The G protein-coupled chemokine receptors CXCR1 and CXCR2 play key roles in inflammatory diseases and carcinogenesis. In inflammation, they activate and recruit polymorphonuclear cells (PMNs) through binding of the chemokines CXCL1 (CXCR1) and CXCL8 (CXCR1 and CXCR2). Structure-activity studies that examined the effect of a novel series of S-substituted 6-mercapto-N-phenyl-nicotinamides on CXCL1-stimulated Ca(2+) flux in whole human PMNs led to the discovery of 2-[5-(4-fluorophenylcarbamoyl)pyridin-2-ylsulfanylmethyl]phenylboronic acid (SX-517), a potent noncompetitive boronic acid CXCR1/2 antagonist. SX-517 inhibited CXCL1-induced Ca(2+) flux (IC50 = 38 nM) in human PMNs but had no effect on the Ca(2+) flux induced by C5a, fMLF, or PAF. In recombinant HEK293 cells that stably expressed CXCR2, SX-517 antagonized CXCL8-induced [(35)S]GTPγS binding (IC50 = 60 nM) and ERK1/2 phosphorylation. Inhibition was noncompetitive, with SX-517 unable to compete the binding of [(125)I]-CXCL8 to CXCR2 membranes. SX-517 (0.2 mg/kg iv) significantly inhibited inflammation in an in vivo murine model. SX-517 is the first reported boronic acid chemokine antagonist and represents a novel pharmacophore for CXCR1/2 antagonism.
The G protein-coupled chemokine receptors CXCR1 and CXCR2 play key roles in inflammatory diseases and carcinogenesis. In inflammation, they activate and recruit polymorphonuclear cells (PMNs) through binding of the chemokines CXCL1 (CXCR1) and CXCL8 (CXCR1 and CXCR2). Structure-activity studies that examined the effect of a novel series of S-substituted 6-mercapto-N-phenyl-nicotinamides on CXCL1-stimulated Ca(2+) flux in whole human PMNs led to the discovery of 2-[5-(4-fluorophenylcarbamoyl)pyridin-2-ylsulfanylmethyl]phenylboronic acid (SX-517), a potent noncompetitive boronic acidCXCR1/2 antagonist. SX-517 inhibited CXCL1-induced Ca(2+) flux (IC50 = 38 nM) in human PMNs but had no effect on the Ca(2+) flux induced by C5a, fMLF, or PAF. In recombinant HEK293 cells that stably expressed CXCR2, SX-517 antagonized CXCL8-induced [(35)S]GTPγS binding (IC50 = 60 nM) and ERK1/2 phosphorylation. Inhibition was noncompetitive, with SX-517 unable to compete the binding of [(125)I]-CXCL8 to CXCR2 membranes. SX-517 (0.2 mg/kg iv) significantly inhibited inflammation in an in vivo murine model. SX-517 is the first reported boronic acid chemokine antagonist and represents a novel pharmacophore for CXCR1/2 antagonism.
The chemokine receptors
CXCR1 and CXCR2 are closely related members
of the class A (rhodopsin-like) family of seven transmembrane G-protein-coupled
receptors.[1] The chemokine CXCL8 (Interleukin-8,
IL8) activates the receptors CXCR1 and CXCR2, whereas the chemokine
CXCL1 (growth related oncogene α, GROα) is a selective
agonist for CXCR2.[2] CXCR1/2 signaling is
sensitive to the pertussis toxin, indicating involvement of the Gαi
subunit in the heterotrimeric G-protein.[3,4] Agonist-induced
changes in receptor conformation uncouple the Gβγ subunit
from the heterotrimeric G-protein complex, activating signaling pathways
that include phospholipase Cβ, phosphatidylinositol-3-kinase,
and mitogen-activated protein kinases. Phospholipase Cβ in turn
generates inositol-1,4,5-triphosphate, which binds to the endoplasmic
reticulum and leads to a release of Ca2+ into the cytoplasm.[5] CXCR1/2 signaling is involved in inflammation,
wound healing, and angiogenesis, and their dysregulation has been
implicated in a myriad of diseases involving acute and chronic inflammation,[6−14] as well as tumorigenesis.[15−20] In particular, CXCR1/2 signaling mediates agonist-induced neutrophil
activation and recruitment to sites of inflammation (i.e., chemotaxis)
and is therefore thought to play an important role in inflammatory
diseases characterized by a significant neutrophil component.Due to the involvement of these receptors in a wide range of inflammatory
diseases and carcinogenesis, CXCR1 and CXCR2 have attracted attention
as targets for small-molecule drug discovery (Figure 1).[21] Reparixin 1 is
a ketoprofen derivative being investigated in trials for the prevention
and treatment of delayed graft function and pancreatic islet transplantation.[22,23] Diarylureas exemplified by 2 (SB225002) have been disclosed
as either selective CXCR2 antagonists[24,25] or dual CXCR1/2
antagonists.[26] The central urea motif in
the diarylureas was later replaced with the cyclic urea bioisostere
3,4-diaminocyclobut-3-ene-1,2-dione to provide potent CXCR2-selective
analogues as represented by SCH527123 (3).[27] The diarylurea SB656933 has advanced into clinical
trials for chronic obstructive pulmonary disease (COPD)[28,29] and cystic fibrosis,[30] and SCH527123
inhibited ozone inhalation-induced sputum neutrophil recruitment in
healthy subjects.[31] AZD-8309 (4) is representative of the bicyclic thiazolopyrimidine class of CXCR2
antagonists,[32] and this antagonist effectively
inhibited the increase of LPS-mediated neutrophil recruitment in the
nasal lavage of healthy subjects.[33]
Figure 1
CXCR1 and CXCR2
receptor antagonists.
CXCR1 and n class="Gene">CXCR2
receptor antagonists.
As disclosed by Cutshall and co-workers, nicotinamide glycolate
esters exemplified by the methyl ester 5 (Figure 1) are antagonists of CXCR2-mediated human neutrophil
chemotaxis and are distinct from the diarylurea and related cyclobutane
classes.[34] We have previously shown that
CXCR2 antagonism by nicotinamide glycolate esters proceeded through
a novel intracellular mechanism that required hydrolytic cleavage
of the ester within the neutrophil for activity.[35] However, the unique pharmacology of this class also led
to rapid degradation in plasma, making it untenable as a therapeutic.
The mechanistic insights gleaned from these studies inspired us to
search for new nicotinamide templates that would directly antagonize
CXCR2 without requiring intracellular hydrolytic activation for their
activity. Herein, we report the discovery, structure–activity
relationship (SAR), in vitro pharmacology, and in vivo biologic activity
of 7 (SX-517). Compound 7 is a potent noncompetitive
CXCR1/2 antagonist resulting from a novel series of S-substituted
6-mercapto-N-phenyl-nicotinamides active in their
native form, and is the first reported boronic acid chemokine antagonist.
Results
and Discussion
The compounds described in this study are
shown in Tables 1–4, and their synthetic
methods are outlined in Schemes 1–4.
Table 1
CXCL1-Inhibitory
Activity of Carboxylate
and Carboxylate Isostere Substituted Thionicotinamides
Each IC50 determination
was performed in triplicate with ≥6 concentrations, and reported
as the mean ± standard error.
Table 4
Effect of N-Substituted
6-(Thiobenzyl-2-borono)-nicotinamides on CXCL1-Induced Ca2+ Flux in Human Neutrophils
Each IC50 determination
was performed in triplicate with ≥6 concentrations and reported
as the mean ± standard error.
Scheme 1
Synthesis of S-Substituted N-(4-Fluoro-phenyl)-6-mercapto-nicotinamides 9–45, 47–48, 52, and 55
Conditions:
(i) 4-fluoroaniline,
EEDQ, DMF, rt; (ii) method A: Br-CH2–R, solid-phase
base (N-methylmorpholino-substituted resin), DMF,
60 °C; method B: Br-CH2–R or Cl-CH2–R, DMF, base (TEA, DIPEA, or), rt; method C: Br-CH2–R, EtOH, 1 N NaOH, reflux.
Scheme 4
Synthesis of N-Substituted 6-(Thiobenzyl-2-borono)-nicotinamides 60–63
Conditions: (i) 4-fluoroaniline,
K2CO3, THF, rt; (ii) N-Boc-2-piperidinecarbaldehyde,
NaBH(OAc)3, DCE, glacial AcOH, rt; (iii) tert-butyl bromoacetate, DIPEA, DMF, 80 °C; (iv) phase-transfer
reaction: bromomethylpyridine, toluene, 50% aq NaOH, TBAH, rt; (v)
6-chloronicotinoyl chloride, DBU, DMF, heat; (vi) NaHS, DMF, heat;
(vii) 2-bromomethyl-phenylboronic acid, EtOH, 1 N aq NaOH, reflux;
(viii) 2-bromomethyl-phenylboronic acid, DMF, TEA; (ix) 4 M HCl, dioxane,
rt; and (x) 90% aq TFA, rt.
Synthetic Strategies and Focused Parallel
Combinatorial Synthesis
The synthesis of S-substituted N-(4-fluorophenyl)-6-mercapto-nicotinamides
(7–53) was carried out as described
in Scheme 1 by first condensing 6-thio-nicotinic
acid with 4-fluoroaniline using the coupling reagent 2-ethoxy-1-ethoxycarbonyl-1,2-dihydroquinoline
(EEDQ). The resulting N-(4-fluorophenyl)-6-mercapto-nicotinamide 64 was then S-alkylated with a commercially available alkyl
halide, typically an alkyl bromide, using one of three methods that
differed primarily in their solvent and base. Method A utilized the
thio-nicotinamide intermediate 64 and bromomethyl building
blocks (Maybridge Chemical Co., Cornwall, U.K.) dissolved in anhydrous
DMF in the presence of resin-bound tertiary amine (4-methylmorpholino
polystyrene resin, NovaBiochem, La Jolla, CA). The reaction mixture
was heated at 60 °C for 1 h, followed by thiol scavenger resin
(mercaptomethyl polystyrene resin, NovaBiochem, La Jolla, CA) addition.
The suspension was heated to 60 °C for 2 h, and then the reaction
mixture was filtered and the filtrate diluted with water. The resulting
precipitate was collected by centrifugation. Purity of the synthesized
compounds was assessed by HPLC, and the identification of compounds
was done by electrospray ionization mass spectroscopy (ESI-MS). In
all cases, the major reaction product was the derivatized thionicotinamide.
Method A was used to initially synthesize compounds 13–43 in moderate to excellent purity by facile
filtration. The synthesized compounds were screened at initial test
concentrations of 5 and 10 μM for antagonism of CXCL1-mediated
intracellular Ca2+ release in isolated human neutrophils
(hPMNs). Compounds that exhibited greater than 50% inhibition at 5
μM were then resynthesized by either methods B or C and re-evaluated
in the assay to obtain IC50 values of these selected compounds
at a higher tested purity. Method B utilized an alkyl halide and a
tertiary amine [typically triethylamine (TEA), diisopropylethylamine
(DIPEA), or N-methylmorpholine (NMM)] to give the
final compounds 8–13, 15, 17, 22, 43, 46, 48, and 50. Method C utilized an alkyl
halide and aqueous NaOH in EtOH at reflux to give the final compounds 7, 30, and 53 in good yield.
