Notch1-MAPK Signaling Axis Regulates CD133+ Cancer Stem Cell-Mediated Melanoma Growth and Angiogenesis

Accepted Manuscript + Notch1-MAPK Signaling Axis Regulates CD133 Cancer Stem Cell-Mediated Melanoma Growth and Angiogenesis Dhiraj Kumar, Santosh Kumar, Mahadeo Gorain, Deepti Tomar, Harshal S. Patil, N.N.V. Radharani, T.V.S. Kumar, Tushar V. Patil, H.V. Thulasiram, Gopal C. Kundu PII:

S0022-202X(16)32232-1

DOI:

10.1016/j.jid.2016.07.024

Reference:

JID 472

To appear in:

The Journal of Investigative Dermatology

Received Date: 7 November 2015 Revised Date:

9 July 2016

Accepted Date: 11 July 2016

Please cite this article as: Kumar D, Kumar S, Gorain M, Tomar D, Patil HS, Radharani N, Kumar T, + Patil TV, Thulasiram HV, Kundu GC, Notch1-MAPK Signaling Axis Regulates CD133 Cancer Stem Cell-Mediated Melanoma Growth and Angiogenesis, The Journal of Investigative Dermatology (2016), doi: 10.1016/j.jid.2016.07.024. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

RI PT

Notch1-MAPK Signaling Axis Regulates CD133+ Cancer Stem Cell-Mediated Melanoma Growth and Angiogenesis

SC

Dhiraj Kumar1, Santosh Kumar2, Mahadeo Gorain1, Deepti Tomar1, Harshal S. Patil3, NNV Radharani1, TVS Kumar1, Tushar V. Patil4, H. V. Thulasiram3 and Gopal C. Kundu1*

1

M AN U

Laboratory of Tumor Biology, Angiogenesis and Nanomedicine Research, National Centre for Cell Science (NCCS), Pune 411007, India

2

Department of Biochemistry and Molecular and Cellular Biology, Georgetown University Medical Center, Washington D.C. 20057, USA

3

Chemistry-Biology Unit, Division of Organic Chemistry, CSIR-National Chemical Laboratory, Pune 411008, India Department of Pathology, YCM Hospital, Pune 411018, India

TE D

4

EP

*Correspondence: Laboratory of Tumor Biology, Angiogenesis and Nanomedicine Research, National Centre for Cell Science (NCCS), Pune 411007, India; Email ID: [email protected].

AC C

Short title: Notch1 drives melanoma-initiating cells

Abbreviations: CSCs, cancer stem cells; VEGF, vascular endothelial growth factor; HUVECs, human umbilical vein endothelial cells; MMP, matrix metalloproteinase; MAPK, mitogen activated protein kinase; NICD1, notch1 intracellular domain

Key Words: CD133, cancer stem cell, Notch1, MAPK, melanoma growth

1

ACCEPTED MANUSCRIPT

ABSTRACT Functional characterization and understanding of the intricate signaling mechanisms in stem like cells is crucial for the development of effective therapies in melanoma. We have studied

RI PT

whether melanoma cells are phenotypically distinct and hierarchically organized according to their tumorigenic nature. We report that melanoma-specific CD133+ cancer stem cells exhibit increased tumor-initiating potential, tumor-endothelial cell interaction and lung metastasis. These cells are able to trans-differentiate into an endothelial-like phenotype when cultured

SC

under endothelial differentiation-promoting conditions. Mechanistically, Notch1 upregulates MAPK activation through CD133, which ultimately controls VEGF and MMP expression in

M AN U

CD133+ stem cells leading to melanoma growth, angiogenesis and lung metastasis. Blockade or genetic ablation of Notch1 and MAPK pathways abolishes melanoma cell migration and angiogenesis. ChIP and reporter assays revealed that NICD1 regulates CD133 expression at the transcriptional level. Andrographolide inhibits NICD1 expression, NICD1-dependent

TE D

CD133-mediated MAPK and AP-1 activation, and epithelial to mesenchymal-specific gene expression, ultimately attenuating melanoma growth and lung metastasis. Human malignant melanoma specimen analyses revealed a strong correlation between NICD1, CD133 and p-

EP

p38 MAPK expression and malignant melanoma progression. Thus, targeting Notch1 and its regulated signaling network may have potential therapeutic implications for the management

AC C

of CSC-mediated melanoma progression.

2

ACCEPTED MANUSCRIPT

INTRODUCTION Malignant melanoma exhibits a high degree of phenotypic plasticity and is highly aggressive and drug-resistant in nature. The phenomenon of cancer stem cell (CSC) is an emerging area

RI PT

of research that provides a potential explanation for aggressiveness, drug-resistance and distant metastasis in various cancers (Li et al., 2007; Frank et al., 2011). Melanoma tumors contain a heterogeneous subpopulation with distinct molecular signatures exhibiting

SC

functional stem cell-like properties (Charafe-Jauffret et al., 2009; Zimmerer et al., 2013). CD133, CD20, CD271, ABCB5 and ALDH1A have been used as markers to identify CSCs

M AN U

in melanoma cell lines and/or patient biopsies (Fang et al., 2005; Monzani et al., 2007; Rappa et al., 2008; Schatton et al., 2008; Civenni et al., 2011; Luo et al., 2012). Recent studies have suggested that CSCs possess the migratory and self-renewal properties necessary for distant organ colonization (Gao et al., 2012). During cancer progression, the epithelial to mesenchymal transition (EMT) contributes to accelerating tumor growth and metastasis

TE D

through activation of Snail, Slug and Zeb (Peinado et al., 2007; Mani et al., 2008). Notch 1-4 and their ligands, Jagged 1, -2, DLL 1, -3, and -4, and multiple effector molecules, such as Hes 1-6, Hey 1 and 2, have been identified in various stem cells in

EP

vertebrates. Notch signaling plays crucial role in regulation of cell fate and differentiation of stem cells derived from brain and glioblastoma (GBM). Notch signaling is involved in

AC C

CD133+ CSC-dependent glioblastoma growth (Fan et al., 2010). Notch signaling promotes proliferation and/or survival in many cancers, including melanoma (Allenspach et al., 2002). Injecting DAPT (γ-secretase inhibitor) along with cisplatin significantly depletes CD133+ cells in lung adenocarcinoma, indicating that the Notch pathway is involved in regulation of the CD133+ subpopulation (Liu et al., 2013). MAPK signaling plays an important role in several cancers (Wagner et al., 2009; Kumar et al., 2010). This pathway is preferentially upregulated in CD133+ cells of colon and liver cancers and regulates tumor progression 3

ACCEPTED MANUSCRIPT

(Wang et al., 2010; Tang et al., 2012). However, the molecular link between CD133 expression and the activation of Notch and MAPK pathways as well as the mechanism underlying the CSC-mediated melanoma growth are not well established.

