You are on page 1of 16

G Model

SBMB 4721 No. of Pages 16

Journal of Steroid Biochemistry & Molecular Biology xxx (2015) xxxxxx

Contents lists available at ScienceDirect

Journal of Steroid Biochemistry & Molecular Biology


journal homepage: www.elsevier.com/locate/jsbmb

Review

Stem cells and the role of ETS transcription factors in the differentiation
hierarchy of normal and malignant prostate epithelium
Leanne K. Archer, Fiona M. Frame, Norman J. Maitland*
Cancer Research Unit, Department of Biology, University of York, York, YO10 5DD, United Kingdom

A R T I C L E I N F O A B S T R A C T

Article history:
Received 19 February 2016 Prostate cancer is the most common cancer of men in the UK and accounts for a quarter of all new cases.
Received in revised form 25 April 2016 Although treatment of localised cancer can be successful, there is no cure for patients presenting with
Accepted 7 May 2016 invasive prostate cancer and there are less treatment options. They are generally treated with androgen-
Available online xxx ablation therapies but eventually the tumours become hormone resistant and patients develop
castration-resistant prostate cancer (CRPC) for which there are no further successful or curative
Keywords: treatments. This highlights the need for new treatment strategies. In order to prevent prostate cancer
Prostate cancer recurrence and treatment resistance, all the cell populations in a heterogeneous prostate tumour must be
Cancer stem cells
targeted, including the rare cancer stem cell (CSC) population. The ETS transcription factor family
ETS factors
members are now recognised as a common feature in multiple cancers including prostate cancer; with
Treatment resistance
aberrant expression, loss of tumour suppressor function, inactivating mutations and the formation of
fusion genes observed. Most notably, the TMPRSS2-ERG gene fusion is present in approximately 50% of
prostate cancers and in prostate CSCs. However, the role of other ETS transcription factors in prostate
cancer is less well understood. This review will describe the prostate epithelial cell hierarchy and discuss
the evidence behind prostate CSCs and their inherent resistance to conventional cancer therapies. The
known and proposed roles of the ETS family of transcription factors in prostate epithelial cell
differentiation and regulation of the CSC phenotype will be discussed, as well as how they might be
targeted for therapy.
2016 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... .......... . . . . . . . . . . . . . . . . . . . . . . . 00


2. Prostate cancer stem cells: the evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... .......... . . . . . . . . . . . . . . . . . . . . . . . 00
2.1. Epithelial composition of the prostate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... .......... . . . . . . . . . . . . . . . . . . . . . . . 00
2.2. Evidence suggests that a basal cell is the cell of origin of prostate cancer . . . . . . . . . ...... .......... . . . . . . . . . . . . . . . . . . . . . . . 00
3. Cancer stem cells and treatment resistance in prostate cancer . . . . . . . . . . . . . . . . . . . . . . . . ...... .......... . . . . . . . . . . . . . . . . . . . . . . . 00
3.1. Drug efux transporters and anti-apoptotic molecules . . . . . . . . . . . . . . . . . . . . . . . . ...... .......... . . . . . . . . . . . . . . . . . . . . . . . 00
3.2. DNA damage response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... .......... . . . . . . . . . . . . . . . . . . . . . . . 00
3.3. The tumour microenvironment and CSC niche . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... .......... . . . . . . . . . . . . . . . . . . . . . . . 00
3.4. Other CSC targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... .......... . . . . . . . . . . . . . . . . . . . . . . . 00
4. Control of cell differentiation and the stem cell phenotype in prostate cancer: the potential role of ETS factors . . . . . . . . . . . . . . . . . . . . . . . 00
4.1. ETS fusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... .......... . . . . . . . . . . . . . . . . . . . . . . . 00
4.2. Aberrant expression of ETS factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... .......... . . . . . . . . . . . . . . . . . . . . . . . 00
4.3. Regulation of epithelial-specic ETS factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... .......... . . . . . . . . . . . . . . . . . . . . . . . 00
4.3.1. PDEF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... .......... . . . . . . . . . . . . . . . . . . . . . . . 00
4.3.2. ELF3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... .......... . . . . . . . . . . . . . . . . . . . . . . . 00
4.3.3. ESE3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... .......... . . . . . . . . . . . . . . . . . . . . . . . 00
5. Targeting ETS factors in cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... .......... . . . . . . . . . . . . . . . . . . . . . . . 00

* Corresponding author.
E-mail address: n.j.maitland@york.ac.uk (N.J. Maitland).

http://dx.doi.org/10.1016/j.jsbmb.2016.05.006
0960-0760/ 2016 Elsevier Ltd. All rights reserved.

Please cite this article in press as: L.K. Archer, et al., Stem cells and the role of ETS transcription factors in the differentiation hierarchy of normal
and malignant prostate epithelium, J. Steroid Biochem. Mol. Biol. (2016), http://dx.doi.org/10.1016/j.jsbmb.2016.05.006
G Model
SBMB 4721 No. of Pages 16

2 L.K. Archer et al. / Journal of Steroid Biochemistry & Molecular Biology xxx (2015) xxxxxx

5.1. Direct targeting of ETS factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00


5.2. Indirect targeting of ETS factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
5.3. Expression of suppressive ETS factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
5.4. Differentiation therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
6. Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00

1. Introduction 2. Prostate cancer stem cells: the evidence

Up to 40% of men with prostate cancer will develop The CSC hypothesis proposes that only a subpopulation of cells
metastases, most commonly to the bone [1]. Locally advanced within a tumour is able to initiate, propagate and maintain tumour
and metastatic prostate cancer are generally treated with growth. In addition to self-renewal and multipotency, CSCs also
pharmacological drugs which aim to deprive the tumour of have dysregulated proliferation and differentiation. Thus a CSC can
androgens which are essential for growth, known collectively as maintain itself, whilst also differentiating into the distinct
androgen-deprivation therapy (ADT). Patients undergoing ADT heterogeneous cell types that constitute the bulk of a tumour
may initially show promising tumour regression and reduced (Fig. 1). These tumour bulk populations are considered non-
prostate-specic antigen (PSA) levels. However, they will tumourigenic, although it should be noted that CSC properties may
eventually become hormone resistant and the recurrent cancer not only be originally acquired by normal tissue SCs but also by a
is known as castration-resistant prostate cancer (CRPC). Therapy more differentiated cell type within the hierarchy. The original cell
for CRPC includes treatment with cytotoxic chemotherapy agents type which is targeted for genetic mutation and transformation,
such as docetaxel or cabazitaxel, which target rapidly dividing whether or not it is a stem cell, is known as the cell of origin. Since
cells. Unfortunately, patients rapidly become resistant to chemo- CSCs are the only cells capable of driving tumour growth, these
therapeutic drugs and the survival benet with current agents is cells therefore require targeting to achieve long term cancer
small [2]. The median life expectancy of a patient with CRPC is therapy. However, consistent with normal tissue SCs, CSCs possess
about 2 years despite the development of next generation inherent resistance mechanisms which allow them to evade
therapies including abiraterone acetate, sipuleucel-T and enza- standard cancer treatments including chemotherapy and radiation
lutamide, which only prolong life expectancy by a matter of [2729].
months relative to a placebo [35]. These new agents target Multiple studies, going back over a hundred years had implied
multiple routes of androgen deprivation, including powerful the existence of CSCs [30]. The rst modern age evidence
blockade of the androgen receptor (enzalutamide) as well as presenting CSCs as the cause and maintenance of cancer was
androgen synthesis (abiraterone acetate) [6]. elegantly demonstrated in acute myeloid leukaemia (AML) by John
Although the principal focus of drug design has been on Dicks laboratory in 1994 [31]. Since then our understanding of
androgen depletion, resistance suggests that androgen receptor CSCs has been rened and their identication in multiple other
(AR) positive cells are not the only cell type within the tumour. leukaemias and solid tumours has supported the CSC hypothesis of
Current prostate cancer drugs target androgen responsive luminal cancer [32]. However, accumulating evidence for the CSC
cells which comprise the bulk of a tumour, whilst overlooking the hypothesis has posed signicant challenges, including the
AR negative basal cell populations which includes a stem-like identication of markers that can accurately distinguish CSCs
population. Cancer stem cells (CSCs), and similar populations, and expedite isolation of these relatively rare cells from a complex
have been identied in a number of cancers, including AML [7], tumour environment. The two functional caveats for identifying
glioblastoma [8], bladder [9], pancreas [10], breast [11], lung [12], CSCs are that they must (1) be tumourigenic, forming heteroge-
colon [13] and prostate [14]. This rare population possesses neous tumours reminiscent of those from which they were derived
similar characteristics to normal tissue stem cells. Stem cells are and (2) be serially transplantable when xenografted in mice. An
able to undergo asymmetric division, resulting in self-renewal array of markers, including CD133, CD44, CD24 and ALDH1, have
and a progeny cell with the ability to differentiate. This gives stem been used to isolate CSCs from different solid tumours. Moreover,
cells the distinct ability to reconstitute a tissue following injury, as there is no single CSC marker for each tumour type, multiple
which poses the hypothesis that a cancer stem cell can markers are generally used to isolate as homogeneous a population
reconstitute a tumour following treatment and thus promote as possible. These markers usually include normal stem cell
resistance and recurrence [1418]. The extended life span of a markers of the same tissue [33]. Whilst it has previously been
stem cell also presents the possibility for accumulation of the hypothesised that CSCs constitute a rare population of tumour
mutations required to develop cancer, but this is currently cells, studies in melanoma have shown that the current models
unproven. A likely candidate founder mutation in prostate cancer, could be signicantly underestimating the size of the CSC
namely fusion of TMPRSS2 to ERG, is found in 50% of prostate population. In NOD/SCID mice the proportion of cells with
cancers and is rarely considered in the context of CSCs [16,19]. tumourigenic capacity was 0.10.0001%, which rose to 25% in
ERG is an ETS family transcription factor, a family which is critical more immunocompromised NOD/SCID/IL2Rgnull mice [34].
for controlling and regulating a variety of cellular processes [20 However, the original hypothesis seems true in tumours of other
22]. The dysregulation of ETS factors via aberrant expression or tissues such as human pancreatic adenocarcinoma, lung squamous
repression, various mutations and their involvement in fusion cell carcinoma, lung adenocarcinoma, and head and neck
genes in both leukaemias and solid tumours, including prostate, squamous cell carcinoma. In these tumours, the CSC population
highlights them as signicant factors inuencing tumourigenesis only accounted for 0.00280.04% of total tumour cells in NOD/
[23]. Furthermore, as many ETS factors are also involved in SCID/IL2Rgnull mice [35]. The proportion of tumour cells with a CSC
epithelial cell differentiation it is important to consider the phenotype is therefore likely to be a quality that is dependent on
consequences of their expression in the individual cell populations tissue of origin and cancer subtype within that tissue, and in some
found in a prostate tumour [2426]. cases tumour grade may also be important [36,37]. Although serial

Please cite this article in press as: L.K. Archer, et al., Stem cells and the role of ETS transcription factors in the differentiation hierarchy of normal
and malignant prostate epithelium, J. Steroid Biochem. Mol. Biol. (2016), http://dx.doi.org/10.1016/j.jsbmb.2016.05.006
G Model
SBMB 4721 No. of Pages 16

L.K. Archer et al. / Journal of Steroid Biochemistry & Molecular Biology xxx (2015) xxxxxx 3

Fig. 1. The cancer stem cell hypothesis and progression of prostate cancer. This model shows the progression from a normal prostate differentiation hierarchy to the
deregulated growth of a prostate tumour. Although it is more likely for the cancer stem cell to arise from a normal stem cell, due to its longevity and self-renewal properties,
this model does not exclude the possibility that a progenitor cell could become the cancer stem cell following a series of mutations. This cancer stem cell is then able to
maintain tumour growth and differentiate into the different cell types that compose the tumour. In addition, it is possible for other mutations to occur within bulk tumour
cells altering their behaviour and contributing to a more aggressive cancer phenotype.

transplantation in mice is considered to be the gold standard as the third population committed to differentiation called committed
means of identifying CSCs, this data also highlights the hosts basal (CB) cells completes the basal hierarchy [43,44]. Signicantly,
immune response and differences in vital growth factors as several studies have shown that both basal and luminal cells derive
limiting factors in studying CSCs in model systems.

2.1. Epithelial composition of the prostate

The normal prostate epithelium is composed of an hierarchy of


cells that can be split into a basal compartment, which is attached
to the basement membrane, and a luminal compartment, which is
the inner layer lining the lumen as central space (Fig. 2). Androgen-
independent basal cells comprise the proliferative compartment of
the adult prostate and can be distinguished by their expression of
specic cytokeratins (CK) including CKs 5 and 14 [38], and other
cell markers such as p63 [39], CD44 [40] and bcl-2 [41]. Fully
differentiated luminal cells are the secretory cells of the prostate
and are androgen-responsive, expressing high levels of the
androgen receptor (AR) and prostate specic antigen (PSA).
Luminal cells also express CKs 8 and 18 [38] and CD57 [42].
Distinct differentiation states have been found within the basal
compartment, adding another level of complexity and heteroge-
neity to the prostate epithelium. The base of the hierarchy consists
Fig. 2. Architecture of the benign prostate. Immunouorescent staining of a tissue
of normal basal SCs which are largely quiescent [43]. Some SCs
section from benign prostatic hyperplasia (BPH) tissue highlighting the organised
asymmetrically divide into a proliferative, intermediate progenitor structure of individual glands and the surrounding stroma. Blue = DAPI (nuclei),
cell population which is termed as transit-amplifying (TA) cells. A red = p63 (basal cell nuclear marker) and green = Nkx3.1 (nuclear luminal marker).

