Mitogen-activated protein kinase pathway and AP-1 are activated during cAMP-induced melanogenesis in B-16 melanoma cells.

In mammalian melanocytes, melanin synthesis is controlled by tyrosinase, the critical enzyme in the melanogenic pathway. We and others showed that the stimulation of melanogenesis by cAMP is due to an increased tyrosinase expression at protein and mRNA levels. However, the molecular events connecting the rise of intracellular cAMP and the increase in tyrosinase activity remain to be elucidated. In this study, using B16 melanoma cells, we showed that cAMP-elevating agents stimulated mitogen-activated protein (MAP) kinase, p44. This effect was mediated by the activation of MAP kinase kinase. cAMP-elevating agents induced a translocation of p44 to the nucleus and an activation of the transcription factor AP-1. cAMP-induced AP-1 contained FOS-related antigen-2 in association with JunD, while after phorbol ester stimulation AP-1 complexes consist mainly of JunD/c-Fos heterodimers. In an attempt to connect these molecular events to the control of tyrosinase expression that appears to be the pivotal point of melanogenesis regulation, we hypothesized that following its activation by cAMP, p44 activates AP-1. Then AP-1 could stimulate tyrosinase expression through the interaction with specific DNA sequences present in the mouse tyrosinase promoter.

In mammalian melanocytes, melanin synthesis is controlled by tyrosinase, the critical enzyme in the melanogenic pathway. We and others showed that the stimulation of melanogenesis by cAMP is due to an increased tyrosinase expression at protein and mRNA levels. However, the molecular events connecting the rise of intracellular cAMP and the increase in tyrosinase activity remain to be elucidated. In this study, using B16 melanoma cells, we showed that cAMPelevating agents stimulated mitogen-activated protein (MAP) kinase, p44 mapk . This effect was mediated by the activation of MAP kinase kinase. cAMP-elevating agents induced a translocation of p44 mapk to the nucleus and an activation of the transcription factor AP-1. cAMP-induced AP-1 contained FOS-related antigen-2 in association with JunD, while after phorbol ester stimulation AP-1 complexes consist mainly of JunD/c-Fos heterodimers. In an attempt to connect these molecular events to the control of tyrosinase expression that appears to be the pivotal point of melanogenesis regulation, we hypothesized that following its activation by cAMP, p44 mapk activates AP-1. Then AP-1 could stimulate tyrosinase expression through the interaction with specific DNA sequences present in the mouse tyrosinase promoter.
In melanocytes and melanoma cells, melanin synthesis is controlled by a cascade of enzymatic reactions regulated at the level of tyrosinase. This enzyme synthesizes dopaquinone from tyrosine and appears to control the rate-limiting step of melanogenesis. Melanin synthesis is stimulated by a large array of effectors including 1-oleyl-2-acetyl-glycerol (1), ultraviolet B radiations (2), and cAMP-elevating agents (forskolin, IBMX, 1 ␣-MSH) (3)(4)(5). Few data are available concerning the molecular mechanisms triggered by these melanogenic agents. Protein kinase C was thought to be involved in the induction of mela-nogenesis by 1-oleyl-2-acetyl-glycerol and ultraviolet B radiations (6,7). However, a recent report of Carsberg et al. (8) has shown that the stimulation of melanogenesis by these agents was not affected by RO485, a potent inhibitor of protein kinase C. While the role of protein kinase C in the induction of melanogenesis remains controversial, compelling data have shown that cAMP-elevating agents stimulate melanogenesis in both melanocytes and melanoma cells, indicating that the cAMP pathway plays a key role in the regulation of melanogenesis (3)(4)(5). The effect of cAMP on melanogenesis is due to a stimulation of tyrosinase activity. This appears to be the consequence of an augmentation of enzymatic activity of preexisting tyrosinase (4,9) following post-translational modifications such as (i) phosphorylation or glycosylation (10), (ii) association with an activator (11,12), and (iii) dissociation from an inhibitor (13). Alternatively, cAMP was shown to increase tyrosinase mRNA (14,15), resulting in an augmentation of tyrosinase amount, suggesting that cAMP stimulates tyrosinase transcription (16). However, the molecular events connecting the stimulation of tyrosinase activity or the activation of tyrosinase gene expression to the rise of cellular cAMP remain to be identified.
