisms
100 could be involved if multiple pathways are affected by treRNA. It is more probable that treRNA could be altering DNA damage repair in cells. It would be useful to determine the kinetics of DNA damage in the cell lines. A time course for DNA damage foci,
γH2AX and 53BP1, using fluorescence microscopy could reveal if OSUCLL treRNA cells initially have similar amounts of damage induced which is then repaired more quickly. This would also suggest if a particular type of break is preferentially repaired, as
53BP1 is preferentially involved at double stranded breaks which undergo non- homologous end joining repair232. To elucidate how treRNA is decreasing DNA damage, gene expression profiling and phospho-protein arrays following DNA damage in the
OSUCLL pRetro and treRNA expressing cell lines could identify potential pathways regulating the damage response. It would also be useful to identify nuclear and cytoplasmic proteins associated with treRNA using an RNA pulldown assay followed by mass spectrometry.
101
Appendix A: The effects of D-2-hydroxyglutarate on cells in the acute myeloid leukemia microenvironment.
Introduction
Acute myeloid leukemia (AML) is leukemia of the precursor cells committed to the myeloid lineage. AML is diagnosed by the presence of ≥20% blast forms from the myeloid lineage in the bone marrow. AML is highly heterogeneous disease. Certain subtypes of AML are characterized by a chromosomal structural abnormality, such as core binding factor AML, PML-RARA, and MLL rearrangements. However, approximately 40-50% of AML is cytogenetically normal233,234. In this subset, recurrent gene mutations have been identified as important drivers of the disease. Progression to
AML may require multiple aberrations to occur and it is hypothesized that two types of pathways must be altered, cell proliferation and/or survival (type I mutations) and cell differentiation (type II mutations)235,236. A third class of mutations (type III mutations) has also been proposed, mutations that result in altered epigenetic patterns in the cell237.
Mutations in the proteins isocitrate dehydrogenase (IDH) 1 and 2 occur in approximately
10-30% of cytogenetically normal AML have been shown experimentally to impair hematopoietic differentiation consistent with a type II mutation238-241 as well as alter the epigenetic profile of the cell consistent with a type III mutation237.
102
The IDH class of proteins includes IDH1, IDH2, and IDH3. IDH enzymes, canonically IDH3, are involved in the tricarboxylic acid cycle (TCA). IDH1 and IDH2 function as homodimers to catalyze the reversible oxidative decarboxylation of isocitrate to alpha-ketoglutarate (α-KG) via the reduction of NADP+ to NADPH. IDH1 activity occurs in the cytoplasm and IDH2 in the mitochondria. In contrast, IDH3 functions as a heterodimer in the mitochondria, is NAD+ dependent, and structurally distinct from
IDH1 and IDH2242,243 (Figure 21).
IDH1 and IDH2 are recurrently mutated in AML as well as other cancer types such as glioblastoma, chondrosarcoma, cholangiocarcinoma, and angioimmunoblastic T cell lymphoma243. IDH1 mutations occur predominately at the Arg132 codon. IDH2 mutations occur at the Arg140 or Arg172 codons. These residues are in the active sites of the proteins where isocitrate binds242. Mutations in the active site reduce the ability of
IDH1/2 to convert isocitrate to α-KG242. These mutations are gain of function in that they cause the enzymes to favor a reaction that converts α-KG to D-2-hydroxyglutarate
(D2HG)244 (Figure 22).
D2HG has been demonstrated to be an oncometabolite chiefly through competitively inhibiting α-KG dependent reactions242,245 (Figure 22). The class of proteins, α-KG dependent dioxygenases, requires α-KG as a substrate to function. These dioxygenases include epigenetic modifiers such as the ten-eleven translocation (TET) family which catalyze the conversion of 5’methylcytosine to hydroxymethylcytosine, a key step in the reversal of DNA methylation246. Jumonji histone demethylases are also dependent on α-KG. D2HG interference with these proteins promotes a hypermethylated
103 phenotype240,245. Other proteins dependent on α-KG include the prolyl hydroxylase proteins (PHDs) which regulate hypoxia-inducible transcription factors 1/2α (HIF1/2A) and the hypoxic response243. D2HG has been reported to both induce and diminish hypoxic response; this discrepancy is currently unresolved245,247. Collagen folding is also
α-KG dependent and misfolded collagen can result in increased levels of ER stress248.
D2HG can function as a biomarker in AML. D2HG accumulates to high concentrations in the serum, urine, and other body fluids of patients with IDH mutations249. D2HG levels can predict for the presence of IDH mutations250,251. Levels of
D2HG fluctuate with therapy and can predict for therapy resistance and for minimal residual disease249-251. What is not known is if the presence of the oncometabolite D2HG in body fluids can affect cells of the microenvironment. We sought to determine if extracellular D2HG produced from AML cells can enter into other cell types and if so, does it produce alterations in the phenotype of those cells.
Methods
Sample processing and culturing
Blood was obtained from AML patients with written informed consent in accordance with the Declaration of Helsinki and under a protocol approved by the
Institutional Review Board of the Ohio State University. Cryopreserved AML cells, previously isolated using ficoll density gradient centrifugation, were thawed from cryovials and cultured in StemSpan SFEM (Stemcell Technologies) containing 10% human serum with or without 1x CC100 cytokine cocktail (Stemcell Technologies).
104
AML cell lines were cultured in RPMI 1640 with 10% FBS supplemented with penicillin, streptomycin, and L-glutamine. T cells were isolated from normal donors using ficoll density gradient centrifugation (Ficoll-Plaque Plus; Amersham Biosciences) and enriched for T cells using the Rosette-Sep negative selection kit (StemCell Technologies) according to the manufacturer’s protocol. T cell purity was confirmed via flow cytometry using antibodies CD3-FITC (BD biosciences), CD45-PE (BD biosciences) and Near-IR
Live/Dead (Life Technologies). After isolation T cells were cultured in StemSpan SFEM or in filtered StemSpan SFEM supernatant removed from primary AML samples kept in culture for 96 hours. HS5 cells were cultured in StemSpan SFEM supernatant removed from primary AML samples or in StemSpan SFEM + 10% human serum.
Mass spectrometry
The mass spectrometry (MS) assay to measure D2HG was developed and performed by the OSUCC Pharmacoanalytical Shared Resource laboratory. D-2-
Hydroxyglutarate powder or L-2-Hydroxyglutarate powder (Sigma) was dissolved in water to get a 5mM stock solution (free base). Calibration intermediate working solutions were created by serial dilution from stock solution aliquots using 50% MeOH.
Calibration intermediate working solutions ranged in concentration from 1.00 to 500uM for both D and L-2-Hydroxyglutarate. Quality control (QC) intermediate working solutions were prepared in the same way from stock solution aliquot with concentrations at 3.00, 15.0 and 150uM. 2-Hydroxyglutaric Acid-d3 powder (Santa Cruz) was dissolved in water to get a 1.00 mg/mL stock solution (free base). 4000ng/mL internal standard (IS) working solution was diluted from stock solution aliquot with 50% MeOH.
105
For calibration standard and QC samples, 10uL IS, 10uL calibration or QC working solution and 100uL blank media solution were mixed in microcentrifuge tubes.
For cell pellets samples, 100uL water was added and mixed well, then 10uL IS and 10uL
50% MeOH were added in the tube and mixed. For supernatant samples, 100uL supernatant samples, 10uL IS and 10uL 50% MeOH were mixed in microcentrifuge tubes. After protein precipitation, the supernatant was transferred to a new tube for evaporation under a gentle stream of nitrogen. The derivative was prepared by treating the dry residue with 100uL of freshly made 50 g/L DATAN in dichloromethane-acetic acid (4:1 by volume). The vial was capped and heated at 75 ºC for 30 min. After the tube was cooled to room temperature, the mixture was evaporated to dryness under a gentle stream of nitrogen.
Dried sample residues were reconstituted in 200uL reconstitution solution. The reconstitutions were centrifuged (4ºC) at 13,500 rpm for 10 minutes and 30 uL was injected for liquid chromatography mass spectrometry/mass spectrometry (LC-MS/MS) analysis. The LC system used was the Thermo Dionex Ultimate 3000 RSLC system.
Retention times for D2HG and L2HG were 5.1 and 5.7 minutes, respectively. The MS system used in this analysis was the TSQ Quantum Ultra EMR mass spectrometer. The calibration curve was linear in the range of 0.1-50uM with a lower limit of quantification of 0.1uM for both D and L2HG in cell media. The correlation coefficient (r2) was consistently 0.99 or better. The reference standard samples and QC samples also meet the acceptance criteria.
106
Viability
Cells were stained with PI then viability was measured using a Tali image based cytometer (Thermo Fisher).
Viable cells/ml
HS5 cells were plated with 50,000 cells per well in a 24-well plate and adhered overnight. The following day cells were drugged with 100nM, 1uM, or 10uM octyl-
D2HG (Cayman Chemicals). Viable cells per mL were counted 72 hours and 7 days later using a Tali cytometer and PI staining.
Proliferation
After isolation, CD3+ T cells were re-suspended in pre-warmed PBS at a final concentration of 1x106 cells/mL. 1uM carboxyfluorescein succinimidyl ester (CFSE,
ThermoFisher) was added, mixed well, and incubated for 10 minutes in a 37ºC cell culture incubator. Staining was quenched by adding 5 volumes of ice-cold culture media to the cells then incubating on ice for 5 minutes. Cells were pelleted and washed three times with media then re-suspended in fresh media. Cells were plated with or without plate bound anti-CD3 (10ug/mL, eBioscience) and soluble anti-CD28 (1ug/mL, eBioscience). Stimulated and unstimulated conditions were either untreated or treated with 200uM D2HG. Flow cytometry for Near-IR Live/Dead (Life Technologies) and
CFSE was run 6 days after stimulation. qRT-PCR
RNA was extracted by phenol chloroform isolation using TRIzol reagent
(Invitrogen) then purified using the MirVANA kit total RNA isolation procedure
107
(Ambion). cDNA was prepared with SuperScript First-Strand Synthesis System
(Invitrogen). qRT-PCR was performed using TaqMan primers for XBP1, ERN1,
HSPA5, GAPDH, and TaqMan master mix (Applied Biosystems). Detection was performed using an ABI Prism 7700 detection system (Applied Biosystems). Differences in gene expression were analyzed using the ddCT method.
T cell activation
2x106 T cells per condition were stimulated with plate bound anti-CD3 and soluble anti-CD28 while cultured in conditioned StemSpan SFEM media without cytokines collected from primary AML samples after 96 hours of incubation. Media was filtered after removal from AML samples prior to incubating T cells. T cells were incubated for 72 hours then collected and stained with Near IR-Live/Dead (Life
Technologies) and the activation marker CD69-PE (Beckman Coulter) then ran on a
Gallios flow cytometer (Beckman Coulter). Results were analyzed using Kaluza software.
