Novel 3-chlorooxazolidin-2-ones as antimicrobial agents

Article abstract of DOI:10.1016/j.bmcl.2011.03.036

Antimicrobial resistance against many known therapeutics is on the rise. We examined derivatives of 3-chlorooxazolidin-2-one 1a (X = H) as antibacterial and antifungal agents. The key findings were that the activity and apparent in vitro cytotoxicity coul

Full text of DOI:10.1016/j.bmcl.2011.03.036

Bioorganic & Medicinal Chemistry Letters 21 (2011) 3025–3028
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Novel 3-chlorooxazolidin-2-ones as antimicrobial agents
Timothy P. Shiau, Eric D. Turtle, Charles Francavilla, Nichole J. Alvarez, Meghan Zuck, Lisa Friedman,
⇑
Donogh J. R. O’Mahony, Eddy Low, Mark B. Anderson, Ramin (Ron) Najafi, Rakesh K. Jain
Novabay Pharmaceuticals, 5980 Horton St. Suite 550, Emeryville, CA 94608, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Antimicrobial resistance against many known therapeutics is on the rise. We examined derivatives of 3-
chlorooxazolidin-2-one 1a (X = H) as antibacterial and antifungal agents. The key findings were that the
activity and apparent in vitro cytotoxicity could be controlled by the substitution of charged solubilizers
at the 4- and 5- positions. These changes both significantly increase the antifungal potency and decrease
cytotoxicity. Particularly effective were trialkylammonium groups which led to 400- to 600-fold
increases in the antifungal therapeutic index when compared to their unsubstituted counterparts.
Ó 2011 Elsevier Ltd. All rights reserved.
Received 13 January 2011
Accepted 10 March 2011
Available online 21 March 2011
Keywords:
Antibacterial
Antifungal
Antimicrobial
Oxazolidinone
Topical
Background and significance: Antimicrobial resistance is a rap-
idly evolving issue for which we felt could be addressed through
the development of new organic chloroamines such as N-chloro-
oxazolidinone 1a. While this compound has broad-spectrum anti-
microbial activity it has generated limited interest beyond
disinfecting water1 or as an additive in cleaning agents.2 Because
chloroamines can rapidly react with a variety of targets which
are essential to the invasive microbe, the potential for them to de-
velop effective resistance to this oxidant is low. Our interest in
these compounds as therapeutic agents for the treatment of topical
infections such as conjunctivitis, otitis, wound care, impetigo, and
skin and soft tissue infections (SSTI) have led us to more fully ex-
plore the structure–activity relationship (SAR) and structure-sta-
bility relationship (SSR) of N-chlorooxazolidinones.
Our mechanism of action experiments with N,N-dichloroamines
(published separately) suggests that chloroamines act by oxidizing
sulfur in methionines and cysteines, denaturing proteins. Based on
our understanding, we anticipated N-chlorooxazolidinones to have
N–Cl bonds that could be electronically controlled by the proxi-
mate functionalities, giving them more versatility, selectivity and
other advantageous properties over simple inorganic chlorinating
agents.3 Moreover, they should exhibit efficacy and toxicity pro-
files which are chemically distinct from other therapeutic chloro-
amines such as N-chlorotaurine. Because the antimicrobial
activity and oxidation potential is sensitive to the electronic and
steric structure around the N–Cl bond, we systematically explored
the antimicrobial activity by altering the electrophilic nature of the
organic structure, using the heterocycle as a targeting moiety. We
were also very interested in determining the role played by chang-
ing the hydrophilic or lipophilic nature of these groups on the
properties and activities of the new heterocycles.
In this paper we describe a systematic chemical study of various
3-chlorooxazolidin-2-one derivatives to elucidate the effect of
water-solubilizing and chemical structural features on the physical
properties of the resulting analogs as well as their in vitro antimi-
crobial activity against Gram-positive bacteria, Gram-negative bac-
teria, and fungi. These derivatives of 3-chloro-oxazolidin-2-one
showed remarkable microbicidal activity against all strains tested
while maintaining a suitable safety profile against mammalian
cells.
Materials and methods: 4-Methyl-4-sulfonyl analogs were syn-
thesized from the chloride 3.4 Displacement of the chloride with
sodium sulfite, followed by N-chlorination of the oxazolidinone
ring, gave sulfonate-solubilized 1b. Similar displacement with eth-
anedithiol gave a mixture of disulfides 5a (n = 1) and 5b (n = 2).
