Chemico-Biological Interactions 183 (2010) 462471 [631790]

Chemico-Biological Interactions 183 (2010) 462–471
Contents lists available at ScienceDirect
Chemico-Biological Interactions
journal homepage: www.elsevier.com/locate/chembioint
Structure–toxicity relationship of phenolic analogs as anti-melanoma agents:
An enzyme directed prodrug approach
Nikhil M. Vada, Prabodh K. Kandalaa, Sanjay K. Srivastavaa, Majid Y. Moridania,b,∗
aDepartment of Pharmaceutical Sciences, School of Pharmacy, Texas Tech University Health Sciences Center, 1406 Coulter Drive, Amarillo, TX 79106, U SA
bDepartment of Pediatrics, School of Medicine, Texas Tech University HSC, Amarillo, TX 79106, USA
article info
Article history:
Received 1 August 2009Received in revised form19 November 2009Accepted 19 November 2009Available online 26 November 2009
Keywords:4-HydroxyanisoleTyrosinaseMelanomaB16-F0SK-MEL-28Ethyl 4-hydroxybenzoateabstract
The aim of this study was to identify a phenolic prodrug compound that is minimally metabolized by
rat liver microsomes, but yet could form quinone reactive intermediates in melanoma cells as a resultof its bioactivation by tyrosinase. In current work, we investigated 24 phenolic compounds for theirmetabolism by tyrosinase, rat liver microsomes and their toxicity towards murine B16-F0 and humanSK-MEL-28 melanoma cells. A linear correlation was found between toxicities of phenolic analogs towardsSK-MEL-28 and B16-F0 melanoma cells, suggesting similar mechanisms of toxicity in both cell lines. 4-HEB was identified as the lead compound. 4-HEB (IC
5048 h, 75 /H9262M) showed selective toxicity towards
five melanocytic melanoma cell lines SK-MEL-28, SK-MEL-5, MeWo, B16-F0 and B16-F10, which expressfunctional tyrosinase, compared to four non-melanoma cells lines SW-620, Saos-2, PC3 and BJ cells andtwo amelanotic SK-MEL-24, C32 cells, which do not express functional tyrosinase. 4-HEB caused sig-nificant intracellular GSH depletion, ROS formation, and showed significantly less toxicity to tyrosinasespecific shRNA transfected SK-MEL-28 cells. Our findings suggest that presence of a phenolic group in4-HEB is critical for its selective toxicity towards melanoma cells.
Published by Elsevier Ireland Ltd.
1. Introduction
Skin cancer is one of the most common types of cancer in the
United States [1]. Although rare in comparison to basal and squa-
mous cell carcinoma of the skin, melanoma is the most lethalcancer of the skin. Human epidemiological studies have revealeda significant increase in the incidence of melanoma in Westernpopulations; the number of cases worldwide has doubled in thepast twenty years [2]. In its early stages malignant melanoma
can be cured by surgical resection, but once it has progressed to
Abbreviations: DETAPAC, diethylenetriaminepentaacetic acid; FBS, fetal bovine
serum; MEM, minimum essential medium alpha; DMEM, Dulbecco’s modifiedeagle medium; DTNB, 5,5
/prime-dithiobis-(2-nitrobenzoic acid); AA, ascorbic acid;
MTT, (3-(4,5-dimethylthiazolyl-2)-2,5-diphenyl tetrazolium bromide); shRNA,short hairpin RNA; PMSF, phenylmethylsulphonyl fluoride; SSA, sulphosalicylicacid; ROS, reactive oxygen species; MPT, membrane permeability transition;4-HA, 4-hydroxyanisole; 4-HT, 4-hydroxytoluene; 4-HBAL, 4-hydroxybenzylalcohol; 4-HAP, 4-hydroxyacetophenone; 4-HBA, 4-hydroxybenzoic acid; 4-HEB,ethyl 4-hydroxybenzoate; 3-HA, 3-hydroxyanisole; 3-HT, 3-hydroxytoluene;3-HBAL, 3-hydroxybenzyl alcohol; 3-HAP, 3-hydroxyacetophenone; 3-HBA, 3-hydroxybenzoic acid; 3-HEB, ethyl 3-hydroxybenzoate; 2-HA,2-hydroxyanisole; 2-HT, 2-hydroxytoluene; 2-HBAL, 2-hydroxybenzyl alco-hol; 2-HAP, 2-hydroxyacetophenone; 2-HBA, 2-hydroxybenzoic acid; 2-HEB, ethyl2-hydroxybenzoate; E-4MB, ethyl 4-methoxybenzoate.
∗Corresponding author. Tel.: +1 806 356 4750×225; fax: +1 806 356 4770.
E-mail address: majid.moridani@ttuhsc.edu (M.Y. Moridani).the metastatic stage it is refractory to treatment and does not
respond to the currently available therapies [3]. Novel chemother-
apeutic agents for the treatment of patients with disseminatedmalignant melanoma are urgently needed. Because the majorityof chemotherapy leads to severe systemic toxicity [4], strategies
that can enhance selective drug delivery to melanoma are highlydesirable. Because tyrosinase enzyme is expressed abundantlyand upregulated in melanoma cells [5], it is considered to be a
suitable molecular target for an enzyme directed prodrug bioac-tivation approach for selective drug delivery to melanoma [6].
It has also been reported that tyrosinase is present in relativelyhigh concentrations in human melanoma tissue [7]. In a study of
tyrosinase prepared from normal skin and metastatic melanomafrom the same patient, 46–95 units/mg were found in the skin and17,000–19,500 units/mg in the melanoma [7]. In this approach,
phenolic-based prodrugs are bioactivated to catechols and thento quinones in melanoma cells as a result of their oxidation bytyrosinase ( Fig. 1 ). Using this approach, 4-hydroxyanisole (4-HA)
was previously investigated in melanoma clinical trials [8–10] .
However these clinical trials were terminated later because 4-HAcaused serious liver toxicity [11] . Previously, we reported that 4-HA
undergoes metabolism by liver P450 enzymes to form p-quinonewhich is cytotoxic to isolated rat hepatocytes [10] . Based on these
findings, we hypothesized that P450 mediated bioactivation of4-HA in hepatocytes contributed in 4-HA induced liver toxic-ity. Recently, using the tyrosinase directed prodrug bioactivation
0009-2797/$ – see front matter. Published by Elsevier Ireland Ltd.
doi: 10.1016/j.cbi.2009.11.020

N.M. Vad et al. / Chemico-Biological Interactions 183 (2010) 462–471 463
Fig. 1. Postulated biochemical mechanism of action for phenolic prodrug toxicity in melanoma cells.
approach; we investigated ethyl 4-hydroxybenzoate (4-HEB),
acetylsalicylic acid (ASA), N-acetyl-para-aminophenol (APAP) asanti-melanoma agents and investigated their mechanism of tox-icity in human SK-MEL-28 melanoma cells [12–14] . Our findings
suggested that the mechanism of 4-HEB, ASA, APAP toxicity inSK-MEL-28 cells included o-quinone formation, intracellular GSHdepletion, ROS formation, and mitochondrial toxicity [12–14] .
