Design of Helicobacter pylori glutamate racemase inhibitors as selective antibacterial agents: A novel pro-drug approach to increase exposure

Article abstract of DOI:10.1016/j.bmcl.2008.06.092

High-throughput screening uncovered a pyrazolopyrimidinedione hit as a selective, low micromolar inhibitor of Helicobacter pylori glutamate racemase (MurI). Variation of the substituents around the scaffold led to low nanomolar inhibitors and improved antibacterial activity. The challenge in this program was to translate excellent enzyme inhibition into potent antibacterial activity and pharmacokinetics suitable for oral therapy. Compounds were profiled for MurI inhibition, activity against H. pylori, microsomal stability, and pharmacokinetics in mice. Iterative cycles of analog synthesis and biological testing led to compounds with substituents optimized for both low MICs (≤2 μg/ml) and good microsomal stability. In order to achieve high bioavailability, a novel pro-drug approach was implemented wherein a solubilizing sulfoxide moiety is oxidized in vivo to a sulfone.

Full text of DOI:10.1016/j.bmcl.2008.06.092

Bioorganic & Medicinal Chemistry Letters 18 (2008) 4716–4722
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design of Helicobacter pylori glutamate racemase inhibitors as selective
antibacterial agents: A novel pro-drug approach to increase exposure
a
a
b
Gregory S. Basarab a, , Pamela J. Hill , Abdullah Rastagar , Peter J. H. Webborn
*
a Cancer and Infection Research Area, AstraZeneca R&D Boston, 35 Gatehouse Drive, Waltham, MA 02451, USA
Physical Metabolic Sciences, AstraZeneca R&D Charnwood, Bakewell Road, Loughborough, Leicestershire LE11 5RH, UK
b
a r t i c l e i n f o
a b s t r a c t
Article history:
High-throughput screening uncovered a pyrazolopyrimidinedione hit as a selective, low micromolar
inhibitor of Helicobacter pylori glutamate racemase (MurI). Variation of the substituents around the scaf-
fold led to low nanomolar inhibitors and improved antibacterial activity. The challenge in this program
was to translate excellent enzyme inhibition into potent antibacterial activity and pharmacokinetics suit-
able for oral therapy. Compounds were profiled for MurI inhibition, activity against H. pylori, microsomal
stability, and pharmacokinetics in mice. Iterative cycles of analog synthesis and biological testing led to
compounds with substituents optimized for both low MICs (62 lg/ml) and good microsomal stability. In
order to achieve high bioavailability, a novel pro-drug approach was implemented wherein a solubilizing
sulfoxide moiety is oxidized in vivo to a sulfone.
Received 30 April 2008
Revised 26 June 2008
Accepted 26 June 2008
Available online 2 July 2008
Keywords:
Glutamate racemase
MurI
Pro-drug
Pyrazolopyrimidinedione
Ó 2008 Elsevier Ltd. All rights reserved.
The 2005 Nobel Prize in Physiology and Medicine was awarded
to Barry Marshall and Robin Warren for the discovery of the bacte-
rium Helicobacter pylori and for the determination that its coloniza-
tion of the stomach mucosa causes gastritis and peptic ulcers.1 An
association has been made between infection by the bacterium and
a predisposition to gastric cancer.2 The diseases caused by H. pylori
can be cured by eradication of the pathogen with various antibac-
terial therapies;3 however, their shortcomings including undesired
intestinal side effects due to disruption of commensal bacteria
prompt the quest for a novel, selective anti-H. pylori therapy. The
recent emergence of resistant H. pylori strains to the commonly
used antibiotics further supports the need for novel therapies with
new modes of action.4 Selectivity versus H. pylori would also re-
move resistance pressures otherwise imposed by broad-spectrum
agents on quiescent bacterial pathogens.
