Discovery of small-molecule inhibitors of tyrosinase

Article abstract of DOI:10.1016/j.bmcl.2007.10.014

To identify novel inhibitors of tyrosinase, a fluorescent assay was developed which is suitable for high-throughput screening. In the assay, oxidation of the substrate by tyrosinase leads to the release of a fluorescent coumarin. Several small molecules w

Full text of DOI:10.1016/j.bmcl.2007.10.014

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 17 (2007) 6871–6875
Discovery of small-molecule inhibitors of tyrosinase
Juris P. Germanas,a,*, Shugauang Wang,b Andrew Miner,a
Wayne Haob and Joseph M. Readyc,*
aDepartment of Dermatology, University of Texas Southwestern Medical Center, Dallas, TX 75390-9038, USA
bHigh Throughput Screening Facility, University of Texas Southwestern Medical Center, Dallas, TX 75390-9038, USA
cDepartment of Biochemistry, University of Texas Southwestern Medical Center, Dallas, TX 75390-9038, USA
Received 10 August 2007; revised 28 September 2007; accepted 5 October 2007
Available online 12 October 2007
Abstract—To identify novel inhibitors of tyrosinase, a fluorescent assay was developed which is suitable for high-throughput screen-
ing. In the assay, oxidation of the substrate by tyrosinase leads to the release of a fluorescent coumarin. Several small molecules were
identified that inhibited mushroom tyrosinase in vitro and human tyrosinase in cell culture. These compounds may represent lead
structures for therapies targeted at disorders of hyperpigmentation.
ꢀ 2007 Elsevier Ltd. All rights reserved.
CO2H
CO2H
Tyrosinase
The enzyme tyrosinase catalyzes the aerobic hydroxyl-
ation of L-tyrosine to L-dopa and the subsequent oxida-
tion of L-dopa to L-dopaquinone (Eq. 1).1 This sequence
marks the initiation of melanin biosynthesis. For this
reason, inhibitors of tyrosinase hold promise as treat-
ments for disorders associated with overproduction of
melanin, including cutaneous hyperpigmentation2 and
ocular retinitis pigmentosa.3 Indeed, tyrosinase inhibi-
tors are found in current treatments for melasma,
postinflammatory hyperpigmentation, solar lentigo and
Addison’s disease.4 However, currently available tyrosi-
nase inhibitors suffer from toxicity and/or a lack of effi-
cacy. For example, the most popular current
medications include hydroquinone as a tyrosinase inhib-
itor.5 Hydroquinone is a substrate analogue of tyrosine,
and, like tyrosine, can undergo oxidation. The product
of hydroquinone oxidation, benzoquinone, is a known
mutagen6 and can damage DNA.7 It is cytotoxic to
hepatocytes and melanocytes8 and may be carcino-
genic.9 For these reasons, the use of hydroquinone in
cosmetic products has been banned in the European Un-
ion10 and is under increasing scrutiny by the United
States Food and Drug Administration.11
Tyrosinase
O2
NH2
NH2
HO
O
HO
OH
L-dopa
ð1Þ
CO2H
NH2
Melanin
O
L-dopaquinone
Alternatives to hydroquinone include kojic acid12 and
arbutin.13 Unfortunately neither substance offers a safe
and effective replacement. In fact concerns regarding
the carcinogenicity of kojic acid spurred Japanese offi-
cials to ban its use in skin treatments recently,14 while
arbutin is simply a hydroquinone glycoside.
Aware of the health and efficacy concerns related to
kojic acid and hydroquinone-derived inhibitors, we con-
sidered possible approaches to discover alternative treat-
ments for hyperpigmentation. To facilitate this search,
we first focused our attention on the development of
an assay suitable for high-throughput screening
(HTS).15 Known in vitro16 or cell-based assays17 can
evaluate dozens of compounds, but present problems
in the context of large screening campaigns.
Keywords: High-throughput screening; Fluorescence; Tyrosinase inhi-
bitors; Pigmentation; Thiazoles.
To develop an assay of tyrosinase activity suitable for
HTS, we planned to take advantage of the promiscuity
of tyrosinase toward phenols and catechols.18 As a gen-
eral design feature, we sought a fluorogenic substrate
*
Corresponding authors. Tel.: +1 214 648 0313; fax: +1 214 648 0320
(J.M.R.); e-mail: joseph.ready@utsouthwestern.edu
Present address: VA Maryland Health Care System, Department of
Managed Care, 10 N Green St., Baltimore, MD 21201, USA.
