3-Arylpiperidines as potentiators of existing antibacterial agents

Article abstract of DOI:10.1016/S0960-894X(01)00330-4

Important resistance patterns in Gram-negative pathogens include active efflux of antibiotics out of the cell via a cellular pump and decreased membrane permeability. A 3-arylpiperidine derivative (1) has been identified by high-throughput assay as a pote

Full text of DOI:10.1016/S0960-894X(01)00330-4

Bioorganic & Medicinal Chemistry Letters 11 (2001) 1903–1906
3-Arylpiperidines as Potentiators of Existing Antibacterial Agents
Atli Thorarensen,a,* Alice L. Presley-Bodnar,a Keith R. Marotti,b Timothy P. Boyle,b
Charlotte L. Heckaman,c Michael J. Bohanon,c Paul K. Tomich,c Gary E. Zurenko,b
Michael T. Sweeneyb and Betty H. Yagib
aMedicinal Chemistry, Pharmacia, Kalamazoo, MI 49007-4940, USA
bBiology I, Pharmacia, Kalamazoo, MI 49007-4940, USA
cDiscovery Technologies, Pharmacia, Kalamazoo, MI 49007-4940, USA
Received 18 October 2000; accepted 7 May 2001
Abstract—Important resistance patterns in Gram-negative pathogens include active efflux of antibiotics out of the cell via a cellular
pump and decreased membrane permeability. A 3-arylpiperidine derivative (1) has been identified by high-throughput assay as a
potentiator with an IC50 ꢀ90 mM. This report details the evaluation of the tether length, aryl substitution and the importance of the
fluorine on antibiotic accumulation. Evaluation of various tether lengths demonstrated that the two-carbon tethered analogues are
optimal. Removal of the fluorine has a modest effect on antibiotic accumulation and the defluorinated analogue 17 is equally potent
to the original lead 1. # 2001 Elsevier Science Ltd. All rights reserved.
The emergence of mutant bacterial pathogens resistant
against a wide range of available antibiotics has resulted
in what some researchers call the ‘post-antimicrobial
era’ when a simple bacterial infection can result in a life-
threatening infection.1 Resistance patterns in bacteria
can be classified into a few separate groups:2 (1) Bac-
terial modification of the antimicrobial target in such a
way that the drug no longer has affinity for binding. (2)
Bacterial production of enzymes that deactivate the
drug. (3) Decrease in the permeability of the bacterial
membrane or active efflux of the drug, resulting in less
drug availability inside the cell.
broad substrate specificity. A large number of anti-
biotics are extruded by these pumps including b-lactams
and tetracycline.8 AcrAB/TolC in Escherichia coli,
which is probably the efflux pump mainly responsible
for resistance in E. coli, is composed of three proteins
that in combination form the functional pump.
Efflux pumps are attractive targets for antibacterial
research since inhibition of them would render many
currently available antibiotics considerably more effec-
tive. There has not been an intense effort in that regard,
and only a handful of reports have appeared in the lit-
erature. The screening of inhibitors of tetracycline-spe-
cific efflux pumps has been described by Yamaguchi et
al.9 They discovered several indane inhibitors of the
pump as monitored by an accumulation of radiolabeled
tetracycline. Another approach to discover novel anti-
biotics has been described by Lewis, which demon-
strated that compounds considered inactive as
antibacterial agents, such as Berberine, have activity in
mutants lacking the efflux pump. This approach allows
for the potential identification of alternative classes of
antibiotics if the compound could be subsequently
altered to minimize efflux.10 Researchers from Micro-
cide have described arginine derivatives such as MC-
207110 as effective efflux pump inhibitors, thus increas-
ing the sensitivity of numerous existing antibiotics.11 An
alternative approach to decrease the resistance to
These patterns of resistance have resulted in increased
research activity in both academic and industrial set-
tings for the discovery of antibiotics that overcome this
challenge of multidrug resistance.3À5 The identification
of efflux pumps in microorganisms as an important
pathway for generating resistance against antibiotics is a
relatively new discovery.6 A diverse set of pumps have
been discovered in a wide variety of both Gram-positive
and Gram-negative organisms. The various transport
proteins in microorganisms have been characterized
based on their origins.7 These pumps show remarkably
*Corresponding author. Tel.: +1-616-833-3962; fax: +1-616-833-
2232; e-mail: atli.thorarensen@pharmacia.com
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00330-4

