University of Southern California Dissertations and Theses

Epigenetic dysregulation in acute myeloid leukemia (AML) with MLL1 aberrations

(USC Thesis Other)

Epigenetic dysregulation in acute myeloid leukemia (AML) with MLL1 aberrations

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Epigenetic Dysregulation in Acute Myeloid Leukemia (AML) with MLL1 Aberrations
by
Yongxiang Wang
A Thesis Presented to the
FACULTY OF THE USC KECK SCHOOL OF MEDICINE
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(BIOCHEMISTRY AND MOLECULAR MEDICINE)
August 2024
Copyright 2024 Yongxiang Wang

ii
Acknowledgments
I sincerely appreciate my mentor, Dr. Yali Dou, for her help and guidance during the past two
years. I could not finish this thesis without her help. It is my pleasure to become a member of her
lab during my Master’s work. I also want to give my thanks to the Dou lab members for their
kindness in sharing the information and teaching me new techniques.
I would like to give my appreciation to my committee members, Dr. Yifan Liu and Dr. Judd
Rice, for their valuable time and suggestions for my project and thesis.
Over the past two years, many significant events have occurred, leaving a lasting impact. These
experiences have become an unforgettable part of my life journey. I hope to express my sincere
gratitude to my friends, Ethan Pan, Tang Tu, Xueqing Wang, Ruichen Wang, and Zihe Shi.
Thank you for your company and encouragement during this special time.
Finally, I would like to thank my parents for their support. It is impossible for me to finish my
thesis work without them.
The more it costs, the better value it should be, without doubt
------Jim Hacker

iii
Table of Contents
Acknowledgments........................................................................................................................... ii
List of Figures................................................................................................................................ iv
List of Abbreviations ...................................................................................................................... v
Abstract.......................................................................................................................................... vi
Chapter 1: Introduction................................................................................................................... 1
1.1 MLL1 is a crucial histone modification enzyme .............................................................. 1
1.2 MLL1-r is an oncodriver for leukemogenesis .................................................................. 2
1.3 MLL1-r affect MLL1 H3K4me3 histone modification function...................................... 5
1.4 The role of MLL1 in cancer through regulation of gene expression ................................ 5
Chapter 2: Methods......................................................................................................................... 8
2.1 ChIP-seq data analysis...................................................................................................... 8
2.2 Cut & Run data analysis ................................................................................................... 8
2.3 Identify MLL1 gene targets.............................................................................................. 9
2.4 Differential Binding Analysis........................................................................................... 9
2.5 Peak file visualization....................................................................................................... 9
2.6 Peak overlap and extension............................................................................................. 10
2.7 Pathway enrichment analysis.......................................................................................... 11
2.8 Protein network analysis................................................................................................. 11
Chapter 3: Histone H3K4 methylation by MLL1 in Three Different Leukemia Cell Lines ........ 12
3.1 MLL1 and H3K4me3 in MLL1-r leukemia.................................................................... 12
3.2 MLL1 and H3K4me3 in MLL1 wild type leukemia ...................................................... 16
3.3 MLL1 and H3K4me3 in MLL1-amplified leukemia...................................................... 18
3.4 Comparing MLL1 and H3K4me3 in three leukemia cell lines with different MLL1
states...................................................................................................................................... 20
Chapter 4: MLL1-Dependent Regulatory Network in Leukemia................................................. 25
4.1 Gene list acquired by peak files...................................................................................... 25
4.2 Common MLL1 targets in three leukemia cell lines ...................................................... 27
4.3 Network analyses in three leukemia cell lines................................................................ 29
Chapter 5: Function of MLL1 in Gene Expression in MLL1 Amplified Leukemia .................... 31
MLL1 shows different binding and methylation activity in three leukemia cell lines. These
enrichment regions are associated with promoters and enhancers of many genes, potentially
regulating their expression levels. In particular, we are interested in demonstrating that
MLL1 binding and HMT activity mediate gene expression and drive the transcription
program in MLL1 amplified leukemia. ................................................................................ 31
5.1 RNA-seq analysis for UOC-M1 cells with or without MLL1 knockout ........................ 31
5.2 Expression changes of MLL1 target genes after MLL1 knockout ................................. 35
5.3 MLL1 targets non-coding RNAs.................................................................................... 39
Chapter 6: Discussion and Future Directions ............................................................................... 40
References..................................................................................................................................... 42

iv
List of Figures
Figure1: The structure and activity of the MLL1 complex................................................................4
Figure2: MLL aberrations in cancer................................................................................................7
Figure3...........................................................................................................................................13
Figure4...........................................................................................................................................15
Figure5...........................................................................................................................................17
Figure6...........................................................................................................................................19
Figure7...........................................................................................................................................22
Figure8...........................................................................................................................................24
Figure9...........................................................................................................................................26
Figure10.........................................................................................................................................28
Figure11: Protein-protein interaction analysis...............................................................................30
Figure12: The KEGG pathway analysis........................................................................................34
Figure13.........................................................................................................................................37

v
List of Abbreviations
STRING Search Tool for the Retrieval of Interacting Genes/Proteins
MLL1 Mixed-Lineage Leukemia 1
H3K4me3 histone H3 lysine 4 trimethylation
H3K4me2 histone H3 lysine 4 dimethylation
MACS2 Model-based Analysis of ChIP-Seq 2 generation
KEGG Kyoto Encyclopedia of Genes and Genomes
ChIP-seq Chromatin Immunoprecipitation Sequencing
RNA-seq RNA sequencing
CSC Cancer Stem cell
SET Suppressor of variegation, Enhancer of Zeste, and Trithorax
GOF Gain of Function
HSCs Hematopoietic stem cells

vi
Abstract
Mixed lineage leukemia 1 (MLL1, also known as MLL or KMT2A) is a crucial transcription
cofactor required for early embryonic development. MLL1 also plays an essential role in the
development and maintenance of hematopoietic stem cells (HSCs). It is frequently mutated in
cancer, including acute leukemia. Our research here compared MLL1 direct target genes in
leukemia with different MLL1 aberrations. We revealed the difference in MLL1 binding in three
leukemia cell lines with MLL1 rearrangement, MLL1 amplification, and wild type MLL1,
respectively. By analyzing histone 3 lysine 4 trimethylation and gene expression in these cell
lines, we found that oncogene Myc and its interaction network were the main MLL1 targets. We
further showed that MLL1 regulated additional genes highly related to cancer development and
leukemogenesis. We found that MLL1 amplification regulated not only Myc, Smad3 and Wnt5b,
but also other oncogenes, such as SOX21-AS1, and a non-coding micro-RNA that have not been
previously reported. Our research highlights unique and common pathways directly regulated by
MLL1 in AML with different MLL1 aberrations.

1
Chapter 1: Introduction
1.1 MLL1 is a crucial histone modification enzyme
Genomic studies have shown the importance of epigenetic regulation in cancer development.
Epigenetic regulations include histone modifications, DNA methylation, non-coding RNA, and
ATP-dependent chromosome remodeling[1]. Histone modification enzymes covalently modify
histone tails, depositing methylation, acetylation, phosphorylation, ubiquitylation or
sumoylation. Histone modifications regulated a broad range of cellular processes in cells,
including transcription, DNA repair, DNA replication, and cell cycle control. Dysregulation of
histone modifications greatly contributes to cancer development and maintenance, which have
been extensively investigated in recent years [1]. A comprehensive understanding of histone
modifications, how they are regulated, and their downstream effects, will help to elucidate the
molecular mechanism of tumorigenesis and provide potential new therapeutic strategies.
As a histone modifying enzyme, Mixed lineage leukemia 1 (MLL1) has long been implicated in
cancer and other human developmental syndromes. MLL1 resides in a large, multi-subunit
complex [18]. MLL1, together with WD repeat domain 5 (WDR5), retinoblastoma binding
protein 5 (RBBP5), Ash2 (absent, small, or homeotic)-like (ASH2L), Dpy-30 homolog (DPY30),
multiple endocrine neoplasia I (MEN1), and other regulation factors, form a multi-subunit
complex that binds to nucleosome core particle and deposits H3K4 methylation (Figure 1-a).
MLL1 activity is regulated by multiple factors including both proteins and non-coding RNA[2].
MLL1 was first identified in childhood acute leukemia and therapy-related secondary leukemia
[3][4]. It is a histone H3 lysine 4 (H3K4) methyltransferase (figue1-b). In most cases, high
H4K4me3 at transcription start site correlates with gene activation. MLL1 is required for early

