View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Spiral - Imperial College Digital Repository

The characterisation of a functional interaction between Ikaros and Foxp1 in pre-B-cells and acute lymphoblastic leukaemia

Jonathan Bond

A thesis submitted for the degree of Doctor of Philosophy at Imperial College London Division of Clinical Sciences April 2013

1

Abstract

B-lymphopoiesis is characterised by orderly progression through a series of stages of differentiation, each of which is associated with specific patterns of expression and cellular proliferation. These processes require a high degree of regulation by the coordinated action of a series of transcription factors that are critical for appropriate stage-specific expression.

We describe a novel interaction between Ikaros and Foxp1, two transcription factors that are essential for normal B-lymphopoiesis, and which frequently exhibit abnormal expression in B-lymphoid malignancy. By co-immunoprecipitation, we have demonstrated physical association between the DNA-binding domains of these in vivo , which is recapitulated in a cell-free system in vitro . In functional experiments, ectopic overexpression of Foxp1 by retroviral infection of pre-B-cells resulted in an increased proportion of cells in the G2 phase of the cell cycle. This phenotype was associated with increased transcription of Gpr132 , a gene which encodes the G2A protein that is known to induce G2/M cell cycle arrest in mammalian cells.

The phenotypic effects of Foxp1 overexpression were abolished by co- infection with Ikaros, which abrogated induction of Gpr132 expression by Foxp1, and caused displacement of Foxp1 from the Gpr132 gene, as shown by chromatin immunoprecipitation. Co-infection of IK6 (an Ikaros isoform which lacks the Foxp1 interaction domain and which is frequently expressed in B-cell acute lymphoblastic leukaemia) had no effect on Gpr132 expression or Foxp1 occupancy of the Gpr132 gene, and failed to abolish the phenotypic effects of Foxp1 overexpression.

2

Our results contribute to a growing understanding of interplay during normal lymphopoiesis, and provide mechanistic insights into the roles of Ikaros and Foxp1 in lymphocyte cell cycle homeostasis. In addition, these findings may contribute to the understanding of cell cycle dysregulation and chemotherapeutic resistance that characterise certain subtypes of acute leukaemia that exhibit aberrant expression of these proteins.

3

Acknowledgements

I’m very grateful to everyone who helped to make my three years at the CSC so educational and productive. Thanks to all the Gene Regulation and Chromatin Group, in particular my supervisor Niall Dillon for guiding the project, and Renae

Domaschenz, who made the initial discovery of Ikaros-Foxp1 interaction upon which all this work is based. Special thanks also to my clinical mentor Irene Roberts for all her excellent advice and guidance.

On a personal note, I’m very grateful to all my family and friends for non-scientific support over the years; more than anything else, thanks and love to my wife Kate.

Declaration of Originality

I declare that all work presented in this thesis was performed by me, unless otherwise indicated or referenced.

4

Table of Contents

List of Illustrations ...... 8 Abbreviations ...... 10

Chapter One: Introduction ...... 12 Regulation of ...... 13 Gene Transcription ...... 13 Regulation of Gene Transcription ...... 16 Transcription Factor Structure ...... 18 The Cell Cycle ...... 20 The Eukaryotic Cell Cycle...... 22 Regulation of the Cell Cycle ...... 23 Cell Cycle Dysregulation and Malignancy ...... 24 B-Cell Development ...... 26 B-Cell Differentiation ...... 28 Transcriptional Regulation of B-Cell Differentiation ...... 33 Ikaros ...... 36 Ikaros Protein Structure and Isofoms ...... 37 Ikaros Function ...... 41 Ikaros Dysfunction in Human Leukaemia ...... 45 Forkhead Box Protein P1 (FoxP1) ...... 46 Foxp1 Protein Structure and Isoforms ...... 47 Functions of Foxp1 ...... 49 FOXP1 Dysfunction in Human Lymphoid Malignancy ...... 51 B-cell Acute Lymphoblastic Leukaemia (B-ALL) ...... 53 Genetic Abnormalities in B-ALL ...... 55 Philadelphia-positive (BCR-ABL+) ALL ...... 58 BCR-ABL+ ALL and IKZF1 deletion ...... 61 Aim of the Project ...... 63

5

Chapter Two: Materials and Methods ...... 64 Cell Culture ...... 64 Plasmid Constructs ...... 65 Transfection of HEK 293T cells ...... 69 Preparation of Protein Extracts ...... 69 Reagents used for protein co-immunoprecipitation and immunoblotting. .... 70 Protein co-immunoprecipitation ...... 71 GST Pulldown Assays ...... 72 Retroviral Infection of B3 cells ...... 74 Cell Cycle Analysis ...... 74 Analysis of Cellular Proliferation ...... 75 Quantitative PCR ...... 75 Chromatin Immunoprecipitation ...... 77

Chapter Three: Characterisation of a Physical Interaction between Ikaros and Foxp1 ...... 81 Murine Ikaros and Foxp1 interact in vivo ...... 83 Murine Ikaros and Foxp1 interact in vitro ...... 85 Ikaros and Foxp1 interaction is mediated by the DNA-binding domains of both proteins ...... 89 The human IK6 deletion abolishes interaction with human FOXP1 ...... 95 Summary ...... 100

Chapter Four: Role of the Ikaros/ Foxp1 interaction in regulating the pre- cycle ...... 101 Overexpression of Foxp1 is associated with G2 cell cycle arrest ...... 105 Overexpression of Foxp1 is associated with reduced cellular proliferation108 Foxp1-mediated cell cycle arrest is associated with increased expression of Gpr132 ...... 110 Summary ...... 113

6

Chapter Five: Verification of a lack of functional interaction between IK6 and Foxp1 in murine pre-B cells ...... 114 Co-expression of IK6 does not abrogate Foxp1-induced G2 arrest in pre-B cells ...... 115 Co-infection with IK6 does not alter Foxp1-induced Gpr132 expression in pre-B-cells ...... 121 Summary ...... 124

Chapter Six: Analysis of Foxp1 binding to the murine Gpr132 ...... 125 Identification of gene regulatory elements at the Gpr132 locus ...... 126 Examination of Foxp1 binding to the Gpr132 locus by Chromatin immunoprecipitation (ChIP) ...... 130 Summary ...... 133

Chapter Seven: Discussion ...... 136 Functional consequences of the Ikaros-Foxp1 interaction ...... 138 Ikaros and Foxp1 regulate the cell cycle of pre-B-cells ...... 141 Loss of interaction between Ikaros and FOXP1 may contribute to chemotherapeutic resistance in B-ALL ...... 143

References ...... 147

7

List of Illustrations Figures Figure 1.2.1 Schematic representation of the cell cycle. Figure 1.3.1 Schematic representation of B-cell development. Figure 1.4.1 Schematic representation of Ikaros protein structure. Figure 1.4.2 Schematic representation of Ikaros isoforms 1 – 6. Figure 1.5.1 Schematic representation of Foxp1 protein structure. Figure 1.5.2. Schematic representation of Foxp1 splice variants. Figure 1.6.1 The incidence of recurrent structural chromosomal abnormalities in childhood and adult ALL. Figure 1.6.2 The Philadelphia . Figure 2.1 Maps of Expression Vectors used in transfection and infection experiments. Figure 3.1 Ikaros and Foxp1 interact in murine fetal liver pre-B-cells. Figure 3.2 Murine Ikaros and Foxp1 interact in vivo , when ectopically expressed in 293T cells. Figure 3.3 Coomassie stain of a mock GST pulldown. Figure 3.4 Murine Ikaros and Foxp1 interact in vitro . Figure 3.5 Schematic representation of Ikaros protein and deletion mutants. Figure 3.6 Mapping of the interaction domain of Ikaros. Figure 3.7 Schematic representation of Foxp1 protein and deletion mutants. Figure 3.8 Mapping of Foxp1 interaction domain. Figure 3.9 Co-IP of murine Ikaros and Foxp1 proteins is not DNA-dependent. Figure 3.10 Human IK6 does not interact with human FOXP1 in a 293T cell expression system. Figure 3.11 Human IK6 does not interact with human FOXP1 in vitro . Figure 3.12 Anti-FOXP1 IP of Ikaros isoforms in human leukaemia cell lines. Figure 4.1 Model of the hypothetical effects of changes in the amounts of Ikaros and/or Foxp1 on expression of target . Figure 4.2 Rates of retroviral infection of B3 cells. Figure 4.3 Cell Cycle Profile of infected B3 cells.

8

Figure 4.4 Expression of FLAG-Foxp1 and HA-tagged Ikaros in retrovirally infected B3 cells. Figure 4.5 Measurement of proliferation of retrovirally infected B3 cells. Figure 4.6 Expression of Gpr132 in retrovirally infected B3 cells. Figure 5.1 Rates of retroviral infection of B3 cells. Figure 5.2 Cell Cycle Profile of infected B3 cells. Figure 5.3 Expression of FLAG-Foxp1 and HA-tagged Ikaros in infected B3 cells. Figure 5.4 Expression of Gpr132 in infected B3 cells. Figure 6.1 Location of the genomic regions amplified in Gpr132 ChIP. Figure 6.2 Genomic Sequences of Gpr132 ChIP Regions A and B. Figure 6.3 ChIP analysis of binding of Foxp1 to the Gpr132 gene. Figure 6.4 Updated model of the effects of protein imbalance on expression of gene targets.

Tables Table 2.1 Primer Sequences used in Gpr132 QPCR assays. Table 2.2 Primers used in QPCR analysis of ChIP experiments.

9

Abbreviations ALL Acute Lymphoblastic Leukaemia DNA Deoxyribonculeic acid RNA Ribonucleic Acid RNA Pol RNA Polymerase Enzyme tRNA Transfer RNA rRNA Ribosomal RNA TBP TATA-binding protein bHLH Basic helix-loop-helix FoxP1 Forkhead Box Protein P1 CDK Cyclin-dependent kinase APC/C Anaphase-promoting complex/ Cyclosome ATM Ataxia telangiectasia-mutated ATR Ataxia telangiectasia and Rad3-related Rb INK4 Inhibitors of Kinase 4 SAC Spindle Assembly Checkpoint Ig Immunoglobulin BCR B-cell HSC Haematopoietic stem cell ELP Early lymphoid progenitor CLP Common lymphoid progenitor RAG Recombination activating gene SLC Surrogate light chain IK6 Ikaros isoform 6 NuRD Nucelosome remodelling and deacetylase ES cell HPC Haematopoietic Precursor Cell SNP Single Nucleotide Polymorphism BCR Breakpoint Cluster Region ABL Abelson

10

HEK Human Embryonic Kidney GST Glutathione-S-transferase GSH Glutathione PI Propidium iodide BrDU Bromodeoxyuridine QPCR Quantitative polymerase chain reaction ChIP Chromatin immunoprecipitation Co-IP Co-immunoprecipitation WT Wild-type MSCV Murine stem cell virus IRES Internal ribosomal entry site GFP Green fluorescent protein Gpr132 G-protein coupled receptor 132 SEM Standard error of the mean TSS Transcriptional start site

11

Chapter 1

Introduction

The evolutionary success of virtually all of the species that currently exist on earth is predicated upon the ability to combat a range of infectious agents that threaten an organism’s survival. In vertebrates, defence against infection is mediated by the innate and adaptive immune systems. The generation of mature lymphocytes that possess the capacity to recognise a diverse repertoire of foreign antigens is a prerequisite for the functioning of the adaptive immune system in (Janeway and Medzhitov, 2002).

Normal lymphoid differentiation is characterised by progression through a series of phenotypically-distinct developmental stages and the regulation of stage- specific gene expression by transcription factors is critical for orderly maturation

(Busslinger, 2004). It is important to understand the mechanisms that underpin these processes, as disorders of lymphopoiesis result in serious human pathology in the form of lymphoid malignancy, immune deficiency or autoimmune disease (Knight and Cunningham-Rundles, 2006; Pui et al., 2004).

In this introduction I will first discuss the basic processes involved in gene transcription (Section 1.1) and the cell cycle (Section 1.2). The next segments will include a general description of B-lymphopoiesis and its transcriptional regulation

(Section 1.3), and fuller discussion of the roles of the Ikaros and Foxp1 transcription factors in lymphoid development (Sections 1.4 and 1.5). In the final part of the introduction (Section 1.6), I will discuss how acute lymphoblastic leukaemia (ALL) results from dysregulation of gene transcription during lymphoid differentiation.

12

1.1 Regulation of Gene Expression

As all cells of the human body share identical DNA sequences, phenotypic differences arise from diversity in the repertoires of genes expressed by individual cells (Levine and Tjian, 2003). Multiple layers of regulation govern the transcription of DNA into RNA, while translation of RNA into protein is in turn subject to several additional layers of control (Brivanlou and Darnell, 2002). In this section of the introduction, I will discuss the processes that directly mediate RNA synthesis (Section

1.1.1) and the regulation of transcription initiation (Section 1.1.2).

1.1.1 Gene Transcription

Gene transcription in eukaryotes is mediated by three RNA polymerase enzymes (RNA Pol I, II and III). RNA Pol I only transcribes ribosomal RNA (rRNA), while RNA Pol III transcribes genes that encode transfer RNA (tRNA), rRNA

(specifically 5SrRNA) and other small RNA molecules including microRNAs (Oettel et al., 1998). Discussion in this section will be confined to the actions of RNA Pol II, which transcribes all other eukaryotic genes (Cramer, 2002).

The first step in gene transcription is the opening and unwinding of a small portion of the DNA helix, which allows binding of the transcriptional machinery

(described below). Following partial unwinding of the DNA molecule, one of the exposed strands acts as template for RNA synthesis. Using nucleotide substrates and energy obtained from the hydrolysis of high-energy bonds, RNA Pol II synthesises nucleotides that are complementary to the DNA strand template (Sawadogo and

Sentenac, 1990). RNA Pol II extends RNA one base at a time (elongation), while also

13 catalysing the formation of phosphodiester bonds that link nucleotides together. As

RNA Pol II moves along the DNA molecule, the helix is unwound just ahead of active site, thereby exposing new stretches of sequence for complementary base pairing.

RNA is released almost immediately from the template DNA strand once transcription of the complementary bases is completed (Boeger et al., 2005).

Many additional proteins (comprising at least 100 individual protein subunits) are required for initiation of transcription by RNA Pol II (Hahn, 2004). Along with

RNA Pol II, these proteins constitute the ‘preinitiation complex’. These proteins include ‘general’ transcription factors that help to position RNA Pol II correctly at the promoter, assist in the DNA strand unwinding, and mediate the release of RNA Pol II from the promoter upon commencement of elongation (Dillon, 2006; Matsui et al.,

1980). The largest general transcription factor is TFIID, which comprises a 38kDa

TATA-binding protein (TBP) subunit (the function of which is discussed in the following paragraph) and 13 TBP-associated factors. Other general transcription factors include TFIIH, one subunit of which has DNA helicase activity that is necessary for DNA strand unwinding. A separate TFIIH subunit mediates phosphorylation of the C-terminal domain of RNA Pol II that is necessary for RNA elongation. An additional protein complex called Mediator is required for formation of the preinitiation complex. Mediator contains approximately 30 protein subunits, some of which bind RNA Pol II, while others bind to the transactivation domains of transcriptional activator proteins (discussed in Section 1.1.2) (Reviewed in Kornberg,

2005).

14

Assembly of the preinitiation complex occurs at the core promoter, which is the minimal DNA sequence required to initiate transcription (Smale and Kadonaga,

2003). The core promoter includes the transcriptional start site, a binding site for RNA

Pol II, and binding sites for the general transcription factors. An example of the latter is the TATA box, which is a sequence rich in thymine and adenine nucleotides that is present in 10-20% of all eukaryotic promoters, typically at approximately 25 bases upstream of the transcriptional start site. Binding by the TBP subunit of TFIID causes distortion of TATA box, which facilitates subsequent steps involved in transcription

(Juven-Gershon and Kadonaga, 2010).

An additional factor that influences gene transcription is the physical and chemical state of both DNA and associated histone molecules (collectively known as chromatin). The basic structural unit of chromatin is the nucleosome, which consists of approximately 147 bases of DNA, wrapped around core histone proteins (H2A,

H2B, H3 and H4) (Luger et al., 1997). Individual nucleosomes are joined by the linker histone H1, and are folded into tertiary structures that have variable degrees of compaction (Thoma and Koller, 1977). Compression of nucleosomes into dense

‘heterochromatin’ results in the inhibition of transcription through a variety of mechanisms, including blockage of the formation of the preinitation complex on DNA

(Elgin, 1996). Some genes are located in regions of chromatin that are more ‘open’ than others (‘euchromatin’), and transcription therefore proceeds more readily. For example, sequences that are flanked by the histone variant H2A.Z are more accessible to the transcriptional machinery, and can also be disassembled more rapidly (Jin and

Felsenfeld, 2007). Local DNA accessibility is also critically affected by chemical

15 modification of the histone tails. For instance, histone acetylation impedes nucleosome compaction and causes the chromatin to remain in a more ‘open’, transcriptionally-accessible state (Reviewed in Haberland et al., 2009). Conversely, trimethylation of lysine at position 9 of the tail of histone H3 (H3K9me3) is associated with heterochromatin formation, and inhibition of gene expression

(Reviewed in Greer and Shi, 2012). The action of many transcriptional activators and (discussed in the next section) is mediated by recruitment of chromatin- remodelling complexes and histone-modifying enzymes that alter local chromatin accessibility (Strahl and Allis, 2000).

1.1.2 Regulation of Gene Transcription

The expression of eukaryotic protein-coding genes is governed by the actions of multiple DNA-binding proteins (transcription factors) that bind to specific nucleotide sequences, and either activate or repress transcription (Bulger and

Groudine, 2010). The promoter region of a gene comprises the core promoter and upstream elements that include binding sites for transcriptional activators and repressors. Many promoters contain recurrent structural features. For example, CpG

‘islands’ are stretches of 20-50 CpG dinucleotides that are often found within 100 bases of the transcriptional start site of eukaryotic genes, and methylation of these sequences is associated with transcriptional repression (Gardiner-Garden and

Frommer, 1987). Transcriptional activators and repressors bind to transcriptional

‘enhancer’ elements, which typically range from 200-1000 bases in length and often contain binding sites for several different transcription factors (Arnosti and Kulkarni,

16

2005). Enhancer regions can be located within 1kb (either upstream or downstream) of the core promoter, but they can also be found at great distances from the transcriptional start site (50 kb or more). In these cases, contact with the promoter is mediated by looping of flexible intervening regions of DNA (de Wit and de Laat,

2012; Dillon et al., 1997). A locus control region (LCR) is the operational definition used to describe the full set of positively acting elements (including enhancers) that are able to give position-independent and copy-dependent expression in a transgenic assay (Li et al., 2002).

The primary function of transcriptional activators is to regulate the recruitment, positioning and modification of RNA Pol II, general transcription factors and the

Mediator complex at the gene promoter, in order to stimulate transcription initiation.

Binding of transcriptional activators to enhancer sequences results in recruitment of

RNA Pol II, and increases transcription rates by approximately 1000-fold (Kadonaga,

2004). Activators are also involved in the alteration of chromatin structure adjacent to the promoter through recruitment of chromatin-remodelling and histone-modifying factors, all of which contributes to increased local accessibility of DNA, which facilitates assembly of the preinitiation complex (Narlikar et al., 2002).

Transcriptional repressors may act in several different ways, including sequestration of transcriptional activators and/ or general transcription factors by direct binding, and recruitment of chromatin-remodelling complexes or histone- modifying enzymes that generate ‘repressive’ chromatin modifications. In addition, direct competition for DNA binding sites and occupancy of gene enhancer elements

17 by repressors can prevent binding of transcriptional activators and subsequent transcription initiation (Courey and Jia, 2001; Sabbattini et al., 2001).

1.1.2.1 Transcription Factor Structure

Transcription factors usually comprise at least two major functional domains.

DNA binding commonly occurs through regions which contain alpha helices and/ or beta sheets that bind to the major groove of DNA. For example, the basic helix-loop- helix (bHLH) motif contains two alpha helices that bind DNA through conserved basic amino acid residues that interact with nucleotide phosphate groups (Murre et al.,

1989). Factors also typically contain a transactivation domain, which usually consists of 30-100 amino acids that interact directly with the preinitiation complex and mediate acceleration of transcription initiation (Pabo and Sauer, 1992).

DNA-binding of transcription factors is frequently mediated by ‘’ domains. Ikaros is one such zinc finger transcription factor (discussed in Section 1.5).

The description of a zinc finger is based on the original diagrammatic representation of the structure, however in vivo it appears that the presence of a central zinc ion

(Zn 2+ ) causes compaction of the protein through internal folding, thereby allowing insertion of the polypeptide into the major groove of DNA (Cox and McLendon,

2000). The C2H2-type zinc finger (also known as Krüppel-type, after the transcription factor in which it was initially described) is the most common type of DNA-binding domain in humans. It comprises a 23-26 residue consensus sequence that contains two conserved cysteine (C) and two conserved histidine (H) residues, whose side chains bind one Zn 2+ ion (Krishna et al., 2003). DNA-binding zinc-finger proteins usually

18 contain several zinc fingers that can interact with successive groups of base pairs within the major groove of DNA, as the protein wraps around the double helix. The extent and strength of the protein-DNA interface may also be increased by homodimerisation of zinc finger proteins, and this is typically required for optimum transcriptional activity (Klug, 1999).

‘’ motifs are also frequently found in eukaryotic transcription factors, and are involved in both protein dimerisation and DNA binding. The ‘zipper’ consists of two alpha helices which form a short coiled-coil domain that is held together by interactions between hydrophobic leucine residues (Alber, 1992). Leucine zipper proteins always bind DNA as dimers. Distal to the dimerisation interface, the two alpha helices separate to form a Y-shaped structure that allows side chains to contact and grip the major groove of DNA (Vinson et al., 1989). Forkhead box protein

P1 (Foxp1) (Section 1.5) contains a region rich in leucine zipper motifs that is believed to mediate protein dimerisation, in addition to a forkhead domain that binds

DNA (discussed in Section 1.5.1) (Wang et al., 2003).

19

1.2 The Cell Cycle

The cell cycle describes the processes by which a cell duplicates its entire genome and organelles (interphase), and divides into two identical daughter cells

(mitosis, or ‘M’ phase). Interphase comprises a period of DNA synthesis (‘S’), and two ‘gap’ phases (G1 and G2) of increased cellular amino acid and RNA biosynthesis, that precede S phase and M phase respectively. Cell cycle progression is mediated by cyclin-dependent kinases (CDKs) that are enzymatically inactive unless bound to a regulatory cyclin, which also confers substrate specificity (Reviewed in Nigg, 2001).

Timely phosphorylation of target molecules is therefore largely governed by fluctuations in cyclin concentration at different phases of the cycle (Pavletich, 1999).

At least 13 CDKs and 29 cyclins (and cyclin-related proteins) exist in humans, however only a subset are important for cell cycle progression - the interphase CDKs

(CDK2, CDK4 and CDK6), the ‘mitotic’ CDK (CDK1) and four classes of cyclin (D-,

E-, A- and B-type) (Malumbres and Barbacid, 2005).

This section will describe the eukaryotic cell cycle (Section 1.2.1) and the regulatory mechanisms that govern its progression (Section 1.2.2). A schematic overview of the cycle is depicted in Figure 1.2.1. In Section 1.2.3, I will discuss how cell cycle regulation is frequently disrupted in human malignancies.

20

Cyclin B Cyclin D CDK1 CDK4, CDK6 M G1

Spindle Assembly and Spindle Position Checkpoints APC/C

DNA Damage Checkpoint Cyclin A ATM, ATR, CDK1 G2

Cyclin E S CD K2

Cyclin A CDK2

Figure 1.2.1 Schematic representation of the cell cycle. Phases of the cycle are listed within the solid blue arrows. The main cyclins and cyclin-dependent kinases (CDKs) involved in cycle progression are shown in green, outside the circle. The proteins involved in the major cycle checkpoints are listed in red inside the circle. APC/C = Anaphase-promoting complex/ Cyclosome. ATM = ataxia telangiectasia- mutated. ATR = ataxia telangiectasia and Rad3-related.

21

1.2.1 The Eukaryotic Cell Cycle

A key event of interphase is the replication of DNA during S phase (Reviewed in Diffley, 2004). During G1, mitogenic signals promote expression of D-type cyclins, which activate CDK4 and CDK6 (Malumbres and Barbacid, 2001). Phosphorylation of retinoblastoma (Rb) protein (which normally represses the activity of transcription factors) is the key event in progression to S phase. Phosphorylation by

CDK4 and CDK6 inhibits Rb interaction with E2F, allowing upregulation of genes involved in DNA synthesis (Ekholm and Reed, 2000). E2F factors also induce expression of E-type cyclins, which promote CDK2-mediated phosphorylation of key regulatory sites in the DNA prereplication complex that in turn results in initiation of

DNA replication. CDK2 is subsequently activated by A-type cyclins during the later stages of S phase, and remains active until the end of G2. A-type cyclins then activate

CDK1 to stimulate the onset of mitosis (Enserink and Kolodner, 2010).

At the end of S phase, duplicated DNA molecules are disentangled and condensed into pairs of sister chromatids that are held together by cohesin proteins

(Zou, 2011). Formation of a protein structure called a kinetochore at the chromatid centromere is necessary for further cell cycle progression (Kline-Smith et al., 2005).

Following dissolution of the nuclear envelope, A-type cyclins are degraded and CDK1 is activated by Cyclin B for the duration of mitosis. Sister chromatids align at the central metaphase plate and are attached (via kinetochores) to the mitotic spindle by microtubules which emanate from opposing poles of the cell (Lew and Burke, 2003).

Separation of chromatids during anaphase requires destruction of securin, which otherwise inhibits the separase enzyme that mediates proteolysis of cohesin and

22 release of sister chromatids (Uhlmann et al., 1999). Securin is destroyed through the action of the anaphase promoting complex/ cyclosome (APC/C) (Peters, 2006).

After sister chromatids are pulled apart, the mitotic spindle is disassembled and the newly segregated are packaged into two separate nuclei during telophase. This phase is followed by cytokinesis, which involves separation of the duplicated material into two daughter cells (each of which contain nuclei and all necessary cellular organelles) through the action of a central contractile ring of actin and myosin (Glotzer, 2005). Cytokinesis is triggered by dephosphorylation of CDK targets due to APC/C-mediated inactivation of CDKs. Following cytokinesis, the nuclear envelope re-forms around segregated chromosomes, and the cell enters a stable G1 state of low CDK activity until a new cycle begins (Jeffrey et al., 1995).

1.2.2 Regulation of the Cell Cycle

The activities of cyclin-CDK complexes are constrained by two major classes of protein. Firstly, three related CDK inhibitors (CKIs) – p21 CIP , p27 KIP1 and p57 KIP2 are most active in late G1 and S-phases, although they also bind later cyclin-CDK complexes. The other main group of proteins is the INK4 ( IN hibitors of Kinase 4) family that interact only with CDK4 and CDK6 in mid-G1. The activities of these inhibitory molecules are largely dictated by fluctuations in levels of cyclin-CDK complexes at different stages of the cell cycle (Sherr, 2001).

Cell cycle progression is punctuated by several checkpoints that are critical for replicative fidelity. Activation of these checkpoints results in cycle arrest until the relevant defect is repaired, or cell death (or senescence) if the damage cannot be

23 rectified (Bartek et al., 2004). The first main checkpoint occurs at the G1/S transition, when the presence of unreplicated or damaged DNA blocks commitment to cell cycle entry. This DNA damage checkpoint involves the sensor kinases ataxia telangiectasia- mutated (ATM) and ataxia telangiectasia and Rad3-related (ATR) proteins, and the checkpoint kinases CHK1 and CHK2. Activation of these pathways results in increased levels of p21 CIP , and inhibition of CDK activators such as Cdc25 phosphatase. The p53 molecule is a critical mediator of the DNA damage response, and is phosphorylated (and therefore protected from Mdm2-mediated degradation) by

ATM and ATR. In turn, p53 also promotes p21 CIP transcription (Levine, 1997).

