PHF6 Modulates the Chromatin Landscape in B-Cell Leukemia
By
Jordan Michael Elizabeth Bartlebaugh
Bachelor of Science, Biological Sciences
University of Missouri, 2013
MASSACHUSETTS INSTITUTE
OF TECHNOLOGY
DEC 06 !02
LIBRARIES
ARCHIVES
SUBMITTED TO THE DEPARTMENT OF BIOLOGY
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
IN BIOLOGY
AT THE
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
FEBRUARY 2018
@2018 Massachusetts Institute of Technology.
All rights reserved.
Signature of Author_
Certified by_
Accepted by
Signature redacted.
I V/
Signature
Signature
redactec
Jordan M.E. Bartlebaugh
Department of Biology
September 20, 2017
Michael T. Hemann
Associate Professor of Biology
Thesis Supervisor
red acted
Amy Keating
Professor of Biology
Chair, Biology Graduate Committee
2
PHF6 Modulates the Chromatin
Landscape in B-Cell Leukemia
by Jordan M.E. Bartlebaugh
Submitted to the Department of Biology on September 20, 2017
in partial fulfillment of the requirements for the degree of
Doctor of Philosophy in Biology.
ABSTRACT
Developmental and lineage plasticity have been observed in numerous malignancies,
and have been correlated with tumor progression and drug resistance. However, little is
known about the molecular mechanisms that enable such plasticity to occur. Here, we
describe the function of the Plant Homeodomain Finger Protein 6 (PHF6) in leukemia
and define its role in regulating chromatin accessibility to lineage-specific transcription
factors. We show that loss of Phf6 in B-cell leukemia results in systematic changes in
gene expression via alteration of the chromatin landscape at the transcriptional start
sites of B- and T-cell specific factors. Additionally, Phf6KO cells show significant down-
regulation of genes involved in the development and function of normal B-cells, up-
regulation of genes involved in T-cell signaling, and give rise to mixed-lineage
lymphoma in vivo. Engagement of divergent transcriptional programs results in
phenotypic plasticity that leads to altered disease presentation in vivo, tolerance of
aberrant oncogenic signaling, and differential sensitivity to frontline and targeted
therapies. These findings suggest that active maintenance of a precise chromatin
landscape is essential for sustaining proper leukemia cell identity, and that loss of a
single factor (PHF6) can cause focal changes in chromatin accessibility and
nucleosome positioning that render cells susceptible to lineage transition.
Thesis Supervisor: Michael Hemann
Title: Associate Professor of Biology
3
4
ACKNOWLEDGEMENTS
It was a wild ride. So much gratitude for...
My family...
- for the never-ending FaceTime calls that got me through these years.
- being my "happy" when I needed it the most.
My lab family...
- Faye, Susy, Yunpeng, Luis, Simona, Yadira, Holly, Eric, Christian, Bo, Pete,
Eleanor, Rana, Nina, Emanuel.
- for making lab a place to laugh and learn.
My scientific advisors...
- Mike Hemann, Jackie Lees, and Tyler Jacks.
- for your mentorship and guidance over the years.
Faye, Susy, Luis, and Yunpeng...
- for tolerating, and maybe even enjoying, my reality TV shows.
- for having my back and always showing true sincerity, character, respect, and so
much love.
Yadira and Yunpeng...
- for the hard work and passion that got us here, as well as the many laughs and late
nights in lab.
Biograd 2013...
- Dawson, Emma, Marissa, Chris, Danielle, Helen, Abe, Dan, Simona.
- for making my first year at MIT one for the books: shotgun season, pocket balls,
landmines, flannel Fridays, and accepting "Auntie JuJu" from the very start at the
BBQ.
Faye, Jake, and Erwin...
- because some people are sent to you when you need it the most, and these
friendships gave me life.
Darts Squad...
- Faye, Jake, Erwin, and Andre.
- for making our darts nights some of my fondest memories.
My Lady Loves...
- Danielle, Faye, Susy, Emma, Helen, Marissa, Taylor, Sarah, Joanna.
- because I am constantly in awe of your strength, courage, intelligence, support,
savviness, and always leading by example.
Frank Solomon...
- for picking up the phone, answering the emails, and always going the extra mile.
VPR Office...
- for teaching me so much about myself and life.
Faye, Danielle...
5
TABLE OF CONTENTS
Abstract 3
Acknowledgements 5
Table of Contents 6
Abbreviations 10
Chapter 1 - Introduction 13
Part 1 - Lymphopoiesis and the Development of B-cell and T-cell Leukemias
THE HIERARCHY OF HEMATOPOIESIS 14
*Early Fate Decisions in Lymphopoiesis 14
>,B-cell Development 16
*T-cell Development 18
PLASTICITY IN HEMATOPOIESIS 20
>Natural Plasticity in Committed Progenitors 20
*Plasticity Upon Modulation of Transcription Factor Networks 21
*Plasticity in B-cell Leukemias 23
THE BIOLOGY OF ACUTE LEUKEMIAS 24
*B-cell Acute Lymphoblastic Leukemia (B-ALL) 24
- BCR-ABL1+ B-ALL 26
- Mouse Model of BCR-ABL1 + B-ALL 27
y T-cell Acute Lymphoblastic Leukemia (T-ALL) 28
LINEAGE INFIDELITY IN LEUKEMIA 29
oMixed-Phenotype Acute Leukemia (MPAL) 29
>CD19 CAR-T Therapy Relapsed Leukemia 30
Part 2 - The Plant Homeodomain Finger Protein 6 (PHF6)
GENE AND PROTEIN STRUCTURE 34
EXPRESSION 34
GERMLINE MUTATIONS IN PHF6 36
SOMATIC MUTATIONS IN PHF6 36
MPHF6 as a Tumor Suppressor 36
- T-cell Acute Lymphoblastic Leukemia 36
- Acute Myeloid Leukemia 37
- Other Cancer Types 37
- MicroRNAs and DNA Methylation 38
- Cooperating Mutations 39
).PHF6 as an Oncogene 39
6
- T-cell Lymphomas 39
- B-cell Acute Lymphoblastic Leukemia 40
DIFFERENCES IN MUTATION TYPE: BFLS VS. T-ALL & AML 40
FUNCTIONAL ROLES OF PHF6 41
>Homology with PHD-Containing Proteins May Hint at Possible Functions
of PHF6 41
>PHF6 Interacts with Components of the Nucleosome Remodeling and
Deacetylation Complex 42
>PPHF6 Associates with the PAF1 Transcriptional Elongation Complex 44
*PHF6 Interacts with the rDNA Transcription Factor UBTF and Suppresses
rRNA Transcription 44
*PHF6 can Bind dsDNA but Not Histones 45
*PHF6 is a Putative Phosphorylation Target 46
Part 3 - The Epigenetic Regulation of Gene Expression
GENOMIC ORGANIZATION 48
METHYLATION OF DNA 48
HISTONE POST-TRANSLATIONAL MODIFICATIONS 50
oAcetylation 50
>PMethylation 50
*Ubiquitination and Phosphorylation 52
oHistone Readers 52
PHD DOMAINS 53
*PHD Domains and Human Disease 57
>PHD Domains and Cancer 58
EPIGENETIC LANDSCAPE OF HEMATOPOIESIS AND CANCER 59
LINEAGE PROMOTION THROUGH TRANSCRIPTION FACTORS 59
*Canonical Transcription Factors 60
*Pioneer Transcription Factors 60
oTerminal Selectors 62
LINEAGE RESTRICTION 62
>Bivalent Promoters 63
>ATP-dependent Chromatin Remodeling Complexes 64
- SWI/SNF 64
- ISWI 65
- IN080 65
- Nucleosome Remodeling and Deacetylation (NuRD) Complex 65
- NuRD in Hematopoiesis 66
" NuRD in Leukemia 69
References 70
7
Chapter 2 - Results 83
PHF6 regulates phenotypic plasticity through chromatin organization within lineage-
specific genes
Abstract 84
Introduction 85
Results 87
Loss of Phf6 decreases the leukemogenic potential of B-ALL cells and results
in the development of mixed-lineage lymphoma in vivo 87
Phf6-deficient B-ALL cells have altered expression of genes involved in
lineage specificity 92
Global genomic binding profiles suggest that PHF6 does not bind in a
sequence-specific manner 96
PHF6 does not act in a transcription factor complex 96
Global genomic binding profiles suggest that PHF6 interacts with chromatin
102
Loss of PHF6 results in focal changes in chromatin accessibility 102
The most significant changes in chromatin accessibility occur at genes crucial
for lineage specification 104
PHF6 plays a vital role in nucleosome positioning 107
PHF6 is necessary for maintenance of chromatin organization at lineage-
specific genes 110
Instability in the chromatin landscape allows for aberrant lineage signaling 114
Phenotypic plasticity underlies differential responses to anti-cancer treatments
in vivo 116
Loss of Phf6 leads to downregulation of B-cell developmental genes in
multiple acute leukemias 120
Discussion 121
Appendix 1 124
The PHF6 protein has distinct domains important for lymphoid and neuronal
contexts 124
Appendix 2 126
In-depth characterization of BCR-ABL1 driven B-cell acute lymphoblastic
leukemia shows reproducible disease homing and kinetics 126
Supplemental Figures 129
Supplemental Tables 133
Acknowledgements 139
Materials and Methods 140
Plasmids and cloning 140
Antibodies 140
Cell culture 141
Drug response analysis 141
Western blotting 141
Co-immunoprecipitation (Co-IP) 141
8
Quantitative PCR (qPCR) 141
Genomic DNA isolation and sequencing 142
Animal experiments 142
Organ processing and cell preparation for flow cytometry 142
Immunostaining 142
Cell cycle analysis 143
Flow cytometry 143
Histology and immunohistochemistry 143
Determination of yH2AX levels 143
RNA-Sequencing (RNA-Seq) library preparation 144
RNA-Sequencing data analysis 144
Chromatin immunoprecipitation-Sequencing (ChIP-Seq) 145
ChIP-Sequencing data analysis 146
Chromatin immunoprecipitation-qPCR (ChIP-qPCR) 147
Chromatin accessibility by ATAC-Sequencing (ATAC-Seq) 148
Statistical analysis 148
References 149
Chapter 3 - Discussion and Future Directions 153
THE ROLE OF PHF6 IN LYMPHOCYTES 154
>Leukemic Plasticity 154
- Determining Engagement of the T-cell Program 154
- Dosage of Transcription Factors 156
- Lineage Infidelity In Vivo 156
- Testing the Lineage-Promiscuity Potential in B-ALL 158
*Discerning the Role of PHF6 in Hematopoiesis 160
- Generating Knockouts of Phf6 in Specific Cell Populations 160
- Studying the Effect of Phf6KO in Development and Hematopoiesis 162
*Interrogating the Binding Profile of PHF6 165
- Histone Arrays 165
- Validating Possible Pioneer TF or Chromatin Boundary Regulator
Behavior 166
>oHow PHF6 Exerts Transcriptional Control Through Interaction with Other
Proteins 168
- BirA Biotinylation Followed by Mass-Spectrometry 169
THE IMPORTANCE OF THE CHROMATIN LANDSCAPE IN LEUKEMIA 171
*Plasticity as an Emerging Mechanism of Resistance to Targeted Therapies
171
>oThe Contribution of the Driving Oncogene to the Degree of Cellular
Plasticity 174
CONCLUDING REMARKS 177
References 178
9
ABBREVIATIONS
Acute lymphoblastic
leukemia
Acute myeloid leukemia
Assay for transposase-
accessible chromatin
Alpha-thalassemia and
mental retardation X-linked
syndrome
B-cell acute lymphoblastic
leukemias
B-cell receptor
B6rjeson-Forssman-
Lehmann Syndrome
DLBCLs Diffuse large B-celllymphomas
DN
DNMTs
DP
Double-negative
DNA methyl-transferases
Double-positive
E
EMSAs
EMT
ePHD
EtBr
ETPs
Electrophoretic mobility shift
assays
Epithelial to mesenchymal
Elongated and atypical PHD
domains
Ethidium bromide
Early T-cell progenitors
Bone marrow EV Empty vector
G
Chimeric antigen receptor
Comparative genome
hybridization
Chromatin
immunoprecipitation
Common lymphoid
progenitor
Chronic myeloid leukemia
Common myeloid
progenitors
Co-immunoprecipitation
Cytokine release syndrome
Chromatin state regulators
Cyclophosphamide,
vincristine, doxorubicin,
dexamethasone
Droplet digital PCR
Differentially expressed
genes
GC
GMPs
GSEA
Germinal center
Granulocyte-monocyte
progenitors
Gene set enrichment
analysis
H
H3K4me3 Histone H3, lysine 4, tri-methylated
HATs
HDACs
HSCs
I
ICA
IP/MS
iPSCs
Histone acetyltransferases
Histone deacetylases
Hematopoietic stem cells
Independent component
analysis
Immunoprecipitation followed
by mass spectrometry
Induced pluripotent stem
cells
L
10
A
ALL
AML
ATAC
ATRX
B
B-ALLs
BCR
BFLS
BM
C
CAR
CGH
ChIP
CLP
CML
CMPs
co-IP
CRS
CSRs
CVAD
D
ddPCR
DEG
LMPPs
LNs
M
MBD
MDS
MEPs
miRNAs
MPALs
MPPs
N
NICD
NLS
NoLS
NSCLC
NuRD
P
PAF1
PHF6
Phf6KO
Phf6wT
POI
PRC1/2
PTMs
R
rDNA
Rme
rRNA
Lymphoid-primed multipotent
progenitors
Lymph nodes
Methyl-CpG binding domains
Myelodysplastic syndrome
Megakaryocyte-erythroid
progenitors
MicroRNAs
Mixed-phenotype acute
leukemias
Multipotent progenitors
NOTCH1 intracellular
domain
Nuclear localization signals
Nucleolar localization signals
Non-small-cell lung cancer
Nucleosome Remodeling
and Deacetylation complex
RNA Polymerase 11-
Associated Factor 1
transcription elongation
complex
Plant Homeodomain Finger
Protein 6
Phf6 knockout
Phf6 wild-type
Protein of interest
Polycomb Repressive
Complexes 1 or 2
Post-translational
modifications
S
SCLC
SEC
shPhf6
SP
T
T-ALL
TCR
TECs
TFs
TKIs
TS
TSSs
U
UTRs
Z
ZaP
Ribosomal DNA
Methylarginine
Ribosomal RNA
Small-cell lung cancer
Super elongation complex
shRNA targeting Phf6
Single positive
T-cell acute lymphoblastic
leukemias
T-cell receptors
Thymic epithelial cells
Transcription factors
Tyrosine kinase inhibitors
Tumor Suppressor
Transcriptional start sites
Untranslated regions
Zinc knuckle+atypical PHD
11
12
Chapter 1
Introduction
Part 1 - Lymphopoiesis and the Development of B-cell
and T-cell Leukemias
All of the cells in the blood are generated from a hematopoietic stem cell through
a stepwise developmental process called hematopoiesis. Simply put, hematopoiesis
can be broken down into lymphopoiesis and myelopoiesis. In this section, I will focus on
the promotion of gene expression programs (called lineage specification) and restricting
the potential to differentiate to other lineages (called commitment) of B- and T-
lymphocytes. When these tightly regulated processes go amiss, cells can undergo
transformation and give rise to hematopoietic malignancies, such as acute
lymphoblastic leukemias. Additionally, I will discuss the plasticity within these highly
ordered developmental programs.
13
THE HIERARCHY OF HEMATOPOIESIS
Hematopoiesis is an ordered developmental process that results in two major
lineages: lymphoid and myeloid. The lymphoid lineage gives rise to B-cells, T-cells, and
natural killer cells. The myeloid lineage consists of monocytes, macrophages,
erythrocytes, megakaryocytes, mast cells, and granulocytes. As differentiation
progresses, lineage restriction occurs concomitantly until a cell is committed to its final
fate. Fate commitment and lineage specification are driven by specific transcription
factors acting in a precise temporal manner, along with epigenetic modifications of
chromatin, solidifying fate decisions. The process of hematopoiesis is very similar
between humans and mice. For simplicity, this section will focus on murine
hematopoiesis. The hierarchy of differentiation is summarized in Figure 1.1. Briefly,
multipotent hematopoietic stem cells (HSCs) have the ability to self-renew or develop
into multipotent progenitors (MPPs). Commitment to the myeloid or lymphoid lineages
occurs next, with the generation of common myeloid progenitors (CMPs) or lymphoid-
primed multipotent progenitors (LMPPs). Myeloid differentiation proceeds through
megakaryocyte-erythroid progenitors (MEPs) giving rise to erythrocytes and platelets,
and granulocyte-monocyte progenitors (GMPs) giving rise to granulocytes and
macrophages. LMPPs have lost all ability to develop into erythrocytes and
megakaryocytes, yet still retain some myeloid potential. Lymphoid priming progresses
through a common lymphoid progenitor (CLP), ultimately differentiating into B-cells, T-
cells, and natural killer cells [Cedar et al. 2011]. In the next section, I will review the
transcription factors that drive these developmental stages and transcriptional
programs, focusing on B- and T-cell lineage commitment.
>-Early Fate Decisions in Lymphopoiesis
Lymphoid progenitors that retain the ability to give rise to both B- and T-cells
include LMPPs and CLPs. Expression of the transcription factors PU.1 and Ikaros
define these progenitor subsets. The earliest steps through both myeloid and lymphoid
differentiation depend on transcription networks driven by PU.1. Modulation of PU.1
expression subsequently determines whether MPPs continue along a track towards
myeloid differentiation (high PU.1 levels) or lymphoid differentiation (low PU.1 levels)
[DeKoter et al. 2000]. PU.1 activates expression of 117r during early lymphoid lineage
14
specification [DeKoter et al. 2002; Medina et al. 2004]. Ikaros suppresses target genes
involved in multipotency, stem cell self-renewal, and erythropoiesis in LMPPs, thus
coordinating the earliest restriction steps in lymphoid development [Georgopoulos et al.
1994; Ng et al. 2009]. Ikaros proteins can be either transcriptional activators or
repressors and often associate with chromatin modifying complexes, such as the SWI/
SNF and NuRD complexes [O'Neill et al. 2000; Kim et al. 1999]. These functions
include promoting early lymphopoiesis through activation of Ft3, //7 and Rag1, while
also repressing erythroid and myeloid genes (further detailed in Part 3 and Figure 1.22)
OMPPG IT
PU.1 (high) PU1 (low)
+ karos
C/EBPPNOTCH1
TCF1
GATA3
60'o
EP MkP GPE2A BCL11B
EBFI
PAX5
Erythrocytes Platelets Granulocyte NK cell
44 T
Macrophage B-cell T-cell
Figure 1.1. Depiction of the hierarchical process of hematopoiesis. Shown in red text are the major transcription
factors that drive lineage specification. Myeloid lineage cells not pictured: neutrophils, eosinophils, dendritic cells,
mast cells, basophils. Adapted from Cedar et al. 2011.
15
[Yoshida et al. 2006; Dege et al. 2004]. E2A, or Tcf3, is another important regulator in
LMPPs. E2A is responsible for priming the expression of genes involved in both B- and
T-cell development, including Rag1, Notch1, 117r, and Dntt [Dias et al. 2008]. From here,
the transcriptional regulators determining B- and T-lineage fates diverge, as I will
discuss in the following two sections.
>-B-cell Development
Mature B-cells are important components of the adaptive immune system,
responsible for producing antibodies upon antigen recognition via the B-cell receptor
(BCR). Post-natal B-cell development occurs in the bone marrow, after which mature B-
cells enter the circulation and travel to the spleen to test for self-reactivity [Nutt et al.
2007]. In mature B-cells, signals to activate proliferation are conferred through the BCR.
Before the B-cell receptor is formed, these signals emanate from IL7R, a cytokine
receptor whose signaling is necessary to activate the JAK/STAT5a pathway [Hagman et
al. 2014; Ahsberg et al. 2010]. PU.1 activates expression of 117r during early
differentiation steps and later drives the expression of Ebfl and Cd45 (B220) [DeKoter
et al. 2002; Medina et al. 2004; Kikuchi et al. 2008]. After the concerted action of Ikaros
and E2A in priming progenitor cells for lymphoid differentiation, the initial steps of
generating a BCR in B-cell progenitors begins through the combined effort of multiple
transcription factors: EBF1, FOXO1, and PAX5. E2A modulates the chromatin
landscape and induces the expression of EBF1 [Hagman 2014; Lin et al. 2010].
Together, EBF1 and E2A drive V(D)J recombination at the lgG heavy chain and /gG
light chain loci in pro-B and pre-B cells, respectively (Figure 1.3) [Bain et al. 1999].
Contributing to the recombination process are other factors including PAX5, STAT5, and
Ikaros [Schwickert et al. 2014; Heizmann et al. 2013; Ochiai et al. 2012]. These TFs,
along with FOXO1, then orchestrate a B-cell specific transcriptional program, activating
genes such as Cd19, Vpreb2, Blk, Cd79a/b, Ragi, Ezh2, and Lefi and repressing
genes such as Notch1, Tcf7, Ergi, Ets2, and Gata3 (Figure 1.2) [Lin et al. 2010]. Figure
1.3 details this process, depicting the fine-tuned expression of transcription factors. This
B-cell specific transcriptional program is then sustained through the action of PAX5.
Ebfl and Pax5 are constitutively expressed throughout committed B-cell development,
16
silenced only upon terminal differentiation of mature B-cells to antibody-secreting
plasma cells [Rothenberg 2014].
The action of EBF1 is necessary for commitment to the B-cell lineage. Murine
knockout studies show that loss of Ebfl results in complete loss of B-lymphocytes,
whereas loss of Spil (PU.1), Ikaros, or Tcf3 (E2A) have deficits in multiple lineages, not
just B-cell precursors [Hagman et al. 2014]. EBF1 and PAX5 are both implicated in
repressing the myeloid and T-cell fates [Nutt et al. 1999; Rolink et al. 1999; Cobaleda et
al. 2007 Pongubala et al. 2008]. The process of B-cell differentiation is regulated by
intertwining feedback and feedforward loops, as depicted in Figure 1.2. In summary, the
hierarchy of TFs that drive this process are PU.1, Ikaros, E2A, EBF1, and PAX5. The
combinatorial network of transcription factors is interdependent upon each other to drive
B-cell development, through activation and repression of specific target genes. Lineage
specification, described as the activation of lineage-specific transcriptional programs,
appears to be largely driven by PU., Ikaros and E2A. Lineage commitment, defined as
the repression of other fate programs, is driven by Ikaros, EBF1, and PAX5 [Nutt et al.
2007].
PU. JAK/ST 5
1L7R BCR
13-Cell PragrM
FOXO1
MYCN
Ekaros E. BF1 O'Pax5 Ragl/2 LEF1
E2A/CF3 .36BLNK
Foxo1 CD79a
Notch1, FLT3, M-CSFR RAGI/2
CD19
B220
CD4 Runx1
C/EBPa, PU.1, 1d2, TCF1
Figure 1.2. Feedback and feedforward loops promoting B-cell differentiation. Blue text represents transcription
factors necessary before commitment to the B-cell lineage. Red text represents transcription factors necessary after
lineage commitment.
17
ST-cell Development
B-cell and T-cell precursors are derived from the same progenitor population in
the bone marrow. Similar to B-cells, T-cells recognize antigens through their T-cell
receptors (TCR) and transmit signals through CD3, along with CD4 and CD8 co-
receptors, depending on whether the effector cell is a helper or cytotoxic T-cell,
respectively. In order for T-cell development to proceed, lymphoid progenitors must
travel to the thymus, and upon doing so, lose their dependency on IL7 and the capacity
to develop into B-lineage cells [Allman et al. 2003]. Interactions with thymic epithelial
cells (TECs) activates NOTCH1 through juxtacrine signaling with Delta-class Notch
ligands, specifically DLL4. This signaling is required throughout T-cell development and
is unique in its extrinsic dependence on the thymic epithelium, needing constant
interaction to sustain signaling in a cell non-autonomous manner [Reizis et al. 2002;
Maillard et al. 2006; Georgescu et al. 2008]. Lymphoid progenitors that seed to the
thymus are considered early T-cell progenitors (ETPs). ETPs proceed through many
developmental stages (DN1, DN2a, DN2b, DN3, DN4, DP) until finally giving rise to a
single positive CD4 or CD8 T-cell. DN denotes stages that are "double-negative" in
which progenitors express neither CD4 nor CD8. DP denotes an advanced
developmental stage in which cells express both receptors until they undergo restriction
to a single positive fate [Thompson et al. 2011].
Unlike B-cell development where a series of defined master transcription factors
drive the transcriptional program through feedforward loops, T-cell development is
driven mainly by NOTCH1 activation of numerous target genes and the action of three
main transcription factors: TCF1 (Tcf7), GATA3, and BCL11B [Verbeek et al. 1995;
Oosterwegel et al. 1991; Hattori et al. 1996; Li et al. 2010; Albu et al. 2007]. As seen in
Figure 1.3, the transcriptional progression towards T-cell lineage commitment consists
of the gradual activation of genes such as Tcfl2 (HEB), Hesi, Gfil, Lefi, Cd3, Etsl,
and Ets2 with simultaneous loss of B-cell and myeloid lineage TF expression (Spil,
Cebpa, Meisl) [Rothenberg 2014]. Shortly after leaving the bone marrow and entering
the thymus, T-cell progenitors lose the ability to give rise to B-cells. NOTCH1 signaling
inhibits B-cell genes, like Ebfl and Pax5, thereby repressing the B-cell fate. Myeloid and
dendritic cell repression occurs later during the DN2b stage, marking T-cell commitment
18
Stable Ikaros, E2A
Down- PU.1, Kit, Bcllla, Erg
regulated C/EPBa, Meis1, Lmo2, Flt3
Etsl, Ets2, Notch3
Up- HEB (TCF12), Gfil, LEF1, CD3regulated
Notchi, GATA3, TCF1, Bclllb
VDJO VDJa
CD4
Thyus ETP DNft DNRb DNft DP CDl'
x Treg
Commitment Pre-TCR Positivecheckpoint selection
Commitment
VDJ-heavy VDJ-light
H90 MPP LMPP CLP Pro pr- n- .m-t-
lkaros a d
PU.1
IL7R
E2A
STAT5
EBF1
FOX01
PAX5
BCR signaling
Figure 1.3. B-cell and T-cell developmental programs driven by lineage-specific transcription factors. Colored
bars indicate the duration of action of indicated TFs. Adapted from Rothenberg 2014 and Georgescu et al. 2008.
and the initiation of TCR loci rearrangement [Rothenberg 2011]. Similar to B-cells, T-
cells must also express Ragl/Rag2 in order to achieve successful gene rearrangement
at the TCRP/y/6 loci to produce a functional TCR. E2A proteins drive the expression of
Ragl/2 genes in both the B- and T-cell contexts, and the expression of E2A (Tcf3) is
very important in both B- and T-cells, albeit found at much higher levels in B-cell
precursors. The feedback and feedforward transcriptional networks that define T-cell
differentiation are simpler than B-cell development and are depicted in Figure 1.4.
19
Gata T-Cell PogrMM
CD3
HEB
E2A Etsl
BC111b * - Notch1 TCF1 Ets2
IF PU.1,
ld2 C/EBPa/
lkaros, E2A, EBF1, PAX5
Figure 1.4. Feedback and feedforward loops promoting T-cell differentiation. Blue text represents transcription
factors necessary before commitment to the B-cell lineage. Red text represents transcription factors necessary after
lineage commitment.
PLASTICITY IN HEMATOPOIESIS
The ability of one cell type to acquire characteristics of another differentiated cell
type - such as phenotype, cell surface marker expression, functionality, or
transcriptional program engagement - is termed cellular plasticity. Hematopoietic stem
cells and their progenitors show a diverse capacity for cellular plasticity. HSCs are able
to differentiate into many non-hematopoietic lineages under distinct culture conditions,
including skeletal muscle, heart, liver, brain, and epithelial tissues [Graf 2002]. As
differentiation proceeds, hematopoietic progenitors lose the ability to give rise to non-
hematopoietic tissue types, yet retain fluidity within the blood lineages. In this section, I
will focus on the cellular plasticity observed specifically in lymphoid cells, both
committed and progenitor cells, detailing the reprogramming of wild-type, genetically
altered, and transformed cell types.
>Natural Plasticity in Committed Progenitors
The canonical hierarchy of hematopoiesis with two distinct and concrete arms
separating the myeloid and lymphoid lineages is misleading. Instead, fluidity between
the lymphoid and myeloid lineages occurs well into lymphocyte development, with both
committed B- and T-cell progenitors able to give rise to myeloid cells under specific
20
conditions. For example, both common lymphoid progenitors and pro-B progenitors can
give rise to functional macrophages that express CD11b and exhibit phagocytic activity
simply upon culture with specific cytokines: IL-3, IL-6, IL-2, SCF, and GM-CSF
[Montecino-Rodriguez et al. 2001; Hsu et al. 2006; Kondo et al. 2000]. Further, the T-
cell lineage also shows the capacity to develop into the myeloid lineage. Both early and
late T-cell progenitors can give rise to myeloid lineage cells either in vitro under the
correct culture conditions or in vivo through reconstitution experiments [Bell &
Bhandoola 2008; Wada et al. 2008; Lee et al. 2001]. However, B- and T-cell progenitors
cannot be coerced to differentiate into the opposite lymphoid lineage through the
mechanisms described above. Instead, the cellular conversion capacity needs to be
amplified by modulation of transcription networks, as I will describe in the following
section (and summarized in Figure 1.5).
*Plasticity Upon Modulation of Transcription Factor Networks
The previous section reviewed the natural capacity of both B- and T-lymphocytes
to differentiate into cells of the myeloid lineage, often through extrinsic cytokine cues.
Conversion between the B- and T-cell lineages, however, requires intrinsic rewiring.
Such reprogramming events occur upon the forced expression or genetic ablation of
lineage-specific transcription factors in progenitor or committed cells. Loss of Pax5 may
impart the greatest amount of cellular plasticity in B-cell precursors, such that Pax5-/-
pro-B cells can differentiate into functional macrophages, osteoclasts, dendritic cells,
granulocytes, and natural killer cells [Nutt et al. 1999]. Pax5-/- and Ebfl-/- pro-B cells
are able to reconstitute the T-cell lineage in vivo and upon culture with OP9-DL1 stromal
cells in vitro. Interestingly, recent studies show that modulation of transcription factor
levels below a certain threshold may be enough to promote cellular plasticity. Pro-B
cells that have trans-heterozygous deletions in both Pax5 and Ebfl (Pax5+/-Ebf1+/-)
can differentiate into cells of the T-lineage [Ungerback et al. 2015; Somasundaram et al.
2016]. Despite multiple examples of B- to T-lineage conversions, T-cell precursors
undergo switches to the B-cell lineage in only a limited number of circumstances. These
include the overexpression of 117r, Stat5a, or the genetic ablation of Gata3, in ETPs and
pro-T cells, enabling differentiation to the B-lineage (Figure 1.5) [Goetz et al. 2005;
Scripture-Adams et al. 2014].
21
CLPs & pro-B cells
pre-B cell lines l
CD45R-/CD19+ B-cell
precursor
Pax5./- pro-B-cells
lkzf1-/- pro-B cells
Ebf1-/- pro-B
M ature Ab-producing B cell
Pax5-/- mature B cell
Myeloid/
Macrophage
Culture with IL-2 and GM-CSF
O/E of C/EBPa or C/EBPP
In vivo reconstitution
Hsu et al 2006
Kondo et al. 2000
Xie at al. 2004
v-fms Infection, CSF-1 Borzillo et al. 1990
Culture with IL-3, IL-6, SCF, and Montecino-Rodriguez et al.
GM-CSF 2001
Culture with M-CSF or stromal Nutt et al. 1999
ST2 cells Mikkola et al. 2002
Culture with M-CSF on S17 Reynaud at al. 2008stromal cellsRenueta.20
Culture with Flt3L, SCF, M-CSF Pongubala et al. 2008
and GM-CSF
O/E of C/EBPa or C/EBPf Xedt al. 2007
O/E of C/EBPa or C/EBPP Xie et al. 2004
In vvo econtitbonRolink et al 1999
Pax5-/- pro-B In VIVO reconstitution oa et al. 2002
Culture with OP9-DL1 cells Hbflinger et al. 2004
Culture with OP9-DL1 cells Pongubala et al. 2008
Ebfl-/- pro-B T-lneage In vivo reconstitution Nechanitzky et al. 2013
Pax5+/-, Ebfl+/- pro B cells O/E of intracellular Notchi UngerbAck et al 2015Culture with OP9-DL1 cells Somasundaram et al. 2016
Pax5-/- mature B cell In vivo reconstitution Cobaleda at al. 2007
Myeloid/
Macrophage
Culture with stromal OP9 cells
In vivo thymus
Bell et al. 2008
Culture with stromal PA6 cells I Wada et al. 2008
In vivo reconstitution W
Culture with stromal TFGD cells Lee et al. 2001
or M-CSF, IL-6, and IL-7
O/E of C/EBPa, C/EBPp, or PU.1 Laiosa at al. 2006
ETP O/E of IL7R or Stal:5 Goetz at al. 2005
Gata3-/- pro-T & DN2 cells - Culture with OP9 cells Scripture-Adams et al. 2014
Figure 1.5. Lineage plastiCity in lymphoid cells. Red text denotes natural capability of lymphoid
progenitors to undergo lineage conversion. O/E, overexpression. Ab, antibody.
Similarly, modulation of transcription factors also allows conversion to myeloid
cells. Pro-B cells that are null for Ikzfl, Ebfl, or Pax5 undergo efficient reprogramming
to the myeloid lineage through culturing in conditions supportive of myeloid cell growth.
Fully-differentiated, antibody producing B-cells may also transdifferentiate to
macrophages via forced expression of the myeloid transcription factors C/EBPa or C/
EBPP [Xie et al. 2004; Nutt et al. 1999; Mikkola et al. 2002; Reynaud et al. 2008;
Pongubala et al. 2008; Cobaleda et al. 2007]. Similarly, pre-T cells also undergo such a
switch upon overexpression of C/EBPa, C/EBPP, or PU.1 (Figure 1.5) [Laiosa et al.
2006].
22
I I
.=I
DN1 and DN2 precursors
T-Uineage pro-T
DN3 p re-T cells
*Plasticity in B-cell Leukemias
A prime example of cellular plasticity in somatic cells is the generation of induced
pluripotent stem cells (iPSCs) from fully differentiated fibroblasts through the
overexpression of four defined transcription factors (OSKM): Oct4, Sox2, KIf4, and Myc
[Graf et al. 2009]. In malignant cells, plasticity is not uncommon, especially in solid
tumors that undergo the epithelial-to-mesenchymal (EMT) transition. B-cell leukemias,
in particular, have a surprising degree of plasticity [Somasundaram et al. 2017]. In
humans, the ability of B-cell leukemias to differentiate into non-malignant macrophages
is being investigated as a potential therapeutic strategy. Along these lines, human B-cell
leukemia and lymphoma cell lines can undergo robust lineage switching by forced
expression of C/EBPa or PU.1. Further, BCR-ABLI+ B-ALLs appear to have enhanced
conversion capability simply by culturing in myeloid-favorable conditions [McClellan et
al. 2015; Rapino et al. 2013]. B-cell leukemias that arise from Pax5+/-Ebfl+- mice can
transdifferentiate into non-leukemic macrophages or T-cell leukemias by overexpression
of C/EBPa and C/EBPP, or intracellular NOTCH1, respectively [Somasundaram et al.
2016]. The information presented here is summarized in Figure 1.6. Perturbation of
transcription factor networks allows cells to obtain a more plastic cellular state,
seemingly beneficial for cancer cells. In the next section, I will detail the mechanisms of
lymphoid tumorigenesis, often due to dysregulation of key lineage-specific transcription
factors and transcriptional programs.
Cel Tpe Developmental Stage Reprogrammed Ty pe of Manipulation Human or Reference
Pax5+/-, Ebf1+/- B-ALL O/E of C/EBPa or C/EBPP Mouse Somnasundaram et
BCR-ABL1+ B-ALLs Nonleukemnic Culture with IL-3, M-CSF, and McClellan et al. 2015
Macrophages GM-CSF
-Human
>20 B-lineage leukemia O/E of C/EBPa Rapino et al. 2013
B-lineage and lymphoma cell lines
Pre-B lymphoma cell line 5-azacytidine treatment Boyd et al. 1982
Macrophage-
Eu-myc B-lymphoma Like v-Raf Infection Mouse Kinken et al. 1988linesMos
Pax5+/-, Ebfl+/- B-ALL T-ALL O/E of intracellular Notchi Somasundaram et
I_ I Culture with OP9-DL1 cells al. 2016
Figure 1.6. Lineage plasticity in B-cell malignancies. O/E, overexpression.
23
THE BIOLOGY OF ACUTE LEUKEMIAS
Acute lymphoblastic leukemia (ALL) is a clonal disease of rapidly proliferating
immature lymphoid precursor cells, most often of B-cell lineage (85%) versus T-cell
lineage (15%). ALL is a heterogeneous disease that occurs in both children and adults,
and the frequencies of the genetic subtypes vary with age, as seen in Figure 1.7A-B.
Although each subtype has a unique mutational landscape, the general mechanisms
that give rise to leukemic disease are the same. This includes activation of proliferation
pathways through mutation or chromosomal translocations, a block in differentiation
through loss or gain of differentiation-specific transcription factors, an increase in self-
renewal capacity and resistance to cell death (Figure 1.7C) [Pui et al. 2004]. Here I will
briefly review B-cell acute lymphoblastic leukemias (B-ALLs), T-cell acute lymphoblastic
leukemias (T-ALLs), and mixed-phenotype acute leukemias (MPALs), focusing
specifically on BCR-ABLI+ driven malignancies.
-B-cell Acute Lymphoblastic Leukemia (B-ALL)
The B-cell lineage is the most common lymphoid subtype that undergoes
leukemic transformation, and precursor B-cell acute lymphoblastic leukemia (B-ALL) is
the most common childhood cancer. The 5-year overall survival rate is upwards of 80%
for pediatric cases, but less than 45% for adult ALLs [Liu et al. 2016; Pui et al. 2004].
Subtypes of B-ALL are classified based on their driving oncogenes. The most common
classifications include: hyperdiploid, BCR-ABL1+, BCR-ABL1-like, ETV6-RUNXI (TEL-
AMLI), TCF3-PBXI (E2A-PBX1), and hypodiploid. Recent sequencing efforts have also
identified a new subset involving ZNF384 fusions that have transcriptomic profiles with
increased expression of transcription factors important in the T-cell and myeloid
contexts (GA TA3, CEBPa and CEBP#) [Liu et al. 2016]. The frequency of these B-ALL
subtypes and other mutations are detailed in Figure 1.7.
Forty percent of B-ALL cases have mutations in genes that are involved in B-cell
development (PAX5, IKZF1, EBFI), highlighting the fact that this disease is heavily
caused by disruptions in differentiation pathways. Additionally, cooperative mutations
commonly disrupt the p53 and RB tumor suppressor pathways (CDKN2A/2B), and Ras
and Janus kinase pathways (FLT3, KRAS, NRAS, JAK1/2) [Zhang et al. 2011b; Liu et
al. 2016; Inaba et al. 2013; Mullighan et al. 2007; Mullighan et al. 2012; Hunger &
24
B
Childhood ALL subtVDes
ETP 2
LYL1 1.5
TLX1 (HOX11) 0.7
MLL-ENL 0.3
Hyperdiploid 25
TEL-AML1 (ETV6-RUNX1) 22
BCR-ABLI-like 15.3
MLL rearrangments 8
E2A-PBX (TCF3-PBX) 5
BCR-ABL 3
MYC 2
Hypodiploid <1
A
C
Block In Uncontrolled
Progenitor Coll Differentiation Prolferation Resistance to Full-Blown
(B-cell, T-cell, Increased Capacity (BCR-ABLI+, Death Signals Leukemic
CLP) f MLL-rearranged, (CDKN2AI2B) Disease(PAX5, IKZF1, ETVO-RUNXI)EBFI)
0
Figure 1.7. Tables depicting the major subtypes and frequencies of (A) childhood and (B) adult acute
lymphoblastic leukemias [Pui et al. 2004; Roberts et al. 2014; Inaba et al. 2013; Liu et al. 2016; Mullighan et al.
2012]. (C) General mechanisms of leukemic transformation.
25
I
T-ALL
B-ALL
| Adult ALL subtypes
TAL1 12
HOX11 8
T-ALL LYL1 2.5
HOX11L2 I
MLL-ENL 0.5
BCR-ABL 25
BCR-ABLI-like 10
MLL rearrangments 10
Hyperdiploid 7
B-ALL E2A-PBX 3(TCF3-PBX)
MYC 4
TEL-AMLi (ETV6-RUNX1) 2
Hypodiploid 2
Mullighan, 2015]. BCR-ABL1 translocation-driven B-ALL will be discussed in detail
below. The other most common translocations, TEL-AML1, E2A-PBXI, and MLL-
rearrangements, all function by interfering with HOX gene expression or activity to drive
leukemogenesis [Pui et al. 2004]. The frequencies of these B-ALL mutations are
detailed in Figure 1.8A.
- BCR-ABL1+ B-ALL
The Philadelphia chromosome, or BCR-ABLI translocation, was discovered by
Nowell and Hungerford in 1960 in chronic myeloid leukemia (CML) patients. It wasn't
until 1973 that Rowley identified the Philadelphia chromosome as a reciprocal
translocation between chromosomes 9 and 22 [Nowell et al. 1960; Rowley 1973]. This
oncogenic fusion protein results in constitutive tyrosine kinase activity. BCR-ABLI+
ALLs constitute 25-30% of adult ALL cases and this translocation is associated with very
poor prognoses in children [Chalandon et al. 2015]. Two isoforms of the fusion protein
exist, p190 and p210, differing in the location of the breakpoint cluster within the BCR
gene. The p190 and p210 BCR-ABL proteins are often referred to as the minor and
major subtypes, respectively. Acute leukemias often harbor the p190 isoform, whereas
the p210 isoform is found in 90% of CMLs [Chalandon et al. 2015]. Expression of BCR-
ABLI in hematopoietic stem cells alone is enough to cause a CML-like disease, yet
additional lesions are necessary to develop full-blown acute lymphoblastic leukemia.
These common cooperating mutations include loss of IKZF1 (84%), PAX5 (51%), and
CDKN2A/B (53.5%) [Liu et al. 2016; Mullighan et al. 2007].
Before the advent of targeted tyrosine kinase inhibitors (TKIs), BCR-ABLI B-ALL
patients had dismal prognoses, with survival rates less than 20%. The discovery of
imatinib changed the course of treatment for this disease, with remission rates
approaching 60%. Current treatment strategies for B-cell leukemias include standard
multi-drug chemotherapy regimens, specifically hyperCVAD (cyclophosphamide,
vincristine, doxorubicin, dexamethasone). Often, targeted TKIs, like imatinib, will be
added to chemotherapy treatment courses to increase on-target cellular killing and
reduce chemotherapy-related toxicity [Chalandon et al. 2015]. Despite childhood B-ALL
cases having relatively good prognoses, pediatric BCR-ABLI+ cases, along with MLL-
AF4 driven tumors, have significantly worse survival rates, compared to hyperdiploid,
26
E2A-PBXI, and TEL-A-MLI cases [Pui et al. 2008]. Disease relapse continues to be an
issue, with cancer cells acquiring mutations within BCR-ABLI itself as a common
mechanism of resistance. To target these resistant cells, second- and third-generation
BCR-ABLI inhibitors were designed to inhibit the evolved, resistant subclones, namely
dasatinib, nilotinib, and ponatinib. One mutation in particular, T3151, confers resistance
to first and second generation inhibitors (imatinib, dasatinib, nilotinib). The third-
generation inhibitor, ponatinib, was designed to overcome this gatekeeper mutation
[Reddy et al. 2012].
It should be noted that the BCR-ABLI translocation is not exclusive to B-ALL and
CML malignancies. On rare occasion, it can be found in acute myeloid leukemias
(<3%), B-cell lymphomas, and myelomas [Soupir et al. 2007; Martiat et al. 1990].
Additionally, it is more common in a subset of leukemias called mixed-phenotype acute
leukemias (MALPs), which will be discussed in greater detail later [Chan et al. 2017].
- Mouse Model of BCR-ABLI+ B-ALL
In order to study B-cell ALL in a manner that recapitulates the human disease,
the Hemann Lab utilizes a transplantable mouse model of BCR-ABLI+ acute
lymphoblastic leukemia developed by Williams and Sherr [Williams et al. 2006; Williams
et al. 2007]. Briefly, the bone marrow from the long bones of C57BL/6 Arf-- mice was
isolated and retrovirally transduced with a vector expressing the human p190 isoform of
the BCR-ABL1 oncogene. These cells were then grown for 7-8 days on IL-7 producing
stromal feeder cells, favoring the outgrowth of pre-B cells. Subsequently, the population
was found to be B220+, CD19+, CD24+, BP-1+, IgM-, Scal-, cKit-, Gr-, and had a
doubling time of 18 hours [Williams et al. 2006]. Proliferation in an IL-7-independent
manner is the result of BCR-ABL1 oncogenic signaling, whereas inactivation of Arf is
necessary to evade oncogene-induced apoptosis [Randle et al. 2001; McLaughlin et al.
1987]. Transplant of these cells into immunocompetent syngeneic recipient mice leads
to a highly reproducible leukemic disease that closely resembles the human pathology.
As few as twenty cells can be transplanted and still give rise to aggressive leukemia
within three weeks, suggesting that within a population, the majority of cells have
leukemia initiating potential [Williams et al. 2007].
27
In addition to allowing for the study of leukemia in the context of a normal
microenvironment and a healthy, functioning immune system, this model is also
amenable for screening purposes. It allows for the genetic modification of tumor cells ex
vivo to assess the importance of specific genes on disease development in vivo. The
Hemann Lab has previously demonstrated the power of this model by successfully
performing an in vivo genome-wide RNAi screen in these BCR-ABLI+ precursor B-ALL
cells. From this screen and subsequent work, we discovered that suppression of Phf6
was detrimental to B-cell leukemia growth [Meacham et al. 2015; Soto-Feliciano 2016].
A B
Subtype/Fusion
BCR-ABL 26 9-12
BCR-ABL1-like 10 5.3-15
MLL rearrangments 7.6 3.6-6
Hyper/Hypo-diploid 0-6.5 0.9-30
E2A-PBX 9.8 2-7.2(TCF3-PBX)
TEL-AML1
(ETV6-RUNX1) 0 15-25
10ZNF384 fusions
PAX5 31.7
B-cell Development IKZF1 15-80+ 68
EBF1 3
p531RB Tumor CDKN2A/2B, TP53, 54Suppressor Pathway RB1
Ras Kinase Pathway FLT3, KRAS, NRAS 50
Janus Kinase JAK1, JAK2 11Pathway II
5.7
Notch Pathway NOTCH1 >60
CDKN2A/2B 70
Cell Cycle Defect FBXW7 30
TAL1 60
LMO2 45
Oncogenic TFs
TLX1 30
TLX3 20
WT1 10
-. LEF1 10-15
Tumor Suppressive TFs BCL11B 10
RUNX1 10-20
GATA3 5
EZH2 10-15
SUZ12 10
Epigenetic Factors EED 10
PHF6 40
Figure 1.8. Recurrent mutations in (A) B-ALL and (B) T-ALL [Liu et al. 2016; Inaba et al. 2013; Mullighan et
al. 2007; Mullighan et al. 2012; Hunger et al. 2015; Zhang et al. 2011; Van Vlierberghe et al. 2012].
> T-cell Acute Lymphoblastic Leukemia (T-ALL)
T-cell acute lymphoblastic leukemia results from the uncontrolled proliferation of
T-cell precursors and is less common than B-ALL but more aggressive. Average survival
rates for childhood T-ALL reaches 85%, whereas adult disease has a worse prognosis
with only a 35% average survival rate. Also, T-ALL occurs in men more often than
women [Peirs et al. 2015]. Activating mutations in NOTCHI and deletion of p16NK4A and
p14ARF genes are found in greater than 50% and 70% of T-cell acute lymphoblastic
28
leukemia cases, respectively [Weng et al. 2004; Peirs et al. 2015]. Whereas B-ALLs
often lose the expression of B-cell specific transcription factors in order to drive
leukemogenesis, T-ALLs acquire translocations that activate the expression of T-cell TF
oncogenes. T-ALLs can be divided into subgroups based off of these TFs including
LYL1, HOXA, TLXI, TLX3, NKX2-1, NKX2-2, and TALI [Van Vlierberghe et Ia. 2012].
Mutations and their respective frequencies are detailed in Figure 1.7 and Figure 1.8B. In
addition to activating the Notch pathway, T-cell leukemias commonly acquire mutations
that disrupt pathways involved in the cell cycle and epigenetic control. T-ALLs are also
treated with aggressive chemotherapy courses [Van Vlierberghe et al. 2010].
LINEAGE INFIDELITY IN LEUKEMIA
As detailed previously, hematopoietic cells retain varying degrees of plasticity in
their genomes, decreasing in potential as differentiation progresses. In this section, I will
detail the plasticity and promiscuity of leukemias, focusing on a subtype of disease that
characteristically expresses receptors of multiple lineages (mixed-phenotype acute
leukemia) and lineage switching that occurs upon treatment with targeted therapy.
>Mixed-Phenotype Acute Leukemia (MPAL)
Mixed-phenotype acute leukemias, or MPALs, are malignancies that express
markers of multiple lineages, have very poor prognoses, and are often driven by BCR-
ABLI or MLL rearrangements. Previously, they were described as mixed lineage
leukemias, hybrid acute leukemias, bi-lineal leukemias, or bi-phenotypic leukemias.
Now they are collectively classified as mixed-phenotype leukemias, modified in 2008 by
the WHO classification system for hematopoietic and lymphoid tumors [Weinberg et al.
2010; Borowitz 2014; Borowitz et al. 2008; Yan et al. 2012] (see Figure 1.9A). It should
be noted that bi-phenotypic and bi-lineal leukemias describe two different phenomena.
Bi-phenotypic refers to tumors that express immunophenotypic features of multiple
lineages within the same cell, whereas bi-lineal describes a disease with two
populations of cells, or possible subclones, that individually express different lineage
markers, perhaps evolving from the same leukemia initiating cell [Wolach et al. 2015].
MPALs are rare, constituting only 2-5% of leukemias [Weinberg et al. 2010; Eckstein et
al. 2016]. Commonly, they express surface markers characteristic of the myeloid lineage
and either B- or T-lineages. MPALs that co-express markers of both the B- and T-cell
29
lineages are infrequent. The BCR-ABL1 oncoprotein drives 20% of MPALs [Chan et al.
2017]. The landscape of MPAL mutations is broad, with no single mutation common to
all cases. To demonstrate, the mutations with highest frequency in MPALs are IKZF1,
EZH2, ASXLI, ETV6, CDKN2A, NOTCHI, and TET2 ranging in mutational frequency
from 3% to 13% of cases [Yan et al. 2012]. One study claims that 33% of MPALs harbor
mutations in DNMT3A [Eckstein et al. 2016]. Intriguingly, PHF6 mutations are found in
8% of MPALs, specifically in malignancies co-expressing B- and T-lineage markers or T-
and myeloid markers [Eckstein et al. 2016].
A B
CD19 CAR T-cell CD19+ B-ALL
CD19 Cytoplasmic CD3 pesB22n(Myeloperoxidase) B220
CD79a Surface CD3 CD11c
Cytoplasmic CD22 CD14
CD10 CD64
Lysozyme
NSE
Mutate CD19 receptor/ Lose CD19 expression
Additional markers tested - not necessary for classIfication alternative splicing
Cytoplasmic IgM TCR (a/0 or y/6) CD117
CD10 CD2 CD13 6
CD20 CD5 CD33
TdT CTransform to alternative
CD24 CD10 CD14 CD19+ B-ALL B220+ B-ALL
TdT CD15
CD7 CD64
CD1a
CD33
Figure 1.9. (A) Guidelines for Mixed-Phenotype Acute Leukemia classification as determined by the 2008 WHO
clinical guidelines. Tumors containing multiple markers from 2 different lineages are considered MPALs. [Weinberg
et al. 20141. (B) Illustration of CD19-CAR-T relavse mechanisms in B-ALL.
*CD19 CAR-T Therapy Relapsed Leukemia
Genetically engineered T-cells expressing a chimeric antigen receptor (CAR)
specific for CD19 is a promising therapy achieving durable remissions used in B-ALL
patients that are resistant to frontline chemotherapy. However, numerous patients still
relapse, through three main mechanisms: (1) mutating the CD19 receptor; (2) losing
expression of CD19; or (3) pathological transformation to an alternative disease with
myeloid-like characteristics (see Figure 1.9B). A subset of patients treated with CD19
CAR-T targeted therapy relapse within 1 month with a disease that no longer expresses
30
CD19, but now expresses myeloid markers like CD11b, CD33, and MPO [Jacoby et al.
2016; Gardner et al. 2016]. The mechanism of resistance is still largely unknown and
warrants further investigation. As demonstrated with MPALs and the relapsed disease
that results after CAR-T therapy, cancer cell plasticity is emerging as an important
mechanism both in general tumorigenesis and within specific contexts, such as therapy
response.
31
32
Chapter 1
Introduction
Part 2 - The Plant Homeodomain Finger Protein 6
(PHF6)
As described in Part 1, B-cell and T-cell leukemias are distinct malignancies with
unique mutational profiles, disease presentations, prognoses, and treatment strategies.
Previous work from the Hemann Lab demonstrated that genes can have context-
specific and lineage-specific roles. In an effort to identify in vivo modulators of leukemia
cell growth, we identified Plant Homeodomain Finger Protein 6 (PHF6) as a gene
implicated in having lineage-specific functions: acting as a positive regulator of B-cell
leukemia growth and a negative regulator of T-cell leukemia growth [Meacham et al.
2015; Soto-Feliciano 2015]. In this thesis, I focus on further studying the role of PHF6 in
B-ALL. In this section, I will review the current knowledge surrounding PHF6 and how
disruption of this gene is implicated in a neuronal disorder and hematological
malignancies.
33
GENE AND PROTEIN STRUCTURE
Plant Homeodomain Finger Protein 6 (PHF6) is a gene located on the X
chromosome. The gene spans 90 kb and consists of 11 exons, of which the first and last
constitute the 5' and 3' untranslated regions (UTRs), respectively. The gene is
transcribed into a 4.5 kb mRNA and encodes a protein of 365 amino acids (41 kDa).
Two main isoforms exist, differing only in the splicing of the tenth intron, which results in
a longer 3'UTR (Figure 1.10A). PHF6 is highly conserved among vertebrates, with mice
and humans sharing 97.5% amino acid identity [Lower et al. 2002].
The PHF6 protein has six main structural features. It contains two elongated and
atypical PHD domains (ePHD), also referred to as ZaP motifs (Zinc knuckle+atypical
PHD). ZaP motifs are derived from a more common motif called a PZP motif (PHD
domain+zinc knuckle+atypical PHD). These domains stabilize zinc ions, usually for the
purpose of binding and reading histones. The composition of each motif is summarized
and depicted in Figure 1.10B-C. [Sanchez et al. 2011]. Additionally, the PHF6 protein
has 2 nuclear localization signals (NLS) and 2 nucleolar localization signals (NoLS).
Localization to both organelles was confirmed by immunocytochemistry, mass
spectrometry, immunofluorescence, and sub-cellular fractionation [Landais et al. 2005;
Lower et al. 2002; Voss et al. 2007; Vallee et al. 2004; Todd & Picketts 2012; Wang et
al. 2013; Zhang et al. 2013] (Figure 1.1OD).
EXPRESSION
Expression analysis has been performed in both murine and adult human
tissues. In adult human tissues, PHF6 expression is highest in the thymus, ovaries,
thyroid, and lymph nodes [Van Vlierberghe et al. 2010a; Hajjari et al. 2015] (Figure
1.10E). Additionally, Phf6 is expressed throughout murine differentiation, with highest
expression in the brain, thymus, kidney, and spleen of adult tissues [Voss et al. 2007;
Zhang et al. 2013]. Upon separation of cell populations from the murine thymus, Phf6
expression in hematopoietic cells is ubiquitous, with higher levels of expression in
lymphoid cells compared to HSCs or myeloid progenitors. Relative expression is highest
in pre-B cells and double positive (CD4+/CD8+) T-cells, with Phf6 transcript levels in pre-
B cells almost double that of any other population [Van Vlierberghe et al. 2010b].
34
Chr. X
1 2 3 4 5 6 7 8 9 10 11
B
Cys2His2 PHD domain Cys4HisICys3  -
Zinc knuckle Cys2His1Cysi Zinc knuckle Cys2His1Cys1
atypical PHD domain Cys4His1CyS2His1 atypical PHD domain Cys4His1Cys2His1
Binds UBF I I Binds RBBP4 Doesn't bind chromatin
-tpia Zinc
Knuckle
Adult Brain-Cerebellum Lymph Node Spleen
IPA
Ovary Thyroid Kidney
Figure 1.10. (A) Schematic of the PHF6 gene. Blue boxes, exons of the open reading frame (1095 bp). White
boxes, 5' and 3' UTRs. Red line represents splice site for the alternative transcript isoform. (B) Comparison of
domains that are derivative of that found in PHF6. Numbers denote the amount of amino acid residues. Cys,
cysteine. His, histidine. PHD, plant homeodomain. (C) Representation of the PHF6 protein and the domains
implicated in protein or DNA interactions. Orange, NLS. Yellow, NoLS. (D) Intracellular localization of PHF6 in the
nucleus and nucleolus of human cell lines. Top to bottom: A-431 (epidermoid carcinoma), U2OS (osteosarcoma),
U-251 MG (glioblastoma). Red, microtubules. Blue, Nucleus. Green, PHF6. Data from the Human Protein Atlas
[Uhlen et al. 2010]. (E) PHF6 protein expression analysis in indicated human tissues. Data from the Human Protein
Atlas [Uhlen et al. 20101.
35
A
C L
UZincKnuckle
D E
I
I
GERMLINE MUTATIONS IN PHF6
PHF6 was first discovered as the gene mutated in patients diagnosed with
B6rjeson-Forssman-Lehmann Syndrome (BFLS), an X-linked intellectual disability
disorder. Elucidating the gene responsible for BFLS was the focus of multiple labs from
the 1960s through the 1980s. Early linkage mapping studies by these groups narrowed
down the locus responsible for the disorder, with further refinement and gene
identification occurring in 2002 by Lower and colleagues [B6rjeson et al. 1962; Ardinger
et al. 1984; Matthews et al. 1989; Turner et al. 1989; Gedeon 1996; Lower et al. 2002].
Sequencing of the PHF6 locus in affected families showed that mutations are
predominantly missense mutations [Lower et al. 2002]. Affected males, with only one X-
chromosome, harbor germline mutations in their only copy of PHF6. Females, with two
copies of the gene, are often carriers and will experience substantial skewed X-
inactivation of the chromosome carrying the mutant PHF6 gene (>90% of cells have
silenced the mutant copy). BFLS is a rare disorder, with less than 30 unrelated familial
cases documented in the literature [Chao et al. 2010]. The disorder is characterized by
intellectual deficits, microcephaly, malformed fingers and toes, hypogonadism, short
stature, epilepsy, obesity, and gynecomastia [Gedeon et al. 1996; Voss et al. 2007].
SOMATIC MUTATIONS IN PHF6
In 2010, work from the laboratory of Adolfo Ferrando revealed that a significant
portion of T-cell acute lymphoblastic leukemias have mutations in PHF6 [Van
Vlierberghe et al. 2010a]. Since then, information from studies probing the mutational
status of PHF6 across a variety of tumor types has been steadily accumulating, with
most information having been collected for T-ALL and AML. The following information is
summarized in Figure 1.11.
>-PHF6 as a Tumor Suppressor
- T-cell Acute Lymphoblastic Leukemia
T-ALL malignancies have a skewed gender distribution with three times more
men than women diagnosed with the disease. Motivated to explain this phenomenon,
the Ferrando Lab set out to identify novel tumor suppressor genes located on the X-
chromosome. In doing so, they discovered PHF6 to be mutated in a substantial fraction
of patients. Approximately 16% of pediatric cases and 38% of adult T-ALL cases have
36
mutations within the PHF6 gene. These mutations were overwhelmingly frameshift or
truncation mutations, while any missense mutations were enriched in the second ZaP
domain of the protein. Frameshift and truncation mutations readily result in complete
loss of function of the protein, supporting a role for PHF6 as a tumor suppressor in T-
ALL [Van Vlierberghe et al. 201 Oa]. Established T-ALL cell lines that are mutant in PHF6
include DND41, HBP-ALL, SUP-T1, and T-ALL1 [Van Vlierberghe et al. 2010a; Kalender
Atak et al. 2012]. Hairpin-mediated knockdown of Phf6 is shown to be neutral or
advantageous in multiple mouse models of T-cell lymphoma, while overexpression is
detrimental to T-ALL cell growth [Meacham et al. 2015; Mavrakis et al. 2011] (Figure
1.11).
- Acute Myeloid Leukemia
After the discovery that PHF6 was mutated in a substantial number of T-ALLs,
Van Vlierberghe and colleagues then investigated the status of PHF6 in acute myeloid
leukemia (AML) tumor samples and discovered mutations in 3% of patient samples
analyzed. As seen in T-ALLs, AMLs have an enrichment of frameshift and truncation
mutations (70% of mutations documented) with most missense mutations located in the
second ZaP domain as well. Finally, hairpin-mediated knockdown of Phf6 is shown to be
especially advantageous in a murine mouse model of AML in vivo [Meacham et al.
2015] (Figure 1.11).
- Other Cancer Types
Screening of other hematological malignancies uncovered that 2.47% of chronic
myeloid leukemia (CML) patients in blast crisis have PHF6 frameshift mutations [Li et al.
2012]. Furthermore, myelodysplastic syndromes (MDS), myeloid neoplasms with
predisposition to develop AML, were found to have PHF6 deletions in 2-10% of cases
[Haferlach et al. 2013].
Since PHF6 is moderately expressed in many adult human tissue types, other
labs have investigated the mutational status of PHF6 in solid tumors and other
malignancies [Van Vlierberghe et al. 2010a; Yoo et al. 2012]. Yoo and colleagues found
that 2.6% of hepatocellular carcinoma patients harbored truncation mutations in PHF6,
demonstrating that PHF6 may also be acting as a tumor suppressor in this malignancy
[Yoo et al. 2012]. In a study aimed to identify the top mutated epigenetic regulators in 21
37
different pediatric cancers, sequencing of 633 genes across 1,000 tumors showed that
PHF6 was the second most frequently mutated gene, surpassed only by mutations in
the gene that encodes histone H3.3 itself [Huether et al. 2014] (Figure 1.11).