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.10.014

6872
J. P. Germanas et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6871–6875
that, following tyrosinase-mediated oxidation, would re-
lease a detectable fluorophore. This process could be
monitored spectroscopically, and small molecules that
minimized fluorophore release could be considered po-
tential enzyme inhibitors (Scheme 1). Inspired by a
pro-drug strategy developed in the Osborn group,19,20
we prepared several acylated amino coumarins as fluo-
rogenic substrates for tyrosinase. Candidates were eval-
uated in terms of spectral characteristics, performance in
enzyme assays, solubility, and ease of synthesis. From
these initial studies urea PAP-AMC emerged as the
most promising candidate for a reliable fluorescent as-
say. It is synthesized in a one-pot, two step sequence
as shown in Eq. 2.
ence of 0.2 mM kojic acid, a known tyrosinase inhibitor,
the fluorescence increase was markedly attenuated
(Fig. 1a).
Two studies of the kinetics of the reaction support the
contention that the time-dependent fluorescence in-
crease results from a tyrosinase-catalyzed oxidation of
the fluorogenic substrate. First, the rate of fluorescence
increase is linearly related to enzyme concentration
(Fig. 1b). The reproducibility of the assay is also illus-
trated in Figure 1b. Second, the reaction shows satura-
tion kinetics with respect to substrate PAP-AMC
(Fig. 1c). Finally, free 7-amino-4-methyl-coumarin was
observed by HPLC/MS analysis of the crude reaction
mixtures. Taken together, the evidence shows that tyros-
inase-catalyzed oxidation of PAP-AMC results in an in-
crease in fluorescence signal, and that this signal is
diminished in the presence of enzyme inhibitors.
Accordingly, the assay conditions were optimized for a
format involving robotic liquid handling and data
acquisition in a 384-well plate format.
OH
CH3
Cl2CO, EtOAc;
CH3
HO
O
ð2Þ
ð3Þ
N
H
N
H
O
O
NH2
PAP-AMC
H2N
O
O
66% yield
AMC
CH3
CH3
HO
tyrosinase
air
O
Using the fluorescence assay described above, an in-
house library of compounds was screened at approxi-
mately 5 lM. By defining a potential hit as a well read-
ing at least three standard deviations below the global
mean, the assay yielded a hit rate of approximately
0.6% (1181 initial hits from 200,000 compounds). Initial
hits were reevaluated at approximately 15, 5.0, and
1.7 lM. Thus we identified 421 small molecules showing
a reproducible inhibition of mushroom tyrosinase. As a
testament to the efficiency of the fluorescence assay, the
entire library was screened in less than one month using
less than 10 mg of substrate PAP-AMC.
N
H
N
H
O
O
H2N
O
O
PAP-AMC
AMC
High Fluorescence
Low Fluroescence
As anticipated, substantial differences in fluorescence
are observed between the free coumarin and its acylated
counterpart (Fig. SI1). Thus, if oxidation with tyrosi-
nase promoted release of free coumarin, this transfor-
mation should be accompanied by a large increase in
fluorescence (Eq. 3).21 Indeed, monitoring a reaction
mixture containing fluorogenic substrate PAP-AMC
and mushroom tyrosinase revealed a steady increase in
fluorescence as a function of time. Critically, in the pres-
In the next stage of analysis, the 421 candidates were
evaluated for toxicity and their ability to decrease pig-
mentation in a human melanocyte cell line. Accordingly,
96-well plates were inoculated with 10,000 MNT-1
cells22 and treated with compound (15 lM, in duplicate),
DMSO (negative control), or phenylthiourea (250 lM,
positive control). After incubation for 4 days at 37 ꢁC,
the wells were treated with CellTiter-Glo reagent and
the absorbance of the mixture at 410 nm was deter-
mined. Absorbance at this wavelength was found to be
proportional to melanin concentration. In addition,
Inhibitor
O
Tyrosinase
Fluorophore
Fluorophore
HO
O
free fluorophore
high fluorescence
fluorogenic tyrosine analogue
low fluorescence
oxidized substrate
Scheme 1. Strategy for fluorogenic reporter of tyrosinase activity.
