Preparation of novel antibacterial agents. Replacement of the central aromatic ring with heterocycles

Article abstract of DOI:10.1016/j.bmcl.2006.12.055

Discovery of novel antibacterial agents is a significant challenge. We have recently reported on our discovery of novel antibacterial agents in which we have rapidly optimized potency utilizing a parallel chemistry approach. These advanced leads suffer from high affinity for human serum albumin (HSA). In an effort to decrease the affinity for HSA we have prepared a series of heterocyclic analogs, which retained antibacterial activity and demonstrated reduced affinity for HSA.

Full text of DOI:10.1016/j.bmcl.2006.12.055

Bioorganic & Medicinal Chemistry Letters 17 (2007) 2347–2350
Preparation of novel antibacterial agents. Replacement of the
central aromatic ring with heterocycles
Jianke Li, Brian D. Wakefield, J. Craig Ruble, Cory M. Stiff, Donna L. Romero,
Keith R. Marotti, Michael T. Sweeney, Gary E. Zurenko,
Douglas C. Rohrer and Atli Thorarensen*
Medicinal Chemistry and Infectious Diseases Biology, Pharmacia Corporation, 301 Henrietta Street, Kalamazoo, MI 49001, USA
Received 21 November 2006; accepted 13 December 2006
Available online 21 December 2006
Abstract—Discovery of novel antibacterial agents is a significant challenge. We have recently reported on our discovery of novel
antibacterial agents in which we have rapidly optimized potency utilizing a parallel chemistry approach. These advanced leads suffer
from high affinity for human serum albumin (HSA). In an effort to decrease the affinity for HSA we have prepared a series of het-
erocyclic analogs, which retained antibacterial activity and demonstrated reduced affinity for HSA.
Ó 2006 Elsevier Ltd. All rights reserved.
Bacterial resistance is a significant problem in the treat-
ment of bacterial infections.1–3 This emerging resistance
has fueled a continuous search for new antibiotics
resulting in numerous commercially available products.
The pharmaceutical industry rapidly took advantage
of the wealth of novel targets available as a result of
the genomic revolution. Despite the expectation that
these new targets would decrease the hurdle in identify-
ing novel classes of antibacterial agents, discovery of
compounds that act via novel mechanisms remains a sig-
nificant challenge. The major obstacle appears to be the
identification of novel drugable chemical matter.4 In the
past several years, we have had an extensive research
program utilizing a range of approaches in search of
novel antibacterials. This includes our work directed to-
ward specific targets such as efflux pumps,5 PDF,6 as
well as the identification of whole cell active compounds
coupled with a reverse pharmacology approach to try to
identify the targets against which such compounds are
acting.7
collection was screened utilizing a transcription transla-
tional assay8 which resulted in the identification of a hit
that was a complex mixture exhibiting both whole cell
antibacterial activity and inhibition in the transcription
translation screening assay. After extensive fraction-
ation, the structure of the active component was eluci-
dated and determined to be 1 (Fig. 1). Our medicinal
chemistry group rapidly transformed this complex unat-
tractive compound into the lead anthranilic amide mol-
ecule 2, which is very attractive for lead optimization
utilizing array chemistry. We have recently reported9
on the optimization of 2 resulting in the identification
of the advanced compound 3 which displayed potent
broad-spectrum antibacterial activity, but not in vivo
CO2H
Br
CO2H
CO2H
A
B
NH
O
NH
O
O
O
O
O
O2S
N
O2S
N
An attractive target for antibacterial drug discovery is
bacterial protein synthesis. The Pharmacia compound
3
2
O2S
N
Keywords: Anthranilic acid; Antibacterial; Protein binding; Human
serum albumin.
1
Cl
*
Corresponding author at present address: Pfizer, 700 Chesterfield
Parkway West AA2G, Chesterfield, MO 63017, USA. Tel.: +1 636
247 3962; fax: +1 636 247 0250; e-mail: atli.thorarensen@pfizer.com
Figure 1. Progression in lead optimization.
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.12.055

2348
J. Li et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2347–2350
Table 1. Antibacterial activity of initial lead matter
Compound
MIC (lg/mL)a
EFAEd
SAURb
SAURc
SEPIe
SPNEf
HINFg
1
16
64
ND
32
16
16
4
8
>128
>128
128
2
>128
>128
2.0
128
32
64
4
3
Vancomycin
0.5
0.5
1.0
1.0
0.25
0.015
a Minimal inhibitory concentration.
b S. aureus UC 9218.
c S. aureus UC 9218 + 5% 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 then filtered using a 0.2 micron filtration system.
d Enterococcus faecalis UC 9217.
e S. epidermidis UC 12084.
f S. pneumoniae UC 9912.
g Haemophilus influenzae 30063.
tuent in place of the sulfonamide due to availability of
CN
1) a) (COCl)2/DCM, b) KOt-Bu/THF/
the appropriate building blocks.
0ο C 93%, 2) Raney Nickel H2 99%,
CO2H
The analogs were prepared as outlined in Scheme 2,
acylation followed by acidic deprotection provided the
final analogs in high yields and purities, Table 2.
3) NIS/DMF/ 0 C 88% 4)
CuCN/NMP/ 110 C 75%
CO2t-Bu
NO2
4
NH2
5
Scheme 1.
In general most of the building blocks contained a com-
parable substitution pattern. Table 2 contains selected
examples and activities of the various 5- and 6-mem-
bered heterocyclic analogs that we prepared, and Table 3
contains selected examples and activities of fused bicy-
clic analogs. In monitoring progress, we decided early
on that improving potency in the whole cell assay versus
S. aureus would be the driving factor instead of potency
in the transcription translation assay. Furthermore,
after evaluating many approaches to measuring albumin
affinity, we decided the best measure of progress in
reducing protein binding was to run the S. aureus assay
in the presence of 10% serum. Utilizing this gauge it is
clear that several heterocyclic replacements were
extremely successful replacements for the B-phenyl ring.
For example in the monocyclic series (Table 2), isoxaz-
ole 7, thiazole 10, and oxazole 11 provided potent anti-
bacterial agents with decreased affinity for HSA. In the
bicyclic series (Table 3), the indole 17, benzisoxazole 18,
and benzthiazole derivative 20 demonstrated very prom-
ising antibacterial activity.
activity in a standard Staphylococcus aureus mouse
model (Table 1). The lack of in vivo activity was attrib-
uted to the high human serum albumin (HSA) binding
affinity of compound 3. In this letter, we will describe
our approach for replacing the central ring with a vari-
ety of heterocycles.
Previous optimization of the A-ring led to the identifica-
tion of several 5-substituents with good activity.9 For
the purposes of this work, we elected to utilize a 5-cyano
substituent.10 In order to rapidly evaluate the replace-
ment of the B-ring with various heterocycles, we re-
quired a suitable A-ring building block where the acid
is protected as an acid-labile ester to simplify the final
deprotection. Nitrile 5 was synthesized via a simple
4-step procedure (Scheme 1). Conversion of 2-nitroben-
zoic acid to the corresponding acid chloride utilizing
oxalyl chloride was uneventful; subsequent treatment
with potassium tert-butoxide in THF cleanly afforded
the desired tert-butyl ester. This acylation has been
scaled uneventfully to >60 g. Reduction followed by
iodination provided an intermediate suitable for the
installment of the cyanide. The iodide was converted
to the nitrile utilizing CuCN in variable yield. We uti-
lized this route to generate over 25 g of the desired ni-
trile. Subsequently this route was improved by
Pharmacia’s process research group to provide kg quan-
tity in a reliable manner.11
The identification of several potent heterocyclic replace-
ments for the B-ring phenyl of 3 validated our approach
which was designed to improve the potency of new ana-
logs, while reducing lipophilicity and maintaining or
reducing the molecular weight. In order to further
validate this approach, we elected to prepare more
specific analogs containing one of the promising new
O
Criteria for the selection of building blocks included (a)
commercial availability, (b) existence in our compound
collection, or (c) easy preparation from known
procedures. Furthermore, previous SAR studies had
illustrated that the sulfonamide attached to the B-ring
could be replaced with a variety of groups such as a
phenyl ring. Therefore, we elected to use a phenyl substi-
NC
OH
NH
O
1) (COCl)2, DCM
2) 5, CH2Cl2/ pyr
OH
R
3) TFA/DCM
O
R
Scheme 2.

J. Li et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2347–2350
2349
Table 2. Antibacterial activity of selected 5- and 6-membered hetero-
cyclic derivatives
Table 3. Antibacterial activity of selected heterobicyclic derivatives
Compound Yield %
MIC (lg/mL)a
SAURb 10%c Serum
R
R
Compound Yield %
MIC (lg/mL)a
SAURb 10%c Serum
17
36
2
32
H
N
N
6
37
4
32
Ph
N
N
O
O
Ph
Ph
N
18
19
20
21
ND
27
0.5
2
16
32
8
7
8
76
70
0.25
8
O
N
O
64
>128
O
N
N
Ph
9
45
41
42
17
37
8
32
16
8
30
0.25
4
S
S
S
S
N
N
N
O
10
11
12
13
1
48
32
Ph
N
0.5
0.5
0.5
O
Ph
22
23
10
17
>128
>128
O
Cl
N
8
Ph
N
32
Ph
16
64
O
N
S
H3C
a Minimal inhibitory concentration.
b S. aureus UC 9218.
O
N
14
12
8
4
>128
N
c S. aureus UC 9218 + 5% serum.
SBn
N
N
1) nBuLi, THF, -20οC
2) BnS-SBn
N
Ph
15
16
60
17
>128
>128
H
TMS
BnS
TMS
24
Me
TBAF, THF
33%, 2 steps
N
N
128
O
O
NaNO2
NH2
NOH
H
26
SBn
EtO
a Minimal inhibitory concentration.
b S. aureus UC 9218.
c S. aureus UC 9218 + 10% serum.
EtO
25
HCl
HCl
54%
Et3N, tol, 90ο C
51%
Cl
O
O
LiOH
SBn
SBn
X-ray
heterocycles, but with a more optimal substitution pat-
tern. We had previously identified benzyl thioether as
a favorable substituent which resulted in very potent
analogs. Therefore, we elected to prepare the ben-
zylthioether analog of isoxazole 7. The synthesis relies
on a cycloaddition to form the isoxazole core (Scheme 3).
The requisite benzyl ethynyl sulfide 25 was prepared
by the addition of lithium acetylide to a disulfide form-
ing 24, which was followed by deprotection of the acet-
ylene. The isoxazole 27 was then prepared by a [3+2]
cycloaddition.12 The cycloaddition reaction was exten-
sively optimized to minimize the formation of isomeric
and dimeric products.
EtO
HO
THF/H2O
82%
N
O
N
28
O
27
O
NC
1) (COCl)2, DCM/DMF
2) 5, DCM, pyr
OH
NH
3) TFA, DCM
57%
SBn
O
29
N
O
Scheme 3.
heated toluene solution containing both reaction
partners.13 In order to secure the assignment of the reg-
iochemistry, an X-ray structure was obtained of the
The optimal reaction conditions afforded a single regio-
isomeric product 27 when Et3N was added dropwise to a

2350
J. Li et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2347–2350
Table 4. Antibacterial properties of 29
activity with reduced affinity for HSA. These efforts will
be disclosed in due course.
Compound
MIC (lg/mL)a
SAURb 10%c Serum SPNEd EFAEe SEPIf
0.125 >128 >128 0.125
29
4
References and notes
a Minimal inhibitory concentration.
b S. aureus UC 9218.
1. Cohen, M. L. Science 1992, 257, 1050.
c S. aureus UC 9218 + 10% 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 then filtered using a 0.2 micron filtration
system
2. (a) Allen, N. E. In Progress in Medicinal Chemistry Ellis,
G. P.; Luscombe, D. K., Eds.; Elsevier Science, New
York, 1995; Vol. 32, Chapter 4, pp 157–238; (b) Hayes, J.
D.; Wolf, C. R. Biochem. J. 1990, 272, 281; (c) Spratt, B.
G. Science 1994, 264, 388; (d) Nikaido, H. Science 1994,
264, 382.
3. (a) Russell, A. D. In Progress in Medicinal Chemistry Ellis,
G. P.; Luscombe, D. K.; Oxford, A. W., Eds.; Elsevier
Science, New York, 1998; Vol. 35, Chapter 4; (b) Fuchs,
T. M. Naturwissenschaften 1998, 85, 99; (c) Cunha, B. A.
Drugs of Today 1998, 34, 691.
d Streptococcus pneumoniae UC 9912.
e Enterococcus faecalis UC 9217.
f S. epidermidis UC 12084.
corresponding acid 28. The subsequent conversion was
performed as described for 7 affording 29 in modest
yield.
4. Harris, C. R.; Thorarensen, A. Curr. Med. Chem. 2004, 11,
2213.
5. Thorarensen, A.; Presley-Bodnar, A. L.; Marotti, K. R.;
Boyle, T. P.; Heckaman, C. L.; Bohanon, M. J.; Tomich,
P. K.; Zurenko, G. E.; Sweeney, M. T.; Yagi, B. H.
Bioorg. Med. Chem. Lett. 2001, 11, 1903.
6. Thorarensen, A.; Deibel, M. R., Jr.; Rohrer, D. C.;
Vosters, A. F.; Yem, A. W.; Marshall, V. D.; Lynn, J. C.;
Bohanon, M. J.; Tomich, P. K.; Zurenko, G. E.; Sweeney,
M. T.; Jensen, R. M.; Nielsen, J. W.; Seest, E. P.; Dolak,
L. A. Bioorg. Med. Lett. 2001, 11, 1355.
7. (a) Thorarensen, A.; Zurenko, G. E.; Sweeney, M. T.;
Marotti, K. R.; Boyle, T. P. Bioorg. Med. Chem. Lett.
2001, 11; (b) Deibel, M. R., Jr.; Bodnar, A. L.; Yem, A.
W.; Wolfe, C. L.; Heckaman, C. L.; Bohanon, M. J.;
Mathews, W. R.; Sweeney, M. T.; Zurenko, G. E.;
Marotti, K. R.; Boyle, T. P.; Thorarensen, A. Bioconjugate
Chem 2004, 15, 333.
8. Murray, R. W.; Melchior, E. P.; Hagadorn, J. C.
Antimicrob. Agents Chemother. 2001, 45, 1900.
9. 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, 6173.
10. Ruble, J. C. unpublished results.
11. WO 2004/018428 A1.
12. (a) Sakamoto, T.; Uchiyama, D.; Kondo, Y.; Yamanaka,
H. Chem. Pharm. Bull. 1993, 41, 478; (b) Sakamoto, T.;
Kondo, Y.; Uchiyama, D.; Yamanaka, H. Tetrahedron
1991, 47, 5111.
13. Humhrey, G. R.; Wright, S. H. B. J. Heterocyclic Chem.
1989, 26, 23.
14. Lipinski, C. A.; Lombardo, F.; Dominy, Beryl W.;
Feeney, P. J Adv. Drug Deliv. Rev. 1997, 23, 3.
Compound 29 demonstrated potent antibacterial activi-
ty against S. aureus both in presence and absence of ser-
um. Despite this potent activity, this compound was not
broad-spectrum and there were several strains of impor-
tant pathogens where 29 demonstrated sub-optimal
activity (Table 4). The improvement of spectrum of
activity against medically important pathogens required
further optimization in this series of novel antibacterial
agents.
In summary, we have prepared a diverse set of heterocy-
clic B-ring replacements. Several of these replacements
have potent antibacterial activity with significantly de-
creased affinity for HSA as measured by adding 10% ser-
um to the assay. This has been accomplished by
reducing the molecular weight by 40% from MW 625
for 1 to MW 380 for 29. Simultaneously, we were suc-
cessful in reducing the lipophilicity of the compounds.
Thus, we were successful in transforming a non-drug-
like screening hit, 1, into a lead series with promising
drug-like properties.14 Through the synthesis of 29 con-
taining a more optimal benzyl ether substituent, we have
demonstrated that heterocyclic B-ring replacements are
a viable option for further exploration. Several of these
heterocyclic B-ring replacements were subsequently
selected for further optimization, by variation of the
B-ring substitution and the anthranilic acid substitution
in an effort to improve the spectrum of antibacterial