University of Southern California Dissertations and Theses

Identification of CBP/FOXM1 as a molecular target in triple negative breast cancer

(USC Thesis Other)

Identification of CBP/FOXM1 as a molecular target in triple negative breast cancer

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content 1

Identification of CBP/FOXM1 as a Molecular Target in Triple Negative
Breast Cancer

Alexander Ring

Cancer Biology and Genomics
Doctor of Philosophy

USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA

Degree conferral date
August 2017

2

To my family

Who always believed in me
So I could believe in my dreams
3
Table of contents

1. INTRODUCTION ...................................................................................................... 11
1.1. Breast cancer ....................................................................................................... 11
1.1.1. Breast cancer as a public health issue........................................................ 11
1.1.2. Characterizing breast cancer in the clinical setting – grading and staging 11
1.1.3. Diagnosis, treatment and biomarkers in breast cancer .............................. 12
1.1.4. Limitation of “traditional” biomarkers ...................................................... 13
1.1.5. Molecular classification............................................................................. 14
1.1.6. Triple negative breast cancer ..................................................................... 16
1.2. Cancer stem cells (CSCs) .................................................................................... 19
1.3. Molecular driver of stem cells biology – The Wnt signaling pathway ............... 21
1.3.1. CREB-binding protein (CBP) ................................................................... 22
1.3.2. Wnt and cancer stem cells ......................................................................... 23
1.3.3. Wnt and breast cancer ............................................................................... 24
1.4. FOXM1 in physiology and disease ..................................................................... 25
1.4.1. FOXM1 and breast cancer ......................................................................... 27
1.5. Convergence of FOXM1 and Wnt signaling in cancer and CSC biology ........... 28
2. METHODS ................................................................................................................. 29
2.1. Cell culture .......................................................................................................... 29
2.2. Cell passaging ...................................................................................................... 30
2.3. Transient transfection .......................................................................................... 31
2.4. TOPFLASH/FOPFLASH assay .......................................................................... 32
2.5. Survivin and FOXM1 reporter assay ................................................................... 33
2.6. FOXM1 overexpression ...................................................................................... 34
2.7. siRNA mediated gene knockdown ...................................................................... 34
2.8. Animal experiments ............................................................................................ 34
2.8.1. Tumor cell line subcutaneous (s.c.) injection ............................................ 35
2.8.2. Patient tumor acquisition and PDX establishment .................................... 36
2.8.3. Xenograft line propagation ........................................................................ 37
2.9. Xenograft cell preparation for in vitro assays ..................................................... 37
2.10. In vitro and in vivo treatment procedures .......................................................... 38
2.10.1. Drug preparation and in vitro application ............................................... 38
2.10.2. Establishing Paclitaxel resistant triple negative breast cancer cells ........ 39
4
2.10.3. In vivo treatment of mice bearing xenografts.......................................... 39
2.10.4. ICG-001 application via osmotic pump implantation ............................. 39
2.10.5. Intraperitoneal (i.p.) injections of Paclitaxel ........................................... 40
2.11. Sample harvest and analysis .............................................................................. 41
2.11.1. RNA extraction ........................................................................................ 41
2.11.2. Reverse transcription of RNA to cDNA .................................................. 42
2.11.3. Quantitative reverse transcription polymerase chain reaction (qRT-PCR)
............................................................................................................................. 42
2.12. RNA Seq library preparation ............................................................................. 44
2.13. RNA Seq data analysis....................................................................................... 45
2.14. Biological RNA Seq data interpretation – Ingenuity Pathway analysis (IPA) .. 46
2.15. Public breast cancer gene expression data repositories ..................................... 46
2.16. Fluorescence Activated Cell sorting (FACS)/ flow cytometry .......................... 47
2.17. Side population assay using FACS .................................................................... 48
2.18. Protein extraction ............................................................................................... 49
2.19. Western blotting ................................................................................................. 50
2.20. Co-immunoprecipitation (CoIP) ........................................................................ 52
2.21. Tissue Micro Arrays (TMAs) ............................................................................ 53
2.22. Statistical analysis .............................................................................................. 55
3. RESULTS ................................................................................................................... 57
3.1. CREB binding protein (CBP) and CBP/β-catenin interaction as a potential target
in triple negative breast cancer (TNBC) .............................................................. 57
3.2. FOXM1 and its target genes are over-expressed in TNBC ................................. 60
3.3. A chemical-genomic approach using ICG-001 identifies FOXM1 as downstream
effector of CBP-signaling in TNBC .................................................................... 61
3.4. CBP/FOXM1 bind in TNBC cells; and ICG-001 effects this binding and as well
as FOXM1 and CTNNB1 protein levels ............................................................. 64
3.5. Treatment with ICG-001 effects FOXM1 transcriptional activity and target gene
expression ............................................................................................................ 66
3.6. Gene knockdown (KD) and overexpression mimic ICG-001 effects on FOXM1
targets gene expression in TNBC cells ................................................................ 71
3.7. Canonical Wnt (TCF/CTNNB1) signaling activity is low in TNBC .................. 71
3.8. Paclitaxel treatment increases FOXM1 expression, as well as drug resistant and
stem like cancer cell populations TNBC cells ..................................................... 76
5
3.9. Treatment with ICG-001 in combination with Paclitaxel eliminates the drug
resistant CSC like cell population ....................................................................... 78
3.10. Treatment with ICG-001 reduces tumor initiation of TNBC cells in vitro........ 80
3.11. Treatment with ICG-001 sensitizes TNBC cells to Paclitaxel treatment .......... 84
3.12. Treatment of MDA-MB-468 TNBC cell line xenografts .................................. 86
3.13. Treatment of PDX of TNBCs ............................................................................ 87
3.14. Tissue micro array (TMA) covariance analysis of correlation between FOXM1
and CBP protein expression and clinical parameters .......................................... 91
4. DISCUSSION ........................................................................................................... 100
5. REFERENCE ............................................................................................................ 116

Table of Figures

Figure 1: Workflow diagram for RNA Seq analysis pipeline using Partek Flow. ....... 45
Figure 2A: Four publicly available data sets show genetic alterations in CBP in breast
cancer (BCCRC – British Columbia Cancer Research Center, TCGA – The
Cancer Genome Atlas, METABRIC – Molecular Taxonomy of Breast
Cancer International Consortium, MSKCC – Memorial Sloan Kettering
Cancer Center. …………………………………………………...………... 57
Figure 2B: CBP RNA levels in normal compared to breast cancer tissue....…...…........57
Figure 2C: CBP RNA expression in TNBC compared to other breast cancer subtypes. 57
Figure 2D: CBP, survivin and β-catenin protein levels in two TNBC cell lines (MDA-
MB-231, MDA-MB-468) and non-tumorigenic epithelial breast cell line
MCF10a (1 – cytosolic fraction, 2 – nuclear fraction)…….....………….… 58
Figure 2E: Chemical structure of ICG-001……………………………..………..……. 59
Figure 2F: Co-immunoprecipitation (CoIP) of CBP/β-catenin in three TNBC cell lines
(MDA231, MDA468 and SUM149) under DMSO vehicle control conditions
and after treatment with 20uM ICG-001 for 24h. Asterisk represents
statistical significance…………… …….……………………….......….….. 59
Figure 2G: Survivin-promoter driven luciferase reporter activity in three TNBC cell lines
(MDA-MB-231, MDA-MB-468 and Hs578T) treated for 24h with 10uM
ICG-001 or DMSO vehicle control. Asterisk represents statistical
significance (N=3 per cell line per treatment) ………………………....…. 60
Figure 2H: Western blot for survivin expression in MDA-MB-231 treated for 24h with
ICG-001 or DMSO vehicle control. ………………………………………. 60
Figure 3A: TCGA breast cancer data dataset visualization via the Santa Cruz Cancer
Genome Browser showing FOXM1 regulated gene expression in different
6
breast cancer subtypes and normal breast tissue (817 samples, 82 TNBC).
The heatmap shows individual samples in rows and genes in columns (red -
up-regulated, blue - down-regulated)….…………………..……………..... 62
Figure 3B: Functional network of FOXM1 target expression and involvement of CBP
via interaction with FOXM1………………………………………………. 62
Figure 3C: FOXM1 RNA levels in normal compared to breast cancer tissue..……...... 62
Figure 3D: FOXM1 RNA expression in TNBC compared to other breast cancer subtypes
……………………………………………………………………………... 62
Figure 4A and B: Heatmap and volcano plot showing 1300 differentially expressed genes
in RNA Seq in MDA-MB-231 treated with 10μM ICG-001 or DMSO vehicle
control (FDR ≤ 0.05, ≤ |2-fold| change)………………………………….... 63
Figure 4C: Ingenuity pathway analysis of RNA Seq data and identification of upstream
regulatory factor FOXM1 (N=2 per treatment condition)...……………….. 63
Figure 5A: CoIP of CBP with CTNNB1 or FOXM1 in MDA-MB-231 treated with
20μM ICG-001 or DMSO vehicle control (N=1)..……………………….... 64
Figure 5B: Western blot showing protein levels of FOXM1 and CTNNB1 4h after
treatment of MDA-MB-231 with ICG-001 or DMSO (N=1).…………...... 65
Figure 5C: CoIP of CBP with CTNNB1 or FOXM1 in MDA-MB-231 treated for 24h
with 20uM ICG-001 or DMSO vehicle control (N=1)…………….………. 65
Figure 5D: Western blot showing protein levels of FOXM1 and CTNNB1 after 24h
treatment of MDA-MB-231 with ICG-001 or DMSO (N=1).…….………. 66
Figure 5E: CoIP for CBP pull-down and staining for associated FOXM1 and β-catenin,
as well as FOXM1 pull down and staining for associated β-catenin (N=1).. 66
Figure 5F: FOXM1 levels in MDA-MB-468 TNBC cells after 24h treatment with 20μM
ICG-001 or DMSO control (N=1)…..……………………….……….…..... 66
Figure 6A: RNA Seq. differential gene expression (RPKM) in FOXM1 target genes in
MDA-MB-231 cells treated for 48h with either 10uM ICG-001 or DMSO
vehicle control (ICG-001 vs. DMSO, N=2 per time point per condition).... 67
Figure 6B: Expression changes (qPCR) in FOXM1 target genes in MDA-MB-231 cells
treated for 4 or 24h with either 10μM ICG-001 or DMSO vehicle control
(N=3 per time point per condition)……………………………………..…. 68
Figure 6C: FOXM1-driven luciferase reporter in MDA-MB-231 transfected for 24h and
treated for either 4 or 24h with 10μM ICG-001 or DMSO vehicle control
(bars represent relative light units (RLUs) per 20,00 cells, N=3 per time point
per condition)………………………………………………………..…….. 69
Figure 6D: FOXM1 target gene expression in three TNBC cell lines (Hs578T, MDA-
MB-436, treated for 24h with 10μM ICG-001 or DMSO vehicle control.... 70
Figure 6E: FOXM1luc
Firefly/Renilla
reporter for three TNBC cell lines (MDA-MB-231,
MDA-MB-468, Hs578T) treated for 24h with either 10μM ICG-001 or
7
DMSO vehicle control (bars represent normalized ratios of FOXM1 Firefly
luciferase to control vector Renilla luciferase expression; N=3 per condition
per cell line)……………………………………………………………....... 70
Figure 7A: Transient siRNA gene knockdown (KD) (48h) in MDA-MB-231 cells and
qPCR for FOXM1 target (N=2 per condition)……...……………………... 72
Figure 7B: FOXM1 target gene expression in MDA-MB-231 cells transiently transfected
with FOXM1
wt
overexpression or empty vector control (24h) and subsequent
24h treatment with 10uM ICG-001 or DMSO control (N=2 per condition). 73
Figure 8A: TOPFLASH as a readout for canonical Wnt signaling in TNBC cell lines.
The colorectal cancer cell lines SW480 and HCT116 served as positive
controls. Cells were treated for 6h with 10mM LiCl (bars represent ratio of
TOPFLASH to FOPFLASH luciferase RLUs, N=3 per cell line per
condition)………………………………………………………………….. 74
Figure 8B: FOXM1 target gene expression (qPCR) in MDA-MB-231 cells treated with
either 10μM ICG-001, 1μM WNT974, 1μM (red) or 10μM (green) DMSO
vehicle control for 24h (N=2 per condition)………………………………. 75
Figure 9A: Paclitaxel resistant MDA-MB-468 cell colony outgrowth after 5d treatment
with 10nM Paclitaxel………………………………………………………. 77
Figure 9B: FOXM1 protein levels are increased in Paclitaxel resistant MDA-MB-468
compared to treatment naïve cells…………………………………………. 77
Figure 9C and D: TCGA breast cancer data set showing increased FOXM1 RNA
expression (RSEM log2) in tumors treated with chemotherapy compared to
untreated tumors…………………………………………………...………. 77
Figure 9E: CSC like CD44
High
CD24Low cells in naïve vs. Paclitaxel resistant MDA-MB-
468 cells (N=3)……………………………………………………………. 77
Figure 9F: Side population (SP) cells in naïve vs. Paclitaxel resistant MDA-MB-468
cells. ……………………………………………………………………….. 77
Figure 10A: Effect of treatment with Paclitaxel plus ICG-001 on resistant cell outgrowth
compared to Paclitaxel only (N=3 per condition)………………………..... 79
Figure 10B: Effect of re-treatment of Paclitaxel resistant MDA-MB-468 (post 5d
Paclitaxel) with Paclitaxel only or Paclitaxel plus ICG-001 (N=3 per
condition)…………………………………………………………………... 79
Figure 10C: Gene expression in Paclitaxel resistant MDA-MB-468 (post 5d treatment)
treated with 10uM ICG-001 or DMSO control (N=3 per condition per time
point)……………………………………………………………………..... 79
Figure 10D: Effect of 24h 10uM ICG-001 on SP cells in MDA-MB-468, MDA-MB-231
and patient derived xenograft (PDX) TNBC cells (N=3 per cell model and
condition). Asterisk represents statistical significance (* p < 0.05, *** p <
0.001)…..………………………………………………………………...... 79
8
Figure 11A: Mammosphere formation of MDA-MB-231 cells treated for 7 days with
10uM ICG-001 or DMSO control (20,000 cells per well, N=3 per condition).
(scale bar = 100 m). Asterisk represents statistical significance (** p <
0.01)………………………………...……………………………………… 80
Figure 11B: First generation mammosphere culture of MDA-MB-231 cells pre-treated for
48h with 10uM ICG-001 or DMSO vehicle and cultured for 7 days (20,000
cells per well, N=3 per condition)………………………………………..... 81
Figure 11C: Second generation 7 days mammosphere culture of MDA-MB-231 (pre-
treated for 48h with ICG-001 or DMSO and cultured for 7 days) without
addition of treatment (20,000 cells per well, N=3 per condition). (scale bar =
100μm). Asterisk represents statistical significance (** p < 0.01)……...…. 81
Figure 11D First generation and 11E second generation 7 day mammosphere culture of
MDA-MB-231 after 48h siRNA KD (F – FOXM1, B – β-catenin, C – CBP)
(20,000 cells per well, N=3 per condition). Asterisk represents statistical
significance (* p < 0.05)………………………………………………….... 82
Figure 11F: Mammosphere formation in TNBC cell lines (MDA-MB-436, CAL51,
Hs578T and SUM149) after 7 days in culture, treated with 10μM ICG-001 or
DMSO vehicle control. (N=3 per cell line per condition). Large spheres <
50μm diameter, small sphered > 50μm (scale bar = 100μm). Asterisk
represents statistical significance (** p < 0.01, *** p < 0.001). ………….. 83
Figure 12A: Cell count after treatment of MDA-MB-231 for 48h (N=3 per condition).
Asterisk represents statistical significance (* p < 0.05)………………….... 85
Figure 12B: Live cell quantification in MDA-MB-231 after 48h siRNA KD and 48h
treatment with 10nM Paclitaxel or PBS vehicle control (N=3 per KD
condition per treatment). Asterisk represents statistical significance (* p <
0.05, *** p < 0.001). ………………………………………………...…..... 85
Figure 13A: MDA-MB-468 TNBC cell xenograft in female Nod/SCID/gamma (NSG)
mice treated for 26 days (N=5 mice per treatment condition)……...……... 86
Figure 13B: Secondary implantation of MDA-MB-468 tumors from primary xenografts.
Tumors were dissociated after termination of the experiment and implanted
into new female NSG mice without further treatment (N=5 mice per
treatment condition). …………………………………………………….... 87
Figure 14A: FOXM1 protein expression and proportion of side population cells in cancer
cells from two TNBC patient…………………………………………….... 88
Figure 14B: PDX patient 1 in female Nod/SCID/gamma (NSG) mice treated for 24 days
(N=5 mice per treatment condition)……………………………………...... 89
Figure 14C: Secondary implantation of PDX patient 1. Tumors were dissociated after
termination of the experiment and implanted into new female NSG mice
without further treatment (N=5 mice per treatment condition)……………. 89
9
Figure 14D: PDX patient 1 in female Nod/SCID/gamma (NSG) mice treated for 24 days
(N=5 mice per treatment condition)……………………………………...... 90
Figure 14E: Secondary implantation of PDX patient 1. Tumors were dissociated after
termination of the experiment and implanted into new female NSG mice
without further treatment (N=5 mice per treatment condition)…………..... 90
Figure 15: Representative cores evaluated for FOXM1 and CBP staining. Staining
scoring 0 = negative, 1 = weakly positive, 2 = strongly positive (nuclear
staining)…………………………………………………………………..... 92
Figure 16: Receiver operating characteristic (ROC) curve for FOXM1 and tumor grade
III as predictors of TNBC………………………………………………….. 95
Figure 17A: TMA Kaplan Meier overall survival analysis for FOXM1 for all cases
(N=817)…………………………………………………………………..... 97
Figure 17B: TMA Kaplan-Meier overall survival analysis for FOXM1 for TNBC cases
(N=82)………………………………………………...………………….... 97
Figure 18A: TMA Kaplan Meier overall survival analysis for CBP for all cases (N=817).
……………………………………………………………………………... 98
Figure 18B: TMA Kaplan-Meier overall survival analysis for CBP for TNBC cases
(N=82)……………………………………………………………………... 98
Figure 19A: TMA Kaplan Meier overall survival analysis for FOXM1 for TNBC cases
(N=82)……………………………………………………………...……… 99
Figure 19B: TMA Kaplan-Meier overall survival analysis for FOXM1 for all cases
(N=817)……………………………………………………………………. 99

Tables

Table 1: Cell lines and culture media. ........................................................................... 30
Table 2: Transfection conditions. .................................................................................. 32
Table 3: Genes and primer sequences for qPCR. .......................................................... 43
Table 4: Clinical variables used for association studies with FOXM1 and CBP protein
levels. .............................................................................................................. 56
Table 5: Differential FOXM1-driven gene expression in MDA-MB-231 after 48h
siRNA mediated gene knockdown.................................................................. 72
Table 6: Differential FOXM1-driven gene expression in MDA-MB-231 after 24h
FOXM1 overexpression and subsequent 24h treatment with 10μM ICG1-001.
......................................................................................................................... 73
Table 7: Differential FOXM1-driven gene expression in MDA-MB-231 after ICG1-
001 or WNT974 treatment. ............................................................................. 75
Table 8: Mammosphere size difference after 48h siRNA gene knockdown. ............... 82
10
Table 9: Mammosphere count in TNBC cell lines treated with ICG-001 or DMSO
control. ............................................................................................................ 84
Table 10: Comparison of tumor volume post 24 days; primary implantation MDA-MB-
468 xenograft (N=4 mice per group) ............................................................. 86
Table 11: Comparison of tumor volume post 24 days; secondary implantation MDA-
MB-468 xenograft (N=5 mice per treatment condition). ................................ 87
Table 12: Clinical characteristic for patient 1 and 2 used for PDX models of TNBC in
NGS mice (NeoChemo – neaoadjuvant chemotherapy, AC – Doxorubicin and
Cyclophosphamide, T – Taxol (Pacliatxel), MRM – modified radical
mastectomy) .................................................................................................... 88
Table 13: Comparison of tumor volume post 24 days; primary implantation patient 1
PDX (N=4 mice for ICG-001, N=5 per each other treatment conditions). .... 89
Table 14: Comparison of tumor volume post 24 days; secondary implantation patient 1
PDX (N=3 mice for ICG-001, N=4 per each other treatment condition). ...... 89
Table 15: Comparison of tumor volume post 24 days; primary implantation patient 2
PDX (N=4 mice each for PBS and ICG-001, N=6 mice each for Paclitaxel +
PBS and Paclitaxel + ICG-001) ...................................................................... 90
Table 16: Comparison of tumor volume post 24 days; secondary implantation patient 2
PDX (N=5 mice per group)............................................................................. 90
Table 17: Breast cancer subtype distribution .................................................................. 92
Table 18: Staining results for FOXM1 and CBP ............................................................ 92
Table 19: Statistical analysis (Chi square statistics – FOXM1 dependent variable) ...... 93
Table 20: Chi square statistics (CBP dependent variable) .............................................. 93
Table 21: Multivariate Logistic Regression (FOXM1 dependent variable) ................... 94
Table 22: Cox Regression analysis (vital status plus overall survival in months) .......... 94
Table 23: Multivariate Logistic Regression (TNBC dependent variable) ...................... 95
Table 24: TMA data set all cases mean OS based on FOXM1 expression..................... 97
Table 25: Statistical analysis of mean OS all cases based on FOXM1 expression ......... 97
Table 26: TMA data set TNBC cases mean OS based on FOXM1 expression .............. 97
Table 27: Statistical analysis of mean OS TNBC based on FOXM1 expression ........... 97
Table 28: TMA data set all cases mean OS based on CBP expression........................... 98
Table 29: Statistical analysis of mean OS all cases based on CBP expression............... 98
Table 30: TMA data set TNBC cases mean OS based on CBP expression .................... 98
Table 31: Statistical analysis of mean OS TNBC based on CBP expression ................. 98
Table 32: TCGA data set all cases mean OS based on FOXM1 expression ................... 99
Table 33: Statistical analysis of mean OS TNBC based on FOXM1 expression ........... 99
Table 34: TCGA data set TNBC cases mean OS based on FOXM1 expression ............ 99
Table 35: Statistical analysis of mean OS all cases based on FOXM1 expression ......... 99

11
1. INTRODUCTION
1.1. Breast cancer
1.1.1. Breast cancer as a public health issue
Breast cancer is the most commonly diagnosed malignancy in women with 1,700,000
patient diagnoses representing 25% of all new cancer cases in women, and more than half
a million deaths worldwide in 2012 [1]. Women in the United States have a 12% lifetime
risk of developing breast cancer [2] and invasive breast cancer is the second leading
cause of cancer related deaths second only to lung cancer [1], placing women in the U.S.
at a 3% lifetime risk for dying of breast cancer [3]. For 2017 the American Cancer
Society projects 252,710 new cases of invasive breast cancer and 40,610 fatalities in the
United State [3].
Although the current study focuses on female breast cancer, it should be mentioned that
breast cancer also occurs in men, albeit 100 times less commonly then in women.
The socioeconomic impact due to loss in productivity and health care expenses results in
costs of more than $209 billion per year in the United States alone. Breast cancer is one
of the most formidable challenges in health care today, with an urgent need for intensive
research efforts to ameliorate the suffering and costs of this devastating disease.
1.1.2. Characterizing breast cancer in the clinical setting – grading and staging
Historically, breast cancer classification in the clinic is based the modified Bloom-
Richardson-Elston histopathological grading system to determine aggressiveness [4-6], as
well as anatomical staging using features such as tumor size and spread of the cancer to
12
distant metastatic sites [7]. Breast cancer staging follows the TMN classification, which
categorizes the disease into early and late stage, depending on tumor size, lymph node
and distant organ involvement [8]. The histopathological grading distinguishes tissue
specific differentiation, number of mitoses and nuclear pleomorphism [6]. The most
commonly encountered histological type of breast cancer are carcinomas, meaning that
these cancers are of epithelial cell origin; with invasive ductal carcinomas accounting for
up to 80% of the breast cancers [9].
1.1.3. Diagnosis, treatment and biomarkers in breast cancer
The clinical management of breast cancer is based on prognostic and predictive
biomarkers. While the choice of surgical treatment of breast cancer is dependent mainly
on staging (tumor size, local and distant spread), the choice of anti-cancer drugs depends
on biological characteristics and biomarker expression.
Prognostic biomarkers determine the course of the disease at the time of diagnosis
independent of treatment strategies and include the clinical characteristics mentioned
above [10, 11], as well as proliferation markers, ethnicity and age [12-14]. Predictive
biomarkers on the other hand are patient characteristics used to determine what effect a
particular therapeutic strategy will have on treatment response [15]. In breast cancer the
by far most commonly used biomarkers are the hormone receptors for estrogen (ER) and
progesterone (PR), as well as the human epidermal growth factor receptor (HER2) [14].
ER, PR and HER2 are present in 75, 65-75 and 13-20% of breast cancers respectively
[16-18]. The expression levels of ER and PR are determined via immunohistochemistry
(IHC). HER2 levels are determined via IHC, fluorescence in-situ hybridization (FISH), or
13
both, depending on whether results of the initial test are unequivocal as defined by the
latest American Association for Clinical Oncology (ASCO) guidelines [19]. The
introduction of therapies targeting these three receptors, such anti-hormonal therapies
(e.g. tamoxifen, aromatase inhibitors) and anti-HER2 drugs (e.g. trastuzumab, lapatinib,
pertuzumab), together with improvements in surgical approaches and widespread
implementation of screening programs has revolutionized the management of breast
cancer and substantially improved the survival outcomes in patients with tumors that
harbor these markers [2, 20-23]. It is noteworthy that the assessment of all three markers
is subject to debate due to analytical variability between central and local laboratories
[24, 25]. Reproducibility issues in biomarker validation studies led to recommendations
to ensure standardization of reporting criteria [26-28]. The latest American Association
for Clinical Oncology (ASCO) guidelines recommend the use of ER, PR and HER2,
given that strict testing standards are maintained [19]. Other biomarkers (e.g. the cell
proliferation marker Ki67) can provide additional information and help determine
prognosis and therapeutic decisions [29], but are not recommended due to problems with
reproducibility (e.g. inter-rater variability, laboratory IHC variability) [19].
1.1.4. Limitation of “traditional” biomarkers
Certain subtypes of breast cancers lacking expression of ER, PR and HER2 (triple
negative breast cancer) do not respond to targeted therapies that are dependent on
expression of ER or HER2 [30]. Initially receptor positive tumors can change under the
selective pressure of therapy, or relapse with a different biology as metastatic disease [31-
34]. Overall, 20-30% of patients develop incurable metastatic disease [35-37],
14
highlighting the urgent need to expand the repertoire of predictive and prognostic
biomarkers and to develop novel methods to determine potential druggable targets,
particularly in the metastatic setting. Better matching of a drug to a patient’s disease
would improve outcomes and reduce unnecessary application of ineffective and
potentially toxic drugs. This would not only help improve outcomes for individual
patients, but also improve cost-effectiveness [38].
1.1.5. Molecular classification
Advances over the last two decades have revealed an enormous complexity and
heterogeneity of breast cancer on the molecular level. Particularly the advent of high
throughput genetic sequencing has enabled a much deeper understanding of the many
aberrations that characterize breast cancer [39-47]. In 2000, Chuck Perou and colleagues
published the first molecular subtype classification of breast cancer [48]. Analyzing gene
expression of 38 breast tumors via cDNA microarrays, the group identified four
molecular subtypes: luminal, HER2, basal-like and normal like. Further work by the
same group led to the distinction of the luminal subtype into luminal A and B [46].
Expansion of this work by several other groups confirmed the findings put forward by
these seminal studies [49-52].
Several prognostic genetic tests have been developed based on these studies and gained
FDA approval, albeit for limited indications. The Prosigna assay (NanoString
Technologies, Seattle WA), based on The Prediction Analysis of Microarray 50 (PAM50)
[50] has been validated [53] and approved to estimate recurrence free survival in early
stage ER-positive cancers treated with adjuvant anti-hormonal therapy. The MammaPrint
15
assay (Agendia, Irvine, CA) uses a 70-gene signature to differentiate low- and high-risk
patients with ER-positive cancers, and the value of adding cytotoxic therapy to anti-
hormonal treatment [54]. The Oncotype DX (Genomic Health, Redwood City, CA) uses
a 21-gene expression score to determine the 10-year distant recurrence risk after
treatment in ER-positive cancers and has been approved by ASCO for risk assessment in
ER-positive cancers only [55]. Predictive genetic tests have been developed as well, such
as the SYMPHONY (Agendia) breast cancer profile and BreastNext (Ambry Genetics,
Aliso Viejo, CA), none of which are FDA approved at this time [56]. Comparative
studies have shown that significant discordance exists when using different tests on the
same specimen [57], again stressing the importance of proper reporting criteria for
biomarkers (see REMARK guidelines [27]).
Other investigators have further expanded the breast cancer molecular subtype
classifications that have not been translated into clinical practice yet. In 2012 Curtis et al.
published a seminal paper analyzing 2000 breast tumors via genomic and transcriptional
profiling. This expansive study revealed a total of 10 so called integrative clusters based
on DNA-RNA profiles [41]. Using publicly available micro array data sets from 21
studies (14 for the training set, and 7 for the validation set), as well as multiple TNBC
cell lines, Lehmann et al. showed that basal-like TNBC can be distinguished into distinct
subtypes based on gene expression signatures, with consequences for prognosis and
treatment response [58, 59].
The advent of “genomic” medicine and detailed molecular characterization of breast
cancer holds great potential to improve diagnostic accuracy and facilitate the choice of
efficient treatment option, based more on the specific biological characteristics of an
16
individual patient’s cancer, rather than the limited subset of biomarkers used to date. The
potential benefit of this approach warrants further expansion and validation of “genetic
biomarkers” in clinical trials.
1.1.6. Triple negative breast cancer
The subset of breast cancer that is negative for all three markers is called TNBC,
representing 10-20% of all breast cancers [60-66]. Compared to receptor positive breast
cancers, TNBCs occur more frequently in younger, pre-menopausal women, and show a
propensity for certain ethnicities, such as African-American, Hispanic and Caribbean
women [64, 67-70]. About 95% of TNBCs are invasive ductal carcinomas with few
distinctive histological characteristics [71], but can be distinguished by the expression of
basal cytokeratins (e.g. CK8 and 18) and high expression of the proliferation marker
Ki67 [71, 72]. These cancers are characterized by aggressive growth, higher rates of
recurrence (locoregional or distant) – most likely within the first 5 years of diagnosis –
and poor 5-year survival rates [65, 73]. The highly aggressive course compared to other
subtypes makes this subtype responsible for a disproportionate number of breast cancer
deaths [62, 74, 75]. Due to the lack of receptors (ER and HER2) for targeted approaches,
non-targeted cytotoxic chemotherapy targeting proliferating cells (mostly taxane and
anthracycline based) are the mainstays in the clinical management of TNBC [76-78].
TNBCs show the high initial response rates to chemotherapy, particularly in the neo-
adjuvant (before surgery) setting [77, 79-82]. Unfortunately, treatment effects are
oftentimes short-lived, resulting in shortened disease-free interval [83], while
significantly reducing quality of life due to severe toxic side effects [84]. Once TNBC
17
recurs as metastatic disease, less than 30% of women survive past 5 years, and virtually
all patients succumb to their disease [85].
Detailed molecular characterizations of TNBCs using “omics” analyses have greatly
enhanced our understanding of the heterogeneity within this breast cancer subtypes. RNA
studies demonstrated that most TNBCs have basal-like gene expression patterns [86, 87],
and over 90% of basal-like breast cancers are TNBC [88]. A groundbreaking study by
Lehmann et al. led to further granularity in the molecular characterization of TNBC.
Using in silico analysis, the group identified four distinct subtypes of TNBC based on
gene expression (basal-like 1 and 2 – BL1 and BL2, mesenchymal – M and luminal
androgen receptor – LAR) [58, 59]. A comparison with the PAM50 classifier showed that
roughly 80% of the tumors analyzed corresponded to the basal-like breast cancer subtype,
with very high overlap in BL1 (99%), BL2 (95%) and M (97%) [89]. In vitro studies in
cell lines representing all subtypes demonstrated therapeutic implications of this novel
classification [58]. Retrospective analysis of clinical trial data established the potentially
predictive value of this classification in response to neoadjuvant therapy in a subset of
patients [90]. Given the finding of androgen receptor expression in TNBC in these and
other studies [58, 59, 91], several trials testing anti-androgen therapies in TNBC are
underway [92]. The study by Curtis et al. mentioned previously, which subdivided breast
cancer into 10 integrative clusters based on gene-expression and mutations, showed that
basal-like cancer heterogeneously fall into clusters 4 and 10, with implications for patient
prognosis [41].
DNA-sequencing studies showed that TNBCs exhibit potentially actionable mutations
with subtype specific prevalence. For example, TP53 has the highest mutation rate in
18
TNBCs globally, but is twice as common in basal-like TNBC [40, 93]. PIK3CA
mutations are found in about 10% of TNBC, with a 10-fold higher prevalence in luminal
type (LAR) TNBC [94]. Interestingly, cell lines harboring PIK3CA mutations are
sensitive to anti-androgen treatment [58, 95], and potential benefit of either single agent
PIK3CA inhibition (NCT01623349) or dual blockade is being explored in a clinical trials
(NCT02457910). Other potentially actionable targets (e.g. KRAS, BRAF, EGFR and
MET) have been identified with large degrees of variation [40, 41, 93]. Mutational
studies revealed that roughly 10-20% of TNBCs exhibits BRCA1/2 mutations [40, 93,
96, 97]. Both genes are considered tumor suppressor genes that are critical for the repair
of DNA double strand breaks [98]. The term “BRCAness” has been applied to tumors
with mutated BRCA1/2 and is highly correlated with basal-like TNBCs, with
approximately 70% of BRCA1 and 20% of BRCA2 mutated breast cancers are being
TNBC [99]. The finding of DNA repair deficiency led to the implementation of DNA
damaging agent (e.g. platinum based drugs) as well as PARP inhibitors (which lead to
“synthetic lethality” by causing double strand breaks [100, 101] for BRCA mutant
patients – with mixed results in early (phase II) trials [102, 103]. Several phase III trials
are ongoing in neo-adjuvant and metastatic settings listed on
https://www.clinicaltrials.gov (NCT02000622, NCT02163694, NCT019005592,
NCT01945775).
Studies showed that tumors deficient in DNA damage repair (e.g. BRCA-mutant tumors)
are more susceptible to immune modulatory therapy, such as PD-1 blockade [104]. Basal-
like and TNBC exhibit disproportionately high expression of PDL1 of ~20% [105, 106].
Two early trials (phase I) showed ~19% response rates [107, 108]. A potential
19
explanation for the relatively low response rates could be the significant assay and
scoring heterogeneity for PDL1 [109], again emphasizing the urgent need for
standardization of potential biomarkers.
1.2. Cancer stem cells (CSCs)
The cancer stem cell concept postulates that tumor evolution follows a similar concept as
organismal development, where a group of stem-like cells form the “root” of tumors and
give rise to the bulk of rapidly proliferating, aberrantly or only partially differentiated
tumor cells [110-113]. CSCs are thought to be responsible not only for tumor initiation
and maintenance, but also drug resistance and metastasis, characteristics that established
these cells as an attractive target for cancer therapy [114, 115].
The term CSC is less an indicator of the cell of origin as it is descriptive of functional
properties that resemble stem cells [116]. For example, CSCs express telomerase for
replicative immortality [117, 118]. They are capable of symmetric or asymmetric cell
division, allowing them to self-replicate, or give rise to more differentiated progenitors,
respectively [119].
The origin of CSCs has been controversial, with speculations about their origin ranging
from transforming mutations in tissue stem cells to cell fusions and cell de-differentiation
[120]. Mutations that block differentiation (e.g. BRCA1) have been identified in breast
cancer [121, 122], and have been linked to stem-like cells of origin in breast cancer [123].
De-differentiation of epithelial cells occurs when these cells undergo a process of
epithelial-to-mesenchymal transition (EMT), which in cancer cells is associated with the
20
acquisition of CSC characteristics [124-126]. Indeed, several EMT-related genes have
been found to be de-regulated in cancers [127].
A hallmark of CSCs is their capability to initiate tumors from single cells or very low cell
numbers compared to “bulk” tumor cells [128], and the terms tumor initiating cells
(TICs) and CSCs have been used interchangeably. The tumor initiating capacity of single
cancer cells was first postulated 80 years ago, when Furth and Kahn described the
transmission of leukemia from a single cell in mice [129]. John Dick’s group revived the
concept of rare cancer initiating cells about 20 years ago, again in leukemia [130, 131].
CSCs/TICs have subsequently been described in numerous solid malignancies [128, 132-
135]. The identification of CSCs is based on cell surface markers or marker combinations
[136], such as CD24 and CD44 [137, 138], CD133 [135, 139], CD34 and CD38 [130],
CD90 [140, 141], aldehyde dehydrogenase (ALDH) [142]. CSCs can also be identified
by the expression of ATP-binding cassette transporters (e.g. MDR1, or ABCG2), which
can transport various substrates across cell membranes [143, 144]. These proteins
function as drug efflux pumps, rendering CSCs resistant to conventional chemotherapy
[145-147].
Al-Hajj et al. first described CSCs cells in breast cancer in 2003 as a CD44
high
CD24
low

