Repeated UVB irradiations do not have the same potential to promote stimulation of melanogenesis in cultured normal human melanocytes

The main physiological stimulus for human melanogenesis is ultraviolet radiation. Darkening of human skin due to increased melanin pigmentation after exposure to sunlight or to UV from artificial sources is commonly known as tan-ning. This reaction results from a combination of immedi-ate pigment darkening (IPD) caused by UVA, due to photo-oxidation of preformed melanins and delayed pigment darkening (DPD) occurring approximately 72 hours after UV exposure, which is optimally stimulated by UVB and to a lesser extent by UVA and visible radiation. DPD is accompanied by increases in the number of DOPA-positive melanocytes, and in the synthesis of melanosomes and is associated with changes in the functional state of melanocytes (Szabo, 1967; Pathak, 1985; reviewed by Ortonne, 1990).

Tyrosinase is considered to be the rate-limiting enzyme for the biosynthesis of melanins in epidermal melanocytes, and to represent the major regulatory step in melanogene-sis. This enzyme catalyses both hydroxylation of tyrosine to DOPA and oxidation of DOPA to dopaquinone (Hear-ing, 1987; Prota, 1988; Hearing and Jimenez, 1989). The possibility that tyrosinase also catalyses a third step in the melanin pathway, the oxidation of 5,6-dihydroxyindole to indole-5,6-quinone has been proposed (Pawelek and Lerner, 1978). Tyrosinase is synthesized on ribosomes as a de novo form of 55 kDa, and goes through an extensive process of post-translational modification in the Golgi apparatus where after glycosylation it reaches a mature form of about 70 kDa. Whether the regulation of melanogenesis occurs at transcriptional and/or post-transcriptional levels is not clearly established.

Histoenzymological studies of UV-irradiated skin demonstrates an increased DOPA-oxidase activity (Jimbow et al., 1974; Rosen et al., 1987). Burchill et al. (1990) have shown that, following psoralen-ultraviolet A (PUVA) ther-apy for several weeks, there was an increase in the amounts of tyrosinase in skin biopsies. However, few direct data are available concerning the catalytic activities of tyrosinase, the abundance of tyrosinase molecules and tyrosinase mRNA levels in UV-irradiated skin.

The availability of techniques allowing pure melanocyte cultures has allowed a better understanding of the direct effects of UV radiations on melanocytes. This led to the demonstration that UV radiations directly induce melanin pigment production in cultured human melanocytes (Fried-mann and Gilchrest, 1987; Halaban et al., 1988; Ramirez-Bosca et al., 1992). These studies were performed using a rather wide spectrum of wavelengths with both UVA and UVB. Different schedules of irradiations have been employed ranging from repeated irradiations at low doses to a single irradiation at high dose. However, human skin is naturally submitted to chronic exposure to sunlight. Taking this fact and the critical role of UVB in the DPD response into account, this study seeks to examine the specific effect of repeated doses of UVB on melanogenesis in vitro. We show that all the UVB irradiations are not effec-tive in inducing stimulation of tyrosinase activity and de novo synthesis of melanin in normal human melanocytes. In contrast, we show that daily irradiations of S-91 murine melanoma cells led to increasing tyrosinase activity and de novo synthesis of melanin in response to each irradiation. To fully interpret these results, we discuss the regulation of tyrosinase at the levels of expression and enzyme activity.

Biochemicals and radiochemicals were from the following sources: DMEM and MCDB 153 media, insulin, hydrocortisone, phorbol 12-myristate 13-acetate (PMA) and forskolin from Sigma (St. Louis, MO, USA). Trypsin, ethylenediaminetetraacetic acid (EDTA) and dispase were from Boehringer (Mannheim, FRG), and fetal calf serum (FCS) from Gibco (New York, NY, USA). L-[ring-3,5-³H]tyrosine, (40-60 Ci/mmol; 1 mCi/ml); L-3,4-dihy-droxyphenyl-[3-¹⁴C]alanine (7-12 mCi/mol, 50 μCi/ml); [α-^32P]dCTP (3000 Ci/mmol); [³⁵S]methionine in vivo cell-labelling grade (1000 Ci/mmole, 10 mCi/ml) were from Amersham (Buck-inghamshire, England).