T-cell acute lymphoblastic leukemia
(T-ALL)
Pediatric
16%
5.4%
13%
38%
18.6%
39.5%
28%
11%
20.7%
19.4%
27.1%
Van Vlierberghe et al. 2010a
Wang et al. 2011
Grossmann et at. 2013
Huh et al. 2013
Neumann et al. 2015
Yuan et al. 2015
Vicente et al. 2015
Li et al. 2016
3% Van Vlierberghe et al. 2010Ob2Acute myeloid leukemia 3% Patel et al. 2012(AML) 2% de Rooij et al. 2016
Chronic myeloid leukemia 2.47% Li et al. 2012(CML)
Hepatocellular carcinoma 2.6% YoO et al. 2011
Myelodysplastic syndromes (MDS) 2-10% Haferlach et al. 2013
T-cell Non-Hodgkin's lymphoma 19% Renedo et al. 2001
Oncogene
(Gain of function/ Van Vlierberghe et al. 2010a
expression) B-cell acute lymphoblastic leukemia 0% Yoo et al. 2011
(B-ALL) Li et al. 2016
Meacham et al. 2015
Figure 1.11. Table summarizing the human neoplasms which harbor mutations in
of mutations, frequency of mutation, and associated referenced studies.
PHF6. Shown are the types
- MicroRNAs and DNA Methylation
Downregulation of tumor suppressor (TS) gene expression can be accomplished
through a variety of mechanisms, including through microRNAs (miRNAs) and
hypermethylation of target gene promoters. Mavrakis and colleagues discovered a
network of microRNAs controlling suppression of TS gene expression in T-ALLs,
showing that miR-20a and miR-26a specifically target PHF6. These oncogenic miRNAs,
or oncomirs, also regulate the expression of another well-known T-ALL tumor
suppressor, PTEN [Mavrakis et al. 2011]. Additionally, another study focusing
specifically on identifying novel oncomirs for PHF6 through an unbiased miRNA screen
38
Tumor Suppressor
(Loss of function/
expression)
found miR-128 to be an important and unique suppressor of PHF6 [Mets et al. 2014].
Further, Franzoni and colleagues separately identified and validated PHF6 as a target of
miR-128 in neurons, implicating the two factors in neuronal migration and outgrowth. In
line with this microRNA-mediated regulation in the brain, computational analysis of
mRNA-microRNA interaction networks in temozolomide-resistant glioblastoma cells
identified PHF6 as a potential factor of resistance through dysregulation by miR-143,
miR-93, miR-183, and miR-214 [Hiddingh et al. 2014]. These studies highlight that
microRNA mediated modulation of PHF6 expression is important in both neuronal and
lymphoid contexts [Franzoni et al. 2015; Mets et al. 2014; Mavrakis et al. 2011]. Another
mechanism of gene expression suppression is through hyper-methylation of CpG
islands. Kraszewska and colleagues found PHF6 to be hypermethylated in 10% of
pediatric T-ALL cases [Kraszewska et al. 20111.
- Cooperating Mutations
Since its identification as a novel tumor suppressor in T-ALL and AML,
considerable work has been done to further identify cooperating mutations with PHF6.
The T-ALL oncogenes, TLX1 and TLX3, are both associated with loss-of-function
mutations in PHF6 [Van Vlierberghe et al. 2010a; Vicente et al. 2015]. Additionally,
multiple groups show that NOTCH1, JAK1, IL7R, and SET-NUP214 rearrangements are
frequently associated with PHF6 mutations, with 80% of PHF6-mutant T-ALLs also
harboring mutations in NOTCHI [Wang et al. 2011; Huh et al. 2013; Yuan et al. 2015;
Vicente et al. 2015; Li et al. 2016]. Transcription factor binding experiments show that
PHF6 is a direct target of both NOTCHI and TLX1 [Palomero et al. 2006; de
Keersmaecker et al. 2010]. Conversely, a negative association between TALI/LMO2 T-
ALL subtypes and PHF6 mutations is observed [Vicente et al. 2015]. In AML, mutations
in IDHI, IDH2, ASXL1, FLT3, CEBPA, and RUNXI are often co-mutated with PHF6
[Van Vlierberghe et al. 2010b; Patel et al. 2012].
.PHF6 as an Oncogene
- T-cell Lymphomas
Early comparative genome hybridization (CGH) studies in T-cell non-Hodgkin's
lymphomas found recurrent amplifications of the X-chromosome at positions that we
now know correspond to the PHF6 locus. Approximately 19% of human T-cell
39
lymphomas were found to harbor these amplifications [Renedo et al. 2001]. Further, the
study of murine T-cell lymphomagenesis due to viral integration of the RadVNL3 virus
showed integration into the locus of putative oncogenes Phf6 and Kis2, in addition to
loci of established oncogenes such as Notch1 and c-Myc [Landais et al. 2005]. Finally,
recent meta-analysis of PHF6 genomic, transcriptomic, and proteomic status uncovered
significant overexpression of PHF6 in human lymphomas, gliomas, cervical, ovarian,
and colorectal cancers [Hajjari et al. 2015].
- B-cell Acute Lymphoblastic Leukemia
As discussed above, PHF6 is often mutated in hematological malignancies.
However, mutations in PHF6 are never found in B-ALL patients (n=148) [Van
Vlierberghe et al. 2010a; Yoo et al. 2012; Li et al. 2016]. Accordingly, an unbiased
shRNA screen performed in a murine B-cell leukemia model discovered that significant
loss of Phf6 expression impaired tumor cell growth in vivo. This study demonstrated that
knockdown also impaired in vivo disease progression of Eu-Myc B-cell lymphoma cells
[Meacham et al. 2016]. These results contrast with the mutations observed in T-ALL and
AML, in which loss of gene expression is beneficial to leukemogenesis, and therefore
highlight that PHF6 could have lineage-specific roles in normal hematopoietic cells, as
well as in malignancies.
DIFFERENCES IN MUTATION TYPE: BFLS VS. T-ALL & AML
The vast majority of mutations observed in BFLS are missense mutations, while
those observed in T-ALL and AML are frameshift or truncating mutations. This suggests
that complete loss of PHF6 protein is necessary for leukemogenesis in T-ALL and AML.
Likewise, some protein function appears to be maintained in BFLS patients, accounting
for the absence of any recorded hematopoietic deficits [Van Vlierberghe et al. 2010a].
Large deletions have recently been documented in female BFLS carriers, suggesting
that these types of mutations are lethal to male fetal development when there is only
one copy of the X-chromosome [Berland et al. 2011]. Consistent with this, the missense
mutations observed in T-ALL patients, despite being small nucleotide substitutions, still
result in no detectable protein product if they are located outside of the second ZaP
domain. Missense mutations within the second ZaP domain of PHF6 still have protein
expression, suggesting that these mutations are enough to interfere with lymphoid
40
tumor suppressor function [Van Vlierberghe et al. 2010a]. However, these point
mutations in T-ALL and AML are often in the zinc-coordinating residues or buried
hydrophobic residues, and have been shown to disrupt protein stability. BFLS point
mutations were not shown to affect protein structure and are therefore hypothesized to
disrupt protein-protein interactions [Liu et al. 2014].
BFLS patients carry germline mutations PHF6, yet are not predisposed to
developing leukemia. Only 2 male BFLS patients have been diagnosed with
hematopoietic malignancies. A 16-year-old male with BFLS was diagnosed with
Hodgkins Lymphoma. The mutation in PHF6 affected exon 8, resulting in a R257G
missense mutation [Carter et al. 2009]. Additionally, a 9-year-old child with BFLS
developed T-ALL. The truncation mutation was located close to the C-terminal end of
the protein in exon 10 (R342X) [Chao et al. 2010; Holmfeldt et al. 2010]. Both mutations
affected arginines in the second ZaP domain of PHF6. The lack of any predispositions
to cancer in BFLS patients again suggests that these BFLS-specific PHF6 mutations do
not result in complete loss of protein function.
FUNCTIONAL ROLES OF PHF6
>.Homology with PHD-Containing Proteins May Hint at Possible Functions of
PHF6
Sequencing studies demonstrated that mutations in PHF6 had catastrophic
consequences leading to intellectual disabilities and leukemogenesis, suggesting that
PHF6 could have important roles in neurogenesis and hematopoiesis. Briefly, PHD
domains have documented roles in binding modified and unmodified histone H3 tails
and DNA. Proteins with PHD domains are involved in recruiting protein complexes to
chromatin, gene expression regulation, and mediating protein-protein interactions
[Sanchez et al. 2011; Todd & Picketts 2012]. A more detailed review of PHD domain
function will be discussed later in Part 3. Based on these well-characterized proteins
and functional studies, the imperfect PHD-like domains of PHF6, along with a lack of
any enzymatic domains, suggests that it could have similar roles to canonical PHD
domain-containing proteins, perhaps in chromatin recognition or recruitment of
associated proteins. Besides protein homology and a handful of intriguing studies by
multiple groups, the cellular function of PHF6 remains widely unknown. Next, I will
41
review the three documented interactions of PHF6 (NuRD complex, PAF1 complex, and
UBTF), which implicate possible roles for PHF6 in transcriptional regulation. Interactors
identified through immunoprecipitation followed by mass spectrometry analyses are
summarized in Figure 1.12.
Nulosm 40S60 0. S 0 0.
CHD4* H1.2 SNRNP200 RPLPO LYAR FKBP3 PAF1* HATI
CHD3 H2B.1 PRPF8 RPS18 FBL .uIE CDC73* DDB1
RBBP4* H2A.Z HNRNPA2B1 RPS25 NPM1 CCDC86 CTR9* VPRBP
RBBP7 H3.1 HNRNPH3 RPL22 UBTF* RCL1 LEO1*
HDAC1** HNRNPU RPL4 RPS3 GTF2B
RALY RPL30 CHDI FEN
HNRNPA3 CHD2 ANP32B
PCBP2
HNRNPC
MAGOH
Figure 1.12. Table summarizing the results of PHF6 IP/MS data. *denotes that there was confirmation of
interaction by co-IP. **was not detected in IP/MS but confirmed via co-IP. Grey boxes are results from [Todd &
Plcketts 2012]. Red boxes are results from [Zhang et al. 2013]. Green Boxes are results from [Wang et al. 2013].
Underline denotes a hit in two separate IPIMS datasets.
oPHF6 Interacts with Components of the Nucleosome Remodeling and
Deacetylation Complex
The first study investigating the function of PHF6 was performed by Todd and
Picketts [2012]. They discovered through immunoprecipitation followed by mass
spectrometry (IP/MS) analyses that PHF6 interacted with proteins of (1) the
Nucleosome Remodeling and Deacetylation (NuRD) complex, (2) histone proteins, (3)
splicing factors, and (4) ribosomal proteins (Figure 1.12). They further interrogated the
association between PHF6 and components of the NuRD complex and confirmed an
interaction between PHF6 and RBBP4, HDAC1, and CHD4 by co-immunoprecipitation.
Other components of the complex, CHD3 and RBBP7, were significant hits in the IP/MS
results but were not included in co-immunoprecipitation validation experiments. These
interactions occur solely in the nucleoplasm, demonstrating that PHF6 has separate and
distinct functions in the nucleolus. Additionally, the interaction with RBBP4 is mediated
through the region housing the nucleolar localization signal in the PHF6 protein, further
42
00. I
uwA
92)
Figure 1.13. (A) Model for interaction between PHF6 and the NuRD complex in the nucleoplasm, involved in
chromatin recognition, complex recruitment, and possible nucleosome remodeling. (B) Model for interaction
between PHF6 and the PAF1 transcriptional elongation complex in the brain, driving the expression of genes
involved in neuronal migration, like NGC/CSPG5. (C) Schematic showing the interaction between PHF6 and UBTF,
sequestering the rDNA transcription factor from driving the expression of ribosomal genes.
43
A
B
eurl
k Af-
1 IOWM3
supporting an interaction between PHF6 and NuRD exclusively in the nucleoplasm [Liu
et al. 2014; Liu et al. 2015] (Figure 1.13A). Since the NuRD complex has well-defined
roles in gene regulation during neurogenesis and hematopoiesis and can be
dysregulated in certain cancers (detailed in Part 3), this interaction between PHF6 and
NuRD has many intriguing implications for the definitive function of PHF6 in
lymphocytes [Todd & Picketts 2012].
>PHF6 Associates with the PAF1 Transcriptional Elongation Complex
PHF6 has important implications in neurons and hematopoietic cells. Zhang et al.
focused their study on the function of PHF6 in neurons and the pathogenesis of
Borjeson-Forssman-Lehmann Syndrome (BFLS). They discovered through IP/MS
analyses that PHF6 interacted with 4 out of the 5 members of the PAF1 transcriptional
elongation complex: CDC73, CTR9, LE01, and PAF1 (Figure 1.12, Figure 1.13B).
Through these interactions, they drive expression of genes important for neuronal
migration, such as NGC/CSPG5, a gene implicated in susceptibility to schizophrenia.
Knockdown of either PHF6 or PAF1 resulted in impaired neuronal migration and loss of
NGC/CSPG5 expression [Zhang et al. 2013]. While this study solidified the importance
of the PAF1 transcriptional elongation complex in neurons, this interaction has not been
confirmed in hematopoietic cells.
>-PHF6 Interacts with the rDNA Transcription Factor UBTF and Suppresses
rRNA Transcription
Intrigued by the nucleolar localization of PHF6, Wang and colleagues focused
their studies on this specific organelle. Through IP/MS analysis, they discovered that
PHF6 associates with the RNA upstream binding transcription factor, also known as
UBTF (Figure 1.12, Figure 1.13C). Interestingly, UBTF was also a significant hit in the
IP/MS results by Zhang and colleagues [Zhang et al. 2013]. UBTF drives the
transcription of ribosomal DNA (rDNA). They found that the N-terminal portion of the
protein, specifically the first ZaP domain and nucleolar localization signal, was
necessary for localization to the nucleolus and protein-protein interactions with UBTF
(Figure 1.10C) [Zhang et al. 2013; Bao et al. 2015]. According to the model proposed by
Wang and colleagues, PHF6 binds to UBTF at the rDNA promoter and interferes with its
function, thereby inhibiting the transcriptional activation of rDNA genes. They go on to
44
postulate that the loss of PHF6 through somatic mutations in T-ALL and AML causes
increased transcription of rDNA genes and that this rRNA forms RNA-DNA hybrids at
the rDNA locus. This in turn results in genomic instability and DNA double strand
breaks, perhaps beneficial to leukemogenesis. To support this model, they show that
hairpin-mediated knockdown of PHF6 leads to increases in UBTF protein and rRNA
transcript levels, and increased levels of yH2AX. Furthermore, knockdown of UBTF
decreases PHF6 occupancy at the promoter of rDNA, suggesting that UBTF recruits
PHF6 [Zhang et al. 2013; Wang et al. 2013] (Figure 1.13C).
-PHF6 can Bind dsDNA but Not Histones
Protein homology of PHD-like domains and association with the NuRD complex
strongly suggests that PHF6 should have the ability to bind nucleosomes or recognize
certain histone marks [Sanchez et al. 2011]. Additionally, the second ZaP domain of
PHF6 appears to be important for hematopoietic function since T-ALL and AML
mutations are exclusively clustered within this region [Van Vlierberghe et al. 2010].
Therefore, Liu and colleagues focused their attention and biochemical analyses on the
second PHD-like domain of PHF6 [Liu et al. 2014]. Despite resembling PHD domains
that are known to bind histones and recognize lysine residues, structural analysis
showed that the second ZaP domain of PHF6 is lacking characteristics necessary for
histone binding. First, the protein structure lacks aromatic and acidic residues that are
needed to recognize methylated lysines via the aromatic cage. Second, the second ZaP
domain has an extended a-helix that sterically hinders binding of a histone tail to the
PHD-like domain. Consistent with these structural deviations, Liu and colleagues were
unable to detect binding of histones to the second ZaP domain of PHF6 [Liu et al.
2014]. On the other hand, the a-helix fold of PHF6 contains a region of positively
charged amino acids that was shown to bind dsDNA in a sequence-independent
manner (Figure 1.10C).
There are in-depth characterization studies of other proteins that contain atypical
PHD domains. For example, many proteins lack the aromatic cage in their PHD
domains necessary for recognition of H3K4me3, yet are still shown to have roles in
histone recognition and DNA binding. Such proteins, like BHC80 and DNMT3b, bind to
unmodified H3K4 [Lan et al. 2007; Ooi et al. 2007]. Additionally, DPF3B and CHD4 bind
45
acetylated residues on H3K14 and H3K9, respectively, and CHD4 can also bind
H3K9me [Lange et al. 2008; Musselman et al. 2009]. Finally, the cageless-PHD
domains of BRFP1 and BRFP2 are shown to bind both unmodified and methylated
H3K4, as well as dsDNA [Lalonde et al. 2013; Liu et al. 2012]. Most of these proteins
rely on stretches of acidic residues to coordinate histone recognition and binding, while
ATRX uses a structure rich in polar residues to bind H3K9me3 [lwase et al. 2011].
Similarly, PHF6 also contains polar amino acids, but lacks the stretch of acidic residues,
implicating that it could be acting in a manner similar to ATRX and biding H3K9me3 [Liu
et al. 2014]. Thus, further investigation into possible interactions between PHF6 and
histones is warranted.
>PHF6 is a Putative Phosphorylation Target
The use of targeted studies to elucidate the function of PHF6 has gleaned
valuable information about associations with transcriptional regulatory machinery, like
PAF1 and the NuRD complex. On the other hand, broad-scale proteomic studies hint at
another possible function of PHF6 as a substrate for phosphorylation in a variety of
contexts. First, PHF6 was identified as a potential phosphorylation target of ATM and
ATR after DNA damage, suggesting that PHF6 may play a role in the DNA damage
response [Matsuoka et al. 2007]. Further, knockdown of PHF6 results in increased
levels of yH2AX [Wang et al. 2013; Van Vlierberghe et al. 2010a]. Additionally, two
independent quantitative phosphoproteomic studies found PHF6 residues S145, S154,
and S155 to be phosphorylated during the cell cycle and in response to T-cell receptor
signaling [Dephoure et al. 2008; Mayya et al. 2009]. Further investigation into these
possible signaling events warrants further investigation.
46
Chapter 1
Introduction
Part 3 - The Epigenetic Regulation of Gene Expression
Although the function of Plant Homeodomain Finger Protein 6 (PHF6) in
lymphocytes is largely unknown, biochemical analyses suggest that it may have
important roles as an epigenetic regulator, contributing to the regulation of transcription.
Throughout development and lineage specification, establishment and regulation of
gene expression programs is vital for proper cellular development and function.
Therefore, in this section, I will broadly review the mechanisms of epigenetic regulation,
chromatin recognition, and nucleosome remodeling complexes. Then, I will widely
examine the epigenetics of lineage restriction and lineage promotion in hematopoiesis,
emphasizing how these epigenetic factors or their functions are commonly co-opted by
cancer cells during tumorigenesis. I will specifically focus the discussion on PHD
domains and the Nucleosome Remodeling and Deacetylation (NuRD), due to their
relationship with PHF6.
47
GENOMIC ORGANIZATION
The eukaryotic genome is organized into discrete units called nucleosomes,
consisting of -146bp of DNA wrapped around a core of histone proteins. The core is
made up of two copies each of histones H2A, H2B, H3, and H4. Histone H1 binds the
linker DNA between nucleosomes. The histone proteins fold in such a way that the N-
terminal portion of the protein protrudes as the "histone tail" while the C-terminus
assumes a globular formation that constitutes the core octamer. The structure of the
eukaryotic genome is tightly regulated, involving many different types of proteins
including histone- and DNA-modifying enzymes, chromatin remodeling complexes, and
transcription factors that contribute to heterochromatic and euchromatic formation,
regulation of transcription, DNA replication, and nuclear organization [Lanouette et al.
2016].
METHYLATION OF DNA
Methylation of CpG dinucleotides within CpG islands, often located at gene
promoters, results in transcriptional repression. DNA methyl-transferases (DNMTs)
deposit the methyl marks, with maintenance of the marks at gene promoters controlled
by DNMT1 and deposition of de novo marks to DNA by DNMT3A and DNMT3B [Lei et
al. 1996; Okano et al. 1999; Schermelleh et al. 2007; Gore et al. 2016]. Removal of
methylation is controlled by TET enzymes. DNA methylation is recognized by C2H2 zinc
fingers or methyl-CpG binding domains (MBD), like the MBD1-3 subunits found in the
NuRD complex [Challen et al. 2011].
The delicate balance between methylation and demethylation of target genomic
loci is crucial for proper hematopoiesis and cell differentiation. Both the activation of
lineage-specific genes (through demethylation) and repression of other fates (through
methylation) occurs during terminal differentiation. For example, demethylation is
observed in the T-cell gene Lck and the B-cell gene Pouafl during lymphoid
development, while methylation of Dachi, a myeloid progenitor specific gene,
contributes to repression of myelopoiesis. Further, the progenitor-specific genes Meisl
and Hoxa9 undergo silencing via DNA methylation in both myeloid and lymphoid fates
[Ji et al. 2010; Borgel et al. 2010]. Hematopoietic stem cells rely on DNMT1, DNMT3A,
and DNMT3B for proper self-renewal, multi-potency, and differentiation. For example,
48
loss of Dnmtl abolishes HSC activity (including self-renewal, BM niche homing, and
proper multipotent differentiation) and deletion of Dnmt3a and Dnmtl leads to improper
expression of multi-potency genes and aberrant repression of differentiation factors
[Challen et al. 2011; Trowbridge et al. 2009].
Consistent with this, changes in the methylation landscape are common
occurrences in hematopoietic malignancies. Global hypomethylation occurs in many
cancer types, while targeted hypermethylation often occurs at tumor suppressor loci
[Shen et al. 2013]. Acute myeloid malignancies commonly have loss-of-function
mutations in DNMT3A. DNMT3A mutations are also prevalent in B- and T-cell
lymphomas and myelodysplastic syndrome (MDS) patients [Kretzmer et al. 2015;
Couronne et al. 2012; Ley et al. 2010; Yan et al. 2011; Gore et al. 2016] (Figure 1.14).
DNA
Methyltransferase DNMT3A AML 20-22%
Ley et a. 2010
Yan et at. 2011
MPN 13% Teffer et al. 2009
DNA TET2 MDS 26% Langemeijer et al. 2009
Demethylase CMML 50% Kosmider et al. 2009
IDHI/2 AML 6-8% Figueroa et al. 2010
MLL Various Leukemias 80% infant Aplan, 2006
Histone 5-10% adults Muntean et al. 2010
Methyltransferase MLL2 DLBCL 22-32% Pasqualucci et al. 2011
Follicular Lymphoma 89% Morin et al. 2011
Demethylase KDM2B DLBCL 7.4% Pasqualucci et al. 2011
Histone CREBBP/ Follicular Lymphoma 41% Pasqualucci et al. 2011Histytoanee CEP DLBCL 39%Acetyltransferase EP300 ALL 17% Mullighan et al. 2011
Chromatin SWI/SNF's
ns AIRIDIAI Burkitt Lymphoma 25% Love et al. 2012SMARC4
MDS/MPNs LOF - 12% Ernst et al. 2010
Polycomb EZH2 T-ALL LOF - 25% Ntziarchristos et al. 2012
PRC2 EPTALL LOF - 48% Zhang et al. 2012
DLBCL GOF - 20% Morin et al. 2010
Histone Variant HISTIH1C Non-Hodgkin's 7% Morin et al. 2011Lymphoma
Figure 1.14. Table chronicling common mutations in epigenetic factors found in hematological malignancies. AML,
acute myeloid leukemia. MPN, myeloproliferative neoplasms. MDS, myelodysplastic syndrome. CMML, chronic
myelomonocytic leukemia. DLBCL, diffuse large B-cell lymphoma. ALL, acute lymphoblastic leukemia. EPT, early T-
cell precursor. Adapted from Timp et al. 2013.
49
I
HISTONE POST-TRANSLATIONAL MODIFICATIONS
Transcriptional regulation is a highly complex process that involves integration of
a variety of epigenetic signals for one final output. This process is controlled by the
addition, removal, or recognition of post-translational modifications (PTMs) on histone
tails by epigenetic "writers," "erasers," and "readers," respectively. Histone tails can be
modified at discrete locations, undergoing most commonly acetylation, methylation, and
phosphorylation. These modifications affect the conformation of DNA binding to histone
proteins and can be recognized by distinct protein domains that recruit chromatin
modifying complexes, resulting in compaction or decompaction of chromatin. This, in
turn, is the basis of transcriptional activation or repression [Yang 2016]. While all core
histone protein tails can be modified, I will focus the next sections on reviewing the
modifications that occur on Histone H3. Figure 1.15 details the characterization of these
marks further, including their writers, erasers, and readers. Briefly, the marks of
transcriptional activation are H3K27ac, H3K4me3, H3K36me1/2/3, and H3K79mel/2/3,
while the most common markers of transcriptional repression are H3K9me3 and
H3K27me3.
>-Acetylation
Acetylation of lysine residues neutralizes the positive charge of the amino acid
side chain, abrogating the interaction with negatively charged DNA. This results in
relaxation of DNA binding to histones, allowing increased DNA accessibility for binding
of transcription factors and transcriptional machinery. Acetylation is exclusively a marker
of transcriptional activation and is controlled by the enzymatic activity of histone
acetyltransferases (HATs), like p300, and histone deacetylases (HDACs). Often the
H3K27 and H3K9 residues undergo acetylation at enhancers and promoters of target
genes [Yang 2016] (Figure 1.15).
>Methylation
Unlike acetylation, the addition of up to three methyl groups to a lysine side chain
has no effect on the charge, but rather prevents other post-translational modifications
from occurring at that residue or neighboring residues. Also, methylation of certain
residues can result in transcriptional activation or repression. Methylation of H3K4
corresponds with transcriptional activation, while tri-methylation at H3K27 is a marker of
50
repression. Poised promoters, or promoters of developmental genes whose expression
can quickly be turned on, have both the activation mark H3K4me3 and repressive mark
H3K27me3.
methylation,
H3K9me3 denotes heterochromatin and is often associated with DNA
reinforcing transcriptional repression [Shen et al. 2013]. The DNA
A
*w 0 M
Fact
36 27 9 4
H2A H0 fm3
I KDM6 JMJD2C SETDB1
EZH2 JMJD1A SUV39H1
Repressive Chromatin
36 27 9 4
H2A H3 eco M034
HDACS
HATs (p300, CBP) LSD1 ML
Active Chromatin
H3K9ac
H3K27ac
H3K4mel
H3K4me2
H3K4me3
Enhancers, Promoters
Enhancers
5' ends of transcribed genes
Promoters of poised and
actively transcribed genes
Activation
Activation
Activation
Activation
Activation
Bivalency
PCAF, TAF,
p300/CBP,
GCN5
ML3/4
SETD1A1B
MLL1, SET1
HDACs
KDMia
LSD2
JARID1A/
KDM1A
BRD4,
BRD8,
SMARCA4
(Bromo/
PHD)
MLL BPTF,
PHFII8
Yang et al. 2016
Lanouette et al.
2016
Garcia et al. 2016
Shen at al. 2013
H3K9me3 Heterochromatin Repression SETDB1I2 KDM(3/4)A-D HP1 Eisseberg et al.8UV39-11/2 KM34AD H12016
Enhancers,
H3K27me3 Poised promoters (with Repesion EZH1/2 UTX, PHF8, Polycomb Shen at al. 2013
H3K4me3) Bivalency JMJD3 CBX7
5Vend of genes/mideoding Activation SETD2 KDM2A/B BRPF1
H3K36me1/2/3 regions/3' end of genes, Elongation NSD1-3 KDM4A-C DNMT3A Garcia et al. 2016
euchromatic regions
H3K79me1I2/3 Coding genes, Euchromatin Activation DOT1L TP53BPI Shen at al. 2013
H3R2me2a Repression PRMT6 Tudor, PHD Haghandish et al.
H3R8me2s PRMT5 2016
H3R2me2s
H3R1 7me2a Activation
PRMT5/7
CARMI/
PRMT4
Nagarajan et al.
2016
H31Oh medat ery esone Activation MSK1-2, PP 1Phosphorylation H3S0ph Immediate ear resnseIKK 14-3-3 Watson at al. 2016
H3S28ph Activation MSK1, JIL1
Ubiquitinatdon H2Bub1/2 Gene Bodies Activation Nagara 1t al.
Figure 1.15. (A) Diagram representing the most common histone H3 post-translational modifications (PTMs).[Flavahan et al. 2017]. (B). Table detailing the modifications, location of the marks, and their respective writers,
readers, and erasers. ac, acetyl. me, methyl. ph, phosphorylation. ub, ubiquitin. K, lysine. R, arginine. S, serine.
51
B
Acetylation
Methylatlon
i
I
methyltransferase subunit DNMT3B recognizes H3K9me3 through its PHD domain and
is able to methylate neighboring cytosine residues, depositing de novo methyl groups
onto DNA [Ooi et al. 2007]. Whereas control of the acetylation mark is shared between
many writers (HATs) and erases (HDACs), specific methylated residues have dedicated
writers and erasers. For example, the writer and eraser of H3K27me3 marks are EZH2
and UTX, respectively, while H3K4me3 marks are deposited and removed by MLL1 and
KDMI1A [Shen et al. 2013; Lanouette et al. 2016] (Figure 1.15).
>Ubiquitination and Phosphorylation
The mono-ubiquitination of histone H2B within gene bodies is essential for
modulating DNA accessibility. The addition of the -8kD peptide to K120 further unravels
DNA from chromatin and allows transcriptional machinery to enter promoter regions
[Fierz et al. 2011; Nagarajan et al. 2016]. Additionally, the phosphorylation of histone H3
at serine residues S10 and S28 within enhancer regions contributes to the activation of
gene expression [Watson et al. 2016] (Figure 1.15).
>-Histone Readers
Epigenetic readers recognize and dock to histone PTMs, often culminating in the
recruitment or stabilization of associated protein effectors. Readers often have
specificity for a certain histone mark, as detailed in Figure 1.16. Methylation can occur
on lysine and arginine residues. Recognition of methylated lysines is achieved by a
myriad of domains, all of which have variations on an aromatic cage structure that is
necessary for binding. The most well-characterized domains include chromodomains,
MBT, PHD, and Tudor domains. Study of methylarginine (Rme) readers is limited but
includes ADD, Tudor, and WD40 domains, all of which have narrow aromatic pockets
responsible for Rme recognition. The bromodomain is the most well-characterized
reader of lysine acetylation, utilizing a hydrophobic pocket to coordinate binding. In
addition to methylation and acetylation recognition, readers can also bind
phosphorylated serine or threonine, and unmodified histones [Musselman et al. 2012].
PHD domain-containing proteins have the ability to recognize a variety of histone PTMs.
In the next section, I will detail more extensively what is known about this flexible
domain, creating a framework for thinking about the atypical PHD domains found in
PHF6.
52
ADD
Ankyrin:
MBT
PHD
Unmodified Histone H3 ADD
WD40
PHD so
TTD DPF
Me Me Me Me Me
18 23 27
R T K QTAR K S TGG K AP R K QLAT K AAR K S APATGGV K P
2 4 9 14
PBIR
(f *~L
ADD
Ankyrin
BAH
Chromo-
barrel
Chromo-
domain
DCD
MBT
PHD
PWWP
TTD
Tudor
WD40
zf-CW
ADD
Tudor
VVD40
Bromodomain
DBD
DPF
Double PHD
PHD
14-3-3
BIR
Tandem BRCT
ADD
PHD
WD40
P
Figure 1.16. (A) Schematic and (B) table showing all of the histone modification reader domains and what
modification(s) they specifically recognize. Adapted from Musselman et al. 2012.
PHD DOMAINS
PHD domains are structural modules found in over 200 proteins with three main
functions: (1) as epigenetic readers in chromatin regulation by binding to particular
histone modifications and DNA; (2) in recruiting multi-protein chromatin remodeling
complexes and transcription factors; and (3) in mediating protein-protein interactions
53
A
Histone AH3
B
[Sanchez et al. 2011; Todd & Picketts 2012; Gatchalian & Kutateladze, 2015]. PHD-
domain proteins can be categorized based on the histone modifications that they
recognize, including binding to:
(1) H3K4me3 (like proteins BPTF and ING2) [Wysocka et al. 2006; Shi et al.
2006; Pefa et al. 2006]
(2) unmodified histone H3 tails (like proteins BHC80 and AIRE) [Lan et al. 2007;
Org et al. 2008]
(3) H3K9me3 (like CHD4) [Mansfield et al. 2011; Musselman et al. 2009]
(4) H3K14ac (like proteins DPF3b and CHD4) [Lange et al. 2008; Zeng et al.
2010; Musselman et al. 2009]
(5) double stranded DNA (like BRPF2) [Liu et al. 2012; Lalonde et al. 2013].
More detailed characterization of PHD-proteins and their method of recognition is found
in Figure 1.17 [Musselman & Kutateladze, 2011].
The first PHD domain was described in 1993 by Schindler and colleagues in the
Arabidopsis protein HAT3.1 [Schindler et al. 1993]. In 2006, it was discovered that PHD
domains bind to H3K4me3 marks through the study of BPTF and ING2 [Wysocka et al.
2006; Shi et al. 2006]. PHD domains are small modules of 50-80 amino acids that bind,
and are stabilized by, two zinc ions through a Cys4-His-Cys3 motif [Sanchez et al. 2011].
This structure forms an aromatic cage necessary for reading the H3K4me3 mark.
Variations of atypical PHD domains exist, allowing for recognition of other substrates
(Figure 1.17B). For example, the PHD domains of BHC80, AIRE, or DNMT3L lack the
aromatic cage, but instead harbor acidic and hydrophobic residues that bind exclusively
to unmodified H3K4. Similar conformations of acidic and hydrophobic residues drive the
binding of DPF3b and CHD4 to acetylated residues on H3K14 and H3K9 [Lange et al.
2008; Zeng et al. 2010; Musselman et al. 2009]. Complete lack of an aromatic cage is
also documented in PHD-proteins, like BRFP2 and PHF6, shown to nonspecifically bind
double stranded DNA [Liu et al. 2012; Lalonde et al. 2013].
Combinations of chromatin recognizing domains, enzymatic domains, and
protein-protein interaction domains within a single protein allow for diverse substrate
54
Modification 0, x3 9 xS
recognized by me mem
PHD domains
Histone H3
2 4 9 14 18 23 27 36
B
Canoil PHD B Unmodified H3K WyO et al. 2006
ZaP ATRX H3K9me3 Dhayalan et al. 2011
(variations of ZaP) PHF6 Unmodified H3K4 iwase et al. 2011dsDNA Liu et al. 2014
DPF3b H3K14Ac Zeng et al. 2010
Double PHD finger AIRE Unmodified H3 Org et al. 2008
CHD4 H3K9me3/Ac Musselman et al. 2009
BRFP1 Unmodified H3K4 Qin et al. 2011
PZP BPF H3K4 all PTMs Lalonde et al. 2013
AF10 dsDNA Liu et al. 2012AF1 unmodified H3K27 Chen et al. 2015
C
Activation Repression Remodelin
Figure 1.17. (A) Schematic depicting all of the histone H3 modifications recognized by PHD domains. (B) Table
showing variations of PHD domains that exist and the modifications that they recognize. (C) Schematic of PHD
domain coordination and downstream effects. Adapted from Osley et al. 2008 and Musselman & Kutateladze 2011.
55
recognition, protein-cofactor recruitment, and leads to a complex regulatory system of
transcriptional outputs (Figure 1.17C). For example, PHD-proteins can recognize the
same mark but have opposing transcriptional outcomes. In the case of ING2 and ING5,
both proteins recognize H3K4me3. However, the protein complexes that they recruit
result in transcriptional repression or activation, respectively. ING2 is associated with
the mSin3a histone deacetylase complex and ING5 associates with the MOZ/MORF
histone acetyltransferase complex [Doyon et al. 2006; Musselman & Kutateladze, 2011].
In the case of PHF8, the PHD domain binds H3K4me3, which aligns the neighboring
Jumonji domain to demethylate H3K9me1/2 [Feng et al. 2010]. Tandem combinations of
different domains within the same protein allow for catalytic specificity.
PHD domains also have other documented functions besides recognizing histone
modifications or dsDNA. These include binding to non-histone proteins and
phosphatidylinositol phosphates. For example, the PHD domain in KAP1 interacts with
an E2 ligase, UBC9, resulting in the sumoylation of KAP1 residues, that, in turn, recruit
CHD3 of the NuRD complex and SETDB1 [Morrison et al. 2016; Ivanov et al. 2007;
Feng et al. 2008]. Additionally, SP140, a leukocyte-specific protein, associates with
PIN1, a peptidyl isomerase, through its PHD domain that catalyzes a change in SP140
structure [Morrison et al. 2016; Zucchelli et al. 2013]. Furthermore, many proteins,
including ING2, TAF3, and ATX1, have been shown to bind of PIP5, nuclear
phosphatidylinositol-5-phosphate, via their PHD domains. However, the purpose of this
interaction and the role of PHD domains as novel PIP5 sensor are still being
investigated [Gozani et al. 2003; Kaadige & Ayer, 2006; Stif-Bultsma et al. 2015;
Alvarez-Venegas et al. 2006]. Many well-characterized proteins have PHD domains of
uncharacterized function, including PHF1, AIRE, and KDM5B, pointing out potentially
novel functions of PHD domains that warrant further investigation [Qin et al. 2013; Org
et al. 2008; Klein et al. 2014]. Figure 1.18 summarizes the proteins with known PHD
domains and their function in transcription. The roles of PHD domains presented here
highlight the flexibility that this protein domain has in both structure and function,
allowing for a variety of mechanisms of transcriptional regulation.
56
-PT Protein Host Complex Complex Function Outcome
BPTF NURF ATP-dependent chromatin
remodeler
Nucleosome mobility
Transcription activation
INGI mSin3a/HDAC1 Histone deacetylase Transcription repression
ING2 mSln3a/HDACI Histone deacetylase Transcription repression
ING3 NuA4/Ip8O Histone acstyltransferase Transciption activation
ING4 HBO1 Histone acetyltransferase Transcription activation
ING5 MOZIMORF Histone acetyltransferase Transcription activation
1-101
JARID1A
(KDM5A)
H3K4ms3 KDM7AKIAA1718
MLL1 MLL1
(Histone demethylase)
(Histone demethylase)
(Histone demethylase)
Histone methyltransferase
Histone methyltransferase
Transcription repression
Transcription activation
Transcription activation
PHF2 (Histone demethylase) Transcription activation
PHF8 (Histone demethylase) Transcription activation
PHO23 Rpd3 Histone deacetylase Transcription repression
PYGO/2 PYGO1I2IBCL9 Transcription factor -Wht Transcription activation
___________ signaling
RAG2 RAG12 V(D)J Recombinase Recombination
TAF3 TFID Transcription factor Transcription activation
YNG1 NuA3 Histone acetyltransferase Transcription activation
YNG2 NuA4 Histone acetyitransferase Transcription activation
AIRE i(Transcription facto) Transcription activation
ATRX (ATP-dependent chromatin Chromatin remodeling
remodeer) Heterochromatin formation
BHC80 LSD1 Histone demethylase Transcription repression
CHD4 NURD (ATPase) Transcription repression
(PHD1 ATP-dependent chromatin Chromatin remodeling
and remodeler
K3YA PHD2) Histone deacetylase
DNMT3A DNA methyltransferase) Transcription repression
DNMT3L (Regulatory factor of DNA Transcription repression
methyltransferase)
DPF3
JADEI
TRIM24
BAF
1-101
Chromatin remodeling
Histone acetyltransferase
(Transcriptional
intermediary factor)
Transcription activation
Transcription activation
Transcription activation/
repression
CHD4 NURD (ATPase) Transcription repression
(PHD2) ATP-dependent chromatin Chromatin remodeling
remodeler
H3K9_3 eHistone deacetylase
Lid2 (s.p.) (Histone demethylase) Transcription repression
SMCX (Histone demethylase) Transcription repression
ECM5 NuA3 (putative histone
H3K366e NTO1 BAF demethytase)
Histone acetyitransferase
MK4 DPF3 Chromatin remodeling Transcription activation(DPF) I
CHD4 NURD Chromatin remodeling Transcription repression
HKAc (PHD2) Transcription activation
Chromatin remodeling
Figure 1.18. Table detailing PHD-containing proteins. Adapted from Musselman & Kutateladze, 2011.
*PHD Domains and Human Disease
Misinterpretation of the histone code can have catastrophic effects including
cancer, neurological diseases, and immunodeficiencies. For example, mutations in the
PHD domain of RAG2 that interfere with H3K4me3 recognition results in the
development of certain autoimmune disorders, specifically T-B-SCID or Omenn
Syndrome [Schwarz et al. 1996; Sobacchi et al. 2006; Matthews et al. 2007]. As
detailed in Part 2, germline mutations in PHF6 cause the X-linked intellectual disability
57
I I
disorder Borjeson-Forssman-Lehmann Syndrome (BFLS). Other neurological disorders
that develop from germline mutations in the PHD domains of histone recognizing
proteins are summarized in Figure 1.19. Like PHF6, ATRX is an X-linked PHD-protein
that lacks an aromatic cage and whose inactivation causes Alpha-thalassemia X-linked
intellectual disability syndrome (a syndrome which shares many symptoms with BFLS).
These mutations affect the binding of ATRX to H3K4me0 (Figure 1.18-1.19) [lwase et al.
2011; Baker et al. 2008].
>-PHD Domains and Cancer
Chromosomal translocations are common oncogenic drivers of hematopoietic
cancers and translocations involving the PHD domain of proteins are frequent. Such is
the case for translocations involving PHF23 or JARID1A, in which the PHD domains are
often found fused to the transactivation domain of NUP98 in AMLs. This results in
targeting of the transcriptional activator NUP98 to lineage-specific progenitor genes by
the PHD domain of PHF23 or JARID1A. This binding has the potential to sustain
transcriptional activation of target genes, but it also blocks binding of repressive
complexes, like PRC2 [Wang et al. 2009a]. Fusions of NUP98 have also been
documented with NSD1, a protein that contains five PHD domains [Baker et al. 2008].
Autoimmune RAG2 T-B-SCID, Omenn Syndrome Germline
Disorders AIRE APECED Germline
INGI Breast cancer, melanoma, esophageal carcinoma, head and Somaticneck cancer
JARIDIA AML Translocation
Cancer PHF23 AML Translocation
NSDI, NSD3 AML Translocation
MLL AML, B-ALL Translocation/Deletion of PHD
PHFI Endometrial stromal sarcoma Translocation
NSDI Sotos Syndrome, Weaver Syndrome Germline
Neurologil ATRX Alpha-Thalassemia and Mental Retardation, X-linked GermlineNeurological ATXSyndrome (ATRX) Grln
Disorders CBP Rebenstein-Taybi Syndrome Germline
PHF6 Borjeson-Forssman-Lehmann Syndrome Germline
Figure 1.19. Table detailing human diseases that are caused by mutations in PHD-containing proteins. Adapted
from Baker et al. 2008.
58
Mutations in other PHD containing-genes implicated in cancer are summarized in Figure
1.19 [Baker et al. 2008].
EPIGENETIC LANDSCAPE OF HEMATOPOIESIS AND CANCER
As detailed in Part 1, cancer is often considered a genetic disease. However,
epigenetic mechanisms also heavily contribute to tumor initiation, progression,
metastasis, and treatment response. Neoplastic lymphocytes often sustain very few
mutations, yet still give rise to advanced, aggressive diseases [Alexandrov et al. 2013].
Hematological cancers have one of the lowest mutation rates with roughly 1 mutation
per megabase [Watson et al. 2013; Vogelstein et al. 2013; Quesada et al. 2011]. This
suggests that epigenetic factors may also be involved in the transformation of leukemia
cells. The discovery of SMARCAB1/SNF5 mutations in rhabdoid tumors initiated the
critical study of dysregulated epigenetics as a major mechanism of tumorigenesis
[Versteege et al. 1998]. Since then, major sequencing studies have identified that
approximately 50% of human cancers have mutations in epigenetic or chromatin-
associated proteins, highlighting the global importance of epigenetic control in all tissue
types [You et al. 2012, Shen et al. 2013].
In the following section, I will review the important role that epigenetics plays in
hematopoiesis and how cancer cells co-opt these regulatory mechanisms during
neoplastic transformation in hematological malignancies. I will organize the discussion
into two main parts, lineage promotion and lineage restriction, and highlight the ways in
which hematological malignancies often disrupt the processes that control lineage
restriction. Figure 1.14 highlights the many mutations in epigenetic regulators often
observed in hematological malignancies [Timp et al. 2013].
LINEAGE PROMOTION THROUGH TRANSCRIPTION FACTORS
Lineage promotion is the activation of specific transcriptional programs that
define a cell type. Transcription factors both drive and maintain gene expression
programs at discrete developmental stages. The process of gene regulation is dynamic,
and requires machinery at every step to coordinate the opening and compaction of
chromatin, recruitment of transcriptional machinery, and anchoring of large protein
complexes to DNA. Many different types of transcription factors exist, based on their
function and engagement with DNA and histones, including pioneer factors, terminal
59
selectors, and chromatin state regulators (summarized in Figure 1.20). Here I will
summarize how transcription factors can have very distinct chromatin and DNA binding
preferences, and how the binding of these factors is often ordered due to these
qualities.
>Canonical Transcription Factors
Transcription factors are individual proteins that recognize and bind 6-12 bp long
specific DNA sequences called motifs. They can bind cis-regulatory elements in DNA,
including transcription start sites, promoters, enhancers, silencers, and insulators. The
timing of TF expression is important during development, as discussed in Part 1;
however, the timing of DNA accessibility and TF occupancy are just as vital. Histones
and TFs compete for access to DNA, demonstrating that motif availability influenced by
chromatin remodelers and nucleosome occupancy underlies transcription factor action.
Additionally, transcriptional outputs can be governed by TF concentration and
cooperative binding of multiple factors to DNA [Spitz et al. 2012].
>Pioneer Transcription Factors
Transcription factors are most often thought to bind to open DNA, recruit
transcriptional machinery, and thus activate expression of target genes. But how do
silent genes become activated? This process and the coordination of chromatin to allow
for binding of de novo TFs is performed by "pioneer transcription factors." Pioneer
transcription factors bind DNA that is inaccessible and work in three main ways:
(1) They can act directly on chromatin which results in a permissive landscape
around de novo TF binding motifs;
(2) They can be the first TF in a complex to bind to DNA and increase the
cooperativity of sequential and subsequent factors to bind to DNA; or
(3) They can act passively by binding to genes and poising them for activation by
reducing the number of TFs that need to bind and activate transcription, thus
allowing for more rapid gene expression.
Pioneer factors differ from canonical transcription factors because they often have
nucleosome binding abilities, rather than just DNA-binding properties [Zaret & Carroll
2011]. Additionally, they are able to recognize partial or degenerate motifs on
nucleosome surfaces [Soufi et al. 2015]. Pioneer factors are implicated in lineage
60
reprogramming due to their role in cellular differentiation by priming genes for activation
during cell fate specification [Morris 2016; Iwafuchi-Doi & Zaret 2014; Iwafuchi-Doi et al.
2016]. Expression of pioneer factors along with another co-factor is often enough to
activate gene expression programs in the reprogramming of fibroblasts, whereby the
pioneer factor establishes competency for expression and the co-factor(s) drive
activation [Soufi et al. 2012; Iwafuchi-Doi et al. 2016].
B-cell development pioneer transcription factors include PU.1, and possibly
PAX5, EBF1, E47 and NF-kB [McManus et al. 2011; Treiber et al. 2010; Hayden et al.
2012; Sherwood et al. 2014; Maier et al. 2004; Hagman 2014; Heinz et al. 2010]. PU.1
coordinates nucleosome positioning and post-translational modifications of histones to
allow for binding of other transcription factors. For example, PU.1 acts at the enhancer
of the Pax5 gene, allowing for the binding of IRF4, IRF8, and NF-kB [Decker et al. 2009;
Ghisletti et al. 2010; Heinz et al. 2010]. Further, PU.1 works together with RUNX1 at the
c-fms/csf1R locus to activate transcription [Krysinska et al. 2007; Hoogenkamp et al.
2009]. In non-hematopoietic cells, ectopic expression of PU.1 leads to activation of
macrophage enhancers and local increases in H3K4mel, truly demonstrating the
pioneer activity of this TF regardless of the cell type [Ghisletti et al. 2010; Feng et al.
2008; Barozzi et al. 2014]. PAX5 displays pioneering activity by recruiting chromatin
remodeling complexes, histone modifying complexes, and other transcription factors to
target genes to induce an active chromatin state. Target genes include Rag2, Cd19,
EbfI, and Btg2 [McManus et al. 2011].
*Chromatin State Regulators and Settler Factors
Pioneer factors govern the establishment of open chromatin, but the
maintenance of the open chromatin state is controlled by chromatin state regulators
(CSRs) [Zaret & Carroll 2011; Morris et al. 2013; Iwafuchi-Doi & Zaret 2014]. Currently,
methods to differentiate between TFs that establish open chromatin (such as pioneer
factors) and those that maintain open chromatin (such as CSRs) are limited [Lamparter
et al. 2017; Sherwood et al. 2014]. Sherwood and colleagues coined the terms 'settler'
and 'migrant' transcription factors to define a group of factors that bind their motif only in
open landscapes or upon recruitment by other TFs, respectively. Settler and migrant
TFs do not coordinate the decompaction of chromatin themselves [Sherwood et al.
61
2014]. According to this study, PAX5, EBF1, and NF-kB were classified as settler
factors, rather than pioneer factors as others have suggested previously.
*Terminal Selectors
Terminal selectors are a special type of transcription factor usually discussed in
reference to maintaining expression of neuron-specific genes in Caenorhabditis elegans
[Hobert 2008]. This subgroup of TFs is necessary for controlling the expression of
terminal effector genes needed for mature cell identity. Since terminal selectors exert
control later in the hierarchy of differentiation, loss of these genes has relatively minor
effects on cellular phenotype [Hobert 2008; Morris 2016]. PAX5 is proposed to be a
terminal selector of B-cell identity, due to its important role in establishing and
maintaining gene expression programs in both progenitors and terminally differentiated
B-cells [Holmberg & Perlmann 2012; Cobaleda et al. 2007]. However, mutations in
PAX5 have catastrophic effects, found mutated in 30% of sporadic precursor B-cell
leukemia cases [Shah et al. 2013]. As I have highlighted here, transcription factors can
bind and function in discrete ways, with the chromatin environment vastly influencing
their effects. In the next section, I will detail the protein complexes that are responsible
for remodeling chromatin.
Pioneer TF Pioneer TF
Migrant TF
Settler TF
Figure 1.20. Schematic depicting the action of different types of transcription factors.
LINEAGE RESTRICTION
The process of cellular differentiation from a stem cell to a fully-committed
differentiated cell occurs through simultaneous activation of specific gene sets and
repression of alternate lineages, giving rise to a unique lineage-specific transcriptional
program. In this section, I will review the epigenetic changes, specifically modulation of
62
the chromatin landscape, that occur upon differentiation from a hematopoietic stem cell.
This process is accompanied by a loss in pluripotency and a gain in cell type-specific
functions. Conrad Waddington proposed a model whereby differentiated cells become
increasingly restricted from other lineages by following distinct developmental paths.
The further a cell progresses down a developmental path, the higher the "canal walls"
become. These walls represent the commitment to one cellular fate and the restriction
from other developmental fates governed by epigenetic processes. Recent work
suggests that cancer cells can raise or lower the height of these developmental walls,
overtaking epigenetic systems to become more restricted or permissive (Figure 1.21)
[Waddington et al. 1957].
HSC
B-lineage T-lineage Myelold lineage
Figure 1.21. Epigenetic landscape of lymphoid development. [Waddington et al. 1957]. Conrad Waddington
proposed a model whereby differentiated cells become increasingly restricted from other lineages, denoted by the
height of the "canal walls."
>Bivalent Promoters
Genes important for development and differentiation often have bivalent
domains, meaning they harbor both transcriptionally active (H3K4me3) and repressive
(H3K27me3) signals. These genes are silenced, but poised for rapid activation. Upon
63
differentiation, genes will lose this bivalency and harbor only one histone mark.
Expression of lineage-specific genes is activated by the removal of H3K27me3 signals
while simultaneous loss of H3K4me3 signals occurs at genes of alternate lineages,
reinforcing repression. Stem cells often have more genes with bivalent promoters, the
number decreasing as terminal differentiation proceeds [Bernstein et al. 2006; Cui et al.
2009]. For example, genes with bivalent domains in HSCs include Meisi, Lefi, Gata6,
Pax5, Runxl, Lmo2, and Ebfl [Bernstein et al. 2006; Adli et al. 2010].
>ATP-dependent Chromatin Remodeling Complexes
Silencing of pluripotency genes by nucleosome compaction is essential in
differentiation. Through lineage specification, chromatin accessibility becomes
increasingly restricted as spurious genetic lineages are repressed [Perino et al. 2016;
Stergachis et al. 2013; Fisher et al. 2011]. Often, multi-protein complexes of various
writers, erasers, and readers assemble to form chromatin remodeling complexes
(CRCs). Modulation of the chromatin landscape is essential for transcriptional
regulation, either by increasing DNA accessibility and allowing for the binding of
transcription factors and transcriptional machinery, or through the compaction of DNA,
thereby blocking such binding events. In the establishment of cell identity, it is therefore
necessary to promote expression of lineage-specific genes while simultaneously
repressing all other fates. In mammals, there are four main ATP-dependent chromatin
modifying families: SWI/SNF (switch/sucrose non-fermentable), ISWI (imitation SWI),
INO80, and the chromodomain-containing (CHD) Nucleosome Remodeling and
Deacetylation (NuRD) complex. Each of these ATPase families are comprised of multi-
protein units that harness the power of ATP hydrolysis to slide nucleosomes into
positions that favor transcriptional activation or repression. They are specific in function
but rely on association with factors to be recruited to specific genomic loci. In this
section I will briefly review the SWI/SNF, ISWI, and IN080 complexes, and discuss the
NuRD complex in more depth, focusing specifically on the roles of CRCs in
hematopoiesis and leukemia.
- SWI/SNF
In mammals, the SWI/SNF ATPase enzyme is either BRG1 (Smarca4) or BRM
(Smarc2) and the complex, made up of 15 or more proteins, is referred to as the BRG1/
64
BRM associated factor (or BAF) complex [Hota et al. 2016]. Within lymphocyte
development, BRG1 and BAF155 subunits drive B-cell development through
transcriptional activation of Ebfl and II7ra [Choi et al. 2012]. In CD8+ T-cell
differentiation, BRG1 and BAF57 activate CD8 expression, while silencing the Cd4
locus through recruitment of the repressor RUNX1 [Chi et al. 2002]. However, in the
context of CD4+ T-cells, BRG1 promotes Cd4 expression through chromatin remodeling
at the Cd4 locus [Jani et al. 2008].
- ISWI
Unlike the large -1.5 MDa SWI/SNF complexes, the ISWI family of ATPase
chromatin complexes is very small, with only 2-4 accessory proteins. There are two
ISWI family ATPases, SNF2H and SNF2L. SNF2H regulates open/permissive chromatin
and is found in the following complexes: ACF, CHRAC, RSF, WICH, and NoRC. SNF2L,
associated with facilitating chromatin compaction, is found in the NURF and CERF
complexes [Hota et al. 2016]. During thymocyte development, the cofactor SRF recruits
the NURF complex to the Ergi locus through direct interactions with BTPF [Landry et al.
2011].
- IN080
Three main complexes constitute the IN080 family of ATPases: ION080,
SRCAP, and P400/TIP60. The function of these chromatin facilitators is to control the
distribution of histone variant H2A.Z into nucleosomes. It has important roles in
maintaining pluripotency by facilitating open chromatin at stem-cell specific genes
[Papamichos-Chronakis et al. 2011; Mizuguchi et al. 2004; Krogan et al. 2003]. Further,
these complexes are important in somatic cell reprogramming by transcription factors
OCT4, SOX2, and NANOG by maintaining an open chromatin structure and recruiting
Mediator and Pol II to pluripotency target genes [Wang et al. 2014].
- Nucleosome Remodeling and Deacetylation (NuRD) Complex
As the name suggests, the Nucleosome Remodeling and Deacetylation (NuRD)
complex has additional enzymatic activities that set it apart from the other chromatin
remodeling complexes discussed above. The Mi-2/NuRD complex has the ability to
deacetylate histones and recruit regulatory proteins, in addition to mobilizing
nucleosomes. The main components of this complex include the ATPases CHD3 and
65
CHD4, histone deacetylases HDAC1 and HDAC2, methyl-CpG binding factors MBD2
and MBD3, protein-protein interaction factors MTA1-3, and histone-interacting
chaperones RBBP4 and RBBP7 [Ramirez & Hagman, 2009; Bowen et al. 2004]. This
complex often has roles in repressing gene expression. However, additional proteins
can associate with these core factors and affect the transcriptional output, sometimes
resulting in activation of gene expression. The NuRD complex is involved in activation of
lymphoid-specific transcriptional programs, shown to bind gene targets and contribute to
the promotion of gene expression. While this is counterintuitive to the NuRD's
deacetylation activity, HDACs have been shown to have important roles in supporting
transcriptional elongation and preventing promiscuous initiation by removing specific
acetyl marks. Global analysis of HDAC1 and HDAC2 binding reveals that these
deacetylases are enriched at actively transcribed genes, but not silent genes [Wang et
al. 2009b]. Additionally, the interaction of the NuRD complex with associated effector
proteins inhibits HDAC activity. In this case, it is postulated that the NuRD complex is
bound to developmental genes, albeit inhibited in deacetylase activity, as another
mechanism of rapidly priming genes for repression during progression through the
differentiation process to a mature, effector cell [Zhang et al. 2012]. Finally, the post-
translational modification of NuRD subunits can also affect transcriptional outcome. As
seen with MTA1, demethylation results in recruitment of the activator NuRF-trithorax
complex [Nair et al. 2013]. Thus, the NuRD complex is a flexible transcriptional
modulator, influenced by both the genomic context and by its interacting partners.
* NuRD in Hematopoiesis
The NuRD complex has well documented roles in hematopoietic stem cell
maintenance and throughout lineage specification. Targeting of the complex to target
genes via associating factors has vital implications for transcriptional activation or
inhibition during all stages of hematopoiesis, as detailed in Figure 1.22. In the
hematopoietic stem cell, conditional deletions of CHD4 results in depletion of the
hematopoietic stem cell pool and stalling of lymphoid and myeloid differentiation, while
erythrocyte numbers increase. This suggests that the NuRD complex has important
roles in maintaining the stem cell pool, promoting lymphocyte differentiation, and
repressing the erythrocyte lineage in early fate decisions [Yoshida et al. 2008].
66
Later in differentiation, the NuRD complex is essential for proper erythrocyte
development, associating with the co-repressors FOG1/2 through MTA1 to suppress
other lineages via downregulation of GATA-1 transcription factors [Hong et al. 2005;
Harju-Baker et al. 2008; Gao et al. 2009]. Suppression of Hesi by the interaction of
GATA-1, IKAROS, PRC2, and the NuRD complex occurs in multi-potent progenitors
during erythropoiesis [Roche et al. 2008; Hong et al. 2005; Ross et al. 2012].
Focusing on lymphocytes, association of factors such as Ikaros, Aiolos, and
Helios direct the NuRD complex to specific gene targets to maintain the expression of
lymphocyte-specific genes. Loss of Ikzfl results in decreased expression of lymphoid-
specific genes and a block in differentiation, beneficial for leukemogenesis [Zhang et al.
2011a; Kim et al. 1999; Koipally et al. 1999; Sridharan & Smale 2007]. In B-cells, the
NuRD complex, specifically CHD4 and MBD2, coordinates the repression of CD79a, an
immunoglobulin important for B-cell receptor signaling [Gao et al. 2009; Maier et al.
2003; Sigvardsson et al. 2002; Ramirez et al. 2012]. In later stages of development,
such as germinal center (GC) B-cells, the NuRD complex associates with BCL6 via
MTA3 to repress terminal differentiation gene targets (Cc13, Prdml, and Sdcl) [Fujita et
al. 2004; Jaye et al. 2007].
The NuRD complex modulates Cd4 expression during CD4+ versus CD8+
thymocyte fate determination by exerting both transcriptional activation and repression
activity. Briefly, the NuRD complex coordinates the repression of Cd4 in double-negative
(DN) thymocytes and activates transcription in double-positive (DP) thymocytes [Zhang
et al. 2012]. Specifically, the complex activates the expression of Cd4 by recruiting the
transcription factor HEB and the histone acetyltransferase p300 to the Cd4 enhancer.
Simultaneously, the NuRD complex represses the intergenic silencer via recruitment by
MOZ [Williams et al. 2004; Naito et al. 2007].
Post-translational modification of NuRD subunits contributes to transcriptional
activation. For example, acetylation of MTA1 recruits p300 and RNA Pol 11 to the
promoter of Pax5, activating transcription in B-cells [Balasenthil et al. 2007]. A
transcription-independent role for the NuRD complex has been described in rapidly
proliferating lymphocytes, in which the complex binds to pericentric heterochromatin
aiding in chromatin assembly during DNA replication [Lai et al. 2011; Helbling Chadick
67
et al. 2009]. The action of NuRD components in B- and T-cells highlights the plasticity
that this chromatin modifying complex can have in association with cell-type specific
interacting proteins, in varying cellular contexts, and upon post-translational
modification.
HSC
4< CHD4
MPP
KAROS/CHD4
IKAROS/CHD4 
-- IE 3)
FOGI/2-MTA1 EP(GATAI) BT N
IKAROSICHD4/MBD2(CD79a)
Plsa C 4-4- Qre OP DP
BCL6/MTA3 f m-lKAROS-jB41T3 CHD4 -
BCLIIb/MTAI D
Figure 1.22. Schematic depicting the action of the NuRD complex throughout hematopoiesis. Red text denotes
inhibition activity. Green text denotes activating activity. Parenthesis signify the specific genes that are being
regulated. Adapted from Cedar et al. 2011 and Dege et al. 2014.
68
e NuRD in Leukemia
As detailed above, the NuRD complex has essential and complicated roles
during hematopoiesis and lineage specification. Cancer cells have thus co-opted the
function of the NuRD complex to contribute to tumorigenesis. Specifically, leukemia cells
recruit the NuRD complex to target genes to aberrantly repress gene expression,
resulting in a block in differentiation. For example, the oncogenic fusion protein PML-
RARa recruits the NuRD complex to the promoters of tumor suppressors, like RARP2,
to suppress their expression and cause a block in differentiation in acute promyelocytic
leukemia [Morey et al. 2008]. In mouse models of acute myeloid leukemia driven by
Setbp1 overexpression, Runxl gene expression is repressed by recruitment of the
NuRD by SETBP1 to the Runxl promoter. Treatment with HDAC inhibitors reverses this
inhibition and induces differentiation in AML cells [Vishwakarma et al. 2015]. Similarly,
Gao and colleagues showed that SALL4 recruits the NuRD complex to inhibit the
expression of the tumor suppressor Pten in AML [Gao et al. 2013]. This mechanism is
also seen in B-cell malignancies, with overexpression of the NuRD subunit MTA3 in
diffuse large B-cell lymphoma (DLBCL) associating with BCL-6 to repress plasma-cell
lineage genes [Fujita et al. 2004; Jaye et al. 2007]. As mentioned above, acetylated-
MTA1 and the NuRD complex are implicated in activating expression of Pax5. MTA1
was also shown to also be overexpressed in DLBCLs [Balasenthil et al. 2007]. This
demonstrates that the aberrant localization of the NuRD complex often contributes to
tumorigenesis in leukemia cells.
