Novel Roles for Human Setd1a Histone Methyltransferase in Regulation of Canonical Wnt Signaling Pathway, Cellular Prolif...


Material Information

Novel Roles for Human Setd1a Histone Methyltransferase in Regulation of Canonical Wnt Signaling Pathway, Cellular Proliferation, and Metastasis
Physical Description:
1 online resource (121 p.)
Salz, Tal H
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Medical Sciences, Biochemistry and Molecular Biology (IDP)
Committee Chair:
Committee Co-Chair:
Committee Members:


Subjects / Keywords:
breast -- cancer -- colorectal -- epigenetics -- h3k4me3 -- metastasis -- signaling -- wnt
Biochemistry and Molecular Biology (IDP) -- Dissertations, Academic -- UF
Medical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation


Changes in epigenetic signatures are characteristic of all cancers. One such phenomenon is perturbation in methylation patterns of histone H3 lysin 4 (H3K4) during cancer progression. The drivers and function of such changes are not clear. H3K4 methylation is enacted by the COMPASS family of histone methyltransferases, which includes two SETD1 and four MLL enzymatic complexes. While SETD1A regulates promoter associated H3K4 trimethylation, MLLs regulate enhancer associated H3K4 mono and dimethylation, and demonstrate only minor effect on global gene expression profile. The purpose of this study is to elucidate the role of human SETD1A in tumorigenesis and explore related underlying mechanisms. Here, we report that hSETD1A, and hSETD1B in partial, are up regulated in colorectal and breast cancer cells and specimens. Depletion of hSETD1A led to a dramatic decrease in bulk H3K4me3 and to retarded cellular proliferation, invasion, and migration in colorectal or breast cancer cells. In parallel, ablation of hSETD1A suppressed tumor growth and cell metastases in nude mice. To identify genes affected by depletion of hSETD1A, we utilized a genome wide microarray analysis approach, which revealed a significant decrease in Wnt signaling target gene expression. Mutations in this pathway are prevalent in in colorectal cancer, and are critical for the manifestation of the disease. We found that hSETD1A interacts with Beta catenin, the master regulator of the Wnt signaling pathway, and that these interactions facilitate the localization of hSETD1A to beta catenin responsive promoters, H3K4 trimethylation enrichment, TAF3 occupancy, and subsequent increase in Wnt target gene expression. These results were supported by the significant overlap in target genes of hSETD1A, TAF3, and beta catenin in colorectal cancer cells, and by the correlation between elevated levels of hSETD1A transcript and increased levels of Wnt target genes and promoter associated H3K4 trimethylation in colorectal tumors. Finally, increased levels of hSETD1A and hSETD1B, but not MLLs, associated with poor survival in lymph node positive breast cancer, suggesting that these enzymes could serve as good biomarkers. Taken together, this work provides novel biologic and mechanistic perspectives into the role of hSETD1A in cancer, and provides a potential candidate for cancer therapeutic applications.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Tal H Salz.
Thesis (Ph.D.)--University of Florida, 2014.
Co-adviser: QIU,YI.

Record Information

Source Institution:
Rights Management:
Applicable rights reserved.
lcc - LD1780 2014
System ID:

This item is only available as the following downloads:

Full Text




2014 Tal Hila Salz


This dissertation is dedicated to the memory of my grandfa ther, Jack Salz, without whom I would most certainly be lost


4 ACKNOWLEDGMENTS I have been fortunate to attend the University of Florida and to be surrounded by a number of influential people who have guided and supported me through graduate school a nd made it possible for me to reach this milestone in my academic career. I would first like to express my appreciation to my committee chair and advisor, Professor Suming Huang, who supported my graduate education, introduced me to epigenetic s, and direc ted me in selecting the f inal theme for this research. He allowed me to explore my true scientific interests and ensured that I take the right steps towards becoming an independent researcher and towards a future career. Without his guidance and patience t his dissertation would not have been possible. I would like to acknowledge my fantastic dissertation committee members, Professors Kevin Brown, Yi Qiu, Jorg Bungert, Peter Sayeski, and Maria Zajac Kaye for their constructive criticism s and intellectual con tributions, patience, caring and kindness I would also like to thank Dr. Frederic Kaye who kindly agreed to serve as my co mentor and whose questions always ignited a healthy rethinking process in me. I am indebted to Professor Kevin Brown, with whom I o ften debated experimental issues, and whose enthusiasm for cancer research and ruthless honesty had a lasting effect on me Special thanks go to Dr. Lei Zhou, Peter Sayeski, Jorg Bungert, Maria Zajac Kaye, Frederic Kaye and Yi Qiu who provided me with let ters of support for numerous grants and other applications Finally, I thank my lab mates for fun times and for experimental support. Most importantly, I would like to express my deepest appreciation to my wonderful family and friends for their uncondition al support and cheering throughout my


5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 ABSTRACT ................................ ................................ ................................ ................... 12 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 14 Cancer Epigenetic ................................ ................................ ................................ .. 14 The COMPASS Family of Histone Methyltransferases ................................ .... 15 Roles for H3K4me3 in Transcription Activation ................................ ................ 17 Roles for H3K4me3 in Cancer ................................ ................................ .......... 19 Roles for SETD1A in Development and Cancer ................................ ............... 20 Colorectal Cancer and the Wnt signaling Pat hway ................................ ................. 21 Colorectal Cancer Epidemiology ................................ ................................ ...... 22 The Wnt signaling Pathway ................................ ................................ .............. 23 Epigenetic Regulation of Wnt Target Genes ................................ .................... 25 The Role of Wnt signaling Pathway in Colorectal Cancer ................................ 26 Regulation of Intestina l Stem Cells (ISC) By the Wnt signaling Pathway ......... 27 Can We Win WNT? ................................ ................................ .......................... 28 Breast Cancer and Metastasis ................................ ................................ ................ 30 Molecular and Biologic al Basics of Metastasis ................................ ................. 31 Epithelial Mesenchymal Transition (EMT) ................................ ........................ 33 Rol e for Epigenetic in EMT ................................ ................................ ............... 34 2 MATERIALS AND METHODS ................................ ................................ ................ 41 Cell Culture and Reagents ................................ ................................ ...................... 41 ShRNA mediated Knock Down ................................ ................................ ............... 42 Antibodies ................................ ................................ ................................ ............... 43 Western Immunoblotting ................................ ................................ ......................... 43 Co immunoprecipitation ................................ ................................ .......................... 44 Immunofluorescence ................................ ................................ ............................... 45 Immunohistochemistry ................................ ................................ ............................ 46 Chromatin Immunoprecipitation (ChIP) ................................ ................................ ... 46 Real Time PCR ................................ ................................ ................................ ....... 47 Growth Assays ................................ ................................ ................................ ........ 48 Migration Assay ................................ ................................ ................................ ...... 49 Invasion Assay ................................ ................................ ................................ ........ 49 Wound Healing Assay ................................ ................................ ............................ 49 Cell Cycle Analysis ................................ ................................ ................................ 50


6 Microarray Analysis ................................ ................................ ................................ 50 Colorectal and Breast Patient Samples ................................ ................................ .. 51 Mouse Xenograft Model ................................ ................................ .......................... 51 In vivo Metastasis ................................ ................................ ................................ ... 52 Statistical Analysis ................................ ................................ ................................ .. 52 3 HUMAN SETD1A AND H3K4ME3 ARE UP REGULATED IN COLORECTAL CANCER AND CONTROL CELLULAR GROWTH ................................ ................. 55 Introductory Remarks ................................ ................................ .............................. 55 Results ................................ ................................ ................................ .................... 55 hSETD1A and H3K4me3 are Up regulated in Human Colorectal Cancer ........ 55 hSETD1A Impacts Proliferation of Colorectal Cancer Cells ............................. 56 Closing Remarks ................................ ................................ ................................ .... 57 4 HUMAN SETD1A REGULATES THE WNT SIGNALING PATHWAY ..................... 64 Introductory Remarks ................................ ................................ .............................. 64 Results ................................ ................................ ................................ .................... 64 hSETD1A Regulates Transcription of Wnt Targe t Genes ................................ 64 hSETD1A Associates with catenin and Activates Wnt Target Genes ............ 65 Recruitment of hSETD1A by catenin is Critical for catenin mediated Transcriptional Activation ................................ ................................ .............. 67 Levels of hSETD1A are Positively Correlated with Activation of Wnt Target Genes and Tumor Growth in Colorectal Cancer. ................................ .......... 68 Closing Remarks ................................ ................................ ................................ .... 69 5 HUMAN SETD1A REGULATES METASTASIS IN BREAST CANCER .................. 82 Introductory Remarks ................................ ................................ .............................. 82 Results ................................ ................................ ................................ .................... 83 hSETD1A and hSETD1B are Up regulated in Breast Cancer .......................... 83 Increased Levels of hSETD1A and hSETD1B Are Associated with Poor Survival in Lymph nod e Positive Breast Cancer ................................ ........... 84 Perturbed hSETD1A Reduces Metastasis ................................ ........................ 85 Closing Remarks ................................ ................................ ................................ .... 86 6 SUMMARY AND DISCUSSION ................................ ................................ .............. 92 Significance ................................ ................................ ................................ ............ 92 Clinical Relevance ................................ ................................ ................................ .. 97 Future Direction ................................ ................................ ................................ .... 101 APPENDIX: THE LONG NON CODING RNA HEMOLINC IS UP REGULATED IN MESENCHYMAL BREAST CANCER CELLS ................................ ...................... 105


7 LIST OF REFERENCES ................................ ................................ ............................. 108 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 120


8 LIST OF TABLES Table page 2 1 Human shRNAs ................................ ................................ ................................ ..... 53 2 2 RT PCR Primer Sequences ................................ ................................ ................... 53 2 3 ChIP Primer Sequences ................................ ................................ ......................... 54 4 1 Expression Profile of Wnt target genes in hSETD1A KD HCT116 cells ................. 72 5 1 hSETD1A gene copy analysis in various types of human cancers ......................... 87


9 LIST OF FIGURES Figure page Figure 1 1. The COMPASS/Trithorax family of H3K4 methyltrasnferases .................... 36 Figure 1 2. Multi stage progression model of colorectal cancer. ................................ .. 37 Figure 1 3. Schematic representation of canonical Wnt signaling pathway ................. 38 Figure 1 4. Cellular and structural organization at the intestinal crypt .......................... 39 Figure 1 5. Schematic representation of the metastasi s process ................................ 40 Figure 3 1. hSETD1A and H3K4me3 are up regulated in hum an colorectal cancer .... 59 Figure 3 2. hSETD1A exclusively affects global H3K4me3 ................................ .......... 61 Figure 3 3. hSETD1A impacts cellular growth ................................ .............................. 62 Figure 3 4. Depletion of hSETD1A is not associated with cell death ............................. 63 Figure 4 1. hSETD1A regulates a subset of Wnt signaling target genes ..................... 70 Figure 4 catenin ................................ ......................... 73 Figure 4 3. hSETD1A mediates promoter H3K4me3 at Wnt target loci ....................... 74 Figure 4 4. catenin recruits hSETD1A to regulate Wnt target gene expression ......... 75 F igure 4 catenin levels lead to an increase in H3K4me3 ......................... 77 Figure 4 6. TAF3 and hSETD1A co regulate Wnt target genes ................................ ... 78 Figure 4 7. hSETD1A levels associate with H3K4me3 and Wnt target gene expression in human colorecta l tumors ................................ ............................. 79 Figure 4 catenin in regulation of Wnt ta rget genes in colorectal cancer .......................... 81 Figure 5 1. hSETD1A, hSETD1B, and H3K4me3 are up regulated in breast cancer ... 88 Figure 5 2. hSETD1A and h SETD1B levels predict clinical outcomes for lymph node positive breast cancer patients ................................ ................................ ........... 89 Figure 5 3. Depletion of hSETD1A reduces cell invasion and migration ...................... 90 Figure 5 4. KD of hSE TD1A reduces metastasis in mice ................................ ............. 91 Figure A 1. Hemolinc expression in breast cancer. ................................ .................... 107


10 LIST OF ABBREVIATIONS Ash 2L CDH1 COMPASS CRC EMT Epigenetic H3K4 H3K4me1 H3K4me2 H3K4me3 HCFC1 HDAC HMT JARID1A KD LEF1 LSD1 MET MLL MMP7 MYC PHD qRT PCR Three prime Untranslated R egion Absent, Small, or Homeotic Discs 2 L ike cadherin 1, type 1 Complex proteins associated with Set 1 Colorectal Cancer Epithlial Mesenchymal Transition over, above genetic Histone 3 Lysin 4 Mono methylation of Histone 3 Lysin 4 Di methylation of Histone 3 Lysin 4 Tri methylation of Histone 3 Lysin 4 Host Cell Factor C1 Histone Deacetylase H istone Methyltransferase Jumonji, AT Rich Interactive Domain 1B Knock Down Lymphoid Enhancer binding F actor 1 Lysine specific histone demethylase 1A Mesenchymal Epithelial Transition M ixed Lineage Leukemia matrix metalloproteinase 7 Myelocytomatosis viral oncogene Plant Homeo Domain Quantitative Reverse Transcription Polymerase Chain Reaction


11 Rb BP5 RNAP II RT PCR SD SETD1A SETD1 B SNAI 1 /Snail SNAI2/S lug TAF3 TAL1 TCF7L2/TCF4 TERT TFIID TSS TWIST2 VEGFA WB WDR5 Wnt eta catenin Retinoblastoma Binding Protein 5 RNA Polymerase II Reverse Transcription Polymerase Chain Reaction Standard Deviation SET domain containing protein 1 A SET domain containing protein 1 B Snail Homolog 1 Snail Homolog 2 TATA box binding protein (TBP) Associated Factor 3 T cell acute lymphoblastic leukemia protein 1 Transcription Factor 7 Lik e 2 / T Cell Specific Factor 7 Like 2 Telomerase reverse transcriptase Transcription Factor II D Transcription Start Site Twist Family Basi c Helix L oop Helix Transcription Factor 2 Vascular endothelial growth factor A Western Blot WD Repeat Containing Protein 5 Wingless / h omolog of int 1 C adherin associated protein beta 1


12 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosop hy NOVEL ROLES FOR HUMAN SETD1A HISTONE METHYLTRANSFERASE IN REGULATION OF CANONICAL WNT SIGNALING PATHWAY, CELLULAR PROLIFERATION, AND METASTASIS By TAL SALZ May 2014 Chair: Suming Huang Major: Medical Sciences L arge and small scale changes in e pig enetic signatures are characteristic of all cancers. One such phenomenon is perturbation in methylation patterns of histone H3 lysin 4 (H3K4) during progression of cancer The drivers and functions of such changes are not clear. H3K4 methylation is enacte d by the COMPASS family of histone methyltransferases which includes MLL1 4 and S ET D 1A/B. While SET D 1A/ B regulate s promoter associated H3K4me3 and affect s global gene expression, MLLs are mainly responsible for enhancer associated H3K4me1 / 2 and demonstrat e only minor effect on global gene expression profile The purpose of this study is to elucidate the role of human SET D 1 A (hSETD1A) in oncogenic functions and explore related underlying mechanisms. Here, we report that hSET D 1A and hSETD1B in partial, are up regulated in col orectal and breast cancer cells and specimens Depletion of hSET D 1A led to a dramatic decrease in bulk H3K4me3 in colorectal cancer cells and to retarded cellular proliferation, invasion, and migration in colorec tal and /or breast cancer cells In parallel, ablation of hSET D 1A suppressed tumor growth and cell metastases in nude mice To


13 identify gene s affected by hSETD1A KD, we utilized a g enome wide microarray analysis approach, which revealed a significant decrease in expression levels of Wnt signaling target genes Mutations in this pathway are prevalent in many cancers, especially in colorectal cancer, and are critical for the manifestation of the disease. We found that hSET D 1A interacts with catenin, the master regulator of the Wnt signaling pathway, and that these interactions facilitate the localization of hSET D 1A to Wnt/ catenin responsive promoters, H3K4 me3 enrichment TAF3 occupancy and subsequent increase in Wnt target gene s These results were supported by the si gnificant overlap in target genes of hSET D 1A, TAF3, and caten in in colorectal cancer cells and by the correlation between elevated levels of hSETD1A transcript and increased levels of Wnt target genes and promoter H3K4me3 in colorectal tumors. Finally, increased l evels of hSETD1A and hSETD1B, but not MLLs, associated with poor survival in ly mph node positive breast cancer, suggesting that these enzymes could serve as good bio marker s Taken together, this work pro vides novel b iologic and mechanistic pers pective s into the role of hSET D 1A/B in cancer and provides a potential candidate for cancer therapeutic applications.


14 CHAPTER 1 INTRODUCTION Cancer Epigenetic Eukaryotic DNA is a negatively charged molecule that is compacted by positively charged histone s [ 1 ] Together they form nucleosomes [ 1 ] Each nucleosome molecule consists of a core particle containing 146 DNA base pairs wrapped around two tetramers each co nsisted of histone H2A, H2B, H3 and H4 and a linker DNA containing histone H1 [ 1 ] Despite the need for DNA compaction, DNA must remain accessible to maintain functionality and allow for appropriate gene transcription and DNA replication to occur One important mechanism impacting DNA accessibility is histone modifica tion s [ 2 4 ] Each histone contains u nstructured amino terminal tail that protrudes out of th e nucleosomal core [ 1 ] The histone tails are prone to various post translationa l modifications including acetylation methylation, phosphorylation sumoylation, ADP ribosylation, ubiquitination and proline isomerization [ 2 4 ] These epigenet ic marks which are subjected to dynamic reading writing by transcription regulators, act upon chromatin structure and regulate cellular function s such as gene expression, splicing, DNA replication, DNA damage response telomere lengt h, and life span [ 2 4 ] Cancer is a disorder driven by genomic instability [ 5 ] It comprises six hallmarks which are acquired during progression: enhanced proliferation, inhibition of growth suppression, resistance to apoptosis, immortalization, angiogenesis, and metastasi s [ 6 ] Even prior to the awareness of important epigenetic modifications in cancer, conventional tumor biology focused on disruption in gene expression underlying tumor growth and metastasis. Perhaps the earliest link between epigenetic and cancer was


15 the discovery of DNA hypermethylation of tumor suppressor genes an d hypomethy lation of oncogenes [ 7 9 ] L ater on post transla tional modifications of histone were acknowledged to facilitate abnormal gene expression in cancer cells [ 2 10 11 ] This includes dynamic small and large scale changes i n histone modifications affecting heterochromatin and euchromatin domains [ 11 ] The absolute role of b ulk changes in histone modifications in cancer cells is not entirely clear, but it seems to play a role in allow ing for greater plasti city, rapid adaptation to new environments and selection of cancer favorite attributes [ 10 12 ] These functions support phenotypic alterations in the context o f both development and tumorigenesis [ 10 11 ] M utations and altered expression of various epigenetic modifie rs have been evident in cancer [ 2 11 ] The s e discoveries provide a potential mechanism for the chaotic chromatin landscape observed in cancer cells and portray epigenetic regulators as potential target s for cancer therapy. Additionally, identification of epigenetic changes associated with cancer progression early on could serve as a biomarker and enable better progn osis [ 13 ] Together, understanding the epigenetic changes occurring in cancer cells and identifying their source provides powerful clinical applications including diagnosis, prognosis, and th erapy [ 2 13 ] The COMPASS Family of Histone Methyltransferases One important and highly studied hi stone modification is the methylation of histone H3 on lysine 4 (H3K4). This histone mark occurs in three states: mono di and tri methyl ation It is mainly abundant at the transcription start site (TSS), enhancers, and gene body of active genes, and re gulates transcription [ 14 18 ] The first H3K4 methyltransferase was identified in Saccharomyces cerevisiae and was termed SETD1A /COMPASS [ 18 20 ] SETD1A /COMPASS homologs were later identified in


16 higher eukaryotes ( Figure 1 1) [ 18 ] F or example three SETD1A /COMPASS homologs were found in Drosophila melanogaster, while six family members were found in Homo Sapien s [ 17 18 ] All family members are similar in their capacity to methylate H3K4, but different in their ability to Mono Di or Tri methylate H3K4 (H3K4me1, H3K4me2, H3K4me3 ), and in their affinity to various loci [ 18 21 24 ] The human COMPASS family includes the SET1 /COMPASS genes hSETD1A and h SETD1 B, the COMPASS like genes MLL1 and MLL2 and the COMPASS related genes MLL3 and MLL4 (Figure 1 1) [ 18 ] COMPASS HMTs operate as l arge enzymatic complexes, each containing several common subunits and few unique subunits all of which are required for their enzymatic activity (Figure 1 1) [ 18 ] The subunit s Ash2L, RbBP5, WDR5 and DPY30 are present in all COMPASS complexes and are therefore perceived as core subunits (Figure 1 1) [ 18 ] The subunits HCF 1, Wdr82, and cxxc1, are unique to the h SETD1A/B complexes (Figure 1 1) [ 18 ] In Saccharomyces cerevisiae null mutants of anyone of the SETD1A /COMPASS subunits led to growth retardation and sensitivity to h ydroxyurea [ 19 ] The only yeast SET1 /COMPASS HMT complex is unique in its ability to mono di and trime thylate H3K4 [ 19 ] In eukaryotic cells, there is a division of labor among the trithorax members; SETD1 complexes regulate the bulk promoter H3K4me3, while the MLL complexes regulate H3K4me1 and H3K4me2 at gene enhancer elements [ 18 21 24 ] Ablation of SETD1A proteins in mammalian cells heavily affected the genome wide ex pression profile while ablation of MLL proteins had only minor and localized effect on