Phylogenetic analysis across a range of bacterial species places
the H. pylori MurI gene encoding the enzyme glutamate racemase
in a divergent niche relative to other MurI genes, and it was sug-
gested that selective inhibition could be achieved.5 As b-lactam
antibiotics operate by blocking cell wall peptidoglycan cross-link-
ing, inhibition of other cell wall biosynthesis targets for drug de-
sign follows.6 Glutamate racemase (MurI) from Gram-positive
and Gram-negative bacteria catalyzes the interconversion of
acterization of MurI from Escherichia coli,8 and the determination
of a 2.3 Å resolution crystal structure of MurI from Aquifex pyrophi-
lus.9 MurI has been targeted for inhibitor design with the synthesis
of glutamate analogs that bind at the active site and show a mea-
sure of antibacterial activity.10 A selective H. pylori MurI inhibitor
could show promise as a drug therapy if inhibition translates into
antibacterial activity and pharmacokinetic (PK) properties lead to
sustained blood plasma levels sufficiently high enough to eliminate
the pathogen. We envisioned that an anti-H. pylori drug would
need to be administered orally to gain widespread acceptance by
the medical community.
N
N
N
O
4
O
N
3
Cl
N5
6
N
N
N
7
N
2
N
1
O
N
O
N
1
MurI IC50 = 1400 nM
MICs: H. pylori (wild-type) = 8 µg/ml
2
MurI IC50 = 26 nM
MICs: H. pylori (wild-type) = 0.5 µg/ml
hefC-
= 0.25 µg/ml
L- and D-glutamate, the latter being a necessary component of cell
wall peptidoglycan. Previous publications had described the ratio-
(
)
nale for targeting MurI for antibacterial therapy,7 the kinetic char-
Human microsomal Clint < 14 µl/min/mg
Mouse microsomal Clint = 50 µl/min/mg
bioavailabilty (mouse) = 7.5%
* Corresponding author. Tel.: +1 781 839 4689; fax: +1 781 839 4220.
E-mail address: greg.basarab@astrazeneca.com (G.S. Basarab).
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.06.092

G. S. Basarab et al. / Bioorg. Med. Chem. Lett. 18 (2008) 4716–4722
4717
oxides, and sulfonamides. These functionalities can be
accommodated sterically by the crystallographic model and offer
the potential to form hydrogen bonds with enzyme side chains
or bridging water molecules. Substitution at the pyrrole 4-position
can be mimicked with substitution on other aromatic 5-membered
rings affording a similar 1,3-relationship to the attachment of the
pyrazolopyrimidinedione scaffold. Examples of substitution on
methyl pyrrole, methyl imidazole, and furan are examined here.
Pyrazolopyrimidinediones 5 can be synthesized efficiently by
condensation of pyrimidinedione hydrazide 3 with aldehydes to
form intermediate hydrazones 4 (Scheme 1).13 A second condensa-
tion with a second aldehyde with heating is followed by cyclization
and rearrangement to the pyrazolopyrimidinedione scaffold in a
single reaction sequence. Piperidine serves as an efficient catalyst
for the second condensation. Hydrazide 3 is readily prepared by
methods set out in the literature.5,14 Since the 5-position methyl
and the 7-position cyclopropylmethyl afford high inhibition po-
tency and antibacterial activity, the two substituents were main-
tained as constants for the purpose of this investigation.
Condensations to form hydrazones 4 were investigated with three
commercial aldehydes (5-chloroindole-3-carboxaldehyde, 5-
chloro-1-methylindole-3-carboxaldehyde, and 6-chloroquinoline-
4-carboxaldehyde) en route to compounds 5 with inhibitory
potencies (IC50s) consistently below 100 nM. Subsequent conden-
sations and rearrangements to compounds 5 were carried out with
a variety of furan, pyrrole and imidazole carboxaldehydes. The
reaction sequence was carried out in a single reaction vessel by
room temperature stoichiometric addition of the first aldehyde to
form hydrazones 4 followed by addition of the second aldehyde
and piperidine with heating to 80ꢀC to afford pyrazolopyrimidined-
iones 5.
