Functionalization of gold nanoparticles with two aminoalcohol-based quinoxaline derivatives for targeting phosphoinositide 3-kinases (PI3Kα)

Article abstract of DOI:10.1039/c8nj04314k

Quinoxaline derivatives have attracted considerable attention due to their vast range of applications that includes electroluminescence and biomedicine. Concerning the latter, the literature has shown that compounds with a quinoxaline motif bind quite efficiently to phosphatidylinositol-4,5-bisphosphate 3-kinases (PI3Ks), which are enzymes found to be overexpressed in some types of neoplasms. In the present study, gold nanoparticles (AuNPs) were easily functionalized with 2,3-diethanolminoquinoxaline (DEQX) and 2-(2,3-dihydro-[1,4]oxazino[2,3-b]quinoxalin-4-yl)ethanol (OAQX). We made use of glycerol in alkaline media as reducing agent and the quinoxalines served as capping ligands to stabilize the AuNPs. This is the first report on the modification of a nanostructure with quinoxalines. Functionalization confers nanoparticles the required specificity to target only cancer cells, which opens possibilities for phototherapy since the modified AuNPs would concentrate in the tumor tissue as a consequence of PI3Kα overexpression. Molecular dynamics simulations have shown that DEQX and OAQX are potential inhibitors of PI3Kα since they bind to the active site of the enzyme in a way similar to other known inhibitors.

Full text of DOI:10.1039/c8nj04314k

NJC
View Article Online
View Journal
PAPER
Functionalization of gold nanoparticles with two
aminoalcohol-based quinoxaline derivatives for
targeting phosphoinositide 3-kinases (PI3Ka)†
Cite this:DOI: 10.1039/c8nj04314k
Janine Araujo,a Fabrıcio G. Menezes,a Heloiza F. O. Silva,a Davi S. Vieira,a
´
´
b
Sergio R. B. Silva,
Adailton J. Bortoluzzi,c Celso Sant’Anna,d Mateus Eugenio,d
Jannyely M. Neria and Luiz H. S. Gasparotto
*
a
Quinoxaline derivatives have attracted considerable attention due to their vast range of applications that
includes electroluminescence and biomedicine. Concerning the latter, the literature has shown that
compounds with a quinoxaline motif bind quite efficiently to phosphatidylinositol-4,5-bisphosphate
3-kinases (PI3Ks), which are enzymes found to be overexpressed in some types of neoplasms. In the present
study, gold nanoparticles (AuNPs) were easily functionalized with 2,3-diethanolminoquinoxaline (DEQX) and
2-(2,3-dihydro-[1,4]oxazino[2,3-b]quinoxalin-4-yl)ethanol (OAQX). We made use of glycerol in alkaline media
as reducing agent and the quinoxalines served as capping ligands to stabilize the AuNPs. This is the first
report on the modification of a nanostructure with quinoxalines. Functionalization confers nanoparticles
the required specificity to target only cancer cells, which opens possibilities for phototherapy since the
modified AuNPs would concentrate in the tumor tissue as a consequence of PI3Ka overexpression.
Molecular dynamics simulations have shown that DEQX and OAQX are potential inhibitors of PI3Ka
since they bind to the active site of the enzyme in a way similar to other known inhibitors.
Received 23rd August 2018,
Accepted 16th December 2018
DOI: 10.1039/c8nj04314k
rsc.li/njc
has been implicated in that severe type of neoplasm.5,6 The
clarification of the molecular mechanism of cancer signaling
1 Introduction
Cancer is a multiple-factor malady caused by accumulation of allowed the development of pharmacological inhibitors that
genetic mutations that lead to abnormal proliferation of tissues. An target specific proteins to halt or cease tumor cell progression.7
important consequence of decoding the human genome sequence As an example, monoclonal antibodies have been considerably
was the facilitation of the analysis of cancer genomes.1 This advance successful for cancer therapy in recent years,8 as they can be
was of paramount importance for locating the most frequently precisely engineered to fit a specific protein’s active site in order
mutated genes that are directly correlated to overexpressed proteins to inhibit its action. However, the production of monoclonal
upon tumor growth.2 As examples, there are E-cadherin and antibodies is a laborious and time-consuming technology.9
b-catenin, which are proteins that play important roles in cell–cell An interesting approach to circumvent these features is the
adhesion and whose overexpression is associated with colorectal design of molecules that have high affinity to the desired
carcinomas.3,4 In the case of brain tumor, atypical augmenta- protein’s site and are easily synthesized through simple
tion of phosphatidylinositol-4,5-bisphosphate 3-kinases (PI3Ks) routes.10 Quinoxaline derivatives have attracted considerable
attention due to their vast range of applications that includes
luminescence11 and biomedicine.12 Concerning the latter, com-
a Biological Chemistry and Chemometrics Research Group, Institute of Chemistry,
pounds with a quinoxaline motif are quite efficient antitumor13
and anticancer agents14,15 as they inhibit enzymes involved in
Federal University of Rio Grande do Norte, Lagoa Nova, 59072-970, Natal, RN,
Brazil. E-mail: lhgasparotto@ufrnet.br; Tel: +55 84 3342 2323
b Brain Institute, Federal University of Rio Grande do Norte, Morro Branco,
cancer proliferation, motility and differentiation. Mielcke et al.16
found that quinoxaline-derived calchones were able to inhibit
glioma cell proliferation through binding with the PI3K enzyme.