17 gene expression. However, all COMPASS family members play an important role in normal development and various malignancies [ 18 ] Roles for H3K4me3 i n Transcription Activation The importance of H3K4me3 as a prominent mark first recognized w hen chr omati n immunoprecipitation (ChIP) combined with rapid genome wide sequencing techniques became available [ 14 16 25 ] Immunopre cipitation of eukaryotic chromatin with antibodies against H3K4me3 revealed specific enrichment of H3K4me3 at transcription start sites (TSS) of active genes [ 14 16 25 ] Other histone modifications, such as H3K27me3 and H3K9me2 were shown to act as repressive mark s at heterochromatin regions [ 14 ] Moreover, H3K4me3 was shown present, along with H3K27me3 at promoters of developmental genes in embryonic stem cells [ 2 6 27 ] These domains were named 'bivalent domains' As investigators soon realized, H3K4me3 not only decorate d gene promoters but also played an active role in transcription activation by serving as a scaffold for transcription factors For example, H3K4me3 can mediate the recruitment of the basal transcription machinery to gene promoters followed by transcription initiation [ 28 ] The mechanism for H3K4me3 dependent transcription initiation was elucidated by showing that TAF3 (TATA box binding protein associated factor 3 ) compon ent of the TFIID general transcription factor, interacts with H3K4me3 mark [ 28 29 ] This interaction was captured by NMR and revealed that the conserved plant homeodomain (PHD) finger o f TAF3 recognize s the specific trimethylation state of H3K4 [ 30 ] Association of TAF3 with H3K4me3 directs the TFIID complex to genes promoter orchestrate assem bly of preinitiation complex (PIC) [ 28 ]


18 More insight into the relationship of H3K4me3 with transcriptional activity has come from the identification of severa l other proteins that recognize and bind H3K4me3 such as CHD1, BPTF, JMJD2A, RAG2 WDR5, and ING [ 31 34 ] For example, the nucleosome remodeling factors, CHD1 a nd BPTF, interact with H3K4me3 and facilitate transcription initiation and elongation. CHD1 and BPTF recognize H3K4me 3, but not naked H3, through their double chromodomains and PHD finger, respectively [ 31 34 ] It was later demonstrated that the interaction of CHD1 with H3K4me3 also facilitates spliceosome anchoring and pre mRNA splicing on active genes suggest ing a novel role for H3K4me3 in pre mRNA maturation [ 32 ] Similar example is the binding of BPTF, a subunit of NURF, ATP dependen t chromatin remodeling complex [ 34 ] The recognition of H3K4me3 and H3K4me2 by WDR5 is an interesting ca se since WDR5 is a core subunit of all COMPASS HMT complexes [ 33 ] The loss of WDR5 leads to a decrease in binding of COMPASS HMTs to H3K4me2 and to a global decrease in H3K4me3. Therefore, it was proposed that WDR5 participate s in reading H3K4me2 prior to writing H3K4me3. One important question remained to be answered is how H3K4me3 is regulated in vivo Although the enzymes catalyzing H3K4me3 are largely known, how these enzymes are targeted to genes promoters is not fully understood. Never theless, it has been shown that H3K4 HMTs interact not only with transcription factors, but also with various histone modifications which affect the positioning of H3K4me3 to gene promoters. For example, transcription couples H2B ubiquitylation and H3 acetylation were sh own to cross talk with H3K4me3 and play an important role in its positioning [ 29 35 36 ] Specific targeting of H3K4me3 HMTs to gene promoters by various


19 transcription factors was also reported. For example, USF1 transcription factor interacts with h SETD1A and target s it to the HOXB4 promoter [ 37 ] Another example is the targeting of h SETD1A catenin [ 38 ] and to p53 target genes by p300 and p53 [ 29 ] The regulation of H3K4me3 is e xtremely complex, but further identification of H3K4me3 regulators will shed light on novel H3K4me3 dependent transcription mechanisms and enable to intervene with its functions. Roles for H3K4me3 i n Cance r The profound role of H3K4me3 in transcription led investigators to explore possible roles of H3K4me3 in different diseases including cancer. Indeed, levels of H3K4me3 were shown to be globally and locally alter ed in different types of canc er and in different developmental stages of the disease and serve as a predictor for recurrence [ 39 46 ] In addition, H3K4me3 was also shown to be involved in cellular plasticity [ 45 47 ] dependent epithelial mesenchymal transition (EMT) of hepatocytes led to a bulk increase in H3K4me3 and dramatic loss of H3K9me2 [ 4 5 ] Hence, H3K4me3, along with other histone modifications, was suggested to serve as potential cancer prognosis marker [ 48 ] Nevertheless, low levels of H3K4me3 were observed in some types of cancer. For example, H3K4me3 levels decrease in late stages of renal cell carci noma [ 40 ] Future studies are needed in order to determine if H3K4me3 mark is clinically relevant for patient assessment Despite limited publications regarding global changes in H3K4me3 levels large number of studies shown position specific alterations in H3K4me3 in cancer cells, especially at promoters of oncogenes and tumor suppressors, such as c Myc and p53 [ 29 49 ] U nderstanding of H3K4me3 regulation in normal and cancer cells will provide opportunities for epigenetic inter ventions for cancer therapeutic.


20 Roles for SETD1A in Development a nd Cancer The function of SETD1 A is not as nearly characterized as tho se of MLL The role of MLL1 gene in leukemia is well established [ 50 51 ] The most prominent mechanism for ML L1 dependent leukemia is frequent chromosomal translocations [ 50 51 ] In contrast to MLLs, the role of SETD1 A in development and cancer is not well understood. Nevertheless, recent work from our lab (unpublished) and other labs demonstrated that the SETD1A knockout (KO) mice are embryoni c lethal at E11.5 [ 52 ] These results present a non redundant role for SETD1 A in early development of mammals Additi onal In vitro studies from our lab showed that ablation of SETD1 A in mouse Embryonic Stem Cells (mESC) le a d to a decrease in expression of mesoderm markers and inhibition of lineage differentiation [ 37 ] providing further explanation for the SETD1A KO phenotype Few studies showed that hSETD1 A impact s genes inv olved in cancer. For example, it regulate the MYC promoter through its binding to BORIS and BAT3 proteins [ 53 ] the TERT promoter through its binding to catenin [ 38 ] and the HOXB4 promoter by its binding to USF1 [ 37 ] The true biological effect of those changes in gene expression is not well established. Another study demonstrated that intraperitoneal injection of hSET D 1A antisense RNA into mice bearing colorectal tumor xenografts, le a d s to a decrease in tumor size [ 54 ] However, those experiments lacked evidence for h SETD1 A KD in the xenografts tumors and failed to explain the mechanism for antisense RNA systemic absorption. The hSETD1A gene is often mutated in human cancers ( ) [ 55 ] Interestingly, hSETD1A mutations are abundant in cancers of the e ndometrium (4.98%), large intestines (4.26 %), and lung s (3.6 %) [ 55 ]


21 In contrast very small number or mutations in hSET D1A were identified in cancers of the liver, p ancreas cervix, p rostate k idney or he matopoietic and lymphoid tissue [ 55 ] The function of hSETD1A mutations is unknown but their tendency to be m issense and their absence from the methyltransferase domain, may suggest retention and even enhancement of the methyltras ferase activity of hSETD1A in cancer. Moreover, g ain in hSETD1A gene copy number was detected in 45% (352/782) of breast cancer cases 18.9 % of CRC cases (92/486), and 18% of kidney cancer cases (54/300), while loss in gene copy was rare [ 55 ] The i nteractions between h SETD1A and catenin are especially interesting [ 38 ] catenin is the master regulator of the Wnt signaling pathway, which has been strongly implicated in CRC and other t ypes of cancer [ 56 61 ] It was further shown that h SETD1 A and H3K4me3 occupancy at the TERT pr omoter decrease in the absence of functional catenin [ 38 ] However, an overall effect of h SETD1 A on the Wnt signaling pathway was not show n at the time Colorectal Cancer a nd t he Wnt signaling Pathway Colorectal cancer is the cancer of the colon and rectum [ 62 ] Conventional treatments, such as chemotherapy produce poor response rates [ 62 64 ] However, detailed mapping of molecular changes initiating or escorting colo n transformation yielded insights which contribute to targeted therapy [ 65 ] One prominent molecular event characterizing most human CRCs is dysregulation of the Wnt signaling pathway [ 57 58 60 65 67 ] Therefore, for decades, the Wnt pathway has been the focus of CRC research.


22 Color ectal Cancer Epidemiology Colorectal cancer is the t hird most common cancer in both men and women, and the third leading cause of cancer related deaths in the USA [ 62 64 ] The American Cancer Society esti mated 136,830 new cases and 50,31 0 deaths of CRC in 2014 in the US A [ 62 ] Fac tors affecting i ncidence and mortality rates include age diet, behaviors such as drinking and smoking, p ersonal history of inflammatory bowel disease ethnicity, and sex [ 62 64 ] Although m ost CRC s are sporadic, a small percentage of patients develop s CRC as a result of genetic predisposition, such as familial adenomatous polyposis (FAP) and hereditary non polyposis CRC (HNPCC) [ 56 62 64 68 69 ] Colorectal cancer is a multi stage disease (Figure 1 2). The progression of the disease is extremely slow and painful symptoms such as rectal bleeding, occur only at late stages. Beca use it is often only in late stages that one become aware of the symptoms, regular screening is critical and early detection of CRC is the key for survival [ 64 ] The recent decrease in CRC incidence and mortality indeed correspond to the increase in number of colonoscopie s performed each year [ 64 ] Determining the choice of treatment is heavily dependent on the cancer stage Polypectomy remain the most common treatment for colorectal cancer (CRC) although it is only effective at early stages of adenomatous polyps [ 70 71 ] Chemotherapy is used at late stages of adenocarcinoma and metastasis [ 63 ] Despite availability of conventional treat ment, call for targeted therapy is urging especially considering the impressive knowledge existing about the mo lecular changes that characterize different stages of the disease [ 63 72 ]


23 The Wnt signaling Pathway The Wnt signa ling pathway is an evolutionarily conserved pathway that is required for prop er development of all metazoans [ 58 66 ] It was named after the secreted Wnt glycoprotein ligands that activate it [ 58 66 73 ] Misregulation of the Wnt pathway is linked to many human diseases, making this pathway one of the most intensely studies signaling pathways in academia and industry in the last 30 years [ 58 73 74 ] One of the earliest events that led to the discovery of Wnt signaling pathway w as the identification of a wingless Drosophila that possessed mutation in the wingless locus [ 73 ] Later, the wingless mouse homolog was discovered and named w nt Interestingly, mutation s in wnt induced mouse mammary tumors [ 75 ] Experiments in Xenopus embryos and Drosophila further indicated that the Wnt pathway plays a critical role in axis formation [ 76 ] These studies implicated the Wnt signaling pathway in both development and carcinogenesis. Because Wnt ligand activ ate s cellular events of different contexts, a conscious distinction between canonical Wnt signaling pathway and non canonical Wnt signaling pathways was made early on Non canonical Wnt signaling pathways affect signaling in cytoplasm, and include the plan ar cell polarity (PCP) pathway and the Wnt/calcium pathway [ 77 78 ] The canonical Wnt signaling pathway begins in the cytoplasm, but convey to the nucleus where it influences gene transcription [ 58 ] In opposed to non canonical Wnt signaling pathways, activation and repression of the canonical Wnt signaling pathway is dependent on the protein levels of its notorious master regulator and primary effector catenin, a mammalian homolog of the Drosophila armadillo [ 58 66 ] Normally, cytoplasmic levels of catenin are kept low [ 58 66 79 ]


24 ( Figure 1 3). This complex includes the Adenomatous Poliposis Coli (APC) tumor suppressor, GSK3 CK1 Axin, and catein [ 58 66 ] GSK3 and CK1 kinases phosphorylate cate nin, leading to its association with TRCP E3 ubiquitin ligase and to its degradation by the 2 6S proteosome [ 58 66 ] (Figure 1 3). Tradit ional activation of the Wnt signaling pathway relies on the association of extracellular Wnt ligand with the extracellular portion of its cognate receptors: G protein coupled receptor Frizzled and the l ow density lipoprotein receptor LRP6 [ 58 66 ] Binding of Wnt ligand to Frizzled and LRP6 results in the recruitment of Axin and its associated and CK1 [ 58 79 ] (Figure 1 3). Phosphorylated LRP6 has a high er affinity for Axin and further increase recruitment of Axin to the cell surface [ 58 66 ] Thi s step creates a positive feedback loop and is known as the Wnt pathway [ 58 66 ] It prev catenin in several ways [ 58 66 ] First, it occupies GSK3 and CK1 which instea d of phosphorylating catenin are now phosphorylating LRP6 [ 58 66 ] to the plasma membrane sequesters it from binding to additional molecules of cytosolic catenin [ 58 66 79 ] And third, o nce the destruction complex associate s with LRP6, the phosphorylated cytoplasmic domain of LRP6 directly inhibit s GSK3 activity, thereby catenin ph osp horylation and subsequent proteo somal degradation [ 58 66 79 ] ( Figure 1 3). catenin translocates from the cytoplasm into the nucleus independently of the classical I mportin/RanGTPase importing system [ 80 ] In the nucleus, catenin associates with the TCF/LEF family of


25 transcription factors and activates gene transcription [ 58 66 ] (Figure 1 3). TCF/LEF catenin to genes that possess a specific consensus sequence [ 58 66 ] Those genes are known as catenin/Wnt target genes [ 58 66 ] One of t he most studied Wnt target gene is the oncogne MYC [ 58 81 ] Activation of the Wnt signaling pathway leads to an increase in the expressi on of MYC and other genes and to an accelerated cellular proliferation [ 58 81 ] Other Wnt target genes, such as SNA I1, SNAI2, TWIST, MMPs VEGF and CDH1 are involved in processes important for cellular plasticity, such as migration, i nvasion, and angiogenesis [ 58 74 ] Both cellu lar proliferation and motility are not only cance rous related functions, but are also essential for embryogenesis and adult tissue maintenance [ 82 84 ] Increased expression of SNAI1, SNAI2, and TWIST upon Wnt signaling activation lead s to a decrease in E cadherin ( CDH1 ) [ 85 ] The repression of E cadherin allows cells to detach from the tight epithelial junctions and migrate [ 85 ] Additional important Wnt target genes are OCT4 NANOG and SOX2 which are three developmental transcription factors supporting embryonic stem cell self renewal [ 86 ] Epigenetic Regulation of Wnt Target Genes While the upstream events leading to activation of Wnt target genes are fairly understood, less is known about how transcription activation occurs Notably, chromatin surrounding the promoters of Wnt target genes is heavily regulated by histone modifications [ 49 87 88 ] S tudies showed that targeted promoters bound by Tcf were H3 hyperacetylation and H3K4 hypertrimethylated, w hile nonresponsive promoter s were devoid of active marks and enriched with the repressive mark H3K27me3 [ 89 ] These observations were supported by studies showing that catenin interacts with h SETD1 A, MLL1/2 and p300, which could explain the H3 acetylation and H3K4 methylation at Wnt


26 responsive promoters [ 38 49 90 91 ] In addition, it was more recently shown that wnt3a can stimulate interactions between the LEF1/TCF complex and SET 8 HMT, resulting in the recruitment of SET8 to Wnt target genes and subsequent H4K20me1 mediated transcription activation [ 92 ] The Role o f Wnt signali ng Pathway i n Colorectal Cancer Constitutive activation of the Wnt signaling pathway is a common feature of solid tumors, particularly colorectal tumors [ 57 58 60 65 67 ] Dysregulation of the Wnt pathway is largely attributed to somatic and genetic mutations in va rious components, most commonly in APC CTNNB1 ( catenin), and AXIN2 [ 58 60 61 ] Notably, g er mline mutations in the APC tumor suppressor are the direct cause for the familial adenomatous polyposis (FAP) syndrome, hereditary form of colon cancer wherein patients develop thousands of polyps in the intestines [ 56 93 ] Similar to the FAP phenotype, heterozygous nonsense mutation in the APC gene resulted in polyps throughout t he intestinal tract in hundred percent of Apc Min mice [ 94 95 ] The majority of these mice do not survive beyond 120 days due to invasive colorectal carcinoma [ 94 95 ] Overall, more than 70% of all CRCs render germline or somatic mut ation s in APC [ 61 96 98 ] catenin destruction nd to activation of the Wnt pathway [ 58 61 79 ] Mutations in CTNNB1 are present in less than 10% of sporadic CRCs [ 58 61 99 100 ] Most mutations in CTNNB1 are concentrated on exon 3 the stability of catenin [ 99 101 ] D eletion of the GSK3 phosphorylati on site in the CTNNB1 locus leads to a phenotype similar to the one observed in Apc Min mice [ 59 94 95 ]


27 The presence of mutations in Wnt signaling components in human CRC and our abil ity to recapitulate disease state by genetic manipulations of wnt components in mouse models are the most prominent evidence s for the importance of Wnt signaling pathway in CRC [ 59 61 94 95 98 ] However, it is important to keep in mind that mutations in Wnt signaling components are also present in other cancers [ 74 ] The downstream effect of constitutively activated Wnt pathway is up regulation in catenin protein levels and overexpressi on of its target onc ogenes, such as c Myc and others involved in self renewal of the colonic epithelium [ 102 106 ] These changes lead to hyperproliferation of u ndifferentiated intestinal cells and formation of intestinal polyps [ 102 106 ] Regulation of Intestinal Stem Cells (ISC) By t he Wnt signaling Pathway The Wnt signaling pathway controls embryonic development and homeostasis of adult self renewing intestinal tissues [ 102 ] The lining of the intestine consist s of severa l unique niches each which is populated with different types of cells (Figure 1 4) [ 102 ] The crypts are invaginations of the epithelium into the unde rlying co nnective tissue. At the bottom of the crypt, intestinal stem cells (ISC) proliferate to replenish sources of undifferentiated cells [ 102 ] These dividing ISCs a lso differentiate and migrate from the very bottom of the crypts upwards to the surface of the lumen ( vill i) to renew the epithelium layer (Figure 1 4) [ 102 ] It has been sugges ted that intestinal polyps rise from excessive proliferation of ISCs [ 102 ] Studies identified the Wnt signaling Notch, and Eph/ephrin pathwa ys as important in s elf r enewal and differentiation of ISC s [ 102 106 ] T he Wnt signaling pathway plays a critical role in maintaining the ISC population at the crypt bottom [ 103


28 106 ] The Wnt pathway is active in the crypt while inactive in the villus and m utations in Wnt componenets lead to irregular distribution of cel l types along the crypt villus axis and results in aberrant crypt foci [ 102 107 ] While hyper activated Wnt pathway results in enlarged crypt s hypo activated W nt pathway leads to disappearance of crypt s and embryonic lethality [ 103 107 ] Expansion of crypts and disappearance of villus upon activation of the Wnt pathway are some of the first transparent phenotypes prior to the development of intestinal polyps [ 102 107 ] Can We W in WNT? Genetic and epigenetic alterations activating the Wnt signaling pathway contribute to the development of variety of human cancers including CRC, pancreatic cancer, lu ng cancer, prostate cancer, leukemia and others [ 57 58 60 65 67 ] Therefore, v arious approaches have been developed as an attempt to target different steps of the Wnt signaling catenin [ 57 74 ] Some approaches focus on upstream components APC, GSK3 Axin2, etc. while others approaches focus on downstream components such as catenin TCF4, and other catenin partners [ 57 74 ] The advantage of inhibiting downstream components versus upstream components is the broader applicability of the therapeutic application to CRC cases regardless of mutations type For example, reactivation catenin is mutated and attenuating wnt secretion by inhibition of porcupine will not be effective in cases of APC mutations On the other hand, direct targeting catenin will be, in theory, effective in most cases of CRC associated with misregulated Wnt pathway.


