Dimethyl Itaconate Reduces α-MSH-Induced Pigmentation via Modulation of AKT and p38 MAPK Signaling Pathways in B16F10 Mouse Melanoma Cells

Dimethyl itaconate (DMI) exhibits an anti-inflammatory effect. Activation of nuclear factor erythroid 2-related factor 2 (NRF2) is implicated in the inhibition of melanogenesis. Therefore, DMI and itaconic acid (ITA), classified as NRF2 activators, have potential uses in hyperpigmentation reduction. The activity of cyclic adenosine monophosphate (cAMP) response element-binding protein (CREB), an important transcription factor for MITF gene promoter, is regulated by glycogen synthase kinase 3β (GSK3β) and protein kinase A (PKA). Here, we investigated the inhibitory effect of ITA and DMI on alpha-melanocyte-stimulating hormone (α-MSH)-induced MITF expression and the modulatory role of protein kinase B (AKT) and GSK3β in melanogenesis in B16F10 mouse melanoma cells. These cells were incubated with α-MSH alone or in combination with ITA or DMI. Proteins were visualized and quantified using immunoblotting and densitometry. Compared to ITA, DMI treatment exhibited a better inhibitory effect on the α-MSH-induced expression of melanogenic proteins such as MITF. Our data indicate that DMI exerts its anti-melanogenic effect via modulation of the p38 mitogen-activated protein kinase (MAPK) and AKT signaling pathways. In conclusion, DMI may be an effective therapeutic agent for both inflammation and hyperpigmentation.


Introduction
The skin, the largest organ of the human body, provides protection against harmful external stresses, such as ultraviolet (UV) radiation and environmental pollutants [1,2]. Various cell types, including keratinocytes, melanocytes, fibroblasts, and skin-resident macrophage cells (Langerhans cells), mingle to form the skin [3]. Located in the outermost layer of the human skin, keratinocytes are mounted above melanocytes outnumbering them by approximately 10 to 1 [4]. Compared to keratinocytes, melanocytes are resistant to apoptosis owing to the presence of high levels of B-cell lymphoma 2 protein, which is also highly expressed in cancer cells [5,6]. Melanocytes proliferate slowly under normal conditions [4]. Microphthalmia-associated transcription factor (MITF) regulates melanin synthesis (melanogenesis) through timely and controlled expression of melanogenic genes such as tyrosinase (TYR) along with its own phosphorylation [7,8]. In response to stress such as UV radiation, keratinocytes secrete paracrine factors such as alpha-melanocytestimulating hormone (α-MSH) and stem cell factor (SCF). The binding of α-MSH and SCF to melanocortin-1-receptor (MC1R) and c-Kit, respectively, initiates signal transduction for melanogenesis [9,10]. The incoming signals from different receptors then converge to control the transcription and activity of MITF [8,11]. Melanin produced by melanocytes is eventually transferred to surrounding keratinocytes and exhibits characteristic pigmentation on the skin surface [12].
four-fold in the α-MSH-only sample. Melanin production was unaffected by treatment with ITA at all the tested concentrations, while DMI inhibited melanin production by 28,36, and 42% at 20, 40, and 80 μM concentrations, respectively ( Figure 1b). All samples showed >110% cell viability except for the 80 μM DMI-treated sample in which the cell viability was 97% (Figure 1c). These data indicated that ITA and DMI did not significantly affect cell viability at all the tested concentrations. The cell viability for all treatment samples was over 110% except 80 μM DMI at which the cell viability was about 97%. The data were presented as the mean ± standard deviation of three independent experiments; * p < 0.05 and ** p < 0.01 compared with α-MSH only and untreated cells for melanin assay and cell viability assay, respectively.