Please cite this article in press as: L.K. Archer, et al., Stem cells and the role of ETS transcription factors in the differentiation hierarchy of normal
and malignant prostate epithelium, J. Steroid Biochem. Mol. Biol. (2016), http://dx.doi.org/10.1016/j.jsbmb.2016.05.006
G Model
SBMB 4721 No. of Pages 16

4 L.K. Archer et al. / Journal of Steroid Biochemistry & Molecular Biology xxx (2015) xxxxxx

from the same rare stem cell precursors found in the basal signalling into the basal and luminal epithelial cells of mice,
compartment, providing further evidence for the presence of a Lawson et al. demonstrated that whilst luminal cells were
differentiation hierarchy in the prostate (Fig. 3) [43,45]. unresponsive, the basal population was able to form fully
One controversy surrounding prostate CSCs is their cell type of differentiated tumours reminiscent of those seen in humans
origin. As mentioned above, CSCs are not always necessarily [51]. Signicantly, evidence from human prostate cells also
derived from normal SCs. In the normal prostate, the ratio of basal indicates a basal SC origin of prostate cancer. The
to luminal cells is around 50:50 [46,47], whereas the majority CD44+/CD133+/a2b1hi markers in basal cells identify normal
(>99%) of a prostate tumour consists of terminally differentiated, prostate SCs [52,53]. These cells showed self-renewal and
androgen-responsive luminal cells which are the target for current differentiation properties, were highly proliferative and recon-
therapy strategies including ADT (Fig. 3) [48]. Since prostate stituted prostate glands in vivo [53]. To isolate primary prostate
adenocarcinoma is histologically characterised by an expansion of CSCs which constitute approximately 0.1% of tumour cells from
luminal cells and loss of basal cells, it has previously been generally prostate cancer biopsies the same markers were used; cells were
assumed that luminal cells were the cell of origin of prostate rst selected for high a2b1 integrin expression by rapid adhesion to
cancer. However, given the limited proliferative potential and short collagen I-coated plates. CD133+ cells were then enriched from
life-span of normal luminal cells, they may be an unlikely a2b1hi cells [14]. These CD133+/a2b1hi cells exhibited enhanced
candidate (Fig. 4) [49,50]. As the study of prostate cancer relies proliferative potential and secondary colony-forming efciency in
partially on mouse models, the difference between the mouse and vitro compared to other populations, i.e. they were able to expand
human prostates can present a disadvantage when trying to and self-renew through several generations as the proportion of
elucidate the cell of origin of prostate cancer. CD133+ cells remained constant. As well as possessing the
fundamental SC properties, the cells also demonstrated classic
2.2. Evidence suggests that a basal cell is the cell of origin of prostate cancer cell characteristics of invasion and anchorage-independent
cancer growth, conrming that the cells were of tumour origin.
Furthermore, CD133+/a2b1hi cells were able to differentiate into
Substantial evidence exists for a basal SC origin of prostate AR+ luminal cells, recapitulating the characteristics of a primary
cancer. By introducing ERG expression and constitutive PI3K prostate tumour [14]. It should also be noted that these cells are

Fig. 3. Differentiation hierarchy and composition of the normal and cancerous prostate. The normal human prostate epithelium is composed of an hierarchical differentiation
pathway. Stem cells differentiate into rapidly dividing transit amplifying cells by asymmetric division and thus also maintain themselves. Multipotent transit amplifying cells
differentiate into committed basal cells, which then commit to terminal differentiation into secretory luminal cells. This tightly regulated pathway generates a highly
organised glandular structure consisting of the luminal layer, the basal layer attached to the basement membrane and the surrounding stroma. In prostate cancer this
differentiation pathway is deregulated, resulting in the mass proliferation and generation of luminal cells and a skewed ratio of basal to luminal cells compared to the normal
prostate.

Please cite this article in press as: L.K. Archer, et al., Stem cells and the role of ETS transcription factors in the differentiation hierarchy of normal
and malignant prostate epithelium, J. Steroid Biochem. Mol. Biol. (2016), http://dx.doi.org/10.1016/j.jsbmb.2016.05.006
G Model
SBMB 4721 No. of Pages 16

L.K. Archer et al. / Journal of Steroid Biochemistry & Molecular Biology xxx (2015) xxxxxx 5

Fig. 4. Prostate cancer cell of origin. In the prostate, evidence suggests that the cell of origin is likely to be a stem cell within the basal compartment. Should the cell of origin of
prostate cancer be a stem cell the resulting tumour would be heterogeneous with the ability to continuously adapt to the microenvironment following treatment. If mutations
accumulate in the more differentiated cells with a less long-lived proliferative potential (i.e. transit amplifying, committed basal and luminal cells) then the cells ability to
initiate heterogeneous, adaptable tumours would be more limited. These tumours are likely to spontaneously cease in growth or be cleared by the immune system or
conventional cancer treatments.

normally studied at as low a passage as possible, minimising any compared to human. It is therefore possible that the development
changes that may occur due to long-term culture. Patrawala et al. of prostate cancer differs between the two species and this CARN
also demonstrated that primary human prostate cancer CD44+ progenitor could be a unique cell found in mice. However, a
basal cells had increased tumourigenic and proliferative capabili- putative luminal CSC was found in the BM18 xenograft model
ties in vitro and in vivo than CD44 cells [54]. Then, similarly to derived from a human prostate cancer bone metastasis [62]. A
Lawson, Goldstein et al. conrmed that human benign basal cells, quiescent population of cells co-expressing stem-like (NANOG or
but not luminal cells, could produce tumours with a luminal ALDH1A1) and luminal (NKx3.1 and CK18) markers was able to
phenotype in NOD/SCID/IL2Rgnull mice following AKT, ERG and AR reconstitute a tumour in the presence of androgen following
expression [55]. Furthermore, it was recently shown by high- previous castration. However, these specic cells were not selected
throughput RNA sequencing that a basal stem cell phenotype was and serially transplanted into different mice, so it remains unclear
associated with more aggressive types of prostate cancer [18]. whether the cells truly have tumour-initiating properties. It should
Various combinations of markers including CD133, CD49f, CD44 also be noted that this xenograft model only represents one patient
and Trop2 are commonly used to detect prostate CSCs [5456]. and it is well known that prostate cancer is a very heterogeneous
In contrast, Wang et al. identied castration-resistant Nkx3.1- disease. Furthermore, the xenograft had been passaged many times
expressing cells (CARNs), which were dened as a subset of since it was established in 2005 and is likely to have evolved
luminal cells in the mouse prostate [57]. These cells comprised 1% signicant changes since then [63]. Critically, a luminal CSC has yet
of cells in the total prostate and were able to survive following to be isolated from a primary human prostate tumour [14].
androgen deprivation. CARN cells were proposed to be the origin of
prostate cancer, due to the presence of high grade PIN and 3. Cancer stem cells and treatment resistance in prostate cancer
carcinoma lesions following PTEN deletion and their ability to form
organoids in culture [57,58]. Similarly, luminal progenitors have The CSC hypothesis also logically explains the development of
also been found in other mouse models of prostate cancer [5961]. metastases. Complex sequencing studies have shown that the
However, the normal mouse and human prostates display cancer cell clones present in a primary prostate tumour are also
signicant differences. For instance, the murine prostate consists present in metastases [64,65]. Theoretically, in order to maintain
of a single layer of epithelia where luminal cells contact the the founder mutations present in the original tumour through to
basement membrane directly, and has a lower basal cell content metastasis, this must be carried out by a primitive, long-lived cell

Please cite this article in press as: L.K. Archer, et al., Stem cells and the role of ETS transcription factors in the differentiation hierarchy of normal
and malignant prostate epithelium, J. Steroid Biochem. Mol. Biol. (2016), http://dx.doi.org/10.1016/j.jsbmb.2016.05.006
G Model
SBMB 4721 No. of Pages 16

6 L.K. Archer et al. / Journal of Steroid Biochemistry & Molecular Biology xxx (2015) xxxxxx

that is capable of both clonal expansion and the ability to fractionated from benign and cancerous prostate tissue showed
accumulate further abnormalities that allow the cell to develop that SCs and CSCs also have signicantly higher expression of IAP
independence from the extracellular matrix and migrate to family members survivin and BIRC6 [16].
extraprostatic sites [66]. Prostate CSCs possess enhanced invasive
qualities in comparison to other tumour cell populations [18,67]. 3.2. DNA damage response
Successfully targeting the CSC population alongside cytotoxic
hormone therapies to kill the rapidly dividing tumour bulk may be Enhanced DNA damage response is another characteristic of
curative, and should prevent the initiation of lethal secondary CSCs, which allows them to survive exposure to radiation and
metastases (Fig. 5). However, heterogeneity within the CSC chemotherapy. Under normal circumstances, when DNA damage is
population itself makes targeted drug design even more challeng- irreversible, cell death pathways are activated, resulting in the
ing [68,69]. Furthermore, CSCs are thought to be the cause of upregulation of pro-apoptotic molecules. However, given the DNA
cancer recurrence following radiation and chemotherapy, and the damage repair mechanisms and upregulation of cell cycle
emergence of CRPC following ADT [47,70]. This seems logical, given checkpoint molecules in CSCs, they are able to survive such
that a characteristic of normal SCs is their robustness and inherent genotoxic stresses [81]. In glioma, CD133+ CSCs were more
resistance to toxic substances. The quiescent nature of stem cells resistant to radiation than the CD133 tumour cell populations.
poses a further challenge, since most cancer therapies target Inhibitors of checkpoint kinases Chk1 and Chk2 induced radio-
actively dividing cells. So, whilst conventional cancer therapies sensitivity, demonstrating that glioma CSCs have a permissive DNA
reduce the bulk population of tumour cells, these treatments damage checkpoint in response to radiation [82]. In cell line
consequently have the potential to enrich for CSCs [7173]. models, prostate CSCs also have increased expression of DNA
damage repair molecules [83,84]. In human prostate tissue the
3.1. Drug efux transporters and anti-apoptotic molecules CD133+/a2b1hi CSCs exhibited higher levels of heterochromatin,
which rendered them more resistant to the lethal double strand
Normal SCs and CSCs express drug efux transporters of the breaks induced by radiation. [85]. Radio-sensitivity could be
ABC transporter superfamily that can efciently pump chemo- induced by combination treatment with the histone deacetylase
therapeutic drugs out of the cells [74]. ABCG2 has been shown to be (HDAC) inhibitor Trichostatin A, i.e. CSC therapy resistance in this
expressed in prostate CSCs and is more highly expressed in case is dened at the chromatin level. Radiotherapy produces
prostate tumours which have recurred following treatment reactive oxygen species (ROS), causing oxidative stress and
compared to non-recurrent tumours, suggesting that the trans- subsequently DNA damage in cells. Breast CSCs possess low levels
porter does indeed play a part in drug resistance in prostate cancer of ROS compared to normal SCs due to increased expression of free
[75,76]. There have been numerous clinical trials for ABC radical scavengers and thus survive following radiation [86].
inhibitors, but outcomes have been poor due to high toxicity
and low efcacy [77]. Many CSCs also express anti-apoptotic 3.3. The tumour microenvironment and CSC niche
molecules, essentially allowing them to bypass signals which
would ordinarily lead to cell death. This includes members of the As well as possessing several mechanisms that confer drug
Bcl2 family [78,79] and inhibitor of apoptosis (IAP) family [80]. resistance, CSCs may also persist due to signals from the local
Affymetrix gene-expression arrays comparing the gene expression microenvironment. The normal SC niche is a distinct area within a
of human primary prostate CSCs and committed basal cells tissue that supports and provides SCs with the factors they require

Fig. 5. Prostate cancer cell response and resistance to treatment. Whilst current prostate cancer therapies can effectively eradicate the differentiated luminal cells which
compose the bulk of the tumour, cancer stem cells (CSCs) can survive via extensive therapy resistance mechanisms. The CSCs are then able to initiate the growth of new
tumours and metastases and may themselves have evolved to thrive in the post-treatment microenvironment. The discovery of novel CSC-directed therapies may be curative
and eliminate all tumour cells in conjunction with conventional therapies.