The PI3Ka enzyme, which is a subclass of PI3Ks, is regarded as
potential therapeutic target due its role in several types of
cancer.17,18 In this context, Wu et al.19,20 reported the design
and synthesis of a series of biologically active 2,3-dissubstituted
quinoxaline derivatives and also made use of computational
59056-450, Natal, RN, Brazil
c
´
ˆ
´
Laboratorio de Bioinorganica e Cristalografia, Departamento de Quımica,
´
Universidade Federal de Santa Catarina, 88040-900 Florianopolis-SC, Brazil
d Laboratory of Microscopy Applied to Life Science – Lamav, National Instituto of
Metrology, Quality and Tecnology – Inmetro, Duque de Caxias, 25250-020,
Rio de Janeiro, RJ, Brazil
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c8nj04314k
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019
New J. Chem.

View Article Online
Paper
NJC
chemistry to suggest binding modes between the PI3Ka active Covalent bonds involving hydrogen were restrained by LINCS
site and the quinoxaline derivatives.
algorithm35 while SPC water molecules geometries were con-
Nowadays, targeted therapies may be substantially enhanced strained by the SETTLE36 algorithm. Motion equations were
by bringing nanotechnology to the field, as it has been employed integrated by a leap-frog37 algorithm with time steps of 2.0 fs.
successfully in imaging,21,22 diagnosing23 and cancer treatment.24 The long-range interactions were computed using particle-
However, nanomaterials usually lack specificity and must be mesh Ewald sum, PME,38 with a cutoff of 10 Å also used for
previously functionalized in order to become a task-specific van der Waals interactions. An energy minimization step using
hybrid.25 In particular, the surface of gold nanoparticles (AuNPs) steepest descent39 was employed before the equilibration
may be modified with a variety of compounds such as proteins, phase. The properties related with structural dynamics were
DNA, and antibodies,26 to confer them the specificity required monitored by root-mean-square deviations (RMSD) using Ca
for targeted therapies.
atoms and the initial structure of the simulation (t = 0 ns) as
Herein, the functionalization and stabilization of AuNPs reference structure. The hydrogen bonds (HB) between amino
with two synthetic quinoxaline derivatives, 2,3-diethanolmino- acids residues and ligands were identified based on geometric
quinoxaline (DEQX) and 2-(2,3-dihydro-[1,4]oxazino[2,3-b]quinoxalin- criteria, the distance between acceptor and donor o 3.5 Å, and
4-yl)ethanol (OAQX) were described. To the best of our knowledge, the angle donor-hydrogen-acceptor o 301.40 The interaction
this is the first study on the functionalization of AuNPs with potential energy (IPE) for ligand-residue and ligand-water
Nj
Ni
quinoxaline-based molecules. Another point worth mentioning
is that, despite the low solubility of heterocyclic units in water,
we were able to functionalize the gold nanoparticles with both
quinoxaline-based molecules in aqueous solution, which is quite
an important aspect for medicinal purposes.27 Computational
studies suggested good affinity to the active site of the PI3Ka. In
this way, based on these preliminary results, the quinoxaline-
AuNPs conjugates are envisaged as dual-function hybrid units
that (i) may prevent the proliferation of tumor cells that are
dependent on the PI3Ka enzyme and (ii) could be suitable for
cancer phototherapy since the modified AuNPs concentrates in
the tumor tissue as a consequence of the PI3Ka overexpression.
P P
molecules can be computed according to
Eij, where Eij
i
j
is the interaction energy between a group of atoms from ligand
(i) and a group of atoms from residue j, and Ni and Nj are
the total number of atoms on molecules/residues i and j,
respectively. The cutoff for this calculation was set to 6 Å from
each ligand.
Surface electrostatic potentials were generated using the
USCF Chimera software with Adaptative Poisson–Boltzmann
Solver (APBS) interface41 and the images were produced by
VMD visualization software.42
2.2 Materials and instrumentation
All chemicals were purchased from commercial sources (Sigma-
Aldrich and Tedia) and used without further purification. Melting
2 Experimental section
´
points were measured in a Microquımica MQAPF-301 and the values
2.1 Computational methodology
were not corrected. Elemental analysis of carbon, hydrogen and
oxygen was performed in a Carlo Erba E-110 instrument. UV-vis
analyses were carried out with an Evolution 60S UV-visible spectro-
photometer (Thermo Scientific). Infrared spectra were acquired in a
283 Perkin-Elmer instrument. Nuclear magnetic resonance (NMR)
analyses were conducted in a Varian Mercury Plus-400 MHz. Trans-
mission electron microscopy (TEM) images were acquired with a FEI
Tecnai G2 Spirit BioTWIN microscope operating at 120 kV.
The association of DEQX and OAQX with the PI3Ka binding site
was first studied via docking using the Autodock Vina software.28
The tridimensional structure of PI3Ka enzyme was obtained
from a 2.8 Å resolution X-ray structure (PDB entry 3HHM).29
The structure of PI3Ka enzyme was prepared for docking
with AutoDockTools 1.5.6 (ADT).30 The three-dimensional struc-
tures of DEQX and OAQX were built with Avogadro software.31
A 25 Â 30 Â 30 Å3 grid was centered near the center of the binding
pocket, and the exhaustiveness parameter was set to 32.