cell population [137]. Subsequent studies demonstrated the presence of CSC/TIC
populations in many breast cancer cell lines [148]. More aggressive, poorly differentiated
basal-like breast cancers have been shown to contain higher percentages of CSC
populations [149] and exhibit stem-like gene expression signatures [150, 151], which are
related to poor outcome [152]. The process of EMT has also been linked to basal like
21
breast cancers [153, 154], aggressive clinical behavior and increased metastatic potential
[155-157].
Insights from studies in developmental and cancer biology revealed that developmentally
conserved molecular pathways regulating physiological stem cell biology also govern
CSC behavior, such as Hedgehog [158], Notch [159] and TGF-b [160], and perhaps most
prominently the Wnt signaling pathway [153, 161].
1.3. Molecular driver of stem cell biology – The Wnt signaling pathway
The Wnt signaling pathway is an evolutionarily conserved signaling cascade that has
fundamental importance in organismal development and tissue homeostasis [162]. Wnt
signaling plays a pivotal role in early embryonic development, such as axis polarization
[163], as well as adult (i.e. tissue) stem cell behavior [164]. Pioneering work by Clevers
et al. demonstrated the importance of Wnt signaling for the maintenance of stem cells in
the small intestines [165]. Many subsequent studies, established the importance of Wnt in
the regulation of stem cell biology and regeneration in multiple organ systems [164, 166].
Deregulation of Wnt signaling has been implicated in malformations as well as other
forms of disease, particularly cancer [167-169]. The so-called canonical Wnt pathway is
mediated through the transcriptional activity of β-catenin [170]. Without Wnt signaling,
cellular β-catenin levels are kept at relatively low levels through phosphorylation by a
multi-protein destruction complex (consisting of GSK-3β, APC and Axin) and
subsequent degradation via the proteasome [171-175].
Wnt signaling is activated through the binding of a family of 19 Wnt glycoproteins [169]
to a family of seven so called frizzled (Fz) transmembrane cell surface receptors [176,
22
177]. Further binding of co-receptors, such as low-density lipoprotein (LDL) receptor-
related protein (LPR5/6), is required to fully activate the Wnt signaling cascade [178-
181].
This receptor activation leads to the recruitment of Dishevelled (Dsh) to the cell
membrane where it binds to and sequesters Axin, leading to the disassembly of the
destruction complex [182]. Consequently, β-catenin remains un-phosphorylated and
accumulates in the cytoplasm, leading to the translocation of β-catenin to the cell nucleus,
where it forms transcriptional complex with transcription factors (TF) and co-activator
proteins. Usually, canonical Wnt signaling is defined by the binding of β-catenin to TFs
of the TCF/lef family [183, 184], but numerous other β-catenin binding factors have been
identified (e.g. Forkhead transcription factors (e.g. FOXM1), Nuclear Receptors, Sox,
Smad, Oct4), many of which play important roles in stem cell biology. As mentioned, the
transcriptional complex is only complete with the addition of co-activators, an important
representative of which are CREB-binding protein (CBP), and E1A-associated cellular
p300 (EP300).
1.3.1. CREB-binding protein (CBP)
CBP is a large (~300 KDa) multi-domain proteins that functions primarily as a histone/
lysine acetyl transferase (HAT, KAT) (e.g. H3K122ac, H3K27ac, H3K4me1) and
chromatin remodeler [185, 186], as well as acetyltransferase for non-histone targets (e.g.
p53, GATA-1) [187, 188]. CBP also has important functions as scaffolding protein in
various transcriptional complexes [189-193]. It is involved in G-protein coupled
23
signaling via cyclic adenosine monophosphate (cAMP) regulated gene expression and
binding to phosphorylated CREB protein [194-196].
Physiologically CBP is involved in cellular proliferation, cell cycle regulation, apoptosis
and differentiation [189, 197-200]. CBP has also been involved in DNA damage repair
via their interaction with BRCA1 [201]. Genetic knockout studies demonstrated that
homozygous of CBP results in embryonic lethality and phenotypic defects, and revealed
haplo-insufficiency of CBP heterozygous deletion [202, 203]. Several studies
demonstrated non-redundant functions for CBP in organism development [200, 202-205].
We and others have shown that CBP has a critical role in stem cell biology, such as
embryonic stem cell pluripotency [206] and hematopoietic stem cell differentiation and
self-renewal [207, 208]. Germline mutations in CBP lead to a rare developmental
disorder in humans called Rubinstein-Taybi syndrome [209]. Somatic mutations in CBP
associated with cancer are rare but have been described [210-214].
As an important co-activator in the Wnt signaling pathway, CBP binding to β-catenin has
been shown to play a critical role in cancer etiology and the occurrence of cells with CSC
phenotype [215, 216].
1.3.2. Wnt and cancer stem cells
The intimate connection between Wnt signaling and CSC biology is further strengthened
by findings that several stem cells markers are Wnt targets (e.g. CD44 [217], CD24
[218], CD133 [219], MDR1/ABCB1 [220, 221] and EpCAM [222]).
Wnt is an important regulator of EMT [223] and several EMT genes that confer CSC
properties are Wnt targets [125]. Other stemness and CSC-related genes, such ID2 and
24
Cdx-1, are β-catenin driven [224, 225]. Telomerase expression is Wnt-dependent [226].
Re-location of β-catenin to the nucleus increases CD44
high
CD24
low
cell population [227].
Wnt/β-catenin signaling important for CSC in leukemia [161], and the Kahn lab has
shown that inhibiting CBP/β-catenin signaling can eliminate leukemic stem cells [215].
Finally, it has been shown that inhibition of Wnt signaling effects metastasis and cancer
cell phenotype in breast cancer [228].
1.3.3. Wnt and breast cancer
Numerous studies have established the important role canonical Wnt signaling in cancer
development and progression [229, 230]. One of the first transgenic mouse models for
mammary tumors was the MMTV-Wnt1 mouse, which established that ectopic
expression of the Wnt-1 proto-oncogene induces adenocarcinomas in mice [231, 232].
Interestingly, these tumors have later been found to exhibit a stem cell-like gene
expression signature [233].
De-regulation of Wnt signaling components is a common phenomenon in breast cancer.
For example, 10% of mice deficient in APC (APC
Min/+
) develop spontaneous mammary
cancers, and 90% of these mice are susceptible to carcinogen-induced mammary cancer
[234]. LOH (23-40%), mutation (6-18%), and hypermethylation of APC lead to a loss of
expression in approximately 36-50% of breast tumors, and has been described as a
potential prognostic biomarker [235-237]. Down-regulation of other negative regulators
of Wnt, such as secreted FRP1 (SFRP1), is one of the most frequent alterations in breast
cancer [238]. The positive Wnt-regulatory protein disheveled (Dsh) is amplified or up
regulated in 50% of ductal breast cancers [239].
25
Studies have demonstrated crosstalk between the Wnt pathway and estrogen receptor
(ER), as well as epidermal growth factor receptor (EGFR/HER2) signaling [240, 241].
Estrogen increases survivin, a bona-fide Wnt target gene [230, 242, 243]. Survivin has
been linked to teratoma formation in embryonic stem cells [244] and has been suggested
to be a prognostic biomarker in breast cancer, although with conflicting results [244-246].
HER2 signaling has been demonstrated to induce and promote mammary tumors via
interaction with β-catenin [247]. Wnt/β-catenin antagonist targets CSC in HER2 breast
cancer [248].
Wnt activity, as measured by nuclear β-catenin, is up-regulated in aggressive basal type
tumors [249-251], which as mentioned above are intricately linked to CSC biology.
Interestingly, large genomic studies in breast cancer identified FOXM1, a known binding
partner of β-catenin, as an important driver in basal type and TNBC [40].
1.4. FOXM1 in physiology and disease
The Forkhead (FOX) family encompasses evolutionary conserved transcriptionally active
proteins with a conserved DNA binding winged-helix consensus domain (TAACA) [252-
254]. In humans, over 19 subfamilies (FOXA-R) have been described, each of which can
contain several subfamily members (indicated by numbers 1, 2 and so forth), bringing the
number of known human FOX genes to 50 [255].
FOX proteins have important physiological functions in embryonic development [256,
257], organ function [258, 259], regeneration [260], the immune system [261], apoptosis
and longevity [262-264].
26
FOX-family members also play an important role in disease development [265]. In
cancer they have been found to either function as bona fide tumor suppressors (e.g.
FOXO) [266] or oncogenes, with one of the most prominent members being FOXM1
[267].
FOXM1 was first identified in HeLa cells [268]. Its genetic locus has been mapped to
chromosome 12p13 [254], consisting of 10 Exons (I-X), of which Va (A1) and VIIa (A2)
are alternatively spliced to create three variants (FOXM1a, b and c). Post-translational
modifications include phosphorylation at Thr residue 596, which is important for the
subsequent recruitment of CBP/p300 as transcriptional co-activators [269]. A study by
Halasi et al. suggested that FOXM1 increased its own expression via a positive feedback
loop, binding its own promoter, although the exact mechanism remains elusive [270].
FOXM1 is critically important for cell cycle progression [271], in particular G1-S and
G2-M phase transitions [272, 273]. Knockout studies in mice demonstrated that FOXM1
-
/-
is embryonically lethal due to mitotic failure and structural defects [257, 274, 275]. It is
expressed in most embryonic tissue and cycling cells, and only low or absent expression
in quiescent or terminally differentiated cells [271, 276-278].
Transcription factors are overrepresented among oncogenes [279], making them a
desirable target for cancer therapy [280]. FOXM1 is one of the most commonly up-
regulated transcripts in many cancer types [281-285] and has been identified as early
marker in cancer development [286], including breast cancer [287]. It has been shown to
play a key role in cancer progression and therapy resistance [288-290], including many
commonly used standard of care (SOC) non-targeted drugs such as platinum-based
compounds [291], anthracyclines [292] and taxanes (in particular Paclitaxel) [293]. Study
27
investigating the molecular basis of resistance to compounds targeting HER2-
overexpressing (e.g. anti-EGFR target drugs such as lapatinib, trastuzumab) and ER-
positive breast cancers have revealed an important role of FOXM1 in the development of
drug resistance to these target compounds [293, 294].
FOXM1 was implicated in conferring a CSC phenotype (self-renewal capacity,
replicative immortality) [295-297]. It is an important factor in cancer cell metastasis
[298-302], and has been shown to regulate the expression of metastasis related genes
(such as MMP’s) [303, 304].
In summary, the above-described findings have placed FoxM1 into the spotlight of
cancer research. FOXM1 has been called the “Achilles’ heel” of cancer and was named
“2010 Molecule of the Year” due to its considerable potential as a critical target in cancer
therapy [305] (International Society for Molecular and Cell Biology and Biotechnology
Protocols and Research 2010 Molecule of the Year).
1.4.1. FOXM1 and breast cancer
Amplification of the FOXM1 locus is frequently observed in breast cancer [306].
FOXM1 activity is intricately linked to hormone receptor signaling [307] and regulation
of ER expression [308], while ER directly regulates FOXM1 expression [309, 310].
Resistance to target therapy in hormone receptor positive and/or HER2 over-expressing
tumors has been correlated with high expression levels of FOXM1 [293, 294, 310]. A
study by Park et al. showed that FOXM1 is among the highest ranked survival-associated
factors in patients with breast cancer [289]. A meta-analysis of 36 clinical studies
encompassing nearly 5000 patients with malignant solid tumors by Dai et al. revealed a
28
certain discrepancy regarding the prognostic value of FOXM1 as a cancer biomarker
[311]. FOXM1 expression levels were correlated with larger tumor size, lymph node
metastasis and higher stage [312]. Subtype specific analysis show that over-expression of
FOXM1 as an independent factor of poor prognosis only ER-positive tumors [294, 311,
313]. Notwithstanding these findings, overexpression of FOXM1 in TNBC has been
demonstrated by other investigators [40, 314], and has been implicated in the
development of basal like BC [315, 316].
1.5. Convergence of FOXM1 and Wnt signaling in cancer and CSC biology
FOXM1 and Wnt signaling both play a critical role in stem cell biology (see individual
chapters). The over-expression and -activation of both FOXM1 and Wnt/β-catenin
signaling is also intricately linked to cancer biology [229, 301, 305, 317-319].
Many studies have demonstrated that both Wnt signaling and FOXM1 are linked to cell
migration, invasiveness, as well the proliferation of cancer cells, CSC phenotype and
self-renewal [294, 320-322].
A study in 2011 by Zhang et al. demonstrated that canonical Wnt signaling and FOXM1
are intricately linked in glioma tumorigenesis [297, 323]. These authors demonstrated
that Wnt ligands (e.g. Wnt3a) stabilized FOXM1, leading to the conclusion that FOXM1
might be a direct downstream target of Wnt signaling. Furthermore, these authors found
that FOXM1 is necessary for the nuclear translocation of β-catenin and is recruited to the
TCF/LEF/β-catenin transcriptional complex, resulting in increased target gene
expression.
29
At this point it remains unclear whether active canonical Wnt/β-catenin/TCF signaling is
required for FOXM1 expression, whether Wnt/β-catenin and FOXM1 signaling converge
in breast cancer, and what role other components of the Wnt signaling pathway, such as
co-activator proteins (e.g. CBP), play in this complex signaling network.

Hypothesis
The current study hypothesized that CBP signaling plays an important role in TNBC
biology and may provide a novel therapeutic target.

2. METHODS
2.1. Cell culture
All cell lines and appropriate culture media composition are listed in Table 1. SUM149
was obtained from Asterand Bioscience (Detroit, MI). CAL51 was obtained from The
Leibniz Institute DSMZ - German Collection of Microorganisms and Cell Cultures
GmbH (Braunschweig, Germany). MDA-MB-231, MDA-MB-468 and Hs578T were
obtained from the American Type Culture Collection (ATCC) (Manassas, VA). Cell
cultures were maintained and propagated as regular 2D-cell adherent monolayer cultures
in T25, T75 or T175 cell culture flasks (Corning Inc., Corning, NY) to minimize chances
of contamination. Before experimental studies, the cells were passaged into 6-well, 24-
well or 96-well culture plates (Corning Inc.) to facilitate subsequent treatment and access
30
with cell scrapers at the harvesting time point. All culture vessels were kept in a
humidified incubator at 37°C and 5% CO2.
For 3D-suspension mammosphere culture, the 20,000 cells per well were re-suspended in
Mammocult culture medium (STEMCELL Technologies, Vancouver, Canada) and
placed into ultra-low adherence 6-well culture plates (Corning Inc.). The plates were
carefully placed in a cell culture incubator without any further disturbance for up to 7
days to allow sphere formation.

Table 1: Cell lines and culture media.
Cell line Culture medium
MDA-MB-231
Hs578T
MDA-MB-436
SUM149
Dulbecco's Modified Eagle Medium (DMEM)
(Thermo Fisher Scientific, Canoga Park, CA) +
10% FBS (Atlanta Biologicals, Flowery Branch,
GA) + 1% Antibiotic-Antimycotic mix (Thermo
Fisher Scientific)
MDA-MB-468
1:1 mixture of DMEM and Ham’s F12 + 10% FBS
+ 1% Antibiotic-Antimycotic
CAL51 DMEM + 20% FBS + 1% Antibiotic-Antimycotic

2.2. Cell passaging
At about 80% confluence, culture medium was siphoned off using a vacuum pump
system in a laminar flow hood. The cell layer was rinsed once with PBS (Thermo Fisher
Scientific (Canoga Park, CA) and subsequently incubated with 0.025% EDTA/ trypsin
solution (Thermo Fisher Scientific) for 5min at 37°C in a cell culture incubator. After
ensuring cell detachment under an inverted phase contrast microscope (Axiovert 200,
31
Zeiss, Oberkochen, Germany) the digestion was stopped by adding complete cell culture
medium at a ratio of 2:1 to the cell suspension. The cell suspension was transferred to
15mL or 50mL conical tubes (BD Biosciences, San Jose, CA) and spun for 5min at
1000rpm. Supernatant was removed without disturbing the cell pellet and fresh culture
medium was added to distribute the cells into new cell culture vessels. Usually cells were
split at a ratio of 1:4 for propagation. The culture flasks or Petri dishes containing freshly
passaged cells were placed in an incubator and not disturbed for at least 24h to guarantee
proper cell attachment. 3D-suspension cultures were first collected and centrifuged at
300rpm for 5min to pellet spheres and remove cell culture medium supernatant. The
pelleted spheres were incubated with Trypsin/ EDTA and re-suspended as described
above.
2.3. Transient transfection
For transient transfection of the cell lines, Lipofectamine 2000 Reagent (Thermo Fisher
Scientific) was used for plasmid DNA, and Lipofectamine RNAiMAX (Thermo Fisher
Scientific) was used for siRNA-mediated transient gene knockdown according to the
manufacturers protocol. Table 2 lists the amount of transfection reagent, plasmid DAN
and culture medium used. Before transfection, cells were seeded at 3×10
5
cells per well in
6-well plates or 1×10
5
cells per well in 24-well plates and incubated at 37°C and 5% CO2
overnight. To prepare the transfection mix, two separate 1.5mL Eppendorf tubes (Thermo
Fisher Scientific) were prepared with Opti-MEM (Thermo Fisher Scientific) transfection
medium. Plasmid DNA or siRNA and Lipofectamine were added to individual tubes and
pipetted several times to ensure homogenous mixing. Both tubes were combined
32
thereafter and incubated for 5min at room temperature to allow DNA/Lipofectamine or
siRNA/Lipofectamine complex formation. The mix was added to the cells and incubated
for up to 48h.

Table 2: Transfection conditions.
Component per well 24-well plate 6-well plate
Lipofectamine 2000 2.5μL 6μL
Plasmid DNA 500ng 2500ng
Lipofectamine RNiMAX 1.5μL
siRNA 25pmol

2.4. TOPFLASH/FOPFLASH assay
The TOPFLASH (TCF Reporter Plasmid) contains several TCF binding sites, which
drive a luciferase expression when bound by active TCF transcription factor complexes
[184], and serves as readout for the activation of canonical Wnt signaling. The
FOPFLASH construct serves as a negative control, where the TCF binding sites have
been mutated [324]. Both constructs were a kind gift of the Moon lab at the University of
Washington (Seattle, WA). Cells were plated in 6-well plates at 3×10
5
cells per well and
transfected using Lipofectamine 2000 as described above. After 24h, the cells were
passaged and seeded at 20,000 per well into opaque 96-well plates (BD Biosciences), and
allowed to adhere for 24h hours. On the day of assay performance, culture medium was
removed and 50μL per well of fresh medium was added. To stimulate Wnt signaling,
lithium chloride (LiCl) is added at a concentration of 10mM to both TOPFLASH and
33
FOPFLASH transfected cells for 6h before the assay. Non-stimulated cells were prepared
in parallel for both constructs, which received culture medium only, without the addition
of LiCl. After incubation (ca. 48h post transfection), 50μL of Bright-Glo (Promega,
Madison, WI) was added to each well and the samples were incubated for 5-10min on a
plate rocker. The signal was read using a SpectraMax M3 spectrophotometer (Molecular
Devices). All samples were prepared in triplicate, unless otherwise specified.
2.5. Survivin and FOXM1 reporter assay
For quantification of survivin expression and FOXM1 transcriptional activity, the TNBC
cell lines MDA-MB-231, MDA-MB-468 and Hs578T were transfected for 24h with
pGL3b-6270 survivin promoter Firefly luciferase reporter construct [243] or pGL3b-
FOXM1-luciferase reporter [272], respectively. For background and transfection
efficiency normalization, cells were simultaneously transfected with pRL3 Renilla
luciferase control vector (Promega, Madison, WI). The next day, the transfected cells
were passaged and plated at 20,000 cells per well in opaque cell culture grad 96-well
plates (Corning Inc.). After a 24h adherence period, fresh culture medium was added to
each well containing either 10μM ICG-001 or DMSO vehicle control. The cells were
treated in triplicate for 24h and luciferase activity quantified using the Dual-Glo
Luciferase Assay System (Promega) according to the manufacturer’s instructions. The
luciferase signal was read using a PerkinElmer EnVision Multilabel plate reader
(PerkinElmer, Waltham, MA).

34
2.6. FOXM1 overexpression
A pCDNA3-Flag-HA-FOXM1 or empty vector control were transfected into cells as
described above using Lipofectamine 2000. After 24h transfection medium was replaced
with regular culture medium containing the appropriate treatment compound for down-
stream analysis.
2.7. siRNA mediated gene knockdown
All siRNAs transient gene knockdowns for FOXM1, CBP, β-catenin and negative control
were performed using Silencer Select siRNAs (Thermo Fisher Scientific) and
Lipofectamine RNiMAX (Thermo Fisher Scientific). Samples were prepared in duplicate
as described above. As a control for transfection efficiency, the cells were simultaneously
transfected with BLOCK-iT Alexa Flour Red Flourescent Oligo (Thermo Fisher
Scientific) and red fluorescence expression was validated after 24h transfection using a
Zeiss Axiovert 200 inverted microscope. Gene knockdown was validated via qPCR.
2.8. Animal experiments
Female NOD scid gamma mice (NOD-scid IL2Rgamma
null
, The Jackson Laboratory)
were used for all xenograft experiments. The animals were housed at the USC Zilkha
Neurogenetic Institute vivarium, according to IACUC requirements (protocol number
11204). Breeding colonies were established to supply animals for experimental studies in
house. Female mice between 2 and 4 months of age were used for all xenograft
experiments. The animals were randomized to treatment groups before the initiation of
any experimental procedure.
35
For all invasive procedures, including tumor cell injection, tumor piece and osmotic
pump implantation, the mice were anesthetized with 2.5-4% Isoflurane (Santa Cruz
Biotechnology) under continuous infusion via a nose cone. A pain stimulus (pinching the
mouse foot) was used to test reactivity of the mouse and ensure proper anesthesia. Prior
to injections or implantations, the graft site was shaved and disinfected using alcohol
swaps to facilitate access and prevent infections. After each procedure, the mice could
recover and remained under observation until they reach full consciousness and
ambulation (about 5-10min).
2.8.1. Tumor cell line subcutaneous (s.c.) injection
For cell line xenograft experiments, MDA-MB-468 cells were prepared at a concentration
of 1×10
6
cells per 100μL in a 1:1 mix of 75μL cell culture medium (DMEM/ F12) and
75μL Matrigel basement membrane (BD Biosciences), for a final volume of 150μL
slurry. The slurry was aspirated into hypodermic syringes (BD Biosciences) and placed
on ice until injection. After the mice reached a sufficiently deep level of anesthesia (no
pain reaction, deep steady breathing as judged by thorax movement), cells were injected
subcutaneously into the left hind flank region. Using a pair of forceps, a small skinfold
was grabbed and lifted to prevent intramuscular injection. A 100μL bolus was injected
into an individual mouse, and the puncture site was sealed using degradable tissue
adhesive (3M Health Care, St. Paul, MN) to prevent leakage.

36
2.8.2. Patient tumor acquisition and PDX establishment
With the permission of the Internal Review Board (IRB) of the University of Southern
California, surgical specimens were obtained from patients with triple negative breast
cancer. Immediately after surgery, the tumor specimens were harvested into ice cold
DMEM supplemented with 10% FBS and 1% Penicillin Streptomycin (P/S) (Thermo
Fisher Scientific), placed on ice and transported to the laboratory for immediate
processing. All processing of the specimen was done in a laminar flow hood to minimize
the chance for contamination. Specimens were divided, depending on the overall size,
into samples for direct implantation as well as storage of p0 (no passage in mouse). For
storage, the samples were cut into 3mm
3
pieces and placed into freezing medium (90%
FBS plus 10% dimethyl sulfoxide (DMSO) (Sigma-Aldrich, St. Louis, MO)). The
samples were then gradually frozen using Mr. Frosty freezing container (Thermo Fisher
Scientific) overnight at -80°C, and subsequently transferred to liquid nitrogen for long-
term storage. The remaining sample pieces were placed into Matrigel basement
membrane (BD Bioscience) and kept on ice until implantation. Tumor pieces were
implanted into the upper mammary fat pad region of at least one female NSG mice, either
uni- or bilaterally, depending on the number of tumor pieces available. All surgical
procedures on the mice were performed under anesthesia as described above. A skin
incision was made using surgical scissors, and wound size was kept as small as
implantation allowed (ca. 7mm length). The tumor pieces were removed from Matrigel
using forceps and inserted subcutaneously into to incision site. Three sutures using 6-0
silk filaments (Ethicon Inc., Somerville, NJ) were placed for wound closure. Tissue
adhesive (3M Health Care) was applied over the incision to ensure complete wound
37
closure. On average, visible tumor outgrowth for successful xenografts were observed
after 1-3 months’ post implantation.
2.8.3. Xenograft line propagation
After tumors reached 100mm extension in the largest axis, mice were sacrificed and
tumors removed under sterile conditions. The tumors were again cut into 3mm
3
pieces
and implanted, as well as viably stored as described above. Each subsequent generation
of xenografts received a passage number (e.g. p1m – first passage in mice, p2m – second
passage in mice, etc.).
2.9. Xenograft cell preparation for in vitro assays
Explanted tumor xenografts were subjected to a dissociation protocol to obtain single
cells for in vitro culture or cell characterization. All steps were performed under sterile
conditions in a laminar flow hood. Immediately after removal from the mouse body
tumor tissue was transferred to 50mL conical tubes containing 15ml pre-chilled complete
culture medium (DMEM + 10% FBS + 1% P/S) and placed on ice. Digestion medium
was prepared freshly each time, consisting of DMEM + 10% FBS + 1% P/S, 200U/ml
Collagenase Type IV (Thermo Fisher Scientific), 0.6U/ml Dispase (STEMCELL
Technologies). Initially the tumor pieces were placed into a p100 culture dish (Corning
Inc.), covered with 1-2mL digestion medium and mechanically minced with razor blades,
to yield pieces of roughly 1mm
3
. The minced tissue and digestion medium were
transferred to a 50mL conical tube containing 10-15mL digestion medium and placed in a
37°C water bath for 30min. The tissue slurry was triturated every 5-10min using 5mL
38
serological pipettes (Thermo Fisher Scientific). The progress of cell dissociation was
evaluated by placing a small aliquot of tissue digest onto a hemocytometer and
visualization under a Zeiss Axiovert 200 phase contrast microscope. After incubation, the
tissue digest was passed through 100μM cell strainers (BD Biosciences) and spun down
at 2000rpm for 5min at 4°C. Two washing steps with 30mL complete culture medium
were performed to remove all enzymes. Due to relatively high rates of cell death (up to
90% of cells), a DNAse I (0.1mg/ml) (Thermo Fisher Scientific) treatment was
performed for 10min at room temperature to remove cell clumps. Following another
washing step with complete culture medium, the cell solution was passed through 70μM
cell strainer (BD Biosciences) and centrifuged at 1000rpm for 5min at 4°C. A viable cell
count was obtained using a hemocytometer and Trypan Blue (Thermo Fisher Scientific)
dye exclusion.
2.10. In vitro and in vivo treatment procedures
2.10.1. Drug preparation and in vitro application
ICG-001 was prepared as 100mM stock solution in DMSO (Sigma-Aldrich) and stored at
-20°C. ICG-001 was used at a final concentration of 10 or 20μM. Paclitaxel
(SupremeMed, Van Nuys, CA) was obtained from the USC Norris Comprehensive
Cancer Center pharmacy, in vials containing 6mg/mL, and kept at 4°C in the dark.
Treatment of Paclitaxel was done at a final concentration of 10nM or 1μM. For vehicle
control, equimolar amounts of DMSO were added to control dishes for each experiment.
All treatment and control incubation were performed using the same complete culture
39
medium used for every culture system. Medium was changed every other day until the
desired treatment time was reached.
For treatment of cells in 3D-suspension culture, all compounds were added upon
distribution of the cell into ultra-low adherence plates (Corning Inc.). Cells were
incubated in treatment medium for up to 7 days, before samples were harvested or fresh
medium was added. In vitro treatment was performed in biological triplicates, until
otherwise specified.
2.10.2. Establishing Paclitaxel resistant triple negative breast cancer cells
MDA-MB-468 triple negative breast cancer cells were treated with 10nM Paclitaxel or
vehicle control for up to 6 days. After initial cell death of most cells, drug resistant cell
colonies emerged after 5-6 days’ post treatment.
2.10.3. In vivo treatment of mice bearing xenografts
The treatment of xenograft bearing mice was initiated after palpable tumors head
developed. Tumor growth was monitored weekly using a digital caliper (VWR, Radnor,
PA). Tumor volume was calculated using the formula Volume = (Width × 2 × Length)/2
[325]. Four to five female NSG mice were used per treatment group (vehicle control
only, Paclitaxel only, ICG-001 only, Paclitaxel plus ICG-001) for all in vivo experiments.
2.10.4. ICG-001 application via osmotic pump implantation
For the in vivo treatment of mice with ICG-001 or vehicle control, a continuous perfusion
approach via 2ML4 Osmotic Pumps (ALZET, Cupertino, CA) was chosen. The pumps
40
were filled with 300mM phospho-ICG-001 (a water-soluble version of the compound,
solubilized in PBS) or PBS vehicle control and placed in PBS in a cell culture incubator
at 37°C overnight to prime the pumps. The following day, one pump per mouse was
implanted s.c. on the back, using surgical procedures described above. The osmotic
pumps continuously release compound at a steady rate of 2.4μL per hour over a 28-day
period. Every other day, the pumps were mobilized manually under the skin of the mice
to prevent adhesions.
2.10.5. Intraperitoneal (i.p.) injections of Paclitaxel
For in vivo Paclitaxel treatment of mice bearing TNBC xenografts a weekly i.p. injection
protocol was chosen (resembling clinical application of the drug). The drug was prepared
in sterile PBS to yield a final concentration of 10μg/ KG body weight, and a volume of
100μL per mouse was injected every week for the duration of the experiment. Before
injection, the injection site was cleaned with alcohol pads to prevent infections. The body
weight of each mouse was measured prior to injection to determine the required amount
of Paclitaxel, and served as an indicator for the overall health of the animals. A weight
loss of more than 20% body weight over the course of the treatment compared to the
body weight pre-treatment was considered as a termination point for individual mice
[326]. Furthermore, tumor xenograft size of >1.2cm in the largest extension was set as a
termination point [327]. Both are in accordance with the IRB approved IACUC protocol.

