IJC International Journal of Cancer
The metastatic microenvironment: Claudin-1 suppresses the malignant phenotype of melanoma brain metastasis Sivan Izraely1, Orit Sagi-Assif1, Anat Klein1, Tsipi Meshel1, Shlomit Ben-Menachem1, Assaf Zaritsky2, Marcelo Ehrlich1, Victor G. Prieto3, Menashe Bar-Eli4, Christine Pirker5, Walter Berger5, Clara Nahmias6,7,8, Pierre-Olivier Couraud6,7,8, Dave S.B. Hoon9 and Isaac P. Witz1 1
Department of Cell Research and Immunology, George S. Wise Faculty of Life Sciences, Tel Aviv University, Tel Aviv, Israel Blavatnik School of Computer Science, Tel Aviv University, Tel Aviv, Israel 3 Department of Pathology, The University of Texas M.D. Anderson Cancer Center, Houston, TX 4 Department of Cancer Biology, The University of Texas MD Anderson Cancer Center, Houston, TX 5 Institute of Cancer Research, Department of Medicine I, Medical University Vienna, Vienna, Austria 6 Inserm, U1016, Institut Cochin, Paris, France 7 Cnrs, UMR8104, Paris, France 8 University Paris Descartes, UMR-S 1016, Paris, France 9 Department of Molecular Oncology, John Wayne Cancer Institute, Saint John’s Health Center, Santa Monica, CA
Cancer Cell Biology
2
Brain metastases occur frequently in melanoma patients with advanced disease whereby the prognosis is dismal. The underlying mechanisms of melanoma brain metastasis development are not well understood. Identification of molecular determinants regulating melanoma brain metastasis would advance the development of prevention and therapy strategies for this disease. Gene expression profiles of cutaneous and brain-metastasizing melanoma variants from three xenograft tumor models established in our laboratory revealed that expression of tight junction component CLDN1 was lower in the brain-metastasizing variants than in cutaneous variants from the same melanoma. The objective of our study was to determine the significance of CLDN1 downregulation/loss in metastatic melanoma and its role in melanoma brain metastasis. An immunohistochemical analysis of human cells of the melanocyte lineage indicated a significant CLDN1 downregulation in metastatic melanomas. Transduction of melanoma brain metastatic cells expressing low levels of CLDN1 with a CLDN1 retrovirus suppressed their metastatic phenotype. CLDN1overexpressing melanoma cells expressed a lower ability to migrate and adhere to extracellular matrix, reduced tumor aggressiveness in nude mice and, most importantly, eliminated the formation of micrometastases in the brain. In sharp contrast, the ability of the CLDN1-overexpressing cells to form lung micrometastases was not impaired. CLDN1-mediated interactions between these cells and brain endothelial cells constitute the mechanism underlying these results. Taken together, we demonstrated that downregulation or loss of CLDN1 supports the formation of melanoma brain metastasis, and that CLDN1 expression could be a useful prognostic predictor for melanoma patients with a high risk of brain metastasis.
Malignant melanoma has a high tendency to develop brain metastasis. Almost 40% of melanoma patients are treated for complications due to brain metastasis, and in additional 30–40% of patients CNS lesions are detected at autopsy.1,2 Treatment options
for melanoma patients with cerebral metastasis are limited and depend on the number and size of the lesions. In view of the poor survival rates of these patients, it is necessary to develop novel treatment strategies for the therapy of melanoma brain metastasis.