2.3 Synthesis and spectroscopic characterization of DEQX
and OAQX
Molecular dynamics (MD) simulations were then performed
with the structures from docking procedures. The minimum
distance of the quinoxaline-PI3Ka complex to the box wall was
2.3.1 2,3-Dichloroquinoxaline (DCQX). This starting material
at least 12 Å, hence the dimensions of the simulation box was prepared by a two-step protocol according to.43 The protocol
were chosen accordingly. The systems were neutralized with involves the acid catalyzed cyclization of o-phenylenediamine and
sodium and chloride ions until reaching the concentration of oxalic acid to generate 1,4-dihydroquinoxaline-2,3-dione, which
0.15 mol LÀ1. The AMBER ff03 force field32 was employed with reacts with thionyl chloride to afford the starting material DCQX
the with the GROMACS 5.0 software package. The topology and at 79% yield.
parameters for the ligands were generated with the General
2.3.2 2,3-Diethanolminoquinoxaline (DEQX). A solution of
Amber Force Fields with structure and charges calculated using DCQX (0.50 g, 2.5 mmol) in ethanolamine (5 ml) was kept at
HF/6-31G(d) as the level of theory and RESP fit.33 All quantum 100 1C for three hours. Afterwards, 25 ml of cold water were
mechanics calculations were performed using GAMESS. MD simula- added and the final aqueous solution was maintained in ice
tions were carried out at neutral pH in NVT ensemble at a tempera- bath for 8 h. The resulting solid was filtered and recrystallized
ture of 310 K controlled by the V-rescale thermostat algorithm.34 from water to give the pure compound.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
NJC
Paper
Mp 179 1C (Lit. 175–177 1C44). IR (KBr) nmax (cmÀ1): 3413, comprised thirty two neighboring residues, the results revealed
3199, 2974, 2921, 1555, 1514, 1465, 1047, 765. 1H NMR that the following twelve interacted with both quinoxalines only
(400 MHz, CDCl3) d (ppm) 7.38 (m, 2H), 7.15 (m, 2H), 7.03 slightly, with energies no lower than À0.1 kcal molÀ1: Ile771,
(t, J = 5.0 Hz), 4.83 (t, J = 5.4 Hz), 3.66 (m, 4H), 3.56 (m, 4H). Leu779, Glu798, Leu807, Asp810, Leu814, Gly837, Cys838, Ser919,
13C NMR (100 MHz, DMSO-D6) d (ppm) 143.8, 136.4, 124.5, Lys924, His931, and Asp933. The other twenty residues are listed
123.2, 59.4, 43.6. Elemental analysis for C12H16N4O2: C, 58.05; in Fig. 1A with their respective IPE for both quinoxalines. Con-
H, 6.50; N, 22.57; found: C, 58.12; H, 6.63; N, 22.68. Yield: 85%. sidering an arbitrary cutoff of À1.0 kcal molÀ1, only six residues
2.3.3 2-(2,3-Dihydro-[1,4]oxazino[2,3-b]quinoxalin-4-
yl)ethanol (OAQX). Prepared similarly to DETQ, except by the for both quinoxalines. Gln859, Val850, Ser773 are hotspots for
use of diethanolamine instead ethanolamine. DEQX, while Asn853 and Trp780 are hotspots for OAQX. On the
(Ile932, Met922, Ser854, Val851, Ile800, Met772) are hotspots
Mp 160 1C (Lit. 159–161 1C45). IR (KBr) nmax (cmÀ1): 3284, 2992, other hand, nine residues (Phe930, Thr856, His855, Arg852,
2862, 1658, 1566, 1456, 1342, 1058, 771. 1H NMR (400 MHz, Glu849, Ile848, Tyr836, Ser774 and Pro778) interact with neither
CDCl3) d (ppm) 7.54–7.51 (m, 2H), 7.40–7.37 (m, 1H), 7.29–2.26 quinoxalines. Fig. 1B and C present a visualization of the local
(m, 1H), 4.83 (t, J = 5.2 Hz, 1H), 4.47–4.45 (m, 2H), 3.77–3.71 structure of the hotspots for (A) DEQX and (B) OAQX.
(m, 6H). 13C NMR (100 MHz, DMSO-D6) d (ppm) 147.6, 142.8,
Val851 and Ser854 deserve further attention due to their high
139.2, 134.9, 127.7, 126.1, 124.7, 123.8, 64.3, 60.0, 49.9, 46.1. quinoxaline-P3Kia IPE (Fig. 1A). In the case of Val851, the more
Elemental analysis for C12H13N3O2: C, 62.33; H, 5.67; N, 18.17; attractive IPE is due to the hydrogen bond between its amide
found: C, 62.40; H, 5.80; N, 18.26. Yield: 81%.
backbone moiety and the hydroxyl group of the quinoxalines. This
type of interaction is reported to be crucial to kinase-inhibitor
complexes.48,49 Similarly, the Ser854 residue has an amide back-
bone and a side-chain hydroxyl conferring it high potential to
form hydrogen bonds with ligands.19 The short distance of these
hydrogen bonds, (around 2.22 Å and 2.05 Å for DEQX and OAQX,
respectively) and the high occurrence frequency (100% and
around 89.5% for DEQX and OAQX, respectively) strongly suggests
that the Ser854 is also a key residue for the specific inhibition
process. In both cases, the intermolecular hydrogen bonds
are also favored by the flexibility of 2-hydroxyethyl group
(–CH2CH2OH) of DEQX and OAQX, respectively.