3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H- tetrazolium (MTS) assay
5,000 HS5 cells per well were plated in a 96 well plate in StemSpan + 10% human serum and allowed to adhere overnight. Media was pipetted off the next day and
100uL of fresh blank media or media containing 100nM, 1uM, or 10uM of octyl-D2HG was added. After 72 hours or 7 day incubations MTS reagent was added and plates were incubated for an additional 4-6 hours. Plates were read by spectrophotometry at 492nm in a Labsystems 96-well plate reader (Fisher Scientific).
108
PD-1 expression
Cryopreserved AML samples were thawed in pre-warmed RPMI. 2x106 cells per condition were first stained with Violet Live/Dead (Life Technologies) stain at room temperature for 20 minutes in PBS. Cells were washed, then stained for CD3, CD4, CD8
(all BD biosciences), and PD-1 or isotype control (Biolegend) for 20 minutes on ice in flow buffer. Cells were washed, re-suspended in flow buffer, then ran on a Beckman
Coulter Gallios flow cytometer. Data was analyzed using Kaluza software.
Results
Validation of the MS assay
To validate the quantification of the MS assay, various concentrations of synthetic
D2HG were spiked into media then sent to blinded MS operators for processing and running the samples. MS accurately discriminated spike-ins of 10, 30, and 100uM (Table
13). We further validated the assay’s ability to identify mutant AML samples by measuring D2HG intracellularly as well as D2HG secreted in the cell culture supernatant.
IDH wildtype AML accumulated 7ng/2 x106 cells D2HG intracellularly and 0.117uM
D2HG in the supernatant. In contrast, mutant IDH1/2 patient samples accumulated between ~300-2600ng/2 x106 cells D2HG intracellularly and approximately 10uM in cell culture supernatant (Table 14). Additionally, we tested AML cell lines transfected with retroviral plasmids expressing mutant IDH1 or IDH2, as well as wildtype IDH, or empty vector. Despite IDH mutant cell lines growing poorly and having low expression of the mutant protein (data not shown), we were able to detect accumulation of D2HG
109 intracellularly compared to wildtype IDH or empty vector expressing cells (Table 15).
The accumulation of D2HG intracellularly and extracellularly was lower than what was seen in primary AML samples and expression of the mutant construct resulted in poor cell viability, thus these models were not utilized further. Finally, we tested the stability of D2HG produced from mutant AML cells in culture over time. Supernatant was removed from AML cells and kept in cell culture incubators for 7 days; D2HG remained stable out to 7 days (Figure 23).
Detecting intracellular D2HG in microenvironment cells
We next sought to investigate if D2HG secreted in the supernatant of AML cell cultures was able to enter into other cells. Cell membranes have been reported to be impermeable to synthetic D2HG245, however, this does not address if a tumor cell producing D2HG may have transport mechanisms allowing transfer of the molecule to other cell types. To determine if D2HG produced from AML cells could enter into other cell types we cultured primary AML samples for 96 hours to enrich the media with
D2HG. The supernatant was then removed from the AML cells, filtered, and then used to culture other cell types. We cultured normal donor T cells in fresh media, supernatant from wildtype AML or supernatant from IDH mutant AML in the presence or absence of
T cell stimuli, plate-bound anti-CD3 and soluble anti-CD28. After 72 hours cells and supernatant were collected for MS analysis. We found low levels, 2.5-4.4ng/2 x106 cells, of D2HG accumulation only in activated T cells incubated with mutant IDH2 supernatant
(Table 16). To determine if this accumulation could occur with synthetic D2HG we cultured activated T cells with increasing concentrations of synthetic D2HG for 72 hours.
110
Intracellular D2HG, ~2.2ng/2 x106 cells, was detected only when 100uM D2HG was present (Table 17). In contrast, this amount of D2HG accumulation in activated T cells occurred with mutant AML supernatant when only ~10uM D2HG present (Table 16).
In additional to normal T cells we also examined if D2HG could accumulate in stromal cells. We cultured the HS5 human fibroblast stromal cell line for 72 hours and 7 days in supernatant from wild type and IDH mutant AML patients. A slightly higher amount of D2HG was detected in HS5 cells cultured with supernatant containing D2HG, however, D2HG was detected in stromal cells cultured with wildtype AML supernatant as well as media alone (Table 18). Low concentrations of D2HG have been detected in cells without IDH mutations250.
Finally, we examined if D2HG secreted from a primary IDH mutant AML could enter into primary wildtype AML cells. After 72 hours incubation in cell culture supernatant from wildtype or IDH mutant AML samples no differences were observed in intracellular D2HG accumulation in the primary wildtype AML sample (Table 19).
D2HG does not alter HS5 phenotype
To determine if the small concentrations of D2HG detected in stromal cells could have a physiological effect over time we drugged HS5 cells with a cell permeabilized
D2HG, ocytl-D2HG. HS5 cells were incubated with 100nM, 1uM, and 10uM octyl-
D2HG for 7 days. No differences were observed in mitochondrial activity, viable cells per mL, or viability between vehicle controls and octyl-D2HG conditions (Figure 24).
We also examined if incubation with D2HG could induce ER stress due to misfolded collagen proteins, a downstream effect of D2HG accumulation248. Gene expression of ER
111 stress markers, XBP1, ERN1, and HSPA5 were not elevated in HS5 cells incubated with
D2HG containing supernatants (Figure 25).
D2HG does not alter normal T cell phenotype
To determine if D2HG had effects on normal T cells we incubated activated T cells with synthetic D2HG or with supernatant from mutant IDH cell cultures. Activated
T cells incubated in up to 200uM synthetic D2HG for 6 days had comparable proliferation and viability to activated T cells in media alone (Figure 26). We also compared the levels of T cell activation by measuring upregulation of CD69 in the presence of D2HG. T cells were stimulated with anti-CD3 and anti-CD28 then cultured in AML supernatants or media alone for 72 hours. CD69 was upregulated beyond the media alone condition for all samples incubated in AML conditioned media. This upregulation showed a slightly higher trend for samples in conditioned media taken from
IDH mutant AML (Figure 27). However, when we stimulated T cells in media alone with or without 100uM synthetic D2HG no differences were observed in CD69 expression
(Figure 28), suggesting the differences seen with supernatant from primary AML samples is unrelated to the presence of D2HG.
We also examined expression of Programmed cell death protein 1 (PD-1) on the surface of T cells from AML patient samples to determine expression differed between
IDH mutant patient samples compared to wildtype AML samples. PD-1 in T cells from
AML patients has previously been shown to be silenced via DNA methylation 252,253.
Since D2HG interferes with TET proteins involved in DNA demethylation240,245, we hypothesized this may result in increased silencing of PD-1 in T cells from IDH mutant
112 patients. However, we did not observe a significant difference of surface expression of
PD-1 in IDH mutant versus wild type patients (Figure 29).
Discussion
Here we developed a MS assay for quantitative measurement of D2HG. Our assay was able to accurately measure spiked in D2HG, as well as detect elevated concentrations of the metabolite in samples from IDH mutant patients and cell lines. In comparison to what has previously been reported our concentrations of D2HG in primary AML samples were lower. Mutant AML cells cultured for 14 days have been reported to accumulate between 1,529-19,247 ng/2 x106 cells of D2HG 254, our cells accumulated between 302-
2656 ng/2 x106 cells of D2HG. However, our cells were cultured for only 4 days. In serum, the median concentration of D2HG reported is 21.2uM 251. Our mutant AML samples accumulated less extracellular D2HG, ~12uM in cell culture supernatant.
We sought to determine if the high concentrations of D2HG which accumulates in the body fluids of patients with IDH mutants was capable of entering into other cells types and resulting in changes in phenotype. We detected small quantities of D2HG in cell pellets from activated normal T cells as well as HS5 stromal cells. D2HG was below the limits of detection for resting normal T cells, and was not elevated in wildtype primary AML sample incubated with D2HG containing supernatant. Cell pellets were washed three times in PBS prior to D2HG extraction, however, we cannot conclude with certainty that the D2HG measured in activated T cells and HS5 cells is truly intracellular, as it is possible that D2HG may adhere to the cell membrane.
113
Despite detectable levels of D2HG in the cell pellets we observed no overt changes in phenotype in T cells or HS5 cells. D2HG may need to accumulate to higher concentrations before having an effect. In tumor cells D2HG accumulates to very high concentrations, and it has been reported that interference of TET activity in an in vitro enzymatic assay required millimolar concentrations of D2HG 245. Our experiments were performed in culture, and may not accurately reflect an in vivo situation. Additionally, our experiments had a significant limitation in that primary AML samples viability in cell culture was highly variable, thus limiting our ability to interpret phenotypic differences between cells incubated in media from wildtype or IDH mutant AML samples. Another limitation imposed by the primary cell viability is that long term studies are not possible.
In myeloid mouse models with mutant IDH, phenotypic changes can take weeks to months to manifest in the myeloid cells241,255. Our inability to detect phenotypic changes due to D2HG exposure in cells present in the microenvironment may be because prolonged exposure is required.
To overcome these challenges cells could be isolated from freshly collected AML samples. For example, T cells from peripheral blood could be sorted from AML patients then measured for the presence of D2HG. These cells could also be used to look for differences between T cells from IDH wildtype versus IDH mutant patients that may be attributable to D2HG, such as changes in DNA methylation patterns, levels of histone markers, or increased expression of HIF1A target genes. Another way to study the effects of D2HG on the microenvironment is to utilize an IDH mutant mouse model. Sasaki et al. have described a conditional knock-in mouse model of IDH1-R132H in the myeloid
114 lineage of cells241. While expression of mutant IDH is not sufficient to cause leukemia, they did observe in older animals a change in hematopoietic progenitors and alterations in
T cell number in the spleen and bone marrow. D2HG is detectable in these mice and the downstream effects on DNA and histone methylation are observed in the myeloid cells.
This system, or a similar model, could be used to isolate T cells to detect intracellular
D2HG and look for changes in methylation, gene expression, or T cell function in comparison to wild type mice.
Effects of D2HG on the cells in the microenvironment may be dependent on the presence of leukemia cells. IDH mutant mice crossed with FMS-like tyrosine kinase 3 internal tandem duplication (Flt3-ITD) mice results in AML256,257, and could be utilized to study changes in cells of the microenvironment. One downfall with this experiment is that the experimental control mice, Flt3-ITD alone, typically develop myeloproliferative neoplasms and not AML256. To compare the effects of IDH mutations on an AML background, T cell characterization and function could be examined in compound mutant
AML model, such as Ftl3-ITD/NPM1 258, which is then crossed with mutant IDH1241 or
IDH2257. Intracellular D2HG could be measured in T cells from these mice. These experiments would indicate if the presence of an IDH mutation alters T cell phenotype and if D2HG can be detected intracellular. However, this does not address if the changes in the T cells are a direct result of the presence of D2HG.