Here I have discussed the importance of epigenetic regulation throughout
lymphoid development and the benefits gained by neoplastic lymphocytes when these
processes are disrupted. PHF6 has tumor essential and tumor suppressive roles in B-
and T-cell leukemias, respectively, and may be doing so through epigenetic processes.
The work of this thesis aims to specifically elucidate how leukemia cells benefit from the
action or loss of this single gene.
69
REFERENCES
Adli M, Zhu J, Bernstein BE. 2010. Genome-wide chromatin maps derived from limited numbers of hematopoietic
progenitors. Nat Meth 7: 615-618.
Adolfsson J, Mbnsson R, Buza-Vidas N, Hultquist A, Liuba K, Jensen CT, Bryder D, Yang L, Borge O-J, Thoren LAM,
et al. 2005. Identification of Flt3+ Lympho-Myeloid Stem Cells Lacking Erythro-Megakaryocytic Potential. Cell
121: 295-306.
Albu DI, Feng D, Bhattacharya D, Jenkins NA, Copeland NG, Liu P, Avram D. 2007. BCL11B is required for positive
selection and survival of double-positive thymocytes. J Exp Med 204: 3003-3015.
Alexandrov LB, Nik-Zainal S, Wedge DC, Aparicio SAJR, Behjati S, Biankin AV, Bignell GR, Bolli N, Borg A,
Borresen-Dale A-L, et al. 2013. Signatures of mutational processes in human cancer. Nature 500: 415-421.
Allman D, Sambandam A, Kim S, Miller JP, Pagan A, Well D, Meraz A, Bhandoola A. 2003. Thymopoiesis
independent of common lymphoid progenitors. Nature Immunology 4: 168-174.
Alvarez-Venegas R, Sadder M, Hlavacka A, Baluska F, Xia Y, Lu G, Firsov A, Sarath G, Moriyama H, Dubrovsky JG,
et al. 2006. The Arabidopsis homolog of trithorax, ATX1, binds phosphatidylinositol 5-phosphate, and the two
regulate a common set of target genes. Proc NatlAcad Sci USA 103: 6049-6054.
Aplan PD. 2006. Chromosomal translocations involving the MLL gene: Molecular mechanisms. DNA Repair 5: 1265-
1272.
Ardinger H, Hanson J, Zellweger H. 1984. Borjeson-Forssman-Lehmann syndrome: further delineation in five
cases. Am. J. Med. Genet. 19: 653-664.
Ahsberg J, Tsapogas P, Qian H, Zetterblad J, Zandi S, Mansson R, Jonsson J-1, Sigvardsson M. 2010. Interleukin-7-
induced Stat-5 acts in synergy with Flt-3 signaling to stimulate expansion of hematopoietic progenitor cells. The
Journal of Biological Chemistry 285: 36275-36284.
Ahsberg J, Ungerback J, Strid T, Welinder E, Stjernberg J, Larsson M, Qian H, Sigvardsson M. 2013. Early B-cell
factor 1 regulates the expansion of B-cell progenitors in a dose-dependent manner. The Journal of Biological
Chemistry 288: 33449-33461.
Bain G, Romanow WJ, Albers K, Havran WL, Murre C. 1999. Positive and negative regulation of V(D)J recombination
by the E2A proteins. Journal of Experimental Medicine 189: 289-300.
Baker LA, Allis CD, Wang GG. 2008. PHD fingers in human diseases: Disorders arising from misinterpreting
epigenetic marks. Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis 647: 3-12.
Balasenthil S, Gururaj AE, Talukder AH, Bagheri-Yarmand R, Arrington T, Haas BJ, Braisted JC, Kim I, Lee NH,
Kumar R. 2007. Identification of Pax5 as a target of MTA1 in B-cell lymphomas. Cancer Research 67: 7132-
7138.
Bao Y, Liu Z, Zhang J, Wu J, Shi Y. 2015. 1H, 13C and 15N resonance assignments and secondary structure of the
human PHF6-ePHD1 domain. Biomolecular NMR Assignments 10: 1-4.
Barozzi 1, Simonatto M, Bonifacio S, Yang L, Rohs R, Ghisletti S, Natoli G. 2014. Coregulation of Transcription Factor
Binding and Nucleosome Occupancy through DNA Features of Mammalian Enhancers. Molecular Cell 54: 844-
857.
Bell JJ, Bhandoola A. 2008. The earliest thymic progenitors for T cells possess myeloid lineage potential. Nature 452:
764-767.
Berland S, Alme K, Brendehaug A, Houge G, Hovland R. 2011. PHF6 Deletions May Cause Borjeson-Forssman-
Lehmann Syndrome in Females. Mol Syndromol 1: 294-300.
Bernstein BE, Mikkelsen TS, Xie X, Kamal M, Huebert DJ, Cuff J, Fry B, Meissner A, Wernig M, Plath K, et al. 2006. A
Bivalent Chromatin Structure Marks Key Developmental Genes in Embryonic Stem Cells. Cell 125: 315-326.
Beguelin W, Popovic R, Teater M, Jiang Y, Bunting KL, Rosen M, Shen H, Yang SN, Wang L, Ezponda T, et al. 2013.
EZH2 Is Required for Germinal Center Formation and Somatic EZH2 Mutations Promote Lymphoid
Transformation. Cancer Cell 23: 677-692.
Boer den ML, van Slegtenhorst M, De Menezes RX, Cheok MH, Buijs-Gladdines JG, Peters ST, Van Zutven LJ,
Beverloo HB, Van der Spek PJ, Escherich G, et al. 2009. A subtype of childhood acute lymphoblastic leukaemia
with poor treatment outcome: a genome-wide classification study. Lancet Oncology 10: 125-134.
Borgel J, Guibert S, Li Y, Chiba H, Schubeler D, Sasaki H, Forne T, Weber M. 2010. Targets and dynamics of
promoter DNA methylation during early mouse development. Nat Genet 42: 1093-1100.
Borowitz MJ. 2014. Mixed phenotype acute leukemia. Cytometry 86: 152-153.
Borowitz MJ, Bene M-C, Harris NL, Porwit A, Matutes E. 2008. Acute leukaemias of ambiguous lineage. In: Swerdlow
S, Campo E, Harris NL, et al, editors. WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues.
Lyon: IARC, 149-155.
Borzillo GV, Ashmun RA, Sherr CJ. 1990. Macrophage lineage switching of murine early pre-B lymphoid cells
expressing transduced fms genes. Molecular and Cellular Biology 10: 2703-2714.
Bowen NJ, Fujita N, Kajita M, Wade PA. 2004. Mi-2/NuRD: multiple complexes for many purposes. Biochimica et
Biophysica Acta (BBA) - Gene Structure and Expression 1677: 52-57.
Boyd AW, Schrader JW. 1982. Derivation of macrophage-like lines from the pre-B lymphoma ABLS 8.1 using 5-
azacytidine. Nature 297: 691-693.
B6rjeson M, Forssman H, Lehmann 0. 1962. An X-linked, recessively inherited syndrome characterized by grave
mental deficiency, epilepsy, and endocrine disorder. Acta Med. Scand. 171: 13-21.
Bua DJ, Martin GM, Binda 0, Gozani 0. 2013. Nuclear phosphatidylinositol-5-phosphate regulates ING2 stability at
discrete chromatin targets in response to DNA damage. Sci Rep 3: 493-9.
70
Carter MT, Picketts DJ, Hunter AG, Graham GE. 2009. Further clinical delineation of the Borjeson-Forssman-
Lehmann syndrome in patients with PHF6mutations. Am J Med Genet 149A: 246-250.
Cedar H, Bergman Y. 2011. Epigenetics of haematopoietic cell development. Nat Rev Immunol 11: 478-488.
Chalandon Y, Thomas X, Hayette S, Cayuela JM, Abbal C, Huguet F, Raffoux E, Leguay T, Rousselot P, Lepretre S,
et al. 2015. Randomized study of reduced-intensity chemotherapy combined with imatinib in adults with Ph-
positive acute lymphoblastic leukemia. Blood 125: 3711-3719.
Challen GA, Sun D, Jeong M, Luo M, Jelinek J, Berg JS, Bock C, Vasanthakumar A, Gu H, Xi Y, et al. 2011. Dnmt3a
is essential for hematopoietic stem cell differentiation. Nat Genet 44: 23-31.
Chan 0, Jamil AR, Millius R, Kaur R, Anwer F. 2017. Mixed phenotype acute leukemia with t(9;22): success with
nonacute myeloid leukemia-type intensive induction therapy and stem cell transplantation. Clin Case Rep 5:
435-439.
Chao MM, Todd MA, Kontny U, Neas K, Sullivan MJ, Hunter AG, Picketts DJ, Kratz CP. 2010. T-cell acute
lymphoblastic leukemia in association with Borjeson-Forssman-Lehmann syndrome due to a mutation in PHF6.
Pediatr Blood Cancer 55: 722-724.
Chase A, Cross NCP. 2011. Aberrations of EZH2 in Cancer. Clinical Cancer Research 17: 2613-2618.
Chen J, Santillan DA, Koonce M, Wei W, Luo R, Thirman MJ, Zeleznik-Le NJ, Diaz MO. 2008. Loss of MLL PHD
Finger 3 Is Necessary for MLL-ENL-Induced Hematopoietic Stem Cell Immortalization. Cancer Research 68:
6199-6207.
Chen S, Yang Z, Wilkinson AW, Deshpande AJ, Sidoli S, Krajewski K, Strahl BD, Garcia BA, Armstrong SA, Patel DJ,
et al. 2015. The PZP Domain of AF10 Senses Unmodified H3K27 to Regulate DOT1L-Mediated Methylation of
H3K79. Molecular Cell 60: 319-327.
Chi TH, Wan M, Zhao K, Taniuchi 1, Chen L, Littman DR, Crabtree GR. 2002. Reciprocal regulation of CD4/CD8
expression by SWI/SNF-like BAF complexes. Nature 418: 195-199.
Choi J, Ko M, Jeon S, Jeon Y, Park K, Lee C, Lee H, Seong RH. 2012. The SWI/SNF-like BAF complex is essential
for early B cell development. J Immunol 188: 3791-3803.
Cismasiu VB, Adamo K, Gecewicz J, Duque J, Lin Q, Avram D. 2005. BCL11B functionally associates with the NuRD
complex in T lymphocytes to repress targeted promoter. Oncogene 24: 6753-6764.
Cobaleda C, Jochum W, Busslinger M. 2007a. Conversion of mature B cells into T cells by dedifferentiation to
uncommitted progenitors. Nature 449: 473-477.
Cohen A, Petsche D, Grunberger T, Freedman MH. 1992. Interleukin 6 induces myeloid differentiation of a human
biphenotypic leukemic cell line. Leuk Res. 16(8):751-760.
Comet I, Riising EM, Leblanc B, Helin K. 2016. Maintaining cell identity: PRC2-mediated regulation of transcription
and cancer. 16: 803-810.
Couronne L, Bastard C, Bernard OA. 2012. TET2 and DNMT3A mutations in human T-cell lymphoma. N Eng J Med
366: 95-96.
Coustan-Smith E, Mullighan CG, Onciu M, Behm FG, Raimondi SC, Pei D, Cheng C, Su X, Rubnitz JE, Basso G, et
al. 2009. Early T-cell precursor leukaemia: a subtype of very high-risk acute lymphoblastic leukaemia. Lancet
Oncology 10: 147-156.
Cui K, Zang C, Roh T-Y, Schones DE, Childs RW, Peng W, Zhao K. 2009. Chromatin Signatures in Multipotent
Human Hematopoietic Stem Cells Indicate the Fate of Bivalent Genes during Differentiation. Stem Cell 4: 80-93.
De Keersmaecker K, Real PJ, Gatta GD, Palomero T, Sulis ML, Tosello V, Van Vlierberghe P, Barnes K, Castillo M,
Sole X, et al. 2010. The TLX1 oncogene drives aneuploidy in T cell transformation. Nat Med 16: 1321-1327.
de Rooij JDE, van den Heuvel-Eibrink MM, van de Rijdt NKAM, Verboon LJ, de Haas V, Trka J, Baruchel A, Reinhardt
D, Pieters R, Fornerod M, et al. 2016. PHF6 mutations in paediatric acute myeloid leukaemia. British Joumal of
Haematology 175: 967-971.
Decker T, di Magliano MP, McManus S, Sun Q, Bonifer C, Tagoh H, Busslinger M. 2009. Stepwise Activation of
Enhancer and Promoter Regions of the B Cell Commitment Gene Pax5 in Early Lymphopoiesis. Immunity 30:
508-520.
DeKoter RP, Lee H-J, Singh H. 2002. PU.1 regulates expression of the interleukin-7 receptor in lymphoid progenitors.
Immunity 16: 297-309.
DeKoter RP, Singh H. 2000. Regulation of B lymphocyte and macrophage development by graded expression of PU.
1. Science 288: 1439-1441.
Dephoure N, Zhou C, Villen J, Beausoleil SA, Bakalarski CE, Elledge SJ, Gygi SP. 2008. A quantitative atlas of mitotic
phosphorylation. Proc Natl Acad Sci USA 1-12.
Dhayalan A, Tamas R, Bock I, Tattermusch A, Dimitrova E, Kudithipudi S, Ragozin S, Jeltsch A. 2011. The ATRX-ADD
domain binds to H3 tail peptides and reads the combined methylation state of K4 and K9. Human Molecular
Genetics 20: 2195-2203.
Dias S, MAnsson R, Gurbuxani S, Sigvardsson M, Kee BL. 2008. E2A Proteins Promote Development of Lymphoid-
Primed Multipotent Progenitors. Immunity 29: 217-227.
Doulatov S, Notta F, Laurenti E, Dick JE. 2012. Hematopoiesis: A Human Perspective. Stem Cell 10: 120-136.
Doyon Y, Cayrou C, Ullah M, Landry A-J, C6te V, Selleck W, Lane WS, Tan S, Yang X-J, C6te J. 2006. ING Tumor
Suppressor Proteins Are Critical Regulators of Chromatin Acetylation Required for Genome Expression and
Perpetuation. Molecular Cell 21: 51-64.
Eckstein OS, Wang L, Punia JN, Kornblau SM, Andreeff M, Wheeler DA, Goodell MA, Rau RE. 2016. Mixed-
phenotype acute leukemia (MPAL) exhibits frequent mutations in DNMT3A and activated signaling genes.
Experimental Hematology 44: 740-744.
Ernst T, Chase AJ, Score J, Hidalgo-Curtis CE, Bryant C, Jones AV, Waghorn K, Zoi K, Ross FM, Reiter A, et al.
2010. Inactivating mutations of the histone methyltransferase gene EZH2 in myeloid disorders. Nat Genet 42:
722-726.
71
Feng R, Desbordes SC, Xie H, Tillo ES, Pixley F, Stanley ER, Graf T. 2008. PU.1 and C/EBPalpha/beta convert
fibroblasts into macrophage-like cells. Proc NatlAcad Sci USA 105: 6057-6062.
Feng W, Yonezawa M, Ye J, Jenuwein T, Grummt 1. 2010. PHF8 activates transcription of rRNA genes through
H3K4me3 binding and H3K9me1/2 demethylation. 17: 445-450.
Fierz B, Chatterjee C, McGinty RK, Bar-Dagan M, Raleigh DP, Muir TW. 2011. Histone H2B ubiquitylation disrupts
local and higher-order chromatin compaction. Nature Chemical Biology 7: 113-119.
Figueroa ME, Abdel-Wahab 0, Lu C, Ward PS, Patel J, Shih A, Li Y, Bhagwat N, Vasanthakumar A, Fernandez HF, et
al. 2010. Leukemic IDH1 and IDH2 Mutations Result in&nbsp;a Hypermethylation Phenotype, Disrupt TET2
Function, and Impair Hematopoietic Differentiation. Cancer Cell 18: 553-567.
Fisher CL, Fisher AG. 2011. Chromatin states in pluripotent, differentiated, and reprogrammed cells. Current Opinion
in Genetics & Development 21: 140-146.
Flavahan WA, Gaskell E, Bernstein BE. 2017. Epigenetic plasticity and the hallmarks of cancer. Science 357:
eaal2380-1 0.
Franzoni E, Booker SA, Parthasarathy S, Rehfeld F, Grosser S, Srivatsa S, Fuchs HR, Tarabykin V, Vida I, Wulczyn
FG. 2015. miR-128 regulates neuronal migration, outgrowth and intrinsic excitability via the intellectual disability
gene Phf6. eLife 4: 207-23.
Fujita N, Jaye DL, Geigerman C, Akyildiz A, Mooney MR, Boss JM, Wade PA. 2004. MTA3 and the Mi-2/NuRD
Complex Regulate Cell Fate during B Lymphocyte Differentiation. Cell 119: 75-86.
Fujita N, Jaye DL, Kajita M, Geigerman C, Moreno CS, Wade PA. 2003. MTA3, a Mi-2/NuRD complex subunit,
regulates an invasive growth pathway in breast cancer. Cell 113: 207-219.
Gao C, Dimitrov T, Yong KJ, Tatetsu H, Jeong HW, Luo HR, Bradner JE, Tenen DG, Chai L. 2013. Targeting
transcription factor SALL4 in acute myeloid leukemia by interrupting its interaction with an epigenetic complex.
Blood 121: 1413-1421.
Gao H, Lukin K, Ramirez J, Fields S, Lopez D, Hagman J. 2009. Opposing effects of SWI/SNF and Mi-2/NuRD
chromatin remodeling complexes on epigenetic reprogramming by EBF and Pax5. Proc Nat/Acad Sci USA 106:
11258-11263.
Gardner R, Wu D, Cherian S, Fang M, Hanafi L-A, Finney 0, Smithers H, Jensen MC, Riddell SR, Maloney DG, et al.
2016. Acquisition of a CD19-negative myeloid phenotype allows immune escape of MLL-rearranged B-ALL from
CD19 CAR-T-cell therapy. 127: 2406-2410.
Gatchalian J, Kutateladze TG. 2015. PHD Fingers as Histone Readers. In Histone Recognition, pp. 27-47, Springer
International Publishing, Cham.
Gedeon AK, Kozman HM, Robinson H, Pilia G, Schlessinger D, Turner G, Mulley JC. 1996. Refinement of the
background genetic map of Xq26-q27 and gene localisation for B6rjeson-Forssman-Lehmann Syndrome. Am J
Med Genet 64: 63-68.
Georgescu C, Longabaugh WJR, Scripture-Adams DD, David-Fung E-S, Yui MA, Zarnegar MA, Bolouri H,
Rothenberg EV. 2008. A gene regulatory network armature for T lymphocyte specification. Proc NatI Acad Sci
USA 105: 20100-20105.
Georgopoulos K, Bigby M, Wang JH, Molnar A, Wu P, Winandy S, Sharpe A. 1994. The Ikaros gene is required for
the development of all lymphoid lineages. Cell 79: 143-156.
Ghisletti S, Barozzi 1, Mietton F, Polletti S, De Santa F, Venturini E, Gregory L, Lonie L, Chew A, Wei C-L, et al. 2010.
Identification and Characterization of Enhancers Controlling the Inflammatory Gene Expression Program in
Macrophages. Immunity 32: 317-328.
Girardi T, De Keersmaecker K. 2015. T-ALL: ALL a matter of Translation? Haematologica 100: 293-295.
Goetz CA, Harmon IR, O'Neil JJ, Burchill MA, Johanns TM, Farrar MA. 2005. Restricted STAT5 activation dictates
appropriate thymic B versus T cell lineage commitment. J Immunol 174: 7753-7763.
Gole B, Wiesmuller L. 2015. Leukemogenic rearrangements at the mixed lineage leukemia gene (MLL)-multiple
rather than a single mechanism. Front Cell Dev Biol 3: 1-13.
Gore AV, Weinstein BM. 2016. DNA methylation in hematopoietic development and disease. Experimental
Hematology 44: 783-790.
Gozani 0, Karuman P, Jones DR, Ivanov D, Cha J, Lugovskoy AA, Baird CL, Zhu H, Field SJ, Lessnick SL, et al.
2003. The PHD Finger of the Chromatin-Associated Protein ING2 Functions as a Nuclear Phosphoinositide
Receptor. Cell 114: 99-111.
Graf T. 2002. Differentiation plasticity of hematopoietic cells. 99: 3089-3101.
Graf T, Enver T. 2009. Forcing cells to change lineages. Nature 462: 587-594.
Grossmann V, Haferlach C, Weissmann S, Roller A, Schindela S, Poetzinger F, Stadler K, Bellos F, Kern W,
Haferlach T, et al. 2013. The molecular profile of adult T-cell acute lymphoblastic leukemia: Mutations in
RUNX1and DNMT3Aare associated with poor prognosis in T-ALL. Genes Chromosom Cancer 52: 410-422.
Haferlach T, Nagata Y, Grossmann V, Okuno Y, Bacher U, Nagae G, Schnittger S, Sanada M, Kon A, Alpermann T, et
al. 2013. Landscape of genetic lesions in 944 patients with myelodysplastic syndromes. Leukemia 28: 241-247.
Haghandish N, Cote J. 2016. Chapter 3 - The Role of Histone Mark Writers in Chromatin Signaling: Protein Arginine
Methyltransferases. Elsevier Inc.
Hagman J. 2014. Chapter 3 - Transcriptional Regulation of Early B Cell Development. Second Edition. Elsevier Ltd.
Hajjari M, Salavaty A, Crea F, Kee Shin Y. 2015. The potential role of PHF6 as an oncogene: a genotranscriptomic/
proteomic meta-analysis. Tumor Biol 37: 5317-5325.
Harju-Baker S, Costa FC, Fedosyuk H, Neades R, Peterson KR. 2008. Silencing of Agamma-globin gene expression
during adult definitive erythropoiesis mediated by GATA-1-FOG-1-Mi2 complex binding at the -566 GATA site.
Molecular and Cellular Biology 28: 3101-3113.
Hattori N, Kawamoto H, Fujimoto S, Kuno K, Katsura Y. 1996. Involvement of transcription factors TCF-1 and GATA-3
in the initiation of the earliest step of T cell development in the thymus. Journal of Experimental Medicine 184:
1137-1147.
72
Hayden MS, Ghosh S. 2012. NF- B, the first quarter-century: remarkable progress and outstanding questions. Genes
Dev 26: 203-234.
Heinz S, Benner C, Spann N, Bertolino E, Lin YC, Laslo P, Cheng JX, Murre C, Singh H, Glass CK. 2010a. Simple
Combinations of Lineage-Determining Transcription Factors Prime cis-Regulatory Elements Required for
Macrophage and B Cell Identities. Molecular Cell 38: 576-589.
Heizmann B, Kastner P, Chan S. 2013. Ikaros is absolutely required for pre-B cell differentiation by attenuating IL-7
signals. Journal of Experimental Medicine 210: 2823-2832.
Helbling Chadwick L, Chadwick BP, Jaye DL, Wade PA. 2009. The Mi-2/NuRD complex associates with
pericentromeric heterochromatin during S phase in rapidly proliferating lymphoid cells. Chromosoma 118: 445-
457.
Hermann A, Goyal R, Jeltsch A. 2004. The Dnmt1 DNA-(cytosine-C5)-methyltransferase Methylates DNA
Processively with High Preference for Hemimethylated Target Sites. Joumal of Biological Chemistry 279:
48350-48359.
Hiddingh L, Raktoe RS, Jeuken J, Hulleman E, Noske DP, Kaspers GJL, Vandertop WP, Wesseling P, Wurdinger T.
2014. Identification of temozolomide resistance factors in glioblastoma via integrative miRNA/mRNA regulatory
network analysis. Sci Rep 4: 987-10.
Hobert 0. 2008. Regulatory logic of neuronal diversity: terminal selector genes and selector motifs. Proc Nat/ Acad
Sci USA 105: 20067-20071.
Holmberg J, Perlmann T. 2012. Maintaining differentiated cellular identity. 1-11.
Holmfeldt L, Mullighan CG. 2010. PHF6 Mutations in T-Lineage Acute Lymphoblastic LeukemiaPHF6 Mutations in T-
Lineage Acute Lymphoblastic Leukemia. Pediatr Blood Cancer 55: 595-596.
Hong W, Nakazawa M, Chen Y-Y, Kori R, Vakoc CR, Rakowski C, Blobel GA. 2005. FOG-1 recruits the NuRD
repressor complex to mediate transcriptional repression by GATA-1. The EMBO Journal 24: 2367-2378.
Hoogenkamp M, Lichtinger M, Krysinska H, Lancrin C, Clarke D, Williamson A, Mazzarella L, Ingram R, Jorgensen H,
Fisher A, et al. 2009. Early chromatin unfolding by RUNX1: a molecular explanation for differential requirements
during specification versus maintenance of the hematopoietic gene expression program. 114: 299-309.
Hori S. 2010. c-Rel: A pioneer in directing regulatory T-cell lineage commitment? European Joumal of Immunology
40: 664-667.
Hota SK, Bruneau BG. 2016. ATP-dependent chromatin remodeling during mammalian development. Development
143: 2882-2897.
H6flinger S, Kesavan K, Fuxa M, Hutter C, Heavey B, Radtke F, Busslinger M. 2004. Analysis of Notch1 function by in
vitro T cell differentiation of Pax5 mutant lymphoid progenitors. J Immunol 173: 3935-3944.
Hsu C-L, King-Fleischman AG, Lai AY, Matsumoto Y, Weissman IL, Kondo M. 2006. Antagonistic effect of CCAAT
enhancer-binding protein-alpha and Pax5 in myeloid or lymphoid lineage choice in common lymphoid
progenitors. Proc NatlAcad Sci USA 103: 672-677.
Hu D, Shilatifard A. 2016. Epigenetics of hematopoiesis and hematological malignancies. Genes Dev 30: 2021-2041.
Huether R, Dong L, Chen X, Wu G, Parker M, Wei L, Ma J, Edmonson MN, Hedlund EK, Rusch MC, et al. 2014. The
landscape of somatic mutations in epigenetic regulators across 1,000 paediatric cancer genomes. Nature
Communications 5: 1-7.
Huh HJ, Lee SH, Yoo KH, Sung KW, Koo HH, Jang JH, Kim K, Kim SJ, Kim WS, Jung CW, et al. 2013. Gene
mutation profiles and prognostic implications in Korean patients with T-lymphoblastic leukemia. Ann Hematol 92:
635-644.
Hunger SP, Mullighan CG. 2015. Redefining ALL classification: toward detecting high-risk ALL and implementing
precision medicine. 125: 3977-3987.
Inaba H, Greaves M, Mullighan CG. 2013. Acute lymphoblastic leukaemia. The Lancet 381: 1943-1955.
Ivanov AV, Peng H, Yurchenko V, Yap KL, Negorev DG, Schultz DC, Psulkowski E, Fredericks WJ, White DE, Maul
GG, et al. 2007. PHD Domain-Mediated E3 Ligase Activity Directs Intramolecular Sumoylation of an Adjacent
Bromodomain Required for Gene Silencing. Molecular Cell 28: 823-837.
Iwafuchi-Doi M, Zaret KS. 2016. Cell fate control by pioneer transcription factors. Development 143: 1833-1837.
Iwafuchi-Doi M, Zaret KS. 2014. Pioneer transcription factors in cell reprogramming. Genes Dev 28: 2679-2692.
Iwase S, Bin Xiang, Ghosh S, Ren T, Lewis PW, Cochrane JC, Allis CD, Picketts DJ, Patel DJ, Li H, et al. 2011. ATRX
ADD domain links an atypical histone methylation recognition mechanism to human mental-retardation
syndrome. 18: 769-776.
Jacoby E, Nguyen SM, Fountaine TJ, Welp K, Gryder B, Qin H, Yang Y, Chien CD, Seif AE, Lei H, et al. 2016. CD19
CAR immune pressure induces B-precursor acute lymphoblastic leukaemia lineage switch exposing inherent
leukaemic plasticity. Nature Communications 7: 1-10.
Jani A, Wan M, Zhang J, Cui K, Wu J, Preston-Hurlburt P, Khatri R, Zhao K, Chi T. 2008. A novel genetic strategy
reveals unexpected roles of the Swi-Snf-like chromatin-remodeling BAF complex in thymocyte development.
Journal of Experimental Medicine 205: 2813-2825.
Jaye DL, lqbal J, Fujita N, Geigerman CM, Li S, Karanam S, Fu K, Weisenburger DD, Chan WC, Moreno CS, et al.
2007. The BCL6-associated transcriptional co-repressor, MTA3, is selectively expressed by germinal centre B
cells and lymphomas of putative germinal centre derivation. J Pathol 213: 106-115.
Ji H, Ehrlich LIR, Seita J, Murakami P, Doi A, Lindau P, Lee H, Aryee MJ, Irizarry RA, Kim K, et al. 2010.
Comprehensive methylome map of lineage commitment from haematopoietic progenitors. Nature 467: 338-342.
Kaadige MR, Ayer DE. 2006. The polybasic region that follows the plant homeodomain zinc finger 1 of Pfl is
necessary and sufficient for specific phosphoinositide binding. Journal of Biological Chemistry 281: 28831-
28836.
Kaji K, Caballero IM, MacLeod R, Nichols J, Wilson VA, Hendrich B. 2006. The NuRD component Mbd3 is required
for pluripotency of embryonic stem cells. Nat Cell Biol 8: 285-292.
73
Kalender Atak Z, De Keersmaecker K, Gianfelici V, Geerdens E, Vandepoel R, Pauwels D, Porcu M, Lahortiga 1, Brys
V, Dirks WG, et al. 2012. High Accuracy Mutation Detection in Leukemia on a Selected Panel of Cancer Genes
ed. J.D. Hoheisel. PLoS ONE 7: e38463-11.
Kikuchi K, Kasai H, Watanabe A, Lai AY, Kondo M. 2008. IL-7 Specifies B Cell Fate at the Common Lymphoid
Progenitor to Pre-ProB Transition Stage by Maintaining Early B Cell Factor Expression. J Immunol 181: 383-
392.
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.
Klein BJ, Piao L, Xi Y, Rincon-Arano H, Rothbart SB, Peng D, Wen H, Larson C, Zhang X, Zheng X, et al. 2014. The
Histone-H3K4-Specific Demethylase KDM5B Binds to Its Substrate and Product through Distinct PHD Fingers.
Cel/Reports 6: 325-335.
Klinken SP, Alexander WS, Adams JM. 1988. Hemopoietic lineage switch: v-raf oncogene converts Emu-myc
transgenic B cells into macrophages. Cell 53: 857-867.
Koipally J, Renold A, Kim J, Georgopoulos K. 1999. Repression by Ikaros and Aiolos is mediated through histone
deacetylase complexes. The EMBO Journal 18: 3090-3100.
Kondo M, Scherer DC, Miyamoto T, King AG, Akashi K, Sugamura K, Weissman IL. 2000. Cell-fate conversion of
lymphoid-committed progenitors by instructive actions of cytokines. Nature 407: 383-386.
Kosmider 0, Gelsi-Boyer V, Ciudad M, Racoeur C, Jooste V, Vey N, Quesnel B, Fenaux P, Bastie JN, Beyne-Rauzy
0, et al. 2009. TET2 gene mutation is a frequent and adverse event in chronic myelomonocytic leukemia.
Haematologica 94: 1676-1681.
Kraszewska MD, Dawidowska M, Larmonie NSD, Kosmalska M, Sedek L, Szczepaniak M, Grzeszczak W, Langerak
AW, Szczepanski T, Witt M. 2011. DNA methylation pattern is altered in childhood T-cell acute lymphoblastic
leukemia patients as compared with normal thymic subsets: insights into CpG island methylator phenotype in T-
ALL. Leukemia 26: 367-371.
Kretzmer H, Bernhart SH, Wang W, Haake A, Weniger MA, Bergmann AK, Betts MJ, Carrillo-de-Santa-Pau E, Doose
G, Gutwein J, et al. 2015. DNA methylome analysis in Burkitt and follicular lymphomas identifies differentially
methylated regions linked to somatic mutation and transcriptional control. Nat Genet 47: 1316-1325.
Krogan NJ, Keogh M-C, Datta N, Sawa C, Ryan OW, Ding H, Haw RA, Pootoolal J, Tong A, Canadien V, et al. 2003.
A Snf2 family ATPase complex required for recruitment of the histone H2A variant Htzl. Molecular Cell 12:
1565-1576.
Krysinska H, Hoogenkamp M, Ingram R, Wilson N, Tagoh H, Laslo P, Singh H, Bonifer C. 2007. A Two-Step, PU.1-
Dependent Mechanism for Developmentally Regulated Chromatin Remodeling and Transcription of the c-fms
Gene. Molecular and Cellular Biology 27: 878-887.
Lai AY, Wade PA. 2011. Cancer biology and NuRD: a multifaceted chromatin remodelling complex. Nat Rev Cancer
11: 588-596.
Laiosa CV, Stadtfeld M, Xie H, de Andres-Aguayo L, Graf T. 2006. Reprogramming of committed T cell progenitors to
macrophages and dendritic cells by C/EBP alpha and PU.1 transcription factors. Immunity 25: 731-744.
Lalonde M-E, Avvakumov N, Glass KC, Joncas F-H, Saksouk N, Holliday M, Paquet E, Yan K, Tong Q, Klein BJ, et al.
2013. Exchange of associated factors directs a switch in HBO1 acetyltransferase histone tail specificity. Genes
Dev 27: 2009-2024.
Lamparter D, Marbach D, Rueedi R, Bergmann S, Kutalik Z. 2017. Genome-Wide Association between Transcription
Factor Expression and Chromatin Accessibility Reveals Regulators of Chromatin Accessibility ed. T. Wang.
PLoS Comput Biol13: e1005311-19.
Lan F, Collins RE, De Cegli R, Alpatov R, Horton JR, Shi X, Gozani 0, Cheng X, Shi Y. 2007. Recognition of
unmethylated histone H3 lysine 4 links BHC80 to LSD1-mediated gene repression. Nature 448: 718-722.
Landais S, Quantin R, Rassart E. 2005. Radiation leukemia virus common integration at the Kis2 locus: simultaneous
overexpression of a novel noncoding RNA and of the proximal Phf6 gene. Joumal of Virology 79: 11443-11456.
Landry JW, Banerjee S, Taylor B, Aplan PD, Singer A, Wu C. 2011. Chromatin remodeling complex NURF regulates
thymocyte maturation. Genes Dev 25: 275-286.
Lange M, Kaynak B, Forster UB, Tonjes M, Fischer JJ, Grimm C, Schlesinger J, Just S, Dunkel 1, Krueger T, et al.
2008. Regulation of muscle development by DPF3, a novel histone acetylation and methylation reader of the
BAF chromatin remodeling complex. Genes Dev 22: 2370-2384.
Langemeijer SMC, Kuiper RP, Berends M, Knops R, Aslanyan MG, Massop M, Stevens-Linders E, van Hoogen P,
van Kessel AG, Raymakers RAP, et al. 2009. Acquired mutations in TET2 are common in myelodysplastic
syndromes. Nat Genet 41: 838-842.
Lanouette S, Haddad J, Zhang P, Couture JF. 2016. Chapter 2 - Impacts of Histone Lysine Methylation
on&nbsp;Chromatin. Elsevier Inc.
Le Gallo M, O'Hara AJ, Rudd ML, Urick ME, Hansen NF, O'Neil NJ, Price JC, Zhang S, England BM, Godwin AK, et
al. 2012. Exome sequencing of serous endometrial tumors identifies recurrent somatic mutations in chromatin-
remodeling and ubiquitin ligase complex genes. Nat Genet 44: 1310-1315.
Lee CK, Kim JK, Kim Y, Lee MK, Kim K, Kang JK, Hofmeister R, Durum SK, Han SS. 2001. Generation of
Macrophages from Early T Progenitors In Vitro. J Immunol 166: 5964-5969.
Lei H, Oh SP, Okano M, Juttermann R, Goss KA, Jaenisch R, Li E. 1996. De novo DNA cytosine methyltransferase
activities in mouse embryonic stem cells. Development 122: 3195-3205.
Ley TJ, Ding L, Walter MJ, McLellan MD, Lamprecht T, Larson DE, Kandoth C, Payton JE, Baty J, Welch J, et al.
2010. DNMT3A mutations in acute myeloid leukemia. N Eng J Med 363: 2424-2433.
Li M, Xiao L, Xu J, Zhang R, Guo J, Olson J, Wu Y, Li J, Song C, Ge Z. 2016. Co-existence of PHF6 and NOTCH1
mutations in adult T-cell acute lymphoblastic leukemia. Oncol Lett 1-8.
Li P, Burke S, Wang J, Chen X, Ortiz M, Lee SC, Lu D, Campos L, Goulding D, Ng BL, et al. 2010. Reprogramming of
T Cells to Natural Killer-Like Cells upon Bc11b Deletion. Science 329: 85-89.
74
Li X, Yao H, Chen Z, Wang Q, Zhao Y, Chen S. 2013. Somatic mutations of PHF6in patients with chronic myeloid
leukemia in blast crisis. Leuk Lymphoma 54: 671-672.
Lin YC, Jhunjhunwala S, Benner C, Heinz S, Welinder E, Mansson R, Sigvardsson M, Hagman J, Espinoza CA,
Dutkowski J, et al. 2010. A global network of transcription factors, involving E2A, EBF1 and Foxol, that
orchestrates B cell fate. Nature Immunology 11: 635-643.
Liu L, Qin S, Zhang J, Ji P, Shi Y, Wu J. 2012. Solution structure of an atypical PHD finger in BRPF2 and its
interaction with DNA. Journal of Structural Biology 180: 165-173.
Liu YF, Wang BY, Zhang WN, Huang JY, Li BS, Zhang M, Jiang L, Li JF, Wang MJ, Dai YJ, et al. 2016. Genomic
Profiling of Adult and Pediatric B-cell Acute Lymphoblastic Leukemia. EBIOM 8: 173-183.
Liu Z, Li F, Ruan K, Zhang J, Mei Y, Wu J, Shi Y. 2014b. Structural and functional insights into the human Borjeson-
Forssman-Lehmann syndrome-associated protein PHF6. The Journal of Biological Chemistry 289: 10069-
10083.
Liu Z, Li F, Zhang B, Li S, Wu J, Shi Y. 2015. Structural basis of plant homeodomain finger 6 (PHF6) recognition by
the retinoblastoma binding protein 4 (RBBP4) component of the nucleosome remodeling and deacetylase
(NuRD) complex. The Journal of Biological Chemistry 290: 6630-6638.
Love C, Sun Z, Jima D, Li G, Zhang J, Miles R, Richards KL, Dunphy CH, Choi WWL, Srivastava G, et al. 2012. The
genetic landscape of mutations in Burkitt lymphoma. Nat Genet 44: 1321-1325.
Lower KM, Turner G, Kerr BA, Mathews KD, Shaw MA, Gedeon AK, Schelley S, Hoyme HE, White SM, Delatycki
MB, et al. 2002. Mutations in PHF6 are associated with B6rjeson-Forssman-Lehmann syndrome. Nat Genet 32:
661-665.
Maier H, Colbert J, Fitzsimmons D, Clark DR, Hagman J. 2003. Activation of the Early B-Cell-Specific mb-1 (Ig-)
Gene by Pax-5 Is Dependent on an Unmethylated Ets Binding Site. Molecular and Cellular Biology 23: 1946-
1960.
Maier H, Ostraat R, Gao H, Fields S, Shinton SA, Medina KL, Ikawa T, Murre C, Singh H, Hardy RR, et al. 2004.
Early B cell factor cooperates with Runx1 and mediates epigenetic changes associated with mb-1 transcription.
Nature Immunology 5: 1069-1077.
Maillard I, Tu L, Sambandam A, Yashiro-Ohtani Y, Millholland J, Keeshan K, Shestova 0, Xu L, Bhandoola A, Pear
WS. 2006. The requirement for Notch signaling at the beta-selection checkpoint in vivo is absolute and
independent of the pre-T cell receptor. Journal of Experimental Medicine 203: 2239-2245.
Mansfield RE, Musselman CA, Kwan AH, Oliver SS, Garske AL, Davrazou F, Denu JM, Kutateladze TG, Mackay JP.
2011. Plant homeodomain (PHD) fingers of CHD4 are histone H3-binding modules with preference for
unmodified H3K4 and methylated H3K9. The Journal of Biological Chemistry 286: 11779-11791.
Martiat P, Mecucci C, Nizet Y, Stul M, Philippe M, Cassiman J, et al. 1990. P190 BCR/ABL transcript in a case of
Philadelphia-positive multiple myeloma. Leukemia 4:751-754.
Matsuoka S, Ballif BA, Smogorzewska A, McDonald ER 111, Hurov KE, Luo J, Bakalarski CE, Zhao Z, Solimini N,
Lerenthal Y, et al. 2007. ATM and ATR Substrate AnalysisReveals Extensive Protein NetworksResponsive to
DNA Damage. Science 1-8.
Matthews AGW, Kuo AJ, Ram6n-Maiques S, Han S, Champagne KS, lvanov D, Gallardo M, Carney D, Cheung P,
Ciccone DN, et al. 2007. RAG2 PHD finger couples histone H3 lysine 4 trimethylation with V(D)J recombination.
Nature 450: 1106-1110.
Matthews KD, Ardinger HH, Nishimura DY, Buetow KH, Murray JC, Bartley JA. 1989. Linkage localization of
Borjeson-Forssman-Lehmann syndrome. Am. J. Med. Genet. 34: 470-474.
Mavrakis KJ, Van der Meulen J, Wolfe AL, Liu X, Mets E, Taghon T, Khan AA, Setty M, Rondou P, Vandenberghe P, et
al. 2011. A cooperative microRNA-tumor suppressor gene network in acute T-cell lymphoblastic leukemia (T-
ALL). Nat Genet 43: 673-678.
Mayya V, Lundgren DH, Hwang S-1, Rezaul K, Wu L, Eng JK, Rodionov V, Han DK. 2009. Quantitative
Phosphoproteomic Analysis of T Cell Receptor Signaling Reveals System-Wide Modulation of Protein-Protein
Interactions. Science Signaling 1-17.
Mansson R, Hultquist A, Luc S, Yang L, Anderson K, Kharazi S, Al-Hashmi S, Liuba K, Thor6n L, Adolfsson J, et al.
2007. Molecular Evidence for Hierarchical Transcriptional Lineage Priming in Fetal and Adult Stem Cells and
Multipotent Progenitors. Immunity 26: 407-419.
McClellan JS, Dove C, Gentles AJ, Ryan CE, Majeti R. 2015. Reprogramming of primary human Philadelphia
chromosome-positive B cell acute lymphoblastic leukemia cells into nonleukemic macrophages. Proc Nat Acad
Sci USA 112: 4074-4079.
McLaughlin J, Chianese E, Witte ON. 1987. In vitro transformation of immature hematopoietic cells by the P210 BCR/
ABL oncogene product of the Philadelphia chromosome. Proc NatlAcad Sci USA 84: 6558-6562.
McManus S, Ebert A, Salvagiotto G, Medvedovic J, Sun Q, Tamir I, Jaritz M, Tagoh H, Busslinger M. 2011. The
transcription factor Pax5 regulates its target genes by recruiting chromatin-modifying proteins in committed B
cells. The EMBO Journal 30: 2388-2404.
Meacham CE, Lawton LN, Soto-Feliciano YM, Pritchard JR, Joughin BA, Ehrenberger T, Fenouille N, Zuber J,
Williams RT, Young RA, et al. 2015. A genome-scale in vivo loss-of-function screen identifies Phf6 as a lineage-
specific regulator of leukemia cell growth. Genes Dev 29: 483-488.
Medina KL, Pongubala JMR, Reddy KL, Lancki DW, DeKoter R, Kieslinger M, Grosschedl R, Singh H. 2004.
Assembling a Gene Regulatory Network for Specification of the B Cell Fate. Developmental Cell 7: 607-617.
Meier-Stiegen F, Schwanbeck R, Bernoth K, Martini S, Hieronymus T, Ruau D, Zenke M, Just U. 2010. Activated
Notch1 Target Genes during Embryonic Cell Differentiation Depend on the Cellular Context and Include Lineage
Determinants and Inhibitors ed. J.-A.L. Stanton. PLoS ONE 5: e11481-18.
Mets E, Van Peer G, Van der Meulen J, Boice M, Taghon T, Goossens S, Mestdagh P, Benoit Y, De Moerloose B, Van
Roy N, et al. 2014. MicroRNA-128-3p is a novel oncomiR targeting PHF6 in T-cell acute lymphoblastic leukemia.
Haematologica 99: 1326-1333.
75
Mikkola 1, Heavey B, Horcher M, Busslinger M. 2002. Reversion of B cell commitment upon loss of Pax5 expression.
Science 297: 110-113.
Mills AA. 2010. Throwing the cancer switch: reciprocal roles of polycomb and trithorax proteins. Nat Rev Cancer 10:
669-682.
Mitani K, Sato Y, Tojo A, Ishikawa F, Kobayashi Y, Miura Y, Miyazono K, Urabe A, Takaku F. 1990. Philadelphia
chromosome positive B-cell type malignant lymphoma expressing an aberrant 190 kDa bcr-abl protein. British
Journal of Haematology 76: 221-225.
Mizuguchi G, Shen X, Landry J, Wu W-H, Sen S, Wu C. 2004. ATP-driven exchange of histone H2AZ variant
catalyzed by SWR1 chromatin remodeling complex. Science 303: 343-348.
Montecino-Rodriguez E, Leathers H, Dorshkind K. 2001. Bipotential B-macrophage progenitors are present in adult
bone marrow. Nature Immunology 2: 83-88.
Morey L, Brenner C, Fazi F, Villa R, Gutierrez A, Buschbeck M, Nervi C, Minucci S, Fuks F, Di Croce L. 2008. MBD3,
a Component of the NuRD Complex, Facilitates Chromatin Alteration and Deposition of Epigenetic Marks.
Molecular and Cellular Biology 28: 5912-5923.
Morin RD, Johnson NA, Severson TM, Mungall AJ, An J, Goya R, Paul JE, Boyle M, Woolcock BW, Kuchenbauer F,
et al. 2010. Somatic mutations altering EZH2 (Tyr641) in follicular and diffuse large B-cell lymphomas of
germinal-center origin. Nat Genet 42: 181-185.
Morin RD, Mendez-Lago M, Mungall AJ, Goya R, Mungall KL, Corbett RD, Johnson NA, Severson TM, Chiu R, Field
M, et al. 2011. Frequent mutation of histone-modifying genes in non-Hodgkin lymphoma. Nature 476: 298-303.
Morinaga R, Okamoto I, Furuta K, Kawano Y, Sekijima M, Dote K, Satou T, Nishio K, Fukuoka M, Nakagawa K. 2007.
Sequential occurrence of non-small cell and small cell lung cancer with the same EGFR mutation. Lung Cancer
58: 411-413.
Morris SA. 2016b. Direct lineage reprogramming via pioneer factors; a detour through developmental gene regulatory
networks. Development 143: 2696-2705.
Morris SA, Baek S, Sung M-H, John S, Wiench M, Johnson TA, Schiltz RL, Hager GL. 2013. Overlapping chromatin-
remodeling systems collaborate genome wide at dynamic chromatin transitions. Nat Struct Mol Biol 21: 73-81.
Morrison EA, Musselman CA. 2016. Chapter 7 - The Role of PHD Fingers in Chromatin Signaling: Mechanisms and
Functional Consequences of the Recognition of Histone and Non-histone Targets. Elsevier Inc.
Mullighan CG. 2012. Molecular genetics of B-precursor acute lymphoblastic leukemia. J Clin Invest 122: 3407-3415.
Mullighan CG, Goorha S, Radtke I, Miller CB, Coustan-Smith E, Dalton JD, Girtman K, Mathew S, Ma J, Pounds SB,
et al. 2007. Genome-wide analysis of genetic alterations in acute lymphoblastic leukaemia. Nature 446: 758-
764.
Mullighan CG, Zhang J, Kasper LH, Lerach S, Payne-Turner D, Phillips LA, Heatley SL, Holmfeldt L, Collins-
Underwood JR, Ma J, et al. 2011. CREBBP mutations in relapsed acute lymphoblastic leukaemia. Nature 471:
235-239.
Muntean AG, Giannola D, Udager AM, Hess JL. 2008. The PHD fingers of MLL block MLL fusion protein-mediated
transformation. Blood 112: 4690-4693.
Muntean AG, Tan J, Sitwala K, Huang Y, Bronstein J, Connelly JA, Basrur V, Elenitoba-Johnson KSJ, Hess JL. 2010.
The PAF Complex Synergizes with MLL Fusion Proteins at HOX Loci to Promote Leukemogenesis. Cancer Cell
17: 609-621.
Musselman CA, Kutateladze TG. 2011. Handpicking epigenetic marks with PHD fingers. Nucleic Acids Research 39:
9061-9071.
Musselman CA, Lalonde M-E, C6te J, Kutateladze TG. 2012. Perceiving the epigenetic landscape through histone
readers. Nat Struct Mol Biol 19: 1218-1227.
Musselman CA, Mansfield RE, Garske AL, Davrazou F, Kwan AH, Oliver SS, O'Leary H, Denu JM, Mackay JP,
Kutateladze TG. 2009. Binding of the CHD4 PHD2 finger to histone H3 is modulated by covalent modifications.
Biochem J 423:179-187.
Nagarajan 5, Johnsen SA. 2016. Chapter 12 - Crosstalk Between Histone Modifications Integrates Various Signaling
Inputs to Fine-Tune Transcriptional Output. Elsevier Inc.
Nair SS, Li D-Q, Kumar R. 2013. A Core Chromatin Remodeling Factor Instructs Global Chromatin Signaling through
Multivalent Reading of Nucleosome Codes. Molecular Cell 49: 704-718.
Naito T, G6mez-del Arco P, Williams CJ, Georgopoulos K. 2007. Antagonistic Interactions between Ikaros and the
Chromatin Remodeler Mi-2p Determine Silencer Activity and Cd4 Gene Expression. Immunity 27: 723-734.
Nechanitzky R, Akbas D, Scherer S, Gyory 1, Hoyler T, Ramamoorthy S, Diefenbach A, Grosschedl R. 2013.
Transcription factor EBF1 is essential for the maintenance of B cell identity and prevention of alternative fates in
committed cells. Nature Immunology 14: 867-875.
Neumann M, Vosberg S, Schlee C, Heesch S, Schwartz S, G6kbuget N, Hoelzer D, Graf A, Krebs S, Bartram I, et al.
2015. Mutational spectrum of adult T-ALL. Oncotarget 6: 2754-2766.
Ng SY-M, Yoshida T, Zhang J, Georgopoulos K. 2009. Genome-wide Lineage-Specific Transcriptional Networks
Underscore Ikaros-Dependent Lymphoid Priming in Hematopoietic Stem Cells. Immunity 30: 493-507.
Nowell PC, Hungerford. 1960. A minute chromosome in human chronic granulocytic leukemia. Science 142:1497.
1960. Science 1-1.
Ntziachristos P, Tsirigos A, Vlierberghe PV, Nedjic J, Trimarchi T, Flaherty MS, Ferres-Marco D, da Ros V, Tang Z,
Siegle J, et al. 2012. Genetic inactivation of the polycomb repressive complex 2 in T cell acute lymphoblastic
leukemia. Nat Med 18: 298-303.
Nutt SL, Heavey B, Rolink AG, Busslinger M. 1999. Commitment to the B-lymphoid lineage depends on the
transcription factor Pax5. Nature 401: 556-562.
Nutt SL, Kee BL. 2007. The Transcriptional Regulation of B Cell Lineage Commitment. Immunity 26: 715-725.
O'Neill DW, Schoetz SS, Lopez RA, Castle M, Rabinowitz L, Shor E, Krawchuk D, Goll MG, Renz M, Seelig HP, et al.
2000. An ikaros-containing chromatin-remodeling complex in adult-type erythroid cells. Molecular and Cellular
Biology 20: 7572-7582.
76
Ochiai K, Maienschein-Cline M, Mandal M, Triggs JR, Bertolino E, Sciammas R, Dinner AR, Clark MR, Singh H.
2012. A self-reinforcing regulatory network triggered by limiting IL-7 activates pre-BCR signaling and
differentiation. Nature Immunology 13: 300-307.
Oguro H, Yuan J, Ichikawa H, Ikawa T, Yamazaki S, Kawamoto H, Nakauchi H, Iwama A. 2010. Poised Lineage
Specification in Multipotential Hematopoietic Stem and Progenitor Cells by the Polycomb Protein Bmi1. Stem
Cell 6: 279-286.
Okano M, Bell DW, Haber DA, Li E. 1999. DNA methyltransferases Dnmt3a and Dnmt3b are essential for de novo
methylation and mammalian development. Cell 99: 247-257.
Olson MO, Dundr M, Szebeni A. 2000. The nucleolus: an old factory with unexpected capabilities. Trends in Cell
Biology 10: 189-196.
Ooi SKT, Qiu C, Bernstein E, Li K, Jia D, Yang Z, Erdjument-Bromage H, Tempst P, Lin S-P, Allis CD, et al. 2007.
DNMT3L connects unmethylated lysine 4 of histone H3 to de novo methylation of DNA. Nature 448: 714-717.
Oosterwegel M, van de Wetering M, Dooijes D, Klomp L, Winoto A, Georgopoulos K, Meijlink F, Clevers H. 1991.
Cloning of murine TCF-1, a T cell-specific transcription factor interacting with functional motifs in the CD3-
epsilon and T cell receptor alpha enhancers. Journal of Experimental Medicine 173: 1133-1142.
Org T, Chignola F, Hetenyi C, Gaetani M, Rebane A, Liiv I, Maran U, Mollica L, Bottomley MJ, Musco G, et al. 2008.
The autoimmune regulator PHD finger binds to non-methylated histone H3K4 to activate gene expression.
EMBO reports 9: 370-376.
Palomero T, McKenna K, O-Neil J, Galinsky 1, Stone R, Suzukawa K, Stiakaki E, Kalmanti M, Fox EA, Caligiuri MA, et
al. 2006. Activating mutations in NOTCH1 in acute myeloid leukemia and lineage switch leukemias. Leukemia
20: 1963-1966.
Papaemmanuil E, Gerstung M, Malcovati L, Tauro S, Gundem G, Van Loo P, Yoon CJ, Ellis P, Wedge DC, Pellagatti
A, et al. 2013. Clinical and biological implications of driver mutations in myelodysplastic syndromes. Blood 122:
3616-27- quiz 3699.
Papamichos-Chronakis M, Watanabe S, Rando OJ, Peterson CL. 2011. Global Regulation of H2A.Z Localization by
the IN080 Chromatin-Remodeling Enzyme is Essential for Genome Integrity. Cell 144: 200-213.
Pasqualucci L, Dominguez-Sola D, Chiarenza A, Fabbri G, Grunn A, Trifonov V, Kasper LH, Lerach S, Tang H, Ma J,
et al. 2011. Inactivating mutations of acetyltransferase genes in B-cell lymphoma. Nature 471: 189-195.
Patel JP, G6nen M, Figueroa ME, Fernandez H, Sun Z, Racevskis J, Van Vlierberghe P, Dolgalev I, Thomas S,
Aminova 0, et al. 2012. Prognostic relevance of integrated genetic profiling in acute myeloid leukemia. N Eng J
Med 366: 1079-1089.
Peirs S, Van der Meulen J, Van de Walle 1, Taghon T, Speleman F, Poppe B, Van Vlierberghe P. 2015. Epigenetics in
T-cell acute lymphoblastic leukemia. Immunol Rev 263: 50-67.
Pelayo R, Dorantes-Acosta E, Vadillo E, and Fuentes-Panana E. 2012. From HSC to B-lymphoid cells in normal and
malignant hematopoiesis. Advances in Hematopoietic Stem Cell Research, Dr. Rosana Pelayo (Ed.), ISBN:
978-953-307-930-1, InTech, Available from: http://www.intechopen.com/books/advances-in-hematopoietic-stem-
cell-research/from-hsc-to-b-lymphoidcells-in-normal-and-malignant-hematopoiesis. 277-296.
Peha PV, Davrazou F, Shi X, Walter KL, Verkhusha VV, Gozani 0, Zhao R, Kutateladze TG. 2006. Molecular
mechanism of histone H3K4me3 recognition by plant homeodomain of ING2. Nature 419: 407-5.
Perino M, Veenstra GJC. 2016. Chromatin Control of Developmental Dynamics and Plasticity. Developmental Cell 38:
610-620.
Piunti A, Shilatifard A. 2016. Epigenetic balance of gene expression by Polycomb and COMPASS families. Science
352: aad9780-17.
Polli M, Dakic A, Light A, Wu L, Tarlinton DM, Nutt SL. 2005. The development of functional B lymphocytes in
conditional PU.1 knock-out mice. Blood 106: 2083-2090.
Pongubala JMR, Northrup DL, Lancki DW, Medina KL, Treiber T, Bertolino E, Thomas M, Grosschedl R, Allman D,
Singh H. 2008. Transcription factor EBF restricts alternative lineage options and promotes B cell fate
commitment independently of Pax5. Nature Immunology 9: 203-215.
Pridans C, Holmes ML, Polli M, Wettenhall JM, Dakic A, Corcoran LM, Smyth GK, Nutt SL. 2008. Identification of
Pax5 Target Genes in Early B Cell Differentiation. J Immunol 180: 1719-1728.
Pui C-H, Downing JR, Relling MV. 2004. Acute Lymphoblastic Leukemia. 1-14.
Pui C-H, Robison LL, Look AT. 2008. Acute lymphoblastic leukaemia. The Lancet 371: 1030-1043.
Pui CH, Relling MV. 2000. Topoisomerase I inhibitor-related acute myeloid leukaemia. British Journal of
Haematology 109: 13-23.
Qin S, Guo Y, Xu C, Bian C, Fu M, Gong S, Min J. 2013. Tudor domains of the PRC2 components PHF1 and PHF19
selectively bind to histone H3K36me3. Biochemical and Biophysical Research Communications 430: 547-553.
Qin S, Jin L, Zhang J, Liu L, Ji P, Wu M, Wu J, Shi Y. 2011. Recognition of Unmodified Histone H3 by the First PHD
Finger of Bromodomain-PHD Finger Protein 2 Provides Insights into the Regulation of Histone
Acetyltransferases Monocytic Leukemic Zinc-finger Protein (MOZ) and MOZ-related factor (MORF). Journal of
Biological Chemistry 286: 36944-36955.
Quesada V, Conde L, Villamor N, Ord6iez GR, Jares P, Bassaganyas L, Ramsay AJ, Bea S, Pinyol M, Martinez-
Trillos A, et al. 2011. Exome sequencing identifies recurrent mutations of the splicing factor SF3B1 gene in
chronic lymphocytic leukemia. Nat Genet 44: 47-52.
Ramirez-Carrozzi VR, Nazarian AA, Li CC, Gore SL, Sridharan R, Imbalzano AN, Smale ST. 2006. Selective and
antagonistic functions of SWI/SNF and Mi-2beta nucleosome remodeling complexes during an inflammatory
response. Genes Dev 20: 282-296.
Ramirez J, Hagman J. 2009. The Mi-2/NuRD complex: a critical epigenetic regulator of hematopoietic development,
differentiation and cancer. Epigenetics 4: 532-536.
Randle DH, Zindy F, Sherr CJ, Roussel MF. 2001. Differential effects of p19(Arf) and p16(Ink4a) loss on senescence
of murine bone marrow-derived preB cells and macrophages. Proc NatlAcad Sci USA 98: 9654-9659.
77
Rapino F, Robles EF, Richter-Larrea JA, Kallin EM, Martinez-Climent JA, Graf T. 2013. C/EBP&alpha; Induces Highly
Efficient Macrophage Transdifferentiation of B Lymphoma and Leukemia Cell Lines and Impairs Their
Tumorigenicity. Cel/Reports 3: 1153-1163.
Reddy EP, Aggarwal AK. 2012. The Ins and Outs of Bcr-AbI Inhibition. Genes & Cancer 3: 447-454.
Reizis B, Leder P. 2002. Direct induction of T lymphocyte-specific gene expression by the mammalian Notch signaling
pathway. Genes Dev 16: 295-300.
Renedo M, Martinez-Delgado B, Arranz E, Garcia MJ, Urioste M, Martinez-Ramirez A, Rivas C, Cigudosa JC, Benitez
J. 2001. Chromosomal changes pattern and gene amplification in T cell non-Hodgkin?s lymphomas. Leukemia
15: 1627-1632.
Reynaud D, A Demarco I, L Reddy K, Schjerven H, Bertolino E, Chen Z, Smale ST, Winandy S, Singh H. 2008.
Regulation of B cell fate commitment and immunoglobulin heavy-chain gene rearrangements by Ikaros. Nature
Immunology 9: 927-936.
Roberts KG, Li Y, Payne-Turner D, Harvey RC, Yang Y-L, Pei D, McCastlain K, Ding L, Lu C, Song G, et al. 2014.
Targetable Kinase-Activating Lesions in Ph-like Acute Lymphoblastic Leukemia. N Eng J Med 371: 1005-1015.
Roche AE, Bassett BJ, Samant SA, Hong W, Blobel GA, Svensson EC. 2008. The zinc finger and C-terminal domains
of MTA proteins are required for FOG-2-mediated transcriptional repression via the NuRD complex. J Mol Cell
Cardiol 44: 352-360.
Rolink AG, Nutt SL, Melchers F, Busslinger M. 1999. Long-term in vivo reconstitution of T-cell development by Pax5-
deficient B-cell progenitors. Nature 401: 603-606.
Ross J, Mavoungou L, Bresnick EH, Milot E. 2012. GATA-1 Utilizes Ikaros and Polycomb Repressive Complex 2 To
Suppress Hes1 and To Promote Erythropoiesis. Molecular and Cellular Biology 32: 3624-3638.
Rothenberg EV. 2011. T cell lineage commitment: identity and renunciation. J Immunol 186: 6649-6655.
Rothenberg EV. 2014. Transcriptional Control of Early T and B Cell Developmental Choices. Annu Rev Immunol 32:
283-321.
Rowley JD. 1973. Letter: A new consistent chromosomal abnormality in chronic myelogenous leukaemia identified by
quinacrine fluorescence and Giemsa staining. Nature 243: 290-293.
