Bisphosphonated fluoroquinolone esters as osteotropic prodrugs for the prevention of osteomyelitis

Article abstract of DOI:10.1016/j.bmc.2008.09.010

Osteomyelitis is a difficult to treat bacterial infection of the bone. Delivering antibacterial agents to the bone may overcome the difficulties in treating this illness by effectively concentrating the antibiotic at the site of infection and by limiting the toxicity that may result from systemic exposure to the large doses conventionally used. Using bisphosphonates as osteophilic functional groups, different forms of fluoroquinolone esters were synthesized and evaluated for their ability to bind bone and to release the parent antibacterial agent. Bisphosphonated glycolamide fluoroquinolone esters were found to present a profile consistent with effective and rapid bone binding and efficient release of the active drug moiety. They were assessed for their ability to prevent bone infection in vivo and were found to be effective when the free fluoroquinolones were not.

Full text of DOI:10.1016/j.bmc.2008.09.010

Bioorganic & Medicinal Chemistry 16 (2008) 9217–9229
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Bisphosphonated fluoroquinolone esters as osteotropic prodrugs
for the prevention of osteomyelitis
Kelly S. E. Tanaka, Tom J. Houghton, Ting Kang, Evelyne Dietrich, Daniel Delorme, Sandra S. Ferreira,
Laurence Caron, Frederic Viens, Francis F. Arhin, Ingrid Sarmiento, Dario Lehoux, Ibtihal Fadhil,
Karine Laquerre, Jing Liu, Valérie Ostiguy, Hugo Poirier, Gregory Moeck, Thomas R. Parr Jr., Adel Rafai Far
*
Targanta Therapeutics Inc., 7170 Avenue Frederick Banting, Saint Laurent, Qué., Canada H4S 2A1
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 29 July 2008
Revised 4 September 2008
Accepted 5 September 2008
Available online 9 September 2008
Osteomyelitis is a difficult to treat bacterial infection of the bone. Delivering antibacterial agents to the
bone may overcome the difficulties in treating this illness by effectively concentrating the antibiotic at
the site of infection and by limiting the toxicity that may result from systemic exposure to the large doses
conventionally used. Using bisphosphonates as osteophilic functional groups, different forms of fluoro-
quinolone esters were synthesized and evaluated for their ability to bind bone and to release the parent
antibacterial agent. Bisphosphonated glycolamide fluoroquinolone esters were found to present a profile
consistent with effective and rapid bone binding and efficient release of the active drug moiety. They
were assessed for their ability to prevent bone infection in vivo and were found to be effective when
the free fluoroquinolones were not.
Keywords:
Osteomyelitis
Fluoroquinolone
Bisphosphonates
Drug delivery
Prodrug
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
circumvent these difficulties. The use of beads impregnated with
antibiotics is being examined for such a purpose.5 The inconve-
Osteomyelitis refers to an inflammatory process often accompa-
nied by bone necrosis which results from an underlying microbial
infection,1 primarily caused by Staphylococcus aureus.2 In general, it
is established as a result of trauma, bone surgery, joint replace-
ment or in cases of reduced vascularization, such as in diabetic
and elderly patients. This is a notoriously challenging illness to
treat with a need for surgical intervention and sometimes amputa-
tion,3 as well as frequent relapses.4 There are no marketed antibi-
otics approved with a Gram positive osteomyelitis indication, and
treatment is often a result of the preferences of the attending
physicians.
It is generally believed that the difficulties in treating osteomy-
elitis stem from the sheltered physiological environment offered to
the bacteria, poorly accessible to immune responsive elements and
to therapeutic agents. In such settings, bacterial growth is likely
very slow or even halted, and thus more resistant to antibacterial
treatment. These are difficult conditions for effective treatment
requiring a large concentration of the antibiotic to be maintained
in the infected bone over a long period of time, and therefore fre-
quent intravenous administration of high doses of antibiotics is
typically required. The ability to deliver an antibacterial agent to
the bone where it would be released over time could conceivably
nience of this approach stems from the need to implant these
materials surgically which severely impedes the repeat applica-
tions that are often necessary in the treatment of the disease.
Delivery of agents to bone has been achieved after systemic
administration by the use of bisphosphonates.6 These pyrophos-
phate analogs demonstrate strong affinity for hydroxyapatite, the
calcium phosphate bone mineral, and rapidly diffuse to osseous
tissues in vivo. Bisphosphonates have been used to deliver small
molecule therapeutics,7 ligands for radioisotope imaging8 and even
proteins9 to osseous tissues. Recently, a report has described the
use of bisphosphonates for the delivery of ciprofloxacin to bone.10
Although the ciprofloxacin-bisphosphonate conjugate described
appeared to bind efficiently to bone, it was unable to release of
the ciprofloxacin moiety (unpublished results). This would limit
the potential of this agent, as it is reasonable to propose that be-
cause the target of ciprofloxacin is located within the cell and be-
cause the hydrophilic bisphosphonate moiety is likely unable to
cross the hydrophobic cell membrane, the release of the bioactive
moiety is necessary to achieve a positive therapeutic outcome. Bac-
teria are not very likely to be impacted by an antibacterial agent
which is essentially irreversibly bound to the bone surface. Like-
wise, bacteria are not known to possess bone-resorptive activity
which is required to absorb bisphosphonates once they are bound
on bone. The relationship between resorptive activity and efficacy
of therapeutic bisphosphonates has been demonstrated in the case
* Corresponding author. Tel.: +1 514 496 6667; fax: +1 514 332 6033.
E-mail address: afar@targanta.com (A. Rafai Far).
0968-0896/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.09.010

9218
K. S. E. Tanaka et al. / Bioorg. Med. Chem. 16 (2008) 9217–9229
of osteoporosis treatments.11 Although the local acidity12 brought
on by the infecting organism might be associated with a measure
of bisphosphonate release from bone, the efficiency of such a pro-
cess in providing a sufficient concentration of the antimicrobial
agent is doubtful.
Bisphosphonate prodrugs have been described for the delivery
of small molecules, such as diclofenac,7a,7b prostaglandins,7c ste-
roids,7d carboxyfluorescein7e,7f, tryptophan7g and camptothecin7g
to bone. We have therefore opted for a similar approach. Unlike
the strategy that was reported for ciprofloxacin,10 the antibiotic-
bisphosphonate conjugates reported here were designed to release
the drug moiety and thus to exploit localized activity of free
antibiotic.
Fluoroquinolones are a well-defined class of wholly synthetic
antibacterial agents that have proven to be successful clinically.13
They possess a set of desirable attributes, being generally safe
and efficacious following oral and parenteral administration, with
a broad antimicrobial spectrum that includes the bacterial patho-
gens that are frequently encountered in osteomyelitis. They are
bactericidal at clinically achievable doses, a feature which is most
desirable in treating osteomyelitis given the sheltered environ-
ment of the microorganism and the high levels of recurrence asso-
ciated with the disease. While drug efflux and mutations in the
drug target have been associated with resistance to fluoroquino-
lones, such resistance occurs at a rate low enough that these agents
continue to provide clinical utility when used in monotherapy. Flu-
oroquinolones have demonstrated some efficacy in animal models
of osteomyelitis.14 Not unimportantly, they are relatively easy to
synthesize and modify.
with trimethylsilylbromide, to provide fluoroquinolone–bis-
phosphonate conjugates 9a–b.
The typically more labile thioesters 12a–b were prepared by
coupling the protected fluoroquinolones with thiol 10 obtained
from the similar iodide 7c (Scheme 3), while coupling with phenol
1315 gave, after deprotection, aryl ester 15 (Scheme 4).
Given the potential of glycolamide esters as prodrugs,16 a set of
such compounds were prepared for evaluation. Alkylation of the
carboxylates of 2a–c with bromoacetamide 167c followed by
deprotection provides glycolamide esters 18a–c (Scheme 5). Alkyl-
ation of tetramethyl methylenebisphosphonate with p-nitrobenzyl
bromide followed by reduction of the nitro group and acylation of
the resulting aniline provided bromoacetamide 21 which was used
similarly to prepare glycolamide esters 23a–b (Scheme 6). Glycola-
mide esters 28a–b were likewise obtained from bromoacetamide
26 which was synthesized from N-methyl benzylamine (Scheme
7).
The moxifloxacin thioglycolamide ester 32 was produced by
coupling 2a to thiol 30 followed by deprotection (Scheme 8). This
thiol was itself prepared by the direct condensation of mercaptoa-
cetic acid with amine 29.8c
The preparation of prodrug 38 started from protected lysine 33,
through coupling with 29, deprotection to the diamine 35, and
acylation to afford the bis(bromoacetamide) 36 (Scheme 9). This
later compound was condensed with 2b and deprotected to pro-
vide 38.
3. In vitro evaluation of prodrugs
Therefore, the preparation and evaluation of bisphosphonated
fluoroquinolone prodrugs appeared to be of significant interest
for the treatment and the prevention of osteomyelitis. This report
describes our efforts in preparing such prodrugs using the carboxy
group of the fluoroquinolones as a synthetic handle.
The antibacterial activities of the prodrugs against S. aureus
ATCC13709 were measured by broth microdilution as per
guidelines of the Clinical and Laboratory Standards Institute (re-
sults not shown). The minimum inhibitory concentrations (MICs)
were >4
l
g/mL, significantly higher than those of moxifloxacin
g/mL), gatifloxacin (0.06 g/mL) and ciprofloxacin
g/mL). Given that at least a portion of the antibacterial activ-
(0.03
(0.12
l
l
l
2. Chemistry
ity of the prodrugs is likely to come from parent drug that is re-
leased from the bisphosphonate conjugate during the course of
the 24 h MIC assay, it is clear that the prodrugs lack significant
antibacterial activity of their own.
In this process, moxifloxacin 1a, gatifloxacin 1b and ciprofloxa-
cin 1c were selected as starting materials for the prodrugs. These
can be easily protected as the Boc carbamates (2a–c) under stan-
dard conditions to leave the carboxylate as a readily modifiable
functionality (Scheme 1).
To demonstrate in vivo activity, the bisphosphonated prodrugs
likely need not only to bind to bone but also to release the active
fluoroquinolone. These two requirements were evaluated in vitro
to predict the therapeutic potential of these compounds.
Bone-bound prodrugs are immobilized on a insoluble matrix,
and therefore the affinity of fluoroquinolone prodrugs for osseous
tissues can be estimated by measuring the amount of unbound pro-
drug in supernatant by fluorescence spectroscopy relative to the to-
tal amount of input prodrug following exposure to bone powder in
phosphate buffered saline (PBS) at 37 °C over 1 h (Table 1).
The release of the parent fluoroquinolone from these prodrugs
immobilized on bone powder can similarly be determined by mea-
suring the appearance of fluorescence in the supernatant over
time. This was done in PBS and 50% rat or human serum, to evalu-
ate the potential for enzymatic cleavage (Table 1). In addition to
fluorescence measurements, the nature of the released entity was
also confirmed by determining its antibacterial activity, and in all
cases, this activity was identical to that of the parent fluoroquino-
lone (data not shown).
Simple bisphosphonated aliphatic esters can be prepared from
iodides 6a–b which are obtained through the alkylation of tetra-
ethyl methylenebisphosphonate with protected
x-bromoalcohols
4a–b followed by deprotection and iodination (Scheme 2). These
iodides 6a–b undergo a displacement reaction with the carboxyl-
ate of 2a, followed by the simultaneous deprotection of the phos-
phonates and removal of the Boc protecting group on the amine
O
N
O
O
N
O
F
F
a
OH
OH
R2
R3
R1
R1
H
H
N
BocN
N
1
2
1a R =OMe, R =
2a R1=OMe, R3 =
HN
H
H
1b R1=OMe, R2 =
N
2b R1=OMe, R3 =
N
The results from these in vitro assays suggest several trends.
First, the fluoroquinolone–bisphosphonate prodrugs are very effi-
cient at binding bone powder, being taken up at >80% and gener-
ally >90% over 1 h, while the parent drugs are at best negligibly
bound. In fact, it is reasonable to assume that at least part of the
unbound fraction results from cleavage of the prodrug during the
HN
HN
BocN
R3 =
1c R1=H,
R2 =
2c R1=H,
N
N
BocN
Scheme 1. Reagents: (a) Boc2O, NaOH, THF, H2O.