41
2.11. Sample harvest and analysis
2.11.1. RNA extraction
Treated samples were prepared in biological duplicates or triplicates in 6-well culture
dishes and 1×10
6
cells were used per sample, unless otherwise indicated. At each
experimental endpoint, culture medium was removed and cells were washed twice with
PBS. After the second wash, all supernatant was completely removed and TRIzol reagent
(Thermo Fisher Scientific) was added at 1mL per well. Samples were pipetted at least 5
times to lyse cell membranes and homogenize the solution, and then placed into 1.5mL
Eppendorf tubes and stored at -80°C until further use. Subjecting the suspension to one
freeze-thaw cycle further helped to break up cell membranes to increase nucleic acid
yield. RNA extraction from TRIzol was performed according to the manufacturers
protocol. Briefly, a phase separation is performed by adding 200μL of chloroform per
1mL TRIzol and centrifugation of the sample at 12,000 × g for 15min at 4°C. The
aqueous phase, which contains RNA, was transferred to a fresh Eppendorf tube, mixed
with 0.5mL of 100% isopropanol, and incubated at room temperature (RT) for 10
minutes. A subsequent centrifugation step was performed at 12,000 × g for 10min at 4°C,
to pellet RNA. After two washing steps using 1mL of 75% ethanol and centrifugation at
7500 × g for 5min at 4°C, all supernatant was removed and the RNA pellet allowed to try
for 5min at RT. RNA was eluded in 20μL of 10mM Tris-HCl, pH 8.0, 1mM EDTA (TE)
buffer and quality checked via a NanoDrop 2000 device (NanoDrop Products,
Wilmington, DE). Samples were considered good quality with A260/280 ratio >1.8 after
42
NanoDrop reading. All RNA aliquots were stored at -80°C, and repeated freeze-thaw
cycles were avoided to ensure RNA integrity.
2.11.2. Reverse transcription of RNA to cDNA
Total RNA was processed for gene expression analysis and first strand synthesis was
performed using the qScript cDNA Supermix (Quantabio, Beverly, MA) according to
manufacturer’s instruction. Per reverse transcription reaction, 1μg of total RNA was used
and total reaction volume was 20μL. The following reaction conditions were used on a
T100 thermal cycler (Bio-Rad, Irvine, CA): 25°C for 5min, 42°C for 30min, 85°C for
5min. The cDNA was stored at -20°C until further use.
2.11.3. Quantitative reverse transcription polymerase chain reaction (qRT-PCR)
For gene expression quantification, a SYBR Green RT-PCR master mix was used
(Quantabio). Primer sequences were obtained via the Harvard primer bank [328] and
synthesized by ValueGene (San Diego, CA). To ensure gene-specific priming, the primer
sequences were validated using the NCBI BLAST tool [329, 330] and only exon-junction
spanning primers were used. All probes were stored as 100nM stock solutions at -20°C
and diluted to a final concentration of 25pM before each experiment. All primer
sequences and corresponding genes are listed in Table 3.
Reaction were performed in 96-well plates (Bio-Rad), with 25μL per reaction containing
the following components: 2μL cDNA, 12.5μL SYBR Green master mix, 1μL each
forward and reverse primer probes, 8.5μL DNAse and RNAse free deionized water. The
43
following reaction conditions were used on a MyiQ or CFX96 Real-Time system (Bio-
Rad): 95°C for 3min, 40 cycles of 95°C for 20s, 60°C for 20s 72°C for 30s.

Table 3: Genes and primer sequences for qPCR.
NCBI Gene symbol Forward primer (5’ -> 3’) Reverse primer (5’ -> 3’)
FOXM1 CGTCGGCCACTGATTCTCAAA GGCAGGGGATCTCTTAGGTTC
CREBBP (CBP) CAACCCCAAAAGAGCCAAACT CCTCGTAGAAGCTCCGACAGT
C T N N B 1/ ( β-catenin) CATCTACACAGTTTGATGCTGCT GCAGTTTTGTCAGTTCAGGGA
AURKB CAGTGGGACACCCGACATC GTACACGTTTCCAAACTTGCC
ABCG2 ACGAACGGATTAACAGGGTCA CTCCAGACACACCACGGAT
BIRC5 (survivin) AGCCCTTTCAAGGACCAC GCACTTTCTTCGCAGTTTCC
CCNA1 ACATGGATGAACTAGAGCAGGG GAGTGTGCCGGTGTCTACTT
CCNB2 CCGACGGTGTCCAGTGATTT TGTTGTTTTGGTGGGTTGAACT
CD24 CTCCTACCCACGCAGATTTATTC AGAGTGAGACCACGAAGAGAC
CD44 CTGCCGCTTTGCAGGTGTA CATTGTGGGCAAGGTGCTATT
CDC2 (CHEK1) ATATGAAGCGTGCCGTAGACT TGCCTATGTCTGGCTCTATTCTG
CDC25 GGATGTGCTTATGCAGGATTCC TCCAGGAGCAGGTTTAACATTTT
CDK1 CATGTACTGACCAGGAGGGATAG CATGTACTGACCAGGAGGGATAG
CENPA CTCCCATCAACACAGTCGGC GAAGTCCACACCACGAGTGA
CENPF CTCTCCCGTCAACAGCGTTC GTTGTGCATATTCTTGGCTTGC
MDR1 TTGCTGCTTACATTCAGGTTTCA AGCCTATCTCCTGTCGCATTA
NEK2 TGCTCCGTGAACTGAAACATCC CCAGAGTCAACTGAGTCATCACT
NES CTGCTACCCTTGAGACACCTG GGGCTCTGATCTCTGCATCTAC
PLK1 AAAGAGATCCCGGAGGTCCTA GGCTGCGGTGAATGGATATTTC
SKP ATGCCCCAATCTTGTCCATCT CACCGAAGTGATAGGTGT
GAPDH GGTGCTGAGTATGTCTGTGA ACAGTCTTCTGGGTGGCAGT
PPIA CCTAAAGCATACGGGTCCTG TTTCACTTTGCCAAACACCA

44
2.12. RNA Seq library preparation

MDA-MB-231 samples for whole transcriptome RNA Sequencing (RNA Seq) were
prepared in biological duplicates after treatment for 48h with 10μM ICG-001 or DMSO
vehicle control. For RNA amplification from cDNA and sequencing the Ovation RNA-
Seq System V2 (NuGEN Technologies, Inc., San Carlos, CA) and Ovation Ultralow
Library System V2 (NuGEN) were used, respectively. The V2 kit is based on a Single
Primer Isothermal Amplification (SPIA) method utilizing 3’ end and random whole
transcript priming, followed by an amplification step. As starting material 100ng of total
RNA were used. Initially, RNA is reverse transcribed to yield first strand cDNA. Next,
the single strand cDNA is used for second strand cDNA synthesis and subsequent
amplification. The amplified product is purified using Agencourt RNAClean XP beads
(Beckman Coulter, Inc., Brea, CA). A NanoDrop 2000 (Thermo Fisher Scientific) was
used to measure final product quantity and quality (A260/280 ratio). Before library
preparation, SPIA cDNA was sheared using a Covaris S220 Focused-ultrasonicator
(Covaris, Woburn, MA) with settings optimized to produce 200-300bp fragments. For
quality control, the fragmentation was validated using a 2100 Bioanalyzer Instrument
(Agilent Technologies, Santa Clara, CA). Following end repair of the cDNA fragments,
hexamer adapters are ligated to the fragments, which allow for multiplexing of samples in
the flow cell. The ligated products were purified and subjected to a final PCR
amplification, followed by a final purification step with. The libraries were validated
using a 2100 Bioanalyzer and sequencing in the laboratory of Dr. Knowles at the USC
Zilkha Neurogenetic institute on an Illumina HiSeq 2000 (Illumina, San Diego, CA)
45
performing paired-end RNA Seq with a desired read depth per sample of 40-60x
coverage.
2.13. RNA Seq data analysis
FastQ files [331] were processed using Partek Flow software (Partek Inc., Chesterfield,
MO). Figure 1 shows a flow diagram of the workflow for complete analysis from Pre-
alignment quality control (QC) to final differential gene expression output.

The FastQ raw files were uploaded to the USC High-Performance Computing Cluster
(HPCC) for access via Partek Flow. After pre-alignment QC all bases with a Phred
quality score [331-333] of 30 or higher (meaning that there is a 1 in 1000 probability for
incorrect base call, leading to a base call accuracy of 99.99%) were trimmed and aligned

Unaligned
reads
Pre-alignment
QC

Trim bases

Alignment/QC
Quantification/
Differential gene expression
Visualization Raw data (FastQ) input/ QC

Trim bases

Align reads

Aligned
reads

Post-alignment QC
Coverage report
Quantification
Gene
analysis
Feature
list
Hierarchical
clustering
Figure 1: Workflow diagram for RNA Seq analysis pipeline using Partek Flow.
46
to Human International Human Genome Sequencing Consortium [334] human reference
genome (hg) 19 using the STAR aligner [335]. The aligned reads were annotated and
quantified using the Ensembl gene annotation database [336]. From this feature list a
differentially expressed gene list was created using false discovery rate (FDR) [337]
adjusted p < 0.05 and a cut off of at least 2 fold expression change, comparing MDA-
MB-231 treated with ICG-001 to DMSO vehicle control treated cells. The gene list was
used for unsupervised hierarchical clustering and to generate a heatmap for visualization.
Results were exported in text file format for further investigation of biological function
impacted by those changes.
2.14. Biological interpretation of RNA Seq data – Ingenuity Pathway analysis (IPA)
The gene lists created as described above were uploaded into IPA (Qiagen, Hilden,
Germany) and a core analysis run.
2.15. Public breast cancer gene expression data repositories
cBioPortal [338, 339] was developed at the Memorial Sloan-Kettering Cancer Center as a
data visualization and analysis tool for large genomic data sets. The web-based portal was
used to analyze FOXM1 gene expression and associated gene expression in the TCGA
breast cancer data set [40, 340]. cBioPortal was used you to extract TCGA reverse phase
protein array (RPPA) data for the breast cancer cohort from the same data set. A total of
817 cases were available for analysis. The TNBC Sunset included 82 cases and was
analyzed separately. RPPA data was available only for FOXM1, but not CBP expression.
The survival OS survival data was generated using cBioPortal and downloaded for
47
further analysis. The dataset was adjusted for 5-year OS and XLSTAT was used to
perform Kaplan-Meier survival analysis.
The UCSC Cancer Genome Browser [341, 342] is a web-based application that analysis
and visualization of genomic data from datasets from TCGA, Cancer Cell Line
Encyclopedia (CCLE), Connectivity Map and Therapeutically Applicable Research to
Generate Effective Treatments (TARGET). The tool was used to explore the relevance of
FOXM1 expression in breast cancer in a subtype dependent manner.
Oncomine, a cancer microarray database and integrated data-mining platform [343] was
used to query the TCGA breast cancer dataset for comparison of CBP and FOXM1
expression in breast cancer specimen compared to normal breast tissue, as well as
comparison of expression in TNBC subtype compared to other subtypes. The data set
contains 596 breast cancer microarray gene expression samples.
2.16. Fluorescence Activated Cell sorting (FACS)/ flow cytometry
The cell culture flasks were rinsed once with PBS and incubated with non-enzymatic cell
dissociation buffer for 5-10min in the cell culture incubator, after which the complete
culture medium containing serum was added at a ratio of 2:1. The cell solution was
harvested into 15mL conical tubes and centrifuged for 5min at 300rpm. After removal of
supernatant, the cell pellets were re-suspended in FACS washing buffer (PBS + 2% FBS)
and centrifuged as before. The washing buffer was removed and cell pellets were re-
suspended in 100μL FACS washing buffer per 1×10
6
cells. The samples were placed on
ice and 10μL each of allophycocyanin (APC)-conjugated mouse anti-human CD44
antibody (BD Biosciences) and fluorescein isothiocyanate (FITC)-conjugated mouse anti-
48
human CD24 antibody (BD Biosciences) were added per sample. Samples were briefly
vortexed and incubated for 15min on ice. Before analysis, 10μL of a 1μg/mL 4',6-
diamidino-2-phenylindole (DAPI) (Sigma-Aldrich) nuclear dye solution was added and
samples were immediately analyzed. FACS was performed on a BD LSRFortessa (BD
Biosciences).
2.17. Side population assay using FACS
The cell number was adjusted to 1×10
6
cells per mL and all samples were prepared as
biological triplicates, unless otherwise specified. Cells were re-suspended in 1mL
complete culture medium (DMEM + 10% FBS + 1% pen strep). Per sample, 5μg of
Hoechst 33342 dye (Sigma Aldrich) was added. For negative controls samples were
prepared in the same way, plus the addition of either 100μM verapamil (Sigma-Aldrich)
or 10μM fumitremorgin C (FTC) (Sigma-Aldrich). Both drugs block ABC-cassette
protein pumps which efflux Hoechst 33342 and produce the dim side population in
FACS. The samples were incubated for 2h at 37°C in a cell culture incubator. Gentle
mixing was performed every 30min. Following incubation, the samples were
immediately placed on ice to prevent further activity. To remove remaining Hoechst dye,
the samples were centrifuged at 1000rpm at 4°C for 5min and supernatant was removed.
Cell pellets were re-suspended in 300μL cold PBS containing 5μg/mL propidium iodide
(PI) for dead cell exclusion, and placed on ice until analysis. The samples were analyzed
using a BD LSRFortessa (BD Biosciences), equipped with an ultra violet laser needed for
Hoechst 33342 excitation. The gating strategy included dead cell exclusion via PI
positive cells. The gate for side population cells was established using the negative
49
controls. Side population positive cells were identified as a dim population in the Hoechst
blue/red plot, which was not present in the negative control.
2.18. Protein extraction
Cells were harvested at about 80% growth density. Two p150 Petri dishes were used per
treatment condition for each CoIP experiment. After treatment, cells were washed twice
with ice-cold PBS and subsequently scraped ice-cold PBS plus protease inhibitor (Merk
Millipore, Billerica, MA) and 1μM Dithiothreitol (DTT) (Sigma-Aldrich). Cells were
pelleted at 1000rpm for 5min and stored and -80°C until further use.
For protein extraction, the cell pellets were thawed on ice and processed using the Pierce
NE-PER Nuclear and Cytoplasmic Extraction Kit (Thermo Fisher Scientific) according to
the manufacturer’s instructions. Briefly, cell pellets are re-suspended in 100μL CERI
buffer per 10μL cell pellet and incubated for 10min on ice, with intermittent vortexing to
break up cell membranes. CERII buffer was added at 5μL per 100μL CERI buffer, and
after brief, vigorous vortexing, the solution was centrifuged at 14,000 × g for 10min at
4°C to pellet nuclei. Supernatant, which contains the cytoplasmic protein fraction was
transferred into 1.5mL Eppendorf tubes and stored at -80°C until further use. The
remaining cell nuclei pellet was re-suspended in 50μL NER solution and incubated for
45min on ice. Vortexing at the highest setting was performed for 15s every 10min to
mechanically break up nuclear membranes. After incubation, the solutions were
centrifuged at 14,000 × g for 5min at 4°C, to pellet cellular debris. The clear supernatant,
containing the nuclear protein fraction, was transferred to fresh 1.5mL Eppendorf tubes
and stored at -80°C until further use.
50
2.19. Western blotting
Protein extracts prepared as described above were used for protein quantification via
western blotting. To ensure loading of equal amounts of protein for each sample,
colorimetric protein quantification was performed using a Bio-Rad Protein Assay (Bio-
Rad). Briefly, 10μL of nuclear, cytoplasmic or whole cell protein lysate were loaded in
duplicate into a clear bottom 96-well plate. For standard curve values, an albumin
standard (Thermo Fisher Scientific) dilution serious was used (0, 125, 250 and 500μg).
Using a multichannel pipette, 200μL of Bio-Rad Protein Assay (Bio-Rad) were added to
each well and incubated on a nutating mixer (VWR) for 5min at RT. The plate was then
placed in a SpectraMax M3 spectrophotometer (Molecular Devices, Sunnyvale, CA) that
produced a read out of protein quantification for each sample compared to standard curve
values. Depending on the protein of interest, 10-30μL of protein extract per sample was
used for western blot analysis. The input values varied due to protein abundance and
antibody binding efficiency, and optimization of loading conditions was performed.
Protein extract was mixed 1:1 with 2x Laemmli Sample Buffer (Bio-Rad) containing 5%
of 2-mercaptoethanol (Sigma-Aldrich, St. Louis, MO) to yield a final volume of 40μL per
sample. The mixture was boiled for 5min to denature proteins, and subsequently loaded
onto pre-cast 4-20% gradient gels (PAGEr Gold Precast gels, Lonza, Basel, Switzerland).
For size determination after gel-electrophoresis, a Precision Plus Protein Dual Color
Standard (Bio-Rad) was loaded in two wells of each gel, flanking the samples of interest.
Gels were placed into running chamber (Bio-Rad) filled with the pre-specified amount of
running buffer (25mM Tris HCl, 193mM glycine, 0.1% Sodium dodecyl sulfate (SDS),
pH 8.3). Electrophoresis performed at 80V for 5min, followed by 130V for 45-60min.
51
After separation of protein fractions, the gels were carefully removed and placed on
nitrocellulose membranes (Bio-Rad) and fixed with sandwich plates. The plates were put
into transfer chamber (Bio-Rad) filled with cold transfer buffer (80% Tris-Borate buffer
[0.089M Tris, 0.089M Borate, pH 8.3], 20% methanol, 25mM Tris HCl, 193mM glycine)
for protein transfer from gel to membrane. The transfer was performed at 30V (PowerPac
Basic, Bio-Rad) overnight in a cold room at 4°C. The membranes were removed the next
day washed for 30min in washing buffer (Tris-buffered saline, 0.1% Tween 20), followed
by incubation in blocking solution (washing buffer plus 5% skim milk) for 60min.
Following blocking, primary antibody incubation was performed in blocking solution
over night at 4°C. To allow for the detection of several proteins simultaneously, the
membranes were cut according the proteins of interest and the membrane parts were
incubated with different primary antibodies. The following antibodies were used at the
indicated dilutions: mouse anti-human β-catenin (610153, BD Biosciences), mouse anti-
human FOXM1 (A-11, Santa Cruz Biotechnologies), rabbit anti-human CBP (A-22,
Santa Cruz Biotechnologies), mouse anti-human survivin (NB-500-205, Novus
Biologicals, LLC, Littleton, CO), mouse anti-human ACTB (H-102, Santa Cruz
Biotechnologies) mouse anti-human GAPDH (6C5, Santa Cruz Biotechnology) mouse
anti-human ABCG2 (MAB4155, Millipore, Billerica, MA), α-tubulin (B-5-1-2, Santa
Cruz Biotechnology) and mouse anti-human Lamin A/C (346, Santa Cruz
Biotechnologies). After primary antibody incubation, membranes were washed for 30min
and incubated with secondary antibody for 60min at RT in washing buffer. Secondary
antibodies used were: goat anti-mouse IgG HRP conjugated, goat anti-rabbit IgG HRP
conjugated (all from Santa Cruz Biotechnologies). A final washing step was performed
52
for 30min, followed by incubation of the membranes with HRP substrate (GE Healthcare
Life Sciences, Pittsburgh, PA) for 5min. Immediately afterwards, the membranes were
either placed in developing chamber containing GeneMate Blue lite films (VWR), or
placed into a ChemiDoc MP gel imaging system (Bio-Rad) to detect and record
chemiluminescence. For quantification of film, films were scanned using Adobe
Photoshop and images were pixelated using ImageJ software [344, 345]. Quantification
of gel dock images was performed with the Image Lab software (Bio-Rad) provided with
the device.
2.20. Co-immunoprecipitation (CoIP)
A total of 250μg of nuclear lysate was used per antibody and treatment condition for IP.
The lysates were diluted 1:10 in complete CoIP buffer (20nM Tris HCL pH 8.0, 137mM
NaCl, 10% glycerol, 1% NP-40, 1mM EDTA, 1mM MgCl2, 1mM CaCl2) (supplemented
with Protease inhibitor and DTT). 2μg of antibody were added per sample and incubated
at 4°C overnight on a nutating mixer (VWR). The antibodies used were: CBP (A-22,
Santa Cruz Biotechnology) and normal rabbit IgG (Santa Cruz Biotechnology). The next
day, a 50% slurry of Immobilized Protein A Resin (GBioscience, St. Louis, MO) was
added to each sample and incubated for 1h at 4°C und constant nutation. Samples were
than washed five times with complete CoIP buffer. Washing time was 5 minutes and
resin beads were pelleted 60s at 1000rpm between washing steps to remove supernatant.
After the final washing step, samples were boiled in 30μL Laemmli sample buffer (Bio-
Rad) for 5min. The resin beads were removed from the denatured protein solution by
spinning in filter columns (GE Healthcare) for 1min at 6000rpm. The protein-buffer
53
solution from each sample was loaded onto agarose gradient (4-20%) gels (PAGEr Gold
Precast gels, Lonza) and protein electrophoresis was done at 150V for 90min. Gels were
plotted onto PDVF membranes (Immun-Blot, Bio-Rad) overnight. The next day,
membranes were washed for 30min wish washing buffer. After blocking in 5% skim milk
for 1h, primary was added in 10mL blocking solution and incubated at 4°C overnight
under constant rotation. The primary antibodies used were: mouse anti-human β-catenin
(610154, BD Biosciences) at a 1:2000 dilution, and mouse anti-human FOXM1 (A-11,
Santa Cruz Biotechnology) at a 1:1000. The following day, membranes were washed for
30min using washing buffer as described above and incubated with goat anti-mouse
secondary antibody conjugated with HRP (sc-2005, Santa Cruz Biotechnology) for 1h at
room temperature. Membranes were washed once more for 30min to remove unbound
antibody and subsequently incubated for 5min with ECL Select Western Blotting
Detection Reagent (GE Healthcare). Chemiluminescent bands were visualized using
GeneMate Blue lite films (VWR) or a ChemiDoc MP gel imaging system (Bio-Rad).
Quantification of protein pull-down was done by densitometric analysis and area under
the curve (AUC) calculation using ImageJ software.

2.21. Tissue Micro Arrays (TMAs)
TMAs were acquired from the Mid-Atlantic Division of the Cooperative Human Tissue
Network (CHTN) based at the University of Virginia. Cancer Diagnosis Program (CDP)
Breast Cancer Stage II Prognostic Tissue Microarrays were specifically designed by the
National Cancer institute (NCI) to examine potential prognostic markers in stage II non-
54
metastatic breast cancer with to high statistical power. TMA distribution was dependent
on an approval process by the pathology department at the University of Virginia based
on the merit of the research project. A total of 8 slides (duplicate cases per two slides)
containing (number) cases were provided, containing 430 breast cancer cores, as well as
normal breast tissue, breast cancer and non-breast cancer cell controls. Relevant de-
identified clinical annotation was available but not limited to the following parameters:
age, hormone receptor status, tumor size (in cm), tumor stage (T), histopathological
grade, lymph node metastasis (N), distant metastasis (M) and site, radiation therapy,
hormone therapy, chemotherapy, type of surgery, vital status, cancer status at death and
overall survival.
Due to loss of sections and missing information on hormone receptor expression, 316
cases were included in the final analysis. After titration and antibody dilution
optimization the following staining conditions were chosen: monoclonal mouse anti-
human FOXM1 (A-11, Santa Cruz Biotechnology) at 1:120 dilution, and monoclonal
mouse anti-human CBP (C-20, Santa Cruz Biotechnology) at 1:100. Staining was
performed on 4 slides per antibody using a BOND-III Automated IHC/ISH Stainer
(Leica, Buffalo Grove, IL) by the USC Clinical Immunohistochemistry laboratory.
A three-tier scoring system was devised in consultation with Dr. Michael Press, a
translational pathologist at USC: 0 - no staining, 1 - weakly positive staining, 2 - strongly
positive staining. Only tumors with strong staining (score 2) were counted as positive for
either FOXM1 or CBP in the subsequent analysis.

55
2.22. Statistical analysis
GraphPad PRISM (GraphPad Software Inc., La Jolla, CA) was used for statistical
analysis of all experiments, except for TMA analysis. Non-normal distribution was
assumed and appropriate non-parametric statistical test were used; for experiments with
single variables and groups of two Mann-Whitney test was used, for three or more groups
non-parametric one-way ANOVA (Kruskal-Wallis and Dunn’s multiple comparison) was
used. For two variables and groups of two or more two-way ANOVA. Statistical
hypothesis testing (Sidak test) was used to correct for multiple comparison. Data for
repeats are presented as means and standard deviation.
The TMAs were analyzed using chi-square (GraphPad PRISM) to explore statistically
significant association of dependent and independent variable pairs. Table 4 lists the
clinical factors for the TMA cases that were used for analysis, classified into categorical
and numerical variables. Statistically significant pairs were used further statistical
analysis with XLSTAT (Addinsoft, New York, NY). A multivariate logistic regression
model was used to calculate odds ratios. Cox regression analysis was used to determine
hazard ratios. The TMA data set was adjusted for 5-year overall survival and analyzed for
association between FOXM1 and CBP protein expression and survival outcome in all
cases (n=316), as well as the TNBC Sunset (n=53). Kaplan-Meier analysis in XLSTAT
was used to generate survival curves.
A two-tailed statistical significance level (alpha) of 0.05 for all statistical tests was
considered meaningful to reject the null hypothesis.

56

Table 4: Clinical variables used for association studies with FOXM1 and CBP protein levels.
Variable Values Type
Marker (FOXM1, CBP)
1 = positive
0 = negative
Categorical

Subtype
TNBC
ER/PR+
ER+
HER2+
Age
≤ 50 years
> 50 years
Stage
T1
T2
T3
Grade
Grade I
Grade II
Grade III
N (lymph node metastasis

N0 = no lymph node metastasis
N1 = yes lymph node metastasis
Chemotherapy

Yes
No
Radiation therapy

Yes
No
Vital status
Alive
Dead
Tumor size cm
Numerical
Overall survival Days

57
3. RESULTS
3.1. CREB binding protein (CBP) and CBP/β-catenin interaction as a potential
target in triple negative breast cancer (TNBC)

Querying publicly available genomic data sets from all breast cancer studies available via
cBioPortal [338, 339] showed that CBP amplification is a common phenomenon in breast
cancer (Figure 1A). Oncomine analysis demonstrated overexpression of CBP RNA in
breast cancer tissue compared to normal breast (Figure 2B), and higher expression in
TNBC subtype compared to other subtypes (Figure 2C) (1.3-fold change, 593 samples, p
= 6.24E-7).

25%
20%
15%
10%
0%
Deletion
Amplification
Mutation
5%
Alteration frequency
Breast
cancer normal TNBC
Other
subtypes
Log2 median-centered intensity
2
0
-3
2.5
0
-1.5
p = 6.24E-7
1.3-fold change
593 samples
2A
2B 2C
Figure 2A: Four publicly available data sets show genetic alterations in CBP in breast cancer (BCCRC –
British Columbia Cancer Research Center, TCGA – The Cancer Genome Atlas, METABRIC – Molecular
Taxonomy of Breast Cancer International Consortium, MSKCC – Memorial Sloan Kettering Cancer Center.
2B: CBP RNA levels in normal compared to breast cancer tissue 2C: CBP RNA expression in TNBC
compared to other breast cancer subtypes.

58
Western blot protein quantification in two TNBC cell lines (MDA-MB-231 and MDA-
MB-468) showed high expression levels of CBP compared to the non-tumorigenic breast
epithelial cell line MCF10a (Figure 2D). The Kahn lab had previously demonstrated that
survivin (BIRC5) is a direct target of CBP/β-catenin transcription [243]. The western blot
in shows that survivin is highly expressed in MDA-MB-231 and MDA-MB-468 cells,
compared to MCF10a (Figure 2D). The same figure also shows β-catenin protein levels,
which did not show any association with the grade of malignancy of all three cell lines
(Figure 2D).

CoIP demonstrated that CBP binds to β-catenin in three TNBC cell lines (MDA-MB-231,
MDA-MB-468 and SUM149) under DMSO control conditions and can be disrupted after
24h treatment with 20μM ICG-001 (Figure 2E), a specific CBP-binding small molecule
inhibitor [346] (measurements given as area under the curve (AUC) DMSO 6961 ± 1528
vs. ICG-001 1095 ± 947, n=3, p = 0.031) (Figure 2E).

Figure 2D: CBP, survivin and β-catenin protein levels in two TNBC cell lines (MDA-MB-231, MDA-
MB-468) and non-tumorigenic epithelial breast cell line MCF10a (1 – cytosolic fraction, 2 – nuclear
fraction).
2 1 2 1 2 1

CBP
Lamin A/C
Survivin
CTNNB1
2D
59

Using 6kb survivin-promoter driven luciferase reporter transfected into TNBC breast
cancer cells (MDA-MB-231, MDA-MB-468 and Hs578T) as a read out for CBP/β-
catenin driven gene expression [243] showed that 24h treatment with 10μM ICG-001
statistically significantly down-regulates reporter activity in three cell lines (results given
as mean difference in ratio of firefly to Renilla luciferase: MDA-MB-231 DMSO 0.18 ±
0.006 vs. ICG-001 0.002 ± 0.0016, p < 0.01; MDA-MB-468 DMSO 0.57 ± 0.0006 vs.
ICG-001 0.16 ± 0.009, p < 0.0001; Hs578T DMSO 1.86 ± 0.1 vs. ICG-001 0.63 ± 0.03, p
< 0.0001) (Figure 2G).
Western blot protein quantification demonstrated that treatment with 10μM ICG-001 for
24h leads to down-regulation of survivin protein levels compared to DMSO vehicle
control in MDA-MB-231 cells (Figure 2H).

ICG-001
Figure 2E: Chemical structure of ICG-001. 2F: Co-immunoprecipitation (CoIP) of CBP/β-catenin in
three TNBC cell lines (MDA231, MDA468 and SUM149) under DMSO vehicle control conditions and
after treatment with 20μM ICG-001 for 24h. Asterisk represents statistical significance (* p < 0.05)

DMSO ICG-001
0
4000
8000
12000
Area under the curve
*

4000
8000
12000
Area under the curve
DMSO ICG-001
0
2E 2F
60

3.2. FOXM1 and its target genes are over-expressed in TNBC
Since neither CBP nor β-catenin have DNA-binding domains and interact directly as
transcription factors, a query of TCGA breast cancer RNA Seq data was performed and
showed that TNBC breast cancers are characterized by the overexpression of a FOXM1
Figure 2G: Survivin-promoter driven luciferase reporter activity in three TNBC cell lines (MDA-MB-
231, MDA-MB-468 and Hs578T) treated for 24h with 10uM ICG-001 or DMSO vehicle control. Asterisk
represents statistical significance (N=3 per cell line per treatment condition (** p < 0.01, **** p < 0.0001)

DMSO ICG-001
Survivin
GAPDH

Figure 2H: Western blot for survivin expression MDA-MB-231 treated for 24h with ICG-001 or DMSO
vehicle control.
MDA-MB-231 MDA-MB-468 Hs578T
0.0
0.1
0.2
0.3
0.5
1.0
1.5
2.0
2.5
normalized ratio
DMSO
ICG-001

MDA231 MDA468 HS578T
DMSO
ICG-001

****
****
**
Normalized ratio
2G
2H
61
driven gene network compared to other molecular subtypes. The data is visualized via the
University of California Santa Cruz (UCSC) Cancer Genome Browser in the heatmap in
figure 3A, with individual breast cancer cases represented in rows, and genes represented
in columns. A network analysis using cBioPortal further confirmed the central role of
FOXM1 in the expression of gene transcripts over-expressed in TNBC and showed the
potential role for CBP in FOXM1 driven gene expression (Figure 3B). Oncomine
analysis demonstrated overexpression of FOXM1 RNA in breast cancer tissue compared
to normal breast (4.2-fold change, 593 samples, p = 3.51E-30) (Figure 3C), and higher
expression in TNBC subtype compared to other subtypes (Figure 3D).
3.3. A chemical-genomic approach using ICG-001 identifies FOXM1 as downstream
effector of CBP-signaling in TNBC

A chemical-genomic approach was used to investigate potential TFs involved in the
CBP/β-catenin signaling. To this end, MD-MBA-231 cells treated with either ICG-001 or
DMSO control were compared using unbiased whole transcriptome analysis via RNA
Seq. Whole transcriptome RNA Seq analysis of duplicate samples of MDA-MB-231
treated for 48h with 10μM ICG-001 or DMSO vehicle control was used as an unbiased
approach to study global gene expression changes upon ICG-001 treatment in TNBC
cells. Differential gene expression analysis and unsupervised hierarchical clustering after
48h treatment revealed that 1339 genes are differentially expressed between treatment
and control conditions (DMSO vs. ICG-001 729 genes up-regulated, 610 genes down-
62

Figure 3A: TCGA breast cancer data dataset visualization via the Santa Cruz Cancer Genome Browser showing FOXM1 regulated gene expression in
different breast cancer subtypes and normal breast tissue (817 samples, 82 TNBC). The heatmap shows individual samples in rows and genes in columns
(red – up-regulated, blue down-regulated) 3B: Functional network of FOXM1 target expression and involvement of CBP via interaction with FOXM1. 3C:
FOXM1 RNA levels in normal compared to breast cancer tissue 3D: FOXM1 RNA expression in TNBC compared to other breast cancer subtypes.
TNBC
ER/PR+
ER/PR/HER2+
Normal breast

ER PR HER2
up
BIRC5
SKP2
FOXM1
CCNE1
CENPA
PLK1
CHEK2
AURKB
CENPF
CCNA2
CCNB1
NEK2
CCNB2
CDK1

down
3A
3B
3C 3D
Normal
Breast
cancer TNBC
Other
subtypes
Log2 median-centered intensity
Log2 median-centered intensity
0
2
-5.5
1
0
-6
63
regulated; FDR ≤ 0.05, ≤ |2-fold| change) (Figures 4A and B). Analysis of this
differentially expressed gene-signature using the Ingenuity Pathway Analysis (IPA) tool
revealed FOXM1 as a potential upstream-regulator of the gene expression changes
observed in RNA Seq after treatment of MDA-MB-231 cells with ICG-001 compared to
DMSO vehicle control cells (Figure 4C).