If differences are observed in T cells from IDH mutated animals it will be important to determine if these alterations are due to D2HG uptake or if the IDH mutant leukemic cells induce these changes through other mechanisms. One possibility is to
115 drug an AML mouse model without mutant IDH with D2HG to determine if similar changes in T cell phenotype are observed. Alternatively, a conditional D-2- hydroxyglutarate dehydrogenase (the enzyme that converts D2HG back to α-KG) knockout mouse would result in increased levels of D2HG independent of IDH. The knockout could be myeloid specific and crossed with an AML leukemia model to observe if changes in T cells are specific to a downstream effect of D2HG. This knockout could also be made specifically in T cells to characterize the effects attributable to intracellular
D2HG in T cells. While the experiments described above focus on T cells, these approaches could also be used to study other cells of the microenvironment. These types of experiments would address if the effects seen in microenvironment cells are due to uptake of extracellular D2HG within those cells or if the changes to the microenvironment are driven by the alternative mechanisms induced by leukemic cells.
116
Measured (uM, average of Spiked in (uM) two measurements) 10 12.1475 30 30.326 100 109.7325
Table 13. Measured values of synthetic D2HG culture spike-ins. Mass spectrometry measurements of D2HG spiked in AML media.
117
IDH status Supernatant D2HG Intracellular (uM) D2HG (ng/2x106 cells) WT 0.117 7.11 IDH1 mutant 10.608 567.5 IDH2 mutant 1 9.699 2656.3 IDH2 mutant 2 12.549 302.2
Table 14. D2HG measurements in cultured primary AML samples. AML samples with IDH mutations accumulate D2HG intracellularly and in the supernatant. Samples were cultured for 96 hours. Intracellular measurements were generated by extracting D2HG from approximately 2x106 cells.
118
Supernatant D2HG Intracellular D2HG Cell line (uM) (ng/2x106 cells) MV411 parental 0.642 2.3 MV411 parental dox 0.595 1.76 MV411-IDH1 0.602 2.77 MV411-IDH1 dox 0.594 2.46 MV411-IDH2 R140Q 0.619 3.38 MV411-IDH2 R140Q dox 0.695 80.3
Supernatant D2HG Intracellular D2HG Cell line (uM) (ng/2x106 cells) THP1-vector 48h 0.692 6.26 THP1-vector 96h 0.751 7.57 THP1-IDHI 48h 0.706 4.46 THP1-IDHI 96h 0.854 7.99 THP1-IDH1 R132H 48h 2.21 267.4 THP1-IDH1 R132H 96h 6.4 576.3
Table 15. D2HG accumulates in AML cell lines expressing IDH mutant constructs. (A) MV411 transfected with stable doxycycline inducible retroviral plasmid containing wildtype IDH1 or mutant IDH2-R140Q. Parental cells were used as a negative control. Cells were either un-induced or induced with doxycycline (dox) for 48 hours then supernatant and pellet collected for MS. (B) THP1 cell line was transfected with stable constitutively expressing plasmid containing wilt-type IDH1 or mutant R132H IDH1. Empty vector was used as a negative control. Cell pellets and supernatant were collected 48 and 96 hours after cells were split in fresh media.
119
Unstimulated Stimulated
Supernatant Pellet Supernatant Pellet (ng/2x106 Sample (uM) (ng/2x106 cells) (uM) cells) D1 T cells + 11.418 BLQ 11.28 BLQ IDH1 mut sup D1 T cells + 11.755 BLQ 12.214 3.07 IDH2 mut 1 sup D1 T cells + 0.119 BLQ 0.121 BLQ WT sup D1 T cells media 0.044 BLQ 0.051 BLQ alone D2 T cells + 11.35 BLQ 10.768 BLQ IDH1 mut sup D2 T cells + 12.934 BLQ 12.784 4.42 IDH2 mut 1 sup D2 T cells + 12.868 BLQ 12.001 2.34 IDH2 mut 2 sup D2 T cells + 0.125 BLQ 0.136 BLQ WT sup D2 T cells media BLQ BLQ 0.048 BLQ alone D3 T cells + 12.333 BLQ 12.98 2.69 IDH2 mut 1 sup D3 T cells + 11.973 BLQ 11.95 2.46 IDH2 mut 2 sup D3 T cells + 0.084 BLQ 0.105 BLQ WT sup D3 T cells media BLQ BLQ BLQ BLQ alone
Table 16. D2HG is detectable in cell pellets following culture of activated T cells in supernatant from primary AML samples. Normal T cells, resting and activated, were incubated for 72 hours in supernatant taken from primary AML cells cultured for 96 hours. Abbreviations, D, donor, BLQ, below limit of quantitation, mut, mutant.
120
Synthetic D2HG Donor 1 Intracellular Donor 2 Intracellular (uM) in culture D2HG (ng/2x106 cells) D2HG (ng/2x106 cells) 10 BLQ BLQ 30 BLQ BLQ 100 2.31 2.19
Table 17. Synthetic D2HG is able to enter normal T cells when 100uM is present. To determine if synthetic D2HG can enter T cells, T cells were activated in the presence of 10, 30, and 100uM of synthetic D2HG for 72 hours. BLQ, below limit of quantitation.
121
D2HG in HS5 after 72 hr D2HG in HS5 after 7 day incubation incubation Intracellular Intracellular Supernatant Supernatant Media from (ng/2x106 (ng/2x106 (uM) (uM) cells) cells) IDH2 24.499 1.54 27.011 4.96 mutant 1 IDH2 11.138 BLQ 12.708 4.03 mutant 2
WT 0.117 BLQ 0.169 2.5
media 0.06 BLQ 0.085 2.73
Table 18. HS5 stromal cells may take up small amounts of D2HG. HS5 stromal cells were incubated in supernatant taken from primary AML samples in culture. D2HG in supernatant and cell pellet was measured by MS after 72 hours and 7 days. BLQ, below limit of quantitation
122
D2HG in IDH wildtype primary AML after 72 hour incubation
Media from Supernatant Intracellular (uM) (ng/2x106 cells) IDH2 mutant 1 22.663 4.96 IDH2 mutant 2 11.394 7.65 WT 0.22 7.03 media 0.151 3.23
Table 19. Wildtype AML sample did not accumulate D2HG. A wildtype AML samples was incubated for 72 hours in supernatant taken from primary AML cells cultured for 96 hours.
123
Figure 21. IDH family members’ roles in the TCA cycle. Diagram of the TCA cycle shows differences in IDH1, IDH2, and IDH3 function. Figure taken from Losman & Kaelin, 2013 242.
124
Figure 22. D2HG is a competitive inhibitor of α-ketoglutarate dependent dioxygenases. Figure taken from Yang et al., JCI 2013 259.
125
18 16 14
12
10
(uM) 8 IDH1- 6 R132H 4 IDH2_ D2HG concentration D2HG concentration 2 R140Q 0 Baseline 24 hours 48 hours 72 hours 7 days
Figure 23. D2HG remains stable in culture for 7 days. Primary AML cell with mutant IDH were cultured for 96 hours (baseline). The supernatant was removed from the cells and kept in a cell culture incubator for 7 days, samples were taken at 24, 48, 72, and 7 days to measure D2HG stability over time.
126
(a) MTS
1.5
1
0.5
0
relative DMSO to relative DMSO 100nM 1uM 10uM
% mitochondrial activity mitochondrial % Octyl-D2HG (n=2)
(b) Viable cells/mL 1.00E+07 1.00E+06 1.00E+05 1.00E+04 1.00E+03 1.00E+02
Viable cells/ml Viable 1.00E+01 1.00E+00 DMSO 100nM 1uM 10uM Octyl-D2HG (n=3)
(c) Viability
100 80 60 40 20 0
% viable (PI negative) viable%(PI DMSO 100nM 1uM 10uM Octyl-D2HG (n=3)
Figure 24. HS5 stromal cells incubated with octyl-D2HG for 7 days did not alter (a) mitochondrial activity, (b) proliferation, or (c) viability.
127
ER stress markers 16.000 14.000 12.000 10.000 8.000 6.000 XBP1 4.000 2.000 ERN1
0.000 HSPA5 Foldchange (relativemedia) to
Conditioned media
Figure 25. D2HG does not increase ER stress in HS5 cells. Fold change relative to media alone for three markers of ER stress. HS5 cells were incubated for 72 hours in conditioned media taken from primary AML samples with or without IDH mutations. Thapsigargin was used as a positive control.
128
(a)
(b)
Figure 26. Synthetic D2HG does not alter T cell proliferation or viability. Normal T cells stimulated (stim) with anti-CD3 and anti-CD28 were incubated in synthetic D2HG for 6 days then using flow cytometry (a) proliferation was measured by CFSE and (b) viability was measured by live/dead stain.
129
Figure 27. T cell activation is enhanced with supernatant from AML patients. T cells (4 donors) were stimulated with anti-CD3 and anti-CD28 while cultured in media taken from IDH mutant (n=4) or wildtype AML (n=3) patient cells, activation was measured 72 hours later. Activation was measured by CD69 upregulation.
130
Figure 28. Synthetic D2HG does not alter T cell activation. Normal donor T cells were stimulated with anti-CD3 and anti-CD28 while cultured in media with or without 100uM synthetic D2HG for 72 hours. Activation was measured by CD69 upregulation.
131
Figure 29. PD-1 expression does not differ between IDH wildtype and mutant AML. PD-1 expression was measured on CD4 and CD8 live T cell populations in patients with IDH mutant (n=4) or wildtype AML (n=6).
132
References
1. Howlader N, Noone AM, Krapcho M, et al. SEER cancer statistics review, 1975-2013. . Updated 2015. Accessed 04/19, 2016.
2. Hallek M, Cheson BD, Catovsky D, et al. Guidelines for the diagnosis and treatment of chronic lymphocytic leukemia: A report from the international workshop on chronic lymphocytic leukemia updating the national cancer institute-working group 1996 guidelines. Blood. 2008;111(12):5446-5456.
3. Matutes E, Owusu-Ankomah K, Morilla R, et al. The immunological profile of B-cell disorders and proposal of a scoring system for the diagnosis of CLL. Leukemia. 1994;8(10):1640-1645.
4. Kitada S, Andersen J, Akar S, et al. Expression of apoptosis-regulating proteins in chronic lymphocytic leukemia: Correlations with in vitro and in vivo chemoresponses. Blood. 1998;91(9):3379-89.
5. Danilov AV, Danilova OV, Klein AK, Huber BT. Molecular pathogenesis of chronic lymphocytic leukemia. Curr Mol Med. 2006;6(6):665-675.
6. Granziero L, Ghia P, Circosta P, et al. Survivin is expressed on CD40 stimulation and interfaces proliferation and apoptosis in B-cell chronic lymphocytic leukemia. Blood. 2001;97(9):2777-2783.
7. Burger JA, Tsukada N, Burger M, Zvaifler NJ, Dell'Aquila M, Kipps TJ. Blood- derived nurse-like cells protect chronic lymphocytic leukemia B cells from spontaneous apoptosis through stromal cell-derived factor-1. Blood. 2000;96(8):2655-2663.
8. Herishanu Y, Perez-Galan P, Liu D, et al. The lymph node microenvironment promotes B-cell receptor signaling, NF-kappaB activation, and tumor proliferation in chronic lymphocytic leukemia. Blood. 2011;117(2):563-574.
9. Burger JA. Nurture versus nature: The microenvironment in chronic lymphocytic leukemia. American Society of Hematology Education Program. 2011;2011:96-103.