Please cite this article in press as: L.K. Archer, et al., Stem cells and the role of ETS transcription factors in the differentiation hierarchy of normal
and malignant prostate epithelium, J. Steroid Biochem. Mol. Biol. (2016), http://dx.doi.org/10.1016/j.jsbmb.2016.05.006
G Model
SBMB 4721 No. of Pages 16

L.K. Archer et al. / Journal of Steroid Biochemistry & Molecular Biology xxx (2015) xxxxxx 7

to maintain stemness [87]. A special niche may also exist for CSCs other members of the ETS family in prostate cancer are emerging in
[88]. For instance, CSCs are often found in niches in close proximity the literature [110115].
to vasculature, where VEGF-secreting endothelial cells can The ETS family of transcription factors are critical for the control
promote CSC-induced angiogenesis and metastasis [8991]. The and regulation of a variety of cellular processes such as
stroma surrounding solid tumours is altered from its normal state, haematopoiesis, differentiation, survival, and the immune re-
sometimes referred to as reactive stroma in the prostate [92,93]. sponse [2022,116]. Notably, ETS factors are also important in the
The reactive stroma (or cancer-associated broblasts, CAFs) maintenance of SCs of different tissues [117,118]. Many ETS factors
isolated from primary prostate tumours was sufcient in inducing are ubiquitously expressed in several different cell types, whilst
tumorigenesis in selected basal cells from the benign BPH-1 cell some factors are restricted to specic cell lineages and tissues
line, resulting in tumour formation in immunocompromised mice [119]. ETS proteins are exclusively found in metazoans and there
[94]. Additionally, with help from the reactive stroma, prostate are at least 30 human genes, which have been divided into specic
cancer cells are able to continuously adapt to their environment subfamilies according to the gene homology of their characteristic
promoting epithelial-to-mesenchymal transition (EMT) which ETS domains (Table 1). All ETS family members contain the unique
allows cells to become motile and disseminate. This inevitably 85 amino acid ETS DNA-binding domain (Fig. 6) which consists of a
results in invasion and metastasis with a particular tropism for winged-helix-turn-helix (WHTH) structure and binds to core
bone [93,9597]. The innate plasticity of CSCs, which allows them purine-rich sequences of GGAA/T in DNA target genes. This domain
to thrive in distant environments from the original tumour also is also involved in protein-protein interactions which regulate DNA
makes them a difcult therapeutic target. Treatments to target the binding [120]. ETS factors vary widely outside the ETS domain and
surrounding microenvironment may also be required to treat variations within the ETS domain itself between the transcription
prostate cancer more effectively [98]. factors can affect the proteins that can bind and also how they are
regulated, demonstrating the vast potential for regulation within
3.4. Other CSC targets this family.
A subset of ETS factors including ETS1, ETS2, ETV6 and ELF3 also
Given the extensive therapy resistance mechanisms of CSCs possess a pointed (PNT) domain at their N-terminal which is
discussed above, it is therefore likely that both the CSC and clonal important for protein-protein interactions and can further inu-
evolution models occur in unison in prostate cancer [99]. Whilst ence the transcriptional regulation of target genes (Fig. 6). For
the tumour is probably initiated by a long-lived CSC, following instance, the PNT domain of ETS1 acts as a docking site for
treatment, the selection pressures will dictate which CSCs will signalling molecule ERK2 [121]. Also, the PNT domain of ETV6
mutate and thrive and cause metastases by clonal evolution. This allows it to form oligomers which are important for the function of
makes the continually adapting CSCs an essential but challenging fusion proteins found in many leukaemias [122,123]. Overlapping
target. Inhibitors of key developmental pathways such as Wnt, functions and redundancy in promoter occupancy have been found
Notch and Hedgehog, all crucial for CSC maintenance, are currently between ETS proteins, making investigation of native biological
being tested [27,28,100]. For instance galiellalactone, a STAT3 interactions challenging [119,124].
inhibitor which blocks the binding of activated STAT3 to target
gene DNA, decreased the proportion of ALDH+ cells in prostate 4.1. ETS fusions
cancer cell lines, another commonly used CSC marker [101]. Other
inhibitors of the STAT3 pathway have also proved successful in The TMPRSS2-ERG fusion gene is formed in 50% of prostate
human primary prostate cancer cultures and xenograft models cancers between the androgen-regulated promoter of the trans-
[102]. Siltuximab (anti-IL-6) and LLL12, a specic inhibitor of membrane protease TMPRSS2 and the ETS factor (ERG) gene, both
activated STAT3, supressed the colony forming ability of prostate encoded on chromosome 21 [125,126]. There have been 14 distinct
CSCs. Following ex vivo treatment, LLL12 also prevented tumour variants of the TMPRSS2-ERG transcript reported in literature,
initiation of a xenograft derived from a castrate-resistant patient which can result in the translation of either normal full-length
[102]. ERG, various N-terminal truncations of ERG and one which encodes
It should be noted that since these pathways are also crucial in a TMPRSS2-ERG fusion protein [127129]. The most common
many normal cell types, including SCs, the off-target effects and isoform is that between TMPRSS2 exon 1 and ERG exon 4.
toxicity must be carefully monitored. An ongoing phase 2 clinical Signicantly, most of these transcripts retain the ETS DNA binding
trial targeting CSCs in late stage pancreatic cancer (ALPINE trial) domain of ERG, together with the PNT domain and C-terminal
has so far provided disappointing results [103]. The lack of success transactivation domain, ultimately resulting in the overexpression
with tarextumab, a Notch2 and Notch3 receptor inhibitor, to of a functional ERG protein. The incidence of the fusion in
produce a benecial response in patients is most likely due to the precancerous prostate intraepithelial neoplasia (PIN) lesions
off-target side effects of blocking this critical signalling pathway. suggests it could be an early event in prostate cancer development
However, encouragingly, NF-kB pathway blockade by parthenolide
treatment has been shown to decrease the viability of human
primary prostate CSCs without affecting the normal SC population, Table 1
ETS Factor Subfamilies.
highlighting that selective cell death of prostate CSCs may be
possible [16]. ETS Subfamily Family Members (alternative names)
ERG ERG, FLI1, FEV
4. Control of cell differentiation and the stem cell phenotype in ERF ERF (PE2), ETV3 (PE1)
prostate cancer: the potential role of ETS factors ETS ETS1, ETS2
ELG GABPa
PDEF PDEF (SPDEF, PSE)
Many ETS factors were discovered due to their integral roles in TEL ETV6 (TEL), ETV7 (TEL2)
leukaemias, and a number of ETS family members have since been SPI PU.1 (SPI), SPIB, SPIC
linked to various different cancers [104108]. In prostate cancer, ESE ELF3 (ESE1, ESX), ESE2 (ELF5), ESE3 (EHF)
ERG is the most overexpressed oncogene in patient tumour ELF ELF1, ELF2 (NERF), ELF4 (MEF)
TCF ELK1, ELK4 (SAP1), ELK3 (NET, SAP2)
samples and the TMPRSS2-ERG fusion is a suspected driver of PEA3 ETV4 (PEA3, E1AF), ETV5 (ERM), ETV1 (ER81), ETV2 (ER71)
tumourigenesis and progression [109]. Additional key roles for

Please cite this article in press as: L.K. Archer, et al., Stem cells and the role of ETS transcription factors in the differentiation hierarchy of normal
and malignant prostate epithelium, J. Steroid Biochem. Mol. Biol. (2016), http://dx.doi.org/10.1016/j.jsbmb.2016.05.006
G Model
SBMB 4721 No. of Pages 16

8 L.K. Archer et al. / Journal of Steroid Biochemistry & Molecular Biology xxx (2015) xxxxxx

Fig. 6. Domain structure of ETS factors which have proposed roles in prostate cancer. The amino acid number and positions of known domains of ETS factors which have been
shown to have a role in prostate cancer are displayed above. ETS factors are characterised by a unique 85 amino acid long sequence known as the ETS DNA binding domain
(ETS). A subset of ETS factors also possess a pointed domain (Pointed) involved in protein-protein interactions. The transactivation domain (TAD) is essential for activation of
transcription factor activity. Other regions important for DNA binding have also been identied such as the serine-rich region (SRR) of ETS1 and the AT hook domain (AT) of
ELF3. ELF3 also contains a serine- and aspartic acid-rich domain (SAR) which may play a role in cellular transformation. It should also be noted that some ETS factors such as
ERG, ETS1 and ELF3 have been found to possess auto-inhibitory regions anking the ETS domain. N = N-terminus, C = C-terminus.

[130]. Some uorescent in situ hybridisation (FISH) and immuno- required concomitant PTEN deletion to observe a disease pheno-
histochemistry (IHC) studies have correlated fusion positive type [145,152], probably due to the different genetic backgrounds
prostate tumours with poor prognosis and disease recurrence of the mice used. Since the TMPRSS2-ERG fusion is suspected to be
[131133]. Increased copy number of the TMPRSS2 or ERG loci, as an early event in prostate cancer initiation and evidence from
well as fusion formation by deletion rather than translocation, has mouse models suggests that other mutation events such as PTEN
also been linked to poor prognosis [134136]. To add to the loss are also required for cancer progression, it is reasonable to
complexity, prostate cancer is a heterogeneous, multifocal disease hypothesise that these initial events are likely to occur in long-
[137139]. Reverse transcriptase (RT) PCR and FISH studies of lived, self-renewing cells where mutations can accumulate. Casey
prostate tumours have revealed that multiple independent et al. produced a bacterial articial chromosome (BAC) by
TMPRSS2-ERG rearrangements can occur in a single patient in recombination to express the ERG gene under the control of the
separate tumour foci [126,127,140,141]. TMPRSS2 promoter [153]. Once again PIN lesions and progression
The TMPRSS2-ERG fusion results in overexpression of ERG to carcinoma were only evident in mice with an additional
oncoprotein. Whilst ERG is not expressed in the normal prostate, it heterozygous PTEN deletion. They observed a more genuine, cell
is consistently overexpressed in a large subset of prostate tumours, specic expression pattern of TMPRSS2-ERG which revealed
corresponding to the frequency of fusion positive patients [142]. expression in not only the EpCAM+/Sca-1/NKx3.1+ luminal cells
ERG overexpression has been proposed to be an inducer of but also in a fraction of the EpCAM+/Sca-1+/p63+ basal/progenitor
epithelial-to-mesenchymal transition (EMT), increased migration cell population. The Sca-1+ (stem cell antigen-1) basal cells from
and invasion, and subsequently metastasis through the upregu- TMPRSS2-ERG mice possessed greater sphere-forming and colony-
lation of associated target genes [143145]. ERG target genes forming ability than their counterparts from wild-type mice,
include activation of FZD4, MMP1, VEGFR2 and repression of E- illustrating that TMPRSS2-ERG positive stem-like cells have
cadherin, all of which directly promote EMT and invasion [109]. increased self-renewal capabilities.
Additionally, ERG stimulates the expression of the chemokine The study by Casey et al. highlighted the importance of
receptor CXCR4. Its ligand, CXCL12, is highly expressed in bone investigating critical genetic aberrations in the individual cell
marrow, resulting in the chemoattraction of prostate cancer cells populations of the prostate as opposed to total epithelial cells,
to bone, the most common site of metastasis [145147]. Two ChIP- which may mask any variations found in the rarer basal epithelial
seq studies in prostate cancer cell lines showed that ERG disrupts cell populations, including stem cells. Polson et al. examined this
the expression of androgen-regulated genes such as PSA in further by growing epithelial cells from human prostate biopsies in
cooperation with epigenetic-modifying enzymes HDAC1 and culture and fractionating the cell populations into SC, TA and CB
EZH2 [148,149]. ERG overexpression also promotes c-myc expres- populations based on CD133 and a2b1 expression as discussed
sion, as well as downregulating prostate-specic differentiation previously [14,19]. Using FISH, RT-PCR and southern blotting the
associated genes [150]. This disruption of androgen-regulation status and expression of TMPRSS2-ERG in the stem cell fraction
retains cells in a de-differentiated progenitor state that is thought from primary patient samples was conrmed. They also showed
to promote invasion and metastasis (Fig. 7). that prostate basal epithelial cells, particularly SCs, do not express
Initial mouse models to investigate the effects of TMPRSS2-ERG AR or the oestrogen receptor (ER), and therefore demonstrated that
utilised ERG cDNA driven by the prostate-specic modied TMPRSS2-ERG must be regulated in these cells in an androgen-
probasin promoter. Although some studies observed the develop- independent manner [19,48]. There was also heterogeneity in the
ment of PIN lesions following ERG overexpression [142,151], others expression patterns of TMPRSS2-ERG. For instance, in one patient

Please cite this article in press as: L.K. Archer, et al., Stem cells and the role of ETS transcription factors in the differentiation hierarchy of normal
and malignant prostate epithelium, J. Steroid Biochem. Mol. Biol. (2016), http://dx.doi.org/10.1016/j.jsbmb.2016.05.006
G Model
SBMB 4721 No. of Pages 16

L.K. Archer et al. / Journal of Steroid Biochemistry & Molecular Biology xxx (2015) xxxxxx 9

Fig. 7. Proposed mechanisms of ETS factor-induced de-differentiation of prostate cancer epithelial cells. The majority of the discussed ETS factors may play a role in prostate
cell differentiation. It appears that in prostate cancer these factors are regulated in a manner which disrupts the normal differentiation pathway and allows cells to remain in a
progenitor state which allows them to possess stem-like qualities promoting cancer growth and spread. (I) ERG is overexpressed in prostate cancer patients who possess the
TMPRSS2-erg fusion gene. ERG has been shown to both activate genes which initiate EMT and repress luminal-associated genes, characteristics associated with cancer stem
cells. (II) PDEF expression may be lost in terminally differentiated luminal cells due to miR-204 expression in cancer cells. PDEF loss results in an invasive cell phenotype
associated with less differentiated cells. (III) Microarray data shows ELF3 is more highly expressed in differentiated committed basal cells compared to stem cells from both
benign and cancerous primary prostate tissue. This suggests ELF3 expression may promote prostate SC differentiation. (IV) Knockdown of ESE3 results in the upregulation of
EMT markers and downregulation of luminal cell marker Nkx3.1. Together with characteristic SC functions such as colony forming ability, this suggests that ESE3 loss may
promote a SC phenotype. SC = stem cell, TA = transit amplifying cell, CB = committed basal cell.