Gene Expression Function Predicted Effect
FOXM1 -1.697 Transcriptional regulator Inhibited

Figure 4A and B: Heatmap and volcano plot showing 1300 differentially expressed genes in RNA Seq in
MDA-MB-231 treated with 10uM ICG-001 or DMSO vehicle control (FDR ≤ 0.05, ≤ |2-fold| change). 4C:
Ingenuity pathway analysis of RNA Seq data and identification of upstream regulatory factor FOXM1
(N=2 per treatment condition).

ICG-001 DMSO
-0.92 1.12
Up-regulated
Down-regulated
Not significant

p-value
Fold change
ICG-001 vs. DMSO

1e-2 1e-6 1e-4 1e-1 1

0.05

-2 -8 -16 -32 8 16 32 4 2 n/c -4
Ingenuity pathway analysis

4A 4B
4C
64
3.4. CBP/FOXM1 bind in TNBC cells; and ICG-001 effects this binding and as well
as FOXM1 and CTNNB1 protein levels

Co-immunoprecipitation (CoIP) is a technique used to identify protein-protein
interactions under physiological conditions [347]. Using an antibody against CBP for
protein pull-down and staining for associated FOXM1 demonstrated that both proteins
interact in MDA-MB-231 cells (N=2 total experiments, one post 24h treatment time
points and one post 4h treatment time point). Treatment with ICG-001 disrupted
CBP/FOXM1 (measurements given as background corrected AUC: DMSO 2,051 vs.
ICG-001 405, 80% reduction) and CBP/β-catenin (DMSO 13,747 vs. ICG-001 7,939,
42% reduction) interaction after 4h of treatment (Figure 5A). Staining for FOXM1 and β-
catenin levels at the 4h time point showed no difference in protein expression levels
between ICG-001 and DMSO vehicle control treated cells (Figure 5B).

Area under the curve
(AUC)
IgG
FOXM1

CTNNB1

DMSO ICG-001 DMSO ICG-001

IgG
4000
3000
2000
1000
5000
0
1000
2000
3000
4000
0
Figure 5A: CoIP of CBP with CTNNB1 or FOXM1 in MDA-MB-231 treated with 20μM ICG-001 or
DMSO vehicle control (N=1).
5A
65

After 24h treatment, on top of the disruption of CBP/FOXM1 and CBP/β-catenin (Figure
5C), protein levels of FOXM1 (and β-catenin) were substantially reduced in ICG-001
treated MDA-MB-231 cells compared to control (Figure 5D). CoIP also demonstrated
that FOXM1 and β-catenin bind in MDA-MB-231, and that treatment with ICG-001 does
not affect FOXM1/β-catenin binding, demonstrating they specificity to CBP (Figure 5E).
FOXM1 levels are also reduced in MDA468 (N=1) (Figure 5F)

IgG
FOXM1
IgG
CTNNB1
DMSO ICG-001 DMSO ICG-001
AUC

4000
3000
2000
1000
5000
0
1000
2000
3000
4000
0
Figure 5B: Western blot showing protein levels of FOXM1 and CTNNB1 4h after treatment of MDA-
MB-231 with ICG-001 or DMSO (N=1).
DMSO ICG-001
0
5000
10000
15000
20000

DMSO ICG-001
0
2000
4000
6000

+
-
+
- +
-
+
-
DMSO
ICG-001
Cytoplasmic Nuclear
FOXM1
α-tubulin
Lamin A/C
CTNNB1
7% 
AUC
AUC
DMSO
CTNNB1
FOXM1
DMSO ICG-001
ICG-001
<1% 

6000
4000
2000
0
5000
20000
15000
10000
0
Figure 5C: CoIP of CBP with CTNNB1 or FOXM1 in MDA-MB-231 treated for 24h with 20uM ICG-
001 or DMSO vehicle control (N=1)
5B
5C
66

0
50
100
150

DMSO 20uM ICG-001
0
1000
2000
3000
4000

Cytoplasmic Nuclear
DMSO
ICG-001
FOXM1
α-tubulin
Lamin A/C
CTNNB1
90% 
DMSO ICG-001
AUC
FOXM1
AUC
CTNNB1
47% 
DMSO ICG-001
+
-
+
- +
-
+
-
4000
3000
2000
0
1000
1000
500
1500
0
DMSO ICG-001
0
50
100
150
relative FOXM1 protein level

FOXM1
ACTB
+
-
+
-
DMSO
ICG-001
DMSO ICG-001
AUC
50% 
100
50
150
0
CTNNB1
FOXM1 IgG
FOXM1
Input
CBP
IgG
Input
CTNNB1
+
-
+
- +
-
+
-
DMSO
ICG-001
+
-
+
-
IP/WB:
IP/WB:
Figure 5F: FOXM1 levels in MDA-MB-468 TNBC cells after 24h treatment with 20uM ICG-001 or
DMSO control (n=1).
Figure 5E: CoIP for CBP pull-down and staining for associated FOXM1 and b-catenin, as well as
FOXM1 pull down and staining for associated β-catenin (N=1).
Figure 5D: Western blot showing protein levels of FOXM1 and CTNNB1 after 24h treatment of MDA-
MB-231 with ICG-001 or DMSO (N=1).

5D
5E
5F
67
3.5. Treatment with ICG-001 effects FOXM1 transcriptional activity and target
gene expression

A comparison of the TCGA data set with the differential expression signature after ICG-
001 treatment in MDA-MB-231 cells showed substantial overlap of gene transcripts
over-expressed in TNBC and down-regulated in ICG-001 treated cells (Figure 6A).
Quantitative real time polymerase chain reaction (qPCR) for FOXM1 target genes
identified in TCGA revealed that while 4h treatment of MDA-MB-231 with ICG-001 did
not change gene expression compared to DMSO control, 24h treatment with ICG-001
caused a significant down-regulation of all FOXM1-dependent genes (Figure 6B).

FOXM1
PLK1
BIRC5
SKP2
CCNE1
AURKB
CENPF
CHEK2
CENPA
NEK2
CCNA2
CCNB1
CCNB2
CDK1
CDC25
0
5
10
15
20
20
30
40
50
50
100
150
RPKM
DMSO
ICG-001
CCNA2
CDC2
PLK1
BIRC5
FOXM1
CCNB1
CCNB2
CDC25
SKP2
CCNE1
CENPF
ARUKB
CENPA
CHEK2
NEK2
DMSO
ICG-001
150
100
50
40
30
20
15
10
0
5
RPKM
Figure 6A: RNA Seq. differential gene expression (RPKM) in FOXM1 target genes in MDA-MB-231
cells treated for 48h with either 10μM ICG-001 or DMSO vehicle control (ICG-001 vs. DMSO, N=2
per time point per condition).
6A
68

Changes in gene expression were only observed after 24h, at which time FOXM1 protein
levels were reduced significantly (Figure 5B). While CBP/FOXM1 binding was disrupted
already after 4h treatment with ICG-001 (Figure 5A), protein levels of FOXM1 and
FOXM1 target gene expression remained at similar levels as controls. These results
suggested that disruption of CBP/FOXM1 is not sufficient to cause the observed down-
regulation, and that transcriptional changes take place only after decrease of FOXM1
levels. Using a FOXM1-driven luciferase reporter showed that transcriptional activity of
FOXM1 in MDA-MB-231 is significantly inhibited only after 24h treatment with ICG-
001 compared to DMSO control (relative light units (RLUs) are adjusted for 20,000 cells
per well per condition: 4h treatment DMSO 43240 ± 501 vs. ICG-001 51986 ± 1265, n.s.;
24h treatment DMSO 43853 ± 7011 vs. ICG-001 8453 ± 300, p < 0.0001) (Figure 6C).

Figure 6B: Expression changes (qPCR) in FOXM1 target genes in MDA-MB-231 cells treated for 4 or
24h with either 10μM ICG-001 or DMSO vehicle control (N=3 per time point per condition)
CDC2 CCNA2 CCNB1 AURKB CDC25C CCND1 PLK1 FOXM1 ABCG2 BIRC5
0.0
0.5
1.0
1.5
2.0
2.5
4h treatment
24h treatment
relative gene expression
MDA-MB-231

AURKB
CCNA2
CCNB1
FOXM1
PLK1
CDC2
CDC25
ABCG2
CCND1

4h
24h

Relative gene expression
0
0.5
1
2
1.5
2.5

BIRC5
6B
69

Treatment of n=3 TNBC cell lines (MDA-MB-468, MDA-MB-453, Hs578T) showed
down-regulation of FOXM1 target genes in cells treated with 10μM ICG-001 compared
to DMSO control (Figure 6D). Three TNBC cell lines (MDA-MB-231, MDA-MB-468
and Hs578T) transfected for 24h with FOXM1-driven Firefly luciferase reporter and
empty vector Renilla luciferase and subsequently treated for 24h with either 10μM ICG-
001 or DMSO vehicle control showed statistically significant inhibition of FOXM1
transcriptional activity (numbers are normalized ratios of FOXM1/empty vector RLUs:
MDA-MB-231 DMSO 1.7 ± 0.06 vs. ICG-001 1.1, p < 0.01; MDA-MB-468 6.7 ± 0.17
vs. ICG-001 3.13 ± 0.06 p < 0.0001; Hs578T DMSO 18.4 ± 0.46 vs. ICG-001 12.7 ± 0.3,
p < 0.0001) (Figure 6E)

Figure 6C: FOXM1-driven luciferase reporter in MDA-MB-231 transfected for 24h and treated for
either 4 or 24h with 10μM ICG-001 or DMSO vehicle control (bars represent relative light units
(RLUs) per 20,00 cells, N=3 per time point per condition). Asterisk represents statistical significance
(** p < 0.01)
4 24
0
20000
40000
60000
RLU/ 20,000 cells
time in hours
ICG-001
DMSO
DMSO
ICG-001

60000
40000
20000
0
RLU per 20000 cells
4h 24h
**
6C
70

Figure 6E: FOXM1luc
Firefly/Renilla
reporter for three
TNBC cell lines (MDA-MB-231, MDA-MB-468,
Hs578T) treated for 24h with either 10μM ICG-001 or
DMSO vehicle control (bars represent normalized ratios
of FOXM1 Firefly luciferase to control vector Renilla
luciferase expression; N=3 per condition per cell line).
Asterisk represents statistical significance (* p < 0.05,
**** p < 0.0001).

Figure 6D: FOXM1 target gene expression in three
TNBC cell lines treated for 24h with 10μM ICG-
001 or DMSO vehicle control (N=2 per cell per
condition).
CDC2 CCNA2 CCNB1 AURKB CDC25C PLK1 FOXM1 BIRC5 CDK1
0.0
0.5
1.0
1.5
relative gene expression
MDA-MB-468
DMSO
ICG-001
MDA-MB-468

CCNA2
CCNB1
AURKB
CDC25C
FOXM1
CDK1
PLK1
CDC2
BIRC5

Relative gene expression
0
0.5
1
1.5

ABCG2 AURKB BIRC5 CBP CCNA2 CCNB1 CTNNB1 FOXM1 MDR1 PLK1
0.0
0.5
1.0
1.5
DMSO
ICG-001
relative gene expression
Hs578T
Hs578T

AURKB
BIRC5
CBP
CCNA2
CCNB1
FOXM1
PLK1
CTNNB1
ABCG2
MDR1

Relative gene expression
0
0.5
1
1.5

AURKB BIRC5 CBP CCNA2 CCNB1 CTNNB1 FOXM1 PLK1
0.0
0.5
1.0
1.5
2.0
relative gene expression
MDA-MB-453
DMSO
ICG-001
MDA-MB-453

AURKB
BIRC5
CBP
CCNA2
CCNB1
FOXM1
PLK1
CTNNB1

Relative gene expression
0
0.5
1
2

1.5
MDA-MB-231 MDA-MB-468 Hs578T
0.0
0.5
1.0
1.5
2.0
2
4
6
8
12
16
20
normalized ratio
DMSO
ICG-001

MDA231 MDA468 HS578T
DMSO
ICG-001

*
****
****

20
16
12
8
6
4
2
1.5
1.0
0.5
0.0
Normalized ratio
6D
6E
71
3.6. Gene knockdown (KD) and overexpression mimic ICG-001 effects on FOXM1
target gene expression in TNBC cells

Since CBP, FOXM1 and β-catenin form a transcriptional complex in TNBC cells and
Western blot showed that reduction in FOXM1 levels also affect β-catenin levels,
combination conditional gene KD via siRNA was performed followed by qPCR for
FOXM1 target genes. Figure 6A and table 5 shows that compared to scramble transfected
cells, simultaneous KD of FOXM1 and β-catenin had the strongest effect on gene
expression recapitulating ICG-001 treatment, followed by CBP and β-catenin
simultaneous KD (Figure 7A). Over-expression of FOXM1 in MDA-MB-231
demonstrated a partial rescue of gene expression changes induced by ICG-001 compared
to empty vector-transfected cells (Figure 7B), which was statistically significant in one-
way ANOVA (vector DMSO vs. vector ICG-001 p = 0.005, FOXM1
Wt
DMSO vs.
FOXM1
Wt
ICG-001 n.s.), but not in two-way ANOVA analysis (Table 6).
3.7. Canonical Wnt (TCF/β-catenin) signaling activity is low in TNBC
Wnt pathway is mediated via binding of β-catenin to TCG/lef TFs. The
TOPFLASH/FOPFLASH is a widely-used read-out for TCF/β-catenin transcriptional
activity [168]. Two colorectal cancer cell lines were used as positive controls, SW480
with known high TOPFLASH activity, and HCT116 with relative low baseline activity
[348]. The results for TOP/FOP ratio for all seven TNBC cell lines fall in the low activity
category like HCT116 (Figure 8A). Since TOPFLASH activity is dependent on β-catenin
stabilization and translocation to the nucleus, all cells were treated with Lithium

72

Gene/Condition
FOXM1+
CTNNB1
FOXM1+
CBP
CBP+
CTNNB1
CBP+FOXM1+
CTNNB1
CBP n.s. p < 0.0001 p < 0.001 p < 0.0001
CTNNB1 p < 0.0001 n.s. p < 0.0001 p < 0.0001
FOXM1 p < 0.0001 n.s. p < 0.0001 p < 0.0001
ARUKB p < 0.001 n.s. n.s. n.s.
CCNA2 p < 0.0001 p < 0.0001 p < 0.01 p < 0.0001
CCNB1 p < 0.0001 n.s. p < 0.001 p < 0.01
BIRC5 p < 0.0001 p < 0.01 n.s. p < 0.001
PLK1 p < 0.01 n.s. n.s. n.s.

Table 5: Differential FOXM1-driven gene expression in MDA-MB-231 after 48h siRNA mediated gene
knockdown.
Figure 7A: Transient siRNA gene knockdown (KD) (48h) in MDA-MB-231 cells and qPCR for FOXM1 target
(N=2 per condition).
CBP CTNNB1 FOXM1 AURKB CCNB1 CCNA2 BIRC5 PLK1
0.0
0.5
1.0
1.5
2.0
SCRBL
FOXM1 / CTNNB1
FOXM1 / CBP
CBP / CTNNB1
FOXM1 / CBP / CTNNB1
relative gene expression

CBP
CTNNB1
FOXM1
AURKB
CCNB1
CCNA2
BIRC5
PLK1

Relative gene expression
0
0.5
1
1.5
2

Scramble
FOXM1 + CTNNB1
FOXM1 + CBP
CBP + CTNNB1
FOXM1 + CBP + CTNNB1
7A
73

Table 6: Differential FOXM1-driven gene expression in MDA-MB-231 after 24h FOXM1 overexpression
and subsequent 24h treatment with 10μM ICG1-001.
Gene/Condition
Empty vector
DMSO vs. ICG-001
FOXM1 vector
DMSO vs. ICG-001
ARUKB p < 0.0001 p < 0.01
CCNA2 p < 0.0001 p < 0.0001
CCNB1 p < 0.001 p < 0.0001
BIRC5 p < 0.0001 p < 0.001
PLK1 p < 0.0001 p < 0.0001

Figure 7B: FOXM1 target gene expression in MDA-MB-231 cells transiently transfected with FOXM1
wt

overexpression or empty vector control (24h) and subsequent 24h treatment with 10uM ICG-001 or DMSO
control (N=2 per condition)
AURKB
BIRC5
CCNA2
CCNB1
PLK1
0.00
0.25
0.50
0.75
1.00
1.25
1.50
1.75
relative gene expression
Vector control DMSO
Vector control ICG-001
FOXM1
wt
DMSO
FOXM1
wt
ICG-001

Vector control DMSO
Vector control ICG-001
FOXM1
Wt
DMSO
FOXM1
Wt
ICG-001

AURKB BIRC5 CCNA2 CCNB1 PLK1

Relative gene expression
1.75
1.5
1.25
1.00
0.75
0.5
0.25
0
7B
74
Chloride (LiCl), which mimics activation of Wnt signaling via the inhibition of GSK3
[349]. Treatment with 10mM LiCl for 6h showed no significant increase in TOPFLASH
activity in neither colorectal cancer nor TNBC cell lines or, except for Hs578T, which
showed a 2.3-fold (p < 0.001) increase in TOP/FOP ration (Figure 8A). Accordingly, the
activity of Wnt/TCF transcriptional activity was judged to be relatively low in TNBC.
Treatment of MDA-MB-231 with a pan-Wnt inhibitor WNT974 [350] had no effect on
FOXM1 target genes and did not recapitulate changes in gene expression seen upon ICG-
001 treatment (Figure 8B and Table 7).

Figure 8A: TOPFLASH as readout for canonical Wnt signaling in TNBC cell lines. The colorectal
cancer cell lines SW480 and HCT116 served as positive controls. Cells were treated for 6h with 10mM
LiCl (bars represent ratio of TOPFLASH to FOPFLASH luciferase RLUs, N=3 per cell line per
condition).
SW480 HCT116 MDA231 MDA468 BT20 HCC1837 CAL51 Hs578T MDA453
0
2
4
6
8
10
No LiCl LiCl
TOPflahs/FOPflash ratio

SW480
HCT116
MDA-MB-231
MDA-MB-468
MDA-MB-453
HCC1837
CAL51
Hs578T
BT20

No LiCl

LiCl

0
2
4
6
8
10
TOP/FOPFLASH ratio
8A
75

Table 7: Differential FOXM1-driven gene expression in MDA-MB-231 after ICG1-001 or WNT974
treatment. Asterisk represents statistical significance (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p <
0.0001).
Gene/ Inhibitor ICG-001 WNT974
CBP p < 0.001 n.s.
CTNNB1 p < 0.05 n.s.
FOXM1 p < 0.01 n.s.
AURKB p < 0.05 n.s.
CCNB1 p < 0.001 n.s.
CCNA2 p < 0.001 n.s.
BIRC5 p < 0.001 n.s.

CBP CTNNB1 FOXM1 AURKB CCNB1 CCNA2 BIRC5 PLK1 ABCG2
0.0
0.5
1.0
1.5
2.0
DMSO 20uM
ICG-001 20uM
DMSO 1uM
WNT974 1uM
relative gene expression

CTNNB1
FOXM1
AURKB
CCNB1
CCNA2
BIRC5
PLK1

Relative gene expression
0
0.5
1
1.5
2
CBP

DMSO
ICG-001
DMSO
WNT974

Figure 8B: FOXM1 target gene expression (qPCR) in MDA-MB-231 cells treated with either 10μM ICG-
001, 1μM WNT974, 1μM (red) or 10μM (green) DMSO vehicle control for 24h (N=2 per condition).
8B
76
3.8. Paclitaxel treatment increases FOXM1 expression, as well as drug resistant and
stem like cancer cell populations TNBC cells

Treatment of MDA-MB-468 TNBC cells with 10nM Paclitaxel over a time course of 5 to
6 days showed that after initial reduction of cell numbers, drug resistant cell colonies
recurred (Figure 9A). These drug resistant MDA-MB-468 cells showed increased levels
of FOXM1 protein expression compared with treatment naïve cells (Figure 9B). A query
of TCGA RNA Seq breast cancer data via cBioPortal showed that tumors from patient
that had received chemotherapy express higher levels of FOXM1 RNA compared to
tumors from patient without prior treatment for all subtypes (Figure 9C), and especially
pronounced in patients with triple negative disease (Figure 9D).
A hallmark of drug resistant cancer cells is the expression of CSC markers such as the
Hoechst side population (with high expression of efflux proteins MDR1 and ABCG2
identified via a fluorescent dye efflux FACS assay [351] ), as well as CD44
high
CD24
low

phenotype, both of which have been associated with increased FOXM1 levels [294].
These cells have been demonstrated to be highly drug resistant. Paclitaxel resistant
MDA-MB-468 cells showed statistically significantly higher numbers of CSC like cell, as
demonstrated by both higher numbers of side population cells (untreated 0.63 ± 0.03%
vs. resistant 3.4 ± 0.06, p < 0.0001) (Figure 9E) as well as a higher percentage of the
CD44
High
CD24
Low
population (p < 0.0001) (Figure 9F).

77

FOXM1
ACTB
Figure 9A: Paclitaxel resistant MDA-MB-468 cell colony outgrowth after 5d treatment with 10nM Paclitaxel. 9B: FOXM1
protein levels are increased in Paclitaxel resistant MDA-MB-468 compared to treatment naïve cells. 9C and D: TCGA breast
cancer data set showing increased FOXM1 RNA expression in tumors treated with chemotherapy compared to untreated
tumors. 9E: FACS analysis of CSC like CD44
High
CD24
Low
cells in naïve vs. Paclitaxel resistant MDA-MB-468 cells (N=3).
9F: FACS analysis of side population (SP) cells in naïve vs. Paclitaxel resistant MDA-MB-468 cells. Asterisk represents
statistical significance (*** p < 0.001) (N=3).

9B
9C
9F

CD24
Unstained Untreated Resistant

CD44
untreated resistant
0
2
4
6
8
% CD44
high
CD24
low
cells

Untreated Paclitaxel
resistant
2
4
0
6
8
% CD44
High
CD24
Low
cells
untreated resistant
0
1
2
3
4 ***
% SP cells

Paclitaxel resistant

Control SP

SP% positive cells
0
1
2
3
4
Untreated Paclitaxel
resistant
Untreated
9A
9E
DMSO Paclitaxel
2 days 5 days
9D
Yes No
14
13
12
11
10
9
8
7
6
5
mRNA expression
(RNA Seq V2 RSEM) (log2)
Yes No
All subtypes TNBC
78
3.9. Treatment with ICG-001 in combination with Paclitaxel eliminates the drug
resistant CSC like cell population

Treatment of MDA-MB-468 with 10nM Paclitaxel plus 10uM ICG-001 completely
abolished the occurrence of drug resistant cell colonies compared to Paclitaxel treatment
alone (DMSO 31 ± 4 colonies vs. ICG-001 0 colonies, p < 0.05) (Figure 10A).
Retreatment of Paclitaxel resistant CSC like MDA-MB-468 cells (post 5d 10nM
Paclitaxel treatment) with combination therapy of 10nM Paclitaxel plus 10μM ICG-001
leads to a significant reduction of drug resistant cells, while Paclitaxel alone does not
affect the number of drug resistant cells (DMSO 29 ± 4 colonies vs. ICG-001 3 ± 1, 8.9-
fold reduction, p < 0.05) (Figure 10B). Gene expression analysis via qPCR of drug
resistance and stem cell markers (MDR1, ABCG2, CD24, CD44, MMP9) showed
resistant MDA-MB-468 cells after 6d of treatment with 10nM Paclitaxel overexpress
these genes compared to treatment naïve cells, and that treatment with 10uM ICG-001 for
24h or 48h significantly reduced expression of these genes (Figure 10C).
Treatment with ICG-001 reduces CSC side population cells in MDA-MB-231 (DMSO
0.93 ± 0.06% vs. ICG-001 0.23 ± 0.06%, p +0.0002) and MDA-MB-468 (DMSO 2.07 ±
0.28% vs. ICG-001 1.07 ± 0.2%, p = 0.012) TNBC cells lines, as well as patient derived
xenograft (PDX) TNBC cells (DMSO 5.1% DMSO vs. ICG-001 0.13 ± 0.06%, p =
0.0003) compared to DMSO vehicle control (Figure 9D).

79

Figure 10A: Effect of 5-day treatment with Paclitaxel plus ICG-001 on resistant cell outgrowth compared to Paclitaxel only
(N=3 per condition). 10B: Effect of re-treatment of Paclitaxel resistant MDA-MB-468 (post 5 days Paclitaxel) with Paclitaxel
only or Paclitaxel plus ICG-001 (N=3 per condition). 10C: Gene expression in Paclitaxel resistant MDA-MB-468 (post 5-day
treatment) treated with 10μM ICG-001 or DMSO control (n=3 per condition per time point). 10D: Effect of 24h 10μM ICG-001
on SP cells in MDA-MB-468, MDA-MB-231 and patient derived xenograft (PDX) TNBC cells (N=3 per cell model and
condition). Asterisk represents statistical significance (* p < 0.05, *** p < 0.001).

10C
10D
5 days
Paclitaxel
Paclitaxel + ICG-001
Taxol Taxol+ICG-001
0
10
20
30
40
Number of colonies
Taxol Taxol+ICG-001
0
10
20
30
40
Number of colonies
Number of colonies

Paclitaxel Paclitaxel
+ ICG-001
Paclitaxel Paclitaxel
+ ICG-001

Cell number * 10E4
40
30
20
10
0
40
30
20
10
0
No
colonies
10A
10B
*
*
control DMSO ICG-001
0
1
2
3
% SP cells
DMSO ICG-001 Control
MDA-MB-468

SP% positive cells
3
2
1
0
*
MDA231 side pop 24h treatment
verapamil DMSO ICG-001
0.0
0.5
1.0
1.5
% SP cells
MDA-MB-231
DMSO ICG-001 Control

SP% positive cells
1.5
1.0
0.5
0
*

control DMSO ICG-001
0
2
4
6
% SP cells
PDX cells
DMSO ICG-001 Control

SP% positive cells
6
4
2
0
*** ***
MMP19
MDR1
ABCG2
STMN
CD24
CD44
NES
0
2
4
6
12
16
20
fold change
Paclitaxel vs. untreated
24 ICG-001 vs. Paclitaxel
48h ICG-001 vs. Paclitaxel
80
3.10. Treatment with ICG-001 reduces tumor initiation of TNBC cells in vitro

The mammosphere assay is used to determine the tumor initiating capacity of cancer cells
in suspension culture. MDA-MB-231 cells were cultured in triplicate (20,000 cells per
well) in 6-well ultra-low adherence dishes for 7 days with either 10μM ICG-001 or
DMSO vehicle control. Mammospheres were counted based on size (small < 50μm, large
> 50μm). Cells treated with ICG-001 showed a statistically significantly reduced number
of mammospheres compared to vehicle control treated cells (Large spheres DMSO 43 ± 3
vs. ICG-001 0, p < 0.0001; small spheres DMSO 36 ± 4 vs. ICG-001 22 ± 3, p < 0.001)
(Figure 11A).

large small
0
10
20
30
40
50
DMSO ICG-001
# of mammospheres
Large Small

# mammospheres
50
40
30
20
10
0
DMSO

ICG-001
No
spheres
**
DMSO ICG-001
Figure 11A: Mammosphere formation of MDA-MB-231 cells treated for 7 days with 10μM ICG-001
or DMSO control (20,000 cells per well, N=3 per condition) (scale bar = 100μm). Asterisk represents
statistical significance (** p < 0.01).
11A
81
To exclude whether effects on mammosphere formation were due to toxic effects or
effects on overall cell proliferation, MDA-MB-231 cells were pre-treated for 48h with
either 10μM ICG-001 or DMSO and subsequently cultured for 7 days in suspension
culture to form the first generation of mammospheres without further treatment. After 7
days’ cells were dissociated and equal cell numbers (20,000 cells per well) plated in
triplicate per condition (ICG-001 or DMSO) for the second generation of mammospheres
without further treatment. While no significant difference in the number of first
generation mammospheres was observed between ICG-001 and DMSO pre-treated cells
(DMSO 39 ± 6 vs. ICG-001 41 ± 6, n.s.) (Figure 11B), cells pre-treated with ICG-001
showed a statistically significantly reduced number of second-generation mammospheres
compared to DMSO pre-treated cells (large spheres DMSO 20 ± 3 vs. ICG-001 12 ± 2, p
< 0.01; small spheres DMSO 36 ± 3 vs. ICG-001 22 ± 2, p < 0.001)) (Figure 11C).

Figure 11B: First generation mammosphere culture of MDA-MB-231 cells pre-treated for 48h with 10μM ICG-
001 or DMSO vehicle and cultured for 7 days (20,000 cells per well, N=3 per condition). 11C: Second generation
7 days mammosphere culture of MDA-MB-231 (pre-treated for 48h with ICG-001 or DMSO and cultured for 7
days) without addition of treatment (20,000 cells per well, N=3 per condition) (scale bar = 100μm). Asterisk
represents statistical significance (** p < 0.01).
large small
0
10
20
30
40
DMSO ICG-001
# of mammospheres
passaging
0
10
20
30
40
50
DMSO ICG-001
# of mammospheres
1
st
generation 2
nd
generation
# mammospheres
# mammospheres
large
small
50
40
30
20
10
0
40
30
20
10
0
DMSO ICG-001
DMSO ICG-001
**
*
large small
11B
11C
82
Conditional siRNA knockdown of FOXM1, β-catenin and CBP demonstrated a reduction
in primary and secondary mammosphere formation compared to scramble transfected
cells (primary spheres: scramble 76 ± 6, F 65 ± 5, F + B 54 ± 4, F + C 53 ± 3, F + B + C
48 ± 4, C + B 47 ± 3, p < 0.05) (Figure 10D) (secondary large spheres: scramble 66 ± 5,
F 48 ± 4, F + B 44 ± 4, F + C 34 ± 2, F + B + C 41 ± 4, C + B 43 ± 3,) (Figure 11E and
Table 8).

Knockdown/Size Large Small
SCR vs. F p < 0.0001 p < 0.0001
SCR vs. F+B p < 0.0001 p < 0.0001
SCR vs. F+C p < 0.0001 p < 0.0001
SCR vs. F+B+C p < 0.0001 p < 0.0001
SCR vs. C+B p < 0.0001 p < 0.0001
SCR F F+B F+C F+B+C C+B
0
20
40
60
80
100
Large Small
# of mammospheres
Scramble F F+B F+C F+B+C C+B

2
nd
generation

100
80
60
40
20
0
Figure 11D first generation and 11E second-generation 7-day mammosphere culture of MDA-MB-231 after
48h siRNA KD (F – FOXM1, B – b-catenin, C – CBP) (20,000 cells per well, N=3 per condition). Asterisk
represents statistical significance (* p < 0.05).

Scarmble F F+B F+C F+B+C C+B
0
20
40
60
80
100
1
st
generation
# mammospheres
Scramble F F+B F+C F+B+C C+B

100
80
60
40
20
0
*
11D
11E
Table 8: Mammosphere size difference after 48h
siRNA gene knockdown.
83
small large
0
10
20
30
40
DMSO
ICG-001
# of mammospheres
**
Hs578T
DMSO ICG-001
Large Small
DMSO
ICG-001
# mammospheres
40
30
20
10
0
Treatment of BT20, CAL51, MDA-MB436, Hs578T, MDA-MB-453, HCC1987 and
MDA-MB-231 cells with ICG-001 compared with DMSO control over seven and
fourteen days significantly reduced or eliminated the number of mammospheres (figure
11F and table 9).