The activity of two spindle checkpoints prevents unequal segregation of genetic material that would otherwise result in aneuploidy. The spindle assembly checkpoint

(SAC) monitors correct microtubule attachment to the chromatid kinetochore.

Incorrect attachment results in inhibition of CDK1 and Cdc20 phosphatase (which is required for APC/C activity), thereby preventing entry into anaphase (Musacchio and

Salmon, 2007). Finally, the spindle position checkpoint relies on feedback from cytoplasmic microtubules, and inhibits progression of mitosis (via inhibition of the

Cdc14 phosphatase) until the correct position is attained (Moore et al., 2009).

1.2.3 Cell Cycle Dysregulation and Malignancy

Human cancers are typically characterised by uncontrolled cellular proliferation and genomic instability, and it is predictable that mutations (both inherited and acquired) in key cell cycle molecules are frequently found. Detailed discussion of the plethora of abnormalities that have been described lies beyond the

24 scope of this thesis. However discussion of some examples will illustrate several mechanisms of oncogenic cell cycle dysregulation.

Deregulation of cyclin-CDK activity by acquired gene mutation has been demonstrated in many human malignancies (Ortega et al., 2002). Several inherited cancer predisposition syndromes are caused by mutations in genes encoding proteins of the DNA damage checkpoint, including ataxia telangiectasia ( ATM ), Seckel syndrome ( ATR ) and Li-Fraumeni syndrome ( CHK2 ) (Caldecott, 2008). Acquired mutations in these genes, and in the downstream targets of these molecules

(particularly p53) are very frequent (Kastan and Bartek, 2004). SAC components

(notably APC/C) are also commonly mutated (Kops et al., 2005). In the context of haematological malignancy, recurrent deletions in the INK4 locus are found in childhood leukaemia (see Section 1.6).

In addition, abnormal prolongation of gap phases can allow cancer cells extra time to repair DNA damage caused by chemotherapy, thereby allowing them to evade checkpoint-mediated cell death. This effect is seen in subsets of leukaemia that exhibit

G2 arrest following radiation and chemotherapy, which is believed to contribute to chemotherapeutic resistance in these cases (Bedi et al., 1995). The results of experiments described in this thesis have particular relevance for elucidation of this phenomenon.

25

1.3 B-Cell Development

B-lymphocytes (henceforth referred to as ‘B-cells’) are blood cells which express clonally diverse surface immunoglobulin (Ig) receptors that recognise antigen as part of the adaptive immune response (LeBien and Tedder, 2008). Historically, the prefix ‘B’ denoted the bursa of Fabricius, which was originally described as the location of Ig-producing cells in chickens (Cooper et al., 1965). In humans, ‘B’ more correctly denotes ‘bone marrow’, which is the predominant site of B-lymphopoieis in post-natal life (Raff et al., 1976).

All mature B-cells express a structurally unique Ig that forms part of the cell surface B-cell receptor (BCR). Somatic rearrangement of Ig gene loci during B- lymphopoiesis (discussed in Section 1.3.1) generates a vast repertoire of Ig molecules that enables recognition of a hugely diverse range of antigen configurations. Binding of Ig by its cognate antigen results in activation of a hitherto ‘naïve’ B-cell, which proliferates and either differentiates into Ig-secreting plasma cells, or participates in a germinal centre reaction in secondary lymphoid organs (i.e. lymph nodes and spleen).

This germinal centre reaction produces the high affinity B-cell clones that constitute the memory compartment of the adaptive immune system, which is critical for a secondary immune response following antigen re-exposure (Jacob et al., 1991).

In addition to antigen recognition, B-cells perform several other important immune functions (Reviewed in Danilova, 2012), including (but not confined to) antigen presentation to T-cells (Janeway et al., 1987), immunomodulatory cytokine secretion (Lampropoulou et al., 2008) and regulation of T-cell activation (Bouaziz et al., 2007).

26

In common with all blood cells, B-cells are derived from immature precursors which, in humans, reside in the fetal liver in pre-natal life, and in the bone marrow thereafter. B-cell development proceeds through a series of stages of increasing lineage restriction and diminishing differentiation potential. Each of these stages is characterised by distinct patterns of cellular proliferation, surface antigen expression and recombination status of Ig genes (Busslinger, 2004). These processes will be discussed in depth in the remainder of this section of the introduction.

It should also be noted that two distinct categories of mature B-cells (B-1 and

B-2) have been defined in the mouse, although the existence of these subsets has not been proven in humans. B-2 cells are considered to be mediators of adaptive immunity

(as described in the preceding paragraphs), while B-1 cells form part of the innate immune system, have distinct phenotype and function, and probably differentiate from separate haematopoietic precursors (Hardy and Hayakawa, 2001). As most of these discrepancies only become relevant following cellular exit from the murine bone marrow, developmental differences between B-cell subsets will not be discussed in this introduction (Reviewed in Montecino-Rodriguez and Dorshkind, 2012).

27

1.3.1 B-cell differentiation

All blood cells originate ultimately from haematopoietic stem cells (HSCs), which are characterised by their dual capacities for self-renewal and differentiation into any haematopoietic cell type. HSCs are defined immunophenotypically by expression of c-kit (CD117) and absence of markers of lineage differentiation (Lin-)

(Spangrude et al., 1988). HSCs give rise to a variety of multipotent progenitors which then develop into mature blood cells via a series of intermediate developmental stages.

The precise differentiation capacity of many of these progenitor cells remains a source of ongoing debate. Traditional assumptions regarding an early lymphoid/ myeloid dichotomy are now contentious, particularly in the light of recent data showing that relatively ‘late’ precursors retain multilineage potential (Doulatov et al., 2010). The immunophenotypic and genetic characteristics of most stages of B-cell development are however considered to be well-defined. A schematic representation of the current consensus is shown in Figure 1.3.1.

The most primitive lymphoid-specific precursors which develop from HSCs are termed ‘early lymphoid progenitors’ (ELP). Precise characterisation of this stage of development remains elusive (Hardy et al., 2007). ELPs are usually defined by retention of an HSC immunophenotype (CD117+, Lin-) and expression of ‘lymphoid- type’ gene transcripts such as DNTT and RAG (Igarashi et al., 2002). It has been suggested that ELPs may also be discriminated from other early progenitors by an increased sensitivity to oestrogen (Medina et al., 2001). Commitment to a lymphoid fate has traditionally been considered to occur with progression to a ‘common lymphoid progenitor’ (CLP) stage, defined by acquisition of IL-7 responsiveness

28

(indicated by CD127/ IL7R α positivity) and Sca-1 expression. CLPs have been shown to generate all major lymphocyte categories (B-, T-, Natural Killer (NK) and dendritic cells), and do not possess myeloid potential (Kondo et al., 1997). It now appears that most mature T-cells develop from immature precursors with dual T-cell/myeloid potential, and that the majority of the CLP compartment is probably destined for B- cell development (Allman et al., 2003; Rumfelt et al., 2006).

Entry into an exclusively B-lymphoid differentiation pathway occurs with the onset of expression of the B-cell specific surface antigen, B220 (also known as

CD45R) at the pro-B-cell stage (Gounari et al., 2002). Pro-B-cells also express CD19, which continues to be expressed for all remaining stages of B-lymphopoieis. Some minor differences in nomenclature have been used by groups who have described alternative paradigms of B-cell development, using similar immunophenotypic methods. For example, the commonly-used schema originally devised by Hardy

(Hardy et al., 1991) also designates a pre-pro-B stage which discriminates precursors that have not commenced transcription of Ig-related genes, and lack expression of

CD19 (see Figure 1.3.1).

Developmental milestones from and inclusive of the pro-B-cell stage onwards are predominantly defined by the rearrangement status of the Ig heavy chain ( Igh ) and light chain ( Igl ) loci. The expression of recombination activating genes 1 and 2 ( RAG1 and RAG2 ), which generate the DNA strand breaks necessary for Ig gene rearrangement, is critical to this process (Oettinger et al., 1990; Schatz et al., 1989).

Targeted knockout of either of these loci causes a complete absence of mature

29 lymphocytes in genetically-modified mice, with differentiation arrest at the pro-B-cell stage (Mombaerts et al., 1992; Shinkai et al., 1992).

The RAG enzymes generate double-stranded DNA breaks that facilitate recombination of the Ig gene segments. Gene rearrangement is initiated with recombination of the diversity (D H) and joining (J H) regions of the Igh locus.

Productive rearrangement of a variable (V H) region to the newly-formed D HJH segment results in synthesis of µ heavy chain (µH) protein, which forms part of the pre-B-cell receptor (pre-BCR). Successful recombination also leads to inhibition of recombination of the second IgH allele (allelic exclusion) (Schlissel, 2002). The structure of the pre-BCR is analogous to that of mature Ig: as the mature light chain binds µH to form mature Ig, so the surrogate light chain (SLC) binds µH to form the pre-BCR. The SLC is a heterodimer of VpreB1 (also known as CD179a) and λ5

(mouse)/ λ14.1 (human) peptides (Sakaguchi and Melchers, 1986). Successful rearrangement of Igh and expression of the pre-BCR is a defining feature of progression to the large pre-B-cell stage, which is also characterised by expression of the CD25 antigen.

Transition to the large pre-BII stage is accompanied by increased cellular proliferation, which is mediated by constitutive pre-BCR signalling. Following this phase of proliferative expansion, pre-BCR signalling causes downregulation of expression of λ5 and Vpreb1 in a negative feedback loop, inhibiting proliferation and causing cell cycle exit and differentiation to small pre-B cells (Martensson et al.,

2002). The role of Ikaros family proteins in the transition of pre-BCR proliferative function is discussed in Section 1.4.2. At the small pre-B-cell stage of development,

30 productive rearrangement of the Igl locus generates a functional light chain. The light- chain associates with µH to form the B-cell receptor (BCR), which is expressed on the surface of immature B-cells. Following loss of CD25 and acquisition of IgD, CD21 and CD22 expression, immature B-cells then exit the bone marrow to mature fully in the secondary lymphoid organs (Loder et al., 1999).

31

CD117+ HSC Lin-

CD117+ ELP Lin- DNTT, RAG +

CD117+/- IL7Rα+ CLP DNTT, RAG + B220+/-

Pre - Pro -B Large Large Small Immature Pro-B Pre-B-I Pre-B-II Pre-B B

CD117- IL7Rα+ IL7Rα+ IL7Rα+ IL7Rα+/- IL7Rα- IL7Rα+ B220+ B220+ B220+ B220+ B220+ B220+ CD19+ CD19+ CD19+ CD19+ CD19+

CD19- IgH D HJH IgH V HDJ H IgH V HDJ H IgH V HDJ H IgH V HDJ H

SLC SLC SLC IgL V L-JL IgL V L-JL Pre-BCR+ Pre-BCR+ Pre-BCR- BCR+

Figure 1.3.1 Schematic representation of B-cell development. Key surface antigen and transcript expression are listed in black, and Ig gene rearrangement status and expression of B-cell receptor (BCR) components are denoted in red. HSC, Haematopoietic Stem Cell. ELP, Early Lymphoid Progenitor. CLP, Common

Lymphoid Progenitor. IgH, IgL Immunoglobulin Heavy and Light Chains. D,

Diversity Segment. J, Joining Segment. V, Variable Segment.

32

1.3.2 Transcriptional regulation of B-cell differentiation

The complexity of the processes that underpin B-cell development mandates tight transcriptional regulation in order to ensure appropriate stage-specific gene expression. This control is provided by the coordinated action of a series of transcription factors (Nutt and Kee, 2007). Most of our knowledge regarding the function of specific transcription factors comes from observation of the phenotype of genetically modified mice with targeted knockdown of individual proteins. From this work, it is evident that PU.1, Bcl11a, E2a, Early B-cell factor 1 (Ebf1), Pax-5, Ikaros and Foxp1 all perform critical roles during normal B-lymphopoiesis. The roles of the latter two proteins will be discussed more fully in Sections 1.4 and 1.5, while relevant data from murine knockout models of the other transcription factors will be discussed in this section of the introduction.

Similar phenotypes are seen with mice which lack PU.1, Bcl11a, E2a or Ebf1.

Mice that have a deficiency of any of these factors all have blocks at very early stages of B-cell development. A mouse model in which the DNA-binding domain of PU.1 is disrupted has impairments in HSC homing, / monocyte and lymphocyte maturation and developmental arrest at the pro-B-cell stage (DeKoter et al., 2002).

The defect in B-cell development can be specifically rescued by restoration of Ebf1 expression, indicating that the primary role of PU.1 in pro-B-cells is the stimulation of

Ebf1 transcription (Medina et al., 2004). Homozygous Bcl11a mutant mice also show a block in early B-cell development. Lymphoid-specific transcripts (including Ebf1 ,

Pax5, and Cd19 ) are not detectable, and productive Igh rearrangement is absent, indicating a block at the pre-pro-B-cell stage (Liu et al., 2003). Targeted knockdown

33 of either E2a or Ebf1 results in similar phenotypes, as the murine lymphoid series shows developmental arrest prior to the onset of D HJH rearrangement (Bain et al.,

1994; Medina et al., 2004). E2a and Ebf1 proteins act together to promote transcription of many B-cell specific genes, including Pax-5, which is considered to be the master regulator of B-cell development (Medvedovic et al., 2011; Sigvardsson et al., 1997). Pax-5 null mice show arrest of B-cell development prior to any evidence of B-lymphoid commitment in the fetal liver (Urbanek et al., 1994), although the developmental block seen in post-natal mice occurs at a later (pro-B-cell) stage. Pax-5 is critical for the maintenance of the B-cell state, mediating upregulation of B-cell specific genes (Nutt et al., 1998), and repression of non-lineage appropriate genes

(Souabni et al., 2002). Conditional inactivation of Pax-5 in a mouse model resulted in re-expression of many Pax-5-repressed genes (Delogu et al., 2006). In a proportion of cells, this derepression led to de-differentiation to the pro-B-cell stage, and some of these cells can acquire T-cell differentiation capacity in vivo (Cobaleda et al., 2007).

Of note, a subset of mice in which Pax-5 was conditionally inactivated developed leukaemia, while in humans, somatic deletion of PAX-5 is a common recurrent finding in childhood leukaemia (Mullighan et al., 2007).

A growing body of evidence indicates that B-cell development is controlled by the interdependent action of multiple transcription factors in a series of positive and negative feedback networks. For example, Pax-5 is a direct transcriptional target of

Ebf1, while Pax-5 in turn upregulates expression of Ebf1 (in company with many other B-cell-specific genes) in a positive feedback loop. In addition, Pax-5 inhibits expression of Notch-1, which is a of Ebf1 (Reviewed in Mandel and

34

Grosschedl, 2012). Ebf1 expression is also regulated by E2a, and by Ebf1 protein itself (Roessler et al., 2007). It is increasingly clear that dissection of this factor interplay will be critical for the full elucidation of transcriptional regulation of B- lymphopoiesis.

35

1.4 Ikaros

The Ikaros family of transcription factors comprises three members (Ikaros,

Aiolos and Helios) with predominantly haematopoietic expression (Molnar et al.,

1996; Schmitt et al., 2002) and the more widely-expressed Eos and Pegasus (Perdomo et al., 2000). The protein structure of all family members is characterised by the presence of two highly conserved Krüppel-type C2H2 zinc finger domains that mediate interaction with both DNA and other proteins. In Section 1.4.1, I will discuss how the structure of Ikaros (and its splice variants) contributes to protein function, cellular localisation and transcriptional activity.

Ikaros was the founder member of the protein family, having initially been discovered to bind to regulatory elements of genes involved in lymphoid development, including Dntt and Cd3d (Georgopoulos et al., 1992; Lo et al., 1991). Transcripts are detected at the very earliest stages of extra-embryonic (the Day 8 yolk sac), and levels of expression remain high in haematopoietic progenitors throughout fetal and adult life. Ikaros is present in erythroid and myeloid progenitors, but expression is greatly reduced following terminal differentiation of these cells.

Conversely, the levels of Ikaros expression remain high throughout all stages of lymphoid development, including the most mature forms (Georgopoulos et al., 1997).

This broad extent of expression at different stages of development in multiple cellular lineages suggests that Ikaros plays a key role in transcriptional regulation of normal haematopoetic differentiation (and lymphopoiesis in particular), and this hypothesis has been validated extensively in multiple murine models that are described in Section

1.4.2. Despite a wealth of experimental data, the precise means by which Ikaros

36 proteins act to regulate gene expression during haematopoieis remains poorly understood, and this issue will also be explored in this section of the introduction.

Importantly (in the context of this project), accumulating evidence supports the idea that Ikaros dysfunction is a key factor that contributes to poor treatment outcome in human leukaemia, and I will discuss these data in Section 1.4.3.

1.4.1 Ikaros protein structure and isoforms

Ikaros is encoded by the Ikzf1 gene (Georgopoulos et al., 1994), which consists of seven highly conserved coding exons (John et al., 2009). As with all Ikaros family members, DNA-binding occurs via four central zinc fingers, while protein dimerisation (which is critical for full activity) is mediated by a C-terminal region containing two zinc fingers (Georgopoulos et al., 1992; Hahm et al., 1994). A schematic representation of Ikaros protein structure is shown in Figure 1.4.1.

37

Figure 1.4.1 Schematic representation of Ikaros protein structure. Black boxes represent zinc fingers, and blue boxes demarcate coding exons. The known functions of protein structural domains are indicated.

38

Alternative splicing of Ikzf1 pre-mRNA leads to expression of at least ten isoforms of

Ikaros during normal lymphoid development (Mullighan et al., 2008), and the structure of some of these variants is depicted schematically in Figure 1.4.2.

Figure 1.4.2 Schematic representation of Ikaros isoforms 1 – 6. All isoforms retain two C-terminal zinc fingers, but differ in the number of central DNA-binding zinc fingers, as indicated.

39

IK1 and IK2 are the isoforms that are expressed most abundantly in mature lymphocytes and lymphoid precursors, while IK4 levels are highest in immature thymocytes; all other isoforms are expressed at low levels in haematopoietic cells

(Sun et al., 1996). The number of central zinc fingers varies between isoforms, and this confers differing DNA-binding properties. Importantly, splice variants which have fewer than two central zinc fingers do not bind DNA (Molnar and Georgopoulos,

1994).

All Ikaros isoforms retain the C-terminal domain, allowing dimerisation between different variants, and this may alter cellular localisation and transcriptional activity. Isoforms which normally exhibit predominantly cytoplasmic expression (i.e. all but IK1 and IK2, which are mainly nuclear) translocate to the nucleus following heterodimerisation with IK1 or IK2. Molecules which include DNA-binding isoforms exclusively (for example IK1/IK3 heterodimers) have equivalent transcriptional activity to IKI or IK2 homodimers. Critically, dimerisation of IK1 or IK2 with non-

DNA-binding isoforms (such as IK6, which lacks all four central zinc fingers) exerts a dominant negative effect, by sequestering DNA-binding proteins in complexes that cannot activate transcription (Sun et al., 1996).

Other described structural regions include a bipartite activation domain, which is contained within the first 81 amino acids encoded by exon 7, upstream of the C- terminal dimerisation region. This domain consists of two distinct acidic and hydrophobic regions, and is believed to mediate the interaction of Ikaros with the preinitiation complex, when Ikaros is bound to DNA (Georgopoulos et al., 1997). In addition, a 20 amino acid region located upstream of the central zinc fingers in Ikaros

40 isoform IK-H (which to date has been described to exist only in humans), appears to enhance chromatin binding of other isoforms following heterodimerisation with IK-H

(Ronni et al., 2007).

1.4.2 Ikaros function

Experiments in genetically modified mice, which express Ikaros proteins that lack specific functional domains, have illustrated its critical role at multiple levels of haematopoietic development. Several of these murine models have been described, and I will discuss three key examples.

The first model to be described involved a targeted deletion of Ikzf1 exons 3 and 4 (which encode three central zinc fingers), that results in expression of a non-

DNA-binding dominant negative protein ( Ikaros DN ). These mice have a complete block in lymphoid development, involving T-, B- and Natural Killer (NK) cells and their respective progenitors, including the very earliest forms (Georgopoulos et al.,

1994). Mice that are heterozygous for the same deletion develop acute T-cell leukaemia/ lymphoma, which is fatal at an early age. Of note, T-lymphocytes isolated from pre-leukaemic animals are hyperproliferative in vitro , while progression to overt leukaemia is associated with somatic loss of the remaining wild-type Ikzf1 allele

(Winandy et al., 1995).

Deletion of the C-terminal dimerisation domain (which causes protein instability and results in effective absence of Ikaros expression) causes a complete block in fetal and adult B-lymphopoiesis, and fetal T-lymphopoiesis. These mice do however have normal total numbers of T-lymphocytes in adult life, although with

41 abnormal distribution of T-cell subsets (including a skew towards CD4+CD8- expression and a gross reduction in thymic dendritic antigen presenting cells). These mice also demonstrate hyperproliferative T-cell receptor responses in vitro (Wang et al., 1996).

Myeloid and erythroid development are largely uncompromised in these two mouse models, however a strain ( Plastic ) in which a point mutation disrupts DNA- binding has severe defects in fetal erythropoietic differentiation, which is fatal in utero

(Papathanasiou et al., 2003). These mice have also subsequently been shown to have defective self-renewal of the long-term reconstituting haematopoietic stem cell (HSC) compartment (Papathanasiou et al., 2009).

Although these critical roles for Ikaros in haematopoiesis are well-established, the precise mechanisms by which its transcriptional effects are exerted remain poorly elucidated. Few direct targets of Ikaros have been described, though these include several genes that are critical for lymphopoiesis such as Dntt (Georgopoulos et al.,

1992), Rag (Yannoutsos et al., 2004) and Igll1, which encodes the λ5 component of the surrogate light chain expressed at the large pre-B-cell stage of development

(Sabbattini et al., 2001). In addition, Ikaros has been shown to negatively regulate the

G1/S phase transition of the cell cycle of pre-B lymphocytes. This effect has been shown to be mediated by binding of Ikaros to the c- promoter, which results in gene silencing and subsequent inhibition of cellular proliferation that is necessary for progression to the small pre-B-cell stage (Ma et al., 2010).

Ikaros has been shown to be involved in both activation and repression of genes in vivo and in transfection assays (Harker et al., 2002; Molnar and

42

Georgopoulos, 1994; Sabbattini et al., 2001; Sun et al., 1996). This divergent activity even extends to individual genetic elements, as Ikaros can either activate or repress transcription of the upstream regulatory element of Sfpi1 (which encodes PU.1), depending on cellular context (Zarnegar and Rothenberg, 2011). The repressive effects of Ikaros have been suggested to be mediated by interaction with chromatin modifiers, as physical interactions with the Mi-2/nu cleosome remodeling and deacetylase (NuRD) complex (Kim et al., 1999) and the PYR complex (O'Neill et al.,

2000) have been demonstrated. In addition, Ikaros has been shown to co-localise with silent genes at transcriptionally inactive pericentric heterochromatin (Brown et al.,

1997), although the mechanistic relevance of this finding remains uncertain, as known transcriptional activators (such as C/EBP) have been shown to exhibit similar distribution (Tang et al., 1999). An additional factor that probably contributes to transcriptional repression in vivo is the observed overlap of Ikaros binding sites with those of transcriptional activators. For example, Ikaros occupancy of the Dntt promoter blocks binding of the transcriptional activator Ets1 at the same location, and disruption of Ikaros binding by mutagenesis of this locus prevents downregulation of

Dntt promoter activity in stable transfection assays (Trinh et al., 2001).

A further level of complexity is provided by the fact that Ikaros family proteins can combine to regulate gene expression in a developmental stage-specific manner.

This is exemplified by the regulation of Igll1 expression during B-lymphopoiesis. At the large pre-B-cell stage, the extent of Igll1 expression is governed by the relative ratios of Ikaros (repressive) and Ebf1 (activating) proteins. Exit from this stage of development is dependent on binding of the Ikaros family member Aiolos to the Igll1

43 promoter, which results in transcriptional silencing, exit from the cell cycle and transition to the small pre-B-cell stage (Thompson et al., 2007).

44

1.4.3 Ikaros dysfunction in human leukaemia

Evidence from the aforementioned murine models confirmed Ikaros as a lymphoid tumour suppressor in mice. Reports of Ikaros dysfunction in human leukaemia were initially confined to reports of several small case series, that identified aberrant Ikaros isoform expression in subsets of human B-ALL (Nakase et al., 2000;

Ruiz et al., 2004) and acute myeloid leukaemia (Yagi et al., 2002). These cases were found to overexpress the non-DNA-binding dominant negative isoform IK6, and this was originally thought to occur as a result of abnormal RNA splicing. More recent work (predominantly from genome-wide analysis of primary human ALL samples using single nucleotide polymorphism (SNP) arrays) has however shown that IK6 expression predominantly occurs as a result of genomic deletions of IKZF1 . These highly important findings (which have critical implications for the results detailed in this thesis) will be discussed in depth in Section 1.6.2.

45

1.5 Forkhead Box Protein P1 (Foxp1)

Forkhead box (Fox) proteins are a family of evolutionarily conserved transcription factors which are characterised by a 110 amino acid forkhead (also known as ‘winged helix’) region which mediates DNA binding (Carlsson and

Mahlapuu, 2002). This domain was originally defined by homology with the protein product of the Drosophila forkhead gene (Weigel and Jackle, 1989), and consists of three alpha helices, three beta sheets and two loops or turns which form a HLH-like motif. DNA binding is believed to occur primarily through the third alpha helix (to the

DNA major groove) and the second wing (to the DNA minor groove) (Clark et al.,

1993).

The Fox protein family currently comprises 19 subfamilies (A to S), which are categorised by the position of certain conserved amino acids in the forkhead domain

(Wijchers et al., 2006). Fox proteins perform key functions in a diverse range of cellular processes, including metabolism, differentiation, proliferation and

(Myatt and Lam, 2007). The four members of the Foxp subfamily (Foxp1-4) play essential roles in development and immunity in mammals (Santos et al., 2011).

Forkhead box protein P1 (Foxp1 in mice, FOXP1 in humans) was the first member of the subfamily to be identified, having been cloned from a mouse B-cell leukaemia line

(Lai et al., 1993). It is widely expressed in mammalian tissues (with highest levels in lymphocytes, neural tissue and gut) and is a critical transcriptional regulator during B- lymphopoiesis (Banham et al., 2001; Shu et al., 2001; Wang et al., 2003). The other

FoxP subfamily members have primary roles in central nervous system development

46

(Foxp2), Regulatory T-lymphocyte function (Foxp3) and pulmonary and gut development (Foxp4) (Lu et al., 2002; Shu et al., 2007; Yagi et al., 2004).

In this part of the introduction, I will describe how the structure of Foxp1 and its splice variants contributes to protein function (Section 1.5.1). I will then discuss the experimental evidence that has revealed a critical role for Foxp1 in haematopoiesis, and the potential means by which its transcriptional effects are exerted (Section 1.5.2). In Section 1.5.3, I will review the increasing body of data which implicates FOXP1 in the pathogenesis of human lymphoid malignancy.

1.5.1 Foxp1 protein structure and isoforms

The Foxp1 gene consists of 16 coding exons, and several structural domains have been described. A schematic representation of Foxp1 protein structure is shown in Figure 1.5.1.