a
b 500
c 1.5
[enz]
1000
800
600
400
200
0
(μg/mL)
400
300
200
100
0
72
1.0
36
V
0.5
18
9
+inhibitor
4.5
0.0
0
5
10
15
20
0
1000 2000 3000 4000 5000 6000
Time (s)
0
250
500
750 1000 1250
Time (s)
[PAP-AMC] μM
Figure 1. Fluorescence monitoring of the mushroom tyrosinase-catalyzed oxidation of PAP-AMC. (a) Oxidation of PAP-AMC in the presence or
absence of 0.2 M kojic acid. (b) Reactions in the presence of varying concentrations of tyrosinase. Duplicate experiments are shown. Inset: plot of
initial rate versus [tyrosinase]. (c) Initial rates as a function of [PAP-AMC]. Inset: Lineweaver–Burk plot. For all experiments, kexcitation = 350 nm;
kemission = 460 nm.

J. P. Germanas et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6871–6875
6873
the luminescence of each well was measured to deter-
mine cytotoxicity.
(entry 5). Although no currently available treatment for
hyperpigmentation includes these types of inhibitors be-
cause of toxicity concerns,24 their identification validates
the fluorescence assay. Second, the structures shown in
Table 1 were not isolated hits; rather, multiple members
of each class were identified.
Representative results from the in vitro and cell-based
assays are shown in Table 1. A number of features of
the data warrant comment. First, several classes of
known tyrosinase inhibitors were identified. These com-
pounds feature thiourea moieties (entries 1–4)23 or could
decompose under the assay conditions to yield thioureas
Several classes of small molecules were potent inhibitors
in the high-throughput screen, but upon closer inspec-
Table 1. Selected hits from high-throughput screening and cell-based assays
Entry
Compound
Related hitsb
% Inhibitiona
Mushroom tyrosinase
NMT-1 cells
15 lM
1.7 lM
15 lM
S
N
HN
1
21
21
46
16
6
54
41
44 (3)
34 (1)
56 (4)
38 (9)
19 (20)
48 (3)
OPh
O
O
O
N
O
S
HN
2
3
4
5
6
43
48
52
40
51
40
23
44
25
47
N
H
O
CH3
O
O
O
S
N
S
N
NHPh
H
S
NH
NH
N
H2N
O
N
O
Cl
S
S
N
N
NH
H
O
O
H
N
OC5H11
17
N
O
S
NH2
7
8
3
66
46
47
36
ꢀ15 (10)
N
H
O
N
S
H
N
N
12
Toxic
N
N
H
O
a Values are reported as [(Flꢀ ꢀ Fl+)/(Flꢀ) · 100] (mushroom) or [(Absꢀ ꢀ Abs+)/(Absꢀ) · 100] (MNT-1 cells), where Flꢀ and Absꢀ are fluorescence
or absorbance in the absence of compound. Fl+ and Abs+ are fluorescence or absorbance in the presence of compound. Values in parentheses are
standard deviations.
b Compounds identified in HTS assay with similar structures.

6874
J. P. Germanas et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6871–6875
tion appeared less promising. For instance, a variety of
quinones were identified, but these are likely non-specific
anti-oxidants (entry 6). Other classes of initial hits did
not inhibit pigmentation in cell culture, either due to
low activity (entry 7) or toxicity (entry 8). The differential
activity observed in the cell culture and the in vitro assays
could arise from differences between mushroom and hu-
man tyrosinase, cell penetration, or metabolism.
The biological properties of selected 2-amino-4-aryl-
thiazoles are summarized in Table 2. Methyl substitu-
tion of the pyridyl ring enhances inhibitory potency in
cell culture, although the position is not critical (entries
2–4). While resorcinol rings occupied the 4-position of
most thiazoles (entries 1–4), this substructure was not
necessary for activity. For example, thiazoles 11 and
12 were identified by high-throughput screening and
lack the resorcinol moiety. Inhibitory potency is lost rel-
ative to compound 8, but certainly not abolished, by
replacing the resorcinol ring with a toluene (11) or imi-
dazopyridine ring (12). Further, thiazole 13 possesses
the 4-(resorcinyl)-thiazole core, but is inactive in cell cul-
ture. Finally, resorcinol itself was found to inhibit mush-
room tyrosinase, but with substantially decreased
potency relative to 7 and 8 (IC50 ca. 20 lM).
