INTRODUCTION

In the epidermis, melanocytes synthesize melanin, which is responsible for skin pigmentation. Melanin synthesis is carried out by a cell-specific enzymatic pathway controlled by tyrosinase (EC 1.14.18.1), the enzyme that catalyzes the initial two rate-limiting reactions of this process, the hydroxylation of tyrosine to dopa and its subsequent oxidation to dopaquinone (1, 2, 3, 4). In vivo, melanogenesis is induced mainly by ultraviolet A and B radiation of sunlight and -melanocyte-stimulating hormone (-MSH)¹ (5) which binds to a specific G protein-coupled receptor. In cultured melanocytes or in melanoma cells, melanogenesis can be induced by ultraviolet A and B radiation and by a large array of effectors including -MSH (4) and pharmacological agents such as forskolin, cholera toxin, or isobutylmethylxanthine (6, 7, 8, 9). These agents increase the intracellular cAMP content, thereby indicating the importance of the cAMP pathway in melanogenesis. The stimulation of melanogenesis by cAMP-elevating agents seems to occur through the induction of tyrosinase expression and stimulation of its intrinsic enzymatic activity ensuing post-translational modifications (10). However, few data are available concerning molecular mechanisms that connect the cAMP signaling pathway and tyrosinase regulation. Recently, we have shown in B16 melanoma cells that cAMP-elevating agents stimulate ERK1 activity and induce its translocation to the nucleus (9), whereas in the majority of cell systems, cAMP has been described to inhibit this kinase (11). Furthermore, concomitantly to the stimulation of ERK1 and melanogenesis, cAMP induces a morphological differentiation characterized by dendrite outgrowth (9) and an inhibition of B16 melanoma cell proliferation. Similar effects including ERK1 activation, neurite outgrowth, and cell growth inhibition have been observed during cAMP-induced differentiation of rat pheochromocytoma PC12 cells (12, 13), which, like melanocytes, are derived from the neural crest. The mechanisms of differentiation in PC12 cells have been thoroughly investigated. Recently, two reports have shown that wortmannin (14), a potent phosphatidylinositol-3-kinase (PI3-K) inhibitor, as well as dominant negative mutants of PI3-K, clearly inhibit nerve growth factor-induced neurite outgrowth in PC12 cells, thus demonstrating a positive involvement of PI3-K in PC12 cell differentiation (15, 16).

PI3-K belongs to a family of signal transducer heterodimeric enzymes composed of a 85-kDa regulatory subunit (p85  or ) containing SH2 and SH3 domains and a 110-kDa catalytic subunit (p110  or ) (17, 18) that phosphorylates the D3 hydroxyl in the inositol ring of phosphatidylinositol. PI3-K is activated after association of the p85 regulatory subunit with tyrosine-phosphorylated proteins, including activated tyrosine kinase receptors, non-receptor tyrosine kinases (19, 20, 21, 22, 23), and docking proteins such IRS-1 (24). The role of PI3-K in transducing mitogenic signals is currently clearly confirmed, and recently the kinase has been found to be implicated in differentiation (15, 16). Other studies have involved PI3-K in membrane transport and intracellular traffic through the regulation of membrane and cytoskeleton rearrangements occurring in response to growth factor stimulation (14, 25, 26, 27).

Recent studies have demonstrated that the serine/threonine kinase, p70^S6-kinase (p70^S6K) acts downstream of PI3-K (28, 29). P70^S6K phosphorylates its main target, the 40 S ribosomal protein S6, which is involved in translational up-regulation of an essential family of mRNAs, including transcripts for ribosomal proteins and elongation factors. Inhibition of the p70^S6K activity by the immunosupressant rapamycin or microinjection of neutralizing antibodies severely block cell cycle progression at the middle G₁ phase, indicating that p70^S6K is necessary for cells to enter the S phase (28).

