Discovery and initial development of a novel class of antibacterials: Inhibitors of Staphylococcus aureus transcription/translation

Article abstract of DOI:10.1016/j.bmcl.2006.09.044

The novel bacterial transcription/translation (TT) inhibitor 1 was identified through a combination of high throughput screening and exploratory medicinal chemistry. Initial optimization of the anthranilic acid moiety and sulfonamide amine diversity was accomplished via 1- and two-dimensional solution phase libraries, resulting in an improvement in the MIC of the lead from 64 to 8 μg/mL (compound 4l). Subsequent modification of the central aromatic ring and further refinement of the sulfonamide amines required the development of a solid phase route on Wang resin. The resulting libraries generated a number of potent antibacterials with MICs of ≤1 μg/mL (e.g., 10b, 12, and 13). During the course of this work, it became apparent that the antibacterial activity of the series is not fully correlated with TT inhibition, suggesting that at least one additional mechanism of action is operative.

Full text of DOI:10.1016/j.bmcl.2006.09.044

Bioorganic & Medicinal Chemistry Letters 16 (2006) 6173–6177
Discovery and initial development of a novel class of antibacterials:
Inhibitors of Staphylococcus aureus transcription/translation
Scott D. Larsen,* Matthew R. Hester, J. Craig Ruble, Gregg M. Kamilar,
Donna L. Romero, Brian Wakefield, Earline P. Melchior,
Michael T. Sweeney and Keith R. Marotti
Medicinal Chemistry and Infectious Diseases Biology, Pharmacia Corporation, 301 Henrietta Street, Kalamazoo, MI 49001, USA
Received 14 July 2006; revised 7 September 2006; accepted 14 September 2006
Available online 5 October 2006
Abstract—The novel bacterial transcription/translation (TT) inhibitor 1 was identified through a combination of high throughput
screening and exploratory medicinal chemistry. Initial optimization of the anthranilic acid moiety and sulfonamide amine diversity
was accomplished via 1- and two-dimensional solution phase libraries, resulting in an improvement in the MIC of the lead from 64
to 8 lg/mL (compound 4l). Subsequent modification of the central aromatic ring and further refinement of the sulfonamide amines
required the development of a solid phase route on Wang resin. The resulting libraries generated a number of potent antibacterials
with MICs of 61 lg/mL (e.g., 10b, 12, and 13). During the course of this work, it became apparent that the antibacterial activity of
the series is not fully correlated with TT inhibition, suggesting that at least one additional mechanism of action is operative.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Bacterial protein synthesis has proven to be a fruitful
target for antibiotic discovery.1 A number of marketed
antibiotics inhibit bacterial growth through inhibition
of prokaryotic RNA transcription and protein transla-
tion. Rifampin is a potent inhibitor of bacterial RNA
polymerase, and the macrolides, lincosamides, amino-
glycosides, tetracyclines, and oxazolidinones all have
protein translation as their site of action. Unfortunately,
the increase in the antibiotic resistance of Gram-positive
bacteria threatens to reduce the effectiveness of these
and other antibiotics. These concerns act as an incentive
to discover new and more effective transcription and
protein translation inhibitors.
(SAR) of this lead through parallel medicinal chemistry.
A straightforward 2-step solution phase route was
developed that was suitable for preparing 1- and two-
dimensional libraries of 50–150 compounds from
CO2H
NH
O
SO2N(nPr)2
1
Through a combination of high throughput screening
and exploratory medicinal chemistry,2 1 was identified
as a novel inhibitor of bacterial transcription/transla-
tion3 (TT) with modest antibacterial activity against
Staphylococcus aureus.4 The presence of easily modified
carboxamide and sulfonamide bonds encouraged us to
rapidly expand the structure–activity relationships
HO2C
i, ii
HO2C
SO2NR1R2
SO2Cl
iii, iv
G
3
2
R3
NH
O
SO2NR1R2
4
Keywords: Staphylococcus aureus; Athranilic acid.
*
Scheme 1. Reagents and conditions: (i) R1R2NH2, MeOH, 0 ꢁC, rt,
3 h; (ii) aq HCl; (iii) SOCl2, DMF, 60 ꢁC, 3 h or (CO)2Cl2, cat. DMF,
CH2Cl2, rt, 2 h; (iv) (R3)(G)PhNH2, rt, 3 h.
Corresponding author. Present address: Pfizer, 2800 Plymouth Road,
28/2026E, Ann Arbor, MI 48105, USA. Tel.: +1 7346222535; fax: +1
7346221407; e-mail: scott.d.larsen@pfizer.com
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.09.044

6174
S. D. Larsen et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6173–6177
Table 1. Biological activity of selected library analogs 4
Compound
G
R3
R1
R2
% TT inha
SAUR MICb (lg/mL)
1
2-COOH
H
2-OH
4,5-fused Ph
4,5-fused Ph
4,5-fused Ph
4-Br
n-Pr
n-Pr
n-Pr
n-Pr
(CH2)5
n-Pr
Me
n-Pr
n-Pr
n-Pr
n-Pr
30
10
10
61
35
0
64
>128
>128
8
4a
4b
4c
4d
4e
4f
2-COOH
2-COOH
2-COOH
2-COOH
2-COOH
2-COOH
2-COOH
2-COOH
2-COOH
2-COOH
2-Me
4-Br
16
>128
32
4,6-diOMe
4-Br
4-Br
n-Pr
4-(MeO)Ph
3-(CF3)Ph
3-(CF3)Ph
3-(CF3)Ph
3-(CN)Ph
MeO–(CH2)2
4-(Cl)Ph
4-(Cl)Ph
4-(Cl)Ph
Ph
50
65
35
55
30
10
75 80c
12
45
50
80
90
34
85
47
10
30
75
20
4
4g
4h
4i
H
8
4-OMe
4,5-diF
4-Br
H
>128
>128
64
H
4j
H
4k
4l
4-Br
4-Br
MeO–(CH2)2
Me
Me
32
16 8c
>128
16
4m
4n
4o
4p
4q
4r
4s
4t
4-Br
4-Br
2-COOH
2-COOH
2-COOH
2-COOH
2-COOH
2-COOH
2-COOH
3-COOH
2-COOH
2-COOH
2-COOH
2-COOMe
H
4-Cl
4-Br
n-Pr
n-Pr
n-Pr
n-Pr
Me
32
16
Ph
Ph
4-I
16
4-Me
4-Cl
5-Cl
Ph
128
16
4-(Cl)Ph
4-(Cl)Ph
4-(Cl)Ph
4-(Cl)Ph
3-(BnO)Ph
3-(MeO)Ph
3-(BnO)Ph
Me
Me
64
4u
4v
4w
4x
4y
4-Cl
4-F
>128
>128
8
Me
H
4-Br
4-Br
H
16
>128
4-Br
H
a Percent inhibition of S. aureus coupled transcription/translation at 100 lM (Ref. 3).
b Minimum Inhibitory Concentration against S. aureus (UC 9218, ATCC 29213) (Ref. 4).
c Values are for resynthesized singleton.
commercially available 4-(chlorosulfonyl)benzoic acid 2
and diverse amines and anilines (Scheme 1). The method
was sufficiently robust that 282 compounds out of 362
possible targets were pure enough for biological assay
(av purity 80%) without HPLC purification (av yield
72% overall).
O
O
Br
Br
i
O
O
N
H
O
NH2
5
Scheme 2. Reagents and conditions: (i) Wang resin, DMF, DMAP,
60 ꢁC, 24 h.
Most of the aniline diversity was selected to include one
carboxyl or hydrogen-bonding group (G = CO2H,
CO2Me, OH, OMe, CN, and PhCO) and at least one
other substituent (R3 = halogen, alkyl, Ph, OMe, OH,
fused phenyl). Amine diversity was designed to span a
range of hydrophilicities and MWs (R1, R2 = H, alkyl,
alkyl-OMe, morpholine, piperidine, NHAr, NMeAr,
NCH2Ar) with varying substitution on the Ar groups.
Representative SAR is presented in Table 1.
HO2C
HO2C
i
R3
R3
ARYL
ARYL
SO2Cl
6
7
O
O
Br
O
Br
iv
O
NH
ii, iii
NH
The carboxylic acid, positioned ortho to the NH, proved
to be essential for activity (compare 1, 4a, 4b, 4s, 4u, 4w,
and 4y). The naphthyl group of the lead could be suc-
cessfully replaced by 4-haloanthranilic acid (4c), and
the halide was optimally placed at the 4-position (4s,
4t). The lipophilic halides Br and I were superior to
the more hydrophilic Cl, F, Me or OMe (4g, 4h, 4o–
4r). With regard to the sulfonamide substituents, sec-
ondary amines were slightly better than primary (4l,
4n) and aromatics could replace aliphatic. Lipophilic
amines were necessary for good TT inhibition and anti-
bacterial activity (e.g., 4g, 4j). One of the best com-
pounds from the library was resynthesized and found
to inhibit TT by 80% at 100 lM with an MIC of
8 lg/mL (4l).
O
R3
ARYL
SO2Cl
O
R3
SO2NR1R2
ARYL
9
8
O
Br
OH
NH
v
O
R3
SO2NR1R2
ARYL
10
Scheme 3. Reagents and conditions: (i) ClSO3H, 0–80 ꢁC, 2 h; (ii)
(ClCO)2, cat. DMF, CH2Cl2, rt, 18 h; (iii) resin 5, CH2Cl2, pyr, rt, 4 h;
(iv) R1R2NH2, Et3N, CH2Cl2, rt, 15 h; (v) TFA, CH2Cl2, rt, 1 h.

S. D. Larsen et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6173–6177
6175
Table 2. Biological activity of selected validation library analogs 10 (ARYL = 1,4-disubstitued Ph; R3 = H)
a
Compound
NR1R2
% TT inh at 100 lM (IC50
)
SAUR MIC (lg/mL)
N
10a
71 (80)
1
Cl
Cl
N
N
10b
10c
10d
10e
10f
96 (40)
87 (55)
90 (60)
88 (60)
77
1
2
N
4
HN
4
4
N
OH
10g
10h
10i
22
8
N
N
96 (55)
84 (60)
57
8
Ph
N
N
N
N
8
10j
8
10k
10l
34
16
16
16
32
>128
98 (28)
96 (30)
62
Ph
Ph
N
N
10m
10n
Ph
N
H
N
OH
10o
19
a IC50 values are in lM.
To facilitate an evaluation of the central aromatic ring, a
solid phase route was developed that would also permit
additional modification of the sulfonamide amine,
potentially allowing the evaluation of a full two-dimen-
sional array. Because of the clear superiority of the
5-bromoanthranilic acid in the solution phase libraries,
this diversity element was fixed in the solid phase li-
braries. Commercially available 5-bromoisatoic anhy-
dride was loaded onto Wang resin with DMAP
catalysis, affording resin-bound bromoanthranilic acid
5 (Scheme 2).5 Commercially available aryl carboxylic
acids 6 were chlorosulfonylated with chlorosulfonic
acid, and the resulting sulfonyl chlorides 7 were loaded
onto resin 5 via the corresponding acid dichlorides
(Scheme 3). Consistent with literature precedent, the res-
in-bound aniline reacted preferentially with the carbonyl
chloride vs the sulfonyl chloride.6 The resulting resin-
bound sulfonyl chlorides 8 were reacted with diverse
amines to afford sulfonamides 9. Cleavage from the res-
in was then effected with TFA, giving the desired ana-
logs 10.
A one-dimensional validation library was first prepared
as in Scheme 3 by combining the aryl sulfonyl chloride

6176
S. D. Larsen et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6173–6177
derived from 2, resin 5, and 46 diverse amines that were
selected based on the SAR of the solution phase li-
braries. Average overall yield of the resulting analogs
10 was 78% with 25 final compounds being pure enough
(>70%) for initial biological assay. Representative SAR
is presented in Table 2.
Cl
N
HN
N
MeN
Cl
Cl
c
b
d
a
HN
HN
O
O
HN
N
N
CF3
N
Bn
O
It is clear from these initial library results that reducing
the lipophilicity of the sulfonamide amine, either
through addition of polar functionality or simply lower-
ing the molecular weight, attenuates antibacterial activ-
ity. Interestingly, fused aromatic bicyclic systems (e.g.,
indoline and isoindoline) were superior to aromatic
substituted cycloalkylamines. It is noteworthy that this
library revealed that some analogs in this series possess
antibacterial activity without significant TT inhibition
(e.g., 10g), suggesting that an additional mechanism
for antibacterial activity is likely operative. Neverthe-
less, a significant advancement in both TT and antibac-
terial activity was realized in this validation library
relative to the best compound from the solution phase
libraries (10b vs 4l).
h
g
f
e
HN
HN
HN
HN
HN
Cl
PO3Et2
NEt2
Cl
F
l
k
i
j
BnN
N
HN
NH2
NBu2
HN
HN
N
NMe2
p
o
n
m
N
HN
N
N
HN
HN
S
NMe2
Cl
t
q
u
s
r
N
EtN
NBn
NH(CH2)2NEt2
N
Encouraged by the results of the validation library, a
larger two-dimensional array was designed and put
into production. A total of 12 central aromatic rings
and 45 amines were selected as diversity elements.
The central rings are presented in Figure 1. Amines
were selected to maintain some of the successful ele-
ments of the previous libraries (bicyclics, aromatics)
along with additional attempts to reduce lipophilicity
and incorporate some new diversity. It was assumed
that production on solid phase would facilitate the
isolation of more analogs with more divergent physi-
cochemical properties.
O
v
w
Figure 2. Successful amine diversity elements in two-dimensional
library.
L
K
J
Aromatic
Rings
I
H
G
F
E
D
C
B
A
% Inhibition
80-100
60-80
40-60
20-40
0-20
The success rate of the larger library was more modest
than those of the solution phase or solid phase valida-
tion libraries, presumably due to the production for-
mat (96-well plates) and the wider diversity of the
amines, which spanned a broad range of reactivities.
Although all 12 acid chloride diversity elements got
successfully incorporated into products, only 23
amines yielded products (Fig. 2). A total of 156 com-
pounds were obtained in purities sufficient for biolog-
ical assay (>70%).
a b c d e f g h i j k l m n o p q r s t u v w
Amines
Figure 3. Contour plot of % TT inhibition (at 30 lM) vs amine and
aromatic ring diversity elements for two-dimensional library.
L
K
J
Aromatic
Rings
I
H
G
Log 1000/MIC
F
3-4
2-3
1-2
0-1
CO2H
E
D
C
B
CO2H
CO2H
CO2H
CO2H
MeO
SO2Cl
Br
SO2Cl
SO2Cl
CO2H
SO2Cl
D
B
C
a b c d e f g h i j k l m n o p q r s t u v w A
A
Br
S
CO2H
S
Amines
ClO2S
HO2C
Figure 4. Contour plot of log1000/MIC versus amine and aromatic
ring diversity elements for two-dimensional library.
SO2Cl
SO2Cl
SO2Cl
E
F
G
H
OMe
S
CO2H
CO2H
CO2H
Key SAR from this library is summarized in the Excelꢂ
contour plots depicted in Figures 3 and 4. These plots
use color to depict levels of activity, as indicated by
the legends in the figures. It is important to note that
there are only 156 data points contained within the
Cl
ClO2S
SO2Cl
CO2H
SO2Cl
J
L
F
SO2Cl
Figure 1. Central aromatic ring diversity.

S. D. Larsen et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6173–6177
6177
Br
CO2H
NH
line (c) and 2-aminoindane (q), while the best central
aromatic ring is F.
Br
CO2H
NH Br
O
O
Despite the apparent disconnect between TT and anti-
bacterial activity, this library was successful in identify-
ing four compounds with MICs less than 1 lg/mL
(Fig. 5). One of these compounds (15) proved to be iden-
tical to one of the best compounds from the validation
library (10b) that was included as a control.
SO2
SO2
HN
HN
12
13
In summary, through a series of both solution and solid
phase libraries, we were successful in rapidly converting
a modest bacterial TT inhibitor with weak antibacterial
activity (MIC = 64 lg/mL) into a number of highly po-
tent antibacterials (MICs = 0.25–1 lg/mL). In the pro-
cess we discovered that the antibacterial activity within
the series is not fully coupled with TT inhibition. This
apparent disconnect is suggestive of additional mecha-
nisms of antibacterial activity. Studies are underway to
better understand this.
MIC = 0.25 ug/mL
MIC = 0.25 ug/mL
TT inh (30 uM) = 20%
TT inh (30 uM) = 30%
Br
CO2H
Br
CO2H
NH Br
NH
O
O
SO2
SO2
N
HN
O
Cl
O
Acknowledgments
14
15 (10b)
MIC = 0.5 ug/mL
TT inh (30 uM) = 93%
MIC = 0.5 ug/mL
We are grateful to Scott Collibee, David Jenkins, and
Allison Walter of Albany Molecular Institute for the
production and plating of the two-dimensional solid
phase library.
TT inh (30 uM) = 15%
Figure 5. Most potent antibacterial compounds from the two-dimen-
sional library.
References and notes
12 · 23 = 276 possible grid intersections depicted in
these plots. This is because not all possible combinations
were tested due to insufficient purity or simple absence
of product. Thus these plots are only useful for identifying
pockets of significant activity corresponding to particular
diversity elements. Regions where no activity is apparent
can either be due to genuine lack of activity or to simple
lack of testing data. As can be seen in Figure 3, the
amines associated with good inhibition of TT are c, e,
m, and q (identified in Fig. 2). The aromatic rings most
associated with good TT inhibition are F, I, and J. In
Figure 4 it can be seen that the amines associated with
the best antibacterial activity were c, e, h, and q, while
the best central aromatic rings were A and F. Although
the contour plots bear some resemblance, indicating a
correlation of TT inhibition with antibacterial activity,
there are significant areas of diversion where antibacte-
rial activity exists without significant TT inhibition
(e.g., for amine h). It is apparent from the two contour
plots that the amines associated with the best combina-
tion of TT and antibacterial activity are 6-chloroindo-
1. (a) Walsh, C. Antibiotics: Actions, Origins, Resistance;
ASM Press: Washington, DC, 2003, Chapter 4; (b) Bodde-
ker, N.; Bahador, G.; Gibbs, C.; Mabery, E.; Wolf, J.; Xu,
L.; Watson, J. RNA 2002, 8, 1120, and references therein.
2. Bundy, G. L.; Banitt, L. S.; Palmer, J. R. unpublished
results.
3. Murray, R. W.; Melchior, E. P.; Hagadorn, J. C.; Marotti,
K. R. Antimicrob. Agents Chemother. 2001, 45, 1900.
4. National Committee for Clinical Laboratory Standards,
Approved Standard. Methods for Dilution Antimicrobial
Susceptibility Tests for Bacteria that Grow Aerobically,
5th ed.; NCCLS document M7-A4: Wayne, Pennsylvania,
2000.
5. (a) Heyman, D. A. J. Heterocycl. Chem. 1978, 15, 1131; (b)
Blagbrough, I. S.; Coates, P. A.; Hardick, D. J.; Lewis, T.;
Rowan, M. G.; Wonnacott, S.; Potter, B. V. L. Tetrahedron
Lett. 1994, 35, 8705; (c) Detailed procedures for the
preparation of two-dimensional libraries on solid phase
are included in WO 2004/018414.
6. Fielding, H. C.; Shirley, I. M. J. Fluorine Chem. 1992,
59, 15.