RI PT

CSCs in melanoma exhibit higher resistance to conventional therapies (Lev et al., 2003). Although chemotherapeutic agents, such as dacarbazine (DTIC), doxorubicin (Dox), dabrafenib and trametinib, are widely used in melanoma, they are however relatively less

SC

curative and exhibit several side effects (Lev et al., 2003; Vorobiof et al., 2003; Flaherty et al., 2012). Hence, there is a need to identify novel, effective and selective therapeutic agents Several groups have reported that

M AN U

that specifically target melanoma stem cells.

Andrographolide (Andro), which is derived from Andrographis paniculata, is a novel anticancer agent that attenuates tumor growth in multiple cancer models and inhibits CSCmediated multiple myeloma growth (Gunn et al., 2011; Tung et al., 2013; Zhang et al., 2014).

TE D

In this study, we report that Notch1 transcriptionally regulates CD133 expression that preferentially activates MAPK and regulates MMP-2/-9 and VEGF expression in CD133+ cells, leading to melanoma growth, angiogenesis and lung metastasis. Moreover, CD133+

EP

cells exhibit the EMT and are able to trans-differentiate into endothelial-like phenotypes. Furthermore, Andro suppresses melanoma progression through attenuation of the Notch1

AC C

signaling pathway. RESULTS

CD133+ mouse and human melanoma cells express high levels of Oct3/4, Nanog and Sox10 To investigate CSC-specific heterogeneity in melanoma, the expression of stem cell markers in mouse (B16F10, B16F1) and human (A375, SK-MEL-2, SK-MEL-28) melanoma cell lines was analyzed by flow cytometry. The results indicated that the majority of these

4

ACCEPTED MANUSCRIPT

markers are present in a heterogeneous manner (Figure 1a). Several studies have demonstrated that CD133+ cells exhibit stem cell properties in various cancers (Monzani et al., 2007; Fan et al., 2010). Therefore, to study the stem-like properties of CD133+ cells in

RI PT

B16F10, we isolated and analyzed the purity and the expression of stem cell-specific transcription factors in these cells. The data revealed that CD133+ cells exhibit increased Oct3/4, Nanog and Sox10 expression compared with CD133- cells (Figure 1b, c and

SC

Supplementary Figure S1a). Additionally, CD133+ cells derived from A375, SK-MEL-2 or SK-MEL-28 exhibit increased Sox10 and Oct3/4 expression (Supplementary Figure S1a). We

M AN U

next examined whether CD133+ cells exhibit dual positivity with CD20 or CD166 populations in B16F10 cells. The results revealed that CD133+ cells are 0.4% and 7.4% positive for CD20 and CD166, respectively (Supplementary Figure S1b). Moreover, CD133cells did not exhibit any overlap with CD271+ cells (Supplementary Figure S1c). The CD133+ subpopulation expresses a distinct gene expression profile

TE D

To determine the distinct gene expression profile of CD133+ with respect to the CD133melanoma subpopulation, the global gene expression profile was analyzed using Illumina Mouse WG6 whole genome microarray. The scatter plots and unsupervised hierarchical

EP

cluster analysis revealed a clear demarcation between these two groups, indicating that the CD133+ subpopulation has a distinct molecular signature (Figure 2a and b). A total of 411

AC C

genes were upregulated and 355 genes were downregulated in CD133+ compared with CD133- cells as represented by volcano plots and heat map analysis (Figure 2c and Supplementary Figure S2a). Further, Gene Ontology (GO) data suggested that many of the upregulated genes in CD133+ cells were associated with angiogenesis, adhesion, differentiation and metastasis. However, several downregulated genes were associated with apoptosis (Supplementary Figure S2b and Supplementary Table S1). These data prompted us

5

ACCEPTED MANUSCRIPT

to functionally validate these findings with respect to the EMT, tumor growth, angiogenesis and metastasis in melanoma. CD133+ cells exhibit increased tumor-initiating and long-term tumorigenic potential

RI PT

We further investigated the tumorigenic potential of CD133+ cells using an in vivo model. CD133+ cells exhibited increased tumor growth in NOD/SCID (Figure 2d-g) and C57BL/6J mice models (Supplementary Figure S2c-e). Moreover, B16F10 CD20+ and SK-MEL-28

SC

CD133+ cells also exhibit increased tumorigenic properties compared with CD20- and CD133- cells respectively (Supplementary Figure S2f and g). Furthermore, in vivo growth

M AN U

kinetics under limiting dilution revealed that CD133+ cells exhibit increased tumor-initiating potential and tumor growth compared with CD133- cells (Figure 2h and Supplementary Table S2). We next evaluated the long-term tumorigenic potential of the CD133+ subpopulation by serial implantation of CD133+ and CD133- cells into NOD/SCID mice. The results revealed that CD133+ cells exhibit increased tumorigenic potential in primary and secondary isografts

TE D

(Figure 2i-k). The expression of CD133 in primary cultures derived from CD133+ primary and secondary isografts was analyzed, and the percentage of cells expressing CD133 remained similar. Moreover, CD133- cells did not revert into the CD133+ subpopulation in

EP

CD133- primary and secondary isografts (Supplementary Figure S3a).

AC C

CD133+ cells exhibit high drug efflux capacity and survival potential Previous reports have demonstrated that acquisition of chemoresistance is associated with increased expression of multidrug resistance (MDR) proteins, a characteristic feature of the side population (SP). Interestingly, our data showed that CD133+ exhibits a higher percentage of SP (15.7%) compared with the CD133- subpopulation (1.6%) (Figure 3a). Furthermore, the effect of this observed drug efflux on cell viability was investigated using dacarbazine (DTIC), doxorubicin (Dox), dabrafenib and trametinib. Our results indicated that CD133+

6

ACCEPTED MANUSCRIPT

cells exhibit increased cell viability compared with CD133- cells in response to these drugs (Figure 3b). Moreover, enhanced colony formation and decreased cleaved caspase 3 levels indicate that CD133+ cells exhibit increased cell survival (Supplementary Figure S3b and c).

chemoresistance properties.