Key words: melanoma, brain metastasis, micrometastasis, claudin-1 Abbreviations: Ab: antibody; ANGPTL4: angiopoietin-like 4; BSA: bovine serum albumin; CFDA-SE: carboxyfluorescein diacetate Nsuccinimidyl ester; CLDN1: claudin-1; DAPI: 40 ,6-diamidino-2-phenylindole dihydrochloride; ECM: extracellular matrix; FCS: fetal calf serum; FITC: fluorescein isothiocyanate; H&E: hematoxylin and eosin; hCMEC/D3: human brain microvascular endothelial cells; HPMEC: human pulmonary microvascular endothelial cells; IFNg: interferon gamma; MCM: melanoma conditioned medium; OD: optical density; PBS: phosphate-buffered saline; PTGS2: prostaglandin-endoperoxide synthase 2; qRT-PCR: quantitative real-time PCR; RS9: ribosomal protein S9; RT: room temperature; SEM: standard error of the mean; TBS: tris-buffered saline; TGFb1: transforming growth factor beta 1; TJ: tight junction; TNFa: tumor necrosis factor alpha; UB: universal buffer Additional Supporting Information may be found in the online version of this article. Grant sponsor: The Dr. Miriam and Sheldon G. Adelson Medical Research Foundation (Needham, MA) DOI: 10.1002/ijc.29090 History: Received 19 Dec 2013; Accepted 7 July 2014; Online 21 Jul 2014 Correspondence to: Isaac P. Witz, Department of Cell Research and Immunology, George S. Wise Faculty of Life Sciences, Tel-Aviv University, Tel-Aviv 69978, Israel, Tel.: 1972-3-640-6979, Fax: 1972-3-640-6974, E-mail: [email protected]
C 2014 UICC Int. J. Cancer: 136, 1296–1307 (2015) V
1297
Izraely et al.
Tumor cells with the potential to metastasize to and colonize the brain may express distinctive molecular determinants, which promote metastasis formation in the brain.1 Prevention strategies for brain metastasis could be developed if molecules involved in this process are identified and their precise role deciphered. To better understand the molecular mechanisms underlying brain metastasis and to identify molecular targets, we have previously performed a gene expression analysis of cutaneous and brain-metastasizing human melanoma cells using exon microarrays,3 revealing that the tight junction (TJ) component claudin-1 (CLDN1) is downregulated in brain-metastasizing variants, compared to cutaneous variants of the same genetic background, of three human melanoma xenograft models. TJs, located along the cell membrane, are the most apical cell–cell contacts, and are important for barrier function in epithelial and endothelial cells.4 TJs create a physiological intercellular barrier separating the apical and basolateral spaces, as well as regulating the paracellular permeability of various solutes.5 Moreover, cell–cell adhesion mediated by TJs is important in maintaining epithelial morphology and tightly linked to cell proliferation and migration.6 Claudins are the key component of TJs and form the backbone of the TJ strands. The claudin family consists of at least 24 members,6 which interact with each other through homophilic and heterophilic interactions, maintaining the cellular barrier.5 In addition, the C-terminal regions of claudins interact with cytosolic proteins such as zonula occludens (ZO-1, 2 and 3), which are linked to the actin cytoskeleton, and are involved in signal transduction.5 CLDN1 and CLDN2 were identified as the main integral components of the TJ strands.7 Highly metastatic tumor cells often exhibit loss of TJ structure and function.4,5 The loss of TJs is thought to promote cancer cell proliferation by allowing constitutive accessibility of cancer cells to nutrients and growth factors.5 Moreover, during metastasis, cancer cells undergo a number of distinct steps. The initial step involves disruption of cell–cell junctions with concomitant changes in the expression of junctional complex proteins, leading to loss of cell–cell adhesion,6 and subsequently to the intravasation of tumor cells into the vasculature.8 All of the above can promote cancer cell migration and metastasis. Loss or downregulation of claudins is frequently observed in tumors isolated from various tissues, such as colon, breast, pancreas, prostate, ovary and melanoma.5,7 Overexpression of C 2014 UICC Int. J. Cancer: 136, 1296–1307 (2015) V
different claudins was reported to negatively regulate the invasiveness and migration of various types of cancer cells.5,9 For example, overexpression of CLDN2 reduced gastric carcinoma cell migration.5 In contrast, several authors reported that CLDN1 expression is elevated in cases of colon carcinoma and melanoma.10,11 We hypothesized that downregulation or loss of CLDN1 occurring in brain-metastasizing melanoma cells plays an important role in conferring a metastatic phenotype upon such cells. Therefore, the aim of our study was to determine if upregulating the expression of this protein in brain-metastasizing melanoma cells would ameliorate their malignancy phenotype and specifically their capacity to metastasize to the brain.