1904
A. Thorarensen et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1903–1906
pumps is increasing the permeability of the bacterial
membrane. The interplay between membrane perme-
ability and efflux pumps has recently been demonstrated
in Pseudomonas aeruginosa.12 This sensitizing approach
should improve the activity of existing antibiotics, and
such an approach has recently been described utilizing
cholic acid derivatives.13
ꢀ90 mM. In order to simplify the preparation of ana-
logues of 1, we elected to evaluate the importance of the
individual functional groups contained in (Æ)-1. In this
report, we describe the synthesis of analogues specifi-
cally prepared to evaluate the importance of the tether
length and the 4-substituted fluorine.
Alkylation of 4-ketopiperidine is an attractive approach
to access the different tethered aryl analogues of 1 along
with the corresponding desfluoro analogue.15 The
desired alkylating side chains were prepared by a multi-
step synthesis similar to reported literature procedures
(Scheme 1).16 Deprotonation of the ketoester 5 with
various metal hydrides in a range of solvents afforded
the desired alkylated product along with recovered
starting material. The optimal reaction conditions
which were deprotonation of 5 with NaH in 30%
DMSO–THF resulted in a clear reaction mixture; addi-
tion of 3 gave the alkylated product in quantitative yield
(Scheme 2). Decarboxylation of 5 in refluxing 6 N HCl
afforded the decarboxylated product 6 in 64% yield.
Reprotection of 6, followed by hydride reduction, pro-
vided the corresponding alcohols. These diastereomers
were easily separated by silica gel chromatography. Boc
removal provided the cis alcohol 7b and the trans
alcohol 8b.
Lead Identification
In an effort aimed at discovering novel efflux pump
inhibitors, our compound collection was screened in a
high-throughput assay with a sub-lethal concentration
of novobiocin. Any compound that exhibited anti-
bacterial activity against E. coli at the same concentra-
tion (100 mM) in the absence of novobiocin and
compounds that had been reported in the literature as
antibacterial agents were disregarded as hits. Any
remaining compound that inhibited growth against
E. coli in the presence of novobiocin was considered a
putative inhibitor of the pump. In this assay, 1 was
identified as a putative inhibitor (Fig. 1). An accumula-
tion assay of 14C labeled linezolid has been utilized as a
secondary assay for a direct measurement of increased
intracellular concentration of antibiotics in the presence
of putative pump inhibitors.14 In this accumulation
assay it has been determined that (Æ)-1 has an IC50
The alcohols (Æ)-7a/8a proved to be prone to elimina-
tion during fluoride displacement via the activated
intermediate (Scheme 3). Utilizing different bases or
absence of base had no effect upon this result. The sur-
prising discovery was made that the product isolated
from elimination of the trans alcohol was not the
expected cyclic alkene, rather the primary amine 12
resulting from fragmentation. The fragmentation can be
explained in that the activated hydroxyl has the correct
stereoelectronic overlap with the nitrogen, allowing the
nitrogen to assist in the fragmentation, resulting after
workup in the primary amine 12, which was extensively
characterized by NMR and MS.
Figure 1. The piperidine derivative 1 and the Microcide compound
MC-207110.
Numerous attempts to alkylate keto-ester 5 with a two-
carbon electrophile were unsuccessful. Since it was
imperative for us to gain access to the desfluoro ana-
logue of 1, attention was directed to an alternative bond
disconnection utilizing a Wittig olefination as the key
event. The challenge of this route is that hydrogenation
of an alkene in the presence of an aryl bromide is very
difficult. Exploration of the literature revealed only a
Scheme 1.
Scheme 3.
Scheme 2.

