Reprogramming to developmental plasticity in cancer stem cells

Developmental Biology xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Developmental Biology journal homepage: www.elsevier.com/locate/developmentalbiology

Reprogramming to developmental plasticity in cancer stem cells ⁎

Caitlin O’Brien-Ball, Adrian Biddle

Blizard Institute, Barts and The London School of Medicine and Dentistry, Queen Mary University of London, UK

A R T I C L E I N F O

A BS T RAC T

Keywords: Development Plasticity EMT Cancer Stem Reprogramming

During development and throughout adult life, sub-populations of cells exist that exhibit phenotypic plasticity – the ability to differentiate into multiple lineages. This behaviour is important in embryogenesis, is exhibited in a more limited context by adult stem cells, and can be re-activated in cancer cells to drive important processes underlying tumour progression. A well-studied mechanism of phenotypic plasticity is the epithelial-tomesenchymal transition (EMT), a process which has been observed in both normal and cancerous cells. The epigenetic and metabolic modifications necessary to facilitate phenotypic plasticity are first seen in development and can be re-activated both in normal regeneration and in cancer. In cancer, the re-activation of these mechanisms enables tumour cells to acquire a cancer stem cell (CSC) phenotype with enhanced ability to survive in hostile environments, resist therapeutic interventions, and undergo metastasis. However, recent research has suggested that plasticity may also expose weaknesses in cancer cells that could be exploited for future therapeutic development. More research is needed to identify developmental mechanisms that are active in cancer, so that these may be targeted to reduce tumour growth and metastasis and overcome therapeutic resistance.

1. Introduction As tumours grow, they evolve through selection to adapt to their environment. Traditionally, this has been viewed as depending on genetic evolution (Greaves, 2015), but the role of phenotypic plasticity in driving tumour adaptation is increasingly recognised. The term ‘phenotypic plasticity’ describes the ability of cells to differentiate into multiple lineages, otherwise known as multipotency. In cancer, this ability is re-acquired by lineage restricted cells through reprogramming of their epigenetic state. This reprogramming may involve re-activation of developmental programs that can drive tumour adaptation, which will be the focus of this review. The precise molecular modifications underlying the epigenetic changes that enable re-acquisition of multipotency are reviewed elsewhere (Easwaran et al., 2014). Recent work has shed light on the relative contributions to tumour development of genetic selection versus phenotypic plasticity (Sottoriva et al., 2015; Williams et al., 2016). These studies demonstrated that, whilst some tumours are heavily dependent on genetic evolution, others undergo no further genetic selection after the early tumour-initiating mutational events and any further adaptation must occur through phenotypic plasticity. Reprogramming to plasticity by tumour-initiating mutations, and consequent reactivation of developmental programs, thus represents a potent mechanism whereby tumours may reversibly adapt to different environmental challenges in the absence of genetic evolution (Fig. 1).

⁎

The idea of phenotypic plasticity in cancer is now relatively well established, although the exact mechanisms behind this behaviour are still not well understood. It is believed that some cancerous cells undergo epigenetic reprogramming to induce metabolic and phenotypic changes which are often linked to behaviours giving the tumour the ability to become more invasive and resistant to treatment (Biddle and Mackenzie, 2012; Gupta et al., 2009; Lee et al., 2016), as well as increasing their lineage potential – the number of possible phenotypes that could arise from a cell. This behaviour is often described as phenotypic plasticity; the ability of a cell to change its phenotype, and in some cases do this multiple times (Biddle et al., 2011; Roesch et al., 2010). A well-studied example of plasticity is the ability to undergo epithelial-to-mesenchymal transition (EMT), a process normally seen in the developing embryo, where cells in a tumour of epithelial origin acquire the ability to express markers and behaviours associated with mesenchymal cells, becoming more invasive and less polarised (Hay, 2005; Yang et al., 2008). There is significant evidence for the model of a cancerous tumour consisting of a heterogeneous population of different cell types (Fisher et al., 2013; Heppner, 1984). A particular subset of cells, the cancer stem cells (CSCs), has the ability to divide symmetrically and asymmetrically in order to initiate and maintain tumour growth (Dalerba et al., 2007), and is a source of phenotypic plasticity in the tumour (Lee et al., 2016). Within this loose definition of a CSC, cells that have undergone

Correspondence to: Centre for Cell Biology and Cutaneous Research, Blizard Institute, 4 Newark Street, London E1 2AT, UK. E-mail address: [email protected] (A. Biddle).

http://dx.doi.org/10.1016/j.ydbio.2017.07.025 Received 13 March 2017; Received in revised form 26 May 2017; Accepted 30 July 2017 0012-1606/ © 2017 Elsevier Inc. All rights reserved.

Please cite this article as: O’Brien-Ball, C., Developmental Biology (2017), http://dx.doi.org/10.1016/j.ydbio.2017.07.025

Developmental Biology xxx (xxxx) xxx–xxx

C. O’Brien-Ball, A. Biddle

Fig. 1. Two different models for genetic versus epigenetic contributions to tumour progression (Sottoriva et al., 2015).

ment, so perhaps it is not surprising that cancer cells hijack these mechanisms to drive their own development. Developing embryos rely on the ability of cells to change phenotype and alter their epigenetic state. Many crucial processes in development require these behaviours; for example in neural tube formation (Green et al., 2015). There is also some continuing limited phenotypic plasticity in adult human tissues. In many adult tissues, there exists a population of stem cells whose purpose is to steadily replicate and produce the differentiated cells required in that particular tissue (Lei et al., 2014) whilst retaining some limited regenerative capacity. This has been observed in the intestinal epithelium (Buczacki et al., 2013), skeletal muscle (Moss and Leblond, 1970) and breast tissue (Shackleton et al., 2006) to name a few examples. Although mammals are not capable of the extraordinary regenerative ability of organisms such as salamanders and hydra, they do have the ability to regenerate some tissues when damage occurs, for example in the liver, where stem cell-mediated regeneration can occur depending on severity of injury (Riehle et al., 2011). Some specialised mammalian cells do however have the ability to reprogram their epigenetics with the goal of returning to a totipotent state – these are the gametocytes. The global epigenetic landscape is considerably altered in these specialised cells so that the resulting fertilised egg can go on to generate an entire organism (Cantone and Fisher, 2013; Teperek et al., 2016).

EMT have received particular attention owing to their therapeutic resistance (Gupta et al., 2009) and their ability to survive in a hostile environment to more efficiently seed tumours in immunocompetent mouse models (Gjerdrum et al., 2010). However, it is clear that stem cell characteristics such as self-renewal, phenotypic plasticity and tumour-initiating potential are also shared (if to a lesser degree) by epithelial tumour cell sub-populations that have not undergone EMT (Biddle et al., 2011). The phenotypic plasticity of both EMT and epithelial sub-populations enables regeneration of each by the other, as has been seen in the spontaneous production of EMT cells by isolated epithelial sub-populations from mammary tumours (Chaffer et al., 2011). This is sometimes referred to as ‘de-differentiation’, although it is unlikely to involve terminally differentiated cells re-acquiring a stem cell state. Instead, it is more likely the case that partially-differentiated sub-populations retaining some degree of stem cell characteristics can act to regenerate more undifferentiated stem cell sub-populations. Even within single genetic clones, CSCs diverge on an epigenetic and phenotypic level, leading to multiple phenotypic sub-populations within a single tumour (Biddle et al., 2011; Hermann et al., 2007; Kreso et al., 2013). Plastic CSCs may play an important role in tumour progression and therapeutic resistance, as they have an increased ability to adapt to challenges presented by drug therapy, and the tumour microenvironment (e.g. hypoxia) (Biddle et al., 2016; Gammon et al., 2013; Kreso et al., 2013). Plasticity may also enable resistance to stresses encountered during metastasis, including detachment from the ECM and increased oxidative stress (Piskounova et al., 2015). Once at a metastatic site, plasticity enables restoration of the cellular heterogeneity characteristic of the primary tumour (Thiery, 2002). Therefore, plastic CSCs have become an attractive target for cancer therapy (Biddle et al., 2016) and for assessment of patient prognosis (Lee et al., 2016). Phenotypic plasticity is essential for successful human develop-

2. Phenotypic plasticity and lineage potential 2.1. Plasticity in development During development, the single fertilised egg from which the embryo will form is said to be totipotent – able to produce any cell type, both embryonic and extra-embryonic (Morgani and Brickman, 2014). As embryogenesis progresses, the zygote develops into a 2