Although Ile932 is not as energetic as Val851 and Ser854, it
qualifies as a hotspot due to its non polar nature that allows for
hydrophobic interactions with the aromatic ring portion of the
quinoxalines. Furthermore, the absence of a Ile932 residue would
lead to total loss of enzyme activity.49 Val850, Ile848, Ile800,
Met922, Phe930, and Ile932 residues are constituents of ATP
binding site interacting as hotspots via hydrophobic contacts with
DEQX and OAQX. The other residues presented in the Fig. 1A play
a collaborative role as secondary hotspots. Concerning electrostatic
2.4 Preparation of quinoxaline-functionalyzed AuNPs
AuNPs were prepared through a previously reported method46,47
that was slightly modified in this study. In the method, glycerol in
alkaline medium is used as reducing agent and the modification
consisted in replacing polyvinylpirrolidole (the stabilizing agent)
with the quinoxalines synthesized in this work. The concentra-
tions of gold ions, glycerol and NaOH were kept at 0.5 mmol LÀ1
,
0.10 mol LÀ1 and 0.10 mol LÀ1, respectively. The amounts
of DEQX and OAQX varied as such to generate the following
Au3+/quinoxaline molar ratios: 1 : 1, 1 : 0.5, 1 : 0.2, 1 : 0.1, 1 : 0.05,
1 : 0.01, and 1 : 0.005. The reaction was carried out by adding
a 5 mL solution of glycerol + NaOH into another 5 mL solution
that contained Au3+ and a quinoxaline. The final mixtures had
red-to-violet colors characteristic of gold nanoparticles.
3 Results and discussion
3.1 Molecular modeling
PI3Ka are well established enzymes as promising targets for interactions, it is known that they may guide binding processes in
cancer therapy in humans. Small molecules based on quinoxaline biological environments.41 In Fig. 1D and E, electrostatic surfaces
moiety have their anti-tumor activity hypothesized in terms of of the P3KIa in the presence of DEQX and OAQX clearly show that
binding to P3Kia.19 It is suggested that hydrogen bonds between the hydroxyl group is highly adaptive in biological environment
oxygen atoms of amine moieties and Val851 and Ser854 residues due to its amphoteric functionality.
are crucial to the activity, with some other possible interactions
MD simulations have shown that DEQX and OAQX are
involving arylsulfonyl units. The present investigation provides an potential inhibitors of PI3Ka since they bind to the active site
analysis initially based on molecular docking models between of the enzyme in a way similar to other known inhibitors.48–50
PI3Ka enzyme and DEQX and OAQX molecules and then the The correct protein–ligand binding is designed by long range inter-
best structures from docking were used for MD simulations. actions optimizing hydrogen bonds and charge complementarity.
Since host–guest intermolecular interactions could be a primary Approximately 67% of the protein–ligand (small organic mole-
condition for a successfully structure-based drug design, we have cules) complexes have hydrogen bonds mediating recognition
analyzed the binding site of the receptor by interaction potential through NH (backbone) groups of a residue and oxygen atoms
energy (IPE) from MD simulations to obtain the individual con- of the ligand.51
tribution of each amino acid residue, ranking the most relevant
interactions among the molecules.
3.2 Synthesis and characterization of the target quinoxalines
The 6 Å sphere centered on each quinoxaline molecule Classically, quinoxaline cores and their derivatives are
was used for the calculation of IPE. Although such a sphere synthesized through reactions of aromatic o-diamines and
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019
New J. Chem.

View Article Online
Paper
NJC
Fig. 1 (A) Interaction potential energy (IPE) for each residue. The values are shown for DEQX (dashed bars) and OAQX (filled bars). The dashed red line
indicates the arbitrary limit for the binding process. Structural representation of (B) DEQX and (C) OAQX inside the binding pocket of PI3Ka. Water
molecules not shown. Electrostatic surface representation of the binding pocket of PI3Ka containing (D) DEQX and (E) OAQX.
a-dialdehydes/a-diketones in the presence of a catalyst.52 Other The DEQX and OAQX molecules were fully characterized through
1
synthetic routes that involve reaction of DCQX with nucleophilic elemental analysis, infrared spectroscopy, NMR and 13C NMR
species are also available in literature. The structural features spectroscopies. Also, OAQX had its structure confirmed by X-ray
of this substrate allow the generation of innumerable 2- and/or crystallography (Fig. 2 – inset), as shown in ESI.† Both synthetic
2,3-substituted quinoxalines, including biologically relevant quinoxaline derivatives were soluble enough in water to allow
compounds such as those that present anti-cancer activity due the synthesis of AuNPs in aqueous medium.