Two inhibitors emerged from our screening experiments
that warranted further study. A bis-resorcinol-substi-
tuted thiazole (7) was found to inhibit mushroom tyros-
inase with an IC50 of around 2 lM using either the
fluorescence assay (Fig. SI2) or the common Bestman’s
hyrazone assay.16 Among compounds screened, 7 was
the most potent inhibitor of human tyrosinase, with
treatment at 15 lM resulting in an 86–97% reduction
in pigmentation in MNT-1 cells (Table 2). An amino-
pyridyl-thiazole (8) was discovered that showed the
most potent activity against mushroom tyrosinase
(IC50 ꢁ 0.1 lM, Fig. SI2) and retained substantial po-
tency in cell culture: a 78–88% reduction in pigmenta-
tion was observed in the presence of 15 lM
compound.25
As concerns regarding toxicity and efficacy compromise
the future utility of current treatments for hyperpigmen-
tation, we initiated a program to identify novel inhibi-
tors of tyrosinase. We succeeded in identifying a
fluorogenic tyrosine analogue that allowed us to develop
an assay for tyrosinase activity. High-throughput
screening revealed approximately 400 compounds that
inhibited mushroom tyrosinase. Further tests using a
human melanoma cell line revealed a collection of about
50 small molecules that minimized pigmentation in cell
culture without displaying toxicity. Two of these com-
pounds—both diaryl thiazoles—may represent valuable
lead structures for the development of safer treatments
for hyperpigmentation.
Table 2. Inhibition of pigmentation in MNT-1 cells with selected
thiazolesa
Entry
Compound
Inhibition of
pigmentation (%)
S
N
OCH3
1
91 (7)
83 (7)
54 (12)
74 (5)
75 (4)
48 (10)
ꢀ6 (3)
H3CO
HO
HO
HO
HO
OH
7
Acknowledgments
S
N
NH
N
Funding from Galderma and UTSW are gratefully
acknowledged. J.M.R. is a Southwestern Medical Foun-
dation Scholar in Biomedical Research. Anand Ganesan
(UTSW) provided NMT-1 cells.
2
3
4
5
6
7
CH3
OH
8
S
N
NH
N
Supplementary data
OH
OH
9
S
N
Experimental procedures for preparation of PAP-AMC
and enzyme and cell-based assays; spectral data; dose–
response data. Supplementary data associated with this
article can be found, in the online version, at
doi:10.1016/j.bmcl.2007.10.014.
NH
N
CH3
10
S
NH
N
N
CH3
11
H3C
H3C
References and notes
S
N
NH
1. Reviews (a) Mayer, A. M.; Harel, E. Phytochemistry 1979,
18, 193; (b) Hearing, V. J., Jr. Methods Enzymol. 1987,
142, 154; (c) Gerdemann, C.; Eicken, C.; Krebs, B. Acc.
Chem. Res. 2002, 35, 183; (d) Seo, S. Y.; Sharma, V. K.;
Sharma, N. J. Agric. Food Chem. 2003, 51, 2837.
2. Grimes, P.; Nordlund, J. J.; Pandya, A. G.; Taylor, S.;
Rendon, M. P.; Ortonne, P. J. Am. Acad. Dermatol. 2006,
54, S255.
N
N
N
CH3
12
S
N
NH
CO2H
HO
OH
13
a Values are reported as [(Absꢀ ꢀ Abs+)/(Absꢀ) · 100], where Absꢀ
and Abs+ are absorbance in the absence or presence of compound,
respectively. Values in parentheses are standard deviations.
3. Hartong, D. T.; Berson, E. L.; Dryja, T. P. Lancet 2006,
368, 1795.
4. Pandya, A. G.; Guevara, I. L. Dermatol. Clin. 2000, 18, 91.

J. P. Germanas et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6871–6875
6875
5. Briganti, S.; Camera, E.; Picardo, M. Pigment Cell Res.
2003, 16, 101.
6. The MSDS for benzoquinone lists it as a mutagen (human
bone marrow, lymphocytes, murine embryos and lympho-
cytes) and a possible carcinogen. Intravenous LD50
(rat) = 25 mg/kg.
