Molecular modeling and LC–MS-based metabolomics of a glutamine-valproic acid (Gln-VPA) derivative on HeLa cells

Article abstract of DOI:10.1007/s11030-020-10089-z

Abstract: Glutaminase plays an important role in carcinogenesis and cancer cell growth. This biological target is interesting against cancer cells. Therefore, in this work, in silico [docking and molecular dynamics (MD) simulations] and in vitro methods (antiproliferative and LC–MS metabolomics) were employed to assay a hybrid compound derived from glutamine and valproic acid (Gln-VPA), which was compared with 6-diazo-5-oxo-l-norleucine (DON, a glutaminase inhibitor) and VPA (contained in Gln-VPA structure). Docking results from some snapshots retrieved from MD simulations show that glutaminase recognized Gln-VPA and DON. Additionally, Gln-VPA showed antiproliferative effects in HeLa cells and inhibited glutaminase activity. Finally, the LC–MS-based metabolomics studies on HeLa cells treated with either Gln-VPA (IC₆₀ = 8?mM) or DON (IC₅₀ = 3.5?mM) show different metabolomics behaviors, suggesting that they modulate different biological targets of the cell death mechanism. In conclusion, Gln-VPA is capable of interfering with more than one pharmacological target of cancer, making it an interesting drug that can be used to avoid multitherapy of classic anticancer drugs. Graphic abstract: [Figure not available: see fulltext.].

Full text of DOI:10.1007/s11030-020-10089-z

Molecular Diversity
https://doi.org/10.1007/s11030-020-10089-z
ORIGINAL ARTICLE
Molecular modeling and LC–MS‑based metabolomics
of a glutamine‑valproic acid (Gln‑VPA) derivative on HeLa cells
M. J. Fragoso‑Vázquez1,2 · D. Méndez‑Luna2 · M. C. Rosales‑Hernández2 · G. R. Luna‑Palencia2 · A. Estrada‑Pérez2 ·
Benedicte Fromager3 · I. Vásquez‑Moctezuma4 · J. Correa‑Basurto2
Received: 7 January 2020 / Accepted: 11 April 2020
© Springer Nature Switzerland AG 2020
Abstract
Glutaminase plays an important role in carcinogenesis and cancer cell growth. This biological target is interesting against
cancer cells. Therefore, in this work, in silico [docking and molecular dynamics (MD) simulations] and in vitro methods (anti-
proliferative and LC–MS metabolomics) were employed to assay a hybrid compound derived from glutamine and valproic
acid (Gln-VPA), which was compared with 6-diazo-5-oxo-l-norleucine (DON, a glutaminase inhibitor) and VPA (contained
in Gln-VPA structure). Docking results from some snapshots retrieved from MD simulations show that glutaminase recog-
nized Gln-VPA and DON. Additionally, Gln-VPA showed antiproliferative efects in HeLa cells and inhibited glutaminase
activity. Finally, the LC–MS-based metabolomics studies on HeLa cells treated with either Gln-VPA (IC60 =8 mM) or DON
(IC50 =3.5 mM) show diferent metabolomics behaviors, suggesting that they modulate diferent biological targets of the cell
death mechanism. In conclusion, Gln-VPA is capable of interfering with more than one pharmacological target of cancer,
making it an interesting drug that can be used to avoid multitherapy of classic anticancer drugs.
Graphic abstract
DON
Gln-vpa
No
treatment
DON
Gln-VPA
No treatment
LC-MS metabolomic on
HeLa cells
DON
No treatment
Gln-VPA
Keywords Glutaminase · Gln-VPA · Dual-target inhibitor · Anti-proliferative LC–MS-based metabolomic
Extended author information available on the last page of the article
Vol.:(0123456789)
1 3

Molecular Diversity
explore the intracellular pathways afected in the presence
of target compounds [16].
Introduction
Based on the overexpression of the glutaminase in cancer
[7], we designed a hybrid compound, named Gln-VPA, that
covalently couples VPA with glutamine (Scheme 1) [14]. In
this work, we expand our prior research to investigate the in
silico and in vitro efect of Gln-VPA on glutaminase.
Furthermore, Gln-VPA, 6-diazo-5-oxo-L-norleucine
(DON), and glutamine were investigate using docking stud-
ies, and snapshots of glutaminase from molecular dynam-
ics (MD) studies showed that both ligands are capable of
reaching the glutaminase catalytic binding site. Additionally,
Gln-VPA and DON show antiproliferative efects in HeLa
cells. In addition, LC–MS-based metabolomics studies of
HeLa cells treated with either Gln-VPA or DON showed up-
and down-regulated metabolites that depict the intracellular
signaling pathways associated with cell death.
Under aerobic conditions, cells metabolize glucose to
yield pyruvate via glycolysis in the cytoplasm and sub-
sequently CO2 into the mitochondria. In contrast, under
anaerobic conditions, glycolysis is the preferred meta-
bolic pathway for cancer cell yielding little pyruvate and
generated lactate with decreased oxygen consumption [1,
2]. In the 1920s, Otto Warburg observed that cancer cells
metabolize glucose in a distinct manner that difers from
normal cells. Even in the presence of oxygen, cancer cells
can rewire their glucose metabolism through a process that
limits their energy metabolism to glycolysis, triggering the
cancer cells to enter a state that has been termed “aerobic
glycolysis” [3–6]. Aside from the Warburg efect, glutami-
nolysis in cancer has emerged as a new opportunity to
design drugs with antiproliferative activity [6]. Glutami-
nase deaminates glutamine, yielding glutamic acid, which
is then converted to α-ketoglutarate to be incorporated into
the energy source cycle. There are three isoforms of glu-
taminase: the kidney isoform (KGA or GLS1), the liver
isoform (LGA or GLS2), and a splice variant of KGA (glu-
taminase C or GAC) [7, 8]. Activation of KGA by TGF-β
and Rho GTPase deregulates several pathways that trigger
cancer progression [9]. Most of the cancer types that afect
the human body depend largely on the glutamine supply
to trigger cell growth and proliferation. Furthermore, tar-
geting glutamine metabolism in cancer cells is a potential
therapeutic strategy against cancer [10–12].
Materials and methods
Molecular dynamics simulation
Since glutaminase KGA is strongly associated with cancer,
its crystal structure (PDB ID: 3VP0) was selected for in
silico simulations. First, water molecules and ions were
removed from the glutaminase. MD simulations were
performed with pmemd.cuda AMBER 12 [17] to obtain
several structural conformers of glutaminase. The glu-
taminase topology was built employing the Leap module
setting of the f99SB force feld, solvating and neutral-
izing the system with the appropriate number of water
molecules and NaCl ions, respectively [18]. Glutaminase
was minimized and equilibrated with the sander module
according to the following steps: The energy of the enzyme
was minimized by 10,000 steps with position restraints on
the enzyme to allow the relaxation of water molecules.
Afterward, the enzyme was slowly heated under the NVT
ensemble through a process of two sequential runs from
0 to 300 K for 1 ns, keeping the enzyme atoms restrained.
Subsequently, 1-ns simulation under an NPT ensemble at
300 K and 1 bar pressure with total heavy atoms restrained
was conducted to equilibrate the system. After that, 1 ns
Recently, several eforts have been conducted to obtain
more selective and potent drugs to treat cancer with fewer
side efects. Due to the toxic efects of VPA and its low
potency as an antiproliferative compound, our research
group has designed new VPA analogs by modifying the
chemical structure of the carboxylic acid, and these have
a better antiproliferative activity on diferent cancer cells
than that of VPA [13–15]. It is important to note that the
drug design structure of the target or ligand base could
yield selective drugs with better desired pharmacological
efects. However, to corroborate a pharmacological mecha-
nism, it is necessary to employ metabolomics studies to
Scheme 1 Synthesis of Gln-VPA
1 3

