Università degli Studi di Trieste

Graduate School in MOLECULAR BIOMEDICINE

PhD Thesis

Prolyl-isomerase Pin1 controls normal and cancer stem cells of the breast by counteracting

the Fbxw7-oncosuppressive barrier on the Notch signalling pathway

Alessandro Zannini

XXVI ciclo – Anno Accademico 2012-2013

“Non troverai mai arcobaleni se guardi in basso”

- C. Chaplin

Dedico questa tesi ai miei genitori, a mia mamma per essere il mio pilastro,

a mio papà per essere la mia roccia.

Table of contents

1. TABLE OF CONTENTS

1. Table of contents ___________________________________________________________ 1

2. Abstract___________________________________________________________________3

3. Introduction _______________________________________________________________ 4

3.1. Breast cancer: a heterogeneous disease___________________________________ 4

3.1.1. Classification of breast cancer subtypes: advances and pitfalls________4

3.1.2. Normal breast stem cells_____________________________________ 7

3.1.3. Breast cancer stem cells______________________________________9

3.1.3.1 Breast CSCs: origin and history___________________________ 10

3.1.3.2 Breast CSCs isolation and functional characterisation____________ 11

3.1.3.3 CSCs underlie heterogeneity of human breast cancers____________ 13

3.1.3.4 CSCs are responsible for the failure of chemotherapeutic treatment__15

3.1.3.5 EMT and metastases are linked to CSC traits__________________17

3.2. The prolyl-isomerase Pin1: a critical hub of phosphorylation-dependent

pathway cross-talk___________________________________________________18

3.2.1. Pin1 structure and functional domains__________________________18

3.2.2 Regulation of Pin1 levels and activity __________________________ 20

3.2.3 Biological functions of Pin1__________________________________ 21

3.2.3.1 Mechanisms through which Pin1 impinges on cellular processes ____21

3.2.3.2 Cell-cycle control______________________________________23

3.2.3.3 Pin1 at the crossroads between stemness and differentiation________24

3.2.3.4 Pin1 pinpoints the cellular response to environmental stresses______26

3.2.4. Pin1 in cancer_____________________________________________28

3.3. The Notch signalling pathway__________________________________________31

3.3.1. Notch receptors and ligands__________________________________31

3.3.2. The Notch pathway activation________________________________ 33

3.3.3. The Notch pathway regulation________________________________33

3.3.4. Biological functions of the Notch pathway______________________ 36

3.3.5. Notch and cancer__________________________________________ 37

4. Aim of the thesis___________________________________________________________ 40

Table of contents

5. Results___________________________________________________________________ 41

5.1 The prolyl-isomerase Pin1 is required for the self-renewal of normal

mammary stem cells___________________________________________________ 41

5.2 Pin1 inhibition impairs mouse and human cancer stem cells traits________________44

5.3 Pin1 promotes breast CSCs maintenance mostly through up-regulation

of the Notch signalling_________________________________________________ 47

5.4 Suppression of Pin1 sensitizes breast CSCs to chemotherapy in vitro and in vivo___ 51

5.5 Pin1 protects N1- and N4-ICDs from proteasomal degradation_________________ 53

5.6 Pin1 protects N1- and N4-ICDs from degradation by the ubiquitin-ligase Fbwx7α__ 56

5.7 Pin1 protects N1- and N4-ICDs from interaction with Fbwx7α__________________59

5.8 Pin1 and Fbwx7α interplay affects Notch-dependent stem cells traits_____________62

5.9 Pin1 sustains Notch signalling in primary breast cancers in spite

of Fbxw7α expression _________________________________________________ 67

6. Discussion________________________________________________________________ 72

7. Materials and Methods______________________________________________________77

8. Bibliography______________________________________________________________ 87

9. Appendix________________________________________________________________ 100

10. Acknowledgements

Abstract

2. ABSTRACT

Cancer stem cells (CSCs) are proposed to be responsible for breast cancer heterogeneity,

chemotherapeutic treatment failure, metastatic spread and disease recurrence. The precise

identification of the molecular bases that govern the induction and maintenance of CSCs and

their aggressive phenotypes is of utmost importance, since it may provide the rational to develop

effective therapeutic strategies. In particular there is a considerable effort in finding common

pathways, mutations or histological features that might be targeted for therapy, overcoming

breast cancer heterogeneity.

Here we now demonstrate that CSC self-renewal, chemoresistance, tumour growth and

metastases formation capabilities’ are under direct control of Pin1’s enzymatic activity on the

Notch signalling pathway. In particular Pin1 protects the nuclear activated forms of Notch1 and

Notch4 (N1/4-ICD) from their E3-ubiquitin-ligase Fbxw7α, thereby boosting their protein levels

and transcriptional activity. Fbxw7α acts as a potent inhibitor of CSCs maintenance by

promoting protein degradation of N1- and N4-ICD, and, as a consequence, this ubiquitin-ligase

strongly decreased tumour growth and metastases dissemination in vivo. Interestingly,

concomitant over-expression of Pin1 almost completely recovered all these aggressive breast

cancer traits.

In tissues from breast cancer patients, we observed Notch signalling over-activation despite

presence of the negative regulator Fbxw7α, which relied on high Pin1 protein levels. Notably,

activation of the Notch-Pin1 axis correlated with poor prognosis in these patients.

As a consequence of our findings, suppression of Pin1 holds promise in reverting aggressive

phenotypes in breast cancer through shrinkage of CSCs number and a concomitant gain in

chemosensitivity, carrying important implications for breast cancers therapy.

Introduction

3. INTRODUCTION

3.1. Breast cancer: a heterogeneous disease

Breast cancer is the most common cause of cancer death among women (522 000 deaths in

2012) and the most frequently diagnosed tumour type in 140 out of 184 countries worldwide

(Bray et al., 2013). Despite death rates have dropped by 34% from 1990 to 2013 in the United

States due to an earlier diagnosis, better histopathological classifications and more specific

pharmacological treatments (Desantis et al., 2014; Weigelt et al., 2005), a significant percentage

of these patients still die. This is partly due to lifestyle changes that cause increased incidence,

but also because of treatment failures and the inability to hinder dissemination and growth of

metastases (Polyak, 2011; Steeg and Theodorescu, 2008). From a clinical point of view, the main

goal of treatment is eradication of the tumour and prevention of metastatic growth. Even though

therapeutic strategies have evolved, the three main pillars on which clinicians still rely are

represented by surgery, radio- and chemotherapy. Unfortunately, despite increasing advances,

when a tumour enters the metastatic phase, the probability of a cure is extremely reduced (Longo

et al., 2011). The main cause of the mild improvement of the clinical management of malignant

breast cancers resides in their intrinsic inter-tumoural and intra-tumoural heterogeneity. Breast

cancer is not a single disease but a collection of breast diseases, with different genetic variations,

gene expression profiles, histopathological features, pharmacological response and clinical

outcomes (Kreso and Dick, 2014; Stingl and , 2007; Vargo-Gogola and Rosen, 2007).

The possibility to associate mutations or cancer properties to a specific cell type or stage of

differentiation could contribute to better classify similar tumours, and to propose tumour type

specific interventions. Indeed, breast cancer heterogeneity suggests the hypothesis that breast

cancers can initiate in different cell types, either breast epithelial stem cells or their progeny.

However, the cellular hierarchy of the mammary gland and the cell of origin of different classes

of breast tumours are still open questions that await elucidation.

3.1.1 Classification of breast cancer subtypes: advances and pitfalls

Gene expression profiles and histological staging offered the opportunity to classify human

breast cancers in 5 different major subtypes: normal-like, luminal A, luminal B, HER2+, basal-

Introduction

like and claudin-low (Eroles et al., 2012; Perou et al., 2000; Sorlie et al., 2001; Sorlie et al.,

2003; Sotiriou et al., 2003). This type of classification is useful not only for assessment of

prognosis but also for the best fitting therapeutic regime.

The luminal A breast cancer is the most common subtype, decteted in almost 50% of breast

cancer patients. They are characterized by expression of estrogen receptor (ER), its target genes

and progesterone receptor (PR) in the luminal epithelium, low rate of proliferation, low

histological grade, good prognosis and they respond to ER modulator drugs.

The luminal B subtype is less present than the A type but these tumours are more aggressive,

with high proliferation rate, worse prognosis and they do not have a specific treatment regime as

not always responding to ER modulators.

The HER2 positive subtype represents al least 15-20% of all breast cancers, and they are

characterized by a high expression of the HER2 gene or amplification of HER2 gene, or up-

regulation of modulators of its pathway characterizes them. These tumours display high

proliferation rate, high histological grade and worse prognosis. From a clinical point of view the

discovery of anti-HER2 molecules strongly changed the median survival of these patients,

reducing the metastases formation but also curbing primary tumours.

The Basal-like subtype represents 10-20% of all breast carcinomas that expresses genes usually

present in normal breast myoepithelial cells. This subtype of cancer is characterized by large

tumour size, high histological grade, high frequency of lymph node affection and poor overall

survival. They are characterized by lack of expression of the 3 key molecular targets present in

the great proportion of breast tumours, namely ER, PR and HER2. For this reason, a large

percentage of Basal-like breast tumours are also called Triple Negative breast cancers, or TNBC.

Moreover, since there is no specific molecular target to date, the therapeutic regime for these

patients still relies on chemotherapy.

The Claudin-low subtype represents 12-15% of breast tumours and it is defined by low

expression of genes involved in the regulation of tight junctions, inter-cellular adhesion (claudin-

3, -4, -7) and epithelial markers (E-Cadherin). They are similar to basal-like tumours, but they

differ by the fact that the claudin-low subtype is strongly infiltrated by immune system cells and

moreover they are characterized by overexpressions of genes linked to mesenchymal

differentiation and epithelial-to-mesenchymal transition (EMT).

This subtype classification is not only relevant to define the most fitting therapeutic approach,

but it is also useful to predict the clinical outcome and to provide, at times, possible explanation

Introduction

to the failure of the pharmacological regime. In addition, the recent discovery of other breast

cancer subtypes (Eroles et al., 2012; Sims et al., 2007), as a consequence of more comprehensive

genomic and transcriptomic analyses, unveiled that a definitive classification is still far to come.

However, such classifications are inherently problematic, since they are based on molecular

analyses on a heterogeneous population, and they don’t take into account that, in a tumour cell

population, there is a collection of various statuses of differentiation, cellular commitment, and

mutational landscapes (Gerlinger et al., 2012). Moreover, the heterogeneity in breast cancer is

even more complex. For example a HER2 breast cancer subtype is classified on the basis of the

percentage of cells in the tumour mass display amplification of its gene –or up-regulation of its

pathway– that exceeds a particular cut-off. Thus, even if a tumour is classified as HER2, not all

tumour cells exhibit high levels of this receptor (Wolff et al., 2013), leading to the existence of

genetic and epigenetic intra-tumour variability (Marusyk and Polyak, 2010; Polyak, 2011). Thus,

even though personalized medicine using targeted agents is considered a more promising way to

target individual tumours than simple chemio- or radiotherapy, intra-tumour heterogeneity

represents an important hurdle to the achievement of a complete patients’ response.

Different approaches in cancer research are converging to provide more insights into the

underlying mechanisms of tumour heterogeneity and uncovering how these are linked to therapy

resistance, tumour progression, and recurrence. In this regard, therapies aimed at targeting

“cancer stem cells” have gained much attention (Figure 1) (Kreso and Dick, 2014).

Figure 1. Stemness as a guiding principle that governs therapeutic response Three fields in biology—cancer genetics, epigenetics, and microenvironment— are coming together to provide increasing clarity to the processes that determine stemness and in turn influence clinical outcome. These three factors can influence stemness simultaneously, but they can also act independently over time. Through evolutionary time, different forces can impact a cell’s stemness properties and thereby shape tumor progression and therapeutic response (Kreso and Dick, 2014).

Introduction

It is increasingly acknowledged that a subpopulation of cancer cells, termed cancer stem cells

(CSCs) play a major role in cancer growth, metastasis formation and chemoresistance in breast

cancer (Dean et al, 2005; Stingl & Caldas, 2007; Visvader & Lindeman, 2012). Like their

normal counterpart, CSCs are able to self-renew and maintain a reservoir of cancer-initiating

cells, which may produce a more differentiated progeny of cells and contribute to intra-tumour

heterogeneity (Stingl & Caldas, 2007). Current breast cancer research aims at better classifying

breast cancer subtypes and defining the molecular circuitries of subtype-specific CSCs, trying to

find common pathways, mutations or histological features that might be targeted for therapy,

overcoming breast cancer heterogeneity. Indeed, there is emerging evidence, that despite the

multiplicity of molecular signatures, receptors expression, histopathological features and intra-

tumour heterogeneity, a common characteristic that could underlie these differences, are the

CSCs (Kreso and Dick, 2014; Magee et al., 2012; Pece et al., 2010; Stingl and Caldas, 2007;

Visvader and Lindeman, 2012)

3.1.2. Normal stem cells of the mammary gland

The mammary gland is a tubulo-alveolar gland, composed by three lineages of epithelial cells,

i.e. luminal, myoepithelial and alveolar epithelial cells that derive from luminal progenitors and

are differentiated for synthesizing milk proteins. The mammary gland undergoes a sequence of

dynamic changes throughout the female life cycle (Gjorevski and Nelson, 2011), and the

capacity of the mammary gland epithelium to expand and remodel during puberty and repeated

cycles of pregnancy is highly suggestive of the existence of resident mammary gland stem cells

(SCs) in the adult (Visvader, 2009).

The first evidence of the existence of mammary stem cells derived from experiments in which a

transplantation of different dissected parts of the mammary gland into the fat pad of syngenic

rodents, gave rise to fully developed and functional mammary glands (Hoshino, 1964). Other

studies carried out with the purpose to identify a stem cell population able to regenerate the

mammary gland (Gudjonsson et al., 2002; Rudland et al., 1997), but the isolation and

characterisation of a true stem cell population from both humans and mice, has been obtained

only few years later, using in vitro and in vivo reconstitution transplantation experiments (Dontu

et al., 2003a; Dontu et al., 2003b; Stingl et al., 1998; Stingl et al., 2001; Stingl et al., 2006).

Afterwards, the generation of the different mammary epithelial cells by resident mammary stem

cells was demonstrated by two different groups through lineage-tracing experiments in mice, as

depicted in the scheme of Figure 2 (Rios et al., 2014; Van Keymeulen et al., 2011; Visvader and

Lindeman, 2011).

Introduction

Figure 2. Model of the epithelial differentiation hierarchy in the mammary gland There is recent evidence for a heterogeneous compartment of stem cells, in which the stem cell at the apex is multipotent and likely has higher self‐renewing capacity than the unipotent stem cells (myoepithelial‐stem cell, Myo‐SC; luminal‐stem cell, Lum‐SC) and the ‘shorter‐term’ mammary repopulating cell that emerges in pregnancy (Preg‐SC). Alternatively, there may be one true stem cell and a hierarchy of descendant progenitor cells (indicated by overlap between the stem and progenitor cell compartments) that include long‐ and short‐lived lineage‐restricted progenitors (Visvader and Lindeman, 2011).

Adult stem stem cells are defined as undifferentiated and multipotent or bi/unipotent cells that

have the potential to differentiate and generate the specialized end cells of that tissue. Moreover

they are able to self-renew, so to maintain in this way the stem cells’ compartment for many

passages, and to generate some committed progenitors with tissue maintaining potential. This

ability is a consequence of their peculiar way of cell division. Indeed they are able to divide

giving rise to one stem cell, identical to the initial stem cell, and a daughter committed

progenitor cell, which undergoes a differentiation process. This type of division is called

asymmetric. By contrast, in specific conditions, stem cells are able to divide also symmetrically

and generate two stem cells, without the differentiated one, expanding in such way the stem cell

pool (Liu et al., 2005; Stingl and Caldas, 2007).

Seminal work performed by Shackleton and Stingl has provided proper means to isolate and

characterize a stem cell population capable of reconstituting a functional mammary gland in

mice, called mammary repopulating units (MRUs) (Shackleton et al., 2006; Stingl et al., 2006).

MRU cells are purified from the mammary gland after digestion and sorted by both excluding

cells expressing hematopoietic and endothelial markers and including the expression of CD24,

α6-integrin (CD49f), Sca-1 and β1-integrin (CD29). In particular, cells with a

CD24med/CD49fhigh profile characterize the MRU population and it is present in a proportion of 1

MRU out of 1400 total cells in mice.

Introduction

Moreover, by using an in vitro colony assay, other groups have demonstrated that a small

subpopulation of mammary gland stem cells (MaSC) is able to give rise to luminal and

myoepithelial lineages, suggesting the existence of bipotent stem cells (Dontu et al., 2003; Guo

et al., 2012). Another in vitro experiment, that is now a widely used as a surrogate stem cell

assay, is based on the ability of cells to generate clonal mammospheres in non-adherent, serum-

free conditions and it evaluates the stem cells’ content in a breast cell line or tissue. Notably,

these mammospheres are enriched in multi-lineage progenitors and display a transcriptional

profile enriched in stemness factors and pathways (Dontu et al., 2003).

The ability to prospectively isolate and analyze the stem cell population of the mammary gland

gives the opportunity to identify molecular signatures and factors required for their own

maintenance. In fact, thanks to these methods, many groups had dissected the molecular bases of

stem cells from both mouse and human breast tissues delineating the main factors involved in

stemness (Ben-Porath et al., 2008; Dontu et al., 2003; Pece et al., 2010; Raouf et al., 2008; Spike

et al., 2012)

In summary, a huge number of data allowed identifying a stem cell population that is able to

generate, in vitro, mammospheres and colonies composed of both lineages and, more

importantly, an entire functional mammary gland in vivo. Notably the isolation and

characterisation of human mammary stem cells lacks a real in vivo validation, but the in vitro

assays are well established and reliable.

3.1.3 Breast cancer stem cells

Breast cancer stem cell (CSCs) research is one of the most investigated fields in breast cancer

research in the last years. Like normal stem cells that are able to regenerate an organ, also CSCs

are considered to be responsible for re-generation of the tumour, as demonstrated by sequential

transplantation assays. From a clinical point of view breast CSCs are associated to

aggressiveness in breast cancer since they are chemoresistant, they have been linked to

metastases formation and associated to the undifferentiated, poor outcome breast cancer subtypes

(Stingl and Caldas, 2007; Valent et al., 2012; Visvader and Lindeman, 2012). Interestingly, the

most aggressive and high-grade breast cancer subtypes display an enriched CSCs population and

a stem cell-like signature (Ben-Porath et al., 2008; Pece et al., 2010). As a matter of fact, a better

knowledge of their properties will be important in order to find possible strategies to curb their

tumourigenic potential.

Introduction

3.1.3.1 Breast CSCs: origin and history

There is increasing evidence that normal stem cells might be the targets of transformation during

tumourigenesis. Tumours are believed to arise through a series of mutations that occur over

many years. Adult stem cells are slowly dividing, long-lived cells, which by their very nature are

exposed to damaging agents for long periods. There are two possible theories concerning the

formation of CSCs. The first, also called the CSC hypothesis, claims that in some cancers CSCs

have arisen from normal stem cells that, through acquisition of genetic and epigenetic

modifications, become cancerous and display perturbed self-renewal. Another possibility is that

terminally differentiated epithelial cells, or restricted transient amplifying progenitors, through

clonal evolution, acquire tumourigenic features and, through reprogramming or acquisition of

plasticity, they lose their committed fate and become more stem-like. Although no clear in vivo

evidence of the CSC origin has been yet found, in favour of the first hypothesis comes the

observation that normal stem cells and cancer stem cells share several important features (Dontu

and Wicha, 2005; Liu et al., 2005; Stingl and Caldas, 2007; Visvader and Lindeman, 2012).

Using different experimental approaches, several investigators proved how only a minority of

cells in human cancers are capable to self-renew. This has been most convincingly demonstrated

by examining the ability of subpopulations of tumour cells identified by cell surface markers to

form tumours when transplanted into immunosuppressed NOD/SCID mice. This approach was

first successfully used to demonstrate the existence of leukemic stem cells (Dick, 1996) and,

some years later, also for breast cancer stem cells (Al-Hajj et al., 2003). Using the cell surface

markers CD44 and CD24, Al-Hajj and colleagues have been able to isolate a putative CSCs

population since a small number of injected cells was able to generate a tumour in a host mice.

There is now evidence for the existence of CSCs in many other tumours (Valent et al., 2012;

Visvader and Lindeman, 2012), thus suggesting that CSCs are present in a wide number of

cancers and, as a consequence, they could be considered as a hallmark of cancer.

The identification of pathways that govern CSCs’ maintenance is important to develop targeted

therapies. The developmental Notch pathway, Bmi-1, Hedgehod and Wnt signalling, p53, c-myc,

TAZ, ErbB2 and regulators of epithelial-to mesenchymal transition such as Slug, are the main

regulators of stemness in breast cancer (Mani et al., 2008; Valent et al., 2012). Tumour-specific

reactivation of some of these pathways or abrogation such as for p53, is able to re-program a

committed or progenitor cell to a more dedifferentiated cell with stem cell features. Since most

of these factors are also involved in cancer development it is evident how some genetic or

Introduction

epigenetic mutations, that drive tumourigenesis, are able to promote formation of a stem cell

population.

3.1.3.2 Breast CSCs isolation and functional characterisation

Nowadays no clear and unique assay has been validated and demonstrated to isolate breast

CSCs. A multi-approach experimental detection and analysis of CSCs, indeed will be more

predictive and reliable (Stingl and Caldas, 2007; Visvader and Lindeman, 2012). The most

important and well-validated assay is the in vivo transplantation in immunodeficient mice. This

assay evaluates the most peculiar trait of CSC, which is the ability of these cells to re-generate a

tumour. For this reason CSC are often called tumour-initiating cells. In particular, while a CSC is

able to generate a tumour, the differentiated counterpart of the tumour is unable to (Valent et al.,

2012). This assay allows also to compare the CSCs’ content of different tumours: indeed,

injection of low numbers of cells gives rise to tumours when derived from breast tumour/cell

lines with a high content of putative CSCs, while the same amount of cells from tumours

displaying a low CSCs’ content are not able to generate tumours with the same frequency.

However, despite the reliability of this in vivo assay, there are some limitations: i) being host

mice immunodeficient the innate and cell-based immune reactions as well as the content of

cytokines is deranged, hence, depending on the injected cells, tumour development in these mice

may be either facilitated or suppressed ii) the host mice lack the tumour microenvironment that

is required as a supportive scaffold for tumours and CSCs, iii) the presence of both stem and

progenitor cells requires second-generation recipient animals to better distinguish true stem cells

from early progenitors which have the potential to generate some tumours in the first but not in

the second transplanted animal (Valent et al., 2012).

The isolation of CSCs with specific markers is widely used but also this type of analysis has its

own limitations. As a fact, no single marker to isolate/identify the CSC has yet been found. The

first demonstration of the breast CSCs has been made by prospective immunophenotypic cell

isolation using the surface markers CD44 and CD24. In particular, CSC are enriched in the

CD44+/CD24low/- population, as demonstrated by generation of tumours in host mice when

injected in vivo. This immunophenotypic analysis is widely used but CD44 and CD24 cannot be

considered as universal markers, indeed their usage does not allow to selectively evaluate and

enrich for CSCs in ER- and triple negative breast cancers, as demonstrated by the fact that CSCs

were found in both the CD44+/CD24- and CD44+/CD24+ (Meyer et al., 2010).

Introduction

Ginestier and colleagues have discovered another breast CSC marker some years later. In

particular they unveiled that both breast normal and cancer human mammary epithelial cells with

increased aldehyde dehydrogenase activity (ALDH1) have stem/progenitor properties.

Moreover, the ALDH1+ cells isolated from human breast tumours contain the CSCs population

enabled with the capacity to self-renew and to generate tumours that recapitulate the original

heterogeneity of the breast cancer from which they derive. Moreover, by performing a double

flow-cytometry analysis of xenografted tumours with the stem cell markers ALDH1 and

CD44+/CD24-, they identified only a small overlapping cell fraction, representing 1.2% of the

entire population. This doubly recognized CSCs population displayed a higher tumourigenic

capacity with respect to the single populations positive for one of the two markers. Thus, a

double analysis of the same tumour population is strongly suggested to better define the CSCs

subpopulation (Ginestier et al., 2007).

By taking advantage of the knowledge derived from neurospheres, Dontu and her colleagues

established an in vitro assay called mammosphere assay. Analogously to primary neural cells,

human mammary epithelial cells form spherical colonies, termed non-adherent mammospheres,

when cultured in suspension on non-adherent surfaces in the presence of specific growth factors.

These mammospheres are enriched in cells with functional characteristics of stem/progenitor

cells. Transcriptional profiling of mammosphere-derived cells demonstrated significant overlap

with the genetic programs of other stem cells. The use of this culture system facilitates the

isolation and characterisation of mammary stem cells and it is useful to test their self-renewal

ability as well as to understand the pathways that govern their self-renewal and differentiation

(Dontu et al., 2003a). Almost all the research groups that have studied human and mouse stem

cells from both normal and tumour breast tissues have extensively used this assay. Such assays

have obvious caveats as factors important to the in vivo growth and self-renewal may not be

provided by the in vitro conditions. As a result cells may be selected anomalously, or they are

induced to differentiate or, more importantly, the true stem cells may be not selected. Moreover

while CSCs have an unlimited proliferative potential, their growth in vitro is limited to some

weeks/month. Therefore, results from mammosphere assays must be paralleled by in vivo

evidence (Valent et al., 2012).

More recently other markers of human normal and breast CSCs have been described. The work

was based on the intrinsic ability of CSCs to divide with lower frequency than their

differentiated progeny. PKH-26 is a fluorescent dye that binds to cell membranes and segregates

in daughter cells after each cell division, hence the intensity of staining inversely correlates with

Introduction

the number of cell divisions. Notably PKH-26high, slowly dividing cells were able to regenerate

both an entire mammary gland and mammospheres with high efficiency. Transcriptional

profiling of these cells, derived from human reductive mammoplasty, allowed determining a

normal breast stem cell specific signature able to predict biological and molecular features of

breast cancers. Indeed, in high-grade breast cancers the number of PKH26pos cells is increased,

indicating a higher CSCs content. By contrast, PKH-26neg, highly dividing cells, were unable to

generate mammary glands and mammospheres and they displayed markers of terminal

differentiation. Thus, PKH26pos cells are defined as stem cells (Cicalese et al., 2009; Pece et al.,

2010).

Taken one by one, all the techniques used to isolate and analyze stem cells display some

weaknesses; a combination of them will be more precise and clinically relevant for a better

evaluation of the CSCs population. In addition, the in vivo experiments are the most reliable ones

and may facilitate the future development of methods for interrogating CSCs in patient samples

for clinical purposes.

3.1.3.3 CSCs underlie heterogeneity of human breast cancers

Breast cancer inter/intra-tumour heterogeneity represents a major problem in defining the fitting

pharmacological regime for each patient. Indeed, despite classification of breast cancer subtypes

has a prognostic value and can properly suggest the better therapy for each single patient,

considerable heterogeneity in response to therapy still exists. As a result, there has been

considerable effort in the breast cancer research community to identify biomarkers that more

accurately predict patients outcome (Stingl and Caldas, 2007). For many reasons CSCs are

suggested as responsible for this heterogeneity, as follows.

- Cancer stem cells boost tumour formation and give rise to a progeny with the same

mutational features. Notwithstanding, CSCs and the more differentiated progeny display

different phenotypic traits, as demonstrated for example by a different sensitivity to

chemotherapeutic agents, contributing to intra-tumour heterogeneity. This aspect is even more

exacerbated if more differentiated cells within a tumour, due to their genomic instability and

increased proliferation, could acquire other epigenetic or genetic mutations. In this way the

tumour will display different cell clones concurring to the intra-tumour heterogeneity, as

depicted in Figure 3 (Stingl and Caldas, 2007; Kreso and Dick, 2014).

Introduction

Figure 3. Unified model of clonal evolution and CSCs Upper panel shows that acquisition of favorable mutations can result in clonal expansion of the founder cell. In parallel, another cell may gain a different mutation that allows it to form a new subclone. Over time, genetic mutations accumulate and subclones evolve in parallel. Lower panel shows that it may be that CSCs are not static entities but can evolve over the lifetime of a cancer as genetic changes can influence CSC frequency. Some subclones may contain a steep developmental hierarchy (left), where only few self-renewing CSCs exist among a large number of non-CSCs. Other subclones (middle) may contain an intermediate hierarchy, where the number of CSCs is relatively high but a hierarchy still exists. Some subclones may have the genetic alterations that confer high self-renewal potential, where most cells are tumorigenic. In this scenario, applying the CSC concept to such homogeneous subclones is not warranted because most cells can self-renew and few non-CSC progeny are generated (Kreso and Dick, 2014).

- It has been demonstrated that cancers can even harbour heterogeneous and biologically

distinct populations of CSCs, as shown for leukaemia, colorectal and skin tumours (Goardon et

al., 2011; Visvader and Lindeman, 2012). Intriguingly, distinct CSCs within a tumour have also

the potential to interconvert, as demonstrated in skin squamous cell carcinomas. Since the

leukaemia model is closely related to that of breast cancer, it is conceivable that also in breast

cancers different pools of CSCs coexist and are able to interconvert (Visvader and Lindeman,

2012).

- As mentioned above, there is no great overlap between CSCs from the same breast cancer,

when selected by different markers, suggesting presence of different types of CSCs highlighted

the presence of genes associated with stemness features and, notably, increased aggressiveness

and tumour grade can partially be ascribed to a higher CSCs content (Ben-Porath et al., 2008;

Pece et al., 2010). Hence, these findings imply that increased percentage and heterogeneous

CSCs may contribute to tumour heterogeneity explaining the different types of breast cancers.

Introduction

3.1.3.4 CSCs are responsible for the failure of chemotherapeutic treatment

Clinically, chemoresistance is probably the major problem for treatment of breast cancers. Many

studies have been carried out to better understand the molecular basis of chemoresistance in

tumours. Collectively, these studies provide evidence that even within a single genetic clone,

cancer cells are heterogeneous in their ability to survive chemotherapeutic insults, and point out

both CSCs and tumour heterogeneity as major responsible for the failure of chemotherapeutic

treatment (Figure 4) (Kreso and Dick, 2014; Dean, 2009; Dean et al., 2005).

In 1996 Goodell and colleagues, while staining murine bone marrow cells with the vital dye

Hoechst 33342, discovered a population that contained low -or none- levels of fluorescence

(Goodell et al., 1996). They called this population “side population” (SP), a “negatively stained”

pool of cells that displayed hematopoietic stem cells traits since, upon transplantation into

irradiated mice, they were able to reconstitute the bone marrow. Subsequent studies unveiled the

presence of SPs in many tissues including brain, breast, lung, heart, pancreas, testes, skin, and

liver that were enriched in stem cells. This technique leads to the identification of SP also in

tumours and in particular in breast, lung, and neural tumours. The fact that this SP does not

retain the dye is due to its intrinsic ability to exclude dyes through specific ABC transporters

(ABCB1; ABCG2; ABCC1), promiscuous transporters of both hydrophobic and hydrophilic

compounds such as dyes and drugs. Interestingly the SP in mammary tissues identify a stem cell

population, suggesting that stem cells could promote efflux of dyes and, more importantly,

drugs, underling a related chemoresistant properties of these cells. Indeed CSCs expressed high

levels of these drug efflux pumps and are resistant to chemotherapy (Sarkadi et al., 2006).