Sanchez R, Zhou M-M. 2011. The PHD finger: a versatile epigenome reader. Trends in Biochemical Sciences 36:
364-372.
Schebesta A, McManus S, Salvagiotto G, Delogu A, Busslinger GA, Busslinger M. 2007. Transcription Factor Pax5
Activates the Chromatin of Key Genes Involved in B Cell Signaling, Adhesion, Migration, and Immune Function.
Immunity 27: 49-63.
Schermelleh L, Haemmer A, Spada F, R~sing N, Meilinger D, Rothbauer U, Cardoso MC, Leonhardt H. 2007.
Dynamics of Dnmt1 interaction with the replication machinery and its role in postreplicative maintenance of DNA
methylation. Nucleic Acids Research 35: 4301-4312.
Schindler U, Beckmann H, Cashmore AR. 1993. HAT3.1, a novel Arabidopsis homeodomain protein containing a
conserved cysteine-rich region. Plant J 4: 137-150.
Schwanbeck R, Martini S, Bernoth K, Just U. 2011. The Notch signaling pathway: Molecular basis of cell context
dependency. European Joumal of Cell Biology 90: 572-581.
Schwarz K, Gauss GH, Ludwig L, Pannicke U, Li Z, Lindner D, Friedrich W, Seger RA, Hansen-Hagge TE, Desiderio
S, et al. 1996. RAG mutations in human B cell-negative SCID. Science 274: 97-99.
Schwickert TA, Tagoh H, GOltekin S, Dakic A, Axelsson E, Minnich M, Ebert A, Werner B, Roth M, Cimmino L, et al.
2014. Stage-specific control of early B cell development by the transcription factor Ikaros. Nature Immunology
15: 283-293.
Scripture-Adams DD, Damle SS, Li L, Elihu KJ, Qin S, Arias AM, Butler RR, Champhekar A, Zhang JA, Rothenberg
EV. 2014. GATA-3 dose-dependent checkpoints in early T cell commitment. J Immunol 193: 3470-3491.
Shah S, Schrader KA, Waanders E, Timms AE, Vijai J, Miething C, Wechsler J, Yang J, Hayes J, Klein RJ, et al.
2013. A recurrent germline PAX5 mutation confers susceptibility to pre-B cell acute lymphoblastic leukemia. Nat
Genet 45: 1226-1231.
Shen H, Laird PW. 2013. Interplay between the Cancer Genome and Epigenome. Cell 153: 38-55.
Sherwood RI, Hashimoto T, O'Donnell CW, Lewis S, Barkal AA, van Hoff JP, Karun V, Jaakkola T, Gifford DK. 2014.
Discovery of directional and nondirectional pioneer transcription factors by modeling DNase profile magnitude
and shape. Nat Biotechnol 32: 171-178.
Shi X, Hong T, Walter KL, Ewalt M, Michishita E, Hung T, Carney D, Peia P, Lan F, Kaadige MR, et al. 2006. ING2
PHD domain links histone H3 lysine 4 methylation to active gene repression. Nature 1-5.
Sigvardsson M, Clark DR, Fitzsimmons D, Doyle M, Akerblad P, Breslin T, Bilke S, Li R, Yeamans C, Zhang G, et al.
2002. Early B-Cell Factor, E2A, and Pax-5 Cooperate To Activate the Early B Cell-Specific mb-1 Promoter.
Molecular and Cellular Biology 22: 8539-8551.
Slany RK. 2009. The molecular biology of mixed lineage leukemia. Haematologica 94: 984-993.
Smith SG, Zhou MM. 2016. Chapter 5 - The Bromodomain as an Acetyl-Lysine Reader&nbsp;Domain. Elsevier Inc.
Sneeringer CJ, Scott MP, Kuntz KW, Knutson SK, Pollock RM, Richon VM, Copeland RA. 2010. Coordinated
activities of wild-type plus mutant EZH2 drive tumor-associated hypertrimethylation of lysine 27 on histone H3
(H3K27) in human B-cell lymphomas. Proc Natl Acad Sci USA 107: 20980-20985.
Sobacchi C, Marrella V, Rucci F, Vezzoni P, Villa A. 2006. RAG-dependent primary immunodeficiencies. Human
Mutation 27: 1174-1184.
Somasundaram R, Ahsberg J, Okuyama K, Ungerback J, Lilljebj6rn H, Fioretos T, Strid T, Sigvardsson M. 2016.
Clonal conversion of B lymphoid leukemia reveals cross-lineage transfer of malignant states. Genes Dev 30:
2486-2499.
Somasundaram R, Strid T, Sigvardsson M. 2017. Cellular plasticity in B-cell leukemia. Cell Cycle 16: 495-496.
Soto-Feliciano Y. 2016. PHF6 is a novel regulator of B-cell identity in acute lymphoblastic leukemia.
Soufi A, Donahue G, Zaret KS. 2012. Facilitators and Impediments of the Pluripotency Reprogramming Factors' Initial
Engagement with the Genome. Cell 151: 994-1004.
78
Soufi A, Garcia MF, Jaroszewicz A, Osman N, Pellegrini M, Zaret KS. 2015. Pioneer Transcription Factors Target
Partial DNA Motifs on Nucleosomes to Initiate Reprogramming. Cell 161: 555-568.
Soupir CP, Vergilio J-A, Cin PD, Muzikansky A, Kantarjian H, Jones D, Hasserjian RP. 2007. Philadelphia
Chromosome-Positive Acute Myeloid Leukemia. American Journal of Clinical Pathology 127: 642-650.
Spitz F, Furlong EEM. 2012. Transcription factors: from enhancer binding to developmental control. Nat Rev Genet
13: 613-626.
Sridharan R, Smale ST. 2007. Predominant interaction of both Ikaros and Helios with the NuRD complex in immature
thymocytes. Joumal of Biological Chemistry 282: 30227-30238.
Stergachis AB, Neph S, Reynolds A, Humbert R, Miller B, Paige SL, Vernot B, Cheng JB, Thurman RE, Sandstrom R,
et al. 2013. Developmental Fate and Cellular Maturity Encoded in Human Regulatory DNA Landscapes. Cell
154: 888-903.
Stijf-Bultsma Y, Sommer L, Tauber M, Baalbaki M, Giardoglou P, Jones DR, Gelato KA, van Pelt J, Shah Z,
Rahnamoun H, et al. 2015. The Basal Transcription Complex Component TAF3 Transduces Changes in Nuclear
Phosphoinositides into Transcriptional Output. Molecular Cell 58: 453-467.
Tao W, Levine AJ. 1999. P19(ARF) stabilizes p53 by blocking nucleo-cytoplasmic shuttling of Mdm2. Proc Nat/ Acad
Sci USA 96: 6937-6941.
Tefferi A, Pardanani A, Lim K-H, Abdel-Wahab 0, Lasho TL, Patel J, Gangat N, Finke CM, Schwager S, Mullally A, et
al. 2009. TET2 mutations and their clinical correlates in polycythemia vera, essential thrombocythemia and
myelofibrosis. Leukemia 23: 905-911.
Thirman MJ, Gill HJ, Burnett RC, Mbangkollo D, McCabe NR, Kobayashi H, Ziemin-van der Poel S, Kaneko Y,
Morgan R, Sandberg AA. 1993. Rearrangement of the MLL gene in acute lymphoblastic and acute myeloid
leukemias with 11q23 chromosomal translocations. N Engl J Med 329: 909-914.
Thompson PK, Zuriiga-Pflucker JC. 2011. On becoming a T cell, a convergence of factors kick it up a Notch along the
way. Seminars in Immunology 23: 350-359.
Timp W, Feinberg AP. 2013. Cancer as a dysregulated epigenome allowing cellular growth advantage at the expense
of the host. Nat Rev Cancer 13: 497-510.
Todd M, Ivanochko D, Picketts D. 2015. PHF6 Degrees of Separation: The multifaceted Roles of a Chromatin Adaptor
Protein. Genes 6: 325-352.
Todd MAM, Picketts DJ. 2012. PHF6 Interacts with the Nucleosome Remodeling and Deacetylation (NuRD) Complex.
J Proteome Res 11: 4326-4337.
Treiber T, Mandel EM, Pott S, Gyory 1, Firner S, Liu ET, Grosschedl R. 2010. Early B Cell Factor 1 Regulates B Cell
Gene Networks by Activation, Repression, and Transcription- Independent Poising of Chromatin. Immunity 32:
714-725.
Trowbridge JJ, Snow JW, Kim J, Orkin SH. 2009. DNA Methyltransferase 1 Is Essential for and Uniquely Regulates
Hematopoietic Stem and Progenitor Cells. Stem Cell 5: 442-449.
Turner G, Gedeon A, Mulley J, Sutherland G, Rae J, Power K, Arthur 1. 1989. Borjeson-Forssman-Lehmann
syndrome: clinical manifestations and gene localization to Xq26-27. Am. J. Med. Genet. 34: 463-469.
Uhlen M, Oksvold P, Fagerberg L, Lundberg E, Jonasson K, Forsberg M, Zwahlen M, Kampf C, Wester K, Hober S,
et al. 2010. Towards a knowledge-based Human Protein Atlas. Nat Biotechnol 28: 1248-1250.
Ungerback J, Ahsberg J, Strid T, Somasundaram R, Sigvardsson M. 2015. Combined heterozygous loss of Ebf1 and
Pax5 allows for T-lineage conversion of B cell progenitors. J Exp Med 212: 1109-1123.
Vallee D, Chevrier E, Graham GE, Lazzaro MA, Lavigne PA, Hunter AG, Picketts DJ. 2004. A novel PHF6 mutation
results in enhanced exon skipping and mild B6rjeson-Forssman-Lehmann syndrome. Journal of Medical
Genetics 41: 778-783.
van den Ancker W, Westers TM, de Leeuw DC, van der Veeken YFCM, Loonen A, van Beckhoven E, Ossenkoppele
GJ, van de Loosdrecht AA. 2013. A threshold of 10% for myeloperoxidase by flow cytometry is valid to classify
acute leukemia of ambiguous and myeloid origin. Cytometry 84B: 114-118.
Van Vlierberghe P, Ferrando A. 2012. The molecular basis of T cell acute lymphoblastic leukemia. J Clin Invest 122:
3398-3406.
Van Vlierberghe P, Palomero T, Khiabanian H, Van der Meulen J, Castillo M, Van Roy N, De Moerloose B, Philippe J,
Gonzelez-Garcia S, Toribio ML, et al. 2010a. PHF6 mutations in T-cell acute lymphoblastic leukemia. 42: 338-
342.
Van Vlierberghe P, Patel J, Abdel-Wahab 0, Lobry C, Hedvat CV, Balbin M, Nicolas C, Payer AR, Fernandez HF,
Tallman MS, et al. 2010b. PHF6 mutations in adult acute myeloid leukemia. Leukemia 25: 130-134.
Verbeek S, Izon D, Hofhuis F, Robanus-Maandag E, Riele te H, van de Wetering M, Oosterwegel M, Wilson A,
MacDonald HR, Clevers H. 1995. An HMG-box-containing T-cell factor required for thymocyte differentiation.
Nature 374: 70-74.
Versteege I, Sevenet N, Lange J, Rousseau-Merck MF, Ambros P, Handgretinger R, Aurias A, Delattre 0. 1998.
Truncating mutations of hSNF5/INI1 in aggressive paediatric cancer. Nature 394: 203-206.
Vicente C, Schwab C, Broux M, Geerdens E, Degryse S, Demeyer S, Lahortiga 1, Elliott A, Chilton L, La Starza R, et
al. 2015. Targeted sequencing identifies associations between IL7R-JAK mutations and epigenetic modulators in
T-cell acute lymphoblastic leukemia. Haematologica 100: 1301-1310.
Vishwakarma BA, Nguyen N, Makishima H, Hosono N, Gudmundsson KO, Negi V, Oakley K, Han Y, Przychodzen B,
Maciejewski JP, et al. 2015. Runx1 repression by histone deacetylation is critical for Setbpl-induced mouse
myeloid leukemia development. 30: 200-208.
Vogelstein B, Papadopoulos N, Velculescu VE, Zhou S, Diaz LA, Kinzler KW. 2013. Cancer genome landscapes.
Science 339: 1546-1558.
Voss AK, Gamble R, Collin C, Shoubridge C, Corbett M, G cz J, Thomas T. 2007. Protein and gene expression
analysis of Phf6, the gene mutated in the B?rjeson?Forssman?Lehmann Syndrome of intellectual disability and
obesity. Gene Expression Patterns 7: 858-871.
79
Voss TC, Hager GL. 2013. Dynamic regulation of transcriptional states by chromatin and transcription factors. 15: 69-
81.
Wada H, Masuda K, Satoh R, Kakugawa K, Ikawa T, Katsura Y, Kawamoto H. 2008. Adult T-cell progenitors retain
myeloid potential. Nature 452: 768-772.
Waddington C, Kacser H. 1957. The Strategy of the Genes: A Discussion of Some Aspects of Theoretical Biology. 1-
274.
Wang GG, Song J, Wang Z, Dormann HL, Casadio F, Li H, Luo J-L, Patel DJ, Allis CD. 2009a. Haematopoietic
malignancies caused by dysregulation of a chromatin-binding PHD finger. Nature 459: 847-851.
Wang J, Leung JWC, Gong Z, Feng L, Shi X, Chen J. 2013. PHF6 Regulates Cell Cycle Progression by Suppressing
Ribosomal RNA Synthesis. Journal of Biological Chemistry 288: 3174-3183.
Wang L, Du Y, Ward JM, Shimbo T, Lackford B, Zheng X, Miao Y-L, Zhou B, Han L, Fargo DC, et al. 2014. IN080
Facilitates Pluripotency Gene Activation in Embryonic Stem Cell Self-Renewal, Reprogramming, and Blastocyst
Development. Stem Cell 14: 575-591.
Wang Q, Qiu H, Jiang H, Wu L, Dong S, Pan J, Wang W, Ping N, Xia J, Sun A, et al. 2011. Mutations of PHF6 are
associated with mutations of NOTCH1, JAK1 and rearrangement of SET-NUP214 in T-cell acute lymphoblastic
leukemia. Haematologica 96:1808-1814.
Wang Y, Zhang H, Chen Y, Sun Y, Yang F, Yu W, Liang J, Sun L, Yang X, Shi L, et al. 2009b. LSD1 Is a Subunit of the
NuRD Complex and Targets the Metastasis Programs in Breast Cancer. Cell 138: 660-672.
Wang Z, Zang C, Cui K, Schones DE, Barski A, Peng W, Zhao K. 2009c. Genome-wide Mapping of HATs and HDACs
Reveals Distinct Functions in Active and Inactive Genes. Cell 138: 1019-1031.
Watson IR, Takahashi K, Futreal PA, Chin L. 2013. Emerging patterns of somatic mutations in cancer. Nat Rev Genet
14: 703-718.
Watson NA, Higgins JMG. 2016. Chapter 4 - Histone Kinases and Phosphatases. Elsevier Inc.
Weinberg OK, Arber DA. 2010. Mixed-phenotype acute leukemia: historical overview and a new definition. Leukemia
24:1844-1851.
Weinberg OK, Seetharam M, Ren L, Alizadeh A, Arber DA. 2014. Mixed Phenotype Acute Leukemia: A Study of 61
Cases Using World Health Organization and European Group for the Immunological Classification of
Leukaemias Criteria. American Journal of Clinical Pathology 142: 803-808.
Weng AP, Ferrando AA, Lee W, Morris JP, Silverman LB, Sanchez-Irizarry C, Blacklow SC, Look AT, Aster JC. 2004.
Activating mutations of NOTCH1 in human T cell acute lymphoblastic leukemia. Science 306: 269-271.
Williams CJ, Naito T, Arco PG-D, Seavitt JR, Cashman SM, De Souza B, Qi X, Keables P, Andrian Von UH,
Georgopoulos K. 2004. The chromatin remodeler Mi-2beta is required for CD4 expression and T cell
development. Immunity 20: 719-733.
Williams RT, Besten den W, Sherr CJ. 2007. Cytokine-dependent imatinib resistance in mouse BCR-ABL+, Arf-null
lymphoblastic leukemia. Genes Dev 21: 2283-2287.
Williams RT, Roussel, Sherr CJ. 2006. Arf gene loss enhances oncogenicity and limits imatinib response in mouse
models of Bcr-Abl-induced acute lymphoblastic leukemia. Proc NatlAcad Sci USA 1-6.
Wolach, Stone RM. 2015. How I treat mixed-phenotype acute leukemia. 1-10.
Wysocka J, Swigut T, Xiao H, Milne TA, Kwon SY, Landry J, Kauer M, Tackett AJ, Chait BT, Badenhorst P, et al. 2006.
A PHD finger of NURF couples histone H3 lysine 4 trimethylation with chromatin remodelling. Nature 419: 407-
6.
Xie H, Ye M, Feng R, Graf T. 2004. Stepwise reprogramming of B cells into macrophages. Cell 117: 663-676.
Yan L, Ping N, Zhu M, Sun A, Xue Y, Ruan C, Drexler HG, MacLeod RAF, Wu D, Chen S. 2012. Clinical,
immunophenotypic, cytogenetic, and molecular genetic features in 117 adult patients with mixed-phenotype
acute leukemia defined by WHO-2008 classification. Haematologica 97: 1708-1712.
Yan X-J, Xu J, Gu Z-H, Pan C-M, Lu G, Shen Y, Shi J-Y, Zhu Y-M, Tang L, Zhang X-W, et al. 2011. Exome
sequencing identifies somatic mutations of DNA methyltransferase gene DNMT3A in acute monocytic leukemia.
Nat Genet 43: 309-315.
Yang XJ. 2016. Chapter 1 - Histone Acetyltransferases, Key Writers of the Epigenetic Language. Elsevier Inc.
Yap DB, Chu J, Berg T, Schapira M, Cheng S-WG, Moradian A, Morin RD, Mungall AJ, Meissner B, Boyle M, et al.
2011. Somatic mutations at EZH2 Y641 act dominantly through a mechanism of selectively altered PRC2
catalytic activity, to increase H3K27 trimethylation. Blood 117: 2451-2459.
Yoo NJ, Kim YR, Lee SH. 2012. Somatic mutation of PHF6 gene in T-cell acute lymphoblatic leukemia, acute
myelogenous leukemia and hepatocellular carcinoma. Acta Oncol 51: 107-111.
Yoshida T, Hazan I, Zhang J, Ng SY, Naito T, Snippert HJ, Heller EJ, Qi X, Lawton LN, Williams CJ, et al. 2008. The
role of the chromatin remodeler Mi-2 in hematopoietic stem cell self-renewal and multilineage differentiation.
Genes Dev 22: 1174-1189.
Yoshida T, Yao-Ming Ng S, ZOniga-Pflucker JC, Georgopoulos K. 2006. Early hematopoietic lineage restrictions
directed by Ikaros. Nature Immunology 7: 382-391.
You JS, Jones PA. 2012. Cancer genetics and epigenetics: two sides of the same coin? Cancer Cell 22: 9-20.
Yuan L, Lu L, Yang Y, Sun H, Chen X, Huang Y, Wang X, Zou L, Bao L. 2015. Genetic mutational profiling analysis of
T cell acute lymphoblastic leukemia reveal mutant FBXW7 as a prognostic indicator for inferior survival. Ann
Hematol 94: 1817-1828.
Zaret KS, Carroll JS. 2011. Pioneer transcription factors: establishing competence for gene expression. Genes Dev
25: 2227-2241.
Zeng L, Yap KL, Ivanov AV, Wang X, Mujtaba S, Plotnikova 0, Rauscher FJ 111, Zhou M-M. 2008. Structural insights
into human KAP1 PHD finger-bromodomain and its role in gene silencing. Nat Struct Mol Biol 15: 626-633.
Zeng L, Zhang Q, Li S, Plotnikov AN, Walsh MJ, Zhou M-M. 2010. Mechanism and regulation of acetylated histone
binding by the tandem PHD finger of DPF3b. Nature 466: 258-262.
80
Zhang C, Mejia LA, Huang J, Valnegri P, Bennett EJ, Anckar J, Jahani-Asl A, Gallardo G, Ikeuchi Y, Yamada T, et al.
2013. The X-Linked Intellectual Disability Protein PHF6 Associates with the PAF1 Complex and Regulates
Neuronal Migration in the Mammalian Brain. Neuron 78: 986-993.
Zhang J, Ding L, Holmfeldt L, Wu G, Heatley SL, Payne-Turner D, Easton J, Chen X, Wang J, Rusch M, et al. 2012a.
The genetic basis of early T-cell precursor acute lymphoblastic leukaemia. Nature 481: 157-163.
Zhang J, Ding L, Holmfeldt L, Wu G, Heatley SL, Payne-Turner D, Easton J, Chen X, Wang J, Rusch M, et al. 2012b.
The genetic basis of early T-cell precursor acute lymphoblastic leukaemia. Nature 481: 157-163.
Zhang J, Dominguez-Sola D, Hussein S, Lee J-E, Holmes AB, Bansal M, Vlasevska S, Mo T, Tang H, Basso K, et al.
2015. Disruption of KMT2D perturbs germinal center B cell development and promotes lymphomagenesis. Nat
Med 21: 1190-1198.
Zhang J, Jackson AF, Naito T, Dose M, Seavitt J, Liu F, Heller EJ, Kashiwagi M, Yoshida T, Gounari F, et al. 2011a.
Harnessing of the nucleosome-remodeling-deacetylase complex controls lymphocyte development and prevents
leukemogenesis. Nature Immunology 13: 86-94.
Zhang J, Mullighan CG, Harvey RC, Wu G, Chen X, Edmonson M, Buetow KH, Carroll WL, Chen IM, Devidas M, et
al. 2011b. Key pathways are frequently mutated in high-risk childhood acute lymphoblastic leukemia: a report
from the Children's Oncology Group. 118: 3080-3087.
Zucchelli C, Tamburri S, Quilici G, Palagano E, Berardi A, Saare M, Peterson P, Bachi A, Musco G. 2013. Structure of
human Sp140 PHD finger: an atypical fold interacting with Pin1. FEBS J 281: 216-231.
81
82
Chapter 2 - Results
PHF6 regulates phenotypic plasticity through
chromatin organization within lineage-specific genes
Jordan M.E. Bartlebaugh, Yadira M. Soto-Feliciano, Yunpeng Liu, Francisco J.
Sanchez-Rivera, Arjun Bhutkar, Abraham S. Weintraub, Jason D. Buenrostro, Christine
S. Cheng, Aviv Regev, Tyler E. Jacks, Richard A. Young, and Michael T. Hemann
An abbreviated version of this chapter has been accepted for publication in Genes &
Development on May 15, 2017.
JMEB, YMSF, and MTH designed the study. JMEB, YMSF, YL, and FJSR conducted experiments. YL
performed critical RNA-Seq, ChIP-Seq, ATAC-Seq and mutational analyses. AB performed RNA-Seq and
ICA analyses. ASW and RAY provided reagents and support with ChIP-Seq experiments. JDB, CSC, and
AR provided reagents and support with ATAC-Seq experiments. Jackie Lees and TEJ gave vital
conceptual advice. JMEB, YMSF, YL, and MTH wrote the manuscript.
83
ABSTRACT
Developmental and lineage plasticity have been observed in numerous malignancies,
and have been correlated with tumor progression and drug resistance. However, little is
known about the molecular mechanisms that enable such plasticity to occur. Here, we
describe the function of the Plant Homeodomain Finger Protein 6 (PHF6) in leukemia
and define its role in regulating chromatin accessibility to lineage-specific transcription
factors. We show that loss of Phf6 in B-cell leukemia results in systematic changes in
gene expression via alterations of the chromatin landscape at the transcriptional start
sites of B- and T-cell specific factors. Additionally, Phf6KO cells show significant down-
regulation of genes involved in the development and function of normal B-cells, up-
regulation of genes involved in T-cell signaling, and give rise to a mixed-lineage
lymphoma in vivo. Engagement of divergent transcriptional programs results in
phenotypic plasticity that leads to altered disease presentation in vivo, tolerance of
aberrant oncogenic signaling, and differential sensitivity to frontline and targeted
therapies. These findings suggest that active maintenance of a precise chromatin
landscape is essential for sustaining proper leukemia cell identity, and that loss of a
single factor (PHF6) can cause focal changes in chromatin accessibility and
nucleosome positioning that render cells susceptible to lineage transition.
84
INTRODUCTION
Lymphopoiesis is a carefully orchestrated process where hematopoietic stem
cells differentiate into B- and T-cells through activation of precise gene expression
programs governed by well-described transcription factors. During the process of
lymphocyte differentiation, gene expression must be activated, silenced, or maintained.
In B-cell development, the accessibility of specific genomic loci for binding of
transcription factors (e.g. PU.1, E2A, EBF1, and PAX5) is vital for transcriptional
regulation. However, the critical mechanisms regulating the chromatin landscape and
transcription factor access have yet to be elucidated [Schebesta et al. 2002].
Lymphoid cells are terminally differentiated, but they still possess the capacity to
transdifferentiate into distinct lineages when subject to specific genetic perturbations.
For example, loss of Pax5 in mature B-cells can produce functional T-cells in
immunodeficient mice [Cobaleda et al. 2007]. Similarly, Pax5 deletion in pro-B cells
allows for transdifferentiation into macrophages, granulocytes, osteoclasts, dendritic
cells, and natural killer cells [Nutt et al. 1999]. Overexpression of CEBPa/P can
transform mature B- and T-cells into macrophages [Xie et al. 2004; Laiosa et al. 2006].
In addition, Ebfl loss converts pro-B cells into innate lymphoid cells and T-cells
[Nechanitzky et al. 2013]. Interestingly, these lineage-specific transcription factors are
often found to be altered in B-cell acute lymphoblastic leukemia (B-ALL). These findings
highlight the plasticity of leukemia cells and how aberrant lymphoid developmental
programs can favor leukemogenesis [Somasundaram et al. 2015; Horcher et al. 2001;
Rathert et al. 2015].
Plant homeodomain finger protein 6 (PHF6) was first described as the single
gene mutated in the X-linked intellectual disability disorder called B6rjeson-Forssman-
Lehmann syndrome [Lower et al. 2002]. Inactivating mutations in PHF6 were
subsequently identified in human leukemias of the T- and myeloid lineages, highlighting
its role as a tumor suppressor gene in these malignancies [Van Vlierberghe et al. 2010;
Van Vlierberghe et al. 2011]. Our group recently described a tumor-promoting role for
85
Phf6 in a murine model of BCR-ABL1+ B-ALL [Meacham et al. 2015; Williams et al.
2006]. In this study, we found that hairpin-mediated knockdown of Phf6 leads to
impaired growth of B-ALL cells in vivo. Taken together, these observations suggest that
PHF6 acts as a tumor suppressor or an oncogene in a lineage-dependent manner.
However, the molecular mechanisms underlying PHF6's function in hematological
malignancies remain entirely unknown (Fig. 2.1A).
Despite the knowledge obtained through sequencing studies, only a handful of
functions have been described for PHF6. The protein contains two atypical PHD-like
zinc finger domains, implying the capacity to bind modified histone tails similar to
canonical PHD domains [Wysocka et al. 2006]. However, PHF6 has only been shown to
bind double stranded DNA in vitro [Liu et al. 2014]. In addition, it has been shown to
interact with transcriptional regulatory factors such as the Nucleosome Remodeling and
Deacetylation (NuRD) complex, the RNA Polymerase Il-Associated Factor 1 (PAF1)
transcription elongation complex, and the rRNA transcriptional activator UBF [Todd &
Picketts 2012; Zhang et al. 2013; Wang et al. 2013]. To better understand the function of
PHF6 as a potential chromatin regulator and examine its lineage-specific roles in
hematological malignancies, we decided to thoroughly investigate its role in B-ALL.
Here, through integrated genomics and in vivo studies, we show that PHF6 regulates
the chromatin landscape in B-ALL cells, where it is responsible for maintaining a
chromatin state that enables a transformed pre-B cell identity. PHF6 controls the
transcriptional state of target genes by supporting a chromatin configuration that permits
or blocks the binding of lineage-specific transcription factors. Further, we show that the
associated transcriptional and chromatin state changes that occur in the absence of
Phf6 contribute to an emerging mechanism of drug resistance, termed pathway
indifference [Cooley et al. 2015]. Loss of Phf6 results in chromatin instability and
genomic plasticity that allows malignant cells to reprogram transcriptional outputs and
tolerate aberrant lineage signaling.
86
RESULTS
Loss of Phf6 decreases the leukemogenic potential of B-ALL cells and results in
the development of mixed-lineage lymphoma in vivo
Recent studies suggest that PHF6 can act as a lineage-specific regulator of
tumor growth. However, the molecular mechanisms underlying PHF6's function in
hematological malignancies remains largely unknown (Fig. 2.1A). To evaluate the
effects of complete loss of Phf6 on B-ALL growth, we engineered isogenic Phf6
knockout (Phf6KO) B-ALL cells using CRISPR-Cas9 [Ran et al. 2013; Sanchez-Rivera &
Jacks 2013; Soto-Feliciano 2016] (Supp. Fig. S2.1A-D). To maximize the chance of
functionally inactivating the PHF6 protein, we employed a domain-focused targeting
strategy by utilizing an sgRNA designed to target the second conserved plant
homeodomain (ePHD2), which is identified as a mutational hotspot in T -ALL and AML
[Shi et al. 2015; Van Vlierberghe et al. 2010; Van Vlierberghe et al. 2011] (Supp. Fig.
S2.1 B-D). The Phf6KO cells were extensively characterized with regard to their in vitro
growth properties, showing no significant differences in their proliferation rates and cell
cycle profiles (Supp. Fig. S2.1E-F) [Soto-Feliciano 2016].
To determine the effects of complete loss of Phf6 on B-ALL growth in vivo, we
performed syngeneic transplants into immunocompetent recipient mice. Tumor
formation in mice injected with 106 Phf6KO cells was significantly delayed compared to
mice transplanted with (1) Phf6WT cells, (2) cells expressing an shRNA targeting Phf6
(shPhf6), and (3) Phf6KO cells rescued with Phf6 cDNA (Fig. 2.1B-C; Supp. Fig. S2.1G-
H). When the number of transplanted cells was reduced 1000-fold, Phf6-deficient cells
failed to develop detectable leukemia in recipient animals (Fig. 2.1B). Thus, complete
loss of Phf6 reduces the fitness and leukemia-initiating capacity of B-ALL cells in vivo.
Notably, mice transplanted with Phf6wJT and Phf6KO cells developed malignancies
with pronounced differences. BCR-ABL1+ B-cell leukemia reproducibly presents with an
enlarged spleen and disseminated disease that infiltrates the bone marrow and
peripheral blood, as can be seen upon transplantation with Phf6WT cells [Williams et al.
2006]. However, tumors that developed from Phf6KO cells pathologically resembled
87
Bor essential, Tumor su
nknown unkn
3chanism mech
100-
ppressor,
own
anism
:3
W
2
_
a)
a-
50-
0
0
A----------------e
+phg - 103 (n=6)
+phfe - 103 (n=1 0)
I PhOW-10(n=14)
PhW0 -108 (n=18)
20 40
Time after transplantation (days)
Genotype of cellsinjected into mice
Phf6WT
shPhf6
Phf6<0
Phf6wT + cDNA
Phf6<O+ cDNA
Median survival
(days)
12
11
16.5
10
10
Significant by
log-rank?
No 0.1952
Yes < 0.0001 18
Yes 0.0162
Yes 0.0162
Figure 2.1. Complete loss of Phf6 reduces the fitness and leukemia-initiating potential of B-ALL cells in
vivo. (A) Representation of outstanding questions in the field regarding role of PHF6 in B-ALL vs T-ALL. (B) Kaplan-
Meier survival analysis of mice injected with either 103 (dotted) or 106 (solid) Phf6wr (blue) and Phf6KO (red) B-ALL
cells. The number (n) of mice per genotype analyzed is shown. Statistical analysis (log-rank test, Mantel-Cox) was
performed for the different groups in comparison to mice injected with Phf6wr cells. P value is shown for the
comparison. (C) Differences in median survival between mice injected with 106 B-ALL cells of different Phf6
genotypes. Statistical analysis (log-rank test, Mantel-Cox) was performed for the different groups in comparison to
mice injected with 106 Phf6wTr cells. [Soto-Feliciano et al. 2017; Soto-Feliciano 2016].
88
A
Tum
u
m
B-ALL T-ALL
C
p<O.0001
P<0.0001
60
Pvalue Number ofmice
14
10
5
5
lymphoma, showing prominent lymphadenopathy with marked blood vessels rooting
from and to the lymph nodes, which completely effaced the normal lymph node
histology in recipient mice (Fig. 2.2A-B). These tumors resulted in size and combined
organ weights that were more than 200% greater than those of wild-type B-ALL
recipients (Fig. 2.2B). Conversely, tumor burden in the blood and spleen size of mice
injected with knockout cells were significantly reduced relative to those from mice
injected with wild-type cells (Fig. 2.2C). These changes were confirmed in a second
isogenic Phf6KO B-ALL cell line (Supp. Fig. S2.2A-B). Moreover, reintroduction of Phf6
cDNA into Phf6KO B-ALL cells resulted in leukemia that was indistinguishable in latency
from that produced by Phf6WT B-ALL cells, ruling out off-target effects (Fig. 2.1C; Supp.
Fig. S2.1G-H) [Soto-Feliciano 2016]. All together, the complete loss of Phf6 in B-ALL
results in a drastic alteration of the natural ontology of the malignancy, changing from a
leukemic/disseminated disease to a lymphoma, and a consequential extension in
survival of recipient mice.
To further investigate the marked differences in disease presentation between
Phf6KO and Phf6WT cells, we performed immunophenotypic analyses on tumor cells
isolated from the lymph nodes (LNs) and bone marrow (BM) of recipient mice (Fig.
2.3A; Supp. Fig. S2.3A). As expected, Phf6wT tumor cells isolated from the BM and LNs
expressed the B-cell markers CD19 and B220, and lacked IgM [Williams et al. 2006]. In
contrast, tumor cells isolated from Phf6KO B-ALL recipient mice showed modest but
significant reduction of CD19 and B220 levels (Fig. 2.3B). Strikingly, these tumor cells
also had high levels of CD4 on the cell surface, indicating that a significant proportion of
Phf6KO B-ALL cells aberrantly expressed a surface marker characteristic of the T-cell
lineage (Fig. 2.3C) [Janeway et al. 1988; Zhu et al. 2010; Rothenberg et al. 2011].
Additional immunophenotypic profiling demonstrated that leukemia cells from both
genotypes did not express markers of additional lineages, including surface or
intracellular CD3 or the myeloid/dendritic cell marker CD11b [Borst et al. 1987; Wada et
al. 2008; Bell & Bhandoola 2008] (Fig. 2.3D-E). As observed previously, loss of Phf6
expression in leukemia cells is detrimental exclusively in the in vivo setting [Meacham et
al. 2015]. Consistent with this, Phf6KO cells cultured in vitro did not express CD4 (Supp.
Fig. S2.3B). Importantly, the changes in disease presentation and aberrant expression
89
A PhfI
0 0p
E
B LN fromPhf6WT
) 400-
~0
300-
c 20
100-
100-E00
Phf6WT
Phf6C< mass
F,
~%
LN from
Phf6C< C
-26
Phf6CO
80.
o 60.
t 240-
o) 20.
E
CD
C
0
0
Phf6W Phf6CO
400-
350
300-
Phf6"T Phf6K
Figure 2.2. In vivo loss of Phf6 results in a drastic alteration of the natural ontology of B-ALL. (A)
Representative haematoxylin and eosin (H&E) (top) and immunohistochemistry (bottom) staining of serial sections
from lymph nodes and lymphomas (mass) of recipient mice injected with Phf6wrr and Phf6KO cells. mCherry
immunochemistry demarcates tumor cells. LN, lymph nodes. All scale bars represent 600pm. (B) (Top) Size
comparison of representative lymph nodes from Phf6wr (left) and Phf6KO (right) recipient mice. (Bottom)
Quantification of combined lymph node weight of Phf6wr (blue, n=5) and Phf6KO recipients (red, n=5). (C) (Top)
Tumor burden in the blood of Phf6wr (blue, n=7) and Phf6KO (red, n=8) recipient mice. mCherry demarcates tumor
cells. (Bottom) Quantification of spleen weights of Phf6wr (blue) and Phf6KO (red) recipients (n=5, mean SD, two-
sided Student's t-test). Data represent the mean standard deviation (SD) in B-C. Statistics were calculated with
two-sided Student's t-test: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, n.s. = not significant. [Soto-Feliciano
et al. 2017; Soto-Feliciano 2016].
90
METRI 1 2
Phf6KO LNsh Phf6
btl Fj
CD19 B220
B-cell markers
Bone marrow
95.
90.
100.9
Phf6w' Ph6KO
n.s.
95-
0@
8&-
Phf6 Phf6KO
Bone marrowliz-
2 n.s
Phf6wT Phf6<O
i5-4n
2
n.s
Phfb" Phf6KO
U
.3
+
0
95.
90.
85,
CD3 CD4
T-cell markers
Lymph node
** i
00w U
U
Phf6wr Phf6KO
*
100.
95-
90-
85
Phfi6T
5.
-4.
S3.
+, 2-
01.
(a
C.,0
0
Cu
I
Lymph nodes
n.s
III
Phf6WT Phf6KO
4
2. n.s
PhfU"'
CD11b
Myeloid markers
10
0
a,
0
0
40-
20-
60-
40-
20-
Phf6KO
EA
.0
Bone marrow
Phfew Phf6KO
Lymph nodes
I
Phff-yT Phf6KO
5 Bofe marrow
*
2. I I
PhfdWa
5-
.4-
3.
S2-
PhOO
Phf6KO
Lymph nodes
n.s
II
Phff'T PhfKO
Figure 2.3. Phf6KO cells aberrantly express the T-cell surface marker CD4 in vivo. (A) Schematic depicting the panel of cell
surface markers used in immunophenotyping experiments testing for expression of B-cell, T-cell, and myeloid markers. (B)
Frequency of B220+ (top) and CD19+ (bottom) cells among mCherry+ leukemia cells from bone marrow (left) and lymph nodes
(right) of recipients injected with Phf6wr (blue, n=9) and Phf6KO (red, n=5) cells. (C) Bar graphs showing the percentage of the
CD4+ fraction among mCherry+ cells isolated from Phf6WT (blue, n=9) and Phf6KO (red, n=5) tumors in bone marrow (top) and lymph
nodes (bottom). (D) Frequency of surface (top) and intracellular (bottom) CD3+ cells among mCherry+ leukemia cells isolated from
bone marrow (left) and lymph nodes (right) of recipients injected with Phf6wT (blue, n=3) and Phf6KO (red, n=3) cells. (E) Frequency
of CD1 1 b+ cells among mCherry+ leukemia cells isolated from bone marrow (top) and lymph nodes (bottom) of recipients injected
with Phf6wvr (blue, n=3) and Phf6KO (red, n=3) cells. Data represent the mean SD in B-E. Statistics were calculated with two-sided
Student's t-test: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, n.s. = not significant. [Soto-Feliciano et al. 2017; Soto-Feliciano
2016].
91
A
B
C)
C'.)
C)
D
C
+
C
C
i
1
of CD4 require the complete loss of function of Phf6, as they cannot be recapitulated via
RNAi-mediated suppression of Phf6 [Meacham et al. 2015] (Fig. 2.1C, Supp. Fig.
S2.3C-E). To rule out potential effects associated with single cell cloning, we generated
a clonal line from the Phf6wT population and performed similar survival and
immunophenotyping experiments (Supp. Fig. S2.3F-G). As expected, mice transplanted
with the Phf6wr clone did not show a significantly different median survival compared to
Phf6wvT non-clonal cells. In addition, we did not observe CD4 overexpression in cells
isolated from the LNs and BM of mice injected with the wild-type clone, mirroring the
disease established by mice transplanted with Phf6wT non-clonal cells [Soto-Feliciano
2016]. Finally, we observe a significant reduction in CD4 expression upon expression of
Phf6 cDNA in Phf6KO cells (Supp. Fig. 2.3H). All together, these results indicate that
PHF6 is critical for B-ALL growth in vivo as a disseminated disease, and suggest a role
for PHF6 in maintaining a stable B-cell identity.
Phf6-deficient B-ALL cells have altered expression of genes involved in lineage
specificity
Intrigued by the underlying phenotypic plasticity observed in Phf6KO B-ALL
tumors, we sought to explore any potential transcriptional changes behind this
phenotype by performing RNA sequencing (RNA-Seq) (Fig. 2.4-2.5, Supp. Fig. S2.4).
Pairwise differential expression analysis between wild-type and knockout transcriptome
profiles showed that Phf6KO cells had a markedly distinct gene expression pattern when
compared to Phf6wT cells (Fig. 2.4A; Supp. Fig. S2.4A; Supp. Table Si). Differentially
expressed genes between Phf6KO and Phf6wT cells were significantly enriched in
pathways involved in cellular and developmental processes of lymphocytes, suggesting
that complete loss of Phf6 in B-ALL cells selectively affects expression of genes
involved in lymphoid differentiation and function (Fig. 2.4B). Gene set enrichment
analysis (GSEA) of pairwise comparisons revealed that Phf6KO cells had a significant
decrease in the expression of genes associated with B-cell development and function,
specifically NF-kB/TNFa, RUNX1, BCR-ABL, and STAT5 signaling pathways
[Subramanian et al. 2005; Mori et al. 2008; Hoffmann et al. 2002] (Fig. 2.4C-D; Supp.
Table S1-S2). Importantly, the expression of many of the genes that formed the leading-
92
A Phf6wT Phf6CO B
Developmental process
Cell differentiation
Regulation of signal transduction
Cell surface receptor signaling
Regulation of immune system
Inflammatory respons
Leukocyte activation
Regulation of leukocyte migration
Leukocyte mediated immunity
0 2 4 6 8 10
-log(P value)
-1.0 -0.5 0.0 0.5 1.0 1.5
Standardized log2(Expression)
CNF-B/TNFasignaling UP D Pre-BI lymphocyteUP
a RUNX1 targetsDN 0.1 NES . -1.7484
BCR-ABL snangUP 0.001
-1.4 STATS signalingUP -D -
0 B-cell developmentUP E -0.4
0 -1.55
W -1.6
.7U p in Phf9( 0  Down in Phf6<O
E * PhU+EV
2.5 ***_**_h__ cN
U Phi i + EV
2.5- Ph6KO +cDNA
Phf6 Cd22 Cd74 II4ra Lyn Ly86 +1k
Figure 2.4. Loss of Phf6 in B-ALL promotes a novel transcriptional program with downregulation of pathways
important for B-cells. (A) Heatmap showing differentially expressed (DE) genes (FC > 4, FDR < 0.05) in pairwise
comparisons between Phf6w (left) and Phf6KO (right) cells as determined by RNA-Seq. Each column represents a replicate
sample. The scale corresponds to row-wise standardized log2-transformed expression values for each gene. (B) Top Gene
ontology (GO) and Panther terms found to be enriched in Phf6KO cells. The P value for each term is plotted as -log1o(P
value). (C) Enriched GSEA gene sets in Phf6KO cells compared to Phf6wT cells. The plot represents NES against nominal P
values (-log) and the red line denotes threshold for significantly enriched gene sets integral for B-cell function (P < 0.05).
DN, down-regulated; UP, up-regulated. (D) GSEA plot depicting significant (P < 0.001) changes in pre-B lymphocyte
signature genes upon Phf6 deletion, as compared to Phf~'wT cells. NES, normalized enrichment score; FDR, false discovery
rate. (E) qPCR analysis of Phf6wT (blue) and Phf6KO (red) cells transduced with empty vector (EV, solid) or a vector
expressing Phf6 cDNA (cDNA, dotted). Relative mRNA levels for B-cell associated genes are shown: Phf6, Cd22, Cd74,
II4ra, Lyn, Ly86 and 81k. Data represent the mean standard deviation (SD). Statistics were calculated with two-sided
Student's t-test: *PO < 0.05, ** < Q.Q1, ***P < Q.001, ****p < 0.0001, n.s. = not significant. [Soto-Feliciano et al. 2017; Soto-
Feliciano 2016].
93
edge cluster of the B-cell development signature (Cd74, IL4ra, Lyn, Ly86, Bik) was
restored by introduction of the Phf6 cDNA, indicating that these changes in gene
expression are caused by the absence of Phf6 in the knock-out cells and not an off-
target effect (Fig. 2.4E).
Using Independent Component Analysis (ICA), we identified a statistically
significant gene signature specific to Phf6KO replicate samples (Fig. 2.5A). GSEA of the
Phf6KO-Specific signature revealed that Phf6KO cells have a notable enrichment of gene
sets associated with T-cell signal transduction and function, including CD4+ T-cell-
specific pathways [Jeffrey et al. 2006; Hutcheson et al. 2008; Hahtola et al. 2006;
Tokoyoda et al. 2009; Abbas et al. 2009] (Fig. 2.5A-C; Supp. Table S1-S2). These
analyses suggest that Phf6KO B-ALL cells adopt a transcriptional program similar to
CD4+ T-cells upon complete loss of Phf6. To directly compare murine CD4+ T-cells and
Phf6KO B-ALL cells, we applied ICA signature analysis to RNA-Seq data from Phf6wr,
shPhf6, Phf6KO B-ALL cells, and murine CD4+-single positive (SP) T-cells [Miyazaki et
al. 2015] (Fig. 2.5D). This analysis unveiled a comparable expression profile between
Phf6KO cells and CD4+ T-cells, which was markedly distinct from shPhf6 and Phf6fiW
cells (Fig. 2.5D-E). These results suggest that complete loss of Phf6 in B-ALL cells
promotes a transcriptional program that partially resembles CD4+ T-cells. Further, in
comparison to the transcriptomes of 12 hematopoietic cell subsets, the transcriptome of
Phf6KO B-ALL cells distinctly clusters with that of T-cells, diverging away from B-cell
subsets (Fig. 2.5F). It should be noted that the shPhf6 used in our previous study
affects in vivo proliferation of B-ALL cells, but it is unable to recapitulate the breadth of
transcriptional changes observed upon complete loss of Phf6 [Meacham et al. 2015;
Soto-Feliciano 2016] (Supp. Fig. S2.4A; Fig. 2.5E). All together, these findings further
suggest that PHF6 is critical for the maintenance of B-cell identity through the positive
transcriptional regulation of genes important for the B-cell state, and the negative
regulation of genes important for the T-cell state.
94
Ap=0.01
Independent components (ICs)
Na ws m ory CD4 T-ceflDN
Srnmory CI4 T-cel vs. B-mlDN
* CD4 T-w w. myebid_DN
* Na CD4 vs. PBME CD4 T-ol DN
. CD4 T-ol w. B-coLDN
* 0 CD4 T-cel vs. na B-colDN
S .Nalive CD4 T-ml vs. monocyteDN
0
6
D
, E
RNA-Seq data
I ndependent component analysis
(ICA)
*0* * * E . s
U . 6.0 . 0. .
* U U * U
* ** * *
* U aIs
Phfa00 -CD4+
specific
* U U
..
N 0
U ... - .
- --
U a U U .
* * U U
* U E U . . . . . .
,-,-1 M
1 2 I 4 5 omp e 10 11 12
Independent components (ICs)
PhV0 CD4 SP-Control Phf6#T shPhf6
.
-CD
0 1
Nominal -log(P value)
T-cell Signal Transduction_.UP
NES = 1.4391
p-value - 0.08
FDR - 03602
--i 
-
-D 
-
-
up in Phf6CO Down In PhfO
F
- .* 0
* gee S
* 
%
* .*
* 0
- 20
5* 0
Figure 2.5. Loss of Phf6 in B-ALL promotes a novel transcriptional program with upregulation of pathways important for T-cells. (A)
Schematic representation of ICA used to identify differential expression signatures in the integrated RNA-Seq dataset comprised of Phf6wT,
shPhf6, and Phf6KO cells. Hinton diagram represents ICA-derived signatures with columns denoting signatures and rows denoting samples. Colors
denote relative directionality of gene expression (red = up-regulation, green = down-regulation) and the size of each square represents the
magnitude of the contribution of each sample to the respective IC. Each signature is two-sided. IC2 identified a Phf6KO-specific gene signature(p=0.01, Mann-Whitney test). (B) Enriched GSEA gene sets in Phf8KO cells compared to Phf6wT cells. The plot represents NES against nominal P
values (-log) and the red line denotes threshold for significantly enriched gene sets (P < 0.05) integral for T-cell function. DN, down-regulated; UP,
up-regulated. (C) GSEA plot depicting (P = 0.08) enrichment in T-cell signal transduction signature upon Phf6 deletion, as compared to Phf6wT
cells. NES, normalized enrichment score; FDR, false discovery rate. Data represent the mean SD. (D) Hinton plot of ICA signatures derived from
integrated signature analysis of murine B-ALL cell transcriptomes in comparison to murine CD4+-SP cells. IC4 identified a gene signature that
highlights similarities between murine Phf6KO B-ALL cells and SP-CD4+ cells. (E) Heatmap of Phf6KO-CD4+ specific (IC4) gene signature
comparing B-ALL cells of different genotypes with murine SP-CD4+ cells. Each column represents a replicate sample. The scale corresponds to
row-wise standardized log2-transformed expression values for each gene. (F) Principle component analysis was performed on RNA-Seq data from
hematopoietic cell subsets listed, Phf6wT (blue circle) and Phf6KO (red circle) B-ALL cells. Data was assembled by the ImmGen consortium [Heng
et al. 2008].SC, stem cell. MLP, multi-lineage progenitor. preB, precursor B-cell. proB, precursor B-cell. B-1a, mature B-cell, produce circulating
IgM antibodies. B-1b, mature B-cell, produce circulating IgM antibodies. preT, precursor T-cell. T, single positive T-cell. Tgd, T-cell gamma delta.
MF, macrophage. NKT natural killer T-cell. [Soto-Feliciano et al. 2017; Soto-Feliciano 2016].
95
CD4-P.Control 1
CD4-8P_COntrMl 2 -
CD44P_Cntrol 3
D4-SP-Control 4
c04-P.contro 5
iPhd-r1 -
PhWi-r2-
ShPh-r2 -
Fwo-r3
Phf6w-rl
Phf6"T-r2
shPhf6-rl
shPhf6-rl
shPhf6-r2
Phf6KO-rl
Phf9( 0-r2
Phf6C0-r3
B
-1.4'
-1.6.
E
2>'a
j z
C
*0.5
~0.4
0.3
E 0.2
0.1
So.o
B
Bla
Bib
MF
MLP
NKT
PHF6_KO
PHF6_Wr
preB
preT
pro
sc
T
Tgd
Global genomic binding profiles suggest that PHF6 does not bind in a sequence-
specific manner
We next sought to determine if the marked differences in gene expression
profiles between knockout and wild-type cells were due to chromatin regulation by
PHF6. To gain insight into the distribution of PHF6-bound sites across the B-ALL
genome, we performed chromatin immunoprecipitation-sequencing (ChIP-Seq) with an
antibody against PHF6 [Lee et al. 2006] (Fig. 2.6; Fig. 2.9). We found that PHF6 binds
to both gene bodies and proximal promoter/enhancer regions of many annotated genes
(Fig. 2.6A; Supp. Table S1, S4). We observed that genes that are differentially
expressed in knockout cells had more than 2-fold enrichment of PHF6 binding at their
transcriptional start sites (TSSs) (Fig. 2.6B). In addition, we find that there is enrichment
of PHF6 binding within the gene body of lowly expressed genes in Phf6wT cells (Fig.
2.6C). Together, these data suggest that the genomic location of PHF6 binding might
influence the transcriptional output in B-ALL cells.
Previous studies have proposed that PHF6 has the ability to bind both DNA and
histone proteins [Liu et al. 2014; Todd & Picketts 2012]. Given these two very different
binding specificities, we decided to examine both modalities, first asking whether PHF6
has any DNA sequence-specific binding properties in B-ALL cells. To do this, we
performed motif enrichment analysis of the DNA sequences surrounding the PHF6
binding peak summits near promoters of all genes and differentially expressed genes
[Bailey & Elkan 1994] (Fig. 2.6D). We found that no significant de novo motifs could be
identified for PHF6 itself (Fig. 2.6D, left column), and that the motifs that passed the
significance threshold bear resemblance to motifs bound by known transcription factors
(TFs) (Fig. 2.6D, right column). Similar results were obtained from a motif search of
genome-wide PHF6 binding peaks (Fig. 2.6F). This is consistent with previous studies in
which the second ZaP domain of PHF6 was found to bind double stranded DNA
nonspecifically [Liu et al. 2014].
PHF6 does not act in a transcription factor complex
Given that PHF6 lacked any sequence-specific binding properties, we next
investigated whether PHF6 is part of a transcription factor complex that could potentially
96
EC
C
4)0)
0
0O
A
D
-5 -4 -3 -2 -1 0 1 2 3 4 5
Distance from TSS (kb)
Differentially expressed (DE)
Unchanged
Genomic
De novo De novo Motif Most similar
motif rank motif matches known motif
F De novo De novo motif
1
2
3
4.
4
5In
15 (NHLH1)DBD
235 MA0132.1 (Pdxl)
12 MA0144.2 (Stat3)
Pvalue
2.1467E-04
5.5544E-04
6.5821 E-05
1 MA0528.1 (Znf263) 5.9297E-05
Most similar
known motif
MA0528.1
(Znf263)
Asc2 DBD
MAsc.2)
MA0095.2
(Yy1)
Motif
'
P value
-2000 TSS 33% 66% TrS 2000
Genomic region (5'-3')
E
PU.1
NF-KB
EGR1 J
EBF1
TCF3
TCF12 - l
0 5 10 15 20
-log(P value)
1.1219E-04
7.9361 E-05
2.6042E-05
IP Input IP Input 'P
-EtBr W :LAI B:
+-EtBrPHF6
IgG TCF12 IgG NF-kB IgG PHF6
Figure 2.6. PHF6 does not behave as a canonical transcriptional factor nor In a transcription factor complex. (A) Pie chart showing
the distribution of 77,749 PHF6-binding sites across genomic regions in B-ALL cells. TSS, transcription start site; TTS, transcription
termination site; UTR, untranslated region. (B) Metagene analysis of PHF6 binding fold enrichment plotted for DE (red), unchanged (green)
and all genes (black), assessed at TSSs 5kb genomic regions of DE genes. (C) Promoter-Gene Body plot for PHF6 in B-ALL cells
illustrates increasing levels of PHF6 occupancy in genes that are DE when Phf6 is lost. Lines represent average profiles for five gene groups
ordered by mRNA expression levels by RNA-Seq, defined as "high", "medium", "low", "differentially expressed" and "unchanged". Promoter,
-2kb from TSS; TSS, transcription start site; TTS, transcription termination site; 33%-66%, 1/ and % of gene body, respectively. (D) De novo
DNA sequence motifs identified in PHF6-bound regions at promoters of DE genes with associated P values. Shown are sequence logos of
de novo position-weight matrices found by the MEME motif discovery tool (left) or that of known transcription factors whose motifs are found
to be most similar to the de novo motif discovery results by the TOMTOM software (right). (E) Putative co-binding partners of PHF6 predicted
by TF binding motifs present at TSSs 1kb genomic region of DE genes in Phf6KO cells with strong PHF6 ChIP-Seq signal. Sequences
obtained from PWMEnrich were randomized and significant (P < 0.01) TFs were selected. (F) De novo DNA sequence motifs identified in
PHF6-bound regions at top genomic peaks with associated P values. Shown are sequence logos of de novo position-weight matrices found
by the MEME motif discovery tool (left) or that of known transcription factors whose motifs are found to be most similar to the de novo motif
discovery results by the TOMTOM software (right).(G) Endogenous co-immunoprecipitation (co-IP) assay of TCF12, NF-KB, and PHF6 in the
absence (top) or presence (bottom) of ethidium bromide (EtBr). (Cell lysates from Phf6WT +cDNA lines). Input is 7% of IP lysate. [Soto-
Feliciano et al. 2017; Soto-Feliciano 2016].
97
Promoter-TSS = 2.6%
5'-UTR = 6.4%
Exon = 14.2%
Intron =42.7%
3'-UTR = 4.2%
TTS = 0.5%
Intergenic = 29.4%
G Input
co-bind and co-regulate gene expression in B-ALL cells. To do this, we obtained mouse
sequences for the 1 kb regions adjacent to the TSSs of differentially expressed genes
and assessed the enrichment of known motifs (those with P value < 0.01 that also have
strong PHF6 signal in ChIP-Seq) [Stojnic et al. 2016]. Ultimately, 28 motifs
(corresponding to 24 unique TFs) were identified as binding sites of putative interacting
partners of PHF6 (Supp. Table S1, S4). These motifs were enriched for binding sites of
transcription factors important during hematopoiesis and lymphopoiesis, including motifs
for PU.1, NF-kB, EGR1, EBF1, E2A (Tcf3), and HEB (Tcf12) [Schebesta et al. 2002;
Wildt et al. 2007, Wang et al. 2008; Carlson et al. 2006; Luo et al. 2004; Arenzana et al.
2009; Soto-Feliciano 2016] (Fig. 2.6E; Supp. Table S1, S4). Through endogenous co-
immunoprecipitation (co-IP) experiments, we confirmed that PHF6 interacts with TCF12
and NF-kB in B-ALL cells (Fig. 2.6G, top). However, these interactions are mediated, in
most part, via binding to DNA rather than via direct protein-protein interactions, as
demonstrated by a notable reduction in immunoprecipitated protein upon addition of the
DNA intercalating agent ethidium bromide (EtBr) (Fig. 2.6G, bottom). Together, these
results show that PHF6 is unable to recognize specific DNA binding motifs/sequences
and is not interacting directly with TCF12 or NF-kB. Therefore, PHF6 is not acting as a
canonical transcription factor in B-ALL cells.
We next wanted to investigate the relationship between PHF6 and the DNA-
dependent TF interactors, TCF12 and NF-kB, by modulating the expression levels (via
overexpression of knockdown) of each TF. First, we examined whether the TFs could
compensate for the loss of Phf6. To do this, we overexpressed the cDNA of Tcfl2 and
Nfkb in B-ALL cells and studied the effects in vivo (Fig. 2.7A). A block in differentiation is
a hallmark of leukemia cells [Mullighan et al. 2007]. As seen in Phf6wT B-ALL cells
transduced with Tcf12 or Nfkb cDNA, the overexpression of transcription factors that
drive B-cell differentiation are not well-tolerated, as seen by the percentage of cells that
have silenced the vector (%GFP+) after sorting for pure populations (Fig. 2.7B, blue).
On the other hand, the plasticity previously observed in Phf6KO B-ALL cells renders
them more permissive to such signaling (Fig. 2.7B, red). Previously, we showed that
expression of an exogenous Phf6 cDNA rescues (1) the differences in survival of Phf6KO
cells; (2) the expression of B-cell target genes; and (3) significantly reduces the levels of
98
Lrl
Compensation?
30-
020
10
0 5 10 15 20
Time after transplantation (days)
-. Phf6wr MSCV-EV (n=5)
PhISKO MSCV-EV (n=6)
n.s. Ph6wM MSCV-TCF12 (n=6)
Phf6Ko MSCV-TCF12 (n=6)
F
40-
4 30-
20-
10-
0 5 10 15 20
B Phf6Sta
EVv
tus
TCF12 WT_
cDNA KO.
NFKB WT
CDNA |KO
BTG21 WT ------
cDNA Ko R RRRRR9
n.s. - n.s.
-* *n.s.
BM LN
U
I
SP
*99.6
*97
55
77.1
64.70
79.9
77.6
72
O %GFP+
- n.s.
U
THY
13 Ph6Wr MSCV-EV (n=5)
3 Phf6KO MSCV-EV (n=6)
C] Phf65 MSCV-TCF12 (n=5)
C Ph6Ko MSCV-TCF12 (n=5)
n.s. n.s. n.s. n.s.
n.s. ** - n.s. n.s.
A
A
BM I NI
U
THV
.4
Time after transplantation (days) C3 "' S*1*
Phf6w MSCV-EV (n=5) C3 Phwr5 MSCV-EV (n=5)n.s. PhKO MSCV-EV (n=6) PhMKo MSCV-EV (n=6)
** ' Phf6wrMSCV-NFkB(n=5) I| PhtiKOMSCV-NFkB(n=5)
PhISKOMSCV-NFkB(n=5) 0 PhSKOMSCVNFkB(n=5)
Figure 2.7. Overexpression of transcription factors that Interact In a DNA-dependent manner with PHF6 does not rescue the in
vivo Phf6KO phenotype. (A) Schematic showing a possible mode of gene expression activation by the DNA-dependent binding of
several transcription factors upstream of a "B-cell identity gene." Overexpression of the interactors TCF12 and NFKB was tested to see if
they could compensate for the loss of PHF6. (B) Percentage of cultured cells expressing indicated MSCV-GFP-cDNA vector in Phf6wT(blue) and Phf6KO (red) cells upon injection into mice. (C) Kaplan-Meier survival analysis of mice injected with 108 Phf6wIr (blue) or Phf6KO
(red) B-ALL cells infected with a vector expressing empty vector (MSCV-EV, solid) or Tcf12 cDNA (MSCV-TCF12, dashed). The number
(n) of mice per genotype analyzed is shown. Statistical analysis (log-rank test, Mantel-Cox) was performed for the different groups in
comparison to mice injected with Phf6wT MSCV-EV cells. (D) Bar graphs showing the percentage of the CD4+ fraction among mCherry+
cells isolated from tumors from Phf6T (blue, n=9) and Phf6KO (red, n=5) cells infected with a vector expressing empty vector (MSCV-EV)
(unfilled) or TCF12 cDNA (MSCV-TCF12) (pattemed). The number (n) of mice per genotype analyzed is shown. (E) Kaplan-Meier
survival analysis of mice injected with 106 Phf6wT (blue) or Phf6KO (red) B-ALL cells infected with a vector expressing empty vector
(MSCV-EV, solid) or NfKb cDNA (MSCV-NFKB, dashed). The number (n) of mice per genotype analyzed is shown. Statistical analysis
(log-rank test, Mantel-Cox) was performed for the different groups in comparison to mice injected with Phf6wT MSCV-EV cells. (F) Bar
graphs showing the percentage of the CD4+ fraction among mCherry+ cells isolated from tumors from Phf6wT (blue, n=9) and Phf6KO
(red, n=5) cells infected with a vector expressing empty vector (MSCV-EV) (unfilled) or NfKb cDNA (MSCV-NFKB) (patterned). The
number (n) of mice per genotype analyzed is shown. BM, bone marrow; LN, lymph node; SP, spleen; THY, thymus. Data represent the
mean SD in D,F. Statistics were calculated with two-sided Student's t-test: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, n.s. = not
significant.
