Chemical subtleties in small-molecule modulation of peptide receptor function: The case of CXCR3 biaryl-type ligands

Article abstract of DOI:10.1021/jm301240t

The G protein-coupled chemokine receptor CXCR3 plays a role in numerous inflammatory events. The endogenous ligands for the chemokine receptors are peptides, but in this study we disclose small-molecule ligands that are able to activate CXCR3. A class of biaryl-type compounds that is assembled by convenient synthetic routes is described as a new class of CXCR3 agonists. Intriguingly, structure-activity relationship and structure-function relationship studies reveal that subtle chemical modifications on the outer aryl ring (e.g., either the size or position of a halogen atom) result in a full spectrum of agonist efficacies on CXCR3. Quantum mechanics calculations and nuclear Overhauser effect spectroscopy NMR studies suggest that the biaryl dihedral angle and the electronic nature of ortho-substituents play an important role in determining agonist efficacies. Compounds 38 (VUF11222) and 39 (VUF11418) are the first reported nonpeptidomimetic agonists on CXCR3, rendering them highly useful chemical tools for detailed assessment of CXCR3 activation as well as for studying downstream CXCR3 signaling.

Full text of DOI:10.1021/jm301240t

Article
pubs.acs.org/jmc
Chemical Subtleties in Small-Molecule Modulation of Peptide
Receptor Function: The Case of CXCR3 Biaryl-Type Ligands
Maikel Wijtmans,†,§ Danny J. Scholten,†,§ Luc Roumen,†,§ Meritxell Canals,†,∥ Hans Custers,†
Marjolein Glas,† Marlies C. A. Vreeker,† Frans J. J. de Kanter,‡ Chris de Graaf,† Martine J. Smit,†
Iwan J. P. de Esch,† and Rob Leurs*,†
†Leiden/Amsterdam Center for Drug Research, Division of Medicinal Chemistry, Faculty of Sciences, VU University Amsterdam,
The Netherlands
‡Division of Organic and Inorganic Chemistry, Faculty of Sciences, VU University Amsterdam, The Netherlands
S
* Supporting Information
ABSTRACT: The G protein-coupled chemokine receptor
CXCR3 plays a role in numerous inflammatory events. The
endogenous ligands for the chemokine receptors are peptides,
but in this study we disclose small-molecule ligands that are
able to activate CXCR3. A class of biaryl-type compounds that
is assembled by convenient synthetic routes is described as a
new class of CXCR3 agonists. Intriguingly, structure−activity
relationship and structure−function relationship studies reveal
that subtle chemical modifications on the outer aryl ring (e.g.,
either the size or position of a halogen atom) result in a full
spectrum of agonist efficacies on CXCR3. Quantum mechanics
calculations and nuclear Overhauser effect spectroscopy NMR
studies suggest that the biaryl dihedral angle and the electronic
nature of ortho-substituents play an important role in determining agonist efficacies. Compounds 38 (VUF11222) and 39
(VUF11418) are the first reported nonpeptidomimetic agonists on CXCR3, rendering them highly useful chemical tools for
detailed assessment of CXCR3 activation as well as for studying downstream CXCR3 signaling.
one example being the frequently used antagonist 1 (NBI-
74330).7 Small-molecule agonism of CXCR3 has been studied
less widely, likely because evidence for the therapeutic value of
CXCR3 agonism is scarce and relatively preliminary.8,9 Another
explanation for this apparent deficiency might be that it is easier
to inhibit peptide receptor activation by small molecules than to
mimic the specific action of the endogenous peptide agonist.
Only one paper disclosing small-molecule agonists of CXCR3
has appeared, reporting on two agonist chemotypes which both
clearly bear peptidomimetic character (2a,b and 3, Figure 1).10
In fact, structure−activity relationship (SAR) studies have
revealed that the basic amino acid side chain is crucial in
activating CXCR3,10 and conformational calculations on 2b
have shown that this compound probably mimics the “30s
loop” of CXCL10 which includes Arg38.11 Several of these
peptidomimetics have since been pharmacologically studied in
vitro and in vivo.8,12,13 Even though several X-ray structures for
the related chemokine receptor CXCR4 have been elucidated
recently,14 ligand-based information such as the SAR and
structure−function relationship (SFR) is still crucial to provide
insights into molecular determinants of ligand-induced (chemo-
1. INTRODUCTION
G protein-coupled receptors (GPCRs) form one of the
therapeutically most relevant protein families, and around
one-third of the available small-molecule drugs target GPCRs.1
Despite this therapeutic importance, it is often still unclear
which structural determinants of the small molecules underlie
the functional outcome of ligand−GPCR binding. This
question is perhaps most intriguing in the case of peptide
GPCRs, which signal as a result of the binding of large peptides.
A well-known class of peptide GPCRs are the chemokine
receptors, 19 subtypes of which are currently known.2
Chemokine receptors are activated by chemokines, which are
chemotactic cytokine peptides that control leukocyte homeo-
stasis, as well as leukocyte recruitment to sites of inflammation
or infection. As such, the chemokine system is a key player in
the human immune system.3 The chemokine receptor CXCR3
is expressed on various immune cells, including T-cells, and
recognizes three endogenous agonist ligands, CXCL9 (11.7
kDa), CXCL10 (8.6 kDa), and CXCL11 (8.3 kDa).
Upregulation of CXCR3 and its ligands is linked to a variety
of immune-related disorders. The notion of CXCR3 as an
interesting therapeutic target is reflected in the high number of
reports4−6 describing the discovery and optimization of small-
molecule CXCR3 antagonists from many different chemotypes,
Received: August 28, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jm301240t | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 1. Chemical structures of selected CXCR3 ligands.
a
Scheme 1. Synthesis of Building Blocks
a
Reagents and conditions: (a) H2, Pd/C, MeOH, 23 h, rt, 72%; (b) (1) H2NMe, MeOH/EtOH, 1−20 h, rt; (2) NaBH4, 2 h, rt, 9: 50%, 10: 14% as
HCl salt; (c) Pd(PPh3)4, 4-CHOPhB(OH)2, Na2CO3, DME, H2O, 60−90 min, microwave, 56−91%; (d) Pd(PPh3)4, 4-BrPhCHO (18), Na2CO3,
DME, H2O, 7 h, reflux, 85%; (e) BH3·THF, THF, 2.5 h, rt; (f) MnO2, DCM, 5 d, rt, 65% from 19.
kine) GPCR modulation.15,16 As such, studies aimed at
unraveling ligand-mediated CXCR3 activation and its cellular
consequences would greatly benefit from studies on non-
peptidomimetic small-molecule activators of CXCR3.
We have recently reported on biaryl-type CXCR3 ligands (4,
Figure 1), obtained after several rounds of modification
following a medium-throughput screen performed in our
laboratories.17 The SAR of scaffold 4 showed varying effects
in affinity upon modulation of the “outer” aryl group R. Key
members of this chemotype were found to behave as
antagonists.17 In the current paper, we describe how these
compounds can be turned into CXCR3 (full) agonists using
intriguingly subtle activation triggers such as manipulation of
the size of a halogen substituent as well as of the halogen
position. These compounds display the full spectrum of
functional efficacies in CXCR3 activation and include the first
reported nonpeptidomimetic full agonists on CXCR3. We
present pharmacological, nuclear Overhauser effect spectrosco-
py NMR (NOESY NMR), and quantum mechanics (QM)
studies to identify the molecular features that are responsible
for the specific functional activity.
B
dx.doi.org/10.1021/jm301240t | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 2. Synthesis of Final Compounds
a
Reagents and conditions: (a) NaBH(OAc)3, DCE, rt, 4 h to 2 d, 43−92%; (b) NaBH(OAc)3, DCE, rt, 20 h, 21: 71% as HCl salt, 22: 78%; (c)
Pd(PPh3)4, RB(OH)2, Na2CO3, H2O, DME or iPrOH/toluene, microwave, 60−90 min, 34−92%; (d) DCM, MeI, 1−5 d, rt, 22−88%.
which the native coupling of CXCR3 to G proteins is left
intact, 6 proved to be a partial agonist (α = 0.73) with
micromolar potency (pEC50 = 5.8) (Table 1, Figure 2). This
partial agonist activity could be abolished by addition of
CXCR3-selective antagonist 17 (Figure S1A, Supporting
Information) and by our previously reported antagonist 27
(VUF10990)17 (data not shown). Moreover, no functional
response for 6 was detected in cell membranes that did not
express CXCR3 (Figure S1C), suggesting that the observed
functional effect for 6 was a consequence of CXCR3 receptor
activation. These interesting findings on a nonpeptidomimetic
scaffold, which is readily assembled through straightforward
synthesis, inspired us to an in-depth investigation of the
molecular determinants for activation of the CXCR3 receptor
by this ammonium salt scaffold. The corresponding tertiary
amines were not investigated because their affinities are
expected to be considerably lower compared to those of the
ammonium salts, as generally observed in this class.17
2.2.1. Exploration of Biaryl Substitution. The newly
synthesized ammonium salts bind to CXCR3 with affinities
ranging from pKi = 5.7 to pKi = 7.3 in a radioligand binding
displacement assay ([125I]CXCL10). Figure 2A shows
representative binding curves. All compounds were also
analyzed in the [35S]GTPγS functional assay to investigate
potency (pEC50) and efficacy (α) at CXCR3 (Table 1).
Concentration ranges in the [35S]GTPγS functional assay were
chosen such that sufficient receptor occupancy was ensured.
The profile of 6 was compared to those of a small set of
ammonium salts previously reported by us (23−27).17
Compounds 23 and 24 (lacking the biaryl moiety) as well as
25 and 26 (having an undecorated (un)constrained biaryl
system) showed little to no agonistic response (α ≤ 0.1).
Remarkably, a shift of the Cl atom in 6 to the meta-position
(giving 27) abolished the agonistic activity. These findings
indicate the ortho-position on the outer phenyl ring of the biaryl
system to be important for receptor activation. With the Cl
atom being maintained at this position, the effect of an
additional Cl substituent was explored (28−30). This yielded
partial agonists, which are all weaker than compound 6 with α
values ranging from 0.31 to 0.64 (Table 1). Recognizing
hindered rotation around the biaryl axis as a potential key factor
for CXCR3 agonism, three compounds with a Cl atom on the
ortho-position of the inner ring were also investigated (31−33).
However, only 33 showed substantial agonism (α = 0.30)
(Table 1), arguably because it contains an o-Cl moiety on the
2. RESULTS AND DISCUSSION
2.1. Synthesis. All final salts 6 and 28−46 were prepared as
outlined in Schemes 1 and 2. The syntheses of analogues 5 and
23−27 through similar routes have been reported by us
previously.17 Hydrogenation of key amine building block 717
provided 8, with 2D NMR spectroscopy indicating an endo
geometry. Stereoisomeric analogue 9 and regioisomeric
analogue 10 were prepared from the corresponding aldehydes
11 and 12.18,19 Biaryl aldehydes 13−17 were prepared through
a Suzuki reaction of an appropriate boronic acid and a
complementary halogenated precursor. This “reversed” pre-
assembly20 of the biaryl moiety in 13−17 was initially pursued
because we observed that 2-bromophenylboronic acid and 2,6-
dichlorophenylboronic acid proved difficult to couple to 21 in
the alternative Suzuki strategy depicted in Scheme 2.
Thereafter, the former procedure was also applied to several
other substrates. Substantial amounts of disubstituted product
were obtained in the formation of 15 despite the 5-fold excess
of 1,2-diiodobenzene. Chlorine-containing building block 20
was prepared through a reduction of the corresponding
carboxylic acid 19 to the alcohol followed by an oxidation.
The syntheses of all final compounds (Scheme 2) proceeded
mostly along the general routes previously published.17 Thus,
tertiary amine precursors 47−66 were obtained through two
approaches. One approach uses a reductive amination to
connect a preassembled biarylaldehyde (13−17) to building
blocks 7−10. The other approach is a two-step process, starting
with reductive alkylation of 7 using aldehyde 18 or 20 to give
bromides 21 and 22, respectively. This was followed by a
microwave Suzuki reaction of 21 and 22 and an appropriate
boronic acid. Tertiary amine precursors 47−66 were then
methylated with MeI to the final ammonium salts 6 and 28−46
in moderate to excellent yield.
2.2. Structure−Function Relationships. In the SAR
efforts on scaffold 4, the difference in affinity between o-F
compound 5 and its meta-analogue had previously indicated
meta-substitution of the outer phenyl ring as a means to achieve
affinity.17 More recently, we prepared and tested the o-Cl
derivative 6. Interestingly, when subjected to our previously
used17 functional CXCR3 assay recording [3H]-
inositolphosphate accumulation after stimulation with
CXCL10, 6 did not show any inhibition of CXCL10 responses
despite full receptor occupancy (not shown), suggesting that 6
may possess residual agonist activity. Indeed, in a [35S]GTPγS
(guanosine 5′-O-[γ-thio]triphosphate) functional assay, in
C
dx.doi.org/10.1021/jm301240t | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 1. General Scan of Substituents on the Outer and Inner Phenyl Rings
a
Measured by [125I]CXCL10 displacement binding experiments with membranes prepared from HEK293 cells stably expressing the CXCR3
b
receptor. Measured by using a [35S]GTPγS functional assay with membranes prepared from HEK293 cells stably expressing the CXCR3 receptor.
c
d
α represents the relative efficacy of a ligand compared to the endogenous agonist CXCL11 (which is set at α = 1.0). Could not be determined due
e
to the too low functional assay window. The point of attachment is colored red. All values shown are the mean SEM of at least three independent
experiments.
outer ring as well. Subsequent SAR/SFR studies focused on the
ortho-position of the outer ring and aimed to introduce groups
differing in electronic properties and size. Introduction of one
(34) or two (35) methyl groups resulted in weak partial
agonism (α ≈ 0.2). An o-OH group (36) induced no significant
functional agonism, and the affinity for this compound was the
lowest in this series (pKi = 5.7) (Table 1). Moderate to weak
partial agonism was observed with an o-OMe (37; α = 0.36) or
o-F (5; α = 0.20) substituent. However, the introduction of
bigger halogen atoms, such as o-Br (38) or o-I (39), resulted in
CXCR3 agonists that are able to reach the same maximum level
of receptor activation (α = 0.95 and 0.99, respectively) as the
endogenous peptide ligand CXCL11 (Figure 2B). Both 38 and
39 were comparably effective in terms of potency and efficacy,
matching those of peptidomimetic agonist 2a (VUF10661).13
In contrast, mere shifting of the Br atom in 38 to the meta-
position (as in 40) led to a complete loss of agonism, as had
also been observed in the couple 6 versus 27 (Figure 2C). In
all, the bigger halogens at the ortho-position of the outer ring
were clearly preferred to provide full functional agonists.
2.2.2. Exploration of the Bicycloaliphatic Group. A selected
set of compounds was subsequently prepared to explore the
“left-hand” bicycloaliphatic side of the molecule. We have
previously reported on the preference of CXCR3 for certain
polycycloaliphatic groups,21 and within this class of biaryl
ligands, the myrtenyl seems an attractive group.17 Given the
subtle changes in activation pattern encountered in the biaryl
moiety, we aimed for subtle changes in the bicycloaliphatic side
of the molecule as well. Toward this end, saturated endo-
myrtanyl analogues (41, 42) as well as derivatives with either
D
dx.doi.org/10.1021/jm301240t | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 2. Radioligand binding curves and functional assay curves for selected compounds: (A) binding displacement assay using [125I]CXCL10 for
key compounds, (B, C) [35S]GTPγS functional assay for (B) compounds containing different (halogen) atoms on the ortho-position of the outer
phenyl ring and for (C) compounds with a chloro or bromo group on the ortho- or meta-position of the outer phenyl ring, (D) CRE-luciferase
reporter gene functional assay with key compounds on HEK293 cells transfected with the CXCR3 receptor and the CRE-luciferase reporter gene
DNA. The graphs shown are the result of grouped data of at least three independent experiments. The dashed line (at 1.0) indicates basal levels of
[35S]GTPγS accumulation, whereas the dashed/dotted line (at 2.8) represents the maximal response (Emax) of the endogenous peptide agonist
CXCL11 (α = 1). Compounds 1 and 2a are included as a reference antagonist and agonist, respectively.
regioisomeric (43, 44) or stereoisomeric (45, 46) forms of the
myrtenyl group were prepared (Table 2). A m-Cl or o-I moiety
was kept as a “representative” biaryl portion of the antagonist
and agonist subseries of this class, respectively. The results for
41−46 (Table 2) clearly show that the nature of the
bicycloaliphatic group had only little effect on the affinity
and, more importantly, on the functional profile of both the
antagonist and the agonist subseries (compare 39 to 42/44/46
and 27 to 41/43/45). The most notable change observed was
the small decrease in efficacy of 42 (α = 0.83) compared to 39
(α = 0.99) as a result of double bond reduction. The collective
results not only suggest that these types of polycycloaliphatic
architectures are indeed preferred anchors within this chemo-
type21 but also reinstate the biaryl portion as the major
determinant for functional efficacy.
Functional effects on CXCR3 were inspected in more detail
for key compounds 27 (no agonism) and 39 (full agonism).
Activation of CXCR3 by 39 could be inhibited by addition of
antagonist 1 and was absent in mock cells (Figure S1B,C,
Supporting Information). We also used a CRE-luciferase
reporter gene assay as an alternative to the [35S]GTPγS
functional assay. In this assay, 27 again did not show any
agonism, whereas 39 showed full agonism comparable to
peptidomimetic 2a (Figure 2D). All this unambiguously shows
that the observed functional effect by 39 is elicited through
specific interaction with the CXCR3 receptor. Collectively, our
data on this nonpeptidomimetic class of ligands suggest a biaryl
o-halogen switch from antagonism to full agonism on CXCR3,
which is not significantly influenced by subtle changes in the
bicycloaliphatic moiety.
2.2.3. QM- and NOESY NMR-Based Studies. The SFR study
reveals a link between the size and the location of the halogen
substituent and the efficacy of the compound (e.g., 5, 6, 31, 38,
and 39). As such, it is tempting to (partially) attribute this
agonist switch to the change of the dihedral biaryl angle. To
validate this hypothesis, QM calculations were performed to
derive the optimal dihedral angle for each of the biaryl systems
by performing a potential energy scan (key structures in Table
3, full listing in Table S1 and Figure S3, Supporting
Information). Results from these simulations correspond to
those from similar studies performed on biaryl systems.22,23
The angle of unsubstituted and meta-substituted compounds
remains at ca. 40° (25, 27, 40). Due to steric hindrance,
ortho,ortho-disubstituted compounds possess a typical dihedral
angle of 90° (30, 33, 35), whereas the ortho-monosubstituted
compounds possess a dihedral angle in the range of 45−70°
depending on the size of the substituent (5, 6, 29, 34, 36, 38,
39). To correlate these QM-derived angles to experimentally
determined angles, we resorted to quantitative NOESY NMR
on selected compounds. Larger dihedral angles between the
aromatic rings will result in a larger distance between the
closest set of ortho-protons on adjacent rings (designated H′
and H″, Table 4 and Figure S2A, Supporting Information). An
indication for mean interatom distances can be obtained using
E
dx.doi.org/10.1021/jm301240t | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
considered a reasonable approach for approximating the mean
calculated angles. The distance Hd−He was measured as a
control, as we expected Y not to have a significant influence on
this distance. Selected ortho-substituted compounds with
varying efficacies were subjected to NOESY NMR in CDCl3
(5, 6, 25, 29, 34, and 38). Severe overlap of 1H NMR signals in
the aromatic region prohibited the inspection of key compound
39. A representative assignment and integration with
compound 38 is shown in the Supporting Information (Figure
S10). The results (Table 4) confirm that the reference Hd−He
distance does not change with varying Y, while the H′−H″
distance depends on the size of the Y group: the larger Y
groups induce a larger distance and hence a larger angle. The
correlation between calculated and experimental data (Figure
S2B) is good with minor deviations for the larger angles
(interference of H′−H‴, vide supra). This indicative
correlation between theory and experiment shows that the
calculated angles are representative of the actual dihedral angles
under the measurement conditions and strengthens the
confidence in the quantum mechanics calculations for the
biaryl compounds.
Table 2. Exploration of the Bicycloaliphatic Group
With the confirmed calculated angles at hand, it is evident
that, on average, ligands with a considerable efficacy (α > 0.5)
seem to require a dihedral angle of around 60° (Figure 3).
However, from the experimental results it is clear that the angle
cannot be the sole determinant for agonism (e.g., compare 6 vs
31 or 6 vs 34). More particularly, the biaryl ortho-switch most
likely has dual contributions from the spatial arrangement as
well as from the electronic properties of the biaryl moiety. The
possibility of a halogen bond, known to be especially applicable
to the bigger halogens, also emerges in these consider-
ations.27,28 As such, the electrostatic potential surfaces were
calculated for the biaryl portion of the compounds (key
structures in Table 3, full listing in Table S1, Supporting
Information) as well as the electrostatic potential of the ortho-
substituent. On the inner aryl ring, the electronegativity of the
ortho-substituent influences the electrostatic potential close to
the halogen (top views of 25 vs 27 and 38 vs 40) due to the
proximity of the halogen p-orbital and the inner aryl ring at the
optimal angle. On the outer aryl ring, the electrostatic potential
(ESP) of the compounds depends on the amount of halogen
atoms attached to the aryl ring. All compounds with only one
halogen atom (5, 6, 27, 38, 39, and 40) possess an ESP of
reciprocal magnitude (yellow surface), and all compounds with
two halogen atoms (28, 29, and 30) possess an ESP of
reciprocal magnitude (green surface). However, the value for
the latter case (two halogen atoms) on the center of the aryl
ring is more positive due to the electron-withdrawing properties
of the additional halogen atom. Finally, the o-F group possesses
a full negative hemisphere, whereas the heavier halogens
introduce a positive potential (the so-called σ-hole28) at the
extension of the C−X axis (Table 3). With the increase of the
halogen atom, the σ-hole becomes progressively stronger (5 < 6
< 38 < 39). Interestingly, this trend corresponds to the order of
efficacies of these compounds (Table 3), suggesting another
contributor to the agonism switch.
a
Measured by [125I]CXCL10 displacement binding experiments with
membranes prepared from HEK293 cells stably expressing the CXCR3
b
receptor. Measured by a [35S]GTPγS functional assay with
membranes prepared from HEK293 cells stably expressing the
CXCR3 receptor. α represents the relative efficacy of a ligand
compared to the endogenous agonist CXCL11 (which is set at α =
1.0). Could not be determined due to the too low functional assay
c
d
window for this low-efficacy compound. All values shown are the mean
SEM of at least three independent experiments.
quantitative NOESY NMR through an r−6 dependence,24,25 and
use of an absolute calibration distance elsewhere in the
molecule allows reasonable approximation of unknown
distances. For distance calibration we used the normalized
sum of the Ha−Hc and Hb−Hc NOESY signals (1H NMR shifts
Ha and Hb overlap), since these are far away from the Y group
and the rigidity of the myrtenyl group bodes well for using
distances extracted from a crystal structure of an unrelated
myrtenyl-containing compound.26 The NOESY signal between
the ortho-protons is a sum of the H′−H″ and H′−H‴ NOESY
effects, but at angles substantially smaller than 90° the r−6
dependence predicts that the measured signal can be attributed
almost exclusively to the NOESY signal between the closest set
of ortho-protons H′ and H″. Indeed, our calculations (Table 4)
suggest that the deviation in the experimentally determined
H′−H″ distance as a result of interference of the more distant
H′−H‴ pair does not exceed 3.5% for the set of selected
compounds, meaning that the NOESY measurements are
We compared the efficacy (α) with both the dihedral angle
(φ, deg) of the optimal ligand conformation and the ESP (kcal/
mol) at the van der Waals radius of the o-halogen (Figure 4A).
A regression analysis, shown in Figure 4B, gave a significant
correlation with α = 0.024(ESP) − 0.010φ + 0.979 (R = 0.89,
R2 = 0.79, Radj = 0.71, SE = 0.17).
F
dx.doi.org/10.1021/jm301240t | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 3. Electron Density Surfaces Mapped with the Electrostatic Potential for Selected Compounds
Table 4. Experimental and Calculated Distances of Several Proton Pairs Based on QM Calculations and on Quantitative
a
NOESY NMR Experiments
Hd−He
H′−H″
H′−H″
H′−H‴
compd no.
Y
(Å) (NOESY)
(Å) (NOESY)
(Å) (QM)
(Å) (QM)
5
6
F
2.34
2.32
2.33
2.33
2.32
2.33
2.37
2.58
2.40
2.66
2.65
2.63
2.40
2.68
2.40
2.69
2.72
2.76
4.41
4.16
4.42
4.16
4.12
4.09
Cl
H
25
29
34
38
2,5-di-Cl
Me
Br
a
The graph depicts the hydrogen atom correlations used.
The results for the SFR, electrostatic potential surfaces and
stabilizing an active state of the CXCR3 receptor. The current
case arguably illustrates the concept of “mechanism cliffs”
(recently discussed29) with subtle structural changes inducing
switches in functional efficacies of small-molecule ligands.
regression analysis collectively demonstrate that the o-halogen
may be involved in halogen bonding and that this may play a
role in the switch from antagonism to full agonism by possibly
G
dx.doi.org/10.1021/jm301240t | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
enlargement of the halogen atom size on the ortho-position
of the outer ring from F to Br (38) or I (39) furnishes full
CXCR3 agonists, whereas moving the halogen group to the
meta-position completely abolishes agonism. In contrast,
modifications in the bicycloaliphatic group only had relatively
minor effects on the agonist efficacies. QM and NOESY NMR
were used to interrogate these phenomena. An important role
is suggested for an optimal dihedral angle within the biaryl
system of around 60°, and halogen bonding is hypothesized to
be a key player in explaining the intriguing halogen effect.
Collectively, we have described how an ensemble of chemical
tools can be applied to explain a relevant chemical biology
event. Practically, we present 38 (VUF11222) and 39
(VUF11418) as useful tools in CXCR3 research.
Figure 3. Ligands with considerable efficacy (greater than 0.5) have an
angle of 60° on average. The calculated optimal angle between the two
aryl rings is plotted against the efficacy measured in [35S]GTPγS
functional assays. The filled black squares highlight ligands that show
an efficacy greater than 0.5, whereas the filled gray circles indicate
ligands with an efficacy lower than 0.5. Data points are accompanied
by the corresponding compound codes.
3. EXPERIMENTAL SECTION
3.1. Pharmacological Assays. 3.1.1. General Information. All
cell culture materials were purchased from PAA Laboratories GmbH
(Pasching, Austria). [125I]CXCL10 radioligand (2200 Ci/mmol) was
obtained from Perkin-Elmer Life and Analytical Sciences (Boston,
MA). Unlabeled CXCL11 was from Peprotech (Rocky Hill, NJ). All
mentioned chemicals were purchased from Sigma-Aldrich (St. Louis,
MO).
3.1.2. Cell Culture and Transfection. HEK293 cells stably
expressing CXCR3 were a kind gift from Dr. K. Biber (University
Medical Center, Groningen, The Netherlands) and were cultured as
described previously.13 In the case of the CRE-luciferase reporter gene
assay, the cells were transfected with plasmid containing the reporter
gene CRE-luciferase as described in a previous report from our
laboratory.35
3.1.3. Membrane Preparation and Chemokine Binding Assay.
Membrane preparation and competition radioligand bindings were
performed as described previously13 with the exception that
membranes from HEK293 cells stably expressing CXCR3 were used.
3.1.4. [35S]GTPγS Functional Assay. The protocol used was as
described previously.13 Briefly, 5 μg of membranes per well prepared
from HEK293 cells stably expressing CXCR3 were incubated with
CXCL11 or compounds in assay buffer containing 50 mM HEPES, 10
mM MgCl2, and 100 mM NaCl, pH 7.2, supplemented with 3 μM
GDP and 300 pM [35S]GTPγS (Perkin-Elmer) for 1 h at room
While this phenomenon appears to be more common for
aminergic class A GPCRs,30 mechanism cliffs for chemokine
receptors are less frequently described.31 Notably, for CCR3
and for CCR5 a few subtle activation triggers have been
disclosed.32−34 The structural class described in the present
study includes a trigger that involves the mere size of a single
atom (from F to Br and I) and can therefore be considered as
displaying a very steep mechanism cliff for the chemokine
CXCR3 receptor.
2.3. Conclusion. The molecular details of activation of
peptide GPCRs by small molecules remain of both fundamental
and practical relevance. Here, we describe the first reported
nonpeptidomimetic agonists for the CXCR3 receptor. A total
of 26 compounds, all possessing a bicycloaliphatic group and
biaryl system linked through a quaternary ammonium cation,
were prepared by versatile synthetic routes. A remarkably subtle
trigger for receptor activation is present on the outer aryl ring
of the biaryl system, and this allows for a full spectrum of
functional efficacies to be achieved. Most notably, mere
Figure 4. (A) Representation of the dihedral angle φ and the ESP at the extension of the C−X bond shown for 39. (B) Predictive value of the
regression analysis of the dihedral angle and the ESP to the experimental efficacy. Compound numbers are indicated for each individual data point in
the figure.
H
dx.doi.org/10.1021/jm301240t | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
temperature before harvesting of the membranes on Unifilter GF/B
plates. [35S]GTPγS incorporation was determined using the Microbe-
ta. When the assay was run in antagonist mode, the compound was
added 30 min prior to addition of [35S]GTPγS. The response was
normalized to that of the endogenous agonist CXCL11 (set to α = 1).
As such, this normalized response represents the relative efficacy of a
given ligand presented as the α value.
an Xbridge (C18) 5 μm column (50 mm × 4.6 mm or 100 mm × 4.6
mm). The four eluent programs used (acidic and basic) are listed in
detail in the Supporting Information. Compound purities were
calculated as the percentage peak area of the analyzed compound by
UV detection at (unless mentioned otherwise) 230 nm. Unless
reported otherwise, all compounds have a purity ≥95% as measured by
these LC−MS analyses.
3.1.5. CRE-Luciferase Reporter Gene Assay. This assay was
performed as described previously.35 In brief, culture medium was
aspirated and replaced with serum-free medium supplemented with
0.05% BSA and ligands. After 5 h, stimulation medium was replaced by
25 μL of substrate solution (39 mM Tris−H3PO4, pH 7.8, 39%
glycerol, 2.6% Triton X-100, 860 μM dithiothreitol, 18 mM MgCl2,
825 μM ATP, 77 μM disodium pyrophosphate, 230 μg/mL beetle
luciferin), and luminescence was measured using a Victor3 plate reader
(Perkin-Elmer).
3.3.2. General Procedure A1. A microwave vial was charged with
the aryl bromide (0.90 mmol), the boronic acid (RB(OH)2, 1.80
mmol), toluene (3 mL), i-PrOH (0.8 mL), and a 2.0 M Na2CO3
solution (0.50 mL). This mixture was degassed for 15 min by bubbling
N2 through the mixture using a needle. Next Pd(PPh3)4 (52 mg, 0.045
mmol) was added to the mixture, and the vial was immediately capped
and again degassed for 5 min. The vial was heated in the microwave at
130 °C for (unless mentioned otherwise) 90 min. After this, the
reaction mixture was filtered over Celite, and most of the organic
solvents were evaporated in vacuo. The pH of the residue was adjusted
to >11 with a 1.0 M aq NaOH solution, and the mixture was extracted
with DCM (3×). The combined DCM layers were washed with brine,
dried over Na2SO4, filtered, and concentrated in vacuo. This afforded
the crude product, which was purified by column chromatography.
3.3.3. General Procedure A2. A microwave vial was charged with
the aryl bromide (free base or salt, 1.0 mmol), the boronic acid
(RB(OH)2, 1.1 equiv), Na2CO3 (3.0 equiv in the case of aryl bromide
salt, 2.0 equiv in the case of aryl bromide free base), DME (15 mL),
and water (5 mL). This mixture was degassed for 15 min by bubbling
N2 through the mixture using a needle. Next Pd(PPh3)4 (0.05 equiv)
was added to the mixture, and the vial was immediately capped and
again degassed for 5 min. The vial was heated in the microwave at 120
°C for (unless mentioned otherwise) 60 min. After this, the reaction
mixture was diluted with DCM and 2.0 M aq Na2CO3. Extraction was
performed twice with DCM. The combined organic layers were dried
over Na2SO4, filtered, and concentrated in vacuo. This afforded the
crude product, which was purified by column chromatography.
3.3.4. General Procedure B. The amine (given amount) and
aldehyde (1.0 equiv) were dissolved in DCE (given amount). The
reaction mixture was stirred at rt for 30 min. Then NaBH(OAc)3
(given amount) was added. After being stirred overnight (unless
mentioned otherwise) at rt, the reaction was quenched with aq
Na2CO3 (2.0 M). The mixture was stirred for 30 min at rt. The layers
were separated, and the aqueous layer was extracted twice with DCM.
The combined organic layers were dried over Na2SO4, filtered, and
concentrated in vacuo to give the crude product, which was purified by
column chromatography.
3.1.6. Data Analysis. Nonlinear regression analysis of the data and
calculation of pEC50 and pKi values were performed using Prism 5.0d
(GraphPad Software Inc., San Diego, CA) as described before.13
3.2. Computational Methods. 3.2.1. Quantum Mechanical
Calculations. Quantum mechanical calculations were performed with
Gaussian 0336 using the Becke3LYP functional and the 6-311+G*
basis set. For iodine, an additional SDD basis set was used. To alleviate
computational complexity, only the biaryl portion was used, whereas
the rest of the molecule was omitted from the calculations as this is not
expected to influence the electron density distribution of the outer
substituted phenyl ring. As such, for the remaining biaryl model
system, different conformations were generated by performing a
potential energy scan for the dihedral angle between the two phenyl
rings. Due to ligand symmetry, calculations were performed for a
dihedral angle between 0° and 180° with an interval of 30° (Figure S3,
Supporting Information). Potential energy differences were calculated,
and the lowest energy conformation was optimized with Gaussian to
obtain the corresponding optimal dihedral angle. Subsequently, the
ESP of the ortho-substituent was determined at the van der Waals
radius of the halogen atom in the extension of the C−X bond.
Distances between the ortho-hydrogen atoms of the phenyl rings for
selected compounds were experimentally approximated by NOESY for
the verification of our calculations (see the main text).
3.2.2. Data Regression. A 2D quantitative structure−activity
relationship (QSAR) model was constructed using the ESP value
and the dihedral angle of the optimized ligand conformations for a
multiple linear regression to the efficacy. Data regression was
performed using SPSS (SPSS Inc., Chicago, IL, 2005; http://spss.
com).
3.3.5. General Procedure C. In a round-bottom flask, the tertiary
amine precursor (0.25 mmol) was dissolved in DCM (6.4 mL). The
flask was placed under a nitrogen atmosphere, closed with a septum,
and covered with aluminum foil. Then excess MeI (given amount) was
added to the solution via a syringe. The reaction mixture was allowed
to stir at rt in the dark. TLC analysis was used to monitor the progress
of the reaction. If required, additional quantities of MeI were added
followed by prolonged stirring. Such cycles were applied until
completion of the reaction. The workup (unless mentioned otherwise)
was as follows. The reaction mixture was cooled in an ice bath, and 3
vol equiv of MTBE was added to the reaction mixture slowly via a
dropping funnel while the mixture was stirred. This usually caused
precipitation of the salt. The precipitate was filtered and washed with a
precooled solution of DCM/MTBE (1/3). Typically, this delivers the
product salt in high purity. If required, the salt was either
reprecipitated from DCM/MTBE as just described or recrystallized
from hot DCM/MTBE. Occasionally (for example, in the case of 28
and 33), in the initial precipitation with MTBE the salt does not
precipitate as a solid but as a sticky oil. In those cases, the following
procedure was used. The solution was concentrated in vacuo. A small
amount of DCM was added, the mixture was gently warmed, and
enough MTBE was added dropwise until slight turbidity appeared.
Cooling to rt and then to −30 °C yielded a precipitate. This solid was
filtered and washed with a precooled solution of DCM/MTBE (1/3).
All final ammonium salts were thoroughly dried under high vacuum
3.3. Synthetic Methods. 3.3.1. General Information. 2-
Hydroxybenzeneboronic acid was from Alfa Aesar. All other chemicals
were from Sigma-Aldrich. THF, toluene, and CH2Cl2 were distilled
freshly from CaH2; all other solvents were used as received.
Compound 2a was prepared as described by us.13 Unless indicated
otherwise, all reactions were carried out under an inert atmosphere.
TLC analyses were performed with Merck F254 alumina silica plates
using UV visualization or staining. Microwave reactions were carried
out on a Biotage initiator. Column purifications were carried out
manually using Silicycle UltraPure silica gel or automatically using the
Biotage equipment. All HRMS spectra were recorded on a Bruker
micrOTOF mass spectrometer using ESI in positive ion mode.
Elemental analysis was carried out at Mikroanalytisches Labor Pascher
(Remagen, Germany). The 1H, 13C, and 2D NMR spectra were
recorded on a Bruker 200, 250, 400, or 500 MHz spectrometer.
Infrared spectra were recorded on a Galaxy series FT-IR 6030
instrument. Systematic names for the molecules were generated using
ChemBioDraw Ultra 11. Melting points were taken using the Stanford
Research Systems Optimelt apparatus, and the values given are
uncorrected. Optical rotations were measured with a sodium lamp and
are reported as follows: [α]2D3 (c, g/100 mL, solvent). Purities were
measured with the aid of analytical LC−MS using a Shimadzu LC-
20AD liquid chromatograph pump system with a Shimadzu SPD-
M20A diode array detector with the MS detection performed with a
Shimadzu LCMS-2010EV mass spectrometer. The column used was
I
dx.doi.org/10.1021/jm301240t | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
and stored in the dark. Even then, some salts tended to retain DCM
traces as crystal solvates, but these are considered noninterfering.
3.4. Representative Synthetic Examples. 3.4.1. 1-(2′-Bromo-
biphenyl-4-yl)-N-(((1R,5S)-6,6-dimethylbicyclo[3.1.1]hept-2-en-2-yl)-
methyl)-N-methylmethanamine (58). A vial was charged with
aldehyde 14 (150 mg, 0.57 mmol), amine 7 (95 mg, 0.58 mmol),
NaBH(OAc)3 (180 mg, 0.85 mmol), and DCE (3 mL) under a
nitrogen atmosphere. The mixture was stirred for 16 h at rt. Then satd
aq NaHCO3 (3 mL) was added while the mixture was stirred. After 5
min, the mixture was diluted with DCM and water. Extraction was
performed twice with DCM. The combined organic layers were dried
over Na2SO4, filtered, and concentrated under vacuum. The crude
product was purified by column chromatography (n-heptane/EtOAc/
TEA, 100/0/1 to 80/20/1) to yield a colorless oil (115 mg, 49%). 1H
NMR (CDCl3, 400 MHz): δ 7.66 (d, 1H, J = 7.9 Hz), 7.37−7.31 (m,
6H), 7.21−7.16 (m, 1H), 5.44 (br s, 1H), 3.59 (d, 1H, J = 13.4 Hz),
3.41 (d, 1H, J = 13.4 Hz), 2.97 (d, 1H, J = 13.2 Hz), 2.86 (d, 1H, J =
13.2 Hz), 2.47−2.41 (m, 1H), 2.36−2.21 (m, 3H), 2.17 (s, 3H), 2.15−
2.09 (m, 1H), 1.31 (s, 3H), 1.18 (d, 1H, J = 8.6 Hz), 0.85 (s, 3H). 13C
NMR (CDCl3, 125 MHz): δ 146.4, 142.5, 139.5, 139.1, 133.1, 131.3,
129.2, 128.5, 127.3, 122.6, 119.7, 63.8, 61.3, 44.2, 42.6, 40.9, 37.9, 31.8,
31.4, 26.3, 21.1 (two aromatic carbon signals overlap). HR-MS (m/z):
[M + H]+ for C24H29BrN, calcd 410.1483, found 410.1478 (doublet,
peak at 412 also visible). LC purity: 99+%.
according to the general methylation procedure (procedure C) using
59 (85 mg, 0.19 mmol) and MeI (300 μL, 4.82 mmol). The crude
product was recrystallized from warm MTBE/DCM (3/1). This gave a
white solid (80 mg, 70%). Part of this was recrystallized from warm
MTBE/DCM (2/1) to give a white solid (33 mg) which retained trace
DCM solvate (2.5 mol %) even after prolonged drying. This material
was used for pharmacological studies. 1H NMR (CDCl3, 500 MHz): δ
7.95 (dd, 1H, J = 8.0, 1.0 Hz), 7.74 (d, 2H, J = 8.2 Hz), 7.43−7.38 (m,
3H), 7.25 (dd, 1H, J1 obscured, J2 = 1.7 Hz), 7.07 (app dt, 1H), 6.29
(br s, 1H), 5.13 (s, 2H), 4.49 (d, 1H, J = 12.4 Hz), 4.34 (d, 1H, J =
12.4 Hz), 3.19 (s, 3H), 3.18 (s, 3H), 2.57−2.51 (m, 1H), 2.48−2.35
(m, 3H), 2.20−2.15 (m, 1H), 1.32 (s, 3H), 1.18 (d, 1H, J = 8.8 Hz),
0.88 (s, 3H). 2D NOESY NMR (CDCl3, 500 MHz): see the
Supporting Information. 13C NMR (CDCl3, 125 MHz): δ 146.5,
145.0, 139.7, 136.9, 136.1, 132.9, 130.2, 129.8, 129.4, 128.3, 126.4,
98.1, 69.3, 66.9, 49.0, 48.8, 47.0, 39.6, 38.1, 32.2, 32.0, 25.9, 21.5. HR-
MS (m/z): [M]+ for C25H31IN, calcd 472.1496, found 472.1480. LC
purity: 99+ %. The Supporting Information contains hardcopies of
selected analytical data for 39.
ASSOCIATED CONTENT
* Supporting Information
■
S
Supporting data on CXCR3 activation by key compounds, on
QM calculations, and on NOESY NMR studies, procedures for
NOESY NMR studies, syntheses, chromatographic details, and
characterization of all compounds, representative 1D, 2D, and
high-temperature NMR spectra, and selected LC chromato-
grams. This material is available free of charge via the Internet
at http://pubs.acs.org.
3.4.2. 1-((1R,5S)-6,6-Dimethylbicyclo[3.1.1]hept-2-en-2-yl)-N-((2′-
iodobiphenyl-4-yl)methyl)-N-methylmethanamine (59). A vial was
charged with aldehyde 15 (75 mg, 0.24 mmol), amine 7 (40 mg, 0.24
mmol), NaBH(OAc)3 (80 mg, 0.38 mmol), and DCE (5 mL) under a
nitrogen atmosphere. The mixture was stirred for 16 h at rt. Then satd
aq NaHCO3 (3 mL) was added while the mixture was stirred. After 5
min, the mixture was diluted with DCM and water. Extraction was
performed twice with DCM. The combined organic layers were dried
over Na2SO4, filtered, and concentrated under vacuum. The crude
product was purified by column chromatography (n-heptane/EtOAc/
AUTHOR INFORMATION
Corresponding Author
*Phone: +31 20 59 87579. Fax: +31 20 598 76 10. E-mail: r.
leurs@vu.nl.
■
1
TEA, 100/0/1 to 80/20/1) to yield a colorless oil (85 mg, 77%). H
NMR (CDCl3, 400 MHz): δ 7.92 (d, 1H, J = 8.1 Hz), 7.37−7.22 (m,
6H), 7.01−6.96 (m, 1H), 5.42 (br s, 1H), 3.56 (d, 1H, J = 13.3 Hz),
3.38 (d, 1H, J = 13.3 Hz), 2.94 (d, 1H, J = 13.1 Hz), 2.83 (d, 1H, J =
13.1 Hz), 2.44−2.37 (m, 1H), 2.33−2.18 (m, 3H), 2.15 (s, 3H), 2.12−
2.06 (m, 1H), 1.28 (s, 3H), 1.16 (d, 1H, J = 8.6 Hz), 0.83 (s, 3H).13C
NMR (CDCl3, 125 MHz): δ 146.6, 146.4, 142.6, 139.5, 139.1, 130.1,
129.0, 128.6, 128.5, 128.1, 119.8, 98.7, 63.8, 61.4, 44.2, 42.6, 40.9, 37.9,
31.8, 31.4, 26.3, 21.1. HR-MS (m/z): [M + H]+ for C24H29IN, calcd
458.1339, found 458.1329. LC purity: 98%.
Present Address
∥Drug Discovery Biology, Monash Institute of Pharmaceutical
Sciences, 399 Royal Parade, Parkville, Victoria 3052, Australia.
Author Contributions
§These authors contributed equally to this work.
Notes
The authors declare no competing financial interest.
3.4.3. 1-(2′-Bromobiphenyl-4-yl)-N-(((1R,5S)-6,6-dimethylbicyclo-
[3.1.1]hept-2-en-2-yl)methyl)-N,N-dimethylmethanaminium Iodide
(38). The experimental procedure for this compound was performed
according to the general methylation procedure (procedure C) using
58 (100 mg, 0.24 mmol) and MeI (300 μL, 4.82 mmol). The crude
compound was reprecipitated according to the protocol described in
procedure C. This afforded a white powder (125 mg, ca. 94%). Part of
this product was recrystallized from warm MTBE/DCM. This gave a
white solid (65 mg) which retained trace DCM solvate even after
prolonged drying. This material was used for pharmacological studies.
1H NMR (CDCl3, 500 MHz): δ 7.73 (d, 2H, J = 8.2 Hz), 7.67 (dd,
1H, J = 8.0, 1.1 Hz), 7.49 (d, 2H, J = 8.2 Hz), 7.37 (app dt, 1H), 7.27
(dd, 1H, J = 7.6, 1.7 Hz), 7.23 (app dt, 1H), 6.28 (br s, 1H), 5.13 (s,
2H), 4.48 (d, 1H, J = 12.4 Hz), 4.32 (d, 1H, J = 12.4 Hz), 3.19 (s, 3H),
3.17 (s, 3H), 2.56−2.50 (m, 1H), 2.47−2.34 (m, 3H), 2.19−2.14 (m,
1H), 1.31 (s, 3H), 1.17 (d, 1H, J = 8.8 Hz), 0.87 (s, 3H). 2D NOESY
NMR (CDCl3, 500 MHz): see the Supporting Information. 13C NMR
(CDCl3, 125 MHz): δ 143.5, 141.0, 136.9, 136.1, 133.3, 133.0, 131.1,
130.3, 129.4, 127.6, 126.4, 122.2, 69.1, 66.8, 49.0, 48.8, 47.0, 39.7, 38.1,
32.1, 32.0, 25.9, 21.5. HR-MS (m/z): [M]+ for C25H31BrN, calcd
424.1634, found 424.1629 (doublet, peak at 426 also visible). LC
purity: 99+ %. The Supporting Information contains hardcopies of
selected analytical data for 38.
ABBREVIATIONS USED
■
α, efficacy; DCE, 1,2-dichloroethane; DME, dimethoxyethane;
ESP, electrostatic potential; GPCR, G protein-coupled
receptor; GTPγS, guanosine 5′-O-[γ-thio]triphosphate;
NOESY, nuclear Overhauser effect spectroscopy; QM,
quantum mechanics; rt, room temperature; SAR, structure−
activity relationship; SFR, structure−function relationship
REFERENCES
■
(1) Overington, J. P.; Al-Lazikani, B.; Hopkins, A. L. How many drug
targets are there? Nat. Rev. Drug Discovery 2006, 5, 993−996.
(2) Scholten, D.; Canals, M.; Maussang, D.; Roumen, L.; Smit, M.;
Wijtmans, M.; de Graaf, C.; Vischer, H.; Leurs, R. Pharmacological
modulation of chemokine receptor function. Br. J. Pharmacol. 2012,
165, 1617−1643.
(3) Viola, A.; Luster, A. D. Chemokines and their receptors: drug
targets in immunity and inflammation. Annu. Rev. Pharmacol. Toxicol.
2008, 48, 171−197.
(4) Medina, J. C.; Johnson, M. G.; Collins, T. L. CXCR3 antagonists.
Annu. Rep. Med. Chem. 2005, 40, 215−225.
(5) Wijtmans, M.; Verzijl, D.; Leurs, R.; de Esch, I. J.; Smit, M. J.
Towards small-molecule CXCR3 ligands with clinical potential.
ChemMedChem 2008, 3, 861−872.
3.4.4. 1-((1R,5S)-6,6-Dimethylbicyclo[3.1.1]hept-2-en-2-yl)-N-((2′-
iodobiphenyl-4-yl)methyl)-N,N-dimethylmethanaminium Iodide
(39). The experimental procedure for this compound was performed
J
dx.doi.org/10.1021/jm301240t | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
(6) Wijtmans, M.; de Esch, I. J. P.; Leurs, R. In Chemokine Receptors
as Drug Targets; Smit, M. J., Lira, S. A., Leurs, R., Eds.; Wiley-VCH:
Weinheim, Germany, 2011; pp 301−315.
groups in the CXCR3 receptor with the aid of a modular synthetic
strategy. Bioorg. Med. Chem. Lett. 2009, 19, 2252−2257.
(22) Grein, F. Twist angles and rotational energy barriers of biphenyl
and substituted biphenyls. J. Phys. Chem. A 2002, 106, 3823−3827.
(23) Karpfen, A.; Choi, C. H.; Kertesz, M. Single-bond torsional
potentials in conjugated systems: a comparison of ab initio and density
functional results. J. Phys. Chem. A 1997, 101, 7426−7433.
(24) Lipsitz, R. S.; Tjandra, N. Residual dipolar couplings in NMR
structure analysis. Annu. Rev. Biophys. Biomol. Struct. 2004, 33, 387−
413.
(7) Heise, C. E.; Pahuja, A.; Hudson, S. C.; Mistry, M. S.; Putnam, A.
L.; Gross, M. M.; Gottlieb, P. A.; Wade, W. S.; Kiankarimi, M.;
Schwarz, D.; Crowe, P.; Zlotnik, A.; Alleva, D. G. Pharmacological
characterization of CXC chemokine receptor 3 ligands and a small
molecule antagonist. J. Pharmacol. Exp. Ther. 2005, 313, 1263−1271.
(8) O’Boyle, G.; Fox, C. R.; Walden, H. R.; Willet, J. D.; Mavin, E. R.;
Hine, D. W.; Palmer, J. M.; Barker, C. E.; Lamb, C. A.; Ali, S.; Kirby, J.
A. Chemokine receptor CXCR3 agonist prevents human T-cell
migration in a humanized model of arthritic inflammation. Proc. Natl.
Acad. Sci. U.S.A. 2012, 109, 4598−4603.
(25) Anet, F. A. L.; Bourn, A. J. R. Nuclear magnetic resonance
spectral assignments from nuclear Overhauser effects. J. Am. Chem. Soc.
1965, 87, 5250−5251.
(26) Veyrat, M.; Maury, O.; Faverjon, F.; Over, D. E.; Ramasseul, R.;
Marchon, J.-C.; Turowska-Tyrk, I.; Scheidt, W. R. Chiroporphyrins: an
approach to asymmetric catalysts with stereocenters near the plane of
the porphyrin ring. Angew. Chem., Int. Ed. Engl. 1994, 33, 220−223.
(27) Hardegger, L. A.; Kuhn, B.; Spinnler, B.; Anselm, L.; Ecabert, R.;
Stihle, M.; Gsell, B.; Thoma, R.; Diez, J.; Benz, J.; Plancher, J. M.;
Hartmann, G.; Banner, D. W.; Haap, W.; Diederich, F. Systematic
investigation of halogen bonding in protein-ligand interactions. Angew.
Chem., Int. Ed. 2011, 50, 314−318.
(9) Yates, C. C.; Krishna, P.; Whaley, D.; Bodnar, R.; Turner, T.;
Wells, A. Lack of CXC chemokine receptor 3 signaling leads to
hypertrophic and hypercellular scarring. Am. J. Pathol. 2010, 176,
1743−1755.
(10) Stroke, I. L.; Cole, A. G.; Simhadri, S.; Brescia, M. R.; Desai, M.;
Zhang, J. J.; Merritt, J. R.; Appell, K. C.; Henderson, I.; Webb, M. L.
Identification of CXCR3 receptor agonists in combinatorial small-
molecule libraries. Biochem. Biophys. Res. Commun. 2006, 349, 221−
228.
(11) Anghelescu, A. V.; DeLisle, R. K.; Lowrie, J. F.; Klon, A. E.; Xie,
X.; Diller, D. J. Technique for generating three-dimensional alignments
of multiple ligands from one-dimensional alignments. J. Chem. Inf.
Model. 2008, 48, 1041−1054.
(12) Nedjai, B.; Li, H.; Stroke, I. L.; Wise, E. L.; Webb, M. L.;
Merritt, J. R.; Henderson, I.; Klon, A. E.; Cole, A. G.; Horuk, R.;
Vaidehi, N.; Pease, J. E. Small molecule chemokine mimetics suggest a
molecular basis for the observation that CXCL10 and CXCL11 are
allosteric ligands of CXCR3. Br. J. Pharmacol. 2012, 166, 912−923.
(13) Scholten, D.; Canals, M.; Wijtmans, M.; de Munnik, S.; Nguyen,
P.; Verzijl, D.; de Esch, I.; Vischer, H.; Smit, M.; Leurs, R.
Pharmacological characterisation of a small-molecule agonist for the
chemokine receptor CXCR3. Br. J. Pharmacol. 2012, 166, 898−911.
(14) Wu, B.; Chien, E. Y. T.; Mol, C. D.; Fenalti, G.; Liu, W.;
Katritch, V.; Abagyan, R.; Brooun, A.; Wells, P.; Bi, F. C.; Hamel, D. J.;
Kuhn, P.; Handel, T. M.; Cherezov, V.; Stevens, R. C. Structures of the
CXCR4 chemokine GPCR with small-molecule and cyclic peptide
antagonists. Science 2010, 330, 1066−1071.
(15) Roumen, L.; Sanders, M. P. A.; Vroling, B.; De Esch, I. J. P.; De
Vlieg, J.; Leurs, R.; Klomp, J. P. G.; Nabuurs, S. B.; De Graaf, C. In
silico veritas: the pitfalls and challenges of predicting GPCR-ligand
interactions. Pharmaceuticals 2011, 4, 1196−1215.
(16) Kufareva, I.; Rueda, M.; Katritch, V.; Stevens, R. C.; Abagyan, R.
Status of GPCR modeling and docking as reflected by community-
wide GPCR Dock 2010 assessment. Structure 2011, 19, 1108−1126.
(17) Wijtmans, M.; Verzijl, D.; Bergmans, S.; Lai, M.; Bosch, L.; Smit,
M. J.; de Esch, I. J. Leurs, R. CXCR3 antagonists: quaternary
ammonium salts equipped with biphenyl- and polycycloaliphatic-
anchors. Bioorg. Med. Chem. 2011, 19, 3384−3393.
(28) Parisini, E.; Metrangolo, P.; Pilati, T.; Resnati, G.; Terraneo, G.
Halogen bonding in halocarbon-protein complexes: a structural survey.
Chem. Soc. Rev. 2011, 40, 2267−2278.
(29) Stumpfe, D.; Bajorath, J. Exploring activity cliffs in medicinal
chemistry. J. Med. Chem. 2012, 55, 2932−2942.
(30) Iyer, P.; Stumpfe, D.; Bajorath, J. Molecular mechanism-based
network-like similarity graphs reveal relationships between different
types of receptor ligands and structural changes that determine
agonistic, inverse-agonistic, and antagonistic effects. J. Chem. Inf.
Model. 2011, 51, 1281−1286.
(31) Wise, E.; Pease, J. E. Unravelling the mechanisms underpinning
chemokine receptor activation and blockade by small molecules: a fine
line between agonism and antagonism? Biochem. Soc. Trans. 2007, 35,
755−759.
(32) Anderskewitz, R.; Bauer, R.; Bodenbach, G.; Gester, D.;
Gramlich, B.; Morschhauser, G.; Birke, F. W. Pyrrolidinohydroquina-
zolinesa novel class of CCR3 modulators. Bioorg. Med. Chem. Lett.
2005, 15, 669−673.
(33) Ting, P. C.; Umland, S. P.; Aslanian, R.; Cao, J.; Garlisi, C. G.;
Huang, Y.; Jakway, J.; Liu, Z.; Shah, H.; Tian, F.; Wan, Y.; Shih, N. Y.
The synthesis of substituted bipiperidine amide compounds as CCR3
ligands: antagonists versus agonists. Bioorg. Med. Chem. Lett. 2005, 15,
3020−3023.
(34) Ferain, T.; Hoveyda, H.; Ooms, F.; Schols, D.; Bernard, J.;
Fraser, G. Agonist-induced internalization of CC chemokine receptor
5 as a mechanism to inhibit HIV replication. J. Pharmacol. Exp. Ther.
2011, 337, 655−662.
(35) Watts, A. O.; Scholten, D. J.; Heitman, L. H.; Vischer, H. F.;
Leurs, R. Label-free impedance responses of endogenous and synthetic
chemokine receptor CXCR3 agonists correlate with Gi-protein
pathway activation. Biochem. Biophys. Res. Commun. 2012, 419, 412−
418.
(36) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.;
Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.;
Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.;
Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J.
B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R.
E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.;
Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J.
J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.;
Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman,
J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Cliffordaaa, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
(18) Oyler, K. D.; Coughlin, F. J.; Bernhard, S. Controlling the
helicity of 2,2′-bipyridyl ruthenium(II) and zinc(II) hemicage
complexes. J. Am. Chem. Soc. 2007, 129, 210−217.
(19) Brown, H. C.; Dhokte, U. P. Hydroboration. 90. Synthesis of 2-
isobutyl- and 2-isopropylapopinenes. Rates and stoichiometry of the
hydroboration of 2-organylapopinenes with borane−methyl sulfide
and borane−tetrahydrofuran. J. Org. Chem. 1994, 59, 2025−2032.
(20) Simoni, D.; Giannini, G.; Roberti, M.; Rondanin, R.; Baruchello,
R.; Rossi, M.; Grisolia, G.; Invidiata, F. P.; Aiello, S.; Marino, S.;
Cavallini, S.; Siniscalchi, A.; Gebbia, N.; Crosta, L.; Grimaudo, S.;
Abbadessa, V.; Di Cristina, A.; Tolomeo, M. Studies on the apoptotic
activity of natural and synthetic retinoids: discovery of a new class of
synthetic terphenyls that potently support cell growth and inhibit
apoptosis in neuronal and HL-60 cells. J. Med. Chem. 2005, 48, 4293−
4299.
(21) Wijtmans, M.; Verzijl, D.; van Dam, C. M.; Bosch, L.; Smit, M.
J.; Leurs, R.; de Esch, I. J. Exploring a pocket for polycycloaliphatic
K
dx.doi.org/10.1021/jm301240t | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision
C.02; Gaussian, Inc.: Wallingford, CT, 2004.
L
dx.doi.org/10.1021/jm301240t | J. Med. Chem. XXXX, XXX, XXX−XXX