2
embryonic development and plays an essential role in hematopoiesis by maintaining the HSCs.
The MLL1 methyltransferase complex regulates expressions of many genes, including
transcription factors, such as Hoxa9, Eya1, Prdm16, and Mecom (encoding both MDS-Evi1 and
Evi1). These genes are involved in HSCs stemness maintenance and leukemogenesis [5][6][7].
Dysregulation of MLL1 leads to constitutive or over-expression of certain Hox genes (e.g.,
HoxA9, Eya1), which can drive leukemogenesis [8][9] . During embryonic development, MLL1
is required for de nono hematopoiesis and self-renewal of the stem cells in fetal liver and
spleens. However, the development of adult hematopoietic cells are only modestly affected by
MLL1 knockout. Therefore, MLL1 has a critical role in self-renewal of hematopoietic stem cells
(HSCs) and maintenance of their stemness during development [10].
1.2 MLL1-r is an oncodriver for leukemogenesis
Chromosome translocations of the MLL1 gene, located at chromosome 11q23, are reported in
approximately 75% of acute leukemia in infants and 5–10% cases in children and adults. The
patients can be clinically characterized as acute lymphocytic leukemia (ALL) or acute myeloid
leukemia (AML). Unlike other pediatric ALLs (with a 5-year survival of ~90%), MLL1-
rearranged (MLL1-r) ALL shows a poor prognosis with 5-year survival rates of only 34–39%
[12]. MLL1-r AML patients also have poor outcomes with 5-year survival of ~50% in patients
less than 45 years old and <35% in older patients [11][12]. MLL-r is also frequently reported in
secondary leukemia after conventional chemotherapy [13][14].
MLL1 translocations lead to generation of MLL1 fusion proteins that contain MLL1 N-terminal
domains and C-terminal domains from over 90 fusion partners[15]. The main fusion partners
include ENL (~13%) AF9 (~19%), AF4 (~36%)[19]. Many of the partners are nuclear proteins

3
with transcription regulatory functions, which are recruited to MLL1 target genes and increase
their levels of expression [15][19].

4
Figure 1
The structure and activity of the MLL1 complex. a. The structure of the MLL1 complex
MLL1 is in complex with WD repeat domain 5 (WDR5), retinoblastoma binding protein 5 (RBBP5),
ash2 (absent, small, or homeotic)-like (ASH2L), dpy-30 homolog (DPY30) [17]
b. In vitro histone methyltransferase assay for the MLL family enzymes using nucleosome core
particles (NCP) as the substrate: The MLL family numbers form complexes with multiple factors to
achieve histone methyltransferase (HMT) activities. The DPY30 is one of the factors involved in
regulating their HMT activities [18]

5
1.3 MLL1-r affect MLL1 H3K4me3 histone modification function
MLL1 is a 3696 amino acid multi-domain protein[19]. Starting at N-terminus, MLL1 has three
AT-hooks for DNA binding, followed by DNMTL (or CxxC). The CxxC region is cysteine rich.
It recognizes unmethylated CpG dinucleotides and physically interacts with the transcription
elongation PAF1 complex and regulates MLL1 activity. Following the CxxC domain, there are
four zinc-finger-like motifs (plant homeodomains (PHDs)) with function of recognizing
H3K4me3 and regulated MLL1-mediated transcription activity. The C-terminus has the catalytic
suppressor of variegation, enhancer of Zeste, and Trithorax (SET) domain. It is a homolog of
Drosophila trithorax, which catalyzes the mono-, di-, or tri-methylation of histone 3 at lysine 4
(H3K4) on the chromatin [19].
Most MLL1-FPs retain the majority of the MLL1 N-terminal domains. In contrast, MLL-FPs
lose its C-terminus including the SET domain and PHD domain. The CxxC domain and AT hook
are retained in the MLL1 fusion proteins after chromosomal rearrangement. The MLL1-FPs have
no methylation activity since they do not have the catalytic SET domain. However, MLL1-FPs
are found to co-localize with wild-type MLL1 in leukemia cells [20]. In addition, some of the
fusion protein partners are able to recruit the RNA polymerase II, enhancing gene expression.
The MLL1-FPs mediated overexpression of Hoxa9, Eya1, Evil plays a significant role in AML
pathogenesis[21][22][23].
1.4 The role of MLL1 in cancer through regulation of gene expression
The mechanism of MLL1-r in leukemogenesis has been intensively studied. However, the
function of MLL1 in other cancers is less well studied. Previous reports have shown that MLL1
is required in Wnt-driven intestinal tumorigenesis and cancer stemness [24]. Another report

6
mentioned the MLL1 is not essential for MLL1-AF9-rransformaed cells survival or cell growth
[16]. Furthermore, there was no evidence of differentiation of MLL-AF9- or MLL-AF6-
transformed MLL1-deficient cells [16]. Loss of MLL1 also accelerates MLL2-defencident cell
apoptosis and reduces cell proliferation [16].
MLL1 expression is increased in gain of function TP53 mutation tumor. The p53 GOF mutants
bind to and upregulated MLL1, leading to genome-wide changes in histone methylation, which
in turn promotes uncontrolled cell proliferation. Knocking down the MLL1 gene reduces cancer
cell proliferation, yielding results like those obtained by knocking down the GOF p53 variant
[25].
MLL1 is also involved in non-coding RNA mediated cellular processes. MLL1 is enriched at the
HOTAIR promoter under hypoxia in colon cancer, which in turn contributes to cancer
development[26]. However, MLL1 mediated non-coding RNA dysregulation in leukemia has not
been reported.
Based on previous reports, MLL1 plays different roles by promoting certain expression profiles
and contributing to cancer development and maintenance. Overexpression of MLL1 in GOF
TP53 is highly related with tumorigenesis. The universal and distinct roles of MLL1 in different
KMT2A aberrations remain elusive. In this project, we evaluated MLL1 binding and methylation
activity in three leukemia cell lines by MLL1 ChIP-seq and H3K4me3 cut & run experiments.
Then we conducted MLL1 knockout in UOC-M1 cell model and evaluated expression change.
We found several potential oncogenes directly regulated by MLL1, which may be critical for
leukemogenesis driven by MLL1 amplification.

7
Figure 2
MLL aberrations in cancer. The chromosome rearrangement results in MLL1
fusion protein. The MLL1-fusion proteins lead to abnormal transcription activity
and epigenetic changes in leukemogenesis. Other MLL1 aberrations are MLL1
partial tandem duplication or gene amplification. These MLL1 mutations lead to
differentiation block and increased proliferation.

8
Chapter 2: Methods
2.1 ChIP-seq data analysis
ChIP-seq data processing: After quality control, paired-end sequencing reads were aligned to the
reference genome (hg38) using Bowtie2 (version 2.5.2) with the following parameters: --end-toend, --very-sensitive, --no-mixed, --no-discordant, --phred33, -I 10, -X 700. Then the Sam files
were sorted and duplicated using the Picard java program. The Sam files were converted to Bam
files by samtools. Next, Bam files were converted to bed files by bedtools’ bamtobed function in
pair-end method. Peaks that were shorter than 150 bp or longer than 750 bp were removed. The
spike-in alignment and sequencing depth calculation used the same Bowtie2 parameters and the
reference genome was changed to Drosophila_BDGP6.46. The sequence depth was divided by
two. The final bedgraph was generated by bedtools with sequence depth as scale factor and hg38
size as genome size.
Peak calling: The bedgraph files generated peak files with SEACR(version1.3). The parameters
for SEACR were norm (normalized control file) and relaxed (use relaxed threshold). Peak calling
was conducted based on MLL1 ChIP-seq result and IgG control.
2.2 Cut & Run data analysis
After quality control, the single-end sequencing reads were aligned to the reference genome
(hg38) using Bowtie2 with the same process as described above. After alignment, the Sam file
was converted to Bam file by samtools and sorted by name. The final bed file was generated by
bedtools and peaks shorter than 150 bp or longer than 750 bp were removed. Peak calling: The
peak bed file was generated by MACS2 using q-value 0.01.