Figure 1.5.1 Schematic representation of Foxp1 protein structure. Boxes demarcate defined protein structural regions (described in the lower portion of the figure). The known functions of the indicated domains are listed in the upper part of the figure.

47

In common with other members of the Foxp subfamily, Foxp1 exhibits truncation of the forkhead region at the helices of the first two ‘wings’ of the domain

(Chatila et al., 2000). Several isoforms of the protein are expressed (at least 11 in mice, and at least 9 in humans). The relative distribution of these isoforms varies according to cell type, and this may contribute to tissue-specific transcriptional activity. For example, a Foxp1 splice variant which is expressed specifically in murine embryonic stem (ES) cells preferentially stimulates the transcription of genes required for ES cell pluripotency (Gabut et al., 2011). A schematic representation of the structure of some Foxp1 splice variants is shown in Figure 1.5.2.

Figure 1.5.2. Schematic representation of Foxp1 splice variants . Foxp1A is the predominant Foxp1 isoform expressed during normal lymphoid development. Foxp1C and Foxp1D (which are expressed following B-cell activation) lack one or both of the polyglutamine and glutamine-rich regions which are important for modulation of Foxp1 transcriptional activity.

48

All Foxp1 isoforms contain leucine zipper and zinc finger motifs that mediate protein dimerisation, which in turn is necessary for efficient DNA binding (Wang et al., 2003). The longest isoforms (Foxp 1A and Foxp1B in mice) include an N-terminal polyglutamine region which is involved in modulation of transcriptional repressive activity (Wang et al., 2003). The structure of the C-terminal acidic region predicts a role in protein transactivation (Banham et al., 2001). Patterns of Foxp1 isoform expression also appear to vary during normal B-cell development. For example, cellular levels of shorter isoforms which lack the polyglutamine region (Foxp1C and

Foxp1D in mice, isoforms 3 and 9 in humans) increase following B-cell activation

(Brown et al., 2008).

1.5.2 Functions of Foxp1

In murine models, homozygous deletion of Foxp1 is lethal in utero due to heart valve and cardiac outflow tract defects (Shu et al., 2001; Wang et al., 2004).

Reconstitution experiments using fetal liver haematopoetic precursor cells (HPCs) from Foxp1 -null mice subsequently revealed a critical role for Foxp1 in B- lymphopoiesis. Following sublethal irradiation and reconstitution with Foxp1 -null

HPCs, recipient mice had an arrest of B-lymphopoiesis at the transition from the pro- to large pre-B-cell stages. This developmental block was associated with decreased expression of recombination activating genes 1 and 2 ( Rag1 and Rag2), which are required for immunoglobulin gene recombination and progression to the large pre-B- cell stage. Decreased Rag expression was shown to be linked to a loss of binding of

Foxp1 to the B-cell specific Erag enhancer (Hu et al., 2006).

49

Experimental evidence implicates Foxp1 in the regulation of cell turnover, with effects that appear to be tissue-specific. For example, Foxp1 has been shown to be essential for the generation and maintenance of quiescence in T-lymphocytes.

Conditional deletion of Foxp1 at the CD4+ CD8+ double positive (DP) stage of thymocyte development resulted in premature T-cell activation and acquisition of effector function in a mouse model (Feng et al., 2010). Recently, this effect was shown to be mediated by transcriptional repression of the IL-7 receptor α-chain gene

(IL7r α) by Foxp1, which competes with Foxo1 (a transcriptional activator) for binding to a forkhead recognition site in the IL7r α locus (Feng et al., 2011). Foxp1 has also been shown to inhibit proliferation in developing murine cardiomyocytes by repressing Nkx2.5 transcription (Zhang et al., 2010). Conversely, Foxp1 has been described to stimulate B-lymphocyte proliferation through an interaction with microRNA-34a (miR-34a). Overexpression of miR-34a was shown to block lymphocyte development, resulting in a similar phenotype to that seen with Foxp1 null mice. This effect was shown to be mediated through direct inhibition of Foxp1 expression by mir-34a, while targeted knockdown of miR-34a in a murine model resulted in increased Foxp1 protein expression and increased numbers of mature B- cells (Rao et al., 2010).

Little investigation of the mechanism of Foxp1-mediated transcriptional repression has been reported. Several groups have described physical interaction of

Foxp1 with transcriptional repressor complexes. For example, Foxp1 has been shown to recruit the corepressor protein C-terminal binding protein 1 (CtBP-1) (Li et al.,

2004), and to associate with the SMRT nuclear corepressor at the promoter of the p21

50 gene (Jepsen et al., 2008). Foxp1 (in addition to Foxp2 and Foxp4) was also shown to interact directly with the p66 β component of the NuRD complex, and to repress gene transcription by a mechanism requiring histone deacetylation by the enzymes histone deacetylase 1 and 2 (HDAC1 and HDAC2) (Chokas et al., 2010).

1.5.3 FOXP1 dysfunction in human lymphoid malignancy

Deregulated expression of several subfamilies of FOX proteins is observed in a wide range of neoplasia (Reviewed in Myatt and Lam, 2007). FOXP1 was initially considered to be a tumour suppressor in humans, as it is located in a chromosomal region (3p14.1) which exhibits loss of heterozygosity in several human malignancies.

FOXP1 mRNA expression was found to be diminished (compared with the normal tissue equivalent) in many epithelial carcinomas (Banham et al., 2001), while reduced levels of nuclear FOXP1 were described to be associated with a poor prognosis in breast cancer patients (Fox et al., 2004).

More recent evidence however suggests an opposing (i.e. oncogenic) role for

FOXP1 in the context of lymphoid malignancies. Recurrent translocation of FOXP1

(predominantly involving the IGH locus as a translocation partner) was described in several subtypes of non-Hodgkin’s lymphoma (NHL), including Mucosa-Associated

Lymphoid Tissue (MALT) Lymphoma (Streubel et al., 2005), and Diffuse Large B-

Cell Lymphoma (DLBCL) (Wlodarska et al., 2005). Subsequently, increased expression of shorter FOXP1 isoforms (isoforms 3 and 9, which are analogous to murine Foxp1C and Foxp1D) have been strongly linked to a poor prognosis in the activated B-cell subtype of DLBCL (Brown et al., 2008). FOXP1 has now also been

51 reported as a recurrent (although rare, with only five reported cases) translocation partner of PAX-5 in both adult and childhood B-ALL (Coyaud et al., 2010; Mullighan et al., 2007; Put et al., 2011). Taken together, these data suggest that deregulation of

FOXP1 activity frequently contributes to transcriptional dysfunction in human lymphoid malignancy.

52

1.6 B-cell Acute Lymphoblastic Leukaemia

Leukaemia (literally ‘white blood’, from the Greek) comprises a range of disorders of leukocyte proliferation and differentiation, characterised by the accumulation of malignant white blood cells and compromise of residual bone marrow function that results in life-threatening susceptibility to infection and bleeding

(Pui, 1996). At the broadest level, leukaemias are classified by lymphoid or myeloid origin. ‘Acute’ leukaemia refers to the rapid proliferation of developmentally-arrested immature progenitors (blasts), while ‘chronic’ leukaemia involves the accumulation of more differentiated cells. In this section of the introduction, I will review the current knowledge of the biology of B-lymphoid acute lymphoblastic leukaemia (ALL).

Although ALL is an uncommon disease, with 600-700 cases per year in this country (UK Office for National Statistics), its clinical importance relates to its relatively high incidence in young people, and poor prognosis in older patients. ALL is the commonest malignancy and the leading cause of non-traumatic death in children and young adults (Advani et al., 2009), despite treatment advances in recent decades.

Children diagnosed with leukaemia in the UK currently have at least an 80% chance of long-term cure overall, and survival rates exceed 90% in some cytogenetic subgroups (Bhojwani et al., 2011; Pui and Evans, 2006). The outlook for adult patients is dismal by comparison, with corresponding rates of treatment success of approximately 40% (Pui et al., 2008). In the main, this reflects a higher relative incidence of leukaemia-associated genetic aberrations associated with poor outcome

(discussed in Sections 1.6.1 and 1.6.2), and a poorer tolerance of treatment-related side-effects.

53

Modern categorisation of haematological malignancy (as typified by the current World Health Organisation (WHO) proposed classification system) relies on a combination of cellular morphology, immunophenotype and genetic analysis to discriminate disease subtypes. In the WHO classification system, the immunophenotype of precursor B-ALL corresponds to that of lymphoid progenitors up to and including the immature B-cell stage. Logically, leukaemic blasts may be considered to be the malignant counterparts of haematopoietic precursors in which normal developmental processes have been subverted by acquired genetic anomaly.

Investigation of the downstream molecular effects of these genetic aberrations is therefore critical for understanding the biology of the disease. In this section of the introduction, I will discuss how some of the abnormalities that are most commonly seen in ALL contribute to lymphoid differentiation arrest. The latter part of this section will focus specifically on Philadelphia-positive ALL (that is characterised by the presence of BCR-ABL translocation), as deletions in IKZF1 that typify these cases have direct relevance for the experimental results of this project.

54

1.6.1 Genetic abnormalities in B-ALL

The incidence of recurrent structural chromosomal abnormalities that are seen in childhood and adult B-ALL is depicted in Figure 1.6.1. The vast majority of these involve numerical anomalies (predominantly hyperdiploidy), or translocations that result in aberrant expression or regulation of transcription factors, tyrosine kinases or proto-oncogenes. It is likely that there is much overlap in the downstream effects of these genetic events, and it is evident that blast cell accumulation results from alterations in molecular pathways involved in cellular self-renewal, proliferation, differentiation, or apoptosis (Hanahan and Weinberg, 2000). For example, leukaemias which harbour diverse fusion transcription factors frequently exhibit similar activation of ( HOX ) genes, which are important for cellular differentiation and proliferation (Buske and Humphries, 2000).

55

,

Figure 1.6.1 The incidence of recurrent structural chromosomal abnormalities in childhood and adult ALL. Modified with permission from Pui et al, N Engl J Med 350 ; 15: 1535-48, Copyright Massachusetts Medical Society.

56

It is instructive to consider the mechanism by which transcription factor fusion genes contribute to leukaemogenesis, via subversion of normal protein function. The genes involved typically encode transcription factors that are developmentally regulated and affect cellular differentiation (Armstrong and Look, 2005). A cardinal example is the ETV6-RUNX1 (formerly TEL-AML1 ) fusion gene that results from the t(12;21) (p13;q22) translocation (Speck and Gilliland, 2002). RUNX1 normally acts as a transcriptional activator, and is critical for normal haematopoiesis (Ichikawa et al., 2004). Fusion to ETV6 (a gene that is also essential for haematopoiesis) induces aberrant recruitment of histone deacetylases, thereby converting RUNX1 into a transcriptional repressor (Downing, 2003). Profiling studies have shown that leukaemias that harbour ETV6-RUNX1 translocations have highly similar gene expression patterns, indicating a common mechanism of leukaemogenesis through transcriptional dysregulation that is caused by the fusion protein (Yeoh et al., 2002).

Recent advances in the speed and resolution of profiling technology have allowed comprehensive assessment of genetic alterations in leukaemic cells, thereby revolutionising our understanding of the ALL genome. The vast majority of data generated from these studies have come from analysis of childhood ALL cases using single nucleotide polymorphism (SNP) arrays (Mullighan, 2011). The key finding from these studies has been the identification of recurrent anomalies (deletions, mutations and translocations, that are mostly heterozygous) in genes that encode transcriptional factors which are critical for B-cell development, including PAX5 ,

EBF1 and IKZF1 (Kawamata et al., 2008; Kuiper et al., 2007; Mullighan et al., 2007).

The significance of the IKZF1 deletion is discussed in detail in Section 1.6.2.1. Over

57 two thirds of cases of childhood B-ALL have alterations that are predicted to disrupt differentiation at early stages of lymphoid commitment (see Section 1.3), which is consistent with the most commonly observed leukaemic immunophenotype of pro-B- cell developmental arrest.

These studies have also identified several additional recurrent lesions that are likely to contribute to leukaemogenesis, and treatment response. Deletions involving genes that encode tumour suppressor and cell cycle regulatory proteins ( CDKN2A/B ,

PTEN and RB1 ), lymphoid signalling factors ( CD200 , BTLA , BLNK ) and molecules involved in drug response ( NR3C1 ) have all been described. An additional important finding was that the number of additional (and presumably cooperative) lesions varies markedly with cytogenetic subtype. For example, the incidence of additional genetic lesions in Mixed Lineage Leukaemia gene ( MLL )-rearranged leukaemia is low, indicating that MLL rearrangement in itself may be sufficient to induce leukaemia

(Bardini et al., 2010). In contrast, ETV6-RUNXI + leukaemias typically contain multiple additional abnormalities (Mullighan et al., 2007).

1.6.2 Philadelphia-positive (BCR-ABL+) ALL

The discovery of the Philadelphia (Ph) chromosome in 1960 (by Peter Nowell and David Hungerford, who carried out their research in Philadelphia) was a seminal moment in cancer genetics, as it constituted the first description of a leukaemic fusion oncogene. This derivative chromosome 22 results from the t(9;22) (q34;q11) translocation which causes juxtaposition of the ABL gene on chromosome 9 with the

58 breakpoint cluster region ( BCR) on chromosome 22. The translocation is depicted schematically in Figure 1.5.2. The resultant fusion gene encodes the oncogenic BCR-

ABL tyrosine kinase.

The Ph chromosome is found in the vast majority (at least 95%) of cases of chronic myeloid leukaemia (CML), and 25% of adult ALL (it is much less common in childhood ALL, comprising 3-5% of cases) (Pui, 1996). The location of breakpoint cluster region ( BCR ) breakpoints varies between CML and ALL, and the fusion proteins are either 210kDa (p210) or 190kDa (p190) respectively (Clark et al., 1987).

Tyrosine kinase activity is well-characterised (Van Etten et al., 1989), but the precise function of the normal product of the ABL gene is not known, although it appears to play a role in DNA damage response (Kharbanda et al., 1995). Fusion to

BCR elements results in increased kinase activity, aberrant cytoplasmic localisation

(Lugo et al., 1990), and activation of multiple pro-proliferative pathways, including

RAS/ MAPK and JAK/STAT (Van Etten, 2002).

59

Figure 1.6.2 The Philadelphia Chromosome. The ABL and BCR genes are located on chromosomes 9 and 22, respectively. The t(9;22) translocation results in the formation of the BCR-ABL fusion gene in the derivative chromsome 22 (the Philadelphia chromosome). Adapted by permission from Macmillan Publishers Ltd (Lydon, Nature Medicine 15 , 1153-7), copyright 2009.

BCR-ABL+ ALL has traditionally been associated with a dismal long-term prognosis (Fletcher et al., 1991). The development of targeted tyrosine kinase inhibition with imatinib has provided recent hope of improved cure rates, and early outcome data indicates that addition of this agent to intensive paediatric-based treatment protocols may improve survival appreciably (Hunger, 2011). The precise mechanism by which BCR-ABL positivity contributes to adverse outcome is unclear.

In children, BCR-ABL+ ALL is associated with adverse prognostic features such as high leukocyte count at diagnosis and an increased incidence of central nervous system involvement (Crist et al., 1990). In addition, BCR-ABL has been shown to

60 exert an anti-apoptotic effect via upregulation of Bcl-xL (Skorski, 2002), that is likely to contribute to enhanced cellular survival and chemotherapeutic resistance in vivo .

High-resolution characterisation of the BCR-ABL+ ALL genome has now revealed that IKZF1 deletion is the key event that is associated with poor outcome in this disease.

1.6.2.1 BCR-ABL+ ALL and IKZF1 deletion

SNP microarray profiling has shown that IKZF1 deletions occur in at least 84% of cases of BCR-ABL+ ALL (Iacobucci et al., 2009; Mullighan et al., 2008).

Approximately half of these heterozygous anomalies involve deletion of exons 3-6, and result in expression of the dominant negative IK6 isoform (discussed in Section

1.4.1). The presence of these deletions is associated with an adverse prognosis within the BCR-ABL+ ALL cohort (Martinelli et al., 2009), and were subsequently also found to be associated with poor outcome (with up to three-fold increased risk of relapse) in BCR-ABL negative (BCR-ABL-) ALL (Kuiper et al., 2010; Mullighan et al., 2009b; Yang et al., 2011). Intriguingly, these cases exhibit gene expression profiles that are very similar to IKZF1 -deleted BCR-ABL+ ALL. Increased expression of stem cell-associated genes (e.g. MPL ) and reduced expression of B-cell signalling genes (e.g. CD19 ) suggests that IKZF1 deletion contributes directly to differentiation arrest in these cases (Mullighan and Downing, 2008). It is notable that IKZF1 -deleted

BCR-ABL- cases commonly have additional alterations in genes coding for cytokine receptors (e.g. CRLF2) and tyrosine kinases (including Janus-associated kinases 1-3)

(Harvey et al., 2010; Mullighan et al., 2009a). Mechanistically, this suggests that

61 molecular targets of BCR-ABL are also activated by dysregulation of these pathways, although this has not been proved experimentally as of yet.

Further evidence of a key role for Ikaros dysfunction in human leukaemia is provided by several additional reports. Forced expression of IK6 in human umbilical cord stem cells has been shown to cause impairment of B-lymphoid maturation in vitro (Tonnelle et al., 2001; Tonnelle et al., 2009). In addition, data from genome- wide association studies of several patient cohorts indicates that germline polymorphism of IKZF1 is associated with an increased lifetime risk of ALL development (Papaemmanuil et al., 2009; Prasad et al., 2010; Trevino et al., 2009)

62

1.7 Aim of the Project

As detailed so far, successful B-lymphopoiesis depends upon the actions of a series of interdependent transcription factors that regulate developmental gene expression. The subversion of these processes results in B-ALL, as exemplified by common recurrent genetic abnormalities that directly inhibit normal transcription factor function ( ETV6-RUNX1 ), and that have been shown to contribute directly to differentiation arrest ( IKZF1 deletion).

The aim of this project was to study the interaction between two factors that are critical for B-lymphopoiesis (namely Ikaros and Foxp1), and to elucidate the functional importance of this interaction in normal B-cell development and B-ALL .

63

Chapter 2

Materials and Methods

2.1 Cell cultures

0 All cells were grown at 37 C with 5% atmospheric CO 2.

2.1.1 Human Embryonic Kidney (HEK) 293T cells

293T cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) with

10% fetal calf serum (FCS) (volume/volume) and L-Glutamine (2mM).

2.1.2 B3 cells

B3 cells were originally derived from a lymphoma, which was isolated from a transgenic mouse that expressed the IL-7 gene under the control of the E α (MHC class II) promoter (Fisher et al., 1995). Cells were maintained in Iscove’s Modified

Dulbecco’s Medium (IMDM) supplemented with 10% FCS, L-Glutamine (2mM), β- mercaptoethanol (50µM) and IL-7 (2ng/ml).

2.1.3 Acute Lymphoblastic Leukaemia cell lines

The REH cell line was purchased from Deutsche Sammlung von

Mikroorganismen und Zellkulturen (DSMZ). The SUP-B15 line was a gift from Dr.

Owen Williams and Dr. Jasper de Boer (Institute of Child Health, University College

London). Both cell lines were grown in RPMI with 20% FCS and L-Glutamine

(2mM).

64

2.2 Plasmid constructs

2.2.1 Expression vectors used for 293T cell transfection

Modified PCDNA expression vectors containing two FLAG epitopes (PDF-

DNAIII) or a HA epitope (PCDNA3.1-HA) were gifts from the Lymphocyte

Development Group, MRC Clinical Sciences Centre. Vector maps are shown in

Figure 2.1. A high-fidelity DNA polymerase (PFU Turbo, Stratagene) was used for all

PCR amplifications.

The cDNAs that encode murine WT and mutant Ikaros proteins were amplified by PCR from the corresponding retroviral plasmid (Section 2.2.3) and inserted between the XhoI and XbaI restriction sites of the PCDNA3.1-HA vector (Figure 2.1).

The corresponding human constructs were amplified by PCR from cDNA isolated from the REH cell line. The cloning of human IK6 cDNA involved a two step process .

5’ (corresponding to IKZF1 exons 1 and 2) and 3’ (corresponding to IKZF1 exon7) fragments were amplified separately. Amplification products had complementary sequences which allowed annealing of the 3’end of the exon 1/2 fragment to the 5’ end of the exon 7 fragment. The fragments were annealed and amplified again in order to generate the complete IK6 cDNA. The integrity of all constructs generated by PCR was confirmed by sequencing of the inserts.

The cDNAs of murine WT and mutant Foxp1 proteins were either amplified from a retroviral plasmid or purchased commercially (GeneCust, Luxembourg or Blue

Heron Biotechnology, Washington, USA) and were cloned between the BamHI and

XhoI restriction sites of the PDF-DNAIII vector (Figure 2.1). Human FOXP1 cDNA

65 was amplified by PCR from the REH cell line, and was inserted between the BamHI and XhoI restriction sites of the PDF-DNAIII vector.

2.2.2 Expression vectors used in generation of GST-tagged proteins

cDNA fragments were obtained either by PCR amplification or digestion of the corresponding PCDNA plasmid (Section 2.2.1) and were inserted between the BamHI and XhoI restriction sites of the pGEX-4T vector (GE Healthcare and vector map shown in Figure 2.1).

2.2.3 Retroviral vectors used for B3 cell infection

Retroviral vectors were derived from the pMSCV-IRES-GFP (pMIG) plasmid, and cDNAs were inserted between the BglII and XhoI restriction sites (Figure 2.1).

The pMIG-Ikaros-HA construct was generated by the Lymphocyte Development

Group, MRC Clinical Sciences Centre. The pMIG-FLAG-Foxp1 plasmid was generated following amplification of Foxp1 cDNA from the PDF-DNAIII-Foxp1 vector, while the cDNA of murine IK6 was synthesised commercially (GeneCust,

Luxembourg).

66

A

B

67

C

D

Figure 2.1 Maps of Expression Vectors used in transfection and infection experiments. A. PDF-DNAIII. B. PCDNA3.1-HA. C. PGEX-4T-1. D. PMSCV- IRES-GFP. The locations of restriction enzyme recognition sequences in the multiple cloning site are shown.

68

2.3 Transfection of HEK 293T cells 293T cells were transfected using the calcium phosphate method (Nature

Methods, 2005). Briefly, HEPES Buffered Saline (pH 7.05) was added to a solution containing 4µg of plasmid DNA in the presence of 2.5M CaCl 2 and incubated for 5 minutes at room temperature. For co-transfections, 4µg of each plasmid were added.

The mixture was then added to HEK 293T cells cultured on a 10cm plate at approximately 50% confluence. The transfection mix was replaced with fresh 293T culture medium 24 hours later, and protein lysates were extracted 48 hours after transfection (see Section 2.4).

2.4 Preparation of protein extracts

All lysis buffers were supplemented with Complete Mini, EDTA-free Protease

Inhibitor (Roche). Following extraction, the concentration of protein extracts was determined using the Bradford method (Bradford, 1976), after which the lysate was snap frozen in liquid nitrogen then stored at -80 0C until required.

2.4.1 Whole cell extracts

In order to prepare whole cell extracts, approximately 10 7 cells were washed once with phosphate buffered saline (PBS), and the cell pellet was suspended in lysis buffer (40mM Tris pH 8.0, 200mM NaCl, 0.1% NP40, 40mM NaF, 2mM NaVO 3,

200µM DTT, 20% Glycerol and bovine serum albumen (BSA) 1µg/ml). The solution was snap frozen in liquid nitrogen and then subjected to three consecutive freeze-thaw cycles. Extracts were recovered from the supernatant following centrifugation at

10,000 x g at 4 0C for 10 minutes.

69

2.4.2 Nuclear extracts

Nuclear extracts were prepared by a two-stage method. Approximately 10 7 cells were washed once in PBS. The cell pellet was initially resuspended in a low osmolarity buffer (10mM HEPES pH 7.9, 10mM KCl, 0.1mM EDTA and 0.1mM

EGTA) and allowed to swell on ice for 15 minutes, in order to lyse the cell outer membranes. After addition of NP40 (to a concentration of 0.6%), the solution was mixed vigorously by vortexing, and subjected to centrifugation at 10,000 x g at 4 0C for 30 seconds. The nuclear pellet was recovered and resuspended in a high osmolarity nuclear lysis buffer (20mM HEPES pH 7.9, 400mM NaCl, 1mM EDTA and 1mM

EGTA) and incubated with shaking at 4 0C for 30 minutes. Nuclear extracts were recovered in the supernatant following centrifugation at 10,000 x g at 4 0C for 5 minutes.

2.5 Reagents used for protein co-immunoprecipitation and immunoblotting.

2.5.1 Primary .

The following primary antibodies were used for both protein co- immunoprecipitation (co-IP) and immunoblotting: monoclonal anti-FLAG M2

(Sigma-Aldrich), monoclonal anti-HA (Covance), monoclonal anti-Foxp1 (Abcam), polyclonal anti-Ikaros (generated by the Lymphocyte Development Group, MRC

Clinical Sciences Centre). Normal mouse IgG and normal rabbit IgG (both Santa Cruz

Biotechnologies) were used as controls for non-specific precipitation during protein co-IP (Section 2.6).

70

2.5.2 Secondary antibodies used for immunoblotting.

Goat anti-mouse and goat anti-rabbit horseradish peroxidase-conjugated antibodies (both Santa Cruz Biotechnologies) were used as secondary antibodies for immunoblotting.

2.5.3 Western blotting

Protein extracts were resolved by polycrylamide gel electrophoresis, then transferred to polyvinylidene fluoride membranes and probed with the relevant antibodies. The membranes were developed with Western Lightning Plus ECL chemiluminescence reagent (PerkinElmer) and exposed to Amersham Hyperfilm ECL

(GE Healthcare).

For quantification of proteins according to band intensity, the membrane was developed with ECL reagent and a digital image was taken using an ImageQuant LAS

4000 Mini machine (GE Healthcare). The image was exported to ImageQuant software (Fuji) and band intensity was quantified and normalised to a protein loading control ( β-tubulin).

2.6 Protein co-immunoprecipitation (co-IP)

0.5 µg of antibody was coupled to 50µl of Protein G Dynabeads (Sigma-

Aldrich) by incubation with rotation at room temperature for 10 minutes. The beads were washed with buffer (0.1M Na 2HPO 4, pH 8.0) to remove unbound antibody from the solution. Protein extracts were then incubated with the antibody-coupled beads for

30 minutes at room temperature, after which the beads were washed three times with

PBS using the magnetic separator. In selected experiments, protein lysates were

71 subjected to treatment with either DNase (Roche) or ethidium bromide according to previously described protocols (Lai and Herr, 1992) prior to IP, in order to abolish

DNA-protein interactions.

Immunocomplexes were eluted by boiling the beads in 1x NuPAGE LDS

Sample Buffer (Invitrogen) for 10 minutes at 70 0C with shaking. Eluates were resolved by SDS-polyacrylamide gel electrophoresis, transferred to polyvinylidene fluoride membranes and probed with the relevant antibodies. 5% of the lysate amount that was used for IP was loaded on the same polyacrylamide gel, in order to provide an ‘input’ comparison.

2.7 GST pulldown assays

2.7.1 Preparation of GST fusion proteins

BL21-CodonPlus Competent Cells (Promega) were transformed (according to the manufacturer’s protocols) with pGEX-4T plasmids that contain the relevant cDNA. Several bacterial strains (RPL, RIL and RP) were used, and these had similar levels of ultimate protein recovery. Gene expression was stimulated by addition of

Isopropyl β-D-1-thiogalactopyranoside (IPTG, at a concentration of 1mM) when the culture medium had an optical density of 0.6, in order to coincide with the logarithmic phase of bacterial growth. The cells were then incubated with shaking at either 37 0C for 3 hours, or 16 0C for 24 hours; the latter protocol was found to yield higher levels of stable protein.