In the current study, our aims were: (1) to identify a lead pheno-
lic agent that was minimally metabolized by rat liver microsomalpreparations but yet could form quinone reactive intermediatesin melanoma as a result of its bioactivation by tyrosinase [15] .
Twenty four phenolic analogs of 4-HA and 4-HEB ( Fig. 2 ) were
investigated for their metabolism by tyrosinase, rat liver micro-somal preparation and for their toxicity towards melanoma celllines; (2) to demonstrate the selective toxicity of the lead com-pound towards melanoma cell lines compared to non-melanomacells lines; (3) to investigate the role of phenolic group asan essential functional group in the chemical structure of thelead prodrug compound in drug induced toxicity in SK-MEL-28melanoma cells. A tyrosinase specific shRNA was also used toinvestigate the role of tyrosinase in phenolic-based prodrug bioac-tivation.
2. Materials and methods
2.1. Materials
All materials, solvents and reagents used in this work were ana-
lytical grade with the highest degree of purity and were purchasedeither from Sigma–Aldrich, St. Louis, MO or Fisher-Scientific, Pitts-burgh, PA. Mushroom tyrosinase was used throughout this studybecause purified human tyrosinase is not available commercially.DMSO was used to dissolve the phenolic agents for addition to thecell culture. Because the compounds were dissolved in DMSO, thefinal concentration of DMSO was 0.1% and 1% (v/v) in culture mediaof the cells treated with drugs. Therefore, the media for controlcells contained 0.1% and 1% (v/v) DMSO in the experiment. DETA-PAC (diethylenetriaminepentaacetic acid) was purchased fromSigma–Aldrich, St. Louis, MO. Phenolic agents that were investi-gated in this study were split into four groups ( Fig. 2 ) and included:
(group A prodrugs) 4-hydroxy derivatives: 4-hydroxyanisole(4-HA, I); 4-hydroxytoluene (4-HT, II), 4-hydroxybenzyl alcohol (4-
HBAL, III), 4-hydroxyacetophenone (4-HAP, IV), 4-hydroxybenzoic
acid (4-HBA, V), ethyl 4-hydroxybenzoate (4-HEB, VI); (group B
prodrugs) 3-hydroxy derivatives: 3-hydroxyanisole (3-HA, VII), 3-
hydroxytoluene (3-HT, VIII), 3-hydroxybenzyl alcohol (3-HBAL,IX), 3-hydroxyacetophenone (3-HAP, X), 3-hydroxybenzoic acid
(3-HBA, XI), ethyl 3-hydroxybenzoate (3-HEB, XII); (group C
prodrugs) 2-hydroxy derivatives: 2-hydroxyanisole (2-HA, XIII),
2-hydroxytoluene (2-HT, XIV), 2-hydroxybenzyl alcohol (2-HBAL,
XV), 2-hydroxyacetophenone (2-HAP, XVI), 2-hydroxybenzoic acid
(2-HBA, XVII ), and ethyl 2-hydroxybenzoate (2-HEB, XVIII ); and
(group D prodrugs) control derivatives that lack a hydroxy func-tional group on the aromatic ring: anisole ( XIX), toluene ( XX),
benzyl alcohol ( XXI), acetophenone ( XXII ), benzoic acid ( XXIII ),
and ethyl benzoate ( XXIV ). In addition, 4-methoxyanisole and ethyl
4-methoxybenzoate (E-4MB), which are the methoxy analogs of4-HA and 4-HEB, respectively, were also included in this studyto investigate the role of the 4-hydroxy functional group of 4-HA and 4-HEB in toxicity. All phenolic analogs used in this studywere of analytical grade ( ≥99% purity) and available commer-
cially.
Ribonuclease A (Cat. No. R6513) and propidium iodide (Cat. No.
P4170) were obtained from Sigma–Aldrich, St. Louis, MO. The Sure-Silencing shRNA plasmid (Cat. No. KH01771N for the Neomycinresistance) for human tyrosinase was obtained from SuperAr-ray Bioscience Corporation, Frederick, MD. The anti-tyrosinasemonoclonal antibody (232 /H9262g//H9262L; Cat. No. 05-647) was obtained
from Upstate Innovative Cell Signalling Solutions, Lake Placid,NY.
2.2. Cell lines and culture conditions
Modified Eagle Medium Alpha (MEM) (1 ×) (Cat. No. 32571-
036), fetal bovine serum (FBS) (Cat. No. 10082-139), andpenicillin-streptomycin (10,000 units/mL; Cat. No. 15140-122)were purchased from American Type Culture Collection (ATCC
®),
Manassas, VA. Dulbecco’s Modified Eagle Medium (DMEM) (Cat.No. 11965-092) and Versene (1 ×, 0.2 g EDTA 4Na/L in phosphate-
buffered saline) (1:5000 Cat. No. 15040-066) were purchasedfrom Invitrogen, Grand Island, NY. RPMI medium 1640 (1 ×)
(Cat. No. 11875-119) was obtained from Invitrogen Corporation,Grand Island, NY. Cell lines were obtained from ATCC
®, Manassas,
VA.
2.3. Tyrosinase-mediated GSH depletion assay
A previously described method [12,13,16] using Ellman’s
reagent (DTNB) [17] was employed to measure the extent of GSH
depletion as a result of the enzymatic oxidation of the phenolicagent by tyrosinase/O
2. The experiment was performed in three
replicates.

464 N.M. Vad et al. / Chemico-Biological Interactions 183 (2010) 462–471
Fig. 2. Chemical structures of phenolic analogs.
2.4. Rate of AA and NADH oxidation mediated by metabolic
bioactivation of the phenolic compound in the presence oftyrosinase/O
2
A previously described method [18] was used to measure
the rate of ascorbic acid (AA) and NADH oxidation mediated bymetabolism of the phenolic prodrug compound by tyrosinase/O
2
[12,13,16] . The experiment was performed in three replicates.
2.5. Animal housing and protocol
Adult male Sprague–Dawley rats, 250–300 g, were obtained
from Charles River Laboratories, Wilmington, MA, fed ad libitum ,
were allowed to acclimatize for 1 week on clay chip bedding in aroom with a 12 h light photo cycle, an environmental temperatureof 21–23
◦C and 50–60% relative humidity. The rats were housed
under specific pathogen free conditions in accordance with the ani-mal protocols used in current investigation. The animal protocolsfor rat liver microsomal preparation were reviewed and approvedby Institutional Animal Care and Use Committee at Texas Tech Uni-
versity Health Sciences Center, Amarillo, TX.