Developmental Biology xxx (xxxx) xxx–xxx

C. O’Brien-Ball, A. Biddle

In the intestine, the stem cells are Lgr5+ (Barker et al., 2007) and Buczacki et al. (2013) later classified these Lgr5+ murine intestinal crypt stem cells as either quiescent or rapidly cycling. The quiescent population had partially differentiated towards a Paneth cell lineage but retained the ability to rapidly proliferate after intestinal injury, replenishing all cell types of the intestine despite normally being committed to produce only Paneth cells. They further demonstrated that this quiescent and multipotent sub-population could be identified by staining for CD24. This study demonstrated the ability of adult stem cells to re-acquire multi-lineage potential after injury, a change that could potentially be exploited in cancer. In normal adult skin, plasticity through a limited EMT can be observed after wounding. Keratinocytes have shown the ability to adopt a mesenchymal-like morphology to remodel the matrix and reepithelialise the wound (Leopold et al., 2012). These keratinocytes are able to become motile by activating signalling pathways such as PI3K/Akt/ mTOR and increasing the use of Slug and Twist transcription factors, all traits shared with cells that have undergone EMT in tumours (Leopold et al., 2012). Breast tissue must be able to go through cycles of expansion and involution at critical times during puberty and pregnancy. Expansion is triggered by hormones such as oestrogen, progesterone and prolactin which stimulate the quiescent stem cells to proliferate (Visvader and Stingl, 2014). In the breast tissue two major cell types are required – luminal and myoepithelial – and these can both derive from the single population of breast stem cells (Rios et al., 2014; Stingl, 2009), thus demonstrating the ability for adult stem cells to possess multi-lineage potential. The multipotent breast stem cells can be identified using a CD24+/medCD49fhigh (mouse) or EpCAM+/lowCD49fhigh (human) cell surface marker profile (Stingl, 2009), and normally exist in a quiescent state (Cai et al., 2017). Notably, in this scenario cells negative for either CD24 or EpCAM do not possess multi-lineage potential. In breast cancer, it has been shown that the stem cells are also multipotent, with the ability to form epithelial and mesenchymal progeny (Battula et al., 2010; Mani et al., 2008), in this way resembling the multipotent stem cells normally residing in the breast. This supports the concept that CSCs may have their origin in normal tissue stem cells. It also raises the intriguing question of whether normal tissue stem cells are the ‘cell of origin’ for tumour initiation, becoming dysregulated upon accumulation of oncogenic insults.

blastocyst containing the inner cell mass (ICM), the in vivo tissue from which embryonic stem (ES) cells are derived. These stem cells are pluripotent; capable of self-renewal and the generation of any embryonic – but not extra-embryonic – tissue (Morgani and Brickman, 2014). Continuing through development the lineage potential of the cells decreases as they become more specialised through epigenetic modifications (Waddington, 1957), although a population of stem cells is retained through childhood and adulthood as part of normal tissue maintenance and regeneration (Rumman et al., 2015). The epigenetic modifications imposed during development specify cells into defined lineages, but since they do not change the actual genetic code there is a potential for plasticity, as demonstrated by seminal experiments in which the addition of four transcription factors to differentiated fibroblasts caused them to return to a pluripotent ES-like state (Takahashi and Yamanaka, 2006). The ability of a cell to switch between epithelial and mesenchymal phenotypes is essential for development (Brabletz et al., 2005). Neural crest progenitor cells for example undergo EMT during neural tube formation (Kalcheim, 2015). This process also occurs during gastrulation and heart formation (Thiery and Sleeman, 2006). The reverse process, mesenchymal-to-epithelial transition (MET) is equally important in embryogenesis (Kalcheim, 2015; Thiery and Sleeman, 2006), occurring during the formation of secondary epithelial tissues such as the endocardium (Lim and Thiery, 2012). EMT is a reversible process: during heart development, cells can undergo multiple EMT and MET cycles to switch between epithelial and mesenchymal phenotypes (Lim and Thiery, 2012). Transitions such as EMT and MET are also crucial for the differentiation of precursor cells possessing multi-lineage potential into cells with a fixed cell fate (Kalcheim, 2015). This indicates that some cells naturally possess phenotypic plasticity under the right conditions, and that there may be potential for recapitulation of these developmental programs later in life. An epithelial phenotype is associated with apico-basal polarity, strong cell-cell junctions (Jolly et al., 2015), and expression of Ecadherin (Wheelock et al., 2008). In contrast, a mesenchymal phenotype is marked by the loss of polarity and cell adhesiveness with an increased motility (Vig et al., 2015) and the ability to remodel the surrounding stroma via matrix metalloproteinases (MMPs) (Leopold et al., 2012). EMT is known to be triggered by stromal factors such as TGFβ and TNFα which activate signalling pathways such as the mitogen-associated protein kinase (MAPK) and phosphatidylinositol3-kinase (PI3K) pathways, ultimately leading to repression of genes involved in cell polarisation and activation of Snail, Twist and Zeb transcription factors (Dave et al., 2011; Xu et al., 2009). There is also a ‘cadherin switch’ to N-cadherin expression (Wheelock et al., 2008).

2.3. Plasticity in cancer – cancer stem cells It is now evident that malignant tumours are not simply a homogeneous mass, but instead a heterogeneous population consisting of tumour cells, stromal cells, and endothelial cells, among others (Kreso and Dick, 2014). Within the tumour cells, there is a subset of cells with stem cell-like properties – these are the CSCs (Clarke et al., 2006). Characteristics of CSCs include their ability to initiate tumour formation (Clarke et al., 2006), drive metastasis, and resist therapeutics (Gupta et al., 2009; Li et al., 2008). Thus, they are heavily implicated in the failure of cancer treatment (Medema, 2013). Some CSC sub-populations exist in a relatively quiescent state (Bao et al., 2006; Li et al., 2008) until they are required to regrow the primary tumour, or the secondary tumour after metastasis. They also often possess phenotypic plasticity, similarly to EMT/MET-competent cells in development and stem-cell mediated regeneration (Biddle et al., 2011). Markers associated with CSCs are often shared with adult and embryonic stem cells – for example CD44, the hyaluronic acid receptor (Nishikawa et al., 2015), which has been associated with CSCs in breast cancer (Tam and Weinberg, 2013), oral squamous cell carcinoma (Locke et al., 2005) and prostate cancer (Patrawala et al., 2006), among others. However, markers of CSCs do vary depending on the tissue of origin – for example CD133 in brain (Al-Hajj et al., 2003) and pancreatic (Hermann et al., 2007) tumours, and ALDH1 in breast

2.2. Plasticity in adult tissue regeneration Plasticity can be seen in the normal process of mammalian regeneration, although the regenerative capacity of mammals is much less than that of some other animals such as amphibians (Christen et al., 2010). In adults, the plasticity and lineage potential of somatic cells is greatly reduced as epigenetic modifications cause these cells to become fully differentiated and specialised for their function. However, each tissue is maintained by a population of stem cells (Rumman et al., 2015), and many of these stem cells possess limited phenotypic plasticity which gives them multi-lineage potential (Buczacki et al., 2013; Shackleton et al., 2006). Some stem cell populations exist in a relatively quiescent state (Rumman et al., 2015) until they are required to proliferate and differentiate in response to particular cues – for example injury (Buczacki et al., 2013) and hormones (Asselin-Labat et al., 2010). Upon the correct signal(s) the cells proliferate and subsequently differentiate to form (or regrow) the tissue containing all necessary cell types in the correct structure. Some well-studied examples include the stem cells residing in the intestinal crypts, skin and breast tissue. 3

Developmental Biology xxx (xxxx) xxx–xxx

C. O’Brien-Ball, A. Biddle

Fig. 2. A multipotent stem cell in development (Morrison et al., 1999), regeneration (Cai et al., 2017; Stingl, 2009) and cancer (Biddle et al., 2016; Grosse-Wilde et al., 2015).

