C32 human melanoma cell endogenous lectins: characterization and implication in vesicle-mediated melanin transfer to keratinocytes.

To optimize skin pigmentation in order to help body prevention against UV radiation, the mechanism of melanin pigment transfer from melanocytes to keratinocytes must be elucidated. Melanin transfer to keratinocytes requires specific recognition between keratinocytes and melanocytes or melanosomes. Cell surface sugar-specific receptor (membrane lectin) expression was studied in human C32 melanoma cells, an amelanotic melanoma, by flow cytometry analysis of neoglycoprotein binding as an approach to the molecular specificity. Sugar receptors on melanocytes are mainly specific for alpha-L-fucose. Their expression is enhanced upon treatment by the diacylglycerol analogue 1-oleoyl-2-acetylglycerol, which can induce melanin synthesis in amelanotic human melanoma cells in a dose-dependent manner. Flow cytometry analyses showed a small-sized population of vesicles distinguishable from large cells by their fluorescence properties upon neoglycoprotein binding. Sorting indicated that the small-sized subpopulation is composed of vesicles produced by melanocytic cells. Upon vesicle formation, a selective concentration of sugar receptors specific for 6-phospho-beta-D-galactosides appears in the resulting melanocytic vesicles. Vesicles are recognized and taken up by cultured keratinocytes and a partial inhibitory effect was obtained upon cell incubation in the presence of neoglycoproteins, indicating a possible participation of sugar receptors in this recognition. The validity for such a model to help in understanding the natural melanin transfer by melanosomes is confirmed by electron microscopy, which demonstrates the presence of melanin inside keratinocytic cells upon incubation with melanocytic vesicles.     

Extracellular matrix modulates the function of human melanocytes but not melanoma cells

Normal human epidermal melanocytes are attached to a basement membrane, a specialized form of extracellular matrix (ECM), located between the epithelium and underlying dermal tissues. To determine whether ECM influences pigmented cell behavior in vitro, human epidermal melanocytes and melanoma cells were cultured on uncoated or ECM-coated plastic culture surfaces, and a comparison was made between growth and function in the presence or absence of ECM. Melanocytes cultured on ECM-coated surfaces developed flatter and larger cell bodies and produced more melanin than melanocytes cultured on uncoated surfaces. In the presence of phorbol-myristate-acetate and cholera toxin, the rate of melanocyte replication was increased by ECM. In the absence of these mitogens, ECM significantly enhanced the adhesiveness of nonproliferating melanocytes. ECM had little or no effect on these parameters (morphology, tyrosinase activity, replication) in a pigmented human malignant melanoma cell line. These findings indicate that normal human epidermal pigment cells have the ability to recognize and respond to matrix signals, whereas this capacity appears to be absent in melanoma cells.     

DNA microarrays and likelihood ratio bioinformatic methods: discovery of human melanocyte biomarkers

In this article, some of the advantages and limitations of DNA microarray technologies for gene expression profiling are summarized. As a model experiment, DermArray DNA microarrays were utilized to identify potential biomarkers of cultured normal human melanocytes in two different experimental comparisons. In the first case, melanocyte RNA was compared with vastly dissimilar non-melanocytic RNA samples of normal skin keratinocytes and fibroblasts. In the second case, melanocyte RNA was compared with a primary cutaneous melanoma line (MS7) and a metastatic melanoma cell line (SKMel-28). The alternative approaches provide dramatically different lists of 'normal melanocyte' biomarkers. The most robust biomarkers were identified using principal component analysis bioinformatic methods related to likelihood ratios. Only three of 25 robust biomarkers in the melanocyte-proximal study (i.e. melanocytes vs. melanoma cells) were coincidentally identified in the melanocyte-distal study (i.e. melanocytes vs. non-melanocytic cells). Selected up-regulated biomarkers of melanocytes (i.e. TRP-1, melan-A/MART-1, silver/Pmel17, and nidogen-2) were validated by qRT-PCR. Some of the melanocytic biomarkers identified here may be useful in molecular diagnostics, as potential molecular targets for drug discovery, and for understanding the biochemistry of melanocytic cells.     