Synthesis of S-Substituted N-(4-Fluoro-phenyl)-6-mercapto-nicotinamides 9–45, 47–48, 52, and 55
Conditions:
(i) 4-fluoroaniline,
n class="Chemical">EEDQ, DMF, rt; (ii) method A: Br-CH2–R, solid-phase
base (N-methylmorpholino-substituted resin), DMF,
60 °C; method B: Br-CH2–R or Cl-CH2–R, DMF, base (TEA, DIPEA, or), rt; method C: Br-CH2–R, EtOH, 1 N NaOH, reflux.
Further
chemical modification of intermediates to final test compounds
is shown in Scheme 2. Treatment of 7 by potassium peroxymonosulfate (Oxone), according to the procedure
of Webb and Levy for the hydroxylation of aryl boronic acids,[36] afforded compound 44. Saponification
of the methyl esters 46, 48, and 50 with NaOH afforded carboxylic acids 47, 49, and 51, respectively. The tetrazoles 8 and 52 were prepared via the cycloaddition of the cyano
intermediates with trimethylsilyl azide and dibutyltin oxide[37] in toluene under reflux.
Scheme 2
Synthesis of S-Substituted N-(4-Fluoro-phenyl)-6-mercapto-nicotinamides 8, 44, 47, 49, 51–52
Conditions:
(i) TMS-N3, Bu2SnO, toluene, reflux; (ii) 1
N NaOH, MeOH, rt; and
(iii) oxone, aq NaHCO3, acetone, 4 °C
Synthesis of S-Substituted N-(4-Fluoro-phenyl)-6-mercapto-nicotinamides 8, 44, 47, 49, 51–52
Conditions:
(i) TMS-N3, n class="Chemical">Bu2SnO, toluene, reflux; (ii) 1
N NaOH, MeOH, rt; and
(iii) oxone, aq NaHCO3, acetone, 4 °C
The synthesis of N-substituted 6-(thiobenzyl-2-borono)-nicotinamides 54–59 was performed as in Scheme 3. 6-Thio-nicotinic acid was coupled to 2-bromomethyl-phenylboronic
acid using a modified method B employing TEA, DMF, and elevated temperature
(60 °C) to give intermediate 67. Intermediate 67 was treated with neopentyl glycol in toluene under reflux
to afford the protected neopentyl boronate ester. After cooling to
0 °C, pivaloyl chloride was added with triethylamine (TEA), and
the reaction proceeded for an hour and then with gradual warming to
room temperature for another hour to give the pivaloyl mixed anhydride
intermediate 68, which was used without purification.
Intermediate 68 was then coupled immediately to either
4-trifluoromethoxyaniline, 4-aminobenzoic acid, 5-(4-aminophenyl)-1H-tetrazole,
4-amino-pyridine, 4-hydroxy-aniline, or 2-amino-5-fluoropyridine under
heating to give the corresponding 6-(thiobenzyl-2-neopentyl boronate
ester)-nicotinamides. Exposure to water liberated the boronic acid
moiety from the neopentyl boronate ester, which after purification
by preparative HPLC afforded 54–59.
Scheme 3
Synthesis of Anilide Derivatives of 6-(Thiobenzyl-2-borono)-nicotinamides 54–59
Conditions: (i) 2-bromomethyl-phenylboronic
acid, DMF, TEA, 60 °C; (ii) neopentyl glycol, toluene, reflux;
(iii) pivaloyl chloride, TEA, toluene, 0 °C to rt; and (iv) aniline
derivative, TEA, DMF, 60 °C.
Synthesis of Anilide Derivatives of 6-(Thiobenzyl-2-borono)-nicotinamides 54–59
Conditions: (i) 2-bromomethyl-phenylboronic
acid, n class="Chemical">DMF, TEA, 60 °C; (ii) neopentyl glycol, toluene, reflux;
(iii) pivaloyl chloride, TEA, toluene, 0 °C to rt; and (iv) aniline
derivative, TEA, DMF, 60 °C.
The synthesis
of N,N-disubstituted6-(thiobenzyl-2-borono)-nicotinamides 60–63 was performed through two alternative
routes as described in Scheme 4. In one route, 6-chloronicotinoyl chloride was
amidated with the secondary amines 70a and 70b, using DBU as the base to give the N,N-disubstituted6-chloronicotinamides 71a and 71b, respectively.
The secondary amine 70a was derived from reductive amination
of N-Boc-2-piperidinecarbaldehyde (Chem-Impex, Wood
Dale, IL) with 4-fluoroaniline and sodium triacetoxyborohydride. The
secondary amine 70b was derived from the coupling of
4-fluoroaniline with tert-butyl bromoacetate in DMF
with DIPEA as base. In the alternate route, 6-chloronicotinoyl chloride
was amidated with 4-fluoroaniline in THF in the presence of potassium
carbonate to form intermediate 69. Compounds 72a–72b were synthesized from intermediate 69 under phase transfer catalyzed conditions, utilizing either
2-bromomethylpyridine or 4-bromomethylpyridine, respectively. The
tertiary amides 71a–71b and 72a–72b were then subjected to the same
treatment with sodium hydrogen sulfide (NaHS) in DMF to displace the
pyridinyl chloride and form the corresponding thio-nicotinamide intermediates 73a–73d, respectively. The resultant thio-nicotinamide
intermediates 73a and 73b were alkylated
with 2-bromomethyl-phenylboronic acid using method C to afford 60 and 61, respectively. Method C applied to 73c followed by treatment with 4 M HCl in dioxane deprotected
the Boc-piperidyl moiety to yield 62. Method B applied
to 73d followed by acid hydrolysis of the tert-butyl protected carboxyl group with 90% TFA yielded 63.
Synthesis of N-Substituted 6-(Thiobenzyl-2-borono)-nicotinamides 60–63
Conditions: (i) 4-fluoroaniline,
K2CO3, n class="Chemical">THF, rt; (ii) N-Boc-2-piperidinecarbaldehyde,
NaBH(OAc)3, DCE, glacial AcOH, rt; (iii) tert-butyl bromoacetate, DIPEA, DMF, 80 °C; (iv) phase-transfer
reaction: bromomethylpyridine, toluene, 50% aq NaOH, TBAH, rt; (v)
6-chloronicotinoyl chloride, DBU, DMF, heat; (vi) NaHS, DMF, heat;
(vii) 2-bromomethyl-phenylboronic acid, EtOH, 1 N aq NaOH, reflux;
(viii) 2-bromomethyl-phenylboronic acid, DMF, TEA; (ix) 4 M HCl, dioxane,
rt; and (x) 90% aq TFA, rt.
CXCL1-Stimulated Ca2+ Flux in Human Neutrophils
An increased flux of
intracellular Ca2+ represents a
key signaling event in CXCL1-induced neutrophil activation through
CXCR2.[2] Since nicotinamide-based chemokine
antagonists act intracellularly,[35] the
activity of a compound will depend not only on its target affinity
but also on its ability to transit into the intracellular compartment.
To simultaneously capture both properties in initial SAR evaluations,
all compounds prepared were evaluated for their ability to inhibit
CXCL1-stimulated Ca2+ flux in whole hPMNs. Full dose–response
curves were determined for select compounds using at least seven concentrations,
and the resulting IC50 values listed in Tables 1–4 as part of these
SAR studies are the mean of at least three determinations.
Structure–Activity
Studies
The nicotinamide
glycolate methyl ester 5 antagonized CXCL1-stimulated
Ca2+ flux by an intracellular mechanism that required hydrolytic
cleavage of the ester within the hPMN to liberate the active species.[35] The required hydrolytic cleavage of the ester
led to instability in plasma that was inseparable from the intrinsic
activation mechanism of this pharmacophore class. Since in vivo stability
of a potential therapeutic is essential, novel CXCR2 antagonists were
sought that would directly antagonize the receptor without the need
for hydrolytic activation and, thus, resist in vivo esterase activity.
To this end and based on the above activation mechanism involving
the glycolate ester moiety, new S-substituted nicotinamides and related
congeners were evaluated for antagonism of CXCL1-stimulated Ca2+ flux in whole hPMNs, as summarized in Tables 1–4.Each IC50 determination
was performed in triplicate with ≥6 concentrations, and reported
as the mean ± standard error.Our SAR efforts first involved the bioisosteric replacement
of
the ester/acid moiety of lead nicotinamides 5 and 6 (Table 1). Replacement of the carboxylate
moiety with either a cyclic ester (9), a sulfone (10), a nitro (11), or a phosphite (12) resulted in loss of activity in our assay. We previously demonstrated
that the negatively charged nicotinamide carboxylate 6 could not enter hPMNs, and its activity was only observed after
hPMNs were electropermeabilized.[35] On the
basis of this observation, we hypothesized that the S-methyl-tetrazole 8 might serve as a more lipophilic
bioisostere of the thioglycolate moiety that could passively enter
hPMNs. Its activity (IC50 = 3900 nM) supported this hypothesis,
but its potency was relatively weak in this capacity.In order
to widen our SAR efforts, a focused combinatorial library
screen was implemented. As we discussed above, this successful parallel
synthetic effort was made possible through the use of excess alkyl
bromide reagent and heat to drive the alkylation to completion, as
well as the use of solid phase scavenging resins to aid in product
isolation. The compounds (13–43)
were first screened for activity versus CXCL1-mediated calcium flux
in hPMNs at test concentrations of 5 and 10 μM. In our initial
screen, compounds 13, 15, 17, 20, 22, 30, and 43 exhibited greater than 50% inhibition of CXCL1-mediated intracellular
Ca2+ release at a test concentration of 5 μM and
were therefore identified as potential hits. These compounds were
then resynthesized and retested, and the results are shown in Tables 2 and 3. Upon resynthesis
and retesting, compounds 13, 15, 17, and 22 exhibited IC50 values greater than
5 μM. Inhibitory activity was confirmed for the tetrahydrofuran
derivative 20 (IC50 = 540 nM), the S-benzyl derivative 30 (IC50 = 390
nM), and the S-dichlorobenzyl derivative 43 (IC50 = 1610 nM). The activity of S-benzyl
derivative 30 was similar to the intrinsic activity of
the de-esterified carboxylate species (IC50 = 100–500
nM) of nicotinamide methyl ester 5 reported previously
in electropermeabilized hPMNs.[35] Removal
of the methylene between the sulfur and the aryl ring resulted in
loss of activity (45). The results from this focused
combinatorial chemistry effort indicated that the S-benzyl nicotinamide nucleus held potential as a new CXCR2 inhibitor
template.
Table 2
CXCL1-Inhibitory Activity of Heterocyclyl-
and Heteroalkyl-Substituted Thionicotinamides
Select IC50 determinations
were performed in triplicate with ≥6 concentrations, and reported
as the mean ± standard error.
Table 3
CXCL1-Inhibitory Activity of Aryl-Substituted
Thionicotinamides
Select
IC50 determinations
were performed in triplicate with ≥6 concentrations and reported
as the mean ± standard error.
