Cancer Research Meeting Calendar Frontiers in Basic Cancer Research
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
Cancer Research 68, 715, February 1, 2008. doi: 10.1158/0008-5472.CAN-07-2103
© 2008 American Association for Cancer Research
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary Data
Right arrow Correction (v68,p2051)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Vallacchi, V.
Right arrow Articles by Rodolfo, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Vallacchi, V.
Right arrow Articles by Rodolfo, M.

Cell, Tumor, and Stem Cell Biology

CCN3/Nephroblastoma Overexpressed Matricellular Protein Regulates Integrin Expression, Adhesion, and Dissemination in Melanoma

Viviana Vallacchi1, Maria Daniotti1, Francesca Ratti1, Delia Di Stasi1, Paola Deho1, Annamaria De Filippo1, Gabrina Tragni2, Andrea Balsari3, Antonino Carbone2, Licia Rivoltini1, Giorgio Parmiani1, Noureddine Lazar4, Bernard Perbal4 and Monica Rodolfo1

1 Unit of Immunotherapy of Human Tumors and 2 Unit of Pathology at Fondazione Istituto di Ricovero e Cura a Carattere Scientifico, Istituto Nazionale per lo Studio e la Cura dei Tumori; 3 Institute of Pathology, University of Milan, Milan, Italy; and 4 Laboratoire d' Oncologie Virale et Moleculaire, UFR de Biochimie, Université Paris 7-D. Diderot, Paris, France

Requests for reprints: Monica Rodolfo, Department of Experimental Oncology, Istituto Nazionale Tumori, via G. Venezian 1, 20133 Milan, Italy. Phone: 39-02-23903235; Fax: 39-02-23902154; E-mail: monica.rodolfo{at}istitutotumori.mi.it.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
CCN3/nephroblastoma overexpressed belongs to the CCN family of genes that encode secreted proteins associated with the extracellular matrix (ECM) and exert regulatory effects at the cellular level. Overexpression of CCN3 was shown in metastatic melanoma cells compared with cells of the primary tumor from the same patient. Analysis of short-term cultures from 50 primary and metastatic melanomas revealed a heterogeneous expression pattern of both the 46-kDa full-length cytoplasmic/secreted protein and the 32-kDa nuclear-truncated form. The different protein expression patterns were not associated with gene alterations or polymorphisms. Like the metastatic cells expressing high levels of the 46-kDa CCN3, cells transfected to overexpress CCN3 showed increased adhesion to ECM proteins, whereas inhibition of CCN3 expression by small interfering RNA decreased adhesion to laminin and vitronectin. CCN3 overexpression induced increased expression of laminin and vitronectin integrin receptors {alpha}7β1 and {alpha}vβ5 by increasing their mRNA production. Moreover, CCN3 secreted by melanoma cells acted as an adhesion matrix protein for melanoma cells themselves. Analysis of CCN3 protein expression with respect to melanoma progression detected the protein in all visceral metastases tested and in most nodal metastases from relapsing patients but in only a few nodal metastases from nonrelapsing patients and cutaneous metastases. Consistently, xenotransplantation in immunodeficient mice showed a higher metastatic potential of melanoma cells overexpressing CCN3. Together, these data indicate a role for CCN3 in melanoma cell interaction with the ECM by regulating integrin expression, resulting in altered cell adhesion and leading melanoma progression to aggressive disease. [Cancer Res 2008;68(3):715–23]


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Melanoma local invasion of the derma and metastatic spread by invasion of blood vessels and lymphatics are associated with alterations in cell adhesion (1, 2). Integrins play a major role in cell adhesion and migration by integrating components of the extracellular matrix (ECM) to the cellular cytoskeleton and by coupling to intracellular signaling pathways that control major cellular processes, such as cell proliferation and survival (3). Integrins regulate melanoma cell movement in response to binding to the ECM in the tumor stroma, connective tissue, or plasma membrane receptors expressed on endothelial cells (4). The insoluble ECM, composed of proteins, glycoproteins, proteoglycans, and glycosaminoglycans in complex arrangements, provides structure and biological signals, stores soluble factors, and exerts a mechanical influence on surrounding cells, with a composition that changes in time and space with tissue-type specificity. A wide variety of integrin ligands has been detected (5), as well as different mechanisms that regulate integrin receptor functions from within the cell (6).

CCN3/nephroblastoma overexpressed (NOV) belongs to the CCN (CYR61, CTGF, and NOV) family of growth and differentiation regulators that act on a variety of cell types (7, 8). CCN3 exerts multiple biological functions through various signaling pathways involving cell surface receptors, such as integrins, connexins, fibulins, Notch, and calcium channels (9). The CCN3 gene encodes a secreted 46 to 48 kDa protein, containing an N-terminal signal peptide and four structural domains sharing identity with insulin-like growth factor–binding protein, Von Willebrand factor type C repeat, thrombospondin type 1 repeat, and cystein knot motif (10). A 32-kDa amino-truncated form is also found both in secreted and in the nucleus of different cell types. This short CCN3 isoform was described to lack the N-terminal domain, as determined by reactivity with the K19M antibody recognizing the C terminal (11) and confirmed recently with antibodies NH2 and NH5 specific for the N-terminal or the C-terminal part of CCN3, respectively (12). Protein sequence analysis (13) and lack of evidence for CCN3 alternative splicing suggest that the 32-kDa protein originates from posttranslational processing of the full-length protein (10). Although CCN3 expression was originally associated with differentiation and growth arrest in Wilms' tumors, neuroblastomas, osteosarcomas, chondrosarcomas, and rhabdomyosarcomas, recent data correlate CCN3 expression with an increased proliferative index in prostate and renal carcinomas and with metastatic potential in Ewing's sarcoma (14). Moreover, overproduction of the nuclear truncated form was shown to induce cell growth deregulation (15), probably by interfering with gene expression (16).