Molecular Diversity
of equilibration with a completely unrestricted system was
performed. Equilibration stages were reached from ≅ 20
to 50 ns of MD simulations without position restraints
applying periodic boundary conditions (PBCs) setting
an NPT ensemble at 300 K and 1 bar pressure. All the
electrostatic interactions were computed employing the
PME method [19], delimiting a 10 Å cutof for the van
der Waals interactions. The SHAKE algorithm was used
to tighten the bonds between heavy and hydrogen atoms
[20]. The temperature was held constant using Langevin
dynamics, and the pressure was kept at 1 bar. Furthermore,
the pressure coupling constant was stated as 1 ps, and the
collision frequency was 1.0 ps−1. Finally, the time step
employed was delimited to 2.0 fs, and the protein coordi-
nates retrieved from the MD simulations were previously
saved every 1 ps. The glutaminase conformers from the
MD were extracted every 10 ns to be used for docking
studies.
complexes were analyzed with the PyMOL v0.99 graphi-
cal viewer [25].
Synthesis of Gln‑VPA
The synthesis of Gln-VPA consisted of an amide bond for-
mation by the chemical activation of the carboxyl acid of
VPA with a subsequent addition of Gln to obtain Gln-VPA
[14] (Scheme 1).
Glutaminase activity inhibition assay
The glutaminase activity assay was carried out in two steps.
First, glutamic acid was produced from the deamidation
of glutamine by glutaminase; second, α-ketoglutarate was
formed from glutamic acid catalyzed by glutamate dehy-
drogenase. The frst step included 50 µL of Trizma base
[63 mM], potassium phosphate [206 mM] and EDTA
[0.25 mM] bufer at a pH of 8.6 at 37 °C. Then, 10 µL of
l-glutamine [200 mM] was added and incubated for 5 min.
Afterward, 6 µL of glutaminase [1.5 U] and 6 µL of potas-
sium phosphate bufer [400 mM] and boric acid [10 mM] at
pH 8.0 were added and incubated at 37 °C for 30 min. The
reactions were quenched by adding 6 µL of 3 M HCl and
then cooled in ice for 5 min.
Docking studies and physicochemical
To explore the afnity of Gln-VPA on glutaminase, dock-
ing studies were performed including DON (a glutami-
nase inhibitor) and glutamine (a glutaminase substrate).
DON and glutamine were used for validating purposes.
2D chemical structures of ligands were drawn (Scheme 2)
and 3D preoptimized with ChemBioDraw Ultra 12.0 soft-
ware [21] to achieve geometry and energy optimization
with the AM1 semiempirical method using the Gaussian
09 [22]. The 3D output structures were converted into.
pdb files with Gauss View 5.0 [23]. Docking studies
were conducted with AutoDock 4.2.6. All the input fles
were prepared with MGLTools 1.5.6 [24] by adding polar
hydrogen atoms and total Kollman charges to the enzyme
and Gasteiger charges and fexible bonds to ligands. A
blind docking procedure with a grid box of 126 Å3 and
a grid spacing of 0.375 Å3 with an initial population of
100 randomized structures and a total energy evaluation of
1 × 107 cycles was achieved. A Lamarckian genetic algo-
rithm was selected for scoring sampling and empirical free
energy functions as scoring function which calculate the
binding free energy values [24]. All glutaminase-ligand
The second step was performed in a 96-well plate. First,
150 µL of Trizma base [130 mM] at pH 9.4 and 37 °C was
added to each well; then, 20 µL of β-NAD [100 mM], 4 µL
of adenosine diphosphate (ADP) [20 mM], 2 µL of hydrogen
peroxide (H2O2) (3.0% v/v) and 20 µL of glutamate dehydro-
genase [4.2 U in H2O2] (GDH: Sigma G2626) were added
to each well. Immediately, 4 µL of glutamate [15 mM] was
added, and 4 µL of H2O was added to the blank. Finally,
4 µL of the reaction with glutaminase from the frst step
and 4 µL of the reaction without glutaminase were added
to their respective wells, and the samples were incubated at
37 °C for 40 min. The absorbance at 340 nm was read every
10 min in an ELISA reader (Multiskan, Thermo Scientifc).
MTT assay for DON in HeLa cells
The antiproliferative activity of DON and Gln-VPA were
assayed in HeLa cells in order to compare the experimental
O
OH
O
O
O
O
O
N
N
H
N
H2N
OH
OH
H
NH2
NH2
NH2
H
O
Scheme 2 Chemical structures of DON, a glutaminase inhibitor, glutamine and Gln-VPA
1 3