in development rather than the more limited process occurring in reepithelialisation. These findings indicate that CSCs with the ability to undergo phenotypic transitions are potentially the most dangerous cancer cell subset and hence should be a target for therapeutic intervention. Recent studies have extended this concept by suggesting a direct link between phenotypic plasticity and stemness in cancer (Biddle et al., 2011; Grosse-Wilde et al., 2015; Jolly et al., 2015), where some cells that undergo EMT retain a ‘hybrid’ epithelial/mesenchymal phenotype that shows increased phenotypic plasticity and stem cell-like properties. More recently, it has been demonstrated that maintenance of multi-lineage potential in EMT cells is essential for both their therapeutic resistance and tumour-initiating ability, and that these highly plastic cells are CD44highEpCAMlow/-CD24+ (Biddle et al., 2016). It has also recently been demonstrated that a retained hybrid epithelial/mesenchymal phenotype at the tumour invasive edge is an indicator of poor prognosis in oral squamous cell carcinoma (Jensen et al., 2015) and lung cancer (Andriani et al., 2016). These new findings have developed the link between EMT and stemness into a bidirectional differentiation hierarchy, where a hybrid and highly plastic multipotent CSC can differentiate into both epithelial and mesenchymal lineages. This multipotent CSC possesses both therapeutic resistance and the ability to drive metastasis (Biddle et al., 2016, 2011; Ocana et al., 2012) and therefore presents an important target for future cancer therapeutic development. The transient, unstable nature of the highly plastic CSC phenotype is suggestive of the situation in ES cells where active signalling is required for maintenance of the undifferentiated phenotype (Festuccia et al., 2016). In this sense, these cells can be considered to be sitting on a knife edge; primed to differentiate at any moment. Indeed, the inner cell mass (the in vivo source of ES cells) is a transient cell state that quickly differentiates; maintenance of undifferentiated ES cells in vitro requires careful application of a defined signalling environment (Festuccia et al., 2016). A special ability of cancer cells to dedifferentiate back into the undifferentiated state (an ability conferred by mutation – see next section) enables maintenance of a small pool of highly plastic CSCs that can act to overcome bottlenecks in tumour progression. This transiency makes it difficult to study these highly plastic CSCs, but a new method using defined signalling mechanisms to enrich and maintain them in culture (analogous to methods used in ES cell culture) promises to aid both the study and therapeutic targeting of highly plastic CSCs (Biddle et al., 2016) (Fig. 3). Given its potential to produce large numbers of highly plastic CSCs in in vitro culture, this method may have great utility for therapeutic screening and enable reduced need for animal models in drug discovery efforts. One way that CSCs maintain plasticity is through their chromatin structure. Evidence suggests that CSCs may use bivalent chromatin – chromatin that is ‘poised’ in a relatively open structure but is not transcribed until the correct transcription factors are present

(Grosse-Wilde et al., 2015) and oral cancers (Clay et al., 2010). CD24, a glycoprotein cell surface adhesion molecule, was initially identified as a negative marker for CSCs in breast cancer (Al-Hajj et al., 2003), as is still often used in this context (Lee et al., 2016). It has however recently gained prominence as a positive CSC marker in lung cancer (Lau et al., 2014), colorectal cancer (Yeung et al., 2010), and triple negative breast cancer (Azzam et al., 2013). Interestingly, CD24 in combination with CD44 (CD24+CD44+) has been shown to mark a transient chemoresistant cell state in lung cancer (Sharma et al., 2010) and breast cancer (Goldman et al., 2015) – this transience suggests an unstable and highly plastic phenotype. Along similar lines, Grosse-Wilde et al. (2015) found in breast cancer that CD24+CD44+ CSCs were enhanced for both plasticity and mammosphere formation. This compares with the previously identified roles for a CD24+ multipotent stem cell in regeneration of normal intestinal and breast tissue (Buczacki et al., 2013; Stingl, 2009) (Fig. 2). It has previously been shown that inducing cancer cells into EMT can increase the number of CSCs in a tumour population (Mani et al., 2008), and also that cells that have undergone EMT display more therapeutic resistance than cells with an epithelial phenotype (Gupta et al., 2009). A link between sequential EMT-MET and metastasis has been hypothesised, driven partly by the fact that most secondary tumours display the epithelial phenotype of the primary tumour (Brabletz et al., 2005). Metastasising tumour cells could undergo EMT to migrate away from the primary tumour and into the circulation or lymphatics, migrate out of these vessels, into a new site, and then undergo MET to revert back to an epithelial phenotype for establishment of a secondary tumour (Reviewed in Biddle and Mackenzie (2012)). This hypothesis is supported by a study on breast cancer which showed that the invasive front of a tumour seemed to contain more mesenchymal-like (EMT) stem cells whereas the tumour interior contained more epithelial-like (MET) stem cells (Liu et al., 2014), and by the demonstration that EMT enhances metastatic seeding of lymph nodes by oral cancer CSCs (Biddle et al., 2011). In mouse models, it has been demonstrated that EMT is required for dissemination from the primary tumour and intra- and extravasation, whereas MET is required for tumour growth at metastatic sites (Ocana et al., 2012; Tsai et al., 2012). Additionally, conditioning of the metastatic site by CSCs in the EMT phase causes fibroblast activation which in turn promotes MET and metastatic growth (Del Pozo Martin et al., 2015). Interestingly, in some cases, drugs used to promote wound healing have been shown to increase metastasis (Leopold et al., 2012), further suggesting that EMT may be involved in metastasis. Moreover, many of the molecular mechanisms of EMT in wound healing are shared with EMT in cancer (Leopold et al., 2012) and developmental processes (Hudson et al., 2009). The EMT in re-epithelialisation and metastasis are however not identical as some factors, such as the transcription factor Snail, are not considered significant in re-epithelialisation (Leopold et al., 2012). Likely, pathological EMT is more akin to the complete EMT witnessed 4

Developmental Biology xxx (xxxx) xxx–xxx

C. O’Brien-Ball, A. Biddle

Fig. 3. CSC sub-populations in oral cancer, showing the relative sizes of different sub-populations under different conditions. Left – in vivo. Middle – in vitro with standard tissue culture medium in 2D culture. Right – in vitro enrichment of plastic EMT CSCs using plastic CSC maintenance medium in 2D culture (Biddle et al., 2016).

potential, and it is therefore unsurprising that no contribution to metastasis was observed as ability to undergo MET is essential to carcinoma metastasis (Tsai et al., 2012). Retention of plasticity appears to be a key consideration and, while it is yet uncertain whether EMT is strictly required for metastasis, it is nevertheless clear that some level of phenotypic plasticity is likely to be essential for successful metastasis.

(Bernstein et al., 2006; Chaffer et al., 2013). This is due to repressive and active histone marks (H3K27me3 and H3K4me3 respectively) being simultaneously present around target genes (Bernstein et al., 2006; Easwaran et al., 2014). Bivalent chromatin is unlike standard repressed chromatin (heterochromatin), which is essentially irreversible and generally imposed during development (Bernstein et al., 2006). It has been shown that genes strongly associated with EMT often exist in a bivalent chromatin state in cancer cells, thus enabling enhanced plasticity of gene expression (Ke et al., 2010). Differences in cell metabolism have also been demonstrated in CSCs. Interestingly, rather than demonstrating increased flexibility in metabolism, it has been suggested that CSCs in pancreatic cancer become somewhat ‘fixed’ in a single metabolic state and hence in this sense are less plastic than their differentiated progeny (Sancho et al., 2015). The same study revealed that pancreatic CSCs can be targeted by the anti-diabetic drug metformin, which inhibits mitochondrial function, because they are dependent on oxidative phosphorylation and have a reduced ability to switch to glycolysis (Sancho et al., 2015). A preference for oxidative phosphorylation and β-oxidation has also been demonstrated elsewhere (Viale et al., 2014). Conversely, other studies have shown that CSCs exhibit enhanced glycolysis matched with reduced mitochondrial respiration (Ciavardelli et al., 2014; Gammon et al., 2013). None of these studies looked specifically at the highly plastic CSC sub-population, and resolution of these apparent inconsistencies in our understanding of CSC metabolism will require a more detailed understanding of the different CSC sub-populations existing within a tumour and their relationships to one another. Finally, it is important to recognise that plasticity through EMT may not be strictly required for cancer cell migration and metastasis. Other possible methods of migration, which may not involve EMT, have been identified and this includes cohesive migration of epithelial cells which may be sufficient for lymphatic but not blood-borne metastasis (Giampieri et al., 2009). In 2015, two papers claimed that EMT did not play a role in metastasis of experimental mouse tumours (Fischer et al., 2015; Zheng et al., 2015). However, these studies were focused on the more differentiated EMT cells that have lost their multi-lineage