2 9 Although great effort s were made over the years to inhibit the Wnt signaling pathway, there is currently no specific Wnt pathway inhibitor approved by the FDA [ 57 74 ] However, recently, four drugs targeting the Wnt signaling have reached clinical trials, t wo which are biological inhibitors developed by OncoMed. Both inhibitors focus on targeting association of Wnt ligands with Frizzled receptor s One of them is the OMP 54F28 inhibitor, a fusion protein of the ligand binding domain of Frizzled 8 and an immunoglobulin Fc region [ 108 ] This fusion protein binds and sequesters soluble Wnt ligand s impairing their recognition by putative receptors The other one is the OMP 18R5 inhibitor, a monoclonal antibody which targets the Frizzled receptors [ 109 ] Another compound developed by Novartis is the LGK 974, which is a small molecule targeting porcupine, a critical enzyme facilitating the lipidation of secreted Wnt ligands [ 110 ] Finally, PRI 724 (Prism Pharma), a small molecule inhibitor targeting the catenin and CBP have recently entered phase I clinical trials for the treatment of solid tumors and leukemia [ 111 ] Other approaches currently being developed are targeting the catenin TCF4 interactions, the proteosome or E3/bTrcp ubiquitin ligase, and Axin2 by using compounds that bind and stabilize it [ 57 74 ] On another level, targeting epigenetic regulators of the Wnt signaling pathway could also contribute to attenuation of the Wnt signaling pathway. For example, inhibition of DNA methylation as aberrant CpG methylation of the APC promoter is observed in 30 % of sporadic CRC cases [ 112 ] Furthermore, inhibition of nuclear epigenetic modifi ers, partners of catenin, could strategically be beneficial. Recent studies including ours, have identified various epigenetic regulators that interact with catenin, such as Brg1, hSETD1A, MLL1/2, and P300, and activate Wnt putative


30 promoters [ 38 49 87 90 91 ] For example the PRI 724 inhibitor mentioned above prevents acetylation and activation of Wnt target genes [ 111 ] Finally, t h e lack of true progress in inhibiting the Wnt signaling pathway over the years is most likely due to the limited number of pathway specific components that are also druggable s and due to toxicity p itfalls [ 57 ] However, the success of few preclinical trials testing new Wnt pathway inhibitors, and recent studies identifying new Wnt pathway effectors is encouraging Breast Cancer and Metastasis Breast cancer is the most common cancer in women, and the second leading cause of cancer related deaths among women in the USA [ 113 114 ] The American Can cer Society esti mated that 232,670 new cases of invasive breast cancer and 62,57 0 new cases of carcinoma in situ breast cancer will be diagnosed in women in the USA, in 2014. In addition, 2,36 0 new cases of breast cancer were estimat ed among men in the USA in 2014 Approximately 40,00 0 women and 430 men are expected t o die from breast cancer in 2014 alone, in the USA [ 113 ] Breast cancer related deaths had slowly increased from1975 to 1990 at a 0.4% annual increase rate [ 113 ] This increase reflected changes in life span, delayed childbearing, reduced number of children, increased use of hormones and increased obesity [ 113 ] Death rates of breast cancer started to decline in 2001 du e to t he d ecreased use of menopausal hormones [ 113 ] Survival rates of breast cancer are dependent on the location of the disease [ 113 ] According to the National Cancer Institute, while there is 99 84% cha nce of five year relative survival for localized and regional disease, there is only 24% chance of five year relative survival for metastatic breast cancer. Therefore, early detection of breast cancer and prevention of meta stasis


31 is critical for survival Understanding of the reasons for breast cancer progression is crucial for developments of new interventions. Like most cancers, breast cancer could be a result of genetic or somatic mutations It is estimated that 5% to 10% of breast cancers are triggere d by genetic mutations including mutations in BRCA1 and BRCA2 TP53 CHECK2 PTEN SDHB SDHD KLLN and other genes [ 115 ] E pigenetic is important for conveying genomic selection and adaptation and for sustention of a phenotype [ 10 12 ] Understanding o f the epigenetic changes that contribute to breast carcinogenesis w ill not only contribute to better di agnosis and prognosis but also allow for new research and development of therapeutic revenues. Molecular and Biologic al Basics of M etastasis Metastati c cancer is defined as cancer that has spread from the primary tumor site to another location to form secondary cancer [ 116 117 ] Virtua lly all cancers can metastasize M etastasis prone tissues are the bones, liver lung s and brain [ 117 ] Tumor cells which populate and flourish in the primary or secondary tissue lead to organ dysfunction and death Still, m e tastasis lies in the heart of therapeutic failure and mortality [ 116 117 ] Researchers have been trying to ta rget metastasis by blocking cer tain functions important for dissemination of the pr imary tumor cells such as migration, invasion a nd angiogenesis. These cellular processes play a critical role in enabling cells to detach from the primary tumor, disseminat e through the blood stream and establish new colonies [ 83 85 116 117 ] In order to metastasize, a cell need s to successfully complete serial of events in a temporal and spatial manner [ 83 85 116 117 ] Both molecular and morphological changes are r equired for such events to occur [ 83 85 116 117 ] First, primary tumor


32 cell s need to detach from the primary tumor and invade into a nearby tissue (local invasion) (Fig ure 1 5) [ 83 ] In order to disseminate to a distant location cells wil l migrate towards a nearby blood vessel and invade into the bloodstream (intravasation) (Fig ure 1 5) [ 83 ] Circulating cells that survived in the blood will eventually reach small capillaries in which the bl ood stream condit ions enable their movement arrest (Figure 1 5) [ 83 ] Cells will then invade through the endothelial wall and migrate into the surrounding tissu e (extravasation) (Figure 1 5) [ 83 ] In order to form a secondary tumor cells nee d to be capable of proliferating in the new environment (Figure 1 5) [ 83 ] Once the secondary tumor reaches a certain size it becomes dependent on increased blood supply for oxygen and nutrie nts [ 118 ] Therefore it will be crucial for the tumor to stimulate the growth of new blood vessels to obtain sufficient blood supply (angiogenesis) [ 118 119 ] The cellular events enabling the dissemination of cells from t heir origin to other locations throughout the body are regulated by molecular events [ 83 85 116 117 ] For example, detachment of cells from the primary tumor is highly dependent on down regulation of factors controlling cell cell contact, such as E cadherin [ 83 ] Inv asion is dependent on up regulation of secreted proteases, such as M atrix metalloproteinases (MMPs ) and Cathepsins for degradation of the extracellular matrix [ 120 121 ] Directional cell migration is dependent on c ell polarity which is regulated by dynamic filaments [ 122 ] The formation of new blood vessels (angiogenesis) is regulated b y angiogenic proteins, such as v ascular endo thelial growth factors (VEGFs), Angiopoietins, and growth factors such as Fibroblast Growth F actors (FGFs) [ 118 119 ] T he molecular and cellular change s characterizing metastatic cells are not exquisitely


33 regulated by the cell itself, but greatly dependent on the cellular envir onment in which the tumor resides in [ 123 ] The tumor microenvironment plays a critical role in providing cancer c ells with extracellular signals of grow th, angiogenesis, invasion, and migration [ 123 ] Epithelial Mesenchymal Transition (EMT) One of the initial changes a tumor cell has to go through in order to migrate is morphological [ 82 8 5 ] There two morphologically different cell types which can interconvert: epithelial and mesenchymal [ 82 85 ] Epithelial mesenchymal transition (EMT) i s a highl y conserved proce ss which was first recognized to be important in early develop ment and cell differentiation [ 82 84 ] That is because mesenchymal features are necessary for migrat ion, a function critical during development [ 82 84 ] Only later it was discovered to be a fundamental process in carcinoma progression and metastasis [ 82 85 ] While epithelial cells are rounded form adh erent and tight junctions and exhibit apical basal polarity, mesenchymal cells are elongated and lack polarization [ 82 84 ] The molecular changes leading to alteration in morphogenesis are partially understood [ 83 85 ] The l oss of E cadherin expression is one of the major changes occurring during EMT [ 83 85 ] E cadherin fun ctions in establishing stable cell cell conta ct through formation of adheren junc catenin and alpha catenin [ 83 85 ] Therefore the loss of E cadherin allows cells to detach from one another and directly influence s the acquisition of the mesenchymal phenotyp [ 83 85 ] The Wnt signaling pathway also controls EMT since it tightly controls the expression of E cadherin through regulation of Snail, Slug, and Twist, negative regulators of E cadherin [ 84 124 125 ] Activation of the Wnt signaling pathway not only repre sses the expression of E cadherin


34 but also sequester s catenin from adherens junctions thereby attenuating cell cell contact [ 58 66 ] Another important positive regulator of EMT is which is also transcriptionally controlled by the Wnt signaling pathway [ 83 ] This cytokine is often used in vitro to induce EMT. Role for Epigenetic in EMT Tumor progression is dependent of genet ic and epigenetic alterations To effectively diagnose and treat advanced disease it is critical to improve our knowledge of genes that regulate metastasis and to identify novel biomarkers G lobal changes in epigenetic marks are observed during EMT, but th e function of such global changes and their source is largely unknown [ 126 ] In contrast modulation s of epigenetics at specific gene loci such as the E cadherin promoter have been studied for dec ades. For example, DNA methylation at the E cadherin promoter was shown to repress transcription of E cadherin and increase EMT [ 85 127 ] In fact, it is extremely difficult to reactivate the E cadherin promoter by molecular manipulations after it has already gone through CpG island methylation This demonstrates the important role of epigenetic modifications in transcription regulation of master regulators of EM T. Another example is the regulation of mesenchymal and epithelial genes such as E cadherin, HIF1 vimentin and N cadherin by HDACs and SIRT1 histone de acetylases [ 85 127 ] to globally increase active marks, such as H3K4me3 and H3K36me3, and decrease repressi ve marks, such as H3K9me2 [ 45 ] Notably, it was recently demonstrated that H3K4 demethylase LSD1 regulates EMT and metastasis [ 45 128 129 ] Ablation of LSD1 led to an increase in cell


35 invasion and migration in vitro and lu ng metastasis in viv o [ 128 129 ] These studies suggest the involvement of H3K4 methylation in stimulation of EMT. Understanding of the epigenetic mechanisms related to metastasis is important and could offer novel approach es for innovative diagnosis and treatment of cancer patients. Future investigations and better understanding of the role of histone modifications du ring progression of cancer could greatly improve our therapeutic strategies.


36 Figure 1 1. The COMPASS /Trithorax family of H3K4 methyltrasnferases [ Adapted from Shilatifard, A. 2012. Annual Review of B iochemistry Page 71 Figure 2 ] [ 18 ]


37 Figure 1 2. Multi stage progression model of colorectal cancer. In situ: cancers that have not yet invaded the wall of th e colon or rectum. Local: Cancers that have grown into the wall of the colon and rectum, but have not extended through the wall to invade nearby tissues. Regional: Cancers that have spread through the wall of the colon or rectum and have invaded nearby tis sue, or that have spread to nearby lymph nodes. Distant: Cancers that have spread to other organs [Adapted from The American Cancer Society, Colorectal Cancer Facts & Figures 2011 2013 page 2, Figure 2]


38 Figure 1 3. Schematic representation of can onical Wnt signaling pathway. In the absence of wnt catenin is bound by a des truction complex and targeted for phosphorylation, ubiqitination, and degradation by the 26S proteosome. Binding of Wnt ligand to a Frizzled/LRP 5/6 receptor complex catenin from the destruction complex, catenin translocates into the nucleus where it interacts with TCF/LEF transcription factors and activates transcription of target genes.


39 Figure 1 4. Cellular and structural organization at the inte stinal crypt Intestinal stem cells lie near the crypt base and differentiate upwards. Above the stem cells are transit amplifying cells (dividing progenitors, some of them already partially differentiated); and above these, in the neck of the crypt and on the villus, lie post mitotic differentiated cells (absorptive cells, goblet cells and enteroendocrine cells. [Adapted from Crosnier C. 2006. Nature reviews. Genetics, Page 350, Figure 1a] [ 102 ]


40 Figure 1 5. Schematic representation of the metastasis process. Cells which have gone through EMT can intravasate into lymph or blood vessels, and circulate until they extravasate through the endothelial layer and esta blish a secondary tumor locally (micrometastasis) or at distant organs (macrometastasis). [ Adapted from Thiery JP 2002. Nat Rev Cancer Page 445 Figure 2 ] [ 83 ]


41 CHAPTER 2 MATERIALS AND METHODS Cell Culture a nd R eagents H uman cell lines were purchased from American Type Cul ture Collection ( ATCC ) u nless otherwise indicated All cell lines were cultured at 37C humidified atmosphere of 5% carbon dioxide (CO 2 ) in air HEK293FT, Pheonix, HCT116, SW48, C2A, MDA MB 231, MCF7, BT 549, and SUM 159 cell lines were maintained in Dulbe Eagle medium ( DMEM ) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin streptomycin (Invitrogen). RKO and HT29 cells were maintained in MEM and RPMI respectively, supplemented with 10% FBS and 1% penicillin streptomycin FHs in t 74 cells, kindly provided by Dr. Maria Zajac Kaye (University of Florida) were cul tured in DMEM supplemented with 10% FBS, 1% penicillin streptomycin 10 g/mL insulin (Sigma) and 30 ng /mL epidermal growth factor ( EGF ; Sigma) The human mammary epithelial cell line s MCF10A and HMEC w ere a kind gift from Dr. Jianrong Lu and Dr. Kevin Brown (University of Florida) respectively, and were cultured in DMEM F 12 medium (Cellgro) supplemente d with 5% horse serum (Sigma), 3 0 ng/mL EGF 10 g/mL insulin, and 0.5 g/mL hydrocortisone (Sigma). Mouse Embryonic Fibroblast s (MEF s 3A CRL purchased from ATTC and c ultured for 4 days in DMEM supplemented with 10% FBS and 1% penicillin streptomycin according to the manufacturer instructions Medium was then collected and replaced with fresh medium for additional 3 days of culturing. Conditioned medium from 4 days of culturing was mixed with the medium from the additional 3 days of culturing at 1:1 ratio to obtain conditioned medium (WCM)


42 ShRNA mediated K nock Down Viruses carrying specific shRNAs were generated by co transfecting shRNA constructs along with viral packaging plasmids PMD2 G (e nvelope p lasmid ; Addgene ) and PsPax2 ( gag/pol packaging plasmid ; Ad dgene ) for lentivirus es, or packaging plasmids gag pol and envelope protein ( Oligoengine) for standard retroviruses For virus propagation, p lasmids were trans f ected into HEK293FT or Pheonix cells using the Calcium Phosphate Transfection method (Promega). Shortly, gag/pol + 2.5 of Env), 55 In a separate tube, was combined with p yrophosphate The CaCl2 DNA solution was added dropwise into the HBS solution while introducing air bubbles through the HBS solution. The solutions mix was incubate d at room temperature for 2 0 minutes and then added drop wise to the host cells (1 2 x10 6 cells / 10 cm dish ). The medium was replaced with 8 mL fresh medium upon 24 hours After additional 48 hours, media containing the virus particles was filtered and collected. To generate stable knockdown (KD) 1x10 6 cells w ere infected with 1 2 mL of the recombinant viruses in the presence of 4 Medium was replaced with fresh puromycin (Calbiochem, 540222) upon 48 hours for stable selection. All stable KD cells were maintained in media containing 1 puromycin. Constructs used for h SET D 1 A KD in HCT116 and C2A cells were generated by subcloning shRNA oligonucleotides into pSuper.retro.puro vector following the Constructs used for hSETD1A KD in other cell lines utilized l entiviral TRC vector harboring shRNAs targeting the h SETD1A gene


43 (Clone ID: TRCN0000152242) or CTNNB1 gene (TRCN0000003846) obtained from the UF Health Cancer Center shRNA library (Thermo Scientific). All shRNA sequences are listed in T able 2 1 Antibodies hSETD1A (A300 289A), Ash2L (A300 489A), MLL1 (A300 086A), TAF3 (A302 359A), and RbBP5 (A300 109A) antibodies were purchased from Bethyl Laboratories. catenin (C2206 ) actin (AC 15), tubulin (T6199) antibodies from Sigma. TCF4 (sc 8631) and MYC (sc 40) antibodies fro m Santa Cruz Biotechnology. WDR5 (Ab56919), histone H3 (Ab1791), and H3K4me1 (Ab8895) antibodies from Abcam. H3K4me3 (MC315), H3K9me2 (05 768), H3K4me2 (07 030), and H3K27me3 (07 449) antibodies from Millipore. Snail antibodies (L70G2) from Cell S ignaling. E cadherin antibodies (610181) from BD Bioscience. Western I mmunoblotting Protein samples were prepared as followed: cells (1x10 6 ) were collected by centrifugation at 1000 rpm for 3 minutes Medium was aspirated and cell pellet was resuspend in ice cold Phosphate Buffered Saline ( PBS ) Cells were collected again by centrifugation at 1000 rpm for 3 minutes followed by a spiration of the PBS. Cells were lysed by r esuspending cell pellet in RIPA buffer [ 140 mM NaCl, 0.5% Triton X 100, 1 % sodium deoxychol ate, 0.025% sodium azide, 10 mM Tris, pH 8.0 ] containing 1 mM PMSF (Sigma, P7626) 1 mM DTT, 1 g/mL P epstatin A (Sigma, P5318) 1 g/mL Leupe ptin (Sigma, L2884) and 1 Aprotinin (Sigma, A1153) Mixture was vortexed and sonicated ( 10 short bits ) Sample was denatured by adding l aemmli buffer and boiling at 95 C for 8 minutes.


44 Protein sample was separated using sodium dodecyl sulfate polyacry lamide gel electrophoresis ( SDS PAGE) Denatured samples (40 80 g) were loaded onto polyacrylamide gel and ran at 100 120 V. Proteins were transferred to PVDF membranes at 350 milliamps for 3 hours using the Wet/Tank Blotting Systems (BioRad) Membrane was blocked with 5% nonfat milk in TBS T for 1 hour at room temperature and then incubated with primary antibody at 4 C o ver night. Membrane was washed with 5% nonfat milk in TBS T for 1 hour at room temperature during which milk was replaced every 15 minutes. Membrane was then i ncubation with species specific horseradish peroxidase conjugated secondary antibodies in 5% nonfat milk in TBS T for 1 h our at room temperature followed by 1 hour wash with TBS T at room temperature during which TBST was repl aced every 15 minutes Immunoreactive bands were visualized by using enhanced chemiluminescence (ECL; Pierce ) and autoradiography. tubulin or a ctin were used as a loading control. Co immunoprecipitation Nuclear extract was prepared as followed: cells (1 2x10 8 ) were collected and washed twice with PBS Cell pellet was resuspended in 5 pcv of cold hypotonic buffer (10 mM HEPES PH 7 .9, 1.5 mM MgCl 2 10mM KCl ] supplemented with protease inhibitors followed by centrifugation (3,000 rpm for 5 minutes at 4C ). Cell pellet was resuspended in 3 pcv hypotonic buffer and incubated on ice for 10 minutes. Cells were then homogenized with 10 18 up and down strokes using type B pestle and centrifuged (4,000 rpm for 15 minutes at 4C). Nuclei pellet was resuspended in pnv of low salt buffer [20 mM HEPES PH 7.9, 25% glycerol, 1.5 mM MgCl 2 0.02 M KCl 0.2 mM EDTA]


45 supplemented with protease inhibi tors, followed by dropwise addition of pnv high salt buffer [20 mM HEPES PH 7.9, 25% glycerol, 1.5 mM MgCl 2 1.2 M KCl 0.2 mM EDTA] supplemented with protease inhibitors under continuous gentle mixing for 30 minutes at 4C. Nuclear extract was centrifuge d (20 minutes at maximum speed at 4C) to remove nuclei debris and supernatant was dialyzed overnight against 50 volume of dialysis buffer [20 mM HEPES PH 7.9, 20% glycerol, 100 m M KCl 0.2 mM EDTA] supplemented with protease inhibitors. Nuclear extract (0 .7 mg ) was precleared with 5 g of imm unoglobulin G (IgG) and Dynabeads Protein A or G. Precleared nuclear extract was immunoprecipitated with antibodies a gainst catenin (C2206) or components of the hSETD1A complex overnight at 4C Beads were thoroughl y washed with RIPA and precipitants were eluted in l aemmli buffer by boiling. Precipitated protein samples were subjected to SDS PAGE, and analyzed by Western blot (WB) analysis as described above Immunofluorescence Immunofluorescence staining was carrie d out by the Molecular Pathology and Immunology Core at the University of Florida T issue samples were formalin fixed and subsequently paraffin embedded. Five micron sections were sequentially deparaffinized and rehydrated through graded ethanol Optimal s taining required 25 minutes of heat antigen retrieval in 10 mM Citrate buffer pH6.0. Sections were blocked with 2% horse catenin antibodies (C2206 sigma) diluted at 1:1200 overnight at 4C. Slides were washed with 1X PBS twice for 5 minutes at RTC, incubated in donkey anti rabbit alexafluor 488 diluted at 1:500 for 60 minutes at RTC in dark and mounted in Vectorshield with DAPI. Images were obtained using a