Taking into account the involvement of the PI3-K pathway in PC12 cell differentiation and in an attempt to search for intracellular signaling pathways involved in melanogenesis, we focused our interest on the implication of PI3-K in the B16 melanoma cell differentiation process. In this report, using the specific PI3-K inhibitor LY294002 (30) and the p70^S6K-specific inhibitor rapamycin (31), we show that the inhibition of the PI3-K/p70^S6K pathway mimics the effect of cAMP-elevating agents and leads to a strong stimulation of melanogenesis in B16 melanoma cells. Furthermore, cAMP triggers a significant inhibition of PI3-K activity and a strong blockage of p70^S6K activity, thus revealing the implication of the PI3-K/p70^S6K pathway in the transmission of the melanogenic effect of cAMP.

EXPERIMENTAL PROCEDURES

Dulbecco's modified Eagle's medium, antibiotics (penicillin and streptomycin), isobutylmethylxanthine, 12-O-tetradecanoylphorbol-13-acetate, forskolin, bovine serum albumin, protein A-Sepharose, 4-(2-aminoethyl)benzenesulfonyl fluoride (AEBSF), aprotinin, leupeptin, L-dopa, benzamidine, PNPP, rapamycin, and phosphatidylinositol (PI) were purchased from Sigma. Fetal calf serum was from Life Technologies, Inc. LY294002 was from Biomol Research Laboratories (Plymouth, UK). The culture cell plates and 96-well plates were from Falcon. Polyclonal rabbit antisera to human tyrosinase (PEP-7) was provided by Dr. V. Hearing (Bethesda, MD), and the secondary fluorescein isothiocyanate-conjugated and peroxidase-conjugated anti-rabbit antibodies were from Dakopatts (Glostrup, Denmark). Antisera anti-PI3-K was provided by Dr. J. Schlessinger (Rochester, NY). Antibodies to p70^S6K and ribosomal extracts were kindly given by Dr. G. Thomas (Basel, Switzerland), and antibodies against ERK1 were from Santa Cruz Biotechnology (Santa Cruz, CA). -[³²P]ATP (3000 Ci/mmol) was from Amersham Corp. (Buckinghamshire, UK). The thin layer chromatography plates, used for the PI3-K assays, were from Whatman.

B-16/F10 murine melanoma cells (from Dr. V. Hearing) were cultured in Dulbecco's modified Eagle's medium with 10% fetal calf serum and penicillin/streptomycin (100 IU/50 µg/ml) in a humidified atmosphere containing 5% CO₂ in air at 37 °C. For melanogenesis experiments (immunofluorescence assays, melanin quantification, enzyme activity assays, RNA extractions, and Western blot experiments) cells were cultured in 6-well plates, and for kinase assays cells were scraped from 10-cm diameter plates.

To increase the cAMP content, we performed 48-h treatments with forskolin (20 µM); LY294002 was used at 15 µM, and rapamycin was used at a 1 nM concentration, always added in serum-containing medium. All agents were diluted in dimethyl sulfoxide. In all experiments the different effects were compared to nonstimulated cells treated only with dimethyl sulfoxide diluted accordingly in all cases.

For immunofluorescence labeling, B16 melanoma cells grown on glass coverslips were rinsed briefly in 1 × PBS, fixed with methanol (20 °C) for 2 min, washed twice in 1 × PBS, and processed. For tyrosinase detection we used the rabbit antibody directed against human tyrosinase PEP7 (32, 33) at a 1/500 dilution. The primary antibody was applied for 45 min at 37 °C, followed by a 10-min wash in PBS and then a 45 min incubation with the secondary antibody (anti-rabbit fluorescein isothiocyanate-conjugated at a 1/50 dilution), followed by a final wash of 10 min in 1 × PBS. Finally, coverslips were mounted using a homemade immunofluorescence mounting medium (glycerol/p-phenylendiamide in PBS) and viewed with the 40× or 63× objectives using a Zeiss-Axiophot microscope equipped with epifluorescence illumination. Photographs were taken using the Kodak T-Max 400 Iso film.

Tyrosinase activity was estimated by measuring the rate of oxidation of L-dopa (34). Cells from a subconfluent monolayer in a 6-well plate well were suspended in 100 µl of phosphate buffer, pH 6.8, containing 1% (w/v) Triton X-100. After vortexing to lyse the cells, the extracts were clarified by centrifugation at 10,000 rpm for 5 min in an Eppendorf Biofuge. The tyrosinase substrate L-dopa (2 mg/ml) was prepared in the same lysis phosphate buffer (without Triton). 40 µl of each extract were put in a 96-well plate, and the enzymatic assay was started by adding 100 µl of L-dopa solution at 37 °C. Control wells contained 40 µl of lysis buffer. Absorbance at 570 nm was read every 10 min for at least 1 h at 37 °C using a microplate reader (Dynatech Laboratories). The blank was removed from each absorbance value, and a plot of absorbance against time was represented for each condition. The final activity was corrected by the total amount of protein of the dish.

