Effect directed synthesis of a new tyrosinase inhibitor with anti-browning activity

Article abstract of DOI:10.1016/j.foodchem.2020.128232

The inhibition of enzymatic browning is an attractive target to elevate the quality of foods. The objective of this work is to describe a novel platform for the discovery of tyrosinase inhibitors, based on (a) one-pot preparation of a library of thiosemicarbazide compounds, (b) biological evaluation using tyrosinase TLC bioautography, (c) inhibitor identification via mass spectrometry coupled to bioautography. During these proof-of-concept experiments, the approach led to the straightforward identification of a new thiosemicarbazone with improved tyrosinase inhibition properties and fresh-cut apple slices antibrowning effect when compared to kojic acid. In conclusion, the platform represents an interesting strategy for the discovery of this type of inhibitors.

Full text of DOI:10.1016/j.foodchem.2020.128232

FoodChemistry341(2021)128232
Contents lists available at ScienceDirect
Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem
Effect directed synthesis of a new tyrosinase inhibitor with anti-browning
activity
⁎
Ignacio Cabezudoa, I. Ayelen Ramalloa, Victoria L. Alonsob, Ricardo L.E. Furlana,
a Farmacognosia, Facultad de Ciencias Bioquímicas y Farmacéuticas, Universidad Nacional de Rosario and Consejo Nacional de Investigaciones Científicas y Técnicas
(CONICET), Suipacha 531, 2000 Rosario, Argentina
b Instituto de Biología Molecular y Celular de Rosario (IBR-CONICET), Facultad de Ciencias Bioquímicas y Farmacéuticas, Universidad Nacional de Rosario, Argentina
A R T I C L E I N F O
A B S T R A C T
Keywords:
The inhibition of enzymatic browning is an attractive target to elevate the quality of foods. The objective of this
work is to describe a novel platform for the discovery of tyrosinase inhibitors, based on (a) one-pot preparation
of a library of thiosemicarbazide compounds, (b) biological evaluation using tyrosinase TLC bioautography, (c)
inhibitor identification via mass spectrometry coupled to bioautography. During these proof-of-concept ex-
periments, the approach led to the straightforward identification of a new thiosemicarbazone with improved
tyrosinase inhibition properties and fresh-cut apple slices antibrowning effect when compared to kojic acid. In
conclusion, the platform represents an interesting strategy for the discovery of this type of inhibitors.
Tyrosinase
Browning
Inhibition
Bioautography
Thiosemicarbazone
1. Introduction
Dong, Yu, & Cao, 2016).
To the best of our knowledge, all the synthetic tyrosinase inhibitors
Enzymatic browning is a major issue in most fresh fruits and ve-
getables, as it causes unfavourable effects on their safety, organoleptic
properties, and nutritional values (Moon et al., 2018). Therefore, the
inhibition of enzymatic browning is an attractive target to elevate the
quality and safety of foods. Browning is caused mainly by polyphenol
oxidases, copper-containing enzymes related to the hydroxylation of
monophenols to o-diphenols and the oxidation of diphenols to qui-
nones, which undergo subsequent reactions that form brown pigments
(Mi Moon, Young Kim, Yeul Ma, & Lee, 2019; Moon et al., 2018).
Tyrosinase (EC 1.14.18.1) is a copper-containing monooxygenase
widely distributed in microorganisms, animals, and plants (Shao et al.,
2018). Tyrosinase inhibitors are usually considered of interest as anti-
browning agents for the industry that processes vegetables (Carcelli
et al., 2020; Xu, Liu, Zhu, Yu, & Cao, 2017).
reported to date have been prepared and biologically evaluated in-
dividually. Alternatively, the one-pot preparation of compound mix-
tures as sources of potential tyrosinase inhibitors has not been explored.
Although efficient from a synthetic perspective, an important drawback
of this strategy is the innate necessity for inhibitor identification from
the mixtures that display interesting properties.
TLC bioautography is an assay format particularly well suited for
the evaluation of enzyme inhibition properties of complex mixtures
(Bräm & Wolfram, 2017). TLC bioautographic methods combine chro-
matographic separation and in situ enzyme activity determination, fa-
cilitating the localisation of inhibitors within a mixture. This assay
format has been used to detect tyrosinase inhibitors from plant extracts
(Hsu, Chan, Chen, Lin, & Cheng, 2018; Zhou, Tang, Wu, & Cheng, 2017)
and chemically engineered extracts as well (Paula García, Salazar,
Ramallo, & Furlan, 2016; Solís, Salazar, Ramallo, García, & Furlan,
2019). Nevertheless, its utility for the evaluation of synthetic libraries
has not been explored. In addition, TLC has demonstrated to be a
straightforward approach to link chromatographic and biological data
with respective mass spectra of bioactive compounds (Azadniya &
Morlock, 2018; Ciura, Dziomba, Nowakowska, & Markuszewski, 2017;
Jamshidi-Aidji & Morlock, 2016; Krüger, Hüsken, Fornasari, Scainelli,
& Morlock, 2017; Ramallo, Salazar, & Furlan, 2015; Zhang et al., 2017).