46 Leica DM2500 microscope equipped with an Opt ronics camera using magnifier software. Immunohistochemistry Immunohistochemistry staining was carried out by the Molecular Pathology and Immunology Core at University of Florida. Briefly, 5 microns serial sections were de paraffinized and incubated with 3% H 2 O 2 /methanol to block endoge nous peroxidase activity. S ections were treated by Citra (Biogenex, Fremont, CA) at 95 C for 30 minutes followed by blocking with Sniper (Biocare Medical, Walnut Creek, CA) to reduce non specific background staining. Sectio ns were incubated with rabbit anti human hSET D 1 A ( IHC 00171 ) at room temperature for 1 hour and then with Mach2 Gt x Rb HRP polymer (Biocare Medical, Walnut Creek, CA) for 30 minutes. Staining was visualized with DAB chromagen (Vector Laboratories, Burling ame, CA) and hematoxylin counterstain. Chromatin I mmunoprecipitation (ChIP) Cells (1x10 8 ) were collected in 10 mL medium and cross linked at room temperature with 1.5 mM Ethylene glycol sulfosuccinimidylsuccinate (EGS; Sigma, E3257) (30 minutes rotation) a nd 1% Formaldehyde (FisherSci, BP531) (10 minutes rotation). Cross linking reaction was quenched by the addition of 0.125 M glycine ( 5 minutes rotation ) Cells were pelleted (780 xg for 10 minutes) and washed twice with cold PBS. Chromatin from frozen t iss ue samples was prepared (30 g tissue /ChIP) by slicing the tissue using a razor blade cross linking with formaldehyde and EGS as described above homogenization and filtration Cells were then lysed with 3 mLs cell lysis buffer [10mM Tris HCl, pH 8.0, 3mM Mg Cl 2 0.4% N P 40] contain ing protease inhibitors, followed by centrifugation (5000 rpm for 5 minutes). Resuspension and centrifugation was repeated up to 3 times to completely lyse the cells. Nuclei were lysed with 1 mL of SDS


47 lysis buffer [50 mM Tris HCl, pH 8, 10 mM EDTA ,1% SDS] containing protease inhibitors and incubat ed on ice for 10 minutes Non SDS lysis buffer (3 mLs ) [50 mM Tris HCl, pH 8, 10 mM EDTA] containing protease inhibitors was added to th e nuclei lysate and the mixture was incubat ed on ice for additional 5 minutes C hromatin was sonicated using Bioruptor TM UCD 200 (Diagenode) for 15 30 minutes to obtain DNA fragment s of 200 500 bp Cellular debris was removed by centrifugation for 30 minutes at maximum speed at 4 C and supernatant chromatin was subjected to IP. Ch romatin immunoprecipitation was performed as followed: chromatin was precle ared by the addition of species specific IgG (5 g/ ChIP) for 2 hours at 4 C with rotation followed by the addition of p rotein A or G dynabeads Precleared chromatin was precipitated overnight at 4 C with 10 g IgG control or other antibodies. Chromatin was centrifuged for 30 minutes at maximum speed an d supernatant was further incubated with p rotein A or G dynabeads for 2 hours at 4 C with rotation Beads were then extensively washed and DNA was eluted from the beads with sodium bicarbonate solution [0.0084 g/mL NaHCO3 1% SDS] followed by Phenol chloro form extraction and ethanol precipitation Specific DNA loci were quantified by real time qPCR. R elative fold enrichment of DNA fragments was determined by normalizing the quantification cycle value ( Cq) of each IP with the Cq value of the input chromatin. The ChIP primers used are listed in Table 2 3 Real Time PCR Total RNA was isolated from cells using the RNeasy Mini Kit (QIAGEN) according to the manufacturer's instructions Total RNA was isolated from tissues by trizol extraction (Ambion ). Shortly, tis sues were homogenized in 1 mL trizol Chloroform (200 L) was added to each sample and mixture was vortexed incubated at room


48 temperature for 3 minutes and centrifuged (12,000 x g for 15 minutes at 4 C ) .The upper phase was collected and mixed with 500 L 100% isopropanol. Mixture was incubated at room temperature for 10 minutes followed by centrifugation (12,000 x g for 10 minutes at 4 C ). Supernatant was discarded and DNA pellet was washed with 75% ethanol followed by centrifugation (7,500 x g for 5 min utes). Supernatant was discarded and DNA pellet was dissolved in water. cDNA was synthesized from 2 g mRNA using the Superscript II Reverse Transcriptase kit (Invitrogen) according to the manufacturer's instructions cDNA was diluted 10 fold and Real time PCR was performed on cDNA templates with a CFX real time P CR detection system (Bio Rad), using SYBR Green [10 L SYBR Green, 4.4 L H 2 O, 0.3 primer mix, 5 L cDNA ]. PCR amplifications were performed in triplicate GAPDH or actin or were used as an internal control. Primers sequences are listed in Table 2 2 Growth A ssays Proliferation assay was pe rformed by seeding cells in 10 mL medium at a density of 2 x 10 5 cells in 10 cm culture dishes. C ell number was determined by cell viability count every other day. Colony formation assays were performed by seed ing cells in 6 cm culture dishes at a density of 1 x 10 3 cells per plate. After 2 3 weeks, c olonies were stained with crystal violet and the number of colonies was determined S oft agar colony formation assays were performed by seeding 1 x 10 4 cells in culture media containing 0.7% agarose medium on t op of a base layer containing 0.5% agarose in medium The number of c olonies (>50 cells) in soft agar was determined after 3 weeks.


49 Migration A ssay shRNA transfected MDA MB 231 cells were harvested, resuspended ( 5 x 10 4 ) in 500 L complete DMEM medium and placed o nto inserts containing 8 m pore membranes (B D Biosciences ) Inserts were placed inside wells (24 well plate) containing 700 uL of complete DMEM medium, serving as an attractant. After 24 h, nonmigrating cells remaining on top of the membranes we re removed with a cotton swabs. C ells that were able to migrate through the membrane onto the underside surface of the membrane were fixed and stained with crystal violet and photos of the entire field were taken Each experiment was carried out in 3 5 re p licates and repeated at least twice. Invasion A ssay This assay was performed as described for the migration assay, with the following modifications. MDA MB 231 c ells (1x10 4 ) were suspended in 500 L serum free medium and seeded onto ba sement membrane matrix coated invasion inserts (BD Biosciences). The insert is covered with a membrane containing 8 coated with a basement membrane matrix (Matrigel). Cells were fixed a fter 17 h ours an d the number of invadin g cells on the entire field was counted by light microscopy. Each experiment was carried out in 3 5 re plicates and repeated at least twice. Wound Healing Assay MDA MB 231 or BT 549 c ells were cultured to confluence or near confluence (>90%) in 6 cm tissue culture dish overnight. Using a sterile 1000 pipet tip, two perpendicular scratch es were applied to the cells. Cells were then r inse d with 1X PBS which was then replaced with complete fresh DMEM medium. Pictures of the scratch at different point marks were taken and retaken after 16 72 hours.


50 Cel l Cycle A nalysis HCT116 c ells were grown to 60% confluence, trypsinized, and washed twice with cold PBS. Cells ( 1 2 x 10 6 ) were collected in in a 15 ml polypropylene tube and Cells were further fixed overnight at 4 C in 1 m L of cold ( 20 C ) 70% ethanol which was added to the cells dropwise while vortexing Cells were then Cell cycle was analyzed wit h a Becton Dickinson cell sorter. Microarray A nalysis Ten micrograms of RNA were reverse transcribed to cDNA using the Applied Biosystems High Capacity cDNA Reverse Transcription Kit according to the manufacturer's instructions (Applied Biosystems). cDNA p roducts were heated to 95 o C for 1min and treat ed with 100ng RNase A for 30min at 37 o C and then purified using the Qiagen PCR purification kit according to the manufacturer's instructions. NimbleGen Human Gene Expression array was purchased from Roche Appl ied Sciences. cDNA samples were labeled, Cy3 hybridized, and processed at the FSU NimbleGen Microarray Facility at Florida State University. The well characterized Wnt target genes analyzed were selected according to the Stanford University Wnt Home Page ( ab/cgi bin/wnt/ ) and according to commercially available SABioscienses Wnt signaling Targets PCR array. The expres sion data were analyzed using the bioconductor packages Oligo [ 130 ] and Limma [ 131 ] Expression was background cor rected and quantile normalized with RMA (robust multi array) analysis, and the empirical Bayes approach in Limma package was applied to call differentially expressed genes. Genes with P value (FDR adjusted) less than 0.005


51 were defi ned as differen tially expressed. The P value of over representation in the Wnt signaling pathway was calculated by a hypergeometric test. The CEO accession number for the microarray data set reported in this article is GSE52230. Colorec tal a nd B reast Patient S amples were obtained from the Cooperative Human Tissue Network (CHTN) in the University of Alabama at Birmingham (IRB No. 031078 and 010294 University of Alabama). All samples were deidentified. The use of all huma n tissues for this study was approved by the Institutional Review Board (IRB 01) of the University of Florida (Non Human Exempt IRB approval No. 24 2012). All procedures were performed according to the regulations and guidelines of the approved protocol. M ouse Xenograft M odel The scramble (Scr) control or shSETD1A (C7) expressed HCT116 cells (1 x 10 6 ) were resuspended in 150 L 50% Matrigel ( BD Biosciences ) and 1X PBS in 1:1 ratio Cell suspension was injected subcutaneously using an insulin syringe into the flank of 5 to 6 week old female athymic nude mice Tumors were harvested 16 days post injection and tumor volume was determined using the formula L 2 W ( /6), L=shortest diameter, W= longest diameter. Dissected tumors were homogenized in RIPA lysis buffer ( Chem Cruz) supplemented with protease inhibitors, tumor lysate was sonicated, and 80 g protein samples were subjected to WB analysis as described above. Mice were purchased from Harlan Labs and housed in the Cancer and Genetics Research Center at the Uni vers ity of Florida Animal handling and procedures were approved and performed according to the guidelines of the Institutional Animal Care and Use Committee of the University of Florida


52 In vivo M etastasis MDA MB 231 cells that have been infected with len tiviruses carrying shRNA targeting firefly luciferase or hSETD1A were injected into the tail vein (1 x 10 6 cells in 200 L PBS ) of 5 6 week old female athymic nude mice ( Harlan Laboratories). After 7 8 weeks, mice were sacrificed and lungs were dissected Lungs were immediately fixed in bouin's solution for 24 hours and washed thoroughly with 70% ethanol. Lungs nodules were counted under a light microscopy Mice were purchased from Harlan Labs and housed in the Cancer and Genetics Research Center at the Uni vers ity of Florida Animal handling and procedures were approved and performed according to the guidelines of the Institutional Animal Care and Use Committee of the University of Florida Statistical A nalysis Quantitative r esults were expressed as means SD Two tailed s test was used for comparison between two different conditions *, p < 0.05; **, p <0.01, ***, P<0.001.






55 CHAPTER 3 HUMAN SETD1A AND H3K4ME 3 ARE UP REGULATED IN COLORECTAL CANCER AND CONTROL CELLU LAR GROWTH Introductory Remarks Colorectal cancer is the t hird most common cancer in both men and women, and the third leading cause of cancer related deaths in men and women, in the USA [ 62 64 ] The American Cancer Society esti mated 136,830 new cases and 50,31 0 deaths of CRC in 2014 in the US A [ 62 ] Despite availability of conventional treatment, calls for targeted therapies are urging. Epige netic modulation s are prevalent in CRC including alteration in levels of DNA methylation and histone modifications [ 39 132 133 ] It has been shown that genome wide enhancer/promoter associated H3K4 methylation pro fi les are signi fi cantly altered in CRC compared with the paired adjacent normal mucosa [ 39 133 ] The data suggest that speci fi c H3K4 methyltransferases may play an important role in the development of colorectal tumors Human SETD1A and SETD1B are predominantly associated with transcribed regions through their interactions with Ser 5 phosphorylated RNA pol II [ 18 ] These HMTs are ma inly responsible for H3K4me3 and globally affect gene expression profile [ 18 21 24 ] Taken together hSETD1A and hSETD1B are potential candidates for regulating bulk H3K4 methylation and cellular growth in CRC. Results hSETD1A and H3K4me3 are U p regulated in H uman C olorectal C ance r To gai n insights into the role of trithorax HMTs in CRC we fi rst examined the expression levels of hSETD1A, hSETD1B, and MLL1 in human CRC. A lthough MLL1 levels remained unchanged, hSETD1A levels markedly increased in all CRC cell lines


56 tested c ompared to the human normal intestinal cell line FHs Int 74 (Fig ure 3 1C ). Moreover, the up regulation of hSETD1A in CRC cells was accompanied by a global increase in H3K4me3 levels ( Figure 3 1C ). In addition the mRNA levels of hSETD1A and hSETD1B were up regulat ed in human CRC specimens compared with adjacent paired normal mucosa collected from 24 patients ( Figure 3 1A and Figure 3 1B ). More than 62% the CRC specimens showed 2 fold or more increase in hSETD1A transcript levels ( 15/24 p atients; Figure 3 1A ), where as only 26% of patients showed 2 fold or more increase in hSETD1B expression (6/23 patients; Figure 3 1B ). Immunohistochemistry staining further con fi rmed that hSETD1A was aberrantly expressed in colorectal tumors in comparison with adjacent normal mucosa (Figure 3 1D ). Importantly, hSETD1A expression also positively correlated with a global increase in H3K4me3 levels in these colorectal tumors ( Figure 3 1D bottom ). More interestingly, t he distribut ion pattern of hSETD1A in normal mucosa indicated that hSE TD1A was highly expressed in the mouse and human intestinal crypt bottom s but under expressed at the top of the crypts ( Figure 3 1E ). The intestinal crypt bottom is populated with highly proliferative and poorly differentiated intestinal stem cells while the crypt top (or villus) is occupied by differentiated enterocytes [ 102 ] Several studies argue that early molecular changes in crypt bottom s drive the initiat ion of colorectal polyps [ 104 105 107 ] T he fact that hSETD1A levels were elevated in human CRC and in normal intestinal crypts implies that hSETD1A activity may be required for intestinal cell proliferation. hSETD1A Impacts Proliferation of Colorectal C ancer C ells To determine the effect of hSETD1A on cellu lar proliferation hSETD1A was stably silenced in the CRC cell line s HCT116 SW48, C2A, and RKO using retrovirus or


57 lentivirus harboring shRNA speci fi c for the hSETD1A gene ( Figure 3 2 and Figure 3 3A ). KD of hSETD1A in HCT116 cells led t o a speci fi c decre ase in global H3K4me3 modi fi cation but did not enact on other histone modi fi cations, such as H3K4me1, H3K4me2, H3K27me3, and H3K9me2 ( Figure 3 2 ). Most importantly, the KD of hSETD1A signi fi cantly reduced the proliferation rate s of HCT116, SW48, C2A, and RKO cells ( Figure 3 3A ). In addition, hSETD1A KD HCT116 and SW48 cells displayed smaller colon ies compared with the scramble control ( Figure 3 3B ). When grown on soft agar, the h SETD1A KD HCT116 cells displayed fewer and smaller colonies with approximately 45% less colonies in the hSETD1A KD cell pool, and average d 70% less colonies in the two individual hSETD1A KD clones compared with the scramble control ( Figure 3 3C ). Furthermo re, sub G1 phase population or increase in cell death were not observed in hS ETD1A KD HCT116 cells using fluorescence activ ated cells sorting (FACS) analysis ( Figure 3 4A ) or T rypan b lue exclusion dye, respectively ( Figure 3 4B) These re sults exclude the possibility that apoptosis is accounted for the decrease in cellular prolife ration in the hSETD1A KD cells Cell cycle analysis rather revealed reduction in S phase population ( Figure 3 4 A ). Taken together we conclude d that hSETD1A and its associated H3K4me3 play an important role in regulating the proliferation of CRC cells, and were prompted to investigate the mechanism. Closing Remarks To investigate dys regulation of H3K4 methylation in human CRC we detected levels of HMTs such as, hSETD1A, hSETD1B, and MLL1 in human colorectal tumors and cell lines by qRT PCR and WB analyses hSETD1A was significantly up regulated in CRC and its depletion resulted in reduced global H3K4me3 and decreased cellular


58 growth, but did not affect cell death. These resu lts imply that increased levels of hSETD1A in CRC could enhance cellular growth.


59 Figure 3 1. hSETD1A and H3K4me3 are up regulated in human colorectal cancer. (A and B) hSETD1A (A) and hSETD1B (B) mRNA levels were up regulated in human colorectal tumors compared with adjacent paired normal mucosa as detected by qRT PCR. Shown are the me an SD of three replicates. *P < 0.05 by s tudent t test. ( C ) hSETD1A and H3K4me3 were up regulated in various human CRC cell lines compared with normal intestinal epi thelial cell line FHs in t 74 as detected by WB analysis (D) I mmunohistochemical (t op) and WB (bottom) analyses revealed an increase in hSETD1A and H3K4me3 levels in human colorectal tumor compared with adjacent n ormal mucosa from patient #2. T = tumor, N = normal. (E) hSETD1A was highly expressed in normal intestinal crypts from mouse (left) a nd human (right). Red arrows indicate crypts and black arrows indicate crypt tops.


60 Figure 3 1. Continued


61 Figure 3 2 hSETD1A exclusively affect s global H3K4me3. HCT116 cells were infected with a retrovirus harboring shRNA speci fi c for hSETD1A o r with scramble control. Two individual clones, C7 and C19, were selected from the hSETD1A KD cell pool. hSETD1A and histone modi fi cation levels were analyzed by WB analysis.


62 Figure 3 3 hSETD1A impacts cellular growth. (A) Depletion of hSETD1A in var ious CRC cell lines led to a decrease in cellular proliferation as measures by cell viability count. (B ) Depletion hSETD1A in HCT116 and SW48 cells led to a decrease in colonies size under adherent conditions. (C) D epletion of hSETD1A in HCT116 led to a de crease in colonies size and number in soft agar, under anchorage independent growth conditions. Shown are the mean SD of three independent experiments P < 0.05 by s tudent t test.


63 Figure 3 4 Depletion of hSETD1A is not associate d with cell death. ( A) Cell cycle analysis of shSETD1A and scramble control HCT116 cells w as carried out using propidium i odide staining. Depletion of hSETD1A did not affect Sub G1 phase. (B) Depletion of hSETD1 did not affect cell death in the presence or absence of Hydrogen Peroxide (H2O2)


64 CHAPTER 4 HUMAN SETD1A REGULATES THE WNT SIGNALING PATHWAY Introductory Remarks Constitutive activation of the Wnt signaling pathway is a common feature of colorectal tumors [ 58 ] The downstream effect of constitutively activated Wnt pathway is up regulation of catenin protein levels a nd its target g ene s, such as c Myc [ 58 81 ] The W nt pathway is highly active in undifferen tiat ed cells that reside at the bott om of in testinal crypt s and regulates their self renewal [ 103 ] hSETD1A is not only over expressed at the crypt bottom ( Figure 3 1E ) but also interacts with catenin the master regulator of the Wnt signaling pathway [ 38 ] Therefore we were pr ompted to investigate whether hSETD1A relates to the Wnt signaling pathw ay. Results hSETD1A Regulates Transcription of Wnt Target Gene s To gain mechanistic insights into the role of hSETD1A in CRC we carried out genome wide microarray analysis in the scramble control and hSETD1A KD HCT116 cells. A total of 6,286 genes were di fferentially expressed (cutoff P value of 0.005) in the hSETD1A KD cells compared with the scramble control cells ( Figure 4 1A ). Among them, 2,863 genes were up regulated, whereas 3,423 genes were down regulated upon loss of hSETD1A ( Figure 4 1A ). Int erest ingly, almost 1% of the total differen tially expressed genes were Wnt signaling target genes (56 of 6, 286; 4 1A ), representing approximately 50% (56 of 113) of the total known Wnt target gene s. Further analysis revealed that the probability of these Wnt ta rgets being affected by the hSETD1A KD was signi fi cantly higher than that of the random distribution of genome wide hSETD1A affected genes (FDR adjusted P value = 4. 67E 8; Figure 4 1B ). Key Wnt signaling


65 target genes involved in proliferation, invasion, mi gration, and angiogenesis were down regulated in the hSETD1A depleted cells (31/56), whereas target genes repressed by this pathway, such as E cadherin ( CDH1 ) and SOX9 were up regulated ( Figure 4 1C and Table 4 1 ). Up regulation of other Wnt target genes could be an indirect effect of hSETD1A KD. The effect of hSETD1A silencing on the expression of selected Wnt target genes and two unaffected genes, CLDN7 and CCNA2 were further validated by qRT PCR in HCT116 and SW48 cells ( Figure 4 1C and Figure 4 1D ). W e also con fi rmed that the KD of hSETD1A in HCT1 16 cells reduced the protein levels of three important Wnt target genes; c Myc ( MYC ), Snail ( SNAI1 ), and TCF4 ( TCF7L2 ) ( Figure 4 2A ). Thus our data indicate that hS ETD1A regulates a subset of Wnt signaling t a rget genes in CRC cells hSETD1A Associates w ith catenin a nd Activates Wnt Target G enes To elucidate the mechanism whereby hSETD1A controls transcription of Wnt targ et gene s we examined the effect of hSETD1A KD on the expression levels of catenin in HCT116 and SW48 cells. The KD of hSETD1A did not h ave an e ffect on the mRNA expression level ( CTNNB1 ; Table 4 1) or protein level (Figs. 4 2 A) of catenin These results indicate that hSETD1A neither regulate s catenin transcription nor catenin protein stability. Interestingly, recent studies demonstr ated that hSETD1A can interact with catenin in the human embry onic kidney cells line HEK293 [ 38 ] To test whether hSETD1A collaborates with catenin in regu lating Wnt ta rget genes in CRC co immunoprecipitation assays were performed in HCT116 or SW48 nuclear extracts using antibodies against catenin or components of the hSETD1A complex As expected, endogenous catenin interacted with hSETD1A and its compo nents in both HCT116 and SW48 cells ( Figure 4 2B and Figure 4 2C ). In


66 addition, catenin was highly expressed in colorectal tumors and in normal crypt bottoms ( Figure 4 2D ) consistent with the role of the Wnt pathway in maintenance and regulation of inte stinal crypt cells [ 103 ] and similar to the hSETD1A expression pattern observed ( Figure 3 1E ). Thus we investigated whether catenin and hSETD1A collaborate to regulate the express ion of Wnt target genes. W e first asked if hSETD1A and catenin complexes are both recruited to the same regions at promoters of the Wnt/ catenin target genes. To that end, we performed ChIP assays using antibodies against the hSETD1 A complex including hSETD1A RbBP5, Ash2L, WDR5, and against the catenin complex including catenin and TCF4. B oth hSETD1A and catenin complexes were enriched at known catenin responsive elements at the MYC and VEGFA loci, R negative control region, in HCT116 cells ( Figure 4 2E and Figure 4 2F ). Moreover, the KD of hSETD1A in HCT116 and SW48 cells invoked a decrease in H3K4me3 enrichment at the TSSs of several Wnt/ catenin target genes that are involved in cell proliferatio n, invasion, migration, and angiogenesis, such as MYC SNAI1 VEGFA MMP7 and TWIST2 In contrast, there was no signi fi cant change in H3K4me3 levels at the TSS of CDH1 ( Figure 4 3A and Figure 4 3B ). Furthermore catenin binding to its target genes was u naffected by hSETD1A KD in HCT116 cells ( Figure 4 3C ). These results suggest that recruitment of the hSETD1A complex and subsequent trimethylation of H3K4 at proximal promoters of Wnt/ catenin target genes are required for catenin action.