For melanin determination, after 48-h treatment with forskolin or the kinase inhibitors, cells from a confluent 3.5-cm diameter plate were solubilized in 100 µl of 1 M NaOH and diluted with 400 µl of distilled water. The samples were incubated at 60 °C for 1 h and vortexed to solubilize the melanin. Absorbance at 405 nm was compared with a standard curve of known concentrations of fungal melanin prepared in a final NaOH concentration of 0.2 M.

The quantification of tyrosinase mRNA was carried out by quantitative reverse transcription PCR. Total cellular RNA was prepared from control and treated mouse melanoma B16 cells according to the modified method of Chomczynski and Sacchi (35) (RNable procedure from Eurobio). In each case, 5 µg of RNA were reverse transcribed using the reverse transcription system from Promega. The cDNA obtained was subjected to 25 cycles of PCR (94 °C, 30 s; 55 °C, 45 s; 72 °C, 1 min) using the following specific primers for the mouse tyrosinase gene: 5-CATTTTTGATTTGAGTGTCT-3 and 5-TGTGGTAGTCGTCTTTGTCC-3, and a 1191-base pair PCR product was amplified. Specific primers for the glyceraldehyde-3-phosphate dehydrogenase (Clontech) were added as a control for the same reverse transcriptase product and gave rise to an amplified PCR product of 983 bp.

Preliminary trials showed that, after 25 cycles of PCR, the reaction remained exponential. The PCR products were electrophoresed on 1% agarose gel and stained with ethidium bromide before visualization using ultraviolet light.

For the tyrosinase immunoblot detection, cells were lysed in phosphate buffer, pH 6.8, containing 1% (w/v) Triton X-100 and 100 IU/ml aprotinin, and 1 mM AEBSF. After vortexing, the extracts were centrifuged at 4 °C at 13,000 rpm in an Eppendorf Biofuge for 5 min, and 20 µl of the solubilized proteins (supernatants) were loaded onto 10% SDS-polyacrylamide gels (30:0.8, acrylamide:bisacrylamide). Gels were blotted into nitrocellulose (Amersham Corp.). The nitrocellulose membranes were saturated with 5% powdered milk in saline buffer, and tyrosinase was detected with the PEP7 polyclonal antibody at a 1/3000 dilution in the saturation buffer and with a secondary peroxidase-conjugated anti-rabbit antibody at a 1/4000 dilution. Three 10-min washes after the primary and secondary antibodies were performed using a washing buffer containing 0.05% Triton X-100, 0.5% powdered milk in a saline buffer. The blot was developed with the ECL system from Amersham Corp.

For p70^S6K detection, we used special 10% polyacrylamide gels (30:0.1, acrylamide:bisacrylamide) to increase resolution. The rest of the process was performed as described above except that the a primary antibody was the M6 polyclonal antibody directed against p70^S6K (36, 37).

For PI3-K activity assays, cells were extracted in the lysis buffer containing 50 mM Hepes, 150 mM NaCl, 10 mM EDTA, 10 mM Na₄P₂O₇, 100 mM NaF, 2 mM vanadate, 1 mM AEBSF, 100 IU/ml Trazylol, 1% w/v Triton X-100, pH 7.4, for 15 min at 4 °C. The extracts were centrifuged at 13,000 rpm for 10 min at 4 °C and were immunoprecipitated with the anti-PI3-K antibody directed against the C-terminal domain of p85 (38) preadsorbed on protein A-Sepharose for 90 min at 4 °C under agitation. The immunoprecipitates were washed twice with each of the following buffers: (i) phosphate-buffered saline (pH 7.4) containing 1% Nonidet-P40; (ii) 100 mM Tris, 0.5 M LiCl, pH 7.4; and (iii) 10 mM Tris, 100 mM NaCl, 1 mM EDTA, pH 7.4 (39). The pellets were resuspended in 30 µl of 20 mM Hepes, 0.4 mM EGTA, 0.4 mM Na2HPO₄, and the kinase reaction was started by addition of phosphatidylinositol (0.2 mg/ml), 10 mM MgCl₂, and 50 µM -[³²P]ATP (10 Ci/mmol). After 15 min under slight agitation at room temperature, the reaction was stopped by addition of 15 µl of 4 M HCl, and the phosphoinositides were extracted with 130 µl of chloroform:methanol (1:1). The phospholipids were analyzed by thin layer chromatography and autoradiography (39).