Protein Kinase C and Melanocyte Transformation

Melanoma cells which have been isolated from metastatic melanoma tissue are able to survive and proliferate in serum supplemented media. In contrast, normal human melanocytes require the presence of growth stimulators if they are to survive in culture. A tumor promotor, 12–0-tetradecanoyl-phorbol-13-acetate (TPA) and substances that increase intracellular levels of cyclic-adenosine-monophosphate (cAMP), such as cholera toxin or isobutylmethyl xanthine, have been widely used for this purpose. The phorbol diester receptor was shown in 1982 to be the phospholipid- and calcium-dependent enzyme protein kinase C (PKC). We therefore directed our studies to the role of PKC regulation in the growth of normal human melanocytes and their transformation. Our studies show that while melanoma cells are inhibited by TPA, the growth of normal melanocytes is stimulated in a dose-dependent manner. The inhibitor, 1-(5-isoquinolinesulfonyl)-2-methyl-piperizine dihydrochloride (H7), which has been found to be the most specific for PKC, had no effect on the growth of normal melanocytes, but inhibited the growth of melanoma cells in a dose-dependent manner. PKC was isolated from the membrane and cytosol of normal melanocytes and melanoma cells. The basal (resting) levels of PKC activity in normal melanocytes was low compared to that measured in melanoma cells, and after short-term (1 hour) treatment with TPA the PKC activity was greatest at the membrane, with the activity decreasing the cytosol. Upon prolonged (48 hours) treatment with TPA, this redistribution of activity continued in normal melanocytes and the total activity increased. In melanoma cells, however, the total PKC activity decreased, particularly in the membrane fraction. A difference in activity and distribution of the enzyme was also seen after short-term (1 hour) treatment with H7. There was very little effect seen on PKC in normal melanocytes; however, the effect on melanoma cells was similar to that seen after 48 hours of exposure to TPA with a decrease in total activity, particularly in the membrane fraction. These results indicate that the regulation of PKC, in particular its activation by TPA, is altered during the transformation of normal human melanocytes     

Comparison of High Performance Liquid Chromatography and Stereological Image Analysis for the Quantitation of Eumelanins and Pheomelanins in Melanoma Cells*

The aim of the study was to compare two methods quantifying eumelanins and pheomelanins, pigments synthesized by melanocytes. One is based on the high performance liquid chromatography (HPLC) quantitation of specific degradation products of each melanin type. The other requires image analysis, transmission election microscopy (TEM), and stereology. This study was carried out in cultured human melanoma cells and for each line, melanins were measured by HPLC and cells were fixed and embedded as pellets for TEM. Ultrathin sections were treated or not by the alkali elution method allowing the elimination of pheomelanins. The obtained micrographs were analyzed with our image analysis program permitting the estimation of several primary parameters. Stereology was used for estimating melanosomal maturation, intracellular melanins content, and number of melanized melanosomes per cell, for total melanin, eumelanins, or pheomelanins. Our results show a good correlation between both methods for total melanin, particularly when using the cytoplasmic volume density of melanin (r=0.93). Moreover, we report that the number of melanized melanosomes per cell and not the melanosomal maturation is responsible for the differences in total melanin content observed between the different cell lines. However, none of the stereological melanization parameters was correlated in the case of eumelanins or pheomelanins. In order to demonstrate the utter relevancy of this stereological approach, utilization of more pigmented melanoma cells, comparative study of HPLC and stereology, in normal epidermal melanocytes and a new evaluation of the alkali elution method in appropriate animal models would help us to explain the present results.     

A method for quantifying melanosome transfer efficacy from melanocytes to keratinocytes in vitro

Several different in vivo and in vitro bioassays are used to evaluate melanosome transfer efficacy from melanocytes to keratinocytes. However, these methods are complicated and time consuming. Here, we report on a simple, rapid, direct, and reliable in vitro method for observing the process of melanosome transfer from melanocytes to keratinocytes. First, we selected and tested a melanoma cell line RPMI-7951 that can normally synthesize melanin and transfer from mature melanosomes to keratinocytes in vitro. We cocultured these cells with a human ovarian teratoma transformed epidermal carcinoma cell line, which is also capable of accepting melanosomes transferred from melanocytes, as in normal keratinocytes. The cells were cocultured for 24-72 h and double labeled with FITC-conjugated antibody against the melanosome-associated protein TRP-1, and with Cy5-conjugated antibody against the keratinocyte-specific marker keratin 14. The cells were examined by fluorescence microscope and flow cytometry. Melanosome transfer from melanocytes to keratinocytes increased in a time-dependent manner. To verify the accessibility of this method, the melanosome transfer inhibitor, a serine protease inhibitor, 4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride, and a melanosome transfer stimulator, alpha-melanocyte-stimulating hormone, were added. The serine protease inhibitor decreased melanosome transfer, and alpha-melanocyte-stimulating hormone increased melanosome transfer, in a dose-dependent manner. In conclusion, this is a simple, rapid, and effective model system to quantify the melanosome transfer efficacy from melanocytes to keratinocytes in vitro.     