Molecular Diversity
results with those from our previous study of Gln-VPA [14].
HeLa cells were source from Dr Efraín Garrido Guerrero,
CINVESTAV IPN. HeLa cells were seeded into 96-well
culture plates at 1×103 cells/well in 100 µL of Dulbecco’s
modifed Eagle’s medium (DMEM Glutamax, Gibco, Life
Technologies, Invitrogen, USA) supplemented with 10%
(v/v) heat-inactivated fetal bovine serum (FBS, Biowest,
Kansas City, MO, USA) and with antibiotic–antimycotic
(Gibco). The cell cultures were maintained at 37 °C in a
humidifed atmosphere containing 5% CO2 for 24 h, and then
the cells were treated for 48 h with fresh medium using dif-
ferent concentrations of DON (0, 0.5, 1, 2, 4, 8 and 16 µM)
(Sigma-Aldrich, St. Louis, MO, USA) and Gln-VPA (0, 2,
4, 6, 8 and 10 mM) [14]. The plates were analyzed for cell
survival using the colorimetric 3-(4,5-dimethyl-thiazol-
2-yl)-2,5-diphenyltetrazolium bromide (MTT) dye reduction
assay (Sigma-Aldrich, St. Louis, MO, USA). The cytotoxic
efect was expressed as a percentage of cell survival relative
to untreated control cells and is defned as [(A594 nm treated
cells − A594 nm medium without cells)/(A594 nm nontreated
cells−A594 nm medium without cells)]×100; nontreated cells
were assumed to have 100% cell survival. The percentage of
HeLa cell survival was expressed as the mean SD of two
independent experiments carried out in triplicate.
part water was then added, and the samples were vortexed.
Then, one part of CHCl3 was added, and the samples were
vortexed and centrifuged (30 min, 5000 rpm, and 4 °C).
The aqueous phase 1 was saved and dried. The organic
phase was split in half, one half (organic phase 1), and the
other half of the sample was used for the acid extraction.
Organic phase 1 and sample from acid extraction samples
were dried.
*Step 3: For the acid extraction, the residual organic
phase from step 2 was resuspended in 2:0.8 MeOH: 1%
formic acid at pH 2. Samples were vortexed and then fol-
lowed the same procedure as step 2. The aqueous phase 2
was saved and dried; the organic phase was separated in
the organic phase 2, and half was used for the basic extrac-
tion. Both samples were dried.
*Step 4: For the basic extraction, the residual organic
phase was resuspended in 2:0.8 MeOH: 2% ammonium
hydroxide at pH 9. Samples were vortexed, and the same
procedure as step 2 was performed. The aqueous phase 3
was saved and dried.
Organic phases 1 and 2 were combined and reconsti-
tuted in MeOH:H2O (5:5) containing 0.1% formic acid for
reversed-phase chromatography.
UHPLC/MS studies
LC–MS‑based metabolomics
For UHPLC-Q-TOF–MS, an Agilent 1290 Infnity II sys-
tem coupled with a 6545 Q-TOF–MS of dual ESI source
(Agilent Technologies, Santa Clara, CA, USA) was used.
The mobile phase was Milli-Q water containing 0.1% for-
mic acid (component A) and methanol with 0.1% formic
acid (component B). For the organic phase analysis, an
Agilent Eclipse XDB-C-8, 150*. A 4.6 mm, 5 μm col-
umn running at 0.5 mL/min was used as standard LCMS
for separation of extracts obtained with a gradient elution
from 10 to 100% B for 20 min followed by 100% B that
was maintained for 5 min. The mobile phase fow rate was
0.5 mL/min after injecting 5 μL. The column temperature
throughout the separation process was 40 °C. Previously to
perform the LC–MS analyses, samples were kept at 4 °C.
The ESI source was operated in a positive ion mode
with the following conditions: The fragmentor voltage
was set at 120 V, the nebulizer gas was set at 35 psi, the
capillary voltage was set at 3500 V, the drying gas fow
rate was set at 10 dm3/min, and the temperature was set
at 300 °C. For MS/MS measurements, a collision energy
ramp ranging from 15 to 40 eV to promote fragmentation
was used. The data were acquired in centroid and pro-
fle modes using the high-resolution mode (4 GHz). The
mass range was set at 50–1000 m/z in the MS and MS/
MS modes. The data were processed with the MassHunter
Workstation LC–MS Data Acquisition B.08.00 Software.
Cell culture, treatment and metabolite extraction
HeLa cells were cultured in DMEM (Sigma-Aldrich, Méx-
ico) supplemented with 5% decomplemented fetal bovine
serum (FBS, Biowest, Kansas City, MO, USA). HeLa cells
were source from Dr. Mario Alberto Rodríguez Rodríguez,
CINVESTAV IPN. HeLa cells were incubated at 37 °C in
a humidifed atmosphere with 5% CO2 in Petri dishes for
10 days to obtain 7×106 cells. HeLa cells were treated with
DON or Gln-VPA using the IC50 values of 3.5 µM for DON
and the IC60 ≅8 mM for Gln-VPA for 48 h.
Steps for metabolite extraction
*Step 1: Petri dishes were placed on wet ice. DMEM was
collected, and cells were washed twice with ice-cold PBS
(DPBS) and frozen on dry ice. MeOH:H2O (2:0.8) (kept
at −80 °C) was added to the dishes and then transferred to
wet ice before scrapping the cells. Cells were collected into
Eppendorf tubes and snap-frozen in liquid nitrogen to then
be stored at −80 °C until processing (step 2).
*Step 2: Samples were sonicated, and one part CHCl3
was added to obtain a total solution ratio of 2:0.8:1 of
MeOH:H2O:CHCl3. Then, the samples were vortexed. One
1 3