The aim of this study was to evaluate the combination of mixture
A large number of tyrosinase inhibitors have been reported in the
literature, from both synthetic and natural sources; nevertheless, most
of them could be unsuitable for practical applications. The inhibition of
metalloenzymes such as tyrosinase, depends in part on the ability to
coordinate the metal ions in the active site (Chen et al., 2019). Thio-
semicarbazones and thiocarbazones are well-known metal-chelating
compounds, such as copper, present in the tyrosinase active site
(Rogolino et al., 2017), and they have been explored for tyrosinase
inhibition properties (Carcelli et al., 2020; Hałdys & Latajka, 2019; Xie,
⁎ Corresponding author.
E-mail addresses: icabezudo@fbioyf.unr.edu.ar (I. Cabezudo), aramallo@fbioyf.unr.edu.ar (I. Ayelen Ramallo), alonso@ibr-conicet.gov.ar (V.L. Alonso),
rfurlan@fbioyf.unr.edu.ar (R.L.E. Furlan).
https://doi.org/10.1016/j.foodchem.2020.128232
Received 8 July 2020; Received in revised form 24 September 2020; Accepted 24 September 2020
Availableonline28September2020
0308-8146/©2020ElsevierLtd.Allrightsreserved.

I. Cabezudo, et al.
FoodChemistry341(2021)128232
library preparation, tyrosinase TLC bioautography, and algorithm aided
mass spectrometry analysis, as a platform for the discovery of tyr-
osinase inhibitors. Using this platform, a new mixed type tyrosinase
(T2) stock solutions were prepared in DMSO (30 mg/mL for A1, A2, A4-
A8, and T1; 15 mg/mL for A3 and T2). Adequate volumes of each
building block solution were mixed and diluted in EtOH (2 mL) to
obtain final concentrations of 0.80 mM for each aldehyde, 3.2 mM for
T1, and 1.6 mM for T2, maintaining a 1:1 proportion between the
equivalents of thiosemicarbazone and aldehyde reactive groups. The
reaction was stirred under reflux for three hours. The library was stored
at 8 °C.
inhibitor with a Ki = 0.29
0.11 µM and interesting anti-browning
effect on apple slices was discovered.
2. Materials and methods
Mushroom tyrosinase, dimethyl sulfoxide (DMSO), L-tyrosine (Tyr),
2-thiophenecarboxaldehyde, 4-hydroxy-3-methoxybenzaldehyde (A1),
3,4,5-trimethoxybenzaldehyde (A2), anthracene-9-carbaldehyde (A3),
5-ethoxyfuran-2-carbaldehyde (A4), 5-hydroxyfuran-2-carbaldehyde
LC-MS analysis of the library was carried out using a reversed-phase
HPLC Agilent Zorbax Eclipse column (50 mm × 3.0 mm XDB C18,
particle size 1.8um) following step-wise gradient: 0 min (water 9:1
ACN); 3 min (water 9:1 ACN); 23–25 min (water 1:9 ACN). Water
contained 0.5% formic acid, column temperature was maintained at
30 °C, and autosampler tray temperature was set at 25 °C.
(A5),
2,3-dihydroxypropanal
(A6),
4-(diethylamino)-2-hydro-
xybenzaldehyde (A7), hexanal (A8), hydrazinecarbothioamide (T1)
and hydrazinecarbothiohydrazide (T2) were purchased from Sigma
Aldrich (St. Louis, MO, USA). Sodium phosphate dibasic and sodium
phosphate monobasic was purchased in Cicarelli (San Lorenzo,
Argentina). Agar was purchased from Britania (Buenos Aires,
Argentina). HPLC-grade acetonitrile was purchased from Carlo Erba
Regents (Milan, Italy). All other reagents were analytical grade, and the
water used was re-distilled and ion-free.
2.3. Tyrosinase bioautography assays
For a 70 cm2 TLC sheet, the staining solution for tyrosinase was
prepared by dissolving agar (135 mg) at 90 °C in sodium phosphate
buffer (20 mM, pH 6.8, 11.2 mL). Whenever the TLC sheets size was
different, the quantities were modified proportionally according to the
technique reported (Paula García & Furlan, 2015). The solution was
allowed to cool to 55 °C, the L-tyrosine solution (2.5 mM, 2.8 mL) was
added, and the whole mixed by hand. At 35 °C, tyrosinase solution
(3800 U/mL, 130 µL) was added, and the obtained solution was mixed
by rotation by hand and then carefully poured on the already developed
TLC sheet (Paula García & Furlan, 2015).
Thin-layer chromatography was carried out on aluminum-backed
silica gel 60 F254 (Merck, Darmstadt, Germany). Chromatograms were
run and processed using equipment from CAMAG (Muttenz,
Switzerland). Separation of synthetic mixtures, controls, and pure
compounds was performed on 7 cm height TLC layers. Samples were
applied in 4 mm bands onto the TLC plate using a CAMAG Automatic
TLC Sampler 4 (ATS 4) under air flux. After solvent evaporation, TLC
images were captured under white light and UV light with a CAMAG
TLC Visualizer. For bioautographic purposes, agar containing enzyme
and medium was distributed evenly over the TLC layer. Colour devel-
opment and control experiments were carried out as described in
Section 2.3, and plate images were captured under white light again.
The chromatograms were scanned using a TLC scanner 4, and optical
density graphs were plotted.
2.3.1. Spiking experiments
The library was diluted 2.4 times using EtOH. Aliquots were spiked
with increasing amounts (1.6 and 3.2 µL) of an A7T1 solution in EtOH
(0.4 mg/mL), for a final addition of 0.015 and 0.036 µg of A7T1 on
each TLC spot. These two samples and a library control were spotted
and eluted with DCM:MeOH (97:3) and evaluated by TLC-tyrosinase
bioautography, according to section 2.3.