RI PT

Taken together, these data indicate that the CD133+ subpopulation may have higher

CD133+ cells exhibit enhanced tube formation, tumor-endothelial cell interaction and

SC

potential to trans-differentiate into endothelial-like cells

To examine the involvement of CD133+ cells in angiogenesis, we performed functional tube

M AN U

formation assays. Our data revealed that CD133+ cells form well-defined tube-like networks compared with CD133- cells (Figure 3c). Furthermore, we examined the role of CD133+ cells in tumor-endothelial cell interactions by co-migration assay. Strikingly, our results showed that human umbilical vein endothelial cells (HUVECs) are highly migratory towards CD133+

TE D

cells compared with CD133- cells (Figure 3d). In addition, conditioned media of CD133+ cells exhibited enhanced tube formation using HUVECs (Figure 3e). Furthermore, our data revealed that CD133+ cells exhibit enhanced expression of VEGF, VEGFR1, VEGFR2,

EP

CD34 and Cox2 (Figure 3f, Supplementary Figure S3d-f). To evaluate the ability of CD133+ cells to trans-differentiate into an endothelial-like

AC C

phenotype, CD133+, CD133- or unsorted cells were cultured in endothelial differentiation conditions. The flow cytometry results revealed that the expression of endothelial-specific markers, such as CD31, VEGFR2 and VEGFR1, was augmented in CD133+ cells at day 14, suggesting that these cells acquired an endothelial-like phenotype (Figure 3g). Additionally, CD133+ tumors also exhibit increased CD31 and VEGF expression (Figure 3h). Taken together, these data demonstrate that CD133+ cells promote angiogenesis in melanoma.

7

ACCEPTED MANUSCRIPT

CD133+ cells undergo the EMT and exhibit enhanced metastatic potential To further examine whether CD133+ cells undergo the EMT, we analyzed the expression of EMT-specific genes. The results revealed significant upregulation of mesenchymal-specific

RI PT

markers, such as N-cadherin, Slug, Snail and Twist, whereas expression of the epithelialspecific marker E-cadherin was downregulated in CD133+ cells derived from B16F10, A375, SK-MEL-2 or SK-MEL-28 (Supplementary Figure S4a-d). Our EMT data prompted us to

SC

examine the metastatic potential of B16F10-derived CD133+ cells. Interestingly, CD133+ cells exhibit increased migratory potential compared with CD133- cells, and this effect was

M AN U

abrogated upon silencing CD133 expression (Figure 4a, b and Supplementary Figure S4e, f). Furthermore, our data revealed that CD133+ cells express high levels of MMP-2/-9 and exhibit enhanced cell attachment (Figure 4c, d and Supplementary Figure S4g). Moreover, when injected through the tail vein, CD133+ cells exhibited enhanced seeding and metastatic

TE D

potential towards lungs but not liver (Figure 4e, f and Supplementary Figure S4h). Notch1 signaling regulates CD133 expression at both transcriptional and posttranscriptional levels

EP

Notch signaling plays a pivotal role in the regulation of differentiation and self-renewal of normal and cancer stem cells. Interestingly, we found that Notch1 pathway-specific genes,

AC C

such as Notch1, NICD1, Hes1 and Jagged1, are highly upregulated in CD133+ compared with CD133- cells (Figure 4g, h and Supplementary Figure S5a). Furthermore, CD133+ tumors exhibit increased Notch1 expression (Supplementary Figure S5b). Notch2, Notch3 and Notch4 expression remained unchanged, indicating that Notch1 signaling is predominant in CD133+ cells (Supplementary Figure S5c). Similarly Notch1 signaling was also activated in CD133+ cells derived from A375, SK-MEL-2 and SK-MEL-28 (Supplementary Figure S5d).

8

ACCEPTED MANUSCRIPT

To elucidate the dynamic inter-relationship between Notch1 signaling and CD133 expression, we treated CD133+ cells independently with two Notch inhibitors, namely GSIIX and GSI-X. Strikingly, these inhibitors suppress CD133 expression in CD133+ cells and

RI PT

increased CD133- population (Figure 4i, j and Supplementary Figure S5e and f). However, GSI-IX does not restore the expression of CD133 in the CD133- subpopulation (Supplementary Figure S5e). Overexpression of Notch1IC increases the expression of CD133

SC

in B16F10, A375 or SK-MEL-2 cells (Figure 4k and Supplementary Figure S5g and h). Moreover, siRNA-mediated knock down of Notch1 downregulates CD133 expression in

M AN U

CD133+ cells, indicating that Notch1 regulates CD133 expression (Supplementary Figure S5i). To further confirm the recruitment of NICD1 at the promoter of CD133, ChIP and Luciferase reporter assays were performed, and the results showed the NICD1 is recruited within -700 to -451 bp upstream of transcription start site in the CD133 promoter (Figure 4ln).

TE D

Notch1 regulates CD133-dependent MAPK and AP-1 activation To further explore whether MAPK signaling is involved in CD133-dependent melanoma progression, the expression of p-MEK3/6, p-p38, c-Fos and c-Jun was analyzed in CD133+

EP

and CD133- cells. The data revealed that the expression of these molecules is upregulated in

AC C

CD133+ cells or tumors (Figure 5a and Supplementary Figure S6a and b). Further, similar results were also observed in A375, SK-MEL-2 and SK-MEL-28 cells (Supplementary Figure S6a). EMSA data revealed enhanced AP-1-DNA binding in CD133+ cells (Figure 5b and Supplementary Figure S6c). These data suggest that MAPK signaling is predominantly active in CD133+ cells. To elucidate the dynamic inter-play between Notch1 and MAPK signaling, CD133+ cells were treated with SB203580 (p38 MAPK inhibitor), and CD133 expression was analyzed. The results showed that SB203580 has no effect on CD133 but attenuates the expression of p-p38, c-Fos and c-Jun, suggesting that CD133 acts upstream of 9

ACCEPTED MANUSCRIPT

the MAPK pathway (Figure 5c and Supplementary Figure S6d). SB203580 also abrogates the expression of p-p38, c-Fos and c-Jun in CD133- cells; however, these cells express very low levels of p38 MAPK and its associated molecules (Supplementary Figure S6e).