Molecular Diversity
LC–MS analysis
Results and discussion
The raw data obtained by UHPLC-Q-TOF/MS were con-
verted into mzData to be analyzed with XCMS Online
version 3.7.1 (https://xcmsonline.scripps.edu/). The
parameters were defined based on the experimental con-
ditions. XCMS showed peak detection, retention time
correction, chromatogram alignment, metabolite feature
annotation, statistical evaluation and putative metabolite
identification through the METLIN database.
Molecular dynamics and docking results
It is known that drug designs can be based on the chemi-
cal properties of either the target or ligand [26]. Tradition-
ally, a rigid structural target is used for docking studies,
sometimes yielding incongruent data [27]. Furthermore,
we have suggested to employ diferent protein conformers
to sample diferent ligand conformations to obtain better
binding poses that could reproduce experimental data [28].
For that reason, in this work, we employed several glu-
taminase structural conformations from MD simulations to
gain insights that explain the biological results of the target
ligands. The RMSD and Rg values of glutaminase show that
it reaches structural stability and equilibration from 20 to
50 ns (Fig. 1a). The RMSF shows greater structural motion
values in those regions that belong to having no defned
Moreover, the data were analyzed with Agilent Mass
Hunter Qualitative Analysis B.07.00, Agilent Mass
Hunter Profinder B.08.00 and Agilent Mass Profiler Pro-
fessional software.
Fig. 1 Molecular dynamics simulation of glutaminase during 50 ns. a RMSD, root-mean-square deviation; b RMSF, root-mean-square fuctua-
tions; c Rg, radius of gyration. RMSD and Rg depict structural equilibration stages from 20 to 50 ns
1 3

Molecular Diversity
Fig. 2 Binding modes of DON
(left column), glutamine (mid-
dle column) and Gln-VPA (right
column) in glutaminase snap-
shots from 0 to 50 ns sampled
every 10 ns from top to bottom.
Hydrogen bonds are shown as
black dashes
secondary structure (Fig. 1b). MD simulations yield difer-
ent glutaminase conformers that were employed for dock-
ing studies using DON, Gln-VPA, and glutamine as small
ligands (Fig. 2). Figure 2 and Table 1 show that all ligands
that were recognized by glutaminase adopted diferent bind-
ing poses and reached diferent residues. In this docking
study, l-glutamine reproduced its co-crystallized coordi-
nated into glutaminase (PDB ID: 3VP0) [29], allowing us
1 3

Molecular Diversity
Table 1 Free energy values
and residues involved in
Ligand
ΔG
ns
0
Reached residues
the enzyme-ligand complex
obtained from docking studies.
Residues formed by hydrogen
bonds are depicted by one or
more asterisks according to the
number of interactions
DON
Gln
Gln-VPA
DON
Gln
Gln-VPA
DON
Gln
Gln-VPA
DON
Gln
Gln-VPA
DON
Gln
Gln-VPA
DON
Gln
−5.59
−5.43
−5.94
−5.3
−5.86
−7.57
−4.44
−4.66
−3.79
−6.97
−5.75
−5.89
−5.29
−5.0
Q285*, S286*, K289*, E381*, N388*, Y414*, C418, V484*
Q285*, S286*, K289, E381, N388*, Y414*, C418, V484*
S286, K289*, N331, M333, V334*, N335*, R387*, C463, Y466, K481
Y249*, Q285, S286*, E381, Y414*, G483, V484, A485
Y249, Q285, E381**, Y414*, C418*, V484*
S286*, K289*, S314, G315, L316, M333, V334, N335*, C463, Y466, D467
Q285*, K289, E381, N388, Y414, C418*
E381***, S384, G386, N388**, Y414*, C418
K289*, N335*, E381*, R387, N388, I391, Y414, C418*
I250, Q34, K255*, P281*, F282, C283*, E421
Y249, K289*, N335, E381**, S384*, N388*, Y414, C418
S286, K289*, N331, M333, V334, N335*, C463, Y466, K481*
K289*, K292, N335, E381*, Y414, C418
Q285, S286*, K289*, E381*
K289*, N331, M333, V334*, N335*, Y466, K481
N311*, E312*, N324*, E325*, H330*
10
20
30
40
50
−5.48
−5.59
−5.34
−2.71
E312, P313*, N324*, E325*, H330
H306, G310, E312, N324*, E325**, D326, H330
Gln-VPA
to validate the docking procedure. Docking studies of glu-
tamine on glutaminase show that glutamine NH3+ has elec-
trostatic interactions with E381 (CO2) and make hydrogen
bonds with C418 (SH) and with N388 (NH2), at 20 ns. Then,
glutamine reached the same binding site but with a diferent
conformation until 40 ns; at 50 ns, it reaches a neighboring
binding site making hydrogen bonds with the backbone of
S286 and electrostatic interactions with E381 (Fig. 2). On
the other hand, in snapshots from 0 to 20 ns, DON reaches
E381 making an electrostatic interaction. At 30 ns, the DON
carboxyl group interacts with K255 making an electrostatic
interaction and hydrogen bonds with C283 and P281, rec-
ognizing a site close to that previously reported [29]. At
40 ns, the ligand again reached E381 making an electrostatic
interaction and hydrogen bonds with N335 and K289, but
at 50 ns, it is shown to be close to the reported site making
an electrostatic interaction with E325 and hydrogen bonds
with N324 and H330 (Fig. 2). However, Gln-VPA on glu-
taminase adopts diferent binding modes that are very close
to the catalytic binding site [29] and make a hydrogen bond
with S286 and making an electrostatic interaction with R387
(Fig. 2). At the 20-ns snapshot, Gln-VPA recognizes resi-
dues very close to the allosteric binding site of glutaminase
in its dimer form making an electrostatic interaction with
K289 and hydrogen bonds with E381, C418 and N335 [29].
At the 50-ns snapshot, Gln-VPA interacts with the exposed
residues, N324 and E325 making hydrogen bonds, in the
dimer conformation, which forms an allosteric binding site
demarked by the interactions established with the BPTES
allosteric inhibitor [30, 31]. These in silico studies show that
Gln-VPA is able to reach the glutaminase catalytic binding
site (Fig. 2), suggesting that Gln-VPA could be a promising
strategy in cancer therapy.
A
B
100
80
60
40
20
0
0
2
4
6
8
10
Gln-VPA [mM]
Fig. 3 a The antiproliferative efect of DON on HeLa cells. b The
antiproliferative efect of Gln-VPA on HeLa cells. HeLa cells were
incubated for 48 h at diferent concentrations and analyzed by the
MTT assay. The data are the mean SD for n≥6 within at least two
independent experiments
1 3