Here, we show that the CCN3 gene is overexpressed in metastatic melanoma cells and assess the functional role of the full-length 46-kDa secreted protein in melanoma cells. Our analysis of CCN3 protein expression points to its functional role in melanoma progression through a mechanism involving modulation of tumor cell adhesion.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Cells and transfections. Melanoma cells obtained from surgical specimens were used to generate short-term cell lines as described (17). Biopsies obtained after written informed consent from 47 patients were used. Cells of melanoma line 3988M were transfected with the pCMV47 vector encoding a V5-HIS-tagged human CCN3 protein (18) or with the pCMV82 vector encoding human CCN3 (11) at the third passage in culture. Stable transfectants were selected in G418 or blasticidin (Invitrogen), respectively. Control mock transfectants were transfected with the empty vector pcDNAV5-HIS (Invitrogen). Melanoma cell line 1568M was transfected with 100 nmol/L of annealed small interfering RNA (siRNA) targeting CCN3 (5'-GGAAAAAAGUGUCUCCGCAtt-3') or a scrambled control (Ambion) using Lipofectamine 2000 (Life Technologies).

Western blot analysis. Proteins obtained from melanoma cells were separated on 4% to 12% bis-Tris precast gels (Invitrogen), transferred to nitrocellulose membranes (Amersham), and blotted with anti-CCN3 rabbit polyclonal antibodies K19M (11) and NH5 specific for the C-terminal CCN3 domain (12), rabbit anti-actin (Sigma-Aldrich), and monoclonal antibodies (mAb) against fibronectin, NBS1, and MHC class I W6/32 (BD Transduction). Reactive proteins were visualized using horseradish peroxidase (HRP)–conjugated secondary antibodies (Amersham) followed by enhanced chemiluminescence (GE Healthcare). Coomassie blue staining was also used as loading control. Subcellular localization of CCN3 was examined in the nuclear fraction (obtained as described; ref. 13), the cytoplasmic fraction (separated from the membranes by centrifugation at 20,800 x g for 1 h), and the membrane fraction (further purified by two centrifugations at 10,000 x g for 10 min). Detection of CCN3 in conditioned culture medium was performed using heparin-conjugated sepharose (Amersham) as described (11).

Immunohistochemical analysis. Formalin-fixed, paraffin-embedded tumor samples were incubated in methanol plus 0.3% H2O2 for 30 min to quench endogenous peroxidase activity, rinsed with water several times, and incubated with anti-CCN3 antibody K19M antibody (1:2,000) for 1 h at room temperature. Antigen retrieval was performed in 0.05 mol/L sodium citrate (pH 6) for 5 min at 95°C in an autoclave, followed by cooling to room temperature and final rinses in PBS. Nonspecific binding sites were blocked by incubation with Ultra V Block (Lab Vision) for 10 min at room temperature. After antibody incubation, slides were washed in PBS plus 0.1% Triton, incubated with Primary Antibody Enhancer (Lab Vision) for 30 min, washed with PBS plus 0.1% Triton, and incubated with HRP Polymer (Lab Vision) for 30 min at room temperature in a humidified atmosphere. The peroxidase enzyme reaction was developed with 3,3'-diaminobenzidine (Dako).

Immunofluorescence. Cells were cultured in Lab-Tek chamber slides (Nunc) for 48 h, fixed in 4% paraformaldehyde in PBS for 30 min at 4°C, rinsed with 30% FCS in PBS, and permeabilized with 0.1% Triton X-100 in PBS at room temperature. Slides were sequentially incubated with (a) 30% FCS in PBS for 30 min, (b) primary antibody for 1 h (immunopurified rabbit anti-CCN3 antibodies NH2 and NH5 diluted at 1:500; ref. 12) and then rinsed thrice with 30% FCS PBS, and (c) secondary antibody (FITC goat anti-rabbit IgG purchased from Abcam) for 1 h and rinsed four times in 30% FCS PBS before mounting in Prolong Gold Antifade Reagent with 4',6-diamidino-2-phenylindole (Invitrogen).

Tumorigenic and metastatic ability in SCID mice. Mice were maintained at constant temperature and humidity, with food and water given ad libitum. Experimental protocols were approved by the Ethics Committee for Animal Experimentation of the Istituto Nazionale Tumori of Milan according to the United Kingdom Coordinating Committee on Cancer Research Guidelines (19). Six-week-old female SCID mice were used (Charles River). Tumor growth rate of CCN3 transfectants was determined after s.c. injection of 5 x 106 cells in the flank and measurement of tumor diameters twice a week by a caliper. Tumor weight was calculated as a2xb / 2. Metastases were induced by i.v. or intrasplenic (i.s.) injection of 5 x 106 cells in five mice per group. After 54 days from i.v. and 47 days from i.s. injection, mice were sacrificed and examined for tumor growth, and organs were fixed in formalin and treated for histologic examination.

Cell adhesion assays. Melanoma cell adhesion to laminin and CCN3 was tested on tissue culture plates, coated overnight at 4°C with 0.2 µg/cm2 human placental laminin, poly-L-lysine (Sigma), and conditioned medium containing CCN3 or recombinant CCN3 (Preprotech), and blocked with 1% bovine serum albumin (Sigma-Aldrich) at room temperature for 1 h. Supernatant from transfectants were concentrated 10-fold by Centricon YM-10 filter devices (Millipore). Cells were plated and incubated at 37°C for 90 min, nonadherent cells were rinsed off with RPMI, and adherent cells were fixed with methanol for 15 min and stained with 0.1% crystal violet. Plates were washed by immersion in tap water and then dried, and absorbance at 550 nm was measured after cell solubilization with a 50:50 mixture of 0.1 mol/L NaH2PO (pH 4.5) in 50% ethanol. CytoMatrix cell adhesion strips coated with human ECM proteins were used according to the manufacturer's instructions (Chemicon). Inhibition of adhesion was tested in cells incubated with antibodies blocking integrin function (see above) for 30 min before plating.