S91-M3 (Flow Laboratories, Irvine, Scotland) murine melanoma cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) with 10% fetal calf serum and penicillin/streptomycin (100 i.u./50 μg per ml) in a humidified atmosphere containing 5% CO₂ in air at 37°C. All the experiments were performed using cells at the same passage.

Human melanocytes were grown by a method modified from that of Eisenger and Marko (1982). Epidermal cell suspensions, obtained from foreskins of caucasoid children by overnight digestion in PBS containing 0.5% dispase grade II at 4°C, followed by a 1 hour digestion in trypsin/EDTA solution (0.05%/0.02% in PBS) at 37°C were grown in MCDB 153 medium supplemented with 5 μg/ml insulin, 0.5 μg/ml hydrocortisone, 16 nM PMA, 10 μM forskolin, 30 μg protein/ml of bovine pituitaries extract (bPE, prepared according to the procedure of Wilkins et al. (1985)) and penicillin/streptomycin (100 i.u./50 μg/ml) in a humidified atmos-phere containing 5% CO₂ in air at 37°C. Medium was changed three times weekly. Pure melanocyte subcultures were trypsinized every 3 weeks and maintained for 5 to 8 passages. Second or third passage melanocytes were used for these experiments.

The source of ultraviolet radiation was a Bio-Sun [R] Vilber Lour-mat stimulator (Marne la Vallée, France) fitted out with a UVB irradiation source composed of Vilber Lourmat tubes T-20.M 312 nm (no UVA, no UVC emission), mercury vapour tubes, low pres-sure, hot cathodes and with a Vilber Lourmat RMX-365/312 radiometer with microprocessor programmable in energy (milli-Joules) with time basis enabling 6 irradiation measurements per second for controlling the energy received by the sample.

S-91 murine melanoma cells were irradiated with 30 mJ/cm² and normal human melanocytes with 100 mJ/cm² each day during a period of 6 days. One day before the beginning of the pho-totreatment, normal human melanocytes were placed in MCDB 153 supplemented with 2% FCS and 30 μg protein/ml of bPE (minimal melanocytes culture medium). Before each irradiation the culture medium was replaced with PBS to avoid formation of medium-derived toxic photoproducts. Immediately after irradia-tion, PBS was replaced with minimal melanocyte culture medium, which was used until the end of the phototreatment to avoid non-specific effect of growth factors, PMA or forskolin. S-91 murine melanoma cells were always maintained in DMEM-10% FCS. Sham-irradiated cells (control cells) were manipulated as irradi-ated cells, except for the step of irradiation itself (each day, medium was replaced by PBS for 2 minutes and cells maintained in the hood during this time).

Following trypsination (0.05% trypsin, 0.02% EDTA in PBS), the number of living cells was counted with an hematocytometer chamber. Dead cells were excluded by the trypan blue viability test. All the results were normalized to cell number or to the amount of protein/assay dish using the Bio-Rad protein kit assay.

The tyrosinase assay uses [ring-3,5-³H]tyrosine (40-60 Ci/mmol) and measures the ³H₂O released as tyrosine is hydroxylated to hydroxyphenylalanine.

Enzymatic activity was estimated by the amount of ³H₂O released into the medium reflecting the accumulated activity of the enzyme over a time lapse of 24 hours, according to Pomerantz (1969) and modified by Oikawa et al. (1972). All values were corrected by subtracting the amount of ³H₂O formed spontaneously from [ring-3,5-³H]tyrosine in culture medium in the absence of cells during the same period. Results were normalized to cell number or pro-tein amount for each determinations made in triplicates.

This method was used to evaluate the enzyme activity in cellular extracts. At the indicated times after irradiation, cells were trypsinized, spun down and the pellets were resuspended in 1% Nonidet P-40/phosphate buffer (0.1 M, pH 6.8) containing pro-tease inhibitors (1 mM phenylmethylsulphonide fluoride (PMSF), 10 μg/ml aprotinin, 50 mM sodium fluoride (NaF) and 1 mM EDTA). After sonication and solubilization, tyrosinase activity was measured according to Pomerantz’s method (1969). Results were normalized to cell number or protein amount for each deter-minations. Each determination was made in triplicate. A unit of tyrosinase was defined as the activity of enzyme that catalyzed the oxydation of 1 μmole of tyrosine in 1 minute.