Molecular Diversity
Fig. 4 Glutaminase activity
inhibition of a DON [μM] and b
Gln-VPA [mM] to measure their
IC50 values
at the highest concentrations tested. These results allowed
us to obtain the IC50 values for DON (IC50 =0.029 mM) and
Gln-VPA (IC50 =1.86 mM). These results showed that Gln-
VPA might be a promising pharmacological compound. It
is due to the presence of the glutamine moiety bound to
VPA do not afect the antiproliferative properties that were
reported elsewhere for VPA [14, 15]. Thus, Gln-VPA could
be a multitarget compound that could be used against can-
cer, similar to other reported promissory compounds [32].
This is because if one protein target mutate, there are other
targets that are able to inhibit and kill cancer cells, avoid-
ing drug resistance [33]; this has some advantages, such as
decreasing the use of multitherapy and, in case of a protein
mutation, stopping cancer due to the action of a multitarget
compound on other targets. The in silico results show that
Gln-VPA and DON can bind to glutaminase whereas that
the in vitro studies show that Gln-VPA and DON inhibit
glutaminase activity. Additionally, Gln-VPA and DON show
antiproliferative properties in HeLa cells. However, their
Cell viability analysis
Antiproliferative assays on cancer cells could be a direct
and frst pharmacological approach to measure the possi-
ble anticancer compounds to validate and support in silico
guide drug design [15]. Therefore, based on our previous
results, Gln-VPA had better antiproliferative efects on HeLa
cells than VPA [14], and this was corroborated again by
employing DON as a positive control. MTT results show a
concentration-dependent response behavior of DON in HeLa
cells with an IC50 =3.5 µM (Fig. 3), whereas for Gln-VPA,
its IC60 ≅8 mM [14].
Enzymatic inhibition of glutaminase by Gln‑VPA
and DON
Gln-VPA was tested as a glutaminase inhibitor using DON
as a positive control. Figure 4 shows that the glutaminase
activity decreased to ~85% by DON and ~55% by Gln-VPA
1 3

Molecular Diversity
Fig. 5 a Cloud plot after data processing using the XCMS online
software of the organic phases of untreated HeLa cells and HeLa
cells treated with DON. b Cloud plot after data processing using the
XCMS online software of the organic phases of untreated HeLa cells
and HeLa cells treated with Gln-VPA
pharmacological cell death mechanisms remain unclear. Fur-
thermore, the metabolomics results could suggest possible
pathways related to cell death in presence of Gln-VPA and
DON as well as the possible pharmacological mechanism.
metabolites. The results show that DON up-regulates 5′-deoxy-
adenosine. This suggests an increase in lipoate production
that disrupts cancer mitochondrial metabolism [34]. Moreo-
ver, DON up-regulates laurate and palmitate, which increases
reactive oxygen species (ROS), afecting mitochondrial depo-
larization and inducing caspase 3 apoptosis [35]. Additionally,
DON down-regulates the metabolites 3-pyridylacetate and
4-(3-pyridyl)-butanoate, which are from nicotine degradation
pathway and stimulate oncogenic and mitogenic signaling
cascades to activate proliferative pathways [36]. Additionally,
nicotine induces the epithelial-mesenchymal transition (EMT),
which facilitates metastases and decreases Chk2 (tumor sup-
pressor) [36]. In addition, DON up-regulated aspirin-triggered
resolvin D and the resolvin D biosynthesis pathway in HeLa
cells. Resolvin is known to be an anti-infammatory mediator
LC–MS‑based metabolomics analysis
LC–MS results show several up- and down-regulated metab-
olites. The focus of this study was on those metabolites asso-
ciated with HeLa cell death in the presence of Gln-VPA or
DON.
The analysis of organic phases showed 2417 metabolites,
of which 358 were signifcantly altered between untreated
HeLa cells and HeLa cells treated with DON. In the cloud
plot (Fig. 5a), there are deregulated (up- or down-regulated)
1 3

Molecular Diversity
Fig. 6 a Up- and down-regu-
lated metabolites of untreated
HeLa cells (left) and HeLa cells
treated with DON (middle)
and Gln-VPA (right). b Venn
diagram from Agilent Mass
Profle Professional software. c
3D PCA scores obtained from
Agilent Mass Profle Profes-
sional software
that suppresses cancer progression [37]. It has been shown that
resolvins suppress EMT, which then inhibits cancer metastasis
[37].
potential, inducing caspase 3 apoptosis. Finally, it increases
fatty acid and oleate uptake that favor lipotoxicity [43].
Finally, Gln-VPA down-regulated fumarate, which may
afect the citric acid cycle that is required for cancer metabo-
lism [44]. Citrate induces a negative feedback on glycoly-
sis, reducing cell proliferation in many tumor cell lines and
inducing apoptosis [45]. Moreover, it is known that Notch
1 pathway is up-regulated in many cancers, such as cervi-
cal cancer. It plays important roles in cancer development,
including cell growth promotion and EMT [46]. Therefore,
down-regulation of fumarate may afect the Notch signaling
pathway, decreasing cell proliferation. [47].
Regarding Hela cells treated with Gln-VPA, analysis of
organic phases showed 2,508 metabolites, of these, 527 were
signifcantly diferent compared with untreated HeLa cells,
as shown in the cloud plot (Fig. 5b). Gln-VPA up-regulates
metabolites involved in the (S)-reticuline biosynthesis path-
way. (S)-reticuline inhibits angiogenesis and cell growth and
promotes apoptosis [38]. Moreover, Gln-VPA up-regulates
arginine-succinate, which is often down-regulated in cancer
cells [39]. Moreover, the up-regulation of arginine succinate
and down-regulation of both fumarate and L-aspartate sug-
gests a decrease in arginine auxotrophy, which is enhanced
in HeLa cells [40] and afects the growth, apoptotic signaling
and mTOR and RAF/MEK/ERK signaling. [41] Decreased
fumarate inhibits tumorigenic processes because it is an onco-
metabolite that favors HIF hydroxylases [42]. Additionally,
Gln-VPA up-regulates laurate and palmitate, which increase
the ROS production and decrease the mitochondrial membrane
Metabolomics analysis of HeLa with Agilent
software
The Profle Plot (Fig. 6a, b) showed that some compounds
are only deregulated in the presence of Gln-VPA, DON, or
both. For example, the compounds in group 1 are down-
regulated when the cells are untreated, whereas compounds
from group 2 are up-regulated in cells that are untreated and
1 3