1H NMR (300 MHz), 13C NMR (75 MHz) were measured on a Bruker
Avance II 300 MHz spectrometer. LC-MS and direct infusion MS ex-
periments were carried out using an Ultimate 3000 RSLC (Dionex,
Thermo Scientific) coupled with an ESI triple quadrupole mass spec-
trometer (TSQ Quantum Access Max (QQQ), Thermo-Scientific). The
following ionization conditions were used: ESI, positive-ion mode;
drying gas (N2) temp., 300 °C; drying gas flow rate, 10 L/min; nebulizer
pressure, 10 UA, and cap. voltage, 4.5 kV.
2.4. BioMSId evaluation of thiosemicarbazone compounds
2.4.1. Sample preparation
For inhibitor detection in the mixture library, three TLC plates were
spotted with 6 μL samples of the library. They were eluted with three
different mobile phases: DCM:MeOH (97:3), hexane:EtOAc (4:6), and
toluene:CHCl3:EtOH (5:5:2). Each TLC was tested for tyrosinase in-
hibition using the described bioautography conditions, and circular
portions of the agar layer from inhibition and background zones were
taken using a glass tube (internal diameter 0.5 cm). Each gel portion
was extracted with EtOH (2 × 1 mL) using an ultrasonic bath for 5 min.
The ethanol was removed under air current; the solid was dissolved in
ACN (0.5 mL) and filtered with 0.45 µm RC syringe filters. The resulting
solutions were infused directly into the ESI chamber at a rate of 25 μL/
min during 5 min for LRMS analysis. The spectra were read with
Thermo Xcalibur 2.2 SP1.48; each mass spectra comprising 0.23 min
data collection were averaged and the mass lists, conformed by m/z and
respective intensity values, were exported and saved in ASCII format.
2.1. Synthesis of individual thiosemicarbazones
Thiosemicarbazones 1 and A7T1 were prepared by refluxing in
ethanol followed by product crystallization. 2-(thiophen-2-ylmethy-
lene)hydrazine-1-carbothioamide (1), was obtained as yellow crystals,
yield rate was 40%. 1H NMR and 13C NMR spectra were coincident with
literature (Xie et al., 2016).
2-(4-(diethylamino)-2-hydroxybenzylidene)
hydrazine-1-car-
bothioamide (A7T1) was obtained as bright yellow crystals, and the
yield rate was 82%. 1H NMR (DMSO‑d6, TMS, 300 MHz): δ (ppm) 11.07
(1H, s, NH), 9.49 (1H, s, Ar-OH), 8.19 (1H, s, CH), 7.87 (1H, s,
CH = N), 7.66 (1H, s, NH2), 7.52 (1H, d, j = 9.0 Hz, Ar-H), 6.22 (1H,
dd, j = 9.0, 2.5 Hz, Ar-H), 6.09 (1H, d, 2.5 Hz, Ar-H), 3.34 (4H, q,
j = 6.5 Hz,CH2), 1.10 (6H, t, j = 6.5 Hz, CH3). 13C NMR (DMSO‑d6,
TMS, 75 MHz): δ (ppm) 178.38 (C]S), 158.09 (Ar-OH), 150.03 (Ar-N),
142.27 (CH = N), 128.94 (CH, Ar), 107.35 (Ar-CHN), 103.91 (CH, Ar),
97.23 (CH, Ar), 43.77 (CH2), 12.51 (CH3). MS (ESI): m/z 267.12
2.4.2. Parameter setting for BioMSId algorithm
BioMSId was implemented according to Ramallo et al. (Ramallo
et al., 2015). BioMSId algorithm was run under MATLAB environment
(R2020a trial use). For spectra comparison, the tolerance parameter
value was set at 0.04 m/z and the analysed mass range was 164–490 m/
z. First, the routine control spectrum of the equipment was subtracted
from background/halo spectra. Second, the spectra were filtered at an
intensity threshold equal to 3.3 × 105 to eliminate low-intensity sig-
nals. Third, the background spectra were subtracted from halo spectra,
to filter signals that were unrelated with the library chemical compo-
nents under the inhibition halo (phosphate buffer, L-tyrosine,
[M + H]+
.
2.2. Thiosemicarbazone library preparation and analysis
Aldehydes A1-A8 and thiosemicarbazide (T1) and thiocarbazide
2

I. Cabezudo, et al.
FoodChemistry341(2021)128232
tyrosinase). Finally, all subtracted spectra were compared, searching
common signals related to the inhibitors.
post hoc Bonferroni's Multiple Comparison Test were performed on the
obtained a* and BI data using the Graph Pad Prism software. A p-value
lower than 0.05 was selected as the decision level for statistically sig-
nificant differences.
2.5. Quantification of tyrosinase inhibitory activity
2.7. Cell viability assay
Enzyme activity was continuously determined at 37 °C during
30 min following the increase in absorbance at 475 nm (ε = 3700 M−1
cm−1) determined by the formation of dopachrome (Choudhary &
Thomsen, 2001) in 96-well microplate using a VersaMax™ microplate
reader (Molecular Devices LLC, California, USA). All measurements
were performed in triplicate using L-tyrosine as substrate, kojic acid as
reference inhibitor, and Na2HPO4-NaH2PO4 as buffer medium
(100 mM, pH 6.8). Wells containing DMSO without inhibitor were used
as control of maximum enzymatic rates. All data sets were processed
using Prism V5.01 software (GraphPadSoftware Inc., La Jolla, CA,
USA).
