Conformational constraint in oxazolidinone antibacterials. Part 2: Synthesis and structure-activity studies of oxa-, aza-, and thiabicyclo[3.1.0] hexylphenyl oxazolidinones

Article abstract of DOI:10.1016/j.bmcl.2005.11.093

A new class of oxazolidinone antibacterials incorporating oxygen-, nitrogen-, or sulfur-containing heterobicyclic C-rings is described. The in vitro potency and in vivo efficacy of these conformationally constrained oxazolidinone analogs are discussed.

Full text of DOI:10.1016/j.bmcl.2005.11.093

Bioorganic & Medicinal Chemistry Letters 16 (2006) 1126–1129
Conformational constraint in oxazolidinone antibacterials. Part 2:
Synthesis and structure–activity studies of oxa-, aza-, and
thiabicyclo[3.1.0]hexylphenyl oxazolidinones
Adam R. Renslo,a,* Hongwu Gao,a Priyadarshini Jaishankar,a Revathy Venkatachalam,a
a
a
b
b
´
Marcela Gomez, Johanne Blais, Michael Huband, J. V. N. Vara Prasad and
Mikhail F. Gordeeva
aPfizer Global Research and Development, 34790 Ardentech Ct., Fremont, CA 94555, USA
bPfizer Global Research and Development, 2800 Plymouth Rd., Ann Arbor, MI 48105, USA
Received 8 November 2005; revised 23 November 2005; accepted 28 November 2005
Available online 4 January 2006
Abstract—A new class of oxazolidinone antibacterials incorporating oxygen-, nitrogen-, or sulfur-containing heterobicyclic C-rings
is described. The in vitro potency and in vivo efficacy of these conformationally constrained oxazolidinone analogs are discussed.
Ó 2005 Elsevier Ltd. All rights reserved.
The oxazolidinones are a promising new class of totally
synthetic antibacterial protein synthesis inhibitors.1
While they share with other antimicrobials a ribosomal
target, the oxazolidinones bind in a distinct region of
23S rRNA near the peptidyl transferase center2 and do
not exhibit significant cross-resistance with the existing
classes of antibacterials. Linezolid (1), the first oxazolid-
inone to receive regulatory approval, has become an
important clinical option in the treatment of serious
Gram-positive bacterial infections, including those
caused by multi-drug resistant pathogens such as
MRSA and VRE.
Other oxazolidinones that have advanced into clinical
trials include the piperazine analog eperezolid (2) and
the thiomorpholine analog PNU-288034 (3). One focus
of our early work in this field included the evaluation
of conformationally constrained analogs of 1–3 in which
the aliphatic C-ring is replaced by a rigid bicy-
clo[3.1.0]hexane ring system. Our initial efforts in this
direction included the design and synthesis of (azabicy-
clo[3.1.0]hexyl)phenyloxazolidinones in which the bicy-
clic ring is connected to the aromatic B-ring via the
pyrrolidine nitrogen atom. Many of these analogs exhib-
ited enhanced antibacterial potency and spectrum,
including activity against clinically relevant fastidious
Gram-negative pathogens.3 This initial success led us
to consider a second generation of bicyclic analogs in
which the cyclopropyl moiety is joined to the aromatic
B-ring. Significantly, this structural modification would
allow the preparation of thia-, oxa-, and azabicyclic ana-
logs (see above, Z = S, N–R, and O) that closely mimic
the steric nature of the unconstrained C-rings in the pro-
genitor oxazolidinones 1–3. In this letter, we describe the
synthesis and antibacterial activities (in vitro and in vivo)
of this new class of oxazolidinones.
Z
Z
F
F
C N
X
O
O
B
H
X
N
N
A O
O
NHAc
1 Z = O; X = H linezolid
NHAc
2 Z = N-C(O)CH2OH; X = H eperezolid
3 Z = SO2; X = F PNU-288034
The bicyclic analogs described herein were synthesized
as shown in Schemes 1–5. All three types of bicyclic
C-rings (e.g., 6–8) were prepared from the key diol inter-
mediates 4a–c (Scheme 1). We recently described a novel
synthesis of diols 4a–c (and 6–8) employing as a key step
Keywords: Oxazolidinone; Antibacterials.
*
Corresponding author. Tel.: +1 510 290 6171; fax: +1 510 653
8009; e-mail: renslo@mindspring.com
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.11.093

A. R. Renslo et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1126–1129
1127
OR
R
a, b
N
F
7b
Z
X
X
O
H
N-acyl analogs
H
Y
RO
a
N
H
Y
O
NHCbz
NHCbz
10b
R = C(O)CH2OH
R = CHO
d
NHAc
16a
16b
16c
16d
16e
16f
16g
16h
16i
8a-c (Z = O)
4a-c (R = H)
R = CN
R = C(O)NH2
15 (R = H)
b or c
c
6a-c (Z = S)
7a-c (Z = NAr)
5a-c (R = Ms)
16
R = C(O)CH2NH2
R = C(O)CH2NHAc
R = C(O)CH2CN
R = C(O)OCH3
R = C(O)CHF2
R = C(O)C(CH3)2OH
N-alkyl analogs
a X, Y = H; b X = F; Y = H; c X, Y = F
16j R = CH2CH2CN
16k R = CH2CH2F
16l R = CH2CH2OH
Scheme 1. Reagents and conditions: (a) Ms2O, Et3N, CH2Cl2; (b)
Na2S, DMSO, 79–88%; (c) NH2Ar (Ar = 4-MeOC6H4CH2–), rt, 63–
92%; (d) 2 equiv n-BuLi, THF, 1.2 equiv MsCl, À78 °C; then 1.2 equiv
n-BuLi, 37–47%.
Scheme 4. Reagents: (a) 3.0 equiv LiO-t-Bu, 2.0 equiv (S)-
ClCH2CH(OAc)CH2NHAc, 2 equiv MeOH, DMF, 64%; (b) H2,
Pd(OH)2/C, MeOH, EtOAc, 95%; (c) for 16d–i, RCOCl, DIEA,
DMF or RCOOH, EDCI, HOBT, DMF, 41–74%; for 16j–l,
RCH2CH2I or CH2CHCN, 16–56%; for 16b CNBr, 41%; for 16c
TMSNCO, 40%; for 16a, HCOOH, Ac2O, 40%.
O
S
X
O
a, b
6a-c
O
H
Y
X
a, b,
or c
N
X, R = H
X = CN; R = H
X = CN; R = CH3
17a
17b
17c
N
R
R
O
N
H
N
9a-c
15
NHAc
O
H
HO
a, c-e
N
X
N
N
N
7a-c
8a-c
d, e
18a
18b
18c
O
R = H
R = 2-CH3
R = 3-CH3
H
Y
N
N
16b
N
O
10a-c
H
NHAc
Scheme
5. Reagents
and
conditions:
(a)
for
17a,
O
X
CbzNHC(@NCbz)SMe, AgOTf, Et3N, then H2, Pd/C, 95%; (b) for
17b, NaN(CN)2, HCl, n-BuOH, reflux, 36%; (c) for 17c, MeNHC(@N–
CN)SMe, AgOTf, Et3N, 33%; (d) NaN3, NH4Cl, DMF, 100 °C, 54%;
(e) MeI, DIEA, DMF (for 18b,c), 30%.
a
O
H
Y
11a-c
N
O
NHAc
the intramolecular cyclopropanation reaction of dia-
zoacetates.4 The conversion of diols 4a–c to bicyclic
intermediates 6–8 is summarized in Scheme 1. Thia-
and azabicyclic compounds 6a–c and 7a–c were pre-
pared in two steps from 4a–c via initial formation of
the bis-mesylates 5a–c followed by reaction with sodium
sulfide in DMSO (for 6a–c) or with neat 4-methoxyben-
zylamine (for 7a–c). The oxabicyclic intermediates 8a–c
were prepared from 4a–c using a one-pot mesylation–cy-
clization reaction.
Scheme 2. Reagents: (a) 3.0 equiv LiO-t-Bu, 2.0 equiv (S)-
ClCH2CH(OAc)CH2NHAc, 2 equiv MeOH, DMF, 41–71%; (b)
AcOOH, THF, 82–92%; (c) for 10b,c: H2, Pd(OH)2/C, MeOH, EtOAc,
for 10a: ClCO2CH(Cl)CH3, MeOH; (d) BnOCH2COCl, Et3N,
CH2Cl2, 74% for two steps; (e) H2, Pd/C, MeOH, CH2Cl2, 95%.
S
F
a
O
6b or 6c
H
X
N
O
Scheme 2 illustrates the synthesis of oxazolidinone ana-
logs with the privileged acetamidomethyl C-5 side chain.
The reaction of thia-, aza-, and oxabicyclic aniline inter-
mediates 6–8 with (1S)-2-(acetylamino)-1-(chlorometh-
yl)ethyl acetate and lithium tert-butoxide in DMF
furnished the desired oxazolidinones (e.g., 9a–c to
11a–c) in a single synthetic step.5 For the thiabicyclic
analogs, a subsequent sulfur oxidation step provided
the sulfone analogs 9a–c. For the azabicyclic analogs,
the 4-methoxybenzyl group was removed via hydrogen-
olysis and the hydroxyacetamide side chain installed in
two steps (see Scheme 2) to provide the desired analogs
10a–c.
12b-c
NHBoc
b-d
F
O
S
O
13/14a
13/14b
13/14c
13/14d
13/14e
13f
R = CH2CH3
R = CH2OH
R = CH2CN
R = CF2CH3
O
H
X
N
O
NH
R = cyclopropyl
R = cyclobutyl
13 (X = H)
14 (X = F)
R
O
Scheme 3. Reagents: (a) 2.5 equiv LiO-t-Bu, 1.3 equiv (S)-
ClCH2CH(OH)CH2NHBoc, DMF, 73–81%; (b), TFA, CH2Cl2; (c)
RCOCl, DIEA, DMF or RCOOH, EDCI, HOBT, DMF; for 13/14b:
AcOCH2COCl, DIEA DMF, then aq LiOH, THF, 49–84% for two
steps; (d) AcOOH, THF, 82–96%.
The synthesis of thiabicyclic analogs with atypical (i.e.,
non-acetamidomethyl) C-5 side chains is illustrated in

1128
A. R. Renslo et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1126–1129
Table 2. Minimum inhibitory concentration (MIC) values for thiabi-
cyclic analogs bearing various C-5 amide side chains
Scheme 3. Reaction of 6b (mono-fluoro B-ring) or 6c
(bis-fluoro B-ring) with tert-butyl (2S)-3-chloro-2-
hydroxypropylcarbamate5 and lithium tert-butoxide
provided the Boc-protected aminomethyl oxazolidinon-
es 12b,c, respectively. Next, the Boc group was removed
and the resulting amine coupled to carboxylic acid or
acid chloride building blocks to provide the desired
amides. A final sulfur oxidation step then completed
the synthesis of oxazolidinone analogs 13a–f and 14a–e.
Compound
MIC (lg/mL)
S.a.
S.p.
E.f.
H.i.
M.c.
9b
4
4
8
8
8
4
8
2
2
4
4
4
2
2
1
4
4
2
2
2
1
1
2
2
1
1
4
2
8
8
8
4
8
2
2
4
4
2
2
8
8
13a
13b
13c
13d
13e
13f
9c
8
8
8
8
8
8
16
4
8
4
To further explore C-ring SAR, the azabicyclic C-ring
intermediate 15 was functionalized with various acyl, al-
kyl, guanidino, or heterocyclic substituents (Schemes 4
and 5). The formyl and cyano analogs 16a,b were pre-
pared from 15 by reaction with HCOOH/Ac2O or cyan-
ogen bromide, respectively. The N-acyl analogs 16d–i
were prepared by reaction of 15 with acid chlorides or
via coupling to carboxylic acids. The N-alkyl analogs
16j–l were prepared from 15 via alkylation or conjugate
addition reactions.
32
8
32
4
14a
14b
14c
14d
14e
4
4
8–16
4–8
16
8
8–16
8–16
8
4
Strains: see Table 1.
Table 3. Minimum inhibitory concentration (MIC) values for azabi-
cyclic analogs with acyl or alkyl C-ring substituents
Scheme 5 illustrates the synthesis of azabicyclic analogs
bearing guanidino or tetrazole functionality. Guanidine
and cyanoguanidine analogs 17a–c were prepared using
established procedures, for example by reaction of 15
with imidothiocarbamate reagents.6 Tetrazole analog
18a was prepared from the cyanamide 16b by reaction
with sodium azide in DMF.7 Methylation of 18a then
provided a separable mixture of 2-methyl and 3-methyl
tetrazole analogs 18b,c.
Compound
MIC (lg/mL)
S.a.
S.p.
E.f.
H.i.
M.c.
15
8
2
2
16
2
2
1
2
4
4
1
4
2
4
2
4
8
32
4
8
8
10b
16a
16b
16c
16d
16e
16f
16g
16h
16i
2
2
1
8
8
2
1
8
4
2
2
8
8
16
4
1
16
32
8
16
32
4
4
1
1
The new bicyclic oxazolidinone analogs were tested
against a panel of Gram-positive and fastidious Gram-
negative bacteria (Tables 1–4). Minimum inhibitory
concentration (MIC) values were determined using stan-
dard broth microdilution methods.8 Inspection of Table
1 reveals interesting SAR trends relating to both B-ring
and C-ring types. All three bicyclic C-ring subtypes are
viable, although thia- and azabicyclic analogs 9a–c and
10a–c were generally more potent than the oxabicyclic
analogs 11a–c. For the thiabicyclic analogs 9a–c, MIC
values improve with increasing degrees of B-ring fluori-
4
2
32
16
32
16
16
8
8
4
1
8
8
2
16
8
16j
2
1
16k
16l
4
1
0.5
8
4
8
Strains: see Table 1.