RI PT

We further studied whether Notch1 signaling regulates CD133 dependent MAPK activation. CD133+ cells were treated with GSI-IX, and expression of NICD1, p-p38, c-Fos and c-Jun was analyzed by western blot. The data showed that GSI-IX inhibits the expression

SC

of these specific molecules (Figure 5d). Moreover, GSI-IX downregulates CD133 expression in CD133+ cells (Figure 4i and j). EMSA results revealed that GSI-IX and SB203580 inhibit

M AN U

AP-1-DNA binding in CD133+ cells (Supplementary Figure S6f). Furthermore, silencing Notch1 or CD133 downregulates p-p38 and c-Jun expression, demonstrating that Notch1 and CD133 regulate MAPK signaling (Figure 5e and f).

Notch1 and MAPK regulate the EMT, migration and angiogenesis in CD133+ cells

TE D

To examine the effects of GSI-IX or SB203580 on tumor-endothelial cell interactions and CD133+ cell migration, co-migration and wound migration assays were performed. The results revealed that blocking Notch1 or MAPK pathway inhibits the CD133+-endothelial cell

EP

interaction and CD133+ cell migration (Figure 5g and Supplementary Figure S6g and h). Silencing Notch1, CD133 or c-Jun attenuates MMP-2/-9 and VEGF expression and inhibits

AC C

the migratory potential of CD133+ cells (Figure 5h and i). Furthermore, Snail and Slug expression was downregulated, and E-cadherin was upregulated in Notch1 or CD133 silenced CD133+ cells (Supplementary Figure S6i and j). These results suggest that Notch1 pathway is involved in the regulation of the CD133+ tumor-endothelial cell interaction and migration in CD133+ cells. Andrographolide

inhibits

CD133+

cell

viability

and

migration

through

downregulation of Notch1-dependent CD133 expression and MAPK activation 10

the

ACCEPTED MANUSCRIPT

Melanoma-specific CSCs are resistant to conventional therapies. However, dacarbazine (DTIC), dabrafenib and trametinib are widely used chemotherapeutic agents for melanoma, but these agents are associated with drug resistance and several side effects (Lev et al., 2003;

RI PT

Vorobiof et al., 2003; Flaherty et al., 2012). Therefore, it is important to identify novel effective and selective therapeutic agents that specifically target melanoma-specific CSCs. Our data showed that Andro reduces cell viability in unsorted, CD133+ and CD133- cells (Figure 6a). Furthermore, unlike DTIC or dabrafenib in combination with trametinib, Andro

SC

selectively downregulates CD133 expression in CD133+ cells (Supplementary Figure S7a and

M AN U

c). Recent studies have demonstrated that differentiated tumor cells (non-CSCs) transform into CSCs upon long term exposure to therapeutic agents (Chaffer et al., 2013). Interestingly, our data suggests that Andro, DTIC or in combination of dabrafenib and trametinib treatment do not restore CD133 expression in CD133- cells (Supplementary Figure S7b and c). In addition, Andro inhibits the expression of CD133 even at a dose of 25 µM in SK-MEL-2 and

TE D

SK-MEL-28 cells (Supplementary Figure S7d).

Our previous data demonstrated that Notch1 and MAPK pathways are highly upregulated in CD133+ cells. Therefore, to examine whether Andro plays a role in Notch1

EP

and MAPK pathways, CD133+ cells were treated with Andro, and the expression and activation of NICD1, Hes1, Hey1, p-p38, p-MEK3/6 and c-Fos were analyzed by western

AC C

blot. Our data showed that Andro downregulates the expression and activation of these signaling molecules (Figure 6b). The attenuation of NICD1 and CD133 expression was also observed even at lower dose (0- 25 µM) of Andro without affecting cell viability in these cells (Figure 6a and Supplementary Figure S7e-g). Thus, our data suggest that the observed inhibition of Notch1 and CD133 signaling is not due to toxicity rather than its inhibitory effect. Moreover, compared with dabrafenib or trametinib, Andro significantly inhibits the

11

ACCEPTED MANUSCRIPT

migratory potential of CD133+ cells through downregulation of MAPK activation (Supplementary Figure S7h and i). Andro suppresses EMT, angiogenesis and metastatic potential of CD133+ cells through

RI PT

abrogation of the MAPK pathway

To further investigate the role of Andro in tumorigenicity, CD133+ cells were injected into C57BL/6J mice subcutaneously, and Andro was administered intraperitoneally. The results

SC

revealed that Andro suppresses tumor growth and attenuates the expression of NICD1, CD133, p-p38, c-Jun and c-Fos in tumors (Figure 6c, d and Supplementary Figure S7j and k).

M AN U

We further examined the role of Andro in the EMT, angiogenesis and metastatic potential of CD133+ cells under in vitro and in vivo conditions. The results revealed that Andro downregulates MET-specific gene expression and lung metastasis in CD133+ cells (Figure 6e, f and Supplementary Figure S8a and b). Moreover, Andro also suppresses VEGF

TE D

expression in these tumors (Figure 6g). Overall, our data revealed that Andro attenuates CD133+ cell-mediated EMT, angiogenesis and metastasis. Notch1, CD133 and MAPK expression are highly correlated with malignant melanoma

EP

progression

We further correlated the in vitro and in vivo findings by examining the expression and co-

AC C

localization of NICD1, p-MEK3/6, p-p38, c-Fos and CD133 in human malignant melanoma specimens by immunofluorescence. The increased expression of NICD1 and CD133 colocalized with p-p38 and p-MEK3/6 in human malignant melanoma compared with peripheral normal tissues (Supplementary Figure S9a and b). Taken together, these results suggest that Notch1 and its regulated MAPK signaling networks may act as potential therapeutic targets for the management of CSC-mediated malignant melanoma (Figure 6h).

12

ACCEPTED MANUSCRIPT

DISCUSSION In this study, we report that melanoma-derived CD133+ cells exhibit stem cell-like phenotypic characteristics that play vital role in the EMT and melanoma progression. The

RI PT

Notch1 pathway regulates the expression of CD133, which preferentially activates MAPK, leading to AP-1-mediated MMP-2/-9 and VEGF expression and contributing to melanoma growth, angiogenesis and metastasis. As a proof of principle, pharmacological inhibitors or

signaling and the tumorigenic potential of CD133+ cells.