Cell viability after treatment was determined by the 3-(4,5-di-
methylthiazol-2-yl)-2,5-diphenyltretazolium bromide (MTT) reduction
assay. Briefly, Vero cells (5.000 cells per well) were incubated in 96-
well plate in the presence of increasing concentrations of A7T1 for 48 h
(9.375 to 300 μM). Then 20 μL of MTT solution (5 mg/mL in PBS) were
added to each well and incubated for 1 h at 37 °C. After this incubation
period, MTT solution was removed and precipitated formazan was so-
lubilized in 100 μL of DMSO. Optical density (OD) was spectro-
photometrically quantified (λ = 540 nm). DMSO was used as blank and
each treatment was performed in triplicates. Values were normalized
using the blank as 100% growth. IC50 was calculated from the log
(concentration) vs. Viability (%) curves using the non-linear fit log(in-
hibitor) vs. response, Variable slope (four parameters) in GraphPad
Prism.
IC50 determinations. Serial DMSO dilutions of kojic acid, 1, and
A7T1 were prepared following equally spaced points. Wells were filled:
10 μL DMSO solution (end plate concentration ranged between 0.048
and 500 μM), and mushroom tyrosinase in buffer phosphate (15.6 U/
mL, end concentration in well). The reaction was started by addition of
L-tyrosine (0.63 mM, end concentration in well) completing 270 μL end
volume in well. The plate was shaken for 10 s, and the increase in
absorbance was immediately monitored, as explained in the previous
section. Percentage of inhibition of tyrosinase activity was calculated as
follows:
3. Results
3.1. In situ inhibition evaluation and MS analysis of a reference tyrosinase
inhibitor
%inhibition = {[(Ct C0) (tt t0)]/(Ct C0)} × 100
In order to evaluate if the TLC bioautography-MS analysis could be
applied for the identification of tyrosinase inhibitors, we used thiose-
micarbazone 1 as a positive control of the method. This potent tyr-
osinase inhibitor was reported recently by the group of Cao (Xie et al.,
2016). Compound 1 was prepared from 2-thiophenecarboxaldehyde
and thiosemicarbazide in refluxing EtOH.
where Ct is the absorbance of a DMSO solution after incubation, C0 is
the absorbance of DMSO before incubation, tt is the absorbance of
sample solution after incubation, and t0 the absorbance of sample so-
lution before incubation.
Inhibition type determination. Five serial DMSO dilutions of kojic
acid, 1, and A7T1 were prepared in equally spaced points. Wells were
filled: 10 μL DMSO solution (end plate concentration ranged between
0.18 and 16.78 μM), and mushroom tyrosinase in buffer phosphate
(15.6 U/mL, end concentration in well). The reaction was started by
addition of L-tyrosine (0.31–1.26 mM end concentration in well) com-
pleting 270 μL end volume in well. The plate was shaken for 10 s and
was immediately monitored by measuring the increase in absorbance.
The tyrosinase inhibition by 1 was evaluated using a TLC-bioauto-
graphy. In this assay format, a TLC sheet is covered by a thin layer of
agar-gel containing the enzyme tyrosinase and the substrate tyrosine
(Paula García & Furlan, 2015). Due to enzyme-catalyzed dye formation,
the gel is colored in reddish-brown, and the presence of enzyme in-
hibitors spotted on the TLC surface can be detected as clear spots. In
order to evaluate the assay sensitivity towards the model compound,
increasing amounts of thiosemicarbazone 1 were spotted on the TLC,
ranging between 0.04 μg and 0.75 μg. The observed inhibition spots
and the optical densities (475 nm, absorption wavelength for tyrosinase
products) were proportional to the amount of inhibitor up to 0.63 μg,
when apparent saturation was observed (Fig. 1).
2.6. Browning colorimetric assays
Granny Smith and Pink Lady apples and eggplants were washed and
cut into small identical slices of 5 mm width using a mandolin slicer,
which were dipped in 300 mL of 15 μM of inhibitor solution, for 3 min
and then drained. Inhibitor solutions were prepared in 20 mM phos-
phate buffer (pH 6.8) (Wu, Cheng, Li, Wang, & Ye, 2008). This proce-
dure was repeated for A7T1 and kojic acid, while control samples were
dipped only in buffer phosphate. Samples were then placed on absor-
bent paper and stored at room temperature, and changes of colour were
measured during 24 h.
To evaluate if compound 1 could be linked to the observed tyr-
osinase inhibition through MS analysis, two gel samples were taken
from the TLC bioautography, one from the inhibition zone and one from
the background, i.e., a gel area without any spots. Samples were ex-
tracted with EtOH, the solvent was evaporated, and the sample was
dissolved in ACN for MS analysis. The analysis was performed using
MatLab, to subtract the background as described in previous work from
our group and lead to a clean signal m/z = 208.00 corresponding to the
[M + Na]+ ion from 1 (Fig. S1). Altogether, the results indicated that
this type of tyrosinase inhibitors can be detected, when spotted on a
TLC plate and that the observed bioactivity could be linked to an MS
signal.
