Aryl urea analogs with broad-spectrum antibacterial activity

Article abstract of DOI:10.1016/j.bmcl.2004.08.059

The synthesis and evaluation of aryl urea analogs as broad-spectrum antibacterial agents is described. The preparation and evaluation of novel aryl urea analogs as broad-spectrum antibacterial agents is described. Numerous compounds showed low micromolar

Full text of DOI:10.1016/j.bmcl.2004.08.059

Bioorganic & Medicinal Chemistry Letters 14 (2004) 5569–5572
Aryl urea analogs with broad-spectrum antibacterial activity
Punit P. Seth,* Ray Ranken, Dale E. Robinson, Stephen A. Osgood, Lisa M. Risen,
Elizabeth L. Rodgers,ꢀ Michael T. Migawa, Elizabeth A. Jefferson and Eric E. Swayze
Ibis Therapeutics, A Division of Isis Pharmaceuticals, Inc., 1891 Rutherford Road, Carlsbad, CA 92008, USA
Received 2 August 2004; revised 20 August 2004; accepted 26 August 2004
Available online 17 September 2004
Abstract—The preparation and evaluation of novel aryl urea analogs as broad-spectrum antibacterial agents is described. Numerous
compounds showed low micromolar minimum inhibitory concentrations (MIC) against both Gram-positive and Gram-negative
bacteria. Selected analogs also exhibited in vivo efficacy in a lethal murine model of bacterial septicemia.
ꢀ 2004 Elsevier Ltd. All rights reserved.
The emergence of drug resistant pathogenic bacteria
continues to be a serious health problem worldwide.1
As a result, it has become critical to identify new struc-
tural classes of antibacterial agents to combat the grow-
ing threat of bacterial resistance.2 During screening of
new compound classes for antibacterial activity, thio-
urea 1 (Fig. 1) was identified as having good activity3
(MIC 3–6lM) against S. aureus in our assay.4 However,
thiourea 1 did not show any antibacterial activity in the
in vitro assay (MIC > 100lM) when tested in the pres-
ence of 4% bovine serum albumin (BSA). Since the loss
of activity in the presence of serum could be attributed
to the high lipophilicity of thiourea 1, we initiated a
study to identify alternate structures with reduced lipo-
philicity and equivalent or improved antibacterial activ-
ity. Here we report the initial SAR study around
thiourea 1 resulting in analogs with good broad-spec-
trum antibacterial activities as well as in vivo efficacy
in a lethal murine model of bacterial septicemia.
We first chose to modify the aryl ring of the thiourea
moiety without altering the 3,5-dibromo substitution
pattern on the N-benzyl substituent. Commercially
available mono-Cbz protected 1,3-diaminopropane was
reacted with Teoc-OSu, followed by removal of the Cbz
group by catalytic hydrogenation to provide mono-Teoc
protected 1,3-diaminopropane 2 (Scheme 1). Reduc-
tive amination of 2 with 3,5-dibromobenzaldehyde,
a, b
TeocHN
NH2
CbzHN
NH2
HCl
2
Br
f, g
c, d, e
H2N
N
Boc
or
3
h, g
Br
S
R
R
Br
Br
N
H
N
H
N
Cl
Cl
OR
H
S
4
Br
N
N
N
Br
H
H
H
O
1
Br
N
H
N
H
N
H
5
Figure 1.
Br
Scheme 1. Reagents and conditions: (a) Teoc-OSu, Et3N, CH2Cl2; (b)
10% Pd/C, H2; (c) 3,5-dibromobenzaldehyde, NaBH3CN, MeOH
(68%, over three steps); (d) Boc2O, CH2Cl2 (95%); (e) 1M TBAF, KF,
CH3CN, 40ꢁC, 10h (75%); (f) ArNCS, CH2Cl2; (g) 30% TFA/CH2Cl2;
(h) ArNCO, CH2Cl2.
Keywords: Aryl urea; Antibacterial; Broad spectrum.
*
Corresponding author. Tel.: +1 760 603 2587; fax: +1 760 603
4653; e-mail: pseth@isisph.com
ꢀ Undergraduate Medicinal Chemistry COOP student (University of
Ottawa, Canada).
0960-894X/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.08.059

5570
P. P. Seth et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5569–5572
protection of the secondary amine with Boc2O and
deprotection of the Teoc group using TBAF/KF pro-
vided amine 3. Subsequent reaction of 3 with commer-
cially available arylisothiocyanates or arylisocyanates
followed by deprotection of the Boc group provided
the corresponding thiourea 4 or urea 5 analogs, respec-
tively, in good overall yield (40–55% after reversed
phase HPLC purification). All final thiourea 4 and urea
5 analogs were tested for antibacterial activity as their
acetate salts.
F3C
O
a, b
BocHN
NH2
F3C
N
N
NH2
H
H
6
7
O
c
N
H
N
N
R
R
H
H
N
8
9
F3C
d
O
In the first SAR set, ꢀ30 thiourea/urea analogs with
electron donating and electron withdrawing groups at
the 2-, 3-, and 4-positions of the aryl ring were prepared
and evaluated. The objective was to identify substitu-
tions that would reduce hydrophobicity while main-
taining activity. Biological screening revealed that
substitution at the 2-position was not tolerated and led
to a substantial decrease in antibacterial activity (Table
1). Substitution at the 3- and 4-positions was tolerated
and optimal activity was seen with small non-polar
groups such as chloro and trifluoromethyl. From this
series urea 5d with a 4-CF3 group exhibited activity sim-
ilar to the lead thiourea 1. Urea analog 5d also had a
lower ClogP value5 (4.89) as compared to thiourea 1
(6.23). As a result we chose urea 5d as the new lead
structure for further optimization.
N
H
N
H
Scheme 2. Reagents: (a) 4-CF3-phenyl isocyanate, CH2Cl2; (b) 30%
TFA, CH2Cl2 (90% over two steps); (c) ArCHO, TMOF, NaBH3CN,
cat. AcOH, MeOH; (d) 40% HCHO, MeOH, cat. AcOH, NaBH3CN.
the final compounds were tested for antibacterial activ-
ity as their acetate salts.
Substitution at the 4-position of the benzylamino
group with a variety of hydrophobic substituents such as
t-butyl, phenyl, phenoxy, and OCF3 provided analogs
that showed good activity against S. aureus (Table 2, en-
tries 4–9). We were also surprised that a 4-dimethyl-
amino group also displayed some antibacterial activity
suggesting that it may be possible to substitute the
4-position with other hydrophilic groups (entry 10).
The best substitution pattern for the halogenated ben-
zylamino analogs was either 3,5-dibromo (8l) or 3,
4-dichloro (8q). 3,5-Dichloro-substitution resulted in a
slight loss of activity while 3,5-difluoro, 3,5-ditrifluoro-
methyl, and 3,5-dimethoxy analogs showed reduced
In the second analog set, we explored the SAR of the
benzylamino substituent. Commercially available N-
Boc-1,3-diaminopropane was reacted with 4-(trifluoro-
methyl)phenyl isocyanate followed by deprotection of
the Boc group to provide urea 7 (Scheme 2). Reductive
aminations with ꢀ40 aromatic aldehydes provided ureas
8 (40–55% after reversed phase HPLC purification). All
Table 1. In vitro antibacterial activity of thiourea/urea analogs
Table 2. In vitro antibacterial activity of benzylamino group modified
analogs
Entry
Compd
R
MIC (lM)4
S. aureus
Entry
Compd
R
MIC (lM)4 S. aureus
1
1
3,4-DiCl
2-F
2-Cl
3–6
1
8a
8b
H
50–100
12–25
6–12
3–6
2
4a
4b
4c
25–50
12–25
>100
12–25
6–12
6–12
6–12
25–50
12–25
6–12
12–25
12–25
6–12
12–25
12–25
25–50
25–50
6–12
3–6
2
2-OCF3
3-Ph
4-OCF3
3
3
8c
8d
4
2-OMe
3-F
3-Cl
4
5
4d
4e
5
8e
4-Ph
3–6
3–6
6
6
8f
8g
4-(40-OMe)-Ph
4-(20-OMe)-Ph
4-OPh
7
4f
3-Br
7
6–12
3–6
8
4g
4h
4i
3-CF3
3-OMe
4-F
8
8h
8i
9
9
4-t-Bu
3–6
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
8j
4-NMe2
4-NHCOMe
3,5-DiBr
3,5-DiCl
3,5-DiF
3,5-DiCF3
3,5-DiOMe
3,4-DiCl
2-OH-3,5-DiBr
2-OEt-3,5-DiBr
2-OEt-5-Br
3,5-DiBr
—
12–25
>100
3–6
4j
4-Cl
4-Br
8k
5d
4k
4l
4-I
8m
8n
6–12
50–100
50–100
50–100
3–6
4m
4n
4o
4p
4q
5a
5b
5c
4-CF3
4-NO2
4-CH3
4-CN
4-OMe
3-CF3
3,4-DiCl
4-Cl
8o
8p
8q
8r
>100
3–6
8s
8t
6–12
>100
3–6
6–12
3–6
9
5d
5e
4-CF3
4-OCF3
—
10a
10b
Linezolid
6–12
3–6
—
—
6–12
3–6
Linezolid

P. P. Seth et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5569–5572
5571
F3C
22–23). An interesting observation is that replacement
of the secondary amino group in the tether with a terti-
ary amine was permitted in case of urea 10b but not for
urea 9 (entry 21). It would appear that the methyl group
in 9 might alter the conformation such that the molecule
is not able to bind its target. In contrast, analog 10b may
be flexible enough to adopt the bioactive conformation
despite the methyl group on the tether amine.
O
O
Br
Br
N
N
N
H
H
H
8s
Br
F3C
O
O
N
N
N
H
H
R
10
A number of analogs were then evaluated against a
broader panel of Gram-positive and Gram-negative bac-
teria (Table 3). For the most part, urea analogs showed
low micromolar activity against S. aureus, S. pyogenes,
and E. faecalis. However, only ureas 8i and 8s showed
good activity against E. faecium. Most of the urea ana-
logs showed slightly reduced activities against E. coli but
good activity against K. pneumoniae. All compounds
showed weak activity against P. vulgaris and no activity
against P. aeruginosa.
Br
F3C
a
O
O
7
Br
N
H
N
H
N
O
H
Br
10a
OMs
11
Br
Br
F3C
O
O
b
Br
N
N
H
N
Selected urea analogs were also evaluated for activity in
the presence of 4% BSA. Most of the analogs tested
showed reduced activity in the presence of serum (Table
3). This was not completely surprising considering the
high ClogP values for these compounds (Table 3). The
greatest reduction in antibacterial activity in the pres-
ence of serum was observed for the halogenated analogs
5d, 8s and 10b. Ureas 8e and 8h gave the best activity in
the presence of serum (4-fold reduction).
H
10b
Br
Scheme 3. Reagents and conditions: (a) 11, CaCO3, DMF, 40ꢁC, 48h;
(b) 40% HCHO, MeOH cat. AcOH, NaBH3CN.
antibacterial activity (entries 12–17). Replacement of the
aryl ring with halogenated heteroaromatic rings was
also not well tolerated (data not shown). A few urea
analogs with substituted and unsubstituted phenolic
groups were also prepared and evaluated (entries 18–
20). Our intention was to incorporate groups that could
potentially provide us with a handle to introduce alkyl
chains with polar substituents. In this series, urea 8s
with a 2-ethoxy substituent showed activity comparable
to the lead urea 5d. Interestingly, analogs 10a–b with ex-
tended tethers (Scheme 3) were prepared and found to
be as active as the 2-ethoxy substituted analog 8s (entries
Analysis of the in vitro data suggests that the urea ana-
logs are binding in a very lipophilic binding pocket as
increasing hydrophobicity improves the activity. The
antibacterial activity of the urea analogs is also reduced
substantially in the presence of 4% bovine serum albu-
min. It is conceivable that the reduced activity is likely
the result of high serum protein binding of this com-
pound class. These observations are consistent with
the previous SAR studies carried out on other urea
Table 3. Broad-spectrum antibacterial activity of urea analogs
F3C
F3C
O
O
O
Br
N
N
N
N
N
N
R
H
H
H
H
H
10b
5d, 8s, 8e, 8h, 8i
Br
Bacteria
MIC (lM)
5d
3,5DiBr
8s
2-OEt-3,5-DiBr
8e
4-Ph
8h
4-OPh
8i
4-t-Bu
10b
Linezolid
S. aureus ATCC 13709
S. pyogenes ATCC 49399
E. faecalis ATCC 29212
E. faecium ATCC 6569
E. coli ATCC 25922
3–6
6–12
3–6
3–6
3–6
3–6
3–6
3–6
3–6
3–6
3–6
3–6
3–6
1.5–3
3–6
6–12
6–12
6–12
12–25
3–6
>100
>100
6–12
6–12
3–6
6–12
3–6
3–6
>100
12–25
6–12
>100
6–12
6–12
12–25
>100
12–25
>100
12–25
3–6
>100
>100
>100
12–25
>100
3–6
6–12
3–6
K. pneumoniae ATCC 13383
P. vulgaris ATCC 8427
P. aeruginosa ATCC 25416
S. aureus ATCC 13709 (+4% BSA)
MTT HUH-7—CC50 48h (lM)
ClogP
12–25
>100
25–50
NT50–100
4.89
12–25
>100
25–50
25–50
5.1
12–25
>100
12–25
>100
12–25
>100
25–50
12–25
12.5–25
4.91
50–100
NT
4.94
NT>100
4.77
5.3
0.58

5572
P. P. Seth et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5569–5572
Table 4. In vivo antibacterial activity of urea analogs13
official endorsement should be inferred. We thank Dr.
Richard H. Griffey for support of this project.
Entry
Compd
Dose (mg/kg)
Mice
alive/total
1
No drug
10b
—
0/10
4/10
7/10
6/10
6/10
4/10
3/10
3/10
3/10
4/10
2/10
0/10
10/10
References and notes
2
75 · 2
37.5 · 2
18.8 · 2
9.4 · 2
4.7 · 2
2.3 · 2
37.5 · 2
18.8 · 2
9.4 · 2
4.7 · 2
2.3 · 2
1 · 2
3
10b
10b
1. Chopra, I.; Hodgson, J.; Metcalf, B.; Poste, G. Antimic-
rob. Agents Chemother. 1997, 41, 497.
2. Wijkmans, J. C. H. M.; Beckett, R. P. DDT 2002, 7, 126.
3. Thiourea 1 showed activity comparable to Linezolid (3–
6lM), the positive control for our in vitro antibacterial
assays.
4. The MIC assays were carried out in a 150lL volume in
duplicate in 96-well clear flat-bottom plates. The bacterial
suspension from an overnight culture growth in the
appropriate medium was added to a solution of test
compound in 0.5% DMSO in water. Final bacterial
inoculum was approximate 103–104 CFU/well. The per-
centage growth of the bacteria in the test wells relative to
that observed for a control well containing no compound
4
5
10b
10b
6
7
10b
8e
8
9
8e
8e
10
11
12
13
8e
8e
Vancomycin
based antibacterial compounds.6–9 While increased
serum binding is not necessarily an undesirable property
for antibiotic compounds, it indicates that higher doses
of compound may be required to elicit a therapeutic re-
sponse in vivo10,11 This may narrow the therapeutic win-
dow and also raise the aqueous solubility threshold for
poorly soluble drugs. During the course of this study
we were able to prepare analogs with reduced lipophilic-
ity and improved aqueous solubility relative to thiourea
1.12 However, there is still scope for reducing the lipo-
philicity and serum protein binding as well as increasing
the potency of this compound class.
was determined by measuring absorbance at 595nm (A595
)
after 20–24h at 37ꢁC. The MIC was determined as a range
of concentrations where complete inhibition of growth
was observed at the higher concentration and the bacterial
cells were viable at the lower concentration. The bacterial
strains used for the assays include S. aureus ATCC 13709,
S. pyogenes ATCC 49399, E. faecalis ATCC 29212, E.
faecium ATCC 6569, E. coli ATCC 25922, K. pneumoniae
ATCC 13383, P. vulgaris ATCC 8427, P. aeruginosa
ATCC 25416.
5. ClogP values were calculated using ChemDraw Ultra 7.0
software.
6. Francisco, G. D.; Li, Z.; Albright, D.; Eudy, N. H.; Katz,
A. H.; Petersen, P. J.; Labthavikul, P.; Singh, G.; Yang,
Y.; Rasmussen, B. A.; Lin, Y.-I.; Mansour, T. S. Bioorg.
Med. Chem. Lett. 2004, 14, 235.
7. Kane, J. L., Jr.; Hirth, B. H.; Liang, B.; Gourlie, B. B.;
Nahill, S.; Barsomian, G. Bioorg. Med. Chem. Lett. 2003,
13, 4463.
8. Wilson, L. J.; Morris, T. W.; Wu, Q.; Renick, P. J.;
Parker, C. N.; Davis, M. C.; McKeever, H. D.; Hersh-
berger, P. M.; Switzer, A. G.; Shrum, G.; Sunder, S.;
Jones, D. R.; Soper, S. S.; Dobson, R. L. M.; Burt, T.;
Morand, K. L.; Stella, M. Bioorg. Med. Chem. Lett. 2001,
11, 1149.
9. Proctor, R. A.; Dalal, S. C.; Kahl, B.; Brar, D.; Peters, G.;
Nichols, W. W. Antimicrob. Agents Chemother. 2002, 46,
2333.
Lastly, to assess the in vivo efficacy of this class, analogs
8e and 10b were advanced for testing in a lethal murine
model of bacterial infection (Table 4).13 During the in
vivo evaluation, urea 10b showed efficacy in the murine
model with 6/10 mice surviving in the 9.4mg/kg (dosed 1
and 3h postinfection) dose group and 7/10 mice surviv-
ing in the 37.5mg/kg (dosed 1 and 3h postinfection)
dose group. Urea 10b, however, was unable to rescue
all mice from infection at the higher doses that were
evaluated. Presumably the toxicity of the drug becomes
an issue at higher drug concentrations.13 Both urea ana-
logs 8e and 10b were not as potent as the vancomycin
positive control, which rescued all the animals at a dose
of 1mg/kg (dosed 1 and 3h postinfection).
10. Bergogne-Berezin, E. Clin. Pharmacokinet. 2002, 41, 741.
11. Actor, P.; Uri, J. V.; Zajac, I.; Guarini, J. R.; Phillips, L.;
Pitkin, D. H.; Berges, D. A.; Dunn, G. L.; Hoover, J. R.
E.; Weisbach, J. A. Antimicrob. Agents Chemother. 1978,
13, 784.
In conclusion, we have described the preparation and
evaluation of novel aryl urea analogs as broad-spectrum
antibacterial agents. Numerous compounds showed low
micromolar activity against both Gram-positive and
Gram-negative bacteria. Selected analogs also exhibited
in vivo efficacy in a lethal murine model (S. aureus) of
bacterial infection.
12. Solubility of 8eÆlactate salt ꢀ8mg/mL, solubility of
10bÆHCl salt >10mg/mL.
13. Mouse protection assay: Ten mice/dose group (ICR-CD-1
female mice 18–20g, Charles River) were infected with a
lethal dose (106 CFU/mouse) of S. aureus (ATCC 13709)
suspended in 7.5% hog Gastric Mucin (IP). The infected
animals were treated at 1 and 3h postinfection with either
compound 8eÆlactate salt, from 37.5mg/kg down to 2.3mg/
kg or compound 10bÆhydrochloride salt, from 75mg/kg
down to 2.3mg/kg (0.1mL/mouse, IV). The positive
control drug was vancomycin 1mg/kg dosed twice at 1
and 3h postinfection. The animals were observed for one
week and mortality was calculated. Ureas 8e and 10b were
toxic when dosed at concentrations above 37.5 and
150mg/kg, respectively.
Acknowledgements
Financial support thanks to USAMRID DAMD717-02-
2-0023. The US Army Medical Research Acquisition
Activity, 820 Chandler Street, Fort Detrick, MD
21702-5014 is the awarding and administering office.
The content of this manuscript does not necessarily re-
flect the position or policy of the Government, and no