What are melanocytes really doing all day long...?

Everyone knows and seems to agree that melanocytes are there to generate melanin - an intriguing, but underestimated multipurpose molecule that is capable of doing far more than providing pigment and UV protection to skin (1). What about the cell that generates melanin, then? Is this dendritic, neural crest-derived cell still serving useful (or even important) functions when no-one looks at the pigmentation of our skin and its appendages and when there is essentially no UV exposure? In other words, what do epidermal and hair follicle melanocytes do in their spare time - at night, under your bedcover? How much of the full portfolio of physiological melanocyte functions in mammalian skin has really been elucidated already? Does the presence or absence of melanocytes matter for normal epidermal and/or hair follicle functions (beyond pigmentation and UV protection), and for skin immune responses? Do melanocytes even deserve as much credit for UV protection as conventional wisdom attributes to them? In which interactions do these promiscuous cells engage with their immediate epithelial environment and who is controlling whom? What lessons might be distilled from looking at lower vertebrate melanophores and at extracutaneous melanocytes in the endeavour to reveal the 'secret identity' of melanocytes? The current Controversies feature explores these far too infrequently posed, biologically and clinically important questions. Complementing a companion viewpoint essay on malignant melanocytes (2), this critical re-examination of melanocyte biology provides a cornucopia of old, but under-appreciated concepts and novel ideas on the slowly emerging complexity of physiological melanocyte functions, and delineates important, thought-provoking questions that remain to be definitively answered by future research.

Dynamics of pigmentation induction by repeated ultraviolet exposures: dose, dose interval and ultraviolet spectrum dependence.

BACKGROUND: The dynamics of ultraviolet (UV)-induced melanogenesis have been well characterized for single UV exposures. However, our knowledge of the effects of repeated UV exposures on the development of new pigmentation is limited. OBJECTIVES: To characterize the dynamics and dose dependence of pigmentation induction by repeated UV exposures using two different UV sources. METHODS: A total of 40 healthy subjects participated in the study: 21 were exposed to a 5% UVB/95% UVA source and 19 were exposed to a 2% UVB/98% UVA source. Skin phototypes 2-3 were represented. Subjects were exposed one to three times per week. The minimal erythemal dose and minimal melanogenic dose of all subjects were determined, and both visual and instrumental observations of the development of pigmentation and erythema were recorded. RESULTS: Dark-brown pigmentation could be produced by a cumulative UV dose of 4200 J m(-2) given as 10 exposures over 5 weeks. However, comparable pigmentation could also be induced by a cumulative dose of 2900 J m(-2) given as eight exposures over 4 weeks. The lowest cumulative dose of 1900 J m(-2) given over 4 weeks produced moderate pigmentation. The 2% UVB source led to earlier and darker pigmentation than the 5% UVB source did for equally erythemogenic doses. CONCLUSIONS: These observations show that the dynamics of melanogenesis induced by repeated exposures depends on UV dose, dose interval and emission spectrum. They also indicate that increasing the UV dose above a certain level of cumulative exposure does not significantly increase the level of UV-induced pigmentation.

Interferons modulate the expression of hormone receptors on the surface of murine melanoma cells.

The effects of IFN-alpha, IFN-beta, and IFN-gamma on the differentiation of murine melanoma cells has been studied, in the presence and absence of melanocyte-stimulating hormone (MSH); the cells were highly responsive to treatment with MSH, which increased the rate of melanin production 25-fold and tyrosinase activity 6-fold within 4 d. Treatment of melanoma cells with IFN-alpha, IFN-beta, or IFN-gamma alone had no stimulatory effect on melanin production, but when the cells were cultured with IFN in the presence of MSH, pigment production was significantly and synergistically increased relative to cells cultured with MSH only. Flow cytometric analysis revealed that levels of tyrosinase in the cells were not affected by MSH or by IFN, which suggests that stimulation of melanogenic activity occurred by activation of a preexisting cellular enzyme. Scatchard analyses showed that the number of MSH receptors on IFN-treated cells was significantly increased (approximately 2.5-fold) relative to untreated cells (approximately 61,000/cell). These findings demonstrate that IFN stimulate differentiation (that is, pigmentation) of melanocytes by increasing the expression of surface MSH receptors; this in turn suggests that such a mechanism may in part be responsible for postinflammatory skin pigmentation, and provides an additional basis for action in the clinical responses of melanoma to IFN treatment.