99
A
TF
Expression: X ++ ++
rter
C
100.-
=
D401 -****
E
100.
a_
n - - s N
-A
SP
A
V
A
IS,
| I
CD4 expression (Fig. 2.1C, Fig. 2.4E, Supp. Fig. S2.1G-H, Supp. Fig. S2.3H). No
significant differences in survival are observed upon injection of wild-type and knockout
cells transduced with Tcfl2 cDNA into syngeneic recipients, as compared to cells
transduced with an empty vector (EV) (Fig. 2.7C). Further, overexpression of Tcfl2
cDNA does not lower the surface CD4 expression levels in the spleen or thymus of
Phf6KO cells (Fig. 2.7D). Similarly, we observed the same trend for cells transduced with
Nfkb cDNA. Although disease onset is accelerated in mice transplanted with Phf6KO
MSCV-NFkB cells to the same extent as Phf6wr cells, the levels of CD4 are
indistinguishable or greater than that from tumors arising from Phf6KO cells (Fig. 2.7E-F).
We next wanted to test the effects of knocking down the expression of the DNA-
dependent TF interactors. To do this, we transduced Phf6wT and Phf6KO cells with
hairpins targeting Tcfl2 or NfKb and achieved significant reduction in the expression of
each gene (Fig. 2.8A-B). While neither Tcfl2 nor NfKb knockdown had a significant
effect on the survival of mice injected with Phf6KO cells, the loss of NfKb expression was
favorable for Phf6KO disease, showing dramatically increased levels of CD4 expression
in the bone marrow, lymph nodes, spleen, and thymus (Fig. 2.8E). This suggests that
loss of NfKb expression increased the number of cells that could undergo conversion to
a T-cell-like state, supporting a model whereby decreased dosage of TFs can promote
cellular plasticity. Surprisingly, recipients of shTcf12 or shNfKb Phf6wT cells did not
succumb to disease during the experimental time course, suggesting that loss of
expression of these factors may be detrimental to B-ALL progression (Fig. 2.8 C-D).
This data demonstrates that the loss of Tcfl2 or Nfkb expression is advantageous for
the Phf6KO mixed-lineage disease state, but is disadvantageous for Phf6wT leukemia.
Through the modulation of TF expression levels, we show that the presence of PHF6 at
target genes appears to be essential for the regulation of gene expression and cannot
be rescued by the overexpression of other transcription factors, like Tcfl2 or Nfkb.
Further, the plasticity observed in Phf6KO cells can be amplified by loss of additional B-
lineage transcription factors. Together, this data led us to hypothesize that PHF6 is
acting prior to the binding of canonical TFs, perhaps by playing a role in chromatin
modulation.
100
Bn.s.
* ~
**Phf6wr
PhI6wr+miRE_TCF!2
PhKO
Phi6KO+rmiRE_TCF12
Phf6KO
z
E
Z
W
X3
1.51
1.0.
0.5
0.01
* Phl6WT
* Phf6w+ miRE p65
* PhfKO
* Phf6KO + miREp65
T
Phf6V7
D
-a
0)
0~
100- --- - -- - -
50
0,
. 5 10 15
Time after transplantation (days)
n.s. - -
n.s. -
20 2o
PhEr+shControl (n=5) n.s.
PhISKO+shControl (n=4)
PhSwr+shNFkB (n=10)
Phf6KO+shNFkB (n=9)
Phf6WT
C
100
50.
0 5 10 15 20 25
lime after transplantation (days)
n.s. Pht6w+shControI (n=5) n.s-I-- PhI6KO+shControl (n=4)
n.s. - Phf6wr+shTCF12 (n=10)
I PhIBKO+shTCF12 (n=5)
n.s. - n.s. n-s.
A
LN SP
I
=n .s.
A=
C3i
Phf6wr+shControl
Phf6KO+shControl
Phf6KO+shTCFI 2
Phf6KO+shNFkB
LI
THY
Figure 2.8. Hairpin mediated knockdown of TCF12 pushes Phf6KO B-ALL cells further into a T-cell like state.(A) Relative Tcf12 mRNA levels in Phf6wr (blue) and Phf6KO (red) cells transduced with a hairpin against Tcfl2.(B)
Relative NfKb mRNA levels in Phf6Wr (blue) and Phf6KO (red) cells transduced with a hairpin against NfKb (p65). (C)
Kaplan-Meier survival analysis of mice injected with 106 Phf6wr (blue) and Phf6KO (red) B-ALL cells transduced with
a vector expressing a control hairpin (solid) or hairpin targeting Tcf12 (dashed). (D) Kaplan-Meier survival analysis
of mice injected with 106 Phf6wr (blue) and Phf6KO (red) B-ALL cells transduced with a vector expressing a control
hairpin (solid) or hairpin targeting NfKb (dashed).(E) Bar graphs showing the percentage of the CD4+ fraction among
mCherry+ cells isolated from tumors from Phf6wT (blue, n=9) and Phf6KO (red, n=5) cells infected with a vector
expressing a control hairpin or a hairpin targeting Tcfl2 or NfKb. The number (n) of mice per genotype is shown.
BM, bone marrow; LN, lymph node; SP, spleen; THY, thymus. Data represents the mean SD in A-B,E. Statistics
were calculated with two-sided Student's t-test). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, n.s. = not
significant.
101
A
1.51
1.0
0.5.
0.0--
z
E
a)
E
60-
40-
20- ImmIf
BM
Global genomic binding profiles suggest that PHF6 interacts with chromatin
Next, we tested whether PHF6 can regulate gene expression by interacting with
histones, and subsequently altering the chromatin state. In fact, we observed that the
PHF6 binding profile is positively correlated with gene expression (Fig. 2.9A, left).
Specifically, PHF6 binding is enriched at regions flanking the TSS of highly expressed
genes, similar to H3K27ac signal location (Fig. 2.9A, right). Conversely, PHF6 signals
tend to be more centered directly at the TSS of genes with lower expression (Fig. 2.9A,
left). In addition, we found that PHF6 binding enrichment is positively correlated with the
presence of activating histone post-translational modifications, including H3K27ac and
H3K4me3, whereas low correlation is observed with the repressive mark H3K27me3
(Fig. 2.9A (right), Fig. 2.9B) [Soto-Feliciano 2016]. Previously, it was demonstrated that
PHF6 can interact with components of the Nucleosome Remodeling and Deacetylation
(NuRD) complex, as well as with several histone proteins (H1.2, H2B.1, H2A.Z, H3.1)
[Todd & Picketts 2012]. Therefore, we decided to investigate whether these interactions
also occur in B-ALL cells. Through endogenous co-IP experiments, we were unable to
confirm interactions between PHF6 and components of the NuRD complex: RBBP4,
HDAC1, or HDAC2 in B-ALL cells (data not shown). However, we were able to confirm a
strong interaction between PHF6 and histone H3 (Fig. 2.9C, top). Importantly, this
interaction is independent of the presence of DNA, as treatment of cell lysates with EtBr
did not disrupt the association of PHF6 with histone H3 (Fig. 2.9C, bottom). All together,
these results show that PHF6 exerts transcriptional control of target genes via direct
protein-protein interactions with histones. We conclude that the interaction between
PHF6 and chromatin has important influences on transcriptional regulation, and that the
loss of Phf6 in B-ALL cells thus results in large transcriptional changes and the
associated phenotypic plasticity in vivo.
Loss of PHF6 results in focal changes in chromatin accessibility
Lineage-specific transcriptional programs are often characterized by differential
chromatin accessibility [Lara-Astiaso et al. 2014]. To determine if the changes in
transcriptional programs and disease presentation that we observe are the result of
altered chromatin organization, we performed an assay for transposase-accessible
102
PHF6
-500 0 500
Distance from TSS (bp)
C4
0~
10I0
H3K27ac
-1000 -500 0 500
Distance from TSS (bp)
Highly Expressed
Genomic
Lowly Expressed
C
V
.... ...
Input IP
-EtBr
+EtBr
IgG H3
-5 -4 -3 -2 -1 0 1 2
Distance from TSS (kb)
H3K27ac
- H3K4me3
H3K27me3
-- Diff. Expressed
-
Genomic
Figure 2.9. PHF6 exerts transcriptional regulation by interacting with histones rather than binding
sequence-specific DNA sites. (A) (Left) Metagene tracks of PHF6 ChIP-Seq signal averaged over all promoter-
TSS tracks grouped by relative expression levels (red-high, green-genomic, purple-low). (Right) Metagene track of
H3K27ac ChIP-Seq signal averaged over all promoter-TSS regions. (B) Metagene tracks of PHF6 ChIP-Seq signal
and correlation with histone marks: H3K27ac (blue), H3K4me3 (yellow, GSE66234), H3K27me3 (green). Pearson
correlation of PHF6 and histone ChIP-Seq signals across 10kb regions spanning the TSS. Differentially expressed
(solid line) and genome-wide (doffed line) genes are shown. Shaded regions around average tracks denote
estimates of 95% confidence interval (CI) of the metagene average signals based on resampling. (C) Endogenous
co-IP assay of histone H3 and PHF6 in the absence (top) or presence (bottom) of ethidium bromide (EtBr). Input is
7% of IP lysate. [Soto-Feliciano et al. 2017; Soto-Feliciano 2016].
103
A
C-
00
C
a)E
M.C N
CM
0
-1000 1000
B
C
0
00
C
a-,
00
0
a' .
n-
*.
WB:
PHF6
345
chromatin with high-throughput sequencing (ATAC-Seq) on Phf6wT and Phf6KO B-ALL
cells (Fig. 2.10-2.14) [Buenrostro et al. 2013]. Phf6 loss results in focal changes in
chromatin accessibility, with approximately 700 regions in the genome showing
statistically significant differences (Fig. 2.10A-B, Supp. Table S1). Further, juxtaposition
of PHF6 binding by ChIP-Seq with chromatin accessibility data from ATAC-Seq shows
that PHF6 binds in two distinct manners: (1) to regions of open chromatin at the
transcription start site (Fig. 2.10C, top); and (2) flanking regions of open chromatin at
the TSS (Fig. 2.10C, bottom). This analysis also demonstrated that changes in
chromatin accessibility at promoter regions are predictive of changes in gene
expression (Fig. 2.11A). Consistent with our phenotypic data, differential chromatin
accessibility analysis revealed that regions that become less accessible upon Phf6 loss
are enriched for binding motifs of TFs required for B-cell development, such as EBF1,
PU.1, E2A, and STAT5 [Schebesta et al. 2002; Dai et al. 2007; Niebuhr et al. 2013] (Fig.
2.11B, left). Conversely, regions that became more accessible upon Phf6 loss are
enriched for motifs of TFs associated with the development of T-ALL and AML, like
ETS1 and ERG [Smeets et al. 2013; Thoms et al. 2011; Eyquem et al. 2004; Kiaii et al.
2013] (Fig. 2.11B, right). Thus, genomic regions important for B- and T-lineage genes,
as well as their regulatory factor binding sites, undergo significant changes in chromatin
accessibility upon complete loss of Phf6 in B-ALL cells [Soto-Feliciano et al. 2017]. This
further explains the inability of Tcfl2 cDNA to rescue the phenotypic changes observed
in Phf6KO cells, due to the TCF12 motif becoming inaccessible after loss of Phf6 (Fig.
2.7, Fig. 2.11B, left). It should be noted that the chromatin accessibility of a TF motif is
precisely regulated in specific cell types [Corces et al. 2016]. Thus, the shifts in
accessibility that we observe in B- and T-cell TF motifs demonstrates a genuine change
in cellular state and disease state.
The most significant changes in chromatin accessibility occur at genes crucial
for lineage specification
Specifying a strict significance threshold on the chromatin accessibility dataset
revealed 29 transcriptional start sites undergoing chromatin compaction or
decompaction (Fig. 2.12A; Supp. Table S5). Notably, the TSSs of two T-cell receptor
104
A hf6w Phfe<OC ATAC Signal PHF6-ChlP
CO,
-0
iR
I I ~ -1kb TS +1kb -1kb TSS +1kb
10g2(ATAC-Seq Signal) Min Max Min Max
B
S4.76% Promoter-TSS
- 3.92% Exon
- 49.30% Intergenic
- 41.18% Intron
Figure 2.10. PHF6 has distinct binding patterns that result in significant changes in chromatin accessibility.
(A) Heatmap showing changes in chromatin accessibility between Phf6fl' and PhfOKO cells. Shown are differentially
accessible region with a fold change of at least 2, as determined using the DESeq2 R package. Each column
represents a replicate sample. The scale corresponds to log2-transformed ATAC-Seq signal for each region. (B)
Distribution of genomic regions undergoing significant changes in chromatin accessibility in Phf6K(O B-ALL cells.
TSS, transcription start site. (C) Density maps of ATAC-Seq and PHF6 ChIP-Seq data from Phf6wr cells, ordered by
ratio of PHF6 signal at TSSs 1ikb over PHF6 signal at TSS flanking regions. Scales correspond to log2-transformed
signal. Vertical axis indicates the peak's rank and horizontal axis shows the position relative to the TSS. [Soto-
Feliciano et al. 2017].
105
DEG Down
p=1.75e-21
Phf6vT Phf6<0
DEG UD
p=2.84e-27
Phf6WT Phf6<0
EBF1-l1 ETS1 distal.
ETS1
ELK1
ERG
IiFL
301o0 2vu 0)
-log(P value)
Up Motifs
I
1.
0 10 20 30 40
-log(P value)
Figure 2.11. Loss of PHF6 results in focal changes in chromatin accessibility at DNA-binding sites for
lineage-specific transcription factors. (A) Violin plots of ATAC-Seq signal in the promoter regions of differentially
expressed genes (DEG) between Phf6wr and Phf6KO cells (DEGDown, differentially expressed genes
downregulated in Phf6KO, left; DEGUp differentially expressed genes upregulated in Phf6KO, right). P values are
shown as returned by the Wilcoxon rank-sum test. (B) Enrichment motif analysis for DNA regions that undergo
significant changes in chromatin accessibility in Phf6KO cells. DNA motifs with decreased chromatin accessibility
(top) and DNA motifs with increased chromatin accessibility (bottom) upon loss of Phf6. [Soto-Feliciano et al. 2017].
106
A
05
C,
0)00<
B Down Motifs
PU.1-
TCF3 (E2A).
STAT1
TCF12 (HEB)-
STAT5-
RUNX1-
EZZZZZJ
[
6
genes, Trbjl-1 (T-cell receptor beta joining 1-1) and Trav2 (T-cell receptor alpha variable
2), undergo increased chromatin accessibility upon loss of Phf6. Conversely, the TSS of
a B-cell related gene, B cell translocation gene 2 (Btg2), experiences significant
chromatin compaction (Fig. 2.12A). Recently, Btg2 has been shown to play an important
role in promoting murine B-cell differentiation. Mice deficient in this gene have reduced
numbers of B220+ B-cells, and additional loss of the related gene Btgl results in co-
expression of CD4 on B220+ B-cells [Tijchon et al. 2016]. Spurred by the similarity of
aberrant T-cell marker expression on B-cells that occurs in Btg1-1- Btg2-'- mice to that of
Phf6KO B-ALL cells, we decided to explore the effect of overexpression of Btg2 in our B-
cell leukemia model (Fig. 2.12B-C). Overexpression of an exogenous Btg2 cDNA in
Phf6WT and Phf6KO B-ALL cells was not readily-tolerated (Fig. 2.7B). Transplantation of
Phf6KO MSCV-BTG2 transduced cells into immunocompetent recipient mice resulted in
a disease latency equivalent to that of Phf6WT MSCV-EV (empty vector), both with a
median survival of 12-days post-transplantation (Fig. 2.12B). Despite the changes in
disease onset, the levels of CD4 expression on Phf6KO MSCV-BTG2 tumors were
sustained, with expression levels greater than or equal to Phf6KO MSCV-EV tumors (Fig.
2.12C). Tumors resulting from Btg2 overexpression followed the same trend as tumors
developing from NfKb overexpression (Fig. 2.7E-F). Once more, we demonstrate that
loss of Phf6 cannot be rescued by the expression of a single transcription factor, but
rather points to a much broader role for PHF6 in transcriptional regulation.
PHF6 plays a vital role in nucleosome positioning
We next asked if nucleosome positions, in addition to chromatin accessibility,
were undergoing significant changes dependent on Phf6 status. To determine
nucleosome positioning with high precision and resolution, we applied the NucleoATAC
algorithm to our dataset [Schep et al. 2015] (Fig. 2.13A-B; Fig. 2.15A). Ranking
nucleosome position by gene expression, we see that there is a marked depletion of
nucleosomes at the TSS with high signal flanking the TSS corresponding to the +/-1
nucleosomes, as expected for highly expressed genes (Fig. 2.13A). Focusing on the
ChIP-Seq profile of PHF6, we can see that there is a pronounced depletion of PHF6
binding at the TSS of the most highly expressed genes (Fig. 2.13A, right) and signal
107
A Phf6wT Phf6<O
Rasa12
Syti
Usp53
Gm9949
Adrb2
Nkd1
281041OL24Rik
Btg2
Sec24d
Creb312
Heyl
Map3k9
Ctbp2
Nespas
Fam20c
Chst7
Jagi
Zc3hl2b
Anol0
Rasilb
Rap2b
Lrrc27
Stk32c
Trav2
Tnnt2
Zfp972
Trbjl-1
Plxdc2
B
1 UL I I
C
0)
0
0)
50
0
n ..
n*s* 
-
1I I
I I
I I
I I
I [LI
5 10 15 20
Time after transplantation (days)
Phf6Nr MSCV-EV (n=5)
PhI6KO MSCV-EV (n=6)
PhEw MSCV-BTG2 (n=5)
PhI6KO MSCV-BTG2 (n=9)
I0.0
lOg2(ATAC-Seq Signal)
--- nas5001 ---- *
1.5
n.s. n.s. -* n.s.
n.s. n.s.
U
A~
U
I~I
Phi6w MSCV-EV (n=5)
PhI6KO MSCV-EV (n=6)
Ph6wr MSCV-BTG2 (n=5)
PhI6KO MSCV-BTG2 (n=5)
BM LN SP THY
Figure 2.12. Expression of Btg2 cannot rescue CD4 expression in vivo. (A) Heatmap showing the most
significant changes (fold change > 4) in chromatin accessibility at the promoters of corresponding genes between
Phf6w' and Phf6KO cells. Each column represents a replicate sample. The scale corresponds to log2-transformed
ATAC-Seq signal for each region. The blue arrow indicates the gene Btg2, used in further analysis. (B) Kaplan-Meier
survival analysis of mice injected with 106 Phf6wr (blue) or Phf6KO (red) B-ALL cells infected with a vector expressing
empty vector (MSCV-EV, solid) or Btg2 cDNA (MSCV-BTG2, dashed). The number (n) of mice per genotype
analyzed is shown. Statistical analysis (log-rank test, Mantel-Cox) was performed for the different groups in
comparison to mice injected with Phf6wr MSCV-EV cells. (C) Bar graphs showing the percentage of the CD4+
fraction among mCherry+ cells isolated from tumors from Phf6wr (blue, n=9) and Phf6KO (red, n=5) cells infected with
a vector expressing empty vector (MSCV-EV) (unfilled) or Btg2 cDNA (MSCV-BTG2) (patterned). The number (n) of
mice per genotype analyzed is shown. BM, bone marrow; LN, lymph node; SP, spleen; THY, thymus. Data
represents the mean SD in A-B,E. Statistics were calculated with two-sided Student's t-test). *P < 0.05, **P < 0.01,
***P < 0.001, ****P < 0.0001, n.s. = not significant. [Soto-Feliciano et al. 2017].
108
-1.5
C
+
0
'0*
40-
30-
20-
10-
mmmftdI a I I I
PHF6-ChIP
B
0i
C
CO
E0
W'0A?
0
z
9
0
High
C:
0
a)
a)
0.
C
a)
4_
0
U)
ow
NucleoATAC
-1000 -500 0 500 1000
Distance from TSS (bp)
Min Max - Phf6T - -- - -- Phf6<0
Figure 2.13. ATAC-Seq analysis of Phf6wr and Phf6KO cells reveals large changes in nucleosomes
positioning. (A) Heatmaps representing the nucleosome positioning called by NucleoATAC (left) and PHF6 binding
determined by ChIP-Seq (right). Both heatmaps are sorted by gene expression (high to low) in the Phf6wr cells, and
are centered at TSSs 1 kb. (B) Metagene analysis of nucleosome positions plotted for global analysis (top), genes
up-regulated upon knockout of Phf6 (middle), and genes down-regulated upon knockout of Phf6 (bottom), assessed
at TSSs 1kb genomic regions. Nucleosome positions are shown for Phf6wr (solid line) and Phf6KO (dotted line) cells,
called by NucleoATAC algorithm and normalized for batch effects using the chromVAR package. -1/+1, +1/-1
nucleosomes flanking the TSS; NFR, nucleosome-free region; grey bars indicate major changes in nucleosome
positioning that corresponds to enriched PHF6 binding. Shaded regions around average tracks denote estimates of
95% confidence interval of the metagene average signals based on resampling. [Soto-Feliciano et al. 2017].
109
A
Global Nucleosome Positions
Differentially Expressed Genes UP
-4 *yE 1
Differentially EXpressed Genes DN
4C4
C4
M
enrichment colocalizes with that of nucleosomes that flank the TSS (Fig. 2.13A, left).
Further, lowly expressed genes show PHF6 biding broadly at the TSS and surrounding
areas. Due to the differential binding patterns of PHF6 at highly and lowly expressed
genes, we next investigated the effect of Phf6 loss on nucleosome positioning in
differentially expressed genes (Fig. 2.13B). Globally, there are few differences in
nucleosome occupancy between Phf6wT (solid line) and Phf6KO (dashed line) cells (Fig.
2.13B, top). In genes that experience up-regulation of expression in Phf6KO cells, there
is a notable depletion of nucleosomes at the TSS (Fig. 2.13B, middle). Further, in genes
that are down-regulated upon deletion of Phf6, nucleosomes converge at the TSS and
regions surrounding the TSS, showing increased occupancy and the +/-1 nucleosome
positions, with the largest changes occurring at the +1 position (Fig. 2.13B, bottom). We
observe consistency in our analyses, with the changes observed in gene expression
reflected in both chromatin accessibility and nucleosome positioning (Fig. 2.4A, Fig.
2.11A, Fig. 2.13B).
PHF6 is necessary for maintenance of chromatin organization at lineage-specific
genes
Given the lineage promiscuity observed upon loss of Phf6 in vivo, we decided to
further compare the transcriptional states of Phf6WT and Phf6KO cells to that of B- and T-
cells. We utilized two pre-curated lists of genes whose expression is unique and
distinguishing to CD19+ B- and CD4+ T-cells [Painter et al. 2011] (Supp. Table S3). As
expected, most of the B-cell signature genes are down-regulated in Phf6KO cells,
whereas the T-cell signature genes showed bi-directional changes (Fig. 2.14A).
Importantly, this observation is consistent with the mixed-lineage malignancy that arises
from syngeneic transplantation of Phf6KO cells in vivo (Fig. 2.2-2.3). Seeing that DNA
accessibility is reshaped at lineage-specific TF biding sites in Phf6KO cells, we decide to
focus more closely on chromatin dynamics at the CD19+ B- and CD4+ T-cell gene sets
(Fig. 2.14B-C; Fig. 2.15). We find that the chromatin accessibility at the promoter
regions of these lineage-specific genes show pronounced changes (Fig. 2.14B-C). In
order to gauge whether PHF6 is directly involved in modulating chromatin accessibility,
we assayed for PHF6 binding at B- and T-cell genes by ChIP-qPCR and indeed found
110
A B-cell gene set
Phf6wi Phf6(O
T-cell gene set
Phf6w Phf(O
Standardized Iog2(Expresslon)
B
B-cell gene set
'5
I-C
0)
U)
T-cell gene set
Phf6wT Phf6/o Phf6wT
p=7.21e-6
C
Cr(D
U)
C/,
U,
i
0
on
C
0
Standardized log(Expreaaion)
Phf(co
n.s.
I I I
-1.0 -0.5 0.0 0.5 1.0
Distance from TSS (kb)
B-cell gene set
-- T-cell gene set
Genome-wide
D 11001
100
M IgG
m Phf6
I ii I IIiII I iII I I I I I I II iJW1- I
B-cell gene set T-cell gene set
Figure 2.14. Genes constituting CD19+ B-cell and CD4+ T-cell gene sets are enriched for PHF6 binding and
subsequent changes in chromatin accessibility after loss of Phf6. (A) Heatmaps comparing gene expression of
curated CD19+ B-cell (left) and CD4+ T-cell gene sets (right) between Phf6wr (left) and Phf6KO (right) cells as
determined by RNA-Seq. Scale corresponds to row-wise standardized log2-transformed expression values for each
gene. Each column represents a replicate sample. (B) Violin plots of ATAC-Seq signals in the promoter regions of
CD19+ B-cell and CD4+ T-cell gene sets in Phf6Yr and Phf6KO cells. P values are shown as returned by the
Wilcoxon rank-sum test. (C) ATAC-Seq signal enrichment at B-cell gene sets, T-cell gene sets, and genome wide,
centered around the TSSs 1kb. Vertical axis indicates ATAC-Seq signal and horizontal axis shows the position
relative to the TSS. (D) ChIP-qPCR analysis of PHF6 and control IgG binding at representative B-cell genes and T-
cell genes. Negative control, PchdO. [Soto-Feliciano et al. 2017].
111
0
0
U-
10
1i
U-. .
SO A) p2 SAGoO- e
O
\Neq sp"
0 of\ -BrO 'c e\y- 000, \,00 Irk' C,&" \AeSA Vr O V
enrichment of PHF6 binding at these targets (Fig. 2.14D). Nucleosome occupancy at
lineage-specific gene sets, in addition to chromatin accessibility, experienced dramatic
shifts upon loss of Phf6 (Fig. 2.15A). Focusing on the CD19+ B-cell gene set, we
observed an increase in nucleosome occupancy at the TSS and within the gene body
upon Phf6 loss (Fig. 2.15A, middle, dashed line). When aligned with the promoter-TSS
binding profile of PHF6 in Phf6wT cells, we observed a large overlap between regions
with enriched PHF6 binding and major shifts in nucleosome occupancy (grey bar). This
finding suggests that PHF6 coordinates nucleosome architecture at B-cell genes in a
way that is favorable for transcriptional activity. When PHF6 is lost, nucleosome
occupancy is enriched and gene expression decreases (Fig. 2.15A, middle; Fig. 2.14A).
Focusing on the CD4+ T-cell gene subset, we observed decreased nucleosome
occupancy surrounding the TSS in Phf6KO cells (dashed line), with preferential depletion
at the -1 nucleosome (Fig. 2.15A, right). Our data is consistent with promoters
undergoing increased chromatin accessibility, where the -1 nucleosome undergoes
more pronounced depletion [Schep et al. 2015]. Furthermore, PHF6 binding is enriched
at sites that are undergoing global shifts in nucleosome occupancy (Fig. 2.15A, right).
The chromatin landscape surrounding the TSS of genes important for CD4+ T-cells
demonstrates fluidity in Phf6KO cells, leading to increased gene expression in a subset
of T-cell genes (Fig. 2.14A), and presentation as a mixed-lineage lymphoma in vivo (Fig.
2.2-2.3). The shifts in nucleosome occupancy that we observe upon complete loss of
Phf6 indicate that B-ALL cells are undergoing lasting and stable transcriptional changes
resulting from alterations in the chromatin architecture, allowing for genomic (as well as
phenotypic) plasticity. Specific genes undergoing changes in chromatin accessibility at
regulatory elements include B cell-specific genes Btg2 and Cd74 (both of which become
less accessible upon Phf6 loss), and T-cell lineage genes Cd4 (distal enhancer) and
Gata3 (which become more accessible upon Phf6 loss) (Fig. 2.15B-C). Importantly, we
confirmed enrichment of PHF6 binding at these target genes through ChIP-qPCR
analysis (Fig. 2.15D-E), suggesting that the observed shifts in chromatin accessibility
are dependent on the presence of PHF6 [Soto-Feliciano et al. 2017]. In conclusion, we
show that PHF6 is necessary for the maintenance of a chromatin state that is
instrumental in promoting B-cell identity, that Phf6 loss leads to the acquisition of a
112
Global Nucleosome Positions
- Phf6wr --- = -Phf6KO
B-cell genes Nucleosome Positions T-cell genes Nucleosome Positions
4) 0!
E-0
CL90
r
0
4) C
N
0
0
E q3
CL .
(0
x
CL
-1 +1
-1000 -500 0 500 1000-1000 -500 0
-1000
-1 +1
- +
Sao 100 -1000 -500 0 500 1000
-500 0 500 1000-1000 -500 0 500 1000 -1000 -500 0 500 1000
Distance from TSS (bp) Distance from TSS (bp) Distance from TSS (bp)
B C
10-1-001 10-1-001
KO KO
-ool Wk PHF8 7A10-2. a. PHFA
10-2001' A - M. HK27AC'P2001 IH3K27Ac
Btg2
0-1.00 , -- - 7
I0"0.3m1~.. MJ.~~~6 PHIPS
[0-2.501 
-. M-V M-- -- I 32AC
Cd4
10-1-001 
-KO
[0-2.20]
=- m -
Cd74
I ] EiIgG *EPhf ..5 3& Gata31sI
Neg Control Btg2 Cd74Neg Oontroi Big C74 aNeg Control Od4 Gta3
Figure 2.15. PHF6 has distinct binding patterns within lineage-specific genes that results in drastic changes In
nucleosome occupancy upon genetic deletion. (A) Metagene analysis of nucleosome positions (top) and fold enrichment of
PHF6 binding (bottom) plotted for global analysis (left), CD19+ B-cell gene set (middle) and CD4+ T-cell gene set (right), assessed
at TSSs *1kb genomic regions. (Top) Nucleosome positions are shown for Phf6wr (solid line) and Phf6KO (dotted line) cells, called
by NucleoATAC algorithm and normalized for batch effects using the chromVAR package. (Bottom) Metagene tracks of PHF6
ChIP-Seq signal averaged over the TSS 1kb in Phf6wr cells. -1/+1, +1/-1 nucleosomes flanking the TSS; NFR, nucleosome-free
region; grey bars indicate major changes in nucleosome positioning that corresponds to enriched PHF6 binding. Shaded regions
around average tracks denote estimates of 95% confidence interval (Cl) of the metagene average signals based on resampling.(B) Loci of B-cell genes (Btg2, top) and (Cd74, bottom) showing ATAC signal in Phf6wr and Phf6KO cells and PHF6/H3K27ac ChIP-
Seq signals in Phf6wr cells. Grey bars highlight changes in chromatin accessibility. Numerical ranges in reads per million (RPM)
are shown in square brackets. (C) Loci of T-cell genes (Cd4, top) and (Gata3, bottom) showing ATAC signal in Phf6wr and Phf6KO
cells and PHF6/H3K27ac ChIP-Seq signals in Phf6wr cells. Grey bars highlight changes in chromatin accessibility and intergenic
regions. (D) ChIP-qPCR analysis of PHF6 and control IgG binding at representative B-cell genes, Btg2 and Cd74. (E) ChIP-qPCR
analysis of PHF6 and control IgG binding at representative T-cell genes, Cd4 and Gata3. Negative control, PchdO. [Soto-
Feliciano et al. 2017].
113
A
D 1E2&
a 20-
E
C
0 
1U-
M
-
I
0*0
r"K"A
permissive/open state at loci of alternate lineages, thus allowing for the engagement of
aberrant transcriptional programs that result in phenotypic plasticity.
Instability in the chromatin landscape allows for aberrant lineage signaling
Given the genomic plasticity observed in our Phf6KO B-ALL cells and the changes
in chromatin accessibility detected at regions of T-cell transcription factor binding sites
upon Phf6 loss, we hypothesized that the genome of Phf6KO cells may now be
permissive to aberrant oncogenic signaling emanating from T-cell lineage factors, such
as NOTCH1. Activating mutations in NOTCHI are well-documented in more than 60%
of T-cell acute lymphoblastic leukemia cases (T-ALLs) [Belver & Ferrando 2016], where
PHF6 is often found to be co-mutated with NOTCHI [Li et al. 2016; Wang et al. 2011].
Clustering analysis of 116 human T-cell leukemia patient samples show that mutational
profiles with NOTCHI and PHF6 alterations cluster together (Fig. 2.16A). Therefore, we
tested the effects of overexpression of the NOTCH1 intracellular domain (NICD) in
Phf6wT and Phf6KO B-ALL cells. As expected, we observed that NOTCH1 signaling
dramatically inhibits the growth of Phf6wT B-ALL cells [Rothenberg 2014] (Fig. 2.16B).
However, Phf6KO cells were viable upon NICD overexpression, with equivalent growth
rates as Phf6vwr and Phf6KO vector control cell lines (Fig. 2.16B). Transplantation of
Phf6KO + NICD cells into immunocompetent syngeneic mice gives rise to a latent
lymphoma with extensive thymic infiltration and elevated levels of CD4 expression,
characteristic of T-cell malignancies (Fig. 2.16C-E) [Soto-Feliciano et al. 2017]. In
conclusion, PHF6 deficiency in B-ALL cells creates a permissive chromatin state that
allows for TFs from alternate lineages to bind to their respective genomic loci, which are
normally in a closed conformation in wild-type cells. Forced expression of NOTCH1 (a
T-cell TF) takes advantage of the chromatin fluidity in knockout cells, where now Phf6KO
tumors are driven more towards a "T-cell like" state instead of a mixed-lineage
phenotype.
114
I N I
I Il
Ill I
SI
H
I lu
Ill
11
I II I'll III
IN iPi
1111111
111111 III IHI I I
I IIIIIIIIIIIIIIIIIIIIIII I I
frameshift / missense
.......
*.
*
Phf6 T Phf6NO Phf6NO
+iControl +iControl .iNICD cDNA
I1111UPl
I I
H
"'ii
lIll
I
B
E2
E
P6
E1C
Ijl
A
1B
R
2
1
1
7
1B3
PHI
4x10 7-
3x107-
2x107-
1x107
100.
50-
non-mutated 0
E
40.
10.
0
1,
6 1 2 3 4 5 6 7
Time (days)
+ Phf6W
+ Phf6e 0
-.- Phf6'iT+ Control
+ PhfOC 0+ Control
+ Phf6w+ NICD cDNA
+ Phf6 0+ NICD cDNA
a
a--i
U------------U
1O 20 3o 40
Time after transplantation (days)
- Phf6wT+ Control (n=14) 1
- Phf6KO+ Control (n=12)
* Phf6KO+ NICD cDNA (n=18)
****
_rwi F
0001
Phf6WT Phf6KO Phf6KO
+Control +Control +NICD cDNA
Lymph nodes
Figure 2.16. Chromatin Instability allows for aberrant T-cell transcription factor signaling. (A) Mutational landscape of T-cell
acute lymphoblastic leukemia tumors from patient samples. Columns represent mutational status of individual patient samples (red,
mutant; tan, non-mutant). Rows represent the indicated gene. Genes are ranked by clustering of mutational profiles, showing
frequency of co-mutations. Data came from Keersmaecker et al. (2012) and Van Vlierberghe et al. (2010). (B) Cell proliferation
assay comparing Phf6wr (blue, up-arrow), Phf6KO (red, up-arrow), Phf6wr + control vector (blue, circle-dotted), Phf6KO + control
vector (red, circle-dofted), Phf6wr + NICD cDNA (blue, down-arrow), and Phf6KO + NICD cDNA (red, down-arrow) cells (n=3). NICD,
NOTCHI intracellular domain. (C) Kaplan-Meier survival analysis of mice injected with 106 B-ALL cells of the indicated genotypes
infected with control vector or activated NOTCH1 intracellular domain (NICD) vector. The number (n) of mice per genotype
analyzed is shown. Statistical analysis (log-rank test, Mantel-Cox) was performed for the different groups in comparison to mice
injected with Phf6wr+control vector cells. P values are shown for the comparisons. (D) Quantification of combined thymus weight of
Phf6wr + control vector (blue, n=4), Phf6KO + control vector (red-unfilled, n=6), and Phf6KO + NICD cDNA (red-patterned, n=7)
recipients. (E) Bar graphs showing the percentage of the CD4+ fraction among mCherry+ cells isolated from Phf6wr + control vector(blue, n=5), Phf6KO + control vector (red-unfilled, n=4), and Phf6KO + NICD cDNA vector (red-pattemed, n=6) tumors from lymph
nodes. Data represent the mean * standard deviation (SD) in D-E. Statistics were calculated with two-sided Student's t-test: *P <
0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. [Soto-Feliciano et al. 2017].
115
A
D
300-
2 200-
100.
C
ci,
C
00
02a..
I
S]n.s.
Phenotypic plasticity underlies differential responses to anti-cancer treatments in
vivo
Clinically, patients with B- and T-cell leukemias and lymphomas undergo different
treatment strategies. BCR-ABL1-driven B-cell leukemia is often treated with targeted
kinase inhibitors, whereas T-cell leukemias and lymphomas are treated with intensive
chemotherapy regimens [Schultz et al. 2009; Litzow & Ferrando 2015]. Therefore, to
further understand the shift from a disseminated leukemia to a CD4+/B220+/CD19+
mixed-phenotype solid lymphoma that results from complete loss of Phf6 in B-ALL cells,
we examined the response of these tumors to targeted therapy (Ponatinib) and front-
line chemotherapy (Doxorubicin) (Fig. 2.17-2.18).
Unbiased transcriptome analysis shows decreased dependency on BCR-ABLI
signaling pathways, the driving oncogene in this B-ALL model [Shamroe et al. 2013;
Williams et al. 2006] (Fig. 2.17A-B). Consistent with this, we observed a marked
difference in the response to Ponatinib (third generation BCR-ABLI inhibitor) between
Phf6wT and Phf6KO tumors in vivo, but not in vitro (Fig. 2.17C-D). Mice transplanted with
Phf6KO cells and treated with Ponatinib have a median survival of 15 days, while their
wild-type counterparts survived more than 60 days post-treatment (Fig. 2.17D).
Moreover, relative expression of the BCR-ABLI transcript is significantly reduced in the
bone marrow and lymph nodes of moribund mice transplanted with Phf6KO cells
compared to those transplanted with Phf6wvr cells (Fig. 2.17E). Focusing on the
composition of cells after treatment, we observe that CD4+ Phf6KO leukemia cells persist
after administration of Ponatinib, with the highest level of CD4+ disease burden residing
in the thymus of recipient mice (Fig. 2.17F). This data suggests that the B-ALL cells
undergoing a "lineage-switch" are inherently more resistant to Ponatinib treatment in
vivo compared to their Phf6WT counterparts. Thus, the lineage plasticity that arises in
vivo upon Phf6 loss drives a significantly decreased response to the targeted kinase
inhibitor Ponatinib.
We next examined the response of Phf6WT and Phf6KO tumors to the
chemotherapeutic agent Doxorubicin. Although there are no changes in response to the
drug in vitro, we find that mice transplanted with Phf6KO cells are exquisitely sensitive to
Doxorubicin treatment in vivo, as demonstrated by an additional 13-days extension in
116
Klin-Targets of BCR-ABLi fusion
NES -I 5IV
FOR-s 0175
Down in Phf6KO
00
-01
-02
-03
-OA
-05
-06
-07
Huang Dasatinib resistanceUP
FOR - 00968
A
B
Down in Phf6KO
-021
-02
-04
-07
DiazChronic myeloid leukemiaDN
NES ;7250
P-58ke,~ 5
Up in PhfBKO Down in Phf6Ko
I, - h
~I R-0 -~~'I , A
Standardized log2(Expression)
D
+ Phf6K
-- Phf6K0
0)
C
0)C-)
0)
a-
100' I...........
50
0.1 1 10 100 1000 10000
[Ponatinib], (nM)
F
-7-
+
0
0 S .0
Phwr M
p<0.01 --- wrP
i_ p<OS .F--- p-KO p
------- -------- a
ockTreated (n-6)
onainib Treated (n=7) Ip01
lock Treated (n-7)
onalinib Treated (n-6) p<.05
15 2040 60 80
Time after transplantation (days)
60.
40-
20-
Phf6WP Phf6K0 Phf6KO Phf6KO
BM m1 BM m2BM ml BM m2 LN m1 LN m2
in vivo
8*8* *88*
BM LN SP THY
E] Ph6wr Mock Treated
"" PhBKO Mock Treated
0 PhIBKO Ponaijnb Treated
n
Figure 2.17. Loss of Phf6 results in decreased dependence on the driving oncogene BCR-ABL1 in vivo. (A) GSEA plots
depicting significant changes in Targets of BCR-ABL1 fusion signature (P=0.0129), Dasatinib resistance signature (P=0.0056), and
Chronic myeloid leukemia (P<0.001) upon Phf6 deletion, as compared to Phf6WT cells. NES, normalized enrichment score; FDR,
false discovery rate. (B) Corresponding heatmaps of genes from GSEA signatures in Phf6wr (top) and Phf6KO (bottom) cells. (C)
Viability curves of Phf6wT and Phf6KO cells after continuous treatment with ponatinib in vitro for 48 hours. (n=3). LoglC50's (Phf6wv
=0.2100) and (Phf6KO =0.1049) are significantly different at P < 0.0001 by an extra sum-of-squares F test. (D) Kaplan-Meier
survival analysis of mice injected with 106 B-ALL cells of the indicated genotypes with subsequent mock or ponatinib treatment (30
mg/kg daily for 4 consecutive days). The number (n) of mice per genotype analyzed is shown. Statistical analysis (log-rank test,
Mantel-Cox) was performed for the different groups in the indicated comparisons. P value is shown for the comparison. (E) qPCR
analysis of cells from the bone marrow or lymph nodes of moribund mice injected with Phf6WT (blue) and Phf6KO (red) cells (n=3).(F) Bar graphs showing the percentage of the CD4+ fraction among mCherry+ cells isolated from Phf6WT (blue, n=9) and Phf6KO(red, n=5) tumors upon treatment with vehicle (unfilled) or ponatinib (patterned). Phf6wT mock treated, n=5. Phf6KO mock treated,
n=5. Phf6KO ponatinib treated, n=3. BM, bone marrow; LN, lymph node; SP, spleen; THY, thymus. Data represent the mean
standard deviation (SD) in E-F. Statistics were calculated with two-sided Student's t-test: *P < 0.05, **P < 0.01, ***P < 0.001, ****P
< 0.0001, n.s. = not significant. [Soto-Feliciano et al. 2017].
117
00
-0 ;,
03
-04
-06
_0 6
-07
2
0
up in PhttflUp in Phf6KO
0=
0
C
()
.0
E
0
12000-
10000
8000-
6000-
4000-
2000-
0
0.01
1.5-E
1.0-
0.5-
0)
cr 0.0.
"" "
An I Ah I M M
-;Phfwr
PhfKo
B
C,@
>
100-
50
0[ 7 1 n-5 ], ( 3 [Doxorubicin], (M)
-- Ph Mock Treated (n=10)
P-- O Phw Doxorubicin Treated (n=10)
p<0.0001 -- Phf6KO Mock Treated (n=9) p<0.001i--' -- i-- PhO Doxorubicin Treated n=8)
FL- -1
10 20 30 40 50
Time after transplantation (days)
D
o* V
I I I U U
h
4
00
50-
40-
30.
20-
n.s.
-- n.s.
BM
BM LN SP THY BM LN SP THY
Phf6"T
Dox Treated
PhfCKO
Dox Treated
U
LN
.3 Phf'r Mock Treated
E PhfSwTDoxorubicin Treated
D3 PhSKO Mock Treated
0 PhSKO Doxorubicin Treated
Fn.s.
n~**
1 ..L..nU U ... ...
1500.
Phf6WT Phf6K Phf6WT Phf6K Phf6WT Phf6K
No IR 2 Gy 10 Gy
n.s. *
.. .
Phf6WT Phf6K Phf6,T PhfGK0 Phf6T Phf46K
BM LN THY
Figure 2.18. Resistance to targeted therapy via pathway indifference uncovers collateral sensitivity to
doxorubicin. (A) Viability curves of Phf6wr and Phf6KO cells after continuous treatment with doxorubicin in vitro for
48 hours. (n=3). LoglC50's (Phf6Wr =-4.190) and (Phf6KO =-4.058) are not significantly different at P=0.06963 by an
extra sum-of-squares F test. (B) Kaplan-Meier survival analysis of mice injected with 106 B-ALL cells of the indicated
genotypes with mock or doxorubicin (single IP dose, 10mg/kg) treatment. The number (n) of mice per genotype
analyzed is shown. Statistical analysis (log-rank test, Mantel-Cox) was performed for the different groups in the
indicated comparisons. P value is shown for the comparison. (C) Tumor burden in the indicated organs of mice
injected with 106 B-ALL Phf6wT (blue, n=5) and Phf6KO (red, n=5) cells. mCherry demarcates tumor cells. (D) Bar
graphs showing the percentage of the CD4+ fraction among mCherry+ cells isolated from Phf6Wr (blue, n=9) and
Phf6KO (red, n=5) tumors upon treatment with vehicle (unfilled) or doxorubicin (patterned). (E) Levels of rH2AX after
Irradiation (0, 2, or 10 Gy) in cultured Phf6wT and Phf6KO cells. (F) Levels of rH2AX after doxorubicin treatment in
indicated organs of recipient mice injected with Phf6WT or Phf6KO cells. BM, bone marrow; LN, lymph node; SP,
spleen; THY, thymus. Data represent the mean standard deviation (SD) in C-F. Statistics were calculated with two-
sided Student's t-test: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, n.s. = not significant.
118
A
100-
1-8
C
100-
s
W
0
E
0-
n.s. n.s.
THY
E
Co
.9
0
U-
C
6000-
4000-
2000-
~
N
.
.
A
:J
median survival (essentially doubling their lifespan) when compared to mice injected
with Phf6WT cells (Fig. 2.18A-B). Upon morbidity, Phf6wT tumor cells are found in the
bone marrow, lymph nodes, and thymus, whereas Phf6KO disease persists exclusively in
the thymus (Fig. 2.18C). This demonstrates that the thymus is a chemo-protective
microenvironment for Phf6KO tumor cells, as was previously documented for bona fide
lymphomas [Gilbert & Hemann, 2010]. We next asked if the amount of CD4+ Phf6KO
leukemia cells was significantly altered after treatment with chemotherapy. We find that
CD4+ Phf6KO cells are indeed more sensitive to Doxorubicin treatment, and that levels
are diminished to a level comparable to Phf6wT disease in the lymph nodes and thymus,
and to a lesser extent in the bone marrow (Fig. 2.18D).
Based on the changes in the chromatin landscape that are observed upon loss of
Phf6, we hypothesized that the sensitivity to the DNA intercalating agent, Doxorubicin,
in Phf6KO tumors may be the result of a more plastic, fluid genome, allowing more
access to DNA for the drug, and thus more damage. To test this, we assayed for
accumulation of yH2AX, a marker of DNA damage, after irradiation in vitro and
treatment with Doxorubicin in vivo. Comparing Phf6wT and Phf6KO cells after irradiation,
we found the original hypothesis to be incorrect, with Phf6WT cells having greater or
equal amounts of DNA damage, as quantified by yH2AX levels (Fig. 2.18E). This trend
is amplified in vivo after Doxorubicin treatment, with Phf6w'T tumors sustaining
significantly greater levels of DNA damage in the bone marrow, lymph nodes, and
thymus (Fig. 2.18F). Therefore, the differential responses to Ponatinib and Doxorubicin
observed between Phf6wNT and Phf6KO malignancies reflect a true shift in disease states.
In conclusion, we find that the complete loss of Phf6 in B-ALL cells not only
changes the chromatin landscape of these tumors, but also the phenotypic state
reflected in the formation of a different disease with vastly distinct responses to anti-
cancer therapies. These findings support a possible underlying mechanism for the
examples of lineage plasticity and drug resistance observed in a variety of
malignancies, and demonstrate that modulation of the chromatin landscape, through the
loss of regulation by a single factor (such as PHF6), may play a crucial role in acquired
resistance.
119
Loss of Phf6 leads to downregulation of B-cell developmental genes in multiple
acute leukemias
Work thus far has demonstrated that genetic ablation of Phf6 leads to a myriad of
changes in the transcriptome, chromatin landscape, disease presentation, and
response to cancer treatments. We sought to extend these findings by exploring the
effect of loss of Phf6 on another model of B-cell leukemia. Using a pro-B-ALL model
driven by the MLL-AF4 oncogene, we generated knockouts of Phf6 using CRISPR-
Cas9 (Fig. 2.19A) [Krivtsov et al. 2008]. PHF6 protein and transcript levels were
diminished to nearly undetectable levels (Fig. 2.19B-C). As seen in the BCR-ABLI
driven B-ALL model, loss of Phf6 in MLL-AF4 driven B-ALL likewise reduces the
expression of genes important for B-cell development (Fig. 2.19D; Fig. 2.4E). This
demonstrates that gene expression regulation by Phf6 is important across multiple B-
cell malignancies, and implicates a possible role for this protein in maintaining or
propagating B-cell identity in somatic cells.
A
gRNA sequence
WT locus 5' - AGCGAGATGAAGAAGATGAGGAGCGAGAGAGTAAAAG-CCGTGGAAGAGTAGCG- 3'
Allele 1 5' - AGCGAGATGAAGAAGATGAGGAGCGAGAGAGTAAAAGTTCCGTGGAAGAGTAGCG- 3' 2bp insertion
Allele 2 5' - AGCGAGATGAAGAAGATGAGGAGCGAGAGAGTAAAAGT-CCGTGGAAGAGTAGCG- 3' 1bp insertion
B sgPhf6 C 1.
< c) CO) LO r-- 00 - - - -~ '
0I 0 0 W 0 0 0 W 000
_J .j 00 0 C00000000 C
_j 00 0 0 00 0 0 0 0 0
0.
HSP90
m PHF6 a.
MLL-AF4 MLL-AF4
Phf6vT Phf6KO
D
1.5 *. * ** ** *** ** n.s. **
C
0
. MLL-AF4 Phf6I
2 1.0 U MLL-AF4 Phf6KO
0.5
0.0-i
IL4ra Rhoq Cd74 Ly86 Lyn Btg2 Cd22 Cd79b Phf6
Figure 2.19. Knockout of Phf6 results In decreased expression of B-cell developmental genes in MLL-AF4 acute leukemia. (A)
Generation of a MLL-AF4 Phf6KO isogenic B-ALL cell line using CRISPR-Cas9. Black line and bolded letters indicate the guide RNA used.
Red denotes the PAM sequence. Blue denotes the indels introduced into each allele. (B) Western blotting analysis of Phf6w" and Phf6KO
cells transiently transfected with pX330 Phf6_sg4. Yellow arrow indicates the cell line used for further analysis. ALLmCh = BCR-ABL1
driven B-ALL. (C) Relative Phf6 mRNA levels in MLL-AF4 Phf6wT (blue) and MLL-AF4 Phf6KO (red) cells. (D) qPCR analysis of MLL-AF4
Phf6wr (blue) and MLL-AF4 Phf6KO (red) cells. Relative mRNA levels for B-cell associated genes are shown: II4ra, Rhoq, Cd74, Ly86, Lyn,
Btg2, Cd22, Cd79b, Phf6. Data represent the mean standard deviation (SD) in C-F. Statistics were calculated with two-sided Student's t-
test: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, n.s. = not significant.
120
DISCUSSION
The transcriptional regulation of lymphocyte development via lineage-specific
factors is well characterized, and the perturbation of this process can promote
leukemogenesis [Mullighan et al. 2007]. However, the dynamic regulation of chromatin
states underlying transcriptional activity and its contribution to cancer are yet to be
elucidated. In this study, we show that PHF6 is a novel chromatin state regulator,
important for the maintenance of a chromatin landscape conducive to B-cell identity in
B-ALL. Through integrated genomics (RNA-Seq, ChIP-Seq, ATAC-Seq) and in vivo
studies, we demonstrate the crucial role that PHF6-mediated chromatin modulation
plays in genomic plasticity, cell identity, leukemogenesis, and response to therapy. Our
data convincingly shows that Phf6 loss leads to a chromatin environment that tolerates
aberrant lineage and oncogenic signaling. Moreover, we demonstrate how malignant
cells can hijack chromatin instability to reprogram transcriptional output and cell identity.
These findings propose the first mechanistic insights into PHF6's role in a lymphoid
malignancy. We show that PHF6 binds to nucleosomes at specific genomic loci, and
orchestrates distinct transcriptional programs. It stabilizes chromatin accessibility at B-
cell specific genes, while T-cell transcription factor binding sites are concealed in
compacted chromatin (Fig. 2.20). Upon loss of this single factor, chromatin architecture
around B- and T-cell specific genes undergoes dramatic remodeling, resulting in focal
genomic plasticity and acquisition of T-cell lineage markers in B-ALL cells (Fig. 2.20).
PHF6 appears to be acting as a novel "chromatin boundary factor" in B-ALL by
promoting a chromatin landscape and transcriptional program that is conducive to B-cell
differentiation. When Phf6 is lost, these chromatin boundaries are lost, resulting in
genomic plasticity.
With 80% of PHF6-mutant T-ALLs also harboring mutations in NOTCHI, and the
normal function of NOTCHI being to activate expression of T-lineage specific
transcription factors, we propose a model by which loss of PHF6 results in a cellular
context in which NOTCHI and subsequent T-cell lineage transcription factors can now
bind and drive transcription of their target genes [Meier-Stiegen et al. 2010;
Schwanbeck et al. 2011]. Along these lines, our data further supports the possibility that
121
a significant portion of T-ALLs (-30%) may have once started as a B-cell progenitor/
precursor that acquired a mutation in PHF6 during early stages of leukemogenesis, and
consequently forced the cell-of-origin into a "pseudo T-cell state" that then presented as
a T-cell malignancy [Van Vlierberghe et al. 2010].
Lineage switching is a rapidly emerging mechanism of resistance to targeted
therapies. A variety of cancers (including lung cancer, prostate cancer, and B-cell
leukemia) can acquire resistance to therapies by relapsing as a histologically distinct
malignancy [Oser et al. 2015; Sequist et al. 2011; Yu et al. 2013; Ku et al. 2017; Mu et
al. 2017; Gardner et al. 2016; Jacoby et al. 2016]. The first functional studies
investigating this process as a driver of drug resistance in prostate cancer identified
epigenetic reprogramming proteins (SOX2 and EZH2) as key effectors of this process
[Ku et al. 2017; Mu et al. 2017]. In our study, we further demonstrate that active
maintenance of chromatin architecture (by PHF6, a novel chromatin regulator) is also
critical in the response to targeted therapy in B-cell leukemia. We postulate that the
chromatin landscape is a pivotal component in the response of many cancer types to
targeted treatments.
Here we identify the function of PHF6 in B-ALL and describe a role in regulating
chromatin accessibility to lineage-specific transcription factors. We show that loss of
Phf6 results in massive disruption of lineage differentiation via focal changes in
chromatin accessibility, engagement of alternative transcriptional programs, altered
disease presentation in vivo, and tolerance of aberrant oncogenic signaling. We
establish that the proper chromatin landscape is essential for maintaining cell identity,
and that loss of PHF6 can cause changes in chromatin accessibility and nucleosome
positioning that result in lineage promiscuity. These findings highlight a mechanism of
gene expression maintenance via chromatin landscape modulation that underlies the
developmental plasticity characteristic of hematopoietic malignancies.
122
Phf6WT
Phf6
y genes: ON
Phf6KO
B-cell identity genes: OFF
T-cell identity genes: ON
Figure 2.20. Model of PHF6 as a chromatin state regulator, permitting transcription factor binding through
chromatin accessibility. In wild-type cells, PHF6 binds to the 1 nucleosome flanking the open TSS of genes,
allowing for B-cell-specific TFs to bind, drive gene expression, and maintain B-cell identity. Conversely, PHF6 binds
nucleosomes surrounding the TSSs of T-cell-specific genes, coordinating chromatin compaction, and thus blocking
the binding of T-cell-specific TFs. In the absence of PHF6, chromatin is no longer maintained in an open state, B-
cell TFs cannot bind, and expression of B-cell identity genes is downregulated. However, T-cell identity genes are
no longer inaccessible, allowing T-cell specific TFs to bind and activate aberrant transcriptional programs. [Soto-
Feliciano et al. 2017].
123
B-cell identit
T-cell identity genes: OFF
jj jj - i ec
APPENDIX 1
The PHF6 protein has distinct domains important for lymphoid and neuronal
contexts
PHF6 was first discovered and characterized as the gene mutated in the
intellectual disability disorder B6rjeson-Forssman-Lehmann syndrome, or BFLS [Lower
et al. 2002]. While BFLS patients largely have germline missense mutations in PHF6, T-
ALL and AML patients almost exclusively sustain truncation or frameshift mutations (Fig.
2.21A) [Todd & Picketts 2012]. Based off of this observation, it appears that complete
loss of PHF6 function is necessary in leukemogenesis, while missense mutations
observed in BFLS may focally affect PHF6 neuronal operations, while leaving domains
important for the lymphoid context intact. This is further supported by the fact that
patients with BFLS are not predisposed to lymphoid malignancies, with only a single
documented case of T-ALL and Hodgkin's lymphoma in two separate male children
[Chao et al. 2010; Carter et al. 2009]. In order to test whether a clinically observed
missense mutation can restore lymphoid function, we generated a BFLS-mutant version
of Phf6 cDNA (Fig. 2.21A) [Lower et al. 2002]. We chose a mutation that targeted amino
acid 243, causing a change from a lysine to a glutamate within the second zinc knuckle
domain. Expression of BFLS-mutant cDNA is similar to that of wild-type Phf6 cDNA (Fig.
2.21 B). At the transcriptome level, BFLS-cDNA rescues the expression of a subset of B-
cell target genes much like wild-type Phf6 cDNA, specifically Cd74 and Lyn (Fig. 2.21C,
Fig. 4E). Upon transplantation into immunocompetent syngeneic mice, BFLS-cDNA also
significantly accelerates the time to disease onset in Phf6wT and Phf6KO recipients,
rescuing the disease latency observed in Phf6KO disease (Fig. 2.21 D). Furthermore, the
levels of CD4 expression in Phf6KO cells are reduced to that from rescue with wild-type
Phf6 cDNA (Fig. 2.21 E). These results demonstrate that mutations that cause Borjeson-
Forssman-Lehmann syndrome and neuronal defects do not affect the lymphoid function
of PHF6, further supporting the hypothesis that PHF6 has context-specific domains and
dependencies.
124
A
BFLS 2 1
T-ALL
AML
ZincKnuckh
mu ,,m'* 1 ii.~J
Zinc
Knuckle II
B
WT BFLS WT BFLS
EV cDNA CDNA EV cDNA cDNA
4 PHF6
HSP90
D
*15
C,)
C
U)
a-
Phf6wT
100.
50-
t~1~
Phf6KO
** **-n.s.
1.0Um
(D ICC 00:1 1 1
U.
Phf6w + EV
Phfw+ cDNA
Phf6r+ BFLS cDNA
Phf6KO + EV
PhMKO + cDNA
PhIUKO + BFLS cDNA
Phf6WT Phf6KO Phf6WT Phf6KO
Cd74 Lyn
E
:I
:I
I
I p<OOO.1
P<00.1
p<OOO.1
AkL 
0000 1<I
+
,eJ
30
20-
10.
** * **
- * ~***
V,
S
A
S
0 5 10 15 20 25
Time after transplantation (days)
Phf6w+EV n=14
Ph6wr+WTcDNA n=5
Phf6wr+BFLS cDNA n=5
Phf6KO+EV n=18
PhKO+WTcDNA n=5
PhfiKO+BFLS cDNA n=4
BM LN SP
[_]+ EV (n=7)
C3+ cDNA (n=3)
:M + BFLS cDNA (n=4)
Figure 2.21. BFLS-mutant PHF6 can rescue lymphoid functions in B-ALL. (A) Schematic of the PHF6 protein, showing
protein domains and location of known mutations found in BFLS, T-ALL, or AML. Red squares represent frameshift and/or
truncating mutations. Blue squares represent missense mutations. The number in each square corresponds to the number of
mutations documented at that location. Nuclear localization signals are shown in orange. Nucleolar localizations are shown in
yellow. Arrow indicates the BFLS mutation that was chosen for further study, nucleotide #700 (A-*G), amino acid #243 (K-*E),
missense mutation in exon 7. Figure adapted from Todd & Picketts (2012). (B) Western blotting analysis of Phf6Yr and Phf6KO
cells expressing Phf6 cDNA or BFLS mutant Phf6 cDNA. Cntrl, control vector. BFLS, B6rjeson-Forssman-Lehmann syndrome. (C)
qPCR analysis of Phf6wr (blue) and Phf6KO (red) cells transduced with empty vector (EV, solid), vector expressing Phf6 cDNA(cDNA, dotted), or vector expressing BFLS mutant Phf6 cDNA (BFLS cDNA, checkered). Relative mRNA levels for B-cell
associated genes are shown: Cd74, Lyn. (D) Kaplan-Meier survival analysis of mice injected with 106 B-ALL cells of the indicated
genotypes. The number (n) of mice per genotype analyzed is shown. Statistical analysis (log-rank test, Mantel-Cox) was
performed for the different groups in comparison to mice injected with Phf6vvr + EV cells. P values are shown for the different
comparisons. (E) Bar graphs showing the percentage of the CD4+ fraction among mCherry+ cells isolated from tumors from Phf6KO(red, n=5) cells infected with a vector expressing empty vector (MSCV-EV) (unfilled) or vector expressing Phf6 cDNA (cDNA,
dotted), or vector expressing BFLS mutant Phf6 cDNA (BFLS cDNA, checkered). The number (n) of mice per genotype analyzed
is shown. BM, bone marrow; LN, lymph node; SP, spleen; THY, thymus. Data represents the mean SD in A-B,E. Statistics were
calculated with two-sided Student's t-test). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, n.s. = not significant.
125
APPENDIX 2
In-depth characterization of BCR-ABL1 driven B-cell acute lymphoblastic
leukemia shows reproducible disease homing and kinetics
In an effort to fully investigate the differences in behavior between Phf6WTr and
Phf6KO disease states, a longitudinal study was performed to characterize each
malignancy over time, rather than solely upon morbidity (Fig. 2.22A). After
transplantation of Phf6wT and Phf6KO cells into immunocompetent mice, recipients were
sacrificed every 3 days and analyzed for disease burden and CD4 expression in the
bone marrow, lymph nodes, and spleen. As seen before, Phf6KO disease has increased
infiltration into the lymph nodes and thymus upon morbidity, as compared to Phf6WT
cells (Fig. 2.22B, Fig. 2.2A-C). Disease kinetics between Phf6WT and Phf6KO cells are
very similar, with disease homing first to the bone marrow (-Day 6) and then the spleen,
with significant burden reaching the lymph nodes by Day 10 (Fig. 2.22C). As disease
develops, the bone marrow becomes almost completely occupied with leukemia cells,
constituting 80% or more of the white blood cells within the bone marrow from Day 9
until morbidity (Fig. 2.22C, left). The lymph nodes also experience heavy infiltration, but
to a greater extent in mice transplanted with Phf6KO cells. Mice injected with Phf6KO cells
show equivalent disease burden in the lymph nodes and bone marrow upon morbidity
(Fig. 2.22C, middle). The spleen has the lowest disease burden, with approximately
50% of cells being leukemic from Day 6 until morbidity (Fig. 2.22C, right).