2.6. Rat liver microsomal preparation
The CYP2E1 induced microsomes were prepared from rats
treated ( i.p.) with inducing agent pyrazole (200 mg/kg/day) by
differential centrifugation as described previously [16] . CYP2E1
induced microsomes were used because our previous study showedthat 4-HA was more cytotoxic towards CYP2E1 induced isolated rathepatocytes in comparison with normal rat hepatocytes [10] . The
experiment was performed in three replicates.
2.7. Enzymatic oxidation of phenolic agent by rat liver
microsomal preparation
A previously described method using Ellman’s reagent (DTNB)
[17] was employed to measure the extent of GSH depletion as
a result of the enzymatic oxidation of the phenolic agent by rat

N.M. Vad et al. / Chemico-Biological Interactions 183 (2010) 462–471 465
Fig. 3. 4-HEB toxicity in the presence and absence of tyrosinase shRNA silencing
plasmid. (A) Tyrosinase protein levels were detected by Western blotting with aspecific anti-tyrosinase monoclonal antibody. Transfection with shRNA3 clone cur-tailed tyrosinase expression. (B) 4-HEB (75 /H9262M) caused significantly less toxicity
in SK-MEL-28 cells transfected with shRNA plasmid directed against tyrosinase.
*Significantly different when compared to control, n=3 .
liver P450 microsomal preparation/NADPH/O 2system [13,16] . The
experiment was performed in three replicates.
2.8. Cell viability
SK-MEL-28, Me Wo, B16-F0, B16-F10, SK-MEL-24, C32, SW-
620, Saos-2, PC-3, and BJ cells were grown as described previously[13,14,19] . All the cells were seeded at 40,000 cells/well except that
SK-MEL-5 cells which were seeded at 4000 cells/well in 24-wellplates. The cell viability assay was performed using yellow tetra-zolium dye (3-(4,5-dimethylthiazolyl-2)-2,5-diphenyl tetrazoliumbromide) (MTT) [13,14,19] . Results from three independent exper-
iments in triplicate were averaged and presented.
2.9. IC
50calculation
The required concentration (IC 50in/H9262M) of the phenolic agent
that can lead to a 50% decrease in cell viability 48 h after incubationwas calculated from the logarithmic regression equation derivedfrom graphing the viability of the cells at 48 h versus the con-centration of the phenolic agent. Results from three independentexperiments in triplicate were averaged and presented.
2.10. Tyrosinase shRNA transfection
Tyrosinase-proficient SK-MEL-28 cells (2 ×10
6) were trans-
fected with the 1 /H9262g plasmid DNAs (of all shRNA clones) for 48 h
using the FuGENE6 reagent (Roche Diagnostics Corporation, Indi-anapolis, IN) as previously described [21,30] .
shRNA3, which was most efficient in curtailing tyrosinase
expression ( Fig. 3 A), was chosen for drug sensitivity experiments.
Briefly, SK-MEL-28 cells were seeded at 160,000 cells/well in 6-well plates. 24 h after seeding, the cells were transfected with theshRNA clone as described in the previous section. 24 h later, the
untreated controls and the transfected cells along with the non-specific shRNA (negative control) transfected cells were treatedwith 4-HEB (75 /H9262M) and incubated further for 48 h. Cell viabil-
ity was assayed by MTT assay [19] by the procedure detailed
above. Results from three experiments were averaged and pre-sented.
2.11. Intracellular GSH measurement
Exponentially growing cells were seeded at 1 ×10
6cells/mL
in MEM Alpha media supplemented with (10%) FBS in 24-wellplates. 3 h after incubation at 37
◦C, the cells were treated with
an additional 1 mL of media with varying concentrations of 4-HEB and E4MB for 1, 2 and 3 h, respectively. Intracellular GSHwas determined as described previously [12,13,22] in melanocytic
SK-MEL-28 and amelanotic C32 melanoma cells [23] . Results from
three experiments were averaged and presented.
2.12. Reactive oxygen species (H
2O2) formation
Melanocytic SK-MEL-28 and amelanotic C 32 cells were seeded
at 50,000 cells/well in 96-well black microplates (Cat. No. 3603,Corning Incorporated, Corning, NY). Reactive oxygen species (ROS)formation was determined using 2
/prime,7/prime-dichlorofluorescein diac-
etate as described previously [12,24] . Results from six replicates
were averaged and presented.
2.13. Cell cycle analyses
0.3×106melanocytic SK-MEL-28 melanoma cells were plated
in a T-75 flask in 16 mL MEM media and allowed to attachovernight. 24 h after seeding the cells, the desired concentrations(75–500 /H9262M) of phenolic prodrug compound were added to the
T-75 flask in 4 mL media. Equal volume of DMSO was added to con-trol. 48 h after incubation, cells were detached using trypsin–EDTAsolution (2–5 mL). Cell cycle distribution was assessed by flowcytometry after staining the cells with propidium iodide asdescribed previously [25] in melanocytic SK-MEL-28 melanoma
cells. Results from three independent experiments were averagedand presented.
2.14. Partition coefficient
Partition coefficient values were estimated using the LogP pro-
gram available at www.logp.com .
2.15. Statistical analysis
Results from three to six replicate experiments have been
reported as mean ±SD. An analysis of variance (ANOVA) was car-
ried out to compare the percentage of surviving cells for eachcompound at various concentrations followed by Bonferoni’s postt-test.
3. Results
3.1. Tyrosinase-mediated GSH depletion assay
Most of the 4-hydroxy derivatives tested in this study
underwent metabolic oxidation by tyrosinase/O
2metaboliz-
ing system more than their corresponding 2-hydroxy and3-hydroxy derivatives ( Table 1 ). As an example, the order of
metabolism for 4-hydroxy derivatives by tyrosinase/O
2was
4-hydroxyanisole /greatermuch4-hydroxytoluene > 4-hydroxybenzyl alco-

466 N.M. Vad et al. / Chemico-Biological Interactions 183 (2010) 462–471
Table 1
Metabolism of phenolic compounds by tyrosinase and rat liver microsomal preparations and their cytotoxicity IC 50(48 h) towards murine B16-F0 and human SK-MEL-28
cell lines.