Analysis of the transcriptional regulation of melanocytic genes by alphaMSH using the cDNA microarray technique.

Studies of mammalian pigmentation have identified many genes involved in the development, migration, and function of melanocytes. Molecular and biochemical mechanisms that switch melanocytes between the production of eumelanin or pheomelanin involve the opposing action of two signaling molecules, alpha-Melanocyte Stimulating Hormone (alphaMSH) and Agouti Signal Protein (ASP). AlphaMSH affects various aspects of melanocyte behavior, stimulating melanocyte dendricity, attachment to extracellular matrix proteins, but also up-regulating the expression of eumelanogenic genes. In response to ASP, melanocytes switch from producing eumelanin to pheomelanin concomitant with the down-regulation of melanogenic genes transcription. Since activation of signaling pathways leads to mRNA expression, microarray technology can provide a quantitative assessment of the consequences of this activation. Significant up/down regulation of all known melanogenic genes by alphaMSH/ASP in cultured melanocytes has been previously reported. We have now used the cDNA microarray technique to screen alphaMSH-treated melanocytes to identify genes that are transcriptionally enhanced by this factor. We report the melanocytic expression and alphaMSH up-regulation of 11 genes spanning 7 functional categories such as apoptosis, body weight control, intracellular transport, signal transduction, oncogenic transformation, transcription factors and genes coding for keratin associated proteins.

Studies on the relationship of the B700 and B50 murine melanoma antigens.

The B50 and B700 proteins of B16 murine melanoma were studied; they were determined to be distinct, unrelated molecules. This was determined by V8 peptide mapping, N-terminal amino acid sequencing, absence of cross-reactivity with specific polyclonal antibodies, and monoclonal antibodies recognizing different epitopes.

Serum and urine protein differences in patients with malignant melanoma.

We used gel electrophoresis to fractionate serum and urine samples of patients with malignant melanoma and of healthy volunteers (controls) and observed that a) all proteins found in the sera of melanoma patients also occurred in the sera of 2 or more healthy volunteers, b) two minor protein bands were present in the sera of 73% of the melanoma patients but in the sera of only 38% of the controls, and c) one major band was present in the sera of 73% of the melanoma patients but in the sera of only 12% of the controls. We characterized these proteins by molecular size and charge. Two proteins, analogous in electrophoretic behavior to those found more frequently in melanoma patients' sera, were observed exclusively in the urine from melanoma patients. The demonstration of such differences in serum and urine proteins warrants the further study of this system as a serologic and/or immunologic test for diagnosis of malignant melanoma.

A unique tumor rejection antigen from the S91 murine malignant melanoma.

We have identified and described the characteristics of a unique tumor rejection antigen (tumor-specific transplantation antigen) obtained from the murine malignant melanoma S91. This antigen is highly restricted to the autologous melanoma and provides striking inhibition of its growth. Previously, we described common or shared tumor-specific transplantation antigens on the murine malignant melanomas B16 F10, K1735, JB/RH, and JB/MS. No cross-reactivity was obtained in this study between S91 and those four other malignant melanomas. The common tumor-specific transplantation antigen resides on a glycoprotein molecule with a molecular weight of 65,000, termed B700, that shares homology with serum albumin as determined by NH2-terminal amino acid sequencing. B700, however, purified from S91 proved to be ineffective as an immunogen.

Modulation of metastatic potential by cell surface urokinase of murine melanoma cells.