The cell cultures were centrifuged at 10,000 x g for 15 minutes, and bacterial pellets were suspended in dilution buffer (20mM Tris pH 7.9, 100mM NaCl, 10%

72 glycerol, 0.2mM EDTA and 0.1% Triton X-100, supplemented with Complete Mini,

EDTA-free Protease Inhibitor (Roche)). GST-tagged proteins were solubilised by sonication using a Branson Digital Sonifier at 20% amplitude for 2 separate sonications of one minute each. GST proteins were then recovered from the supernatant following centrifugation at 10,000 x g at 4 0C for 30 minutes. Relative quantification of GST fusion proteins was determined by Coomassie staining of resolved polyacrylamide gels.

2.7.2 Pulldown assay

In vitro transcription and translation of 250ng of plasmid DNA (using the corresponding PCDNA expression vector) was achieved by incubation with 1µl of 35 S

Methionine (1000 Ci/mmol at 10mCi/ml) and TNT T7 Quick Master Mix (Promega) for one hour at 30 0C.

35 S -labelled proteins were incubated with purified GST proteins in dilution buffer (20mM Tris pH 7.9, 100mM NaCl, 10% glycerol, 0.2mM EDTA and 0.1%

Triton X-100) with rotation overnight at 4 0C. The solution was then added to glutathione-coated paramagnetic beads (Piercenet, 50µl per assay) for one hour at 4 0C.

The beads were washed five times with dilution buffer, and complexes were eluted by boiling in 1x NuPAGE SDS Sample Buffer (Invitrogen) for 3 minutes at 100 0C.

Eluted materials were resolved by polyacrylamide gel electrophoresis, and 1% of the original 35 S-labelled protein solution was run as an ‘input’ comparison. The gel was dried and exposed to x-ray film (GE Healthcare) to detect radioactive signal.

73

2.8 Retroviral infection of B3 cells

HEK 293T cells were co-transfected with pMIG retroviral vectors (containing the relevant cDNA) and pCL-Eco retrovirus packaging vector, as described above

(Section 2.3). The 293T medium was replaced after 24 hours, and viral supernatant was collected at both 48 and 72 hour time-points after transfection.

Virus was added to B3 cells in the presence of HEPES (10mM) and Polybrene

(4µg/ml). Infections were performed in a volume of 4ml per well in 6 well plates, at a cell density of 1 x 10 6 cells/ml. The virus/ cell mix was centrifuged at 700 x g for 60 minutes (‘spinfection’), after which the viral supernatant was replaced by standard B3 cell culture medium. Infection efficiency was determined by flow cytometric analysis of green fluorescent protein (GFP) expression 48 hours after infection.

As retroviral plasmids differed in infection efficiency, the relative volumes of viral supernatant that were used for infections were determined by titration experiments. Relative infection efficiency was quantified using the following formula:

Infective units = (% GFP positive cells * starting number of cells)/ volume of viral supernatant. Equal amounts of infective units were used in co-infections, and the corresponding volume was used for retroviral infection with each plasmid alone.

2.9 Cell cycle analysis (Propidium Iodide Staining)

Analysis of the cell cycle profile of retrovirally infected B3 cells was performed 48 hours after infection. Approximately 10 6 cells were washed once in PBS supplemented with 2% FCS (volume/ volume). The cells were fixed by incubation in cold 70% ethanol (volume/ volume) for 10 minutes on ice. After two washes in PBS/

74

2% FCS, the fixed cells were resuspended in PBS supplemented with propidium iodide (50µg/ml), NP40 (0.05%) and RNase (Roche, 180µg/ml). Flow cytometric analysis was carried out after 30 minutes of incubation.

2.10 Analysis of cellular proliferation (BrdU incorporation)

The rates of proliferation of retrovirally infected B3 cells were assessed by measurement of incorporation of bromodeoxyuridine (BrdU) 48 hours after infection.

BrdU (10µM) was added directly to a culture of 10 6 cells. After one hour’s incubation

0 (at 37 C and 5% CO 2), the cells were washed twice in PBS with 1% BSA then fixed by incubation with 70% ethanol on ice for 30 minutes. Cellular DNA was denatured to single-stranded molecules by addition of 2M HCl with 0.5% Triton X-100. The solution was centrifuged at 500 x g for 10 minutes and the pellet was resuspended in

0.1M Na 2B4O7 (pH 8.5), in order to neutralise the acid. Following further centrifugation (500 x g for 10 minutes), the resultant cell pellet was resuspended in

PBS with 0.5% Tween and 1% BSA. Flow cytometric analysis of BrdU incorporation was performed following incubation with a FITC-conjugated anti-BrdU antibody for

30 minutes at room temperature.

2.11 Quantitative PCR (QPCR)

RNA was extracted from B3 cells 48 hours after retroviral infection using the

Qiagen RNeasy mini kit, according to the manufacturer’s protocol. The Mouse Cell

Cycle RT 2 Profiler TM PCR array was performed according to the manufacturer’s (SA

Biosciences) instructions. All other QPCR assays were carried out following reverse

75 transcription of 1µg of RNA to cDNA using Superscript II Reverse Transcriptase

(Invitrogen), in a final volume of 20µl.

QPCR reactions were performed on the DNA Engine Opticon 2 (MJ Research) using the SYBR Green JumpStart Taq ReadyMix (Sigma-Aldrich). Reactions were carried out in a final volume of 25 µl per well of a 96 well plate. 1 µl of the cDNA reaction was added per well, in the presence of final concentrations of 400µM for each primer. For chromatin immunoprecipitation experiments (Section 2.11), reactions were performed in duplicate. For quantification of Gpr132 expression, reactions were performed in triplicate. In addition, a ‘no reverse transcriptase’ (no-RT) control was run during each Gpr132 quantification assay, in order to ensure that any detected PCR amplification was not caused by contamination with genomic DNA.

The following thermal cycles were used in all QPCR experiments: 10 minutes at 95 0C, followed by 40 cycles of 15 seconds at 95 0C, 15 seconds at 60 0C and 15 seconds at 72 0C. For each experiment, the threshold was set to cross a point at which

PCR amplification was linear.

Gpr132 expression was assessed by PCR amplification with two different primer sets, which gave concordant results in separate experiments. The primer sequences are shown in Table 2.1. Relative expression was calculated by the change- in-threshold (− CT) method (Pfaffl, 2001), following normalisation to a housekeeping gene ( CASC3 ) control. Results were analysed using Student’s paired t- test in order to determine the statistical significance of any observed differences in expression between samples.

76

Table 2.1 Primer Sequences used in Gpr132 QPCR assays .

Name Sequence

Gpr132 Primer Set A Forward GCTACATGGCCGTGGTCTAT

Gpr132 Primer Set A Reverse GGTGAAACGCAGGTAGTGGT

Gpr132 Primer Set B Forward CGTCACCATCTTCCTGGTCT

Gpr132 Primer Set B Reverse ACGTAGATGATGGGGTCAGC

CASC3 Forward AGCTAACGATGCTGCTGATTC

CASC3 Reverse TCCAAGTGCTTAGGGCCTTTT

2.12 Chromatin immunoprecipitation (ChIP)

2.12.1 Chromatin Isolation

Chromatin was isolated from B3 cells 48 hours after retroviral infection.

Approximately 10 7 cells were fixed and cross-linked with formaldehyde (1%) at 37 0C for 10 minutes. The cross-linking reaction was terminated by addition of glycine

(0.125M) and incubation at room temperature for 10 minutes, after which the cells were washed three times in PBS at 4 0C (at 700 x g for 5 minutes each time).

Cells were resuspended in a swelling buffer (25mM HEPES pH 7.9, 1.5mM

MgCl 2, 10mM KCl and 0.1% NP40 supplemented with Complete Mini Protease

Inhibitor (Roche)). This solution was kept on ice for 10 minutes in order to lyse the cells, and was then homogenised using a Dounce homogeniser (50 strokes) and subjected to centrifugation (3000 x g at 4 0C for 8 minutes). The resultant nuclear

77 pellet was resuspended in sonication buffer (50mM HEPES, 140mM NaCl, 1mM

EDTA, 1% Triton X-100, 0.1% sodium deoxycholate and 0.1% sodium dodecylsulphate supplemented with Complete Mini Protease Inhibitor (Roche)) and sonicated at high intensity using a Bioruptur sonicator (Diagenode) for 30 minutes in total (5 seconds on/ 5 seconds off).

The sizes of the chromatin fragments were verified by agarose gel electrophoresis and if satisfactory (a range of 500-1000 base pairs), the chromatin was recovered from the supernatant following centrifugation (10,000 x g at 4 0C for 10 minutes). The chromatin concentration was determined following alkaline lysis (1:10 dilution in 0.1M NaOH) using a spectrophotometer, by measurement of absorbance at

260 nm. Chromatin samples were diluted to equal concentrations and aliquots of equivalent amount were snap frozen in liquid nitrogen and stored at -80 0C until required.

2.12.2 Immunoprecipitation

5µg of anti-FLAG M2 antibody (Sigma-Aldrich) was coupled to 40 µl Protein

G Dynabeads (Sigma-Aldrich) for two hours at 4 0C. Following removal of unbound antibody, 100µg of chromatin was added to the beads and incubated overnight at 4 0C with rotation.

The beads were washed four times at 4 0C in the following sequence: once with sonication buffer; once with wash buffer ‘A’ (50mM HEPES, 500mM NaCl, 1mM

EDTA, 1% Triton X-100, 0.1% sodium deoxycholate and 0.1% sodium dodecylsulphate); once with wash buffer ‘B’ (20mM Tris pH 8.0, 1mM EDTA,

250mM LiCl, 0.5% NP40, 0.5% sodium deoxycholate); and finally with Tris-EDTA

78

(TE, pH 8.0). The beads were resuspended in fresh TE (pH 8.0) and transferred to a new Eppendorf tube, in order avoid subsequent detection of material that bound non- specifically to the walls of the original tube.

Following removal of TE, the beads were suspended in an elution buffer

(50mM Tris pH 8.0, 50mM NaCl, 1mM EDTA, 1% sodium dodecylsulphate supplemented with DNase-free RNase at 20µg/ml) and incubated overnight at 68 0C, in order to reverse the cross-linking of the chromatin. In addition, non-IP’d chromatin equivalent to 10% of the original IP (which would subsequently be used an ‘input’ comparison for QPCR amplification) was reverse cross-linked simultaneously.

Recovered chromatin was treated with Proteinase K (200 µg/ml) for two hours, after which DNA was purified using a PCR purification kit (Qiagen). The concentration of the recovered DNA was calculated using PicoGreen reagent (Invitrogen).

2.12.3 QPCR

The QPCR procedure is described in Section 2.11. Each amplification was performed on equivalent volumes of recovered DNA, and a 10% ‘input’ comparison that facilitated relative quantification of the observed IP. The sequences of the primers that were used to amplify Gpr132 Regions A and B, and ‘background’ regions 1 and 2

(discussed in Section 6.1) are listed in Table 2.2.

79

Table 2.2 Primers used in QPCR analysis of ChIP experiments.

Name Sequence

Gpr132 Region A; Primer Set 1, Forward CGCAAAATCAGCATGAAGAA

Gpr132 Region A; Primer Set 1, Reverse TGGGGCTCTGAAATATTGGT

Gpr132 Region A; Primer Set 2, Forward TACAGACAACCAGGCACCAG

Gpr132 Region A; Primer Set 2, Reverse TCATGCTGATTTTGCGTTTC

Gpr132 Region B; Primer Set 1, Forward GAGGGCTGCTCAGGATGAT

Gpr132 Region B; Primer Set 1, Reverse AGTGGTGGGTACCAGCAGTG

Gpr132 Region B; Primer Set 2, Forward CTCACCCAACCCCTAACAGA

Gpr132 Region B; Primer Set 2, Reverse CACGAACAGCTGCTTTCAGA

Background Region 1, Forward CAGAGGACAGCTAGGCAGAGA

Background Region 1, Reverse GAGCAGACTCATTAGGGCAAA

Background Region 2, Forward TGATGACAGATCAAGCCAGGT

Background Region 2, Reverse TTGTAACATGGGGTCGTG

80

Chapter 3

Characterisation of a physical interaction

between Ikaros and Foxp1

Interplay between transcription factors is known to be important for the regulation of gene expression during B-lymphopoiesis (Busslinger, 2004). As Ikaros and Foxp1 are expressed contemporaneously through several stages of B cell differentiation, this raised the possibility that these proteins might interact functionally. If this is the case, disruption of such an interaction (for example, as a consequence of the acquired genetic changes to Ikaros that are typically observed in

Ph+ B-ALL) could result in changes to B-cell transcriptional programmes.

The above considerations suggested the possibility of a physical association between the two proteins. Prior to my commencing this project, this hypothesis was initially tested in pre-B-cells, as expression of each factor is relatively high at this stage of differentiation. Co-immunoprecipitation (co-IP) experiments were performed on protein lysates obtained from wild-type (WT) murine fetal liver pre-B-cells by Dr.

Renae Domaschenz. Lysates were subjected to IP with anti-Ikaros and anti-Foxp1 antibodies, and immunoblots were performed using the eluates. Figure 3.1 shows an anti-Foxp1 immunoblot (generated by Dr. Domaschenz) that demonstrated the presence of endogeneous Foxp1 and successful IP of Foxp1 using an anti-Foxp1 antibody. Critically, there was also significant pulldown of Foxp1 following IP with the anti-Ikaros antibody, demonstrating co-IP of the proteins in these cells. The results of this co-IP, together with results from repeat confirmatory experiments formed the basis for the work described in this thesis.

81

IP Antibody

Input Ikaros IgG Foxp1

Foxp1A (79kDa)

IgH

Anti-Foxp1 Immunoblot

Figure 3.1 Ikaros and Foxp1 interact in murine fetal liver pre-B-cells. Protein extracts from murine fetal liver pre-B-cells were subjected to IP with anti-Foxp1 and anti-Ikaros antibodies. An anti-Foxp1 immunoblot is shown. An arrow indicates the presence of full-length Foxp1 (Foxp1A) in the input lane. Foxp1A and some shorter splice isoforms were precipitated by both anti-Ikaros and anti-Foxp1 antibodies, but not by normal IgG control antibody. IgH indicates the presence of immunoglobulin heavy chain from the IP antibody.

82

3.1 Murine Ikaros and Foxp1 interact in vivo

The physical interaction between Ikaros and Foxp1 was investigated using an expression system based on Human Embryonic Kidney (HEK) 293T cells (henceforth referred to as ‘293T cells’). 293T cells were transfected with expression vectors that encode murine Ikaros and Foxp1 proteins, containing an N-terminal epitope tag (HA-

Ikaros and FLAG-Foxp1). The addition of an epitope tag to the construct facilitated subsequent IP, using anti-HA or anti-FLAG antibodies.

Cells were co-transfected with expression vectors that encode HA-Ikaros and

FLAG-Foxp1, and IP was performed on protein lysates that were extracted 48 hours after transfection. Figure 3.2 shows the results of an IP carried out with an anti-FLAG antibody, which recognises the FLAG-tagged Foxp1. Probing of the blot with anti-HA showed a significant pulldown of HA-Ikaros, demonstrating the in vivo interaction between the two proteins in this system (Figure 3.2, left panel). The reciprocal experiment was carried out using IP with a polyclonal anti-Ikaros antibody, which I found to be more efficient in IP than the anti-HA antibody. Detection of Foxp1 by

Western blotting of the precipitated proteins with anti-FLAG antibody provided further confirmation of the association between the two proteins (Figure 3.2, right panel).

83

IP Antibody IP Antibody

Input IgG Ikaros Input IgG FLAG

FLAG- Foxp1A Ikaros-HA (79kDa) (60kDa)

IgH

IgH

Anti-FLAG Immunoblot Anti-HA Immunoblot

Figure 3.2 Murine Ikaros and Foxp1 interact in vivo , when ectopically expressed in 293T cells. Protein extracts from 293T cells that were co-transfected with HA- tagged Ikaros and FLAG-tagged Foxp1 were subjected to IP and analysed by Western blotting with anti-HA antibody (right panel) and anti-FLAG antibody (left panel). The antibody used for the IP is indicated above each blot.

84

3.2 Murine Ikaros and Foxp1 proteins interact in vitro

Reciprocal IP of Ikaros and Foxp1 provides strong evidence of an interaction between these proteins in vivo . However, this approach does not differentiate between a direct protein-protein interaction and an indirect association as part of a larger multiprotein complex. As transcription factor activity entails physical contact with a wide range of partner proteins, IP in vivo could result in pulldown of associated complexes. Although a finding that the association is indirect would not necessarily preclude the possibility of a functionally significant interaction, the confirmation of a direct physical association would increase the likelihood that Ikaros and Foxp1 proteins act together in the regulation of gene expression in B-cells.

Evidence of direct interaction can be obtained by co-precipitation of purified proteins in vitro in a cell-free system. These systems typically involve the addition of high-affinity protein tags to the molecules of interest, in order to facilitate subsequent purification. An example of this system entails the fusion of a protein of interest to a high-affinity Glutathione-S-Transferase (GST) tag and expression in bacteria under the control of an Isopropyl β-D-1-thiogalactopyranoside (IPTG)-inducible promoter.

The protein can then be isolated and precipitated by addition to paramagnetic beads that are coated with Glutathione (GSH), the substrate for GST (see Methods Section

2.7). Competent bacteria were transformed with expression vectors that encode GST fusion proteins of either murine Ikaros or Foxp1. Bacteria were also transformed with an ‘empty’ vector that encodes GST only. The latter control is necessary, as inclusion of a GST protein in the subsequent experiments ensures that any detected precipitation does not simply reflect non-specific association with the GST molecule.

85

Following optimisation of the purification procedure, the quality of each fusion protein preparation was assessed by precipitation using GSH-coated paramagnetic beads and analysis by SDS-polyacrylamide gel electrophoresis followed by

Coomassie staining. Precipitation of the bacterially expressed GST-Ikaros fusion protein consistently gave a pulldown that was mainly composed of smaller products, reflecting either protein instability or premature termination of transcription or translation. Attempts at performing pulldowns using this preparation (including attempted precipitation of Ikaros protein as a positive control) were unsuccessful, indicating that the purified GST-Ikaros protein was not suitable for use in these experiments. For this reason, we focused our efforts on assessment of whether there is a direct interaction between GST-Foxp1 and an Ikaros protein which was radio- labelled by coupled in vitro transcription/ translation in the presence of 35 S-methionine

(see Methods Section 2.7).

The 35 S-labelled Ikaros was incubated with either purified GST or GST-Foxp1 fusion protein and was then added to GSH-coated paramagnetic beads. Complexes were eluted and resolved by SDS-polyacrylamide gel electrophoresis. The results showed a clear band for the 35 S-Ikaros in the GST-Foxp1 pulldown, which is not present in the GST lane (Figure 3.4). A Coomassie stain of a mock pulldown performed at the same time showed equivalent elution of GST and GST-Foxp1, indicating that similar amounts of protein were bound to the GSH-coated paramagnetic beads (Figure 3.3, below). These in vitro data confirm the existence of a direct interaction between murine Ikaros and Foxp1 proteins, without the requirement for formation of larger multiprotein complexes that contain both proteins.

86

kDa 250 150 100 GST-Foxp1 75 (104kDa)

50

37.5

25 GST (25kDa)

Purified Purified GST GST-Foxp1

Figure 3.3 Coomassie stain of a mock GST pulldown. A mock pulldown was performed at the same time as the GST pulldown depicted in Figure 3.4 (overleaf) and resolved by polyacrylamide gel electrophoresis. Coomassie staining of the gel showed that similar amounts of GST and GST-Foxp1 were bound to the GSH-coated beads.

87

Pulldown

Input GST GST-Foxp1

Ikaros

Figure 3.4 Murine Ikaros and Foxp1 interact in vitro . 35 S-labelled Ikaros was transcribed and translated in vitro , incubated with either purified GST or GST-Foxp1 fusion protein and then added to GSH-coated paramagnetic beads. Immunocomplexes were eluted and resolved by polyacrylamide gel electrophoresis and the presence of 35 S-labelled protein was detected by autoradiography.

88

3.3 Ikaros and Foxp1 interaction is mediated by the DNA-binding domains of both proteins

In order to discover which regions of each factor are necessary for protein- protein interaction, we generated deletion constructs, which lacked specific domains of the protein that have been previously described to have functional significance

(Cobb et al., 2000; Molnar and Georgopoulos, 1994; Wang et al., 2003). Schematic representations of the structure of each protein and of the deletion mutants are shown in Figures 3.5 (Ikaros) and 3.7 (Foxp1).

As discussed in the previous section, the association between GST-Foxp1 and radio-labelled Ikaros was readily demonstrable in vitro . This method detects direct interaction of the proteins, without the requirement for participation in multiprotein complexes that may be formed in vivo . Therefore this system was used to test the binding of deletion mutants of Ikaros to WT Foxp1. Mutants were transcribed and translated in vitro and incubated with both GST and GST-Foxp1. Most deletion mutants interacted with GST-Foxp1 to a similar extent to that of the WT protein

(Figure 3.6). However, deletion of exons 3-6 (which encompasses the coding region for the central zinc finger domain and therefore corresponds to the non-DNA-binding

IK6 isoform) abrogated the interaction with Foxp1 almost completely (Figure 3.6).

This indicates that a significant part of the interaction is mediated by this region of the

Ikaros protein.

89

C-terminal Central Zinc Finger Domain Zinc Finger Domain

54C 235-362 363-451 N458 IK6

Figure 3.5 Schematic representation of Ikaros protein and deletion mutants. Zinc fingers are represented by filled boxes. Lines indicate the deletions that were tested in in vitro interaction experiments (Figure 3.6). The numbering indicates the amino acid position of each deletion, e.g. 54C encompasses the first 54 amino acids. The IK6 deletion extends from residues 50 to 233. The N458 deletion comprises the final 62 amino acids of the protein.

90

Pulldown GST- Input GST Foxp1

60kDa WT Ikaros

50kDa 54C

37kDa IK6

45kDa 235-362

48kDa 363-451

50 kDa N458

Figure 3.6 Mapping of the interaction domain of Ikaros. WT Ikaros and deletion mutants (depicted schematically along with molecular weights) were transcribed and translated in vitro and pulldown was performed with either GST or GST-Foxp1. Pulldown with GST alone is associated with non-specific precipitation of shorter translated proteins in most experiments. These shorter products are not seen following in vitro translation of the smaller IK6 protein. All experiments were performed at least twice.

91

Due to the difficulties that were encountered when purifying intact GST-Ikaros

(see Section 3.2), the 293T cell system was used to IP FLAG-tagged Foxp1 deletion mutants, using an anti-Ikaros antibody. Interaction of most deletion mutants of Foxp1 was similar to that of the WT protein, however deletion of 543 base pairs that encode the forkhead region resulted in an abolition of the association (Figure 3.8). Similar to the interaction region of Ikaros, this part of Foxp1 has been shown to mediate DNA- binding (Wang et al., 2003).

Poly glutamine Glutamine-Rich Zinc Leucine Forkhead Acidic-Rich Region Region Finger Zipper Domain Region Domain Region 0 261 320 372 477 658 705

Figure 3.7 Schematic representation of Foxp1 protein and deletion mutants. Patterned boxes indicate the labelled structural domains. Deletion mutants lacking specific functional regions were tested in interaction assays shown in Figure 3.8. Foxp1C is an isoform that lacks the polyglutamine and glutamine-rich regions (Amino acids 1-261). All other deletion mutants lack the amino acids that code for the relevant functional region, as shown.

92

IP

Input IgG Ikaros

← IgH Foxp1A 79kDa

← IgH Foxp1C 54kDa

← IgH Zinc Finger Deletion 69kDa

← IgH Leucine Zipper Deletion 59kDa

← IgH Forkhead Deletion 60kDa

Acidic-Rich Deletion 70kDa ← IgH

Figure 3.8 Mapping of Foxp1 interaction domain. 293T cells were co-transfected with WT Ikaros and either WT Foxp1 or Foxp1 deletion mutants that lack specific structural domains (depicted schematically). IP using an anti-Ikaros antibody resulted in pulldown of Foxp1 and most deletion mutants (indicated by block arrows next to the input lane). A Foxp1 mutant which lacks the forkhead region (which mediates DNA-binding) was not precipitated. All experiments were performed at least twice.

93

As each of the regions that were found to mediate protein-protein association has also been shown to be necessary for binding to DNA, we considered the possibility that demonstration of the association was dependent on the presence of

DNA in the IP reaction (and therefore, isoforms that do not bind DNA would not be precipitated). In order to ensure that the preceding results did not simply reflect the isolation of DNA-bound complexes, IP was also performed in the presence of either ethidium bromide (which disrupts DNA-protein interactions (Lai and Herr, 1992)) or

DNase I. Representative results of this analysis are shown in Figure 3.9 and show that the IP was unaffected by ethidium bromide or DNase I.

IP Antibody IP Antibody

Input IgG FLAG Input IgG Ikaros

FLAG- Ikaros-HA Foxp1A (60kDa) → (79kDa) → ← IgH

← IgH

Anti-HA Immunoblot Anti-FLAG Immunoblot

Figure 3.9 Co-IP of murine Ikaros and Foxp1 proteins is not DNA-dependent. Protein lysates were treated with either ethidium bromide or DNaseI, in order to disrupt DNA-protein interactions. At least two experiments were performed using either condition. IP was performed in the manner previously described. A representative IP experiment following DNase treatment of protein lysates is shown.

94

3.4 The human IK6 deletion abolishes interaction with human

FOXP1

The DNA sequences of both Ikaros and Foxp1 are highly conserved between mice and humans (Brown et al., 2008; Cortes et al., 1999), which raised the strong possibility of an equivalent interaction between the human forms of the proteins. In order to test whether the human proteins interact, IP was performed on protein extracts from 293T cells that had been transfected with FLAG-tagged human FOXP1 and either WT human Ikaros, or human Ikaros isoform IK6. The results of this IP are shown in Figure 3.10. WT Ikaros was precipitated by an anti-FLAG antibody, but the same antibody did not pull down IK6. The reciprocal anti-Ikaros IP showed corresponding results (i.e. lack of pulldown of FLAG-Foxp1 by IK6), indicating that the central zinc finger domain of human Ikaros is required for interaction with human

FOXP1. This pattern of association was also observed in an in vitro pulldown assay.

WT Ikaros was pulled down efficiently by GST-FOXP1, whereas IK6 gave almost no detectable pulldown (Figure 3.11).

These findings were further confirmed by performing anti-FOXP1 IP on protein extracts from human B-ALL cell lines that exhibit differing patterns of Ikaros isoform expression. The predominant isoform that is expressed in the REH ALL line is WT Ikaros. IP of protein extracts from this cell line showed precipitation of WT

Ikaros by an anti-FOXP1 antibody (Figure 3.12 (left panel). The SUP-B15 cell line expresses the IK6 isoform, and a reduced amount of WT protein. IP with the anti-

FOXP1 antibody failed to pull down the IK6 isoform in this case (Figure 3.12). In addition, several experimental replicates in this line showed that anti-FOXP1 IP of

95

WT Ikaros was either reduced or absent. This may reflect the reduced levels of WT

protein in the SUP-B15 line (in comparison with REH cells), or may also be caused

by a sequestration effect exerted by IK6 on WT Ikaros (Molnar et al., 1996). Overall,

these results provide further confirmation that the DNA-binding domain of Ikaros is

required for interaction with FOXP1 in vivo and in vitro .