For p70^S6K assays, cells were scraped in a buffer containing 50 mM Tris, pH 8.0, 120 mM NaCl, 20 mM NaF, 1 mM EDTA, 6 mM EGTA, 15 mM Na₄P₂O, 1% Nonidet-P40, 30 mM PNPP, 0.1 mM AEBSF, 1 mM benzamidine, and 0.5 mM dithiothreitol. Next the extracts were sonicated in an ice-water bath for 5 min at maximum power and then centrifuged at 13,000 rpm for 5 min at 4 °C. The lysates were incubated with 1 µl of the anti-P70^S6K antibody M6 for 1 h. The same amount of protein A-Sepharose (CL4B) was added in each tube followed by another incubation of 20 min at 4 °C under agitation. The immunoprecipitates were washed three times in the lysis buffer and one final time in a buffer containing 50 mM MOPS, pH 7.2, 10 mM PNPP, 0.1 mM AEBSF, 0.1% Triton X-100. The kinase reaction was performed by adding a reaction mix (10 µl per point) containing 1 µl of ribosomal extract prepared as described in Lane et al. (40), cold ATP 30 µM final, and 3 µCi of -[³²P]ATP diluted in the kinase buffer (50 mM MOPS, pH 7.2, 10 mM PNPP, 0.1 mM AEBSF, 1% Triton X-100). The reaction was stopped by adding 10 µl of 2 × Laemmli solution, then the tubes were boiled and centrifuged, and the samples were loaded in a 12.5% polyacrylamide gel and exposed for autoradiography.

RESULTS

Since the inhibition of PI3-K activity prevents neurite outgrowth in PC12 cells (15, 16), we first investigated the possible involvement of the PI3-K pathway in the cAMP-induced dendrite outgrowth in B16 melanoma cells. Control and forskolin-treated cells were exposed to PI3-K and p70^S6K inhibitors for 48 h, and cell morphology was observed in immunofluorescence experiments using the anti-tyrosinase PEP7 as a primary antibody (Fig. 1). In control conditions, B16 cells displayed a fibroblastic appearance (Fig. 1a) while forskolin treatment promoted the emergence of small and numerous dendrites from the plasma membrane (Fig. 1b). Considering the data reported in PC12 cells, we next wanted to inhibit the cAMP induced dendrite outgrowth by specifically inhibiting the PI3-K activity with LY294002. The PI3-K inhibitor LY294002 behaves as a competitive inhibitor of the ATP binding site specific for PI3-K and abolishes the activity of this enzyme in vitro and in vivo at low micromolar concentrations but has no effect against PI4-kinase nor a number of intracellular serine/threonine kinases (30, 41). It is worth remarking that LY294002 was chosen because it appears to be a much more stable agent than wortmannin in culture medium (15, 16) (data not shown). Therefore, B16 cells were treated with 15 µM LY294002 alone or simultaneously with 20 µM forskolin and LY294002 for 48 h. Unexpectedly, LY294002 did not inhibit the forskolin-induced dendricity but it rather strengthened the effect of the single forskolin exposure (Fig. 1d). Furthermore, when cells were treated only with LY294002, dendricity was greatly induced, similarly to what is observed upon forskolin stimulation (Fig. 1c). To investigate the implication of the PI3-K signaling pathway in dendrite outgrowth processes, we next looked at the role of p70^S6K, which has been reported to constitute an indirect PI3-K target in the same signaling cascade. As shown in Fig. 1f, rapamycin, the specific p70^S6K inhibitor, was not able to inhibit the cAMP-induced dendricity, but the kinase inhibitor alone did not induce any apparent dendritogenesis (Fig. 1e). In addition to the information concerning cell dendricity, these immunofluorescence experiments showed that in control conditions (Fig. 1a) or in rapamycin-treated cells (Fig. 1e), tyrosinase was located mainly in the cytoplasm surrounding the nucleus. After forskolin (Fig. 1d) or LY294002 treatment (Fig. 1c), the tyrosinase labeling was found in condensed vesicles which appeared to be the melanosomes spreading throughout the dendritic expansions of the membrane. Interestingly, in forskolin-, LY294002-, or rapamycin-treated cells, the immunofluorescence labeling appeared more intense, suggesting that these specific agents increased tyrosinase expression. Previous immunofluorescence experiments using other antibodies directed against tyrosinase confirmed the specificity of the PEP7 antibody for tyrosinase (data not shown).