3. Cell reprogramming 3.1. Reprogramming in reproduction In normal function, epigenetic modifications to the genome are imposed mainly during development; in adult life, where these modifications mostly remain unchanged, somatic cells exist in unipotent lineages (Al-Hajj and Clarke, 2004). Normal stem cells governing tissue maintenance and regeneration are also lineage restricted, though often possess some plasticity of fate choice and may be considered to possess an epigenetic state much more restricted than the pluripotent cells of the early embryo but somewhat more free than their differentiated progeny (Halleux et al., 2001; Shenoy and Blelloch, 2014). As a cell becomes more differentiated, more epigenetic modifications are added to restrict its lineage potential and (usually) prevent it from de-differentiating (Halleux et al., 2001; Messerschmidt et al., 2014). One such modification is DNA methylation, a common epigenetic modification that leads to repression of gene expression. Future germ cells are reprogrammed through demethylation to specify the pluripotent germ cell lineage, then methylation is reinserted during re-differentiation into the specialised phenotypes required for fertilisation. Interestingly, the mature sperm genome in mice contains the highest methylation density in the body, but is globally demethylated after the formation of a zygote, with the new combined genome reaching the lowest methylation status at the early blastocyst stage (Messerschmidt et al., 2014). This rapid demethylation demonstrates how epigenetic reprogramming of highly specialised cells such as spermatozoa and oocytes enables production of a totipotent zygote. 5

Developmental Biology xxx (xxxx) xxx–xxx

C. O’Brien-Ball, A. Biddle

intermediate EMT state also imparts anoikis resistance (Grosse-Wilde et al., 2015; Huang et al., 2013), a key factor in successful metastasis. Thus, targeting plastic CSCs may be key to avoiding relapse after surgery. In addition to driving metastasis, it has been shown that phenotypic plasticity also produces drug resistance. Treatment with TGF-β and retinoic acid enabled enrichment of a population of highly plastic EMT CSCs in cell culture for therapeutic testing (Biddle et al., 2016). These cells were characterised by a CD44highEpCAMlow/−CD24+ cell surface marker profile and enhanced ability to undergo MET. Compared with phenotypically stable populations, they exhibited resistance to the DNA crosslinking agent cisplatin and the EMT-targeting drug salinomycin. These cells however were found to be sensitive to the endoplasmic reticulum stressor and autophagy inhibitor thapsigargin, suggesting that enhanced phenotypic plasticity confers sensitivity to this agent.

3.2. Reprogramming in cancer The normal reprogramming mechanisms can be hijacked by cancer cells in order to generate less differentiated, more stem-like cells (i.e. CSCs). This epigenetic reprogramming occurs in response to the effects of oncogenic mutations. Plasticity is strictly limited in normal adult tissue, whereas tumours possess a heterogeneous mixture of epigenetically different cells with unusually plastic phenotypes (Burrell et al., 2013; Ombrato and Malanchi, 2014), particularly within the CSC subpopulation. Recently, it was shown that in vitro transformation induced fibroblasts to spontaneously acquire CSC properties including multipotency (Scaffidi and Misteli, 2011). In addition, PIK3CA mutation has been demonstrated to induce reprogramming of lineagerestricted progenitors to a multipotent stem-like state in breast tumour initiation in vivo (Koren et al., 2015). The origin of CSCs is unclear, but these findings suggest that plastic CSCs may originate from epigenetic reprogramming of lineage restricted somatic cells, or possibly from the normal adult stem cell compartment (Li and Neaves, 2006). In contrast to normal tissue cells that retain epigenetic modifications imposed during development (Sharma et al., 2010), cancer cells are epigenetically unstable and less well differentiated. The jump from a lineage restricted phenotype to an undifferentiated stem cell phenotype thus becomes more likely. Evidence suggests that the EMT program involves global reprogramming of the epigenome (McDonald et al., 2011). Taking this further, Ombrato and Malanchi have suggested a link between partial EMT (or hybrid E/M phenotype) and cell ‘stemness’, hypothesising that the early EMT program results in epigenetic reprogramming and de-differentiation into a highly plastic stem cell phenotype (Ombrato and Malanchi, 2014). In line with this, it has been shown that the coexpression of epithelial and mesenchymal genes can predict poor outcome in breast cancer (Grosse-Wilde et al., 2015) and ovarian cancer (Huang et al., 2013). Partial EMT, with enhanced plasticity, also imparts therapeutic resistance (Biddle et al., 2016). Therefore, the mechanisms of epigenetic reprogramming could be a crucial target for future cancer therapies.

4.2. Metabolic changes One proposed mechanism of resistance against metastasis-induced cell stress is through changes in cell metabolism. A study in melanoma by Piskounova et al. (2015) determined that the levels of oxidative stress and ROS are much higher in circulating and metastatic tumour cells and that treatment with an antioxidant increased the number of circulating tumour cells and the metastatic disease burden. Successful metastasis required metabolic adaptation to combat oxidative stress, including increased production of NADPH through the folate pathway (Piskounova et al., 2015). The detoxifying enzyme ALDH1 was identified as a key pathway component, and interestingly is also a common CSC marker (Clay et al., 2010; Grosse-Wilde et al., 2015). Targeting these enzymes could perhaps reduce the metastatic burden from plastic CSCs. In a mouse model of squamous cell carcinoma, perivascular TGF-β signalling induced an EMT CSC phenotype with enhanced metastatic capability and resistance to cisplatin-induced apoptosis (Oshimori et al., 2015). Cisplatin resistance in this model was driven by enhanced glutathione metabolism, which promotes inactivation of cisplatin through a conjugation reaction (Kelland, 2007). This metabolic change could therefore be responsible for the cisplatin resistance observed in highly plastic CSCs.

4. Implications for cancer therapy A normal cell must constantly battle against insults from the environment, both internal and external, such as oxidative stress and nutrient deprivation. In cancer, malignant cells must tolerate these stresses in addition to others such as a hypoxia that come with tumour formation (Kreso and Dick, 2014). Additionally, cells that are able to effectively metastasise must be able to cope with the stresses that come from migration and travel through the body to colonise a distant area, recapitulating the migration of cells in normal development. In leaving the primary tumour, they must adapt to the new microenvironment and also, importantly, survive in circulation. The circulatory system is especially hostile to cancer cell survival and thus a key barrier to metastasis (Piskounova et al., 2015). Moreover, CSCs may be exposed to therapeutics in the form of drugs, radiation and surgery. Such mechanisms place a selective pressure on tumours and, in line with the two major models of tumour evolution, may either drive Darwinian selection of genetically resistant clones or induce reversible epigenetic changes in the cell population leading to a rise in therapeutic resistance. It is possible that, since CSCs often ‘reuse’ mechanisms initially required in development (such as EMT), they may adopt a developmental state that is more resistant to cellular stress.

4.3. Quiescence The ability to adopt a quiescent state is a hallmark of plastic CSCs and is often responsible for disease relapse - a patient that is considered disease-free may still harbour thousands of disseminated cells, particularly in organs such as the bone marrow (Braun et al., 2005) where is it currently impossible to remove them. CSCs with the ability to form secondary tumours have been shown to downregulate natural killer cell activator molecules when they transform into a quiescent state, through a mechanism involving autocrine inhibition of Wnt/β-catenin signalling by DKK1, thus avoiding death by cytolysis (Malladi et al., 2016). This finding could also explain why highly plastic CSCs are resistant to salinomycin, as this drug targets EMT cells through inhibition of Wnt/β-catenin signalling (Lu et al., 2011). Radiotherapy and standard chemotherapeutics preferentially target tumour cells due to their high replication rate (Liu et al., 2015). They arrest cell division and can cause apoptosis, cell death by other mechanisms, or cellular senescence (Eriksson and Stigbrand, 2010). Such treatments can be highly effective at targeting the bulk of the tumour population but the few quiescent, slowly (or non-) dividing CSCs are resistant and could subsequently repopulate the tumour leading to relapse (Liu et al., 2015). Thus, quiescence of CSCs can lead to therapeutic resistance. Interestingly, given the finding that thapsigargin targets highly plastic CSCs (Biddle et al., 2016), it has previously been shown that thapsigargin induces apoptosis in a proliferationindependent fashion and can thus effectively target quiescent cells

4.1. Phenotypic plasticity and EMT The primary method by which tumour cells can evade surgical intervention is by migrating away from the primary site (Biddle et al., 2011), where they may utilise plasticity through EMT and MET to travel to and colonise distant tissues (Thiery, 2002). A highly plastic 6

Developmental Biology xxx (xxxx) xxx–xxx

C. O’Brien-Ball, A. Biddle

the replacement of animal models in investigations of the developmental mechanisms driving human tumour progression.