SC

genetic ablation of Notch1 inhibits the expression of CD133 that ultimately abrogates MAPK

M AN U

Our results indicate that melanoma is hierarchically organized into phenotypically distinct subpopulations of tumorigenic and non-tumorigenic cells. We report that CD133+ melanoma-specific CSCs are highly tumorigenic in nature. Several studies on melanoma and breast cancer suggest that functional CSCs are present in these cancer cells with higher

TE D

metastatic potential and a characteristic molecular signature (Charafe-Jauffret et al., 2009; Zimmerer et al., 2013). Our results demonstrate that CD133+ cells derived from melanoma exhibit a distinct gene expression profile compared with CD133- cells. CD133+ cells exhibit

EP

increased drug efflux as indicated by the SP phenotype, which is further supported by enhanced cell viability in response to dacarbazine (DTIC), doxorubicin (Dox), dabrafenib and

AC C

trametinib. In addition to increased cell survival, these results provide a possible explanation for enhanced chemoresistance in CD133+ cells. During tumor development, angiogenesis plays an important role, and various pro-

angiogenic genes are expressed in CSCs derived from solid tumors (Bussolati et al., 2008; Biddle et al., 2011; Frank et al., 2011; Grange et al., 2011; Raja et al., 2014). Our data revealed that CD133+ cells are associated with increased angiogenic activity in vitro and in vivo. The tumor-endothelial cell interaction plays a crucial role in neovascularization (Maity

13

ACCEPTED MANUSCRIPT

et al., 2014). Indeed, our results suggest that endothelial cells exhibit enhanced migratory and tube formation ability towards CD133+ compared with CD133- cells. Several studies have reported that tumor neovascularization is regulated by many angiogenic factors either from

RI PT

pre-existing or recruited endothelial cells (Ping et al., 2011). Our present findings demonstrate that CD133+ but not CD133- cells in melanoma have potential to transdifferentiate into endothelial-like cells, suggesting that neovascularization could be a

SC

functional property of these cells.

Previous studies have identified that CSCs switch between two distinct phenotypes:

M AN U

proliferative and migratory (Biddle et al., 2011). Our data indicates that CD133+ cells are highly migratory in nature as shown by the wound migration assay. Inhibition of the cell cycle upon mitomycin C treatment followed by a cell migration assay suggest that enhanced migration of CD133+ cells is not due to proliferation. The EMT programming is initially required for invasion and dissemination of tumor cells (Ocana et al., 2012). Circulating tumor

TE D

cells must survive with multiple stressors in hematogenous circulation to reach distant organs (Valastyan and Weinberg, 2011). During tumor metastasis, activation of pro-metastatic genes, such as uPA and MMP-2/-9, play an important role in the degradation of ECM

EP

proteins that facilitate tumor invasion (Rangaswami et al., 2006). Further, it has been reported that CSCs are involved in metastasis (Malanchi et al., 2011). Our results revealed

AC C

that CD133+ cells express high levels of MMP-2-/9, leading to enhanced metastasis towards lungs.

We further studied the molecular mechanism by which CD133+ cells regulate

melanoma progression in depth. The Notch1 pathway plays a crucial role in glioma cell survival and proliferation (Purow et al., 2005). Furthermore, Notch1 controls CD133 expression, leading to the regulation of glioblastoma growth (Fan et al., 2010). However, the mechanism by which Notch1 regulates CD133 expression and controls CSCs-mediated 14

ACCEPTED MANUSCRIPT

melanoma growth has not been previously studied. Our data show that Notch1 augments CD133 expression, whereas inhibition of its function substantially attenuates CD133 expression. Furthermore, CD133 promoter analysis revealed the presence of a NICD1

RI PT

binding site that has been reported by Wang et al. (Wang et al., 2011). Our ChIP and luciferase reporter assay data indicate that NICD1 binds within -700 to -451 region of the CD133 promoter and regulates CD133 expression. Taken together, our findings suggest that

SC

Notch1 regulates CD133 expression both at transcriptional and post-transcriptional levels in melanoma.

M AN U

The Notch pathway is activated by MAPK signaling and promotes tumor growth in thyroid follicular cells (Yamashita et al., 2013). However, our findings suggest that Notch1 stimulates the p38 MAPK pathway that preferentially activates AP-1 and upregulates MMP2-/9 and VEGF expression in CD133+ cells. We also observed that inhibition of p38 MAPK did not alter the expression of CD133, suggesting that CD133 is upstream of the p38 MAPK

TE D

pathway. However, Notch1 attenuates migration and the tumor-endothelial cell interaction. Blockade or silencing of the Notch1 pathway inhibits CD133-dependent MAPK signaling and ultimately attenuates CD133+ cell migration and the tumor-endothelial cell interaction,

EP

suppressing melanoma growth, angiogenesis and metastasis.

AC C

Failure to respond to conventional chemotherapy is attributed to chemoresistance of CSCs, and this phenomenon is one of the causes of cancer recurrence (Fusi et al., 1993; Gao et al., 2012). Accordingly, we sought to identify potential therapeutic agent(s) that could specifically and selectively target CSCs in melanoma. Earlier results suggest that Andro acts as anti-cancer agent targeting CSCs in multiple myeloma (Gunn et al., 2011). Our data revealed that Andro blocks the Notch1-dependent CD133 expression and MAPK activation in CD133+ cells leading to suppression of tumor growth, angiogenesis and metastasis in mice melanoma model. We have also shown that Andro inhibits the expression of CD133 in 15

ACCEPTED MANUSCRIPT

human melanoma, SK-MEL-2 and SK-MEL-28 cells. Moreover, additional studies are required to confirm these effects in human melanoma model. All these data suggested that Andro may improve the recurrence-free survival of melanoma patients. In summary, our

RI PT

study demonstrates the crucial role of the Notch1 pathway that drives CD133 expression to promote MAPK activation and ultimately control melanoma growth, angiogenesis and metastasis. Thus, targeting Notch1 and its regulated signaling network may have potential

MATERIALS AND METHODS

M AN U

In vivo tumorigenicity and immunohistochemistry

SC

therapeutic implications in the management of CSC-mediated melanoma progression.

