Highly Selective Y4Receptor Antagonist Binds in an Allosteric Binding Pocket

Article abstract of DOI:10.1021/acs.jmedchem.0c02000

Human neuropeptide Y receptors (Y1R, Y2R, Y4R, and Y5R) belong to the superfamily of G protein-coupled receptors and play an important role in the regulation of food intake and energy metabolism. We identified and characterized the first selective Y4R allosteric antagonist (S)-VU0637120, an important step toward validating Y receptors as therapeutic targets for metabolic diseases. To obtain insight into the antagonistic mechanism of (S)-VU0637120, we conducted a variety of in vitro, ex vivo, and in silico studies. These studies revealed that (S)-VU0637120 selectively inhibits native Y4R function and binds in an allosteric site located below the binding pocket of the endogenous ligand pancreatic polypeptide in the core of the Y4R transmembrane domains. Taken together, our studies provide a first-of-its-kind tool for probing Y4R function and improve the general understanding of allosteric modulation, ultimately contributing to the rational development of allosteric modulators for peptide-activated G protein-coupled receptors (GPCRs).

Full text of DOI:10.1021/acs.jmedchem.0c02000

pubs.acs.org/jmc
Article
Highly Selective Y4 Receptor Antagonist Binds in an Allosteric
Binding Pocket
̈
Corinna Schuß, Oanh Vu, Mario Schubert, Yu Du, Nigam M. Mishra, Iain R. Tough, Jan Stichel,
C. David Weaver, Kyle A. Emmitte, Helen M. Cox, Jens Meiler,* and Annette G. Beck-Sickinger*
Cite This: J. Med. Chem. 2021, 64, 2801−2814
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Human neuropeptide Y receptors (Y1R, Y2R, Y4R, and
Y5R) belong to the superfamily of G protein-coupled receptors and play
an important role in the regulation of food intake and energy
metabolism. We identified and characterized the first selective Y4R
allosteric antagonist (S)-VU0637120, an important step toward
validating Y receptors as therapeutic targets for metabolic diseases. To
obtain insight into the antagonistic mechanism of (S)-VU0637120, we
conducted a variety of in vitro, ex vivo, and in silico studies. These
studies revealed that (S)-VU0637120 selectively inhibits native Y4R
function and binds in an allosteric site located below the binding pocket
of the endogenous ligand pancreatic polypeptide in the core of the Y4R transmembrane domains. Taken together, our studies
provide a first-of-its-kind tool for probing Y4R function and improve the general understanding of allosteric modulation, ultimately
contributing to the rational development of allosteric modulators for peptide-activated G protein-coupled receptors (GPCRs).
NPY.9−11 However, no highly selective small-molecule
antagonists for Y4R have been described to date.
Aiming at the development of small-molecule probes to
selectively target the human Y4R, approximately 77 000
INTRODUCTION
■
The neuropeptide Y system is a multiligand/multireceptor
family consisting of four Y receptors (Y1R, Y2R, Y4R, and Y5R)
and three peptide ligands: neuropeptide Y (NPY), peptide YY
compounds were tested by high-throughput screening
(PYY), and pancreatic polypeptide (PP). The NPY family is
(HTS). We recently reported on the discovery of Y4R positive
allosteric modulators (PAM) from this experiment.12,13 In
well known for its regulation of important physiological
processes, including energy metabolism and food intake.
Therefore, dysregulation of these processes can result in
disorders, such as anorexia and obesity. Whereas the Y1R and
Y5R induce orexigenic signals, the Y2R and Y4R display
anorexigenic functions. As Y receptors have overlapping
preferences for their peptidic agonists, the development of
subtype-selective compounds is essential to enable a better
understanding of the function of these receptors and to explore
their therapeutic potential.1−3 The Y1R and Y2R bind NPY and
PYY with high affinity. The Y4R has a preference for PP
compared to NPY and PYY, and the Y5R is activated by all
three ligands with comparable potency. Thus, it is very
challenging to develop orthosteric ligands with high selectivity
for one receptor subtype versus the other Y receptors. To date,
a number of agonists and antagonists for Y receptors have been
developed.1,2,4 With respect to Y4R, the native ligand PP
represents a potent agonist that induces anorexigenic effects in
humans.5,6 The major challenge in the development of novel
Y4R ligands remains obtaining selectivity with respect to the
evolutionarily most closely related Y1R subtype.7,8 Selective
Y4R agonists have been identified by crosslinking two peptide
fragments derived from the C-terminal pentapeptide of PP/
comparison to PAMs and agonists, the hit rate of negative
allosteric modulators (NAM) and antagonists was very low
(0.022%, Table S1, Supporting Information), reflecting the
challenges in the development of Y4R antagonists.11,14,15
In this study, we present the in vitro, in silico, and ex vivo
characterization of the first selective, small-molecule Y4R
allosteric antagonist (S)-VU0637120 (Figure 1a), including the
identification of its binding mode that allows further preclinical
studies.
RESULTS AND DISCUSSION
■
Chemistry. Synthesis of (S)- and (R)-VU0637120 was
carried out according to Scheme 1. Heteroaryl amine 116 was
coupled to both (S)-1-(tert-butoxycarbonyl)piperidine-2-car-
boxylic acid (2) and (R)-1-(tert-butoxycarbonyl)piperidine-2-
Received: November 30, 2020
Published: February 17, 2021
https://dx.doi.org/10.1021/acs.jmedchem.0c02000
© 2021 American Chemical Society
J. Med. Chem. 2021, 64, 2801−2814
2801

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Figure 1. Characterization of VU0637120. (a) Structure of VU0637120, chiral C-atom is labeled by an asterisk. (b) VU0637120 effectively
decreases Y4R activation by PP in Ca2+ flux assays using stably transfected COS7_hY4R-eYFP_Δ6Gαqi4myr cells. Data represent mean standard
error of the mean (SEM) from N ≥ 2 independent experiments. VU0637120 effect on PP potency (α) and efficacy (β) was quantified using an
operational model of allosterism.17 (c) Effect of increasing concentrations of VU0637120 and its enantiopure isomers on Y4R activation in response
to a submaximal PP (PP EC80) concentration in COS7_hY4R-eYFP_Δ6Gαqi4myr cells. Data are shown as mean SEM from at least N ≥ 2
independent experiments. (d) Investigation of arrestin3 (arr3) recruitment to Y4R in response to 100 nM PP in the presence of 30 μM of
VU0637120. Kinetic bioluminescence resonance energy transfer (BRET) experiments were performed in HEK293 cells transiently expressing Y4R-
Rluc8 and venus-arr3. Rate constant k of arr3 recruitment is summarized as a bar graph. Data represent the mean SEM from N ≥ 4 independent
experiments. Statistical analysis was performed using unpaired t-test, ***P <0.001. (e) VU0637120 reduces the specific binding of 125I-PP to Y4R
membrane preparations. Binding was investigated in competition binding assays with 60 or 180 pM 125I-PP. Unspecific binding was determined in
the presence of 1 μM PP. Data represent the mean SEM from N ≥ 4 independent experiments. (f) PP competition binding in the presence of
VU0637120 using 125I-PP (60 pM) at Y4R membrane preparations. Data are shown as mean SEM from N ≥ 3 independent experiments. (g)
VU0637120 shows high selectivity to the Y4R subtype. Activation of Y1R, Y2R, and Y5R by NPY and Y4R by PP was investigated in the presence of
30 μM of VU0637120. Receptor signaling was measured in Ca2+ flux assays using COS7 cells stably expressing one specific Y receptor (Y1,2,4,5R-
eYFP) and the chimeric G protein Δ6Gαqi4myr. Data represent the mean SEM from N ≥ 3 independent experiments.
carboxylic acid (3) to afford amides 4 and 5, respectively.
Cleavage of the amine-protecting groups under acidic
conditions provided the secondary amines 6 and 7 in high
yield. Finally, the reaction of 6 and 7 with methanesulfonyl
chloride and trimethylamine gave (S)- and (R)-VU0637120,
respectively.
(S)-VU0637120 Selectively Inhibits Y4R Activation by
PP through an Allosteric Mechanism. The compound
VU0637120 was identified as the most potent Y4R small-
molecule antagonist in the HTS and is characterized in detail
within this study. VU0637120 was first investigated for its
effect to modulate the G protein signaling pathway at the Y4R
by performing Ca2+ flux assays. These studies confirmed that
VU0637120 effectively decreases Y4R activation by PP in a
concentration-dependent manner. This is indicated by the
rightward shift of PP concentration−response curves with
increasing VU0637120 concentrations (Figure 1b). The
activity of VU0637120 at the Y4R was quantified using an
operational model of allosterism.17 This model allows the
calculation of individual parameters like the equilibrium
binding constant (KB) of VU0637120 and the effect of
VU0637120 on the modulation of orthosteric agonist potency
(α) and efficacy (β) (Figure 1b). VU0637120 has a negative
allosteric effect on the affinity of PP to its Y4R orthosteric
binding pocket (α = 0.02) and slightly reduces signaling
efficacy (β = 0.77). The affinity (KB) of VU0637120 to its
allosteric binding pocket is in the range of 300−400 nM (pKB
6.4 0.08). VU0637120 has one chiral center (Figure 1a). To
determine whether both stereoisomers are active, we
synthesized enantiopure (S)- and (R)-VU0637120 and tested
them for their ability to reduce a submaximal PP response at
the Y4R (Figure 1c). Here, (S)-VU0637120 was confirmed as
2802
https://dx.doi.org/10.1021/acs.jmedchem.0c02000
J. Med. Chem. 2021, 64, 2801−2814

