Search

Brain metastasis confers an extremely unfavorable prognosis upon melanoma patients. The mechanisms underlying the homing of metastatic melanoma to the brain and survival of metastatic melanoma cells in the brain are unknown. Tumor cells, including melanoma, use chemokine receptor-ligand axes to home to specific organ sites. To identify chemokine receptors that might be involved in brain-targeted melanoma metastasis, we first established a chemokine receptor profile of cultured melanoma cells (3 cell lines of cutaneous melanoma and 5 cell lines of melanoma brain metastasis). The expression of the membrane-bound chemokine CX3CL1 by these lines was also determined. We show that out of 19 receptors tested, cultured melanoma cells express CCR3, CCR4, CXCR3, CXCR7, CX3CR1 and membrane CX3CL1. Utilizing cells from newly created variants of human melanoma xenografts, we found that the expression of CCR4 was significantly higher in one brain metastatic variant compared to its expression in the corresponding local variant. Local and metastatic variants stimulated with the CCR4 ligand, CCL22, showed a differential AKT phosphorylation pattern. These findings may suggest the involvement of CCR4 in the process of brain metastasis in human melanoma, and that CCR4 may be a novel molecular biomarker for the identification of melanoma cells likely to metastasize to the brain.

The dual-specificity phosphatase Pyst2-L was found to be over expressed in leukocytes derived from AML and ALL patients as well as in certain other solid tumors and lymphoblastoid cell lines. Pyst2-L, binds and dephosphorylates both pERKs and pJNKs proteins, and thus, plays a role in regulating the MAP kinase signaling pathway. In the present study, a comparative genomic application was used and sequence analysis of multi-organisms databases were searched in order to identify genes homologous to Pyst2-L. The Xenopus laevis MAP kinase phosphatase X17c gene and the Yeast nitrogen starvation-induced protein phosphatase Yvh1p gene were revealed to be highly homologous with Pyst2-L. Both X17c and Yvh1p genes play a role in cell cycle regulation. A down regulated expression of the Yvh1p gene occurred in Saccharomyces cerevisiae that were synchronized to the G2-phase of the cell cycle by alpha-factor. In conformity with this result, a reduction in Pyst2-L expression levels was observed in G2-phase-synchronized Human K562 cells. Finally, we were able to show that cells in highly crowded cultures express high levels of the Pyst2-L phosphatase. These observations may indicate that low levels of the Pyst2-L phosphatase are essential for the G2-phase of the cell cycle and that this phosphatase might play a role in signaling cascades induced by cellular crowding.

About 70 years ago, Peyton Rous described the progression of cancer towards metastasis formation as "the process whereby tumors go from bad to worse". The interactions of tumor cells with endothelium are pivotal steps in this process. This review focuses on the role played by the selectins and their ligands in these interactions and especially in tumor cell extravasation. The working hypothesis of researchers studying tumor cell extravasation is that the tumor cells follow the extravasation strategy of leukocytes in their migration towards inflammatory sites. A significant portion of this review is, therefore, dedicated to the molecular mechanisms underlying leukocyte extravasation and to a comparison between the extravasation strategy employed by leukocytes and tumor cells. The review also summarizes some of the available data on signals generated by selectin-selectin ligand interactions.

The dual-specificity phosphatase Pyst2-L was found to be highly expressed in leukocytes derived from AML and ALL patients as well as in certain other solid tumors and lymphoblastoid cell lines. Recently, by use of the 5′-RNA ligation-mediated rapid amplification of cDNA ends (5′-RLM-RACE) technique, we sequenced and cloned the entire open reading frame (ORF) of Pyst2-L. In the present study we determined the effect of exogenous overexpression on Erk1/2 phosphorylation. It was demonstrated that overexpression of this phosphatase in HEK293 cells reduced the basal levels of phospho-Erk1/2 as compared to the same cells transfected with the wild-type vector. This reduction was concomitant with a growth retardation of the Pyst2-L-transfected cells. Treating Pyst2-L transfected cells with known activators of the MAPK signaling cascade such as TPA or stimulating them by serum, it was demonstrated that the up regulation of phospho-Erk1/2, caused by these activators, was only partially suppressed by the over expression of the Pyst2-L phosphatase in these cells. These results together with our previous ones showing that the TPA-induced up regulation of Pyst2-L mRNA was only partially inhibited by the use of a specific Mek1/2 inhibitor, lead us to ask whether the Pyst2-L phosphatase has a monogamous relationship with the Erk2 protein. To answer this question, we employed the pull-down method and showed that in addition to phospho-Erk1/2, recombinant Pyst2-L binds the phospho-JNK protein. These findings may raise new perspectives regarding the role played by this phosphatase in malignant cells and in activation processes.