Figure 11F: Mammosphere formation in TNBC cell lines (MDA-MB-436, CAL51, Hs578T and SUM149) after 7 days in
culture, treated with 10μM ICG-001 or DMSO vehicle control. (N=3 per cell line per condition). Large spheres < 50μm
diameter, small sphered > 50μm (scale bar = 100μm). Asterisk represents statistical significance (** p < 0.01, *** p < 0.001).

small large
0
20
40
60
80
DMSO
ICG-001
# of mammospheres
***
MDA-MB-436
No
spheres
DMSO ICG-001
Large Small
DMSO
ICG-001
80
60
40
20
0
# mammospheres
small large
0
10
20
30
40
DMSO
ICG-001
# of mammospheres
CAL51
***
No
spheres
DMSO ICG-001
Large Small
DMSO
ICG-001
# mammospheres
40
30
20
10
0
small large
0
10
20
30
40
DMSO
ICG-001
# of mammospheres
SUM149
No
spheres
DMSO
ICG-001
Large Small
DMSO
ICG-001
# mammospheres
40
30
20
10
0
11F
84

Cell lines
Sphere
size
# of Spheres
DMSO
# of Spheres
ICG-001
MDA-MB-436
Small
Large
66 ± 4
26 ± 3
24 ± 6
0
CAL51
Small
Large
33 ± 3
8 ± 2
9 ± 1
0
Hs578T
Small
Large
28 ± 3
23 ± 6
17 ± 3
3 ± 1
SUM149
Small
Large
32 ± 3
27 ± 2
20 ± 2
0

3.11. Treatment with ICG-001 sensitizes TNBC cells to Paclitaxel treatment

Cell counts performed after treatment of MDA-MB-231 for 48h with 10nM Paclitaxel
with or without the addition of 10μM ICG-001 confirmed that the combination treatment
of Paclitaxel plus ICG-001 is most effective in killing cells (DMSO vs. 10nM Paclitaxel
plus 10μM ICG-001 p = 0.0003) (Figure 12A).

Table 9: Mammosphere count in TNBC cell lines treated with ICG-001 or DMSO control.

85

Conditional siRNA gene KD in MDA-MB-231 for 48h and subsequent treatment for 48h
recapitulated the effects of ICG-001 on live cell numbers (comparison of RLU between
control and 10nM Paclitaxel treatment: F + B p < 0.05, F + C p < 0.001, C + B p < 0.05)
(Figure 12B).

Figure 12A: Cell count after treatment of MDA-MB-231 for 48h (N=3 per condition). Asterisk
represents statistical significance (* p < 0.05)
Figure 12B: Live cell quantification in MDA-MB-231 after 48h siRNA KD and 48h treatment with 10nM
Paclitaxel or PBS vehicle control (N=3 per KD condition per treatment). (F – FOXM1, B – b-catenin, C –
CBP) (10,000 cells per well). Asterisk represents statistical significance (* p < 0.05, *** p < 0.001).

SCR FOXM1 (F) F+CTNNB1 (B) F+CBP (C) F+B+C C+B
0
100
200
300
400
Control
10nM Paclitaxel
*
***
*
RLU

x

Control
Paclitaxel

RLU
400
300
200
100
0
12B
0
20
40
60
80
100
cell number ´ 10
3
10nM taxol
DMSO
1uM taxol
ICG-001
ICG-001+10nM
ICG-001+1uM

*

DMSO
10nM Paclitaxel
1μM Paclitaxel

ICG-001
ICG-001 + 10nM Paclitaxel
ICG-001 + 1μM Paclitaxel

Cell number × 10
3

0
20
40
60
80
100
12A
86
0 10 20 30
0
500
1000
1500
2000
PBS
ICG-001
Paclitaxel + PBS
Paclitaxel + ICG-001
Tumor volume in mm
3
3.12. Treatment of MDA-MB-468 TNBC cell line xenografts
After 24 days of treatment, tumor volume in mice treated with a combination of
Paclitaxel plus ICG-001 showed a statistically significantly reduced tumor burden
compared to either Paclitaxel or ICG-001 alone, as well as compared to PBS vehicle
control treated animals (Figure 13A and Table 10).

To demonstrate the presence of tumor initiation cells after treatment, tumors were
removed, dissociated and implanted into healthy female NOD/SCID/gamma mice,
without further administration of any treatment. Figure 13B shows the results of this
secondary implantation experiment. The tumors previously treated with Paclitaxel plus
ICG-001 developed statistically significantly smaller tumors, even in the absence of any
Treatment
Mean
difference
(mm
3
)
Significance
PBS vs. ICG-001 55.82 n.s.
PBS vs.
Paclitaxel + PBS
548.1 p < 0.01
PBS vs.
Paclitaxel + ICG-001
1097 p < 0.0001
ICG-001 vs.
Paclitaxel + PBS
492.2 p < 0.01
ICG-001 vs.
Paclitaxel + ICG-001
1041 p < 0.0001
Paclitaxel + PBS vs.
Paclitaxel + ICG-001
549.1 p < 0.01
12A
Table 10: Comparison of tumor volume post 24
days; primary implantation MDA-MB-468
xenograft (N=4 mice per group).
Figure 13A: MDA-MB-468 TNBC cell xenograft in female Nod/SCID/gamma (NSG) mice treated for 26 days (N=5
mice per treatment condition).

87
0 10 20 30 40
0
500
1000
1500
2000
2500
3000
3500
4000
PBS
ICG-001
Paclitaxel + PBS
Paclitaxel + ICG-001
Tumor volume in mm
3
treatment, compared to tumors that received either single agent treatment (Paclitaxel or
ICG-001 alone) or PBS vehicle control only (Figure 13B and Table 11).

3.13. Treatment of PDX of TNBCs

Two patient derived xenograft mouse models from triple negative breast cancer patients
were established to demonstrate that FOXM1 expression correlates with the level of
CSC-like side population cells as well as response to chemotherapy and disease
recurrence after serial transplantation in vivo. A patient derived xenograft (Patient 1) with
relatively lower levels of FOXM1 contains only a small fraction of side population cells
Treatment
Mean
difference
(mm
3
)
Significance
PBS vs. ICG-001 -354.6 n.s.
PBS vs.
Paclitaxel + PBS
-683.4 p < 0.05
PBS vs.
Paclitaxel + ICG-001
753.5 p < 0.05
ICG-001 vs.
Paclitaxel + PBS
-328.9 n.s.
ICG-001 vs.
Paclitaxel + ICG-001
1108 p < 0.001
Paclitaxel + PBS vs.
Paclitaxel + ICG-001
1437 p < 0.0001
13B
Table 11: Comparison of tumor volume post 24 days;
secondary implantation MDA-MB-468 xenograft
(N=5 mice per treatment condition).
Figure 13B: Secondary implantation of MDA-MB-468 tumors from primary xenografts. Tumors were dissociated
after termination of the experiment and implanted into new female NSG mice without further treatment (N=5 mice
per treatment condition).

88
(Figure 14A), while a second PDX model (Patient 2) with relatively higher expression of
FOXM1 exhibits a higher proportion of side population cells (Figure 14B). Comparison
of clinical data with FOXM1 expression in tumor samples from both patients confirmed
that high levels of FOXM1 resulted in poor response to standard chemotherapy and poor
survival outcome (Table 12).

Patient Side Histology Stage Grade TNM Treatment Relapse Survival
Patient 1
Right
Infiltrating
Ductal
Carcinoma
IIb
poorly
differentiated
yT3N0Mx
NeoChemo AC->T
Surgery: MRM
Radiation
No Alive
Patient 2 Yes Deceased

Treatment of mice bearing xenografts derived from patient 1’s tumor showed that
Paclitaxel treatment alone was similarily effective as combination therapy of Paclitaxel
plus ICG-001 in reducing tumor burden comapred to PBS vehicle control of ICG-001
only (Figure 14B and Table 13).
0.2%

FoxM1
ACTB
Patient 1 Patient 2

5%
Figure 14A: FOXM1 protein expression and proportion of side population cells in cancer cells from two
TNBC patient.

14A
Table 12: Clinical characteristic for patient 1 and 2 used for PDX models of TNBC in NGS mice ((NeoChemo –
neaoadjuvant chemotherapy, AC – Doxorubicin and Cyclophosphamide, T – Taxol (Pacliatxel), MRM –
modified radical mastectomy)
89
0 6 12 18 24
0
500
1000
1500
2000
2500
PBS
Paclitaxel + PBS
ICG-001
Paclitaxel + ICG-001
days
Tumor volume in mm
3
0 5 8 12 15 19 21 24
0
500
1000
1500
2000
2500
3000
PBS
Paclitaxel + ICG-001
Paclitaxel + PBS
ICG-001
days
Tumor volume in mm
3

Secondary implantion of tumors from mice bearing PDX from Patient 1 and treated with
either PBS, ICG-001, Paclitaxel or Paclitaxel plus ICG-001 into a second cohort of
healthy mice showed that all treatment conditions led to a signficant reduction in tumor
outgrowth compared to PBS without further treatment (Figure 14C and Table 14).

Treatment
Mean
difference
(mm
3
)
Significance
PBS vs. ICG-001 112.1 n.s.
PBS vs.
Paclitaxel + PBS
1206 p < 0.01
PBS vs.
Paclitaxel + ICG-001
1332 p < 0.01
ICG-001 vs.
Paclitaxel + PBS
1094 p < 0.05
ICG-001 vs.
Paclitaxel + ICG-001
1220 p < 0.05
Paclitaxel + PBS vs.
Paclitaxel + ICG-001
126.1 n.s.
Treatment
Mean
difference
(mm
3
)
Significance
PBS vs. ICG-001 753.8 p < 0.01
PBS vs.
Paclitaxel + PBS
656.2 p < 0.01
PBS vs.
Paclitaxel + ICG-001
516 p < 0.05
ICG-001 vs.
Paclitaxel + PBS
-97.64 n.s.
ICG-001 vs.
Paclitaxel + ICG001
-237.8 n.s.
Paclitaxel + PBS vs.
Paclitaxel + ICG-001
-140.2 n.s.
Figure 14B: PDX patient 1 in female Nod/SCID/gamma (NSG) mice treated for 24 days (N=5 mice per treatment condition).

Figure 14C: Secondary implantation of PDX patient 1. Tumors were dissociated after termination of the experiment and
implanted into new female NSG mice without further treatment (N=5 mice per treatment condition).

14B
14C
Table 13: Comparison of tumor volume post 24 days;
primary implantation patient 1 PDX (N=4 mice for
ICG-001, N=5 per each other treatment conditions)
Table 14: Comparison of tumor volume post 24 days;
secondary implantation patient 1 PDX (N=3 mice for
ICG-001, N=4 per each other treatment condition)
90
0 6 12 17 21
0
500
1000
1500
2000
2500
3000
Paclitaxel + ICG-001
PBS
Paclitaxel + PBS
ICG-001
days
Tumor volume in mm
3
1 5 12 19 26
0
500
1000
1500
PBS
Paclitaxel + ICG-001
ICG-001
Paclitaxel + PBS
days
Tumor volume in mm
3

In contrast to patient 1, mice bearing tumors from patient 2 showed no statistically
significant reduction in tumor growth with Paclitaxel treatment alone, compared to
control. Treatment with Paclitaxel plus ICG-001 resulted in a statistically significant
reduction in tumor growth compared to PBS vehicle control (Figure 14D and Table 15).

Treatment
Mean
difference
(mm
3
)
Significance
PBS vs. ICG-001 655.8 p < 0.05
PBS vs.
Paclitaxel + PBS
155.1 n.s.
PBS vs.
Paclitaxel + ICG-001
642.4 p < 0.05
ICG-001 vs.
Paclitaxel + PBS
810.9 p < 0.01
ICG-001 vs.
Paclitaxel + ICG-001
1298 p < 0.0001
Paclitaxel + PBS vs.
Paclitaxel + ICG-001
487.3 n.s.
Treatment
Mean
difference
(mm
3
)
Significance
PBS vs. ICG-001 -74.42 n.s.
PBS vs.
Paclitaxel + PBS
274.7 n.s.
PBS vs.
Paclitaxel + ICG-001
554.1 p < 0.001
ICG-001 vs.
Paclitaxel + ICG-001
628.5 p < 0.0001
ICG-001 vs.
Paclitaxel + PBS
349.1 p < 0.05
Paclitaxel + PBS vs.
Paclitaxel + ICG-001
279.5 n.s.
Figure 14D: PDX patient 1 in female Nod/SCID/gamma (NSG) mice treated for 24 days (N=5 mice per treatment condition).

Figure 14E: Secondary implantation of PDX patient 1. Tumors were dissociated after termination of the experiment and
implanted into new female NSG mice without further treatment (N=5 mice per treatment condition).

14E
14D Table 15: Comparison of tumor volume post 24
days; primary implantation patient 2 PDX (N=4 mice
each for PBS and ICG-001, N=6 mice each for
Paclitaxel + PBS and Paclitaxel + ICG-001)
Table 16: Comparison of tumor volume post 24
days; secondary implantation patient 2 PDX (N=5
mice per group)
91
Re-implantation of tumors from mice bearing PDX from Patient 2 showed that only
tumors pre-treated with combination therapy (Paclitaxel plus ICG-001) showed a
statistically significantly reduced secondary engraftment compared to PBS (Figure 14E
and table 16). Interestingly, ICG-001 alone did not inhibit secondary engraftment (Figure
14E). These experiments further demonstrated that ICG-001 treatment in combination
with Paclitaxel sensitized FOXM1 high tumors to chemotherapeutic treatment and
inhibited disease recurrence in secondary transplantation experiments without further
administration of treatment.
3.14. Tissue micro array (TMA) covariance analysis of correlation between
FOXM1 and CBP protein expression and clinical parameters

All breast cancer cases provided on the CDP Breast Cancer Stage II Prognostic Tissue
Microarray were classified by receptor expression to distinguish TNBC (negative for ER,
PER and HER2) from other subtypes. Table 17 gives an overview of the percentage of
each breast cancer subtypes on the TMAs. After exclusion of cases for which sections
were missing or damaged as well as cases for which receptor expression (ER, PR and
HER2) was not available, a total of 316 tumor cores were evaluated for FOXM1 and CBP
expression via immunohistochemistry staining. The scoring was based on a three-tier
system in consultation with a pathologist at USC (0 and 1 = weakly positive, 2 = strongly
positive protein expression). Figure 15 shows representative cases for FOXM1 and CBP
staining on the TMAs. Only cases with a score 2 were counted as positive for subsequent
analysis (numbers are given in Table 18) and correlated to clinical parameters. Global
scoring results for FOXM1 and CBP are shown in tables 18.
92
Table 17: Breast cancer subtype distribution
Subtype Frequency %
ER+/PR- 24 7.7
ER-/PR+ 11 3.5
ER/PR+ 176 56.2
HER2+ 49 15.7
TNBC 53 16.9

Table 18: Staining results for FOXM1 and CBP
Protein Frequency %
FOXM1
Positive
Negative

95
218

30.4
69.6
CBP
Positive
Negative

136
168

44.7
55.3

Score 0 Score 1 Score 2
CBP FOXM1
Figure 15: Representative cores evaluated for FOXM1 and CBP staining. Staining scoring 0 = negative, 1 =
weakly positive, 2 = strongly positive (nuclear staining).

93
Chi square analysis demonstrated a significant correlation between TNBC subtype (p <
0.001) and ER/PR+ tumors (p = 0.045) and FOXM1 protein expression (Table 19).
Tumor grade and FOXM1 expression were statistically significantly correlated (p =
0.013) (Table 20). A statistically significant correlation was found between CBP protein
levels and HER2 positive status (p = 0.017) as well as positive lymph node metastasis
(N1) (p = 0.034) (Table 20).

Table 19: Statistical analysis (Chi square statistics – FOXM1 dependent variable)
Variable FOXM1 (+) FOXM1 (-) Total % positive p-value
Differentiation
Grade 1
Grade 2
Grade 3

7
41
47

39
110
79

46
151
126

15
27
37

0.013
Subtype
TNBC
ER+
ER/PR+
HER2+

25
4
43
13

28
20
134
30

53
24
177
43

47
17
24
30

< 0.001
0.177
0.045
0.8

Table 20: Chi square statistics (CBP dependent variable)

Variable CBP (+) CBP (-) Total % positive p-value
Subtype
TNBC
ER+
ER/PR+
HER2+

20
12
72
26

33
12
105
17

53
24
177
43

38
50
41
60

0.329
0.521
0.192
0.017
Nodal status
N0
N1

40
96

76
110

116
206

35
47

0.034
94
Logistic regression for multivariate analysis in XLSTAT software was performed for all
statistically significantly associated variables found in the chi square test. Using FOXM1
as the dependent variable, positive correlations were identified for all variables (results
given as odds ratio and [95% CI]: TNBC subtype 2.016 [1.05 to 3.85]; tumor grade II
2.79 [1.15 to 6.75]; tumor grade III 3.3 [1.34 to 8.12]; radiation therapy 1.74 [1.03 to
2.96]) (Table 21). A cox regression model analysis was performed to determine hazard
ratios (HRs) (table 19). Using FOXM1 as the dependent variable, statistically significant
HRs were found for TNBC subtype (HR 1.63 [1.07 to 2.49]), age younger than 50 years
(0.35 [0.23 to 0.53]), tumor stage (1.43 [0.996 to 2.04]) and tumor grade III (1.63 [1 to
2.65]) (Table 22).

Table 21: Multivariate Logistic Regression (FOXM1 dependent variable)
Source p-value Odds ratio
Odds ratio Lower
bound (95%)
Odds ratio Upper
bound (95%)
TNBC 0.034 2.02 1.06 3.85
Differentiation
Grade II
Grade III
0.023
0.01
2.79
3.3
1.15
1.34
6.75
8.12
Radiation therapy 0.04 1.74 1.03 2.96

Table 22: Cox Regression analysis (vital status plus overall survival in months)
Variable p-value Hazard ratio
Odds ratio Lower
bound (95%)
Odds ratio Upper
bound (95%)
TNBC 0.023 1.63 1.07 2.49
Age ≤ 50 years < 0.0001 0.035 0.23 0.53
Differentiation
Grade III
0.051 1.63 1 2.65
Tumor stage
T2
0.053 1.43 1 2.04
95
Using TNBC as the dependent variable revealed that FOXM1 and high tumor grade
(Grade III) are statistically significantly correlated with this subtype (Table 23) and show
potential as a fair biomarker (Figure 16).

Table 23: Multivariate Logistic Regression (TNBC dependent variable)
Source p-value Odds ratio
Odds ratio Lower
bound (95%)
Odds ratio Upper
bound (95%)
FOXM1
positive
0.031 2.03 1.067 3.86
Grade III 0.002 3.79 2.34 44.44

Figure 16: Receiver operating characteristic (ROC) curve for FOXM1 and tumor
grade III combined as predictors of TNBC subtype.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 0.2 0.4 0.6 0.8 1
Sensitivity
Specificity
ROC Curve (AUC = 0.751)
96
Kaplan-Meier 5-year survival analysis of the TMA data for association between FOXM1
levels and overall survival demonstrated no statistically significant link in all breast
cancer samples (N=316) (p = 0.144) (Figure 17A and tables 24 and 25). Stratification of
the data to include only TNBC cases (N=53, 28 cases FOXM1 negative, 25 cases
FOXM1 positive) showed no relation between FOXM1 expression and OS (p = 0.834)
(Figure 17B and tables 26 and 27).
Analysis of the same data set for association between CBP protein expression and overall
survival similarly showed no statistically significant correlation between outcome and
protein expression levels of CBP (all cases p = 0.096, TNBC p = 0.300) (Figure 18 and
tables 28 though 31).
A similar analysis of the TCGA RPPA breast cancer data revealed that high FOXM1
expression shows a trend towards worse 5-year overall survival prognosis (p = 0.059) for
all cases (n=817) irrespective of subtype (Figure 19A and tables 32 and 33). Selecting
TNBC cases only (n=82) revealed a similar trend, albeit not statistically significant (p =
0.214) (Figure 19B and tables 34 and 35). No expression data was available for CBP.

97

Statistic
Observed
value
Critical value p-value alpha
Log-rank 0.044 3.841 0.834 0.050
Statistic
Observed
value
Critical value p-value alpha
Log-rank 2.132 3.841 0.144 0.050
Marker
expression
Mean survival
time (OS)
Standard
deviation
Lower bound
(95%)
Upper bound
(95%)
FOXM1 (+) 50.947 1.389 48.226 53.669
FOXM1 (-) 51.962 0.788 50.418 53.505
Marker
expression
Mean survival
time (OS)
Standard
deviation
Lower bound
(95%)
Upper bound
(95%)
FOXM1 (+) 45.960 3.129 39.827 52.093
FOXM1 (-) 47.793 2.994 41.924 53.661
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 10 20 30 40 50 60 70
months
5-year Survival Distribution (FOXM1 all)
p = 0.144
FOXM1 negative
FOXM1 positive
Table 24: Statistical analysis of mean OS all cases based on FOXM1 expression

Figure 17A: TMA Kaplan Meier overall survival analysis of FOXM1 for all cases (N=817). 17B: TMA Kaplan-Meier overall survival analysis for FOXM1 for TNBC cases (N=82)

17A
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 10 20 30 40 50 60 70
months
5-year Survival distribution (FOXM1 TNBC)
p = 0.834
FOXM1 negative
FOXM1 positive
17B
Table 25: TMA data set all cases mean OS based on FOXM1 expression

Table 26: TMA data set TNBC cases mean OS based on FOXM1 expression

Table 27: Statistical analysis of mean OS TNBC based on FOXM1 expression

98

Marker
expression
Mean survival
time (OS)
Standard
deviation
Lower bound
(95%)
Upper bound
(95%)
CBP (+) 42.433 3.829 34.929 49.937
CBP (-) 49.719 2.560 44.701 54.737
Statistic
Observed
value
Critical value p-value alpha
Log-rank 1.075 3.841 0.300 0.050
Marker
expression
Mean survival
time (OS)
Standard
deviation
Lower bound
(95%)
Upper bound
(95%)
FOXM1 (+) 50.708 1.175 48.405 53.011
FOXM1 (-) 52.657 0.837 51.016 54.298
Statistic
Observed
value
Critical value p-value alpha
Log-rank 2.772 3.841 0.096 0.050
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 10 20 30 40 50 60 70
months
5-year Survival Distribution (CBP all)
CBP negative
CBP positive p = 0.096
18A 18B
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 10 20 30 40 50 60 70
months
5-year Survival Distribution (CBP TNBC)
CBP negative
CBP positive p = 0.300
Figure 18A: TMA Kaplan Meier overall survival analysis of CBP for all cases (N=817). 18B: TMA Kaplan-Meier overall survival analysis for CBP for TNBC cases (N=82)

Table 28: TMA data set all cases mean OS based on CBP expression
Table 30: Statistical analysis of mean OS all cases based on CBP expression
Table 29: TMA data set TNBC cases mean OS based on CBP expression
Table 31: Statistical analysis of mean OS TNBC based on CBP expression
99

Marker
expression
Mean survival
time (OS)
Standard
deviation
Lower bound
(95%)
Upper bound
(95%)
FOXM1 (+) 48.878 3.186 42.635 55.122
FOXM1 (-) 55.302 0.508 54.306 56.299
Marker
expression
Mean survival
time (OS)
Standard
deviation
Lower bound
(95%)
Upper bound
(95%)
FOXM1 (+) 42.889 6.545 30.061 55.718
FOXM1 (-) 50.074 1.955 46.243 53.905
Statistic
Observed
value
Critical value p-value alpha
Log-rank 3.577 3.841 0.059 0.050
Statistic
Observed
value
Critical value p-value alpha
Log-rank 1.543 3.841 0.214 0.050
Figure 19A: TMA Kaplan Meier overall survival analysis of FOXM1 for TNBC cases (N=82). 19B: TMA Kaplan-Meier overall survival analysis for FOXM1 for all cases (N=817)