The CSCs chemoresistance goes beyond the presence of ABC transporters. Although it is well

known that chemoresistance could arise in any cell of a cancer through several mechanisms such

as mutation, such features are more likely to be already present in CSCs. Indeed, CSCs, like

normal stem cells, are relatively quiescent, resistant to apoptosis and they possess an efficient

DNA-repair capacity (Dean et al., 2005).

It is evident that CSCs are the main responsible for the failure of chemotherapeutic treatment,

since they already intrinsically possess a plethora of chemoresistance mechanisms. Even worse,

despite treatment of breast cancer patients with chemotherapy causes an initial shrinkage of the

tumour mass, a concomitant increase in the percentage of resistant cancer stem cells was

observed. Notably, in vitro and in vivo experiments with cells obtained from these treated

tumours demonstrated that these cells eventually regenerated an even more aggressive tumour,

Introduction

phenocopying tumour recurrence after therapy (Figure 4) (Kreso and Dick, 2014; Dean, 2009;

Dean et al., 2005; Korkaya et al., 2009; Yu et al., 2007).

Figure 4. CSCs within subclones impact response to therapy Each clone (depicted by the different colors) contains a mixture of cells that vary with respect to their stemness and/or proliferative ability, including dormant or CSCs (depicted by a central dot in the cell). Together these factors represent the functional diversity present within single genetic subclones. Chemotherapy can reduce tumor burden by eliminating the highly proliferative cells within subclones, while sparing the CSCs; following therapy, these cells can seed a new cancer. Thereby, subclonal diversity can be altered with chemotherapy and can allow for the selection of cells with additional genetic mutations that confer a survival advantage. Adapted from Kreso and Dick, 2014.

Introduction

3.1.3.5 EMT and metastases are linked to CSC traits Aggressive and deadly tumours are characterized by metastatic dissemination and outgrowth of

drug resistant tumours. The ability of the infiltrated tumour cells to survive during latency and to

reinitiate growth at the secondary site, when conditions are favourable, was reminiscent of cells

having a CSC phenotype.

The group of Weinberg found that both normal and cancer stem cells of the breast display

features of epithelial-to-mesenchymal transition (EMT) features. EMT is a process by which

epithelial cells lose their cell polarity and become mesenchymal cells to gain migratory and

invasive properties. This process was first recognized as indispensable for embryogenesis.

Indeed EMT, and its reverse process, the Mesenchymal-to-epithelial transition (MET) are critical

for development of many tissues and organs in the developing embryo (Kong et al., 2011). The

EMT is also involved in cancer, in particular during the metastatic process. Interestingly over-

expression of EMT transcription factors (such as Slug, Snail and Twist) or the down-regulation

of epithelial markers (such as E-cadherin) not only induce an EMT program but also cause a

concomitant enrichment in stem cells, as detected by different in vitro assays (Mani et al., 2008).

In addition, enforced EMT in terminally differentiated mammary cells promotes a

reprogramming to a bipotent stem cell able to generate an entire and functional mammary gland

Accordingly, mammary stem cells have been shown to display mesenchymal traits instead of

epithelial ones (Guo et al., 2012).

The precise identification of the molecular bases that govern the self-renewal of normal

mammary stem cells is important because these same pathways are thought to be required for the

induction and maintenance of the CSCs and their aggressive phenotype. Inhibition of such

signalling pathways has been proposed to be a novel and promising therapeutic strategy. In

particular there is a considerable effort in finding therapies that target pathways, mutations or

histological features that are common to all CSCs, overcoming breast cancer heterogeneity. A

possible candidate as a fine tuner of these stemness traits is the prolyl-isomerase Pin1.

Introduction

3.2. The prolyl-isomerase Pin1: a critical hub of phosphorylation-dependent

pathway cross-talk

Post-translational modifications (PTMs) represent a milestone process in biology (Prabakaran et

al., 2012; Walsh, 2006). They are involved in many cellular processes, such as the control of cell

cycle, apoptosis, transcription, cell commitment, metabolism and DNA repair. There are

thousands of PTMs but, interestingly, phosphorylation on serine/threonine residues (phospho-

S/T) is the most prevalent one and it is a common way to process information in cells (Khoury et

al., 2011). Phosphorylation of a protein involves the enzymatically, and reversible, mediated

addition of a phosphate group to its amino acid side chains. In particular, phosphorylation of

serine or threonine preceding a proline, represents a key signalling mechanism involved in many

physiological and pathological processes (Lu and Zhou, 2007). Phosphorylation of these residues

is a prerogative of Proline-directed protein kinases including glycogen synthase kinase-3, p38

kinases, extracellular signal-regulated kinases and cyclin-dependent protein kinases. Proteins

that have S/T-P sites phosphorylated adopt either a cis or a trans conformation, which implies

that the same protein can perform different functions. The spontaneous conversion from one

isomer to the other, i.e. uncatalyzed cis-trans isomerization is normally very slow and

phosphorylation per se further decreases the already slow isomerization rate of prolines

(Schutkowski et al., 1998). This reaction needs the intervention of the peptidyl-prolyl cis/trans

isomerase Pin1 for a biologically relevant timescale cis/trans conversion to occur (Hunter, 1998).

Pin1 belongs to the parvulin family of Peptidil Prolyl cis/trans Isomerases (PPIases). PPIases are

a class of enzymes that catalyze conformational changes centered around proline residues and

that include three structurally distinct subfamiles: the cyclophilins, the FK506-binding protein

family and the parvulins (Galat, 2003). Notably, the only known human isomerase that

specifically regulates the cis-trans conformation of Prolines only upon phosphorylation of the

preceeding Serine or Threonine residues, is the parvulin Pin1 (Lu et al., 1996; Ranganathan et

al., 1997; Yaffe et al., 1997).

3.2.1 Pin1 structure and functional domains

Human Pin1 is a 18KDa protein of 163 amino acids and its gene is located in chromosome

19p13. Structurally, it is composed of two functional domains: an amino-terminal WW domain

(amino acids 1-39) and a carboxy-terminal PPIase domain (amino acids 45-163) connected by a

Introduction

short flexible linker region (Figure 5). For recognition and interaction with its targets on

phospho-S/T-P sites, Pin1 requires the WW domain that is characterized by two conserved

tryptophan residues and organized in a three-stranded antiparallel antiparallel β-sheet (Sudol and

Hunter, 2000; Sudol et al., 2001). Subsequently the PPIase can interact with the phospho-S/T-P

motifs and induce a cis-trans isomerisation of the peptide bond between the phosphorylated

residue and the Proline. This enzymatic activity is mediated by a specific basic triad, composed

by Lys63-Arg68-Arg69, that recognizes the negative charge present in the phospho-residues

(Zhou et al., 2000).

Figure 5. Pin1 structure and function The two-domain structure of Pin1 is shown in the upper part of the figure. The lower part of the figure shows the conversion of the target protein from cis to trans conformation and vice-versa (Yeh and Means, 2007).

Pin1 was first identificated by a two-hybrid screening performed to identify proteins that interact

with never in mitosis gene A (NIMA), a fundamental mitotic kinase in Aspergillus nidulans (Lu

et al., 1996). Since overexpression of NIMA in yeast induces premature chromosome

condensation and subsequently cell death and concomitant overexpression of Pin1 prevents this

effect, it was suggested that Pin1 might be a regulator of mitosis.

Pin1 is an evolutionarily conserved protein and was found to be fundamental in S. cervisiase, C.

albicans and A. nidulans but not in X. laevis, D. melanogaster, M. musculus where its depletion

is linked to some developmental defects (Crenshaw et al., 1998; Fujimori et al., 1999; Hanes et

Introduction

al., 1989; Hsu et al., 2001; Maleszka et al., 1996; Winkler et al., 2000). Pin1 knock-out mice

(Pin1-/-) have been generated in two different backgrounds. In the mixed background

129SvJae/C57BL/6, Pin1-/- mice diplay premature aging and tissue defects, in particular

decreased body weight, testicular atrophy, a dramatic retinal atrophy and alteration of the

mammary gland that fails to undergo to proliferation during pregnancy (Liou et al., 2002). Mice

in the inbred C57BL/6 background, display a reduced number of germ cells due to an alterated

cell cycle in primordial germ cells (PGC), causing a decreased fertility both in males and females

(Atchison and Means, 2003).

3.2.2 Regulation of Pin1 levels and activity

Even though no amplification or deletion of the PIN1 gene have been reported, the levels and

functions of the protein are finely regulated by several layers of control under physiological and

also pathological conditions. Some polymorphisms have been discovered, one of them,

rs2233678 (-842G>C) in the Pin1 promoter that elicits reduced expression of the gene and is

associated with a reduced risk of cancer in some human sub-populations (Han et al., 2010).

Notably Pin1 expression is strongly associated with cell proliferation and cell cycle progression,

since it is transcriptionally regulated by the E2F transcription factor in response to growth factor

stimuli, Ras, and Her2/Neu (Ryo et al., 2002). In breast cancer cells, its mRNA levels are also

upregulated by the Notch signalling pathway (Rustighi et al., 2009) and the Insulin-like growth

factor (You et al., 2002). Accordingly, Pin1 mRNA and protein levels are increased in

transformed cells (Bao et al., 2004). Recently, it has been reported that the apolipoprotein E4 is

able to positively modulate Pin1 expression in the hippocampus and that this ApoE-Pin1 axis is

involved in the pathogenesis of the Alzheimer disease (Lattanzio et al., 2014). Pin1 is also

controlled by micro-RNA, in particular it is negatively controlled by miR-200b that promotes

anoikis in breast cancer cells and is able to block metastatic progression (Zhang et al., 2013). It

has become evident how pathways involved in pathologies such as cancer and Alzheimer

mediate misregulation of Pin1 transcription. Consequently, Pin1 protein and mRNA levels are

considered a prognostic and predictive marker for many diseases (Lu and Zhou, 2007).

At a post-translational level, the enzymatic activity of Pin1, its cellular localization and turnover

are finely regulated. Indeed a lot of kinases affect Pin1 functions by modifying its

phosphorylation status (Liou et al., 2011). Activation of PKA phosphorylates Pin1 on Ser16 in

the WW domain and inhibits its function by impairing its ability to interact with protein target

and changing the sub-cellular localization (Lu et al., 2002b). This inhibitory phosphorylation on

Introduction

Ser16 is mediated also by treatment with 12-O-tetradecnoylphorbol-13-acetate (TPA), that

promotes an interaction between the 90-kDa ribosomal protein-S6-kinase 2 and Pin1 (Cho et al.,

2012). Another inhibitory phosphorylation occurs on Ser71 in the PPIase domain, and it is

mediated by the Death-associated protein kinase1 (DAPK1). This phosphorylation blocks Pin1’s

catalytic activity and nuclear localization (Lee et al., 2011). Despite a lot of kinases have an

inhibitory effect on Pin1 activity, the mixed-lineage kinase 3 (MLK3) instead promotes Pin1

catalytic activity and nuclear localization by adding phospate group to the Ser138 in the PPIase

domain in breast cancer (Lee et al., 2011; Rangasamy et al., 2012). It is conceivable that an

appropriate evaluation of the phosphorylation status of Pin1 in tissues will be more predictive of

its activity rather than solely total protein levels.

SUMOylation is another important PTM that affects Pin1’s activity, in particular it inhibits its

functions when it occurs on Lys6 or Lys63. This inhinbitory effect is reverted by

deSUMOylation by SENP1 which, in turn, increases Pin1 protein levels (Chen et al., 2013).

3.2.3. Biological functions of Pin1

As a consequence of its enzymatic nature, Pin1 requires a substrate to impact in any given

cellular process. The peculiarity of a Pin1 target is the presence of a docking site constituted by a

phosphorylated-Ser/Thr preceeding a Proline, which represents a key element in the signal

tranduction. Indeed Pin1 interacts with and modifies a large number of proteins, impinging on

several cellular processes such as cell cycle control, transcription, chromatin remodelling, cell-

fate commitment, DNA damage response, and metabolism. A growing body of evidence

demonstrates that on one hand Pin1 acts as a hub that controls the dynamics of physiological cell

behaviour by fine-tuning the cross-talk between phosphorylation signalling and cellular

pathways. But on the other, its dysregulated expression in diseased conditions such as cancer,

critically amplifies perturbed cell signalling, contributing to disease progression.

3.2.3.1 Mechanisms through which Pin1 impinges on cellular processes

The conformational changes induced by Pin1 on its protein substrates can have profound effects

mainly on their stability, but also modulation of the catalytic activity, other PTMs like de-

phosphorylation, protein–protein interactions and subcellular localization have been described

(Lu et al., 2002a).

Pin1 has been shown to control protein turnover of many substrates by either favouring or

blocking their recognition by several E3 ubiquitin-ligases. The phosphorylation of Ser/Thr-Pro

Introduction

constitutes an important motif not only for Pin1 but also, when embedded within the socalled

CDC4-phosphodegron, for specific recognition by the Skp1-Cullin-F-Box (SCF) complex that

promotes the ubiquitination-dependent degradation of critical cellular substrates. Consequently

Pin1 plays a major role in modulating the conformation of these proteins thereby affecting their

interaction with the SCF complex and proteasomal degradation (Gutierrez and Ronai, 2006; Liou

et al., 2011).

Fbwx7α is a component of SCF complex and is also known as CDC4, AGO, SEL10. It is an

important E3 ubiquitin-ligase that controls many cellular processes such as cell proliferation,

stemness and differentiation, metabolism and apoptosis (Buckley et al., 2012; Welcker and

Clurman, 2008; Wertz et al., 2011). It regulates a plethora of targets in many tissues such as

Notch1, Notch4, c-myc, cyclin E, c-jun, MCL-1 and SREBP (Cheng and Li, 2012; Welcker et

al., 2004). Fbwx7α recognizes a phospho-epitope, termed CPD (for Cdc4 Phospho-Degron),

contained within these substrates. Via its F-box domain, Fbwx7α recruits the remainder of an

SCF ubiquitin ligase complex, thus promoting substrate ubiquitination and proteasomal

degradation. The eight WD40 repeats at the C-terminal end form a phospho-epitope binding

pocket that can recognize and bind to phosphorylated CPDs. CPDs consist of a central

phosphorylated threonine immediately followed by a proline (pT-P). Moreover a priming

phosphate in the +4 position is required, that is provided by a phosphorylated serine, or a

phospho-mimetic residue, in order to phosphorylate the central threonine (Welcker and Clurman,

2008).

Once phosphorylated, the CPD might become a Pin1 binding site. Indeed, many of the Fbwx7α

targets are also Pin1 substrates and, consequently, Pin1 is impinging on their ubiquitination and

protein degradation mediated by this ubiquitin-ligase (Liou et al., 2011). Notably the role of Pin1

on the Fbxw7-mediated protein degradation is context- and tissue-dependent. Indeed it has been

shown that on one hand Pin1 promotes protein stabilization of some targets, such as Mcl-1, on

the other it promotes the degradation of c-Myc and cyclin E (Welcker et al., 2004). In addition it

has been recently demonstrated, in colon cancer cells and fibroblasts, that Pin1 promotes auto-

ubiquitination and degradation of Fbwx7 (Min et al., 2012). Thus the effect of Pin1 on Fbxw7

targets seems not to be unidirectional.

Besides the Pin1-Fbwx7α dualism in the control of protein degradation, Pin1 is able to control

the stability of many other proteins, such as cyclin D1, NF-κB, β-catenin, p53 family members,

Nanog, Oct4, ErbB2, Akt, Smad proteins and Tau by different mechanisms (Liou et al., 2011).

Alteration of their protein levels was shown to affect their activities and their ability to interact

and cross-talk with their partners and regulators. As a result Pin1, through modulation of protein

Introduction

stability, is able to impinge on several cellular processes. The best-studied biological function of

Pin1 is undoubtedly regulation of cell cycle progression. Important roles for this enzyme in the

control of a variety of other phosphorylation-dependent processes have been unveiled, some of

which have been studied by us and are described in depth below.

3.2.3.2 Cell-cycle control

Pin1 was discovered as a regulator of mitosis (Lu et al., 1996; Yeh and Means, 2007). Further

studies unveiled a role for Pin1 in all phases of the cell cycle. In fact, during cell cycle

progression there is a precise timing and sequence of activation and inactivation of CDKs which

are known to be phosphorylated on Ser/Thr-Pro motifs and, indeed, Pin1 fits perfectly in this

molecular organization (Lu and Zhou, 2007; Nigg, 2001). The G0/G1-S phase transition dictates

the entrance in the cell cycle and a key regulator of this passage is cyclin D1. Notably Pin1

increases cyclin D1 levels in a bymodal manner, i.e. preventing its protein degradation and

upregulating its trascription (Liou et al., 2002; Wulf et al., 2001). Another important step that

occurs during the G1-S phase, is the coordination of centrosome duplication and DNA synthesis.

Pin1 is localized in the centrosomes and induces their duplication; in breast cancer Pin1 over-

expression promotes chromosome aneuploidy as a consequence of deregulated centrosome

amplification. Conversely, reduction of Pin1 levels strongly impaired centrosome amplification

(Suizu et al., 2006). Accordingly, its protein levels are higher during the G1-S transition (Ryo et

al., 2002) while Pin1-/- mouse embryonic fibroblasts exhibit defects in the G1-S transition

(Fujimori et al., 1999). The role of Pin1 in the regulation of G1-S transition is more complex

since it promotes the degradation of cyclin E at a later stage of the S phase (van Drogen et al.,

2006), thus suggesting that Pin1 is controlling the amplitude of the G1-S transition.

During the S phase the RNA polymerase II is recruited on the promoters of transcribed genes

while it is released in M phase. Notably Pin1 coordinates this dynamic association by directly

binding the RNA polymerase II (Xu and Manley, 2007a). In particular, Pin1 controls also the

G2-M transition, acting on the early mitotic inhibitor-1 and, consequently, on anaphase

promoting complex (APC) thus coordinating the S and M phases (Bernis et al., 2007).

Mitosis is promoted by cyclin B-CDC2 that, also in this case, is controlled by Pin1. In this case

Pin1 indirectly controls cyclin B, acting on CDC25C and WEE1, regulators of the cyclin B-

CDC2 phosphorylation status (Okamoto and Sagata, 2007; Zhou et al., 2000). The impact of

Pin1 in the M phase emerged also by the evidence that it interacts with topoisomerase IIα

impinging on the G2-M phase transition (Xu and Manley, 2007b).

Introduction

All this evidence supports the idea that Pin1 is a fine tuner of the cell cycle progression in which

it calibrates the correct timing and quantity of factors that trigger initiation or termination of each

single phase.

3.2.3.3 Pin1 at the crossroads between stemness and differentiation

The normal development of organisms and the maintenance of tissue integrity are guaranteed by

a correct balance between differentiation and stemness (Chen and Daley, 2008; Keller, 2005;

Yeo and Ng, 2013). During embryogenesis, the proper timing and specification of differentiation

stimuli is fundamental for the commitment of embryonic stem cells to generate the three

embryonic germ layers. A lot of studies have been done to understand the molecular basis of

embryonic stem cells’ biology (ESCs) in light of a better comprehension of embryogenesis, of

the pathways required by the CSCs and, recently, for their application in regenerative medicine

(Yeo and Ng, 2013). In particular, manipulation of these molecular circuitries allowed the

conversion of adult differentiated cells in pluripotent stem cells, called induced pluripotent stem

cells (iPS) (Takahashi and Yamanaka, 2006).

It has been demonstrated that, beyond the up-regulation of important pro-stem factors such as

Oct4, Nanog and Sox2 by transcriptional regulation and chromatin remodelling (Yeo and Ng,

2013),their post-translational modifications, in particular phosphorylation and ubiqitination, are

essential for a correct shift from a pluripotent to a differentiated fate and viceversa (Brill et al.,

2009; Buckley et al., 2012; Van Hoof et al., 2009). These evidences suggest that the ESC

regulation is a plastic process governed by several layers of control in which small changes in

ESC regulators are sufficient for a complete cell reprogramming (Yeo and Ng, 2013). In this

scenario, the prolyl-isomerase Pin1, an important fine tuner of gene expression, protein levels

and post-translational modifications (Liou et al., 2011; Lu and Zhou, 2007) emerged as an

essential regulator of pluripotency and stemness both in ESC and iPS cells (Moretto-Zita et al.,

2010; Nishi et al., 2011). In particular Pin1 is indispensable for the induction and maintenance of

pluripotency in iPS by interacting with phosphorylated Oct4, preventing its ubiquitination,

boosting its protein levels and activity. Accordingly, Pin1 levels are higher in iPS and ablation of

its activity suppresses colony formation in murine ESC cells (Nishi et al., 2011). Similarly, Pin1

levels are higher in ESCs and its genetic or pharmacological ablation curbs their clonogenic

potential. In ESCs Pin1 is boosting stemness by preventing the degradation of Nanog (Moretto-

Zita et al., 2010).

Introduction

The impact of Pin1 in cell fate commitment carries several clinical implications. In particular,

the major problem of iPSs is their intrinsic tumourigenic potential that has prevented their use in

regenerative medicine (Takahashi and Yamanaka, 2006). This tendency could be controlled by

modulation of Pin1 protein levels (Moretto-Zita et al., 2010; Nishi et al., 2011).

The regulation of the transition from an undifferentiated stem cell to a differentiated state and

viceversa mediated by Pin1 is maintained also in adult tissues with several consequences for the

organogenesis. It has been demonstrated that Pin1-/- mice develop normally but display some

defects in many compartments (Atchison and Means, 2003; Fujimori et al., 1999; Liou et al.,

2002; Shen et al., 2013). Different studies unveiled that these defects might be partially

explained by a role of Pin1 as a master regulator of the cell commitment in many cellular types

such as neurons (Ciarapica et al., 2014; Nakamura et al., 2012), skeletal muscle cells (Magli et

al., 2010), primordial germ cells (Atchison and Means, 2003), adipocytes (Uchida et al., 2012),

osteoblasts (Lee et al., 2013a)and dental pulp cells (Lee et al., 2013b). All this evidence

pinpoints Pin1 as a cell-fate determinant.

Interestingly, in adult tissues, Pin1 is not always linked to the maintenance of stem cells. Indeed,

while in skeletal muscles it sustains the maintenance of undifferentiated cells, in the brain Pin1

drives neuronal differentiation. In particular two different groups observed that Pin1 drives

neuronal differentiation and its protein levels are higher in differentiated neurons during

embryonic development, compared to less differentiated ones, without perturbing glia cells’

commitment (Ciarapica et al., 2014; Nakamura et al., 2012). This effect is mediated in a bimodal

fashion: on one hand Pin1 interacts with β-catenin, a master regulator of neurogenesis, boosting

its protein levels, on the other it cooperates with the protein kinase HIPK2 in antagonizing

Gro/TLE1:Hes1, a protein complex involved in the inhibition of neuronal differentiation. As a

result Pin1-/- mice display impaired neonatal motor activity.

The ability of Pin1 in the regulation of the cell-fate commitment is not only restricted to

alterations of its proteins levels but, in some cases, also through its sub-cellular localization

(Magli et al., 2010). In skeletal muscle cells Pin1 is localized in the nucleus and acts as an

inhibitor of differentiation by promoting degradation of the Myocyte Enhancer Factor 2C

(MEF2C). Interestingly, during muscle cells’ differentiation, there is a shuttling of Pin1 from the

nucleus to the cytoplasm with a concomitant rescue of the stability and transcriptional activity of

MEF2C that, consequently, induces a differentiation program.

Introduction

Based on these data, Pin1 can be clearly considered as a cell-fate determinant in the adult tissue,

therefore alteration of its levels, activity or sub-cellular localization could compromise the

integrity of many tissues.

3.2.3.4 Pin1 pinpoints the cellular response to environmental stresses

Cells exposed to genotoxic stress, chemical agents, or stimuli from the enviroment, undergo an

adaptive response involving adjustments of the levels and activity of its genome-protecting

protein machinery. When cell cycle checkpoints are switched on and stimulate DNA repair, cell

cycle arrest, apoptosis or other processes are induced, depending on the stimulus and cell type.

The p53 family members, in particular p53 and p73, are important tumour suppressors and are

the major mediators of the cellular response to genotoxic insults and pharmacological treatment.

Both p53 and p73 are subjected to proline-directed phosphorylation and are Pin1 targets

(Mantovani et al., 2004; Mantovani et al., 2007; Zacchi et al., 2002; Zheng et al., 2002). In

detail, Pin1 is able to prevent the interaction of p53 with its ubiquitin ligase Mdm2 thus enhances

its half-life and, consequently, it increases cellular apoptosis upon genotoxic stress. The ability

of Pin1 to potentiate p53 goes even beyond: on one hand Pin1 is required for a correct loading of

p53 on its target genes and on the other it prevents the association of the apoptosis inhibitor

iASPP (Mantovani et al., 2007). In addition, it has been recently demonstrated that Pin1,

cooperating with HIPK2, promotes both p53 mitochondrial translocation and its transcriptional-

independent apoptotic activity (Sorrentino et al., 2013). Moreover Pin1 boosts also p73 protein

levels and transcriptional activity since it promotes p73 acetylation and subsequent protein

stabilization; this circuitry has an important role in tumours cells lacking p53 and treated with

chemotherapeutic agents (Mantovani et al., 2004). Interestingly Pin1 is not only able to control

the DNA controllers, but it is directly involved in the regulation of DNA integrity by modulation

of DNA double strand break (DSB) repair mechanisms in response to DNA damage. In

particular Pin1 promotes non-homologous end-joining (NHEJ) at the expense of homologous

recombination (HR), directly impinging on the DSBs resection (Steger et al., 2013).

The in vivo impact of these findings is still not so clear since many of these studies have been

performed in primary fibroblasts and cells lines. It will be important to assess this issue, since

Pin1 is often up-regulated in tumours (Bao et al., 2004) while p53 tumour suppressor activity is

frequently abolished (Levine and Oren, 2009). The fact that Pin1 is required for a more efficient

p53-dependent apoptosis following genotoxic stress may have clinical relevance in tumors with

Introduction

concomitatnt high Pin1 levels and wild-type p53 status, since in these tumors its activity might

be restored (Hoe et al., 2014).

Pin1 is not only able to drive the cellular response in genotoxic stress conditions, but also

following oxidative stress. One of the major cellular mediators of this process is p66shc, which

regulates cellular aging, apoptosis induction and the lifespan. As a consequence the knock-out

mice models display longer life span and protection from oxidative stress (Migliaccio et al.,

1999). It has been demonstrated that Pin1 potentiates the mitochondrial accumulation of p66shc

thus inducing apoptosis in response to ROS (Pinton et al., 2007).

Thus, Pin1 is clearly an important modulator of the cellular response to environmental stresses

acting through proteins of the DNA checkpoint and repair machinery and mediators of oxidative

response.

Tumour chemoresistance is emerging as a major problem for the eradication of tumors. Despite

clinical advances in the last years, chemoresistance still remains a major cause of death and an

unresolved issue. The resistance could be intrinsic, since some tumour cells do not respond

properly to the chemoterapeutic regime, or acquired, since they are initially responsive to the

chemotherapy but become resistant later. In both cases cells display common features such as

alteration of the pharmacological target, activation and expression of proteins that reduce the

intracellular drug concentration, acquisition of the ability to change the cellular response by

alterating the apoptotic pathway or DNA repair machinery, or they being tolerant to genotoxic

stresses (Dean et al., 2005).

Pin1 has been directly or indirectly linked to chemoresistance. First of all, Pin1 is overexpressed

in tumours and its protein levels correlate with aggressiveness. Since the most aggressive

tumours are often less responsive to chemotherapy it is conceivable that Pin1 is linked to

chemoresistance (Ayala et al., 2003; Bao et al., 2004; Fukuchi et al., 2006; Wulf et al., 2001).

Second, Pin1 is more expressed in some stem-cell subpopulations (Magli et al., 2010; Moretto-

Zita et al., 2010; Nishi et al., 2011) which are chemoresistant by definition (Visvader and

Lindeman, 2012) thus, as consequence, Pin1 is linked to a chemoresistant subpopulation. Third,

Pin1 interacts with several factors that have been separately linked to chemoresistance such as

mutant-p53 (Muller et al., 2014), Notch1 (Rustighi et al., 2009), Mcl-1 (Ding et al., 2008),

Nanog and Oct4 (Hu et al., 2010; Moretto-Zita et al., 2010; Nishi et al., 2011). Fourth, in breast

cancer cell lines Pin1 has been directly linked to specific chemoresistance mechanisms (Ding et

al., 2008). Indeed, it was discovered that Pin1 protects cancer cells from paclitaxel treatment,

one of the most used and effective chemotherapeutic agents against breast cancer, through up-

Introduction

regulation of Myeloid cell leukemia-1 (Mcl-1), a Bcl-2–like antiapoptotic protein, involved in

chemoresistance and tumourigenesis in several cancers. In particular, Pin1 is able to bind the

Mcl-1 protein, upon phosphorylation by ERK, preventing its degradation and sustaining its

protein levels. Notably Pin1 and Mcl-1 protein levels positively correlate in breast cancer patient

samples. Despite all these publications, a real in vivo evidence of Pin1 in regulating the cellular

response to environmental stresses is still missing.

This bivalent and opposed role of Pin1 in cooperating both with proteins that promote apoptosis

or protect cells from chemotherapy is due to its enzymatic nature and its biological output is

strongly dependent on the availability of its targets. Specifically, tumours often display down-

regulation or loss of tumour-suppressors and concomitant activation of oncogenes and survival

pathways, thus upsetting Pin1 to a more anti-apoptotic and tumour promoting role.

3.2.4. Pin1 in cancer

The first evidence of Pin1 in tumours has been reported in 2001 by two different publications in

which Pin1 was found overexpressed in breast cancer (Ryo et al., 2001; Wulf et al., 2001). Later,

high levels of Pin1 have been found in other tumours such as prostate, lung, colon, esophageal,

human oral squamous cell, glioblastoma, ovary, cervical and melanoma cancers (Atkinson et al.,

2009; Bao et al., 2004; Jin et al., 2011; Miyashita et al., 2003). All these studies indicated a

prominent role of Pin1 in boosting tumourigenesis through up-regulation of key oncogenic

pathways (Lu and Zhou, 2007). Pin1 was revealed not only as a marker of tumourigenesis but it

was also shown to correlate with poor prognosis in some cancer types (Ayala et al., 2003;

Fukuchi et al., 2006; Girardini et al., 2011; Tan et al., 2010; Wulf et al., 2001).