In previous studies, we demonstrated that human neuroblastoma cells are equipped with the machinery to direct their homing to bone marrow. These tumor cells express the CXCR4 receptor for the bone marrow stroma-derived chemokine CXCL12 (SDF-1) and secrete the CXCL12 ligand. The present study was undertaken to explore possible differences in gene-expression patterns between neuroblastoma variants that over-express CXCR4 (designated STH cells) and those which express very little of this receptor (STL cells). The results of the study clearly indicate that these variants show a differential gene-expression profile. They differ in expression of some integrins such as VLA2, VLA3 and VLA6, of neuroendocrine-markers such as CD56 and synaptophysin, in the expression of c-kit and in the secretion of certain cytokines and growth factors such as TNFalpha, SDF-1, VEGF, IL-8, GM-CSF and IP-10. We hypothesize that these differences are due to an autocrine SDF-1alpha-CXCR4 axis.

The extravasation of tumor cells is a pivotal stage in the formation of hematogenous metastasis. An interaction of selectins expressed on endothelial cells and selectin ligands expressed by tumor cells has been implicated to play a role in extravasation. In the present study we used a human-mouse model to prove the hypothesis that the selectin ligand sialyl Lewis-a (sLe-a) is indeed involved in the in vivo extravasation of colorectal carcinoma (CRC) cells. The results indicated that highly metastatic CRC cells expressing high levels of sLe-a extravasate more efficiently than non-metastatic CRC cells expressing low levels of sLe-a. It was also demonstrated that down regulating the expression levels of sLe-a in CRC cells by genetic manipulations, significantly reduced CRC extravasation. Non-specific effects of these manipulations were ruled out. The results of this study indicate that the arrest and adhesion of CRC cells, and possibly of other types of cancer cells as well, to endothelium depend on the expression of the selectin ligand sLe-a by the tumor cells.

Tumor progression is a multistep process in which alterations in the expression of numerous gene products may give rise to highly malignant cellular variants. In the present study, we analyzed the differential expression of several genes in cellular variants of mammary adenocarcinomas with high or low malignancy potential, which originated in a common ancestor. To assess the generality of our findings, high and low malignancy variants were derived from two different mammary adenocarcinoma cell lines, namely DA3 and CSML cells. Of major importance is the fact that the differences between high- and low-malignancy variants observed in one system of mammary adenocarcinoma cells (DA3 cells) were identically reproduced in the other system of mammary adenocarcinoma cells (CSML cells). The high malignancy variants of tumors both DA3-high and CSML-high (previously called CSML-100), expressed higher levels of factors that induce monocyte migration than the low malignancy DA3-low and CSML-low (previously called CSML-0) variants. In addition, it was found that DA3-high and CSML-high cell variants expressed higher levels of monocyte chemoattractant protein-1 (MCP-1), interleukin-6 (IL-6) and matrix metalloproteinases (MMPs) than the low malignancy variants (DA3-low and CSML-low). These results suggest that MCP-1, IL-6 and MMPs potentially contribute to mammary adenocarcinoma progression and that their expression is regulated by a common pathway. The expression of MCP-1, IL-6 and MMPs in both DA3-high and CSML-high cells was up-regulated by tumor necrosis factor alpha (TNFalpha). The fact that TNFalpha exerted similar effects on the expression of these three factors in both cell systems raises the possibility of a coordinated co-regulation of tumor-promoting factors.

The search for mechanisms that regulate tumor progression has motivated the authors' laboratory to establish a unique murine model system, consisting of two lines of DA3 mammary adenocarcinoma cells that were derived originally from a common ancestor but differed in their malignant potential. Studies indicated that the highly malignant phenotype manifested by one of the cell lines (termed Ly-6hi DA3 cells) was associated with high expression of the Ly-6E.1 antigen. To characterize the mechanisms controlling the high malignancy phenotype expressed by Ly-6hi DA3 cells, the study was focussed on the potential contribution of tumor-derived factors to the high malignancy phenotype expressed by these cells. To this end, the expression of CC chemokines, major chemoattractants of monocytes and T cells, by the highly malignant Ly-6hi DA3 cells as compared to the low malignancy Ly-6lo DA3 cells was evaluated. The results indicate that the highly malignant cells express higher levels of factors that induce monocyte migration than the low malignancy cells. Two CC chemokines were shown to be highly produced by Ly-6hi DA3 cells, MIP-1alpha and MCP-1, of which only the latter was shown to contribute to the high migratory activity expressed by the high malignancy Ly-6hi DA3 cells. Since MCP-1 may attract monocytes to tumor sites, these findings suggest that monocyte-derived mediators, such as growth factors or angiogenic cytokines, have pro-malignancy effects that contribute to the high malignancy phenotype expressed by Ly-6hi DA3 cells.

Angiogenic and poorly angiogenic tumor variants were obtained by an intraperitoneal inoculation of cells from clones of polyoma-virus transformed BALB c 3T3 cells into syngeneic mice. The angiogenic tumor cells expressed a higher tumorigenicity phenotype and a higher capacity to produce artificial pulmonary metastases than cells from the poorly angiogenic tumors. The former cells expressed also significantly higher levels of the lymphocyte activation protein Ly-6E.1 than the former cells. The two types of cells did not differ in expression levels of CD44 and of a polyoma-virus specific membrane antigen. These results raise the possibility that the angiogenic phenotype is coregulated with Ly-6. The effect on Ly-6 expression of signal transduction through TNF receptors, functioning as pivotal regulators of angiogenesis was therefore studied. It was found that TNFα and more so antibodies against Fas down-regulate expression levels of Ly-6. This down-regulation seemed to be selective as expression levels of CD44 were not affected by this treatment.