The melanin formation assay uses L-3,4-dihydroxyphenyl-[3-¹⁴C]alanine (7-12 mCi/mmol) as substrate; 0.5 μCi/ml medium was added immediately after irradiation. Twenty-four hours later, the cells were trypsinized, counted, pelleted and the uptake of [¹⁴C]DOPA into newly synthesized melanins was determined as described by Hearing and Ekel (1976). 3T3 fibroblasts were used as negative control. Results were normalized to cell number or amount of protein determined for each assay. Each determination was made in triplicate.

Cells were scraped and homogenized in 1% Triton X-100/phos-phate buffer (0.1 M, pH 7.2) containing 1 mM PMSF, 10 μg/ml aprotinin, 50 mM NaF and 1 mM EDTA. Protein content was determined by Lowry assay (Lowry, 1951). Samples (20 μg of protein) were boiled in sample buffer (60 mM Tris, pH 6.8, 1% SDS, 10% glycerol) and separated by SDS-PAGE (10% acry-lamide gels). Following electroblotting, the nitrocellulose sheets were incubated in a blocking buffer (0.5% gelatine, 3% BSA, 0.1% Tween-20, 1 mM EDTA in 10 mM Tris, pH 7.4, 0.15 M NaCl) to avoid non-specific binding. The blots were then incu-bated with mouse antisera directed against human tyrosinase (Bouchard et al., 1993) and followed by incubation with peroxi-dase-conjugated second antibody. The antibody-antigen complex was revealed using the enhanced chemiluminescence (ECL) west-ern blotting kit (Amersham).

Cells (4 ×10⁶) were pre-incubated for 3 hours at 37°C in methio-nine-free medium with 2% dialyzed FCS. Immediately after irra-diation, melanocytes were metabolically labeled for 4 hours with 100 μCi/ml [³⁵S]methionine in the same medium. Cells were washed three times with ice-cold PBS prior to scraping and sol-ubilized by shaking in TBS (150 mM NaCl, 10 mM Tris, pH 7.6) supplemented with 1% NP-40 and protease inhibitors (1 mM PMSF, 10 μg/ml aprotinin, 50 mM NaF and 1 mM EDTA) for 1 hour at 4°C. Radioactive counts incorporated into proteins were precipitated with trichloroacetic acid.

Total cellular RNAs were purified using the single step isolation method of acid guanidinium thiocyanate-phenol-chloroform RNA extraction described by Chomczynski and Sacchi (1987). For northern blot analysis, total RNA (15 μg per lane) was fraction-ated on 1% formaldehyde agarose gels, stained with ethidium bro-mide, and transferred onto Nytran nylon filters by capillary trans-fer. DNA probes (human tyrosinase, pRHOHT2 generously provided by Dr Shibahara and human or mouse GAPDH) were labelled with [α-³²P]dCTP using a random priming kit (Stratagen, USA).

The extent of pigmentation in humans varies, depending on the genetic background of the individual. This differ-ence can also be expressed in vitro (Halaban et al., 1983). Therefore, we first determined an average effective dose of UVB irradiations by which melanogenesis in human melanocytes from different donors could be significantly stimulated with minimal cytotoxic effect, as determined by the trypan blue viability test. This optimum experimental condition was determined on cultures of human melanocytes (from 8 donors), which were irradiated daily with 10, 30, 50 and 100 mJ/cm² over a period of 1 to 8 days. Phototreatment efficiency was evaluated by measur-ing the formation of melanin. Among the 8 cultures tested, 7 were maximally stimulated after the 6th irradiation ses-sion consisting of 100 mJ/cm², with intensities ranging from 2-to 5-fold as compared to the basal level of non-irradiated cells from the same donor. Melanocytes were therefore irradiated during 6 days at 100 mJ/cm² per day in our experiments.

When human melanocytes were grown in complete growth medium, they appeared mostly elongated and bipo-lar with only a few cells exhibiting some degree of den-dricity, being typical of cells cultured in the presence of phorbol esters (Fig. 1a). When growth factors were with-drawn and melanocytes maintained in minimal melanocyte culture medium, the cells appeared bipolar with shortened cytosolic prolongation and enlarged cytoplasm (Fig. 1b). Following 6 daily UVB exposures at 100 mJ/cm², melanocytes displayed a substantial increase in length and number of dendritic processes. In addition, at this time, the melanocytes always appeared healthy and without vacuo-lation or other apparent evidence of damage (Fig. 1c).

To determine the effects of UVB irradiation on pigment formation, de novo synthesis of melanin determined by uptake of [¹⁴C]DOPA and tyrosinase activity was analyzed in living cells, 24 hours after the 6th irradiation session (Table 1). Treatment consisting of 6 daily exposures of 100 mJ/cm² resulted, for all the cultures tested, in a significant increase in tyrosine hydroxylase activity (determined in situ as the release of tritiated water in the medium). [¹⁴C]DOPA uptake was significantly increased in 6 cultures among the 8 examined. The amount of tyrosinase activity determined in medium samples according to the method of Oikawa et al. (1972), reflects the accumulated enzyme activity that occurred in the melanocytes for a period of 24 hours fol-lowing the irradiation. We also measured tyrosinase activity using the homogenate assay (Fig. 2). Tyrosinase activity was then higher, because of the solubilization of membrane-bound tyrosinase by the detergent NP-40. This discrepancy between in situ and homogenate assays has also been observed by Iozumi et al. (1993) who suggest that melanocytes contain a significant amount of tyrosinase, which in the living cell does not exist in an optimum cat-alytic state. More important however, the homogenate assay revealed that, as soon as 3 hours after the irradiation, tyrosi-nase activity was stimulated, compared to the basal activity of non-irradiated melanocytes. Surprisingly, when tyrosi-nase activity was measured 24 hours after the sixth irradi-ation, we observed a significant decrease, compared to the basal activity. This suggests that tyrosinase activation by UVB is followed by an inhibitory step, which may be inter-preted as a negative regulatory mechanism.

Data in the literature are contradictory concerning the correlation between tyrosinase activity and the abundance of enzyme. To verify if the changes observed in tyrosinase activity after UVB treatment could be related to expression of tyrosinase, northern and western blot analyses of human melanocytes were performed 3 and 24 hours after 6 irradiations. A northern blot of total mRNA from human melanocytes probed for tyrosinase is shown in Fig. 3. There was no clear change in the level of tyrosinase mRNA at 3 hours following 6 UVB irradiations compared to the control, non-irradiated cultures. However, 24 hours later, tyrosi-nase mRNA was hardly detectable, compared to the basal and 3 hours post-irradiation levels. The constant expression of GAPDH mRNA showed that the decrease in tyrosinase mRNA is specific and related to regulation of melanogene-sis. To establish a correlation between the mRNA synthesis and tyrosinase protein produced upon exposure to UVB, we performed western blot analysis of proteins isolated from melanocytes 3 and 24 hours after six irradiations. Fig. 4 shows that the anti-human tyrosinase serum specifically rec-ognized the mature (67-69 kDa) and the newly synthesized (55 kDa) forms of the enzyme. In non-irradiated cells, the newly synthesized form of tyrosinase was constitutively expressed but less than the mature form. Three hours after the sixth irradiation, expression of both forms was markedly increased compared to control melanocytes. Twenty-four hours after the last irradiation, the expression of the newly synthesized form decreased as compared to that of the non-irradiated or 3 hours post-irradiation melanocytes. On the other hand, expression of the mature form was decreased only relative to the 3 hours post-irradiation melanocytes, but remained higher than expression in control cells. The decrease in the expression of tyrosinase observed in irradi-ated cells at this stage was specific to this enzyme and not a consequence of a general decrease in protein synthesis fol-lowing UVB irradiations, as shown by the levels of [³⁵S]methionine incorporation into proteins. Indeed, follow-ing metabolic labelling of the irradiated and sham-irradiated melanocytes for 6 days, cpm in TCA-precipited proteins were identical in both treatment groups (2.656×10⁶±1.065× 10⁶ cpm/μg protein for control cells and 2.674×10⁶±1.113× 10⁶ cpm/μg protein for irradiated cells).

In vivo, DPD is observed 72 hours after UV irradiation and since UVB is responsible for delayed tanning, we next ana-lyzed melanogenic activity after each irradiation. Tyrosi-nase assays were then performed in situ after 24 hours and in cell homogenates after 3 and 24 hours; incorporation of [¹⁴C]DOPA into melanins was also determined (Fig. 5).

Interestingly, we observed that tyrosinase activity in situ and incorporation of [¹⁴C]DOPA into melanins progressively increased after the first 4 irradiations. The melanogenic activity then decreased after 5 irradiations, resumed at 6 irradiations and again slowed down following irradiation 7. This result suggests that repeated expo-sures to UVB induce an ‘on and off’ stimulation of the melanogenic activity.

When tyrosinase activity was measured using the homogenates assay, at 3 hours two main peaks of activity (after 4 and 6 irradiations) were observed and correlated with the increased activity detected by the in situ assay. When the assay was performed 24 hours after irradiation, values were close to or under the basal level for all the ses-sions. This indicates to us that tyrosinase activity is increased in the first hours after an effective irradiation and that this stimulation is followed by a refractory period. In addition, reduced activity of the enzyme after 5 and 7 irra-diations (as shown by tyrosinase activity in situ and incorporation of [¹⁴C]DOPA) was in this case expressed by values identical to the basal enzyme level of sham-irradiated cells.

It has been shown that, in S-91 melanoma cells, UV irra-diations result in an increase in melanin synthesis and a decrease in cellular proliferation (Friedman and Gilchrest, 1987). We further examined these melanoma cells to deter-mine whether tyrosinase activity was regulated by serial UVB irradiations in a way similar to that in normal human melanocytes. A dose-determining experiment was also per-formed for murine melanoma cells and showed that doses of up to 50 mJ/cm² were cytotoxic, resulting in cell death. S-91 melanoma cells, irradiated for 6 days with 30 mJ/cm², showed a slight but significant increase in [¹⁴C]DOPA uptake and a 2-fold increase in tyrosinase activity measured in living cells. The striking difference between these and normal human melanocytes was that tyrosinase activity in the cell extract, measured 24 hours after the last irradiation, exhibited a 4-fold increase, compared to the basal activity of non-irradiated melanoma cells (Fig. 6). We then per-formed a western blot analysis to correlate this increased enzyme activity with levels of tyrosinase expression. Fig. 7 shows that, in contrast to normal human melanocytes, no decrease in the amount of tyrosinase was observed 24 hours after the last irradiation as compared to basal and 3 hours-stimulated levels. In addition, we observed an increase in the expression of the newly synthesized form, 3 and 24 hours after the last irradiation as compared to non-irradi-ated cells. These results were confirmed by northern blot analysis. Levels of tyrosinase mRNA were higher at 3 and 24 hours after the last irradiation as compared to the level in non-irradiated cells (data not shown).

In this study we have shown that the melanogenic activity of the melanocytes stimulated by repeated UVB irradiations and evaluated by [¹⁴C]DOPA uptake and tyrosine hydrox-ylase activity of the tyrosinase measured in intact cells was increased. However, the activity of the tyrosinase assayed in cell extracts, 24 hours after the sixth irradiation, was sig-nificantly repressed (Fig. 2). This inhibition was repro-ducible among the cells of all the patients tested and suggested a negative feedback regulation of the activity of the tyrosinase. This idea was later confirmed in our studies of the melanogenic response at different days of the pho-totreatment. This experiment revealed that not all the irradiation sessions are effective in stimulating tyrosinase activity (Fig. 5). One hypothesis to explain this latency is that a protective mechanism exists that prevents melanocytes from over-producing melanin in response to the uncontrolled stimulation represented by the exposure to external UV. This negative regulation occured at a certain degree of melanization and was reversible.

Two hypotheses have been raised to understand the specific tyrosinase related mechanisms responsible for the stimulation of melanogenesis. An increase in the rate of de novo synthesis of tyrosinase could be one possible explanation (Wong and Pawelek, 1973; Halaban et al., 1984). On the other hand, activation by functional association of immature pre-existing enzyme could also contribute to the increase in tyrosinase activity (Wong and Pawelek, 1975; Fuller et al., 1987; Naeyaert et al., 1991). Our results show that, at least in the case of normal human melanocytes repetitively irradiated with UVB, these two mechanisms are not mutually exclusive. Indeed, mammalian tyrosinase is composed of different forms, which reflect different stages of the post-translational processing of a single polypeptide precursor. Among them, a de novo synthesized form (55,000 Da) and a mature form (67,000-69,000 Da) have been described (Hearing, 1987). When we analyzed the amount of tyrosinase present in the melanocytes 3 and 24 hours after the sixth irradiation, we could clearly dis-tinguish between these two isoforms (Fig. 4). Our experi-ments have shown that, in non-irradiated cells, the mature form is constitutively expressed, which agrees with the presence of a pre-existing mature tyrosinase. In addition, it seemed that the increase in the enzyme activity observed 3 hours after the 6th irradiation correlated with increases in both the newly synthesized and mature forms. The level of tyrosine hydroxylase activity measured in cellular homogenates 24 hours after the sixth irradiation was below background (Fig. 2). However, at this stage, melanocytes possessed the same amount of mature tyrosinase compared to sham-irradiated control melanocytes (Fig. 4). This obser-vation backs up the hypothesis of the existence of a melanogenic inhibitor as described by Kameyama et al. (1989a, 1993). The fact that low tyrosinase activity (under basal level) was associated with a basal amount of mature form agrees with previous reports (Kameyama et al., 1989b; Jimenez et al., 1988; Naeyaert et al., 1991; Kameyama et al., 1993) on the post-translational regulation of melanogenic activity. On the other hand, this decrease in enzyme activity was related to a decrease in the expression of the newly synthesized form of tyrosinase (Fig. 4). This was confirmed by northern blot analysis, which showed that mRNA for tyrosinase was hardly expressed in melanocytes 24 hours after the irradiation (Fig. 3). However, one would expect that the increase in the newly synthesized form at 3 hours after the 6th irradiation (Fig. 4) would correlate with an increase in mRNA for tyrosinase at the same time. An explanation for this apparent discrepancy may be the insta-bility of tyrosinase messenger. Indeed it is possible that the increase in mRNA, which accounts for the increase in the newly synthesized form of the enzyme observed at 3 hours, had occured earlier and was no longer detectable at this time because of a rapid turn over of the messenger. Anyway, the increase in the newly synthesized form observed in Fig. 4 was real since we could confirm it by immunoprecipitation following 4 hours of metabolic label-ing (data not shown).

Taken together, these results show that, in normal human melanocytes, long-term regulation of tyrosinase activity by UVB irradiation is directly related to the amount of the newly synthesized form of the enzyme, while the rela-tionship between enzyme activity and amount of mature form suggests the existence of a melanogenic inhibitor. Another interesting point revealed by this study is that the possible regulatory mechanism described for normal human melanocytes irradiated with repetitive doses of UVB was not observed with S-91 melanoma cells. Indeed, these cells responded to serial stimulation by an increas-ing activity of the tyrosinase without any refractory period (Fig. 6). The finding that these cells exhibit a ‘physiolog-ical’ response to UV has already been reported by Fried-mann and Gilchrest (1987). The new information that we provide is that the stimulation of the enzyme activity is related to an increase in the amount of tyrosinase 3 and 24 hours after six irradiations (Fig. 7). The main change con-cerns the de novo synthesized form while only a slight increase in the expression of the mature form is observed at 24 hours. One hypothesis to explain this phenomenon is that in transformed melanocytes the ability of the cells to protect themselves from repeated irradiation is lost. However, the hypothesis that ‘self-regulated’ melanogen-esis induced by UVB exposures has been specifically developed by human melanocytes cannot be ruled out. Therefore, complementary studies with human melanoma cells are needed.

In conclusion, our results indicate for the first time that during long term UVB treatment not all the irradiations are effective to induce stimulation of tyrosinase activity and de novo synthesis of melanin. This may suggest a regulatory mechanism developed by the melanocytes against over pro-duction of melanin, which seems to involve regulation of the amount of tyrosinase as well as a melanogenic inhibitor.

We thank Dr Bouchard and Dr Shibahara for their generous gifts of tyrosinase antibody and tyrosinase cDNA as well as Dr Marinkovich for critical reading of this manuscript. This work was supported by Hoffman-La Roche, La Ligue Nationale Française Contre Le Cancer, GEFLUC, and the Association Vaincre Le Mélanome.