Co-localization of beta1,6-branched oligosaccharides and coarse melanin in macrophage-melanoma fusion hybrids and human melanoma cells in vitro

Fusion hybrids between normal macrophages and Cloudman S91 melanoma cells were shown earlier to have increased metastatic potential, along with high expression of beta1,6-N-acetylglucosaminyltransferase V and beta1,6-branched oligosaccharides. Curiously, hybrids, but not parental melanoma cells, also produced 'coarse melanin'- autophagic vesicles with multiple melanosomes. As beta1,6-branched oligosaccharides were known to be associated with metastasis, and coarse melanin had been described in invasive human melanomas, we looked for potential relationships between the two. Using lectin- and immunohistochemistry, we analyzed cell lines producing coarse melanin for beta1,6-branched oligosaccharides: gp100/pmel-17 (a melanosomal structural component) and CD63 (a late endosome/lysosome component associated with melanoma and certain other human cancers). Cell lines used in this study were (i) hybrid 94-H48, a highly metastatic, macrophage-melanoma experimental fusion hybrid; (ii) 6(neo) mouse melanoma cells, the weakly metastatic, parental fusion partner; and (iii) SKmel-23, a human melanoma cell line derived from a metastasis. Coarse melanin granules were prominent both in hybrids and in SKmel-23 cells, and co-localized with stains for beta1,6-branched oligosaccharides, gp100/pmel 17, and CD63. This is the first report of this phenotype being expressed in vitro, although co-expression of beta1,6-branched oligosaccharides and coarse melanin was recently shown to be a common and pervasive characteristic in archival specimens of human melanomas, and was most prominent in metastases. The results suggest that pathways of melanogenesis in melanoma may differ significantly from those in normal melanocytes. In vitro expression of this phenotype provides new biological systems for more detailed analyses of its genesis and regulation at the molecular genetic level.     

Monoclonal antibody against a melanosomal protein in melanotic and amelanotic human melanoma cells

BALB/c mice were immunized with tyrosinase, partially purified in two stages from a human melanoma cell line. A hybridoma was obtained which produced monoclonal antibody (MoAb 1C11) reactive with 8/10 melanoma cell lines and 10/10 primary cultures of human melanocytes, neval cells, and melanomas. Immunoreactivity correlated to a certain extent with tyrosinase activity but not with melanin content. No crossreactivity was obtained with neuroblastoma, medulloblastoma, fibroblasts, keratinocytes, lymphoid cells, or murine melanomas. Purification of the antigen directly from cell lysates with a MoAb 1C11 CNBr-Sepharose affinity column gave a green-brown protein of 56 kDa with no detectable tyrosinase activity. This protein was therefore different from 60 kDa active tyrosinase, identified by enzyme activity and Western blotting with a MoAb derived previously (MoAb 5C12). Unlike 5C12, 1C11 reactivity was not destroyed by pretreatment of the antigen with periodate. Immunogold labelling showed that the 1C11-reactive antigen was associated with melanosomes, and there was close correlation between 5C12 and 1C11 reactivity in resistance to trypsin and in staining various melanocytic cell populations. MoAb 1C11 may therefore recognise a polypeptide epitope in a molecule closely linked to melanin biosynthesis.     

Gangliosides of normal and neoplastic human melanocytes

The major ganglioside component isolated from diploid human melanocytes is sialosyllactosylceramide (GM3 86-91% of total sialic acid). The corresponding disialo derivative (GD3) is found as a minor component (2-6% of total sialic acid) in the membranes of these cells. In human melanoma cells, grown in tissue culture, GD3 is the predominant ganglioside component (48-63% of total sialic acid). Withdrawal of TPA from the culture medium of normal melanocytes or addition of TPA to the medium of melanoma cells had no significant effect on GM3/GD3 ratios. We conclude that the difference between the composition of gangliosides is related to the normal vs transformed phenotypes of melanocytes.     

