Activation of Notch1 signaling in stromal fibroblasts inhibits ... - Nature

4 downloads 75 Views 916KB Size Report
Apr 25, 2011 - mice (Charles River, Wilmington, MA, USA) at age 8–10 weeks. ... Davies SR, Watkins G, Mansel RE, Jiang WG. (2007). Differential expression ...
Oncogene (2011) 30, 4316–4326

& 2011 Macmillan Publishers Limited All rights reserved 0950-9232/11 www.nature.com/onc

ORIGINAL ARTICLE

Activation of Notch1 signaling in stromal fibroblasts inhibits melanoma growth by upregulating WISP-1 H Shao1, L Cai1,3, JM Grichnik2,4, AS Livingstone1, OC Velazquez1,2 and Z-J Liu1,2 1 Department of Surgery, Miller School of Medicine, University of Miami, Miami, FL, USA; 2Sylvester Comprehensive Cancer Center, University of Miami, Miami, FL, USA; 3Hangzhou Red-Cross Hospital, Zhejiang Province, China and 4Department of Dermatology & Cutaneous Surgery, Miller School of Medicine, University of Miami, Miami, FL, USA

The tumor microenvironment is emerging as an important target for cancer therapy. Fibroblasts (Fbs) within the tumor stroma are critically involved in promoting tumor growth and angiogenesis through secretion of soluble factors, synthesis of extracellular matrix and direct cell–cell interaction. In this work, we aim to alter the biological activity of stromal Fbs by modulating the Notch1 signaling pathway. We show that Fbs engineered to constitutively activate the Notch1 pathway significantly inhibit melanoma growth and tumor angiogenesis. We determine that the inhibitory effect of ‘Notch-engineered’ Fbs is mediated by increased secretion of Wnt-induced secreted protein-1 (WISP-1) as the effects of Notch1 activation in Fbs are reversed by shRNA-mediated blockade of WISP-1. When ‘Notch-engineered’ Fbs are co-grafted with melanoma cells in SCID mice, shRNAmediated blockade of WISP-1 reverses the tumorsuppressive phenotype of the ‘Notch-engineered’ Fbs, significantly increases melanoma growth and tumor angiogenesis. Consistent with these findings, supplement of recombinant WISP-1 protein inhibits melanoma cell growth in vitro. In addition, WISP-1 is modestly expressed in melanoma-activated Fbs but highly expressed in inactivated Fbs. Evaluation of human melanoma skin biopsies indicates that expression of WISP-1 is significantly lower in melanoma nests and surrounding areas filled with infiltrated immune cells than in the adjacent dermis unaffected by the melanoma. Overall, our study shows that constitutive activation of the Notch1 pathway confers Fbs with a suppressive phenotype to melanoma growth, partially through WISP-1. Thus, targeting tumor stromal Fbs by activating Notch signaling and/or increasing WISP-1 may represent a novel therapeutic approach to combat melanoma. Oncogene (2011) 30, 4316–4326; doi:10.1038/onc.2011.142; published online 25 April 2011 Keywords: tumor microenvironment; melanoma; fibroblasts; Notch; WISP-1

Correspondence: Dr Z-J Liu, Department of Surgery, School of Medicine, Sylvester Comprehensive Cancer Center, University of Miami, RMSB 1046, 1600 NW 10th Avenue, Miami, FL 33136, USA. E-mail: [email protected] Received 11 January 2011; revised 14 March 2011; accepted 19 March 2011; published online 25 April 2011

Introduction Neoplastic cells reside in a complex tissue microenvironment in which numerous infiltrated and recruited host cells interact with cancer cells to contribute to tumor initiation and progression (Lorusso and Ruegg, 2008; Anton and Glod, 2009). These non-neoplastic cells, including (i) endothelial cells and pericytes that together form the tumor vasculature; (ii) infiltrated inflammatory cells, including macrophages and lymphocytes and (iii) fibroblasts (Fbs), which are derived from the local existing Fbs, recruited bone marrow-derived mesenchymal stem cells (MSCs) (Liu et al., 2009) and in some cases, epithelial–mesenchymal transition (Kalluri and Neilson, 2003; Zeisberg and Kalluri, 2008; Kalluri, 2009), constitute the tumor-associated stroma. Fbs are the major components of tumor stroma and critically involved in regulating tumor growth, metastasis and angiogenesis through secretion of soluble factors, synthesis of extracellular matrix and direct cell–cell interaction (Olumi et al., 1999; Lynch and Matrisian, 2002; Allinen et al., 2004; Bhowmick et al., 2004; Midwood et al., 2004). Infiltrated/recruited stromal Fbs are activated in tumor tissue and are named as cancer-associated Fbs (CAFs) (Orimo and Weinberg, 2006). CAFs have distinct biological features that distinguish them from their normal counterparts (Kalluri and Zeisberg, 2006; Orimo and Weinberg, 2006). The importance of CAFs in tumorigenesis has been recently highlighted. For instance, CAFs can render non-tumorigenic cells to gain a permanently transformed phenotype (Hayward et al., 2001). In addition, CAFs extracted from invasive human breast carcinomas are more capable of promoting the growth of mammary carcinoma cells and tumor angiogenesis compared with cells derived from outside of tumor masses (Orimo et al., 2005). Moreover, CAFs can mediate resistance to antiangiogenic therapy (Crawford et al., 2009). CAFs, thus, may represent promising therapeutic targets in the prevention and treatment of cancer. This may be particularly important to melanoma as Fbs make up a significant portion of the tumor mass. Cutaneous melanoma is the most dangerous form of skin cancer and its incidence rate has been rising rapidly in the last several decades (Liu and Herlyn, 2005). Unlike many other cancers, melanoma is notorious for its propensity to metastasize and for its poor response to chemotherapy and radiation. The current clinical

