Genome Assembly and Finishing Alla Lapidus, Ph.D. Associate Professor Fox Chase Cancer Center
Fox Chase Cancer Center Annual Progress Report: 2009 ... Chase Cancer Center... · Fox Chase Cancer...
Transcript of Fox Chase Cancer Center Annual Progress Report: 2009 ... Chase Cancer Center... · Fox Chase Cancer...
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 1
Fox Chase Cancer Center
Annual Progress Report: 2009 Formula Grant
Reporting Period
July 1, 2011 – June 30, 2012
Formula Grant Overview
The Fox Chase Cancer Center received $3,477,323 in formula funds for the grant award period
January 1, 2010 through December 31, 2013. Accomplishments for the reporting period are
described below.
Research Project 1: Project Title and Purpose
Regulation of Genomic and Epigenomic Stability at CpG Sites - The purpose of this project is to
understand how two base excision repair enzymes, thymine DNA glycosylase (TDG) and
methyl-CpG-binding endonuclease 1(MED1)/ methyl binding domain 4 (MBD4), not only
protect CpG sequences from transition mutations and ward off against endogenous deamination
events, but also mediate DNA demethylation and modulate DNA methylation states, chromatin
structure and gene transcription. This dual role in genomic and epigenomic stability of CpG sites
is likely to be important for cancer formation, given the frequent occurrence of CpG to CpA (or
TpG) point mutations and hypermethylation of tumor suppressor genes in cancer. Thus,
knowledge accumulated in this project may lead to novel strategies for cancer prevention.
Anticipated Duration of Project
1/1/2010 - 12/31/2013
Project Overview
The conversion of cytosine into 5-methylcytosine (DNA methylation) is an important epigenetic
modification of the mammalian genome that is potentially mutagenic. In fact, 5-methylcytosine
has a tendency to spontaneously deaminate, generating thymine, which, if not removed from the
G:T mismatch prior to replication, will lead to incorporation of adenine. Indeed, deamination at
CpG sites is estimated to cause nearly one-third of all mutations in both cancer and human
genetic diseases. Work conducted by us and others indicate that in order to maintain genomic
stability at CpG sites, two base excision repair enzymes remove the offending thymine with their
DNA N-glycosylase activity, MED1 (previously identified by us and also known as MBD4) and
TDG.
DNA methyltransferases are known to initiate DNA methylation, whereas it is less clear how the
removal of the methylation, or demethylation, takes place. Evidence in Arabidopsis and
Zebrafish points to the role of the DNA repair machinery in affecting demethylation, either by
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 2
direct removal of 5-methylcytosine, or by its enzymatic conversion by deaminases into thymine,
which is then removed from the resulting G:T mismatch. However, it is still largely unknown
how mammalian cells demethylate the DNA and what functional consequences demethylation
has in mammals. Given the ample remodeling of methylation patterns at CpG sites during
development, and the hypermethylation of tumor suppressor genes and other CpG islands in
human cancer, it is expected that a defect in demethylating activities may lead to a situation of
epigenomic instability and contribute to developmental alterations and tumor formation.
The overall goal of this project is to test the hypothesis that the mammalian DNA repair enzymes
TDG and MED1/MBD4 have dual functions in protection from tumorigenesis and
developmental disorders: 1) Promote genomic stability at CpG sites by suppressing mutations;
and 2) Maintain proper DNA methylation patterns by regulating CpG demethylation.
Additional funding was secured from NIH during the first year of this grant to support the work
outlined below under aims 1 and 2. This project has been expanded to logically continue the
research through the addition of aims 3, 4 and 5. Therefore, the focus of the remainder of this
project will be to analyze the effect of a modification we have recently identified in MED1,
sumoylation, on MED1 DNA damage response and repair capacity; to characterize the structural
and functional properties of the AID-GADD45a-TDG demethylating complex, that we have
recently described in mammalian cells; and to examine the effects of double inactivation of
MED1 and TDG.
Thus, to test the general hypothesis outlined above, we now propose five Specific Aims:
Aim 1: Characterize the requirement of TDG during development and its role in DNA
demethylation and chromatin modification, using knock-out and knock-in mice;
Aim 2: Evaluate the role of MED1 and TDG in tumorigenesis using animal models of intestinal
cancer and analyzing human cancer specimens.
Aim 3: Analysis of MED1 sumoylation in DNA damage response and repair;
Aim 4: Characterization of the structure and function of the TDG demethylating complex;
Aim 5: Characterization of the phenotype of TDG-MED1-double mutant embryos.
Principal Investigators
Alfonso Bellacosa, MD, PhD
Associate Professor
The Fox Chase Cancer Center
333 Cottman Avenue
Philadelphia, PA 19111-2434
Other Participating Researchers
Shannon Dalton, PhD - employed by The Fox Chase Cancer Center
Expected Research Outcomes and Benefits
By evaluating the contribution of pathways of base excision repair (via MED1/MBD4 and TDG)
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 3
to the maintenance of CpG site integrity, we expect to obtain important information of relevance
to several related issues. We anticipate that our studies will uncover the mechanisms by which
endogenous mutagenesis by deamination is counteracted in mammalian cells, and will have
immediate implications for cancer biology. Indeed, failure of these repair systems and
acquisition of a mutator phenotype at CpG sites or loss of the epigenetic control afforded by CpG
methylation may be critical for cancer formation. In the future, such knowledge might suggest
innovative cancer prevention or therapeutic strategies.
Summary of Research Completed
In the timeframe 7/1/2011-6/30/2012, we made significant progress in all the Specific Aims
outlined above.
Specific Aim 1.
We confirmed the important role of the TDG catalytic domain in development. In fact, not only
in homozygosity (embryonic lethality), but also in heterozygosity, the TdgN151A
mice (knock-in
allele with inactivation of TDG glycosylase activity) show mild developmental defects.
In the past year, in order to determine which pathway(s) of active cytosine demethylation are
engaged by TDG during development (e.g deamination-dependent and -independent pathways),
we developed a dot-blot procedure using antibodies directed against 5hmU, 5fC and 5caC.
Interestingly, we detected an enrichment of some of these novel DNA bases in genomic DNA
extracted from TDG knock-out embryos in comparison to wild type embryos, thus suggesting
that TDG plays a role in their removal in vivo during DNA demethylation in development
(unpublished data).
In the past year, in order to determine which pathway(s) of active cytosine demethylation are
engaged by TDG during development (e.g deamination-dependent and -independent pathways),
we developed a dot-blot procedure using antibodies directed against 5hmU, 5fC and 5caC.
Interestingly, we detected an enrichment of some of these novel DNA bases in genomic DNA
extracted from TDG knock-out embryos in comparison to wild type embryos, thus suggesting
that TDG plays a role in their removal in vivo during DNA demethylation in development
(unpublished data).
Specific Aim 2.
In the past year, we confirmed that the detected downregulation of TDG is mediated by
overexpression of a specific microRNA cluster (unpublished data). This Aim also involves an
analysis of the effect of double inactivation of MED1 and TDG on intestinal tumor formation
associated with the APC Min (multiple intestinal neoplasia) mutation. We have initiated the
crosses between the TDG- and MED1-mutant mice with the APC-Min mice and FABP-Cre
transgenic mice in order to generate the TDG+/-
MED1+/-
FABP-Cre mice that will be interbred
with the TDG+/flox
MED1+/-
Apc-Min mice, in turn, to generate the experimental mice.
Specific Aim 3.
In the past year we characterized a novel modification of MED1, i.e. sumoylation, that we
previously discovered. By using an in vitro sumoylation assay and a mutant recombinant SUMO
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 4
protein, we have been able to precisely map the three most important sumoylation sites in MED1
(unpublished data). Through the use of the novel SUMO mutant engineered by us, this strategy
should be applicable to the mapping of sumoylation sites in other target proteins.