The first interactor of Pin1described in cancer is cyclin D1 as demonstrated by in vivo and in

vitro experiments (Wulf et al., 2001). The nuclear localization and proteasomal degradation of

cyclin D1 are controlled by a phosphorylation on Thr-286-Pro that is promoted by glycogen

synthase kinase-3beta (GSK-3β). Pin1 was shown to bind to this motif upon phosphorylation,

preventing nuclear export and subsequent degradation, maintaining high cyclin D1 protein

levels. Notably Pin1 knock-out and cyclin D1 knock-out mice share similar phenotypes in

particular in the mammary epithelium, where both fail to undergo to the massive proliferative

change during pregnancy (Liou et al., 2002).

The ability of Pin1 in boosting cyclin D1 levels goes even beyond an effect on protein

stabilization. Indeed it has been demonstrated that the up-regulation of cyclin D1 induced by

several signalling pathways such as the Ras, Her2/Neu, Wnt and NF-κB, requires the presence of

Introduction

Pin1 which in turn is indispensable for full activity of these pathways (Liou et al., 2002; Lu and

Zhou, 2007; Ryo et al., 2001; Ryo et al., 2003; Wulf et al., 2001). Pin1 can bind to

phosphorylated c-Jun, β-catenin and NF-κB boosting their protein levels and transcriptional

activity that is translated in high cyclin D1 gene expression. Moreover Pin1 is able to protect β-

catenin and NF-κB from their negative regulators, APC (adenomatous polyposis coli gene

product) and IkappaB, respectively.

Besides activation of growth factors receptors, one of the most important signalling pathways

activated in cancer is the mitogen-activated protein kinase (MAPKs) cascade. In response to

growth stimuli, Ras activates the Raf kinase, which in turn controls the MAPKs. As for other

signalling pathways, this kinase cascade must be turned off by a negative feedback mechanism

in which MAPKs phosphorylate and inactivate Raf. Notably Pin1 prevents this negative

feedback, promoting Raf dephosphorylation with a consequent maintenance of an activated

MAPKs cascade (Dougherty et al., 2005).

All these data point out Pin1 as a fine tuner of signalling pathways in cancer, regulating their

activation, counteracting their inhibitors and thereby boosting their mediators. The proof that

Pin1 is strongly required by these pathways emerged also by the discovery that, in turn, they

induce Pin1 gene transcription and that genetic ablation of its gene block, in vivo, their

oncogenic potential. In particular, Her2/Neu and Ras induce E2F transcription and progression

of the cell cycle; E2F is then recruited on the Pin1 gene promoter inducing its transcription (Ryo

et al., 2002). Since E2F is not only regulated by these pathways but it is a marker and promoter

of the G1/S transition, it is conceivable that whatever the stimulus that induces cell cycle

progression, Pin1 transcription can be induced.

Several reports have shown that Pin1 is also linked to aggressive traits of cancers. Indeed it has

been demonstrated that Pin1 protein levels are higher in high-grade and more aggressive breast

cancer subtypes (Wulf et al., 2001). In particular in triple negative breast cancer cell lines, Pin1

cooperates with two other important oncogenes such as Notch1 (Rustighi et al., 2009) and

mutant-p53 (Girardini et al., 2011). Notch1 is a membrane bound heterodimer receptor, normally

inactive, that becomes cleaved and activate by γ-secretase upon interaction with its specific

ligand (Ranganathan et al., 2011). Pin1 interacts with the Notch1 receptor and increases the

cleavage mediated by the γ-secretase, allowing full activation of the pathway and boosting its

oncogenic activity both in vitro and in vivo. Also in this case the oncogene stimulated by Pin1 is

able in turn to transcribe Pin1 itself, since Notch1 is directly recruited on the Pin1 gene

Introduction

promoter. As a consequence, in human breast cancer samples there is a strong correlation

between high levels of activated Notch1 and Pin1 overexpression (Rustighi et al., 2009).

Pin1 has also been linked to mestastasis, aggressiveness and poor prognosis in triple negative

breast cancer through cooperation with mutant-p53, following oncogenic stress. In particular it

has been described that Pin1 binds to a phosphorylated mutant-p53 in breast cancer cell lines

where it co-regulates a transcriptional program, that promotes migration and metastasis

formation in vivo and correlates with poor prognosis in breast cancer patients and. Ablation of

Pin1 lowers the oncogenic gain of function of mutant-p53 with reduced formation of breast

cancer metastases into the lung, suggesting interesting therapeutic opportunities (Girardini et al.,

2011).

Another important Pin1 target involved in the metastatic process is the focal adhesion-associated

non-receptor protein-tyrosine kinase (FAK) involved in cellular adhesion and cell migration.

FAK becomes phosphorylated following oncogenic stress by a Ras-ERK axis on Ser910. This

PTM recruits Pin1 on the FAK protein causing an association with the protein-Tyr phosphatase

(PTP) (PTP-PEST) that in turn dephosphoryates FAK on Th397 inhibiting its activity. The result

is the promotion of a migratory and invasive phenotype (Zheng et al., 2009).

Besides all the evidence that Pin1 is required for full activation of oncogenic pathways, it’s

conceivable that its effects go even beyond. Phosphorylation plays a fundamental role in

regulating many intracellular processes such as growth, proliferation, cell division, apoptosis and

differentiation. It is clear that perturbations of the phosphorylation status in a cell are likely to

drive many of the hallmarks of cancer. Thereby, many cancers display deregulation or

mutational events in kinases and phosphatases, and many of the genes encoding for these

proteins are oncogenes or tumour suppressors. Accordingly, phosphorylation sites mutated by

single nucleotide variants are found in almost 90% of tumours and, moreover, 29% of these

mutations directly abolish phosphorylation or modify kinases’ target sites (Reimand et al., 2013).

It is evident how Pin1 is strongly involved in this scenario, as a consequence of its way of action.

It is unsurprising that modulation of Pin1 activity and levels could be comparable to deregulation

or mutation in kinases and phosphatases, strengthening the idea that Pin1 is an oncogene.

Accordingly, transgenic overexpression of Pin1 in mouse mammary glands leads to mammary

hyperplasia and malignant mammary tumours (Suizu et al., 2006).

Introduction

3.3. The Notch signalling pathway

The Notch pathway is a conserved signalling route indispensable for the development of

multicellular organisms, based on interaction with ligands expressed on neighbouring cells that

are required for its proper activation. Signals that activate this pathway boost initiation and

amplification of molecular differences that promote the cell fate commitment (Artavanis-

Tsakonas et al., 1999). Activation of the pathway promotes the release of the nuclear activated

form of Notch that translocates into the nucleus e transactivates several target genes. This

pathway is involved in the regulation of many cellular processes such as proliferation, cell death,

metabolism, angiogenesis, stemness and cell fate specification during development and renewal

of adult tissues. As a consequence, alteration of this pathway is associated with pathological

situation, such as developmental defects, cardiovascular disease and cancer (Bray, 2006;

Ranganathan et al., 2011).

3.3.1. Notch receptors and ligands

In 1917, Thomas Hunt Morgan discovered a particular mutant of Drosophila with notches at the

end of their wings. This trait was attributed to a partial loss of function of what have

subsequently been described as the Notch gene that transcribes for a type I trans-membrane

receptor. The human Notch homologue gene is located in the 9q34.3 chromosome and was

cloned in the 1985/6 by two groups (Kidd et al., 1986; Wharton et al., 1985). In Drosophila there

are only one single Notch protein and two ligands (Delta and Serrate), by contrast mammals

display four Notch proteins Notch 1-4 and five ligands (Delta-like-1,-3 and -4, and Jagged-1,-2)

(Radtke and Raj, 2003) (Figure 6). Notch receptors are produced as monomeric precursors that

are subsequently processed by Furin protease in the trans-Golgi (called S1 cleavage) and

exposed on the cell surface as heterodimers (Blaumueller et al., 1997). The heterodimer consists

of a non-covalent association between the extra-cellular domain (NECD) and the intracellular

portion that goes across the membrane and protrudes in the extracellular space (NTCIM). The

extra-cellular domain contains both 36 tandemly repeated copies of an epidermal growth factor-

like motif, called EGF-like, required for the interaction with the ligands, a negative regulatory

domain (NRR) composed by cysteins that prevents the inappropriate activation of the receptor in

absence of its ligands and a heterodimerization domain (HD) that interacts with the extracellular

part of NTMIC. The Notch intra-cellular domain, called NICD is released from the membrane

and constitutes is the activated part of the cleaved heterodimer that mediates the majority of the

Introduction

biological processes in which the Notch pathway is involved. NICD contains a RAM domain,

necessary for interaction with transcriptional factors, seven ankyrin repeats, two nuclear

localization signals, a trans-activation domain (TAD) and a PEST domain, enriched in proline,

glutamate, serine and threnonine, that regulates the stability of the protein (Andersson et al.,

2011; Kovall and Blacklow, 2010) .

Figure 6. Structure of Notch proteins and their ligands a) Wing blade of a wild-type Drosophila melanogaster (left), and of a mutant with a partial loss of the Notch gene (right). The notches, which are absent in the wild type, but clearly visible at the border of the wing blade, have given the name to the implicated gene. b) Structure of Notch proteins and their ligands. Drosophila has one Notch receptor (dNotch) and vertebrates have four (Notch1–4), which are presented on the cell surface as heterodimers. The ectodomain of Notch receptors contains epidermal-growth-factor (EGF)-like repeats and a cysteine-rich Notch/Lin12 domain (LN); this is followed by a transmembrane domain, the RAM domain and six ankyrin repeats (ANK, also known as CDC10 repeats), two nuclear-localization signals (NLSs), followed by the transactivation domain (TAD) and a PEST sequence. c) Two transmembrane-bound ligands for Notch have been identified in Drosophila, named Delta (Dl) and Serrate (Ser). The vertebrates possess three Delta homologues, called Delta-like (DLL)-1, -3 and -4, and two Serrate homologues, Jagged 1 (JAG1) and Jagged 2 (JAG2). The ligands harbour an amino-terminal structure called DSL (Delta, Serrate and LAG-2), which is common to all family members, followed by EGF-like repeats. Serrate, Jagged1 and Jagged2 harbour a cysteine-rich domain (CR) following the EGF-like repeats (Radtke and Raj, 2003).

Introduction

3.3.2. Notch pathway activation

Notch signalling is initiated by a receptor-ligand interaction between two neighbouring cells.

Notably it has been recently demonstrated that also soluble forms of ligands forms exist, that can

activate Notch signalling in non neighbouring cells, suggesting other possible levels of cellular

communications (Lu et al., 2013a). The interaction between the ligand and the receptor is an

essential step for the Notch signalling activation, since it induces a conformational change in the

receptor that allows the cleavage and the subsequent release of a citoplasmic activated form that

translocates in the nucleus and activates its target genes. Following interaction with the ligand,

the receptor is cleaved by the disintegrin and metalloprotease ADAM10 or ADAM17 that

constitutes the S2 cleavage (Bray, 2006; Brou et al., 2000; Fortini, 2002). The subsequent

cleavage within the trans-membrane domain of Notch is mediated by the multisubunit protease

called gamma secretase that contains presenilin, nicastrin, PEN2 and APH1. This cleavage (S3

cleavage) results in the release of the Notch intra-cellular domain called NICD (Figure 3).

In the canonical pathway NICD translocates in the nucleus and cooperates with the DNA-

binding protein CSL (CBF1, Su(H) and LAG-1) and its co-activator Mastermind to promote

transcription of its target genes such as HES-1, HEY-L and cyclin D1 (Radtke and Raj, 2003). In

the absence of NICD, CSL already bound to the promoters of Notch target genes, recruits co-

repressors and histone deacetylases with a consequent inhibition of transcription (Kao et al.,

1998). When NICD is activated and present in the nucleus it competes with the repressors for

CSL binding, displaces them and converts CSL from transcriptional repressor to an activator

(Bray, 2006; Fryer et al., 2002).

While Notch mediates a number of biological processes through the canonical pathway, ligand-

or transcription-independent functions are emerging that may have and unprecedented impact in

development and disease (Andersen et al., 2012).

3.3.3 The Notch pathway regulation

Although the intracellular transduction of the Notch signal is remarkably simple, with no

secondary messengers, there is an enormous complexity in its regulation to keep in check the

amount and timing of its activation. Such regulation is essential since Notch controls both the

transcription and the function of genes and proteins involved in many cellular processes. Indeed

alteration of these checkpoints is associated with developmental defects and diseases (Grabher et

al., 2006; Ranganathan et al., 2011).

Introduction

Notch signalling is modulated in many ways during maturation and upon activation such as

through post-translational modifications of both receptors and ligands, endocytosis,

ubiquitination and receptor maturation (Le Borgne et al., 2005) (Figure 7). Numb, a cell fate

determinant is a known inhibitor of the Notch pathway. Numb is an endocytic adaptor protein

that acts as a suppressor of Notch signalling (Knoblich et al., 1995; Uemura et al., 1989).

Mechanistically, Numb has been shown to recruit the E3 ubiquitin ligase Itch, the mammalian

homolog of Drosophila Suppressor of deltex (Su(Dx)), to promote the degradation of the Notch

receptor and to regulate post-endocytic sorting events for Notch (McGill et al., 2009; McGill and

McGlade, 2003). Numb differentially affects various Notch receptors, since it negatively

regulates Notch1 and Notch2 receptors, but not Notch3, during myogenic differentiation (Beres

et al., 2011). In human, six alternatively spliced NUMB isoforms have been characterized to

date. The two most recently identified isoforms, NUMB5 and NUMB6, are less potent

antagonists of Notch signalling (Karaczyn et al., 2010), although it remains to be established if

the difference in biological effects among the different isoforms is strictly due to different effects

on Notch, as Numb also interacts with other signalling proteins, such as p53, Gli1, and

Hedgehog pathway effectors (Colaluca et al., 2008; Di Marcotullio et al., 2006). There is

emerging evidence that the relationship between Numb and Notch is not unidirectional. Indeed it

has been shown that Notch also can influence Numb since high levels of Notch reduce Numb

protein levels (Chapman et al., 2006).

Another important inhibitor of the Notch pathway is the E3-ubiquitin ligase and tumour

suppressor Fbwx7α (Gupta-Rossi et al., 2001; Oberg et al., 2001). As mentioned before,

activation of the Notch signalling promotes the cleavage and the translocation into the nucleus of

the NICD that recruits Mastermind for a proper activation of its target genes. Notably, the trans-

activation complex recruited by NICD on one hand promotes Notch target genes’ transcription,

on the other it promotes its own turn-over. Indeed Mastermind recruits the CDK8/CycC that

promotes NICD phosphorylation thus triggering recognition by Fbwx7α and the subsequent

proteasomal degradation (Figure 7) (Fryer et al., 2002; Gupta-Rossi et al., 2001; Oberg et al.,

2001).

Introduction

Figure 7. Notch pathway activation and regulation a-b) Notch is produced as a monomer and it is glycosylated by O-fucosyltransferase1 (OFUT1) at the endoplasmic reticulum (ER) thus into the Golgi for the subsequent cleavage mediated by Furin protease. c) This results in a heterodimeric receptor with non-covalently associated domains that is transported to the plasma membrane. d) Ligand binding initiates two successive proteolytic cleavages (S2 and S3). The first, mediated by an ADAM proteinase (a disintegrin and metalloproteinase domain), occurs in the extracellular domain. S2 cleavage allows access of the gamma-secretase complex, which is responsible for the second proteolytic cleavage (S3), which occurs within the transmembrane domain and liberates the intracellular domain (ICN) e) The ICN translocates to the nucleus, where it interacts with the CSL transcription factor. Numb suppresses Notch signalling, possibly by preventing nuclear localization and targeting the ICN for degradation through the E3 ligase Itch. g) The stability of nuclear ICN is regulated through its PEST domain (enriched in proline, glutamate, serine and threonine). Binding of MAML to p300 and cyclin-dependent kinase 8 (CDK8) promotes hyper-phosphorylation of the ICN PEST domain to facilitate ubiquitylation, through recruitment of Fbwx7αα E3 ligases, which target the ICN to the proteosome (Grabher et al., 2006). Fbwx7α is an important E3 ubiquitin-ligase that controls, besides N1-ICD, the turn-over of many

other important transcription factors, by recognition of the phosphorylated CPD motif. The

Notch1 CPD is located in the PEST domain, it is constituted by FLTPSPESP residues and it is

centred on threonine 2512. Indeed substitution of this threonine with an alanine disrupts the

Introduction

interaction of N1-ICD with Fbwx7α (Thompson et al., 2007). Since NICD is the final mediator

of activated Notch signalling, the regulation of its amount is of utmost importance as emerged by

the evidence that mouse models knock-out for Notch or Fbwx7α are similar since they die in

utero for the same cardiovascular disease (Swiatek et al., 1994; Tetzlaff et al., 2004; Tsunematsu

et al., 2004). In addition, in about 50% of human T-cell acute leukaemias and lymphomas (T-

ALL) (Weng et al., 2004), some of which disrupt binding to Fbwx7α. Notably, in some cases, T-

ALLs harbouring NOTCH1 activating mutations and treated with the gamma-secretase inhibitor

(GSI), were resistant to the pharmacological treatment because of mutation on Fbwx7α that

impaired its ability to bind N1-ICD (O'Neil et al., 2007; Thompson et al., 2007).

3.3.4. Biological functions of the Notch pathway

Notch biological functions are fundamental during both embryogenesis and regulation of adult

tissues in many organisms, through transcription of its target genes and cooperation with other

signalling pathways (Ranganathan et al., 2011).

While the genetic ablation of Notch1 or Notch2 induces premature embryonic death, the absence

of Notch3 or Notch4 does not induce this phenotype suggesting a non-redundant role for

Notch1/2 (Conlon et al., 1995; Krebs et al., 2003; Swiatek et al., 1994). During embryogenesis

the cooperation with Wnt, SHH and TGFβ signals is required for promotion of an epithelial-to-

mesenchymal transition (EMT) resulting in upregulation of the Snail/SLUG transcriptional

factor and down-regulation of E-cadherin (Grego-Bessa et al., 2004; Timmerman et al., 2004).

Notch is one of the major cell-fate determinants in particular during neurogenesis. Indeed cells

that over-express the Notch ligands will be committed to neuronal differentiation, while cells

that express the receptors will acquire epidermal traits (Artavanis-Tsakonas et al., 1999).

Another compartment in which Notch drives differentiation is the hematopoietic system, in

particular in lymphocytes’ precursors, in which it promotes activation of PDK1 and mTOR with

a consequent differentiation and proliferation as T cells (Kelly et al., 2007). By contrast

lymphocytic precursors in which the Notch pathway is turned off, will acquire B cell features

(Osborne and Minter, 2007). In addition Notch has a role in the cardiovascular system,

regulating vasculo- and artherio-genesis by cooperation with VEGF and HIF1α, respectively

(Holderfield and Hughes, 2008; Pear and Simon, 2005).

The ability of Notch signalling to drive cellular fate persists also during adulthood

(Androutsellis-Theotokis et al., 2006; Molofsky et al., 2004). Indeed the stem cell compartment

of many adult tissues, such as the hematopoietic, breast, muscular and intestinal one requires

Introduction

Notch activation for a proper regulation and of cell-specification. This process is a consequence

of the interplay between Notch with other signalling pathways and an inappropriate deregulation

of this well controlled mechanism could cause an unbalance in the stem cell compartments, as

occurs in cancer (Ranganathan et al., 2011).

The Notch regulators can have both positive and negative effects on the Notch pathway

activation. In this context well known interactors of Notch are the p53 family members (Dotto,

2009). Notch and p53 counteract each other at multiple levels regulating many cellular

processes. In addition, another member of the p53 family, p63 has been demonstrated to curb the

Notch signalling both in mammary stem cells and in epithelial cells where it blocks the Notch-

induced EMT (Yalcin-Ozuysal et al., 2010). On the contrary, enforced Notch activation impairs

p63 cellular functions with a consequent transformation process (Mazzone et al., 2010).

The result is an involvement of the Notch signalling in the regulation of stemness, proliferation,

apoptosis, differentiation, EMT, angiogenesis, tissue homeostasis, which, when alterated, are

associated with tumourigenesis pointing out Notch as an oncogene (Miele et al., 2006;

Ranganathan et al., 2011).

3.3.5. Notch and cancer

The role of Notch in cancer is strongly dependent on the cellular context. Indeed, in the skin,

Notch is an important tumour suppressor, mainly by blocking p63 functions (Dotto, 2008).

Notably in many other tissues, Notch emerges as a master oncogene. The first evidence of a role

of Notch in cancer, has been done in 1987, in human T-cell acute lymphoblastic leukemia (T-

ALL) in which there was a translocation involving the Notch locus on chromosome 7 and the T-

cell-receptor-beta on chromosome 9 (Reynolds et al., 1987). The consequence of this

translocation is the aberrant transcription of a truncated Notch1 protein that corresponds to the

nuclear active form, N1-ICD. Very frequent point mutations have subsequently been found

mainly in T-ALL, in particular on the heterodimerization and PEST domains, causing ligand-

independent activation and protein stabilization, respectively (O'Neil et al., 2007). In T-ALL,

Notch1 regulates two important proteins involved in tumourigenesis such as c-Myc and PTEN,

in an opposite fashion: on one hand Notch1 directly activates c-Myc transcription and on the

other PTEN gene expression is dampened by the transcriptional repressor Hes1, one of the most

important mediators of the Notch pathway (Palomero et al., 2008; Palomero et al., 2006; Weng

et al., 2006). In this way, Notch controls multiple cellular processes such as differentiation,

transformation, metabolism, apoptosis, boosting tumourigenesis. After the discovery of its

Introduction

involvement in T-ALL, Notch signalling has been implicated in various solid tumours, including

breast cancer, medulloblastoma, colorectal cancer, pancreatic tumour and non–small cell lung

carcinoma (NSCLC) (Miele et al., 2006).

With exception of NSCLC, in all these solid tumours deregulated Notch activity results mainly

from over-activation in absence of mutations. In particular it has been demonstrated that up-

regulation of the Notch pathway in 30% of NSCLS is caused by loss of Numb expression, while

in 10% of cases is caused by gain-of-function mutations of the NOTCH-1 gene (Westhoff et al.,

2009) suggesting an important role of the Notch pathway in this tumour type.

The oncogenic potential of Notch activation in solid tumours was first observed in mouse

mammary tumor virus (MMTV)–driven breast cancer. The integration of MMTV in specific loci

of the host genome results in dysregulated expression of adjacent genes and subsequent

outgrowth of tumorigenic clones (Dievart et al., 1999). Characterisation of one of these loci

revealed expression of a truncated constitutively active form of Notch4 (Gallahan and Callahan,

1987; Gallahan et al., 1996). In mouse models, Notch activation can clearly drive mammary

tumours, and in human breast cancers increased expression of Notch and Jagged1 correlates with

poor prognosis (Reedijk et al., 2005). Loss of Numb expression causes over-activation of Notch

signalling and tumourigenesis also in a large part of breast cancers (Pece et al., 2004). In breast

tumours Notch has been demonstrated to functionally cooperate with several proteins for which

a role in breast tumourigenesis has been well enstablished such as ErbB2, Ras signalling, Pin1,

cyclin D1 and c-Myc (Miele et al., 2006; Ranganathan et al., 2011).

The impact of Notch in breast tumours patients, derives from the evidence that Notch plays a

major role in driving malignant phenotypes including EMT features, CSCs, metastases and

chemoresistance, as assessed also by analyses of overall survival (Rizzo et al., 2008).

In normal breast stem cells the role of Notch1 is well established since it was shown to trigger

luminal progenitors’ formation by several groups. Consistent with this function in normal

mammary stem cells and progenitors, the Notch pathway controls also breast CSCs. It seems that

while Notch1 is mainly expressed and required for the early progenitors, Notch4 in more

expressed in the CSC pool (Bouras et al., 2008; Dontu et al., 2004; Harrison et al., 2010).

Accordingly, genetic and pharmacological abrogation of the Notch pathway in breast cancer

cells is effective in curbing the CSCs pool.

The ability of activated Notch receptors to promote also chemoresistance is not simply a

consequence of an increased CSC/progenitor population. Indeed Notch has well-established

roles in different mechanisms of chemoresistance as a consequence of the promotion of cell

Introduction

survival factors such as BIRC5 (SURVIVIN), Bcl-2, and drug efflux pumps such as ABCG2

(Bhattacharya et al., 2007; Ranganathan et al., 2011). Thus abrogation of Notch signalling

sensitizes both cancer cells and CSCs to chemotherapeutic agents (Espinoza et al., 2013).

Notch has been clearly linked to EMT features and metastases in breast cancers with in vitro and

in vivo experiments. It has been demonstrated that Notch induces Slug gene transcription, with a

consequent E-cadherin decrement, that results in an increased migratory and invasive ability

(Leong et al., 2007). Recently it has also been demonstrated that activation of the Notch pathway

affects the breast cancer secretome and promotes the metastases formation in the brain. Notably

pharmacological ablation of Notch activation prevents this effect (Xing et al., 2013).

It is evident that Notch is a candidate druggable pathway in cancer since it is enzymatically

controlled at many levels and it is linked to aggressiveness phenotypes such as chemoresistance,

tumourigenesis and CSCs maintenance and involvement of this pathway in breast CSCs’ biology

warrants further investigation.

Aim of the thesis

4. AIM OF THE THESIS

A major issue of breast cancer management is the failure of therapeutic regimens and disease

relapse. These features are widely suggested to be promoted by cancer stem cells’

subpopulations. It is conceivable that genetic or pharmacological ablation of molecular pathways

required for breast CSCs induction and maintenance, will exhaust this cell population with a

consequent gain in sensitivity to chemotherapy. Moreover, the abrogation of this molecular

signalling will prevent reprogramming of terminal differentiated cells to a more stem-like

phenotype, blocking the selection of a chemoresistant and aggressive cancer cell population.

The aim of my PhD work was to discover factors that might underlie all these aspects in breast

CSCs. A putative candidate for such a role is the prolyl-isomerase Pin1 since i) it is

indispensable for full activation of oncogenic signalling pathways ii) its protein levels are up-

regulated in high-grade breast cancers which exhibit an elevated CSCs content and iii) Pin1 is a

stemness factor in embryonic stem cells and some adult tissues.

The first part of this study will focus on the role of Pin1 in the maintenance of normal mouse

mammary stem cells ex vivo and in vivo. As normal stem cells and CSCs display considerable

similarities in their molecular regulators, the subsequent part will be aimed at dissecting how

Pin1 controls and orchestrates breast CSCs number, maintenance and aggressive traits in vivo

such as tumourigenesis and chemoresistance.

In the second part we will investigate which is the functional mediator of the Pin1 promoted

CSCs maintenance. In particular this study will focus the attention on how Pin1 potentiates the

Notch pathway by safeguarding Notch1 and Notch4 nuclear active forms from their E3-

ubiqutitin-ligase Fbxw7α. Thereafter we will investigate the impact of the Pin1-Notch1/4 axis in

tumourigenesis, CSC maintenance with in vitro and in vivo analyses in mouse models and in

human breast cancer samples.

These studies, collectively, might be relevant for providing the rational for a therapeutic strategy

based on Pin1 inhibition to hit CSCs, restore chemo-sensitiveness and inhibit metastatic spread.

Results

5. RESULTS

5.1 The prolyl-isomerase Pin1 is required for the self-renewal of normal mammary stem cells Pin1 knock-out mice display a variety of developmental defects. In particular the mammary

epithelium of these mice fails to undergo the dynamic changes required for its expansion during

pregnancy (Liou et al., 2002). Since it has been demonstrated that some of these defects are a

consequence of an impaired stem cell function (Atchison and Means, 2003), we hypothesized a

possible function of Pin1 in the regulation the mammary stem cell compartment. For this aim we

decided to evaluate the mammary stem cell number of wild-type (Pin1+/+) and Pin1 knock-out

(Pin1-/-) mice. Mammary tissues from 8-10 weeks old virgin female mice were dissociated,

prepared as single cell suspensions of purified, lineage-depleted epithelial cells (Stingl et al.,

2006) and grown in suspension cultures to form secondary mammospheres (M2) (Dontu et al.,

2003). We detected a significant reduction in the spheres’ formation in absence of Pin1, since

cells obtained from Pin1+/+ mice formed an average of 22.9 (±1.44) M2 mammospheres per

100.000 seeded cells, while those from Pin1-/- cells gave rise to only 13.53 (±2.49), indicating a

40% reduction of M2 formation (Figure 8A).

Figure 8. Ablation of Pin1 reduced the mouse mammary epithelial stem cells number A) Left panel: Number of secondary mammospheres (M2) generated from primary mammary epithelial cells of indicated mice. Right panel: representative M2 microscope images with 200 µm scale bar. B) Serial replating of mammospheres (M1-M5) generated from Pin+/+ mice treated with DMSO or PiB (1,5 µM). Means, standard deviations and p-values (t-test, n=4) are indicated in the histogram It is known that the microenvironment is important to control the stem cell compartment’s

homeostasis (Valent et al., 2012). To determine if the reduced stem cell number in Pin1-/- mice

was due to an effect of Pin1 specifically in the mammary stem cells or in the microenvironment,

we performed with Pin1+/+ mammary epithelial cells an ex vivo mammsophere experiment by

blocking in vitro the enzymatic activity of Pin1. We serially replated cells from M2 to tertiary

Results

and quaternary mammospheres (M3, M4) in presence or absence of the Pin1 inhibitor PiB

(Uchida et al., 2003) (Figure 8B).

As expected in these conditions, we observed a progressive decrease in mammosphere formation

efficiency (MFE) at each passage, due to exhaustion of adult stem cells (Cicalese et al., 2009).

Notably, this effect was exacerbated by almost 50% at the M4 stadium upon addition of the Pin1

small molecule inhibitor PiB. These data suggest that inhibition of Pin1 can directly inhibit the

stem cell self-renewal of mammary stem cells.

This evidence indicates a role for Pin1 in determining self-renewal and replicative potential of

mammary stem cells thus implying alterations of the mammary stem cell compartment in Pin1-/-

mice. To better address this point we analyzed the mammary epithelial stem cells hierarchy in

Pin1+/+ and Pin1-/- mice by detection of specific immune phenotype markers, CD24 and CD49f,

that identify both stem/bipotent progenitors (CD24med/CD49fhigh or mammary repopulating

units, MRU) and luminal progenitors (CD24high/CD49flow or mammary colony forming cells,

Ma-CFCs) (Stingl et al., 2006). In line with our hypothesis, the MRU and Ma-CFC cell

populations from Pin1-/- mammary glands were present at lower proportion as compared to

Pin1+/+ mice (Figure 9A).

Figure 9. Bipotent stem cell and luminal progenitor number is decreased in Pin1-/- mammary tissue Bipotent stem cell and luminal progenitor number is decreased in Pin1-/- mammary tissue. Left panel: representative FACS analyses of mammary epithelial cells from indicated mice. CD24/CD49f plots and gatings for MRU and Ma-CFC populations are indicated. Right panel: histogram of mean counts of MRU and MA-CFC populations from Pin-

/- normalized to Pin1+/+ mice. Means, standard deviations and p-values (t-test, n=3) are indicated.