K. S. E. Tanaka et al. / Bioorg. Med. Chem. 16 (2008) 9217–9229
9219
O
O
O
O
O
P(OR)2
P(OR)2
F
P(OR)2
a
d
O
P(OR)2
n P(OR)2
+
THPO
n Br
X
H
n
O
N
N
O
P(OR)2
R'N
O
MeO
H
8a R = Et, n = 1, R' = Boc
8b R = Et, n = 2, R' = Boc
5a R = Et, n = 1, X = -OTHP
4a n=1
4b n=2
3a R = Et
3b R = iPr
5b R = Et, n = 2, X = -OTHP
5c R = iPr, n = 1, X = -OTHP
e
b
9a-b R = H, R' = H
6a-c X = -OH
7a-c X = -I
c
Scheme 2. Reagents and condition: (a) NaH, THF,
D. (b) pTsOH, MeOH. (c) I2, PPh3, imidazole, CH2Cl2. (d) Compound 2a, K2CO3, DMF. (e) TMSBr, CH2Cl2.
O
O
N
O
F
P(OR)2
S
H
R'
P(OR)2
N
O
N
MeO
f
H
O
O
d,e
11a R = iPr, R' = Boc
12a R = H, R' = H
P(OiPr)2
P(OiPr)2
HS
I
h
P(OiPr)2
P(OiPr)2
O
O
O
O
O
7c
10
g
F
P(OR)2
S
P(OR)2
N
N
O
N
h
MeO
R'
11b R = iPr, R' = Boc
12b R = H, R' = H
Scheme 3. Reagents and conditions: (a) NaH, THF,
D. (b) Amberlyst, MeOH. (c) I2, PPh3, imidazole, CH2Cl2. (d) Thiourea, EtOH, D. (e) NaOH, H2O, D. (f) Compound 2a, 2-fluoro-
1-methylpyridinium tosylate, Et3N, CH2Cl2. (g) Compound 2b, 2-fluoro-1-methylpyridinium tosylate, Et3N, CH2Cl2. (h) TMSBr, CH2Cl2.
of hydrolysis, as can be seen with 23a–b and 38. The use of a ter-
tiary glycolamide, which in most instances of glycolamide pro-
drugs accelerates cleavage,16 also results in a reduction, as can be
seen with 28a–b. Unlike the situation with the simple fluoroquin-
olone–bisphosphonate esters, the use of a thioglycolamide ester,
for compound 32, results in a decrease in the rate of cleavage.
Interestingly, prodrug 38 exemplifies the versatility of the bis-
phosphonate osteotropic moiety, with a single bisphosphonate
being able to target two fluoroquinolones to bone.
OH
O
N
O
F
a
O
O
N
P(OR)2
N
MeO
(EtO)2P P(OEt)2
O O
R'
P(OR)2
O
13
14 R = Et, R' = Boc
15 R = H, R' = H
b
The trends observed with the glycolamide esters would suggest
that the bisphosphonate functionality can participate in the cleav-
age, possibly by serving as an intramolecular acid catalyst. The
more electron rich structures of moxifloxacin 1a and gatifloxacin
1b present carbonyl groups which are more readily protonated
or are more apt to coordinate a Lewis acidic cation in the medium,
and their prodrugs are more readily cleaved than the more electron
poor derivatives of ciprofloxacin 1c. These two factors tend to sug-
gest an intramolecular protonation by the bisphosphonate as a rate
determining event.
Scheme 4. Reagents: (a) Compound 2b, 2-fluoro-1-methylpyridinium tosylate,
Et3N, CH2Cl2. (b) TMSBr, CH2Cl2.
time-course of the assay, thereby under-representing the true
affinity of the prodrugs for bone powder. Second, it is clear that
simple bisphosphonated esters are not able to release the parent
drugs. Compounds 9a–b are not cleaved at all, while the thioesters
12a–b and aryl ester 15 release the parent fluoroquinolones only
very slowly. With all of these compounds, the rates of cleavage
are not impacted by the presence of serum, suggesting that they
are poor serum esterase substrates.
4. In vivo evaluation of prodrugs
In contrast, the glycolamide esters are able to appreciably re-
lease their antibacterial moieties. Compounds 18a–b display simi-
lar rates of cleavage in PBS and in sera, suggesting that the process
at least does not require the participation of enzymes. Compound
18c differs in the sense that it is not hydrolyzed in simple buffer
but only in sera. This result does not however preclude the possi-
bility that this is a result of medium effects rather than enzymatic
cleavage.
Prodrugs 18a and 18b appreciably release their parent fluoro-
quinolones in vitro in a process which does not require the partic-
ipation of an enzyme. Such a feature is attractive since it is
predicted to minimize potential discrepancies during translation
of animal efficacy results to humans. As such, these compounds
were selected for evaluation in vivo.
Compounds 18a–b were tested in a rat model for the prevention
of osteomyelitis, which was adapted from the known parent treat-
ment models.17 Briefly, the animals were injected with a single bo-
Lengthening the distance between the fluoroquinolone glycola-
mide ester and the bisphosphonate functionality decreases the rate

9220
K. S. E. Tanaka et al. / Bioorg. Med. Chem. 16 (2008) 9217–9229
O
N
O
O
N
O
O
O
F
O
a
b
O
F
Br
O
NH
NH
NH
N
H
R'
N
(RO)2P P(OR)2
O O
(EtO)2P P(OEt)2
O O
N
N
MeO
(RO)2P P(OR)2
O O
R'
MeO
H
16
17a R = Et, R' = Boc
18a R = H, R' = H
17b R = Et, R' = Boc
18b R = H, R' = H
d
d
c
O
N
O
O
F
O
NH
N
(RO)2P P(OR)2
O O
N
R'
17c R = Et, R' = Boc
18c R = H, R' = H
d
Scheme 5. Reagents: (a) Compound 2a, Cs2CO3, DMF. (b) Compound 2b, Cs2CO3, DMF. (c) Compound 2c, Cs2CO3, DMF. (d) TMSBr, CH2Cl2.
NO2
NH2
O
a
b
P(OMe)2
O
O
P(OMe)2
P(OMe)2
P(OMe)2
P(OMe)2
P(OMe)2
O
O
O
O
19
20
c
O
O
O
N
O
O
O
F
F
Br
O
O
NH
e
NH
NH
R'
N
N
N
d
N
MeO
N
MeO
R'
O
O
O
P(OMe)2
P(OR)2
P(OR)2
P(OR)2
P(OMe)2
P(OR)2
O
O
O
22a R = Me, R' = Boc
22b R = Me, R' = Boc
21
f
f
23a R = H, R' = H
23b R = H, R' = H
Scheme 6. Reagents: (a) NaH, THF, DMF. (b) H2, PtO2. (c) Bromoacetyl bromide, pyridine, CH2Cl2. (d) Compound 2a, Cs2CO3, DMF. (e) Compound 2b, Cs2CO3, DMF. (f) TMSBr,
CH2Cl2.
Bn
HN
N
Bn
a
b
N
H
(EtO)2P P(OEt)2
O O
(EtO)2P P(OEt)2
O O
24
25
c
O
N
O
O
N
O
O
O
F
O
d
e
O
F
Br
O
N
N
N
H
N
R'
N
(RO)2P P(OR)2
O O
(EtO)2P P(OEt)2
O O
N
N
MeO
(RO)2P P(OR)2
O O
R'
MeO
H
26
27a R = Et, R' = Boc
28a R = H, R' = H
27b R = Et, R' = Boc
28b R = H, R' = H
f
f
Scheme 7. Reagents and conditions: (a) Diethylphosphite, HC(OEt)3,
DMF. (e) Compound 2b, Cs2CO3, DMF. (f) TMSBr, CH2Cl2.
D
. (b) Cyclohexene, Pd/C, EtOH, D. (c) Bromoacetyl bromide, pyridine, CH2Cl2. (d) Compound 2a, Cs2CO3,
lus intravenous dose of the prodrug two days before an infection of
the bone was established by surgically instilling S. aureus bacteria
(ATCC 13709) into the tibiae. The animals were sacrificed 24 h after
challenge and the bacterial titer in the tibiae was determined.
Moxifloxacin was administered 1 h after infection as the positive
control. It should be noted that moxifloxacin is ineffective when

K. S. E. Tanaka et al. / Bioorg. Med. Chem. 16 (2008) 9217–9229
9221
O
O
O
O
F
NH2
HS
S
a
b
NH
NH
H
R'
(EtO)2P P(OEt)2
O O
N
N
N
(EtO)2P P(OEt)2
(RO)2P P(OR)2
O O
MeO
c
O O
H
29
30
31 R = Et, R' = Boc
32 R = H, R' = H
Scheme 8. Reagents and conditions: (a) Mercaptoacetic acid,
D
. (b) Compounds 2a–b, 2-fluoro-1-methylpyridinium tosylate, Et3N, CH2Cl2. (c) TMSBr, CH2Cl2.
O
O
O
P(OEt)2
P(OEt)2
O
O
H
N
O
O
P(OEt)2
P(OEt)2
Br
N
H
BocHN
OH
XHN
N
O
c
a
H
O
NH
Br
NHBoc
NHX
33
34 X = Boc
36
b
35 X = H
d
O
O
O
O
P(OR)2
P(OR)2
H
N
F
O
O
N
N
H
O
O
O
N
N
N
MeO
R'
F
NH
O
37 R = Et, R' = Boc
O
e
N
38 R = H, R' = H
N
MeO
R'
Scheme 9. Reagents: (a) Compound 29, N-(3-dimethylaminopropyl)-N0-ethylcarbodiimide hydrochloride, DMAP, CH2Cl2. (b) TFA, CH2Cl2. (c) Bromoacetyl bromide, pyridine,
CH2Cl2. (d) Compound 2b, Cs2CO3, DMF. (e) TMSBr, CH2Cl2.
These experiments and the in vitro evaluations of the prodrugs
are consistent with the ability of the bisphosphonate moiety to de-
liver the fluoroquinolone to the bone and release it over time to
sustain a concentration sufficient to significantly reduce the bacte-
rial titer in bone beyond a point in time by which the parent drug is
already cleared, and is completely ineffective.
Table 1
Bone binding and regeneration of parent fluoroquinolone from prodrugs
Compound
Amount bound
to bone (%)
Amount of parent drug released over 24 h (%)
PBS
50% rat serum
50% human serum
1a
1b
1c
9a
<l.o.d.a
<l.o.d.
<l.o.d.
93.8
92.7
99.9
98.7
79.2
98.7
91.0
97.2
97.2
88.9
97.3
90.4
99.0
92.8
—
—
—
—
—
—
—
—
—
<l.o.d.
0.01
0.04
0.1
0.2
1.2
2.7
0.01
0.6
0.5
0.3
1.1
0.3
0.2
<l.o.d.
<l.o.d.
0.1
0.1
0.1
1.4
1.3
0.5
1.3
0.6
0.7
0.9
0.4
0.4
<l.o.d.
<l.o.d.
0.1
0.1
0.1
1.4
1.6
0.4
1.2
0.7
0.8
0.6
0.5
0.4
5. Conclusion
9b
12a
12b
15
A number of methods were presented to tether a bisphospho-
nate group to the carboxylic acid functionality of moxifloxacin,
gatifloxacin and ciprofloxacin. The resulting fluoroquinolone bis-
phosphonate conjugates were evaluated for their affinity to bone
and their ability to release their parent drug once bound to bone.
In vitro investigations highlighted the strong affinity of meth-
ylenebisphosphonate derivatives for bone, as opposed to the negli-
gible bone affinities of the parent drugs.
This study presents the successful combination of bisphospho-
nate-based drug delivery and the known glycolamide ester class
of prodrugs. These bisphosphonated esters were shown to release
their parent drugs once they were bound to bone, under circum-
stances where simple bisphosphonated esters and even thioesters
only produced negligible rates of regeneration.
18a
18b
18c
23a
23b
28a
28b
32
38
a
Below the limit of detection.
used 24 h after the infection at that dose, likely a result of its rela-
tively rapid clearance in vivo.
Bisphosphonate fluoroquinolone conjugates 18a–b were further
evaluated in vivo for their ability to prevent infection in a rat mod-
el of osteomyelitis. They were shown to prevent the establishment
of infection while moxifloxacin and gatifloxacin, their respective
parent drugs, were unable to do so.
These prodrugs are being further evaluated in well-established
animal models for their ability to treat chronic osteomyelitis, and
the results of these investigations will be presented in due course.
This experiment was performed with 18a at 15.8 mg/kg of body
weight, and compared with the equivalent dose of moxifloxacin
(10 mg/kg), both administered two days prior to the establishment
of infection (Fig. 1). Whereas the osteotropic prodrug clearly
yielded an efficacious outcome, the parent drug was unable to pre-
vent the infection under these conditions. The same experiment
was performed with 18b at 17.3 mg/kg and compared with gati-
floxacin at 10 mg/kg, with the same conclusions (Fig. 2).

9222
K. S. E. Tanaka et al. / Bioorg. Med. Chem. 16 (2008) 9217–9229
8
7
6
5
4
3
2
1
0
limit of
detection
Untreated
Moxifloxacin
10mg/kg 1h after
Moxifloxacin
10mg/kg
18a 15.8mg/kg
Figure 1. Prophylactic effect by prodrug 18a and moxifloxacin in the rat bone infection model when injected at 15.8 mg/kg and 10 mg/kg, respectively, 2 days prior to the
establishment of infection.
8
7
6
5
4
3
limit of
detection
2
1
0
Untreated
Moxifloxacin
10mg/kg 1h after
Gatifloxacin
10mg/kg
18b 17.3mg/kg
Figure 2. Prophylactic effect by prodrug 18b and gatifloxacin in the rat bone infection model when injected at 17.3 mg/kg and 10 mg/kg, respectively, 2 days prior to the
establishment of infection.
were approved by the institutional animal care and use
committee.
6. Experimental
6.1. General
6.2. Chemistry
1H NMR spectra were recorded on a Varian MercuryTM 400 spec-
trometer. The reported chemical shifts (in parts per million) are
referenced using the signals assigned to the residual non-deuter-
ated solvents or to tetramethylsilane as the internal standard. Mass
spectral analyses were performed on an Agilent mass spectrometer
under electron spray ionization (ESI). Reactions were monitored by
TLC on Silica gel Gel 60 F254 (0.25 mm, Merck). Column chroma-
tography was performed on silica gel (70–230 mesh). All reactions
were carried out under an argon atmosphere with anhydrous sol-
vents, unless otherwise noted. All chemicals, unless otherwise sta-
ted, were obtained from commercial sources.
6.2.1. N-Boc protection of fluoroquinolones
General procedure. A mixture of the fluoroquinolone (2 mmol),
Boc2O (460 mg, 2.1 mmol) and 4.2 mL of 1 M NaOH aqueous solu-
tion in 20 mL of THF was stirred at room temperature overnight.
For 2a–b, after the removal of the organic solvent, the residue
was neutralized with saturated ammonium chloride aqueous
solution. The mixture was extracted with ethyl acetate (3ꢀ) and
dried over anhydrous sodium sulfate. The combined organics
were concentrated in vacuo to provide the product. For 2c, the
product precipitates from the reaction mixture and was collected
by filtration.
Compound purities were evaluated by LC/MS on an Agilent
1100 HPLC system using
a
Waters XTerra RP18 (3.5 l,
6.2.1.1. 7-((4aS,7aS)-1-(tert-Butoxycarbonyl)-octahydropyrrolo-
[3,4-b]pyridin-6-yl)-1-cyclopropyl-6-fluoro-1,4-dihydro-8-metho-
xy-4-oxoquinoline-3-carboxylic acid (2a). Yield: 91%. 1H NMR
(400 MHz, CDCl3): d 0.79–0.86 (m, 1H), 1.03–1.18 (m, 2H), 1.23–
1.34 (m, 2H), 1.44–1.54 (m, 1H), 1.49 (s, 9H), 1.76–1.84 (m, 2H),
2.25–2.29 (m, 1H), 2.89 (t, J = 11.8, 1H), 3.22–3.30 (m, 1H), 3.38
(br s, 1H), 3.57 (s, 3H), 3.88 (dt, J = 2.7, 10.0, 1H), 3.96–4.01 (m,
1H), 4.07–4.12 (m, 2H), 4.79 (br s, 1H), 7.82 (d, J = 13.7, 1H), 8.79
(s, 1H) ppm.
2.1 mm ꢀ 50 mm) column at three different wavelengths using
the following methods:
– Method A. Eluents: A (0.1% HCO2H in H2O) and B (0.1%
HCO2H in acetonitrile). Method: 0–50% B in A over 6 min,
then 50–100% B in A over 2 min then 100% B for 1 min at a
flow rate of 0.3 mL/min.
– Method B. Eluents: A (10 mM NH4OAc) and B (MeOH).
Method: 0–50% B in A over 6 min, then 50–100% B in A over
2 min at a flow rate of 0.3 mL/min.
6.2.1.2. 7-(4-(tert-Butoxycarbonyl)-3-methylpiperazin-1-yl)-1-
cyclopropyl-6-fluoro-1,4-dihydro-8-methoxy-4-oxoquinoline-3-
carboxylic acid (2b). Yield: 95%. 1H NMR (400 MHz, CDCl3): d
0.94–1.04 (m, 2H), 1.19–1.26 (m, 2H), 1.33 (d, J = 6.9, 3H), 1.50 (s,
9H), 3.23–3.37 (m, 3H), 3.44–3.51 (m, 2H), 3.73 (s, 3H), 3.95–
4.03 (m, 2H), 4.36 (br s, 1H), 7.89 (d, J = 11.4, 1H), 8.83 (s, 1H).
– Method C. Eluents: A (10 mM NH4OAc) to B (MeOH).
Method: 20–70% B in A over 6 min, then 70–100% B in A over
2 min at a flow rate of 0.3 mL/min.
All animal experiments followed the Canadian Council of Ani-
mal guidelines in an accredited animal facility, and protocols