Molecular Diversity
10. Yuan LQ, Sheng XG, Willson AK, Roque DR, Stine JE, Guo H,
Jones HM, Zhou CX, Bae-Jump VL (2015) Glutamine promotes
ovarian cancer cell proliferation through the mTOR/S6 path-
way. Endocr-Relat Cancer 22:577–591. https://doi.org/10.1530/
ERC-15-0192
treated with Gln-VPA. This means that these compounds are
specifcally targeted by DON. The Venn diagram (Fig. 6b)
shows us if some entities are only present in one condition
or are shared between two conditions. This helps us to high-
light specifc targets. The PCA scores show that metabolites
grouped according to each sample condition (treated versus
untreated cells) were well separated, and both treatments
showed diferent efects on HeLa cells (Fig. 6c).
11. Mohamed A, Deng X, Khuri FR, Owonikoko TK (2014) Altered
glutamine metabolism and therapeutic opportunities for lung
cancer. Clin Lung Cancer 15:7–15. https://doi.org/10.1016/j.
cllc.2013.09.001
12. Gross MI, Demo SD, Dennison JB, Chen L, Chernov-Rogan T,
Goyal B, Janes JR, Laidig GJ, Lewis ER, Li J, Mackinnon AL,
Parlati F, Rodriguez ML, Shwonek PJ, Sjogren EB, Stanton TF,
Wang T, Yang J, Zhao F, Bennett MK (2014) Antitumor activity
of the glutaminase inhibitor CB-839 in triple-negative breast can-
cer. Mol Cancer Ther 13:890–901. https://doi.org/10.1158/1535-
7163.MCT-13-0870
In conclusion, it is important to note that Gln-VPA is
an inhibitor of glutaminase according to the results from
in silico and in vitro assays. In addition, Gln-VPA showed
antiproliferative efects modulates diferent intracellular
pathways according to metabolomics studies.
13. Luna-Palencia GR, Martinez-Ramos F, Vasquez-Moctezuma I,
Fragoso-Vazquez MJ, Mendieta-Wejebe JE, Padilla-Martinez II,
Sixto-Lopez Y, Mendez-Luna D, Trujillo-Ferrara J, Meraz-Rios
MA, Fonseca-Sabater Y, Correa-Basurto J (2014) Three amino
acid derivatives of valproic acid: design, synthesis, theoretical
and experimental evaluation as anticancer agents. Anti-Cancer
Agent Med Chem 14:984–993. https://doi.org/10.2174/18715
20614666140127113218
Acknowledgments The authors thank the Instituto Politécnico
Nacional (SIP:20170539, 20161454, 20171022, and 20171881), BEIFI,
COFAA-IPN and Grants CB-254600, CB-241339, SEP-CONACYT-
ANUIES-ECOS Francia: 296637 (CONACYT) and Proyecto Insig-
nia IPN-2015. Thanks to PhD Carol Haney Ball (Agilent) for article
revision.
14. Martinez-Ramos F, Luna-Palencia GR, Vasquez-Moctezuma I,
Mendez-Luna D, Fragoso-Vazquez MJ, Trujillo-Ferrara J, Meraz-
Rios MA, Mendieta-Wejebe JE, Correa-Basurto J (2016) Dock-
ing studies of glutamine valproic acid derivative (S)-5- amino-
2-(heptan-4-ylamino)-5-oxopentanoic acid (Gln-VPA) on HDAC8
with biological evaluation in HeLa cells. Anticancer Agents Med
Chem 16:1485–1490. https://doi.org/10.2174/187152061666616
0204111158
Compliance with ethical standards
Conflict of interest All authors inform that there are not confict of
interest.
15. Prestegui-Martel B, Bermudez-Lugo JA, Chavez-Blanco A,
Duenas-Gonzalez A, Garcia-Sanchez JR, Perez-Gonzalez OA,
Padilla M II, Fragoso-Vazquez MJ, Mendieta-Wejebe JE, Cor-
rea-Basurto AM, Mendez-Luna D, Trujillo-Ferrara J, Correa-
Basurto J (2016) N-(2-hydroxyphenyl)-2-propylpentanamide,
a valproic acid aryl derivative designed in silico with improved
anti-proliferative activity in HeLa, rhabdomyosarcoma and breast
cancer cells. J Enzyme Inhib Med Chem 31:140–149. https://doi.
org/10.1080/14756366.2016.1210138
References
1. Hanahan D, Weinberg RA (2011) Hallmarks of cancer: the
next generation. Cell 144:646–674. https://doi.org/10.1016/j.
cell.2011.02.013
2. Jones RG, Thompson CB (2009) Tumor suppressors and cell
metabolism: a recipe for cancer growth. Genes Dev 23:537–548.
https://doi.org/10.1101/gad.1756509
16. Shah NJ, Sureshkumar S, Shewade DG (2015) Metabolomics:
a tool ahead for understanding molecular mechanisms of drugs
and diseases. Indian J Clin Biochem 30:247–254. https://doi.
org/10.1007/s12291-014-0455-z
3. Locasale JW, Cantley LC, Heiden MGV (2009) Cancer’s insatia-
ble appetite. Nat Biotechnol 27:916–917. https://doi.org/10.1038/
nbt1009-916
4. Warburg O (1956) On the origin of cancer cells. Science 123:309–
314. https://doi.org/10.1126/science.123.3191.309
17. Case DA, Cheatham TE 3rd, Darden T, Gohlke H, Luo R, Merz
KM Jr, Onufriev A, Simmerling C, Wang B, Woods RJ (2005)
The Amber biomolecular simulation programs. J Comput Chem
26:1668–1688. https://doi.org/10.1002/jcc.20290
5. Liberti MV, Locasale JW (2016) The warburg efect: how does it
beneft cancer cells? Trends Biochem Sci 41:211–218. https://doi.
org/10.1016/j.tibs.2015.12.001
18. Wang J, Wolf RM, Caldwell JW, Kollman PA, Case DA (2004)
Development and testing of a general amber force feld. J Comput
Chem 25:1157–1174. https://doi.org/10.1002/jcc.20035
19. Darden T, York D, Pedersen L (1993) Particle mesh Ewald—an
N.Log(N) method for Ewald sums in large systems. J Chem Phys
98:10089–10092. https://doi.org/10.1063/1.464397
6. Bensinger SJ, Christofk HR (2012) New aspects of the Warburg
efect in cancer cell biology. Semin Cell Dev Biol 23:352–361.
https://doi.org/10.1016/j.semcdb.2012.02.003
7. Aledo JC, Gomez-Fabre PM, Olalla L, Marquez J (2000) Identif-
cation of two human glutaminase loci and tissue-specifc expres-
sion of the two related genes. Mamm Genome 11:1107–1110.
https://doi.org/10.1007/s003350010190
20. van Gunsteren WF, Berendsen HJC (1977) Algorithms for macro-
molecular dynamics and constraint dynamics. Mol Phys 34:1311–
1327. https://doi.org/10.1080/00268977700102571
8. Curthoys NP, Watford M (1995) Regulation of glutaminase activ-
ity and glutamine metabolism. Annu Rev Nutr 15:133–159. https
://doi.org/10.1146/annurev.nu.15.070195.001025
21. Kimberley R. Cousins (2005) ChemDraw Ultra 9.0. Cam-
bridgeSoft, 100 CambridgePark Drive, Cambridge, MA 02140.
http://www.cambridgesoft.com. See Web site for pricing options. J
Am Chem Soc 127(11):4115–4116. https://doi.org/10.1021/ja041
0237
9. McDermott LA, Iyer P, Vernetti L, Rimer S, Sun J, Boby M,
Yang T, Fioravanti M, O’Neill J, Wang L, Drakes D, Katt W,
Huang Q, Cerione R (2016) Design and evaluation of novel glu-
taminase inhibitors. Bioorg Med Chem 2:1819–1839. https://doi.
org/10.1016/j.bmc.2016.03.009
22. Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA,
Cheeseman JR, Scalmani G, Barone V, Petersson GA, Nakat-
suji H, Li X, Caricato M, Marenich AV, Bloino J, Janesko BG,
1 3