We have carried out enzymatic, immunofluorescence, and surface iodination studies which show that B16 melanoma cells express the single chain form of the urokinase type plasminogen activator (uPA) on their cell surface, and that these cells are capable of plasminogen-dependent fibronectin degradation. The significance of the expression of surface single-chain uPA and uPA activity to the metastatic process was examined by preincubating melanoma cells with uPA modulating agents followed by i.v. injection of the cells into mice and enumeration of pulmonary nodules 17 days later. B16 cells that had been pretreated with anti-uPA immunoglobulins that were inhibitory to uPA activity invariably showed significantly decreased numbers of metastases compared to controls. On the contrary, pretreatment with plasmin, which is not only the product of the uPA catalyzed reaction but is also able to convert single-chain uPA to uPA, significantly increased the numbers of metastases. Control treatments, which included normal rabbit and mouse immunoglobulins, monovalent noninhibitory anti-uPA Fab fragments, and various monoclonal and polyclonal antibodies directed against other B16 cell surface antigens, did not affect the metastatic potential of the cells. Divalent inhibitory anti-uPA F(ab)2 fragments, on the contrary, inhibited metastasis as efficiently as intact IgG. The results support the hypothesis that proteolysis of extracellular matrix components by cell surface-localized uPA may be a critical step during the process of tumor cell invasion and metastasis.

Inhibition of melanoma-associated antigen expression and ecotropic retrovirus production in B16BL6 melanoma cells transfected with major histocompatibility complex class I genes.

We have previously reported that expression of the melanoma-associated antigen (MAA) recognized by MM2-9B6 monoclonal antibody in B16 melanoma was closely associated with C-type ecotropic retroviral particle production. Our present data show that this MAA is encoded by the env gene of an ecotropic retrovirus produced by B16 melanoma cells. Transfection of H-2Kb or H-2Kd genes into two individual clones isolated from B16BL6 melanoma, BL6-8 (H-2Kb-, H-2Db+) and BL6-2 (H-2Kb-, H-2Db-), resulted in a loss of MAA expression. Electron immunohistochemical analysis of melanoma cells and reverse transcriptase assay revealed that the loss of MAA expression in the H-2K gene-transfected cells paralleled the elimination of retroviral particles. In contrast, expression of the endogenous H-2Db gene or transfection with the H-2Dd or H-2Ld gene had no effect on MAA expression or retrovirus production. Northern blot analysis showed equivalent retroviral messages in retrovirus-producing and -nonproducing BL6 melanoma clones. Southern blot analysis revealed that H-2Kb-negative BL6 melanoma cells contain at least four different ecotropic retroviruses with different insertion sites that somatically emerged during malignant transformation or progression. Restriction enzyme analysis showed various changes in proviral DNAs from the H-2Kb- and H2Kd-transfected cells that failed to produce retroviral particles. The observed alterations in the patterns of PstI- and HindIII-digested proviral DNA were found to be due to the appearance of PstI and loss of HindIII restriction sites in the pol region as a result of several nucleotide substitutions. Thus, BL6 melanoma cells produce melanoma-specific ecotropic murine leukemia viruses that encode serologically detectable cell surface MAA. The transfection of BL6 melanoma cells with H-2Kb or H2Kd genes but not H-2Dd or H-2Ld genes resulted in a loss of MAA expression that was attributed to the changes in proviral DNA and loss of melanoma-specific ecotropic retrovirus particle production.

Expression and modulation of a retrovirus-associated antigen by murine melanoma cells.

The antigen recognized by the monoclonal antibody MM2-9B6 is specific for melanomas originating in C57BL/6 mice. It is expressed by three melanomas of independent origin and not by normal or fetal tissues, by a wide variety of other nonmelanoma tumors in C57BL/6 mice, or even by melanomas syngeneic to other strains of mice. We have demonstrated that the expression of the relevant antigen is dependent on the replication and budding of a B-tropic ecotropic murine retrovirus. The relationship between the expression of this virus and carcinogenic progression may yield important insights into the cascade of events leading to neoplastic transformation.