133
10. Meinhardt G, Wendtner CM, Hallek M. Molecular pathogenesis of chronic lymphocytic leukemia: Factors and signaling pathways regulating cell growth and survival. J Mol Med. 1999;77(2):282-293.
11. Rai KR, Sawitsky A, Cronkite EP, Chanana AD, Levy RN, Pasternack BS. Clinical staging of chronic lymphocytic leukemia. Blood. 1975;46(2):219-234.
12. Binet JL, Auquier A, Dighiero G, et al. A new prognostic classification of chronic lymphocytic leukemia derived from a multivariate survival analysis. Cancer. 1981;48(1):198-206.
13. Oronsky BT, Reid T, Knox SJ, Scicinski JJ. The scarlet letter of alkylation: A mini review of selective alkylating agents. Transl Oncol. 2012;5(4):226-229.
14. Leoni LM, Bailey B, Reifert J, et al. Bendamustine (treanda) displays a distinct pattern of cytotoxicity and unique mechanistic features compared with other alkylating agents. Clin Cancer Res. 2008;14(1):309-317.
15. Zong WX, Ditsworth D, Bauer DE, Wang ZQ, Thompson CB. Alkylating DNA damage stimulates a regulated form of necrotic cell death. Genes Dev. 2004;18(11):1272- 1282.
16. Yi C, He C. DNA repair by reversal of DNA damage. Cold Spring Harb Perspect Biol. 2013;5(1):a012575.
17. Yamauchi T, Nowak BJ, Keating MJ, Plunkett W. DNA repair initiated in chronic lymphocytic leukemia lymphocytes by 4-hydroperoxycyclophosphamide is inhibited by fludarabine and clofarabine. Clin Cancer Res. 2001;7(11):3580-3589.
18. Rai, Kanti R.,Hollweg, Angelica,. Fludarabine versus chlorambucil: Is the debate over? Clinical Lymphoma Myeloma and Leukemia. 2011;11:S7-S9.
19. Eichhorst BF, Busch R, Stilgenbauer S, et al. First-line therapy with fludarabine compared with chlorambucil does not result in a major benefit for elderly patients with advanced chronic lymphocytic leukemia. Blood. 2009;114(16):3382-91.
20. Dillman RO, Mick R,McIntyre OR, et al. Pentostatin in chronic lymphocytic leukemia: A phase II trial of cancer and leukemia group B. J Clin Oncol. 1989;7(4):433- 8.
21. Tallman MS, Hakimian D, Zanzig C, Hogan DK, Rademaker A, Rose E,Variakojis D,. Cladribine in the treatment of relapsed or refractory chronic lymphocytic leukemia. J Clin Oncol. 1995;13(4):983-8.
134
22. Plunkett W, Saunders PP. Metabolism and action of purine nucleoside analogs. Pharmacol Ther. 1991;49(3):239-268.
23. Robertson LE, Keating MJ. Fludarabine phosphate in the treatment of chronic lymphocytic leukemia: Biology, clinical impact, and future directions. Cancer Treat Res. 1993;64:105-119.
24. Sorensen JM, Vena DA, Fallavollita A, Chun HG, Cheson BD. Treatment of refractory chronic lymphocytic leukemia with fludarabine phosphate via the group C protocol mechanism of the national cancer institute: Five-year follow-up report. J Clin Oncol. 1997;15(2):458-465.
25. Flinn IW, Neuberg DS, Grever MR, et al. Phase III trial of fludarabine plus cyclophosphamide compared with fludarabine for patients with previously untreated chronic lymphocytic leukemia: US intergroup trial E2997. J Clin Oncol. 2007;25(7):793- 798.
26. Rai KR, Peterson BL, Appelbaum FR, Kolitz J, Elias L, Shepherd L, Hines J, Threatte GA, Larson RA, Cheson BD,Schiffer CA,. Fludarabine compared with chlorambucil as primary therapy for chronic lymphocytic leukemia. N Engl J Med. 2000;343(24):1750-7.
27. Keating MJ, O'Brien S, Lerner S, et al. Long-term follow-up of patients with chronic lymphocytic leukemia (CLL) receiving fludarabine regimens as initial therapy. Blood. 1998;92(4):1165-1171.
28. Lucas DM, Ruppert AS, Lozanski G, et al. Cytogenetic prioritization with inclusion of molecular markers predicts outcome in previously untreated patients with chronic lymphocytic leukemia treated with fludarabine or fludarabine plus cyclophosphamide: A long-term follow-up study of the US intergroup phase III trial E2997. Leuk Lymphoma. 2015;56(11):3031-3037.
29. Hallek M, Fischer K, Fingerle-Rowson G, et al. Addition of rituximab to fludarabine and cyclophosphamide in patients with chronic lymphocytic leukaemia: A randomised, open-label, phase 3 trial. Lancet. 2010;376(9747):1164-74.
30. Catovsky D, Richards S, Matutes E, et al. Assessment of fludarabine plus cyclophosphamide for patients with chronic lymphocytic leukaemia (the LRF CLL4 trial): A randomised controlled trial. Lancet. 2007;370(9583):230-239.
31. Boross P, Leusen JHW. Mechanisms of action of CD20 antibodies. American Journal of Cancer Research. 2012;2(6).
135
32. Huhn D, von Schilling C, Wilhelm M, Ho AD, Hallek M, Kuse R, Knauf W, Riedel U, Hinke A, Srock S, Serke S, Peschel C, Emmerich B,German Chronic Lymphocytic Leukemia Study Group,. Rituximab therapy of patients with B-cell chronic lymphocytic leukemia. Blood. 2001;98(5):1326-31.
33. Tam CS, O'Brien S, Wierda W, et al. Long-term results of the fludarabine, cyclophosphamide, and rituximab regimen as initial therapy of chronic lymphocytic leukemia. Blood. 2008;112(4):975-980.
34. Rai KR, Jain P. Chronic lymphocytic leukemia (CLL)—Then and now. AJH American Journal of Hematology. 2016;91(3):330-340.
35. Byrd JC, Jones JJ, Woyach JA, Johnson AJ, Flynn JM. Entering the era of targeted therapy for chronic lymphocytic leukemia: Impact on the practicing clinician. J Clin Oncol. 2014;32(27):3039-3047.
36. Chaudhry M, Cheson BD. What is the status of novel anti-CD20 antibodies for chronic lymphocytic leukemia and are they set to leave rituximab in the shadows? Expert Rev Hematol. 2015;8(6):733-742.
37. Lozanski G, Heerema NA, Flinn IW, Smith L, Harbison J, Webb J, Moran M, Lucas M, Lin T, Hackbarth ML, Proffitt JH, Lucas D, Grever MR, Byrd JC. Alemtuzumab is an effective therapy for chronic lymphocytic leukemia with p53 mutations and deletions. Blood. 2004;103(9):3278-81.
38. Dal Porto JM, Gauld SB, Merrell KT, Mills D, Pugh-Bernard AE, Cambier J. B cell antigen receptor signaling 101. Mol Immunol. 2004;41(6-7):6-7.
39. Zhong Y, Byrd JC, Dubovsky JA. The B-cell receptor pathway: A critical component of healthy and malignant immune biology. Semin Hematol. 2014;51(3):206-218.
40. Geering B, Cutillas PR, Nock G, Gharbi SI, Vanhaesebroeck B. Class IA phosphoinositide 3-kinases are obligate p85-p110 heterodimers. Proc Natl Acad Sci U S A. 2007;104(19):7809-7814.
41. Cantrell DA. Phosphoinositide 3-kinase signalling pathways. J Cell Sci. 2001;114:1439-45.
42. Kim D, Chung J. Akt: Versatile mediator of cell survival and beyond. J Biochem Mol Biol. 2002;35(1):106-115.
43. Chang L, Karin M. Mammalian MAP kinase signalling cascades. Nature. 2001;410(6824):37-40.
136
44. Herman SE, Gordon AL, Hertlein E, et al. Bruton tyrosine kinase represents a promising therapeutic target for treatment of chronic lymphocytic leukemia and is effectively targeted by PCI-32765. Blood. 2011;117(23):6287-6296.
45. Woyach JA, Johnson AJ, Byrd JC. The B-cell receptor signaling pathway as a therapeutic target in CLL. Blood. 2012;120(6):1175-1184.
46. Honigberg LA, Smith AM, Sirisawad M, et al. The bruton tyrosine kinase inhibitor PCI-32765 blocks B-cell activation and is efficacious in models of autoimmune disease and B-cell malignancy. Proc Natl Acad Sci U S A. 2010;107(29):13075-13080.
47. Byrd JC, Furman RR, Coutre SE, et al. Three-year follow-up of treatment-naive and previously treated patients with CLL and SLL receiving single-agent ibrutinib. Blood. 2015;125(16):2497-2506.
48. Byrd JC, Harrington B, O'Brien S, et al. Acalabrutinib (ACP-196) in relapsed chronic lymphocytic leukemia. N Engl J Med. 2016;374(4):323-332.
49. Chantry D, Vojtek A, Kashishian A, Holtzman DA, Wood C, Gray PW, Cooper JA,Hoekstra MF. P110delta, a novel phosphatidylinositol 3-kinase catalytic subunit that associates with p85 and is expressed predominantly in leukocytes. The Journal of biological chemistry. 1997;272(31):19236-41.
50. O'Brien SM, Lamanna N, Kipps TJ, et al. A phase 2 study of idelalisib plus rituximab in treatment-naive older patients with chronic lymphocytic leukemia. Blood. 2015;126(25):2686-2694.
51. Brown JR, Byrd JC, Coutre SE, et al. Idelalisib, an inhibitor of phosphatidylinositol 3-kinase p110delta, for relapsed/refractory chronic lymphocytic leukemia. Blood. 2014;123(22):3390-3397.
52. Catovsky D, Fooks J, Richards S. Prognostic factors in chronic lymphocytic leukaemia: The importance of age, sex and response to treatment in survival. A report from the MRC CLL 1 trial. MRC working party on leukaemia in adults. Br J Haematol. 1989;72(2):141-149.
53. Fischer K, Bahlo J, Fink AM, et al. Long-term remissions after FCR chemoimmunotherapy in previously untreated patients with CLL: Updated results of the CLL8 trial. Blood. 2016;127(2):208-215.
54. Hallek M, Wanders L, Ostwald M, et al. Serum beta(2)-microglobulin and serum thymidine kinase are independent predictors of progression-free survival in chronic lymphocytic leukemia and immunocytoma. Leuk Lymphoma. 1996;22(5-6):439-447.
137
55. Montserrat E, Sanchez-Bisono J, Vinolas N, Rozman C. Lymphocyte doubling time in chronic lymphocytic leukaemia: Analysis of its prognostic significance. Br J Haematol. 1986;62(3):567-575.
56. Shanafelt TD, Geyer SM, Kay NE. Prognosis at diagnosis: Integrating molecular biologic insights into clinical practice for patients with CLL. Blood. 2004;103(4):1202- 1210.
57. Simonsson B, Wibell L, Nilsson K. Beta 2-microglobulin in chronic lymphocytic leukaemia. Scand J Haematol. 1980;24(2):174-180.