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
a
that increasing concentrations of VU0637120 reduce PP
binding at the Y4R (Figure 1e). Whereas high concentrations
of VU0637120 reduce specific binding of 60 pM 125I-PP (1 ×
KD) down to about 15%, the competition curve observed with
180 pM failed to reach a bottom plateau, and even high
concentrations of the compound were not able to entirely
displace the binding of the orthosteric radioligand. Further
binding studies showed that the Ki of PP increases in the
presence of VU0637120, confirming that the compound
modulates the activity of the endogenous ligand (Figure 1f).
These experiments verify an allosteric mechanism of
VU0637120, as the binding of a NAM to its receptor often
induces a low-affinity state for the binding of the orthosteric
ligand, which can result in reduced radioligand binding in
competition binding studies.21 This finding is also in
agreement with the effect of VU0637120 on the reduction of
PP affinity to its Y4R orthosteric binding pocket (α = 0.02) in
Ca2+ flux assay. VU0637120 is structurally distinct from all
previously described Y receptor antagonists, supporting the
proposition of an allosteric mechanism of VU0637120.
With respect to the different functions of Y receptors in the
regulation of satiety and energy metabolism, a high selectivity
of VU0637120 to the Y4R is important.3,24 We first
investigated the effect of VU0637120 on the Y1R, Y2R, and
Y5R in vitro. A high concentration of 30 μM of VU0637120
has no effect on the efficacy (EMax) and potency (EC50) of
NPY-induced response at the Y1R, Y2R, and Y5R, which
demonstrates the selectivity of VU0637120 to the Y4R subtype
(Figure 1g).
Confirmation of (S)-VU0637120 Activity in Ex Vivo
Assays on Mouse Tissue. In addition to the in vitro
characterization of VU0637120, we investigated the activity of
the enantioselective compounds ex vivo in mucosal prepara-
tions of mouse descending colon. This tissue was shown
previously to express native Y4R.25 Electrophysiological
measurements of vectorial ion transport showed a decrease
in the size of rat PP (rPP) responses after exposure of mucosae
to the increasing concentrations of (S)-VU0637120, exhibiting
an IC50 value of 3.8 μM (pIC50 5.42 0.27). Consistent with
results from the Ca2+ flux assays, the (R)-VU0637120
enantiomer had no effect on rPP signaling (Figure 2a). To
further study the selectivity of VU0637120 in mouse colon
mucosa, PYY-induced activation of Y1R and Y2R, which are
also expressed in these mucosal preparations, was examined.
PYY responses were unaffected by either enantiomer (Figure
2b), showing selectivity for the Y4R, rather than Y1R or Y2R
PYY-induced signaling in these native preparations.
Scheme 1. Synthesis of (S)- and (R)-VU0637120
a
Reagents and conditions: (a) N,N-diisopropylethylamine (DIEA), 1-
[bis(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium
3-oxid hexafluorophosphate (HATU), dimethylformamide (DMF),
86% (4), 85% (5); (b) HCl, dioxane, CH2Cl2, 89% (6), 93% (7); (c)
NEt3, MeSO2Cl, CH2Cl2, 60% ((S)-VU0637120), 59% ((R)-
VU0637120).
the active stereoisomer of the compound, as it is able to
efficaciously decrease PP response at the Y4R with an IC50
value of 2.7 μM (pIC50 5.6
0.02). In contrast, (R)-
VU0637120 is inactive at the Y4R up to the top concentration
tested. To validate the effect of the active enantiomer (S)-
VU0637120, we also tested (S)-VU0637120 on the G protein
signaling pathway at the Y4R using different functional assays:
Ca2+ flux assay, a hemiequilibrium kinetic assay, and inositol
phosphate (IP)-One assay as an equilibrium assay.18 Testing of
the pure (S)-VU0637120 confirmed the antagonistic activity in
Ca2+ flux and IP-One assays. Schild analysis of the data reveals
that the antagonistic effect of (S)-VU0637120 is saturable at
concentrations larger than 10 μM, indicating that the
compound does not act as a pure orthosteric antagonist and
support an allosteric mode of action (Figure S1, Supporting
Information).19
As previous studies have already shown the importance of
the arrestin3 (arr3) pathway for Y4R signaling,20 we
investigated the activity of VU0637120 on arr3 recruitment
as an alternative pathway to G protein signaling. PP-induced
arr3 recruitment to the Y4R was analyzed in kinetic
experiments by detecting arr3 recruitment to the Y4R
constantly over 30 min (Figure 1d). The study demonstrates
that VU0637120 slows arr3 recruitment to the Y4R in response
to PP compared to the dimethyl sulfoxide (DMSO) control,
which can be confirmed by the significantly reduced rate
constant k of arr3 recruitment to the Y4R in the presence of 30
μM of VU0637120. Furthermore, the effect of VU0637120 on
the internalization of the Y4R was investigated (Figure S2,
Supporting Information). Data show that VU0637120 itself
has no effect on the membrane localization of the Y4R as the
amount of quantified cell surface receptors is comparable to
the DMSO control. Stimulation of Y4R with either 10 or 100
nM PP in the presence of DMSO induces internalization of the
Y4R, indicated by a reduced amount of Y4R cell surface
receptors. In contrast, VU0637120 blocks Y4R internalization
induced by 10 nM PP and only slightly decreases Y4R
internalization by 100 nM PP compared to DMSO.
Mutagenesis Studies Identified Y4R Key Residues
Involved in VU0637120 Activity. To characterize the
interaction between VU0637120 and Y4R and to gain insight
into the mechanism of allosteric compounds, we first tested the
activity of VU0637120 at eight Y4R/Y1R chimeras (Figure 3a)
to elucidate Y4R domains, which are relevant for the
interaction with the compound. The activity of VU0637120
at all Y4R/Y1R chimeras is displayed as the reduction of PP
response in the presence of 15 μM antagonist, compared to the
DMSO control (bar graphs, Figure 3b). The results suggest an
important role of transmembrane helix 1 (TM1), TM2, and
extracellular loop 2 (ECL2) for the activity of VU0637120, as
the exchange of these Y4R regions by the corresponding
sequences of the Y1R drastically reduces the antagonistic
activity (Figure 3b).
To further characterize the effect of VU0637120 on the
endogenous ligand PP at the Y4R, radioactive ligand-binding
studies were performed. Competition binding studies revealed
2803
https://dx.doi.org/10.1021/acs.jmedchem.0c02000
J. Med. Chem. 2021, 64, 2801−2814

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
approach was performed previously to investigate the binding
mode of PP (Figure S5, Supporting Information).30 The
resulting docking models were prioritized based on Rosetta
predicted binding energy (Table S4, Supporting Information)
and the proximity of (S)-VU0637120 to the residues that were
shown to be important in mutagenesis experiments (Figures 3
and S3). Initial placement of (S)-VU0637120 inside Y4R was
guided by a significant loss in antagonistic activity of the
compound to the variants Y4(Y1.39A) and Y4(Y1.39F)
(Figures 3c and S3). The best-scoring binding pose (Figures
4 and S6, cluster 1) places the ligand parallel to the membrane
plane inside the transmembrane region. This binding mode is
supported by the sensitivity of VU0637120 to mutation of
F4.60 and A3.29. The loss of VU0637120 function by the
introduction of the bigger, hydrophilic asparagine at position
A3.29 confirms that the space on top of TM3, below the
ECL2, is essential for VU0637120 activity (Figures 3 and S3).
A second round of mutagenesis was conducted to confidently
exclude two alternative binding poses (Figures S6, cluster 2/3
and S7). The data reveal that (S)-VU0637120 docking poses
of cluster 2 and cluster 3 are unlikely, as (S)-VU0637120
activity is insensitive to mutations at positions predicted to be
important for the compound in cluster 2 (D1.27, T2.65, D2.68,
N7.32, and M7.43) or cluster 3 (M2.53, S3.35, V3.36, V5.24,
W6.48, N7.32, and M7.43). The docking model of cluster 1 is
consistent with the mutagenesis data. It shows that (S)-
VU0637120 forms polar contacts with Y1.39, Q3.32, and
H7.39, CH−π interactions with F4.60 and F7.35, and nonpolar
contacts to Y2.64 and L7.36 (Figure 4). Thereby, residues
Y2.64, W2.70, F7.35, and L7.36 form the top surface area of
the binding site. The distances among the hydroxyl group of
Y2.64, the amine group in the indole ring of W2.70, and the
oxygen in the amide group of (S)-VU0637120 are close to the
range of significant polar contacts (Figure 4b). It is possible
that those polar interactions exist, changing the orientation of
the side chain of W2.70 and Y2.64, and destabilize the PP
binding mode.
The model further predicts the role of TM1, TM2, and
ECL2 for Y4R selectivity of VU0637120, which was observed
by the investigation of different Y4R/Y1R chimeras (Figure
3a,b). Mutation of specific fragments and positions identified
regions 1.40−1.43 and Q2.58 to be highly important (Figures
3c and S3). In the model, these positions are in close proximity
and form polar contacts between TM1 and TM2. In addition,
position H7.39 has polar interactions with Y1.39 (Figure 4c,d).
Thus, these residues might be important for the TM1−TM2
and TM1−TM7 orientation to shape the binding pocket of
(S)-VU0637120. The role of the ECL2 was investigated by the
mutation of almost the entire loop sequence (Figure S3,
Supporting Information). However, none of these mutations
had a strong effect on VU0637120 activity. Thus, the loss of
antagonistic activity at the Y4R chimera containing the ECL2
of the Y1R (Figure 3a,b, chimera H) is likely due to
conformational effects, which is in agreement with the high
sequence variation of the loops between Y4R and Y1R (Figure
S8, Supporting Information). The docking model locates (S)-
VU0637120 directly below the conserved disulfide bridge and
suggests a structural role of the ECL2 in the formation of the
binding pocket (Figure 4a).
Figure 2. Investigation of VU0637120 effect ex vivo. (a) (S)-
VU0637120 inhibits optimal rPP (30 nM)-mediated inhibition of ion
transport in a concentration-dependent manner, while 30 μM of (R)-
VU0637120 has no effect on the Y4R agonist in the mucosal
preparations of mouse descending colon. (b) Subsequent PYY signals
(10 nM) are resistant to either enantiomer of VU0637120 at the
different concentrations shown, compared to the DMSO control. Ex
vivo data represent the mean SEM from N = 5 experiments.
Based on these data, we performed an initial screen of 75
variants of Y4R with segmental and single-point mutations
(Figure S3, Supporting Information). For the initial mutant
screen, the racemic mixture of VU0637120 was used, as it is
commercially available and thus less expensive compared to
the enantioselective compounds. Based on these results, the
systematic investigation of the Y4R extracellular domains
identified 11 positions, including Y1.39, Q2.58, T2.61, Y2.64,
W2.70, S3.28, A3.29, Q3.32, F4.60, F7.35, and L7.36 to be
critical for the antagonistic activity of VU0637120 (Figure S3,
Supporting Information). To consider the effect of a mutation
on the functionality of the receptor, all constructs were
characterized for membrane localization (Figure S4, Support-
ing Information) and PP activation (Tables S2 and S3,
Supporting Information).
To more thoroughly characterize the binding pocket of the
active (S)-VU0637120 enantiomer at the Y4R, positions that
have been identified in the initial screen to be important for
VU0637120 were tested with (S)-VU0637120. The antago-
nistic activity of the compound was determined by the
measurement of submaximal receptor activation (PP EC80) in
the presence of increasing (S)-VU0637120 concentrations.
Here, we confirmed the importance of positions Y1.39, Q2.58,
T2.61, Y2.64, W2.70, S3.28, A3.29, Q3.32, F4.60, F7.35, and
L7.36 for (S)-VU0637120 as mutations at these positions lead
to a dramatic loss of antagonistic activity of (S)-VU0637120
(Figure 3c). A summary of the results of the mutagenesis data,
highlighting important positions for VU0637120, is given in
Figure 3d.
Rosetta Docking Identified a Partial Allosteric Bind-
ing Pocket of (S)-VU0637120. To further delineate the
structural determinants of activity, (S)-VU0637120 was
docked using RosettaLigand26−28 into a Y4R homology
model (HM) constructed with RosettaCM.29 A similar
Identification of Y4R Gain-of-Function Variants for
(S)-VU0637120. Based on the docking of (S)-VU0637120 to
the Y4R, we used RosettaDesign to predict Y4R variants that
are favorable for the activity of (S)-VU0637120. The top
2804
https://dx.doi.org/10.1021/acs.jmedchem.0c02000
J. Med. Chem. 2021, 64, 2801−2814

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Figure 3. Investigation of the VU0637120 binding site at the Y4R. (a) Schemes of Y4R/Y1R chimeras, whereby Y4R segments are shown in white
and Y1R segments are colored in black. (b) Importance of TM1, TM2, and ECL2 for VU0637120 activity. Loss of submaximal PP response (PP
EC80) in the presence of 15 μM VU0637120 at Y4R, Y1R, and chimeras investigated in Ca2+ flux assays. Statistical analysis was performed using
one-way analysis of variance (ANOVA) and Dunnett’s post-test: *P < 0.05, **P < 0.01, and ***P < 0.001. (c) Effect of increasing concentrations
of (S)-VU0637120 at important Y4R residues in TM1, TM2, TM3/4, and TM7 in response to a submaximal PP concentration (PP EC80) using
Ca2+ flux assay. Receptor activation was investigated in COS7 cells transiently expressing one specific receptor-eYFP construct and the chimeric G
protein Δ6Gαqi4myr. Data are shown as mean SEM from N ≥ 3 independent experiments. Residues are numbered according to the nomenclature
of Ballesteros−Weinstein.22 (d) Y4R snake plot, residues important for VU0637120 activity are highlighted in green, and residues not important for
VU0637120 are marked in black. Adapted from GPCRdb.org.23
variants predicted by RosettaDesign were inspected visually.
The E6.58R variant was selected because the mutated residue
is predicted to form an additional hydrogen bond interaction
to the methanesulfonamide group of the compound. Addi-
tionally, A3.29V and A3.29P variants were chosen, as both
valine and proline could form better hydrophobic interactions
to the six-member ring of (S)-VU0637120. Those predicted
favorable interactions were confirmed in the docking model of
(S)-VU0637120 to single Y4R variants, which were recon-
structed with RosettaCM29 and RosettaLigand26−28 (Figure
5a).
To check whether we can verify the RosettaDesign
prediction in in vitro assays, we created the Y4R gain-of-
function variants by mutagenesis and tested the effect of (S)-
VU0637120 at these constructs. Here, we demonstrate that the
antagonistic effect of (S)-VU0637120 on PP response is
significantly increased at the variants Y4(E6.58R), Y4(A3.29P),
and Y4(A3.29V) compared to the wild-type Y4R (Figure 5b,c).
These experiments confirm the accuracy of the (S)-
VU0637120 binding mode to the Y4R as we are able to
predict and confirm favorable interactions between distinct
Y4R residues and functional groups of (S)-VU0637120.
Comparison of the Orthosteric Ligand-Binding
Pocket and the Allosteric Binding Site. To map the
binding pocket of the allosteric antagonist VU0637120 in
comparison to the orthosteric agonist PP, mutagenesis
experiments were performed for both ligands (Tables S2 and
S3; Figure S3, Supporting Information). The models of PP
docked to Y4R were reconstructed as we remade Y4R
homology models with a more realistic ECL2 loop and an
updated list of templates (Table S5, Supporting Information),
which include the Y1R crystal structure, discovered in 2018.31
Previous studies and our most recent mutagenesis data
revealed that positions D2.68, W2.70, N6.55, D6.59, N7.32,
F7.35, and M7.43 are important for Y4R activation by PP and
demonstrated that PP binds at the Y4R in a flat binding mode
(Figures 6 and S5).30,32 Additionally, positions T2.61, W2.70,
and F7.35 were identified to be critical for both, the activity of
the orthosteric peptidic agonist PP and the small-molecule
antagonist VU0637120. In contrast, most of the positions that
influence VU0637120 activity, including Y1.39, Q2.58, Y2.64,
S3.28, A3.29, F4.60, and L7.36, are located in a deeper pocket
and fail to affect PP activation (Figure 6). These results
indicate that VU0637120 binds in a secondary, allosteric
binding pocket at the Y4R with a slight overlap with the
binding site of the endogenous ligand PP. Whereas the Y4R-PP
docking model indicates a more open conformation of the Y4R
in complex with PP, the binding of VU0637120 might stabilize
2805
https://dx.doi.org/10.1021/acs.jmedchem.0c02000
J. Med. Chem. 2021, 64, 2801−2814

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Figure 4. Predicted binding mode of (S)-VU0637120. (a) Computational docking of (S)-VU0637120 identified an allosteric binding site located
below the extracellular interface. (b, c) Contacts of (S)-VU0637120 with most of the important residues (side view and top view). (d) Interaction
network among TM1, TM2, TM7, and (S)-VU0637120. Y4R homology models were generated using 10 class A G protein-coupled receptor
(GPCR) X-ray crystal structures (PDB: 5ZBQ, 4S0V, 4ZJC, 5DHG, 4N6H, 4DJH, 4DKL, 5GLH, 3ODU, 5NDD; Table S5, Supporting
Information).
a more closed receptor conformation. Thus, it follows that
VU0637120 binding might create a low-affinity binding state
for the endogeneous ligand PP with reduced agonist binding.
Interestingly, the binding pocket of (S)-VU0637120 at the
Y4R overlaps with that of orthosteric antagonists and NAMs of
other peptide GPCRs (Figure S9, Supporting Information).
The crystal structure of the related Y1R in complex with
orthosteric antagonists revealed that the binding pocket of the
antagonist BMS-193885 extends from the peptide binding site
to the center of the transmembrane domain among TM3,
TM4, TM5, and TM6, and partially overlaps with the binding
site of VU0637120 at its six-member ring and methanesulfo-
namide group.31 In contrast, the binding site of (S)-
VU0637120 stretches horizontally toward TM1, TM2, and
TM7. This region is known as an allosteric pocket in other
rhodopsin-like GPCRs such as C−C chemokine receptor type
5 (CCR5)33 and protease-activated receptor 2 (PAR2).34 The
interaction with these receptor regions might explain the
allosteric character of (S)-VU0637120. Furthermore, positions
in TM1 and TM2 were shown to be critical for the Y4R
selectivity of VU0637120. These findings highlight the
potential to develop highly selective Y4R ligands by targeting
this binding pocket.
The discovery of (S)-VU0637120 defines a novel class of
Y4R antagonists that are smaller and more selective than the
most promising compounds described to date. A small-
molecule compound with weak antagonistic activity at the
Y4R is UR-AK49, but this compound is not Y4R selective and
can also bind to Y1R and Y5R.35 Previous Y4R antagonists were
generated by modifying Y1R antagonists (BIBP3226,36
BIBO330437). However, the most active compounds devel-
oped by this approach still display substantial activity at Y1R
(UR-MEK381) or Y2R (UR MEK388), respectively.14
Recently, a peptide-based Y4R antagonist has been identified
by modifying the structure of a known Y4R hexapeptide
agonist.11 However, peptide ligands that activate peptide
GPCRs often face challenges with dosing and pharmacokinetic
properties such as rapid clearance, poor metabolic stability, and
consequently poor bioavailability, as well as weak membrane
permeability.38 Additionally, it is very challenging to develop
selective orthosteric peptide ligands within multiligand/multi-
receptor systems. Such difficulties have been noted with PP,
the native peptide ligand of the Y4R, which possesses a half-life
in plasma of about 7 min39 and shows a high affinity for Y5R.40
Therefore, allosteric modulators, such as (S)-VU0637120,
provide a promising new class of ligands to target peptide
GPCRs. As these molecules bind to allosteric sites, which are
2806
https://dx.doi.org/10.1021/acs.jmedchem.0c02000
J. Med. Chem. 2021, 64, 2801−2814

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Figure 5. Gain-of-function of (S)-VU0637120 at specific Y4R variants.
(a) Point mutant models made by Rosetta illustrate the improvement
in hydrophobic interactions between (S)-VU0637120 and P3.29 or
V3.29, and the additional hydrogen bond between (S)-VU0637120
and R6.58. Ten class A GPCR templates were used for building the
Y4R homology models (PDB: 5ZBQ, 4S0V, 4ZJC, 5DHG, 4N6H,
4DJH, 4DKL, 5GLH, 3ODU, 5NDD; Table S5, Supporting
Information). (b) Confirmation of P3.29, V3.29, and R6.58 as Y4R
gain-of-function variants in vitro. Shown are concentration−response
curves of (S)-VU0637120, stimulated with a submaximal PP EC80
concentration at specifically identified Y4R residues. (c) Loss of a
submaximal PP concentration in the presence of 15 μM of (S)-
VU0637120 at Y4R gain-of-function variants, data from 5(b). Ca2+
flux assays were performed in COS7 cells transiently expressing one
Figure 6. Comparing the binding pocket of PP and (S)-VU0637120.
(a, b) Superimposition of docking models of PP and (S)-
VU0637120 to Y4R. Residues that are important for the activity of
PP, (S)-VU0637120, and both are colored in green, red, and yellow,
respectively. Group of residues that are represented in the red surface:
(*): Y1.39, E1.42, Q2.58, T2.61, S3.28, A3.29, Q3.32, F4.60, L7.36,
and H7.39. (c) (S)-VU0637120 binds in a secondary, allosteric
binding pocket (red) at the Y4R, but this site might slightly overlap
with the orthosteric binding site of PP (green). Y4R homology models
were generated from the X-ray crystal structure of 10 class A GPCRs
(PDB: 5ZBQ, 4S0V, 4ZJC, 5DHG, 4N6H, 4DJH, 4DKL, 5GLH,
3ODU, 5NDD; Table S5, Supporting Information).
specific Y4R-eYFP variant and the chimeric G protein Δ6Gαqi4myr
.
Data are shown as mean SEM of N ≥ 3 independent experiments.
Comparison of (S)-VU0637120 activity at Y4R and variants was
performed using one-way ANOVA and Dunnett’s post-test: **P <
0.01; ***P < 0.001. Residues are numbered according to the
nomenclature of Ballesteros−Weinstein.22
ligand-binding site, which ultimately strengthens our under-
standing of allosteric modulation of the Y4R.
The proposed binding pose of (S)-VU0637120 highlights
patterns of locations among allosteric pockets across different
GPCR class A subfamilies despite their low sequence identity.
The discovery of small, selective Y4R allosteric antagonists
helps pave the way for validating Y4R as a target for treating
metabolic diseases such as anorexia and obesity by providing
key structural insights that may be extended to other
chemotypes. Looking at a broader picture, this study
demonstrates that targeting this allosteric binding pocket
holds promise for the development of novel, selective
compounds as tools for the investigation of peptide GPCRs,
which represent about 15% of all GPCRs.43
often less conserved compared to orthosteric ligand-binding
sites of receptor subtypes, a higher subtype selectivity can be
achieved.41 Furthermore, allosteric molecules of peptide
GPCRs have the advantage to modulate or fine-tune the
receptor response by the endogeneous peptide and con-
sequently maintain the normal physiological regulation.42 This
study will support the analysis of this relevant class of
modulators and facilitate drug development for peptide
GPCRs in general.
CONCLUSIONS
EXPERIMENTAL SECTION
■
■
In this study, we highlight the characterization of the first-in-
class allosteric antagonist (S)-VU0637120 that exhibits clear
Y4R selectivity in engineered cell lines and in proven native
Y4R-expressing mouse colon mucosa. We obtained a model for
the (S)-VU0637120/Y4R complex that not only agrees with
known mutagenesis data but also predicted three gain-of-
function mutations that increased the activity of (S)-
VU0637120. For the enantioselective (S)-VU0637120 com-
pound, we identified an allosteric binding pocket in the core of
the Y4R transmembrane domains below the endogenous
Cell Culture. COS7 cells were obtained from the American Type
Culture Collection (ATCC). Wild-type COS7 cells were cultured in
Dulbecco’s modified Eagle’s medium (DMEM, Lonza) supplemented
with 10% heat-inactivated fetal bovine serum (FBS, Biochrom). The
COS7 cell lines stably expressing one specific human Y receptor-eYFP
fusion protein (Y1,2,4,5R-eYFP) and the chimeric G protein Δ6Gαqi4myr
were created using pVitro2-hygro-mcs vectors as previously
described44 and cultured in DMEM supplemented with 10% FBS,
133 μg/mL hygromycin (Invivogen), and 1.5 mg/mL G418 sulfate
(Invivogen). HEK293 cells were obtained from Deutsche Sammlung
von Mikroorganismen und Zellkulturen (DSMZ). Wild-type HEK293
2807
https://dx.doi.org/10.1021/acs.jmedchem.0c02000
J. Med. Chem. 2021, 64, 2801−2814

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
cells were cultured in DMEM/Ham’s F12 (1:1 (v/v), Lonza)
supplemented with 15% FBS. Stably transfected HEK293_Y4R-eYFP
cells were generated using a hY4R-eYFP-pVitro2-neo-mcs vector as
described45 and cultured in DMEM/Ham’s F12 (1:1 (v/v), Lonza)
with 15% FBS and 100 μg/mL hygromycin. All cell lines were
cultured in a humidified atmosphere at 37 °C and 5% CO2.
Plasmids. For the Y4R mutant screening, the cDNA of Y4R, Y1R,
Y4R/Y1R chimeras, and Y4R mutants were cloned in a pEYFP_N1
vector (Clontech), C-terminally tagged with an enhanced yellow
fluorescent protein (eYFP), as performed in previous studies.30,32 The
Vanderbilt Discovery Collection, a custom library designed in
collaboration with Vanderbilt, industry-trained medicinal chemists,
Vanderbilt chemoinformaticists, and in collaboration with the vendor,
Life Chemicals. The Vanderbilt Discovery Collection is a diverse,
druglike library, which had been largely depleted of “problem”
compounds (e.g., pan assay interference compounds, PAINS). For the
Ca2+ flux screening, cells were cultured in α minimum essential
medium (MEM) (Corning) with 10% FBS (Invitrogen), 1× glutagro
(Corning), 800 μg/mL G418 sulfate (Cellgro), and 250 μg/mL
zeocin (Life Science) to a confluency of about 90%, then plated in
amine-coated 384-well plates (black, clear bottom, Corning) in 20 μL
cell culture medium/well without antibiotics using a Multidrop
Combi microplate dispenser (ThermoScientific, Thermo Fisher) at a
density of 20 000 cells/well. The cells were incubated for 18 h at 37
°C in the presence of 5% CO2. Following incubation, the medium was
replaced with 20 μL/well fluorescent dye solution, containing 1.0 μM
Fluo-8 AM AAT (Bioquest), 0.01% (v/v) pluronic acid F-127 in assay
buffer (Hanks’ balanced salt solution (HBSS) buffer, Invitrogen, with
20 mM N-(2-hydroxyethyl)piperazine-N′-ethanesulfonic acid
(HEPES), Invitrogen, pH 7.3−7.4) by plate flicking. Following a 60
min incubation at room temperature, fluorescent dye solution was
replaced with 20 μL/well assay buffer as described above. The cell
plates were loaded into Panoptic (Drug Screening System, WaveFront
Biosciences) and imaged at 1 Hz (excitation 480 40 nm, emission
chimeric G protein Δ6Gαqi4myr cDNA was provided by E. Kostenis
46
̈
(Rheinische Friedrich-Wilhelm-Universitat, Bonn, Germany) and
cloned into a pVitro2-mcs-neo vector. In the arr3 recruitment studies,
the Y4R was C-terminally tagged to Renilla luciferase 8 (Rluc8) and
cloned in a pcDNA3 vector. Arr3 was N-terminally fused to venus
fluorescent protein and cloned in a pcDNA3 plasmid.13,20 The
sequence of all constructs was confirmed by DNA sequencing.
Generation of Y4R/Y1R Chimeras and Y4R Mutants. The
chimeric Y1R/Y4R receptors were created using overlap extension
polymerase chain reaction (PCR).47,48 In the first PCR, the N-
terminal cDNA of the donor receptor subtype was amplified using a
sense primer, which annealed before the N-terminal receptor
sequence and contained a specific enzyme restriction site. The
antisense primer annealed at the end of the desired segment and
contained an overhang complementary to the cDNA of the second
receptor subtype. The cDNA of the C-terminal fragment was obtained
in a second PCR using the cDNA from the second receptor subtype.
Here, the sense primer annealed at the beginning of the desired
segment and contained a 5′ overhang sequence complementary to the
cDNA of the N-terminal segment of the donor receptor, which was
created in the first PCR. The antisense primer of the second PCR
contained a restriction site sequence and annealed after the receptor
sequence. The fragments obtained in both PCRs were purified using
agarose gel electrophoresis and Promega Wizard SV Gel and PCR
Clean-Up System (Promega). In a third PCR, the receptor fragments
were combined using the N-terminal and C-terminal fragments with
sense and antisense primers annealing at the beginning or end of the
respective fragments to obtain the chimeric cDNA. Afterwards, the
PCR product was digested using BgIII and XbaI or SalI and NotI
enzymes (ThermoFischer). The desired products were purified by gel
electrophoresis and ligated into a pEYFP_N1 vector using T4 DNA
ligase (ThermoFischer). The mutation of single amino acids or 2−4
adjacent residues in the Y4R was performed by Quick change site-
directed mutagenesis using PCR and appropriate primers. In the PCR,
the Phusion (ThermoFischer) or Pfu Turbo DNA polymerase
(Agilent) was used with 5× HF (Phusion) or 10× reaction (Pfu)
buffer, 100 ng Y4R-eYFP_N1 template, 0.25−0.5 μM sense/antisense
primers, 200 μM deoxyribonucleotide (dNTP) mix, and 1−10%
DMSO. The PCR conditions were adjusted according to the
manufacturer’s protocol of the DNA polymerase. Afterwards, the
parental DNA was digested using DpnI (ThermoFischer). The PCR
products were transformed into chemically competent Escherichia coli
DH5α using heat shock (60 s, 42 °C) and single clones were
separated on Luria broth (LB) medium-agarose plates containing
kanamycin (30 μg/mL, Sigma-Aldrich). The plasmid DNA was
prepared using Promega Wizard SV Gel and PCR Clean-Up System
or Promega Pureyield Midiprep System (Promega) according to the
manufacturer’s protocol. Finally, the sequence of all Y4R/Y1R
chimeras and Y4R mutants was confirmed by DNA sequencing.
High-Throughput Screening Ca2+ Flux Assay. Ca2+ flux
screening assays to identify Y4R allosteric modulators were performed
in a 384-well format using Panoptic (WaveFront Biosciences). The
compound screening was performed with stably transfected HEK293
cells coexpressing human Y4R and Gqi5 α subunit protein (Addgene
plasmid #24501). In a following HTS, 77 021 compounds were tested
for Y4R modulatory effects and 579 primary hits were detected. Forty-
one compounds from the primary hits were reordered as dry
compounds and re-evaluated with different concentrations, also for
nonspecific effects in wild-type HEK293 cells. Seventeen compounds
were confirmed. Seventy-seven thousand compounds were from the
538
40 nm using a three-addition protocol designed to detect
agonists, potentiators, and inhibitors). After collecting 10 s of
baseline, 20 μL/well of 20 μM test compounds in assay buffer were
added to make the final concentration of each compound to be 10
μM. Following a 240 s delay, 10 μL/well of concentration of 5-fold
over the PP EC30 (∼2 nM) in assay buffer containing 0.0125% bovine
serum albumin (BSA) (Sigma-Aldrich) was added. After 240 s, a 12
μL/well addition 5-fold over the PP EC80 (∼20 nM) in assay buffer
containing 0.0125% BSA was performed, imaged for another 240 s.
Ca2+ Flux Assay. The compound characterization was performed
with stably transfected COS7 cells expressing one specific Y receptor
subtype (Y1,2,4,5R-eYFP) and the chimeric G protein Δ6Gαqi4myr as
previously described.13 To investigate Y4R/Y1R chimera and Y4R
variants, COS7 cells were transiently transfected with one specific
receptor construct (C-terminally tagged with eYFP) and the chimeric
G protein Δ6Gαqi4myr using Metafectene Pro transfection reagent.
Therefore, COS7 cells were seeded in 25 cm2 culture flasks and grown
to a confluency of 80%. Each flask was transfected with a total of 4 μg
of plasmid DNA (3 μg of receptor-eYFP and 1 μg of Δ6Gαqi4myr
)
using 15 μL of Metafectene Pro according to manufacturer’s protocol
over night at 37 °C. On the next day, 20 000 cells per well were
reseeded in black, clear bottom 96-well plates (Greiner) and
incubated over night at 37 °C. For the assay, the cell culture medium
was aspirated, and cells were incubated in 100 μL/well assay buffer
(HBSS (Lonza), 20 mM HEPES (Sigma), 2.5 mM probenecid
(Sigma), pH 7.4) containing 2.4 μM Fluo2-AM (Abcam) calcium
dye. Afterwards, the dye solution was replaced by assay buffer and
Ca2+ signals were detected (excitation 485 nm, emission 525 nm) in a
FlexStation III device (Molecular Devices). The compound
VU0637120 and agonists (PP/NPY) were added automatically in a
two-addition protocol. After 20 s baseline detection, DMSO control
or VU0637120 was added to the cells, followed by the addition of PP
or NPY after 110 s in a total run time of 140 s. The Ca2+ signal
responses were analyzed as x-fold over basal values and normalized to
the maximum activation obtained in the presence of 1 μM PP (Y4R,
Y4R/Y1R chimeras and Y4R variants) or NPY (Y1R, Y2R, Y5R).
Y4R Membrane Preparations and Radioligand Binding
Studies. Y4R membranes were generated from stably transfected
HEK293_Y4R-eYFP cells, as previously described.13 Cells were
suspended in Dulbecco’s phosphate-buffered saline (DPBS) and
centrifuged at 1800 rpm, 4 °C for 5 min. The cell pellet was
resuspended in Tris buffer (50 mM Tris (Sigma), 50 μM Pefabloc
(Sigma), pH 7.5) and homogenized in a potter grinder (Potter SB
Braun). The cell suspension was centrifuged at 2400 rpm at 4 °C for
20 min; the supernatant was removed and centrifuged for 60 min at
12 000 rpm, 4 °C. The supernatant was discarded and the membrane
2808
https://dx.doi.org/10.1021/acs.jmedchem.0c02000
J. Med. Chem. 2021, 64, 2801−2814

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
pellet was resuspended in HEPES buffer (25 mM HEPES, 25 mM
CaCl2 (Fluka), 1 mM MgCl2 (Fluka), 50 μM Pefabloc (Sigma), pH
7.4). After the homogenization in the potter grinder, the suspension
was centrifuged at 12 000 rpm at 4 °C for 60 min. The resulting pellet
was resuspended in HEPES buffer and the protein concentration was
determined using Bradford protein assay.49 Membrane preparations
were frozen in liquid nitrogen and stored at −20 °C. For binding
experiments, 0.3 μg of Y4R membrane preparation was suspended in
HEPES buffer containing either DMSO as control or VU0637120. PP
and 125I-PP (human) solutions were prepared in aqua dest.
supplemented with 0.1% BSA (PAA Laboratories). Nonspecific
binding was determined in the presence of 1 μM PP. Radioligand
solutions (60 or 180 pM 125I-PP) were added to all samples and
incubated for 5 h (at room temperature, 200 rpm). Afterwards,
membrane-bound 125I-PP was separated by filtration using a GFC
filter (PerkinElmer), presoaked with 0.1% polyethylenimine (Sigma)
in a MicroBeta 96-well filtermate harvester (PerkinElmer). Mem-
branes were washed with cold DPBS, dried for 15 min at 55 °C, and
treated with MeltiLex scintillation sheets. For the quantification of
radioactivity, membranes were measured in a MicroBeta scintillation
counter (PerkinElmer).
IP-One Assay. The IP-One assay was performed in COS7 cells
stably expressing the Y4R-eYFP and the chimeric G protein
Δ6Gαqi4myr. Therefore, cells were seeded into white 384-well plates
(Greiner) at a density of 8000 cells/well and incubated over night
under standard conditions. For the detection of IP production, the IP-
One Gq assay kit (Cisbio) was used. On the following day, media was
removed, (S)-VU0637120 or DMSO was added followed by the
subsequent stimulation with peptide solution in HBSS + 20 mM LiCl
for 1 h at 37 °C, 5% CO2. Afterwards, cells were lysed with lysis buffer
containing antibody 1 and antibody 2 according to the manufacturer’s
protocol. Signals were detected using a Tecan Spark microplate reader
(Tecan Group AG) measuring emission at 665 and 620 nm.
preparations were dissected from each colon. Mucosae were placed
individually between Ussing chamber halves and voltage-clamped at 0
mV (using a DVC1000; WPI, Sarasota, FL), as described previously.
Mucosae with exposed areas of 0.14 cm2 were bathed in oxygenated
Krebs−Henseleit solution on both sides, at 37 °C, and were voltage-
clamped at 0 mV. A stable basal Isc was reached within 20 min after
which DMSO (0.3%) or antagonist additions (300 nM to 30 μM of
the active enantiomer (S)-VU637120 or 30 μM of inactive (R)-
VU637120) were made to the basolateral reservoir. After an
intermediary addition of the secretagogue VIP (10 nM, for 10−15
min), an optimal single rPP concentration (30 nM)25 was added to
activate epithelial Y4R and reduce Isc levels, followed 15 min later by
PYY (10 nM). Changes in Isc to each drug addition (all to the
basolateral reservoir) were pooled and presented as mean
SEM.
The pIC50 and statistical analysis were performed using GraphPad
Prism software (version 7).
Homology Modeling of Y4R. Sequences of Y4R and the
templates were aligned using the Clustal Omega web server.50
Multiple sequence alignment was then adjusted to ensure that the
helix regions and reserved residues remain aligned, and to remove
gaps within transmembrane α-helices. Y4R homology models were
built using RosettaCM protocol29 of the Rosetta3.9 software suites.51
Addition constraints were set up to account for the disulfide bond
between Cys1.25 and Cys7.38, as well as between Cys3.25 and
Cys5.25. The output Y4R homology models were evaluated by
Rosetta total energy score, which includes knowledge-based energy
terms such as hydrogen bonds, electrostatic interactions, and van der
Waals packing.29 Top 10% of output models, which scored most
favorably by Rosetta, were clustered based on the RMSD cutoff of 2 Å
using BCL::Cluster.52 The clustering procedure was extensively
discussed in the precedent protocol paper.27 Finally, a Y4R model
with the best Rosetta total energy score was chosen from each of the
10 largest clusters.
Arrestin3 Recruitment Assay. Arrestin3 recruitment to the Y4R
was measured using BRET assay in HEK293 cells transiently
transfected with Y4R-Rluc8 and venus-arr3 fusion proteins as
previously described.13 For kinetic measurements, signals were
measured continuously over 30 min after PP (100 nM) addition.
BRET signals were observed as the ratio of fluorescence (venus-arr3)
and luminescence (Y4R-Rluc8). NetBRET signals were calculated by
subtraction of unstimulated controls.
Docking NAM to Y4R. As the experimental data suggested that
the ligand could interact with residues located on TM1, TM2, and
TM7, it is crucial to create enough space among those three
transmembrane regions in the homology models before ligand
docking. The output Y4R homology models from the first round of
RosettaCM did not yield enough space for VU0637120 to approach
and form hydrogen bonds to the residue Y1.39. Hence, the Y4R
models were rehybridized with the ligand placed inside the area
among important residues. To create a set of templates for the second
rounds of RosettaCM, 65 different poses of the ligand were placed
inside Y4R such that it interacted to TM1, TM2, TM7, and in no
more than 4 Å from Y1.39. This way, the interactions between the
ligand and the important residues according to the mutagenesis data
would be weighed in the placement of helices in MC trajectory and
optimization. The resulting models were then filtered for interactions
to important residues and hydrogen bond to Y1.39 and clustered. The
final 10 cluster representative models were selected to be templates
for the docking protocol based on the size of the corresponding
clusters as well as the openness of the binding pocket suggested by the
mutagenesis data.
A set of 106 conformations of VU0637120 was generated using the
ConformerGenerator application of BCL v3.2.2.53 RosettaLigand was
used to dock VU0637120 to Y4R homology models. The docking
protocol is described in a previous publication.27 In addition to
Rosetta total energy score, the interface_delta score, which is relative
to the predicted binding energy between VU0637120 and Y4R, was
also extracted for all 20 000 output docking models. We selected
models with a total Rosetta score of at least 95% of the best total
Rosetta score attained and the top 10% interface delta score. We
applied an experimental filter on the selected docking models. The
filter makes sure that the selected ligand poses that contact 9 out of
the 11 residues were shown to be important to the antagonistic
activity by mutagenesis data. A residue is determined to contact the
ligand if at least one atom of the residue is at most 5 Å away from any
atom of the ligand. Those filtered models were then clustered based
on the ligand position with 3 Å RMSD cutoff. For each of the five
largest clusters, the interaction contact and strength were computed
Fluorescence Microscopy. The internalization of the Y4R was
investigated using fluorescence microscopy in HEK293 cells stably
expressing the Y4R-eYFP fusion protein. Cells were seeded in μ-slide
eight well (IBIDI) chambers at a density of 180 000 cells/well and
incubated over night at 37 °C. For microscopy, the cell culture
medium was aspirated and nuclei of the cells were stained with 300
μL/well OptiMEM (serum-reduced media, Gibco) containing
Hoechst33342 (Sigma) for 30 min at 37 °C. Afterwards, the solution
was aspirated and 150 μL/well OptiMEM containing 30 μM
VU0637120 or DMSO control were added. Typically, 50 μL/well
PP solution (10 or 100 nM PP final concentration) or OptiMEM
(unstimulated samples) was added, and the cells were stimulated for
30 min at 37 °C. The stimulation solution was replaced by 200 μL of
OptiMEM. Microscopy was performed using the AxioVert Observer
Z1 (ZEISS) fluorescence microscope (YFP: filter set 46, 4′,6-
diamidino-2-phenylindole (DAPI): filter set 49, ApoTome, 63×/1.40
oil objective). The images were processed using ZEN2012 (ZEISS)
and Axiovision SE64 (ZEISS) software. The detection of the eYFP
fluorescence was performed with a constant exposure time (400 ms)
to allow the quantification of cell surface receptors as performed
previously.13 The cell surface fluorescence was quantified as pixel
intensity in black/white pictures using ImageJ software and
normalized to the unstimulated DMSO control.
Electrophysiological Measurements. All of the animal experi-
ments performed in this manuscript were conducted in compliance
with the local ethics committee and is covered by the national (U.K.)
Home Office Project license (P6EA0199C) held at KCL. Vectorial
ion transport was measured across mucosal preparations of mouse
descending colon as described previously.13,25 Up to six adjacent
2809
https://dx.doi.org/10.1021/acs.jmedchem.0c02000
J. Med. Chem. 2021, 64, 2801−2814

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
for each cluster such that the scores are higher when an important
residue has high contact frequency to the ligand or favorable binding
energy toward the ligand. The interaction contact score is computed
as
Compound Synthesis and Characterization. The compound
VU0637120 (product ID: F5141-0188) was purchased from Life
Chemicals Inc. (Ukraine). Purity was determined by reversed-phase
high-performance liquid chromatography (RP-HPLC) employing two
different columns (Figure S10, Supporting Information). Purity of
racemic VU0637120 is >95%. The enantiomers (R)- and (S)-
VU0637120 were synthesized and purified individually. 1H NMR
spectra and RP-HPLC were employed to determine purity of the
enantioselective compounds. (R)- and (S)-VU0637120 have a purity
of >95% (Figures S10 and S11, Supporting Information).
Enantiomeric excess (ee) was determined by chiral supercritical
fluid chromatography (SFC) separation performed on a Thar
(Waters) Investigator at 220 nm UV (Figure S12, Supporting
Information).
interaction contact (t) =
ct × 4.7 − ct
∑
i
n
where cti and ctn are the percents of docking poses that have the ligand
contact with residues that have been identified as important and
residues that were not confirmed as critical for binding residue,
respectively. The weight of 4.7 was chosen to balance the impact of 13
critical with 61 noncritical residues. Similarly, the interaction strength
of a cluster t is calculated as
interaction strength (t) =
ddg × (−4.7) + ddg
i n
∑
Synthesis and Purification. Air sensitive reactions were carried
out under a nitrogen atmosphere (Airgas Catalog no. NI UHP300).
The following solvents were employed for chemical reactions:
dichloromethane (99.9%, extra dry, AcroSeal, Acros Organics Catalog
no. 610300010) and N,N-dimethylformamide (anhydrous, 99.8%,
packaged under argon in resealable ChemSeal bottles, Alfa Aesar
Catalog no. 43997). The following solvents were employed for
working up reactions and/or extractions: ethyl acetate (Certified ACS
grade, Fisher Chemical Catalog no. E145-20), ethyl ether (anhydrous,
BHT stabilized, Certified ACS, Fisher Chemical Catalog no. E138-4),
and dichloromethane (not stabilized, HPLC grade, Fisher Chemical
Catalog no. D150-4). Brine was prepared from deionized water and
sodium chloride (Reagent grade, Fisher Chemical Catalog no.
S25541B). Anhydrous sodium sulfate (Lab grade, Fisher Chemical
Catalog no. S25568A) was employed for drying organic extracts.
Thin-layer chromatography (TLC) was conducted on glass plates
coated with silica gel 60 F254 from Millipore Sigma (Catalog no.
1057150001). Normal phase flash chromatography was carried out on
either a CombiFlash EZ Prep or CombiFlash Rf+ automated flash
chromatography system, both from Teledyne ISCO. Normal phase
flash chromatography was carried out using RediSep Rf normal phase
disposable flash columns (40−60 μm) from Teledyne ISCO (Catalog
nos. 69-2203-304, 69-2203-312, 69-2203-324, 69-2203-340, 69-2203-
380, and 69-2203-320). The following solvents were employed for
TLC and normal phase chromatography: hexanes (Certified ACS
grade, Fisher Chemical Catalog no. H292-20), ethyl acetate (Certified
ACS grade, Fisher Chemical Catalog no. E145-20), dichloromethane
(not stabilized, HPLC grade, Fisher Chemical Catalog no. D150-4),
and methanol (HPLC grade, Fisher Chemical, Catalog no. A452-4).
Reverse-phase preparative HPLC was carried out on a CombiFlash
EZ Prep automated flash chromatography system equipped with a
RediSep Prep C18 10 × 250 mm2, 100 Å, 5 μm HPLC preparative
column from Teledyne ISCO (Catalog no. 692203809). The
following solvents were employed for reversed-phase chromatog-
raphy: acetonitrile (HPLC grade, Fisher Chemical Catalog no.
A998SK-4) and water purified using a Milli-Q Advantage A10
Water Purification System from Millipore Sigma.
where ddgi and ddgn are the predicted binding energies between the
ligand and an important residue and noncritical residue, respectively.
In the cluster with the most favorable interaction contact and strength
scores, a model that best explains the mutagenesis data will be chosen
as a putative binding pose of VU0637120 and Y4R. Per residue
contact and interaction strength are listed in Tables S4 and S6,
Supporting Information.
Predicting Gain-of-Function Y4R Variants. The residues that
interact with VU0637120 were redesigned with the Rosetta binding
pocket design protocol.54 Predicted variants that are located on TM3,
TM4, and TM5 are selected only if they specifically interact with the
methylsulfonyl tail and the six-member ring head of the pose of
VU0637120 cluster 1. The single-mutant models of the predicted
gain-of-function variants are then rebuilt with RosettaCM to verify
favorable selective interactions of new residues to VU0637120 cluster
1.
Docking PP to Y4R. PP was docked to Y4R homology models
(built in step 10) through four rounds of RosettaCM and
FlexPepDock with the gradually increasing length of PP (last 5, last
8, and last 23 residues of PP). We removed the ECL2 from the Y4R
homology models (HM) before docking 5mer PP and added the
ECL2 back in after docking 23mer PP. A similar protocol can be
found in a previous publication.31 The docking was guided by
experimental restraints from mutagenesis studies.30 Details of the
restraints are described in Table S7, Supporting Information.
Data Analysis. For data analysis, GraphPad Prism 5.0 (or version
7.0, for electrophysiology) and python matplotlib packages were used.
EC50 and Emax values of concentration−response curves were obtained
by nonlinear regression analysis (curve fit). The quantification of the
effect of VU0637120 in Ca2+ flux assays was performed using an
operational model of allosterism (eq 1).17 The parameters of the
model describe: Emmaximum system response; nrelated to the
slope of response curves; [A] and [B]concentration of agonist (A)
and modulator (B); τA and τBintrinsic agonism (efficacy) of the
agonist (A) and the compound (B); α and βeffect of the modulator
on the affinity (α) or the efficacy (β) of the agonist A. In the
calculations, the KA of the agonist PP (log KA −10.2) was constrained
to values determined in previous binding assays.13
Compound Characterization. All NMR spectra were recorded
on a 300 MHz Bruker Fourier 300HD NMR spectrometer equipped
1
with a dual H and 13C probe with Z-Gradient and automatic tuning
E =
and matching, full computer control of all shims with TopShim, 24-
sample SampleCase automation system, and TopSpin software. All
NMR samples were prepared with either methyl sulfoxide-d6 with
0.03% tetramethylsilane (TMS), 99.8 atom % D, Acros Organics
Catalog no. 360000100 or chloroform-d with 0.03% TMS, 99.8+ atom
Em(τA[A](KB + αβ[B]) + τB[B]KB)n
([A]KB + KAKB + KA[B] + α[A][B])n + (τA[A](KB + αβ[B]) + τB[B]KA)n
(1)
The arr3 recruitment rates k were calculated from kinetic traces
1
% D, Acros Organics Catalog no. 209561000. H and 13C chemical
using the one-phase association equation, as performed previously.13
shifts are reported in δ values in ppm downfield with tetramethylsilane
(TMS) as the internal standard. Data are reported as follows:
chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q =
quartet, b = broad, m = multiplet), integration, coupling constant
(Hz). High-resolution mass spectrometry was conducted on an
Agilent 6230 accurate-mass time-of-flight (TOF) LC/MS with the
ESI source equipped with MassHunter Walkup software. MS
parameters were as follows: fragmentor: 175 V, capillary voltage:
3500 V, nebulizer pressure: 35 psig, drying gas flow: 11 L/min, drying
gas temperature: 325 °C. Samples were introduced via an Agilent
(−k×x)
Y = Y0 + (Em − Y0) × (1 − e
)
(2)
Schild analysis for allosteric antagonist was performed using eq 3,
whereas [B] describes the concentration of the allosteric antagonist,
EC50 refers to the EC50 of the DMSO control, and EC50′ to the EC50
in the presence of (S)-VU0637120.55
Ä
Å
Å
Å
Å
Å
Å
Å
Å
Å
É
Ñ
Ñ
Ñ
Ñ
Ñ
Ñ
Ñ
Ñ
Ñ
i
j
j
j
j
j
y
z
z
EC ′
EC50
[B](1 − α)
50
z
Log
− 1 =Log
z
z
α[B] + K
(3)
B
Å
Ç
Ñ
Ö
k
{
2810
https://dx.doi.org/10.1021/acs.jmedchem.0c02000
J. Med. Chem. 2021, 64, 2801−2814

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
1260 Infinity UHPLC comprised of a G4225A HiP Degasser, G1312B
binary pump, G1367E ALS, G1316A TCC, and G1315C DAD VL+
with a 5 μL semimicro flow cell with a 6 mm path length. UV
absorption was observed at 220 and 254 nm with a 4 nm bandwidth.
Column: Agilent Zorbax SB-C18, Rapid Resolution HT, and 1.8 μm,
2.1 × 50 mm2. Gradient conditions: hold at 5% CH3CN in H2O
(0.1% formic acid) for 1.0 min, 5−95% CH3CN in H2O (0.1% formic
acid) over 5 min, hold at 95% CH3CN in H2O (0.1% formic acid) for
1.0 min, and 0.5 mL/min. Chiral SFC separation was performed on a
Thar (Waters) Investigator at 220 nm (UV). Column: Chiral
Technologies CHIRALPAK IB, 4.6 × 250 mm2, and 5 μm. Isocratic
conditions: 50% MeOH (0.1% diethanolamine (DEA)) in CO2 over 5
min. Flow rate: 3.5 mL/min. Column temperature: 40 °C. System
backpressure: 100 bar. All final analogs were at least 95% pure
according to these analytical methods.
NMR (75 MHz, DMSO-d6) δ 168.73, 156.70, 143.67, 143.08, 142.04,
124.55, 109.28, 108.37, 64.48, 57.44, 43.80, 27.18, 22.03, and 21.54.
HRMS, calcd for C15H18N3O3S+ [M + H], 320.1063; found 320.1067.
(S)-N-(6,7-Dihydro-[1,4]dioxino[2′,3′:4,5]benzo[1,2-d]thiazol-2-
yl)-1-(methylsulfonyl)-piperidine-2-carboxamide ((S)-VU0637120).
Triethylamine (128 μL, 0.936 mmol) was added to a solution of
compound 4 (Scheme 1) in dichloromethane (2 mL) at 0 °C in an ice
bath and stirred for 5 min. Methanesulfonyl chloride (40.0 μL, 0.376
mmol) was added dropwise, and the reaction was warmed to room
temperature and stirred for an additional 1 h. After completion of the
reaction, water (15 mL) was added to the reaction mixture, and it was
extracted with dichloromethane (2×). The crude product was purified
by flash chromatography using ethyl acetate and hexane as eluent, and
the title compound was obtained as an off-white solid (75 mg, 60%).
1H NMR (300 MHz, DMSO-d6): δ 12.39 (br s, 1H), 7.45 (s, 1H),
(S)-VU0637120 and (R)-VU0637120. Starting material 6,7-
dihydro-[1,4]dioxino[2′,3′:4,5]benzo[1,2-d]thiazol-2-amine 1
(Scheme 1) is available commercially from several vendors; however,
it was prepared according to the literature16 utilizing the following
method. N-Bromosuccinimide (1.30 g, 7.30 mmol) was added in
portions to a suspension of NH4SCN (1.10 g, 14.5 mmol) in AcOH
(32 mL) at 10 °C. The reaction mixture was stirred at the same
temperature for 30 min. After filtration to remove the insoluble
materials, the filtrate was added to a solution of 3,4-ethylenedioxyani-
line (Chem-Impex Catalog no. 27084) (1.00 g, 6.62 mmol) in AcOH
(32 mL), and the reaction mixture was stirred at room temperature
for 20 h. After removal of the solvent in vacuo, the residue was diluted
with ethyl acetate (75 mL), washed with water and brine, and
concentrated in vacuo. The crude product obtained was purified by
flash chromatography to afford 1 (Scheme 1) as a yellow solid (600
mg, 44%). 1H NMR (300 MHz, DMSO-d6) δ 9.38 (bs, 2H), 7.41 (s,
1H), 6.96 (s, 1H), and 4.28−4.23 (m, 4H). 13C NMR (75 MHz,
CD3OD) δ 169.79, 146.07, 142.82, 139.86, 122.71, 108.16, 105.57,
64.23, and 64.12. High-resolution mass spectrometry (HRMS), calcd
for C9H9N2O2S+ [M + H], 209.0379; found 209.0378.
7.21 (s, 1H), 4.72 (d, J = 4.6 Hz, 1H), 4.27 (s, 4H), 3.67−3.47 (m,
2H), 2.93 (s, 3H), and 1.89−1.34 (m, 6H). 13C NMR (75 MHz,
CDCl3) δ 169.13, 156.52, 143.55, 142.80, 142.06, 124.86, 108.62,
108.37, 64.31, 64.26, 55.95, 43.34, 39.66, 26.35, 24.29, and 19.73.
+
HRMS, calcd for C16H20N3O5S2 [M + H], 398.0839; found
25
398.0848. [α]D −66° (c 0.098, CHCl3). 99.6% ee (RT = 3.74).
tert-Butyl-(S)-2-((6,7-dihydro-[1,4]dioxino[2′,3′:4,5]benzo[1,2-d]-
thiazol-2-yl)carbamoyl)piperidine-1-carboxylate (5). Intermediate 6
(Scheme 1) was synthesized analogous to intermediate 3 (Scheme 1)
using N-Boc-homoproline (Chem-Impex Catalog no. 05499). 1H
NMR (300 MHz, CDCl3) δ 9.51 (br s, 1H), 7.27 (s, 2H), 5.02 (br s,
1H), 4.30 (s, 4H), 4.06 (m, 1H), 2.89−2.75 (m, 1H), 2.39 (m, 1H),
1.77−1.36 (m, 5H), and 1.51 (s, 9H). 13C NMR (75 MHz, CDCl3) δ
169.80, 156.53, 143.43, 143.07, 141.94, 140.90, 125.07, 108.57,
108.50, 81.52, 64.31, 60.42, 28.36, 24.69, 21.08, 20.57, and 14.22.
HRMS, C20H26N3O5S+ [M + H], 420.1588; found 420.1607.
(S)-N-(6,7-dihydro-[1,4]dioxino[2′,3′:4,5]benzo[1,2-d]thiazol-2-
yl)piperidine-2-carboxamide (7). Intermediate 7 (Scheme 1) was
1
synthesized analogous to intermediate 4 (Scheme 1). H NMR (300
tert-Butyl-(S)-2-((6,7-dihydro-[1,4]dioxino[2′,3′:4,5]benzo[1,2-d]-
thiazol-2-yl)carbamoyl)piperidine-1-carboxylate (4). N,N-Diisopro-
pylethylamine (DIEA) (331 μL, 1.92 mmol) was added to a mixture
of Boc-L-homoproline (Chem-Impex Catalog no. 05135) (242 mg,
1.06 mmol), 1-[bis(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-
b]pyridinium 3-oxid hexafluorophosphate (HATU) (734 mg, 1.92
mmol) in DMF (4 mL) and stirred for 20 min. A solution of 1 (200
mg, 960 mmol) in DMF (2 mL) was added dropwise, and the
solution was allowed to stir for 12 h. After completion of the reaction
as judged by TLC, 15 mL of water was added. The mixture was
extracted with ethyl acetate (2×), washed with brine, and dried over
anhydrous sodium sulfate. The crude product was purified by flash
chromatography using ethyl acetate and hexane as eluent, and the title
MHz, DMSO-d6) δ 9.83 (d, J = 9.8 Hz, 1H), 9.13 (m, 1H), 7.51 (s,
1H), 7.26 (s, 1H), 5.26 (br s, 1H), 4.29 (s, 4H), 4.13 (m, 1H), 3.30
(d, J = 12.0 Hz, 1H), 3.07 (m, 1H), 2.96 (m, 1H), 2.28 (d, J = 11.4
Hz, 1H), and 1.88−1.47 (m, 5H). 13C NMR (75 MHz, DMSO-d6) δ
168.73, 156.71, 143.72, 143.07, 142.05, 124.55, 109.30, 108.38, 64.49,
57.47, 43.81, 27.18, 22.04, and 21.56. HRMS, calcd for C15H18N3O3S+
[M + H], 320.1063; found 320.1075.
(R)-N-(6,7-Dihydro-[1,4]dioxino[2′,3′:4,5]benzo[1,2-d]thiazol-2-
yl)-1-(methylsulfonyl)-piperidine-2-carboxamide ((R)-VU0637120).
1
(R)-VU0637120 was synthesized analogous to (S)-VU0637120. H
NMR (300 MHz, DMSO-d6): δ 12.39 (br s, 1H), 7.45 (s, 1H), 7.21
(s, 1H), 4.72 (d, J = 4.2 Hz, 1H), 4.28 (s, 4H), 3.62−3.47 (m, 2H),
2.93 (s, 3H), and 1.89−1.19 (m, 6H). 13C NMR (75 MHz, CDCl3) δ
168.88, 156.36, 143.57, 142.91, 142.09, 124.95, 108.70, 108.40, 64.34,
64.29, 56.03, 43.43, 39.86, 26.06, 24.26, and 19.82. HRMS, calcd for
1
compound was obtained as an off-white solid (346 mg, 86%). H
NMR (300 MHz, CDCl3) δ 9.50 (br s, 1H), 7.26 (s, 2H), 5.02 (br s,
1H), 4.30 (s, 4H), 4.06 (m, 1H), 2.91−2.75 (m, 1H), 2.37 (m, 1H),
1.78−1.36 (m, 5H), and 1.51 (s, 9H). 13C NMR (75 MHz, CDCl3) δ
171.18, 169.79, 156.48, 143.42, 143.14, 141.92, 125.10, 108.59,
108.49, 81.52, 64.33, 60.41, 28.36, 24.97, 21.08, 20.56, and 14.22.
HRMS, calcd for C20H26N3O5S+ [M + H], 420.1588; found 420.1600.
(S)-N-(6,7-Dihydro-[1,4]dioxino[2′,3′:4,5]benzo[1,2-d]thiazol-2-
yl)piperidine-2-carboxamide (6). Compound 3 (340 mg, 0.811
mmol, Scheme 1) was dissolved in dichloromethane (6 mL) and
cooled to 0 °C in an ice bath. Hydrogen chloride solution (4 mL, 4.0
M in dioxane, 16 mmol) was added dropwise, and the reaction
mixture was stirred at room temperature for 2 h. After completion of
the reaction as judged by TLC, the majority of the solvent was
removed in vacuo, and the product was precipitated by the addition of
diethyl ether. The crude product was filtered, washed with diethyl
ether (2×), dried, and purified by preparative HPLC using acetonitrile
and water as eluent (20% acetonitrile in water). The title compound
was obtained as off-white solid (230 mg, 89%). 1H NMR (300 MHz,
DMSO-d6) δ 9.90 (d, J = 10.0 Hz, 1H), 9.13 (m, 1H), 7.51 (s, 1H),
7.26 (s, 1H), 4.29 (s, 4H), 4.15 (m, 1H), 3.30 (d, J = 11.8 Hz, 1H),
2.96 (m, 1H), 2.29 (d, J = 11.4 Hz, 1H), 1.93−1.41 (m, 5H). 13C
+
25
C16H20N3O5S2 [M + H], 398.0839; found 398.0849. [α]D +73° (c
0.098, CHCl3). >99.9% ee (RT = 2.84).
Peptide Synthesis. The peptides human PP (APLEPVYP
GDNATPEQMAQYAADLRRYINMLTRPRY-NH2) and porcine
NPY (YPSKPDNPGEDAPAEDLARYYSALRHYINLITRQRY-NH2)
were synthesized by automated solid-phase peptide synthesis in an
automated multiple peptide synthesis robot system (Syro, Multi-
SynTech) using the 9-fluorenylmethyloxycarbonyl (Fmoc) strategy as
described previously.43 Peptides were purified by preparative RP-
HPLC on a Aeris PEPTIDE XB-C18 column (100 Å, 5 μm, 250 × 2.1
mm2, Phenomenex). Purity of the peptides (≥95%, Figures S13 and
S14, Supporting Information) was determined by two different
analytical RP-HPLC systems using gradients of 20−70% eluent B
(0.08% TFA in acetonitrile (ACN)) in eluent A (0.1% trifluoroacetic
acid (TFA) in H2O) in 40 min. Identity was confirmed by matrix-
assisted laser desorption/ionization time-of-flight (MALDI-TOF)
mass spectrometry (Ultraflex III MALDI-TOF/TOF, Bruker
Daltonics) (Figures S13 and S14, Supporting Information).
2811
https://dx.doi.org/10.1021/acs.jmedchem.0c02000
J. Med. Chem. 2021, 64, 2801−2814

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
University, Nashville, Tennessee 37235, United States;
orcid.org/0000-0002-6886-1195
Kyle A. Emmitte − Department of Pharmaceutical Sciences,
UNT System College of Pharmacy, University of North Texas
Health Science Center, Fort Worth, Texas 76107, United
States; orcid.org/0000-0002-6643-3947
Helen M. Cox − King’s College London, Wolfson Centre for
Age-Related Diseases, Institute of Psychiatry, Psychology &
Neuroscience, London SE1 1UL, U.K.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.jmedchem.0c02000.
■
sı
Effect of (S)-VU0637120 on G protein activation,
internalization studies, Y4R variant screening for the
investigation of the VU0637120 binding pocket,
RosettaLigand docking clusters, analytics of compounds,
and synthetic peptide ligands (Figures S1−S14); small-
molecule screening data, PP activation of Y4R/Y1R
chimeras and Y4R variants, templates for Y4R homology
modeling, overview of docking parameters (Tables S1−
S7) (PDF)
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.jmedchem.0c02000
Author Contributions
C.S., O.V., and M.S. contributed equally to this work. All
authors have given approval to the final version of the
manuscript.
Coordinates of final homology models (A3.29P-Y4-
NAM.pdb) (PDB)
Coordinates of final homology models (A3.29V-Y4-
NAM.pdb) (PDB)
Coordinates of final homology models (E6.58R-Y4-
NAM.pdb) (PDB)
Coordinates of final homology models (Y4-NAM-
cluster1.pdb) (PDB)
Coordinates of final homology models (Y4-NAM-
cluster2.pdb) (PDB)
Coordinates of final homology models (Y4-NAM-
cluster3.pdb) (PDB)
Notes
The authors declare the following competing financial
interest(s): C.D.W. is an owner of WaveFront Biosciences,
maker of the Panoptic instrument used in high-throughput
screening.
ACKNOWLEDGMENTS
■
̈
The authors thank K. Lobner, J. Schwesinger, and C.
Dammann for technical assistance. We kindly acknowledge
the Vanderbilt HTS facility for the help in the identification of
the Y4R modulators. Dr. A. Rodriguez, B.J. Bender, and N.
Coordinates of final homology models (Y4-PP.pdb)
(PDB)
Molecular formula strings (CSV, TIF)
Kuhn are acknowledged for discussions about data analysis and
̈
work in docking studies. We are grateful to N. Kett of the
Vanderbilt University Warren Center for Neuroscience Drug
Discovery for performing the chiral SFC analysis. We
acknowledge funding by the National Institutes of Health
(Grant R01 DK0973376 to J.M. and C.D.W., grant R01
DA046138 to J.M. and grant 1S10OD021734 for the Panoptic
kinetic imaging plate reader), the Deutsche Forschungsge-
meinschaft (DFG, German Research Foundation) through
CRC 1423, project number 421152132, B1 to A.G.B.-S and A7
to J.M., the European Union and the Free State of Saxony
(Grants 100148835 and 143213128452 to A.G.B.-S.), and the
BBSRC (BB/N006763/1 to H.M.C.).
AUTHOR INFORMATION
Corresponding Authors
■
Jens Meiler − Department of Chemistry and Department of
Pharmacology, Vanderbilt University, Nashville, Tennessee
37235, United States; Institute for Drug Discovery, Leipzig
University, Leipzig 04103, Germany; Email: jens@
meilerlab.org
Annette G. Beck-Sickinger − Institute of Biochemistry,
Leipzig University, Leipzig 04103, Germany; orcid.org/
0000-0003-4560-8020; Email: abeck-sickinger@uni-
leipzig.de
Authors
Corinna Schuß − Institute of Biochemistry, Leipzig University,
̈
ABBREVIATIONS
■
Leipzig 04103, Germany
Oanh Vu − Department of Chemistry, Vanderbilt University,
Nashville, Tennessee 37235, United States; orcid.org/
0000-0003-4704-7538
Mario Schubert − Institute of Biochemistry, Leipzig University,
Leipzig 04103, Germany
Yu Du − Department of Pharmacology, Vanderbilt University,
Nashville, Tennessee 37232, United States
Nigam M. Mishra − Department of Pharmaceutical Sciences,
UNT System College of Pharmacy, University of North Texas
Health Science Center, Fort Worth, Texas 76107, United
States
Iain R. Tough − King’s College London, Wolfson Centre for
Age-Related Diseases, Institute of Psychiatry, Psychology &
Neuroscience, London SE1 1UL, U.K.
ACN, acetonitrile; arr3, arrestin3; CCR5, C−C chemokine
receptor type 5; COS7, African green monkey kidney cells;
DMEM, Dulbecco’s modified Eagle’s medium; dNTP,
deoxyribonucleotide; DPBS, Dulbecco’s phosphate-buffered
saline; ECL, extracellular loop; eYFP, enhanced yellow
fluorescent protein; FBS, fetal bovine serum; GPCR, G
protein-coupled receptor; HBSS, Hank’s balanced saline
solution; HEK293, human embryonic kidney cells; HM,
homology model; IP, inositol phosphate; KB, equilibrium
binding constant; LB, Luria broth; MALDI-TOF, matrix-
assisted laser desorption/ionization time of flight; NAM,
negative allosteric modulator; NPY, neuropeptide Y; PAM,
positive allosteric modulator; PAR2, protease-activated re-
ceptor 2; PP, pancreatic polypeptide; PYY, peptide YY; Rluc8,
Renilla luciferase 8; RP-HPLC, reversed-phase high-perform-
ance liquid chromatography; rPP, rat pancreatic polypeptide;
SEM, standard error of the mean; TM, transmembrane helix;
Jan Stichel − Institute of Biochemistry, Leipzig University,
Leipzig 04103, Germany
C. David Weaver − Department of Chemistry, Department of
Pharmacology, and Institute of Chemical Biology, Vanderbilt
Y
1/2/4/5R, neuropeptide Y receptor 1/2/4/5
2812
https://dx.doi.org/10.1021/acs.jmedchem.0c02000
J. Med. Chem. 2021, 64, 2801−2814