TMPRSS2-ERG was expressed in both the SC and TA populations case in Ewings sarcoma, is an occurrence which should be
and lost in the CB population. Most commonly, whilst TMPRSS2- investigated. This may be an example of redundancy seen in ETS
ERG was expressed in the SC population, in several patients fusion factors, where in the absence of one fusion another may be able to
expression was lost in TA cells and regained in CB cells. This can be exert the same effects.
explained by a phenomenon called monoallelic expression,
whereby one inherited allele of a gene is preferentially expressed 4.2. Aberrant expression of ETS factors
and the other is silenced [154]. In cases of genetic abnormality
where one allele is normal and the other mutated, such as ETS1 is the founding member of the ETS transcription factor
TMRPSS2-ERG, the mechanism of monoallelic expression becomes family [160]. Its expression in cancers other than prostate is
signicant. Silencing of a mutated allele may prevent the early generally associated with negative prognosis and is most
removal of a potential cancer-initiating clone thereby allowing a frequently linked to high grade tumours and metastasis
higher chance of accumulating additional necessary mutations [161,162]. Studies of primary prostate tissue have shown that
required for cancer progression. ETS1 expression is also associated with poorly differentiated, high
TMPRSS2 fusions can also occur with other ETS factors, albeit at Gleason grade prostate tumours as well as faster progression to
much reduced frequencies. Along with ERG, TMPRSS2 fusions to castration-resistant disease following ADT [110,163]. A series of
ETV1 and ETV4 were also discovered by Tomlins et al. in prostate studies by Shaikhibrahim and Wernert also identied metastasis-
cancer cell lines and primary tissue [125,155]. The presence of the associated genes that were differentially expressed in PC3 cells
fusions again coincides with the overexpression of the corre- after ETS1 knockdown. These were then subsequently correlated to
sponding ETS factor and each appears to be mutually exclusive in primary prostate cancer tissue by gene expression microarray
most cases [156]. Unlike ERG, ETV1 and ETV4 are located on [164,165]. There is also evidence that ETS1 contributes to the
different chromosomes from TMPRSS2 and fusions are therefore castration-resistant phenotype by directly interacting with the AR
formed solely by translocation rather than interstitial deletion. and decreasing LNCaP sensitivity to the AR antagonist utamide
This may be relevant when deducing the prognostic signicance of following ETS1 overexpression [166]. Furthermore, ETS1 knock-
different fusions, given that TMPRSS2-ERG fusions formed by down in C4-2 cells (a castration-resistant progression model of
deletion are associated with poor prognosis [135]. The propensity LNCaP cells) resulted in decreased invasion and anchorage-
of ETS factors to form tumourigenic fusions has also led to the independent growth. Angiotensin II is an important mediator of
discovery of several other non-TMPRSS2 ETS factor fusions in angiogenesis and is a potent upregulator of VEGF [167]. Kosaka
prostate cancer [157159]. The incidence of several mutually et al. explored the effects of angiotensin II and angiotensin II type I
exclusive ETS factor fusions in prostate cancer, which is also the receptor blockers (ARBs) in prostate cancer cell lines [168].

Please cite this article in press as: L.K. Archer, et al., Stem cells and the role of ETS transcription factors in the differentiation hierarchy of normal
and malignant prostate epithelium, J. Steroid Biochem. Mol. Biol. (2016), http://dx.doi.org/10.1016/j.jsbmb.2016.05.006
G Model
SBMB 4721 No. of Pages 16

10 L.K. Archer et al. / Journal of Steroid Biochemistry & Molecular Biology xxx (2015) xxxxxx

Interestingly, angiotensin II upregulates ETS1, and is inhibited and invasive phenotype of prostate cancer cell lines and thus have
following treatment with the ARB candesartan. Collectively the tumour suppressor qualities upon its re-expression [112,178,179].
evidence shows that ETS1 is strongly associated with prostate This coincides with evidence found in other cancers showing loss
cancer metastasis and castration-resistant disease. of PDEF protein expression, including colon and breast cancer
Closely related to ETS1, ETS2 expression is also upregulated in [180182]. In prostate cancer, PDEF loss is in part regulated by
prostate cancer. Similar to ETS1, downregulation of ETS2 in Du145 microRNA miR-204 expression [112].
and PC3 cells reduces colony forming ability and cell survival, both Conversely, some studies claim PDEF expression increases in
characteristics of CSCs, whilst increasing apoptosis [111,169]. prostate and breast cancers compared to their normal tissues and
Interestingly, both ETS1 and ETS2 are highly expressed in induces invasiveness of a non-tumourigenic breast epithelial cell
androgen-insensitive and more invasive prostate cancer cell lines line when co-expressed with tyrosine kinases ErbB2 and CSF-1
PC3 and Du145, but are expressed at lower levels in androgen- [183,184]. This dichotomy can be explained since the studies that
sensitive (luminal) cell lines LNCaP and VCaP, providing further show PDEF to be a tumour suppressor in breast cancer are of the
evidence that ETS2 expression is associated with more malignant more invasive basal cell subtype and those that describe it as an
disease [111,170]. An early study by Liu et al. using a series of PCR oncogene are of the more differentiated luminal subtype
primers showed that ETS2 mRNA was expressed in 5 tumour foci (associated with better prognosis). This implies that in different
derived from frozen tissue sections from one patient [171]. On the cancer cell subtypes, PDEF can inuence tumour growth in
other hand, ETS1 was only expressed in one of these tumour foci. different ways. Since PDEF is normally expressed in luminal cells
This suggests that individual ETS factor expression, as well as in the prostate and associated with luminal cell markers such as
distinct TMPRSS2-ERG fusions, may arise in independent tumour PSA, loss of protein expression could potentially result in cells
foci and thus display different expression patterns between dedifferentiating, which would correlate with an increase of
focuses [140,171]. These studies highlight that prostate cancer is invasiveness (Fig. 7). Whether PDEF knockdown in more differen-
a heterogeneous disease with intra- and inter-patient heterogene- tiated (luminal) prostate cancers could be benecial therapeuti-
ity. Interestingly, ETS2 can regulate telomerase action. Telomerase cally remains to be seen, but this seems to be the situation in breast
is a reverse transcriptase which maintains the elongation of cancer [185]. Using this rationale, one would therefore expect PDEF
telomeres on chromosomes permitting an extended life span. expression to be lost in more aggressive, less differentiated
Telomerase is essential during development and remains active in prostate cancers (high Gleason grade) resulting in suppression of
long-lived cells, namely germ cells and SCs, whilst it is down- luminal markers such as PSA. If PDEF loss also promotes a stem-like
regulated in most other cell types. Telomerase is also present in phenotype in prostate cancer cells then EMT and invasion (also
>85% of human cancers [172]. ETS2 is an activator and repressor of properties of CSCs) should be seen. These proposals are yet to be
hTERT, the enzymatic component of telomerase, by binding to experimentally demonstrated.
different ETS binding sites within its promoter [173]. This was
found in breast cancer cell lines, and ETS2 knockdown resulted in 4.3.2. ELF3
decreased cell survival [174]. Since ETS2 is often overexpressed in ELF3 (also known as ESE1, ESX, and Jen) is mainly expressed in
prostate cancer, its regulation of telomerase and therefore epithelial-rich tissues such as the gut, kidneys, bladder and
potential of maintaining cell survival, a CSC characteristic, in prostate [186188]. Two papers have been published describing a
prostate cells could be of importance. role for ELF3 in prostate cancer with conicting views. Shatnawi
et al. described ELF3 as a repressor of prostate cancer. By
4.3. Regulation of epithelial-specic ETS factors interfering with AR DNA binding, ELF3 could repress the
upregulation of AR target genes known to be drivers of prostate
Within the ETS transcription factor family there is a subset of cancer [113]. Knockdown of ELF3 using siRNA induced AR target
genes whose expression is limited to epithelial cells. ESE2, ESE3 gene expression resulting in increased cell migration and
and ELF3 (also known as ESE1) are highly related and form a proliferation of the prostate cancer cell line LNCaP. Furthermore,
distinct gene subfamily. PDEF is another epithelial-specic ETS overexpression of ELF3 repressed tumour growth in an LNCaP
factor, although it differs from the others due to its preferential xenograft model. ELF3 protein loss was also associated with
binding to GGAT core sequences in target genes, as opposed to the prostate cancer progression by IHC from tissue sections, although
more common GGAA [175]. This relatively new category of ETS the patient cohort was relatively small.
factors, the rst of which was discovered in 1997, display In contrast, Longoni et al. described ELF3 as a driver of prostate
important roles in epithelial cell differentiation. Notably, there is cancer [114]. They observed high ELF3 expression at the mRNA
now mounting evidence that several of these factors play a role in level in several patient datasets. Immunohistochemical staining
prostate cancer. for ELF3 in 207 tumours revealed that 63% of tumours signicantly
expressed the protein, whereas ELF3 was poorly expressed in
4.3.1. PDEF normal tissues. ELF3 mRNA was also more highly expressed in
Prostate-derived ETS factor (PDEF) is most highly expressed in metastases compared with primary prostate tumours. The
the prostate but is also found at lower levels in other hormone- mechanism by which ELF3 drives prostate cancer was proposed
regulated organs including the breast, ovaries and salivary glands to be a positive feedback loop with the transcription factor NF-kB, a
and also in the lung and intestine [175,176]. PDEF protein is known upregulator of ELF3 [189,190]. This required pro-inam-
exclusively expressed in the terminally-differentiated secretory matory IL1-b stimulation for NF-kB activation and subsequent
luminal cells of normal prostate epithelium and has a role in the ELF3 induction, which was then able to sustain NF-kB activation.
differentiation of intestinal progenitor cells into the secretory cell ELF3 induction resulted in an increased malignant phenotype of
types of the intestinal epithelium [26,175]. Consistent with this LNCaP and 22RV1 cell lines by increasing colony forming ability in
expression pattern, PDEF was rst described as an upregulator of soft agar, anoikis resistance and cell migration. Xenograft tumours
PSA both in the presence of and independently of androgen [175]. produced by ELF3 over-expressing 22RV1 cells in nude mice
In prostate cancer, loss of PDEF expression is associated with more presented with larger and faster growing tumours and also
aggressive tumours, and patients with PDEF tumours have a produced metastases in the lung compared to control cells.
decreased survival rate compared to PDEF+ patients [112,177]. However, since 22RV1 cells are known to migrate and promote
Several studies have shown PDEF loss to enhance the migratory angiogenesis, other cell lines may have been more appropriate to

Please cite this article in press as: L.K. Archer, et al., Stem cells and the role of ETS transcription factors in the differentiation hierarchy of normal
and malignant prostate epithelium, J. Steroid Biochem. Mol. Biol. (2016), http://dx.doi.org/10.1016/j.jsbmb.2016.05.006
G Model
SBMB 4721 No. of Pages 16

L.K. Archer et al. / Journal of Steroid Biochemistry & Molecular Biology xxx (2015) xxxxxx 11

validate an effect in vivo [191]. It should also be taken into account the expression of Nkx3.1; a tumour-suppressor commonly lost in
that both of these studies did not use a normal cell comparison as a prostate cancer partly by heterozygous deletion of the gene located
control. Elucidating the roles of ELF3 in the normal prostate will aid at chromosome 8p21.2 [200,201]. Knockdown of ESE3 in LNCaP
further experiments in cancer models. and LHS cells (normal prostate epithelial cells [PrECs] which have
ELF3 has oncogenic and suppressive roles in colorectal cancer been immortalised via hTERT and SV40 large T antigen expression)
(CRC). It can upregulate b-catenin and subsequently activate resulted in reduced Nkx3.1 transcription. Since Nkx3.1 is also an
downstream targets of the Wnt/b-catenin signalling pathway to androgen-regulated marker of prostate epithelial cell differentia-
drive colon cancer progression [192]. ELF3 expression can also tion [202], downregulation of ESE3 and upregulation of ELF3 and
directly promote the apoptotic effects of non-steroidal anti- ERG may act in concert to produce similar effects.
inammatory drugs (NSAIDs) in CRC [193]. An explanation for To complement this theory, a study by the same group further
these conicting ndings is that ELF3 could be under similar examined the role of ESE3 suppression in prostate cancer cells
regulation to that of PDEF in cancer where it has different roles in [115]. ESE3 shRNA knockdown in immortalised PrECs and RWPE-1
different cell types. ELF3 also plays major roles in epithelial cell cells resulted in the upregulation of EMT markers vimentin and
differentiation, including the terminal differentiation of keratino- twist-1 as well as downregulation of epithelial cell marker
cytes as well as a high expression in the most differentiated cells of E-cadherin. Furthermore, ESE3 repression induced both colony-
the normal human urothelium [25,188]. Furthermore, a recent forming ability in soft agar and sphere-forming ability, both
study has shown that ELF3 is linked to differentiation of the human hallmarks of stem cells. Gene set enrichment studies of 3
embryonal carcinoma stem cell line NCCIT, after differentiation independent datasets of primary prostate tumours indicated that
was promoted using retinoic acid treatment. This treatment ESE3 suppression is found in approximately 25% of tumours and
resulted in the downregulation of early differentiation transcrip- can occur both cooperatively or independently of ERG over-
tion factor OCT4 but upregulation of ELF3 [24]. ELF3 knockdown expression. The highest upregulated gene sets in tumours with
led to an increase in endogenous OCT4. OCT4 is important for ESE3 suppression include EMT, adhesion and cytoskeleton
maintaining the pluripotency and self-renewal characteristics of remodelling related genes [115]. Therefore, since loss of ESE3
SCs and its expression has been linked to enhanced malignant downregulates luminal cell markers and appears to induce stem-
phenotype of cells and cancer progression in glioma, bladder and like properties it could be speculated that this would allow the de-
prostate cancer [194196]. differentiation of luminal cells and/or block the differentiation of
When Affymetrix gene-expression arrays of 15 benign prostatic basal cells (Fig. 7).
hyperplasia (BPH) and cancer tissues obtained from patients,
comparing prostate SC/CSC (CD133+/a2b1hi) gene expression to the 5. Targeting ETS factors in cancer
committed basal cell population (a2b1lo), 581 genes associated
with cancer and inammation were found to have signicantly 5.1. Direct targeting of ETS factors
altered expression in the CSC population [16]. This information has
allowed investigation of specic targets and pathways to further Various methods of targeting ETS factors have been investigat-
our understanding of prostate cell heterogeneity and CSC biology ed either at the transcription, mRNA or protein levels. For instance,
and identify potential therapeutic targets for prostate cancer [197]. triplex-forming oligonucleotides (TFOs) are single-stranded
Here, ELF3 was found to have consistently higher expression in oligonucleotides that bind to specic sequences in DNA and form
committed basal cells compared to SCs. Since low expression of triple helices, interrupting the transcription of the targeted gene.
ELF3 was found in both benign and malignant SCs across all They bind to long homopurine sequences, which are commonly
patients, ELF3 may be involved in SC differentiation (Fig. 7). Using found in gene promoter regions and enhance both the specicity
the different fractions of primary prostate epithelial cells to and stability of the interaction. A TFO targeted to ETS2 resulted in
investigate ELF3 should provide a more precise account of its roles decreased colony forming ability and induced apoptosis in DU145
in prostate cell differentiation and prostate cancer. cells [169]. Since ETS2 is a regulator of telomerase activation in
breast cancer cells [173], this type of therapy could potentially
4.3.3. ESE3 abrogate the self-renewal capacity of CSCs. The control experi-
ESE3 is an epithelial-specic ETS factor belonging to the same ments in this study revealed that the ETS2 TFO was highly target
subfamily as ELF3. The ESE3 ETS DNA-binding domain shares 84% gene-specic and presented an appealing prospect for future gene-
homology with ELF3 [198]. However, each has a distinct PNT targeted therapies [203]. However, a suitable delivery method and
domain, suggesting that whilst they may bind to similar target the potential for long-term mutagenic effects are signicant
genes, their effects could ultimately vary depending on the limitations.
protein-protein interactions achieved via their PNT domains. This A small molecule inhibitor of ETS factor fusion protein EWS-
was shown for prostate-specic promoters PSA and PSMA FLI1 in Ewings sarcoma has been tested on ERG-fusion positive
(prostate-specic membrane antigen). The PSMA promoter was VCaP cells and ETV1-fusion positive LNCaP cells [204]. The
upregulated 3-fold in the presence of ELF3, whilst ESE3 has no inhibitor, YK-4-279, decreased the migratory and invasive proper-
effect. Conversely, ESE3 was able to activate the PSA promoter over ties associated with ERG and ETV1 overexpression in prostate
2-fold whereas ELF3 repressed the PSA promoter [198]. This study cancer, without altering their protein levels in vitro. The fusion
was carried out in (non-prostate) HEK-293 cells, and so studies in negative cell line (PC3) was unresponsive to treatment, suggesting
more appropriate prostate models should be carried out to the inhibitor is selectively targeting the overexpressed ETS factors.
determine if ESE3 and ELF3 have opposing roles in prostate cell Whether YK-4-279 will also target essential ETS factors and cause
differentiation. off-target toxic effects in vivo still needs to be investigated.
In a microarray analysis of 59 primary prostate tumour samples A major consideration is that direct targeting of a transcription
and 14 normal prostate samples, differential ETS factor protein factor is likely to result in off-target side effects, especially given
expression in prostate cancer has been evaluated. ERG, ELF3 and the wide variety of functions ETS factors perform in different types
ESE3 were the most dysregulated genes, with ERG and ELF3 of cells and tissues (Fig. 8). ETS factor knockouts in mice generally
upregulated and ESE3 downregulated in cancer samples compared result in at best detrimental effects or even lethality [205]. As only
to normal prostate [199]. This study found a tumour-suppressor a limited number of target genes are likely to cause the oncogenic
role for ESE3 in prostate cancer as a result of its ability to regulate effects, or alternatively block tumour suppressors, the toxicity