58. Keating MJ, O'Brien S, Robertson L, Huh Y. Chronic lymphocytic leukemia - correlation of response and survival. Leuk Lymphoma. 1993;11:167.
59. Fayad L, Keating MJ, Reuben JM, et al. Interleukin-6 and interleukin-10 levels in chronic lymphocytic leukemia: Correlation with phenotypic characteristics and outcome. Blood. 2001;97(1):256-263.
60. Gentile M, Mauro FR, Rossi D, et al. Italian external and multicentric validation of the MD anderson cancer center nomogram and prognostic index for chronic lymphocytic leukaemia patients: Analysis of 1502 cases. Br J Haematol. 2014;167(2):224-232.
61. Molica S, Levato D, Cascavilla N, Levato L, Musto P. Clinico-prognostic implications of simultaneous increased serum levels of soluble CD23 and beta2- microglobulin in B-cell chronic lymphocytic leukemia. Eur J Haematol. 1999;62(2):117- 122.
62. Malavasi F, Deaglio S, Damle R, Cutrona G, Ferrarini M, Chiorazzi N. CD38 and chronic lymphocytic leukemia: A decade later. Blood. 2011;118(13):3470-3478.
63. Damle RN, Wasil T, Fais F, et al. Ig V gene mutation status and CD38 expression as novel prognostic indicators in chronic lymphocytic leukemia. Blood. 1999;94(6):1840- 1847.
64. Hamblin TJ, Davis Z, Gardiner A, Oscier DG, Stevenson FK. Unmutated ig V(H) genes are associated with a more aggressive form of chronic lymphocytic leukemia. Blood. 1999;94(6):1848-1854.
65. Damle RN, Wasil T, Fais F, et al. Ig V gene mutation status and CD38 expression as novel prognostic indicators in chronic lymphocytic leukemia. Blood. 1999;94(6):1840- 1847.
138
66. Oscier DG, Gardiner AC, Mould SJ, et al. Multivariate analysis of prognostic factors in CLL: Clinical stage, IGVH gene mutational status, and loss or mutation of the p53 gene are independent prognostic factors. Blood. 2002;100(4):1177-1184.
67. Hamblin TJ, Orchard JA, Ibbotson RE, et al. CD38 expression and immunoglobulin variable region mutations are independent prognostic variables in chronic lymphocytic leukemia, but CD38 expression may vary during the course of the disease. Blood. 2002;99(3):1023-1029.
68. Vasconcelos Y, Davi F, Levy V, et al. Binet's staging system and VH genes are independent but complementary prognostic indicators in chronic lymphocytic leukemia. J Clin Oncol. 2003;21(21):3928-3932.
69. Fais F, Ghiotto F, Hashimoto S, et al. Chronic lymphocytic leukemia B cells express restricted sets of mutated and unmutated antigen receptors. J Clin Invest. 1998;102(8):1515-1525.
70. Fabbri G, Dalla-Favera R. The molecular pathogenesis of chronic lymphocytic leukaemia. Nat Rev Cancer. 2016;16(3):145-162.
71. Vaisitti T, Aydin S, Rossi D, et al. CD38 increases CXCL12-mediated signals and homing of chronic lymphocytic leukemia cells. Leukemia. 2010;24(5):958-969.
72. Deaglio S, Aydin S, Grand MM, et al. CD38/CD31 interactions activate genetic pathways leading to proliferation and migration in chronic lymphocytic leukemia cells. Mol Med. 2010;16(3-4):87-91.
73. Ibrahim S, Keating M, Do KA, et al. CD38 expression as an important prognostic factor in B-cell chronic lymphocytic leukemia. Blood. 2001;98(1):181-186.
74. D'Arena G, Musto P, Cascavilla N, et al. CD38 expression correlates with adverse biological features and predicts poor clinical outcome in B-cell chronic lymphocytic leukemia. Leuk Lymphoma. 2001;42(1-2):109-114.
75. Montillo M, Hamblin T, Hallek M, Montserrat E, Morra E. Chronic lymphocytic leukemia: Novel prognostic factors and their relevance for risk-adapted therapeutic strategies. Haematologica. 2005;90(3):391-399.
76. Rosenwald A, Alizadeh AA, Widhopf G, Simon R, Davis RE, Yu X, Yang L, Pickeral OK, Rassenti LZ, Powell J, Botstein D, Byrd JC, Grever MR, Cheson BD, Chiorazzi N, Wilson WH, Kipps TJ, Brown PO,Staudt LM,. Relation of gene expression phenotype to immunoglobulin mutation genotype in B cell chronic lymphocytic leukemia. J Exp Med. 2001;194(11):1639-47.
139
77. Moreno C ME. New prognostic markers in chronic lymphocytic leukemia. Blood Rev. 2008;22(4):211-9.
78. Orchard J, Ibbotson R, Best G, Parker A, Oscier D. ZAP-70 in B cell malignancies. Leuk Lymphoma. 2005;46(12):1689-1698.
79. Chen L, Widhopf G, Huynh L, et al. Expression of ZAP-70 is associated with increased B-cell receptor signaling in chronic lymphocytic leukemia. Blood. 2002;100(13):4609-4614.
80. Messmer D, Fecteau JF, O'Hayre M, Bharati IS, Handel TM, Kipps TJ. Chronic lymphocytic leukemia cells receive RAF-dependent survival signals in response to CXCL12 that are sensitive to inhibition by sorafenib. Blood. 2011;117(3):882-889.
81. Richardson SJ, Matthews C, Catherwood MA, et al. ZAP-70 expression is associated with enhanced ability to respond to migratory and survival signals in B-cell chronic lymphocytic leukemia (B-CLL). Blood. 2006;107(9):3584-3592.
82. Crespo M, Bosch F, Villamor N, et al. ZAP-70 expression as a surrogate for immunoglobulin-variable-region mutations in chronic lymphocytic leukemia. N Engl J Med. 2003;348(18):1764-1775.
83. Corcoran M, Parker A, Orchard J, et al. ZAP-70 methylation status is associated with ZAP-70 expression status in chronic lymphocytic leukemia. Haematologica. 2005;90(8):1078-1088.
84. Claus R, Lucas DM, Stilgenbauer S, et al. Quantitative DNA methylation analysis identifies a single CpG dinucleotide important for ZAP-70 expression and predictive of prognosis in chronic lymphocytic leukemia. J Clin Oncol. 2012;30(20):2483-2491.
85. Döhner H, Stilgenbauer S, Benner A, et al. Genomic aberrations and survival in chronic lymphocytic leukemia. N Engl J Med. 2000;343(26):1910-6.
86. Van Dyke DL, Werner L, Rassenti LZ, et al. The dohner fluorescence in situ hybridization prognostic classification of chronic lymphocytic leukaemia (CLL): The CLL research consortium experience. Br J Haematol. 2016;173(1):105-113.
87. Vogelstein B, Lane D, Levine AJ. Surfing the p53 network. Nature. 2000;408(6810):307-310.
88. Shiloh Y, Ziv Y. The ATM protein kinase: Regulating the cellular response to genotoxic stress, and more. Nat Rev Mol Cell Biol. 2013;14(4):197-210.
140
89. Lim LP, Lau NC, Garrett-Engele P, et al. Microarray analysis shows that some microRNAs downregulate large numbers of target mRNAs. Nature. 2005;433(7027):769- 773.
90. Cimmino A, Calin GA, Fabbri M, et al. miR-15 and miR-16 induce apoptosis by targeting BCL2. Proc Natl Acad Sci U S A. 2005;102(39):13944-13949.
91. Dierlamm J, Michaux L, Criel A, Wlodarska I, Van den Berghe H, Hossfeld DK. Genetic abnormalities in chronic lymphocytic leukemia and their clinical and prognostic implications. Cancer Genet Cytogenet. 1997;94(1):27-35.
92. Geisler CH, Philip P, Christensen B, et al. In B-cell chronic lymphocytic leukaemia chromosome 17 abnormalities and not trisomy 12 are the single most important cytogenetic abnormalities for the prognosis: A cytogenetic and immunophenotypic study of 480 unselected newly diagnosed patients. Leukemia Research. 1997;21(11-12):1011- 1023.
93. Muthusamy N, Breidenbach H, Andritsos L, et al. Enhanced detection of chromosomal abnormalities in chronic lymphocytic leukemia by conventional cytogenetics using CpG oligonucleotide in combination with pokeweed mitogen and phorbol myristate acetate. Cancer Genet. 2011;204(2):77-83.
94. Put N, Konings P, Rack K, et al. Improved detection of chromosomal abnormalities in chronic lymphocytic leukemia by conventional cytogenetics using CpG oligonucleotide and interleukin-2 stimulation: A belgian multicentric study. Genes Chromosomes Cancer. 2009;48(10):843-853.
95. Struski S, Gervais C, Helias C, Herbrecht R, Audhuy B, Mauvieux L. Stimulation of B-cell lymphoproliferations with CpG-oligonucleotide DSP30 plus IL-2 is more effective than with TPA to detect clonal abnormalities. Leukemia. 2009;23(3):617-619.
96. Decker T, Schneller F, Sparwasser T, et al. Immunostimulatory CpG-oligonucleotides cause proliferation, cytokine production, and an immunogenic phenotype in chronic lymphocytic leukemia B cells. Blood. 2000;95(3):999-1006.
97. Krieg AM. From A to Z on CpG. Trends in Immunology. 2002;23(2):64-65.
98. Liang X, Moseman EA, Farrar MA, et al. Toll-like receptor 9 signaling by CpG-B oligodeoxynucleotides induces an apoptotic pathway in human chronic lymphocytic leukemia B cells. Blood. 2010;115(24):5041-5052.
99. Heerema NA, Byrd JC, Dal Cin PS, et al. Stimulation of chronic lymphocytic leukemia cells with CpG oligodeoxynucleotide gives consistent karyotypic results among
141 laboratories: A CLL research consortium (CRC) study. Cancer Genet Cytogenet. 2010;203(2):134-140.
100. Wu X, Smoley SA, Peterson MA, et al. Biological evaluation of CpG stimulation of normal human B-cells: Implications for B-cell biology and cytogenetic analysis of CLL B-cells. Br J Haematol. 2011;153(3):402-405.
101. Haferlach C, Dicker F, Weiss T, Schnittger S, Beck C, Grote-Metke A, Oruzio D, Kern W, Haferlach T. Toward a comprehensive prognostic scoring system in chronic lymphocytic leukemia based on a combination of genetic parameters. Genes Chromosomes Cancer. 2010;49(9):851-9.
102. Baliakas P, Iskas M, Gardiner A, et al. Chromosomal translocations and karyotype complexity in chronic lymphocytic leukemia: A systematic reappraisal of classic cytogenetic data. Am J Hematol. 2014;89(3):249-255.
103. Rigolin GM, del Giudice I, Formigaro L, et al. Chromosome aberrations detected by conventional karyotyping using novel mitogens in chronic lymphocytic leukemia: Clinical and biologic correlations. Genes Chromosomes Cancer. 2015;54(12):818-826.