67 Recruitment of hSETD1A b y catenin is Critical f or catenin m ediated Transcriptional A ctivation We next sought to further examine the possibility that hSETD1A is recruited by catenin to Wnt target gene promoters. To test this, catenin was silenced in HCT116 cells using two independent lentiviral shRNAs ( Figure 4 4 A). As expected the depletion of catenin invoked a decrease in expression of its target genes ( Figure 4 4C ) and impeded cellular proliferation ( Figure 4 4B ). Interestingly H3K4me3 levels were also sign i fi cant ly decreased at the TSSs of Wnt/ catenin target genes upon loss of catenin in HCT116 cells ( Figure 4 4D ). Moreover i nduc tion of the Wnt signaling pathway and catenin in cells (HEK293) that normally express low levels catenin, resu lted in enrichment of H3K4me3 at promoters of several Wnt target genes and to an increase in their expression levels ( Figure 4 5) T his effect of catenin on H3K4me3 was not due to changes in hSETD1A levels, as catenin KD did not affect hSETD1A protein levels ( Figure 4 4A ). Since hSETD1A interacts with catenin ( Figure 4 2B and Figure 4 2C ), we f urther investigated whether catenin directs hSETD1A to these genomic loci by examining the effect of catenin KD on hSE TD1A occupancy at promoters of selected Wnt/ catenin target gene s. The recruitment of hSETD1A to the proximal promoters of SNAI1 MYC and VEGFA was signi fi cantly reduced upon loss of catenin consistent with the observed decrease in H3K4me3 enrichment at tho se loci ( Figure 4 4E ). While the recrui tment of hSETD1A to the Wnt/ catenin responsive promoters is dependent on catenin ( Figure 4 4E ), the occupancy of catenin at t hese promoters is independent of hSETD1A ( Figure 4 3C ). Th ese results indicate d that catenin recruits hSETD1A to promoters of the Wnt/


68 catenin target genes and relies, at least partially, on hSETD1A mediated H3K4me3 for transcriptional activation. A link between promoter associated H3K4me3 and transcriptional activation was proposed previously to rely on the binding of PHD d omain containing TAF3 to H3K4me3 mark at active promoters [ 28 30 ] Therefore, we examine d whether binding of TAF3 to selected Wnt/ catenin targeted promoters w as impaired by hSETD1A depletion. TAF3 occupancy at the TSSs of MYC SNAI1 and VEGFA was decreased in the hSETD1A depleted HCT116 cells compared with the scramble control ( Figure 4 6 A ). Furthermore we identified Wnt target genes that are coregulated b y T AF3 and hSETD1A in HCT116 cells by comparing our microarray data set for hSETD1A KD in HCT116 cells with the TAF3 KD microarray data set in HCT116 cells obtained from NCBI GEO2R [ 28 30 ] First, w e found total of 991 genes that are coregulated by TAF3 and h SETD1A D1A [5605 (991) 2735] ( Figure 4 6 B ). Secondly hSETD1A and TAF3 share d 14 Wnt target genes including MYC TCF7L2 SNAI2 and MMP7 ( Figure 4 6 C ). These dat a support a novel mechanism for activation of the Wnt signaling pathway, in which catenin activates transcription of Wnt target genes by recruiting the hSETD1A comp lex to deposit H3K4me3 marks and assembly of the preinitiation com plex (PIC). Levels of hSETD1A are Positively Correlate d with Activation of Wnt Target Genes and Tumor G row th in Colorectal C ancer. To examine whether levels of hSETD1A correlate with H3K4me 3 levels at the TSSs of Wnt/ catenin target genes in human CRC we carried out H3K4me3 ChIP anal yses in colorectal tumor tissue s and adjacent normal mucosa from patient #1 which presented high levels of hSET D1A, and from patient #10 which did not show any change in hSETD1A levels compared to the normal mucosa In opposed to the tumor


69 from patient #10, tumor from patient #1 presented high levels of hSETD1A and increased occ upancy of H3K4me3 at TSSs of Wnt/ catenin target genes ( Figure 4 7 A ). M oreover, Wnt/ catenin target genes were highly expressed in tumors presenting high levels of hSETD1A (patient #1 3), bu t not in other tumors (patients #7 and #10; Figure 4 7 B ). These results support our hypothesis th at hSETD1A controls promoter H3K4me3 levels leading to the activation of Wnt/ catenin target genes and subsequently Wnt pathway mediated cellular proliferation. Finally, to assess the role of hSETD1A and its mediate d H3K4me3 in colorectal tumor growth, we employed a mouse xenograft model T he scramble control and hSETD 1A KD HCT116 cells were injected subcutaneously into the flank of female athymic nude mice, and primary tumor volume was evaluated at 2 weeks postinjection. Consistent with the in vitro data, t he KD of hSETD1A partially impaired primary tumor growth ( Figure 4 7 C ). Variations in tumors size may have been attributed to fluctuations in hSETD1A protein levels in the absence of antibiotic selection in vivo ( Figure 4 7D and Figure 4 7 E ). These resul ts suggest th at hSETD1A controls colorectal tumor gro wth in a dose dependent manner Closing Remarks The Wnt signaling pathway is controlled by levels of its master regulator catenin [ 58 ] This protein associates with promoters of Wnt target genes through its interactions with LEF/TCF transcription factors in the nucleus but the question of how catenin activates Wnt target genes is not entirely clear [ 58 ] Nevertheless, recent studies showed that catenin mediates histone modifications at targeted loci by recruiting various histone modifiers, such as p300, Brg1, MLL1, and MLL2 [ 38 49 87


70 90 91 ] These perspectives suggest that activation of the Wnt pathway requires two steps: up regulation in catenin and subsequent elevation in active histone marks. Ultimately, the increase in active marks at gene promoters serves as a nest for the basal transcription machinery and permits transcription. Here we brought to light yet another epigenetic regula tor hSETD1A that interacts with catenin and controls H3K4me3 at TSSs ( Figure 4 8 ) We showed that absence of either hSETD1A or catenin interferes with H3K4me3 at TSS of Wnt target genes and that transcription is down regulated. Like other studies, we demonstrated that drained levels of H3K4me3 leads to a decrease in TAF3 recruitment to TSSs, which explains diminished transcription. But, our discovery is novel in that it shows that hSETD1A can mediate this decrease in H3K4me3 and TAF3. Finally, we demo nstrated association between hSETD1A le vels and expression of Wnt target genes in colorectal tumors from two patients The relation s between hSETD1A and the Wnt pathway could account for the decrease in cellular growth observed in the absence of hSETD 1A F igure 4 1. hSETD1A regulates a subset of Wnt signaling target genes. (A) Genome wide microarray analysis of HCT116 cells depleted of hSETD1A revealed a decrease in the expression levels of Wnt target genes compared to the scramble control cells. Venn diagr am shows the overlap between total differentially expressed genes in the hSETD1A KD and Wnt signaling target genes. The P value (FDR adjusted) < 0.005. (B) Statistical analysis was performed using a hypergeometric distribution test comparing the distributi on of Wnt target genes among the total hSETD1A target genes to the random distribution of hSETD1A target genes. The probability of Wnt target genes that were affected by the hSETD1A KD (N=56; red line) is significantly higher than that of the randomly dist ributed hSETD1A target genes (black curve). The green dashed line indicates the P value=0.05 cutoff expected if the number of representative Wnt target genes would have been N=38. (C and D) validation of microarray analysis by qRT PCR in HCT116 cells (C) a nd SW48 cells (D). Wnt target genes expression is normalized to GAPDH and relative to the scramble control. Shown are the mean SD of two or three independent repeats. P < 0.05 by s tudent t test.




72 Table 4 1 Expression Profile of Wnt target genes in hSETD1A KD HCT116 cells Down regulated (p value<0.005) Up regulated (p value< 0.005) Unaffected (p value>0.005) Gene name P value (FDR) Gene name P value (FDR) Gene name P value (FDR) Gene name P value (FDR) GDNF 4.77E 03 NRP1 7.99E 06 GDF5 1.26E 01 MYC BP 5.18E 03 TCF7L1 3.54E 03 FGF18 1.16E 05 CD44 1.37E 01 EGR1 6.40E 03 SIX1 2.81E 03 RUNX2 1.80E 05 T 2.06E 01 WISP2 6.75E 03 FGF7 2.69E 03 AHR 2.71E 05 DAB2 2.27E 01 ETS2 7.02E 03 SOX2 2.10E 03 DKK1 3.52E 05 SOX17 2.57E 01 TLE1 7.76E 03 TWIST1 2.03E 03 PTGS2 6.07E 05 CUBN 3.65E 01 BIRC5 9.00E 03 MET 2.00E 03 ANTXR1 8.47E 05 NANOG 4.18E 01 HIF1A 9.65E 03 ABCB1 1.74E 03 EPHB3 1.05E 04 FGF4 5.06E 01 SMO 7.02E 03 NRCAM 1.67E 03 IGF2 1.25E 04 MYCN 5.27E 01 LRP1 6.36E 03 VEGF 1.67E 03 KLF5 1.43E 04 WNT3 A 5.69E 01 EFNB1 5.71E 03 MMP2 1.38E 03 CYR61 1.53E 04 MAP3K3 6.61E 01 BTRC 5.03E 03 NTRK2 1.35E 03 PPAP2B 1.56E 04 CTNNB1 6.36E 01 CACNA2D3 8.00E 02 JUN 1.28E 03 GJA1 2.40E 04 PTCH 4.30E 01 GREM1 7.97E 02 LBH 1.10E 03 FGF9 2.48E 04 PLAUR 2.64E 01 ID2 6.11E 02 POU5F1 1.07E 03 CTGF 2.87E 04 CCND3 2.12E 01 IGF1 5.75E 02 MYC 1.05E 03 CCND1 4.38E 04 BMP4 1.93E 01 CDON 3.78E 02 WISP1 9.56E 04 CLDN1 6.54E 04 LGR5 1.64E 01 PITX2 1.50E 02 TERT 8.13E 04 CDH1 7.48E 04 FN1 1.45E 01 EDN1 1.44E 02 SALL4 8.03E 0 4 WNT9A 1.07E 03 DPP10 1.44E 01 ATOH1 1.38E 02 SNAI2 5.41E 04 AXIN2 1.20E 03 IL6 1.40E 01 TNFSF11 1.35E 02 FOSL1 4.70E 04 PDGFRA 1.26E 03 SFRP2 1.13E 01 CEBPD 1.31E 02 SNAI1 4.28E 04 IRS1 1.37E 03 DLK1 3.91E 02 FZD7 1.23E 02 WNT5A 2.97E 04 FZD2 1.71E 0 3 MMP9 4.51E 02 GAST 1.21E 02 MMP7 2.43E 04 SOX9 3.35E 03 TWIST2 5.08E 02 TGFB3 1.09E 02 TIAM1 2.38E 04 FST 4.28E 03 EPHB2 6.37E 02 NF1 1.24E 02 TCF7L2 1.66E 04 FGF20 6.75E 02 TGFB1 1.57E 02 CCND2 1.62E 04 PPARD 7.12E 02 TNFRSF11B 1.71E 02 CDK N2A 8.46E 05 EGFR 3.87E 02 BGLAP 1.76E 02 LEF1 3.67E 05 TCF7 2.20E 02 ANGPTL4 2.05E 05 JAG1 1.94E 05 hSETD1A 7.99E 06


73 Figure 4 catenin.( A) Protein levels of Wnt target genes, catenin, are decreased in hSETD1A KD HCT116 cells. (B and C) co catenin in HCT116 cells (B) and in SW48 cells (C). (D) IF staining of colorectal tumor and paired normal mucosa from patient #2 indicate catenin in the intestinal crypt bottoms. (E and F) hSETD1A complex catenin responsive elements in the VEGFA loci (F) and MYC loci (E), in HCT116 cells. ChIP analysis was performed using antib odies against hSETD1A complex (hSETD1A, Ash2L, catenin and TCF4) to determine their binding to responsive elements compared with a negative control region located 70 Kb downstream of the TAL1 promoter and relative to IgG, in HCT116.


74 Figure 4 3. hSETD1A mediates promoter H3K4me3 at Wnt target loci. (A and B ) H3K4me3 ChIP in HCT116 cells depleted of hSETD1A, C7 and C19 clones. (C) H3K4me3 ChIP in SW48 cells depleted of hSETD1A, cell pool. (D) catenin ChIP in HCT 116 cells depleted of hSETD1A, clone C7. Promoter H3K4me3 (A and B), but not catenin (C), diminished in the absence of IgG. Shown are the mean SD of two or three independent experime nts P < 0.05, by s tudent t test


75 Figure 4 4. catenin recruits hSETD 1A to regulate Wnt target gene expression. catenin was knocked down in HCT116 cells using two independent lentiviral shRNAs E/1 and E/2 (A) Depletion of catenin did not affect h SETD1A protein levels compared to the shLuciferase (shLuc) control, as detected by WB analysis. Depletion of catenin led to a decrease in cellular proliferation (B) and to a decrease in expression of Wnt target gene, as detected by qRT PCR (C). Wnt targe t genes expression is normalized to GAPDH and relative to the s hLuciferase control KD of catenin resulted in reduced occupancy of H3K4me3 (D) and hSETD1A (E) at TSSs of the Wnt target genes compared TAL1 + 70 negative control regions as det ected by ChIP analyses. Shown are the mean SD of three replicates P < 0.05, by s tudent t test


76 Figure 4 4. Continued


77 Figure 4 catenin levels lead to an increase in H3K4me3. (A) Experimental design. MEF cells over expressing the hu L cells ) were cultured and conditioned medium ( WCM) was then collected HEK293 cells, which express low levels of catenin were cultured in WCM and catenin levels in HEK293 cells upon 2 ho urs of WCM treatments. qRT PCR analysis of Wnt target genes (C) and promoter H3K4me3 ChIP (D) in HEK293 cells upon 2 hours of WCM treatments H3K4me3 was enriched at promoters of Wnt target genes and Wnt targets expression levels were elevated upon induct catenin. Shown are the mean SD of three replicates P < 0.05, by student t test


78 F igure 4 6 TAF3 and hSETD1A co regulate Wnt target genes. (A) TAF3 ChIP in hSETD1A KD HCT116 cells. TAF3 occupancy decreases at TSSs of several Wnt target gene s in the absence of hSETD1A and H3K4me3. Shown are the mean SD of three replicates P < 0.05, by s tudent t test. (B) Venn diagram presenting the overlapping among target genes of hSETD1A, Wnt, and TAF3 in HCT116 cells. TAF3 was silenced in HCT116 cells. TAF3 microarray data was analyzed with the online software NCBI GEO2R [ 134 ] using the data set GSE43542 [ 29 ] to call the differentially expressed genes with the cutoff p value (FDR adjusted) of 0.05. Cutoff p value used for the shSETD1A microarray was 0.005 (FDR adjusted). Genes that appear in multiple arrays were identif ied with GeneWeaver. TAF3 target genes significantly overlapped with hSETD1A target genes (p value=4.27E 90) and with Wnt target genes (p value=0.002 ). (C) Shown are Wnt target genes that regulated by both hSETD1A and TAF3.


79 Figure 4 7 hSETD1A levels associate with H3K4me3 and Wnt target gene express ion in human colorectal tumors. (A) H3K4me3 ChIP assays in human colorectal tumors and normal mucosa from two patients; patient # 1, which exhibit high levels of hSETD1A, and patient #10, which did not sho w change in hSETD1A levels compared with the paired normal mucosa. In oppose to colorectal tumor from patient #10, the c olorectal tumor from patient # 1 revealed enrichment of H3K4me3 at TSSs of Wnt target genes compared with the matching normal mucosa. (B ) Wnt target genes are highly active in human colorectal tumors in which hSETD1A is highly expressed (patients #1 3), but are less active in tumors in which hSETD1A levels are unchanged compared with the normal mucosa (patients #7 and #10). Levels (mRNA) o f Wnt target were determined by qRT PCR and normalized to actin. Shown are the mean SD of three replicates P < 0.05, by s tudent t test. (C) Mouse xenograft model. shSETD1A (C7) or Scramble control HCT116 cells (1x10 6 ) were injected s.c. in 50% Matrigel into the flank of 5 6 week old females athymic nude mice Mice were sacrificed at 2 weeks and tumor volume was measured. (D and E) hSETD1A, c MYC, and catenin protein levels were analyzed by WB in cellular extract from t umor s of different sizes hSETD1A levels positively corresponded to tumor xenograft sizes.




81 Figure 4 catenin in regulation of Wnt target genes catenin levels remain low due to its phosphorylation and subsequent degrada tion. In colorectal cancer catenin levels are up catenin to the nucleus enables its interactions with hSETD1A HMT. These interactions facilitate promoter H3K4me3, recruitment of TAF3 (and transcription machi catenin target genes, and cellular growth.