9
2.3 Identify MLL1 gene targets
The peaks target genes were identified by ChIPseeker package in R for binding target. The
ChIPseeker package was used to determine the peak region and annotated to the nearby genes.
The genome library was generated using TxDb.Hsapiens.UCSC.hg38.knownGene. The
distribution was visualized by gglot2.
2.4 Differential Binding Analysis
The differential binding analysis is processed with Diffbind package. For MLL1 and H3K4me3
peak files, the hold categories = DBA_CONDITION, minMembers = 2 was used; the default
mode was used to analyze the difference.
2.5 Peak file visualization
Heatmap: the bam file was converted to bigwig file. Together with bigwig file and peak calling
bed files, CumputeMatrix in lunix generates Matrix file. The calculation result was visualized as
heatmap generated by linux function plotHeatmap
Code: {bash}
for bamFile in "${bamFiles[@]}"; do
# Construct the bigWig file name
bwFile="$outputDir/$(basename "$bamFile" .bam)_raw.bw"
# Convert BAM to bigWig
bamCoverage -b "$bamFile" -o "$bwFile" --numberOfProcessors $cores
done
for bwFile in "$outputDir"/*_raw.bw; do
baseName=$(basename "$bwFile" _raw.bw)

10
matrixFile="$outputDir/${baseName}_matrix.gz"
heatmapFile="$outputDir/${baseName}_heatmap.png"
computeMatrix reference-point -S "$bwFile" \
-R "$bedFile" \
--skipZeros -o "$matrixFile" \
-a 3000 -b 3000 --referencePoint center -p $cores
plotHeatmap -m "$matrixFile" -out "$heatmapFile" \
--sortUsing sum --startLabel "Peak Start" --endLabel "Peak End" \
--xAxisLabel "" --regionsLabel "Peaks" --samplesLabel "$baseName" \
--colorMap Greens
done
Genome level view: The MLL1 peak files and H3K4me3 peak files were entered into IGV to
visualize, reference genome is Hg38.
2.6 Peak overlap and extension
Peak file overlap: Overlap peak files was generated by bedtools’ intersect function. The direct
overlap was directly using the original peak file to generate new peak file from the MLL1 ChIPseq data and H3K4me3 cut&run data. Code: bedtools intersect -a file1.bed -b file2.bed >
intersected.bed. In gap overlap method, new peak extended the start and end points by 500 bp
using bedtools’ slop function with genome size file Hg38. The overlap peak file was generated
by extended peak files with bedtools’ intersect function.

11
2.7 Pathway enrichment analysis
Pathway enrichment was analyzed by the enrichDAVID package or directly used the
David website. The result was visualized by gglot2. Code: {R}
data1 <- read.table("David-KEGG-enrichment.txt", header=TRUE, sep="\t")
combined_data <- rbind(data1)
combined_data$neg_log_pvalue <- -log10(combined_data$PValue)
ggplot(combined_data, aes(x=Source, y=Term, size=Count, color=neg_log_pvalue)) +
geom_point(alpha=0.6) + # Adjust point transparency
scale_color_gradient(low="blue", high="red") + theme_light()
+ theme(axis.text.y = element_text(angle=0, hjust=1)) + labs(x="", y="KEGG
Pathway", title="KEGG Pathway Analysis")
+ guides(size=guide_legend(title="Gene Count"), color=guide_colorbar(title="-
log10(P Value)"))
ggsave("KEGG_Pathway_Dotplot.png", width = 10, height = 10, dpi = 300)
2.8 Protein network analysis
Protein-protein interaction was analyzed by STRING. The protein list was uploaded to STRING
website for multiple protein interaction analysis. The STRING will remove disconnected
proteins. STRING indicates protein-protein interactions include direct (physical) and indirect
(functional) associations. In this research, we used MLL1 potential regulated gene list shared in
three leukemia to analyze potential interaction between these genes.

12
Chapter 3: Histone H3K4 methylation by MLL1 in Three Different
Leukemia Cell Lines
MLL1 is an important transcription cofactor playing crucial role in leukemia. However, whether
its methylation activity differentially regulate gene expression in different KMT2A aberrations is
unclear. We hypothesize that MLL1 affects different methylation regions in different leukemia
with KMT2A aberrations. To investigate the genomic binding and methylation activity of MLL1
in leukemia, we performed ChIP-seq of MLL1 to find the MLL1 binding sites and conducted cut
& run sequencing for H3K4me3 genome-wide distribution. By integrating the MLL1 and
H3K4me3 analysis, we have identified H3K4me3 regions directly deposited by MLL1 in
leukemia cells. To functionally validate the function of MLL1 in different leukemia cells, we
conducted RNA-seq analysis in leukemia cells that have MLL1-r, MLL1-amplificamtion, and
MLL1 wild type, respectively (Figure 3).
We tested three different methods to call peak for the MLL1 ChIP-seq result (MACS2, HOMER
and SEACR) and selected the SEACR as the final method of peak calling. The H3K4me3 cut &
run result was processed to peak file and peak calling was performed by MACS2.
3.1 MLL1 and H3K4me3 in MLL1-r leukemia
The MOLM-13 cell line was established from the peripheral blood of a 20-year-old man with
relapsed MLL1-AF9 acute myeloid leukemia (AML FAB M5a)[27]. The patient exhibited rapid
disease progression, extremely poor prognosis, and drug resistance. This cell line is widely used
in MLL1-r research. We used this cell line to investigate the MLL1 binding and its methylation
activity in MLL1-r leukemia.

13
Figure 3
a. Chromatin Profiling Workflow: the chromatin profiling workflow begins
with mapping sequencing reads to the genome and normalizing the data. The
read distribution is visualized, and peak calling is performed to identify regions
of enrichment. These peaks undergo motif analysis to identify potential
regulatory sequence motifs, followed by pathway analysis or to determine the
biological processes associated with the identified peaks and changes.
b. Gene Expression Analysis Workflow: cDNA sequencing reads are aligned
to the genome. Transcripts are assembled either using a reference-based method
or de novo. The abundance of each transcript is quantified and normalized.

14
We found 12,920 peaks of H3K4me3 and 9,103 peaks for MLL1. We overlapped MLL1 binding
regions and H3K4me3 regions and found 929 joint peak regions. Then we annotated the shared
peaks to nearby genes. We have found a total of 804 genes as MLL1 direct targets (Figure 4-a).
The heatmap was shown in Figure 4-b. About 15% of shared peaks are at the promoter region,
and 41.19% at distal intergenic region. For the MLL1-specific peaks, 13.18% are at the promoter
region, and 49.1% are at distal intergenic region (Figure 4-c).

15
Figure 4 a. Overlap of MLL1 and H3K4me3 peaks in MOLM-13: 12920 peaks of
H3K4me3 cut & run results and 9103 peaks for ChIP-seq results. 11991 peaks
regions are H3K4me3 specific, 8174 peaks regions are MLL1 specific, 929 peak
regions are shared. The shared peaks target 804 genes.
b. Heatmap of MLL1 and H3K4me3 peaks.
c. Distribution of MLL1&H3K4me3 shared peaks and MLL1 specific
peaks: For the shared peaks: 14.99% of them are in the promoter region, and
41.19% in the distal intergenic region. For the MLL1-specific peaks: 13.18% in
the promoter region, and 49.1% in the distal intergenic region

16
3.2 MLL1 and H3K4me3 in MLL1 wild type leukemia
The OCI-AML3 cell line was established from the peripheral blood of a 57-year-old man with
acute myeloid leukemia (AML FAB M4). OCI-AML3 has two copies of wild type MLL1. We
analyzed the MLL1 binding and H3K4me3 in OCI-AML3, like we did for MOLM-13 cells.
There are 25,181 peaks of H3K4me3 and 9,653 peaks for MLL1 in the OCI-AML3 cell line.
There are 4,149 joint peaks, which are assigned to 3,773 genes as MLL1 direct targets (Figure 5-
a). Peaks were visualized in Heatmap shown in Figure 5-b. Of the shared peaks, 17.23% of the
peaks are at the promoter region and 36.36% of the peaks are at the distal intergenic region. In
comparison, about 13.37% of the MLL1-specific binding sites are at the promoter region and
45.8% at the distal intergenic region (Figure 5-c).

