CC chemokine receptors 2 (CCR2) and 5 (CCR5) are involved in many inflammatory diseases; however, most CCR2 and CCR5 clinical candidates have been unsuccessful. (Pre)clinical evidence suggests that dual CCR2/CCR5 inhibition might be more effective in the treatment of such multifactorial diseases. In this regard, the highly conserved intracellular binding site in chemokine receptors provides a new avenue for the design of multitarget ligands. In this study, we synthesized and evaluated the biological activity of a series of triazolopyrimidinone derivatives in CCR2 and CCR5. Radioligand binding assays first showed that they bind to the intracellular site of CCR2, and in combination with functional assays on CCR5, we explored structure-affinity/activity relationships in both receptors. Although most compounds were CCR2-selective, 39 and 43 inhibited β-arrestin recruitment in CCR5 with high potency. Moreover, these compounds displayed an insurmountable mechanism of inhibition in both receptors, which holds promise for improved efficacy in inflammatory diseases.
CC chemokine receptors 2 (CCR2) and 5 (CCR5) are involved in many inflammatory diseases; however, most CCR2 and CCR5 clinical candidates have been unsuccessful. (Pre)clinical evidence suggests that dual CCR2/CCR5 inhibition might be more effective in the treatment of such multifactorial diseases. In this regard, the highly conserved intracellular binding site in chemokine receptors provides a new avenue for the design of multitarget ligands. In this study, we synthesized and evaluated the biological activity of a series of triazolopyrimidinone derivatives in CCR2 and CCR5. Radioligand binding assays first showed that they bind to the intracellular site of CCR2, and in combination with functional assays on CCR5, we explored structure-affinity/activity relationships in both receptors. Although most compounds were CCR2-selective, 39 and 43 inhibited β-arrestin recruitment in CCR5 with high potency. Moreover, these compounds displayed an insurmountable mechanism of inhibition in both receptors, which holds promise for improved efficacy in inflammatory diseases.
CC chemokine receptors 2 (CCR2) and 5
(CCR5) are two membrane-bound
G protein-coupled receptors (GPCRs), which belong to the subfamily
of chemokine receptors. Chemokine receptors are widely expressed in
leukocytes, and thus, they regulate different homeostatic and inflammatory
leukocyte functions upon interaction with their endogenous chemokines.[1,2] In general, chemokine receptors interact with multiple endogenous
chemokines, such as CCL2, CCL7, and CCL8 in the case of CCR2, and
CCL3, CCL4, and CCL5 in the case of CCR5.[1] Furthermore, most chemokines can interact with multiple chemokine
receptors, allowing for a very complex and fine-tuned system.[3,4] Dysregulation of this system has been linked to the development
of several pathophysiological conditions. For example, both CCR2 and
CCR5 have been implicated in many inflammatory and immune diseases
such as rheumatoid arthritis, multiple sclerosis, atherosclerosis,
diabetes mellitus, and psoriasis,[5,6] rendering these
proteins attractive targets for the pharmaceutical industry. As a
result, many efforts have been made to bring CCR2 and CCR5 small-molecule
antagonists into the clinic although with limited success. Only maraviroc,
an HIV-1 entry inhibitor selectively targeting CCR5, has been approved
by the FDA and EMA,[7] while all other drug
candidates have failed in clinical trials.Recently, it has
been suggested that the development of multitarget
drugs (designed to interact with multiple receptors) represents a
more effective approach in the treatment of complex multifactorial
diseases.[8,9] Thus, dual targeting of CCR2 and CCR5 emerges
as a potentially more efficacious strategy in diseases where both
receptors are involved. Indeed, combined CCR2/CCR5 inhibition has
resulted in beneficial effects in several preclinical disease models
and clinical studies, further supporting the use of dual antagonists.[10−12] In this regard, several antagonists with dual CCR2/CCR5 activity
have been reported in the past years, including the first dual antagonist
TAK-779 and the clinical candidate cenicriviroc.[13] All of these antagonists bind to the extracellular region
of CCR2 and CCR5, in a site overlapping with the chemokine’s
binding pocket.[14] Yet the crystal structures
of CCR2 and CCR9 have demonstrated that chemokine receptors can also
be targeted with intracellular allosteric modulators.[15,16] These intracellular ligands offer a number of advantages, such as
noncompetitive binding and, as a consequence, insurmountable inhibition,
which is particularly important due to the high local concentration
of chemokines during pathological conditions.[17,18] In addition, the high conservation of this intracellular site allows
for the design of multitarget antagonists.[18,19] Several high-affinity intracellular ligands have been already identified
for CCR2[20,21] but not for CCR5, although intracellular
compounds developed for CCR2 or CCR4 have been reported to bind CCR5
with much lower potency.[21,22]In the current
study we first report that previously patented CCR2
antagonists with a triazolopyrimidinone scaffold, such as compound 8 (Figure ),[23] bind to the intracellular site of
the receptor with high affinity. In addition, we show that this compound
is able to inhibit CCR5 with moderate activity, suggesting a potential
dual CCR2/CCR5 activity for this class of compounds. Thus, a series
of novel and previously reported triazolopyrimidinone derivatives
were synthesized according to published methods[23] in order to obtain structure–affinity/activity relationships
(SARs) in both CCR2 and CCR5. Radioligand binding assays and functional
assays were used to evaluate their affinity toward CCR2 and activity
toward CCR5. In addition, characterization of two selected compounds
(39 and 43) in a [35S]GTPγS
binding assay demonstrated that these compounds inhibit both receptors
in a noncompetitive, insurmountable manner. Finally, selected compounds
were docked into the CCR2 crystal structure in order to shed light
on the binding mode of these derivatives, in comparison to that of
the crystallized CCR2-RA-[R].[15] In summary, our findings provide some insight on the CCR2/CCR5
selectivity profile of triazolopyrimidinone derivatives, as
well as on the structural requirements for the design of multitarget
or selective intracellular ligands for these receptors.
Figure 1
Chemical structures
of the orthosteric CCR2/CCR5 antagonist TAK-779
and the CCR2 intracellular ligands CCR2-RA-[R], SD-24,
JNJ-27141491 and the triazolopyrimidinone derivative 8. [3H]-CCR2-RA-[R] was used in radioligand
binding assays for CCR2.
Chemical structures
of the orthosteric CCR2/CCR5 antagonist TAK-779
and the CCR2 intracellular ligands CCR2-RA-[R], SD-24,
JNJ-27141491 and the triazolopyrimidinone derivative 8. [3H]-CCR2-RA-[R] was used in radioligand
binding assays for CCR2.
Results and Discussion
Chemistry
Triazolo-pyrimidinone derivatives 6–43 were synthesized using a three-step synthesis
approach as described by Bengtsson et al.[23] (Scheme ). First,
if not commercially available, the β-keto esters 1a–n were synthesized from ethyl acetoacetate 1a and the respective bromo- or iodoalkanes 2f–h,j,k or benzyl bromide 2n. Benzylation of the β-keto esters 1a–n with the corresponding R1-substituted
benzyl bromides (3a–v), at reflux,
resulted in a series of benzylated β-keto esters 4aa–na, 4bb–bq, 4eq–ev in yields between 8% and 97% (Scheme , Table S1). Finally a cyclization reaction of the benzylated
β-keto esters 4aa–na, 4bb–bq, 4eq–ev with the commercially available 3,5-diaminotriazole 5c in ionic liquid BMIM-PF6 (1-butyl-3-methylimidazolium hexafluorophosphate)
at 200 °C under microwave irradiation resulted in final compounds 6, 9–43 in yields ranging
from 4% to 83%. Final compound 7 (R2 = H)
was synthesized using H3PO4 in ethanol conditions
and 8 (R2 = Me) in p-toluenesulfonic
acid monohydrate conditions.
Scheme 1
Synthesis Scheme of the Triazolopyrimidinone
Derivatives 6–43
Reagents and conditions:
(i)
NaH, n-BuLi, THF, overnight, 0 °C to rt (1a–e,i,l,m were commercially available); (ii) DIPEA, LiCl, THF, reflux,
overnight; (iii) (8–43, R2 = NH2) BMIM-PF6, 200 °C, 1 h or (6,
R2 = H) H3PO4, EtOH,
170 °C, 10 h or (7, R2 = Me) p-toluenesulfonic acid monohydrate, 180 °C, 30 min.
Synthesis Scheme of the Triazolopyrimidinone
Derivatives 6–43
Reagents and conditions:
(i)
NaH, n-BuLi, THF, overnight, 0 °C to rt (1a–e,i,l,m were commercially available); (ii) DIPEA, LiCl, THF, reflux,
overnight; (iii) (8–43, R2 = NH2) BMIM-PF6, 200 °C, 1 h or (6,
R2 = H) H3PO4, EtOH,
170 °C, 10 h or (7, R2 = Me) p-toluenesulfonic acid monohydrate, 180 °C, 30 min.
Biology
We have previously identified several CCR2
intracellular ligands belonging to different chemical scaffolds, such
as CCR2-RA-[R], SD-24, and JNJ-27141491 (Figure ).[20,21] In contrast to CCR2 orthosteric ligands, these intracellular ligands
lack a basic nitrogen and have lower molecular weights, unsaturated
systems with haloarenes, and acidic groups capable of forming hydrogen
bonds.[18,20] Other CCR2 antagonists with similar features
have been described in the literature, including the triazolo- or
pyrazolopyrimidinone derivatives described in two different
patents.[23,24] To test whether they also bind to the intracellular
site of the receptor, we synthesized “example 1” from
the patent by Bengtsson et al.,[23] corresponding
to the triazolopyrimidinone derivative 8 in our
study (Figure ). Using
a [3H]-CCR2-RA-[R] binding assay as previously
described,[19] we found that compound 8 fully displaced [3H]-CCR2-RA-[R] binding from U2OS cells stably expressing hCCR2b (U2OS-CCR2) with
high affinity and a pseudo-Hill slope (nH) close to unity, indicating a competitive interaction with [3H]-CCR2-RA-[R] for the intracellular binding
site. 8 displaced [3H]-CCR2-RA-[R] with a pKi of 8.90 ± 0.04 (Ki = 1.3 nM, Figure a and Table ), consistent with its previously reported activity
in a CCR2calcium flux assay (IC50 = 16 nM).[23]
Figure 2
Characterization of ligands in U2OS-CCR2 and U2OS-CCR5.
(a) [3H]-CCR2-RA-[R] displacement by increasing
concentrations of triazolopyrimidinone derivatives 8, 39, and 43 in U2OS-CCR2 at 25 °C.
Data are normalized to specific binding in the absence of compound
(set as 100%). (b) Inhibition of CCL2-stimulated β-arrestin
recruitment in U2OS-CCR2 by increasing concentrations of compounds 8, 39, and 43, after stimulation
with an EC80 concentration of CCL2 (set as 100%). (c) Inhibition
of CCL3-stimulated β-arrestin recruitment in U2OS-CCR5 by increasing
concentrations of compounds 8, 39, and 43, after stimulation with an EC80 concentration
of CCL3 (set as 100%). All data are from single, representative experiments
performed in duplicate.
Table 1
Characterization of Compounds 6–23 in hCCR2 and hCCR5a
hCCR2
hCCR5
compd
R1
R2
pKi ± SEM (Ki, nM)b
pIC50 ± SEM (IC50, nM) or inhibition at 1 μM (%)c
6
3-Cl
H
8.76 ± 0.01 (1.7)
60%
7
3-Cl
Me
8.46 ± 0.05 (3.5)
35%
8
3-Cl
NH2
8.90 ± 0.04 (1.3)
6.24 ± 0.004 (571)
9
H
NH2
8.27 ± 0.10 (5.9)
35%
10
2-Me
NH2
7.81 ± 0.05 (15.7)
28%
11
2-Cl
NH2
7.84 ± 0.03 (14.5)
27%
12
2-OMe
NH2
6.98 ± 0.04 (104.6)
–20%d
13
3-Me
NH2
8.61 ± 0.03 (2.5)
62%
14
3-F
NH2
8.53 ± 0.18 (3.4)
43%
15
3-Br
NH2
9.08 ± 0.06 (0.9)
60%
16
3-I
NH2
9.06 ± 0.02 (0.9)
66%
17
3-OMe
NH2
7.89 ± 0.07 (13.0)
–27%d
18
3-CF3
NH2
8.26 ± 0.09 (5.9)
36%
19
4-Me
NH2
8.46 ± 0.03 (3.5)
–57%d
20
4-F
NH2
8.39 ± 0.03 (4.1)
31%
21
4-Cl
NH2
8.74 ± 0.05 (1.8)
31%
22
4-Br
NH2
8.84 ± 0.02 (1.5)
14%
23
4-OMe
NH2
7.68 ± 0.05 (20.9)
–28%
Data are presented as the mean pKi/pIC50 ± standard error of
the mean (SEM) and mean Ki/IC50 (nM) of at least three independent experiments performed in duplicate.
pKi values
from the displacement of ∼6 nM [3H]-CCR2-RA-[R] from U2OS cells stably expressing CCR2, at 25 °C.
Percent inhibition of β-arrestin
recruitment in U2OS cells stably expressing CCR5 by 1 μM compound,
in the presence of CCL3 (pEC80 = 7.9). pIC50 values were determined for compounds displaying more than 70% inhibition.
% Inhibition values are presented as means values of at least two
independent experiments, performed in duplicate.
No inhibition was observed at the
concentration of 1 μM; instead some CCL3 stimulation was measured.
Characterization of ligands in U2OS-CCR2 and U2OS-CCR5.
(a) [3H]-CCR2-RA-[R] displacement by increasing
concentrations of triazolopyrimidinone derivatives 8, 39, and 43 in U2OS-CCR2 at 25 °C.
Data are normalized to specific binding in the absence of compound
(set as 100%). (b) Inhibition of CCL2-stimulated β-arrestin
recruitment in U2OS-CCR2 by increasing concentrations of compounds 8, 39, and 43, after stimulation
with an EC80 concentration of CCL2 (set as 100%). (c) Inhibition
of CCL3-stimulated β-arrestin recruitment in U2OS-CCR5 by increasing
concentrations of compounds 8, 39, and 43, after stimulation with an EC80 concentration
of CCL3 (set as 100%). All data are from single, representative experiments
performed in duplicate.Data are presented as the mean pKi/pIC50 ± standard error of
the mean (SEM) and mean Ki/IC50 (nM) of at least three independent experiments performed in duplicate.pKi values
from the displacement of ∼6 nM [3H]-CCR2-RA-[R] from U2OS cells stably expressing CCR2, at 25 °C.Percent inhibition of β-arrestin
recruitment in U2OS cells stably expressing CCR5 by 1 μM compound,
in the presence of CCL3 (pEC80 = 7.9). pIC50 values were determined for compounds displaying more than 70% inhibition.
% Inhibition values are presented as means values of at least two
independent experiments, performed in duplicate.No inhibition was observed at the
concentration of 1 μM; instead some CCL3 stimulation was measured.Previous studies have shown that some of these intracellular
ligands
are able to bind and inhibit multiple chemokine receptors, enabling
the design of selective and multitarget inhibitors.[19,21,22] In this regard, CCR5 is the closest homolog
to CCR2, with >90% sequence similarity of their intracellular binding
pockets. From the main interactions of CCR2-RA-[R] to CCR2, only Val2446×36 is exchanged to Leu2366×36 in CCR5[15] (residues named
according to structure-based Ballesteros—Weinstein nomenclature[25]). Thus, we investigated whether compound 8 is also able to inhibit the highly homologous CCR5. However,
the much lower affinity of [3H]-CCR2-RA-[R] for CCR5 compared to CCR2 hindered us from performing radioligand
binding assays.[21] Thus, we assessed the
CCR5 activity of 8 with a functional β-arrestin
recruitment assay after stimulation with CCL3, one of the endogenous
agonists of CCR5. For this assay, we also included the intracellular
ligands CCR2-RA-[R], SD-24 and JNJ-27141491, as well
as the CCR2/CCR5 orthosteric antagonist TAK-779 as a positive control
(Figure ), since it
is a potent CCR5 antagonist in a variety of functional assays.[26,27]In this assay, CCL3 induced β-arrestin recruitment to
U2OS
cells stably expressing hCCR5 (U2OS-CCR5) with a pEC50 of
8.3 ± 0.08 (6 nM) (Figure S1a), similar
to values reported in the literature.[28] As expected, TAK-779 was able to completely inhibit β-arrestin
recruitment induced by an EC80 concentration of CCL3 (pEC80 = 7.9 ± 0.08), when tested at a single concentration
of 1 μM (Figure S1b). In contrast,
none of the intracellular ligands were able to fully inhibit CCL3-induced
β-arrestin recruitment to the same level as TAK-779; in fact,
only compound 8 displayed more than 70% inhibition when
tested at 1 μM (Figure S1b), while
CCR2-RA-[R], SD-24, and JNJ-27141491 led to approximately
50% inhibition or less at the same concentration of 1 μM (Figure S1b). Consistent with this low inhibition
in CCR5, it was previously shown that CCR2-RA-[R],
JNJ-27141491, and SD-24 inhibited inositol phosphate (IP) formation
in CCR5 with 7- to 22-fold lower potency compared to CCR2 inhibition.[21] Preincubation of U2OS-CCR5 cells with increasing
concentrations of TAK-779, before exposure to CCL3, resulted in an
inhibitory potency (IC50) of 6 nM, consistent with previously
reported values (Table S2).[27] Also in agreement with a previous study,[21] the reference intracellular ligand CCR2-RA-[R] inhibited CCL3-induced β-arrestin recruitment with
an IC50 value of 703 nM (Table S2). Moreover, while TAK-779 inhibited CCL3-induced β-arrestin
recruitment with a pseudo-Hill slope close to unity (nH = −1.1), CCR2-RA-[R] inhibition
showed a significantly higher Hill slope (nH = −2.4), indicative of two different binding sites for CCL3
and CCR2-RA-[R] (Table S2).[29]As compound 8 was
the best CCR5 inhibitor in this
assay, displaying an IC50 value of 571 nM and a Hill slope
of −2.2 ± 0.3 (Figure c and Table ), we then synthesized several triazolopyrimidinone
derivatives to explore their structure–affinity/activity relationships
(SARs) in CCR2 and CCR5. All synthesized triazolopyrimidinone
derivatives were evaluated in [3H]-CCR2-RA-[R] binding assays to determine their binding affinity for CCR2 and
in β-arrestin recruitment assays to determine their activity
toward CCR5 (Figure and Tables –3). In CCR5, compounds were first screened at a concentration
of 1 μM, as we were only interested in dual-targeting compounds
with moderate to high potencies (IC50 < 1 μM).
For the same reason, only those that displayed >70% inhibition
at
this concentration were further evaluated in a concentration–inhibition
curve to determine their potency. For better comparison, compounds 8, 39, and 43 were also tested in
a CCR2 β-arrestin recruitment assay as previously described
(Figure b).[20] Finally, we determined the mechanism of inhibition
of 39 and 43 in both CCR2 and CCR5 using
a [35S]GTPγS binding assay (Figure , Table ).
Table 3
Characterization of Compounds 37–43 in hCCR2 and hCCR5a
hCCR2
hCCR5
compd
R1
R3
pKi ± SEM (Ki, nM)b
pIC50 ± SEM (IC50, nM) or inhibition at 1 μM (%)c
37
3,4-diCl
cPr
9.35 ± 0.05 (0.4)
6.67 ± 0.03 (214)
38
3,4-diCl
iPr
9.22 ± 0.05 (0.6)
6.91 ± 0.09 (132)
39
2,3-diCl
iPr
8.81 ± 0.07 (1.6)
7.09 ± 0.07 (84)
40
2,5-diCl
iPr
7.65 ± 0.03 (22.5)
20%
41
3,5-diCl
iPr
8.66 ± 0.05 (2.2)
6.49 ± 0.06 (336)
42
3,5-diBr
iPr
8.68 ± 0.01 (2.1)
64%
43
3-Br, 4-Cl
iPr
9.42 ± 0.02 (0.4)
6.95 ± 0.04 (115)
Data are presented as mean pKi/pIC50 ± standard error of
the mean (SEM) and mean Ki/IC50 (nM) of at least three independent experiments performed in duplicate.
pKi values
from the displacement of ∼6 nM [3H]-CCR2-RA-[R] from U2OS cells stably expressing CCR2, at 25 °C.
Percent inhibition of β-arrestin
recruitment in U2OS cells stably expressing CCR5 by 1 μM compound,
in the presence of CCL3 (pEC80 = 7.9). pIC50 values were determined for compounds displaying more than 70% inhibition.
% Inhibition values are presented as mean values of at least two independent
experiments, performed in duplicate.
Figure 3
Characterization of compounds 39 and 43 as insurmountable, negative allosteric modulators using
a [35S]GTPγS binding assay in hCCR2 and hCCR5. Effect
of
increasing concentrations of 39 and 43 in
a CCL2-stimulated [35S]GTPγS binding in U2OS-CCR2
(a, b) or in a CCL3-stimulated [35S]GTPγS binding
in U2OS-CCR5 (c, d), at 25 °C. Parameters obtained from the concentration–response
curves (pEC50, Emax) are summarized
in Table . Data are
presented as mean ± SEM values of three experiments performed
in duplicate.
Table 4
Effects of Compounds 39 and 43 in Chemokine-Stimulated [35S]GTPγS
Bindinga
receptor
compd
pEC50 ± SEM (EC50, nM)
Emax ±
SEM (%)b
hCCR2
CCL2
8.10 ± 0.06 (8)
107 ± 2
CCL2 + 10 nM 39
7.89 ± 0.04 (13)
91 ± 1**
CCL2 + 30 nM 39
7.60 ± 0.07 (26)**
75 ± 4****
CCL2 + 100 nM 39
7.27 ± 0.10 (56)****
50 ± 3****
CCL2 + 1 nM 43
7.91 ± 0.10 (13)
72 ± 4****
CCL2 + 3 nM 43
7.53 ± 0.12 (32)***
51 ± 5****
CCL2 + 10 nM 43
6.87 ± 0.13 (148)****
33 ± 3****
hCCR5
CCL3
8.42 ± 0.06 (4)
108 ± 2
CCL3 + 100 nM 39
8.35 ± 0.09 (5)
79 ± 5****
CCL3 + 300 nM 39
8.14 ± 0.12 (8)
56 ± 2****
CCL3 + 1000 nM 39
8.14 ± 0.17 (9)
25 ± 4****
CCL3 + 30 nM 43
8.30 ± 0.05 (5)
81 ± 3****
CCL3 + 100 nM 43
8.21 ± 0.05 (6)
58 ± 1****
CCL3 + 300 nM 43
8.05 ± 0.06 (9)*
35 ± 2****
Data represent the mean ± standard
error of the mean (SEM) of three independent experiments performed
in duplicate. One-way ANOVA with Dunnett’s post hoc test was
used to analyze differences in pEC50 and Emax values against CCL2 or CCL3 controls.
Maximum effect (Emax) of CCL2 or CCL3 measured in the absence or presence
of fixed concentrations of compound 39 and 43 in CCR2 or CCR5, respectively.
Characterization of compounds 39 and 43 as insurmountable, negative allosteric modulators using
a [35S]GTPγS binding assay in hCCR2 and hCCR5. Effect
of
increasing concentrations of 39 and 43 in
a CCL2-stimulated [35S]GTPγS binding in U2OS-CCR2
(a, b) or in a CCL3-stimulated [35S]GTPγS binding
in U2OS-CCR5 (c, d), at 25 °C. Parameters obtained from the concentration–response
curves (pEC50, Emax) are summarized
in Table . Data are
presented as mean ± SEM values of three experiments performed
in duplicate.
Structure–Affinity/Activity Relationships (SARs) in CCR2
and CCR5
Analysis of the triazolopyrimidinone derivatives
started by modifying the amino group (R2) of the triazolo
moiety (R2, Table ). Compared to 8, removing the amino group (6) resulted in a similar affinity toward CCR2, in agreement
with the similar reported IC50 values of approximately
20 nM for both compounds, when tested in a calcium flux assay.[23] However, in CCR5 6 displayed a
lower potency, as the inhibition of CCL3-stimulated recruitment of
β-arrestin decreased to 60%, compared to 76% inhibition by 8. The introduction of a methyl group in R2 (7) was less favorable for both receptors, as both affinity
for CCR2 and activity to CCR5 were reduced compared to 8. As compound 8 displayed the highest affinity/activity
for both receptors, we decided to keep the amino group in R2 and explore different phenyl substituents (R1, Table ), taking 8 as the starting point.Compared to 8, the unsubstituted 9 showed a 5-fold decrease in affinity toward CCR2, while
in CCR5 it was only able to inhibit 35% of the receptor response at
1 μM. Next, we investigated the effect of several benzyl modifications,
including the influence of different substituent positions (Table ). In the case of
CCR2, meta-substituted derivatives yielded the highest affinities
in this series of compounds (13–18), whereas ortho-substituted derivatives yielded the lowest (10–12). None of the ortho-substitutions
led to an improvement in affinity over 8 or the unsubstituted 9. Introduction of a methyl (10) or a chloro
(11) group in this position resulted in affinities lower
than 10 nM, while the introduction of an electron-donating methoxy
group further reduced the affinity to 105 nM (12), displaying
the lowest CCR2 affinity in this series (Table ). Moving the methyl group to meta (13) or para (19) position slightly improved the
CCR2 binding affinity compared to 9, achieving the highest
affinity in meta position (19, 3 nM). Similarly, moving
the methoxy group to meta or para position resulted in improved affinities
following the meta > para > ortho order; however, the affinities
remained
lower than 10 nM (17, 13 nM; 23, 21 nM),
with no improvement over 9. This is consistent with functional
data reported in the patent by Bengtsson et al., where similar compounds
with a methoxybenzyl moiety displayed a loss of CCR2 activity compared
to the unsubstituted-phenyl analogue.[23] Substitution of the meta methoxy group by an electron-withdrawing
CF3 group resulted in improved affinity over 17 (18, 6 nM) but no improvement over the unsubstituted 9.The effect of introducing different halogen groups
was first investigated
in the meta position. Overall, an increase in size and lipophilicity
from fluoro to iodo resulted in improved binding affinities toward
CCR2 (F, 14 < Cl, 8 < Br, 15 ≈ I, 16). In fact, compounds 15 and 16 displayed the highest affinities in this series
of derivatives (15, 0.8 nM; 16, 0.9 nM).
Moving the halogen substituents to the para position resulted in a
similar trend in affinity (F, 20 < Cl, 21 < Br, 22); however, their affinities were lower
compared to the meta-substituted analogues. Of note, compounds with
a fluorine atom in meta (14) or para (20) position displayed lower affinities than compounds with a methyl
group in the equivalent position (13 and 19). To gain more insight in a potential relationship between affinity
and lipophilicity as observed in the halogen series, calculated log P values (cLogP) of compounds 8–23, with R1 modifications, were plotted against
their pKi values in CCR2. This analysis
revealed only a slight correlation between these two parameters for
this set of compounds (Figure S2a); however,
this correlation was lost when all synthesized derivatives were included
in this plot (Figure S2b), indicating that
this is not a general trend.In the case of CCR5, meta-substituted
derivatives also outperformed
their ortho- and para-substituted analogues, with some compounds displaying
>60% inhibition at 1 μM; in contrast, ortho- and para-substitution
resulted in compounds with low (≤31%) to marginal efficacy
in CCR5, suggesting that substituents in ortho or para position are
not tolerated in CCR5. Similarly as in CCR2, the introduction of a
methoxy group was unfavorable, as it led to a complete loss of activity
in CCR5 when tested at 1 μM (12, 17, and 23), regardless of the position, whereas electron-withdrawing
groups in meta position (18, R2 = CF3) did not bring any improvement over the unsubstituted 9. Except for compound 14 bearing a meta-fluoro, which showed less than 45% inhibition, all other compounds
bearing halogens in meta position led to >60% inhibition; the same
was achieved when a methyl group was placed in this position (13). Overall, these data indicate that meta-substituents,
especially halogens, are preferred to achieve dual CCR2/CCR5 activity,
while ortho- and para-substituents lead to a lower affinity but higher
selectivity toward CCR2.As none of the other substituents in
R2 led to a significant
improvement in CCR5 activity over compound 8, we decided
to continue with this compound and investigate the effect of replacing
the cyclopropyl moiety in R3. On the basis of the chemical
structure of 8 and CCR2-RA-[R] (Figure ), we hypothesized
that the cyclopropyl group in 8 interacts with Val2446×36 in CCR2 in a similar manner as the cyclohexyl group
of CCR2-RA-[R].[15] Thus,
several triazolopyrimidinone derivatives were synthesized with
different alkyl chains and aromatic groups in this position in order
to investigate their SARs (Table ). Starting with the effect of alkyl substituents,
we observed that increasing the size and flexibility of the alkyl
chain from n = 1 (methyl) to n =
4 (butyl) resulted in a parallel increase in CCR2 affinity (17 nM
for R3 = Me (24); ∼4 nM for R3 = Et (25) and R3 = Pr (26);
2 nM for R3 = Bu (28)). However, further elongation
of the chain length (n = 5–7) led to a progressive
drop in affinity (7 nM for R3 = Pent (30);
22 nM for R3 = Hex (32); 178 nM for R3 = Hept (28)), indicating that linear alkyl chains
longer than five carbons might not fit in this hydrophobic pocket.
The same trend was observed for CCR5 activity, as only the n-propyl (26) and n-butyl
(28) substituted compound led to >60% inhibition,
albeit
without improvement over 8 (28, 519 nM).
Moreover, introduction of a hexyl or heptyl group resulted in CCL3
stimulation instead of inhibition, which was not further investigated.
Increasing bulkiness via branching of alkyl groups or substitution
with aliphatic rings enhanced the affinity toward CCR2, indicating
that these substituents might provide a better interaction with the
receptor. For instance, the introduction of both isopropyl (27) and cyclopropyl (8) groups led to an improvement
in CCR2 affinity compared to the linear analogue 26.
Moreover, compound 27 with an isopropyl substituent also
yielded a 2-fold increase in CCR5 potency compared to the cyclic analogue 8, displaying the highest potency in this series of compounds
(27, 281 nM). In line with this trend, we observed that
replacing the linear pentyl group (30) with a cyclopentyl
group (31) was also beneficial for CCR2, as this derivative
showed a 4.5-fold increased affinity compared to 30 (31, 1.6 nM). In CCR5, 31 inhibited the CCL3-induced
response with a potency of 388 nM, showing a slight improvement over
compound 8. In contrast, the introduction of a 2-ethylbutyl
group (29) resulted in reduced affinity/activity toward
both CCR2 and CCR5. These data suggest that the isopropyl group is
the preferred R3 substituent when designing CCR2/CCR5 dual
antagonists, as this substituent led to the highest potency in CCR5
while maintaining a high affinity for CCR2. Next, inspired by our
work on CCR1/CCR2 selectivity of pyrrolone derivatives,[19] we investigated whether aromatic substituents
are tolerated in this position. As expected from previous studies,[19,30] the introduction of aromatic groups decreased 20-fold (34, 27 nM), 40-fold (36, 52 nM), and 122-fold (35, 159 nM) the affinity for CCR2 compared to 8. When
tested in CCR5, all derivatives showed a complete loss of activity
at 1 μM, indicating that aromatic groups are not favorable for
selectivity or dual activity.
Table 2
Characterization of Compounds 24–36 in hCCR2 and hCCR5a
hCCR2
hCCR5
compd
R3
pKi ± SEM (Ki, nM)b
pIC50 ± SEM (IC50, nM) or inhibition at 1 μM (%)c
8
cPr
8.90 ± 0.04 (1.3)
6.24 ± 0.004 (571)
24
Me
7.78 ± 0.07 (17.2)
–35%
25
Et
8.40 ± 0.07 (4.0)
29%
26
Pr
8.46 ± 0.07 (3.6)
64%
27
iPr
8.72 ± 0.05 (1.9)
6.56 ± 0.05 (281)
28
Bu
8.64 ± 0.03 (2.3)
6.29 ± 0.05 (519)
29
2-EtBu
8.20 ± 0.04 (6.4)
29%
30
Pent
8.14 ± 0.03 (7.2)
38%
31
cPent
8.81 ± 0.04 (1.6)
6.43 ± 0.08 (388)
32
Hex
7.66 ± 0.02 (22.0)
–63%d
33
Hept
6.76 ± 0.05 (178.1)
–265%d
34
Ph
7.64 ± 0.17 (26.7)
–41%d
35
4-MePh
6.81 ± 0.07 (158.8)
–13%d
36
CH2CH2Ph
7.29 ± 0.05 (52.3)
–42%d
Data are presented as the mean pKi/pIC50 ± standard error of
the mean (SEM) and mean Ki/IC50 (nM) of at least three independent experiments performed in duplicate.
pKi values
from the displacement of ∼6 nM [3H]-CCR2-RA-[R] from U2OS cells stably expressing CCR2, at 25 °C.
Percent inhibition of β-arrestin
recruitment in U2OS cells stably expressing CCR5 by 1 μM compound,
in the presence of CCL3 (pEC80 = 7.9). pIC50 values were determined for compounds displaying more than 70% inhibition.
% Inhibition values are presented as mean values of at least two independent
experiments, performed in duplicate.
No inhibition was observed at the
concentration of 1 μM; instead some CCL3 stimulation was measured.
Data are presented as the mean pKi/pIC50 ± standard error of
the mean (SEM) and mean Ki/IC50 (nM) of at least three independent experiments performed in duplicate.pKi values
from the displacement of ∼6 nM [3H]-CCR2-RA-[R] from U2OS cells stably expressing CCR2, at 25 °C.Percent inhibition of β-arrestin
recruitment in U2OS cells stably expressing CCR5 by 1 μM compound,
in the presence of CCL3 (pEC80 = 7.9). pIC50 values were determined for compounds displaying more than 70% inhibition.
% Inhibition values are presented as mean values of at least two independent
experiments, performed in duplicate.No inhibition was observed at the
concentration of 1 μM; instead some CCL3 stimulation was measured.With the aim of finding dual CCR2/CCR5 intracellular
inhibitors,
we kept the isopropyl moiety in R3 and investigated the
effect of having a disubstituted phenyl moiety in R1 by
exploring different positions and combinations of chlorine and bromine
atoms (Table ). First and similar to 8, we
kept the cyclopropyl moiety in R3 and combined it with
dichlorination in meta and para positions (37). Compared
to the monosubstituted analogues 8 and 21, this compound yielded an even higher affinity to CCR2 (37, 0.4 nM); moreover, its ability to inhibit CCL3-induced response
in CCR5 was also improved, as the potency increased to 214 nM. By
replacing the cyclopropyl of 37 with an isopropyl group
(38), we retained affinity for CCR2 (0.6 nM), but the
potency for CCR5 increased by almost 2-fold (132 nM), in agreement
with the higher potency observed in 27 versus 8 (Table ). Moving
one chlorine atom to the ortho position, while keeping one in the
adjacent meta position, yielded compound 39 with slightly
lower affinity for CCR2 but even higher potency in CCR5 (39, 84 nM), indicating that although ortho substituents are not preferred
in monosubstituted derivatives, they are still tolerated when placed
in combination with halogens in other positions. However, placing
the two halogens in the second and fifth positions was clearly detrimental
for both receptors (40); in CCR2, the affinity decreased
by almost 40-fold, while in CCR2, the compound was only able to inhibit
20% of the CCR5 response. Placing the two halogens in the symmetrical
third and fifth positions restored the affinity/activity in both receptors
(41, 2.2 nM in CCR2 and 336 nM in CCR5). Replacing the
two chlorine atoms of 41 by bromine atoms yielded derivative 42, which retained affinity toward CCR2 but led to decrease
in CCR5 activity, as this compound was not able to inhibit >70%
of
the CCL3-induced response. Finally, the combination of a bromo in
meta position with a chloro in para position (42) improved
both the affinity and activity to both receptors to similar levels
as 37, in the case of CCR2, and 38 in the
case of CCR5, indicating that halogens in adjacent positions are more
favorable for activity in these receptors. Of note, compounds 37, 38, and 43 displayed the highest
affinities to CCR2 in this study, while 38, 39, and 43 displayed the highest potencies to CCR5.Data are presented as mean pKi/pIC50 ± standard error of
the mean (SEM) and mean Ki/IC50 (nM) of at least three independent experiments performed in duplicate.pKi values
from the displacement of ∼6 nM [3H]-CCR2-RA-[R] from U2OS cells stably expressing CCR2, at 25 °C.Percent inhibition of β-arrestin
recruitment in U2OS cells stably expressing CCR5 by 1 μM compound,
in the presence of CCL3 (pEC80 = 7.9). pIC50 values were determined for compounds displaying more than 70% inhibition.
% Inhibition values are presented as mean values of at least two independent
experiments, performed in duplicate.It is important to note that so far we are comparing
data not only
between two different receptors but also between two different assays:
(i) a radioligand binding assay for CCR2, in the absence of agonist,
which allows the determination of true affinities (pKi values); (ii) a functional assay for CCR5 in the presence
of an EC80 concentration of CCL3, without further correction
of their IC50 values. To better compare the activities
in both receptors, we selected starting compound 8 as
well as compounds 39 and 43 (with the highest
potency on CCR5 and the highest affinity for CCR2, respectively) and
tested these in a previously described β-arrestin recruitment
assay for CCR2.[20] In this assay, compound 8 inhibited CCL2-stimulated β-arrestin recruitment with
a potency of 10 nM and a Hill slope of −2.7, in agreement with
its allosteric binding mode. Compound 39 inhibited β-arrestin
recruitment in CCR2 with a lower potency of 21 nM, while compound 43 displayed a higher potency of 4 nM, consistent with their
affinities. In addition, their Hill slopes (nH = −2.5 for 39; nH = −3.4 for 43) are also indicative of
a noncompetitive form of inhibition, a further confirmation of their
allosteric binding site located in the intracellular region of CCR2
(Figure b and Table S3). Of note, the Hill slopes in CCR5 were
comparable to those in CCR2 (nH = −3.7
for 39; nH = −4.4
for 43), i.e., indicating an allosteric interaction at
CCR5 as well. Comparing the IC50 values obtained with the
functional assays in both receptors, we observe a 4-fold difference
between CCR2 and CCR5 in the case of 39, making it a
potential dual-antagonist for both receptors. In contrast, the potencies
in CCR2 and CCR5 differ by 29-fold in the case of 43,
indicating a higher selectivity toward CCR2. Yet, network studies
have suggested that partial inhibition by a low-affinity binder might
be sufficient to effectively modulate cellular pathways in
vivo;[31] thus further studies are
needed to establish the optimal activity ratio for these receptors
in order to achieve in vivo efficacy.[32]
Mechanism of Inhibition of Selected Compounds
Selected
compounds 39 and 43 were also tested in
a [35S]GTPγS binding assay in both CCR2 and CCR5
in order to determine their mechanism of inhibition. In the case of
CCR2, we have shown that these ligands fully displace [3H]-CCR2-RA-[R], indicating that triazolopyrimidinone
derivatives bind in the same intracellular binding site. Thus, these
compounds were expected to show noncompetitive, insurmountable antagonism
to (orthosteric) chemokine ligands, as previously demonstrated in
CCR2 with CCR2-RA-[R][20] and JNJ-27141491.[33] To verify this, 39 and 43 were characterized in a previously
described [35S]GTPγS binding assay on U2OS-CCR2 membranes.[20] In this assay, CCL2-stimulation of [35S]GTPγS binding in CCR2 was examined in the absence or presence
of fixed concentrations of 39 and 43 (Table and Figure a,b). In the absence of antagonist, increasing concentrations of
CCL2 induced [35S]GTPγS binding with an EC50 of 8 nM, in line with previously described parameters.[19,20] Co-incubation of CCL2 with 39 or 43 caused
a significant reduction in the maximal response of CCL2 (Emax) at all three antagonist concentrations tested. The
lowest concentrations of antagonist did not affect the potency of
CCL2, while higher concentrations significantly reduced the potency
of CCL2 (Table and Figure a,b). Of note, both
compounds were also tested in the absence of CCL2 at a single concentration
of 1 μM to determine potential inverse agonism. At this concentration
they only reduced the basal [35S]GTPγS binding levels
by 7–8% (Figure S3), providing too
small a window to accurately determine their potencies as inverse
agonists. Previously, we reported that some intracellular pyrrolone
derivatives were also able to decrease the CCR2 basal activity in
this assay; however, the effect seemed to be dependent on the assay
conditions, such as GDP concentrations.[19] Thus, more studies are needed to investigate whether the observed
CCR2 constitutive activity is biologically relevant.Data represent the mean ± standard
error of the mean (SEM) of three independent experiments performed
in duplicate. One-way ANOVA with Dunnett’s post hoc test was
used to analyze differences in pEC50 and Emax values against CCL2 or CCL3 controls.Maximum effect (Emax) of CCL2 or CCL3 measured in the absence or presence
of fixed concentrations of compound 39 and 43 in CCR2 or CCR5, respectively.To confirm our hypothesis that these two compounds
also bind to
an allosteric site in CCR5, i.e., the intracellular binding site,
we next analyzed the effect of 39 and 43 on CCL3-induced [35S]GTPγS binding in U2OS-CCR5
membranes. In agreement with previous studies, CCL3 stimulated [35S]GTPγS binding in CCR5 with a potency of 4 nM.[28] Similarly as in CCR2, the two compounds were
able to significantly suppress the maximal response induced by CCL3
at all concentrations tested (Table and Figure c,d). However, in contrast to CCR2, the potency of CCL3 was
only significantly reduced with the highest concentration of 43 (Table ). Such depression of the maximal response with or without a decrease
of agonist potency is typical of insurmountable antagonists,[34] indicating that 39 and 43 behave as insurmountable antagonists at both CCR2 and CCR5. Of note,
insurmountable antagonism can be generally achieved by two different
mechanisms: allosteric binding or slow binding kinetics, i.e., slow
equilibration of a competitive antagonist.[34] However, insurmountable inhibition due to hemiequilibrium is only
evident in preincubation experiments, where the receptor is preincubated
with the antagonist before exposure to the agonist.[34] In contrast, allosteric binding leads to insurmountable
inhibition in co-incubation experiments, as performed in this study.
These data further support our hypothesis that 39 and 43 bind to an allosteric binding site in CCR5, most probably
located intracellularly.
Docking Study
To further investigate the binding mode
of triazolopyrimidinone derivatives, compounds 8, 39, 40, and 43 were docked
into a CCR2b model based on the crystal structure of CCR2 (PDB code 5T1A, Figures and S4).[15] Due to the close proximity to the
intracellular binding site, several residues from the intracellular
loop 3 (ICL3) had to be modeled based on the crystal structure of
CCR5 (PDB code 4MBS),[35] since they were mutated in the original
CCR2 crystal structure to further stabilize the receptor. As seen
in Figure a, 43 was predicted to adopt a similar binding pose as that of
the previously cocrystallized ligand CCR2-RA-[R].[15] The disubstituted phenyl group of 43 was constrained to overlap with the corresponding phenyl group of
CCR2-RA-[R], since the disubstituted aromatic rings
of JNJ-27141491 and SD-24 (Figure ) were also predicted to overlay with the phenyl group
of CCR2-RA-[R] in a previous study.[21] However, the bromine group of compound 43 is
predicted to form a halogen bond with the backbone of Val1×53, which might contribute to the higher affinity of 43 versus CCR2-RA-[R] (Figure b). The positions of the halogens in this
phenyl group might also explain the results of our SAR study. For
instance, compound 40 with no halogen group on position
3 is not able to form halogen bonds with the receptor; in addition,
one of the chlorine groups seems to point upward toward Tyr7×53, resulting in a sterically unfavorable position and thus in the
observed lower activity (Figure S4 and Table ). Similar to 43, both 8 and 39 contain a chlorine
group in position 3, promoting the formation of a halogen bond with
the backbone of Val1×53, which might result in the
improved activity compared to 40 (Figure S4 and Table ). However, bromine displays a larger σ-hole and therefore
a higher halogen bond strength compared to chlorine,[36] which might result in the higher affinity of 43, containing a bromine group, compared to 8 or 39 containing a chlorine. As the SAR follows a similar trend
in CCR5, these data suggest that the intracellular ligands share similar
interactions between the aromatic group and both receptors (Table ).
Figure 4
Proposed binding mode
of 43 in hCCR2b. (a) Overlay
of 43 with the CCR2 intracellular ligand CCR2-RA-[R], showing that 43 interacts in a similar
manner as CCR2-RA-[R]. (b) Docking of 43, displaying the interactions with CCR2. The amino group in R2 makes an extra hydrogen-bond interaction with E8×48, while the bromine group in R1 makes an extra halogen
bond with the backbone of V1×53, which might contribute
to the improved affinity of 43 to this receptor. Model
of hCCR2 is based on the crystal structure of CCR2 (PDB code 5T1A),[15] and amino acid residues are labeled according to their
structure-based Ballesteros–Weinstein numbers.[25]
Proposed binding mode
of 43 in hCCR2b. (a) Overlay
of 43 with the CCR2 intracellular ligand CCR2-RA-[R], showing that 43 interacts in a similar
manner as CCR2-RA-[R]. (b) Docking of 43, displaying the interactions with CCR2. The amino group in R2 makes an extra hydrogen-bond interaction with E8×48, while the bromine group in R1 makes an extra halogen
bond with the backbone of V1×53, which might contribute
to the improved affinity of 43 to this receptor. Model
of hCCR2 is based on the crystal structure of CCR2 (PDB code 5T1A),[15] and amino acid residues are labeled according to their
structure-based Ballesteros–Weinstein numbers.[25]In addition, the isopropyl group present in compounds 39, 40, and 43 is predicted to bind
in the
same position as the cyclohexyl moiety of CCR2-RA-[R], although it seems to make less interactions with Val6×36 perhaps due to the slightly different ligand orientation (Figure b). Previous studies
have confirmed the crucial role of Val6×36 for binding
affinity of some intracellular ligands in CCR2, as mutation of this
residue to alanine completely abolished binding of CCR2-RA-[R] to the receptor.[21] Moreover,
this residue might be involved in target selectivity, as the main
difference between the intracellular pockets of CCR2 and CCR5 is the
single substitution of Val6×36 by Leu6×36. The steric hindrance introduced by this substitution might thus
be responsible for the reduction in activity of CCR2-RA-[R] and the triazolopyrimidinone derivatives toward CCR5 compared
to CCR2.[21] Indeed, in the case of CCR5,
only small aliphatic groups were tolerated in R3 position,
such as cyclopropyl or isopropyl (Table ), while bigger aliphatic groups resulted
in improved selectivity toward CCR2. In line with the role of Val6×36 as determinant of selectivity, a previous SAR analysis
of pyrrolone derivatives in CCR1, which also contains a leucine in
position 6 × 36, showed that aromatic groups in the equivalent
R3 position provide CCR1 selectivity versus CCR2, as aromatic
groups are not tolerated in this position in CCR2[19] (Table ).The binding pose of 43 seems to be stabilized
by a
network of hydrogen bonds between the triazolopyrimidinone core
and residues E8×48, Lys8×49, F8×50, and R3×50 (Figure b). Although the core of CCR2-RA-[R] and 43 binds with a different orientation,
the carboxy group of both is overlaid in the same position, interacting
with the backbones of Lys8×49 and F8×50. Moreover, the secondary and tertiary amino groups present in the
triazolopyrimidinone core also form hydrogen bonds with the
backbones of Lys8×49 and Glu8×48,
as well as with the side chains of Arg3×50. Finally,
the primary amino group in position R2 of compound 43 also makes an extra hydrogen bond with the side chain of
E8×48. Such extended network of hydrogen bond interactions
is not present with CCR2-RA-[R], and thus it might
also be responsible for the higher affinity of 43 in
CCR2, compared to CCR2-RA-[R]. In addition, the SAR
data suggest that this interaction is also crucial in CCR5, as the
removal of this amino group (compounds 6 and 7) was detrimental for CCR5 activity (Table ). Previous studies have confirmed the importance
of residues 8 × 49 and/or 8 × 50 in chemokine receptors
for the binding of several intracellular ligands. For example, alanine
mutations of Lys8×49 and F8×50 in
CCR2 caused a 10-fold reduction or a complete loss of affinity of
intracellular ligands, respectively, compared to the wild-type receptor.[21] In CXCR2, alanine mutation of Lys8×49 led to a reduced affinity of three different intracellular ligands,
while the mutation F8×50A only affected one of the
ligands tested, indicating a different binding mode.[37] Moreover, Lys8×49 has been suggested as
a key residue for target selectivity between CXCR1 and CXCR2, as it
is exchanged by Asn8×49 in CXCR1.[38] In addition, the crystal structure of CCR9 in complex with
vercirnon[16] also shows a binding interaction
between the ligand and Arg3238×49 and Phe3248×50.Overall, these data suggest that although the intracellular
pockets
of CCR2 and CCR5 are quite conserved, the design of multitarget compounds
is not quite straightforward. Moreover, several of these residues
have been shown to be involved in Gαi coupling in
recent cryoelectron microscopy (cryo-EM)-derived GPCR structures,
including residues 3 × 50, 6 × 29, 6 × 32 to 6 ×
37, 8 × 47, and 8 × 49.[39−41] Similarly, homologous
residues are also involved in direct interactions between rhodopsin
and arrestin,[42] suggesting a direct interference
of these intracellular ligands with the Gαi protein
and β-arrestin binding site, and the possibility of fine-tuning
residue interactions for the design of biased ligands. On the basis
of the SAR analysis and the docking study, these compounds could be
further optimized by exploring the triazolopyrimidinone core.
For instance, pyrazolopyrimidinones have also been described
for CCR2,[24] which might also display CCR5
activity. In addition, exploring other halogen combinations at the
phenyl group in R1 or other small, bulky aliphatic groups
in R3 such as cyclobutyl might lead to compounds with improved
dual activity. Although out of the scope of this manuscript, ligands
such as 39 with high potency in CCR5 could be investigated
as potential tools to further study binding interactions in CCR5,
i.e., by obtaining a radiolabeled tool compound or by obtaining a
CCR5 crystal structure in complex with an intracellular ligand. Finally,
these ligands can be used in future experiments designed to investigate
their functional effects both in vitro and in vivo, to validate the target combination, and to establish
the required level of target modulation.[43]
Conclusions
In this study we first confirmed that the
triazolopyrimidinone
derivative 8 binds to the intracellular pocket of CCR2
in a similar manner as the reference intracellular ligand CCR2-RA-[R]. Moreover, compound 8 was also able to inhibit
CCR5 in a functional β-arrestin recruitment assay; thus, we
took this compound as a starting point for the synthesis of a series
of novel and previously described triazolopyrimidinone derivatives.
Using [3H]-CCR2-RA-[R] binding assays
and functional β-arrestin recruitment assays, we explored structure–affinity/activity
relationships (SARs) in both receptors. Overall, these compounds were
mostly selective toward CCR2; however, CCR5 activity was increased
with the combination of a primary amino group in R2 position,
an isopropyl moiety in R3, and two halogens placed in adjacent
positions at the phenyl group in R1. Overall, these findings
indicate that even though the intracellular pockets of CCR2 and CCR5
are highly conserved, selectivity of intracellular ligands can be
fine-tuned, allowing the design of either selective or multitarget
ligands. Evaluation of compounds 39 and 43 in a [35S]GTPγS binding assay indicates that both
compounds display a noncompetitive, insurmountable mode of inhibition
in CCR2 and CCR5, which might represent a therapeutic advantage in
inflammatory diseases characterized by a high local concentration
of endogenous chemokines, such as multiple sclerosis and rheumatoid
arthritis. Thus, in diseases where selective chemokine receptor antagonists
have been largely unsuccessful, the development of multitarget, intracellular
ligands for CCR2 and CCR5 is warranted to further study the effects
of multitarget versus selective inhibition, as these ligands may represent
a novel therapeutic option in these diseases.
Experimental Section
Chemistry. General Methods
All solvents and reagents
used were of analytical grade and from commercial sources. Demineralized
water was used in all cases, unless stated otherwise, and is simply
referred to as H2O. Microwave-based synthesis was carried
out using a Biotage Initiator equipment (Biotage, Sweden). All reactions
were monitored by thin-layer chromatography (TLC) using aluminum plates
coated with silica gel 60 F254 (Merck), and compounds were
visualized under ultraviolet light at 254 nm or via KMnO4 staining. Column chromatography for compound purification was performed
using silica gel (Merck Millipore) with particle size 0.04–0.63
mm. Chemical identity of final compounds was established using 1H NMR and liquid chromatography–mass spectrometry (LC–MS). 1H NMR spectra were recorded on a Bruker AV 400 liquid spectrometer
(1H NMR, 400 MHz) at room temperature. Compound 39, as one of the most soluble and one of the best CCR2/CCR5 antagonists,
was fully characterized (1H NMR, 13C NMR, and
attached proton test (APT)) on a Bruker AV500 spectrometer at 80 °C.
The corresponding NMR spectra of compound 39 are shown
in Figures S5–S7. Chemical shifts
(δ) are reported in parts per million (ppm), and coupling constants
(J) in Hz. Liquid chromatography–mass spectrometry
(LC–MS) of final compounds was performed using a Thermo Finnigan
Surveyor LCQ Advantage Max LC–MS system and a Gemini C18 Phenomenex
column (50 mm × 4.6 mm, 3 μm). Analytical purity of the
compounds was determined using a Shimadzu high pressure liquid chromatography
(HPLC) equipment with a Phenomenex Gemini column (3 x C18 110A column,
50 mm × 4.6 mm, 3 μm). A flow rate of 1.3 mL/min and an
elution gradient of 10–90% MeCN/H2O (0.1% TFA) were
used. The absorbance of the UV spectrophotometer was set at 254 nm.
All compounds tested in biological assays showed a single peak at
the designated retention time and were ≥95% pure. Sample preparations
for HPLC and LC–MS were as follows unless stated otherwise:
0.3 mg/mL of compound was dissolved in a 1:1:1 mixture of H2O:MeOH:BuOH. Of note, some compounds
required DMSO and heat to ensure proper dissolution. None of the final
compounds were identified as potential pan-assay interference compounds
(PAINS) after assessment with the free ADME-Tox filtering tool (FAF-Drugs4),[44,45] which uses three different PAINS filters based on Baell et al.[46]
General Procedure 1: Synthesis of β-Keto Esters 1f–h,j,k,n[47]
In a flame-dried round-bottom
flask under a nitrogen atmosphere, ethyl acetoacetate (2.53 mL, 20.0
mmol, 1.00 equiv) was added dropwise to a suspension of NaH (880 mg,
22.1 mmol 1.10 equiv) in dry THF (5 mL) at 0 °C while stirring.
After 20 min, n-butyllithium (20 0.0 mmol, 2.50 M
solution in pentane,1.00 equiv) was added dropwise to the mixture
and stirred for further 30 min. The respective alkyl halide 2f–h,j,k or
benzyl bromide 2n (1.20 equiv) was subsequently added
dropwise over a period of 10 min to the dianion solution after which
the solution was allowed to reach rt. After 14 h, the reaction was
quenched by the addition of saturated NH4Cl (aq, 80 mL).
The mixture was subsequently extracted with diethyl ether (2 ×
120 mL). The combined organic fractions were washed with brine (80
mL) and dried over MgSO4 followed by concentration in vacuo.
The crude products were purified by flash chromatography (CH2Cl2/petroleum ether and/or EtOAc/petroleum ether as the
eluent) to give the title compounds 1f–h,j,k,n as oils. Compounds 1a–e,i,l,m were commercially available.
Ethyl-3-oxoheptanoate (1f)[47]
Compound 1f was synthesized according
to general procedure 1, using 1-bromopropane (2f, 2.00
mL, 22.0 mmol, 1.10 equiv) as starting compound. Compound was purified
by silica column chromatography (1–30% EtOAc in petroleum ether).
Yield: 36% (1.25 g) as a colorless oil. 1H NMR (400 MHz,
CDCl3) δ 4.19 (q, J = 7.2 Hz, 2H),
3.44 (s, 2H), 2.54 (t, J = 7.4 Hz, 2H), 1.65–1.55
(m, 2H), 1.38–1.25 (m, 7H), 0.88 (t, J = 6.8
Hz, 3H) ppm.
Ethyl 5-Ethyl-3-oxoheptanoate (1g)
Compound 1g was synthesized according to general procedure
1, using 3-bromopentane (2g, 3.00 mL, 24.2 mmol, 1.21
equiv) as starting compound. Compound was purified by silica column
chromatography (1–30% EtOAc in petroleum ether). Yield: 18%
(630 mg) as a yellow oil. 1H NMR (400 MHz, CDCl3) δ 4.17 (q, J = 7.2 Hz, 2H), 3.40 (s, 2H),
2.43 (d, J = 6.8 Hz, 2H), 1.88–1.73 (m, 1H),
1.33–1.21 (m, 7H), 0.82 (t, J = 7.4 Hz, 6H)
ppm.
Ethyl 3-Oxooctanoate (1h)[47]
Compound 1h was synthesized according
to general procedure 1, using 1-bromobutane (2h, 2.59
mL, 24.1 mmol, 1.21 equiv) as starting compound. Compound was purified
by silica column chromatography (1–30% EtOAc in petroleum ether).
Yield: 38% (1.41 g) as a yellow oil. 1H NMR (400 MHz, CDCl3) δ 4.19 (q, J = 7.2 Hz, 2H), 3.43
(s, 2H), 2.53 (t, J = 7.2 Hz, 2H), 1.65–1.55
(m, 2H), 1.37–1.12 (m, 7H), 0.89 (t, J = 7.2
Hz, 3H) ppm.
Ethyl 3-Oxononanoate (1j)[47]
Compound 1j was synthesized according
to general procedure 1, using 1-iodopentane (2j, 3.13
mL, 24.0 mmol, 1.20 equiv) as starting compound. Compound was purified
by silica column chromatography (1–30% EtOAc in petroleum ether).
Yield: 44% (1.76 g) as a yellow oil. 1H NMR (400 MHz, CDCl3) δ 4.19 (q, J = 7.2 Hz, 2H), 3.43
(s, 2H), 2.53 (t, J = 7.4 Hz, 2H), 1.54–1.64
(m, 2H), 1.33–1.24 (m, 9H), 0.88 (t, J = 6.8
Hz, 3H) ppm.
Ethyl 3-Oxodecanoate (1k)[48]
Compound 1k was synthesized according to general
procedure 1 using 1-bromohexane (2k, 3.36 mL, 25.0 mmol,
1.24 equiv) as starting compound. Compound was purified by silica
column chromatography (30–100% CH2Cl2 in petroleum ether). Yield: 15% (625 mg) as a yellow oil. 1H NMR (400 MHz, CDCl3) δ 4.19 (q, J = 7.2 Hz, 2H), 3.43 (s, 2H), 2.53 (t, J = 7.4 Hz,
2H), 1.64–1.53 (m, 2H), 1.33–1.24 (m, 11H), 0.88 (t, J = 7.5 Hz, 3H) ppm.
Ethyl 3-Oxo-5-phenylpentanoate (1n)[49]
Compound 1n was synthesized
according to general procedure 1, using benzyl bromide (2n, 2.90 mL, 24.4 mmol, 1.22 equiv) as starting compound. Compound
was purified by silica column chromatography (1–30% EtOAc in
petroleum ether). Yield: 20% (1.06 g) as a colorless oil. 1H NMR (400 MHz, CDCl3) δ 7.33–7.23 (m, 2H),
7.23–7.12 (m, 3H), 4.17 (q, J = 7.2 Hz, 2H),
3.42 (s, 2H), 2.98–2.81 (m, 4H), 1.26 (t, J = 7.2 Hz, 3H) ppm.
General Procedure 2. Benzylated β-Keto Esters 4aa–na, 4bb–bq, 4eq–ev[23]
LiCl (1.00 equiv) was slurried in anhydrous THF (1 mL/mmol 1a–n) in a flame-dried round-bottom flask
and under an atmosphere of nitrogen. The desired β-keto ester 1a–n (1.00 equiv) was added and followed
by DIPEA (2.00 equiv) and the respective benzylic halide 3a–v (1.20 equiv). The reaction mixture was refluxed
for 20 h, after which the reaction was completed as indicated by TLC
(5–10% EtOAc in petroleum ether). THF was removed in vacuo,
the crude dissolved in EtOAc (30 mL), and this organic layer was washed
with citric acid (5%, 25 mL) followed by saturated NaHCO3 (25 mL) and brine (25 mL). The organic layer was subsequently dried
over MgSO4, and concentrated in vacuo to afford the crude
product. The crude product was purified by flash column chromatography
(5–10% EtOAc/petroleum ether) to yield the corresponding benzylated
β-keto esters 4aa–na, 4bb–bq, 4eq–ev.
Ethyl 2-(3-Chlorobenzyl)-3-oxobutanoate (4aa)[23]
Compound was synthesized according
to general procedure 2, using the following reagents: ethyl 3-oxobutanoate 1a (0.37 mL, 2.92 mmol, 1.20 equiv), 3-chlorobenzyl bromide 3a (0.32 mmol, 2.43 mmol, 1.00 equiv), DIPEA (0.85 mL, 4.86
mmol, 2.00 equiv), LiCl (103 mg, 2.43 mmol, 1.00 equiv), 5 mL of dry
THF. Yield: 70% (433 mg) as a colorless oil. 1H NMR: (400
MHz, CDCl3): δ 7.21–7.14 (m, 3H), 7.09–7.04
(m, 1H), 4.16 (q, J = 7.2 Hz, 2H), 3.75 (t, J = 8.0 Hz, 1H), 3.17–3.07 (m, 2H), 2.21 (s, 3H),
1.21 (t, J = 7.2 Hz, 3H) ppm.
Compound was synthesized according to general procedure
2, using the following reagents: ethyl 3-oxo-5-phenylpentanoate 1n (0.661 g, 3.00 mmol, 1.00 equiv), 3-chlorobenzyl bromide 3a (0.475 mL, 3.65 mmol, 1.20 equiv), DIPEA (1.05 mL, 6.00
mmol, 2.00 equiv), LiCl (0.128 g, 3.00 mmol, 1.00 equiv), 5 mL of
dry THF. Yield: 67% (0.690 g) as a white solid. 1H NMR
(400 MHz, CDCl3) δ 7.32–7.26 (m, 2H), 7.24–7.19
(m, 3H), 7.16 (t, J = 7.2 Hz, 3H), 4.13 (q, J = 7.2 Hz, 2H), 3.75 (t, J = 7.6 Hz, 1H),
3.13 (d, J = 7.6 Hz, 2H), 2.99–2.85 (m, 3H),
2.75–2.67 (m, 1H), 1.20 (t, J = 7.2 Hz, 3H)
ppm.
Procedure for the Synthesis of 6-(3-Chlorobenzyl)-5-cyclopropyl[1,2,4]triazolo[1,5-a]pyrimidin-7(4H)-one (6)[23,24]
In a sealed microwave tube 3-amino-1,2,4-triazole 5a (66 mg, 0.78 mmol, 1.1 equiv), 4ba (200 mg,
0.71 mmol, 1.00 equiv), and H3PO4 (96 μL,
1.42 mmol, 2.00 equiv) were heated at 170 °C in 1 mL of EtOH
in the microwave for 10 h. The reaction mixture was poured in water
(30 mL), the pH was adjusted to pH = 12 (1 M NaOH aq), and the organics
were extracted with EtOAc (3 × 30 mL). The combined extracts
were dried over MgSO4 and the solvents evaporated in vacuo,
resulting in 165 mg of crude mixture. The pure product was obtained
by column chromatography (5% CH3OH in CH2Cl2) followed by prep HPLC gradient 10–90% CH3CN/water + 0.1%TFA yielding 4% (9 mg) as a white solid. 1H NMR (400 MHz, CDCl3): δ 8.14 (s, 1H), 7.28–7.25
(m, 1H), 7.22–7.15 (m, 3H), 7.17–7.08 (m, 3H), 4.16
(s, 2H), 2.22–2.13 (m, 1H), 1.36–1.30 (m, 2H), 1.20–1.15
(m, 2H) ppm. LC–MS (ESI) m/z calcd for C15H13ClN4O [M + H]+ 301.09, found 301.1. HPLC: 6.566 min, purity 97%.
Procedure for the Synthesis of 6-(3-Chlorobenzyl)-5-cyclopropyl-2-methyl[1,2,4]triazolo[1,5-a]pyrimidin-7(4H)-one (7)
In a sealed microwave tube 3-amino-5-methyl-1,2,4-triazole 5b (126 mg, 1.28 mmol, 1.2 equiv), 4ba (300 mg,
1.07 mmol, 1.0 equiv), and p-toluenesulfonic acid
monohydrate (102 mg, 0.53 mmol, 0.5 equiv) were heated for 30 min
at 180 °C in the microwave. As visualized by TLC, 4ba was consumed and mainly one product was formed (Rf 0.5 in 5% CH3OH in CH2Cl2). The crude product was
purified by column chromatography (3% CH3OH in CH2Cl2) yielding 31% (96 mg) as a white solid. 1H NMR (400 MHz, CDCl3 + drop MeOD): δ 7.23 (s, 1H),
7.21–7.13 (m, 3H), 6.26 (br s, 2H), 4.11 (s, 2H), 2.47 (s,
2H), 2.05–1.96 (m, 1H), 1.14–1.07 (m, 2H), 1.05–1.01
(m, 2H) ppm. LC–MS (ESI) m/z calcd for C16H15ClN4O [M + H]+ 315.10, found 315.1. HPLC: 6.826 min, purity 96%.
General Procedure 3. Triazolopyrimidinones 8–43[23]
The synthesis of
compounds 8–43 was according to the
following procedure: In a microwave tube a mixture of the corresponding
benzylated β-keto ester 4aa–na, 4bb–bq, 4eq–ev (1.00 equiv), triazole 5c (2.00 equiv) and
1-butyl-3-methylimidazolium hexafluorophosphate (BMIM-PF6, 1 mL or
6.00 equiv) was heated at 200 °C in a microwave reactor for an
hour. Afterward, the reaction was allowed to cool to room temperature
and stirred in a mixture of CH2Cl2 (30 mL),
H2O (10 mL), and 5–10% aqueous citric acid (1 mL)
for 20–30 min. The resulting mixture was filtered over a glass
filter, and the residue was washed with hot methanol. Finally, the
precipitate was collected and dried in vacuo to yield
the pure compounds.
Compound was synthesized according to general procedure
3, using the following reagents: 3,5-diamino-4H-1,2,4-triazole 5c (181 mg, 1.82 mmol, 2.00 equiv), 4ma (314
mg, 0.937 mmol, 1.00 equiv), and BMIM-PF6 (1.5 mL, 7.28 mmol, 7.77
equiv). Yield: 33% (108 mg) as a white solid. 1H NMR (400
MHz, DMSO): δ 12.80 (s, 1H), 7.33–7.28 (m, 4H), 7.26–7.14
(m, 2H), 7.05 (s, 1H), 6.96 (d, J = 7.1 Hz, 1H),
6.11 (s, 2H), 3.64 (s, 2H), 2.36 (s, 3H) ppm. LC–MS (ESI) m/z calcd for C19H16ClN5O [M + H]+ 366.20, found 366.2. HPLC: 7.342
min, purity 98%.
Procedure for the synthesis of 2-Amino-6-(3-chlorobenzyl)-5-phenethyl[1,2,4]triazolo[1,5-a]pyrimidin-7(4H)-one (36)
A mixture of 4na (213 mg, 0.57 mmol, 1.00
equiv), 3,5-diamino-4H-1,2,4-triazole 5c (116 mg, 1.16 mmol, 2.00 equiv), and ortho-phosphoric
acid 85% (59.6 μL, 0.906 mmol,1.58 equiv) in EtOH (1 mL) was
stirred for 1 min at 20 °C and then heated at 175 °C under
microwave irradiation for 3 h. The reaction mixture was allowed to
cool to room temperature, and CH2Cl2 (30 mL),
water (10 mL), and aqueous citric acid (5%, 1 mL) were added. The
resulting precipitate was stirred for 20 min, filtered, and the residue
was washed with hot methanol, collected, and dried in vacuo to afford
the title compound. Yield: 22% (49 mg) as a white solid. 1H NMR (400 MHz, DMSO): δ 12.67 (s, 1H), 7.32–7.18 (m,
6H), 7.17–7.10 (m, 3H), 6.06 (br s, 2H), 3.76 (s, 2H), 2.84–2.78
(m, 2H), 2.70–2.66 (m, 2H) ppm. LC–MS (ESI) m/z calcd for C20H18ClN5O [M + H]+ 380.13, found 380.2. HPLC: 7.392
min, purity 97%.
Compound was synthesized according to general procedure
3, using the following reagents: 3,5-diamino-4H-1,2,4-triazole 5c (109 mg, 1.10 mmol, 2.00 equiv), 4ev (200
mg, 0.553 mmol, 1.00 equiv), and BMIM-PF6 (1.00 mL, 4.86 mmol, 8.83
equiv). Yield: 24% (53 mg) as a white solid. 1H NMR (400
MHz, DMSO): δ 7.60 (d, J = 2.0 Hz, 1H), 7.49
(d, J = 8.0 Hz, 1H), 7.21 (dd, J = 8.0, 2.0 Hz, 1H), 6.08 (br s, 2H), 3.87 (s, 2H), 3.22–3.11
(m, 1H), 1.12 (d, J = 7.2 Hz, 6H) ppm. LC–MS
(ESI) m/z calcd for C15H15BrClN5O [M + H]+ 396.02, found
396.1. HPLC: 7.080 min, purity 98%.
Chemicals and Reagents
The human recombinant chemokines
CCL2 and CCL3 were purchased from PeproTech (Rocky Hill, NJ). TAK-779
was obtained from NIH AIDS reagent program (Germantown, MD, catalogue
number 4983). All triazolopyrimidinone derivatives were synthesized
in-house. Guanosine 5′-O-[γ-thio]triphosphate
([35S[GTPγS) (specific activity 1250 Ci/mmol) was
purchased from PerkinElmer (Waltham, MA), while [3H]-CCR2-RA-[R] (specific activity 59.6 Ci mmol–1)
was custom-labeled by Vitrax (Placentia, CA). Bovine serum albumin
was purchased from Sigma-Aldrich (St. Louis, MO). Bicinchoninic acid
(BCA) and Pierce BCA protein assay kit were purchased from Pierce
Biotechnology (Thermo Scientific, Rockford, IL). Tango U2OS cells
stably expressing humanCCR2b (U2OS-CCR2) or humanCCR5 (U2OS-CCR5)
were purchased from Invitrogen (Carlsbad, CA). All other chemicals
were obtained from standard commercial sources.
Cell Culture
Both U2OS-CCR2b and U2OS-CCR5 cells were
cultured in McCoy’s 5A medium supplemented with 10% (v/v) fetal
calf serum, 2 mM glutamine, 0.1 mM nonessential amino acids, 25 mM
HEPES, 1 mM sodium pyruvate, 200 IU/mL penicillin, 200 μg/mL
streptomycin, 100 μg/mL G418, 40–50 μg/mL hygromycin,
and 125 μg/mL zeocin. Cells were grown until 80% confluence
and cultured twice-weekly on 10 or 15 cm ⌀ plates by trypsinization.
Dialyzed fetal calf serum was used when culturing cells for functional
assays or as a last step before membrane preparation.
Membrane Preparation
Membranes from U2OS-CCR2 or U2OS-CCR5
cells were prepared as previously described for CCR2.[20] Briefly, U2OS-CCR2 or U2OS-CCR5 cells were scraped from
confluent 15 cm ⌀ plates using phosphate buffered saline (PBS)
and subsequently centrifuged at 3000 rpm for 5 min. Pellets were then
resuspended in ice-cold Tris buffer (50 mM Tris-HCl, 5 mM MgCl2, pH 7.4) before homogenization with an Ultra Turrax homogenizer
(IKA-Werke GmbH & Co. KG, Staufen, Germany). Membranes and cytosolic
contents were separated using an Optima LE-80 K ultracentrifuge (Beckman
Coulter, Inc., Fullerton, CA) at 31 000 rpm for 20 min at 4
°C. After a second cycle of homogenization and centrifugation,
the final pellet was resuspended and homogenized in ice-cold Tris
buffer, aliquoted, and stored at −80 °C. Finally, membrane
protein concentrations were determined using a BCA protein determination
assay, as described by the manufacturer (Pierce BCA protein assay
kit).[53]
[3H]-CCR2-RA-[R] Binding Assays
For [3H]-CCR2-RA-[R] displacement assays,
U2OS-CCR2b membrane homogenates (15–20 μg of total protein)
were incubated with ∼6 nM [3H]-CCR2-RA-[R] and at least 6 increasing concentrations of competing
ligand in a final volume of 100 μL of assay buffer (50 mM Tris-HCl,
5 mM MgCl2, 0.1% CHAPS, pH 7.4). Ligands were diluted to
the desired concentration with an HP D300 digital dispenser (Tecan,
Giessen, The Netherlands). Total radioligand binding did not exceed
10% of the amount added to prevent ligand depletion, and nonspecific
binding was determined using 10 μM CCR2-RA-[R]. After 2 h at 25 °C, incubation was terminated by rapid filtration
through a 96-well GF/B filterplate on a PerkinElmer FilterMate harvester,
using ice-cold wash buffer (50 mM Tris-HCl buffer supplemented with
5 mM MgCl2 and 0.05% CHAPS, pH 7.4). Filters were washed
10 times with ice-cold wash buffer and subsequently dried at 55 °C
for 30 min. After addition of 25 μL of Microscint scintillation
cocktail (PerkinElmer), the filter-bound radioactivity was measured
by scintillation spectrometry using the P-E 2450 Microbeta2 counter (PerkinElmer).
Tango β-Arrestin Recruitment Assay
β-Arrestin
recruitment was measured using the Tango CCR2-bla or CCR5-bla U2OS cell-based assay (Invitrogen)
according to the manufacturer’s protocol. Briefly, U2OS-CCR2b
or U2OS-CCR5 cells were grown until approximately 80% confluence and
detached by trypsinization. Cells were recovered by centrifugation
at 1000 rpm for 5 min, resuspended in assay medium (FreeStyle Expression
Medium, Invitrogen) to a density of 10 000 cells per well and
seeded into black-wall, clear-bottom, 384-well assay plates (Corning).
For agonist assays, cells were exposed to increasing concentrations
of CCL2 or CCL3 for CCR2 or CCR5, respectively, for 16 h at 37 °C
and 5% CO2. For antagonist assays, compounds were first
diluted in assay medium containing a final DMSO concentration of 0.5%
or lower. Cells were then preincubated with either 1 μM (for
single-point inhibition experiments) or increasing concentrations
of antagonist for 30 min at room temperature, before a 16 hour co-incubation
with an EC80 concentration of CCL2 (5 nM) or CCL3 (14 nM)
at 37 °C and 5% CO2. After 16 h cells were loaded
in the dark with 8 μL of LiveBLAzer-FRET B/G substrate (Invitrogen)
and incubated for 2 h at room temperature. Finally, fluorescence emission
at 460 and 535 nm was measured in an EnVision multilabel plate reader
(PerkinElmer) after excitation at 400 nm. The ratio of emission at
460 and 535 nm was calculated for each well.
[35S]GTPγS Binding Assay
To determine
the mechanism of inhibition [35S]GTPγS binding assays
were performed. In CCR2 the [35S]GTPγS binding assay
was performed as previously described.[19,20] In the case
of CCR5, an amount of 10 μg of U2OS-CCR5 membranes with 0.25
mg/mL saponin was preincubated with 5 μM GDP, increasing concentrations
of CCL3, and three different antagonist concentrations for 30 min
at 25 °C. All dilutions were made in assay buffer containing
50 mM Tris-HCl (pH 7.4), 10 mM MgCl2, 10 mM NaCl, 1 mM
EDTA, and 0.05% BSA. [35S]GTPγS (0.3 nM) was added,
and the mixture was co-incubated for an additional 90 min at 25 °C
before harvesting. Incubation was stopped by dilution with ice-cold
50 mM Tris-HCl buffer with 5 mM MgCl2. Separation of bound
and unbound radioligand was performed by rapid filtration through
a 96-well GF/B filter plate as described in “[3H]-CCR2-RA-[R] Binding Assays”.
Data Analysis
All experiments were analyzed using nonlinear
regression curve fitting program Prism 7 (GraphPad, San Diego, CA).
EC50, EC80, Emax, and IC50 values from functional assays were obtained
by nonlinear regression analysis. All values obtained are the mean
± SEM of at least three separate experiments performed in duplicate,
unless stated otherwise. For radioligand binding assays, Ki values were determined using the Cheng–Prussoff
equation using a KD of 6.3 nM for the
radioligand.[19]
Computational Modeling
The inactive crystal structure
of hCCR2b with BMS-681 and CCR2-RA-[R] (PDB code 5T1A)[15] was used as the basis for docking compounds 8, 39, 40, and 43. Docking
was performed in the Schrodinger suite,[54] as previously described for the docking of CCR2-RA-[R].[19] Before docking, the CCR2b crystal
structure was prepared by replacing the residues between L2265×62 and R2406×32, which correspond to
the M2 muscarinic acetylcholine receptor, with the CCR2b sequence
using prime[55−57] and CCR5 as template (PDB code 4MBS).[35] Induced fit docking, with a substructure restraint on the
right-hand phenyl (R1, SMARTS: “c1ccccc1”)
was used to dock compound 43 in the hCCR2b model.[58,59] Compounds 8, 39, and 40 were
docked using regular docking (Glide-SP) on the basis of the model
of compound 43. Figures a, 4b, and S4 were rendered using PyMOL.[60]
Authors: Annelien J M Zweemer; Julia Bunnik; Margo Veenhuizen; Fabiana Miraglia; Eelke B Lenselink; Maris Vilums; Henk de Vries; Arthur Gibert; Stefanie Thiele; Mette M Rosenkilde; Adriaan P IJzerman; Laura H Heitman Journal: Mol Pharmacol Date: 2014-07-14 Impact factor: 4.436
Authors: Christine Oswald; Mathieu Rappas; James Kean; Andrew S Doré; James C Errey; Kirstie Bennett; Francesca Deflorian; John A Christopher; Ali Jazayeri; Jonathan S Mason; Miles Congreve; Robert M Cooke; Fiona H Marshall Journal: Nature Date: 2016-12-07 Impact factor: 49.962
Authors: Mieke Buntinx; Bart Hermans; Jan Goossens; Dieder Moechars; Ron A H J Gilissen; Julien Doyon; Staf Boeckx; Erwin Coesemans; Guy Van Lommen; Jean P Van Wauwe Journal: J Pharmacol Exp Ther Date: 2008-07-03 Impact factor: 4.030
Authors: Scott L Friedman; Vlad Ratziu; Stephen A Harrison; Manal F Abdelmalek; Guruprasad P Aithal; Juan Caballeria; Sven Francque; Geoffrey Farrell; Kris V Kowdley; Antonio Craxi; Krzysztof Simon; Laurent Fischer; Liza Melchor-Khan; Jeffrey Vest; Brian L Wiens; Pamela Vig; Star Seyedkazemi; Zachary Goodman; Vincent Wai-Sun Wong; Rohit Loomba; Frank Tacke; Arun Sanyal; Eric Lefebvre Journal: Hepatology Date: 2018-01-29 Impact factor: 17.425
Authors: Natalia V Ortiz Zacarías; Kirti K Chahal; Tereza Šimková; Cas van der Horst; Yi Zheng; Asuka Inoue; Emy Theunissen; Lloyd Mallee; Daan van der Es; Julien Louvel; Adriaan P IJzerman; Tracy M Handel; Irina Kufareva; Laura H Heitman Journal: J Med Chem Date: 2021-02-18 Impact factor: 7.446
Figure 1
Scheme 1
Figure 2
Figure 3
Figure 4