82 CHAPTER 5 HUMAN SETD1A REGULATES METASTASIS IN BREAST CANCER Introductory Remarks Breast cancer is the most common cancer in women, and the second leading cause of cancer related death s among women in the USA [ 113 114 ] According to the National Cancer Institute, while there is 99 84% chance fi ve year relative survival for localized and regional disease, there is only 24% chance five year relative survival for metastatic breast cancer [ 113 ] As m etastasis lies in the heart of therapeutic failure and mortality [ 117 ] understanding and preventing metast asis is critical for intervention and survival Metastatic cancer cells are cells that in addition to have acquired uncontrollably proliferative characteristics have also acquired excessive motility and are highly adaptive and resistant to new environment s [ 83 85 116 117 ] Alterations in e pigenetic modifications are important for conveying genomic adaptation, cells plasticity, and sustention of a phenotype [ 10 12 ] Understanding of the epigenetic changes that occur during progression of breast cancer is important not only in making decisions regarding diagnosis and prognosis but also in treatments strategies. Epigen etic regulators are often mutated or over expressed in cancer [ 11 ] and vast epigenetic changes, such as increase in global H3K4me3 have been observed during induced EMT switch [ 45 127 ] Gain in hSETD1A gene copy number was detected in 45% (352/782) of breast cancer, the high est frequency of hSETD1A gain in gene copy among all cancers cases ( ) [ 55 ] Therefore, we investigated whether hSETD1A could affect the metastatic pheno type in breast cancer


83 Results hSETD1A and hSETD1B a re Up regulated in Breast Cancer hSETD1A is up regulated in colorectal carcinoma and controls cellular proliferation [ 135 ] However, the role of hSETD1A in metastasis has not been studied. H3K4me3 has been shown to increase during EMT [ 45 ] suggesting dysregulation of HMTs or/and Histone demethylases. In fact, it was recently demonstrated that H3K4 demethylase LSD1 regulates EMT and metastasis in breast cancer [ 45 128 129 ] Ablation of LSD1 in MDA MB 231 cells invoked an incr ease in cell invasion and migration, in vitro and lung metastasis in vivo Yet, the involvements of H3K4 HMTs in breast cancer metastasis in unknown. First, gain in hSETD1A gene copy number was detected in 45% (352/782) of breast cancer cases compared wi th only 6 .8% loss in gene copy (53/782) (Table 5 1). This frequency of hSETD1A gain of gene copy is the highest among all cancers. On the other hand loss in gene copy of hSETD1B (19.3%), MLL1 (49.4%), MLL2 (16.5%), MLL3 (22.0%), was more prevalent than ga in in copy number, in breast cancer. MLL4 gene copy was lost in 11.4% and gained in 18.8% of breast cancer samples ( ) [ 55 ] Together, this data suggest an increase in hSETD1A expression and a unique role for hSETD1A in breast cancer. To examine this hypothesis we detected protein levels of hSETD1A in breast cancer cell lines compared with the mammary epithelial cell lines MCF10A and HMEC. hSETD1A and H3K4me3 levels were eleva ted in breast cancer cell lines including MCF7, MDA MB 231, BT 459, and SUM159 ( Figure 5 1A). In addition, mRNA levels of both hSETD1A and hSETD1B were detected by qRT PCR in primary and metastatic breast tumors compared with normal breast samples. hSETD1A and hSETD1B levels were elevated in metastatic


84 breast tumors compared with primary breast tumors and normal breast samples ( Figure 5 1B). However, hSETD1A levels in p rimary breast tumors were not significantly elevated compared with normal breast tissues ( Figure 5 1B). Increased Levels of hSET D1A and hSETD1B Are Associated w ith Poor Survival in Lymph node Positive Breast Cancer Since hSETD1A and hSETD1B were up regul ated in metastatic breast tumors, we examined the relations between their expression levels and clinical outcomes in breast cancer. Kaplan Meier survival curves of breast cancer patients who were lymph node positive or lymph node negative were plotted and grouped by low/high levels of hSETD1A or hSETD1B ( [ 136 ] ( Figure 5 2A and Figure 5 2B). Interestingly, increased levels of hSETD1A or hSETD 1B were associated with poor survival in lymph node positive but not in lymph node negative breast cancer. On the other hand, levels of other COMPASS members, such as MLL1, MLL2, or MLL4, were not associated with survival of lymph node positive or lymph no de negative breast cancer patients ( Figure 5 2A and Figure 5 2B). This is consistent with the accumulated knowledge about MLLs, implicating them mainly in hematologic malignancies It also coincides with the overall agreement that dysregulation of MLLs in leukemia is largely through gene rearrangements rather than abnormal transcriptional [ 50 51 ] The association of hSETD1A and hSETD1B with survival in breast cancer suggests that they could serve as good biomarker s for progression of breast cancer in patients who showed metastasis to the sentinel lymph node (Figure 5 2A ) The fact that hSETD1A and hSETD1B were not associated with survival in lymph node negative breast cancer suggests that the significance of their expression levels is most likely relevant to metastasis ( Figure 5 2B).


85 Perturbed hSETD1A Reduces Metastasis The role of hSETD1A in invasion and migration of bre ast cancer cells was investigated in vitro and in vivo hSETD1A was stably knocked down in MDA MB 231 and BT 459 cells using a lentiviral shRNA (Figure 5 3A). These cells express low levels of E cadherin ( Figure 5 1A) and are therefore highly migrat ory and invasive Interestingly, the hSETD1A KD MDA MB 231 cells showed morp hological changes suggesting Mesenchymal Epithelial Transition (MET). For example m esenchymal MDA MD 231 cells beca me round shaped and smaller, characteristics of epithelial cells ( Figure 5 3B). We next assessed the effect of hSETD1A KD on cell migration using wound healing assay ( Figure 5 3C) and transwell migration assay ( Figure 5 3D ). hSETD1A KD cells shown a decrease in their migratory capacity compared to the control cells in both assay s ( Figure 5 3C and Figure 5 3D) In addition, hSETD1A KD MDA MB 231 ce lls showed decreased invasion by a transwell invasion assay ( Figure 5 3E). These results indicate d that hSETD1A is important not only for cell motility but also for inv asion thr ough matrigel which imitates the surrounding tissue. The KD of hSETD1A had only minor effect on cellular proliferation of MDA MD 231 and BT 549 cells (Figure 5 3F). Finally, we also tested the ability of the hSETD1A KD cells to metastesize in vivo MDA MD 2 31 shSETD1A or MDA MD 231 sh L uciferase cells (1 x10 6 ) were injected into the tail vein of nude mice and metastasis in the lungs was assessed after 7 8 weeks. We found a significant decrease in the number of lung nodules detected in mice which were injected with the MDA MD 231 shSETD1A cells compared with mice that were injected with the MDA MD 231 shuciferase cells ( Figure 5 4). The r e fore, we speculated that the up regulation in hSETD1A observed in advanced breast cancer


86 ( Figure 5 1A) could s erve as a regul atory mechanism by which hSETD1A stimulate s metastasis ( Figure 5 4 ) and associate with survival of lymph node positive breast cancer patients (Fig 5 2A) In other words, the increased levels of hSETD1A in primary breast tumors could account for the ability of the tumors to expand their territories and metastasize, thereby affecting clinical outcomes. Nevertheless, the molecular events controlled by hSETD1A in breast cancer are still not clear. Closing Remarks Piling evidences for alteration s in histone modi fication patterns in EMT and metastasis have led to investigation s of histone modifier enzymes during progression of canc er [ 45 85 92 126 129 ] Altered H3K4me3 ( Figure 5 1A ) and hSETD1A gene dupli cation s in breast cancer (Table 5 1) [ 55 ] have prompted us to inspect the biological function s of hSETD1A and its relative hSETD1B in breast cancer. In summary, we found that these enzymes are up regulated in breast cancer cell lines and in metastatic human breast cancer samples compared with normal mammary cells ( Figure 5 1) We f urther discovered association between increased lev els of hSETD1A and hSETD1B and poor survival in lymph node positive breast cancer ( Figure 5 2) More over, depletion of hSETD1A in mesenc hymal breast cancer cells invoked a decrease in metastasis ( Figure 5 4) However, g ene Loci affected by hSETD1A KD have not been identifie d in this study and therefore the molecular mechanism by which hSETD1A control met astasis is currently not clear. One interesting observation is that the hSETD1A KD in MDA MD 231 cells did not affect global H3K4me3 as it did in HCT116 cells. The reason for such difference between cell lines is unknown but could be extremely valuable in understanding the importance of this enzyme in different cell contexts.


87 We previously demonstrated that hSETD1A regulates the Wnt signaling pathway in CRC [ 135 ] This is also a plausible mechanism in breast cancer as the Wnt signaling pathway sur e ly regulates cell invasion and migration. Nevertheless, f urther investigation of how hSETD1A and hSETD1B regulate breast cancer metastasis is needed before we can effectively target them. Yet it seems that those enzymes might serve as goo d markers for clinical outcomes in patients who were positive for breast metastas is in the sentinel lymph node. Table 5 1 hSETD1A gene copy analysis in various types of human cancers Tissue Gain in Copy Number (%) Loss in Copy Number (%) Breast Lung Large intestine Kidney NS Ovary Endometrium 45 22.1 18.9 18 13.6 10.6 10.2 6.8 15.1 4.1 1.3 4.5 30.3 6.1 ( [ 55 ] )


88 Figure 5 1. hSETD1A, hSETD1B, and H3K4me3 are up regulated in breast cancer.(A) Cell extract were prepared for WB from mammary epithelial cell lines : MCF10A and HMEC and the breast c ancer cell line s : SUM159 MCF7 BT 549, and MDA MB231. Levels of hSETD1A and H3K4me3 were up regulated in the breas t cancer cells compared with normal mammary epithelial cell s. E cadherin expression was lower in SUM159 BT 549, and MDA MB231, indicating a pro mesenchymal phenotype. (B) hSETD1A and hSETD1B mRNA levels were detected by qRT PCR in normal breast tissues and in primary and metastatic breast cancer samples (n=6 each). hSETD1A and hSETD1B expression levels were both up regulated in metastatic brea st cancer compared to normal beast and primary breast cancer samples. Shown are the mean SD of three replicates P < 0.05, * P < 0.01, ** P < 0.001 by s tudent t test.


89 Figure 5 2. hSETD1A and hSETD1B level s predict clinical outcomes for l ymph node pos itive breast cancer patients. S hown are Kaplan Meier plots of lymph node positive (A) and l ymph node negative (B) cohorts. Patients were separated into two groups with mRNA levels higher or lower than the median levels of the indicated gene Hazard ratio w ith 95% confidence intervals and logrank P value were calculated. Microarray analysis used the Affymetrix HG U133A, HG U133 Plus 2.0, and HG U133A 2.0 ( ; 2012 version [ 136 ] ). Increased levels of hSETD1A (Aff ID 213202_at) and hSETD1B (Aff ID 213153_at) but not levels of MLLs, was associate d with p oor survival of lymph node positive but not l ymph node negative b reast cancer patients


90 Figure 5 3. Depletion of hSETD1A reduces cell inva sion and migration. (A) Lentivirus mediated KD of hSETD1A in breast cancer cell line s MDA MD 231 and BT 549 (B) hS ETD 1A KD MDA MD 231 cells invoked morp hological changes resembling Mesenchy mal Epithelial Transition (MET) The effect of hSETD1A KD on cell migration was assessed by wound healing assay (C ) and transwell migration assay (D). (E) The impact of hSETD1A KD on cell invasion was assessed by transwell invasion assay. MDA MD 2 31 cells (1x10 4 ) were seeded and invaded cells were counted 17 hours later. Shown are the mean SD of five independent repeats. ** P < 0.01by s tudent t test. (F) Proliferation assay.


91 Figure 5 4 KD of hSETD1A reduces metastasis in mice MDA MD 231 shSETD 1A or MDA MD 231 shLuciferase cells (1x10 6 ) in 1X PBS were injected into the tail vein of female athymic nude mice. Mice were sacrificed 7 8 weeks later and lung nodules were counted P < 0.05 by student t test. shLuc ; n=7, shSETD1A; n=9.


92 CHAPTER 6 SUM MARY AND DISCUSSION Significance Understanding the reasons for the epigenetic turmoil observed in cancer is critical for targeting specific epigenetic regulators and rebooting epigenetic. The goal of this dissertation was to elucidate the roles of hSETD1 H MTs and H3K4me3 in cancer related functions, in solid tumors. The rationale for this work was the observations of up regulation in H3K4me3, hSETD1A and hSETD1B in cancer and the frequent mutations of hSETD1A/B in CRC and breast cancer ( Figure 3 1 and Figu re 5 1) [ 39 46 55 ] The H3K4me3 mark was suggested to play a role in transcription activation by recrui ting transcription regulators to gene promoters [ 28 34 ] H ence misregulation of hSETD1A/B could change normal gene expression patterns and affect cellular func tion W e therefore proposed a model in which the HMTs hSETD1A and hSETD1B play pro oncogenic roles in CRC and breast cancer [ 135 ] Here we showed that h SETD1A and hSETD1B are overexpresses in 62 % and 26% of the human colorectal carcinomas examined respectively [ 135 ] These resul ts led us to focus our efforts o n hSETD1A. W e uncover an imp ortant pro proliferative ro le for hSETD1A in CRC and identified a possible link between this function and the Wnt signaling pathway [ 135 ] Dysfunction of the Wnt pathway plays an essential role in the manifestation and progression of most CRCs [ 65 ] We found that hSETD1A regulates proximal promoters of Wnt target genes through its interactions with catenin a centered player which conveys the Wnt signal [ 135 ] Such interactions have been previously identified in HEK293 cells, but the overall meaning of the se interaction s was not elucidated [ 38 ] Similarly in catenin and relatives of hSE TD1A,


93 such as MLL1 and MLL2 were previously identified and were shown to play a role in activating the c MYC enhancer in CRC [ 49 ] The possibility that MLLs play a similar role to hSETD1A in regulating other Wnt target genes has not been explored. While H3K4me3 occupies proximal promoters, H3K4me1 is rather enriched at enhancers [ 14 25 ] Interestingly, e pigenomic analysis of H3K4me1 indicated changes in enhancer associated H3K4me1 in CRC [ 39 ] I t was further suggested that enhancer associated H3K4me1 can drive a speci fi c transcription program to promote colon carcinogenesis [ 39 ] This common role for H3K4me1 and H3K4me3 in activating transcription of CRC related genes is interesting considering the different mechanism s utilized by MLLs and hSETD1 A/ B to activate gene transcriptio n ; while both MLLs and hSETD1 A/B can activate gene transcription, MLLs do so through reg ulation of enhancer H3K4me1 while hSETD1A/B do so through regulation of proximal promoter H3K4me3 [ 18 21 24 ] In our context, this could imply a clever mechanism whereby catenin amplifies transcription of Wnt target genes: recruitment of hSETD1A to proximal promoters complemented by the recruitment of MLLs to enhancers of Wnt target genes. This concept could be readily tested by ChIP seq of COMPASS members and H3K4 methy lation states in CRC cells. catenin act as a platform to recruit histone mod ifying enzymes such as hSETD1A, CBP/ p300, and MLLs [ 38 49 90 91 ] It was further proposed that the catenin /TCF/LEF complex mediates the assembly of mediator and TFIID complexes that are required for ac tivation of Wnt target genes [ 137 ] It was suggested that TFIID is recruited to these loci in Drosophila through the binding of TAF4 to Pygopus, a protein which associates with the catenin /TCF/LEF complex [ 137 ] Nevertheless, we


94 proposed another mechanism wherein hSETD1A mediates H3K4me3 at Wnt targeted promoters upon activation on the W nt pathway followed by direct binding of TAF3 to H3K4me3 [ 135 ] The H3K4me3 modification was previously shown to interact with the PHD domain of TAF3 a TFIID subunit [ 28 30 ] Our data demonstrated that TAF3 and hSET D1A share a large number of target genes including 14 Wnt target genes, and that hSETD1A dependent H3K4me3 is critical for the recruitment of TAF3 to Wnt/ catenin targeted promoters in HCT116 cells [ 135 ] An o ther interesting finding of this dissertation is that hSETD1A is highly expressed and colocalized with cateni n at intestinal crypts where proliferating ISCs reside and where the Wnt signaling pathway is constitutively active [ 135 ] A ctivation of the Wnt signaling pathway is a key event in maintenance of ISC self renewal [ 102 106 ] It serves as a master switch between the proliferation and differentiation of crypt cells hence involved in intestinal malignancies [ 102 106 ] Thus, aberrant expression of hSETD1A in crypt bottoms may cooperate with catenin in activating Wnt signaling target genes and initiating oncogenic programs Based on our results, we depicted a model dem onstrating collaborative relations between hSETD1A and catenin to activate t ranscription program that could lead to the de velopment of CRC [ 135 ] In this model, catenin and hSETD1A are up regulated and engage in the nucleus. catenin recruits hSETD1A to Wnt putat ive promoters to confer H3K4me3 [ 135 ] This allows for PIC assembly activation of Wnt target genes, and subs equent promotion of cellular growth. Transcriptional activation of Wnt target genes is the last an d most important step of the Wnt signaling pathway [ 58 ] Identifying a nd targeting catenin partners, such as hSETD1A could be an effective therapeutic


95 approach for inhibiting the Wnt signaling pathway in CRC [ 57 74 ] Such targeting strategies showed effective, for example in the case of inhibitors of the interactions catenin [ 111 ] In a ddition to our intention to study the effect of hSETD1A/B on colorectal tumor growth, we were also intrigued to understand their roles in metastasis. Unfortunately, the study of metastatic CRC in vitro is difficult since most CRC cell lines are epithelial However, we found evidences for misregulation of hSETD1A and/or hSETD1B in other cancers especially in breast cancer (Table 5) [ 55 ] Gain in copy number of the hSETD1A gene was present in 45% of breast cancers (Table 5 1) [ 55 ] Consistent with that we showed that hSETD1A and hSETD1B are overexpressed in breast cancer specimens and cell lines (Figure 5 1) Fortunately, there are many mesenchymal breast cancer cell lines available, and in fact breast cancer is a model disease for the study of metastasis. Hence, we further characterize hSETD1A and hSETD1B in partial, in breast cancer metastasis. Increased levels of both hSETD1A and hSETD1B were associated with worse outcomes in ly mph node positive breast cancer (Figure 5 2) There was no clear association between levels of hSETD1 A and survival in cohorts grouped by level s of human epidermal growth factor receptor 2 ( HER2 ) progesterone receptor ( PR ) or estrogen receptors ( ER ) [ 136 ] Nevertheless, increased hSETD1B levels were associated with poor survival of PR/ER double negative cohorts [ 136 ] Together, t hese results support that hSETD1 enzymes could serve as good biomarkers for pro gnosis in breast cancer patients who already diagnosed with at least l ymphatic metastasis. It is n oteworthy that there was even strong er association in lung cancer [ 138 ] Increased


96 levels of hSETD1A (P value = 1.3e 06) was associated with 10 year poor outcomes for lung cancer patient s [ 138 ] On the other hand, the exact opposite association was observed for hSETD1B in lung cancer increased levels of hSETD1B (P value = 3e 07) were associated with favored outcomes [ 138 ] This differential association of hSETD1A and hSETD1B levels with survival in lung cancer, although not directly relevant to the findings of this current work, is interesting because it suggest s that those enzymes are not entirely redundant and emphasize s the need to carefully characterize them independently. Despite statistical significant relations between hSETD1A/B level s and breast cancer outcomes (hSETD1A; *P <0.02, hSETD1B; **P<0.009), it should be noted that they were not dramatically significant (Figure 5 2) In fact, it is likely that the large number of samples that were used for this analysis could have influenced the statistical test ( too much statistical power ) [ 136 ] While i t is difficult to comment on the clinical relevance an d significance of these results, they at least support a role for hSETD1A and hSETD1B in advanced breast cancer. There was no association between COMPASS core components, such as RbBP5 and Ash2L and survival of lymph node positive breast cancer [ 136 ] However HCF1, component of both hSETD1A/B and MLL1/2 complexes, and WDR82, component unique to the hSETD1A/B complexes, showed association with lymph node positive breast cancer [ 136 ] Metastasis relevant functions such as migration and invasion were affected by the depletion of hSETD1A in MDA MD 231 or BT 549 cells (Figure 5 3). Moreover, depletion of hSETD1A im peded metasta si s of MDA MD 231 to the mouse lungs (Figure 5 4) Based on these results we concluded that hSETD1A impacts metastasis but the


97 mechanism was not elucidated Of note is the fact that in contrast to what was observed in HCT116 cells, depletion of hSETD1A in MDA MB 231 cells did not affect global H3K4me3 (data not shown) This could be related to the efficiency of the KD or more interestingly to the cell type. The effect of hSETD1B on metastasis has not been investigated due to technical diffic ulties in obtaining KD and for the lack of commercial antibodies. In summary, versatile roles for hSETD1A and hSETD1B in solid tumors were identified here for the first time. These results indicate that different COMPASS family members, although structura lly similar, can carry out different functions, some which are more prominent in blood cancers and others which are selective for solid cancers. Inhibition studies of SETD1A/B complementing the results obtained here are necessary for determining clinical r elevance and feasibility. Clinical Relevance The catenin signaling pathway was first characterized in the 1970s in Drosophila melanogaster [ 73 ] It was later recognized that constit utive activation of the catenin signaling pathway is a hallmark of numerous malignancies in human, especially gastrointestinal cancers [ 57 58 60 65 67 ] Traditionally, abnormal signaling of this pathway was attributed to genetic or somatic mutations in APC, Axin, and catenin but as layers of complexity of this pathway were revealed it became clear that there are many other mecha nisms affecting this signaling [ 58 ] Despite the many years of research and attempts to develop attractive inhibitors for this pathway, to date there are no therapies meant to attenuate the Wnt pathway in clinic [ 57 74 ] Although in vitro manipulations of the Wnt pathway were promising in theory in practice it has bee n difficult to obtain specific and effective therapeutic s [ 57


98 74 ] The re are several reasons for that: f irst, the Wnt signaling pathway is essential for normal development and regeneration particularly that of intestinal tissue. Therefore, blockage of this pathway is accompanied by unfavorable toxiciti es and side effects. Second potential target Wnt components are also involved in other cellular processes, thereby reducing the specificity of the candidate inhibitors. Third, the small number of enzymatic components in this pathway makes it difficult to target. Although in recent years there has been some progres s in targeting the Wnt pathway, the benefit for CRC is questionable as u pstream antagonists may not be effective in cases where the downstream components have been mutated For example, inhibitors of the Wnt putative receptors LRP and Frizzled or porcupin e (PORCN) are probably going to be ineffective in cases of APC truncations. These insights clearly support a rationale for targeting the pathway downstr eam of the destruction complex. However, i nhibitors targeting t he obvious catenin and TCF downstream were so far only shown useful in cell based assays [ 139 141 ] Nevertheless, small molecule inhibitor (PRI 724 ;Prism Pharma ) targeting the interaction s catenin and CBP/P300 and preventing transcriptio n of Wnt target genes, has recently entered phase I clinical trials for the treatment of solid tumors and leukemia [ 111 ] So far, data from 18 patients with solid tumors demonstrated an acceptable safety profile. These encouraging results suggest that targeting the interactions of catenin with other coactivators in the form of histone modifying enzym es for example hSETD1A, could also be beneficial. The findings of this dissertation not on ly identified hSETD1A as a novel binding partner of catenin in CRC cells but also showed that hSETD1A is up regulated in CRC