Specific Aim 4.
In the past year, in addition to AID and GADD45a, we have identified additional components of
the TDG demethylating complex, including other proteins previously known to play an
important role in DNA repair. We have also initiated a collaboration with Dr. Javier Di Noia
(Institut de Recherches Cliniques de Montréal, Canada), to precisely map the interaction
interfaces between TDG and the deaminase AID.
Specific Aim 5.
We have initiated the crosses between the TDG- and MED1-mutant mice in order to generate the
double heterozygous TDG+/-
MED1+/-
mice that will be interbred, in turn, to generate the
experimental mice/embryos.
Research Project 2: Project Title and Purpose
Estrogen Receptor-Mediated Gene Silencing in Tumorigenesis - The role of estrogen in
contributing to women’s cancers, especially those of the breast and uterus, is well established at
the epidemiologic level. The molecular mechanism through which inappropriate timing and/or
excessive levels of estrogen exposure initiate or promote human tumorigenesis remains largely
unknown, however. The purpose of this study is to test the hypothesis that the activated estrogen
receptor, which functions as a transcription factor, is also capable of silencing genes, through
promoter hypermethylation, that normally function to suppress tumorigenesis in human cells
subject to regulation by estrogen. In addition, we will explore the molecular mechanism of this
methylation-dependent gene silencing.
Anticipated Duration of Project
1/1/2010 - 12/31/2012
Project Overview
The broad objective of this project is to test the hypothesis that the activated estrogen receptor
silences the expression of critical tumor suppressor genes, and further, to test whether this
methylation-mediated gene silencing involves the recruitment of DNA methyl-transferases
(DNMTs) and/or histone deacetylase (HDAC) proteins to the promoter regions of target genes.
A substantial body of preliminary data has been generated to support these hypotheses. These
objectives will be tested through five specific aims. Aim one will be to confirm that treatment of
estrogen receptor (ER)-positive human breast cancer cells with estrogen results in the down-
regulation of expression of the genes encoding lipocalin 2 (LCN2) and corticotropin releasing
hormone (CRH). Aim two will be to determine whether this down-regulation of expression is
associated with hypermethylation of the promoter regions of these two genes, to map the specific
methylation sites, and to determine the time course of methylation in relation to gene silencing.
Aim three will be to test whether this phenomenon is associated with the physical association of
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 5
activated ER with DNMT1 and/or DNMT3B, and the physical association of HDAC1 with this
complex in the promoter regions of LCN2 and CRH. Further, we will test whether deacetylation
and methylation occur at H3-K9 in conjunction with these biophysical events. In aim four,
RNAi approaches will be used to test whether knockdown of DNMTs prevents the ER-activation
associated methylation and silencing of LCN2 and CRH. Finally, aim five will be to determine
whether exposure of human breast cancer cells to recombinant LCN2 or CRH has a growth
inhibitory or apoptotic effect, confirming the physiologic relevance of this phenomenon. These
aims will be accomplished using standard techniques of cell culture, qRT-PCR, western blotting,
real-time MS-PCR, bisulfite sequencing, chromatin immunoprecipitation (ChIP)-PCR, stable
transfection of shRNA expression vectors, and cell proliferation and apoptosis assays.
Principal Investigator
Jeff Boyd, PhD
Senior Vice President, Molecular Medicine
Executive Director, Institute for Personalized Medicine
The Fox Chase Cancer Center
333 Cottman Ave
Philadelphia, PA 19111-2434
Other Participating Researchers
Eric Ariazi, PhD - employed by The Fox Chase Cancer Center
Expected Research Outcomes and Benefits
Successful completion of the aims of this research project will provide additional insight into the
mechanism through which estrogen contributes to human tumorigenesis, e.g., breast and
endometrial cancers. Such an understanding will allow for a more informed approach to the use
of this hormone pharmacologically, in terms of risk-benefit analysis. Furthermore, a
fundamental understanding of the molecular mechanism through which estrogen contributes to
the development or progression of women’s cancers is essential in order to develop more
effective preventive strategies for these cancers, in light of the long-term exposure of females to
this hormone over a reproductive lifetime. Finally, proof of the hypotheses proposed in this
project will lead to a new model for the role of hormone receptor transcription factors in human
carcinogenesis, which would provide greater insight into the process of tumorigenesis in a wide
range of cancers not limited to those associated with estrogen exposure.
Summary of Research Completed
The importance of estrogen and estrogen receptor alpha (ERα) in contributing to women’s
cancers, especially those of the breast and uterus, is well established. Estrogen regulates
expression of ~4% of all genes in hormonally-responsive breast cancer cells. Much is known
about ERα-mediated transcriptional activation, but relatively little is known about mechanisms
of ERα-mediated repression, despite that more genes may be down-regulated than up-regulated
by estrogen. We have generated a substantial body of data supporting the hypothesis that ERα is
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 6
also capable of silencing genes, through promoter hypermethylation. Advanced breast cancer
patients often develop resistance to long-term ER blockade therapy, in which case genes
normally silenced by ERα may be de-repressed and contribute to disease progression such as
metastasis. The broad objective of this project is to test the hypothesis that ERα silences genes by
DNA methylation, and that loss of ERα expression as a result of long-term ER blockade therapy,
leads to de-repression of these otherwise silenced genes and progression to more
aggressive/metastatic breast cancer cells.
We have generated two panels of antihormone-resistant human breast cancer cell line models
based on estrogen-responsive ERα-positive T47D and MCF-7 cells. This was achieved by long-
term blockade of ER activity by either estrogen deprivation (ED, ≥ 1 yr) using phenol red-free
media containing charcoal-stripped fetal bovine serum, or by supplementing media with the
antiestrogen fulvestrant (FUL, 100 nM). This process allowed selection of MCF-7 and T47D-
matched antihormone resistant cell lines (MCF7/ED1, MCF7/FUL, T47D/ED2, T47D/ED3, and
T47D/FUL). Importantly, the T47D and MCF-7 cells responded in a completely opposing
manner regarding ERα expression: antihormone-resistant T47D-based ED2, ED3, and FUL cells
almost completely lost ERα expression (20-200 -fold reduction), whereas MCF-7/ED1 and
MCF-7/FUL either overpressed ERα or maintained ERα expression.
We reasoned that loss of ERα in the antihormone-resistant T47D/ED2, T47D/ED3, and
T47D/FUL cells would allow de-repression of genes normally silenced by ERα in wild-type
T47D. However, the continued presence of ERα in the antihormone-resistant MCF-7-based cells,
despite ER blockade, may or may not allow de-repression of ERα-silenced genes. Hence, our
continued studies primarily employed the T47D-based cells.
To identify genes which inversely associated with ERα expression, we performed a gene
expression microarray study comparing wild-type ERα-positive T47D cells to matched ERα-
negative/low T47D/ED2, T47D/ED3 and T47D/FUL cells. Using the method Significance
Analysis of Microarrays (SAM) (Tusher VG, Tibshirani R, Chu G. Proc Natl Acad Sci U S A.
2001;98:5116-21), differentially expressed genes were identified. There were 822 genes
significantly up-regulated and 432 genes down-regulated in all 3 ERα-negative/low cell lines
compared to wild-type T47D cells. As expected, gene expression arrays showed ERα and
numerous other well known ERα-inducible genes were dramatically down-regulated in
T47D/ED2, ED3, and FUL-resistant cells compared to wild-type T47D cells including PGR
(~362-fold decrease across all 3 cell lines), GREB1 (~113-fold decrease) TFF1 (pS2, ~9-fold
decrease), TFF3 (~31-fold decrease) CXCL12 (SDF-1, ~288-fold decrease) and CA12 (~122-
fold decrease). Likewise, known ERα repressible genes were found up-regulated in all three
resistant cell lines including EGF, EGFR, and ERBB2/HER2.