PIG3V, an immortalized human vitiligo melanocyte cell line, expresses dilated endoplasmic reticulum

Summary Vitiligo is an enigmatic pigmentary disorder of the skin. Factors potentially involved in the progressive loss of melanocytes from the basal layer of the epidermis include genetically determined aberrancies of the vitiligo melanocyte. It follows that analysis of melanocytes cultured from vitiligo donors can contribute to a further understanding of the etiopathomechanism. A setback for vitiligo research has been the limited availability of vitiligo-derived melanocytes. To overcome this limitation, we have generated a vitiligo melanocyte cell line according to a protocol established previously for the immortalization of normal human melanocytes. Vitiligo melanocytes Ma9308P4 were transfected with HPV16 E6 and E7 genes using the retroviral construct LXSN16E6E7. Successful transformants were selected using geneticin and subsequently cloned to ensure genetic homogeneity. The resulting cell line PIG3V has undergone more than 100 cell population doublings ince its establishment as a confluent primary culture, whereas untransfected melanocytes derived from adult skin senesce after a maximum of 50 population doublings. Cells immortalized by this transfection procedure retain lineage-specific characteristics and proliferate significantly faster than parental cells. In this study, the phenotype of PIG3V resembled melanocytes rather than melanoma cells in culture. Tyrosinase was processed properly and melanosomes remained pigmented. Importantly, ultrastructural characterization of PIG3V cells revealed dilated endoplasmic reticulum profiles characteristic of vitiligo melanocytes. An explanation for this dilation may be found in the retention of proteins with molecular weight of 37.5, 47.5, and 56.5 kDa, as determined by gel electrophoresis of microsomal proteins isolated from radiolabeled cells. Presented in part at the Annual Meeting of the Panamerican Society for Pigment Cell Research, Aspen, Colorado, 1998.     

Identification of a cDNA Coding for a Ca-Binding Phosphoprotein (p90), Calnexin, on Melanosomes in Normal and Malignant Human Melanocytes

In order to have a proper biosynthesis and secretion of the melanin-pigment granules (melanosomes) the melanocyte may require a melanosome-associated molecule that provides a signal for assembly and organization of melanogenic enzymes and proteins within the compartment of melanosomes. This study reports the presence of a Ca²⁺-binding phosphoprotein, p90, which can be engaged in such melanogenic function, located on the melanosomal membrane of human melanocytes. A human melanoma cDNA expression library in λ Zap II was screened with a rabbit polyclonal antibody raised against human melanosomes isolated from cultured human melanoma cells, SK MEL 23. A cDNA encoding a melanosomal protein, M_r 90 kDa, was identified through this immunoscreening. A partial sequencing of nucleotides (822 bp from the N-terminal domain) of this clone (3.8 kb) and predicted amino acids showed more than 90% homology with dog calnexin, a previously reported endoplasmic reticulum (ER) transmembrane protein. A fusion protein of this p90 with β-galactosidase expressed in Escherichia coli revealed both the immuno-cross-reactivity with anti-dog calnexin and anti-human melanosome antibodies and the Ca²⁺-binding property. Upon immunohistochemistry, the anti-dog calnexin antibody revealed the positive immunoreactivities with both normal and malignant human melanocytes, showing a much higher expression of antigenic epitope than nonmelanocytic human cells. The laser scanning confocal immunofluorescence, using an anti-body against a human melanosome-specific antigen (HMSA-5), and immunoelectron microscopy, using immunogold, confirmed the major localization of anti-dog calnexin antibody epitope on the melanosomes and ER.     

Roscovitine inhibits differentiation and invasion in a three-dimensional skin reconstruction model of metastatic melanoma

The aim of this study was to investigate the therapeutic potential of a cyclin-dependent kinase inhibitor, roscovitine, in cultured melanoma cells and a three-dimensional skin reconstruction model of metastatic melanoma. The modulatory effects of roscovitine on the growth and survival of normal melanocytes and cultured melanoma cell lines were tested. Additionally, we investigated the potential of roscovitine to regulate the growth and differentiation of a metastatic melanoma cell line (A375) in a three-dimensional skin reconstruction culture consisting of A375 cells admixed with normal human keratinocytes embedded within a collagen-constricted fibroblast matrix. We show that roscovitine is able to induce apoptosis in the melanoma cell lines A375, 888, and 624 but not in normal human cultured epithelial melanocytes. The degree of apoptosis within these cell lines correlated with the accumulation of p53 protein and concomitant reduction of X-linked inhibitor of apoptosis protein, with no change in the proteins Bcl-2 and survivin. We also found that roscovitine inhibited the growth and differentiation of A375 melanoma cells within the dermal layer of the skin. The results of this study show that roscovitine has the potential to inhibit the differentiation and invasion of metastatic melanoma and may be useful as a therapy for the treatment of patients with metastatic melanoma.     