All mice experiments were performed according to the guidelines of Institutional Animal Care and Use Committee (IACUC) of National Centre for Cell Science (NCCS), Pune, India. CD133+, CD133- or unsorted B16F10-Luc cells were mixed with Matrigel (1:1) (BD

TE D

Biosciences) and injected subcutaneously into the dorsal right flank of NOD/SCID or C57BL/6J mice (6-8 weeks old). Tumor length and width were measured twice a week with Vernier Calipers. In vivo bioluminescence imaging was conducted using IVIS (Xenogen Corp.) as previously described (Kumar et al., 2010). B16F10-derived CD133+ or CD133-

EP

cells (1 x 103, 5 x 102 and 1 x 102 cells) were subcutaneously injected with Matrigel into

AC C

C57BL/6J mice. The melanoma-initiating cell frequency was analyzed as previously described (Hu and Smyth, 2009). In separate experiments, sorted CD133+ cells were injected into C57BL/6J mice s.c.

and randomly divided into three groups. Once the tumor appeared, two doses of Andro (50 and 150 mg/Kg body wt) were injected intraperitoneally (i.p.), and tumor volumes were measured. Mice were sacrificed, and tumors were excised, photographed and weighed. Tumor volumes were measured using V= π/6 [(l x b)3/2]. Tumor sections were analyzed by immunohistochemistry using anti-CD31, anti-VEGF or anti-melanoma antibody. 16

ACCEPTED MANUSCRIPT

Immunohistochemistry of human melanoma specimens The formalin fixed specimens of human malignant melanoma and peripheral tissues were collected with the help of a histopathologist from the local hospital with written, informed

RI PT

patient consent as per institutional and hospital ethical committee approvals and guidelines. The specimens were analyzed by immunohistochemistry using confocal microscope (Zeiss). SUPPLEMENTARY INFORMATION

SC

Additional methods can be found in the Supplementary Materials and Methods. CONFLICTS OF INTEREST

ACKNOWLEDGEMENTS

M AN U

No potential conflicts of interest were disclosed by authors.

This work was primarily supported by National Centre for Cell Science (NCCS), Pune, India (to G.C.K.). D.K. was supported by UGC. D.T. and H.S.P. are supported by CSIR,

TE D

Government of India. We thank Ms. Anuradha Bulbule and Ms. Priyanka Ghorpade for

AC C

EP

reading this article and Ms. Poonam R. Pandey for technical help.

17

ACCEPTED MANUSCRIPT

REFERENCES Allenspach EJ, Maillard I, Aster JC, Pear WS. Notch signaling in cancer. Cancer Biol Ther 2002;1:466-76

RI PT

Biddle A, Liang X, Gammon L, Fazil B, Harper LJ, Emich H, et al. Cancer stem cells in squamous cell carcinoma switch between two distinct phenotypes that are preferentially migratory or proliferative. Cancer Res 2011;71:5317-26 Bussolati B, Bruno S, Grange C, Ferrando U, Camussi G. Identification of a tumor-initiating stem cell population in human renal carcinomas. FASEB J 2008;22:3696-705

SC

Chaffer CL, Marjanovic ND, Lee T, Bell G, Kleer CG, Reinhardt F, et al. Poised chromatin at the ZEB1 promoter enables breast cancer cell plasticity and enhances tumorigenicity. Cell 2013;154:61-74

M AN U

Charafe-Jauffret E, Ginestier C, Iovino F, Wicinski J, Cervera N, Finetti P, et al. Breast cancer cell lines contain functional cancer stem cells with metastatic capacity and a distinct molecular signature. Cancer Res 2009;69:1302-13 Civenni G, Walter A, Kobert N, Mihic-Probst D, Zipser M, Belloni B, et al. Human CD271positive melanoma stem cells associated with metastasis establish tumor heterogeneity and long-term growth. Cancer Res 2011;71:3098-109 Fan X, Khaki L, Zhu TS, Soules ME, Talsma CE, Gul N, et al. Notch pathway blockade depletes CD133-positive glioblastoma cells and inhibits growth of tumor neurospheres and xenografts. Stem Cells 2010;28:5-16

TE D

Fang D, Nguyen TK, Leishear K, Finko R, Kulp AN, Hotz S, et al. A tumorigenic subpopulation with stem cell properties in melanomas. Cancer Res 2005;65:9328-37 Flaherty KT, Infante JR, Daud A, Gonzalez R, Kefford RF, Sosman J, et al. Combined BRAF and MEK inhibition in melanoma with BRAF V600 mutations. N Engl J Med 2012;367:1694-703

EP

Frank NY, Schatton T, Kim S, Zhan Q, Wilson BJ, Ma J, et al. VEGFR-1 expressed by malignant melanoma-initiating cells is required for tumor growth. Cancer Res 2011;71:147485

AC C

Fusi S, Ariyan S, Sternlicht A. Data on first recurrence after treatment for malignant melanoma in a large patient population. Plast Reconstr Surg 1993;91:94-8 Gao H, Chakraborty G, Lee-Lim AP, Mo Q, Decker M, Vonica A, et al. The BMP inhibitor Coco reactivates breast cancer cells at lung metastasis sites. Cell 2012;150:764-79 Grange C, Tapparo M, Collino F, Vitillo L, Damasco C, Deregibus MC, et al. Microvesicles released from human renal cancer stem cells stimulate angiogenesis and formation of lung premetastasis niche. Cancer Res 2011;71:5346-56 Gunn EJ, Williams JT, Huynh DT, Iannotti MJ, Han C, Barrios FJ, et al. The natural products parthenolide and andrographolide exhibit anti-cancer stem cell activity in multiple myeloma. Leuk Lymphoma 2011;52:1085-97

18

ACCEPTED MANUSCRIPT

Hu Y, Smyth GK. ELDA: Extreme limiting dilution analysis for comparing depleted and enriched populations in stem cell and other assays. J Immunol Methods 2009;347:70-8 Kumar V, Behera R, Lohite K, Karnik S, Kundu GC. p38 kinase is crucial for osteopontininduced furin expression that supports cervical cancer progression. Cancer Res 2010;70:10381-91

RI PT

Lev DC, Ruiz M, Mills L, McGary EC, Price JE, Bar-Eli M. Dacarbazine causes transcriptional up-regulation of interleukin 8 and vascular endothelial growth factor in melanoma cells: A possible escape mechanism from chemotherapy. Mol Cancer Ther 2003;2:753-63

SC

Li F, Tiede B, Massaque J, Kang Y. Beyond tumorigenesis: cancer stem cells in metastasis. Cell Res 2007;17:3-14 Liu YP, Yang CJ, Huang MS, Yeh CT, Wu AT, Lee YC, et al. Cisplatin selects for multidrug-resistant CD133+ cells in lung adenocarcinoma by activating Notch signaling. Cancer Res 2013;73:406-16