IP Antibody IP Antibody IP Antibody IP Antibody Input IgG FOXP1 Input IgG FOXP1 Input IgG Ikaros Input IgG Ikaros

WT Ikaros FOXP1A (60kDa) → (79kDa) →

← IgH

IK6 (35kDa) →

WT Ikaros IK6 WT Ikaros IK6 Anti-Ikaros Immunoblot Anti-FLAG (FOXP1) Immunoblot

Figure 3.10 Human IK6 does not interact with human FOXP1 in a 293T cell expression system. 293T cells were co-transfected with FLAG-tagged human FOXP1, and either WT human Ikaros, or human IK6. Whole cell extracts were IP’d with anti-Ikaros or anti-FLAG and probed with anti-Ikaros (left-hand panel) or anti- FOXP1 (right-hand panel). Confirmation of the result was seen in a further repeat experiment.

96

The corollary of the observation that the DNA-binding domains of each protein are required for the interaction is that these regions should be sufficient in themselves for association of the factors to occur. We attempted to assess this by using a construct that contains the cDNA which codes for the four DNA-binding zinc fingers of the

Ikaros protein. This was transcribed and translated in vitro , and pulldown with GST-

Foxp1 was assessed. Unfortunately, this truncated protein was not expressed to any great extent using this system and it was therefore not possible to assess its binding.

We therefore cannot formally rule out the possibility that abnormal folding of either or both deletion mutant may have caused false negative results in the interaction assays.

97

WT Pulldown Pulldown Ikaros GST IK6 GST Input GST FOXP1 Input GST FOXP1

WT Ikaros IK6

WT Ikaros IK6

Figure 3.11 Human IK6 does not interact with human FOXP1 in vitro . WT Ikaros and IK6 were transcribed and translated in vitro in the presence of 35 S-methionine and analysed for interaction with GST or GST-FOXP1 as previously described (Figure 3.3). Confirmation of the result was obtained in a further repeat experiment.

98

IP Antibody IP Antibody Input IgG FOXP1 Input IgG FOXP1 WT Ikaros WT Ikaros (60kDa) → (60kDa) →

← IgH ← IgH

IK6 (35kDa) →

REH Cell Line SUP-B15 Cell Line Anti-Ikaros Immunoblot

Figure 3.12 Anti-FOXP1 IP of Ikaros isoforms in human leukaemia cell lines. WT Ikaros was precipitated by an anti-FOXP1 IP of protein extracts from the REH cell line. The same antibody failed to IP the IK6 isoform in lysates obtained from the SUP-B15 cell line. In addition, no IP of WT Ikaros was seen in the SUP-B15 line. IgG IP controls are negative in all cases. Two further repeat experiments showed very weak interaction of WT Ikaros with FOXP1.

99

3.5 Summary The hypothesis that Ikaros and Foxp1 might interact during B-cell development arose from established concepts of transcription factor interplay during haematopoiesis, and the observation that the two factors are expressed at similar stages of differentiation. Co-IP of the proteins in murine fetal liver pre-B-cells provided the first evidence of a physical association that might involve a functional interaction between the two proteins.

In order to investigate this possibility further, the reproducibility of the interaction was confirmed in several different systems. Ikaros and Foxp1 were shown to associate in vivo in a DNA-independent manner when ectopically expressed in

293T cells. The finding that interaction between the purified proteins could also be detected in vitro reinforces this DNA-independence, and suggests that the physical association is a direct one. Experimental assessment of human proteins using the same systems yielded identical results, suggesting that the interaction is evolutionarily conserved.

Elucidation of the interaction region of the relevant proteins showed that contact was mediated through the respective DNA-binding domains. This finding raises important questions about the potential downstream effects of alterations in the relevant amounts of each protein (or their isoforms) on developmentally-regulated gene transcription. In the following chapters, I will discuss how the functional significance of Ikaros-Foxp1 interplay was investigated.

100

Chapter 4

Role of the Ikaros/Foxp1 interaction in

regulating the pre-B cell cycle

We considered that the demonstrated physical interaction between Ikaros and

Foxp1 could affect transcription factor activity by several diverse mechanisms. In one scenario, Ikaros/ Foxp1 association might increase the affinity of one or both factors for target gene sequences (for example through induction of protein conformational changes), thereby stimulating transcription. The converse hypothesis would predict that association of the proteins negatively regulates target gene expression through inhibition of DNA binding. It could also be possible that the consequences of Ikaros/

Foxp1 contact might depend on the presence or absence of additional associated factors, so that the effects of the interaction may vary according to gene locus and/ or cell type.

The fact that contact between Ikaros and Foxp1 is mediated by known DNA- binding regions led us to speculate that one likely outcome of protein-protein interaction would be abrogation of DNA binding of one or both factors. This hypothesis predicts that the amount of each factor that is available for gene regulation is affected by the relative quantity of the interacting molecule, which would otherwise occupy the DNA-binding domain. In this scenario (depicted in Figure 4.1, and which is simplified by assuming mutually exclusive binding of each factor), a reduction in the quantity of one protein would ‘liberate’ the other partner to bind gene regulatory elements and alter transcription.

101

Baseline Imbalance

Target Target Gene Gene Expression Expression

Figure 4.1 Model of the hypothetical effects of changes in the amounts of Ikaros and/or Foxp1 on expression of target genes. A simplified model (predicated on the hypothesis that protein-protein interaction inhibits DNA binding, and that DNA binding of each factor is mutually exclusive) of the effects of changes in relative protein levels is shown. Interacting factors are depicted as coloured pies, and DNA as a spiral. The left side of the figure shows ‘baseline’ levels of yellow proteins, which occupy the DNA-binding domains of the blue proteins; the non-bound fraction is free to bind DNA and alter gene expression. On the right side of the figure, reduced levels of the yellow protein (‘imbalance’) result in increased availability of non-bound blue proteins, which can bind DNA and alter expression of gene targets.

102

Fluctuations in the levels of either factor (during normal development, or as a consequence of pathological derangement) could therefore cause changes in cellular gene expression and phenotype. This would have particular relevance for cases of B-

ALL that have genomic deletions of the IKZF1 gene and expression of the non-

FOXP1-interacting IK6 isoform (Section 1.6). If the model shown in Figure 4.1 is correct, this would be expected to increase the availability of FOXP1 for binding to

DNA, thereby altering expression of FOXP1 gene targets.

In order to examine the effects of altered balance between Ikaros and Foxp1, the relative level of each protein was manipulated by retroviral infection in a pre-B- cell experimental model. The B3 cell line (Fisher et al., 1995) was used, as these cells are highly amenable to retroviral infection, and correspond phenotypically to the large pre-B-cell stage of differentiation.

B3 cells were infected with retroviruses that encode either Ikaros or Foxp1, or with both constructs in a ‘co-infection’. A bicistronic retroviral vector was used, in which the same gene promoter controls expression of both of the gene of interest

(Ikaros or Foxp1) and an internal ribosome entry site (IRES)-green fluorescent protein

(GFP) sequence. This system enables infected cells to be readily detected by measurement of GFP expression. Infection rates were consistently greater than 95%

(Figure 4.2), obviating the need for cell sorting and further manipulation which might otherwise affect experimental results. Assessment of the resultant cellular phenotype was performed 48 hours after infection, and these results are detailed in the following sections.

103

Uninfected

t n u

o C l l e

C

GFP→

Empty Vector Foxp1

t

t 96.4%

97.9% n n u u o

o C C l l l l

e e C C GFP→ GFP→

Ikaros Ikaros + Foxp1

t t 96.7% 95.3% n n u

u o o C C l

l l l e e C C

GFP→ GFP→

Figure 4.2 Rates of retroviral infection of B3 cells. Infection rates were determined by green fluorescent protein (GFP) expression (measured on the x axis) 48 hours after infection. The percentages of GFP-expression (indicating infected cells) are shown in this representative experiment. The experiments described in this chapter were performed four times in total.

104

4.1 Overexpression of Foxp1 is associated with G2 cell cycle arrest As both Ikaros and Foxp1 have been described to affect the progression of the cell cycle in haematopoietic cells (Feng et al., 2009; Ma et al., 2010), we considered the possibility that interaction between the proteins influences the regulation of this process in lymphocytes. This led us to hypothesise that alterations in the relative amounts of either factor would perturb lymphoid cell cycle homeostasis.

In order to test this hypothesis, the cell cycle profiles of retrovirally-infected cells were determined by staining of DNA with propidium iodide (PI). This analysis relies on the fact that cells in G2/M contain twice as much DNA as cells in G1. Flow cytometric assessment therefore shows two peaks of cellular DNA content corresponding to these phases, while all cells that have intermediate amounts are considered to be in S phase. PI staining of retrovirally transduced cells revealed that

Foxp1 infection resulted in a large increase (at least two-fold) in the proportion of cells in the G2 phase of the cycle, compared with an empty vector control (Figure

4.3). Co-infection of Ikaros with Foxp1 consistently and completely abrogated this phenotype with the co-infected cells showing cell cycle profiles that were very similar to the empty vector control. As expected, infection with Ikaros alone resulted in an increased proportion of cells in the G1 phase of the cycle, due to its inhibitory effect on G1/S transition (Ma et al., 2010). Each of these results was highly reproducible across five replicate experiments.

105

G1 G1 t

n Empty Vector u Ikaros o C l l

G2/M e G2/M Cell Count S C S

PI→ PI→ G1 G1

t t n

n Ikaros u

u Foxp1 o o

C + Foxp1 C l l l l e e G2/M G2/M C C S S

PI→ PI→

100% 90%

80% 70% 60% G2/M 50% S 40% G0/G1 30%

20% 10% 0% PMIG Ikaros Foxp1 Ikaros + Foxp1

Figure 4.3 Cell Cycle Profile of infected B3 cells. Cell cycle status was assessed by propidium iodide (PI) staining 48 hours after retroviral infection. The flow cytometric pattern of a representative experiment is shown in the upper part of the figure. A substantial increase in the proportion of cells in the G2 phase of the cycle is observed following infection with Foxp1. A graphical representation of results averaged across four separate experiments is shown in the bottom panel.

106

In order to confirm that levels of ectopically expressed Foxp1 were similar in single infections and co-infections, immunoblotting of protein extracts from infected pre-B-cells was performed. The levels of FLAG-Foxp1 following co-infection with

Ikaros were equivalent to those seen following Foxp1 infection alone. This indicates that the observed abrogation of cell cycle phenotype does not simply reflect a reduction in FLAG-Foxp1 expression due to co-infection with another retrovirus.

Empty Ikaros Vector Ikaros Foxp1 + Foxp1

Anti-FLAG (Foxp1)

Anti-HA (Ikaros)

Anti-Tubulin (Loading Control)

Figure 4.4 Expression of FLAG-Foxp1 and HA-tagged Ikaros in retrovirally infected B3 cells. The levels of ectopic protein expression were determined by immunoblotting of protein extracts from infected pre-B-cells. The levels of FLAG- Foxp1 are similar in cells infected with Foxp1 alone, and cells co-infected with Ikaros and Foxp1.

107

4.1.1 Overexpression of Foxp1 is associated with reduced cellular proliferation

There are two alternative explanations for the observation that increased Foxp1 expression resulted in an increased proportion of cells in the G2 phase of the cell cycle. The first possibility is that higher levels of Foxp1 caused a delay or arrest of cells in G2. Alternatively, overexpression of Foxp1 might have stimulated higher rates of cycling, thereby generating a larger G2 fraction. In order to discriminate between these two possibilities, cellular proliferation was assessed by measurement of bromodeoxyuridine (BrdU) incorporation. This measure quantifies the amount of

BrdU that has been incorporated into newly synthesised DNA during S phase, and thus provides a surrogate measure of cellular proliferation.

The results of a representative experiment are shown in Figure 4.5. Foxp1 infection was associated with a reduction in BrdU incorporation that was consistent with a diminished rate of proliferation relative to an empty vector control (Figure 4.5).

This indicated that the observed cell cycle changes were not due to increased rates of cycling, but occurred because of G2 arrest. As expected from previously published data (Ma et al., 2010), infection with Ikaros also caused a reduction in cellular proliferation. Co-infection of Ikaros with Foxp1 was associated with similar reductions in proliferation to those seen following infection with either factor alone.

108

Empty Vector Ikaros

41.5% 20.4% t t n n u u o o C C l l l l e e C

C

BrdU→ BrdU→

Foxp1 Ikaros + Foxp1

34.2% 28.3% t t n n u u o o C C l l l l e e C C

BrdU→ BrdU→

Figure 4.5 Measurement of proliferation of retrovirally infected B3 cells. Cellular proliferation was assessed by measurement of bromodeoxyuridine (BrdU) incorporation. The percentages of cells that have incorporated BrdU (and are therefore proliferating) are indicated.

109

4.2 Foxp1-mediated G2 cell cycle arrest is associated with increased expression of Gpr132 We reasoned that the mechanistic basis of the observed G2 arrest in Foxp1 overexpressing cells could be transcriptional activation (or repression) of a specific gene (or genes) by Foxp1. Co-infection with Ikaros would abrogate these effects by occupation of the Foxp1 forkhead domain, thereby reducing Foxp1 DNA-binding activity to baseline levels. In order to measure the transcriptional changes that followed Foxp1 overexpression, RNA was extracted from retrovirally infected B3 cells and quantitative PCR (QPCR) analysis was performed. A commercially available

QPCR-based array (Qiagen) was initially used to determine the expression of a range of cell cycle-associated genes (84 in total), with the aim of identifying candidate genes for further analysis. The gene that showed the greatest increase in expression following Foxp1 infection was Gpr132 (2.8-fold, data not shown). This gene encodes the G2A protein, a stress-inducible G protein-coupled receptor which has been described to mediate a variety of effects in blood cells, including monocyte activation and T-cell chemotaxis (Bolick et al., 2009; Osmers et al., 2009). Notably, G2A has previously been described to induce G2 arrest when ectopically expressed in human fibroblasts (Le et al., 2002; Weng et al., 1998) and we therefore considered it to be a prime candidate to mediate the observed cell cycle effects.

The results from the cell cycle array were validated by analysing Gpr132 expression in three separate Foxp1 overexpression experiments. QPCR was carried out using two sets of primers for Gpr132 mRNA, and results were highly concordant between primer sets. Expression levels were calculated using the change-in-threshold

(− CT) method, with normalisation to an internal housekeeping gene control

110

(CASC3 ). The results of the expression analysis are shown in Figure 4.6. Foxp1 infection consistently resulted in increases of between 1.5- and 4-fold in the levels of

Gpr132 relative to infection with an empty vector control. The mean increase in expression in assays of material isolated from three biological replicates was 2.33 ±

0.6 (mean ± SEM).

In cells that were co-infected with Ikaros and Foxp1, the ratio of Gpr132 expression relative to empty vector was 1.08 +/- 0.29, indicating that co-infection with

Ikaros abrogates Foxp1-mediated induction of Gpr132 transcription. The differences in expression that were observed following Foxp1 single infection and Ikaros/ Foxp1 co-infection were analysed using a Student’s paired t-test, and were found to be statistically significant (p < 0.05).

Infection with Ikaros alone resulted in a decreased ratio of Gpr132 relative to empty vector (0.77 ± 0.19), which was likely to be caused by interaction of ectopically-expressed Ikaros with endogeneous Foxp1. According to our model, enhancement of gene expression by Foxp1 would be inhibited as a consequence of occupancy of the DNA-binding domain of the endogeneous protein by ectopically- expressed Ikaros, and these results provide further support for this type of model.

Attempts were made to confirm a corresponding increase in Gpr132 at the protein level, but immunoblotting using a commercially available antibody was unsatisfactory.

111

4 p < 0.05 *

3

2

1 Expression Relative to Empty Vector

Ikaros Gpr132 Ikaros Foxp1 + Foxp1

Figure 4.6 Expression of Gpr132 in retrovirally infected B3 cells. Expression analysis was carried out on RNA extracted from cells 48 hours after retroviral infection. Mean changes in expression from assays of three biological replicates are shown. Levels of Gpr132 are expressed relative to levels in cells infected with the control empty vector. Foxp1 infection consistently resulted in increased Gpr132 expression, while co-infection of Ikaros with Foxp1 was associated with Gpr132 levels that were equivalent to empty vector control. The difference between these values was statistically significant, as shown. Ikaros infection alone resulted in reduced Gpr132 expression. Error bars indicate standard error of the mean (SEM).

112

4.3 Summary

Infection of B3 cells with retroviral constructs that encode either Ikaros or

Foxp1 provided a model for studying the potential effects of quantitative imbalances of either factor. The most striking phenotype was seen following infection with

Foxp1, which consistently resulted in an approximately two-fold increase in the proportion of cells in the G2 phase of the cell cycle. The associated reductions in cellular proliferation indicated that the observed effects occurred as a result of G2 arrest.

Co-infection of Ikaros and Foxp1 was associated with cell cycle profiles that were similar to the profiles obtained with the empty vector control. Ikaros may therefore be considered to have ‘rescued’ the cell cycle phenotype, presumably through inhibition of Foxp1-mediated alterations in gene expression. Initial screening of a range of candidate genes using a QPCR-based array identified Gpr132 as a potential mediator of G2 arrest. Confirmatory analysis showed consistent Foxp1- mediated upregulation of Gpr132 expression, which was ‘rescued’ by co-infection with Ikaros.

Our mapping of the interaction domains that mediate protein-protein association (Chapter 3) led us to postulate that phenotypic rescue by co-infection was dependent on Ikaros occupancy of the Foxp1 forkhead domain, and a reduction in the amount of Foxp1 that is available to bind DNA. This hypothesis was examined further in the experiments described in the following chapters.

113

Chapter 5

Verification of a lack of functional interaction between

IK6 and Foxp1 in murine pre-B-cells

The experimental results described in the preceding chapter led us to reason that the preservation of steady-state ratios of Ikaros and Foxp1 proteins is critical for maintaining cell cycle homeostasis in pre-B-cells, and that regulation of the levels of expression of Gpr132 is central to this requirement. Overexpression of Foxp1 (thereby altering cellular Ikaros/ Foxp1 ratios) led to G2 arrest, which was associated with increases in Gpr132 expression. Restoration of equivalent Ikaros/Foxp1 ratios (by co- expression of similar amounts of each protein) abolished both the phenotype of G2 arrest and induction of Gpr132 expression. We hypothesised that co-infection of an

Ikaros isoform that does not interact with Foxp1 would fail to alter Foxp1-induced transcriptional changes and would not restore a steady-state cell cycle profile.

In order to test this theory, several additional pre-B-cell retroviral infection experiments were performed. B3 cells were infected with either WT Ikaros, Foxp1, or both constructs in a co-infection. Cells were also co-infected with Foxp1 and the

Ikaros isoform IK6, which was shown not to interact with Foxp1 either in vivo or in vitro (Chapter 3). If the model shown in Figure 4.1 is correct, co-infection of Foxp1 with IK6 would be expected to have no effect on the cell cycle phenotype or on

Gpr132 levels.

114

5.1 Co-expression of IK6 does not abrogate Foxp1-induced G2 arrest in pre-B-cells

Infection rates of all constructs (as determined by GFP expression) exceeded

95%, which again obviated the necessity for cell sorting and further manipulation that might otherwise affect experimental results (Figure 5.1). Phenotypic assessment was made 48 hours after infection, as with previous experiments. Consistent with the results described in Chapter 4, infection with Foxp1 was associated with G2 arrest, and this phenotype was ‘rescued’ by co-infection with WT Ikaros. Co-infection of IK6 with Foxp1 resulted in the same G2 arrest phenotype that was seen observed with

Foxp1 infection alone. Infection with IK6 alone had no effect on cell cycle profiles

(Figure 5.2 (a)). These results were reproducible in three separate experiments.

Averages of the observed cell cycle profiles are shown in Figure 5.2 (b).

The observation that IK6 failed to ‘rescue’ the cell cycle phenotype shows that the central zinc finger domain of Ikaros is required for abrogation of Foxp1-related cell cycle effects by WT Ikaros. The fact that this region is also required for the interaction between Ikaros and Foxp1 provides strong evidence that the interaction is directly involved in preventing the G2 arrest that results from Foxp1 overexpression.

Importantly, these results suggest that alterations in the expression pattern of Ikaros isoforms such as IK6 have critical knock-on effects on Foxp1 transcriptional activity in vivo .

115

Uninfected t

n u o C l l e C

GFP→

Empty Vector WT Ikaros + Foxp1

t 97.1% 98.2% n u

o C l l e

C Cell Count

GFP→ GFP→

WT Ikaros IK6

t 97.8% t 95.3% n n u u o o C C l l l l e e C C GFP→ GFP→

Foxp1 IK6 + Foxp1 t 98.4% t 97.9% n n u u o o C C l l

l l e e C C

GFP→ GFP→

Figure 5.1 Rates of retroviral infection of B3 cells. B3 cells were infected with the retroviral constructs, as listed. Infection rates (as determined by GFP expression) are indicated in each panel. All experiments described in this chapter were performed three times.

116

G1 G1

WT Ikaros Empty + Foxp1 Vector G2/M G2/M Cell Count Cell Count S S

PI→ PI→

G1 G1

WT Ikaros IK6

G2/M G2/M Cell Count Cell Count S S

PI→ PI→

G1 G1

IK6 Foxp1 + Foxp1

G2/M G2/M Cell Count Cell Count S S

PI→ PI→

Figure 5.2 (a) Cell Cycle Profile of infected B3 cells. Cell cycle status was assessed by propidium iodide (PI) staining 48 hours after infection. The flow cytometric pattern of a representative experiment is shown. A graphical representation of these results is shown in Figure 5.2 (b).

117

100% 90% 80% 70% 60%

50% G2/M 40% S 30% G0/G1 20% 10% 0% PMIG Ikaros Foxp1 WT IK6 IK6 + Ikaros + Foxp1 Foxp1

Figure 5.2 (b) Cell Cycle Profile of infected B3 cells. A graphical representation of the cell cycle profile of infected B3 cells is shown. These figures represent the average of three separate experiments. A marked increase in the proportion of cells in the G2 phase of the cycle is observed following infection with Foxp1. This effect is abrogated by co-infection with WT Ikaros, but not by co-infection with Ikaros isoform IK6.

118

As with the previous co-expression experiments, it was important to confirm that the levels of ectopic FLAG-Foxp1 expression were similar in single and co- infections. Protein expression was quantified following immunoblotting of extracts from retrovirally infected cells (Figure 5.3). The levels of each protein were determined using ImageQuant software (Fuji), and normalised to a loading control ( β- tubulin). The calculated molar ratios of each protein are shown in Figure 5.3. The levels of FLAG-Foxp1 were almost identical in the single infections and the co- infections with either WT Ikaros (0.96) or IK6 (1.02), indicating that the absence of phenotypic ‘rescue’ is not caused by a differential reduction in FLAG-Foxp1 expression when co-expressed with WT Ikaros or IK6.

In addition, it was critical to verify that the amount of IK6 expression following

IK6/Foxp1 co-infection was at least equivalent to the levels of WT Ikaros that

‘rescued’ the cell cycle phenotype. In fact, the levels of IK6 that were present in IK6/

Foxp1 co-infection (1.51) were in excess of those found in the WT Ikaros/Foxp1 co- infection (1.08). This indicates that the absence of phenotypic ‘rescue’ that is seen with IK6/Foxp1 co-infection is not simply a function of limited IK6 expression, relative to that of WT Ikaros.

119

Empty WT Ikaros IK6 Vector WT Ikaros Foxp1 + Foxp1 IK6 + Foxp1

Anti-FLAG (Foxp1)

WT Anti-HA Ikaros (Ikaros)

IK6

Anti-Tubulin (Loading Control)

Infection Relative HA-Ikaros Expression Relative FLAG-Foxp1 Expression

WT Ikaros 1.00

Foxp1 1.00

WT Ikaros + Foxp1 1.08 0.96

IK6 2.11

IK6 + Foxp1 1.51 1.02

Figure 5.3 Expression of FLAG-Foxp1 and HA-tagged Ikaros in infected B3 cells. The levels of ectopically expressed proteins were determined by immunoblotting of protein lysates extracted from infected B3 cells (upper panel) using ImageQuant software. Molar ratios of each protein (which were calculated following normalisation to an anti-β-tubulin loading control) are shown in the lower panel. Ratios are normalised to the levels of ectopically-expressed protein observed in the relevant single infection (i.e. infection with either WT Ikaros or Foxp1), which are given an arbitrary value of 1.

120

5.2 Co-infection with IK6 does not alter Foxp1-induced Gpr132 expression in pre-B-cells As the cell cycle phenotype mirrored Gpr132 expression levels in the experiments described in Chapter 4, we predicted that the IK6 isoform would be unable to block Foxp1-mediated induction of Gpr132 expression. Levels of Gpr132 were analysed in multiple QPCR assays using RNA from three separate experiments.

Once again, expression of Gpr132 correlated very precisely with cell cycle phenotype

(Figure 5.4), and consistent upregulation (between three- and seven-fold) of transcription was observed after Foxp1 infection. The mean increase in expression in all QPCR assays was 5.3 ± 0.45 (mean ± SEM). Co-infection of WT Ikaros with

Foxp1 was associated with Gpr132 mRNA levels that were very similar (0.85 ± 0.07) to those seen with an empty vector control, consistent with the phenotypic rescue observed in previous experiments. In contrast, co-infection of IK6 with Foxp1 did not affect Gpr132 levels, which were very similar to those seen following single Foxp1 infection (6.96 ± 0.58). Retroviral infection with IK6 alone had no significant effect on Gpr132 expression (0.99 ± 0.04). The differences in Gpr132 expression levels between WT Ikaros/Foxp1 co-infection and both Foxp1 single infection and

IK6/Foxp1 co-infection were analysed using Student’s paired t-test, and were found to be highly statistically significant in each case (p < 0.001).

Several additional observations can be made from the results shown in Figure

5.4. Firstly, a decrease in Gpr132 expression was again seen following infection with

WT Ikaros alone (0.58 ± 0.11), and this presumably reflects similar inhibition of the effects of endogeneous Foxp1 by ectopically-expressed WT Ikaros. Single infection with IK6 alone did not significantly alter Gpr132 expression levels, indicating that

121 any sequestration of endogeneous WT Ikaros by IK6 does not significantly inhibit

Ikaros/ Foxp1 interaction in this setting.

It is also notable that Foxp1-mediated induction of Gpr132 expression was higher in this set of assays (5.28 ± 0.45) than in the experiments described in Chapter

4 (2.33 ± 0.6). It is most likely that these differences arise from improved infection efficiency in the latter set of experiments, and a consequent increase in Foxp1- mediated effects. Although overall rates of GFP positivity were similarly high (at least

95% in each case), the mean fluorescence intensity of the transduced cell population was approximately 10-fold higher in the more recent infections, and this was presumably associated with higher Foxp1 activity. This may also explain an extra phenotypic effect that was observed in the latter experiments, as Foxp1 infection was associated with an increase in the proportion of cells in G1 and a reduction in the proportion of cells in S-phase that had not previously been observed (Figure 5.2). This additional finding indicates potential further roles for Foxp1 in the regulation of genes that are involved in cell cycle progression, that were only evident in the context of further increases in Foxp1 activity. Further investigation of these processes was not possible in the course of the investigation described in this thesis, but it is important to note that the key original phenotypic observation (G2 arrest) was consistently found in all performed experiments.