The H. pylori MurI gene was cloned and overexpressed in E. coli and
the enzyme was harvested for biochemical characterization, inhibi-
tion assay development, and protein crystallization trials.5 High-
throughput screening of the AstraZeneca compound collection
identified pyrazolopyrimidinedione 1 as a low micromolar inhibitor
of H. pylori MurI with selectivity versus isozymes from the Gram-
negative pathogen E. coli and 2 Gram-positive pathogens, Staphylo-
coccus aureus and Enterococcus faecalis.5 Analog work around 1 led
to 2 as a formidable low nanomolar inhibitor of MurI showing an
MIC = 0.5 lg/ml against wild-type H. pylori and 0.25 lg/ml against
an RND-efflux pump debilitated H. pylori strain (hefC-).5,11 Bacterial
efflux oftentimes contributes to increased MICs, and to assess this
contribution, the compounds described herein were screened
against both the wild-type and pump debilitated H. pylori strains.
Despite the encouraging anti-microbial properties of 2, some of
the properties deemed essential for an orally dosed drug are lack-
ing. For example, low aqueous solubility (<0.4 lM) and low bio-
availability (<10%) on oral dosing in mice prevent the plasma
drug concentration from exceeding the MIC. Here, we attempt to
synthesize more polar and water-soluble MurI inhibitors to identify
anti-H. pylori compounds with improved bioavailability in mouse
models. Ultimately, a pro-drug strategy was devised wherein an
in vivo sulfoxide to sulfone conversion delivers plasma levels above
the MIC on oral administration.
We set out to incorporate more polar functionality onto the
fused pyrimidinedione scaffold to increase solubility anticipating
a concomitant improvement in a number of PK properties (e.g.
microsomal stability, plasma protein binding, and absorption)
that contribute toward generating drug exposure in the blood
necessary to eradicate H. pylori. Ideally, the polar functionality
should maintain or improve inhibitory potency so as not to
diminish antibacterial activity. The X-ray structure of the H. py-
A series of furans, pyrroles and imidazole carboxaldehydes (R3-
CHO) used in Scheme 1 to affix the R3 substituent were prepared de
novo. The furan carboxaldehyde for Compound 27 is known in the
literature.15 The chlorinated furan 9 incorporated into 28 was syn-
thesized by formation, lithiation, and chlorination of 7 followed by
hydrolysis and oxidation as outlined in Scheme 2. The synthesis of
the furan 12 (Scheme 3) for incorporation into 29 and 30 began by
reduction of the commercial furylsulfonyl chloride 10 and alkyl-
ation with methyl iodide affording 11 via methods set out in the
literature.16,17 Under the reaction conditions some ester hydrolysis
occurred requiring re-esterification to maximize recovery of 11.
The ester 11 was converted to the aldehyde 12 by reduction to
the alcohol and oxidation.
An imidazole sulfonamide carboxaldehyde 14 for incorporation
into 31 was prepared by silylation of the commercially available
sulfonamide 13 of Scheme 4 followed by lithiation and treatment
with DMF. The silyl group is removed during workup of the subse-
quent cyclization reaction forming 31.
Pyrrole sulfonamides 17a–c were derived from the sulfonyl
chloride 1518 shown in Scheme 5 by amination followed by con-
version of the ester to the aldehyde. The t-butyl protecting group
after formation of the pyrazolopyrimidinedione was removed with
TFA to give 32. The sulfonamides otherwise were incorporated into
Compounds 33–36.
lori glutamate racemase with 1 and
D-glutamate bound has
been determined to 1.9 Å showing that the inhibitor occupying
an allosteric binding site is removed from the substrate.5 The
core scaffold accommodates variation at the 2-, 3-, 5- and 7-
positions (see 1 for numbering) for analog enumeration. The
7-position isobutyl group and the 2-position naphthalene of
the inhibitor are deeply buried in hydrophobic regions of the
inhibitor binding pocket. The pyrazolopyrimidinedione 5-posi-
tion methyl group lies against a hydrophobic surface and other-
wise faces solvent. Without nearby polar residues or backbone
peptide linkages to offer opportunities for an electrostatic com-
plement, replacing the methyl group with polar functionality
does not bode well for enhancing inhibitory potency.
