Spinal cord injuries (SCIs) irreversibly disrupt spinal connectivity, leading to permanent neurological disabilities. Current medical treatments for reducing the secondary damage that follows the initial injury are limited to surgical decompression and anti-inflammatory drugs, so there is a pressing need for new therapeutic strategies. Inhibition of the type 2 lysophosphatidic acid receptor (LPA2) has recently emerged as a new potential pharmacological approach to decrease SCI-associated damage. Toward validating this receptor as a target in SCI, we have developed a new series of LPA2 antagonists, among which compound 54 (UCM-14216) stands out as a potent and selective LPA2 receptor antagonist (Emax = 90%, IC50 = 1.9 μM, KD = 1.3 nM; inactive at LPA1,3-6 receptors). This compound shows efficacy in an in vivo mouse model of SCI in an LPA2-dependent manner, confirming the potential of LPA2 inhibition for providing a new alternative for treating SCI.
Spinal cord injuries (SCIs) irreversibly disrupt spinal connectivity, leading to permanent neurological disabilities. Current medical treatments for reducing the secondary damage that follows the initial injury are limited to surgical decompression and anti-inflammatory drugs, so there is a pressing need for new therapeutic strategies. Inhibition of the type 2 lysophosphatidic acid receptor (LPA2) has recently emerged as a new potential pharmacological approach to decrease SCI-associated damage. Toward validating this receptor as a target in SCI, we have developed a new series of LPA2 antagonists, among which compound 54 (UCM-14216) stands out as a potent and selective LPA2 receptor antagonist (Emax = 90%, IC50 = 1.9 μM, KD = 1.3 nM; inactive at LPA1,3-6 receptors). This compound shows efficacy in an in vivo mouse model of SCI in an LPA2-dependent manner, confirming the potential of LPA2 inhibition for providing a new alternative for treating SCI.
A spinal cord injury (SCI) is defined
as damage to the spinal cord
that provokes a temporary or permanent impairment of its function.
It has negative consequences for the physical and social well-being
of patients and imposes an important economic burden to the individual
and the health care system. SCI can have traumatic or nontraumatic
origins. The former happens when an external physical impact acutely
harms the spinal cord, whereas the latter is associated with disease
development, such as a tumor, an infection, or a neurodegenerative
process. Regardless of the etiology, the primary injury damages cells
and initiates a complex secondary cascade of secondary degeneration
characterized by ischemia, excitotoxicity, and inflammatory processes
that lead to the death of neurons and glial cells. This process is
followed by a reorganization of the structural architecture of the
spinal cord and by the formation of glial scars that, together with
the poor capacity of the central nervous system (CNS) to promote remyelination
and axonal growth, causes irreversible neurological deficits. Considering
the negative impact of SCI, it is clear that prevention of the primary
injury is desirable, as would an efficacious treatment to minimize
secondary injury events to prevent functional impairments. The last
several years have witnessed an important advancement of the field,
with the development of different experimental neuroprotective and
neuroregenerative therapies that have been translated from preclinical
studies into clinical trials.[1−3] However, the current medical reality
is that there is no treatment for acute SCI because methylprednisolone,
which was the standard treatment for acute SCI, is no longer used
for the management of spinal cord trauma in many countries based on
several reports demonstrating its lack of therapeutic efficacy and
its undesirable side effects related to immunosuppression and gastrointestinal
bleeding.[4] Hence, it is evident there is
a crucial need to develop new treatments for SCI. In this regard,
there is a consensus in that primary injury cannot be therapeutically
addressed, but secondary cell damage events that occur after SCI could
be susceptible to therapeutic intervention. Hence, much research effort
has focused on delineation of the receptor pathways responsible for
the irreversible cellular damage that occurs after SCI, because they
could represent new therapeutic targets for novel drug treatments.
In this context, bioactive lipids have recently emerged as major players
in the initiation and maintenance of the pro-inflammatory environment
that prevent tissue repair and recovery of homeostasis.[5] Among them, lysophosphatidic acid (LPA, 1-acyl-sn-glycerol-3-phosphate) has received an increasing attention.[6,7] Although LPA can refer to multiple different species of lysophospholipids
with saturated (16:0, 18:0) and unsaturated (16:1, 18:1, 18:2, 20:4)
acyl chains, in the context of SCI, LPA 18:1 (1-oleoyl-sn-glycerol-3-phosphate) appears to be the most important form.[8] The increase in LPA levels in the CNS after traumatic
injury has detrimental effects, as it has been confirmed by experiments
showing that intraspinal injection of LPA leads to inflammation and
demyelination.[8] However, taking into account
that LPA can activate at least six different receptors (LPA1–6) that belong to the G protein-coupled receptor (GPCR) superfamily,[9−11] the next important step is to determine which specific receptor
subtype(s) is responsible for the deleterious effect of pathological
LPA exposure. In this regard, the importance of LPA1 as
a target for the treatment of SCI has been well established,[8,12] but this receptor does not account for all the effects observed
with LPA. Very recently, LPA2 has been postulated as a
key receptor in mediating the effects of LPA in SCI.[13] However, its validation has been hampered by the lack of
selective antagonists. Currently, only two compounds (C35 and H2L5186303, Figure ) have been characterized
as potent (IC50 values at LPA2 of 0.017 and
0.0089 μM, respectively) and selective LPA2 antagonists
(IC50 values >50 μM at LPA1 and LPA3 for C35 and 1.23 and 27.3 μM at LPA1 and
LPA3 for H2L5186303).[14,15] However, their
selectivity profile versus the other LPA receptors (LPA4–6), pharmacokinetic properties, and in vivo efficacy have not been
studied. Another tool compound widely used to study the effect of
blocking LPA receptor signaling is Ki16425 (Figure ), but although it has good in vitro potency,
this derivative is a nonselective antagonist with submicromolar activity
at LPA1 and LPA3 and lower affinity at LPA2 (IC50 values of 0.34, 0.93, and 6.5 μM,
respectively)[16] and with limited in vivo
activity that may reflect its short half-life.[17]
Figure 1
Structure of the LPA2 antagonists C35, H2L5186303, and
Ki16425.
Structure of the LPA2 antagonists C35, H2L5186303, and
Ki16425.New potent and selective LPA2 antagonists
could enable
the validation of this receptor as a target for the treatment of SCI
and might represent a new therapeutic avenue. Here we report the development
of the most potent and selective LPA2 antagonist described
so far, compound UCM-14216 (54), which has an IC50 value of 1.9 μM as an LPA2 antagonist,
a KD value of 1.3 nM, and a selectivity
over other LPA receptor subtypes (LPA1 and LPA3–6, with 10-fold selectivity in terms of IC50 value with
respect to LPA1 and LPA3 and 10-fold selectivity
versus LPA6 and >50-fold selectivity versus LPA4 and LPA5 in terms of KD).
In addition, this compound significantly improves motor recovery in
an in vivo model of SCI, supporting the importance of LPA2 for the treatment of SCI.
Results and Discussion
Within a broad project focused
on the discovery of new ligands
for LPA receptors,[18] we started our search
of potent and selective LPA2 antagonists through an in-house
screen using a functional assay to detect calcium mobilization in
cells stably transfected with the LPA2 receptor in which
the compounds under study were added at a fixed dose of 10 μM
and the cells were subsequently stimulated with LPA at the same concentration.
We considered active those compounds able to reduce the LPA-mediated
calcium response by at least 30%. Among all tested molecules, compound 1 (Figure ) showed a consistent antagonist signal at LPA2 receptor,
absence of significant agonist activity at this receptor (Figure S1), and selectivity versus LPA1 and LPA3 receptors, so it was selected as our initial
hit. However, its moderate antagonism at LPA2 at 10 μM
(Emax = 48 ± 9%) led us to carry
out a systematic structural exploration of this compound with the
aim of improving its biological activity.
Figure 2
Design of new LPA2 antagonists 2, 3, and 10–13.
Design of new LPA2 antagonists 2, 3, and 10–13.
Structure–Activity Relationship (SAR) Study of Hit 1
First, we tried to establish the relative importance of the different
parts of the molecule for the LPA2 antagonist activity.
We started by studying the influence of the chlorophenoxy group by
removing the whole moiety or just the halogen atom with the synthesis
of compounds 2 and 3 (Figure ).Compound 2 was prepared
from commercially available 1-(2,4-dihydroxyphenyl)ethanone by treatment
with triethyl orthoformate and perchloric acid. Then, resulting hydroxychromone 4 was alkylated with methyl bromoacetate and treated with
an excess of hydrazine to obtain desired derivative 2 (Scheme ), through
opening of the pyrone ring and subsequent formation of pyrazole ring.
The reaction with hydrazine promoted the simultaneous transformation
of the ester group to the corresponding hydrazide, which was hydrolyzed
to obtain the target carboxylic acid. With respect to compound 3, its synthesis started with a Friedel–Crafts acylation
between 2-phenoxyacetyl chloride and resorcinol. Next, the Kostanecki–Robinson
reaction between the resulting ketone 6 and acetic anhydride
afforded chromone 7 in a good yield, which was, after
hydrolysis of the acetyl group in acid media, alkylated with methyl
bromoacetate to obtain intermediate 9. Finally, treatment
with hydrazine gave target compound 3 (Scheme ). Antagonist activity assays
revealed that compound 2 was inactive at LPA2 (Emax = 7 ± 3%) whereas derivative 3 showed a low activity at LPA2 (Emax = 20 ± 7%, Table ), highlighting the need not only of the chlorine atom
but also of the whole phenoxy system for the LPA2 antagonist
activity. Hence, we studied the influence of the position of the chloro
substituent with the synthesis of compounds 10 and 11, where the chlorine atom was located in a meta or para
position, respectively (Figure ). These syntheses were accomplished following a synthetic
route similar to the one previously followed for compound 3 starting from the corresponding chlorophenoxyacetic acid and resorcinol
(Scheme ).
Antagonist Activities of Compounds 1, 3, 10–13,
and Ki16425 at LPA1-3
Emax = maximum blockade effect of the activation induced by 10 μM
of LPA (18:1, 1-oleoyl-sn-glycerol-3-phosphate) at
a concentration of the compound under study of 10 μM.
For Emax > 70%, IC50 values are expressed as mean ± s.e.m,
from a minimum of two independent experiments, performed in triplicate.
N.E., no effect was observed
at
the highest concentration of compound tested (10 μM).
Synthesis
of Compounds 2, 3, and 10–13
Reagents and conditions:
(a)
CH(OEt)3, 70% HClO4, H2O, rt, 13
h, 47%; (b) methyl bromoacetate, K2CO3, acetone,
reflux, 3 h, 24–98%; (c) (i) 65% N2H4·H2O, EtOH, reflux, 30 min, 77–99%; (ii) 2
M NaOH, EtOH, reflux, 12 h, 67–99%; (d) (i) SOCl2, toluene, 110 °C, 16 h, 99%; (ii) resorcinol, BF3·Et2O, DCM, reflux, 4–5 h, 19–25%;
(e) acetic anhydride, Et3N, NaOAc, 140 °C, 2–3
h, 83–99%; (f) conc. HCl, EtOH, reflux, 2 h, 75–99%.Emax = maximum blockade effect of the activation induced by 10 μM
of LPA (18:1, 1-oleoyl-sn-glycerol-3-phosphate) at
a concentration of the compound under study of 10 μM.For Emax > 70%, IC50 values are expressed as mean ± s.e.m,
from a minimum of two independent experiments, performed in triplicate.N.E., no effect was observed
at
the highest concentration of compound tested (10 μM).Determination of the LPA2 antagonist character
of compounds 10 and 11 revealed that whereas
the former did
not improve the antagonist activity of the initial hit 1 [Emax (1) = 48%; Emax (10) = 45%], the latter increased
the LPA2 antagonist activity at the maximum concentration
[Emax (11) = 60%] (Table ). These results suggested
that the chlorine atom was tolerated at the three positions, with
the best result obtained for the para derivative, so it may be possible
that the introduction of a second chlorine atom allowed further improvement
of activity. Accordingly, compounds 12 and 13 were synthesized (Scheme ) and tested for LPA2 activity (Table ). Determination of their antagonist
character revealed that introduction of the 2,4-dichlorophenoxy moiety
yielded an excellent LPA2 antagonist [Emax (13) = 84%; IC50 (13) = 5.5 μM; Figure S2], with similar
LPA2 antagonist activity to Ki16425, used as the reference
ligand (Table ). Also,
to rule out the existence of partial agonism, we measured the agonist
activity of compounds 3 and 10–13 at LPA2 receptors, and none of them was able
to induce any significant activation of the receptor at 10 μM
concentration (see Figure S1 for the result
obtained for compound 13, which is representative of
the rest of the compounds).At this point, we considered that
a detailed study of the molecular
interactions involved in the affinity of compound 13 for
LPA2 could help us to rationalize the activity results
and also shed some light on the binding site of the compound. Hence,
we built a homology model of LPA2 using the disclosed crystal
structure of the LPA1 as a template.[19] The best docking pose of compound 13 in the
LPA2 receptor model (Figure A) suggests that the phenolic hydroxy group interacted
with two hydrogen bonds with arginine 107 and glutamine 108 and the
carboxylic acid group is engaged in two salt bridges with lysines
22 and 278 (Figure A). The dichlorophenoxy moiety lies in a hydrophobic pocket surrounded
by leucine 261, leucine 111, glutamine 108, glycine 257, tryptophan
254, alanine 284, tyrosine 85 and phenylalanine 280. The chlorine
atom in the 2-position points to residues leucine 111 and glutamine
108, while the one in the 4-position points to residues glycine 257
and alanine 284 (Figure A). Also, the oxygen atom of the phenoxy moiety forms a hydrogen
bond with glutamine 108. Compound 3 adopts a similar
pose to compound 13, but its phenoxy moiety does not
completely fill the hydrophobic pocket since it cannot simultaneously
reach glycine 257, alanine 284, and leucine 111 as compound 13 does through its two chlorine atoms (Figure B).
Figure 3
(A) LPA1 (PDB ID 4Z35)-derived homology
model of LPA2R in complex with compounds 3 (in purple) and 13 (in white). The phenolic hydroxy
group of both compounds
is engaged in two hydrogen bonds with arginine 107 and glutamine 108
and the carboxylic acid group is engaged in two salt bridges with
lysines 22 and 278. Also, the oxygen atom of the phenoxy moiety forms
a hydrogen bond with glutamine 108. (B) Phenoxy moiety of the two
compounds lies in the same hydrophobic pocket but compound 3, represented here with a C-purple surface representation, cannot
reach simultaneously residues glycine 257, alanine 284, and leucine
111 as compound 13 does, represented here as C-white
mesh representation.
(A) LPA1 (PDB ID 4Z35)-derived homology
model of LPA2R in complex with compounds 3 (in purple) and 13 (in white). The phenolic hydroxy
group of both compounds
is engaged in two hydrogen bonds with arginine 107 and glutamine 108
and the carboxylic acid group is engaged in two salt bridges with
lysines 22 and 278. Also, the oxygen atom of the phenoxy moiety forms
a hydrogen bond with glutamine 108. (B) Phenoxy moiety of the two
compounds lies in the same hydrophobic pocket but compound 3, represented here with a C-purple surface representation, cannot
reach simultaneously residues glycine 257, alanine 284, and leucine
111 as compound 13 does, represented here as C-white
mesh representation.The importance of the phenolic hydroxy group was
confirmed through
the synthesis of compound 30 (Scheme ), which was obtained starting with a Williamson
reaction between 2-bromo-1-(4-methoxyphenyl)ethanone and 2,4-dichlorophenol
under microwave (MW) irradiation, using 1,8-diazabicyclo[5.4.0]undec-7-ene
(DBU) as a base. Then, treatment of the intermediate 31 with 1,1-dimethoxy-N,N-dimethylethanamine
yielded enaminone 32, which was reacted with hydrazine
to obtain pyrazole 33. Finally, removal of the methoxy
group followed by O-alkylation with methyl bromoacetate
and hydrolysis of the ester gave the target pyrazole 30 (Scheme ), which
was basically inactive at the LPA2 antagonist assay, with
an Emax value of only 11%.
Scheme 2
Synthesis
of Compound 30
Reagents and conditions:
(a)
2,4-dichlorophenol, DBU, DMF, MW, 140 °C, 30 min, 80%; (b) 1,1-dimethoxy-N,N-dimethylethanamine, 90 °C, 4 h,
52%; (c) 65% N2H4·H2O, EtOH,
reflux, 1 h, 44%; (d) BBr3, DCM, −78 °C to
rt, 21 h, 70%; (e) methyl bromoacetate, K2CO3, DMF, −20 °C to rt, 16 h, 32%; (f) 1 M NaOH, 1,4-dioxane,
60 °C, 2 h, 99%.
Synthesis
of Compound 30
Reagents and conditions:
(a)
2,4-dichlorophenol, DBU, DMF, MW, 140 °C, 30 min, 80%; (b) 1,1-dimethoxy-N,N-dimethylethanamine, 90 °C, 4 h,
52%; (c) 65% N2H4·H2O, EtOH,
reflux, 1 h, 44%; (d) BBr3, DCM, −78 °C to
rt, 21 h, 70%; (e) methyl bromoacetate, K2CO3, DMF, −20 °C to rt, 16 h, 32%; (f) 1 M NaOH, 1,4-dioxane,
60 °C, 2 h, 99%.Next, we focused our
attention on the influence of the distance
between the oxygen atom and the carboxylic acid group in compound 13. To determine the optimum length of the methylenic chain
that separates these two moieties, we synthesized compounds 36–38, which have 2–4 methylenes
in the linker (Scheme ).
Scheme 3
Synthesis of Compounds 36–38
Reagents and conditions:
(a)
methyl acrylate, DMAP, MW, 150 °C, 2.5 h, 10%; (b) (i) 65% N2H4·H2O, EtOH, reflux, 30 min, 24–86%;
(ii) 2 M NaOH, EtOH, reflux, 12 h, 86–99%; (c) methyl 4-bromobutanoate
or methyl 5-bromopentanoate, K2CO3, acetone,
reflux, 3–5 h, 80–82%.
Synthesis of Compounds 36–38
Reagents and conditions:
(a)
methyl acrylate, DMAP, MW, 150 °C, 2.5 h, 10%; (b) (i) 65% N2H4·H2O, EtOH, reflux, 30 min, 24–86%;
(ii) 2 M NaOH, EtOH, reflux, 12 h, 86–99%; (c) methyl 4-bromobutanoate
or methyl 5-bromopentanoate, K2CO3, acetone,
reflux, 3–5 h, 80–82%.None
of the synthesized compounds showed any activity as LPA2 agonists at 10 μM concentration and, in all cases,
increasing the distance between the carboxylic acid group and the
rest of the molecule resulted in decreased LPA2 antagonism
activity (Table ).
The worst result was obtained for compound 38, bearing
the longest chain (n = 4) with an Emax value of 21% compared to the 84% of derivative 13. This decrease in activity can be rationalized by the docking
model between compound 38 and LPA2 (Figure ), which shows that
a key salt bridge interaction between the carboxylic acid group and
lysine 22 can take place in compound 13 but not in derivative 38 due to the binding conformation induced by the four-unit
spacer. In addition, a key hydrogen bond established between the phenolic
hydroxy group of compound 13 and arginine 107 is missing
in the binding of compound 38 to LPA2 (Figure ).
Table 2
Antagonist activities of compounds 13, 36–38, and Ki16425 at
LPA1-3
Emax (%)a [IC50 (μM)]b
compd
n
LPA1
LPA2
LPA3
13
1
N.E.c
84 ± 3 [5.5 ± 0.7]
N.E.
36
2
N.E.
67 ± 7
N.E.
37
3
N.E.
43 ± 12
N.E.
38
4
N.E.
21 ± 8
N.E.
Ki16425
97 ± 4 [0.8 ± 0.2]
92 ± 3 [1.2 ± 0.6]
99 ± 3 [1.6 ± 0.5]
Emax = maximum blockade effect of the activation induced by 10 μM
of LPA (18:1, 1-oleoyl-sn-glycerol-3-phosphate) at
a concentration of the compound under study of 10 μM.
For Emax > 70%, IC50 values are expressed as mean ± s.e.m,
from a minimum of two independent experiments, performed in triplicate.
N.E., no effect was observed
at
the highest concentration of compound tested (10 μM).
Figure 4
LPA1 (PDB ID 4Z35)-derived homology model of LPA2 in complex
with compound 13 (in white) and compound 38 (in purple). Docking of compound 13 in the model suggests
that the carboxylic acid of the compound is involved in two salt bridge
interactions with lysine 22 and lysine 278; however, derivative 38 cannot form the salt bridge with lysine 22. Docking results
also suggest that the phenolic hydroxy group of compound 13 establishes a hydrogen bond with arginine 107 and another one with
glutamine 108, whereas derivative 38 can only establish
hydrogen bonds with glutamine 108. For compound 13, salt
bridges are colored in yellow and hydrogen bonds in blue, whereas
for derivative 38, they are colored in orange and pink,
respectively.
Emax = maximum blockade effect of the activation induced by 10 μM
of LPA (18:1, 1-oleoyl-sn-glycerol-3-phosphate) at
a concentration of the compound under study of 10 μM.For Emax > 70%, IC50 values are expressed as mean ± s.e.m,
from a minimum of two independent experiments, performed in triplicate.N.E., no effect was observed
at
the highest concentration of compound tested (10 μM).LPA1 (PDB ID 4Z35)-derived homology model of LPA2 in complex
with compound 13 (in white) and compound 38 (in purple). Docking of compound 13 in the model suggests
that the carboxylic acid of the compound is involved in two salt bridge
interactions with lysine 22 and lysine 278; however, derivative 38 cannot form the salt bridge with lysine 22. Docking results
also suggest that the phenolic hydroxy group of compound 13 establishes a hydrogen bond with arginine 107 and another one with
glutamine 108, whereas derivative 38 can only establish
hydrogen bonds with glutamine 108. For compound 13, salt
bridges are colored in yellow and hydrogen bonds in blue, whereas
for derivative 38, they are colored in orange and pink,
respectively.Further confirmation of the importance of the carboxylic
acid interactions
was obtained with the synthesis of compounds 42–45, where the carboxylic acid moiety was replaced by hydroxy,
methoxy, methyl ester or carboxamide groups, respectively. The synthesis
of these compounds started from chromone 25, which by
reaction with hydrazine yielded pyrazole 42 that was
further methylated to give 43. Alternatively, O-alkylation of chromone 25 with methyl bromoacetate
or bromoacetamide followed by pyrazole ring formation yielded compounds 44 and 45 (Scheme ). Biological evaluation of all these compounds (Table ) revealed that only
the methyl ester derivative 44 showed a good activity
value (Emax = 74 ± 7%; IC50 = 11.6 ± 0.4 μM). To further discard the existence of
partial agonism, derivatives 42–45 were tested for their capacity to activate the LPA2 receptor
and none of them induced any appreciable effect at a concentration
of 10 μM.
Antagonist Activities of Compounds 13, 42–45, and Ki16425 at LPA1-3
Emax = maximum blockade effect of the activation induced by 10 μM
of LPA (18:1, 1-oleoyl-sn-glycerol-3-phosphate) at
a concentration of the compound under study of 10 μM.
For Emax > 70%, IC50 values are expressed as mean ± s.e.m,
from a minimum of two independent experiments, performed in triplicate.
N.E., no effect was observed
at
the highest concentration of compound tested (10 μM).
Synthesis of Compounds 42–45
Reagents and conditions:
(a)
65% N2H4·H2O, EtOH, reflux,
30 min, 65–80%; (b) CH3I, K2CO3, acetone, 65 °C, 6 h, 20%; (c) methyl bromoacetate or bromoacetamide,
K2CO3, acetone, reflux, 3 h, 51–67%;
(d) 2 M NaOH, EtOH, reflux, 12 h, 90%; (e) CH3OH, cat.
H2SO4, reflux, 16 h, 85%.Emax = maximum blockade effect of the activation induced by 10 μM
of LPA (18:1, 1-oleoyl-sn-glycerol-3-phosphate) at
a concentration of the compound under study of 10 μM.For Emax > 70%, IC50 values are expressed as mean ± s.e.m,
from a minimum of two independent experiments, performed in triplicate.N.E., no effect was observed
at
the highest concentration of compound tested (10 μM).We next studied the effect of changes in the pyrazole
ring. Specifically,
we introduced an N-methyl group (compound 47), removed the methyl group located at position 5 of the heterocycle
(compound 48), and replaced the pyrazole by an isoxazole
ring (derivative 49). Direct methylation reaction of 13 provided N-methylated analogue 47 (Scheme ), whereas
the preparation of compound 48 started with the reaction
of intermediate 17 with methanesulfonyl chloride, in
the presence of the boron trifluoride diethyl etherate complex to
obtain the corresponding hydroxychromone 50. Alkylation
of this intermediate with ethyl bromoacetate and reaction with hydrazine,
afforded desired pyrazole 48 (Scheme ). Finally, isoxazole analogue 49 was obtained from chromone 29 using hydroxylamine (Scheme ). Biological evaluation
of the compounds (Table ) revealed the importance of the methyl group in position 5 of the
pyrazole ring for the antagonist activity, since derivative 48 exhibited a moderate Emax value
of 35%. With respect to derivatives 47 and 49, they showed good activity at LPA2 but also a decrease
in selectivity, since they display some antagonist character at LPA3 (Table ).
None of them showed any activity as LPA2 agonists at a
concentration of 10 μM.
Scheme 5
Synthesis of Compounds 47-49
Reagents and conditions:
(a)
CH3I, NaH, THF, rt, 16 h, 31%; (b) CH3SO2Cl, BF3·Et2O, DMF, 100 °C,
1.5 h, 94%; (c) ethyl bromoacetate, K2CO3, acetone,
reflux, 3 h, 91%; (d) (i) 65% N2H4·H2O, EtOH, reflux, 30 min, 99%; (ii) 2 M NaOH, EtOH, reflux,
12 h, 58%; (e) (i) NH2OH·HCl, pyridine, EtOH, 85 °C,
12 h; (ii) p-toluenesulfonic acid, EtOH, 78 °C,
5 h, 40%; (f) 1 M NaOH, 1,4-dioxane, 60 °C, 16 h, 91%.
Table 4
Antagonist Activities of Compounds 13, 47–49, and Ki16425 at LPA1-3
Emax (%)a [IC50 (μM)]b
compd
R
R′
X
LPA1
LPA2
LPA3
13
Me
H
N
N.E.c
84 ± 3 [5.5 ± 0.7]
N.E.
47
Me
Me
N
N.E.
73 ± 10
47 ± 7
48
H
H
N
N.E.
35 ± 8
N.E.
49
Me
H
O
N.E.
58 ± 3
71 ± 4
Ki16425
97 ± 4 [0.8 ± 0.2]
92 ± 3 [1.2 ± 0.6]
99 ± 3 [1.6 ± 0.5]
Emax = maximum blockade effect of the activation induced by 10 μM
LPA (18:1, 1-oleoyl-sn-glycerol-3-phosphate) at a
concentration of the compound under study of 10 μM.
For Emax > 70% and selectivity at LPA2 receptor, IC50 values are expressed as mean ± s.e.m, from a minimum of two
independent experiments, performed in triplicate.
N.E., no effect was observed at
the highest concentration of compound tested (10 μM).
Synthesis of Compounds 47-49
Reagents and conditions:
(a)
CH3I, NaH, THF, rt, 16 h, 31%; (b) CH3SO2Cl, BF3·Et2O, DMF, 100 °C,
1.5 h, 94%; (c) ethyl bromoacetate, K2CO3, acetone,
reflux, 3 h, 91%; (d) (i) 65% N2H4·H2O, EtOH, reflux, 30 min, 99%; (ii) 2 M NaOH, EtOH, reflux,
12 h, 58%; (e) (i) NH2OH·HCl, pyridine, EtOH, 85 °C,
12 h; (ii) p-toluenesulfonic acid, EtOH, 78 °C,
5 h, 40%; (f) 1 M NaOH, 1,4-dioxane, 60 °C, 16 h, 91%.Emax = maximum blockade effect of the activation induced by 10 μM
LPA (18:1, 1-oleoyl-sn-glycerol-3-phosphate) at a
concentration of the compound under study of 10 μM.For Emax > 70% and selectivity at LPA2 receptor, IC50 values are expressed as mean ± s.e.m, from a minimum of two
independent experiments, performed in triplicate.N.E., no effect was observed at
the highest concentration of compound tested (10 μM).In sum, these results indicated that derivative 13 was the best compound identified so far. Hence, we studied
its pharmacokinetic
profile. First, we estimated the membrane permeability using the parallel
artificial membrane permeability assay (PAMPA) and its metabolic stability
in mouse and human liver microsomes (MLMs and HLMs, respectively).
In these assays, compound 13 showed a moderate permeability
value (P) of 0.11 × 10–6 cm/s, considering
as reference values P < 1 × 10–7 cm/s for low permeable compounds and P > 1 ×
10–5 cm/s for highly permeable molecules. The metabolic
stability was also moderate, with a half-life (t1/2) of about 60 min in HLMs and 16 min in MLMs. Hence, it
would be desirable to improve these parameters to obtain an optimized
compound suitable for in vivo efficacy experiments.
Optimization of Compound 13
We initially
addressed the optimization of derivative 13 with the
replacement of chlorine atoms by fluorine in compound 53 (Scheme ), as this
change usually involves an improvement of the pharmacokinetic parameters.[20,21] Also, considering that the free carboxylic acid could be responsible
for the moderate permeability, it was replaced by the (bio)isostere
tetrazole (compounds 54 and 55, Scheme ). Tetrazole is among
the most commonly employed carboxylic acid isosteres[22] because its planarity and acidity closely resemble those
of carboxylic acids (pKa = 4.5–4.9).
In addition, tetrazolate anions are more lipophilic than the corresponding
carboxylates and they exhibit slightly different electrostatic potential
and charge distribution due to the delocalization of the negative
charge over the five-membered ring system. Then, synthesis of difluorinated
derivative 53 was carried out following a similar route
to the one described for compound 13 but starting with
2,4-difluorophenoxyacetic acid (Scheme ). With respect to the tetrazole derivatives 54 and 55, they were prepared by alkylation of
the intermediate chromones 25 and 58, respectively,
with 2-bromoacetonitrile followed by sequential treatment with hydrazine
and sodium azide to build the corresponding pyrazole and tetrazole
rings, respectively (Scheme ).
Reagents and conditions:
(a)
(i) SOCl2, toluene, 110 °C, 16 h, 99%; (ii) resorcinol,
BF3·Et2O, DCM, reflux, 4 h, 10%; (b) acetic
anhydride, Et3N, NaOAc, 140 °C, 2.5 h, 71%; (c) conc.
HCl, EtOH, reflux, 2 h, 99%; (d) methyl bromoacetate, K2CO3, acetone, reflux, 3 h, 99%; (e) 65% N2H4·H2O, EtOH, reflux, 30 min, 21–99%;
(f) 2 M NaOH, EtOH, reflux, 12 h, 61%; (g) bromoacetonitrile, K2CO3, acetone, reflux, 3–5 h, 77–93%;
(h) NaN3, NH4Cl, DMF, reflux, 16 h, 52–72%.Biological evaluation of compounds 53–55 indicated that derivative 54 showed the best
results, being the most potent LPA2 antagonist, with an Emax of 90% and an IC50 value of 1.9
μM (Figure S2), values that are superior
to the ones showed by its analogue 13 (Table ). None of these compounds showed
any agonist activity at LPA2 (see Figure
S1 for the result obtained for compound 54, representative
of the rest of the compounds).
Table 5
Antagonist Activities of Compounds 13, 53–55, and Ki16425 at
LPA1-3
Emax = maximum blockade effect of the activation induced by 10 μM
of LPA (18:1, 1-oleoyl-sn-glycerol-3-phosphate) at
a concentration of the compound under study of 10 μM.
For Emax > 60%, IC50 values are expressed as mean ± s.e.m,
from a minimum of two independent experiments, performed in triplicate.
N.E., no effect was observed
at
the highest concentration of compound tested (10 μM).
Emax = maximum blockade effect of the activation induced by 10 μM
of LPA (18:1, 1-oleoyl-sn-glycerol-3-phosphate) at
a concentration of the compound under study of 10 μM.For Emax > 60%, IC50 values are expressed as mean ± s.e.m,
from a minimum of two independent experiments, performed in triplicate.N.E., no effect was observed
at
the highest concentration of compound tested (10 μM).Docking studies of these compounds showed how the
docking pose
of compound 54 is very similar to that of compound 13 by replacing the carboxylate moiety with its tetrazole
ring (Figure A). In
fact, the tetrazole moiety of compound 54 perfectly reproduces
the interactions of the carboxylic acid of compound 13, substituting salt bridges for hydrogen bonds with lysines 22 and
278. With respect to the replacement of chlorine by fluorine in compounds 53 and 55, the dichloro and difluorophenoxy moieties
lie in the same hydrophobic pocket (Figure B). However, the substitution of chlorine
by fluorine atoms provokes a change in the orientation of the aromatic
ring and hinders the difluorophenoxy moiety from simultaneously reaching
residues leucine 111 and alanine 284 as observed for compounds 13 and 54 through their two chlorine atoms (Figures and 5B). To experimentally validate the proposed docking model,
we carried out point mutation experiments to confirm the importance
of the most relevant residues (lysines 22 and 278, arginine 107, and
glutamine 108). Thus, we transfected McA-RH7777 cells with plasmids
containing the corresponding N terminus HA-tagged LPA2 mutant
(K22A, R107A, Q108A, or K278A). Transfection efficacy was assessed
by flow cytometry using a primary antibody against HA and the appropriate
fluorescent secondary antibody (Figure A) and the antagonist capacity of compound 54 was determined in cells transfected with each mutant. The obtained
results indicate that replacement of any of the four amino acids (lysines
22 and 278, arginine 107 and glutamine 108) by alanine involved the
lost of the antagonist activity of compound 54 (Figure B), thus confirming
the importance of the proposed interactions. The data indicate that
substitution of lysine 22, arginine 107, and glutamine 108 by alanine
basically abolished the capacity of compound 54 to bind
LPA2 receptor (since no significant agonist nor antagonist
activity was observed in the mutant receptors). However, the exchange
of lysine 278 by alanine completely switched the functional activity
of the receptor since compound 54 behaved as an agonist
in this mutant receptor. Overall, these data suggest that amino acids
22, 107, and 108 are important for binding, whereas lysine 278 seems
to be involved in the functional activity of the receptor.
Figure 5
(A) LPA1 (PDB ID 4Z35)-derived homology model of LPA2 in complex with compounds 13 (in white) and 54 (in turquoise). The tetrazole
moiety of compound 54 interacts in a similar manner to
the carboxylic acid group of compound 13. Thus, the carboxylic
acid forms salt bridges with lysines
22 and 278 (in yellow) and the tetrazole ring establishes hydrogen
bonds with these two residues (in blue). The rest key interactions
are maintained in both compounds. (B) LPA1-derived homology
model of LPA2 in complex with compounds 13 (in white), 53 (in orange), 54 (in turquoise),
and 55 (in yellow). The dichlorophenoxy and difluorophenoxy
moieties of compounds 13 and 53–55 lie in the same hydrophobic pocket. However, the substitution
of chlorine for fluorine provokes a change in the orientation of the
aromatic ring and prevents compounds 53 and 55 from reaching residues leucine 111 and alanine 284.
Figure 6
(A) Cell surface expression of each mutant LPA2 receptor
was assessed by flow cytometry using an anti-HA antibody raised in
mice followed by an antimouse antibody conjugated to Alexa 488 in
McA-RH7777 cells transiently transfected with mock plasmid (CTL, control)
or with plasmids containing the indicated LPA2 mutant receptor.
(B) Capacity of compound 54 (10 μM) to block the
activation induced by 10 μM of LPA (18:1, 1-oleoyl-sn-glycerol-3-phosphate) in the indicated mutant LPA2 receptor.
Values are expressed as mean ± s.e.m, from two independent experiments
performed in triplicate. Values obtained for the four point mutations
have differences statistically significant (p <
0.01) with respect to the value obtained for the wild type (WT) receptor. #Compound 54 behaved as an agonist of the LPA2 K278A mutant (able to induce 75 ± 5% activation at 10
μM, being the stimulation produced by 10 μM LPA normalized
to 100%).
(A) LPA1 (PDB ID 4Z35)-derived homology model of LPA2 in complex with compounds 13 (in white) and 54 (in turquoise). The tetrazole
moiety of compound 54 interacts in a similar manner to
the carboxylic acid group of compound 13. Thus, the carboxylic
acid forms salt bridges with lysines
22 and 278 (in yellow) and the tetrazole ring establishes hydrogen
bonds with these two residues (in blue). The rest key interactions
are maintained in both compounds. (B) LPA1-derived homology
model of LPA2 in complex with compounds 13 (in white), 53 (in orange), 54 (in turquoise),
and 55 (in yellow). The dichlorophenoxy and difluorophenoxy
moieties of compounds 13 and 53–55 lie in the same hydrophobic pocket. However, the substitution
of chlorine for fluorine provokes a change in the orientation of the
aromatic ring and prevents compounds 53 and 55 from reaching residues leucine 111 and alanine 284.(A) Cell surface expression of each mutant LPA2 receptor
was assessed by flow cytometry using an anti-HA antibody raised in
mice followed by an antimouse antibody conjugated to Alexa 488 in
McA-RH7777 cells transiently transfected with mock plasmid (CTL, control)
or with plasmids containing the indicated LPA2 mutant receptor.
(B) Capacity of compound 54 (10 μM) to block the
activation induced by 10 μM of LPA (18:1, 1-oleoyl-sn-glycerol-3-phosphate) in the indicated mutant LPA2 receptor.
Values are expressed as mean ± s.e.m, from two independent experiments
performed in triplicate. Values obtained for the four point mutations
have differences statistically significant (p <
0.01) with respect to the value obtained for the wild type (WT) receptor. #Compound 54 behaved as an agonist of the LPA2 K278A mutant (able to induce 75 ± 5% activation at 10
μM, being the stimulation produced by 10 μM LPA normalized
to 100%).Noteworthy, compound 54 kept the receptor
selectivity
versus LPA1 and LPA3 (Table ), so it emerged as an excellent candidate
for in-depth pharmacological characterization.
In-Depth Characterization of Compound 54
First, we determined the membrane permeability using the PAMPA assay
and the in vitro metabolic stability of the compound. The obtained
results showed a good permeability value (P = 6.1
× 10–6 cm/s) and also increased stability in
comparison with analog 13, with t1/2 values of 50 ± 6 and 97 ± 15 min for MLMs and
HLMs, respectively (Table ).
Table 6
In Vitro Pharmacokinetic Profile of
Compounds 13 and 54
stability
(t1/2, min)a
compd
MLMs
HLMs
P (cm/s)b
clogPc
13
16 ± 8
61 ± 10
0.1 × 10–6
4.14
54
50 ± 6
97 ± 15
6.1 × 10–6
3.29
Data for stability in mouse and
human liver microsomes (MLMs and HLMs, respectively) are expressed
as the mean ± s.e.m. from five independent experiments performed
in duplicate.
P = permeability
value; reference values consider P < 10–7 cm/s for low permeability compounds and P >
10–5 cm/s for highly permeable molecules.
Values obtained with the ACDLabs Percepta software (version 6.0).
Data for stability in mouse and
human liver microsomes (MLMs and HLMs, respectively) are expressed
as the mean ± s.e.m. from five independent experiments performed
in duplicate.P = permeability
value; reference values consider P < 10–7 cm/s for low permeability compounds and P >
10–5 cm/s for highly permeable molecules.Values obtained with the ACDLabs Percepta software (version 6.0).These in vitro values correlated with the results
obtained in the
in vivo pharmacokinetic (PK) study, which was carried out to determine
the suitability of the compound to reach the CNS in therapeutically
relevant doses. For this aim, compound 54 was administered
intraperitoneally (i.p.) at a dose of 25 mg/kg. Then, at different
postinjection times (between 0.5 and 4 h), plasma, brain, and spinal
cord samples were taken and the levels of compound 54 were measured using high-performance liquid chromatography coupled
to mass spectrometry (HPLC-MS). These experiments confirmed the presence
of compound 54 at significant levels in spinal cord and
brain, with the maximum levels reached at one hour postadministration.
These data (Table ) indicate that antagonist 54 can readily cross the
blood brain barrier and it is therefore an excellent candidate to
validate the role of LPA2 antagonism, at least as a proof
of principle, in an in vivo model of SCI.
Table 7
In Vivo Levels of Compound 54 at Different Post-injection Times
concentration
of compound 54 (ng/mg
tissue) after the indicated postinjection time (h)a
sample
0.5
1
2
4
plasma
130 ± 50
41 ± 5
38 ± 1
ND
spinal
cord
0.40 ± 0.06
3.3 ± 0.3
NDb
ND
brain
0.10 ± 0.02
28 ± 4
0.08 ± 0.01
ND
Mice received a single injection
of compound 54 (25 mg/kg, i.p.). Samples were taken at
the indicated postinjection times and immediately frozen, and the
levels of compound were then analyzed by HPLC-MS. Data are the means
± s.e.m. from three independent samples.
ND, not detected.
Mice received a single injection
of compound 54 (25 mg/kg, i.p.). Samples were taken at
the indicated postinjection times and immediately frozen, and the
levels of compound were then analyzed by HPLC-MS. Data are the means
± s.e.m. from three independent samples.ND, not detected.In addition, the binding affinity of the compound
for LPA2 was evaluated by means of a free solution assay-compensated
interferometric
reader (FSA-CIR) technique,[23−26] showing a binding equilibrium constant (KD) value of 1.3 nM. As a positive control, LPA showed
a KD value of 6.7 nM to LPA2. The analogous assay carried out for LPA4–6 provided
10-fold selectivity versus LPA6 and >50-fold selectivity
versus LPA4 and LPA5 (Figure
S3), making compound 54 (UCM-14216) the most potent
and selective LPA2 antagonist described so far.
In Vivo Efficacy Study of Compound 54 in an SCI
Mouse Model
Since LPA2 activation plays harmful
actions after SCI, we finally assessed whether compound 54 protects against locomotor deficits in a spinal cord contusion injury
model. It has been established that the LPA2 receptor is
constitutively expressed at very low levels in spinal cord and its
transcripts are up-regulated during the first days after injury, returning
to basal levels by day 7.[13] This suggests
that LPA-LPA2 signaling in the injured spinal cord mainly
occurs during the first week postinjury. Hence, we hypothesized that
administration of compound 54 for 10 days could block
LPA-LPA2 signaling in the injured spinal cord, and consequently,
improve the outcome of SCI, as observed after genetic deletion of Lpa.[13] The in vivo PK study suggested that an i.p. dose of 25 mg/kg was
enough to reach significant levels of the compound one hour after
administration (3.3 ng/mg tissue are equivalent approximately to a
concentration of 2 μM in the spinal cord considering the volume
of the sections used in the study). It is conceivable that this concentration
is even higher in the injured mice, as SCI results in increases permeability
of the blood-spinal cord barrier.[27] Then,
this dose was selected as the minimal capable of potentially eliciting
the sought biological effects and simultaneously avoiding side effects
related with the use of higher concentrations. Accordingly, mice were
treated daily with compound 54 (25 mg/kg, i.p.) starting
at 1 h following lesion and subsequently for 10 consecutive days,
and locomotor performance was assessed by using the Basso Mouse Scale
(BMS). BMS is the gold standard test used for locomotor scoring after
SCI in which two blinder observers score the mouse motor performance
based on a nine-point scale.[28] As shown
in Figure A, mice
treated with compound 54 displayed significant improvement
in locomotor recovery after SCI. Bonferroni’s post hoc analysis
revealed that motor skills were significantly enhanced in the injured
mice that had been treated with compound 54 for 10 days
at 25 mg/kg, from day 35 postinjury onward. At the end of the follow
up (day 50 postinjury) mice treated with vehicle showed plantar placement
of the hind paw but no weight-bearing stepping (BMS score 3.0 ±
0.2). In contrast, mice treated with compound 54 displayed
occasional or frequent stepping (BMS of 4.1 ± 0.3). We do not
discard that the therapeutic actions of the compound 54 could be enhanced with more frequent administration (i.e., twice
a day), longer duration or greater dose of the compound.
Figure 7
Compound 54 significantly improves locomotor recovery
after SCI in LPA2 wild type mice but not in LPA2 deficient mice. Effect of i.p. injection of 54 (25
mg/kg) or vehicle on locomotor recovery in (A) C57bl/6 mice and (B)
LPA2 deficient mice quantified using the 0 (absence of
movement) to 9 (completely normal locomotor behavior) Basso Mouse
Scale (BMS). Data are expressed as mean ± s.e.m. and correspond
to seven animals per group in A and five animals per group in B. * p < 0.05 compared with vehicle-treated group (two-way
repeated measures ANOVA with Bonferroni’s post hoc test for
multiple comparisons).
Compound 54 significantly improves locomotor recovery
after SCI in LPA2 wild type mice but not in LPA2 deficient mice. Effect of i.p. injection of 54 (25
mg/kg) or vehicle on locomotor recovery in (A) C57bl/6 mice and (B)
LPA2 deficient mice quantified using the 0 (absence of
movement) to 9 (completely normal locomotor behavior) Basso Mouse
Scale (BMS). Data are expressed as mean ± s.e.m. and correspond
to seven animals per group in A and five animals per group in B. * p < 0.05 compared with vehicle-treated group (two-way
repeated measures ANOVA with Bonferroni’s post hoc test for
multiple comparisons).Importantly, the observed locomotor improvement
is largely mediated
by the action of the compound at the LPA2 receptor, because
administration of the same dose of compound to LPA2 null
mice undergoing SCI did not induce any significant effect (Figure B). These results
clearly validate the LPA2 receptor as a valuable therapeutic
target for the treatment of SCI.
Conclusions
In this work, we report the synthesis of
the most potent and selective
LPA2 antagonist identified to date, compound 54 (UCM-14216), with functional Emax and
IC50 values of 90% and 1.9 μM, respectively, and
a KD value of 1.3 nM at LPA2 and functional selectivity against other LPA receptors (LPA1,3–6). In addition, compound 54 has a
good pharmacokinetic profile both in vitro and in vivo, reaching pharmacologically
relevant levels in the CNS, where the site of action is located. Furthermore,
it shows efficacy in an acute in vivo mouse model of SCI being inactive
in LPA2 knockout mice undergoing the same model, thus supporting
the involvement of LPA2 in the secondary damage that follows
SCI. SCI mainly affects to young and otherwise healthy adults, who
suffer from a lack of efficacious treatments. Current treatments are
generally palliative, limited to analgesic and anti-inflammatory drugs,
underscoring high medical need for new pharmacological strategies
that might be accessed by LPA2 antagonists to ameliorate
SCI physiopathology and improve neurological outcomes. Further study
of UCM-14216 and other related compounds could provide novel approaches
to treat SCI and possibly other traumatic CNS injuries.
Experimental Section
Synthesis
Unless stated otherwise, starting materials,
reagents, and solvents were purchased as high-grade commercial products
from Sigma-Aldrich, Alfa Aesar, Acros, Fluka, Panreac or Scharlab,
and were used without further purification. Dichloromethane (DCM)
and tetrahydrofurane (THF) were dried using a Pure Solv Micro 100
Liter solvent purification system. Triethylamine and pyridine were
dried over KOH and distilled prior to its use. All nonaqueous reactions
were performed under an argon atmosphere in oven-dried glassware unless
otherwise stated. MW irradiation reactions were carried out on a Biotage
Initiatior 2.5 reactor, using Biotage vials sealed with an aluminum/Teflon
crimp top, which can be exposed to a maximum of 250 °C and 20
bar internal pressure.Analytical thin-layer chromatography
(TLC) was run on Supelco silica gel plates (silica gel 60 F254) with detection by UV light (254 nm) and 5% ninhydrin solution in
ethanol or 10% phosphomolybdic acid solution in ethanol. Products
were purified by flash chromatography on glass columns using silica
gel (60 Å pore size, 230–400 mesh particle size from Supelco)
or using a Varian 971-FP system with cartridges of silica gel (Varian,
50 μm size particle).All compounds were obtained as oils,
except for those whose melting
points (mp) are indicated, which were solids. Mp values were determined
on a Stuart Scientific electrothermal apparatus. Infrared (IR) spectra
were measured on a Bruker Tensor 27 instrument equipped with a Specac
ATR accessory of 5200–650 cm–1 transmission
range; frequencies (ν) are expressed in cm–1.Nuclear magnetic resonance (NMR) spectra were recorded at
rt on
a Bruker Avance III 700 MHz (1H, 700 MHz; 13C, 175 MHz), Bruker Avance 500 MHz (1H, 500 MHz; 13C, 125 MHz) or Bruker DPX 300 MHz (1H, 300 MHz; 13C, 75 MHz) instrument at the Universidad Complutense de Madrid
(UCM) NMR core facility. 19F-NMR spectra were recorded
on a Bruker DPX 300 MHz. Chemical shifts (δ) are expressed in
parts per million relative to the residual solvent peak for 1H and 13C nucleus (CDCl3: δH = 7.26, δC = 77.2; MeOH-d4: δH = 3.31, δC = 49.0;
DMSO-d6: δH = 2.50, δC = 39.5), and coupling constants (J) are
in hertz (Hz). The following abbreviations are used to describe peak
patterns when appropriate: s (singlet), d (doublet), t (triplet),
q (quadruplet), m (multiplet), br (broad), and app (apparent). 2D
NMR experiments (COSY, HMQC, and HMBC) of representative compounds
were carried out to assign protons and carbons of new structures;
for those carbons displaying very broad signals in 13C
NMR spectra, the corresponding chemical shifts were established by
their correlation peaks in HSQC and HMBC spectra (Figure S4 shows the numbered structures used in the structural
characterization by NMR of all final compounds). High-resolution mass
spectrometry (HRMS) was carried out on a FTMS Bruker APEX Q IV spectrometer
in electrospray ionization (ESI) or matrix-assisted laser desorption
ionization (MALDI) mode at UCM’s mass spectrometry facilities.For all final compounds, purity was determined by HPLC-MS and satisfactory
chromatograms confirmed a purity of at least 95%. HPLC-MS analysis
was performed using an Agilent 1200LC-MSD VL instrument. LC separation
was achieved with an Eclipse XDB-C18 (5 μm, 4.6 mm
× 150 mm) or a Zorbax SB-C3 column (5 μm, 2.1
mm × 50.0 mm) together with a guard column (5 μm, 4.6 mm
× 12.5 mm). Mobile phase consisted of A (95:5 water/acetonitrile)
and B (5:95 water/acetonitrile) with 0.1% formic acid as solvent modifier.
Gradients are indicated in Table S1. MS
analysis was performed with an ESI source. The capillary voltage was
set to 3.0 kV and the fragmentor voltage was set at 72 eV. The drying
gas temperature was 350 °C, the drying gas flow was 10 L/min,
and the nebulizer pressure was 20 psi. Spectra were acquired in positive
or negative ionization mode from 100 to 1000 m/z and in UV-mode at four different wavelengths (210, 230,
254, and 280 nm).
General Procedure 1: Friedel–Crafts Acylation
(a) Preparation of the aryloxyacetyl chloride: to a solution of the
corresponding aryloxyacetic acid (1 equiv.) in anhydrous toluene (5.5
mL/mmol) was added thionyl chloride (2.8 mL/mmol) and the reaction
mixture was refluxed for 16 h. After this time, the excess of thionyl
chloride and toluene were evaporated under reduced pressure, affording
the corresponding aryloxyacetyl chloride in quantitative yield. (b)
Friedel–Crafts acylation: to a cooled (0 °C) stirred solution
of the corresponding freshly prepared aryloxyacetyl chloride (1 equiv.)
and resorcinol (1.1 equiv.) in anhydrous DCM (1.5 mL/mmol), boron
trifluoride diethyl etherate (1.3 mL/mmol) was added. The reaction
was stirred at 0 °C for 10 min and then at 90 °C until starting
material was consumed (TLC, 4–5 h). The reaction vessel was
then cooled in an ice bath and the mixture poured into an excess of
ice water. The aqueous phase was extracted with DCM (×2), and
the combined organic layers were washed with brine, dried over Na2SO4, filtered and concentrated under reduced pressure.
The residue was purified by flash chromatography to yield the corresponding
2,4-dihydroxyphenylethanones 6, 14–17, and 56.
General Procedure 2: Synthesis of Chromones by Kostanecki–Robinson
Acylation
A mixture of the corresponding 2,4-dihydroxyphenylethanone
(1 equiv.), freshly distilled acetic anhydride (0.6 mL/mmol), triethylamine
(0.8 mL/mmol), and anhydrous sodium acetate (2.4 equiv.) was stirred
at 140 °C until the reaction was completed (TLC, 2–3 h).
Afterward, cold water was added and the mixture was extracted with
DCM (×2). The combined organic phases were washed with brine,
dried over Na2SO4, filtered, and concentrated
under reduced pressure, affording the corresponding acetoxychromones 7, 18–21, and 57, which were used without further purification.
General Procedure 3: Hydrolysis of Acetoxychromone Derivatives
To a solution of the appropriate acetoxychromone (1 equiv.) in
the minimum amount of absolute ethanol was added conc. HCl (0.6 mL/mmol),
and the reaction was refluxed for 2 h. After cooling to rt, the mixture
was diluted with ethyl acetate and washed with a saturated aqueous
solution of NaHCO3 and brine. The organic phase was dried
over Na2SO4, filtered, and concentrated under
reduced pressure, affording the corresponding hydroxychromones 8, 22–25, and 58, which were used without further purification.
General Procedure 4: Alkylation of Hydroxychromone Derivatives
To a solution of the corresponding hydroxychromone (1 equiv.) in
anhydrous acetone (15 mL/mmol) was added K2CO3 (2 equiv), and the reaction mixture was refluxed for 30 min. Then,
a solution of the appropriate bromoderivative (1.1–4.3 equiv.)
in anhydrous acetone (1 mL/mmol) was added and the mixture was refluxed
until consumption of starting material (TLC, 3–5 h). Next,
cold water was added and acetone was removed under reduced pressure.
The aqueous residue was extracted with DCM (×2), and the combined
organic phases were washed with brine, dried with Na2SO4, filtered, and concentrated under reduced pressure. The residue
was purified by flash chromatography to afford the corresponding alkylated
chromones 5, 9, 26–29, 40, 41, 46, 51, 59, 60, and 61.
General Procedure 5: Synthesis of Pyrazole Derivatives by Reaction
with Hydrazine
A solution of the corresponding chromone or
enaminone (1 equiv.) in absolute ethanol (5 mL/mmol) at 40 °C
was treated with a solution of hydrazine monohydrate (65%, 0.18 mL/mmol)
in absolute ethanol (1.3 mL/mmol), and the mixture was refluxed until
the reaction was completed (TLC, 0.5–2 h). After cooling to
rt, the mixture was concentrated under reduced pressure and the residue
was dissolved with ethyl acetate and acidified with 1 M HCl until
pH 6. The aqueous phase was extracted with ethyl acetate (×2),
and the combined organic layers were washed with brine, dried over
Na2SO4, filtered, and concentrated under reduced
pressure. The residue was purified by flash chromatography to yield
the corresponding pure pyrazoles 33, 42, 45, 62, and 63. For those chromones
or enaminones bearing an ester group, the resulting hydrazide derivative
was taken to next step (General Procedure 6) without further purification.
General Procedure 6: Hydrolysis of Hydrazide Derivatives
To a solution of the corresponding hydrazide obtained according to General Procedure 5 (1 equiv.) in the minimum
amount of 96% ethanol was added 2 M NaOH (0.8 mL/mmol), and the reaction
was refluxed for 12 h. After cooling to rt, the mixture was diluted
with ethyl acetate and acidified with 1 M HCl until pH 6. The aqueous
phase was extracted with ethyl acetate (×2) and the combined
organic layers were washed with brine, dried over Na2SO4, filtered, and concentrated under reduced pressure. The residue
was purified by flash chromatography to yield target carboxylic acid
derivatives 2, 3, 10–13, 36–38, 48, and 53.
General Procedure 7: Synthesis of Tetrazole Derivatives
To a solution of the corresponding nitrile (1 equiv.) in anhydrous
DMF (15 mL/mmol), NH4Cl (1.5 equiv.), and NaN3 (1.5 equiv.) were added, and the reaction was refluxed overnight.
Then, the mixture was filtered to remove salts, and the resulting
solution was acidified until pH 3 with 1 M HCl and extracted with
ethyl acetate (×2). The combined organic phases were washed with
a 1:1 mixture of water/brine, dried with Na2SO4, filtered, and concentrated under reduced pressure. The residue
was purified by flash chromatography to afford the corresponding tetrazoles 54 and 55.
Evaluation of Receptor Activation by Ca2+ Mobilization
Assay
Cells stably expressing the corresponding LPA1–3 receptor were grown as described previously.[18] Changes in intracellular calcium levels were measured by
using the fluorescent calcium sensitive dye Fluo-4 NW (Invitrogen).
RH7777 cells or B103 cells were plated on poly-d-lysine or
collagen coated, respectively, black-wall clear-bottom 96-well plates
(Corning) at a density of 50 000 cells/well and cultured overnight.
The culture medium was then replaced with 100 μL of Fluo-4 NW
dye loading solution containing 2.5 μM of probenecid and incubated
for 30 min at 37 °C followed by an additional 30 min at rt. Then,
20 μL of the test compound from a 6× stock solution in
assay buffer were added and fluorescence was measured during 120 s
after which 10 μM of LPA (18:1, 1-oleoyl-sn-glycerol-3-phosphate) was added and wells were monitored for additional
120 s. Fluorescence changes were registered in a FluoStar Optima instrument
(BMG Labtech) at 525 nm using an excitation wavelength of 494 nm.
Ca2+ transient increase was quantified by calculating the
difference between maximum (stimulation with LPA 10 μM) and
baseline values for each well, and antagonist activity was quantified
by determining the percentage of the signal suppression caused by
the compound under study with respect to the Ca2+ increase
induced by LPA (which was considered 100%). As positive controls,
10 μM LPA and 10 μM ionomycin were included in every experiment.
At this concentration, LPA induced a response about 30–33%
of the one shown by ionomycin, which is in agreement with previously
described results.[29] The data presented
are from two to four independent experiments carried out in triplicate
or quadruplicate. Dose–response curves were generated and IC50 values calculated by nonlinear regression analysis using Prism software version 5 (GraphPad Software Inc., San Diego,
CA, USA).The agonist activity at LPA2 receptors
was determined at 10 μM concentration for all final compounds
as previously described.[18]
Binding Affinity at LPA4–6
A free
solution assay,[23,24] where the lysophosphatidic acid
receptor (LPA2, 4–6) containing nanovesicles
(of 110–130 nm size as measured by dynamic light scattering)
and compound under study are freely moving into solution was prepared
to determine the equilibrium binding constants (KD) in a native environment of the binding partners (ligand/compound–receptor).
The assay was analyzed using a benchtop Compensated Interferometric
Reader (CIR) that measured the light refractive index (ΔRI)
change from binding-induced conformational and/or hydration changes
produced by real time binding events in a sample (receptor containing
nanovesicles plus compound) and compared to a nonbinding reference
(RI matched buffer plus compound).[30] Finally,
the interferometric signal from vector nanovesicles binding (nonspecific)
to compound was subtracted from the LPA2,4–6 containing
nanovesicles binding (total) to compound, to determine the specific
binding interactions of the compounds to LPA2,4–6. The concentration-dependent change in RI (ΔRI) signal from
the compounds to LPA2,4–6 or vector was fitted using
the single site total vs nonspecific binding isotherm using GraphPad Prism. Specific KD values were
determined by fitting the total minus nonspecific signal to a single
site binding isotherm.The detailed free solution assay method
was described previously.[25,26] All the compounds were
dissolved in 100% DMSO, aliquoted, and frozen at −80 °C
for 1 week. The compound dilution series was freshly prepared in 0.5%
DMSO/PBS (pH 7.4) to keep the maximum compound in solution. In the
final assay, total protein concentration was maintained at 25 μg/mL
with 0, 0.08, 0.4, 2, 10, 50, and 250 nM concentrations of the compound
in a final buffer composition of 0.25% DMSO/PBS. Receptor-compound
mixture was incubated at rt for about an hour on a shaker and then
filled in a dropix sample well tray in the format of reference then
sample and finally introduced to CIR using an automated Mitos Dropix
(Dolomite Microfluidics, UK) sample introducer. The detailed description
of the CIR was mentioned elsewhere.[31,32] It is a benchtop
RI reader that combined a compensated interferometer with a Mitos
Dropix (an automated droplet generator) and a syringe pump. The compensated
interferometer, which consisted of a diode laser, one or two mirrors,
one glass capillary, and a CCD camera, measures the RI change from
a solution undergoing conformational and/or hydration alteration compared
to a reference (with no such binding events). ΔRI is measured
by capturing the translational shifts in backscattered light interference
fringes produced from the interaction between an expanded beam profile
of the laser and a capillary filled with droplets of sample-reference
solutions. The positional shift of the backscattered fringes, which
is equivalent to molecular interaction, was quantified using fast
Fourier transform of selected bright fringes captured in a CCD camera.
The data acquisition and analysis were performed using a LabVIEW interface
designed at the laboratory.
In Silico Experiments
Docking calculations were performed
using Autodock4(33) [using:
ga_num_evals (depending on the number of rotatable bonds) = 6 310 000
(for compounds 3, 13, 53, 54, 55) and 25 000 000 (for compound 38), ga_run = 100 and all the other parameters set to their
default values]. The LPA2 receptor model was generated
using SwissModel[34] and the crystal structure
with PDB ID 4Z35(19) as a template. The generated model
was prepared for docking using pdb 2pqr(35,36) with the propka[37,38] protonation option at a pH of 7.4 and the peoepb force field.[39] All the analyzed compounds were modeled using RDKIT (Open-source cheminformatics) and its protonation
state adjusted at pH 7.4 by the ChemAxon cxcalc module (command line
version of ChemAxon’s Calculator Plugins,
v16.10.24.0, 2016). Binding mode pictures were created using PyMOL v2.5.0.
Mutagenesis Experiments
Amino terminal hemagglutinin
(HA)-tagged LPA2 point mutants K22A, R107A, Q108A, and
K278A containing pcDNA3.1 plasmids were provided by GenScript. For
expression of the different constructs, McA-RH7777 (CRL-1601, ATCC)
cells were selected, as they have been previously used for point mutation
experiments of LPA receptors.[40] Cells were
grown in Dulbecco’s modified Eagle’s medium supplemented
with 10% (v/v) fetal bovine serum and 1% (v/v) penicillin-streptomycin
and kept at 37 °C and 5% CO2. Cells were transiently
transfected with the different plasmids using lipofectamine and following
the manufacturer’s procedure. Successful transfection was confirmed
by flow cytometry analysis. For these experiments, 1 × 105 cells were resuspended in 50 μL of PBS with 2 mM EDTA
and 0.5% BSA. Anti-HA antibody (Santa Cruz, sc-7392; 1 μg per
1 × 106 cells) was added and cells were incubated
for 30 min at rt with shaking. Cells were centrifuged, washed with
buffer, and incubated with antimouse Alexa Fluor 488 (Invitrogen,
1:5000) for 30 min at rt with shaking and protected from light. Cells
were centrifuged, washed with buffer, resuspended in 0.3 mL of buffer
and analyzed by flow cytometry in a FACScalibur instrument (Becton
Dickinson) at the UCM’s microscopy and flow cytometry unit.
After confirming the transfection by flow cytometry, calcium mobilization
experiments were carried out as previously described.
Permeability and Microsomal Stability
These studies
were carried out as previously described with minor modifications.[41,42] The assessment of the membrane permeability of synthesized compounds
and propranolol and metoprolol as reference compounds was performed
in a commercially available 96-well Corning Gentest precoated PAMPA
plate system (Cultek S.L.U., Spain). Prior to use, the precoated PAMPA
plate system was warmed to rt for 30 min and 300 μL of 200 μM
solution of tested compound in 2% DMSO in PBS (pH 7.4) were added
into wells in the receiver (donor) plate. Then, 200 μL of PBS
were added into wells in the filter (acceptor) plate. The filter plate
was placed on the receiver plate by slowly lowering the precoated
PAMPA plate until it sits on the receiver plate. The assembly was
incubated at rt for 5 h, and then buffer samples were collected carefully
from each plate. The final concentrations of compound in both donor
and acceptor wells were analyzed by HPLC-MS and quantification was
estimated by using the peak area integration normalized with an internal
standard. Permeability value of the compounds was calculated using
the following formula: P (cm/s) = {−ln[1 – CA(t)/Ceq]}/[A(1/VD +
1/VA)t], where A = filter area (0.3 cm2), VD = donor well volume (0.3 mL), VA = acceptor well volume (0.2 mL), t = incubation
time (s), CA(t) = compound
concentration (μM) in the acceptor well at time t, CD(t) = compound concentration
(μM) in donor well at time t, and Ceq = [CD(t)VD + CA(t)VA]/(VD + VA). Assays were performed
in duplicate, and the compound was tested in two different plates
on different days.For measuring the stability in mouse and
human liver microsomes, compounds were incubated at 37 °C at
a final concentration of 1 or 5 μM in PBS, respectively, together
with a solution of nicotinamide adenine dinucleotide phosphate (NADPH)
in PBS (final concentration of 2 mM) and a solution of MgCl2 in PBS (final concentration of 5 mM). Reactions were initiated by
the addition of a suspension of mouse liver microsomes (MLMs) (male
CD-1 mice pooled, Sigma-Aldrich) or human liver microsomes (HLMs)
(male human pooled, Sigma-Aldrich), respectively, at a final protein
concentration of 1 mg/mL. The solutions were vortexed and incubated
at 37 °C. Aliquots of 100 μL were quenched at time zero
and at seven points ranging to 2 h (MLM) or 4 h (HLM) by pouring into
100 μL of ice-cold acetonitrile. Quenched samples were centrifuged
at 10 000g for 10 min, and the supernatants
were filtered through a polytetrafluoroethylene (PTFE) membrane syringe
filter (pore size of 0.2 μm, 13 mm in diameter, GE Healthcare
Life Sciences). The relative disappearance of the compound under study
over the course of the incubation was monitored by HPLC-MS using SIM
mode. Concentrations were quantified by measuring the area under the
peak ([M + H]+) normalized with an internal standard and
converted to the percentage of compound remaining, using the time
zero peak area value as 100%. The natural logarithm of the remaining
percentage versus time data for each compound was fit to a linear
regression, and the slope was used to calculate the degradation half-life
(t1/2).
Determination of the In Vivo Levels of Compound
Compound 54 was administered intraperitoneally (25 mg/kg) in adult
female 12–16 weeks old C57Bl/6J mice. At 1, 2, and 4 h after
drug administration (n = 3 for each time and sample),
mice were sacrificed and their brains, spinal cords, and blood were
obtained. The brain and spinal cord were immediately frozen and kept
at −80 °C until analysis. Blood was allowed to clot at
rt for 30 min and centrifuged at 4 °C for 10 min at 16 000g. Serum was transferred to a clean polypropylene tube and
stored at −80 °C until analysis. For analysis, a volume
of cold acetonitrile was added to the serum. The sample was incubated
in an ice bath for 10 min and centrifuged at 4 °C for 10 min
at 16 000g. The resulting organic layer was
filtered through a PTFE filter (0.2 μm, 13 mm diameter, Fisher
Scientific) and 20 μL of the sample analyzed by LC-MS/MS at
the UCM’s Mass Spectrometry CAI. Separation was performed using
a Phenomenex Gemini 5 μm C18 110A 150 × 2 mm column (run
time 8 min; flow 0.5 mL/min; gradient: 0.5 min 10% Phase B, 2 min
60% Phase B, 4.5–6 min 100% Phase B, 7–8 min 10% Phase
B; Phase A: water with formic acid 0.1%; Phase B: acetonitrile). The
entire LC eluent was directly introduced to an electrospray ionization
(ESI) source operating in the positive ion mode for LC MS/MS analysis
on a Shimadzu LCMS8030 triple quadrupole mass spectrometer coupled
to UHPLC with an oven temperature of 31.5 °C. The mass spectrometer
ion optics were set in the multiple reaction monitoring mode and the
transition selected for quantification was 432.90 > 215.10 (CE:
−30
V).
Spinal Cord Injury In Vivo Model
All Surgical Procedures
Were Approved by the Universitat Autònoma De Barcelona Animal
Care Committee (CEEAH 4273) and followed the guidelines of the European
Commission on Animal Care (EU Directive 2010/63/EU). Adult female
C57Bl/6J mice (10–12 weeks old) and LPA2 null mice
were anesthetized by intramuscular injection with a mixture of ketamine
and xylazine (90:10 mg/kg). A laminectomy was performed at the 11th
thoracic vertebrae and the exposed spinal cord was contused using
the Infinite Horizon Impactor device (Precision Scientific Instrumentation)
using a force of 60 kdynes. Only mice showing a spinal cord tissue
displacement ranging between 450 and 550 μm were selected. One
hour after injury, compound 54 was injected intraperitoneally
(25 mg/kg) which was then repeated daily for 10 consecutive days.
Authors: James I Fells; Ryoko Tsukahara; Yuko Fujiwara; Jianxiong Liu; Donna H Perygin; Daniel A Osborne; Gabor Tigyi; Abby L Parrill Journal: Bioorg Med Chem Date: 2008-04-18 Impact factor: 3.641
Authors: Darryl J Bornhop; Michael N Kammer; Amanda Kussrow; Robert A Flowers; Jens Meiler Journal: Proc Natl Acad Sci U S A Date: 2016-03-09 Impact factor: 11.205
Authors: Ji Woong Choi; Deron R Herr; Kyoko Noguchi; Yun C Yung; Chang-Wook Lee; Tetsuji Mutoh; Mu-En Lin; Siew T Teo; Kristine E Park; Alycia N Mosley; Jerold Chun Journal: Annu Rev Pharmacol Toxicol Date: 2010 Impact factor: 13.820
Figure 1
Figure 2
Scheme 1
Figure 3
Scheme 2
Scheme 3
Scheme 4
Scheme 5
Scheme 6
Figure 6
Figure 7