122

p < 0.001 ***

7 p < 0.001

6 ***

5

4

3

2 Expression Relative to Empty Vector 1

Gpr132 WT Ikaros Foxp1 WT Ikaros IK6 IK6 + Foxp1 + Foxp1

Figure 5.4 Expression of Gpr132 in infected B3 cells. Expression analysis was carried out as described in Chapter 4. Foxp1 infection resulted in increased levels of Gpr132, which were ‘rescued’ by co-infection with WT Ikaros, but not Ikaros isoform IK6. Statistical significance was calculated using Student’s paired t test and is indicated above the graph columns. Error bars indicate standard error of the mean (SEM). Averages of QPCR assays of three biological replicates are shown.

123

5.3 Summary

The experiments reported in this chapter shed further light on the results described in Chapter 4, and on the mechanism of Ikaros/Foxp1 interaction. Increased

Foxp1 activity in pre-B-cells resulted in clear phenotypic changes in lymphocyte cell cycle profile that were closely correlated with the expression of a key cell cycle molecule (G2A, encoded by the gene Gpr132 ). These effects were abolished by WT

Ikaros (which interacts physically with Foxp1), but not by Ikaros isoform IK6 (which does not interact with Foxp1, as shown in Chapter 3). This demonstrates that rescue of the G2 arrest phenotype by WT Ikaros is not due to non-specific protein expression effects, as relatively higher levels of IK6 had no effect on Foxp1 activity. The correspondence between the requirement for the central zinc finger domain of Ikaros both for the physical interaction between the two proteins and for the block on Foxp1- mediated enhancement of Gpr132 expression supports the idea that Ikaros exerts its effects on transcriptional regulation by Foxp1 through direct physical interaction.

124

Chapter 6

Analysis of Foxp1 binding to the murine Gpr132 locus

The experiments described in Chapters 4 and 5 have shown a close concordance between increases in Gpr132 transcription and the cell cycle phenotype that follows Foxp1 overexpression. The fact that the G2A protein product of the

Gpr132 gene is known to induce G2 arrest when ectopically expressed in mammalian cells (Le et al., 2002; Weng et al., 1998) strongly suggests that the G2 arrest that was observed is a direct consequence of the induction of Gpr132 expression.

We hypothesised that the G2 arrest and associated Gpr132 expression were dependent on binding of Foxp1 to regulatory sequences located within or close to the

Gpr132 gene. The abolition of these effects by WT Ikaros would therefore rely upon direct contact with the Foxp1 forkhead region, which would inhibit DNA binding.

Conversely, IK6 lacks the central zinc finger domain that is necessary for association with Foxp1 (Chapter 3) and so would not affect the amount of Foxp1 that is available for DNA interaction.

In order to test this hypothesis experimentally, chromatin immunoprecipitation

(ChIP) was performed. Chromatin was extracted from retrovirally infected B3 cells

(from the same experiments that are described in Chapter 5) and subjected to IP with an anti-FLAG antibody, which precipitated the ectopically-expressed FLAG-Foxp1.

Following reverse cross-linking and DNA purification, genomic regions in and around the Gpr132 gene were amplified by QPCR in order to determine the extent of Foxp1 binding in the presence and absence of ectopic WT Ikaros and IK6.

125

6.1 Identification of gene regulatory elements at the Gpr132 locus

The Gpr132 gene consists of three coding exons and is located at mouse chromosome 12: 114,089,087-114,099,127 (July 2007 (NCBI37/mm9) Assembly).

The regulatory elements that control Gpr132 expression have not been described in published literature, and it was therefore necessary to identify DNA regions in and around the gene locus at which Foxp1 might bind.

Transcription factors bind to enhancer sequences that are often located close to the transcriptional start site, although regulatory regions may be located at varying distances both upstream and downstream of the gene (see Chapter 1.1). Recent advances in high-throughput genomics have generated large amounts of published experimental data that can be used for the identification of putative gene promoter and enhancer elements. For example, ChIP followed by massive parallel sequencing

(ChIP-sequencing) has provided new evidence of the genome-wide patterns of histone modifications that are associated with enhancer location, including monomethylation of the lysine residue at position 4 of the N-terminal tail of histone H3 (H3K4me1). In addition, the presence of other histone modifications (such as trimethylation of H3K4) indicates active transcription that depends on contiguous transcription factor binding.

Much of this information is freely accessible through publicly available databases such as the UCSC genome browser ( http://genome.ucsc.edu ), and this resource was found to be the most useful in the assessment of potential sites of Foxp1 binding near the Gpr132 locus. This database also integrates additional information that identifies possible transcription factor binding sites, such as locations of DNase hypersensitivity.

126

The strategy that was adopted involved searching the genomic sequence directly for specific putative Foxp1 binding sequences, which have been described to contain the core seven nucleotides TATTTRT (where R = any nucleotide) (Wang et al., 2003). The genomic sequence in and around the Gpr132 gene was searched for this motif, in both the sense and antisense directions. In silico examination of the sequence (depicted schematically in Figure 6.1) identified two particular regions of interest, which were designated Regions ‘A’ and ‘B’. The genomic sequences of these regions are shown in Figure 6.2. Region A lies approximately 4kb downstream of the transcriptional start site, and includes four putative Foxp1 binding sites, according to the previously described consensus sequence. Region B lies approximately 1kb upstream of the transcriptional start site, within the area where a gene enhancer might typically be expected. Region B contains one putative Foxp1 binding site, and has been associated with the presence of H3K4me1 in published ChIP-sequencing data. In addition, each region borders an area of reported DNase hypersensitivity in adult (8 week-old) murine B-cells, which is suggestive of transcription factor binding. The extent of Foxp1 occupancy of these loci in retrovirally-infected B3 cells was assessed in ChIP experiments, which are described in the next section.

127

Mouse Gpr132 Locus

5000 0 +5000 +10000 +15000

Scale TSS

Exon1 Exon2 Exon3

Region A Region B

Figure 6.1 Location of the genomic regions amplified in Gpr132 ChIP . A schematic representation of the Gpr132 locus is shown, with the scale depicted in bases. The locations of ChIP regions A and B are indicated with arrows. Coding exons are represented by filled boxes. TSS = Transcriptional start site.

128

Region A : Mouse Chromosome 12:114,100,053-114,100,552 GGCTTGGGGGTGTTGGCCCCATCCCTGACAGCTCCCAGAGTCCTCAAAACT GAAGGCAAAGCTGGCTTGGTAGCCTGTGTTCAGCTAACACTTTCAGAAGC GCTAGCCGACGAAGCCCAGCTTTACAGAGCAGTTTGAAAGGTCATGACCC AGATAGCAGCTCCCTGCCTAAGTCCAAAGACAGACATGCTCTCTGAGAGT CATGCCAGGCTTCAAGGCTGTGAGGCTGGAGGCCTTTCTTTCTCACCCAAC CCCTAACAGACATTTACATATTTGTGTGGGTGGCAGTACAGTCCCTAGAAA GCTGGCCCTCAGGGGTGTGGGGCAGACAGCTGGCATGCGGGCCTGAAGAT AGTCCATCTTAACTTTCTGAAAGCAGCTGTTCGTGGGCCGTCGCCTTCAAG GAATCTAGTGCACAGTAAGTCCTCCCTGGGCTCTGACCACGATGCCACTCG TACCTGACCTTGGAGGGCAAAGGGGTTTGATGGCTCTGGCACAA

Region B : Mouse Chromosome 12:114,094,726-114,095,225 TACAGACAACCAGGCACCAGGGAGCATCAAAGCATCACCCTGGATACCAG AGTTAAAACTGTTATTTGTTGGTTATGGTTGGCAAAGCTGATGGCACATGA GCTCCCACAGAAGAATTAAGAGCTCCCAAAAGGCGAAGAAACGCAAAAT CAGCATGAAGAACGGTCTCTTTGTTTTGTTTTGTTTTCCTGCCTATCAGGCT GACAAAAATAACTAAATCAGTGTTCTGAAGTCTGAAGACATGGCACCCTG AAGCTACACAACGGGGTGGTCACTAGGTACCAATATTTCAGAGCCCCACT TTGCAGTCAGTTTCCAACACCGTTAAGAAAGTCTGAACTTGCTAAAATTCC TCCTAGGGATTTTTTATTTCTGATTAAAAATAATGACCTGAAAAGTTAAGT GTGTCTGACCTGCTGCCCACCTGCTCCCCCGACACACATGGCTGCCCCTCA GCCAGGCAGATAACCACGGTATTTCCCCAGAGTGCATTGGGCCGC

Figure 6.2 Genomic Sequences of Gpr132 ChIP Regions A and B. The sequences and genomic locations (July 2007 (NCBI37/mm9) Assembly) of Regions A and B that were amplified by QPCR following anti-FLAG ChIP of retrovirally infected B3 cells are shown. Putative Foxp1 binding sites are highlighted.

129

6.2 Examination of Foxp1 binding to the Gpr132 locus by

Chromatin immunoprecipitation (ChIP)

Chromatin was isolated from retrovirally infected B3 cells (see Section 2.12) and subjected to IP with an anti-FLAG antibody. IP of chromatin extracted from cells that had been infected with empty retroviral vector was performed concurrently, in order to provide a control for non-specific antibody precipitation. Following reverse cross-linking, pulldown of the sequences of Regions A and B was assessed by QPCR amplification.

The retroviral infection experiments were associated with high levels of ectopic protein expression, which in turn led to a significant amount of baseline ‘background’ binding of FLAG-Foxp1 to chromatin. Our group has previously found that Foxp1 tends to show relatively broad patterns of chromatin binding in ChIP experiments

(Lieber, 2008 and Domaschenz, unpublished data), while Forkhead box proteins have been described to exhibit high-affinity non-specific binding to nucleosomes (Sekiya et al., 2009). Inclusion of an experimental control to allow for these effects was therefore very important, and was integral to accurate estimation of Foxp1 chromatin occupancy.

Two genomic regions (here arbitrarily designated as Background Regions 1 and 2) were used that to date have exhibited an absence of transcription factor binding in reported ChIP-sequencing data (Natalia Kunowska, personal communication).

Background Region1 is located at mouse chromosome 17:86,190,215-86,190,285, while Region 2 is found at chromosome 4:55,591,773-55,591,842 (July 2007

(NCBI37/mm9) Assembly). These sequences were assayed by QPCR following anti-

130

FLAG ChIP, and the resultant levels of amplification (typically approximately 2% of input, henceforth referred to as ‘background’) were subtracted from the corresponding amplification figures for the test regions, in order to give a quantification of FLAG-

Foxp1 binding to Gpr132 Regions A and B.

The results of the ChIP experiments were concordant with Gpr132 expression levels. Foxp1 bound strongly to both Regions A and B, at levels greatly in excess of background. Figure 6.3 shows the mean % input figures of five QPCR experiments, following subtraction of background. The separate control for non- specific antibody pulldown (i.e. anti-FLAG IP of the empty vector infection) was associated with essentially no pulldown (typically < 10 -5 % input), and these figures are not shown in the graph. In addition, the results of separate QPCR assays using different primer pairs that amplified the same genomic regions were also highly concordant (data not shown).

Co-infection with WT Ikaros reduced the extent of Foxp1 binding to levels that were close to background in a highly reproducible manner (Figure 6.3). Co-infection with IK6 had minimal effects on Foxp1 binding, which remained very similar to the levels seen following Foxp1 infection alone. The differences between the amounts of binding that were seen following each retroviral infection were analysed by Student’s paired t-test and found to be statistically significant for all comparisons with WT

Ikaros/ Foxp1 co-infection (see legend to Figure 6.3).

Taken together, these data provide strong evidence that the reduction of

Gpr132 expression when WT Ikaros is co-expressed with Foxp1 is the result of specific displacement of Foxp1 from enhancer elements a t the Gpr132 gene. This

131 finding provides a mechanistic explanation for the observed abrogation of the G2 arrest phenotype that is seen following co-expression of WT Ikaros.

p< 0.05 6 * p< 0.05 5 * p< 0.05 4 * p< 0.01 ** 3 % Input 2

1

Foxp1 WTIkaros IK6 Foxp1 WTIkaros IK6 + Foxp1 + Foxp1 + Foxp1 + Foxp1 Region A Region B

Figure 6.3 ChIP analysis of binding of Foxp1 to the Gpr132 gene. Chromatin from B3 cells infected with different retroviral constructs (indicated below each bar) was precipitated with anti-FLAG antibody, and Gpr132 regions A and B (Figures 6.1 and 6.2) were amplified by QPCR. Figures for % input were obtained after subtraction of background levels, as described in Section 6.2. The differences between WT Ikaros/Foxp1 co-infections and both Foxp1 single infection and IK6/Foxp1 co- infections were analysed using Student’s paired t-test, and were found to be highly significant, as shown. The averages of QPCR assays of three biological replicates are shown.

132

6.3 Summary The results of these ChIP experiments show that the extent of chromatin occupancy of Foxp1 is affected by altering the levels of its interaction partner, WT

Ikaros. Foxp1 was found to bind specifically to the Gpr132 locus, as the ChIP values for control genomic regions (‘background’) were markedly lower. The effect of expressing WT Ikaros on the binding of Foxp1 to the Gpr132 was specific, as expression of relatively higher quantities of IK6 (Figure 5.3) had no effect on Foxp1 occupancy of the Gpr132 locus. Unfortunately, attempts at demonstrating a similar phenomenon by electrophoretic mobility shift assay (EMSA) were unsuccessful, as poor binding of purified GST-Foxp1 to oligonucleotides was observed in vitro .

These data provide compelling evidence that the interaction between Ikaros and

Foxp1 directly affects the DNA-binding activity of Foxp1, and a coherent mechanism for the cell cycle and gene expression effects that were observed may therefore be suggested. Foxp1 binding to Gpr132 regulatory elements stimulates expression of the

G2A protein which triggers G2 arrest. The effects of Foxp1 excess are abrogated by the presence of equivalent amounts of WT Ikaros, which occupies the DNA-binding domain of Foxp1, thereby preventing the upregulation of Gpr132 expression. IK6 has no effect on gene expression or the cell cycle phenotype, because it does not interact with Foxp1. The original model shown in Figure 4.1 may now be updated to incorporate the findings of the intervening experiments (Figure 6.4).

133

Foxp1 overexpression

Gpr132

+ WT Ikaros + IK6

Gpr132 Gpr132

Figure 6.4 Updated model of the effects of protein imbalance on expression of gene targets. The upper panel depicts the effects of overexpression of Foxp1 (blue pies), which binds to DNA to activate Gpr132 expression. Co-infection of WT Ikaros (lower left panel) inhibits induction of Gpr132 expression through occupancy of the Foxp1 DNA-binding domain. Co-infection of IK6 (lower right panel) has no effect on the amounts of Foxp1 that bind DNA to induce Gpr132 expression.

134

As with the original description of the model in Figure 4.1, this diagram shows one potential ramification of an imbalance in factor levels, and several other outcomes are possible. For example, this model would be invalidated in the context of a gene locus where Ikaros and Foxp1 binding are not mutually exclusive, as corresponding effects on Ikaros regulation of gene targets would be expected to confound the output.

We have however excluded this possibility by examination of ChIP-sequencing data that were published by collaborators in our institute (Ferreiros-Vidal et al., 2013), as

Ikaros was not found to bind to the Gpr132 locus in an identical B3 cell system to that used in the experiments I have described.

135

Chapter 7

Discussion

Ikaros and Foxp1 have been established as key regulators of B-cell differentiation and proliferation (Georgopoulos et al., 1994; Hu et al., 2006).

Mutations and/ or deletions in both proteins are also strongly implicated in lymphoid malignancy, including B-ALL (Brown et al., 2008; Mullighan and Downing, 2008;

Put et al., 2011; Sagaert et al., 2006). The work described in this thesis has characterised a hitherto unrecognised physical and functional interaction between these factors. The physical association between the proteins is direct, and requires the

DNA-binding domains of both molecules, as shown by the results of the experiments reported in Chapter 3.

The normal functions and patterns of expression of each protein and the locations of the domains that mediate the interaction between them raised the possibility that they might affect one another’s functioning. This effect could be synergistic, with the interaction affecting the specificity or degree of DNA binding.

Alternatively, the effect could be antagonistic, with contact between the two binding domains blocking binding of either or both factors to transcriptional regulatory elements. In addition, the interaction between the proteins may be necessary for recruitment of other factors that are involved in the modulation of gene expression, such as transcriptional activators or repressors.

The experimental findings detailed in Chapters 4-6 demonstrate that at least one type of effect is antagonistic. Overexpression of Foxp1 in a pre-B cell line resulted in increased transcription of Gpr132 , which encodes the G-protein coupled

136 receptor G2A protein that is known to cause G2 cell cycle arrest in mammalian cells

(Weng et al., 1998). Foxp1-mediated induction of Gpr132 expression and phenotypic cell cycle arrest were both reversed by co-expression of WT Ikaros, which was shown to block binding of Foxp1 to sequences in and around the Gpr132 gene. This indicates that the abrogation of Foxp1-mediated effects is contingent upon direct inhibition of binding of the protein to DNA regulatory sequences.

Co-expression of the Ikaros isoform IK6 had no effect on the cell cycle phenotype, Gpr132 expression levels, or Foxp1 binding to chromatin. These results strongly suggest that the inhibition of Foxp1-related effects is dependent on direct protein-protein contact that follows restoration of the equilibrium between the relative amounts of Ikaros and Foxp1 proteins. In addition, this finding provides further evidence for the mediation of protein contact through the Ikaros central zinc finger domain in vivo .

These novel findings provide important new insights into the roles of Ikaros and Foxp1 in the maintenance of cell cycle homeostasis in B-lymphocytes. In the final section of this thesis, I will discuss how the functional implications of the interaction might be further elucidated (Section 7.1), and how these results may contribute to the understanding of normal physiological (Section 7.2) and pathological (Section 7.3) lymphoid cell cycle regulation.

137

7.1 Functional consequences of the Ikaros-Foxp1 interaction

As both Ikaros and Foxp1 are expressed at high levels from the early stages of

B-lymphopoiesis (Georgopoulos et al., 1994; Wang et al., 2003), the data described in this thesis strongly suggest that interaction between these factors could contribute to the regulation of a number of genes during the course of B-cell development. By extension, this suggests that fluctuations in the relative amounts (and patterns of isoform expression) of each protein at different stages of differentiation contribute to cellular phenotype.

Intuitively, occupation of the DNA-binding domain of a transcription factor by another protein might be expected to inhibit factor binding to gene targets. In keeping with this hypothesis (as shown in Figures 4.1 and 6.4), we have shown that one effect of the interaction between the proteins is antagonistic, as Ikaros directly blocked binding of Foxp1 to putative regulatory elements of the Gpr132 gene. Our findings do not however preclude the occurrence of different effects at other gene loci. Protein- protein association has been described to induce conformational changes that enhance transcription factor activity (Nagulapalli et al., 2002), and it is also possible that the presence and/ or recruitment of additional molecules at other genomic sites (or in other cell types) may result in enhanced binding of one or both factors to DNA.

Elucidation of the full spectrum of functional consequences of Ikaros-Foxp1 interaction may be provided by experimental approaches that comprehensively examine the global range of transcription factor activity, using unbiased approaches.

For example, gene expression analysis by microarray would clarify the range of

Foxp1 transcriptional effects in B-lymphoid cells, while ChIP-sequencing would

138 provide a comprehensive picture of the genome-wide patterns of Foxp1 binding at differing levels of protein expression. Integration with published measurements of

Ikaros binding patterns (Zhang et al., 2011, Ferreiros-Vidal et al., 2013) should identify further putative targets that may be influenced by the protein-protein association, particularly if the reciprocal effect (i.e. Foxp1 blocking Ikaros binding to

DNA) also affects transcriptional regulation. Further clarification of the influences of

Ikaros-Foxp1 interaction on the dynamics of factor binding would also be provided by experimental models in which the expression of one or both factors is suppressed either by genetic knockout or RNA interference.

There is a precedent for interaction between Ikaros family members and

Forkhead transcription factors, as Eos (a widely-expressed Ikaros family member) and

Foxp3 were found to associate directly in CD4+ Regulatory T-cells in a study that was published during the course of this work. This interaction was shown to be critical for recruitment of transcriptional repressors and Foxp3-mediated gene silencing (Pan et al., 2009). No interaction was observed between Foxp3 and Ikaros in these experiments. In contrast to the antagonistic interaction that was observed between

Ikaros and Foxp1 in this thesis, the interaction between Eos and Foxp3 seems to be synergistic, as Eos was shown to be essential for recruitment of transcriptional repressors and gene silencing by Foxp3.

A key difference between the Eos-Foxp3 interaction and the Ikaros-Foxp1 interaction is the location of the domains of protein association. The region that is necessary for interaction with Eos is located some distance away from the Foxp3 forkhead domain, in the N-terminal region of the protein. Similarly, the C-terminal

139 portion (300 amino acids) of Eos was found to be sufficient for interaction with

Foxp3, indicating that the DNA-binding region is not required for protein-protein contact. This suggests that the Ikaros-Foxp1 and Eos-Foxp3 interactions arose separately during the evolution of immune regulation, and are not related functionally.

It would be worthwhile to investigate the existence of any further interactions between Forkhead proteins and Ikaros family members, as the latter proteins exhibit significant structural homology. In particular, the existence of an interaction with other Ikaros proteins could provide mechanistic insight into the regulation of the cell cycle at differing stages of B-cell differentiation. For example, fluctuations in the amounts of Aiolos are known to directly affect cell cycle exit at the pre-B cell stage

(see Section 1.4.2), and the existence (or otherwise) of an interaction with Foxp1 would have implications for the mechanism of Aiolos-mediated effects.

140

7.2 Ikaros and Foxp1 regulate the cell cycle of pre-B-cells

Regulation of the cell cycle is crucial for orderly progression of B- lymphopoiesis (Jumaa et al., 2005). In particular, cell cycle exit is required for differentiation from the large pre-B-II stage (which is characterised by high rates of cell cycling and proliferation that are mediated by the pre-BCR) to the small, non- cycling pre-B stage. Ikaros is known to modulate the lymphocyte cell cycle at this developmental transition, as Ikaros-mediated repression of c-myc transcription has been shown to be involved in triggering G1 arrest and cell cycle exit (Ma et al., 2010).

To our knowledge, Foxp1 has not previously been described to directly affect the cell cycle of B-lymphocytes, although it has been shown to promote T-cell quiescence through transcriptional repression of the IL-7 receptor α-chain gene

(IL7R α) (Feng et al., 2010; Feng et al., 2011). This does not preclude an alternative function in B-cells, as the effects of Foxp1 on cellular proliferation appear to be tissue-specific, being either pro- (Rao et al., 2010) or anti-proliferative (Zhang et al.,

2010), depending on cellular context, as detailed in Section 1.5.

Further characterisation of the role of Foxp1 in B-lymphocyte cell cycle regulation is necessary, in light of the clear phenotypic effects that were observed following modulation of protein expression in these experiments. The fact that Foxp1- induced G2 arrest can be overcome by co-expression of equivalent amounts of Ikaros suggests that maintenance of homeostatic levels of Ikaros and Foxp1 could be necessary for normal progression of pre-B cells through the G2/M stages of the cell cycle. The additional observation of G1 arrest at higher levels of Foxp1 expression also suggests that the Ikaros-Foxp1 interaction may influence gene expression at

141 several stages of the cell cycle. Again, clarification of the influence of the interaction on the cell cycle phenotype of haematopoietic cells would be provided by knockdown experiments as discussed in Section 7.1. It would also be interesting to examine the later effects of Foxp1 overexpression in pre-B-cells, as all functional analyses that were described in B3 cells were performed 48 hours after infection, at which point all available cells were harvested for testing. In particular, it would be important to clarify whether the described cell cycle effects influence cellular viability and/ or differentiation.

The data described here show that Foxp1-induced G2 arrest is clearly corrrelated with upregulation of Gpr132 expression, and the observed cellular phenotype is consistent with previous reports of this gene’s function. Gpr132 encodes the G2A protein, a G-protein coupled receptor for oxidised free fatty acids

(Kabarowski et al., 2001). G2A is known to induce G2 cell cycle arrest when overexpressed in NIH 3T3 cells, although the consistency of this effect between cell types has not been established. For example, overexpression of the protein in keratinocytes produces a different phenotype, as a G1 block is seen (Obinata and

Izumi, 2009). Normal tissue expression of Gpr132 is highest in haematopoietic cells, including both T- and B-lymphocytes (Weng et al., 1998). G2A has been reported to inhibit lymphocyte proliferation, as G2A-deficient mice have hyperproliferative T- cells and develop an autoimmune syndrome in adult life (Le et al., 2001). This effect is consistent with the observation of decreased cellular proliferation of Foxp1-infected cells in the experiments described in this thesis.

142

Knowledge of the normal function of G2A in haematopoietic cells is lacking, although recent evidence suggests an involvement in modulation of activation (Bolick et al., 2009; Frasch et al., 2011). The downstream effects of G2A

(and any activation of cell cycle related molecular pathways) are poorly understood.

Our results suggest a potential role for G2A in regulating the cell cycle in normal lymphocytes; in particular, the hypothesis that suppression of Gpr132 expression may be necessary for progression of pre-B cells through G2 requires further investigation, and this would be provided by experiments in which the expression of Gpr132 is suppressed by RNA interference. Conversely, manipulation of Gpr132 levels by overexpression would be predicted to trigger cell cycle arrest of pre-B-cells, if the hypotheses described herein are correct.

7.3 Loss of interaction between Ikaros and FOXP1 may contribute to chemotherapeutic resistance in B-ALL

As human Ikaros and FOXP1 exhibit similar patterns of inter-protein association to their murine counterparts (Chapter 3), it is likely that the interaction between the proteins is also important for regulation of cell cycle progression in human B-cells. By extension, disruption of this interaction would be expected to result in altered cell cycle control and perturbation of B-lymphopoiesis. The results described in this thesis therefore have relevance for the elucidation of the downstream effects of quantitative imbalances of transcription factors that characterise lymphoid malignancy.

143

This is especially pertinent in the consideration of Ph+ B-ALL, which is characterised by expression of the BCR-ABL oncogene and by a poor long-term prognosis. This leukaemia is associated with high rates of genomic deletion of IKZF1 , which result in expression of IK6 and reduced levels of full-length WT Ikaros

(Mullighan et al., 2008). Both of these changes would be predicted to increase the amounts of FOXP1 available for induction of expression of target genes such as

GPR132 . We therefore propose that FOXP1-mediated effects could be important for the pathogenesis of these leukaemias. Importantly, IKZF1 deletion is associated with similar gene expression profiles in both Ph+ and non Ph+ B-ALL, suggesting that similar mechanisms of transcriptional dysregulation may contribute to leukaemogenesis, independent of the effects of BCR-ABL (Mullighan and Downing,

2008).

Previously published data have suggested a functional interplay between BCR-

ABL and G2A. GPR132 was originally identified in a search for genes that were transcriptionally upregulated by p185 BCR-ABL, and forced expression of G2A was shown to antagonise the transformative effects of BCR-ABL in vitro (Weng et al.,

1998). In keeping with these observations, G2A was shown to have a negative effect on in vivo leukaemic cell proliferation in a mouse model of Ph+ALL (Le et al., 2002).

The effects of G2A deficiency on cell cycle profile were not reported in this study.

The authors of this paper speculated that increased G2A expression in Ph+ALL was a favourable prognostic indicator, as it should lead to a diminished leukaemic cell mass.