99 [ 135 ] Therefore, theoretically, several therapeutic strategies could be applied: specific in hibition of hSETD1A or inhibition of the interactions between hSETD1A and catenin. Nevertheless, i nhi bition of hSETD1A activity might carry out off target effect s as revealed by the large number of hSETD1A target genes in HCT116 cells ( Figure 4 1A). In a ddition specific inhibition of the hSETD1A SET domain might be difficult due to its high similarity to other COMPASS members. Since KD of hSETD1A only partially inhibited the Wnt signaling pathway, toxicities associated with blocking the Wnt pathway would likely be avoided. The successes of PRI 724 inhibitor in the phase I clinical trials could have be en attributed to the very same reason The i mpact of DNA and histone methylation events on the advanced disease is poorly understood compared with our unders tanding of initial oncogenic events. Metastasis is a hallmark of advanced cancer and the number one reason for therapeutic failure and death [ 6 116 117 ] Metastatic cells require down regulation of genes related to cell cell adhesion and up regulation of genes related to cell motility, angiogenesis, and matrix digestion [ 83 85 116 117 ] Epigenetic changes can greatly impact these characteristics [ 127 ] T herefore, identification of epigenetic regulatory mechanisms is important for potential therapeutic avenues. M is regulation of a large number of histone modifying enz ymes has been established in advanced cancer For examples the polycomb group protein histone lysine N methyltransferase EZH2 is overexpressed in many cancers and its expression correlates with tumor grade [ 142 143 ] The Heterochromatin protein 1 (HP1) is down regulated in metastatic breast cancer and control s cell invasion [ 144 ] LSD1, while promoting metastasis of colon cancer cells [ 145 ] suppress es metastasis of breast cancer [ 128 129 ] The potential reversibility of

PAGE 100

100 epigenetic events make s their regulators an attractive biomarkers and th erapeutic targets of the metastatic disease. The effect of hSETD1A on metastasis and its association with lymph node positive breast cancer found here, are other layer s demonstrating the importance of this enzyme in cancer. Whether hSETD1A also affect s CR C metastasis is currently unknown although likely. The Wnt signaling pathway controls many genes related to metastasis a nd angiogenesis [ 58 ] Therefore, as a regulator of the Wnt signaling pathway hSETD1A is expected to affect metastasis The question of how hSETD1A affect s metastasis is yet to be answered We currently do n ot have evidence that the enzymatic activity of hSET D1A is related to its function in metastasis and we do not know which genes are affected by it. Understanding of the molecular events regulated by hSETD1A during EMT and metastasis in breast cancer could be a key for an effective therapeutic. The identification of critical biomarkers in cancer has led to advances in cancer detection, prognosis, and treatment. hSETD1A and hSETD1B levels were associated with clinical outcomes in lymph node positive but not lymph node negative breast cancer (Figure 5 2) These results suggest that the targeting of hSETD1A in breast cancer will perhaps only be beneficial to patients who already showed signs of metastasis. While ER or PR pos itive breast cancer patient s benefit from endocrine therapy such as tamoxifen and HER2 positive breast cancer patie nts greatly benefit from trastuzamab, triple negative patients (negative for PR, ER, and HER2) representing 10 20% of breast cancers, are not currently a match for any targeted therapies [ 146 148 ] hSETD1A levels were not associated with survival in cohorts

PAGE 101

101 defined by ER, PR, or HER2 levels suggest ing an independent mechanism [ 136 ] On the other hand hSETD1B levels associated with poor 10 year survival of PR and ER negative breast cancer patients (P value = 0.034; n=185 ) [ 136 ] These results indicate that PR and ER negative patients who showed lower levels of hSETD1B, have better chances to survive breast cancer. Although hSETD1A and hSETD1B levels associated with clinical outcomes these results require validation. There were many important database [ 136 ] For example, age at diagnosis genetic predispositions such as BRCA1/2 mutations, pre or post menopausal and childbearing [ 113 ] These characteristics play a critical role in prognosis of breast cancer patients and should be considered when characterizing biomarkers. Future Direction The findings of this study present a novel role for hSETD1A in solid tumors and call for further research. Future s tudies elucidating the effect of hSETD1A in vivo are critical, f or example, testing of the capacity of hSETD1A conditional KO to reduce the burden of colorectal polyp s in mouse model s, such as the Apc Min mouse model [ 94 95 ] In ad dition, the effect of hSETD1A on the development of breast cancer and metastasis needs to be further investigated in vivo While we utilized tail in jection model to study metastasis of MDA MB 231 shSETD1A cells, an orthotopic model might better recapitulate the natural habitat and microenvironment of breast cancer cells [ 149 ] That is implantation of breast cancer cells in the mammary fat pad leadin g to the development of a primary tumor and metastasis to distant sites. Other mouse models could be utilized for genetic manipulations to rigorously study the effect of various genes on the development and progression of breast cancer [ 149 ] For e xample, the MMTV

PAGE 102

102 Neu mouse m odel in which the neu (Erbb2) onco gene was inserted under the transcriptional control o f the mouse mammary tumor virus (MMTV) promoter wherein mice develop mammary tumors at 29 weeks [ 150 ] Crossing these mammary tumor prone mice with hSETD1A conditional KO strain would allow for testing the effect of hSETD1A KO on the onset and progression of breast cancer. Nevertheless it is somewhat difficult to predict which mouse model would be most suitable as the molecular mechanism of how hSETD1A affect breast cancer metastasis is currently unknown. Therefore, in vitro studies of gene expression profile s of hSETD1A /B KD in breast cancer cell line s should be applied first. In addit ion, the utilization of hSETD1A/B as biomarkers in breast cancer sh ould be further studied while considering tumor heterogeneity ; examining the association of hSETD1A/B with survival of patients b e aring breast tumors of different classifications, such as histological type, grade, lymph node status and the presence of predictive markers such as ER /PR/HER2 [ 113 ] In addition other patient characteristics, such as pre/post menopausal, childbearing age, and genetic predispositions such as BRCA1 and BRCA2 mutations, should also be considered [ 113 ] Complementary studies examining the impact of hSETD1A /B KD in different classifications of breast cancer cell lines would be usef ul in further understanding the relation s of hSETD1A /B to already known biomarkers. For example, ER PR HER2 + ( SKBR3 ), ER PR HER2 ( BT549, MDA MB 231 MDA MB 468 ), Luminal A ER + PR +/ HER 2 ( MCF 7, T47D, SUM185 ), Luminal B ER + PR +/ HER 2 + ( BT474, ZR 75 ) [ 151 ] Full hSETD1A KO was sh own by our lab and others to be embryonic lethal, indicating that hSETD1A is essential for normal development [ 52 ] Therefore,

PAGE 103

103 investigating the possible functions of hSETD1A in normal development of intestine and breast tissues is an important one. To that end conditional KO of hSETD1A could be generated Moreover, the effect of hSETD1A on normal intestinal tissue should be carefully studied considering our important finding s revealing that hSETD1A is highly expressed at the intestinal crypt [ 135 ] Th ese results suggest a regulatory role for hSETD1A in self renewal or/and differentiation of intestinal stem cells and regeneration of intestinal tissue. First, the expression of hSETD1A in specific type of crypt cells should be more comprehensively studied Second, hSETD1A dependent mechanisms regul ating self renewal migration, and differentiation of cell s in the crypt dependently or independently of the Wnt signaling pathway should be examined. The normal human intestinal epithelial crypt like (HIEC) cell line and the villus like primary cultures of di fferentiated enterocytes (PCDE) could be resourceful for studying the former in vitro [ 152 ] Increasing evidences indicate that catenin mediates activation of Wnt target genes by recruiting epigenetic modifying enzymes [ 38 49 87 90 91 ] This idea could be more carefully tested though induction of catenin followed by ChIP seq an alys e s of various histone modifications. Understanding of the role of histone modifications in activation of the Wnt signaling pathway could lead to the discovery of novel targets in the form of histone modifying enzymes. Finally, we discovered a link be tween the Long non coding RNA (lncRNA) Hemolinc first discovered in our lab, and breast cancer. This lncRNA is expressed from the HOXB locus and serves a transcription enhancer for HOXB genes. Hemolinc interacts with the hSETD1A complex and affect many target genes ot her than the

PAGE 104

104 HOXB genes, f or example genes associated with EMT. Therefore we investigate d the expression level s of hemolinc in breast cancer cell lines and found that compared with the human mammary epithelial cell lines MCF10A and HMEC, it is over expressed in the mesenchymal breast cancer cell lines MDA MB 231 and SUM159 These preliminary results could suggest that the hemoli n c regulates EMT, perhaps through interactions with hSETD1A. Nevertheless, further research is required to answer t he question of whether or not Hemoli n c affect s progression of breast cancer and how.

PAGE 105

105 APPENDIX THE LONG NON CODING RNA HEMOLINC IS UP REGULATED IN MESENCHYMAL BREAST CANCER CELLS Long non coding RNAs (lncRNAs) are diverse class of transcribed RNA molecul es of 200 or more nucleot ides that do not encode protein However, like protein s they are express ed in a spatial and temporal manner and can regulate variety of cellular processes, including gene transcription. A significant number of lncRNAs have been sho wn to collaborate with epigenetic regulators to modify gene transcription [ 153 ] The long Hemolinc and was shown to transcribed from the HOXB locus (Deng et. al.). HoxB associated Hemolinc governs transcription ac tivation of anterior HoxB genes and promotes hematopoietic differentiation of mouse embryonic stem cells (mESC). Importantly, Hemolinc associates with the conserved SET domain of the Setd1a/MLL1TrxG complexes, in the erythroleukemia cell line K562 and in embryo id b odies (EB). Epithelial plasticity is a common theme during embryonic development and cancer progression [ 82 83 85 ] Like cancer cells, embryonic stem cells undergo EMT during development in order to migrate and travel long distances to form tissues and organs. Because Hemolinc regulates stem cell differentiati on and interact s with the EMT related hSETD1A we investigated the potential link between Hemoli n c and EMT. This idea was further supported by the change s in expression of EMT related genes upon KD of Hemolinc in mESC. Therefore we analyzed the expression levels of Hemolinc in breast cancer cell lines in which hSETD1A was elevated and compared it with human mammary epithelial cell lines. Hemolinc expression increased significantly in the triple negative mesenchymal breast cancer cell lines MDA MB 231 and BT 549, but not in the epithelial breast cancer cell line MCF7 or in the human mammary

PAGE 106

106 epithelial cell lines MCF10A or HMEC ( Figure A 1A ) Interestingly, the expression of Hemolinc was low in the HER2 positive mesenchymal breast cancer cell line SUM159 cells ( Figure A 1A). These results imply a role for Hemolinc in EMT. We further examined the expression of Hemolinc in normal human breast tissues and in primary and metastatic breast cancer samples There was no association between Hemolinc expression levels and progression of breast cancer (Figure A 1B). However, it is difficult to generalize these results considering the large variation in breast tumor subtype s and the small number of samples we tested (n=6 7 each category).

PAGE 107

107 Figure A 1. Hemolinc expres sion in breast cancer. (A) Expression of hemolinc in normal and breast cancer cell lines by RT PCR Hemolinc is over expressed in the mesenchymal breast cancer cell lines MDA MB 231 and SUM159. The bottom figure shows the appropriate controls; No treatment (NT), treatment of RNA with RNase A prior to reverse transcription (RNase), Reverse transcription in the absence of RT polymerase (no RT). actin is shown as a control gene. (B) Expression of Hemolinc in normal breast, breast carcinoma, and metastatic br east samples by RT PCR.

PAGE 108

108 LIST OF REFERENCES 1. Luger, K., et al., Crystal structure of the nucleosome core particle at 2.8 A resolution. Nature, 1997. 389 (6648): p. 251 60. 2. Dawson, M.A. and T. Kouzarides, Cancer epigenetics: from mech anism to therapy. Cell, 2012. 150 (1): p. 12 27. 3. Kouzarides, T., Chromatin modifications and their function. Cell, 2007. 128 (4): p. 693 705. 4. Zentner, G.E. and S. Henikoff, Regulation of nucleosome dynamics by histone modifications. Nature structural & molecular biology, 2013. 20 (3): p. 259 66. 5. Negrini, S., V.G. Gorgoulis, and T.D. Halazonetis, Genomic instability -an evolving hallmark of cancer. Nature reviews. Molecular cell biology, 2010. 11 (3): p. 220 8. 6. Hanahan, D. and R.A. Weinberg, Hallmark s of cancer: the next generation. Cell, 2011. 144 (5): p. 646 74. 7. Feinberg, A.P., H. Cui, and R. Ohlsson, DNA methylation and genomic imprinting: insights from cancer into epigenetic mechanisms. Seminars in cancer biology, 2002. 12 (5): p. 389 98. 8. Fein berg, A.P. and B. Vogelstein, Hypomethylation distinguishes genes of some human cancers from their normal counterparts. Nature, 1983. 301 (5895): p. 89 92. 9. Jones, P.A. and S.B. Baylin, The fundamental role of epigenetic events in cancer. Nature reviews. Genetics, 2002. 3 (6): p. 415 28. 10. Feinberg, A.P., The epigenetic basis of common human disease. Transactions of the American Clinical and Climatological Association, 2013. 124 : p. 84 93. 11. Timp, W. and A.P. Feinberg, Cancer as a dysregulated epigenome allowing cellular growth advantage at the expense of the host. Nature reviews. Cancer, 2013. 13 (7): p. 497 510. 12. Pujadas, E. and A.P. Feinberg, Regulated noise in the epigenetic landscape of development and disease. Cell, 2012. 148 (6): p. 1123 31. 13. Rodriguez Paredes, M. and M. Esteller, Cancer epigenetics reaches mainstream oncology. Nature medicine, 2011. 17 (3): p. 330 9. 14. Barski, A., et al., High resolution profiling of histone methylations in the human genome. Cell, 2007. 129 (4): p. 823 37.

PAGE 109

109 15. Bernstein, B.E., et al., Genomic maps and comparative analysis of histone modifications in human and mouse. Cell, 2005. 120 (2): p. 169 81. 16. Kim, T.H., et al., A high resolution map of active promoters in the human genome. Nature, 2005. 436 (7052): p. 87 6 80. 17. Mohan, M., et al., The COMPASS family of H3K4 methylases in Drosophila. Molecular and cellular biology, 2011. 31 (21): p. 4310 8. 18. Shilatifard, A., The COMPASS family of histone H3K4 methylases: mechanisms of regulation in development and disea se pathogenesis. Annual review of biochemistry, 2012. 81 : p. 65 95. 19. Miller, T., et al., COMPASS: a complex of proteins associated with a trithorax related SET domain protein. Proceedings of the National Academy of Sciences of the United States of Ameri ca, 2001. 98 (23): p. 12902 7. 20. Nagy, P.L., et al., A trithorax group complex purified from Saccharomyces cerevisiae is required for methylation of histone H3. Proceedings of the National Academy of Sciences of the United States of America, 2002. 99 (1): p. 90 4. 21. Ardehali, M.B., et al., Drosophila Set1 is the major histone H3 lysine 4 trimethyltransferase with role in transcription. The EMBO journal, 2011. 30 (14): p. 2817 28. 22. Hallson, G., et al., dSet1 is the main H3K4 di and tri methyltransferase throughout Drosophila development. Genetics, 2012. 190 (1): p. 91 100. 23. Herz, H.M., et al., Enhancer associated H3K4 monomethylation by Trithorax related, the Drosophila homolog of mammalian Mll3/Mll4. Genes & development, 2012. 26 (23): p. 2604 20. 24. Hu, D., et al., The MLL3/MLL4 branches of the COMPASS family function as major histone H3K4 monomethylases at enhancers. Molecular and cellular biology, 2013. 33 (23): p. 4745 54. 25. Heintzman, N.D., et al., Distinct and predictive chromatin signatures of transcriptional promoters and enhancers in the human genome. Nature genetics, 2007. 39 (3): p. 311 8. 26. Azuara, V., et al., Chromatin signatures of pluripotent cell lines. Nature cell biology, 2006. 8 (5): p. 532 8. 27. Bernstein, B.E., et al., A bivalent chromatin structure marks key developmental genes in embryonic stem cells. Cell, 2006. 125 (2): p. 315 26. 28. Vermeulen, M., et al., Selective anchoring of TFIID to nucleosomes by trimethylation of histone H3 lysine 4. Cell, 2007. 131 (1): p. 58 69.

PAGE 110

110 29. Lau berth, S.M., et al., H3K4me3 interactions with TAF3 regulate preinitiation complex assembly and selective gene activation. Cell, 2013. 152 (5): p. 1021 36. 30. van Ingen, H., et al., Structural insight into the recognition of the H3K4me3 mark by the TFIID s ubunit TAF3. Structure, 2008. 16 (8): p. 1245 56. 31. Flanagan, J.F., et al., Double chromodomains cooperate to recognize the methylated histone H3 tail. Nature, 2005. 438 (7071): p. 1181 5. 32. Sims, R.J., 3rd, et al., Recognition of trimethylated histone H 3 lysine 4 facilitates the recruitment of transcription postinitiation factors and pre mRNA splicing. Molecular cell, 2007. 28 (4): p. 665 76. 33. Wysocka, J., et al., WDR5 associates with histone H3 methylated at K4 and is essential for H3 K4 methylation a nd vertebrate development. Cell, 2005. 121 (6): p. 859 72. 34. Wysocka, J., et al., A PHD finger of NURF couples histone H3 lysine 4 trimethylation with chromatin remodelling. Nature, 2006. 442 (7098): p. 86 90. 35. Dover, J., et al., Methylation of histone H3 by COMPASS requires ubiquitination of histone H2B by Rad6. The Journal of biological chemistry, 2002. 277 (32): p. 28368 71. 36. Kim, J., et al., RAD6 Mediated transcription coupled H2B ubiquitylation directly stimulates H3K4 methylation in human cells. Cell, 2009. 137 (3): p. 459 71. 37. Deng, C., et al., USF1 and hSET1A mediated epigenetic modifications regulate lineage differentiation and HoxB4 transcription. PLoS genetics, 2013. 9 (6): p. e1003524. 38. Hoffmeyer, K., et al., Wnt/beta catenin signaling r egulates telomerase in stem cells and cancer cells. Science, 2012. 336 (6088): p. 1549 54. 39. Akhtar Zaidi, B., et al., Epigenomic enhancer profiling defines a signature of colon cancer. Science, 2012. 336 (6082): p. 736 9. 40. Ellinger, J., et al., Prognos tic relevance of global histone H3 lysine 4 (H3K4) methylation in renal cell carcinoma. International journal of cancer. Journal international du cancer, 2010. 127 (10): p. 2360 6. 41. Ellinger, J., et al., Global levels of histone modifications predict pro state cancer recurrence. The Prostate, 2010. 70 (1): p. 61 9. 42. He, C., et al., High expression of trimethylated histone H3 lysine 4 is associated with poor prognosis in hepatocellular carcinoma. Human pathology, 2012. 43 (9): p. 1425 35.

PAGE 111

111 43. Ke, X.S., et al., Genome wide profiling of histone h3 lysine 4 and lysine 27 trimethylation reveals an epigenetic signature in prostate carcinogenesis. PloS one, 2009. 4 (3): p. e4687. 44. Kwon, M.J., et al., Claudin 4 overexpression is associated with epigenetic derepr ession in gastric carcinoma. Laboratory investigation; a journal of technical methods and pathology, 2011. 91 (11): p. 1652 67. 45. McDonald, O.G., et al., Genome scale epigenetic reprogramming during epithelial to mesenchymal transition. Nature structural & molecular biology, 2011. 18 (8): p. 867 74. 46. Zhou, L.X., et al., Application of histone modification in the risk prediction of the biochemical recurrence after radical prostatectomy. Asian journal of andrology, 2010. 12 (2): p. 171 9. 47. Flintoft, L., Epigenomics: Reprogramming in transition. Nature reviews. Genetics, 2011. 12 (8): p. 522. 48. Chervona, Y. and M. Costa, Histone modifications and cancer: biomarkers of prognosis? American journal of cancer research, 2012. 2 (5): p. 589 97. 49. Sierra, J., e t al., The APC tumor suppressor counteracts beta catenin activation and H3K4 methylation at Wnt target genes. Genes & development, 2006. 20 (5): p. 586 600. 50. de Boer, J., V. Walf Vorderwulbecke, and O. Williams, In focus: MLL rearranged leukemia. Leukemi a, 2013. 27 (6): p. 1224 8. 51. Mohan, M., et al., Licensed to elongate: a molecular mechanism for MLL based leukaemogenesis. Nature reviews. Cancer, 2010. 10 (10): p. 721 8. 52. Bledau, A.S., et al., The H3K4 methyltransferase Setd1a is first required at th e epiblast stage, whereas Setd1b becomes essential after gastrulation. Development, 2014. 141 (5): p. 1022 35. 53. Nguyen, P., et al., BAT3 and SET1A form a complex with CTCFL/BORIS to modulate H3K4 histone dimethylation and gene expression. Molecular and c ellular biology, 2008. 28 (21): p. 6720 9. 54. Yadav, S., et al., hSET1: a novel approach for colon cancer therapy. Biochemical pharmacology, 2009. 77 (10): p. 1635 41. 55. Bamford, S., et al., The COSMIC (Catalogue of Somatic Mutations in Cancer) database a nd website. British journal of cancer, 2004. 91 (2): p. 355 8. 56. Gala, M. and D.C. Chung, Hereditary colon cancer syndromes. Seminars in oncology, 2011. 38 (4): p. 490 9.