Molecular Diversity
laurate. FEBS Open Bio 29:485–491. https://doi.org/10.1016/j.
fob.2015.05.008
Gomperts R, Mennucci B, Hratchian HP, Ortiz JV, Izmaylov AF,
Sonnenberg JL, Williams-Young D, Ding F, Lipparini F, Egidi
F, Goings J, Peng B, Petrone A, Henderson T, Ranasinghe D,
Zakrzewski VG, Gao J, Rega N, Zheng G, Liang W, Hada M,
Ehara M, Toyota K, Fukuda R, Hasegawa J, Ishida M, Nakajima
T, Honda Y, Kitao O, Nakai H, Vreven T, Throssell K, Montgom-
ery JA Jr, Peralta JE, Ogliaro F, Bearpark MJ, Heyd JJ, Brothers
EN, Kudin KN, Staroverov VN, Keith TA, Kobayashi R, Normand
J, Raghavachari K, Rendell AP, Burant JC, Iyengar SS, Tomasi J,
Cossi M, Millam JM, Klene M, Adamo C, Cammi R, Ochterski
JW, Martin RL, Morokuma K, Farkas O, Foresman JB, Fox DJ
(2016) Gaussian 09, Revision C.01. Gaussian Inc, Wallingford
23. Dennington R, Keith TA, Millam JM (2016) GaussView, Version
5. Semichem Inc., Shawnee Mission
36. Sanner T, Grimsrud TK (2015) Nicotine: carcinogenicity and
efects on response to cancer treatment. Front Oncol 5:196. https
://doi.org/10.3389/fonc.2015.00196
37. Moro K, Nagahashi M, Ramanathan R, Takabe K, Wakai T (2016)
Resolvins and omega three polyunsaturated fatty acids: clinical
implications in infammatory diseases and cancer. World J Clin
Cases 4:155–164. https://doi.org/10.12998/wjcc.v4.i7.155
38. Chen Y, Qin Y, Li L, Chen J, Zhang X, Xie Y (2017) Morphine
can inhibit the growth of breast cancer MCF-7 cells by arresting
the cell cycle and inducing apoptosis. Biol Pharm Bull 40:1686–
1692. https://doi.org/10.1248/bpb.b17-00215
39. Wu L, Li L, Meng S, Qi R, Mao Z, Lin M (2013) Expression
of arginosuccinate synthetase in patients with hepatocellular
carcinoma. J Gastroenterol Hepatol 28:365–368. https://doi.
org/10.1111/jgh.12043
24. Morris GM, Huey R, Lindstrom W, Sanner MF, Belew RK, Good-
sell DS, Olson AJ (2009) AutoDock4 and AutoDockTools4: auto-
mated docking with selective receptor fexibility. J Comput Chem
30:2785–2791. https://doi.org/10.1002/jcc.21256
40. Dillon BJ, Prieto VG, Curley SA, Ensor CM, Holtsberg FW,
Bomalaski JS, Clark MA (2004) Incidence and distribution of
arginosuccinate synthetase defciency in human cancers: a method
for identifying cancers sensitive to arginine deprivation. Cancer
100:826–833. https://doi.org/10.1002/cncr.20057
25. DeLano WL (2002) The PyMOL Molecular Graphics System.,
0.99. DeLano Scientifc, San Carlos
26. Baig MH, Ahmad K, Roy S, Ashraf JM, Adil M, Siddiqui MH,
Khan S, Kamal MA, Provazník I, Choi I (2016) Computer aided
drug design: success and limitations. Curr Pharm Des 22:572–
581. https://doi.org/10.2174/1381612822666151125000550
27. Poli G, Martinelli A, Tuccinardi T (2016) Reliability analysis and
optimization of the consensus docking approach for the develop-
ment of virtual screening studies. J Enzyme Inhib Med Chem
31:167–173. https://doi.org/10.1080/14756366.2016.1193736
28. Rosales-Hernández MC, Correa-Basurto J (2015) The impor-
tance of employing computational resources for the automation
of drug discovery. Expert Opin Drug Discov 10:213–219. https
://doi.org/10.1517/17460441.2015.1005071
41. Savaraj N, You M, Wu C, Kuo MT, Dinh V, Wangpaichitr M, Feun
L (2012) Targeting arginosuccinate synthetase in cancer therapy.
In: Chatterjee M, Kashf K (eds) Cell signaling & molecular tar-
gets in cancer. Springer, New York, pp 37–51
42. Yang M, Soga T, Pollard PJ, Adam J (2012) The emerging role
of fumarate as an oncometabolite. Front. Oncol. 2:85. https://doi.
org/10.3389/fonc.2012.00085
43. Kitaura Y, Inoue K, Kato N, Matsushita N, Shimomura Y (2010)
Enhanced oleate uptake and lipotoxicity associated with lau-
rate. FEBS Open Bio 5:485–491. https://doi.org/10.1016/j.
fob.2015.05.008
29. Thangavelu K, Chong QY, Low BC, Sivaraman J (2014) Struc-
tural basis for the active site inhibition mechanism of human
kidney-type glutaminase (KGA). Sci Rep 4:3827. https://doi.
org/10.1038/srep03827
44. Asgari Y, Zabihinpour Z, Salehzadeh-Yazdi A, Schreiber F,
Masoudi-Nejad A (2015) Alterations in cancer cell metabolism:
the Warburg efect and metabolic adaptation. 105:275–281. https
://doi.org/10.1016/j.ygeno.2015.03.001
30. Ramachandran S, Pan CQ, Zimmermann SC, Duvall B, Tsuka-
moto T, Low BC, Sivaraman J (2016) Structural basis for explor-
ing the allosteric inhibition of human kidney type glutaminase.
Oncotarget 7:57943–57954. https://doi.org/10.18632/oncotarget
.10791
45. Ren JG, Seth P, Ye H, Gou K, Hanai JI, Husain Z, Sukhatme
VP (2017) Citrate suppresses tumor growth in multiple models
through inhibition of glycolysis, the tricarboxylic acid cycle and
the IGF-1R pathway. Sci Rep 7:4537. https://doi.org/10.1038/
s41598-017-04626-4
31. Thangavelu K, Pan CQ, Karlberg T, Balaji G, Uttamchandani M,
Suresh V, Schuler H, Low BC, Sivaraman J (2012) Structural
basis for the allosteric inhibitory mechanism of human kidney-
type glutaminase (KGA) and its regulation by Raf-Mek-Erk
signaling in cancer cell metabolism. Proc Natl Acad Sci USA
109:7705–7710. https://doi.org/10.1073/pnas.1116573109
32. Chen Z, Han L, Xu M, Xu Y, Qian X (2013) Rationally designed
multitarget anticancer agents. Curr Med Chem 20:1694–1714.
https://doi.org/10.2174/0929867311320130009
46. Sun L, Liu M, Sun GC, Yang X, Qian Q, Feng S, Mackey LV,
Coy DH (2016) Notch signaling activation in cervical cancer cells
induces cell growth arrest with the involvement of the nuclear
receptor NR4A2. J Cancer 7:1388–1395. https://doi.org/10.7150/
jca.15274
47. Chen T, Wang T, Liang W, Zhao Q, Yu Q, Ma CM, Zhuo L,
Guo D, Zheng K, Zhou C, Wei S, Huang W, Jiang J, Liu L, Li S,
He J, Jiang Y (2019) PAK4 phosphorylates fumarase and blocks
TGFβ-Induced cell growth arrest in lung cancer cells. Can Res
79:1383–1397. https://doi.org/10.1158/0008-5472.CAN-18-2575
33. Kucuksayan E, Ozben T (2017) Hybrid compounds as multitar-
get directed anticancer agents. Curr Top Med Chem 17:907–918.
https://doi.org/10.2174/1568026616666160927155515
34. Zachar Z, Marecek J, Maturo C, Gupta S, Stuart SD, Howell K,
Schauble A, Lem J, Piramzadian A, Karnik S, Lee K, Rodríguez
R, Shorr R, Bingham PM (2011) Non-redox-active lipoate deri-
vates disrupt cancer cell mitochondrial metabolism and are potent
anticancer agents in vivo. J Mol Med 89:1137–1148. https://doi.
org/10.1007/s00109-011-0785-8
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional afliations.
35. Kitaura Y, Inoue K, Kato N, Matsushita N, Shimomura Y
(2015) Enhanced oleate uptake and lipotoxicity associated with
1 3