17
Figure5 a. Overlap of MLL1 and H3K4me3 peaks in OCI-AML3: 25,181 peaks of
H3K4me3 cut & run results and 9,653 peaks for ChIP-seq results. 21032 peaks
regions are H3K4me3 specific, 5504 peaks regions are MLL1 specific, 4,149
peak regions are shared. The shared peaks target 3,773 genes.
b. Heatmap of shared peaks.
c. Distribution of MLL1&H3K4me3 shared peak and MLL1-specific peaks:
For shared peaks: 17.23% of them are in the promoter region, and 36.36% in the
distal intergenic region. For MLL1-specific peaks: 13.37% in the promoter
region, and 45.8% in the distal intergenic region

18
3.3 MLL1 and H3K4me3 in MLL1-amplified leukemia
The UOC-M1 cell line has four copies of WT MLL1. It was established from bone marrow of a
68-year-old man with undifferentiated acute myeloid leukemia of the megakaryoblastic lineage
(AML FAB M1). Gene amplification is common in old patients as a result of accumulation of
genetic alterations. The amplification of MLL1 is usually found in old patients [28]. MLL1
amplification sometimes is associated with both GOF and LOF TP53 mutations. GOF of TP53
mutations have been associated with increased expression of MLL1 in solid tumors, which have
no reported MLL1 amplification cases[25].
Gene amplification is a form of genome instability, commonly found in cancer cells [29]. The
function of MLL1 has not been previously studied in MLL1-amplified leukemia. We
investigated the function of MLL1 in UOC-M1 and compared the results to that of MLL1-r and
MLL1 WT leukemia.
There were 19,377 peaks of H3K4me3 and 8,336 peaks for MLL1. There were 4,211 shared
peaks for MLL1 and H3K4me3, corresponding to 3,944 genes (Figure 6-a). The heatmap was
used to visualize the shared peaks (Figure 6-b). Interesting, there are more MLL1 and H3K4me3
signals in UOC-M1 as compared to MOLM-13 and OCI-AML3, consistent with higher copies of
MLL1 gene in this cell line. In UOC-M1 cells, 16.19% of the shared peaks are at the promoter
regions, 37.35% of the shared peaks are at intergenic region. In comparison, 20.16% of MLL1-
specific peaks were at promoter regions and 40.22% of the peaks were at intergenic regions
(Figure 6-c).

19
Figure 6 a. Overlap of MLL1 and H3K4me3 peaks in UOC-M1: 19377 peaks of
H3K4me3 cut & run results and 8336 peaks for MLL1 ChIP-seq results. 15166
peaks regions are H3K4me3 specific, 4125 peaks regions are MLL1 specific,
4211 peak regions are shared. The shared peaks target 3944 genes.
b. Heatmap of shared peaks.
c. Distribution of MLL1&H3K4me3 shared peaks and MLL1-specific
peaks: For shared peaks: 16.19% of them were in promoter regions, and 37.35%
in distal intergenic regions. For MLL1-specific peaks: 20.16% were in promoter
regions, and 40.22% in distal intergenic regions.

20
3.4 Comparing MLL1 and H3K4me3 in three leukemia cell lines with different MLL1
states
After examining MLL1 and H3K4me3 in three different leukemia cell lines, we compared their
distribution patterns. We found that their distributions relative to annotated gene features, (e.g.
gene body, promoter, and enhancer)were similar. However, the shared peak numbers are
significantly lower in MOLM-13 with MLL1-AF9, reflecting lower levels of MLL1 in this cell
line.
To compare the H3k4me3 region on the genome browser view, we used the Diffbind package in
R to analyze the difference. The result shows difference among three different cell lines (figue7-
a). The Diffbind package removed the share region of all peaks and combined the original peak
file to better evaluate the difference among three leukemia cell lines. Consistent with
MLL1/H3K4me3 joint peaks, H3K4me3 peaks are significantly lower in MOLM-13 cells, as
shown in Heatmap, and drastically higher in UOC-M1 cells.
From the genome view, the MLL1 binding sites do not always have high H3K4me3 (Figure 7-
b). We further examined the binding of MLL1 at Myc locus. Myc is a well-known oncogene
involved in multiple human malignancies. Previous studies showed that c-Myc expression is
mediated by MLL1. It plays an important role in leukemia development [30]. The ChIP-seq
result showed that MLL1 binds to the Myc locus (Figure 8-b), near promoters. From pairwise
comparison, we found common H3K4me3 peaks between MOLM-13 and OCI-AML3 as well as
OCI-AML3 and UOC-M1, but not in MOLM-13 and UOC-M1 cells. Interestingly, some
H3K4me3 marked regions had MLL1 binding only in UOC-M1 cells (Figure 7-b). These regions
had H3K4me3 in other cells but no MLL1 binding, suggesting that other KMT2 enzymes are

21
probably responsible. Another observation was that MLL1 had broader peaks in UOC-M1 that
those in other cells.

22
Figure 7 a. Heatmap of H3K4me3 regions from the three different cell lines.
b. Genome view of H3K4me and MLL1 peaks. Blue rectangular box indicates
differential MLL1 binding and H3K4 trimethylation. The left figure is MLL1
binding around Myc gene. The red arrow indicates the region of H3K4me3. Blue
arrow indicates Myc gene.

23
Similar to H3K4me3, we also compared MLL1 binding in three different cell lines (Figure 8- a).
As expected, there were more binding regions of MLL1 in UOC-M1 as compared to other cell
lines, consistent with the high copy number of MLL1 in MLL1 amplified leukemia. We showed
that MLL1 had binding sites around Myc coding region in these three cell lines, but these were
not always shared with H3K4me3 regions.
In conclusion, we have characterized MLL1 and its cognate H3K4me3 in three leukemia cell
lines with MLL1-rearrangement, MLL1 amplification and wild type MLL1, respectively. The
results show that MLL1 binds to similar regions in the genome in all three cell lines, but it
probably has more significant role in MLL1-amplified leukemia. We have also identified regions
that are commonly targeted by MLL1 in two or more cell lines, suggesting that MLL1 maintains
some core leukemia transcriptional program regardless of the genetic alterations.

24
Figure 8 a. Heatmap of MLL1 binding sites in three leukemia cell lines.

25
Chapter 4: MLL1-Dependent Regulatory Network in Leukemia
4.1 Gene list acquired by peak files
In previous experiments, we identified the binding regions of MLL1 and H3K4me3 in three
leukemia cell lines. This allowed us to generate a list of potential direct targets of MLL1. Our
goal is to determine if there are any shared genes among the three different cell lines and to
explore potential interactions between these genes.
To achieve this, we utilized the MLL1 ChIP-seq results to generate peak files of the MLL1
binding region and the H3K4me3 CUT&RUN results to generate peak files of the histone H3
lysine 4 trimethylation region. To identify regions where histone methylation activity is
potentially mediated by MLL1, we employed both direct overlap and gap overlap methods to
determine the final overlap regions.
The direct overlap method directly uses the bed file to generate peak overlap. The gap overlap
method extends the original peak file for 1000bp. Then generate the overlap region with
extended peak files (Figure 9- a). The gap overlap method increased the number of peaks. For
MOLM-13, 929 peaks are increased to 1270 peaks targeting 875 genes. In OCI-AML3, 4149
peaks are increased to 5399 peaks target 4053 genes. In Uoc-M1, 4211 peaks increased to 5975
peaks targeting 4199 genes. Compared with the direct overlap method, the distribution of peaks
relative to gene annotation had not significantly changed

26
Figure 9 a. Different methods for peak overlap generation: Direct overlap: use the
original peak files to find regions shared by both peak files.
Gap overlap: add 1kb to the original peak file. Extend the start and end positions
of each peak in the peak file by 500 bp, and generate a new overlap bed file for
the shared region.
b. The distribution of gap overlap peaks:. For MOLM-13, 929 peaks are
increased to 1270 peaks targeting 875 genes. Of these peaks, 16.61% of them are
in the promoter region, 39.08% are in the distal intergenic region. In OCIAML3, 4149 peaks are increased to 5399 peaks targeting 4053 genes. 18.88% of
them are in the promoter region, 35.3% in the distal intergenic region. The
UOC-M1, 4211 peaks increased to 5975 peaks targeting 4199 genes, and 17.82
of 4211 genes are in the promoter region and 36.49% in the distal intergenic
region.

27
.
4.2 Common MLL1 targets in three leukemia cell lines
The MLL1 target gene list is generated by two different methods (Figure 10-a). We hope to find
out common MLL1 target genes shared in three leukemia. For the direct overlap method, we
found 150 MLL1 direct targets shared by three leukemia cell lines. In comparison, about 2470
genes are unique MLL1 targets in UOC-M1. About 2270 genes are unique MLL1 targets in OCIAML3, and only 336 genes are unique MLL1 targets of MOLM-13. Further analysis showed that
there were only 144 genes shared between UOC-M1 and MOLM-13, 173 genes shared between
OCI-AML3 and MOLM-13, and there were 1180 gene shared between OCI-AML3 and UOCM1 (Figure 10-b). In gap overlap method, we found 175 genes shared in three cell lines. There
were 2638 unique MLL1 targets in UOC-M1. There were 2488 unique MLL1 targets in OCIAML3, and 358 MLL1 unique targets in MOLM-13. There were only 165 MLL1 direct targets
shared in MOLM-13 and UOC-M1,

28
Figure 10 a. Gene count and peak count of different methods to generate overlap peak
region: Direct overlap method: 929 genes in MOLM-13, 4149 genes in OCIAML3, 4211 genes in UOC-M1. Gap overlap method: 1270 genes in MOLM13, 5399 genes in OCI-AML3, 5629 genes in UOC-M1.
b. Venn Diagram for genes overlap and shared in three cell lines: In direct
overlap method: 150 genes are shared targets of MLL1 in three leukemia cell
lines. In gap overlap method: 175 genes are shared targets of MLL1 in three
leukemia cell lines.