To demonstrate that the analyzed populations were indeed MRUs and Ma-CFCs, we sorted these

two populations by FACS and analyzed the expression of specific myoepithelial (SMA, CK14)

or luminal (CK-18, CK-19) markers, that are mainly expressed in the stem cell and progenitor

populations, respectively (Stingl et al., 2006). As shown in Figure 10, Ma-CFC and MRU cells

Results

sorted from the mammary epithelial cell population expressed higher levels of CK-18, CK-19 or

SMA and CK-14, respectively, when compared to unsorted cell population (TOTAL).

Figure 10. Molecular characterisation of the sorted MRU and Ma-CFC mammary epithelial cells. Control qRT-PCR for myoepithelial (SMA, CK14) and luminal (CK18, CK19) markers of mRNA extracted from MRU and Ma-CFC (Stingl et al., 2006) sorted populations relative to the total population, shows accuracy of the FACS sorting procedure.

Many stem cells factors, such as Slug and Pin1 itself, are more expressed and active in the stem

cell compartments (Guo et al., 2012) (Moretto-Zita et al., 2010; Nishi et al., 2011). For this

reason we compared Pin1 mRNA and protein levels in the stem cells’ compartment with their

differentiated counterpart. Interestingly, we found almost three times higher Pin1 levels in the

MRU cell population as compared to the total of mammary epithelial cells (Figure 11).

This evidence confirmed our hypothesis and suggests a prominent role of Pin1 in sustaining the

mammary stem cell compartment in vivo.

Figure 11. Pin1 mRNA and protein levels are upregulated in the mammary stem cell compartment. Left panel: qRT-PCR of endogenous Pin1 mRNA in MRU sorted populations relative to total population. Means, standard deviations and p-values (t-test, n=3) are indicated. Right panel: Western blot analysis of the same cell populations as in the left panel. Fold change in Pin1 protein levels determined by Image J software with respect to actin levels is indicated by a number, Molecular weights in KDa (Mr (K)) are shown on the right.

Results

5.2 Pin1 inhibition impairs mouse and human cancer stem cells traits

Molecular pathways required for normal mammary stem cells’ maintenance are often

indispensable also for CSCs (Valent et al., 2012; Visvader and Lindeman, 2012). Thus many of

these pathways are up-regulated also in the CSCs. Since we demonstrated that Pin1 is

indispensable for the maintenance of normal stem cells, we investigated whether Pin1 could have

a role also in the regulation of breast CSCs.

For this aim we performed mammosphere formation with NOP6 mouse mammary cancer cells,

harbouring a Her2/Neu amplification, for serial passages, in presence or absence of the Pin1

inhibitor PiB. Interestingly Pin1 inhibition strongly impaired the formation of mammospheres at

each passage, indicating that Pin1 is required for the self-renewal of mouse mammary cancer

stem cells (Figure 12).

Figure 12. Pin1 inhibition curbs mouse mammary CSCs maintenance. Serial repleating of mammospheres from M1 to M4 generated from NOP6 cells trated with DMSO or Pib (1,5µM). Mammosphere formation efficiency (%MFE) was calculated as percentage of mammospheres divided by the number of plated cells.

To verify if Pin1 is required also by human breast cancer stem cells, we infected the triple

negative breast cancer cell line MDA-MB-231 with a doxycycline-inducible knockdown

lentiviral construct for Pin1 (pLKO-TetO-shPin1) and tested their ability to generate

mammospheres. As shown in Figure 9A, while MDA-MB-231 MFE remained constant

throughout serial replating to M4, the formation of mammospheres from Pin1 silenced cells

(+DOX) progressively decreased at each passage. Accordingly the content of putative stem cells

was lower following Pin1 silencing or inhibition, as confirmed by the Aldefluor assay, that

valuates the activity of Aldehyde dehydrogenase 1 (Aldh), a marker for breast CSCs (Ginestier

et al., 2007) (Figure 13A). Accordingly, Pin1 pharmacological inhibition curbed not only MDA-

MB-231 mammospheres formation but also the MFE of other breast cancer cell lines such as and

BT-549 and SUM-159 (Figure 13B). These observations could be extended also to two patient-

Results

derived aggressive breast cancers, whose CSCs were reduced by Pin1 inhibition (Figure 13C).

All these data demonstrate that genetic or pharmacological ablation of Pin1 impairs breast CSC

self-renewal and replicative potential in a broad spectrum of breast cancer cells.

Figure 13. Pin1 inhibition curbs human mammary CSCs traits. A) Left panel: MFE of MDA-MB-231-pLKO-shPin1 control cells (Ctrl) compared to shPin1 induced cells (DOX) upon serial passages. Right panel: Quantification of Aldh-positive and Aldh-negative cells from control- and shPin1 induced M4, as assessed by FACS. B) Left: Serial replating of mammospheres (M1-M4) generated from MDA-MB-231 human breast cancer cells and treated with DMSO or PiB (1,5 µM and 3 µM). Mammosphere formation efficiency (%MFE) was calculated as percentage of mammospheres divided by the number of plated cells. Means, standard deviations and p-values are indicated (t-test, n=3, M4). Right: Percentage of M2FE of BT-549 and SUM-159 breast cancer cell lines treated with vehicle (DMSO) or PiB. Means, standard deviations and p-values are indicated (t-test, n=3). C) Secondary and tertiary mammosphere formation efficiency of a grade 2 (G2) and grade 3 (G3) patient-derived breast cancer cells treated with vehicle (DMSO) or 3 µM PiB.

Results

Conversely, overexpression of Pin1 in MDA-MB-231 cells increased M2 formation by almost

two fold and produced an increase of Aldh-positive cells, when compared to empty vector

harbouring cells (Figure 14).

Figure 14. Pin1 over-expression boosts human mammary CSCs traits. Left: Secondary mammosphere formation efficiency of MDA-MB-231 cells transduced with empty or HA-Pin1 expressing vectors. Representative microphotographs and scale bars are shown. Right: Histogram showing percentage of Aldefluor median fluorescence intensity difference (MFI).

To unveil the molecular mechanism underlying the role of Pin1 in promotion of CSCs, we

evaluated the expression of several genes acting within pathways that govern the stemness

phenotypes of breast CSCs (Leong et al, 2007; Yu et al, 2007, 2011; Polyak & Weinberg, 2009;

Cordenonsi et al, 2011; Visvader & Lindeman, 2012) in Pin1-silenced quaternary

mammospheres. Interestingly, as shown in Figure 11, no single pathway emerged alone, as the

expression of all tested factors (Hes1, HeyL, Birc5, CTGF, Slug, ABCG2, Ptch, Bmi-1, HMGA2

and Klf4) decreased by Pin1 knockdown thus suggesting a common down-regulation of

stemness signalling pathways following Pin1 depletion.

Epithelial-mesenchymal plasticity in breast carcinoma has recently been linked to acquisition of

stem cell traits by tumor cells (Mani et al, 2008). We therefore also analysed the impact of Pin1

modulation on this process by analyzing markers of epithelial-to-mesenchymal transition (EMT).

Of note, Pin1 downmodulation caused enhanced mRNA expression of the epithelial marker E-

cadherin (CDH1) while that of mesenchymal markers Vimentin and Fibronectin (VIM1, FN)

was reduced (Fig 2C, left panel), in parallel with a strong recovery of E-cadherin and decay of

Slug and Vimentin at the protein level (Figure 15).

Results

Figure 15. Pin1 depletion reduces stem cell and mesenchymal marchers Left panel: qRT-PCR of the indicated stemness and EMT marker genes from MDA-MB-231-pLKO-shPin1 quaternary mammospheres (M4) upon shPin1 induction (DOX) with respect to control cells (Ctrl). Standard deviations are indicated, P-values * <0.02 (t-test, n = 3). Right panel: Western Blot analysis of EMT markers of the same cells. Molecular weights Mr(K) are indicated in kDa. Representative microphotographs of M4 are shown, 200 lm scale bar is indicated.

All together these results indicate that Pin1 is indispensable for breast CSCs maintenance and its

depletion strongly impairs CSCs number and traits, through the modulation of a specific

stemness and mesenchymal gene expression program.

5.3 Pin1 promotes breast CSCs maintenance mostly through up-regulation of the Notch

signalling

As shown above, breast CSCs depleted of Pin1 display down-modulation of many signalling

pathways. The majority of these genes are directly -or indirectly- controlled by the Notch

signalling pathway (Lee et al., 2008; Li et al., 2012; Ranganathan et al., 2011). The role of the

Notch signalling in breast CSCs and EMT induction is well established, hence we tested if Notch

is the key mediator of Pin1-mediated stemness. In particular we analyzed the protein levels of

both Notch1 and Notch4 since their role in breast CSC is well described (Dontu et al., 2004;

Harrison et al., 2010; Ranganathan et al., 2011). Intriguingly, the levels of their active forms,

N1-ICD and N4-ICD, were strongly deregulated in Pin1 depleted quaternary (M4)

mammospheres (Figure 16A). In a previous work we established that that Pin1 potentiates the

cleavage of the Notch1 receptor, boosting its nuclear levels (Rustighi et al., 2009). However, the

very strong down-regulation of the Notch nuclear active forms observed in this case could not be

totally ascribed to a reduced cleavage of the receptor. Since N-ICD levels are finely tuned by

Results

proteasomal degradation, we speculated another role for Pin1 in boosting Notch signalling,

probably through protein stabilization.

Figure 16. N1- and N4-ICD protein levels are decreased in Pin1-silenced mammospheres Left panel: Western Blot analysis of N1- and N4-lCD protein from MDA-MB-231-pLKO-shPin1 M4 control cells (Ctrl) and shPin1 induced cells (DOX). Molecular weights (Mr) are indicated in kDa. Right panel: histogram representing the percentage of band intensity with respect to actin levels.

To understand if the effect of Pin1 on breast CSC required high N1-ICD levels, we tested the

ability of N1-ICD or of a constitutively stable N1-ICD mutant (dPEST) to boost M2 formation

following Pin1 knockdown. This mutant lacks the cdc4-phosphodegron (CPD) constituting the

consensus for the E3 ubiquitin-ligase Fbwx7α. As expected, Pin1 silencing (+DOX) reduced the

M2FE of MDA-MB-231 pLKO (Figure 16B). Moreover ectopic expression of N1-ICD in

control cells did not affect the formation of mammospheres, since in these cells the endogenous

Notch pathway is already strongly activated (Harrison et al, 2010). Interestingly, N1-ICD

overexpression was not able to rescue M2FE in Pin1 silenced cells. This evidence suggests that

in this context the role of Pin1 in boosting the Notch cleavage seems to be dispensable, since the

ectopically expressed N1-ICD is already cleaved and activated. Instead, overexpression of N1-

ICD-dPEST was able to rescue M2FE. When we analyzed the protein levels of Notch nuclear

active forms we found that protein levels of over-expressed N1-ICD, but not those of N1-ICD-

Results

dPEST, were strongly decreased in Pin1 depleted mammospheres.

These data suggest that Pin1 critically supports breast CSC self-renewal by sustaining high

intracellular levels of N1-ICD. To better address this point, we analysed the endogenous levels

of both Pin1 and N1-ICD in breast CSCs. Indeed we sorted the Aldh-positive cells from MDA-

MB-231 mammospheres or collected patient derived mammospheres and analysed their protein

content. As shown in Figure 13, Pin1 and N1-ICD protein levels were always higher in the stem

cell compartment compared to their differentiated counterpart. Moreover we found that also in

vivo, in breast cancer tissues, the expression of Pin1 co-localized with Aldh-positive cells

(Figure 17, right panel).

Figure 17. Pin1 and N1-ICD are up-regulated in the CSC compartment. Left panel: Comparative Western blot analyses of Aldh-positive (stem) versus Aldh-negative cells (nonstem) sorted from MDA-MB-231 M2 (left panel) and patient-derived breast cancer secondary M2 mammospheres (stem) versus cells cultured in adherence (2D) (right panel). Relative fold changes in Pin1 or N1-ICD protein levels were determined by Image J software with respect to actin levels is indicated by a number. Right panel: Representative bright field microphotographs at 100X magnification of immunohistochemical analyses of serial sections of a breast cancer with the indicated antibodies are shown. Part of the figure indicated with a square is shown by an inset at 400X magnification. Scale bar is indicated.

Over-expression of Pin1 is detected in breast cancers and it is known that ectopic over-

expression of Pin1 is able to disrupt cellular polarity of breast epithelial cells (Ryo et al., 2002),

but it was not investigated yet if Pin1 is able to promote both an EMT program and formation of

secondary mammospheres in normal cells and if the Notch signalling is required for this aspect.

To answer to these questions, we ectopically over-expressed Pin1 in epithelial non transformed

MCF10A cells and observed that Pin1 promoted an enrichment of the stem cell population, as

assessed by the specific CD44high/CD24low immunophenotype (Al-Hajj et al, 2003). Moreover

Pin1 overexpression triggered a huge formation of mammospheres and a concomitant EMT as

Results

demonstrated by down- and up-regulation of the two markers E-cadherin and Slug, respectively

(Figure 18). Crucially, the levels of cleaved Notch1 and Notch4, normally undetectable in these

cells, were up-regulated by Pin1 over-expression and were essential for mammosphere

formation, since block of Notch cleavage by a gamma-secretase inhibitor (GSI) was sufficient to

blunt this process and to reduce the EMT markers as depicted in Figure 18A.

Figure 18. Pin1 over-expression promotes an EMT-stem cells phenotype mostly though the Notch pathway. A) Upper panel: %M2FE of MCF10A breast epithelial cells transduced with empty (pLPC) or HA-Pin1 overexpressing vectors (pLPC-HA-Pin1) and treated with DMSO () or 10 lM GSI. Means, standard deviations and P-values (t-test, n = 3) are indicated. Middle panel: Representative microphotographs of M2 are shown, scale bar of 200 lm is indicated. Lower panel: Western Blot of cell lysates from corresponding MCF10A clones. White and black arrows indicate over-expressed and endogenous Pin1, respectively. Comparative Western blot analyses of Aldh-positive (stem) versus Aldh-negative cells (nonstem) sorted from MDA-MB-231 M2 (left panel) and patient-derived breast cancer secondary M2 mammospheres (stem) versus cells cultured in adherence (2D) (right panel). Relative fold changes in Pin1 or N1-ICD protein levels were determined by Image J software with respect to actin levels is indicated by a number. B) CD44/CD24 FACS analyses of MCF10A cells transduced with pLPC or pLPC-HA-Pin1. Percentage of the different CD44/CD24 populations is indicated on the right.

All together our results demonstrate that Pin1 is a bona fide stem cell factor by promoting EMT,

maintaining a mesenchymal/stem cell fate and self-renewal of CSCs mainly through regulation

of the Notch pathway.

Results

5.4 Suppression of Pin1 sensitizes breast CSCs to chemotherapy in vitro and in vivo

There is increasing evidence that, while chemotherapy reduces the tumour mass by killing the

sensitive cancer cells, it selects a subpopulation of cancer cells that are intrinsically

chemoresistant or alternatively, it induces cells to acquire resistance to chemotherapy by genetic

or epigenetic alterations (Dean et al., 2005). In accordance, it has been demonstrated that

chemotherapy selects and strongly increases the breast CSCs population (Yu et al., 2007). We

also found that uninduced MDA-MB-231-pLKO-shPin1 cells treated with different

chemotherapeutic agents display a slight increase in MFE (Figure 19). Instead, when we induced

Pin1 silencing with doxycyclin (+DOX), this effect was completely lost, since Pin1 depletion

does not only prevent accumulation of CSC following chemotherapy but more importantly it

regains chemosensitivity.

Figure 19. Pin1 depletion sensitizes CSCs to chemotherapeutic agents Pin1 knockdown synergizes with chemotherapy treatment to block breast CSCs’ self-renewal. Percentage of M2FE of control MDA-MB-231-pLKO-shPin1 cells (Ctrl) compared to shPin1-induced cells (DOX) treated with indicated drugs or PBS. Means and standard deviations are indicated, P-values are * = 0.001, **< 0.0003 (t-test, n = 3).

To address the in vivo impact of these findings, we performed a xenograft experiment in

immunocompromised mice by injecting MDA-MB-231 infected with a control- or a Pin1

shRNA. We allowed the tumour to grow till tumour mass became visible, then mice where

randomized in different groups and treated with paclitaxel or PBS. When we analyzed the size of

the tumours, either the treatment with paclitaxel or the silencing of Pin1 similarly reduced the

tumour mass, while the combination had a strong synergistic effect (Figure 20A). Analysis of the

cancer stem cells’ content by evaluating the activity of the Aldh enzyme evidenced that, while

paclitaxel treatment alone reduced the tumour mass but heavily increased the CSC content, the

concomitant Pin1 depletion prevented this effect (Figure 20B). As a result, Pin1 ablation on one

hand is able to reduce the tumour mass and the CSCs content, on the other hand it re-sensitize

CSC to chemotherapy. At the protein level, Western Blot analysis unveiled mesemchymal-to-

Results

epithelial transition in Pin1 depleted tumours as emerged by the reduction in Slug and Vimentin

and the concomitant increase in E-cadherin (Figure 20C). Accordingly, N1-ICD levels were

down-regulated in Pin1 silenced tumours. Moreover the increased cleavage of caspase-3, an

apoptotic marker, was significantly increased in Pin1-silenced tumours treated with paclitaxel

underlining a synergistic effect between paclitaxel treatment and Pin1 ablation in inducing

apoptosis. It has been demonstrated by another group that Pin1 depletion curbed Mcl-1 protein

levels and promoted apoptosis in breast cancer cell lines treated with paclitaxel (Ding et al.,

2008). To verify the impact of Mcl-1 in our condition, we analyzed its protein levels in our

xenografts. As expected Mcl-1 levels decreased in Pin1 depleted tumours, but also in the

paclitaxel treated one, suggesting that Mcl-1 is not directly involved in the CSC accumulation

due to paclitaxel treatment. More importantly Mcl-1 levels did not further decrease in Pin1

silenced tumours treated with the drug, suggesting the involvement of another signalling

pathway that, once inactivated, could give rise to the apoptotic process (Figure 20C).

Figure 20. Pin1 depletion sensitizes CSCs to chemotherapeutic agents in vivo A) Pin1 knockdown synergizes with paclitaxel to block breast cancer growth. Tumor growth of MDA-MB-231 xenografts espressing the indicated shRNAs and treated with paclitaxel (grey bars) or left untreated (PBS) (black bars). Means and standard deviations are indicated, P-values are *** < 0.0003 (t-test, n = 12). B) Pin1 knockdown blocks chemotherapy-induced breast CSCs’ expansion in vivo. Histogram representing the Aldefluor mean fluorescent intensity (MFI) of cells from control- and shPin1 MDA-MB-231 xenografts treated with Paclitaxel or PBS. Means and standard deviations are indicated, P-values are * = 0.001 (t-test, n = 3 for each condition). C) Pin1 knockdown induces reversal of EMT and cell death in breast cancer xenografts in combination with paclitaxel. Western blot analyses of tumor xenografts from A).

Indeed, to verify if Notch was the involved signalling pathway, we performed mammosphere

formation assays with MDA-MB-231 expressing control- or Pin1-specific shRNA and treated

with paclitaxel. Notably, over-expression of N1-ICD-dPEST totally rescued mammospheres’

formation and chemoresistance (Figure 21).

Results

All together these data demonstrate that Pin1 is required for the accumulation of chemoresistant

breast CSCs during chemotherapy and that ablation of its levels or activity curbes the CSCs

population with concomitant gain of chemosensitivity.

Figure 21. N1-ICDs-dPEST over-expression is able to rescue the chemoresistance in Pin1-depleted breast cancer cells. Left: expression of stable N1-ICD-dPEST rescues resistance to paclitaxel treatment in Pin1 silenced cells. Percentage of M2FE of control (black bars) or Pin1 shRNA (grey bars) expressing MDA-MB-231 cells transduced with empty (-) or N1-ICD-dPEST (+) expressing vectors and treated with Paclitaxel (+) or PBS (-). Means and standard deviations are indicated, P-values are **0.0001, ***<0.00003 (t-test, n = 3). Right: the corresponding analysis of protein expression by Western blot is shown. Arrow indicates position of the specific band.

5.5 Pin1 protects N1- and N4-ICD from proteasomal degradation

All the data presented above indicate that N1-ICD protein levels are under direct control of the

polyl-isomerase, moreover the stable mutant N-ICD-dPEST is unaffected by modulation of Pin1

levels, suggesting that Pin1 is impinging on N1-ICD turnover through regulation of its protein

degradation. To test this possibility, we performed protein stability assays. As shown in Figure

22A, Pin1 depletion by siRNA in MDA-MB- 231 breast cancer cells treated with gamma-

secretase inhibitor to block Notch cleavage at the membrane (Rustighi et al, 2009), reduced the

half-life of N1-ICD. Addition of the proteasome inhibitor Lacatacystin rescued N1-ICD levels in

this context, unveiling that this reduced half-life was proteasome-dependent. Pin1 silencing

curbed the half-life of over-expressed N1-ICD protein also in the SK-BR-3 breast cancer cell

line that was strongly reduced from 2 h to 40 min upon Pin1 siRNA treatment with respect to

control silencing (Figure 22B). Moreover, re-introduction of a siRNA resistant Pin1 construct

(Pin1r) in Pin1 silenced cells prevented the reduction of N1-ICD protein levels, demonstrating

that N1-ICD protein levels are specifically dependent on the prolyl-isomerase Pin1.

Results

Figure 22. Pin1 depletion impaired N1-ICD protein half-life A) Pin1 knockdown accelerates the decay of endogenous N1-ICD. Western blot of endogenous N1-ICD following RNA interference (RNAi) with the indicated siRNAs and time points following GSI or GSI plus Lactacystin (+) chase is shown. B) Left panel: Western blot of over-expressed N1-ICD-myc along with indicated siRNAs and empty (-) or siPin1 resistant pCDNA3-HA-Pin1 (Pin1r) expression vectors, treated with DMSO (-) or CHX. Right panel: Graph indicating N1-ICD-myc protein levels versus time of CHX treatment in the three tested conditions (siCtrl, siPin1 and siPin1 + HA-Pin1r). Means, standard deviations and p-values (t-test, n=3) are indicated.

Pin1 is known to bind several substrates on phosphorylated Ser/Thr-Pro residues and to affect

their stability (Liou et al, 2011). In vitro and in vivo binding assay demonstrated that Pin1 binds

to N1-ICD and this interaction was direct. Moreover, domain-mapping analysis showed a

marked reduction in Pin1 binding to N1-ICD when the C-terminal PEST domain containing the

cdc4-phosphodegron was deleted (Figure 23A). Intriguingly, the residues T2512/P2513 within

the phosphodegron in the PEST domain are known to be indispensable for recognition by the

ubiquitin-ligase Fbwx7α when phoshorylated on T2512. To verify if this motif is also a Pin1

binding site, we performed in vitro GST-Pin1 binding assays with N1-ICD or a mutant construct

in which we replaced the threonine residue 2512 with alanine (T2512A) (Figure 23B). This

mutation impaired the interaction between Pin1 and N1-ICD. Accordingly, Pin1 overexpression

was unable to affect the protein stability of this mutant as compared to wild-type N1-ICD

(Figure 23C).

Results

Figure 23. N1-ICD-T2512A mutation hampers binding to Pin1 A) Mapping of the Pin1-N1-ICD interaction. Upper: Schematic of pCDNA3-N1-ICD-myc C-terminal deletion constructs (from d2444 to d1991) used for mapping Pin1 binding domains in N1-ICD. TM: transmembrane, RAM: CSL interacting, ANK: ankyrin, STR: Serine-Threonine rich, TAD: transactivation, PEST: Pest domain. Numbering refers to Swissprot entry P46531. Interaction with Pin1 is indicated next to the constructs: +++ very strong, ++ strong, – no binding. A relevant drop in binding strength is observed by deletion of the PEST domain and the STR. Lower: Representative anti-myc tag Western blot analyses of GST-Pin1 pull down assays of indicated proteins overexpressed in HEK 293T cells is shown. The borders around the panels demarcate juxtaposed parts of same or different gels aligned in function of the molecular weight.. B) Anti-myc tag Western blot analysis of GST-Pin1 or GST pull down assay and input levels of indicated proteins overexpressed in HEK 293T cells is shown. (F) N1-ICD-T2512A mutant half-life is unaffected by Pin1 levels. C) Western blot of CHX chase experiment with N1-ICD-myc and N1-ICD-T2512A-myc mutant with empty (-) or HA-Pin1 expressing vectors. Molecular weights Mr(K) are indicated in KDa

Results

Finally, these results suggest that the T2512/P2513 motif, indispensable for Fbwx7α recognition,

constitutes also the docking site for Pin1, which might exert N1-ICD stabilization by interfering

with Fbwx7α-binding.

5.6 Pin1 protects N1- and N4-ICDs from degradation by the ubiquitin-ligase Fbwx7α

To test if Pin1 was counteracting N1-ICD degradation by hampering the interaction with the E3-

ubiquitin-ligase Fbwx7α, we tested if Pin1 is able to affect N1-ICD ubiquitination. For this aim

we performed an ubiquitination assay in COS-7 cells by a Ni-NTA pull down and, as expected,

over-expression of Fbwx7α in N1-ICD transfected cells, caused an increase in N1-ICD poly-

ubiquitination (Figure 24A). More importantly, when we depleted Pin1 from these cells by

RNA-interference, we found an enhanced N1-ICD poly-ubiquitination. Thus, Pin1 interferes

with Fbwx7α-mediated ubiquitination of N1-ICD. Accordingly when N1-ICD and Fbwx7α were

co-ovexpressed in two different cellular contexts, the concomitant Pin1 over-expression

completely rescued N1-ICD protein levels (Figure 24B-C). Since the final read-out of Notch

pathway activation is the transcription of its target genes, we evaluated if Pin1 was able to boost

Notch dependent transcriptional activity despite presence of high-levels of its negative regulator

Fbwx7α. Interestingly, as emerged from luciferase assays, Pin1 fully rescued the ability of N1-

ICD to promote gene transcription despite presence of its ubiquitin ligase (Figure 24D). These

data clearly show that Pin1 is able to prevent Fbwx7α dependent degradation of N1-ICD,

boosting its protein levels and transcriptional activity.

Results

Figure 24. Pin1 protects N1-ICD from Fbwx7α dependet degradation A) Pin1 depletion enhances Fbwx7α-dependent poly-ubiquitination of N1-ICD. Western blot analysis of high molecular weight N1-ICD-myc products (N1-ICD-myc) from a Ni-NTA pull-down in COS-7 cells transfected with the indicated vectors along with control- or Pin1 siRNA. Input levels of over-expressed proteins are shown. B) Pin1-

/-MEFs transfected with pCDNA3-N1-ICD-myc and increasing amounts of p3xFLAG-Fbwx7α along with either empty vector (-) or pCDNA3-HA-Pin1. Borders around the panels demarcate juxtaposed parts of the same gel. C) Pin1 overexpression rescues N1-ICD levels in presence of Fbwx7α. Western blot analysis of lysates from SK-BR-3 cells over-expressing N1-ICD-myc, Flag-Fbwx7α along with empty (-) or increasing amounts of HA-Pin1 expressing vector, normalized for co-expressed GFP protein. D) Histogram of Luciferase reporter assays in SK-BR-3 cells co-transfected with pGL2-RBPjκ/LUC and empty vector (-) or pCDNA3-N1-ICD-myc, along with p3xFLAG-Fbwx7 and increasing amounts of pCDNA3-HA-Pin1 as indicated by a wedge. Means, standard deviations and p-value (t-test, n=3) are indicated. Cell lysates were analyzed by Western Blot and are shown below.

Results

Figure 16A showed that in Pin1 depleted mammospheres there is a strong decay also in N4-ICD

levels. Since it has been demonstrated that also N4-ICD is a target of Fbwx7α (Wu et al., 2001),

we speculated that also in this case Pin1 was able to prevent N4-ICD protein degradation.

Interestingly, Pin1 silencing in SK-BR-3 strongly increased N4-ICD proteasome dependent

degradation (Figure 25A) and, accordingly, Pin1 overexpression rescued both N4-ICD protein

levels and transcriptional activity in presence of Fbwx7α as depicted in Figure 25B.

Figure 25. Pin1 protects N4-ICD from Fbwx7α dependet degradation A) Western blot of a CHX chase in SK-BR-3 cells transfected with pCDNA4-N4-ICD-HA along with the indicated RNAi, in presence (+) or absence of proteasome inhibitor Lactacystin. B) Histogram of Luciferase reporter assays in SK-BR-3 cells co-transfected with pGL2-RBPjκ/LUC and pCDNA4-N4-ICD-HA alone or with p3xFlag-Fbxw7α and increasing amounts of pCDNA3-HA-Pin1. Means, standard deviations and p-value (t-test) are indicated for three independent experiments. Cell lysates were analyzed by Western blot and are shown below the histograms.

To confirm and to strengthen these evidences in vivo, we analyzed N1- and N4-ICD protein

levels in the epithelial cells of the mammary gland from Pin1+/+ and Pin1-/- mice. As shown in

Figure 26, from the immunoistochemical and Western Blot analyses performed in Pin1-/- mice it

emerged that the Notch pathway is strongly reduced, while the levels of Fbwx7α seemed not to

be altered, when compared to wild-type mammary tissues.

Results

Figure 26. Pin1-/- mammary epithelial cells display reduced N1- and N4-ICDs protein levels A) Microscope images at 400X magnification of sectioned mammary glands from 8 weeks old mice stained with the indicated antibodies by immunohistochemistry and counterstained with haematoxylin. Stainings were performed without primary antibody (only secondary, IIary Ab) for specificity and with anti-Pin1 antibody as genotype control, respectively. Scale bar is indicated. B) Western blot analyses of primary MECs from indicated female mice. Molecular weights Mr(K) are indicated in KDa.

All these experiments clearly show that Pin1 protects the nuclear active forms of both Notch1

and Notch4 from the major nuclear ubiquitin-ligase Fbwx7α, boosting their protein levels and

transcriptional activity.

5.7 Pin1 protects N1- and N4-ICDs from interaction with Fbwx7α

Pin1 induces prolyl-isomerisation and conformational changes on its targets (Lu et al., 2002a).

Thus we analyzed if Pin1 is able to disrupt the interaction between Fbwx7α with NICDs. For this

aim we performed co-immunoprecipitation (Co-IP) experiments with both endogenous and over-

expressed NICDs and Fbwx7α.