K. S. E. Tanaka et al. / Bioorg. Med. Chem. 16 (2008) 9217–9229
9223
6.2.1.3. 7-(4-(tert-Butoxycarbonyl)piperazin-1-yl)-1-cyclopropyl-
6.2.2.5. Tetraethyl
5-hydroxypentylene-1,1-bisphosphonate
6-fluoro-1,4-dihydro-4-oxoquinoline-3-carboxylic acid (2c).
Yield: 78%. 1H NMR (400 MHz, CDCl3) d 1.89–1.23 (m, 2H), 1.37–
1.43 (m, 2H), 1.50 (s, 9H), 3.28 (bd, J = 5.0, 4H), 3.51–3.56 (m,
1H), 3.67 (bt, J = 5.0, 4H), 7.37 (d, J = 7.3, 1H), 8.05 (d, J = 13.0,
1H), 8.78 (s, 1H).
(6b). From 5b using p-toluenesulfonic acid monohydrate as the
acidic catalyst. The eluent for chromatography was 1/15:1/6 meth-
anol/ethyl acetate. Yield: 41% from tetraethyl meth-
ylenebisphosphonate (2 steps). 1H NMR (400 MHz, CDCl3): d
1.24–1.36 (m, 12H), 1.55–1.72 (m, 4H), 1.89–2.03 (m, 2H), 2.16
(br s, 1H), 2.29 (tt, J = 6.1, 24.3, 1H), 3.66 (br s, 2H), 4.11–4.22 (m,
8H).
6.2.2. Iodoalkylene-1,1-bisphosphonate esters
General procedure for the preparation of
x-(2-tetrahydro-2H-pyr-
anyloxy)alkylene-1,1-bisphosphonates. To a suspension of NaH (60%
suspension in mineral oil, 900 mg, 22.0 mmol) in dry THF (20 mL)
was added dropwise the methylenebisphosphonate tetraester 3a
or 3b (22.4 mmol). The resulting clear solution was stirred 15–
45 min at room temperature, and 2-(x-bromoalkoxy)tetrahydro-
2H-pyran 4a or 4b (22.6 mmol) was added dropwise. The reaction
6.2.2.6. Tetraisopropyl 4-hydroxybutylene-1,1-bisphosphonate
(6c). From 5b using amberlyst 15 as the acidic catalyst. The eluent
for chromatography was 0–10% methanol/ethyl acetate. Yield: 48%
from tetraisopropyl methylenebisphosphonate (2 steps). 1H NMR
(400 MHz, CDCl3) d 1.33–1.36 (m, 24H), 1.77–1.83 (m, 1H), 1.96–
2.10 (m, 2H), 2.21 (tt, J = 24.8, 5.4, 1H), 2.31–2.42 (m, 2H), 3.66
(t, J = 5.9 2H), 4.70–4.83 (m, 4H).
mixture was heated to reflux for 6 h and worked up as indicated.
General procedure for the preparation of
x-iodoalkylene-1,1-bis-
6.2.2.1. Tetraethyl
4-(2-tetrahydro-2H-pyranyloxy)butylene-
phosphonates. To a solution of -hydroxyalkylene-1,1-bisphospho-
x
1,1-bisphosphonate (5a)19. From 3a and 4a. After reflux, the reac-
tion mixture was diluted with CH2Cl2 (75 mL) washed with brine
(2ꢀ 50 mL), dried (MgSO4) and evaporated. The crude product
was used as such in the following step. 1H NMR (400 MHz, CDCl3)
d 1.33 (t, J = 7.1, 12H), 1.44–2.14 (m, 10H), 2.35 (tt, J = 24.1, 6.1,
1H), 3.52 (m, 2H), 3.86 (m, 2H), 4.17 (m, 8H), 4.58 (m, 1H). 31P
NMR (162 MHz, CDCl3) d 24.98 (s, 2P).
nate 6a–c (4.39 mmol) in CH2Cl2 (50 mL) were added
triphenylphosphine (1.32 g, 5.033 mmol) and imidazole (0.45 g,
6.61 mmol). The reaction mixture was cooled to 0 °C, before the
addition of iodine (1.22 g, 4.81 mmol). The mixture was then re-
moved from the cooling bath, stirred for 2 h and worked up as
indicated.
6.2.2.7. Tetraethyl 4-iodobutylene-1,1-bisphosphonate (7a). From
6a. The reaction mixture was diluted with hexanes (100 mL) and fil-
tered washing the precipitate with further hexanes (2ꢀ 30 mL). The
filtrate was concentrated and purified by flash chromatography on
silica gel using 0–10% methanol/ethyl acetate as the eluent to give
pure 7a. Yield: 80%. 1H NMR (400 MHz, CDCl3) d 1.32–1.38 (m,
12H), 1.95–2.15 (m, 4H), 2.28 (tt, J = 24.1, 6.1, 1H), 3.18 (t, J = 6.6,
2H), 4.12–4.24 (m, 8H).
6.2.2.2. Tetraethyl 5-(2-tetrahydro-2H-pyranyloxy)pentylene-
1,1-bisphosphonate (5b). From 3a and 4b. After reflux, the reac-
tion mixture was quenched with saturated aqueous solution of
ammonium chloride and a small amount of water to dissolve all
solids. The mixture was extracted with ethyl acetate (3ꢀ), dried
over anhydrous sodium sulfate and concentrated in vacuo. Flash
chromatography on silica gel with 20:1 (v/v) dichloromethane/
methanol as the eluent afforded impure product 5b as a slightly
yellow oil. The material was used directly in the next step without
further purification. Selected 1H NMR signals (400 MHz, CDCl3): d
1.32 (t, J = 7.1,12H), 1.45–2.00 (m, 12H), 2.28 (tt, J = 6.1, 24.3,
1H), 3.44 (m, 2H), 3.80 (m, 2H), 4.17 (m, 8H), 4.57 (m, 1H). 31P
NMR (162 MHz, CDCl3) d 25.08 (s, 2P).
6.2.2.8. Tetraethyl 5-iodopentylene-1,1-bisphosphonate (7b). From
6b. The reaction mixture was concentrated in vacuo and the residue
was taken up in a mixture of ethyl acetate and saturated Na2S2O3 aque-
ous solution. The mixture was stirred until the organic layer turned
pale yellow and the two phases were separated. The organic phase
was dried over anhydrous sodium sulfate and concentrated. Flash chro-
matography on silica gel with 15:1 ethyl acetate/methanol as the elu-
ent afforded the product 7b as a yellow oil. Yield: 68%. 1H NMR
(400 MHz, CDCl3): d 1.36 (t, J = 7.0, 12H), 1.66–1.72 (m, 2H), 1.81–
1.99 (m, 4H), 2.35 (tt, J = 5.9, 24.1, 1H), 3.20 (t, J = 6.9, 2H), 4.17–4.23
(m, 8H).
6.2.2.3. Tetraisopropyl 4-(2-tetrahydro-2H-pyranyloxy)butyl-
ene-1,1-bisphosphonate (5c). From 3b and 4a. After reflux, the
reaction mixture was concentrated, the residue taken up in ethyl
acetate and washed with semi-saturated brine. The aqueous layer
was extracted with ethyl acetate and the combined organics
washed with brine, dried over MgSO4 and concentrated to dryness.
It was used as such in the following step. 1H NMR (400 MHz, CDCl3)
d 1.33–1.36 (m, 24H), 1.45–2.06 (m, 8H), 2.13 (m, 2H), 2.19 (tt,
J = 24.3, 5.9, 1H), 3.51 (m, 2H), 3.86 (m, 2H), 4.59 (m, 1H), 4.67
(m, 4H).31P NMR (162 MHz, CDCl3) d 23.00 (s, 2P).
6.2.2.9. Tetraisopropyl
4-iodobutylene-1,1-bisphosphonate
(7c). From 6c. The reaction mixture was added to hexanes
(300 mL) and filtered washing the precipitate with further hex-
anes (2ꢀ 50 mL). The filtrate was evaporated and purified by
flash chromatography on silica gel eluting with ethyl acetate
to give pure 7c. Yield: 85%. 1H NMR (400 MHz, CDCl3) d 1.33–
1.37 (m, 24H), 1.92–2.23 (m, 5H), 3.18 (t, J = 6.7, 2H), 4.74–
4.83 (m, 4H).
General procedure for the preparation of
bisphosphonates. To a solution of the crude
x
-hydroxyalkylene-1,1-
-(2-tetrahydro-2H-
x
pyranyloxy)alkylene-1,1-bisphosphonate 5a–c (0.5 M) in methanol
was added a catalytic amount of the indicated acid and the mixture
was heated to reflux until consumption of the starting material as
indicated by TLC (4–24 h). If an acidic resin was used, it was fil-
tered off, and the solution was concentrated in vacuo and the res-
idue was purified by silica gel chromatography.
6.2.3. Simple bisphosphonated fluoroquinolone esters
6.2.3.1. 4,4-Bis(diethylphosphono)butyl
7-((4aS,7aS)-1-(tert-
butoxycarbonyl)-octahydropyrrolo[3,4-b]pyridin-6-yl)-1-cyclo-
propyl-6-fluoro-1,4-dihydro-8-methoxy-4-oxoquinoline-3-car-
6.2.2.4. Tetraethyl 4-hydroxybutylene-1,1-bisphosphonate (6a).
From 5a using amberlite IR-120 as the acidic catalyst. The eluent
for chromatography was 5–10% methanol/ethyl acetate. Yield:
34% from tetraethyl methylenebisphosphonate (2 steps). 1H NMR
(400 MHz, CDCl3) d 1.34 (t, J = 7.1, 12H), 1.81 (quint, J = 6.5, 2H),
1.99–2.13 (m, 2H), 2.37 (tt, J = 24.4, 5.6, 1H), 3.66 (q, J = 5.9, 2H),
4.13–4.22 (m, 8H).
boxylate (8a). A mixture of 2a (576 mg, 1.15 mmol), iodo
bisphosphonate 7a (497 mg, 1.09 mmol) and potassium carbonate
(151 mg, 1.09 mmol) in 10 mL anhydrous DMF was stirred at room
temperature for 21 h. Ethyl acetate (100 mL) was added, and the
organics extracted with water (3ꢀ 20 mL) and brine (20 mL), and
dried over MgSO4. Flash chromatography on silica gel with gradi-
ent elution from 5 to 10% methanol/ethyl acetate afforded the pure

9224
K. S. E. Tanaka et al. / Bioorg. Med. Chem. 16 (2008) 9217–9229
product (518 mg, 57%). 1H NMR (400 MHz, CDCl3) d 0.72–0.80 (m,
1H), 0.92–1.10 (m, 1H), 1.16–1.30 (m, 2H), 1.33 (t, J = 7.0, 12H),
1.47 (s, 9H), 1.71–1.83 (m, 4H), 2.05–2.16 (m, 4H), 2.18–2.28 (m,
1H), 2.32–2.50 (m, 1H), 2.81–2.94 (m, 1H), 3.13–3.26 (m, 1H),
3.28–3.42 (m, 1H), 3.55 (s, 3H), 3.78–3.90 (m, 2H), 3.98–4.08 (m,
2H), 4.12–4.23 (m, 8H), 4.25–4.36 (m, 2H), 4.76 (br s, 1H), 7.79
(d, J = 4.3, 1H), 8.51 (s, 1H).
6.2.4. Preparation of bisphosphonated fluoroquinolone
thioesters
6.2.4.1. Tetraisopropyl
5-thiapentylene-1,1-bisphosphonate
(10). To a solution of 7c (3.8 g, 7.4 mmol) in ethanol (20 mL) was
added thiourea (0.59 g, 7.75 mmol). The reaction mixture was re-
fluxed for 18 h, concentrated to dryness and used as such in the fol-
lowing step. 1H NMR (400 MHz, D2O) d 1.35–1.38 (m, 24 H), 1.94–
2.09 (m, 4 H), 2.50–2.67 (m, 1 H), 3.17 (t, J = 6.1, 2 H), 4.70–4.85 (m,
4 H). To this residue in water (30 mL) was added sodium hydroxide
(0.396 g, 9.90 mmol). The reaction mixture was refluxed for 1.5 h,
cooled to 0 °C and acidified with 1 M HCl (10 mL). The product
was extracted with CHCl3 (3ꢀ 50 mL), the organics washed with
brine (70 mL), dried (MgSO4) and concentrated to give a quantita-
tive yield of crude 10 used as such in the following steps. 1H NMR
(400 MHz, CDCl3) d 1.33–1.36 (m, 24 H), 1.88–2.19 (m, 5 H), 2.50–
2.56 (m, 2 H), 4.74–4.83 (m, 4 H).
General procedure for coupling of thiol 10 to fluoroquinolones. To a
solution of N-Boc protected fluoroquinolone (0.40 mmol) in CH2Cl2
(3 mL) was added 2-fluoro-1-methylpyridinium tosylate (0.136 g,
0.48 mmol). The reaction mixture was cooled to 0 °C, and triethyl-
amine (0.20 mL, 1.43 mmol) was added via syringe. After stirring
1 h at 0 °C a solution of thiol 10 (0.208 g, 0.497 mmol) in CH2Cl2
(3 mL) was added. After a further 1 h at 0 °C the reaction was al-
lowed to warm to room temperature overnight. The reaction mix-
ture was diluted with ethyl acetate and washed with ice cold
saturated NH4Cl solution, 5% NaHCO3, and water. After drying
(MgSO4) and concentration under vacuum, the residue was puri-
fied by flash chromatography on silica gel with gradient elution
from 2.5 to 5% methanol/CH2Cl2 to give the product.
6.2.3.2. 5,5-Bis(diethylphosphono)pentyl 7-((4aS,7aS)-1-(tert-
butoxycarbonyl)-octahydropyrrolo[3,4-b]pyridin-6-yl)-1-cyclo-
propyl-6-fluoro-1,4-dihydro-8-methoxy-4-oxoquinoline-3-car-
boxylate (8b). The mixture containing compound 2a (464.7 mg,
0.9265 mmol), iodo bisphosphonate 7 b (435.5 mg, 0.9262 mmol)
and potassium carbonate (129.3 mg, 0.9355 mmol) in 15 mL of
anhydrous DMF was heated at 65 °C for 2d. After cooling to room
temperature, the reaction was diluted with water, extracted with
ethyl acetate (3ꢀ), dried over anhydrous sodium sulfate and con-
centrated. Flash chromatography on silica gel with gradient elu-
tion from 15:1 ethyl acetate/methanol to 8:1 then 5:1 afforded
425.6 mg (54%) product 8b as
(400 MHz, CDCl3): 0.72–0.80 (m, 1H), 1.00–1.08 (m, 2H),
a
yellow foam. 1H NMR
d
1.24–1.36 (m, 13H), 1.48 (s, 9H), 1.63–1.81 (m, 8H), 1.92–2.04
(m, 2H), 2.20–2.28 (m, 1H), 2.36 (tt, J = 5.8, 24.1, 1H), 2.82–2.92
(m, 1H), 3.16–3.24 (m, 1H), 3.30–3.38 (m, 1H), 3.56 (s, 3H),
3.80–3.85 (m, 1H), 3.89–3.94 (m, 1H), 4.01–4.08 (m, 2H), 4.12–
4.21 (m, 8H), 4.29–4.36 (m, 2H), 4.77 (br s, 1H), 7.78 (d,
J = 14.4, 1H), 8.55 (s, 1H).
General procedure for bisphosphonated deprotection. TMSBr
(0.82 mL, 6.21 mmol) was added in one portion to a stirring solu-
tion of the protected bisphosphonate (518 mg, 0.624 mmol) in
CH2Cl2 (50 mL) and the resulting mixture was stirred at room tem-
perature for 1 d. The solvent was removed under reduced pressure
and the solid was dried under high vacuum for 1 h.
6.2.4.2. S-4,4-Bis(diisopropylphosphono)butyl 7-((4aS,7aS)-1-
(tert-butoxycarbonyl)-octahydropyrrolo[3,4-b]pyridin-6-yl)-1-
cyclopropyl-6-fluoro-1,4-dihydro-8-methoxy-4-oxoquinoline-
3-carbothioate (11a). Yield: 67%. 1H NMR (400 MHz, CDCl3) d
0.73–0.82 (m, 1H), 0.97–1.11 (m, 2H), 1.24–1.27 (m, 1H), 1.32–
1.36 (m, 24H), 1.48 (s, 9H), 1.59–2.28 (m, 10H), 2.82–2.93 (m,
1H), 2.97 (t, J = 7.2, 2H), 3.17–3.26 (m, 1H), 3.30–3.43 (m, 1H),
3.55 (s, 3H), 3.78–3.95 (m, 2H), 4.05–4.12 (m, 2H), 4.72–4.85 (m,
5H), 7.84 (d, J = 13.9, 1H), 8.54 (s, 1H).
6.2.3.3. 4,4-Bisphosphonobutyl 7-((4aS,7aS)-octahydropyrrol-
o[3,4-b]pyridin-6-yl)-1-cyclopropyl-6-fluoro-1,4-dihydro-8-meth-
oxy-4-oxoquinoline-3-carboxylate (9a). The solid was then re-
suspended in H2O and the pH was immediately adjusted to pH 7
by the addition of 1 M NaOH, with concomitant dissolution of
the product. The product solution was washed with CHCl3, filtered
and concentrated to dryness to give the product in quantitative
yield. 1H NMR (400 MHz, D2O) d 0.83 (br s, 1H), 0.98 (br s, 1H),
1.08 (br s, 1H), 1.21 (br s, 1H), 1.65–2.15 (m, 9H), 2.61 (br s, 1H),
2.93 (br s, 1H), 3.20–3.35 (m, 1H), 3.43–3.64 (m, 2H), 3.52 (s,
3H), 3.75 (br s, 2H), 3.84–4.17 (m, 2H), 4.31 (br s, 2H), 7.41 (d,
6.2.4.3. S-4,4-Bis(diisopropylphosphono)butyl 7-(4-(tert-butox-
ycarbonyl)-3-methylpiperazin-1-yl)-1-cyclopropyl-6-fluoro-1,4-
dihydro-8-methoxy-4-oxoquinoline-3-carbothioate (11b). Yield:
60% contaminated with a small amount of 2b. 1H NMR (400 MHz,
CDCl3) d 0.88–1.01 (m, 2H), 1.10–1.27 (m, 2H), 1.31–1.39 (m,
27H), 1.49 (s, 9H), 1.85–2.28 (m, 5H), 2.97 (t, J = 7.4, 2H), 3.18–
3.52 (m, 5H), 3.71 (s, 3H), 3.80–4.05 (m, 2H), 4.34 (br s, 1H),
4.72–4.87 (m, 4H), 7.91 (d, J = 12.5, 1H), 8.57 (s, 1H).
J = 12.1,
1H),
8.80
(s,
1H).
ESI-MS:
calculated
for
(C25H34FN3O10P2+H+): 618; found: 618 (M+H+).
6.2.3.4. 5,5-Bisphosphonopentyl 7-((4aS,7aS)-octahydropyrrol-
o[3,4-b]pyridin-6-yl)-1-cyclopropyl-6-fluoro-1,4-dihydro-8-meth-
oxy-4-oxoquinoline-3-carboxylate (9b). The solid was dissolved in
water. The resulting solution was brought to pH 7.4 with 1 N aque-
ous sodium hydroxide and the solvent was removed. The solid was
twice dissolved in water and the solvent removed. The solid ob-
General procedure for deprotection of bisphosphonates. To a solu-
tion of bisphosphonated fluoroquinolone thioester (0.70 mmol) in
CH2Cl2 (50 mL) was added TMSBr (0.93 mL, 7.05 mmol). The reac-
tion mixture was stirred for 65 h, the volatiles removed under re-
duced pressure and the solid dried under high vacuum for 1 h.
The solid was suspended in H2O (200 mL) and the pH was immedi-
ately adjusted to pH 7.5–8 by the addition of 1 M NaOH, with con-
comitant dissolution of the product. The product solution was
filtered washing the insoluble material with water and CHCl3.
The aqueous solution was concentrated to dryness and purified
by reverse-phase chromatography (gradient elution, 100% water–
33% methanol/water) to provide the product.
tained was subjected to
a
WatersÒ C18 Sep-PakTM cartridge
(20 cc) with gradient elution from neat water to 10:1 water/meth-
anol to 5:1 to afford 9b as an off-white solid (211 mg, 70%). 1H
NMR (400 MHz, D2O): d 0.78–0.84 (m, 1H), 0.97–1.10 (m, 2H),
1.17–1.24 (m, 1H), 1.62–2.00 (m, 11H), 2.81 (br s, 1H), 3.04–3.10
(m, 1H), 3.39–3.43 (m, 1H), 3.54 (s, 3H), 3.63–3.68 (m, 2H), 3.80
(dt, J = 3.3, 9.6, 1H), 3.92–3.94 (m, 1H), 4.05–4.08 (m, 2H), 4.25–
4.28 (m, 2H), 7.47 (d, J = 14.1, 1H), 8.76 (s, 1H); 31P NMR
(162 MHz, D2O): d 21.45; 19F NMR (376 MHz, D2O): d ꢁ121.64 (d,
J = 13.8). ESI-MS: calculated for (C26H36FN3O10P2-H+): 630; found:
630 (MꢁH)ꢁ. Purity (method A): 97.2% (254 nm), 98.5% (220 nm),
97.8% (320 nm).
6.2.4.4. S-4,4-Bisphosphonobutyl 7-((4aS,7aS)-octahydropyrrol-
o[3,4-b]pyridin-6-yl)-1-cyclopropyl-6-fluoro-1,4-dihydro-8-metho-
xy-4-oxoquinoline-3-carbothioate (12a). Yield: 47% based on
tetrasodium salt of product. 1H NMR (400 MHz, D2O) d 1.02–1.11
(m, 1H), 1.13–1.22 (m, 2H), 1.27–1.36 (m, 1H), 1.64–1.98 (m,

