There exists an urgent medical need to identify new chemical entities (NCEs) targeting multidrug resistant (MDR) bacterial infections, particularly those caused by Gram-negative pathogens. 4-Hydroxy-2-pyridones represent a novel class of nonfluoroquinolone inhibitors of bacterial type II topoisomerases active against MDR Gram-negative bacteria. Herein, we report on the discovery and structure-activity relationships of a series of fused indolyl-containing 4-hydroxy-2-pyridones with improved in vitro antibacterial activity against fluoroquinolone resistant strains. Compounds 6o and 6v are representative of this class, targeting both bacterial DNA gyrase and topoisomerase IV (Topo IV). In an abbreviated susceptibility screen, compounds 6o and 6v showed improved MIC90 values against Escherichia coli (0.5-1 μg/mL) and Acinetobacter baumannii (8-16 μg/mL) compared to the precursor compounds. In a murine septicemia model, both compounds showed complete protection in mice infected with a lethal dose of E. coli.
There exists an urgent medical need to identify new chemical entities (NCEs) targeting multidrug resistant (MDR) n class="Disease">bacterial infections, particularly those caused by Gram-negative pathogens. 4-Hydroxy-2-pyridones represent a novel class of nonfluoroquinolone inhibitors of bacterial type II topoisomerases active against MDR Gram-negative bacteria. Herein, we report on the discovery and structure-activity relationships of a series of fused indolyl-containing 4-hydroxy-2-pyridones with improved in vitro antibacterial activity against fluoroquinolone resistant strains. Compounds 6o and 6v are representative of this class, targeting both bacterial DNA gyrase and topoisomerase IV (Topo IV). In an abbreviated susceptibility screen, compounds 6o and 6v showed improved MIC90 values against Escherichia coli (0.5-1 μg/mL) and Acinetobacter baumannii (8-16 μg/mL) compared to the precursor compounds. In a murinesepticemia model, both compounds showed complete protection in mice infected with a lethal dose of E. coli.
Infections caused by
drug resistant bacteria represent a major
threat to n class="Species">human health.[1] The Centers for
Disease Control and Prevention (CDC) estimates that in the United
States more than 2 million people acquire serious drug-resistant bacterial
infections resulting in more than 23,000 deaths annually.[2] The financial burden to the U.S. health system
alone has been estimated at billions of dollars annually.[3] Several initiatives, including the Infectious
Diseases Society of America’s “10 × ′20
Initiative”[4] and several public-private
partnerships of the Innovative Medicines Initiative (IMI),[5] have been developed to combat this antimicrobial
resistance threat. At the same time, there has been a large body of
literature highlighting the challenges involved in identifying novel
antibacterial agents.[6] Despite the urgent
clinical need for new chemical entities (NCEs) to treat these serious
infections, there has been a significant decline in the number of
new antibacterial agents developed and approved over the past three
decades.[6c,7] For NCEs that target infections caused by
Gram-negative pathogens, this problem is particularly acute.[4] Despite these challenges, there remains a significant
need for novel antibacterial agents that target these drug resistant
pathogens.
Fluoroquinolones are a major class of broad spectrum
antibacterial
agents that target the type II bacterial Dn class="Chemical">NA topoisomerases, DNA gyrase
and topoisomerase IV (Topo IV).[8] The function
of DNA gyrase and Topo IV is to maintain DNA in a proper topological
state during DNA replication and transcription.[9] Several additional classes of type II bacterial DNA topoisomerase
inhibitors have been reported, including 3-aminoquinazolinediones,[10] 4H-4-oxoquinolizines,[8e,11] isothiazolones,[12] quinolyl-piperidines,[13] and spiropyrimidinetriones.[14]
Recently, we disclosed a novel series of 4-hydroxy-2-pyridones,
e.g., 1a and 1b, that target type II bacterial
Dn class="Chemical">NA topoisomerases and are active against wild-type and resistant
Gram-negative pathogens (Figure ).[15] These compounds were
derived from a series of structurally related bacterial protein synthesis
inhibitors[16] by removal of a benzyl substituent
from the 2-pyridonenitrogen atom. Compound 1b had modest in vitro antibacterial activity and demonstrated efficacy
in a murinesepticemia model.[15] Against
an abbreviated panel of E. coli strains, the MIC90, defined as the lowest concentration inhibiting 90% of the
isolates, was 8 μg/mL.
Figure 1
Structures of novel 4-hydroxy-2-pyridone bacterial
topoisomerase
II inhibitors.
Structures of novel 4-hydroxy-2-pyridone bacterial
topoisomerase
II inhibitors.To expand the chemical
space and explore novel scaffolds with improved
antibacterial activity and pharmaceutical properties, we initially
envisioned modifying the substituted aniline moiety in 1a by cyclizing the methyl substituent onto the aryl ring attached
to the n class="Chemical">4-hydroxy-2-pyridone moiety to form an indole, e.g., 2. The resultant 5-indolyl-4-hydroxy-2-pyridones 2 retained a similar level and spectrum of antibacterial activity.
The 5-indolyl-4-hydroxy-2-pyridones can exist in multiple conformation
states by virtue of the rotational freedom around the bond connecting
the indole and 2-pyridone rings. To determine whether restricting
the rotational freedom could lead to analogues with improved antibacterial
activity, we designed and evaluated a set of compounds, 3, containing a tether between the indole and 2-pyridone rings serving
as a second attachment point.[17] This allowed
us to evaluate the effect of regioisomers, tether length, and constituents
of the tether on the antibacterial activity and pharmaceutical properties.
Herein, we report on the discovery and evaluation of novel conformationally
restricted analogues 3 that have improved minimum inhibitory
concentrations (MICs) and spectrum of antibacterial activity against
Gram-negative strains, and demonstrate efficacy in a murinen class="Disease">septicemia
model.
Results and Discussion
Chemistry
The general synthetic
route to 5-indolyl-4-hydroxy-2-pyridones 2a–2c is outlined in Scheme . n class="Chemical">5-Bromoindoles were converted
to ketones 9 by metal–halogen exchange followed
by reaction with N-methoxy-N-methylbutyramide.
Transformation of ketones 9 to the 5-indolyl-4-hydroxy-2-pyridones 2a–c was carried out by conversion of 9 to the respective t-butyl imines followed
by high temperature annulation with trimethylmethanetricarboxylate
to provide the 4-hydroxy-2-pyridone esters 10. Final
targets 2a–c were obtained by ester
dealkylation.
Scheme 1
Reagents and conditions: (a) n-BuLi (1.2 equiv), THF, −78 °C, then N-methoxy-N-methylbutyramide (1.2 equiv),
THF, −78 °C to rt, 50–75%; (b) t-butylamine (4 equiv), TiCl4 (0.65 equiv), CH2Cl2, 0 °C to rt, overnight; (c) CH(CO2Me)3 (1.7 equiv), Ph2O, 230 °C, 10 min,
21–56% over 2 steps; (d) LiI (3.0 equiv), EtOAc, 65 °C,
1 h, 54–67%.
Reagents and conditions: (a) n-BuLi (1.2 equiv), THF, −78 °C, then N-methoxy-N-methylbutyramide (1.2 equiv),
THF, −78 °C to rt, 50–75%; (b) t-butylamine (4 equiv), TiCl4 (0.65 equiv), CH2Cl2, 0 °C to rt, overnight; (c) CH(CO2Me)3 (1.7 equiv), Ph2O, 230 °C, 10 min,
21–56% over 2 steps; (d) LiI (3.0 equiv), EtOAc, 65 °C,
1 h, 54–67%.An alternative route,
depicted in Scheme , was utilized for the preparation of 5-indolyl-4-hydroxy-2-pyridones 2d–2n containing n class="Chemical">basic amines attached
to the indole C-2 through a methylene spacer. Starting with ethyl
5-bromo-1-methyl-1H-indole-2-carboxylate, DIBAL-H
reduction of the ester and protection of the resultant primary alcohol
with TBSCl gave 11. Conversion of 11 to
the ketone intermediates 12 was accomplished via metal–halogen
exchange of the indole 5-Br, followed by reaction of the 5-indolyllithium
species with Weinreb amides. Transformation to the 5-indolyl-4-hydroxy-2-pyridones 13 was carried out by initial conversion of the ketone to
the 2,4-DMB imine followed by high temperature annulation of the imine
with trimethylmethanetricarboxylate. TBS deprotection to the alcohols 14 followed by LiI-mediated ester dealkylation gave the hydroxy
acid intermediates 15. Alcohol oxidation provided the
indole-2-carboxaldehydes 16. Final targets 2d–n were obtained by reductive amination followed
by TFA-mediated DMB deprotection.
Scheme 2
Reagents and conditions:
(a)
DIBAL-H (2.2 equiv), −78 °C to rt, CH2Cl2; (b) TBSCl (1.3 equiv), imidazole (1.3 equiv), CH2Cl2, 0 °C, 96% over 2 steps; (c) n-BuLi (1.2 equiv), THF, −78 °C, followed by Weinreb amides
(1.2 equiv), THF, −78 °C to rt, 57–92%; (d) 2,4-dimethoxybenzylamine
(1.1 equiv), TiCl4 (0.65 equiv), Et3N (2.7 equiv),
CH2Cl2, 0 °C to rt; (e) CH(CO2Me)3 (1.7 equiv), diphenyl ether, 230 °C, 10 min,
37–56% over 2 steps; (f) TBAF (2.0 equiv), THF, 0 °C to
rt, 70–92%; (g) LiI (3.0 equiv), EtOAc, 65 °C, 1 h, 74–99%;
(h) MnO2 (20 equiv), CH2Cl2, rt,
1 h, 64–85%; (i) amine (2 equiv), AcOH (2 equiv), DCE, rt,
1 h, then NaBH(OAc)3 (2 equiv); (j) TIPS-H, TFA, 65 °C,
1 h, HCl, 12–57% yield over 2 steps.
Reagents and conditions:
(a)
DIBAL-H (2.2 equiv), −78 °C to rt, CH2Cl2; (b) TBSCl (1.3 equiv), imidazole (1.3 equiv), CH2Cl2, 0 °C, 96% over 2 steps; (c) n-BuLi (1.2 equiv), THF, −78 °C, followed by Weinreb amides
(1.2 equiv), THF, −78 °C to rt, 57–92%; (d) 2,4-dimethoxybenzylamine
(1.1 equiv), TiCl4 (0.65 equiv), Et3N (2.7 equiv),
CH2Cl2, 0 °C to rt; (e) CH(CO2Me)3 (1.7 equiv), diphenyl ether, 230 °C, 10 min,
37–56% over 2 steps; (f) TBAF (2.0 equiv), THF, 0 °C to
rt, 70–92%; (g) LiI (3.0 equiv), EtOAc, 65 °C, 1 h, 74–99%;
(h) MnO2 (20 equiv), CH2Cl2, rt,
1 h, 64–85%; (i) amine (2 equiv), AcOH (2 equiv), DCE, rt,
1 h, then NaBH(OAc)3 (2 equiv); (j) TIPS-H, TFA, 65 °C,
1 h, HCl, 12–57% yield over 2 steps.Several synthetic routes were developed to access basic amine-containing
tetracyclic 4-hydroxy-2-pyridones 3. For maximum structural
diversity and to evaluate analogues containing different tether lengths
and/or heteroatoms in the tether between the indole and 2-pyridone
rings, we envisioned utilizing cyclic ketones for the high temperature
annulation analogously to that described previously to provide the
tetracyclic 4-hydroxy-2-pyridones. Cyclic ketones of various ring
sizes were readily prepared by olefin metathesis.To prepare
tetracyclic derivatives bearing an all n class="Chemical">carbon tether
(forming 5–8-membered rings), we utilized a multistep synthetic
sequence to arrive at the TBS-protected indolyl cyclic ketone intermediates 24a–d (Scheme ). 4-Amino-2-chloro-5-iodobenzonitrile[18] was converted to the indole derivative 17 in two steps.[19] TBS protection
of the primary alcohol followed by N-methylation
of the indoleNH and DIBAL-H reduction of the nitrile gave 20. Dienes 22a–d, precursors to the
olefin metathesis, were obtained by Suzuki–Miyaura[20] cross-coupling to give the 6-vinylindole 21, followed by reaction with alkenylmagnesium halides to
give 22a–d. Ring closing olefin metathesis[21] using Grubbs second generation catalyst[22] provided the alkene-alcohols 23a–d. The key cyclic ketone intermediates were
obtained by reduction of the ring double bond followed by oxidation
of the secondary alcohol to provide the cyclic ketone intermediates 24a–d. Conversion of the cyclic ketones
to the 2,4-DMB imines, followed by high temperature annulation provided 25a–d. TBS deprotection followed by LiI-mediated
ester dealkylation gave the hydroxy acid intermediates 26a–d. Alcohol oxidation provided the indole-2-carboxaldehydes27a–d. Reductive amination followed by
TFA-mediated DMB deprotection provided the final targets 4a, 5a, 6a, 6b, 6i, 6n, 6p, 6r, 6v, 6z, 6bb, 6dd, and 7a.
Scheme 3
Reagents and conditions: (a)
propargyl alcohol (1.2 equiv), Et3N (2 equiv), Pd(PPh3)2Cl2 (1 mol %), CuI (2 mol %), CH3CN, 70 °C, 2 h; (b) t-BuOK (2.2 equiv),
DMF, 70 °C, 2 h, 77% over 2 steps; (c) TBSCl (1.2 equiv), imidazole
(1.3 equiv), DMF, rt, 1.5 h, 74%; (d) 60% NaH dispersion (1.4 equiv),
0 °C to rt, 10 min, then MeI (1.4 equiv), 0 °C to rt, 1.5
h, 76%; (e) DIBAL-H (1.2 equiv), CH2Cl2, −78
°C to −15 °C, then Rochelle salt, −40 °C
to rt, 1 h, 91%; (f) potassium vinyltrifluoroborate (1.5 equiv), Pd(OAc)2 (3 mol %), S-Phos ligand (6 mol %), K2CO3 (3 equiv), dioxane/H2O, 90 °C, 5 h, 89%; (g) 1-propenylmagnesium
chloride (n = 0), or allymagnesium chloride (n = 1), or 3-butenylmagnesium bromide (n = 2), or 4-pentenylmagnesium bromide (n = 3), THF,
−78 °C to rt, 90–100%; (h) Grubbs second generation
catalyst (3–5 mol %), toluene, 12–24 h; (i) TBSCl (2.1
equiv), imidazole (2.5 equiv), CH2Cl2, rt, overnight,
76%, then Grubbs second generation catalyst (3 mol %), toluene, 2
h, then TBAF (2.1 equiv), THF, 96 h, 20%, then TBSCl (1.1 equiv),
imidazole (1.2 equiv), THF, 0 °C to rt, 1.5 h, 77%; (j) 10% Pd–C
(n = 0 or 2–3), H2, 1 atm, EtOAc
CH2Cl2, rt, 3 h or 10% PtO2 (n = 1), toluene; (k) NMO, TPAP (n = 0 or
2–3), 4 Å, CH2Cl2, 0 °C to
rt or MnO2 (n = 1), CH2Cl2; (l) 2,4-dimethoxybenzyl amine (1.05 equiv), Et3N, (2.7 equiv), TiCl4 (0.65 equiv), CH2Cl2, 0 °C to rt, 16 h, then CH(CO2Me)3 (1.7 equiv), diphenyl ether, 230 °C, 10 min, 33–63%
over 2 steps; (m) TBAF (2.5 equiv), THF, rt, 2 h, then LiI (2.9 equiv),
EtOAc, 60 °C, 1.5 h, 84–91% over 2 steps; (n) MnO2 (15 equiv), CH2Cl2, rt, 69–82%;
(o) amine (2.0 equiv), AcOH (2.0 equiv), DCE, rt, 1 h, then NaBH(OAc)3 (2.0 equiv), 2 h; (p) TIPS-H, TFA, 60 °C, 2 h, then
HCl, 6–83% over 2 steps.
Reagents and conditions: (a)
propargyl alcohol (1.2 equiv), n class="Chemical">Et3N (2 equiv), Pd(PPh3)2Cl2 (1 mol %), CuI (2 mol %), CH3CN, 70 °C, 2 h; (b) t-BuOK (2.2 equiv),
DMF, 70 °C, 2 h, 77% over 2 steps; (c) TBSCl (1.2 equiv), imidazole
(1.3 equiv), DMF, rt, 1.5 h, 74%; (d) 60% NaH dispersion (1.4 equiv),
0 °C to rt, 10 min, then MeI (1.4 equiv), 0 °C to rt, 1.5
h, 76%; (e) DIBAL-H (1.2 equiv), CH2Cl2, −78
°C to −15 °C, then Rochelle salt, −40 °C
to rt, 1 h, 91%; (f) potassium vinyltrifluoroborate (1.5 equiv), Pd(OAc)2 (3 mol %), S-Phos ligand (6 mol %), K2CO3 (3 equiv), dioxane/H2O, 90 °C, 5 h, 89%; (g) 1-propenylmagnesium
chloride (n = 0), or allymagnesium chloride (n = 1), or 3-butenylmagnesium bromide (n = 2), or 4-pentenylmagnesium bromide (n = 3), THF,
−78 °C to rt, 90–100%; (h) Grubbs second generation
catalyst (3–5 mol %), toluene, 12–24 h; (i) TBSCl (2.1
equiv), imidazole (2.5 equiv), CH2Cl2, rt, overnight,
76%, then Grubbs second generation catalyst (3 mol %), toluene, 2
h, then TBAF (2.1 equiv), THF, 96 h, 20%, then TBSCl (1.1 equiv),
imidazole (1.2 equiv), THF, 0 °C to rt, 1.5 h, 77%; (j) 10% Pd–C
(n = 0 or 2–3), H2, 1 atm, EtOAcCH2Cl2, rt, 3 h or 10% PtO2 (n = 1), toluene; (k) NMO, TPAP (n = 0 or
2–3), 4 Å, CH2Cl2, 0 °C to
rt or MnO2 (n = 1), CH2Cl2; (l) 2,4-dimethoxybenzyl amine (1.05 equiv), Et3N, (2.7 equiv), TiCl4 (0.65 equiv), CH2Cl2, 0 °C to rt, 16 h, then CH(CO2Me)3 (1.7 equiv), diphenyl ether, 230 °C, 10 min, 33–63%
over 2 steps; (m) TBAF (2.5 equiv), THF, rt, 2 h, then LiI (2.9 equiv),
EtOAc, 60 °C, 1.5 h, 84–91% over 2 steps; (n) MnO2 (15 equiv), CH2Cl2, rt, 69–82%;
(o) amine (2.0 equiv), AcOH (2.0 equiv), DCE, rt, 1 h, then NaBH(OAc)3 (2.0 equiv), 2 h; (p) TIPS-H, TFA, 60 °C, 2 h, then
HCl, 6–83% over 2 steps.
Tetracyclic
derivatives fused through a seven-membered ring tn class="Chemical">ether
bearing an oxygen atom at the position adjacent to the indole core
were obtained via the TBS-protected indolyl cyclic ketone intermediate 35 (Scheme ). Compound 28 was obtained by etherification of 3-bromo-4-hydroxybenzaldehyde
followed by condensation with ethyl azidoacetate. Hemetsberger[23] cyclization followed by N-methylation
of the resulting indole gave ethyl indole-2-carboxylate 30. DIBAL-H reduction followed by TBS protection of the resultant alcohol
and metal–halogen exchange of the 5-bromo indole derivative
followed by formylation gave 5-formylindole 32. Vinylmagnesium
bromide addition followed by olefin metathesis gave 34. Oxidation of the alcohol and subsequent hydrogenation of the double
bond gave the cyclic ketone intermediate 35. Transformation
of 35 to the indole-2-carboxaldehydes 39 was carried out via the four-step sequence described previously.
Reductive amination and DMB deprotection provided the final targets 6c, 6f, 6j, 6o, 6q, 6s, 6w, 6aa, 6cc, and 6ee.
Scheme 4
Reagents and conditions:
(a)
allyl bromide (1.05 equiv), K2CO3 (1.1 equiv),
DMF, rt, 16 h, 93%; (b) ethyl azidoacetate (3 equiv), NaOEt (3.0 equiv),
EtOH, −10 °C to rt, 16 h, 62%; (c) xylenes, 140 °C,
1 h, 47%; (d) NaH (1.1 equiv), MeI (1.1 equiv), DMF, 0 °C to
rt, 1 h, 92%; (e) DIBAL-H (2.1 equiv), CH2Cl2, −78 to 0 °C, 30 min, then Rochelle salt, 0 °C
to rt, 1 h, 87%; (f) TBSCl (1.05 equiv), imidazole (1.05 equiv), CH2Cl2, 3 h; (g) n-BuLi (1.5 equiv),
THF, −78 °C, 30 min, then DMF, −78 °C to rt,
81% over 2 steps; (h) vinylmagnesium bromide (1.15 equiv), THF, 0
°C to rt, 1 h, 98%; (i) Grubbs second generation catalyst, toluene,
60 °C, 3 h, 51%; (j) MnO2 (7 equiv in three portions
over 2 h), CH2Cl2, rt; (k) H2, PtO2, EtOH, rt, 3 h, 71% over 2 steps; (l) 2,4-dimethoxybenzyl
amine (1.1 equiv), Et3N, (3.0 equiv), TiCl4 (0.6
equiv), CH2Cl2, 0 °C to rt, 16 h; (m) trimethylmethanetricarboxylate
(2.0 equiv), diphenyl ether, 230 °C, 15 min; 43% over 2 steps;
(n) TBAF (2.1 equiv), THF, rt, 1.5 h, 80%; (o) LiI (3 equiv), EtOAc,
60 °C, 3 h, 98%; (p) MnO2 (16 equiv in three portions
over 2 h), CH2Cl2, 64%; (q) amine (2.0 equiv),
AcOH (2.0 equiv), DCE, rt, 1 h, then NaBH(OAc)3 (2.0 equiv),
2 h; (r) TIPS-H, TFA, 60 °C, 1 h, then HCl/Et2O, 17–72%
over 2 steps.
Reagents and conditions:
(a)
allyl bromide (1.05 equiv), K2CO3 (1.1 equiv),
DMF, rt, 16 h, 93%; (b) ethyl azidoacetate (3 equiv), NaOEt (3.0 equiv),
EtOH, −10 °C to rt, 16 h, 62%; (c) xylenes, 140 °C,
1 h, 47%; (d) NaH (1.1 equiv), MeI (1.1 equiv), DMF, 0 °C to
rt, 1 h, 92%; (e) DIBAL-H (2.1 equiv), CH2Cl2, −78 to 0 °C, 30 min, then Rochelle salt, 0 °C
to rt, 1 h, 87%; (f) TBSCl (1.05 equiv), imidazole (1.05 equiv), CH2Cl2, 3 h; (g) n-BuLi (1.5 equiv),
THF, −78 °C, 30 min, then DMF, −78 °C to rt,
81% over 2 steps; (h) vinylmagnesium bromide (1.15 equiv), THF, 0
°C to rt, 1 h, 98%; (i) Grubbs second generation catalyst, toluene,
60 °C, 3 h, 51%; (j) MnO2 (7 equiv in three portions
over 2 h), CH2Cl2, rt; (k) H2, PtO2, EtOH, rt, 3 h, 71% over 2 steps; (l) 2,4-dimethoxybenzyl
amine (1.1 equiv), Et3N, (3.0 equiv), TiCl4 (0.6
equiv), CH2Cl2, 0 °C to rt, 16 h; (m) trimethylmethanetricarboxylate
(2.0 equiv), diphenyl ether, 230 °C, 15 min; 43% over 2 steps;
(n) TBAF (2.1 equiv), THF, rt, 1.5 h, 80%; (o) LiI (3 equiv), EtOAc,
60 °C, 3 h, 98%; (p) MnO2 (16 equiv in three portions
over 2 h), CH2Cl2, 64%; (q) amine (2.0 equiv),
AcOH (2.0 equiv), DCE, rt, 1 h, then NaBH(OAc)3 (2.0 equiv),
2 h; (r) TIPS-H, TFA, 60 °C, 1 h, then HCl/Et2O, 17–72%
over 2 steps.The azepino-indole 4-hydroxy-2-pyridones 6d, 6g, and 6l were prepared analogously
to the oxepino-n class="Chemical">indole
4-hydroxy-2-pyridones described in Scheme as detailed in the Supporting
Information. Thiepino-indole 4-hydroxy-2-pyridones 6e, 6h, 6m, 6u, and 6y were prepared via straightforward chemistry, first building a masked
4-hydroxy-2-pyridone before constructing the substituted indole moiety
as detailed in the Supporting Information. Compounds 6k, 6t, and 6x, substituted with O in the middle of the three-atom tether, were
prepared as described in the Supporting Information. Regioisomeric tetracyclic 4-hydroxy-2-pyridones 8a and 8b were prepared as described in the Supporting Information.
Lead Optimization
MIC values were measured against
a panel of wild-type and resistant Gram-negative strains, including
permeable Escherichia coli BAS849[24] (n class="Species">E. colip), wild-type E. coli ATCC 25922 (E. coliWT), highly fluoroquinolone resistant E. coli strains
SKM18 (E. coliR1),[25] bearing four mutations (two in gyrA and
two in parC) in the quinolone resistance-determining
region (QRDR), and clinical isolate ELZ4251 (E. coliR2).[26] Additionally, wild-type
strains of Acinetobacter baumannii ATCC BAA-747 (A. bauWT) and Klebsiella pneumoniae ATCC 35657 (K. pneuWT) as well as fluoroquinolone
resistant clinical isolate A. baumannii MMX2240 (A. bauR)[27] were routinely
screened against. Bacterial MICs were determined by following CLSI
guidelines with the exception that organisms were grown in brain–heart–infusion
media.[28] As a surrogate measure of serum
shift, MICs were also determined against E. coliWT in the presence of 40 mg/mL of human serum albumin (HSA).
Previously, we described the discovery and optimization of aminophenyl-substituted
4-hydroxy-2-pyridones active against multidrug resistant (MDR) Gram-negative
bacteria and the requirement of the 4-hydroxy group as a key pharmacophoric
structural feature for imparting activity against n class="Chemical">fluoroquinolone
resistant strains.[15] Compound 1a exhibited good activity against several wild-type Gram-negative
strains including E. coliWT, A.
bauWT, and K. pneuWT (MIC values of 0.78, 1.6, and 3.1 μg/mL, respectively) and
modest activity (MIC = 15.6 μg/mL) against the highly fluoroquinolone
resistant E. coli strain SKM18 (E. coliR1). Similar activity was observed against the fluoroquinolone
resistant E. coli clinical isolate ELZ4251 (E. coliR2) (Table ).
Table 1
In Vitro Antibacterial
Activity of 5-Indolyl-4-hydroxy-2-pyridones
Minimum inhibitory concentration. E. colip = E. coli BAS849; E. coliWT = E. coli ATCC 25922; E. coliR1 = E. coli SKM18; E. coliR2 = E. coli ELZ4251; A. bauWT = A. baumannii ATCC
BAA-747; A. bauR = A. baumannii MMX2240; K. pneuWT = K. pneumoniae ATCC 35657.
Not determined.
Minimum inhibitory concentration. E. colip = n class="Species">E. coli BAS849; E. coliWT = E. coli ATCC 25922; E. coliR1 = E. coli SKM18; E. coliR2 = E. coli ELZ4251; A. bauWT = A. baumannii ATCC
BAA-747; A. bauR = A. baumannii MMX2240; K. pneuWT = K. pneumoniae ATCC 35657.
Not determined.As a part of a lead optimization
campaign focused on improving
activity against resistant Gram-negative strains, we envisioned fusing
the dimethylamino group onto the phenyl ring, thereby replacing the
n class="Chemical">aniline motif with an indole scaffold, obtaining the 5-indolyl-4-hydroxy-2-pyridone 2a. Compared to 1a, compound 2a had
similar activity against the wild-type strains of E. coli, A. baumannii, and K. pneumoniae but exhibited a 2- to 8-fold reduction in MICs against the E. coli and A. baumannii resistant strains.
Adding Me to R1 (2b) had a detrimental effect
on activity against the resistant E. coli strains
while adding an additional Me group at R2 (2c) led to complete loss of activity against the strains evaluated.
Compounds 2a and 2b exhibited a >40-fold
shift in the MIC in the presence of HSA. A greater than 25-fold increase
in MIC values against the wild-type E. coli strain
compared to the permeable strain in compounds 2a and 2b suggest that either limited membrane permeability and/or
drug efflux are contributing factors to the reduction in activity.
Appending a basicnitrogen atom to R1 in 2b (e.g., 2d) resulted in marked improvement in MICs including
an 8-fold reduction in MIC shift between E. coliWT and E. coliP. The basicnitrogen
atom renders 2d zwitterionic, which could result in improved
porin transit.[29] Further, the basicnitrogen
in 2d resulted in a >4-fold reduction in the MIC shift
in the presence of HSA against E. coliWT compared to 2b.[15]
We next evaluated the effect of the R3 substituent size
on antibacterial activity (Table ). Compared to 2d, replacement of R3 with H (2e) resulted in a dramatic reduction
in activity while replacement with Me (2f) exhibited
a 4- to 8-fold reduction in activity against the wild-type strains
but a 2- to 4-fold improvement in activity against the resistant strains.
Compared to 2d, a small increase in substituent size,
e.g., i-Pr (2g) and c-Pr (2h), led to reduced activity. Overall, Et substitution
at R3 provided the best balance of antibacterial activity
and was utilized for further SAR investigations.Additional
aliphatic secondary and tertiary n class="Chemical">basic amines (2i–2n) appended to the indole R2 were next investigated
(Table ). In general,
these compounds displayed similar MICs
(within 4-fold) compared to 2d against the bacterial
panel, exhibiting some tolerance to substituent size at R2. Increased MIC shifts in the presence of HSA were generally 2- to
4-fold. Overall, while compounds 2d and 2i–m exhibited elevated MICs against E.
coliWT, they exhibited far superior activity against
the fluoroquinolone resistant strains E. coliR1 and E. coliR2 compared to the
control compound, ciprofloxacin.
Noting the effect of R3 on antibacterial activity, we
surmised that the dihedral angle between the n class="Chemical">indole and 2-pyridone
rings might play an important role. To gain better insight into the
preferred conformational state of the biaryl system, we generated
an energy profile for a model compound (2b, R1 = Me) encompassing a full 360° rotation around the C–C
bond between the indole and 2-pyridone rings (Figure ).[30] In the lowest
energy conformation, the dihedral angle (θ) between the two
rings is 129°. In this twisted conformation, the steric clash
between the 2-pyridone ethyl group and the indole C-4 and C-6 hydrogens
is reduced, while retaining considerable aromatic conjugation in the
biaryl system. Further, the energy profile revealed two energetically
equivalent rotamers differing in the position of the indolenitrogen
atom facing either “in” or “out” of the
plane as depicted in Figure (see Supplemental Graph 1).
Figure 2
Rotation about
the C–C bond in the 5-indolyl-4-hydroxy-2-pyridones.
Rotation about
the C–C bond in the 5-indolyl-4-hydroxy-2-pyridones.We next restricted the rotation
around the C–C bond joining
the two rings by introducing a tether between the n class="Chemical">pyridone ring ethyl
group and the indole ring as a means to gain an entropic advantage
and potentially increase antibacterial activity. We had two objectives:
(1) determine the optimal tether length conferring maximum activity
and (2) determine whether rotamer I or rotamer II is the preferred
conformation. To address this, we modeled four tetracyclic systems
containing tethers of one to four carbons (forming 5- to 8-membered
rings) between the pyridone C-5 and the indole C-6 ([5,6]-fusion nomenclature)
and determined their equilibrium geometries and the dihedral angle
for those conformations (Table ). As expected, the five-membered ring system is completely
rigid with θ = 180°. Adding an additional carbon atom to
the tether (six-membered ring) allowed for some flexibility with a
predicted dihedral angle of 160° in the most energetically stable
conformation. Further increasing the size of the tether to seven-
and eight-membered rings allowed for several conformations with predicted
optimal θ values of 135° and 123°, respectively. In Figure , each of the optimized
constrained tetracyclic analogues is overlaid with the acyclic analogue 2b in its most energetically stable conformation. The seven-
and eight-membered ring tethered compounds, with optimal dihedral
angles similar to the acyclic analogue, provided the best overlay
of the indole and 2-pyridone rings. The five- and six-membered ring
tethered compounds placed the indole ring in different spatial orientations.
Table 2
In Vitro Antibacterial
Activity of Constrained Analogs: Ring Size
MIC (μg/mL)b
E. coli enzymatic activityc (IC50, μM)
ring size
compd
dihedral
anglea (calcd)
E. coliP
E. coliWT
E. coliR1
DNA gyrase
Topo IV
acyclic
2j
129
0.10
0.39
1.6
0.3
3.3
five-membered
4a
180
>12.5
>62.5
>62.5
>11.1e
>11.1e
six-memberedd
5a
160
3.1
7.8
7.8
0.9
37.7
seven-membered
6a
135
0.10
0.39
0.39
0.1
2.9
eight-membered
7a
123
0.19
0.78
3.1
0.8
15.3
ciprofloxacin
0.006
0.012
>250
0.19
3.5
Calculated energy minimized dihedral
angles. See ref (30).
Minimum inhibitory concentration. E. colip = E. coli BAS849; E. coliWT = E. coli ATCC 25922; E. coliR1 = E. coli SKM18.
See ref (31).
N-Methylamino analogue.
Highest concentration tested.
Figure 3
Structural
overlays of constrained analogs: (A) 2b (green) and five-membered
ring constrained analogue (blue); (B) 2b and six-membered
ring constrained analogue (orange); (C) 2b and seven-membered
ring constrained analogue (red); (D) 2b and eight-membered
ring constrained analogue (yellow).
Structural
overlays of constrained analogs: (A) 2b (green) and five-membered
ring constrained analogue (blue); (B) 2b and six-membered
ring constrained analogue (orange); (C) 2b and seven-membered
ring constrained analogue (red); (D) 2b and eight-membered
ring constrained analogue (yellow).Calculated energy minimized dihedral
angles. See ref (30).Minimum inhibitory concentration. E. colip = n class="Species">E. coli BAS849; E. coliWT = E. coli ATCC 25922; E. coliR1 = E. coli SKM18.
See ref (31).N-Methylamino analogue.Highest concentration tested.To confirm the effect of the predicted
geometries on the antibacterial
activity, we prepared the four tethered n class="Chemical">tetracyclic compounds 4a–7a containing basic amines appended
to the indole R1 (Table ). Good correlation between the dihedral angle and
activity was observed. Compound 4a, most structurally
dissimilar to acyclic compound 2j (R1 = CH2NHEt), was essentially inactive against the E. coli strains. Compounds 6a and 7a, most similar
to acyclic analogue 2j, provided the best activity with
seven-membered ring analogue 6a demonstrating a 4-fold
reduction in MIC value against E. coliR1 compared to 2j. We also determined inhibitory activities
against the target enzymes DNA gyrase and Topo IV.[31] Similar to the effect on MIC values, 6a exhibited
greater potency against the target enzymes than the other tethered
analogues. These results prompted an expanded SAR investigation of
the seven-membered ring tethered tetracyclic series.
We then
assessed whether the active conformation was better defined
by sterically constrained rotamer I or II (Figure ). Compounds 8a and 8b, containing a three-n class="Chemical">carbon tether (seven-membered ring) between
the pyridone C-5 and the indole C-4 ([5,4] fusion), restrict the conformation
in the form of rotamer II (Figure , Table ). Unlike the two rotamers shown for 2b in Figure , the [5,6]- and
[5,4]-regiosomeric tetracycles cannot interconvert (Table ). The data indicate a clear
activity preference toward the [5,6]-fused regioisomers. Compared
to 6a and 6b, the [5,4]-fused regioisomers 8a and 8b resulted in a >30-fold increase
in
MIC values against E. coliP and E. coliWT and a 16-fold increase in MIC values
against E. coliR1, respectively. Compounds 8a and 8b also demonstrated a 5- to 10-fold reduction
in inhibitory activity against the E. coli target
enzymes, DNA gyrase and Topo IV.
Table 3
In Vitro Antibacterial
Activity of Constrained Analogs: [5,6]- vs. [5,4]-Fused Regioisomers
MIC (μg/mL)a
E. coli enzymatic activityb (IC50, μM)
compd
fusion
NRaRb
E. coliP
E. coliWT
E. coliR1
DNA gyrase
Topo IV
6b
[5,6]
NHMe
0.19
0.39
0.39
0.1
3.7
6a
[5,6]
NHEt
0.10
0.39
0.39
0.1
2.9
8a
[5,4]
NHMe
6.3
>12.5
6.3
1.0
27.9
8b
[5,4]
NHEt
6.3
12.5
6.3
n.d.c
19.0
ciprofloxacin
0.006
0.012
>250
0.19
3.5
Minimum inhibitory concentration. E. colip = E. coli BAS849; E. coliWT = E. coli ATCC 25922; E. coliR1 = E. coli SKM18.
See ref (31).
Not determined.
Minimum inhibitory concentration. E. colip = n class="Species">E. coli BAS849; E. coliWT = E. coli ATCC 25922; E. coliR1 = E. coli SKM18.
See ref (31).Not determined.Inspired by the activity profiles of 6a and 6b, we sought to further investigate the tetracyclic
series
in more detail. Having determined the optimal tn class="Chemical">ether length (three
atoms) and preference for [5,6]-fusion, we next investigated two additional
points of diversity: incorporation of heteroatoms in the linker and
modification of the basic amine groups appended to the indole C-2
(Table ). Several
analogs containing secondary and tertiary aliphatic amines (NRaRb) were evaluated. For compounds containing a
three-carbon tether (X,Y = CH2, cyclohepta[1,2-f]indole), secondary amines 6a, 6b, 6i, 6n, and 6p and tertiary
amines 6r and 6v exhibited similar activity
against wild-type strains of E. coli, A.
baumannii, and K. pneumoniae with MIC values
ranging from 0.39 to 1.6 μg/mL. Compared to E. coliWT, MIC values against E. coliR1 and E. coliR2 typically increased 2-
to 4-fold, while activity against A. bauR was generally similar to A. bauWT. These
compounds exhibited a 4- to 16-fold shift in MIC values in the presence
of 40 mg/mL of HSA. Antibacterial activity was generally independent
of the nature of the aliphatic amines evaluated.
Table 4
In Vitro Antibacterial
Activity of Amine Substitution
Minimum inhibitory concentration. E. colip = E. coli BAS849; E. coliWT = E. coli ATCC 25922; E. coliR1 = E. coli SKM18; E. coliR2 = E. coli ELZ4251; A. bauWT = A. baumannii ATCC
BAA-747; A. bauR = A. baumannii MMX2240; K. pneuWT = K. pneumoniae ATCC 35657.
Not determined.
Minimum inhibitory concentration. E. colip = n class="Species">E. coli BAS849; E. coliWT = E. coli ATCC 25922; E. coliR1 = E. coli SKM18; E. coliR2 = E. coli ELZ4251; A. bauWT = A. baumannii ATCC
BAA-747; A. bauR = A. baumannii MMX2240; K. pneuWT = K. pneumoniae ATCC 35657.
Not determined.Replacement of the methylene
group with an oxygen atom at the α-position
of the tn class="Chemical">ether (X = O, Y = CH2, oxepino[3,2-f]indole) resulted in a different SAR profile. For compounds with
small secondary amines, e.g., 6c (Me) and 6f (Et), MIC values generally increased 2- to 4-fold against E. coliWT and E. coliR1 compared to the respective methylene analogues (6b and 6a), while more dramatic shifts in MICs (8- to 16-fold) were
observed against E. coliR2, A.
bauWT, and A. bauR. For oxepino[3,2-f]indoles containing branched
secondary amines, e.g., 6j (i-Pr), 6o (sec-Bu), 6q (1-Me-c-Bu), and tertiary amines 6s (dimethyl) and 6w (pyrrolidine), MIC values were similar to the analogous
cyclohepta[1,2-f]indole analogues. The increased
polarity imparted by the tetheroxygen in the oxepino[3,2-f]indole series resulted in smaller MIC shifts (2- to 4-fold)
in the presence of HSA compared to similar analogues with the all-carbon
tether (4- to 16-fold shifts). Compounds 6o and 6q displayed good antibacterial activity across the panel
of strains (MICs ≤ 2 μg/mL) and exhibited similar or
lower MICs against the resistant strains of E. coli and A. baumannii compared to the corresponding
wild-type strains.
Transposition of the oxygen atom to the β
position of the
tether (X = CH2, Y = O, oxepino[4,3-f]indole)
resulted in an overall reduction in antibacterial activity. Compounds 6k and 6t exhibited higher MIC values against
wild-type E. coli and both the wild-type and resistant
strains of A. baumannii compared to the congener
pairs 6i, 6j and 6r, 6s, respectively. In contrast, against the highly fluoroquinolone
resistant strain E. coliR1, 6k and 6t exhibited similar activity to the related congeners.Introduction of a polar, hydrogen bond donating n class="Chemical">amine (X = NH,
Y = CH2, azepino[3,2-f]indole) at the
α-position of the tether was generally disfavored. Compounds 6d, 6g, and 6l exhibited modest
to weak antibacterial activity against E. coliWT and the clinical isolate E. coliR2 as well as wild-type and resistant A. baumannii and modest activity against K. pneuWT. It should be noted that all three compounds exhibited improved
antibacterial activity against the highly fluoroquinolone resistant
strain E. coliR1 (MIC = 3.9 μg/mL)
compared to the wild-type strain (MIC range 7.8–31.3 μg/mL).
Although the polar nitrogen atom in the tether generally resulted
in no shift in MIC values in the presence of HSA, these compounds
generally exhibited weak antibacterial activity.
Replacement
of the methylene group with the larger, more lipophilic
sulfur atom at the α-position of the tn class="Chemical">ether (X = S, Y = CH2, thiepino[3,2-f]indole) resulted in a 2-
to 4-fold decrease in antibacterial activity (compounds 6e, 6h, 6m, 6u, and 6y). As a consequence of the longer C–S bond lengths (∼1.8
Å vs ∼1.5 Å for C–C bonds), the topology of
the tetracyclic series is altered, effectively lengthening the tether
to between that of a three and four carbon chain. As a result of the
increased lipophilicity, an 8- to 16-fold shift in E. coliWT MICs in the presence of HSA was observed for these
analogues.
We next evaluated the effect on antibacterial activity
of adding
additional heteroatoms to the basic amine moiety in both the n class="Chemical">cyclohepta[1,2-f]indole (X = CH2) and oxepino[3,2-f]indole (X = O) series (6z and 6aa–6ee). Substitution of the alkyl amine moiety with OMe favored
the carbon-fused analogue (6z) compared to the oxygen-fused
analogue (6aa) with no net improvement in activity compared
to the corresponding ethylamino analogues 6a and 6f, respectively. Adding a second basic amine moiety (NMe2) to the alkyl amine group (6bb and 6cc) resulted in elevated MIC values across multiple strains. Methoxy-substituted
pyrrolidine6dd provided activity similar to 6v, while similar substitution of the oxygen-fused analogue (6ee) resulted in increased MIC values across several strains.
To more readily visualize activity differences, data are presented
graphically in the Spotfire plot depicted in Figure in which antibacterial activities against E. coliR1 and n class="Species">A. baumanniiR are compared across four series (acyclic and three cyclic
series) containing ten distinct amine fragments.[32] The seven-membered carbon tethered 4-hydroxy-2-pyridones
(cyclohepta[1,2-f]indoles, blue circles) display
a marked improvement in MICs against both E. coliR1 and A. baumanniiR compared
to the acyclic series (5-indolyl-4-hydroxy-2-pyridones, black circles).
The fused oxepino[3,2-f]indoles (red circles) and
thiepino[3,2-f]indoles (green circles) generally
show a 2- to 4-fold improvement in MICs against E. coliR1, but no obvious overall improvement against A. baumanniiR.
Figure 4
Spotfire plot comparing activities of
acyclic 5-indolyl series
and three fused series containing different amine fragments against E. coli SKM18 (E. coliR1) and A. baumannii MMX2240 (A. baumanniiR).
Spotfire plot comparing activities of
acyclic 5-indolyl series
and three fused series containing different n class="Chemical">amine fragments against E. coli SKM18 (E. coliR1) and A. baumannii MMX2240 (A. baumanniiR).
Preliminary Mechanistic
Evaluation
Compound 2m, an early representative
analogue, was evaluated against wild-type
strains of both E. coli and n class="Species">A. baumannii at several concentrations to determine the mode of bacterial inhibition.
As shown in the time–kill kinetics plot (Figure ), 2m shows concentration-dependent
bactericidal activity against both pathogens. Rapid killing is observed
at concentrations above the MIC against E. coli (Figure a) and above 2-fold
of the MIC of 2m against A. baumannii (Figure b). At these
concentrations, reduction in the number of CFUs (colony forming units)
by more than 3 log10 units occurred within ∼2 h
in the E. coli strain and ∼2 log10 units within the same time period for the A. baumannii strain. This concentration-dependent killing kinetics mirrors what
was observed for the known bactericidal antibiotic ciprofloxacin (Supporting Information, Figure ).
Figure 5
Time–kill kinetics plots for compound 2m against E. coliWT (5a) and A. baumanniiWT (5b).
Time–kill kinetics plots for compound 2m against E. coliWT (5a) and A. baumanniiWT (5b).To explain the enhanced activity
observed for the 4-hydroxy-2-pyridones
against n class="Chemical">fluoroquinolone resistant strains, we compared the activities
of 6a and ciprofloxacin against wild-type E.
coli DNA gyrase (GyrAWT) and DNA gyrase that contain
two mutations in QRDR, S83L and D87Y (GyrASD-LY),
which confer significant resistance to the fluoroquinolones.[25] For 6a, there was a minimal shift
in activity against DNA gyraseWT compared to the DNA gyraseSD-LY, with IC50 values of 0.56 and 0.76
μM, respectively (Table ). Ciprofloxacin exhibited a greater than 13-fold reduction
in activity (IC50 = 2.6 μM) against DNA gyraseSD-LY compared to DNA gyraseWT (IC50 = 0.19 μM). There was a good correlation with the activity
against the E. coli DNA gyrases and antibacterial
activity. MIC values against wild-type E. coli strain
1609 (E. coliWT2)[9] and E. coli containing the S83L and D87YGyrA mutations,
SKM11 (E. coliR3),[24] were both 1.6 μg/mL, reflecting the similar IC50 values against both gyrases (Table ). The reduced activity against DNA gyraseSD-LY for ciprofloxacin is consistent with the 32-fold
increase in MIC observed against E. coliR3, which contains both mutations in GyrA. E. coli SKM18 (E. coliR1) is an isogenic construct
that contains, in addition to S83L and D87Y, two additional mutations
in Topo IV that confer resistance to fluoroquinolones (S80L and E84G).
Against E. coliR1, 6a displayed
a 4-fold decrease in MIC, while these additional mutations in the
QRDR confer a ∼250-fold increase in MIC for ciprofloxacin.
Table 5
E. coli MICs and
DNA Gyrase Activity for Compound 6a and Ciprofloxacin
MIC (μg/mL)a
E. coli DNA gyrase (IC50, μM)
compd
E. coliWT2
E. coliR3
E. coliR1
gyraseWT
gyraseSD-LY
6a
1.6
1.6
0.39
0.56
0.76
ciprofloxacin
0.024
0.78
200
0.19
2.6
Minimum inhibitory concentration. E. coliWT2 = E. coli 1609; E. coliR3 = E. coli SKM11; E. coliR1 = E. coli SKM18.
Minimum inhibitory concentration. E. coliWT2 = E. coli 1609; E. coliR3 = E. coli SKM11; E. coliR1 = E. coli SKM18.
Susceptibility Testing
Several representative compounds
with favorable MIC profiles from the cyclohepta[1,2-f]indole (6a, 6i, 6r, and 6v) and n class="Chemical">oxepino[3,2-f]indole (6o, 6q, 6s, and 6w) series were
selected for susceptibility testing against a panel of 12 E. coli and 12 A. baumannii isolates to
determine abbreviated MIC50 and MIC90 values
(Table ).[33] The panel contained clinical isolates with varying
degrees of resistant phenotypes including six highly fluoroquinolone
resistant E. coli strains, three of which were MDR,
and six fluoroquinolone resistant A. baumannii strains,
three of which were MDR. Greater potency was observed against the E. coli isolates (MIC90 values of 0.5–2
μg/mL) compared to the A. baumannii isolates
(MIC90 values of 8 to >16 μg/mL). The comparator
fluoroquinolone, levofloxacin, exhibited MIC90 values of
32 and 16 μg/mL against the E. coli and A. baumannii isolates, respectively. All compounds shown
in Table exhibited
good activity against E. coli DNA gyrase and Topo
IV as well as good selectivity to human topoisomerase II (see Supplemental Table 1).
Table 6
Susceptibility
Testing against E. coli and A. baumannii Isolates for Selected
Compoundsa
E. coli (μg/mL)
A. baumannii (μg/mL)
compd
MIC90
MIC50
MIC range
MIC90
MIC50
MIC
range
2j
4
1
0.25–4
16
2
0.5–32
6a
1
1
0.25–1
16
2
0.25–32
6i
1
0.5
0.25–1
16
2
0.25–32
6o
1
1
0.12–1
8
2
≤0.06 to >16
6q
1
0.5
0.12–1
8
1
≤0.06–16
6r
1
0.5
0.25–2
>16
2
0.12–16
6s
2
0.5
0.25–8
16
2
0.25 to >16
6v
0.5
0.5
0.12–0.5
16
2
0.12–16
6w
1
0.5
0.25–1
8
2
0.12–16
levofloxacin
32
0.06
0.015–32
16
2
0.06–64
Susceptibility testing was performed
against 12 strains of E. coli and A. baumannii. See ref (33) for
strain information.
Susceptibility testing was performed
against 12 strains of E. coli and A. baumannii. See ref (33) for
strain information.
Pharmacokinetics
and Efficacy Screening
Pharmacokinetic
parameters were measured for representative compounds from the cyclohepta[1,2-f]indole and n class="Chemical">oxepino[3,2-f]indole series
after intravenous administration to CD-1mice (Table ). With the exception of 6i,
the compounds exhibited similar profiles with moderate clearance (35–53
mL/min/kg) and a large volume of distribution (7–13 L/kg) resulting
in a terminal half-life of approximately 2–3 h. Compound 6i exhibited higher clearance, substantially larger volume
of distribution, and a longer half-life. These compounds displayed
oral bioavailability (F, %) in the range of 33–60%,
which is lower than the oral bioavailability observed for most fluoroquinolones
(70–90%).[34] Plasma protein binding
for these compounds exhibited unbound fractions in mouse plasma (fu) of 0.047–0.23.
Table 7
Pharmacokinetic Properties, Activity,
and in Vivo Efficacy of Selected Compounds
pharmacokinetic
properties, mousea
E. coliWT
compd
AUC0–24h (h·μg/mL)
t1/2b (h)
Vdssc (L/kg)
Cld (mL/min/kg)
F (%)e
PPB fuf
MICg (μg/mL)
survivalh,i,j (%)
6a
n.d.k
n.d.k
n.d.k
n.d.k
n.d.k
0.17
0.39
83
6i
1.81
11
69
64
44
0.16
0.78
83
6o
3.44
2.1
7.1
48
41
0.23
0.78
100
6p
4.58l
3.1
7.1
35
65
0.047
1.0
0
6q
4.42
3.3
13
37
50
0.15
1.0
67
6r
3.52
2.9
11
47
42
0.10
0.78
100
6s
2.93l
3.1
7.6
53
n.d.k
0.21
1.0
100
6v
4.05
2.8
8.1
41
33
0.11
0.78
100
6w
n.d.k
n.d.k
n.d.k
n.d.k
n.d.k
0.18
2.0
83
moxifloxacin
0.64m
1.8
3.7
72
n.d.k
0.80
0.024
100n
Compounds
were dosed at 10 mg/kg
iv (n = 3) in 10% DMSO and 1% Tween 80 in pH 8.0
phosphate buffer in CD-1 mice.
Terminal plasma half-life.
Steady-state volume of distribution.
Clearance.
Oral bioavailability.
Mouse
plasma protein binding unbound
fraction determined by equilibrium dialysis (1 μM compound concentration).
Minimum inhibitory concentration
against wild-type E. coli (ATCC 25922).
Percent of mice (n =
6) surviving 96 h postinfection after intravenous administration
of compound in 3% DMSO/saline.
Compounds were dosed at 20 mg/kg
at 1 and 5 h postinfection.
Prior to inoculation with bacteria,
mice were made neutropenic with 150 mg/kg cyclophosphamide (ip) on
day −4 and 100 mg/kg cyclophosphamide on day −1 (ip).
Mice were infected 100 × LD50 (1 × 106 bacteria).
Not determined.
Dosed at 10 mg/kg iv (n = 3) in 20% (2-hydroxypropyl)-β-cyclodextrin in
pH 4.0 citrate
buffer.
Dosed at 3 mg/kg
iv (n = 3) in 3% DMSO/saline.
Dosed at 3 mg/kg at 1 and 5 h postinfection.
Compounds
were dosed at 10 mg/kg
iv (n = 3) in 10% DMSO and 1% Tween 80 in pH 8.0
phosphate buffer in CD-1mice.Terminal plasma half-life.Steady-state volume of distribution.Clearance.Oral bioavailability.Mouse
plasma protein binding unbound
fraction determined by equilibrium dialysis (1 μM compound concentration).Minimum inhibitory concentration
against wild-type E. coli (ATCC 25922).Percent of mice (n =
6) surviving 96 h postn class="Disease">infection after intravenous administration
of compound in 3% DMSO/saline.
Compounds were dosed at 20 mg/kg
at 1 and 5 h postinfection.Prior to inoculation with bacteria,
mice were made neutropenic with 150 mg/kg cyclophosphamide (ip) on
day −4 and 100 mg/kg cyclophosphamide on day −1 (ip).
Mice were infected 100 × LD50 (1 × 106 bacteria).Not determined.Dosed at 10 mg/kg iv (n = 3) in 20% (2-hydroxypropyl)-β-cyclodextrin in
pH 4.0 citrate
buffer.Dosed at 3 mg/kg
iv (n = 3) in 3% DMSO/saline.Dosed at 3 mg/kg at 1 and 5 h postinfection.The compounds were evaluated for
efficacy in a mousen class="Disease">septicemia
model (Table ). Female
CD-1mice were made neutropenic by treatment with cyclophosphamide
and then inoculated by intraperitoneal injection with 100× the
LD50 (lethal dose at which 50% of the mice die) of E. coliWT. The mice were then treated with compounds
by intravenous administration at 20 mg/kg (n = 6
per group) 1 h and again at 5 h postchallenge (i.e., b.i.d., 40 mg/kg
total). Percent survival was determined at 96 h postinfection. All
compounds with the exception of 6p showed >50% protection.
The cyclohepta[1,2-f]indoles 6r and 6v and oxepino[3,2-f]indoles 6o and 6s provided 100% protection. For these concentration-dependent
bactericidal compounds, these results could be rationalized by the
free drug concentrations (e.g., free AUC) as a contributor to the
PK–PD. For example, the congeners 6p and 6q possess equal MICs against E. coliWT and exhibited nearly identical AUC values, but the 3-fold
lower free AUC (resulting from a 3-fold lower fu of 6p compared to 6q) is consistent
with reduced protection of the former compound. The control compound,
moxifloxacin, possesses a lower MIC of 0.024 μg/mL against E. coli compared to the 4-hydroxy-2-pyridone
compounds and showed 100% protection at a dose of 3 mg/kg b.i.d. and
83% protection at a dose of 1.5 mg/kg b.i.d. Mock treatment resulted
in all mice succumbing to infection within 24 h postinfection.
Conclusion
In summary, the synthesis of a novel series 5-indolyl-4-hydroxy-2-pyridones 2, derived from n class="Chemical">phenyl-substituted 4-hydroxy-2-pyridones has
been described. These dual targeting bacterial type-II DNA topoisomerase
inhibitors primarily target wild-type and resistant Gram-negative
bacteria. Further SAR optimization of these compounds was investigated
by adding linking groups to restrict rotational freedom between the
indole and 2-pyridone rings. The length and placement of the linking
groups was found to have a profound effect on antibacterial activity.
Evaluation of compounds containing tethers of one to four methylene
groups linking the 4-hydroxy-2-pyridone C-5 and the indole C-6 positions
revealed that optimal activity was achieved when the linking group
comprised three methylene units (seven-membered ring). In this case,
the calculated dihedral angle for the most energetically stable conformation
between the two rings (e.g., 6a) most closely approximated
that of the acyclic analog 2j (135 vs 129°, respectively).
Replacement of the methylene group in the tether closest to the indole
ring with heteroatoms led to reduced antibacterial activity in the
order of CH2 > O > S > N. In contrast, connecting
a linker
of three methylene groups between the 4-hydroxy-2-pyridone C-5 and
the indole C-4 positions (e.g., 8a) provides similar
torsion between the two rings as in 6a, but the significantly
altered topology resulted in a dramatic reduction in activity against E. coli.
The enhanced activity of the ring constrained
4-hydroxy-2-pyridones
(i.e., 6a) compared to n class="Chemical">ciprofloxacin against fluoroquinolone
resistant strains could be rationalized by the attenuated shift in
activity against E. coli DNA gyrase containing mutations
in the QRDR at positions S83L and D87Y. Susceptibility testing of
compounds 6o and 6v showed moderate to good
MIC90 values against a panel of clinically relevant resistant
phenotypes of E. coli (0.5 and 1 μg/mL, respectively)
and A. baumannii (8 and 16 μg/mL, respectively).
The sterically constrained 4-hydroxy-2-pyridones possess favorable
pharmacokinetic properties. Compounds 6o and 6v were characterized by moderate clearance and moderately high volume
of distribution in mice. In a murinesepticemia model, both compounds
demonstrated complete protection in mice infected with a lethal dose
of E. coli.
Experimental Section
Starting materials and other reagents were purchased from commercial
suppliers and were used without further purification unless otherwise
indicated. Air or moisture sensitive reactions were performed under
either a nitrogen or n class="Chemical">argon atmosphere. Flash chromatography was performed
using silica gel with standard techniques or with silica gel cartridges
on an ISCO Combiflash chromatography instrument. 1HNMR
spectra were recorded at 500 MHz on a Bruker NMR spectrometer. The
chemical shifts are given in parts per million referenced to the deuterated
solvent signal. Coupling constants (J) are recorded
in hertz. LC–MS analyses were performed on a Waters Acquity
UPLC–MS system with an analytical C18 column, and compounds
were detected by UV absorption at 254 nm. All final compounds with
reported biological data were determined to be >95% pure based
on
LC–MS and 1HNMR.
To a solution of 16c (200 mg, 0.41 mmol) in DCE (2.0 mL) was added pyrrolidine (0.07
mL, 0.85 mmol, 2.0 equiv) and HOAc (0.05 mL, 0.82 mmol, 2.0 equiv)
at room temperature. The reaction was stirred for 1 h before NaBH(OAc)3 (174 mg, 0.82 mmol, 2.0 equiv) was added. Upon completion
of the reaction, the volatiles were removed under reduced pressure
followed by the addition of water. The crude product was collected
via filtration and purified by preparative HPLC (40–90% MeCN
in H2O) to afford the reductive amination intermediate.
To a suspension of this intermediate in TIPS-H (1.5 mL) was added
TFA (1.5 mL), and the reaction mixture was heated to 65 °C for
1 h. Upon completion, the volatiles were removed under reduced pressure.
The residue was dissolved in CH2Cl2 (1.5 mL),
then 2 M HCl/Et2O (2.0 mL) was added. The precipitate was
collected by filtration and washed with Et2O (3 mL ×
3) and then dried under a stream of nitrogen to afford 2n as a light yellow solid. LC–MS m/z = 396.1 [M + H]+; 1HNMR (500 MHz,
methanol-d4) δ 1.04 (t, J = 7.25 Hz, 3 H), 1.81–1.86 (m, 4 H), 2.41 (q, J = 7.57 Hz, 2 H), 2.62–2.67 (m, 4 H), 3.84 (s, 3
H), 3.86 (s, 2 H), 6.48 (s, 1 H), 7.18 (d, J = 9.46
Hz, 1 H), 7.45 (d, J = 8.51 Hz, 1 H), 7.55 (s, 1
H).
Compound 6ee was obtained from 39 and 3-methoxypyrrolidine
to afford a white solid in 49% yield. LC–MS m/z = 440.1 [M + H]+; 1HNMR
(500 MHz, DMSO-d6) δ 2.64 (m, 2H),
3.28–3.41 (m, 7H), 3.48–3.65 (m, 2H), 3.81–3.94
(m, 3H), 4.39–4.54 (m, 3H), 4.60–4.77 (m, 2H), 6.95
(s, 1H), 7.44 (s, 1H), 7.87 (s, 1H), 13.03 (br. s, 1H), 13.88 (br.
s, 1H).
1-(1-Methyl-1H-indol-5-yl)butan-1-one (9a)
To a solution of 5-bromo-1-methyl-1H-indole (9.8 g, 46.7 mmol) in THF (40 mL) was added n-BuLi (2.5 M in hexanes, 22.3 mL, 55.8 mmol, 1.2 equiv) at −78
°C dropwise over 15 min. The mixture was stirred for 30 min at
−78 °C before a solution of N-methoxy-N-methylbutyramide (7.35 g, 56.0 mmol, 1.2 equiv) in THF
(10 mL) was added. After stirring at −78 °C for 10 min,
the reaction was quenched by saturatedaq. NH4Cl. The resulting
mixture was extracted by ether (30 mL × 3), and the combined
organic layers were dried over Na2SO4. The crude
product was purified by flash chromatography (0–10% EtOAc in
hexanes) to give 9a (5.40 g, 26.8 mmol, 57%) as a white
solid. LC–MS m/z = 202.2
[M + H]+; 1HNMR (500 MHz, CDCl3)
δ ppm 1.04 (t, J = 7.45 Hz, 3 H), 1.82 (sextet, J = 7.39 Hz, 2 H), 3.04 (t, J = 7.39 Hz,
2 H), 3.84 (s, 3 H), 6.62 (dd, J = 3.15, 0.87 Hz,
1 H), 7.09–7.16 (m, 1 H), 7.35 (d, J = 8.67
Hz, 1 H), 7.92 (dd, J = 8.71, 1.69 Hz, 1 H), 8.32
(dd, J = 1.69, 0.51 Hz, 1 H).
1-(1,2-Dimethyl-1H-indol-5-yl)butan-1-one (9b)
Compound 9b was prepared from 5-bromo-1,2-dimethyl-1H-indole following the procedure used to prepare 9a in
50% yield as an off-white solid. 1HNMR (500 MHz,
CDCl3) δ 1.03 (t, J = 7.41 Hz, 3
H), 1.81 (sextet, J = 7.44 Hz, 2 H), 2.45 (d, J = 0.95 Hz, 3 H), 2.97–3.08 (m, 2 H), 3.70 (s, 3
H), 6.36 (t, J = 0.95 Hz, 1 H), 7.22–7.31
(m, 1 H), 7.85 (dd, J = 8.67, 1.73 Hz, 1 H), 8.20
(d, J = 1.58 Hz, 1 H).
1-(1,2,3-Trimethyl-1H-indol-5-yl)butan-1-one
(9c)
Compound 9c was prepared from
5-bromo-1,2,3-trimethyl-1H-indole following the procedure
used to prepare 9a (75%) as an off-white solid. 1HNMR (500 MHz, CDCl3) δ 1.04 (t, J = 7.41 Hz, 3 H), 1.82 (sextet, J = 7.38
Hz, 2 H), 2.30 (s, 3 H), 2.37 (s, 3 H), 2.99–3.07 (m, 2 H),
3.68 (s, 3 H), 7.21–7.26 (m, 1 H), 7.84 (dd, J = 8.67, 1.73 Hz, 1 H), 8.18 (d, J = 1.58 Hz, 1
H).
To a solution of 9a (5.40
g, 26.8 mmol) in CH2Cl2 (30 mL) was added n class="Chemical">t-butylamine (11.3 mL, 107.1 mmol, 4.0 equiv). The mixture
was cooled to 0 °C before TiCl4 (1.0 M in CH2Cl2, 17.4 mL, 17.4 mmol, 0.65 equiv) was added via syringe
pump over 30 min. The reaction was warmed to room temperature and
stirred overnight. The reaction mixture was quenched by saturatedaq. NaHCO3 solution (20 mL) then extracted by CH2Cl2 (25 mL × 5) using a phase separator (75 mL).
The combined organic layers were dried over Na2SO4 and then concentrated to give 6.70 g of the imine intermediate (LC–MS m/z = 257.3 [M + H]+), which
was used directly in the next step without further purification. To
a suspension of the imine intermediate (6.70 g, 26.1 mmol) in Ph2O (40 mL) was added trimethylmethanetricarboxylate (8.44 g,
44.4 mmol, 1.7 equiv). The reaction was heated with stirring at 230
°C for 10 min, removing the methanol with a distillation apparatus
attached to the flask. The mixture was cooled to room temperature,
and the precipitate was filtered and then washed with diethyl ether
to afford the 10a as a yellow solid (4.85 g, 14.9 mmol,
56% over 2-step). LC–MS m/z = 327.2 [M + H]+; 1HNMR (500 MHz, CDCl3) δ 1.14 (t, J = 7.33 Hz, 3 H), 2.49
(q, J = 7.33 Hz, 2 H), 3.87 (s, 3 H), 4.02 (s, 3
H), 6.58 (dd, J = 3.11, 0.83 Hz, 1 H), 7.18 (d, J = 3.07 Hz, 1 H), 7.24 (dd, J = 8.47,
1.69 Hz, 1 H), 7.44 (d, J = 8.51 Hz, 1 H), 7.68 (dd, J = 1.69, 0.59 Hz, 1 H), 13.88 (s, 1 H).
Compound 10b was prepared
from 9b following the procedure used to prepare 10a in 27% yield as an off-white solid. LC–MS m/z = 339.2 [M – H]−, 341.1 [M + H]+.
To a solution of ethyl
5-bromo-1-methyl-1H-indole-2-carboxylate (9.0 g,
31.9 mmol) in CH2Cl2 (80 mL) was added DIBAL-H
(1.0 M in hexanes, 70.0 mL, 2.2 equiv) at −78 °C over
15 min. After stirring for 1 h at −78 °C, the reaction
was quenched by the addition of 1 NHCl (20 mL) at −78 °C
and then warmed to room temperature and stirred for an additional
30 min to break up the aluminum emulsion. The biphasic mixture was
extracted with ether/EtOAc (1:1, 50 mL × 3). The combined organic
layers were dried over Na2SO4 and then concentrated
to give (5-bromo-1-methyl-1H-indol-2-yl)methanol
(ca. 7.8 g, quant).To a solution of (5-bromo-1-methyl-1H-indol-2-yl)methanol (7.8 g, ca. 31.3 mmol) in CH2Cl2 (80 mL) was added imidazole (2.8 g, 40.7 mmol, 1.3
equiv) followed by TBSCl (6.1 g, 40.7 mmol, 1.3 equiv) at 0 °C.
After 1 h, the reaction was quenched with water and then extracted
with CH2Cl2 (50 mL × 3). The solvent was
concentrated to give crude product, which was purified by flash column
chromatography (50% CH2Cl2 in hexanes) to afford 11 (10.7 g, 96% yield over two steps) as an off-white solid.
LC–MS m/z = 354.0, 356.0
[M + H]+; 1HNMR (500 MHz, CDCl3)
δ 0.07 (s, 6 H), 0.90 (s, 9 H), 3.77 (s, 3 H), 4.79–4.85
(m, 2 H), 6.30–6.35 (m, 1 H), 7.17 (d, J =
8.67 Hz, 1 H), 7.27–7.31 (m, 1 H), 7.69 (dd, J = 1.89, 0.47 Hz, 1 H).
Compound 12a was prepared from 11 and N-methoxy-N-methylacetamide in 92% yield according
to the procedure used to prepare 12c. LC–MS m/z = 318.3 [M + H]+.
To
a solution of 11 (16.6 g, 46.7 mmol) in THF (60 mL) was
added n-BuLi (2.5 M in hexanes, 22.3 mL, 55.8 mmol)
at −78 °C dropwise over 15 min. The mixture was stirred
for 30 min at −78 °C before a solution of N-methoxy-N-methylbutyramide (7.35 g, 56.0 mmol)
in THF (10 mL) was added. After stirring at −78 °C for
10 min, the reaction was quenched with saturatedaq. NH4Cl. The resulting mixture was extracted with ether (30 mL ×
3), and the combined organic layers were dried over Na2SO4. The crude product was purified by flash chromatography
(0–10% EtOAc in hexanes) to give 12c (9.3 g, 57%)
as a white solid. 1HNMR (500 MHz, CDCl3) δ
0.08 (s, 6 H), 0.91 (s, 9 H), 1.03 (t, J = 7.41 Hz,
3 H), 1.74–1.88 (m, 2 H), 2.97–3.07 (m, 2 H), 3.82 (s,
3 H), 4.85 (s, 2 H), 6.46–6.52 (m, 1 H), 7.32 (d, J = 8.67 Hz, 1 H), 7.90 (dd, J = 8.67, 1.73 Hz, 1
H), 8.26 (d, J = 1.18 Hz, 1 H).
Compound 12d was prepared from 11 and N-methoxy-N-methylisobutyramide in 85%
yield according to the procedure used to prepare 12c.
LC–MS m/z = 360.2 [M + H]+.
Compound 12e was prepared from 11 and N-methoxy-N-methylcyclopropanecarboxamide
in 72% yield according to the procedure used to prepare 12c. LC–MS m/z = 358.3 [M +
H]+.
To a stirred solution of 12c (4.14 g, 12.0 mmol) in n class="Chemical">CH2Cl2 (12 mL) was
added (2,4-dimethoxybenzylamine (1.98 mL, 13.2 mmol) and Et3N (4.5 mL, 32.4 mmol) sequentially at 0 °C. Then TiCl4 (7.8 mL, 1.0 M in CH2Cl2, 7.8 mmol, 0.65 equiv)
was added to the reaction mixture via syringe pump over 30 min. The
reaction was warmed to room temperature and stirred overnight. The
mixture was quenched by satd. NaHCO3 solution and extracted
with CH2Cl2 (30 mL × 5). The combined organic
layers were dried over Na2SO4 and then concentrated
under reduced pressure to give the crude product (5.94 g), which was
carried into the next step without further purification. The crude
imine (5.94 g, ca. 12.0 mmol) was dissolved in Ph2O (20
mL) and trimethylmethanetricarboxylate (3.88 g, 20.4 mmol) was added.
A distillation apparatus was attached to the flask containing the
reaction mixture. The reaction was heated to 230 °C for 10 min.
The heating was removed after the distillation of methanol ceased.
The mixture was cooled to room temperature and then purified by flash
column chromatography (0–50% EtOAc in CH2Cl2) to give 13c (4.17 g, 6.72 mmol) in 56% yield.
LC–MS m/z = 621.3 [M + H]+; 1HNMR (500 MHz, CDCl3) δ 0.08–0.14
(m, 6 H), 0.85–0.94 (m, 12 H), 2.13 (m, 2 H), 3.16 (s, 3 H),
3.76 (s, 3 H), 3.78–3.81 (m, 3 H), 3.99–4.03 (m, 3 H),
4.78–4.87 (m, 4 H), 6.13 (d, J = 2.29 Hz,
1 H), 6.30 (s, 1 H), 6.39 (dd, J = 8.39, 2.40 Hz,
1 H), 6.79 (d, J = 7.49 Hz, 1 H), 6.84 (d, J = 8.43 Hz, 1 H), 7.06 (s, 1 H), 7.22 (d, J = 8.51 Hz, 1 H).
To a suspension of 14c (3.13
g, 6.18 mmol) in n class="Chemical">EtOAc (15 mL) was added LiI (2.48 g, 18.5 mmol) at
room temperature. The mixture was heated to 65 °C and then stirred
for 1 h. The reaction mixture was diluted with EtOAc (30 mL) and then
quenched by satd. Na2S2O3 (30 mL).
The organic phase was separated and the aqueous layer was extracted
with EtOAc (30 mL × 4). The combined organic layers were dried
over Na2SO4 and then concentrated to give 15c (2.89 g, 5.87 mmol) in 95% yield, which was carried on
to the next step without further purification. LC–MS m/z = 493.2 [M + H]+; 1HNMR (500 MHz, CDCl3) δ 0.88–0.99 (m, 3
H), 2.11–2.27 (m, 2 H), 3.29 (s, 3 H), 3.78 (s, 3 H), 3.85
(s, 3 H), 4.81–4.87 (m, 2 H), 4.90 (d, J =
15.76 Hz, 1 H), 4.99 (d, J = 15.84 Hz, 1 H), 6.21
(d, J = 2.36 Hz, 1 H), 6.35–6.46 (m, 2 H),
6.69 (d, J = 8.43 Hz, 1 H), 6.85 (dd, J = 8.43, 1.66 Hz, 1 H), 7.16 (d, J = 1.26 Hz, 1
H), 7.29–7.33 (m, 1 H), 13.97 (s, 1 H), 15.96 (s, 1 H).
To a solution of 4-amino-2-chloro-5-iodobenzonitrile[18] (51.73 g, 186 mmol) in n class="Chemical">CH3CN (270
mL) was added propargyl alcohol (13.3 mL, 223 mmol) and NEt3 (52 mL, 373 mmol). The mixture was degassed with argon before Pd(PPh3)2Cl2 (1.30 g, 1.85 mmol) and CuI (0.70
g, 3.67 mmol) were added. The reaction was heated at 70 °C for
2 h until starting material was completely consumed. After cooling
to room temperature, the solvent was concentrated. Water (300 mL)
was added, and the precipitate was filtered, washed with H2O (200 mL × 2), and dried with a N2 flow. Crude alkyne
intermediate was obtained as tan solid (40.0 g) and was used directly
in the next step. To a solution of the intermediate (36.70 g, 178
mmol) in DMF (450 mL) was added t-BuOK (44.0 g, 392
mmol), and the mixture was heated at 70 °C for 2 h under an argon
atmosphere. Upon cooling to room temperature, the mixture was carefully
poured into a mixture of ice (∼800 mL) and conc. HCl (50 mL).
The precipitate was filtered and washed with H2O (300 mL
× 2) and dried affording 17 as a brownish solid
(28.50 g, 77% over 2 steps). LC–MS m/z = 207.1, 209.1 [M + H]+; 1HNMR
(500 MHz, DMSO-d6) δ 4.63 (d, J = 5.7 Hz, 2 H), 5.47 (t, J = 5.7 Hz,
1 H), 6.44 (s, 1 H), 7.58 (s, 1 H), 8.12 (s, 1 H), 11.79 (br. s, 1
H).
A solution
of 17 (32.66 g, 158 mmol) and imidazole (14.0 g, 205
mmol) in n class="Chemical">DMF (450 mL) was stirred at room temperature for 5 min before
TBSCl (29.0 g 192 mmol) was added in one portion. The reaction was
stirred at room temperature for 1.5 h and then poured into ice H2O (total final volume ≈ 1700 mL). A dark brown oil
was formed, which solidified upon the addition of pentane (∼100
mL). The resulting solid was filtered, washed with H2O
(300 mL × 2), and dried overnight. The solid was washed with
pentane (300 mL × 2) and then suspended in 800 mL of CH2Cl2. The resulting mixture was stirred vigorously at room
temperature for 1 h and then filtered through Celite. The mother liquor
was concentrated to yield 18 as a brownish solid (37.42
g, 74%), which was used in the next step without further purification.
LC–MS m/z = 321.2, 323.2
[M + H]+; 1HNMR (500 MHz, CDCl3)
δ 0.13 (s, 6 H), 0.94 (s, 9 H), 4.88 (s, 2 H), 6.36 (s, 1 H),
7.49 (s, 1 H), 7.89 (s, 1 H), 8.63 (br. s, 1 H).
To a
solution of crude 18 (37.50 g, 117 mmol) in DMF (400
mL) at 0 °C was added n class="Chemical">NaH (60%, 6.7 g, 168 mmol) in portions.
Upon completion of the addition, the mixture was warmed to room temperature
with stirring for 10 min. The reaction mixture was cooled to 0 °C,
and MeI (10.5 mL, 169 mmol) was added. The reaction mixture was warmed
to room temperature and stirred for 1.5 h and then poured into ice
H2O and 100 mL of 1 M HCl (final volume ∼1600 mL).
The precipitate was filtered, washed with H2O (200 mL ×
3), and dried overnight. The solid was then washed with pentane (200
mL × 2). The crude product, obtained as a brownish solid (39.0
g), was purified by column chromatography (CH2Cl2/hexanes, 50–100%) to yield 19 as a pale orange
solid (30.0 g, 76% yield). LC–MS m/z = 335.2, 337.2 [M + H]+; 1HNMR
(500 MHz, CDCl3) δ 0.08 (s, 6 H), 0.90 (s, 9 H),
3.79 (s, 3 H), 4.82 (s, 2 H), 6.44 (s, 1 H), 7.41 (s, 1 H), 7.90 (s,
1 H).
To a
solution of 19 (4.15 g, 12.4 mmol) in CH2Cl2 (50 mL) at −78 °C was added n class="Chemical">DIBAL-H (1 M in CH2Cl2, 15.0 mL, 15.0 mmol) dropwise over 10 min.
The reaction was stirred at this temperature for 10 min and slowly
warmed to −15 °C over ∼2 h. The reaction mixture
was then cooled to −40 °C and quenched upon the addition
of Rochelle salt solution (aq. satd., 20 mL). The resulting emulsion
was warmed to room temperature and vigorously stirred for ∼1
h. The organic phase was separated, and the aqueous layer was extracted
with CH2Cl2 (50 mL). The combined organic layers
were washed sequentially with 1 M HCl (50 mL), NaHCO3 (aq.
satd., 50 mL), and brine (50 mL) and then dried over Na2SO4. After solvent removal in vacuo, the crude product
was purified by column chromatography (EtOAc/hexanes, 5–15%)
affording 20 as an off-white solid (3.80 g, 91%). LC–MS m/z = 338.2, 340.3 [M + H]+; 1HNMR (500 MHz, CDCl3) δ 0.08 (s, 6 H), 0.90
(s, 9 H), 3.79 (s, 3 H), 4.82 (s, 2 H), 6.49 (s, 1 H), 7.33 (s, 1
H), 8.21 (s, 1 H), 10.50 (s, 1 H).
Compound 20 (3.80 g, 11.24 mmol), potassium vinyltrifluoroborate (2.30
g, 17.17 mmol), Pd(OAc)2 (76 mg, 0.34 mmol, 0.03 equiv),
S-Phos ligand (280 mg, 0.68 mmol, 0.06 equiv), and K2CO3 (4.70 g, 34.0 mmol) were mixed together in a 100 mL round-bottom
flask. The flask was purged and backfilled with argon before dioxane
(45 mL) and H2O (7.5 mL) were added. The reaction was heated
at 85–90 °C for 5 h and then cooled to room temperature.
Water (30 mL) was added, and the product was extracted with CH2Cl2 (80 mL × 3). The combined organic phases
were washed with brine (80 mL) and dried over Na2SO4. After removal of the solvents, purification by column chromatography
(EtOAc/hexanes, 0–10% gradient) afforded 21 as
a white solid (3.32 g, 89%). LC–MS m/z = 330.3 [M + H]+; 1HNMR (500 MHz,
CDCl3) δ 0.09 (s, 6 H), 0.91 (s, 9 H), 3.84 (s, 3
H), 4.85 (s, 2 H), 5.42 (dd, J = 10.7, 1.7 Hz, 1
H), 5.69 (dd, J = 17.2, 1.7 Hz, 1 H), 6.51 (s, 1
H), 7.43 (s, 1 H), 7.75 (dd, J = 17.3, 10.7 Hz, 1
H), 8.07 (s, 1 H), 10.24 (s, 1 H).
To a solution
of 21 (8.23 g, 24.98 mmol) in THF (50
mL) at −78 °C was added 3-butenylmagnesium bromide (0.5
M in THF, 60.0 mL, 30.0 mmol) dropwise over ∼10 min. The reaction
was stirred at this temperature for 10 min and slowly allowed to warm
to −10 °C. The reaction was quenched by the addition of
NH4Cl (aq. satd., 80 mL) and was extracted with EtOAc (150
mL × 4). The combined organic layers were washed with brine (100
mL) and dried over Na2SO4. Upon removal of the
solvent, the product was obtained as a pale-yellow oil (9.60 g, quant),
which solidified under high vacuum. Compound 22c (>95%
purity) was used directly in the next step without purification. LC–MS m/z = 386.3 [M + H]+; 1HNMR (500 MHz, CDCl3) δ 0.06 (s, 3 H), 0.07 (s,
3 H), 0.90 (s, 9 H), 1.86–1.99 (m, 2 H), 2.11–2.31 (m,
2 H), 3.80 (s, 3 H), 4.82 (s, 2 H), 4.99 (d, J =
10.4 Hz, 1 H), 5.06 (dd, J = 17.2, 1.7 Hz, 1 H),
5.11 (dd, J = 7.3, 5.7 Hz, 1 H), 5.29 (dd, J = 10.4, 1.7 Hz, 1 H), 5.65 (dd, J = 17.2,
1.7 Hz, 1 H), 5.83–5.93 (m, 1 H), 6.34 (s, 1 H), 7.23 (dd, J = 17.2, 10.9 Hz, 1 H), 7.40 (s, 1 H), 7.66 (s, 1 H).
A solution of 23c (24.98 mmol) in EtOAc (105 mL) and
n class="Chemical">CH2Cl2 (15 mL) was hydrogenated over Pd/C (10%,
880 mg) under a H2-filled balloon (1 atm) for ∼3
h. The catalyst was filtered and washed with EtOAc. The filtrate was
concentrated, and the intermediate obtained as a brown solid was taken
directly into the next step. To activated 4 Å molecular sieves
(6.2 g, 250 mg/mmol) was added a solution of the intermediate obtained
above (24.98 mmol) in CH2Cl2 (125 mL). The mixture
was cooled to 0 °C before NMO (4.45 g, 37.99 mmol) and TPAP (445
mg, 1.26 mmol, 0.05 equiv) were added sequentially. The reaction was
stirred at 0 °C. Upon complete consumption of starting material
(∼1.5 h), the molecular sieves were filtered off and washed
with CH2Cl2. The filtrate was concentrated,
and the residue was purified by column chromatography (EtOAc/hexanes,
0–25% gradient) to provide 24c as an off-white
solid (7.47 g, 84% over 4 steps). LC–MS m/z = 358.3 [M + H]+; 1HNMR (500 MHz,
CDCl3) δ 0.06 (s, 6 H), 0.90 (s, 9 H), 1.76–1.84
(m, 2 H), 1.88–1.95 (m, 2 H), 2.74–2.77 (m, 2 H), 3.05
(t, J = 6.6 Hz, 2 H), 3.79 (s, 3 H), 4.82 (s, 2 H),
6.43 (s, 1 H), 7.06 (s, 1 H), 8.04 (s, 1 H).
To a solution of 3-bromo-4-hydroxybenzaldehyde
(10.0 g, 50 mmol) in DMF (50 mL) were added K2CO3 (7.6 g, 55 mmol) and allyl bromide (4.5 mL, 52.5 mmol). The reaction
mixture was stirred at room temperature for 16 h. The reaction mixture
was then poured into H2O (100 mL) and extracted with Et2O (100 mL × 2). The combined organic extracts were washed
with brine, dried over MgSO4, filtered, and concentrated
to afford 4-(allyloxy)-3-bromobenzaldehyde (11.2 g, 93%) as a clear
oil (LC–MS m/z = 243.1 [M
+ H]+), which was used directly in the next step. To a
solution of 4-(allyloxy)-3-bromobenzaldehyde (11.2 g, 46.5 mmol) and
ethyl 2-azidoacetate (19.0 g, 140 mmol) in EtOH (100 mL), cooled to
−10 °C, was added NaOEt (50 mL, 2.76 M) dropwise over
20 min. The reaction mixture was then warmed to 5 °C and stirred
for 16 h at which point the reaction mixture was cooled to 0 °C
and H2O was added. The precipitate was then filtered and
washed with H2O to provide 28 (10.1 g, 62%)
as a beige powder. 1HNMR (500 MHz, DMSO-d6) δ 1.29–1.35 (m, 3 H), 4.31 (q, J = 7.09 Hz, 2 H), 4.62 (s, 2 H), 5.32 (dq, J = 10.60, 1.62 Hz, 1 H), 5.52 (dq, J = 17.26, 1.79
Hz, 1 H), 6.10 (ddt, J = 17.25, 10.61, 4.83, 4.83
Hz, 1 H), 7.15 (d, J = 8.75 Hz, 1 H), 7.78–7.90
(m, 2 H), 8.15–8.22 (m, 1 H).
To a solution of 29 (5.5 g,
16.9 mmol) in DMF (17 mL), cooled to 0 °C, was added NaH (60%
oil dispersion, 0.75g, 18.6 mmol). Gas evolution was observed, and
the mixture was stirred for 30 min at which point MeI (1.2 mL, 18.6
mmol) was added, and the solution was warmed to room temperature.
After stirring for 1 h, saturated NH4Cl was added, and
the mixture was poured into H2O and extracted with Et2O. The organic extracts were washed with brine, dried over
MgSO4, filtered, and concentrated to give 30 (5.24 g, 92%) as a white solid. LC–MS m/z = 340.1 [M + H]+; 1HNMR (500 MHz,
CDCl3) δ 1.31 (t, J = 7.13 Hz, 3
H), 3.86–3.91 (m, 3 H), 4.26 (q, J = 7.17
Hz, 2 H), 4.55 (dt, J = 4.95, 1.59 Hz, 2 H), 5.25
(dq, J = 10.63, 1.47 Hz, 1 H), 5.46 (dq, J = 17.26, 1.66 Hz, 1 H), 5.97–6.08 (m, 1 H), 6.63
(s, 1 H), 7.05 (d, J = 0.79 Hz, 1 H), 7.71 (s, 1
H).
To a solution of 30 (5.24
g, 15.4 mmol) in CH2Cl2 (40 mL), cooled to −78
°C, was added DIBAL-H (32.4 mL, 1 M). After stirring at −78
°C for 30 min, the reaction was warmed to 0 °C, and a saturated
solution of Rochelle salt (30 mL) was added followed by CH2Cl2 (100 mL). The solution was then warmed to room temperature
and stirred for 1 h. The organic layer was separated and washed with
brine, dried over Na2SO4, filtered, and concentrated
to afford 31 (4.0 g, 87%) as a white solid. LC–MS m/z = 298.0 [M + H]+; 1HNMR (500 MHz, CDCl3) δ 3.66 (s, 3 H), 4.60 (dt, J = 5.04, 1.62 Hz, 2 H), 4.73 (s, 2 H), 5.27 (dd, J = 10.56, 1.50 Hz, 1 H), 5.49 (dd, J =
17.22, 1.62 Hz, 1 H), 6.04–6.14 (m, 1 H), 6.20 (d, J = 0.63 Hz, 1 H), 6.74 (s, 1 H), 7.66 (s, 1 H).
To a
solution of 31 (4.8 g, 16 mmol) in CH2Cl2 (30 mL) were added n class="Chemical">imidazole (1.2 g, 17 mmol) and TBSCl (2.6
g, 17 mmol). After stirring for 3 h, the solution was washed with
H2O, brine, dried with Na2SO4, filtered,
and concentrated. The crude residue was purified on silica gel (20%
EtOAc/hexanes) to afford the TBS protected alcohol (5.58 g, 85%) as
a white solid (LC–MS m/z:
410.2 [M + H]+). To a solution of this intermediate (0.2
g, 0.5 mmol) in THF (5 mL), cooled to −78 °C was added n-BuLi (300 μL, 2.5 M solution). After stirring at
−78 °C for 30 min, DMF (100 μL) was added, and the
solution was warmed to room temperature. After stirring for 30 min,
the reaction was quenched with saturated NH4Cl (2 mL) and
poured into H2O. The aqueous layer was extracted with Et2O (20 mL × 2), and the combined organic phases were washed
with brine, dried with MgSO4, filtered, and concentrated
to afford 32 (165 mg, 92%) as a white solid. LC–MS m/z = 360.2 [M + H]+; 1HNMR (500 MHz, CDCl3) δ −0.03–0.02
(m, 6 H), 0.80–0.85 (m, 9 H), 3.67 (s, 3 H), 4.63 (dt, J = 5.10, 1.55 Hz, 2 H), 4.71 (d, J = 0.32
Hz, 2 H), 5.28 (dd, J = 10.56, 1.42 Hz, 1 H), 5.40–5.47
(m, 1 H), 6.00–6.15 (m, 1 H), 6.33 (d, J =
0.55 Hz, 1 H), 6.64 (s, 1 H), 8.03 (s, 1 H), 10.45 (s, 1 H).
To
a solution of compound 32 (5.2 g, 14.6 mmol) in THF (60
mL) cooled to 0 °C was added n class="Chemical">vinylmagnesium bromide (16.0 mL,
1 M). After stirring at 0 °C for 30 min, the solution was warmed
to room temperature and stirred for an additional 30 min at which
point saturated NH4Cl (10 mL) was added. The crude reaction
mixture was poured into H2O and extracted with Et2O (100 mL × 2). The combined organic phases were washed with
brine, dried over MgSO4, filtered, and concentrated to
afford 33 (5.5 g, 98%) as a yellow oil, which was used
immediately in the subsequent step without further purification. 1HNMR (500 MHz, DMSO-d6) δ
−0.02–0.03 (m, 6 H), 0.80–0.86 (m, 9 H), 3.63–3.68
(m, 3 H), 4.59 (dt, J = 4.93, 1.60 Hz, 2 H), 4.76
(s, 2 H), 4.88–4.94 (m, 1 H), 5.12 (d, J =
4.81 Hz, 1 H), 5.13–5.19 (m, 1 H), 5.22–5.27 (m, 1 H),
5.40–5.48 (m, 1 H), 5.91–6.01 (m, 1 H), 6.04–6.14
(m, 1 H), 6.26 (s, 1 H), 6.94 (s, 1 H), 7.43 (s, 1 H).
To a solution of compound 33 (5.5 g, 14.3
mmol) in toluene (200 mL) was added Grubbs second generation catalyst
(350 mg), and the reaction mixture was heated at 60 °C for 3
h. The reaction mixture was then cooled to room temperature and concentrated,
and the crude residue was purified on n class="Chemical">silica gel (20% EtOAc/hexanes)
to provide 34 as a light green solid (2.7 g, 51%). LC–MS m/z = 342.3 [M – H2O]−; 1HNMR (500 MHz, DMSO-d6) δ −0.03–0.03 (m, 6 H), 0.79–0.86
(m, 9 H), 3.64 (s, 3 H), 4.17–4.28 (m, 1 H), 4.65–4.74
(m, 1 H), 4.76 (s, 2 H), 5.23–5.34 (m, 1 H), 5.47 (d, J = 5.60 Hz, 1 H), 5.69–5.80 (m, 2 H), 6.30 (d, J = 0.55 Hz, 1 H), 7.08 (s, 1 H), 7.41 (d, J = 0.63 Hz, 1 H).
To a solution
of 38 (1.25 g, 2.5 mmol) in CH2Cl2 (25 mL) was added MnO2 in three batches over 2 h (1.5,
1.0, and 1.0 g). The reaction mixture was then filtered through Celite
and concentrated to afford the product as a brown foam (0.81 g, 64%),
which was used without further purification.
Authors: Jason A Wiles; Akihiro Hashimoto; Jane A Thanassi; Jijun Cheng; Christopher D Incarvito; Milind Deshpande; Michael J Pucci; Barton J Bradbury Journal: J Med Chem Date: 2006-01-12 Impact factor: 7.446
Authors: Edmund L Ellsworth; Tuan P Tran; H D Hollis Showalter; Joseph P Sanchez; Brian M Watson; Michael A Stier; John M Domagala; Stephen J Gracheck; E Themis Joannides; Martin A Shapiro; Steve A Dunham; Debra L Hanna; Michael D Huband; Jeffrey W Gage; Joel C Bronstein; Jia Yeu Liu; Dai Q Nguyen; Rajeshwar Singh Journal: J Med Chem Date: 2006-11-02 Impact factor: 7.446
Authors: Q Li; D T Chu; A Claiborne; C S Cooper; C M Lee; K Raye; K B Berst; P Donner; W Wang; L Hasvold; A Fung; Z Ma; M Tufano; R Flamm; L L Shen; J Baranowski; A Nilius; J Alder; J Meulbroek; K Marsh; D Crowell; Y Hui; L Seif; L M Melcher; J J Plattner Journal: J Med Chem Date: 1996-08-02 Impact factor: 7.446
Figure 1
Scheme 1
Scheme 2
Scheme 3
Scheme 4
Figure 2
Figure 3
Figure 4
Figure 5