Co-IP in MDA-MB-231 unveiled that the binding between N1- and N4-ICD with Fbwx7α is

increased when Pin1 activity was blocked with PiB (Figure 27A). Accordingly, we recapitulated

these results in another cell line by over-expressing N1-ICDs and Fbwx7α following control- or

Pin1-silencing (Figure 27B). Similar experiments performed in Pin1-/- fibroblasts unveiled that

the catalytic activity of Pin1 was required for this effect. Indeed while over-expression of Pin1 is

able to prevent the binding between N1-ICD and Fbwx7α, the catalytic inactive mutant Pin1S67E

was unable to (Figure 27C).

Results

Figure 27. Pin1 prevents the interaction between N1/N4-ICDs and Fbwx7α A) Inhibition of Pin1 increases interaction between endogenous Fbwx7α and N1-ICD proteins. Western blot analysis of co-immunoprecipitation (Co-IP) experiments between endogenous Fbwx7α and both N1-ICD and N4-ICD from MDA-MB-231 cells treated with DMSO (-) or PiB (+). Anti-Fbwx7α or non-related antibody (NRA) immunoprecipitates (IP) were recognized with anti-N1-ICD (Val1744) and anti-N4-ICD (Val1432) antibody and after stripping with an anti-Fbwx7α antibody. Input levels are shown below.   B) Left Panel: depletion of Pin1 increases Fbwx7α-N1-ICD interaction. Representative Western blot analysis of Co-IP experiments between over-expressed N1-ICD-myc and Flag-Fbwx7α in SK-BR-3 cells. Over-expressed N1-ICD-myc was immunoprecipitated (IP) and subjected to anti-Flag Western Blot to reveal Flag-Fbwx7α Co-IP. Input levels of over-expressed or silenced proteins are shown below. Right Panel: Graph depicting the mean of Flag-Fbwx7α Co-IP levels, normalized to immunoprecipitated N1-ICD. Standard deviations and p-value (t-test, n=3) are shown. C) Pin1 catalytic activity is required to uncouple N1-ICD from Fbwx7α. Co-IP as in (B) in Pin1-/- embryo fibroblasts transduced with the indicated vectors. A)-C) Molecular weights Mr(K) are indicated in KDa.   All these data demonstrated that Pin1 is boosting Notch1/4 nuclear levels by impairing

recognition and subsequent induced degradation by the E3-ubiquitin ligase Fbwx7α.

Results

Figure 28. PP2A is required for the Pin1-dependet N1-ICD detachment from Fbwx7α . A) Inhibition of PP2A enforces Fbxw7a-N1-ICD interaction. Co-IP as above in figure 24 SK-BR-3 cells treated with DMSO (-) or okadaic acid and transduced with the indicated vectors. B) Inhibition of PP2A accelerates the half-life of endogenous N1-ICD. Western blot analysis of a GSI chase of MDA-MB-231 cells treated with vehicle (DMSO) or okadaic acid. Anti-p21Cip1Waf1 immunoblot was added as control for Okadaic acid functioning (Park et al, 2001). C) PP2A is required for Pin1-dependent N1-ICD detachment from Fbxw7a. Co-IP as above in SK-BR-3 cells transduced with the indicated vectors and treated with DMSO (-) or okadaic acid.

Moreover we wanted to uncover the mechanism by which Pin1 is protecting N1/4-ICDs from

Fbxw7α. It is known that prolyl-isomerization of specific phospho-S/T-P sites leads, on certain

substrates, to recognition and subsequent dephosphorylation by the trans-specific phosphatase

PP2A (Liou et al, 2011). Insights from a nuclear Notch1 interactome revealed that PP2A binds to

N1-ICD (Yatim et al., 2012). Hence, we investigated if, following isomerization by Pin1, the

Notch cdc4-phosphodegron could be dephosphorylated by PP2A, thus eluding recognition by

Fbxw7α. To evaluate this possibility we performed co-imunoprecipitation assay between N1-

ICD and Fbxw7α in SK-BR-3 cells treated with the PP2A inhibitor okadaic acid. Interestingly

the interaction between N1-ICD and Fbxw7α increased when cells were treated with okadaic

acid (Figure 28A). Accordingly, when we analyzed the protein half-life of N1-ICD in MDA-

Results

MB-231 we found that okadaic treatment enhanced N1-ICD protein dregradation (Figure 28B).

We next analyzed the dynamics of this interaction following Pin1 overexpression in the same

conditions as above. As shown in Figure 28C Pin1 expression consistently reduced the

interaction between N1-ICD and Fbxw7α but only in the presence of functionally active PP2A,

indicating that PP2A is required for Pin1-dependent N1-ICD accumulation.

5.8 Pin1 and Fbwx7α interplay affects Notch-dependent stem cells traits

Pin1 and Fbwx7α are able to tightly control the Notch signalling pathway at the nuclear level.

Since the role of both Notch1 and Notch4 in the breast CSCs maintenance is well known, we

investigated if the Pin1-Fbwx7α interplay on the Notch pathway can affect the CSC population.

For this aim we performed mammosphere assays with MDA-MB-231 by over-expressing Pin1

and Fbwx7α. Interestingly Fbwx7α over-expression had a great effect in reducing the

mammosphere formation, the percentage of putative CSCs as detected by the reduced number of

Aldh+ cells and also the nuclear levels of both N1- and N4-ICDs and the transcriptional target

HES1 (Figure 29). This inhibitory effect on the CSC populations mediated by Fbwx7α is almost

completely abolished by a concomitant over-expression of Pin1 that, through stabilization of

N1/N4-ICDs, recovered the mammosphere formation ability and the percentage of Aldh+ cells.

To verify the contribution of the Notch signalling in this context, we treated cells in all the

experimental points with the Notch inhibitor GSI, that abolished the CSCs population.

Results

Figure 29. Pin1 and Fbwx7 modulate Notch breast CSCs activity in vitro A) %M2FE of MDA-MB-231 cells overexpressing the indicated vectors. Black and grey bars indicate DMSO or GSI treated mammospheres, respectively. Means, standard deviations and P-values (t-test, n = 3), are indicated. B) Representative microphotographs of M2 are shown, 200 lm scale bar is indicated. C) Percentage of Aldh positive cells of clones from Figure 7A grown as mammospheres, calculated as Median flurescence intensity (MFI) difference between samples containing BAAA and that of the same samples containing also DEAB. D) Western Blot of cell lysates from M2. White and black arrows indicate over-expressed and endogenous Pin1, respectively. Molecular weights (Mr) are indicated in kDa.

Since Fbwx7α acts on several important targets (Wang et al, 2011), we assessed the relevance of

Notch activity for this phenotype. As shown in Figure 30, down-modulation of Fbwx7α caused a

consistent increase of N1-ICD levels, boosted M2FE while treatment with GSI elicited a strong

reduction of both mammosphere formation and Notch pathway activation, thus demonstrating

that among several Fbwx7α targets, Notch is the critical one in this cellular context.

Results

Figure 30. Fbwx7 depletion increased mammospheres formation through Notch pathway Left panel: Histogram showing %M2FE of MDA-MB-231 cells with the indicated siRNAs and treated with vehicle (black bars) or GSI (grey bars). Means, standard deviations and P-values (t-test, n = 3), are indicated. Right panel: WesternBlot of cell lysates from M2. Molecular weights (Mr) are indicated in kDa.

We next addressed the Pin1-Fbwx7α interplay in CSCs maintenance in an in vivo experiment by

testing the tumour initiation properties. In particular, we performed a limiting dilution assay by

injecting increasingly diluted MDA-MB-231 cells over-expressing Fbwx7α, or Fbwx7α plus

Pin1, in the inguinal mammary glands of SCID mice. As shown in Table 1, while 300.000

injected control cells gave rise to a tumour in all mice, Fbwx7α strongly impaired the tumour

initiating properties, since only 2 out of 9 mice harboured palpable tumours. Notably,

concomitant Pin1 over-expression significantly rescued the tumour initiation capability in Fbxw7

over-expressing MDA-MB-231 to 7 out of 9 mice. At the subsequent dilution the same trend

could be appreciated, demonstrating in vivo a potent counteraction of Pin1 and Fbwx7α in

controlling the CSCs population.

Table 1. Mice were transplanted with decreasing numbers of MDA-MB-231 cells overexpressing empty, Fbwx7α, or Fbwx7α + Pin1 vectors (number of injected cells is indicated). Results are shown as the number of tumors per number of injected mice (upper panel). CSC frequencies (estimates and upper/lower limits) were calculated by limiting dilution analysis, as described in Materials and methods. Differences in CSC frequencies are indicated for each sample against the empty vector and for Fbwx7α + Pin1 against Fbwx7α only. Their significance is indicated by a P value (lower panel).

1.000.000 600.000 300.000 150.000

empty 6/6 9/9 9/9 12/15

Fbwx7α 6/6 9/9 2/9 1/15

Fbwx7α + Pin1 6/6 7/8 7/9 5/15

Estimate (0.95) (Lower-Upper) P value P value

empty 1:144232 (1:84380-1:49364)

Fbwx7α 1:776940 (1:485184-1:302989) 1.58 x 10-6

Fbwx7α + Pin1 1:466587 (1:302232-1:195771) 1.04 x 10-2 0.0239

Results

The content of CSCs in a tumour is linked to several aggressive features, such as tumour growth

and metastasis formation. To gauge for these features, we performed xenograft experiments by

injecting 1.000.000 of cells/flank with the same MDA-MB-231 clones used above. Notably

Fbwx7α exerted a potent tumour suppressor activity as emerged by the fact that, compared to

empty vector infected cells, MDA-MB-231 over-expressing Fbxw7α gave rise to small tumour

with an impaired Notch signalling, a reversal of EMT, and a reduced number of metastases to the

nearby lymph nodes and to the lungs. By contrast tumours that co-expressed Pin1 and Fbwx7α

were growing faster, with a re-activation of the Notch signalling, a rescue of mesenchymal

markers and dissemination of metastases (Figure 31).

In summary, we uncovered a potent role of Fbwx7α in reducing the breast CSC traits through a

down-regulation Notch1/4 activity by promoting degradation of their nuclear active forms. In

this context, concomitant Pin1 over-expression reverted all these effects boosting N1/4-ICDs

protein stability and levels.

Results

Figure 31. Pin1 and Fbwx7α modulate Notch breast CSCs activity in vivo A) Tumour volume of orthotopic xenografts in SCID mice obtained from the indicated MDA-MB-231 cell clones. B) qRT-PCR of Fbwx7α and Fbwx7α+Pin1 tumor xenografts relative to control tumors (empty) explanted at the end of the experiment. C-D) Western blot of Fbwx7α and Fbwx7α + Pin1 tumor xenografts relative to control tumors (empty) explanted at the end of the experiment E-F) Representative images of colonized lymph nodes E) and hematoxylin and eosin stained pulmonary sections F) are shown. Rulers and Scale bars (1 mm) are indicated for calibration, arrows indicate metastatic areas. A), B), E), F) Means, standard errors of the mean and P-values (t-test, n = 6), are indicated. C), D) Molecular weights Mr(K) are indicated in kDa.

Results

5.9 Pin1 sustains Notch signalling in primary breast cancers in spite of Fbxw7a expression

According to CONAN, Cosmic and TCGA databases and recent publications, FBXW7 and

NOTCH a rarely mutated in breast cancer (Byrd et al., 2008; Ibusuki et al., 2011; Mao et al.,

2008; Santarpia et al., 2012), but its known that Notch pathway is over-activated in many breast

cancers (Farnie et al., 2007; Han et al., 2011; Pece et al., 2004; Reedijk et al., 2005; Rizzo et al.,

2008; Rustighi et al., 2009; Xu et al., 2012). We have discovered a mechanism by which, despite

absence of mutational events and presence of the ubiquitin-ligase Fbwx7α, N1- and N4-ICD are

up regulated because of high Pin1 protein levels.

This hypothesis prompted us to analyze breast cancer tissues from patients to address if our

mechanism occurs also in human patients. We took 43 breast cancer tissues from which the

protein levels of N1-ICD had been already detected by IHC (Verzemovic et al., submitted). We

extracted the total RNA from these tissues and analyzed the mRNA levels of Pin1 and Fbwx7α;

notably it has been demonstrated that their mRNA levels provide a direct correspondence to

protein levels. The tissues were grouped, based on the N1-ICD protein levels, into high or low

expressing (Figure 32). Interestingly, 11/17 tissues displaying high N1-ICD protein levels,

expressed also high FBXW7α mRNA levels and more importantly all of them expressed high

PIN1 levels. Moreover while FBXW7α did not correlate with any of the categories, PIN1 levels

were higher in the high N1-ICD expressing group. This evidence suggests that high N1-ICD

protein levels can coexist with its negative regulator Fbwx7α because of high Pin1 expression.

Figure 32. Association of FBWX7Α or PIN1 expression in N1-ICD high and low breast cancers A) Heatmap representing the expression of PIN1 and FBWX7 (detected at mRNA expression level) and N1-ICD (detected at protein level) in a cohort of 43 breast cancer patients. The color in each block represents if the mRNA or protein is above (red) or below (blue) the average value of the samples for each gene or was scored high or low by immunohistochemistry, respectively B) Left panel: Box-plot representations of gene expression values (of FBWX7 and PIN1, upper and lower panels, respectively) in the breast cancer cohort of Figure S8A. Patients were binned in low or high categories for weak or strong N1-ICD staining (detected by IHC, see Methods for detail). Finally, to test the associations between N1-ICD staining and expression of FBWX7 and PIN1, Wilcoxon rank sum tests were performed. The results showed no association for FBWX7 (W = 192, p-value = 0.8998) and very strong association for PIN1 (W = 35, p-value = 2.967 x 10-6). Right panel: Contingency table showing percentage of each category calculated on the precedent category of patients from A).

Results

To strengthen our observation we analyzed another cohort of breast cancer tissue from 38

patients (Figure 33). We selected in particular tissues of triple negative breast cancer subtypes,

which have been described to be chemoresistant, highly metastatic and of grade 3, suggesting

that they are also enriched in CSCs (Pece et al., 2010). We analyzed the protein levels of Pin1,

N1-ICD ad Fbwx7α by immunohistochemical analysis. We found that 22/38 tissues display high

N1-ICD nuclear protein levels and in 72.7% of them there was also a strong nuclear

accumulation of Fbwx7α. Again, a large part of this group presented also high Pin1 protein

levels as shown in Figure 32.

Figure 33. High N1-ICD levels in human breast cancers coexist with Fbwx7α thanks to high Pin1 expression Left panel: Heatmap representing the protein levels of Pin1, Fbwx7α and N1-ICD in a cohort of 38 breast cancer patients. The colours represent high (red) or low (blue) protein levels according to protein expression scores (see supplementary Methods). Right panel: Contingency table showing percentage of each category calculated on the precedent category of patients; chi-square test was performed for independence between the variables and the P-value = 105.

Finally we analyzed a medataset of la arge cohort of 3254 breast cancer patients collected from

19 independent studies (see Material and Methods). We stratified patients into high or low levels

of FBXW7α and PIN1 expressing patients (Adorno et al., 2009). Since the Notch mRNA is not a

representative of N1-ICD protein levels, we generated a signature, that reflects the Notch

pathway activation, since it is was created by selecting Notch direct transcriptional targets (NDT

signature) (see Material and Methods). 48% of all samples exhibited high NDT expression and

correlated with poor overall survival (Figure 34). Notably 51.7% of these samples displayed also

high FBWX7α mRNA levels and again the majority of this group presented high PIN1 levels.

Results

Figure 34. High N1-ICD levels in human breast cancers coexist with Fbwx7α thanks to high Pin1 expression Upper panel: heatmap representing the contingency table frequencies of samples classified as having high or low levels of FBWX7Α, of PIN1 and of the NDT gene signature. Number of samples in each category is indicated on the left. The association among high levels of NDT gene signature, PIN1, and FBWX7Α resulted statistically significant (P < 0.001; chi-square test). Lower panel left: Contingency table showing percentage of each category calculated on the precedent category of patients. : Expression correlation between NDT and PIN1 and FBWX7Α mRNA levels. Average expression of NDT gene signature in breast cancer samples stratified according to high or low expression of PIN1 and FBWX7Α mRNA. Data are shown as mean and standard error of the mean (s.e.m.). Lower panel right: Kaplan-Meier graphs representing the probability of survival in 1173 breast cancer patients from the meta-dataset of the Upper panel, with outcome information on survival. Tumor samples were classified as high or low Notch-dependent Direct Target (NDT) gene signature using the classifier described (Adorno et al., 2009). The log-rank test p value reflects the significance of the association between NDT gene signature low and longer survival.

In summary, all the 3 analyses confirmed our hypothesis that in breast cancer patients N1-ICD

could coexist with its ubiquitin-ligase thanks to high Pin1 expression. Moreover the category

displaying high NDT and Fbwx7α expression but low Pin1 is the less represented compared to

Results

any other category. Accordingly, the average expression value of NDT was contingent on Pin1,

while the Fbwx7α status was not influential, strengthening the idea that Pin1 is dominant, over

Fbwx7α, in modulating Notch pathway in breast cancers (Figure 34).

In this work we discovered a functional interaction between Pin1, N1-ICD and N4-ICD in breast

CSCs and aggressive breast tumour traits. Since all these features are connected to poor clinical

outcome in patients, we analyzed in our cohort from figure 34 the survival of patients. Notably,

as shown in Figure 35, we found that in the grade 3 breast cancers of this cohort, high Pin1

levels correlated with a worse outcome only in patients with activated Notch signalling. By

contrast, in patients with low Notch signalling, Pin1 is unable to stratify patients, thus indicating

that Pin1 requires Notch activation to promote aggressiveness and poor clinical outcome.

Figure 35. Pin1 levels correlate with poor clinical outcome only in patients with high NDT signature. A) Kaplan–Meier survival curve is indicated for high NDT signature, grade 3 breast cancer patients of the metadataset in function of high or low PIN1 mRNA levels. P-value and the number of subjects at risk at each time point is indicated below. B) Kaplan–Meier survival curve is indicated for low NDT signature, grade 3 breast cancer patients of the metadataset in function of high or low PIN1 mRNA levels. P-value and the number of subjects at risk at each time point is indicated below. To understand which molecular pathways are over-represented in patients with high Pin1/high

NDT compared to patients with high Pin1/low NDT, we performed a Gene set enrichment

analysis. Interestingly, as depicted in Table 2, we found enrichment of several stem cell pathway

signature genes in this group of patients, supporting our findings of a pro-CSCs program induced

by the cooperation between Pin1 and the Notch pathway.

Results

Table 2. Enrichment of gene signatures in the list of genes preferentially expressed in NDT signature HIGH/PIN1 HIGH versus NDT signature HIGH/PIN1 LOW tumors in the metadataset (GRADE 3). Differentially expressed genes (n=1460, q-value <0.01 and fold change >1.5) were subjected to Gene set enrichment analysis (GSEA) as already described (Montagner et al., 2012). Enrichment has been determined using a Fischer’s exact test on all 254 signaling pathways in the database. P-values have been corrected using BH procedure

SIGNATURES

(BIOCARTA PATHWAY + OTHER

SIGNATURES)

ENRICHMENT in

NDT HIGH/PIN1 HIGH

(q-value)

ES1

ES.like

stem_tumorig

Mut.p53_up

0.008710

0.021817

0.029714

0.034324

Discussion

6. DISCUSSION

Cancer stem cells represent a major problem for breast cancer treatment, as they are responsible

for tumour heterogeneity, recurrence, chemoresistance metastatic dissemination. Many studies

have been carried out to understand the molecular bases of their sustenance, that have led to the

identification of several involved signalling pathways, transcriptional factors and micro-

environmental stimuli showing that CSCs may take advantage of the normal stem cell molecular

machinery to maintain their self-renewal. Here we demonstrate that the prolyl-isomerase Pin1 is

required for both normal and cancer stem cell biology. These Pin1 pro-stemness functions are

mediated through the maintenance of both N1 and N4-ICD protein levels, protecting them from

the tumour suppressor and E3-ubiquitin ligase Fbwx7α. Pin1 and Notch signalling promote each

other since Notch1 directly induces Pin1 transcription(Rustighi et al., 2009). Thus the Pin1-

Notch1/4 axis promotes a pro-CSCs program by up-regulating EMT markers, stemness

signalling pathways and proteins involved in chemoresistance, regardless of Fbwx7α, as depicted

by the model in Figure 36. As a consequence, genetic or pharmacological ablation of Pin1

suppresses CSC number and self-renewal, reduces the number of tumour initiating cells, the

metastases dissemination and, more importantly, recovers sensitivity to chemotherapy both in

vitro and in vivo.

Figure 36. Schematic model that depicts the role of Pin1 in sustaining CSCs through Notch1 and Notch4 by antagonizing Fbxw7α-mediated destruction. In G3 breast cancer patients, high Pin1 levels are associated to poor clinical outcome when

Notch signalling is activated. Therefore, since it has been demonstrated that G3 are enriched in

CSCs, it is conceivable that a major mechanism by which Pin1 is involved in cancer progression

might be its promotion of the CSCs number. In addition, Pin1 has not yet been linked to any

particular subtype or unique mutational event in breast cancer, nevertheless its levels increase in

parallel with the tumour grade and it cooperates with several oncogenic pathways. In particular it

cooperates with several oncogenes or oncogene-induced factors, that are involved in almost all

Discussion

the breast cancer subtypes such as estrogen receptor alpha (ERα+) tumors (Rajbhandari et al.,

2012), progesterone receptor (PR) (Yi et al., 2005), Her2/Neu (Wulf et al., 2003), mutp53

(Girardini et al., 2011) and Notch1/4. Moreover Pin1-/- mice display reduction of the stem cell

compartment without perturbing the luminal or myoepithelial fate commitment, as suggested by

an unchanged proportion of the myoepithelial and luminal populations in Pin1-/- mice (Figure

9). Therefore, it seems that Pin1 goes beyond the heterogeneity of breast cancer, since it

impinges on the heart of basic oncogenic mechanisms, and it could therefore be an important

molecular target for breast cancer therapy. Indeed, ablation of Pin1 can reduce the CSCs

population, the specific subtype driven breast cancer oncogenes and elicit sensitivity to

chemotherapy.

A role for both Pin1 and Notch pathway in the promotion of chemoresistance has been well

demonstrated (Ding et al., 2008; Ranganathan et al., 2011). Of note, the role of Pin1 is

controversial, since on one hand it promotes apoptosis through p53 following genotoxic stress

(Wulf et al., 2002; Zacchi et al., 2002), on the other it promotes chemoresistance acting on Mcl-1

stability (Ding et al., 2008). Nevertheless, to date no in vivo experiment has been done to address

how Pin1 dictates the cellular response to pharmacological treatments. Here we provide a clear

demonstration of the pro-survival role of Pin1 in presence of chemotherapy in breast cancers

with in vitro and in vivo experiments. In particular breast tumour cells and CSCs depleted for

Pin1 display reduction of the Notch signalling, of the drug efflux pump ABCG2 (Sarkadi et al.,

2006), of the pro-survival factor SURVIVIN (Fukuda and Pelus, 2006), of the detoxification

enzyme Aldh (Ginestier et al., 2009), and other factors for which a role in chemoresistance is

already known. This evidence coupled with the fact Pin1 promotes exhaustion of CSC, point out

Pin1 as an important mediator of chemoresistance in breast cancers through many mechanisms.

Notch1 and Notch4 receptors have been linked to many diseases such as cancer (Ranganathan et

al., 2011). Notch1 has been associated for the first time to tumourigenesis in human T-cell acute

lymphoblastic leukemia (T-ALL) due to mutational events that unleash high levels of N1-ICD in

tumour cells and, accordingly, in 50% of T-ALL there are mutational events on NOTCH1 or

FBXW7α (Ferrando, 2009). Moreover, in leukemia cells, high levels of Notch signalling are

important for the CSCs maintenance (Liu et al., 2013). Intriguingly Notch proteins are also over-

activated in breast cancer and important for normal and cancer stem cells; while Notch1 seems to

be more important for luminal progenitors, Notch4 seems to be more related to bipotent stem

cells (Dontu et al., 2004; Harrison et al., 2010; Raouf et al., 2008). Despite the high percentage

Discussion

of mutational events in T-ALL, in breast cancer there are rare mutations on FBXW7α and

NOTCH1 which can not provide molecular explanation for the great percentage of mammary

tumours displaying high levels of N1/4-ICD protein levels. Our findings unveil a mechanism by

which Pin1 is responsible for elevated Notch signalling in breast cancers without the need of

mutational events on NOTCH and FBXW7α. Thus, an over-activated Notch-Pin1 axis can

promote normal and cancer stem cells despite high levels of functional Fbwx7α.

New insights for the tumour suppressor Fbxw7α in breast cancer

There is only limited evidence in the literature that links Fbxw7 to breast tumours. Nevertheless

this ubiquitin-ligase is a potent tumour suppressor in many other tumours. Overall,

approximately 6% of tumours harbour mutations in FBW7, but there is substantial variation

among tumour types. The highest mutation rates (approximately 30%) were found in

cholangiocarcinoma and T‑cell acute lymphoblastic leukaemias (T-ALL) whereas

gastrointestinal cancers (pancreatic and gastric) as well as prostate and endometrial cancers had

mutation frequencies in the range of 4–15% (Welcker and Clurman, 2008). By contrast, many

tumour types do not exhibit or only rarely exhibit mutations in FBW7, in particular breast

cancers. Rather, very often the oncogenic substrate proteins of Fbxw7, such as c-Myc or Notch1,

have the CPD mutated in order to escape Fbxw7-mediated recognition and proteasomal

degradation.

In tumours with low mutations on Fbxw7 gene, deregulation of its protein levels is not

documented and the role of Fbw7α inactivation by dominant oncogenes in human tumours is not

known, with exception of human colon cancers where Fbxw7α protein levels were found

decreased due to high levels of Pin1 (Min et al., 2012). By our observations in mice and human

breast cancer samples, we did not find any correlation between Fbxw7α and Pin1 levels.

Notably, and unexpectedly, also N1-ICD nuclear levels did not correlate with those of its

ubiquitin-ligase Fbxw7. Notwithstanding, there might be a proportion of breast cancer patients

where Pin1 and Fbxw7 are inversely correlated due to inactivation of Fbxw7 by Pin1, suggesting

a further level of cooperation between Pin1 and several oncogenes in cancer. Here, we have

described that a quite high number of breast cancer patients have high N1-ICD levels despite

high levels of Fbxw7, due to the concomitant presence of Pin1 that protects N1/4-ICDS from

degradation. Ablation of Pin1 allowed Fbxw7 to interact with and turn-off the Notch pathway

strongly impacting on tumourigenesis. Indeed in vitro and in vivo experiments demonstrated that

Discussion

Fbxw7α over-expression strongly impaired the number and the self-renewal of breast CSCs,

curbed tumourigenesis formation and metastases dissemination. Interestingly Fbxw7α controls

the number and maintenance of stem cells in many other cells such as neurons, colon and

hematopoietic cells mostly through regulation of the Notch signalling (Wang et al., 2012). It will

be interesting to investigate if, in absence of mutational events on NOTCH or FBXW7 genes,

also in these tissues Pin1 could cooperate with the Notch signalling pathways in the maintenance

of normal and cancer stem cells.

The interplay of the Notch-Pin1 axis with other signalling pathways

The interplay between Pin1 and N1-ICD is not simply restricted to NICDs protein stabilization

and Pin1-induced transcription. Indeed the Notch pathway is highly inter-connetected with

several signalling pathways such as PI3K/Akt, Jak/STAT, NF-kappaB, ErbB2, Wnt, and HIF1

(Ranganathan et al, 2011). Strikingly, considering that most of them are also Pin1 substrates

(Kim et al, 2008; Liou et al, 2011), this isomerase may contribute to increase Notch induced

oncogenesis also by amplifying the signalling emerging from these cross-talks. In addition, the

death associated protein kinase 1 (DAPK1), a tumour suppressor that negatively controls Pin1

activity (Lee et al., 2011), was shown to be repressed by N1-ICD (Li et al., 2008), suggesting

that Notch signalling is not only boosting Pin1 transcription but also its activity.

Another important interactor of both Notch signalling and Pin1, is the tumour suppressor p53

(Beverly et al., 2005; Kim et al., 2007; Sun et al., 2011; Wulf et al., 2002; Zacchi et al., 2002).

Interestingly wild-type p53 (wtp53) has a fundamental role in the regulation of the breast stem

cell compartment, since its depletion leads to an uncontrolled symmetric expansion boosting the

number of stem cells (Bonizzi et al., 2012; Cicalese et al., 2009; Insinga et al., 2013). Notably,

pharmacologic inhibition of the Notch signalling in p53 depleted CSCs is able to revert the

phenotype with a consequent reduction of CSCs number (Tao et al., 2011). The role of wtp53 in

our findings is not evident, raising the question on whether in the mammary stem cell

compartment Pin1 could act by modulating p53 and Notch functions in concert, unveiling novel

intersections between these two pathways. While in vitro the activity of Pin1 on p53 has been

well characterized in the context of genotoxic stress signalling acting on the wtp53, their

interplay in the stem cell compartment remains to be elucidated. The fact that Pin1-/- mice have

reduced mammary stem cells (Figures 1 and 2), in contrast with what has been observed in the

TP53-/- genotype (Bonizzi et al., 2012; Cicalese et al., 2009), allows speculating a more complex

scenario for a wtp53-Pin1 interplay in these cells.

Discussion

On the other hand, also for oncogenic gain of function (GOF) mutant p53 proteins, there is

growing evidence that supports a role in promoting cellular reprogramming and EMT and

inciting expansion of mammary epithelial stem cells giving rise to mammary tumours (Chang et

al., 2011; Dong et al., 2013; Lu et al., 2013b; Sarig et al., 2010). Nevertheless no in vivo

experiment has been done yet to understand the molecular processes controlled by mutp53 in

breast stem cells. In this context, we have recently demonstrated a key role of Pin1 in promoting

mutp53 GOF in breast cancer (Girardini et al., 2011). On these bases, it is conceivable that in

CSCs lacking wtp53 or expressing oncogenic mutp53, increased levels of Pin1, due to a

hyperactivated Notch pathway, may foster stem cell traits by acting both on mutp53 oncogenic

properties and on Notch itself. In support of this hypothesis, our analysis revealed that poorly

differentiated grade 3 and NDT-high/PIN1-high breast cancers are enriched for the expression of

stem cell and mutp53 signature genes.

In conclusion, this study has two major impacts. First, our findings provide explanation to

understand how, in the absence of mutations in NOTCH genes, in a consistent proportion of

breast cancer patients, activated Notch1 can exist in spite of the presence of its major constrain,

the tumour suppressor Fbxw7α. The data presented, in fact, suggest that deregulated activity of

Pin1, at least in breast cancer, where mutational events in NOTCH1 and FBXW7 are rare, could

reduce the selective pressure for mutations providing an alternative mechanism for Notch

oncogenic activation.

Second, our findings pinpoint Pin1 as a crucial target in aggressive breast cancers, providing the

rational for a therapeutic strategy based on Pin1 inhibition to hit CSCs, restore chemo-

sensitiveness and inhibit metastatic spread.

Material and Methods

7. MATERIAL AND METHODS

Cell lines and treatments.