K. S. E. Tanaka et al. / Bioorg. Med. Chem. 16 (2008) 9217–9229
9225
9H), 2.42–2.52 (m, 1H), 2.59–2.69 (m, 1H), 2.74–2.84 (m, 1H),
2.95–3.25 (m, 1H), 3.34–3.44 (m, 1H), 3.56–3.70 (m, 2H), 3.61 (s,
3H), 3.83–3.96 (m, 2H), 4.08–4.18 (m, 2H), 7.53 (d, J = 13.7, 1H),
ylenediphosphonate (5.8 g, 25 mmol) in DMF (40 mL). After
30 min a solution of 4-nitrobenzylbromide (5.00 g, 23.1 mmol) in
THF (5 mL) was added and the resulting mixture was stirred at
room temperature for 4.5 h. The reaction was quenched by the
addition of saturated aqueous NH4Cl (20 mL). After the addition
of water (100 mL) the product was extracted with EtOAc and the
combined organics were washed with brine, dried over MgSO4, fil-
tered and concentrated at reduced pressure. The crude product
was purified by silica gel chromatography (0% to 10% MeOH in
EtOAc) resulting in 19 as a colorless solid (2.55 g, 30%). 1H NMR
(400 MHz, CDCl3) d 2.65 (tt, J = 6.5, 23.8, 1H), 3.31 (dt, J = 6.5,
16.5, 2H), 3.73 (d, J = 7.0, 6H), 3.75 (d, J = 7.0, 6H), 7.42 (d, J = 8.9,
2H), 8.15 (d, J = 8.9, 2H).
8.59 (s, 1H).19F (376 MHz, D2O) d ꢁ121.38 (d, J = 13.7, 1F).31
P
(162 MHz, D2O)
d 20.74 (s, 2P). ESI-MS calculated for
(C25H34FN3O9P2S+H+): 634; found: 634 (M+H+). Purity (method
A): 99.7% (254 nm), 99.7% (220 nm), 99.4% (320 nm).
6.2.4.5. S-4,4-Bisphosphonobutyl 7-(3-methylpiperazin-1-yl)-1-
cyclopropyl-6-fluoro-1,4-dihydro-8-methoxy-4-oxoquinoline-3-
carbothioate (12b). Yield: 20% based on tetrasodium salt of
product. 1H NMR (400 MHz, D2O) d 0.96–1.04 (m, 2H), 1.16–1.25
(m, 2H), 1.33 (d, J = 6.3, 3H), 1.76–2.02 (m, 5H), 2.99–3.08 (m,
2H), 3.16–3.59 (m, 7H), 3.76 (s, 3H), 4.06–4.14 (m, 1H), 7.34 (d,
J = 12.0, 1H), 8.66 (s, 1H). 19F (376 MHz, D2O) d ꢁ121.26 (d,
J = 12.0, 1F). 31P (162 MHz, D2O) d 20.80 (s, 2P). MS: (MH+) 608.1.
ESI-MS calculated for (C23H32FN3O9P2S+H+): 608; found: 608
(M+H+). Purity (method B): 99.2% (254 nm), 100% (220 nm), 100%
(290 nm).
6.2.6.1.2. Dimethyl
2-(4-aminophenyl)-1-(dimethoxyphospho-
ryl)ethylphosphonate (20). A mixture of 19 (1.01 g, 2.75 mmol)
and PtO2 (0.035 g, 0.15 mmol) in EtOH (40 mL, 95%) was shaken
in a Parr apparatus under 3.8 hPa of H2 for 14 h. The catalyst was
removed by filtration through glass fiber filter paper and the sol-
vent was removed under reduced pressure to give 20 as a pale yel-
low solid (0.959 g, 103%) that was used without purification. 1H
NMR (400 MHz, CDCl3) d 2.62 (tt, J = 6.3, 23.9, 1H), 3.12 (dt,
J = 6.3, 16.2, 2H), 3.70 (d, J = 1.9, 6H), 3.73 (d, J = 1.9, 6H), 6.61 (d,
J = 8.5, 2H), 7.04 (d, J = 8.5, 2H).
6.2.5. Bisphosphonated gatifloxacin aryl ester
6.2.5.1. 4-(Bis(diethylphosphono)methyl)phenyl 7-(4-(tert-butox-
ycarbonyl)-3-methylpiperazin-1-yl)-1-cyclopropyl-6-fluoro-1,4-
dihydro-8-methoxy-4-oxoquinoline-3-carboxylate (14). 2-Flu-
oro-1-methylpyridinium tosylate (0.360 g, 1.26 mmol) was added
to a stirring solution of 2b (500 mg, 1.05 mmol) in CH2Cl2 that
was cooled in an ice-bath. Triethylamine (0.586 g, 4.2 mmol) was
then added dropwise and the resulting mixture was stirred at that
temperature for 70 min. A solution of 13 (0.40 g, 1.05 mmol) in
CH2Cl2 (1 mL) was then added and the resulting solution was stir-
red while warming to room temperature over 18 h. After diluting
with EtOAc, the organic layer was washed with 10% aqueous HCl,
brine, 5% aqueous bicarbonate, brine then dried over Na2SO4. The
crude product was purified by silica gel chromatography (0–30%
MeOH in EtOAc) resulting in 14 as a colourless solid (0.318 g,
36%). 1H NMR (400 MHz, CDCl3) d 0.96 (t, J = 3.9, 2H), 1.14–1.20
(m, 2H), 1.17 (t, J = 6.8, 6H), 1.29 (t, J = 6.8, 6H), 1.34 (d, J = 6.7,
3H), 1.50 (s, 9H), 3.20–3.25 (m, 3H), 3.40–3.47 (m, 2H), 3.74 (s,
3H), 3.91–4.00 (m, 3H), 4.02–4.16 (m, 8H), 4.35 (br s, 1H), 7.20
(d, J = 8.5, 2H), 7.40 (d, J = 8.5, 2H), 7.93 (d, J = 12.4, 1H), 8.71 (s,
1H). 19F (376 MHz, D2O) d ꢁ121.16 (d, J = 12.4, 1F). 31P (162 MHz,
CDCl3) d 19.35 (s, 2P).
6.2.6.1.3. Dimethyl 2-{4-[(bromoacetyl)amino]phenyl}-1-(dimeth-
oxyphosphoryl)ethylphosphonate (21). A solution of 22 (0.959 g,
2.87 mmol) and pyridine (349
was cooled in an ice-bath while stirring. A solution of bro-
moacetylbromide (250 L, 2.87 mmol) in CH2Cl2 (5 mL) was added
lL, 4.31 mmol) in 20 mL of CH2Cl2
l
drop-wise and the resulting mixture was stirred for 4 h at that
temperature. The reaction was quenched by the addition of water
and the product was extracted with CH2Cl2. The combined organic
layers were washed with brine, dried over sodium sulfate, filtered,
and concentrated under reduced pressure. The crude yellow solid
was purified by silica gel chromatography resulting in 21 as a col-
orless solid (0.897 g, 67%). 1H NMR (400 MHz, CDCl3) d 2.65 (tt,
J = 6.2, 24.4, 1H), 3.22 (dt, J = 6.2, 17.4, 2H), 3.72 (d, J = 3.7, 6H),
3.75 (d, J = 3.7, 6H), 4.01 (s, 2H), 7.26 (d, J = 8.6, 2H), 7.47 (d,
J = 8.6, 2H), 8.15 (br s, 1H). 31P (162 MHz, CDCl3) d 26.33 (s, 2P).
6.2.6.1.4. Tetraethyl
N-benzyl-N-methyl-1-aminomethylenebis-
phosphonate (24). Triethyl orthoformate (13.8 g, 93.3 mmol),
diethyl phosphite (32.2 g, 233 mmol) and N-benzylmethyl amine
(9.42 g, 77.7 mmol) were heated in a 100 mL round bottom flask
fitted with a distillation apparatus. The reaction was heated to a
temperature of 180–190 °C for 3 h under Ar at which time the evo-
lution of EtOH was complete. The reaction mixture was cooled to
room temperature, diluted with CHCl3 (400 mL), washed with
aqueous NaOH (1 M) and brine then dried over Na2SO4. The solvent
was removed at aspirator pressure resulting in the colourless oil 24
(31.7 g, 100%). 1H NMR (400 MHz, CDCl3) d 1.34 (t, J = 7.1, 6H), 1.35
(t, J = 7.1, 6H), 2.66 (s, 3H), 3.48 (t, J = 24.9, 1H), 3.99 (s, 2H), 4.07–
4.24 (m, 8H), 7.24–7.39 (m, 5H).
6.2.6.1.5. Tetraethyl N-methyl-1-aminomethylenebisphosphonate
(25). Compound 24 (12.4 g, 30.4 mmol) was dissolved in EtOH
(150 mL) followed by the addition of palladium on carbon (10%,
5 g) and cyclohexene (9.0 mL, 88.7 mmol). The resulting mixture
was heated to reflux under argon for 16 h. The cooled solution
was filtered through glassfiber filter paper and concentrated at re-
duced pressure to give 25 as a pale yellow oil (8.7 g, 90%), which
was used directly in the next step without further purification.
1H NMR (400 MHz, CDCl3) d 1.36 (t, J = 7.4, 12H), 2.69 (s, 3H),
3.41 (m, 1H), 4.20–4.31 (m, 8H). 31P (162 MHz, CDCl3) d 19.44 (s,
2P).
6.2.5.2. 4-(Bisphosphonomethyl)phenyl 7-(3-methylpiperazin-
1-yl)-1-cyclopropyl-6-fluoro-1,4-dihydro-8-methoxy-4-oxoquin-
oline-3-carboxylate (15). TMSBr (370 lL, 2.84 mmol) was added in
one portion to a stirring solution of 14 (155 mg, 0.185 mmol) in
CH2Cl2. After 18 h the solvent was removed at reduced pressure
and the yellow solid was re-suspended in H2O and the pH was ad-
justed to 7.4 by the addition of NaOH. The crude product was puri-
fied by Waters C18 Sep-PakTM column (40 mL H2O) to give 15 as a
colourless solid (80 mg, 61%). 1H NMR (400 MHz, D2O) d 1.24 (d,
J = 6.2, 3H), 1.21–1.30 (m, 2H), 1.36–1.45 (m, 2H), 2.51–2.59 (m,
1H), 2.86 (br s, 1H), 3.04–3.14 (m, 2H), 3.20–3.26 (m, 1H), 3.42 (t,
J = 23.2, 1H), 3.46–3.56 (m, 2H), 3.65 (s, 3H), 4.02 (br s, 1H), 7.27
(d, J = 7.7, 2H), 7.37 (d, J = 12.1, 1H), 7.67 (d, J = 7.0, 2H), 8.96 (s,
1H). 19F (376 MHz, D2O) d ꢁ121.76 (d, J = 12.1, 1F). 31P (162 MHz,
D2O) d 16.74 (s, 1P). ESI-MS calculated for (C26H30FN3O10P2+H+):
626; found: 626.1 (M+H+). Purity (method B): 100% (254 nm),
100% (220 nm), 100% (290 nm).
6.2.6. Bisphosphonated fluoroquinolone glycolamide esters
6.2.6.1. Bisphosphonated bromoacetamides
6.2.6.1.6. Tetraethyl N-(bromoacetyl)-N-methyl-1-aminomethyl-
6.2.6.1.1. Dimethyl 1-(dimethoxyphosphoryl)-2-(4-nitrophenyl)eth-
ylphosphonate (19). Sodium hydride (1.02 g, 25.4 mmol) was
added in portions to a stirring solution of tetramethyl meth-
enebisphosphonate (26).
(1.48 mL, 17.0 mmol) in CH2Cl2 (1 mL) was added dropwise to a
stirred, cooled (ice-bath) solution of 25 (4.5 g, 14 mmol) and pyri-
A solution of bromoacetyl bromide