Despite having similar disease homing behavior, Phf6KO disease consistently and
aberrantly expresses the T-cell lineage surface marker CD4. Cells expressing this
marker are highest in the bone marrow, followed by the lymph nodes and the spleen
(Fig. 2.22D). Rather than upon morbidity, the highest levels of CD4 expression
surprisingly seem to peak in the days preceding sacrifice, with a peak at Day 9 or Day
10 (Fig. 2.22D). Preliminary data suggests that a fraction of the Phf6KO cells have the
capacity to undergo a lineage switch. When performing limited dilution experiments, the
percentage of cells expressing CD4 increases as the number of cells transplanted
deceases (data not shown). This suggests that cells that have undergone a lineage
126
switch may have slower proliferation rates compared to Phf6WT counterparts in vivo, and
that limiting the number of conversion-incapable cells allows for the majority of disease
to originate from CD4+ Phf6KO cells upon morbidity. In line with this, dedifferentiation and
transdifferentiation processes are documented as occurring very slowly [Cobaleda et al.
2007; Ungerback et al. 2015].
127
n=3 n=3 n=3 n=3 n=3 n=3
Phf6w V V V V
Phf6KO V
Time after
transplantation 0
(Days)
3 6 9 10 12
Disease/ burden
(%mCherry)
CD4
Expression
n"s- 
-
- m Phf6W
I_ 
_* 
PhI6KO
BM LN SP
Bone Marrow
THY
Lymph Node
100.
80.
+
60-
40-E
0
20.
* Ph6KO ~% S.
a 6 9 1'2
Time after transplantation (days)
S
I
3 6 9
0*
12 0
Time after transplantation (days)
I
0
I
3 6 9 12
Time after transplantation (days)
D
15-
0-
Bone Marrow
0 Phfw
* Phf6KO
I
0
0 3 6 9 12
Time after transplantation (days)
15.
10.
5.
I
Lymph Node
10.
0
I 3 6 9 12
Time after transplantation (days)
6.
4-
Spleen
U
lin
a
0 3 6 9 12
Time after transplantation (days)
00
Figure 2.22. B-ALL has distinct homing patterns and disease kinetics. (A) Schematic showing the experiment
time course. 1 million Phf6wr or Phf6KO cells were injected into immunocompetent mice. Mice were sacrificed
every 3 days or upon morbidity. Disease burden and expression of CD4 was determined. The number (n) of mice
per genotype analyzed is shown. (B) Disease burden upon morbidity in the indicated organs of mice transplanted
with Phf6wr (blue) and Phf6KO (red) cells. (C) Disease burden (%mCherry cells) upon a longitudinal time course(3, 6, 9, 10, 12 days after transplantation) in the indicated organs. (D) Graphs showing the percentage of the CD4+
fraction among population P3 isolated from tumors from Phf6wr (blue, n=3) and Phf6KO (red, n=3) cells. BM, bone
marrow; LN, lymph node; SP, spleen; THY, thymus. Data represent the mean standard deviation (SD) in B.
Statistics were calculated with two-sided Student's t-test: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, n.s. =
not significant.
128
A
B 100.
0
E
0.
C Spleen
n 1 -1
SUPPLEMENTAL FIGURES
A pX458 * Single cell Crtiin DNA sequencing
B11110- Characterization Protein expression
BcR-ABL ALL scils clones
froma male mosuse
mnRNA levels
B
*PHDI * *
132 239
sgRNA
33o 365aa
T C-term Ab
273aa
D
I***
N-term.
PHF6
HSP90
C-term.
gRNA sequence
WT locus 5' - GAGATTTCGATATrAAAACTGTACTrCAGGAGATTAAGCGAGGAAAAAGGATGAA- 3'
Allele In PhfOKO 5' - GAGATrCGATATTAAAACTGTACTTCAGGAGATTA-GCGAGGAAAAAGGATGAA- 3'
Phf6KO
F
1.5x10
7
-
1.x107 -
E 5.Ox10e-
z
]n.s
o.oDa
Days
** n.
50. r i r i r- r i rai rILL
-" Ph.f6 w
- Phf6KO 40-
S20-
. E
mm1
GO/G1
5
X OX OX
PHF6
HSP90
H
z
E
4e
CO0)
Io
2.5-
2.0.
1.5.
1.0.
0.5.
0.0.
Phf6T Phf6o
+ EV
+cDNA
+ EV
+ cDNA
Supp. Figure S2.1. Generation Phf6-null B-ALL cells and characterization of its effect on B-cell lineage leukemia in vitro. (A)
Cells were transiently transfected with non-integrating CRISPR-Cas9 plasmid that expressed a Phf6-specific sgRNA molecule along
with Cas9. Isogenic lines were generated and characterized by DNA sequencing, western blotting and qPCR. (B) (Left) Diagram of
PHF6 protein and corresponding Cas9-sgRNA targeting site (ePHD2). (Right) Western blotting analysis of Phf6wr and Phf6KO cells
with antibodies (Ab) that recognize the N- and C-termini of PHF6 protein. aa, amino acid; NLS, nuclear localization signal; NoLS,
nucleolar localization signal; ePHD, extended PHD domain. (C) qPCR analysis of Phf6 mRNA levels in Phf6wr (blue, n=3) and
Phf6KO (red, n=3) cells. (D) DNA sequence of X-linked Phf6 mutant allele (in male Phf6KO cells) in comparison to wild-type allele in B-
ALL cells. Blue line, deleted base; red bases, protospacer adjacent motif (PAM) sequence; black line, sgRNA sequence. (E) Results
of a cell proliferation assay are shown for Phf6wlr (blue) and Phf6KO (red) (n=4). (F) Cell cycle analysis by BrdU incorporation of
Phf6wr (blue) and Phf6KO (red) cells in vitro (n=3). (G) Western blotting analysis of Phf6wvr and Phf6KO cells expressing Phf6 cDNA.
EV, empty vector. (H) qPCR analysis of Phf6 mRNA levels in Phf6wr (blue) and Phf6KO (red) cells expressing Phf6 cDNA (n=3). Data
represent the mean SD in FH. Statistics were calculated with two-sided Student's t-test: *P < 0.05, **P < 0.01, ***P < 0.001, ****P
< 0.0001, n.s. = not significant. [Soto-Feliciano et al. 2017; Soto-Feliciano 2016].
129
N( 0
1 42
ter
N-term Ab
C
LD
C1
E
1.2
1.0
o.0
0.6
0.4.
0.2
Phfr6'
E
PhfS%+ EV
Ph6KO + EV
Phf6T+ cDNA
Ph6KO + cDNA
G
G2/M
600.
4(0
200-
**
-U
0)
Phf6wr phf6KO(Clone 2)
Phf6Wr Phf6o(cIone 2)
Bone Marrow 2
U)
0
0
400
3001
200
0
20-
10-
- E
Phf6"r Phf6KO(Clone 2)
0 U
alo
Phf6wI Phf6K(cIone 2)
Lymph Nodes
Supp. Figure S2.2. Phenotypic plasticity is consistent across multiple isogenic Phf6 knockout B-ALL cell
lines. (A) Representative image of recipient mouse transplanted with 106 B-ALL cells of a second Phf6-knockout
isogenic cell line. Arrows indicate grossly enlarged lymph nodes. (B) Quantification of combined LN (left) and spleen
(right) weights of Phf6Wr (blue, n=5) and Phf6KO(Clone 2) (black, n=4) recipients. (C) Bar graphs showing the
percentage of the CD4+ fraction among mCherry+ cells isolated from Phf6vvr (blue, n=9) and Phf6KO(Clone 2) (black,
n=5) tumors in bone marrow (left) and lymph nodes (right). Data represents the mean SD in B-C. Statistics were
calculated with two-sided Student's t-test). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, n.s. = not significant.
[Soto-Feliciano et al. 2017; Soto-Feliciano 2016].
130
A
C
E
0
0
C
::3
W.
AI
I
I
I
9 TS 6 % 
c o t s9 P 0 4
in ~ 846% 2.89%
90.5% 44.4%
F5C.A ---- FSC-W --- 55cW- FSC-A --- FSCA --
C
M PHF6
4 - HSP90
0
I'
121
10D
0.8.
061
o.2
D
(L
Phf6wT cells PhfeC0 cells
1LKj
IA14K
0
0
CD4 CD4
- Isotype control
- CD4
100-
50-
Phf6"l' shPhf6
Phf6Wn=14
-- shPhfB n=10
n.s
5 10 15 20
Time after transplantation (days)
E 0.
..
o 20-
0
Bone marrow
n.s
r _1
Phfi6vv
n.s
40
o20
shFhf6
Sn.s n.s
10C0--
75-
25m
B220+ CD19+
Lymph nodes
n.s
Phf6W
U
U
F
50-
CL
shPhf6
Phf6w BM
Ph5"d done BM
Phl6w LN
Phflr done LN
H
25-
20.
+ 15.
10.
ns na
m m_
M Y
CD4+
n..s
p 4 8 12 
lime after transplantation (days)
- Phfd clone n=4
- Pht6wrn=14
16
"" Phl6'T+EV (n=5)
C3 PhrKO+EV (n=7)
al Ph16w+cDNA (n=3)
C3N Phl6KO+cDNA (n=3)
Phf6wT Phf6CO Phf6NT Phf6CO
Supp. Figure S2.3. Effect of Phf6 loss on B-cell lineage leukemia in vivo is not a result of clonal effects. (A) Gating
schemes and representative plots for flow cytometric and cell surface staining analyses for B-ALL cells (Phf6wr-top and Phf6KO-
bottom): gating of lymphocytes based on morphology (left) and then gating of leukemia cells of mCherry+ cells (middle) and
analysis of surface expression of CD19, B220, IgM and CD4 on those cells. (B) Histograms of CD4 cell surface expression on
Phf6wr (left) and Phf6KO (right) B-ALL cells maintained in culture. In each plot, pink histogram represents staining with isotype
control and grey histogram represents staining with anti-CD4 in wild-type (left) and knockout (right) cells. (C) (Left) Western
blotting analysis of Phf6wr and shPhf6 cells to determine PHF6 protein levels. (Right) qPCR analysis of Phf6 mRNA levels in
Phf6wr and shPhf6 cells (n=3). (D) Kaplan-Meier survival analysis of mice injected with 106 Phf6wT (blue) and shPhf6 (green) B-
ALL cells. The number (n) of mice per genotype analyzed is shown. Statistical analysis (log-rank test, Mantel-Cox) was
performed for recipients injected with shPhf6 cells in comparison to recipients injected with Phf6wr cells. (E) Frequency of
surface CD4+ cells among mCherry+/Phf6wr (blue) and mCherry+/GFP+ shPhf6 (green) leukemia cells isolated from bone
marrow and lymph nodes of recipients. (F) Kaplan-Meier survival analysis of mice injected with 106 clonal (grey) and non-clonal(blue) populations of Phf6wvr B-ALL cells. Statistical analysis (log-rank test, Mantel-Cox) was performed for the clonal cell
recipients in comparison to non-clonal cell recipients. The number (n) of mice per genotype is shown. (G) Frequency of B220+,
CD1 9+ and CD4+ cells among mCherry+ leukemia cells isolated from bone marrow and lymph nodes of recipients injected with
Phf6wr clonal (grey, n=3) and non-clonal (blue, n=3) cells. (H) Bar graphs showing the percentage of the CD4+ fraction among
mCherry+ cells isolated from Phf6wrr (blue) and Phf6KO (red) tumors in the bone marrow after transduction with a vector
expression empty vector (EV) or Phf6 cDNA. Data represents the mean SD in G. Statistics were calculated with two-sided
Student's t-test). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, n.s. = not significant. [Soto-Feliciano et al. 2017; Soto-
Feliciano 2016].
131
,,In =A - b-
0
A Phf6O Phf6O shPhf6 Phf6WT(Clone 2) (Clone 1)
-10 -05 0.0 0.5 1.0 16
Standardized log2(Expression)
Supp. Figure S2.4. Complete loss of Phf6 in B-ALL promotes a novel transcriptional program. (A)
Heatmap showing differentially expressed genes (FC > 4, FDR < 0.05) in pairwise comparisons comparing B-
ALL cells of different genotypes determined by RNA-Seq. Each column represents a replicate sample. The scale
corresponds to row-wise standardized log2-transformed expression values for each gene. [Soto-Feliciano et al.
2017; Soto-Feliciano 2016].
132
SUPPLEMENTAL TABLES
RNA-Seq GSE77456
ChIP-Seq GSE74878
ATAC-Seq GSE98716
B-cell Development Gene Sets
MORIPRE_BLLYMPHOCYT MORILARGEPREBILLYMPHOCY MORIIMMATUREBLYMPH MORILMATUREBLYMPHO HOFFMANNLARGETO SMALLPRE_11_LYM
EDN TEDN OCYTEUP CYTEUP PHOCYTEUP
B lymphocyted eeop ntl > Down-regulated genes in the B > Up-regulated genes in the B > Up-regulated genes in the B
signature, based on lymphocyte developmental signature, lymphocyte developmental lymphocyte developmental
expression prfilng of based on expression profiling of signature based on expression signature, based on > Genes up-regulated during differentiation from
lymphomas from the Emu-myc lymphomas from the Emu-myc profiling of lymphomas from the expression profiling of large pre-Bil to small pre-Bit lymphocyte.
troasgenic mice: the Pre-BI transgenic mice: the Large Pre-BiI Emu-myc transgenic mice: the lymphomas from the Emu-myc
stage. stage. immature B stage. transgenic mice: the mature B
ABLIMI
ADCY7
AIMI
AKIRINI
ATP6API
ATP6VOB
BIRC3
BLK
BTG1
BTG2
C22orfI 3
C4orf34
CAPN1
CASP9
CCNDBP1
CD2
CD22
CD72
CD74
CD79B
CLKI
CNR2
CYTIP
DIP2B
DTXI
EROLB
ERP29
FAM107B
FAM3C
FBXO11
FCGR2B
FCRLA
GNS
GPCPD1
HLA-G
IAH1
ILIORA
IL4R
JARID2
KLF2
LPGATI
LSP1
LY86
LYN
MCART6
MCL1
MS4AI
NBR1
NEATI
PTK2B
PTPRC
RALGPS2
RBM39
RGS14
RHOQ
SIPR1
SAMHD1
SAT1
TANK
TIRAP
TK2
TMEM50B
TMEM71
TRAF3
TRAF5
TRNT1
TSNAX
UBE2H
UBLCP1
UBPI
ZCCHC6
ZFP14
ZFP36
ADCY7
AIM1
ALDH2
BIRC3
BTGI
CAPG
CAPN1
CD22
CD40
CD74
CD79B
CD83
CNR2
CTSH
CTSS
CXCR5
DOCK2
DTX1
FAM5B
GGA2
GMIP
GNS
GPR65
HIP1R
HLA-DMA
HLA-DMB
HLA-DQAI
ILIORA
IL2RG
IL4R
IRF8
ITGAV
LAT2
LCPI
LPP
LTB
LY86
MACF1
MCARTS
MCL1
MS4A1
NEAT1
PIP4K2A
PTPN6
PTPRC
PXK
RASA3
RGS14
SAMHD1
SOATI
TAPBP
UBA7
VCAM1
WIPFI
ADCY7
ATP6VOB
BIRC3
BTGI
C22orfl3
CAPG
CAPN1
CD40
CD72
CD74
CD83
CNR2
CTSS
DTX1
DUSPS
EIF4A1
ERO1LB
GDPD5
GGA2
GMIP
GNS
GPCPD1
GPR65
GRN
HCK
HLA-DMA
HLA-DMB
IF130
ILIORA
JARID2
LAT2
LIMS1
LY6D
LY86
LYN
MCART6
MCLI
MS4AI
NEATI
PIP4K2A
PML
PTPN6
PTPRC
PXK
RGS14
RHOQ
SAMHD1
SEMA4D
SEPW1
ABCA1
ADCY7
ADDI
ADRB2
AIM1
APOE
ARHGEF3
B4GALNT1
BARX1
C12orf57
C19of48
CD40
CD74
CD83
CERI
CMTM6
CNOT6L
CR2
CTSH
CTSS
CXCR5
DGCR6
EBS13
ETSI
FAM102A
FAM55B
FAM76B
FCER2
FOXD4
GGA2
GNPNATI
GPR65
HIATL1
HLA-DMA
HLA-DMB
HLA-DOA
HLA-DOB
HLA-DQA1
HLA-DRB1
HLA-G
ID3
IFNGR1
ILI ORA
IL1ORB
IL2RG
IRF8
JAK2
LAT2
LCP1
MAP2K1
MAP3K8
MCL1
MEF2C
MS4AI
MYCBP2
NCF4
NEAT1
OSBPLS
PEA15
PHTF2
PIP4K2A
PISD
POUGFI
PPM1M
PTPN6
PTPRC
PXK
RAB8A
RAPIGDS1
RASA3
RGS14
RMI11
SAMHD1
SDF2
SELL
SGPP1
SH3BP2
SLC4A7
SMAP2
SORLI
TBCID14
TMOD3
VPS28
WDR43
YTHDC1
ZFP36LI
AARS
ABR
ACTN4
ACY1
ANLN
ANP32A
ANP32B
ANP32E
ARLI
ATP5D
AURKA
AURKB
AZINI
BRCA2
BUBI
CALM2
CAMP
CASP3
CBX1
CBX5
CCDC99
CCNA2
CCNBI
CCNB2
CCNE1
CDC20
CDC25C
CDC45
CDCA3
CDCA5
CDK1
CDKNIA
CDKN2C
CDKN3
CENPA
CENPE
CENPL
CHAF1B
CIT
CKAP5
CKS1B
CKS2
COIL
CRIP2
CTC1
DBF4
DCTPP1
DDX19A
DLGAP5
ECT2
EDN2
EIF2AK2
ENO1
FAM111A
FENI
FLiI
GPHN
GSN
GSPT1
H2AFX
HAUS6
HELLS
HES1
HLA-DQB1
HLA-DRBI
HMGB2
HMGB3
HMMR
HOXC5
HSPA8
HrT
IDE
IL4
INCENP
IP05
ITGB7
KIAA0101
KIAA1524
KIF20A
KIF22
KIF23
KIF2C
KIF4A
KIN
KPNA2
LDHA
LIG1
MCM10
MCM5
MCMB
MELK
MIS18BP1
MKI67
MMP14
MTHFD2
MYEF2
MYH11
NASP
NELF
NME2
NT5DC2
NUCKS1
NUDC
NUSAPI
PCK2
PHF17
PIHIDI
PLK4
POLAI
PPIA
PPP1CA
PPP2CA
PRC1
PSATI
PSMC3
RACGAP1
RAD51
RAD54L
RAMPI
RAN
RBL1
RFC3
RFC5
RPAI
RPL10
RRM1
RRM2
SGOL1
SLC29AI
SMC2
SMC4
SNX2
SSX21P
STIM1
STMN1
TK1
TMPO
TOP2A
TRIM46
TRIP13
TTK
TUBA1A
TUBAIB
TUBA3C
TUBB
TUBB3
TUBB4B
133
I i I
NFKB1
NFKBIZ
PDE7A
PNRC2
SEMA4D
SERPINB1
SIPAI1
SNAPIN
SMAP2
TAPBP
TMEM50B
TRAF3
LPP
LY8O
MACF
MAN1AI
DNA2
DNMT1
DUT
E2F8
NCAPH
NCAPH2
NEDD4
NEK2
U2AF1
UBE2C
UBE2S
UBE2T
VPS72
WYDHD1
XPo1
ZNF358
BCR-ABL Signaling Gene Sets
KLEINTARGETSOF_BCRABLIJUSION HUANGDASATINIBRESl8TANCEUP DIAZ-CHRONICMEYLOGENOUS_-LEUKEMIADN
> Genes changed in pre-B lymphoblastic > Genes whose expression positively correlated with > Genes down-regulated in CD34+ [GenelD=947] cells Isolated from bone marrow
leukemia cells with BCR-ABL1 fusion sensitivity of breast cancer cell lines to dasatinib of CML (chronic myelogenous leukemia) patients, compared to those from nornal
[GenelD613, 25] vs normal pre-B lymphocytes. [PubChem=3062316]. donors.
ANPEP AGPS MAP7D1 ACTGI HAL RPL32
BLNK AKRIC3 MET ADA HERC5 RPL39
BTK ANTXR2 MSN ADAP2 HLA-B RPL41
CBFA2T2 ANXA1 MYO10 ADD3 HLA-C RPLP1
CD19 ARHGAP29 NDC8O ANK3 HLA-DMB RPS11
CD33 BTN3A2 NNMT ANXA5 HMHB1 RPS12
CD38 BTN3A3 OSMR BCL6 IF130 RPS15
CD40 CALD1 PALM2-AKAP2 BCR IL1 ORA RPS17
CD79A CAST PARVA BLNK IL13RA1 RPS24
CSF1R CAV PCDH7 BTG2 IL16 S10OA8
CSF3R CAV2 PDGFC CAPN3 IL7R SlOOA9
CXCR4 CCDC5O PGM1 CCR2 INSR SATBI
EBF1 CDC42EP3 PLAU CCR7 IRF7 SELL
GATA1 COL5A1 POPDC3 CD53 IRF8 SGCB
GYPC COTLI PPPIRI8 CDH2 ISG20 SH3BP4
IGBP1 DCBLD2 PRNP CEBPD JAM2 SLC16A7
IGHMBP2 DPYD PSMB8 CECRI KCNE1L SLC1A3
IKZF1 DUSPI0 PSMB9 CLEC4A KLF2 SORLI
IL3RA EGFR PTRF CLEC5A KLF4 STAG3L1
IL7R ELK3 RAC2 CRHBP LHFP TCF7
IRAK1 ELL2 RAl14 CRYGD LILRB3 TERT
IRF4 EPHA2 REXO2 CSF1R LILRB4 TGFBI
ITGAL EPHB2 RIPK4 CST7 LPL TLR2
MLF2 EXTI SAMD9L CSTA LY86 VAMP5
MNDA F2RL1 SH3KBP1 DACT1 MAFB VPREB1
MPO F3 SNA12 DFNA5 MCL1 ZNF124
MYD88 FSTL1 SOCS5 DNTT ME1
MZF1 FXYD5 ST5 EEF1A1 MN1
NCF4 GBP3 TFPI ELANE MNDA
NUP98 GNG12 TGFBI EPB41L3 MPO
PAX5 IF116 TGFBR2 F13A1 MS4A4A
PML IFIT3 TNFRSF21 FAAH MS4ABA
POU2AF1 IL15RA UBE2E3 FGFR1 MVP
PROM INPP1 UPPI FGL2 MYLK
PTPRC rTGA5 ZNF559 FLNB MYO5C
RAG1 JAG1 ZYX FLT3 MYOF
RAG2 KCTD12 FZD2 NINJ1
RUNX1 LAMB3 GADD45B NKG7
SEPT9 LARP6 GAPDH OAS2
SIGLECS LAYN GAS7 PRNP
SLC22A2 LEPREL1 GATA3 PROMI
SPIB LIMAl GLIPRI PTPRE
SYK LOC346887 GLTSCR2 RPLI8A
TCF3 LYN GPR137B RPL23A
TNFRSF13B MAML2 GPX3 RPL3O
NF-kB Signaling STAT5 Signaling RUNXI targets
HINATANFKBTARGETSKERAT SANA_TNF_SIGNAUNG UP WIERENGASTAT5A_TARGETSGROUP1 TONKSTARGET_0F.RUNX1_RUNX1TFUSION_HSCDN
INOCYTEUP
> Genes up-regulated in primary > Genes up-regulated in five > Genes up-regulated to their maximal levels in
keratinocytes by expression of p50 primary endothelial cell types CD34+ [GenelD=947] cells by Intermediate activity > Genes down-regulated In normal hematopoietic progenitors by
(NFKB1) and p65 (RELA) (lung, aortic, iliac, dermal, and levels of STAT5A [GenelD=67761; predominant long- RUNX1-RUNX1T1 [GenelD=861;862] fusion.
[GenelD=4790;5970] components of colon) by TNF [GenelD=7124]. term growth and self-renewal phenotype.
NFKB. ____________ ___________________________________________
ARHGDIA
ARHGDIB
BCL2A1
BIRC2
BMP2
CCL2
CCL20
CCL5
CCNH
CD83
KRT14
KRT6B
KRT7
LITAF
MAP2K3
MAPK9
MMP1
MMP14
MMP9
MT3
ANO9
APOL
APOL3
ATP13A3
BIRC3
BPGM
BST2
C1QTNF1
Ci s
CASP1
INHBA
ITGAV
KIAA1147
LAMB3
LGALS3BP
LGALS9
LIPG
MMPIO
MMP3
MX1
A4GALT
ADM
AGPAT9
AMACR
ARL4A
BATF3
BBX
C14orf49
Clorf51
C5orf32
FNDC3B
GATA2
GBP4
GGT5
GPRI14
GYPB
HBZ
HERC2P2
HSPA2
HSPA6
PLIEKHA4
PLVAP
PLXNA3
POU2F1
PPFIA4
PRPS1
RABGAP1
RAPIGAP
RASGEF1A
RHAG
ABC04
ACSL1
ACTN1
ADCY7
ALOX5AP
ANK1
ANPEP
APOBR
AQP3
ARHGEF1 0
CYPIB1l
CYP51A1
DOK2
DUSP10
DYRK2
E2F8
EGR3
EHD3
ELOVL6
EMR2
JUND
KBTBD11
KCNH2
KCNK5
KIAA0182
KLF1
LAPTM5
LAT2
LCP2
LDLRAP1
RMNDDB
RNF125
RPS6KA1
RUNX3
S100A4
S10OA9
S100B
SASH3
SELPLG
SETX
134
CD9
CDH3
CDKN1A
CFLAR
CSF2
CXCL1O
CXCLI1
CXCL3
CXCL6
CYP27B1
EFNAI
EGFR
ERCC1
ESMI
FNI
GABARAPLI1
GADD45A
GBP1
GPRC5B
GSTO1
ICAM1
IFNAR2
IL15
ILIA
ILI B
ILIRN
IL32
IL6
IL7R
IL8
IRFI
ITGB6
KRT10
MTSSI
NFKB1
NFKBIA
NPR1
OLR1
PDPN
PLAT
PLAU
PNRC1
PTX3
RACI
RAC2
RARRESI
RELA
RND3
SAA1
SERPINB2
SERPINE1
SLC7A2
SOD1
SOD2
SPRR1B
STAT5A
TAPI
TIAM1
TNC
TNF
TNFAIP2
TNFAIP3
TNFAIP6
TNFAIP8
TNIP1
TRAF1
VEGFA
CCL11
CCL2
CCL20
CCL7
CCLO
CMPK2
CSF1
CTHRC1
CX3CL1
CXCL10
CXCLI1
CXCL2
CXCL3
CXCL6
CXCR7
DDX60
DNAJAI
DRAM1
FTH1
GBP1
HLA-A
HLA-B
HLA-C
HSD17B11
ICAM1
ICOSLG
IF130
IF144L
IFIHI
IFIT1
IL32
IL7R
IL8
MX2
NCEH1
NFKBIA
OASI
OAS2
OAS3
'OXR1
PARP14
PLA1A
RABL3
RAC3
RHOB
RIPK2
SAMD9L
SAMHD1
SATI
SERPINE1
SLC15A3
SLC7A2
SMAD3
SOD2
SPAG9
SSPN
TAPBP
TLR2
TNFAIP2
TNFAIP3
TNIP1
UBD
VCAM1
C7orlS8 IGFBP4
CALB2 IGFBP5
CBS IL18RAP
CCDC102A IL3RA
CD1A IL411
CD40LG IL4R
CDHR1 ITGA2
CDKN1C KIAA0226
CERCAM KISSIR
CFH KRBA1
CISH LINCO0341
CPXMI1 LTB
CST7 MAPIA
CTNS MEX3D
CTSH MIIP
CTSZ MOSPD3
CTXNI MTHFD1
DAB2 MUCI1
DAPK1 MYOIO
DCPS NAAA
DDIT4 NAALADLI
DDN NCS1
DHRS3 NDRGI
DPYSL4 NOD1
DUSP5 NT5C3L
ENC1 NUDT4P1
ENPP3 OSBP2
F2RL1 P2RX4
FAM126B P4HA2
FAM129B PCSK9
FAM131A PFKFB4
FAM150B PHACTR1
FBXOB PHLDA1
FCER2 PIK3P1
FGF11 PIMI
RHEBL1
SAMSN1
SCHIP1
SEMA6C
SH3BP4
SLC25A37
SLC25A4
SLC2A1O
SLC6A9
SLC9A7
SMAP2
SNTA1
SOS1
SPAG4
ST3GAL4
ST3GAL5
STK17B
TARP
TGFBR1
TGM2
TJP2
TLE3
TMEFF2
TMEM116
TNFAIP8L3
TNFRSF4
TP531NP2
TRIB3
TRIM46
TRIM58
TULP3
UBAC1
UPKIA
VASN
XIRP1
ZNF609
CD4 T-cells Gene Sets
GSEI32 CD4_TCEL GSE12_CD4_CELL GSE132.CD4 TCELL GSIEJ07_NAIVE_V_ GSE398JMEMORYC GSEII07_NAIVECD4 GSE228&0_NAIVECD4
_SE110325S EL _SB0 C L _VSE 4_M lEI MEMORY CD4_JCELL 04_TCELL_ VS_BCELL _VS_PBMCCD4_TCEL _TCELLVSMONOCYL.VS-BCELL-DN VYSB1CELLDN _VSJMYIELOID_DN D NLNTD
> Genes down-regulated > Genes down-regulated > Genes down-regulated > Genes down-regulated > Genes down-regulated > Genes down-regulatedin comparison of healthy in comparison of naive In comparison of healthy > Genes down-regulated In comparison of In comparison of naive T in comparison of naive
cets vnersshelh CD4 [Genel0D920] T CD4 [GenelD=920] T In compas n T memory CD4 cells versus peripheral C14 [GenelD=920] T
C15 [GenelD=920] B cells versus naive B cells versus healthy cells [GenelD=920] T cells blood mononuclear cells cells versus day 0C1 Bcells. myeloid cells. c versus B cells. (PBMC). monocytes.
ABCB4 KIAA0182
ADAM19 KIAA0226L
ADAM28 KMO
ADK LAMC1
AIM2 LARGE
ALOX5 LAT2
ANXA4 LHFPL2
ARHGAP1 LMO2
7 L0
ARHGAP2 LOC10029
4 0557
ATPD LOCI0029ATID 3440
84GALT1 LOC652493
BANK1 LOC96610
BASPI LRMP
BCL11A LTA4H
BCL7A LY86
BENDS LYL1
BLK LYN
BLNK MAP3K14
BTK MARCH1
C11orf24 MARCKS
CACNA1A MEF2A
CD180 MEF2C
CD19 METTL7A
CD1C MICAL3
CD1D MIR600HG
CD200 MS4A1
CD22 MTSS1
CD24 MZB1
CD40 NCF1C
ADAM19 1L60
ADRBK2 L0010029ADB2 3871
ANKRD34 LOC72814
C 2
APIG1 LY86
ATP6VOA1 MAGEA3
ATXN3 MAP1B
BANKI 3-Mar
BCL11A MCM2
BET1 MCTP2
BLNK MEF2C
BTK MGC2889
C12orf49 MIR600HG
C14orf118 MKLN1
C1Sorf29 MOXD1
ClorfI29 MPZL1
C7orl68 MS4A1
CCP11O MTMRIO
CD19 MTMR9
CD200 MYEF2
CD22 MYO1B
CD72 MYD1E
CD79B MYOZ3
CD80 NEBL
CDK14 NEK4
CHAT NIPSNAP3CHT B
CHL NODI
CLCN4 PARG
CLEC4A PARMI
CNR1 PDE1SA
ACTGI LRPI
ADM LTA4H
ADORA2B LYBe
ADRBK2 LYLl
ALDH2 LYN
ALOX5 LYST
ANXA4 MAP2K3
APLP2 MARCH1
AQP9 MARCKS
ATP10D MEF2C
ATP6VOAI MEGF9
BASP1 METTL7A
BCL6 MSRI
BTK MTMR11
Clorf38 MTMR14
C20orr24 MYD88
C5orfl3 NAGK
CARD9 NCF1C
CBFA2T3 NCF4
CCDC88A NCOA4
CCRI NFE2
CDIC NFKBIE
C01D NOTCH2
CD86 NPL
CDK2API NUDT3
CDKN1A OAZ2
CDKN1C OPN3
CEBPB ORA13
CHST15 OSSPL11
ACOT9 MAF
ACTB MAP3K5
ACVR1 MCOLN2
ADAMIS MLAT
AHNAK MIB1
AHR MIS18BP1
AKIRIN2 MLF1
ALCAM MYBL1
AMMECRI MYLe
ANXA1 MY01F
ANXA2 MYO5A
ANXA4 NBEAL2
AQP3 NCOA7
ARHGAP1 N0D2
ARHGAP3 BFC2A
ARLEIP1 OBFC2B
ARPCIB OGOH
ATP2B1 OPTN
ATP2B4 OSBPL3
ATXN1 PDIA6
AUTS2 PDPI
CAPN2 PEA15
CASK PFKP
CBLL1 PHACTR2
CCR6 PHTF2
CD58 PIEZO1
CD74 PLP2
CD82 PLXNC1
CD84 PMAIP1
ACP5 LMO2
ADD1 LOC65249
ADK LY86
ADO LYN
AGPAT5 LYST
ALDH2 MAP4K2
ALOX5 MAPK12
ARHGAP1
7 MCM5
ARHGEF7 MED14
BACH2 MEF2C
BANKI METTL7A
BASPI MICAL3
BCL11A MICALLI
BHLHE41 MRPL15
BLNK MRPL40
BMP2K MS4A1
BTN2A2 MZBI
C20or1O3 NADK
C~orf15 NASP
C7orf23 NT5E
CA2 NUMB
CACNA1A OAS1
CAT OPN3
CCNBIIP1 ORCS
CCT2 OSBPL1O
CD180 P2RX1
CD1D P2RX5
CD200 PAPSS1
CD38 PARM1
ACER3 MCTP1
ACTB MEIS1
ACTN4 MIR22HG
ADAM8 MPP1
ADAMTSL MPST
4 MS
AHR MTHFD2
AIF1 MYL6
AKIRIN2 NAGA
ANXA4 NAGK
AP3BI NAMPT
ARHGAP2 NBN
ARPCIB NFIL3
ARSB NOTCH2
ATP2A2 OAZ2
ATP6VOB OSBPLS
B3GNT5 PDE4A
B4GALT5 PDGFC
BCKDK PDLIMI
C11orf82 PEA15
C22orf13 PGD
C3AR1 PHKB
C9orr72 PIPSKIB
CAPG PLEKHG3
CAPNS1 PLEKHO1
CARD9 PLSCRI
CAT PLXND1
CATSPER POU2AF1
CCDC50 PPP4C
CD244 PRR8L
ABHD5 KCTD12
ACSL4 KIAA1033
ADAP2 KLF10
AKR1A1 KLF4
ALOX5 LAT2
ANXA2 LGALS3
ANXA2P2 LILRA2
AP1S2 LILRB1
APOBEC3
A LILRB2
ARHGAP2
AH LILRB3
ARHGEF1
01 LY96
ASAH1 LYZ
ATP10D MANBA
ATP6VOB MARCKS
ATP6VOD1 MEDS
ATP6V1A MEGF9
BCL6 METTL9
BID MFSD1
C11or75 MGATI
CSARI MICAL2
CAPG MNDA
CAPZA1 MTMR14
CASP1 NADK
CCRI NAGK
CD14 NCFIC
CD36 NCF2
CD93 NCOA4
CDC42EP NEU
CEBPD NFE2
135
ASRGL1
BCOR
BIN2
BIRC3
C11oI21
C15orf39
CI8or1I
CAMK1
CAPG
CBFA2T3
CCL2
CCNA1
CCNDI
CCND3
CCR7
CD244
CD33
CD36
CD48
CD52
CD55
CDB2
CEBPA
CEBPE
CFLAR
CHST12
CIAOI
CKAP4
CLC
CRCP
CSF1
CSF2RB
CST7
CSTA
CTSG
CXCR4
CYBB
EPB49
EXOSC4
FADS1
FADS2
FAM30A
FCER1A
FCGR2A
FHL2
FLOT2
FPR2
FUT7
GALNT6
GATA1
GATA2
GFIIB
GMPR
GPR183
GPR35
GSR
HBB
HBBP1
HBD
HDC
IGFBP2
IGFBP4
IGFBP5
IL1ORA
IL17RA
IL1 RN
ILSR
IL7
IL7R
IL9R
IRF8
ITGA2B
ITGA6
ITGAL
LGALS1
LSTl
LTB
MAGEFI
MBP
MGAT1
MICAL2
MMP2
MS4A3
MYCN
MYL4
NCF4
NEFH
NFE2
NKG7
NOD2
OGG1
P2RX5
PDLIMI
PDXK
PIK3CB
PLAC8
PLCH1
PLP2
PLXNC1
PMP22
PPAP2A
PRG2
PRKCD
PTPN12
PTPN22
RAMP1
RARA
RASSF2
RBM38
RGS14
RGS14
SH3BGRL3
SLA
SLCI6A3
SLC16A7
SLC1A4
SLC27A2
SLC43A3
SLC48A1
SLC7A5
SLCO3A1
SLCO4C1
SMOX
SOCS1
SPit
STAB1
STK10
TARP
TEAD3
TESC
TFR2
TGFBI
TIMP3
TMEM166
TNFSF13
TPCN1
TPM1
TRAF4
TRBC1
TSC22D1
TUBA4A
TYROBP
XK
XYLT1
ZBTB16
ZFP36L2
ZNF787
CD72 NT5C
CD79A OPN3
CD79B ORA13
CD83 OSBPL10
CDK14 P2RX5
CHST15 PAX6
CITA PCDH9
CLCN6 PDLIM1
CLIC4 PKIG
CLIP2 PLAC8
COCH PLCG2
CR2 PLEKHF2
CSDA PNOC
CXCRS POU2AF1
CYBB PRKCB
DDAH2 PTPN6
DENND3 PTPRK
DHTKDl QRSL1
DNMBP RAB30
DTX4 RABEP2
DUS2L RASGRP3
DYRK4 RFX5
EAF2 RHBDF2
EHD3 RHOBTB2
ENTPD1 RHOO
FADS3 RNASE6
FAM30A RNF141
FAM3C RRAS2
FCER2 SCARB1
FCGR2B SCPEPI
FCGR2C SELIL3
FCHSD2 SETBP1
FCRL2 SH2B2
GFOD1 SHMT2
GGA2 SIPA1Ll
GM2A SKAP2
GNG7 SLC1SA3
GUSBP11 SLC17A9
H2BFS SLC2A6
HDAC9 SLC7A7
HHEX SMAGP
HIST1H2B SNX3K SX
HLA-DMA SP140
HLA-DMB SPIB
HLA-DOB STAG3
HLA-DPAI STAP1
HLA-DPB1 STX7
HLA-DQA1 SWAP70
HLA-DOBI SYBU
HLA-DRA SYK
HLA-DRB1 SYNGR2
HLA-DRB6 SYNPO
IFNGR2 TBC1D5
IGHAl TBC1D9
IGHD TCF3
IGHM TCF4
IGJ TCLlA
IGKC TCTN1
IGKV1D-1 TFEB
IGKV4-1 TIMELESS
IGLO TLE1
IGLJ3 TLR7
IGLL3P TPD52
IGLV1-44 TRAKI
IGLV3-19 TSPAN13
IRF4 TSPAN3
IRF5 TUBB6
IRF8 VAV3
JUN VPREB3
KIAAOD40 WASFI
KIAA0125 ZMIZ2
CNR2 PDLIM1
COLEC10 PEGlS
COLEC12 PHLPP2
CORO2B PHTF1
CPBl PIPSKIB
CPEB3 PLCG2
CR2 POU2AF1
CSDA PPEF2
CTNNALI PPMID
CXorf57 PRPH2
CYP4B1 PTGSI
DCAF17 PTPRK
DDRI RAD51D
DNASE1L RHOBTB1I3
DPF3 RIC3
DPPA4 RMI1
DRAM1 RNF219
DTX4 RPS27
EHD3 RRAS2
ENTPD1 SBNOI
EV15 SCD5
FAM65A SEMA4F
FBX04 SERPINBE
FGFR1DP SETBP1
FMO3 SLC15A3
FMOE SLC2A5
FPR2 SLC35F2
FRAS1 SLC4A4
FZD10 SLC6Al6
GABBR1 SLC9A7
GALNT3 SMADE
GATM SMPX
GGA1 SOBP
GM2A SOCS5
GPM6A SPOCK1
GSTA1 STRN3
GSTA4 SYK
GTPBP2 TBC1D312
GTPBP3 TCF4
HAUSS TCLIA
HDAC9 TFP12
HESXI TGFBR1
HHEX TIAM2
HIST1H4D TIMELESS
HLA-DMA TLL1
HLA-DMB TMOD1
HLA-DPB1 TNS3
HLA-OQBI TOMM34
HLA-DRA TREML2
HLA-DRBI TRIO
HMGCS1 TRMT2B
HYDIN TSHB
ID12-ASI TSPAN13
IER31P1 TSPYL5
IGHM TULP2
IL24 UGODH
IQCB1 USP2
IRF4 UST
IRF6 UTP14C
IRF8 VAV3
ITGB3 WEE1
KCTD20 WWAX
KIAA0125 ZC3H7B
KIAA0889 ZNF235
KLHL35 ZNF468
KRT19P2 ZNF532
KYNU ZNF536
LARGE ZNF552
LHFPL2 ZNF682
LINCO0339 ZNF821
LOC0028 ZNRD1-
CLCNE P2RX1
CLEC7A P2RY13
CLIC4 PAD12
CLPB PEA15
COL9A2 PHC2
CORO1C PLAUR
CSDA PLCG2
CSF1R PLEK
CSF2RB PLEKHO1
CTBP2 PLXNCI
CTNNA1 PRCP
CTSH PRKCD
CXCR2 PSEN2
CYBB PTGS1
CYP1B1 QPRT
DENND3 RAB31
DGKG RAB32
DHRS9 RBM47
DSE REPS2
DYSF RGS2
EAF2 RHBDF2
EMR2 RHOG
ENCI RIN2
ENTPD1 RNASE6
FAM110B RNF141
FAM114A1 RNPEP
FAM49A ROGDI
FCGRIB RPH3A
FCGR2A RUSC2
FGL.2 RXRA
FGR SAT1
FIG4 SCARBI
FLJ11235 SCPEP1
FTH1PE SCRNI
GAB2 SERPINF1
GCNTi SH3BP2
GFOD1 SIDT2
GM2A SIRPA
GSN SKAP2
HCK SLC1lAI
HEXB SLC1EA3
HHEX SLC16A3
HLA-DMA SLC17A9
HLA-DMB SLC2A6
HLA-DPAl SLC31A2
HLA-DPBI SLC43A3
HLA-DGB1 SLC6A12
HLA-DOB2 SLC7A7
HLA-DRB1 SNX1O
HLA-DRB6 SNX27
HMGB3 SPINTI
HSBPI STX7
HSPA6 SUOX
HTRA1 SYK
ICAM1 TBC1D9
IF130 TESC
IFNGR2 TLR5
IL13RAl TLR6
IL22RA1 TLR7
IRF TNFSF13
IRFE TNS3
ITGAM TRAK1
ITGAX TRIBI
KCNMB1 TSPAN4
KMO TUBBS
KYNU UPK3A
LAT2 VENTX
LHFPL2 VNN1
LILRA2 WARS
LILRB3 ZDHHC7
LMO2 ZNF193
CD99 PRCI
CDC42EP PRDX1
CDCA7 PREX1
CHST11 PRKCD
CHST7 PSMD8
CLDND1 PTTG1
CLICI PYHIN1
CORO1B RAB11FIP
COTL1 RAB27A
RAFIGAP
CRIP1 2
CRYBG3 RAPGEF1
CTSA REEP3
CD72 PARP1
CD74 PAX5
CD79B PDLIM1
CDK14 PECAM1
CEPT1 PIPSK1B
CHFR PKIG
CKAP4 PLCG2
CLEC4A POU2AFI
COBLLI PSEN2
COL9A3 PSMB6
COQ2 PTK2B
COX15 PTPN6
CXCR3 REEP5 I COX17 PUSi
DEGS1 RFTN1
DHRS7 RILPL2
DIAPH2 RNF135
DUSP10 RNF139
DUSPE RNF19B
EEPD1 RNF214
EHD4 RORA
ELOVLE RSUl
EMB SIOAII
ENDOD1 S100A4
EPHA4 SEC11C
EPS15 SELIL3
ETV6 SH2D1A
EVI2B SH3BGRL
EZR SIAH2
FAM129A SLC2A3
FAM164A SLC35D2
FAM45A SLCSA6
FAR2 SLC9A3R1
FAS SLCA9
FBXL8 SMAD3
GATA3 SMAPI
GLBI SMO
GLIPRI SNX1O
GNA12 SNX24
GOLGA7 SPPL2A
GOLGBI SORDL
GPR183 SRGN
GPRIN3 SSR3
GSTK1 STSSIA1
GZMK STOM
HERPUD1 SYNE2
HLA-DRBI TAGLN2
HMGB2 TANK
HN1 TBCB
HNRPLL TBK1
IF116 TGFBR3
IL10RA TIFA
IL2RB TIGIT
IOGAP1 TIMP1
ICGAP2 TMEMS4
ITGB1 TNFRSF1
KIAA0182 TNFRSF4
KIAA0247 TNIP1
KIAA1324 TOM1
KIFIB TOR3A
KLF6 TPE31NP1
KLRBI TPM4
LDHA TRADD
LGALS1 TTC39C
LGALS3 TTYH2
LIMS1 TUBB4B
LIMS3 UBL3
LOC10028
8693 VDAO1
LOC0050 W
7564 WEEl
LOC33862 YWHAH
LPP ZBPi
LRIG1 ZBTB38
COX4NB QRSLI1
CSDA RAB31
CSNK2A1 RABGEF1
CTNNA1 RASGRP3
CTSH REXO4
CYBB RHOB
DGKD RUFYI
DLAT S10OA8
DUS2L SAV1
E2F5 SCN3A
EAF2 SCO2
EIF2AK3 SCPEP1
EIF2B1 SEC23B
ERCC1 SEL1L3
FAM30A SERPINF1
FCER2 SHMT2
FCGR2C SIDT1
GABARAP SIDT2
GM2A SIPAIL3
GNA12 SLC15A2
GNG7 SLC15A3
GSN SLC37A1
GUSBP11 SNRNP25
HESX1 Sp110
HHEX SPIB
HIST1H2A SQLEC
HISTIH2B STAPIH
HIST1H2B STX7K
HLA-DMA SWAP70
HLA-DMB SYK
HLA-DOB SYPL1
HLA-DQB1 TBCID9
HPS5 TCF4
HS3ST1 TERF2
HSDl7B4 TFEB
IGHAl TLE1
IGHD TLR7
IGHM TMEM14B
IGJ TMEM156
IGKC TMEM62
IGKVID-1 TMEM70
IGKV4-1 TMUB2
IGLO TNFRSF17
IGLL3P TOR3A
IL13RAl TRAK1
IL4R TSPAN13
ING1 TSPAN3
IRAK1 UBE2J1
IRF4 UROD
IRF8 USP22
ISG20 WARS
ITPRl XIST
JUN ZBTB5
KCNN4 ZDHHC14
KIAADD40 ZFP106
KIAA0125 ZNF165
KIAA0226 ZNF232
LAT2 ZNF318
CD9 PTGS1
CDK2AP1 PTPN12
CDKNIA QSOX1
CEBPB RAB1O
CEBPD RAB27A
CFD RABBA
CHP RALS
CLTA RASGRP4
CLTC REL
COPA REPS2
CSF2RB RGS2
CTSA RHOC
CTSZ RHOU
CYBA RNF130
CYPIBI RNF144B
DENN03 RNF149
DGAT2 RTN1
DGKG SAP30
DNAJB11 SCO2
DPYD SECTMI
DUSP22 SESTD1
DUSPE SETBPl
DYSF SGPL1
EFHD2 SH2B3
SH3BGRLEMP3 3
ENG SLCIA4
EPAS1 SLC22A15
FABP5 SLC31A2
FEZ2 SLC37A2
FNDC3B SLCA
FTH1 SLC7A7
FTHIPS SLFNll
FUCA2 SNAP23
FUT7 SNDI
GABARAP SPOCKI
GALC SRC
GALNT1O SRXN1
GD12 SSR3
GLUL STOM
GNA12 STS
GNGT2 STXBP2
GNLY SYT17
GRAMD4 TBClD8
GRB2 TEP1
GSR TGFBlI1
GSTO1 TKT
GSTP1 TLR6
GUCY1B3 TM6SF1
HAL TM9SF2
HERPUDI TMCC3
HISTIH2A TMED5
HLA-DPAl TMEM154
HVCN1 TMEM40
IL18RI TMEM59
IRF2 TNFSF13
JAK2 TP531NP1
KIAA0182 TPST2
LAP3 TRAK1
LILRA6 TTYH3
LILRB3 UBE2E1
LIMS1 UBE2J1
LOC10012 USP159034 USE
LOC10028 VAMP38617 AP
LOCIO050 VCL7632 VO
LOC20327 WDFY44
LSPI WDR11
LTA4H YWHAG
LY96 ZDHHC7
LYST ZMPSTE2LYT 4
MAP3K3 ZNF385A
MAP3K8 ZNF710
.5 .5 I ___________
CHST15
CKAP4
CLCN6
CLEC4A
CLEC7A
COQ2
CORO1C
CREG
CSDA
CSF2RB
CSF3R
CSTA
NIT1
NOD2
NOTCH2
NPC2
NPTN
NSF
NUP62
OAZI
PCTP
PDLIMS
PGD
PILRA
CTBP2 PISD
CTNNA1
CTSA
CTSB
CTSH
CTSS
CYBB
DAPK1
DIAPH2
DMXL2
DOK1
DPYD
DUSP3
DUSP6
E2F3
EIF4E2
EMR2
EXOC1
FAM49A
FBP1
FCERlG
FCGR2A
FCN1
FEZ2
FGL2
FGR
FKBP15
FLVCR2
PLBD1
PLCG2
PLSCRI
PLXNB2
PLXNC1
PPFIA1
PRKCD
PSAP
PTGS1
PTPRE
PTTG11P
QPCT
RAB31
RAB32
RIN2
RNF130
RXRA
S10OAll
S100A12
SIOAS
S100A9
SCPEPI
SDCBP
SEMA4A
SH2B3
SIGLEC9
SIRPA
FOS SLCllAl
FPR1
FTL
FUT4
GAB2
GABARAIP
GAPDH
GCA
GLB1i
GLRX
GLRX2
GLUL
GNG10
GNS
GPX1
GRN
HCK
HEXB
HK3
HSBP1
HSPA1A
IF130
IFNGR1
IFNGR2
IGSF6
ILER
ILK
IRAK3
IRF8
ITGAM
KCNMBl
SLC1SA3
SLC16A6
SLC27A3
SLC31A2
SLC7A7
SNX10
SNX11
STX12
STX7
SYK
TALDOI
TBC1D12
TBC1D2
TIMMIB
TIMP2
TLR2
TNFAIP2
TNFRSF10
B
TNFSF13
TNS3
TPP1
TYMP
UBA3
UBE2Dl
VAMP3
VCAN
VNN2
ZCCHC6
ZMIZI
136
6895 ASII
CDI 9+ B-cell Gene Set CD4+ T-celI gene set
Igi-VI 6330407a03ik Rhoq Cd39 Stesal AdCYG
H2-Abl Fcuda A530032dl5Rik Ud247 Ppml h Tnfrsfl a
Ly~d Btk Marcks Cd3d Icos SIC9a9
Ms4al 11411 A0324046 117r 221 DOI0c17Rik Btbdl4a
H2-Aa Ud180 Atp6va1 Tcau Art2b Cd200rI
H2-EbI PId4 Ccbp2 Trati Cd8bl Ud226
Sodi Hek Ud93 Igb4A830023pl2Rk A130090k4Rfk
Cd74 MOW~b Lrlcl 2610019t03Rik Fasi F2rll
Sink Ceacam2 Rkib E430004n04ik AdhI
We8 Myol e Myadm A530021J07 Cd4
Cr2 142-oa H2-DMa 2310032103Rk Gpr83
H2-DMbl Snx8 LrrclB Itk Tcrb-J
Lyn Hs3stl SWPep Prkcli Ldhb
PlacS na1 Hesi Tcf7 Tglbr3
SWk3 Rasgefl b Caspi Bcllb Cd3e
Fcer2a Irf5 Hipi r Let Tnfrsf25
Napsa Swap7O Tyrsbp Tcrb-vl 3 Ubesh3a
Rasgrp3 Rassf4 Igh-la Thyl P2r
Fa~m3 Wc4 Kcnk5 1700025g04Rik 2310010m24Rk
2010001m09ik 1112a DO300l1olORik Tnfrsl7 lkbke
Hhex Haws Bfsp2 Fyb ligpl
Banki A530023o14Rik Irf4 Bc021614 Tox
Tntrsfl 3c Bcar3 Z"088 Cd8 SytI2
RnaseS Ud36 U38 Ampdl Lck
Ctsh Ceacami Bhlhb2 Ppmlj Zfpnl1a2
Lrrk2 Sdc4 Snx9 Lcp2 A830098a13Rik
Cyp4fl 8 Plek Li~rb3 Als2cI Lefi
Serpnbia 10oSI Rgsl8 Cd5 Zdhhcl5
Ebfl Dok3 Ebi3 Nsg2 CodcS4
Syk Pou2afl OoSPl Stfnl Rab4a
C21e Ferl 13 39873 Txk ftm2a
Spib Cd4O Lynxl Art~c Cst7
Cd79a H2-ob mgstl HnrpI Vps28b
Cybb Serpina3g Chsyl Trrde Tbcld2b
Lmo2 Lat2 Nuak2 Sh2dla Ggt~al
Fgd2 Rab3O Cybasc3 Sloo3al Tanci
Fcull S100a BcD04728 Camk*4 4930431 b09Rik
Vpreb3 Myol c Gem Tesc 4632428n05Rik
Mefc Igi Slarfif Rampl C630004h02RIk
Unc93bl Fc15 Loc380823 MM&aO Tmem86
Pkig Ncfl Dennd3 Dzlpl S~c12a7
Cd24a Mpegl Casp4 AtlIha CtSW
EJ13 115ra AyNl 11 90002h23R~k A430lO7pOgRnk
Lyll PaxS Tm~sfl H2413 Arhgap29
Fcgr2b 5031414d18OR~k Ldl Kenmb4
Sig"ec Gentl Gata Rtn4il
Wd22 Plk3apl Emb TnfalpSll
Kmo Igh-%1558 SIcOaS Znrf3
Ncr2 5430427o19Rik K~kS Kirdl
1300014I08Rlk Cardl2 PdIlm4 Ms4a~b
Pipil Ogfwll Elovf7 Akrlcl2
Suplemnta.Tale.4 .-Enrche F Bidn Moif at Sk regon (fakn S~ of DE
Rank Target ID Raw Score P value Top Motif Prop
5 Mypg Mmusculus-JASPAR_20I4-Myog-MAO500.1 4.060423899 3.46E-24 0.225249773
6 Tclhp2a Mmusculus..jolma2O13-Tcfap2a-2 4.389340118 2.93E-22 0.198001817
7 AscI2 Mmuscszlus-UnIPROBE-AscI2.UP00099 4.293191424 4.25E-21 0.151880291
10 Tcfl 2 Mmusculus-JASPAR-2014-Tcfl2-MA0521.1 3.850738835 7.15E-19 0.204359873
11 Myodl Mmusculus-JASPARk214-Myodl-MA0499.1 4.267201711 8.18E-19 0.185288104
14 KIN4 Mmusculus-JASPAR-CORE-Kf4-MA0039.2 13.180162850 1.02E-17 0.202543143
15 Tef3 Mmusculus-JASPAR 2014-Tc3-MA0522.1 4.539535860 1.12E-17 0.185286104
16 EBFI Mmusculus-JASPARCORE-EBF-MA01 54.1 2.658401148 1.25E-17 0.144414169
18 Zfx Mmusculus-JASPAR-CORE-Zfx-MiA014.1 7.419709134 3.26E-1 5 0.181653043
19 Zfx Mmusculus-JASPAR_2014-Zfx-MA0148.2 7.413481155 5.20E-15 0.188194389
22 E2F3 Mmusculus-JASPAR;214-E2F3-MA0489.1 18.247273354 1.84E-11 0.181653043
23 Sp4 Mmusculus-UnPROBE-Sp.UPOOO2 21 .395855342 3.02E-11 0.158038147
25 TWep2Il Mmusculus-JASPAR-CORE-Tccp2I -MA0145.1 2.802808241 4.81E-11 0.124432334
137
Tc(cp211
Tcfe2a
Ascd2
EgrI
KIf7
NF-kappaB
Zfp740
Zfp161
Zfp740
Zbtb7b
Zic1
Spil
Zic2
Zbtb3
Mmusculus-JASPAR_- 2014-Tcfcp2il-MA0145.2
Mmusculus-UnIPROBE-Tcfe2a.UP00046
Mmusculus-jolma2013-AscI2
Mmusculus-UnIPROBE-Egrl.UP00007
Mmusculus-UnIPROSE-KIf7.UP00093
Mmusculus-JASPARCORE-NF-kappaB-MA0081.1
Mmusculus-joIma2O3-Zfp740
Mmusculus-JASPA_.2014-Ktfi-MA0493.1
Mmusculus-UnIPROBE-Zfpl6l.UP00085
Mmusculus-UnIPROBE-Zfp740.UP00022
Mmusculus-UniPROBE-Zbtb7b.UP00047
Mmusculus-UnIPROBE-ZicI.UP00102
Mmusculus-JASPAR_2014-Spi-MAOO8O.3
Mmusculus-UnIPROBE-ZIc2.UP00057
Mmusculus-UniPROBE-Zbtb3.UP00031
2.819675394
1.772525492
3.985152446
10.254008721
13.053070601
2.169104844
7.155212991
6.085493598
6.196856008
40.586223479
1.647896944
4.804087474
8.139399354
4.356653944
1.632787074
6.OOE-11
1.46E-10
1.60E-09
1.89E-09
6.84E-08
7.01 E-08
2.66E-07
2.55E-06
6.49E-05
0.003273146
0.007151548
0.007577186
0.008608833
0.009628762
0.009960612
0.132606721
0.122615804
0.137148047
0.158046412
0.156221617
0.116257947
0.151680291
0.118982743
0.111716621
0.110808356
0.079019074
0.171862125
0.089918258
0.17075386
0.108991826
Su pem na Ta l 5 ot ifcn ATA -e Hits_____
Gene P Value Bound by PHF6? LOg2 Fold Change
Zfp972 6.38E-15 -1.403906763
Pxdc2 1.05E-18 Y -1.362605695
Trbj1-1 1.62E-13 -1.213707541
Tnnt2 4.10E-13 Y -1.069853829
Trav2 2.85E-10 -1.04925368
Rasa72 1.62E-09 Y 1.006382536
syta 4.33E-10 1.01863963
Nkd1 1.96E-09 Y 1.039766243
Fam24c 1.52E-09 Y 1.04813747
Ctbp2 1.49E-09 Y 1.055835278
Adrb2 1.43E-09 Y 1.068249311
Chst7 9.52E-09 Y 1.073794828
Nespas 4.94E-10 Y 1.08375316
Rasil1b 1.46E-10 Y 1.088895071
Gm9949 5.25E-10 1.095099944
Creb312 2.28E-11 1.186955003
Usp53 2.87E-15 1.20613516
Map3k9 7.71E-12 Y 1.210159337
281041oL24Rik 1.78E-16 1.334650647
Trib2 2.14E-16 Y 1.349098408
Sec24d 6.19E-20 Y 1.472343845
Btg2 2.90E-23 Y 1.516190724
Hey1 8.06E-20 1.635179179
Jag1 2.39E-25 1.658741832
Rap2b 6.93E-30 1.699432974
Ano10 3.25E-50 Y 2.259049264
Zc3h12b 7.34E-40 2.324187644
Lrrc27 5.21 E-56 Y 2.461441613
Stk32c 6.28E-59 2.493524645
138
27
28
30
31
34
35
36
39
40
43
44
45
46
47
48
ACKNOWLEDGEMENTS
The work in this thesis is the result of a wonderful collaboration with Yadira Soto-
Feliciano and Yunpeng Liu. Many elements from the beginning of the chapter can also
be found in the thesis of Yadira Soto-Feliciano [2016]. This majority of this work was
also published in Genes & Development [Soto-Feliciano et al. 2017].
We thank S. Levine and S. Motola for RNA-Seq support, T. Volkert and S. Gupta for
ChIP-Seq support, K. Cormier from histology facility for technical support, and G.
Paradis, M. Griffin, M. Jennings and M. Saturno-Cond6n for flow cytometry analysis and
sorting support. We thank Alexey Soshnev (Allis lab) for help in the preparation of the
model figure. This work was supported by the Ludwig Center for Molecular Oncology at
MIT and the Cancer Center Support (Core) grant (P30-CA1405143). Y.M.S.F. was
supported by the National Cancer Institute of the National Institutes of Health under
Award Number F31-CA183405. J.M.E.B. was supported by the National Science
Foundation Graduate Research Fellowship under Award Number 1122374. YL. was
supported by the MIT Department of Biology training grant. This work was supported in
part by the Ludwig Center for Molecular Oncology at MIT, the Bridge Project (a
collaboration between the Koch Institute and the Dana-Farber/Harvard Cancer Center),
and Koch Institute Support (core) Grant P30-CA14051 from the National Cancer
Institute. The content is solely the responsibility of the authors and does not necessarily
represent the official views of the National Institutes of Health or National Science
Foundation. All RNA-Seq and ChIP-Seq datasets have been deposited into the NCBI
Gene Expression Omnibus (GEO) repository under BioProjectlD GSE74878. ATAC-Seq
dataset and processed TSS nucleosome signals are available in the GEO database
with accession number GSE98716, under super-series GSE77457.