MDA-MB-231, SK-BR-3, BT-549, and SUM-159 are human breast carcinoma cells, HEK 293T is a

human embryonic kidney cell line with SV40 large T, immortalized Pin1-/- fibroblasts have been obtained

by spontaneous immortalization from Mouse Embryo Fibroblasts of C57BL6/129Sv mixed background

(Rustighi et al., 2009). COS-7 are monkey kidney cells immortalized with SV40 large T antigen. NOP6 is

a mouse mammary tumor cell line (Yang et al., 2009). All cells were cultured in DMEM, supplemented

with 10% Fetal Bovine Serum (Gibco) and Penicillin/Streptomycin, as described (Rustighi et al., 2009).

For NOP6, in addition Insulin supplements were added (Yang et al., 2009). MCF10A cells were

maintained in DMEM:F12 Ham’s (1:2; Sigma), supplemented with 5% horse serum (Gibco), insulin (10

µg ml–1; Sigma), hydrocortisone (0.5 µg ml–1) and EGF (20 ng ml–1; Peprotech). Primary breast cancer

cell lines were maintained in F12 Ham’s supplemented with 10% FCS and insulin, hydrocortisone and

EGF as above. Transient transfections, retroviral or lentiviral infections were performed by standard

procedures, as described (Rustighi et al., 2009). pEGFP was included as control for transfection

efficiency. For creation of stable clones, a selection corresponding to the expressed vectors was applied

for 2 weeks to infected cells at the concentrations of 2 µgr/ml for Puromycin and 0,5mg/ml for

Blasticidin. Cycloheximide (Sigma) for chase experiments was used at 50 nM concentration. Treatment

with gamma-secretase inhibitor DAPT (Sigma), Pin1 inhibitor PiB (Calbiochem) or proteasome inhibitor

lactacystin (Sigma) have been described previously (Rustighi et al., 2009). Protein phosphatase inhibitor

Okadaic Acid (Sigma) was dissolved in DMSO and used at a final concentration of 200 nM for 6h.

Isolation and purification of mammary epithelial cells.

Mammary glands from 8 to 12-week-old virgin female mice were enzymatically digested and single cell

suspensions of purified mammary epithelial cells were obtained, as described (Sleeman et al, 2006; Stingl

et al, 2006). Briefly, Mammary glands from 8 to 12-week-old virgin female mice were digested for 1–2 h

at 37°C in Epi-Cult-B medium (StemCell Technologies Inc, Vancouver, Canada) with 600 U/ml

collagenase (Sigma-Aldrich, St. Louis, MO, USA) and 200 U/ml hyaluronidase (Sigma). After lysis of

the red blood cells with NH4Cl, the remaining cells were washed with PBS/0.02% w/v EDTA to allow

cell-cell contacts begin to break down. Cells were then dissociated with 2 ml trypsin 0.25%w/v, 0.2% w/v

EDTA for 2 min by gentle pipetting, then incubated in 5 mg/ml Dispase II (Sigma) plus 1 lg/ml DNase I

(Sigma) for 5 min followed by filtration through a 40 lM cell strainer (BD Falcon, San Jose, CA, USA).

Mammary epithelial cells were then purified using the EasySep Mouse Mammary Stem Cell Enrichment

Kit (StemCell Technologies Inc).

Material and Methods

Flow cytometric analyses and sorting (FACS)

Mouse mammary epithelial lineage-depleted cells, pre-enriched using the EasySep Mouse Mammary

Stem Cell Enrichment Kit (see above), were analysed by FACS and sorted to near purity (85%) with

antibodies against CD49f and CD24. FACS analysis and sorting from mammospheres based on aldehyde

dehydrogenase (Aldh) activity were performed using the Aldefluor kit (StemCell Technologies Inc)

following the manufacturer’s instructions and are detailed in Supporting Information. Briefly, cells were

incubated with active substrate (BAAA) in presence or absence of a specific adehyde dehydrogenase

inhibitor (DEAB) for 30-60 min at 37°C, to allow conversion of BAAA into a fluorescent product

(BAA). Fluorescence was measured by flow cytometry (FACS Calibur) using adjusted FSC and SSC

voltages to center the nucleated and viable cell population. DEAB treated vs. non treated cells were used

to identify Aldefluor positive cells. Median fluorecence intensity (MFI) was derived from each sample

and the baseline fluorescence, used to identify the Aldefluor positive and negative populations, was

calculated by the difference between the MFI of sample containing BAAA and the MFI of the same

sample containing also DEAB. Sorting of the populations of interest was performed on ARIA II cell

sorter (Beckton Dickinson) to near purity (85%). CD44/CD24 flow cytometric analysis was performed

with mouse anti-human PE conjugated anti-CD44 and FITC conjugated anti-CD24 antibodies (BD).

Plasmids, transfection, retroviral and lentiviral transfection

For DNA trasfection Lipofectamine 2000 (Invitrogen) was used and for siRNA trasfection double-

stranded RNA oligos (10pml/cm2) were transfected using Lipofectamine RNAiMax (Invitrogen)

according to manufacter’s instructions. Retroviruses were made by transient trasfection of 293GP

packaging cells with the appropriate plasmids with pMD2-ENV coding for envelope proteins, using

standard calcium-phosphate metod. After 48h incubation at 32°C, the supernatants containing viral

particles were collacted and infection was performed as described (Rustighi et al., 2009) pCDNA3-N1-

ICD-myc, pCDNA3-N1-ICD-deltaPEST(d2444)-myc, pCDNA3-N1-ICD-T2512A-myc, were generated

by PCR and standard cloning procedures starting from pCDNA3-NdeltaE-myc (Rustighi et al., 2009)

constructs as templates. pCDNA3-HA-Pin1 and pCDNA3-HA-S67E, retroviral pLPC-HA-Pin1,

pCDNA3-HA-Pin1r, were already described (Rustighi et al., 200; Girardini et al., 2011). Retroviral

pMSCV-3X-Flag-Fbxw7α was obtained by PCR and subcloning from p3X-Flag-Fbxw7α coding for the

human alpha isoform. pCDNA4-N4-ICD-HA was a kind gift of I. Prudovsky and was already described

(MacKenzie et al., 2004). For lentiviral infection an shPin1 corrsponding to the siRNA sequence Pin1#1

was cloned into the Doxicyclin-inducible pLKO-Tet-ON (Addgene) following the “All-in-one” system

described by D.Wiederschain (dmitri.wiederschain@novartis.com). For preparation of viral particles

psPAX2 and pMD2.G were used in HEK 293T cells.

Material and Methods

Oligonucleotides for cloning and mutagenesis:

Plasmids Forward 5’-3’ Reverse 5’-3’

pLKO-TetO-shPin1 CCGGCGGGAGAGGAGGACTTTGACTCGAGTC

AAAGTCCTCCTCTCCCGTTTTT

AATTAAAAACGGGAGAGGAGGACTTTGAC

TCGAGTCAAAGTCCTCCTCTCCCG

pCDNA3-N1-ICD

Threonine 2512 to

Alanine

CACCCCTTCCTCGCGCCGTCCCCTGAGTCCCC

TGAC

GAGGAAGGGGTGCTCAGGCACCTGTAG

pMSCV-3XFlag-Fbxw7α TATAAAGATCTTATGATCGACTACAAAGACG

AT

GAGCGGATCCATGGCACGCAAGCGCCGG

siRNA sequences:

Mammosphere cultures

To obtain mammospheres, cells from monolayer cultures were enzymatically disaggregated (0.05%

trypsin–EDTA, Gibco) to a single cell suspension, passed though a 40 lm cell strainer (BD Falcon),

plated at clonogenic density (2500 cells/cm2), and grown in nonadherent culture conditions, as described

(Dontu et al, 2003). In detail, cells were grown for 7–10 days in DMEM:F12 (1:1) supplemented with

B27 (Invitrogen Corporation, Carlsbad, CA, USA), 20 ng/ml EGF (PROSPEC, East Brunswick, NJ,

USA), 20 ng/ml bFGF (BD Biosciences, San Jose, CA, USA), 4 µg/ml heparin (StemCell Technologies

Inc.), 0.5 µg/ml hydrocortisone (Sigma) and 5 µg/ml Insulin (Sigma) in low attachment 24 or 96 well

plates (Coroning) in a humified incubator at 37°C, 5% CO2. Primary mammospheres (≥200 µm) were

obtained, collected, counted and again enzymatically disaggregated as above to re-plate cells at

clonogenic densities to obtain secondary mammospheres. The same procedure was applied starting from

secondary mammospheres to proceed to tertiary and quaternary mammospheres. Percentages of

mammosphere forming efficiencies (%MFE) were calculated as number of mammospheres divided by

the plated cell number and multiplied by a hundred. Mammospheres were counted with a 20× objective

on an Olympus CK30 microscope (Olympus Italia Srl, Milan, Italy).

Quantitative Real Time PCR (qRT–PCR) analysis

Total RNA from cell lines and xenografts was extracted with QIAzol Lysis Reagent (Qiagen Srl-Italy,

Milan, Italy). Total RNA from formalin-fixed, paraffin-embedded samples of breast cancer patients was

extracted starting from 2 to 3 20 lm slices with the HighPure RNA paraffin Kit (Roche SpA, Monza,

siRNA Sequence 5’-3’ Ref

siFbxw7α ACCTTCTCTGGAGAGAGAAATGC Welcker et al., 2004

siPin1 #1 CGGGAGAGGAGGACUUUGA Girardini et al., 2011

siPin1 #2 GCCAUUUGAAGACGCCUCG Girardini et al., 2011

siNotch1 GUGUCUGAGGCCAGCAAGA

siCtrl Allstar NEGATIVE control, Qiagen

Material and Methods

Italy) and cDNA was transcribed with QuantiTect (Qiagen) in accordance with the manufacturer’s

protocols, then amplified on a StepOne Plus cycler (Applied Biosystems, Life Technologies Europe BV,

Monza, Italy), using SYBR Green Universal PCR Master Mix (Applied Biosystems). Histone H3 and

GAPDH mRNA were used as internal controls.

Oligonucleotides for quantitative real-time PCR:

All oligonucleotides were supplied by MWG.

Antibodies for Western blot, Far Western, Immunoprecipitations, and Immunohistochemistry.

The following antibodies were used: rabbit and goat polyclonal anti-Notch1 (C-20: sc-6014, and S-20: sc-

23304, SantaCruz), rabbit polyclonal anti-N1-ICD Val1744 (#2421, Cell Signalling), rabbit polyclonal

anti-N4-ICD Val1432 (SAB4502023, Sigma), rabbit polyclonal anti-Notch4 (H-225, sc-5594), rabbit

polyclonal anti-Pin1 (Rustighi et al., 2009) and mouse monoclonal anti-Pin1 (G-8: sc-46660, Santa Cruz),

rabbit polyclonal anti-HES-1 (#AB5702, Millipore), rabbit polyclonal anti-Slug (#9585S, Cell Signaling),

mouse monoclonal anti-Vimentin (ab8069, Abcam), mouse monoclonal anti-E-cadherin (610182, BD),

rabbit monoclonal anti-cleaved Caspase-3 (5A1E, Cell Signaling), rabbit polyclonal anti-Mcl-1 (S-19: sc-

Gene symbol Forward 5’-3’ Reverse 5’-3’ Ref

Mouse Pin1 GTCTCAGGGATGGGGCTTTT TGGTGGGGCTCAGAGGTATT

Mouse SMA TGATCACCATTGGAAACGAACG TGGTTTCGTGGATGCCCGCT Stingl et al., 2006

Mouse CK14 TGAGAGCCTCAAGGAGGAGC TCTCCACATTGACGTCTCCAC Stingl et al., 2006

Mouse CK18 CTTGCTGGAGGATGGAGAAG CTGCCATCCACGATCTTACGG Stingl et al., 2006

Mouse CK19 CAGGCTGGAGCAGGAGATCG TGGTAGCTCAGATGGCCTTGG Stingl et al., 2006

Mouse GAPDH TTCACCACCATGGAGAAGGC CCCTTTTGGCTCCACCCT Li et al., 2009

Human HES1 GAGAAAAGACGAAGAGCA TGTGCTCAGCGCAGCCGT

Human HEYL TCCCCACTGCCTTTGAG GGCACTCTTCCCAGGAT Leong et al., 2007

Human BIRC5 GCCCAGTGTTTCTTCTGCTT CCGGACGAATGCTTTTTATG Cheung et al., 2009

Human CTGF GTCCGCGTCGCCTTCGTGGTC GAGCACCATCTTTGGCGGTGCAC

Human SLUG AGATGCATATTCGGACCCAC CCTCATGTTTGTGCAGGAGA Leong et al., 2007

Human ABCG2 TTTCCAAGCGTTCATTCAAAAA TACGACTGTGACAATGATCTGAGC

Human GLI1 CCCCAGGGGCTGAGTCCTCC TCCAGAGCTGCCCCGCTGAT

Human PTCH CTGAGCAACACTCTGATGAA CAGTTAATGACTCCCAAGCA

Human BMI-1 GTCCAAGTTCACAAGACCAGACC ACAGTCATTGCTGCTGGGCATCG Liu et al., 2006

Human VIM-1 GAGAACTTTGCCGTTGAAGC GCTTCCTGTAGGTGGCAATC Casas et al., 2011

Human HMGA2 AGCAGAAGCCACTGGAGAAA TCTTCGGCAGACTCTTGTGA

Human KLF4 ATTACGCGGGCTGCGGCAAAA TTTTTGGCACTGGAACGGGCGG

Human DKK GGGAATTACTGCAAAAATGGAATA ATGACCGGAGACAAACAGAAC Butler., 2010

Human CDH1 TGCCCAGAAAATGAAAAAGG GTGTATGTGGCAATGCGTTC Casas et al., 2011

Human Pin1 CTGGAGCTGATCAACGGCTACATCC GCAGCGCAAACGAGGCGTCT

Human Fbxw7 GTGGACCTGCCCGTTCACCAACTCT CGGACCTCAGAACCATGGTCCAACT

Human GAPDH GAGTCAACGGATTTGGTCGT GACAAGCTTCCCGTTCTCAG

Human H3 GTGAAGAAACCTCATCGTTACAGGC

CTGGT

CTGCAAAGCACCAATAGCTGCACTC

TGGAA

Material and Methods

819, Santa Cruz), for immunoprecipitation and Western blot of endogenous Fbxw7α mouse monoclonal

(Abnova MO2, 3D1) and rabbit polyclonal (Abcam ab12292) antibodies were used, respectively. For

tagged proteins rabbit polyclonal and mouse monoclonal (9B11) anti-myc (#2272 and #2276 respectively,

Cell Signaling), mouse monoclonal anti-Flag clone M2 (F3165, Sigma), anti-GFP rabbit polyclonal

serum was raised against GST-GFP fusion protein expressed in bacteria, affinity purified and used

1:1000, mouse monoclonal (12CA5) and rabbit polyclonal (Y-11, sc-805 Santa Cruz) anti-HA. For

immunohistochemical stainings anti-cleaved Notch1 (Abcam 8925) anti-Notch1 (Santa-Cruz C-20), anti-

Notch4 (Santa Cruz H-225), anti-Fbxw7α (Abnova MO2, 3D1), and anti-Pin1 home made anti-rabbit

(Rustighi et al. 2009) were used.

Western blot, in vitro binding, immunoprecipitation.

In vitro binding assays, immuno- and co-immunoprecipitations and Western blot analyses were

performed by standard procedures, as described (Rustighi et al., 2009). Briefly, for GST and GST-Pin1

pull-down analysis using cells were lysed in GST pull-down buffer (150 mM NaCl, 50 mM Tris/HCl pH

7.5, 10% glycerol, 0.1% Nonidet P-40, Sigma), supplemented with inhibitors of phosphatase (1mM

sodium orthovanadate, 5mM NaF, Sigma) and protease (phenylmethylsulfonylfluoride (PMSF) 1mM and

chymostatin, leupeptin, antipain, pepstatin 10 µg ml-1 each, Sigma). For Notch1 immunoprecipitations

cells were treated in all cases with proteasome inhibitor lactacystin for 8 hrs and collected in GST pull-

down buffer as above. Cell lysates were cleared with proteinA Sepharose by rocking for 30 min, then

Protein A/G sepharose (GE Healthcare) cross-linked antibodies, precleared with 10 mg/ml BSA (Sigma),

were added. Binding reactions were left for a minimum of 4 hours to over night rocking at 4°C. Then

beads were washed and bound proteins were loaded and separated in SDS-PAGE, followed by Western

blotting on Nitrocellulose membranes (Scleicher & Schuell). For β-mercaptoethanol stripping of the

primary antibody membranes were shortly boiled in 0.1 mM β-mercaptoethanol, 1% SDS, 50 mM

Tris/HCl pH 6.9, then washed by shaking at RT with PBS, followed by blocking in Blotto-tween (PBS,

0.2% Tween-20, not fat dry milk 5%) or with TBST (Tris/HCl 25 mM pH7.5) plus 5% BSA (Sigma)

depending on the antibody.

Purified GST-Pin1 protein for Far Western analysis was obtained by immobilization, after production in

bacteria, on glutathione sepharose 4B beads (GEhealthcare) followed by elutions using reduced GSH as a

competitor in Tris/HCl pH8 100mM and NaCl 100mM. The eluted protein was subsequently purified by

dialysis.

Immunohistochemical analyses.

Mammary tissue from Pin1+/+ and Pin1-/- mice was collected, formalin fixed and embedded in paraffin. 3

µm sections were stained with anti-Notch1 (Santa Cruz C-20, 1:50), anti-Notch4 (Santa Cruz H-225,

1:100), anti-Fbxw7α (Abnova MO2, 3D1, 1:250), and anti-Pin1 (home made anti-rabbit, 1:200, Rustighi

et al. 2009) antibodies. For immunohistochemical analysis of human breast cancers, anti-cleaved Notch1

Material and Methods

(Abcam 8925, 1:200), anti-Pin1 (Santa Cruz mouse monoclonal G-8: sc-46660) and the abovementioned

anti-Fbxw7α antibody were used. Briefly, stainings were performed according to standard procedures for

paraffin embedded tissue. Slides were deparaffinized, rehydrated and antigen retrieved in a calibrated

steam pressure cooker with citrate buffer (pH 6.0). Endogenous peroxidase was blocked with peroxidase

block from an EnVision™ Kit (DakoCytomation, Glostrup, Denmark) for 15 minutes. The slides were

blocked in 5% nonfat dry milk or PBS + 5% BSA for 20 minutes to 1 hour at room temperature to

minimize nonspecific binding due to hydrophobic interaction. The slides were then incubated in blocking

buffer without any antibody (negative control) or with the indicated primary antibodies for 1 hour at 37°C

or overnight at 4°C. After washing, slides were incubated with secondary universal antibody (Vectastain)

for 45 minutes at room temperature. Colorimetric detection was completed with diaminobenzidine and

hydrogen peroxidase for 6 minutes and counterstained with hematoxylin (Sigma-Aldrich, St. Louis, MO,

USA). The immunostaining was scored semiquantitatively. In the evaluation of activated/cleaved Notch1

and Fbxw7α only nuclear staining was considered. The scores for IHC were: 0, no staining; 1, few

nuclear staining of tumor cells (<5% of positive tumor cells); 2, 5-10% of positive tumor cells; 3, >10%

positive tumor cells. For pairwise comparisons, the scores were collapsed to low (score, 1–2) versus high

(score, 3) expression, excluding not interpretable samples.

Data acquisition, Image processing, Equipment and settings.

Western blot films were scanned with an Epson Stylus DX7450 scanner at a grey scale 600 d.p.i.

resolution and saved as Tiff files. Single panels were cropped to obtain individual layers by Photpshop

software. Densitometric values of protein levels in Western blot analyses were obtained by Image J

software. Where necessary, for the sake of clarity, parts of the same or different gels with equal molecular

weights were cropped and juxtaposed, demarcated with borders and indicated in the figure legend. Images

of mouse and human immunohistochemical analyses were obtained using Leica DM4000B Microscope

with DFC420C photocamera. Pictures of Lymph node metastases were taken with an Olympus Super

bright Zoom Lens F1.8, C-4040200M, 4.1 Megapixel, 7.5 digital zoom. Images of haematoxylin and

eosin stained sections of pulmonary metastases were obtained with an Olympus BX40 microscope and

Leica DFC295 CH-9435 camera and were used for Computer-aided assessment of percentage of lung

tissue area occupied by metastases. Microscope image files were obtained through Leica Application

Suite LAS 4.1 software.

Breast cancer data collection and processing

We collected 21 datasets comprising microarray data of breast cancer samples and annotations on

patients’ clinical outcome. All data were measured on Affymetrix arrays and have been downloaded from

Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo) and ArrayExpress

(http://www.ebi.ac.uk/arrayexpress/). The complete list of datasets is provided in Table S1. Prior to

analysis, we reorganized the datasets eliminating duplicate samples and samples without outcome

Material and Methods

information and renamed any original study after the medical center where patients were recruited.

Briefly, the original studies have been modified as follows:

-Stockholm has been used as it is and re-named as KI_Stockholm (Karolinska Institutet Stockholm); -

EMC-286 and EMC-58 have been merged into EMC-344 (Erasmus Medical Center); -MSK has been

used as is and re-named as MSKCC (Memorial Sloan-Kettering Cancer Center); -Uppsala-Miller,

Ivshina-Miller, and Loi datasets (GSE3494, GSE4922, and GSE6532) included samples derived from

Uppsala University Hospital, John Radcliffe Hospital in Oxford, and Guys Hospital in London.

Moreover, a comparison of the hybridization dates on the raw files and of the patients’ clinical

information revealed that, although deposited twice, GSE3494 and GSE4922 were identical. As such, the

3 series have been split into KI_Uppsala, comprising all 253 unique patients of the Uppsala University

Hospital, OXF composed of the 178 samples collected at the John Radcliffe Hospital in Oxford, and

GUY composed of the 87 samples from the Guys Hospital in London;

-Sotiriou is entirely included in GSE6532;

-Desmedt has been used as is and renamed as TRANSBIG (after the consortium of cancer centers where

samples have been collected);

-Schmidt has been used as is and re-named as Mainz (Johannes Gutenberg University in Mainz);

-Veridex has been used as is and re-named as Veridex_MultiCenter (after the three European and one US

institutions where samples have been collected);

-Tamoxifen has been added to GUY;

-Chin and Zhou have been merged into UCSF (University of California, San Francisco) after removing

samples deposited in both GEO (GSE7378) and ArrayExpress (E-TABM-158);

-TOP TRIAL has been re-named as IJB_TOP (Institut Jules Bordet /Trial of Principle) after removal of

13 samples lacking of outcome information;

-GSE19615 has been used as is and re-named as US_NCI (US National Cancer Institute);

-IPC has been re-named as CRCM (Centre de cancérologie de Marseille) after removal of 14 samples

lacking of outcome information;

-KFSYSCC has been used as is and re-named as KOOF (Koo Foundation SYS Cancer Center);

-GSE31519 has been used as is and re-named as Goethe (Goethe-University, Frankfurt) after removal of

2 samples lacking of outcome information;

-Hatzis included samples derived from 4 cohorts, i.e., I-SPY-1 (Investigation of Serial Studies to Predict

Your Therapeutic Response With Imaging and Molecular Analysis), LBJ_INEN_GEICAM (Lyndon B.

Johnson Hospital, Instituto Nacional de Enfermedades Neoplásicas, and Grupo Español de Investigación

en Cáncer de Mama), USO-02103 (US Oncology), and MDACC (M. D. Anderson Cancer Center,

Houston). The orginal data has been split into I-SPY-1 comprising 83 samples, LBJ_INEN_GEICAM

comprising 58 samples, MDACC comprising 313 samples, and USO-02103 comprising 54 samples.

This re-organization resulted in a meta-dataset comprising 3254 unique samples from 19 independent

cohorts (Table S1). According to Cordenonsi et al. (2011), we standardized clinical information among

Material and Methods

the various datasets redefining the outcome descriptions based on the clinical annotations of each

individual study. Specifically, we defined survival as death because of cancer and includes overall

survival, disease free survival, and disease specific survival. Raw expression data (i.e., CEL files)

obtained from different platforms have been integrated using an approach inspired by geometry and probe

content of HG-U133 Affymetrix arrays (Fallarino et al., 2010). Briefly, probes with the same

oligonucleotide sequence, but located at different coordinates on different type of arrays, may be arranged

in a virtual platform grid. As for any other microarray geometry, this virtual grid may be used as a

reference to create a virtual Chip Definition File (virtual-CDF), containing the probes shared among the

various HG-U133 platforms and their coordinates on the virtual platform, and a virtual-CEL files

containing the fluorescence intensities of the original CEL files properly re-mapped on the virtual grid.

Once defined the virtual platform through the creation of the virtual-CDF and transformed the CEL files

into virtual-CELs, raw data, originally obtained from different HG-U133 arrays, are homogeneous in

terms of platform and can be preprocessed and normalized adopting standard approaches, as RMA

(Irizarry et al., 2003). Specifically, expression values were generated from intensity signals using the

virtual-CDF, obtained merging HG-U133A, HG-U133AAofAV2, and HG-U133 Plus2 original CDFs,

and the transformed virtual-CEL files. Intensity values for a total of 21981 meta-probe sets have been

background adjusted, normalized using quantile normalization, and gene expression levels calculated

using median polish summarization (RMA). The entire procedure has been implemented as an R script.

To identify two groups of tumor samples with either high or low levels of the Notch-dependent Direct

Target (NDT) gene signature we used the classifier described in Adorno et al (2009). Briefly, we defined

a classification rule based on summarizing the standardized expression levels of each gene in NDT

signature into a combined score with zero mean. Tumors were then classified as NDT signature Low if

the combined score was negative and as NDT signature High if the combined score was positive.

Similarly, we defined tumors as expressing high or low levels of Fbxw7 and Pin1 mRNA if the

standardized expression signal of Fbxw7 and Pin1 probes was positive or negative, respectively. This

classification was applied to log2 expression values obtained using RMA on the meta-dataset described

above.

Survival analysis

To evaluate the prognostic value of the NDT signatures, we estimated, using the Kaplan-Meier method

(Kalbfleisch and Prentice), the probabilities that patients would remain free of death (survival). To

confirm these findings, the Kaplan-Meier curves were compared using the log-rank or Mantel-Haenszel

test (Harrington and Fleming). P-values were calculated according to the standard normal asymptotic

distribution. When comparing “NDT signature High” and “NDT signature Low” groups, the group with

low NDT levels displayed a significantly higher probability (at a significance level α=5x10-2) of a

reduced survival (Figure S8B). Survival analysis and Kaplan-Meier plots were obtained using R

Material and Methods

survcomp package. Kaplan-Meyer curves have been compared using the log-rank test of the surv_test

\function (coin R package).

Breast cancer re-organized cohorts comprised in the meta-dataset analyzed in this study.

Cohort Affymetrix platform Samples Data source References

KI_Stockholm HG-U133 A 159 GSE1456 Pawitan et al., 2005

EMC-344 HG-U133A 344 GSE2034; GSE5327 Wang et al., 2005; Minn et al., 2007

MSKCC HG-U133A 82 GSE2603 Minn et al., 2005

KI_Uppsala HG-U133A 253 GSE3494; GSE4922; GSE6532

Loi et al, 2008; Ivshina et al, 2006; Miller et al, 2005

OXF HG-U133A 178 GSE6532 Ivshina et al., 2006 TransBIG HG-U133A 198 GSE7390 Desmedt et al., 2007 Mainz HG-U133A 200 GSE11121 Schmidt et al., 2008 Veridex_MultiCenter HG-U133A 136 GSE12093 Zhang et al., 2009

GUY HG-U133 Plus 2.0 164 GSE6532; GSE9195 Loi et al., 2007; Loi et al., 2008; Loi et al., 2010

UCSF HG-U133AAofAV2 166 E-TABM-158; GSE7378

Merritt et al., 2008; Zhou T et al., 2007; Yau C et al., 2008

IJB_TOP HG-U133 Plus 2.0 107 GSE16446

Desmedt Cet al., 2011; Li Y et al., 2010; Juul N et al., 2010

US_NCI HG-U133 Plus 2.0 115 GSE19615 Li Y t al., 2010

CRCM HG-U133 Plus 2.0 252 GSE21653 Sabatier R et al., 2011

KOOF HG-U133 Plus 2.0 327 GSE20685 Kao KJ et al., 2011

Goethe HG-U133A 65 GSE31519 Rody A. et al., 2011; Karn T. et al., 2011; Karn T. et al., 2012

MDACC HG-U133A 313 GSE25066 Hatzis C. et al., 2012 I-SPY-1 HG-U133A 83 GSE25066 Hatzis C. et al., 2012 LBJ_INEN_GEICAM HG-U133A 58 GSE25066 Hatzis C. et al., 2012

USO-02103 HG-U133A 54 GSE25066 Hatzis C. et al., 2012

Material and Methods

Notch-dependent Direct Target (NDT) gene signature.