104. Badoux XC, Keating MJ, Wang X, et al. Cyclophosphamide, fludarabine, alemtuzumab, and rituximab as salvage therapy for heavily pretreated patients with chronic lymphocytic leukemia. Blood. 2011;118(8):2085-2093.
105. Badoux XC, Keating MJ, Wang X, et al. Fludarabine, cyclophosphamide, and rituximab chemoimmunotherapy is highly effective treatment for relapsed patients with CLL. Blood. 2011;117(11):3016-3024.
106. Woyach JA, Lozanski G, Ruppert AS, Lozanski A, Blum KA, Jones JA, Flynn JM, Johnson AJ, Grever MR, Heerema NA,Byrd JC,. Outcome of patients with relapsed or refractory chronic lymphocytic leukemia treated with flavopiridol: Impact of genetic features. Leukemia. 2012;26(6):1442-4.
107. Jaglowski SM, Ruppert AS, Heerema NA, et al. Complex karyotype predicts for inferior outcomes following reduced-intensity conditioning allogeneic transplant for chronic lymphocytic leukaemia. Br J Haematol. 2012;159(1):82-87.
108. Maddocks KJ, Ruppert AS, Lozanski G, et al. Etiology of ibrutinib therapy discontinuation and outcomes in patients with chronic lymphocytic leukemia. JAMA Oncol. 2015;1(1):80-87.
109. Thompson PA, O'Brien SM, Wierda WG, et al. Complex karyotype is a stronger predictor than del(17p) for an inferior outcome in relapsed or refractory chronic
142 lymphocytic leukemia patients treated with ibrutinib-based regimens. Cancer. 2015;121(20):3612-3621.
110. Lander ES, Linton LM, Birren B, et al. Initial sequencing and analysis of the human genome. Nature. 2001;409(6822):860-921.
111. Kapranov P, Cheng J, Dike S, et al. RNA maps reveal new RNA classes and a possible function for pervasive transcription. Science. 2007;316(5830):1484-1488.
112. Rinn JL, Chang HY. Genome regulation by long noncoding RNAs. Annu Rev Biochem. 2012;81:145-166.
113. Ulitsky I, Bartel DP. lincRNAs: Genomics, evolution, and mechanisms. Cell. 2013;154(1):26-46.
114. Guttman M, Amit I, Garber M, et al. Chromatin signature reveals over a thousand highly conserved large non-coding RNAs in mammals. Nature. 2009;458(7235):223-7.
115. Iyer MK, Niknafs YS, Malik R, et al. The landscape of long noncoding RNAs in the human transcriptome. Nat Genet. 2015;47(3):199-208.
116. Cabili MN, Trapnell C, Goff L, et al. Integrative annotation of human large intergenic noncoding RNAs reveals global properties and specific subclasses. Genes Dev. 2011;25(18):1915-27.
117. Clamp M, Fry B, Kamal M, et al. Distinguishing protein-coding and noncoding genes in the human genome. Proc Natl Acad Sci U S A. 2007;104(49):19428-19433.
118. Prensner JR, Iyer MK, Sahu A, et al. The long noncoding RNA SChLAP1 promotes aggressive prostate cancer and antagonizes the SWI/SNF complex. Nat Genet. 2013;45(11):1392-1398.
119. Ørom UA, Derrien T, Beringer M, et al. Long noncoding RNAs with enhancer-like function in human cells. Cell. 2010;143(1):46-58.
120. Hung T, Wang Y, Lin MF, et al. Extensive and coordinated transcription of noncoding RNAs within cell-cycle promoters. Nat Genet. 2011;43(7):621-9.
121. Kino T, Hurt DE, Ichijo T, Nader N, Chrousos GP. Noncoding RNA gas5 is a growth arrest- and starvation-associated repressor of the glucocorticoid receptor. Science signaling. 2010;3(107).
143
122. Poliseno L, Salmena L, Zhang J, Carver B, Haveman WJ, Pandolfi PP. A coding- independent function of gene and pseudogene mRNAs regulates tumour biology. Nature. 2010;465(7301):1033-8.
123. Zhao J, Ohsumi TK, Kung JT, et al. Genome-wide identification of polycomb- associated RNAs by RIP-seq. Mol Cell. 2010;40(6):939-953.
124. Khalil AM, Guttman M, Huarte M, et al. Many human large intergenic noncoding RNAs associate with chromatin-modifying complexes and affect gene expression. Proc Natl Acad Sci U S A. 2009;106(28):11667-11672.
125. Chu C, Quinn J, Chang HY. Chromatin isolation by RNA purification (ChIRP). J Vis Exp. 2012;(61). pii: 3912. doi(61):10.3791/3912.
126. Spitale RC, Tsai MC, Chang HY. RNA templating the epigenome: Long noncoding RNAs as molecular scaffolds. Epigenetics. 2011;6(5):539-43.
127. Wang KC, Chang HY. Molecular mechanisms of long noncoding RNAs. Mol Cell. 2011;43(6):904-14.
128. Tsai MC, Spitale RC, Chang HY. Long intergenic noncoding RNAs: New links in cancer progression. Cancer Res. 2011;71(1):3-7.
129. Gupta RA, Shah N, Wang KC, et al. Long non-coding RNA HOTAIR reprograms chromatin state to promote cancer metastasis. Nature. 2010;464(7291):1071-1076.
130. Zhang S, Chen S, Yang G, et al. Long noncoding RNA HOTAIR as an independent prognostic marker in cancer: A meta-analysis. PLoS One. 2014;9(8):e105538.
131. Kogo R, Shimamura T, Mimori K, et al. Long noncoding RNA HOTAIR regulates polycomb-dependent chromatin modification and is associated with poor prognosis in colorectal cancers. Cancer Res. 2011;71(20):6320-6326.
132. Li D, Feng J, Wu T, et al. Long intergenic noncoding RNA HOTAIR is overexpressed and regulates PTEN methylation in laryngeal squamous cell carcinoma. The American journal of pathology. 2013;182(1):64-70.
133. Rinn JL, Kertesz M, Wang JK, et al. Functional demarcation of active and silent chromatin domains in human HOX loci by noncoding RNAs. Cell. 2007;129(7):1311-23.
134. Chu C, Qu K, Zhong FL, Artandi SE, Chang HY. Genomic maps of long noncoding RNA occupancy reveal principles of RNA-chromatin interactions. Mol Cell. 2011;44(4):667-78.
144
135. Lejeune J, Maunoury C, Prieur M, Van den Akker J. A jumping translocation (5p;15q), (8q;15q), and (12q;15q) (author's transl). Ann Genet. 1979;22(4):210-213.
136. Berger R, Bernard OA. Jumping translocations. Genes Chromosomes Cancer. 2007;46(8):717-723.
137. Najfeld V, Hauschildt B, Scalise A, et al. Jumping translocations in leukemia. Leukemia. 1995;9(4):634-639.
138. Reis MD, Dube ID, Pinkerton PH, et al. "Jumping" translocations involving band 3q13.3 in a case of acute monocytic leukemia. Cancer Genet Cytogenet. 1991;51(2):189- 194.
139. Wlodarska I, Mecucci C, De Wolf-Peeters C, et al. "Jumping" translocation of 9q in a case of follicular lymphoma. Cancer Genet Cytogenet. 1994;76(2):140-144.
140. Fan YS, Rizkalla K, William BF, Engel CJ. Jumping translocations of 11q in acute myeloid leukemia and 1q in follicular lymphoma. Cancer Genet Cytogenet. 2000;118(1):35-41.
141. Haltrich I, Kost-Alimova M, Kovacs G, et al. Jumping translocation of 17q11 approximately qter and 3q25 approximately q28 duplication in a variant philadelphia t(9;14;22)(q34;q32;q11) in a childhood chronic myelogenous leukemia. Cancer Genet Cytogenet. 2006;164(1):74-80.
142. Manola KN, Georgakakos VN, Stavropoulou C, et al. Jumping translocations in hematological malignancies: A cytogenetic study of five cases. Cancer Genet Cytogenet. 2008;187(2):85-94.
143. Najfeld V, Tripodi J, Scalise A, et al. Jumping translocations of the long arms of chromosome 1 in myeloid malignancies is associated with a high risk of transformation to acute myeloid leukaemia. Br J Haematol. 2010;151(3):288-291.
144. Callet-Bauchu E, Salles G, Gazzo S, et al. Translocations involving the short arm of chromosome 17 in chronic B-lymphoid disorders: Frequent occurrence of dicentric rearrangements and possible association with adverse outcome. Leukemia. 1999;13(3):460-468.
145. Muthusamy N, Breidenbach H, Andritsos L, et al. Enhanced detection of chromosomal abnormalities in chronic lymphocytic leukemia by conventional cytogenetics using CpG oligonucleotide in combination with pokeweed mitogen and phorbol myristate acetate. Cancer Genet. 2011;204(2):77-83.
145
146. Shaffer LG, Slovak ML, Campbell LJ, eds. ISCN (2009): An international system for human cytogenetic nomenclature. Basel: S. Karger; 2009.
147. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. Journal of the American statistical association. 1958;53(282):457-481.
148. Rai KR, Sawitsky A, Cronkite EP, Chanana AD, Levy RN, Pasternack BS. Clinical staging of chronic lymphocytic leukemia. Blood. 1975;46(2):219-34.
149. Rosenwald A, Ott G, Katzenberger T, et al. Jumping translocation of 1q as the sole aberration in a case of follicular lymphoma. Cancer Genet Cytogenet. 1999;108(1):53-6.
150. Sawyer JR, Tricot G, Mattox S, Jagannath S, Barlogie B. Jumping translocations of chromosome 1q in multiple myeloma: Evidence for a mechanism involving decondensation of pericentromeric heterochromatin. Blood. 1998;91(5):1732-41.
151. Vogelstein B, Lane D, Levine AJ. Surfing the p53 network. Nature. 2000;408(6810):307-310.
152. Dohner H, Fischer K, Bentz M, et al. p53 gene deletion predicts for poor survival and non-response to therapy with purine analogs in chronic B-cell leukemias. Blood. 1995;85(6):1580-1589.
153. Dohner H, Stilgenbauer S, Benner A, et al. Genomic aberrations and survival in chronic lymphocytic leukemia. N Engl J Med. 2000;343(26):1910-1916.
154. Byrd JC, Gribben JG, Peterson BL, et al. Select high-risk genetic features predict earlier progression following chemoimmunotherapy with fludarabine and rituximab in chronic lymphocytic leukemia: Justification for risk-adapted therapy. J Clin Oncol. 2006;24(3):437-443.
155. Grever MR, Lucas DM, Dewald GW, et al. Comprehensive assessment of genetic and molecular features predicting outcome in patients with chronic lymphocytic leukemia: Results from the US intergroup phase III trial E2997. J Clin Oncol. 2007;25(7):799-804.
156. Haferlach C, Dicker F, Schnittger S, Kern W, Haferlach T. Comprehensive genetic characterization of CLL: A study on 506 cases analysed with chromosome banding analysis, interphase FISH, IgV(H) status and immunophenotyping. Leukemia. 2007;21(12):2442-2451.