The proline-directed serine/threonine kinases of the MAP kinase family (p44 mapk and p42 mapk ) are activated upon phosphorylation on both threonine 192 and tyrosine 194 by the dual specificity MAP kinase kinase (MEK) (20). MEK is itself phosphorylated and activated by Raf-1 kinase (20) or by a recently identified MAP kinase kinase kinase (MEK kinase) (21). MAP kinases were shown to be involved in the control of cell growth (17), in the regulation of some metabolic processes such as glycogen synthesis (18,19), and more recently in the regulation of pheochromocytoma and adipocytes differentiation (22,23). In melanocytes and melanoma cells the induction of melanogenesis is associated with cell differentiation. Thus, we hypothesized that MAP kinases could be activated during cAMPinduced melanogenesis. Using the well characterized mouse melanoma cells B-16, we demonstrated that cAMP-elevating agents such as forskolin, IBMX, and 4-norleucine 7-D-phenylalanine-␣-melanocyte stimulating hormone ([Nle 4 ,D-Phe 7 ]␣-MSH), a potent analog of ␣-MSH, stimulated p44 mapk through the activation of the MEK. Further investigations demonstrated a translocation of p44 mapk to the nucleus and an activation of the transcription factor AP-1 by cAMP-elevating agents. In this condition the AP-1 complex contained predominantly JunD and Fra-2. Our results provide meaningful clues concerning the molecular mechanisms triggered by cAMP in B-16 melanoma cells and suggest that the MAP kinase pathway and AP-1 could play a role in melanogenesis regulation by cAMP.
Determination of Tyrosinase Activity and Melanin Synthesis-Tyrosinase activity in living cells was estimated by the amount of 3 H 2 O released into the culture medium during the hydroxylation of [3, H]tyrosine to hydroxyphenylalanine, according to Oikawa et al. (24). The melanin formation assay uses L-3,4-dihydroxyphenyl 3-[ 14 C]alanine (7-12 mCi/mmol) as precursor (25).
Western Blot Analysis-B-16 cell lysates were separated by SDS-PAGE (10% acrylamide gels) and transferred to Hybond-C extra membranes. The blots were probed with PEP-7 antibody directed to the C terminus part of tyrosinase (26), and then proteins were visualized by the Amersham ECL system and quantified by image analysis.
Northern Blot Analysis-Total cellular RNAs were purified using the method described by Chomczynski and Sacchi (27). For Northern blot analysis, total cellular RNA (25 g/lane) were fractionated on 1% agarose, 0.66 M formaldehyde gels and transferred onto nylon membranes. Mouse tyrosinase cDNA probe (generously provided by Dr. B. Bouchard) and human glyceraldehyde-3-phosphate deshydrogenase cDNA probe were labeled with [␣-32 P]dCTP using a random priming kit (Stratagene). Membranes were autoradiographed and quantified by image analysis.
Image Analysis-Morphometric measurements were performed using a Biocom 500 (BIOCOM SA, Les Ulis, France) image analysis system coupled to a CCD video camera and a Nikon TMS inverted light microscope. After treatment, cells were viewed using a 20 ϫ phase contrast objective and projected onto the video screen. Outlines of 100 cells of each experimental condition were acquired manually, quantitative measurements of area (A) and perimeter (P) being performed using the Mima software. Then the dendricity factor was evaluated as P 2 (4A) Ϫ1 according to Lacour et al. (28).