to interactions with PI3Ka.19,20 In this study, DEQX and OAQX
3.3 Production of DEQX-AuNPs and OAQX-AuNPs
were obtained from reactions of DCQX43 with ethanolamine and
diethanolamine, respectively (Fig. 2). DEQX originated from In this study, Au3+ ions were reduced by glycerol in alkaline
the reaction of DCQX with two equivalents of ethanolamine medium at room temperature. As it is biodegradable under
via double N-nucleophilic attack, while OAQX was generated aerobic conditions, non-toxic, and relatively inexpensive,
by reacting one equivalent of diethanolamine via cyclization glycerol is a ‘‘greener’’ option for generating nanoparticles when
processes involving simultaneous N- and O-nucleophilic attacks. compared to common reducing chemicals such as formamide,
Fig. 2 Synthetic protocol of DEQX and OAQX from DCQX – conditions: ethanolamine or diethanolamine, 100 1C, 3 h. Inset: ORTEP illustration of OAQX
determined by X-ray crystallography.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
NJC
Paper
sodium borohydride and hydrazine.53 It is well known that a decreased with the augmentation of DEQX concentration. In
suitable stabilizing agent is also required in order to control contrast, the concentration of OAQX impacted only slightly on
size and aggregation of nanoparticles in solution.54 To this end, the amount of AuNPs freshly produced, as evidenced by similar
we investigated the ability of DEQX and OAQX to stabilize gold level of absorbances regardless of the Au3+/OAQX molar ratio
nanoparticles. In this way, not only would the AuNPs be (Fig. 3B). This may be explained by the formation of a relatively
appropriately available in solution, but they would also be stable complex between gold ions and DEQX that competes
unquestionably functionalized with the quinoxalines employed with the reduction of Au3+. In an experiment devised to check
in this study. Fig. 3 presents UV-vis spectra of AuNPs capped on this hypothesis, a Au3+ solution was mixed with another of
with DEQX (A) and OAQX (B) acquired at different Au3+/qui- DEQX in absence of glycerol and NaOH and UV-vis spectra were
noxaline molar ratios. The Au3+ concentration was fixed at recorded at different times as presented in Fig. 4A. The shoulder at
0.50 mmol LÀ1. Upon increasing the amount of quinoxalines, 410 nm that arises with time is attributed to a gradual formation of a
the UV-vis spectra tended to acquire the shape typically related yellow-to-orange complex between gold ions and DEQX. Khan et al.56
to spherical AuNPs,46 as seen in Fig. 3A and B. The shape of reported the formation of a stable yellow-orange complex from
AuNPs will be confirmed later by TEM. The signal observed in the the reaction between cetyltrimethylammonium bromide (CTAB)
UV-vis experiments reflects the surface plasmon resonance (SPB) and gold ions. In their study the same shoulder at 410 nm evolved
of oscillating electrons on the surface of the nanoparticles.55 An over time upon mixing CTAB with gold ions. Furthermore, a yellow
important difference between DEQX and OAQX is that, compara- coordination complex has been recently isolated from the reaction
tively, a much lower concentration of the former is sufficient to of KAuCl4 and quinoxaline,57 which is a further evidence that Au3+
generate AuNPs. A DEQX concentration of 25 mmol LÀ1 (which is prone to complexation with at least some types of quinoxalines.
corresponds to the molar ratio of 1 : 0.05) is sufficient for the On the other hand, similar behaviour is not observed for OAQX at
generation of AuNPs, while at least 100 mmol LÀ1 of OAQX is the same interval. Upon mixing Au3+ with OAQX, a slight intensi-
required for the same purpose. Based on these results, DEQX fication of the pale yellow is observed only after 30 min, which
was more efficient than OAQX for the formation of AuNPs.
Another important feature worth mentioning is the relative
might suggest weak complexation to some degree.
It is important to note that after 24 h the absorbance of the
amount of AuNPs generated with DEQX and OAQX. In Fig. 3A, DEQX-stabilized AuNPs solutions increased for Au3+/DEQX
the lower absorbance over the Au3+/DEQX molar ratio range molar ratios between 1 : 0.05 and 1 : 0.5 (Fig. 5A), meaning that
of 1 : 0.005–1 : 0.5 indicates that the amount of nanoparticles in the 24 hour interval the Au3+-DEQX complex slowly released
Fig. 3 UV-vis spectra of AuNPs stabilized with (A) DEQX and (B) OAQX at different Au3+/quinoxalines molar ratios.
Fig. 4 UV-vis spectra of solutions containing (A) Au3+ + DEQX and (B) Au3+ + OAQX acquired at different times. Glycerol and NaOH are absent. In both
cases the concentrations of Au3+ and quinoxalines were 0.50 mmol LÀ1 and 0.25 mmol LÀ1, respectively.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019
New J. Chem.

View Article Online
Paper
NJC
Fig. 5 UV-vis spectra of AuNPs stabilized with (A) DEQX and (B) OAQX at different Au3+/quinoxalines molar ratios. Spectra were acquired 24 h after the synthesis.
gold ions for further formation of AuNPs. Fig. 5A also shows that AuNPs. A closer inspection of Fig. 5B reveals that there was a
Au3+/DEQX molar ratios of 1 : 0.01 and 1 : 0.005 were not suffi- decrease in absorbance for Au3+/OAQX molar ratios other than
cient for stabilization of AuNPs. On the other hand, absorbances 1 : 1, which implies that only in the latter condition stable AuNPs
from OAQX-capped AuNPs (Fig. 5B) varied only marginally, could be obtained with OAQX as capping agent.
which corroborates the hypothesis that OAQX binds weakly to
Fig. 6 presents TEM images of AuNPs capped with DEQX (A)
Au3+ and, therefore, gold ions were promptly available to form and OAQX (B) at Au3+/quinoxaline molar ratios of 1 : 0.2 and
Fig. 6 TEM images of AuNPs capped with (A) DEQX and (B) OAQX synthesized at Au3+/quinoxaline molar fractions of 1 : 0.2 and 1 : 1, respectively.
Circularity analysis of (C) DEQX-capped AuNPs and (D) OAQX-capped AuNPs.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
NJC
Paper
Fig. 7 FTIR in ATR mode of (A) DEQX-adsorbed AuNPs and (B) OAQX-adsorbed AuNPS. The spectra of the pure quinoxalines are also shown for
comparison.