29
4.3 Network analyses in three leukemia cell lines
For the 175 MLL1 target genes shared in three cell lines, we use STRING to investigate
potential interaction between these proteins. We found that 142 out of 175 genes are presented in
the protein interaction network (Figure 11). Myc was at the center of the interaction network.
Myc is a transcription factor. It can interact with WDR5, a core component of the MLL1
complex. Our study showed Myc as a direct target of MLL1. MLL1 binds at the Myc gene
promoter and deposit H3K4me3, potentially contributing to tumorigenesis [30]. In addition,
deregulated Myc is found in about half of human tumors and is more prevalent in hematological
neoplasms. Myc plays key roles at different stages of the antigen-producing B-cell differentiation
process. This process is often dysregulated in lymphoid neoplasms. Myc expression is decreased
after MLL1 knockout, indicating an essential role of MLL1 in controlling Myc expression. Myc
inhibitor of Myc has been shown to have anti-leukemic effects in MLL-r leukemia [32]. The
inhibitors also activate T, NK and iNKT cells, increase the expression of immunostimulatory
cytokines like interferon γ (IFNγ), decrease the expression of both programmed cell death-1 (PD1) on the immune effector cells and the respective ligand (PD-L1) on the target cells, leading to a
more efficient killing of MLL-r cells[32].
Based on our result, we conclude that the MLL1 directly target the expression of multiple genes,
including key oncogenes like Myc. We found 175 MLL1 target genes shared in three leukemia
cell line. These genes form an interaction network including curial oncogene Myc. We also want
to find out whether MLL1 mediates expression of these genes, which will be explored in the last
chapter.

30
Figure 11
Protein-protein interaction analysis: There are 175 MLL1 target genes shared
in three cell lines, 32 of them were removed as they have no functional
annotation in the database. The blue arrow indicates Myc. The interaction
network is highly connected to Myc.

31
Chapter 5: Function of MLL1 in Gene Expression in MLL1
Amplified Leukemia
MLL1 shows different binding and methylation activity in three leukemia cell lines. These
enrichment regions are associated with promoters and enhancers of many genes, potentially
regulating their expression levels. In particular, we are interested in demonstrating that MLL1
binding and HMT activity mediate gene expression and drive the transcription program in MLL1
amplified leukemia.
5.1 RNA-seq analysis for UOC-M1 cells with or without MLL1 knockout
To investigate MLL1 functions in UOC-M1, we constructed a MLL1-knockout UOC-M1 model
and conducted RNA-seq analysis. By using different thresholds, we generated two lists of
differentially expressed genes. With the threshold set at log2fold greater than 0.9 and p-value
after adjustment to be smaller than 0.05, 1418 genes showed significant gene expression
changes. With the threshold set at log2change greater than 1 and p-value after adjustment to be
smaller than 0.05, 1212 genes showed significant gene expression changes. We next used the
DAVID database to analyze the pathway enrichment. KEGG pathway analysis showed that 307
out of 1212 genes, or 351 out of 1418 genes were enriched in pathways.
MLL1 direct targets with expression changes were enriched in the Rap1 signaling pathway,
pathway in cancer, PI3K-Akt signaling pathway, and hematopoietic cell lineage pathway (Figure
13). These pathways are highly related to cancer initiation and maintenance.

32
The Rap1 signaling pathway is related to cell adhesion, migration and polarity. Ras-associated
protein1 (Rap1) is a member of the Ras family of small GTPases and functions as a molecular
switch. In some solid tumors, the activated Rap1 leads to integrin-mediated cell adhesion in
cancer via EGFR and Src/FAK pathway. Src may activate MAPK/ERK to increase VEGF
expression levels in the tumor cells, ultimately leading to angiogenesis. Src may also induce
p130Cas and PD-L1 expression in the tumor cells, avoiding immune destruction. Interaction with
MMP also accelerates metastasis [33]. The Rap1 signaling pathway has crosstalk with the PI3KAkt signaling pathway and Rap1 directly activates PI3K. From our analysis, we found that
RAPGEF2, which is MLL1 target only in UOC-M1 cells, was down-regulated after MLL1
knockout. RAPGEF2 plays an important role in embryonic hematopoiesis [34]. It directly
activates the Rap1 pathway by accelerating the transition of Rap1 from its GDP-bound inactive
state to its GTP-bound active state. Rap1 activation may lead to PI3K-Akt signaling mediated
cell proliferation.
There are 52 gene involved in the pathway for cancer from KEGG analyses. Their expression
levels were altered after MLL1 knockout, 34 of which were down-regulated. They include:
AGTR1, CYCS, DCC, DLL1, ESR2, FGF17, FGFR4, FLT4, GLI2, GNG2, GSTA4, GSTM3,
HES1, KIT, LAMA3, LAMB3, LAMC3, LEF1, MITF, MYC, NOTCH3, NTRK1, PLCB4,
PLD2, PLEKHG5, PTGER3, SMAD3, SMO, STAT4, TERT, TRAF3, WNT10B, WNT3 and
WNT5B.
For these gens, we found some of them are members of the Notch signaling pathway. Notch
activation is important in genesis and maintenance of lymphoid malignancies [36]. Previous
studies have reported that activated Notch signaling pathway promotes AML cell proliferation

33
and migration in vitro and accelerates AML progression in vivo [54]. The GSTs family is related
to drug resistance in leukemia (human leukemia cell line: HL60). Cancer cells with
overexpression of GSTA4 resist better to Sulforaphane and Curcumin induced toxicity[37].

34
Figure 12
The KEGG pathway analysis: The Rap1 signaling pathway, pathway in cancer, PI3K-Akt
signaling pathway and hematopoietic cell lineage were highlighted.

35
5.2 Expression changes of MLL1 target genes after MLL1 knockout
We next studied whether MLL1 target genes are related to the Rap1 pathway, pathway in cancer
or PI3K-Akt pathway. For MLL1 target genes shared in three cell lines, 13 out of 175 genes have
expression changes after MLL1 knockout. The six genes that were downregulated included:
CA3-AS1, PSD3, EXT1, FABP5, CTSB and Myc.
Table1 Function of MLL1 shared target
For MLL1 unique targets genes in UOC-M1 cells, 274 genes showed expression changes after
MLL1 knockout. They were enriched in the Wnt and TGF-beta pathway. The Wnt signaling
pathway is an ancient and evolutionarily conserved pathway that regulates crucial aspects of cell
fate determination, cell migration, cell polarity, neural patterning and organogenesis during
embryonic development. Overexpression of non-canonical Wnt signaling will lead to the
survival of cancer stem cells (CSCs) [35]. The Wnt/Ca2+ pathway is one of the non-canonical
Wnt pathway in which Wnt5b is involved [35].
Gene Function in Cancer Reference
CA3-AS1 Overexpression induce cell apoptosis in gastric cancer [43]
PSD3 Promotes proliferation, migration, invasion [44]
EXT1 Tumor suppressor in AML [45]
FABP5 AML cell survival [46]
CTSB Inhibition of CTSB activity induce apoptosis in AML [47]
Myc Oncogene in myeloid neoplasias [30]

36
Among 274 genes, 141 genes were downregulated following MLL1 knockout, indicating that
MLL1 activates their transcription. From the downregulated genes, 16 are enriched in cancerrelated pathways, including NTRK1, WNT5B, OLR1, MYC, PLEKHG5, ZNF276, SMAD3,
SCAMP2, MITF, LAMA3, DLL1, WDR1, PTGER3, NOTCH3, and ANKHD1. Additionally,
we found that MLL1 also regulates non-coding RNAs (e.g. CA3-AS1, SOX21-AS1).
Gene Function in Cancer Reference
NTRK1 Overexpression related with cancer development [48]
WNT5B Convert pre-leukemic stem cells into leukemia stem cells [35]
OLR1 Pan-Cancer prognostic and immunotherapeutic predictor [49]
FABP5 Overexpression promotes tumour growth. [50]
Myc Oncogene in myeloid neoplasias [30]
ZNF276 Promote breast cancer cell proliferation [51]
SMAD3 Overexpression suggest poor prognosis [40]
SCAMP2 Diagnostic and prognostic marker for AML [52]
MITF Overexpression in chronic myeloid leukemia (CML) patients [53]
DLL1 Overexpression activate Notch signal pathway [54]
NOTCH3 Essential in lymphoid malignancies, related with AML
proliferation and AML progression in vivo
[36]

37
Figure 13
The KEGG pathway analysis of MLL1 & H3K4me3 joint targets that were significantly
changed upon MLL1 deletion in UOC-M1 cells. Hematopoietic cell lineage pathway, with
selection hold fold>=1, the CSF1, TFRC had expression down regulated and ITGA2 up
regulated after loss function of MLL1. With selection hold fold>=0.9, the CR2 was identified
as down regulated genes.
The MLL1 & H3K4me3 shared target genes showed TGF-beta signaling pathway enrichment
with RNA-seq data selection hold fold>=0.9. It has high confidence level. The Myc is
involved in TGF-beta pathway.