(Pinski et al., 2001). Thapsigargin is not selective for cancer cells, but recent efforts to modify it as a tumour-targeted pro-drug have greatly improved the toxicity profile and it has now progressed into phase II trials (Mahalingam et al., 2016). When combined with standard cytotoxic agents, it may prove an effective CSC-targeted drug.

6. Conclusion Whilst the underlying genetic mutations, cell of origin and microenvironment to a large extent determine the behaviour of a tumour, it is increasingly apparent that this behaviour is driven by the reactivation of developmental mechanisms. Through understanding mechanisms of normal development, we can thus gain valuable insight into the key processes driving tumour development. In particular, future identification of plastic stem cell phenotypes in development and tissue regeneration, together with their cell surface marker profiles, will serve to inform studies of plasticity in cancer. Likewise, the huge research effort currently targeted at understanding tumour development may also yield valuable insights into normal human development. An interesting outstanding question relates to the role of CD24 as a marker of plastic stem cells in development, regeneration and cancer. CD24+ plastic stem cells have been identified as playing important roles in both regeneration and cancer (Biddle et al., 2016; Buczacki et al., 2013; Grosse-Wilde et al., 2015; Stingl, 2009), and this suggests the exciting possibility of an as yet unidentified CD24+ plastic stem cell in development (Fig. 2). Some research suggests the possibility of such a cell type in development of the central nervous system where CD24 was found to be transiently expressed during neuronal development, and re-expressed on some neuronal tumours, especially the less well-differentiated types (Poncet et al., 1996). Such a stem cell, with the ability to migrate, survive in hostile environments, and undergo multi-lineage differentiation, could be hypothesised to play a crucial role in both embryonic development and cancer.

5. Future outlook Identification of more specific markers will be very important for the detailed investigation of CSC sub-populations - distinction of relatively subtle phenotypic differences such as partial EMT and full EMT will likely require equal subtlety in marker profile. Cell surface proteins alone may be insufficient for this purpose; intracellular proteins and RNA, or complex glycosylation modifications of cell surface proteins, may be essential to aid accurate identification of CSC sub-populations. Recent technological innovations (e.g. SmartFlare) will help with this. However, even with the emergence of more sophisticated cellular markers, it is possible that not all of the important CSC phenotypes will be discrete sub-populations with totally separate behaviours and traits – rather, some may be transient states that become stabilised in response to particular environments or therapeutic interventions. Working to understand how phenotypic plasticity is regulated under the control of the tumour environment, metastasis, and during the application of therapeutics, will be key to developing future treatments with higher success than those currently available. Additionally, we need to understand the underlying genetic and epigenetic properties that allow tumours to adopt phenotypic plasticity and thus become resistant to therapeutics. Identifying key gene expression changes, epigenetic modifications, and mutations (and understanding how these change cell behaviour) could help predict how a tumour will progress, and hence how best to target it. Understanding the biology and molecular drivers of highly plastic CSCs will enable identification of their Achilles heel, through which these cells can be directly targeted with drugs that either kill them directly or sensitize them to standard therapies. Progress has already been made on this front. Huang et al. (2013) identified an intermediate mesenchymal phenotype of cells characterised by high N-cadherin and ZEB1 expression, which were spheroidogenic and resistant to anoikis. Treatment of these cells with a Src kinase inhibitor forced the cells to produce E-cadherin, an epithelial marker, and reduced spheriodogenesis (Huang et al., 2013). The identification of Thapsigargin as an agent that targets highly plastic CSCs is also a promising development in this area (Biddle et al., 2016), although the mechanism through which Thapsigargin specifically targets highly plastic CSCs remains to be determined. However, these methods have their limitations; targeting any individual molecule, such as Src kinase, will likely lead to rapid adaptive resistance. In addition, Src may only be required by the highly plastic CSCs in a sub-set of tumours with particular genetic and epigenetic backgrounds. Thapsigargin, on the other hand, exhibits broad non-specific toxicity that will likely limit its potential as a cancer drug. If developmental mechanisms could be identified that drive tumour growth and spread whilst being silent in normal adult tissue, then these could represent powerful targets for cancer therapeutics. New human developmental models will greatly aid our understanding of developmental mechanisms in both normal human development and cancer. The booming field of organoid culture, in which miniature human organs are grown from ES cells or tissue-specific stem cells in vitro, has already begun to enhance our understanding of how human organs such as brain (Lancaster et al., 2013) and liver (Huch et al., 2015) develop. This same technology is also being applied to cancer, so that human tumour development can be observed in vitro (Cheung et al., 2013). This technology will likely prove a vital tool for the dissection of developmental mechanisms that may drive human cancer. Given their greater experimental tractability and recapitulation of human biology, these organoid models also have great potential for

Acknowledgements Adrian Biddle is supported by Animal Free Research UK, as part of the Animal Replacement Centre of Excellence at Queen Mary University of London. The work from our laboratory discussed in this review was supported by the National Centre for the 3Rs (NC3Rs) through Grant G0900799/1 and Fellowship NC/K500495/1. References Al-Hajj, M., Clarke, M.F., 2004. Self-renewal and solid tumor stem cells. Oncogene 23, 7274–7282. Al-Hajj, M., Wicha, M.S., Benito-Hernandez, A., Morrison, S.J., Clarke, M.F., 2003. Prospective identification of tumorigenic breast cancer cells. Proc. Natl. Acad. Sci. USA 100, 3983–3988. Andriani, F., Bertolini, G., Facchinetti, F., Baldoli, E., Moro, M., Casalini, P., Caserini, R., Milione, M., Leone, G., Pelosi, G., et al., 2016. Conversion to stem-cell state in response to microenvironmental cues is regulated by balance between epithelial and mesenchymal features in lung cancer cells. Mol. Oncol. 10, 253–271. Asselin-Labat, M.L., Vaillant, F., Sheridan, J.M., Pal, B., Wu, D., Simpson, E.R., Yasuda, H., Smyth, G.K., Martin, T.J., Lindeman, G.J., Visvader, J.E., 2010. Control of mammary stem cell function by steroid hormone signalling. Nature 465, 798–802. Azzam, D.J., Zhao, D., Sun, J., Minn, A.J., Ranganathan, P., Drews-Elger, K., Han, X., Picon-Ruiz, M., Gilbert, C.A., Wander, S.A., et al., 2013. Triple negative breast cancer initiating cell subsets differ in functional and molecular characteristics and in γsecretase inhibitor drug responses. EMBO Mol. Med. 5, 1502–1522. Bao, S., Wu, Q., McLendon, R.E., Hao, Y., Shi, Q., Hjelmeland, A.B., Dewhirst, M.W., Bigner, D.D., Rich, J.N., 2006. Glioma stem cells promote radioresistance by preferential activation of the DNA damage response. Nature 444, 756–760. Barker, N., van Es, J.H., Kuipers, J., Kujala, P., van den Born, M., Cozijnsen, M., Haegebarth, A., Korving, J., Begthel, H., Peters, P.J., Clevers, H., 2007. Identification of stem cells in small intestine and colon by marker gene Lgr5. Nature 449, 1003–1007. Battula, V.L., Evans, K.W., Hollier, B.G., Shi, Y., Marini, F.C., Ayyanan, A., Wang, R.Y., Brisken, C., Guerra, R., Andreeff, M., Mani, S.A., 2010. Epithelial-mesenchymal transition-derived cells exhibit multilineage differentiation potential similar to mesenchymal stem cells. Stem Cells 28, 1435–1445. Bernstein, B.E., Mikkelsen, T.S., Xie, X., Kamal, M., Huebert, D.J., Cuff, J., Fry, B., Meissner, A., Wernig, M., Plath, K., et al., 2006. A bivalent chromatin structure marks key developmental genes in embryonic stem cells. Cell 125, 315–326. Biddle, A., Gammon, L., Liang, X., Costea, D.E., Mackenzie, I.C., 2016. Phenotypic

