Research Paper

Anti-melanogenic effects of Aster spathulifolius extract in

UVB-exposed C57BL/6J mice and B16F10 melanoma cells
through the regulation of MAPK/ERK and AKT/GSK3b
signalling
Ga Yeon Hwanga and Se-Young Chounga,b
a
Department of Preventive Pharmacy and Toxicology, College of Pharmacy, Kyung Hee University and bDepartment of Life and
Nanopharmaceutical Sciences, College of Pharmacy, Kyung Hee University, Seoul, Korea

Keywords Abstract
melanin synthesis, Aster spathulifolius
maxim; MAPK/ERK; Akt/GSK3b; B16F10; Objectives Pharmacological studies of Aster spathulifolius Maxim(AS) have
C57bl/6J demonstrated its anti-allergy, anti-viral and anti-obesity effects, however, its anti-
melanogenic effects is still unclear. In this study, the effects of AS extract (ASE)
Correspondence
on the inhibition of melanin synthesis were investigated in vitro and in vivo.
Se-Young Choung, Department of Life and
Nanopharmaceutical Sciences, College of
Methods To perform this study, the contents of melanin and tyrosinase activity
Pharmacy, Kyung Hee University, 26, were analysed in B16F10 melanoma cells. Western blotting was carried out to
Kyunghee-daero, Dongdaemun-gu, Seoul, determine the underlyling mechanism. Additionally, we investigated the effect of
Korea. this extract on hyperpigmentation in C57bL/6J mice induced by 3, 6 and 9 weeks
E-mail: .sychoung@[Link] of UVB irradiation.
Key findings AS extract led to reduced melanin synthesis through the regulation
Received November 6, 2015
of MITF and its downstream signals. Furthermore, ASE increased the phosphory-
Accepted January 14, 2016
lation of MAPK/ERK and Akt/GSK3b signalling pathway components. In vivo
doi: 10.1111/jphp.12524 study, hypopigmentation effects were also observed. The melanocyte activity and
the distribution of melanin granules were decreased in UVB-irradiated mice trea-
ted with ASE.
Conclusions These results suggest that the ASE may be promising as an active
anti-melanogenic component, and further investigations should be performed
regarding its potential as a whitening agent in the field of cosmetics.

TRP2[3]. The two consecutive steps in this pathway that are

Introduction
regulated by tyrosinase are rate-limiting because the
Melanin is a protective component of the skin that is essen- remainder of the melanin synthesis pathway can proceed
tial for defending skin from toxic chemicals and ultraviolet naturally at a physiological pH[4]. Therefore, tyrosinase is
ray-induced effects, such as premature ageing and skin can- considered a target for whitening agents and has been rec-
cer[1]. However, abnormal hyperpigmentation can result in ognized as the most critical regulator in melanin biosynthe-
pigmentary disorders, such as chloasma, freckles, moles, sis[5,6]. Furthermore, the upregulation of tyrosinase
senile lentigines and allergic contract dermatitis[2]. The expression increases melanin levels[7].
melanogenic pathway is an enzymatic cascade consisting of Microphthalmia-associated transcription factor (MITF)
various melanogenic enzymes, such as tyrosinase, tyrosi- regulates the transcription of melanogenic genes[8]. MITF-
nase-related protein 1 (TRP-1) and dopachrome tau- M, which is expressed in melanocytes and melanoma cells,
tomerase (DCT; TRP2). Tyrosinase is the rate-limiting binds to the tyrosinase promoter and transactivates tyrosi-
enzyme of melanogenesis that catalyses the hydroxylation nase and TRP-1 during melanogenesis[9]. Following sig-
of tyrosine to 3,4-dihydroxyphenylalanine (DOPA) and, nalling pathways play important roles in regulating MITF
subsequently, the oxidation of DOPA to DOPA–quinone. expression, namely, the mitogen-activated protein kinase
The oxidation of DOPA is further catalysed by TRP1 and (MEK)/extracellular signal-regulated protein kinase (ERK)

© 2016 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 68 (2016), pp. 503–513 503
Melanogenesis inhibitory effects of ASE Ga Yeon Hwang and Se-Young Choung

and phosphatidylinositol 3-kinase(PI3K)/Akt/glycogen syn- bovine serum (FBS) and penicillin-streptomycin were
thase kinase 3b (GSK3b) pathways[10,11]. The ERK path- purchased from Hyclone Laboratories Inc. (Logan, UT,
way, which is one of the mitogen-activated protein kinase USA). Synthetic melanin, 3,4- dihydroxy-L-phenylalanine
(MARK) pathways, responds to extracellular stress stimuli (L-DOPA), mushroom tyrosinase, bovine serum albumin
and controls several cellular activities, such as gene expres- (BSA), kojic acid, MTT and dimethyl sulfoxide (DMSO)
sion, differentiation and proliferation[12]. Increased ERK were purchased from Sigma-Aldrich (St. Louis, MO, USA).
phosphorylation can lead to the degradation of MITF. Sodium bromide was purchased from Junsei Chemical Co.,
MITF is known to be phosphorylated at serine 73(S73) by Ltd (Tokyo, Japan). Phosphatase and protease inhibitor cock-
ERK, leading to ubiquitin-dependent proteasomal degrada- tails were purchased from Roche (Mannheim, Germany).
tion; moreover, decreasing MITF phosphorylation leads to Antibodies to b-actin(sc-47778), MITF(sc-56725), tyrosinase
decreased of melanin synthesis[13]. Activated Akt/GSK3b (sc-7833), TRP1(sc-10448), and TRP2(sc-10451) were pur-
signalling also downregulates MITF[14,15], resulting in the chased from Santa Cruz Biotechnology (Santa Cruz, CA,
decreased transcription of tyrosinase and TRP-1. USA). Antibodies to MEK(#9122), p-MEK(#9154), ERK
Many researchers have focused on identifying skin-light- (#9102), p-ERK(#9101), p38(#9212), p-p38(#9211), Akt
ening compounds to treat disorders characterized by hyper- (#9272), p-Akt(#9271), GSK-3b(#9315), p-GSK-3b(#9336),
pigmentation of the skin and hair. Two common whitening PD98059 and LY294002 were purchased from Cell Signaling
compounds include arbutin and kojic acid. However, arbu- Technology (Danvers, MA, USA).
tin promotes pigmentation at concentrations in the range
of 0.5–8 mM in cultured human melanocytes[16], and kojic
Preparation of Aster spathulifolius Maxim
acid induces cancer[17]. Therefore, the needs for identifica-
extract
tion of natural whitening agents are growing rapidly. A vari-
ety of natural sources, including crude extracts and isolated The AS was obtained from the coasts of Ulleung Island in
compounds from medicinal plants, have well-established Korea. The AS was authenticated by a manager at Ulleung-
anti-melanogenic properties and no significant side effects; gun Agriculture Technology Center (Ulleung Island,
such compounds and extracts are therefore promising can- Republic of Korea). The leaves were washed and dried at
didates for safe and effective depigmenting products[18]. 25°C for one day, and the dried leaves were ground into a
The AS (Asteraceae family) is a traditional herbal medi- fine power. The power was refluxed with 50% ethanol at
cine that is widely used in Korea and can be found along 60°C for 4 h; these steps were repeated at least three times.
the eastern and southern coasts of South Korea. The AS has After being filtered, the extract was evaporated using a
been used to treat asthma and diuresis[19], and pharmaco- rotary evaporator (Eyela, Tokyo, Japan) under vacuum and
logical studies have demonstrated its anti-allergy[20], anti- then freeze-dried at
70°C. The powder was stored at this
viral[21] and anti-obesity effects[22]. A variety of these phar- temperature until use (yield: 27.1%).
macological effects were mainly influenced by antioxidant
and free radical scavenging effects[23,24]. But, to our knowl-
High-performance liquid chromatography -
edge, this is the first time study to investigate anti-melano-
electrospray ionization-tandem mass
genesis of ASE by using B16F10 melanoma cells and in a
spectrometry analysis
C57bL/6J mouse model of UVB radiation-induced hyper-
pigmentation. AS extract was studied by high-performance liquid chro-
Previous in vitro study, Asteraceae family is a potential matography (HPLC), Dionex model P680 HPLC pump,
candidate for the treatment for oxidative stress and hyper- ASI100 autosampler, and PDA detector (Dionex Softron
pigmentation[25]. But, further studies will be needed to Gmbh, Germering, Germany) using the Atlantis T3
understand to mechanism of anti-melanogenesis. So we 4.6 9 150 mm, 3 lm column (Water, Milford, MA, USA).
investigated the anti-melanogenic effects of ASE, as well as The monitoring wavelength was set to 280 nm. The mobile
the underlying molecular mechanisms, of ASE. Further- phase was comprised of acidified acetonitrile as solvent A
more, we examined the anti-melanogenesis of ASE having and distilled water as solvent B with 0.1% acetic acid at a
inhibitory effects of hyperpigmentation in vivo. flow rate was 0.7 ml/min using a commercial splitter[26].
ASE was dissolved with 50% ethanol to 20 mg/ml concen-
tration and the injection volume was 10 ll. The gradient
Materials and Methods
program was 0 min, 10% of solvent A; 20 min, 30% of sol-
vent A; 30 min, 50% of solvent A; 40 min, 100% of solvent
Materials
A.
Dulbecco’s modified Eagle’s Medium (DMEM) was pur- The mass spectrometer was an AccuTOF single-reflec-
chased from Welgene (Daegu, Republic of Korea). Fetal tron time-of-flight mass spectrometer equipped with an

504 © 2016 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 68 (2016), 503–513
Ga Yeon Hwang and Se-Young Choung Melanogenesis inhibitory effects of ASE

Electrospray ionization (ESI) source (JEOL, Peabody, MA, mination of protein concentration using a BSA protein
USA) and was operated with Mass Center System version assay kit. A lytic solution was added to the wells of a 96-well
1.3.7b (JEOL)[27]. In the positive ion mode, the typical val- plate, and the absorbance at 405 nm was measured with an
ues were set to the following: orifice 1 = 80 V, ring ELISA reader to evaluate melanin content. Melanin pro-
lens = 10 V, orifice 2 = 5 V, the ion guide potential volt- duction was calculated as the lg of melanin/lg of total pro-
age = 2000 V, and detector voltage = 2300 V[27]. ESI teins in the cell extract.
parameters were set as follows: needle electrode = 2000 V,
nitrogen gas was used to desolvate as a nebulizer, and their
Assay of mushroom tyrosinase activity
flow rate was 1 and 3 l/min, desolvating chamber tempera-
ture = 250°C, and orifice 1 temperature = 80°C[27]. For Many researchers used tyrosinase extracted from the edible
accurate mass measurements and calculations of the ele- Agaricus bisporus due to high homology with the mam-
mental composition, mass scale calibration was accom- malian enzyme that is well-suited as a model for studying
plished with YOKUDELNA calibration kit (JEOL)[27]. MS melanogenesis[29]. In this study, L-DOPA is used as sub-
acquisition was set with a scan range of m/z 100–1500[28]. strate to investigate the effect of inhibitor compounds on
the oxidation of L-DOPA by tyrosinase[30]. Mushroom
tyrosinase solution was prepared by dissolving 1000 units
Cell culture
mushroom tyrosinase in 0.02 M sodium phosphate buffer
The B16F10 melanoma cells (CRL6323) were purchased (pH 6.5). For the mushroom tyrosinase activity assay,
from the American Type Culture Collection (Manassas, 220 ll 0.02 M sodium phosphate buffer (pH 6.5), 20 ll of
VA, USA). The cells were cultured in Dulbecco’s modified various concentrations of ASE or extract solvent and 20 ll
Eagle medium (DMEM) with 10% heat-inactivated fetal 10 mM L-DOPA were added to the wells of a 96-well plate.
bovine serum (FBS, Hyclone) and 1% penicillin-streptomy- Lastly, 40 ll mushroom tyrosinase or 0.02 M sodium phos-
cin. The cells were incubated in a humidified incubator phate buffer was added to the wells. The mixture was then
under 5% CO2 at 37°C. The culture medium was replaced incubated at 37°C for 30 min. Kojic acid was used as a pos-
every 2–3 days. itive control. To evaluate the inhibition of tyrosinase activ-
ity, absorbance was measured at 475 nm with an ELISA
reader. Tyrosinase activity was calculated using the follow-
Cell viability assay
ing equation:
Cell viability was evaluated by performing the 3-(4,5-
dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide Tyrosinase activity ð%Þ ¼ ðA
A1 =B
B1 Þ  100
(MTT) assay. B16F10 cells were seeded at a concentration
of 2 9 104 cells/well on a 24-well plate. The cells were where A and A1 are the absorbances of the samples with
allowed to attach to the plates for 24 h. And then the or without tyrosinase, respectively, and B and B1 are the
B16F10 melanoma cells were treated with various concen- absorbances of the solvent of samples with or without tyr-
trations of ASE (0–300 lg/ml) and incubated for another osinase,respectively.
48 h. The media was removed and replaced with 0.5 mg/
ml MTT in serum-free medium. The cells were incubated
Measurement of cellular tyrosinase activity
for 3–4 h, after which the medium was removed from each
well. The formazan blue in the cells was dissolved by adding B16F10 melanoma cells were seeded on a 6-well plate at a
1 ml DMSO per well. The absorbance was measured at density of 1 9 105 per well. After 24 h incubation, the cells
540 nm using an ELISA reader. were treated with various concentrations (50, 100, 200 lg/
ml) of ASE. Then, the cells were incubated for another 48 h.
The cells were then washed twice with ice-cold PBS. The
Measurement of melanin content in B16F10
cells were lysed by adding 200 ll lysis buffer consisting of
melanoma cells
20 mM Tris–HCI (pH 7.4), 0.32 mM sucrose, 1 mM PMSF,
B16F10 melanoma cells (5 9 104 cells/well) were seeded 0.5 M EDTA (pH 8.0), 1 mM NaF, 1 mM Na3NO4 and pro-
on a 12-well plate. The next day, the cells were treated with tease inhibitor cocktail[31]. The cell pellets were centrifuged
various concentrations of ASE (50, 100 and 200 lg/ml) or for 15 min at 14 930g. The protein concentration in the cell
the positive control kojic acid (400 lM). The cells were then homogenates was quantified using the BCA Protein Assay
incubated for another 48 h. The cells were washed twice Kit. For the cellular tyrosinase assay, 100 ll cell homoge-
with warm PBS and resuspended in 170 ll 1 N NaOH with nates and 20 ll 10 mM L-DOPA was added to the wells of a
10% DMSO. After harvesting, the cell pellets were heated at 96-well plate. The cells were incubated at 37°C for 1 h and
90°C for 1 h. A portion of the lysate was used for the deter- cellular tyrosinase activity was determined by measuring

© 2016 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 68 (2016), pp. 503–513 505
Melanogenesis inhibitory effects of ASE Ga Yeon Hwang and Se-Young Choung

absorbance at 475 nm using an ELISA reader. Cellular

Staining
tyrosinase activity was expressed relative to control (%).
After sacrificing the mice, the ears of the mice were
incubated for 2 h at 37°C in 2 N NaBr. To cleanly
Western blot analysis
remove the hair of the ears, each specimen was soaked
To investigate the signalling pathways affected by ASE, in 0.9% saline for 30 min. The epidermis was separated
B16F10 cells were first treated with various concentrations of from the dermis isolate the epidermis. The epidermis
the extract (50, 100, 200 lg/ml). After treatment, the cells was fixed in 4% formaldehyde containing 0.1 M sodium
were washed with cold-ice PBS and lysed in a lysis buffer on phosphate buffer for 20 min. The epidermal tissues were
ice to obtain protein. All of the cell lysates were kept on ice incubated for 1 h at 37°C in 0.1% L-DOPA containing
for 30 min and centrifuged 14 930g for 15 min at 4°C. Equal 0.1 M sodium phosphate buffer. In addition, epidermal
amounts of protein (30 lg) per lane were separated on a tissue was incubated 0.1% L-DOPA in 0.1 M sodium
SDS-polyacrylamide gel (8%) by electrophoresis and trans- phosphate buffer for 2 h at 37°C. DOPA-positive epider-
ferred onto polyvinylidene fluoride (PVDF) membranes. The mal melanocytes were stained, and the tissues were
membranes were then incubated in a blocking solution of mounted with mounting medium.
5% dried milk in Tris-buffered saline containing 0.1% Tween The other ears of the mice were fixed in buffered 4% for-
20 for 1 h at room temperature. The membranes were then malin, sectioned at 4 lm, and stained using the Fontana–
incubated overnight at 4°C with the appropriate primary Masson technique for histological observation.
antibodies at a dilution of 1 : 1000. The membranes were
washed with Tris-buffered saline containing 0.1% Tween 20
Statistical analysis
for 1 h. After adding the appropriate horseradish peroxidase
conjugated secondary antibody (1 : 5000), the membranes All of the values obtained from at least three independent
were incubated for 2 h and then washed with 1 9 TBST for experiments were used to determine the mean  standard
1 h. Bound antibodies were detected using Enhanced Chemi- error. Statistical analysis was carried out using the Kruskal–
luminescence Plus kits and a detection system (Fuji Film, Wallis test with the Mann–Whitney U post-hoc test using
Tokyo, Japan). b-actin was used as an internal control. the Statistical Packages for the Social Sciences (SPSS) soft-
ware (version 21.0; SPSS Inc., Chicago, IL, USA). Signifi-
cant significance for effects in the each figures given as
Animals
follows: *P < 0.05, **P < 0.01 and ***P < 0.001.
Seven-week-old C57bl/6J (20–22 g) mice were purchased
from RaonBio Inc. (KyungKeeDo, YoungIn, Korea). All of
Results
the animals were housed under controlled conditions of
23  3°C (55  5% humidity) and a 12-h light/dark cycle.
Identification of ASE by HPLC-ESI-MS
All of the animals had access to water and commercial food
for the duration of the experimental period. To induce pig- An HPLC chromatogram of ASE recorded at 254 nm[32]
mentation, the negative control, positive control and treat- and we conducted HPLC-ESI-MS analysis to verify them
ment groups were exposed to UVB irradiation (wave length more exactly. The molecular ions [M+H]+ and the major
312 nm) at a dose of 100 mJ/cm2 for 10 days; oral admin- fragments ions of the four peaks are illustrated in Figure 1.
istration of ASE was performed daily for 9 weeks. As a result, one flavonoid and three hydroxyl cinnamoly
We divided the mice into six groups; a normal group quinic acids were identified, respectively. The ESI-MS spec-
(not irradiated), a control group (UVB-irradiated group), a tra of peak 1 showed the protonated molecule [M+H]+ at
positive control group and three ASE treatment (35, 70, m/z 355 and fragments ions at m/z 163 and m/z 191 was
140 mg/kg) groups. ASE was orally administered by gastric proposed as chlorogenic acid. The peak 2 showed the pro-
intubation using an animal-feeding needle daily for 9 weeks. tonated molecule [M+H]+ at m/z 465 and aglycone ion at
The normal and control groups were orally administered m/z 303, which identified as isoquercitrin, respectively. The
0.5% carboxymethyl cellulose (CMC). The positive control protonated molecule [M+H]+ at m/z 517 and fragments
group was administered ascorbic acid (70 mg/kg) saturated ions at m/z 499 were identical analysed to peak 3 and 4
in 0.5% CMC. At 3, 6 and 9 weeks, all mice were sacrificed which were proposed as 3,4-di-O-caffeoylquinic acid and
to evaluate the anti-pigmentation effect of ASE, and the hair 3,5-di-O-caffeoylquinic acid by the elution sequence,
of the irradiated ears was removed using hair removal respectively. The characters of these phytochemicals have
creams. All of the mice were treated according to protocols been formerly identified in arnica[28] and chrysanthemum
approved by the Institutional Animal Care and Use Com- flower[32]and these plants belong in the family of
mittees (Approval number. KHP–2014–11-2). Asteraceae.

506 © 2016 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 68 (2016), 503–513
Ga Yeon Hwang and Se-Young Choung Melanogenesis inhibitory effects of ASE

Figure 1 (a) High-performance liquid chromatography chromatogram of the AS extract (b) positive ion ESI-MS spectrum of the four peaks:
1. Chlorogenic acid; 2. Isoquercitrin; 3. 3,4-di-O-caffeoylquinic acid; 4. 3,5-di-O-caffeoylquinic acid.

compared at similar concentrations. We also found that

The cytotoxicity and melanogenic effects of
ASE inhibited cellular tyrosinase activity (Figure 2d). The
ASE in B16F10 melanoma cells
relative levels of cellular tyrosinase after treatment with
We evaluated whether various concentrations of ASE (0, ASE were 87%, 84% and 72% at extract concentrations of
12.5, 25, 50, 100, 200 and 300 lg/ml) is cytotoxic to B16F10 50, 100 and 200 lg/ml, respectively. Taken together, these
melanoma cells. After 24 h incubation, the cells were treated results indicate that ASE is a potent inhibitor of the
with ASE and incubated for an additional 48 h. Cytotoxicity melanin synthesis cascade and acts by blocking tyrosinase
in B16F10 melanoma cell was measured using the MTT activity.
assay. The ASE had no apparent cytotoxicity at concentra-
tions up to 200 lg/ml. However, cell viability decreased by
The effects of ASE on the expression levels
72% at the highest concentration (Figure 2a). Therefore, for
of melanin synthesis-related proteins and
subsequent melanin synthesis experiments, the cells were
signalling pathway components
treated with sub-cytotoxic concentrations of the extract (50,
100 and 200 lg/ml). Kojic acid was used as a positive con- To determine whether ASE has anti-melanogenic effects,
trol. Dose-dependent decreases in the melanin content were we examined the protein expression levels of melanin syn-
observed. Specifically, treatment with 200 lg/ml ASE thesis-related proteins, including tyrosinase, TRP1, TRP2
resulted in a 30% reduction in melanin content in melano- and MITF. To this end, cell lysates from cells treated with
cytes and was not significantly toxic to the cells. These various concentrations of extract (50, 100 and 200 lg/ml)
results showed that ASE has anti-pigmentation effects in were analysed using Western blot analysis. The results indi-
B16F10 melanoma cells (Figure 2b). cate that ASE downregulated the expression of tyrosinase,
TRP1 and MITF; however, the expression of TRP2 was not
significantly altered (Figure 3a).
Inhibitory effects of ASE on tyrosinase
activity
The effects of cotreatment with MAPK/ERK
To investigate whether ASE affects tyrosinase activity, we
or Akt/GSK-3b inhibitors and ASE on melanin
measured cell-free mushroom tyrosinase and cellular
synthesis
tyrosinase activity. The results in Figure 2c show that
mushroom tyrosinase activity was dose-dependently The results shown in Figure 3 suggest that MAPK/ERK and
decreased by ASE treatment. Interestingly, the inhibitory Akt/GSK3b signalling may mediate the hypopigmentation
effect of ASE was greater than that observed for kojic acid effects of ASE. To address this possibility, we tested the

© 2016 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 68 (2016), pp. 503–513 507
Melanogenesis inhibitory effects of ASE Ga Yeon Hwang and Se-Young Choung

Figure 2 The effects of melanogenesis in B16F10 cells following treatment with various concentrations of AS extract (ASE). The cells were trea-
ted with ASE for 48 h. The effect of ASE on cell viability in B16 F10 melanoma cells (a). Cellular melanin content following treatment with ASE
was evaluated as a percentage of melanin content compared with control cells (b). Mushroom tyrosinase activity (c). Cellular tyrosinase activity (d).
The statistical results are expressed as means  SD of at least three independent experiments (*P < 0.05, **P < 0.01 and ***P < 0.001).

effects of cotreatment with ASE and PD98059 (a MAPK/ C57BL/6J mice were subjected to a UVB irradiation-
ERK-specific inhibitor) and LY294002 (a Akt/GSK3b-speci- induced pigmentation protocol for 3, 6 and 9 weeks
fic). The results demonstrated that melanoma cells treated (6 weeks data was not shown). In this study, many active
with PD98059 and LY294002, as well as cells overexpressing melanocytes were visualized using DOPA staining. After
dominant negative mutants, exhibit increased melanin syn- the UVB irradiation was stopped, both the number of and
thesis due to upregulated tyrosinase expression. However, size of the melanocytes was increased. Moreover, after UVB
ASE suppressed this increase in melanin levels and tyrosi- irradiation for 10 days staining, the number of melanin
nase activity. It was observed that ASE increased levels of granules was increased fivefold (data not shown), as deter-
both phosphorylated MAPK/ERK and phosphorylated Akt/ mined by DOPA staining (Figure 5aI).
GSK 3b in a dose-dependent manner. Based on these data, Melanin pigmentation on the ears was dose-dependently
we conclude that ASE exerts its anti-melanogenic activity decreased after oral administration with ASE for 3, 6 and
by activating the ERK cascade, leading to the increased 9 weeks. The number and size of the melanocytes were
degradation of MITF. Thus, the effects of ASE are closely reduced in the 35 mg/kg treatment group to a lesser extent
connected to the known mechanisms of melanogenesis than for the other groups but reduced melanin pigmentation
suppression. Additionally, it is concluded that ASE inhibits and melanocytes size were still observed compared with the
melanogenesis by activating both the MAPK/ERK and Akt/ UVB irradiation group (Figure 5bI, II at control and experi-
GSK3b signalling pathways (Figure 4). mental I group). The hypopigmentation effect of 70 mg/kg
ASE was similar to the effect observed with ascorbic acid;
and the concentration of 140 mg/kg had an even stronger
Decreased melanocyte activity and
effect on melanin pigmentation and melanocyte size. In this
distribution of melanin granules with the
study, a significant decrease compared to the UVB-exposed
oral application ASE, as determined by
group at 9 weeks of 140 mg/kg treatment was not observed
DOPA and Fontana–Masson staining
in the dermis but in all layers of the epidermis. It is because
In addition to in vitro experiments, we also observed the that melanin is synthesized in specific cells within melano-
whitening effects of ASE using an in vivo model. Specifically somes, which are membrane-bound organelles.

508 © 2016 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 68 (2016), 503–513
Ga Yeon Hwang and Se-Young Choung Melanogenesis inhibitory effects of ASE

Figure 3 The effects of AS extract (ASE) on the expression of proteins related to melanogenesis in B16F10 cells. The cells were treated with
non-cytotoxic concentrations of ASE (below 300 lg/ml) for 48 h. Total cell lysates were used for Western blotting to determine expression levels
of tyrosinase, TRP-1, DCT and MITF. The results are relative to b-actin, which was used as the loading control (a). Western blotting showed that
treatment with different concentrations of ASE changed the protein expression levels of MEK, p-MEK, ERK, p-ERK, p38, pp38, Akt, p-Akt, GSK-3b
and p-GSK-3b (b). The statistical results are expressed as means  SD of at least three independent experiments (*P < 0.05, **P < 0.01 and
***P < 0.001).

We found melanin granules in all layers of the epidermis acid and both the 35 and 70 mg/kg ASE groups, showed
using Fontana-Masson staining, and the distribution of significantly decreased epidermal thicknesses compared to
melanin granules increased after UVB irradiation in all lay- the UVB-exposed group at 9 weeks.
ers of the epidermis (Figure 5a, bII). However, the extent
of the UVB irradiation-induced increase in melanin distri-
Discussion
bution was significantly decreased in the groups treated
with ASE. This UVB irradiation-induced increase in mela- Melanogenesis can be regulated at multiple points within
nin distribution was irregularly reduced in portions of the its synthesis pathway, such as by the inhibition of tyrosinase
epidermis in the ascorbic acid positive control groups fol- expression or the direct inactivation of tyrosinase the
lowing 3, 6 and 9 weeks of treatment. Reduced melanin enzyme[33]. During melanin synthesis, MITF plays a critical
distribution was observed in the 70 mg/kg ascorbic acid role as a transcriptional activator of genes encoding
treatment group; a significant effect was observed in the melanogenesis-related proteins, such as tyrosinase and
epidermis with an ASE concentration of 140 mg/kg. TRP-1 in mammalian cells[9]. MITF-M cannot transactivate
the TRP-2 promoter; therefore, tyrosinase and TRP-1 are
expressed before TRP2 during embryogenesis[34] and it is
Observation of decreased epidermal
assumed that the mechanisms related to expression of
thickness
TRP2 are fundamentally different from those regulating
To examine changes of the epidermal thickness following TRP1 expression[35]. In this study, ASE mechanistically
treatment with ASE, other tissues were stained with H&E. appears to downregulate the expression of tyrosinase, TRP-
The thickness of the epidermis in the control group (UVB- 1 and not much for TRP-2.
exposed group) was remarkably increased compared to the In particular, various pathways were involved in the
normal group (non-UVB-exposed) after UVB irradiation melanogenesis signalling cascade. Previous studies have
for 10 days (Figure 5a). After 9 weeks of treatment, the found that the activation of the MAPK/ERK and Akt/
epidermis in the 140 mg/kg ASE treatment group was GSK3b signalling pathways can regulate melanin produc-
much thinner compared with the UVB-exposed group. The tion through the downregulation of MITF activity[36].
other treatment groups, including the 70 mg/kg ascorbic Accordingly, we investigated the levels of phosphorylated

© 2016 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 68 (2016), pp. 503–513 509
Melanogenesis inhibitory effects of ASE Ga Yeon Hwang and Se-Young Choung

Figure 4 The effects of cotreatment with AS extract (ASE) and specific inhibitors. B16F10 melanoma cells were pretreated for 1 h with or with-
out 20 lM PD98059 (a MEK/ERK inhibitor) or LY294002 (an Akt/GSK3b inhibitor) and treated with ASE. The cells were further incubated for 48 h,
after which the melanin content and cellular tyrosinase activity were investigated. The statistical results are expressed as means  SD of at least
three independent experiments (*P < 0.05, **P < 0.01 and ***P < 0.001: ASE or each inhibitor treatment compared to untreated blank;
###P < 0.001: ASE cotreated with each inhibitor compared to treatments with each inhibitor alone).

MAPK/ERK, Akt, and GSK3b by Western blotting to In previous studies, it have found that p38 MAPK is a
strengthen our understanding of the anti-melanogenic major intracellular signalling molecule that is essential for
effects of ASE. We observed that the levels of phosphory- pigmentation. The activation of p38 MAPK signalling
lated MEK were increased by ASE, which in turn increased increases MITF expression, leading to increased expression
the levels of phosphorylated ERK. The activation of ERK of tyrosinase and melanin synthesis[40]. Therefore, the p38
results in the phosphorylation of MITF at S73 and the MAPK inhibition can be considered to downregulate
recruitment of the transcriptional coactivator, p300. This melanogenesis. However, in the present study, no signifi-
process induces the ubiquitin-dependent proteasomal cant increase was observed in the levels of phosphorylated
degradation of MITF[37]. Additionally, it is reported that (inactive) p38 MAPK. We conclude that the P38 MAPK
Akt and GSK3b, which are under downstream effectors of signalling pathway does not significantly participate in
the PI3K-mediated signalling pathway, regulate melanogen- ASE-mediated depigmentation. Accordingly, it is con-
esis in G361 and B16 melanoma cells[15,38]. PI3K phospho- cluded that the hypopigmentation effects of ASE are pri-
rylates Akt, and activated Akt then phosphorylates GSK3b marily mediated by MAPK/ERK and Akt/GSK3b
on Ser9, leading to GSK3b inactivation. Unphosphorylated phosphorylation.
GSK3b phosphorylates MITF on Ser298, increasing MITF- Besides in vitro experiment, we also observed the whiten-
M localization to M-box in the tyrosinase promoter and ing effects of ASE on hyperpigmentation in C57bL/6J mice
leading to the induction of melanogenesis[39]. Therefore, induced by 3, 6 and 9 weeks through UV irradiation. It is
ASE activates Akt/GSK3b signalling, leading to reduced reported that the number of melanocytes was observed to
levels of MITF Ser298 phosphorylation. This effect results increase 2 days after approximately 14 days of repeated
in decreased binding of MITF-M to the M-box and reduced UVB irradiation exposure. When the UV irradiation was
expression of tyrosinase and TRP-1[35]. stopped, the number of melanocytes on the ears increased

510 © 2016 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 68 (2016), 503–513
Ga Yeon Hwang and Se-Young Choung Melanogenesis inhibitory effects of ASE

Figure 5 Histological observations of C57bl/6J mouse tissues following the hyperpigmentation protocol and administration with AS extract
(ASE). Scale bar 200 lm I: DOPA-stained images for the examination of melanocyte activity, 9100. II: Fontana–Masson stain images of skin sec-
tions for the examination of melanin granule distribution, 9400. III: H&E stain images of skin sections for the examination of the thickness of the
epidermis. 9400. (a): Images after oral administration for 3 weeks (b): Images after oral administration for 9 weeks. The mice were divided into 6
groups: N (CMC 0.5% treated normal group); C (UVB-irradiated control group); experiment 1 (ASE, 35 mg/kg); experiment 2 (ASE 70 mg/kg);
experiment 3 (ASE, 140 mg/kg).

© 2016 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 68 (2016), pp. 503–513 511
Melanogenesis inhibitory effects of ASE Ga Yeon Hwang and Se-Young Choung

approximately fourfold compared to before irradiation,

and the melanocytes were spread over the epidermis. The Conclusions
induced pigmentation was maintained for 20 weeks[41]. In this study, the results show a clear in vivo hypopigmenta-
The UV-induced increase in melanocyte number indicates tion effect of ASE and confirm the in vitro results through
that the melanocytes underwent mitosis and did not simply the regulation of MAPK/ERK and AKT/GSK3b signalling.
become more active[42]. Furthermore, by Fontana–Masson Consistent with these in vitro and in vivo data, ASE could
staining, it is observed that the pigment foci in the melano- be useful as a melanogenesis inhibitor in the field of cos-
cytes were not decreased by ASE treatment in the dermis metics. From this perspective, we believe that it deserves
but in the epidermis. Melanocytes in the epidermis produce further development as novel skin-whitening agent.
melanin upon stimulation with growth factors, which are
released from keratinocytes following UV irradiation[43].
The ethanol extract of AS contains chlorogenic acid, iso- Declarations
quercitrin, luteolin-7-O-rutinoside, 3,4-O-caffeoylquinic
acid and 3,5,-O-caffeolylquinic acid (Figure 1). These phy- Conflict of interest
tochemicals have inhibitory effects of melanogenesis[3,44] The Author(s) declare(s) that they have no conflicts of
and related show anti-oxidant[45,46] and inflammatory interest to disclose.
effects. Lastly, chlorogenic acid has been investigated to
have effects of inhibitory melanogenesis caused by regulat-
ing activity of tyrosinase[44]. Consistent with these findings, Funding
it is observed that ASE inhibits activity of tyrosianse. In This research received no specific grant from any funding
addition to, methyl 3,5-dicaffeoyl quinate lead activation of agency in the public, commercial, or not-for-profit sectors.
phospho-ERK and inhibition of phospho-p38 MAPK,
which plays important role of melanin synthesis[3]. Consis-
tent with these studies, our results support that ASE mainly Acknowledgements
inhibits melanogenesis by regulating ERK, Akt and GSK3b We thank Myung-Ah Kang, Lee-Bom Lee and Hyang-Yu
levels. Therefore, we assume that the inhibitory effects of Kim (Kyung Hee University, College of Pharmacy) for the
melanogenesis by ASE may be synergistic interactions of help with taking care of the mice.
ASE contained various phytochemicals. Therefore, ASE is
valued for further development as complementary agent in
the near future.

References 5. Naish-Byfield S, Riley PA. Tyrosinase pigmentation. Pigment Cell Res 2000;
kinetics: failure of acceleration in oxi- 13: 60–69.
1. Costin GE, Hearing VJ. Human skin dation of ring-blocked monohydric 11. Han E et al. Melanogenesis inhibi-
pigmentation: melanocytes modulate phenol substrate. Pigment Cell Res tory effect of aerial part of Pueraria
skin color in response to stress. 1998; 11: 94–97. thunbergiana in vitro and in vivo.
FASEB J 2007; 21: 976–994. 6. Chang TS. An updated review of Arch Dermatol Res 2015; 307: 57–
2. Briganti S et al. Chemical and instru- tyrosinase inhibitors. Int J Mol Sci 72.
mental approaches to treat hyperpig- 2009; 10: 2440–2475. 12. Wu LC et al. Antimelanogenic effect
mentation. Pigment Cell Res 2003; 16: 7. Park HY et al. MITF mediates cAMP- of c-phycocyanin through modulation
101–110. induced protein kinase C-beta expres- of tyrosinase expression by upregula-
3. Shen T et al. Involvement of the p38 sion in human melanocytes. Biochem tion of ERK and downregulation of
MAPK and ERK signaling pathway in J 2006; 395: 571–578. p38 MAPK signaling pathways. J
the anti-melanogenic effect of methyl 8. Tachibana M. MITF: a stream flowing Biomed Sci 2011; 18: 74.
3,5-dicaffeoyl quinate in B16F10 for pigment cells. Pigment Cell Res 13. Wu M et al. c-Kit triggers dual phos-
mouse melanoma cells. Chem Biol 2000; 13: 230–240. phorylations, which couple activation
Interact 2012; 199: 106–111. 9. Steingrimsson E et al. Melanocytes and degradation of the essential mela-
4. Halaban R et al. Abnormal acidifica- and the microphthalmia transcription nocyte factor Mi. Genes Dev 2000; 14:
tion of melanoma cells induces tyrosi- factor network. Annu Rev Genet 2004; 301–312.
nase retention in the early secretory 38: 365–411. 14. Kono M et al. Role of the mitogen-
pathway. J Biol Chem 2002; 277: 10. Busca R, Ballotti R. Cyclic AMP a key activated protein kinase signaling
14821–14828. messenger in the regulation of skin pathway in the regulation of human

512 © 2016 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 68 (2016), 503–513
Ga Yeon Hwang and Se-Young Choung Melanogenesis inhibitory effects of ASE

melanocytic antigen expression. Mol ate: involvement of oxidative stress thesis and tyrosinase activity through
Cancer Res 2006; 4: 779–792. and the JNK and ERK signaling path- activating the MEK/ERK and AKT/
15. Oka M et al. Regulation of melano- ways in melanoma cells. PLoS ONE GSK3beta signaling pathways. J Agri
genesis through phosphatidylinositol 2014; 9: e95186. Food Chem 2012; 60: 4859–4864.
3-kinase-Akt pathway in human G361 26. Lee E et al. Lilium lancifolium Thunb. 36. Kim DS et al. Transforming growth
melanoma cells. J Invest Dermatol extract attenuates pulmonary inflam- factor-beta1 decreases melanin synthe-
2000; 115: 699–703. mation and air space enlargement in a sis via delayed extracellular signal-
16. Nakajima M et al. Arbutin increases the cigarette smoke-exposed mouse model. regulated kinase activation. Int J Bio-
pigmentation of cultured human mela- J Ethnopharmacol 2013; 149: 148–156. chem Cell Biol 2004; 36: 1482–1491.
nocytes through mechanisms other 27. Lee H et al. Galla Chinensis attenuates 37. Hemesath TJ et al. MAP kinase links
than the induction of tyrosinase activ- cigarette smoke-associated lung injury the transcription factor Microph-
ity. Pigment Cell Res 1998; 11: 12–17. by inhibiting recruitment of inflam- thalmia to c-Kit signalling in melano-
17. Takizawa T et al. Enhancement of matory cells into the lung. Basic Clin cytes. Nature 1998; 391: 298–301.
hepatocarcinogenesis by kojic acid in Pharmacol Toxicol 2015; 116: 222–228. 38. Busca R et al. Inhibition of the phos-
rat two-stage models after initiation 28. Lin LZ, Harnly JM. Identification of phatidylinositol 3-kinase/p70(S6)-
with N-bis(2-hydroxypropyl)nitrosa- hydroxycinnamoylquinic acids of kinase pathway induces B16 mela-
mine or N-diethylnitrosamine. Toxicol arnica flowers and burdock roots noma cell differentiation. J Biol Chem
Sci 2004; 81: 43–49. using a standardized LC-DAD-ESI/MS 1996; 271: 31824–31830.
18. Yen FL et al. Melanogenesis inhibitor profiling method. J Agri Food Chem 39. Khaled M et al. Glycogen synthase
(s) from Phyla nodiflora extract. Evid 2008; 56: 10105–10114. kinase 3beta is activated by cAMP and
Based Complement Alternat Med 2012; 29. Song KK et al. Inhibitory effects of 4- plays an active role in the regulation
2012: 867494. vinylbenzaldehyde and 4-vinylbenzoic of melanogenesis. J Biol Chem 2002;
19. Lee SO et al. Cytotoxic terpene acid on the activity of mushroom 277: 33690–33697.
hydroperoxides from the aerial parts tyrosinase. J Enzyme Inhib Med Chem 40. Saha B et al. Activation of the Mitf
of Aster spathulifolius. Arch Pharm Res 2005; 20: 239–243. promoter by lipid-stimulated activa-
2006; 29: 845–848. 30. Gerdemann C et al. The crystal struc- tion of p38-stress signalling to CREB.
20. Lee JH et al. Camellia japonica sup- ture of catechol oxidase: new insight Pigment Cell Res 2006; 19: 595–605.
presses immunoglobulin E-mediated into the function of type-3 copper pro- 41. Rosdahl IK. Melanocyte mitosis in
allergic response by the inhibition of teins. Acc Chem Res 2002; 35: 183–191. UVB-irradiated mouse skin. Acta
Syk kinase activation in mast cells. 31. Bang CY, Choung SY. Enzogenol Derm Venereol 1978; 58: 217–221.
Clin Exp Allergy 2008; 38: 794–804. improves diabetes-related metabolic 42. Seiler N, Knodgen B. Effects of ultra-
21. Won JN et al. Antiviral activity of the change in C57BL/KsJ-db/db mice, a violet light on epidermal polyamine
plant extracts from Thuja orientalis, model of type 2 diabetes mellitus. J metabolism. Biochem Med 1979; 21:
Aster spathulifolius, and Pinus thun- Pharm Pharmacol 2014; 66: 875–885. 168–181.
bergii against influenza virus A/PR/8/ 32. Lin L-Z, Harnly JM. Identification of 43. Yamada T et al. Wnt/beta-catenin and
34. J Microbiol Biotechnol 2013; 23: the phenolic components of chrysan- kit signaling sequentially regulate mel-
125–130. themum flower (Chrysanthemum anocyte stem cell differentiation in
22. Kim SJ, Choung SY. Inhibitory effects morifolium Ramat). Food Chem 2010; UVB-induced epidermal pigmenta-
of Aster spathulifolius extract on adi- 120: 319–326. tion. J Invest Dermatol 2013; 133:
pogenesis and lipid accumulation in 33. Ando H et al. Correlation between the 2753–2762.
3T3-L1 preadipocytes. J Pharm Phar- number of melanosomes, tyrosinase 44. Li HR et al. Effect of chlorogenic acid
macol 2016; 68: 107–118. mRNA levels, and tyrosinase activity on melanogenesis of B16 melanoma
23. Ferreira JF et al. Flavonoids from in cultured murine melanoma cells in cells. Molecules 2014; 19: 12940–12948.
Artemisia annua L. as antioxidants response to various melanogenesis 45. Xu JG et al. Antioxidant and DNA-
and their potential synergism with regulatory agents. J Cell Physiol 1995; protective activities of chlorogenic
artemisinin against malaria and can- 163: 608–614. acid isomers. J Agri Food Chem 2012;
cer. Molecules 2010; 15: 3135–3170. 34. Yasumoto K et al. Functional analysis 60: 11625–11630.
24. Mukherjee AK et al. Advances in can- of microphthalmia-associated tran- 46. Konczol A et al. Antioxidant activity-
cer therapy with plant based natural scription factor in pigment cell-speci- guided phytochemical investigation of
products. Curr Med Chem 2001; 8: fic transcription of the human Artemisia gmelinii Webb. ex Stechm.:
1467–1486. tyrosinase family genes. J Biol Chem isolation and spectroscopic challenges
25. Peng HY et al. The melanogenesis 1997; 272: 503–509. of 3,5-O-dicaffeoyl (epi?) quinic acid
alteration effects of Achillea mille- 35. Lin YS et al. Nicotinic acid hydroxa- and its ethyl ester. J Pharm Biomed
folium L. essential oil and linalyl acet- mate downregulated the melanin syn- Anal 2012; 59: 83–89.

© 2016 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 68 (2016), pp. 503–513 513