Gene name ENTREZ ID Reference

HES1 3280 Grabher et al., 2006

HEY1 23462 Grabher et al., 2006

HEY2 23493 Grabher et al., 2006

HEYL 26508 Grabher et al., 2006

MYC 4609 Palomero et al., 2006; Weng et al., 2006

BUB1B 701 Palomero et al., 2006

BUB3 9184 Palomero et al., 2006

CDC25A 993 Palomero et al., 2006

PHB 5245 Palomero et al., 2006

RBL1 5933 Palomero et al., 2006

RPL3 6122 Palomero et al., 2006

USP5 8078 Palomero et al., 2006

PLAU 5328 Shimizu et al., 2011

SHQ 55164 Chadwick et al., 2009

CCND1 595 Ronchini and Capobianco, 2001

GATA3 2625 Amsen et al., 2007

SKP2 6502 Sarmento et al., 2005

ERBB2 2064 Chen et al., 1997

CDKN1A 1026 Rangarajan et al., 2001

SNAI1 6615 Sahlgren et al., 2008

SNAI2 6591 Leong et al., 2007

NFKB2 4791 Oswald et al., 1998

BIRC5 332 Lee et al., 2008

NOTCH1 4851 Weng et al., 2006; Hamidi et al., 2011

NOTCH3 4854 Weng et al., 2006; Hamidi et al., 2011

NOTCH4 4855 Hamidi et al., 2011

IFRD2 7866 Palomero et al., 2006

ING3 54556 Palomero et al., 2006

PTCRA 171558 Grabher et al., 2006

CD3D 915 Palomero et al., 2006

Bibliography

8. BIBLIOGRAPHY

Adorno, M., Cordenonsi, M., Montagner, M., Dupont, S., Wong, C., Hann, B., Solari, A., Bobisse, S., Rondina, M. B., Guzzardo, V., et al. (2009). A Mutant-p53/Smad complex opposes p63 to empower TGFbeta-induced metastasis. Cell 137, 87-98. Al-Hajj, M., Wicha, M. S., Benito-Hernandez, A., Morrison, S. J., and Clarke, M. F. (2003). Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad Sci U S A 100, 3983-3988. Andersen, P., Uosaki, H., Shenje, L. T., and Kwon, C. (2012). Non-canonical Notch signaling: emerging role and mechanism. Trends Cell Biol 22, 257-265. Andersson, E. R., Sandberg, R., and Lendahl, U. (2011). Notch signaling: simplicity in design, versatility in function. Development 138, 3593-3612. Androutsellis-Theotokis, A., Leker, R. R., Soldner, F., Hoeppner, D. J., Ravin, R., Poser, S. W., Rueger, M. A., Bae, S. K., Kittappa, R., and McKay, R. D. (2006). Notch signalling regulates stem cell numbers in vitro and in vivo. Nature 442, 823-826. Artavanis-Tsakonas, S., Rand, M. D., and Lake, R. J. (1999). Notch signaling: cell fate control and signal integration in development. Science 284, 770-776. Atchison, F. W., and Means, A. R. (2003). Spermatogonial depletion in adult Pin1-deficient mice. Biol Reprod 69, 1989-1997. Atkinson, G. P., Nozell, S. E., Harrison, D. K., Stonecypher, M. S., Chen, D., and Benveniste, E. N. (2009). The prolyl isomerase Pin1 regulates the NF-kappaB signaling pathway and interleukin-8 expression in glioblastoma. Oncogene 28, 3735-3745. Ayala, G., Wang, D., Wulf, G., Frolov, A., Li, R., Sowadski, J., Wheeler, T. M., Lu, K. P., and Bao, L. (2003). The prolyl isomerase Pin1 is a novel prognostic marker in human prostate cancer. Cancer Res 63, 6244-6251. Bao, L., Kimzey, A., Sauter, G., Sowadski, J. M., Lu, K. P., and Wang, D. G. (2004). Prevalent overexpression of prolyl isomerase Pin1 in human cancers. Am J Pathol 164, 1727-1737. Ben-Porath, I., Thomson, M. W., Carey, V. J., Ge, R., Bell, G. W., Regev, A., and Weinberg, R. A. (2008). An embryonic stem cell-like gene expression signature in poorly differentiated aggressive human tumors. Nat Genet 40, 499-507. Beres, B. J., George, R., Lougher, E. J., Barton, M., Verrelli, B. C., McGlade, C. J., Rawls, J. A., and Wilson-Rawls, J. (2011). Numb regulates Notch1, but not Notch3, during myogenesis. Mech Dev 128, 247-257. Bernis, C., Vigneron, S., Burgess, A., Labbe, J. C., Fesquet, D., Castro, A., and Lorca, T. (2007). Pin1 stabilizes Emi1 during G2 phase by preventing its association with SCF(betatrcp). EMBO Rep 8, 91-98. Beverly, L. J., Felsher, D. W., and Capobianco, A. J. (2005). Suppression of p53 by Notch in lymphomagenesis: implications for initiation and regression. Cancer Res 65, 7159-7168. Bhattacharya, S., Das, A., Mallya, K., and Ahmad, I. (2007). Maintenance of retinal stem cells by Abcg2 is regulated by notch signaling. J Cell Sci 120, 2652-2662. Blaumueller, C.M., Qi, H., Zagouras, P. & Artavanis-Tsakonas, S. Intracellular cleavage of Notch leads to a heterodimeric receptor on the plasma membrane. Cell 90, 281-291 (1997). Bonizzi, G., Cicalese, A., Insinga, A., and Pelicci, P. G. (2012). The emerging role of p53 in stem cells. Trends Mol Med 18, 6-12. Bouras, T., Pal, B., Vaillant, F., Harburg, G., Asselin-Labat, M. L., Oakes, S. R., Lindeman, G. J., and Visvader, J. E. (2008). Notch signaling regulates mammary stem cell function and luminal cell-fate commitment. Cell Stem Cell 3, 429-441. Bray, F., Ren, J. S., Masuyer, E., and Ferlay, J. (2013). Global estimates of cancer prevalence for 27 sites in the adult population in 2008. Int J Cancer 132, 1133-1145.

Bibliography

Bray, S.J. Notch signalling: a simple pathway becomes complex. Nat Rev Mol Cell Biol 7, 678-689 (2006). Brill, L. M., Xiong, W., Lee, K. B., Ficarro, S. B., Crain, A., Xu, Y., Terskikh, A., Snyder, E. Y., and Ding, S. (2009). Phosphoproteomic analysis of human embryonic stem cells. Cell Stem Cell 5, 204-213. Buckley, S. M., Aranda-Orgilles, B., Strikoudis, A., Apostolou, E., Loizou, E., Moran-Crusio, K., Farnsworth, C. L., Koller, A. A., Dasgupta, R., Silva, J. C., et al. (2012). Regulation of pluripotency and cellular reprogramming by the ubiquitin-proteasome system. Cell Stem Cell 11, 783-798. Byrd, K. N., Huey, B., Roydasgupta, R., Fridlyand, J., Snijders, A. M., and Albertson, D. G. (2008). FBXW7 and DNA copy number instability. Breast Cancer Res Treat 109, 47-54. Chang, C. J., Chao, C. H., Xia, W., Yang, J. Y., Xiong, Y., Li, C. W., Yu, W. H., Rehman, S. K., Hsu, J. L., Lee, H. H., et al. (2011). p53 regulates epithelial-mesenchymal transition and stem cell properties through modulating miRNAs. Nat Cell Biol 13, 317-323. Chapman, G., Liu, L., Sahlgren, C., Dahlqvist, C., and Lendahl, U. (2006). High levels of Notch signaling down-regulate Numb and Numblike. J Cell Biol 175, 535-540. Chen, C. H., Chang, C. C., Lee, T. H., Luo, M., Huang, P., Liao, P. H., Wei, S., Li, F. A., Chen, R. H., Zhou, X. Z., et al. (2013). SENP1 deSUMOylates and regulates Pin1 protein activity and cellular function. Cancer Res 73, 3951-3962. Chen, L., and Daley, G. Q. (2008). Molecular basis of pluripotency. Hum Mol Genet 17, R23-27. Cheng, Y., and Li, G. (2012). Role of the ubiquitin ligase Fbw7 in cancer progression. Cancer Metastasis Rev 31, 75-87. Cho, Y. S., Park, S. Y., Kim, D. J., Lee, S. H., Woo, K. M., Lee, K. A., Lee, Y. J., Cho, Y. Y., and Shim, J. H. (2012). TPA-induced cell transformation provokes a complex formation between Pin1 and 90 kDa ribosomal protein S6 kinase 2. Mol Cell Biochem 367, 85-92. Ciarapica, R., Methot, L., Tang, Y., Lo, R., Dali, R., Buscarlet, M., Locatelli, F., del Sal, G., Rota, R., and Stifani, S. (2014). Prolyl isomerase Pin1 and protein kinase HIPK2 cooperate to promote cortical neurogenesis by suppressing Groucho/TLE:Hes1-mediated inhibition of neuronal differentiation. Cell Death Differ 21, 321-332. Cicalese, A., Bonizzi, G., Pasi, C. E., Faretta, M., Ronzoni, S., Giulini, B., Brisken, C., Minucci, S., Di Fiore, P. P., and Pelicci, P. G. (2009). The tumor suppressor p53 regulates polarity of self-renewing divisions in mammary stem cells. Cell 138, 1083-1095. Colaluca, I. N., Tosoni, D., Nuciforo, P., Senic-Matuglia, F., Galimberti, V., Viale, G., Pece, S., and Di Fiore, P. P. (2008). NUMB controls p53 tumour suppressor activity. Nature 451, 76-80. Conlon, R.A., Reaume, A.G. & Rossant, J. Notch1 is required for the coordinate segmentation of somites. Development 121, 1533-1545 (1995). Cordenonsi, M., Zanconato, F., Azzolin, L., Forcato, M., Rosato, A., Frasson, C., Inui, M., Montagner, M., Parenti, A. R., Poletti, A., et al. (2011). The Hippo transducer TAZ confers cancer stem cell-related traits on breast cancer cells. Cell 147, 759-772. Crenshaw, D. G., Yang, J., Means, A. R., and Kornbluth, S. (1998). The mitotic peptidyl-prolyl isomerase, Pin1, interacts with Cdc25 and Plx1. EMBO J 17, 1315-1327. Dean, M. (2009). ABC transporters, drug resistance, and cancer stem cells. J Mammary Gland Biol Neoplasia 14, 3-9. Dean, M., Fojo, T., and Bates, S. (2005). Tumour stem cells and drug resistance. Nat Rev Cancer 5, 275-284. Desantis, C., Ma, J., Bryan, L., and Jemal, A. (2014). Breast cancer statistics, 2013. CA Cancer J Clin 64, 52-62. Di Marcotullio, L., Ferretti, E., Greco, A., De Smaele, E., Po, A., Sico, M. A., Alimandi, M., Giannini, G., Maroder, M., Screpanti, I., and Gulino, A. (2006). Numb is a suppressor of Hedgehog signalling and targets Gli1 for Itch-dependent ubiquitination. Nat Cell Biol 8, 1415-1423. Dick, J. E. (1996). Normal and leukemic human stem cells assayed in SCID mice. Semin Immunol 8, 197-206.

Bibliography

Dievart, A., Beaulieu, N., and Jolicoeur, P. (1999). Involvement of Notch1 in the development of mouse mammary tumors. Oncogene 18, 5973-5981. Ding, Q., Huo, L., Yang, J. Y., Xia, W., Wei, Y., Liao, Y., Chang, C. J., Yang, Y., Lai, C. C., Lee, D. F., et al. (2008). Down-regulation of myeloid cell leukemia-1 through inhibiting Erk/Pin 1 pathway by sorafenib facilitates chemosensitization in breast cancer. Cancer Res 68, 6109-6117. Dong, P., Karaayvaz, M., Jia, N., Kaneuchi, M., Hamada, J., Watari, H., Sudo, S., Ju, J., and Sakuragi, N. (2013). Mutant p53 gain-of-function induces epithelial-mesenchymal transition through modulation of the miR-130b-ZEB1 axis. Oncogene 32, 3286-3295. Dontu, G., Abdallah, W. M., Foley, J. M., Jackson, K. W., Clarke, M. F., Kawamura, M. J., and Wicha, M. S. (2003). In vitro propagation and transcriptional profiling of human mammary stem/progenitor cells. Genes Dev 17, 1253-1270. Dontu, G., Jackson, K. W., McNicholas, E., Kawamura, M. J., Abdallah, W. M., and Wicha, M. S. (2004). Role of Notch signaling in cell-fate determination of human mammary stem/progenitor cells. Breast Cancer Res 6, R605-615. Dontu, G., Al-Hajj, M., Abdallah, W. M., Clarke, M. F., and Wicha, M. S. (2003b). Stem cells in normal breast development and breast cancer. Cell Prolif 36 Suppl 1, 59-72. Dontu, G., and Wicha, M. S. (2005). Survival of mammary stem cells in suspension culture: implications for stem cell biology and neoplasia. J Mammary Gland Biol Neoplasia 10, 75-86. Dougherty, M. K., Muller, J., Ritt, D. A., Zhou, M., Zhou, X. Z., Copeland, T. D., Conrads, T. P., Veenstra, T. D., Lu, K. P., and Morrison, D. K. (2005). Regulation of Raf-1 by direct feedback phosphorylation. Mol Cell 17, 215-224. Dotto, G. P. (2008). Notch tumor suppressor function. Oncogene 27, 5115-5123. Dotto, G. P. (2009). Crosstalk of Notch with p53 and p63 in cancer growth control. Nat Rev Cancer 9, 587-595. Eroles, P., Bosch, A., Perez-Fidalgo, J. A., and Lluch, A. (2012). Molecular biology in breast cancer: intrinsic subtypes and signaling pathways. Cancer Treat Rev 38, 698-707. Espinoza, I., Pochampally, R., Xing, F., Watabe, K., and Miele, L. (2013). Notch signaling: targeting cancer stem cells and epithelial-to-mesenchymal transition. Onco Targets Ther 6, 1249-1259. Farnie, G., Clarke, R. B., Spence, K., Pinnock, N., Brennan, K., Anderson, N. G., and Bundred, N. J. (2007). Novel cell culture technique for primary ductal carcinoma in situ: role of Notch and epidermal growth factor receptor signaling pathways. J Natl Cancer Inst 99, 616-627. Ferrando, A. A. (2009). The role of NOTCH1 signaling in T-ALL. Hematology Am Soc Hematol Educ Program, 353-361. Fryer, C.J., Lamar, E., Turbachova, I., Kintner, C. & Jones, K.A. Mastermind mediates chromatin-specific transcription and turnover of the Notch enhancer complex. Genes Dev 16, 1397-1411 (2002). Fujimori, F., Takahashi, K., Uchida, C., and Uchida, T. (1999). Mice lacking Pin1 develop normally, but are defective in entering cell cycle from G(0) arrest. Biochem Biophys Res Commun 265, 658-663. Fukuchi, M., Fukai, Y., Kimura, H., Sohda, M., Miyazaki, T., Nakajima, M., Masuda, N., Tsukada, K., Kato, H., and Kuwano, H. (2006). Prolyl isomerase Pin1 expression predicts prognosis in patients with esophageal squamous cell carcinoma and correlates with cyclinD1 expression. Int J Oncol 29, 329-334. Fukuda, S., and Pelus, L. M. (2006). Survivin, a cancer target with an emerging role in normal adult tissues. Mol Cancer Ther 5, 1087-1098. Galat, A. (2003). Peptidylprolyl cis/trans isomerases (immunophilins): biological diversity--targets--functions. Curr Top Med Chem 3, 1315-1347. Gallahan, D., and Callahan, R. (1987). Mammary tumorigenesis in feral mice: identification of a new int locus in mouse mammary tumor virus (Czech II)-induced mammary tumors. J Virol 61, 66-74.

Bibliography

Gallahan, D., Jhappan, C., Robinson, G., Hennighausen, L., Sharp, R., Kordon, E., Callahan, R., Merlino, G., and Smith, G. H. (1996). Expression of a truncated Int3 gene in developing secretory mammary epithelium specifically retards lobular differentiation resulting in tumorigenesis. Cancer Res 56, 1775-1785. Gerlinger, M., Rowan, A. J., Horswell, S., Larkin, J., Endesfelder, D., Gronroos, E., Martinez, P., Matthews, N., Stewart, A., Tarpey, P., et al. (2012). Intratumor heterogeneity and branched evolution revealed by multiregion sequencing. N Engl J Med 366, 883-892. Ginestier, C., Hur, M. H., Charafe-Jauffret, E., Monville, F., Dutcher, J., Brown, M., Jacquemier, J., Viens, P., Kleer, C. G., Liu, S., et al. (2007). ALDH1 is a marker of normal and malignant human mammary stem cells and a predictor of poor clinical outcome. Cell Stem Cell 1, 555-567. Ginestier, C., Wicinski, J., Cervera, N., Monville, F., Finetti, P., Bertucci, F., Wicha, M. S., Birnbaum, D., and Charafe-Jauffret, E. (2009). Retinoid signaling regulates breast cancer stem cell differentiation. Cell Cycle 8, 3297-3302. Girardini, J. E., Napoli, M., Piazza, S., Rustighi, A., Marotta, C., Radaelli, E., Capaci, V., Jordan, L., Quinlan, P., Thompson, A., et al. (2011). A Pin1/mutant p53 axis promotes aggressiveness in breast cancer. Cancer Cell 20, 79-91. Gjorevski, N., and Nelson, C. M. (2011). Integrated morphodynamic signalling of the mammary gland. Nat Rev Mol Cell Biol 12, 581-593. Goardon, N., Marchi, E., Atzberger, A., Quek, L., Schuh, A., Soneji, S., Woll, P., Mead, A., Alford, K. A., Rout, R., et al. (2011). Coexistence of LMPP-like and GMP-like leukemia stem cells in acute myeloid leukemia. Cancer Cell 19, 138-152. Goodell, M. A., Brose, K., Paradis, G., Conner, A. S., and Mulligan, R. C. (1996). Isolation and functional properties of murine hematopoietic stem cells that are replicating in vivo. J Exp Med 183, 1797-1806. Grabher, C., von Boehmer, H., and Look, A. T. (2006). Notch 1 activation in the molecular pathogenesis of T-cell acute lymphoblastic leukaemia. Nat Rev Cancer 6, 347-359. Grego-Bessa, J., Diez, J., Timmerman, L. & de la Pompa, J.L. Notch and epithelial-mesenchyme transition in development and tumor progression: another turn of the screw. Cell Cycle 3, 718-721 (2004). Guo, W., Keckesova, Z., Donaher, J. L., Shibue, T., Tischler, V., Reinhardt, F., Itzkovitz, S., Noske, A., Zurrer-Hardi, U., Bell, G., et al. (2012). Slug and Sox9 cooperatively determine the mammary stem cell state. Cell 148, 1015-1028. Gudjonsson, T., Villadsen, R., Nielsen, H. L., Ronnov-Jessen, L., Bissell, M. J., and Petersen, O. W. (2002). Isolation, immortalization, and characterization of a human breast epithelial cell line with stem cell properties. Genes Dev 16, 693-706. Gupta-Rossi, N., et al. Functional interaction between SEL-10, an F-box protein, and the nuclear form of activated Notch1 receptor. J Biol Chem 276, 34371-34378 (2001). Gutierrez, G. J., and Ronai, Z. (2006). Ubiquitin and SUMO systems in the regulation of mitotic checkpoints. Trends Biochem Sci 31, 324-332. Han, C. H., Lu, J., Wei, Q., Bondy, M. L., Brewster, A. M., Yu, T. K., Buchholz, T. A., Arun, B. K., and Wang, L. E. (2010). The functional promoter polymorphism (-842G>C) in the PIN1 gene is associated with decreased risk of breast cancer in non-Hispanic white women 55 years and younger. Breast Cancer Res Treat 122, 243-249. Han, J., Hendzel, M. J., and Allalunis-Turner, J. (2011). Notch signaling as a therapeutic target for breast cancer treatment? Breast Cancer Res 13, 210. Hanes, S. D., Shank, P. R., and Bostian, K. A. (1989). Sequence and mutational analysis of ESS1, a gene essential for growth in Saccharomyces cerevisiae. Yeast 5, 55-72. Harrison, H., Farnie, G., Howell, S. J., Rock, R. E., Stylianou, S., Brennan, K. R., Bundred, N. J., and Clarke, R. B. (2010). Regulation of breast cancer stem cell activity by signaling through the Notch4 receptor. Cancer Res 70, 709-718. Hoe, K. K., Verma, C. S., and Lane, D. P. (2014). Drugging the p53 pathway: understanding the route to clinical efficacy. Nat Rev Drug Discov 13, 217-236.

Bibliography

Holderfield, M. T., and Hughes, C. C. (2008). Crosstalk between vascular endothelial growth factor, notch, and transforming growth factor-beta in vascular morphogenesis. Circ Res 102, 637-652. Hoshino, K. (1964). Regeneration and Growth of Quantitatively Transplanted Mammary Glands of Normal Female Mice. Anat Rec 150, 221-235. Hsu, T., McRackan, D., Vincent, T. S., and Gert de Couet, H. (2001). Drosophila Pin1 prolyl isomerase Dodo is a MAP kinase signal responder during oogenesis. Nat Cell Biol 3, 538-543. Hu, L., McArthur, C., and Jaffe, R. B. (2010). Ovarian cancer stem-like side-population cells are tumourigenic and chemoresistant. Br J Cancer 102, 1276-1283. Hunter, T. (1998). Prolyl isomerases and nuclear function. Cell 92, 141-143. Ibusuki, M., Yamamoto, Y., Shinriki, S., Ando, Y., and Iwase, H. (2011). Reduced expression of ubiquitin ligase FBXW7 mRNA is associated with poor prognosis in breast cancer patients. Cancer Sci 102, 439-445. Insinga, A., Cicalese, A., Faretta, M., Gallo, B., Albano, L., Ronzoni, S., Furia, L., Viale, A., and Pelicci, P. G. (2013). DNA damage in stem cells activates p21, inhibits p53, and induces symmetric self-renewing divisions. Proc Natl Acad Sci U S A 110, 3931-3936. Jin, H., Jiang, J., Sun, L., Zheng, F., Wu, C., Peng, L., Zhao, Y., and Wu, X. (2011). The prolyl isomerase Pin1 is overexpressed in human esophageal cancer. Oncol Lett 2, 1191-1196. Karaczyn, A., Bani-Yaghoub, M., Tremblay, R., Kubu, C., Cowling, R., Adams, T. L., Prudovsky, I., Spicer, D., Friesel, R., Vary, C., and Verdi, J. M. (2010). Two novel human NUMB isoforms provide a potential link between development and cancer. Neural Dev 5, 31. Keller, G. (2005). Embryonic stem cell differentiation: emergence of a new era in biology and medicine. Genes Dev 19, 1129-1155. Kelly, A.P., et al. Notch-induced T cell development requires phosphoinositide-dependent kinase 1. EMBO J 26, 3441-3450 (2007). Khoury, G. A., Baliban, R. C., and Floudas, C. A. (2011). Proteome-wide post-translational modification statistics: frequency analysis and curation of the swiss-prot database. Sci Rep 1. Kidd, S., Kelley, M. R., and Young, M. W. (1986). Sequence of the notch locus of Drosophila melanogaster: relationship of the encoded protein to mammalian clotting and growth factors. Mol Cell Biol 6, 3094-3108. Kim, M. R., Choi, H. S., Heo, T. H., Hwang, S. W., and Kang, K. W. (2008). Induction of vascular endothelial growth factor by peptidyl-prolyl isomerase Pin1 in breast cancer cells. Biochem Biophys Res Commun 369, 547-553. Kim, S. B., Chae, G. W., Lee, J., Park, J., Tak, H., Chung, J. H., Park, T. G., Ahn, J. K., and Joe, C. O. (2007). Activated Notch1 interacts with p53 to inhibit its phosphorylation and transactivation. Cell Death Differ 14, 982-991. Knoblich, J. A., Jan, L. Y., and Jan, Y. N. (1995). Asymmetric segregation of Numb and Prospero during cell division. Nature 377, 624-627. Kong, D., Li, Y., Wang, Z., and Sarkar, F. H. (2011). Cancer Stem Cells and Epithelial-to-Mesenchymal Transition (EMT)-Phenotypic Cells: Are They Cousins or Twins? Cancers (Basel) 3, 716-729. Korkaya, H., Paulson, A., Charafe-Jauffret, E., Ginestier, C., Brown, M., Dutcher, J., Clouthier, S. G., and Wicha, M. S. (2009). Regulation of mammary stem/progenitor cells by PTEN/Akt/beta-catenin signaling. PLoS Biol 7, e1000121. Kovall, R. A., and Blacklow, S. C. (2010). Mechanistic insights into Notch receptor signaling from structural and biochemical studies. Curr Top Dev Biol 92, 31-71. Krebs, L.T., et al. Characterization of Notch3-deficient mice: normal embryonic development and absence of genetic interactions with a Notch1 mutation. Genesis 37, 139-143 (2003) Kreso, A., and Dick, J. E. (2014). Evolution of the Cancer Stem Cell Model. Cell Stem Cell 14, 275-291.

Bibliography

Lattanzio, F., Carboni, L., Carretta, D., Rimondini, R., Candeletti, S., and Romualdi, P. (2014). Human apolipoprotein E4 modulates the expression of Pin1, Sirtuin 1, and Presenilin 1 in brain regions of targeted replacement apoE mice. Neuroscience 256, 360-369. Le Borgne, R., Bardin, A., and Schweisguth, F. (2005). The roles of receptor and ligand endocytosis in regulating Notch signaling. Development 132, 1751-1762. Lee, C. W., Simin, K., Liu, Q., Plescia, J., Guha, M., Khan, A., Hsieh, C. C., and Altieri, D. C. (2008). A functional Notch-survivin gene signature in basal breast cancer. Breast Cancer Res 10, R97. Lee, S. H., Choi, Y. H., Kim, Y. J., Choi, H. S., Yeo, C. Y., and Lee, K. Y. (2013a). Prolyl isomerase Pin1 enhances osteoblast differentiation through Runx2 regulation. FEBS Lett 587, 3640-3647. Lee, T. H., Chen, C. H., Suizu, F., Huang, P., Schiene-Fischer, C., Daum, S., Zhang, Y. J., Goate, A., Chen, R. H., Zhou, X. Z., and Lu, K. P. (2011). Death-associated protein kinase 1 phosphorylates Pin1 and inhibits its prolyl isomerase activity and cellular function. Mol Cell 42, 147-159. Lee, Y. M., Shin, S. Y., Jue, S. S., Kwon, I. K., Cho, E. H., Cho, E. S., Park, S. H., and Kim, E. C. (2013b). The Role of PIN1 on Odontogenic and Adipogenic Differentiation in Human Dental Pulp Stem Cells. Stem Cells Dev. Leong, K. G., Niessen, K., Kulic, I., Raouf, A., Eaves, C., Pollet, I., and Karsan, A. (2007). Jagged1-mediated Notch activation induces epithelial-to-mesenchymal transition through Slug-induced repression of E-cadherin. J Exp Med 204, 2935-2948. Levine, A. J., and Oren, M. (2009). The first 30 years of p53: growing ever more complex. Nat Rev Cancer 9, 749-758. Li, X., Gounari, F., Protopopov, A., Khazaie, K., and von Boehmer, H. (2008). Oncogenesis of T-ALL and nonmalignant consequences of overexpressing intracellular NOTCH1. J Exp Med 205, 2851-2861. Li, Y., Hibbs, M. A., Gard, A. L., Shylo, N. A., and Yun, K. (2012). Genome-wide analysis of N1ICD/RBPJ targets in vivo reveals direct transcriptional regulation of Wnt, SHH, and hippo pathway effectors by Notch1. Stem Cells 30, 741-752. Liou, Y. C., Ryo, A., Huang, H. K., Lu, P. J., Bronson, R., Fujimori, F., Uchida, T., Hunter, T., and Lu, K. P. (2002). Loss of Pin1 function in the mouse causes phenotypes resembling cyclin D1-null phenotypes. Proc Natl Acad Sci U S A 99, 1335-1340. Liou, Y. C., Zhou, X. Z., and Lu, K. P. (2011). Prolyl isomerase Pin1 as a molecular switch to determine the fate of phosphoproteins. Trends Biochem Sci 36, 501-514. Liu, S., Dontu, G., and Wicha, M. S. (2005). Mammary stem cells, self-renewal pathways, and carcinogenesis. Breast Cancer Res 7, 86-95. Liu, N., Zhang, J., and Ji, C. (2013). The emerging roles of Notch signaling in leukemia and stem cells. Biomark Res 1, 23. Longo, D., Fauci, A., Kasper, D., Hauser, S., Jameson, J., and Loscalzo, J. (2011). Harrison's Principles of Internal Medicine, 18th Edition: McGraw-Hill Education). Lu, J., Ye, X., Fan, F., Xia, L., Bhattacharya, R., Bellister, S., Tozzi, F., Sceusi, E., Zhou, Y., Tachibana, I., et al. (2013a). Endothelial cells promote the colorectal cancer stem cell phenotype through a soluble form of Jagged-1. Cancer Cell 23, 171-185. Lu, K. P., Hanes, S. D., and Hunter, T. (1996). A human peptidyl-prolyl isomerase essential for regulation of mitosis. Nature 380, 544-547. Lu, K. P., Liou, Y. C., and Zhou, X. Z. (2002a). Pinning down proline-directed phosphorylation signaling. Trends Cell Biol 12, 164-172. Lu, K. P., and Zhou, X. Z. (2007). The prolyl isomerase PIN1: a pivotal new twist in phosphorylation signalling and disease. Nat Rev Mol Cell Biol 8, 904-916. Lu, P. J., Zhou, X. Z., Liou, Y. C., Noel, J. P., and Lu, K. P. (2002b). Critical role of WW domain phosphorylation in regulating phosphoserine binding activity and Pin1 function. J Biol Chem 277, 2381-2384.

Bibliography

Lu, X., Liu, D. P., and Xu, Y. (2013b). The gain of function of p53 cancer mutant in promoting mammary tumorigenesis. Oncogene 32, 2900-2906. Magee, J. A., Piskounova, E., and Morrison, S. J. (2012). Cancer stem cells: impact, heterogeneity, and uncertainty. Cancer Cell 21, 283-296. Magli, A., Angelelli, C., Ganassi, M., Baruffaldi, F., Matafora, V., Battini, R., Bachi, A., Messina, G., Rustighi, A., Del Sal, G., et al. (2010). Proline isomerase Pin1 represses terminal differentiation and myocyte enhancer factor 2C function in skeletal muscle cells. J Biol Chem 285, 34518-34527. Maleszka, R., Hanes, S. D., Hackett, R. L., de Couet, H. G., and Miklos, G. L. (1996). The Drosophila melanogaster dodo (dod) gene, conserved in humans, is functionally interchangeable with the ESS1 cell division gene of Saccharomyces cerevisiae. Proc Natl Acad Sci U S A 93, 447-451. Mani, S. A., Guo, W., Liao, M. J., Eaton, E. N., Ayyanan, A., Zhou, A. Y., Brooks, M., Reinhard, F., Zhang, C. C., Shipitsin, M., et al. (2008). The epithelial-mesenchymal transition generates cells with properties of stem cells. Cell 133, 704-715. Mantovani, F., Piazza, S., Gostissa, M., Strano, S., Zacchi, P., Mantovani, R., Blandino, G., and Del Sal, G. (2004). Pin1 links the activities of c-Abl and p300 in regulating p73 function. Mol Cell 14, 625-636. Mantovani, F., Tocco, F., Girardini, J., Smith, P., Gasco, M., Lu, X., Crook, T., and Del Sal, G. (2007). The prolyl isomerase Pin1 orchestrates p53 acetylation and dissociation from the apoptosis inhibitor iASPP. Nat Struct Mol Biol 14, 912-920. Mao, J. H., Kim, I. J., Wu, D., Climent, J., Kang, H. C., DelRosario, R., and Balmain, A. (2008). FBXW7 targets mTOR for degradation and cooperates with PTEN in tumor suppression. Science 321, 1499-1502. Marusyk, A., and Polyak, K. (2010). Tumor heterogeneity: causes and consequences. Biochim Biophys Acta 1805, 105-117. Mazzone, M., et al. Dose-dependent induction of distinct phenotypic responses to Notch pathway activation in mammary epithelial cells. Proc Natl Acad Sci U S A 107, 5012-5017 (2010). McGill, M. A., Dho, S. E., Weinmaster, G., and McGlade, C. J. (2009). Numb regulates post-endocytic trafficking and degradation of Notch1. J Biol Chem 284, 26427-26438. McGill, M. A., and McGlade, C. J. (2003). Mammalian numb proteins promote Notch1 receptor ubiquitination and degradation of the Notch1 intracellular domain. J Biol Chem 278, 23196-23203. Miele, L., Golde, T., and Osborne, B. (2006). Notch signaling in cancer. Curr Mol Med 6, 905-918. Migliaccio, E., Giorgio, M., Mele, S., Pelicci, G., Reboldi, P., Pandolfi, P. P., Lanfrancone, L., and Pelicci, P. G. (1999). The p66shc adaptor protein controls oxidative stress response and life span in mammals. Nature 402, 309-313. Min, S. H., Lau, A. W., Lee, T. H., Inuzuka, H., Wei, S., Huang, P., Shaik, S., Lee, D. Y., Finn, G., Balastik, M., et al. (2012). Negative regulation of the stability and tumor suppressor function of Fbw7 by the Pin1 prolyl isomerase. Mol Cell 46, 771-783. Miyashita, H., Mori, S., Motegi, K., Fukumoto, M., and Uchida, T. (2003). Pin1 is overexpressed in oral squamous cell carcinoma and its levels correlate with cyclin D1 overexpression. Oncol Rep 10, 455-461. Molofsky, A.V., Pardal, R. & Morrison, S.J. Diverse mechanisms regulate stem cell self-renewal. Curr Opin Cell Biol 16, 700-707 (2004). Moretto-Zita, M., Jin, H., Shen, Z., Zhao, T., Briggs, S. P., and Xu, Y. (2010). Phosphorylation stabilizes Nanog by promoting its interaction with Pin1. Proc Natl Acad Sci U S A 107, 13312-13317. Muller, P. A., Trinidad, A. G., Caswell, P. T., Norman, J. C., and Vousden, K. H. (2014). Mutant p53 regulates Dicer through p63-dependent and -independent mechanisms to promote an invasive phenotype. J Biol Chem 289, 122-132. Nakamura, K., Kosugi, I., Lee, D. Y., Hafner, A., Sinclair, D. A., Ryo, A., and Lu, K. P. (2012). Prolyl isomerase Pin1 regulates neuronal differentiation via beta-catenin. Mol Cell Biol 32, 2966-2978.