Please cite this article in press as: L.K. Archer, et al., Stem cells and the role of ETS transcription factors in the differentiation hierarchy of normal
and malignant prostate epithelium, J. Steroid Biochem. Mol. Biol. (2016), http://dx.doi.org/10.1016/j.jsbmb.2016.05.006
G Model
SBMB 4721 No. of Pages 16

12 L.K. Archer et al. / Journal of Steroid Biochemistry & Molecular Biology xxx (2015) xxxxxx

Fig. 8. The benets and disadvantages of inhibiting ETS factors in cancer. The inhibition of aberrantly expressed ETS transcription factors in cancer would prevent their
binding to the promoter regions of target genes. This would subsequently lead to the repression of cancer-associated target genes normally upregulated by said ETS factor in
cancer. Alternatively, this may also alleviate the suppression of repressed tumour suppressor genes. This will lead to many benecial, tumour-specic effects. However, given
the importance of ETS factors in normal tissue homeostasis, there will also be toxic, off-target effects to normal cells.

induced by blocking transcription factors must therefore be surrounding tissues, the induced transcriptome of the factor is
carefully monitored. Targeting unique cancer-associated mole- an essential tool. The ability to evaluate precisely which target
cules such as TMPRSS2-ERG mRNA would avoid such complica- genes are important for cancer initiation and progression, and
tions and specically target cancer cells. A good example of this in identifying key co-regulators would be instrumental in developing
practice is imatinib, a small molecule inhibitor of the BCR-ABL ETS factor targeting therapies. Although time-consuming, this is
fusion protein which has revolutionised the treatment of fusion the only way, short of nding a cancer-specic and targetable
positive chronic myeloid leukaemia (CML) as well as other cancers molecule, to maintain the equilibrium between positive and
[206]. TMPRSS2-ERG is a complex target since the mRNA presents negative ETS factor function. ChIP-chip technology, where whole
as multiple different splice variants in each patient and the genome analysis can identify all the binding sites of a specic ETS
translated protein formed only consists of ERG. Shao et al. designed factor within the genome in different types of cells would be a
siRNA molecules specic for the unique junctions of two isoforms primary tool for this analysis [218].
of TMPRSS2-ERG mRNA, including the most common isoform
TMPRSS2 exon 1-ERG exon 4 and TMPRSS2 exon 2-ERG exon 4 5.3. Expression of suppressive ETS factors
which is associated with aggressive disease [207]. The efciency in
knocking down the fusion gene was tested in vivo via liposomal With regards to prostate cancer, PDEF and ESE3 are candidate
nanovectors, which have also been successfully tested in murine tumour suppressors that could be exploited for gene therapy. The
models of breast and ovarian cancer [208,209]. The siRNAs were classical example of a tumour suppressor is transcription factor
able to reduce growth of established VCaP xenograft tumours with p53, which is mutated in over 50% of all human cancers, making it
no effect on normal ERG levels and no toxicity observed in the an attractive therapy target. Re-introduction of wildtype p53 has
mice. Since prostate CSCs express TMPRSS2-ERG, knocking it down been investigated using non-replicative adenoviruses in multiple
might inhibit cell growth. Whilst this approach is appealing, its cancers including prostate [219221]. Clinical trials have produced
effect on the function and longevity of fusion positive prostate CSCs promising results in different cancers particularly in combination
would need to be monitored before it could be considered curative. with conventional chemotherapy and radiotherapy [222,223].
Furthermore, CSC liposome uptake into every tumour cell and their Notably, loss of functional p53 has also been linked to enhance-
potential to degrade the siRNA through resistance mechanisms ment of the CSC phenotype [224]. Therefore, reintroduction of p53
would also need to be investigated. may abrogate the CSC population and could explain the success of
the viral transduction approach in combination with traditional
5.2. Indirect targeting of ETS factors cancer therapies. Re-expression of PDEF or ESE3 would only be
benecial for the subtype of prostate cancers that have lost PDEF or
An alternative and almost certainly less toxic method of ESE3 expression. If PDEF and ESE3 loss are truly maintaining cells
preventing the aberrant effects of ETS factors in cancer is to in a de-differentiated state then their re-expression, e.g. via a viral
indirectly target the specic factor and thus target the oncogenic vector, could promote differentiation of the tumour cells. This
effect. ETS factors are known to regulate gene expression via should also differentiate the CSCs into a cell type that is more
interactions with co-regulatory proteins [210]. Since multiple ETS susceptible to conventional treatments.
factors are able to bind to the same target genes, targeting the
protein-protein interactions should be a more selective way of 5.4. Differentiation therapy
inhibiting their effects in cancer without interfering with the
normal binding of other ETS factors. Several synthetic molecules In order to cure prostate cancer, we must take into account the
have been constructed which interfere with the interaction cellular heterogeneity of the disease, a feature which current
between ELF3 and one of its co-activators, Sur2 [211213]. The therapy strategies have disregarded. This is reected in the limited
ELF3-Sur2 interaction is required for the activation of the growth survival rates of patients prescribed the next-generation drugs
factor HER2 [214]. The use of ELF3-Sur2 inhibitors downregulated which typically only target luminal cells. The effects of the
HER2 expression with a concomitant decrease in cell proliferation aberrantly expressed epithelial-specic ETS factors in prostate
and increase in apoptosis in cell line models of breast cancer, cancer appear to maintain the cancer cells in a more stem cell-like
gastric cancer and head and neck squamous cell carcinoma state. We and others have observed this stem-like phenotype in
(HNSCC). Furthermore, the ELF3-Sur2 inhibitors sensitised breast advanced metastatic tumours [17,18,225]. This will confer drug-
cancer cell lines to tamoxifen and also worked synergistically with resistance mechanisms and inevitably tumour recurrence follow-
the currently used treatments for gastric cancer and HNSCC [215 ing treatment, as these cells retain their proliferative potential and
217]. Thus, determining the interacting partners of ETS factors hence tumour renewal potency. The premise of differentiation
which are important in driving tumour progression may be a more therapy is to manipulate CSCs and progenitor cells to drive them
viable route for cancer therapeutics. towards differentiation and lose their stem-like abilities, resulting
In order to inhibit the negative effects of a specic ETS factor in in differentiated cells with limited proliferative capacity charac-
cancer without hindering its normal roles in target and teristic of the bulk of a prostate tumour. Although an attractive

Please cite this article in press as: L.K. Archer, et al., Stem cells and the role of ETS transcription factors in the differentiation hierarchy of normal
and malignant prostate epithelium, J. Steroid Biochem. Mol. Biol. (2016), http://dx.doi.org/10.1016/j.jsbmb.2016.05.006
G Model
SBMB 4721 No. of Pages 16

L.K. Archer et al. / Journal of Steroid Biochemistry & Molecular Biology xxx (2015) xxxxxx 13