For colorimetric determinations, standardized images of apples
were obtained using CAMAG TLC visualizer. Then images were im-
ported into the Fiji software and converted to CIE L*a*b* space, where
L* indicates lightness, a* indicates chromaticity on a green (-) to red
(+) axis, and b* chromaticity on a blue (–) to yellow (+) axis. In order
to compare the apple slice samples, L*, a*, and b* coordinates were
obtained, and from them, the browning index (BI) was calculated using
the formula (Perez-Gago, Serra, Alonso, Mateos, & del Río, 2005):
3.2. Library preparation and analysis
In order to establish conditions for library preparation and analysis,
the model thiosemicarbazone 1 was now prepared in ethanol from
DMSO solutions (30 mg/mL) of 2-thiophenecarboxaldehyde and thio-
semicarbazide (T1). Product formation was followed by TLC-bioauto-
graphy analysis through both, naked eye TLC observation and TLC
scanning at 475 nm (Fig. 2). At 3 h of reaction, the product reached a
BI = [100(X 0.31)]/0.17
Where X= (a + 1.75L )/(5.645 L + a
3.012b )
For differences among control and the two inhibitor treatments
applied to apple slices, one-way analysis of variance (ANOVA) with the
3

I. Cabezudo, et al.
FoodChemistry341(2021)128232
Fig. 1. Analysis of tyrosinase inhibition produced by increasing amounts of thiosemicarbazone 1: a) tyrosinase bioautography; b) tyrosinase bioautography scan at
475 nm.
The most intense inhibition zone was observed in the intermediate
polarity region (Rf between 0.35 and 0.55), which could be indicative
of one (or more) thiocarbazone or thiosemicarbazone inhibitor present
in the library (Fig. 3c). This inhibition spot showed similar intensity to
that of the spot produced by 0.51 µg of the known inhibitor 1. Taking
into consideration the library sample applied to the TLC for analysis,
the compound responsible for tyrosinase inhibition in that zone could
have similar potency to 1.
3.4. Tyrosinase inhibitor identification using BioMSId
Identification of the library member responsible for the inhibition
spot by MS analysis of a sample from the bioautography can be more
complicated than in the previous analysis with model compound 1
because of the number of compounds present. In addition to MS signals
from bioassay related reagents, now the co-extraction of inactive library
members and starting materials with similar TLC migration properties
could increase further the complexity of the MS spectrum obtained.
Therefore, to gain insight into the identity of the compound re-
sponsible for the observed activity in the library, the BioMSId
(BIOautography coupled to Mass Spectrometry for the IDentification of
compounds) strategy was applied (Ramallo et al., 2015). To implement
BioMSId, library samples were chromatographed on TLC using three
different solvent systems, and the three TLC plates were then tested for
tyrosinase activity (Fig. S4). Samples from the inhibition zones and the
background were taken from each TLC plate and subjected to MS
analysis. The obtained spectra were processed with a MATLAB algo-
rithm that compares the spectra searching for common signals that
could be linked to the structure of the active compound (García,
Ramallo, Salazar, & Furlan, 2016; Ramallo et al., 2015).
Fig. 2. Tyrosinase activity-based detection of product A9T1 in the reaction
mixture at different times: a) tyrosinase bioautography; b) tyrosinase bioauto-
graphy scan at 475 nm.
constant concentration. Product formation could not be detected by
TLC analysis under UV light at 254 and 365 nm, indicating lower
sensitivity than the bio-based detection (Fig. S2).
The reaction conditions and building-block concentrations used for
the preparation of compound 1 were then used to prepare a library of
thiosemicarbazones and thiocarbazones, Fig. 3a. Adequate volumes of
DMSO solutions of aldehydes Ai (15 mg/mL for A3 and 30 mg/mL for
the rest), thiosemicarbazide T1 (30 mg/mL), and thiocarbazide T2
(15 mg/mL) were mixed and diluted in EtOH to a final volume of 2 mL
(Fig. 3b). The reaction mixture was stirred for 3 h under reflux and
analyzed by LC-MS, detecting the presence of 43 thiocarbazone and
thiosemicarbazone products out of the 52 possible products (Table S1).
One MS signal was linked by BioMSId to the inhibition spot, with an
m/z value of 267.12 that could correspond to the [M + H]+ signal of
compound A7T1 (Fig. 4c). This is the only library member with MW of
266, the two compounds with closest MW values are A6T2A8
([M + H]+ m/z = 261.13) and A2T1 (m/z = 270.08), undoubtedly
different from the one observed. The signals corresponding to the active
compound were picked out from a crowded zone of the spectra (Fig. 4a
and 4b). Nevertheless, they could be correctly assigned to the com-
pound responsible for the observed activity (Fig. 4c).
3.3. Bioautography detection of tyrosinase inhibitors
Tyrosinase inhibition by the library was evaluated by TLC-bioau-
tography using DCM:MeOH (97:3) as mobile phase. Mixtures of
building blocks A1-A8, and T1 and T2 were also spotted, separately, as
controls. The assay showed three potentially interesting inhibition
zones: two weak inhibition zones with Rf values between 0.02 and 0.18
and 0.72–0.88, and one more intense inhibition zone of intermediate
polarity in the Rf range 0.35–0.55 (Fig. 3c). Inhibition observed in the
more polar zone could be produced by T1 and/or T2 since a similar
inhibition was observed in the TLC lane for these building blocks (Fig.
S3). Similarly, comparison of the tyrosinase inhibition based chroma-
tograms also suggested that the least polar inhibition zone, observed at
Rf 0.72–0.88, was due to some of the aldehyde building blocks since a
similar inhibition was detected in the A1-A8 control mixture (Fig. S3).