Fluorescence-activated cell sorting analysis. Expression of laminin and vitronectin receptors was determined by flow cytometry with a FACScan (BD) by considering 10,000 events and using the following mAbs: mouse anti-integrin {alpha}1 (FB12), {alpha}3 (P1B5), {alpha}5β1 (HA5), {alpha}2β1 (BHA2.1), {alpha}vβ5 (P1F6; all from Chemicon), {alpha}IIb (VM16a; Abcam), {alpha}vβ3 (23C6; BD PharMingen), 67-kDa laminin receptor (MLuC5; kindly provided by Dr. Serenella Pupa, Istituto Nazionale Tumori), rat anti-β1 (mAb13) and anti-{alpha}6 (GoH3; BD PharMingen), and anti-{alpha}7 mAbs O26 and 3C12 (generously provided by Dr. Stephen Kaufman from University of Illinois and Dr. Klaus von der Mark from University of Erlangen, respectively). Secondary FITC-labeled antibodies were antimouse IgG (Biosource) or antirat immunoglobulin (BD PharMingen).

Real-time reverse transcription–PCR analysis. Quantitative real-time reverse transcription–PCR (RT-PCR) was carried out by TaqMan TM; total RNA was reverse-transcribed using the high-capacity cDNA archive kit, and reactions were carried out in triplicate on an ABI PRISM 7900 machine using gene expression assays (Applied Biosystem). Samples were amplified in singleplex PCR reactions using the assays for ITGA7 (Hs00174397_m1), ITGB1 (Hs00559595_m1), ITGAV (Hs00233808_m1), and ITGB5 (Hs00174435_m1) labeled with FAM and for the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) labeled with VIC. Negative control without cDNA template was run with each assay. Data analysis was done using the SDS version 2.1 software. The relative mRNA expression was expressed as 2–{Delta}{Delta}CT, wherein {Delta}CT = CT target genes – CT GAPDH, and {Delta}{Delta}CT = {Delta}CT sample – {Delta}CT calibrator (mock trasfectant sample).

Statistical methods. The results of adhesion assays were evaluated by two-tailed unpaired t tests, and two-sided Fisher's exact test was used to calculate P values in Table 1 . Statistical significance was annotated as indicated in the figure legends.


View this table:
[in this window]
[in a new window]

  		Table 1. CCN3 isoform expression in short-term melanoma cell lines

 

   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Expression of CCN3 protein forms in melanoma. Analysis of a cDNA library obtained by subtractive hybridization from the global transcripts of a primary and a metastatic melanoma cell line derived from the same patient indicated up-regulation of CCN3 cDNA in the metastatic cells (not shown). A cDNA microarray analysis of gene expression profiles in the same melanoma samples, as well as Northern analysis and RT-PCR (not shown), confirmed this result. Indeed, higher CCN3 protein levels were detected in the metastatic compared with the primary tumor cells (Fig. 1A, left ).


Figure 1
View larger version (80K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. CCN3 expression in melanoma cells as detected by antibody K19M. A, expression of both the 46-kDa full-length and of the 32-kDa truncated CCN3 in 20842P (P) and 20842M2 (M; left) and in 3988M (1) and 1568M (2; right) short-term melanoma cell lines as detected in the whole-cell lysate (WCL) and in the culture medium (CM). Actin, fibronectin (FN) and Coomassie blue staining (CBB) were used as loading controls. B, expression of CCN3 isoforms in a panel of short-term melanoma cell lines established from lymph node metastases (LN) of patients at stage IIIB-C; LS, long survival patients; DOD, died of disease. C, CCN3 expression in tumor biopsies. Formalin-fixed and paraffin-embedded 3988M (a) and 1568M (d) cultured cells and in the corresponding cutaneous (b and c) and lung metastases (e and f) are shown as representative cases. Magnifications, 40x (b and e) and 60x (a, c, d, and f).

 
To assess the potential role of CCN3 in melanoma progression, expression was analyzed in a panel of 50 short-term melanoma cell cultures. Because CCN3 expression levels in cultured cells are dependent on cell density (20), samples were obtained in standard conditions at culture confluence. Differing levels of CCN3 gene expression detected by RT-PCR and confirmed by quantitative PCR (not shown) were associated with a heterogeneous pattern of CCN3 protein expression, with some melanoma samples expressing CCN3 at high levels, whereas others did not express it or expressed it only at low levels (Fig. 1B). Both the 46–48-kDa full-length protein doublet and the 32-kDa N-terminal truncated isoform were also expressed at various levels in the different samples, with a few samples expressing only one or the other form. In different cases tested, expression detected by immunohistochemistry on tumor specimens from which the cell lines were derived was similar to that of the cell lines (Fig. 1A, right and C), indicating that the different levels of expression occurring in vivo are maintained in cultured cells.

Four of seven short-term melanoma cell lines derived from primary tumors expressed the full-length protein. Assessment in metastases revealed the full-length CCN3 in 9 of 21 lymph node, 4 of 15 cutaneous, and 7 of 7 visceral metastases (Table 1), a distribution suggesting a role for CCN3 in the hematogenous dissemination required to establish visceral melanoma metastases. Note that in lymph node metastases, CCN3 protein expression was higher in melanoma cells derived from stage IIIB-C patients with short survival compared with cells derived from patients with long survival (Fig. 1B), suggesting a relationship between CCN3 expression and disease progression. The differential expression pattern was also observed when the 32-kDa short form of the protein was considered (Table 1). Furthermore, retrieval of the available clinical data for the relapsing patients with CCN3-positive nodal metastases revealed that, for most of them, disease progression occurred with visceral or cerebral metastases.

The role of CCN3 in establishment of visceral metastases is supported by data obtained in SCID mice after i.s. and i.v. injection of melanoma cells stably expressing CCN3 full-length protein after gene transfer. In fact, as shown in Fig. 2 , mice injected i.s. with CCN3-expressing cells developed more hepatic metastases compared with mice injected with control cells (mean liver weight, 3.124 g versus 1.844 g). Furthermore, in mice injected i.v., metastatic nodules were found in 8 of 10 adrenal glands after injection of CCN3-expressing cells compared with 1 of 10 in control mice injected with CCN3-negative cells, whereas subcutaneous tumor growth rate seemed similar.


Figure 2
View larger version (35K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. CCN3 full-length overexpression promotes visceral metastatic dissemination of melanoma cells in SCID mice. A, top left, expression of CCN3 full-length protein in whole-cell lysate and in culture medium obtained from 3988M melanoma cells transfected with empty plasmid (mock) or with pCMV47 plasmid (CCN3/1) as resulting from immunoblotting with K19M antibody. Top right, growth curves of tumors induced by mock ({blacksquare}) and CCN3/1 ({blacktriangleup}) transfectants after s.c. injection of 5 x 106 cells. Points, mean tumor weight (MTW); bars, SD. Bottom left, mean liver weight determined 47 d after i.s. injection of 5 x 106 cells is significantly greater in mice injected with CCN3/1 cells than in mice injected with mock. *, P < 0.05 by two-tailed unpaired t test. Bottom right, percentage of adrenals displaying melanoma metastases 54 d after i.v. injection of mock or CCN3/1 cells. B, histologic examination of liver tissue from mice injected with mock or CCN3/1 cells (a, b), showing a different level of invasion of liver tissue. Magnification, 4x. The appearance of liver from representative mice injected with mock or CCN3/1 cells (c, d). Low (4x, e) and high (40x, f) magnification of a representative section of adrenals obtained from a mouse injected with CCN3/1 cells showing massive tumor cell infiltration.

 
Immunohistochemical staining indicated a prevalent cytoplasmic and membrane localization of CCN3 in vivo (Fig. 1C). Analysis of cell fractions performed with the C-terminal domain specific antibody K19M on several melanoma cell lines established that the 46-kDa protein was mainly localized in the cytoplasm and secreted by melanoma cells in the culture medium, whereas the 32-kDa form had a nuclear localization (Fig. 3A and B ). Furthermore, immunofluorescence analysis performed with the antibody NH2, specific for the CCN3 N-terminal domain, and with the antibody NH5, specific for the CCN3 C-terminal domain, showed that only the latter detected the nuclear 32-kDa form, demonstrating that this truncated form of CCN3 maintained the C-terminal and lacked the N-terminal domains (Fig. 3C). Both CCN3 forms were present in the medium of melanomas expressing both protein forms as shown in Fig. 1A (right). The secreted full-length CCN3 was also detected in some melanomas whose lysates stained positive only for the truncated form. The use of NH5 confirmed this finding (Fig. 3B). Thus, in these cells, the bulk of the full-length CCN3 protein seems to be efficiently secreted upon biosynthesis and is not maintained in the cytoplasm. Rapid secretion of the full-length CCN3 in cultured cells, as well as the detection of variable amounts of the 46-kDa form within the cells, have been reported (11). Interestingly, some melanomas showing only the short form at the cellular level lacked a secreted full-length protein (Fig. 3B). This finding raises the possibility that mechanisms other than posttranslational processing of the full-length protein give rise to the 32-kDa form. However, CCN3 protein patterns are not due to gene alteration as resulted from the sequence analysis of the entire coding region of the CCN3 transcript (Supplementary data).


Figure 3
View larger version (27K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Cellular localization and secretion of CCN3 in melanoma cells. A, whole-cell lysate, nuclear (NF), membrane (MF), and cytoplasmic (CF) fractions obtained from 1568M (1), 20842P (2), and 20842M2 (3) melanoma cells and immunoblotted with antibody K19M. B, Western blot analysis of CCN3 protein isoforms as detected by antisera K19M and NH5 in whole-cell lysate and the corresponding culture medium (CM) from 9460M, 14362M, 1568M, and 5810P melanoma cells (lanes 1–4). C, immunofluorescence analysis of mock and CCN3/1 transfectants and of 1568M (1), 30966M (2), 2211M (3), and 26256M (4) melanoma cells stained with NH2 and NH5 antibodies, showing that the CCN3 C-terminal domain, recognized by NH5, is localized both in the cytoplasm and in the nucleus whereas the CCN3 N-terminal domain, recognized by NH2 antibody, is localized only in the cytoplasm.

 
Secreted CCN3 acts as an ECM protein for melanoma cells. Several studies have shown the involvement of CCN3 in cell adhesion. Recombinant CCN3 was shown to induce adhesion of vascular smooth muscle cells, endothelial cells, fibroblasts, and skeletal muscle cells through interactions with integrin cell surface receptors (21–24). In melanocytes, CCN3 expression determined adhesion to collagen type IV and localization to the basement membrane (25). Analysis to determine whether secreted CCN3 can act as an adhesive substrate protein for melanoma cells revealed higher adhesion to wells of plates coated with CCN3-containing conditioned medium compared with control medium; similar results were obtained using recombinant CCN3-coated plates (Fig. 4A, left ), indicating that adhesion was due to CCN3 and not to other factors in the conditioned medium. Of the integrins known to bind CCN3, namely {alpha}vβ3, {alpha}5β1, and {alpha}6β1, only {alpha}vβ3 is expressed by melanoma cells. Blocking experiments indicated that cell adhesion to CCN3 was mediated by multiple integrin receptors expressed by melanoma cells, including {alpha}2β1, {alpha}3, {alpha}7, β1, {alpha}vβ3, and {alpha}vβ5 (Fig. 4A, right).


Figure 4
View larger version (22K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. CCN3 increases adhesion of melanoma cells. A, left, adhesion of melanoma cells to CCN3. Mock transfectant cells were plated in microtiter wells coated with conditioned medium obtained from transfectants producing CCN3 or not (+ and –) or with increasing concentrations of recombinant CCN3 (rCCN3), and adhesion was measured as absorbance (A at 550 nm). PL, polylysine; CM, conditioned medium; CCM, concentrated conditioned medium. Right, melanoma cell adhesion to CCN3 is mediated by different integrin receptors. Mock transfectant cells were added to wells coated with concentrated conditioned medium obtained from CCN3/1 cells in the presence of function-blocking mAb directed to the indicated integrin receptors. Bars, SD. B, CCN3-induced increase in melanoma cell adhesion to ECM proteins. Adhesion in melanoma cells 20842M2 (M) expressing higher levels of 46-kDa CCN3 compared with autologous 20842P (P) cells (left), in 1568M and 3988M cells (middle), and in 3988M parental cells compared with transfectants CCN3/1, CCN3/2, and mock (right). Cells were plated in wells coated with fibronectin, collagen type I (CI), collagen type IV (CIV), laminin (LN), and vitronectin (VN) proteins, and adhesion was measured as absorbance at 550 nm. C, effect of CCN3 silencing on melanoma cell adhesion to laminin and vitronectin. siRNA-mediated knockdown of CCN3 46-kDa protein expression as detected in both the whole-cell lysate and the culture medium (CM) of 1568M cells 72 h after transfection (left) led to decreased adhesion of the cells (right). The different molecular weight of the secreted CCN3 form is the result of extensive glycosylation (11). Percentage of adhesion was calculated as (mean absorbance in treated samples / mean absorbance in control) x 100. *, P < 0.001 by two-tailed unpaired t test.

 
CCN3 protein alters the adhesive capacity of melanoma cells. The influence of CCN3 expression on the adhesive properties of melanoma cells was analyzed with respect to different ECM substrates. The adhesion capacity of cells from a primary tumor differed from that of cells from the autologous metastasis expressing the full-length CCN3 at different levels, thus suggesting that the 46-kDa full-length protein was involved in cell adhesion capacity (Fig. 4B, left). Melanoma cells that lacked expression of the full-length CCN3 showed poor adhesion compared with melanoma expressing it at high levels. In addition, stable transfectants expressing the full-length CCN3 protein showed increased adhesion to collagen I (2-fold to 3-fold), vitronectin (2-fold to 5-fold), and especially to laminin (7-fold to 14-fold) compared with nontransfected parental cells and mock transfectants; no difference in the adhesion capacity to fibronectin and collagen IV was observed (Fig. 4B, middle and right). Similar data were obtained in analysis of 10 different cellular clones obtained from CCN3 transfectants (not shown). To confirm that the increased adhesion to laminin and vitronectin was related to CCN3 expression, adhesion was tested in 1568M melanoma cells, which constitutively express high levels of the full-length protein, after transfection with siRNA targeting CCN3. In these transfectants, levels of the CCN3 46-kDa protein were reduced to 15% and 25% as shown by Western blot analysis of cell lysates and conditioned medium, respectively, and adhesion to both ECM proteins was decreased (Fig. 4C).

Adhesion assays using cells preincubated with CCN3-containing conditioned medium or with recombinant CCN3 revealed no detectable increase in adhesion (not shown), confirming the requirement for the protein secreted by melanoma cells for the increase in adhesive capacity.

CCN3 regulates expression of integrin receptors. CCN3 has been shown to bind directly to integrin receptors (26). Because integrin expression profiles associated with a migratory and invasive phenotype change during melanoma progression, we tested whether CCN3 expression might correlate with that of integrin receptors. Among the integrin receptors expressed by melanoma cells during progression, {alpha}1β1, {alpha}2β1, {alpha}3β1, {alpha}6β1, and {alpha}7β1 bind laminin whereas {alpha}vβ3, {alpha}vβ5, and {alpha}IIbβ3 bind vitronectin. Transfectants expressing the CCN3 46-kDa protein also expressed higher levels of both {alpha}7 and β1 compared with cells not expressing CCN3 (Fig. 5A ) whereas {alpha}1β1 and {alpha}6β1 were not detected. Both parental and CCN3-transfected cells expressed high levels of {alpha}2β1 and {alpha}3 and moderate levels of the 67-kDa laminin receptor (not shown). With respect to vitronectin receptors, transfectants expressed higher levels of {alpha}vβ5 than did parental cells whereas {alpha}vβ3 expression was similar in both cell types and integrin {alpha}IIbβ3 was not expressed (not shown). Quantitative real-time RT-PCR analysis revealed that higher expression of integrin {alpha}7, β1, {alpha}v, and β5 in transfectant cells expressing CCN3 were correlated to higher levels of the correspondent mRNA and particularly of {alpha}7, indicating that CCN3 modifies the level of integrin mRNA in melanoma cells (Fig. 5B).


Figure 5
View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5. CCN3-induced melanoma cell adhesion to laminin and to vitronectin is integrin-dependent. A, fluorescence-activated cell sorting analysis of integrin expression in 3988M cells transfected with the empty vector (mock) or with the pCMV47 vector encoding CCN3 (CCN3/1) showing increased {alpha}7, β1, and {alpha}vβ5 integrin expression in the CCN3/1 transfectants and similar expression of {alpha}3 and {alpha}vβ3. Negative and positive staining is shown. B, relative expression of {alpha}7, β1, {alpha}v, and β5 mRNA in mock and CCN3/1 cells as detected by real-time PCR. Each sample was tested in triplicate. Columns, mean; bars, SD. C and D, CCN3/1 cells were added to laminin-coated or vitronectin-coated wells in the presence or absence of function-blocking mAbs directed against the indicated receptors and adhesion was measured as absorbance at 550 nm. Relative cell adhesion was calculated as mean absorbance in samples treated with mAb/mean absorbance in control x 100. Control (C) unrelated isotype-matched mAbs were anti-{alpha}vβ5 for laminin and anti-{alpha}3 for vitronectin. Adhesion of mock control cells is also shown. *, P < 0.0001 compared with absorbance measured in controls.

 
Adhesion assays using cells preincubated with integrin-blocking antibodies indicated that only {alpha}7β1 was involved in the laminin adhesion, and both {alpha}vβ3 and {alpha}vβ5 were involved in the adhesion to vitronectin (Fig. 5C and D).


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Our study shows that CCN3 expression in melanoma affects tumor cell adhesion and dissemination. CCN3 was expressed in ~60% of the melanomas tested but was differentially expressed in metastases from diverse localizations; high expression was detected in visceral metastases, whereas only a few cutaneous metastases were positive. Most lymph node metastases expressing CCN3 were from patients relapsing after surgery and progressing to visceral disease. Up-regulation of CCN3 in epidermal melanocytes upon interaction with keratinocytes has been reported, as well as the requirement for this up-regulation for the spatial localization of melanocytes to the basement membrane (25). In that study, CCN3 overexpression increased adhesion to collagen type IV, whereas its knockdown dissociated melanocytes from the basement membrane and mislocalized them in the epidermis or in the dermis in skin reconstructs (25). Our results suggest that among the transformed melanocytic cells spreading from the primary tumor, those that express CCN3 are more prone to metastasize hematogenously to visceral organs. Consistent with a role in melanoma metastases, data obtained in SCID mice indicate an increase in experimental metastatic potential of melanoma cells expressing CCN3.

Our analysis of the pattern of CCN3 isoform expression and their transcript sequences in a large set of samples revealed gene expression in all melanomas tested, but different patterns of protein expression were assessed at the cellular level and in the culture supernatant. Of the 50 melanomas tested, 16 showed no CCN3 protein expression when antibody K19M, which has a relatively low sensitivity (27), was used for detection. Another group of 18 samples showed expression of both the 46-kDa and the 32-kDa protein forms at the cellular level, as well as in the culture medium, and a third group of 16 samples displayed only the full-length form (6 of 16 samples) or only the short form with or without protein secretion (5 of 16 each). The latter samples showing only the 32-kDa form in the cell lysate and absence of CCN3 in the culture medium, thus, contain cells in which only the short form is produced.

These results suggest that different mechanisms, including posttranscriptional regulation, maintenance in the cytoplasm of the 46-kDa long form or its complete extracellular release, absence of proteolytic cleavage of the full-length CCN3, and, lastly, synthesis of amino-truncated variants, all contribute to the modulation of CCN3 expression in melanoma cells. It has been hypothesized that the 32-kDa isoform originates from posttranslational processing of the full-length protein at the membrane site or in the ECM followed by direct or indirect internalization through specific interaction with a receptor or a transporter (10). Consistent with this model, we detected the 32-kDa species in the supernatant of melanoma cells that displayed both forms at the cellular level (Fig. 1A, right). However, the absence of the full-length protein in the culture medium, as revealed using either antibody K19M or the more sensitive NH5 antibody (12), suggests that mechanisms other than posttranslational proteolysis of the secreted protein are also at play. Different patterns of CCN3 protein expression have also been detected in chronic myeloid leukemia (28) and in other tumor types (Perbal, unpublished results). Kyurkchiev et al. (27) have suggested that an association of CCN3 with specific protein partners might regulate protein conformations, isoforms expressed, additional binding partners, and functional states.

Functional properties for both the CCN3 forms have been described, with the full-length form originally described as antiproliferative, whereas the amino-truncated nuclear form was shown to stimulate proliferation and act as an oncogene (16). Although we did not study the short form at the functional level, its nuclear localization in melanoma cells (Fig. 3A and C) is consistent with other studies and with its regulatory effect on transcription (29). Our results do, however, clearly show that the CCN3 full-length protein is associated with high adhesiveness of melanoma cells to ECM proteins.

Previous single-nucleotide polymorphism (SNP) array analysis of melanoma cell lines revealed no gene amplifications or homozygous deletions in the 8q24 chromosomal region (30) despite the gains in 8q that are reportedly common in primary melanomas as detected by comparative genomic hybridization analysis (31). The CCN3 transcript was detectable in all melanoma samples; however, we found no evidence to suggest alternative splicing in the samples showing only the short form, consistent with previous studies (10, 13). Moreover, mutational analysis of the CCN3 transcript gave no indication that sequence alterations underlie the different protein expression patterns, although our sequencing analysis of the cDNA included ~60 nucleotides upstream and ~300 nucleotides downstream of the translated region but not the remaining terminal 1,000 nucleotides of the mRNA, which might contain unknown regulatory sequences (Supplementary data). Of the different SNPs detected, the rs11538929 polymorphism of exon 4 and the rs2279112 polymorphism of exon 2 were equally distributed in the above-mentioned three groups of samples.

The mechanisms underlying different CCN3 expression patterns in melanoma remain unknown. Expression of other CCN genes has been shown to be regulated by specific growth factors and by environmental changes, such as hypoxia and biomechanical stimuli, and to include posttranscriptional regulatory mechanisms involving message stability (32). CCN2, in which gene regulation has been studied in detail, contains CAESAR sequences in the 3' untranslated region (UTR) region (32); however, we did not detect this sequence in the CCN3 gene 3' UTR region. Transforming growth factor-β1 (TGF-β1) reportedly regulates CCN3 expression in adrenocortical cells, glial cells, leiomyoma, and myometrium (33, 34). In melanocytes, CCN3 expression is induced by interleukin-1 (IL-1) and by other undefined factors produced by keratinocytes (25). Whether CCN3 expression depends on the autocrine production of IL-1, TGF-β, or other cytokines by melanoma cells awaits further study. The potential involvement of TGF-β1 in the regulation of CCN3 expression in melanoma is particularly intriguing because TGF-β–type signaling up-regulates genes that express vasculogenic, ECM remodeling factors, and Wnt signal inhibitors and is associated with a high metastatic propensity in melanoma (35).

Metastatic melanoma progression results from the integration of genetic and epigenetic events, with multiple genetic abnormalities determining epigenetic variations that are shaped by the tumor microenvironment and immune response. In this context, we assessed CCN3 expression in 39 melanomas previously characterized (17) for gene mutations frequently found in melanoma, i.e., in the CDKN2A, PTEN, BRAF, NRAS, CDK4, and TP53 genes. Although the pattern of CCN3 production could not be associated to a defined profile, p53 mutations were more significantly frequent (P < 0.005) in the group expressing high levels of CCN3 protein (9 of 14) compared with melanomas that were negative (1 of 13) or expressed only the long or only the short protein form (1 of 12). Whereas TP53 mutations are detectable in only 10% to 25% melanomas (36), these data strongly suggest that alteration of this pathway can influence CCN3 expression. Whether and how p53 mutations are associated with high CCN3 production remains unknown.

CCN3 acts as an adhesive signaling molecule in several cell types, including vascular smooth muscle cells (21), endothelial cells (22, 37), myoblasts (24), and fibroblasts (23). Purified CCN3 has been shown to support adhesion, migration, and proliferation/survival in endothelial cells and fibroblasts (22, 23). CCN3-induced neovascularization in vivo and promotion of proangiogenic activities in endothelial cells have also been described (22).