It therefore seemed best to target the region of the binding
pocket surrounding the pyridine of 1 or, analogously, the pyrrole
of 2 for the incorporation of polar functionality. Replacing 1 with
2 in the inhibitor binding region shows the pyrrole methyl group
surrounded by a nicely complementary hydrophobic environment
whereas the region surrounding the cyano substituent is lined with
the hydrogen bonding atoms of a series of more polar side-chain
functionalities (Ser152, Gln248, Arg247, and Trp 252) and reaches
outward to solvent.12 Here, we prepared compounds where the cy-
ano group is replaced with the sulfur substitutions: sulfones, sulf-
R3
O
O
N
O
N
R3CHO
N
R2CHO
N
N
N
N
NH2
N
R2
R2
O
N
O
N
O
N
H
H
N
H
3
4
5
Scheme 1.

4718
G. S. Basarab et al. / Bioorg. Med. Chem. Lett. 18 (2008) 4716–4722
NH
1) n-BuLi
2) Cl2FCCF2Cl
S
H
N
Cl
O
S
S
O
O
n
3) H3O+
O
O
N
O
H2O
N
98%
30%
H
H
8
9
n = 0
2
6
7
m-CPBA
80%
Scheme 2.
O
S
O
O
O
O
O
1) Na2SO3
2) CH3I
O
O
O
S
S
1) LAH
Cl
2) MnO2
3) MeOH
HCl
O
O
O
O
O
H
35%
54%
10
11
Scheme 3.
12
rings reflects improved inhibitory potencies. Higher potencies with
concomitant improved antimicrobial activity are further seen if a
small methyl substituent is positioned on the 5-membered aro-
matic ring in a 1–2 relationship to the bond connecting to the scaf-
fold as seen in 2. This is explained by the substituent filling a
hydrophobic area of the MurI inhibitor binding pocket and favoring
a staggered conformation as corroborated by X-ray crystallography
where the dihedral angle around the connecting bond is 63 deg.
This is quite accessible energetically from the optimal unbound
ground-state conformation of 49 deg. The methyl group can be
incorporated onto a pyrrole ring as in 2, imidazole as in 31, and fur-
an as in 29. The chlorine atom of furan 28 serves a similar role to
the methyl group increasing potency 25-fold relative to 27, which
lacks a substituent that would favor the staggered conformation
(Table 1).
Potency is further improved when polar groups are situated in a
1–3 relationship with the linkage to the scaffold. The highest po-
tency in the series was found for the quite polar sulfonamide sub-
stituent of 32 (Table 2). However, the polarity raises the MICs
versus H. pylori presumably due to diminished membrane perme-
ability. The diminished membrane permeability translates to
greater susceptibility to efflux by the RND transporter demon-
strated by considerably lower MICs against the hefCÀ strain. Adding
substituents onto the sulfonamide slightly decreases inhibitory po-
tency (compare 33 and 35 with 32) but can improve MICs through
the increase in lipophilicity. Better MICs were achieved with incor-
poration of less polar functionalities such as the nitrile of 2 and the
methyl sulfones of compounds 28–30 (Table 1) and 41–43 (Table
3). In line with increased polarity mitigating antibacterial activity,
imidazoles increased MICs relative to pyrroles (compare 37 and 38
Si
HN
NH2
1) TBDMS-Cl
2) -BuLi, DMF
S
O
O
S
N
N
O
n
O
N
N
O
60%
H
13
14
Scheme 4.
Imidazole sulfone 21 for the preparation of 37 and 38 was syn-
thesized via the sequence outlined in Scheme 6. Imidazole carbox-
aldehyde 18 is protected as the acetal and doubly brominated. The
bromine nearest to the imidazole methyl group is removed by lith-
ium halogen exchange followed by proteolysis. Metal halogen ex-
change of the second bromine was followed by sulfenation
affording 20. Oxidation to the sulfone and hydrolysis of the ketal
gave aldehyde 21.
Compounds 39–44 utilized pyrrole sulfone and sulfoxide carb-
oxaldehydes prepared by protection of 1-methyl-2-formyl-5-
bromopyrrole 2219 as the aminal 23, conversion to the sulfides,
hydrolysis to aldehydes 24, and oxidation as diagrammed in
Scheme 7.
High inhibitory potencies (<100 nM) were recorded for the
majority of compounds listed in Tables 1–3. As inhibitory poten-
cies versus the H. pylori isozyme increased, selected compounds
evaluated for inhibition against the MurI isozymes of E. coli, S. aur-
eus, and E. faecalis showed IC50s > 400 lM reinforcing a superb
selectivity. The focus on 5-membered aromatic rings at the 3-posi-
tion of the pyrazolopyrimidinedione scaffold over 6-membered
R
R
N
N
Cl
RNH2, Et3N
S
1) LAH
S
S
N
N
N
O
O
O
O
O
O
2) MnO2
O
H
O
90-100%
O
O
O
40-50%
17a R = Me
17b
17c
OMe
-Bu
15
16
t
Scheme 5.

G. S. Basarab et al. / Bioorg. Med. Chem. Lett. 18 (2008) 4716–4722
4719
Br
S
1) HC(OEt)3
TsOH
Br
S
O
1) m-CPBA
2) H3O+
1)
2)
n
-BuLi, H+
-PrMgBr,
N
N
N
N
O
N
N
N
i
2) NBS
N
MeSSMe
O
O
O
O
O
H
O
Et
83%
H
70%
36%
Et
Et
Et
18
20
19
21
Scheme 6.
NH
H
N
1)
n-BuLi
Br
N
S
2) RSSR
Br
N
H
N
O
n
3) H3O+
99%
N
O
O
H2O
N
88%
H
24 n = 0
22
23
NaIO4, 50%
25
1
m-CPBA, 88%
26
2
Scheme 7.
Table 1
Influence of 3-position furans on IC50, MIC, and microsomal clearance (Clint
)
O
R3
H3C
O
N
N
R2
N
N
a
n
R3
R2
IC50 (nM)
1500 100
MICsb
(l
g/ml)
Clint
Human
(ll/min/mg)
Wild-type
hefC-
Mouse
CH3
Cl
Cl
S
27
28
>64
>64
4
17
19
14
O
O
N
N
O
CH3
S
Cl
59
51
4
5
4
4
22
61
O
O
O
CH3
Cl
O
S
29
30
H3C
—
22
O
O
O
N
CH3
Cl
O
S
H3C
57
5
1
0.5
>100
>100
N
O
H3C
a
IC50s and standard errors from duplicate determinations.
Helicobacter pylori MICs measured by NCCLS guidelines with a twofold variance.
b
with 41 and 42, respectively, Table 3). Lower inhibitory potencies
for imidazoles relative to pyrroles may also contribute to the high-
er MICs.
The compounds in Tables 1–3 were further assessed for poten-
tial in vivo stability by incubation in vitro with human and mouse
microsomes. Many of the compounds proved unstable to micro-
somes eliminating their utility as drugs. All the compounds con-
taining an indole R2 substituent suffered from instability to
human and mouse microsomes, mitigating their utility in vivo.
Across the range of analogs, the chloroquinoline group slightly de-
creased potency; however, it was necessary (but not sufficient) for
metabolic stability. Furthermore, the slight decrease in potencies
generally did not diminish antimicrobial activities. Two com-
pounds (Compounds 2 and 41, vide infra for 41) were evaluated
for in vivo PK properties in the mouse due to the combination of
lower MICs and lower intrinsic clearances. Though both com-
pounds showed low in vivo clearances (18 and 15 ml/min/kg),
respectively, both failed to achieve sufficiently high bioavailability

4720
G. S. Basarab et al. / Bioorg. Med. Chem. Lett. 18 (2008) 4716–4722
Table 2
Influence of 3-position pyrroles and imidazoles with sulfonamide substituents on IC50, MIC, and microsomal clearance (Clint
)
O
R3
H3C
O
N
N
R2
N
N
R3
R2
IC50 (nM)
MICsb
Wild-type
(
lg/ml)
Clint (
Human
ll/min/mg)
Mouse
a
n
hefC-
NH2
Cl
S
31
H3C
38
16
44
5
1
2
>32
32
0.5
56
>100
<14
O
N
N
N
O
H
N
NH2
Cl
Cl
S
32
33
H3C
H3C
H3C
0.25
22
O
O
N
H3C
NH
16
1
>100
>100
S
O
N
N
N
O
H3C
Cl
Cl
NH
34
35
36
27
36
55
1
2
2
16
0.5
>100
>100
S
O
N
H
O
H3C
O
NH
8
8
ND
ND
S
H3C
O
N
N
H
O
H3C
O
Cl
NH
16
4
ND
ND
S
H3C
O
N
N
O
a
IC50s and standard errors from duplicate determinations.
Helicobacter pylori MICs measured by NCCLS guidelines with a twofold variance.
b
to afford concentrations above the MICs. This is perhaps due to
poor solubility of the compounds preventing efficient absorption.
The data in Table 3 suggest a correlation between the clearances
of 39 and 41, direct analogs with a more polar sulfoxide and a less
polar sulfone functionality. Note that the two compounds are
nearly equipotent (37 and 34 nM, respectively) against the target
enzyme yet show an eightfold difference in MICs. The polar sulfox-
ide decreases membrane permeability mitigating the expression of
antibacterial activity in accordance with the discussion above.
From the viewpoint of microsomal stability, the sulfoxide serves
as a handle for rapid clearance, while the sulfone survives intact.
It follows that the sulfoxide might be converted to the sulfone by
microsomes and indeed, when 39 was incubated with both human
and mouse microsomes, there was quantitative conversion to 41
(Figs. 1 and 2).
MIC of 2 lg/ml versus H. pylori (Fig. 3). Oral administration of a
40 mg/kg dose of sulfoxide 39 afforded peak plasma levels of sul-
fone 41 of 3 g/ml, exceeding the MIC of 2 g/ml for 4 h. The resid-
ual sulfoxide detected in the plasma reached a peak concentration
of 1.4 g/ml. The data demonstrate that total drug levels (sulfox-
l
l
l
ide + sulfone) significantly increase on oral dosing of the former
relative to the latter, affording a bioavailability of 60% for the sul-
fone and 10% for the sulfoxide. Hence, conversion of the sulfoxide
to the sulfone is realized in vivo as predicted by in vitro incubation
with microsomes. Equilibrium solubilities of the sulfoxide and sul-
fone (120 and 4 lM, respectively) correlate affirmatively with their
abilities to achieve overall higher drug exposure. The non-steroidal
anti-inflammatory drug sundilac represents another example of a
more water-soluble sulfoxide compound serving as a pro-drug to
improve oral bioavailability.20 In this case, the sulfoxide is reduced
in vivo to the pharmacologically active sulfide form rather than
being oxidized to the sulfone.21 The free drug concentration cor-
rected for plasma protein binding ultimately governs the capability
for eradication of disease in vivo. Despite the achievement of good
Hence, the microbiologically less potent sulfoxide was envi-
sioned to serve as a pro-drug for the sulfone. The oral bioavailabil-
ity of 41 was low (9%) when administered to mice at 40 mg/kg
achieving maximum plasma levels of 1.2 lg/ml, lower than the

G. S. Basarab et al. / Bioorg. Med. Chem. Lett. 18 (2008) 4716–4722
4721
Table 3
Influence of 3-position pyrroles and imidazoles with sulfone and sulfoxide substituents on IC50, MIC and microsomal clearance (Clint
)
O
R3
H3C
N
N
R2
N
O
N
a
n
R3
R2
IC50 (nM)
MICsb
Wild-type
(
lg/ml)
Clint
Human
(ll/min/mg)
hefC-
Mouse
CH3
Cl
Cl
S
37
H3C
160 10
16
16
16
16
2
4
23
56
O
N
N
N
N
N
N
N
O
N
N
N
CH3
S
38
39
40
41
42
43
H3C
H3C
H3C
H3C
H3C
H3C
31
37
23
34
26
33
4
4
1
3
2
1
2
49
>100
>100
>100
<14
64
O
N
O
H
CH3
Cl
Cl
S
0.5
>100
>100
20
O
N
CH3
S
4
O
N
H
CH3
Cl
Cl
S
0.5
0.5
4
O
O
O
O
N
CH3
S
4
49
N
O
H
CH3
Cl
S
16
23
56
N
O
H3C
a
IC50s and standard errors from duplicate determinations.
Helicobacter pylori MICs measured by NCCLS guidelines with a twofold variance.
b
2.5
2.5
2.0
1.5
1.0
0.5
0.0
2.0
39
39
41
1.5
41
1.0
0.5
0.0
0
10
20
30
0
5
10
15
20
25
30
Incubation Time (min)
Incubation Time (min)
Figure. 1. Transformation of 39–41 in mouse microsmes.
Figure 2. Transformation of 39–41 in human microsomes.
bioavailability through this pro-drug approach, the protein binding
of 41 in mouse plasma is sufficiently high (98%) to abrogate testing
of 39 for efficacy in H. pylori infected mice.
In summary, a series of potent inhibitors of the H. pylori MurI
enzyme are described. In order to translate target potency to
effectiveness as a drug, chemical–physical properties need to be
optimized to allow sufficient membrane permeability for the
expression of MICs. For oral bioavailability, chemical–physical
properties also need to be optimized to achieve appropriate solu-

4722
G. S. Basarab et al. / Bioorg. Med. Chem. Lett. 18 (2008) 4716–4722
References
S
S
u
uSS
uSSfuu
uSSfuuollff
l
l
f
ufu
ollff
noo
o
ollff
noo
enn
x
xoo
enn
4ee
i
i
d
dxx
4ee
144
e
ieidd
144
(11
3
3ee
9
933
(0(44
(0(44
m00 m
(
(99
4
4(
0
m
m00 mm
gkg
g//kk
g//kkggp
g
g
/
/gg
k
k
g
g//kk
foo
opp)oo))
p
o
opp
)
)oo))
S
Su
l
lf
o
n
e
4
1
(
(11
4
40
m
m00
g
/
/
k
g
ogog
f
3
9
933
p
p99o
o
)
)
1. Megraud, F. Cell 2005, 123, 975.
3500
3000
2500
2000
1500
1000
500
S
Su
l
lf
o
n
e
4
1
(
4
40
m
g
/
k
g
p
o
)
2. (a) Peek, R. M., Jr. Springer Semin. Immunopathol. 2005, 2, 197; (b) Blaser, J. J.;
Ahterton, J. C. J. Clin. Invest. 2004, 113, 321; (c) Uemura, N.; Okamoto, S.;
Yamamoto, S.; Matsumura, N.; Yamaguchi, S.; Yamakido, M.; Taniyama, K.;
Sasaki, N.; Schlemper, R. J. N. Eng. J. Med. 2001, 345, 784; (d) Nomura, A. M. Y.;
Lee, J.; Stemmermann, G. N.; Nomura, Y.; Perez-Perez, G. I.; Blaser, M. J. J. Infect.
Dis. 2002, 186, 1138.
MIC
3. Rubinstein, G.; Dunkin, K.; Howard, A. J. J. Antimicrob. Chemother. 1994, 24,
1028.
4. Suerbaum, S.; Michetti, P. N. Engl. J. Med. 2001, 11, 784.
5. Lundqvist, T.; Fisher, S. L.; Kern, G.; Folmer, R. H. A.; Xue, Y.; Newton, D. T.;
Keating, T. A.; Alm, R.; de Jonge, B. L. M. Nature 2007, 447, 817.
6. (a) Zoeiby, A. E.; Sanschagrin, F.; Levesque, R. C. Mol. Microbiol. 2003, 47, 1; (b)
Sobral, R. G.; Kimura, K.-I. ; Bugg, T. D. H. Nat. Prod. Rep. 2003, 20, 252; (c) Lazar,
K.; Walker, S. Curr. Opin. Chem. Biol. 2002, 6, 786; (d) Lambert, P. A.; Allison, D.
G.; Gilbert, P. Mol. Med. Microbiol. 2002, 1, 591; Katz, A. H.; Caufield, C. E. Curr.
Pharm. Des. 2003, 9, 857.