p44 mapk Assay-Following a 2-h serum starvation period, B-16 cells were incubated with effectors. Cells were washed with cold phosphatebuffered saline and lysed in a solubilization buffer (1% Triton X-100, 50 mM Hepes, pH 7.4, 150 mM NaCl, 10 mM EDTA, 10 mM Na 4 P 2 O 7 , 2 mM sodium orthovanadate, 100 mM NaF with 100 IU/ml aprotinin, 20 mM leupeptin, 1 mM AEBSF). Solubilized proteins were incubated for 2 h at 4°C with antibody to p44 mapk preadsorbed to protein A-Sepharose. Immune complexes were washed twice with solubilization buffer and twice with HNTG buffer (50 mM Hepes, 150 mM NaCl, 0.1% (v/v) Triton X-100, 10% (v/v) glycerol with 0.2 mM sodium orthovanadate) and resuspended in 50 l of the same buffer supplemented with 30 mM magnesium acetate, 0.2 mg/ml myelin basic protein, and [␥-32 P]ATP (15 M; specific activity 30 Ci/mmol). The phosphorylation reaction was allowed to proceed for 45 min at room temperature and was stopped by spotting the sample onto 3MM Chr Whatman papers, which were then dropped into 10% (v/v) trichloroacetic acid containing 5 mM pyrophosphate. After three washes the radioactivity was determined by Cerenkov counting.
MEK and Raf-1 Assay-MEK activity was measured in a reconstitution assay as the ability of immunopurified MEK to phosphorylate in vitro p44 mapk immunoprecipitated from nonstimulated cells. Pellets containing immunoprecipitated MEK and p44 mapk were washed as described above, mixed, and resuspended in 50 l of HNTG buffer supplemented with 10 mM MnCl 2 . Raf-1 activity was measured with the same procedure using MEK immunoprecipitated from nonstimulated cells as substrate for Raf-1. The phosphorylation reaction was performed in 20 mM Hepes, pH 7.4, 10 mM MgCl 2 , 1 mM MnCl 2 , 1 mM dithiothreitol, 10 mM p-nitrophenyl phosphate. Reactions were initiated by the addition of [␥-32 P]ATP (10 M, 60 Ci/mmol). After incubation at room temperature for 20 min, assays were stopped by the addition of Laemmli sample buffer. The samples were analyzed by SDS-PAGE (10% acrylamide) and autoradiography.
Immunofluorescence Studies-After stimulation, B-16 cells were washed with PBS and fixed at Ϫ20°C for 10 min with methanol/acetone (3:7, v/v). After a 10-min rehydration at 25°C in PBS containing 3% BSA (PBS/BSA), fixed cells were incubated with the primary antibody directed to the C terminus part of p44 mapk (1:500) for 60 min at 25°C. Cells were then washed five times with PBS and incubated in PBS/BSA for 60 min at 25°C with fluorescein isothiocyanate-conjugated secondary antibody (anti-rabbit, 1:100). Finally, cells were washed five times with PBS and examined under confocal laser scanning microscopy.
Nuclear Extracts and Gel Mobility Shift Assay-Serum-starved B-16 cells were stimulated for 2 h with the different effectors, and the nuclear extracts were prepared essentially as described by Dignam (29). In vitro binding reaction of AP-1 in a total volume of 25 l was performed by incubation of 10 g of nuclear extract in a binding buffer containing 10 mM Hepes, pH 7.8, 50 mM KCl, 2 mM dithiothreitol, 1 mM EDTA, 5 mM MgCl 2 , 10% glycerol, 3 mM AEBSF, 2 g of poly(dI-dC), 1 mg/ml BSA. After 10 min of preincubation on ice, 50,000 -100,000 cpm of 32 P end-labeled oligonucleotide probe was added and incubated at 25°C for 20 min. Then DNA-protein complexes were resolved by electrophoresis on 4% polyacrylamide gels in TAE buffer (10 mM Tris, 9 mM sodium acetate/acetic acid, 275 M EDTA) for 8 h at 120 volts, dried, and subjected to autoradiography. When indicated, an excess of cold competitor oligonucleotides was added during preincubation. For antibody supershift assays, 1 l of the corresponding antisera were preincubated with the nuclear extracts for 1 h on ice in the binding reaction buffer before adding the labeled probe. ]␣-MSH plus IBMX (M ϩ I) stimulated tyrosinase activity, melanin synthesis, tyrosinase expression at the protein, and RNA messenger levels (Table I). These results suggest that the stimulation of tyrosinase gene expression plays a key role in the control of melanogenesis by cAMP-elevating agents. Simultaneously, we observed morphological differentiation characterized by numerous and arborescent dendrite outgrowths. These changes were quantified by image analysis as described under "Experimental Procedures." The results presented in Table I show that the dendricity factor is markedly increased after 24 and 48 h of incubation with M ϩ I.  Solubilized proteins were submitted to immunoprecipitation with a specific antibody raised against the C terminus domain of p44 mapk . Kinase activity was then measured using myelin basic protein as substrate and quantified by filter paper assay ( Effect of cAMP-elevating Agent on MEK and on Raf-1 Kinase-In an attempt to understand the mechanism by which cAMP stimulated p44 mapk , we studied the effect of M ϩ I on MAP kinase kinase and Raf-1 kinase activities.