The cell adhesion molecules used by endothelial cells and fibroblasts to adhere and migrate to CCN3 include integrins and heparan sulfate proteoglycans (21–23, 37). Despite the absence of an RGD sequence motif, CCN3 associates to vitronectin receptor {alpha}vβ3 and to integrins {alpha}5β1 and {alpha}6β1 (22, 37), possibly through the C-terminal domain of the protein (24). We found that melanoma cells adhered to the CCN3 protein when used as an adhesion substrate through multiple integrin receptors, including {alpha}vβ3, {alpha}vβ5, {alpha}2β1, {alpha}3β1, and {alpha}7β1, consistent with the involvement of a cell type–dependent subset of adhesion molecules. Moreover, CCN3 expression in melanoma cells was associated with a high adhesive capacity to different ECM proteins, with CCN3-overexpressing cells displaying increased adhesion to laminin and vitronectin whereas siRNA-transfected cells showed decreased adhesion. Adhesion to laminin and vitronectin was mediated by integrin receptors {alpha}7β1 and {alpha}vβ5, whose expression increased in CCN3-transfected melanoma cells. By contrast, Benini and coworkers (18) reported the reduced expression of {alpha}2β1 and the absence of alteration of adhesive capacities to ECM proteins in TC-71 Ewing's sarcoma cells transfected to overexpress CCN3. Similarly, CCN3-transfected glioblastoma cells showed increased cell migration, but not increased adhesion to ECM proteins (38). These differing results further point to the role of cell type context in determining the effects of CCN3 on cellular functions. In particular, laminin receptor {alpha}7β1, first described in human MeWo melanoma cells (39), was shown to be expressed in melanoma cell lines and to induce motility on laminin upon gene transfer (40) but was not detected in epidermal melanocytes which show poor adhesion to laminin (41).

Integrin function and expression can be modulated by interaction with other integrins and other membrane proteins through complex external and internal means (6). Our data suggest that, in melanoma cells, CCN3 affects the expression and function of integrin receptors by up-regulating their expression and activation status, possibly through both extracellular interaction with the secreted protein and through intracellular means.

Our findings in melanoma add to the growing list of tumors displaying altered CCN3 expression and functional effects associated with this alteration. Our data point to a role for CCN3 expression in melanoma progression to aggressive disease and hematogenous dissemination to visceral organs.


   Acknowledgments
 
Grant support: Associazione Italiana per la Ricerca sul Cancro. Work performed in BP laboratory was funded by EU grant PROTHETS LSH-CT2004-5030306, and N. Lazar was supported by the same grant.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Tiziana Ranzani, Gloria Sovena, Donata Penso, and Monica Tortoreto for their excellent technical contribution and Dr. A. Addis for his help with experiments in mice.


   Footnotes
 
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

Received 6/ 6/07. Revised 11/13/07. Accepted 11/27/07.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Johnson JP. Cell adhesion molecules in the development and progression of malignant melanoma. Cancer Metastasis Rev 1999;18:345–57.[CrossRef][Medline]
  2. Haass NK, Herlyn M. Normal human melanocyte homeostasis as a paradigm for understanding melanoma. J Investig Dermatol Symp Proc 2005;10:153–63.[CrossRef][Medline]
  3. van der Flier A, Sonnenberg A. Function and interactions of integrins. Cell Tissue Res 2001;305:285–98.[CrossRef][Medline]
  4. Kuphal S, Bauer R, Bosserhoff AK. Integrin signaling in malignant melanoma. Cancer Metastasis Rev 2005;24:195–222.[CrossRef][Medline]
  5. Humphries JD, Byron A, Humphries MJ. Integrin ligands at a glance. J Cell Sci 2006;119:3901–3.[Free Full Text]
  6. Ginsberg MH, Partridge A, Shattil SJ. Integrin regulation. Curr Opin Cell Biol 2005;17:509–16.[CrossRef][Medline]
  7. Perbal B, Takigawa M. The CCN family of proteins: an overview. In: Perbal B, Takigawa M, editors. The CCN Proteins: A New Family of Cell Growth and Differentiation Regulators. London: Imperial College Press; 2005. p. 1–18.
  8. Gupta R, Hong D, Iborra F, Sarno S, Enver T. NOV (CCN3) functions as a regulator of human hematopoietic stem or progenitor cells. Science 2007;316:590–3.[Abstract/Free Full Text]
  9. Perbal B. Related CCN proteins: multifunctional signalling regulators. Lancet 2004;363:62–4.[CrossRef][Medline]
  10. Perbal B. NOV (nephroblastoma overexpressed) and the CCN family of genes: structural and functional issues. Mol Pathol 2001;54:57–79.[Abstract/Free Full Text]
  11. Chevalier G, Yeger H, Martinerie C, et al. novH: differential expression in developing kidney and Wilm's tumors. Am J Pathol 1998;152:1563–75.[Abstract]
  12. Lazar N, Panque N, Perbal B. New domain-specific CCN3 antibodies as unique tools for structural and functional studies. Cell Comm Signal. In press. 2008.
  13. Perbal B. Nuclear localization of NOVH protein: a potential role for NOV in the regulation of gene expression. Mol Pathol 1999;52:84–91.[Abstract]
  14. Perbal B. The CCN3 protein and cancer. Adv Exp Med Biol 2006;587:23–40.[Medline]
  15. Joliot V, Martinerie C, Dambrine G, et al. Proviral rearrangements and overexpression of a new cellular gene (nov) in myeloblastosis-associated virus type 1-induced nephroblastomas. Mol Cell Biol 1992;12:10–21.[Abstract/Free Full Text]
  16. Planque N, Perbal B. A structural approach to the role of CCN (CYR61/CTGF-/NOV) proteins in tumourigenesis. Cancer Cell Int 2003;3:15.[CrossRef][Medline]
  17. Daniotti M, Oggionni M, Ranzani T, et al. BRAF alterations are associated with complex mutational profiles in malignant melanoma. Oncogene 2004;23:5968–77.[CrossRef][Medline]
  18. Benini S, Perbal B, Zambelli D, et al. In Ewing's sarcoma CCN3(NOV) inhibits proliferation while promoting migration and invasion of the same cell type. Oncogene 2005;24:4349–61.[CrossRef][Medline]
  19. United Kingdom Co-ordinating Committee on Cancer Research (UKCCCR) guidelinesfor the welfare of animals in experimental neoplasia. 2nd ed. Br J Cancer 1998;77:1–10.[Medline]
  20. Bleau AM, Planque N, Lazar N, et al. Antiproliferative activity of CCN3: involvement of the C-terminal module and post-translational regulation. J Cell Biochem 2007;101:1475–91.[CrossRef][Medline]
  21. Ellis PD, Chen Q, Barker PJ, Metcalfe JC, Kemp PR. Nov gene encodes adhesion factor for vascular smooth muscle cells and is dynamically regulated in response to vascular injury. Arterioscler Thromb Vasc Biol 2000;20:1912–9.[Abstract/Free Full Text]
  22. Lin CG, Leu SJ, Chen N, et al. CCN3 (NOV) is a novel angiogenic regulator of the CCN protein family. J Biol Chem 2003;278:24200–8.[Abstract/Free Full Text]
  23. Lin CG, Chen CC, Leu SJ, Grzeszkiewicz TM, Lau LF. Integrin-dependent functions of the angiogenic inducer NOV (CCN3): implication in wound healing. J Biol Chem 2005;280:8229–37.[Abstract/Free Full Text]
  24. Lafont J, Thibout H, Dubois C, Laurent M, Martinerie C. NOV/CCN3 induces adhesion of muscle skeletal cells and cooperates with FGF2 and IGF-1 to promote proliferation and survival. Cell Commun Adhes 2005;12:41–57.[CrossRef][Medline]
  25. Fukunaga-Kalabis M, Martinez G, Liu ZJ, et al. CCN3 controls 3D spatial localization of melanocytes in the human skin through DDR1. J Cell Biol 2006;175:563–9.[Abstract/Free Full Text]
  26. Lau LF, Lam SC. The CCN family of angiogenic regulators: the integrin connection. Exp Cell Res 1999;248:44–57.[CrossRef][Medline]
  27. Kyurkchiev S, Yeger H, Bleau AM, Perbal B. Potential cellular conformations of the CCN3(NOV) protein. Cell Commun Signal 2004;2:9.[CrossRef][Medline]
  28. McCallum L, Price S, Planque N, et al. A novel mechanism for BCR-ABL action: stimulated secretion of CCN3 is involved in growth and differentiation regulation. Blood 2006;108:1716–23.[Abstract/Free Full Text]
  29. Planque N, Long Li C, Saule S, Bleau AM, Perbal B. Nuclear addressing provides a clue for the transforming activity of amino-truncated CCN3 proteins. J Cell Biochem 2006;99:105–16.[CrossRef][Medline]
  30. Stark M, Hayward N. Genome-wide loss of heterozygosity and copy number analysis in melanoma using high-density single-nucleotide polymorphism arrays. Cancer Res 2007;67:2632–42.[Abstract/Free Full Text]
  31. Curtin JA, Fridlyand J, Kageshita T, et al. Distinct sets of genetic alterations in melanoma. N Engl J Med 2005;353:2135–47.[Abstract/Free Full Text]
  32. Leask A, Abraham DJ. All in the CCN family: essential matricellular signaling modulators emerge from the bunker. J Cell Science 2006;119:4803–10.[Abstract/Free Full Text]
  33. Lafont J, Laurent M, Thibout H, et al. The expression of novH in adrenocortical cells is down-regulated by TGF-β 1 through c-Jun in a Smad-independent manner. J Biol Chem 2002;277:41220–9.[Abstract/Free Full Text]
  34. Luo X, Ding L, Chegini N. CCNs, fibulin-1C and S100A4 expression in leiomyoma and myometrium: inverse association with TGF-β and regulation by TGF-β in leiomyoma and myometrial smooth muscle cells. Mol Hum Reprod 2006;12:245–56.[Abstract/Free Full Text]
  35. Hoek KS, Schlegel NC, Brafford P, et al. Metastatic potential of melanomas defined by specific gene expression profiles with no BRAF signature. Pigment Cell Res 2006;19:290–302.[CrossRef][Medline]
  36. Rodolfo M, Daniotti M, Vallacchi V. Genetic progression of metastatic melanoma. Cancer Lett 2004;214:133–47.[CrossRef][Medline]
  37. Ellis PD, Metcalfe JC, Hyvonen M, Kemp PR. Adhesion of endothelial cells to NOV is mediated by the integrins {alpha}vβ3 and {alpha}5β1. J Vascular Res 2003;40:234–43.[CrossRef][Medline]
  38. Laurent M, Martinerie C, Thibout H, et al. NOVH increases MMP3 expression and cell migration in glioblastoma cells via a PDGFR-{alpha}-dependent mechanism. FASEB J 2003;17:1919–21.[Abstract/Free Full Text]
  39. Kramer RH, McDonald KA, Vu MP. Human melanoma cells express a novel integrin receptor for laminin. J Biol Chem 1989;264:15642–9.[Abstract/Free Full Text]
  40. Echtermeyer F, Schober S, Poschl E, von der Mark H, von der Mark K. Specific induction of cell motility on laminin by {alpha}7 integrin. J Biol Chem 1996;27:2071–5.
  41. Kramer RH, Vu MP, Cheng YF, Ramos DM, Timpl R, Waleh N. Laminin-binding integrin {alpha}7β1: functional characterization and expression in normal and malignant melanocytes. Cell Regul 1991;2:805–17.[Medline]



This article has been cited by other articles:


Home page
 Cancer Res.Home page
Correction: CCN3 Increases Integrin Expression and Adhesion
Cancer Res., March 15, 2008; 68(6): 2051 - 2051.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary Data
Right arrow Correction (v68,p2051)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Vallacchi, V.
Right arrow Articles by Rodolfo, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Vallacchi, V.
Right arrow Articles by Rodolfo, M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online