Effects of H-2Kb gene on expression of melanoma-associated antigen and lectin-binding sites on BL6 melanoma cells.

An H-2Kb negative BL6 melanoma clone (BL6-8) was transfected with plasmids containing either the class I H-2Kb or class II H-2IAk gene in combination with the neor gene. The effects of the transfected genes on the expression of the melanoma-associated antigen (MAA) recognized by the monoclonal antibodies MM2-9B6 and MM2-3C6 and the cell surface carbohydrates recognized by 15 different lectins were studied. The original H-2Kb- clone or clones transfected with neor or class II H-2IAk genes expressed high levels of MAA and very low levels of soybean agglutinin (SBA), Griffonia simplicifolia I-B4 (GSIB4), and peanut agglutinin (PNA) lectin-binding sites. In contrast, clones that expressed high levels of the transfected H-2Kb gene completely lost the expression of MAA. In addition, these clones were characterized by the appearance of high levels of expression of the sugars specifically reacting with SBA, GSIB4, and PNA lectins. When the original BL6-8 clone was transfected with the H-2Kd gene, 25 clones subsequently isolated had relatively low expression of the transfected H-2Kd gene but high expression of the endogenous H-2Kb gene accompanied by an alteration in expression of the MAA and lectin binding identical with patterns common for H-2Kb+ melanoma cells. These changes were not due to the transfection, plasmid construction, or place of insertion, since similar phenotypic characteristics were found in H-2Kb+ but not H-2Kb- clones isolated from the N-methyl-N'-nitro-N-nitrosoguanidine-treated BL6T2 or parental BL6 melanoma lines. In total, 73 BL6 melanoma clones were investigated and all of the 41 H-2Kb+ clones displayed loss of MAA and appearance of SBA, GSIB4, and PNA-binding sugars. None of the 32 H-2Kb- clones showed these changes. This study indicates that the class I H-2Kb gene product might alter several phenotypic properties of BL6 melanoma cells. The mechanisms of these changes remain unknown. We consider that these effects of the class I H-2Kb gene are indirect, involving interactions with the B-tropic ecotropic retrovirus specific for melanomas of C57BL/6 mice origin.

Intracellular localization of tumor-associated antigens in murine and human malignant melanoma.

Cross-reacting tumor-associated antigens have been demonstrated to be present in the cytoplasm of melanocytes in malignant melanoma. We have utilized both the leukocyte migration inhibition and the lymphocyte stimulation assays to determine the intracellular location(s) of these antigens in both human malignant melanoma and B-16 murine melanoma. The results of both assays indicate that this antigen(s) is associated with the membranous organelles, particularly the melanin granules, and is not found in the soluble components of the cytoplasm, as suggested by other studies.

Role of estradiol and 2-hydroxyestradiol in melanin formation in vitro.

Melanin formation from 3,4-dihydroxyphenylalanine (dopa) was studied in the presence of estradiol and 2-hydroxyestradiol by use of a tyrosinase isolated from B16-F10 melanoma cells grown in C57 black female mice. Both steroids were found incorporated into melanin, but the 2-hydroxy compound was incorporated to a higher extent. The melanin was also able to bind substantial amounts of the two steroids, and the more highly oxidized compound showed higher binding. Melanin isolated from incubates of dopa with mushroom tyrosinase has the ability to bind the steroids and to incorporate small amounts into its structure. It is suggested that melanin in mammalian tissues may function as a depository for estrogens, particularly for those which are more highly oxidized.

Specific incorporation of 2-thiouracil into biological melanins.