M AN U

Luo Y, Dallaglio K, Chen Y, Robinson WA, Robinson SE, McCarter MD, et al. ALDH1A isozymes are markers of human melanoma stem cells and potential therapeutic targets. Stem Cells 2012;30:2100-13 Maity G, Mehta S, Haque I, Dhar K, Sarkar S, Banerjee SK, et al. Pancreatic tumor cell secreted CCN1/Cyr61 promotes endothelial cell migration and aberrant neovascularization. Sci Rep 2014;4:4995

TE D

Malanchi I, Santamaria-Martinrez A, Susanto E, Peng H, Lehr HA, Delaloye JF, et al. Interactions between cancer stem cells and their niche govern metastatic colonization. Nature 2011 ;481:85-9 Mani SA, Guo W, Liao MJ, Eaton EN, Ayyanan A, Zhou AY, et al. The epithelialmesenchymal transition generates cells with properties of stem cells. Cell 2008;133:704-15

EP

Monzani E, Facchetti F, Galmozzi E, Corsini E, Benetti A, Cavazzin C, et al. Melanoma contains CD133 and ABCG2 positive cells with enhanced tumorigenic potential. Eur J Cancer 2007;43:935-46

AC C

Ocana OH, Corcoles R, Fabra A, Moreno-Bueno G, Acloque H, Vega S, et al. Metastatic colonization requires the repression of the epithelial mesenchymal transition inducer Prrx1. Cancer Cell 2012;22:709-24 Peinado H, Olmeda D, Cano A. Snail, Zeb and bHLH factors in tumour progression: an alliance against the epithelial phenotype? Nat Rev Cancer 2007;7:415-28 Ping YF, Bian XW. Cancer stem cells switch on tumor neovascularization. Curr Mol Med 2011;11:69-75 Purow BW, Haque RM, Noel MW, Su Q, Burdick MJ, Lee J, et al. Expression of Notch-1 and its ligands, Delta-like-1 and Jagged-1, is critical for glioma cell survival and proliferation. Cancer Res 2005;65:2353-63

19

ACCEPTED MANUSCRIPT

Raja R, Kale S, Thorat D, Soundararajan G, Lohite K, Mane A, et al. Hypoxia-driven osteopontin contributes to breast tumor growth through modulation of HIF1α-mediated VEGF-dependent angiogenesis. Oncogene 2014;33:2053-64 Rangaswami H, Bulbule A, Kundu GC. Osteopontin: role in cell signaling and cancer progression. Trends Cell Biol 2006;16:79-87

RI PT

Rappa G, Fodstad O, Lorico A. The stem cell-associated antigen CD133 (Prominin-1) is a molecular therapeutic target for metastatic melanoma. Stem Cells 2008;26:3008-17 Schatton T, Murphy GF, Frank NY, Yamaura K, Waaga-Gasser AM, Gasser M, et al. Identification of cells initiating human melanomas. Nature 2008;451:345-9

SC

Tang KH, Ma S, Lee TK, Chan YP, Kwan PS, Tong CM, et al. CD133(+) liver tumorinitiating cells promote tumor angiogenesis, growth, and self-renewal through neurotensin/interleukin-8/CXCL1 signaling. Hepatology 2012;55:807-20

M AN U

Tung YT, Chen HL, Tsai HC, Yang SH, Chang YC, Chen CM. Therapeutic potential of Andrographolide isolated from the leaves of Andrographis paniculata nees for treating lung adenocarcinomas. Evid Based Complement Alternat Med 2013;2013:305898. Valastyan S, Weinberg RA. Tumor metastasis: molecular insights and evolving paradigms. Cell 2011;147:275-92 Vorobiof DA, Rapoport BL, Mahomed R, Karime M. Phase II study of pegylated liposomal doxorubicin in patients with metastatic malignant melanoma failing standard chemotherapy treatment. Melanoma Res 2003;13:201-3

TE D

Wagner EF, Nebreda AR. Signal integration by JNK and p38 MAPK pathways in cancer development. Nat Rev Cancer 2009;9:537-49 Wang H, Zou J, Zhao B, Johannsen E, Ashworth T, Wong H, et al. Genome-wide analysis reveals conserved and divergent features of Notch1/RBPJ binding in human and murine Tlymphoblastic leukemia cells. Proc Natl Acad Sci U S A 2011;108:14908-13

EP

Wang YK, Zhu YL, Qiu FM, Zhang T, Chen ZG, Zheng S, et al. Activation of Akt and MAPK pathways enhances the tumorigenicity of CD133+ primary colon cancer cells. Carcinogenesis 2010;31:1376-80

AC C

Yamashita AS, Geraldo MV, Fuziwara CS, Kulcsar MA, Friguglietti CU, da Costa RB, et al. Notch pathway is activated by MAPK signaling and influences papillary thyroid cancer proliferation. Transl Oncol 2013;6:197-205 Zhang QQ, Ding Y, Lei Y, Qi CL, He XD, Lan T, et al. Andrographolide suppress tumor growth by inhibiting TLR4/NF-κB signaling activation in insulinoma. Int J Biol Sci 2014;10:404-14 Zimmerer RM, Korn P, Demougin P, Kampmann A, Kokemüller H, Eckardt AM, et al. Functional features of cancer stem cells in melanoma cell lines. Cancer Cell Int 2013;13:78

20

ACCEPTED MANUSCRIPT

FIGURE LEGENDS: Figure 1. Melanoma cells constitute a heterogeneous CSCs subpopulation. (a) Bar graph represents the flow cytometry analyses of a heterogeneous subpopulation of CSCs using

RI PT

specific markers (CD133, CD20, CD271, ABCG2, ABCG5, VEGFR1, CD166 and CD44) in mouse melanoma B16F10 and B16F1 as well as human melanoma A375, SK-MEL-2 and SK-MEL-28 cells. Data are presented as the mean ± SEM of three independent experiments. (b) FACS analysis of CD133+ and CD133- subpopulations and their purity in B16F10 cells.

SC

(c) Immunoblot analyses of stemness specific markers (Oct3/4, Nanog and Sox10) in unsorted, CD133+ and CD133- B16F10 cells.