7. Smith, M. T.; Yager, J. W.; Steinmetz, K. L.; Eastmond,
D. A. Environ. Health Perspect. 1989, 82, 23.
8. Rossi, L.; Moore, G. A.; Orrenius, S.; O’Brien, P. J. Arch.
Biochem. Biophys. 1986, 251, 25.
17. Nakumura, K.; Yoshida, M.; Uchiwa, H.; Kawa, Y.;
Mizoguchi, M. Pigment Cell Res. 2003, 16, 494.
18. Cushing, M. L. J. Am. Chem. Soc. 1948, 70,
1184.
19. (a) Jordan, A. M.; Khan, T. K.; Malkin, H.; Osborn, H.
M. I. Bioorg. Med. Chem. 2002, 10, 2625; (b) Knaggs, S.;
Malkin, H.; Osborn, H. M. I.; Williams, N. A. O.;
Yaqoob, P. Org. Biomol. Chem. 2005, 3, 4002.
20. Osborn, H. M. I.; Williams, N. A. O. Org. Lett. 2004, 6,
3111.
9. (a) DeCaprio, A. P. Crit. Rev. Toxicol. 1999, 29, 283; (b)
Kasraee, B.; Handjani, F.; Aslani, F. S. Dermatology 2003,
206, 289.
10. EEC Cosmetics Directive 84/415/EEC, enclosure III bans
the use of hydroquinone. Directive 85/391/EEC bans specific
hydroquinone ethers. For a list of EEC cosmetics directives,
see: http://www.leffingwell.com/cosmetics/vol_1en.pdf.
11. Federal Register, Proposed Rule, 2006, 71, Docket No.
1978N-0065.
21. (a) Hoffmann, K.; Brosch, G.; Loidl, P.; Jung, M.
Pharmazie 2000, 55, 601; (b) Ninkovic, M.; Riester, D.
L.; Wirsching, F.; Dietrich, R.; Schwienhorst, A. Anal.
Biochem. 2001, 292, 228; (c) Li, J.; Chen, L.; Cui, Y.; Luo,
Q.; Li, J.; Nan, F.; Ye, Q. Biochem. Biophys. Res.
Commun. 2003, 307, 172; (d) Ramarao, M. K.; Murphy,
E. A.; Shen, M. W. H.; Wang, Y.; Bushell, K. N.; Huang,
N.; Pan, N.; Williams, C.; Clark, J. D. A. Anal. Biochem.
2005, 343, 143; (e) Heltweg, B.; Trapp, J.; Jung, M.
Methods 2005, 36, 321.
12. Saruno, R.; Kato, F.; Ikeno, T. Agric. Biol. Chem. 1979,
43, 1337.
22. Fryer, J. P.; Oetting, W. S.; Brott, M. J.; King, R. A.
J. Invest. Dermatol. 2001, 117, 1261.
13. (a) Funayama, M.; Arakawa, H.; Yamamoto, R.; Nish-
ino, T.; Shin, R.; Murao, S. Biosci. Biotech. Biochem.
1995, 59, 143; (b) Maeda, K.; Arakawa, H. J. Pharm. Exp.
Ther. 1996, 276, 765; (c) Nihei, K.; Kubo, I. Bioorg. Med.
Chem. Lett. 2003, 13, 2409.
23. (a) Andrawis, A.; Kahn, V. Biochem. J. 1986, 235, 91; (b)
Yamamoto, S. Japanese Patent 91-76587, 1993; (c) Doo-
ley, T. P.; Curto, E. V. WO 2001064206, 2001.
24. (a) Tukamoto, T. T.; Taniguchi, T. J. Invest. Dermatol.
1958, 30, 305; (b) Mars, U.; Larsson, B. S. Pigment Cell
Res. 1995, 8, 194.
25. Saida, M.; Inoe, T.; Tagami, Y.; Betsupu, K.; Shimozono,
J.; Mukai, M.; Oota, S. Japanese Patent 05222015 A2,
1993.
14. Fuyuno, I. Nature 2004, 432, 938.
15. Ni-Komatsu, L.; Keung, J. K.; Williams, D.; Min, J.;
Khersonsky, S. M.; Chang, Y.-T.; Orlow, S. J. Pigment
Cell Res. 2005, 18, 446.
16. Mazzocco, F.; Pifferi, P. G. Anal. Biochem. 1976, 72, 643.