Orphan Receptor GPR88 as a potential therapeutic target for CNS Disorders – an in silico approach

Abstract

The G-protein-coupled receptors are a part of the largest and most physiologically relevant family of membrane proteins. One third of the medications, now on the market, target the GPCR receptor family, which is one of the most important therapeutic targets for many disorders. In the reported work, we have focused on orphan GPR88 receptor which is a part of the GPCR protein family and a potential target for central nervous system disorders. GPR88 is known to show the highest expression in the striatum, which is a key region in motor control and cognitive functions. Recent studies have reported that GPR88 is activated by two agonists, 2-PCCA and RTI-13951-33. In this study, we have predicted the three-dimensional protein structure for the orphan GPR88 using the homology modeling approach. We then used shape-based screening techniques based on known agonists and structure-based virtual screening methods employing docking to uncover novel GPR88 ligands. The screened GPR88-ligand complexes were further subjected to molecular dynamics simulation studies. The selected ligands could fasten the development of novel treatments for the vast list of movement disorders.

Full text

Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=tbsd20
Journal of Biomolecular Structure and Dynamics
ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/tbsd20
Orphan receptor GPR88 as a potential therapeutic
target for CNS disorders – an in silico approach
Vasavi Garisetti, Anantha Krishnan Dhanabalan & Gayathri Dasararaju
To cite this article: Vasavi Garisetti, Anantha Krishnan Dhanabalan & Gayathri
Dasararaju (2023): Orphan receptor GPR88 as a potential therapeutic target for CNS
disorders – an in silico approach, Journal of Biomolecular Structure and Dynamics, DOI:
10.1080/07391102.2023.2222820
To link to this article: https://doi.org/10.1080/07391102.2023.2222820
View supplementary material
Published online: 12 Jun 2023.
Submit your article to this journal
View related articles
View Crossmark data

Orphan receptor GPR88 as a potential therapeutic target for CNS disorders –an
in silico approach
Vasavi Garisetti, Anantha Krishnan Dhanabalan and Gayathri Dasararaju
Centre of Advanced Study in Crystallography and Biophysics, University of Madras, Chennai, Tamil Nadu, India
Communicated by Ramaswamy H. Sarma
ABSTRACT
The G-protein-coupled receptors are a part of the largest and most physiologically relevant family of
membrane proteins. One-third of the medications, now on the market, target the GPCR receptor fam-
ily, which is one of the most important therapeutic targets for many disorders. In the reported work,
we have focussed on orphan GPR88 receptor which is a part of the GPCR protein family and a poten-
tial target for central nervous system disorders. GPR88 is known to show the highest expression in the
striatum, which is a key region in motor control and cognitive functions. Recent studies have reported
that GPR88 is activated by two agonists, 2-PCCA and RTI-13951-33. In this study, we have predicted
the three-dimensional protein structure for the orphan GPR88 using the homology modeling
approach. We then used shape-based screening techniques based on known agonists and structure-
based virtual screening methods employing docking to uncover novel GPR88 ligands. The screened
GPR88-ligand complexes were further subjected to molecular dynamics simulation studies. The
selected ligands could fasten the development of novel treatments for the vast list of movement and
central nervous system disorders.
ARTICLE HISTORY
Received 16 January 2023
Accepted 2 June 2023
KEYWORDS
GPCR; GPR88; virtual
screening; molecular
docking; molecular
dynamics simulation; central
nervous system disorders
Introduction
G-protein-coupled receptors (GPCRs) are the largest and the
most diverse family of membrane proteins that regulate the
majority of cellular reactions to hormones and neurotrans-
mitters. The GPCRs also translate extracellular signals into
significant physiological effects. Numerous signalling path-
ways are activated when various ligands bind to GPCRs.
These ligands are diverse and vary from photons, ions, odor-
ants, tiny molecules including biogenic amines, amino acids,
nucleotides and peptides (Acher, 2006). Numerous disorders,
including type 2 diabetes mellitus, obesity, depression, can-
cer and Alzheimer’s disease have been linked to GPCRs.
The most well-researched pharmacological targets are the
GPCRs, and in humans there are over 800 GPCRs that control
several physiological functions. About one-fourth of all thera-
peutic medications sold worldwide are GPCR-targeting phar-
maceuticals. Therefore, the primary goal of drug discovery
research is to continue to create physiologically relevant and
reliable procedures to look for new GPCR ligands or modula-
tors. The global market for GPCRs which was estimated as
US$2.6 Billion in the year 2020, is projected to reach a
revised size of US$3.7 Billion by 2026. This highlights that
GPCRs are the largest family of targets in the market which
have been in top consideration for approved drugs (GPCRs -
Global Market Trajectory & Analytics).
The GPCRs in vertebrates are typically grouped into five
families; rhodopsin (family A), secretin (family B), glutamate
(family C), adhesion and frizzled/taste2. Of these groups, the
rhodopsin family is by far the most numerous and diversified
and its members are distinguished by conserved sequence
motifs that suggest shared structural characteristics and acti-
vation processes (Kobilka, 2007; Lameh et al., 1990). GPCRs
have a canonical structure, sometimes termed heptahelical
or seven-transmembrane receptors. They are characterized
by the presence of seven membrane-spanning a-helical seg-
ments separated by alternating intracellular and extracellular
loop regions (Latorraca et al., 2017), which are distinguished
and represented in Figure 1.
Orphan GPCR proteins
The term ‘orphan GPCR’refers to a GPCR that has been
found but whose endogenous ligand is still unknown and
whose associated pairing has not been approved by the
International Union of Basic and Clinical Pharmacology
(IUPHAR; Wacker et al., 2017). The letters ‘GPR’and a specific
number are frequently used to identify GPCR orphan recep-
tors. The orphan GPCR is deorphanized once its ligand is
identified. The class of orphan GPCRs has rapidly expanded
to more than 122 new members since 1986 as a result of
the advancement of genetic cloning and nucleic acid-based
homology screening to known receptors. These include 87
CONTACT Gayathri Dasararaju drdgayathri@gmail.com Centre of Advanced Study in Crystallography and Biophysics, University of Madras, Guindy
Campus, Chennai 600025, Tamil Nadu, India
Supplemental data for this article can be accessed online at https://doi.org/10.1080/07391102.2023.2222820.
ß2023 Informa UK Limited, trading as Taylor & Francis Group
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS
https://doi.org/10.1080/07391102.2023.2222820

class A (rhodopsin-like GPCRs family) which includes GPR88;
8 class C (glutamate family) - GPR156, GPR158, GPR179,
GPRC5A, GPRC5B, GPRC5C, GPRC5D, GPRC6; 27 adhesion
members (large extracellular region, similar to the class B
GPCR, linked to the 7 TM region by a GPCR autoproteolysis-
inducing domain; Libert et al., 1991).
GPR88 and its pathway
Mizushima et al., (2000) were the first to discover the GPR88
gene using the differential display screening for region-spe-
cific transcripts in rat brain. They also reported that the
GPR88 gene showed strict striatum-specificity in the expres-
sion patterns in the experiments conducted. GPR88 is one of
the many GPCRs that has been designated as an orphan
GPCR receptor (Ye et al., 2019). GPR88’s enrichment in the
striatum, sensitivity to antidepressant medications and famil-
ial association with schizophrenia have sparked interest in its
functions in striatal physiology and behaviours involving this
brain region.
The nucleus accumbens, the olfactory tubercle, the thal-
amus, the cortex and the inferior olive, are essential parts of
the brain that have high GPR88 expression; and GPR88 alters
glutamatergic and GABA-dependent signalling. Lovinger,
(2012) investigated the function of the GPR88 receptor in
synaptic transmission, action regulation and action learning
using gene-targeted mice lacking GPR88 and viral-based re-
expression of the protein. The GPR88-knockout mice showed
increased locomotion, decreased performance and impaired
learning when a motor skill test was conducted.
When GPCRs are activated, the receptor undergoes con-
formational changes that ultimately activate heterotrimeric G
proteins by exchanging GDP for GTP in the G
a
subunit. The
G
a
subunit and G
bc
subunits undergo rearrangement which
lead to many different signalling events through a variety of
effectors (Rasmussen et al., 2011). Following the activation of
associated G proteins, the receptor can be phosphorylated
by GPCR kinases, allowing for the recruitment of b-arrestin to
the GPCR (Lefkowitz & Shenoy, 2005). Recruitment of
b-arrestin can lead to receptor desensitization and receptor
downregulation from the membrane, as well as diverse
b-arrestin.
In GPR88 signalling pathway, seen in Figure 2, it was
studied that GPR88 associates with Ga
i/o
G proteins, which
leads to the inhibition of adenylyl cyclase and hence
reducing the cAMP production and signalling (Dzierba et al.,
2015). When studies were carried out in striatal membranes
from GPR88 knockout mice, loss of GPR88 receptor expres-
sion was observed which inhibited the function of opioid
(d/l) and muscarinic acetylcholine receptors (M1/M4) which
couple with G
i
/G
o
proteins, this also led to the inhibition of
other GPCRs at cellular levels (Meirsman et al., 2016;Ye
et al., 2019).
GPR88 in CNS disorders
Recent research using GPR88 knockout mice revealed that
medium spiny neurons had more glutamatergic excitation
and lower GABAergic inhibition. This raised neuronal firing
rates in vivo and caused hyperactivity, poor motor coordin-
ation and decreased cue-based learning in the GPR88 knock-
out mice. These results demonstrate that modulators of
GPR88 activity in vivo may have therapeutic utility for treat-
ing CNS diseases (Bi et al., 2015). CNS diseases include
Alzheimer’s, Parkinson’s, brain cancer and stroke which have
costly treatments. Therefore, there is an urgent need for
newer therapeutic medicines based on new targets.
GPR88 agonists
There is an urge to develop potent and selective ligands for
GPR88 owing to its important and fast emerging recognition
as an important pharmaceutical target. The two major types
of the known agonists that have been taken into study in
this paper are 2-PCCA [(1R,2R)-N-[(2S,3S)-2-amino-3-methylpen-
tyl]-N-[4-(4-propylphenyl)phenyl]-2-pyridin-2-ylcyclopropane-1-
carboxamide] and RTI-13951-33 [(1R,2R)-N-[(2R,3R)-2-amino-3-
methoxybutyl]-N-[4-[4-(methoxymethyl)phenyl]phenyl]-2-pyridin-
2-ylcyclopropane-1-carboxamide].
2-PCCA is a synthetic small molecule which has an effect
on GPR88 activity. Jin et al., (2014) demonstrated that GPR88
couples to the Ga
i
subunits and is activated by 2-PCCA in
both transient and stable cells that express the GPR88 pro-
tein. The 2-PCCA agonist inhibited isoproterenol-stimulated
cAMP accumulation with an EC
50
of 603 nM in a GloSensor
cAMP assay using stable HEK293-GPR88-pGloSensor22F cells,
suggesting GPR88 is coupled to Ga
i
proteins (Decker et al.,
2017). Later, the team also reported the discovery of RTI-
13951-33, which was the first selective and brain-penetrant
GPR88 agonist. In an in vitro cAMP functional experiment,
RTI-13951-33 had an EC
50
of 25 nM and showed no discern-
ible off-target activity at any of the 38 GPCRs, ion channels,
or neurotransmitter transporters that were examined (Jin
et al., 2018). Figure 3 shows the structure of the mentioned
known agonists.
Considering the importance of GPR88 and the current
need to address the lack of GPR88-specific ligands, we have
created GPCR and CNS-targeted small molecule dataset and
screened against the GPR88 target. In this study, virtual
screening techniques were used to shortlist the potential
leads that showed good binding with the GPR88 protein.
Figure 1. Pictorial representation of GPCR in membrane bilayer.
2 V. GARISETTI ET AL.

Materials and methods
Structure Prediction
A versatile tool named GPCRM (Miszta et al., 2018) is a struc-
ture modelling server which predicts the structure of GPCRs.
In principle, GPCRM builds a GPCR model using a
MODELLER-based homology modeling procedure (Jain et al.,
2015). GPCRM integrates a number of tools and methodolo-
gies to create the final model, including: i) template detec-
tion and alignment generation (alignment with all template
sequences in the GPCRM database using MUSCLE (Edgar,
2004) and ClustalW2 (Larkin et al., 2007)); ii) model building
with MODELLER; iii) loop refinement (the top 10 models
were chosen for a loop refinement in ROSETTA); and iv)
refinement using the all-atom ROSETTA force field (Bonneau
et al., 2001). GPCRM incorporates a Z-coordinate-based filter
to generate only such GPCR models in which extra- and
intra-cellular loops do not penetrate the membrane. The best
predicted structure was used for further studies.
Virtual screening pipeline
Working dataset
Millions of chemical compounds can be computationally
chosen from virtual databases for further study based on
structural similarity with known agonists and complementar-
ity with target binding site of a biological macromolecule.
GPR88 being a CNS-targeted protein, we combined data-
bases of GPCR-related libraries and CNS-targeted drug
libraries from the ChemDiv Database Library. We created a
working database of 4,81,250 compounds by combining
70,371 compounds from CNS-targeted focus libraries and
4,10,879 compounds from GPCR-targeted libraries. These
compounds were first screened based on the scaffold shape
of the two known agonists, 2-PCCA and RTI-13951-33 and
the screened compounds were then subjected to structure-
based virtual screening using the active site of the GPR88
protein. Figure 4 gives an overview of the workflow of the
present study.
Shape-based screening using PHASE
The shape-based screening was performed using the PHASE
module of Schr€
odinger suite (Schr€
odinger Release 2022-23;
Dixon et al., 2006). We used the predicted 3D structure of
GPR88 as the query structure. The pharmacophore volume
score setting, which treats each chemical as a collection of
pharmacophore properties, was used to screen the com-
pounds. These features are aromatic group, hydrogen bond
acceptor (HBA) group, hydrogen bond donor (HBD) group,
hydrophobic group, positive charge group and negative
charge group. A pharmacophoric tolerance sphere of radius
2 Å was generated. For each molecule, 100 conformers were
created throughout the screening process and the top 10
conformers were selected based on their pharmacophore
similarity. PHASE shape similarity score was computed based
on the maximum aligned features for these conformers.
Figure 2. GPR88 signalling pathway.
Figure 3. Structures of known agonists, 2-PCCA and RTI-13951-33.
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 3

Compounds which showed a shape similarity of more than
55% was further subjected to structure-based screening.
Structure-based screening using induced fit docking
The shortlisted compounds, after shape-based screening of
the agonists, were docked with target GPR88 protein to
study the binding affinity of the protein–ligand complex
using induced-fit docking (IFD) protocol implemented in
Schr€
odinger (Schr€
odinger Release 2022-23; Sherman et al.,
2006). The IFD module is the most powerful and accurate
method to account for both ligand and receptor flexibilities.
For this method, initially the ligands were docked to the
rigid protein with van der Waals radii scaling of 0.5 for the
protein and ligand. The entire protein molecule was sub-
jected to constrained minimization with OPLS-2005 force
field with an implicit solvation model. The initial docking was
carried out using Standard Precision mode and the number
of poses generated was set for further steps of refinements.
Each structure from the previous step was subjected to
prime side-chain minimization. Then, the receptor structures
that were within 30 kcal/mol of the minimal energy structure
were passed through for a final round of glide docking and
scoring. In the final step, each ligand was redocked into
every refined low-energy receptor structure that is produced
in the second step using glide XP at default settings. An IFD
score, that explains the protein ligand interaction and the
total energy of the system, was calculated and used to rank
the IFD poses. The top-ranked poses were analyzed based
on their docking score, glide energy and the interactions
formed with the binding pocket. The docking score repre-
sents the predicted binding affinity of a ligand to a protein
based on the shape complementarity and electrostatic inter-
actions between the two molecules. It is a widely used
metric in docking studies and provides a quick estimate of
ligand binding affinity. On the other hand, glide energy is a
more accurate measure of ligand binding affinity that takes
into account the internal energy of the ligand and the pro-
tein-ligand interactions (Friesner et al., 2004). Further, the
best poses of protein-ligand docked complexes were ana-
lyzed by using PyMOL (PyMOL), Chimera (Pettersen et al.,
2004) and Maestro view in Schrodinger (Schr€
odinger Release
2022-23).
Molecular dynamics simulations (MDS)
In silico system composition
1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC)
bilayer was generated using the CHARMM-GUI Membrane
Builder web tool (Jo et al., 2009). The GPR88 predicted struc-
ture in complex with the screened compounds and the ago-
nists was oriented using the OPM (Orientations of Proteins in
Membranes) database server (Lomize et al., 2012), where it
orients the membrane proteins with respect to the hydrocar-
bon core of the lipid bilayer. The oriented structure was then
embedded into a lipid bilayer consisting of 256 POPC mole-
cules with 128 lipids in the upper leaflet and 128 lipids in
the lower leaflet. The system was solvated with TIP3P water
and neutralised with 0.15 M of potassium chloride ions
(Bruzzese et al., 2018; Im & Roux, 2002). The potassium ions,
chloride ions and water molecules were added accordingly
to the system.
Molecular dynamics simulation protocols
All simulations, including minimization, equilibration and pro-
duction runs were performed using GROMACS (Abraham
et al., 2015) with the CHARMM36 force field. The protein-lipid
Figure 4. Workflow of the study.
4 V. GARISETTI ET AL.

bilayer system was minimized using 5000 steps of steepest
descent minimization followed by 5000 steps of conjugate
gradient minimization. The minimized system was subjected
to NVT and NPT ensembles and underwent a total of six
equilibration steps. The initial three steps lasted 125 ps each,
while the following three steps lasted 250 ps each. A force-
based switching function was used to turn off the van der
Waals interactions smoothly at 10-12 Å. Temperature was
maintained at 303 K using Langevin dynamics with a cou-
pling coefficient of 1 ps
1
. The Nos
e-Hoover Langevin-piston
method (Bussi et al., 2009) was used to maintain constant
pressure at 1 bar with a piston period of 50 fs and a piston
decay of 25 fs. The Nos
e-Hoover thermostat is widely used in
membrane simulations because it produces a correct kinetic
ensemble and allows for fluctuations that produce more nat-
ural dynamics. The pressure was controlled in semi-isotropic
scaling with constant surface tension using Parrinello-
Rahman barostat with target pressure of 1 bar (Berendsen
et al., 1984). We began the equilibration with Berendsen
thermostat and switched to Nos
e-Hoover for the final pro-
duction of 150 ns MD. Simulation trajectory was saved in
100 ps intervals which was used for further analysis.
QikProp analysis
In order to forecast the ADMET (absorption, distribution,
metabolism, excretion and toxicity) characteristics of organic
molecules, we utilized the QikProp module within the
Schr€
odinger suite (Schr€
odinger Release 2022-23). QikProp uti-
lizes a range of physically significant descriptors and pharma-
ceutically relevant properties to estimate these characteristics
of organic molecules. Also, QikProp offers a comparison
range for each molecule’s properties in relation to those of
95% of known drugs.
Results and discussions
Structure Prediction
The GPR88 sequence (Uniprot ID: Q9GZN0) was used as the
target sequence in GPCRM to predict its three-dimensional
protein structure. The sequence was aligned with all the
template sequences in the GPCRM database by applying the
MUSCLE (Edgar, 2004) and ClustalW2 (Larkin et al., 2007)
sequence alignment algorithms. The final template used for
structure prediction was PDB ID: 2Y02 (Warne et al., 2011)
(b
1
adrenergic receptor with a bound co-crystal ligand)
which showed 18.066% similarity with GPR88 sequence. The
sequence reported in the PDB ID: 2Y02 has 6 mutations and
three truncations (1-44 N-terminus, 232-287 cytoplasmic loop
3 and 339-440 C-terminus). In order to compare the
sequence of the target with the template, we have used the
sequence of b
1
adrenergic receptor using the Uniprot ID:
P07700. Figure 5 shows the alignment of the target and the
template sequences (16.66% similarity) which was created
using the ESPript server (ESPript 3). GPCRM uses MODELLER
for model building, the ten best predicted models were then
selected for loop refinement in ROSETTA. After the loop
refinement, ten best models were predicted and the best
GPR88 model was selected based on the ROSETTA score. The
predicted GPR88 model structure and transmembrane archi-
tecture highlighting the seven-spanning transmembrane,
which is the GPCR canonical structure, is shown in Figure 6.
Virtual screening pipeline
In the present study, we screened ChemDiv database
libraries which were specifically related to GPCR and CNS-
targeted drugs. We combined 70,371 compounds from CNS-
targeted libraries and 4,10,879 compounds from GPCR-targeted
libraries, resulting in a working dataset of 4,81,250 com-
pounds. The compounds were foremost subjected to
shape-based screening then followed by structure-based
screening.
Shape-based screening approach
In shape-based screening, each conformer of the initial data-
set was aligned to the query ligands and a similarity was cal-
culated based on the conformer shape of the known two
agonists. The known agonists, 2-PCCA and RTI-13951-33,
were used as the query ligands for shape screening against
the dataset. The compounds recovered from the PHASE
screening were further sorted using the PHASE similarity
scores. In the present study, a total of 73,536 compounds,
which showed a shape similarity of more than 55% were
obtained and these molecules were then carried forward for
further analysis.
Structure-based screening approach
The 73,536 compounds were further subjected to structure-
based screening based on the active site of the GPR88 pro-
tein structure. In Class A GPCRs, the endogenous ligand
binding pocket is typically in the transmembrane (TM)
domain, near the extracellular (EC) region. The ligand binding
pocket in majority of class A GPCRs is specifically the cavity
formed between the TM domains 3, 5, 6 and 7 (Yanamala &
Klein-Seetharaman, 2010). The crystal structure of the tem-
plate showed a bound ligand carmoterol and the binding
site of the ligand was considered as the active site for the
docking studies of the agonists and lead compounds with
GPR88. In the crystal structure of 2Y02, the co-crystal ligand
specifically binds between 3, 5, 6 and 7 TM domains. Figure
7highlights the transmembrane domains and co-crystal
binding pocket showing the orthosteric binding site made
up of TM domains 3, 5, 6 and 7 as mentioned (Cantarini
et al., 2023). Figure 7 also shows the seven transmembrane
domains (TM1-TM7), the extracellular loops (EL1 and EL2)
and the intracellular loop (IL1). The other connecting loops
(EL3, IL2 and IL3) are missing in the template crystal struc-
ture (PDB id: 2Y02). On superposition of the GPR88 modelled
structure with the template (Figure 7c), the ligand binding
pocket of the template showed a high degree of structural
similarity with GPR88. Based on these observations, the tem-
plate’s ligand binding site was selected as a potential active
site for GPR88 where IFD of the screened compounds was
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 5

done using the GLIDE module. From the results, 11 com-
pounds were shortlisted on analysing the active site, hydro-
gen bonds, the docking score and glide energies. Table S1
shows the two-dimensional structures and the IUPAC names
of the two known agonists and shortlisted 11 compounds
and Table 1 shows the docking score, glide energy, the
hydrogen bond interactions and hydrophobic interactions of
the docked compounds with the protein receptor (GPR88).
The 11 compounds were further analyzed by studying the
active site residues, hydrogen bonds between the ligand and
the protein, docking score and glide energy.
As glide energy is a more accurate measure to study the
binding affinity of the ligand with the protein, we have
observed that GPR88-lead complexes showed good glide
energy when compared with the GPR88-known agonists.
GPR88-lead1 complex showed a glide energy of
71.90 kcal/mol and GPR88-lead2 complex showed
77.83 kcal/mol which are better than that of the GPR88-
known agonist complexes. Coming to the known agonists, 2-
PCCA and RTI-13951-33 showed glide energies of
62.73 kcal/mol and 65.35 kcal/mol, respectively. Figure 8
highlights the binding pocket of GPR88 where the two
Figure 5. Sequence alignment of the template (b
1
adrenergic receptor –B1AR) sequence with the target GPR88 sequence.
Figure 6. (a) Crystal structure of the template (PDB id: 2Y02) with the bound ligand (b) three-dimensional predicted model for GPR88 with the transmembrane
(TM) helices (TM1–TM7) highlighted from the top view of the predicted model.
6 V. GARISETTI ET AL.

shortlisted lead compounds and the two agonists bind.
Figure 9 shows the protein-ligand interactions at the active
site of the protein highlighting the amino acid residues
involved in hydrogen bonding and hydrophobic interactions.
The IUPAC names from Table S1 specify that lead1 is a
furan and quinazoline derivative and lead2 is a quinazoline
derivative. Both furan and quinazoline-based compounds
have been used as target drugs for central nervous system
disorders. Furan is a heterocyclic organic compound and it
consists of a five-membered aromatic ring with four carbon
atoms and one oxygen. It is an essential class of heterocyclic
compound that has important biological properties. Furan
has the ability to infiltrate different organs and pass through
the biological membranes (Sen et al., 2010). Quinazolines are
one of the most significant heterocycles in medicinal chemis-
try and have a wide range of biological properties. Recent
studies have highlighted that quinazolines are regarded to
be a potential class of versatile bioactive heterocyclic chemi-
cals in design and production of novel CNS-active medicines
(Jafari et al., 2016). Figure 10 highlights the furan and quina-
zoline moieties in the two lead compounds.
Molecular dynamics simulation
To assess the stability of the GPR88-agonist/lead complexes,
a 150 ns simulation run was performed using the GROMACS
software. The stability of the complexes was measured based
on the root mean square distance (RMSD), root mean square
fluctuation (RMSF), radius of gyration (RGYR) and the solvent
accessible surface area (SASA) graphs. MD simulation was
carried out for four GPR88 complexes and the apo (without
ligand) form of GPR88. The four complexes include GPR88-
RTI-13951-33, GPR88-2-PCCA, GPR88-lead1 and GPR88-lead2.
During the 150 ns MD simulation, the RMSD of the backbone
atoms of GPR88-lead1 and lead2 complexes showed lesser devia-
tions when compared to GPR88-agonist complexes. From Figure
11a, it can be seen that lead1 and lead2 showed the least devia-
tions and a converging graph over the last 50ns indicating its
stability. The average RMSD values for the backbone atoms of
the GPR88-lead1 and lead2 complexes over the last 50ns were
0.41 nm and 0.41 nm, respectively, which were lower than the
RMSD values obtained for the GPR88-agonist complexes RTI-
13951-33 (0.55 nm) and 2-PCCA (0.49 nm). This suggests that the
lead compounds have a comparatively stronger and stable inter-
action with the protein than the known agonists. The ligand
RMSD graph in Figure 11b shows how the RMSD values of the
ligand change over time, helping us to assess the stability of the
protein-ligand complex. The ligand RMSD graphs of the lead
complexes remain relatively stable throughout the simulation,
indicating that the ligand is tightly bound to the protein, and
the GPR88 complexes with the lead compounds are stable.
When we looked at the average RMSD values for the ligand
atoms over the last 50 ns, we found that the lead compounds
had lower values (0.15 nm for lead1 and 0.12 nm for lead2) than
the agonists (0.18 nm for RTI-13951-33 and 0.17 nm for 2-PCCA).
These findings suggest that the lead compounds have a stronger
interaction with the protein and are stable at the active site of
theGPR88receptor.TheRMSFgraphhighlightstheresidualfluc-
tuations in the four complexes and the apo form of GPR88. In
Figure 11c, the graph shows that the amino acid residues of
GPR88-lead complexes have lower fluctuations compared to the
GPR88-known agonist complexes. This indicates that lead com-
pounds may stabilize the protein complex better. The key resi-
dues forming the active site have been marked with arrows in
the RMSF graph and it can be noticed that they show less fluctu-
ations suggesting that the residues remained in a relatively fixed
position throughout the simulation. This observation throws light
on the structural stability of the binding site.
The RGYR graph shows the compactness of the protein
molecule over time, while the SASA graph provides informa-
tion about the protein’s surface area exposed to solvent.
Together, they give an idea of the protein’s shape and its
interaction with the surrounding solvent during the simula-
tion. From the graph in Figure 12a, the radius of gyration of
GPR88-lead complexes show a relatively stable trend com-
pared to the GPR88-known agonist complexes; it suggests
that the protein is in a stable conformation and is not under-
going any significant structural changes during the
Figure 7. (a) Seven-transmembrane domain architecture of 2Y02; TM: transmembrane domain, EL: extracellular domain, IL: intracellular domain (b) residues at the
orthosteric binding site of the co-crystal ligand specifically interacting with TM 3, 5, 6 and 7 (c) superposition of the template crystal structure (2Y02) with the pre-
dicted GPR88 model (in grey).
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 7

simulation. Figure 12b shows the SASA trends for the apo
form, GPR88-lead complexes and the GPR88-known agonist
complexes. The average SASA over the last 50 ns was calcu-
lated to be 237.91 nm
2
for GPR88-RTI-13951-33 complex and
230.95 nm
2
for GPR88-2-PCCA and for the lead compounds,
the average SASA was lower. Over the last 50 ns, GPR88-
lead1 complex showed an average SASA of 217.17 nm
2
and
GPR88-lead2 complex showed 228.43 nm
2
. Particularly, the
SASA of the protein when in complex with the lead 1 shows
the least average SASA over the last 50 ns and also shows a
steady decreasing trend in the graph indicating that lead1 is
interacting more strongly with the protein, becoming more
buried in the protein’s binding pocket, and forming stronger
interactions with surrounding residues.
Figure 13 shows the graphs which give the total number of
hydrogen bond contacts made by the ligand with the protein
over the 150 ns simulation run. It can be seen that a good num-
ber of hydrogen bonds are being maintained for all the four
complexes. Particularly, the consistency of hydrogen bonding
observed in GPR88-lead complexes suggests the stable interac-
tions of the lead1 and lead2 with the protein throughout the
simulation. The hydrogen bond interactions can also be observed
in Figure 14, where the conformational snapshots of the com-
plexes at 50, 100 and 150ns have been presented.
Table 1. Molecular docking analysis of agonists and lead compounds.
S.No. Compound
Docking Score
(kcal/mol)
Glide Energy
(kcal/mol)
Hydrogen bond interactions
Hydrophobic interactionsResidues Distance (Å)
12-PCCA (Agonist) 212.37 262.73 [ASN230] N-H …N 2.94
[Arg116] N-H …N 2.98
Trp84, Leu120, Leu124, Pro177,
Ala180, Pro181, Pro183, Tyr212,
Leu213, Phe231, His235,
Leu300, Val301, Ser304,
Gln318, Ser321, Cys325
2RTI-13951-33
(Agonist)
212.04 265.35 N-H …O [Ser304] 2.86
[Asn230] N-H …N 3.08
Arg116, Leu120, Leu124,
Pro181, Arg182, Ala185,
Ala186, Cys211, Tyr212,
Leu213, Ile215, Val216, Ser227,
Phe231, Leu234, Leu300,
Val301
3 C498-0437 11.90 61.31 [Arg116] N-H …O 3.18
[Arg116] N-H …O 3.04
N-H …O [Ser321] 2.97
N-H …O [Cys211] 2.97
Gly117, Ley120, Leu176, Pro177,
Al180, Leu209, His210, Leu213,
Val226
4C487-0582
(lead1)
211.24 271.90 [Arg116] N-H …O 2.83
[Asn230] N-H …O 2.97
[Ser304] O-H …O 2.87
Trp84, Leu115, Leu120, Gly121,
Leu124, Leu213, Val223,
Val226, Ser227, Leu234, His235,
Leu300, Ser321, Leu324, Cys325
5 E136-0992 10.88 54.16 [Arg116] N-H …O 2.79
[Leu213] N-H …O 3.04
N-H …O [Gln318] 2.90
[Ser321] O-H …O 2.83
Gly118, Leu120, Gly121, Leu124,
Tyr212, Leu300, Pro314, Val317,
Leu324
6C487-0086
(lead2)
210.72 277.83 N-H …O [Ser304] 3.15 Trp84, Arg116, Gly117, Gly118,
Leu120, Gly121, Leu176,
Ala180, Pro181, Arg182,
Pro183, Leu213, Gly214,
Val216, Val226, Asn230,
Leu300, Ser321, Leu324
7 C487-0185 10.64 73.34 N-H …O [Cys211] 2.88
[Asn230] N-H …O 2.80
Arg116, Gly117, Pro177, Ala180,
Pro181, Pro183, Leu209, His210,
Tyr212, Leu213, Gly214, Val216,
Val226, Ser227, His235, Leu300,
Val301, Ser304, Leu324
8 F389-0505 10.57 74.34 [Arg116] N-H …O 2.82
[Ser227] O-H …O 2.92
N-H …O [Asn230] 2.88
N-H …O [Ser321] 2.99
Gly117, Leu176, Pro177, Ala180,
Pro181, Arg182, His210, Tyr212,
Leu213, Val226, Leu300, Val301,
Ser304, Val317
9 E136-0431 10.49 54.68 [Arg116] N-H …O 3.09
[Arg116] N-H …O 2.95
[Leu213] N-H …O 2.89
N-H …O [Gln318] 3.07
Leu120, Gly121, Leu124, Tyr212,
His235, Leu300, Val313, Pro314,
Val317, Ser321, Leu324
10 E136-0411 10.21 56.55 [Arg116] N-H …O 2.79
[Leu213] N-H …O 2.86
N-H …O [Gln318] 2.99
Pro95, Tyr212, Leu300, Pro314,
Val317, Ser321
11 E136-0691 10.17 57.60 [Arg116] N-H …O 2.91
[Leu213] N-H …O 2.88
N-H …O [Gln318] 2.79
Leu120, Gly121, Leu124, Tyr212,
Leu300, Pro312, Val317, Ser321,
Leu324
12 F295-0048 9.51 60.46 [Arg116] N-H …O 2.97
[Arg116] N-H …O 2.92
[Gly121] N-H …O 2.89
O-H …O [Ser321] 2.72
Trp84, Leu120, Leu124, Ala180,
Val226, Asn230, His235, Leu300,
Val317, Leu324, Cys325
13 E136-0695 8.88 48.55 [Arg116] N-H …O 2.66
[Leu213] N-H …O 2.86
N-H …O [Gln318] 3.13
Leu115, Gly121, Leu124, Cys211,
Tyr212, Pro314, Val317, Ser321,
Leu324
8 V. GARISETTI ET AL.

Conformational snapshots at regular intervals from MD
trajectory
Conformational snapshots were taken at regular intervals from
the 150 ns MD trajectory of the GPR88-agonists complexes, as
well as the GPR88-lead complexes, to investigate the poses and
interactions with the protein. All compounds were observed to
remain within the binding pocket of the protein throughout the
simulation. The agonist RTI-13951-33 initially made interactions
with SER 304 and ASN 230, forming hydrogen bonds with ASN
230 in the 50 ns frame and with ARG 217 and SER 304 in the
100 ns frame. However, no interactions were observed in the
150 ns frame. Similarly, the agonist 2-PCCA initially interacted
with ASN 230 and ARG 116, but over the time it showed no
hydrogen bond interactions with the protein in the 100 and
150 ns frames. In contrast, lead1 consistently made interactions
with ASN 230 at all intervals, suggesting a stable binding mode.
Lead2 also made interactions with ARG 116 and GLY 117 at 50,
100, 150 ns intervals, indicating a stable binding mode and
potential for strong interactions with the protein’s catalytic resi-
dues throughout the simulation. Thus, the lead compounds were
found to make interactions with one or more key residues within
Figure 8. Superposition of the docked complexes of known agonists 2-PCCA (blue) and RTI-13951-33 (pink) and the two best lead compounds, lead1(yellow) and
lead2 (green).
Figure 9. Interactions of the known agonists (a) 2-PCCA, (b) RTI-13951-33 (c) lead1 (d) lead2 at the active site of GPR88.
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 9

the binding pocket during the majority of the 150ns MD simula-
tion (Figure 14).
From the MD simulation run, it was understood that
GPR88 in complex with the identified lead molecules showed
a stable system when compared to the known agonists.
Steady binding of the lead compounds was observed over
the simulation, and the hydrogen bonds and the binding
energy further support our study.
MM/PBSA-based free energy calculations from MD
trajectory
MM/PBSA-based free energy calculations were also per-
formed for the four complexes, two with the known agonists;
GPR88-RTI-13951-33 and GPR88-2-PCCA and two with the
lead compounds; GPR88-lead1 and GPR88-lead2. The results
show that lead1 has higher binding affinity than the two
Figure 10. 2D structures of lead1 and lead2 compounds, the furan and quinazoline moieties are highlighted in dotted circles.
Figure 11. (a) shows the RMSD graph (backbone) of GPR88 bound with lead compounds, agonists and the apo form of GPR88 (b) shows the ligand RMSD graph
(c) shows the RMSF graph of the residual fluctuation.
10 V. GARISETTI ET AL.

known agonists, as validated by its more negative total bind-
ing energy. The total binding energies were 151.55,
130.64, 140.13 and 138.708 kJ/mol for lead1, lead2, RTI-
13951-33 and 2-PCCA, respectively. The total binding energy
for lead1 indicates that lead1 forms more stable protein-lig-
and complex.
Furthermore, the individual energy components also sup-
port the higher binding affinity of the lead compounds.
Lead1 and lead2 exhibit stronger van der Waals and polar
solvation interactions, as well as slightly weaker electrostatic
interactions, with the protein than the agonists. The SASA
values for the lead compounds are also lower than those for
the agonists, indicating a tighter binding of the lead com-
pounds to the protein.
In conclusion, the MM/PBSA binding free energy calcula-
tions suggest that lead1 has higher binding affinity and sta-
bility with the protein than lead2 and the known agonists,
RTI-13951-33 and 2-PCCA. Though lead2 showed slightly
Figure 12. (a) Radius of gyration and (b) SASA graph of the GPR88 apo form, GPR88 with the two agonists and two lead compounds.
Figure 13. Number of hydrogen bond interactions formed during the MD simulation in the case of GPR88-known agonists and GPR88-lead compounds complexes
over the 150 ns MD run.
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 11

higher total binding energy, it showed better van der Waals,
electrostatic and polar solvation energies when compared to
the known agonists. Overall, these results support the poten-
tial of both lead1 and lead2 as promising candidates for fur-
ther development as GPR88 agonists (Table 2).
QikProp analysis
The physicochemical properties of a drug are essential fac-
tors in the process of drug discovery and development. In
our study, we evaluated the physicochemical properties of
the two lead compounds, lead1 and lead2 through QikProp
analysis. We compared the results with the known agonists
and presented the pharmacologically relevant properties in
Table 3. Additionally, we provided a range of properties that
meet the 95% range of known drugs for comparison pur-
poses. From Table 3, it can be seen that both lead1 and
lead2 have higher QPlogPw values when compared to the
known agonists. QPlogPw is the measure of lipophilicity,
thus a higher value indicates a higher potential to cross lipid
membranes and reach their target sites. Additionally, the
lead compounds also show moderate CNS activity, same as
that of the known agonists. Lead1 showed a medium HOA
value, indicating moderate levels of oral absorption, and
lead2 showed low HOA value. Furthermore, both lead
Figure 14. Binding poses of two known agonists and the lead compounds during 150 ns MD simulation at every 50 ns time interval.
Table 2. Binding free energy calculation using MM/PBSA in GROMACS.
S.No. Compounds
van der Waals
(kJ/mol)
Electrostatic
(kJ/mol)
Polar solvation
(kJ/mol)
SASA
(kJ/mol)
Total binding
energy (kJ/mol)
1 2-PCCA 244.24 9.49 143.70 28.68 138.708
2 RTI-13951-33 234.30 41.10 185.44 27.79 140.13
3 Lead1 304.77 56.39 245.09 29.48 151.55
4 Lead2 280.51 51.03 230.96 30.06 130.64
Table 3. QikProp results of the lead compounds and known agonists.
Compound molMW SASA Donor HB Accpt HB CNS HOA QplogPoct QplogPw QplogPo/w QplogS QplogBB QplogKp
Reference
range
130–725 500–2000 0.0–6.0 2.0–20.0 2: inactive
to þ2: active
1: low
2: medium
3: high
8.0–35.0 4.0–45.0 2.0–6.5 6.5–0.5 3.0–1.2 8.0–1.0
RTI-13951-33 479.75 718.54 4 9.6 1 2 25.96 14.87 2.16 0.42 0.01 6.04
2-PCCA 475.80 807.44 4 6.2 1 3 25.08 11.58 4.14 2.75 0.14 5.62
Lead1 526.50 705.39 6 15.5 1 2 31.54 17.14 3.58 2 0.44 6.01
Lead2 528.60 709.92 6 15.5 1 2 31.77 20.22 2.69 2 0.18 6.05
12 V. GARISETTI ET AL.

compounds have acceptable QPlogPo/w and QplogPoct val-
ues within the reference range, indicating they have suitable
lipophilicity to be drug candidates. They also have accept-
able QplogKp values, suggesting they have good permeabil-
ity across the blood-brain barrier, which is important for CNS
drugs. Overall, the QikProp values of the lead compounds
suggest they have favourable drug-like properties and could
potentially be developed as agonists. However, further stud-
ies such as in vitro and in vivo assays are necessary to con-
firm their activity and efficacy.
The QikProp descriptors are molecular weight (molMW),
total solvent accessible surface area (SASA), estimated num-
ber of hydrogen bond donor (donorHB) and acceptor
(accptHB) in aqueous solution, Predicted central nervous sys-
tem activity (CNS) and Human Oral Adsorption (HOA).
QplogPoct, QplogPw, QplogPo/w, QplogS, QplogBB and
QplogKp are predicted partition coefficients of octanol/gas,
water/gas, octanol/water, aqueous solubility, brain/blood,
whereas QplogKp is predicted skin permeability. Numbers in
the reference range are for 95% of known drugs or recom-
mended value.
Conclusions
GPR88 is an orphan GPCR protein, which has high expression
in the brain striatum. Experiments reported in literature
involving GPR88-knockout mice have shown that they show
increased locomotor activity in response to dopaminergic
compounds and also, studies have stated that GPR88 is a
crucial moderator of CNS signalling pathways associated with
mental illness. All the findings imply that regulating GPR88
activity may be therapeutically useful in the management of
a variety of CNS-related illnesses. 2-PCCA and RTI-13951-33
are two known agonists of GPR88 which have known to
influence the GPR88 function and its activation. In the pre-
sent study, we have put together the databases of GPCR-
related libraries, and CNS-targeted drug libraries from the
ChemDiv Database Library and created a working database
of 4,81,250 compounds. Homology modelling was used to
predict the GPR88 structure; shape and structure-based
screening was carried out, after which two compounds lead1
and lead2 were shortlisted based on favourable glide energy
and key interactions at the active site. A 150 ns molecular
dynamics simulation was carried out for four GPR88 com-
plexes, with two agonists, the two lead compounds and the
apo form of GPR88. The simulation result analysis provided
us valuable information on the stability of the complexes
and showed that the lead compounds exhibited stable bind-
ing and favourable dynamics over the 150 ns simulation
time. MM/PBSA free energy calculations after the MD run
showed that lead1 particularly showed better van der Waals
and total binding energy. Furthermore, we performed ADME
analysis using QikProp on the lead compounds and found
them to have promising drug-like properties. Thus, the lead
compounds could be potential agonists and add to the list
of synthetic ligands which would further help to study and
elucidate the potential of GPR88 in CNS disorders.
Acknowledgements
Vasavi Garisetti thanks the Department of Science and Technology,
Government of India, for the INSPIRE fellowship (IF160596). Anantha
Krishnan Dhanabalan thanks the Indian Council of Medical Research,
Government of India for Senior Research Fellowship (Grant No:
ISRM/11(69)/2017). Authors thank DST-FIST, Government of India for the
computational facilities sanctioned to the department. Authors are also
thankful to the Schr€
odinger team for providing a trial version of the
Schr€
odinger suite for the research work.
Disclosure statement
Authors declare no conflicts of interest.
Funding
The candidates have received the fellowship from Department of
Science and Technology, Ministry of Science and Technology, India; and
Indian Council of Medical Research, Government of India, but the
research presented in this paper was conducted without any funding
support from any agencies.
References
Abraham, M. J., Murtola, T., Schulz, R., P
all, S., Smith, J. C., Hess, B., &
Lindah, E. (2015). Gromacs: High performance molecular simulations
through multi-level parallelism from laptops to supercomputers.
SoftwareX,1-2,19–25. https://doi.org/10.1016/j.softx.2015.06.001
Acher, F. C. (2006). Orthogonal ligand-receptor pairs. Comprehensive
Medicinal Chemistry II,3, 215–234. https://doi.org/10.1016/b0-08-
045044-x/00084-5
Berendsen, H. J. C., Postma, J. P. M., Van Gunsteren, W. F., Dinola, A., &
Haak, J. R. (1984). Molecular dynamics with coupling to an external
bath. The Journal of Chemical Physics,81(8), 3684–3690. https://doi.
org/10.1063/1.448118
Bi, Y., Dzierba, C. D., Fink, C., Garcia, Y., Green, M., Han, J., Kwon, S.,
Kumi, G., Liang, Z., Liu, Y., Qiao, Y., Zhang, Y., Zipp, G., Burford, N.,
Ferrante, M., Bertekap, R., Lewis, M., Cacace, A., Westphal, R. S., …
Macor, J. E. (2015). The discovery of potent agonists for GPR88, an
orphan GPCR, for the potential treatment of CNS disorders. Bioorganic
& Medicinal Chemistry Letters,25(7), 1443–1447. https://doi.org/10.
1016/j.bmcl.2015.02.038
Bonneau, R., Tsai, J., Ruczinski, I., Chivian, D., Rohl, C., Strauss, C. E. M., &
Baker, D. (2001). Rosetta in CASP4: Progress in ab initio protein struc-
ture prediction. Proteins: Structure, Function, and Genetics,45(S5), 119–
126. https://doi.org/10.1002/prot.1170
Bruzzese, A., Gil, C., Dalton, J. A. R., & Giraldo, J. (2018). Structural
insights into positive and negative allosteric regulation of a G pro-
tein-coupled receptor through protein-lipid interactions. Scientific
Reports,8(1), 1–14. https://doi.org/10.1038/s41598-018-22735-6
Bussi, G., Zykova-Timan, T., & Parrinello, M. (2009). Isothermal-isobaric
molecular dynamics using stochastic velocity rescaling. The Journal of
Chemical Physics,130(7), 074101–074109. https://doi.org/10.1063/1.
3073889
Cantarini, M., Rusciano, D., Amato, R., Canovai, A., Cammalleri, M., Monte,
M. D., Minnelli, C., Laudadio, E., Mobbili, G., Giorgini, G., & Galeazzi, R.
(2023). Structural basis for agonistic activity and selectivity toward
melatonin receptors hMT1 and hMT2. International Journal of
Molecular Sciences,24(3), 2863. https://doi.org/10.3390/ijms24032863
ESPript - ESPript 3. https://espript.ibcp.fr/ESPript/ESPript/esp_citing.php
Decker, A. M., Gay, E. A., Mathews, K. M., Rosa, T. C., Langston, T. L.,
Maitra, R., & Jin, C. (2017). Development and validation of a high-
throughput calcium mobilization assay for the orphan receptor
GPR88. Journal of Biomedical Science,24(1), 1–9. https://doi.org/10.
1186/s12929-017-0330-3
Dixon, S. L., Smondyrev, A. M., & Rao, S. N. (2006). PHASE: A novel
approach to pharmacophore modeling and 3D database searching.
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 13

Chemical Biology & Drug Design,67(5), 370–372. https://doi.org/10.
1111/j.1747-0285.2006.00384.x
Dzierba, C. D., Bi, Y., Dasgupta, B., Hartz, R. A., Ahuja, V., Cianchetta, G.,
Kumi, G., Dong, L., Aleem, S., Fink, C., Garcia, Y., Green, M., Han, J.,
Kwon, S., Qiao, Y., Wang, J., Zhang, Y., Liu, Y., Zipp, G., …Macor, J. E.
(2015). Design, synthesis, and evaluation of phenylglycinols and phe-
nyl amines as agonists of GPR88. Bioorganic & Medicinal Chemistry
Letters,25(7), 1448–1452. https://doi.org/10.1016/j.bmcl.2015.01.036
Edgar, R. C. (2004). MUSCLE: Multiple sequence alignment with high
accuracy and high throughput. Nucleic Acids Research,32(5), 1792–
1797. https://doi.org/10.1093/nar/gkh340
Friesner, R. A., Banks, J. L., Murphy, R. B., Halgren, T. A., Klicic, J. J., Mainz,
D. T., Repasky, M. P., Knoll, E. H., Shelley, M., Perry, J. K., Shaw, D. E.,
Francis, P., & Shenkin, P. S. (2004). Glide: A new approach for rapid,
accurate docking and scoring. 1. Method and assessment of docking
accuracy. Journal of Medicinal Chemistry,47(7), 1739–1749. https://doi.
org/10.1021/jm0306430
GPCRs - Global Market Trajectory & Analytics. https://www.researchand-
markets.com/reports/1227782/gprotein_coupled_receptors_gpcrs_
global
Im, W., & Roux, B. (2002). Ions and counterions in a biological channel: A
molecular dynamics simulation of ompf porin from Escherichia coli in
an explicit membrane with 1 M KCl aqueous salt solution. Journal of
Molecular Biology,319(5), 1177–1197. https://doi.org/10.1016/S0022-
2836(02)00380-7
Jafari, E., Khajouei, M. R., Hassanzadeh, F., Hakimelahi, G. H., &
Khodarahmi, G. A. (2016). Quinazolinone and quinazoline derivatives:
Recent structures with potent antimicrobial and cytotoxic activities.
Research in Pharmaceutical Sciences,11(1), 1–14. https://journals.lww.
com/rips/Fulltext/2016/11010/Quinazolinone_and_quinazoline_deriva-
tives__recent.1.aspx
Jain, C. K., Gupta, M., Prasad, Y., Wadhwa, G., & Sharma, S. K. (2015).
Homology modelling and molecular dynamics simulations of a pro-
tein serine/threonine phosphatase stp1 in Staphylococcus aureus
N315: A potential drug target. Molecular Simulation,41(7), 592–599.
https://doi.org/10.1080/08927022.2014.902535
Jin,C.,Decker,A.M.,Huang,X.P.,Gilmour,B.P.,Blough,B.E.,Roth,B.L.,
Hu,Y.,Gill,J.B.,&Zhang,X.P.(2014). Synthesis, pharmacological charac-
terization, and structure –Activity relationship studies of small molecular
agonists for the orphan GPR88 receptor. ACS Chemical Neuroscience,5(7),
576–587. https://doi.org/10.1021/cn500082p
Jin, C., Decker, A. M., Makhijani, V. H., Besheer, J., Darcq, E., Kieffer, B. L.,
& Maitra, R. (2018). Discovery of a potent, selective, and brain-pene-
trant small molecule that activates the orphan receptor GPR88 and
reduces alcohol intake. Journal of Medicinal Chemistry,61(15), 6748–
6758. https://doi.org/10.1021/acs.jmedchem.8b00566
Jo,S.,Lim,J.B.,Klauda,J.B.,&Im,W.(2009). CHARMM-GUI membrane
builder for mixed bilayers and its application to yeast membranes.
Biophysical Journal,97(1), 50–58. https://doi.org/10.1016/j.bpj.2009.04.013
Kobilka, B. K. (2007). G protein coupled receptor structure and activation.
Biochimica et Biophysica Acta,1768(4), 794–807. https://doi.org/10.
1016/j.bbamem.2006.10.021
Lameh, J., Cone, R. I., Maeda, S., Philip, M., Corbani, M., N
adasdi, L.,
Ramachandran, J., Smith, G. M., & Sad
ee, W. (1990). Structure and
function of G protein coupled receptors. Pharmaceutical Research,
7(12), 1213–1221. https://doi.org/10.1023/A:1015969301407
Larkin, M. A., Blackshields, G., Brown, N. P., Chenna, R., Mcgettigan, P. A.,
McWilliam, H., Valentin, F., Wallace, I. M., Wilm, A., Lopez, R.,
Thompson, J. D., Gibson, T. J., & Higgins, D. G. (2007). Clustal W and
Clustal X version 2.0. Bioinformatics (Oxford, England),23(21), 2947–
2948. https://doi.org/10.1093/bioinformatics/btm404
Latorraca, N. R., Venkatakrishnan, A. J., & Dror, R. O. (2017). GPCR dynam-
ics: Structures in motion. Chemical Reviews,117(1), 139–155. https://
doi.org/10.1021/acs.chemrev.6b00177
Lefkowitz, R. J., & Shenoy, S. K. (2005). Transduction of receptor signals
by b-arrestins. Science (New York, N.Y.),308(5721), 512–517. https://
doi.org/10.1126/science.1109237
Libert, F., Vassart, G., & Parmentier, M. (1991). Current developments in
G-protein-coupled receptors. Current Opinion in Cell Biology,3(2), 218–
223. https://doi.org/10.1016/0955-0674(91)90142-L
Lomize, M. A., Pogozheva, I. D., Joo, H., Mosberg, H. I., & Lomize, A. L.
(2012). OPM database and PPM web server: Resources for positioning
of proteins in membranes. Nucleic Acids Research,40(Database issue),
D370–D376. https://doi.org/10.1093/nar/gkr703
Lovinger, D. M. (2012). New twist on orphan receptor GPR88 function.
Nature Neuroscience,15(11), 1469–1470. https://doi.org/10.1038/nn.
3244
Meirsman, A. C., Le Merrer, J., Pellissier, L. P., Diaz, J., Clesse, D., Kieffer,
B. L., & Becker, J. A. J. (2016). Mice lacking GPR88 show motor deficit,
improved spatial learning, and low anxiety reversed by delta opioid
antagonist. Biological Psychiatry,79(11), 917–927. https://doi.org/10.
1016/j.biopsych.2015.05.020
Miszta, P., Pasznik, P., Jakowiecki, J., Sztyler, A., Latek, D., & Filipek, S.
(2018). GPCRM: A homology modeling web service with triple mem-
brane-fitted quality assessment of GPCR models. Nucleic Acids
Research,46(W1), W387–W395. https://doi.org/10.1093/nar/gky429
Mizushima, K., Miyamoto, Y., Tsukahara, F., Hirai, M., Sakaki, Y., & Ito, T.
(2000). A novel G-protein-coupled receptor gene expressed in stri-
atum. Genomics,69(3), 314–321. https://doi.org/10.1006/geno.2000.
6340
Pettersen, E. F., Goddard, T. D., Huang, C. C., Couch, G. S., Greenblatt,
D. M., Meng, E. C., & Ferrin, T. E. (2004). UCSF Chimera –A visualiza-
tion system for exploratory research and analysis. Journal of
Computational Chemistry,25(13), 1605–1612. https://doi.org/10.1002/
jcc.20084
PyMOL. The PyMOL molecular graphics system. Version 1.2r3pre.
Schr€
odinger, LLC.
Rasmussen, S. G. F., Devree, B. T., Zou, Y., Kruse, A. C., Chung, K. Y.,
Kobilka, T. S., Thian, F. S., Chae, P. S., Pardon, E., Calinski, D.,
Mathiesen, J. M., Shah, S. T. A., Lyons, J. A., Caffrey, M., Gellman, S. H.,
Steyaert, J., Skiniotis, G., Weis, W. I., Sunahara, R. K., & Kobilka, B. K.
(2011). Crystal structure of the b
2
adrenergic receptor-Gs protein com-
plex. Nature,477(7366), 549–555. https://doi.org/10.1038/nature10361
Schr€
odinger Release 2023-1. (2021). Schr€
odinger, LLC., NY.
Sen, S., De, B., Panday, L., Easwari, T. S., & Nagarajan, A. S. S. (2010).
Physicochemical and pharmacological study of some newly synthe-
sized furan imine derivatives. Journal of Chemical and Pharmaceutical
Research,2(2), 469–477. https://www.jocpr.com/abstract/physico-
chemical-and-pharmacological-study-of-some-newly-synthesized-furan-
imine-derivatives-133.html
Sherman, W., Beard, H. S., & Farid, R. (2006). Use of an induced fit recep-
tor structure in virtual screening. Chemical Biology & Drug Design,
67(1), 83–84. https://doi.org/10.1111/j.1747-0285.2005.00327.x
Wacker, D., Stevens, R. C., & Roth, B. L. (2017). How ligands illuminate
GPCR molecular pharmacology. Cell,170(3), 414–427. https://doi.org/
10.1016/j.cell.2017.07.009
Warne, T., Moukhametzianov, R., Baker, J. G., Nehm
e, R., Patricia, C., Leslie,
A.G.W.,Schertler,G.F.X.,&Tate,C.G.(2011). The structural basis for
agonist and partial agonist action on a b1 -adrenergic receptor. Nature,
469(7329), 241–244. https://doi.org/10.1038/nature09746
Yanamala, N., & Klein-Seetharaman, J. (2010). Allosteric modulation of g
protein coupled receptors by cytoplasmic, transmembrane and extra-
cellular ligands. Pharmaceuticals (Basel, Switzerland),3(10), 3324–3342.
https://doi.org/10.3390/ph3103324
Ye,N.,Li,B.,Mao,Q.,Wold,E.A.,Tian,S.,Allen,J.A.,&Zhou,J.(2019).
Orphan receptor GPR88 as an emerging neurotherapeutic target. ACS
Chemical Neuroscience,10(1), 190–200. https://doi.org/10.1021/acschem-
neuro.8b00572
14 V. GARISETTI ET AL.