theory, the design of effective differentiation therapies has been [7] D. Bonnet, J.E. Dick, Human acute myeloid leukemia is organized as a
limited [47]. The most notable success has been with acute hierarchy that originates from a primitive hematopoietic cell, Nat. Med. 3 (7)
(1997) 730737.
promyelotic leukaemia (APL) which is characterised by an [8] S.K. Singh, et al., Identication of a cancer stem cell in human brain tumors,
accumulation of progenitor cells caused by a block in differentia- Cancer Res. 63 (18) (2003) 58215828.
tion [226,227]. Differentiation is restored in APL via all-trans [9] K.S. Chan, et al., Identication: molecular characterization, clinical prognosis,
and therapeutic targeting of human bladder tumor-initiating cells, Proc. Natl.
retinoic acid treatment which is now part of the standard APL Acad. Sci. U. S. A. 106 (33) (2009) 1401614021.
therapy regime [228,229]. If differentiation therapy were to be [10] C. Li, et al., Identication of pancreatic cancer stem cells, Cancer Res. 67 (3)
successful in prostate cancer it would have to be combined with (2007) 10301037.
[11] M. Al-Hajj, et al., Therapeutic implications of cancer stem cells, Curr. Opin.
conventional cytotoxic treatments to kill the bulk tumour cells as Genet. Dev. 14 (1) (2004) 4347.
well as the differentiated and now vulnerable CSCs, which would [12] A. Eramo, et al., Identication and expansion of the tumorigenic lung cancer
eradicate all tumour cells and prevent recurrence. stem cell population, Cell Death Differ. 15 (3) (2008) 504514.
[13] L. Ricci-Vitiani, et al., Identication and expansion of human colon-cancer-
initiating cells, Nature 445 (7123) (2007) 111115.
6. Concluding remarks [14] A.T. Collins, et al., Prospective identication of tumorigenic prostate cancer
stem cells, Cancer Res. 65 (23) (2005) 1094610951.
The majority of studies which have investigated ETS factors in [15] L. Li, R. Bhatia, Stem cell quiescence, Clin. Cancer Res. 17 (15) (2011)
49364941.
prostate cancer understandably focus on TMPRSS2-ERG, given its [16] R. Birnie, et al., Gene expression proling of human prostate cancer stem cells
high incidence. However, whilst our knowledge of the roles of reveals a pro-inammatory phenotype and the importance of extracellular
other ETS factors in prostate cancer is in its infancy by matrix interactions, Genome Biol. 9 (5) (2008).
[17] D. Pellacani, et al., DNA hypermethylation in prostate cancer is a consequence
comparison, from the investigations discussed here it is clear of aberrant epithelial differentiation and hyperproliferation, Cell Death
that the regulation of these ETS transcription factor family Differ. 21 (5) (2014) 761773.
members could play a major role in prostate tumourigenesis and [18] B.A. Smith, et al., A basal stem cell signature identies aggressive prostate
cancer phenotypes, Proc. Natl. Acad. Sci. U. S. A. (2015).
progression. These ETS factors control target genes involved in [19] E.S. Polson, et al., Monoallelic expression of TMPRSS2/ERG in prostate cancer
EMT, invasion, metastasis and epithelial cell differentiation. The stem cells, Nat. Commun. 4 (2013) 1623.
accumulated evidence implies that several ETS factors inhibit [20] P. Kastner, S. Chan, PU.1: a crucial and versatile player in hematopoiesis and
leukemia, Int. J. Biochem. Cell Biol. 40 (1) (2008) 2227.
androgen-regulated genes and encourage cells to remain in a de-
[21] L. Russell, L.A. Garrett-Sinha, Transcription factor Ets-1 in cytokine and
differentiated state deemed optimal for EMT; all of which results chemokine gene regulation, Cytokine 51 (3) (2010) 217226.
in a more invasive cellular phenotype which is also characteristic [22] A. Kar, A. Gutierrez-Hartmann, Molecular mechanisms of ETS transcription
factor-mediated tumorigenesis, Crit. Rev. Biochem. Mol. Biol. 48 (6) (2013)
of CSCs.
522543.
The collective data also further strengthens the evidence that [23] A. Seth, D.K. Watson, ETS transcription factors and their emerging roles in
prostate CSCs are of basal origin. Deducing the expression pattern human cancer, Eur. J. Cancer 41 (16) (2005) 24622478.
of ETS factors in prostate CSCs is integral to elucidating whether [24] S.W. Park, et al., Transcriptional regulation of OCT4 by the ETS transcription
factor ESE-1 in NCCIT human embryonic carcinoma cells, Biochem. Biophys.
these factors could ultimately be manipulated to improve prostate Res. Commun. 450 (2) (2014) 984990.
cancer treatment. The principal line of investigation should now be [25] A. Cabral, et al., Distinct functional interactions of human Skn-1 isoforms
to look more closely at the specic functions of these ETS factors in with Ese-1 during keratinocyte terminal differentiation, J. Biol. Chem. 278
(20) (2003) 1779217799.
the more primitive basal cell populations including CSCs, as well as [26] A. Gregorieff, et al., The ets-domain transcription factor Spdef promotes
their contribution to cell differentiation, and to examine the maturation of goblet and paneth cells in the intestinal epithelium,
network of genes affected by their dysregulation. Whilst estab- Gastroenterology 137 (4) (2009) 13331345 e13.
[27] M.R. Alison, et al., Cancer stem cells: in the line of re, Cancer Treat. Rev. 38
lished cell lines are easily manipulated and provide a suitable (6) (2012) 589598.
means to explore mechanisms and theories, they do not fully [28] R. Morrison, et al., Targeting the mechanisms of resistance to chemotherapy
represent the complexity of a tumour and how the cells interact and radiotherapy with the cancer stem cell hypothesis, J. Oncol. 2011 (2011)
941876.
and differentiate. The possible dual roles of some of the ETS factors [29] H. Ishii, et al., Cancer stem cells and chemoradiation resistance, Cancer Sci. 99
discussed in different subtypes of cancer also adds complexity, and (10) (2008) 18711877.
emphasises that in order to truly deduce the roles of these factors [30] N.J. Maitland, A.T. Collins, The hallmarks of prostate cancer stem cells, in: V.K.
Rajasekhar (Ed.), Cancer Stem Cells, John Wiley & Sons, Inc., 2014.
we must look at each of the different cell types present in a tumour
[31] T. Lapidot, et al., A cell initiating human acute myeloid leukaemia after
population. Utilising the fractionated cell populations from transplantation into SCID mice, Nature 367 (6464) (1994) 645648.
primary patient tumours will provide a more accurate account [32] J.E. Visvader, G.J. Lindeman, Cancer stem cells in solid tumours: accumulating
of ETS factor expression, and the consequences of their aberrant evidence and unresolved questions, Nat. Rev. Cancer 8 (10) (2008) 755768.
[33] P.C. Park, et al., Stem cell enrichment approaches, Semin. Cancer Biol. 17 (3)
expression in normal and malignant prostate epithelial cells. (2007) 257264.
[34] E. Quintana, et al., Efcient tumour formation by single human melanoma
Acknowledgements cells, Nature 456 (7222) (2008) 593598.
[35] K. Ishizawa, et al., Tumor-initiating cells are rare in many human tumors, Cell
Stem Cell 7 (3) (2010) 279282.
This work was supported by Prostate Cancer UK (grant number [36] P. Marcato, et al., Aldehyde dehydrogenase activity of breast cancer stem cells
S13-016) and Yorkshire Cancer Research (grant number Y257PA). is primarily due to isoform ALDH1A3 and its expression is predictive of
metastasis, Stem Cells 29 (1) (2011) 3245.
[37] C. Thirant, et al., Clinical relevance of tumor cells with stem-like properties in
References pediatric brain tumors, PLoS One 6 (1) (2011) e16375.
[38] J.A. Schalken, G. van Leenders, Cellular and molecular biology of the prostate:
[1] H. Beltran, et al., New therapies for castration-resistant prostate cancer: stem cell biology, Urology 62 (5 Suppl. 1) (2003) 1120.
efcacy and safety, Eur. Urol. 60 (2) (2011) 279290. [39] S. Signoretti, et al., p63 is a prostate basal cell marker and is required for
[2] B. Seruga, I.F. Tannock, Chemotherapy-based treatment for castration- prostate development, Am. J. Pathol. 157 (6) (2000) 17691775.
resistant prostate cancer, J. Clin. Oncol. 29 (27) (2011) 36863694. [40] H.J. Terpe, et al., CD44 variant isoforms are preferentially expressed in basal
[3] J.S. de Bono, et al., Abiraterone and increased survival in metastatic prostate epithelial of non-malignant human fetal and adult tissues, Histochemistry
cancer, N. Engl. J. Med. 364 (21) (2011) 19952005. 101 (2) (1994) 7989.
[4] P.W. Kantoff, et al., Sipuleucel-T immunotherapy for castration-resistant [41] T.J. McDonnell, et al., Expression of the protooncogene bcl-2 in the prostate
prostate cancer, N. Engl. J. Med. 363 (5) (2010) 411422. and its association with emergence of androgen-independent prostate
[5] H.I. Scher, et al., Increased survival with enzalutamide in prostate cancer after cancer, Cancer Res. 52 (24) (1992) 69406944.
chemotherapy, N. Engl. J. Med. 367 (13) (2012) 11871197. [42] A.Y. Liu, et al., Cell-cell interaction in prostate gene regulation and
[6] A. Heidenreich, et al., EAU guidelines on prostate cancer. Part II: treatment of cytodifferentiation, Proc. Natl. Acad. Sci. U. S. A. 94 (20) (1997) 1070510710.
advanced, relapsing, and castration-resistant prostate cancer, Eur. Urol. 65 (2) [43] E.E. Oldridge, et al., Prostate cancer stem cells: are they androgen-
(2014) 467479. responsive? Mol. Cell. Endocrinol. 360 (1-2) (2012) 1424.

Please cite this article in press as: L.K. Archer, et al., Stem cells and the role of ETS transcription factors in the differentiation hierarchy of normal
and malignant prostate epithelium, J. Steroid Biochem. Mol. Biol. (2016), http://dx.doi.org/10.1016/j.jsbmb.2016.05.006
G Model
SBMB 4721 No. of Pages 16

14 L.K. Archer et al. / Journal of Steroid Biochemistry & Molecular Biology xxx (2015) xxxxxx

[44] J.T. Isaacs, D.S. Coffey, Etiology and disease process of benign prostatic [81] M. Maugeri-Sacca, M. Bartucci, R. De Maria, DNA damage repair pathways in
hyperplasia, Prostate Suppl. 2 (1989) 3350. cancer stem cells, Mol. Cancer Ther. 11 (8) (2012) 16271636.
[45] F.M. Frame, et al., Development and limitations of lentivirus vectors as tools [82] S. Bao, et al., Glioma stem cells promote radioresistance by preferential
for tracking differentiation in prostate epithelial cells, Exp. Cell Res. 316 (19) activation of the DNA damage response, Nature 444 (7120) (2006) 756760.
(2010) 31613171. [83] J. Yan, D. Tang, Prostate cancer stem-like cells proliferate slowly and resist
[46] M. El-Alfy, et al., Unique features of the basal cells of human prostate etoposide-induced cytotoxicity via enhancing DNA damage response, Exp.
epithelium, Microsc. Res. Tech. 51 (5) (2000) 436446. Cell Res. 328 (1) (2014) 132142.
[47] J.K. Rane, D. Pellacani, N.J. Maitland, Advanced prostate cancera case for [84] Y.S. Kim, M.J. Kang, Y.M. Cho, Low production of reactive oxygen species and
adjuvant differentiation therapy, Nat. Rev. Urol. 9 (10) (2012) 595602. high DNA repair: mechanism of radioresistance of prostate cancer stem cells,
[48] N.J. Maitland, et al., Prostate cancer stem cells: do they have a basal or luminal Anticancer Res. 33 (10) (2013) 44694474.
phenotype? Horm. Cancer 2 (1) (2011) 4761. [85] F.M. Frame, et al., HDAC inhibitor confers radiosensitivity to prostate stem-
[49] D.L. Hudson, et al., Proliferative heterogeneity in the human prostate: like cells, Br. J. Cancer 109 (12) (2013) 30233033.
evidence for epithelial stem cells, Lab. Invest. 80 (8) (2000) 12431250. [86] M. Diehn, et al., Association of reactive oxygen species levels and
[50] C. Zenzmaier, G. Untergasser, P. Berger, Aging of the prostate epithelial stem/ radioresistance in cancer stem cells, Nature 458 (7239) (2009) 780783.
progenitor cell, Exp. Gerontol. 43 (11) (2008) 981985. [87] L. Li, T. Xie, Stem cell niche: structure and function, Annu. Rev. Cell Dev. Biol.
[51] D.A. Lawson, et al., Basal epithelial stem cells are efcient targets for prostate 21 (2005) 605631.
cancer initiation, Proc. Natl. Acad. Sci. U. S. A. 107 (6) (2010) 26102615. [88] J.B. Sneddon, Z. Werb, Location, location, location: the cancer stem cell niche,
[52] A.T. Collins, et al., Identication and isolation of human prostate epithelial Cell Stem Cell 1 (6) (2007) 607611.
stem cells based on alpha(2)beta(1)-integrin expression, J. Cell Sci. 114 (Pt 21) [89] B. Beck, et al., A vascular niche and a VEGF-Nrp1 loop regulate the initiation
(2001) 38653872. and stemness of skin tumours, Nature 478 (7369) (2011) 399403.
[53] G.D. Richardson, et al., CD133, a novel marker for human prostatic epithelial [90] A.B. Alvero, et al., Stem-like ovarian cancer cells can serve as tumor vascular
stem cells, J. Cell Sci. 117 (Pt 16) (2004) 35393545. progenitors, Stem Cells 27 (10) (2009) 24052413.
[54] L. Patrawala, et al., Highly puried CD44+ prostate cancer cells from [91] A. Veeravagu, et al., The cancer stem cell-vascular niche complex in brain
xenograft human tumors are enriched in tumorigenic and metastatic tumor formation, Stem Cells Dev. 17 (5) (2008) 859867.
progenitor cells, Oncogene 25 (12) (2006) 16961708. [92] O. Dakhova, et al., Global gene expression analysis of reactive stroma in
[55] A.S. Goldstein, et al., Identication of a cell of origin for human prostate prostate cancer, Clin. Cancer Res. 15 (12) (2009) 39793989.
cancer, Science 329 (5991) (2010) 568571. [93] G.N. Thalmann, et al., Human prostate broblasts induce growth and confer
[56] M. Trerotola, et al., CD133, Trop-2 and alpha2beta1 integrin surface receptors castration resistance and metastatic potential in LNCaP cells, Eur. Urol. 58 (1)
as markers of putative human prostate cancer stem cells, Am. J. Transl. Res. 2 (2010) 162171.
(2) (2010) 135144. [94] R.A. Taylor, et al., Human epithelial basal cells are cells of origin of prostate
[57] X. Wang, et al., A luminal epithelial stem cell that is a cell of origin for prostate cancer: independent of CD133 status, Stem Cells 30 (6) (2012) 10871096.
cancer, Nature 461 (7263) (2009) 495500. [95] S. Josson, et al., Tumor-stroma co-evolution in prostate cancer progression
[58] C.W. Chua, et al., Single luminal epithelial progenitors can generate prostate and metastasis, Semin. Cell Dev. Biol. 21 (1) (2010) 2632.
organoids in culture, Nat. Cell Biol. 16 (10) (2014) 951961 14. [96] C. Scheel, R.A. Weinberg, Phenotypic plasticity and epithelial-mesenchymal
[59] H.et al. Korsten, Accumulating progenitor cells in the luminal epithelial cell transitions in cancer and normal stem cells? Int. J. Cancer 129 (10) (2011)
layer are candidate tumor initiating cells in a Pten knockout mouse prostate 23102314.
cancer model, PLoS One 4 (5) (2009) e5662. [97] G. van der Pluijm, Epithelial plasticity: cancer stem cells and bone metastasis
[60] J. Liu, et al., Regenerated luminal epithelial cells are derived from preexisting formation, Bone 48 (1) (2011) 3743.
luminal epithelial cells in adult mouse prostate, Mol. Endocrinol. 25 (11) [98] S. Bracarda, et al., Current and emerging treatment modalities for metastatic
(2011) 18491857. castration-resistant prostate cancer, BJU Int. 107 (Suppl. 2) (2011) 1320.
[61] S. Agarwal, et al., Identication of different classes of luminal progenitor cells [99] M. Greaves, C.C. Maley, Clonal evolution in cancer, Nature 481 (7381) (2012)
within prostate tumors, Cell Rep. (2015). 306313.
[62] M. Germann, et al., Stem-like cells with luminal progenitor phenotype [100] N. Takebe, et al., Targeting cancer stem cells by inhibiting Wnt, Notch, and
survive castration in human prostate cancer, Stem Cells 30 (6) (2012) Hedgehog pathways, Nat. Rev. Clin. Oncol. 8 (2) (2011) 97106.
10761086. [101] R.et al. Hellsten, Galiellalactone inhibits stem cell-like ALDH-positive
[63] D.R. McCulloch, et al., BM18: a novel androgen-dependent human prostate prostate cancer cells, PLoS One 6 (7) (2011) e22118.
cancer xenograft model derived from a bone metastasis, Prostate 65 (1) [102] P. Kroon, et al., JAK-STAT blockade inhibits tumor initiation and clonogenic
(2005) 3543. recovery of prostate cancer stem-like cells, Cancer Res. 73 (16) (2013)
[64] M.K. Hong, et al., Tracking the origins and drivers of subclonal metastatic 52885298.
expansion in prostate cancer, Nat. Commun. 6 (2015) 6605. [103] OncoMed Provides Update on Tarextumab Phase 2 Pancreatic Cancer ALPINE
[65] M.C. Haffner, et al., Tracking the clonal origin of lethal prostate cancer, J. Clin. Trial, Oncomed Pharmaceuticals, 2016. ([cited 2016 1st Feb]; Available from:)
Invest. 123 (11) (2013) 49184922. https://globenewswire.com/news-release/2016/01/25/804294/0/en/
[66] G. Gundem, et al., The evolutionary history of lethal metastatic prostate OncoMed-Provides-Update-on-Tarextumab-Phase-2-Pancreatic-Cancer-
cancer, Nature 520 (7547) (2015) 353357. ALPINE-Trial.html.
[67] G.J. Klarmann, et al., Invasive prostate cancer cells are tumor initiating cells [104] A. Zelent, M. Greaves, T. Enver, Role of the TEL-AML1 fusion gene in the
that have a stem cell-like genomic signature, Clin. Exp. Metastasis 26 (5) molecular pathogenesis of childhood acute lymphoblastic leukaemia,
(2009) 433446. Oncogene 23 (24) (2004) 42754283.
[68] G. Guenechea, et al., Distinct classes of human stem cells that differ in [105] W.A. May, et al., The Ewings sarcoma EWS/FLI-1 fusion gene encodes a more
proliferative and self-renewal potential, Nat. Immunol. 2 (1) (2001) potent transcriptional activator and is a more powerful transforming gene
7582. than FLI-1, Mol. Cell. Biol. 13 (12) (1993) 73937398.
[69] M. Greaves, Cancer stem cells: back to Darwin? Semin. Cancer Biol. 20 (2) [106] I. Panagopoulos, et al., Fusion of the FUS gene with ERG in acute myeloid
(2010) 6570. leukemia with t(16;21)(p11;q22), Genes Chromosomes Cancer 11 (4) (1994)
[70] F.M. Frame, N.J. Maitland, Cancer stem cells, models of study and implications 256262.
of therapy resistance mechanisms, Adv. Exp. Med. Biol. (2011) 105118. [107] T.R. Golub, et al., Oligomerization of the ABL tyrosine kinase by the Ets protein
[71] S.J. Dylla, et al., Colorectal cancer stem cells are enriched in xenogeneic TEL in human leukemia, Mol. Cell. Biol. 16 (8) (1996) 41074116.
tumors following chemotherapy, PLoS One 3 (6) (2008) e2428. [108] F. de Nigris, et al., Induction of ETS-1 and ETS-2 transcription factors is
[72] D.P. Freitas, et al., Therapy-induced enrichment of putative lung cancer stem- required for thyroid cell transformation, Cancer Res. 61 (5) (2001)
like cells, Int. J. Cancer 134 (6) (2014) 12701278. 22672275.
[73] X. Li, et al., Intrinsic resistance of tumorigenic breast cancer cells to [109] P. Adamo, M.R. Ladomery, The oncogene ERG: a key factor in prostate cancer,
chemotherapy, J. Natl. Cancer Inst. 100 (9) (2008) 672679. Oncogene 35 (4) (2015) 403414.
[74] K. Moitra, H. Lou, M. Dean, Multidrug efux pumps and cancer stem cells: [110] B. Li, et al., Overexpression of ETS-1 is associated with malignant biological
insights into multidrug resistance and therapeutic development, Clin. features of prostate cancer, Asian J. Androl. 14 (6) (2012) 860863.
Pharmacol. Ther. 89 (4) (2011) 491502. [111] V.I. Sementchenko, et al., ETS2 function is required to maintain the
[75] L.E. Pascal, et al., Molecular and cellular characterization of ABCG2 in the transformed state of human prostate cancer cells, Oncogene 17 (22) (1998)
prostate, BMC Urol. 7 (2007) p6. 28832888.
[76] E. Guzel, et al., Differential expression of stem cell markers and ABCG2 in [112] D.P. Turner, et al., Mechanisms and functional consequences of PDEF protein
recurrent prostate cancer, Prostate 74 (15) (2014) 14981505. expression loss during prostate cancer progression, Prostate 71 (16) (2011)
[77] J.I. Fletcher, et al., ABC transporters in cancer: more than just drug efux 17231735.
pumps, Nat. Rev. Cancer 10 (2) (2010) 147156. [113] A. Shatnawi, et al., ELF3 is a repressor of androgen receptor action in prostate
[78] Z. Madjd, et al., CD44+ cancer cells express higher levels of the anti-apoptotic cancer cells, Oncogene 33 (7) (2014) 862871.
protein Bcl-2 in breast tumours, Cancer Immun. 9 (2009) p4. [114] N. Longoni, et al., ETS transcription factor ESE1/ELF3 orchestrates a positive
[79] M. Konopleva, et al., The anti-apoptotic genes Bcl-X(L) and Bcl-2 are over- feedback loop that constitutively activates NF-kB and drives prostate cancer
expressed and contribute to chemoresistance of non-proliferating leukaemic progression, Cancer Res. 73 (14) (2013) 45334547.
CD34+ cells, Br. J. Haematol. 118 (2) (2002) 521534. [115] D. Albino, et al., ESE3/EHF controls epithelial cell differentiation and its loss
[80] G. Liu, et al., Analysis of gene expression and chemoresistance of CD133+ leads to prostate tumors with mesenchymal and stem-like features, Cancer
cancer stem cells in glioblastoma, Mol. Cancer 5 (2006) 67. Res. 72 (11) (2012) 28892900.