It is however also possible that these anti-proliferative effects actually confer a negative prognosis, as G2A may increase leukaemic cell resistance to

144 chemotherapeutic agents which are preferentially toxic to rapidly-dividing cells. In addition, G2 cell cycle arrest is believed to contribute to chemotherapeutic resistance in haematological malignancy, as delay of G2/M transition provides additional time for repair of DNA damage that might otherwise trigger cell death at the mitotic checkpoint (Bedi et al., 1995). The mechanism by which G2 arrest occurs in these leukaemias is currently poorly understood. Attempts were made to recapitulate this described G2 arrest by treating ALL cell lines with a number of chemotherapeutic agents, however the global cell cycle toxicity that was observed precluded targeted analysis.

Our results suggest that upregulation of GPR132 as a result of increased

FOXP1 activity that is consequent upon IKZF1 deletion could result in G2 arrest in these leukaemias, which in turn might contribute to treatment resistance and poor prognosis. If this is correct, it should be possible to demonstrate increased levels of

GPR132 expression in IKZF1 -deleted ALL. Preliminary assays performed by our laboratory and collaborators have however failed to show any correlation between

IKZF1 deletion and GPR132 levels in either ALL cell lines or primary patient samples

(data not shown). This might suggest that the expression of GPR132 is influenced by additional factors in ALL cells. Murine and human G2A are known to differ in activity, particularly with respect to ligand specificity (Radu et al., 2005), and it is probable that the regulation of GPR132 expression differs between the species. It may also be the case that these results fail to show a correlation between IKZF1 deletion and G2A expression because of the timing of the analysis. QPCR was performed on material isolated from diagnostic samples, and it is possible that FOXP1-mediated

145

GPR132 induction may only become pathologically relevant under conditions of cellular stress (for example, following chemotherapy). These results do not preclude a functional importance for G2A in IKZF1 -deleted ALL, and further work is necessary to examine the potential role of this molecule in the regulation of the pathologically subverted lymphoid cell cycle. In particular, analysis of the cell cycle of IKZF1 wild- type and deleted ALL cells (and correlation with any changes in FOXP1-mediated gene regulation and DNA-binding patterns that might exist) would provide insight into the effects of Ikaros protein levels on this aspect of cellular phenotype. The work described in this thesis will hopefully provide the impetus for new avenues of research in this area that may ultimately lead to better treatments for this disease.

146

References

• Calcium phosphate-mediated transfection of eukaryotic cells. Nat Meth 2, 319- 320. • Advani, A.S., Hunger, S.P., and Burnett, A.K. (2009). Acute in adolescents and young adults. Semin Oncol 36 , 213-226. • Alber, T. (1992). Structure of the leucine zipper. Curr Opin Genet Dev 2, 205- 210. • Allman, D., Sambandam, A., Kim, S., Miller, J.P., Pagan, A., Well, D., Meraz, A., and Bhandoola, A. (2003). Thymopoiesis independent of common lymphoid progenitors. Nat Immunol 4, 168-174. • Armstrong, S.A., and Look, A.T. (2005). Molecular genetics of acute lymphoblastic leukemia. J Clin Oncol 23 , 6306-6315. • Arnosti, D.N., and Kulkarni, M.M. (2005). Transcriptional enhancers: Intelligent enhanceosomes or flexible billboards? J Cell Biochem 94 , 890-898. • Bain, G., Maandag, E.C., Izon, D.J., Amsen, D., Kruisbeek, A.M., Weintraub, B.C., Krop, I., Schlissel, M.S., Feeney, A.J., van Roon, M. , et al. (1994). E2A proteins are required for proper B cell development and initiation of immunoglobulin gene rearrangements. Cell 79 , 885-892. • Banham, A.H., Beasley, N., Campo, E., Fernandez, P.L., Fidler, C., Gatter, K., Jones, M., Mason, D.Y., Prime, J.E., Trougouboff, P. , et al. (2001). The FOXP1 winged helix transcription factor is a novel candidate on chromosome 3p. Cancer Res 61 , 8820-8829. • Bardini, M., Spinelli, R., Bungaro, S., Mangano, E., Corral, L., Cifola, I., Fazio, G., Giordan, M., Basso, G., De Rossi, G. , et al. DNA copy-number abnormalities do not occur in infant ALL with t(4;11)/MLL-AF4. Leukemia 24 , 169-176. • Bartek, J., Lukas, C., and Lukas, J. (2004). Checking on DNA damage in S phase. Nat Rev Mol Cell Biol 5, 792-804. • Bedi, A., Barber, J.P., Bedi, G.C., el-Deiry, W.S., Sidransky, D., Vala, M.S., Akhtar, A.J., Hilton, J., and Jones, R.J. (1995). BCR-ABL-mediated inhibition

147

of apoptosis with delay of G2/M transition after DNA damage: a mechanism of resistance to multiple anticancer agents. Blood 86 , 1148-1158. • Bernstein, B.E., Meissner, A., and Lander, E.S. (2007). The mammalian epigenome. Cell 128 , 669-681. • Bhojwani, D., Pei, D., Sandlund, J.T., Jeha, S., Ribeiro, R.C., Rubnitz, J.E., Raimondi, S.C., Shurtleff, S., Onciu, M., Cheng, C., et al. ETV6-RUNX1- positive childhood acute lymphoblastic leukemia: improved outcome with contemporary therapy. Leukemia 26 , 265-270. • Boeger, H., Bushnell, D.A., Davis, R., Griesenbeck, J., Lorch, Y., Strattan, J.S., Westover, K.D., and Kornberg, R.D. (2005). Structural basis of eukaryotic gene transcription. FEBS Lett 579 , 899-903. • Bolick, D.T., Skaflen, M.D., Johnson, L.E., Kwon, S.C., Howatt, D., Daugherty, A., Ravichandran, K.S., and Hedrick, C.C. (2009). G2A deficiency in mice promotes macrophage activation and atherosclerosis. Circ Res 104 , 318-327. • Bouaziz, J.D., Yanaba, K., Venturi, G.M., Wang, Y., Tisch, R.M., Poe, J.C., and Tedder, T.F. (2007). Therapeutic B cell depletion impairs adaptive and autoreactive CD4+ T cell activation in mice. Proc Natl Acad Sci U S A 104 , 20878-20883. • Bradford, M.M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72 , 248-254. • Branco, M.R., Ficz, G., and Reik, W. Uncovering the role of 5- hydroxymethylcytosine in the epigenome. Nat Rev Genet 13 , 7-13. • Brivanlou, A.H., and Darnell, J.E., Jr. (2002). Signal transduction and the control of gene expression. Science 295 , 813-818. • Brown, K.E., Guest, S.S., Smale, S.T., Hahm, K., Merkenschlager, M., and Fisher, A.G. (1997). Association of transcriptionally silent genes with Ikaros complexes at centromeric heterochromatin. Cell 91 , 845-854. • Brown, P.J., Ashe, S.L., Leich, E., Burek, C., Barrans, S., Fenton, J.A., Jack, A.S., Pulford, K., Rosenwald, A., and Banham, A.H. (2008). Potentially 148

oncogenic B-cell activation-induced smaller isoforms of FOXP1 are highly expressed in the activated B cell-like subtype of DLBCL. Blood 111 , 2816- 2824. • Bulger, M., and Groudine, M. Enhancers: the abundance and function of regulatory sequences beyond promoters. Dev Biol 339 , 250-257. • Buske, C., and Humphries, R.K. (2000). Homeobox genes in leukemogenesis. Int J Hematol 71 , 301-308. • Busslinger, M. (2004). Transcriptional control of early B cell development. Annu Rev Immunol 22 , 55-79. • Caldecott, K.W. (2008). Single-strand break repair and genetic disease. Nat Rev Genet 9, 619-631. • Carlsson, P., and Mahlapuu, M. (2002). Forkhead transcription factors: key players in development and metabolism. Dev Biol 250 , 1-23. • Chatila, T.A., Blaeser, F., Ho, N., Lederman, H.M., Voulgaropoulos, C., Helms, C., and Bowcock, A.M. (2000). JM2, encoding a fork head-related protein, is mutated in X-linked autoimmunity-allergic disregulation syndrome. J Clin Invest 106 , R75-81. • Chokas, A.L., Trivedi, C.M., Lu, M.M., Tucker, P.W., Li, S., Epstein, J.A., and Morrisey, E.E. Foxp1/2/4-NuRD interactions regulate gene expression and epithelial injury response in the lung via regulation of interleukin-6. J Biol Chem 285 , 13304-13313. • Clark, K.L., Halay, E.D., Lai, E., and Burley, S.K. (1993). Co-crystal structure of the HNF-3/fork head DNA-recognition motif resembles histone H5. Nature 364 , 412-420. • Clark, S.S., McLaughlin, J., Crist, W.M., Champlin, R., and Witte, O.N. (1987). Unique forms of the abl tyrosine kinase distinguish Ph1-positive CML from Ph1-positive ALL. Science 235 , 85-88. • Cobaleda, C., Jochum, W., and Busslinger, M. (2007). Conversion of mature B cells into T cells by dedifferentiation to uncommitted progenitors. Nature 449 , 473-477.

149

• Cobb, B.S., Morales-Alcelay, S., Kleiger, G., Brown, K.E., Fisher, A.G., and Smale, S.T. (2000). Targeting of Ikaros to pericentromeric heterochromatin by direct DNA binding. Genes Dev 14 , 2146-2160. • Cooper, M.D., Peterson, R.D., and Good, R.A. (1965). Delineation of the Thymic and Bursal Lymphoid Systems in the Chicken. Nature 205 , 143-146. • Cortes, M., Wong, E., Koipally, J., and Georgopoulos, K. (1999). Control of lymphocyte development by the Ikaros gene family. Curr Opin Immunol 11 , 167-171. • Courey, A.J., and Jia, S. (2001). Transcriptional repression: the long and the short of it. Genes Dev 15 , 2786-2796. • Cox, E.H., and McLendon, G.L. (2000). Zinc-dependent protein folding. Curr Opin Chem Biol 4, 162-165. • Coyaud, E., Struski, S., Dastugue, N., Brousset, P., Broccardo, C., and Bradtke, J. PAX5-AUTS2 fusion resulting from t(7;9)(q11.2;p13.2) can now be classified as recurrent in B cell acute lymphoblastic leukemia. Leuk Res 34 , e323-325. • Cramer, P. (2002). Multisubunit RNA polymerases. Curr Opin Struct Biol 12 , 89-97. • Crist, W.M., Carroll, A.J., Shuster, J.J., Behm, F.G., Whitehead, M., Vietti, T.J., Look, A.T., Mahoney, D., Ragab, A., Pullen, D.J. , et al. (1990). Poor prognosis of children with pre-B acute lymphoblastic leukemia is associated with the t(1;19)(q23;p13): a Pediatric Oncology Group study. Blood 76 , 117- 122. • Danilova, N. The evolution of adaptive immunity. Adv Exp Med Biol 738 , 218-235. • de Wit, E., and de Laat, W. A decade of 3C technologies: insights into nuclear organization. Genes Dev 26 , 11-24. • DeKoter, R.P., Lee, H.J., and Singh, H. (2002). PU.1 regulates expression of the interleukin-7 receptor in lymphoid progenitors. Immunity 16 , 297-309.

150

• Delogu, A., Schebesta, A., Sun, Q., Aschenbrenner, K., Perlot, T., and Busslinger, M. (2006). Gene repression by Pax5 in B cells is essential for blood cell homeostasis and is reversed in plasma cells. Immunity 24 , 269-281. • Diffley, J.F. (2004). Regulation of early events in chromosome replication. Curr Biol 14 , R778-786. • Dillon, N., Trimborn, T., Strouboulis, J., Fraser, P., and Grosveld, F. (1997). The effect of distance on long-range chromatin interactions. Mol Cell 1, 131- 139. • Dillon, N. (2006). Gene regulation and large-scale chromatin organization in the nucleus. Chromosome Res 14 , 117-126. • Doulatov, S., Notta, F., Eppert, K., Nguyen, L.T., Ohashi, P.S., and Dick, J.E. Revised map of the human progenitor hierarchy shows the origin of and dendritic cells in early lymphoid development. Nat Immunol 11 , 585-593. • Downing, J.R. (2003). The core-binding factor : lessons learned from murine models. Curr Opin Genet Dev 13 , 48-54. • Ekholm, S.V., and Reed, S.I. (2000). Regulation of G(1) cyclin-dependent kinases in the mammalian cell cycle. Curr Opin Cell Biol 12 , 676-684. • Elgin, S.C. (1996). Heterochromatin and gene regulation in Drosophila. Curr Opin Genet Dev 6, 193-202. • Enserink, J.M., and Kolodner, R.D. An overview of Cdk1-controlled targets and processes. Cell Div 5, 11. • Feng, X., Ippolito, G.C., Tian, L., Wiehagen, K., Oh, S., Sambandam, A., Willen, J., Bunte, R.M., Maika, S.D., Harriss, J.V., et al. Foxp1 is an essential transcriptional regulator for the generation of quiescent naive T cells during thymocyte development. Blood 115 , 510-518. • Feng, X., Wang, H., Takata, H., Day, T.J., Willen, J., and Hu, H. Transcription factor Foxp1 exerts essential cell-intrinsic regulation of the quiescence of naive T cells. Nat Immunol 12 , 544-550.

151

• Ferreiros-Vidal, I., Carroll, T., Taylor, B., Terry, A., Liang, Z., Bruno, L., Dharmalingam, G., Khadayate, S., Cobb, B.S., Smale, S.T. , et al. Genome- wide identification of Ikaros targets elucidates its contribution to mouse B-cell lineage specification and pre-B-cell differentiation. Blood 121 , 1769-1782. • Fisher, A.G., Burdet, C., Bunce, C., Merkenschlager, M., and Ceredig, R. (1995). Lymphoproliferative disorders in IL-7 transgenic mice: expansion of immature B cells which retain macrophage potential. Int Immunol 7, 415-423. • Fletcher, J.A., Lynch, E.A., Kimball, V.M., Donnelly, M., Tantravahi, R., and Sallan, S.E. (1991). Translocation (9;22) is associated with extremely poor prognosis in intensively treated children with acute lymphoblastic leukemia. Blood 77 , 435-439. • Fox, S.B., Brown, P., Han, C., Ashe, S., Leek, R.D., Harris, A.L., and Banham, A.H. (2004). Expression of the forkhead transcription factor FOXP1 is associated with alpha and improved survival in primary human breast carcinomas. Clin Cancer Res 10 , 3521-3527. • Frasch, S.C., Fernandez-Boyanapalli, R.F., Berry, K.Z., Leslie, C.C., Bonventre, J.V., Murphy, R.C., Henson, P.M., and Bratton, D.L. Signaling via macrophage G2A enhances efferocytosis of dying neutrophils by augmentation of Rac activity. J Biol Chem 286 , 12108-12122. • Gabut, M., Samavarchi-Tehrani, P., Wang, X., Slobodeniuc, V., O'Hanlon, D., Sung, H.K., Alvarez, M., Talukder, S., Pan, Q., Mazzoni, E.O. , et al. An alternative splicing switch regulates embryonic stem cell pluripotency and reprogramming. Cell 147 , 132-146. • Gardiner-Garden, M., and Frommer, M. (1987). CpG islands in vertebrate genomes. J Mol Biol 196 , 261-282. • Garvie, C.W., and Wolberger, C. (2001). Recognition of specific DNA sequences. Mol Cell 8, 937-946. • Georgopoulos, K., Bigby, M., Wang, J.H., Molnar, A., Wu, P., Winandy, S., and Sharpe, A. (1994). The Ikaros gene is required for the development of all lymphoid lineages. Cell 79 , 143-156.

152

• Georgopoulos, K., Moore, D.D., and Derfler, B. (1992). Ikaros, an early lymphoid-specific transcription factor and a putative mediator for T cell commitment. Science 258 , 808-812. • Georgopoulos, K., Winandy, S., and Avitahl, N. (1997). The role of the Ikaros gene in lymphocyte development and homeostasis. Annu Rev Immunol 15 , 155-176. • Glotzer, M. (2005). The molecular requirements for cytokinesis. Science 307 , 1735-1739. • Gounari, F., Aifantis, I., Martin, C., Fehling, H.J., Hoeflinger, S., Leder, P., von Boehmer, H., and Reizis, B. (2002). Tracing lymphopoiesis with the aid of a pTalpha-controlled reporter gene. Nat Immunol 3, 489-496. • Greer, E.L., and Shi, Y. Histone methylation: a dynamic mark in health, disease and inheritance. Nat Rev Genet 13 , 343-357. • Grewal, S.I., and Jia, S. (2007). Heterochromatin revisited. Nat Rev Genet 8, 35-46. • Haberland, M., Montgomery, R.L., and Olson, E.N. (2009). The many roles of histone deacetylases in development and physiology: implications for disease and therapy. Nat Rev Genet 10 , 32-42. • Hahm, K., Ernst, P., Lo, K., Kim, G.S., Turck, C., and Smale, S.T. (1994). The lymphoid transcription factor LyF-1 is encoded by specific, alternatively spliced mRNAs derived from the Ikaros gene. Mol Cell Biol 14 , 7111-7123. • Hahn, S. (2004). Structure and mechanism of the RNA polymerase II transcription machinery. Nat Struct Mol Biol 11 , 394-403. • Hanahan, D., and Weinberg, R.A. (2000). The hallmarks of cancer. Cell 100 , 57-70. • Hardy, R.R., Carmack, C.E., Shinton, S.A., Kemp, J.D., and Hayakawa, K. (1991). Resolution and characterization of pro-B and pre-pro-B cell stages in normal mouse bone marrow. J Exp Med 173 , 1213-1225. • Hardy, R.R., and Hayakawa, K. (2001). B cell development pathways. Annu Rev Immunol 19 , 595-621.

153

• Hardy, R.R., Kincade, P.W., and Dorshkind, K. (2007). The protean nature of cells in the B lymphocyte lineage. Immunity 26 , 703-714. • Harker, N., Naito, T., Cortes, M., Hostert, A., Hirschberg, S., Tolaini, M., Roderick, K., Georgopoulos, K., and Kioussis, D. (2002). The CD8alpha gene locus is regulated by the Ikaros family of proteins. Mol Cell 10 , 1403-1415. • Harvey, R.C., Mullighan, C.G., Chen, I.M., Wharton, W., Mikhail, F.M., Carroll, A.J., Kang, H., Liu, W., Dobbin, K.K., Smith, M.A. , et al. Rearrangement of CRLF2 is associated with mutation of JAK kinases, alteration of IKZF1, Hispanic/Latino ethnicity, and a poor outcome in pediatric B-progenitor acute lymphoblastic leukemia. Blood 115 , 5312-5321. • Hatch, C.L., and Bonner, W.M. (1990). The human histone H2A.Z gene. Sequence and regulation. J Biol Chem 265 , 15211-15218. • Hu, H., Wang, B., Borde, M., Nardone, J., Maika, S., Allred, L., Tucker, P.W., and Rao, A. (2006). Foxp1 is an essential transcriptional regulator of B cell development. Nat Immunol 7, 819-826. • Hunger, S.P. Tyrosine kinase inhibitor use in pediatric Philadelphia chromosome-positive acute lymphoblastic anemia. Hematology Am Soc Hematol Educ Program 2011 , 361-365. • Iacobucci, I., Storlazzi, C.T., Cilloni, D., Lonetti, A., Ottaviani, E., Soverini, S., Astolfi, A., Chiaretti, S., Vitale, A., Messa, F. , et al. (2009). Identification and molecular characterization of recurrent genomic deletions on 7p12 in the IKZF1 gene in a large cohort of BCR-ABL1-positive acute lymphoblastic leukemia patients: on behalf of Gruppo Italiano Malattie Ematologiche dell'Adulto Acute Leukemia Working Party (GIMEMA AL WP). Blood 114 , 2159-2167. • Ichikawa, M., Asai, T., Chiba, S., Kurokawa, M., and Ogawa, S. (2004). Runx1/AML-1 ranks as a master regulator of adult hematopoiesis. Cell Cycle 3, 722-724. • Igarashi, H., Gregory, S.C., Yokota, T., Sakaguchi, N., and Kincade, P.W. (2002). Transcription from the RAG1 locus marks the earliest lymphocyte progenitors in bone marrow. Immunity 17 , 117-130.

154

• Jacob, J., Kelsoe, G., Rajewsky, K., and Weiss, U. (1991). Intraclonal generation of antibody mutants in germinal centres. Nature 354 , 389-392. • Janeway, C.A., Jr., and Medzhitov, R. (2002). Innate immune recognition. Annu Rev Immunol 20 , 197-216. • Janeway, C.A., Jr., Ron, J., and Katz, M.E. (1987). The B cell is the initiating antigen-presenting cell in peripheral lymph nodes. J Immunol 138 , 1051-1055. • Jeffrey, P.D., Russo, A.A., Polyak, K., Gibbs, E., Hurwitz, J., Massague, J., and Pavletich, N.P. (1995). Mechanism of CDK activation revealed by the structure of a cyclinA-CDK2 complex. Nature 376 , 313-320. • Jepsen, K., Gleiberman, A.S., Shi, C., Simon, D.I., and Rosenfeld, M.G. (2008). Cooperative regulation in development by SMRT and FOXP1. Genes Dev 22 , 740-745. • Jin, C., and Felsenfeld, G. (2007). Nucleosome stability mediated by histone variants H3.3 and H2A.Z. Genes Dev 21 , 1519-1529. • John, L.B., Yoong, S., and Ward, A.C. (2009). Evolution of the Ikaros gene family: implications for the origins of adaptive immunity. J Immunol 182 , 4792-4799. • Jumaa, H., Hendriks, R.W., and Reth, M. (2005). B cell signaling and tumorigenesis. Annu Rev Immunol 23 , 415-445. • Juven-Gershon, T., and Kadonaga, J.T. Regulation of gene expression via the core promoter and the basal transcriptional machinery. Dev Biol 339 , 225-229. • Kabarowski, J.H., Zhu, K., Le, L.Q., Witte, O.N., and Xu, Y. (2001). Lysophosphatidylcholine as a ligand for the immunoregulatory receptor G2A. Science 293 , 702-705. • Kadonaga, J.T. (2004). Regulation of RNA polymerase II transcription by sequence-specific DNA binding factors. Cell 116 , 247-257. • Kastan, M.B., and Bartek, J. (2004). Cell-cycle checkpoints and cancer. Nature 432 , 316-323. • Kawamata, N., Ogawa, S., Zimmermann, M., Kato, M., Sanada, M., Hemminki, K., Yamatomo, G., Nannya, Y., Koehler, R., Flohr, T. , et al.

155

(2008). Molecular allelokaryotyping of pediatric acute lymphoblastic leukemias by high-resolution single nucleotide polymorphism oligonucleotide genomic microarray. Blood 111 , 776-784. • Kharbanda, S., Ren, R., Pandey, P., Shafman, T.D., Feller, S.M., Weichselbaum, R.R., and Kufe, D.W. (1995). Activation of the c-Abl tyrosine kinase in the stress response to DNA-damaging agents. Nature 376 , 785-788. • Kim, J., Sif, S., Jones, B., Jackson, A., Koipally, J., Heller, E., Winandy, S., Viel, A., Sawyer, A., Ikeda, T. , et al. (1999). Ikaros DNA-binding proteins direct formation of chromatin remodeling complexes in lymphocytes. Immunity 10 , 345-355. • Kline-Smith, S.L., Sandall, S., and Desai, A. (2005). Kinetochore-spindle microtubule interactions during mitosis. Curr Opin Cell Biol 17 , 35-46. • Klug, A. (1999). Zinc finger peptides for the regulation of gene expression. J Mol Biol 293 , 215-218. • Knight, A.K., and Cunningham-Rundles, C. (2006). Inflammatory and autoimmune complications of common variable immune deficiency. Autoimmun Rev 5, 156-159. • Kondo, M., Weissman, I.L., and Akashi, K. (1997). Identification of clonogenic common lymphoid progenitors in mouse bone marrow. Cell 91, 661-672. • Kops, G.J., Weaver, B.A., and Cleveland, D.W. (2005). On the road to cancer: aneuploidy and the mitotic checkpoint. Nat Rev Cancer 5, 773-785. • Kornberg, R.D. (2005). Mediator and the mechanism of transcriptional activation. Trends Biochem Sci 30, 235-239. • Krishna, S.S., Majumdar, I., and Grishin, N.V. (2003). Structural classification of zinc fingers: survey and summary. Nucleic Acids Res 31 , 532-550. • Kuiper, R.P., Schoenmakers, E.F., van Reijmersdal, S.V., Hehir-Kwa, J.Y., van Kessel, A.G., van Leeuwen, F.N., and Hoogerbrugge, P.M. (2007). High- resolution genomic profiling of childhood ALL reveals novel recurrent genetic lesions affecting pathways involved in lymphocyte differentiation and cell cycle progression. Leukemia 21 , 1258-1266. 156

• Kuiper, R.P., Waanders, E., van der Velden, V.H., van Reijmersdal, S.V., Venkatachalam, R., Scheijen, B., Sonneveld, E., van Dongen, J.J., Veerman, A.J., van Leeuwen, F.N. , et al. IKZF1 deletions predict relapse in uniformly treated pediatric precursor B-ALL. Leukemia 24 , 1258-1264. • Lai, E., Clark, K.L., Burley, S.K., and Darnell, J.E., Jr. (1993). Hepatocyte nuclear factor 3/fork head or "winged helix" proteins: a family of transcription factors of diverse biologic function. Proc Natl Acad Sci U S A 90 , 10421- 10423. • Lai, J.S., and Herr, W. (1992). Ethidium bromide provides a simple tool for identifying genuine DNA-independent protein associations. Proc Natl Acad Sci U S A 89 , 6958-6962. • Lampropoulou, V., Hoehlig, K., Roch, T., Neves, P., Calderon Gomez, E., Sweenie, C.H., Hao, Y., Freitas, A.A., Steinhoff, U., Anderton, S.M. , et al. (2008). TLR-activated B cells suppress T cell-mediated autoimmunity. J Immunol 180 , 4763-4773. • Le, L.Q., Kabarowski, J.H., Weng, Z., Satterthwaite, A.B., Harvill, E.T., Jensen, E.R., Miller, J.F., and Witte, O.N. (2001). Mice lacking the orphan G protein-coupled receptor G2A develop a late-onset autoimmune syndrome. Immunity 14 , 561-571. • Le, L.Q., Kabarowski, J.H., Wong, S., Nguyen, K., Gambhir, S.S., and Witte, O.N. (2002). Positron emission tomography imaging analysis of G2A as a negative modifier of lymphoid leukemogenesis initiated by the BCR-ABL oncogene. Cancer Cell 1, 381-391. • LeBien, T.W., and Tedder, T.F. (2008). B lymphocytes: how they develop and function. Blood 112 , 1570-1580. • Levine, A.J. (1997). p53, the cellular gatekeeper for growth and division. Cell 88 , 323-331. • Levine, M., and Tjian, R. (2003). Transcription regulation and animal diversity. Nature 424 , 147-151. • Lew, D.J., and Burke, D.J. (2003). The spindle assembly and spindle position checkpoints. Annu Rev Genet 37 , 251-282. 157

• Li, Q., Peterson, K.R., Fang, X., and Stamatoyannopoulos, G. (2002). Locus control regions. Blood 100 , 3077-3086. • Li, S., Weidenfeld, J., and Morrisey, E.E. (2004). Transcriptional and DNA binding activity of the Foxp1/2/4 family is modulated by heterotypic and homotypic protein interactions. Mol Cell Biol 24 , 809-822. • Liber, D. (2008) Mechanisms for establishing transcriptional competence of tissue-specific genes in pluripotent stem cells. PhD Thesis, Imperial College London. • Liu, P., Keller, J.R., Ortiz, M., Tessarollo, L., Rachel, R.A., Nakamura, T., Jenkins, N.A., and Copeland, N.G. (2003). Bcl11a is essential for normal lymphoid development. Nat Immunol 4, 525-532. • Lo, K., Landau, N.R., and Smale, S.T. (1991). LyF-1, a transcriptional regulator that interacts with a novel class of promoters for lymphocyte-specific genes. Mol Cell Biol 11 , 5229-5243. • Loder, F., Mutschler, B., Ray, R.J., Paige, C.J., Sideras, P., Torres, R., Lamers, M.C., and Carsetti, R. (1999). B cell development in the spleen takes place in discrete steps and is determined by the quality of B cell receptor-derived signals. J Exp Med 190 , 75-89. • Lu, M.M., Li, S., Yang, H., and Morrisey, E.E. (2002). Foxp4: a novel member of the Foxp subfamily of winged-helix genes co-expressed with Foxp1 and Foxp2 in pulmonary and gut tissues. Mech Dev 119 Suppl 1 , S197-202. • Luger, K., Mader, A.W., Richmond, R.K., Sargent, D.F., and Richmond, T.J. (1997). Crystal structure of the nucleosome core particle at 2.8 A resolution. Nature 389 , 251-260. • Lugo, T.G., Pendergast, A.M., Muller, A.J., and Witte, O.N. (1990). Tyrosine kinase activity and transformation potency of bcr-abl oncogene products. Science 247 , 1079-1082. • Ma, S., Pathak, S., Mandal, M., Trinh, L., Clark, M.R., and Lu, R. Ikaros and Aiolos inhibit pre-B-cell proliferation by directly suppressing c-Myc expression. Mol Cell Biol 30 , 4149-4158.