38
The Wnt5b is a member of the Wnt family of proteins, its structure is related to Wnt5a. It is an
MLL1 target in UOC-M1. It is important during hematopoiesis, bone marrow colonization and
hematopoietic stem cell (HSC) maturation via β-catenin-independent signaling. In chronic
lymphocytic leukemia (CLL), Wnt5b is expressed at higher than that of normal B cells. During
leukemia development, it is crucial to convert pre-leukemic stem cells into leukemia stem
cells[35]. Furthermore, Wnt/Ca2+ is related to AML drug resistance [35].
The transforming growth factor-β (TGF-β) signaling pathway regulates cell proliferation,
differentiation, migration, and cell survival. In the hematopoiesis process, TGF-beta is one of the
main regulators that keep hematopoietic stem cells (HSCs) in quiescence to maintain stemness. It
also promotes differentiation or induce apoptosis for certain condition [39]. Deregulation of the
TGF-beta pathway greatly contributes to cancer development. Multiple therapeutic strategies
target this pathway to improve prognosis[38] [39].
The Smad3 is one of the unique targets of MLL1 in UOC-M1. It is downregulated after MLL1
knockout. During TGF-beta signaling, Smad2/3 are phosphorylated by activated TβRI,
translocated into the nucleus and regulates the transcription of TGF-β–responsive genes by
forming complex with multiple factors [39]. In leukemia, this process is disrupted by different
mechanisms. The PML-RARα prevents the phosphorylation of Smad2/3. Onco-drivers such as
AML/ETO, AML/EVI-1, and Evi-1, inhibit Smad3 DNA binding and recruit the transcription
repressor. Tax blocks the formation of Smad2/3/4 complexes [40]. TGF-β/SMAD3 signaling in
solid tumors plays a role in tumor suppression during the early phase of tumorigenesis [40], but
it also could promote epithelial-to-mesenchymal transition of tumor cells at later stage for tumor
invasion and metastasis. The dysregulation of TGF-β/SMAD3 in leukemia has been found in

39
many cases [40]. High expression level of SMAD3 together with SMAD7 suggest poor
prognosis in AML patients. These studies suggest that overexpression of Smad3 could contribute
to leukemia development. Interestingly, they also found a correlation between high expression
level of Smad3 and TP53 or RunX1 mutation [41].
5.3 MLL1 targets non-coding RNAs
MLL1 regulates expression of non-coding RNAs to modulate cellular processes. The SRY-box
transcription factor 21 antisense divergent transcript 1 (SOX21-AS1) is a long non-coding RNA
and is a potential oncogene [42]. It can inhibit downstream microRNAs. SOX21-AS1 is involved
in multiple cellular processes including TGF-beta signaling, Wnt signaling and PI3K/AKT
signaling pathway [42]. Overexpression of SOX21-AS1 is involved in epithelial mesenchymal
transition (EMT), invasion, migration, apoptosis, and cell cycle arrest [42]. The SOX21-AS1
also has been considered as a prognosis biomarker in multiple cancers including lung cancer,
breast cancer and leukemia [42]. The exact mechanism of how SOX21-AS1 is involved in these
processes still remains elusive. SOX21-AS1 is downregulated after MLL1 knockout and it is a
unique target of MLL1 in UOC-M1.
In this chapter, we constructed MLL1 knockout UOC-M1 cell model and evaluated expression
changes. We found 141 genes are downregulated by MLL1 knockout and some of them are
known to contribute to leukemogenesis. We conclude that MLL1 promotes expression of
potential oncogenes in leukemia with MLL1 amplification.

40
Chapter 6: Discussion and Future Directions
Our study investigates the MLL1 binding and its methylation activity in three leukemia cell
lines. We also evaluate their expression change after MLL1 knockout in UOC-M1. The result
provides insights on MLL1 function in MLL1-amplified leukemia.
We reveal that MLL1 is associated with elevated H3K4me3 in UOC-M1 as compared with the
other two cell lines. There are 175 MLL1 direct targets with high level of H3K4me3 that are
shared in three cell lines. Based on the STRING analyses, the MLL1 direct targets form a
network centered around Myc. This is consistent with MLL1 directly regulating Myc expression,
which is confirmed by our RNA-seq analyses. These results support that Myc is one of the major
effectors of MLL1 in MLL1-amplified leukemia.
There are 141 genes downregulated after MLL1 knockout. It suggests that MLL1 is the main
transcription cofactor for these genes. 102 of them are MLL1 unique targets in UOC-M1, 6 are
MLL1 target genes shared in three cell lines. Some of these have well established roles in cancer.
For example, Wnt5b, Smad3 and SOX21-AS1 are known to promote cancer development,
although their functions in leukemia are less well understood.
The SOX21-AS1 is a MLL1 unique target in UOC-M1, but it is not clear whether the
amplification of MLL1 causes the overexpression of SOX21-AS1. Future studies may allow us
to demonstrate that amplification of MLL1 causes SOX21-AS1 overexpression and drives

41
leukemia development. It is also valuable to investigate if the MLL1 targets SOX21-AS1 in
other cancers and contributes to their pathogenesis. In addition, SOX21-AS1 is a non-coding
RNA, and MLL1 may target more non-coding genes and mediate their expression in leukemia or
other type of cancers. Therefore, future research for MLL1’s role in non-coding RNA regulation
may be of value.
MLL1 amplification is often found in aged patients, and these patients show unique clinical
characteristics [56]. We reveal that MLL1 promotes the expression of potential oncogenes in
leukemia with MLL1 amplification. But the sequence of molecule events that eventually lead to
leukemogenesis with MLL1 amplification remains to be elucidated.
The functions of MLL1 target regions outside the promoter, specifically in the distal intergenic
regions, remain underexplored. For MLL1, more than 35% of the MLL1/H3K4me3 joint peak
regions in three different cell lines are in distal intergenic regions. While the potential functions
of these regions are not yet clear, they might be related to the transcription of genes that are far
apart in the linear sequence but close in three-dimensional space. Currently, 3D genome data for
UOC-M1 and OCI-AML3 are limited. Therefore, research focusing on MLL1-mediated 3D
genome changes in different cell lines may be warranted.

42
References
[1] Neganova ME, Klochkov SG, Aleksandrova YR, Aliev G. Histone modifications in epigenetic
regulation of cancer: Perspectives and achieved progress. Seminars in cancer biology.
2022;83:452-471. doi:10.1016/j.semcancer.2020.07.015
[2] Zhu G, Luo H, Feng Y, et al. HOXBLINC long non-coding RNA activation promotes
leukemogenesis in NPM1-mutant acute myeloid leukemia. Nature communications. 2021;12(1).
doi:10.1038/s41467-021-22095-2
[3] Tkachuk DC, Kohler S, Cleary ML. Involvement of a homolog of Drosophila trithorax by 11q23
chromosomal translocations in acute leukemias. Cell. 1992;71(4):691-700. doi:10.1016/0092-
8674(92)90602-9
[4] Felix CA, Hosier MR, Winick NJ, Masterson M, Wilson AE, Lange BJ. ALL-1 Gene
Rearrangements in DNA Topoisomerase II Inhibitor-Related Leukemia in Children. Blood.
1995;85(11):3250-3256. doi:10.1182/blood.V85.11.3250.bloodjournal85113250
[5] Gan T, Jude CD, Zaffuto K, Ernst P. Developmentally induced Mll1 loss reveals defects in
postnatal haematopoiesis. Leukemia. 2010;24(10):1732-1741. doi:10.1038/leu.2010.171
[6] Jude CD, Climer L, Xu D, Artinger E, Fisher JK, Ernst P. Unique and Independent Roles for
MLL in Adult Hematopoietic Stem Cells and Progenitors. Cell stem cell. 2007;1(3):324-337.
doi:10.1016/j.stem.2007.05.019
[7] Artinger EL, Mishra BP, Zaffuto KM, et al. MLL-dependent network sustains hematopoiesis.
Proceedings of the National Academy of Sciences - PNAS. 2013;110(29):12000-12005.
doi:10.1073/pnas.1301278110
[8] Kataoka K, Kurokawa M. Ecotropic viral integration site 1, stem cell self‐renewal and
leukemogenesis. Cancer science. 2012;103(8):1371-1377. doi:10.1111/j.1349-
7006.2012.02303.x
[9] Lawrence HJ, Christensen J, Fong S, et al. Loss of expression of the Hoxa-9 homeobox gene
impairs the proliferation and repopulating ability of hematopoietic stem cells. Blood.
2005;106(12):3988-3994. doi:10.1182/blood-2005-05-2003
[10]McMahon KA, Hiew SYL, Hadjur S, et al. Mll Has a Critical Role in Fetal and Adult
Hematopoietic Stem Cell Self-Renewal. Cell stem cell. 2007;1(3):338-345.
doi:10.1016/j.stem.2007.07.002
[11]Chan AKN, Chen CW. Rewiring the epigenetic networks in MLL-rearranged leukemias:
Epigenetic dysregulation and pharmacological interventions. Frontiers in cell and developmental
biology. 2019;7(May):81-81. doi:10.3389/fcell.2019.00081
[12]Pediatric acute leukemia fact sheet.
https://www.epizyme.com/wpcontent/uploads/2014/05/EPZM_Acute_Leukemia_Fact_Sheet_5.9
.pdf. Accessed 1 Jan 2021