Previous results obtained in our laboratory suggested that natural antibodies reactive with L5178Y lymphoma cells play a role in the induction of lung tumors by the chemical carcinogen urethane. In order to characterize some of the naturally-occurring L5178Y reactive antibodies we prepared hybridomas that secreted natural monoclonal IgM antibodies reactive with L5178Y lymphoma cells. In the present study we characterized some of these antibodies and provided further proof as to their role in urethane carcinogenesis. One hybridoma secreted a cytotoxic antibody that reacted only with mouse lymphoma cell lines. Other non-cytotoxic monoclonal L5178Y-reactive antibodies showed various degrees of cross-reactivity with syngeneic, allogeneic and xenogeneic cells of normal or malignant origin. One of these antibodies reacted much better with activated T cells than with resting ones. Four groups of mice were treated with urethane. Three groups were injected twice a week during 5 months with different IgM preparations of natural monoclonal antibodies. The mice in the fourth group were not treated with IgM and served as controls. Five months after the urethane treatment the mice were sacrificed and the number of tumor foci in the lungs of each mouse was determined. The results show that the group treated with the cytotoxic monoclonal antibody 1.80 had a significant decrease, while the group treated with the IgM myeloma protein 104E had a significant increase in the number of tumor foci compared to urethane-treated mice that did not receive any IgM treatment. The tumor incidence in the group treated with the 104E IgM was significantly higher than that in the control group.

Different clones of mouse hybridomas, derived from the fusion of unstimulated mouse peritoneal cells with mouse myeloma cells, producing IgM monoclonal antibodies directed against the membrane of bromelain-treated mouse erythrocytes (MRBC(Br)) have been previously established. We have recently shown that one of these hybridomas produce, in ascites, antibodies cross-reacting with phosphorylcholine derivatives (trimethylammonium (TMA) derivatives). In this work the cross-reactivity for TMA derivatives of the monoclonal antibodies produced by 4 anti-MRBC(Br) hybridomas have been studied at the cell level (plaque-forming cells). Phosphorylcholine, choline bromide and p -aminophenyl-trimethylammonium were found to be potent specific inhibitors of plaque formation (anti MRBC(Br)). The hemolytic activities of ascites and tissue culture supernatants were studied and their inhibition by TMA derivatives was determined. Immunoglobulins from ascites purified on TMA immunoadsorbant column were analyzed by two-dimensional gel electrophoresis, their spectrotype was compared to the spectrotype of immunoglobulins from tissue culture supernatants from the same hybridoma radioactively tagged by internal incorporation of [ 14 C]leucine. It could be shown without ambiguity that the PTMA column retained an IgM with the same characteristics as the IgM secreted in vitro.

Monoclonal IgM anti-erythrocyte autoantibody produced by a NZB-derived hybridoma has been found to react with trimethylammonium-containing compounds. Such compounds are able to prevent the lysis of bromelain-treated mouse erythrocytes (BrMRBC) by those autoantibodies. Using a column of insolubilized betaine hydrazide (BH) the monoclonal anti-erythrocyte antibody has been specifically retained. Elution of this antibody was accomplished by 0.15 M choline (Ch).

Radioimmunoassays (RIA) have confirmed previous studies that trimethylammonium (TMA) or its derivatives constitute part of the determinant recognized by naturally-occurring antibodies (NOA) with the ability to interact with bromelain-treated mouse erythrocytes (BrMRBC). Further studies on this determinant revealed its presence on erythrocytes from several species in addition to mice. In most cases (except in chickens) the determinant was cryptic and could be exposed only after proteolytic treatment of the erythrocytes. The determinant was also found on certain murine lymphoma cells. We also found that bromelain was not the only enzyme that could be used to expose the determinant. Papain, but not trypsin, was able to unmask the determinant on mouse erythrocytes. Rabbit antibodies directed against the idiotypes of four different monoclonal BrMRBC-binding NOA were prepared. Direct RIA assays and inhibition assays showed that the different monoclonal BrMRBC-binding NOA shared a common idiotype specific to such antibodies. The common idiotype was detected in the serum from several mouse strains and in wild mice.

A natural IgM monoclonal antibody, 1.67, was generated from apparently healthy unstimulated BALB/c mice. This antibody reacted with L5178Y murine T cell lymphoma, with human Raji cells, and with several normal cells. Further analysis of its ligand binding capacity disclosed strong binding to single-stranded DNA (ssDNA). However, this naturally-occurring monoclonal antibody binds to different epitopes on cell membranes and on DNA than another anti-DNA monoclonal antibody (18/103/1) from human origin. This conclusion was based on competition assays. Furthermore, NOA 1.67 lacks the 16/6 idiotype expressed on the 18/103/1 antibody. The 16/6 idiotype is shared by human and mouse lupus monoclonal autoantibodies that bind simultaneously to lymphoid cells and DNA. This is a first report on a natural autoantibody that binds to malignant and to normal cell membrane(s) as well as to ssDNA. It may have regulatory functions controlling malignancy and or autoimmunity.