Select IC50 determinations
were performed in triplicate with ≥6 concentrations, and reported
as the mean ± standard error.Select
IC50 determinations
were performed in triplicate with ≥6 concentrations and reported
as the mean ± standard error.Elaboration of the benzyl ring with a carboxyl group
resulted in
activity, with the 2-position (51, IC50 =
1480 nM) favored over both the 3-position (49, IC50 = 2120 nM) and the 4-position (47, IC50 > 10000 nM). We hypothesized that the reduced activity of 51 was the result of poor accumulation in the hPMN as a result
of the negatively charged carboxylate adversely affecting uptake,
efflux, or both. To examine this hypothesis further and because we
have shown previously that the nicotinamide methyl ester 5 can be de-esterified within hPMNs to liberate high concentrations
of the corresponding acid,[35] we examined
the uncharged S-methyl-benzoyl methyl esters 46, 48, and 50 as potential intracellular
precursors to the corresponding acids. There was a 2-fold improvement
in potency in the 2-methyl ester 50 relative to the corresponding
free acid 51, but the 3- and 4-methyl esters 48 and 46 were inactive. It is possible that the limited
improvement in potency arose because benzoyl methyl esters are poorer
substrates for intracellular esterase activity than the thioglycolatemethyl ester 5.To explore bioisosteric replacements
for the carboxylate in our
intermediate lead compound, the tetrazole derivative was synthesized
and evaluated for activity. Compound 52 (IC50 = 170 nM) exhibited an almost 10-fold increase in activity as compared
to the 2-carboxylate derivative 51. Our SAR efforts then
turned to carboxyl replacements at the 2-position. An important functionality
being explored in recent pharmaceutical development is the boronic
acid moiety.[38,39] Inclusion of a boronic acid at
the 2-postion of the S-benzyl nicotinamide scaffold
resulted in compound 7 exhibiting potent inhibition of
CXCR2 activation with an IC50 of 38 nM. Phenylboronic acid
has a pKa of 8.9,[40] and at neutral pH, compound 7 is expected to be mostly
uncharged and readily transit into the cell interior. However, to
test the possibility that greater lipophilicity would further increase
potency, we prepared the corresponding pinacol ester 53. Although acyclic and unhindered cyclic esters of boronic acids
are rapidly hydrolyzed in water, hydrolysis is slowed considerably
for hindered cyclic aliphatic esters such as the pinacol ester.[40] Remarkably, the pinacol ester derivative 53 retained the same order of activity as the parent boronic
acid, with its activity modestly reduced by approximately 7-fold (IC50 = 275 nM). Whether the pinacol ester was active at the target
or served to liberate the boronic acid inside the cell is unknown,
but importantly the enhanced lipophilicity of the boronyl ester offered
no advantage in potency over the underivatived boronyl group.Concluding that the S-benzyl-2-borono scaffold
in N-(4-fluorophenyl)-6-(thiobenzyl-2-borono)-nicotinamide 7 provided optimal nanomolar potency against CXCL1-induced
Ca2+ flux, we undertook separate focused SAR studies of
the apical 4-fluorophenyl-carboxamido domain (Table 4). The 4-fluoroanilide
moiety was previously investigated and optimized for a related class
of CXCR2 antagonists,[41] but our focus was
to introduce polar and/or ionizable substitutions that could potentially
increase aqueous solubility while retaining potent chemokine antagonism.
To accomplish this, the SAR of the apical 4-fluorophenyl-carboxamido-
moiety in 7 was explored in compounds 54–59, a series of R2 N-monosubstituted6-(thiobenzyl-2-borono)-nicotinamides
(Table 4). The results clearly revealed the
importance of the 4-fluoroanilide. Its replacement with carboxyl (55) or tetrazolyl (56) abrogated all activity.
Its replacement with trifluoromethoxy (54), a pyridinylnitrogen (57), hydroxyl (58), or inclusion
of a ring nitrogen (59) resulted in a 7-, 14-, 3-, and
8-fold reduction in activity, respectively. The N,N-disubstituted6-(thiobenzyl-2-borono)-nicotinamides 60–63 were explored by leaving the N-4-fluorophenyl R2 moiety constant and elaborating the R1 N-substitution. The N-methyl-2-pyridyl
(60) was found to be well-tolerated (IC50 =
78 nM), yielding an activity nearly equivalent to 7 and
providing evidence that the pyridyl group, which would be expected
to be uncharged at neutral pH, does not preclude entry of the molecule
into the cell. Surprisingly, the potency of the closely related congener 61 employing N-methyl-4-pyridyl was 11-fold
lower than 60, suggesting a model where the pyridyl moieties
are engaged in a highly structured and restricted environment. N-methyl-2-piperidyl 62 was explored as a nonplanar
and nonaromatic analog of 60. It was completely devoid
of activity, possibly because it was not accommodated at the target
or because its positive charge prevented passive diffusion into the
cell. A similar finding was obtained for the N-2-acetic
acid analog 63. None of the compounds 60–63 exhibited significantly improved aqueous
solubility relative to 7, despite attempts to prepare
conjugate salts (data not shown).Each IC50 determination
was performed in triplicate with ≥6 concentrations and reported
as the mean ± standard error.On the basis of its optimal in vitro potency in inhibiting
CXCL1-stimulated
Ca2+ release in the above SAR studies, compound 7 was further evaluated with respect to its in vitro signaling pharmacology
and in vivo biologic activity as described below.
G-Protein Coupling
to CXCR2
CXC receptors transduce
signals to the interior of the cell through activation of a coupled
heterotrimeric G-protein, the most proximal signaling event after
agonist binding to the receptor.[3,4] CXCR2 is bound and activated
by both CXCL1 and CXCL8.[2] The effect of
compound 7 on G-protein coupling to CXCR2 was evaluated
by [35S]GTPγS binding in CXCR2 membranes prepared
from HEK293 cells that stably expressed the human receptor (Figure 2). Compound 7 potently inhibited [35S]GTPγS binding in response to 10 nM CXCL8 with an
IC50 of 60 ± 7 nM (mean + S.E.).
Figure 2
Inhibition of chemokine-stimulated
[35S]GTPγS
binding by compound 7. HEK293 cells stably expressing
CXCR2 were incubated with different concentrations of compound (from
10–10 to 10–4 M) at 37 °C
for 60 min, and then the membranes were prepared by lysis and centrifugation.
The membranes were then incubated in buffer containing 50 μM
GDP, 8 nM [35S]GTPγS, and 10 nM CXCL8 at 30 °C
for another 60 min. Membranes were harvested by rapid filtration and
membrane-bound [35S]GTPγS quantitated. Data, expressed
as a percentage of basal [35S]GTPγS bound in the
absence of CXCL8, are the mean ± SE from three independent experiments,
each done in triplicate.
Inhibition of chemokine-stimulated
[35S]GTPγS
binding by compound 7. HEK293 cells stably expressing
CXCR2 were incubated with different concentrations of compound (from
10–10 to 10–4 M) at 37 °C
for 60 min, and then the membranes were prepared by lysis and centrifugation.
The membranes were then incubated in buffer containing 50 μM
GDP, 8 nM [35S]GTPγS, and 10 nM CXCL8 at 30 °C
for another 60 min. Membranes were harvested by rapid filtration and
membrane-bound [35S]GTPγS quantitated. Data, expressed
as a percentage of basal [35S]GTPγS bound in the
absence of CXCL8, are the mean ± SE from three independent experiments,
each done in triplicate.
CXCL8 Binding at CXCR2
We have shown that nicotinamide
glycolates act away from the orthosteric chemokine binding site to
antagonize CXCR2 through an intracellular mechanism.[35] Maximal inhibition of radioligand binding by an allosteric
antagonist can be observed in displacement experiments with radioligand
concentrations much lower than the Kd value.[42] We examined the ability of compound 7 to displace binding of [125I]-CXCL8 from CXCR2 membranes
using a radioligand concentration (25 pM) that was >10-fold below
the Kd for CXCL8 binding to CXCR2.[43] Although compound 7 potently inhibited
functional CXCR2 signaling by CXCL1 (IC50 = 38 nM, CXCL1-stimulated
Ca2+ flux) and CXCL8 (IC50 = 60 nM, CXCL8-stimulated
[35S]GTPγS binding), up to 10 μM failed to
compete the binding of [125I]-CXCL8 to CXCR2 membranes
(Figure 3). In parallel controls, unlabeled
CXCL8 isopotently displaced its homologous radioligand (IC50 = 28 pM). Collectively, these data support a model where compound 7 acts as a noncompetitive, intracellular allosteric inhibitor
of CXCR2. These findings mirrored those of the noncompetitive allosteric
inhibitor reparixin 1, which inhibited CXCR1 responses
with no effect on CXCL8 binding to CXCR1.[22,23]
Figure 3
Competition
binding assay with 7 and CXCL8 at human
CXCR2. Membranes from recombinant HEK293-CXCR2 cells were incubated
with the compound at the indicated concentrations and 25 pM [125I]-CXCL8. Radioligand binding to the membranes was measured
by scintillation. Data show the mean ± SD (n = 3) radioligand binding, expressed as the percent of control specific
radioligand binding with vehicle.
Competition
binding assay with 7 and CXCL8 at humanCXCR2. Membranes from recombinant HEK293-CXCR2 cells were incubated
with the compound at the indicated concentrations and 25 pM [125I]-CXCL8. Radioligand binding to the membranes was measured
by scintillation. Data show the mean ± SD (n = 3) radioligand binding, expressed as the percent of control specific
radioligand binding with vehicle.
Cell Surface Expression of CXCR2
CXCR2 activation by
chemokines is followed by receptor phosphorylation and subsequent
down-regulation; events that are accompanied by receptor internalization.[44] We considered the possibility that compound 7 may antagonize CXCR2 in whole cells at least partially through
a similar mechanism of sequestering receptor away from the cell membrane
signaling machinery. We therefore evaluated the effect of compound 7 on the CXCR2 surface expression in stably transfected HEK293
cells using a fluorescently labeled antibody to the receptor and fluorescence-activated
cell sorting. As shown in Figure 4, 60 min
exposure to 10 μM compound 7 did not significantly
alter the cell surface expression of CXCR2. These data together with
the data showing inhibition of CXCL8-stimulated [35S]GTPγS
binding are most consistent with a mechanism of antagonism involving
direct blockade of receptor activation.
Figure 4
Effect of compound 7 on the cell surface expression
of CXCR2. HEK293 cells stably expressing CXCR2 were pretreated with
1% DMSO (vehicle) or 10 μM compound (cpd. 7) for
60 min. HEK293 cells not expressing CXCR2 served as a negative isotype
control (isotype). All cells were then incubated with R-phycoerythrin (PE)-conjugated antihuman CXCR2 mouse
monoclonal antibody at 4 °C for 60 min. Cells were washed, fixed
in 2% formaldehyde in PBS, and subjected to flow cytometric fluorescence-activated
cell sorting (FACS) analysis of the PE signal. Results are representative
of three independent experiments.
Effect of compound 7 on the cell surface expression
of CXCR2. HEK293 cells stably expressing CXCR2 were pretreated with
1% DMSO (vehicle) or 10 μM compound (cpd. 7) for
60 min. HEK293 cells not expressing CXCR2 served as a negative isotype
control (isotype). All cells were then incubated with R-phycoerythrin (PE)-conjugated antihumanCXCR2mouse
monoclonal antibody at 4 °C for 60 min. Cells were washed, fixed
in 2% formaldehyde in PBS, and subjected to flow cytometric fluorescence-activated
cell sorting (FACS) analysis of the PE signal. Results are representative
of three independent experiments.
CXCR2 MAPK Signaling
CXCR1 and CXCR2 mediate downstream
signaling in part through MAPK activation.[4,45] To
assess the ability of compound 7 to inhibit MAPK in HEK293
cells that stably expressed CXCR2, we measured ERK1/2 phosphorylation
in response to CXCL8 in the presence of vehicle or 10 μM compound 7. As shown in the vehicle arm in Figure 5, CXCR2 induced a time-dependent phosphorylation of ERK1/2
upon activation by CXCL8. Maximum response was obtained at 15 min.
In contrast, compound 7 completely blocked CXCR2-mediated
phosphorylation of ERK1/2 by 30 min. There was no effect on the amount
of ERK1/2 present in the lysates (Figure 5,
lower panel) as assessed by an anti-ERK1/2 antibody recognizing ERK1/2
irrespective of its phosphorylation state. These data are consistent
with receptor blockade by compound 7 as the common event
that abrogates signaling pathways downstream of CXCR2 involving G-proteins,
MAP kinases, and intracellular Ca2+.
Figure 5
Effect of Compound 7 on CXCR2-induced ERK1/2 phosphorylation.
HEK293 cells stably expressing CXCR2 were stimulated with CXCL8 (100
ng/mL) for 0–30 min in the presence of vehicle or 10 μM
compound 7. Phosphorylated ERK1/2 and total ERK were
determined by Western blotting using antiphospho-ERK1/2 (P-ERK1/2) and antitotal ERK1/2 (T-ERK1/2) antibodies, respectively.
Data are representative of three independent experiments performed
in triplicate.
Effect of Compound 7 on CXCR2-induced ERK1/2 phosphorylation.
HEK293 cells stably expressing CXCR2 were stimulated with CXCL8 (100
ng/mL) for 0–30 min in the presence of vehicle or 10 μM
compound 7. Phosphorylated ERK1/2 and total ERK were
determined by Western blotting using antiphospho-ERK1/2 (P-ERK1/2) and antitotal ERK1/2 (T-ERK1/2) antibodies, respectively.
Data are representative of three independent experiments performed
in triplicate.
Selectivity of Antagonism
CXCR1 and CXCR2 are expressed
in equal numbers on the surface of hPMNs.[46] Whereas CXCL8 activates both CXCR1 and CXCR2, CXCL1 is a selective
agonist for CXCR2.[2] Compound 7 inhibited both CXCL1- and CXCL8-stimulated Ca2+ flux
in hPMNs with IC50 values of 38 and 36 nM, respectively
(Table 5).
Table 5
Effect of 7 on Agonist-Induced
Ca2+ Flux
IC50 (nM)a
agonist
hPMN
HEK293-CXCR1
HEK293-CXCR2
CXCL1
38 ± 3
NDb
ND
CXCL8
36 ± 11
880 ± 90
210 ± 40
C5a
>5000
PAFc
>5000
fMLFd
>5000
Each IC50 determination
was performed in triplicate with ≥6 concentrations and reported
as the mean ± standard error.
ND = not determined.
Platelet activating factor.
Formyl-Met-Leu-Phe.
Each IC50 determination
was performed in triplicate with ≥6 concentrations and reported
as the mean ± standard error.ND = not determined.Platelet activating factor.Formyl-Met-Leu-Phe.This
inhibition was not partial, since a concentration of 1 μM
compound 7 was sufficient to completely eliminate CXCL8-stimulated
Ca2+ flux in hPMNs. The total antagonism of CXCL8-induced
Ca2+ flux by compound 7 presumably reflected
a dual blockade of both CXCR1 and CXCR2 signaling. The antagonism
of CXCR1 and CXCR2 by compound 7 in hPMNs was found to
be selective; however, as concentrations up to 5 μM failed to
inhibit Ca2+ flux induced by optimal concentrations of
the chemokines C5a, PAF, and fMLF (Table 5).Although the inhibition of CXCL8-induced Ca2+ flux in
hPMNs indicated dual antagonism of CXCR1 and CXCR2, the coexpression
of these receptors in hPMNs precluded evaluating the relative inhibitory
potency at each receptor. Additional experiments were therefore performed
to separately evaluate the potency of compound 7 at CXCR1
and CXCR2. For these determinations, CXCL8-induced Ca2+ flux was measured in recombinant HEK293 cells that stably expressed
CXCR1 (HEK293-CXCR1) or CXCR2 (HEK293-CXCR2) in the presence of vehicle
and different concentrations of compound 7 (Table 5). For CXCL8-induced Ca2+ flux, the data
showed that compound 7 exhibited a 4-fold selectivity
in inhibition for CXCR2 (IC50 = 210 nM) over CXCR1 (IC50 = 880 nM) in these recombinant systems. As a positive control,
the CXCR2-selective antagonist SB225002 inhibited CXCL8-induced Ca2+ flux in the HEK293-CXCR2 cells with an IC50 of
40 nM, a value consistent with its previously reported IC50 value of 30 nM for CXCL1-induced Ca2+ flux in hPMNs.[24] The observed discrepancy in the potency of 7 when evaluated in isolated hPMNs versus HEK cells is not
fully understood but may be attributed to intracellular differences
between the native (hPMN) and artificially engineered (HEK) systems.
Effect on Inflammation in Vivo
To confirm the in vitro
effects on the neutrophil function in vivo, compound 7 was evaluated in the murine air-pouch model of inflammation (Figure 6). An air-pouch was induced on the backs of male
CD1 Swiss mice as described.[47] Cohorts
of five animals each were given vehicle (negative and positive control
cohorts) or compound 7 dissolved in vehicle (0.02 mg/kg
and 0.20 mg/kg test cohorts) by intravenous injection. Three hours
afterward, the pouches in the negative control cohort were injected
with sterile phosphate-buffered saline. Inflammation was induced in
the pouches of the remaining three cohorts by injection with 2% carrageenan,
which causes an inflammatory infiltrate consisting predominantly of
neutrophils. The total cell count in the inflammatory infiltrate from
each pouch was quantitated. At a dose of 0.2 mg/kg of compound 7, there was a significant reduction in cell count in the
pouches of treated animals compared to the positive control cohort
(**p < 0.01, Student’s t-test).
Figure 6
An air-pouch was formed on the backs of 10–15 week old,
male CD1 Swiss mice in four cohorts (n = 5 animals
per cohort). Compound dissolved in vehicle (0.02 mg/kg and 0.20 mg/kg
cohorts) or vehicle alone (positive and negative cohorts) was administered
intravenously. After 3 h, each air pouch was injected with 1 mL of
PBS (negative cohort) or 2% carrageenan in PBS (positive, 0.02 mg/kg
and 0.20 mg/kg cohorts). After 4 h, the pouch fluid was collected
and combined with an additional 2 mL PBS wash of the pouch. The cells
in the combined fluid were stained with trypan blue and manually counted
on a hemocytometer. Data show the mean ± SE of the absolute pouch
cell count per cohort. Student’s t-test: **p < 0.01 vs positive cohort.
An air-pouch was formed on the backs of 10–15 week old,
male CD1 Swiss mice in four cohorts (n = 5 animals
per cohort). Compound dissolved in vehicle (0.02 mg/kg and 0.20 mg/kg
cohorts) or vehicle alone (positive and negative cohorts) was administered
intravenously. After 3 h, each air pouch was injected with 1 mL of
PBS (negative cohort) or 2% carrageenan in PBS (positive, 0.02 mg/kg
and 0.20 mg/kg cohorts). After 4 h, the pouch fluid was collected
and combined with an additional 2 mL PBS wash of the pouch. The cells
in the combined fluid were stained with trypan blue and manually counted
on a hemocytometer. Data show the mean ± SE of the absolute pouch
cell count per cohort. Student’s t-test: **p < 0.01 vs positive cohort.
Conclusion
The results reported here describe SAR studies
that examined the
effect of a novel series of S-substituted 6-mercapto-N-phenyl-nicotinamides on CXCL1-stimulated Ca2+ flux in
whole hPMNs. The SAR data established that an ortho-modified S-benzyl substituent played a critical role in determining
activity, with an ortho-boronyl group being optimal. The fluorine
in the 4-fluorophenyl-carboxamido- moiety was also an important requirement
for optimal inhibitory activity. Among the derivatives exhibiting
the most potent antagonism of CXCL1 described here, 7 was selected for further evaluation of its in vitro pharmacology
and in vivo biologic activity.Compound 7 was found
to be a dual CXCR2/1 antagonist
with a modest preference for CXCR2. It inhibited CXCL1- and CXCL8-induced
Ca2+ flux (IC50 = 38 and 36 nM, respectively)
in hPMNs. In response to CXCL8 stimulation, compound 7 directly antagonized [35S]GTPγS binding (IC50 = 60 nM) and ERK1/2 phosphorylation in HEK293 cells that
stably expressed CXCR2. In an in vivo murine model of inflammation
characterized by an inflammatory infiltrate predominated by neutrophils,
compound 7 significantly reduced total cell count at
an intravenous dose of 0.2 mg/kg.Early work by others to develop
small-molecule chemokine inhibitors
focused on evolving compounds that were highly specific for either
CXCR1[22,23] or CXCR2.[24,25] These efforts
were later followed by compounds with dual-activity at CXCR1 and CXCR2[26,48] in recognition that these homologous receptors mediate inflammation
through unique and overlapping pathways. For example, whereas both
CXCR1 and CXCR2 mediate CXCL8-induced chemotaxis,[23,49] myeloperoxidase release from hPMNs occurs mainly through CXCR1 activation.[50,51] CXCL8-mediated neutrophil chemotaxis is most effectively inhibited
by dual CXCR1 and CXCR2 blockade.[52] Compared
to specific antagonists, dual CXCR2/1 inhibition by compound 7 may therefore offer a more complete therapeutic strategy
in a number of inflammatory diseases where the CXCR2 pathway is involved
specifically or in conjunction with CXCR1 signaling.G-protein
coupled receptors (GPCRs) are thought to exist in multiple
conformational states, and activation by an agonist mediates signaling
to the cell interior through poorly understood conformational changes
in the receptor. It has been proposed that allosteric sites in the
transmembrane domain of GPCRs may represent high-value targets for
noncompetitive inhibitors that block agonist-induced G-protein activation.[53] Indeed, studies have found that the antagonist
reparixin (1),[22,23] and the CXCR2-specific
antagonists SB265610, SB332235, and SCH527123 (3), act
as allosteric inhibitors.[43,49,54−56] These inhibitors achieve receptor selectivity by
exploiting amino acid differences between these homologous receptors,
with reparixin (1) differentially binding the extracellular
half of the seven-transmembrane core[23] and
SB332235 differentially binding an intracellular pocket made up of
several transmembrane helices.[55] We previously
demonstrated that the nicotinamide glycolate methyl ester 5 inhibited CXCL1-induced effects in hPMNs through an intracellular
mechanism.[35] In the studies herein, potent
antagonism of CXCL8-stimulated [35S]GTPγS binding
by compound 7 localized inhibition to a proximal signaling
event involving the CXCR2 receptor and its G-protein. Further data
demonstrated compound 7 was CXCR2/1 specific, with no
inhibition of fMLF-, C5a-, or PAF-induced Ca2+ flux through
their corresponding GPCRs. We therefore speculate that compound 7 acts at an intracellular pocket, involving at least the
CXCR2/1 receptor to lock the receptor in a conformation unable to
activate downstream signaling. Studies to explore this postulated
mechanism are underway in our group.Consistent with being a
noncompetitive inhibitor, compound 7 potently inhibited
functional CXCR2 signaling by CXCL1 and
CXCL8 but failed to compete with the binding of [125I]-CXCL8
to CXCR2 at concentrations ∼100-fold larger than its IC50 for inhibiting CXCR2 ligand-triggered Ca2+ flux.
Others have earlier recognized the possibility of identifying allosteric
small-molecule chemokine inhibitors that disrupt receptor function
without displacing the chemokine from its orthosteric binding site.[57] The SAR studies herein demonstrate that our
whole-cell functional assay for CXCL1-induced Ca2+ flux
in hPMNs beneficially allowed for the selection of a novel class of
potent noncompetitive inhibitors that would have otherwise been missed
by a competitive displacement screen. Our findings mirror those of
the noncompetitive allosteric inhibitor reparixin, which likewise
inhibited CXCR1 responses with no effect on CXCL8 binding to CXCR1.[22,23]Compound 7 is notable for being the first reported
boronic acid chemokine antagonist. Due to the unique chemical similarities
and differences betweenboron and carbon, boronic acid based inhibitors
have been the subject of considerable recent interest as a new source
of small-molecule therapeutics.[38] The clinical
utility of the boronyl moiety has been validated with the U.S. and
European approval of bortezomib (Velcade), and other boronic acids
have been previously disclosed as inhibitors of serine proteases,
proteasomes, arginase, nitric oxide synthase, and transpeptidases.[39] The effectiveness of boronic acids as inhibitors
of proteosomes[58,59] and serine proteases[60−63] has been ascribed to the ability of the boronyl group to form a
reversible tetrahedral adduct with an active-site nucleophile that
has most often been serine but has also included histidine.[60] On the basis of crystallographic and NMR studies
of boronic acid inhibitors in complex with their target enzymes, this
adduct closely mimics the putative tetrahedral intermediate or transition-state
formed between the enzyme and its substrates. Although CXCR1 and CXCR2
are not thought to possess intracellular enzymatic activity, we speculate
that the marked increase in potency observed upon incorporation of
the boronyl group at the 2-position of the S-benzyl
scaffold in compound 7 may be due to an analogous reversible
boronyl-mediated adduct with a receptor-based nucleophile in the putative
intracellular binding pocket. The potential benefit of such a postulated
adduct in prolonging pharmacodynamic effects in subjects remains to
be investigated. However, like other CXCR2/1 inhibitors with a low
off-rate (e.g., SCH527123),[49] the benefit
will likely be limited by the kinetics of new receptor biosynthesis
and the 1–2 day tissue lifespan of the neutrophil itself.Current efforts are underway in our laboratory to increase the
oral bioavailability of compound 7 and to further investigate
the interaction of this class of compounds with CXCR1 and CXCR2 receptors.
Compound 7 (SX-517) represents a novel boronic acid containing
pharmacophore for the antagonism of CXCR1/2 chemokine receptors and
may prove useful in the treatment of inflammatory diseases with a
significant neutrophil component.
Experimental
Methods
Pharmacology and Biology
Materials and Reagents
Chemicals and carrageenan were
obtained from Sigma-Aldrich Chemical Co. (St. Louis, MO). Chemokines
were from PeproTech (Rocky Hill, NJ) or from R&D Systems (Minneapolis,
MN). Control CXCR2 antagonists (SB265610 and SB225002) were from Tocris
Biosciences (Ellisville, MO). The HEK293 cell line was from ATCC (Manassas,
VA). Erk1/2 and phospho-Erk (p-Erk) antibodies were from Santa Cruz
Biotechnology (Santa Cruz, CA). Cellulose nitrate membrane filters
were from Whatman Inc. (Piscataway, NJ). [35S]GTPγS
(1250 Ci/mmol), 125I-CXCL8 (2200 Ci/mmol), and Unifilter
GF/C 96-well filter plates were from PerkinElmer Life and Analytical
Sciences (Waltham, MA).
HEK293 Cell Stably Expressing Human CXCR1
and CXCR2
HEK293 cells were cultured in DMEM (Dulbecco’s
modified Eagle’s
medium) supplemented with 50 units/mL penicillin, 50 μg/mL streptomycin,
3 mM glutamine, 10% heat-inactivated fetal bovine serum (Atlanta Biologicals,
Lawrenceville, GA) at 37 °C, 5% CO2. Transfection
with a humanCXCR1[3] or CXCR2 plasmid[44] was performed with Lipofectamine (Invitrogen
Life Technology) based on the manufacturer’s protocol. For
selection of stable polyclonal cell lines, 800 μg/mL G418 (Sigma,
St Louis, MO) were added 24 h after transfection and cells were maintained
in DMEM medium containing 800 μg/mL G418 through subculture
procedures until a pooled, stable cell line was established. Surface
expression of CXCR1 and CXCR2 was confirmed using R-phycoerythrin (PE)-conjugated antihumanCXCR1 or CXCR2mouse monoclonal
antibody (BD Pharmingen, San Diego, CA) and fluorescence-activated
cell sorting as described below.
Isolation of Human Neutrophils
Blood was collected
from healthy donors in accordance with a protocol approved by the
Institutional Review Board at Montana State University. Human polymorphonuclear
leukocytes (hPn class="CellLine">MNs) were purified from the blood using dextran sedimentation,
followed by Histopaque 1077 gradient separation and hypotonic lysis
of red blood cells, as described previously.[64] hPMN preparations were routinely >95% pure, as determined by
light
microscopy, and >98% viable, as determined by trypan blue exclusion.
Isolated hPMNs were washed twice and resuspended in RPMI containing
10% fetal bovine serum (FBS).
Animals
Male CD1mice were obtained from Charles River
Laboratories (Wilmington, MA). Animals were housed and acclimatized
for 1 week under controlled temperature (20 ± 2 °C), humidity
(55 ± 10%), and lighting (7 a.m. to 7 p.m.). Standard sterilized
food and water were supplied ad libitum during acclimatization
and experiments. All procedures and protocols were approved by the
Institutional Animal Care and Use Committee and were carried out in
accordance with NIH guidelines for the handling and use of laboratory
animals.
Calcium Flux Assay
hPMNs (or cells
expressing either
CXCR1 or CXCR2) were suspended in HBSS– (Hank’s
balanced salt solution without Ca2+ and Mg2+) containing 10 mM HEPES and FLIPR Calcium 3 dye (3.1 × 107 cells in total volume 1.7 mL). Cells were aliquoted (200
μL of the cell suspension per tube, 8 tubes total), and 2 μL
of the designated compound (with appropriate dilutions) were added
to each of 6 tubes. As controls, 2 μL of DMSO (1% final concentration)
were added to two other tubes. Cells were incubated at 37 °C
for 30 min. After dye loading, tubes were centrifuged at 6000 rpm
for 1 min, supernatant was removed, and the cell pellet was resuspended
in 200 μL of HBSS+ (with Ca2+ and Mg2+), containing 10 mM HEPES. The test compound or DMSO (control)
were added again at the same concentrations that were used during
cell loading. The cell suspension was aliquoted into a 96-well Reading
Plate (Corning) in a volume of 90 μL (105 cells/well). The Compound
Plate contained agonist in HBSS–) or HBSS– (control). After 15 s of reading the basal level of fluorescence
by FlexStation II, 10 μL of agonist or HBSS– were automatically transferred from the compound plate into the
reading plate. The agonists used and their final concentrations were
25 nM CXCL1, 1 nM CXCL8, 10 nM N-formyl-l-methionyl-l-leucyl-l-phenylalanine (fMLF), and
50 nM C5a. Changes in fluorescence were monitored (λex = 485 nm, λem = 525 nm) every 5 s for 240 to 500
s at room temperature. The maximum change in fluorescence, expressed
in arbitrary units over baseline (max–min), was used to determine
the agonist response. The effect of each compound on the agonist response
was normalized and expressed as a percent of the DMSO control, which
was designated as “100% response.” Curve fitting and
calculation of the compound inhibitory concentration that reduced
the level of the agonist response by 50% (IC50) was determined
by nonlinear regression analysis of the dose–response curves
generated using Prism 4 (GraphPad Software, Inc., San Diego, CA).
[35S]GTPγS Assay
[35S]GTPγS
assays were performed as previously described[65] with the following modifications: HEK293 cells stably expressing
hCXCR2, pretreated with different concentrations of compound, were
lysed in buffer containing 5 mM Tris-HCl, pH 7.5, 5 mM EDTA, and 5
mM EGTA, and the cell lysate was centrifuged at 30000g for 10 min. Protein concentration in membrane preparations was determined
using the BioRad Protein Determination assay 18 from Bio-Rad (Hercules,
CA). Membranes containing 50 μg of protein were incubated in
50 mM Tris-HCl, pH 7.5, 5 mM MgCl2, 1 mM EGTA, 100 mM NaCl,
50 μM GDP, 8 nM [35S]GTPγS, 10 nM CXCL8 in
a total volume of 0.1 mL at 30 °C for 1 h. The reaction was terminated
by dilution into phosphate-buffered saline and rapid filtration through
Unifilter GF/C 96-well filter plates pretreated with 0.3% polyethylenimine
and washed three times with ice-cold wash buffer (50 mM Na2HPO4 and 50 mM KH2PO4, pH 7.4).
Bound radioactivity was determined using a MicroBeta counter (PerkinElmer
Life and Analytical Sciences). Basal binding was assessed in the absence
of CXCL8, and nonspecific binding was determined in the presence of
10 μM GTPγS. The percentage of CXCL8-stimulated [35S]GTPγS binding was calculated as [cpmCXCL8 – cpmnonspecific]/[cpmbasal –
cpmnonspecific]. Curve fitting and calculation of the compound
inhibitory concentration that reduced the percentage of CXCL8-stimulated
[35S]GTPγS binding by 50% (IC50) was determined
by nonlinear regression analysis of the dose–response curves
generated using Prism 4 (GraphPad Software, Inc., San Diego, CA).
Competition 125I-CXCL8 Binding Assay
This
was performed according to White et al. using HEK293-hCXCR2 membranes.[24] Briefly, assays were performed in 96-well microtiter
plates where the reaction mixture contained 1.0 μg/mL membrane
protein in 20 mM Bis-trispropane, pH 8.0, with 1.2 mM MgSO4, 0.1 mM EDTA, 25 mM NaCl, and 0.03% CHAPS and compound (100 μM
stock in DMSO) added at the indicated concentrations, the final DMSO
concentration was <0.5% under standard binding conditions. Binding
was initiated with 25 pM [125I]-CXCL8. Nonspecific binding
was determined with 30 nM CXCL8. After 1 h incubation at room temperature,
membranes were harvested by rapid filtration. The filter was dried
and counted with a liquid scintillation counter. Specific CXCL8 binding
to the receptors was defined as the difference between the total binding
and the nonspecific binding determined in the presence of an excess
of unlabeled CXCL8. The results were expressed as a percent of control
specific binding: (specific binding with compound)/(control specific
binding) × 100. IC50 for unlabeled CXCL8 was determined
by nonlinear regression analysis of the concentration–response
curve generated with mean replicate values fitted to the Hill equation.
CXCR2 Cell Surface Expression by Flow Cytometry
HEK293
cells stably expressing CXCR2 were pretreated with 1% DMSO (vehicle)
or compound (10 μM) for 60 min. HEK293 cells not expressing
CXCR2 served as a negative isotype control. All cells were then incubated
with a 1:100 dilution of R-phycoerythrin (PE)-conjugated
antihumanCXCR2mouse monoclonal antibody (BD Pharmingen, San Diego,
CA) at 4 °C for 60 min. Cells were washed, fixed in 2% formaldehyde
in phosphate-buffered saline, and quantitated by fluorescence-activated
cell sorting (FACS) using a FACScan flow cytometer equipped with CellQuest
software (BD Biosciences, Mountain View, CA).
ERK1/2 Phosphorylation
HEK293 cells expressing CXCR2
were harvested and plated in equal number to 60 mm plates (5 ×
106 cells/plate). The cells were then incubated with PBS
containing 1% DMSO (vehicle) or 10 μM compound for 60 min at
37 °C, followed by the addition of 100 ng/mL CXCL8. All cells
in a plate were then lysed at 0, 5, 15, and 30 min after agonist-treatment
by adding lysis buffer: 0.1% SDS, 0.5% sodium deoxycholate, 1% Triton
X-100, 10 mM Tris (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1 mM phenylmethylsulfonyl
fluoride, 10 μg/mL aprotinin, and 10 μg/mL leupeptin.
Lysates containing equal amounts of protein (∼50 μg)
were resolved by 10% SDS–polyacrylamide gel electrophoresis,
transferred to a nitrocellulose membrane, and probed with antibody
against phospho-ERK1/2 or ERK1/2. Visualization was carried out with
a horseradishperoxidase-conjugated secondary antibody.
Murine Air-Pouch
Model of Inflammation
An air-pouch
was induced on the backs of 10–15 week old, male CD1 Swiss
mice by subcutaneous injection (2 mL) of air as described.[47] An additional subcutaneous injection of air
(1.5 mL) was given to reinflate the pouch the following day. The compound
was solubilized in vehicle consisting of a 20:20:80 mixture of PEG400:DMF:saline.
Compound dissolved in vehicle, or vehicle alone was administered via
iv tail vein injection at different doses according to body weight.
After 3 h, either sterile phosphate-buffered saline (PBS, 1 mL) or
2% carrageenan in sterile PBS (1 mL) was injected into the air pouch.
After 4 h, the mice were sedated with ketamine and sacrificed by cervical
dislocation. The pouch fluid was then collected by syringe and combined
with an additional 2 mL PBS wash of the pouch. The cells in the combined
fluid were stained with trypan blue and manually counted on a hemocytometer
under 20× magnification.
Chemistry
Materials
and Reagents
General chemicals and reagents
for synthesis were purchased from Sigma-Aldrich (Milwaukee, WI), and
solvents were purchased from VWR International (West Chester, PA)
and used without further purification. Commercial synthetic precursors
and intermediates were from Acros Organics (Pittsburgh, PA), Sigma-Adrich
Chemical (Milwaukee, WI), Avocado Research (Lancashire, U.K.), Bionet
(Cornwall, U.K.), Boron Molecular (Research Triangle Park, NC), Combi-Blocks
(San Diego, CA), Eastman Organic Chemicals, Eastman Kodak Company
(Rochester, NY), Fisher Scientific Co. (Pittsburgh, PA), Frontier
Scientific (Logan, UT), ICN Biomedicals, Inc. (Costa Mesa, CA), Lancaster
Synthesis (Windham, NH), Maybridge Chemical Co. (Cornwall, U.K.),
Pierce Chemical Co. (Rockford, IL), Riedel de Haen (Hannover, Germany),
Santa Cruz Biotechnology (Dallas, TX), Spectrum Quality Product, Inc.
(New Brunswick, NJ), TCI America (Portland, OR), and Wako Chemicals
USA, Inc. (Richmond, VA). Solid phase scavenger resins were from NovaBiochem
(La Jolla, CA).
General Analytical Procedures
Synthetic
reaction progression
was monitored by thin-layer chromatography (TLC) using precoated aluminum-backed
plates with silica gel with fluorescent indicator (precoated F254
Macherey-Nagel plates, EMD Chemicals); the spots were examined with
UV light. Chromatographic purification was performed with 230–400
mesh (32–63 μm) flash silica gel (Dynamic Adsorbents,
Norcross, GA) or by preparative high-performance liquid chromatography
(HPLC) using a Waters Delta Prep 4000 HPLC fitted with a Phenomenex
Gemini 250 × 21 mm, 10 μm, C18 column and monitored
at 254 nm. Retention time (RT) is reported in minutes (min), and purity
as measured by UV absorbance is reported as a percentage of all peak
areas. HPLC analyses were performed using the following gradients
and systems:
Gradient A
Water:n class="Chemical">acetonitrile:formic
acid (95:5:0.1)
to water:acetonitrile:formic acid (5:95:0.1) at 35 °C over 12
min, on a Shimadzu HPLC system, with a Phenomenex Gemini 50 ×
2 mm, 5 μm, C18 column, monitored at 254 nm.
Gradient B
Water:n class="Chemical">acetonitrile:formic
acid (95:5:0.1)
to water:acetonitrile:formic acid (5:95:0.1) at 30 °C over 30
min, on an Agilent 1100 Series HPLC system, with a Phenomenex Gemini
50 × 2 mm, 5 μm, C18 column, monitored at 254
nm.
Gradient C
Water:n class="Chemical">acetonitrile:trifluoroacetic
acid
(95:5:0.1) to water:acetonitrile:trifluoroacetic acid (5:95:0.1) at
35 °C over 0.5 min, held at 5:95:0.1 for 6.5 min, on an Agilent
1100 Series HPLC system, with Phenomenex Gemini 50 × 2 mm, 5
μm, C18 column, monitored at 254 nm.
Gradient D
Water:n class="Chemical">acetonitrile:formic
acid (95:5:0.1)
to water:acetonitrile:formic acid (5:95:0.1) at 35 °C over 4
min, held at 5:95:0.1 for 5 min, on a Shimadzu HPLC system, with Phenomenex
Gemini 50 × 2 mm, 5 μm, C18 column, monitored
at 254 nm.
Electrospray ionization mass spectrometric analysis
(ESI-MS) was performed using a Micromass Quattro II mass spectrometer
with MassLynx 4.0. 1H NMR spectra were obtained on a Bruker
AVance (300 or 500 MHz, 1H) and are reported as parts per
million downfield from tetramethylsilane with number of protons, multiplicities,
and coupling constants in Hertz indicated parenthetically. Elemental
analyses were performed by Atlantic Microlab (Norcross, GA).
General Procedures for the Synthesis of S-Substituted N-(4-Fluoro-phenyl)-6-mercapto-nicotinamides (Scheme 1)
Accomplished by one of the following
three methods. If required, the crude was purified by either flash
n class="Chemical">silica gel chromatography or preparative HPLC. Final product was characterized
by HPLC, ESI-MS, and 1H NMR where indicated.
Method A
A solution of the bromomethyl
derivative (2.5
equiv) in anhydrous DMF (5 mL/mmol) was added to thionicotinamide 64 (1 equiv) and N-methylmorpholino-substituted
polystyrene resin (5 equiv) and heated at 60 °C for 2 h in a
screw cap glass vial. Sulfhydryl-bearing scavenger resin (5 equiv)
was then added to the reaction mixture and heated at 60 °C for
a further 4 h. After cooling, the organic reaction solution was filtered,
and then diluted into water (100 mL/mmol) to precipitate the product.
The resulting suspension was then centrifuged at 5000 rpm for 15 min,
the aqueous supernatant was decanted, and the product dried in a vacuum
oven overnight at 50 °C. Compounds prepared by this method were
used without further purification.
Method B
Thionicotinamide 64 (1 equiv)
and the corresponding bromomethyl or chloromethyl derivative (1 equiv)
was dissolved in anhydrous DMF (2 mL/mmol). To the solution, a tertiary
amine base (diisopropylethylamine, triethylamine, or N-methylmorpholine, 1 equiv) was added. The reaction was allowed to
proceed at room temperature and monitored by either TLC or LC–MS
until complete (1–18 h). The crude product was then precipitated
out of solution by the addition of water (5–50 mL/mmol).
Method C
Thionicotinamide 64 (1 equiv)
and the corresponding bromomethyl derivative (1 equiv) were suspended
in EtOH (5 mL/mmol). To the suspension, 1 N NaOH (1 equiv) was added,
and the reaction mixture was brought to gentle reflux and monitored
by either TLC or LC–MS until complete (0.5–2 h). The
crude product was then precipitated out of solution by the addition
of water (5–20 mL/mmol).
Prepared via method C using
thionicotinamide 64 (1.27 g, 5.10 mmol) and 2-bromomethyl-phenylboronic
acid (1.09 g, 5.10 mmol) suspended in EtOH (50 mL). To the suspension,
1 N NaOH (5.1 mL, 5.10 mmol) was added and the reaction mixture heated
to gentle reflux for 2 h. Then water (50 mL) was added to the reaction
mixture while still hot. Upon cooling, a white precipitate formed
and this was filtered, washed with 50% aqueous EtOH, and then water
and dried in an oven to yield 1.53 g (78%) of 7 (SX-517)
as an off-white solid. ESI-MS m/z = 383.1 [M + H]+. 1H NMR (500 MHz, DMSO-d6): δ 10.39 (s, 1H), 8.99 (s, 1H), 8.23
(s, 2H), 8.13–8.11 (m, 1H), 7.80–7.77 (m, 2H), 7.55
(d, J = 7.0 Hz, 1H), 7.45 (d, J = 8.0 Hz, 2H), 7.30
(t, J = 7.0 Hz, 1H), 7.22 (t, J =
8.8 Hz, 3H), 4.69 (s, 2H). 13C NMR (100.6 MHz, DMSO-d6): d 163.6, 162.7, 159.6, 157.2, 148.6, 141.5,
136.2, 135.5, 135.3, 134.0, 129.3, 129.0, 126.2, 122.2, 122.1, 120.7,
115.4, 115.1. Anal. Calcd for C19H16BFN2O3S: C, 59.71%; H, 4.22%; N, 7.33%; S, 8.39%. Found:
C, 59.54%; H, 4.38%; N, 7.48%; S, 8.49%. HPLC (gradient B): RT = 18.45
min, purity 96.1%.
The cyano
intermediate was prepared via method B using thionicotinamide 64 and n class="Chemical">chloroacetonitrile. This intermediate was used without
further purification. The cyano-intermediate (190 mg, 0.66 mmol),
dibutyltin oxide (33 mg, 0.14 mmol) and trimethylsilyl azide (174
μL, 1.32 mmol) were suspended in toluene (50 mL) and refluxed
for 24 h. The mixture was allowed to cool to room temperature, and
the resulting precipitate was filtered and washed with toluene to
yield 160 mg (73%) of 8 as a light yellow solid. TLC
(1% AcOH/ethyl acetate): R = 0.32. ESI-MS m/z = 331.3
[M + H]+. 1H NMR (500 MHz, DMSO-d6): δ 10.41 (s, 1H), 8.97 (s, 1H), 8.18 (d, J = 8.0 Hz, 1H), 7.77 (t, J = 6.0 Hz, 6.6
Hz, 2H), 7.58 (d, J = 8.2 Hz, 1H), 7.21 (t, J = 8.2 Hz, 2 H), 4.81 (s, 2H). HPLC (gradient B): RT =
14.52 min, purity 96.4%.
Prepared via method A using thionicotinamide 64 and n class="Chemical">2-[2-(bromomethyl)phenyl]thiophene and screened without
purification. ESI-MS: m/z = 421.1
[M + H]+. HPLC (gradient C): RT = 7.31 min, purity 73.7%.
The boronic acid 6 (200 mg)
was dissolved in 18 mL of THF/acetone/0.1 N NaOH (1:1:1) and cooled
to −3 °C. OXONE (1.0 eq, 311 mg) dissolved in 0.4 mM EDTA
(2.1 mL) was added, according to the procedure of Webb and Levy for
the hydroxylation of aryl boronic esters or boronic acids.[36] The reaction was then stirred for 1 h and quenched
with sodium bisulfite (3 mL, 5 M aqueous), while being stirred for
10 min. The solvent was removed by reduced pressure rotary evaporation,
leaving an aqueous suspension. The suspension was extracted 3×
with ethyl acetate, and the combined ethyl acetate layers were washed
with 0.01 N HCl, water, 0.01 N NaOH, and brine. They were dried over
MgSO4, filtered, and dried under vacuum. The residue was
dissolved in THF/MeOH and adhered to 2 g of silica gel. The adhered
silica was loaded onto 20 g of silica gel, eluting the product with
3:1 hexanes:ethyl acetate, which was dried under vacuum to provide
67 mg of compound 44 (36% yield). ESI-MS: m/z = 355.4 [M + H]+. 1H NMR
(500 MHz, DMSO-d6): δ 9.75 (s, 1H),
9.01 (d, J = 2.1 Hz, 1H), 8.14 (dd, J = 8.7 Hz, 2.6 Hz, 1H), 7.80–7.77 (m, 3H), 7.48 (d, J = 8.3 Hz, 1H), 7.33 (dd, J = 7.4 Hz,
1.3 Hz, 1H), 7.23–7.20 (m, 2H), 7.11–7.07 (m, 1H), 6.85
(dd, J = 8.1 Hz, 0.8 Hz, 1H), 6.75–6.72 (m,
1H), 4.42 (s, 2H). Anal. Calcd for C19H15FN2O2S: C, 64.39%; H, 4.27%; N, 7.90%. Found: C, 64.50%;
H, 4.47%; N, 7.72%. HRMS: calcd for C19H16FN2O2S [M + 1]+ 355.0911, found 355.0901;
calcd for C19H15FN2NaO2S [M + Na]+ 377.0730, found 377.0718. HPLC (gradient B):
RT = 20.18 min, purity 97.4%.
The cyanobenzyl intermediate 66 (363 mg, 1 mmol) was
suspended in anhydrous toluene (25 mL). Dibutyltin oxide (25 mg, 0.1
mmol) was added, followed by trimethylsilyl azide (131 μL, 1
mmol). The mixture was then heated to reflux for 18 h. The dark yellow
brown solution was allowed to cool to room temperature, at which time
a brown precipitate was seen to form. The precipitate was filtered
and washed with toluene to yield 190 mg (47%) of compound 52. ESI-MS: m/z = 407.2 [M + H]+. 1H NMR (500 MHz, DMSO-d6): δ 10.39 (s, 1H), 8.93 (s, 1H), 8.11 (dd, J = 8.6 Hz, 1.6 Hz, 1H), 7.80–7.75 (m, 4H), 7.55–7.49
(m, 2H), 7.42 (d, J = 8.3 Hz, 1H), 7.22 (t, J = 8.6 Hz, 2H), 4.90 (s, 2H). HRMS: calcd for C20H16FN6OS [M + 1]+ 407.1085, found 407.1078;
calcd for C20H15FN6NaOS [M + Na]+ 429.0904, found 429.0896. Purity determined by HPLC (gradient
B): RT = 18.81 min, purity 94.6%.
According to method C, N-(4-fluorophenyl)-N-pyridin-2-yl-methyl-6-mercapto-nicotinamide
intermediate 73a (2.9 g, 8.7 mmol) and 2-bromomethyl-phenylboronic
acid (2.9 g, 8.7 mmol) were suspended in EtOH (100 mL). One N NaOH
(8.7 mL, 8.7 mmol) was added and the suspension brought to reflux.
After 2 h, the mixture was concentrated in vacuo and
partioned between ethyl acetate and water. The ethyl acetate was extracted
3× with water and evaporated to yield 2.9 g (70%) of compound 60 as a white solid. ESI-MS: m/z = 473.9 [M + H]+. 1H NMR (500 MHz, DMSO-d6): δ 8.52 (d, J = 4.4
Hz, 1H), 8.37 (s, 1H), 8.17 (s, 2H), 7.77 (t, J =
8.0 Hz, 7.4 Hz, 1H), 7.51–7.47 (m, 3H), 7.32–7.23 (m,
5H), 7.21–7.16 (m, 2H), 7.11 (t, J = 8.6 Hz,
2H), 5.15 (s, 2H), 4.56 (s, 2H). HPLC (gradient B): RT = 16.08 min,
purity 95.9%.
Thiol alkylation was achieved according
to method C by suspending the thionicotinamide intermediate 73c (90 mg, 1 equiv) and 2-bromomethylphenyl boronic acid
(50 mg, 1.1 equiv) in EtOH (1 mL). Then 1 N NaOH (0.2 mL, 1 equiv)
was added and the solution gently refluxed for 2 h. Alkylated product 74a was extracted with ethyl acetate (3 × 5 mL), washed
with water (3 × 5 mL), brine (3 × 5 mL), and dried over
MgSO4. The organic layer was filtered to remove MgSO4, and then removed by rotary evaporation and dried in vacuo
to afford 90 mg of the Boc-protected piperidine boronic acid. The
Boc group was then removed by adding 4 M HCl in dioxane (1 mL) to
the Boc-protected piperidine boronic acid (20 mg). The resulting piperidine
compound was then purified using preparative HPLC (80:20:0.1 water:acetonitrile:formic
acid to 70:30:0.1 water:acetonitrile:formic acid over 30 min) to yield
5 mg (29%) of 62 as a white solid. ESI-MS: m/z = 480.1 [M + H]+. 1H NMR
(500 MHz, D2O): δ 8.35 (s, 1H), 8.08 (s, 1H), 7.41–7.37
(m, 2H), 7.28–7.15 (m, 4H), 7.05 (d, J = 7.6
Hz, 1H), 6.96–6.92 (m, 2H), 4.30 (s, 3H), 3.87 (dd, J = 14.8 Hz, 3.7 Hz, 1H), 3.38 (d, J =
10.8 Hz, 1H), 3.28 (br s, 1H), 2.88–2.84 (m, 1H), 1.78 (d, J = 7.7 Hz, 3H), 1.61–1.56 (m, 1H), 1.48–1.43
(m, 1H), 1.39–1.36 (m, 1H). HPLC (gradient B): RT = 19.29 min,
purity 90.1%.
The thionicotinamide 73d (110 mg, 0.3 mmol) and n class="Chemical">2-bromomethyl-phenylboronic acid were coupled
using method B to yield the tert-butyl esterthionicotinamide
intermediate 74b (138 mg, 93%). TLC (ethyl acetate): R = 0.6; ESI-MS: m/z = 496.9 [M + H]+. The intermediate
was dissolved in 90% aq TFA and incubated at room temperature for
2 h. The TFA was removed by rotary evaporation to and the crude material
purified by preparative HPLC (65:35:01 water:acetonitrile:formic acid
isocratic) to yield 74 mg (61%) of compound 63 as a white
hygroscopic solid. ESI-MS: m/z =
440.9 [M + H]+. 1H NMR (500 MHz, DMSO-d6): δ 12.95 (br s, 1H), 8.37 (s, 1H),
8.30 (s, 1H), 7.53 (d, J = 5.9 Hz, 1H), 7.41–6.92
(br m, 10H), 4.57 (s, 2H), 4.46 (s, 2H). HPLC (gradient B): RT = 16.75
min, purity 96.6%.
N-(4-Fluoro-phenyl)-6-mercaptonicotinamide
(64)
To 150 mL DMF was added and stirred 6-mercapto-nicotinic
acid (50 mmol, 7.76 g). To the stirred solution, n class="Chemical">4-fluoroaniline (50
mmol, 4.8 mL) was added followed by the addition of 2-ethoxy-1-ethoxycarbonyl-1,2-dihydroquinoline
(EEDQ, 50 mmol, 12.36 g). The dark brown mixture was stirred for 12
h. The mixture was then diluted with water (500 mL), and the precipitate
collected by filtration and washed repeatedly with water. The off-white
solid was dried in an oven (50 °C) for 72 h to afford 6.42 g
(52%) of the thionicotinamide product. ESI-MS: m/z = 248.9 [M + H]+. 1H NMR (500 MHz,
DMSO-d6): δ 13.87 (s, 1H), 10.23
(s, 1H), 8.32 (s, 1H), 7.89 (dd, J = 8.8 Hz, 2.2
Hz, 1H), 7.74–7.71 (m, 2H), 7.38 (d, J = 9.3
Hz, 1H), 7.19 (t, J = 8.7 Hz, 2H). TLC (ethyl acetate/hexanes/MeOH
1:1:0.1): R = 0.7.
Thionicotinamide 64 (248 mg. 1.00 mmol)
dissolved in anhydrous n class="Chemical">DMF (3 mL) with chloroacetonitrile (63 μL,
1.0 mmol) was then added, followed by triethylamine (140 μL,
1.00 mmol). The reaction was heated to 100 °C for 30 min. After
cooling to room temperature, the solution was diluted into water (60
mL), and the resulting precipitate was filtered and washed with water
to yield the cyano-intermediate 65 as a white solid (259
mg, 90%). ESI-MS: m/z = 288.2 [M
+ H]+. 1H NMR (300 MHz, DMSO-d6): δ 10.47 (s, 1H), 9.06 (dd, J = 2.4 Hz, 0.9 Hz, 1H), 8.22 (dd, J = 8.4, 2.4 Hz,
Hz, 1H), 7.78 (dd, J = 9.0 Hz, 4.8 Hz, 2H), 7.64
(dd, J = 8.4 Hz, 0.6 Hz, 1H), 7.22 (t, J = 8.7 Hz, 2 H), 4.36 (s, 2H).
Intermediate 67 (290 mg, 1 mmol) and neopentyl glycol (1 mmol) were suspended in
anhydrous toluene (5 mL) and set to reflux over 24 h. Without isolation,
the resulting mixture containing the boronic ester was cooled in an
ice bath under argon. Pivaloyl chloride (120 μL, 1 mmol) and
triethylamine (140 μL, 1 mmol) were then added to the cooled
solution. The reaction was allowed to proceed at the lowered temperature
for 1 h, and then warmed to room temperature for an additional hour.
The resulting crystalline triethylammonium salt was filtered away,
and the reaction mixture concentrated by rotary evaporation. The resulting
oil containing 70 was diluted with anhydrous DMF (3 mL).
This solution was used without further isolation.
N-(4-Fluoro-phenyl)-6-chloro-nicotinamide (69)
4-Fluoroaniline (5.8 mL, 60 mmol) was stirred
in n class="Chemical">THF (300 mL) and potassium carbonate (16.6 g, 120 mmol) was added,
followed by 6-chloronicotinoyl chloride (10.6 g, 60 mmol) and stirred
overnight. The salts were removed by filtration, and the organic solution
was cooled in an ice bath and diluted with water while stirring. The
resulting white precipitate was filtered to yield 8.3 g (55%) of 69. TLC (ethyl acetate/hexanes 1:1) R = 0.51. ESI-MS: m/z = 251.0 [M + H]+. 1H NMR (500 MHz,
DMSO-d6): δ 10.53 (s, 1H), 8.95
(d, J = 2.5 Hz, 1H), 8.37 (dd, J = 8.4 Hz, 2.5 Hz, 1H), 7.81–7.70 (m, 3H), 7.26–7.18
(m, 2H). HPLC (gradient B): RT = 13.8 min, purity 94.3%.
N-Boc-2-piperidinecarbaldehyde (1.1 g,
1 n class="Chemical">eq., Combi-Blocks, San Diego, CA) and 4-fluoroaniline (481 μL,
1 equiv) were dissolved in DCE (20 mL). Under an inert atmosphere,
NaBH(OAc)3 (1.5 g, 1.4 equiv) and glacial acetic acid (294
μL, 1 equiv) were added, and the reaction was monitored by TLC
and LC–MS. The mixture was diluted with ethyl acetate and washed
with 10% citric acid (×3), brine (×1) and dried over MgSO4. The mixture was filtered and the solvent removed by rotary
evaporation and dried in vacuo to afford 1.4 g of the secondary aniline 70a (90% yield). The intermediate was carried forward without
further characterization. ESI-MS: m/z = 309.1 [M + H]+.
tert-Butyl
2-(4-fluorophenylamino)acetate (70b)
A solution
of 4-fluoroaniline (1.00 mL, 10 mmol)
and n class="Chemical">DIPEA (1.74 mL, 10 mmol) in anhydrous DMF (10 mL) was warmed to
80 °C, and then a solution of tert-butyl bromoacetate
(1.47 mL, 10 mmol) in anhydrous DMF (10 mL) was added dropwise over
1 h. After addition, the mixture was kept at 80 °C for 4 h. The
mixture was then concentrated by rotary evaporation and then partitioned
between ethyl acetate and water. The organic layer was washed with
water and then evaporated to yield 70b as a dark brown
liquid (1.96 g, 87%). A sample was purified by flash silica gel chromatography
(1:10 ethyl acetate/hexanes). TLC (ethyl acetate/hexanes, 1:4) R = 0.31. ESI-MS: m/z = 225.9 [M + H]+. 1H NMR
(500 MHz, DMSO-d6): δ 6.94–6.90
(m, 2H), 6.55–6.52 (m, 2H), 5.85 (s, 1H), 3.75 (d, J = 4.3 Hz, 2H), 1.41 (s, 9H). HPLC (gradient B): RT = 18.78
min, purity 86.5%.
The secondary aniline intermediate 70a (0.5 g, 1 equiv) was stirred in DMF (5 mL) at 80 °C
followed by the addition of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU,
240 μL, 1 equiv) and 6-chloronicotinoyl chloride (0.3 g, 1 equiv).
The reaction was monitored by TLC and LC–MS and was stirred
overnight at 80 °C. The mixture was then diluted with ethyl acetate
(25 mL), washed with water, saturated NaHCO3, water, 10%
citric acid, water, saturated NaCl, dried over Na2SO4, and filtered. The organic solvent was then removed by rotary
evaporation and dried in vacuo to afford 0.56 g of the N-substituted N-(4-fluorophenyl)-6-chloro-nicotinamide 71a (78% yield). The intermediate was carried forward without
further purification. ESI-MS: m/z = 448.2 [M + H]+.
N-(4-fluorophenyl)-6-chloro-nicotinamide
69 (1.0 g, 4.0
mmol) and n class="Chemical">2-bromomethyl-pyridine hydrobromide (1.01 g, 4.0 mmol) were
suspended in toluene (5 mL). 50% aqueous NaOH (3.0 mL) was added to
the mixture, followed by tetra-n-butyl-ammonium hydroxide
(TBAH, 100 μL). The biphasic reaction mixture was vigorously
stirred overnight and the aqueous layer removed by pipet. The organic
layer was diluted with ethyl acetate and washed with water (3 ×
25 mL) and saturated aqueous NaCl (3 × 25 mL). The water washes
were back-extracted once with ethyl acetate (25 mL). The combined
organic layers were dried over Na2SO4, filtered,
and dried by rotary evaporation to yield 1.3 g (96%) of N-(4-fluorophenyl)-N-pyridin-2-ylmethyl-6-chloro-nicotinamide
intermediate 72a as an off-white solid. ESI-MS: m/z = 341.9, 343.9 [M + H]+. 1H NMR (500 MHz, DMSO-d6): δ
8.52 (d, J = 4.3 Hz, 1H), 8.35 (d, J = 2.3 Hz, 1H), 7.79–7.73 (m, 2H), 7.50 (d, J = 5.4 Hz, 1H), 7.46 (d, J = 8.2 Hz, 1H), 7.33–7.27
(m, 3H), 7.10 (t, J = 8.6 Hz, 2H), 5.16 (s, 2H).
N-(4-fluorophenyl)-N-pyridin-2-ylmethyl-6-chloro-nicotinamide
intermediate 72a (3.4 g, 10 mmol) and anhydrous sodium
hydrogen sulfide (1.1 g, 20 mmol) were suspended in anhydrous DMF
(30 mL). The suspension was heated to reflux, and the mixture turned
a deep green color. After 2 h, the mixture was diluted with ethyl
acetate (150 mL) and extracted with water. The aqueous layer was acidified
with glacial AcOH to pH 6–7 and back extracted with ethyl acetate
(4 × 50 mL), and the organic layers were combined, dried over
Na2SO4, filtered, and evaporated by rotary evaporation
to yield 2.9 g (87%) of N-(4-fluorophenyl)-N-pyridin-2-ylmethyl-6-mercapto-nicotinamide intermediate 73a as a dark yellow oil. TLC (ethyl acetate/hexanes, 2:1): R = 0.1; ESI-MS: m/z = 339.9 [M + H]+.
Following
the same procedure described for 73a and starting from
the N-(4-fluorophenyl)-N-pyridin-4-ylmethyl-6-chloro-nicotinamide
intermediate 72b (2.5 g, 7.4 mmol), workup yielded 2.8
g (quant.) of compound crude 72b, which was carried forward
without further purification. TLC (ethyl acetate/hexanes, 2:1): R = 0.1; ESI-MS: m/z = 339.9 [M + H]+.
Thiolation of 71a was achieved
by suspending the tertiary amide (0.56 g, 1 equiv) and sodium hydrogen
sulfide (0.14 g, 2 equiv) in anhydrous DMF (2 mL) under an inert atmosphere.
The reaction was gently refluxed for 2 h and then diluted with ethyl
acetate (30 mL) and extracted with water (3 × 20 mL). The aqueous
layer was then acidified with 10% citric acid and stored at 2–8
°C. The fine white precipitate was then collected by filtration
to afford 90 mg of thionicotinamide intermediate 73c (16%
yield). The intermediate was carried forward without further purification.
The 6-chloronicotinamide 71b (211 mg, 0.58 mmol) and n class="Chemical">sodium hydrogen sulfide (74 mg) were dissolved
in DMF (1 mL). The mixture was heated to 85 °C for 15 min, and
then allowed to cool to room temperature. The mixture was diluted
in ethyl acetate, washed with water, saturated NaHCO3,
water, 10% citric acid, water, saturated NaCl, and dried over Na2SO4. The crude material was filtered and evaporated
to yield 123 mg (58%) of 73d as a yellow oil, which was
carried forward without further purification. TLC (ethyl acetate): R = 0.53; ESI-MS: m/z = 362.9 [M + H]+.
Authors: Michael P Dwyer; Younong Yu; Jianping Chao; Cynthia Aki; Jianhua Chao; Purakkattle Biju; Viyyoor Girijavallabhan; Diane Rindgen; Richard Bond; Rosemary Mayer-Ezel; James Jakway; R William Hipkin; James Fossetta; Waldemar Gonsiorek; Hong Bian; Xuedong Fan; Carol Terminelli; Jay Fine; Daniel Lundell; J Robert Merritt; Laura L Rokosz; Bernd Kaiser; Ge Li; Wei Wang; Tara Stauffer; Lynne Ozgur; John Baldwin; Arthur G Taveras Journal: J Med Chem Date: 2006-12-28 Impact factor: 7.446
Authors: David J Nicholls; Nick P Tomkinson; Katherine E Wiley; Anne Brammall; Lorna Bowers; Caroline Grahames; Alasdair Gaw; Premji Meghani; Philip Shelton; Tracey J Wright; Philip R Mallinder Journal: Mol Pharmacol Date: 2008-08-01 Impact factor: 4.436
Authors: Dean Y Maeda; Mark T Quinn; Igor A Schepetkin; Liliya N Kirpotina; John A Zebala Journal: J Pharmacol Exp Ther Date: 2009-09-24 Impact factor: 4.030
Authors: Waldemar Gonsiorek; Xuedong Fan; David Hesk; James Fossetta; Hongchen Qiu; James Jakway; Motasim Billah; Michael Dwyer; Jianhua Chao; Gregory Deno; Art Taveras; Daniel J Lundell; R William Hipkin Journal: J Pharmacol Exp Ther Date: 2007-05-11 Impact factor: 4.030
Authors: O Holz; S Khalilieh; A Ludwig-Sengpiel; H Watz; P Stryszak; P Soni; M Tsai; J Sadeh; H Magnussen Journal: Eur Respir J Date: 2009-07-30 Impact factor: 16.671
Authors: Petra de Kruijf; Jane van Heteren; Herman D Lim; Paolo G M Conti; Miranda M C van der Lee; Leontien Bosch; Koc-Kan Ho; Douglas Auld; Michael Ohlmeyer; Martin J Smit; Jac C H M Wijkmans; Guido J R Zaman; Martine J Smit; Rob Leurs Journal: J Pharmacol Exp Ther Date: 2009-02-03 Impact factor: 4.030
Authors: Dean Y Maeda; Angela M Peck; Aaron D Schuler; Mark T Quinn; Liliya N Kirpotina; Winston N Wicomb; Richard L Auten; Rambabu Gundla; John A Zebala Journal: Bioorg Med Chem Lett Date: 2015-04-23 Impact factor: 2.823
Authors: Aaron D Schuler; Courtney A Engles; Dean Y Maeda; Mark T Quinn; Liliya N Kirpotina; Winston N Wicomb; S Nicholas Mason; Richard L Auten; John A Zebala Journal: Bioorg Med Chem Lett Date: 2015-07-30 Impact factor: 2.823
Figure 1
Scheme 1
Scheme 4
Scheme 2
Scheme 3
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6