19A 19B
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 10 20 30 40 50 60 70
months
5-year Survival Distribution (FOXM1 TNBC)
p = 0.214
FOXM1 negative
FOXM1 positive
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 10 20 30 40 50 60 70
months
5-year Survival Distribution (FOXM1 all)
p = 0.059
FOXM1 negative
FOXM1 positive
Table 35: TCGA data set TNBC cases mean OS based on FOXM1 expression
Table 34: Statistical analysis of mean OS TNBC based on FOXM1 expression
Table 33: TCGA data set all cases mean OS based on FOXM1 expression
Table 32: Statistical analysis of mean OS all cases based on FOXM1 expression
100
4. DISCUSSION
CBP is an important co-activator in β-catenin (CTNNB1) driven transcription, including
the Wnt signaling pathway [352, 353], which has been implicated in TNBC biology
[354]. The Kahn lab has developed a specific CBP-binding small molecule inhibitor,
ICG-001 [346]. We hypothesized that CBP-signaling plays an important role in TNBC
biology and may provide a novel therapeutic target.
Publicly available data sets and protein quantification in TNBC cell line models
demonstrated CBP over-expression in breast cancer, and particularly in TNBC.
Luciferase reporter assays showed that gene transcription in TNBC is CBP/β-catenin-
dependent, but not Wnt/β-catenin/TCF-dependent, and can be decreased via ICG-001.
RNA Seq analysis of TNBC cells treated with ICG-001 discovered Forkhead box M1
(FOXM1) as a potential downstream regulator of CBP signaling, as well as down-
regulation of FOXM1 target genes over-expressed in TNBC. CoIP demonstrated that
CBP and FOXM1 interact in TNBC cells. Treatment with ICG-001 disrupted
CBP/FOXM1 binding, and led to a reduction in protein levels of FOXM1 and β-catenin.
qPCR gene expression and luciferase reporter activity analysis demonstrated a decrease
in FOXM1-driven gene expression after treatment with ICG-001 in multiple TNBC cell
lines. Transient gene knockdown (KD) and artificial over-expression confirmed that the
effect of ICG-001 is FOXM1 dependent. In vitro viability assays demonstrated that
treatment with ICG-001, as well as FOXM1 and β-catenin KD sensitized TNBC cells to
Paclitaxel treatment, and eliminated drug resistant CSC cell populations.
PDX mouse models of TNBC demonstrated that FOXM1 expression correlated with
response to chemotherapy and disease recurrence in vivo. Treatment with ICG-001
101
sensitized FOXM1
high
tumors to Paclitaxel treatment, and reduced tumor growth in serial
transplantation experiments.
Staining of clinically annotated TMAs for FOXM1 and CBP showed that high levels
FOXM1 protein were associated with TNBC subtype, high tumor grade and radiation
therapy, while CBP was associated with HER2+ subtype and lymph node metastasis.
Kaplan-Meier analysis of the TMAs and TCGA RPPA data showed a trend towards high
expression of FOXM1 and CBP protein expression and worse overall survival outcome.
The Kahn lab and others have demonstrated that disrupting CBP/β-catenin with the small
molecule inhibitor ICG-001 has beneficial effects in various disease states, such as
fibrosis [355, 356] and heart function [357], but in particular in cancer and cancer stem
cells [215, 248, 358-360]. Since neither CBP nor β-catenin directly binds to DNA, which
is ultimately necessary for transcriptional activity and to drive gene expression, the
current study expanded the mechanistic understanding how ICG-001 effects CBP-driven
gene expression.
ICG-001 was shown to not only affect CBP binding to β-catenin, but also binding of CBP
to the transcription factor FOXM1. These results do not challenge the specificity of ICG-
001 binding to CBP since the CoIP results demonstrated that FOXM1/β-catenin binding
was not disrupted with ICG-001 treatment.
Transcription factors are overrepresented among oncogenes [279], making them a
desirable target for cancer therapy [280]. Yet specifically targeting TFs has proven to be
formidably difficult [361, 362], calling the druggability of TFs into question [363, 364].
Other oncoproteins, such as kinases, have specific substrate binding pockets and have
been successfully targeted via small molecule inhibitors based on a “lock-and-key”
principle. No such pockets exist on most TFs. Additionally, their hydrophobic surfaces
102
render the development of targeted drugs a significant challenge [364]. The DNA binding
domain (DBD) of TFs has been explored as a potential target site for small molecule
inhibitors, but is usually too large and too similar between different TF families to be
targeted selectively and specifically.
Hence, approaches to target TFs have focused either on inhibition via gene knock-down
using siRNA, or by targeting protein-protein interactions between TFs and essential co-
activator proteins [364-367], dimerization partners [368, 369], or chaperone and shuttling
proteins [370], successfully inhibiting tumorigenicity [371].
FOXM1 has been targeted in in vitro models using a variety of the aforementioned
approaches. RNAi mediated knockdown of FOXM1 has been shown to effect cancer cell
growth and reduce metastatic potential in vitro [281, 282, 372], and can affect FOXM1
function in vivo [373, 374]. Proteasome inhibitors, such as siomycin A and thiostrepton,
have shown promising potential inhibiting FOXM1 transcriptional activity and oncogenic
function in pre-clinical studies [375-377]. Despite success in in vitro and in vivo studies
of inhibiting FOXM1 driven tumorigenesis, drug resistance and metastatic potential of
cancer cells by these targeted approaches, major obstacles remain. For example,
thiostrepton, a compound originally developed as an antibiotic, inhibits mitochondrial
translation, with deleterious effects on metabolism in all cells non-specifically [378].
RNAi-based drugs also face the danger of off-target effects, as well as inflammatory
reactions and the overall challenge of efficient delivery in vivo [379].
A potential alternative strategy to targeting FOXM1 directly would be via the disruption
of protein-protein interaction with important co-factors. The role of CBP binding with
FOXM1 has been shown previously to be of importance for FOXM1 transcriptional
activity [269]. This study demonstrated the validity of this approach by demonstrating
103
that disruption of CBP/FOXM1 interaction leads to reduction in FOXM1 transcriptional
activity.
Detailed investigation of the mechanistic effect of ICG-001 on FOXM1-driven gene
transcription demonstrated that although FOXM1/CBP-binding is disrupted already after
4h ICG-001 treatment, gene expression was not affected. Only after 24h, when FOXM1
protein levels were reduced, did expression of target genes change significantly. These
results indicate that CBP not only acts as an activator of FOXM1 transcriptional activity,
but also plays a role in maintaining FOXM1 levels, which is critical for FOXM1 activity.
CBP has been shown to bind to hundreds of proteins, resulting in a complexity of
functional consequences that could be investigated using the chemical-genomic approach
applied in this study. It would be of interest to further investigate the role DNA-binding
TFs play in the transcriptional complex formed by CBP and β-catenin in other
malignancies or pathologies.
The Kahn lab and others have previously published on the effect of ICG-001 on CSC cell
populations [215, 216, 358]. TICs have been shown to be responsible for mammosphere
formation [380, 381]. Since ICG-001 has not been shown to be cytotoxic per se, it could
potentially affect the mode of division of CSCs. ICG-001 potentially forces CSC
differentiation by switching the cell-division program from asymmetric or symmetric
non-differentiative cell division, which would maintain or increase stem cell numbers,
respectively, to symmetric differentiation, which would deplete stem cells over time
[382-384]. Testing ICG-001 in isolated CSC populations and comprehensive
investigation (e.g. whole transcriptome RNA Seq) of the underlying mechanism could
shed further light on how ICG-001 affects CSC populations, and possibly yield a better
understanding of CSC biology more generally.
104
In the current study the ICG-001 was tested in two different conditions, continuous
treatment for 7 days, or pre-treatment for 48h without the addition of drug for the
remaining culture period. ICG-001 inhibited mammosphere formation more efficiently in
continuous treatment compared to pre-treated populations.
A possible explanation could be that ICG-001 has cytostatic effects in the continuous
culture on top of effects on CSC mode of division. Furthermore, in the pre-treated cells,
TIC cells must be depleted in the first-generation culture. This could explain why reduced
mammospheres were only observed in the second generation after pre-treatment. The
effect could also be less pronounced in the pre-treatment conditions due to residual CSC
after only 48h treatment. Wash out of ICG-001 could lead to ineffective drug dosage with
reestablishment of FOXM1/CBP binding and baseline FOXM1 levels, allowing for the
re-emergence and maintenance of CSC/TIC populations in subsequent culture.
Additionally, cancer cells have been shown to possess high plasticity [385, 386] and
recapitulate parental cell phenotype after sorting into different populations over time in
culture [387].
It would be of interest to further elucidate the mechanism by which ICG-001 eliminates
CSC populations. Future studies could explore treatment schedules for ICG-001 to
potentially completely eradicate CSC populations, or to investigate if reemergence of
CSC can be suppressed over long periods of time, or even abolished.
Transient knockdown of FOXM1, CBP and β-catenin demonstrated that all combinations
had similar effects on mammosphere formation in the first and second generation. In the
first generation, the effect was not as pronounced as with continuous ICG-001 treatment.
This could be due to the same reason as stated for 48h pre-treatment with ICG-001, with
the transient knockdown effect wearing off relative to the duration of the culture (7 days).
105
Additionally, gene knockdown supposedly does not affect the interaction of FOXM1 and
CBP, which might be important for the CSC phenotype.
It is interesting that CBP and β-catenin combined knockdown had the strongest effect on
mammosphere formation. Perhaps other TFs binding in this complex play a role in
CSC/TIC phenotype and reduced levels of both proteins affect their function. This could
be the subject of future studies to gain more comprehensive insides into CSC-related
transcriptional processes.
A hallmark of drug resistant cancer cells and CSC is the up-regulation of efflux proteins
on the cell surface such as MDR1 and ABCG2, also known as BCRP (breast cancer
resistance protein) and can be identified via a fluorescent dye efflux FACS assay [351].
Furthermore, the CD44
high
CD24
low
phenotype has been described to mark CSC like cell
populations in breast cancer [137]. These cells have been demonstrated to be highly drug
resistant, and capable of tumor initiation at very low numbers [388-390]. The CSC/TIC
phenotype has been associated with high FOXM1 levels in breast cancer [294].
The current study showed that the CSC phenotype is increased in Paclitaxel resistant
TNBC cells, which was associated with increased FOXM1 levels. Treatment with ICG-
001 in combination with Paclitaxel prevented the occurrence of drug resistant CSC-like
cell population and re-sensitized pre-treated resistant cells to Paclitaxel. In vivo studies
using a TNBC cell line showed that combination treatment with Paclitaxel plus ICG-001
was most efficient in controlling tumor outgrowth during treatment, as well as secondary
outgrowths, which was used as a surrogate for disease recurrence [391].
Nevertheless, tumors re-grew in all treatment conditions, including combination therapy.
It would be interesting in future studies to further investigate whether outgrowth after
secondary implantation was indeed due to mostly CSC-like persisting cells and if
106
different therapeutic regimen in combination with ICG-001 could completely eradicate
secondary outgrowth (e.g. recurrence).
Two PDX models of TNBC demonstrated that sensitivity to Paclitaxel treatment
appeared to be associated with FOXM1 protein levels. Tumors expressing high levels of
FOXM1 (Patient 2) were more resistant to Paclitaxel only treatment and presented higher
recurrence rates (secondary implantation). Treatment in combination with ICG-001
ameliorated these unfavorable outcomes.
The limited number of cell lines and PDX models restrict the power of the current study
to extrapolate the results for broader clinical implications. For the in vivo studies, one
TNBC cell line xenograft (MDA-MB-468) and two TNBC patient PDX were used in
NSG mice to establish the effect of CBP inhibition via ICG-001 in TNBC on drug
resistance and recurrence. Given the heterogeneity of TNBC, it would be valuable to
expand these studies using additional cell lines, potentially based on the classification by
Lehmann et al [58, 59] to obtain more generalizable results.
For both patients, whose tumors were used for PDX models in this study, FOXM1
expression levels correlated with clinical outcome. Both patients were diagnosed with
TNBC (right breast, infiltrating ductal carcinoma, T3N0Mx (stage IIb), poorly
differentiated) and received similar treatments (neoadjuvant chemotherapy with
Doxorubicin and Cyclophosphamide followed by taxane treatment, modified radical
mastectomy and radiation therapy). While patient 1 with lower FOXM1 levels is alive,
patient 2 died despite the same diagnosis and treatment. The clinical outcome data
together with results obtained here in PDX mouse models of both patients’ disease
support the notion of the potential benefit of implementing FOXM1-targeting agents into
107
TNBC treatment. A larger study with additional tumors from patients with TNBC could
firmly establish the data produced here.
A potential limitation of this study is the use of only a single chemotherapeutic agent in
our study (Paclitaxel). Other compounds have been used or are currently under
investigation in TNBC, such as DNA-damaging agents [392, 393]. FOXM1 has been
associated with resistance to these drugs [289, 291, 394] and the effect of ICG-001 in
combination with DNA damaging drugs could be a potentially attractive line of research.
The TMA analysis demonstrated a correlation between radiation therapy and FOXM1
protein levels in breast cancer. Radiation therapy is a standard of care procedure in
patients who receive breast-conserving surgery [395]. FOXM1 has been shown to
mediate resistance to radiation therapy [295]. Investigating the effect of ICG-001
treatment in ameliorating resistance of cancer cells radiation would be of interest to
potentially improve outcome for this line of treatment, reducing the rate of recurrence and
necessity for more aggressive and repeated treatment.
Many targeted therapies have been developed for various cancers [396], including breast
cancer (such as ER and HER2 targeted [22, 397, 398] approaches). These strategies had
tremendous positive effects on the clinical management and patient outcome.
Nevertheless, targeted approaches are often of limited success due to the occurrence of
mutations in the targeted genes and proteins [399].
The limited scope of the current study does not allow for the evaluation of the potential
occurrence of CBP mutations that might render cancer cells resistant to treatment with
ICG-001. Several studies have described mutations in CBP, including germline mutations
(in AML [210], and mixed lineage leukemia (MLL) [213]), as well as somatic mutations
in breast, colorectal, ovarian and pancreatic cancer [212, 214, 400, 401]. Future studies
108
could aim at a broader investigation of specific CBP mutations in cell line models and
their effect on ICG-001 binding to CBP.
Canonical Wnt signaling has been shown to be an important driver various cancers,
particularly in colorectal cancer [162, 319, 321]. Despite this well-established fact,
Kuroda et al. have previously demonstrated that the TOP/FOPFLASH ratio as a readout
for Wnt/TCF/β-catenin signaling varies greatly between colorectal cancer (CRC) cell
lines, such as SW480 and HCT116. These authors demonstrated that SW480 has a high
ratio of TOP/FOPFLASH (62.1), while HCT116 showed a comparatively low ration of
1.7 [348]. These two cell lines have different alterations in the Wnt pathway, SW480 has
a mutation in APC which renders the β-catenin destruction complex dysfunctional, while
HCT116 carries a mutation in β-catenin that allows it to escape targeting for degradation.
Both alterations lead to accumulation and nuclear translocation of β-catenin. Since
TOPFLASH only measures TCF transcriptional activity, it is an indirect readout for β-
catenin /co-activator/TF complex. The results in this study indicated that β-catenin could
bind to different TF, such as FOXM1, for transcriptional activity. It would be interesting
to investigate other cell types, such as HCT116 for potential other TF binding partners,
and potentially broaden the definition of Wnt signaling, as defined by accumulation and
transcriptional activity of β-catenin.
Both Wnt signaling and FOXM1 transcriptional activity have been previously shown to
drive CSC phenotype and drug resistance in several malignancies [228, 296, 320, 322,
359]. A study by Zhang et al. in glioblastoma had established the combined role of both
FOXM1 and TCF in Wnt signaling [297, 323]. The authors of this study demonstrated
that FOXM1 was important to shuttle β-catenin into the nucleus, and that both FOXM1
and TCF were important in mediating Wnt signaling.
109
The current study showed that in TNBC, FOXM1 seems to be closely linked to β-catenin
levels, but that transcriptional activity of FOXM1 in complex with β-catenin and CBP
might be sufficient to drive CSC phenotype and drug resistance, without involvement of
TCF, as judged by low TOPFLASH activity and lack of response to a Wnt targeting
small molecule inhibitor (WNT974) [350]. Further studies could use gene knockdown for
TCF or transfection of a dominant negative TCF construct to investigate potential effects
on FOXM1-driven gene expression. Dissecting this interaction could elucidate further
how Wnt signaling (as defined by increased β-catenin levels and transcriptional activity)
is linked to TCF-signaling in TNBC.
One major limitation of the current study is a lack of explanation as to how FOXM1
levels are regulated by CBP and treatment with ICG-001. Halasai et al. proposed that
FOXM1 levels are regulated via an auto-regulatory loop in which FOXM1 binds to its
own promoter in a positive feedback loop [317]. In a preliminary study (data not shown)
for the present work, two FOXM1 consensus (TATA) binding sequences were detected
in the proximal promoter region of FOXM1 (within 1.5kb upstream of the ATG start site)
using BLAT on the NCBI website [402].
A future study should investigate the role of FOXM1, CBP and potentially β-catenin
binding in FOXM1 expression. A ChIP-PCR could be performed in cells treated with
ICG-001 or DMSO control and pull-down with FOXM1, CBP and β-catenin to see
whether these factors bind in the proximal FOXM1 promoter region, how treatment with
ICG-001 affects binding, and how this relates to changes in FOXM1 levels and activity.
Alternatively, Clustered regularly interspaced short palindromic repeats (CRISPR)-
mediated deletion of the consensus sites could be used to delete FOXM1 binding sites in
promoter regions of oncogenes or genes over-expressed in breast cancer.
110
The effects of ICG-001 on FOXM1 genome-wide binding could be investigated using
ChIP Seq in different TNBC cell lines. Coupled with ChIP Seq for CBP, these studies
could shed further light on the role of FOXM1/CBP-binding, target gene expression and
druggability of these target genes via ICG-001. Based on these results, in combination
with the mechanism proposed in this study, clinically actionable targets could be defined
and used to stratify patients for potential benefit of treatment with ICG-001, or inhibition
of FOXM1 more generally.
Although RNA Seq analysis of only one TNBC cell line treated with ICG-001 compared
to DMSO control yielded information about the importance of CBP/FOXM1 signaling in
TNBC, only a subset of genes was further investigated in a variety of TNBC cell lines
models using qPCR. TCGA data analysis demonstrated that the top 100 down-regulated
genes upon ICG-001 treatment in MDA-MB-231 are up-regulated in 99.8% of TNBC
and 100% of basal breast cancer in this data set.
The TNBC classification by Lehmann et al. revealed the tremendous heterogeneity within
this breast cancer subtype [58, 59]. The initial seven-subtype classification was
subsequently reduced to four, due to sample contamination with immune and stromal
cells in the initial analysis, perhaps demonstrating the caveats such studies entail. Despite
this issue, the findings put forward by this group were shown to have predictive and
prognostic validity [89]. Further investigation of how ICG-001 affects gene expression at
the transcriptome level in various TNBC cell lines representing all seven TNBC subtypes
is warranted. Such studies could elucidate the overlap between ICG-001 sensitivity and
TNBC subtypes, and potentially allow the definition of a genetic signature as a biomarker
for candidate patients for ICG-001, or by association FOXM1 targeted therapy.
111
FOXM1 has been associated with various clinical characteristics in breast cancer, such as
tumor stage and nodal status [312]. FOXM1 has also been shown to be associated with
ER+ tumors [310] as well as TN subtype [312, 403]. In the current study, the importance
of FOXM1 in drug resistance and CSC phenotype in TNBC was shown to be dependent
on FOXM1/CBP binding and can be ameliorated via ICG-001 small molecule mediated
disruption of this interaction.
Using TMA staining for FOXM1 demonstrated an association with TNBC subtype and
high tumor grade. Using FOXM1 and tumor grade as biomarkers, TNBC can be
predicted with 75% accuracy. Although negative receptor expression (for ER, PR and
HER2) can identify TNBC, no information can be gained about possible targeted
therapeutic options with this approach. In combination with the results from in vitro and
in vivo studies showing the beneficial effect of targeting FOXM1 in TNBC, using
FOXM1 as a biomarker for high grade TNBC could form the basis for anti-FOXM1
targeted therapy in these patients. A comparison of FOXM1 expression and
chemotherapy or survival did not yield statistically significant results in the TMA
analysis. Given that only 320 tumors were available, of which 52 were of TNBC subtype,
the numbers might be too small to establish an association.
The TMA analysis also confirmed an association between FOXM1 and ER+ tumors, and
combination treatment with ICG-001 could represent an interesting line of investigation
for this breast subtype, for example regarding endocrine resistance [294].
No association was found between CBP expression and TNBC subtype. Instead a
statistically significant association was found between high levels of CBP and HER2+
subtype. Wnt signaling has been shown to be a potential driver in HER2+ breast tumors
[241, 248], and ICG-001 treatment in a HER2-driven mouse model of breast cancer was
112
shown to target CSCs [248]. These results warrant further investigation of the potential
role of CBP/β-catenin and the effect of ICG-001 treatment in the biology of this HER2+
breast cancer.
Potentially the sample size in the TMA did not allow for establishment of the role of both
CBP and TNBC or FOXM1 and HER2+ tumors. Previous studies have demonstrated the
importance of FOXM1 in HER2+ tumors [404, 405]. Another possibility is that CBP is
important in HER2+ tumors, but mediates its transcriptional activity via binding to TFs
other than FOXM1.
While the survival analysis correlating FOXM1 or CBP protein expression with OS in the
TMA and TCGA data sets did not show drastically significant association, a trend
towards high expression of both markers with worse survival outcome was observed. The
larger number of samples in the TCGA data set compared with the TMAs for TNBC (82
vs. 53) and cases of all subtypes (817 vs. 316) showed that the larger data set (TCGA)
further increased the trend towards a statistically significant correlation between high
FOXM1 protein levels and worse OS outcome, with a p of 0.059 for all subtypes trending
towards statistical significance. No data was available for CBP in the TCGA data set and
analysis of the TMA data for an association between CBP expression and OS outcome
did not reach statistical significance.
It could be speculated that a larger number of breast cancer samples, and particularly
TNBC samples, could yield a statistically significant (p < 0.05) correlation between high
levels of FOXM1 or CBP protein expression and worse OS outcome. The above analysis
could also be extended to specific TNBC subtypes identified by Lehmann ER al. [58, 59]
and could reveal subtype specific differences in FOXM1 and CBP protein expression and
OS.
113
A further limitation is that FOXM1 was investigated as potential predictive biomarker in
the preclinical studies presented here. It is noteworthy that not all biomarkers are
simultaneously predictive and prognostic [14, 15]. For example, axillary lymph node
metastasis has been shown to prognostic of patient survival [406], but cannot make any
predictions about response to therapy. Mutations in epidermal growth factor receptor
(EGFR) on the other hand can predict therapy response [407].
A limitation of the current study is the focus on non-metastatic breast cancer. Metastasis
is the major cause of death in breast cancer patients. Metastatic breast cancer (MBC) is
considered incurable, with 5-year survival rates of less than 30% [408]. An analysis of
randomized clinical trials by Tevaarwerk at el. showed no improvement in survival for
MBC patients over a 30-year period [409]. TNBC have a disproportionately high
propensity to relapse and spread to distant organ sites (e.g. distant metastasis) [410]
Currently more than 80% of breast cancer patients receive adjuvant chemotherapy, but
only 40% relapse and die, leading to overtreatment with sometimes severe side effects
(e.g. neurotoxicity, cardiotoxicity) [411]. As mentioned, the only reliable clinical grade
biomarkers in breast cancer to date are ER, PR and HER2 [28]. MBC show up to 48%
discordance in the expression of ER, PR and HER2 between primary tumor and
metastatic sites [412-415]. These findings highlight the importance to identify reliable
biomarkers and potential molecular targets.
Studies have shown that cell with metastatic potential acquire stem cell (CSC) properties,
and that Wnt signaling in these cells controls stemness and metastatic colonization [416,
417]. Furthermore, FOXM1 is intricately related to tumor metastasis [301].
114
Taken the findings of the current study that FOXM1 is an important player in TNBC
biology and CSC stem cell phenotype, further pre-clinical studies are warranted
investigate the potential benefit of FOXM1-targeted therapy in MBC.
Accurate assessment of prognostic and predictive biomarkers (e.g., ER, PR, HER2) plays
a critical role in the clinical management of breast cancer. TNBCs lack the expression of
all three targets, and no targetable molecular pathways have been identified to date.
Hence, TNBCs are treated primarily with multiple non-targeted, cytotoxic
chemotherapeutic agents (e.g. Paclitaxel), and are characterized by high rates of drug
resistance and metastatic relapse.
This study describes a novel molecular driver of TNBC that can be targeted via a specific
small molecule inhibitor, ICG-001. In vitro and in vivo studies demonstrated the potential
benefit of targeting FOXM1/CBP to address important issue in TNBC clinical
management, such as drug resistance and disease recurrence. The results showed that
FOXM1/CBP is related to CSC biology, and targeting of CSC has been postulated as a
critical target for anti-cancer therapy [115, 418]. A search on
https://www.clinicaltrials.gov for triple negative breast cancer yielded 373 phase 1-3
studies as of November 2016, highlighting the need and immense effort undertaken to
improve therapeutic outcome and prognosis for patients with TNBC. Interestingly, no
trial specifically explores the potential of targeting cancer stem cells. The current study
could provide a rationale for implementing CSC targeted therapy into TNBC trials.
This study demonstrated the validity of using an innovative approach of chemical-
genomics to identify a novel molecular target in TNBC. The results demonstrated that
CBP/FOXM1 targeted therapy can ameliorate the most pressing issues to date in TNBC
management, namely drug resistance and disease recurrence.
115
CBP and FOXM1 could potentially be of value as predictive biomarkers in TNBC and
could provide a clinical-translational rationale for patient stratification in future clinical
trials exploring the therapeutic potential of ICG-001 in combination with chemotherapy.
ICG-001, or rather the clinically used equivalent PRI-724, has been shown to not be toxic
in several animal studies in mice and dogs. The drug has been or is currently being
investigated in several clinical trials listed on https://www.clinicaltrials.gov, including a
phase I trial conducted at USC [419]. No negative side effects were observed in the
patients receiving clinically relevant doses of PRI-724.
The preclinical results together with the low toxicity profile of ICG-001 in patients could
offer an exciting new strategy in TNBC therapy.