1 : 1, respectively. It is important to note that while the histo- quinoxaline were spherical in shape, slight differences in the
gram of DEQX-stabilized AuNPs is centered at around 8.0 nm, proportion of non-spherical AuNPs were observed between
the one related to OAQX-stabilized AuNPs is shifted to 11.5 nm, samples. In the range 0.70–0.79, OAQX and DEQX delivered
meaning that even at a lower concentration DEQX is more 20% and 8.7% of non-spherical nanoparticles, respectively.
efficient than OAQX in delivering smaller nanoparticles. Fig. 6C Therefore, not only is DEQX more efficient in delivering smaller
and D depict circularity analysis of the AuNPs produced in this nanoparticles, it provides more homogeneous ones in terms of
study. A circularity value of 1 means a perfectly symmetrical morphology as well.
circle and values equal or below 0.90 denote non-spherical
Fig. 7 presents FTIR spectra of DEQX, OAQX the respective
morphologies. While over 60% of AuNPs produced by either quinoxaline-adsorbed AuNPs in ATR mode. Differences in the
Fig. 8 Proposed mechanism for the formation of AuNPs with (A) DEQX and (B) OAQX.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019
New J. Chem.

View Article Online
Paper
NJC
range of 1450–1600 cmÀ1, which is related to CQN, CQC and the nanoparticle surface. In addition, there is no need of
N–H bonds, suggest that DEQX (Fig. 7A) binds to the AuNPs employing a further stabilizing agent (e.g.: polymer) since both
surface through nitrogen atoms. The bands at 1039 cmÀ1 (C–O quinoxalines may act as stabilizers at appropriate concentrations.
from alcohol) and 1053 cmÀ1 (C–N from amine) for pure DEQX MD simulations indicated that DEQX and OAQX bind to PI3Ka
become broader when the latter is adsorbed onto the AuNPs, through a variety of amino-acid residues, a feature that makes
suggesting that those functional groups are involved in the those quinoxalines eligible to inhibit PI3Ka enzymes. The combi-
interaction with gold. In the case of OAQX (Fig. 7B), the nation of synthetic organic chemistry with nanoscience and MD
spectrum for functionalized AuNPs is far less defined than that simulations is an interesting approach to conceiving new
for DEQX-AuNPs. This might be an indicative that OAQX weakly potential platforms for cancer diagnostic and treatment.
binds to AuNPs and, therefore, is removed from the metallic
surface during the centrifugation process. UV-vis and TEM
results already showed that OAQX is less efficient than DEQX Conflicts of interest
in stabilizing AuNPs. The general features for the spectrum are
similar to the DEQX, including the region comprising CQN
There are no conflicts to declare.
and CQC that becomes broader when OAQX is adsorbed onto
AuNPs. Considerable alterations in bands associated to the
ether group at 1056 cmÀ1 and 1203 cmÀ1 indicated that the
oxygen of the saturated cycle also contributes to the interaction
with the gold surface. For both quinoxaline-adsorbed AuNPs,
the band at 1039 cmÀ1 and 1022 cmÀ1 for DEQX and OAQX,
respectively, related to the alcoholic group remains at the same
frequency, suggesting that it is not involved in the ligand-
metal coordination. In both spectra, absorptions due to water
bending at approximately 1635 cmÀ1 denote that AuNPs are
hydrated to some extent.58,59
Acknowledgements
The authors gratefully acknowledge CNPq, FAPERN, CAPES
and FINEP for the financial support of this work. This research
was supported by High Performance Computing Center at
NPAD/UFRN.
References
1 D. A. Wheeler and L. Wang, Genome Res., 2013, 23,
1054–1062.
3.4 Mechanism of AuNPs formation
2 R. T. Dorsam and J. S. Gutkind, Nat. Rev. Cancer, 2007, 7,
79–94.
3 E. Tsanou, D. Peschos, A. Batistatou, A. Charalabopoulos and
K. Charalabopoulos, Anticancer Res., 2008, 28, 3815–3826.
In light of the results from UV-Vis and IR-ATR, we propose the
following mechanism for the formation of AuNPs stabilized
by the quinoxalines produced in the current study (Fig. 8):
firstly, the quinoxaline derivatives bind to Au3+ generating a
coordination complex that releases gold ions (quickly or slowly
depending on the quinoxaline) which are then reduced by
glycerol to form AuNPs. At this point, the strong alkaline
medium induces deprotonation of DEQX, which, in turn, binds
to the gold surface more strongly than OAQX since the anionic
NQC–NÀ atomic arrangement of the former is more basic than
the NQC–O system of the latter.
Molecular simulation studies have investigated the molecular
interactions between peptides and Au-Pd nanoparticles60 and the
stabilization effect of organic molecules on Au nanoparticles.61
Heinz and coworkres60 have demonstrated that short peptides (up
to 12 residues) prefer to bind in vacant sites on fcc lattice of the Au
surface molecules present structural/chemical requirements to bind
preferably in vacant sites on Au fcc lattice at the {111} surface.
Concerning the orientation of the molecule on the gold surface, for
both systems, the binding of quinoxalines to the gold surface is
more likely to take place perpendicularly, which is consistent with
previously reports involving pyridine derivatives where vertical
adsorption is more energetically favorable than flat adsorption62,63
ˇ
´
ˇ
´
4 S. Kapitanovic, T. Cacev, M. Antica, M. Kralj, G. Cavric,
´
K. Pavelic and R. Spaventi, Exp. Mol. Pathol., 2006, 80,
91–96.