Stimulation of Melanogenesis in B-16 Melanoma Cells-We
The kinase activity of MEK was monitored in a cell-free system, after immunoprecipitation with specific antibody, using as substrate p44 mapk extracted from unstimulated cells (Fig. 2). Lane 1 shows the basal autophosphorylation of unstimulated p44 mapk at 44 kDa. A faint band at 45 kDa in lanes 3 and 4 indicates that MEK autophosphorylation was stimulated by TPA and M ϩ I compared with the basal autophosphorylation (lane 2). When phosphorylation was performed in the presence of both MEK and p44 mapk , we observed a strong phosphorylation of a protein at 44 kDa, indicating that MEK phosphorylated p44 mapk (lane 5). This phosphorylation was increased when MEK was extracted from TPA or M ϩ I-treated cells (lanes 6 and 7), demonstrating that MEK was stimulated by TPA and [Nle 4 , D-Phe 7 ]␣-MSH plus IBMX in B-16 melanoma cells.
Since Raf-1 was described as operating immediately upstream of MEK, similar experiments were performed to examine the effect of M ϩ I on Raf-1 activity. Raf-1 was isolated from B-16 melanoma cells stimulated as described above, and its kinase activity was evaluated using as substrate MEK immunoprecipitated from unstimulated cells (Fig. 3). In the first lane, we observed a band at 45 kDa corresponding to the basal autophosphorylation of MEK. The other bands appeared to be nonspecific, since they were precipitated by preimmune serum (not shown). With Raf-1 incubated alone, no autophosphorylation was observed (lanes 2-4). When MEK was added to Raf-1, no significant increase in the basal phosphorylation of MEK was observed with Raf-1 precipitated from control or M ϩ I-treated cells (lanes 5 and 7). In contrast, MEK phosphorylation was markedly increased in the presence of Raf-1 extracted from TPA-treated cells (lane 6). These results indicate that cAMP-elevating agents did not stimulate Raf-1 activity in B-16 cells. Thus, the effect of [Nle 4 , D-Phe 7 ]␣-MSH plus IBMX on p44 mapk is mediated by MEK that is activated by an unknown mechanism independently on Raf-1 kinase stimulation.
Translocation of p44 mapk to the Nucleus-Stimulation of tyrosinase gene expression plays a key role in the regulation of melanogenesis by cAMP. Thus we hypothesized that the cAMP signal should be transmitted to transcription factors present in the nucleus. To verify this hypothesis, we studied the effect of [Nle 4 , D-Phe 7 ]␣-MSH plus IBMX on p44 mapk localization in B-16 melanoma cells.