2-Thiouracil (TU), an antithyroid drug, is generally recognized as a highly specific melanoma seeker owing to its capability of being selectively accumulated into active melanin-producing tissues. We recently reported evidence that in vitro TU is capable of reacting with dopaquinone (DQ), an early intermediate in melanin biosynthesis, to give an addition product characterized as 6-S-thiouracildopa (TD). However, several aspects of the mechanism of the uptake of TU into melanin in vivo still need to be clarified. We report here the extremely rapid incorporation of [2-14C]thiouracil into melanoma tumors growing subcutaneously in mice and show its selective accumulation into melanin by isolation and purification of the pigment fraction. Formation of the TD adduct in the tumor was examined by HPLC analysis of the soluble fractions of the tissue homogenates: however, no trace of TD could be detected on account of its rapid oxidation by the melanogenic enzyme tyrosinase, as evidenced by in vitro kinetic measurements. Monitoring the course of the tyrosinase-catalyzed oxidation of 5,6-dihydroxyindole (DHI) and 5,6-dihydroxyindole-2-carboxylic acid (DHICA) in the presence of TU, at various molar ratios, provided evidence for the ability of the drug to affect melanogenesis by interaction with biosynthetic intermediates beyond the DQ stage, suggesting other possible modes for its chemical binding to the growing pigment.

Homology of the B50 murine melanoma antigen to the Ro/SS-A antigen of human systemic lupus erythematosus and to calcium-binding proteins.

B50 is a murine melanoma-associated antigen found in tight association with B700, a melanoma-specific antigen. B700-like molecules are produced by all melanomas tested to date, including those of murine, human, swine and hamster origin. We have used rabbit antibodies to B50 to determine whether B50 expression is also restricted to melanomas. The results demonstrate that B50 is a commonly occurring protein, or is immunologically cross-reactive to a commonly occurring protein; 29 of 29 cell lines tested bound anti-B50 antibodies. N-terminal amino acid sequence analysis indicates that B50 has significant homology to the Ro/SS-A antigen of human systemic lupus erythematosus and to calcium binding proteins; hence B50 is likely to be an RNA and/or calcium-binding protein.

Further studies of the therapeutic effects of murine melanoma-specific monoclonal antibodies.

The results presented here further characterize four murine monoclonal antibodies (mAb) that recognize melanoma-specific antigens (9B6, T97, 2-3-1 and 2-3-3). These melanoma-specific mAbs are of the IgG2b isotype and are significantly therapeutic when administered systemically against established pulmonary melanoma metastases. Here we show a consistent and significant inhibition of the growth of melanoma lung metastases by all four mAbs and the existence of a time 'window' at days 5-8 after tumor inoculation for optimal therapy. Since these mAbs were found not to be cytotoxic or cytolytic in vitro, we looked for host immune response regulation as being responsible for the therapeutic effects. Natural killer (NK) cells were implicated as one arm of the host immune system involved in this response since depletion of NK cells in vivo by alpha asialoGM1 or alpha NK1.1 antibodies partially abrogated the inhibitory effect of the mAbs. The observed antimetastatic effects could also be partially abrogated using antibodies directed against the T-cell subset surface markers, CD4+ and CD8+. Intramuscular melanoma tumor growth was also found to be suppressed by mAb 2-3-1, but only if administered in the area of tumor growth and only if multiple inoculations are administered over a 13-day period. The beneficial effect of mAb antimetastatic therapy was found to be useful against several syngeneic melanomas, including JB/MS, B16 and several sublines of the B16 F10 melanoma.

Mammalin tyrosinase. Stoichiometry and measurement of reaction products.