5 F. E. Bleeker, S. Lamba, C. Zanon, R. J. Molenaar, T. J. M.
Hulsebos, D. Troost, A. A. van Tilborg, W. P. Vandertop,
S. Leenstra, C. J. F. van Noorden and A. Bardelli, BMC
Cancer, 2014, 14, 1–12.
6 F. E. Bleeker, R. J. Molenaar and S. Leenstra, J. Neurooncol.,
2012, 108, 11–27.
7 P. J. Roberts and C. J. Der, Oncogene, 2007, 26, 3291–3310.
8 L. M. Weiner, R. Surana and S. Wang, Nat. Rev. Immunol.,
2010, 10, 317–327.
9 J. K. H. Liu, Ann. Med. Surg., 2014, 3, 113–116.
10 D.-F. Shi, T. D. Bradshaw, S. Wrigley, C. J. McCall,
P. Lelieveld, I. Fichtner and M. F. G. Stevens, J. Med. Chem.,
1996, 39, 3375–3384.
11 A. J. Hallett, B. M. Kariuki and S. J. A. Pope, Dalton Trans.,
2011, 40, 9474–9481.
12 P. Marques-Gallego, M. A. Gamiz-Gonzalez, F. R. Fortea-Perez,
M. Lutz, A. L. Spek, A. Pevec, B. Kozlevcar and J. Reedijk,
Dalton Trans., 2010, 39, 5152–5158.
13 M. N. Noolvi, H. M. Patel, V. Bhardwaj and A. Chauhan,
Eur. J. Med. Chem., 2011, 46, 2327–2346.
4 Conclusion
We have shown that AuNPs can be easily functionalized with 14 S.-H. Lee, N. Kim, S.-J. Kim, J. Song, Y.-D. Gong and
OAQX and DEQX, with the latter adsorbing more strongly onto S.-Y. Kim, J. Cancer Res. Clin. Oncol., 2013, 139, 1279–1294.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
NJC
Paper
15 V. Desplat, M. Vincenzi, R. Lucas, S. Moreau, 38 T. Darden, D. York and L. Pedersen, J. Chem. Phys., 1993, 98,
S. Savrimoutou, N. Pinaud, J. Lesbordes, E. Peyrilles, 10089–10092.
M. Marchivie, S. Routier, P. Sonnet, F. Rossi, L. Ronga 39 G. Arfken, Mathematical Methods for Physicists, Academic
and J. Guillon, Eur. J. Med. Chem., 2016, 113, 214–227. Press, Orlando, FL, 3rd edn, 1985.
16 T. R. Mielcke, A. Mascarello, E. Filippi-Chiela, R. F. Zanin, 40 D. van der Spoel, P. J. van Maaren, P. Larsson and
ˆ
N. Tımneanu, J. Phys. Chem. B, 2006, 110, 4393–4398.
G. Lenz, P. C. Leal, L. D. Chirardia, R. A. Yunes, R. J. Nunes,
A. M. O. Battastini, F. B. Morrone and M. M. Campos, Eur. 41 N. A. Baker, D. Sept, S. Joseph, M. J. Holst and J. A.
J. Med. Chem., 2012, 48, 255–264.
McCammon, Proc. Natl. Acad. Sci. U. S. A., 2001, 98,
17 S. I. Chiosea, J. R. Grandis, V. W. Y. Lui, B. Diergaarde,
10037–10041.
J. H. Maxwell, R. L. Ferris, S. W. Kim, A. Luvison, M. Miller 42 W. Humphrey, A. Dalke and K. Schulten, J. Mol. Graphics,
and M. N. Nikiforova, BMC Cancer, 2013, 13, 1–7. 1996, 14(33–38), 27–38.
18 Y. Samuels, Z. Wang, A. Bardelli, N. Silliman, J. Ptak, 43 F. da Silva Miranda, A. M. Signori, J. Vicente, B. de Souza,
S. Szabo, H. Yan, A. Gazdar, S. M. Powell, G. J. Riggins,
J. K. V. Willson, S. Markowitz, K. W. Kinzler, B. Vogelstein
and V. E. Velculescu, Science, 2004, 304, 554.
J. P. Priebe, B. Szpoganicz, N. S. Gonçalves and A. Neves,
Tetrahedron, 2008, 64, 5410–5415.
44 D. E. Ames and M. I. Brohi, J. Chem. Soc., Perkin Trans. 1,
1980, 1384–1389.
45 C. A. Obafemi, W. Pfleiderer and F. Taiwo, Molbank, 2006,
2006, M508.
19 P. Wu, Y. Su, X. Liu, B. Yang, Q. He and Y. Hu, Bioorg. Med.
Chem., 2012, 20, 2837–2844.
20 P. Wu, Y. Su, X. Liu, L. Zhang, Y. Ye, J. Xu, S. Weng, Y. Li,
T. Liu, S. Huang, B. Yang, Q. He and Y. Hu, Eur. J. Med. 46 L. H. S. Gasparotto, A. C. Garcia, J. F. Gomes and
Chem., 2011, 46, 5540–5548. G. Tremiliosi-Filho, J. Power Sources, 2012, 218, 73–78.
21 A. M. Maley, Y. Terada, S. Onogi, K. J. Shea, Y. Miura and 47 A. G. Garcia, P. P. Lopes, J. F. Gomes, C. Pires, E. B. Ferreira,
R. M. Corn, J. Phys. Chem. C, 2016, 120, 16843–16849.
R. G. M. Lucena, L. H. S. Gasparotto and G. Tremiliosi-Filho,
22 K. M. G. Lima, R. F. A. Junior, A. A. Araujo, A. L. C. S.
New J. Chem., 2014, 38, 2865–2873.