7

Developmental Biology xxx (xxxx) xxx–xxx

C. O’Brien-Ball, A. Biddle

Kuestner, R.E., del Sol, A., Walters, K.A., Huang, S., 2015. Stemness of the hybrid epithelial/mesenchymal state in breast cancer and its association with poor survival. PLoS One 10, e0126522. Gupta, P.B., Onder, T.T., Jiang, G., Tao, K., Kuperwasser, C., Weinberg, R.A., Lander, E.S., 2009. Identification of selective inhibitors of cancer stem cells by highthroughput screening. Cell 138, 645–659. Halleux, C., Sottile, V., Gasser, J.A., Seuwen, K., 2001. Multi-lineage potential of human mesenchymal stem cells following clonal expansion. J. Musculoskelet. Neuron. Interact. 2, 71–76. Hay, E.D., 2005. The mesenchymal cell, its role in the embryo, and the remarkable signaling mechanisms that create it. Dev. Dyn. 233, 706–720. Heppner, G.H., 1984. Tumor heterogeneity. Cancer Res. 44, 2259–2265. Hermann, P.C., Huber, S.L., Herrler, T., Aicher, A., Ellwart, J.W., Guba, M., Bruns, C.J., Heeschen, C., 2007. Distinct populations of cancer stem cells determine tumor growth and metastatic activity in human pancreatic cancer. Cell Stem Cell 1, 313–323. Huang, R.Y., Wong, M.K., Tan, T.Z., Kuay, K.T., Ng, A.H., Chung, V.Y., Chu, Y.S., Matsumura, N., Lai, H.C., Lee, Y.F., et al., 2013. An EMT spectrum defines an anoikis-resistant and spheroidogenic intermediate mesenchymal state that is sensitive to e-cadherin restoration by a src-kinase inhibitor, saracatinib (AZD0530). Cell Death Dis. 4, e915. Huch, M., Gehart, H., van Boxtel, R., Hamer, K., Blokzijl, F., Verstegen, M.M., Ellis, E., van Wenum, M., Fuchs, S.A., de Ligt, J., et al., 2015. Long-term culture of genomestable bipotent stem cells from adult human liver. Cell 160, 299–312. Hudson, L.G., Newkirk, K.M., Chandler, H.L., Choi, C., Fossey, S.L., Parent, A.E., Kusewitt, D.F., 2009. Cutaneous wound reepithelialization is compromised in mice lacking functional Slug (Snai2). J. Dermatol. Sci. 56, 19–26. Jensen, D.H., Dabelsteen, E., Specht, L., Fiehn, A.M., Therkildsen, M.H., Jønson, L., Vikesaa, J., Nielsen, F.C., von Buchwald, C., 2015. Molecular profiling of tumour budding implicates TGFβ-mediated epithelial-mesenchymal transition as a therapeutic target in oral squamous cell carcinoma. J. Pathol. 236, 505–516. Jolly, M.K., Boareto, M., Huang, B., Jia, D., Lu, M., Ben-Jacob, E., Onuchic, J.N., Levine, H., 2015. Implications of the hybrid epithelial/mesenchymal phenotype in metastasis. Front. Oncol. 5, 155. Kalcheim, C., 2015. Epithelial-mesenchymal transitions during neural crest and somite development. J. Clin. Med., 5. Ke, X.S., Qu, Y., Cheng, Y., Li, W.C., Rotter, V., Øyan, A.M., Kalland, K.H., 2010. Global profiling of histone and DNA methylation reveals epigenetic-based regulation of gene expression during epithelial to mesenchymal transition in prostate cells. BMC Genom. 11, 669. Kelland, L., 2007. The resurgence of platinum-based cancer chemotherapy. Nat. Rev. Cancer 7, 573–584. Koren, S., Reavie, L., Couto, J.P., De Silva, D., Stadler, M.B., Roloff, T., Britschgi, A., Eichlisberger, T., Kohler, H., Aina, O., et al., 2015. PIK3CA(H1047R) induces multipotency and multi-lineage mammary tumours. Nature 525, 114–118. Kreso, A., Dick, J.E., 2014. Evolution of the cancer stem cell model. Cell Stem Cell 14, 275–291. Kreso, A., O'Brien, C.A., van Galen, P., Gan, O.I., Notta, F., Brown, A.M., Ng, K., Ma, J., Wienholds, E., Dunant, C., et al., 2013. Variable clonal repopulation dynamics influence chemotherapy response in colorectal cancer. Science 339, 543–548. Lancaster, M.A., Renner, M., Martin, C.A., Wenzel, D., Bicknell, L.S., Hurles, M.E., Homfray, T., Penninger, J.M., Jackson, A.P., Knoblich, J.A., 2013. Cerebral organoids model human brain development and microcephaly. Nature 501, 373–379. Lau, A.N., Curtis, S.J., Fillmore, C.M., Rowbotham, S.P., Mohseni, M., Wagner, D.E., Beede, A.M., Montoro, D.T., Sinkevicius, K.W., Walton, Z.E., et al., 2014. Tumorpropagating cells and Yap/Taz activity contribute to lung tumor progression and metastasis. EMBO J. 33, 468–481. Lee, G., Hall, R.R., Ahmed, A.U., 2016a. Cancer stem cells: cellular plasticity, niche, and its clinical relevance. J. Stem Cell Res. Ther., 6. Lee, K.M., Lee, M., Lee, J., Kim, S.W., Moon, H.G., Noh, D.Y., Han, W., 2016b. Enhanced anti-tumor activity and cytotoxic effect on cancer stem cell population of metforminbutyrate compared with metformin HCl in breast cancer. Oncotarget 7, 38500–38512. Lei, J., Levin, S.A., Nie, Q., 2014. Mathematical model of adult stem cell regeneration with cross-talk between genetic and epigenetic regulation. Proc. Natl. Acad. Sci. USA 111, E880–E887. Leopold, P.L., Vincent, J., Wang, H., 2012. A comparison of epithelial-to-mesenchymal transition and re-epithelialization. Semin. Cancer Biol. 22, 471–483. Li, L., Neaves, W.B., 2006. Normal stem cells and cancer stem cells: the niche matters. Cancer Res. 66, 4553–4557. Li, X., Lewis, M.T., Huang, J., Gutierrez, C., Osborne, C.K., Wu, M.F., Hilsenbeck, S.G., Pavlick, A., Zhang, X., Chamness, G.C., et al., 2008. Intrinsic resistance of tumorigenic breast cancer cells to chemotherapy. J. Natl. Cancer Inst. 100, 672–679. Lim, J., Thiery, J.P., 2012. Epithelial-mesenchymal transitions: insights from development. Development 139, 3471–3486. Liu, H., Lv, L., Yang, K., 2015. Chemotherapy targeting cancer stem cells. Am. J. Cancer Res. 5, 880–893. Liu, S., Cong, Y., Wang, D., Sun, Y., Deng, L., Liu, Y., Martin-Trevino, R., Shang, L., McDermott, S.P., Landis, M.D., et al., 2014. Breast cancer stem cells transition between epithelial and mesenchymal states reflective of their normal counterparts. Stem Cell Rep. 2, 78–91. Locke, M., Heywood, M., Fawell, S., Mackenzie, I.C., 2005. Retention of intrinsic stem cell hierarchies in carcinoma-derived cell lines. Cancer Res. 65, 8944–8950. Lu, D., Choi, M.Y., Yu, J., Castro, J.E., Kipps, T.J., Carson, D.A., 2011. Salinomycin inhibits Wnt signaling and selectively induces apoptosis in chronic lymphocytic