M AN U

Figure 2. Molecular profile, tumorigenicity and growth kinetics of CD133+ cells. (a-c) Scatter plots, cluster analysis and volcano plot of differentially expressed genes in CD133+ vs. CD133- cells. (d-f) CD133+, CD133- or unsorted B16F10-Luc cells were injected subcutaneously (s.c.) into NOD/SCID mice and tumor images were captured using IVIS,

TE D

quantified and represented in mean flux (n=6 mice). *P<0.007, **P<0.0007. Tumor volumes were measured twice weekly. *P<0.007. (g) CD133 expression was analyzed by immunoblot in tumor lysates. (h) In vivo limiting dilution analyses of indicated cells in C57BL/6J mice

EP

(n=10 mice). *P<0.00004, **P<0.000006, ***P<0.0000007. (i, j) CD133+ and CD133- cells were injected s.c. into NOD/SCID mice to generate primary isografts. Further, primary

AC C

isograft-derived cultures were re-implanted for secondary isografts. *P<0.05, **P<0.02 (n=6 mice). (k) Schematic representation of experimental outline for Figure 2i and j. Data are ± SEM.

Figure 3. Chemoresistance and angiogenic properties of CD133+ cells. (a) SP analysis of CD133+ and CD133- cells. (b) Effect of dacarbazine, doxorubicin, dabrafenib and trametinib on CD133+ and CD133- cells viability. *P<0.004 (n=3). (c) Vasculogenic mimicry in CD133+ and CD133- cells. Scale bars = 30 µm. Bar graph represents the number of tube junction/hpf;

21

ACCEPTED MANUSCRIPT

*P<0.0001 (n=3). (d) Migration of HUVECs towards CD133+ or CD133- cells and their quantitation in a co-migration assay. Scale bars = 30 µm. *P<0.00005 (n=3). (e) Tube formation using HUVECs (1 x 104) in the presence of conditioned medium (CM) of CD133+

RI PT

or CD133- cells and their quantitation. Scale bars = 30 µm. *P<0.00004 (n=3). (f left) qRTPCR analysis of VEGF in indicated cells (n=3). (f right) Immunoblot of VEGF in CD133+ and CD133- cells. (g) Flow cytometry analysis of CD31, VEGFR2 and VEGFR1 in unsorted,

SC

CD133+ and CD133- cells. (h) CD31 and VEGF expression in tumor sections derived from CD133+ and CD133- cells based on immunofluorescence and data quantitation. Scale bars =

M AN U

20 µm. *P<0.0005 (n=3). Data are mean ± SEM.

Figure 4. Role of CD133+ cells in metastasis and transcriptional regulation of CD133 by Notch1. (a) Wound migration at 0 and 12 h in indicated cells. *P<0.0006 (n=3). (b) Wound migration in CD133 silenced CD133+ cells. *P<0.00008 (n=3). (c) qRT-PCR of MMP-2/-9 (n=3). (d) Zymography in indicated cells. (e) CD133+ and CD133- B16F10-Luc cells were

TE D

injected through the tail vein. Metastatic sites were captured by IVIS. (f) Immunostaining of anti-melanoma antigens.

Scale

bars

=

100

µm.

(g and

h)

Immunoblot and

immunofluorescence of Notch1 pathway-associated proteins. Scale bars = 20 µm. (i and j)

EP

Effect of GSI-IX (γ-secretase inhibitor) on CD133 expression as shown by flow cytometry and immunoblot. Blue indicates CD133+ and red represents CD133- subpopulations. (k) Flow

AC C

cytometry analysis of CD133 in NICD1-overexpressing B16F10 cells. (l and m) Schematic representation of NICD1 binding sites on CD133 promoter. Chromatin immunoprecipitation was performed using CD133+ cells with anti-NICD1 antibody and PCR amplified with CD133 promoter specific primers. (n) CD133 promoter activity by luciferase reporter assay in cells and conditions as indicated. *P<0.009; #P<0.003 (n=3). Data are mean ± SEM. Figure 5. Role of Notch1 signaling on CD133-dependent MAPK activation. (a) Immunoblots of p-MEK3/6, p-p38, c-Fos and c-Jun in unsorted, CD133+ or CD133- cells. (b) 22

ACCEPTED MANUSCRIPT

AP-1-DNA binding in cells as indicated in 5a by EMSA. (c) Effect of SB203580 (p38 MAPK inhibitor) on p-p38, c-Jun, c-Fos and CD133 expression in CD133+ cells as shown by immunoblots. (d) Effect of GSI-IX on NICD1, p-p38, c-Jun and c-Fos levels in CD133+ cells.

RI PT

(e and f) Immunoblot of p-p38, c-Jun, Notch1 and CD133 in Notch1- or CD133-silenced CD133+ cells. (g) Effect of GSI-IX and SB203580 on migration of HUVECs towards CD133+ cells. The data are mean ± SEM (n=3). *P<0.005; **P<0.0005.

(h) Wound

migration assay at indicated conditions. The data are mean ± SEM (n=3). *P<0.002;

SC

**P<0.0006. (i) Notch1, CD133 or c-Jun were silenced by their specific siRNA in CD133+

M AN U

cells, and the expression of metastasis- and angiogenesis-specific genes was analyzed by qRT-PCR. Bar graph represents mean ± SEM (n=2).

Figure 6. Andrographolide abrogates tumorigenic and metastatic potential of CD133+ cells by targeting Notch1-dependent MAPK pathway. (a) A cell viability assay was performed in CD133+, CD133- and unsorted B16F10 cells treated with Andro at indicated

TE D

doses. Bar graph denotes mean ± SEM (n=3). *P<0.008 and **P<0.0008 compared with untreated cells. (b) Immunoblots of Notch1 regulated signaling molecules in Andro-treated CD133+ cells. (c) Effect of Andro on CD133+ cell-derived tumors in C57BL/6J mice (n=6).

EP

Data are mean ± SEM. *P<0.005; **P<0.0007; ***P<0.0006. (d) Immunoblots of NICD1, pp38, c-Jun and c-Fos from mice tumors lysate treated with Andro. (e) CD133+ B16F10-Luc

AC C

(1 x 103) cells were injected through the tail vein of NOD/SCID mice, and mice were treated with Andro intraperitoneally. Metastatic sites were captured by IVIS. (f) H & E staining of lung metastases. Scale bars = 100 µm. (g) Immunofluorescence of VEGF in tumor sections as indicated. Scale bars = 20 µm. (h) Schematic representation of Notch1 signaling that regulates CD133-mediated MAPK activation. MAPK activation leads to AP1-dependent VEGF and MMP-2/-9 expression, which contributes to melanoma growth, angiogenesis and metastasis.

23

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT