Synthesis, Pharmacological, and Biological Evaluation of 2-Furoyl-Based MIF-1 Peptidomimetics and the Development of a General-Purpose Model for Allosteric Modulators (ALLOPTML)

Article abstract of DOI:10.1021/acschemneuro.0c00687

This work describes the synthesis and pharmacological evaluation of 2-furoyl-based Melanostatin (MIF-1) peptidomimetics as dopamine D2 modulating agents. Eight novel peptidomimetics were tested for their ability to enhance the maximal effect of tritiated N-propylapomorphine ([3H]-NPA) at D2 receptors (D2R). In this series, 2-furoyl-l-leucylglycinamide (6a) produced a statistically significant increase in the maximal [3H]-NPA response at 10 pM (11 ± 1%), comparable to the effect of MIF-1 (18 ± 9%) at the same concentration. This result supports previous evidence that the replacement of proline residue by heteroaromatic scaffolds are tolerated at the allosteric binding site of MIF-1. Biological assays performed for peptidomimetic 6a using cortex neurons from 19-day-old Wistar-Kyoto rat embryos suggest that 6a displays no neurotoxicity up to 100 μM. Overall, the pharmacological and toxicological profile and the structural simplicity of 6a makes this peptidomimetic a potential lead compound for further development and optimization, paving the way for the development of novel modulating agents of D2R suitable for the treatment of CNS-related diseases. Additionally, the pharmacological and biological data herein reported, along with >20a000 outcomes of preclinical assays, was used to seek a general model to predict the allosteric modulatory potential of molecular candidates for a myriad of target receptors, organisms, cell lines, and biological activity parameters based on perturbation theory (PT) ideas and machine learning (ML) techniques, abbreviated as ALLOPTML. By doing so, ALLOPTML shows high specificity Sp = 89.2/89.4%, sensitivity Sn = 71.3/72.2%, and accuracy Ac = 86.1%/86.4% in training/validation series, respectively. To the best of our knowledge, ALLOPTML is the first general-purpose chemoinformatic tool using a PTML-based model for the multioutput and multicondition prediction of allosteric compounds, which is expected to save both time and resources during the early drug discovery of allosteric modulators.

Full text of DOI:10.1021/acschemneuro.0c00687

pubs.acs.org/chemneuro
Research Article
Synthesis, Pharmacological, and Biological Evaluation of 2‑Furoyl-
Based MIF‑1 Peptidomimetics and the Development of a General-
Purpose Model for Allosteric Modulators (ALLOPTML)
́
́
́
Ivo E. Sampaio-Dias,* Jose E. Rodríguez-Borges, Víctor Yan
̃
ez-Perez, Sonia Arrasate, Javier Llorente,
́
Jose M. Brea, Harbil Bediaga, Dolores Vin
̃
a, María Isabel Loza, Olga Caamano, Xerardo García-Mera,*
̃
́
and Humberto Gonzalez-Díaz*
Cite This: https://dx.doi.org/10.1021/acschemneuro.0c00687
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: This work describes the synthesis and pharmacological evaluation of 2-furoyl-based Melanostatin (MIF-1)
peptidomimetics as dopamine D2 modulating agents. Eight novel peptidomimetics were tested for their ability to enhance the
maximal effect of tritiated N-propylapomorphine ([3H]-NPA) at D2 receptors (D2R). In this series, 2-furoyl-L-leucylglycinamide
(6a) produced a statistically significant increase in the maximal [3H]-NPA response at 10 pM (11 1%), comparable to the effect of
MIF-1 (18 9%) at the same concentration. This result supports previous evidence that the replacement of proline residue by
heteroaromatic scaffolds are tolerated at the allosteric binding site of MIF-1. Biological assays performed for peptidomimetic 6a
using cortex neurons from 19-day-old Wistar-Kyoto rat embryos suggest that 6a displays no neurotoxicity up to 100 μM. Overall, the
pharmacological and toxicological profile and the structural simplicity of 6a makes this peptidomimetic a potential lead compound
for further development and optimization, paving the way for the development of novel modulating agents of D2R suitable for the
treatment of CNS-related diseases. Additionally, the pharmacological and biological data herein reported, along with >20 000
outcomes of preclinical assays, was used to seek a general model to predict the allosteric modulatory potential of molecular
candidates for a myriad of target receptors, organisms, cell lines, and biological activity parameters based on perturbation theory
(PT) ideas and machine learning (ML) techniques, abbreviated as ALLOPTML. By doing so, ALLOPTML shows high specificity
Sp = 89.2/89.4%, sensitivity Sn = 71.3/72.2%, and accuracy Ac = 86.1%/86.4% in training/validation series, respectively. To the best
of our knowledge, ALLOPTML is the first general-purpose chemoinformatic tool using a PTML-based model for the multioutput
and multicondition prediction of allosteric compounds, which is expected to save both time and resources during the early drug
discovery of allosteric modulators.
KEYWORDS: Allosteric modulators, artificial neural networks, big data, ChEMBL, machine learning, Melanostatin, multitarget models,
perturbation theory
1. INTRODUCTION
Dopamine receptors belong to a complex monoaminergic family
Received: October 22, 2020
of G protein-coupled receptors (GPCRs) represented by five
distinct receptors (D1−5 receptors),1 which are grouped into D1-
like (related to excitatory neurotransmission and composed by
D1 and D5 isoforms) and D2-like receptors (associated with
inhibitory neurotransmission, comprising D2, D3, and D4
Accepted: December 7, 2020
https://dx.doi.org/10.1021/acschemneuro.0c00687
© XXXX American Chemical Society
ACS Chem. Neurosci. XXXX, XXX, XXX−XXX
A

ACS Chemical Neuroscience
pubs.acs.org/chemneuro
Research Article
isoforms).1 These two subgroups differ in their signal trans-
duction, binding profile, and physiological effects.2 Among the
D1−5 receptors, D2-like receptors are considered of utmost
importance due to their intrinsic role in the pathophysiology of
neurological and psychiatric disorders.3 Alternative splicing
leads to generation of two distinct D2 receptors (D2Rs): D2 short
(D2S) and D2 long (D2L) isoforms, which are associated with
presynaptic and postsynaptic populations of D2R, respectively.4
The difference between these two splicing isoforms consist of 29
amino acid residues located in the III intracellular loop,
responsible for the G protein coupling.4 The design of ligands
with receptor specificity is considered a herculean task due to the
high structural homology shared by the members of D1-like or
D2-like receptors.3 For example, while D2R is the main target to
control the positive symptoms of schizophrenia, none of
antipsychotics approved are capable to discriminate D2 from
D3 receptors.3 In this sense, allosteric regulation is considered a
key pharmacological approach to selectively modulate the
responsiveness of a specific receptor subtype with negligible or
absent interference on the other subtypes. For this reason,
allosteric modulators may contribute to the development of new
selective and less toxic therapies than those focused on the
orthosteric sites.5 Allosteric modulators bind to a topological
region of the receptor (i.e., an allosteric site) different from the
binding site used by endogenous ligands (i.e., the orthosteric
site). Upon binding, allosteric modulators induce structural
modification that can either stabilize the active form of the
receptor (positive modulation), inactivate it (negative modu-
lation), or have no influence on the ligand−receptor binding
(silent or neutral modulation).6
Positive allosteric modulators (PAMs) operate by either
lowering the energy barrier necessary to achieve the active
conformation of a given protein, by stabilizing its active
conformation (preventing it from adopting nonactive con-
formations), or both.7 PAMs do not display any activity or
pharmacological effect in the absence of the endo/exogenous
agonists, but when combined with an orthosteric agonist, they
increase its efficacy, thus minimizing the overall side-effect
profile of the agonist.7
An example of a PAM with intrinsic D2 selectivity is the
neuropeptide L-prolyl-L-leucylglycinamide,8−12 also known as
Melanostatin8 or as melanocyte stimulating hormone release
inhibiting factor 1 (MIF-1, Figure 1).11
show that MIF-1 potentiates the binding of [3H]-NPA and
[3H]-quinpirrole in a dose-dependent manner to D2L, D2S, and
D4 receptor subtypes to give a bell-shaped dose−response
curve,15 without enhancing the binding to D1 and D3
receptors.8,14 Mechanistically, it is suggested that MIF-1
stabilizes D2R into a high-affinity state, making this receptor
more sensitive to agonists.14,16,17 It is speculated that MIF-1 may
have a direct interaction with a domain of the receptor other
than its orthosteric site or that it may display an indirect
mechanism associated with the increase of the rate of GTP
hydrolysis promoted by G protein.18 However, this last
hypothesis seems less likely as the modulatory effect is not
observed in other G protein-coupled receptors, such as α2
adrenergic receptors, or with Gs-coupled receptors, such as
D1R family, thus showing D2R specificity.8
Regarding its PAM activity and D2R selectivity, MIF-1 has
been proposed as an attractive drug candidate for further
development in the treatment of CNS-related neurodegener-
ative diseases associated with dopamine receptor sensitivity,
such as Parkinson’s disease (PD), schizophrenia, tardive
dyskinesia, and Gilles de la Tourette syndrome.19 For example,
MIF-1 is widely acknowledged for alleviating the behavioral
changes in animal models of PD.20,21 Furthermore, this
neuropeptide also has the ability to interact with other drugs
by modifying their effects22 (e.g., MIF-1 antagonizes morphine-
induced catalepsy as well as dyskinesia induced by antipsychotic
drugs).23−25 X-ray crystallography26 and structure−activity
relationship studies using constrained peptidomimetics27
suggest that MIF-1 adopts a type-II β-turn formed by
intramolecular hydrogen bonding between the carbonyl oxygen
atom of the L-proline residue and the trans amide proton of the
glycinamide residue (depicted as dashed blue line in Figure 1).
The presence of the C-terminal carboxamide is considered a key
pharmacophore for the modulatory activity of MIF-1 and its
active peptidomimetics, however an extended bioactive
conformation is also suggested.16 Modifications in the proline
residue of MIF-1 have long been recognized to render bioactive
peptidomimetics,28 indicating that this residue may be suitable
for chemical derivatization and the development of novel
bioactive PAMs.10−12
In this work, we aimed to explore 2-furoic acid (2-Fu) as
proline surrogate for the assembly and pharmacological
evaluation of a novel class of MIF-1 peptidomimetics as PAMs
of D2R. In addition, we intended to use the experimental data to
construct a general-purpose computational model based on
perturbation theory (PT) ideas and machine learning (ML)
techniques to predict allosteric modulators for a myriad of
receptor targets, organisms, cell lines, and biological activity
parameters, abbreviated as ALLOPTML.
2. RESULTS AND DISCUSSION
Design and Chemistry. Proline has long been established
as a nonessential amino acid residue for the activity of MIF-1
analogues,28 thereby offering the possibility for the design of
novel MIF-1 peptidomimetics with PAM activity and D2
selectivity. For example, picolinic acid has been successfully
used as proline surrogate for the assembly of MIF-1 analogues
with PAM activity and suitable toxicological profile.11 These
findings led us to hypothesize that heterocyclic scaffolds may be
tolerated at the putative allosteric binding site of D2R. In this
sense, to gain further insights into the architecture of D2R
allosteric pocket, the use of different heteroaromatic scaffolds
may allow to further rationalize the chemical features required
Figure 1. Melanostatin (MIF-1) structure denoting a type-II β-turn as
the bioactive conformation assisted by an intramolecular hydrogen
bond (dashed blue line).
MIF-1 is formed by exocyclic cleavage of oxytocin hormone
and displays several biological activities within the central
nervous system (CNS),13 acting mainly as a neuronal
modulating agent in the nigrostriatal pathway.14 Pharmacolog-
ically, MIF-1 interacts with D2R increasing the effect of
dopaminergic agonists on these receptors such as N-
propylnorapomorphine (NPA), apomorphine, and 2-amino-
6,7-dihydroxy-3,4-tetrahydronaphthalene (ADTN).8 Studies
B
https://dx.doi.org/10.1021/acschemneuro.0c00687
ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

ACS Chemical Neuroscience
pubs.acs.org/chemneuro
Research Article
a
Scheme 1. Synthesis of 2-Furoyl-Based Peptidomimetics 4−7(a,b)
a
Reagents and conditions: (i) DIEA, TBTU, CH2Cl2; (ii) LiOH, MeOH/H2O followed by H2SO4 1 M; (iii) NH3 (g), MeOH.
for the development of pharmacologically active PAMs.
Considering that proline contains a pyrrolidine heterocycle,
five-membered heteroaromatic rings such as furan scaffolds may
represent suitable proline bioisosters for the design of MIF-1
peptidomimetics and thus providing useful information about
the size and the nature of the hydrogen bonding at the D2R
allosteric binding site. To the best of our knowledge, the
application of furan moieties for the development of MIF-1
derivatives have not yet been reported. In this sense, 2-Fu was
investigated for the assembly of a novel class of MIF-1
peptidomimetics.
For the preparation of 2-furoyl-based MIF-1 peptidomi-
metics, a diversity-oriented synthesis (DOS) strategy was
employed to create a set of structure-related amino acid
combinations at the central and C-terminal positions. For this
purpose, L-leucine was replaced by L-valine, whereas glycine was
substituted by L-alanine. The synthetic route for the preparation
of 2-furoyl peptidomimetics is depicted in Scheme 1.
detection of acyclic side products (TLC, NMR) under this
protocol, as corroborated by the NMR data (1H, 13C{1H}, and
DEPT-135) shown in Figures S5−S8 of Supporting Informa-
tion. Carboxylic acids 3(a,b) were then coupled with either
glycine (a) or L-alanine (b) methyl esters using the same peptide
coupling protocol, delivering 4(a,b)/5(a,b) in good to very
good yields (79−94%).
Finally, pseudopeptides 4(a,b)/5(a,b) were converted into
the corresponding C-terminal carboxamides 6(a,b)/7(a,b)
upon treatment with gaseous ammonia in methanol, prepared
in situ by reacting Ca(OH)2 and NH4Cl at 100 °C. The reactions
proceeded cleanly, albeit it took about 48 h for completion
(TLC, NMR). After the removal of the volatiles, C-terminal
carboxamides 6(a,b)/7(a,b) were obtained in practically
quantitative yields (94−100%), without the need of chromato-
graphic purifications. Following this synthetic route, 2-furoyl
peptidomimetics were successfully prepared in high global yields
(69−84%).
Pharmacology. Eight novel MIF-1 peptidomimetics were
tested for their ability to potentiate the maximum binding of the
radiolabeled agonist N-propylnorapomorphine, [3H]-NPA, to
cloned human dopamine D2S receptors, and their activity was
compared to that of MIF-1 as described in the literature.8
Peptidomimetics 4−7(a,b) were tested at eight different
concentrations (between 1 pM and 10 μM) following the
protocol previously reported in our research group.10,12 The
whole set of peptidomimetics was evaluated in silico to rule out
the possibility of these ligands acting as promiscuous pan-assay
interference compounds (PAINS) and/or aggregators.31 The
experimental results obtained for the binding assays are shown in
Table 1 and Figure 2.
Starting from 2-Fu, the preparation of amides 2(a,b) was
performed by peptide coupling with L-leucine and L-valine,
respectively (Scheme 1). For that purpose, TBTU was used as
the peptide coupling reagent in the presence of Hunig’s
̈
base.10,12,29 After chromatographic purification, compounds
2(a,b) were obtained in high yields (96%).
Next, compounds 2(a,b) were saponified using LiOH
followed by acidic workup with 1 M H2SO4 to obtain the
corresponding carboxylic acids 3(a,b). Because furan system is
sensitive to low pH, at which may result in acid-catalyzed ring
opening phenomena,30 special attention was required during the
acidic workup. Therefore, the pH was carefully adjusted to 4 to
prevent the formation of side products. Both peptide products
3(a,b) were isolated in very good yields (91−93%) with no
C
https://dx.doi.org/10.1021/acschemneuro.0c00687
ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

ACS Chemical Neuroscience
pubs.acs.org/chemneuro
Research Article
host cells used for expressing human D2S receptors, which differ
from the ones originally reported.8
Table 1. Experimental Results for MIF-1 Derivatives 4-7(a,b)
in the D2R Binding Assays
The C-terminal carboxamide is widely recognized as a key
pharmacophore for the activity of MIF-1 and its peptidomi-
metics,33,9 which is crucial for these compounds to adopt the
postulated type-II β-turn bioactive conformation.33,9 However,
Saitton and co-workers suggested the possibility for an extended
bioactive conformation of MIF-1 through the development of
bioactive peptidomimetics incorporating a 2,3,4-trisubstituted
pyridine scaffold as leucyl mimetic preventing it from adopting
the postulated type-II β-turn.16 Like MIF-1, peptidomimetic 6a
displays the L-leucylglycinamide dipeptide motif; therefore, a
type-II β-turn conformation is envisioned for this compound.
Interestingly, the simple substitution of proline residue by 2-Fu
proved to effectively mimic the MIF-1 modulatory activity at
D2R allosteric binding site. The activity of 6a is thus considered
of utmost interest, because it endorses previous findings that
achiral and heteroaromatic scaffolds are tolerated in the putative
MIF-1 binding pocket, establishing proline as a key residue for
derivatization in the development of MIF-1 analogues with PAM
activity.
In Vitro Neurotoxicity Profile of 2-Furoyl MIF-1
Peptidomimetics. Low nonspecific cytotoxicity is an
important feature of lead compounds during early drug
development. Therefore, peptidomimetics 4−7(a,b) were
evaluated to assess their neurotoxic profiles using culture of
motor cortex neurons from 19-day-old Wistar-Kyoto (WKY) rat
embryos.11,34 Hydrogen peroxide, H2O2, was selected as a
standard control for neurocytotoxity.48,72,35 In fact, the majority
of the in vitro neurotoxicity studies carried out in cortical
Cpnd
% [3H]-NPAmax
AUC
Log (1/Cmax)
4a
4b
5a
5b
6a
6b
7a
7b
11
9
11
6
53.72
34.35
27.25
9.36
12.98
0
2.76
0
44.19
10
12
10
12
11
nd
12
nd
10
11
nd
a
3
nd
22
MIF-1
a
nd = not detected.
A statistically significant enhancement (P < 0.05; ANOVA
test; post hoc Dunnet T3 test) of the [3H]-NPA response was
observed for peptidomimetic 6a (Figure 2). The maximum
specific binding of [3H]-NPA elicited by 6a was 11 1% at 10
pM, while MIF-1 (positive control) displays [3H]-NPA binding
of 18 9% at the same concentration and maximum effect of 22
9% at 100 pM, as shown in Table 1 and Figure 2. In respect
with the other peptidomimetics, no statistically significant
increase in the [3H]-NPA binding was observed, therefore no
conclusion can be drawn for these compounds. In this work it
was found that MIF-1 displays an increase in the [3H]-NPA
binding at lower concentrations than those previously reported
by Verma and co-workers.8 Because allosteric modulators are
sensitive to environmental changes elicited by endogenous
agonists on GPCRs,32 this deviation may be attributed to the
Figure 2. Modulation of the [3H]-NPA binding by 2-furoyl-based MIF-1 peptidomimetics 4−7(a,b) at eight different concentrations. Points represent
the mean standard deviation (vertical bars) of three independent experiments carried out with duplicates. *P < 0.05 (ANOVA test; posthoc Dunnet
T3 test).
D
https://dx.doi.org/10.1021/acschemneuro.0c00687
ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

ACS Chemical Neuroscience
pubs.acs.org/chemneuro
Research Article
neurons uses H2O2 as an effective inducer of oxidative stress.36
The use of H2O2 creates an imbalance between reactive oxygen
species (ROS) levels and scavenging processes.36 Excessive cell
exposure to free radicals subsequently leads to neuronal death.
In this assay, cellular viability was estimated based on the ability
of cells to reduce MTT after a 24 h incubation period with the
test compounds (100 μM). The percentage of cell viability was
compared with the dimethyl sulfoxide (DMSO, negative
proteins. For this purpose, the pharmacological and biological
data disclosed in this work for peptidomimetics 4−7(a,b) was
used to train the model along with the 21 439 cases from 8984
different ChEMBL assays of 8957 allosteric compounds for 79
different proteins. For this task 12 different variables (c0−11
)
were considered from the ChEMBL database big data (BD) sets
of compounds covering 38 biological activity parameters (c0;
e.g., efficacy, potency, intrinsic activity, IC50, Ki, Km), organisms
of the protein target (c1; e.g., human, rat, mouse), organism of
assay (c2; e.g., rat, swine, herpesvirus 5), different cellular lines
(c3; e.g., CHO, HEK293), and so on. The general workflow used
in this paper to obtain the PTML model is depicted in Figure 4.
control). All results are expressed as mean
independent experiments (Figure 3).
SEM of five
Figure 3. Neurotoxicity assays for peptidomimetics 4−7(a,b) in
cultured neurons from embryonic motor cortices of fetal Wistar-Kyoto
rats. Data are the mean SEM of n = 5 independent experiments at 100
μM. Negative control = DMSO; positive control = 100 μM H2O2.
*Level of statistical significance = P < 0.05 vs DMSO condition by t test.
Figure 4. General workflow used to obtain the PTML model.
So far, finding allosteric binding sites and lead compounds has
been largely serendipitous, accomplished by medicinal chemists
through preclinical assays performing high-throughput screen-
ing with a high number of combinations of experimental
conditions (ca,j). In addition, the existence of allo-network drugs,
i.e., allosteric compounds that do not act directly over the
pharmacological target but over a second protein which in turn
interacts directly or indirectly with the target protein, makes this
task even more challenging.38−40 In addition to experimental
techniques, computational techniques can be used to predict
new drugs for different targets.41 In this sense, many researchers
have developed cheminformatics models for the discovery of
new bioactive compounds. Nevertheless, almost all these models
are designed for homologous series of compounds, one target,
and/or a single cell line. Other models, instead, use
heterogeneous series of compounds, although they are unable
to incorporate information about the target, cell line, or
organism of assay, among other features.42−57 This drawback
can be partially faced by ML techniques through which diverse
molecular descriptors, encoding the chemical structure of
different drugs, can be calculated.58−74 ML techniques have
already been carried out to predict allosteric pockets on
proteins.75 Unfortunately, this method failed to account for
large BD sets of preclinical assays, such as the ChEMBL
database, which are difficult to study due to the high complexity
of the data in addition to the large volume of highly
heterogeneous preclinical assays.76,77
As expected, the stress inducer (100 μM H2O2) reduced the
cell survival to 78.51 4.97% compared to the nontreated cells
control (P < 0.05). In this assay compounds 5a and 5b display a
statistically significant decrease in neuronal survival (91.34
1.50 and 91.13 3.20%, respectively, P < 0.05), in comparison
with DMSO condition, suggesting these peptidomimetics
exhibit a marginal neurotoxicity. The cell viability was not
affected by the presence of the remaining test compounds when
compared to the control (DMSO). These results indicate that 2-
furoyl-based MIF-1 analogues display a safe profile up to 100
μM.
General-Purpose Model for Allosteric Modulators
(ALLOPTML). The application and combination of computa-
tional methods and predictive software tools in medicinal
chemistry are considered of utmost interest, enabling one to
identify, prioritize, and optimize novel chemotypes with desired
pharmacological activity profiles.37 Computational models and
algorithms capable to make predictions on the pharmacological
and biological activity using different conditions of assay with
minimal computational cost are thus highly desired. These
powerful computer-aided models are useful to guide molecular
design and decision making in drug discovery, boosting the
identification of lead compounds avoiding time-consuming
assays and saving resources; therefore, it is considered an
essential pillar of modern drug research.37
In the absence of a robust and general-purpose model for the
multioutput and multicondition prediction of allosteric
compounds in medicinal chemistry, we sought to fill this gap.
To achieve this goal, we developed and implemented
ALLOPTML, abbreviation used to describe a novel predictive
model for allosteric modulators based on perturbation theory
(PT) ideas and machine learning (ML) techniques, which may
be used for several types of systems and for a hundred different
To tackle this problem we have associated PT and ML
modeling.70,78 In fact, PTML (PT + ML) models have been
used in many fields of knowledge such as medicinal chemistry,
proteomics, or nanotechnology for modeling large data sets with
BD features.34,79−91 Probably due to the potential benefits of
allosteric drugs, a renewed interest in allostery has led to the
development of a number of computational approaches to
understanding allostery.92 Nevertheless, to the best of our
E
https://dx.doi.org/10.1021/acschemneuro.0c00687
ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

ACS Chemical Neuroscience
pubs.acs.org/chemneuro
Research Article
Table 2. PTML Predictive Study of New Compounds
c0
EC50 (nM)
E
max (%)
% max
inhibition (%)
assay
c1
c2
c3
c4
4a
4b
5a
5b
6a
6b
H. sapiens
H. sapiens
HEK293
H. sapiens
H. sapiens
CHO
R. norvegicus
R. norvegicus
null
mGluR2
0.404
0.388
0.413
0.397
0.397
0.378
0.403
0.392
0.368
H. sapiens
H. sapiens
HEK293
mGluR5
0.796
0.785
0.802
0.791
0.791
0.778
0.796
0.787
0.770
a
a
mGluR5
mGluR2
compound
0.042
0.039
0.043
0.041
0.041
0.038
0.042
0.040
0.036
0.087
0.082
0.090
0.085
0.085
0.078
0.087
0.083
0.076
7a
7b
MIF-1
a
mGluR5 = metabotropic glutamate receptor 5 (CHEMBL2564 in rat and CHEMBL3227 in human); mGluR2 = metabotropic glutamate receptor
2 (CHEMBL5137 in human and CHEMBL2581 in rat).
knowledge, general-purpose PTML models for multiple
preclinical assays of allosteric compounds have not yet been
reported.
In training series, explored in detail in Supporting
Information, the model presented high values of specificity, Sp
= 89.2%; sensitivity, Sn = 71.3%; and accuracy, Ac = 86.1%.
Likewise, the MMA model was also reliable in external validation
series: Sp = 89.4%, Sn = 72.2%, and Ac = 86.4%. These values are
in the range considered as useful for classification models with
application in medicinal chemistry.93
In Table 2, the results obtained for the selected examples
including different c0 = activity parameter (units) such as
effective 50% concentration EC50 (nM), maximun effect Emax
(%), and inhibition (%) = inhibition (%) are listed. The
examples also involve two different organisms (human and rat)
expressing the protein target and organisms of assays (c1 and c2).
We also changed the condition c3 = cell line of assay (HEK293,
CHO) and the condition c4 = target proteins (mGluR5 and
mGluR2). Interestingly, the model predicts a similar behavior
for all the series of compounds in the same conditions, which is
reasonably for a homogeneous series of compounds. The model
predicts low probabilities of having interesting EC50 (nM)
values for mGluR5 but high probabilities of inhibition (%) for
the same receptor. This may be indicative that this set of
peptidomimetics display activity on human mGluR5 but
perhaps at higher concentrations.
Computational Study on the 2-Furoyl-Based MIF-1
Peptidomimetics. After training and validating the AL-
LOPTML model (Supporting Information), we decided to
illustrate how to use it by providing a detailed step-by-step guide
to calculate the posterior probabilities with a practical example.
For this purpose, the series of peptidomimetics 4−7(a,b) were
studied in different assay conditions cj (target proteins, cell lines,
assay organisms). First, the chemical structures of peptidomi-
metics 4−7(a,b) were drawn using ChemDraw software to
obtain their SMILE codes. Then, the file containing all SMILE
codes was created and uploaded to Virtual Computational
Chemistry Laboratory (VCCLAB, http://www.vcclab.org,
2005)94 in order to calculate the molecular descriptors Dk
(Log P, PSA) for each compound. Because the goal was to
illustrate how to operate the model, only a few assay conditions
were selected. After that, the function of reference f(vij)ref
(expected probability of activity) values for each assay were
obtained from Supporting Information. Using this file, the
expected or average values <Dk(cj)> of the molecular descriptors
Dk were also obtained for all compounds assayed on these
conditions. With these values in hand, we calculated the MMA
operators ΔDk(cj) for each compound vs every set of
experimental conditions selected. Subsequently, f(vij)ref and
ΔDk(cj) values were substituted into the equation of the model
to obtain the values of the output function f(vij)calc. Lastly, these
values were used to calculate the posterior probabilities
p(f(vij)pred = 1), with which the compounds may be considered
as candidates for assay according to our criteria. The equation
used to calculate these probabilities was a sigmoid function
p(f(vij)pred = 1) = 1/(1 + (π0/π1)·Exp(−f(vij)calc). Recall that π1
= 1 − π0 are the probabilities (defined before to seek the model)
with which a compound may be considered as active f(vij)pred = 1
or nonactive and f(vij)pred = 0, a priori.
3. CONCLUSIONS
The use of 2-Fu for the assembly of MIF-1 peptidomimetics is
unprecedented. In this work, a small library of MIF-1
peptidomimetics bearing 2-Fu as proline surrogate was
successfully obtained and pharmacologically evaluated toward
D2R. Remarkably, compound 6a stands out as a D2R modulator
capable of increasing the maximal [3H]-NPA response at 10 pM
(11 1%), comparable to the effect of MIF-1 (18 9%) at the
same concentration and with no neurotoxic effect up to 100 μM
in cortex neurons of Wistar-Kyoto rat embryos. These findings
corroborate that D2R allosteric binding site can accommodate
heteroaromatic rings at N-terminal position of MIF-1, establish-
ing furan as a suitable heterocyclic scaffold to be used as proline
mimetic for the assembly of MIF-1 analogues. Consequently, 6a
is considered a valid lead compound for further optimization,
paving the way for the discovery of adequate candidates to treat
CNS-related diseases. Exploration of other furan motifs is
currently underway in our research group and will be disclosed
in due time.
The topology of a complex network of allosteric modulators
was constructed and studied delivering a predictive model,
ALLOPTML, which shows remarkable specificity Sp = 89.2/
89.4%, sensitivity Sn = 71.3/72.2%, and accuracy Ac = 86.1%/
86.4% in training/validation series. The linear model with
multicondition MAs proved to be superior to a linear model with
simple condition MAs. To the best of our knowledge,
ALLOPTML is the first general-purpose medicinal chemistry
F
https://dx.doi.org/10.1021/acschemneuro.0c00687
ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

ACS Chemical Neuroscience
pubs.acs.org/chemneuro
Research Article
performed in a rotatory evaporator to afford the desired compound
without further purifications.
tool using PTML model for the multioutput and multicondition
prediction of allosteric compounds, which is expected to be a
valuable chemoinformatic tool to assist in the early drug
discovery, saving both time and resources.
General Protocol for C-Terminal Carboxamide: Synthesis of
Pseudotripeptides 6(a,b) and 7(a,b). Pseudotripeptide methyl
ester (1 equiv) was charged in a reaction flask and dissolved in MeOH
p.a. (40 mL). The solution was cooled at 0 °C using an ice bath and
gaseous ammonia (generated in situ by mixing NH4Cl and Ca(OH)2
2:1 (w/w) at 100 °C using an oil bath) was bubbled into the reaction
flask. The ammonia stream was maintained for 1 h, and the reaction was
left stirring at room temperature (ca. 25 °C) until completion (ca. 48 h,
TLC). Removal of the volatiles were performed in a rotatory evaporator
to afford the desired compound without the need of chromatographic
purification.
Methyl 2-Furoyl-L-leucinate (2a). Following the general protocol
for peptide coupling, DIEA (4.66 mL, 26.8 mmol), TBTU (3.15 g, 9.81
mmol), and methyl L-leucinate hydrochloride (1.94 g, 10.7 mmol) was
added to a solution of 2-Fu (1.00 g, 8.92 mmol) in anhydrous CH2Cl2
(30 mL). After the typical workup, the crude oil was chromatographed
using EtOAc as eluent, affording 2.05 g of 2a as a white solid. Yield:
96%. m.p.: 83−87 °C. Rf: 0.84 in EtOAc. [α]2D4: +8.3 (c1, CHCl3). 1H
NMR (CDCl3, 400 MHz) δ ppm: 7.56−7.36 (m, 1H, H-5), 7.23−7.00
(m, 1H, H-3), 6.77 (d, J = 8.0 Hz, 1H, CONH), 6.59−6.38 (m, 1H, H-
4), 4.82 (dt, J = 8.6, 4.9 Hz, 1H, HLeu-2), 3.76 (s, 3H, CO2CH3), 1.89−
1.52 (m, 3H, HLeu-3 + HLeu-4), 1.11−0.82 (m, 6H, HLeu-5). 13C NMR
and DEPT (CDCl3, 101 MHz) δ ppm: 173.4 (C, CO2CH3), 158.1 (C,
CONH), 147.6 (C, C-2), 144.2 (CH, C-5), 114.8 (CH, C-3), 112.3
EXPERIMENTAL METHODS
■
Chemistry. All chemicals were of reagent grade and used without
further purifications: 2-Furoic acid was obtained from Alfa Aesar; O-
(Benzotriazol-1-yl)-N,N,N′,N′-tetramethyluronium tetrafluoroborate
(TBTU), H-L-Leu-OMe, H-L-Val-OMe, H-L-Ala-OMe, and H-Gly-
OMe were obtained from Bachem, and N,N-diisopropylethylamine
(DIEA) was obtained from Sigma-Aldrich. All air sensitive reactions
were carried out under argon atmosphere. Analytical TLC was carried
out on precoated silica gel plates (Merck 60 F254, 0.25 mm) using UV
light and an ethanolic solution of phosphomolybdic acid (followed by
gentle heating) for visualization. Flash chromatography was performed
on silica gel (Merck 60, 230−240 mesh).
Apparatus. Mass spectra were recorded on a LTQ OrbitrapTM XL
hybrid mass spectrometer (Thermo Fischer Scientific, Bremen,
Germany) controlled by LTQ Tune Plus 2.5.5 and Xcalibur 2.1.0
(Centro de Materiais da Universidade do Porto, CEMUP). The
capillary voltage of the electrospray ionization source (ESI) was set to
3.1 kV. The capillary temperature was set at 275 °C. The sheath gas was
set at 6 (arbitrary unit as provided by the software settings). The
capillary voltage was set at 35 V and the tube lens voltage set at 110 V.
1H and 13C NMR spectra were recorded at CEMUP with a Bruker
Avance III 400 at 400.15 and 100.62 MHz, respectively. The 1H NMR
spectra were calibrated using the residual protic signals from the
deuterated solvents while 13C NMR spectra were calibrated directly
from the deuterated solvents (CDCl3: δH = 7.26, δC = 77.16; CD3OD:
δH = 3.31, δC = 49.00; and DMSO-d6: δH = 2.50, δC = 39.52)95 and are
reported in parts per million (ppm). The nomenclature used for the
assignment of protons and/or carbons for each α-amino acid residue in
peptidomimetics was made using a single letter system in subscript for
the amino acid residue (Leu = L-leucine; Val = L-valine, Ala = L-alanine;
Gly = glycine) and indicating the proton (or group of protons) and/or
the carbons in the structures by starting the numeration at the carbonyl
carbon of the main chain of each α-amino acid residue. Optical
rotations were measured on a JASCO P-2000 thermostated polarimeter
using a sodium lamp and are reported as follows: αθD expressed in (deg)
(dm−1) (g−1), in which θ is the temperature in Celsius and c (g/100 mL,
solvent). Melting points were determined using a STUART Scientific,
model SMP1, and are not corrected. Solvents were evaporated in a
(CH, C-4), 52.4 (CH3, CO2CH3), 50.4 (CH, CLeu-2), 41.9 (CH2, CLeu
-
3), 25.0 (CH, CLeu-4), 22.9 (CH3, CLeu-5), 22.0 (CH3, CLeu-5). HRMS
+
(ESI-TOF) m/z: [M + H]+ Calcd for C12H18NO4 , 240.123 03; found,
240.123 12.
Methyl 2-Furoyl-L-Valinate (2b). Following the general protocol for
peptide coupling, DIEA (7.00 mL, 40.2 mmol), TBTU (4.7340 g,
14.743 mmol), and methyl L-valinate hydrochloride (2.6960 g, 16.084
mmol) was added to a solution of 2-Fu (1.5022 g, 13.403 mmol) in
anhydrous CH2Cl2 (30 mL). After the typical workup, the crude oil was
chromatographed using EtOAc as eluent, affording 2.8981 g of 2b as a
white solid. Yield: 96%. m.p.: 75−78 °C. Rf: 0.89 in EtOAc. [α]2D1: +9.0
(c1.015, CHCl3). 1H NMR (CDCl3, 400 MHz) δ ppm: 7.45 (dd, J =
1.8, 0.8 Hz, 1H, H-5), 7.10 (dd, J = 3.5, 0.8 Hz, 1H, H-3), 6.80 (d, J = 8.3
Hz, 1H, CONH), 6.48 (dd, J = 3.5, 1.8 Hz, 1H, H-4), 4.70 (dd, J = 9.0,
5.0 Hz, 1H, HVal-2), 3.74 (s, 3H, CO2CH3), 2.24 (dhept, J = 6.9, 5.1 Hz,
1H, HVal-3), [0.98 (d, J = 6.9 Hz, 3H), 0.95 (d, J = 6.9 Hz, 3H), HVal-4].
13C NMR and DEPT (CDCl3, 101 MHz) δ ppm: 172.4 (C, CO2CH3),
158.2 (C, CONH), 147.7 (C, C-2), 144.2 (CH, C-5), 114.8 (CH, C-3),
112.3 (CH, C-4), 56.8 (CH3, CO2CH3), 52.3 (CH, CVal-2), 31.6 (CH,
CVal-3), 19.1 (CH, CVal-4), 17.9 (CH, CVal-4). HRMS (ESI-TOF) m/z:
Buchi rotavapor model R-210.
̈
+
General Protocol for Peptide Coupling: Synthesis of 2(a,b),
4(a,b), and 5(a,b). The appropriate carboxylic acid (1 equiv) was
charged in a round-bottom flask and dissolved in anhydrous CH2Cl2
(50 mL), under argon atmosphere, followed by iterative addition of
DIEA (3 equiv) and TBTU (1.1 equiv), with stirring for 30 min at RT.
Finally, the appropriate amine (1.2 equiv) was added, and the reaction
was left stirring for approximately 1 h. After completion of the reaction
(TLC), CH2Cl2 was removed in vacuo, and the crude oil was dissolved
with EtOAc (100 mL) and transferred to a separatory funnel. The
organic layer was washed with saturated solutions of NaHCO3 (3 × 100
mL) and NaCl (100 mL) and dried over anhydrous Na2SO4. After
filtration, the solvent was removed in vacuo, and the crude oil was
chromatographed as specified for each peptide product.
[M + H]+ Calcd for C11H16NO4 , 226.107 38; found, 226.107 38.
2-Furoyl-L-leucine (3a). Following the general protocol for
saponification, LiOH (0.2414 g, 10.08 mmol) was added to a solution
of 2a (0.8034 g, 3.360 mmol) in MeOH (30 mL). After the typical
workup, 0.7038 g of 3a was obtained as a white solid. Yield: 93%. m.p.:
81−84 °C. Rf: 0.09 in EtOAc. [α]2D4: +8.3 (c1, CHCl3). H NMR
1
(CDCl3, 400 MHz) δ ppm: 8.95 (br s, 1H, CO2H), 7.41 (d, J = 1.3 Hz,
1H, H-5), 7.10 (d, J = 3.4 Hz, 1H, H-3), 7.06 (d, J = 7.9 Hz, 1H,
CONH), 6.42 (dd, J = 3.4, 1.7 Hz, 1H, H-4), 4.76−4.61 (m, 1H, HLeu
-
2), 1.80−1.57 (m, 3H, HLeu-3 + HLeu-4), 1.10−0.80 (m, 6H, HLeu-5).
13C NMR and DEPT (CDCl3, 101 MHz) δ ppm: 177.6 (C, CO2H),
158.7 (C, CONH), 147.3 (C, C-2), 144.6 (CH, C-5), 115.3 (CH, C-3),
112.3 (CH, C-4), 51.6 (CH, CLeu-2), 41.5 (CH2, CLeu-3), 25.0 (CH,
CLeu-4), 23.1 (CH3, CLeu-5), 21.8 (CH3, CLeu-5). HRMS (ESI-TOF)
General Protocol for Saponification: Synthesis of Pseudodi-
peptides 3(a,b). Pseudodipeptide methyl ester (1 equiv) was charged
in a round-bottom flask, dissolved in MeOH (50 mL), and the solution
was cooled at 0 °C using an ice bath with magnetic stirring. Then, LiOH
(3 equiv) was slowly added during a period of 30 min and left stirring
until complete consumption of the starting material (TLC). The
solvent was removed in vacuo (at 40 °C), and the crude solid was then
dissolved in water (20 mL), cooled at 0 °C using an ice bath with
magnetic stirring. After that, 1 M H2SO4 was carefully added until pH =
4. The solvent was distilled off (at 40 °C), and the crude solid was then
triturated with hot CHCl3 and filtered. Removal of the volatiles were
−
m/z: [M − H]− Calcd for C11H14NO4 , 224.092 83; found, 224.094 82.
2-Furoyl-L-valine (3b). Following the general protocol for
saponification, LiOH (0.2558 g, 10.68 mmol) was added to a solution
of 2b (0.8019, 3.560 mmol) in MeOH (30 mL). After the typical
workup, 0.6843 g of 3b was obtained as a white solid. Yield: 91%. m.p.:
65−70 °C. Rf: 0.05 in EtOAc. [α]2D0: +67.5 (c1.09, CHCl3). 1H NMR
(CDCl3, 400 MHz) δ ppm: 10.12 (br s, 1H, CO2H), 7.42 (d, J = 0.9 Hz,
1H, H-5), 7.10 (d, J = 3.3 Hz, 1H, H-3), 7.04 (d, J = 8.5 Hz, 1H,
CONH), 6.42 (dd, J = 3.4, 1.7 Hz, 1H, H-4), 4.60 (dd, J = 8.5, 5.1 Hz,
1H, HVal-2), 2.37−2.13 (m, 1H, HVal-3), 0.95 (d, J = 7.1 Hz, 3H, HVal
-
G
https://dx.doi.org/10.1021/acschemneuro.0c00687
ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

ACS Chemical Neuroscience
pubs.acs.org/chemneuro
Research Article
4), 0.93 (d, J = 7.1 Hz, 3H, HVal-4). 13C NMR and DEPT (CDCl3, 101
MHz) δ ppm: 176.2 (C, CO2H), 158.8 (C, CONH), 147.3 (C, C-2),
(0.8381 g, 2.610 mmol), and methyl L-alaninate hydrochloride (0.3975
g, 2.848 mmol) were added to a solution of 3b (0.5012 g, 2.373 mmol)
in anhydrous CH2Cl2 (30 mL). After the typical workup, the crude oil
was chromatographed using EtOAc as eluent, affording 0.6258 g of 5b
as a white solid. Yield: 89%. m.p.: 130−132 °C. Rf: 0.69 in EtOAc. [α]2D0:
−8.1 (c1.01, DMSO). 1H NMR (CDCl3, 400 MHz) δ ppm (rotamers
present 60:40): 7.44−7.40 (m, 1H, H-5), 7.40−7.26 [7.38 (d, J = 6.6
Hz, major), 7.29 (d, J = 6.9 Hz, minor), 1H, CONH], 7.19−6.99 [7.13
(d, J = 9.0 Hz), H-3 + CONH], 6.50−6.39 (m, 1H, H-4), 4.68−4.46
(m, 2H, HVal-2 + HAla-2), [3.71 (s, major), 3.65 (s, minor), 3H,
CO2CH3], 2.26−2.04 (m, 1H, HVal-3), [1.40 (d, J = 7.2 Hz, minor),
144.6 (CH, C-5), 115.3 (CH, C-3), 112.3 (CH, C-4), 57.9 (CH, CVal
-
2), 31.3 (CH, CVal-3), 19.3 (CH, CVal-4), 17.9 (CH, CVal-4). HRMS
−
(ESI-TOF) m/z: [M − H]− Calcd for C10H12NO4 , 210.077 18; found,
210.077 77.
Methyl 2-Furoyl-L-leucylglycinate (4a). Following the general
protocol for peptide coupling, DIEA (2.06 mL, 11.8 mmol), TBTU
(1.3933 g, 4.3393 mmol), and methyl glycinate hydrochloride (0.5942
g, 4.733 mmol) were added to a solution of 3a (0.8767 g, 3.944 mmol)
in anhydrous CH2Cl2 (30 mL). After the typical workup, the crude oil
was chromatographed using EtOAc as eluent, affording 1.1394 g of 4a
as a white solid. Yield: 94%. m.p.: 135−136 °C. Rf: 0.58 in EtOAc. [α]2D0:
−8.5 (c1, CHCl3). 1H NMR (CDCl3, 400 MHz) δ ppm: 7.44−7.40 (m,
1H, H-5), 7.26−7.12 (m, 1H, CONH), 7.12−7.06 (m, 1H, H-3), 7.04−
6.86 (m, 1H, CONH), 6.46 (dt, J = 4.8, 1.7 Hz, 1H, H-4), 4.82−4.64
(m, 1H, HLeu-2), 4.13−3.87 (m, 2H, HGly-2), 3.76−3.63 [3.70 (s), 3.69
(s), 3H, CO2CH3], 1.87−1.53 (m, 3H, HLeu-3 + HLeu-4), 1.04−0.79
(m, 6H, HLeu-5). 13C NMR e DEPT (CDCl3, 101 MHz) δ ppm: [172.5
(C), 172.4 (C), 170.2 (C), CONH + CO2CH3], 158.5 (C, CONH),
147.4 (C, C-2), 144.5 (CH, C-5), 115.0 (CH, C-3), 112.3 (CH, C-4),
1.36 (d, J = 7.3 Hz, major), 3H, HAla-3], 1.04−0.90 (m, 6H, HVal-4). 13
C
NMR and DEPT (CDCl3, 101 MHz) δ ppm (rotamers present):
[173.3 (C), 173.0 (C), 170.9 (C), 170.7 (C), CO2CH3 + CONH],
[158.3 (C), 158.3 (C), CONH], [147.6 (C), 147.6 (C), C-2], [144.4
(CH), 144.4 (CH), C-5], [114.7 (CH), 114.7 (CH), C-3], [112.2
CH), 112.2 (CH), C-4], [57.8 (CH), 57.8 (CH), CVal-2], [52.4 (CH3),
52.4 (CH3), CO2CH3], [48.2 (CH), 48.1 (CH), CAla-2], [31.9 (CH),
31.7 (CH), CVal-3], [19.3 (CH3), 19.1 (CH3), 18.3 (CH3), 18.2 (CH3),
18.1 (CH3), 17.8 (CH3), CVal-4 + CAla-3]. HRMS (ESI-TOF) m/z: [M
+
+ H]+ Calcd for C14H21N2O5 , 297.144 50; found, 297.144 63.
52.4 (CH3, CO2CH3), 51.3 (CH, CLeu-2), 41.3 (2 CH2, CGly-2 + CLeu
-
Methyl 2-Furoyl-L-leucylglycinamide (6a). Following the general
protocol for the synthesis of primary amide, gaseous ammonia was
bubbled into a solution of 4a (0.3090 g, 1.043 mmol) in MeOH p.a. (20
mL), and the system was left stirring for 48 h at rt. After the typical
workup, 0.2921 g of 6a was obtained as a beige solid. Yield: quantitative.
3), 24.8 (CH, CLeu-4), 23.0 (CH3, CLeu-5), 22.1 (CH3, CLeu-5). HRMS
+
(ESI-TOF) m/z: [M + H]+ Calcd for C14H21N2O5 , 297.144 50; found,
297.143 77.
Methyl 2-Furoyl-L-valylglycinate (4b). Following the general
protocol for peptide coupling, DIEA (3.70 mL, 21.3 mmol), TBTU
(2.5089 g, 7.8138 mmol), and methyl glycinate hydrochloride (1.0701
g, 8.5242 mmol) were added to a solution of 3b (1.5004 g, 7.1035
mmol) in anhydrous CH2Cl2 (30 mL). After the typical workup, the
crude oil was chromatographed using EtOAc as eluent, affording 1.5842
g of 4b as a white solid. Yield: 79%. m.p.: 133−136 °C. Rf: 0.60 in
EtOAc. [α]2D2: −38.4 (c1.015, CHCl3). 1H NMR (CDCl3, 400 MHz) δ
ppm: 7.50−7.40 (m, 1H, H-5), 7.32 (br s, 1H, CONH), 7.20−7.00
[7.11 (dd, J = 6.9, 4.2 Hz), 7.09 (br s), H-3 + CONH], 6.49 (dd, J = 3.4,
m.p.: 90−95 °C. Rf: 0.05 in EtOAc. [α]2D0: +161.2 (c1, CHCl3). H
1
NMR (CDCl3, 400 MHz) δ ppm: 8.01−7.77 (m, 1H, CONH), 7.49−
7.32 (m, 2H, H-5 + CONH), 7.12−7.03 (m, 1H, H-3), 6.94 (br s, 1H,
CONH2), 6.48 (br s, 1H, CONH2), 6.42 (dd, J = 3.1, 1.5 Hz, 1H, H-4),
4.74−4.50 (m, 1H, HLeu-2), 3.98 (dd, J = 16.9, 6.1 Hz, 1H, HGly-2),
3.87−3.66 (m, 1H, HGly-2), 1.80−1.58 (m, 3H, HLeu-3 + HLeu-4), 0.99−
0.81 (m, 6H, HLeu-5). 13C NMR and DEPT (CDCl3, 101 MHz) δ ppm:
[173.3 (C), 173.3 (C), 172.5 (C), 159.1 (C), 2 × CONH + CONH2],
147.2 (C, C-2), 144.9 (CH, C-5), 115.1 (CH, C-3), 112.3 (CH, C-4),
52.3 (CH, CLeu-2), 42.9 (CH2, CLeu-3), 41.0 (CH2, CGly-2), 24.9 (CH,
CLeu-4), 23.0 (CH3, CLeu-5), 21.9 (CH3, CLeu-5). HRMS (ESI-TOF)
1.7 Hz, 1H, H-4), 4.63−4.53 (m, 1H, HVal-2), 4.20−3.90 (m, 2H, HGly
-
2), 3.73 (s, 3H, CO2CH3), 2.28−2.15 (m, 1H, HVal-3), 1.12−0.92 (m,
6H, HVal-4). 13C NMR and DEPT (CDCl3, 101 MHz) δ ppm: [171.6
(C), 170.2, CO2CH3 + CONH], 158.5 (C, CONH), 147.5 (C, C-2),
+
m/z: [M + H]+ Calcd for C13H20N3O4 , 282.144 83; found, 282.144 17.
Methyl 2-Furoyl-L-valylglycinamide (6b). Following the general
protocol for the synthesis of primary amide, gaseous ammonia was
bubbled into a solution of 4b (0.3819 g, 1.353 mmol) in MeOH p.a. (20
mL), and the system was left stirring for 48 h at rt. After the typical
workup, 0.3611 g of 6b was obtained as a white solid. Yield:
quantitative. m.p.: 180−183 °C. Rf: 0.20 in EtOAc. [α]2D1: +14.6
(c1.025, DMSO). 1H NMR (DMSO-d6, 400 MHz) δ ppm: 8.30 (t, J =
5.8 Hz, 1H, CONH), 8.04 (d, J = 8.4 Hz, 1H, CONH), 7.87−7.82 (m,
1H, H-5), 7.29−7.15 [7.23 (br s), 7.21 (d, J = 3.5 Hz), 2H, CONHH +
H-3], 7.04 (br s, 1H, CONHH), 6.63 (dd, J = 3.5, 1.8 Hz, 1H, H-4),
4.25 (dd, J = 8.1, 7.7 Hz, 1H, HVal-2), [3.66 (dd, J = 16.7, 6.0 Hz), 3.65
(dd, J = 16.7, 5.5 Hz), 2H, HGly-2], 2.18−2.04 (m, 1H, HVal-3), [0.91 (d,
J = 6.8 Hz), 0.90 (d, J = 6.7 Hz), 6H, HVal-4]. 13C NMR and DEPT
(DMSO-d6, 101 MHz) δ ppm: [171.1 (C), 170.7 (C), 157.8 (C), 2 ×
CONH + CONH2], 147.4 (C, C-2), 145.2 (CH, C-5), 114.0 (CH, C-
3), 111.9 (CH, C-4), 58.3 (CH, CVal-2), 41.8 (CH2, CGly-2), 30.1 (CH,
CVal-3), [19.3 (CH3), 18.6 (CH3), CVal-4]. HRMS (ESI-TOF) m/z: [M
144.5 (CH, C-5), 114.8 (CH, C-3), 112.2 (CH, C-4), 58.1 (CH, CVal
-
2), 52.4 (CH3, CO2CH3), 41.2 (CH2, CGly-2), 31.5 (CH, CVal-3), 19.3
(CH, CVal-4), 18.3 (CH, CVal-4). HRMS (ESI-TOF) m/z: [M + H]+
+
Calcd for C13H19N2O5 , 283.128 85; found, 283.128 78.
Methyl 2-Furoyl-L-leucyl-L-alaninate (5a). Following the general
protocol for peptide coupling, DIEA (3.92 mL, 22.5 mmol), TBTU
(2.6506 g, 8.2550 mmol), and methyl L-alaninate hydrochloride
(1.2570 g, 9.0054 mmol) were added to a solution of 3a (1.6678 g,
7.5045 mmol) in anhydrous CH2Cl2 (30 mL). After the typical workup,
the crude oil was chromatographed using EtOAc as eluent, affording
1.9797 g of 5a as a white solid. Yield: 85%. m.p.: 135−136 °C. Rf: 0.58 in
EtOAc. [α]2D2: +83.7 (c1.16, CHCl3). 1H NMR (CDCl3, 400 MHz) δ
ppm (rotamers present 60:40): [7.43 (dd, J = 1.7, 0.8 Hz, minor), 7.41
(dd, J = 1.7, 0.8 Hz, major), 1H, H-5], [7.11 (dd, J = 3.5, 0.7 Hz, minor),
7.09 (dd, J = 3.5, 0.7 Hz, major), 7.07 (br s), 2H, H-3 + CONH], [6.95
(d, J = 8.6 Hz, major), 6.91 (d, J = 8.6 Hz, minor), 1H, CONH], 6.51−
6.41 (m, 1H, H-4), 4.78−4.66 (m, 1H, HLeu-2), 4.60−4.47 (m, 1H,
HAla-2), [3.72 (s, major), 3.66 (s, minor), 3H, CO2CH3], 1.78−1.59
(m, 3H, HLeu-3 + HLeu-4), [1.40 (d, J = 7.2 Hz, minor), 1.36 (d, J = 7.2
Hz, major), 3H, HAla-3], 0.94−0.90 (m, 6H, HLeu-5). 13C NMR and
DEPT (CDCl3, 101 MHz) δ ppm: [173.3 (C), 173.1 (C), 171.8 (C),
171.6 (C), CONH + CO2CH3], [158.4 (C), 158.3 (C), CONH],
[147.6 (C), 147.5 (C), C-2], 144.4 (CH, C-5), [114.9 (CH), 114.8
(CH), C-3], [112.3 (CH), 112.2 (CH), C-4], 52.5 (CH3, CO2CH3),
[51.3 (CH), 51.2 (CH), CLeu-2], [48.2 (CH), 48.2 (CH), CAla-2], [41.7
(CH2), 41.5 (CH2), CLeu-3], [24.9 (CH), 24.8 (CH), CLeu-4], [23.0
(CH3), 23.0 (CH3), CAla-3], [22.2 (CH3), 22.2 (CH3), 18.2 (CH3),
18.0 (CH3), CLeu-5]. HRMS (ESI-TOF) m/z: [M + H]+ Calcd for
+
+ H]+ Calcd for C12H18N3O4 , 268.129 18; found, 268.128 84.
Methyl 2-Furoyl-L-leucyl-L-alaninamide (7a). Following the
general protocol for the synthesis of primary amide, gaseous ammonia
was bubbled into a solution of 5a (0.5552 g, 1.788 mmol) in MeOH p.a.
(20 mL), and the system was left stirring for 48 h at rt. After the typical
workup, 0.5279 g of 7a was obtained as a white solid. Yield: quantitative.
m.p.: 85−90 °C. Rf: 0.27 in EtOAc. [α]2D3: −36.6 (c1.015, CHCl3). 1H
NMR (CDCl3, 400 MHz) δ ppm (rotamers present 70:30): [7.84 (d, J
= 7.8 Hz, minor), 7.76 (d, J = 7.5 Hz, major), 1H, CONH], 7.47−7.20
[7.44 (dd, J = 1.7, 0.7 Hz), 2H, H-5 + CONH], [7.13 (dd, J = 3.5, 0.6
Hz, major), 7.10 (dd, J = 3.6, 0.6 Hz, minor), 1H, H-3], [7.02 (br s,
minor), 6.94 (br s, major), 1H, CONHH], 6.63−6.19 [6.42 (br s,
minor), 6.38 (br s, major), 2H, H-4 + CONHH], 4.81−4.63 (m, 1H,
HLeu-2), [4.53 (p, J = 7.0 Hz), 4.52 (p, J = 7.1 Hz), 1H, HAla-2], 1.76−
1.63 (m, 3H, HLeu-3 + HLeu-4), [1.39 (d, J = 7.1 Hz, minor), 1.33 (d, J =
+
C15H23N2O5 , 311.160 15; found, 311.158 26.
Methyl 2-Furoyl-L-valyl-L-alaninate (5b). Following the general
protocol for peptide coupling, DIEA (1.24 mL, 7.12 mmol), TBTU
H
https://dx.doi.org/10.1021/acschemneuro.0c00687
ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

ACS Chemical Neuroscience
pubs.acs.org/chemneuro
Research Article
7.1 Hz, major), 3H, HAla-3], 0.98−0.87 (m, 6H, HLeu-5). 13C NMR and
DEPT (CDCl3, 101 MHz) δ ppm (rotamers present): [176.3 (C),
176.2 (C), 173.4 (C), 173.2 (C), 159.8 (C), 159.5 (C), 2 × CONH +
CONH2], [148.2 (C), 148.0 (C), C-2], [145.5 (CH), 145.4 (CH), C-
5], [115.9 (CH), 115.8 (CH), C-3], 113.1 (CH, C-4), [53.0 (CH), 52.6
(CH), CLeu-2], [49.7 (CH), 49.6 (CH), CAla-2], [42.3 (CH2), 42.1
(CH2), CLeu-3], [25.7 (CH), 25.7 (CH), CLeu-4], [23.8 (CH3), 23.7
(CH3), 22.9 (CH3), 22.8 (CH3), CLeu-5], [18.8 (CH3), 18.7 (CH3),
when [3H]-NPA was coincubated with 10 μM (+)-butaclamol.
ANOVA analysis was carried out to evaluate significant
differences using SPSS software (V15.0). Statistical significance
was set at P < 0.05.
Biological Assays. Animals. Five female Wistar-Kyoto
(WKY) rats (Iffa-Credo, L’Arbresele, Lyon, France), purchased
from Criffa (Barcelona, Spain), were used throughout this study.
They were housed (groups of five) in Makrolon cages (Panlab,
Barcelona, Spain) on poplar shaving bedding (B&K Universal;
G. Jordi, Barcelona, Spain) in a standard experimental animal
Bioclean room, illuminated from 8:00 AM to 8:00 PM (12 h/12
h/light/dark cycle) and maintained at a temperature of 22 to 24
°C. Animals had free access to food pellets (B&K Universal) and
drinking fluid (tap water). All animal experiments performed
were conducted in compliance with institutional guidelines and
European regulations on the protection of animals (Directive
2010/63/UE), the Spanish Real Decreto 53/2013 (1 February),
and the Guide for the Care and Use of Laboratory Animals as
adopted and promulgated by the US.
Culture of Rat Cortical Neurons and Glial Cells. Pregnant
rats (19−20 days) were killed by CO2 inhalation. Embryos were
immediately extracted from the womb by caesarean section, and
their brains were carefully dissected out. Meninges were
removed, and a portion of motor cortex was isolated after the
dissection of the brain. Fragments obtained from several
embryos were subjected to mechanic disintegration. Neurobasal
medium supplemented with 2% B-27 was used to seed the cells
in 48-well plates at a density of 100 000 cells/mL. Neuronal
cultures were allowed to grow for 8−10 days keeping in an
incubator (Forma Direct Heat CO2, Thermo Electron
Corporation, Madrid, Spain) under saturated humidity at a
partial pressure of 5% CO2 in air at 37 °C. Once a dense
neuronal network could be observed, motor cortex cultures were
treated with the compounds 4−7(a,b) at 100 μM, H2O2 (100
μM), or DMSO (1%) in order to evaluate their cytotoxic effects.
After incubation for 24 h under the conditions described aboved,
cell viability was evaluated using MTT as follows: 0.5 mg/mL
MTT was added to each well, and incubation was performed at
37 °C for 2 h. Medium was removed, and formazan salt formed
was dissolved in DMSO. The colorimetric determination was
performed at 540 nm. Percent viability (% viability) was
+
CAla-3]. HRMS (ESI-TOF) m/z: [M + H]+ Calcd for C14H22N3O4 ,
296.160 48; found, 296.160 33.
Methyl 2-Furoyl-L-valyl-L-alaninamide (7b). Following the general
protocol for the synthesis of primary amide, gaseous ammonia was
bubbled into a solution of 5b (0.4012 g, 1.353 mmol) in MeOH p.a. (20
mL), and the system was left stirring for 48 h at rt. After the typical
workup, 0.3580 g of 7b was obtained as a white solid. Yield: 94%. m.p.:
202−204 °C. Rf: 0.14 in Et2O. [α]2D1: +4.3 (c1.04, DMSO). 1H NMR
(CD3OD, 400 MHz) δ ppm (rotamers present 60:40): 7.75−7.58 (m,
1H, H-5), 7.25−7.07 (m, 1H, H-3), 6.66−6.50 [6.59 (dd, J = 3.5, 1.8
Hz), 6.58 (dd, J = 3.5, 1.8 Hz), 1H, H-4], 4.46−4.16 (m, 2H, HVal-2 +
HAla-2), 2.25−2.10 (m, 1H, HVal-3), 1.42−1.33 [m, 3H, HAla-3], 1.08−
0.93 (m, 6H, HVal-4). 13C NMR and DEPT, (CD3OD, 101 MHz) δ
ppm (rotamers present): [177.6 (C), 177.3 (C), 173.6 (C), 173.1 (C),
160.9 (C), 160.6 (C), 2 × CONH + CONH2], [148.5 (C), 148.5 (C),
C-2], 146.7 (CH, C-5), [115.9 (CH), 115.9 (CH), C-3], [113.1 CH),
113.0 (CH), C-4], [60.8 (CH), 59.9 (CH), CVal-2], [50.1 (CH), 50.0
(CH), CAla-2], [32.2 (CH), 31.6 (CH), CVal-3], [19.8 (CH3), 19.6
(CH3), 19.3 (CH3), 18.9 (CH3), 18.3 (CH3), 18.0 (CH3), CVal-4 +
+
CAla-3]. HRMS (ESI-TOF) m/z: [M + H]+ Calcd for C13H20N3O4 ,
282.144 83; found, 282.144 37.
PHARMACOLOGY
■
Dopamine D2 Receptor Binding Assay. Chinese hamster
ovary (CHO) cells expressing short isoform of human D2S
receptors were grown in 150 mm Petri dishes in Dulbeco’s
Modified Eagle Medium (DMEM) supplemented with 10% FBS
and 2 mM L-Glutamine. When cells were confluent, medium was
removed, and cells were washed twice with buffer A (5 mM
Tris−HCl pH = 7.4, 2 mM EDTA). Cells were scrapped and
homogenized twice in a Polytron. Cell suspension was
centrifuged (300 g, 10 min, 4 °C). Pellet was discarded, and
supernatant was centrifuged (48 400 g; 4 °C; 60 min). Pellet was
resuspended in buffer B (50 mM Tris−HCl; pH = 7.4), and
protein quantity was measured by using Bradford method. The
binding of [3H]-NPA to the membrane preparation was assayed
in duplicate in 96-well plates. Membranes (30 μg/well)
expressing human D2R were incubated with 0.25 nM [3H]-
NPA and test compounds for 60 min at 25 °C in a 96-well
polypropylene microplate with incubation buffer (50 mM Tris−
HCl, pH = 7.4; 120 mM NaCl, 5 mM KCl, 4 mM MgCl2, 1 mM
EDTA) up to a total volume of 250 μL. Nonspecific binding was
defined in the presence of 1 μM (+)-butaclamol. After
incubation time, 200 μL was transferred to a multiscreen FC
microplate (Millipore) pretreated with 0.5% polyethilenimine,
and samples were filtered and washed four times with 250 μL of
wash buffer (50 mM Tris−HCl, pH = 7.4; 0.9% NaCl). Filters
were dried, and 35 μL of scintillation cocktail (Universol) was
added to each well; radioactivity was detected in a microplate
beta scintillation counter (Microbeta Trilux). Data are expressed
as the increase of specific binding following the formula:
calculated as mean
SEM of five determinations from five
independent cultures.
ASSOCIATED CONTENT
■
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acschemneuro.0c00687.
Copies of the 1D (1H, 13C{1H}, DEPT-135) and 2D
(COSY, HSQC) NMR spectra for all the compounds
reported and details for the development of the
ALLOPTML model (PDF)
Full lists of the values of MA and MMA, along with the
data set used and the results of the MMA model for each
case, including compound code, molecular descriptors,
and the assay conditions (XLSX)
(X − NSB) × 100
% increase =
− 100
BT − NSB
AUTHOR INFORMATION
■
Corresponding Authors
where X is the radioactivity detected in the test well, BT is the
radioactivity detected when [3H]-NPA was incubated in the
absence of any compound, and NSB is the radioactivity detected
Ivo E. Sampaio-Dias − LAQV/REQUIMTE, Dept. of
Chemistry and Biochemistry, Faculty of Sciences, University of
I
https://dx.doi.org/10.1021/acschemneuro.0c00687
ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

ACS Chemical Neuroscience
pubs.acs.org/chemneuro
Research Article
Porto, 4169-007 Porto, Portugal; orcid.org/0000-0002-
9071-6197; Email: idias@fc.up.pt
Xerardo García-Mera − Dept. of Organic Chemistry, Faculty of
Pharmacy, University of Santiago de Compostela, 15782
Santiago de Compostela, Spain; Email: xerardo.garcia@
usc.es
grants PTDC/BIA-MIB/29059/2017 and PEst-OE/QUI/
UI0674/2013. This work was also supported by the
Collaborative Project of Genomic Data Integration (CICL-
OGEN). The authors thank the USEF Drug Screening Platform
at University of Santiago de Compostela for the support to carry
out this work. The support of Ikerbasque, Basque Foundation
for Science and the research grants from Ministry of Economy
and Competitiveness, MINECO, Spain (FEDER CTQ2016−
74881-P), and Basque government (IT1045−16) are also
acknowledged. The financial support (ED431G 2019/02) from
́
Humberto Gonzalez-Díaz − Dept. of Organic Chemistry II,
University of Basque Country (UPV-EHU), 48940 Leioa,
Spain; Basque Center for Biophysics (CSIC UPV/EHU),
University of Basque Country (UPV-EHU), 48940 Leioa,
Spain; IKERBASQUE, Basque Foundation for Science, 48011
Bilbao, Spain; orcid.org/0000-0002-9392-2797;
Email: humberto.gonzalezdiaz@ehu.es
́
the Xunta de Galicia (Centro singular de investigacion de
Galicia accreditation 2019−2022) and the European Union
(European Regional Development FundERDF) is gratefully
acknowledged. I.E.S.-D. thanks FCT for the doctoral grant
SFRH/BD/93632/2013. X.G.-M. thanks Xunta de Galicia for
financial support (GPC2014/003).
Authors
́
Jose E. Rodríguez-Borges − LAQV/REQUIMTE, Dept. of
Chemistry and Biochemistry, Faculty of Sciences, University of
Porto, 4169-007 Porto, Portugal
Notes
The authors declare no competing financial interest.
́ ́
Víctor Yanez-Perez − Dept. of Organic Chemistry II, University
̃
of Basque Country (UPV-EHU), 48940 Leioa, Spain
Sonia Arrasate − Dept. of Pharmacology, Faculty of Medicine
and Nursing, University of Basque Country (UPV-EHU),
48940 Leioa, Spain; orcid.org/0000-0003-2601-5959
Javier Llorente − Dept. of Pharmacology, Faculty of Medicine
and Nursing, University of Basque Country (UPV-EHU),
48940 Leioa, Spain; Dept. of Pharmacology, University of
Santiago de Compostela, 15782 Santiago de Compostela,
Spain
ACKNOWLEDGMENTS
■
Thanks are due to the Portuguese NMR network (RNRMN) for
supporting the Laboratory for Structural Elucidation (LAE) of
the Materials Centre of the University of Porto (CEMUP).
I.E.S.-D. would like to thank Tiago Carrola for the valuable
discussions about this work.
ABBREVIATIONS USED
13C NMR, carbon-13 nuclear magnetic resonance; H NMR,
■
́
Jose M. Brea − Innopharma Screening Platform, Biofarma
1
Research group, Centre of Research in Molecular Medicine and
Chronic Diseases CIMUS, University of Santiago de
Compostela, 15782 Santiago de Compostela, Spain
Harbil Bediaga − Dept. of Organic Chemistry II, University of
Basque Country (UPV-EHU), 48940 Leioa, Spain; Dept. of
Physical Chemistry, University of Basque Country (UPV-
EHU), 48940 Leioa, Spain
proton magnetic nuclear resonance; 2-Fu, 2-furoic acid; Ac,
accuracy; BD, big data; CEMUP, Centro de Materiais da
Universidade do Porto; CHO, Chinese hamster ovary cells;
D1R, dopamine D1 receptor; DIEA, N-ethyl-N,N-diisopropyl-
amine; DMSO, dimethyl sulfoxide; DMSO-d6, deuterated
dimethyl sulfoxide; DOS, diversity-oriented synthesis; ESI,
electrospray ionization; GPCRs, G-protein-coupled receptors;
HEK293, Human Embryonic Kidney 293 cells; HRMS, high-
resolution mass spectrometry; LDA, Linear Discriminant
Analysis; MA, moving average operators; mGluR, metabotropic
glutamate receptor; MIF-1, Melanostatin; ML, machine
learning; MMA, multiple moving average operators; NPA, N-
propylapomorphine; PAM, positive allosteric modulator; PSA,
polar surface area; PT, perturbation theory; PTML, perturba-
tion theory and machine learning model; ROS, reactive oxygen
species; Sn, sensitivity; Sp, specificity; TBTU, O-(benzotriazol-
1-yl)-N,N,N′,N′-tetramethyluronium tetrafluoroborate; TLC,
thin-layer chromatography; VCCLAB, Virtual Computational
Chemistry Laboratory; WKY, Wistar-Kyoto
Dolores Vina − Dept. of Pharmacology, Faculty of Pharmacy
̃
and Centre of Research in Molecular Medicine and Chronic
Diseases CIMUS, University of Santiago de Compostela,
15782 Santiago de Compostela, Spain
María Isabel Loza − Innopharma Screening Platform, Biofarma
Research group, Centre of Research in Molecular Medicine and
Chronic Diseases CIMUS, University of Santiago de
Compostela, 15782 Santiago de Compostela, Spain
̃
Olga Caamano − Dept. of Organic Chemistry, Faculty of
Pharmacy, University of Santiago de Compostela, 15782
Santiago de Compostela, Spain
Complete contact information is available at:
https://pubs.acs.org/10.1021/acschemneuro.0c00687
REFERENCES
■
Author Contributions
(1) Mishra, A., Singh, S., and Shukla, S. (2018) Physiological and
Functional Basis of Dopamine Receptors and Their Role in
Neurogenesis: Possible Implication for Parkinson’s disease. J. Exp.
Neurosci. 12, 1−8.
(2) Beaulieu, J. M., and Gainetdinov, R. R. (2011) The physiology,
signaling, and pharmacology of dopamine receptors. Pharmacol. Rev. 63
(1), 182−217.
(3) Platania, C. B. M., Salomone, S., Leggio, G. M., Drago, F., and
Bucolo, C. (2012) Homology modeling of dopamine D2 and D3
receptors: molecular dynamics refinement and docking evaluation.
PLoS One 7 (9), e44316.
(4) Lindgren, N., Usiello, A., Goiny, M., Haycock, J., Erbs, E.,
Greengard, P., Hokfelt, T., Borrelli, E., and Fisone, G. (2003) Distinct
roles of dopamine D2L and D2S receptor isoforms in the regulation of
I.E.S.-D. and H.G.-D. conceived and designed the experiments.
I.E.S.-D. completed the experimental work. H.G.-D., S.A., J.L.,
V.Y.-P., and H.B. completed the development of the model.
I.E.S.-D., X.G.-M., J.E.R.-B., H.G.-D., and O.C. analyzed the
data. J.M.B. and M.I.L. completed the pharmacological assays.
D.V. completed the biological assays. I.E.S.-D., X.G.-M., and
H.G.-D. wrote the paper. All authors have given approval to the
final version of the manuscript.
Funding
̂
This research was funded by Fundaca̧ o para a Ciencia e
Tecnologia (FCT, Portugal), through grants UIDB/50006/
2020 (to LAQV-REQUIMTE Research Unit) and for project
̃
J
https://dx.doi.org/10.1021/acschemneuro.0c00687
ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

ACS Chemical Neuroscience
pubs.acs.org/chemneuro
Research Article
protein phosphorylation at presynaptic and postsynaptic sites. Proc.
Natl. Acad. Sci. U. S. A. 100 (7), 4305−4359.
glycinamide and its peptidomimetic, PAOPA, attenuate haloperidol
induced vacuous chewing movements in rat: A model of human tardive
dyskinesia. Peptides 24 (2), 313−319.
(5) Nussinov, R., and Tsai, C. J. (2013) Allostery in disease and in
drug discovery. Cell 153 (2), 293−305.
(24) Mycroft, F. J., Bhargava, H. N., and Wei, E. T. (1987)
Pharmacological activities of the MIF-1 analogues Pro-Leu-Gly, Tyr-
Pro-Leu-Gly and pareptide. Peptides 8 (6), 1051−1055.
(25) Chiu, P., Rajakumar, G., Chiu, S., Johnson, R. L., and Mishra, R.
K. (1985) Mesolimbic and striatal dopamine receptor supersensitivity:
Prophylactic and reversal effects of L-prolyl-L-leucyl-glycinamide
(PLG). Peptides 6 (2), 179−183.
(6) Khan, R. S., Yu, C., Kastin, A. J., He, Y., Ehrensing, R. H., Hsuchou,
H., Stone, K. P., and Pan, W. (2010) Brain Activation by Peptide Pro-
Leu-Gly-NH2 (MIF-1). Int. J. Pept. 2010, 537639.
(7) Allegretti, M., Cesta, M. C., and Locati, M. (2016) Allosteric
Modulation of Chemoattractant Receptors. Front. Immunol. 7, 170.
(8) Verma, V., Mann, A., Costain, W., Pontoriero, G., Castellano, J. M.,
Skoblenick, K., Gupta, S. K., Pristupa, Z., Niznik, H. B., Johnson, R. L.,
Nair, V. D., and Mishra, R. K. (2005) Modulation of agonist binding to
human dopamine receptor subtypes by L-prolyl-L-leucyl-glycinamide
and a peptidomimetic analog. J. Pharmacol. Exp. Ther. 315 (3), 1228−
1236.
(26) Reed, L. L., and Johnson, P. L. (1973) Solid state conformation of
the C-terminal tripeptide of oxytocin, L-Pro-L-Leu-Gly-NH2.0.5H2O.
J. Am. Chem. Soc. 95 (22), 7523−7524.
́
(27) Aizpurua, J. M., Palomo, C., Balentova, E., Jimenez, A., Andreieff,
E., Sagartzazu-Aizpurua, M., Miranda, J. I., and Linden, A. (2013)
Chirality-Driven Folding of Short β-Lactam Pseudopeptides. J. Org.
Chem. 78 (2), 224−237.
(9) Vartak, A. P., Skoblenick, K., Thomas, N., Mishra, R. K., and
Johnson, R. L. (2007) Allosteric modulation of the dopamine receptor
by conformationally constrained type VI beta-turn peptidomimetics of
Pro-Leu-Gly-NH2. J. Med. Chem. 50 (26), 6725−6729.
(28) Johnson, R. L., Rajakumar, G., Yu, K. L., and Mishra, R. K. (1986)
Synthesis of Pro-Leu-Gly-NH2 analogues modified at the prolyl residue
and evaluation of their effects on the receptor binding activity of the
central dopamine receptor agonist, ADTN. J. Med. Chem. 29 (10),
2104−2107.
́
(10) Ferreira da Costa, J., Caamano, O., Fernandez, F., García-Mera,
̃
X., Sampaio-Dias, I. E., Brea, J. M., and Cadavid, M. I. (2013) Synthesis
and allosteric modulation of the dopamine receptor by peptide analogs
of l-prolyl-l-leucyl-glycinamide (PLG) modified in the l-proline or l-
proline and l-leucine scaffolds. Eur. J. Med. Chem. 69 (SupplementC),
146−158.
(29) Sampaio-Dias, I. E., Sousa, C. A. D., Silva-Reis, S. C., Ribeiro, S.,
García-Mera, X., and Rodríguez-Borges, J. E. (2017) Highly efficient
one-pot assembly of peptides by double chemoselective coupling. Org.
Biomol. Chem. 15 (36), 7533−7542.
(30) Liang, X., Haynes, B. S., and Montoya, A. (2018) Acid-Catalyzed
Ring Opening of Furan in Aqueous Solution. Energy Fuels 32 (4),
4139−4148.
(31) Open-source cheminformatics, http://zinc15.docking.org/
patterns/home (accessed 2020-07-10).
(32) Christopoulos, A., and Kenakin, T. (2002) G protein-coupled
receptor allosterism and complexing. Pharmacol. Rev. 54 (2), 323−374.
(33) Bhagwanth, S., Mishra, R. K., and Johnson, R. L. (2013)
Development of peptidomimetic ligands of Pro-Leu-Gly-NH(2) as
allosteric modulators of the dopamine D(2) receptor. Beilstein J. Org.
Chem. 9, 204−214.
(11) Sampaio-Dias, I. E., Silva-Reis, S. C., García-Mera, X., Brea, J.,
Loza, M. I., Alves, C. S., Algarra, M., and Rodríguez-Borges, J. E. (2019)
Synthesis, Pharmacological, and Biological Evaluation of MIF-1
Picolinoyl Peptidomimetics as Positive Allosteric Modulators of D2R.
ACS Chem. Neurosci. 10 (8), 3690−3702.
(12) Sampaio-Dias, I. E., Sousa, C. A. D., Garcia-Mera, X., Ferreira da
Costa, J., Caamano, O., and Rodriguez-Borges, J. E. (2016) Org. Biomol.
Chem. 14, 11065−11069.
(13) Celis, M. E., Taleisnik, S., and Walter, R. (1971) Regulation of
formation and proposed structure of the factor inhibiting the release of
melanocyte-stimulating hormone. Proc. Natl. Acad. Sci. U. S. A. 68 (7),
1428−1433.
(14) Srivastava, L. K., Bajwa, S. B., Johnson, R. L., and Mishra, R. K.
(1988) Interaction of l-Prolyl-l-Leucyl Glycinamide with Dopamine D2
Receptor: Evidence for Modulation of Agonist Affinity States in Bovine
Striatal Membranes. J. Neurochem. 50 (3), 960−968.
(15) Chiu, S., Paulose, C. S., and Mishra, R. K. (1981) Effect of L-
prolyl-L-leucyl-glycinamide (PLG) on neuroleptic-induced catalepsy
and dopamine/neuroleptic receptor bindings. Peptides 2 (1), 105−111.
(16) Saitton, S., Del Tredici, A. L., Mohell, N., Vollinga, R. C.,
(34) Alonso, N., Caamano, O., Romero-Duran, F. J., Luan, F.,
Cordeiro, M. N. D. S., Yanez, M., Gonzalez-Diaz, H., and Garcia-Mera,
X. (2013) Model for High-Throughput Screening of Multitarget Drugs
in Chemical Neurosciences: Synthesis, Assay, and Theoretic Study of
Rasagiline Carbamates. ACS Chem. Neurosci. 4 (10), 1393−1403.
(35) Luan, F., Cordeiro, M. N. D. S., Alonso, N., García-Mera, X.,
́
Caamano, O., Romero-Duran, F. J., Yanez, M., and Gonzalez-Díaz, H.
̃
̃
(2013) TOPS-MODE model of multiplexing neuroprotective effects of
drugs and experimental-theoretic study of new 1,3-rasagiline derivatives
potentially useful in neurodegenerative diseases. Bioorg. Med. Chem. 21
(7), 1870−1879.
̈
Bostrom, D., Kihlberg, J., and Luthman, K. (2004) Design, synthesis
and evaluation of a PLG tripeptidomimetic based on a pyridine scaffold.
J. Med. Chem. 47 (26), 6595−6602.
(17) Reed, G. W., Olson, G. A., and Olson, R. D. (1994) The Tyr-
MIF-1 family of peptides. Neurosci. Biobehav. Rev. 18 (4), 519−525.
(18) Mishra, R. K., Makman, M. H., Costain, W. J., Nair, V. D., and
Johnson, R. L. (1999) Modulation of agonist stimulated adenylyl
cyclase and GTPase activity by L-pro-L-leu-glycinamide and its
peptidomimetic analogue in rat striatal membranes. Neurosci. Lett.
269 (1), 21−24.
(36) Ricart, K. C., and Fiszman, M. L. (2001) Hydrogen Peroxide-
Induced Neurotoxicity in Cultured Cortical Cells Grown in Serum-Free
and Serum-Containing Media. Neurochem. Res. 26 (7), 801−808.
(37) Schneider, G. (2011) Computational medicinal chemistry.
Future Med. Chem. 3 (4), 393−394.
(38) Csermely, P., Nussinov, R., and Szilagyi, A. (2013) From
allosteric drugs to allo-network drugs: state of the art and trends of
design, synthesis and computational methods. Curr. Top. Med. Chem. 13
(1), 2−4.
(39) Szilagyi, A., Nussinov, R., and Csermely, P. (2013) Allo-network
drugs: extension of the allosteric drug concept to protein- protein
interaction and signaling networks. Curr. Top. Med. Chem. 13 (1), 64−
77.
(40) Csermely, P., Korcsmaros, T., Kiss, H. J., London, G., and
Nussinov, R. (2013) Structure and dynamics of molecular networks: a
novel paradigm of drug discovery: a comprehensive review. Pharmacol.
Ther. 138 (3), 333−408.
(41) Taylor, P. J. (1990) Comprehensive Medicinal Chemistry, 1990 ed.,
Vol. 4, p 241−294, Pergamon Press, Oxford.
(42) Wang, J., Yun, D., Yao, J., Fu, W., Huang, F., Chen, L., Wei, T., Yu,
C., Xu, H., Zhou, X., Huang, Y., Wu, J., Qiu, P., and Li, W. (2018)
(19) Birtwistle, J., and Baldwin, D. (1998) Role of dopamine in
schizophrenia and Parkinson’s disease. Br. J. Nurs. 7 (14), 832−834.
836, 838−841.
(20) Smith, J. R., and Morgan, M. (1982) The effects of prolyl-leucyl-
glycine amide on drug-induced rotation in lesioned rats. Gen.
Pharmacol. 13 (3), 203−207.
(21) Marcotte, E. R., Chugh, A., Mishra, R. K., and Johnson, R. L.
(1998) Protection Against MPTP Treatment by an Analog of Pro-Leu-
Gly-NH2 (PLG, MIF-1). Peptides 19 (2), 403−406.
(22) Barbeau, A., Roy, M., and Kastin, A. J. (1976) Double-blind
evaluation of oral L-prolyl-Lleucyl-glycine amide in Parkinson’s disease.
Can. Med. Assoc. J. 114 (2), 120−122.
(23) Sharma, S., Paladino, P., Gabriele, J., Saeedi, H., Henry, P.,
Chang, M., Mishra, R. K., and Johnson, R. L. (2003) Pro-Leu-
K
https://dx.doi.org/10.1021/acschemneuro.0c00687
ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

ACS Chemical Neuroscience
pubs.acs.org/chemneuro
Research Article
Design, synthesis and QSAR study of novel isatin analogues inspired
Michael acceptor as potential anticancer compounds. Eur. J. Med. Chem.
144, 493−503.
QSAR study of anticancer ellipticine analogues. J. Chem. Inf. Comput.
Sci. 38 (2), 189−199.
(58) Han, L., Cui, J., Lin, H., Ji, Z., Cao, Z., Li, Y., and Chen, Y. (2006)
Recent progresses in the application of machine learning approach for
predicting protein functional class independent of sequence similarity.
Proteomics 6 (14), 4023−4037.
(59) Chou, K. C., and Shen, H. B. (2010) Plant-mPLoc: a top-down
strategy to augment the power for predicting plant protein subcellular
localization. PLoS One 5 (6), e11335.
(60) Chou, K. C., and Shen, H. B. (2008) Cell-PLoc: a package of Web
servers for predicting subcellular localization of proteins in various
organisms. Nat. Protoc. 3 (2), 153−162.
(61) Cai, Y. D., and Chou, K. C. (2005) Predicting enzyme subclass by
functional domain composition and pseudo amino acid composition. J.
Proteome Res. 4 (3), 967−971.
̇
́
́
(43) Pogorzelska, A., Sławinski, J., Zołnowska, B., Szafranski, K.,
́
Kawiak, A., Chojnacki, J., Ulenberg, S., Zielinska, J., and Bączek, T.
(2017) Novel 2-(2-alkylthiobenzenesulfonyl)-3-(phenylprop-2-
ynylideneamino)guanidine derivatives as potent anticancer agents -
Synthesis, molecular structure, QSAR studies and metabolic stability.
Eur. J. Med. Chem. 138, 357−370.
̇
́
́
(44) Sławinski, J., Szafranski, K., Pogorzelska, A., Zołnowska, B.,
Kawiak, A., Macur, K., Belka, M., and Bączek, T. (2017) Novel 2-
benzylthio-5-(1,3,4-oxadiazol-2-yl)benzenesulfonamides with anti-
cancer activity: Synthesis, QSAR study, and metabolic stability. Eur. J.
Med. Chem. 132, 236−248.
(45) Murahari, M., Kharkar, P. S., Lonikar, N., and Mayur, Y. C.
(2017) Design, synthesis, biological evaluation, molecular docking and
QSAR studies of 2,4-dimethylacridones as anticancer agents. Eur. J.
Med. Chem. 130, 154−170.
(46) Ruddarraju, R. R., Murugulla, A. C., Kotla, R., Chandra Babu
Tirumalasetty, M., Wudayagiri, R., Donthabakthuni, S., Maroju, R.,
Baburao, K., and Parasa, L. S. (2016) Design, synthesis, anticancer,
antimicrobial activities and molecular docking studies of theophylline
containing acetylenes and theophylline containing 1,2,3-triazoles with
variant nucleoside derivatives. Eur. J. Med. Chem. 123, 379−396.
(47) Singh, H., Kumar, R., Singh, S., Chaudhary, K., Gautam, A., and
Raghava, G. P. (2016) Prediction of anticancer molecules using hybrid
model developed on molecules screened against NCI-60 cancer cell
lines. BMC Cancer 16, 77.
(48) Pingaew, R., Prachayasittikul, V., Worachartcheewan, A.,
Nantasenamat, C., Prachayasittikul, S., Ruchirawat, S., and
Prachayasittikul, V. (2015) Novel 1,4-naphthoquinone-based sulfona-
mides: Synthesis, QSAR, anticancer and antimalarial studies. Eur. J.
Med. Chem. 103, 446−459.
(49) Anwer, Z., and Gupta, S. P. (2013) A QSAR study on some series
of anticancer tyrosine kinase inhibitors. Med. Chem. (Sharjah, United
Arab Emirates) 9 (2), 203−212.
(50) Toropov, A. A., Toropova, A. P., Benfenati, E., Gini, G.,
Leszczynska, D., and Leszczynski, J. (2011) SMILES-based QSAR
approaches for carcinogenicity and anticancer activity: comparison of
correlation weights for identical SMILES attributes. Anti-Cancer Agents
Med. Chem. 11 (10), 974−982.
(51) Huang, X. Y., Shan, Z. J., Zhai, H. L., Li, L. N., and Zhang, X. Y.
(2011) Molecular design of anticancer drug leads based on three-
dimensional quantitative structure-activity relationship. J. Chem. Inf.
Model. 51 (8), 1999−2006.
(52) Benfenati, E., Toropov, A. A., Toropova, A. P., Manganaro, A.,
and Gonella Diaza, R. (2011) coral Software: QSAR for Anticancer
Agents. Chem. Biol. Drug Des. 77 (6), 471−476.
(62) Shen, H. B., and Chou, K. C. (2009) QuatIdent: a web server for
identifying protein quaternary structural attribute by fusing functional
domain and sequential evolution information. J. Proteome Res. 8 (3),
1577−1584.
(63) Chou, K. C., and Shen, H. B. (2006) Large-scale predictions of
gram-negative bacterial protein subcellular locations. J. Proteome Res. 5
(12), 3420−3428.
(64) Rodriguez-Soca, Y., Munteanu, C. R., Dorado, J., Pazos, A.,
́
Prado-Prado, F. J., and Gonzalez-Díaz, H. (2010) Trypano-PPI: a web
server for prediction of unique targets in trypanosome proteome by
using electrostatic parameters of protein-protein interactions. J.
Proteome Res. 9 (2), 1182−1190.
́
(65) Munteanu, C. R., Vazquez, J. M., Dorado, J., Sierra, A. P.,
́
́
́
Sanchez-Gonzalez, A., Prado-Prado, F. J., and Gonzalez-Díaz, H.
(2009) Complex network spectral moments for ATCUN motif DNA
cleavage: first predictive study on proteins of human pathogen
parasites. J. Proteome Res. 8 (11), 5219−5228.
́
(66) Gonzalez-Díaz, H., Saiz-Urra, L., Molina, R., Santana, L., and
Uriarte, E. (2007) A model for the recognition of protein kinases based
on the entropy of 3D van der Waals interactions. J. Proteome Res. 6 (2),
904−908.
́
(67) Gonzalez-Díaz, H., Prado-Prado, F., García-Mera, X., Alonso, N.,
́
Abeijon, P., Caaman
̃
o, O., Yan
̃
ez, M., Munteanu, C. R., Pazos, A., Dea-
́
Ayuela, M. A., Gomez-Mun
̃
oz, M. T., Garijo, M. M., Sansano, J., and
Ubeira, F. M. (2011) MIND-BEST: Web server for drugs and target
discovery; design, synthesis, and assay of MAO-B inhibitors and
theoretical-experimental study of G3PDH protein from Trichomonas
gallinae. J. Proteome Res. 10 (4), 1698−1718.
(68) Concu, R., Dea-Ayuela, M. A., Perez-Montoto, L. G., Bolas-
́
Fernandez, F., Prado-Prado, F. J., Podda, G., Uriarte, E., Ubeira, F. M.,
́
and Gonzalez-Díaz, H. (2009) Prediction of enzyme classes from 3D
structure: a general model and examples of experimental-theoretic
scoring of peptide mass fingerprints of Leishmania proteins. J. Proteome
Res. 8 (9), 4372−4382.
́
́
(53) Gonzalez-Díaz, H., Bonet, I., Teran, C., De Clercq, E., Bello, R.,
García, M. M., Santana, L., and Uriarte, E. (2007) ANN-QSAR model
for selection of anticancer leads from structurally heterogeneous series
of compounds. Eur. J. Med. Chem. 42 (5), 580−585.
(69) Aguero-Chapin, G., Varona-Santos, J., de la Riva, G. A., Antunes,
̈
́
́
A., Gonzalez-Villa, T., Uriarte, E., and Gonzalez-Díaz, H. (2009)
Alignment-free prediction of polygalacturonases with pseudofolding
topological indices: experimental isolation from Coffea arabica and
prediction of a new sequence. J. Proteome Res. 8 (4), 2122−2128.
(70) Martínez-Arzarte, S. G., Tenorio-Borroto, E., Pliego, A. B., Díaz-
́
(54) Gonzalez-Díaz, H., Vina, D., Santana, L., de Clercq, E., and
̃
Uriarte, E. (2006) Stochastic entropy QSAR for the in silico discovery
of anticancer compounds: prediction, synthesis, and in vitro assay of
new purine carbanucleosides. Bioorg. Med. Chem. 14 (4), 1095−1107.
́
́
́
Albiter, H. M., Vazquez-Chagoyan, J. C., and Gonzalez-Díaz, H. (2017)
PTML Model for Proteome Mining of B-cell Epitopes and Theoretic-
Experimental Study of Bm86 Protein Sequences from Colima, Mexico.
J. Proteome Res. 16 (11), 4093−4103.
́
́
(55) Gonzalez-Díaz, H., Gia, O., Uriarte, E., Hernadez, I., Ramos, R.,
Chaviano, M., Seijo, S., Castillo, J. A., Morales, L., Santana, L., Akpaloo,
D., Molina, E., Cruz, M., Torres, L. A., and Cabrera, M. A. (2003)
Markovian chemicals ″in silico″ design (MARCH-INSIDE), a
promising approach for computer-aided molecular design I: discovery
of anticancer compounds. J. Mol. Model. 9 (6), 395−407.
́
́
(71) Fernandez, M., Caballero, F., Fernandez, L., Abreu, J. I., and
Acosta, G. (2008) Classification of conformational stability of protein
mutants from 3D pseudo-folding graph representation of protein
sequences using support vector machines. Proteins: Struct., Funct., Genet.
70 (1), 167−175.
(72) Fernandez, L., Caballero, J., Abreu, J. I., and Fernandez, M.
(2007) Amino Acid Sequence Autocorrelation Vectors and Bayesian-
Regularized Genetic Neural Networks for Modeling Protein Conforma-
tional Stability: Gene V Protein Mutants. Proteins: Struct., Funct., Genet.
67, 834−852.
(56) Jung, M., Kim, H., and Kim, M. (2003) Chemical genomics
strategy for the discovery of new anticancer agents. Curr. Med. Chem. 10
(9), 757−762.
(57) Shi, L. M., Fan, Y., Myers, T. G., O’Connor, P. M., Paull, K. D.,
Friend, S. H., and Weinstein, J. N. (1998) Mining the NCI anticancer
drug discovery databases: genetic function approximation for the
́
́
L
https://dx.doi.org/10.1021/acschemneuro.0c00687
ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

ACS Chemical Neuroscience
pubs.acs.org/chemneuro
Research Article
́
́
(73) Caballero, J., Fernandez, L., Abreu, J. I., and Fernandez, M.
(2006) Amino Acid Sequence Autocorrelation vectors and ensembles
of Bayesian-Regularized Genetic Neural Networks for prediction of
conformational stability of human lysozyme mutants. J. Chem. Inf.
Model. 46 (3), 1255−1268.
target QSAR model for the in silico discovery of anti-breast cancer
agents. Eur. J. Pharm. Sci. 47 (1), 273−279.
(89) Cordeiro, M. N. D. S., and Speck-Planche, A. (2013) Computer-
aided drug design, synthesis and evaluation of new anti-cancer drugs.
Curr. Top. Med. Chem. 12 (24), 2703−2704.
(90) Speck-Planche, A., Kleandrova, V. V., Luan, F., and Cordeiro, M.
N. D. S. (2011) Multi-target drug discovery in anti-cancer therapy:
fragment-based approach toward the design of potent and versatile anti-
prostate cancer agents. Bioorg. Med. Chem. 19 (21), 6239−6244.
(91) Speck-Planche, A., and Cordeiro, M. N. D. S. (2017) Advanced
In Silico Approaches for Drug Discovery: Mining Information from
Multiple Biological and Chemical Data Through mtk- QSBER and pt-
QSPR Strategies. Curr. Med. Chem. 24, 1687−1704.
(74) Perez-Riverol, Y., Audain, E., Millan, A., Ramos, Y., Sanchez, A.,
̈
Vizcaíno, J. A., Wang, R., Muller, M., Machado, Y. J., Betancourt, L. H.,
́
́
Gonzalez, L. J., Padron, G., and Besada, V. (2012) Isoelectric point
optimization using peptide descriptors and support vector machines. J.
Proteomics 75 (7), 2269−2274.
(75) Greener, J. G., and Sternberg, M. J. E. (2015) AlloPred:
prediction of allosteric pockets on proteins using normal mode
perturbation analysis. BMC Bioinf. 16, 335.
(92) Collier, G., and Ortiz, V. (2013) Emerging computational
approaches for the study of protein allostery. Arch. Biochem. Biophys.
538 (1), 6−15.
(76) Gaulton, A., Hersey, A., Nowotka, M., Bento, A. P., Chambers, J.,
Mendez, D., Mutowo, P., Atkinson, F., Bellis, L. J., Cibrian-Uhalte, E.,
Davies, M., Dedman, N., Karlsson, A., Magarinos, M. P., Overington, J.
P., Papadatos, G., Smit, I., and Leach, A. R. (2017) The ChEMBL
database in 2017. Nucleic Acids Res. 45 (D1), D945.
(77) Gaulton, A., Bellis, L. J., Bento, A. P., Chambers, J., Davies, M.,
Hersey, A., Light, Y., McGlinchey, S., Michalovich, D., Al-Lazikani, B.,
and Overington, J. P. (2012) ChEMBL: a large-scale bioactivity
database for drug discovery. Nucleic Acids Res. 40 (D1), D1100−D1107.
(78) Gonzalez-Díaz, H., Arrasate, S., Gomez-SanJuan, A., Sotomayor,
N., Lete, E., Besada-Porto, L., and Ruso, J. M. (2013) General Theory
for Multiple Input-Output Perturbations in Complex Molecular
Systems. 1. Linear QSPR Electronegativity Models in Physical, Organic,
and Medicinal Chemistry. Curr. Top. Med. Chem. 13 (14), 1713−1741.
(79) Blazquez-Barbadillo, C., Aranzamendi, E., Coya, E., Lete, E.,
Sotomayor, N., and Gonzalez-Díaz, H. (2016) Perturbation theory
model of reactivity and enantioselectivity of palladium-catalyzed Heck-
́
̃
́
(93) Marrero-Ponce, Y., Siverio-Mota, D., Galvez-Llompart, M.,
̀
́
Recio, M. C., Giner, R. M., García-Domenech, R., Torrens, F., Aran, V.
J., Cordero-Maldonado, M. L., Esguera, C. V., de Witte, P. A., and
Crawford, A. D. (2011) Discovery of novel anti-inflammatory drug-like
compounds by aligning in silico and in vivo screening: the
nitroindazolinone chemotype. Eur. J. Med. Chem. 46 (12), 5736−5753.
(94) Tetko, I. V., Gasteiger, J., Todeschini, R., Mauri, A., Livingstone,
D., Ertl, P., Palyulin, V. A., Radchenko, E. V., Zefirov, N. S., Makarenko,
A. S., Tanchuk, V. Y., and Prokopenko, V. V. (2005) Virtual
computational chemistry laboratory − design and description. J.
Comput.-Aided Mol. Des. 19 (6), 453−463.
́
́
(95) Gottlieb, H. E., Kotlyar, V., and Nudelman, A. (1997) NMR
Chemical Shifts of Common Laboratory Solvents as Trace Impurities. J.
Org. Chem. 62 (21), 7512−7515.
́
́
Heck cascade reactions. RSC Adv. 6 (45), 38602−38610.
́ ́
̃
(80) Casanola-Martin, G. M., Le-Thi-Thu, H., Perez-Gimenez, F.,
́
́
Marrero-Ponce, Y., Merino-Sanjuan, M., Abad, C., and Gonzalez-Díaz,
H. (2016) Multi-output Model with Box-Jenkins Operators of
Quadratic Indices for Prediction of Malaria and Cancer Inhibitors
Targeting Ubiquitin-Proteasome Pathway (UPP) Proteins. Curr.
Protein Pept. Sci. 17 (3), 220−227.
́
(81) Romero-Duran, F. J., Alonso, N., Yanez, M., Caamano, O.,
̃ ̃
́
García-Mera, X., and Gonzalez-Díaz, H. (2016) Brain-inspired
cheminformatics of drug-target brain interactome, synthesis, and
assay of TVP1022 derivatives. Neuropharmacology 103, 270−278.
́
(82) Kleandrova, V. V., Luan, F., Gonzalez-Díaz, H., Ruso, J. M.,
Speck-Planche, A., and Cordeiro, M. N. D. S. (2014) Computational
Tool for Risk Assessment of Nanomaterials: Novel QSTR-Perturbation
Model for Simultaneous Prediction of Ecotoxicity and Cytotoxicity of
Uncoated and Coated Nanoparticles under Multiple Experimental
Conditions. Environ. Sci. Technol. 48 (24), 14686−14694.
́
(83) Luan, F., Kleandrova, V. V., Gonzalez-Díaz, H., Ruso, J. M., Melo,
A., Speck-Planche, A., and Cordeiro, M. N. D. S. (2014) Computer-
aided nanotoxicology: assessing cytotoxicity of nanoparticles under
diverse experimental conditions by using a novel QSTR-perturbation
approach. Nanoscale 6 (18), 10623−10630.
(84) Speck-Planche, A., and Cordeiro, M. N. D. S. (2017) Fragment-
based in silico modeling of multi-target inhibitors against breast cancer-
related proteins. Mol. Diversity 21 (3), 511−523.
(85) Speck-Planche, A., Kleandrova, V. V., Luan, F., and Cordeiro, M.
N. D. S. (2013) Unified multi-target approach for the rational in silico
design of anti-bladder cancer agents. Anti-Cancer Agents Med. Chem. 13
(5), 791−800.
(86) Speck-Planche, A., Kleandrova, V. V., Luan, F., and Cordeiro, M.
N. D. S. (2012) Chemoinformatics in multi-target drug discovery for
anti-cancer therapy: in silico design of potent and versatile anti-brain
tumor agents. Anti-Cancer Agents Med. Chem. 12 (6), 678−685.
(87) Speck-Planche, A., Kleandrova, V. V., Luan, F., and Cordeiro, M.
N. D. S. (2012) Rational drug design for anti-cancer chemotherapy:
multi-target QSAR models for the in silico discovery of anti-colorectal
cancer agents. Bioorg. Med. Chem. 20 (15), 4848−4855.
(88) Speck-Planche, A., Kleandrova, V. V., Luan, F., and Cordeiro, M.
N. D. S. (2012) Chemoinformatics in anti-cancer chemotherapy: multi-
M
https://dx.doi.org/10.1021/acschemneuro.0c00687
ACS Chem. Neurosci. XXXX, XXX, XXX−XXX