9226
K. S. E. Tanaka et al. / Bioorg. Med. Chem. 16 (2008) 9217–9229
dine (1.78 mL, 21.3 mmol) in CH2Cl2 (25 mL). The reaction was stir-
red for 18 h while slowly warming to room temperature. After
quenching with water the product was extracted with CH2Cl2
and the combined organics were washed with 10% aqueous HCl,
brine, dried over sodium sulfate and concentrated at reduced pres-
sure. The crude yellow oil was purified by silica gel chromatogra-
phy on an automated flash chromatography system (0–10%
MeOH in EtOAc) resulting in 26 as a pale yellow liquid (2.93 g,
47%). 1H NMR (400 MHz, CDCl3) d 1.32 (t, J = 7.1, 12H), 3.38 (s,
3H), 3.92 (s, 2H), 4.15–4.25 (m, 8H), 5.69 (t, J = 24.5, 1H). 31P
(162 MHz, CDCl3) d 16.93 (s, 2P).
1H), 0.99–1.09 (m, 2H), 1.22–1.29 (m, 1H), 1.31–1.37 (m, 12H),
1.46–1.51 (m, 1H), 1.48 (s, 9H), 1.75–1.83 (m, 3H), 2.22–2.27 (m,
1H), 2.86–2.91 (m, 1H), 3.19–3.23 (m, 1H), 3.34–3.38 (m, 1H),
3.56 (s, 3H), 3.83 (dt, J = 1.9, 10.0, 1H), 3.87–3.92 (m, 1H), 4.02–
4.09 (m, 2H), 4.20–4.38 (m, 8H), 4.73–4.82 (br s, 1H), 4.83 (d,
J = 15.8, 1H), 4.91 (d, J = 15.8, 1H), 5.22 (dt, J = 10.0, 23.0, 1H),
7.75 (d, J = 14.1, 1H), 8.49 (s, 1H), 9.24 (d, J = 9.2, 1H).
6.2.6.2.2. (1,1-Bis(diethylphosphono)methylcarbamoyl)methyl 7-
(4-(tert-butoxycarbonyl)-3-methylpiperazin-1-yl)-1-cyclopropyl-6-flu-
oro-1,4-dihydro-8-methoxy-4-oxoquinoline-3-carboxylate (17b). From
2b and 16. Chromatography eluent: 0–5% MeOH in CH2Cl2. Yield:
62%. 1H NMR (400 MHz, CDCl3) d 0.90–0.95 (m, 2H), 1.13–1.19
(m, 2H), 1.32–1.36 (m, 15H), 1.50 (s, 9H), 3.19–3.46 (m, 5H), 3.72
(s, 3H), 3.89–3.97 (m, 2H), 4.21–4.37 (m, 9H), 4.87 (s, 2H), 5.21
(dt, J = 9.9, 22.9, 1H), 7.81 (d, J = 12.4, 1H), 8.53 (s, 1H), 9.10 (d,
J = 9.9, 1H). MS: (MH+) 819.3.
6.2.6.1.7. Tetraethyl
1-(N-(N-a,e-di-(tert-butoxycarbonyl)lysi-
noyl)amino)methylenebisphosphonate (34). To a solution of Boc-
Lys(Boc)-OH dicyclohexylamine salt (1.57 g, 2.97 mmol) in CH2Cl2
(12 mL) was added amine 29 (900 mg, 2.97 mmol), N-(3-dimethyl-
aminopropyl)-N0-ethylcarbodiimide
hydrochloride
(626 mg,
3.26 mmol) and DMAP (36 mg, 0.30 mmol). The mixture was stirred
for 18 h, after which the precipitate was removed by filtration and
washed with a portion of CH2Cl2. The combined filtrates were
washed with 1 M HCl, saturated NaHCO3 solution and brine. The or-
ganic layer was dried over Na2SO4, filtered and concentrated to dry-
ness. The residue was purified by flash chromatography on silica gel
using a gradient of 0–15% MeOH/EtOAc. Amide 34 was obtained as a
white foam (1.35 g, 72%). 1H NMR (400 MHz, CDCl3) d 1.30–1.34 (m,
12H), 1.43 (s, 18H), 1.38–1.53 (m, 4H), 1.59–1.69 (m, 1H), 1.79–1.87
(m, 1H), (3.07–3.12 (m, 2H), 4.11–4.24 (m, 8H), 4.71 (br s, 1H), 4.97
(dt, J = 21.4, 10.0, 1H), 5.11 (br s, 1H), 6.76 (d, J = 10.0, 1H).
6.2.6.2.3. (1,1-Bis(diethylphosphono)methylcarbamoyl)methyl 7-
(4-(tert-butoxycarbonyl)piperazin-1-yl)-1-cyclopropyl-6-fluoro-1,4-dihy-
dro-4-oxoquinoline-3-carboxylate (19c). From 2c and 16. Chroma-
tography eluent: 0–7% MeOH in CH2Cl2. Yield: 90%. 1H NMR
(400 MHz, CDCl3) d 1.12–1.17 (m, 2H), 1.31–1.36 (m, 14H), 1.50
(s, 9H), 3.26 (bt, J = 4.9, 4H), 3.41–3.47 (m, 1H), 3.66 (bt, J = 4.9,
4H), 4.20–4.38 (m, 8H), 4.88 (s, 2H), 5.22 (dt, J = 10.3, 22.6, 1H),
7.31 (d, J = 7.2, 1H), 7.99 (d, J = 13.2, 1H), 8.49 (s, 1H), 9.21 (d,
J = 9.8, 1H). 19F (376 MHz, CDCl3) d ꢁ123.57 (dd, J = 7.2, 13.2, 1F).
31P (162 MHz, CDCl3) d 17.35 (s, 2P).
6.2.6.2.4. (4-(2,2-Bis(dimethylphosphono)ethyl)phenylcarbamoyl)-
methyl 7-((4aS,7aS)-1-(tert-butoxycarbonyl)-octahydropyrrolo[3,4-
b]pyridin-6-yl)-1-cyclopropyl-6-fluoro-1,4-dihydro-8-methoxy-4-oxo-
quinoline-3-carboxylate (24a). From 2a and 21. Chromatography
eluent: 0–8% MeOH in CH2Cl2. Yield: 65%. 1H NMR (400 MHz,
CDCl3) d 0.73–0.84 (m, 1H), 0.97–1.11 (m, 2H), 1.20–1.28 (m, 2H),
1.41–1.52 (m, 1H), 1.46 (s, 9H), 1.70–1.85 (m, 2H), 2.19–2.29 (m,
1H), 2.67 (tt, J = 6.3, 23.7, 1H), 2.82–2.93 (m, 1H), 3.20 (dt, J = 6.2,
16.0, 3H), 3.36 (br s, 1H), 3.55 (s, 3H), 3.69 (d, J = 3.1, 6H), 3.72
(d, J = 3.1, 6H), 3.74–3.95 (m, 2H), 4.01–4.13 (m, 2H), 4.77 (br s,
1H), 4.89 (AB q, J = 14.9, 2H), 7.24 (d, J = 8.4, 2H), 7.88 (d, J = 8.4,
2H), 7.89 (d, J = 14.1, 1H), 8.49 (s, 1H), 11.0 (s, 1H).
6.2.6.1.8. Tetraethyl
1-(N-lysinoylamino)methylenebisphospho-
nate, trifluoroacetate salt (35). To carbamate 34 (1.35 g,
2.14 mmol) was added a solution of TFA/CH2Cl2 (11 mL, 40% v/v).
After stirring for 18 h, the reaction mixture was concentrated to
dryness and co-evaporated several times with Et2O. The resulting
deprotected material, a yellowish oil (2.1 g, >quant), was used
without further purification. 1H NMR (400 MHz, DMSO-d6) d
1.22–1.28 (m, 12H), 1.34–1.40 (m, 2H), 1.48–1.54 (m, 2H), 1.68–
1.74 (m, 2H), 2.71–2.76 (m, 2H), 3.93–3.97 (m, 1H), 4.03–4.15
(m, 8H), 4.82 (dt, J = 22.4, 9.8, 1H), 7.77 (br s, 3H), 8.25 (bd,
J = 4.2, 3H), 9.27 (d, J = 9.8, 1H).
6.2.6.1.9. Tetraethyl 1-(N-(N-
methylenebisphosphonate (36). To TFA salt 35 (max 2.14 mmol) in
CH2Cl2 (27 mL) at 0 °C was added pyridine (1.73 mL, 21.4 mmol)
and bromoacetyl bromide (390 lL, 4.49 mmol). The mixture was
stirred for 1.5 h at 0 °C after which it was diluted with CH2Cl2 and
washed with 5% HCl, saturated NaHCO3 solution and brine. The or-
ganic layer was dried over MgSO4, filtered and concentrated to dry-
ness. Purification by flash chromatography on silica gel, using a
gradient of 0–20% MeOH/EtOAc provided 36 as a white foam
(574 mg, 40%). 1H NMR (400 MHz, CDCl3) d 1.30–1.37 (m, 12 H),
1.38–1.46 (m, 2H), 1.53–1.61 (m, 2H), 1.72–1.81 (m, 1H), 1.86–
1.93 (m, 1H), 3.25–3.34 (m, 2H), 3.87 (s, 2H), 3.88 (s, 2H), 4.13–
4.25 (m, 8H), 4.53 (q, J = 6.2, 1H), 4.97 (dt, J = 21.7, 9.8, 1H), 6.96
(br s, 1H), 7.01 (bd, J = 9.8, 1H), 7.17 (d, J = 7.8, 1H).
a
,
e
-di-(bromoacetyl)lysinoyl)amino)-
6.2.6.2.5. (4-(2,2-Bis(dimethylphosphono)ethyl)phenylcarbamoyl)-
methyl 7-(4-(tert-butoxycarbonyl)-3-methylpiperazin-1-yl)-1-cyclo-
propyl-6-fluoro-1,4-dihydro-8-methoxy-4-oxoquinoline-3-carboxylate
(24b). From 2b and 21. Chromatography eluent: 0–8% MeOH in
CH2Cl2. Yield: 70%. 1H NMR (400 MHz, CDCl3) d 0.92–1.00 (m,
2H), 1.16–1.25 (m, 2H), 1.35 (d, J = 6.8, 3H), 1.50 (s, 9H), 2.69 (tt,
J = 6.4, 24.3, 1H), 3.17–3.50 (m, 7H), 3.71 (d, J = 2.5, 6H), 3.74 (s,
3H), 3.75 (d, J = 2.5, 6H), 3.93–3.99 (m, 2H), 4.36 (br s, 1H), 4.92
(s, 2H), 7.26 (d, J = 8.8, 2H), 7.89 (d, J = 8.8, 2H), 7.98 (d, J = 12.1,
1H), 8.57 (s, 1H), 10.90 (s, 1H). 19F (376 MHz, CDCl3) d ꢁ120.65
(d, J = 12.1, 1F). 31P (162 MHz, CDCl3) d 26.5 (s, 2P).
6.2.6.2.6. (N-Methyl-1,1-bis(diethylphosphono)methylcarbamoyl)-
methyl 7-((4aS,7aS)-1-(tert-butoxycarbonyl) -octahydropyrrolo[3,4-
b]pyridin-6-yl)-1-cyclopropyl-6-fluoro-1,4-dihydro-8-methoxy-4-oxo-
quinoline-3-carboxylate (29a). From 2a and 26. Chromatography
eluent: 0% to 10% MeOH in CH2Cl2. Yield: 25%. 1H NMR
(400 MHz, CDCl3) d 0.74–0.83 (m, 1H), 0.96–1.15 (m, 2H), 1.18–
1.25 (m, 1H), 1.31–1.39 (m, 14H), 1.48 (s, 9H), 1.75–1.81 (m, 2H),
2.20–2.27 (m, 1H), 2.88 (bt, J = 8.7, 1H), 3.21 (br s, 1H), 3.31 (s,
3H), 3.36 (br s, 1H), 3.54 (s, 3H), 3.80–3.91 (m, 2H), 4.01–4.08 (m,
2H), 4.13–4.26 (m, 8H), 4.77 (br s, 1H), 4.99 (AB q, J = 15.8, 2H),
5.70 (t, J = 24.8, 1H), 7.81 (d, J = 7.8, 1H), 8.61 (s, 1H). 31P
(162 MHz, CDCl3) d 17.10 (s, 2P).
6.2.6.2. Preparation of glycolamide esters. General procedure for
coupling of bromoacetamides with N-Boc fluoroquinolones. A solution
of N-Boc fluoroquinolone (2.1 mmol), bromoacetamide (2.1 mmol)
and Cs2CO3 (0.76 g, 2.3 mmol) in 10 mL of DMF was stirred at room
temperature for 12 h. The mixture was then diluted with H2O and ex-
tracted with CH2Cl2. The organics were washed with brine, dried over
sodium sulfate, filtered and concentrated at reduced pressure result-
ing in a brown oil that was purified by silica gel chromatography.
6.2.6.2.1. (1,1-Bis(diethylphosphono)methylcarbamoyl)methyl 7-
((4aS,7aS)-1-(tert-butoxycarbonyl)-octahydropyrrolo[3,4-b]pyridin-6-yl)-
1-cyclopropyl-6-fluoro-1,4-dihydro-8-methoxy-4-oxoquinoline-3-carbox-
ylate (17a). From 2a and 16. Chromatography eluent: 0–8% MeOH
in CH2Cl2. Yield: 72%. 1H NMR (400 MHz, CDCl3) d 0.75–0.80 (m,
6.2.6.2.7.
(N-Methyl-1,1-bis(diethylphosphono)methylcarba-
moyl)methyl 7-(4-(tert-butoxycarbonyl)-3-methylpiperazin-1-yl)-1-
cyclopropyl-6-fluoro-1,4-dihydro-8-methoxy-4-oxoquinoline-3-car-
boxylate (29b). From 2b and 26. Chromatography eluent: 0–10%
MeOH in CH2Cl2. Yield: 30%. 1H NMR (400 MHz, CDCl3) d 0.87 (bt,