Phenolic agent % GSH depletion (2 h) IC 50(48 h) LogP
Tyrosinase/O 2 Rat liver microsomes B16-F0 (mM) SK-MEL-28 (mM)
4-Hydroxy analogs4-Hydroxytoluene 88 ±46 0 ±7 1.9 2.0 1.93
4-Hydroxyanisole 95 ±3
†44±8 0.06 0.05 1.56
4-Hydroxybenzyl alcohol 70 ±61 8 ±2 2.2 2.0 0.58
4-Hydroxyacetophenone 56 ±61 9 ±3 3.0 8.0 1.63
4-Hydroxybenzoic acid 30 ±51 1 ±4 4.2 11.2 1.36
Ethyl 4-hydroxybenzoate 24 ±41 5 ±2 0.190 0.075 2.11
3-Hydroxy analogs
3-Hydroxytoluene 79 ±85 1 ±4 2.1 2.5 1.88
3-Hydroxyanisole 24 ±32 1 ±3 0.043 3.4 1.54
3-Hydroxybenzyl alcohol 29 ±38 ±2 4.6 5.0 0.37
3-Hydroxyacetophenone 60 ±54 8 ±4 1.4 2.0 1.50
3-Hydroxybenzoic acid 10 ±12 4 ±3 1.8 2.3 1.29
Ethyl 3-hydroxybenzoate 9 ±31 6 ±4 1.9 2.1 1.97
2-Hydroxy analogs
2-Hydroxytoluene 11 ±33 9 ±4 3.2 3.2 1.80
2-Hydroxyanisole 25 ±22 4 ±5 0.051 2.2 1.52
2-Hydroxybenzyl alcohol 0 0 1.1 1.5 0.772-Hydroxyacetophenone 82 ±55 0 ±6 3.5 4.0 2.02
2-Hydroxybenzoic acid 0 12 ±3 5.2 5.0 1.88
Ethyl 2-hydroxybenzoate 22 ±24 ±1 2.6 2.0 2.48
Non-hydroxy analogs
Toluene 26 ±21 1 ±1 >10 10 2.66
Anisole 21 ±32 0 ±3 >5 >12 2.03
Benzyl alcohol 0 36 ±4 8.0 >10 1.24
Acetophenone 44 ±31 8 ±3 >12 >10 1.74
Benzoic acid 12 ±27 ±
2 11 >10 1.63
Ethyl benzoate 26 ±34 5 ±7 >10 7.0 2.30
†The GSH depletion measurement conducted at 30 min.
The LogP values were estimated using the LogP program available at www.logp.com
hol > 4-hydroxyacetophenone > 4-hydroxybenzoic acid and ethyl
4-hydroxybenzoate (4-HEB).
In order to investigate the role of phenolic functional group
in the oxidation of these compounds, ethyl 4-methoxybenzoate(E4MB), in which the 4-hydroxy functional group is masked by amethoxy group was also investigated for its metabolism by tyrosi-nase. When metabolized by tyrosinase at 2 h incubation time,4-HEB depleted 24% of GSH, which translates to 0.48 mol GSHper mol of 4-HEB oxidation whereas E4MB underwent negligiblemetabolism by tyrosinase/O
2at 2 h incubation ( Table 3 ). Negligible
GSH depletion occurred in the absence of the enzyme. Our find-
Fig. 4. Toxicity correlation between IC 50of phenolic agents in murine B16-F0 and
human SK-MEL-28 melanoma cell lines. 2-Hydroxyanisole, 3-hydroxyanisole, 4-hydroxybenzoic acid, 4-hydroxyacetophenone were considered as outliers, n=3 .ings suggests that the presence of the phenolic functional group in
4-HEB played a central role in 4-HEB metabolic bioactivation to acatechol and then a quinone by tyrosinase, which could then reactwith GSH in the solution ( Figs. 1 and 5 ).
3.2. The rate of AA and NADH oxidation
4-HEB and E4MB were investigated for their ability to oxidize
ascorbic acid (AA) and NADH in the presence of tyrosinase. Therate of AA and NADH oxidation as a result of 4-HEB metabolismby tyrosinase were 143 nM of AA and 389 nM of NADH per 4-HEB(/H9262M) per tyrosinase (unit/mL) per min, respectively ( Fig. 5 ). How-
ever, the rate of AA and NADH oxidation mediated by tyrosinasein the presence of E4MB, which lacks the phenol functional group,were 8 nM of AA and 50 nM of NADH per E4MB ( /H9262M) per tyrosinase
(unit/mL) per min, respectively ( Fig. 5 ;Table 3 ), which were found
to be significantly lower than 4-HEB. These findings indicate thatbecause 4-HEB possesses a phenolic functional group it can readilyundergo bioactivation by tyrosinase to a quinone, which can in turnoxidize AA and NADH, rapidly.
3.3. Metabolism by rat liver P450 microsomal
preparation/NADPH/O
2
Previously, we showed that CYP2E1 induced isolated rat hepato-
cytes were more susceptible to the toxic effects of phenolic-basedcompounds [26] . CYP2E1 induced rat liver microsomes were used
to investigate the metabolism of these phenols. There was no dis-tinctive order of metabolism observed for the 2-, 3- and 4-hydroxyderivatives when the compounds were incubated by the CYP2E1induced rat liver microsomal preparations ( Table 1 ). 4-HEB and 4-
hydroxybenzoic acid were found to be metabolized the least by rat

N.M. Vad et al. / Chemico-Biological Interactions 183 (2010) 462–471 467
Fig. 5. Enzymatic oxidation of 4-HEB and E4MB at pH 7.4 by tyrosinase. AA and NADH oxidations were monitored at 266 and 340 nm, respectively. Ethyl 4-hydroxybe nzoate
(4-HEB); ethyl 4-methoxybenzoate (E4MB); n=3 .
liver microsomal preparation. Negligible GSH depletion occurred
in the absence of the enzyme.
3.4. Cell viability
The cell viability was measured using trypan blue exclusion test
[20] and was always greater than 90% before seeding the cells.
3.5. Anti-proliferative effects in cell lines
The required concentrations of the compound that can cause
50% decrease in cell viability 48 h after incubation were reportedas (IC
50(/H9262M or mM), 48 h) and are given in Table 1 . Our findings
indicated that the toxicity of the compounds with phenolic func-tional group were dose- and time-dependent. The compounds thatlack the hydroxy functional group were used as negative controls.The negative controls showed no dose- and time-dependency atconcentrations between 1 and 5 mM.