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
(18) Bdioui, S.; Verdi, J.; Pierre, N.; Trinquet, E.; Roux, T.; Kenakin,
T. Equilibrium assays are required to accurately characterize the
activity profiles of drugs modulating Gq-protein-coupled receptors.
Mol. Pharmacol. 2018, 94, 992−1006.
REFERENCES
■
(1) Pedragosa-Badia, X.; Stichel, J.; Beck-Sickinger, A. G. Neuro-
peptide Y receptors: how to get subtype selectivity. Front. Endocrinol.
2013, 4, No. 5.
(19) Christopoulos, A.; Kenakin, T. G protein-coupled receptor
allosterism and complexing. Pharmacol. Rev. 2002, 54, 323−374.
(2) Yulyaningsih, E.; Zhang, L.; Herzog, H.; Sainsbury, A. NPY
receptors as potential targets for anti-obesity drug development. Br. J.
Pharmacol. 2011, 163, 1170−1202.
̈
(20) Wanka, L.; Babilon, S.; Burkert, K.; Morl, K.; Gurevich, V. V.;
Beck-Sickinger, A. G. C-terminal motif of human neuropeptide Y4
receptor determines internalization and arrestin recruitment. Cell.
Signalling 2017, 29, 233−239.
(3) Zhang, L.; Bijker, M. S.; Herzog, H. The neuropeptide Y system:
pathophysiological and therapeutic implications in obesity and cancer.
Pharmacol. Ther. 2011, 131, 91−113.
(21) Burford, N. T.; Watson, J.; Bertekap, R.; Alt, A. Strategies for
the identification of allosteric modulators of G-protein-coupled
receptors. Biochem. Pharmacol. 2011, 81, 691−702.
(4) Sato, N.; Ogino, Y.; Mashiko, S.; Ando, M. Modulation of
neuropeptide Y receptors for the treatment of obesity. Expert Opin.
Ther. Pat. 2009, 19, 1401−1415.
(22) Ballesteros, J. A.; Weinstein, H. [19] Integrated methods for the
construction of three-dimensional models and computational probing
of structure-function relations in G protein-coupled receptors.
Methods Neurosci. 1995, 25, 366−428.
(5) Batterham, R. L.; Le Roux, C. W.; Cohen, M. A.; Park, A. J.;
Ellis, S. M.; Patterson, M.; Frost, G. S.; Ghatei, M. A.; Bloom, S. R.
Pancreatic polypeptide reduces appetite and food intake in humans. J.
Clin. Endocrinol. Metab. 2003, 88, 3989−3992.
(23) Isberg, V.; Mordalski, S.; Munk, C.; Rataj, K.; Harpsoe, K.;
Hauser, A. S.; Vroling, B.; Bojarski, A. J.; Vriend, G.; Gloriam, D. E.
GPCRdb: an information system for G protein-coupled receptors.
Nucleic Acids Res. 2016, 44, D356−D364.
(24) Zhang, L.; Riepler, S. J.; Turner, N.; Enriquez, R. F.; Lee, I. C.;
Baldock, P. A.; Herzog, H.; Sainsbury, A. Y2 and Y4 receptor signaling
synergistically act on energy expenditure and physical activity. Am. J.
Physiol.: Regul., Integr. Comp. Physiol. 2010, 299, R1618−R1628.
(25) Tough, I. R.; Holliday, N. D.; Cox, H. M. Y(4) receptors
mediate the inhibitory responses of pancreatic polypeptide in human
and mouse colon mucosa. J. Pharmacol. Exp. Ther. 2006, 319, 20−30.
(26) Meiler, J.; Baker, D. ROSETTALIGAND: protein-small
molecule docking with full side-chain flexibility. Proteins 2006, 65,
538−548.
(27) Combs, S. A.; Deluca, S. L.; Deluca, S. H.; Lemmon, G. H.;
Nannemann, D. P.; Nguyen, E. D.; Willis, J. R.; Sheehan, J. H.; Meiler,
J. Small-molecule ligand docking into comparative models with
Rosetta. Nat. Protoc. 2013, 8, 1277−1298.
(28) Lemmon, G.; Meiler, J. Rosetta ligand docking with flexible
XML protocols. Methods Mol. Biol. 2012, 819, 143−155.
(29) Song, Y.; DiMaio, F.; Wang, R. Y.; Kim, D.; Miles, C.; Brunette,
T.; Thompson, J.; Baker, D. High-resolution comparative modeling
with RosettaCM. Structure 2013, 21, 1735−1742.
(6) Jesudason, D. R.; Monteiro, M. P.; McGowan, B. M.; Neary, N.
M.; Park, A. J.; Philippou, E.; Small, C. J.; Frost, G. S.; Ghatei, M. A.;
Bloom, S. R. Low-dose pancreatic polypeptide inhibits food intake in
man. Br. J. Nutr. 2007, 97, 426−429.
(7) Hofmann, S.; Frank, R.; Hey-Hawkins, E.; Beck-Sickinger, A. G.;
Schmidt, P. Manipulating Y receptor subtype activation of short
neuropeptide Y analogs by introducing carbaboranes. Neuropeptides
2013, 47, 59−66.
(8) Parker, E. M.; Babij, C. K.; Balasubramaniam, A.; Burrier, R. E.;
Guzzi, M.; Hamud, F.; Mukhopadhyay, G.; Rudinski, M. S.; Tao, Z.;
Tice, M.; Xia, L.; Mullins, D. E.; Salisbury, B. G. GR231118
(1229U91) and other analogues of the C-terminus of neuropeptide Y
are potent neuropeptide Y Y1 receptor antagonists and neuropeptide
Y Y4 receptor agonists. Eur. J. Pharmacol. 1998, 349, 97−105.
(9) Balasubramaniam, A.; Mullins, D. E.; Lin, S.; Zhai, W.; Tao, Z.;
Dhawan, V. C.; Guzzi, M.; Knittel, J. J.; Slack, K.; Herzog, H.; Parker,
E. M. Neuropeptide Y (NPY) Y4 receptor selective agonists based on
NPY(32-36): development of an anorectic Y4 receptor selective
agonist with picomolar affinity. J. Med. Chem. 2006, 49, 2661−2665.
(10) Kuhn, K. K.; Ertl, T.; Dukorn, S.; Keller, M.; Bernhardt, G.;
Reiser, O.; Buschauer, A. High affinity agonists of the neuropeptide Y
(NPY) Y4 receptor derived from the C-terminal pentapeptide of
human pancreatic polypeptide (hPP): synthesis, stereochemical
discrimination, and radiolabeling. J. Med. Chem. 2016, 59, 6045−
6058.
(30) Pedragosa-Badia, X.; Sliwoski, G. R.; Dong Nguyen, E.;
Lindner, D.; Stichel, J.; Kaufmann, K. W.; Meiler, J.; Beck-Sickinger,
A. G. Pancreatic polypeptide is recognized by two hydrophobic
domains of the human Y4 receptor binding pocket. J. Biol. Chem.
2014, 289, 5846−5859.
(11) Konieczny, A.; Braun, D.; Wifling, D.; Bernhardt, G.; Keller, M.
Oligopeptides as neuropeptide Y Y4 receptor ligands: Identification of
a high-affinity tetrapeptide agonist and a hexapeptide antagonist. J.
Med. Chem. 2020, 63, 8198−8215.
(31) Yang, Z.; Han, S.; Keller, M.; Kaiser, A.; Bender, B. J.; Bosse,
̈
M.; Burkert, K.; Kogler, L. M.; Wifling, D.; Bernhardt, G.; Plank, N.;
(12) Sliwoski, G.; Schubert, M.; Stichel, J.; Weaver, D.; Beck-
Sickinger, A. G.; Meiler, J. Discovery of small-molecule modulators of
the human Y4 receptor. PLoS One 2016, 11, No. e0157146.
(13) Schubert, M.; Stichel, J.; Du, Y.; Tough, I. R.; Sliwoski, G.;
Meiler, J.; Cox, H. M.; Weaver, C. D.; Beck-Sickinger, A. G.
Identification and characterization of the first selective Y4 receptor
positive allosteric modulator. J. Med. Chem. 2017, 60, 7605−7612.
(14) Keller, M.; Kaske, M.; Holzammer, T.; Bernhardt, G.;
Buschauer, A. Dimeric argininamide-type neuropeptide Y receptor
antagonists: chiral discrimination between Y1 and Y4 receptors.
Bioorg. Med. Chem. 2013, 21, 6303−6322.
Littmann, T.; Schmidt, P.; Yi, C.; Li, B.; Ye, S.; Zhang, R.; Xu, B.;
Larhammar, D.; Stevens, R. C.; Huster, D.; Meiler, J.; Zhao, Q.; Beck-
Sickinger, A. G.; Buschauer, A.; Wu, B. Structural basis of ligand
binding modes at the neuropeptide Y Y1 receptor. Nature 2018, 556,
520−524.
̈
̈
(32) Merten, N.; Lindner, D.; Rabe, N.; Rompler, H.; Morl, K.;
̈
Schoneberg, T.; Beck-Sickinger, A. G. Receptor subtype-specific
docking of Asp6.59 with C-terminal arginine residues in Y receptor
ligands. J. Biol. Chem. 2007, 282, 7543−7551.
(33) Tan, Q.; Zhu, Y.; Li, J.; Chen, Z.; Han, G. W.; Kufareva, I.; Li,
T.; Ma, L.; Fenalti, G.; Li, J.; Zhang, W.; Xie, X.; Yang, H.; Jiang, H.;
Cherezov, V.; Liu, H.; Stevens, R. C.; Zhao, Q.; Wu, B. Structure of
the CCR5 chemokine receptor-HIV entry inhibitor maraviroc
complex. Science 2013, 341, 1387−1390.
(34) Cheng, R. K. Y.; Fiez-Vandal, C.; Schlenker, O.; Edman, K.;
Aggeler, B.; Brown, D. G.; Brown, G. A.; Cooke, R. M.; Dumelin, C.
E.; Dore, A. S.; Geschwindner, S.; Grebner, C.; Hermansson, N. O.;
Jazayeri, A.; Johansson, P.; Leong, L.; Prihandoko, R.; Rappas, M.;
Soutter, H.; Snijder, A.; Sundstrom, L.; Tehan, B.; Thornton, P.;
Troast, D.; Wiggin, G.; Zhukov, A.; Marshall, F. H.; Dekker, N.
Structural insight into allosteric modulation of protease-activated
receptor 2. Nature 2017, 545, 112−115.
(15) Kuhn, K. K.; Littmann, T.; Dukorn, S.; Tanaka, M.; Keller, M.;
Ozawa, T.; Bernhardt, G.; Buschauer, A. In Search of NPY Y4R
antagonists: Incorporation of carbamoylated arginine, aza-amino
acids, or d-amino acids into oligopeptides derived from the C-termini
of the endogenous agonists. ACS Omega 2017, 2, 3616−3631.
(16) Hirose, W.; Sato, K.; Matsuda, A. Fluorescence properties of 5-
(5,6-dimethoxybenzothiazol-2-yl)-2-deoxyuridine (dbtU) and oligo-
deoxyribonucleotides containing dbtU. Eur. J. Org. Chem. 2011,
6206−6217.
(17) Leach, K.; Sexton, P. M.; Christopoulos, A. Allosteric GPCR
modulators: taking advantage of permissive receptor pharmacology.
Trends Pharmacol. Sci. 2007, 28, 382−389.
2813
https://dx.doi.org/10.1021/acs.jmedchem.0c02000
J. Med. Chem. 2021, 64, 2801−2814

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
(35) Ziemek, R.; Schneider, E.; Kraus, A.; Cabrele, C.; Beck-
Sickinger, A. G.; Bernhardt, G.; Buschauer, A. Determination of
affinity and activity of ligands at the human neuropeptide Y Y4
receptor by flow cytometry and aequorin luminescence. J. Recept.
Signal Transduction 2007, 27, 217−233.
(36) Rudolf, K.; Eberlein, W.; Engel, W.; Wieland, H. A.; Willim, K.
D.; Entzeroth, M.; Wienen, W.; Beck-Sickinger, A. G.; Doods, H. N.
The first highly potent and selective non-peptide neuropeptide Y Y1
receptor antagonist: BIBP3226. Eur. J. Pharmacol. 1994, 271, R11−
R13.
Conference on Computational Advances in Bio and Medical Sciences
(ICCABS), 2011; pp 13−18.
(53) Kothiwale, S.; Mendenhall, J. L.; Meiler, J. BCL::Conf: small
molecule conformational sampling using a knowledge based rotamer
library. J. Cheminf. 2015, 7, No. 47.
(54) Moretti, R.; Bender, B. J.; Allison, B.; Meiler, J. Rosetta and the
design of ligand binding sites. Methods Mol. Biol. 2016, 1414, 47−62.
(55) Kenakin, T. P. Allosteric Drug Antagonism. A Pharmacology
PrimerTheory. Applications, and Methods, 2nd ed.; Elsevier
Academic Press: Amsterdam, Boston, 2006; pp 143−145.
(37) Wieland, H. A.; Engel, W.; Eberlein, W.; Rudolf, K.; Doods, H.
N. Subtype selectivity of the novel nonpeptide neuropeptide Y Y1
receptor antagonist BIBO 3304 and its effect on feeding in rodents.
Br. J. Pharmacol. 1998, 125, 549−555.
(38) Craik, D. J.; Fairlie, D. P.; Liras, S.; Price, D. The future of
peptide-based drugs. Chem. Biol. Drug Des. 2013, 81, 136−147.
(39) Adrian, T. E.; Greenberg, G. R.; Besterman, H. S.; Bloom, S. R.
Pharmacokinetics of pancreatic polypeptide in man. Gut 1978, 19,
907−909.
(40) Gerald, C.; Walker, M. W.; Criscione, L.; Gustafson, E. L.;
Batzl-Hartmann, C.; Smith, K. E.; Vaysse, P.; Durkin, M. M.; Laz, T.
M.; Linemeyer, D. L.; Schaffhauser, A. O.; Whitebread, S.; Hofbauer,
K. G.; Taber, R. I.; Branchek, T. A.; Weinshank, R. L. A receptor
subtype involved in neuropeptide-Y-induced food intake. Nature
1996, 382, 168−171.
(41) Gao, Z. G.; Jacobson, K. A. Allosteric modulation and
functional selectivity of G protein-coupled receptors. Drug Discovery
Today: Technol. 2013, 10, e237−e243.
(42) Wootten, D.; Christopoulos, A.; Sexton, P. M. Emerging
paradigms in GPCR allostery: implications for drug discovery. Nat.
Rev. Drug Discovery 2013, 12, 630−644.
(43) Wu, F.; Song, G.; de Graaf, C.; Stevens, R. C. Structure and
function of peptide-binding G protein-coupled receptors. J. Mol. Biol.
2017, 429, 2726−2745.
̈
(44) Made, V.; Babilon, S.; Jolly, N.; Wanka, L.; Bellmann-Sickert,
̈
K.; Diaz Gimenez, L. E.; Morl, K.; Cox, H. M.; Gurevich, V. V.; Beck-
Sickinger, A. G. Peptide modifications differentially alter G protein-
coupled receptor internalization and signaling bias. Angew. Chem., Int.
Ed. 2014, 53, 10067−10071.
̈
̈
(45) Bohme, I.; Stichel, J.; Walther, C.; Morl, K.; Beck-Sickinger, A.
G. Agonist induced receptor internalization of neuropeptide Y
receptor subtypes depends on third intracellular loop and C-terminus.
Cell. Signalling 2008, 20, 1740−1749.
(46) Kostenis, E.; Degtyarev, M. Y.; Conklin, B. R.; Wess, J. The N-
terminal extension of Gαq is critical for constraining the selectivity of
receptor coupling. J. Biol. Chem. 1997, 272, 19107−19110.
(47) Bryksin, A. V.; Matsumura, I. Overlap extension PCR cloning: a
simple and reliable way to create recombinant plasmids. BioTechniques
2010, 48, 463−465.
(48) Ho, S. N.; Hunt, H. D.; Horton, R. M.; Pullen, J. K.; Pease, L.
R. Site-directed mutagenesis by overlap extension using the
polymerase chain reaction. Gene 1989, 77, 51−59.
(49) Bradford, M. M. A rapid and sensitive method for the
quantitation of microgram quantities of protein utilizing the principle
of protein-dye binding. Anal. Biochem. 1976, 72, 248−254.
(50) Sievers, F.; Wilm, A.; Dineen, D.; Gibson, T. J.; Karplus, K.; Li,
̈
W.; Lopez, R.; McWilliam, H.; Remmert, M.; Soding, J.; Thompson, J.
D.; Higgins, D. G. Fast, scalable generation of high-quality protein
multiple sequence alignments using Clustal Omega. Mol. Syst. Biol.
2011, 7, No. 539.
(51) Bender, B. J.; Cisneros, A., 3rd; Duran, A. M.; Finn, J. A.; Fu,
D.; Lokits, A. D.; Mueller, B. K.; Sangha, A. K.; Sauer, M. F.; Sevy, A.
M.; Sliwoski, G.; Sheehan, J. H.; DiMaio, F.; Meiler, J.; Moretti, R.
Protocols for molecular modeling with Rosetta3 and RosettaScripts.
Biochemistry 2016, 55, 4748−4763.
(52) Alexander, N.; Woetzel, N.; Meiler, J. In Bcl::Cluster: A Method
for Clustering Biological Molecules Coupled with Visualization in the
Pymol Molecular Graphics System, 2011 IEEE 1st International
2814
https://dx.doi.org/10.1021/acs.jmedchem.0c02000
J. Med. Chem. 2021, 64, 2801−2814