Table 4. Minimum inhibitory concentration (MIC) values for azabi-
cyclic analogs with guanidino or heterocyclic C-ring substituents
Compound
MIC (lg/mL)
Table 1. Minimum inhibitory concentration (MIC) values for linezolid
(1) and thia-, aza-, and oxabicyclic analogs 9–11 against Gram-positive
and fastidious Gram-negative bacteriaa
S.a.
S.p.
E.f.
H.i.
M.c.
17a
17b
17c
18a
18b
18c
64
1
4
1
64
2
64
8
16
8
Compound
MIC (lg/mL)
4
2
4
16
64
16
32
32
64
8
S.a.
S.p.
E.f.
H.i.
M.c.
64
2
64
1
32
2
Linezolid 1
9a
9b
4
4
4
2
4
2
4
8
4
8
2
4
2
1
4
2
2
8
2
4
4
8
16
16
8
8
8
8
4
4
8
8
8
4
8
2
1
2
8
4
Strains: see Table 1.
9c
10a
2
8
4
8
nation, 9c showing the best potency. In contrast, a
mono-fluorinated B-ring is preferred in the case of
aza- and oxabicyclic analogs (cf. 10/11b and 10/11a,c).
These B-ring SAR trends hold for both the Gram-posi-
tive and Gram-negative strains examined. The bis-fluoro
B-ring thiabicyclic analog 9c exhibited the best overall
spectrum and potency among these initial acetamide
analogs. Notably, five analogs in Table 1 (9b,c, 10a,b,
and 11b) had improved in vitro activity against Hae-
10b
10c
2
4
4
16
32
8
11a
11b
16
4
11c
8
32
a Strains: S.a.: Staphylococcus aureus UC-76 SA-1; S.p.: Streptococcus
pneumoniae SVI SP-3; E.f.: Enterococcus faecalis MGH-2 EF1-1;
H.i.: Haemophilus influenzae HI-3542; M.c.: Moraxella catarrhalis
BC-3531.

A. R. Renslo et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1126–1129
1129
Table 5. In vivo efficacy of selected analogs in a systemic mouse
infection model
mophilus influenzae as compared to linezolid, and the
most potent of these—azabicyclic analog 10b—was
fourfold more potent (H. influenzae MIC = 4 lg/mL).
Compound
Administration
route
S. aureus UC 9213
a
ED50 (mg/kg)
SAR of the oxazolidinone ring C-5 side chain was eval-
uated in the context of the thiabicyclic C-ring series
(analogs 13a–f and 14a–e, Scheme 3 and Table 2). The
majority of the C-5 groups examined were well tolerat-
ed. Propionamide analogs 13a and 14a equaled or bet-
tered the activity of the corresponding acetamides 9b,c.
Difluoropropionamide, hydroxyacetamide, and cya-
noacetamide analogs were somewhat less potent than
the propionamides, particularly against the Gram-nega-
tive pathogens (cf. 14a vs. 14b–d). The cyclopropyl
amides 13e and 14e exhibited a combination of Gram-
positive and Gram-negative activities at least as good
as the corresponding acetamides 9b,c. However, the
slightly larger cyclobutyl amide 13f was notably less ac-
tive against the Gram-negative strains. The extent of B-
ring fluorination was again important. Hence, bis-fluoro
B-ring analogs 14a–e were typically 2- to 4-fold more
potent than mono-fluoro analogs 13a–e. In total, nine
analogs in Table 2 had improved in vitro activity against
H. influenzae as compared to linezolid.
9a
9b
po
po
po
po
po
7.2 (L 4.6)
4.2 (L 2.5)
3.1 (L 2.8)
2.2 (L 2.8)
15.3 (L 4.4)
9c
10b
16c
a ED50 is the amount of drug required to cure 50% of infected mice.
Value for linezolid control is given in parentheses (L = linezolid).
corresponding urea derivative 16c was notably less ac-
tive in vivo (ED50 = 15.3 mg/kg).
In summary, conformationally constrained thia-, oxa-,
and azabicyclo[3.1.0]hexane heterocycles are valid bio-
isosteres of thiomorpholine, morpholine, and piperazine
ring systems as applied to the oxazolidinone class of
antibacterials. Analogs bearing these novel heterocycles
possess in vitro and in vivo activities comparable and in
some cases superior to those of the progenitor uncon-
strained analogs, including linezolid.
The impact of C-ring substitution was examined via the
introduction of acyl, alkyl, or heterocyclic groups in a
series of azabicyclic analogs (Tables 3 and 4). The ana-
log 10b represents a benchmark compound for this ser-
ies in that it contains the privileged hydroxyacetamide
side chain of eperezolid. Indeed, 10b displayed improved
potency and spectrum as compared to linezolid. Among
the various side chains examined, those containing a ni-
trile substituent often produced analogs with significant-
ly improved activity against the Gram-positive strains
(e.g., 16b,f, and j). In contrast, analogs possessing a pri-
mary or secondary amine were much less active (e.g., 15
and 16d). A tertiary amine however is surprisingly well
tolerated (e.g., N-alkyl analogs 16j–l). Analogs incorpo-
rating more lipophilic side chains (e.g., 16g–i) generally
had reduced activity against H. influenzae.
References and notes
1. For recent reviews, see: (a) Brickner, S. J. Curr. Pharm. Des.
1996, 2, 175; (b) Barbachyn, M. R.; Ford, C. W. Angew.
Chem., Int. Ed. 2003, 42, 2010; (c) Hutchinson, D. K. Curr.
Top. Med. Chem. 2003, 3, 1021; (d) Nilus, A. M. Curr.
Opin. Invest. Drugs 2003, 4, 149.
2. (a) Shinabarger, D. Expert Opin. Invest. Drugs 1999, 8,
1195; (b) Colca, J. R.; McDonald, W. G.; Waldon, D. J.;
Thomasco, L. M.; Gadwood, R. C.; Lund, E. T.; Cavey, G.
S.; Mathews, W. R.; Adams, L. D.; Cecil, E. T.; Pearson, J.
D.; Bock, J. H.; Mott, J. E.; Shinabarger, D. L.; Xiong, L.;
Mankin, A. S. J. Biol. Chem. 2003, 278, 21972.
3. Renslo, A. R.; Jaishankar, P.; Venkatachalam, R.; Hack-
barth, C.; Lopez, S.; Patel, D. V.; Gordeev, M. F. J. Med.
Chem. 2005, 48, 5009.
4. Renslo, A. R.; Gao, H.; Jaishankar, P.; Venkatachalam, R.;
Hackbarth, C.; Lopez, S.; Gordeev, M. F. Org. Lett. 2005,
7, 2627.
Table 4 presents MIC data for azabicyclic analogs with
guanidine and tetrazole substituents. Strongly basic or
acidic functionality is clearly not tolerated; guanidine
17a and tetrazole 18a were essentially inactive. Antimi-
crobial activity could be restored however by attenuat-
ing the basicity as in cyanoguanidines 17b,c or by
alkylation of the acidic tetrazole ring (N-Me analogs
18b,c). No significant difference in activity was observed
for regioisomeric tetrazoles 18b,c.
5. (a) Perrault, W. R.; Pearlman, B. A.; Godrej, D. B. U.S.
Patent 6,887,995, 2005; (b) Perrault, W. R.; Pearlman, B.
A.; Godrej, D. B.; Jeganathan, A.; Yamagata, K.; Chen, J.
J.; Lu, C. V.; Herrinton, P. M.; Gadwood, R. C.; Chan, L.;
Lyster, M. A.; Maloney, M. T.; Moeslein, J. A.; Greene, M.
L.; Barbachyn, M. R. Org. Process Res. Dev. 2003, 7, 533.
6. (a) Gadekar, S. M.; Nibi, S.; Cohen, E. J. Med. Chem.
1968, 11, 811; (b) DeMong, D. E.; Williams, R. M. J. Am.
Chem. Soc. 2003, 125, 8561.
7. Garbrecht, W. L.; Herbst, R. M. J. Org. Chem. 1953, 18,
1003.
The in vivo efficacy of selected thia- and azabicyclic ana-
logs was evaluated in a murine septicemia infection
model (Table 5). Thiabicyclic analogs 9a–c demonstrat-
ed oral efficacy similar to that of linezolid. Azabicyclic
analog and eperezolid isostere 10b was the most effica-
cious analog examined (ED50 = 2.2 mg/kg), while the
8. NCCLS (National Committee for Clinical Laboratory
Standards). 2000. Methods for dilution antimicrobial sus-
ceptibility tests for bacteria that grow aerobically—5th ed.;
Approved Standard. NCCLS Document M7-A5, Vol. 20,
No 2.