139
MATERIALS AND METHODS
Plasmids and cloning
A mouse Phf6 sgRNA (5'-GTACTTCAGGAGATTAAGCG-3') was designed using the
Broad Institute sgRNA Designer [Doench et al. 2014] and cloned into pX458 vector
(pSpCas9(BB)-2A-GFP, Addgene #48138) (Ran et al. 2013). For generating knockouts
in the MLL-AF4 model, Phf6 sgRNA was cloned into pX330_mCherry vector (U6-
ChimericBB-CBh-hSpCas9, Addgene #42230) (5'-CACCGAGCGAGAGAGTAA
AAGCCG-3'). Mouse Phf6 cDNA, Tcfl2 cDNA, and NfKb cDNA were obtained from
OriGene (#MC203493, #MGC5345693, #MGC3157927), Btg2 cDNA (NM_007570.2)
was ordered via geneBlock, and sub-cloned into pMSCV-PGK-GFP. BFLS-mutant cDNA
was generated by site-directed mutagenesis, primers used (Forward/Reverse): (5'-
ATTTAATGCCGAGAAGGCAGC)/(5'-ATATGCAATTTCC CTCTTG). The NOTCH1-ICD
overexpression vectors were obtained from Addgene: pLPCX-NICD (Addgene #44471)
and pLPCX-IRES-GFP (Addgene #65436). The Phf6 shRNA (5'- GCAAGGGATTTACAT
GGTTTA-3', 97mer: 5'-TGCTGTTGACAGTGAGCGAGCAAGGGATTTACATGGTTTATA
GTGAAGCCACAGATGTATAAACCATGTAAATCCCTTGCGTGCCTACTGCCTCGGA-3')
was cloned into the mir30 backbone [Dickins et al. 2005, Meacham et al. 2015].
Antibodies
Western blotting: anti-HSP90 (610419, BD Biosciences), anti-PHF6 N-terminus
(A301-450A, Bethyl), anti-PHF6 C-terminus (A301-452A, Bethyl), anti-Histone H3
(C-16) (sc-8654, Santa Cruz Biotechnology), and anti-PHF6 (68262, Novus). Secondary
antibodies: anti-rabbit IgG, HRP-linked (7074, Cell Signaling Technology), anti-mouse
IgG, HRP-linked (7076, Cell Signaling Technology) and anti-goat IgG, HRP-linked
(sc-2350, Santa Cruz Biotechnology). Immunoprecipitation: anti-PHF6 (68262, Novus),
anti-TCF12 (11825, Cell Signaling Technology), anti-NF-kB (p65) (sc-372, Santa Cruz
Biotechnology), anti-Histone H3 (A300-823A, Bethyl), and normal rabbit IgG (sc-2027,
Santa Cruz Biotechnology). Flow cytometry: APC anti-mouse B220 (553092 Clone
RA3-6B2, BD Biosciences), V450 anti-mouse IgM (560575 Clone R6-60.2, BD
Biosciences), FITC anti-mouse CD19 (11-0193-81 Clone eBiolD3, eBiosciences), APC
anti-CD4 (553051 Clone RM4-5, BD Biosciences), V450 anti-CD4 (560468 Clone
RM4-5, BD Biosciences), FITC anti-CD3 (11-0032 Clone 17A2, eBiosciences) and APC
anti-CD11b (101212 Clone M1/70, Biolegend). Flow cytometry isotype controls: V450
anti-IgG2a, K (560377 Clone R35-95, BD Biosciences), FITC anti-IgG2b, K (11-4031-82
Clone eB149/10H5, eBiosciences), APC anti-IgG2a, K (553932 Clone R35-95, BD
Biosciences), FITC anti-IgG2a, K (11-4321 Clone eBR2a, eBiosciences) and APC anti-
IgG2b, K (400612 Clone RTK4530, Biolegend). IHC staining: anti-RFP (600-401-379,
Rockland). ChIP experiments: anti-PHF6 (A301-451A, Bethyl), anti-H3K27ac (ab4729,
abcam), anti-H3K27me3 (ab6002, abcam), anti-PHF6 (68262, Novus), and normal
rabbit IgG (sc-2027, Santa Cruz Biotechnology). yH2AX experiments: anti-yH2AX
140
(Upstate 05-636 or Cell Signaling CS9718S), and FITC-anti-mouse (BD 553443, BD
Biosciences) or FITC-anti-rabbit (BD 554020, BD Biosciences).
Cell culture
Murine BCR-ABL1+; p19-'-; mCherry+ B-ALL cells [Williams et al. 2006] were cultured in
RPMI-1640 (Corning) supplemented with 10% FBS, 4mM L-glutamine, 50pM 2-
mercaptoethanol, 100U/mL penicillin and 100pg/mL streptomycin. All retrovirus was
generated using the Phoenix cell system (G. Nolan, Stanford University). Briefly,
calcium phosphate transfection was performed with 10pg of plasmid DNA and 5pg of
helper plasmid (pCMV-Gag-Pol). Transduction efficiencies were assessed 48 hours post
infection. To generate Phf6KO cells, sgPhf6_1-pX458 and sgPhf6_4-pX330 plasmids
were transfected (Lipofectamine 3000, Life Technologies) into B-ALL cells and single
cell sorted into 96-well plates based on GFP expression. The cell proliferation assay
was performed by plating B-ALL cells in triplicate (1,00,000 cells/well in 3mL of media in
a 6-well plate), then counting every 24 hours by trypan blue exclusion and replating at
the initial density. This was performed for 7 days.
Drug response analysis
10,000 cells/well were plated in 96-well plates and media containing Ponatinib
(AP24534, LC Laboratories) or Doxorubicin (D1515, Sigma-Aldrich) were added to
achieve the indicated final concentrations. Cell viability was assessed 48 hours post-
treatment by propidium iodide (81845, Sigma-Aldrich) exclusion and flow cytometry.
Inhibitor dilutions were made in cell medium immediately before use.
Western blotting
Cells were lysed with RIPA buffer (BP-115, Boston BioProducts) supplemented with 1X
protease inhibitor solution (cOmplete EDTA-free, 11873580001, Roche). Protein
concentration of cell lysates was determined by Pierce BCA Protein Assay (23225,
ThermoFisher Scientific). Total protein (50-70pg) was separated on 4-12% Bis-Tris
gradient SDS-PAGE gels (Life Technologies) and then transferred to PVDF membranes
(IPVH00010, EMD Millipore) for blotting.
Co-immunoprecipitation (Co-IP)
Antibodies were pre-cleaned before attaching to magnetic beads using the Pierce
Antibody Clean-Up Kit (44600, ThermoFisher Scientific). A total of 10pg of antibody
were covalently attached to magnetic beads (88828, ThermoFisher Scientific) and
incubated with 1mg of protein for 2 hours at room temperature while rotating (or
overnight at 4'C). Immunoprecipitation (IP) was performed according to manufacturer's
directions (Pierce Direct Magnetic IP/Co-IP Kit, Thermo Fisher Scientific). For DNA-
dependent interactions, ethidium bromide was added (50ug/mL) and lysates were
incubated on ice for 30 mins. Samples were centrifuged and the supernatant was used
for IPs. 70ug (or 7%) of protein was used for input on the western blot.
Quantitative PCR (aPCR)
RNA was prepared using RNeasy Mini Kit (Qiagen). Synthesis of cDNA was performed
using M-MLV Reverse Transcriptase (28025, Life Technologies) with oligo(dT)20 primer.
141
qPCR was done in Applied Biosystems StepOnePlus machine with TaqMan Fast
Universal PCR Master Mix (4352042, Life Technologies). Data were analyzed using the
comparative ACT method, and were normalized to the levels of Gapdh or Actin. TaqMan
gene expression assays (Life Technologies) used: Phf6 (Mm00804415_ml), Blk
(Mm00432077_ml), Cd74 (Mm00658576_ml), lL4ra (Mm0l275139_ml), Ly86
(Mm00440240_ml), Lyn (Mm01217488_ml), Gapdh (Mm99999915_g1), Cd22
(Mm00515432_ml), Rhoq (Mm00467435_ml), Cd79b (Mm00434143_ml). The
following qPCRs were performed with Fast SYBER Green qPCR Master MIX (4385612,
Applied Biosciences). Primer sequences (Forward/Reverse):
BCR-ABLI (5'-CTGGCCCAACGATGGCGA)/(5'-CACTCAGACCCTGAGGCTCAA),
Tcfl2 (5'-ATGTGCTACGAAACCATGCAG)/(5'-GCCATTGAGACTGACTGAATCTT),
NfKb (5'-AGGCTTCTGGGCCTTATGTG)/(5'-TGCTTCTCTCGCCAGGAATAC),
NotchI (5'-CCAACTGAGGACAGACGGAC)/(5'-GGGATCAGAGGCCACATAGC),
mGAPDH (5'-AGAACATCATCCCTGCATCC)/(5'-CACATTGGGG GTAGGAACAC).
Genomic DNA isolation and sequencing
Genomic DNA from B-ALL cells was isolated using the Blood & Cell Culture DNA Mini
Kit (Qiagen). For sequencing of the Phf6 locus, PCR products were amplified using
Herculasell Fusion DNA polymerase (Agilent), gel purified, cloned into pCR4-TOPO TA
vector (Life Technologies) and subsequently sequenced. Primers for DNA sequencing:
F: 5'-AGCTGGGTATTAGCTCCAGTTG-3'/R: 5'-TGCACTCCACTGATCCTCTC-3'.
Animal experiments
All animal studies were approved by the MIT Institutional Animal Care and Use
Committee. For syngeneic transplants, 103 or 106 cells murine BCR-ABL1+p19-- B-ALL
cells were injected via tail-vein into syngeneic C57BL/6J (000664, The Jackson
Laboratory) recipient mice (age of 6-8 weeks). Treatment with Ponatinib (30mg/kg,
administered via oral gavage, treated once daily for 4-consecutive days), Doxorubicin
(10mg/kg, administered via IP, treated once, NovaPlus), or vehicle (25mM citric acid
buffer) was initiated upon significant disease presentation. Mice were observed daily
and sacrificed when moribund (dehydration, ruffled fur, respiratory distress, poor
motility). Survival curves were generated using GraphPad Prism version 6.0 software,
and the Mantel-Cox test was applied to pairwise comparisons of survival data.
Organ grocessing and cell preDaration for flow cytometry
Upon morbidity, mice were sacrificed and lymphoid organs were collected. Bone marrow
cells were obtained by flushing the femora and tibiae with B-ALL media. Lymph nodes,
spleens, and thymi were isolated, weighed for mass measurements, ground by frosted
glass slides, and filtered to obtain single cell suspension. Red blood cell lysis was
performed with ACK Lysing Buffer (Al 0492-01, Life Technologies).
Immunostaining
Leukemia cell suspensions were collected and treated with indicated antibodies. Briefly,
1-2x106 cells in 1OOpL of 10% FBS in PBS were incubated with antibody (1:100) for 1
hour at room temperature in the dark. Stained samples were analyzed by flow
cytometry. For intracellular immunophenotyping, leukemia cell suspensions (cultured
142
and isolated from mice) were collected, re-suspended in PBS and fixed with 4%
formaldehyde (1008A, Tousimis) solution for 10 minutes at 370C. Cells were
permeabilized with ice-cold methanol for 30 minutes on ice.
Cell cycle analysis
FITC BrdU Flow Kit (559619, BD Biosciences) was used for cell cycle analysis. Briefly,
1x106 cells B-ALL cells plated in triplicate were labeled with 10pM BrdU for 30 minutes.
Cells were subsequently fixed and stained with anti-BrdU and 7-AAD and analyzed by
flow cytometry (LSRFortessa, BD Biosciences) [Soto-Feliciano 2016].
Flow cytometry
A FACSAria (BD Biosciences) was used for fluorescence-activated cell sorting and
LSRFortessa (BD Biosciences) was used for flow cytometry analysis. Data were
analyzed with FlowJo software (version 10, TreeStar).
Histology and immunohistochemistry
For histological and immunohistochemistry analyses, mice were euthanized by carbon
dioxide asphyxiation. Lymphoid tissues were harvested, fixed overnight with 10%
neutral buffered formalin (VWR), transferred to 70% ethanol solution and subsequently
embedded in parafilm. Sections were cut at a thickness of 10pm and stained with
haematoxylin and eosin (H&E) for pathological assessment. Immunohistochemistry
(IHC) was performed on a Thermo Autostainer 360 machine. Slides were antigen
retrieved using Thermo citrate buffer, pH6.0 in the pre-treatment module. Sections were
treated with Biocare rodent block, primary antibody, and anti-rabbit HRP-polymer
(Vector Labs). The slides were developed with Thermo Ultra DAB and counterstained
with haematoxylin in a Thermo Gemini stainer and coverslips added using the Thermo
Consul coverslipper [Soto-Feliciano 2016].
Determination of yH2AX levels
Cells were irradiated with a dose of 2Gy or 1OGy and analyzed 1 hour post-irradiation.
Mice were treated with 10mg/kg of doxorubicin (IP) upon significant disease
presentation and tissues were harvested 12-hour post administration. Levels of yH2AX
were determined by intracellular FACS as detailed by Huang and Darzynkiewicz [2006].
Briefly, 5x 106 cells were centrifuged, resuspended in 500 uL of PBS, and 4.5 mLs of
ice-cold 1% methanol-free formaldehyde solution was added. Cells were kept on ice for
15 minutes, centrifuged, and washed in 4.5 mLs PBS. Cells were centrifuged,
resuspended in 500uL PBS, and 4.5 mLs of ice-cold 70% ethanol was added. Samples
were kept at -20 for 2 hours (or overnight), centrifuged, and washed in 2mLs of BSA-T-
PBS (1% BSA and 0.2% Triton-X). Pellets were resuspended in 2mLs BSA-T-PBS and
incubated for 5 mins. Cells were centrifuged, resuspended in 10uL of BSA-T-PBS, and
0.75uL of anti-yH2AX antibody was added (Upstate 05-636 or Cell Signaling CS9718S)
and incubated for 2 hours at RT. Cells were washed with 2mLs of BSA-T-PBS,
centrifuged, and resuspended in 1OOuL of BSA-T-PBS. 0.75uL of secondary antibody
was added (BD 553443 or BD 554020) and incubated for 1 hours. Cells were washed in
143
3mLs of BSA-T-PBS, centrifuged, and resuspended in a final volume of 200uL of PBS
for FACS analysis.
RNA-Sequencing (RNA-Sea) library preparation
cDNA for RNA-Seq libraries was prepared using TruSeq mRNA library Prep Kit
(Illumina). Illumina libraries were produced using the SPRlworks (Beckman-Coulter
Genomics) with a 200-400 bp size selection and enriched with 14 cycles of PCR.
Library quality was determined by qPCR and on the Fragment Analyzer (Advanced
Analytical) and loaded on two lanes of 40-nucleotides (nt) single end sequencing on a
HiSeq2000 system (Illumina) [Soto-Feliciano 2016].
RNA-Sequencing data analysis
Illumina HiSeq2000 40-nt single-ended reads were mapped to the UCSC mm9 mouse
genome build (http://genome.ucsc.edu/) using RSEM [Li & Dewey 2011]. Raw estimated
expression counts were upper-quartile normalized to a count of 1000 [Bullard et al.
2010]. Given the complexity of the dataset in terms of a mixture of different conditions, a
high-resolution signature discovery approach was employed to characterize global gene
expression profiles. Independent Component Analysis (ICA), an unsupervised blind
source separation technique, was used on this discrete count-based expression dataset
to elucidate statistically independent and biologically relevant signatures. ICA is a signal
processing and multivariate data analysis technique in the category of unsupervised
matrix factorization methods. Conceptually, ICA decomposes the overall expression
dataset into independent signals (gene expression patterns) that represent distinct
signatures. High-ranking positively and negatively correlated genes in each signature
represent gene sets that drive the corresponding expression pattern (in either direction).
Signatures were visualized using the sample-to-signature correspondence schematic
afforded by Hinton plots where colors represent directionality of gene expression (red,
up-regulated; green, down-regulated) and the size of each rectangle quantifies the
strength of a signature (column) in a given sample (row). Each signature is two-sided,
allowing for identification of up-regulated and down-regulated genes for each signature
within each sample. Utilizing input data consisting of a genes-samples matrix, ICA uses
higher order moments to characterize the dataset as a linear combination of statistically
independent latent variables. These latent variables represent independent components
based on maximizing non-Gaussianity, and can be interpreted as independent source
signals that have been mixed together to form the dataset under consideration. Each
component includes a weight assignment to each gene that quantifies its contribution to
that component. Additionally, ICA derives a mixing matrix that describes the contribution
of each sample towards the signal embodied in each component. This mixing matrix
can be used to select signatures among components with distinct gene expression
profiles across the set of samples. The R implementation of the core JADE algorithm
(Joint Approximate Diagonalization of Eigenmatrices) [Nordhausen et al. 2015; Rutledge
& Bouveresse 2013; Biton et al. 2016] was used along with custom R utilities. The P
value for the KO-specific (IC2) signature is: P= 0.0119 (Mann-Whitney U test directional
P value). Targeted pairwise differential expression analyses were conducted using
EBSeq v1.4.0 [Leng et al. 2013]. Differentially expressed (DE) genes from the pairwise
144
comparison between Phf6KO and Phf6wT B-ALL cells were determined using a
significance level of FDR < 0.05 and fold change (FC) greater than 1.5. All RNA-Seq
analyses were conducted in the R Statistical Programming language (http://www.r-
project.org/). Gene set enrichment analysis (GSEA) was carried out using the pre-
ranked mode with default settings [Subramanian et al. 2005]. Heatmaps were generated
using the Heatplus package in R. Integrated RNA-Seq signature analysis of B-ALL cells
and a data set from CD4+-single positive (SP) cells [Miyazaki et al. 2015] (GEO
accession code GSE64779, samples GSM1580493-GSM1580497) [Edgar et al. 2002;
Barrett et al. 2012] was performed using ICA. A two-dimensional expression principal
component (PCA) plot for each group of cells was generated by plotting the first two
principal components of expression (explaining 75% and 11% of expression variance
respectively). Gene ontology analysis was done using DAVID and top GO- and Panther-
terms were plotted against their corresponding P values (log-scale) [Soto-Feliciano
2016].
Chromatin immunoprecipitation-Seauencing (ChIP-Sea)
ChIP was performed as described before [Lee et al. 2006] with a few adaptations. A
total of 150x10 6 cells, grown to a density of 1x106 cells/mL, were cross-linked for 10
minutes at room temperature by the addition of one-tenth of the volume of 11%
formaldehyde solution (11% formaldehyde, 50 mM HEPES pH 7.3, 100 mM NaCl, 1 mM
EDTA pH 8.0, 0.5 mM EGTA pH 8.0) to the growth media followed by 5 minutes
quenching with 125 mM glycine. Cells were washed twice with PBS, then the
supernatant was aspirated and the cell pellet was flash frozen in liquid nitrogen. Frozen
cross-linked cells were stored at -80*C. ProteinG Dynabeads (100pL, Life
Technologies) were blocked with 0.5% BSA (w/v) in PBS. Magnetic beads were bound
with 10pg of anti-H3K27ac (ab4729, abcam), anti-H3K27me3 (ab6002, abcam) and
anti-PHF6 (A301-451A, Bethyl). Nuclei were isolated as previously described [Lee et al.
2006], and sonicated in lysis buffer (20 mM Tris-HCI pH 8.0, 150 mM NaCl, 2 mM EDTA
pH 8.0, 0.1 % SDS and 1 % Triton X-1 00) on a Misonix 3000 sonicator for 10 cycles at 30
seconds each on ice (18-21W) with 60 seconds on ice between cycles. Sonicated
lysates were cleared once by centrifugation and incubated overnight at 40C with
magnetic beads bound with antibody to enrich for DNA fragments bound by the
indicated factor. Beads were washed with wash buffer A (50 mM HEPES-KOH pH 7.9,
140 mM NaCI, 1 mM EDTA pH 8.0, 0.1% sodium deoxycholate, 1% Triton X-100 and
0.1% SDS), buffer B (50 mM HEPES-KOH pH 7.9, 500 mM NaCl, 1 mM EDTA pH 8.0,
0.1% sodium deoxycholate, 1% Triton X-100 and 0.1% SDS), buffer C (20 mM Tris-HCI
pH 8.0, 250 mM LiCI, 1 mM EDTA pH 8.0, 0.5% sodium deoxycholate, 0.5% IGEPAL
CA-630 and 0.1% SDS) and buffer D (TE with 50 mM NaCI), sequentially. DNA was
eluted in elution buffer (50 mM Tris-HCI pH 8.0, 10 mM EDTA and 1% SDS). Cross-links
were reversed overnight. RNA and protein were digested using RNaseA and Proteinase
K, respectively and DNA was purified with phenol-chloroform extraction and ethanol
precipitation. Purified ChIP DNA was used to prepare Illumina multiplexed sequencing
libraries. Libraries were prepared following the TruSeq DNA Sample Prep v2 Kit
(Illumina). Amplified libraries were size-selected using a 2% gel cassette in the Pippin
Prep system (Sage Science) set to capture fragments between 200-400bp. Libraries
were quantified by qPCR using the Illumina Library Quantification Kit (KAPA
145
Biosystems) according to manufacturer guidelines. Libraries were sequenced on the
Illumina HiSeq2500 for 40-nt in single read mode [Soto-Feliciano 2016].
ChIP-Sequencing data analysis
To compare RNA-Seq with ChIP-Seq datasets, normalized RNA-Seq tag counts for
each gene were used to plot a heatmap showing expression levels of these genes,
ordered by fold change values. Among these genes, 1,127 symbols were mappable to
ENSEMBL mouse gene IDs (unmappable gene symbols mostly correspond to those
that are not well characterized) and were used for mapping to genomic coordinates.
Sequencing reads were mapped to the UCSC mm9 mouse genome using Bowtie2
[Langmead & Salzberg 2012; Langmead et al. 2009] with at most 1 mismatch allowed in
seed alignments. Unmapped reads were removed and alignments were output in BAM
files for peak calling with the MACS software [Zhang et al. 2008]. PHF6 and histone
modification peaks were called using a P value threshold of 1x1O- 6. The fraction of
reads falling within the called peaks (FRiP) for PHF6 ChIP-Seq data is 25.9%
(8,333,082 out of 32,143,465 uniquely mappable reads), showing high ChIP quality. To
examine whether PHF6 binding strength correlates with gene expression levels, RNA-
Seq data from Phf6wT cells was sorted and grouped by expression into three bins
corresponding to high, medium and low expression by separating into 3 equally sized
bins. Average log2 fold change values over input signals were plotted as a metagene
track for each gene set. Plots for H3K27ac signal enrichment were also generated to
confirm promoter activity for each gene set. To determine PHF6 binding at promoter/
enhancer regions of the DE genes, genomic sequences were obtained for the 5kb
region surrounding the transcription start site (TSS) for these genes from the UCSC
mm9 genome assembly and binned sequences into 50bp bins. ChIP-Seq RPM (reads
per million) values for PHF6, H3K27ac, H3K27me3 and H3K4me3 [Wang et al. 2015]
(publically available ChIP-Seq dataset was obtained as SRA-lite files, GEO accession
code GSE66234, datasets GSM1617788-GSM1617789) [Edgar et al. 2002; Barrett et
al. 2012] at these promoter bins were divided by the corresponding ChIP input (control)
reads with a pseudocount of 1 and then log2-transformed to show fold enrichment. To
show correlation between PHF6 and histone modification signals, Pearson's correlation
coefficients between corresponding pairs of columns in the heatmaps were calculated.
The MEME suite [Bailey & Elkan, 1994] was used for de novo motif discovery of
sequences surrounding PHF6 binding peak summits. A 100-nt window up- and down-
stream of the peak summits was set to obtain sequences from the UCSC mm9 genome
assembly. Among all 77,749 PHF6 summits called by the MACS pipeline [Zhang et al.
2008], 6,037 fell within the proximal regulatory regions ( 5kb surrounding the TSS) of
DE genes, covering a total of 1,104 (out of 1,123) unique promoters. For each promoter
region the PHF6 summit was selected with the highest average peak signal for motif
analysis, resulting in 1,044 unique summits and their surrounding sequences as input
for the MEME motif discovery tool [Bailey & Elkan, 1994]. To see if PHF6 displays any
bias in sequence specificity at the DE gene promoters, the top 1,044 among all PHF6
summits were selected and ran motif discovery in parallel. A motif length of 8 was used
and selected the top 10 significant motifs returned by MEME [Bailey & Elkan, 1994].
The motifs thus discovered were then tested for similarity with known motifs (as in the
146
JASPAR CORE 2014 vertebrates [Mathelier et al. 2013], UniProbe Mouse [Hume et al.
2015] and other human and mouse motifs [Jolma et al. 2013] using the Tomtom tool
[Gupta et al. 2007], with a minimum motif overlap of 1 and Pearson distance as the
distance metric for motif similarity. To identify other transcription factors (TFs) that co-
bind with PHF6, mm9 sequences were obtained for 1kb regions around the TSS of DE
genes and tested for known motif enrichment using the Bioconductor PWMEnrich
package [Stojnic & Diez 2016]. A default Mus musculus genomic background and the
MotifDb database of motifs [Shannon 2017] included in the package were used, and a P
value threshold of 0.01 was set for reporting significant TFs whose motifs were enriched
in the selected regions. Input sequences were randomly shuffled 100 times and
performed motif enrichment for each shuffled sequence set, and considered motifs that
came up as significant (P < 0.001) in more than 5% of the shuffled sets as false positive
enrichments. A total of 28 motifs corresponding to 24 unique TFs passed the filtering
step and were reported as potential co-binding partners of PHF6 at the promoters of
differentially expressed genes (Supp. Table S1). Gene functions were determined by
using the PANTHER Classification System [Thomas et al. 2003; Mi et al. 2004;Soto-
Feliciano 2016].
Chromatin immunoprecipitation-aPCR (ChIP-aPCR)
Chromatin immunoprecipitation was performed as above. qPCR reactions were
performed as described above. Quantities were normalized to input, calculated using
the AACT method, and expressed as fold enrichment relative to a negative binding
control (Pchd10). Primer sequences (forward/reverse):
Spib (5'-TGCCTCACTCACCTCCTTCT) / (5'-ACTTGGGCCCTTGTTCTCTT),
Lmo2 (5'-TCTCAGAAGGGCTGTGTGTG)/(5'-TTTGTGTGGACGTGGAAAAA)
Hes1 (5'-CCCACCTCTCTCTTCTGACG)/(5'-AGGCGCAATCCAATATGAAC)
Hck (5'-CAGGTGTGGACTTTCCCACT)/(5'-GCGATCGGTGAGAACTTAGC)
Cd74 (5'-GATGGCTACTCCCTTGCTGA) / (5'-GATGGCTACTCCCTTGCTGA),
Lyn (5'-GCAGAGAGAAGTGGCCTGAG)/(5'-CTTAGAGAGGGGCTGGGTTC)
Ly86 (5'-AGTGGGGGCTTGGAAGTAGT)/(5'-AAAGCACGGAGAGGCTGATA)
Phf6 (5'-TTCCAAGACATTCCGTCCTC)/(5'-CTCCTTTCGCCTGATTTCAC)
B/k (5'-ACAGGGACAATGCCTACCTG)/(5'-TCCCTGGAGCTACTGCACTT)
Btg2 (5'-GACACTGACAGAGCCGTTCA) / (5'ACACTCCTCCCACCAAACAG),
Gata3 (5'-AGCCCAGGACTGACTAAGCA) / (5'-TGTTTGGGGGTTTGTTTGTT),
Lefi (5'-GAGGGCCAGAGACACTGAAC)/(5'-CCTGTGTGTCTTCCCTTGGT)
Lat (5'-CAATAGAGATGCGGTGCAGA)/(5'-GGGCCCTTCTTCTCCTGTAG)
Tcf7 (5'-GAAGTGTGGGAGAGGCTCAG)/(5'-ATGTTTGTGGGGAGCACATT)
Cd4 (5'-TCTCCTGCTTCAGGGTCAGT) / (5'-AGGGACACCCTGTTTCTGTG),
Jag2 (5'-AGCCACAGCACACTGAACAC)/(5'-CACAGGTAAGCTGGGGTGAT)
Runx3 (5'-CTCCAGCCCGAGACTACAAG)/(5'-GGGATGCACAGCTAGAGAGG)
HesI (5'-CCCACCTCTCTCTTCTGACG)/(5'-AGGCGCAATCCAATATGAAC)
Lck (5'-AGAGTCAACCCTCAGCCTCA)/(5'-GGAGCCGAAGTAGACACAGC)
Zbtb7b (5'-CCCAGAGGCGTCTATCAGAG)/(5'-TGGGGGTCAGAACTTACTGG)
LmoI (5'-GACCTGCTGCCTGTCTTAGG)/(5'-AGTGGTTTGAGGGTGACCTG)
Cd5 (5'-CTAGCTTGGCCAGTCCTTTG)/(5'-CCTCTTGAGCCTCAGACACC)
Pchd10 (5'-CCCTGGACTTGCTGACTAGC) / (5'-GTGCAAACCAGAACATGGTG).
147
Chromatin accessibility by ATAC-Sequencing (ATAC-Seq)
ATAC-Seq libraries from Phf6wT and Phf6KO B-ALL cells were constructed as previously
reported [Buenrostro et al. 2013]. For the ATAC-Seq data analysis, mapped ATAC reads
were summarized into 500bp bins. The DESeq2 package in R [Love et al. 2014] was
then applied to the binned reads to analyze for differential chromatin accessibility
between Phf6 wild-type and knockout cells. To call differentially accessible chromatin
regions, we used a fold-change cutoff of >2 and a FDR cutoff of <0.001. GREAT
enrichment tool [McLean et al. 2010] was used to analyze for enriched gene sets and
other genomic features in the regions that were more or less accessible in Phf6KO cells
separately. To test enrichment of known and de novo sequence motifs within these
differentially accessible chromatin regions, we used the HOMER motif analysis toolkit
[Heinz et al. 2010] on these regions using all mappable ATAC regions as background
and the UCSC mm9 genome, with all other parameters set to default values.
The NucleoATAC software was used with default parameters to call nucleosome signals
from raw ATAC-Seq signals in Phf6WT cells [Schep et al. 2015]. Background-corrected
nucleosome signals 1kb from transcription start sites (TSSs) annotated in the mm9
genome were then stacked into a heatmap with rows corresponding to individual TSSs
ordered by gene expression in Phf6wT cells. To visualize PHF6 occupancy at these
sites, a heatmap of normalized PHF6 signals in the same row order was also
generated. ATAC-Seq raw data and processed TSS nucleosome signals are available in
the GEO database with accession number GSE98716 under super-series GSE77457.
Statistical analysis
Results are expressed as means standard deviation (SD). Statistical significance was
determined with Prism software (version 6.0, GraphPad) by comparison of mean values
by the two-tailed, unpaired Student's t-test on two experimental conditions, with P <
0.05 considered statistically significant. Statistical significance levels are designated as
follows: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, n.s. = not significant.
Survival curves were generated using Prism software (version 6.0, GraphPad), and the
Mantel-Cox test was applied to pairwise comparisons of survival data.
148
REFERENCES
Abbas AR, Wolslegel K, Seshasayee D, Modrusan Z, Clark HF. 2009. Deconvolution of Blood Microarray Data
Identifies Cellular Activation Patterns in Systemic Lupus Erythematosus ed. P. Tan. PLoS ONE 4: e6098-16.
Arenzana TL, Smith-Raska MR, Reizis B. 2009. Transcription factor Zfx controls BCR-induced proliferation and
survival of B lymphocytes. Blood 113: 5857-5867.
Bailey TL, Elkan C. 1994. Fitting a mixture model by expectation maximization to discover motifs in biopolymers. Proc
Int Conf Intell Syst Mol Biol 2: 28-36.
Barrett T, Wilhite SE, Ledoux P, Evangelista C, Kim IF, Tomashevsky M, Marshall KA, Phillippy KH, Sherman PM,
Holko M, et al. 2012. NCBI GEO: archive for functional genomics data sets--update. Nucleic Acids Research 41:
D991-D995.
Bell JJ, Bhandoola A. 2008. The earliest thymic progenitors for T cells possess myeloid lineage potential. Nature
452: 764-767.
Belver L, Ferrando A. 2016. The genetics and mechanisms of T cell acute lymphoblastic leukaemia. Nature Reviews
Cancer 16: 494-507.
Biton A, Zinovyev A, Barillot E, Radvanyi F. 2016. MinelCA: Independent component analysis of transcriptomic data.
R package version 1.16.0.
Borst J, van de Griend RJ, van Oostveen JW, Ang SL, Melief CJ, Seidman JG, Bolhuis RL. 1987. A T-cell receptor
gamma/CD3 complex found on cloned functional lymphocytes. Nature 325: 683-688.
Buenrostro JD, Giresi PG, Zaba LC, Chang HY, Greenleaf WJ. 2013. Transposition of native chromatin for fast and
sensitive epigenomic profiling of open chromatin, DNA-binding proteins and nucleosome position. Nature
Methods 10: 1213-1218.
Bullard JH, Purdom E, Hansen KD, Dudoit S. 2010. Evaluation of statistical methods for normalization and differential
expression in mRNA-Seq experiments. BMC Bioinformatics 11: 94-13.
Carlson CM, Endrizzi BT, Wu J, Ding X, Weinreich MA, Walsh ER, Wani MA, Lingrel JB, Hogquist KA, Jameson SC.
2006. Kruppel-like factor 2 regulates thymocyte and T-cell migration. Nature 442: 299-302.
Carter MT, Picketts DJ, Hunter AG, Graham GE. 2009. Further clinical delineation of the B6rjeson-Forssman-
Lehmann syndrome in patients with PHF6mutations. Am J Med Genet 149A: 246-250.
Chao MM, Todd MA, Kontny U, Neas K, Sullivan MJ, Hunter AG, Picketts DJ, Kratz CP. 2010. T-cell acute
lymphoblastic leukemia in association with Bbrjeson-Forssman-Lehmann syndrome due to a mutation in PHF6.
Pediatr Blood Cancer 55: 722-724.
Cobaleda C, Jochum W, Busslinger M. 2007. Conversion of mature B cells into T cells by dedifferentiation to
uncommitted progenitors. Nature 449: 473-477.
Cooley M, Fang P, Fang F, Nephew KP, Chien J. 2015. Molecular determinants of chemotherapy resistance in
ovarian cancer. Pharmacogenomics 16, 1763-1767.
Corces MR, Buenrostro JD, Wu B, Greenside PG, Chan SM, Koenig JL, Snyder MP, Pritchard JK, Kundaje A,
Greenleaf WJ, et al. 2016. Lineage-specific and single-cell chromatin accessibility charts human hematopoiesis
and leukemia evolution. Nat Genet 48: 1193-1203.
Dai X, Chen Y, Di L, Podd A, Li G, Bunting KD, Hennighausen L, Wen R, Wang D. 2007. Stat5 Is Essential for Early B
Cell Development but Not for B Cell Maturation and Function. Joumal of Immunology 179: 1068-1079.
De Keersmaecker K, Atak ZK, Li N, Vicente C, Patchett S, Girardi T, Gianfelici V, Geerdens E, Clappier E, Porcu M,
et al. 2012. Exome sequencing identifies mutation in CNOT3 and ribosomal genes RPL5 and RPL10 in T-cell
acute lymphoblastic leukemia. Nat Genet 45: 186-190.
Dickins RA, Hemann MT, Zilfou JT, Simpson DR, Ibarra I, Hannon GJ, Lowe SW. 2005. Probing tumor phenotypes
using stable and regulated synthetic microRNA precursors. Nature Genetics 1-8.
Doench JG, Hartenian E, Graham DB, Tothova Z, Hegde M, Smith 1, Sullender M, Ebert BL, Xavier RJ, Root DE.
2014. Rational design of highly active sgRNAs for CRISPR-Cas9-mediated gene inactivation. Nature
Biotechnology 32: 1-8.
Edgar R, Domrachev M, Lash AE. 2002. Gene Expression Omnibus: NCBI gene expression and hybridization array
data repository. Nucleic Acids Research 30: 207-210.
Eyquem S, Chemin K, Fasseu M, Bories J-C. 2004. The Ets-1 transcription factor is required for complete pre-T cell
receptor function and allelic exclusion at the T cell receptor beta locus. Proc Nat Acad Sci USA 101: 15712-
15717.
Gardner R, Wu D, Cherian S, Fang M, Hanafi L-A, Finney 0, Smithers H, Jensen MC, Riddell SR, Maloney DG, et al.
2016. Acquisition of a CD19-negative myeloid phenotype allows immune escape of MLL-rearranged B-ALL from
CD19 CAR-T-cell therapy. Blood 127: 2406-2410.
Gilbert LA, Hemann MT. 2010. DNA Damage-Mediated Induction of a Chemoresistant Niche. Cell 143: 355-366.
Gupta S, Stamatoyannopoulos JA, Bailey TL, Noble WS. 2007. Quantifying similarity between motifs. Genome
Biology 8: R24.
Hahtola S, Tuomela S, Elo L, Hakkinen T, Karenko L, Nedoszytko B, Heikkila H, Saarialho-Kere U, Roszkiewicz J,
Aittokallio T, et al. 2006. Th1 response and cytotoxicity genes are down-regulated in cutaneous T-cell lymphoma.
Clinical Cancer Research 12: 4812-4821.
Heinz S, Benner C, Spann N, Bertolino E, Lin YC, Laslo P, Cheng JX, Murre C, Singh H, Glass CK. 2010. Simple
Combinations of Lineage-Determining Transcription Factors Prime cis-Regulatory Elements Required for
Macrophage and B Cell Identities. Molecular Cell 38: 576-589.
149
Hoffmann R, Seidl T, Neeb M, Rolink A, Melchers F. 2002. Changes in gene expression profiles in developing B cells
of murine bone marrow. Genome Research 12: 98-111.
Horcher M, Souabni A, Busslinger M. 2001. Pax5/BSAP maintains the identity of B cells in late B lymphopoiesis.
Immunity 14: 779-790.
Huang X, Darzynkiewicz Z. 2006. Cytometric assessment of histone H2AX phosphorylation: a reporter of DNA
damage. Methods Mol Biol 314: 73-80.
Hume MA, Barrera LA, Gisselbrecht SS, Bulyk ML. 2015. UniPROBE, update 2015: new tools and content for the
online database of protein-binding microarray data on protein-DNA interactions. Nucleic Acids Research 43:
D117-D122.
Hutcheson J, Scatizzi JC, Siddiqui AM, Haines GK 111, Wu T, Li Q-Z, Davis LS, Mohan C, Perlman H. 2008. Combined
Deficiency of Proapoptotic Regulators Bim and Fas Results in the Early Onset of Systemic Autoimmunity.
Immunity 28: 206-217.
Jacoby E, Nguyen SM, Fountaine TJ, Welp K, Gryder B, Qin H, Yang Y, Chien CD, Seif AE, Lei H, et al. 2016. CD19
CAR immune pressure induces B-precursor acute lymphoblastic leukaemia lineage switch exposing inherent
leukaemic plasticity. Nature Communications 7: 1-10.
Janeway CA, Carding S, Jones B, Murray J, Portoles P, Rasmussen R, Rojo J, Saizawa K, West J, Bottomly K. 1988.
CD4+ T cells: specificity and function. Immunological Reviews 101: 39-80.
Jeffrey KL, Brummer T, Rolph MS, Liu SM, Callejas NA, Grumont RJ, Gillieron C, Mackay F, Grey S, Camps M, et al.
2006. Positive regulation of immune cell function and inflammatory responses by phosphatase PAC-1. Nature
Immunology 7: 274-283.
Jolma A, Yan J, Whitington T, Toivonen J, Nitta KR, Rastas P, Morgunova E, Enge M, Taipale M, Wei G, et al. 2013.
DNA-Binding Specificities of Human Transcription Factors. Cell 152: 327-339.
Kiaii S, Clear AJ, Ramsay AG, Davies D, Sangaralingam A, Lee A, Calaminici M, Neuberg DS, Gribben JG. 2013.
Follicular Lymphoma Cells Induce Changes in T-Cell Gene Expression and Function: Potential Impact on
Survival and Risk of Transformation. Joumal of Clinical Oncology 31: 2654-2661.
Krivtsov AV, Feng Z, Lemieux ME, Faber J, Vempati S, Sinha AU, Xia X, Jesneck J, Bracken AP, Silverman LB, et al.
2008. H3K79 Methylation Profiles Define Murine and Human MLL-AF4 Leukemias. Cancer Cell 14: 355-368.
Ku S-Y, Rosario S, Wang Y, Mu P, Seshadri M, Goodrich ZW, Goodrich MM, Labb6 DP, Gomez EC, Wang J, et al.
2017. Rbl and Trp53 cooperate to suppress prostate cancer lineage plasticity, metastasis, and antiandrogen
resistance. Science 355: 78-83.
Laiosa CV, Stadtfeld M, Xie H, de Andres-Aguayo L, Graf T. 2006. Reprogramming of committed T cell progenitors to
macrophages and dendritic cells by C/EBP alpha and PU.1 transcription factors. Immunity 25: 731-744.
Langmead B, Salzberg SL. 2012. Fast gapped-read alignment with Bowtie 2. Nature Methods 9: 357-359.
Langmead B, Trapnell C, Pop M, Salzberg SL. 2009. Ultrafast and memory-efficient alignment of short DNA
sequences to the human genome. Genome Biology 10: R25.
Lara-Astiaso D, Weiner A, Lorenzo-Vivas E, Zaretsky 1, Jaitin DA, David E, Keren-Shaul H, Mildner A, Winter D, Jung
S, et al. 2014. Chromatin state dynamics during blood formation. Science 345: 943-949.
Lee TI, Johnstone SE, Young RA. 2006. Chromatin immunoprecipitation and microarray-based analysis of protein
location. Nature Protocols 1: 729-748.
Leng N, Dawson JA, Thomson JA, Ruotti V, Rissman Al, Smits BMG, Haag JD, Gould MN, Stewart RM, Kendziorski
C. 2013. EBSeq: an empirical Bayes hierarchical model for inference in RNA-seq experiments. Bioinformatics
29:1035-1043.
Li B, Dewey CN. 2011. RSEM: accurate transcript quantification from RNA-Seq data with or without a reference
genome. BMC Bioinformatics 12: 323.
Li M, Xiao L, Xu J, Zhang R, Guo J, Olson J, Wu Y, Li J, Song C, Ge Z. 2016. Co-existence of PHF6 and NOTCH1
mutations in adult T-cell acute lymphoblastic leukemia. Oncology Letters 12: 16-22.
Litzow MR, Ferrando AA. 2015. How I treat T-cell acute lymphoblastic leukemia in adults. Blood 126: 833-841.
Liu Z, Li F, Ruan K, Zhang J, Mei Y, Wu J, Shi Y. 2014. Structural and functional insights into the human B6rjeson-
Forssman-Lehmann syndrome-associated protein PHF6. The Joumal of Biological Chemistry 289: 10069-
10083.
Love MI, Huber W, Anders S. 2014. Moderated estimation of fold change and dispersion for RNA-seq data with
DESeq2. Genome Biology 15: 31-62.
Lower KM, Turner G, Kerr BA, Mathews KD, Shaw MA, Gedeon AK, Schelley S, Hoyme HE, White SM, Delatycki
MB, et al. 2002. Mutations in PHF6 are associated with B6rjeson-Forssman-Lehmann syndrome. Nature
Genetics 32: 661-665.
Luo Q, Ma X, Wahl SM, Bieker JJ, Crossley M, Montaner LJ. 2004. Activation and Repression of Interleukin-12 p40
Transcription by Erythroid Kruppel-like Factor in Macrophages. Joumal of Biological Chemistry 279: 18451-
18456.
Mathelier A, Zhao X, Zhang AW, Parcy F, Worsley-Hunt R, Arenillas DJ, Buchman S, Chen C-Y, Chou A, lenasescu
H, et al. 2013. JASPAR 2014: an extensively expanded and updated open-access database of transcription
factor binding profiles. Nucleic Acids Research 42: D142-D147.
McLean CY, Bristor D, Hiller M, Clarke SL, Schaar BT, Lowe CB, Wenger AM, Bejerano G. 2010. GREAT improves
functional interpretation of cis-regulatory regions. Nature Biotechnology 28: 495-501.
Meier-Stiegen F, Schwanbeck R, Bernoth K, Martini S, Hieronymus T, Ruau D, Zenke M, Just U. 2010. Activated
Notchi Target Genes during Embryonic Cell Differentiation Depend on the Cellular Context and Include Lineage
Determinants and Inhibitors ed. J.-A.L. Stanton. PLoS ONE 5: e11481-18.
Meacham CE, Lawton LN, Soto-Feliciano YM, Pritchard JR, Joughin BA, Ehrenberger T, Fenouille N, Zuber J,
Williams RT, Young RA, et al. 2015. A genome-scale in vivo loss-of-function screen identifies Phf6as a lineage-
specific regulator of leukemia cell growth. Genes & Development 29: 483-488.
Mi H. 2004. The PANTHER database of protein families, subfamilies, functions and pathways. Nucleic Acids
Research 33: D284-D288.
150
Miyazaki M, Miyazaki K, Chen S, Chandra V, Wagatsuma K, Agata Y, Rodewald H-R, Saito R, Chang AN, Varki N, et
al. 2015. The E-ld protein axis modulates the activities of the P13K-AKT-mTORC1-Hif1a and c-myc/p19Arf
pathways to suppress innate variant T FHcell development, thymocyte expansion, and lymphomagenesis.
Genes & Development 29: 409-425.
Mori S, Rempel RE, Chang JT, Yao G, Lagoo AS, Potti A, Bild A, Nevins JR. 2008. Utilization of Pathway Signatures
to Reveal Distinct Types of B Lymphoma in the E -myc Model and Human Diffuse Large B-Cell Lymphoma.
Cancer Research 68: 8525-8534.
Mu P, Zhang Z, Benelli M, Karthaus WR, Hoover E, Chen C-C, Wongvipat J, Ku S-Y, Gao D, Cao Z, et al. 2017.
SOX2 promotes lineage plasticity and antiandrogen resistance in TP53- and RB1-deficient prostate cancer.
Science 355: 84-88.
Mullighan CG, Goorha S, Radtke 1, Miller CB, Coustan-Smith E, Dalton JD, Girtman K, Mathew S, Ma J, Pounds SB,
et al. 2007. Genome-wide analysis of genetic alterations in acute lymphoblastic leukaemia. Nature 446: 758-
764.
Nechanitzky R, Akbas D, Scherer S, Gy6ry 1, Hoyler T, Ramamoorthy S, Diefenbach A, Grosschedl R. 2013.
Transcription factor EBF1 is essential for the maintenance of B cell identity and prevention of alternative fates in
committed cells. Nature Immunology 14: 867-875.
Niebuhr B, Kriebitzsch N, Fischer M, Behrens K, Gunther T, Alawi M, Bergholz U, Muller U, Roscher S, Ziegler M, et
al. 2013. Runx1 is essential at two stages of early murine B-cell development. Blood 122: 413-423.
Nordhausen K, et al., JADE: Blind Source Separation Methods Based on Joint Diagonalization and Some BSS
Performance Criteria. 2015. (available at https://cran.r-project.org/web/packages/JADE/index.html).
Nutt SL, Heavey B, Rolink AG, Busslinger M. 1999. Commitment to the B-lymphoid lineage depends on the
transcription factor Pax5. Nature 401: 556-562.
Oser MG, Niederst MJ, Sequist LV, Engelman JA. 2015. Transformation from non-small-cell lung cancer to small-cell
lung cancer: molecular drivers and cells of origin. Lancet Oncology 16: e165-172.
Painter MW, Davis S, Hardy RR, Mathis D, Benoist C, The Immunological Genome Project Consortium. 2011.
Transcriptomes of the B and T Lineages Compared by Multiplatform Microarray Profiling. Joumal of Immunology
186: 3047-3057.
Przychodzen, B.P., Polprasert, C., Hosono, N., Dienes, B., Miyano, S., Shiraishi, Y., Yoshida, K., Ogawa, S.,
Saunthararajah, Y., Maciejewski, J.P. and Makishima, H., 2013. Somatic PHF6 Mutations In Myeloid
Malignancies. Blood, 122(21), pp.2514-2514.
Ran FA, Hsu PD, Wright J, Agarwala V, Scott DA, Zhang F. 2013. Genome engineering using the CRISPR-Cas9
system. Nature Protocols 8: 2281-2308.
Rathert P, Roth M, Neumann T, Muerdter F, Roe J-S, Muhar M, Deswal S, Cerny-Reiterer S, Peter B, Jude J, et al.
2015. Transcriptional plasticity promotes primary and acquired resistance to BET inhibition. Nature 525: 543-
547.
Rothenberg EV. 2011. T Cell Lineage Commitment: Identity and Renunciation. Joumal of Immunology 186: 6649-
6655.
Rothenberg EV. 2014. Transcriptional Control of Early T and B Cell Developmental Choices. Annual Reviews
Immunoogy 32: 283-321.
Rutledge DN, Bouveresse DJ-R. 2013. Independent Components Analysis with the JADE algorithm. Trends in
Analytical Chemistry 50: 22-32.
Senchez-Rivera FJ, Jacks T. 2015. Applications of the CRISPR-Cas9 system in cancer biology. Nature Reviews
Cancer 15: 387-395.
Schebesta M, Heavey B, Busslinger M. 2002. Transcriptional control of B-cell development. Current Opinion in
Immunology 14: 216-223.
Schep AN, Buenrostro JD, Denny SK, Schwartz K, Sherlock G, Greenleaf WJ. 2015. Structured nucleosome
fingerprints enable high-resolution mapping of chromatin architecture within regulatory regions. Genome
Research 25: 1757-1770.
Schep AN, Wu B, Buenrostro JD, Greenleaf WJ. 2017. chromVAR: Inferring transcription factor variation from single-
cell epigenomic data. bioRxiv doi: 10.1101/110346
Schultz KR, Bowman WP, Aledo A, Slayton WB, Sather H, Devidas M, Wang C, Davies SM, Gaynon PS, Trigg M, et
al. 2009. Improved Early Event-Free Survival With Imatinib in Philadelphia Chromosome-Positive Acute
Lymphoblastic Leukemia: A Children's Oncology Group Study. Joumal of Clinical Oncology 27: 5175-5181.
Schwanbeck R, Martini S, Bernoth K, Just U. 2011. The Notch signaling pathway: Molecular basis of cell context
dependency. European Journal of Cell Biology 90: 572-581.
Sequist LV, Waltman BA, Dias-Santagata D, Digumarthy S, Turke AB, Fidias P, Bergethon K, Shaw AT, Gettinger S,
Cosper AK, et al. 2011. Genotypic and histological evolution of lung cancers acquiring resistance to EGFR
inhibitors. Science Translational Medicine 3: 75ra26-75ra26.
Shamroe CL, Comeau JM. 2013. Ponatinib: A New Tyrosine Kinase Inhibitor for the Treatment of Chronic Myeloid
Leukemia and Philadelphia Chromosome-Positive Acute Lymphoblastic Leukemia. Annals of Pharmacotherapy
47:1540-1546.
Shannon P. 2017. MotifDb: An Annotated Collection of Protein-DNA Binding Sequence Motifs. R package version
1.18.0.
Shi J, Wang E, Milazzo JP, Wang Z, Kinney JB, Vakoc CR. 2015. Discovery of cancer drug targets by CRISPR-Cas9
screening of protein domains. Nature Biotechnology 33: 661-667.
Smeets MFMA, Chan AC, Dagger S, Bradley CK, Wei A, Izon DJ. 2013. Fli-1 Overexpression in Hematopoietic
Progenitors Deregulates T Cell Development and Induces Pre-T Cell Lymphoblastic Leukaemia/Lymphoma ed.
C.H. Kim. PLoS ONE 8: e62346-12.
Somasundaram R, Sigvardsson M. 2015. Transcription factor networks in B-cell differentiation link development to
acute lymphoid leukemia. Blood 126:144-152.
151
Soto-Feliciano, Yadira. 2016. PHF6 is a novel regulator of B-cell identity in acute lymphoblastic leukemia. PhD
dissertation. Massachusetts Institute of Technology. http://hdl handle net/1 721 1/103165.
Stojnic R. 2016. Overview of the PWMEnrich package. R package version 4.12.0.
Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, Paulovich A, Pomeroy SL, Golub TR,
Lander ES, et al. 2005. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-
wide expression profiles. Proc Natl Acad Sci USA 102: 15545-15550.
Thomas PD, Campbell MJ, Kejariwal A, Mi H, Karlak B, Daverman R, Diemer K, Muruganujan A, Narechania A. 2003.
PANTHER: a library of protein families and subfamilies indexed by function. Genome Research 13: 2129-2141.
Thoms JAI, Birger Y, Foster S, Knezevic K, Kirschenbaum Y, Chandrakanthan V, Jonquieres G, Spensberger D,
Wong JW, Oram SH, et al. 2011. ERG promotes T-acute lymphoblastic leukemia and is transcriptionally
regulated in leukemic cells by a stem cell enhancer. Blood 117: 7079-7089.
Tijchon E, van Ernst L, Yuniati L, van Ingen Schenau D, Havinga J, Rouault J-P, Hoogerbrugge PM, van Leeuwen
FN, Scheijen B. 2016. Tumor suppressors BTG1 and BTG2 regulate early mouse B-cell development.
Haematologica 101: e272-6.
Todd MAM, Picketts DJ. 2012. PHF6 Interacts with the Nucleosome Remodeling and Deacetylation (NuRD) Complex.
Journal of Proteome Research 11: 4326-4337.
Tokoyoda K, Zehentmeier S, Hegazy AN, Albrecht I, GrUn JR, LOhning M, Radbruch A. 2009. Professional Memory
CD4+ T Ly mphocytes Preferentially Reside and Rest in the Bone Marrow. Immunity 30: 721-730.
Ungerback J, Ahsberg J, Strid T, Somasundaram R, Sigvardsson M. 2015. Combined heterozygous loss of Ebf1 and
Pax5 allows for T-lineage conversion of B cell progenitors. J Exp Med 212: 1109-1123.
Van Vlierberghe P, Palomero T, Khiabanian H, Van der Meulen J, Castillo M, Van Roy N, De Moerloose B, Philippe J,
Gonz~lez-Garcia S, Toribio ML, et al. 2010. PHF6 mutations in T-cell acute lymphoblastic leukemia. Nature
Genetics 42: 338-342.
Van Vlierberghe P, Patel J, Abdel-Wahab 0, Lobry C, Hedvat CV, Balbin M, Nicolas C, Payer AR, Fernandez HF,
Tallman MS, et al. 2011. PHF6 mutations in adult acute myeloid leukemia. Leukemia 25: 130-134.
Wada H, Masuda K, Satoh R, Kakugawa K, Ikawa T, Katsura Y, Kawamoto H. 2008. Adult T-cell progenitors retain
myeloid potential. Nature 452: 768-772.
Wang E, Kawaoka S, Roe J-S, Shi J, Hohmann AF, Xu Y, Bhagwat AS, Suzuki Y, Kinney JB, Vakoc CR. 2015. The
transcriptional cofactor TRIM33 prevents apoptosis in B lymphoblastic leukemia by deactivating a single
enhancer. eLife 4: e06377.
Wang GG, Song J, Wang Z, Dormann HL, Casadio F, Li H, Luo J-L, Patel DJ, Allis CD. 2009. Haematopoietic
malignancies caused by dysregulation of a chromatin-binding PHD finger. Nature 459: 847-851.
Wang J, Leung JWC, Gong Z, Feng L, Shi X, Chen J. 2013. PHF6 Regulates Cell Cycle Progression by Suppressing
Ribosomal RNA Synthesis. Joumal of Biological Chemistry 288: 3174-3183.
Wang L, Wildt KF, Castro E, Xiong Y, Feigenbaum L, Tessarollo L, Bosselut R. 2008. The Zinc Finger Transcription
Factor Zbtb7b Represses CD8-Lineage Gene Expression in Peripheral CD4+ T Cells. Immunity 29: 876-887.
Wang Q, Qiu H, Jiang H, Wu L, Dong S, Pan J, Wang W, Ping N, Xia J, Sun A, et al. 2011. Mutations of PHF6 are
associated with mutations of NOTCH1, JAK1 and rearrangement of SET-NUP214 in T-cell acute lymphoblastic
leukemia. Haematologica 96:1808-1814.
Wildt KF, Sun G, Grueter B, Fischer M, Zamisch M, Ehlers M, Bosselut R. 2007. The Transcription Factor Zbtb7b
Promotes CD4 Expression by Antagonizing Runx-Mediated Activation of the CD4 Silencer. Journal of
Immunology 179: 4405-4414.
Williams RT, Roussel, Sherr CJ. 2006. Arf gene loss enhances oncogenicity and limits imatinib response in mouse
models of Bcr-Abl-induced acute lymphoblastic leukemia. Proc Natl Acad Sci USA 1-6.
Wysocka J, Swigut T, Xiao H, Milne TA, Kwon SY, Landry J, Kauer M, Tackett AJ, Chait BT, Badenhorst P, Wu P, Allis
CD. 2006. A PHD finger of NURF couples histone H3 lysine 4 trimethylation with chromatin remodelling. Nature
442, 86-90
Xie H, Ye M, Feng R, Graf T. 2004. Stepwise reprogramming of B cells into macrophages. Cell 117: 663-676.
Yu HA, Arcila ME, Rekhtman N, Sima CS, Zakowski MF, Pao W, Kris MG, Miller VA, Ladanyi M, Riely GJ. 2013.
Analysis of Tumor Specimens at the Time of Acquired Resistance to EGFR-TKI Therapy in 155 Patients with
EGFR-Mutant Lung Cancers. Clinical Cancer Research 19: 2240-2247.
Zhang C, Mejia LA, Huang J, Valnegri P, Bennett EJ, Anckar J, Jahani-AsI A, Gallardo G, Ikeuchi Y, Yamada T, et al.
2013. The X-Linked Intellectual Disability Protein PHF6 Associates with the PAF1 Complex and Regulates
Neuronal Migration in the Mammalian Brain. Neuron 78: 986-993.
Zhang Y, Liu T, Meyer CA, Eeckhoute J, Johnson DS, Bernstein BE, Nusbaum C, Myers RM, Brown M, Li W, et al.
2008. Model-based analysis of ChIP-Seq (MACS). Genome Biology 9: R1 37.
Zhu J, Yamane H, Paul WE. 2010. Differentiation of Effector CD4 T Cell Populations. Annual Reviews Immunology
28: 445-489.
152
Chapter 3
Discussion and Future
Directions
While PHF6 was known to be a novel tumor suppressor in T-cell leukemia since
2010, it wasn't identified as an important modulator of B-cell leukemia cell growth until
2015, through an unbiased genome-wide RNAi screen [Van Vlierberghe et al. 2010a;
Meacham et al. 2015]. Since then, the action of PHF6 in B- or T-lymphocytes has
remained largely unknown. Here, we have demonstrated that PHF6 is vital for
maintaining the proper chromatin landscape for B-cell identity, and disruption of this
genomic organization has widespread consequences including dysregulated
transcriptional programs, altered disease presentation, lineage promiscuity, tolerance of
spurious pathway activation, and altered responses to anti-cancer therapies. The
changes in nucleosome positioning and chromatin organization upon loss of Phf6 allow
for B-cell leukemias to exhibit plasticity, leading to extensive alterations in malignant
cells. Below I will discuss in greater detail the broader significance of this work, as well
as future experimental directions. These discoveries have exciting implications in two
main fields of research: (1) the role of PHF6 in developing lymphocytes and throughout
hematopoiesis; and (2) the importance of the chromatin landscape in cancer and how
cellular plasticity affects the response to cancer therapy (Figure 3.1). I will examine the
153
significance of these findings and discuss future directions for the project and the field
below.
THE ROLE OF PHF6 IN LYMPHOCYTES
The study of Phf6 has mainly been restricted to malignant cells, specifically B-
and T-cell leukemias, due to the work of the Ferrando and Hemann Labs [Van
Vlierberghe et al. 2010; Meacham et al. 2015]. We have demonstrated that loss of Phf6
in B-ALL cells results in dramatic reprogramming processes, similar to that seen upon
deletion of Pax5 or Ebfl in nonmalignant cells (Figure 1.5-1.6). Below I will begin by
discussing future lines of investigation that could further detail the role of Phf6 in B-ALL
cells. Subsequently, I will describe experiments investigating the contribution of Phf6 in
normal hematopoiesis in nonmalignant B-cell precursors.
*Leukemic Plasticity
- Determining Engagement of the T-cell Program
Upon transplantation of Phf6KO B-ALL cells into recipient mice, disease presents
with characteristics of a mixed-lineage lymphoma, expressing the T-cell marker CD4
(Figure 2.2-2.3). We also observe increased chromatin accessibility at T-cell receptor
gene loci, specifically Trbjl-1 (T-cell receptor beta joining 1-1), and Trav2 (T-cell
receptor alpha variable 2) (Chapter 2 - Figure 2.12A). Rearrangement of the T-cell
receptor occurs before and during developmental stages that correspond with CD4
expression [Rothenberg 2014]. Since Phf6KO B-ALL cells resemble T-cells at the
transcriptome and chromatin level, I hypothesize that engagement of T-cell
development processes may have occurred, specifically rearrangement of TCR genes.
In order to determine if TCR gene rearrangements occurred upon knockout of Phf6,
established PCR based protocols exist [Gartner et al. 1999]. Briefly, combinations of
specific PCR primers will amplify regions of the TCRP locus. The amplification products
can be run on an agarose gel and visualized by ethidium bromide. Differences in
amplification product length suggest that a recombination event occurred. The
experimental design can be found in Figure 3.2, and will report the extent to which
Phf6KO B-ALL cells (preferably in vivo tumor cells) travel along the road of T-cell
development.
154
0
Generating Phf6KO H
in Nonmalignant Re
CLPs and pro-B P
Cells 
W
Transcription Factor
Dosage Testing the Potential
of B-ALL to Convert
Lineages
Discerni
PHF6 in
Determining the
ngagement of the Leukemia Plasticity
T-cell Program
Ig
Hi
Studying
ematopoietic
constitution in
)hf6Ko HSCs
Identifying
Interacting Partners
with the BirA System
Determining Pioneer
TF Activity
the Role of
ematopoiesis Histone
Binding Profile of PHF6
0
THE ROLE OF PHF6 IN LYMPHOCYTES
THE IMPORTANCE OF THE CHROMATIN LANDSCAPE IN LEUKEMIA
Plasticity as an The Contribution
Emerging Mechanism of Oncogenic Epigenetic
of Resistance to Signaling to Inhibitors After
Targeted Therapies Cellular Plasticity CAR-T
Resistance to CD1 9- Therapy
CAR-T Therapy in Rb, PtenB-ALL p53 in
B-ALL Murine Models Relapsed
Driven by Different R-ALI
DNA-Seq
ATAC-Seq of
Relapsed Samples
Oncogenes
0
Contribution of IL-6 to
Lineage Switching
Figure 3.1. Summary of Future Directions. Arms represent experiments with similar overarching themes. Branch location
represents ease of experimental procedure. Size of circle represents personal opinion of importance. TF, transcription factor. B-
ALL, B-cell acute lymphoblastic leukemia.