Please cite this article in press as: L.K. Archer, et al., Stem cells and the role of ETS transcription factors in the differentiation hierarchy of normal
and malignant prostate epithelium, J. Steroid Biochem. Mol. Biol. (2016), http://dx.doi.org/10.1016/j.jsbmb.2016.05.006
G Model
SBMB 4721 No. of Pages 16

L.K. Archer et al. / Journal of Steroid Biochemistry & Molecular Biology xxx (2015) xxxxxx 15

[116] E.W. Scott, et al., Requirement of transcription factor PU.1 in the development [150] C. Sun, et al., TMPRSS2-ERG fusion: a common genomic alteration in prostate
of multiple hematopoietic lineages, Science 265 (5178) (1994) 15731577. cancer activates C-MYC and abrogates prostate epithelial differentiation,
[117] H. Hock, et al., Tel/Etv6 is an essential and selective regulator of adult Oncogene 27 (40) (2008) 53485353.
hematopoietic stem cell survival, Genes Dev. 18 (19) (2004) 23362341. [151] O. Klezovitch, et al., A causal role for ERG in neoplastic transformation of
[118] C. Chen, et al., ERM is required for transcriptional control of the prostate epithelium, Proc. Natl. Acad. Sci. U. S. A. 105 (6) (2008) 21052110.
spermatogonial stem cell niche, Nature 436 (7053) (2005) 10301034. [152] J.C. King, et al., Cooperativity of TMPRSS2-ERG with PI3-kinase pathway
[119] T. Oikawa, T. Yamada, Molecular biology of the Ets family of transcription activation in prostate oncogenesis, Nat. Genet. 41 (5) (2009) 524526.
factors, Gene 303 (2003) 1134. [153] O.M. Casey, et al., TMPRSS2- driven ERG expression in vivo increases self-
[120] A.D. Sharrocks, The ETS-domain transcription factor family, Nat. Rev. Mol. renewal and maintains expression in a castration resistant subpopulation,
Cell Biol. 2 (11) (2001) 827837. PLoS One 7 (7) (2012) e41668.
[121] J.J. Seidel, B.J. Graves, An ERK2 docking site in the pointed domain [154] A. Chess, Mechanisms and consequences of widespread random monoallelic
distinguishes a subset of ETS transcription factors, Genes Dev. 16 (1) (2002) expression, Nat. Rev. Genet. 13 (6) (2012) 421428.
127137. [155] S.A. Tomlins, et al., TMPRSS2:ETV4 gene fusions dene a third molecular
[122] C. Jousset, et al., A domain of TEL conserved in a subset of ETS proteins denes subtype of prostate cancer, Cancer Res. 66 (7) (2006) 33963400.
a specic oligomerization interface essential to the mitogenic properties of [156] M.A. Svensson, et al., Testing mutual exclusivity of ETS rearranged prostate
the TEL-PDGFR beta oncoprotein, EMBO J. 16 (1) (1997) 6982. cancer, Lab. Invest. 91 (3) (2011) 404412.
[123] C.D. Mackereth, et al., Diversity in structure and function of the Ets family [157] J.D. Barros-Silva, et al., Novel 50 fusion partners of ETV1 and ETV4 in prostate
PNT domains, J. Mol. Biol. 342 (4) (2004) 12491264. cancer, Neoplasia 15 (7) (2013) 720726.
[124] P.C. Hollenhorst, et al., Genome-wide analyses reveal properties of redundant [158] B.E. Helgeson, et al., Characterization of TMPRSS2:ETV5 and SLC45A3:ETV5
and specic promoter occupancy within the ETS gene family, Genes Dev. 21 gene fusions in prostate cancer, Cancer Res. 68 (1) (2008) 7380.
(15) (2007) 18821894. [159] S.A. Tomlins, et al., Distinct classes of chromosomal rearrangements create
[125] S.A. Tomlins, et al., Recurrent fusion of TMPRSS2 and ETS transcription factor oncogenic ETS gene fusions in prostate cancer, Nature 448 (7153) (2007)
genes in prostate cancer, Science 310 (5748) (2005) 644648. 595599.
[126] J. Clark, et al., Complex patterns of ETS gene alteration arise during cancer [160] V. Laudet, et al., Evolution of the ets gene family, Biochem. Biophys. Res.
development in the human prostate, Oncogene 27 (14) (2008) 19932003. Commun. 190 (1) (1993) 814.
[127] J. Clark, et al., Diversity of TMPRSS2-ERG fusion transcripts in the human [161] N. Takai, et al., Expression of c-Ets1 is associated with malignant potential in
prostate, Oncogene 26 (18) (2007) 26672673. endometrial carcinoma, Cancer 89 (10) (2000) 20592067.
[128] J.J. Tu, et al., Gene fusions between TMPRSS2 and ETS family genes in prostate [162] T. Nakayama, et al., Expression of the ets-1 proto-oncogene in human
cancer: frequency and transcript variant analysis by RT-PCR and FISH on colorectal carcinoma, Mod. Pathol. 14 (5) (2001) 415422.
parafn-embedded tissues, Mod. Pathol. 20 (9) (2007) 921928. [163] G. Alipov, et al., Overexpression of Ets-1 proto-oncogene in latent and clinical
[129] J. Wang, et al., Expression of variant TMPRSS2/ERG fusion messenger RNAs is prostatic carcinomas, Histopathology 46 (2) (2005) 202208.
associated with aggressive prostate cancer, Cancer Res. 66 (17) (2006) [164] Z. Shaikhibrahim, et al., Comprehensive gene expression microarray analysis
83478351. of Ets-1 blockade in PC3 prostate cancer cells and correlations with prostate
[130] B.S. Carver, et al., ETS rearrangements and prostate cancer initiation, Nature cancer tissues: insights into genes involved in the metastatic cascade, Int. J.
457 (7231) (2009) E1 (discussion E23). Mol. Med. 27 (6) (2011) 811819.
[131] S. Perner, et al., TMPRSS2:ERG fusion-associated deletions provide insight into [165] Z. Shaikhibrahim, N. Wernert, ETS transcription factors and prostate cancer:
the heterogeneity of prostate cancer, Cancer Res. 66 (17) (2006) 83378341. the role of the family prototype ETS-1 (review), Int. J. Oncol. 40 (6) (2012)
[132] A.B. Rajput, et al., Frequency of the TMPRSS2:ERG gene fusion is increased in 17481754.
moderate to poorly differentiated prostate cancers, J. Clin. Pathol. 60 (11) [166] A.M. Smith, et al., ETS1 transcriptional activity is increased in advanced
(2007) 12381243. prostate cancer and promotes the castrate-resistant phenotype,
[133] K.C. Huang, et al., ERG expression is associated with increased risk of Carcinogenesis 33 (3) (2012) 572580.
biochemical relapse following radical prostatectomy in early onset prostate [167] X. Chen, et al., Angiotensin II type 1 receptor antagonists inhibit cell
cancer, Clin. Transl. Oncol. 16 (11) (2014) 973979. proliferation and angiogenesis in breast cancer, Cancer Lett. 328 (2) (2013)
[134] S.W. Fine, et al., TMPRSS2-ERG gene fusion is associated with low Gleason 318324.
scores and not with high-grade morphological features, Mod. Pathol. 23 (10) [168] T. Kosaka, et al., Ets-1 and hypoxia inducible factor-1alpha inhibition by
(2010) 13251333. angiotensin II type-1 receptor blockade in hormone-refractory prostate
[135] R. Mehra, et al., Characterization of TMPRSS2-ETS gene aberrations in cancer, Prostate 70 (2) (2010) 162169.
androgen-independent metastatic prostate cancer, Cancer Res. 68 (10) [169] G.M. Carbone, et al., Triplex DNA-mediated downregulation of Ets2
(2008) 35843590. expression results in growth inhibition and apoptosis in human prostate
[136] G. Attard, et al., Duplication of the fusion of TMPRSS2 to ERG sequences cancer cells, Nucleic Acids Res. 32 (14) (2004) 43584367.
identies fatal human prostate cancer, Oncogene 27 (3) (2008) 253263. [170] Z. Shaikhibrahim, et al., Differential expression of ETS family members in
[137] R. Arora, et al., Heterogeneity of Gleason grade in multifocal adenocarcinoma prostate cancer tissues and androgen-sensitive and insensitive prostate
of the prostate, Cancer 100 (11) (2004) 23622366. cancer cell lines, Int. J. Mol. Med. 28 (1) (2011) 8993.
[138] D.G. Bostwick, et al., Independent origin of multiple foci of prostatic [171] A.Y. Liu, et al., Identication of differentially expressed prostate genes:
intraepithelial neoplasia: comparison with matched foci of prostate increased expression of transcription factor ETS-2 in prostate cancer, Prostate
carcinoma, Cancer 83 (9) (1998) 19952002. 30 (3) (1997) 145153.
[139] A.M. Wise, et al., Morphologic and clinical signicance of multifocal prostate [172] E. Hiyama, K. Hiyama, Clinical utility of telomerase in cancer, Oncogene 21 (4)
cancers in radical prostatectomy specimens, Urology 60 (2) (2002) 264269. (2002) 643649.
[140] C.S. Cooper, et al., Analysis of the genetic phylogeny of multifocal prostate [173] J.M. Dwyer, J.P. Liu, Ets2 transcription factor, telomerase activity and breast
cancer identies multiple independent clonal expansions in neoplastic and cancer, Clin. Exp. Pharmacol. Physiol. 37 (1) (2010) 8387.
morphologically normal prostate tissue, Nat. Genet. 47 (4) (2015) 367372. [174] D. Xu, et al., Ets2 maintains hTERT gene expression and breast cancer cell
[141] R. Mehra, et al., Heterogeneity of TMPRSS2 gene rearrangements in proliferation by interacting with c-Myc, J. Biol. Chem. 283 (35) (2008) 23567
multifocal prostate adenocarcinoma: molecular evidence for an independent 23580.
group of diseases, Cancer Res. 67 (17) (2007) 79917995. [175] P. Oettgen, et al., PDEF: a novel prostate epithelium-specic ets transcription
[142] S.A. Tomlins, et al., Role of the TMPRSS2-ERG gene fusion in prostate cancer, factor, interacts with the androgen receptor and activates prostate-specic
Neoplasia 10 (2) (2008) 177188. antigen gene expression, J. Biol. Chem. 275 (2) (2000) 12161225.
[143] O.et al. Leshem, TMPRSS2/ERG promotes epithelial to mesenchymal [176] T.K. Noah, et al., SAM pointed domain ETS factor (SPDEF) regulates terminal
transition through the ZEB1/ZEB2 axis in a prostate cancer model, PLoS One 6 differentiation and maturation of intestinal goblet cells, Exp. Cell Res. 316 (3)
(7) (2011) e21650. (2010) 452465.
[144] S. Gupta, et al., FZD4 as a mediator of ERG oncogene-induced WNT signaling [177] A. Ghadersohi, et al., Prostate-derived Ets transcription factor (PDEF) is a
and epithelial-to-mesenchymal transition in human prostate cancer cells, potential prognostic marker in patients with prostate cancer, Prostate 71 (11)
Cancer Res. 70 (17) (2010) 67356745. (2011) 11781188.
[145] B.S. Carver, et al., Aberrant ERG expression cooperates with loss of PTEN to [178] T.R. Johnson, et al., Loss of PDEF, a prostate-derived Ets factor is associated
promote cancer progression in the prostate, Nat. Genet. 41 (5) (2009) 619624. with aggressive phenotype of prostate cancer: regulation of MMP 9 by PDEF,
[146] R. Singareddy, et al., Transcriptional regulation of CXCR4 in prostate cancer: Mol. Cancer 9 (2010) 148.
signicance of TMPRSS2-ERG fusions, Mol. Cancer Res. 11 (11) (2013) [179] X. Gu, et al., Reduced PDEF expression increases invasion and expression of
13491361. mesenchymal genes in prostate cancer cells, Cancer Res. 67 (9) (2007)
[147] J. Cai, et al., Androgens induce functional CXCR4 through ERG factor 42194226.
expression in TMPRSS2-ERG fusion-positive prostate cancer cells, Transl. [180] O. Moussa, et al., PDEF is a negative regulator of colon cancer cell growth and
Oncol. 3 (3) (2010) 195203. migration, J. Cell. Biochem. 108 (6) (2009) 13891398.
[148] J. Yu, et al., An integrated network of androgen receptor: polycomb, and [181] J.J. Steffan, H.K. Koul, Prostate derived ETS factor (PDEF): a putative tumor
TMPRSS2-ERG gene fusions in prostate cancer progression, Cancer Cell 17 (5) metastasis suppressor, Cancer Lett. 310 (1) (2011) 109117.
(2010) 443454. [182] R.J. Feldman, et al., Pdef expression in human breast cancer is correlated with
[149] K.R. Chng, et al., A transcriptional repressor co-regulatory network governing invasive potential and altered gene expression, Cancer Res. 63 (15) (2003)
androgen response in prostate cancers, EMBO J. 31 (12) (2012) 28102823. 46264631.