Characterization of human melanoma cell lines and melanocytes by proteome analysis

We have analyzed the proteomes of two human melanoma cell lines (A375 and 526), and of the human melanocytes, (FOM 78), by two-dimensional electrophoresis (2D-PAGE) and liquid chromatography – tandem mass spectrometry (LC-MS/MS). Our comparative proteomic analysis revealed that six proteins were over-expressed in both melanoma cell lines as compared to melanocytes: galectin-1, inosine-5'-monophosphate dehydrogenase 2, serine/threonine-protein phosphatase 2A 65 kDa regulatory subunit A alpha isoform, protein DJ-1, cyclophilin A and cofilin-1. We show, for the first time, that only specific isoforms of these molecules are over-expressed in melanoma. Different protein profiles were also found between each individual melanoma cell line and the melanocytes. s-Methyl-5-thioadenosine phosphorylase, ubiquitin and ribosomal protein S27 a precursor, the basic form of protein DJ-1, annexin a1, proliferation associated protein 2g4, isoform alfa-enolase of alfa-enolase, protein disulfide-isomerase precursor, and elongation factor 2 were more strongly expressed in A375 cells compared to melanocytes. In 526 cells, 60s acidic ribosomal protein p1 and calreticulin precursor were more highly expressed than in melanocytes. These molecular differences may help in better understanding melanoma development and its different responsiveness to therapies. The identified proteins could be exploited as biomarkers or therapeutic targets for melanoma.     

Crosstalk in skin: melanocytes,keratinocytes, stem cells,and melanoma

In the vertebrate embryo, melanocytes arise from the neural crest, migrate to and colonize the basal layer within the skin and skin appendages. Post-migratory melanocytes are securely attached to the basement membrane, and their morphology, growth, adhesion, and migration are under control of neighboring keratinocytes. Melanoma is a malignant tumor originated from melanocytes or their progenitor cells. During melanocyte transformation and melanoma progression, melanocytes lose their interactions with keratinocytes, resulting in uncontrolled proliferation and invasion of the malignant cells. Melanoma cells at the advanced stages often lack melanocytic features and resemble multipotent progenitors, which are a potential melanocyte reservoir in human skin. In this mini-review, we will summarize findings on cell-cell interactions that are responsible for normal melanocyte homeostasis, stem cell self-renewal, and differentiation. Our ultimate goal is to define molecules and pathways, which are essential for normal cell-cell interactions but deregulated in melanoma formation and progression.     

Expression of Ca(2+)-independent and Ca(2+)-dependent phospholipases A(2) and cyclooxygenases in human melanocytes and malignant melanoma cell lines

We provide novel evidence that human melanoma cell lines (M10, M14, SK-MEL28, SK-MEL93, 243MEL, 1074MEL, OCM-1, and COLO38) expressed, at mRNA and protein levels, either Ca(2+)-independent phospholipase A(2) (iPLA(2)) or cytosolic phospholipase A(2) (cPLA(2)) and its phosphorylated form. Normal human melanocytes contained the lowest levels of both PLA(2)s. Cyclooxygenase-1 and -2 (COX-1 and COX-2) were also expressed in cultured tumor cells as measured by Western blots. The most pronounced overexpression of iPLA(2) and COX-1 was found in two melanoma-derived cells, M14 and COLO38. Normal human melanocytes and the M10 melanoma cell line displayed no COX-2 expression. Using subcellular fractionation, Western blot and confocal microcopy analyses, in paradigmatic SK-MEL28 and SK-MEL93 cells we showed that iPLA(2), COX-1 and even cPLA(2) were equally located in the cytosol, membrane structures and perinuclear region while COX-2 was preferentially associated with the cytosol. Specific inhibitors of these three enzymes significantly reduced the basal proliferation rate either in melanocytes or in melanoma cell lines. These results, coupled with the inhibition of the cell proliferation by electroporation of melanoma cells with cPLA(2) or COX-2 antibodies, demonstrate that a possible correlation between PLA(2)-COX expression and tumor cell proliferation in the melanocytic system does exist. In addition, the high expression level of both PLA(2)s and COXs suggests that eicosanoids modulate cell proliferation and tumor invasiveness.     