Therefore, these results show that, neither LY294002 nor rapamycin prevented forskolin-induced dendricity. However, PI3-K inhibition, but not p70^S6K inhibition, was sufficient to mimic the stimulatory effect of forskolin in promoting dendrite outgrowth.

To further investigate the effects of PI3-K and p70^S6K inhibition on melanogenesis, we measured melanin synthesis and tyrosinase activity in B16 cells after the treatment with LY294002 or rapamycin. The color of cell pellets and melanin content quantification using a colorimetric assay are shown in Fig. 2. Forskolin, LY294002, and rapamycin clearly promoted cell darkening. Forskolin and LY294002 induced approximately a 6-fold increase in pigment content, while with rapamycin a 3-fold stimulation was detected. Next we quantified the DOPA oxidase activity of tyrosinase, which constitutes the second reaction catalyzed by the enzyme in the melanin synthesis cascade (Fig. 3). The DOPA oxidase activity was increased approximately 5-6-fold by forskolin treatment. When cells were exposed to the LY294002 PI3-K inhibitor or to the rapamycin p70^S6K inhibitor, a similar increase in tyrosinase activity was observed.

To determine whether these melanogenic agents might affect directly the tyrosinase expression, tyrosinase mRNA and protein were quantified in control and treated cells. Reverse transcription PCR assays on RNA extracted from control B16 cells using tyrosinase-specific primers produced a 1191-bp fragment corresponding to the tyrosinase mRNA. An increased amount of this PCR fragment was observed with RNA from forskolin, LY294002, and rapamycin treated cells (Fig. 4B), indicating that these agents increased the levels of tyrosinase gene expression compared to nonstimulated cells. A control of PCR amplification, using specific primers for the glyceraldehyde-3-phosphate dehydrogenase transcript, gave similar amounts of a 983-bp PCR product in each condition (Fig. 4A).

Finally, a Western blot detection of tyrosinase was carried out in order to measure the amount of tyrosinase protein in cells exposed to the different agents. Forskolin, LY294002, and rapamycin markedly increased the amount of tyrosinase protein (detected as a band of 70 kDa) compared to the nearly undetectable level of expression in the nonstimulated cells (Fig. 5). This result indicated that the inhibition of PI3-K or p70^S6K led to a stimulation of tyrosinase expression. The detection of the 44-kDa ERK1 protein ensured that each lane was loaded with the same amount of protein. Taken together these data clearly demonstrate that PI3-K or p70^S6K inhibition induce an increase of tyrosinase gene expression and amount of protein, thus leading to a stimulation of melanin synthesis.

Since PI3-K inhibition by LY294002 and p70^S6K inhibition by rapamycin appear to trigger the same melanogenic effects as forskolin, we next evaluated the action of this cAMP-elevating agent on PI3-K and P70^S6K activities. PI3-K was immunoprecipitated from B16 cells and an in vitro phosphorylation assay with a mix of PIs as substrate was performed. Then the phosphatidylinositol phosphates were separated by thin layer chromatography and visualized by autoradiography. After treatment of B16 cells with serum for 30 min, we observed a moderate but significant increase in PI phosphorylation (PI-3-P) (Fig. 6). This phosphorylation was completely inhibited in the presence of LY294002 in the kinase reaction. When PI3-K was measured in cells exposed concomitantly to serum and forskolin for 15, 20, or 30 min, we detected a diminished PIs phosphorylation compared to serum-stimulated cells (Fig. 6). Quantification of two independent experiments showed a 40% decrease of PIs phosphorylation in extracts from forskolin-treated cells. We therefore conclude that cAMP induced a weak but significant inhibition of the PI3-K activity in B16 cells.