The order of toxicity (IC
50at 48 h) for 4-hydroxy deriva-
tives in both human SK-MEL-28 and murine B16-F0 melanomacell lines was identical. For instance, the order of toxicity inhuman SK-MEL-28 melanoma cells was: 4-hydroxyanisole andethyl 4-hydroxybenzoate ≫4-hydroxytoluene, 4-hydroxybenzyl
alcohol /greatermuch4-hydroxyacetophenone > 4-hydroxybenzoic acid. A lin-
ear correlation IC
50 (SK-MEL-28) = 0.98 ×IC50 (B16-F0) + 0.17; n= 20,
R2= 0.94 was found between the toxicity of phenolic analogs
towards human SK-MEL-28 and murine B16-F0 melanomacells, suggesting similar mechanisms of toxicity may exist inboth cell lines ( Fig. 4 ). 2-Hydroxyanisole, 3-hydroxyanisole,
4-hydroxybenzoic acid, and 4-hydroxyacetophenone were notincluded in the correlation as they were considered as outliersbecause they significantly showed more toxicity towards murineB16-F0 melanoma cell line than in human SK-MEL-28, suggestingthat their mechanism of toxicity may differ in these two cell lines.
The IC
50(48 h) for E4MB in human SK-MEL-28 cells melanoma
cells was found to be 750 /H9262M, demonstrating a 10-fold decrease
in E4MB toxicity when compared to 4-HEB. These findings indi-cate that the presence of 4-hydroxy functional group in chemicalstructure of 4-HEB is essential for its enhanced toxicity towardsmelanoma cells.
When tested at 75 /H9262M concentration, 4-HEB showed consid-
erable toxicity in five melanocytic human SK-MEL-28, SK-MEL-5,MeWo cells and melanocytic murine B16-F0 and B16-F10melanoma cell lines resulting in 52 ±4%, 57 ±5%, 55 ±8%, 62 ±11%,
and 60 ±8% anti-proliferative effects, respectively ( Fig. 6 ). How-
ever, 4-HEB (75 /H9262M) showed 24 ±3% and 28 ±4% anti-proliferative
effects towards amelanotic SK-MEL-24 and C32 cells, which do notexpress functional tyrosinase [23] , respectively, whereas it showed
52±4% anti-proliferative effect in human melanocytic SK-MEL-28
melanoma cells ( Fig. 7 ), which expresses functional tyrosinase [23] .In addition, 4-HEB (75 /H9262M) demonstrated significantly lower tox-
icity when tested against non-melanoma cell lines such as PC-3prostate cells (viability, 85 ±3%), BJ skin fibroblast cells (viabil-
ity, 89 ±7%), Saos-2 osteosarcoma cells (viability, 86 ±1%) and
non-melanoma SW-620 colorectal adenocarcinoma cells (viability,89±4%), which do not express tyrosinase ( Fig. 6 ).
3.6. Drug sensitivity assay using tyrosinase shRNA transfected
cells
4-HA (50 /H9262M) and 4-HEB (75 /H9262M) toxicity in SK-MEL-28 cells
was investigated in tyrosinase shRNA transfected cells. A non-
Fig. 6. 4-HEB toxicity in melanocytic melanoma cells (black bars) and non-
melanoma (open bars) cell lines.*Significantly different, n=3 .
Fig. 7. 4-HEB toxicity in melanocytic melanoma and amelanotic melanoma cell
lines.*Significantly different, n=3 .

468 N.M. Vad et al. / Chemico-Biological Interactions 183 (2010) 462–471
Fig. 8. Intracellular GSH depletion in human melanocytic SK-MEL-28 and amelanotic C32 melanoma cells. (A) 4-HEB (75–300 /H9262M) showed dose- and time-dependent
intracellular GSH depletion in human melanocytic SK-MEL-28 melanoma cells. (B) 4-HEB (75–300 /H9262M) caused significantly less GSH depletion in amelanotic C32 melanoma
cells. (C and D) E4MB (0.75, 1.0, and 1.5 mM) caused significantly less intracellular GSH depletion in SK-MEL-28 and C32 melanoma cells. The IC 50(48 h) for 4-HEB and E4MB
in SK-MEL-28 cells was 75 and 750 /H9262M, respectively. Ethyl 4-hydroxybenzoate (4-HEB); ethyl 4-methoxybenzoate (E4MB), n=3 .
specific shRNA (NS-shRNA) plasmid was used as control ( Fig. 3 A).
Transfection with the shRNA clone 3 which curtailed tyrosinaseexpression by 50% prevented the toxicity of 4-HEB (75 /H9262M) for
24±7% in SK-MEL-28 melanoma cells ( Fig. 3 B). Cells transfected
with non-specific shRNA plasmid (NS-shRNA) did not offer anyprotection against 4-HEB toxicity in SK-MEL-28 melanoma cells.
3.7. Intracellular GSH measurement
Our findings indicate that 4-HEB (75, 150, and 300 /H9262M) depleted
37%, 41%, and 45% of intracellular GSH, respectively, at 2 h incuba-tion with human melanocytic SK-MEL-28 melanoma cells ( Fig. 8 A),
whereas 4-HEB (75, 150, and 300 /H9262M) depleted only 8%, 13%, and
16% of intracellular GSH, respectively, at 2 h incubation with humanamelanotic C32 melanoma cells ( Fig. 8 B). Additionally, E4MB were
investigated for its effect on intracellular GSH depletion. Our find-ings indicate that E4MB (0.75, 1, and 1.5 mM) depleted 10%, 13%,and 15% of intracellular GSH, respectively, at 2 h incubation withhuman melanocytic SK-MEL-28 melanoma cells ( Fig. 8 C), and
depleted 6%, 7%, and 10% of intracellular GSH, respectively, at 2 hincubation with human amelanotic C32 melanoma cells ( Fig. 8 D).
3.8. Reactive oxygen species (H
2O2) formation
Our results indicate that 4-HEB at concentrations ranging
from 75 to 1500 /H9262M at various time points showed a dose- and
time-dependent increase in ROS formation in human melanocyticSK-MEL-28 melanoma cells ( Fig. 9 A). ROS formation observed in
amelanotic C32 melanoma cells ( Fig. 9 B), which do not express
functional tyrosinase [23] was significantly less than ROS forma-
tion in SK-MEL-28 cells, which express functional tyrosinase [23] .Our results indicate that when tested at even higher concentrations
(750/H9262M to 5 mM), ethyl 4-methoxybenzoate (E4MB) caused signif-
icantly lower ROS formation in melanocytic SK-MEL-28 melanomacells in comparison to 4-HEB, and caused negligible ROS formationin amelanotic C32 melanoma cells ( Fig. 9 C and D).
3.9. Cell cycle analyses of 4-HEB treatment on SK-MEL-28 cells
As summarized in Table 2 , SK-MEL-28 cells treated with 4-HEB
(75, 150, 250, and 500 /H9262M) caused corresponding increase in the
percentage of sub G
0/G1cells undergoing apoptosis and an increase
in the percentage of cells in the S phase in comparison to non-treated control cells. The total percentage increase in cells in the subG
0/G1and S phase was accompanied by a corresponding decrease
in cells in the G 1phase.