3.5. Synthesis and inhibitory properties of compound A7T1
In order to confirm the identity and further study tyrosinase in-
hibition and anti-browning properties, compound A7T1 was prepared
in 82% yield from the aldehyde A7 and thiosemicarbazide (T1) in re-
fluxing ethanol. As expected, A7T1 showed the same MS signal ob-
served previously after BioMSId application to the library (m/
z = 267.12, Fig. S5). To corroborate that A7T1 played an important
role in the tyrosinase inhibition produced by the library, a library
4

I. Cabezudo, et al.
FoodChemistry341(2021)128232
Fig. 3. Preparation and biological analysis of a mixture library of thiosemicarbazones and thiocarbazones: a) library; b) building blocks; c) TLC-UV-tyrosinase
bioautography analysis of the library.
sample was spiked with A7T1 and evaluated by TLC-bioautography
observing a clear increase in the inhibition spot intensity in tyrosinase
bioautography (Fig. S6).
3.6. Anti-browning effect
To confirm the potential of tyrosinase inhibitory activity of A7T1 in
anti-browning applications, Granny Smith apple slices were chosen as
food models. Slices were dipped in solutions of A7T1 and kojic acid (a
reported anti-browning agent), and the enzymatic browning reaction
was measured. Readings of images were transformed to a CIELAB scale
(L*, a*, b*), and browning index (BI) was calculated. Earlier research
studies have found that a* value was more sensitive to browning in
fresh-cut apple (Rojas-Graü, Sobrino-López, Tapia, & Martín-Belloso,
2006). The browning index (BI), calculated from L*, a* and b*, is de-
fined as brown-color purity, usually used as an indicator of the extent of
browning in apple-cuts (Perez-Gago et al., 2005).
The IC50 value for A7T1 is 0.46
0.05 μM, significantly lower to
the IC50 of kojic acid (22.84
0.26 µM) and similar to that of com-
pound 1 (IC50 = 0.36
0.05 µM) determined using the same con-
ditions. The IC50 plots of A7T1, 1, and kojic acid are included in SI (Fig.
S7).
The Michaelis-Menten plot for the enzymatic reaction in the pre-
sence of different concentrations of compound A7T1 is shown in
Fig. 5a. This inhibitor presented the best fit using a mixed-type in-
hibition model, according to the Graph Pad Prism software. Compound
A7T1 binds not only to the enzyme (competing with the substrate) but
also to the substrate-enzyme complex. The affinity of the inhibitor for
Therefore, the browning of the three groups at 6 h and 24 h was
compared through a* and BI (Fig. 6a and b). Apple slices treated with
15 µM A7T1 possessed the slower browning degree than control and
15 µM kojic acid during the entire storage period. At 24 h, the mean BI
value in untreated slices almost doubles that in the slices treated with
A7T1. This effect is most pronounced in a* whose mean value in the
control quadruples that observed for A7T1 treated slices. This is con-
sistent with the results of surface color observations (Fig. 6c), and
supported statistically though one-way analysis of variance (ANOVA)
with the post hoc Bonferroni's Multiple Comparison Test, that found
significant differences between A7T1 treated slices vs control slices for
a* and BI (p < 0.05) (Tables S2 and S3 respectively). Similar browning
reducing effects were observed in slices of eggplants and Pink Lady
the free enzyme (Ki = 0.29 µM
0.11 µM) is higher than the affinity
of the inhibitor for the substrate-enzyme complex (Kis = αKi, with
α = 5.40, therefore Kis = 1.56 µM). The Lineweaver-Burk plot for
A7T1 is shown in Fig. 5b and slope and intercept of the Lineweaver-
Burk plot can be found in Supplementary material, Fig. S8.
Two aminobenzaldehyde-thiosemicarbazones have been reported as
tyrosinase inhibitors: 4-dimethylaminobenzaldehyde-thiosemicarba-
zone (IC50 = 1.54 µM, mix-type inhibitor) and 4-dimethylamino-
benzaldehyde-N-phenyl-thiosemicarbazone (IC50 = 1.78 µM, non-
competitive inhibitor) (Yang et al., 2013). Compared to A7T1, both
compounds exhibited higher IC50 values.
5

I. Cabezudo, et al.
FoodChemistry341(2021)128232
Fig. 4. BioMSId identification of A7T1 as the tyrosinase inhibitor in the library: a) raw LRMS spectrum; b) amplification of the area from which the compound is
finally identified; c) algorithm output.
apples (Fig. S9). The ANOVA and the post hoc Bonferroni's Multiple
Comparison Test found again significant differences between A7T1
treated slices vs control slices for a* and BI (p < 0.05) (Tables S4–S7).
4. Conclusions
For the first time, library-preparation, TLC bioautography, and mass
spectrometry were applied in tandem for the discovery of a bioactive
compound. The strategy led to the straightforward identification of a
new tyrosinase inhibitor with interesting inhibitory properties and anti-
browning effects that deserves further studies in order to evaluate its
applicability
3.7. Cellular assay
The cytotoxicity of compound A7T1 against Vero Cells was eval-
uated using an MTT assay. The IC50 value was 145.00
1.28 µM
Since the approach involves bioguided synthesis, milligram scale
preparation and characterization is carried out only after a bioactive
molecule has been identified. Going beyond this proof-of-principle
(Table S4 and Fig. S10). This concentration is more than 300 times
higher than the concentration required for inhibition of 50% of the
enzyme activity (0.46
0.05 µM).