The localization of p44 mapk in cells treated or not treated with [Nle 4 , D-Phe 7 ]␣-MSH plus IBMX was studied by immunofluorescence and confocal laser scanning microscopy (Fig. 4). Using an antipeptide to the C terminus part of p44 mapk , we observed, in the absence of M ϩ I, a strong perinuclear and a weak nuclear labeling. After a 60-min exposure to M ϩ I, the cytoplasm and the nucleus appeared equally labeled, indicating that p44 mapk translocated to the nucleus. This phenomenon was transient, since after 150 min in presence of M ϩ I, nucleus labeling decreased, suggesting that p44 mapk returned to the cytoplasm.  4 and 7). Then MEK was phosphorylated alone (lanes 2-4) or in the presence of p44 mapk (lanes 5-7). p44 mapk , isolated from unstimulated cells, was also phosphorylated alone (lane 1). Phosphorylated proteins were analyzed by SDS-PAGE and autoradiography. Molecular masses, indicated on the left, are expressed in kilodaltons.

Activation of Transcription Factors by cAMP-elevating
Agents-Since p44 mapk is activated and translocated to the nucleus by M ϩ I, it is tempting to propose that these phenomena are followed by an activation of transcription factors. Among the transcription factors activated by p44 mapk , AP-1 was reported to stimulate gene expression through the binding to TRE sequences. The presence of TRE-like sequences in the mouse tyrosinase promoter prompted us to study the activation of AP-1 by [Nle 4 , D-Phe 7 ]␣-MSH plus IBMX. B-16 melanoma cells were stimulated either with TPA as positive control or with M ϩ I. Then nuclear extracts were prepared, and the gel retardation assay was performed with a labeled oligonucleotide containing the TRE consensus sequence (TGACTCA) (Fig. 5A). In control, TPA, and M ϩ I conditions, two strong bands with fast electrophoretic mobility were observed. These bands were totally displaced by a 10-fold excess of unlabeled cAMP-responsive element (CRE) (TGACGTCA), indicating that these bands corresponded to a nonspecific binding of CRE binding protein (CREB) to the TRE probe. With TPA or M ϩ I, a third band with a slower electrophoretic mobility appeared. This band was identified as a classical AP-1 complex, since it was displaced more efficiently by unlabeled TRE than by unlabeled CRE. These data demonstrate that [Nle 4 , D-Phe 7 ]␣-MSH plus IBMX activated the transcription factor AP-1 in B-16 melanoma cells. Additionally, an activation of the transcription factor AP-2 by cAMP but not by TPA (Fig. 5B) was observed in B-16 melanoma cells.
Characterization of AP-1 Complexes Induced by cAMP and TPA-Interestingly, the AP-1 complex induced by M ϩ I migrated more slowly than that observed with TPA, suggesting that these AP-1 complexes contain different components. AP-1 consists either in homodimers of Jun family proteins or in heterodimers of Jun/Fos family proteins. Three Jun proteins (c-Jun, JunB, JunD) and at least four Fos proteins (c-Fos, FosB, Fra-1, Fra-2) were found in AP-1 complexes. The composition of AP-1 complexes in both TPA and M ϩ I conditions was investigated by supershift experiments using specific antibodies to c-Jun, JunB, JunD, c-Fos, and Fra-2 (Fig. 6)  types (30). Indeed, neither Raf-1, which is not activated by cAMP, nor MEK kinase, which is not detected in B-16 melanoma cells, is apparently involved in MEK activation by cAMP. The involvement of another member of the Raf kinase family, i.e. A-Raf or B-Raf that is mainly expressed in neuronal cells (31) may be suggested. However, the inhibition of B-Raf kinase activity by cAMP observed in PC12 cells (32), makes this hypothesis unlikely. It remains possible that in B-16 melanoma cells cAMP activates an isoform of MEK kinase, different from that previously described by Lange-Carter (21). Alternatively, inhibition by cAMP of phosphatase 2A activity, which was reported to dephosphorylate and deactivate MEK (33), can be also suggested.