43
[13]Gill Super HJ, McCabe NR, Thirman MJ, et al. Rearrangements of the MLL Gene in TherapyRelated Acute Myeloid Leukemia in Patients Previously Treated With Agents Targeting DNATopoisomerase II. Blood. 1993;82(12):3705-3711. doi:10.1182/blood.V82.12.3705.3705
[14]Felix CA. Secondary leukemias induced by topoisomerase-targeted drugs. BBA - Gene Structure
and Expression. 1998;1400(1):233-255. doi:10.1016/S0167-4781(98)00139-0
[15]Vedadi M, Blazer L, Eram MS, Barsyte‐Lovejoy D, Arrowsmith CH, Hajian T. Targeting human
SET1/MLL family of proteins. Protein science : a publication of the Protein Society.
2017;26(4):662-676. doi:10.1002/pro.3129
[16]Chen Y, Anastassiadis K, Kranz A, et al. MLL2, Not MLL1, Plays a Major Role in Sustaining
MLL-Rearranged Acute Myeloid Leukemia. Cancer cell. 2017;31(6):755-770.e6.
doi:10.1016/j.ccell.2017.05.002
[17]Park SH, Ayoub A, Lee YT, et al. Cryo-EM structure of the human MLL1 core complex bound
to the nucleosome. Nature communications. 2019;10(1):5540-5540. doi:10.1038/s41467-019-
13550-2
[18]Ayoub A, Park SH, et al. Mechanism for DPY30 and ASH2L intrinsically disordered regions to
modulate the MLL/SET1 activity on chromatin. Nature communications. 2021;12(1):2953-2953.
doi:10.1038/s41467-021-23268-9
[19]Li X, Song Y. Structure, function and inhibition of critical protein–protein interactions involving
mixed lineage leukemia 1 and its fusion oncoproteins. Journal of hematology and oncology.
2021;14(1):56-56. doi:10.1186/s13045-021-01057-7
[20]Milne TA, Kim J, Wang GG, et al. Multiple Interactions Recruit MLL1 and MLL1 Fusion
Proteins to the HOXA9 Locus in Leukemogenesis. Molecular cell. 2010;38(6):853-863.
doi:10.1016/j.molcel.2010.05.011
[21]Ayton PM, Cleary ML. Transformation of myeloid progenitors by MLL oncoproteins is
dependent on Hoxa7 and Hoxa9. Genes & development. 2003;17(18):2298-2307.
doi:10.1101/gad.1111603
[22]Wang Q fei, Wu G, Mi S, et al. MLL fusion proteins preferentially regulate a subset of wild-type
MLL target genes in the leukemic genome. Blood. 2011;117(25):6895-6905. doi:10.1182/blood2010-12-324699
[23]Arai S, Yoshimi A, Shimabe M, et al. Evi-1 is a transcriptional target of mixed-lineage leukemia
oncoproteins in hematopoietic stem cells. Blood. 2011;117(23):6304-6314. doi:10.1182/blood2009-07-234310
[24]Grinat J, Heuberger J, Vidal RO, et al. The epigenetic regulator Mll1 is required for Wnt-driven
intestinal tumorigenesis and cancer stemness. Nature communications. 2020;11(1):6422-6422.
doi:10.1038/s41467-020-20222-z

44
[25]Zhu J, Sammons MA, Donahue G, et al. Gain-of-function p53 mutants co-opt chromatin
pathways to drive cancer growth. Nature (London). 2015;525(7568):206-211.
doi:10.1038/nature15251
[26]Bhan A, Deb P, Shihabeddin N, Ansari KI, Brotto M, Mandal SS. Histone methylase MLL1
coordinates with HIF and regulate lncRNA HOTAIR expression under hypoxia. Gene.
2017;629:16-28. doi:10.1016/j.gene.2017.07.069
[27]Matsuo Y, MacLeod RAF, Uphoff CC, et al. Two acute monocytic leukemia (AML-M5a) cell
lines (MOLM-13 and MOLM-14) with interclonal phenotypic heterogeneity showing MLL-AF9
fusion resulting from an occult chromosome insertion, ins(11;9)(q23;p22p23). Leukemia.
1997;11(9):1469-1477. doi:10.1038/sj.leu.2400768
[28]Tang G, DiNardo C, Zhang L, et al. MLL gene amplification in acute myeloid leukemia and
myelodysplastic syndromes is associated with characteristic clinicopathological findings and
TP53 gene mutation. Human pathology. 2015;46(1):65-73. doi:10.1016/j.humpath.2014.09.008
[29]Mondello C, Smirnova A, Giulotto E. Gene amplification, radiation sensitivity and DNA doublestrand breaks. Mutation research. 2010;704(1):29-37.doi:10.1016/j.mrrev.2010.01.008
[30]Dolores Delgado M, León J. Myc roles in hematopoiesis and leukemia. Genes & cancer.
2010;1(6):605-616. doi:10.1177/1947601910377495
[31]Thomas LR, Wang Q, Grieb BC, Phan J, Foshage AM, Sun Q, et al. Interaction with WDR5
promotes target gene recognition and tumorigenesis by MYC. Mol Cell (2015) 58:440–52. doi:
10.1016/j.molcel.2015. 02.028
[32]Fitzel R, Secker-Grob KA, Keppeler H, et al. Targeting MYC in combination with epigenetic
regulators induces synergistic anti-leukemic effects in MLLr leukemia and simultaneously
improves immunity. Neoplasia (New York, NY). 2023;41:100902-100902.
doi:10.1016/j.neo.2023.100902
[33]Looi CK, Hii LW, Ngai SC, Leong CO, Mai CW. The role of ras-associated protein 1 (Rap1) in
cancer: Bad actor or good player? Biomedicines. 2020;8(9):334-.
doi:10.3390/BIOMEDICINES8090334
[34]Satyanarayana A, Gudmundsson KO, Chen X, et al. RapGEF2 is essential for embryonic
hematopoiesis but dispensable for adult hematopoiesis. Blood. 2010;116(16):2921-2931.
doi:10.1182/blood-2010-01-262964
[35]Láinez-González D, Alonso-Aguado AB, Alonso-Dominguez JM. Understanding the Wnt
Signaling Pathway in Acute Myeloid Leukemia Stem Cells: A Feasible Key against Relapses.
Biology (Basel, Switzerland). 2023;12(5):683-. doi:10.3390/biology12050683
[36]McCarter AC, Wang Q, Chiang M. Notch in Leukemia. Vol 1066. Springer International
Publishing :; 2018:355-394. doi:10.1007/978-3-319-89512-3_18
[37]Sharma R, Ellis B, Sharma A. Role of alpha class glutathione transferases (GSTs) in
chemoprevention: GSTA1 and A4 overexpressing human leukemia (HL60) cells resist