K. S. E. Tanaka et al. / Bioorg. Med. Chem. 16 (2008) 9217–9229
9227
J = 3.8, 2H), 1.13 (d, J = 7.5, 2H), 1.30–1.39 (m, 15H), 1.50 (s, 9H),
3.19–3.27 (m, 3H), 3.31 (s, 3H), 3.42 (bt, J = 13.4, 2H), 3.70 (s,
3H), 3.84–3.90 (m, 1H),3.94 (d, J = 12.2, 1H), 4.11–4.26 (m, 8H),
4.33 (br s, 1H), 4.97–5.01 (m, 2H), 5.70 (t, J = 24.8, 1H), 7.88 (d,
J = 12.8, 1H), 8.63 (s, 1H). 19F (376 MHz, CDCl3) d ꢁ121.61 (d,
J = 12.8, 1F). 31P (162 MHz, CDCl3) d 17.11 (s, 2P).
2.62–2.72 (m, 1H), 2.90–3.17 (m, 3H), 3.31–3.40 (m, 1H), 3.44–
3.63 (m, 5H), 3.65–3.72 (m, 1H), 3.75–3.99 (m, 3H), 4.80 (s, 2H),
7.25–7.39 (m, 5H), 8.57 (s, 1H). 19F (376 MHz, D2O) d ꢁ121.42 (d,
J = 14.0, 1F). 31P (162 MHz, D2O) d 20.25 (d, J = 22.4, 2P). ESI-MS cal-
culated for (C31H37FN4O11P2+H+): 723; found: 723.2 (M+H+). Purity
(method A): 95.7% (254 nm), 95.4% (220 nm), 96% (290 nm).
6.2.6.2.8. Bis(gatifloxacin ester) conjugate 39. From 2b and 36
(½ mole equivalent). The crude product was purified by reverse
phase flash chromatography on a C18 column, using a gradient of
20–100% MeCN/H2O, followed by flash chromatography on silica
gel using a gradient of 0–10% MeOH/CH2Cl2. Yield: 47%. 1H NMR
(400 MHz, CDCl3) d 0.93–0.99 (m, 4H), 1.14–1.19 (m, 4H), 1.25–
1.34 (m, 18H), 1.49 (s, 9H), 1.50 (s, 9H), 1.55–1.61 (m, 1H), 1.65–
1.77 (m, 3H), 1.99–2.06 (m, 2H), 3.20–3.35 (m, 9H), 3.39–3.47
(m, 4H), 3.74 (s, 6H), 3.91–3.98 (m, 4H), 4.11–4.18 (m, 8H), 4.34
(br s, 2H), 4.49–4.68 (m, 5H), 5.03 (dt, J = 21.9, 10.2, 1H), 7.12 (d,
J = 10.2, 1H), 7.86 (d, J = 12.3, 2H), 8.47–8.50 (m, 2H), 8.72 (t,
J = 5.6, 1H), 9.51 (d, J = 8.2, 1H). ESI-MS calculated for
C67H95F2N9O21P2 1461, found 1460 (M-H)ꢁ.
General procedure for deprotection of bisphosphonates. TMSBr
(3.0 mL, 23 mmol) was added in one portion to a stirring solution
of protected bisphosphonated fluoroquinolone (1.50 mmol) in
CH2Cl2. After 18 h the solvent was removed at reduced pressure.
The yellow solid resuspended in H2O and the pH adjusted to 7.4 by
the addition of NaOH. The resulting solution was the subjected to
chromatography on a Waters C18 Sep-PakTM to furnish the product.
6.2.6.2.9. (1,1-Bisphosphonomethylcarbamoyl)methyl 1-cyclopro-
pyl-6-fluoro-1,4-dihydro-7-((4aS,7aS)-octahydropyrrolo[3,4-b]pyridin-
6-yl)-8-methoxy-4-oxoquinoline-3-carboxylate (18a). Compound
18a eluted in H2O then 5% MeOH/H2O. Yield: 69%. 1H NMR
(400 MHz, D2O) d 0.70–0.78 (m, 1H), 0.86–1.00 (m, 2H), 1.03–
1.10 (m, 1H), 1.67–1.78 (m, 4H), 2.61 (br s, 1H), 2.90–2.95 (m,
1H), 3.23–3.26 (m, 1H), 3.40 (s, 3H), 3.51–3.65 (m, 3H), 3.75 (br
s, 1H), 3.83–3.87 (m, 1H), 3.92–3.97 (m, 1H), 4.19 (t, J = 18.0, 1H),
4.74 (s, 2H), 7.24 (d, J = 15.0, 1H), 8.77 (s, 1H). 19F (376 MHz,
D2O) d ꢁ121.76 (d, J = 15.0, 1F). 31P (162 MHz, D2O) d 13.90 (s,
2P). ESI-MS calculated for (C24H31FN4O11P2+H+): 633; found:
633.0 (M+H+). Purity (method B): 99.4% (254 nm), 99.8%
(220 nm), 99.7% (290 nm).
6.2.6.2.13. (4-(2,2-Bisphosphonoethyl)phenylcarbamoyl)methyl 7-
(3-methylpiperazin-1-yl)-1-cyclopropyl-6-fluoro-1,4-dihydro-8-meth-
oxy-4-oxoquinoline-3-carboxylate (23b). Compound 23a eluted
with 0% to 10% MeOH in H2O. Yield: 43%. 1H NMR (400 MHz,
D2O) d 0.98–1.01 (m, 2H), 1.15–1.18 (m, 2H), 1.22 (d, J = 6.3, 3H),
2.20 (tt, J = 6.7, 21.0, 1H), 3.07–3.49 (m, 9H), 3.80 (s, 3H), 4.05–
4.11 (m, 1H), 4.92 (s, 2H), 7.46 (AB q, J = 8.4, 4H), 7.55 (d, J = 12.3,
1H), 8.80 (s, 1H). 19F (376 MHz, D2O) d ꢁ121.16 (d, J = 12.3, 1F).
31P (162 MHz, D2O)
d 20.23 (s, 2P). ESI-MS calculated for
(C29H35FN4O11P2+H+): 697; found: 697.2 (M+H+). Purity (method
B): 89.3% (254 nm), 91.4% (220 nm), 94.0% (290 nm).
6.2.6.2.14. (N-Methyl-1,1-bisphosphonomethylcarbamoyl)methyl
1-cyclopropyl-6-fluoro-1,4-dihydro-7-((4aS,7aS)-octahydropyrrolo[3,
4-b]pyridin-6-yl)-8-methoxy-4-oxoquinoline-3-carboxylate(28a). Com-
pound 28a eluted with 0% to 10% MeOH in H2O. Yield: 27%. 1H NMR
(400 MHz, D2O) d 0.89–0.96 (m, 1H), 1.06–1.17 (m, 2H), 1.20–1.26
(m, 1H), 1.83–1.93 (m, 4H), 2.78 (br s, 1H), 3.10 (br s, 1H), 3.16 (s,
1/3-3H), 3.27 (s, 2/3-3H), 3.39 (br s, 1H), 3.55 (s, 1/3-3H), 3.57 (s,
2/3-3H), 3.67–3.83 (m, 3H), 3.88–4.14 (m, 3H), 4.92 (t, J = 21.9,
1H), 5.12 (AB q, J = 15.7, 2/3-2H), 5.16 (AB q, J = 15.7, 1/3-2H), 7.37
(d, J = 14.0, 2/3-1H), 7.44 (d, J = 14.0, 1/3-1H), 8.96 (s, 1H). 19F
(376 MHz, D2O) d ꢁ121.92 (d, J = 14.0, 2/3-1F), –121.84 (d, J = 14.0,
1/3-1F). 31P (162 MHz, D2O) d 12.31 (s, 1/3-2P), 13.08 (s, 2/3-2P).
ESI-MS calculated for (C25H33FN4O11P2+H+): 647; found: 647.1
(M+H+). Purity (method B): 98.4% (254 nm), 99.2% (220 nm), 98.9%
(320 nm).
6.2.6.2.15. (N-Methyl-1,1-bisphosphonomethylcarbamoyl)methyl
7-(3-methylpiperazin-1-yl)-1-cyclopropyl-6-fluoro-1,4-dihydro-8-
methoxy-4-oxoquinoline-3-carboxylate (28b). Compound 28b
eluted with 0–5% MeOH in H2O. Yield: 54%. 1H NMR (400 MHz,
D2O) d 1.06–1.11 (m, 2H), 1.19–1.23 (m, 2H), 1.33–1.36 (m,
3H), 3.16 (s, 1/3-3H), 3.27 (s, 2/3-3H), 3.21–3.30 (m, 1H),
3.32–3.36 (m, 1H), 3.40–3.48 (m, 1H), 3.53–3.64 (m, 3H), 3.74
(s, 2/3-3H), 3.79 (s, 1/3-3H), 3.89 (t, J = 21.0, 1/3-1H), 4.13–
4.17 (m, 1H), 4.92 (t, J = 21.0, 2/3-1H), 5.13 (s, 2/3-2H), 5.19 (s,
1/3-2H), 7.45 (bd, J = 12.0, 2/3-1H), 7.59 (bd, J = 12.0, 1/3-1H),
9.02 (s, 2/3-1H), 9.03 (s, 1/3-1H). 19F (376 MHz, D2O) d ꢁ121.83
(d, J = 12.0, 1/3-1F), -122.02 (d, J = 12.0, 2/3-1F). 31P (162 MHz,
D2O) d 12.32 (s, 1/3-2P), 13.11 (s, 2/3-2P). ESI-MS calculated for
(C23H31FN4O11P2+H+): 621; found: 621.1 (M+H+). Purity (method
B): 98.0% (254 nm), 97.3% (220 nm), 97.4% (290 nm).
6.2.6.2.10. (1,1-Bisphosphonomethylcarbamoyl)methyl 7-(3-meth-
ylpiperazin-1-yl)-1-cyclopropyl-6-fluoro-1,4-dihydro-8-methoxy-4-oxo-
quinoline-3-carboxylate (18b). Compound 18b eluted in H2O then
5% MeOH/H2O. Yield: 75%. 1H NMR (400 MHz, D2O) d 1.02–1.06 (m,
2H), 1.18–1.23 (m, 2H), 1.38 (d, J = 6.7, 3H), 3.31–3.43 (m, 2H),
3.49–3.69 (m, 5H), 3.81 (s, 3H), 4.14–4.20 (m, 1H), 4.49 (t,
J = 20.1, 1H), 4.94 (s, 2H), 7.68 (d, J = 12.2, 1H), 9.00 (s, 1H). 19F
(376 MHz, D2O) d ꢁ119.20 (d, J = 12.2, 1F). 31P (162 MHz, D2O) d
16.41 (s, 2P). ESI-MS calculated for (C22H29FN4O11P2+H+): 606;
found: 606.0 (M+H+). Purity (method B): 99.5% (254 nm), 99.5%
(220 nm), 99.7% (290 nm).
6.2.6.2.16. Bis(gatifloxacin ester) conjugate 38. To a solution of
protected conjugate 39 (391 mg, 0.27 mmol) in CH2Cl2 (5 mL) was
added 2,6-lutidine (1.55 mL, 13.4 mmol). The mixture was cooled
6.2.6.2.11. (1,1-Bisphosphonomethylcarbamoyl)methyl 1-cyclo-
propyl-6-fluoro-1,4-dihydro-4-oxo-7-(piperazin-1-yl)quinoline-3-car-
boxylic acid (18c). Compound 18c eluted with H2O then 5%
MeOH/H2O. Yield: 66%. 1H NMR (400 MHz, D2O) d 1.18–1.22 (m,
2H), 1.36–1.41 (m, 2H), 3.15 (m, 4H), 3.25 (m, 4H), 3.52 (br s,
1H), 4.22 (t, J = 18.7, 1H), 4.79 (s, 2H), 7.36 (d, J = 7.0, 1H), 7.60
(d, J = 12.2, 1H), 8.80 (s, 1H). 19F (376 MHz, D2O) d ꢁ123.72 (dd,
J = 7.0, 12.2, 1F). 31P (162 MHz, D2O) d 13.91 (s, 2P). ESI-MS calcu-
lated for (C20H25FN4O10P2+H+): 563; found: 563.1 (M+H+). Purity
(method A): 97.9% (254 nm), 97.4% (220 nm), 98.2% (290 nm).
6.2.6.2.12. (4-(2,2-Bisphosphonoethyl)phenylcarbamoyl)methyl 1-
cyclopropyl-6-fluoro-1,4-dihydro-7-((4aS,7aS)-octahydropyrrolo[3,4-b]-
pyridin-6-yl)-8-methoxy-4-oxoquinoline-3-carboxylate (23a). Com-
pound 23a eluted with 0% to 10% MeOH in H2O. Yield: 9%. 1H
NMR (400 MHz, D2O) d 0.60–0.69 (m, 1H), 0.93–1.07 (m, 2H),
1.12–1.21 (m, 1H), 1.72–1.96 (m, 4H), 2.16 (tt, J = 6.9, 20.6, 1H),
to -78 °C and trimethylsilylbromide (882 lL, 6.68 mmol) was added
slowly. The mixture was brought to room temperature and stirred
for 18 h, then was concentrated to dryness. Crude product was puri-
fied by 2 consecutive reverse phase flash chromatographies on a C18
column, using a gradient of 5–60% MeCN/50 mM Et3NH2CO3 buffer,
pH 7 for the first column, then a gradient of 5–50% MeCN/50 mM
Et3NH2CO3 buffer, pH 7 for the second column. Lyophilization of
the combined pure fractions provided conjugate 40 as a white solid
(16 mg, 5%). 1H NMR (400 MHz, DMSO-d6+TFA) d 0.97 (br s, 4H),
1.07–1.10 (m, 4H), 1.17 (t, J = 7.4, 9H), 1.26 (d, J = 6.4, 6H), 1.34–
1.49 (m, 4H), 1.61–1.64 (m, 1H), 1.72–1.76 (m, 1H), 3.10 (q, J = 7.4,
6H), 3.17–3.26 (m, 4H), 3.38–3.52 (m, 10H), 3.81 (s, 6H), 4.01, 4.06
(m, 2H), 4.45–4.64 (m, 5H), 7.67 (d, J = 12.1, 1H), 7.68 (d, J = 12.1,
1H), 8.52 (s, 1H), 8.53 (s, 1H). ESI-MS calculated for
(C49H63F2N9O17P2ꢁH+): 1148; found: 1148.2 (M-H)-. Purity (meth-
od C): 98.1% (254 nm), 97.6% (220 nm), 99.1% (320 nm).