Please cite this article in press as: L.K. Archer, et al., Stem cells and the role of ETS transcription factors in the differentiation hierarchy of normal
and malignant prostate epithelium, J. Steroid Biochem. Mol. Biol. (2016), http://dx.doi.org/10.1016/j.jsbmb.2016.05.006
G Model
SBMB 4721 No. of Pages 16

16 L.K. Archer et al. / Journal of Steroid Biochemistry & Molecular Biology xxx (2015) xxxxxx

[183] R.N. Gunawardane, et al., Novel role for PDEF in epithelial cell migration and [208] C.N. Landen Jr., et al., Therapeutic EphA2 gene targeting in vivo using neutral
invasion, Cancer Res. 65 (24) (2005) 1157211580. liposomal small interfering RNA delivery, Cancer Res. 65 (15) (2005)
[184] A.K. Sood, et al., Expression characteristics of prostate-derived Ets factor 69106918.
support a role in breast and prostate cancer progression, Hum. Pathol. 38 (11) [209] J. Halder, et al., Focal adhesion kinase targeting using in vivo short interfering
(2007) 16281638. RNA delivery in neutral liposomes for ovarian carcinoma therapy, Clin.
[185] G. Buchwalter, et al., PDEF promotes luminal differentiation and acts as a Cancer Res. 12 (16) (2006) 49164924.
survival factor for ER-positive breast cancer cells, Cancer Cell 23 (6) (2013) [210] R. Li, H. Pei, D.K. Watson, Regulation of Ets function by proteinprotein
753767. interactions, Oncogene 19 (55) (2000) 65146523.
[186] M.J. Tymms, et al., A novel epithelial-expressed ETS gene, ELF3: human and [211] H. Shimogawa, et al., A wrench-shaped synthetic molecule that modulates a
murine cDNA sequences, murine genomic organization, human mapping to transcription factor-coactivator interaction, J. Am. Chem. Soc. 126 (11) (2004)
1q32.2 and expression in tissues and cancer, Oncogene 15 (20) (1997) 34613471.
24492462. [212] S. Asada, Y. Choi, M. Uesugi, A gene-expression inhibitor that targets an
[187] P. Oettgen, et al., Isolation and characterization of a novel epithelium-specic alpha-helix-mediated protein interaction, J. Am. Chem. Soc. 125 (17) (2003)
transcription factor: ESE-1, a member of the ets family, Mol. Cell. Biol. 17 (8) 49924993.
(1997) 44194433. [213] L.W. Lee, et al., Inhibition of ErbB2(Her2) expression with small molecule
[188] M. Bock, et al., Identication of ELF3 as an early transcriptional regulator of transcription factor mimics, Bioorg. Med. Chem. Lett. 19 (21) (2009) 6233
human urothelium, Dev. Biol. 386 (2) (2014) 321330. 6236.
[189] S. Rudders, et al., ESE-1 is a novel transcriptional mediator of inammation [214] M.M. Moasser, Targeting the function of the HER2 oncogene in human cancer
that interacts with NF-kappa B to regulate the inducible nitric-oxide therapeutics, Oncogene 26 (46) (2007) 65776592.
synthase gene, J. Biol. Chem. 276 (5) (2001) 33023309. [215] J.M. Nam, et al., Dithiiranylmethyloxy azaxanthone shows potent anti-tumor
[190] F.T. Grall, et al., The Ets transcription factor ESE-1 mediates induction of the activity via suppression of HER2 expression and HER2-mediated signals in
COX-2 gene by LPS in monocytes, FEBS J. 272 (7) (2005) 16761687. HER2-overexpressing breast cancer cells, Eur. J. Pharm. Sci. 50 (2) (2013)
[191] K. Stieler, et al., XMRV induces cell migration, cytokine expression and tumor 181190.
angiogenesis: are 22Rv1 cells a suitable prostate cancer model? PLoS One 7 [216] H.L. Kim, et al., A-62176: a potent topoisomerase inhibitor, inhibits the
(7) (2012) pe42321. expression of human epidermal growth factor receptor 2, Cancer Lett. 325 (1)
[192] J.L. Wang, et al., Elf3 drives beta-catenin transactivation and associates with (2012) 7279.
poor prognosis in colorectal cancer, Cell Death Dis. 5 (2014) pe1263. [217] M. Zhang, et al., Genetic and chemical targeting of epithelial-restricted with
[193] S.H. Lee, et al., ESE-1/EGR-1 pathway plays a role in tolfenamic acid-induced serine box reduces EGF receptor and potentiates the efcacy of afatinib, Mol.
apoptosis in colorectal cancer cells, Mol. Cancer Ther. 7 (12) (2008) Cancer Ther. 12 (8) (2013) 15151525.
37393750. [218] M.J. Buck, J.D. Lieb, ChIP-chip: considerations for the design, analysis, and
[194] Z. Du, et al., Oct4 is expressed in human gliomas and promotes colony application of genome-wide chromatin immunoprecipitation experiments,
formation in glioma cells, Glia 57 (7) (2009) 724733. Genomics 83 (3) (2004) 349360.
[195] D.E. Linn, et al., A role for OCT4 in tumor initiation of drug-resistant prostate [219] B. Hong, et al., Targeting tumor suppressor p53 for cancer therapy: strategies,
cancer cells, Genes Cancer 1 (9) (2010) 908916. challenges and opportunities, Curr. Drug Targets 15 (1) (2014) 8089.
[196] C.C. Chang, et al., Oct-3/4 expression reects tumor progression and regulates [220] A.H. Stegh, Targeting the p53 signaling pathway in cancer therapythe
motility of bladder cancer cells, Cancer Res. 68 (15) (2008) 62816291. promises: challenges and perils, Expert Opin. Ther. Targets 16 (1) (2012)
[197] N.J. Maitland, A.T. Collins, Prostate cancer stem cells: a new target for therapy, 6783.
J. Clin. Oncol. 26 (17) (2008) 28622870. [221] C. Yang, et al., Adenovirus-mediated wild-type p53 expression induces
[198] K. Kas, et al., ESE-3: a novel member of an epithelium-specic ets apoptosis and suppresses tumorigenesis of prostatic tumor cells, Cancer Res.
transcription factor subfamily, demonstrates different target gene specicity 55 (19) (1995) 42104213.
from ESE-1, J. Biol. Chem. 275 (4) (2000) 29862998. [222] Y.S. Xie, et al., Synergistic gastric cancer inhibition by chemogenetherapy
[199] P. Kunderfranco, et al., ETS transcription factors control transcription of EZH2 with recombinant human adenovirus p53 and epirubicin: an in vitro and in
and epigenetic silencing of the tumor suppressor gene Nkx3.1 in prostate vivo study, Oncol. Rep. 24 (6) (2010) 16131620.
cancer, PLoS One 5 (5) (2010) pe10547. [223] Y. Li, et al., In vitro and clinical studies of gene therapy with recombinant
[200] C. Bowen, et al., Loss of NKX3.1 expression in human prostate cancers human adenovirus-p53 injection for oral leukoplakia, Clin. Cancer Res. 15
correlates with tumor progression, Cancer Res. 60 (21) (2000) 61116115. (21) (2009) 67246731.
[201] C.D. Vocke, et al., Analysis of 99 microdissected prostate carcinomas reveals a [224] V.V. Prabhu, et al., Therapeutic targeting of the p53 pathway in cancer stem
high frequency of allelic loss on chromosome 8p 1221, Cancer Res. 56 (10) cells, Expert Opin. Ther. Targets 16 (12) (2012) 11611174.
(1996) 24112416. [225] J.K. Rane, et al., MicroRNA expression prole of primary prostate cancer stem
[202] T. Iwata, et al., MYC overexpression induces prostatic intraepithelial cells as a source of biomarkers and therapeutic targets, Eur. Urol. 67 (1)
neoplasia and loss of Nkx3.1 in mouse luminal epithelial cells, PLoS One 5 (2) (2015) 710.
(2010). [226] H. de The, et al., The PML-RAR alpha fusion mRNA generated by the t(15;17)
[203] G.M. Carbone, et al., Selective inhibition of transcription of the Ets2 gene in translocation in acute promyelocytic leukemia encodes a functionally altered
prostate cancer cells by a triplex-forming oligonucleotide, Nucleic Acids Res. RAR, Cell 66 (4) (1991) 675684.
31 (3) (2003) 833843. [227] F. Grignani, et al., The acute promyelocytic leukemia-specic PML-RAR alpha
[204] S.et al. Rahim, YK-4-279 inhibits ERG and ETV1 mediated prostate cancer cell fusion protein inhibits differentiation and promotes survival of myeloid
invasion, PLoS One 6 (4) (2011) e19343. precursor cells, Cell 74 (3) (1993) 423431.
[205] F.O. Bartel, T. Higuchi, D.D. Spyropoulos, Mouse models in the study of the Ets [228] M.E. Huang, et al., Use of all-trans retinoic acid in the treatment of acute
family of transcription factors, Oncogene 19 (55) (2000) 64436454. promyelocytic leukemia, Blood 72 (2) (1988) 567572.
[206] E. Nadal, E. Olavarria, Imatinib mesylate (Gleevec/Glivec) a molecular- [229] M.A. Sanz, et al., Management of acute promyelocytic leukemia:
targeted therapy for chronic myeloid leukaemia and other malignancies, Int. recommendations from an expert panel on behalf of the European
J. Clin. Pract. 58 (5) (2004) 511516. LeukemiaNet, Blood 113 (9) (2009) 18751891.
[207] L. Shao, et al., Highly specic targeting of the TMPRSS2/ERG fusion gene using
liposomal nanovectors, Clin. Cancer Res. 18 (24) (2012) 66486657.

Please cite this article in press as: L.K. Archer, et al., Stem cells and the role of ETS transcription factors in the differentiation hierarchy of normal
and malignant prostate epithelium, J. Steroid Biochem. Mol. Biol. (2016), http://dx.doi.org/10.1016/j.jsbmb.2016.05.006

You might also like