L-DOPA Oxidation
Tyrosinase (TYR) functions as a rate-limiting enzyme in melanin production [38]. Oxidation of L-DOPA by intracellular cell lysate is frequently used to estimate TYR activity [39][40][41][42]. However, the studies by Schallreuter et al. [43], Land et al. [44], and Plonka et al. [45] reported that tyrosinases require L-DOPA for their own activation and may not produce dopaquinone via DOPA. We first investigated whether DMI exerted its inhibitory effect on α-MSH-induced melanin production through the inhibition of L-DOPA oxidation. B16F10 cells were treated with ITA and DMI at different concentrations (20,40, and 80 μM), and melanin production was induced by adding α-MSH. Relative to the untreated control, α-MSH induced a 3.3-fold increase in L-DOPA oxidation. ITA treatment did not cause any statistically significant inhibition of L-DOPA oxidation (98, 100, and 99% L-DOPA oxidation at 20, 40, and 80 μM ITA, respectively) ( Figure 2). In contrast, DMI treatment at 20, 40, and 80 μM concentrations inhibited α-MSH-induced L-DOPA oxidation by 0.5, 8, and 14%, respectively ( Figure 2). Our data indicated that the melanin content was simultaneously decreased with the decrease in α-MSH-induced L-DOPA oxidation. itaconate showed a significant reduction in α-MSH-induced melanin production relative to itaconate acid. (c) The cell viability for all treatment samples was over 110% except 80 µM DMI at which the cell viability was about 97%. The data were presented as the mean ± standard deviation of three independent experiments; * p < 0.05 and ** p < 0.01 compared with α-MSH only and untreated cells for melanin assay and cell viability assay, respectively.

Time-Dependent Activation of AKT, GSK3β and MAPK Signaling Pathways in Response to α-MSH Treatment
According to the review by Cargnello and Roux, the three MAPK enzymes of extracellular signal-regulated kinase (ERK), c-Jun N-terminal kinase (JNK), and p38 regulate melanogenesis by relaying extracellular signals to intracellular responses [46]. For example, the binding of α-MSH to MC1R transactivates c-Kit/Ras/Raf/MEK1/2/ERK/RSK signaling [23]. In another instance, the phosphorylation of MITF at Ser73 and Ser409 by ERK and RSK enhances both transcriptional activity and degradation of MITF [24,25].
Previous studies showed the inhibitory role of the PI3K/AKT signaling pathway in melanogenesis. Shin et al. [30] suggested that NRF2 exerted an anti-melanogenic effect through the activation of PI3K/AKT/mTOR signaling in normal human epidermal melanocytes [30], Mosca et al. [47] demonstrated negative feedback on melanogenesis via the activation of the α-MSH-induced PI3K pathway [47], and Oka et al. [48] reported that PI3K inhibition increases melanin production while constitutively active mutant of AKT inhibits melanogenesis.

Time-Dependent Activation of AKT, GSK3β and MAPK Signaling Pathways in Response to α-MSH Treatment
According to the review by Cargnello and Roux, the three MAPK enzymes of extracellular signal-regulated kinase (ERK), c-Jun N-terminal kinase (JNK), and p38 regulate melanogenesis by relaying extracellular signals to intracellular responses [46]. For example, the binding of α-MSH to MC1R transactivates c-Kit/Ras/Raf/MEK1/2/ERK/RSK signaling [23]. In another instance, the phosphorylation of MITF at Ser73 and Ser409 by ERK and RSK enhances both transcriptional activity and degradation of MITF [24,25].
Due to the significant involvement of MAPK, AKT, and GSK3β signaling pathways in the α-MSH-induced transcription of MITF and melanogenesis, we investigated the timedependent phosphorylation of MAPK proteins, AKT, and GSKβ in B16F10 melanoma cells at five time points (0, 15, 30, 60, and 120 min) after α-MSH treatment. When normalized to the expression level of phosphorylated ERK (p-ERK) to that of total ERK and compared to the basal expression level at time zero, the relative phosphorylation levels of ERK were approximately 568, 542, 347, and 243% at 15, 30, 60, and 120 min after α-MSH stimulus, respectively ( Figure 4a). The phosphorylation of p38 induces phosphorylation of CREB which in turn stimulates the expression of MITF [27]. When normalized to that of p38 and compared to the basal expression level at time zero, the relative phosphorylation levels of p38 were approximately 347, 284, 258, and 145% at 15, 30, 60, and 120 min after α-MSH treatment, respectively (  [26,51]. The time-course analysis of the phosphorylation of JNK showed that the phosphorylation level of JNK relative to that at time zero was increased by approximately 5% at 15 min post-α-MSH treatment but decreased to approximately 82, 42, and 51% at 30, 60, and 120 min, respectively ( Figure 4c).  The data were presented as the mean ± standard deviation of three independent experiments; * p < 0.05, ** p < 0.01, *** p < 0.005, # p < 0.05 compared with α-MSH only.

Effect of ITA and DMI on the α-MSH-Induced Activation of MAPK, AKT, and GSK3β Signlaing Pathways
From the time-course experiment, it was found that 15 min after α-MSH treatment, the phosphorylation of MAPKs peaked although the phosphorylation of AKT was maximally inhibited at 30 min after α-MSH treatment. Therefore, this time point (15 min) was used to assess the effect of ITA and DMI on α-MSH-induced phosphorylation of MAPKs (ERK, JNK, and p38) and AKT. The band intensities for phosphorylated proteins were normalized to their respective total protein levels. The band intensities for MITF, p-GSK3β, and GSK3β were normalized to β-actin. The phosphorylation levels of MAPK, MITF, p-GSK3β, and GSK3β were compared with those of α-MSH only. GSK3β. (f) The relative protein expression of MITF. The data were presented as the mean ± standard deviation of three independent experiments; * p < 0.05, ** p < 0.01, *** p < 0.005, # p < 0.05 compared with α-MSH only.
In addition to α-MSH-triggered activation of MAPK proteins, we investigated the timedependent phosphorylation of AKT and GSK in response to α-MSH. The phosphorylation level of AKT at each time point was normalized to that of total AKT and compared to that at time zero. In line with the research by Mosca et al. [47], the expression level of p-AKT decreased by approximately 45 and 63% at 15 and 30 min after α-MSH treatment, respectively (Figure 4d). However, the relative phosphorylation level of AKT at 1 h bounced back to that at time zero and further increased to approximately 130% at 2 h after α-MSH treatment (Figure 4d). The time-dependent increase or decrease in GSK3β phosphorylation following α-MSH treatment was similar to that of AKT (Figure 4d,e). When normalized to β-actin, the α-MSH-induced phosphorylation level of GSKβ relative to that at time zero was approximately 59, 70, 100, and 110% at 15, 30, 60, and 120 min, respectively, while the protein expression level of GSK3β relative to a zero point was approximately 79, 89, 100, and 115% at their respective time points (Figure 4e). When compared to those at time zero, the relative expression levels of p-GSKβ to GSKβ were approximately 75, 78, 100, and 96% at 15, 30, 60, and 120 min after α-MSH treatment (Figure 4e).
The study by Kim et al. [52] indicated that the inhibition of AKT by α-MSH is concomitant with the down-regulation of phosphorylated mTOR (p-mTOR). In addition, the study by Hah et al. showed that mTOR inhibition by rapamycin induces melanogenesis in human MNT-1 melanoma cells [32]. Taken together, these studies indicated that α-MSH treatment inhibits the activation of AKT/mTOR signaling and induces melanogenesis. In addition to the inhibitory role of GSK3β in the DNA binding activity of CREB, p-AKTmediated GSK3β inhibition promotes melanin production [49]. In agreement with the results of the aforementioned studies, our results showed that after α-MSH treatment there was a short period of time within which the α-MSH-induced expression level of p-AKT, p-GSK3β, and GSK3β was below those at time zero (Figure 4d,e), which may contribute to α-MSH-induced melanogenesis.
MAPK, AKT, and GSK3β signaling pathways converge to initiate MITF transcription [10,20]. When normalized to that of β-actin and compared to the basal expression level at time zero, the relative expression level of MITF increased by 1.11-, 1.26-, 1.33-, and 1.9-fold at 15, 30, 60, and 120 min after α-MSH treatment, respectively (Figure 4f). The relative expression level of MITF continued to increase over the tested time points.

Effect of ITA and DMI on the α-MSH-Induced Activation of MAPK, AKT, and GSK3β Signlaing Pathways
From the time-course experiment, it was found that 15 min after α-MSH treatment, the phosphorylation of MAPKs peaked although the phosphorylation of AKT was maximally inhibited at 30 min after α-MSH treatment. Therefore, this time point (15 min) was used to assess the effect of ITA and DMI on α-MSH-induced phosphorylation of MAPKs (ERK, JNK, and p38) and AKT. The band intensities for phosphorylated proteins were normalized to their respective total protein levels. The band intensities for MITF, p-GSK3β, and GSK3β were normalized to β-actin. The phosphorylation levels of MAPK, MITF, p-GSK3β, and GSK3β were compared with those of α-MSH only.
The data indicate that DMI may inhibit α-MSH-induced pigmentation via modulation of p38 signaling. noma cells. As expected, at 15 min after the α-MSH treatment, α-MSH suppressed p-A expression. Our data showed that DMI at 40 and 80 μM concentrations exerted a bet stimulatory effect on p-AKT expression relative to both α-MSH-only treatment as well α-MSH and ITA co-treatments (Figure 6a). Neither ITA nor DMI has a statistically sign icant effect on the α-MSH-induced expression of p-GSK3β and GSK3β (Figure 6b).  (c) Effect of ITA and DMI on the α-MSH-induced expression of p38. The data were presented as the mean ± standard deviation of three independent experiments; * p < 0.05, ** p < 0.01, *** p < 0.005 compared with α-MSH only.
In addition to MAPK signaling, we examined the depigmentation effect of ITA and DMI through PI3K/AKT/mTOR axis and GSK3β signaling in α-MSH-treated B16 melanoma cells. As expected, at 15 min after the α-MSH treatment, α-MSH suppressed p-AKT expression. Our data showed that DMI at 40 and 80 µM concentrations exerted a better stimulatory effect on p-AKT expression relative to both α-MSH-only treatment as well as α-MSH and ITA co-treatments (Figure 6a). Neither ITA nor DMI has a statistically significant effect on the α-MSH-induced expression of p-GSK3β and GSK3β (Figure 6b NRF2 activators exhibit their hypopigmentation effects via modulation of the PI3K/AKT/mTOR/autophagy axis [30][31][32][33]. p-AKT inactivates GSK3β by phosphorylation at Ser9 [49], leading to the enhanced transcription of MITF. In addition, GSK3β inhibits CREB DNA binding activity while GSK3β inhibition promotes melanogenesis in B16 melanoma and normal human melanocytes [49,50]. Taken together, the activation of mTOR and GSK3β signaling pathways are associated with the inhibition of melanogenesis. To further evaluate the depigmentation effect of ITA and DMI via α-MSH-activated AKT and GSK3β signaling pathways, B16F10 cells were cultured in the presence or absence of α-MSH in combination with ITA or DMI (40 μM) for 6 h. The expression level of p-AKT was normalized to that of total AKT. The expression levels of p-GSK3β, GSK3β, and MITF were normalized to those of β-actin. The relative expression level of each protein was then compared to the basal expression level in untreated or α-MSH-stimulated cells. Compared to the untreated control cells, the treatment of ITA and DMI either alone or in combination with α-MSH increased the relative expression levels of p-AKT, p-GSK3β, and GSK3β (Figure 7a,b,c). However, when compared to the co-treatment group of ITA and α-MSH, the co-treatment group of DMI and α-MSH showed stronger and weaker inhibitory effects on the α-MSH-induced expression of MITF and p-AKT, respectively (Figure 7a,d). Our data showed that the increased expression of p-GSK3β and GSK3β correlated positively with the increased expression of p-AKT (Figure 7a,b,c). The results thus indicated that DMI at 40 μM exerted its inhibitory effect on α-MSH-induced MITF expression via modulation of AKT and GSK3β signaling pathways. NRF2 activators exhibit their hypopigmentation effects via modulation of the PI3K/ AKT/mTOR/autophagy axis [30][31][32][33]. p-AKT inactivates GSK3β by phosphorylation at Ser9 [49], leading to the enhanced transcription of MITF. In addition, GSK3β inhibits CREB DNA binding activity while GSK3β inhibition promotes melanogenesis in B16 melanoma and normal human melanocytes [49,50]. Taken together, the activation of mTOR and GSK3β signaling pathways are associated with the inhibition of melanogenesis.
To further evaluate the depigmentation effect of ITA and DMI via α-MSH-activated AKT and GSK3β signaling pathways, B16F10 cells were cultured in the presence or absence of α-MSH in combination with ITA or DMI (40 µM) for 6 h. The expression level of p-AKT was normalized to that of total AKT. The expression levels of p-GSK3β, GSK3β, and MITF were normalized to those of β-actin. The relative expression level of each protein was then compared to the basal expression level in untreated or α-MSH-stimulated cells. Compared to the untreated control cells, the treatment of ITA and DMI either alone or in combination with α-MSH increased the relative expression levels of p-AKT, p-GSK3β, and GSK3β (Figure 7a-c). However, when compared to the co-treatment group of ITA and α-MSH, the co-treatment group of DMI and α-MSH showed stronger and weaker inhibitory effects on the α-MSH-induced expression of MITF and p-AKT, respectively (Figure 7a,d). Our data showed that the increased expression of p-GSK3β and GSK3β correlated positively with the increased expression of p-AKT (Figure 7a-c). The results thus indicated that DMI at 40 µM exerted its inhibitory effect on α-MSH-induced MITF expression via modulation of AKT and GSK3β signaling pathways. The data were presented as the mean ± standard deviation of three independent experiments; * p < 0.05, ** p < 0.01, *** p < 0.005, **** p < 0.001 compared with untreated cells.

Discussion
NRF2 activators such as ITA and DMI provide cells with protection against oxidative stress. In addition, it is suggested that the overexpression of NRF2 may inhibit melanogenesis through the PI3K3/AKT/mTOR axis/autophagy axis [30][31][32][33]. While the anti-inflammatory properties of ITA and DMI have been well examined [34][35][36][37], the depigmentation effects of these two NRF2 activators have not been studied and this is the first to do so. The expression and activity of MITF are modulated through MAPK, AKT, and GSK3β signaling pathways [11,19,20,[46][47][48][49][50]. Our time-course analysis of the phosphorylation of AKT and GSK3β signaling molecules in response to α-MSH treatment showed that there The data were presented as the mean ± standard deviation of three independent experiments; * p < 0.05, ** p < 0.01, *** p < 0.005, **** p < 0.001 compared with untreated cells.

Discussion
NRF2 activators such as ITA and DMI provide cells with protection against oxidative stress. In addition, it is suggested that the overexpression of NRF2 may inhibit melanogenesis through the PI3K3/AKT/mTOR axis/autophagy axis [30][31][32][33]. While the anti-inflammatory properties of ITA and DMI have been well examined [34][35][36][37], the depigmentation effects of these two NRF2 activators have not been studied and this is the first to do so. The expression and activity of MITF are modulated through MAPK, AKT, and GSK3β signaling pathways [11,19,20,[46][47][48][49][50]. Our time-course analysis of the phos-phorylation of AKT and GSK3β signaling molecules in response to α-MSH treatment showed that there was a brief period of time within which the α-MSH-induced expression of p-AKT, p-GSK3β, and GSK3β was below that at time zero and this may enhance MITF transcription and its activity. In the follow-up experiments, the inhibitory effect of DMI and ITA on the α-MSH-induced expression of melanogenesis enzymes and the regulatory role of AKT and GSK3β in melanogenesis were assessed. Our data showed that DMI downregulated the α-MSH-induced expression of melanin synthetic enzymes such as MITF in a concentration-dependent manner in B16F10 cells via modulation of AKT and p38 signaling pathways. DMI exhibited visibly evident depigmentation effects in α-MSH-treated B16F10 melanoma cells, whereas ITA did not. The results of this study are in line with those by Kim et al. [52] that DMI exerted its anti-melanogenic effect by up-regulation of p-AKT and by down-regulation of p38 in B16F10 melanoma cells. As reported by Smalley and Eisen [53], Bellei et al. [54], and Kim et al. [55], our results suggested pigmentation through p38 activation. However, p38 silencing ultimately increases melanin production [54]. Thus, further research on the role of p38 in melanogenesis is required. In addition, L-DOPA oxidation assay alone would not provide accurate measurements for tyrosinase activity [43][44][45]. For this reason, the combined use of L-DOPA oxidation assay and other tyrosinase assay methods such as mushroom tyrosinase-based enzyme inhibition assay and tyrosinase zymography [56] would confirm the anti-tyrosinase activities of DMI. In conclusion, DMI may be an effective therapeutic agent for both inflammation and hyperpigmentation.

Cell Culture
B16 culture medium was composed of 450 mL of DMEM, 50 mL of heat-inactivated (at 56 • C for 30 min) FBS, and 5 mL of NaPy. B16F10 cells (2.5 × 10 5 cells) were plated in 10 mL culture medium. Sub-cultivation was performed every two days by washing once with PBS and a brief rinse with trypsin-EDTA. Trypsin-EDTA was then removed from the plate and incubated at 37 • C and 5% CO 2 . Once the cell layer was dispersed, the cells were collected in the B16 culture medium and centrifuged at 1000 RPM for 3 min. The resulting pellet was resuspended in B16 culture medium, and the cell number was counted.

MTT Assay
The cells (2 × 10 3 cells) were seeded in 200 µL of culture medium per well in a 96-well plate and cultured at 37 • C and 5% CO 2 for 18 h. Further, the cells were treated with α-MSH alone or in combination with three different concentrations (20,40, and 80 µM) of ITA and DMI. After incubation for 48 or 72 h at 37 • C and 5% CO 2 , the culture medium was replaced with B16 medium containing 0.6 mg/mL 3-(4,5-dimehtylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT) and incubated for 1 h at room temperature. DMSO (200 µL per well) was added to the resulting formazan crystals which were then put on an orbital shaker at 200 rpm for 5 min. The color intensities of the plates were measured at an absorbance wavelength of 590 nm. The obtained absorbance values were compared to those of the α-MSH-only samples and converted to percentages.

Melanin Quantification
After seeding in a 96-well plate as described above, the cells were treated with α-MSH alone or in combination with ITA or DMI for 72 h. The absorbance of melanin was measured at 405 nm wavelength. The obtained absorbance values were compared to those of the α-MSH-only samples and converted to percentages.

L-DOPA Oxidation
L-DOPA oxidation was measured as previously described methods with slight modifications [39][40][41][42]. B16F10 cells (5 × 10 4 cells) were seeded at 2 mL per well in two 6-well plates and incubated at 37 • C and 5% CO 2 for 18 h. The cells were then treated with α-MSH (200 nM) alone or with three different concentrations (20,40, and 80 µM) of ITA or DMI and further incubated for 72 h at 37 • C and 5% CO 2 . After one wash in PBS, the cells were lysed in protease inhibitor cocktail-added RIPA buffer and mixed on a rocking shaker for 2 h at 4 • C. The cell lysate was then collected in a 2 mL tube and centrifuged in a refrigerated benchtop centrifuge at 12,000 rpm for 30 min. The supernatant was transferred to a new tube, and the proteins in the tube were quantified using bicinchoninic acid (BCA) assay. L-DOPA powder was solubilized in 0.1 M sodium phosphate buffer (pH 6.8) at a final concentration of 2 mg/mL. The samples were diluted to 1 µg/µL using the previously used PI-added 1X RIPA buffer. Cell lysate (20 µL) and diluted L-DOPA (80 µL) were added to the wells of a 96-well microplate and wrapped in aluminum foil. Following a one-min shaking on an orbital shaker, the plate was incubated at 37 • C until the color difference between the untreated control and sample treatments became clear. The absorbance values were measured at 490 nm using a microplate reader [57].

Immunoblotting
Equal amounts of protein were loaded onto each well in a gel (8% resolving gel, 4% stacking gel), and the proteins were separated by molecular weight. The proteins on the gel were then transferred to a polyvinylidene difluoride membrane using the Trans-Blot Turbo RTA Transfer Kit and Trans-Blot Turbo Transfer System (2.5 A constant, up to 25 V, 13 min). The blot was washed once with 1X TBST and blocked with 5% w/v skim milk-added 1X TBST buffer for 1 h. After washing thrice for 10 min each, the blot was incubated in primary antibody buffer for 18 h on a shaker at 4 • C. After collecting the primary antibody buffer for reuse, the blot was washed twice for 10 min each and incubated in secondary antibody buffer with shaking at room temperature for 1 h. The blots were subsequently washed three times for 10 min each. Further, the proteins present in the blots were visualized using Western ECL substrates and a LAS 4000 MINI Image Reader. Immunoblot bands were quantified using the ImageJ software (NIH, United States). The band intensities for phosphorylated proteins were normalized to their respective total protein levels. The band intensities for MITF, p-GSK3β, and GSK3β were normalized to β-actin