3.10. Partition coefficient
The LogP value of 4-hydroxy phenolic compounds ( Table 1 ) fol-
lowed an order of ethyl 4-hydroxybenzoates > 4-hydroxyltoluene >4-hydroxyacetophenone and 4-hydroxyanisole > 4-hydroxy-benzoic acid > 4-hydroxybenzyl alcohol which corresponds tophenols with the following functional groups at the para posi-tion to the phenolic group: –COOEt > –CH
3> –C(CH 3)O, –OCH 3,
>–COOH > –CH 2OH. It should be noted that benzoic acid derivatives
are ionized more than 99.99% at pH 7.4 and therefore LogP hasa limited value in the prediction of toxicity for these phenoliccompounds towards melanoma cell lines. 2-Hydroxy derivativeswith –COOEt, –C(CH
3)O, –COOH and –CH 2OH functional groups
had a higher LogP value than 4-hydroxy and 3-hydroxy derivativesfor phenolic compounds mainly due to an intramolecular hydrogen

N.M. Vad et al. / Chemico-Biological Interactions 183 (2010) 462–471 469
Fig. 9. ROS formation in human melanocytic SK-MEL-28 and amelanotic C32 melanoma cells. (A) 4-HEB (75–1500 /H9262M) led to a time- and dose-dependent escalation in
ROS formation in human melanocytic SK-MEL-28 cells. (B) 4-HEB (75–1500 /H9262M) led to significantly less ROS formation in amelanotic C32 melanoma cells. (C and D) E4MB
(750/H9262M to 5 mM) leads to significantly less ROS formation in human melanocytic SK-MEL-28 and amelanotic C32 melanoma cells, n=6 .
binding which hinders the molecule from interaction with aqueous
environment. Toluene, anisole, benzyl alcohol, acetophenone, ben-zoic acid and ethyl benzoate that lacked a phenol demonstratedhigher LogP value than their corresponding hydroxy derivatives.
4. Discussion
Riley’s group had previously investigated the structure–activity
relationship of tyrosinase-dependent cytotoxicity of a series of sub-stituted linear alkoxyphenols [6]. However, their study did not
investigate the metabolic bioactivation of alkoxyphenols by liverrat microsomal preparations enzymes. Moreover the influence ofhydrophilic functional groups such as alcohol and carboxylic acidon the metabolism of phenols by melanoma tyrosinase was notaddressed previously. Such information is invaluable in the designof a safer anti-melanoma phenolic-based prodrug.In this study, 24 phenolic analogs were investigated with
heterogeneous functional groups (2-, 3-, and 4-hydroxy deriva-tives of anisole, toluene, benzyl alcohol, acetophenone, benzoicacid and ethyl benzoate) ( Fig. 2 ) to identify compounds that
are bioactivated by tyrosinase but not by rat liver microsomesand hence limiting their toxicity towards liver. In this study, wehave found that 4-hydroxy derivatives generally demonstrateda higher metabolic bioactivation with tyrosinase with a lowermetabolic bioactivation with CYP2E1 induced rat liver micro-somal preparations. Our findings showed that 4-hydroxylbenzylalcohol, 4-hdyroxybenzophenone, 4-hydroxybenzoic acid and itsethyl ester were metabolized substantially less than 4-HA and4-hydroxytoluene by CYP2E1 induced rat liver microsomal prepa-ration indicating that the introduction of a water soluble functionalgroup such as hydroxyl group or carboxy group hindered theirmetabolic bioactivation by rat liver microsomal preparation sub-
Table 2
SK-MEL-28 cell cycle analysis after treatment with 4-HEB (48 h). 4-HEB (75, 150, 250, and 500 /H9262M) caused corresponding increase in the percentage of sub G 0/G1cells
undergoing apoptosis and an increase in the percentage of cells undergoing S phase cell cycle arrest in comparison to non-treated control cells as ana lyzed by flow cytometry.
SK-MEL-28 cells
Treatment Apoptosis (sub G 0/G1)G 1 SG 2/M
Control 0.1% DMSO 1.6 ±0.1 76.9 ±0.3 11.1 ±0.3 10.2 ±0.2
4-HEB 75 /H9262M 2.9 ±0.3*74.7±0.8 12.4 ±0.7 9.9 ±0.1
150/H9262M 3.1 ±0.2*72.1±0.7 13.2 ±1.2 10.6 ±0.9
250/H9262M 5.0 ±0.9*63.4±1.8*20.1±2.6*10.7±0.7
500/H9262M 6.2 ±0.8*59.8±1.6*22.1±0.9*10.9±0.4
*Significantly different, n=3 .

470 N.M. Vad et al. / Chemico-Biological Interactions 183 (2010) 462–471
Table 3
Summary data for 4-HEB and E4MB. 4-HEB selectively undergoes metabolic bioactivation by tyrosinase, and leads to significant GSH depletion, NADH and AA oxidation
mediated by tyrosinase bioactivation, and selective ROS and intracellular GSH depletion and cytotoxicity in SK-MEL-28 cells when compared to E4MB, which lacks a hydroxy
functional group.
Compounds Tyrosinase metabolism
mediated % GSHdepletion (2 h)Tyrosinase
Kinetics
*IC50(48 h)
SK-MEL-28 cellsICG†depletion
(2 h) at IC 50Fold ROS†
formation(40 min) at IC
50
NADH AA
4-HEB 24±4% 389 nM*143 Nm* 75/H9262MSK-MEL-28 cellsa37% 24
C32 cellsa8% 6
E4MB 5±3% 50 nM**8n M** 750/H9262MSK-MEL-28 cellsb10% 9
C32 cellsb6% 4
a4-HEB was tested at its IC 50(48 h) of 75 /H9262M.
bE4MB was tested at its IC 50(48 h) of 750 /H9262M.
†ICG, intracellular GSH; ROS, reactive oxygen species.
*nM of AA or NADH/4-HEB ( /H9262M)/tyrosinase (units/mL)/min.
**nM of AA or NADH/E4MB ( /H9262M)/tyrosinase (units/mL)/min.
stantially. CYP2E1 induced microsomes were used because our
previous study showed that 4-HA was more cytotoxic towardsCYP2E1 induced isolated rat hepatocytes in comparison with nor-mal rat hepatocytes [10] . However, since induced rat microsomes
were used as a source of CYP2E1, these results should be inter-preted with caution as there may be species differences in thebioactivation of prodrugs by CYP2E1. Use of commercially availablehuman CYP2E1 (supersomes) for investigation of bioactivation ofprodrugs would constitute an important avenue of such researchin the future.