The cAMP effect on p70^S6K activity was also examined. Initially the p70^S6K activity toward its substrate, the ribosomal protein S6, was evaluated after immunoprecipitation of the enzyme from control serum-starved and serum-treated B16 cells (Fig. 7A). A 30-min serum stimulation induced a strong increase in the P70^S6K activity as shown by the high S6 phosphorylation levels. In the presence of LY294002 or rapamycin, the stimulatory effect of serum was completely inhibited. Interestingly, forskolin also inhibited the serum-induced P70^S6K activation since a dramatic diminution of S6 phosphorylation was already observed after 15 min of forskolin treatment. After 30 min, this inhibition was almost complete. It is now well established that the activity of p70^S6K is regulated by multiple phosphorylation of the enzyme (42). Thus we studied the effect of forskolin on the phosphorylation status of p70^S6K. When p70^S6K is activated, it is highly phosphorylated and presents a slower electrophoretic mobility. In contrast, the inactive kinase is less phosphorylated and has a faster electrophoretic migration. We examined the p70^S6K electrophoretic mobility in Western blot experiments using the antibody M6 (36, 37). In starved cells, p70^S6K appeared as a two principal bands with high electrophoretic mobility representing the unphosphorylated forms of the enzyme. In serum-treated cells we observed two additional bands with higher apparent molecular weights corresponding to the activated p70^S6K. This effect of serum was reversed by LY294002 or rapamycin which increased the mobility of the protein. When serum-treated cells were exposed for 15, 20, or 30 min to forskolin, the low electrophoretic mobility forms gradually disappeared, indicating diminished phosphorylation (Fig. 7B). Hence, cAMP inhibited the p70^S6K activity as a result of an inhibition of its phosphorylation.

It is worth mentioning that while forskolin induced an increase in the MAP kinase activity in B16 cells as reported (9), neither LY294002 nor rapamycin affected the activity of this kinase measured by in vitro phosphorylation assays (not shown).

DISCUSSION

In the present work we provide evidence for the role of the PI3-K/p70^S6K pathway in the control of B16 melanoma cell differentiation which is characterized by a stimulation of melanin synthesis and dendrite outgrowth. In PC12 cells, the inhibition of PI3-K activity has been shown to prevent nerve growth factor-induced neurite outgrowth (15, 16). Since PC12 cells share numerous features with B16 cells, we therefore expected that the inhibition of PI3-K would block the cAMP-induced dendrite outgrowth in this melanoma cell line. Strikingly, our studies revealed that, in B16 cells, the inhibition of PI3-K by LY294002 did not block forskolin effects but was sufficient by itself to induce a strong cell dendricity. However, the inhibition of p70^S6K that functions downstream PI3-K (29) did not lead to any morpholological changes. This observation suggests the existence of PI3-K targets acting upstream or independently of p70^S6K that might play a regulatory role in the induction of these morphological modifications. The protein kinase B/Akt which is a serine/threonine protein kinase encoded by the proto-oncogene akt (43) has been found to function downstream PI3-K and to participate to the p70^S6K activation (44); it is tempting to propose that other unknown targets of protein kinase B/AKT could be involved in the induction of the dendritic phenotype. Similarly, some members of the protein kinase C family such as protein kinase C , , , and , which are directly activated by phosphatidylinositol phosphate (45, 46, 47) are also putative candidates that could participate in this phenomenon. Moreover, it has recently been reported that proteins Rho and Rac, which belong to the superfamily of small G proteins, are indirectly regulated by PI3-K and mediate cytoskeleton rearrangements triggering events such as membrane ruffling and vesicle reorganization (48, 49, 50, 51). This points to the putative involvement of these molecules in eliciting dendrite outgrowth processes after cAMP or LY294002 stimulation. Future research is certainly required to identify the pathway leading to this morphological modification (dendricity) in B16 melanoma cells.

On the other hand, our work demonstrates that the inhibition of PI3-K led to a stimulation of melanin synthesis that appeared to result from an increased tyrosinase activity and expression. In addition, similar effects were observed with rapamycin, indicating that the inhibition of p70^S6K is sufficient to induce melanogenesis. However, rapamycin, which is as potent as forskolin or LY294002 in stimulating tyrosinase activity, is markedly less efficient in inducing cell tanning. A possible explanation of this finding might reside in the fact that the synthesis of black melanin (eumelanin) involves other specific regulated enzymes such as tyrosinase-related proteins 1 and 2 (52, 53, 54). Thus, we can hypothesize that rapamycin may somehow inhibit the expression or the activity of these tyrosinase-related enzymes or down-regulate other steps downstream of tyrosinase in the melanin production pathway.

It is worth emphasizing that forskolin, LY294002, and rapamycin produced similar effects on melanogenesis. This supports the hypothesis that the induction of melanogenesis by cAMP could be due to a down-regulation of the PI3-K/p70^S6K pathway. Indeed an inhibition of PI3-K activity by cAMP has been already described in T lymphocytes (55) and in neutrophils (56). Furthermore, cAMP-elevating agents have been shown to inhibit p70^S6K activity in Swiss 3T3 cells (57) and in T lymphocytes (55). In B16 melanoma cells, forskolin partially inhibits PI3-K activity and completely blocks p70^S6K activity. In addition, this cAMP-elevating agent as well as -MSH (not shown), a physiological melanogenic agent that increases the intracellular cAMP levels, induced a dramatic inhibition of p70^S6K phosphorylation which constitutes an essential event for the activation of the kinase. Since the inhibition of PI3-K activity by cAMP appears rather weak, this could not entirely explain the strong inhibition of p70^S6K activity found after the treatment with cAMP elevating agents. Therefore it can be proposed that cAMP could act through the regulation of the activity of protein kinase B or other kinases involved in phosphorylation and activation of p70^S6K.

Interestingly, we observed a significant diminution in the cell number and in DNA synthesis, without cytotoxic effects, after forskolin, LY294002, or rapamycin treatments (not shown), revealing an arrest of cell growth. This could be due to the inhibitory effects of forskolin, LY294002, or rapamycin on p70^S6K activity, which has been found essential for progression through the G₁ phase of the cell cycle (40). In contrast, serum or growth factors, such as 12-O-tetradecanoylphorbol-13-acetate or epidermal growth factor that activate p70^S6K, display an inhibitory effect on melanogenesis in B16 cells (not shown). These findings suggest that the induction of melanogenesis and morphological differentiation in B16 cells might require the arrest of cell growth at the G₁ phase. A similar conclusion has been reached concerning the induction of neurite outgrowth in the neuroblastoma cell line NE1-115 (58).

The precise molecular mechanisms that connect the inhibition of the PI3-K/p70^S6K pathway to the induction of melanogenesis and dendrite outgrowth remain to be elucidated. Nevertheless, the existence of a nuclear isoform of the S6 kinase (p85^S6K) (59) is consistent with the idea that this kinase could modulate the activity of transcription factors. Indeed, cAMP-responsive element modulator  has been reported to be phosphorylated and activated by p70^S6K (60). On the other hand, microphthalmia, a tissue-specific transcription factor that has been recently shown to play a key role in the stimulation of tyrosinase gene expression by cAMP (61), could also be a target of p70^S6K. Thus the inhibition of this kinase by cAMP would lead to a decreased activation of microphthalmia, or other transcription factors involved in B16 cell differentiation, and thereby regulate their transcriptional activities.

Taken together our results demonstrate the involvement of the PI3-K/p70^S6K pathway in B16 melanoma cells differentiation and indicate for the first time that the inhibition of this signaling cascade by cAMP is likely to be a key event for the cAMP-induced melanogenesis and dendrite outgrowth. Furthermore, the novel finding that the inhibition of the PI3-K/p70^S6K pathway displays a positive effect on B16 melanoma cell differentiation whereas the PI3-K inhibition blocks the differentiation process in PC12 cells, reveals that the study of several cell systems is required to fully understand the mechanisms of cell differentiation.