Following its activation by cAMP, we observed a transient translocation of p44 mapk to the nucleus. Similar observations were reported in serum-treated fibroblast (34) or in NGF-stimulated PC12 cells (35). In the nucleus, p44 mapk is thought to phosphorylate and activate numerous transcription factors such as p62 TCF (36), c-Myc (37), and AP-1 (38). In B-16 melanoma cells, we showed that cAMP stimulated AP-1 binding to an oligonucleotide containing a TRE sequence. cAMP-induced AP-1 contained mainly JunD and Fra-2 components, while in TPA-induced AP-1, we found JunB, c-Jun, JunD, c-Fos, and Fra-2, JunD and c-Fos being the major components of these AP-1 complexes. Recently Tamir et al. (39) reported the activation of AP-1 by cAMP in lymphocyte and ascribed the activation of AP-1 by cAMP to the inactivation of the AP-1 inhibitory protein, IP-1, upon phosphorylation by cAMP-dependent kinase (protein kinase A) (40). However, AP-1 can be also activated following the phosphorylation of serines 63 and 73 of the N terminus domain of Jun proteins. These sites are phosphorylated by Jun N-terminal kinases (41,42) and by MAP kinases (38), suggesting that MAP kinases are involved in AP-1 activation. Additionally, a recent report indicates that MAP kinases are involved in the regulation of the expression of Fos family proteins, leading thereby to the stimulation of AP-1 activity (43). Thus, it is tempting to propose that p44 mapk through JunD phosphorylation or Fra-2 up-regulation is accountable for AP-1 activation by cAMP in B-16 melanoma cells.
In this study we showed that melanin synthesis, tyrosinase activity, and amount were simultaneously increased by [Nle 4 , D-Phe 7 ]␣-MSH plus IBMX. These effects appear to be the consequence of the augmentation of tyrosinase mRNA. These observations confirmed previous reports (14,15) suggesting that the control of tyrosinase mRNA expression is a key step in the cAMP-mediated stimulation of melanogenesis in B-16 melanoma cells. Usually, regulation of gene expression by cAMP is mediated by CRE through the binding of CREB family transcription factors that are phosphorylated and activated by protein kinase A (44). However, no canonical CRE was found in the mouse tyrosinase promoter. The presence of two TRE-like sequences (2.1-and 0.18-kilobase upstream transcription start site) in the mouse tyrosinase promoter suggests that the stimulation of AP-1 by cAMP could lead to an increased tyrosinase gene expression. AP-2, another transcription factor, was also shown to mediate the effect of cAMP on gene expression (45). The presence of a putative AP-2 binding site in the mouse tyrosinase promoter and its activation by cAMP suggest that AP-2 could participate, in coordination with AP-1, in the regulation of mouse tyrosinase gene expression. Interestingly, TPA and cAMP display a common set of cellular responses, i.e. activation of p44 mapk and of AP-1, but they promote opposite effects on melanogenesis (14,46). This could be explained by the respective nature of TPA and cAMP-induced AP-1 complexes, suggesting that JunD/Fra-2 would transactivate tyrosinase gene expression while JunD/c-Fos would inhibit, directly or indirectly, tyrosinase gene transcription.
Dendritogenesis, another feature of melanocyte differentiation is stimulated during cAMP-induced melanogenesis in B-16 melanoma cells. Interestingly, cAMP-elevating agents induce in PC12 a differentiated phenotype characterized by neurite outgrowth and an activation of p44 mapk (47)(48)(49). Further, the transfection of these cells with a constitutively active MEK leads to spontaneous neuritogenesis (22), demonstrating that the MAP kinase pathway plays a pivotal role in the regulation of PC12 differentiation. Since dendritogenesis and neuritogenesis are closely related processes, we hypothesize that MAP kinase could play a critical role in the control of differentiation in neural crest-derived cells.
In summary, the data gathered in this study demonstrate that the MAP kinase pathway and AP-1 are activated during cAMP-induced melanogenesis. The role of p44 mapk and that of AP-1 in the regulation of melanogenesis remain to be proved. Nevertheless, we would like to suggest that p44 mapk , possibly through the regulation of AP-1, plays a pivotal role in the control of tyrosinase gene expression and thereby in the regulation of melanogenesis by cAMP in B-16 melanoma cells.