plasticity determines cancer stem cell therapeutic resistance in oral squamous cell carcinoma. EBioMedicine 4, 138–145. Biddle, A., Liang, X., Gammon, L., Fazil, B., Harper, L.J., Emich, H., Costea, D.E., Mackenzie, I.C., 2011. Cancer stem cells in squamous cell carcinoma switch between two distinct phenotypes that are preferentially migratory or proliferative. Cancer Res. 71, 5317–5326. Biddle, A., Mackenzie, I.C., 2012. Cancer stem cells and EMT in carcinoma. Cancer Metastas. Rev.. Brabletz, T., Jung, A., Spaderna, S., Hlubek, F., Kirchner, T., 2005. Opinion: migrating cancer stem cells - an integrated concept of malignant tumour progression. Nat. Rev. Cancer 5, 744–749. Braun, S., Vogl, F.D., Naume, B., Janni, W., Osborne, M.P., Coombes, R.C., Schlimok, G., Diel, I.J., Gerber, B., Gebauer, G., et al., 2005. A pooled analysis of bone marrow micrometastasis in breast cancer. N. Engl. J. Med. 353, 793–802. Buczacki, S.J., Zecchini, H.I., Nicholson, A.M., Russell, R., Vermeulen, L., Kemp, R., Winton, D.J., 2013. Intestinal label-retaining cells are secretory precursors expressing Lgr5. Nature 495, 65–69. Burrell, R.A., McGranahan, N., Bartek, J., Swanton, C., 2013. The causes and consequences of genetic heterogeneity in cancer evolution. Nature 501, 338–345. Cai, S., Kalisky, T., Sahoo, D., Dalerba, P., Feng, W., Lin, Y., Qian, D., Kong, A., Yu, J., Wang, F., et al., 2017. A quiescent Bcl11b high stem cell population is required for maintenance of the mammary gland. Cell Stem Cell 20, 247–260, (e245). Cantone, I., Fisher, A.G., 2013. Epigenetic programming and reprogramming during development. Nat. Struct. Mol. Biol. 20, 282–289. Chaffer, C.L., Brueckmann, I., Scheel, C., Kaestli, A.J., Wiggins, P.A., Rodrigues, L.O., Brooks, M., Reinhardt, F., Su, Y., Polyak, K., et al., 2011. Normal and neoplastic nonstem cells can spontaneously convert to a stem-like state. Proc. Natl. Acad. Sci. USA 108, 7950–7955. Chaffer, C.L., Marjanovic, N.D., Lee, T., Bell, G., Kleer, C.G., Reinhardt, F., D'Alessio, A.C., Young, R.A., Weinberg, R.A., 2013. Poised chromatin at the ZEB1 promoter enables breast cancer cell plasticity and enhances tumorigenicity. Cell 154, 61–74. Cheung, K.J., Gabrielson, E., Werb, Z., Ewald, A.J., 2013. Collective invasion in breast cancer requires a conserved basal epithelial program. Cell 155, 1639–1651. Christen, B., Robles, V., Raya, M., Paramonov, I., Izpisúa Belmonte, J.C., 2010. Regeneration and reprogramming compared. BMC Biol. 8, 5. Ciavardelli, D., Rossi, C., Barcaroli, D., Volpe, S., Consalvo, A., Zucchelli, M., De Cola, A., Scavo, E., Carollo, R., D'Agostino, D., et al., 2014. Breast cancer stem cells rely on fermentative glycolysis and are sensitive to 2-deoxyglucose treatment. Cell Death Dis. 5, e1336. Clarke, M.F., Dick, J.E., Dirks, P.B., Eaves, C.J., Jamieson, C.H., Jones, D.L., Visvader, J., Weissman, I.L., Wahl, G.M., 2006. Cancer stem cells–perspectives on current status and future directions: AACR Workshop on cancer stem cells. Cancer Res. 66, 9339–9344. Clay, M.R., Tabor, M., Owen, J.H., Carey, T.E., Bradford, C.R., Wolf, G.T., Wicha, M.S., Prince, M.E., 2010. Single-marker identification of head and neck squamous cell carcinoma cancer stem cells with aldehyde dehydrogenase. Head Neck 32, 1195–1201. Dalerba, P., Cho, R.W., Clarke, M.F., 2007. Cancer stem cells: models and concepts. Annu. Rev. Med. 58, 267–284. Dave, N., Guaita-Esteruelas, S., Gutarra, S., Frias, À., Beltran, M., Peiró, S., de Herreros, A.G., 2011. Functional cooperation between Snail1 and twist in the regulation of ZEB1 expression during epithelial to mesenchymal transition. J. Biol. Chem. 286, 12024–12032. Del Pozo Martin, Y., Park, D., Ramachandran, A., Ombrato, L., Calvo, F., Chakravarty, P., Spencer-Dene, B., Derzsi, S., Hill, C.S., Sahai, E., Malanchi, I., 2015. Mesenchymal cancer cell-stroma crosstalk promotes niche activation, epithelial reversion, and metastatic colonization. Cell Rep. 13, 2456–2469. Easwaran, H., Tsai, H.C., Baylin, S.B., 2014. Cancer epigenetics: tumor heterogeneity, plasticity of stem-like states, and drug resistance. Mol. Cell 54, 716–727. Eriksson, D., Stigbrand, T., 2010. Radiation-induced cell death mechanisms. Tumour Biol. 31, 363–372. Festuccia, N., Gonzalez, I., Navarro, P., 2016. The epigenetic paradox of pluripotent ES cells. J. Mol. Biol.. Fischer, K.R., Durrans, A., Lee, S., Sheng, J., Li, F., Wong, S.T., Choi, H., El Rayes, T., Ryu, S., Troeger, J., et al., 2015. Epithelial-to-mesenchymal transition is not required for lung metastasis but contributes to chemoresistance. Nature 527, 472–476. Fisher, R., Pusztai, L., Swanton, C., 2013. Cancer heterogeneity: implications for targeted therapeutics. Br. J. Cancer 108, 479–485. Gammon, L., Biddle, A., Heywood, H.K., Johannessen, A.C., Mackenzie, I.C., 2013. Subsets of cancer stem cells differ intrinsically in their patterns of oxygen metabolism. PLoS One 8, e62493. Giampieri, S., Manning, C., Hooper, S., Jones, L., Hill, C.S., Sahai, E., 2009. Localized and reversible TGFbeta signalling switches breast cancer cells from cohesive to single cell motility. Nat. Cell Biol. 11, 1287–1296. Gjerdrum, C., Tiron, C., Hoiby, T., Stefansson, I., Haugen, H., Sandal, T., Collett, K., Li, S., McCormack, E., Gjertsen, B.T., et al., 2010. Axl is an essential epithelial-tomesenchymal transition-induced regulator of breast cancer metastasis and patient survival. Proc. Natl. Acad. Sci. USA 107, 1124–1129. Goldman, A., Majumder, B., Dhawan, A., Ravi, S., Goldman, D., Kohandel, M., Majumder, P.K., Sengupta, S., 2015. Temporally sequenced anticancer drugs overcome adaptive resistance by targeting a vulnerable chemotherapy-induced phenotypic transition. Nat. Commun. 6, 6139. Greaves, M., 2015. Evolutionary determinants of cancer. Cancer Discov. 5, 806–820. Green, S.A., Simoes-Costa, M., Bronner, M.E., 2015. Evolution of vertebrates as viewed from the crest. Nature 520, 474–482. Grosse-Wilde, A., Fouquier d'Herouel, A., McIntosh, E., Ertaylan, G., Skupin, A.,