146
157. Scheurlen WG, Schwabe GC, Seranski P, et al. Mapping of the breakpoints on the short arm of chromosome 17 in neoplasms with an i(17q). Genes Chromosomes Cancer. 1999;25(3):230-240.
158. Fink SR, Smoley SA, Stockero KJ, et al. Loss of TP53 is due to rearrangements involving chromosome region 17p10 approximately p12 in chronic lymphocytic leukemia. Cancer Genet Cytogenet. 2006;167(2):177-181.
159. Barbouti A, Stankiewicz P, Nusbaum C, et al. The breakpoint region of the most common isochromosome, i(17q), in human neoplasia is characterized by a complex genomic architecture with large, palindromic, low-copy repeats. Am J Hum Genet. 2004;74(1):1-10.
160. Stankiewicz P, Lupski JR. Genome architecture, rearrangements and genomic disorders. Trends Genet. 2002;18(2):74-82.
161. Stankiewicz P, Cheung SW, Shaw CJ, Saleki R, Szigeti K, Lupski JR. The donor chromosome breakpoint for a jumping translocation is associated with large low-copy repeats in 21q21.3. Cytogenet Genome Res. 2003;101(2):118-123.
162. Sawyer JR, Tian E, Heuck CJ, et al. Jumping translocations of 1q12 in multiple myeloma: A novel mechanism for deletion of 17p in cytogenetically defined high-risk disease. Blood. 2014;123(16):2504-2512.
163. Woyach JA, Heerema NA, Zhao J, et al. Dic(17;18)(p11.2;p11.2) is a recurring abnormality in chronic lymphocytic leukaemia associated with aggressive disease. Br J Haematol. 2010;148(5):754-759.
164. Jain P, O'Brien S. Richter's transformation in chronic lymphocytic leukemia. Oncology (Williston Park). 2012;26(12):1146-1152.
165. Honigberg LA, Smith AM, Sirisawad M, et al. The bruton tyrosine kinase inhibitor PCI-32765 blocks B-cell activation and is efficacious in models of autoimmune disease and B-cell malignancy. Proc Natl Acad Sci U S A. 2010;107(29):13075-13080.
166. Byrd JC, Furman RR, Coutre SE, et al. Targeting BTK with ibrutinib in relapsed chronic lymphocytic leukemia. N Engl J Med. 2013;369(1):32-42.
167. Byrd JC, Brown JR, O'Brien S, et al. Ibrutinib versus ofatumumab in previously treated chronic lymphoid leukemia. N Engl J Med. 2014;371(3):213-223.
147
168. O'Brien S, Furman RR, Coutre SE, et al. Ibrutinib as initial therapy for elderly patients with chronic lymphocytic leukaemia or small lymphocytic lymphoma: An open- label, multicentre, phase 1b/2 trial. Lancet Oncol. 2014;15(1):48-58.
169. Woyach JA, Furman RR, Liu TM, et al. Resistance mechanisms for the bruton's tyrosine kinase inhibitor ibrutinib. N Engl J Med. 2014;370(24):2286-2294.
170. Liu TM, Woyach JA, Zhong Y, et al. Hypermorphic mutation of phospholipase C, gamma2 acquired in ibrutinib-resistant CLL confers BTK independency upon B-cell receptor activation. Blood. 2015;126(1):61-68.
171. Jain P, Keating M, Wierda W, et al. Outcomes of patients with chronic lymphocytic leukemia after discontinuing ibrutinib. Blood. 2015;125(13):2062-2067.
172. Maddocks KJ, Ruppert AS, Lozanski G, et al. Etiology of ibrutinib therapy discontinuation and outcomes in patients with chronic lymphocytic leukemia. JAMA Oncol. 2015;1(1):80-87.
173. Preetesh Jain M, O'Brien S. Richter's transformation in chronic lymphocytic leukemia. Oncology. 2012;26(12):1146.
174. Scandurra M, Rossi D, Deambrogi C, et al. Genomic profiling of richter's syndrome: Recurrent lesions and differences with de novo diffuse large B-cell lymphomas. Hematol Oncol. 2010;28(2):62-67.
175. Chigrinova E, Rinaldi A, Kwee I, et al. Two main genetic pathways lead to the transformation of chronic lymphocytic leukemia to richter syndrome. Blood. 2013;122(15):2673-2682.
176. Rossi D, Cerri M, Capello D, et al. Biological and clinical risk factors of chronic lymphocytic leukaemia transformation to richter syndrome. Br J Haematol. 2008;142(2):202-215.
177. Rossi D, Spina V, Deambrogi C, et al. The genetics of richter syndrome reveals disease heterogeneity and predicts survival after transformation. Blood. 2011;117(12):3391-3401.
178. Deambrogi C, De Paoli L, Fangazio M, et al. Analysis of the REL, BCL11A, and MYCN proto-oncogenes belonging to the 2p amplicon in chronic lymphocytic leukemia. Am J Hematol. 2010;85(7):541-544.
148
179. Chapiro E, Leporrier N, Radford-Weiss I, et al. Gain of the short arm of chromosome 2 (2p) is a frequent recurring chromosome aberration in untreated chronic lymphocytic leukemia (CLL) at advanced stages. Leuk Res. 2010;34(1):63-68.
180. Fabris S, Mosca L, Cutrona G, et al. Chromosome 2p gain in monoclonal B-cell lymphocytosis and in early stage chronic lymphocytic leukemia. Am J Hematol. 2013;88(1):24-31.
181. Edelmann J, Holzmann K, Miller F, et al. High-resolution genomic profiling of chronic lymphocytic leukemia reveals new recurrent genomic alterations. Blood. 2012;120(24):4783-4794.
182. Rinaldi A, Mian M, Kwee I, et al. Genome-wide DNA profiling better defines the prognosis of chronic lymphocytic leukaemia. Br J Haematol. 2011;154(5):590-599.
183. Chigrinova E, Rinaldi A, Kwee I, et al. Two main genetic pathways lead to the transformation of chronic lymphocytic leukemia to richter syndrome. Blood. 2013;122(15):2673-2682.
184. Bea S, Lopez-Guillermo A, Ribas M, et al. Genetic imbalances in progressed B-cell chronic lymphocytic leukemia and transformed large-cell lymphoma (richter's syndrome). Am J Pathol. 2002;161(3):957-968.
185. Landau DA, Tausch E, Taylor-Weiner AN, et al. Mutations driving CLL and their evolution in progression and relapse. Nature. 2015;526(7574):525-530.
186. Ott G, Kalla J, Ott MM, et al. Blastoid variants of mantle cell lymphoma: Frequent bcl-1 rearrangements at the major translocation cluster region and tetraploid chromosome clones. Blood. 1997;89(4):1421-1429.
187. Kaneko Y, Rowley JD, Variakojis D, et al. Prognostic implications of karyotype and morphology in patients with non-hodgkin's lymphoma. Int J Cancer. 1983;32(6):683- 692.
188. Schultz RA, Delioukina M, Gaal K, et al. Evaluation of chronic lymphocytic leukemia by BAC-based microarray analysis. Mol Cytogenet. 2011;4(1):4-8166-4-4.
189. Storchova Z, Kuffer C. The consequences of tetraploidy and aneuploidy. J Cell Sci. 2008;121(Pt 23):3859-3866.
190. Gray RJ. A class of K-sample tests for comparing the cumulative incidence of a competing risk. The Annals of statistics. 1988:1141-1154.
149
191. Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. Journal of the American statistical association. 1999;94(446):496-509.
192. Calin GA, Ferracin M, Cimmino A, et al. A MicroRNA signature associated with prognosis and progression in chronic lymphocytic leukemia. N Engl J Med. 2005;353(17):1793-1801.
193. Cui B, Chen L, Zhang S, et al. MicroRNA-155 influences B-cell receptor signaling and associates with aggressive disease in chronic lymphocytic leukemia. Blood. 2014;124(4):546-554.
194. Guinn D, Ruppert AS, Maddocks K, et al. miR-155 expression is associated with chemoimmunotherapy outcome and is modulated by bruton's tyrosine kinase inhibition with ibrutinib. Leukemia. 2015;29(5):1210-1213.
195. Mraz M, Chen L, Rassenti LZ, et al. miR-150 influences B-cell receptor signaling in chronic lymphocytic leukemia by regulating expression of GAB1 and FOXP1. Blood. 2014;124(1):84-95.
196. Calin GA, Dumitru CD, Shimizu M, et al. Frequent deletions and down-regulation of micro- RNA genes miR15 and miR16 at 13q14 in chronic lymphocytic leukemia. Proc Natl Acad Sci U S A. 2002;99(24):15524-15529.
197. Cimmino A, Calin GA, Fabbri M, et al. miR-15 and miR-16 induce apoptosis by targeting BCL2. Proc Natl Acad Sci U S A. 2005;102(39):13944-13949.
198. Ulitsky I, Bartel DP. lincRNAs: Genomics, evolution, and mechanisms. Cell. 2013;154(1):26-46.
199. Geng YJ, Xie SL, Li Q, Ma J, Wang GY. Large intervening non-coding RNA HOTAIR is associated with hepatocellular carcinoma progression. J Int Med Res. 2011;39(6):2119-2128.
200. Ji P, Diederichs S, Wang W, et al. MALAT-1, a novel noncoding RNA, and thymosin beta4 predict metastasis and survival in early-stage non-small cell lung cancer. Oncogene. 2003;22(39):8031-8041.
201. Gutschner T, Hammerle M, Eissmann M, et al. The noncoding RNA MALAT1 is a critical regulator of the metastasis phenotype of lung cancer cells. Cancer Res. 2013;73(3):1180-1189.
202. Blume CJ, Hotz-Wagenblatt A, Hullein J, et al. p53-dependent non-coding RNA networks in chronic lymphocytic leukemia. Leukemia. 2015.
150
203. Isin, Mustafa, Ozgur, Emre, Cetin, Guven, Erten, Nilgun, Aktan, Melih, Gezer, Ugur,Dalay, Nejat,. Investigation of circulating lncRNAs in B-cell neoplasms. CCA Clinica Chimica Acta. 2014;431:255-259.
204. Gumireddy K, Li A, Yan J, et al. Identification of a long non-coding RNA- associated RNP complex regulating metastasis at the translational step. EMBO J. 2013;32(20):2672-2684.
205. Flinn IW, Neuberg DS, Grever MR, et al. Phase III trial of fludarabine plus cyclophosphamide compared with fludarabine for patients with previously untreated chronic lymphocytic leukemia: US intergroup trial E2997. J Clin Oncol. 2007;25(7):793- 798.
206. Hertlein E, Beckwith KA, Lozanski G, et al. Characterization of a new chronic lymphocytic leukemia cell line for mechanistic in vitro and in vivo studies relevant to disease. PLoS One. 2013;8(10):e76607.
207. Tarlton JF, Knight PJ. Comparison of reflectance and transmission densitometry, using document and laser scanners, for quantitation of stained western blots. Anal Biochem. 1996;237(1):123-128.