Bibliography

Nigg, E. A. (2001). Mitotic kinases as regulators of cell division and its checkpoints. Nat Rev Mol Cell Biol 2, 21-32. Nishi, M., Akutsu, H., Masui, S., Kondo, A., Nagashima, Y., Kimura, H., Perrem, K., Shigeri, Y., Toyoda, M., Okayama, A., et al. (2011). A distinct role for Pin1 in the induction and maintenance of pluripotency. J Biol Chem 286, 11593-11603. Oberg, C., et al. The Notch intracellular domain is ubiquitinated and negatively regulated by the mammalian Sel-10 homolog. J Biol Chem 276, 35847-35853 (2001). O'Neil, J., Grim, J., Strack, P., Rao, S., Tibbitts, D., Winter, C., Hardwick, J., Welcker, M., Meijerink, J. P., Pieters, R., et al. (2007). FBW7 mutations in leukemic cells mediate NOTCH pathway activation and resistance to gamma-secretase inhibitors. J Exp Med 204, 1813-1824. Okamoto, K., and Sagata, N. (2007). Mechanism for inactivation of the mitotic inhibitory kinase Wee1 at M phase. Proc Natl Acad Sci U S A 104, 3753-3758. Osborne, B. A., and Minter, L. M. (2007). Notch signalling during peripheral T-cell activation and differentiation. Nat Rev Immunol 7, 64-75. Palomero, T., Dominguez, M., and Ferrando, A. A. (2008). The role of the PTEN/AKT Pathway in NOTCH1-induced leukemia. Cell Cycle 7, 965-970. Palomero, T., Lim, W. K., Odom, D. T., Sulis, M. L., Real, P. J., Margolin, A., Barnes, K. C., O'Neil, J., Neuberg, D., Weng, A. P., et al. (2006). NOTCH1 directly regulates c-MYC and activates a feed-forward-loop transcriptional network promoting leukemic cell growth. Proc Natl Acad Sci U S A 103, 18261-18266. Pear, W. S., and Simon, M. C. (2005). Lasting longer without oxygen: The influence of hypoxia on Notch signaling. Cancer Cell 8, 435-437. Pece, S., Serresi, M., Santolini, E., Capra, M., Hulleman, E., Galimberti, V., Zurrida, S., Maisonneuve, P., Viale, G., and Di Fiore, P. P. (2004). Loss of negative regulation by Numb over Notch is relevant to human breast carcinogenesis. J Cell Biol 167, 215-221. Pece, S., Tosoni, D., Confalonieri, S., Mazzarol, G., Vecchi, M., Ronzoni, S., Bernard, L., Viale, G., Pelicci, P. G., and Di Fiore, P. P. (2010). Biological and molecular heterogeneity of breast cancers correlates with their cancer stem cell content. Cell 140, 62-73. Perou, C. M., Sorlie, T., Eisen, M. B., van de Rijn, M., Jeffrey, S. S., Rees, C. A., Pollack, J. R., Ross, D. T., Johnsen, H., Akslen, L. A., et al. (2000). Molecular portraits of human breast tumours. Nature 406, 747-752. Pinton, P., Rimessi, A., Marchi, S., Orsini, F., Migliaccio, E., Giorgio, M., Contursi, C., Minucci, S., Mantovani, F., Wieckowski, M. R., et al. (2007). Protein kinase C beta and prolyl isomerase 1 regulate mitochondrial effects of the life-span determinant p66Shc. Science 315, 659-663. Polyak, K. (2011). Heterogeneity in breast cancer. J Clin Invest 121, 3786-3788 Prabakaran, S., Lippens, G., Steen, H., and Gunawardena, J. (2012). Post-translational modification: nature's escape from genetic imprisonment and the basis for dynamic information encoding. Wiley Interdiscip Rev Syst Biol Med 4, 565-583. Radtke, F., and Raj, K. (2003). The role of Notch in tumorigenesis: oncogene or tumour suppressor? Nat Rev Cancer 3, 756-767. Rajbhandari, P., Finn, G., Solodin, N. M., Singarapu, K. K., Sahu, S. C., Markley, J. L., Kadunc, K. J., Ellison-Zelski, S. J., Kariagina, A., Haslam, S. Z., et al. (2012). Regulation of estrogen receptor alpha N-terminus conformation and function by peptidyl prolyl isomerase Pin1. Mol Cell Biol 32, 445-457. Ranganathan, P., Weaver, K. L., and Capobianco, A. J. (2011). Notch signalling in solid tumours: a little bit of everything but not all the time. Nat Rev Cancer 11, 338-351. Ranganathan, R., Lu, K. P., Hunter, T., and Noel, J. P. (1997). Structural and functional analysis of the mitotic rotamase Pin1 suggests substrate recognition is phosphorylation dependent. Cell 89, 875-886.

Bibliography

Rangasamy, V., Mishra, R., Sondarva, G., Das, S., Lee, T. H., Bakowska, J. C., Tzivion, G., Malter, J. S., Rana, B., Lu, K. P., et al. (2012). Mixed-lineage kinase 3 phosphorylates prolyl-isomerase Pin1 to regulate its nuclear translocation and cellular function. Proc Natl Acad Sci U S A 109, 8149-8154. Raouf, A., Zhao, Y., To, K., Stingl, J., Delaney, A., Barbara, M., Iscove, N., Jones, S., McKinney, S., Emerman, J., et al. (2008). Transcriptome analysis of the normal human mammary cell commitment and differentiation process. Cell Stem Cell 3, 109-118. Reedijk, M., Odorcic, S., Chang, L., Zhang, H., Miller, N., McCready, D. R., Lockwood, G., and Egan, S. E. (2005). High-level coexpression of JAG1 and NOTCH1 is observed in human breast cancer and is associated with poor overall survival. Cancer Res 65, 8530-8537. Reimand, J., Wagih, O., and Bader, G. D. (2013). The mutational landscape of phosphorylation signaling in cancer. Sci Rep 3, 2651. Reynolds, T. C., Smith, S. D., and Sklar, J. (1987). Analysis of DNA surrounding the breakpoints of chromosomal translocations involving the beta T cell receptor gene in human lymphoblastic neoplasms. Cell 50, 107-117. Rios, A. C., Fu, N. Y., Lindeman, G. J., and Visvader, J. E. (2014). In situ identification of bipotent stem cells in the mammary gland. Nature 506, 322-327. Rizzo, P., Osipo, C., Foreman, K., Golde, T., Osborne, B., and Miele, L. (2008). Rational targeting of Notch signaling in cancer. Oncogene 27, 5124-5131. Rudland, P. S., Barraclough, R., Fernig, D. G., and Smith, J. A. (1997). Mammary stem cells in normal development and cancer. Stem Cells 147. Rustighi, A., Tiberi, L., Soldano, A., Napoli, M., Nuciforo, P., Rosato, A., Kaplan, F., Capobianco, A., Pece, S., Di Fiore, P. P., and Del Sal, G. (2009). The prolyl-isomerase Pin1 is a Notch1 target that enhances Notch1 activation in cancer. Nat Cell Biol 11, 133-142. Ryo, A., Liou, Y. C., Wulf, G., Nakamura, M., Lee, S. W., and Lu, K. P. (2002). PIN1 is an E2F target gene essential for Neu/Ras-induced transformation of mammary epithelial cells. Mol Cell Biol 22, 5281-5295. Ryo, A., Nakamura, M., Wulf, G., Liou, Y. C., and Lu, K. P. (2001). Pin1 regulates turnover and subcellular localization of beta-catenin by inhibiting its interaction with APC. Nat Cell Biol 3, 793-801. Ryo, A., Suizu, F., Yoshida, Y., Perrem, K., Liou, Y. C., Wulf, G., Rottapel, R., Yamaoka, S., and Lu, K. P. (2003). Regulation of NF-kappaB signaling by Pin1-dependent prolyl isomerization and ubiquitin-mediated proteolysis of p65/RelA. Mol Cell 12, 1413-1426. Santarpia, L., Qi, Y., Stemke-Hale, K., Wang, B., Young, E. J., Booser, D. J., Holmes, F. A., O'Shaughnessy, J., Hellerstedt, B., Pippen, J., et al. (2012). Mutation profiling identifies numerous rare drug targets and distinct mutation patterns in different clinical subtypes of breast cancers. Breast Cancer Res Treat 134, 333-343. Sarig, R., Rivlin, N., Brosh, R., Bornstein, C., Kamer, I., Ezra, O., Molchadsky, A., Goldfinger, N., Brenner, O., and Rotter, V. (2010). Mutant p53 facilitates somatic cell reprogramming and augments the malignant potential of reprogrammed cells. J Exp Med 207, 2127-2140. Sarkadi, B., Homolya, L., Szakacs, G., and Varadi, A. (2006). Human multidrug resistance ABCB and ABCG transporters: participation in a chemoimmunity defense system. Physiol Rev 86, 1179-1236. Schutkowski, M., Bernhardt, A., Zhou, X. Z., Shen, M., Reimer, U., Rahfeld, J. U., Lu, K. P., and Fischer, G. (1998). Role of phosphorylation in determining the backbone dynamics of the serine/threonine-proline motif and Pin1 substrate recognition. Biochemistry 37, 5566-5575. Shackleton, M., Vaillant, F., Simpson, K. J., Stingl, J., Smyth, G. K., Asselin-Labat, M. L., Wu, L., Lindeman, G. J., and Visvader, J. E. (2006). Generation of a functional mammary gland from a single stem cell. Nature 439, 84-88. Shen, Z. J., Hu, J., Ali, A., Pastor, J., Shiizaki, K., Blank, R. D., Kuro-o, M., and Malter, J. S. (2013). Pin1 null mice exhibit low bone mass and attenuation of BMP signaling. PLoS One 8, e63565.

Bibliography

Sims, A. H., Howell, A., Howell, S. J., and Clarke, R. B. (2007). Origins of breast cancer subtypes and therapeutic implications. Nat Clin Pract Oncol 4, 516-525. Sotiriou, C., Neo, S. Y., McShane, L. M., Korn, E. L., Long, P. M., Jazaeri, A., Martiat, P., Fox, S. B., Harris, A. L., and Liu, E. T. (2003). Breast cancer classification and prognosis based on gene expression profiles from a population-based study. Proc Natl Acad Sci U S A 100, 10393-10398. Sorlie, T., Perou, C. M., Tibshirani, R., Aas, T., Geisler, S., Johnsen, H., Hastie, T., Eisen, M. B., van de Rijn, M., Jeffrey, S. S., et al. (2001). Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci U S A 98, 10869-10874. Sorlie, T., Tibshirani, R., Parker, J., Hastie, T., Marron, J. S., Nobel, A., Deng, S., Johnsen, H., Pesich, R., Geisler, S., et al. (2003). Repeated observation of breast tumor subtypes in independent gene expression data sets. Proc Natl Acad Sci U S A 100, 8418-8423. Sorrentino, G., Mioni, M., Giorgi, C., Ruggeri, N., Pinton, P., Moll, U., Mantovani, F., and Del Sal, G. (2013). The prolyl-isomerase Pin1 activates the mitochondrial death program of p53. Cell Death Differ 20, 198-208. Spike, B. T., Engle, D. D., Lin, J. C., Cheung, S. K., La, J., and Wahl, G. M. (2012). A mammary stem cell population identified and characterized in late embryogenesis reveals similarities to human breast cancer. Cell Stem Cell 10, 183-197. Steeg, P. S., and Theodorescu, D. (2008). Metastasis: a therapeutic target for cancer. Nat Clin Pract Oncol 5, 206-219. Steger, M., Murina, O., Huhn, D., Ferretti, L. P., Walser, R., Hanggi, K., Lafranchi, L., Neugebauer, C., Paliwal, S., Janscak, P., et al. (2013). Prolyl isomerase PIN1 regulates DNA double-strand break repair by counteracting DNA end resection. Mol Cell 50, 333-343. Stingl, J., and Caldas, C. (2007). Molecular heterogeneity of breast carcinomas and the cancer stem cell hypothesis. Nat Rev Cancer 7, 791-799. Stingl, J., Eaves, C. J., Kuusk, U., and Emerman, J. T. (1998). Phenotypic and functional characterization in vitro of a multipotent epithelial cell present in the normal adult human breast. Differentiation 63, 201-213. Stingl, J., Eaves, C. J., Zandieh, I., and Emerman, J. T. (2001). Characterization of bipotent mammary epithelial progenitor cells in normal adult human breast tissue. Breast Cancer Res Treat 67, 93-109. Stingl, J., Eirew, P., Ricketson, I., Shackleton, M., Vaillant, F., Choi, D., Li, H. I., and Eaves, C. J. (2006). Purification and unique properties of mammary epithelial stem cells. Nature 439, 993-997. Sudol, M., and Hunter, T. (2000). NeW wrinkles for an old domain. Cell 103, 1001-1004. Sudol, M., Sliwa, K., and Russo, T. (2001). Functions of WW domains in the nucleus. FEBS Lett 490, 190-195. Suizu, F., Ryo, A., Wulf, G., Lim, J., and Lu, K. P. (2006). Pin1 regulates centrosome duplication, and its overexpression induces centrosome amplification, chromosome instability, and oncogenesis. Mol Cell Biol 26, 1463-1479. Sun, Y., Klauzinska, M., Lake, R. J., Lee, J. M., Santopietro, S., Raafat, A., Salomon, D., Callahan, R., and Artavanis-Tsakonas, S. (2011). Trp53 regulates Notch 4 signaling through Mdm2. J Cell Sci 124, 1067-1076. Swiatek, P.J., Lindsell, C.E., del Amo, F.F., Weinmaster, G. & Gridley, T. Notch1 is essential for postimplantation development in mice. Genes Dev 8, 707-719 (1994). Takahashi, K., and Yamanaka, S. (2006). Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126, 663-676. Tan, X., Zhou, F., Wan, J., Hang, J., Chen, Z., Li, B., Zhang, C., Shao, K., Jiang, P., Shi, S., et al. (2010). Pin1 expression contributes to lung cancer: Prognosis and carcinogenesis. Cancer Biol Ther 9, 111-119. Tao, L., Roberts, A. L., Dunphy, K. A., Bigelow, C., Yan, H., and Jerry, D. J. (2011). Repression of mammary stem/progenitor cells by p53 is mediated by Notch and separable from apoptotic activity. Stem Cells 29, 119-127. Tetzlaff, M. T., Yu, W., Li, M., Zhang, P., Finegold, M., Mahon, K., Harper, J. W., Schwartz, R. J., and Elledge, S. J. (2004). Defective cardiovascular development and elevated cyclin E and Notch proteins in mice lacking the Fbw7 F-box protein. Proc Natl Acad Sci U S A 101, 3338-3345.

Bibliography

Thompson, B.J., et al. The SCFFBW7 ubiquitin ligase complex as a tumor suppressor in T cell leukemia. J Exp Med 204, 1825-1835 (2007). Timmerman, L.A., et al. Notch promotes epithelial-mesenchymal transition during cardiac development and oncogenic transformation. Genes Dev 18, 99-115 (2004). Tsunematsu, R., Nakayama, K., Oike, Y., Nishiyama, M., Ishida, N., Hatakeyama, S., Bessho, Y., Kageyama, R., Suda, T., and Nakayama, K. I. (2004). Mouse Fbw7/Sel-10/Cdc4 is required for notch degradation during vascular development. J Biol Chem 279, 9417-9423. Uchida, T., Furumai, K., Fukuda, T., Akiyama, H., Takezawa, M., Asano, T., Fujimori, F., and Uchida, C. (2012). Prolyl isomerase Pin1 regulates mouse embryonic fibroblast differentiation into adipose cells. PLoS One 7, e31823. Uchida, T., Takamiya, M., Takahashi, M., Miyashita, H., Ikeda, H., Terada, T., Matsuo, Y., Shirouzu, M., Yokoyama, S., Fujimori, F., and Hunter, T. (2003). Pin1 and Par14 peptidyl prolyl isomerase inhibitors block cell proliferation. Chem Biol 10, 15-24. Uemura, T., Shepherd, S., Ackerman, L., Jan, L. Y., and Jan, Y. N. (1989). numb, a gene required in determination of cell fate during sensory organ formation in Drosophila embryos. Cell 58, 349-360. Valent, P., Bonnet, D., De Maria, R., Lapidot, T., Copland, M., Melo, J. V., Chomienne, C., Ishikawa, F., Schuringa, J. J., Stassi, G., et al. (2012). Cancer stem cell definitions and terminology: the devil is in the details. Nat Rev Cancer 12, 767-775. van Drogen, F., Sangfelt, O., Malyukova, A., Matskova, L., Yeh, E., Means, A. R., and Reed, S. I. (2006). Ubiquitylation of cyclin E requires the sequential function of SCF complexes containing distinct hCdc4 isoforms. Mol Cell 23, 37-48. Van Hoof, D., Munoz, J., Braam, S. R., Pinkse, M. W., Linding, R., Heck, A. J., Mummery, C. L., and Krijgsveld, J. (2009). Phosphorylation dynamics during early differentiation of human embryonic stem cells. Cell Stem Cell 5, 214-226. Van Keymeulen, A., Rocha, A. S., Ousset, M., Beck, B., Bouvencourt, G., Rock, J., Sharma, N., Dekoninck, S., and Blanpain, C. (2011). Distinct stem cells contribute to mammary gland development and maintenance. Nature 479, 189-193. Vargo-Gogola, T., and Rosen, J. M. (2007). Modelling breast cancer: one size does not fit all. Nat Rev Cancer 7, 659-672. Visvader, J. E. (2009). Keeping abreast of the mammary epithelial hierarchy and breast tumorigenesis. Genes Dev 23, 2563-2577. Visvader, J. E., and Lindeman, G. J. (2012). Cancer stem cells: current status and evolving complexities. Cell Stem Cell 10, 717-728. Visvader, J. E., and Lindeman, G. J. (2011). The unmasking of novel unipotent stem cells in the mammary gland. EMBO J 30, 4858-4859. Walsh, C. (2006). Posttranslational modification of proteins: expanding nature's inventory: Roberts and Company Publishers). Wang, Z., Inuzuka, H., Fukushima, H., Wan, L., Gao, D., Shaik, S., Sarkar, F. H., and Wei, W. (2012). Emerging roles of the FBW7 tumour suppressor in stem cell differentiation. EMBO Rep 13, 36-43. Wharton, K.A., Johansen, K.M., Xu, T. & Artavanis-Tsakonas, S. Nucleotide sequence from the neurogenic locus notch implies a gene product that shares homology with proteins containing EGF-like repeats. Cell 43, 567-581 (1985). Weigelt, B., Peterse, J. L., and van 't Veer, L. J. (2005). Breast cancer metastasis: markers and models. Nat Rev Cancer 5, 591-602. Welcker, M., and Clurman, B. E. (2008). FBW7 ubiquitin ligase: a tumour suppressor at the crossroads of cell division, growth and differentiation. Nat Rev Cancer 8, 83-93. Welcker, M., Orian, A., Jin, J., Grim, J. E., Harper, J. W., Eisenman, R. N., and Clurman, B. E. (2004). The Fbw7 tumor suppressor regulates glycogen synthase kinase 3 phosphorylation-dependent c-Myc protein degradation. Proc Natl Acad Sci U S A 101, 9085-9090.

Bibliography

Weng, A. P., Ferrando, A. A., Lee, W., Morris, J. P. t., Silverman, L. B., Sanchez-Irizarry, C., Blacklow, S. C., Look, A. T., and Aster, J. C. (2004). Activating mutations of NOTCH1 in human T cell acute lymphoblastic leukemia. Science 306, 269-271. Weng, A. P., Millholland, J. M., Yashiro-Ohtani, Y., Arcangeli, M. L., Lau, A., Wai, C., Del Bianco, C., Rodriguez, C. G., Sai, H., Tobias, J., et al. (2006). c-Myc is an important direct target of Notch1 in T-cell acute lymphoblastic leukemia/lymphoma. Genes Dev 20, 2096-2109. Wertz, I. E., Kusam, S., Lam, C., Okamoto, T., Sandoval, W., Anderson, D. J., Helgason, E., Ernst, J. A., Eby, M., Liu, J., et al. (2011). Sensitivity to antitubulin chemotherapeutics is regulated by MCL1 and FBW7. Nature 471, 110-114. Westhoff, B., Colaluca, I. N., D'Ario, G., Donzelli, M., Tosoni, D., Volorio, S., Pelosi, G., Spaggiari, L., Mazzarol, G., Viale, G., et al. (2009). Alterations of the Notch pathway in lung cancer. Proc Natl Acad Sci U S A 106, 22293-22298. Winkler, K. E., Swenson, K. I., Kornbluth, S., and Means, A. R. (2000). Requirement of the prolyl isomerase Pin1 for the replication checkpoint. Science 287, 1644-1647. Wolff, A. C., Hammond, M. E., Hicks, D. G., Dowsett, M., McShane, L. M., Allison, K. H., Allred, D. C., Bartlett, J. M., Bilous, M., Fitzgibbons, P., et aml. (2013). Recommendations for human epidermal growth factor receptor 2 testing in breast cancer: American Society of Clinical Oncology/College of American Pathologists clinical practice guideline update. J Clin Oncol 31, 3997-4013. Wu, G., Lyapina, S., Das, I., Li, J., Gurney, M., Pauley, A., Chui, I., Deshaies, R. J., and Kitajewski, J. (2001). SEL-10 is an inhibitor of notch signaling that targets notch for ubiquitin-mediated protein degradation. Mol Cell Biol 21, 7403-7415. Wulf, G., Ryo, A., Liou, Y. C., and Lu, K. P. (2003). The prolyl isomerase Pin1 in breast development and cancer. Breast Cancer Res 5, 76-82. Wulf, G. M., Liou, Y. C., Ryo, A., Lee, S. W., and Lu, K. P. (2002). Role of Pin1 in the regulation of p53 stability and p21 transactivation, and cell cycle checkpoints in response to DNA damage. J Biol Chem 277, 47976-47979. Wulf, G. M., Ryo, A., Wulf, G. G., Lee, S. W., Niu, T., Petkova, V., and Lu, K. P. (2001). Pin1 is overexpressed in breast cancer and cooperates with Ras signaling in increasing the transcriptional activity of c-Jun towards cyclin D1. EMBO J 20, 3459-3472. Xing, F., Kobayashi, A., Okuda, H., Watabe, M., Pai, S. K., Pandey, P. R., Hirota, S., Wilber, A., Mo, Y. Y., Moore, B. E., et al. (2013). Reactive astrocytes promote the metastatic growth of breast cancer stem-like cells by activating Notch signalling in brain. EMBO Mol Med 5, 384-396. Xu, K., Usary, J., Kousis, P. C., Prat, A., Wang, D. Y., Adams, J. R., Wang, W., Loch, A. J., Deng, T., Zhao, W., et al. (2012). Lunatic fringe deficiency cooperates with the Met/Caveolin gene amplicon to induce basal-like breast cancer. Cancer Cell 21, 626-641. Xu, Y. X., and Manley, J. L. (2007a). Pin1 modulates RNA polymerase II activity during the transcription cycle. Genes Dev 21, 2950-2962. Xu, Y. X., and Manley, J. L. (2007b). The prolyl isomerase Pin1 functions in mitotic chromosome condensation. Mol Cell 26, 287-300. Yaffe, M. B., Schutkowski, M., Shen, M., Zhou, X. Z., Stukenberg, P. T., Rahfeld, J. U., Xu, J., Kuang, J., Kirschner, M. W., Fischer, G., et al. (1997). Sequence-specific and phosphorylation-dependent proline isomerization: a potential mitotic regulatory mechanism. Science 278, 1957-1960. Yalcin-Ozuysal, O., Fiche, M., Guitierrez, M., Wagner, K. U., Raffoul, W., and Brisken, C. (2010). Antagonistic roles of Notch and p63 in controlling mammary epithelial cell fates. Cell Death Differ 17, 1600-1612. Yatim, A., Benne, C., Sobhian, B., Laurent-Chabalier, S., Deas, O., Judde, J. G., Lelievre, J. D., Levy, Y., and Benkirane, M. (2012). NOTCH1 nuclear interactome reveals key regulators of its transcriptional activity and oncogenic function. Mol Cell 48, 445-458. Yeh, E. S., and Means, A. R. (2007). PIN1, the cell cycle and cancer. Nat Rev Cancer 7, 381-388. Yeo, J. C., and Ng, H. H. (2013). The transcriptional regulation of pluripotency. Cell Res 23, 20-32.

Bibliography

Yi, P., Wu, R. C., Sandquist, J., Wong, J., Tsai, S. Y., Tsai, M. J., Means, A. R., and O'Malley, B. W. (2005). Peptidyl-prolyl isomerase 1 (Pin1) serves as a coactivator of steroid receptor by regulating the activity of phosphorylated steroid receptor coactivator 3 (SRC-3/AIB1). Mol Cell Biol 25, 9687-9699. You, H., Zheng, H., Murray, S. A., Yu, Q., Uchida, T., Fan, D., and Xiao, Z. X. (2002). IGF-1 induces Pin1 expression in promoting cell cycle S-phase entry. J Cell Biochem 84, 211-216. Yu, F., Yao, H., Zhu, P., Zhang, X., Pan, Q., Gong, C., Huang, Y., Hu, X., Su, F., Lieberman, J., and Song, E. (2007). let-7 regulates self renewal and tumorigenicity of breast cancer cells. Cell 131, 1109-1123. Zacchi, P., Gostissa, M., Uchida, T., Salvagno, C., Avolio, F., Volinia, S., Ronai, Z., Blandino, G., Schneider, C., and Del Sal, G. (2002). The prolyl isomerase Pin1 reveals a mechanism to control p53 functions after genotoxic insults. Nature 419, 853-857. Zhang, X., Zhang, B., Gao, J., Wang, X., and Liu, Z. (2013). Regulation of the microRNA 200b (miRNA-200b) by transcriptional regulators PEA3 and ELK-1 protein affects expression of Pin1 protein to control anoikis. J Biol Chem 288, 32742-32752. Zheng, H., You, H., Zhou, X. Z., Murray, S. A., Uchida, T., Wulf, G., Gu, L., Tang, X., Lu, K. P., and Xiao, Z. X. (2002). The prolyl isomerase Pin1 is a regulator of p53 in genotoxic response. Nature 419, 849-853. Zheng, Y., Xia, Y., Hawke, D., Halle, M., Tremblay, M. L., Gao, X., Zhou, X. Z., Aldape, K., Cobb, M. H., Xie, K., et al. (2009). FAK phosphorylation by ERK primes ras-induced tyrosine dephosphorylation of FAK mediated by PIN1 and PTP-PEST. Mol Cell 35, 11-25. Zhou, X. Z., Kops, O., Werner, A., Lu, P. J., Shen, M., Stoller, G., Kullertz, G., Stark, M., Fischer, G., and Lu, K. P. (2000). Pin1-dependent prolyl isomerization regulates dephosphorylation of Cdc25C and tau proteins. Mol Cell 6, 873-883.

Appendix

9. APPENDIX

During the period of my PhD I have been collaborating in the following publication: Rustighi, A.*, Zannini, A.*, Tiberi, L., Sommaggio, R., Piazza, S., Sorrentino, G., Nuzzo, S., Tuscano, A., Eterno, V., Benvenuti, F., et al. (2014). Prolyl-isomerase Pin1 controls normal and cancer stem cells of the breast. EMBO Mol Med 6, 99-119. (*= these authors equally cotributed to the work)

10. ACKNOWLEDGEMENTS

Voglio ringraziare il professor Giannino Del Sal per avermi dato l’opportunità di lavorare a

questo progetto, per avermi fatto crescere sia scientificamente sia professionalmente.

Voglio ringraziare la dottoressa Alessandra Rustighi per aver contribuito in modo fondamentale

alla mia formazione, alla pubblicazione di questo lavoro di tesi e alla sua stesura. Un

ringraziamento speciale va fatto anche per la sua disponibilità ad ascoltarmi e aiutarmi ogni

qualvolta ne avessi avuto bisogno e per avermi sempre messo alla prova affrontando i miei

limiti.

Un ringraziamento va anche alle dottoresse Anna Comel, Elena Campaner, Fiamma Mantovani e

Marta Milan, per i loro suggerimenti e per il loro supporto sia scientifico che umano durante

tutto il periodo di PhD. Ringrazio anche il dottor Giovanni Sorrentino per aver sperimentalmente

contribuito a questo progetto. Ringrazio anche il mio amico Sante per avermi aiutato durante la

stesura della tesi.

Ringrazio i miei genitori, perché indubbiamente sono stati i miei pilastri da ventotto anni a

questa parte e per avermi sempre supportato e sopportato. Grazie perché per me ci siete sempre e

mi fate sentire sempre speciale e prioritario rispetto a tutto il resto.

Ringrazio i miei nonni per essere sempre presenti e affettuosi anche se riusciamo a vederci poco.

Ringrazio tutti i miei amici perché a volte la famiglia non è solamente quella che ti trovi ma

soprattutto quella che ti costruisci; ringrazio in particolare Alessandro, Andrea, Anna, Clara,

Elena, Luca, Marco S., Marco P., Michael, Stefania.