Our study suggests that the oxidation state, presence of an elec-
tron donating/withdrawing group and position of the functionalgroup on the phenol all play a major role in determining the extentof metabolism by tyrosinase and rat liver microsomal preparations,and the toxicity of these compounds towards melanoma, which isin keeping with earlier reports in literature [27,28] .
In cytotoxicity studies, although 4-hydroxybenzoic acid was
not toxic towards the melanoma cell lines its ethyl ester deriva-tive was the second most toxic phenolic analog tested in thestudy after 4-HA. 4-Hydroxybenzoic acid showed an IC
50(48 h) of
11.2 mM in SK-MEL-28 cells, while its ethyl ester prodrug ethyl4-hydroxybenzoate was demonstrated to have an IC
50(48 h) of
75/H9262M. The lack of induced toxicity could be due to the extensive
ionization of 4-hydroxybenzoic acid in culture media at pH 7.4,thereby limiting its entry into the melanoma cells. The presenceof an ester group in chemical structure of ethyl 4-hydroxybenzoateprevents the ionization and thereby the molecule can diffuse acrossthe cell unhindered and contribute towards cell toxicity leading toa 150-fold increase in the toxicity of ethyl 4-hydroxybenzoic acidin comparison to its carboxylic acid derivative.
In an earlier study, we observed a direct relationship between
toxicity towards murine B16-F0 cells and the degree of lipophilicitywas observed via a one parameter quantitative structure–activityrelationship (QSTR) equation [18] . However, in the current study
some of the compounds did not follow the general rule of thegreater lipid solubility the greater cytotoxicity. For instance eventhough 4-hydroxytoluene has similar lipid solubility and metabolicbioactivation by tyrosinase metabolism profiles to that of 4-HA, 4-hydroxytoluene was 40-fold less toxic towards both murine andhuman melanoma cell lines than 4-HA. The reason that 4-HA wasthe more cytotoxic than 4-hydroxytoluene could be attributed tothe presence of an electron donating group of methoxy at paraposition in 4-HA.
Despite being metabolized less by tyrosinase, ethyl 4-
hydroxybenzoate demonstrated a comparable toxicity to that of4-HA towards the melanoma cell lines. One reason could bethat ethyl 4-hydroxybenzoate possesses a higher degree of lipidsolubility than 4-HA ( Table 1 ) which makes it easier for ethyl 4-hydroxybenzoate to enter the cells. The second reason could be
that we have used mushroom tyrosinase for in vitro metabolicbioactivation because purified human tyrosinase is not availablecommercially. In addition, factors other than lipophilicity, elec-tron donating/withdrawing groups, and metabolic bioactivation bytyrosinase may be responsible for toxicity of these phenolic com-pounds towards B16-F0 and SK-MEL-28 cells.
Additionally, in order to investigate the role of the 4-hydroxy
functional group in cytotoxicity towards melanoma cells and theirbioactivation with tyrosinase, ethyl 4-methoxybenzoate (E4MB)and 4-methoxyanisole (4-MA), in which the 4-hydroxy functionalgroup is masked by a methoxy group, were used as controls for 4-HEB and 4-HA, respectively, because they lack hydroxy functionalgroup. Table 3 , summarizes the difference between 4-HEB and
E4MB, and highlights that 4-HEB selectively undergoes metabolicbioactivation by tyrosinase, and leads to significant GSH depletion,NADH and AA oxidation mediated by tyrosinase bioactivation, andselective ROS and intracellular GSH depletion and cytotoxicity inSK-MEL-28 cells when compared to E4MB, which lacks a hydroxylfunctional group.
In order to investigate the role of tyrosinase, the cytotoxicity
of 4-HEB was also investigated in amelanotic C32 and SK-MEL-24melanoma cell lines, which do not express tyrosinase [23] .I tw a s
found that 4-HEB demonstrated significantly less toxicity towardsamelanotic C32 and SK-MEL-24 melanoma cell lines than towardsmelanocytic SK-MEL-28 melanoma cells, which express tyrosinase[23] . 4-HEB also leads to significantly less ROS formation and ICG
depletion in amelanotic C32 than in melanocytic SK-MEL-28. More-over, in an earlier report, Smit et al. [29] showed that prolonged
exposure to agents that exhibited tyrosinase-dependent toxicityresulted in reduction of cell numbers and cell damage, which wasdetectable by various methods. In addition, 4-HEB showed selectivetoxicity towards five melanocytic melanoma cell lines SK-MEL-28,SK-MEL-5, MeWo, B16-F0 and B16-F10 compared to four non-melanoma cells lines SW-620, Saos-2, PC3, and BJ cells. Our resultsalso indicate that 4-HA and 4-HEB showed significantly less celltoxicity when tyrosinase was silenced by shRNA plasmid, thus pro-viding further evidence that melanocytotoxicity was mediated inpart by tyrosinase.
In summary, ethyl 4-hydroxybenzoate was identified to be the
most suitable cytotoxic phenol from this series against murine B16and human SK-MEL-28 melanoma cells with minimal metabolismby rat liver microsomal preparations. Therefore, similar to our pre-vious work [12] , our data suggests that ethyl 4-hydroxybenzoate
may have potential as a lead anti-melanoma compound. However,further investigation into its in vivo efficacy against melanomatumor model in mice and in vivo metabolism, toxicology and phar-macokinetic profiles is required. One should also note that 4-HEB

N.M. Vad et al. / Chemico-Biological Interactions 183 (2010) 462–471 471
may also cause toxicity towards cells having high tyrosinase activ-
ity including dopaminergic neurons, pigmented retinal epithelialcells, and melanocytes which should be also investigated.
Conflict of interest
The authors state no conflict of interest.
Acknowledgments
This work was supported partly by a grant from NIH
(1R15CA122044-01A1) and the TTUHSC, School of Pharmacy.
References
[1] L.A.G. Ries, M.P. Eisner, C.L. Kosary, B.F. Hankey, B.A. Miller, L. Clegg, A. Mari-
otto, E.J. Feuer, B.K. Edwards (Eds.), SEER Cancer Statistics Review, 1975–2001,National Cancer Institute, Bethesda, MD, 2004, http://seer.cancer.gov/csr/
1975
2001/ .
[2] C.M. Anderson, A.C. Buzaid, S.S. Legha, Systemic treatments for advanced
cutaneous melanoma, Oncology (Williston Park) 9 (11) (1995) 1149–1158 (dis-cussion 1163–1164, 1167–1168).
[3] P.A. Riley, Hydroxyanisole depigmentation: in-vitro studies, J. Pathol. 97 (2)
(1969) 193–206.
[4] I. Ojima, Guided molecular missiles for tumor-targeting chemotherapy—case
studies using the second-generation taxoids as warheads, Acc. Chem. Res. 41(1) (2008) 108–119.