Melanin has a role in Ca2+ homeostasis in human melanocytes

We have examined whether melanin affects Ca2+ homeostasis in cultured normal human melanocytes. Intracellular Ca2+ concentrations ([Ca2+]i), were measured in four Caucasian and in three Negroid melanocyte cultures. Under resting conditions [Ca2+]i was around 100 nM in all cultures, but differences between cells within cultures were observed. All cultures responded to endothelin-1 (ET-1) with increases in [Ca2+]i and there were no differences between Caucasian and Negroid cultures. However, large differences in responses between cells within cultures were observed, indicating that melanocyte cultures are very heterogeneous. The addition of 2.5 mM CaCl2 to melanocytes kept in Ca2+-free medium resulted in rapid and transient increases in [Ca2+]i of up to 1500 nM. These increases were on average more than two times smaller in melanocyte cultures established from Negroid donors compared with Caucasian cultures. In addition, well melanized Caucasian melanocytes, cultured in the presence of 400 microM tyrosine and 10 mM NH4Cl, showed a reduced increase in cytoplasmic Ca2+ concentration upon the addition of extracellular Ca2+. The difference in maintaining Ca2+ homeostasis between poorly and well melanized melanocytes may be the result of the clearance of cytoplasmic Ca2+ into melanosomes and the greater capacity for this in the more pigmented melanocytes.     

Expression of α2-macroglobulin receptor-associated protein in normal human epidermal melanocytes and human melanoma cell lines

α₂-Macroglobulin receptor/low-density lipoprotein receptor-related protein is a multifunctional cell surface receptor known to bind and internalize a large number of ligands. α₂-Macroglobulin receptor-associated protein acts as an intracellular “chaperone” for this receptor, and it has been shown to inhibit binding of all its known ligands. In this paper, we characterize the expression of the receptor-associated protein in both normal human epidermal melanocytes and in six different human melanoma cell lines, by the use of flow cytometry and Western blotting analysis. We show that all the melanoma cell lines and the normal melanocytes express the receptor-associated protein at similar levels, with most located intracellularly. No receptor-associated protein was detected at the cell surface in the melanocytes or in three of the cell lines. However, in two of the melanoma cell lines, large amounts of receptor-associated protein were found on the cell surface, these having the largest amounts of it reported to date; in a further melanoma cell line, there was a small amount at the cell surface. We have also shown that the melanocytes and all the melanoma cell lines express the receptor itself at a wide range of levels, the highest levels of both the cell surface receptor and the cell surface receptor-associated protein being found in one particular melanoma cell line. By growing the cell lines under controlled conditions, we have demonstrated that, although the total cellular content of the receptor is markedly increased at high cell culture density, this treatment has no effect on the level of expression of the receptor-associated protein. J. Cell. Biochem. 71:149–157, 1998. © 1998 Wiley-Liss, Inc.     

Convenient methodology for extraction and subsequent selective propagation of mouse melanocytes in culture from adult mouse skin tissue

Mouse melanoma B16-BL6 cells are useful cells for cancer metastatic studies. To understand the metastatic principle at molecular levels, it is necessary to carry out experiments in which cancer cells and their normal counterparts are compared. However, unlike normal human melanocytes, preparation of normal mouse melanocytes is quite difficult due to the lack of marketing and insufficient information on an established protocol for primary culture of mouse melanocytes. In this study, we aimed to establish a convenient method for primary culture of mouse melanocytes on the basis of the protocol for human melanocytes. The main obstacles to preparing pure mouse melanocytes are how to digest mouse skin tissue and how to reduce the contamination of keratinocytes and fibroblasts. The obstacles were overcome by collagenase digestion for skin specimens, short time trypsinization for separating melanocytes and keratinocytes, and use of 12-O-Tetradecanoylphorbol 13-acetate (TPA) and cholera toxin in the culture medium. These supplements act to prevent the proliferation of keratinocytes and fibroblasts, respectively. The convenient procedure enabled us to prepare a pure culture of normal mouse melanocytes. Using enriched normal mouse melanocytes and cancerous B16-BL6 cells, we compared the expression levels of melanoma cell adhesion molecule (MCAM), an important membrane protein for melanoma metastasis, in the cells. The results showed markedly higher expression of MCAM in B16-BL6 cells than in normal mouse melanocytes.