208. Bichi R, Shinton SA, Martin ES, et al. Human chronic lymphocytic leukemia modeled in mouse by targeted TCL1 expression. Proc Natl Acad Sci U S A. 2002;99(10):6955-6960.
209. Costinean S, Zanesi N, Pekarsky Y, et al. Pre-B cell proliferation and lymphoblastic leukemia/high-grade lymphoma in E(mu)-miR155 transgenic mice. Proc Natl Acad Sci U S A. 2006;103(18):7024-7029.
210. Zou YS, Ouahchi K, Lu Y, et al. Jumping translocations of 3q21 in an acute monocytic leukemia (M5) patient reveal mechanisms of multistage telomere shortening in pathogenesis of AML. Leukemia Research Leukemia Research. 2012;36(1):e31-e33.
211. Wan TS, Ma SK, Chow EY, Li YH, Lin SY, Chan LC. Pathogenesis of jumping translocations: A molecular cytogenetics study. Leuk Res. 2004;28(10):1075-1079.
212. Gray BA, Bent-Williams A, Wadsworth J, Maiese RL, Bhatia A, Zori RT. Fluorescence in situ hybridization assessment of the telomeric regions of jumping translocations in a case of aggressive B-cell non-hodgkin lymphoma. Cancer Genet Cytogenet. 1997;98(1):20-7.
151
213. Mathieu N, Pirzio L, Freulet-Marrière M, Desmaze C, Sabatier L. Telomeres and chromosomal instability. Cellular and Molecular Life Sciences CMLS. 2004;61(6):641- 656.
214. O'Sullivan RJ, Karlseder J. Telomeres: Protecting chromosomes against genome instability. Nature reviews. 2010;11(3):171-181.
215. Roos G, Kröber A, Grabowski P, et al. Short telomeres are associated with genetic complexity, high-risk genomic aberrations, and short survival in chronic lymphocytic leukemia. Blood. 2008;111(4):2246-52.
216. Lin TT, Letsolo BT, Jones RE, et al. Telomere dysfunction and fusion during the progression of chronic lymphocytic leukemia: Evidence for a telomere crisis. Blood. 2010;116(11):1899-1907.
217. Ramsay AJ, Quesada V, Foronda M, et al. POT1 mutations cause telomere dysfunction in chronic lymphocytic leukemia. Nat Genet. 2013;45(5):526-530.
218. MacKinnon RN, Campbell LJ. The role of dicentric chromosome formation and secondary centromere deletion in the evolution of myeloid malignancy. Genetics research international. 2011;2011.
219. Storchova Z, Pellman D. From polyploidy to aneuploidy, genome instability and cancer. Nature reviews. 2004;5(1):45-54.
220. King RW. When 2 2= 5: The origins and fates of aneuploid and tetraploid cells. Biochimica et Biophysica Acta (BBA)-Reviews on Cancer. 2008;1786(1):4-14.
221. Ouillette P, Collins R, Shakhan S, et al. The prognostic significance of various 13q14 deletions in chronic lymphocytic leukemia. Clin Cancer Res. 2011;17(21):6778- 6790.
222. Wang L, Pal S, Sif S. Protein arginine methyltransferase 5 suppresses the transcription of the RB family of tumor suppressors in leukemia and lymphoma cells. Mol Cell Biol. 2008;28(20):6262-6277.
223. Yin XY, Grove L, Datta NS, Katula K, Long MW, Prochownik EV. Inverse regulation of cyclin B1 by c-myc and p53 and induction of tetraploidy by cyclin B1 overexpression. Cancer Res. 2001;61(17):6487-6493.
224. Davoli T, Denchi EL, de Lange T. Persistent telomere damage induces bypass of mitosis and tetraploidy. Cell. 2010;141(1):81-93.
152
225. Davoli T, de Lange T. Telomere-driven tetraploidization occurs in human cells undergoing crisis and promotes transformation of mouse cells. Cancer cell. 2012;21(6):765-776.
226. Brugat T, Nguyen-Khac F, Grelier A, Merle-Beral H, Delic J. Telomere dysfunction-induced foci arise with the onset of telomeric deletions and complex chromosomal aberrations in resistant chronic lymphocytic leukemia cells. Blood. 2010;116(2):239-249.
227. Baerlocher GM, Vulto I, de Jong G, Lansdorp PM. Flow cytometry and FISH to measure the average length of telomeres (flow FISH). Nature protocols. 2006;1(5):2365- 2376.
228. Rufer N, Dragowska W, Thornbury G, Roosnek E, Lansdorp PM. Telomere length dynamics in human lymphocyte subpopulations measured by flow cytometry. Nat Biotechnol. 1998;16(8):743-747.
229. Podhorecka M, Skladanowski A, Bozko P. H2AX phosphorylation: Its role in DNA damage response and cancer therapy. J Nucleic Acids. 2010;2010:10.4061/2010/920161.
230. Kufe DW, Pollock RE, Weichselbaum RR, et al. Holland-Frei Cancer Medicine, 6th edition. 2003.
231. Galmarini C, Mackey J, Dumontet C. Nucleoside analogues: Mechanisms of drug resistance and reversal strategies. Leukemia (08876924). 2001;15(6).
232. Panier S, Boulton SJ. Double-strand break repair: 53BP1 comes into focus. Nature reviews Molecular cell biology. 2014;15(1):7-18.
233. Gaidzik V, Döhner K. Prognostic implications of gene mutations in acute myeloid leukemia with normal cytogenetics. . 2008;35(4):346-355.
234. Mrózek K, Heerema NA, Bloomfield CD. Cytogenetics in acute leukemia. Blood Rev. 2004;18(2):115-136.
235. Kelly LM, Gilliland DG. Genetics of myeloid leukemias. Annual review of genomics and human genetics. 2002;3(1):179-198.
236. Dohner K, Dohner H. Molecular characterization of acute myeloid leukemia. Haematologica. 2008;93(7):976-982.
237. Shih AH, Abdel-Wahab O, Patel JP, Levine RL. The role of mutations in epigenetic regulators in myeloid malignancies. Nature Reviews Cancer. 2012;12(9):599-612.
153
238. Patel KP, Ravandi F, Ma D, et al. Acute myeloid leukemia with IDH1 or IDH2 mutation: Frequency and clinicopathologic features. Am J Clin Pathol. 2011;135(1):35- 45.
239. Marcucci G, Maharry K, Wu YZ, et al. IDH1 and IDH2 gene mutations identify novel molecular subsets within de novo cytogenetically normal acute myeloid leukemia: A cancer and leukemia group B study. J Clin Oncol. 2010;28(14):2348-2355.
240. Figueroa ME, Abdel-Wahab O, Lu C, et al. Leukemic IDH1 and IDH2 mutations result in a hypermethylation phenotype, disrupt TET2 function, and impair hematopoietic differentiation. Cancer cell. 2010;18(6):553-567.
241. Sasaki M, Knobbe CB, Munger JC, et al. IDH1 (R132H) mutation increases murine haematopoietic progenitors and alters epigenetics. Nature. 2012;488(7413):656-659.
242. Losman JA, Kaelin WG,Jr. What a difference a hydroxyl makes: Mutant IDH, (R)- 2-hydroxyglutarate, and cancer. Genes Dev. 2013;27(8):836-852.
243. Cairns RA, Mak TW. Oncogenic isocitrate dehydrogenase mutations: Mechanisms, models, and clinical opportunities. Cancer Discov. 2013;3(7):730-741.
244. Dang L, White DW, Gross S, et al. Cancer-associated IDH1 mutations produce 2- hydroxyglutarate. Nature. 2009;462(7274):739-744.
245. Xu W, Yang H, Liu Y, et al. Oncometabolite 2-hydroxyglutarate is a competitive inhibitor of α-ketoglutarate-dependent dioxygenases. Cancer cell. 2011;19(1):17-30.
246. Wu H, Zhang Y. Mechanisms and functions of tet protein-mediated 5- methylcytosine oxidation. Genes Dev. 2011;25(23):2436-2452.
247. Koivunen P, Lee S, Duncan CG, et al. Transformation by the (R)-enantiomer of 2- hydroxyglutarate linked to EGLN activation. Nature. 2012;483(7390):484-488.
248. Sasaki M, Knobbe CB, Itsumi M, et al. D-2-hydroxyglutarate produced by mutant IDH1 perturbs collagen maturation and basement membrane function. Genes Dev. 2012;26(18):2038-2049.
249. Fathi AT, Sadrzadeh H, Borger DR, et al. Prospective serial evaluation of 2- hydroxyglutarate, during treatment of newly diagnosed acute myeloid leukemia, to assess disease activity and therapeutic response. Blood. 2012;120(23):4649-4652.
154
250. DiNardo CD, Propert KJ, Loren AW, et al. Serum 2-hydroxyglutarate levels predict isocitrate dehydrogenase mutations and clinical outcome in acute myeloid leukemia. Blood. 2013;121(24):4917-4924.
251. Janin M, Mylonas E, Saada V, et al. Serum 2-hydroxyglutarate production in IDH1- and IDH2-mutated de novo acute myeloid leukemia: A study by the acute leukemia french association group. J Clin Oncol. 2014;32(4):297-305.
252. Orskov AD, Treppendahl MB, Skovbo A, et al. Hypomethylation and up-regulation of PD-1 in T cells by azacytidine in MDS/AML patients: A rationale for combined targeting of PD-1 and DNA methylation. Oncotarget. 2015;6(11):9612-9626.
253. Schnorfeil FM, Lichtenegger FS, Emmerig K, et al. T cells are functionally not impaired in AML: Increased PD-1 expression is only seen at time of relapse and correlates with a shift towards the memory T cell compartment. Journal of hematology & oncology. 2015;8(1):1-12.
254. Gross S, Cairns RA, Minden MD, et al. Cancer-associated metabolite 2- hydroxyglutarate accumulates in acute myelogenous leukemia with isocitrate dehydrogenase 1 and 2 mutations. J Exp Med. 2010;207(2):339-344.
255. Chaturvedi A, Cruz MM, Jyotsana N, et al. Enantiomer-specific and paracrine leukemogenicity of mutant IDH metabolite 2-hydroxyglutarate. Leukemia. 2016.
256. Kats LM, Reschke M, Taulli R, et al. Proto-oncogenic role of mutant IDH2 in leukemia initiation and maintenance. Cell stem cell. 2014;14(3):329-341.
257. Shih AH, Shank KR, Meydan C, et al. AG-221, a small molecule mutant IDH2 inhibitor, remodels the epigenetic state of IDH2-mutant cells and induces alterations in self-renewal/differentiation in IDH2-mutant AML model in vivo. Blood. 2014;124(21):437-437.
258. Sportoletti P, Varasano E, Rossi R, et al. Npm1 and flt3-itd mutations cooperate to impair hematopoietic stem/progenitor cells differentiation and deregulate gata1 in a mouse model of AML. Haematologica. 2014;99:256-256.
259. Yang M, Soga T, Pollard PJ. Oncometabolites: Linking altered metabolism with cancer. J Clin Invest. 2013;123(9):3652-3658.
155