Material and Methods Cell culture
The production and maintenance of cutaneous human melanoma variants RKTJ.C and YDFR.C; brain macrometastasizing variants RKTJ.CB2, YDFR.CB2 and YDFR.CB3 and variants spontaneously micrometastasizing to the brain (YDFR.SB1, YDFR.SB2 and YDFR.SB3) and lungs (YDFR.CSL3) were previously described.3 Human embryonic kidney 293T cells were grown in Dulbecco’s modified Eagle’s medium supplemented with 10% heat-inactivated fetal calf serum (FCS), 2 mmol/ml Lglutamine, 100 U/ml penicillin, 0.1 mg/ml streptomycin and 12.5 U/ml nystatin (Biological Industries, Kibbutz Beit Haemek, Israel). Immortalized human brain microvascular endothelial cells (hCMEC/D3) were maintained as previously described.12 Immortalized human pulmonary endothelial cells (HPMECs) were kindly provided by Dr. Vera KrumpKonvalinkova (Institute of Pathology, Johannes-Gutenberg University, Mainz, Germany) and were maintained as previously described.13 Cells were routinely cultured in humidified air with 5% CO2 at 37 C. The cultures were tested and found to be free of mycoplasma. Flow cytometry
Cells were plated in growth medium for 48 hr. CLDN1 expression was analyzed by monoclonal anti-human CLDN1 antibody (Ab) (1 lg/sample, R&D Systems, Minneapolis, MA, USA), using flow cytometry as previously described.3 Fluorescein isothiocyanate (FITC)-conjugated goat anti-rat IgG (1:50, Jackson ImmunoResearch Laboratories, West
Cancer Cell Biology
What’s new? Melanoma often metastasizes to the brain, but if researchers could find out how it does so, perhaps they could prevent it. Cells poised to metastasize often reveal themselves by molecular clues, and indeed, melanoma cells likely to infiltrate the brain express less of the tight-junction protein CLDN1 than other melanoma cells. In this study, the authors showed that when melanoma cells express extra CLDN1, they could not form micro-metastases in the brain – though their ability to metastasize to the lungs was not impaired. Thus, CLDN1 expression could help predict the likelihood of brain metastasis, and targeting cells expressing low levels of the protein could help prevent or treat this deadly complication.
1298
Grove, PA, USA) was used as secondary Ab. Dead cells were gated out from analysis.
Cancer Cell Biology
CLDN1 immunostaining of tissue microarrays
Immunohistochemical staining of CLDN1 was performed on 5-mm sections of formalin-fixed paraffin-embedded progressive melanoma tissue microarray. The microarray contained a total of 94 melanocytic lesions including 19 benign melanocytic nevi, 21 dysplastic nevi, 23 primary malignant melanoma and 31 melanoma metastases (12 subcutaneous, nine lymph node, eight lungs, five brain, four bowel and two bone metastases). Each of the blocks contained two controls, one nevus and one melanoma. Histological diagnosis was confirmed by a consecutive section staining with hematoxylin and eosin (H&E). To determine CLDN1 expression, the sections were reacted overnight at 4 C with a polyclonal rabbit Abs against CLDN1 (1:50, Zymed Laboratories, San Francisco, CA, USA), then reacted with a biotin-conjugated anti-rabbit (1:200) at 37 C for 1 hr, followed by a peroxidase-conjugated streptavidin complex. Antigen retrieval was performed by microwave heating in citrate buffer (pH 6.0) for 10 min. The slides were developed with diaminobenzidine and lightly counterstained with hematoxylin. Immunoreactivity in tumor cells was graded by intensity (0– 31) and by percentage of reactive cells (0–5, 6–25, 26–75 and >75%), as previously described.14 Cases that showed cytoplasmic or membranous staining of at least 21 in >25% of the cells were considered positive for CLDN1 expression. Results from the different groups were compared using the v2 method and Kruskal–Wallis test. A p-value 25% lesional cells expressing CLDN1 with moderate or marked intensity). Staining was performed with anti-CLDN1, DAB and light hematoxylin. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
the upper side of the apical chamber was scraped gently with cotton swab to remove nonmigrating cells. The membranes were fixed with 4% paraformaldehyde, washed with PBS and mounted with Dapi Fluoromount-G (SouthernBiotech, Birmingham, AL, USA). The number of melanoma cells transmigrating to the underside of the membrane was determined by counting five high-power fields under fluorescence microscopy in duplicates (Nikon eclipse TE 2000-S, Nikon). Preparation of melanoma-conditioned medium
A total of 1 3 106 human melanoma cells (YDFR.CB3) were cultured for 24 hr. The cells were washed and supplemented with RPMI-1640 medium containing 0.5% FCS for 48 hr. Melanomaconditioned medium (MCM) was collected, centrifuged twice for 7 min at 1,200 rpm and filtered (0.45 lm; Whatman GmbH). A total of 4 3 105 hCMEC/D3 cells, cultured 24 hr before collecting the MCM, were washed with medium containing 0.5% FCS and then treated with MCM for 48 hr. hCMEC/D3 cells treated with 0.5% FCS medium served as control. Cells were then analyzed for CLDN1 expression using qRT-PCR. Statistical analysis
Paired or unpaired Student’s t-test was used to compare in vitro and in vivo results. Survival analysis was computed by the Kaplan–Meier method and compared by the log-rank test.
Results CLDN1 is differentially expressed in cutaneous and metastatic melanoma variants
We demonstrated previously that at the transcriptional level, CLDN1 expression was lower in brain-metastasizing human
melanoma variants than in cutaneous variants of the same genetic background.3 To determine whether cutaneous melanoma cells and cells that metastasize to the brain or lungs differ in CLDN1 protein expression, flow cytometry was performed with Abs against human CLDN1. The results showed that the percentage of CLDN1-positive cells was significantly lower (p < 0.05) in the brain-metastasizing and micrometastasizing variants than in the corresponding cutaneous variants of the human melanomas (Fig. 1a). However, there was no significant difference in the percentage of CLDN1-positive cells between the spontaneous lung micrometastatic variant YDFR.CSL3 and the corresponding cutaneous variant (p 5 0.12). The percentage of CLDN1-positive cells was significantly higher (p < 0.05) in the lung micrometastatic variant than in the five brain metastatic or micrometastatic variants of the YDFR melanoma, in which a negligible percentage of cells expressed CLDN1 (
The metastatic microenvironment: Claudin-1 suppresses the malignant phenotype of melanoma brain metastasis Sivan Izraely1, Orit Sagi-Assif1, Anat Klein1, Tsipi Meshel1, Shlomit Ben-Menachem1, Assaf Zaritsky2, Marcelo Ehrlich1, Victor G. Prieto3, Menashe Bar-Eli4, Christine Pirker5, Walter Berger5, Clara Nahmias6,7,8, Pierre-Olivier Couraud6,7,8, Dave S.B. Hoon9 and Isaac P. Witz1 1
Department of Cell Research and Immunology, George S. Wise Faculty of Life Sciences, Tel Aviv University, Tel Aviv, Israel Blavatnik School of Computer Science, Tel Aviv University, Tel Aviv, Israel 3 Department of Pathology, The University of Texas M.D. Anderson Cancer Center, Houston, TX 4 Department of Cancer Biology, The University of Texas MD Anderson Cancer Center, Houston, TX 5 Institute of Cancer Research, Department of Medicine I, Medical University Vienna, Vienna, Austria 6 Inserm, U1016, Institut Cochin, Paris, France 7 Cnrs, UMR8104, Paris, France 8 University Paris Descartes, UMR-S 1016, Paris, France 9 Department of Molecular Oncology, John Wayne Cancer Institute, Saint John’s Health Center, Santa Monica, CA
Cancer Cell Biology
2
Brain metastases occur frequently in melanoma patients with advanced disease whereby the prognosis is dismal. The underlying mechanisms of melanoma brain metastasis development are not well understood. Identification of molecular determinants regulating melanoma brain metastasis would advance the development of prevention and therapy strategies for this disease. Gene expression profiles of cutaneous and brain-metastasizing melanoma variants from three xenograft tumor models established in our laboratory revealed that expression of tight junction component CLDN1 was lower in the brain-metastasizing variants than in cutaneous variants from the same melanoma. The objective of our study was to determine the significance of CLDN1 downregulation/loss in metastatic melanoma and its role in melanoma brain metastasis. An immunohistochemical analysis of human cells of the melanocyte lineage indicated a significant CLDN1 downregulation in metastatic melanomas. Transduction of melanoma brain metastatic cells expressing low levels of CLDN1 with a CLDN1 retrovirus suppressed their metastatic phenotype. CLDN1overexpressing melanoma cells expressed a lower ability to migrate and adhere to extracellular matrix, reduced tumor aggressiveness in nude mice and, most importantly, eliminated the formation of micrometastases in the brain. In sharp contrast, the ability of the CLDN1-overexpressing cells to form lung micrometastases was not impaired. CLDN1-mediated interactions between these cells and brain endothelial cells constitute the mechanism underlying these results. Taken together, we demonstrated that downregulation or loss of CLDN1 supports the formation of melanoma brain metastasis, and that CLDN1 expression could be a useful prognostic predictor for melanoma patients with a high risk of brain metastasis.
Malignant melanoma has a high tendency to develop brain metastasis. Almost 40% of melanoma patients are treated for complications due to brain metastasis, and in additional 30–40% of patients CNS lesions are detected at autopsy.1,2 Treatment options
for melanoma patients with cerebral metastasis are limited and depend on the number and size of the lesions. In view of the poor survival rates of these patients, it is necessary to develop novel treatment strategies for the therapy of melanoma brain metastasis.
Key words: melanoma, brain metastasis, micrometastasis, claudin-1 Abbreviations: Ab: antibody; ANGPTL4: angiopoietin-like 4; BSA: bovine serum albumin; CFDA-SE: carboxyfluorescein diacetate Nsuccinimidyl ester; CLDN1: claudin-1; DAPI: 40 ,6-diamidino-2-phenylindole dihydrochloride; ECM: extracellular matrix; FCS: fetal calf serum; FITC: fluorescein isothiocyanate; H&E: hematoxylin and eosin; hCMEC/D3: human brain microvascular endothelial cells; HPMEC: human pulmonary microvascular endothelial cells; IFNg: interferon gamma; MCM: melanoma conditioned medium; OD: optical density; PBS: phosphate-buffered saline; PTGS2: prostaglandin-endoperoxide synthase 2; qRT-PCR: quantitative real-time PCR; RS9: ribosomal protein S9; RT: room temperature; SEM: standard error of the mean; TBS: tris-buffered saline; TGFb1: transforming growth factor beta 1; TJ: tight junction; TNFa: tumor necrosis factor alpha; UB: universal buffer Additional Supporting Information may be found in the online version of this article. Grant sponsor: The Dr. Miriam and Sheldon G. Adelson Medical Research Foundation (Needham, MA) DOI: 10.1002/ijc.29090 History: Received 19 Dec 2013; Accepted 7 July 2014; Online 21 Jul 2014 Correspondence to: Isaac P. Witz, Department of Cell Research and Immunology, George S. Wise Faculty of Life Sciences, Tel-Aviv University, Tel-Aviv 69978, Israel, Tel.: 1972-3-640-6979, Fax: 1972-3-640-6974, E-mail: [email protected]
C 2014 UICC Int. J. Cancer: 136, 1296–1307 (2015) V
1297
Izraely et al.
Tumor cells with the potential to metastasize to and colonize the brain may express distinctive molecular determinants, which promote metastasis formation in the brain.1 Prevention strategies for brain metastasis could be developed if molecules involved in this process are identified and their precise role deciphered. To better understand the molecular mechanisms underlying brain metastasis and to identify molecular targets, we have previously performed a gene expression analysis of cutaneous and brain-metastasizing human melanoma cells using exon microarrays,3 revealing that the tight junction (TJ) component claudin-1 (CLDN1) is downregulated in brain-metastasizing variants, compared to cutaneous variants of the same genetic background, of three human melanoma xenograft models. TJs, located along the cell membrane, are the most apical cell–cell contacts, and are important for barrier function in epithelial and endothelial cells.4 TJs create a physiological intercellular barrier separating the apical and basolateral spaces, as well as regulating the paracellular permeability of various solutes.5 Moreover, cell–cell adhesion mediated by TJs is important in maintaining epithelial morphology and tightly linked to cell proliferation and migration.6 Claudins are the key component of TJs and form the backbone of the TJ strands. The claudin family consists of at least 24 members,6 which interact with each other through homophilic and heterophilic interactions, maintaining the cellular barrier.5 In addition, the C-terminal regions of claudins interact with cytosolic proteins such as zonula occludens (ZO-1, 2 and 3), which are linked to the actin cytoskeleton, and are involved in signal transduction.5 CLDN1 and CLDN2 were identified as the main integral components of the TJ strands.7 Highly metastatic tumor cells often exhibit loss of TJ structure and function.4,5 The loss of TJs is thought to promote cancer cell proliferation by allowing constitutive accessibility of cancer cells to nutrients and growth factors.5 Moreover, during metastasis, cancer cells undergo a number of distinct steps. The initial step involves disruption of cell–cell junctions with concomitant changes in the expression of junctional complex proteins, leading to loss of cell–cell adhesion,6 and subsequently to the intravasation of tumor cells into the vasculature.8 All of the above can promote cancer cell migration and metastasis. Loss or downregulation of claudins is frequently observed in tumors isolated from various tissues, such as colon, breast, pancreas, prostate, ovary and melanoma.5,7 Overexpression of C 2014 UICC Int. J. Cancer: 136, 1296–1307 (2015) V
different claudins was reported to negatively regulate the invasiveness and migration of various types of cancer cells.5,9 For example, overexpression of CLDN2 reduced gastric carcinoma cell migration.5 In contrast, several authors reported that CLDN1 expression is elevated in cases of colon carcinoma and melanoma.10,11 We hypothesized that downregulation or loss of CLDN1 occurring in brain-metastasizing melanoma cells plays an important role in conferring a metastatic phenotype upon such cells. Therefore, the aim of our study was to determine if upregulating the expression of this protein in brain-metastasizing melanoma cells would ameliorate their malignancy phenotype and specifically their capacity to metastasize to the brain.
Material and Methods Cell culture
The production and maintenance of cutaneous human melanoma variants RKTJ.C and YDFR.C; brain macrometastasizing variants RKTJ.CB2, YDFR.CB2 and YDFR.CB3 and variants spontaneously micrometastasizing to the brain (YDFR.SB1, YDFR.SB2 and YDFR.SB3) and lungs (YDFR.CSL3) were previously described.3 Human embryonic kidney 293T cells were grown in Dulbecco’s modified Eagle’s medium supplemented with 10% heat-inactivated fetal calf serum (FCS), 2 mmol/ml Lglutamine, 100 U/ml penicillin, 0.1 mg/ml streptomycin and 12.5 U/ml nystatin (Biological Industries, Kibbutz Beit Haemek, Israel). Immortalized human brain microvascular endothelial cells (hCMEC/D3) were maintained as previously described.12 Immortalized human pulmonary endothelial cells (HPMECs) were kindly provided by Dr. Vera KrumpKonvalinkova (Institute of Pathology, Johannes-Gutenberg University, Mainz, Germany) and were maintained as previously described.13 Cells were routinely cultured in humidified air with 5% CO2 at 37 C. The cultures were tested and found to be free of mycoplasma. Flow cytometry
Cells were plated in growth medium for 48 hr. CLDN1 expression was analyzed by monoclonal anti-human CLDN1 antibody (Ab) (1 lg/sample, R&D Systems, Minneapolis, MA, USA), using flow cytometry as previously described.3 Fluorescein isothiocyanate (FITC)-conjugated goat anti-rat IgG (1:50, Jackson ImmunoResearch Laboratories, West
Cancer Cell Biology
What’s new? Melanoma often metastasizes to the brain, but if researchers could find out how it does so, perhaps they could prevent it. Cells poised to metastasize often reveal themselves by molecular clues, and indeed, melanoma cells likely to infiltrate the brain express less of the tight-junction protein CLDN1 than other melanoma cells. In this study, the authors showed that when melanoma cells express extra CLDN1, they could not form micro-metastases in the brain – though their ability to metastasize to the lungs was not impaired. Thus, CLDN1 expression could help predict the likelihood of brain metastasis, and targeting cells expressing low levels of the protein could help prevent or treat this deadly complication.
1298
Grove, PA, USA) was used as secondary Ab. Dead cells were gated out from analysis.
Cancer Cell Biology
CLDN1 immunostaining of tissue microarrays
Immunohistochemical staining of CLDN1 was performed on 5-mm sections of formalin-fixed paraffin-embedded progressive melanoma tissue microarray. The microarray contained a total of 94 melanocytic lesions including 19 benign melanocytic nevi, 21 dysplastic nevi, 23 primary malignant melanoma and 31 melanoma metastases (12 subcutaneous, nine lymph node, eight lungs, five brain, four bowel and two bone metastases). Each of the blocks contained two controls, one nevus and one melanoma. Histological diagnosis was confirmed by a consecutive section staining with hematoxylin and eosin (H&E). To determine CLDN1 expression, the sections were reacted overnight at 4 C with a polyclonal rabbit Abs against CLDN1 (1:50, Zymed Laboratories, San Francisco, CA, USA), then reacted with a biotin-conjugated anti-rabbit (1:200) at 37 C for 1 hr, followed by a peroxidase-conjugated streptavidin complex. Antigen retrieval was performed by microwave heating in citrate buffer (pH 6.0) for 10 min. The slides were developed with diaminobenzidine and lightly counterstained with hematoxylin. Immunoreactivity in tumor cells was graded by intensity (0– 31) and by percentage of reactive cells (0–5, 6–25, 26–75 and >75%), as previously described.14 Cases that showed cytoplasmic or membranous staining of at least 21 in >25% of the cells were considered positive for CLDN1 expression. Results from the different groups were compared using the v2 method and Kruskal–Wallis test. A p-value 25% lesional cells expressing CLDN1 with moderate or marked intensity). Staining was performed with anti-CLDN1, DAB and light hematoxylin. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
the upper side of the apical chamber was scraped gently with cotton swab to remove nonmigrating cells. The membranes were fixed with 4% paraformaldehyde, washed with PBS and mounted with Dapi Fluoromount-G (SouthernBiotech, Birmingham, AL, USA). The number of melanoma cells transmigrating to the underside of the membrane was determined by counting five high-power fields under fluorescence microscopy in duplicates (Nikon eclipse TE 2000-S, Nikon). Preparation of melanoma-conditioned medium
A total of 1 3 106 human melanoma cells (YDFR.CB3) were cultured for 24 hr. The cells were washed and supplemented with RPMI-1640 medium containing 0.5% FCS for 48 hr. Melanomaconditioned medium (MCM) was collected, centrifuged twice for 7 min at 1,200 rpm and filtered (0.45 lm; Whatman GmbH). A total of 4 3 105 hCMEC/D3 cells, cultured 24 hr before collecting the MCM, were washed with medium containing 0.5% FCS and then treated with MCM for 48 hr. hCMEC/D3 cells treated with 0.5% FCS medium served as control. Cells were then analyzed for CLDN1 expression using qRT-PCR. Statistical analysis
Paired or unpaired Student’s t-test was used to compare in vitro and in vivo results. Survival analysis was computed by the Kaplan–Meier method and compared by the log-rank test.
Results CLDN1 is differentially expressed in cutaneous and metastatic melanoma variants
We demonstrated previously that at the transcriptional level, CLDN1 expression was lower in brain-metastasizing human
melanoma variants than in cutaneous variants of the same genetic background.3 To determine whether cutaneous melanoma cells and cells that metastasize to the brain or lungs differ in CLDN1 protein expression, flow cytometry was performed with Abs against human CLDN1. The results showed that the percentage of CLDN1-positive cells was significantly lower (p < 0.05) in the brain-metastasizing and micrometastasizing variants than in the corresponding cutaneous variants of the human melanomas (Fig. 1a). However, there was no significant difference in the percentage of CLDN1-positive cells between the spontaneous lung micrometastatic variant YDFR.CSL3 and the corresponding cutaneous variant (p 5 0.12). The percentage of CLDN1-positive cells was significantly higher (p < 0.05) in the lung micrometastatic variant than in the five brain metastatic or micrometastatic variants of the YDFR melanoma, in which a negligible percentage of cells expressed CLDN1 (