L. Oliveira and L. H. S. Gasparotto, Sens. Actuators, B, 2014, 48 Z. A. Knight, B. Gonzalez, M. E. Feldman, E. R. Zunder,
196, 306–313.
D. D. Goldenberg, O. Williams, R. Loewith, D. Stokoe,
A. Balla, B. Toth, T. Balla, W. A. Weiss, R. L. Williams and
K. M. Shokat, Cell, 2006, 125, 733–747.
23 I. N. Brigger, C. Dubernet and P. Couvreur, Adv. Drug
Delivery Rev., 2012, 64(supplement), 24–36.
24 X. Huang, I. H. El-Sayed, W. Qian and M. A. El-Sayed, J. Am. 49 E. H. Walker, M. E. Pacold, O. Perisic, L. Stephens,
Chem. Soc., 2006, 128, 2115–2120.
25 R. A. Sperling and W. J. Parak, Philos. Trans. R. Soc., A, 2010,
368, 1333–1383.
P. T. Hawkins, M. P. Wymann and R. L. Williams, Mol. Cell,
2000, 6, 909–919.
50 R. Marone, V. Cmiljanovic, B. Giese and M. P. Wymann,
Biochim. Biophys. Acta, 2008, 1784, 159–185.
26 M. C. Daniel and D. Astruc, Chem. Rev., 2004, 104, 293–346.
27 K. T. Savjani, A. K. Gajjar and J. K. Savjani, ISRN Pharm., 51 K. Chen and L. Kurgan, PLoS One, 2009, 4, e4473.
2012, 2012, 195727.
52 M. M. Heravi, S. Taheri, K. Bakhtiari and H. A. Oskooie,
28 O. Trott and A. J. Olson, J. Comput. Chem., 2010, 31, 455–461.
Catal. Commun., 2007, 8, 211–214.
29 D. Mandelker, S. B. Gabelli, O. Schmidt-Kittler, J. Zhu, I. Cheong, 53 J. F. Gomes, L. H. S. Gasparotto and G. Tremiliosi-Filho,
C.-H. Huang, K. W. Kinzler, B. Vogelstein and L. M. Amzel, Proc.
Natl. Acad. Sci. U. S. A., 2009, 106, 16996–17001.
30 G. M. Morris, R. Huey, W. Lindstrom, M. F. Sanner,
Phys. Chem. Chem. Phys., 2013, 15, 10339–10349.
54 E. B. Ferreira, J. F. Gomes, G. Tremiliosi-Filho and
L. H. S. Gasparotto, Mater. Res. Bull., 2014, 55, 131–136.
R. K. Belew, D. S. Goodsell and A. J. Olson, J. Comput. Chem., 55 I. H. El-Sayed, X. Huang and M. A. El-Sayed, Nano Lett.,
2009, 30, 2785–2791. 2005, 5, 829–834.
31 M. D. Hanwell, D. E. Curtis, D. C. Lonie, T. Vandermeersch, 56 M. N. Khan, T. A. Khan, S. A. Al-Thabaiti and Z. Khan,
E. Zurek and G. R. Hutchison, J. Cheminf., 2012, 4, 1–17. Spectrochim. Acta, Part A, 2015, 149, 889–897.
ˇ ´
´
32 Y. Duan, C. Wu, S. Chowdhury, M. C. Lee, G. Xiong, W. Zhang, 57 B. Ð. Glisic, B. War˙zajtis, N. S. Radulovic, U. Rychlewska and
R. Yang, P. Cieplak, R. Luo, T. Lee, J. Caldwell, J. Wang and
P. Kollman, J. Comput. Chem., 2003, 24, 1999–2012.
33 C. I. Bayly, P. Cieplak, W. Cornell and P. A. Kollman, J. Phys.
Chem., 1993, 97, 10269–10280.
34 G. Bussi, D. Donadio and M. Parrinello, J. Chem. Phys., 2007,
126, 014101.
M. I. Djuran, Polyhedron, 2015, 87, 208–214.
´
´
´
58 A. I. Lopez-Lorente, M. Sieger, M. Valcarcel and B. Mizaikoff,
Anal. Chem., 2014, 86, 783–789.
59 J.-W. Park and J. S. Shumaker-Parry, J. Am. Chem. Soc., 2014,
136, 1907–1921.
60 H. H. Heinz, B. L. Farmer, R. B. Pandey, J. M. Slocik,
S. S. Patnaik, R. Pachter and R. R. Naik, J. Am. Chem. Soc.,
2009, 131, 9704–9714.
35 B. Hess, H. Bekker, H. J. C. Berendsen and J. G. E. M. Fraaije,
J. Comput. Chem., 1997, 18, 1463–1472.
36 S. Miyamoto and P. A. Kollman, J. Comput. Chem., 1992, 13, 61 E. L. da Rocha, G. F. Caramori and C. R. Rambo, Phys. Chem.
952–962. Chem. Phys., 2013, 15, 2282–2290.
37 W. F. Van Gunsteren and H. J. C. Berendsen, Mol. Simul., 62 K. Tonigold and A. Groß, J. Chem. Phys., 2015, 132, 224701.
1988, 1, 173–185. 63 V. J. Gandubert and B. Lennox, Langmuir, 2005, 21, 6532.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019
New J. Chem.