45
sulforaphane and curcumin induced toxicity. Phytotherapy research. 2011;25(4):563-568.
doi:10.1002/ptr.3297
[38]Huang F, Wan J, Hu W, Hao S. Enhancement of Anti-Leukemia Immunity by LeukemiaDerived Exosomes Via Downregulation of TGF-β1 Expression. Cellular physiology and
biochemistry. 2017;44(1):240-254. doi:10.1159/000484677
[39]Vaidya A, Kale VP. TGF-β signaling and its role in the regulation of hematopoietic stem cells.
Systems and synthetic biology. 2015;9(1-2):1-10. doi:10.1007/s11693-015-9161-2
[40]Brown KA, Pietenpol JA, Moses HL. A tale of two proteins: Differential roles and regulation of
Smad2 and Smad3 in TGF-β signaling. Journal of cellular biochemistry. 2007;101(1):9-33.
doi:10.1002/jcb.21255
[41]Zhang J, Zhang L, Cui H, et al. High expression levels of SMAD3 and SMAD7 at diagnosis
predict poor prognosis in acute myeloid leukemia patients undergoing chemotherapy. Cancer
gene therapy. 2019;26(5-6):119-127. doi:10.1038/s41417-018-0044-z
[42]Lu Z, Xu J, Cao B, Jin C. Long non-coding RNA SOX21-AS1: A potential tumor oncogene in
human cancers. Pathology, research and practice. 2023;249:154774-154774.
doi:10.1016/j.prp.2023.154774
[43]Zhang XY, Zhuang HW, Wang J, et al. Long noncoding RNA CA3-AS1 suppresses gastric
cancer migration and invasion by sponging miR-93-5p and targeting BTG3. Gene therapy.
2022;29(9):566-574. doi:10.1038/s41434-020-00201-1
[44]Jin L, Zheng D, Bhandari A, et al. PSD3 is an oncogene that promotes proliferation, migration,
invasion, and G1/S transition while inhibits apoptotic in papillary thyroid cancer. Journal of
Cancer. 2021;12(18):5413-5422. doi:10.7150/jca.60885
[45]Gözdaşoğlu S, Uysal Z, Kürekçi AE, et al. HEREDITARY MULTIPLE EXOSTOSES AND
ACUTE MYELOID LEUKEMIA: An Unusual Association. Pediatric hematology and oncology.
2000;17(8):707-711. doi:10.1080/08880010050211439
[46]Liang X, Fu W, Peng Y, Duan J, Zhang T. Lycorine induces apoptosis of acute myeloid
leukemia cells and inhibits triglyceride production via binding and targeting FABP5. Annals of
hematology. 2023;102(5):1073-1086. doi:10.1007/s00277-023-05169-7
[47]Pandey G. Prognostic and therapeutic relevance of cathepsin B in pediatric acute myeloid
leukemia. American journal of cancer research. 2019;9(12):2634-2649.
[48] Khotskaya YB, Holla VR, Farago AF, Mills Shaw KR, Meric-Bernstam F, Hong DS. Targeting
TRK family proteins in cancer. Pharmacology & therapeutics. 2017;173:58-66.
doi:10.1016/j.pharmthera.2017.02.006
[49] New Findings from Wuhan University in the Area of Cancer Described ( OLR1 Is a Pan-Cancer
Prognostic and Immunotherapeutic Predictor Associated with EMT and Cuproptosis in HNSCC).
NewsRX LLC; 2023:3021-.

46
[50]George Warren W, Osborn M, Yates A, Wright K, O’Sullivan SE. The emerging role of fatty
acid binding protein 5 (FABP5) in cancers. Drug discovery today. 2023;28(7):103628-103628.
doi:10.1016/j.drudis.2023.103628
[51]Lei T, Zhang W, He Y, et al. ZNF276 promotes the malignant phenotype of breast carcinoma by
activating the CYP1B1-mediated Wnt/β-catenin pathway. Cell death & disease. 2022;13(9):781-
781. doi:10.1038/s41419-022-05223-8
[52]Yue C, Xie S, Zhong J, et al. SCAMP2/5 as diagnostic and prognostic markers for acute myeloid
leukemia. Scientific reports. 2021;11(1). doi:10.1038/s41598-021-96440-2
[53]Aggoune D, Sorel N, Bonnet ML, Goujon JM. Bone marrow mesenchymal stromal cell (MSC)
gene profiling in chronic myeloid leukemia (CML) patients at diagnosis and in deep molecular
response induced by tyrosine kinase inhibitors (TKIs). Leukemia research. 2017;60:94-102.
doi:10.1016/j.leukres.2017.07.007
[54]Li F, Han Y, Chen R, et al. MicroRNA-143 acts as a tumor suppressor through Musashi2/DLL1/Notch1 and Musashi-2/Snail1/MMPs axes in acute myeloid leukemia. Journal of
translational medicine. 2023;21(1). doi:10.1186/s12967-023-04106-6
[55] Bedtools function: https://bedtools.readthedocs.io/en/latest/content/tools/intersect.html
[56] Cuthbert G, Thompson K, McCullough S, et al. MLL amplification in acute leukaemia: a United
Kingdom Cancer Cytogenetics Group (UKCCG) study. Leukemia : official journal of the
Leukemia Society of America, Leukemia Research Fund, UK. 2000;14(11):1885-1891.
doi:10.1038/sj.leu.2401919

Abstract (if available)

Abstract Mixed lineage leukemia 1 (MLL1, also known as MLL or KMT2A) is a crucial transcription cofactor required for early embryonic development. MLL1 also plays an essential role in the development and maintenance of hematopoietic stem cells (HSCs). It is frequently mutated in cancer, including acute leukemia. Our research here compared MLL1 direct target genes in leukemia with different MLL1 aberrations. We revealed the difference in MLL1 binding in three leukemia cell lines with MLL1 rearrangement, MLL1 amplification, and wild type MLL1, respectively. By analyzing histone 3 lysine 4 trimethylation and gene expression in these cell lines, we found that oncogene Myc and its interaction network were the main MLL1 targets. We further showed that MLL1 regulated additional genes highly related to cancer development and leukemogenesis. We found that MLL1 amplification regulated not only Myc, Smad3 and Wnt5b, but also other oncogenes, such as SOX21-AS1, and a non-coding micro-RNA that have not been previously reported. Our research highlights unique and common pathways directly regulated by MLL1 in AML with different MLL1 aberrations.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Developing MLL1 degraders as therapeutics for leukemia

PDF

Cancer-recurrent missense mutations in SET domain alter KMT2A methyltransferase activity

PDF

Evaluating the therapeutic potential of targeting CBX8 in MLLr leukemia

PDF

Mixed lineage leukemia proteins (MLLs), their effect as coregulators on target gene expression and global histone methylation

PDF

Collecting epigenetic information with DNA methylation footprinting

PDF

Epigenetic regulation of non CPG island gene promoters

PDF

Do the ZFX and ZFY transcription factors have redundant or unique functions?

PDF

Elucidation of MBNL1 function in the nervous system of myotonic dystrophy type 1

PDF

MSI2 level alters histone transcription rate in HepG2 cells

PDF

A novel screening approach to identify regulators of polycomb-dependent gene silencing

PDF

Role of apolipoprotein C2 in acute myeloid leukemia: clinical, functional and mechanistic study

PDF

Generation of an epigenetic toggle switch to test LINC00261 function on lung adenocarcinoma cellular response to the chemotherapeutics oxaliplatin and carboplatin

PDF

Characterizing ZFX-mediated gene regulation to reveal possible candidates for clinical intervention

PDF

Understanding ancestry-specific disease allelic effect sizes by leveraging multi-ancestry single-cell RNA-seq data

PDF

Transcription errors induce proteotoxic stress in mammalian cells

PDF

Runx2 interactions with the osteoblast genome

PDF

RNA polymerase III-dependent transcription is repressed under prolonged hypoxic conditions

PDF

Investigation of a causal role of transposable element activation in vertebrate aging

PDF

The role of microRNAs in cancer

PDF

Defining the functional region of LINC00261 in lung adenocarcinoma

Asset Metadata

Creator Wang, Yongxiang (author)

Core Title Epigenetic dysregulation in acute myeloid leukemia (AML) with MLL1 aberrations

School Keck School of Medicine

Degree Master of Science

Degree Program Biochemistry and Molecular Medicine

Degree Conferral Date 2024-08

Publication Date 07/16/2024

Defense Date 06/28/2024

Publisher Los Angeles, California (original), University of Southern California (original), University of Southern California. Libraries (digital)

Tag AML,ChIP-seq,epigentic,KMT2A,RNA-seq

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Dou, Yali (committee chair), Liu, Yifan (committee member), Rice, Judd (committee member)

Creator Email yongxiangwang0528@outlook.com,ywang863@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC113997X8M

Unique identifier UC113997X8M

Identifier etd-WangYongxi-13245.pdf (filename)

Legacy Identifier etd-WangYongxi-13245

Document Type Thesis

Format theses (aat)

Rights Wang, Yongxiang

Internet Media Type application/pdf

Type texts

Source 20240718-usctheses-batch-1184 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu