The role of endogenous IFN-beta in cutaneous melanoma: Consequences of epidermal hyperplasia

The purpose of this study was to characterize epidermal hyperplasia overlying malignant melanoma, to determine the mitogenic factor responsible for the induction of this hyperplasia and to investigate its biological consequence. Whether increased keratinocyte proliferation overlying melanoma is due to production of growth factors by the tumor cells or to other mechanisms is unknown. Epidermal hyperplasia overlying human melanoma was found overlying thick (>4.0mm), but not thin (<1.0mm) tumors. Immunostaining of the sections for growth factors related to angiogenesis revealed that epidermal hyperplasia was associated with loss of IFN-β production by the keratinocytes directly overlying the tumors. Since previous studies from our laboratory have demonstrated that exogenous administration of IFN-β negatively regulates angiogenesis, we hypothesize that tumors are able to produce growth factors which stimulate the proliferation of cells in the surrounding tissues. This hyperplasia leads to a decrease in the endogenous negative regulator of angiogenesis, IFN-β. ^ The human melanoma cell line, DM-4 and several of its clones were studied to identify the mitogenic factor for keratinocytes. The expression of TGF-α directly correlated with epidermal hyperplasia in the DM-4 clones. A375SM, a human melanoma cell line that produces high levels of TGF-α, was transfected with a plasmid encoding full-length antisense TGF-α. The parental and transfected cells were implanted intradermally into nude mice. The extent of epidermal hyperplasia directly correlated with expression of TGF-α and decreased production of IFN-β, hence, increased angiogenesis. ^ In the next set of experiments, we determined the role of IFN-β on angiogenesis, tumor growth and metastasis of skin tumors. Transgenic mice containing a functional mutation in the receptor for IFN α/β were obtained. A375SM melanoma cells were implanted both s.c. and i.v. into IFN α/βR −/− mice. Tumors in the IFN α/β R −/− mice exhibited increased angiogenesis and metastasis. IFN α/βR −/− mice were exposed to chronic UV irradiation. Autochthonous tumors developed earlier in the transgenic mice than the wild-type mice. ^ Collectively, the data show that TGF-α produced by tumor cells induces proliferation of keratinocytes, leading to epidermal hyperplasia overlying malignant melanoma associated with loss of IFN-β and enhanced angiogenesis, tumorigenicity and metastasis. ^

NRAS and BRAF mutations in primary cutaneous melanoma: A comparison of mutation rates between radial and vertical growth phases in individual tumors

Primary cutaneous melanoma is a cancer arising from melanocytes in the skin. In recent decades the incidence of this malignancy has increased significantly. Mortality rates are high for patients with tumors measuring over a few millimeters in thickness. Response rates to conventional radiation and chemotherapy are very low in patients with metastatic melanoma. New therapies targeting melanoma’s aberrant cell signaling pathways such as the MAP Kinase pathway are being developed. Mutations of NRAS and BRAF genes are quite common in cutaneous melanoma and lead to constitutive activation of the MAP Kinase pathway. This study tests the hypothesis that NRAS and BRAF mutations increase as a tumor progresses from the noninvasive radial growth phase (RGP) to the invasive vertical growth phase (VGP). Laser capture microdissection was used to obtain separate, pure tumor DNA samples from the RGP and VGP of thirty primary cutaneous melanomas. PCR was used to amplify NRAS exon 2 and BRAF exon 15 tumor DNA. The amplified DNA was sequenced and analyzed for mutations. An overall mutation rate of 74% was obtained for the twenty-three melanomas in which there were complete sequence results. With the exception of one melanoma NRAS and BRAF mutations were mutually exclusive. All seven NRAS exon 2 mutations involved codon 61. Three of these melanomas had mutations in both the RGP and VGP. The remaining four tumors were wild type for NRAS exon 2 in the RGP but mutated in the VGP. Of the fifteen BRAF exon 15 mutated melanomas all but one involved codon 600. Twelve of the fifteen BRAF exon 15 mutations were the T1799A type. Nine of the fifteen BRAF mutated tumors had the same mutation in both the RGP and VGP. Five of fifteen melanomas had wild type RGP DNA and BRAF exon 15 mutated VGP DNA. A single melanoma had BRAF exon 15 mutated DNA in the RGP and wild type DNA in the VGP. Overall, these results suggest a trend toward the acquisition of NRAS and BRAF mutations as cutaneous melanomas change from a noninvasive to an invasive, potentially deadly cancer.^

The role and mechanism of supressor of cytokine signaling 1 in brain metastasis

Brain metastasis, which occurs in 40%-60% of patients with advanced melanoma, has led directly to death in the majority of cases. Unfortunately, little is known about the biological and molecular basis of melanoma brain metastases. In our previous study, we developed a model to study human melanoma brain metastasis and found that Stat3 activity was increased in human brain metastatic melanoma cells when compared with that in cutaneous melanoma cells. The increased activation of Stat3 is also responsible for affecting melanoma angiogenesis in vivo and melanoma cell invasion in vitro and significantly affecting the expression of bFGF, VEGF, and MMP-2 in vivo and in vitro. Interestingly, a member of a new family of cytokine-inducible inhibitors of signal transduction, termed suppressors of cytokine signaling 1 (SOCS1) was found to negatively regulate the Janus kinase signal transducer and activator of transcription (Jak/STAT) signaling cascade. Here we report that restoration of SOCS1 expression by transfecting of SOCS1-expressing vector effectively inhibited melanoma brain metastasis through inhibiting Stat3 activation and further affecting melanoma angiogenesis and melanoma cell invasion in vitro, and significantly affected the expression of vascular endothelial growth factor (VEGF) and matrix metalloproteinase-2 (MMP-2) in vitro and in vivo. In addition, we used cDNA array to compare mRNA expression in the SOCS1-transfected and vector-transfected cell lines and found some genes are tightly correlated to the restoration of SOCS1. One of them is Caveolin-1 (Cav-1). Cav-1 was reported to function as a tumor suppressor gene by several groups. Finally, the Cav-1 expression is up-regulated in SOCS1-overexpressing cell line. Further study found the regulation of Cav-1 by SOCS1 occurs through inhibiting Stat3 activation. Activated Stat3 binds directly to Cav-1 promoter and the Cav-1 promoter within -575bp is essential for active Stat3 binding. My studies reveal that Stat3 activation and SOCS1 expression play important roles in melanoma metastases. Moreover, the expression between SOCS1, Stat3 and Cav-1 forms a feedback regulation loop. ^

Fas ligand and soluble Fas -opposing molecules in melanoma lung metastasis

Deregulation of apoptotic cell death can result in aberrant accumulation of cells and increased tumor incidence. Fas (CD95) and Fas ligand (FasL) are a receptor-ligand pair whose activation induces apoptosis in many cell types. Previously, we demonstrated that low metastatic, Fas+ K1735-P murine melanoma cells spontaneously metastasize to the lung following orthotopic injection into FasL-deficient (gld) mice compared to wild-type (wt) controls. We further demonstrated that the expression of the Fas antagonist soluble Fas (sFas) directly correlates with disease stage in patients with melanoma, breast, and colon cancer. These findings document a role for host-derived FasL, in the control of metastatic disease and suggest a role for tumor-associated sFas in acquiring metastatic potential. To directly test whether FasL expressed on lymphocytes or on lung stromal cells restricts metastasis, bone marrow chimeras were generated between C3H wt and C3H gld mice. Chimeric animals were injected subcutaneously with 5 × 105 K1735-P and the incidence and number of spontaneous lung metastases scored. The data show that wt mice receiving gld marrow had a greater number of lung metastases (median 9.5, range 2–31) than gld mice reconstituted with wt marrow (median 1, range 0–31; p < 0.016). Interestingly, both groups had fewer metastases compared to gld controls (median 18.5, range 0–46) but more than wt controls (median 2, range 0–7). These observations provide the first evidence that both hematopoietic- and nonhematopoietic-host derived FasL, are important in the control of melanoma metastasis to the lung. To directly test whether tumor-associated sFas expression can enhance metastasis, K1735-P cells were transfected with three isoforms of sFas (Exo4Del, Exo6Del, and Exo3, 4, 6Del). RT-PCR and ELISA analysis confirmed the expression of sFas RNA and protein respectively. Following intravenous injection of 5 × 104 cells, sFas transfected cells formed significantly more experimental lung metastases [Exo6Del clone 3 (median 22, range 0–36), Exo6Del clone 7 (median 31, range 4–50), Exo3, 4, 6Del (median 22.5, range 13–48)] compared to vector control cells (median 6.5, range 3–29). Together, these data provide the first evidence that sFas is sufficient to enhance the metastatic potential of Fas+ melanoma cells. ^

MCAM/MUC18 REGULATES MELANOMA PROGRESSION BY MODULATING THE EXPRESSION OF ID-1

The acquisition of the metastatic melanoma phenotype is associated with increased expression of the melanoma cell adhesion molecule MCAM/MUC18 (CD146). However, the mechanism by which MUC18 contributes to melanoma metastasis remains unclear. Herein, we stably silenced MUC18 expression utilizing lentivirus-incorporated small hairpin RNA, in two metastatic melanoma cell lines, A375SM and C8161, and conducted cDNA microarray analysis. We identified and validated that the transcriptional regulator, Inhibitor of DNA Binding-1 (Id-1), previously shown to function as an oncogene in several malignancies, was downregulated by 5.6-fold following MUC18 silencing. Additionally, we found that MUC18 regulated Id-1 expression at the transcriptional level via ATF-3. Interestingly, ATF-3 was upregulated by 6.9 fold in our cDNA microarray analysis following MUC18 silencing. ChIP analysis showed increased binding of ATF-3 to the Id-1 promoter after MUC18 silencing, while mutation of the ATF-3 binding site on the Id-1 promoter increased Id-1 promoter activity in MUC18-silenced cells. These Data suggest that MUC18 silencing promotes inhibition of Id-1 expression by increasing ATF-3 expression and binding to the Id-1 promoter. Rescue of MUC18 reverted the expression of Id-1 and ATF-3, thus validating that they are not off-target effects of MUC18. To further assess the role of Id-1 in melanoma invasion and metastasis, we overexpressed Id-1 in MUC18-silenced cells. Overexpression of Id-1 in MUC18-silenced cells resulted in increased cell invasion, as well as increased expression and activity of MMP-2. Our data further reveal that Id-1 regulates MMP-2 at the transcriptional level through Sp1 and Ets-1. This is the first report to demonstrate that MUC18 does not act exclusively in cell adherence, but is also involved in cell signaling that regulates the expression of genes, such as Id-1 and ATF-3, thus contributing to the metastatic melanoma phenotype.

Loss of activator protein-2alpha results in overexpression of the thrombin receptor and correlates with the malignant phenotype of human melanoma

Increasing evidence demonstrates that the thrombin receptor (protease activated receptor-1, PAR-1) plays a major role in tumor invasion and contributes to the metastatic phenotype of human melanoma. We demonstrate that the metastatic potential of human melanoma cells correlates with overexpression of PAR-1. The promoter of the PAR-1 gene contains multiple putative AP-2 and Sp1 consensus elements. We provide evidence that an inverse correlation exists between the expression of AP-2 and the expression of PAR-1 in human melanoma cells. Re-expression of AP-2 in WM266-4 melanoma cells (AP-2 negative) resulted in decreased mRNA and protein expression of PAR-1 and significantly reduced the tumor potential in nude mice. ChIP analysis of the PAR-1 promoter regions bp −365 to −329 (complex 1) and bp −206 to −180 (complex 2) demonstrates that in metastatic cells Sp1 is predominantly binding to the PAR-1 promoter, while in nonmetastatic cells AP-2 is bound. In vitro analysis of complex 1 demonstrates that AP-2 and Sp1 bind to this region in a mutually exclusive manner. Transfection experiments with full-length and progressive deletions of the PAR-1 promoter luciferase constructs demonstrated that metastatic cells had increased promoter activity compared to low and nonmetastatic melanoma cells. Our data shows that exogenous AP-2 expression decreased promoter activity, while transient expression of Sp1 further activated expression of the reporter gene. Mutational analysis of complex 1 within PAR-1 luciferase constructs further demonstrates that the regulation of PAR-1 is mediated through interactions with AP-2 and Sp1. Moreover, loss of AP-2 in metastatic cells alters the AP-2 to Sp1 ratio and DNA-binding activity resulting in overexpression of PAR-1. In addition, we evaluated the expression of AP-2 and PAR-1 utilizing a tissue microarray of 93 melanocytic lesions spanning from benign nevi to melanoma metastasis. We report loss of AP-2 expression in malignant tumors compared to benign tissue while PAR-1 was expressed more often in metastatic melanoma cells than in benign melanocytes. We propose that loss of AP-2 results in increased expression of PAR-1, which in turn results in upregulation of gene products that contribute to the metastatic phenotype of melanoma. ^

Galectin-3 Enhances the Malignant Melanoma Phenotype by Regulating Autotaxin

In melanoma patient specimens and cell lines, the over expression of galectin-3 is associated with disease progression and metastatic potential. Herein, we have sought out to determine whether galectin-3 affects the malignant melanoma phenotype by regulating downstream target genes. To that end, galectin-3 was stably silenced by utilizing the lentivirus-incorporated small hairpin RNA in two metastatic melanoma cell lines, WM2664 and A375SM, and subjected to gene expression microarray analysis. We identified and validated the lysophospholipase D enzyme, autotaxin, a promoter of migration, invasion, and tumorigenesis, to be down regulated after silencing galectin-3. Silencing galectin-3 significantly reduced the promoter activity of autotaxin. Interestingly, we also found the transcription factor NFAT1 to have reduced protein expression after silencing galectin-3. Electrophoretic mobility shift assays from previous reports have shown that NFAT1 binds to the autotaxin promoter in two locations. ChIP analysis was performed, and we observed a complete loss of bound NFAT1 to the autotaxin promoter after silencing galectin-3 in melanoma cells. Mutation of the NFAT1 binding sites at either location reduces autotaxin promoter activity. Silencing NFAT1 reduces autotaxin expression while over expressing NFAT1 in NFAT1 negative SB-2 melanoma cells induces autotaxin expression. These data suggest that galectin-3 silencing reduces autotaxin transcription by reducing the amount of NFAT1 protein expression. Rescue of galectin-3 rescues both NFAT1 and autotaxin. We also show that the re-expression of autotaxin in galectin-3 shRNA melanoma cells rescues the angiogenic phenotype in vivo. Furthermore, we identify NFAT1 as a potent inducer of tumor growth and experimental lung metastasis. Our data elucidate a previously unidentified mechanism by which galectin-3 regulates autotaxin and assign a novel role for NFAT1 during melanoma progression.

The molecular cloning and characterization of human heparanase cDNA and the immunochemical localization of heparanase in metastatic melanomas

Heparanase, an endo-$\beta$-D-glucuronidase, has been associated with melanoma metastasis. Polyclonal antibodies directed against the murine N-terminal heparanase peptide detected a M$\sb{\rm r}\sim 97,000$ protein upon SDS-polyacrylamide gel electrophoresis of mouse melanoma and human melanoma cell lysates. In an indirect immunocytochemical study, metastatic human A375-SM and mouse B16-BL6 melanoma cells were stained with the anti-heparanase antibodies. Heparanase antigen was localized in the cytoplasm of permeabilized melanoma cells as well as at the cell surface of unpermeabilized cells. Immunohistochemical staining of frozen sections from syngeneic mouse organs containing micrometastases of B16-BL6 melanoma demonstrated heparanase localized in metastatic melanoma cells, but not in adjacent normal tissues. Similar studies using frozen sections of malignant melanomas resected from patients indicated that heparanase is localized in invading melanoma cells, but not in adjacent connective tissues.^ Monoclonal antibodies directed against murine heparanase were developed and characterized. Monoclonal antibody 10E5, an IgM, precipitated and inhibitated the enzymatic activity of heparanase. A 2.6 kb cDNA was isolated from a human melanoma $\lambda$gt11 cDNA library using the monoclonal antibody 10E5. Heparan sulfate cleavage activity was detected in the lysogen lysates from E. Coli Y1089 infected with the $\lambda$gt11 cDNA and this activity was inhibited in the presence of 10-fold excess of heparin, a potent inhibitor of heparanase. The nucleotide sequence of the cDNA was determined and insignificant homology was found with the gene sequences currently known. The cDNA hybridized to a 3.2-3.4 kb mRNA in human A375 melanoma, WI-38 fibroblast, and THP-1 leukemia cells using Northern blots.^ Heparanase expression was examined using Western and Northern blots. In comparison to human A375-P melanoma cells, the quantity of 97,000 protein recognized by the polyclonal anti-heparanase antibodies doubled in the metastatic variant A375-SM cells and the quantity of 3.2-3.4 kb mRNA doubled in A375MetMix, a metastatic variant similar to A375-SM cells. In B16 murine melanoma cell, the intensity of the 97,000 protein increased more than 2 times comparing with B16-F1 cells. The extent in the increase of the protein and the mRNA levels is comparable to the change of heparanase activity observed in those cells.^ In summary, the studies suggest that (a) the N-terminus of the heparanase molecule in mouse and human is antigenically related; (b) heparanase antigens are localized at the cell surface and in the cytoplasm of metastatic human and mouse melanoma cells; (c) heparanase antigens are localized in invasive and metastatic murine and human melanomas in vivo, but not in adjacent normal tissues; (d) heparanase molecule appeared to be differentially expressed at the transcriptional as well as at the translational level; and (e) the size of human heparanase mRNA is 3.2-3.4 kilobase. ^

BAYESIAN STATISTICAL METHODS IN GENE-ENVIRONMENT AND GENE-GENE INTERACTION STUDIES

Complex diseases such as cancer result from multiple genetic changes and environmental exposures. Due to the rapid development of genotyping and sequencing technologies, we are now able to more accurately assess causal effects of many genetic and environmental factors. Genome-wide association studies have been able to localize many causal genetic variants predisposing to certain diseases. However, these studies only explain a small portion of variations in the heritability of diseases. More advanced statistical models are urgently needed to identify and characterize some additional genetic and environmental factors and their interactions, which will enable us to better understand the causes of complex diseases. In the past decade, thanks to the increasing computational capabilities and novel statistical developments, Bayesian methods have been widely applied in the genetics/genomics researches and demonstrating superiority over some regular approaches in certain research areas. Gene-environment and gene-gene interaction studies are among the areas where Bayesian methods may fully exert its functionalities and advantages. This dissertation focuses on developing new Bayesian statistical methods for data analysis with complex gene-environment and gene-gene interactions, as well as extending some existing methods for gene-environment interactions to other related areas. It includes three sections: (1) Deriving the Bayesian variable selection framework for the hierarchical gene-environment and gene-gene interactions; (2) Developing the Bayesian Natural and Orthogonal Interaction (NOIA) models for gene-environment interactions; and (3) extending the applications of two Bayesian statistical methods which were developed for gene-environment interaction studies, to other related types of studies such as adaptive borrowing historical data. We propose a Bayesian hierarchical mixture model framework that allows us to investigate the genetic and environmental effects, gene by gene interactions (epistasis) and gene by environment interactions in the same model. It is well known that, in many practical situations, there exists a natural hierarchical structure between the main effects and interactions in the linear model. Here we propose a model that incorporates this hierarchical structure into the Bayesian mixture model, such that the irrelevant interaction effects can be removed more efficiently, resulting in more robust, parsimonious and powerful models. We evaluate both of the 'strong hierarchical' and 'weak hierarchical' models, which specify that both or one of the main effects between interacting factors must be present for the interactions to be included in the model. The extensive simulation results show that the proposed strong and weak hierarchical mixture models control the proportion of false positive discoveries and yield a powerful approach to identify the predisposing main effects and interactions in the studies with complex gene-environment and gene-gene interactions. We also compare these two models with the 'independent' model that does not impose this hierarchical constraint and observe their superior performances in most of the considered situations. The proposed models are implemented in the real data analysis of gene and environment interactions in the cases of lung cancer and cutaneous melanoma case-control studies. The Bayesian statistical models enjoy the properties of being allowed to incorporate useful prior information in the modeling process. Moreover, the Bayesian mixture model outperforms the multivariate logistic model in terms of the performances on the parameter estimation and variable selection in most cases. Our proposed models hold the hierarchical constraints, that further improve the Bayesian mixture model by reducing the proportion of false positive findings among the identified interactions and successfully identifying the reported associations. This is practically appealing for the study of investigating the causal factors from a moderate number of candidate genetic and environmental factors along with a relatively large number of interactions. The natural and orthogonal interaction (NOIA) models of genetic effects have previously been developed to provide an analysis framework, by which the estimates of effects for a quantitative trait are statistically orthogonal regardless of the existence of Hardy-Weinberg Equilibrium (HWE) within loci. Ma et al. (2012) recently developed a NOIA model for the gene-environment interaction studies and have shown the advantages of using the model for detecting the true main effects and interactions, compared with the usual functional model. In this project, we propose a novel Bayesian statistical model that combines the Bayesian hierarchical mixture model with the NOIA statistical model and the usual functional model. The proposed Bayesian NOIA model demonstrates more power at detecting the non-null effects with higher marginal posterior probabilities. Also, we review two Bayesian statistical models (Bayesian empirical shrinkage-type estimator and Bayesian model averaging), which were developed for the gene-environment interaction studies. Inspired by these Bayesian models, we develop two novel statistical methods that are able to handle the related problems such as borrowing data from historical studies. The proposed methods are analogous to the methods for the gene-environment interactions on behalf of the success on balancing the statistical efficiency and bias in a unified model. By extensive simulation studies, we compare the operating characteristics of the proposed models with the existing models including the hierarchical meta-analysis model. The results show that the proposed approaches adaptively borrow the historical data in a data-driven way. These novel models may have a broad range of statistical applications in both of genetic/genomic and clinical studies.

Cutaneous malignant melanoma: A study of skin cancer in the United States from 1973--1997

Cutaneous malignant melanoma (CMM) is the cancer of the melanocytes, the cells that produce the pigment melanin, and is an aggressive skin cancer that is most prevalent in the white population. Although most cases of malignant melanoma are white, black and other non-white populations also develop this disease. However, the etiologic factors involved in the development of melanoma in these lower-risk populations are not well known. Generally, survival rates of malignant melanoma have been found to be lower in blacks than for whites with similar stage of disease at diagnosis. ^ This study presents an analysis of the differences in survival between black and white cases with malignant melanoma of the skin as the only or first primary cancer, found in the National Cancer Institute Surveillance, Epidemiology and End Results (SEER) cancer registry from 1973 to 1997. A total of 54,193 cases of CMM were diagnosed in black and white patients between 1973 and 1997. Black patients tended to be older, with a mean age of 64.46 years, compared to 53.14 years for white patients. Eighty-nine percent of patients were diagnosed with CMM as the only cancer. (Abstract shortened by UMI.)^

Protease-activated receptor-1 signaling plays a major role in melanoma growth and metastasis

Overexpression of the thrombin receptor (Protease-Activated-Receptor-1), PAR-1, in cell lines and tissue specimens correlates with the metastatic potential of human melanoma. Utilizing lentiviral shRNA to stably silence PAR-1 in metastatic melanoma cell lines results in decreased tumor growth and lung metastasis in vivo. Since the use of viral technology is not ideal for clinical therapies, neutral liposomes (DOPC) were utilized as a delivery vehicle for PAR-1 siRNA. Our data suggest that PAR-1 siRNA-DOPC treatment by systemic delivery significantly decreases tumor growth and lung metastasis in nude mice. Concomitant decreases in angiogenic and invasive factors (IL-8, VEGF, MMP-2) were observed in PAR-1 siRNA-DOPC-treated mice. Utilizing a cDNA microarray platform, several novel PAR-1 downstream target genes were identified, including Connexin 43 (Cx-43) and Maspin. Cx-43, known to be involved in tumor cell diapedesis and attachment to endothelial cells, is decreased after PAR-1 silencing. Furthermore, the Cx-43 promoter activity was significantly inhibited in PAR-1-silenced cells suggesting transcriptional regulation of Cx-43 by PAR-1. ChIP analysis revealed a reduction in SP-1 and AP-1 binding to the Cx-43 promoter. Moreover, melanoma cell attachment to HUVEC was significantly decreased in PAR-1-silenced cells as well as in Cx-43 shRNA transduced cells. As both SP-1 and AP-1 transcription factors act as positive regulators of Cx-43, our data provide a novel mechanism for the regulation of Cx-43 expression by PAR-1. Maspin, a serine protease inhibitor with tumor-suppressor function, was found to be upregulated after PAR-1 silencing. Our results indicate that PAR-1 transcriptionally regulates Maspin, as the promoter activity was significantly increased after PAR-1 silencing. ChIP analysis revealed that silencing PAR-1 increased binding of Ets and c-Jun to the Maspin promoter. As Maspin was recently found to be a tumor-suppressor in melanoma by reducing the invasive capacity of melanoma cells, invasion assays revealed a decrease in invasion after PAR-1 silencing and in cells transduced with a Maspin expression vector. We propose that PAR-1 is key to the progression and metastasis of melanoma in part by regulating the expression of Cx-43 and Maspin. Taken together, we propose that PAR-1 is an attractive target for the treatment of melanoma.^

The role of reactive astrocytes in the resistance of melanoma brain metastasis to chemotherapy

Brain metastasis is resistant to chemotherapy while the leaky blood-brain-barrier in brain metastasis can not be the underlying reason. Metastatic tumor cells (“seed”) exploit the host microenvironment (“soil”) for survival advantages. Astrocytes which maintain the homeostasis of the brain microenvironment become reactive subsequent to brain damages and protect neurons from various injuries. We observed reactive astrocytes surrounding and infiltrating into brain metastasis in both clinical specimen and experimental animal model, thus raising a possibility that reactive astrocytes may protect tumor cells from cytotoxic chemotherapeutic drugs. ^ To test this hypothesis, we first generated an immortalized astrocyte cell line from H-2Kb-tsA58 mice. The immortal mouse astrocytes expressed specific markers including GFAP. Scanning electron microscopy demonstrated that astrocytes formed direct physical contact with tumor cells. Moreover, the expression of GFAP by astrocytes was up-regulated subsequent to co-culture with tumor cells, indicating that the co-culture of astrocytes and tumor cells may serve as a model to recapitulate the pathophysiological situation of brain metastasis. ^ In co-culture, astrocytes dramatically reduced apoptosis of tumor cells produced by various chemotherapeutic drugs. This protection effect was not because of culturing cells from different species since mouse fibroblasts did not protect tumor cells from chemotherapy. Furthermore, the protection by astrocytes was completely dependent on a physical contact. ^ Gap junctional communication (GJC) served as this physical contact. Tumor cells and astrocytes both expressed the major component of gap junctional channel—connexin 43 and formed functional GJC as evidenced by the “dye transfer” assay. The blockage of GJC between tumor cells and astrocytes by either specific chemical blocker carbenoxolone (CBX) or by genetically knocking down connexin 43 on astrocytes reversed the chemo-protection. ^ Calcium was the signal molecule transmitted through GJC that rescued tumor cells from chemotherapy. Accumulation of cytoplasmic calcium preceded the progress of apoptosis in tumor cells treated with chemotherapeutic drugs. Furthermore, chelation of accumulated cytoplasmic calcium inhibited the apoptosis of tumor cells treated with chemotherapeutic drugs. Most importantly, astrocytes could “shunt” the accumulated cytoplasmic calcium from tumor cells (treated with chemotherapeutic drug) through GJC. We also used gene expression micro-array to investigate global molecular consequence of tumor cells forming GJC with astrocytes. The data demonstrated that astrocytes (but not fibroblasts), through GJC, up-regulated the expressions of several well known survival genes in tumor cells. ^ In summary, this dissertation provides a novel mechanism underlying the resistance of brain metastasis to chemotherapy, which is due to protection by astrocytes through GJC. Interference with the GJC between astrocytes and tumor cells holds great promise in sensitizing brain metastasis to chemotherapy and improving the prognosis for patients with brain metastasis. ^

Role and regulation of c-KIT gene expression in tumor growth and metastasis of human melanoma

The purpose of this study was to investigate the role of the c-KIT receptor in the progression of human melanoma and the mechanism(s) for the regulation of c-KIT gene expression in human melanoma.^ The molecular changes associated with the transition of melanoma cells from radial growth phase (RGP) to vertical growth phase (VGP) (metastatic phenotype) are not well-defined. Expression of the tyrosine-kinase receptor c-KIT progressively decreases during local tumor growth and invasion of human melanomas. To provide direct evidence that the metastasis of human melanoma is associated with the loss of c-KIT expression, highly metastatic A375SM cells, which express very low or undetectable levels of c-KIT, were tranduced with the human c-KIT gene. We demonstrated that enforced c-KIT expression in highly metastatic human melanoma cells significantly suppressed their tumorigenicity and metastatic propensity in nude mice. In addition, we showed that the ligand for c-KIT, SCF, induces apoptosis in human melanoma cells expressing c-KIT under both in vitro and in vivo conditions. These results suggest that loss of c-KIT receptor may allow malignant melanoma cells to escape SCF/c-KIT-mediated apoptosis, thus contributing to tumor growth and eventually metastasis.^ Furthermore, we investigated the possible mechanism(s) for the down-regulation of c-KIT gene expression in malignant melanoma. Sequence analysis of the c-KIT promoter indicated that this promoter contains several consensus binding-site sequences including three putative AP2 and two Myb sites. Although Myb was shown to be associated with c-KIT expression in human hemotopoietic cells, we found no correlation between c-KIT expression and Myb expression in human melanoma cell lines. In contrast, we showed that c-KIT expression directly correlates with expression of AP2 in human melanoma cells. We found that highly metastatic cells do not express the transcription factor AP2. Expression of AP2 in A375SM cells (c-KIT-negative and AP2-negative) was enough to restore luciferase activity driven by the c-KIT promoter in a dose-dependent manner. On the other hand, co-expression of the dominant-negative form of AP2 (AP2B) in Mel-501 cells (c-KIT-positive and AP2-positive) resulted in two-fold reduction in luciferase activity. Electrophoretic mobility shift assays revealed that the c-KIT promoter contains functional AP2 binding sites which could associate with AP2 protein. Endogenous c-KIT gene expression levels were elevated in AP2 stably-transfected human melanoma A375SM cells. Expression of exogenous AP2 in A375SM cells inhibited their tumorigenicity and metastatic potential in nude mice. The c-KIT ligand, SCF, also induced apoptosis in the AP2 stably-transfected A375SM cells. The identification of AP2 as an important regulator for c-KIT expression suggests that AP2 may have tumor growth and metastasis inhibitory properties, possibly mediated through c-KIT/SCF effects on apoptosis of human melanoma cells. Since AP2 binding sites were found in the promoters of other genes involved in the progression of human melanoma, such as MMP2 (72 kDa collagenase), MCAM/MUC18 and P21/WAF-1, our findings suggest that loss of AP2 expression might be a crucial event in the development of malignant melanoma. ^

ROLES FOR BRAF KINASE ACTIVATING MUTATIONS IN MELANOMA: MICROENVIRONMENTAL IMMUNOSUPPRESSION

The BRAF oncogene demonstrates a characteristic mutation (V600E) in a significant fraction of cutaneous melanomas, leading to constitutive activation of the MAP kinase pathway. This genetic lesion endows tumor cells with proliferative and survival advantages, and metastatic melanoma patients treated with the BRAF(V600E)-specific inhibitor, Vemurafenib, have shown dramatic clinical responses. Here, I show that BRAF(V600E) induces transcription of the IL-1α and IL-1β genes in both melanocytes and melanoma cell lines and that this upregulation is specifically abrogated by targeted BRAF(V600E) inhibitors. Furthermore, treatment of melanoma tumor-associated fibroblasts (TAFs) with IL-1α/β significantly enhanced the ability of TAFs to suppress the proliferation and function of melanoma antigen-specific cytotoxic T cells. IL-1α/β treatment of TAFs upregulated multiple immunosuppressive factors, including COX-2 and the PD-1 ligands PD-L1 and PD-L2. Specific BRAF(V600E) inhibitors largely abrogated the ability of melanoma cells to confer T cell-suppressive properties on TAFs. These results support a model in which BRAF(V600E) promotes immune suppression in the melanoma tumor environment through an IL-1-mediated mechanism involving resident stromal fibroblasts. Based on these findings, combination therapies involving targeted BRAF inhibition and T cell-based immunotherapies are warranted.

Host factors in metastasis: Direct lysis of bloodborne metastatic tumor cells by murine vascular endothelial cells

During the process of cancer metastasis, the majority of circulating tumor cells arrest in microcapillary beds and then rapidly die. To study whether vascular endothelial cells can directly lyse tumor cells, we isolated vascular endothelial cells by perfusion of lungs from immunocompetent or nude mice. The cells were grown in culture, and then cloned and characterized. Cloned endothelial cells were incubated with several lymphokines and cytokines. Cells incubated with IFN-$\gamma$ and TNF lysed a variety of tumor cells with different metastatic potential. Mouse skin and lung fibroblasts treated with the same cytokines did not. Endothelial cell mediated tumor cell lysis was not due to different binding ability of tumor cells to cytokine treated and untreated endothelial monolayers. Kinetic studies demonstrated that the continuous presence of cytokines in the tumor-endothelial cocultures was necessary to produce maximal lysis of tumor cells. Target cell lysis was not due to the direct effects of IFN-$\gamma$ or TNF, since vascular endothelial cells isolated from the lung of nude mice lysed human melanoma cells that are sensitive or resistant to TNF. Cytokine treated endothelial cells produced a high level of nitric oxide, which is known to be cytotoxic to a variety of target cells. The level of nitric oxide production was directly correlated with the degree of tumor cell lysis. A specific inhibitor of nitric oxide synthesis(N$\sp{\rm G}$-monomethyl-L-arginine), completely inhibited production of nitric oxide and tumor cell lysis. Treatment of cytokine activated endothelial cells with dexamethasone also inhibited tumor cell lysis. This inhibition was independent of tumor-endothelial adhesion but correlated with inhibition of nitric oxide production. Collectively, these results suggest that vascular endothelial cells can directly destory tumor emboli and thus play an active role in the pathogenesis of cancer metastasis. ^