116
5. REFERENCE
1. Torre, L.A., et al., Global cancer statistics, 2012. CA Cancer J Clin, 2015. 65(2):
p. 87-108.
2. Howlader N, N.A., Krapcho M, Miller D, Bishop K, Altekruse SF, Kosary CL,
Yu M, Ruhl J, Tatalovich Z, Mariotto A, Lewis DR, Chen HS, Feuer EJ, Cronin
KA (eds), SEER Cancer Statistics Review, 1975-2013, National Cancer Institute.
Bethesda, MD, http://seer.cancer.gov/csr/1975_2013/, based on November 2015
SEER data submission, posted to the SEER web site, April 2016. 2016.
3. https://www.cancer.org/cancer/breast-cancer/about/how-common-is-breast-
cancer.html.
4. Bloom, H.J. and W.W. Richardson, Histological grading and prognosis in breast
cancer; a study of 1409 cases of which 359 have been followed for 15 years. Br J
Cancer, 1957. 11(3): p. 359-77.
5. Rakha, E.A., et al., Breast cancer prognostic classification in the molecular era:
the role of histological grade. Breast Cancer Res, 2010. 12(4): p. 207.
6. Elston, C.W. and I.O. Ellis, Pathological prognostic factors in breast cancer. I.
The value of histological grade in breast cancer: experience from a large study
with long-term follow-up. Histopathology, 1991. 19(5): p. 403-10.
7. Elston, C.W., I.O. Ellis, and S.E. Pinder, Prognostic factors in invasive
carcinoma of the breast. Clin Oncol (R Coll Radiol), 1998. 10(1): p. 14-7.
8. Sobin, L.H. and I.D. Fleming, TNM Classification of Malignant Tumors, fifth
edition (1997). Union Internationale Contre le Cancer and the American Joint
Committee on Cancer. Cancer, 1997. 80(9): p. 1803-4.
9. Weigelt, B., F.C. Geyer, and J.S. Reis-Filho, Histological types of breast cancer:
how special are they? Mol Oncol, 2010. 4(3): p. 192-208.
10. Henson, D.E., et al., Relationship among outcome, stage of disease, and
histologic grade for 22,616 cases of breast cancer. The basis for a prognostic
index. Cancer, 1991. 68(10): p. 2142-9.
11. Mirza, A.N., et al., Prognostic factors in node-negative breast cancer: a review of
studies with sample size more than 200 and follow-up more than 5 years. Ann
Surg, 2002. 235(1): p. 10-26.
12. Cianfrocca, M. and L.J. Goldstein, Prognostic and predictive factors in early-
stage breast cancer. Oncologist, 2004. 9(6): p. 606-16.
13. Taneja, P., et al., Classical and Novel Prognostic Markers for Breast Cancer and
their Clinical Significance. Clin Med Insights Oncol, 2010. 4: p. 15-34.
14. Weigel, M.T. and M. Dowsett, Current and emerging biomarkers in breast
cancer: prognosis and prediction. Endocr Relat Cancer, 2010. 17(4): p. R245-62.
15. Ballman, K.V., Biomarker: Predictive or Prognostic? J Clin Oncol, 2015. 33(33):
p. 3968-71.
16. Anderson, W.F., et al., Estrogen receptor breast cancer phenotypes in the
Surveillance, Epidemiology, and End Results database. Breast Cancer Res Treat,
2002. 76(1): p. 27-36.
17. Dai, X., et al., Cancer Hallmarks, Biomarkers and Breast Cancer Molecular
Subtypes. J Cancer, 2016. 7(10): p. 1281-94.
18. Slamon, D.J., et al., Human breast cancer: correlation of relapse and survival
with amplification of the HER-2/neu oncogene. Science, 1987. 235(4785): p. 177-
82.
117
19. Harris, L.N., et al., Use of Biomarkers to Guide Decisions on Adjuvant Systemic
Therapy for Women With Early-Stage Invasive Breast Cancer: American Society
of Clinical Oncology Clinical Practice Guideline. J Clin Oncol, 2016. 34(10): p.
1134-50.
20. Blanks, R.G., et al., Effect of NHS breast screening programme on mortality from
breast cancer in England and Wales, 1990-8: comparison of observed with
predicted mortality. BMJ, 2000. 321(7262): p. 665-9.
21. DeSantis, C., et al., Breast cancer statistics, 2013. CA Cancer J Clin, 2014. 64(1):
p. 52-62.
22. Early Breast Cancer Trialists' Collaborative, G., Effects of chemotherapy and
hormonal therapy for early breast cancer on recurrence and 15-year survival: an
overview of the randomised trials. Lancet, 2005. 365(9472): p. 1687-717.
23. Swedish Organised Service Screening Evaluation, G., Reduction in breast cancer
mortality from the organised service screening with mammography: 2. Validation
with alternative analytic methods. Cancer Epidemiol Biomarkers Prev, 2006.
15(1): p. 52-6.
24. McCullough, A.E., et al., Central pathology laboratory review of HER2 and ER
in early breast cancer: an ALTTO trial [BIG 2-06/NCCTG N063D (Alliance)]
ring study. Breast Cancer Res Treat, 2014. 143(3): p. 485-92.
25. Viale, G., et al., High concordance of protein (by IHC), gene (by FISH; HER2
only), and microarray readout (by TargetPrint) of ER, PgR, and HER2: results
from the EORTC 10041/BIG 03-04 MINDACT trial. Ann Oncol, 2014. 25(4): p.
816-23.
26. Hammond, M.E., et al., American Society of Clinical Oncology/College Of
American Pathologists guideline recommendations for immunohistochemical
testing of estrogen and progesterone receptors in breast cancer. J Clin Oncol,
2010. 28(16): p. 2784-95.
27. McShane, L.M., et al., REporting recommendations for tumour MARKer
prognostic studies (REMARK). Br J Cancer, 2005. 93(4): p. 387-91.
28. Wolff, A.C., et al., Recommendations for human epidermal growth factor
receptor 2 testing in breast cancer: American Society of Clinical
Oncology/College of American Pathologists clinical practice guideline update. J
Clin Oncol, 2013. 31(31): p. 3997-4013.
29. Yerushalmi, R., et al., Ki67 in breast cancer: prognostic and predictive potential.
Lancet Oncol, 2010. 11(2): p. 174-83.
30. Bianchini, G., et al., Triple-negative breast cancer: challenges and opportunities
of a heterogeneous disease. Nat Rev Clin Oncol, 2016. 13(11): p. 674-690.
31. Kuukasjarvi, T., et al., Loss of estrogen receptor in recurrent breast cancer is
associated with poor response to endocrine therapy. J Clin Oncol, 1996. 14(9): p.
2584-9.
32. Broom, R.J., et al., Changes in estrogen receptor, progesterone receptor and Her-
2/neu status with time: discordance rates between primary and metastatic breast
cancer. Anticancer Res, 2009. 29(5): p. 1557-62.
33. Liedtke, C., et al., Prognostic impact of discordance between triple-receptor
measurements in primary and recurrent breast cancer. Ann Oncol, 2009. 20(12):
p. 1953-8.
34. Lower, E.E., et al., HER-2/neu expression in primary and metastatic breast
cancer. Breast Cancer Res Treat, 2009. 113(2): p. 301-6.
118
35. Gonzalez-Angulo, A.M., F. Morales-Vasquez, and G.N. Hortobagyi, Overview of
resistance to systemic therapy in patients with breast cancer. Adv Exp Med Biol,
2007. 608: p. 1-22.
36. Phillips, C., R. Jeffree, and M. Khasraw, Management of breast cancer brain
metastases: A practical review. Breast, 2017. 31: p. 90-98.
37. Zahreddine, H. and K.L. Borden, Mechanisms and insights into drug resistance in
cancer. Front Pharmacol, 2013. 4: p. 28.
38. Greenberg, D., et al., When is cancer care cost-effective? A systematic overview
of cost-utility analyses in oncology. J Natl Cancer Inst, 2010. 102(2): p. 82-8.
39. Beerman, H., et al., Flow cytometric analysis of DNA stemline heterogeneity in
primary and metastatic breast cancer. Cytometry, 1991. 12(2): p. 147-54.
40. Cancer Genome Atlas, N., Comprehensive molecular portraits of human breast
tumours. Nature, 2012. 490(7418): p. 61-70.
41. Curtis, C., et al., The genomic and transcriptomic architecture of 2,000 breast
tumours reveals novel subgroups. Nature, 2012. 486(7403): p. 346-52.
42. Ignatiadis, M. and C. Sotiriou, Understanding the molecular basis of histologic
grade. Pathobiology, 2008. 75(2): p. 104-11.
43. Komaki, K., N. Sano, and A. Tangoku, Problems in histological grading of
malignancy and its clinical significance in patients with operable breast cancer.
Breast Cancer, 2006. 13(3): p. 249-53.
44. Polyak, K., Heterogeneity in breast cancer. J Clin Invest, 2011. 121(10): p. 3786-
8.
45. Russnes, H.G., et al., Insight into the heterogeneity of breast cancer through next-
generation sequencing. J Clin Invest, 2011. 121(10): p. 3810-8.
46. Sorlie, T., et al., Gene expression patterns of breast carcinomas distinguish tumor
subclasses with clinical implications. Proc Natl Acad Sci U S A, 2001. 98(19): p.
10869-74.
47. Wild, P., et al., Laser microdissection and microsatellite analyses of breast
cancer reveal a high degree of tumor heterogeneity. Pathobiology, 2000. 68(4-5):
p. 180-90.
48. Perou, C.M., et al., Molecular portraits of human breast tumours. Nature, 2000.
406(6797): p. 747-52.
49. Hoadley, K.A., et al., Multiplatform analysis of 12 cancer types reveals molecular
classification within and across tissues of origin. Cell, 2014. 158(4): p. 929-44.
50. Parker, J.S., et al., Supervised risk predictor of breast cancer based on intrinsic
subtypes. J Clin Oncol, 2009. 27(8): p. 1160-7.
51. van 't Veer, L.J., et al., Gene expression profiling predicts clinical outcome of
breast cancer. Nature, 2002. 415(6871): p. 530-6.
52. van de Vijver, M.J., et al., A gene-expression signature as a predictor of survival
in breast cancer. N Engl J Med, 2002. 347(25): p. 1999-2009.
53. Wallden, B., et al., Development and verification of the PAM50-based Prosigna
breast cancer gene signature assay. BMC Med Genomics, 2015. 8: p. 54.
54. Cardoso, F., et al., 70-Gene Signature as an Aid to Treatment Decisions in Early-
Stage Breast Cancer. N Engl J Med, 2016. 375(8): p. 717-29.
55. Harris, L., et al., American Society of Clinical Oncology 2007 update of
recommendations for the use of tumor markers in breast cancer. J Clin Oncol,
2007. 25(33): p. 5287-312.
119
56. Raman, G., E.E. Avendano, and M. Chen, in Update on Emerging Genetic Tests
Currently Available for Clinical Use in Common Cancers. 2013: Rockville (MD).
57. Prat, A., et al., Concordance among gene expression-based predictors for ER-
positive breast cancer treated with adjuvant tamoxifen. Ann Oncol, 2012. 23(11):
p. 2866-73.
58. Lehmann, B.D., et al., Identification of human triple-negative breast cancer
subtypes and preclinical models for selection of targeted therapies. J Clin Invest,
2011. 121(7): p. 2750-67.
59. Lehmann, B.D., et al., Refinement of Triple-Negative Breast Cancer Molecular
Subtypes: Implications for Neoadjuvant Chemotherapy Selection. PLoS One,
2016. 11(6): p. e0157368.
60. Blows, F.M., et al., Subtyping of breast cancer by immunohistochemistry to
investigate a relationship between subtype and short and long term survival: a
collaborative analysis of data for 10,159 cases from 12 studies. PLoS Med, 2010.
7(5): p. e1000279.
61. Boyle, P., Triple-negative breast cancer: epidemiological considerations and
recommendations. Ann Oncol, 2012. 23 Suppl 6: p. vi7-12.
62. Carey, L.A., et al., Race, breast cancer subtypes, and survival in the Carolina
Breast Cancer Study. JAMA, 2006. 295(21): p. 2492-502.
63. Fan, C., et al., Concordance among gene-expression-based predictors for breast
cancer. N Engl J Med, 2006. 355(6): p. 560-9.
64. Kohler, B.A., et al., Annual Report to the Nation on the Status of Cancer, 1975-
2011, Featuring Incidence of Breast Cancer Subtypes by Race/Ethnicity, Poverty,
and State. J Natl Cancer Inst, 2015. 107(6): p. djv048.
65. Voduc, K.D., et al., Breast cancer subtypes and the risk of local and regional
relapse. J Clin Oncol, 2010. 28(10): p. 1684-91.
66. Yang, X.R., et al., Associations of breast cancer risk factors with tumor subtypes:
a pooled analysis from the Breast Cancer Association Consortium studies. J Natl
Cancer Inst, 2011. 103(3): p. 250-63.
67. Amirikia, K.C., et al., Higher population-based incidence rates of triple-negative
breast cancer among young African-American women : Implications for breast
cancer screening recommendations. Cancer, 2011. 117(12): p. 2747-53.
68. Howlader, N., et al., US incidence of breast cancer subtypes defined by joint
hormone receptor and HER2 status. J Natl Cancer Inst, 2014. 106(5).
69. Sineshaw, H.M., et al., Association of race/ethnicity, socioeconomic status, and
breast cancer subtypes in the National Cancer Data Base (2010-2011). Breast
Cancer Res Treat, 2014. 145(3): p. 753-63.
70. Stead, L.A., et al., Triple-negative breast cancers are increased in black women
regardless of age or body mass index. Breast Cancer Res, 2009. 11(2): p. R18.
71. Weigelt, B. and J.S. Reis-Filho, Histological and molecular types of breast
cancer: is there a unifying taxonomy? Nat Rev Clin Oncol, 2009. 6(12): p. 718-
30.
72. Rakha, E.A., I.O. Ellis, and J.S. Reis-Filho, Immunohistochemical heterogeneity
of breast carcinomas negative for estrogen receptors, progesterone receptors and
Her2/neu (basal-like breast carcinomas). Mod Pathol, 2008. 21(8): p. 1060-1;
author reply 1061-2.
120
73. Millar, E.K., et al., Prediction of local recurrence, distant metastases, and death
after breast-conserving therapy in early-stage invasive breast cancer using a five-
biomarker panel. J Clin Oncol, 2009. 27(28): p. 4701-8.
74. Lund, M.J., et al., Race and triple negative threats to breast cancer survival: a
population-based study in Atlanta, GA. Breast Cancer Res Treat, 2009. 113(2): p.
357-70.
75. Rakha, E.A., et al., Basal phenotype identifies a poor prognostic subgroup of
breast cancer of clinical importance. Eur J Cancer, 2006. 42(18): p. 3149-56.
76. Cortazar, P., et al., Pathological complete response and long-term clinical benefit
in breast cancer: the CTNeoBC pooled analysis. Lancet, 2014. 384(9938): p. 164-
72.
77. Early Breast Cancer Trialists' Collaborative, G., et al., Comparisons between
different polychemotherapy regimens for early breast cancer: meta-analyses of
long-term outcome among 100,000 women in 123 randomised trials. Lancet,
2012. 379(9814): p. 432-44.
78. Henderson, I.C., et al., Improved outcomes from adding sequential Paclitaxel but
not from escalating Doxorubicin dose in an adjuvant chemotherapy regimen for
patients with node-positive primary breast cancer. J Clin Oncol, 2003. 21(6): p.
976-83.
79. Carey, L.A., et al., The triple negative paradox: primary tumor chemosensitivity
of breast cancer subtypes. Clin Cancer Res, 2007. 13(8): p. 2329-34.
80. Citron, M.L., et al., Randomized trial of dose-dense versus conventionally
scheduled and sequential versus concurrent combination chemotherapy as
postoperative adjuvant treatment of node-positive primary breast cancer: first
report of Intergroup Trial C9741/Cancer and Leukemia Group B Trial 9741. J
Clin Oncol, 2003. 21(8): p. 1431-9.
81. Gluz, O., et al., Triple-negative high-risk breast cancer derives particular benefit
from dose intensification of adjuvant chemotherapy: results of WSG AM-01 trial.
Ann Oncol, 2008. 19(5): p. 861-70.
82. Liedtke, C., et al., Response to neoadjuvant therapy and long-term survival in
patients with triple-negative breast cancer. J Clin Oncol, 2008. 26(8): p. 1275-81.
83. Dent, R., et al., Triple-negative breast cancer: clinical features and patterns of
recurrence. Clin Cancer Res, 2007. 13(15 Pt 1): p. 4429-34.
84. Conlin, A.K. and A.D. Seidman, Taxanes in breast cancer: an update. Curr Oncol
Rep, 2007. 9(1): p. 22-30.
85. Bonotto, M., et al., Measures of outcome in metastatic breast cancer: insights
from a real-world scenario. Oncologist, 2014. 19(6): p. 608-15.
86. Foulkes, W.D., I.E. Smith, and J.S. Reis-Filho, Triple-negative breast cancer. N
Engl J Med, 2010. 363(20): p. 1938-48.
87. Prat, A., et al., Molecular characterization of basal-like and non-basal-like triple-
negative breast cancer. Oncologist, 2013. 18(2): p. 123-33.
88. Cheang, M.C., et al., Defining breast cancer intrinsic subtypes by quantitative
receptor expression. Oncologist, 2015. 20(5): p. 474-82.
89. Lehmann, B.D. and J.A. Pietenpol, Identification and use of biomarkers in
treatment strategies for triple-negative breast cancer subtypes. J Pathol, 2014.
232(2): p. 142-50.
121
90. Masuda, H., et al., Differential response to neoadjuvant chemotherapy among 7
triple-negative breast cancer molecular subtypes. Clin Cancer Res, 2013. 19(19):
p. 5533-40.
91. Cochrane, D.R., et al., Role of the androgen receptor in breast cancer and
preclinical analysis of enzalutamide. Breast Cancer Res, 2014. 16(1): p. R7.
92. Barton, V.N., et al., Anti-androgen therapy in triple-negative breast cancer. Ther
Adv Med Oncol, 2016. 8(4): p. 305-8.
93. Shah, S.P., et al., The clonal and mutational evolution spectrum of primary triple-
negative breast cancers. Nature, 2012. 486(7403): p. 395-9.
94. Lehmann, B.D., et al., PIK3CA mutations in androgen receptor-positive triple
negative breast cancer confer sensitivity to the combination of PI3K and
androgen receptor inhibitors. Breast Cancer Res, 2014. 16(4): p. 406.
95. Gonzalez-Angulo, A.M., et al., Androgen receptor levels and association with
PIK3CA mutations and prognosis in breast cancer. Clin Cancer Res, 2009. 15(7):
p. 2472-8.
96. Foulkes, W.D., et al., Germline BRCA1 mutations and a basal epithelial
phenotype in breast cancer. J Natl Cancer Inst, 2003. 95(19): p. 1482-5.
97. Gonzalez-Angulo, A.M., et al., Incidence and outcome of BRCA mutations in
unselected patients with triple receptor-negative breast cancer. Clin Cancer Res,
2011. 17(5): p. 1082-9.
98. Yun, M.H. and K. Hiom, CtIP-BRCA1 modulates the choice of DNA double-
strand-break repair pathway throughout the cell cycle. Nature, 2009. 459(7245):
p. 460-3.
99. Metzger-Filho, O., et al., Dissecting the heterogeneity of triple-negative breast
cancer. J Clin Oncol, 2012. 30(15): p. 1879-87.
100. Farmer, H., et al., Targeting the DNA repair defect in BRCA mutant cells as a
therapeutic strategy. Nature, 2005. 434(7035): p. 917-21.
101. Turner, N.C., et al., A synthetic lethal siRNA screen identifying genes mediating
sensitivity to a PARP inhibitor. EMBO J, 2008. 27(9): p. 1368-77.
102. Gelmon, K.A., et al., Olaparib in patients with recurrent high-grade serous or
poorly differentiated ovarian carcinoma or triple-negative breast cancer: a phase
2, multicentre, open-label, non-randomised study. Lancet Oncol, 2011. 12(9): p.
852-61.
103. Tutt, A., et al., Oral poly(ADP-ribose) polymerase inhibitor olaparib in patients
with BRCA1 or BRCA2 mutations and advanced breast cancer: a proof-of-
concept trial. Lancet, 2010. 376(9737): p. 235-44.
104. Le, D.T., et al., PD-1 Blockade in Tumors with Mismatch-Repair Deficiency. N
Engl J Med, 2015. 372(26): p. 2509-20.
105. Ali, H.R., et al., PD-L1 protein expression in breast cancer is rare, enriched in
basal-like tumours and associated with infiltrating lymphocytes. Ann Oncol,
2015. 26(7): p. 1488-93.
106. Mittendorf, E.A., et al., PD-L1 expression in triple-negative breast cancer.
Cancer Immunol Res, 2014. 2(4): p. 361-70.
107. Emens LA, B.F., Cassier P, et al, Inhibition of PD-L1 by MPDL3280A leads to
clinical activity in patients with metastatic triple-negative breast cancer. 2015
AACR Annual Meeting. Abstract 2859, 2015.
122
108. Nanda, R., et al., Pembrolizumab in Patients With Advanced Triple-Negative
Breast Cancer: Phase Ib KEYNOTE-012 Study. J Clin Oncol, 2016. 34(21): p.
2460-7.
109. Cree, I.A., et al., PD-L1 testing for lung cancer in the UK: recognizing the
challenges for implementation. Histopathology, 2016. 69(2): p. 177-86.
110. Clevers, H., The cancer stem cell: premises, promises and challenges. Nat Med,
2011. 17(3): p. 313-9.
111. Jordan, C.T., M.L. Guzman, and M. Noble, Cancer stem cells. N Engl J Med,
2006. 355(12): p. 1253-61.
112. Kreso, A. and J.E. Dick, Evolution of the cancer stem cell model. Cell Stem Cell,
2014. 14(3): p. 275-91.
113. Reya, T., et al., Stem cells, cancer, and cancer stem cells. Nature, 2001.
414(6859): p. 105-11.
114. Adorno-Cruz, V., et al., Cancer stem cells: targeting the roots of cancer, seeds of
metastasis, and sources of therapy resistance. Cancer Res, 2015. 75(6): p. 924-9.
115. Chen, K., Y.H. Huang, and J.L. Chen, Understanding and targeting cancer stem
cells: therapeutic implications and challenges. Acta Pharmacol Sin, 2013. 34(6):
p. 732-40.
116. Pardal, R., M.F. Clarke, and S.J. Morrison, Applying the principles of stem-cell
biology to cancer. Nat Rev Cancer, 2003. 3(12): p. 895-902.
117. Armanios, M. and C.W. Greider, Telomerase and cancer stem cells. Cold Spring
Harb Symp Quant Biol, 2005. 70: p. 205-8.
118. Shay, J.W. and W.E. Wright, Telomeres and telomerase in normal and cancer
stem cells. FEBS Lett, 2010. 584(17): p. 3819-25.
119. Al-Hajj, M. and M.F. Clarke, Self-renewal and solid tumor stem cells. Oncogene,
2004. 23(43): p. 7274-82.
120. Bjerkvig, R., et al., Opinion: the origin of the cancer stem cell: current
controversies and new insights. Nat Rev Cancer, 2005. 5(11): p. 899-904.
121. Foulkes, W.D., BRCA1 functions as a breast stem cell regulator. J Med Genet,
2004. 41(1): p. 1-5.
122. Liu, S., et al., BRCA1 regulates human mammary stem/progenitor cell fate. Proc
Natl Acad Sci U S A, 2008. 105(5): p. 1680-5.
123. Molyneux, G., et al., BRCA1 basal-like breast cancers originate from luminal
epithelial progenitors and not from basal stem cells. Cell Stem Cell, 2010. 7(3):
p. 403-17.
124. Friedmann-Morvinski, D. and I.M. Verma, Dedifferentiation and reprogramming:
origins of cancer stem cells. EMBO Rep, 2014. 15(3): p. 244-53.
125. Mani, S.A., et al., The epithelial-mesenchymal transition generates cells with
properties of stem cells. Cell, 2008. 133(4): p. 704-15.
126. Morel, A.P., et al., Generation of breast cancer stem cells through epithelial-
mesenchymal transition. PLoS One, 2008. 3(8): p. e2888.
127. Moreno-Bueno, G., F. Portillo, and A. Cano, Transcriptional regulation of cell
polarity in EMT and cancer. Oncogene, 2008. 27(55): p. 6958-69.
128. Zhou, B.B., et al., Tumour-initiating cells: challenges and opportunities for
anticancer drug discovery. Nat Rev Drug Discov, 2009. 8(10): p. 806-23.
129. Furth, J., Kahn, M. C. and Breedis, C., The Transmission of Leukemia of Mice
with a Single Cell. Cancer Res, 1937. 31(2): p. 276-82.
123
130. Bonnet, D. and J.E. Dick, Human acute myeloid leukemia is organized as a
hierarchy that originates from a primitive hematopoietic cell. Nat Med, 1997.
3(7): p. 730-7.
131. Lapidot, T., et al., A cell initiating human acute myeloid leukaemia after
transplantation into SCID mice. Nature, 1994. 367(6464): p. 645-8.
132. Li, C., et al., Identification of pancreatic cancer stem cells. Cancer Res, 2007.
67(3): p. 1030-7.
133. O'Brien, C.A., et al., A human colon cancer cell capable of initiating tumour
growth in immunodeficient mice. Nature, 2007. 445(7123): p. 106-10.
134. Schatton, T., et al., Identification of cells initiating human melanomas. Nature,
2008. 451(7176): p. 345-9.
135. Singh, S.K., et al., Identification of human brain tumour initiating cells. Nature,
2004. 432(7015): p. 396-401.
136. Klonisch, T., et al., Cancer stem cell markers in common cancers - therapeutic
implications. Trends Mol Med, 2008. 14(10): p. 450-60.
137. Al-Hajj, M., et al., Prospective identification of tumorigenic breast cancer cells.
Proc Natl Acad Sci U S A, 2003. 100(7): p. 3983-8.
138. Ke, J., et al., A subpopulation of CD24(+) cells in colon cancer cell lines possess
stem cell characteristics. Neoplasma, 2012. 59(3): p. 282-8.
139. Collins, A.T., et al., Prospective identification of tumorigenic prostate cancer
stem cells. Cancer Res, 2005. 65(23): p. 10946-51.
140. Blair, A., et al., Lack of expression of Thy-1 (CD90) on acute myeloid leukemia
cells with long-term proliferative ability in vitro and in vivo. Blood, 1997. 89(9):
p. 3104-12.
141. Jordan, C.T., et al., The interleukin-3 receptor alpha chain is a unique marker for
human acute myelogenous leukemia stem cells. Leukemia, 2000. 14(10): p. 1777-
84.
142. Cheung, A.M., et al., Aldehyde dehydrogenase activity in leukemic blasts defines
a subgroup of acute myeloid leukemia with adverse prognosis and superior
NOD/SCID engrafting potential. Leukemia, 2007. 21(7): p. 1423-30.
143. Hirschmann-Jax, C., et al., A distinct "side population" of cells with high drug
efflux capacity in human tumor cells. Proc Natl Acad Sci U S A, 2004. 101(39): p.
14228-33.
144. Moshaver, B., et al., Identification of a small subpopulation of candidate
leukemia-initiating cells in the side population of patients with acute myeloid
leukemia. Stem Cells, 2008. 26(12): p. 3059-67.
145. Abdullah, L.N. and E.K. Chow, Mechanisms of chemoresistance in cancer stem
cells. Clin Transl Med, 2013. 2(1): p. 3.
146. Creighton, C.J., et al., Residual breast cancers after conventional therapy display
mesenchymal as well as tumor-initiating features. Proc Natl Acad Sci U S A,
2009. 106(33): p. 13820-5.
147. Dean, M., T. Fojo, and S. Bates, Tumour stem cells and drug resistance. Nat Rev
Cancer, 2005. 5(4): p. 275-84.
148. Charafe-Jauffret, E., et al., Breast cancer cell lines contain functional cancer stem
cells with metastatic capacity and a distinct molecular signature. Cancer Res,
2009. 69(4): p. 1302-13.
124
149. Hennessy, B.T., et al., Characterization of a naturally occurring breast cancer
subset enriched in epithelial-to-mesenchymal transition and stem cell
characteristics. Cancer Res, 2009. 69(10): p. 4116-24.
150. Herschkowitz, J.I., et al., Identification of conserved gene expression features
between murine mammary carcinoma models and human breast tumors. Genome
Biol, 2007. 8(5): p. R76.
151. Neve, R.M., et al., A collection of breast cancer cell lines for the study of
functionally distinct cancer subtypes. Cancer Cell, 2006. 10(6): p. 515-27.
152. Hussenet, T., et al., An adult tissue-specific stem cell molecular phenotype is
activated in epithelial cancer stem cells and correlated to patient outcome. Cell
Cycle, 2010. 9(2): p. 321-7.
153. DiMeo, T.A., et al., A novel lung metastasis signature links Wnt signaling with
cancer cell self-renewal and epithelial-mesenchymal transition in basal-like
breast cancer. Cancer Res, 2009. 69(13): p. 5364-73.
154. Sarrio, D., et al., Epithelial-mesenchymal transition in breast cancer relates to the
basal-like phenotype. Cancer Res, 2008. 68(4): p. 989-97.
155. Horimoto, Y., et al., Combination of Cancer Stem Cell Markers CD44 and CD24
Is Superior to ALDH1 as a Prognostic Indicator in Breast Cancer Patients with
Distant Metastases. PLoS One, 2016. 11(10): p. e0165253.
156. Li, F., et al., Beyond tumorigenesis: cancer stem cells in metastasis. Cell Res,
2007. 17(1): p. 3-14.
157. Sampieri, K. and R. Fodde, Cancer stem cells and metastasis. Semin Cancer Biol,
2012. 22(3): p. 187-93.
158. Liu, S., et al., Hedgehog signaling and Bmi-1 regulate self-renewal of normal and
malignant human mammary stem cells. Cancer Res, 2006. 66(12): p. 6063-71.
159. Harrison, H., et al., Regulation of breast cancer stem cell activity by signaling
through the Notch4 receptor. Cancer Res, 2010. 70(2): p. 709-18.
160. Bhola, N.E., et al., TGF-beta inhibition enhances chemotherapy action against
triple-negative breast cancer. J Clin Invest, 2013. 123(3): p. 1348-58.
161. Wang, Y., et al., The Wnt/beta-catenin pathway is required for the development of
leukemia stem cells in AML. Science, 2010. 327(5973): p. 1650-3.
162. Clevers, H., Wnt/beta-catenin signaling in development and disease. Cell, 2006.
127(3): p. 469-80.
163. Kimura-Yoshida, C., et al., Canonical Wnt signaling and its antagonist regulate
anterior-posterior axis polarization by guiding cell migration in mouse visceral
endoderm. Dev Cell, 2005. 9(5): p. 639-50.
164. Nusse, R., Wnt signaling and stem cell control. Cell Res, 2008. 18(5): p. 523-7.
165. Barker, N., et al., Identification of stem cells in small intestine and colon by
marker gene Lgr5. Nature, 2007. 449(7165): p. 1003-7.
166. Clevers, H., K.M. Loh, and R. Nusse, Stem cell signaling. An integral program
for tissue renewal and regeneration: Wnt signaling and stem cell control.
Science, 2014. 346(6205): p. 1248012.
167. Clevers, H. and R. Nusse, Wnt/beta-catenin signaling and disease. Cell, 2012.
149(6): p. 1192-205.
168. Korinek, V., et al., Constitutive transcriptional activation by a beta-catenin-Tcf
complex in APC-/- colon carcinoma. Science, 1997. 275(5307): p. 1784-7.
169. MacDonald, B.T., K. Tamai, and X. He, Wnt/beta-catenin signaling: components,
mechanisms, and diseases. Dev Cell, 2009. 17(1): p. 9-26.
125
170. Komiya, Y. and R. Habas, Wnt signal transduction pathways. Organogenesis,
2008. 4(2): p. 68-75.
171. Aberle, H., et al., beta-catenin is a target for the ubiquitin-proteasome pathway.
EMBO J, 1997. 16(13): p. 3797-804.
172. Hart, M.J., et al., Downregulation of beta-catenin by human Axin and its
association with the APC tumor suppressor, beta-catenin and GSK3 beta. Curr
Biol, 1998. 8(10): p. 573-81.
173. Ikeda, S., et al., Axin, a negative regulator of the Wnt signaling pathway, forms a
complex with GSK-3beta and beta-catenin and promotes GSK-3beta-dependent
phosphorylation of beta-catenin. EMBO J, 1998. 17(5): p. 1371-84.
174. Orford, K., et al., Serine phosphorylation-regulated ubiquitination and
degradation of beta-catenin. J Biol Chem, 1997. 272(40): p. 24735-8.
175. Stamos, J.L. and W.I. Weis, The beta-catenin destruction complex. Cold Spring
Harb Perspect Biol, 2013. 5(1): p. a007898.
176. Bhanot, P., et al., A new member of the frizzled family from Drosophila functions
as a Wingless receptor. Nature, 1996. 382(6588): p. 225-30.
177. Yang-Snyder, J., et al., A frizzled homolog functions in a vertebrate Wnt signaling
pathway. Curr Biol, 1996. 6(10): p. 1302-6.
178. He, X., et al., LDL receptor-related proteins 5 and 6 in Wnt/beta-catenin
signaling: arrows point the way. Development, 2004. 131(8): p. 1663-77.
179. MacDonald, B.T. and X. He, Frizzled and LRP5/6 receptors for Wnt/beta-catenin
signaling. Cold Spring Harb Perspect Biol, 2012. 4(12).
180. Pinson, K.I., et al., An LDL-receptor-related protein mediates Wnt signalling in
mice. Nature, 2000. 407(6803): p. 535-8.
181. Tamai, K., et al., LDL-receptor-related proteins in Wnt signal transduction.
Nature, 2000. 407(6803): p. 530-5.
182. Cliffe, A., F. Hamada, and M. Bienz, A role of Dishevelled in relocating Axin to
the plasma membrane during wingless signaling. Curr Biol, 2003. 13(11): p. 960-
6.
183. Behrens, J., et al., Functional interaction of beta-catenin with the transcription
factor LEF-1. Nature, 1996. 382(6592): p. 638-42.
184. Molenaar, M., et al., XTcf-3 transcription factor mediates beta-catenin-induced
axis formation in Xenopus embryos. Cell, 1996. 86(3): p. 391-9.
185. Bannister, A.J. and T. Kouzarides, The CBP co-activator is a histone
acetyltransferase. Nature, 1996. 384(6610): p. 641-3.
186. Ogryzko, V.V., et al., The transcriptional coactivators p300 and CBP are histone
acetyltransferases. Cell, 1996. 87(5): p. 953-9.
187. Boyes, J., et al., Regulation of activity of the transcription factor GATA-1 by
acetylation. Nature, 1998. 396(6711): p. 594-8.
188. Gu, W. and R.G. Roeder, Activation of p53 sequence-specific DNA binding by
acetylation of the p53 C-terminal domain. Cell, 1997. 90(4): p. 595-606.
189. Chan, H.M. and N.B. La Thangue, p300/CBP proteins: HATs for transcriptional
bridges and scaffolds. J Cell Sci, 2001. 114(Pt 13): p. 2363-73.
190. Hottiger, M.O., L.K. Felzien, and G.J. Nabel, Modulation of cytokine-induced
HIV gene expression by competitive binding of transcription factors to the
coactivator p300. EMBO J, 1998. 17(11): p. 3124-34.
126
191. Merika, M., et al., Recruitment of CBP/p300 by the IFN beta enhanceosome is
required for synergistic activation of transcription. Mol Cell, 1998. 1(2): p. 277-
87.
192. Wathelet, M.G., et al., Virus infection induces the assembly of coordinately
activated transcription factors on the IFN-beta enhancer in vivo. Mol Cell, 1998.
1(4): p. 507-18.
193. Yie, J., K. Senger, and D. Thanos, Mechanism by which the IFN-beta
enhanceosome activates transcription. Proc Natl Acad Sci U S A, 1999. 96(23):
p. 13108-13.
194. Arias, J., et al., Activation of cAMP and mitogen responsive genes relies on a
common nuclear factor. Nature, 1994. 370(6486): p. 226-9.
195. Kwok, R.P., et al., Nuclear protein CBP is a coactivator for the transcription
factor CREB. Nature, 1994. 370(6486): p. 223-6.
196. Mayr, B. and M. Montminy, Transcriptional regulation by the phosphorylation-
dependent factor CREB. Nat Rev Mol Cell Biol, 2001. 2(8): p. 599-609.
197. Eckner, R., et al., Interaction and functional collaboration of p300/CBP and
bHLH proteins in muscle and B-cell differentiation. Genes Dev, 1996. 10(19): p.
2478-90.
198. Giordano, A. and M.L. Avantaggiati, p300 and CBP: partners for life and death.
J Cell Physiol, 1999. 181(2): p. 218-30.
199. Goodman, R.H. and S. Smolik, CBP/p300 in cell growth, transformation, and
development. Genes Dev, 2000. 14(13): p. 1553-77.
200. Kawasaki, H., et al., Distinct roles of the co-activators p300 and CBP in retinoic-
acid-induced F9-cell differentiation. Nature, 1998. 393(6682): p. 284-9.
201. Pao, G.M., et al., CBP/p300 interact with and function as transcriptional
coactivators of BRCA1. Proc Natl Acad Sci U S A, 2000. 97(3): p. 1020-5.
202. Kung, A.L., et al., Gene dose-dependent control of hematopoiesis and
hematologic tumor suppression by CBP. Genes Dev, 2000. 14(3): p. 272-7.
203. Yao, T.P., et al., Gene dosage-dependent embryonic development and
proliferation defects in mice lacking the transcriptional integrator p300. Cell,
1998. 93(3): p. 361-72.
204. Partanen, A., J. Motoyama, and C.C. Hui, Developmentally regulated expression
of the transcriptional cofactors/histone acetyltransferases CBP and p300 during
mouse embryogenesis. Int J Dev Biol, 1999. 43(6): p. 487-94.
205. Wang, J., et al., CBP histone acetyltransferase activity regulates embryonic
neural differentiation in the normal and Rubinstein-Taybi syndrome brain. Dev
Cell, 2010. 18(1): p. 114-25.
206. Miyabayashi, T., et al., Wnt/beta-catenin/CBP signaling maintains long-term
murine embryonic stem cell pluripotency. Proc Natl Acad Sci U S A, 2007.
104(13): p. 5668-73.
207. Chan, W.I., et al., The transcriptional coactivator Cbp regulates self-renewal and
differentiation in adult hematopoietic stem cells. Mol Cell Biol, 2011. 31(24): p.
5046-60.
208. Rebel, V.I., et al., Distinct roles for CREB-binding protein and p300 in
hematopoietic stem cell self-renewal. Proc Natl Acad Sci U S A, 2002. 99(23): p.
14789-94.
209. Petrij, F., et al., Rubinstein-Taybi syndrome caused by mutations in the
transcriptional co-activator CBP. Nature, 1995. 376(6538): p. 348-51.
127
210. Borrow, J., et al., The translocation t(8;16)(p11;p13) of acute myeloid leukaemia
fuses a putative acetyltransferase to the CREB-binding protein. Nat Genet, 1996.
14(1): p. 33-41.
211. Garraway, L.A. and E.S. Lander, Lessons from the cancer genome. Cell, 2013.
153(1): p. 17-37.
212. Gayther, S.A., et al., Mutations truncating the EP300 acetylase in human cancers.
Nat Genet, 2000. 24(3): p. 300-3.
213. Ida, K., et al., Adenoviral E1A-associated protein p300 is involved in acute
myeloid leukemia with t(11;22)(q23;q13). Blood, 1997. 90(12): p. 4699-704.
214. Muraoka, M., et al., p300 gene alterations in colorectal and gastric carcinomas.
Oncogene, 1996. 12(7): p. 1565-9.
215. Gang, E.J., et al., Small-molecule inhibition of CBP/catenin interactions
eliminates drug-resistant clones in acute lymphoblastic leukemia. Oncogene,
2014. 33(17): p. 2169-78.
216. Zhao, Y., et al., CBP/catenin antagonist safely eliminates drug-resistant
leukemia-initiating cells. Oncogene, 2016. 35(28): p. 3705-17.
217. Wielenga, V.J., et al., Expression of CD44 in Apc and Tcf mutant mice implies
regulation by the WNT pathway. Am J Pathol, 1999. 154(2): p. 515-23.
218. Shulewitz, M., et al., Repressor roles for TCF-4 and Sfrp1 in Wnt signaling in
breast cancer. Oncogene, 2006. 25(31): p. 4361-9.
219. Katoh, Y. and M. Katoh, Comparative genomics on PROM1 gene encoding stem
cell marker CD133. Int J Mol Med, 2007. 19(6): p. 967-70.
220. Correa, S., et al., Wnt/beta-catenin pathway regulates ABCB1 transcription in
chronic myeloid leukemia. BMC Cancer, 2012. 12: p. 303.
221. Yamada, T., et al., Transactivation of the multidrug resistance 1 gene by T-cell
factor 4/beta-catenin complex in early colorectal carcinogenesis. Cancer Res,
2000. 60(17): p. 4761-6.
222. Yamashita, T., et al., Activation of hepatic stem cell marker EpCAM by Wnt-beta-
catenin signaling in hepatocellular carcinoma. Cancer Res, 2007. 67(22): p.
10831-9.
223. Wu, Z.Q., et al., Canonical Wnt signaling regulates Slug activity and links
epithelial-mesenchymal transition with epigenetic Breast Cancer 1, Early Onset
(BRCA1) repression. Proc Natl Acad Sci U S A, 2012. 109(41): p. 16654-9.
224. Lickert, H., et al., Wnt/(beta)-catenin signaling regulates the expression of the
homeobox gene Cdx1 in embryonic intestine. Development, 2000. 127(17): p.
3805-13.
225. Russell, R.G., et al., Id2 drives differentiation and suppresses tumor formation in
the intestinal epithelium. Cancer Res, 2004. 64(20): p. 7220-5.
226. Hoffmeyer, K., et al., Wnt/beta-catenin signaling regulates telomerase in stem
cells and cancer cells. Science, 2012. 336(6088): p. 1549-54.
227. Bhat-Nakshatri, P., et al., SLUG/SNAI2 and tumor necrosis factor generate breast
cells with CD44+/CD24- phenotype. BMC Cancer, 2010. 10: p. 411.
228. Jang, G.B., et al., Blockade of Wnt/beta-catenin signaling suppresses breast
cancer metastasis by inhibiting CSC-like phenotype. Sci Rep, 2015. 5: p. 12465.
229. Polakis, P., Wnt signaling in cancer. Cold Spring Harb Perspect Biol, 2012. 4(5).
230. Zhang, T., et al., Evidence that APC regulates survivin expression: a possible
mechanism contributing to the stem cell origin of colon cancer. Cancer Res, 2001.
61(24): p. 8664-7.
128
231. Kwan, H., et al., Transgenes expressing the Wnt-1 and int-2 proto-oncogenes
cooperate during mammary carcinogenesis in doubly transgenic mice. Mol Cell
Biol, 1992. 12(1): p. 147-54.
232. Nusse, R., et al., The Wnt-1 (int-1) oncogene promoter and its mechanism of
activation by insertion of proviral DNA of the mouse mammary tumor virus. Mol
Cell Biol, 1990. 10(8): p. 4170-9.
233. Lim, E., et al., Transcriptome analyses of mouse and human mammary cell
subpopulations reveal multiple conserved genes and pathways. Breast Cancer
Res, 2010. 12(2): p. R21.
234. Moser, A.R., et al., ApcMin, a mutation in the murine Apc gene, predisposes to
mammary carcinomas and focal alveolar hyperplasias. Proc Natl Acad Sci U S
A, 1993. 90(19): p. 8977-81.
235. Muller, H.M., et al., DNA methylation in serum of breast cancer patients: an
independent prognostic marker. Cancer Res, 2003. 63(22): p. 7641-5.
236. Virmani, A.K., et al., Aberrant methylation of the adenomatous polyposis coli
(APC) gene promoter 1A in breast and lung carcinomas. Clin Cancer Res, 2001.
7(7): p. 1998-2004.
237. Zhou, D., et al., Association between aberrant APC promoter methylation and
breast cancer pathogenesis: a meta-analysis of 35 observational studies. PeerJ,
2016. 4: p. e2203.
238. Ugolini, F., et al., WNT pathway and mammary carcinogenesis: loss of expression
of candidate tumor suppressor gene SFRP1 in most invasive carcinomas except of
the medullary type. Oncogene, 2001. 20(41): p. 5810-7.
239. Nagahata, T., et al., Amplification, up-regulation and over-expression of DVL-1,
the human counterpart of the Drosophila disheveled gene, in primary breast
cancers. Cancer Sci, 2003. 94(6): p. 515-8.
240. Kouzmenko, A.P., et al., Wnt/beta-catenin and estrogen signaling converge in
vivo. J Biol Chem, 2004. 279(39): p. 40255-8.
241. Yamaguchi, H., et al., Signaling cross-talk in the resistance to HER family
receptor targeted therapy. Oncogene, 2014. 33(9): p. 1073-81.
242. Frasor, J., et al., Profiling of estrogen up- and down-regulated gene expression in
human breast cancer cells: insights into gene networks and pathways underlying
estrogenic control of proliferation and cell phenotype. Endocrinology, 2003.
144(10): p. 4562-74.
243. Ma, H., et al., Differential roles for the coactivators CBP and p300 on TCF/beta-
catenin-mediated survivin gene expression. Oncogene, 2005. 24(22): p. 3619-31.
244. Blum, B., et al., The anti-apoptotic gene survivin contributes to teratoma
formation by human embryonic stem cells. Nat Biotechnol, 2009. 27(3): p. 281-7.
245. Kennedy, S.M., et al., Prognostic importance of survivin in breast cancer. Br J
Cancer, 2003. 88(7): p. 1077-83.
246. Youssef, N.S., I.H. Hewedi, and N.M. Abd Raboh, Immunohistochemical
expression of survivin in breast carcinoma: relationship with clinicopathological
parameters, proliferation and molecular classification. J Egypt Natl Canc Inst,
2008. 20(4): p. 348-57.
247. Schroeder, J.A., et al., ErbB-beta-catenin complexes are associated with human
infiltrating ductal breast and murine mammary tumor virus (MMTV)-Wnt-1 and
MMTV-c-Neu transgenic carcinomas. J Biol Chem, 2002. 277(25): p. 22692-8.
129
248. Hallett, R.M., et al., Small molecule antagonists of the Wnt/beta-catenin signaling
pathway target breast tumor-initiating cells in a Her2/Neu mouse model of breast
cancer. PLoS One, 2012. 7(3): p. e33976.
249. Howe, L.R. and A.M. Brown, Wnt signaling and breast cancer. Cancer Biol Ther,
2004. 3(1): p. 36-41.
250. Khramtsov, A.I., et al., Wnt/beta-catenin pathway activation is enriched in basal-
like breast cancers and predicts poor outcome. Am J Pathol, 2010. 176(6): p.
2911-20.
251. Mukherjee, N., et al., Subtype-specific alterations of the Wnt signaling pathway in
breast cancer: clinical and prognostic significance. Cancer Sci, 2012. 103(2): p.
210-20.
252. Clark, K.L., et al., Co-crystal structure of the HNF-3/fork head DNA-recognition
motif resembles histone H5. Nature, 1993. 364(6436): p. 412-20.
253. Clevidence, D.E., et al., Identification of nine tissue-specific transcription factors
of the hepatocyte nuclear factor 3/forkhead DNA-binding-domain family. Proc
Natl Acad Sci U S A, 1993. 90(9): p. 3948-52.
254. Korver, W., et al., The human TRIDENT/HFH-11/FKHL16 gene: structure,
localization, and promoter characterization. Genomics, 1997. 46(3): p. 435-42.
255. Jackson, B.C., et al., Update of human and mouse forkhead box (FOX) gene
families. Hum Genomics, 2010. 4(5): p. 345-52.
256. Costa, R.H., et al., Transcription factors in liver development, differentiation, and
regeneration. Hepatology, 2003. 38(6): p. 1331-47.
257. Kim, I.M., et al., The forkhead box m1 transcription factor is essential for
embryonic development of pulmonary vasculature. J Biol Chem, 2005. 280(23): p.
22278-86.
258. Papanicolaou, K.N., Y. Izumiya, and K. Walsh, Forkhead transcription factors
and cardiovascular biology. Circ Res, 2008. 102(1): p. 16-31.
259. Uhlenhaut, N.H. and M. Treier, Forkhead transcription factors in ovarian
function. Reproduction, 2011. 142(4): p. 489-95.
260. Zhang, H., et al., The FoxM1 transcription factor is required to maintain
pancreatic beta-cell mass. Mol Endocrinol, 2006. 20(8): p. 1853-66.
261. Coffer, P.J. and B.M. Burgering, Forkhead-box transcription factors and their
role in the immune system. Nat Rev Immunol, 2004. 4(11): p. 889-99.
262. Bella, L., et al., FOXM1: A key oncofoetal transcription factor in health and
disease. Semin Cancer Biol, 2014. 29: p. 32-9.
263. Myatt, S.S. and E.W. Lam, The emerging roles of forkhead box (Fox) proteins in
cancer. Nat Rev Cancer, 2007. 7(11): p. 847-59.
264. Partridge, L. and J.C. Bruning, Forkhead transcription factors and ageing.
Oncogene, 2008. 27(16): p. 2351-63.
265. Benayoun, B.A., S. Caburet, and R.A. Veitia, Forkhead transcription factors: key
players in health and disease. Trends Genet, 2011. 27(6): p. 224-32.
266. Greer, E.L. and A. Brunet, FOXO transcription factors at the interface between
longevity and tumor suppression. Oncogene, 2005. 24(50): p. 7410-25.
267. Wierstra, I., FOXM1 (Forkhead box M1) in tumorigenesis: overexpression in
human cancer, implication in tumorigenesis, oncogenic functions, tumor-
suppressive properties, and target of anticancer therapy. Adv Cancer Res, 2013.
119: p. 191-419.
130
268. Westendorf, J.M., P.N. Rao, and L. Gerace, Cloning of cDNAs for M-phase
phosphoproteins recognized by the MPM2 monoclonal antibody and
determination of the phosphorylated epitope. Proc Natl Acad Sci U S A, 1994.
91(2): p. 714-8.
269. Major, M.L., R. Lepe, and R.H. Costa, Forkhead box M1B transcriptional activity
requires binding of Cdk-cyclin complexes for phosphorylation-dependent
recruitment of p300/CBP coactivators. Mol Cell Biol, 2004. 24(7): p. 2649-61.
270. Halasi, M. and A.L. Gartel, A novel mode of FoxM1 regulation: positive auto-
regulatory loop. Cell Cycle, 2009. 8(12): p. 1966-7.
271. Korver, W., J. Roose, and H. Clevers, The winged-helix transcription factor
Trident is expressed in cycling cells. Nucleic Acids Res, 1997. 25(9): p. 1715-9.
272. Fu, Z., et al., Plk1-dependent phosphorylation of FoxM1 regulates a
transcriptional programme required for mitotic progression. Nat Cell Biol, 2008.
10(9): p. 1076-82.
273. Laoukili, J., et al., FoxM1 is required for execution of the mitotic programme and
chromosome stability. Nat Cell Biol, 2005. 7(2): p. 126-36.
274. Korver, W., et al., Uncoupling of S phase and mitosis in cardiomyocytes and
hepatocytes lacking the winged-helix transcription factor Trident. Curr Biol,
1998. 8(24): p. 1327-30.
275. Krupczak-Hollis, K., et al., The mouse Forkhead Box m1 transcription factor is
essential for hepatoblast mitosis and development of intrahepatic bile ducts and
vessels during liver morphogenesis. Dev Biol, 2004. 276(1): p. 74-88.
276. Kalin, T.V., V. Ustiyan, and V.V. Kalinichenko, Multiple faces of FoxM1
transcription factor: lessons from transgenic mouse models. Cell Cycle, 2011.
10(3): p. 396-405.
277. Yao, K.M., et al., Molecular analysis of a novel winged helix protein, WIN.
Expression pattern, DNA binding property, and alternative splicing within the
DNA binding domain. J Biol Chem, 1997. 272(32): p. 19827-36.
278. Ye, H., et al., Hepatocyte nuclear factor 3/fork head homolog 11 is expressed in
proliferating epithelial and mesenchymal cells of embryonic and adult tissues.
Mol Cell Biol, 1997. 17(3): p. 1626-41.
279. Furney, S.J., et al., Structural and functional properties of genes involved in
human cancer. BMC Genomics, 2006. 7: p. 3.
280. Darnell, J.E., Jr., Transcription factors as targets for cancer therapy. Nat Rev
Cancer, 2002. 2(10): p. 740-9.
281. Chan, D.W., et al., Over-expression of FOXM1 transcription factor is associated
with cervical cancer progression and pathogenesis. J Pathol, 2008. 215(3): p.
245-52.
282. Kim, I.M., et al., The Forkhead Box m1 transcription factor stimulates the
proliferation of tumor cells during development of lung cancer. Cancer Res, 2006.
66(4): p. 2153-61.
283. Li, Q., et al., Critical role and regulation of transcription factor FoxM1 in human
gastric cancer angiogenesis and progression. Cancer Res, 2009. 69(8): p. 3501-9.
284. Liu, M., et al., FoxM1B is overexpressed in human glioblastomas and critically
regulates the tumorigenicity of glioma cells. Cancer Res, 2006. 66(7): p. 3593-
602.
131
285. Pilarsky, C., et al., Identification and validation of commonly overexpressed genes
in solid tumors by comparison of microarray data. Neoplasia, 2004. 6(6): p. 744-
50.
286. Uddin, S., et al., Genome-wide expression analysis of Middle Eastern colorectal
cancer reveals FOXM1 as a novel target for cancer therapy. Am J Pathol, 2011.
178(2): p. 537-47.
287. Kretschmer, C., et al., Identification of early molecular markers for breast cancer.
Mol Cancer, 2011. 10(1): p. 15.
288. Li, X., et al., FOXM1 mediates resistance to docetaxel in gastric cancer via up-
regulating Stathmin. J Cell Mol Med, 2014. 18(5): p. 811-23.
289. Park, Y.Y., et al., FOXM1 mediates Dox resistance in breast cancer by enhancing
DNA repair. Carcinogenesis, 2012. 33(10): p. 1843-53.
290. Wang, Y., et al., FOXM1 confers resistance to gefitinib in lung adenocarcinoma
via a MET/AKT-dependent positive feedback loop. Oncotarget, 2016. 7(37): p.
59245-59259.
291. Kwok, J.M., et al., FOXM1 confers acquired cisplatin resistance in breast cancer
cells. Mol Cancer Res, 2010. 8(1): p. 24-34.
292. Millour, J., et al., ATM and p53 regulate FOXM1 expression via E2F in breast
cancer epirubicin treatment and resistance. Mol Cancer Ther, 2011. 10(6): p.
1046-58.
293. Carr, J.R., et al., FoxM1 mediates resistance to herceptin and paclitaxel. Cancer
Res, 2010. 70(12): p. 5054-63.
294. Bergamaschi, A., et al., The forkhead transcription factor FOXM1 promotes
endocrine resistance and invasiveness in estrogen receptor-positive breast cancer
by expansion of stem-like cancer cells. Breast Cancer Res, 2014. 16(5): p. 436.
295. Lee, Y., et al., FoxM1 Promotes Stemness and Radio-Resistance of Glioblastoma
by Regulating the Master Stem Cell Regulator Sox2. PLoS One, 2015. 10(10): p.
e0137703.
296. Yang, N., et al., FOXM1 recruits nuclear Aurora kinase A to participate in a
positive feedback loop essential for the self-renewal of breast cancer stem cells.
Oncogene, 2017.
297. Zhang, N., et al., FoxM1 promotes beta-catenin nuclear localization and controls
Wnt target-gene expression and glioma tumorigenesis. Cancer Cell, 2011. 20(4):
p. 427-42.
298. Chu, X.Y., et al., FOXM1 expression correlates with tumor invasion and a poor
prognosis of colorectal cancer. Acta Histochem, 2012. 114(8): p. 755-62.
299. Luo, X., et al., FOXM1 promotes invasion and migration of colorectal cancer
cells partially dependent on HSPA5 transactivation. Oncotarget, 2016. 7(18): p.
26480-95.
300. Park, H.J., et al., Deregulation of FoxM1b leads to tumour metastasis. EMBO
Mol Med, 2011. 3(1): p. 21-34.
301. Raychaudhuri, P. and H.J. Park, FoxM1: a master regulator of tumor metastasis.
Cancer Res, 2011. 71(13): p. 4329-33.
302. Yang, D.K., et al., Forkhead box M1 expression in pulmonary squamous cell
carcinoma: correlation with clinicopathologic features and its prognostic
significance. Hum Pathol, 2009. 40(4): p. 464-70.
132
303. Chandran, U.R., et al., Gene expression profiles of prostate cancer reveal
involvement of multiple molecular pathways in the metastatic process. BMC
Cancer, 2007. 7: p. 64.
304. Dai, B., et al., Aberrant FoxM1B expression increases matrix metalloproteinase-2
transcription and enhances the invasion of glioma cells. Oncogene, 2007. 26(42):
p. 6212-9.
305. Radhakrishnan, S.K. and A.L. Gartel, FOXM1: the Achilles' heel of cancer? Nat
Rev Cancer, 2008. 8(3): p. c1; author reply c2.
306. Spirin, K.S., et al., p27/Kip1 mutation found in breast cancer. Cancer Res, 1996.
56(10): p. 2400-4.
307. Sanders, D.A., et al., Genome-wide mapping of FOXM1 binding reveals co-
binding with estrogen receptor alpha in breast cancer cells. Genome Biol, 2013.
14(1): p. R6.
308. Madureira, P.A., et al., The Forkhead box M1 protein regulates the transcription
of the estrogen receptor alpha in breast cancer cells. J Biol Chem, 2006. 281(35):
p. 25167-76.
309. Cicatiello, L., et al., A genomic view of estrogen actions in human breast cancer
cells by expression profiling of the hormone-responsive transcriptome. J Mol
Endocrinol, 2004. 32(3): p. 719-75.
310. Millour, J., et al., FOXM1 is a transcriptional target of ERalpha and has a
critical role in breast cancer endocrine sensitivity and resistance. Oncogene,
2010. 29(20): p. 2983-95.
311. Dai, J., et al., Prognostic Value of FOXM1 in Patients with Malignant Solid
Tumor: A Meta-Analysis and System Review. Dis Markers, 2015. 2015: p.
352478.
312. Saba, R., et al., The Role of Forkhead Box Protein M1 in Breast Cancer
Progression and Resistance to Therapy. Int J Breast Cancer, 2016. 2016: p.
9768183.
313. Ahn, H., et al., Increased expression of forkhead box M1 is associated with
aggressive phenotype and poor prognosis in estrogen receptor-positive breast
cancer. J Korean Med Sci, 2015. 30(4): p. 390-7.
314. Craig, D.W., et al., Genome and transcriptome sequencing in prospective
metastatic triple-negative breast cancer uncovers therapeutic vulnerabilities. Mol
Cancer Ther, 2013. 12(1): p. 104-16.
315. Ellis, M.J. and C.M. Perou, The genomic landscape of breast cancer as a
therapeutic roadmap. Cancer Discov, 2013. 3(1): p. 27-34.
316. Prat, A., et al., Genomic analyses across six cancer types identify basal-like
breast cancer as a unique molecular entity. Sci Rep, 2013. 3: p. 3544.
317. Halasi, M. and A.L. Gartel, Targeting FOXM1 in cancer. Biochem Pharmacol,
2013. 85(5): p. 644-52.
318. Koo, C.Y., K.W. Muir, and E.W. Lam, FOXM1: From cancer initiation to
progression and treatment. Biochim Biophys Acta, 2012. 1819(1): p. 28-37.
319. Polakis, P., Wnt signaling and cancer. Genes Dev, 2000. 14(15): p. 1837-51.
320. Cai, C. and X. Zhu, The Wnt/beta-catenin pathway regulates self-renewal of
cancer stem-like cells in human gastric cancer. Mol Med Rep, 2012. 5(5): p.
1191-6.
321. Holland, J.D., et al., Wnt signaling in stem and cancer stem cells. Curr Opin Cell
Biol, 2013. 25(2): p. 254-64.
133
322. Quan, M., et al., The roles of FOXM1 in pancreatic stem cells and
carcinogenesis. Mol Cancer, 2013. 12: p. 159.
323. Gong, A. and S. Huang, FoxM1 and Wnt/beta-catenin signaling in glioma stem
cells. Cancer Res, 2012. 72(22): p. 5658-62.
324. van de Wetering, M., et al., Armadillo coactivates transcription driven by the
product of the Drosophila segment polarity gene dTCF. Cell, 1997. 88(6): p. 789-
99.
325. Faustino-Rocha, A., et al., Estimation of rat mammary tumor volume using
caliper and ultrasonography measurements. Lab Anim (NY), 2013. 42(6): p. 217-
24.
326. in Guide for the Care and Use of Laboratory Animals. 2011: Washington (DC).
327. Workman, P., et al., Guidelines for the welfare and use of animals in cancer
research. Br J Cancer, 2010. 102(11): p. 1555-77.
328. Wang, X., et al., PrimerBank: a PCR primer database for quantitative gene
expression analysis, 2012 update. Nucleic Acids Res, 2012. 40(Database issue):
p. D1144-9.
329. Altschul, S.F., et al., Basic local alignment search tool. J Mol Biol, 1990. 215(3):
p. 403-10.
330. Altschul, S.F., et al., Gapped BLAST and PSI-BLAST: a new generation of protein
database search programs. Nucleic Acids Res, 1997. 25(17): p. 3389-402.
331. Cock, P.J., et al., The Sanger FASTQ file format for sequences with quality
scores, and the Solexa/Illumina FASTQ variants. Nucleic Acids Res, 2010. 38(6):
p. 1767-71.
332. Ewing, B., et al., Base-calling of automated sequencer traces using phred. I.
Accuracy assessment. Genome Res, 1998. 8(3): p. 175-85.
333. Ewing, B. and P. Green, Base-calling of automated sequencer traces using phred.
II. Error probabilities. Genome Res, 1998. 8(3): p. 186-94.
334. Lander, E.S., et al., Initial sequencing and analysis of the human genome. Nature,
2001. 409(6822): p. 860-921.
335. Dobin, A., et al., STAR: ultrafast universal RNA-seq aligner. Bioinformatics,
2013. 29(1): p. 15-21.
336. Curwen, V., et al., The Ensembl automatic gene annotation system. Genome Res,
2004. 14(5): p. 942-50.
337. Benjamini, Y.a.H.Y., Controlling the False Discovery Rate: A Practical and
Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society.
Series B (Methodological), 1995. Vol. 57(No. 1): p. 289-300.
338. Cerami, E., et al., The cBio cancer genomics portal: an open platform for
exploring multidimensional cancer genomics data. Cancer Discov, 2012. 2(5): p.
401-4.
339. Gao, J., et al., Integrative analysis of complex cancer genomics and clinical
profiles using the cBioPortal. Sci Signal, 2013. 6(269): p. pl1.
340. Ciriello, G., et al., Comprehensive Molecular Portraits of Invasive Lobular Breast
Cancer. Cell, 2015. 163(2): p. 506-19.
341. Zhu, J., et al., The UCSC Cancer Genomics Browser. Nat Methods, 2009. 6(4): p.
239-40.
342. Goldman, M., et al., The UCSC Cancer Genomics Browser: update 2015. Nucleic
Acids Res, 2015. 43(Database issue): p. D812-7.
134
343. Rhodes, D.R., et al., ONCOMINE: a cancer microarray database and integrated
data-mining platform. Neoplasia, 2004. 6(1): p. 1-6.
344. Schneider, C.A., W.S. Rasband, and K.W. Eliceiri, NIH Image to ImageJ: 25
years of image analysis. Nat Methods, 2012. 9(7): p. 671-5.
345. Collins, T.J., ImageJ for microscopy. Biotechniques, 2007. 43(1 Suppl): p. 25-30.
346. Emami, K.H., et al., A small molecule inhibitor of beta-catenin/CREB-binding
protein transcription [corrected]. Proc Natl Acad Sci U S A, 2004. 101(34): p.
12682-7.
347. Phizicky, E.M. and S. Fields, Protein-protein interactions: methods for detection
and analysis. Microbiol Rev, 1995. 59(1): p. 94-123.
348. Kuroda, T., S.D. Rabkin, and R.L. Martuza, Effective treatment of tumors with
strong beta-catenin/T-cell factor activity by transcriptionally targeted oncolytic
herpes simplex virus vector. Cancer Res, 2006. 66(20): p. 10127-35.
349. Stambolic, V., L. Ruel, and J.R. Woodgett, Lithium inhibits glycogen synthase
kinase-3 activity and mimics wingless signalling in intact cells. Curr Biol, 1996.
6(12): p. 1664-8.
350. Liu, J., et al., Targeting Wnt-driven cancer through the inhibition of Porcupine by
LGK974. Proc Natl Acad Sci U S A, 2013. 110(50): p. 20224-9.
351. Golebiewska, A., et al., Critical appraisal of the side population assay in stem
cell and cancer stem cell research. Cell Stem Cell, 2011. 8(2): p. 136-47.
352. Takemaru, K.I. and R.T. Moon, The transcriptional coactivator CBP interacts
with beta-catenin to activate gene expression. J Cell Biol, 2000. 149(2): p. 249-
54.
353. Hecht, A., et al., The p300/CBP acetyltransferases function as transcriptional
coactivators of beta-catenin in vertebrates. EMBO J, 2000. 19(8): p. 1839-50.
354. King, T.D., M.J. Suto, and Y. Li, The Wnt/beta-catenin signaling pathway: a
potential therapeutic target in the treatment of triple negative breast cancer. J
Cell Biochem, 2012. 113(1): p. 13-8.
355. Hao, S., et al., Targeted inhibition of beta-catenin/CBP signaling ameliorates
renal interstitial fibrosis. J Am Soc Nephrol, 2011. 22(9): p. 1642-53.
356. Henderson, W.R., Jr., et al., Inhibition of Wnt/beta-catenin/CREB binding protein
(CBP) signaling reverses pulmonary fibrosis. Proc Natl Acad Sci U S A, 2010.
107(32): p. 14309-14.
357. Sasaki, T., et al., The small molecule Wnt signaling modulator ICG-001 improves
contractile function in chronically infarcted rat myocardium. PLoS One, 2013.
8(9): p. e75010.
358. Chan, K.C., et al., Therapeutic targeting of CBP/beta-catenin signaling reduces
cancer stem-like population and synergistically suppresses growth of EBV-
positive nasopharyngeal carcinoma cells with cisplatin. Sci Rep, 2015. 5: p. 9979.
359. Lenz, H.J. and M. Kahn, Safely targeting cancer stem cells via selective catenin
coactivator antagonism. Cancer Sci, 2014. 105(9): p. 1087-92.
360. Wend, P., et al., Wnt/beta-catenin signalling induces MLL to create epigenetic
changes in salivary gland tumours. EMBO J, 2013. 32(14): p. 1977-89.
361. Koehler, A.N., A complex task? Direct modulation of transcription factors with
small molecules. Curr Opin Chem Biol, 2010. 14(3): p. 331-40.
362. Redmond, A.M. and J.S. Carroll, Defining and targeting transcription factors in
cancer. Genome Biol, 2009. 10(7): p. 311.
135
363. Hopkins, A.L. and C.R. Groom, The druggable genome. Nat Rev Drug Discov,
2002. 1(9): p. 727-30.
364. Mapp, A.K., R. Pricer, and S. Sturlis, Targeting transcription is no longer a
quixotic quest. Nat Chem Biol, 2015. 11(12): p. 891-4.
365. Arkin, M.R., Y. Tang, and J.A. Wells, Small-molecule inhibitors of protein-
protein interactions: progressing toward the reality. Chem Biol, 2014. 21(9): p.
1102-14.
366. Berg, T., Modulation of protein-protein interactions with small organic
molecules. Angew Chem Int Ed Engl, 2003. 42(22): p. 2462-81.
367. Fletcher, S., J. Turkson, and P.T. Gunning, Molecular approaches towards the
inhibition of the signal transducer and activator of transcription 3 (Stat3) protein.
ChemMedChem, 2008. 3(8): p. 1159-68.
368. Issaeva, N., et al., Small molecule RITA binds to p53, blocks p53-HDM-2
interaction and activates p53 function in tumors. Nat Med, 2004. 10(12): p. 1321-
8.
369. Zhan, C., et al., An ultrahigh affinity d-peptide antagonist Of MDM2. J Med
Chem, 2012. 55(13): p. 6237-41.
370. Avadisian, M., et al., Artificially induced protein-membrane anchorage with
cholesterol-based recognition agents as a new therapeutic concept. Angew Chem
Int Ed Engl, 2011. 50(28): p. 6248-53.
371. Xie, X., et al., Targeting HPV16 E6-p300 interaction reactivates p53 and inhibits
the tumorigenicity of HPV-positive head and neck squamous cell carcinoma.
Oncogene, 2014. 33(8): p. 1037-46.
372. Chen, H., et al., Adenovirus-mediated RNA interference targeting FOXM1
transcription factor suppresses cell proliferation and tumor growth of
nasopharyngeal carcinoma. J Gene Med, 2012. 14(4): p. 231-40.
373. Wang, M. and A.L. Gartel, The suppression of FOXM1 and its targets in breast
cancer xenograft tumors by siRNA. Oncotarget, 2011. 2(12): p. 1218-26.
374. Yang, C., et al., Inhibition of FOXM1 transcription factor suppresses cell
proliferation and tumor growth of breast cancer. Cancer Gene Ther, 2013. 20(2):
p. 117-24.
375. Bhat, U.G., M. Halasi, and A.L. Gartel, Thiazole antibiotics target FoxM1 and
induce apoptosis in human cancer cells. PLoS One, 2009. 4(5): p. e5592.
376. Kwok, J.M., et al., Thiostrepton selectively targets breast cancer cells through
inhibition of forkhead box M1 expression. Mol Cancer Ther, 2008. 7(7): p. 2022-
32.
377. Pandit, B. and A.L. Gartel, New potential anti-cancer agents synergize with
bortezomib and ABT-737 against prostate cancer. Prostate, 2010. 70(8): p. 825-
33.
378. Zhang, L., et al., Antibiotic susceptibility of mammalian mitochondrial
translation. FEBS Lett, 2005. 579(28): p. 6423-7.
379. Aagaard, L. and J.J. Rossi, RNAi therapeutics: principles, prospects and
challenges. Adv Drug Deliv Rev, 2007. 59(2-3): p. 75-86.
380. Dontu, G. and M.S. Wicha, Survival of mammary stem cells in suspension
culture: implications for stem cell biology and neoplasia. J Mammary Gland Biol
Neoplasia, 2005. 10(1): p. 75-86.
381. Dontu, G., et al., In vitro propagation and transcriptional profiling of human
mammary stem/progenitor cells. Genes Dev, 2003. 17(10): p. 1253-70.
136
382. Morrison, S.J. and J. Kimble, Asymmetric and symmetric stem-cell divisions in
development and cancer. Nature, 2006. 441(7097): p. 1068-74.
383. Kahn, M., Symmetric division versus asymmetric division: a tale of two
coactivators. Future Med Chem, 2011. 3(14): p. 1745-63.
384. Kahn, M., Can we safely target the WNT pathway? Nat Rev Drug Discov, 2014.
13(7): p. 513-32.
385. Meacham, C.E. and S.J. Morrison, Tumour heterogeneity and cancer cell
plasticity. Nature, 2013. 501(7467): p. 328-37.
386. Tang, D.G., Understanding cancer stem cell heterogeneity and plasticity. Cell
Res, 2012. 22(3): p. 457-72.
387. Gupta, P.B., et al., Stochastic state transitions give rise to phenotypic equilibrium
in populations of cancer cells. Cell, 2011. 146(4): p. 633-44.
388. Harris, M.A., et al., Cancer stem cells are enriched in the side population cells in
a mouse model of glioma. Cancer Res, 2008. 68(24): p. 10051-9.
389. Ho, M.M., et al., Side population in human lung cancer cell lines and tumors is
enriched with stem-like cancer cells. Cancer Res, 2007. 67(10): p. 4827-33.
390. Chiba, T., et al., Side population purified from hepatocellular carcinoma cells
harbors cancer stem cell-like properties. Hepatology, 2006. 44(1): p. 240-51.
391. Whittle, J.R., et al., Patient-derived xenograft models of breast cancer and their
predictive power. Breast Cancer Res, 2015. 17: p. 17.
392. Comen, E.A. and M. Robson, Poly(ADP-ribose) polymerase inhibitors in triple-
negative breast cancer. Cancer J, 2010. 16(1): p. 48-52.
393. Berrada, N., S. Delaloge, and F. Andre, Treatment of triple-negative metastatic
breast cancer: toward individualized targeted treatments or chemosensitization?
Ann Oncol, 2010. 21 Suppl 7: p. vii30-5.
394. Zona, S., et al., FOXM1: an emerging master regulator of DNA damage response
and genotoxic agent resistance. Biochim Biophys Acta, 2014. 1839(11): p. 1316-
22.
395. Early Breast Cancer Trialists' Collaborative, G., et al., Effect of radiotherapy after
breast-conserving surgery on 10-year recurrence and 15-year breast cancer
death: meta-analysis of individual patient data for 10,801 women in 17
randomised trials. Lancet, 2011. 378(9804): p. 1707-16.
396. Aggarwal, S., Targeted cancer therapies. Nat Rev Drug Discov, 2010. 9(6): p.
427-8.
397. Incorvati, J.A., et al., Targeted therapy for HER2 positive breast cancer. J
Hematol Oncol, 2013. 6: p. 38.
398. Puhalla, S., S. Bhattacharya, and N.E. Davidson, Hormonal therapy in breast
cancer: a model disease for the personalization of cancer care. Mol Oncol, 2012.
6(2): p. 222-36.
399. Groenendijk, F.H. and R. Bernards, Drug resistance to targeted therapies: deja vu
all over again. Mol Oncol, 2014. 8(6): p. 1067-83.
400. Ohshima, T., T. Suganuma, and M. Ikeda, A novel mutation lacking the
bromodomain of the transcriptional coactivator p300 in the SiHa cervical
carcinoma cell line. Biochem Biophys Res Commun, 2001. 281(2): p. 569-75.
401. Bryan, E.J., et al., Mutation analysis of EP300 in colon, breast and ovarian
carcinomas. Int J Cancer, 2002. 102(2): p. 137-41.
402. Kent, W.J., BLAT--the BLAST-like alignment tool. Genome Res, 2002. 12(4): p.
656-64.
137
403. Engebraaten, O., H.K. Vollan, and A.L. Borresen-Dale, Triple-negative breast
cancer and the need for new therapeutic targets. Am J Pathol, 2013. 183(4): p.
1064-74.
404. Bektas, N., et al., Tight correlation between expression of the Forkhead
transcription factor FOXM1 and HER2 in human breast cancer. BMC Cancer,
2008. 8: p. 42.
405. Francis, R.E., et al., FoxM1 is a downstream target and marker of HER2
overexpression in breast cancer. Int J Oncol, 2009. 35(1): p. 57-68.
406. Jatoi, I., et al., Significance of axillary lymph node metastasis in primary breast
cancer. J Clin Oncol, 1999. 17(8): p. 2334-40.
407. Brugger, W., et al., Prospective molecular marker analyses of EGFR and KRAS
from a randomized, placebo-controlled study of erlotinib maintenance therapy in
advanced non-small-cell lung cancer. J Clin Oncol, 2011. 29(31): p. 4113-20.
408. Siegel, R.L., K.D. Miller, and A. Jemal, Cancer statistics, 2016. CA Cancer J
Clin, 2016. 66(1): p. 7-30.
409. Tevaarwerk, A.J., et al., Survival in patients with metastatic recurrent breast
cancer after adjuvant chemotherapy: little evidence of improvement over the past
30 years. Cancer, 2013. 119(6): p. 1140-8.
410. Tseng, L.M., et al., Distant metastasis in triple-negative breast cancer.
Neoplasma, 2013. 60(3): p. 290-4.
411. Azim, H.A., Jr., et al., Long-term toxic effects of adjuvant chemotherapy in breast
cancer. Ann Oncol, 2011. 22(9): p. 1939-47.
412. Amir, E., et al., Tissue confirmation of disease recurrence in breast cancer
patients: pooled analysis of multi-centre, multi-disciplinary prospective studies.
Cancer Treat Rev, 2012. 38(6): p. 708-14.
413. Amir, E., et al., Prospective study evaluating the impact of tissue confirmation of
metastatic disease in patients with breast cancer. J Clin Oncol, 2012. 30(6): p.
587-92.
414. Curigliano, G., et al., Should liver metastases of breast cancer be biopsied to
improve treatment choice? Ann Oncol, 2011. 22(10): p. 2227-33.
415. Regitnig, P., et al., Change of HER-2/neu status in a subset of distant metastases
from breast carcinomas. J Pathol, 2004. 203(4): p. 918-26.
416. Malanchi, I., et al., Interactions between cancer stem cells and their niche govern
metastatic colonization. Nature, 2011. 481(7379): p. 85-9.
417. Tenbaum, S.P., et al., beta-catenin confers resistance to PI3K and AKT inhibitors
and subverts FOXO3a to promote metastasis in colon cancer. Nat Med, 2012.
18(6): p. 892-901.
418. Dragu, D.L., et al., Therapies targeting cancer stem cells: Current trends and
future challenges. World J Stem Cells, 2015. 7(9): p. 1185-201.
419. Ko, A.H.e.a., A phase Ib dose-escalation study of PRI-724, a CBP/beta-catenin
modulator, plus gemcitabine (GEM) in patients with advanced pancreatic
adenocarcinoma (APC) as second-line therapy after FOLFIRINOX or FOLFOX.
J Clin Oncol, 2016. 34.