PAGE 112

112 57. Garber, K., Drugging the Wnt pathway: problems and progress. Journal of the Natio nal Cancer Institute, 2009. 101 (8): p. 548 50. 58. Giles, R.H., J.H. van Es, and H. Clevers, Caught up in a Wnt storm: Wnt signaling in cancer. Biochimica et biophysica acta, 2003. 1653 (1): p. 1 24. 59. Harada, N., et al., Intestinal polyposis in mice with a dominant stable mutation of the beta catenin gene. The EMBO journal, 1999. 18 (21): p. 5931 42. 60. Sparks, A.B., et al., Mutational analysis of the APC/beta catenin/Tcf pathway in colorectal cancer. Cancer research, 1998. 58 (6): p. 1130 4. 61. Suraweera N., et al., Mutations within Wnt pathway genes in sporadic colorectal cancers and cell lines. International journal of cancer. Journal international du cancer, 2006. 119 (8): p. 1837 42. 62. Colorectal Cancer Facts & Figures 2014 2016 2014: Atlanta. 63. Hagan, S., M.C. Orr, and B. Doyle, Targeted therapies in colorectal cancer an integrative view by PPPM. The EPMA journal, 2013. 4 (1): p. 3. 64. Siegel, R., C. Desantis, and A. Jemal, Colorectal cancer statistics, 2014. CA: a cancer journal for clinicians, 2014. 64 (2): p. 104 17. 65. Fearon, E.R., Molecular genetics of colorectal cancer. Annual review of pathology, 2011. 6 : p. 479 507. 66. Eisenmann, D.M., Wnt signaling. WormBook : the online review of C. elegans biology, 2005: p. 1 17. 67. Markowitz, S.D. a nd M.M. Bertagnolli, Molecular origins of cancer: Molecular basis of colorectal cancer. The New England journal of medicine, 2009. 361 (25): p. 2449 60. 68. Kerber, R.A., et al., Frequency of familial colon cancer and hereditary nonpolyposis colorectal canc er (Lynch syndrome) in a large population database. Familial cancer, 2005. 4 (3): p. 239 44. 69. Taylor, D.P., et al., Population based family history specific risks for colorectal cancer: a constellation approach. Gastroenterology, 2010. 138 (3): p. 877 85. 70. Tolliver, K.A. and D.K. Rex, Colonoscopic polypectomy. Gastroenterology clinics of North America, 2008. 37 (1): p. 229 51, ix. 71. Zauber, A.G., et al., Colonoscopic polypectomy and long term prevention of colorectal cancer deaths. The New England jour nal of medicine, 2012. 366 (8): p. 687 96.

PAGE 113

113 72. Comprehensive molecular characterization of human colon and rectal cancer. Nature, 2012. 487 (7407): p. 330 7. 73. Sharma, R.P. and V.L. Chopra, Effect of the Wingless (wg1) mutation on wing and haltere developm ent in Drosophila melanogaster. Developmental biology, 1976. 48 (2): p. 461 5. 74. Anastas, J.N. and R.T. Moon, WNT signalling pathways as therapeutic targets in cancer. Nature reviews. Cancer, 2013. 13 (1): p. 11 26. 75. van 't Veer, L.J., et al., Molecular cloning and chromosomal assignment of the human homolog of int 1, a mouse gene implicated in mammary tumorigenesis. Molecular and cellular biology, 1984. 4 (11): p. 2532 4. 76. McMahon, A.P. and R.T. Moon, Ectopic expression of the proto oncogene int 1 in Xenopus embryos leads to duplication of the embryonic axis. Cell, 1989. 58 (6): p. 1075 84. 77. Seifert, J.R. and M. Mlodzik, Frizzled/PCP signalling: a conserved mechanism regulating cell polarity and directed motility. Nature reviews. Genetics, 2007. 8 (2) : p. 126 38. 78. Kohn, A.D. and R.T. Moon, Wnt and calcium signaling: beta catenin independent pathways. Cell calcium, 2005. 38 (3 4): p. 439 46. 79. Stamos, J.L. and W.I. Weis, The beta catenin destruction complex. Cold Spring Harbor perspectives in biolog y, 2013. 5 (1): p. a007898. 80. Fagotto, F., U. Gluck, and B.M. Gumbiner, Nuclear localization signal independent and importin/karyopherin independent nuclear import of beta catenin. Current biology : CB, 1998. 8 (4): p. 181 90. 81. He, T.C., et al., Identif ication of c MYC as a target of the APC pathway. Science, 1998. 281 (5382): p. 1509 12. 82. Nieto, M.A., Epithelial plasticity: a common theme in embryonic and cancer cells. Science, 2013. 342 (6159): p. 1234850. 83. Thiery, J.P., Epithelial mesenchymal tran sitions in tumour progression. Nature reviews. Cancer, 2002. 2 (6): p. 442 54. 84. Thiery, J.P., et al., Epithelial mesenchymal transitions in development and disease. Cell, 2009. 139 (5): p. 871 90. 85. De Craene, B. and G. Berx, Regulatory networks definin g EMT during cancer initiation and progression. Nature reviews. Cancer, 2013. 13 (2): p. 97 110.

PAGE 114

114 86. Zhang, X., et al., Gene regulatory networks mediating canonical wnt signal directed control of pluripotency and differentiation in embryo stem cells. Stem c ells, 2013. 31 (12): p. 2667 79. 87. Barker, N., et al., The chromatin remodelling factor Brg 1 interacts with beta catenin to promote target gene activation. The EMBO journal, 2001. 20 (17): p. 4935 43. 88. Shi, Y., et al., Coordinated histone modifications mediated by a CtBP co repressor complex. Nature, 2003. 422 (6933): p. 735 8. 89. Wohrle, S., B. Wallmen, and A. Hecht, Differential control of Wnt target genes involves epigenetic mechanisms and selective promoter occupancy by T cell factors. Molecular and cellular biology, 2007. 27 (23): p. 8164 77. 90. Hecht, A., et al., The p300/CBP acetyltransferases function as transcriptional coactivators of beta catenin in vertebrates. The EMBO journal, 2000. 19 (8): p. 1839 50. 91. Miyagishi, M., et al., Regulation of Lef mediated transcription and p53 dependent pathway by associating beta catenin with CBP/p300. The Journal of biological chemistry, 2000. 275 (45): p. 35170 5. 92. Li, Z., et al., Histone H4 Lys 20 monomethylation by histone methylase SET8 mediates Wnt ta rget gene activation. Proceedings of the National Academy of Sciences of the United States of America, 2011. 108 (8): p. 3116 23. 93. Kinzler, K.W. and B. Vogelstein, Lessons from hereditary colorectal cancer. Cell, 1996. 87 (2): p. 159 70. 94. Moser, A.R., H.C. Pitot, and W.F. Dove, A dominant mutation that predisposes to multiple intestinal neoplasia in the mouse. Science, 1990. 247 (4940): p. 322 4. 95. Su, L.K., et al., Multiple intestinal neoplasia caused by a mutation in the murine homolog of the APC gen e. Science, 1992. 256 (5057): p. 668 70. 96. Powell, S.M., et al., APC mutations occur early during colorectal tumorigenesis. Nature, 1992. 359 (6392): p. 235 7. 97. Miyoshi, Y., et al., Somatic mutations of the APC gene in colorectal tumors: mutation cluste r region in the APC gene. Human molecular genetics, 1992. 1 (4): p. 229 33. 98. Miyaki, M., et al., Characteristics of somatic mutation of the adenomatous polyposis coli gene in colorectal tumors. Cancer research, 1994. 54 (11): p. 3011 20.

PAGE 115

115 99. Polakis, P., The oncogenic activation of beta catenin. Current opinion in genetics & development, 1999. 9 (1): p. 15 21. 100. Samowitz, W.S., et al., Beta catenin mutations are more frequent in small colorectal adenomas than in larger adenomas and invasive carcinomas. C ancer research, 1999. 59 (7): p. 1442 4. 101. Johnson, V., et al., Exon 3 beta catenin mutations are specifically associated with colorectal carcinomas in hereditary non polyposis colorectal cancer syndrome. Gut, 2005. 54 (2): p. 264 7. 102. Crosnier, C., D. Stamataki, and J. Lewis, Organizing cell renewal in the intestine: stem cells, signals and combinatorial control. Nature reviews. Genetics, 2006. 7 (5): p. 349 59. 103. Hirata, A., et al., Dose dependent roles for canonical Wnt signalling in de novo crypt formation and cell cycle properties of the colonic epithelium. Development, 2013. 140 (1): p. 66 75. 104. Korinek, V., et al., Depletion of epithelial stem cell compartments in the small intestine of mice lacking Tcf 4. Nature genetics, 1998. 19 (4): p. 379 83. 105. Kuhnert, F., et al., Essential requirement for Wnt signaling in proliferation of adult small intestine and colon revealed by adenoviral expression of Dickkopf 1. Proceedings of the National Academy of Sciences of the United States of America, 2004 101 (1): p. 266 71. 106. van de Wetering, M., et al., The beta catenin/TCF 4 complex imposes a crypt progenitor phenotype on colorectal cancer cells. Cell, 2002. 111 (2): p. 241 50. 107. Boman, B.M., et al., Colonic crypt changes during adenoma development in familial adenomatous polyposis: immunohistochemical evidence for expansion of the crypt base cell population. The American journal of pathology, 2004. 165 (5): p. 1489 98. 108. David C. Smith, M.G., Wells Messersmith, Rashmi Chugh, David Mendelson, Jako b Dupont, Robert Stagg, Ann M. Kapoun, Lu Xu, Rainer K. Brachmann, Antonio Jimeno. A first in human Phase 1 study of anti cancer stem cell (CSC) agent OMP 54F28 (FZD8 Fc) targeting the WNT pathway in patients with advanced solid tumors [Abstract] 2013. 10 9. Gurney, A., et al., Wnt pathway inhibition via the targeting of Frizzled receptors results in decreased growth and tumorigenicity of human tumors. Proceedings of the National Academy of Sciences of the United States of America, 2012. 109 (29): p. 11717 2 2.

PAGE 116

116 110. Liu, J., et al., Targeting Wnt driven cancer through the inhibition of Porcupine by LGK974. Proceedings of the National Academy of Sciences of the United States of America, 2013. 110 (50): p. 20224 9. 111. Anthony B. El Khoueiry, Y.N., Dongyun Yang, Sarah Cole, Michael Kahn, Marwan Zoghbi, Jennifer Berg, Masamoto Fujimori, Tetsuhi Inada, Hiroyuki Kouji, Heinz Josef Lenz. A phase I first in human study of PRI 724 in patients (pts) with advanced solid tumors in 2013 ASCO Annual Meeting 2013 Universit y of Southern California Norris Comprehensive Cancer Center, Los Angeles, CA; University of Southern California, Los Angeles, CA; PRISM Pharma Co, Yokohama, Japan. 112. Kamory, E., J. Olasz, and O. Csuka, Somatic APC inactivation mechanisms in sporadic col orectal cancer cases in Hungary. Pathology oncology research : POR, 2008. 14 (1): p. 51 6. 113. Breast Cancer Facts and Figures 2013 2014 2013, American Cancer Society: Atlanta. 114. DeSantis, C., et al., Breast cancer statistics, 2013. CA: a cancer journa l for clinicians, 2014. 64 (1): p. 52 62. 115. Apostolou, P. and F. Fostira, Hereditary breast cancer: the era of new susceptibility genes. BioMed research international, 2013. 2013 : p. 747318. 116. Sethi, N. and Y. Kang, Unravelling the complexity of metas tasis molecular understanding and targeted therapies. Nature reviews. Cancer, 2011. 11 (10): p. 735 48. 117. Chiang, A.C. and J. Massague, Molecular basis of metastasis. The New England journal of medicine, 2008. 359 (26): p. 2814 23. 118. Ferrara, N., VEG F and the quest for tumour angiogenesis factors. Nature reviews. Cancer, 2002. 2 (10): p. 795 803. 119. Carmeliet, P., Mechanisms of angiogenesis and arteriogenesis. Nature medicine, 2000. 6 (4): p. 389 95. 120. Tan, G.J., et al., Cathepsins mediate tumor me tastasis. World journal of biological chemistry, 2013. 4 (4): p. 91 101. 121. Gialeli, C., A.D. Theocharis, and N.K. Karamanos, Roles of matrix metalloproteinases in cancer progression and their pharmacological targeting. The FEBS journal, 2011. 278 (1): p. 16 27. 122. Rorth, P., Collective cell migration. Annual review of cell and developmental biology, 2009. 25 : p. 407 29.

PAGE 117

117 123. Swartz, M.A., et al., Tumor microenvironment complexity: emerging roles in cancer therapy. Cancer research, 2012. 72 (10): p. 2473 8 0. 124. 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. Proceedings of the National Academy of Sciences of the United States of A merica, 2012. 109 (41): p. 16654 9. 125. Yook, J.I., et al., Wnt dependent regulation of the E cadherin repressor snail. The Journal of biological chemistry, 2005. 280 (12): p. 11740 8. 126. Tam, W.L. and R.A. Weinberg, The epigenetics of epithelial mesenchy mal plasticity in cancer. Nature medicine, 2013. 19 (11): p. 1438 49. 127. Wang, Y. and Y. Shang, Epigenetic control of epithelial to mesenchymal transition and cancer metastasis. Experimental cell research, 2013. 319 (2): p. 160 9. 128. Li, Q., et al., Bind ing of the JmjC demethylase JARID1B to LSD1/NuRD suppresses angiogenesis and metastasis in breast cancer cells by repressing chemokine CCL14. Cancer research, 2011. 71 (21): p. 6899 908. 129. Wang, Y., et al., LSD1 is a subunit of the NuRD complex and targe ts the metastasis programs in breast cancer. Cell, 2009. 138 (4): p. 660 72. 130. Lemieux, S., Probe level linear model fitting and mixture modeling results in high accuracy detection of differential gene expression. BMC bioinformatics, 2006. 7 : p. 391. 131 Smyth, G.K., J. Michaud, and H.S. Scott, Use of within array replicate spots for assessing differential expression in microarray experiments. Bioinformatics, 2005. 21 (9): p. 2067 75. 132. Lao, V.V. and W.M. Grady, Epigenetics and colorectal cancer. Natur e reviews. Gastroenterology & hepatology, 2011. 8 (12): p. 686 700. 133. Enroth, S., et al., Cancer associated epigenetic transitions identified by genome wide histone methylation binding profiles in human colorectal cancer samples and paired normal mucosa. BMC cancer, 2011. 11 : p. 450. 134. Davis, S. and P.S. Meltzer, GEOquery: a bridge between the Gene Expression Omnibus (GEO) and BioConductor. Bioinformatics, 2007. 23 (14): p. 1846 7. 135. Salz, T., et al., hSETD1A regulates Wnt target genes and controls t umor growth of colorectal cancer cells. Cancer research, 2013.

PAGE 118

118 136. Gyorffy, B., et al., An online survival analysis tool to rapidly assess the effect of 22,277 genes on breast cancer prognosis using microarray data of 1,809 patients. Breast cancer researc h and treatment, 2010. 123 (3): p. 725 31. 137. Cadigan, K.M. and M. Peifer, Wnt signaling from development to disease: insights from model systems. Cold Spring Harbor perspectives in biology, 2009. 1 (2): p. a002881. 138. Gyorffy, B., et al., Online surviva l analysis software to assess the prognostic value of biomarkers using transcriptomic data in non small cell lung cancer. PloS one, 2013. 8 (12): p. e82241. 139. Wang, W., et al., A diterpenoid derivative 15 oxospiramilactone inhibits Wnt/beta catenin signa ling and colon cancer cell tumorigenesis. Cell research, 2011. 21 (5): p. 730 40. 140. Lepourcelet, M., et al., Small molecule antagonists of the oncogenic Tcf/beta catenin protein complex. Cancer cell, 2004. 5 (1): p. 91 102. 141. Gonsalves, F.C., et al., A n RNAi based chemical genetic screen identifies three small molecule inhibitors of the Wnt/wingless signaling pathway. Proceedings of the National Academy of Sciences of the United States of America, 2011. 108 (15): p. 5954 63. 142. Deb, G., V.S. Thakur, an d S. Gupta, Multifaceted role of EZH2 in breast and prostate tumorigenesis: epigenetics and beyond. Epigenetics : official journal of the DNA Methylation Society, 2013. 8 (5): p. 464 76. 143. Simon, J.A. and C.A. Lange, Roles of the EZH2 histone methyltrans ferase in cancer epigenetics. Mutation research, 2008. 647 (1 2): p. 21 9. 144. Norwood, L.E., et al., A requirement for dimerization of HP1Hsalpha in suppression of breast cancer invasion. The Journal of biological chemistry, 2006. 281 (27): p. 18668 76. 14 5. Ding, J., et al., LSD1 mediated epigenetic modification contributes to proliferation and metastasis of colon cancer. British journal of cancer, 2013. 109 (4): p. 994 1003. 146. Banin Hirata, B.K., et al., Molecular Markers for Breast Cancer: Prediction o n Tumor Behavior. Disease markers, 2014. 2014 : p. 513158. 147. Cornejo, K.M., et al., Theranostic and molecular classification of breast cancer. Archives of pathology & laboratory medicine, 2014. 138 (1): p. 44 56. 148. Lehmann, B.D. and J.A. Pietenpol, Ide ntification and use of biomarkers in treatment strategies for triple negative breast cancer subtypes. The Journal of pathology, 2014. 232 (2): p. 142 50.

PAGE 119

119 149. Fantozzi, A. and G. Christofori, Mouse models of breast cancer metastasis. Breast cancer research : BCR, 2006. 8 (4): p. 212. 150. Guy, C.T., et al., Expression of the neu protooncogene in the mammary epithelium of transgenic mice induces metastatic disease. Proceedings of the National Academy of Sciences of the United States of America, 1992. 89 (22): p 10578 82. 151. Holliday, D.L. and V. Speirs, Choosing the right cell line for breast cancer research. Breast cancer research : BCR, 2011. 13 (4): p. 215. 152. Beaulieu, J.F. and D. Menard, Isolation, characterization, and culture of normal human intestina l crypt and villus cells. Methods in molecular biology, 2012. 806 : p. 157 73. 153. Marchese, F.P. and M. Huarte, Long non coding RNAs and chromatin modifiers: Their place in the epigenetic code. Epigenetics : official journal of the DNA Methylation Society 2013. 9 (1).

PAGE 120

120 BIOGRAPHICAL SKETCH Tal Hila Salz was born (1984) in Israel to Nili and David Salz. She holds a n American and Israeli dual citizenship Tal a ttended Mevo HaGalil, the private regional elementary school (1991 199 7), located at kibbutz Ayelet HaShahar and graduated from the Har Vagai private junior and senior regional high school (1997 2002). She joined the women professional basketball leag ue at the age of 15 (1998 2008 ) and was recruited to the youth U 20, and adult Israeli national team. She was the captain of U 20 Israeli national basketball team ( age 18 21) with she has travelled to the youth Olympic games in Spain and played the Euro league. Tal has also completed a mandatory army se rvice (2002 2004) in which she w as awarded a prestigious athlete status that allowed her to continue playing professional basketball under min imal service conditions. Tal graduated from the University of Bar Ilan in Ramat Gan Israel, with a Bachelor of Science degree in B iotechnology (2005 2008). She then retired from basketball and relocated to the United States to work as a scientist at the laboratory of Dr. Karl Drlica at the Public Health Research Institute Center, University of Medicine and Dentistry of New Jersey (UMDNJ) S he co authored one scientif ic paper ( Cell Rep. 2013 ) and a book ( Antibiotic Discovery and Development, 2009 ) discussing Fluoroquinolone resistance In 2009, Tal was admitted to the PhD program in Biomedical Sciences at the University of Florida, Gainesville, Fl. She joined the la boratory of Dr. Suming Huang and completed two curriculums ; Bio chem istry and Molecular Biology and Clinical and Translational Science. At the laboratory of Dr. Huang Tal studied epigenetic regulation in cancer and co authored several scientific papers

PAGE 121

121 ( PLoS Genet. 2013, Leukemia. 2013 ) and one first author paper ( C ancer Res. 201 4 ) During her graduate years she was awarded several internal incentive and travelling awards, as well as a training grant in clinical and translational science TL1. She was also invited to speak at an international symposium on Cancer epigenetic (Keystone S ymposia, Santa Fe, NM, 2013 ) and part took in poster sessions in other international and internal meetings. Tal recei ved her doctoral degree in May 2014.