7. (a) Tanner, M. E.; Miao, S. Tetrahedron Lett. 1994, 35, 4073; (b) Ruzheinikov,
Sergey N.; Taal, Makie A.; Sedelnikova, Svetlana E.; Baker, Patrick J.; Rice, David
W. Structure 2005, 11, 1707.
0
0
2
4
6
Time (hr)
Figure 3. Mouse plasma levels of 39 and 41.
8. (a) Glavas, S.; Tanner, M. E. Biochemistry 2001, 21, 6199; (b) Tanner, M. E.; Gallo,
K. A.; Knowles, J. R. Biochemistry 1993, 15, 3998; Gallo, K. A.; Tanner, M. E.;
Knowles, J. R. Biochemistry 1993, 15, 3991.
9. Hwang, K. Y.; Cho, C.-S. ; Kim, S. S.; Sung, H.-C. ; Yu, Y. G.; Cho, Y. Nat. Struct.
Biol. 1999, 5, 422.
10. De Dios, A.; Prieto, L.; Martin, J. A.; Rubio, A.; Ezquerra, J.; Tebbe, M.; Lopez de
Uralde, B.; Martin, J.; Sanchez, A.; LeTourneau, D. L.; McGee, J. E.; Boylan, C.;
Parr, T. R., Jr.; Smith, M. C. J. Med. Chem. 2002, 45, 4559.
11. Kutschke, A.; de Jonge, L. M. B. Antimicrob. Agents Chemother. 2005, 7, 3009.
12. PDB accession number of the H. pylori MurI complex with 1: 2jfz.
13. Yamada, F.; Higuchi, M.; Natamura, T. J. Am. Chem. Soc. 1974, 96, 5607.
14. (a) Naka, T.; Furukawa, Y. Chem. Pharm. Bull. 1979, 6, 1328; (b) Isobe, Y.; Hirota,
K. Chem. Pharm. Bull. 2003, 12, 1451; (c) Basarab, G.; Eyermann, J.;
Gowravaram, M.; Green, O.; MacPherson, L.; Morningstar, M.; Nguyen, T.
Chem. Abstr. 2004, 139, 89821. USP 7285558; (d) Cheshire, D.; Cooper, M.;
Donald, D.; Thorne, P. Chem. Abstr. 2000, 129, 302656. USP 6028074.
15. Chou, S.-S. P.; Shen, C.-H. Tetrahedron Lett. 1997, 36, 6407.
16. Urban, F. J.; Jasys, V. J.; Raggon, J. W.; Buzon, R. A.; Hill, P. D.; Eggler, J. F.;
Weaver, J. D. Synth. Commun. 2003, 12, 2029.
bility and permeability for intestinal absorption. The most potent
antibacterial compounds that maintained stability to microsomes
failed to achieve blood plasma levels in the mouse, approaching
the respective MICs. It was only when the more polar, yet less po-
tent sulfoxide 39 was dosed orally that higher plasma levels of
drug were achieved. The sulfoxide was largely converted to the
more potent sulfone 41, achieving nearly a sevenfold increase in
bioavailability relative to oral administration of the sulfone. The
sulfoxide thus serves as a pro-drug handle for the active sulfone
metabolite as the increase in polarity and solubility enhances the
absorption necessary for an oral drug.
Acknowledgments
17. Brown, R. W. J. Org. Chem. 1991, 16, 4974.
18. Nagel, Arthur Adam. DE 78-2820411.
19. He, R. H.-Y. ; Jiang, X.-K. J. Chem. Res., Synop. 1998, 12, 786.
20. Shen, T.; Winter, C. Adv. Drug Res. 1977, 12, 89–295.
21. For a recent review on pro-drugs, see Ratio, J.; Kumpulaninen, H.; Heimbach,
T.; Oliyai, R.; Oh, D.; Jävinen, T.; Savolainen, J. Nat. Rev. Drug Disc. 2008, 7, 255.
The authors thank the AstraZeneca Infection Discovery Microbi-
ology Department for the MICs, the DMPK Department for measur-
ing in vitro and in vivo PK parameters, and the Biochemistry
Department for determining IC50s.