Abstract (if available)

Abstract Background: Accurate assessment of prognostic and predictive biomarkers (ER, PR, HER2) plays a critical role in the clinical management of breast cancer. Triple negative breast cancers (TNBCs) lack the expression of all three targets, and no targetable molecular pathways have been identified to date. Hence, TNBCs are treated with non-targeted, cytotoxic chemotherapeutic agents (e.g. paclitaxel), and are characterized by high rates of drug resistance and disease relapse. CREB binding protein (CBP) has been implicated in cell growth and malignant transformation in various cancers. CBP is an important co-activator in the Wnt/β-catenin signaling pathway, which has been implicated in driving TNBC and cancer stem cell (CSC) biology. Cancer stem cells are responsible for tumor initiation, drug resistance and tumor recurrence. The Kahn lab has developed a specific CBP-binding small molecule inhibitor, ICG-001. The hypothesis for this study was that CBP signaling plays an important role in TNBC biology and may provide a novel therapeutic target. ❧ Methods: A chemico-genomic approach using ICG-001 in combination with whole transcriptome RNA Seq was utilized to characterize CBP-driven gene expression in TNBC cell line models. Co-immunoprecipitation (CoIP), protein and gene expression studies, and gene knockdown were used for validation. In vitro drug resistant cell line models as well as in vivo TNBC cell line and patient derived xenograft (PDX) NOD SCID gamma mouse models were used to determine the effect of CBP inhibition on tumor initiation, CSC populations and drug resistance. Tissue micro arrays (TMAs) were used to investigate the potential of CBP and associated proteins as prognostic and predictive biomarkers in TNBC. ❧ Results: RNA Seq analysis revealed that gene expression in TNBC is CBP dependent. Pathway analysis revealed the involvement of FOXM1 in CBP-driven gene expression in TNBC. CoIP demonstrated that CBP formed a transcriptional complex with FOXM1 and β-catenin. CBP/FOXM1 binding was critical for FOXM1 expression, and conferred an aggressive phenotype in TNBC, including drug resistance in vitro and in vivo, cancer stem cell phenotype and tumor initiation. Comparison of clinical data with FOXM1 expression in TNBC patient’s tumor samples demonstrated that high expression levels of FOXM1 were associated with disease relapse and poor survival outcome. PDX mouse models demonstrated that high levels of FOXM1 resulted in paclitaxel resistance and disease recurrence in TNBC PDX models. Treatment with ICG-001 led to down-regulation of FOXM1, eliminated CSCs and sensitized FOXM1 high tumors to paclitaxel treatment. Immunohistochemistry of TMAs demonstrated a significant correlation between FOXM1 expression and TNBC subtype. ❧ Conclusion: FOXM1 expression is dependent on CBP-binding. The CBP/FOXM1 transcriptional complex drives aggressive behavior of TNBC characterized by a CSC phenotype, drug resistance, as well as poor clinical outcome. Targeting CBP/FOXM1 via ICG-001 can ameliorate this aggressive phenotype and target CSC populations. FOXM1 expression levels could potentially serve as a biomarker in TNBC. These results could provide a clinical-translational rational for patient stratification in future clinical trials exploring the therapeutic potential of CBP/FOXM1 inhibition via ICG-001 in combination with chemotherapy.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The cancer stem-like phenotype: therapeutics, phenotypic plasticity and mechanistic studies

PDF

RNA methylation in cancer plasticity and drug resistance

PDF

The role of LGR5 in the pathogenesis of Ewing sarcoma: a marker of aggressive disease and a contributor to the malignant phenotype

PDF

Exploration of the roles of cancer stem cells and survivin in the pathogenesis and progression of prostate cancer

PDF

Applying multi-omics in cancer liquid biopsy for improved patient monitoring and biomarker discovery

PDF

Estrogen receptor-β characterization in breast cancer: development of a reliable assay for measuring expression

PDF

TLR8-transferred miR-192 acts as a tumor suppressor in neuroblastoma by inhibiting CTCF

PDF

Cisplatin activates mitochondrial oxphos leading to acute treatment resistance in bladder cancer

PDF

Context-dependent role of androgen receptor (AR) in estrogen receptor-positive (ER+) breast cancer

PDF

Identification of novel epigenetic biomarkers and microRNAs for cancer therapeutics

PDF

Ectopic expression of a truncated isoform of hair keratin 81 in breast cancer alters biophysical characteristics to promote metastatic propensity

PDF

Role of cancer-associated fibroblast secreted annexin A1 in generation and maintenance of prostate cancer stem cells

PDF

Targeting glioma cancer stem cells for the treatment of glioblastoma multiforme

PDF

Investigating the complexity of the tumor microenvironment's role in drug response

PDF

Creating a multiple micrornia expression vector to target GRP78, an ER chaperone and signaling regulator in cancer

PDF

Targeting molecular signals involved in the development of castration resistant prostate cancer

PDF

Differential role of two coactivators, CCAR1 and CARM1, for dysregulated beta-catenin activity in colorectal cancer cell growth and gene expression

PDF

Temporal and spatial characterization of cisplatin treatment and emerging acute resistance in bladder cancer cells

PDF

Exploring the effects of CXCR4 inhibition on circulating tumor cell populations in metastatic prostate cancer

PDF

Stable expression of human B7-H4 in a mouse mammary tumor model as a target for cancer immunotherapy

Asset Metadata

Creator Ring, Alexander (author)

Core Title Identification of CBP/FOXM1 as a molecular target in triple negative breast cancer

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Cancer Biology and Genomics

Publication Date 07/14/2017

Defense Date 05/30/2017

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

Tag OAI-PMH Harvest,triple negative breast cancer (TNBC), cancer stem cells (CSCs), Wnt signaling, CBP, FOXM1, small molecule, ICG-001, targeted therapy, tissue micro array (TMA), biomarker

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Yu, Min (committee chair), Goldkorn, Amir (committee member), Lang, Julie Eileen (committee member), Stallcup, Michael (committee member)

Creator Email aleq.ring@gmail.com,ringa@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-398410

Unique identifier UC11265597

Identifier etd-RingAlexan-5506.pdf (filename),usctheses-c40-398410 (legacy record id)

Legacy Identifier etd-RingAlexan-5506.pdf

Dmrecord 398410

Document Type Dissertation

Rights Ring, Alexander

Type texts

Source 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 a...

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