155
Lineage
Infidelity
In Vvo
*A
E
RNA-
,
- Dosage of Transcription Factors
Significant hairpin-mediated knockdown of Phf6 is detrimental to leukemia cell
growth, yet complete knockout of Phf6 allows for cells to undergo lineage conversion,
as shown in this thesis [Meacham et al. 2015; Soto-Feliciano 2015]. Hairpin-mediated
reduction in Phf6 expression did not drastically affect the transcriptome of B-ALL cells
(Figure 2.5E and Supp. Figure S2.4A). Although we did not perform ATAC-Seq analysis
on shPhf6 cells, I hypothesize that there will not be wide-spread changes in the
chromatin landscape, reflective of the lack of changes seen in the transcriptome.
Therefore, the growth disadvantage that we observe in shPhf6 B-ALL cells in vivo may
be the results of a partial loss of B-cell identity, yet not complete dysregulation of the
chromatin architecture that allows plasticity and lineage conversion to occur, as seen
with Phf6KO B-ALL cells. Therefore I want to test the effects of combinatorial knockdown
of other epigenetic factors (such as Pax5, Ebf1, E2a, Runxl) in shPhf6 B-ALL cells to
see if global decreases in TF dosage could push cells to obtain plasticity and the
identity of another lineage. This is indeed true for trans-heterozygous deletion of
Pax5+/- and Ebfl+/- CD19+ precursor B-cells and leukemias, which show potential to
switch to both the T- and myeloid-lineages, whereas heterozygous deletion of each TF
singularly has only subtle changes [Somasundaram et al. 2016; Ungerback et al. 2015].
Our preliminary data suggests that interruption of transcription factor dosage levels is
generally detrimental to leukemia cell growth, as seen previously with knockdown of
Phf6 and described here with reduction of Tcfl2 and NfKb in B-ALL cells leading to
increased survival of leukemia recipient mice (Figure 2.8). Whereas reduction of single
TF levels may be deleterious to leukemia cells, combined reduction of multiple factors
may allow leukemia cells to adopt programs of other lineages, releasing the brakes set
up during lineage restriction and allowing leukemia cells to thrive again.
- Lineage Infidelity In Vivo
The large-scale integrated genomic analyses executed in this body of work were
all performed in cultured cells, including RNA-Seq, ChlP-Seq, and ATAC-Seq. This
highlights that loss of Phf6 results in an underlying lineage plasticity that can be
detected in vitro, and upon transplant into immunocompetent recipient mice (and
156
TCRP Locus
VV -
=L* vn-S
,,E1 110100
I I- I" IMENNEN
Vn Vn Vn D1 J1
I 1111121
C1 D2 J2
V,D, or J? Primer Name Sequence (5'-3')
TCRB-D1 U-S GAGGAGCAGCTTATCTGGTGGTTT
D 
_-5' GTAGGCACTGTGGGGAAGAAACT
TCRB-JID-A CACAACCCCTCCAGTCAGAAATG
J-3' TCRB-J2D-A TGAGAGCTGTCTCCTACTATCGATT
Figure 3.2. (A) Schematic of the murine TCR3 locus. V, D, and J segments are indicated. C, constant region. Red arrows
indicated forward PCR primers. Blue arrows indicate reverse PCR primers. Partial TCR rearrangement is defined as D-J joining,
whereas complete rearrangement results in V-DJ recombination. (B) Table listing the PCR primer sequences. Adapted from
GArtner et al. 1999.
157
A
D1-LS
5, I
J1-D-A D2-U-S J2-D-AA-
B
3'
C2
AAATGAGACGGTGCCCAGTCGTTTCRB-V1-S
TCRB-V2-S TCCTGGGGACAAAGAGGTCAAATC
TCRB-V3-S GAAAAACGATTCTCTGCTGAGTGTCC
TCRB-V4-S AGCTATCAAAAACTTATGGACAATCAG
TCRB-V5-S CAGCAGATTCTCAGTCCAACAGTTT
TCRIB-V-S AAGGCGATCTATCTGAAGCTATGA
TCRB-V7-S AGCTGATTTATATCTCATACGATGTTG
TCRB-V8-S TATATGTACTGGTATCGGCAGGACA
TCRB-V9-S TTCCAATCCAGTCGGCCTAACAAT
TCRB-VIO-S GCGCTTCTCACCTCAGTCTTCAG
TCRB-V11-S TTCTCAGCTCAGATGCCCAATCAG
TCRB-V12-S AGCTGAGATGCTAAATTCATCCTTC
TCRB-V13-S CTGCTGTGAGGCCTAAAGGAACTAA
TCRB-V14-S AGAGTCGGTGGTGCAACTGAACCT
TCRB-V15-S CCCATCAGTCATCCCAACTTATCC
TCRB-Vi6-S GATTTTAGGACAGCAGATGGAGTTTC
TCRB-V17-S TCGAAATGAAGAAATTATGGAACAAAC
TCRB-V18-S CCGGCCAAACCTAACATTCTCAAC
TCRB-V19-S CTACAAGAAACCGGGAGAAGAACTC
CTGGTATCAACAAAAGCAGAGCAAATCRB-V20-S
interaction with cytokines, immune cells, stroma, etc.) are capable of lineage conversion
by acquiring T-cell characteristics. The striking differences we see in tumor presentation,
disease states, and response to chemotherapy between Phf6wT and Phf6KO tumors
suggests that repeating the large-scale genomic analyses from in vivo tumor cells
collected from lymphoid organs (bone marrow, lymph nodes, thymus) may provide
additional valuable insight into the mechanism of lineage conversion upon Phf6 loss,
and call attention to possible mediators that are controlling this process.
In Chapter 2 (Figure 2.11), we show that loss of Phf6 leads to increased
accessibility of DNA that is enriched for binding motifs of TFs associated with T-ALL and
AML. This suggests that the observed lineage conversion is a result of T-cell TFs
binding DNA and driving a T-lineage transcriptional program. In order to support this
model, we need to demonstrate the binding of a T-cell TF to its target gene, activating
its expression in B-ALL cells. The TF motif analysis from our current ATAC-Seq data is a
starting point to determine factors that may be driving the lineage conversion (ETS1,
ELK1, ERG, FL11, ETV1). Further, transcriptome analysis from in vivo tumors may
identify a subset of TFs and their targets that become upregulated in vivo during
tumorigenesis. Finally, performing ChIP-qPCR of these TFs at the promoters of the
target genes from in vivo tumors would reinforce the model thus presented.
- Testing the Lineage-Promiscuity Potential in B-ALL
In Chapter 2 we provide evidence that loss of Phf6 in precursor B-cell acute
lymphoblastic leukemia cells allows for adoption of T-lineage characteristics. As
discussed above, these results are reminiscent of that seen for B-cell leukemias with
heterozygous mutations in B-cell master transcription factors Pax5 and Ebfl
[Ungerback et al. 2015]. Like Phf6KO B-ALL cells, Pax5+/-Ebfl+/- B-ALL cells
downregulate expression of B-cell genes and can give rise to CD4 expressing cells in
vivo. Interestingly, this conversion is dependent on NOTCHI signaling. As we have
shown in Figure 2.16, Phf6KO B-ALL cells are also responsive to exogenous expression
of NOTCHI. Therefore, the plasticity of Phf6KO cells can be further tested in vitro by
recapitulating T-cell promoting conditions as done by the Sigvardsson Group
[Ungerback et al. 2015; Somasundaram et al. 2016; Schmitt & ZOhiga-Pflucker, 2002;
Holmes & ZOhiga-Pflucker, 2009]. Briefly, Phf6KO B-ALL cells can be cultured on OP9-
158
DL1 stromal cells in the presence of specific cytokines or injected into Rag1-- recipient
mice. Subsequent immunophenotyping and transcriptomic analyses will be performed to
test for lineage conversion to T-lineage cells, as described in Figure 3.3. Additionally,
Phf6KO cells can be grown under conditions that favor myeloid cell growth (IL-3, IL-6,
SCF, GM-CSF) and myeloid-plasticity can be assessed. Providing conclusive evidence
that Phf6 loss phenocopies that seen during loss of master B-cell transcription factors
will allow us to gain confidence that Phf6 does indeed play important roles in
lymphocyte development. The next step is to definitively test the role of Phf6 in
lymphoid lineage specification using nonmalignant precursors, as discussed in more
detail below.
T-lineage
conditions
OP9-DL1 stromal cells
+ Kit, Flt3, and 1L7
~14 days
+-O/E of
NOTCH1
sr
Phf6KO B-
ALL cells
Myeloid
q B 
conditions
OP9 stromal cells + Kit, FIt3,
1L7, 1L3, 1L6, SCF, GM-CSF
- 14 days
)etermine In vivo
reconstitution
6-9 weeks post-
transplantation:
pleen, thymus, BM
+1- O/E of
CEBPa or CEBPP
TranSarin* A nnsi@
Immunobhenotvoina Iripi nalysis Immunophenoing = U A
CD3, CD4, CD8, TCRP CD3e, Pre-Ta, Lck, Gata3, CD11b, Mac, Gr-1 PU.1, Mpo, 116, H10,CD27 H12p40, TNFd
Figure 3.3. Testing the plasticity of Phf6KO B-ALL cells. In order to test if Phf6KO B-ALL cells can be pushed to the T-cell
lineage or myeloid lineage without further genetic perturbations, the cells can be cultured in specific conditions that foster either
myeloid or T-cell growth (as specified). These cells can also be injected into Rag1-- mice. In order to test for lineage conversion,
immunophenotyping and transcript analyses should be performed as indicated. Overexpression (O/E) of NOTCH1 or CEBPa/
CEBP3 can also be done to further push cells to another lineage. Culture conditions described in Ungerbtck et al. 2015,
Somasundaram et al. 2016, Schmitt & ZOfiga-Pflucker, 2002, Holmes & Zfhiga-Pflucker, 2009.
159
I I I i
>-Discerning the Role of PHF6 in Hematopoiesis
- Generating Knockouts of Phf6 in Specific Cell Populations
Above we discussed the ability of Phf6KO B-leukemia cells to transfer its
malignant states to other lineages, giving rise to T- or myeloid-lineage-like leukemias.
However, it is vital to elucidate the function of PHF6 outside of the leukemia context in
nonmalignant cells during hematopoiesis and lymphoid differentiation. As discussed in
Chapter 1 - Part 1, hematopoietic cells have varying degrees of plasticity, especially
upon genetic deletion of master transcription factors like Ikzfl, Pax5, and Ebfl (Figure
1.5) [Nutt et al. 1999; Mikkola et al. 2002; Reynaud et al. 2008; Pongubala et al. 2008;
Xie et al. 2004; Cobaleda et al. 2007; Rolink et al. 1999; H6flinger et al. 2004;
Nechanitzky et al. 2013]. In these studies, the labs determined the degree of plasticity to
the T- and myeloid- lineages by first genetically ablating the gene in question in
common lymphoid progenitor cells (CLPs), or more often in pro-B-cells, and analyzing
significant changes in cellular populations. Using CRISPR-Cas9 technology we can
easily generate genetic knockouts in these specific cell populations. The experimental
design is detailed in Figure 3.4. In the case of Phf6, pro-B cells (Lin-, CD19+, B220+,
IgM-) can be sorted from the bone marrow of donor male mice (having only 1 copy of
Phf6) and transiently transfected with a vector expressing the sgRNA targeting Phf6 and
the Cas9 nuclease. Alternatively, pro-B cells from a Cas9-expressing mouse can be
infected with the sgRNA. Cells that received the sgRNA can then be sorted into single
cells and grown up to produce isogenic cell lines [Nutt et al. 1999]. After identifying lines
that have successfully deleted Phf6 (either by Western blot or droplet digital PCR if
material is limited), cells can then be analyzed for plasticity capacity either in vitro (via
culture conditions with OP9-DL1 stromal cells or addition of IL-3, IL-6, SCF, GM-CSF) or
in vivo (determining reconstitution of cell lineages upon transplantation). Additionally, the
overexpression of lineage-specific transcription factors (Notch1, CEBPa, CEBP3) is
documented to cause lineage conversion in mature B-cells and B-cell precursors, in
both wild-type cells or cells deleted for the following genes: Pax5--, Ebfl-, Pax5+/-
Ebf+l-, and Ikzfl-I- [Xie et al. 2004; Cobaleda et al. 2007; Ungerback et al. 2015;
Hoflinger et al. 2004; Nechanitzky et al. 2013]. Due to the fact that knockout of Phf6
shows lineage fluidity in pre-B ALL cells, I believe that a Phf6-knockout in an earlier
160
Male Cas9 mouse
or WTC57BL6 -----------
mouse
....-.. FACS sort for
pro-B cells
.0
.0 *
I- --- Infect with Phf6sgRNA or transientlytransfect with All-In-One Cas9/sgRNA
vector
Pro-B cells can be
cultured on OP9
cells with IMDM...........
media + IL-7
[Nutt et al 1999]
Continue with
confirmed Phf6KO.
pro-B clones
T
Culture with OP9-
DL1 storma cells
T
~I --- -
..----
M
Single cell sort for
GFP-positive cells
Confirm genetic KO
by Western Blot or
ddPCR
M T/M
4
CEBPa
UCEBP
Rag1-- mice
O/E of NOTCH1 O/E ofCEBPa or CEBPP
Culture with IL-3,
IL-6, SCF, GM-CSF
In vivo
Reconstitution
Figure 3.4. Experimental design to test the plasticity of nonmalignant B-cell precursors using CRISPR/Cas9 technology. pUSCG is
the sgRNA expression vector that is GFP-tagged. pX458 is the All-In-One Cas9/sgRNA vector that is GFP-tagged. O/E,
overexpression. T, testing for T-cell lineage plasticity. M, testing for myeloid lineage plasticity. ddPCR, droplet digital PCR.
161
Pro-B cells
Lin-
CD19+
B220+ -----------
CD43hi
IgM-
pUSCG
or pX458
F* NICD
precursor population may phenocopy the results seen in pro-B cells that have lost
master transcription factors like PAX5, EBF1, or Ikaros. Preliminary data in a pro-B cell
leukemia model driven by MLL-AF4 also shows that Phf6 appears to play an important
role in maintaining B-cell transcriptional programs in an earlier precursor population
(Figure. 2.19). The experiments detailed up to this point focus on the dysregulation that
occurs in a specific cellular subset (CLPs vs pro-B vs pre-B cells). Although much
information can be obtained through these studies, we need to investigate the effect of
Phf6 loss on global hematopoiesis, as discussed below.
- Studying the Effect of Phf6KO in Development and Hematopoiesis
To date, there are no published studies concerning the role of Phf6 in
hematopoiesis or murine development. In order to conclusively study the importance of
Phf6 in hematopoiesis or embryonic development, animal knockout and reconstitution
experiments are necessary. Preliminary results from the Speleman Lab showed that in
vitro differentiation of human hematopoietic progenitor cells experience accelerated T-
cell maturation with enrichment of CD4+CD8+ double-positive T-cells upon knockdown
of PHF6 [Loontiens et al. 2017, data unpublished]. Further, a TALEN-based PHF6
knockout zebrafish model from the same group showed increased expression of Imo2
and c-myb, with decreased expression of ragi during zebrafish embryonic development
[Janssens et al. 2014, data unpublished]. This suggests that PHF6 may be important in
lymphocyte precursors to silence HSC genes and activate expression of Ragi in order
to facilitate B-cell receptor or T-cell receptor gene rearrangements. Additionally, these
results suggest that PHF6 may also be involved in suppressing T-cell development, and
knockdown allows for accelerated T-lineage development. I hypothesize that Phf6 plays
an important role in hematopoiesis, possibly by propagating B-cell lineage
transcriptional programs while simultaneously repressing T-cell programs. Examining
this more closely, the binding profile of PHF6 has a large overlap with that of B-cell
master transcription factors EBF1 and PAX5. Approximately 42.6% of activated and
36% of repressed EBF1 target genes, and 19.5% of putative PAX5 target genes are
also bound by PHF6 [Trieber et al. 2010; Schebesta et al. 2007; Pridans et al. 2008;
McManus et al. 2011]. Thus, in order to determine the contribution of PHF6 to the
162
process of hematopoiesis, it is necessary to perform a hematopoietic stem cell
reconstitution experiment with Phf6 deficient HSCs [Cheng et al. 2013].
Performing a hematopoietic reconstitution experiment with Phf6KO HSCs can be
broken down into 4 steps: (1) isolation of the HSC pool from the bone marrow cells of
donor mice (CD45.1); (2) generating Phf6KO hematopoietic stem cells and single cell
sorting; (3) transplanting single HSCs with host (CD45.2) bone marrow to aid in
engraftment; and (4) assessing bone marrow reconstitution and lineage makeup 16
weeks to 6 months post-transplantation. The experiment is outlined in Figure 3.5. A key
challenge here is to ensure that CD45.1 donor HSCs did indeed sustain a knockout in
Phf6. Modification of a given allele with CRISPR/Cas9 in mouse ESCs is greater than
70% [Li et al. 2013; Yang et al. 2014]. Using male mice can increase the rate of gene
editing by needing modification in only one copy of the X-linked gene Phf6. Further,
47% of highly purified Lin-c-kit+Sca-1+CD34-CD150+CD48- HSCs have the ability to
achieve long-term bone marrow reconstitution in multiple lineages of irradiated mice
[Kiel et al. 2005; Cheng et al. 2013]. Therefore, at least 20 mice should be injected with
single Phf6KO HSCs and analyzed for bone marrow reconstitution for each experimental
cohort [Charan & Kantharia, 2013]. To test for HSC self-renewal capability, serial
transplantation of HSCs into recipient mice can also be performed. Use of constitutively
Cas9-expressing murine hematopoietic stem cells can also be used to increase gene
editing rates [Platt et al. 2014]. Any hematopoietic cell population deficits or increases
will provide novel insights into the influence of PHF6 during lymphoid differentiation.
Due to the promotion of T-cell characteristics upon genetic deletion of Phf6 in B-cell
leukemia cells, I hypothesize that single-Phf6KO HSC reconstitution experiments will
result in depletion of B-lymphocytes, and possibly an increase in the T-lineage arm of
lymphoid development. Failure to reconstitute any cell type would suggest that PHF6
also has important roles in HSCs, and more refined experiments would need to be
performed to study this possibility, such as the generation of a Phf6KO mouse [Yang et
al. 2014].
163
Male donor mouse
(CD45. 1)
Sort for HSC oool:
Lin-
c-kit+
Sca-1+
CD34-
CD150+
CD48-
Transfect with Cas9 + Phf6 sgRNA
or
Mock transfect
44Single-cell sort for single
guide-containing cells
Add 200,000 Ly5.2
donor BM cells
Irradiate donor mice
Male host mouse (CD45.2)
Inject into lethally irradiated mice
Blood collections
Weeks after transplant
-- - ----------- --------- ----------- ---------------
.11 . _' " _. 1.Ulannor
---------------------------
1. I31 1111 l M Myeloid lineageT-cell lineageM B-cell lineage
Individual Mice
-- -- -- - -- - - - - -- - - - -
...... ........... pil I~u I
Figure 3.5. Hematopoietic reconstitution experiment to test the role of Phf6 in hematopoiesis. Addition of bulk donor bone marrow
(BM) cells will increase engraftment success. Irradiate donor mice (-10 Gy) less than 12 hours prior to transplantation. Monitor
engraftment every 4 weeks. Reconstitution is achieved upon when >0.1% of hematopoietic cells are CD45.1+. Full reconstitution
can occur after at least 16 weeks - 6months post injection is optimal. Confirm Phf6 knockout status in reconstituted mice. In single
cell reconstitution experiments, 20/30 mice show some sort of reconstitution. Adapted from Cheng et al. 2013.
164
4
I.-)
I
E
(D
E) L6
A
*Interrogating the Binding Profile of PHF6
- Histone Arrays
The work in this thesis has shown that PHF6 plays an integral role in modulating
the chromatin landscape through organizing nucleosome positions at regions that are
enriched for B- and T-lineage transcription factors (Figure 2.11). Additionally, we show
that PHF6 has direct protein-protein interactions with histones, specifically histone H3
(Figure 2.9). Biochemical analyses performed by various groups have demonstrated
that the PHD domains found in PHF6 are atypical and have only been found to bind
double-stranded DNA in a non-specific manner [Liu et al. 2014] (also reviewed in
Chapter 1 - Part 2). However, histone proteins were found to be prevalent hits in an IP/
MS experiment, and taken together with our data, suggest that PHF6 has bona fide
histone-binding capabilities [Todd & Picketts 2012; Soto-Feliciano et al. 2017]. The
atypical PHD domains of PHF6 lack the aromatic cage that is necessary for recognition
of H3K4me3, the PHD domain's canonical substrate, and are also missing the stretch of
acidic residues that are important in binding unmodified histones [Musselman et al.
2012]. However, these properties are similar to other PHD-domain containing proteins,
ATRX and BRPF2, which also lack these features but have well-defined roles in binding
to H3K9me3, unmodified H3K4, and dsDNA [Dhayalan et al. 2011; Iwase et al. 2011; Liu
et al. 2012]. Interestingly, mutations in ATRX also are the cause of an intellectual
disability disorder (alpha-thalassemia and mental retardation X-linked syndrome),
similar to PHF6's causal role in BFLS [Baker et al. 2008]. Therefore, further
investigation into the binding properties of PHF6 to histones and histone post-
translational modifications are necessary.
A first-step, unbiased way to gauge the ability of PHF6 to bind histones, specific
histone post-translational modifications, or to observe how neighboring modifications
affect PHF6 binding is through the use of a histone peptide array to screen binding
specificity of recombinant PHF6 (Figure 3.6). The histone peptide array contains
hundreds of combinations of PTMs (acetylation, methylation, phosphorylation, and
citrullination) of histone proteins H2A, H2B, H3 and H4, and can be commercially
bought (i.e. Active Motif - Cat. No. 13001). Briefly, PHF6 would be incubated with a pre-
blocked histone peptide array. After washing away excess protein, the binding of PHF6
165
to histones can be determined by using an anti-PHF6 antibody, similar to Western
blotting procedures, and detected by ECL and film exposure. The location and intensity
of any binding events, or spots, can be quantified, revealing potential binding specificity
of PHF6 to histone proteins or PTMs. Subsequently, any results from the peptide screen
would have to be validated individually.
The extent of PHF6 binding to histones and the specific mechanism of
recognition is still largely unknown. If PHF6 does not bind a specific histone PTM,
further work is necessary to investigate the process that allows for protein-protein
interactions to occur. The following sections will describe experiments that can begin to
answer these questions, by exploring the capacity of PHF6 as a novel pioneer factor or
chromatin regulator.
Incubation with *4- H3K9ac?
Antibodies:
Primary: anti-PHF6
Blocking Secondary: anti-
Po *lcigRabbit-HRP
Incubation with 3K4me3?
PHF6 Wash
Wash Detection by ECL
S.-H3K4?
op- Visualization by film Histon Peptide
Array
Histon Peptide
Array
Figure 3.6. Histone peptide array to elucidate any specific histone recognition by PHF6. Any candidate substrates will require
further validation. Adapted from Nady et al. 2008.
- Validating Possible Pioneer TF or Chromatin Boundary Regulator Behavior
In Chapter 1 - Part 3, I discussed the different types of transcription factors:
canonical TFs, pioneer TFs, migrant TFs, settler TFs, and chromatin state regulators.
Each of these subsets of transcriptional regulators interacts with chromatin and DNA in
distinct and specialized ways. For example, pioneer transcription factors bind silenced
genes in closed chromatin and facilitate the activation of gene expression by
166
coordinating a permissive, accessible environment. In line with this ability, pioneer
factors have well documented roles in cell reprogramming [Iwafuchi-Doi et al. 2014].
Our work has uncovered a role for PHF6 in maintaining B-cell identity, and that loss of
this factor can lead to reprogramming to a T-cell-like state. Additionally, ChIP-Seq and
ATAC-Seq data suggest that PHF6 binds to regions of both open and closed chromatin.
Investigating the nature of PHF6 binding and its influence on chromatin accessibility in
both B- and T-cells will help to distinguish how PHF6 regulates B-cell identity. The
hallmark activity of pioneer TFs is to bind to target genes in heterochromatin and
facilitate its opening to a transcriptionally permissive state. Testing pioneering activity
can be done in the following two ways [Iwafuchi-Doi et al. 2014]:
(1) Observe the chromatin accessibility status of PHF6 target sites before and
after PHF6 is expressed. Top PHF6 target genes from ChIP-Seq analysis (Table S1)
include B-cell genes Spib, Adcy7, and Rhoq, T-cell genes Zbtb7b and Hes5, and
myeloid gene C/ebp#. PHF6 binding behavior in T-ALL has also been interrogated, with
top target genes including RUNXI, DMNT3A, NOTCH1 and JAGI [Meacham et al.
2015]. By expressing an exogenous Phf6 cDNA in Phf6KO B-ALL and T-ALL cells, we
can assess chromatin accessibility at target genes before and after Phf6 expression by
ATAC-qPCR. Further, we can parse out the context-specific roles for Phf6 by looking at
the B-ALL and T-ALL targets in the opposite cell type after cDNA expression, or in a
non-hematopoietic lineage (i.e. fibroblasts). If Phf6 is acting as a pioneer factor, we
would expect increased accessibility at target genes, regardless of lineage context. If
Phf6 does not affect the state of accessibility after expression, it is then more likely that
Phf6 is acting as a chromatin state regulator, propagating and maintaining the chromatin
landscape after the action of a pioneer factor. Although we do not yet know how
expression of Phf6 affects chromatin accessibility, we do know how gene targets
behave upon loss of Phf6. On average, approximately 9% of the top PHF6 target genes
from ChIP-Seq analyses undergo significant changes in chromatin accessibility upon
loss of Phf6, including II4ra, Btg2, Erg, Lefi, and Tnnt2 (Figure 3.7A).
(2) Determine if direct biding occurs between PHF6 and reconstituted
mononucleosomes by EMSA. End-labelled specific or non-specific DNA fragments can
be used to coat mononucleosomes. Interactions between PHF6 and mononucleosomes
167
can be determined by electrophoretic mobility shift assays (EMSAs) [Soufi et al. 2015].
These studies will provide crucial insight as to how PHF6 exerts transcriptional control
at the nucleosomal level. However, due to the lack of an enzymatic domain in PHF6, I
hypothesize that this chromatin landscape control is largely mediated through
interactions with other chromatin modifying/remodeling proteins and complexes. In the
following section I will describe an experimental method to determine possible PHF6
lymphocyte interactors.
A B
Phf6wr B-ALL cells Phf6KO B-ALL cells 
35000
7.36614399 7.06932747 9183167 5.5700101 S.75439727 Btg2 30000Open to 7.20512643 7.26418424 .68860Un 5.6437532 6.23850889 ll4raClose 7.5509531 7.03088735 5.96706652 6.09310136 5.61524034 Erg
Close to 6.70752664 6.39201063 7.99246716 .17160216 .06373992 Tnnt2 25000fa
Open 6.97767442 6.86596484 7A648581 D1464s317 7.9995465 Lef 1 20
-is0.01.9200000
10g2(ATAC-Seq Signal)
Figure 3.7. (A) Heatmap of ATAC-signal at B- and T-cell genes in Phf6wr and Phf6KO B-ALL cells. (B). Nucleosome free regions
(NFR) in Phf6wT and Phf6KO B-ALL cells from ATAC-Seq data.
>How PHF6 Exerts Transcriptional Control Through Interaction with Other
Proteins
Previous analyses have described and validated interactions between PHF6 and
members of the Nucleosome Remodeling and Deacetylation Complex (NuRD), the
PAF1 transcriptional elongation complex, and the rDNA transcription factor UBTF
(Chapter 1- Figure 1.12) [Todd & Picketts, 2012; Zhang et al. 2013; Wang et al. 2013].
While these studies describe vital roles for PHF6 in transcriptional regulation in
combination with well-characterized protein complexes, they do not address the function
of PHF6 in the lymphoid context. Additionally, we could not confirm an interaction
between PHF6 and members of the NuRD complex in B-ALL cells, warranting further
168
investigation into lymphocyte-specific protein interactors of PHF6. Rather than antibody-
mediated immunoprecipitation experiments, use of the BirA biotin ligase allows for
quick, efficient, and gentle purification of PHF6 and potential interacting proteins,
detailed in Figure 3.8.
- BirA Biotinylation Followed by Mass-Spectrometry
Fusion of a small biotin-acceptor peptide (-23 amino acids) to a protein of
interest (POI), coupled with expression of the bacterial BirA biotin ligase in cells can
achieve high-quality purification in a single step using streptavidin beads. Proteins that
interact with a POI can subsequently be identified by mass spectrometry. Fusion of the
tag to the POI is shown to have no interference with the protein's DNA- or protein-
binding capability, and the high biotin/streptavidin affinity (Kd = 10-14 M) allows for
efficient transcription factor complex purification [de Boer et al. 2003; Green, 1963]. This
method has been utilized in vivo for Ikaros, Gata-1, and Pax5 to aid in purification
studies followed by mass spectrometry or sequencing of associated DNA [de Boer et al.
2003; Geimer Le Lay et al. 2014; McManus et al. 2011; Rodriguez et al. 2005].
Therefore, generation of a knock-in biotin-tag to the endogenous Phf6 allele and
expression of the BirA ligase in lymphoid cells, followed by streptavidin purification and
mass spectrometry is an efficient way to identify high-confidence PHF6 interacting
partners.
Alternatively, rather than tagging PHF6 itself, the biotin-BirA system can be co-
opted to tag proximal and interacting proteins within the vicinity of PHF6. Subsequent
purification with streptavidin beads will isolate the partners of PHF6 itself, thus
decreasing the potential loss of proteins during purification steps [Roux et al. 2012].
This technique, named BiolD, involves fusion of a promiscuous version of BirA (R118G)
to a protein of interest. Rather than selectively biotinylating only the "biotin tag," the
promiscuous version of BirA will nonspecifically biotinylate all proteins that are
neighboring the POI. Fusion of promiscuous BirA, a 35-kDa protein, to Phf6, along with
addition of biotin to tissue culture medium, results in biotinylation of PHF6-interacting
partner proteins within a cell. These proteins can then be efficiently pulled down via
streptavidin beads and identified by mass spectrometry. The use of the biotin/BirA
tagging system proves to be an exciting and easily amenable method to identifying
169
Biotinylated PHF6
Biotin
Promiscuous BirA System
Transcription Factors
S
I
I
4 Ali
Chromatin Remodeling Complex
Streptavidin
Beads
I
I
-------
~3~)
m - - - m - - - - m - - m - - -
I
I
I
~22 0*
.0 0
(UW I
0<
Mass Spectrum
Figure 3.8. Using the BirA ubiquitin ligase to identify interacting protein partners with PHF6. (Left) Depiction of PHF6-biotin tag
fusion protein that is biotinylated by BirA. Proteins in complex with PHF6 will be purified upon pulldown of PHF6 by streptavidin
beads and subsequently identified by mass spectrometry. (Right) Depiction of PHF6 fused to the promiscuous ubiquitin ligase
version of BirA, in which the neighboring interaction partner will become biotinylated, specifically pulled down and identified by
mass spectrometry.
170
Blotin Acceptor Tag System
Biotin Acceptor Tag
4
4
4
I
U
I
I
M
**T**-O
partners of PHF6, as depicted in Figure 3.8. These experiments aim to discover how
PHF6 contributes to the process of transcriptional control, how PHF6 is recruited to
target genes, and to parse out the direct and indirect consequences of Phf6 loss. These
experiments will also help determine specifically how PHF6 exerts transcriptional control
at target genes by identifying important factors within the lymphoid context.
THE IMPORTANCE OF THE CHROMATIN LANDSCAPE IN LEUKEMIA
In this thesis, we have demonstrated that the maintenance of a proper chromatin
landscape is essential for maintaining B-cell identity. More importantly, we have
revealed how disruption of the chromatin landscape can drastically change a tumor's
response to chemotherapy and targeted treatments. It is becoming increasingly evident
that malignant cells are rewiring their genomes through epigenetic mechanisms in order
to evolve and escape therapeutic suppression. This section will focus on developing an
understanding of exactly how the chromatin landscape contributes to tumorigenesis and
resistance channels observed clinically.
>Plasticity as an Emerging Mechanism of Resistance to Targeted Therapies
Recently, a variety of cancers acquire resistance to targeted therapeutics by
relapsing as a histologically different malignancy. For example, patients with non-small-
cell lung cancer (NSCLC) treated with a targeted EGFR inhibitor can relapse via
pathological transformation to small-cell lung cancer (SCLC), a distinct malignancy with
poor response to tyrosine kinase inhibitors (TKIs) [Oser et al. 2015; Sequist et al. 2011;
Yu et al. 2013]. Additionally, prostate cancers also acquire resistance to anti-androgen
therapies by relapsing as a discrete malignancy with altered expression of
neuroendocrine markers [Ku et al. 2017; Mu et al. 2017]. Recently, B-cell leukemias
have also been shown to undergo a lineage switch to evade CD19 CAR-T cell therapy.
CD19+ B-ALLs relapse as a myeloid-like leukemia that now expresses the myeloid
marker CD33 [Gardner et al. 2016; Jacoby et al. 2016]. Furthermore, work from the
Winslow Lab demonstrates that modulation of the chromatin landscape is implicated in
metastasis. Specifically, global increases in chromatin accessibility mediated by Nfib
promotes metastasis in small cell lung cancer [Denny et al. 2016]. Changes in the
chromatin landscape during metastasis may not be hugely surprising, since the process
of metastasis (EMT) is largely defined as a process of cellular reprogramming, with
171
reprogramming often requiring significant changes in genome accessibility. However,
cancer cells redefining their chromatin landscape through epigenetic mechanisms in
order to evade cytotoxic therapy represents a previously uncharacterized process. Like
metastatic cells, Phf6KO B-ALL cells also have a global increase in genome accessibility
compared to Phf6WT cells (Figure 3.7B). These observations suggest that lineage
plasticity, described as gaining phenotypic characteristics of an alternate lineage that is
no longer dependent on the drug target, is emerging as a major mechanism of
resistance [Mu et al. 2017]. As shown by our work and that of the Winslow Lab,
promotion of an open chromatin architecture allowing for the atypical binding of
transcription factors supports a model of reprogramming that drives metastasis, lineage
infidelity, or loss of sensitivity to targeted therapy.
In order to test whether drastic changes to the chromatin landscape drive
resistance to targeted therapies through lineage switching, patient samples need to be
collected before and after treatment with the targeted inhibitor. For simplicity, treatment
of B-cell leukemias with CD19-targeted CAR-T cell therapy will be described here and
depicted in Figure 3.9. If enough sample can be collected, leukemia cells should be split
between 3 main integrated genomic analyses: RNA-Seq, ATAC-Seq, and DNA-
sequencing. Data from ATAC-Seq will be able to distinguish if any global changes to the
chromatin architecture occur after therapy treatment, and if DNA accessibility changes
are enriched at transcription factor binding sites. Data from RNA-Seq (single-cell or bulk
cell population) may uncover underlying activation or repression of developmental
programs that contribute to resistance. Since the relapse of CAR-T treated B-ALLs is
rapid (less than 1 month), the mechanism of resistance is thought to be epigenetic-
based rather than the outgrowth of a mutant clone [Gardner et al. 2016]. However,
sequencing of tumor DNA may still represent a valuable tool for analyses and identify
mutations in epigenetic regulators that underlie this process. I hypothesize that relapsed
samples will have a more permissive and open chromatin landscape, and due to the
acquisition of myeloid phenotypes, will be enriched for increased accessibility of myeloid
TF binding sites. This suggests that targeting the factor(s) that control or maintain the
chromatin landscape can undermine this mechanism of resistance and prove to be a
novel target for inhibition. Whether this be through targeting of global mediators, like
172
histone acetyltransferases (anacardic acid/MG149/C646), or individual factors, like Nfib
or specific myeloid TFs, understanding the contribution of the chromatin landscape
could help to suppress metastasis and resistance [Dekker et al. 2014; Soto-Feliciano et
al. 2017; Denny et al. 2016]. Since lineage switching is observed in relapsed non-small
cell lung cancers and prostate cancers, patient samples can also be collected before
and after treatment with anti-EGFR or anti-androgen therapy, respectively. Only 5,000
cells are needed for ATAC-Seq in order to get reliable, accurate results, making this
/
RNA-Seq
Gene X
Gene Y
0-)
FACS sort leukemia cells
ATAc-Seq
11tCACkAC
*u..u*u.
DNA-Seq
I
How do the transcriptomes of Are there global Changes in chromatin Does treatment select
relapsed tumors compare to changes in chromatin accessibility enriched for a pre-existing
the initial disease? accessibility upon for.... subclone?
Upregulation of other relapse? (opening) Myeloid TFs? Mutation in an
developmental programs? (closing) B-cell TFs? epigenetic regulator?
Figure 3.9. Investigating the mechanism of lineage switching in relapsed human B-ALLs after CD1 9 CAR-T therapy.
173
TI I1
Human B-ALL samples
Pro-Treatment
Human B-ALL samples
Lineage-Switched Relapse
approach and validation in multiple cancer types feasible and can confirm if this
mechanism is consistent across different cancer types [Corces et al. 2016].
>The Contribution of the Driving Oncogene to the Degree of Cellular Plasticity
Unlike prostate and lung cancers treated with targeted therapies, CAR-T cell
therapies target a cell type (as opposed to a driving oncogene). This makes CD1 9 CAR-
T therapy applicable to treat many types of B-cell malignancies driven by a variety of
oncogenes. However, it is unclear whether the oncogene driving a given B-cell
malignancy has any contribution to resistance via transdifferentiation. Clinically, it has
been shown that MLL rearranged ALLs undergo a switch to AMLs and cases of CLL
relapse as plasmablastic lymphomas after exposure to CAR-T therapy [Gardner et al.
2016; Evans et al. 2015]. Additionally, human BCR-ABLI cell lines have the capacity to
switch to myeloid-like cells [McClellan et al. 2015]. Further, a murine B-ALL model
driven by the E2a:PBX translocation has also been shown to undergo a switch to
myeloid leukemia after CAR-T treatment [Jacoby et al. 2016]. These examples highlight
the plasticity of B-cell malignancies and show that leukemias driven by many different
oncogenes are able to undergo lineage switches. However, it is not known what
molecular subtypes of B-ALL are prone to lineage switching in order to evade CD19
CAR-T therapy. Identification and stratification of patients into groups that will respond
or relapse after treatment is clinically important and warrants further investigation. To do
this, many models of murine B-ALL driven by different oncogenes (ETV6-RUNXI, E2a-
PBX1, E2a-HLF BCR-ABL, MLL-AF4, MLL-AF9, STAT5, Eu-RET) can be tested for
their ability to switch lineages after treatment with CD19-directed CAR-T cells, as
described in Figure 3.10 [Bernardin et al. 2002; BijI et al. 2005; Smith et al. 2002;
Williams et al. 2006; Krivtsov et al. 2008; Thiel et al. 2010; Heltemes-Harris et al. 2011;
Zeng et al. 1998]. It has previously been shown that the rate of lineage switching in
leukemia-bearing mice is substantial (-80% of mice) following treatment with murine
CD19 CAR-T cells [Jacoby et al. 2016; Kochenderfer et al. 2010]. Using this approach,
different murine B-ALL models can be transplanted into recipient hosts, conditioned with
radiation, treated with CD19-CAR-T cells and then analyzed for relapse by lineage
switching. Variations of this protocol can include modulation of the leukemia cells ex
174
vivo, or during relapse, via administration of therapies/inhibitors after adoptive transfer
of CAR-T cells.
In addition to identifying molecular subtypes of B-ALL that are more susceptible
to relapse by this novel mechanism, the experimental design described above allows for
further investigation into the mechanism of resistance. For example, patients that
relapse often experience a cytokine release syndrome (CRS), with elevated levels of
IL-6 [Perna et al. 2016; Gardner et al. 2016]. Therefore, the influence of IL-6 on the
relapse process can be determined by inhibiting IL-6 and IL-6 signaling throughout the
treatment course by administration of IL-6R antibodies (15A7, BioXCell), or AG490
(Jak2/3 inhibitor). Additionally, use of C57BL16 1L6-/- mice can be used to test the effects
of full IL-6 ablation on the relapse process [Bent et al. 2016; Gilbert et al. 2010].
The work set forth by Zhu, Mu, Ku and colleagues suggests that anti-androgen
resistance via lineage plasticity is mediated by loss of PTEN, RBI, and TP53, which
results in derepression of the epigenetic factor EZH2 and reprogramming factor SOX2.
Use of prostate cancer genomic data sets, in vivo mouse models, and human prostate
cancer cell lines support this model [Zou et al. 2017; Mu et al. 2017; Ku et al. 2017].
Further, loss of RB1 is found in 100% of relapsed lung cancer samples [Niederst et al.
2015]. After the identification of B-ALLs that undergo lineage switching, modulation of
Pten, p53, Rbl, Sox2, and Ezh2 expression in cell lines before transplantation can
determine the extent of any similarity between B-ALL versus prostate and lung cancer
resistance mechanisms. Additionally, as discussed at length in this thesis, the chromatin
landscape and epigenetic changes likely play important roles in the evolution of drug
resistance. Therefore, targeting factors with broad roles in epigenetic regulation (such
as HATs and HDACs) is an intriguing experiment that could potentially uncover novel
combination therapies for a variety of cancers, and is easily integrated into the
experimental design set forth in Figure 3.10.
175
A
Generation of murine CD19
CAR-T cells
Inject with leukemia Condition
with radiation
(500 cGy)
Adoptive transfer of
CAR-T cells
Day: 0
4'
4
Experimental Intervention
5 -100-300 days
Experimental Intervention
Molecular
Pten KD
p53 KD
Rbl KD
Sox2 O/E
Ezh2 O/E
1L6-/- mice
Treatment deplete 1L6HAT/HDAC inhibitors
ETV6-RUNXI C57BL/6 Cdkn2a-/- Bemardin et al. 2002(TEL-A MLl)___________ _____
E2a-PBX1 C57BL/6 x CH3 CD3-/- Biji et al. 2005
E2a-HLF BALB/c Bcl-2 Smith et al. 2002
BCR-ABL C57BL/6 Arf-/- Williams et al. 2006
MLL-AF4 BL6/129 - Krstov et al. 2008
MLL-AF9 C57BL6 x C57B6-SJL - Thiel et al. 2009
Stat5 C57BL/6 Ebfl+/-, Pax5+/- Heltemes-Harris et al.
Eu-RET BALB/c I Zeng et al. 1998
Figure 3.10. (A) Schematic depicting the experimental process to study lineage conversion after CD19 CAR-T cell therapy.
Generation of murine CAR-T cells is described by Kochenderfer et al. 2010. Modification of cells before treatment can help
elucidate the role of certain genes in contributing to lineage switching. Therapeutic intervention after adoptive cell transfer may
uncover novel combined therapeutic strategies to suppress lineage switched relapses. (B) Table listing murine B-cell leukemias
driven by various oncogenes that can be used in the experiment. [Jacoby et al. 2014; Jacoby et al. 2016]
176
Relapse
B
73
CONCLUDING REMARKS
In conclusion, we have identified the function of PHF6 in B-cell acute
lymphoblastic leukemia, describing an essential role in organizing the chromatin
architecture and modulating accessibility to lineage-specific transcription factors. We
characterized the wide-spread changes that occur within B-ALL cells upon loss of Phf6,
including large downregulation of B-cell developmental programs, upregulation of T-cell
programs, acquisition of T-cell lymphoma tumor characteristics, and focal modification of
the accessible chromatin landscape. Importantly, we described how these changes are
due to enhanced developmental plasticity, and demonstrated how this plasticity can be
co-opted by leukemia cells for their benefit. This includes engaging and tolerating
aberrant lineage programs in order to survive. Further, we examined how our findings
with Phf6 in B-ALL cells translate to the global importance of the chromatin landscape in
many different types of cancer, and how dysregulation of chromatin accessibility may be
the underlying driver of resistance to targeted therapy in B-ALLs, lung cancers and
prostate cancers, as well as represent a global mechanism of resistance for any cancer
treated with a targeted therapy. The field of chromatin biology and its contribution to
cancer initiation, development, metastasis, and therapy response is rapidly expanding.
Here we identified and defined a role for PHF6 within this framework, and demonstrate
an exciting future for continued study of this protein in lymphocyte development and the
broader investigation of chromatin landscape maintenance/dysregulation in cancer.
177
REFERENCES
Baker LA, Allis CD, Wang GG. 2008. PHD fingers in human diseases: Disorders arising from misinterpreting
epigenetic marks. Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis 647: 3-12.
Bent EH, Gilbert LA, Hemann MT. 2016. A senescence secretory switch mediated by P13K/AKT/mTOR activation
controls chemoprotective endothelial secretory responses. Genes Dev 30: 1811-1821.
Bernardin F, Yang Y, Cleaves R, Zahurak M, Cheng L, Civin Cl, Friedman AD. 2002. TEL-AML1, expressed from
t(12;21) in human acute lymphocytic leukemia, induces acute leukemia in mice. Cancer Research 62: 3904-
3908.
BijI J, Sauvageau M, Thompson A, Sauvageau G. 2005. High incidence of proviral integrations in the Hoxa locus in a
new model of E2a-PBX1-induced B-cell leukemia. Genes Dev 19: 224-233.
Charan J, Kantharia ND. 2013. How to calculate sample size in animal studies? J Pharmacol Pharmacother4: 303-4.
Cheng H, Liang PH, Cheng T. 2013. Mouse Hematopoietic Stem Cell Transplantation. In Lymphoma (ed. R.
KOppers), Vol. 976 of Methods in Molecular Biology, pp. 25-35, Humana Press, Totowa, NJ.
Cobaleda C, Jochum W, Busslinger M. 2007. Conversion of mature B cells into T cells by dedifferentiation to
uncommitted progenitors. Nature 449: 473-477.
Corces MR, Buenrostro JD, Wu B, Greenside PG, Chan SM, Koenig JL, Snyder MP, Pritchard JK, Kundaje A,
Greenleaf WJ, et al. 2016. Lineage-specific and single-cell chromatin accessibility charts human hematopoiesis
and leukemia evolution. Nat Genet 48: 1193-1203.
de Boer E, Rodriguez P, Bonte E, Krijgsveld J, Katsantoni E, Heck A, Grosveld F, Strouboulis J. 2003. Efficient
biotinylation and single-step purification of tagged transcription factors in mammalian cells and transgenic mice.
Proc Natl Acad Sci USA 100: 7480-7485.
Dekker FJ, van den Bosch T, Martin NI. 2014. Small molecule inhibitors of histone acetyltransferases and
deacetylases are potential drugs for inflammatory diseases. Drug Discovery Today 19: 654-660.
Denny SK, Yang D, Chuang C-H, Brady JJ, Lim JS, GrOner BM, Chiou S-H, Schep AN, Baral J, Hamard C, et al.
2016. Nfib Promotes Metastasis through a Widespread Increase in Chromatin Accessibility. Cell 166: 1-56.
Dhayalan A, Tamas R, Bock I, Tattermusch A, Dimitrova E, Kudithipudi S, Ragozin S, Jeltsch A. 2011. The ATRX-ADD
domain binds to H3 tail peptides and reads the combined methylation state of K4 and K9. Human Molecular
Genetics 20: 2195-2203.
Evans AG, Rothberg PG, Burack WR, Huntington SF, Porter DL, Friedberg JW, Liesveld JL. 2015. Evolution to
plasmablastic lymphoma evades CD19-directed chimeric antigen receptor T cells. British Journal of
Haematology 171: 205-209.
Gardner R, Wu D, Cherian S, Fang M, Hanafi L-A, Finney 0, Smithers H, Jensen MC, Riddell SR, Maloney DG, et al.
2016. Acquisition of a CD19-negative myeloid phenotype allows immune escape of MLL-rearranged B-ALL from
CD19 CAR-T-cell therapy. 127: 2406-2410.
Gartner F, Alt FW, Monroe R, Chu M, Sleckman BP, Davidson L, Swat W. 1999. Immature thymocytes employ distinct
signaling pathways for allelic exclusion versus differentiation and expansion. Immunity 10: 537-546.
Geimer Le Lay A-S, Oravecz A, Mastio J, Jung C, Marchal P, Ebel C, Dembele D, Jost B, Le Gras S, Thibault C, et al.
2014. The tumor suppressor Ikaros shapes the repertoire of notch target genes in T cells. Science Signaling 7:
ra28-ra28.
Gilbert LA, Hemann MT. 2010. DNA Damage-Mediated Induction of a Chemoresistant Niche. Cell 143: 355-366.
Green NM. 1963. Avidin. 1. The use of (14-C) Biotin for Kinetic Studies and for Assay. Biochem J 89: 585-591.
Heltemes-Harris LM, Willette MJL, Ramsey LB, Qiu YH, Neeley ES, Zhang N, Thomas DA, Koeuth T, Baechler EC,
Kornblau SM, et al. 2011. Ebf1 or Pax5 haploinsufficiency synergizes with STAT5 activation to initiate acute
lymphoblastic leukemia. J Exp Med 208: 1135-1149.
Holmes R, Zuhiga-Pflucker JC. 2009. The OP9-DL1 System: Generation of T-Lymphocytes from Embryonic or
Hematopoietic Stem Cells In Vitro. Cold Spring Harbor Protocols 2009: pdb.prot5l56-pdb.prot5l56.
H6flinger S, Kesavan K, Fuxa M, Hutter C, Heavey B, Radtke F, Busslinger M. 2004. Analysis of Notch1 function by in
vitro T cell differentiation of Pax5 mutant lymphoid progenitors. J Immunol 173: 3935-3944.
lwafuchi-Doi M, Zaret KS. 2014. Pioneer transcription factors in cell reprogramming. Genes Dev 28: 2679-2692.
Iwase S, Bin Xiang, Ghosh S, Ren T, Lewis PW, Cochrane JC, Allis CD, Picketts DJ, Patel DJ, Li H, et al. 2011. ATRX
ADD domain links an atypical histone methylation recognition mechanism to human mental-retardation
syndrome. 18: 769-776.
Jacoby E, Chien CD, Fry TJ. 2014. Murine models of acute leukemia: important tools in current pediatric leukemia
research. Front Oncol 4: 95.
Jacoby E, Nguyen SM, Fountaine TJ, Welp K, Gryder B, Qin H, Yang Y, Chien CD, Seif AE, Lei H, et al. 2016. CD19
CAR immune pressure induces B-precursor acute lymphoblastic leukaemia lineage switch exposing inherent
leukaemic plasticity. Nature Communications 7: 1-10.
Kaplan N, Dekker J. 2013. High-throughput genome scaffolding from in vivo DNA interaction frequency. 31: 1143-
1147.
Kiel MJ, Yilmaz OH, Iwashita T, Yilmaz OH, Terhorst C, Morrison SJ. 2005. SLAM Family Receptors Distinguish
Hematopoietic Stem and Progenitor Cells and Reveal Endothelial Niches for Stem Cells. Cell 121: 1109-1121.
Kochenderfer JN, Yu Z, Frasheri D, Restifo NP, Rosenberg SA. 2010. Adoptive transfer of syngeneic T cells
transduced with a chimeric antigen receptor that recognizes murine CD19 can eradicate lymphoma and normal
B cells. Blood 116: 3875-3886.
178
Krivtsov AV, Feng Z, Lemieux ME, Faber J, Vempati S, Sinha AU, Xia X, Jesneck J, Bracken AP, Silverman LB, et al.
2008. H3K79 Methylation Profiles Define Murine and Human MLL-AF4 Leukemias. Cancer Cell 14: 355-368.
Ku S-Y, Rosario S, Wang Y, Mu P, Seshadri M, Goodrich ZW, Goodrich MM, Labbe DP, Gomez EC, Wang J, et al.
2017. Rbl and Trp53 cooperate to suppress prostate cancer lineage plasticity, metastasis, and antiandrogen
resistance. Science 355: 78-83.
Li D, Qiu Z, Shao Y, Chen Y, Guan Y, Liu M, Li Y, Gao N, Wang L, Lu X, et al. 2013. Heritable gene targeting in the
mouse and rat using a CRISPR-Cas system. Nat Biotechnol 31: 681-683.
Liu L, Qin S, Zhang J, Ji P, Shi Y, Wu J. 2012. Solution structure of an atypical PHD finger in BRPF2 and its
interaction with DNA. Journal of Structural Biology 180: 165-173.
Liu Z, Li F, Ruan K, Zhang J, Mei Y, Wu J, Shi Y. 2014. Structural and functional insights into the human B6rjeson-
Forssman-Lehmann syndrome-associated protein PHF6. The Journal of Biological Chemistry 289: 10069-
10083.
McClellan JS, Dove C, Gentles AJ, Ryan CE, Majeti R. 2015. Reprogramming of primary human Philadelphia
chromosome-positive B cell acute lymphoblastic leukemia cells into nonleukemic macrophages. Proc Nat Acad
Sci USA 112: 4074-4079.
McManus S, Ebert A, Salvagiotto G, Medvedovic J, Sun Q, Tamir I, Jaritz M, Tagoh H, Busslinger M. 2011. The
transcription factor Pax5 regulates its target genes by recruiting chromatin-modifying proteins in committed B
cells. The EMBO Journal 30: 2388-2404.
Meacham CE, Lawton LN, Soto-Feliciano YM, Pritchard JR, Joughin BA, Ehrenberger T, Fenouille N, Zuber J,
Williams RT, Young RA, et al. 2015. A genome-scale in vivo loss-of-function screen identifies Phf6 as a lineage-
specific regulator of leukemia cell growth. Genes Dev 29: 483-488.
Mikkola 1, Heavey B, Horcher M, Busslinger M. 2002. Reversion of B cell commitment upon loss of Pax5 expression.
Science 297: 110-113.
Mu P, Zhang Z, Benelli M, Karthaus WR, Hoover E, Chen C-C, Wongvipat J, Ku S-Y, Gao D, Cao Z, et al. 2017.
SOX2 promotes lineage plasticity and antiandrogen resistance in TP53- and RB1-deficient prostate cancer.
Science 355: 84-88.
Musselman CA, Lalonde M-E, C6te J, Kutateladze TG. 2012. Perceiving the epigenetic landscape through histone
readers. Nat Struct Mol Biol 19: 1218-1227.
Nady N, Min J, Kareta MS, Chedin F, Arrowsmith CH. 2008. A SPOT on the chromatin landscape? Histone peptide
arrays as a tool for epigenetic research. Trends in Biochemical Sciences 33: 305-313.
Nechanitzky R, Akbas D, Scherer S, Gy6ry 1, Hoyler T, Ramamoorthy S, Diefenbach A, Grosschedl R. 2013.
Transcription factor EBF1 is essential for the maintenance of B cell identity and prevention of alternative fates in
committed cells. Nature Immunology 14: 867-875.
Niederst MJ, Sequist LV, Poirier JT, Mermel CH, Lockerman EL, Garcia AR, Katayama R, Costa C, Ross KN, Moran
T, et al. 2015. RB loss in resistant EGFR mutant lung adenocarcinomas that transform to small-cell lung cancer.
Nature Communications 6: 1-10.
Nutt SL, Heavey B, Rolink AG, Busslinger M. 1999. Commitment to the B-lymphoid lineage depends on the
transcription factor Pax5. Nature 401: 556-562.
Oser MG, Niederst MJ, Sequist LV, Engelman JA. 2015. Transformation from non-small-cell lung cancer to small-cell
lung cancer: molecular drivers and cells of origin. Lancet Oncol 16: e165-72.
Perna F, Sadelain M. 2016. Myeloid leukemia switch as immune escape from CD19 chimeric antigen receptor (CAR)
therapy. Transl Cancer Res 5: S221-S225.
Platt RJ, Chen S, Zhou Y, Yim MJ, Swiech L, Kempton HR, Dahlman JE, Parnas 0, Eisenhaure TM, Jovanovic M, et
al. 2014. CRISPR-Cas9 Knockin Mice for Genome Editing and Cancer Modeling. Cell 159: 440-455.
Pongubala JMR, Northrup DL, Lancki DW, Medina KL, Treiber T, Bertolino E, Thomas M, Grosschedl R, Allman D,
Singh H. 2008. Transcription factor EBF restricts alternative lineage options and promotes B cell fate
commitment independently of Pax5. Nature Immunology 9: 203-215.
Pridans C, Holmes ML, Polli M, Wettenhall JM, Dakic A, Corcoran LM, Smyth GK, Nutt SL. 2008. Identification of
Pax5 Target Genes in Early B Cell Differentiation. J Immunol 180: 1719-1728.
Reynaud D, A Demarco I, L Reddy K, Schjerven H, Bertolino E, Chen Z, Smale ST, Winandy S, Singh H. 2008.
Regulation of B cell fate commitment and immunoglobulin heavy-chain gene rearrangements by Ikaros. Nature
Immunology 9: 927-936.
Rodriguez P, Bonte E, Krijgsveld J, Kolodziej KE, Guyot B, Heck AJR, Vyas P, de Boer E, Grosveld F, Strouboulis J.
2005. GATA-1 forms distinct activating and repressive complexes in erythroid cells. The EMBO Journal 24:
2354-2366.
Rolink AG, Nutt SL, Melchers F, Busslinger M. 1999. Long-term in vivo reconstitution of T-cell development by Pax5-
deficient B-cell progenitors. Nature 401: 603-606.
Rothenberg EV. 2014. Transcriptional Control of Early T and B Cell Developmental Choices. Annu Rev Immunol 32:
283-321.
Roux KJ, Kim DI, Raida M, Burke B. 2012. A promiscuous biotin ligase fusion protein identifies proximal and
interacting proteins in mammalian cells. J Cell Biol 196: 801-810.
Schebesta A, McManus S, Salvagiotto G, Delogu A, Busslinger GA, Busslinger M. 2007. Transcription Factor Pax5
Activates the Chromatin of Key Genes Involved in B Cell Signaling, Adhesion, Migration, and Immune Function.
Immunity 27: 49-63.
Schmitt TM, Z6higa-Pflucker JC. 2002. Induction of T cell development from hematopoietic progenitor cells by delta-
like-1 in vitro. Immunity 17: 749-756.
Sequist LV, Waltman BA, Dias-Santagata D, Digumarthy S, Turke AB, Fidias P, Bergethon K, Shaw AT, Gettinger S,
Cosper AK, et al. 2011. Genotypic and histological evolution of lung cancers acquiring resistance to EGFR
inhibitors. Sci Transl Med 3: 75ra26-75ra26.
Smith KS, Rhee JW, Cleary ML. 2002. Transformation of bone marrow B-cell progenitors by E2a-Hlf requires
coexpression of Bcl-2. Molecular and Cellular Biology 22: 7678-7687.
179
Somasundaram R, Ahsberg J, Okuyama K, Ungerback J, Lilljebj6rn H, Fioretos T, Strid T, Sigvardsson M. 2016.
Clonal conversion of B lymphoid leukemia reveals cross-lineage transfer of malignant states. Genes Dev 30:
2486-2499.
Soto-Feliciano YM, Bartlebaugh JME, Liu Y, SAnchez-Rivera FJ, Bhutkar A, Weintraub AS, Buenrostro JD, Cheng
CS, Regev A, Jacks TE, et al. 2017. PHF6 regulates phenotypic plasticity through chromatin organization within
lineage-specific genes. Genes Dev 1-18.
Soufi A, Garcia MF, Jaroszewicz A, Osman N, Pellegrini M, Zaret KS. 2015. Pioneer Transcription Factors Target
Partial DNA Motifs on Nucleosomes to Initiate Reprogramming. Cell 161: 555-568.
Thiel AT, Blessington P, Zou T, Feather D, Wu X, Yan J, Zhang H, Liu Z, Ernst P, Koretzky GA, et al. 2010. MLL-AF9-
Induced Leukemogenesis Requires Coexpression of the Wild-Type Mll Allele. Cancer Cell 17: 148-159.
Todd MAM, Picketts DJ. 2012. PHF6 Interacts with the Nucleosome Remodeling and Deacetylation (NuRD) Complex.
J Proteome Res 11: 4326-4337.
Treiber T, Mandel EM, Pott S, Gyory I, Firner S, Liu ET, Grosschedl R. 2010. Early B Cell Factor 1 Regulates B Cell
Gene Networks by Activation, Repression, and Transcription- Independent Poising of Chromatin. Immunity 32:
714-725.
Ungerback J, Ahsberg J, Strid T, Somasundaram R, Sigvardsson M. 2015. Combined heterozygous loss of Ebf1 and
Pax5 allows for T-lineage conversion of B cell progenitors. J Exp Med 212: 1109-1123.
Van Vlierberghe P, Palomero T, Khiabanian H, Van der Meulen J, Castillo M, Van Roy N, De Moerloose B, Philippe J,
Gonz;lez-Garcia S, Toribio ML, et al. 2010. PHF6 mutations in T-cell acute lymphoblastic leukemia. 42: 338-
342.
Wang J, Leung JWC, Gong Z, Feng L, Shi X, Chen J. 2013. PHF6 Regulates Cell Cycle Progression by Suppressing
Ribosomal RNA Synthesis. Journal of Biological Chemistry 288: 3174-3183.
Williams RT, Roussel, Sherr CJ. 2006. Arf gene loss enhances oncogenicity and limits imatinib response in mouse
models of Bcr-Abl-induced acute lymphoblastic leukemia. Proc Nat/ Acad Sci USA 1-6.
Xie H, Ye M, Feng R, Graf T. 2004. Stepwise reprogramming of B cells into macrophages. Cell 117: 663-676.
Yang H, Wang H, Jaenisch R. 2014. Generating genetically modified mice using CRISPR/Cas-mediated genome
engineering. Nat Protoc 9: 1956-1968.
Yu HA, Arcila ME, Rekhtman N, Sima CS, Zakowski MF, Pao W, Kris MG, Miller VA, Ladanyi M, Riely GJ. 2013.
Analysis of Tumor Specimens at the Time of Acquired Resistance to EGFR-TKI Therapy in 155 Patients with
EGFR-Mutant Lung Cancers. Clinical Cancer Research 19: 2240-2247.
Zeng XX, Zhang H, Hardy RR, Wasserman R. 1998. The fetal origin of B-precursor leukemia in the E-mu-ret mouse.
Blood 92: 3529-3536.
Zhang C, Mejia LA, Huang J, Valnegri P, Bennett EJ, Anckar J, Jahani-Asl A, Gallardo G, Ikeuchi Y, Yamada T, et al.
2013. The X-Linked Intellectual Disability Protein PHF6 Associates with the PAF1 Complex and Regulates
Neuronal Migration in the Mammalian Brain. Neuron 78: 986-993.
Zou M, Toivanen R, Mitrofanova A, Floch N, Hayati S, Sun Y, Le Magnen C, Chester D, Mostaghel EA, Califano A, et
al. 2017. Transdifferentiation as a Mechanism of Treatment Resistance in a Mouse Model of Castration-
Resistant Prostate Cancer. Cancer Discovery 7: 736-749.
180