Bacterial translation inhibitors, 1-acylindazol-3-ols as anthranilic acid mimics

Article abstract of DOI:10.1016/j.bmcl.2007.08.041

The discovery and initial optimization of a novel anthranilic acid derived class of antibacterial agents has been described in a recent series of papers. This paper describes the discovery of 1-acylindazol-3-ols as a novel bioisostere of an anthranilic acid. The synthesis and structure-activity relationships of the indazol bioisosteres are described herein.

Full text of DOI:10.1016/j.bmcl.2007.08.041

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 6293–6297
Bacterial translation inhibitors, 1-acylindazol-3-ols
as anthranilic acid mimics
Cory Stiff,a David R. Graber,a Atli Thorarensen,a Brian D. Wakefield,a Keith R. Marotti,c
Earline P. Melchior,c Michael T. Sweeney,c Fusen Han,b Douglas C. Rohrer,b
Gary E. Zurenkoc and Donna L. Romeroa,*
aMedicinal Chemistry, 301 Henrietta Street, Kalamazoo, MI 49001, USA
bStructural Analytical and Medicinal Chemistry, 301 Henrietta Street, Kalamazoo, MI 49001, USA
cInfectious Diseases Biology, Pharmacia Corporation, 301 Henrietta Street, Kalamazoo, MI 49001, USA
Received 14 May 2007; revised 17 August 2007; accepted 21 August 2007
Available online 25 August 2007
Abstract—The discovery and initial optimization of a novel anthranilic acid derived class of antibacterial agents has been described
in a recent series of papers. This paper describes the discovery of 1-acylindazol-3-ols as a novel bioisostere of an anthranilic acid.
The synthesis and structure–activity relationships of the indazol bioisosteres are described herein.
Ó 2008 Published by Elsevier Ltd.
We have previously reported the discovery of a novel
series of antibacterial agents stemming from HTS hit
compound 1.1 Such compounds were identified via a
transcription/translation assay (T/T assay). In-depth
structure–activity relationship studies were conducted
using a combination of parallel and traditional synthesis
leading to a number of highly active compounds, such as
compound 2. The acidic proton of the anthranilic acid
portion of this class of compounds was shown to be crit-
ical for activity.2 Analogs with the acid replaced with
alternate groups such as sulfonamide, acetamide or
hydrogen exhibited no activity versus Staphylococcus
aureus (MICs >128 lg/mL).
retain activity similar to the corresponding parent car-
boxylic acid. However, protein binding of these bioisos-
teres was not significantly impacted.2
Since our interest in this series of novel translation
inhibitors3 was very high, we continued to try to develop
replacements for the acid that improved potency and
reduced the affinity for human serum albumin (HSA).4,5
Capillary electrophoresis measurement of the pKa of this
type of anthranilic acid indicates the presence of an
intramolecular hydrogen bond, which pre-organizes
the acid into a favorable bioactive conformation
(Fig. 2, I).6 It appeared to us that a 1-acylindazol-3-
Advanced compounds in this series failed to show activ-
ity in a standard mouse bacteremia model of infection,
likely due to their high protein binding. Since human
serum albumin is known to bind aromatic carboxylic
acids such as salicylates and ibuprofen, we undertook
a study to attempt to replace the carboxylic acid with
bioisosteres to reduce the extent of protein binding. Tra-
ditional five-membered ring bioisosteres, such as tetra-
zole 3 (Fig. 1), were synthesized and were shown to
N
N
Br
CO2H
CO2H
HN
N
NC
NH
O
NH
O
NH
O
O2S
N
O2S
N
2
1
O2S
N
3
H3C
Keywords: Bioisosteres; Translation inhibitors; Antibacterials; Indazol;
Protein binding.
Cl
Cl
Figure 1. Initial SAR of HTS anthranilic acid screening lead 1.
*
Corresponding author. Tel.: +1 636 536 6487; e-mail: donna.l.romero
@sbcglobal.net
0960-894X/$ - see front matter Ó 2008 Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2007.08.041

6294
C. Stiff et al. / Bioorg. Med. Chem. Lett. 18 (2008) 6293–6297
pKa ~ 4 - 5
a
1. t-BuLi, -78oC
2. CO2, -78o to rt
OH
N
OH
R
COOH
NH2
R
X
O
NH
N
H
3. THF/3M HCl
N
R = CF3, SO2Me, OCF3
R
boc
60-70%
7
O
O
R
8
I
II
b
Figure 2. Comparison of indazole to anthranilic acid.
Br
1. Chloral hydrate
hydroxylamine
O
O
ols, II could closely approximate the desired conforma-
tion of the anthranilic acid, while still providing an
acidic proton (pKa ꢀ 4–5). Therefore, we hypothesized
that conformational restraint in the form of an indazol
would lead to a novel anthranilic acid bioisostere that
may have improved properties. This report describes
the synthesis and SAR of a series of novel indazol-3-ol
translation inhibitor analogs.
+
O
O
Br
NH2
2. H2SO4
Br
N
N
H
H
68%
9
10
11
1:1
COOH
KOH/H2O2
84%
Br
NH2
12
Scheme 2. Synthesis of anthranilic acids.
1H-Indazol-3-ols (indazol, henceforth) are a well-known
class of compounds and numerous syntheses have been
reported.7 The desired 5-substituted indazols are readily
available via a single pot procedure starting from the
corresponding anthranilic acid (Scheme 1). Diazotiza-
tion of the aniline with sodium nitrite in hydrochloric
acid is followed by reduction to the hydrazine. After
heating the hydrazine with additional acid, the indazol
is simply filtered away from the solution.
OH
N
NC
O
Cl
OH
N
NC
N
pyridine
+
N
O
SO2
N
reflux, 3h
87%
H
O2S
N
14
13
15
Scheme 3. First generation synthesis of 1-acylindazol-3-ols.
In a few instances the anthranilic acid was not commer-
cially available and was synthesized from the aniline
using one of two methods. The majority of the anthra-
nilic acids were obtained through an ortho-directed lith-
iation and subsequent quenching with carbon dioxide
(Scheme 2, route a). The 4-Br analog was synthesized
via the isatin, given its incompatibility with the metalla-
tion conditions (route b).
the indazol was performed. Fortunately, treatment of
the acid (16) with 1,10-carbonyldiimidazole (CDI), fol-
lowed by the desired indazol, provided the coupled
products 17 in good yields and typically as the sole prod-
ucts (Scheme 4).9 Thirty analogs varying both the ind-
azol ring substitution (9 indazols) and the aryl ring
substitution (13 bottom pieces) were synthesized.6 Se-
lected compounds and their T/T inhibition and antibac-
terial activity are shown in Table 1.
The parent indazol compounds are potentially reactive
at both of the ring nitrogens as well as the hydroxyl
group, so regioselectivity was an issue when developing
amide formation conditions. Conditions reported to
limit reaction to the 1-position,8 treatment of the cyano
indazol 14 with an acid chloride in refluxing pyridine
(Scheme 3), provided the desired analog in good yield.
However, purification of the final product was trouble-
some, and when these conditions were applied to other
analogs, multiple products formed (TLC).
Several of the analogs synthesized possessed good anti-
bacterial potency such as compounds 19 and 47. Com-
pound 19, in particular, exhibited potency that closely
matched that of its parent anthranilic acid congener.10
Many of the indazols synthesized exhibited modest anti-
bacterial activity (e.g., 20–30, 44–45). Testing of antibac-
terial activity in the presence of serum indicated that
these analogs were still highly protein bound. Even
though the indazol core can successfully replace the
anthranilic acid, it does not appear to offer any advan-
tage in terms of reducing protein binding.
In order to resolve the problem of multiple products and
to eliminate the need to form the acid chloride, a survey
of conditions that would couple a carboxylic acid with
1. CDI, CH2Cl2
O
OH
N
R
OH
R
COOH
1. NaNO2/HCl
R
OH
N
R
OH
N
aryl
R
COOH
aq. HCl 80oC
NH2
N
NH2
O
aryl
2. Na2SO3
N
H
N
N
75-80%
H
16
17
H
6
CH2Cl2, RT
5
4
Scheme 1. Synthesis of indazol-3-ol fragment.
Scheme 4. Second-generation synthesis of 1-acylindazol-3-ols.

C. Stiff et al. / Bioorg. Med. Chem. Lett. 18 (2008) 6293–6297
Table 1. Activity of compounds synthesized according to Scheme 4
6295
Aryl-
OH
N
R
5
6
SO2
N
SO2
N
SO2
N
SO2
N
SO2
N
Br
SO2
N
N
H3C
H3C
HN
O
aryl
O
Cl
Cl
Cl
C
A
B
E
D
Cl
F
N
N
N
N
N
H3C
PhO2S
HN
N
N
O
N
S
S
S
Ph
PhO2S
Ph
Ph
G
L
H
M
J
N
O
K
Compound
R
Aryl
% T/T Inhibition
at 100 lMa
SAUR MIC
(lg/mL)b
SAUR
MIC + 10%
serum (lg/mL)c
Parent anthranilic acid
MIC (lg/mL)
(0% and 10% serum)
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
5-CN
5-CN
5-Br
5-Br
5-Br
5-Cl
B
C
A
B
C
A
B
A
A
A
A
A
B
D
E
F
F
F
G
H
H
I
5
13
20
56
52
80
61
49
78
62
67
78
86
55
29
3
64
>128
8
128
1, 64
0.5
4
0.125, 8
0.25, 64
1, 128
1, 64
8
>128
128
128
8
4
16
16, 128
NT
5-Cl
5-Me
>128
>128
>128
>128
>128
>128
>128
>128
>128
>128
>128
64
16
4
16, 128
NT
5-OCF3
5-CF3
6-Br
8
8, 128
32, 128
NT
8
6-Cl
6-Cl
8
8
NT
5-Br
5-Br
>128
>128
>128
>128
64
1, 128
0.25, 128
4, 64
5-Br
5-Cl
5
4, 64
1, 8
1, 32
5-CN
5-Br
10
49
22
5
>128
>128
>128
32
>128
>128
>128
>128
>128
>128
>128
>128
>128
128
5-CN
5-Br
5-Br
0.125, 4
0.125, 16
16, 128
64, 128
0.25, 16
2, 32
28
7
5-CN
5-Br
I
>128
16
64
J
42
99
49
NT
NT
NT
NT
5-CN
5-CN
5-CN
5-Br
K
L
M
M
N
N
64
16
0.25, 4
0.125, 1
0.125, 16
0.125, 1
0.125, 8
16
4
5-CN
5-Br
64
128
2
a S. aureus coupled transcription translation assay.12
b Minimum Inhibitory concentration versus Staphylococcus aureus UC9218.
c Staphylococcus aureus UC9218 + 10% Pooled Human Serum. Human serum (male, from Sigma) was thawed at room temperature, then placed in a
56 °C water bath for 30 min. The serum was filtered using a 0.2 micron filtration system.
In an attempt to better understand the SAR of the inda-
zol series relative to the anthranilic acid series, small
molecule X-ray crystal structures of an indazol analog
47 (Fig. 3) and a highly potent anthranilic acid analog
(Fig. 4) were obtained.11 It was hoped that the solid-
state conformation would shed some light on the bound
conformation. Interestingly, the structure of anthranilic
acid 48 indicated that the northern two aromatic rings,
the anthranilic acid and the indole ring, adopt a planar
or nearly planar orientation (Fig. 4) because of the pres-
ence of an intramolecular hydrogen bond. In contrast,
the indazol 47 (Fig. 3) is twisted significantly out of
plane relative to the aromatic B-ring (Fig. 5). Com-
pound 47 lacks the ability to form an intramolecular
H-bond to the amide carbonyl because of the N-methyl
indole substituent. Steric interactions between this

6296
C. Stiff et al. / Bioorg. Med. Chem. Lett. 18 (2008) 6293–6297
OH
N
Br
N
O
N
S
O
47
O
Figure 3. X-ray structure of indazol 47.
Figure 5. Structural overlay of 47 (white) and Anthranilic acid 48
(green).
OH
N
ols are readily prepared via the synthetic routes outlined
above. An X-ray structure of a potent anthranilic acid
demonstrated that the upper two rings adopt a planar
or nearly planar conformation (Fig. 2). In contrast,
indazol analogs such as 47 (Fig. 1) are twisted out of
plane, providing a hypothesis regarding their reduced
activity relative to the anthranilic acids. Further investi-
gation of the SAR around the acylindazol core and the
B-ring structure with groups designed to favor a planar
conformation may provide access to a more active series
of compounds.
O
NH
O
HN
HN
O
48
Figure 4. X-ray crystal structure of potent anthranilic acid 48.
References and notes
1. Unpublished results and Larsen, S. D.; Hester, M. R.;
Ruble, J. C.; Kamilar, G. M.; Romero, D. L.; Wakefield,
B.; Melchior, E. P.; Sweeney, M. T.; Marotti, K. R.
Bioorg. Med. Chem. Lett. 2006, 16, 6173.
2. Ruble, J. C.; Wakefield, B. D.; Kamilar, G. M.; Marotti,
K. R.; Melchior, E. P.; Sweeney, M. T.; Zurenko, G. E.;
Romero, D. L. Bioorg. Med. Chem. Lett. 2007, 17, 4040.
3. Translation inhibition was demonstrated by using mRNA
(no DNA) in the assay described in Ref. 12.
4. Li, J.; Wakefield, B. D.; Ruble, J. C.; Stiff, C. M.; Romero,
D. L.; Marotti, K. R.; Sweeney, M. T.; Zurenko, G. E.;
Rohrer, D. C.; Thorarensen, A. Bioorg. Med. Chem. Lett.
2007, 17, 2347.
5. Thorarensen, A.; Wakefield, B. D. C. M.; Romero, D. L.;
Marotti, K. R.; Sweeney, M. T.; Zurenko, G. E.; Rohrer,
D. C.; Han, F.; Bryant, G. L. Bioorg. Med. Chem. Lett.
2007, 17, 2823.
methyl substituent and the amide carbonyl may also
cause the twisting. In contrast, anthranilic acids bearing
indole N-1 methyl substitution (Table 1, analogs of
compound 44 or 46) are still highly potent. Modeling
studies support the adoption of a planar conformation
similar to compound 48 likely due to a steric interaction
between the N-methyl indole substituent and the amide
NH. The indazole ring may not provide enough bulk to
bias the conformation. If the out of plane conformation
of the indazols is the cause of their reduced activity rel-
ative to the anthranilic acid counterparts (Fig. 5), future
indazol analogs should be designed with an aryl B-ring
that can intramolcular H-bond to the indazol ring nitro-
gen or to the carbonyl of the amide. This would allow
for the adoption of a planar conformation and subse-
quent proper positioning of the B-ring substituent.
6. Stiff, C. M.; Zhong, M.; Sarver, R. W.; Gao, H.; Ho, A.
M.; Sweeney, M. T.; Zurenko, G. E.; Romero*, D. L.
Bioorg. Med. Chem. Lett. 2007, 17, 2823.
The SAR of the indazol series of analogs deviates from
that of the anthranilic series of analogs in one other way.
Substitution at the 5-position of the anthranilic acid
(corresponding to the 6-position on the indazol) is not
well tolerated,1 whereas in the indazol series the 6-
substituted analogs are equipotent to the 5-substituted
compounds (Table 1, 28 vs 20, 29 vs 23, 30 vs 24). The
indazol upper ring does not overlap exactly with the
anthranilic acid when the southern half of the analog
is held fixed (Fig. 5). This may cause the indazol to ro-
tate or be positioned differently in the binding site allow-
ing for this alternate substitution pattern.
7. Baiocchi, L.; Corsi, G.; Palazzo, G. Synthesis 1978, 633.
8. Anderson, R. J. Chem. Res., Synop. 1985, 376.
9. Compound 26: 1-{3-[(5-chloro-2,3-dihydro-1H-indol-1-
yl)sulfonyl]benzoyl}-5-(trifluoromethoxy)-1H-indazol-3-ol.
Carboxylic acid (337 mg, 1.0 mmol) was placed in a
scintillation vial and suspended in 10 mL of CH2Cl2.
CDI was added (200 mg, 1.25 mmol) and the vial was
capped and agitated for 4 h. Then the indazol (218 mg,
1.0 mmol) was added and the reaction mixture was
allowed to agitate overnight (ꢀ18 h). The reaction mixture
was diluted with 50 mL of CH2Cl2. The organic layer was
washed with 1 N HCl, water, and brine, and dried over
Na2SO4. After concentration, the product was purified by
flash chromatography (1% MeOH in CH2Cl2) affording
403 mg of an off-white solid (75% yield). Analytical sample
obtained by recrystallization from MeOH. 1H NMR
In conclusion, we have discovered that an indazol ring
can serve as a novel anthranilic acid bioisostere. Indaz-

C. Stiff et al. / Bioorg. Med. Chem. Lett. 18 (2008) 6293–6297
6297
(400 MHz, DMSO) 2.96 (t, 2H), 4.02 (t, 2H), 7.24–7.26
(m, 2H), 7.47 (d, 1H), 7.73–7.77 (m, 2H), 7.86 (s, 1H), 8.01
(d, 1H), 8.22 (d, 1H), 8.37 (s, 1H), 8.47 (d, 1H), 12.45 (s,
1H). Anal. Calcd for C23H15ClF3N3O5S: C, 51.36; H, 2.81;
N, 7.81. Found: C, 51.41; H, 2.89; N, 7.79.
Marotti, K. R.; Miller, A. A. J. Antimicrob. Chemother.
2008, 62, 720.
11. Crystallographic data (excluding structure factors) for the
structures in the paper have been deposited at the
Cambridge Crystallographic Data Centre as CCDC
646068 & 646069.
12. Murray, R. W.; Melchior, E. P.; Hagadorn, J. C.; Marotti,
K. R. Antimicrob. Agents Chemother. 2001, 45,
1900.
10. Antibacterial activity does not fully correlate with activity
in the T/T assay, indicating that the compounds may
operate by an additional mechanism. See: Mott, J. E.;
Shaw, B. A.; Smith, J. F.; Bonin, P. A.; Romero, D. L.;