9228
K. S. E. Tanaka et al. / Bioorg. Med. Chem. 16 (2008) 9217–9229
6.2.7. Preparation of bisphosphonated moxifloxacin
thioglycolamide ester
Standards Institute (formerly the National Committee for Clinical
Laboratory Standards).18
6.2.7.1. Tetraethyl 1-(N-3-thiapropionylamino)methylenebis-
phosphonate (30). A mixture of amine 31 (691 mg, 2.28 mmol)
6.3.2. Determination of levels of bone binding in vitro
and mercaptoacetic acid (200
l
L, 2.89 mmol) was heated to 140–
Individual compounds were dissolved in PBS and added, at a fi-
nal concentration of 1 mg/mL, to a slurry of bone meal powder
(Now Foods, Bloomingdale, Illinois, USA) in PBS at 10 mg/mL. The
suspension of drug/prodrug in bone meal powder was incubated
at 37 °C for 1 h to allow for binding, and centrifuged at
13,000 rpm for 2 min, before recovering the supernatant. The bone
meal powder pellet was then washed three times with 1 ml of PBS.
All supernatants were saved and assessed for fluoroquinolone con-
tent by fluorescence measurements at excitation/emission wave-
lengths of 280/465 nm. The amount of fluoroquinolone was
determined from standard curves generated for each experiment.
The amount of drug/prodrug bound to bone powder was deduced
from the difference between the input amount (typically 1 mg)
and the amount recovered in the supernatants after binding. In
all binding experiments, >99% of input drug was recovered in the
supernatant for the parent drugs.
150 °C under a continuous argon purge. When steam evolution ap-
peared complete, the residue was purified by flash chromatogra-
phy on silica gel eluting with 5% methanol/CH2Cl2 to give 30
(0.321 g, 37%). 1H NMR (400 MHz, CDCl3) d 1.339 (t, J = 7.0, 6H),
1.344 (t, J = 7.0, 6H), 1.99 (t, J = 8.8, 1H), 3.25–3.35 (m, 2H), 4.11–
4.30 (m, 8H), 4.97 (td, J = 21.4, J = 10.1, 1H).
6.2.7.2. S-(1,1-Bis(diethylphosphono)methylcarbamoyl)methyl
7-((4aS,7aS)-1-(tert-butoxycarbonyl)
-octahydropyrrolo[3,4-
b]pyridin-6-yl)-1-cyclopropyl-6-fluoro-1,4-dihydro-8-methoxy-
4-oxoquinoline-3-carbothioate (31). To a solution of 2a (427 mg,
0.851 mmol) in CH2Cl2 (6.5 mL) was added 2-fluoro-1-methylpy-
ridinium tosylate (0.292 g, 1.03 mmol). The reaction mixture was
cooled to 0 °C, and triethylamine (0.43 mL, 3.09 mmol) was added
via syringe. After stirring 1 h at 0 °C a solution of thiol 30 (0.32 g,
0.85 mmol) in CH2Cl2 (10 mL) was added. After a further 10 min
at 0 °C the reaction was allowed to warm to room temperature
overnight. The reaction mixture was diluted with CH2Cl2 and
washed with ice cold saturated NH4Cl solution, ice cold 5% NaH-
CO3, water and brine. After drying (MgSO4) and concentration to
dryness, the residue was purified by flash chromatography on silica
gel eluting with 4% methanol/CH2Cl2 to give almost pure 31
(0.418 g, 57%) as a yellow foam, which was used without additional
purification. 1H NMR (400 MHz, CDCl3) d 0.75–0.85 (m, 1H), 0.99–
1.17 (m, 2H), 1.21–1.39 (m, 13H), 1.48 (s, 9H), 1.61 (s, 2H), 1.75–
1.83 (m, 2H), 2.21–2.30 (m, 1H), 2.82–2.94 (m, 1 H), 3.17–3.28
(m, 1H), 3.30–3.44 (m, 1H), 3.57 (s, 3H), 3.72 (s, 2H), 3.81–3.88
(m, 1H), 3.91–3.98 (m, 1H), 4.01–4.28 (m, 10H), 4.70–4.86 (br s,
1H), 4.98 (td, J = 21.6, 9.9, 1H), 7.10 (d, J = 10.3, 1H), 8.59 (s, 1H).
6.3.3. Regeneration of parent drug from bone-bound prodrug
Washed bone powder-bound prodrugs from the protocol de-
scribed in Section 6.3.2 were resuspended in 400
lL PBS or in
400 L 50% (v/v in PBS) human or rat serum. The suspension was
l
incubated for 24 h at 37 °C, centrifuged at 13,000 rpm for 2 min
and the supernatant was recovered. Methanol (5ꢀ volume relative
to supernatant) was added to each supernatant and the mixture
was vortexed on a floor model vortex for 15 min to extract freed
fluoroquinolone. The mixture was then centrifuged at 10,000 rpm
for 15 min to pellet the insoluble material. The supernatant con-
taining the extracted fluoroquinolone was recovered and evapo-
rated to dryness in
a speed vac. The dried pellets were
resuspended in PBS and the amount of fluoroquinolone was deter-
mined by fluorescence measurements as described previously. The
percentage of drug regenerated was deduced from the difference
between amount of bound prodrug and the amount of regenerated
drug. The identity of regenerated drug was deduced by determina-
tion of the minimum inhibitory concentration (as per the guide-
lines of the Clinical and Laboratory Standards Institute, guideline
M7–A7), which always matched those of parent drugs but not
those of prodrugs.
6.2.7.3. S-(1,1-Bisphosphonomethylcarbamoyl)methyl 1-cyclo-
propyl-6-fluoro-1,4-dihydro-7-((4aS,7aS)-octahydropyrrolo[3,4-
b]pyridin-6-yl)-8-methoxy-4-oxoquinoline-3-carbothioate (32).
To a solution of 31 (418 mg, 0.486 mmol) in CH2Cl2 (30 mL) was
added TMSBr (0.64 mL, 4.8 mmol). The reaction mixture was stirred
for 41 h, the solvent removed under reduced pressure and the solid
dried under high vacuum for 1 h. The solid was suspended in H2O
(100 mL) and the pH was immediately adjusted to pH 7 by the addi-
tion of 1 M NaOH, with concomitant dissolution of the product. The
product solution was washed with CHCl3 (2ꢀ 50 mL), filtered, con-
centrated to dryness and purified by reverse-phase chromatography
(gradient elution, 0–15% methanol in water) to furnish 32 as a
slightly yellowish solid (90 mg, 25% recovery based on tetrasodium
salt of product). 1H NMR (400 MHz, D2O) d 0.93–1.03 (m, 1H), 1.03–
1.19 (m, 2H), 1.19–1.29 (m, 1H), 1.78–2.01 (m, 4H), 2.79 (br s, 1H),
3.02–3.12 (m, 1H), 3.36–3.44 (m, 1H), 3.61 (s, 3H), 3.57–3.85 (m,
3H), 3.78 (br s, 2H), 3.89–3.95 (m, 1H), 4.02–4.16 (m, 2H), 4.27 (t,
J = 18.7, 1H), 7.40 (d, J = 13.7, 1H), 8.59 (s, 1H). 19F (376 MHz, D2O)
d ꢁ94.67 (d, J = 13.7, 1F). 31P (162 MHz, D2O) d 14.08 (d, J = 22.0,
6.3.4. Prophylactic use of prodrugs in a rat model of bone
infection
Staphylococcus aureus ATCC 13709 cells were grown overnight
at 37 °C in brain heart infusion broth (BHIB). Cells were subcul-
tured into fresh BHIB and incubated for 4–5 h at 37 °C. The cells
were washed twice with PBS and resuspended in BHIB supple-
mented with 10% (vol./vol.) fetal bovine serum at a density of
approximately 1010 colony forming units (CFU)/ml (based upon
turbidimetry). The suspension was aliquoted and a portion was
plated to check the CFU count. The culture was stored frozen
(ꢁ80ꢀC) and was used without subculture. For use as an inoculum
the culture was thawed, diluted in PBS and kept in an ice bath until
it was used.
Animals were injected with the drugs intravenously and, after
the indicated duration, infected as described17b to generate the
bone infection. Female CD rats (age 57–70 days; n = 5/group;
Charles River, St-Constant, Canada) were anaesthetized by isofluo-
rane before and during the surgery. Following complete induction
of anesthesia, the rat was placed ventral side up and hair was
shaved from the surgical site. The skin over the leg was cleaned
and disinfected (proviodine-ethanol 70%). A longitudinal incision
below the knee joint was made in the sagital plane. The incision
was made over the bone below the ‘’knee joint’’ (tibia head or con-
1P),
13.95
(d,
J = 22.0,
1P).
ESI-MS
calculated
for
(C24H31FN4O10P2S+H+): 649; found: 649 (M+H+). Purity (method
A): 99.1% (254 nm), 99.6% (220 nm), 99.0% (290 nm).
6.3. Biology
6.3.1. Determination of minimum inhibitory concentrations
against S. aureus ATCC 13709
The antimicrobial activities of the compounds were tested
against S. aureus strain ATCC 13709. Minimum inhibitory concen-
tration (MIC) testing was performed by the microdilution method
according to M7–A7 guidelines set by the Clinical and Laboratory

K. S. E. Tanaka et al. / Bioorg. Med. Chem. 16 (2008) 9217–9229
9229
1068; (e) Joostena, U.; Joist, A.; Goshegerc, G.; Liljenqvist, U.; Brandt, B.; von
Eiff, C. Biomaterials 2005, 26, 5251–5258.
6. (a) Hirabayashi, H.; Fujisaki, J. Clin. Pharmacokinet. 2003, 42, 1319–1330; (b)
Uludag, H. Curr. Pharm. Des. 2002, 8, 1929–1944; (c) Vepsäläinen, J. J. Curr. Med.
Chem. 2002, 9, 1201–1208.
7. (a) Hirabayashi, H.; Takahashi, T.; Fujisaki, J.; Masunaga, T.; Sato, S.; Hiroi, J.;
Tokunaga, Y.; Kimura, S.; Hata, T. J. Control. Release 2001, 70, 183–191; (b)
Hirabayashi, H.; Sawamoto, T.; Fujisaki, J.; Tokunaga, Y.; Kimura, S.; Hata, T.
Biopharm. Drug Dispos. 2002, 23, 307–315; (c) Gil, L.; Han, Y.; Opas, E. E.; Rodan,
G. A.; Ruel, R.; Seedor, J. G.; Tyler, P. C.; Young, R. N. Bioorg. Med. Chem. 1999, 7,
901–919; (d) Page, P. C. B.; McKenzie, M. J.; Gallagher, J. A. J. Org. Chem. 2001,
66, 3704–3708; (e) Fujisaki, J. J.; Tokunaga, Y.; Takahashi, T.; Hirose, T.;
Shimojo, F.; Kagayama, A.; Hata, T. J. Drug Target. 1995, 3, 273–282; (f) Fujisaki,
J.; Tokunaga, Y.; Sawamoto, T.; Takahashi, T.; Kimura, S.; Shimojo, F.; Hata, T. J.
Drug Target. 1996, 4, 117–123; (g) Erez, R.; Ebner, S.; Attali, B.; Shabat, D. Bioorg.
Med. Chem. Lett. 2008, 18, 816–820.
dyle) but not completely extending to the ankle. A high speed drill
fitted with a 2-mm bulb bit was used to drill a hole into the medul-
lar cavity of the tibia. Rats were injected intra-tibially with 0.05 ml
5% sodium morrhuate (sclerosing agent) and then with 0.05 ml of
S. aureus suspension (ca. 5 ꢀ 105 CFU/rat). The hole was sealed by
applying a small amount of dry dental cement which immediately
absorbs fluids and adheres to the site. The wound was closed using
three metal skin clips. Moxifloxacin 1a (as a positive control) was
injected once at 10 mg/kg intravenously 1 h postinfection in saline,
while the test compounds (prepared in 0.9% saline) were injected
as a single intravenous bolus dose at the indicated time points
prior to the infection.
Infected rats were sacrificed by CO2 asphyxiation 24 h postin-
fection to monitor the bacterial CFU count. Infected tibiae were re-
moved, dissected free of soft tissue, and weighed. The bones were
ground using a metal ball mill, resuspended in 5 ml 0.9% NaCl, seri-
ally diluted and processed for quantitative cultures. Treatment effi-
cacy was measured in terms of Log viable bacteria (Log CFU per
gram of bone). The results obtained for each group of rats were
evaluated by calculating the mean Log CFU and standard deviation.
The limit of detection is 2Log CFU/g of bone. Statistical compari-
sons of viable bacterial counts for the different treated and un-
treated groups were performed with Dunnett’s multiple-
comparison test or Kruskal–Wallis analysis. Differences were con-
sidered significant when the P value was <0.05 when comparing
treated infected animals to the untreated infected ones.
8. (a) Adzamli, I. K.; Gries, H.; Johnson, D.; Blau, M. J. Med. Chem. 1989, 32, 139–
144; (b) Ogawa, K.; Mukai, T.; Arano, Y.; Ono, M.; Hanaoka, H.; Ishino, S.;
Hashimoto, K.; Nishimura, H.; Saji, H. Bioconjugate Chem. 2005, 16, 751–757; (c)
Kubíek, V.; Rudovsk, J.; Kotek, J.; Hermann, P.; Vander Elst, L.; Muller, R. N.;
Kolar, Z. I.; Wolterbeek, H. Th.; Peters, J. A.; Luke, I. J. Am. Chem. Soc. 2005, 127,
16477–16485.
9. (a) Bansal, G.; Wright, J. E. I.; Kucharski, C.; Uludag˘, H. Angew. Chem. Int. Ed.
ˇ
ˇ
2005, 44, 3710–3714; (b) Wanga, D.; Miller, S. C.; Kopecková, P.; Kopecek, J.
Adv. Drug Delivery Rev. 2005, 57, 1049–1076; (c) Wright, J. E. I.; Gittens, S. A.;
Bansal, G.; Kitov, P. I.; Sindrey, D.; Kucharski, C.; Uludag˘, H. Biomaterials 2006,
27, 769–784.
10. Herczegh, P.; Buxton, T. B.; McPherson, J. C., III; Kovács-Kulyassa, Á.; Brewer, P.
D.; Sztaricskai, F.; Stroebel, G. G.; Plowman, K. M.; Farcasiu, D.; Hartmann, J. F. J.
Med. Chem. 2002, 45, 2338–2341.
11. Rogers, M. J.; Gordon, S.; Benford, H. L.; Coxon, F. P.; Luckman, S. P.;
Monkkonen, J.; Frith, J. C. Cancer 2000, 88, 2961–2978.
12. Henderson, B.; Nair, S. P. Trends Microbiol. 2003, 11, 570–577.
13. (a) Mitscher, L. A. Chem. Rev. 2005, 105, 559–592; (b) Van Bambeke, F.;
Michot, J.-M.; Van Eldere, J.; Tulkens, P. M. Clin. Microbiol. Infect. 2005, 11,
256–280.
14. (a) Greenberg, R. N.; Newman, M. T.; Shariaty, S.; Pectol, R. W. Antimicrob.
Agents Chemother. 2000, 44, 164–166; (b) Shirtliff, M. E.; Calhoun, J. H.; Mader,
J. T. J. Antimicrob. Chemother. 2001, 48, 253–258; (c) Shirtliff, M. E.; Calhoun, J.
H.; Mader, J. T. Antimicrob. Agents Chemother. 2002, 46, 231–233.
15. Vovk, A. I.; Kalchenko, V. I.; Cherenok, S. A.; Kukhar, V. P.; Muzychka, O. V.;
Lozynsky, M. O. Org. Biomol. Chem. 2004, 2, 3162–3166.
References and notes
1. Lew, D. P.; Waldvogel, F. A. Lancet 2004, 364, 369–379.
2. Mandal, S.; Berendt, A. R.; Peacock, S. J. J. Infect. 2002, 44, 143–151.
3. Peter, K.; Henke, P. K.; Blackburn, S. A.; Wainess, R. W.; Cowan, J.; Alicia
Terando, A.; Proctor, M.; Wakefield, T. W.; Upchurch, G. R., Jr.; Stanley, J. C.;
Greenfield, L. J. Ann. Surg. 2005, 241, 885–894.
4. (a) Tice, A. D.; Hoaglund, P. A.; Shoultz, D. A. Am. J. Med. 2003, 114, 723–728; (b)
Tice, A. D.; Hoaglund, P. A.; Shoultz, D. A. J. Antimicrob. Chemother. 2003, 51,
1261–1268.
16. Nielsen, N. M.; Bundgaard, H. J. Pharm. Sci. 1988, 77, 285–298.
17. (a) Mader, J. T. Am. J. Med. 1985, 78, 213–217; (b) O’Reilly, T.; Kunz, S.; Sande,
E.; Zak, O.; Sande, M. A.; Täuber, M. G. Antimicrob. Agents Chemother. 1992, 36,
2693–2697.
5. For examples of release of antibiotics from impregnated materials: (a) Nelson,
C. L.; McLaren, S. G.; Skinner, R. A.; Smeltzer, M. S.; Thomas, J. R.; Olsen, K. M. J.
Orthop. Res. 2002, 20, 643–647; (b) Baro, M.; Sánchez, E.; Delgado, A.; Perera, A.;
Évora, C. J. Control. Release 2002, 83, 353–364; (c) Castro, C.; Sánchez, E.;
Delgado, A.; Soriano, I.; Núñez, P.; Baro, M.; Perera, A.; Évora, C. J. Control.
Release 2003, 93, 341–354; (d) Mäkinen, T. J.; Veiranto, M.; Lankinen, P.; Moritz,
N.; Jalava, J.; Törmälä, P.; Aro, H. T. J. Antimicrob. Chemother. 2005, 56, 1063–
18. [CLSI] Clinical and Laboratory Standards Institute. 2006a. Methods for Dilution
Antimicrobial Susceptibility Tests for Bacteria That Grow Aerobically;
Approved Standard-Seventh Edition. CLSI document M7-A7 (ISBN 1-56238-
587-9). Clinical and Laboratory Standards Institute, 940 West Valley Road,
Suite 1400, Wayne, Pennsylvania 19087-1898 USA.
19. Gourves, J.-P.; Couthon, H.; Sturtz, G. Phosphorus Sulfur Silicon Relat. Elem. 1998,
132, 219–230.