[5] U. Vielkind, W. Schlage, F. Anders, Melanogenesis in genetically determined
pigment cell tumors of platyfish and platyfish-swordtail hybrids: correlationbetween tyrosine activity and degree of malignancy, Cancer Res. Clin. Oncol.90 (3) (1977) 285–299.
[6] A.M. Jordan, T.H. Khan, H. Malkin, H.M. Osborn, A. Photiou, P.A. Riley,
Melanocyte-directed enzyme prodrug therapy (MDEPT): development of sec-ond generation prodrugs for targeted treatment of malignant melanoma,Bioorg. Med. Chem. 9 (6) (2001) 1549–1558.
[7] Y.M. Chem, W. Chavin, Tyrosinase activity in highly pigmented human
melanoma and in Negro skin, Proc. Soc. Exp. Biol. Med. 145 (1974) 695–698.
[8] S. Naish, C.J. Cooksey, P.A. Riley, Initial mushroom tyrosinase-catalysed oxida-
tion product of 4-hydroxyanisole is 4-methoxy-ortho-benzoquinone, PigmentCell Res. 1 (6) (1988) 379–381.
[9] S. Naish, J.L. Holden, C.J. Cooksey, P.A. Riley, Major primary cytotoxic product
of 4-hydroxyanisole oxidation by mushroom tyrosinase is 4-methoxy orthobenzoquinone, Pigment Cell Res. 1 (6) (1988) 382–385.
[10] M.Y. Moridani, S.S. Cheon, S. Khan, P.J. O’Brien, Metabolic activation of 4-
hydroxyanisole by isolated rat hepatocytes, Drug Metab. Dispos. 30 (10) (2002)1063–1069.
[11] G.J. Rustin, M.R. Stratford, A. Lamont, et al., Phase I study of intravenous 4-
hydroxyanisole, Eur. J. Cancer 28A (8–9) (1992) 1362–1364.
[12] N.M. Vad, I.H. Shaik, R. Mehvar, M.Y. Moridani, Metabolic bioactivation and
toxicity of ethyl 4-hydroxybenzoate in human SK-MEL-28 melanoma cells, J.Pharm. Sci. 97 (5) (2008) 1934–1945.[13] N.M. Vad, G. Yount, D. Moore, J. Weidanz, M.Y. Moridani, Biochemical mech-
anism of acetaminophen (APAP) induced toxicity in melanoma cell lines, J.Pharm. Sci. 98 (4) (2009) 1409–1425.
[14] N.M. Vad, G. Yount, M.Y. Moridani, Biochemical mechanism of acetylsalicylic
acid (aspirin) selective toxicity toward melanoma cell lines, Melanoma Res. 18(6) (2008) 386–399.
[15] S. Naish-Byfield, C.J. Cooksey, A.M. Latter, C.I. Johnson, P.A. Riley, In vitro
assessment of the structure-activity relationship of tyrosinase-dependentcytotoxicity of a series of substituted phenols, Melanoma Res. 1 (4) (1991)273–287.
[16] M.Y. Moridani, Biochemical basis of 4-hydroxyanisole induced cell toxicity
towards B16-F0 melanoma cells, Cancer Lett. 243 (2) (2006) 235–245.
[17] G.L. Ellman, Tissue sulfhydryl groups, Arch. Biochem. Biophys. 82 (1) (1959)
70–77.
[18] M.Y. Moridani, M. Moore, R.A. Bartsch, Y. Yang, S. Heibati-Sadati, Structural
toxicity relationship of 4-alkoxyphenols’ cytotoxicity towards murine B16-F0melanoma cell line, J. Pharm. Pharm. Sci. 8 (2) (2005) 348–360.
[19] X. Wu, H. Zeng, X. Zhang, et al., Phosphatase of regenerating liver-3 promotes
motility and metastasis of mouse melanoma cells, Am. J. Pathol. 164 (6) (2004)2039–2054.
[20] M.Y. Moridani, S.S. Cheon, S. Khan, P.J. O’Brien, Metabolic activation of 3-
hydroxyanisole by isolated rat hepatocytes, Chem. Biol. Interact. 142 (3) (2003)317–333.
[21] I.E. Johansen, Intron insertion facilitates amplification of cloned virus cDNA in
Escherichia coli while biological activity is reestablished after transcription in
vivo, Proc. Natl. Acad. Sci. U.S.A. 93 (22) (1996) 12400–12405.
[22] I.H. Shaik, R. Mehvar, Rapid determination of reduced and oxidized glutathione
levels using a new thiol-masking reagent and the enzymatic recycling method:application to the rat liver and bile samples, Anal. Bioanal. Chem. 385 (1) (2006)105–113.
[23] K. Hirosaki, T. Yamashita, I. Wada, H.Y. Jin, K. Jimbow, Tyrosinase and
tyrosinase-related protein 1 require Rab7 for their intracellular transport, J.Invest. Dermatol. 119 (2) (2002) 475–480.
[24] A.G. Siraki, T.S. Chan, P.J. O’Brien, Application of quantitative structure–toxicity
relationships for the comparison of the cytotoxicity of 14 p-benzoquinone con-geners in primary cultured rat hepatocytes versus PC12 cells, Toxicol. Sci. 81(1) (2004) 148–159.
[25] M. Lu, A. Strohecker, F. Chen, et al., Aspirin sensitizes cancer cells to TRAIL-
induced apoptosis by reducing survivin levels, Clin. Cancer Res. 14 (10) (2008)3168–3176.
[26] M.Y. Moridani, A. Siraki, P.J. O’Brien, Quantitative structure–toxicity relation-
ships for phenols in isolated rat hepatocytes, Chem. Biol. Interact. 145 (2) (2003)213–223.
[27] B. Kasraee, Peroxidase-mediated mechanisms are involved in the melanocy-
totoxic and melanogenesis-inhibiting effects of chemical agents, Dermatology205 (4) (2002) 329–339.
[28] S. Passi, M. Nazzaro-Porro, Molecular basis of substrate and inhibitory speci-
ficity of tyrosinase: phenolic compounds, Br. J. Dermatol. 104 (6) (1981)659–665.
[29] N.P. Smit, K. Peters, W. Menko, W. Westerhof, S. Pavel, P.A. Riley, Cytotoxicity
of a selected series of substituted phenols towards cultured melanoma cells,Melanoma Res. 2 (5–6) (1992) 295–304.
[30] N.M. Vad, S.K. Kudugunti, D. Graber, N. Bailey, K. Srivenugopal, M.Y. Moridani,
Efficacy of acetaminophen in skin B16-F0 melanoma tumor-bearing C57BL/6mice, Int. J. Oncol. 35 (1) (2009) 193–204.