The substrates and intermediates involved in the conversion of tyrosine or 3,4-dihydroxyphenylalanine into melanin by autooxidation, or tyrosinases (monophenol, dihydroxyphenylalanine:oxygen oxidoreductases, EC 1.14.18.1) of mushroom or mammalian melanocyte origin, was studied by a variety of enzymic assays, and by amino acid analysis. It was found that the classic pathway of melanin formation was followed, and that the proposed alternate pathway involving formation of the intermediate 3,4,6-trihydroxyphenylalanine was not a functional route, since nascent trihydroxyphenylalanine was not detectable. The ability of isolated mammalian tyrosinases to convert tyrosine into dihydroxyphenylalanine was unequivocably demonstrated. The polymerization of monomers into melanin was followed by the use of specifically labelled precursors, and the data indicate that uncyclized and carboxylated derivatives are not incorporated into the polymer in vitro. It was found that although in most respects the melanin produced from tyrosine by mushroom and mammalian tyrosinses are similar, the control mechanisms involved in the expression of melanin formation in these organisms must differ greatly.

Production of monoclonal antibodies against the B700 murine melanoma antigen and their antimetastatic properties.

Two unique murine melanoma antigens, termed B700 and B50, have been identified and isolated from several different murine melanoma cell lines. Both antigens can be detected on the cell surface, are actively shed in culture, and are often found in close association intracellularly. In previous studies, the antigen B700, which is related to serum albumin by biochemical and immunological criteria, was shown to function as a melanoma-specific tumor rejection antigen. We have also shown that animals sensitized to irradiated JB/RH melanoma cells produce antibodies which recognize B700 and/or B50, with B700 evoking the stronger humoral response. Animals testing positive by ELISA for antibody production to B700 or B50 were used for preparation of hybridomas and four different murine monoclonal antibodies have been produced whose specificities should facilitate epitope mapping. Clones have been used to generate ascites fluid in nude mice; the antibodies specifically recognize B700 and intact murine melanoma cells, but not B50. Two of these monoclonal antibodies have been administered systemically to C57Bl/6 mice bearing 5 day pulmonary metastases of the JB/MS melanoma, and significant inhibition of metastatic growth was observed for both antibodies.

Cross-reactivity between murine melanoma antigen B700 and a human melanoma-associated antigen (M-66) recognized by autologous antibody: evidence suggesting shared epitopes.

We have previously reported the purification and partial characterization of a human melanoma-associated antigen (M-66) recognized by autologous antibody. This antigen was found to be an unusually acidic 66 kDa glycoprotein. In studies of murine melanoma, a 67-kDa albumin-like melanoma-associated antigen (MAA) isolated from B16 melanoma cells has also been reported by our laboratories. Because the murine MAA, B700, has a molecular weight that is nearly the same as M-66, we sought to determine what similarities and differences existed between these two antigens. Human sera S150, which is known to recognize M-66, was found to bind to murine melanoma cell line B16. The addition of purified M-66 inhibited binding of S150 to B16 cells. Binding by S150 was not noted against murine melanoma cell line S91, which is known not to express cell surface B700. Conversely, reactivity of S150 against Y-Mel 84:420, known to express M-66, could be inhibited by preincubation with B16 cells. Four monoclonal antibodies known to recognize B700 were evaluated for-binding against murine B16 and human melanoma cell line Y-Mel 84:420. Binding was noted against both B16 and Y-Mel 84:420 which could be inhibited by the addition of M-66. Binding of S150 was also noted against purified B700 as tested by ELISA. While a comparison of the amino acid composition of the two antigens revealed similarities, M-66 contained 2.8 times as much serine and 0.4 times as much proline as B700. B700 has been reported to be related to serum albumin, which is not the case for M-66.(ABSTRACT TRUNCATED AT 250 WORDS)

Mammalian tyrosinase. Structural and functional interraltionship of isozymes.

The isozymes of tyrosinase from normal and malignant melanocytes were studied; the data indicates that each consists of a basic tyrosinase polypeptide, and differs by post-translational modifications. T3 represents the de novo form of the enzyme; it is converted to T1 in vivo by the addition of sialic acids and neutral sugars, and in turn, to T4 by complexing with mealanosomal membrane constituents. The T2 isomer is suggested to be an artefact of the electrophoretic procedure, and due to deamidation of T3. It is shown that the apparent kinetics of enzyme activity are unafffected by any of these modifications.