R. J. Kerns et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2109–2112
2111
descarboxy ciprofloxacin monomers, respectively (Table
1). Compound 10, which contains pipemidic acid in
place of one ciprofloxacin in the dimer, is more potent
or equipotent than ciprofloxacin dimer 9 against all
strains of drug-resistant S. aureus tested. In contrast,
incomplete dimer 11, which contains descarboxy cipro-
floxacin in place of one ciprofloxacin,6 is less active than
9 against all strains. To our surprise, although the des-
carboxy ciprofloxacin dimer 11 is generally less active
than parent dimer 9, 11 does not have increased MICs
against SA 1199-3 or SA 1199B.
In summary, we have shown that potent activity of
piperazinyl-linked ciprofloxacin dimers against drug-
resistant strains of S. aureus requires the presence of a
carboxyl group on both halves of the dimer structure.
These results indicate that each half of the dimer might
participate in equivalent-type interactions at the homo-
dimeric topoisomerase interface; however, unique or
non-equivalent interactions with the topoisomerase, the
DNA, or the putative topoisomerase–DNA complex are
possible. The observation that all piperazinyl-linked
dimers and incomplete dimers maintain their level of
activity against SA 1199 across strains SA 1199-3 and
SA 1199B indicates that piperazine substitution renders
these agents refractory to the NorA efflux pump and the
topoisomerase IV point mutation present in SA 1199B.9
These strains possess varied levels of FQ resistance
where SA 1199-3 is a laboratory-derived mutant of SA
1199 that inducibly overexpresses the norA encoded
efflux pump and has no DNA gyrase or topoisomerase
IV mutations, and SA 1199B is a derivate of SA 1199
that constitutively overexpresses norA and harbors a
topoisomerase IV A subunit substitution (A116E)
known to correlate with raised FQ MICs.7 Therefore,
while the increased potency of dimers 9, 10 and 14 over
ciprofloxacin against SA 1199-3 and SA 1199B requires
both carboxyl groups, this is not a requirement to sim-
ply maintain activity against the fluoroquinolone resis-
tant strains. This observation is more remarkable with
incomplete dimer analogues 12–13 and 15–16, where
only piperazine or fluorophenyl piperazine are linked to
ciprofloxacin. In general, the inherent antibacterial
activity of these analogues is lower than that of the full
dimers (e.g., 9 and 14) and even ciprofloxacin itself,
however, this inherent level of activity against SA 1199
is maintained against SA 1199-3 and SA 1199B, indi-
cating these simplified analogues are still refractory to
the efflux-mediated and mutation-mediated resistance
mechanisms (Table 1).8,9
References and Notes
1. (a) Hooper, D. C. Biochem. Biophys. Acta 1998, 1400, 45.
(b) Gootz, T.; Brighty, K. Med. Res. Rev. 1996, 16, 433. (c)
Hooper, D. C. Emerg. Infect. Dis. 2001, 7, 337. (d) Acar, J.
Clin. Inf. Dis. 1997, 24, S67.
2. Kerns, R. J.; Rybak, M. J.; Kaatz, G. W.; Vaka, F.; Cha,
R.; Grucz, R. G.; Diwadkar, V. U.; Ward, T. D. Bioorg. Med.
Chem. Lett 2003, 13, 1745.
3. Synthesis of 7 and 8 was achieved by heating ciprofloxacin
with 5 equivalents of bis-alkyl chloride in DMF/acetonitrile at
80 ꢀC for 12–24 h in the presence of powdered NaHCO3. The
product was precipitated from the cooled reaction mixture
with excess diethyl ether, centrifuged, decanted and the white
solid separated by flash silica gel chromatography (20:1;
CH2Cl2:CH3OH). Products were characterized by 1H NMR
and mass spectroscopic analysis.
4. Incomplete dimers 11–13, 15, 16 were prepared by heating
7 or 8 with the corresponding piperazine derivatives in aceto-
nitrile at 60–85 ꢀC for 12 to 48 h. When complete by TLC, the
reaction mixture was cooled, filtered, precipitated with excess
diethyl ether, centrifuged and the recovered white solid dried.
Purification was achieved using semi-preparative reversed-
phase (C-18) HPLC employing a linear gradient of acetonitrile
in water (0.1% trifluoroacetic acid), affording separation of
the desired products as a single peak. Acetonitrile was evapo-
rated and the remaining aqueous samples lyophilized provid-
ing products in good to excellent isolated yield (50–90%).
1
Products were characterized by H NMR, mass spectroscopic
analysis, and analytical HPLC, showing a single peak at
l=220 nm and l=254 nm. Dimers 9, 10 and 14 were pre-
pared as previously reported.2
5. MICs were determined by broth microdilution following
NCCLS guidelines using 2-fold dilutions of test compound.
Error for MICs is thereforeÆone dilution.
6. Decarboxylated ciprofloxacin was prepared as previously
1
reported and the structure confirmed by H NMR and mass
spectroscopic analysis; Reuman, M.; Eissenstat, M. A.; Wea-
ver, J. D., III Tetrahedron Lett. 1994, 35, 8303.
7. Ng, E.; Trucksis, M.; Hooper, D. Antimicrob. Agents Che-
mother. 1996, 40, 1881.
8. We have previously shown that symmetric fluoroquinolone
dimers do not interfere with NorA-mediated efflux of ethidium
bromide in SA 1199-B at concentrations up to five times the
MIC and are not competitive substrates for this pump.
Therefore it is likely that the incomplete dimers here similarly
are not substrates for the NorA pump, accounting for their
ability to evade this efflux mechanism.
9. (a) The primary target of ciprofloxacin in S. aureus is
topoisomerase IV. One explanation for the observed activity
Scheme 1. Synthetic approach employed for the synthesis of piper-
azinyl-linked ciprofloxacin analogues described in Table 1.3,4