Fig. 5. Tyrosinase activity assay: a) Michaelis-Menten plot: effects of different concentrations of A7T1 on the reaction rate as a function of substrate concentration; b)
Lineweaver-Burk plot for A7T1.
6

I. Cabezudo, et al.
FoodChemistry341(2021)128232
Software, Validation, Formal analysis, Data curation, Writing - review &
editing, Visualization. Victoria L. Alonso: Investigation, Formal ana-
lysis. Ricardo L.E. Furlan: Conceptualization, Writing - review &
editing, Resources, Supervision, Project administration, Funding ac-
quisition.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgements
Financial support for this work was provided by Fondo para la
Investigación Científica
y Tecnológica (Argentina, grant number
PICT2015-3574), and Universidad Nacional de Rosario (Argentina,
grant number 80020180300114UR). The authors thank the Instituto de
Química de Rosario (IQUIR, CONICET-UNR) for providing access to
NMR equipment. Ignacio Cabezudo thanks CONICET for his fellowship.
I. Ayelen Ramallo, Victoria L. Alonso, and Ricardo L. E. Furlan are
CONICET researchers.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.foodchem.2020.128232.
References
Azadniya, E., & Morlock, G. E. (2018). Bioprofiling of Salvia miltiorrhiza via planar
chromatography linked to (bio)assays, high resolution mass spectrometry and nu-
clear magnetic resonance spectroscopy. Journal of Chromatography A, 1533, 180–192.
https://doi.org/10.1016/j.chroma.2017.12.014.
Bräm, S., & Wolfram, E. (2017). Recent Advances in Effect-directed Enzyme Assays based
on Thin-layer Chromatography. Phytochemical Analysis, 28(2), 74–86. https://doi.
org/10.1002/pca.2669.
Carcelli, M., Rogolino, D., Bartoli, J., Pala, N., Compari, C., Ronda, N., ... Fisicaro, E.
(2020). Hydroxyphenyl thiosemicarbazones as inhibitors of mushroom tyrosinase
and antibrowning agents. Food Chemistry, 303, Article 125310. https://doi.org/10.
1016/j.foodchem.2019.125310.
Chen, A. Y., Adamek, R. N., Dick, B. L., Credille, C. V., Morrison, C. N., & Cohen, S. M.
(2019). Targeting Metalloenzymes for Therapeutic Intervention. Chemical Reviews,
119(2), 1323–1455. https://doi.org/10.1021/acs.chemrev.8b00201.
Ciura, K., Dziomba, S., Nowakowska, J., & Markuszewski, M. J. (2017). Thin layer
chromatography in drug discovery process. Journal of Chromatography A, 1520, 9–22.
https://doi.org/10.1016/j.chroma.2017.09.015.
García, P., & Furlan, R. L. E. (2015). Multiresponse Optimisation Applied to the
Development of a TLC Autography for the Detection of Tyrosinase Inhibitors.
Phytochemical Analysis, 26(4), 287–292. https://doi.org/10.1002/pca.2562.
García, P., Ramallo, I. A., Salazar, M. O., & Furlan, R. L. E. (2016). Chemical diversifi-
cation of essential oils, evaluation of complex mixtures and identification of a xan-
thine oxidase inhibitor. RSC Advances, 6(62), 57245–57252. https://doi.org/10.
1039/C6RA05373D.
García, P., Salazar, M. O., Ramallo, I. A., & Furlan, R. L. E. (2016). A New Fluorinated
Tyrosinase Inhibitor from a Chemically Engineered Essential Oil. ACS Combinatorial
Science, 18(6), 283–286. https://doi.org/10.1021/acscombsci.6b00004.
Hałdys, K., & Latajka, R. (2019). Thiosemicarbazones with tyrosinase inhibitory activity.
MedChemComm, 10(3), 378–389. https://doi.org/10.1039/C9MD00005D.
Hsu, K.-D., Chan, Y.-H., Chen, H.-J., Lin, S.-P., & Cheng, K.-C. (2018). Tyrosinase-based
TLC Autography for anti-melanogenic drug screening. Scientific Reports, 8(1), 401.
https://doi.org/10.1038/s41598-017-18720-0.
Fig. 6. Browning of apple slices in control samples and samples treated using
15 μM of kojic acid and A7T1: a) comparison using a*; b) comparison using BI;
c) some images used in the analysis of apple slice browning.
Jamshidi-Aidji, M., & Morlock, G. E. (2016). From Bioprofiling and Characterization to
Bioquantification of Natural Antibiotics by Direct Bioautography Linked to High-
Resolution Mass Spectrometry: Exemplarily Shown for Salvia miltiorrhiza Root.
Analytical Chemistry, 88(22), 10979–10986. https://doi.org/10.1021/acs.analchem.
6b02648.
study with anti-browning thiosemicarbazones, the effect-directed based
platform can be applied for the discovery of other bioactive compounds
relevant for food. Using appropriate high-yield simple chemical reac-
tions and existing TLC-bioautography targets, novel antioxidants, gly-
cosidase and lipase inhibitors, or antimicrobials could be identified.
Krüger, S., Hüsken, L., Fornasari, R., Scainelli, I., & Morlock, G. E. (2017). Effect-directed
fingerprints of 77 botanical extracts via a generic high-performance thin-layer
chromatography method combined with assays and mass spectrometry. Journal of
Chromatography A, 1529, 93–106. https://doi.org/10.1016/j.chroma.2017.10.068.