Oxidation of the mixture gave sulfone-extended sulfonate 6, and
N-chlorination yielded 1c. Displacement of the alkyl chloride with
potassium thioacetate gave thioacetate 7; alkaline hydrolysis of
the thioester, alkylation followed by oxidation and N-chlorination
gave other sulfone-extended solubilizers such as carboxylates 1d
(Scheme 1).
The 4-methyl-4-aminomethyl analogs were also synthesized
from 3, but displacement of the chloride was not straightforward.
Reaction with dialkylamines gave two isomeric products which
were separable through silica gel chromatography, namely oxazo-
lidinones 9a–d and 2-aminodihydrooxazoles 10a–d. While the
reaction with dimethylamine (entry a) has been previously re-
⇑
Corresponding author. Tel.: +1 510 899 8871.
E-mail address: rjain@novabaypharma.com (R.K. Jain).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.03.036

3026
T. P. Shiau et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3025–3028
ported in the literature5, the chemical characterization was limited
to IR and elemental analysis, neither of which could determine if
the reported sample was a pure compound or a mixture of regioi-
somers (Scheme 2).
The structures of 9a and 10a were established by HMBC exper-
iments. A crosspeak between the dimethyl signal and the imidate
carbon atom in compound 10a verified that the lower Rf isomer
was the dihydrooxazole 10a rather than a carbamate. Likewise, a
crosspeak between the dimethyl signal and an alkyl carbon atom
in compound 9a verified that the dimethylamino group was linked
through the methylene chain (Fig. 1).
O
H
H
O
H
H
HN
H
O
N
N
N
OH
10a
H
9a
Figure 1. Non-trivial crosspeaks observed in the HMBC spectra of 9a and 10a.
10a
to
O
Cl
HO
N
O
to 10a
We propose that compound 10a comes from addition of the
dimethylamine into the carbamate carbonyl rather than through
the expected SN2 displacement of chloride 3.6 The urea intermedi-
ate can then cyclize to give the isourea7 (Scheme 3).
O
HN
NH
O
N
N
H
N
Cl
to 9a
OH
Scheme 3. Hypothesized mechanism of rearrangement to give 10a.
Addition of iodide to the reaction mixture to form the alkyl io-
dide in situ did improve the isolated yield of the desired oxazolid-
inone, but did not reduce the isolated yield of the dihydrooxazole.
In addition, the isolated yields of oxazolidinone from different sec-
ondary amines was reasonably constant. However, the reaction
with tert-butylamine caused decomposition of the starting mate-
rial with neither the oxazolidinone nor the dihydrooxazole isolated
(Table 1).
Oxazolidinones 9a and c were further alkylated to give ammo-
nium-solubilized oxazolidinones 11a–d and N-chlorination gave
1e–h (Scheme 2).
Synthesis of the 5-sulfonylmethyl analog was accomplished
through the sequence shown in Scheme 4. A Henry reaction be-
tween 2-nitropropane and benzyloxyacetaldehyde gave nitro alco-
Table 1
Reaction conditions to form oxazolidinones 9a–d and dihydrooxazoles 10a–d
Entry
R1, R2
Additive
Yield (9) (%)
Yield (10) (%)
a
a
b
b
c
–Me, –Me
–Me, –Me
–Et, –Et
None
NaI
none
NaI
NaI
NaI
30
40
3
39
35
49
Trace
41
41
19
51
60
36
Trace
–Et, –Et
–(CH2)4–
–(CH2)5–
–H, –tBu
d
e
None
hol 12, which was reduced with hydrogen to give amino alcohol
13. Surprisingly, the benzyloxy group was not cleaved under these
conditions. Cyclization to oxazolidinone 14, debenzylation with a
palladium catalyst, followed by transformation of the alcohol to
the sulfonic acid 17, through a thioacetate intermediate 16, pro-
O
O
Cl
b
HN
HN
O
N
O
a
SO3Na
SO3Na
1b
4
O
O
O
O
Cl
b
N
O
HN
O
O
O
O
HN
H
O
O
c
d
O
a
b
e
SO3H
CO2H
S
SO3H
OBn
S
5a-b
O
S
Cl
3
S
O2N
d
OBn
H2N
HN
OBn
H
n
O
13OH
1O2 H
O
O1c
6
O
O
O
e
Cl
c
HN
O
O
N
O
HN
f-h
b
O
O
HN
HN
O
OBn
OH
O
SAc
SAc
7
S
CO2H
S
O O
14
15
16
3
3
1d
O O
8
O
O
g
f
Cl
O
HN
N
Scheme 1. Synthesis of 4-substituted anionic oxazolidinones 1b–d. Reagents and
conditions: (a) Na2SO3, DMF/H2O, 50 °C, quant.; (b) t-BuOCl, MeOH, 0 °C, 40–90%
over 1–2 steps; (c) EDT, DMF, 90 °C; (d) mCPBA, DCM, 0 °C, 33% over two steps; (e)
KSAc, DMF, 90 °C, 46%; (f) NaOH, MeOH, rt; (g) Br(CH2)3CO2Et, DMF, 80 °C, 71% over
two steps; (h) LiOH, MeOH/H2O, rt, 18%.
SO3H
SO3H
17
2a
Scheme 4. Synthesis of 5-substituted anionic oxazolidinone derivative 2a.
Reagents and conditions: (a) TMG, THF, rt, 84%; (b) H2, Raney Ni, MeOH, 89%; (c)
triphosgene, NEt3, DCM, 0 °C, 78%; (d) H2, Pd/C, MeOH/THF, 99%; (e) PPh3, DIAD,
HSAc, THF, 0 °C, 87%; (f) H2O2, HCO2H, rt, 90%; (g) t-BuOCl, MeOH, 0 °C, 53%.
O
O
NR1R2
O
a
HN
HN
O
O
N
+
OH
N
OH
N
O2N
H2N
NO2
c
NO2
O
a
b
Cl
NR1R2
9a-d
OH
3
10a-d
18
19
20
21
b
O
O
O
O
O
d
e
f
Cl
O
HN
N
O
O
HN
Cl
c
O
O
N
HN
Cl-
+ Cl-
+ Cl-
N
N
N
N+R1R2R3
N+R1R2R3
22
23
2b
Cl-
11a-d
1e-h
Scheme 5. Synthesis of 5-substituted cationic oxazolidinone derivative 2b.
Reagents and conditions: (a) mCPBA, DCM, 87%; (b) HNMe2, THF/H2O, 42%; (c)
H2, Raney Ni, MeOH, 76%; (d) triphosgene, NEt3, DCM, 0 °C, 18%; (e) MeI, MeOH;
Ag2O; HCl, 71%; (f) t-BuOCl, MeOH, 0 °C, 48%.
Scheme 2. Synthesis of cationic oxazolidinones 1e–h. Reagents and conditions: (a)
HNR1R2, NaI, THF, 85 °C, 35–49%; (b) R3I, MeOH, 25–70 °C; Ag2O; HCl; (c) t-BuOCl,
MeOH, 15–48% over two steps.

T. P. Shiau et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3025–3028
3027
vided the 5-substituted sulfonic acid, which was N-chlorinated to
O
O
O
Cl
give oxazolidinone 2a.
Cl
Cl
N
O
N
O
N
O
Synthesis of the 5-aminomethyl derivative was accomplished
through epoxidation of alkene 188, followed by ring opening with
dimethylamine, reduction of the nitro group, and cyclization to
the oxazolidinone. Alkylation and N-chlorination afforded com-
pound 2b (Scheme 5).
Two des-methyl analogs of 2c and d were synthesized through
the route in Scheme 5. Starting with commercially available chlo-
ride 24, the alkyl chloride was either replaced with a sulfonate
(26) through the thioacetate 25, or by displacement with pyridine
to give its pyridinium analog 27 (Scheme 6). Unfortunately, N-chlo-
rinated analogs 2c–d were not sufficiently stable to test in an MBC
assay. It is our view that the gem-dimethyl group has a stabilizing
effect on the N-chlorooxazolidinones by either blocking dehydro-
chlorination (similar to the stabilizing effect of a gem-dimethyl
group on N,N-dichlorotaurine9) or oxazolidinone hydrolysis. Fur-
ther studies are expected to elucidate the importance of the influ-
ence of the 4,4-disubstitution on stability.
Increasing the steric bulk at the 4-position afforded better
chemical stability than 2c–d (data not shown). The spirocycle
2810 was alkylated with methyl iodide to give the spirocyclic
ammonium derivative 29; N-chlorination gave compound 2e
which showed much better stability than did 2c–d (Scheme 7).
In vitro antimicrobial activity of analogs 1a–h and 2a–e (Fig. 2)
was conducted against Escherichia coli ATCC 25922, Staphylococcus
aureus ATCC 29213, and Candida albicans ATCC 10231 using a mod-
ified CLSI M26-A protocol where Mueller–Hinton broth was re-
placed with phosphate-buffered saline (pH 7) and the residence
time of the compound was reduced to 1 h to generate minimum
bactericidal concentrations (MBC) and minimum fungicidal con-
centrations (MFC). In vitro cytotoxicity was measured against
L929 (mouse fibroblast) cells after exposure to the compounds
for 1 h. After exposure, the test article was removed and cells were
incubated overnight. Viability was measured using the Dojindo
Cell Proliferation Assay (Dojindo Laboratories, Mashiki, Kumamoto
prefecture, Japan).
R
X
Cl-
R
X
N+
Entry
R
X
2e
Entry
X
2a Me SO3Na
1a
1b
1c
H
SO3Na
Me NMe3+Cl-
2b
SO2(CH2)2SO3Na
2c
H
H
SO3Na
NC5H5+Cl-
1d SO2(CH2)3CO2H
2d
1e NMe3+Cl-
NMe2Et+Cl-
NMe2Pr+Cl-
1f
1g
1h N(Me)(CH2)4+Cl-
Figure 2. Reported compounds.
to 50-fold reduction in cytotoxicity results. Also, up to 64-fold
increase in antifungal activity is observed, and the antibacterial
activity was retained with this added functionality (Table 2, en-
try 1g).
Compared to parent oxazolidinone 1a, sulfonate-solubilized 1b
showed a 15-fold reduction in in vitro cytotoxicity with a corre-
sponding reduction of activity reflected against bacteria and fungi,
with 2- to 256-fold reductions in antimicrobial activity. Somewhat
surprisingly, extending the water-solubilizing group with a sulfone
(1c–d) increased the antimicrobial activity relative to 1b without
significantly affecting the CT50
.
Table 2
Antimicrobial activities, cytotoxicities, and calculated in vitro therapeutic indices
against bacteria and fungi for reported compounds
Entry E. coli S. aureus C. albicans CT50
TIE.
TIC.
albicans
coli./S. aureus
1a
1b
1c
1d
1e
1f
1g
1h
2a
2b
2e
1
4
512
1024
128
>1024
16
8
8
16
1024
64
0.07
0.8
0.6
5.2
1.4
8.5
14
0.02
0.25
1.5
0
n.d.
28
31
19
1.0
2.7
0.70
256
16
16
2
128
32
8
0.52
n.d.
4
4
2
4
n.d.
2
2
4
0.86⁄ 78
0.9⁄
122
1.13⁄ 76
Antibacterial therapeutic indices were calculated based on the
geometric mean of the MBCs against E. coli and S. aureus, antifungal
therapeutic indices were calculated based on the MFC against C.
albicans.
256
16
16
512
32
32
4.3
2.9
7.6
7.9
0.69
256
0.7⁄
Antimicrobial activities (MBCs and MFCs) are given in lg/ml, CT50s are given in mM,
n.d. = not determined, ⁄ = Testing solution was pH 4 rather than pH 7. Results for
other compounds (data not shown) show that CT50 values are comparable at the
two pHs.
Results and discussion: We showed that by appending
a
water-solubilizing group to either the 4- or 5- position, a 10-
O
O
O
Most remarkably, cationic groups (1e–h) showed greatly im-
proved antifungal activity while the cytotoxicity was only slightly
worse than sulfonate 1b. The most potent antifungal compounds,
1f–g, were 64-fold better than the parent 1a.
This trend was also shown for 5-substituted oxazolidinones
2a–b, e. Compared to parent 1a, the 5-sulfonylmethyl derivative
2a showed significantly reduced cytotoxicity, but with the cost of
reduced antimicrobial activity. Replacing the sulfonate solubilizer
with a cationic group (2b, e) yielded more potent antimicrobial
activity (16-fold across both bacteria and fungi) at only a 7-fold
cost in cytotoxicity relative to the sulfonate.
c
b
Cl
O
O
HN
HN
c
N O
a
O
SAc
SO3H
O
SO3H
25
O
26
2c
O
HN
d
Cl-
N+
Cl
Cl
Cl-
N+
O
HN
O
N
24
27
2d
Scheme 6. Synthesis of 4,4-unsubstituted oxazolidinones 2c–d. Reagents and
conditions: (a) KSAc, DMF, 50 °C, 53%; (b) H2O2, HCO2H, quant.; (c) t-BuOCl, MeOH,
0 °C, 35–40%; (d) pyridine, 115 °C, 36%.
Therapeutic indices (ratios of activity to toxicity) were calcu-
lated for bacteria as well as fungi. The bacterial therapeutic index
(mathematically defined as CT50/MBC), increased 0- to 20-fold
across the series, while the fungal therapeutic index (CT50/MFC) in-
creased 12- to 600-fold across the series (with the exception of
compound 1d which did not show antifungal activity). This
remarkable transformation of compound 1a (whose threshold for
antifungal activity is 50-fold higher than its threshold for cytotox-
icity) to compound 1g (whose threshold for antifungal activity is
O
O
O
Cl
b
O
N
a
O
HN
O
HN
N+
Cl-
N+
Cl-
NH
HCl
28
29
2e
Scheme 7. Synthesis of spirocycle oxazolidinone 2e. Reagents and conditions: (a)
MeI, Cs2CO3, DMF; Ag2O; HCl, 92%; (b) t-BuOCl, MeOH, 0 °C, 13%.

3028
T. P. Shiau et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3025–3028
now 30-fold lower than its threshold for cytotoxicity) validates this
approach for making safer, more potent antimicrobial agents. Re-
search is ongoing to further elucidate the SXR by careful control
of the physiochemical properties. These data together with our
increasing understanding of the mechanism of action of these
and similar agents should result in new safe and effective human
and animal antimicrobials.
Acknowledgements
The authors wish to acknowledge Alcon Laboratories, Inc. for
their financial support, and our thanks to Dr. Joaquin Randle at
Acorn NMR for helpful discussions with respect to the 2-D NMR
experiments.
Conclusions: Changing the electronic structure around the N–Cl
bond greatly affects the antimicrobial activity as well as the cyto-
toxicity. This offers the opportunity to better control the processes
involved. This finding also supports our general assertion that
these molecules are not non-specific chlorinating agents, but
rather that they are capable of hitting multiple chemical targets
with some degree of selectivity which ultimately results in micro-
bial death.
The use of charged functionality for water solubilization of
N-chlorooxazolidinones is a simple and effective tool for reducing
the cytotoxicity of powerful antimicrobial agents such as 4,4-di-
methyl-3-chlorooxazolidin-2-one. Addition of trialkylammonium
solubilizers at either the 4- or 5- positions both increases the anti-
fungal activity as well as reduces cytotoxicity as compared to
unsolubilized oxazolidinones, thereby greatly increasing their
therapeutic index.
References and notes
1. Williams, D. E.; Worley, S. D.; Wheatley, W. B.; Swango, L. J. Appl. Environ.
Microbiol. 1985, 49, 637–643.
2. Rathore, O. European Patent No. 1,327,893 B1.
3. Williams, D. E.; Elder, E. D.; Worley, S. D. Appl. Environ. Microbiol. 1988, 54,
2583–2585.
4. Homeyer, A.H. United States Patent No. 2,399,118.
5. Rosnati, V.; Misiti, D. Tetrahedron 1960, 9, 175–182.
6. (a) Najer, k. et al Bull. Soc. Chim. Fr. 1957, 1069–1071; (b) Rossi, F. M.; Powers, E.
T.; Yoon, R.; Rosenberg, L.; Meinwald, J. Tetrahedron 1996, 52, 10279–10286.
7. (a) Johnston, T. P.; McCaleb, G. S.; Opliger, P. S.; Montgomery, J. A. J. Med. Chem.
1966, 9, 892–911; (b) Johnston, T. P.; McCaleb, G. S.; Opliger, P. S.;
Montgomery, J. A. J. Liebigs Ann. Chem. 1974, 593–607.
8. Tanikaga, R.; Sugihara, H.; Tanaka, K.; Kzuhiko, K.; Kaji, A. Synthesis 1977, 299–
301.
9. Wang, L.; Khosrovi, B.; Najafi, M. R. Tetrahedron Lett. 2008, 49, 2193–2195.
10. Smith, P. W.; Cooper, A. W.; Bell, R.; Beresford, I. J.; Gore, P. M.; McElroy, A. B.;
Pritchard, J. M.; Saez, V.; Taylor, N. R.; Sheldrick, R. L. J. Med. Chem. 1995, 38,
3772–3779.