Targeting melanoma by manipulating Notch in fibroblasts H Shao et al

4317

treatment regimens, except early surgical removal of primary lesion, are, in general, ineffective. Therefore, it is necessary to develop novel, effective therapeutics to conquer melanoma. The Notch pathway is an evolutionarily conserved signaling cascade that regulates cellular activities in a variety of cells, including Fbs (Artavanis-Tsakonas et al., 1999). For instance, Notch pathway activation through either overexpression of the Notch1 intracellular domain (NIC) or stabilization of NIC by ablation of Sel-10 (Fbxw7), a negative regulator of Notch signaling, resulted in cell-cycle arrest and apoptosis in mouse embryonic Fbs (Ishikawa et al., 2008). Consistently, inhibition of Notch signaling by soluble forms of the Delta-like-1 and Jagged1 ligands was able to induce Fb growth factor receptor-dependent transformation of NIH 3T3 Fbs in vitro (Urs et al., 2008). These studies pointed to Notch signaling having a negatively regulatory role in controlling the cellular behavior of Fbs and suggested that the cellular activity of Fbs can be regulated by manipulation of Notch signaling. We aim to develop a novel approach to control melanoma growth and metastasis. In this study, we tested an innovative strategy to target the melanoma microenvironment by modulating stromal Fbs through manipulation of Notch signaling. We found that engineered stromal Fbs, in which the Notch pathway is constitutively activated, were able to attenuate melanoma growth and suppress tumor angiogenesis, partially through upregulating Wnt-induced secreted protein-1 (WISP-1). Our results unveil Notch signaling and its downstream functional mediator, WISP-1, as novel targets for therapeutic modulation of stromal Fbs in the treatment of melanoma.

Results Activation of Notch1 pathway confers Fbs with a suppressive phenotype that inhibits melanoma cell growth in vitro To investigate the effect of ‘Notch-engineered’ Fbs on melanoma cell growth, the Notch1 pathway was constitutively activated by transduction of FF2441 with NIC-GFP/lentivirus. Green fluorescent protein (GFP)/ lentivirus-transduced cells were used as control. The transduced cells were named as NIC-GFP/Fbs and GFP/Fbs accordingly. A 495% of the transduced Fbs expressed GFP as observed by fluorescence microscopy (data not shown). Expression of NIC in NIC-GFP/Fbs was confirmed by immunoblotting (Figure 1a). The effect of NIC-engineered Fbs on melanoma cell growth was tested by Fb–melanoma cell co-culture experiments. NIC-GFP/Fbs or GFP/Fbs were mixed with 1205Lu at a ratio of 1:1, 2:1, 3:1 and 5:1. After 2 days, cells were harvested for fluorescence-activated cell sorting analysis to determine the numbers of melanoma cells (red) and Fbs (green). Melanoma cell growth was significantly inhibited when co-cultured with NIC-GFP/Fbs versus GFP/Fbs (Figure 1a). Inhibition was more effective at higher ratios compared with that at 1:1. The 2:1 ratio appeared to be an optimal ratio as 3:1 and 5:1 did not achieve stronger inhibition. More apoptotic melanoma cells were observed in the presence of NIC-GFP/Fbs versus GFP/Fbs. Interestingly, many of the apoptotic melanoma cells were not in contact with the co-cultured NIC-GFP/Fbs, suggesting that the Fbs -mediated inhibitory effect may be cell–cell contact-independent (at least partly) and might derive from soluble factor(s) secreted from the Fbs (Figure 1b). To investigate such a NIC-GFP/Fb + 1205Lu

N

IC

-G G FP FP

GFP/Fb + 1205Lu

NIC β-actin 1.2

*

*

*

0.73

0.60

0.60

0.61

(1:1)

(2:1)

(3:1)

(5:1)

0.4

0.0 (1:1)

GFP/Fb+1205Lu

NIC-GFP/Fb + 1205Lu

P=0.0017 Relative growth rate

*

0.8 1.00

Cell growth rate

* P