Further examination of changes in gene expression revealed that loss of ERα led to decreased
luminal differentiation markers (KRT18, KRT8), increased basal cell markers (KRT17, KRT6A,
KRT5, CRYAB), increased epithelial-mesenchymal transition (EMT) markers (LCN2,
SNAI2/Slug, PLAU, MMP9, FOXC1, BMP7, TGFB3, RELB, MET, FGFR2, JAG1,
SMO,WNT3A, HEY1, CAV1), and increased cancer/stem cell markers (ALDH1A1, SOX2,
KLF5, IFI16, HAS3, LGR4, CYP26A, FGFR3) in all three resistant T47D-derived cell lines
compared to wild-type cells. Therefore, ERα was a major driver of maintaining luminal cell
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 7
differentiation, and loss of ERα led to de-differentiation / trans-differentiation to a basal-like cell
type characterized by EMT and stem cell markers.
To further discriminate which of the genes up-regulated in all three T47D/ED2, T47D/ED3 and
T47D/FUL2 cell lines that were likely direct ER targets, the T47D/ED2 cells were re-exposed to
1 nM 17β-estradiol (E2) for 24 weeks to derive T47D/ED2/E2 cells. The T47D/ED2 cells were
chosen since they still expressed 5% the level of ERα as in the parental wild-type cells, whereas
the T47D/ED3 cells expressed an even lower level (1.4%) of ERα as parental cells. The
T47D/ED2/E2 cells were examined by gene expression microarrays. Comparing T47D/ED2/E2
to parental T47D/ED2 cells showed 926 genes were significantly up-regulated including PGR,
TFF3, and CA12, indicating that ERα was functional despite its low level. However, GREB1,
TFF1, and CXCL12 were not significantly induced. Several overlapping possibilities exist to
explain the lack of induction of these known E2-responsive genes: 1) ERα levels were below a
threshold level to regulate these genes; 2) these genes were irreversibly silenced; and 3) the
differentiation context of the T47D/ED2 cells changed the spectrum of genes regulated by ERα.
Thus, some but not all well-known ERα-target genes were responsive to E2 in T47D/ED2/E2
cells. Comparison of genes up-regulated in T47D/ED2, T47D/ED3 and T47D/FUL2 cells vs.
wild-type T47D cells, and repressed by E2 in T47D/ED2/E2 cells vs. T47D/ED2 cells identified
253 genes likely directly targeted by ER.
Next, to determine which genes were likely regulated epigenetically by DNA methylation, wild-
type T47D cells were treated for 96 hours with compound vehicle only (control) or with 1 μM
decitabine (DEC; 5-aza-2'-deoxycytidine), which is a cytidine analog that inhibits DNA
methyltransferases. Gene expression microarrays identified 934 genes de-repressed (or induced)
by the hypomethylating agent DEC vs. control. Interestingly, no genes were down-regulated by
DEC according to the SAM method. This was consistent with the expectation that
hypermethylated genes were repressed, and demethylation allowed their de-repression.
Cross-referencing 1) genes up-regulated in T47D/ED2, T47D/ED3 and T47D/FUL2 cells vs.
wild-type T47D cells, 2) genes down-regulated in T47D/ED2/E2 vs. T47D/ED2 cells, and 3)
genes up-regulated by DEC vs. control treatment in wild-type T47D identified 38 genes (Table
1) in common. Therefore, these genes represent high-value candidates silenced by ER-mediated
DNA methylation. There were also 74 genes which were up-regulated in all three ER-
negative/low cell lines and induced by DEC, but not repressed by E2 in T47D/ED2/E2 cells
(Table 2). Some if not most of these 74 genes may yet have been silenced in an ER-dependent
manner in wild-type T47D cells because some expected E2-responsive genes were not induced in
T47D/ED2/E2 cells (i.e. GREB1, TFF1 and CXCL12). Of the 38 high-value candidate ER-
methylated genes (Table 1), LCN2 (lipocalin 2) was the most repressed by E2 in T47D/ED2/E2
cells compared to wild-type T47D cells.
In annual reports from prior years, we characterized time-dependent loss of ERα expression and
concomitant de-repression of LCN2 expression during long-term estrogen deprivation of
T47D/ED2 cells. We also showed by pyrosequencing that LCN2’s promoter near the
transcriptional start site became progressively hypomethylated with increasing time of estrogen
deprivation. In the current year, we verified these results in another independently derived ED-
resistant cell line, T47D/ED3, and in antiestrogen-resistant T47D/FUL cells. We previously
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 8
mapped methylation of CpG sites in the LCN2 promoter and identified CpG sites near the
transcriptional start site (CpGs #9 - 14; Fig. 1A) which showed decreased methylation upon loss
of ERα expression. Again we found CpGs #9 - 14 were hypomethylated in T47D/ED3 and
T47D/FUL cells compared to wild-type T47D cells in a similar manner as T47D/ED2 cells (Fig.
1B). Additionally, long-term E2 treatment of T47D/ED2 cells led to increased CpG methylation
in the LCN2 promoter, most prominently of CpG site #9 (Fig. 1C). We further verified that
LCN2 protein expression progressively increased during long-term estrogen deprivation in
T47D/ED3 cells (Fig. 2A) and during fulvestrant exposure in T47D/FUL cells (Fig. 2B). Thus
LCN2 methylation inversely correlated with its expression and positively correlated with ERα
expression in three independent T47D-derived cell lines (including the T47D/ED2 cells from
prior years). Others have shown that ectopic LCN2 overexpression in MCF-7 cells induced EMT
via increasing expression of the EMT transcription factor Slug (SNAI2) ( Yang J, Bielenberg
DR, Rodig SJ, Doiron R, Clifton MC, Kung AL, et al. Proc Natl Acad Sci U S A.
2009;106:3913-8.). Also in the gene expression microarray studies, Slug was up-regulated in all
three ERα-negative/low T47D-based cell lines compared to wild-type T47D cells. Therefore, we
examined Slug protein expression and found it increased after increases in LCN2 during long-
term estrogen deprivation and fulvestrant treatment in T47D/ED3 and T47D/FUL cells,
respectively (Fig 2A-B). Further, Slug protein expression decreased following decreases in
LCN2 expression during long-term E2 treatment of T47D/ED2/E2 cells (Fig. 2C). These results
are consistent with LCN2 regulating Slug expression.
Taken together, our data suggest a model whereby ERα enforces luminal cell differentiation at
least in part by silencing EMT genes via DNA hypermethylation. These data have important
implications in antihormone-resistant breast cancer. Loss of ERα expression due to long-term
estrogen deprivation or fulvestrant treatment leads to hypomethylation and hence de-repression
of EMT genes such as LCN2 which promotes a shift from luminal to basal-like and stem-like
breast cancer cells with worse clinical outcome.
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 9
Table 1. Candidate genes silenced by ER-mediated DNA methylation.
T47D/ED2
> wt T47D
T47D/ED3
> wt T47D
T47D/FUL
> wt T47D
T47D/ED2/E2
< wt T47D/ED2
DEC > CON
in wt T47D
SYMBOL GENE NAME
Fold
Change
Fold
Change
Fold
Change
Fold
Change
Fold
ChangeLCN2 lipocalin 2 149.51 132.40 42.29 0.04 4.73IFI27 interferon, alpha-inducible protein 27 484.07 327.79 130.79 0.13 5.32SERPINB5 serpin peptidase inhibitor, clade B (ovalbumin), member 5 66.63 4.94 34.13 0.15 2.81BATF2 basic leucine zipper transcription factor, ATF-like 2 30.42 6.88 3.13 0.16 2.96ISG15 ISG15 ubiquitin-like modifier 65.18 27.05 11.75 0.24 4.44SLC15A3 solute carrier family 15, member 3 11.48 8.17 2.24 0.26 5.67LAMP3 lysosomal-associated membrane protein 3 47.65 45.54 27.87 0.27 2.35INHBA inhibin, beta A 17.87 9.35 8.86 0.31 2.64REC8 REC8 homolog (yeast) 6.58 6.02 5.32 0.32 6.67KCNG1 potassium voltage-gated channel, subfamily G, member 1 66.19 9.52 4.75 0.32 4.40MSLN mesothelin 22.81 2.06 4.77 0.33 2.51TRIM29 tripartite motif containing 29 9.41 5.82 19.86 0.34 6.98CES1 carboxylesterase 1 2.76 2.11 3.57 0.35 9.71MAFB v-maf musculoaponeurotic fibrosarcoma oncogene homolog B 6.29 5.02 29.02 0.36 3.33ANKRD65 ankyrin repeat domain 65 4.64 3.21 7.92 0.37 6.99CTNND2 catenin (cadherin-associated protein), delta 2 16.83 8.15 20.77 0.37 3.46DLK2 delta-like 2 homolog (Drosophila) 23.51 19.07 7.69 0.38 4.12CALML5 calmodulin-like 5 6.72 2.59 4.11 0.38 7.11GSTP1 glutathione S-transferase pi 1 9.31 2.45 5.86 0.38 150.12IRF7 interferon regulatory factor 7 15.47 11.52 5.50 0.38 4.65KRT14 keratin 14 249.24 38.23 24.15 0.39 5.46LAMB3 laminin, beta 3 18.43 7.23 19.17 0.40 8.61EFNB2 ephrin-B2 32.14 15.81 5.64 0.43 2.71KRT17 keratin 17 232.56 40.10 22.61 0.43 5.68RAMP1 receptor (G protein-coupled) activity modifying protein 1 19.19 2.67 3.38 0.43 3.25ARAP3 ArfGAP with RhoGAP domain, ankyrin repeat and PH domain 3 28.50 5.90 35.92 0.44 5.87ETV7 ets variant 7 5.51 2.10 2.70 0.45 2.99GPR110 G protein-coupled receptor 110 122.95 30.81 127.97 0.47 2.75SMPD3 sphingomyelin phosphodiesterase 3, neutral membrane 13.65 5.10 11.50 0.47 3.06WNT6 wingless-type MMTV integration site family, member 6 25.26 4.63 4.92 0.48 5.43IGFBP3 insulin-like growth factor binding protein 3 25.53 240.32 2.19 0.50 2.65EPHB2 EPH receptor B2 10.75 12.66 18.73 0.51 2.24NGFR nerve growth factor receptor 11.23 2.40 4.97 0.51 6.18HAS3 hyaluronan synthase 3 7.65 2.86 6.03 0.53 4.84S100A8 S100 calcium binding protein A8 2.57 30.73 35.61 0.53 17.92D4S234E DNA segment on chromosome 4 (unique) 234 expressed sequence 8.54 5.92 10.06 0.56 4.76ANXA8L2 annexin A8-like 2 187.37 10.86 183.42 0.56 4.96GPR87 G protein-coupled receptor 87 19.18 10.36 40.53 0.59 3.10
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 10
Table 2. Additional methylated genes de-repressed by long-term ER blockade in three T47D-
based cell lines but not repressed by long-term E2 in T47D/ED2/E2 vs. T47D/ED2 cells.
T47D/ED2
> wt T47D
T47D/ED3
> wt T47D
T47D/FUL
> wt T47D
DEC > CON
in wt T47D
SYMBOL GENE NAME
Fold
Change
Fold
Change
Fold
Change
Fold
Change
CIB2 calcium and integrin binding family member 2 43.01 58.40 44.74 17.23MRC2 mannose receptor, C type 2 3.79 2.06 8.16 16.86CHST6 carbohydrate (N-acetylglucosamine 6-O) sulfotransferase 6 2.91 1.54 1.84 16.66S100P S100 calcium binding protein P 297.56 199.74 154.11 13.30GTSF1 gametocyte specific factor 1 22.71 25.77 3.37 12.76TPM4 tropomyosin 4 2.58 2.37 2.53 12.37FOXJ1 forkhead box J1 16.27 4.04 2.18 10.71S100A9 S100 calcium binding protein A9 22.85 44.14 37.74 8.27GPRC5B G protein-coupled receptor, family C, group 5, member B 16.46 22.41 4.10 8.25ANXA1 annexin A1 72.91 30.03 44.92 7.96TGFBI transforming growth factor, beta-induced, 68kDa 9.95 2.72 6.91 7.86PLAU plasminogen activator, urokinase 131.30 81.55 141.63 7.23TPM2 tropomyosin 2 (beta) 3.93 3.97 2.34 6.62KCNN4 potassium intermediate/small conductance calcium-activated channel,subfamily N, member 4 4.26 1.77 30.79 5.96RINL Ras and Rab interactor-like 3.63 3.41 3.51 5.92PCDH7 protocadherin 7 22.26 44.67 12.51 5.85PGF placental growth factor 2.79 3.79 3.16 5.68CRYAB crystallin, alpha B 47.79 35.88 10.40 5.27FBLN1 fibulin 1 11.05 8.26 6.35 5.07LOC440335 uncharacterized LOC440335 3.72 7.98 2.86 4.98SH3PXD2A SH3 and PX domains 2A 11.58 4.32 15.47 4.23IFI30 interferon, gamma-inducible protein 30 4.46 5.81 2.44 4.19RAB36 RAB36, member RAS oncogene family 4.34 7.69 3.22 4.00TNNC1 troponin C type 1 (slow) 2.76 2.05 3.67 3.87BMP7 bone morphogenetic protein 7 73.95 51.67 118.52 3.73OSTalpha organic solute transporter alpha 48.94 44.06 16.67 3.72LRG1 leucine-rich alpha-2-glycoprotein 1 11.08 23.59 6.45 3.69MMP9 matrix metallopeptidase 9 (gelatinase B, 92kDa gelatinase, 92kDa type IV collagenase) 3.39 2.22 1.63 3.65S100A2 S100 calcium binding protein A2 47.05 2.52 19.81 3.56SLC37A2 solute carrier family 37 (glycerol-3-phosphate transporter), member 2 10.36 13.96 7.97 3.42HOXB6 homeobox B6 6.95 7.37 2.48 3.34ARHGDIB Rho GDP dissociation inhibitor (GDI) beta 4.74 9.24 19.32 3.33PSCA prostate stem cell antigen 119.33 19.61 479.57 3.32TRIM2 tripartite motif containing 2 9.05 4.40 5.88 3.25C15orf33 chromosome 15 open reading frame 33 2.75 1.89 1.80 3.24NTN4 netrin 4 2.80 3.58 9.99 3.23PLBD1 phospholipase B domain containing 1 5.48 7.38 7.42 3.23TM4SF1 transmembrane 4 L six family member 1 9.72 4.05 71.00 3.18KRTAP3-1 keratin associated protein 3-1 3.16 7.98 3.43 3.06ID4 inhibitor of DNA binding 4, dominant negative helix-loop-helix protein 2.15 3.57 1.75 3.04WNT3A wingless-type MMTV integration site family, member 3A 3.39 4.88 7.14 3.02ABCA1 ATP-binding cassette, sub-family A (ABC1), member 1 9.60 4.65 18.77 3.02NMB neuromedin B 2.78 1.87 2.00 2.96FAM27A family with sequence similarity 27, member A 1.76 2.89 1.59 2.94MIR205HG MIR205 host gene (non-protein coding) 3.61 2.83 3.06 2.92PRRX2 paired related homeobox 2 7.45 1.72 3.89 2.85TMEM37 transmembrane protein 37 2.97 2.56 1.90 2.85LMO2 LIM domain only 2 (rhombotin-like 1) 35.03 22.87 23.31 2.83MYCL1 v-myc myelocytomatosis viral oncogene homolog 1, lung carcinoma derived (avian) 6.14 12.60 11.30 2.81RCOR2 REST corepressor 2 6.53 4.73 2.98 2.81MDK midkine (neurite growth-promoting factor 2) 10.61 7.81 6.34 2.80TMEM173 transmembrane protein 173 5.76 2.28 2.54 2.71OLFML2A olfactomedin-like 2A 4.85 4.10 2.28 2.71FAM59B family with sequence similarity 59, member B 2.86 2.28 1.94 2.69TUBB8 tubulin, beta 8 class VIII 1.94 2.27 1.77 2.68MALL mal, T-cell differentiation protein-like 37.01 36.93 124.51 2.64MAGEA10 melanoma antigen family A, 10 3.87 2.04 6.09 2.63EMP1 epithelial membrane protein 1 25.55 2.91 4.36 2.59CD68 CD68 molecule 2.43 3.12 2.35 2.52ELF3 E74-like factor 3 (ets domain transcription factor, epithelial-specific ) 7.03 7.17 6.61 2.48CPNE2 copine II 5.66 5.16 3.08 2.43MET met proto-oncogene (hepatocyte growth factor receptor) 23.08 14.47 7.11 2.42LARP6 La ribonucleoprotein domain family, member 6 3.88 3.45 1.69 2.42HSD17B11 hydroxysteroid (17-beta) dehydrogenase 11 5.32 3.00 3.62 2.42CYB5R2 cytochrome b5 reductase 2 2.24 1.85 2.02 2.40FGFR3 fibroblast growth factor receptor 3 2.91 3.74 3.80 2.40SMOX spermine oxidase 6.97 2.85 1.54 2.37TMPRSS4 transmembrane protease, serine 4 28.45 4.10 8.65 2.36PALM3 paralemmin 3 4.45 2.88 1.88 2.36MFAP3L microfibrillar-associated protein 3-like 4.11 4.60 3.24 2.35EFEMP2 EGF containing fibulin-like extracellular matrix protein 2 1.92 2.17 1.88 2.34TMEM40 transmembrane protein 40 4.82 4.49 8.58 2.34HEY1 hairy/enhancer-of-split related with YRPW motif 1 313.48 71.50 4.25 2.33HTRA1 HtrA serine peptidase 1 6.05 4.97 4.56 2.23
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 11
Figure 1. DNA methylation of the LCN2 promoter. (A) Map of the LCN2 promoter. CpG sites are
indicated. (B) Hypomethylation of the LCN2 promoter in ERα-negative/low expressing T47D/ED2,
T47D/ED3, and T47D/FUL compared to wild-type (wt) T47D cells. Methylation of CpG sites #9 -
#14 decreased in all three ERα-negative/low cell lines compared to wt ERα-positive cells. (C)
Increased methylation of the LCN2 promoter due to long-term E2 treatment in ERα-low T47D/ED2
cells. CpG Methylation analysis of CpG sites #9-#14 was determined by pyrosequencing of bisulfite-
treated DNA.
Figure 2. LCN2and Slug (SNAI2) protein expression inversely correlates with (A-B) ERα
expression or (C) E2 stimulation. As ERα expression decreased, first LCN2 and then Slug expression
increased in (A) T47D/ED3 and (B) T47D/FUL cells. (C) Long-term E2 treatment of T47D/ED2/E2
cells led to progressive decreases first in LCN2 and then in Slug expression. Protein levels were
measured by immunoblotting and visualized using the Odyssey Infrared Imaging System (Li-Cor
Biosciences).
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 12
Research Project 3: Project Title and Purpose
Markers of Adult Epithelial Stem Cells - The purpose of this project is to identify novel stem cell
markers in adult epithelial tissues. Currently, no markers are available that consistently and
robustly enable the in vivo identification and purification of adult stem cells from normal tissues
or tumors. Moreover, the difficulty associated with in vivo identification of stem cells has
hampered attempts to conduct functional analyses of genes thought to regulate stem cell function
or contribute to the formation of tumor cells with self-renewing properties. We aim to determine
the molecular identity and function of a new epithelial stem cell marker in the fly and assess its
functional conservation in mammalian epithelial tissues. The information provided in this study
will be a foundation for the development of agents that promote or target stem cell function for
use in replacement therapies and anti-cancer treatment.
Anticipated Duration of Project
1/1/2010 – 12/31/2013
Project Overview
Substantial progress has been made in the identification of markers used to isolate and identify
stem cells from adult tissues. In some cases, markers are relatively stem-cell specific (e.g. Oct4).
In other cases, broadly expressed proteins exhibit differential levels of expression in stem cells
vs. non-stem cell daughters. Despite these advancements, purification of adult tissue stem cells,
identification of stem cells in situ, and the functional relevance of stem cell markers currently
used in isolation protocols remain challenges for this field.
Our goal is to identify molecules that control stem cell identity and maintenance. Since
identifying novel signals that regulate stem cell function is technically challenging in mammalian
tissues, we are focusing on Drosophila ovarian Follicular epithelium Stem Cells (FSCs) as a
model system for this analysis. In this system, epithelial stem cell formation can be visualized
directly in vivo. Moreover, genetic mutational analysis allows us to pinpoint functional roles for
specific genes in the cellular events required for stem cell commitment and self-renewal. The
striking conservation of stem cell control signals in mammals and flies suggests that novel stem
cell regulatory mechanisms identified in the fly system will generally be applicable for
understanding epithelial stem cell regulation in mammalian tissues.
To identify novel stem cell regulators, we have initiated a screen for antibodies that specifically
recognize FSCs in situ. Using this approach, we identified one antibody that recognizes a
nuclear protein that is highly upregulated in FSCs and their differentiated progeny cells. We
propose that this follicle cell-nuclear antigen (FC-NA) is a novel epithelial stem cell marker that
is important for promoting FSC commitment and function. To test this hypothesis, we will 1)
determine the molecular identity of FC-NA using biochemical approaches and molecular
genetics, 2) define how FC-NA regulates FSCs in vivo through mutational analysis and imaging,
and 3) define the protein expression patterns of FC-NA in mammalian epithelial tissues using
immunostaining techniques. This project has the potential to uncover novel stem cell markers
with broad application in stem cell purification and identification as well as in functional analysis
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 13
of normal and tumor stem cells.
Principal Investigator
Alana M. O’Reilly, PhD
Assistant Professor
The Fox Chase Cancer Center
333 Cottman Avenue
Philadelphia PA 19111-2434
Other Participating Researchers
None
Expected Research Outcomes and Benefits
The promise of stem cell research is two-fold. On the one hand, stem cell replacement therapies
may provide cures for a diverse group of disorders, including cancer, diabetes, and traumatic
injury. On the other hand, the identification of cancer cells with self-renewing properties that
promote tumor growth and evade conventional anti-cancer therapies may lead to the
development of potent new approaches to cancer treatment. Substantial progress has been made
in the development of in vitro technologies that allow stem cells to be grown in culture and
stimulated with factors that promote their differentiation into specific cell types. Far less
progress has been made in identifying adult stem cells and their mechanisms of regulation in
functioning adult tissues.
In this project, we aim to identify new markers of adult epithelial stem cells. Our preliminary
results indicate that an unidentified nuclear protein is highly upregulated in epithelial stem cells
in the Drosophila ovary, suggesting that this protein may be an important new stem cell marker.
If the results of the experiments proposed support our hypothesis that this nuclear antigen is a
novel regulator of epithelial stem cell function in flies and in mammals, a new tool for the
identification, purification, and in vivo analysis of stem cells will be available. In addition, this
study has clear implications for understanding the role of specific genes in promoting stem cell
self-renewal and, importantly in the genesis of epithelial stem cells from uncommitted precursor
cells. Conservation of the normal developmental mechanisms described in this study may have
an important impact on the ability to recognize newly forming cancer cells with stem cell
properties. Such a discovery would open new therapeutic avenues for treating cancer in its
earliest stages. Finally, this tool may advance our understanding of the mechanisms regulating
stem cells in their normal environments, information that is critical for the success of stem cell
replacement therapies.
Summary of Research Completed
The goal of this work is to identify novel markers of epithelial stem cells. During the past year,
we focused primarily on how the Hedgehog (Hh) signaling pathway controls epithelial stem cell
proliferation. This month, we submitted a manuscript describing our identification of
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 14
environmental signals that promote Hh signaling to control proliferation of follicle stem cells, an
epithelial stem cell population in the Drosophila (fruit fly) ovary. Currently, we are evaluating
conservation of our recently defined mechanism in maintaining human cancers and protecting
them from drug treatments. In addition, we are in the process of completing characterization of
two additional epithelial stem cell markers that we identified in previous screens and continue
screening for novel genes that control the stem cell response to Hh signaling. Our overall goal is
to develop reagents that will enable the unambiguous identification of epithelial stem cells in
vivo so that determining the mechanisms of action of specific genes and the tracking of
developing cancer stem cells will become possible.
Epithelial stem cell proliferation is controlled by the environment
This past year, we focused on defining how Hedgehog molecules are released from Boi. We
reasoned that changes in environmental factors that influence egg production might contribute to
control of Hedgehog release and follicle stem cell proliferation. Changes in food availability,
mating, and light cycle all contribute to control of egg laying in flies, suggesting that any of these
environmental factors might trigger signals that promote egg production. After careful analysis,
we found that a specific nutrient in fly food affects the ability of Boi to sequester and release
Hedgehog molecules and defined the series of proteins that controls this event. This manuscript
has been submitted for publication this past month. Conservation of this mechanism for
proliferation control of mammalian stem cell populations may have important implications for
studies on aging and cancer, where the benefits of low calorie and low cholesterol diets are
beginning to emerge. Moreover, manipulation of growth factor levels in the stem cell niche via
controlled sequestration and release may improve strategies for regenerating aging or damaged
tissues or inhibiting the availability of growth factors to tumor cells with stem cell properties.
To continue to advance our understanding of the role of Hedgehog signaling in epithelial stem
cell regulation, we have initiated studies designed to define how Hedgehog molecules are
delivered from the Boi-expressing cells to the follicle stem cell niche and identify proteins that
are important for the response to Hedgehog signaling within the stem cells. A genetic screen has
already been initiated in our lab that has uncovered two novel suppressors and one novel
enhancer of Hedgehog signaling within follicle stem cells. Defining the mechanism of action of
these genes in follicle stem cell control will be a major focus for the upcoming year.
Mechanistic conservation of Hh sequestration and release in human cancer
In the past year, we conducted experiments designed to define the role of the Hedgehog pathway
in epithelial ovarian cancer, the most common and most deadly type of malignant ovarian cancer.
Although components of the Hedgehog pathway are expressed in the ovarian surface epithelium,
our analysis failed to reveal a critical role for Boi homologs or Hedgehog sequestration in the
etiology or progression of epithelial ovarian cancer in a transgenic mouse model. In contrast to
ovarian cancer, where roles for Hedgehog signaling have not been clearly defined previously,
cancers of the skin, pancreas, prostate, and brain all depend on Hedgehog for their maintenance
and progression. Currently, we are expanding our analysis of the role of Hedgehog sequestration
and release in control of additional epithelial cancers, with a focus on the expression patterns and
functions of the mammalian homologs of Boi, Cell adhesion molecule-related/down-regulated
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 15
by oncogenes (Cdon, pronounced “kiddo”) and Brother-of-Cdon (BOC) in tumor cells and the
surrounding stroma.
Novel stem cell markers
In addition to our recently completed study on the role of external factors in regulation of
Hedgehog sequestration and release, we have molecularly defined a novel follicle stem cell
marker and assessed its functional role in follicle stem cell proliferation and maintenance of
morphology. A manuscript describing these findings is currently in preparation. Moreover, our
attempts to partially purify FC-NA, an antigen that is highly expressed in follicle stem cells that
was the initial focus of this proposal, yielded a short list of candidate antigens. We have
observed clear defects in follicle stem cell proliferation and morphology upon mutation of our
top candidate, suggesting that this gene encodes FC-NA. We are in the process of confirming
the molecular identification of FC-NA and its mechanism of action in follicle stem cell
proliferation control. Together with the genetic screen described above, we hope to define new
epithelial stem cell markers and assess their roles in stem cell regulation in the fly and in
mammalian tissues.
Research Project 4: Project Title and Purpose
Role of a Novel Zinc Finger Factor in Hematopoietic Development - The purpose of this project
is to understand the role of a novel erythroid factor (G4) in hematopoietic development using the
zebrafish model system. Since the molecular basis for G4 function in controlling erythropoiesis
is completely unknown, we will determine where g4 fits in the hierarchy of transcription factors
controlling hematopoietic development, identify its human ortholog, and perform
structure/function analysis to identify the domains of g4 that are essential for its function. Results
from these studies will provide insights into how normal blood cell development is controlled
and how those processes are perturbed in cancer.
Duration of Project
1/1/2010 - 12/31/2011
Project Overview
A network of transcriptional regulators coordinates the cell fate choices and progressive
differentiation of hematopoietic stem cells, a process that is incompletely understood.
Hematopoiesis is conserved throughout vertebrate evolution, and we have employed the facile
zebrafish experimental model system to identify a novel molecular effector, called g4, that is
critical for erythroid development. G4 is part of a cluster of highly homologous genes in
zebrafish, and the corresponding human gene cluster encodes zinc finger factors of unknown
function. The objective of this project is to investigate the role of g4 in hematopoietic
development using the zebrafish model system. We intend to do so according to the following
aims:
Aim 1. Use epistasis analysis to determine where g4 function is positioned within the hierarchy
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 16
of known regulators of erythroid development. This will be accomplished by performing
morpholino knockdown of known molecular effectors of primitive hematopoiesis (g4, scl, gata1,
pu.1), determining how this affects g4 expression, and assessing whether forced expression of g4
mRNA rescues the developmental blockade.
Aim 2. Utilize a complementation strategy to investigate the specificity of function among g4
family members and identify human paralogs. We will knock down g4 expression and determine
if forced expression of the other members of the zebrafish gene cluster or the human paralogs are
able to restore development. Because the zebrafish family members are highly similar, this may
inform the structure/function analysis if differences among family members are identified.
Aim 3. Perform structure/function analysis to identify domains or motifs of g4 that are essential
for function. We will do so by determining if mutant g4 constructs lacking conserved domains or
motifs are able to rescue the developmental arrest caused by g4 knockdown.
Together, these studies will provide insight into the function of g4 in normal development of
erythroid precursors as well as to how its misregulation might contribute to their transformation.