158

• Malumbres, M., and Barbacid, M. (2001). To cycle or not to cycle: a critical decision in cancer. Nat Rev Cancer 1, 222-231. • Malumbres, M., and Barbacid, M. (2005). Mammalian cyclin-dependent kinases. Trends Biochem Sci 30 , 630-641. • Mandel, E.M., and Grosschedl, R. Transcription control of early B cell differentiation. Curr Opin Immunol 22 , 161-167. • Martensson, I.L., Rolink, A., Melchers, F., Mundt, C., Licence, S., and Shimizu, T. (2002). The pre-B cell receptor and its role in proliferation and Ig heavy chain allelic exclusion. Semin Immunol 14 , 335-342. • Martinelli, G., Iacobucci, I., Storlazzi, C.T., Vignetti, M., Paoloni, F., Cilloni, D., Soverini, S., Vitale, A., Chiaretti, S., Cimino, G. , et al. (2009). IKZF1 (Ikaros) deletions in BCR-ABL1-positive acute lymphoblastic leukemia are associated with short disease-free survival and high rate of cumulative incidence of relapse: a GIMEMA AL WP report. J Clin Oncol 27 , 5202-5207. • Matsui, T., Segall, J., Weil, P.A., and Roeder, R.G. (1980). Multiple factors required for accurate initiation of transcription by purified RNA polymerase II. J Biol Chem 255 , 11992-11996. • Medina, K.L., Garrett, K.P., Thompson, L.F., Rossi, M.I., Payne, K.J., and Kincade, P.W. (2001). Identification of very early lymphoid precursors in bone marrow and their regulation by estrogen. Nat Immunol 2, 718-724. • Medina, K.L., Pongubala, J.M., Reddy, K.L., Lancki, D.W., Dekoter, R., Kieslinger, M., Grosschedl, R., and Singh, H. (2004). Assembling a gene regulatory network for specification of the B cell fate. Dev Cell 7, 607-617. • Medvedovic, J., Ebert, A., Tagoh, H., and Busslinger, M. Pax5: a master regulator of B cell development and leukemogenesis. Adv Immunol 111 , 179- 206. • Molnar, A., and Georgopoulos, K. (1994). The Ikaros gene encodes a family of functionally diverse zinc finger DNA-binding proteins. Mol Cell Biol 14 , 8292- 8303.

159

• Molnar, A., Wu, P., Largespada, D.A., Vortkamp, A., Scherer, S., Copeland, N.G., Jenkins, N.A., Bruns, G., and Georgopoulos, K. (1996). The Ikaros gene encodes a family of lymphocyte-restricted zinc finger DNA binding proteins, highly conserved in human and mouse. J Immunol 156 , 585-592. • Mombaerts, P., Iacomini, J., Johnson, R.S., Herrup, K., Tonegawa, S., and Papaioannou, V.E. (1992). RAG-1-deficient mice have no mature B and T lymphocytes. Cell 68 , 869-877. • Montecino-Rodriguez, E., and Dorshkind, K. B-1 B cell development in the fetus and adult. Immunity 36 , 13-21. • Moore, J.K., Magidson, V., Khodjakov, A., and Cooper, J.A. (2009). The spindle position checkpoint requires positional feedback from cytoplasmic microtubules. Curr Biol 19 , 2026-2030. • Mullighan, C., and Downing, J. (2008). Ikaros and acute leukemia. Leuk Lymphoma 49 , 847-849. • Mullighan, C.G. Single nucleotide polymorphism microarray analysis of genetic alterations in cancer. Methods Mol Biol 730 , 235-258. • Mullighan, C.G., Collins-Underwood, J.R., Phillips, L.A., Loudin, M.G., Liu, W., Zhang, J., Ma, J., Coustan-Smith, E., Harvey, R.C., Willman, C.L. , et al. (2009a). Rearrangement of CRLF2 in B-progenitor- and Down syndrome- associated acute lymphoblastic leukemia. Nat Genet 41 , 1243-1246. • Mullighan, C.G., Goorha, S., Radtke, I., Miller, C.B., Coustan-Smith, E., Dalton, J.D., Girtman, K., Mathew, S., Ma, J., Pounds, S.B. , et al. (2007). Genome-wide analysis of genetic alterations in acute lymphoblastic leukaemia. Nature 446 , 758-764. • Mullighan, C.G., Miller, C.B., Radtke, I., Phillips, L.A., Dalton, J., Ma, J., White, D., Hughes, T.P., Le Beau, M.M., Pui, C.H. , et al. (2008). BCR-ABL1 lymphoblastic leukaemia is characterized by the deletion of Ikaros. Nature 453 , 110-114. • Mullighan, C.G., Su, X., Zhang, J., Radtke, I., Phillips, L.A., Miller, C.B., Ma, J., Liu, W., Cheng, C., Schulman, B.A. , et al. (2009b). Deletion of IKZF1 and prognosis in acute lymphoblastic leukemia. N Engl J Med 360 , 470-480. 160

• Murre, C., McCaw, P.S., Vaessin, H., Caudy, M., Jan, L.Y., Jan, Y.N., Cabrera, C.V., Buskin, J.N., Hauschka, S.D., Lassar, A.B. , et al. (1989). Interactions between heterologous helix-loop-helix proteins generate complexes that bind specifically to a common DNA sequence. Cell 58 , 537-544. • Musacchio, A., and Salmon, E.D. (2007). The spindle-assembly checkpoint in space and time. Nat Rev Mol Cell Biol 8, 379-393. • Myatt, S.S., and Lam, E.W. (2007). The emerging roles of forkhead box (Fox) proteins in cancer. Nat Rev Cancer 7, 847-859. • Nagulapalli, S., Goheer, A., Pitt, L., McIntosh, L.P., and Atchison, M.L. (2002). Mechanism of e47-Pip interaction on DNA resulting in transcriptional synergy and activation of immunoglobulin germ line sterile transcripts. Mol Cell Biol 22 , 7337-7350. • Nakase, K., Ishimaru, F., Avitahl, N., Dansako, H., Matsuo, K., Fujii, K., Sezaki, N., Nakayama, H., Yano, T., Fukuda, S. , et al. (2000). Dominant negative isoform of the Ikaros gene in patients with adult B-cell acute lymphoblastic leukemia. Cancer Res 60 , 4062-4065. • Narlikar, G.J., Fan, H.Y., and Kingston, R.E. (2002). Cooperation between complexes that regulate chromatin structure and transcription. Cell 108 , 475- 487. • Nigg, E.A. (2001). Mitotic kinases as regulators of cell division and its checkpoints. Nat Rev Mol Cell Biol 2, 21-32. • Nutt, S.L., Morrison, A.M., Dorfler, P., Rolink, A., and Busslinger, M. (1998). Identification of BSAP (Pax-5) target genes in early B-cell development by loss- and gain-of-function experiments. EMBO J 17 , 2319-2333. • Nutt, S.L., and Kee, B.L. (2007). The transcriptional regulation of B cell lineage commitment. Immunity 26 , 715-725. • O'Neill, D.W., Schoetz, S.S., Lopez, R.A., Castle, M., Rabinowitz, L., Shor, E., Krawchuk, D., Goll, M.G., Renz, M., Seelig, H.P. , et al. (2000). An ikaros- containing chromatin-remodeling complex in adult-type erythroid cells. Mol Cell Biol 20 , 7572-7582.

161

• Obinata, H., and Izumi, T. (2009). G2A as a receptor for oxidized free fatty acids. Prostaglandins Other Lipid Mediat 89 , 66-72. • Oettel, S., Kober, I., and Seifart, K.H. (1998). The activity binding to the termination region of several pol III genes represents a separate entity and is distinct from a novel component enhancing U6 snRNA transcription. Nucleic Acids Res 26 , 4324-4331. • Oettinger, M.A., Schatz, D.G., Gorka, C., and Baltimore, D. (1990). RAG-1 and RAG-2, adjacent genes that synergistically activate V(D)J recombination. Science 248 , 1517-1523. • Osmers, I., Smith, S.S., Parks, B.W., Yu, S., Srivastava, R., Wohler, J.E., Barnum, S.R., and Kabarowski, J.H. (2009). Deletion of the G2A receptor fails to attenuate experimental autoimmune encephalomyelitis. J Neuroimmunol 207 , 18-23. • Ortega, S., Malumbres, M., and Barbacid, M. (2002). Cyclin D-dependent kinases, INK4 inhibitors and cancer. Biochim Biophys Acta 1602 , 73-87. • Pabo, C.O., and Sauer, R.T. (1992). Transcription factors: structural families and principles of DNA recognition. Annu Rev Biochem 61 , 1053-1095. • Pan, F., Yu, H., Dang, E.V., Barbi, J., Pan, X., Grosso, J.F., Jinasena, D., Sharma, S.M., McCadden, E.M., Getnet, D. , et al. (2009). Eos mediates Foxp3- dependent gene silencing in CD4+ regulatory T cells. Science 325 , 1142-1146. • Papaemmanuil, E., Hosking, F.J., Vijayakrishnan, J., Price, A., Olver, B., Sheridan, E., Kinsey, S.E., Lightfoot, T., Roman, E., Irving, J.A. , et al. (2009). Loci on 7p12.2, 10q21.2 and 14q11.2 are associated with risk of childhood acute lymphoblastic leukemia. Nat Genet 41 , 1006-1010. • Papathanasiou, P., Attema, J.L., Karsunky, H., Hosen, N., Sontani, Y., Hoyne, G.F., Tunningley, R., Smale, S.T., and Weissman, I.L. (2009). Self-renewal of the long-term reconstituting subset of hematopoietic stem cells is regulated by Ikaros. Stem Cells 27 , 3082-3092. • Papathanasiou, P., Perkins, A.C., Cobb, B.S., Ferrini, R., Sridharan, R., Hoyne, G.F., Nelms, K.A., Smale, S.T., and Goodnow, C.C. (2003). Widespread

162

failure of hematolymphoid differentiation caused by a recessive niche-filling allele of the Ikaros transcription factor. Immunity 19 , 131-144. • Pavletich, N.P. (1999). Mechanisms of cyclin-dependent kinase regulation: structures of Cdks, their cyclin activators, and Cip and INK4 inhibitors. J Mol Biol 287 , 821-828. • Perdomo, J., Holmes, M., Chong, B., and Crossley, M. (2000). Eos and pegasus, two members of the Ikaros family of proteins with distinct DNA binding activities. J Biol Chem 275 , 38347-38354. • Peters, J.M. (2006). The anaphase promoting complex/cyclosome: a machine designed to destroy. Nat Rev Mol Cell Biol 7, 644-656. • Pfaffl, M.W. (2001). A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 29 , e45. • Prasad, R.B., Hosking, F.J., Vijayakrishnan, J., Papaemmanuil, E., Koehler, R., Greaves, M., Sheridan, E., Gast, A., Kinsey, S.E., Lightfoot, T. , et al. Verification of the susceptibility loci on 7p12.2, 10q21.2, and 14q11.2 in precursor B-cell acute lymphoblastic leukemia of childhood. Blood 115 , 1765- 1767. • Pui, C.H. (1996). Acute leukemia in children. Curr Opin Hematol 3, 249-258. • Pui, C.H., and Evans, W.E. (2006). Treatment of acute lymphoblastic leukemia. N Engl J Med 354 , 166-178. • Pui, C.H., Relling, M.V., and Downing, J.R. (2004). Acute lymphoblastic leukemia. N Engl J Med 350 , 1535-1548. • Pui, C.H., Robison, L.L., and Look, A.T. (2008). Acute lymphoblastic leukaemia. Lancet 371 , 1030-1043. • Put, N., Deeren, D., Michaux, L., and Vandenberghe, P. FOXP1 and PAX5 are rare but recurrent translocations partners in acute lymphoblastic leukemia. Cancer Genet 204 , 462-464. • Radu, C.G., Nijagal, A., McLaughlin, J., Wang, L., and Witte, O.N. (2005). Differential proton sensitivity of related G protein-coupled receptors T cell

163

death-associated gene 8 and G2A expressed in immune cells. Proc Natl Acad Sci U S A 102 , 1632-1637. • Raff, M.C., Megson, M., Owen, J.J., and Cooper, M.D. (1976). Early production of intracellular IgM by B-lymphocyte precursors in mouse. Nature 259 , 224-226. • Rao, D.S., O'Connell, R.M., Chaudhuri, A.A., Garcia-Flores, Y., Geiger, T.L., and Baltimore, D. MicroRNA-34a perturbs B lymphocyte development by repressing the forkhead box transcription factor Foxp1. Immunity 33 , 48-59. • Roessler, S., Gyory, I., Imhof, S., Spivakov, M., Williams, R.R., Busslinger, M., Fisher, A.G., and Grosschedl, R. (2007). Distinct promoters mediate the regulation of Ebf1 gene expression by interleukin-7 and Pax5. Mol Cell Biol 27 , 579-594. • Ronni, T., Payne, K.J., Ho, S., Bradley, M.N., Dorsam, G., and Dovat, S. (2007). Human Ikaros function in activated T cells is regulated by coordinated expression of its largest isoforms. J Biol Chem 282 , 2538-2547. • Ruiz, A., Jiang, J., Kempski, H., and Brady, H.J. (2004). Overexpression of the Ikaros 6 isoform is restricted to t(4;11) acute lymphoblastic leukaemia in children and infants and has a role in B-cell survival. Br J Haematol 125 , 31- 37. • Rumfelt, L.L., Zhou, Y., Rowley, B.M., Shinton, S.A., and Hardy, R.R. (2006). Lineage specification and plasticity in CD19- early B cell precursors. J Exp Med 203 , 675-687. • Sabbattini, P., Lundgren, M., Georgiou, A., Chow, C., Warnes, G., and Dillon, N. (2001). Binding of Ikaros to the lambda5 promoter silences transcription through a mechanism that does not require heterochromatin formation. EMBO J 20 , 2812-2822. • Sagaert, X., de Paepe, P., Libbrecht, L., Vanhentenrijk, V., Verhoef, G., Thomas, J., Wlodarska, I., and De Wolf-Peeters, C. (2006). Forkhead box protein P1 expression in mucosa-associated lymphoid tissue lymphomas predicts poor prognosis and transformation to diffuse large B-cell lymphoma. J Clin Oncol 24 , 2490-2497.

164

• Sakaguchi, N., and Melchers, F. (1986). Lambda 5, a new light-chain-related locus selectively expressed in pre-B lymphocytes. Nature 324 , 579-582. • Santos, M.E., Athanasiadis, A., Leitao, A.B., DuPasquier, L., and Sucena, E. Alternative splicing and gene duplication in the evolution of the FoxP gene subfamily. Mol Biol Evol 28 , 237-247. • Sawadogo, M., and Sentenac, A. (1990). RNA polymerase B (II) and general transcription factors. Annu Rev Biochem 59 , 711-754. • Schatz, D.G., Oettinger, M.A., and Baltimore, D. (1989). The V(D)J recombination activating gene, RAG-1. Cell 59 , 1035-1048. • Schlissel, M. (2002). Allelic exclusion of immunoglobulin gene rearrangement and expression: why and how? Semin Immunol 14 , 207-212; discussion 225- 206. • Schmitt, C., Tonnelle, C., Dalloul, A., Chabannon, C., Debre, P., and Rebollo, A. (2002). Aiolos and Ikaros: regulators of lymphocyte development, homeostasis and lymphoproliferation. Apoptosis 7, 277-284. • Sekine, S., Tagami, S., and Yokoyama, S. Structural basis of transcription by bacterial and eukaryotic RNA polymerases. Curr Opin Struct Biol 22 , 110-118. • Sekiya, T., Muthurajan, U.M., Luger, K., Tulin, A.V., and Zaret, K.S. (2009). Nucleosome-binding affinity as a primary determinant of the nuclear mobility of the pioneer transcription factor FoxA. Genes Dev 23 , 804-809. • Sherr, C.J. (2001). The INK4a/ARF network in tumour suppression. Nat Rev Mol Cell Biol 2, 731-737. • Shinkai, Y., Rathbun, G., Lam, K.P., Oltz, E.M., Stewart, V., Mendelsohn, M., Charron, J., Datta, M., Young, F., Stall, A.M. , et al. (1992). RAG-2-deficient mice lack mature lymphocytes owing to inability to initiate V(D)J rearrangement. Cell 68 , 855-867. • Shu, W., Lu, M.M., Zhang, Y., Tucker, P.W., Zhou, D., and Morrisey, E.E. (2007). Foxp2 and Foxp1 cooperatively regulate lung and esophagus development. Development 134 , 1991-2000.

165

• Shu, W., Yang, H., Zhang, L., Lu, M.M., and Morrisey, E.E. (2001). Characterization of a new subfamily of winged-helix/forkhead (Fox) genes that are expressed in the lung and act as transcriptional repressors. J Biol Chem 276 , 27488-27497. • Sigvardsson, M., O'Riordan, M., and Grosschedl, R. (1997). EBF and E47 collaborate to induce expression of the endogenous immunoglobulin surrogate light chain genes. Immunity 7, 25-36. • Skorski, T. (2002). BCR/ABL regulates response to DNA damage: the role in resistance to genotoxic treatment and in genomic instability. Oncogene 21 , 8591-8604. • Smale, S.T., and Kadonaga, J.T. (2003). The RNA polymerase II core promoter. Annu Rev Biochem 72 , 449-479. • Souabni, A., Cobaleda, C., Schebesta, M., and Busslinger, M. (2002). Pax5 promotes B lymphopoiesis and blocks T cell development by repressing Notch1. Immunity 17 , 781-793. • Spangrude, G.J., Heimfeld, S., and Weissman, I.L. (1988). Purification and characterization of mouse hematopoietic stem cells. Science 241 , 58-62. • Speck, N.A., and Gilliland, D.G. (2002). Core-binding factors in haematopoiesis and leukaemia. Nat Rev Cancer 2, 502-513. • Strahl, B.D., and Allis, C.D. (2000). The language of covalent histone modifications. Nature 403 , 41-45. • Streubel, B., Vinatzer, U., Lamprecht, A., Raderer, M., and Chott, A. (2005). T(3;14)(p14.1;q32) involving IGH and FOXP1 is a novel recurrent chromosomal aberration in MALT lymphoma. Leukemia 19 , 652-658. • Sun, L., Liu, A., and Georgopoulos, K. (1996). Zinc finger-mediated protein interactions modulate Ikaros activity, a molecular control of lymphocyte development. EMBO J 15 , 5358-5369. • Tang, Q.Q., Jiang, M.S., and Lane, M.D. (1999). Repressive effect of Sp1 on the C/EBPalpha gene promoter: role in adipocyte differentiation. Mol Cell Biol 19 , 4855-4865.

166

• Thoma, F., and Koller, T. (1977). Influence of histone H1 on chromatin structure. Cell 12 , 101-107. • Thompson, E.C., Cobb, B.S., Sabbattini, P., Meixlsperger, S., Parelho, V., Liberg, D., Taylor, B., Dillon, N., Georgopoulos, K., Jumaa, H. , et al. (2007). Ikaros DNA-binding proteins as integral components of B cell developmental- stage-specific regulatory circuits. Immunity 26 , 335-344. • Tonnelle, C., Bardin, F., Maroc, C., Imbert, A.M., Campa, F., Dalloul, A., Schmitt, C., and Chabannon, C. (2001). Forced expression of the Ikaros 6 isoform in human placental blood CD34(+) cells impairs their ability to differentiate toward the B-lymphoid lineage. Blood 98 , 2673-2680. • Tonnelle, C., Dijon, M., Moreau, T., Garulli, C., Bardin, F., and Chabannon, C. (2009). Stage specific over-expression of the dominant negative Ikaros 6 reveals distinct role of Ikaros throughout human B-cell differentiation. Mol Immunol 46 , 1736-1743. • Trevino, L.R., Yang, W., French, D., Hunger, S.P., Carroll, W.L., Devidas, M., Willman, C., Neale, G., Downing, J., Raimondi, S.C., et al. (2009). Germline genomic variants associated with childhood acute lymphoblastic leukemia. Nat Genet 41 , 1001-1005. • Trinh, L.A., Ferrini, R., Cobb, B.S., Weinmann, A.S., Hahm, K., Ernst, P., Garraway, I.P., Merkenschlager, M., and Smale, S.T. (2001). Down-regulation of TDT transcription in CD4(+)CD8(+) thymocytes by Ikaros proteins in direct competition with an Ets activator. Genes Dev 15 , 1817-1832. • Uhlmann, F., Lottspeich, F., and Nasmyth, K. (1999). Sister-chromatid separation at anaphase onset is promoted by cleavage of the cohesin subunit Scc1. Nature 400 , 37-42. • Urbanek, P., Wang, Z.Q., Fetka, I., Wagner, E.F., and Busslinger, M. (1994). Complete block of early B cell differentiation and altered patterning of the posterior midbrain in mice lacking Pax5/BSAP. Cell 79 , 901-912. • Van Etten, R.A. (2002). Studying the pathogenesis of BCR-ABL+ leukemia in mice. Oncogene 21 , 8643-8651.

167

• Van Etten, R.A., Jackson, P., and Baltimore, D. (1989). The mouse type IV c- abl gene product is a nuclear protein, and activation of transforming ability is associated with cytoplasmic localization. Cell 58 , 669-678. • Vinson, C.R., Sigler, P.B., and McKnight, S.L. (1989). Scissors-grip model for DNA recognition by a family of leucine zipper proteins. Science 246 , 911-916. • Wang, B., Lin, D., Li, C., and Tucker, P. (2003). Multiple domains define the expression and regulatory properties of Foxp1 forkhead transcriptional repressors. J Biol Chem 278 , 24259-24268. • Wang, B., Weidenfeld, J., Lu, M.M., Maika, S., Kuziel, W.A., Morrisey, E.E., and Tucker, P.W. (2004). Foxp1 regulates cardiac outflow tract, endocardial cushion and myocyte proliferation and maturation. Development 131 , 4477-4487. • Wang, J.H., Nichogiannopoulou, A., Wu, L., Sun, L., Sharpe, A.H., Bigby, M., and Georgopoulos, K. (1996). Selective defects in the development of the fetal and adult lymphoid system in mice with an Ikaros null mutation. Immunity 5, 537-549. • Weigel, D., and Jackle, H. (1989). Novel homeotic genes in Drosophila melanogaster. Biochem Cell Biol 67 , 393-396. • Weng, Z., Fluckiger, A.C., Nisitani, S., Wahl, M.I., Le, L.Q., Hunter, C.A., Fernal, A.A., Le Beau, M.M., and Witte, O.N. (1998). A DNA damage and stress inducible G protein-coupled receptor blocks cells in G2/M. Proc Natl Acad Sci U S A 95 , 12334-12339. • Wijchers, P.J., Burbach, J.P., and Smidt, M.P. (2006). In control of biology: of mice, men and Foxes. Biochem J 397 , 233-246. • Winandy, S., Wu, P., and Georgopoulos, K. (1995). A dominant mutation in the Ikaros gene leads to rapid development of leukemia and lymphoma. Cell 83 , 289-299. • Wlodarska, I., Veyt, E., De Paepe, P., Vandenberghe, P., Nooijen, P., Theate, I., Michaux, L., Sagaert, X., Marynen, P., Hagemeijer, A. , et al. (2005). FOXP1, a gene highly expressed in a subset of diffuse large B-cell lymphoma, is recurrently targeted by genomic aberrations. Leukemia 19 , 1299-1305. 168

• Yagi, H., Nomura, T., Nakamura, K., Yamazaki, S., Kitawaki, T., Hori, S., Maeda, M., Onodera, M., Uchiyama, T., Fujii, S. , et al. (2004). Crucial role of FOXP3 in the development and function of human CD25+CD4+ regulatory T cells. Int Immunol 16 , 1643-1656. • Yagi, T., Hibi, S., Takanashi, M., Kano, G., Tabata, Y., Imamura, T., Inaba, T., Morimoto, A., Todo, S., and Imashuku, S. (2002). High frequency of Ikaros isoform 6 expression in acute myelomonocytic and monocytic leukemias: implications for up-regulation of the antiapoptotic protein Bcl-XL in leukemogenesis. Blood 99 , 1350-1355. • Yang, Y.L., Hung, C.C., Chen, J.S., Lin, K.H., Jou, S.T., Hsiao, C.C., Sheen, J.M., Cheng, C.N., Wu, K.H., Lin, S.R. , et al. IKZF1 deletions predict a poor prognosis in children with B-cell progenitor acute lymphoblastic leukemia: A multicenter analysis in Taiwan. Cancer Sci. • Yannoutsos, N., Barreto, V., Misulovin, Z., Gazumyan, A., Yu, W., Rajewsky, N., Peixoto, B.R., Eisenreich, T., and Nussenzweig, M.C. (2004). A cis element in the recombination activating gene locus regulates gene expression by counteracting a distant silencer. Nat Immunol 5, 443-450. • Yeoh, E.J., Ross, M.E., Shurtleff, S.A., Williams, W.K., Patel, D., Mahfouz, R., Behm, F.G., Raimondi, S.C., Relling, M.V., Patel, A. , et al. (2002). Classification, subtype discovery, and prediction of outcome in pediatric acute lymphoblastic leukemia by gene expression profiling. Cancer Cell 1, 133-143. • Zarnegar, M.A., and Rothenberg, E.V. Ikaros represses and activates PU.1 cell- type-specifically through the multifunctional Sfpi1 URE and a myeloid specific enhancer. Oncogene. • Zhang, J., Jackson, A.F., Naito, T., Dose, M., Seavitt, J., Liu, F., Heller, E.J., Kashiwagi, M., Yoshida, T., Gounari, F. , et al. Harnessing of the nucleosome- remodeling-deacetylase complex controls lymphocyte development and prevents leukemogenesis. Nat Immunol 13 , 86-94. • Zhang, Y., Li, S., Yuan, L., Tian, Y., Weidenfeld, J., Yang, J., Liu, F., Chokas, A.L., and Morrisey, E.E. Foxp1 coordinates cardiomyocyte proliferation

169

through both cell-autonomous and nonautonomous mechanisms. Genes Dev 24 , 1746-1757. • Zou, H. The sister bonding of duplicated chromosomes. Semin Cell Dev Biol 22 , 566-571.

170