8

Developmental Biology xxx (xxxx) xxx–xxx

C. O’Brien-Ball, A. Biddle

Sancho, P., Burgos-Ramos, E., Tavera, A., Bou Kheir, T., Jagust, P., Schoenhals, M., Barneda, D., Sellers, K., Campos-Olivas, R., Graña, O., et al., 2015. MYC/PGC-1α balance determines the metabolic phenotype and plasticity of pancreatic cancer stem cells. Cell Metab. 22, 590–605. Scaffidi, P., Misteli, T., 2011. In vitro generation of human cells with cancer stem cell properties. Nat. Cell Biol. 13, 1051–1061. Shackleton, M., Vaillant, F., Simpson, K.J., Stingl, J., Smyth, G.K., Asselin-Labat, M.L., Wu, L., Lindeman, G.J., Visvader, J.E., 2006. Generation of a functional mammary gland from a single stem cell. Nature 439, 84–88. Sharma, S., Kelly, T.K., Jones, P.A., 2010a. Epigenetics in cancer. Carcinogenesis 31, 27–36. Sharma, S.V., Lee, D.Y., Li, B., Quinlan, M.P., Takahashi, F., Maheswaran, S., McDermott, U., Azizian, N., Zou, L., Fischbach, M.A., et al., 2010b. A chromatinmediated reversible drug-tolerant state in cancer cell subpopulations. Cell 141, 69–80. Shenoy, A., Blelloch, R.H., 2014. Regulation of microRNA function in somatic stem cell proliferation and differentiation. Nat. Rev. Mol. Cell Biol. 15, 565–576. Sottoriva, A., Kang, H., Ma, Z., Graham, T.A., Salomon, M.P., Zhao, J., Marjoram, P., Siegmund, K., Press, M.F., Shibata, D., Curtis, C., 2015. A Big Bang model of human colorectal tumor growth. Nat. Genet. 47, 209–216. Stingl, J., 2009. Detection and analysis of mammary gland stem cells. J. Pathol. 217, 229–241. Takahashi, K., Yamanaka, S., 2006. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126, 663–676. Tam, W.L., Weinberg, R.A., 2013. The epigenetics of epithelial-mesenchymal plasticity in cancer. Nat. Med. 19, 1438–1449. Teperek, M., Simeone, A., Gaggioli, V., Miyamoto, K., Allen, G.E., Erkek, S., Kwon, T., Marcotte, E.M., Zegerman, P., Bradshaw, C.R., et al., 2016. Sperm is epigenetically programmed to regulate gene transcription in embryos. Genome Res. 26, 1034–1046. Thiery, J.P., 2002. Epithelial-mesenchymal transitions in tumour progression. Nat. Rev. Cancer 2, 442–454. Thiery, J.P., Sleeman, J.P., 2006. Complex networks orchestrate epithelial-mesenchymal transitions. Nat. Rev. Mol. Cell Biol. 7, 131–142. Tsai, J.H., Donaher, J.L., Murphy, D.A., Chau, S., Yang, J., 2012. Spatiotemporal regulation of epithelial-mesenchymal transition is essential for squamous cell carcinoma metastasis. Cancer Cell 22, 725–736. Viale, A., Pettazzoni, P., Lyssiotis, C.A., Ying, H., Sánchez, N., Marchesini, M., Carugo, A., Green, T., Seth, S., Giuliani, V., et al., 2014. Oncogene ablation-resistant pancreatic cancer cells depend on mitochondrial function. Nature 514, 628–632. Vig, N., Mackenzie, I.C., Biddle, A., 2015. Phenotypic plasticity and epithelial-tomesenchymal transition in the behaviour and therapeutic response of oral squamous cell carcinoma. J. Oral. Pathol. Med. 44, 649–655. Visvader, J.E., Stingl, J., 2014. Mammary stem cells and the differentiation hierarchy: current status and perspectives. Genes Dev. 28, 1143–1158. Waddington, C.H., 1957. The Strategy of the Genes. A Discussion with Some Aspects of Theoretical Biologys. George Allen & Unwin, Ltd., London. Wheelock, M.J., Shintani, Y., Maeda, M., Fukumoto, Y., Johnson, K.R., 2008. Cadherin switching. J. Cell Sci. 121, 727–735. Williams, M.J., Werner, B., Barnes, C.P., Graham, T.A., Sottoriva, A., 2016. Identification of neutral tumor evolution across cancer types. Nat. Genet. 48, 238–244. Xu, J., Lamouille, S., Derynck, R., 2009. TGF-beta-induced epithelial to mesenchymal transition. Cell Res. 19, 156–172. Yang, M.H., Wu, M.Z., Chiou, S.H., Chen, P.M., Chang, S.Y., Liu, C.J., Teng, S.C., Wu, K.J., 2008. Direct regulation of TWIST by HIF-1alpha promotes metastasis. Nat. Cell Biol. 10, 295–305. Yeung, T.M., Gandhi, S.C., Wilding, J.L., Muschel, R., Bodmer, W.F., 2010. Cancer stem cells from colorectal cancer-derived cell lines. Proc. Natl. Acad. Sci. USA 107, 3722–3727. Zheng, X., Carstens, J.L., Kim, J., Scheible, M., Kaye, J., Sugimoto, H., Wu, C.C., LeBleu, V.S., Kalluri, R., 2015. Epithelial-to-mesenchymal transition is dispensable for metastasis but induces chemoresistance in pancreatic cancer. Nature 527, 525–530.

leukemia cells. Proc. Natl. Acad. Sci. USA 108, 13253–13257. Mahalingam, D., Wilding, G., Denmeade, S., Sarantopoulas, J., Cosgrove, D., Cetnar, J., Azad, N., Bruce, J., Kurman, M., Allgood, V.E., Carducci, M., 2016. Mipsagargin, a novel thapsigargin-based PSMA-activated prodrug: results of a first-in-man phase I clinical trial in patients with refractory, advanced or metastatic solid tumours. Br. J. Cancer 114, 986–994. Malladi, S., Macalinao, D.G., Jin, X., He, L., Basnet, H., Zou, Y., de Stanchina, E., Massague, J., 2016. Metastatic latency and immune evasion through autocrine inhibition of WNT. Cell 165, 45–60. 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. McDonald, O.G., Wu, H., Timp, W., Doi, A., Feinberg, A.P., 2011. Genome-scale epigenetic reprogramming during epithelial-to-mesenchymal transition. Nat. Struct. Mol. Biol. 18, 867–874. Medema, J.P., 2013. Cancer stem cells: the challenges ahead. Nat. Cell Biol. 15, 338–344. Messerschmidt, D.M., Knowles, B.B., Solter, D., 2014. DNA methylation dynamics during epigenetic reprogramming in the germline and preimplantation embryos. Genes Dev. 28, 812–828. Morgani, S.M., Brickman, J.M., 2014. The molecular underpinnings of totipotency. Philos. Trans. R. Soc. Lond. B Biol. Sci., 369. Morrison, S.J., White, P.M., Zock, C., Anderson, D.J., 1999. Prospective identification, isolation by flow cytometry, and in vivo self-renewal of multipotent mammalian neural crest stem cells. Cell 96, 737–749. Moss, F.P., Leblond, C.P., 1970. Nature of dividing nuclei in skeletal muscle of growing rats. J. Cell Biol. 44, 459–462. Nishikawa, S., Konno, M., Hamabe, A., Hasegawa, S., Kano, Y., Fukusumi, T., Satoh, T., Takiguchi, S., Mori, M., Doki, Y., Ishii, H., 2015. Surgically resected human tumors reveal the biological significance of the gastric cancer stem cell markers CD44 and CD26. Oncol. Lett. 9, 2361–2367. Ocana, O.H., Corcoles, R., Fabra, A., Moreno-Bueno, G., Acloque, H., Vega, S., BarralloGimeno, A., Cano, A., Nieto, M.A., 2012. Metastatic colonization requires the repression of the epithelial-mesenchymal transition inducer Prrx1. Cancer Cell 22, 709–724. Ombrato, L., Malanchi, I., 2014. The EMT universe: space between cancer cell dissemination and metastasis initiation. Crit. Rev. Oncog. 19, 349–361. Oshimori, N., Oristian, D., Fuchs, E., 2015. TGF-β promotes heterogeneity and drug resistance in squamous cell carcinoma. Cell 160, 963–976. Patrawala, L., Calhoun, T., Schneider-Broussard, R., Li, H., Bhatia, B., Tang, S., Reilly, J.G., Chandra, D., Zhou, J., Claypool, K., et al., 2006. Highly purified CD44+ prostate cancer cells from xenograft human tumors are enriched in tumorigenic and metastatic progenitor cells. Oncogene 25, 1696–1708. Pinski, J., Parikh, A., Bova, G.S., Isaacs, J.T., 2001. Therapeutic implications of enhanced G(0)/G(1) checkpoint control induced by coculture of prostate cancer cells with osteoblasts. Cancer Res. 61, 6372–6376. Piskounova, E., Agathocleous, M., Murphy, M.M., Hu, Z., Huddlestun, S.E., Zhao, Z., Leitch, A.M., Johnson, T.M., DeBerardinis, R.J., Morrison, S.J., 2015. Oxidative stress inhibits distant metastasis by human melanoma cells. Nature 527, 186–191. Poncet, C., Frances, V., Gristina, R., Scheiner, C., Pellissier, J.F., Figarella-Branger, D., 1996. CD24, a glycosylphosphatidylinositol-anchored molecules is transiently expressed during the development of human central nervous system and is a marker of human neural cell lineage tumors. Acta Neuropathol. 91, 400–408. Riehle, K.J., Dan, Y.Y., Campbell, J.S., Fausto, N., 2011. New concepts in liver regeneration. J. Gastroenterol. Hepatol. 26 (Suppl 1), 203–212. Rios, A.C., Fu, N.Y., Lindeman, G.J., Visvader, J.E., 2014. In situ identification of bipotent stem cells in the mammary gland. Nature 506, 322–327. Roesch, A., Fukunaga-Kalabis, M., Schmidt, E.C., Zabierowski, S.E., Brafford, P.A., Vultur, A., Basu, D., Gimotty, P., Vogt, T., Herlyn, M., 2010. A temporarily distinct subpopulation of slow-cycling melanoma cells is required for continuous tumor growth. Cell 141, 583–594. Rumman, M., Dhawan, J., Kassem, M., 2015. Concise review: quiescence in adult stem cells: biological significance and relevance to tissue regeneration. Stem Cells 33, 2903–2912.

9