Principal Investigator
Jennifer Rhodes, PhD
Assistant Professor
The Fox Chase Cancer Center
333 Cottman Avenue
Philadelphia, PA 19111-2434
Other Participating Researchers
Wittaya Pimtong, PhD – employed by the Rhodes lab, Fox Chase Cancer Center
Expected Research Outcomes and Benefits
Defects in the normal ontogeny of the hematopoietic system can contribute to the pathogenesis
of these cells, leading to blood cell cancers or anemia. Furthermore, the continued advancement
of molecularly targeted therapies to treat hematopoietic diseases is reliant on the identification of
new, relevant factors and our understanding of the underlying biology. This study focuses on the
in vivo function of a novel erythroid factor, and other highly homologous family members
localized adjacent to one another in the genome. The potential importance of these genes in
hematopoietic disease is evidenced by amplification of the locus encompassing this gene cluster
in human hematologic cancers. Thus, results from this project will reveal insights into the
function of a novel factor that regulates erythroid development and may also reveal a potential
molecular target or generate animal models that could be exploited for diagnostic and therapeutic
purposes.
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 17
Summary of Research Completed
The summary below describes the research completed from July 1, 2011 through December 31,
2011.
Aim 1. Use epistasis analysis to determine where g4 function is positioned within the hierarchy
of known regulators of erythroid development.
We last reported that since g4 function has never been examined and the pu.1-gata1 regulatory
network is highly conserved, the identification of g4 as a new participant in this genetic pathway
is significant and we were focused on completing the aim 1 analyses for the purpose of
publication. A manuscript resulting from our investigation has been prepared for publication and
is undergoing final editing prior to submission.
In total, we discovered that g4 is essential for embryonic erythropoiesis and is transiently
necessary for embryonic myeloid cell development; however, it was not clear how g4 fit in to the
established hierarchy of hematopoietic regulators. Our approach to address this question was to
perform antisense morpholino knockdown of known molecular effectors of primitive
hematopoiesis and determine how this affects g4 expression, and assess whether forced
expression of g4 mRNA rescues the developmental blockade. Here, our studies focus on the
interplay of the novel gene g4 with the pu.1-gata1 axis regulating hematopoiesis. g4, gata1 and
pu.1 are all expressed in developing hematopoietic progenitor cells. Moreover, mammalian and
zebrafish models have been used to establish essential roles for pu.1 in white blood cell
development and gata1 in red cell development, defining a cross-antagonistic function for these
factors in determining hematopoietic progenitor cell fate. This is evidenced by morpholino
knockdown of zebrafish gata1 which blocks hematopoietic progenitor cell differentiation to the
erythroid lineage and instead causes the progenitor cells to transfate to the myeloid lineage.
Similarly, pu.1 morpholino knockdown causes a loss of myeloid cells and transfates the
progenitor cells to the erythroid lineage. Erythropoiesis occurs exclusively in the posterior
hematopoietic region in the zebrafish embryo, while myelopoiesis occurs in both the posterior
and anterior hematopoietic compartments. pu.1 morpholino-injected (morphant) embryos display
ectopic erythroid development in the anterior of the embryo. The reciprocal antagonism between
pu.1 and gata1 is conserved between vertebrates and the regional compartmentalization of
myeloid and erythroid development make zebrafish a very convenient model to examine the role
of g4 in this pathway.
Our work found that the knockdown of gata1 leads to decreased numbers of g4-expressing cells.
However, in this reporting period, we found that enforced expression of g4 is not sufficient to
rescue the erythroid deficiency in gata1-depleted embryos. Likewise, depletion of g4 leads to
decreased expression of g4 and gata1. While the erythroid and myeloid defects in g4-deficient
embryos (morphants) could be reverted back to a wild-type state by co-injecting g4 mRNA with
the g4 morpholino, we found that enforced expression of gata1 was not sufficient to rescue g4
morphants. These data indicate that gata1 and g4 do not participate in a linear pathway and
suggest that these factors have functions in addition to supporting the expression of each other.
Turning to the myeloid lineage, embryos depleted of pu.1 displayed increased numbers of
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 18
embryonic g4-expressing cells that was consistent with the increased erythropoiesis observed in
the pu.1 morphants. Thus, g4 expression tracks closely with the development of erythroid cells.
In total, since stem cell markers are not affected by the knockdown of g4, these results suggest
that pu.1-gata1 govern the fate of cells that express g4, however g4 functions in a linear pathway
or upstream of the molecular axis controlling progenitor cell fate decisions. Since the g4
morphants display a paucity of both embryonic myeloid and erythroid cell, one possibility is that
g4 is essential for progenitor cell survival. However, using acridine orange staining and antibody
detection of activated Caspase 3 to detect dead/dying cells we found that the g4 morphants were
indistinguishable for control embryos, indicating that loss of g4 does not induce cell death and
suggesting an alternate mechanism underlies the essential role of g4 in hematopoietic
development.
Aim 2. Utilize a complementation strategy to investigate the specificity of function among g4
family members and identify human paralogs.
To address this aim we planned to knock down g4 and determine if forced expression of the
other members of the zebrafish gene cluster or the human paralogs are able to restore
development. Indeed, depletion of g4 leads to decreased expression of the other gene family
members; however, enforced expression of the founding member of this family did not rescue
the erythroid or myeloid cell deficiencies in g4 morphants. This indicates that g4 has a specific
and essential role in hematopoiesis. This also indicates that the three regions of G4 that are not
homologous to the other gene family members have important roles in the function of the
protein.
Next, we turned our focus to determining whether the potential human homologs are functionally
equivalent to g4 in zebrafish. Previously, we reported generating in vitro transcribed mRNA of
g4 and the two human paralogs to determine whether forced expression of these genes can rescue
the defects observed in g4-deficient embryos. In this reporting period, as stated above, we found
that co-injection of g4 mRNA (not targeted by the morpholino) with the g4-targeting morpholino
(which targets endogenous transcripts) resulted in wild-type embryos. This result indicates that
the defects induced by g4 knockdown are due to decreased activity of this gene. By contrast,
forced expression of the putative human homologs did not alter the g4 deficiency-induced
defects. This may be due to species-specific functions of g4 or that the functional equivalent of
g4 was not correctly identified in our preliminary analysis. To better address the latter
possibility, we consulted with experts in the FCCC Bioinformatics and Molecular Modeling
Facilities. Zebrafish G4 contains multiple consecutive Cys2-His2 (C2H2) zinc-finger domains
and the gene is strongly expressed in embryonic hematopoietic stem and progenitor cells, but
expression is barely detectable in mature myeloid cells. We used bioinformatics to identify
human genes that had similar patterns of expression and were homologous to G4 protein. We
identified 15 putative human g4 homologs. Synteny, homology, and expression analyses did not
make obvious which one of these genes is most likely to be the g4 homolog, but several
candidates reside in chromosomal regions that can be found amplified in acute myeloid (AML)
and megakaryoblastic leukemias. Oncomine expression data indicates that many of the putative
g4 human homologs have aberrant expression levels in a variety of leukemias, although
expression levels were often normal in pre-leukemic (MDS) samples. The pathogenic
contributions of the putative g4 homologs have not been examined. A National Institutes of
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 19
Health grant has been submitted in part to examine the activity of this refined cohort of g4
homologs.
Aim 3. Perform structure/function analysis to identify domains or motifs of g4 that are essential
for function.
Our ability to evaluate the activity of wild-type g4 is essential for aim 3 studies. To do this, we
have developed in vitro transcriptional reporter assays, in vivo rescue analysis (described above)
and generated a novel antibody that recognizes the zebrafish G4 protein. In this reporting period,
we have successfully used the G4 antibody to detect the protein using western analysis of a cell
line transfected to express the protein and in whole mount zebrafish embryos, although the use of
confocal microscopy to determine the protein localization within cells is ongoing. The initial in
vitro transcriptional reporter assays were done using a reporter construct in which a short
fragment upstream of the g4 gene was controlling expression of a luciferase reporter gene.
Unfortunately, we obtained variable results using this construct. We subsequently subcloned the
entire region upstream of the ATG containing exon of g4 into the luciferase reporter vector.
Preliminary studies indicate that g4 can activate transcription from this reporter, however a more
complete study is necessary to better establish this assay. Concurrently, we have confirmed that
we can detect protein following transfection of the expression constructs (G4 and Gata1) and
have begun to generate deletion derivatives that remove the G4-specific regions, identified in
aim 2.
Research Project 5: Project Title and Purpose
Reducing the Burden of Hepatocellular Carcinoma: Identification of New Targets against
Hepatitis B virus (HBV) for Drug Development – A reduction in the burden of liver cancer
depends on the development of effective therapies that prevent or arrest the growth of
hepatocellular carcinomas. The purpose of the project is i) to develop rapid screening assays for
primary cultures derived from hepatocellular carcinoma for drugs and drug combinations that
induce cell death and to identify serine and tyrosine kinases that are required for the survival of
tumor cells in the presence and absence of already available anticancer drugs and ii) to identify
novel targets for antiviral therapy against hepatitis B virus.
Duration of Project
1/1/2010 – 6/30/2012
Project Overview
The principal research objective of this project is to reduce the burden of hepatocellular
carcinoma (HCC). HCC is the third leading cause of cancer mortality worldwide. So far,
treatment options for patients with HCC are sparse and available medicines exhibit insufficient
therapeutic efficacy to significantly prolong life. The overall goal of this project to develop a
method for the rapid identification of new targets for the development of drugs that can prevent,
arrest growth of HCC or even cure the disease. We will test the hypothesis that the survival of
liver tumors depends on the expression of genes belonging to the cellular serine and tyrosine
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 20
kinases and that inactivation of certain kinases will arrest tumor growth by increasing the rate of
cell death of liver tumor cells. This hypothesis is supported by observations made with certain
breast and lung cancers as well as by observations in our laboratory with an established cell line
derived from a cervical carcinoma. Moreover, we will test the hypothesis that hepatitis B virus
replication depends on cellular enzymes required for nucleotide excision repair of damaged
DNA. If successful, our investigations will identify novel cellular targets for the development of
antiviral therapies against hepatitis B virus, a major cause for liver cancer in the world.
Our goal to identify kinases required for survival of liver tumors will be achieved through two
specific research aims: 1) to develop conditions for the screening of primary cell cultures
derived from human liver cancers and 2) to use RNA interference technology to identify kinases
that are required for survival of liver tumor cells and 3) to identify cellular enzymes required for
the life cycle of hepatitis B virus.
Principal Investigator
Christoph Seeger, PhD
Professor
The Fox Chase Cancer Center
333 Cottman Avenue
Philadelphia, PA 19111-2497
Other Participating Researchers
Ronald Matteotti, MD – employed by The Fox Chase Cancer Center
Expected Research Outcomes and Benefits
The expected outcome of this project is the development of conditions to establish primary cell
cultures from primary hepatocellular carcinomas and the identification of serine or tyrosine
kinases that are essential for the development, growth and survival of these cancers. In addition,
an expected outcome of this project is the identification of an endonuclease, polymerase and/or
ligase required for hepatitis B virus replication. The benefits of the research project will be the
identification of new targets for the development of drugs that in combination with already
available medicines can prevent, arrest or cure liver cancers, and increase the survival rate of
patients struck with this deadly disease.
Summary of Research Completed
Aim 3.1: Identification of the polymerase required for CCC DNA synthesis.
We have continued our efforts to identify the polymerase required for the synthesis of covalently
closed circular (CCC) DNA in hepatitis B viruses (Figure 1). We confirmed our previous
observation indicating a role for the Y-DNA polymerase, DNA polymerase kappa (pol ) in the
formation of CCC DNA of duck hepatitis B virus (DHBV). Inhibition of expression of the
polymerase with several shRNAs led to a significant reduction in CCC DNA formation in 293T
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 21
cells. To validate these observations, we set out to create a HEK293T cell line that is deficient in
the expression of polκ. To accomplish this goal we purchased a zinc finger nuclease (ZFN) that
specifically cleaves the human polκ gene. With the help of this reagent, we were successful in
isolating two cell clones (293P1, 293P2) with mutations in both alleles of the polκ gene. To
investigate whether CCC DNA could be formed in these clones, we transfected them together
with wildtype 293T cells with a DHBV expression vector and determined the levels of
cytoplasmic and nuclear viral DNA. The results showed that CCC DNA could still accumulate in
the mutant cell line, albeit at lower levels compared to the parental cells. While these results
were consistent with our previous observation resulting from experiments with shRNAs, they
also indicated that other polymerases could carry out CCC DNA formation or substitute for polκ
in polκ-deficient cells.
We have initiated experiments to identify additional polymerases that play a role in CCC DNA
synthesis. We continued to focus on Y-DNA polymerases, because we confirmed that CCC
DNA synthesis is resistant to aphidicolin, an inhibitor of B-family polymerases (i.e. polymerases
delta, epsilon, zeta). Moreover, we found that CCC DNA was resistant to treatment of cells with
hydroxyurea, which is known to inhibit ribonucleotide reductase leading to a lower concentration
of the nucleotide pool, which inhibits the enzymatic activity of B-, but not Y-family DNA
polymerases. We have purchased siRNAs to inhibit expression of the two remaining Y-family
DNA polymerases, eta and iota that could substitute for polκ in our mutant cell lines.
Experiments to determine whether suppression of more than one Y-DNA polymerase inhibits
CCC DNA synthesis are underway.
Aim 3.2: Identification of the exonuclease and DNA ligase required for CCC DNA synthesis.
Using the same approach as described for the polymerase, we continued experiments aimed at
identification of the relevant endonuclease and DNA ligase required for the repair process. While
we have obtained evidence for a role of an endonuclease in CCC DNA formation, our efforts to
identify the relevant DNA ligase have not yet yielded firm results. The mammalian genome
encodes only three known DNA ligases, DNA ligase I, III and IV. We have been able to inhibit
expression of each of the three ligases with shRNA and siRNAs, but could not demonstrate
inhibition of CCC DNA synthesis, even when we suppressed two ligases simultaneously (Figure
2). These results were unexpected. It is conceivable that suppression of one or all ligases was not
complete. Moreover, it is known that DNA ligases I and III can substitute for each other. To
address these problems, we will be using a gene knockout approach as we have used for DNA
polκ. . We are in the process of constructing Talen nucleases specific for each of the three DNA
ligases.
_____________________________________________________________________________________________
Pennsylvania Department of Health – 2011-2012 Annual C.U.R.E. Report
Fox Chase Cancer Center – 2009 Formula Grant – Page 22
Figure 1. Model for CCC DNA formation. A) Physical
structure of relaxed circular (RC) and CCC DNA. B) The
figure depicts the three enzymatic steps required for the
formation of CCC DNA. Filled circle; viral reverse
transcriptase, half circle; capped RNA primer, X;
cleavage site for endonuclease.
Figure 2. DHBV CCC DNA synthesis in 293T cells. A) Cells were
transfected with siRNAs against DNA ligases I,II and IV. The left panel
shows cytoplasmic core DNA (RC, DSL,SS), the right panel nuclear CCC
DNA The inset shows a western blot developed with Ab to each of the
three DNA ligases. B) Same as A, except that 293T cells were infected
with a lentivirus vector expressing shRNAs against DNA ligase IV prior to
transfection with DHBV DNA.