Mi Moon, K., Young Kim, C., Yeul Ma, J., & Lee, B. (2019). Xanthone-related compounds
as an anti-browning and antioxidant food additive. Food Chemistry, 274, 345–350.
https://doi.org/10.1016/j.foodchem.2018.08.144.
CRediT authorship contribution statement
Moon, K. M., Lee, B., Cho, W.-K., Lee, B.-S., Kim, C. Y., & Ma, J. Y. (2018). Swertiajaponin
as an anti-browning and antioxidant flavonoid. Food Chemistry, 252, 207–214.
Ignacio Cabezudo: Methodology, Validation, Investigation, Data
curation, Writing - original draft. I. Ayelen Ramallo: Methodology,
7

I. Cabezudo, et al.
FoodChemistry341(2021)128232
https://doi.org/10.1016/j.foodchem.2018.01.053.
9b00064.
Perez-Gago, M. B., Serra, M., Alonso, M., Mateos, M., & del Río, M. A. (2005). Effect of
whey protein- and hydroxypropyl methylcellulose-based edible composite coatings
on color change of fresh-cut apples. Postharvest Biology and Technology, 36(1), 77–85.
https://doi.org/10.1016/j.postharvbio.2004.10.009.
Wu, J.-J., Cheng, K.-W., Li, E. T. S., Wang, M., & Ye, W.-C. (2008). Antibrowning activity
of MRPs in enzyme and fresh-cut apple slice models. Food Chemistry, 109(2),
379–385. https://doi.org/10.1016/j.foodchem.2007.12.051.
Xie, J., Dong, H., Yu, Y., & Cao, S. (2016). Inhibitory effect of synthetic aromatic het-
erocycle thiosemicarbazone derivatives on mushroom tyrosinase: Insights from
fluorescence, 1H NMR titration and molecular docking studies. Food Chemistry, 190,
709–716. https://doi.org/10.1016/j.foodchem.2015.05.124.
Ramallo, I. A., Salazar, M. O., & Furlan, R. L. E. (2015). Thin layer chromatography-
autography-high resolution mass spectrometry analysis: Accelerating the identifica-
tion of acetylcholinesterase inhibitors. Phytochemical Analysis, 26(6), 404–412.
https://doi.org/10.1002/pca.2574.
Xu, J., Liu, J., Zhu, X., Yu, Y., & Cao, S. (2017). Novel inhibitors of tyrosinase produced by
the 4-substitution of TCT. Food Chemistry, 221, 1530–1538. https://doi.org/10.1016/
j.foodchem.2016.10.140.
Rogolino, D., Cavazzoni, A., Gatti, A., Tegoni, M., Pelosi, G., Verdolino, V., ... Carcelli, M.
(2017). Anti-proliferative effects of copper(II) complexes with hydroxyquinoline-
thiosemicarbazone ligands. European Journal of Medicinal Chemistry, 128, 140–153.
https://doi.org/10.1016/j.ejmech.2017.01.031.
Yang, M.-H., Chen, C.-M., Hu, Y.-H., Zheng, C.-Y., Li, Z.-C., Ni, L.-L., ... Chen, Q.-X.
(2013). Inhibitory kinetics of DABT and DABPT as novel tyrosinase inhibitors. Journal
of Bioscience and Bioengineering, 115(5), 514–517. https://doi.org/10.1016/j.jbiosc.
2012.11.019.
Rojas-Graü, M. A., Sobrino-López, A., Tapia, M. S., & Martín-Belloso, O. (2006). Browning
Inhibition in Fresh-cut‘Fuji’Apple Slices by Natural Antibrowning Agents. Journal of
Food Science, 71(1), S59–S65. https://doi.org/10.1111/j.1365-2621.2006.tb12407.x.
Shao, L.-L., Wang, X.-L., Chen, K., Dong, X.-W., Kong, L.-M., Zhao, D.-Y., ... Zhou, T.
(2018). Novel hydroxypyridinone derivatives containing an oxime ether moiety:
Synthesis, inhibition on mushroom tyrosinase and application in anti-browning of
fresh-cut apples. Food Chemistry, 242, 174–181. https://doi.org/10.1016/j.foodchem.
2017.09.054.
Zhang, L., Shi, J., Tang, J., Cheng, Z., Lu, X., Kong, Y., & Wu, T. (2017). Direct coupling of
thin-layer chromatography-bioautography with electrostatic field induced spray io-
nization-mass spectrometry for separation and identification of lipase inhibitors in
lotus leaves. Analytica Chimica Acta, 967, 52–58. https://doi.org/10.1016/j.aca.
2017.03.008.
Zhou, J., Tang, Q., Wu, T., & Cheng, Z. (2017). Improved TLC Bioautographic Assay for
Qualitative and Quantitative Estimation of Tyrosinase Inhibitors in Natural Products.
Phytochemical Analysis, 28(2), 115–124. https://doi.org/10.1002/pca.2666.
Solís, C. M., Salazar, M. O., Ramallo, I. A., García, P., & Furlan, R. L. (2019). A Tyrosinase
Inhibitor from a Nitrogen-Enriched Chemically Engineered Extract. ACS
Combinatorial Science, 21(9), 622–627. https://doi.org/10.1021/acscombsci.
8