Molecular Diversity
Afliations
M. J. Fragoso‑Vázquez1,2 · D. Méndez‑Luna2 · M. C. Rosales‑Hernández2 · G. R. Luna‑Palencia2 · A. Estrada‑Pérez2 ·
Benedicte Fromager3 · I. Vásquez‑Moctezuma4 · J. Correa‑Basurto2
* M. C. Rosales-Hernández
Innovation), Escuela Superior de Medicina, Instituto
Politécnico Nacional, Plan de San Luis y Díaz Mirón,
Mexico City 11340, Mexico
marcrh2002@yahoo.com.mx
* J. Correa-Basurto
3
4
corrjose@gmail.com
Lycée Saint Jean Hulst in Versailles, Versailles City, France
Laboratorio de Bioquímica de la Escuela Superior
de Medicina, Instituto Politécnico Nacional, Plan de
San Luis y Díaz Mirón, Delegación Miguel Hidalgo,
Mexico City 11340, Mexico
1
Departamento de Química Orgánica, Escuela Nacional
de Ciencias, Biológicas, Instituto Politécnico Nacional,
Prolongación de Carpio y Plan de Ayala, Col. Casco de
Santo Thomas, Delegación Miguel Hidalgo, Mexico City,
Mexico
2
Laboratorio de Diseño y Desarrollo de Nuevos Fármacos
e Innovación Biotécnológica (Laboratory for the Design
and Development of New Drugs and Biotechnological
1 3