A. Thorarensen et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1903–1906
1905
few examples of a styryl olefin being saturated in the
presence of an aryl bromide.17 The Wittig reagent was
easily prepared by refluxing in toluene triphenylpho-
sphine and the arylbromide 3, affording the pure ylide
(Scheme 4).18 The ylide was deprotonated with NaH in
DMSO at 60 ꢁC followed by the addition of the alde-
hyde 13 to afford the desired alkene 14 as a ꢀ2/1 (E/Z)
mixture. The alkene 14 was then saturated under an
atmosphere of hydrogen in EtOH utilizing Adams cat-
alyst affording the alkane 16 in excellent yield with
minimal dehalogenation. This approach afforded access
to a range of different aryl substituted two-carbon
analogues (Table 2).
(Scheme 5). In this instance, the displacement of alcohol
19 with fluoride proved to be very problematic due to
the predomination of alkene formation (alkene/fluoride,
7/3) affording desired analogue, 20 in poor yield.
Each of the analogues were evaluated in the accumula-
tion assay, which measures the accumulation of 14C-
labeled linezolid within wild-type E. coli K12 cells in the
presence of the potentiator.19
Using these assays it was determined that the lead
compound 1 has an IC50 ꢀ90 mM. As seen in Table 1,
one-carbon tether analogues displayed marginal activ-
ity. The desfluoro analogues 17 and 15 have comparable
activity to the lead compound, while the three-carbon
tether analogue 20 had a lower inhibitory effect than the
lead compound. The importance of aryl substitution was
evaluated in a desfluoro series (Table 2). The halogens
are essential for activity since 4-EtO, 4-CN, 3,5-Me and
2-Cl-4-(2MeOPh)Ph substitution afforded inactive
compounds. The Br in 17 could be replaced with Cl
without affecting activity. In general, the 3,4 and 2,5
disubstitution affords the most active compounds, with
activity comparable to the initial lead. Several of the
analogues display very weak antibacterial activity of
their own, but there is no correlation between their
antibacterial activity and their potency as potentiators
(compare compounds 24, 27, and 29, Table 2). The MIC
of linezolid against E. coli in the presence of three
potentiators is displayed in Figure 2. In the absence of
the potentiator, the MIC of linezolid is >128 mg/mL.
The three-carbon analogue was prepared in a similar
fashion as previously described for the preparation of 9
Scheme 4.
Table 1. Accumulation data and representative minimum inhibitory
concentration (MIC) data for selected compounds
Compd
Accumulation
IC50 (mM)
MIC
In this report, we described the discovery of a novel
piperidine derivative with the ability to increase
S. aureusa
E. colib
(mg/mL)
(mg/mL)
MC207110
1
730
90
—
64
256
64
6
>900
>900
>900
>900
600
480
140
185
400
>128
>128
>128
>125
128
>128
32
64
>256
>256
>256
256
256
128
64
128
—
7b
8b
9
10
11
15
17
20
—
aS. aureus strain ATCC 29213.
bE. coli strain ATCC 25922.
Scheme 5.
Table 2. Accumulation data and representative minimum inhibitory concentration (MIC) data for selected two-carbon analogues
Compd
Accumulation IC50 (mM)
MIC
S. aureusa (mg/mL)
E. colib (mg/mL)
Ph
4-EtOPh
4-CNPh
2,5-ClPh
2,5-MePh
3,5-CF3Ph
2,6-ClPh
3,4-ClPh
2-Cl-4-(2-MeOPh)Ph
21
22
23
24
25
26
27
28
29
>900
>900
>900
195
>900
300
290
>128
>128
>128
64
>128
64
>128
128
32
>128
>128
>128
128
>128
64
>128
128
64
200
>900
aS. aureus strain ATCC 29213.
bE. coli strain ATCC 25922.

1906
A. Thorarensen et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1903–1906
7. (a) Saier, M. H., Jr.; Tam, R.; Reizer, A.; Reizer, J. Mol.
Microbiol. 1994, 11, 841. (b) Saier, M. H., Jr.; Paulsen, I. T.;
Sliwinski, M. K.; Pao, S. S.; Skurray, R. A.; Nikaido, H.
FASEB J. 1998, 12, 265.
8. (a) Li, M.-Z.; Srikumar, R.; Poole, K. Antimicrob. Agents
Chemother. 1998, 42, 399. (b) Schnappinger, D.; Hillen, W.
Arch. Microbiol. 1996, 165, 359.
9. Hirata, T.; Wakatabe, R.; Nielsen, J.; Satoh, T.; Nihira, S.-
I.; Yamaguchi, A. Biol. Pharm. Bull. 1998, 21, 678.
10. Hsieh, P.-C.; Siegel, S. A.; Rogers, B.; Davis, D.; Lewis,
K. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 6602.
11. Renau, T. E.; Leger, R.; Flamme, E. M.; Sangalang, J.;
She, M. W.; Yen, R.; Gannon, C. L.; Griffith, D.; Chamber-
land, S.; Lomovskaya, O.; Hecker, S. J.; Lee, V. J.; Ohta, T.;
Nakayama, K. J. Med. Chem. 1999, 42, 4928.
12. Li, X.-Z.; Zhang, L.; Poole, K. J. Antimicrob. Chemother.
2000, 45, 433.
13. Rehman, A. u.; Li, C.; Budge, L. P.; Street, S. E.; Savage,
P. B. Tetrahedron Lett. 1999, 40, 1865.
Figure 2. MIC of linezolid against E. coli K12 in the presence of three
potentiators. In the absence of the potentiator, the MIC of linezolid is
>128 mg/mL. The MIC of all of the potentiators is equal or greater
than 128 mg/mL for this strain of E. coli. The concentration of the
compounds and the MICs of linezolid are reported in mg/mL.
14. (a) Thanassi, D. G.; Suh, G. S. B.; Nikaido, H. J. Bacter-
iol. 1995, 177, 998. (b) E. coli K12 was the bacterial strain used
in these studies. Growth medium was LB plus 0.2% glucose.
Assay medium was 50 mM KPO4 pH 7.0, 1 mM MgSO4, 0.2%
glucose. Bacteria were grown at 37 ꢁC. Cells were rinsed twice
in Assay medium and re-suspended to an OD530 of 8.0. For
each assay, 1 mL of cell suspension was pre-incubated at 37 ꢁC
for 10 min and the putative pump inhibitor was added. After
30 min, radiolabeled linezolid (14C-100766) was added to a
final concentration of 30 mM, and the sample incubated an
additional 15 min. As a positive control, the known pump
inhibitor, CCCP, was added at a final concentration of
100 mM. An aliquot of cells was removed from the sample and
layered over an oil cushion of a 7:3 mix of Dow Corning 550
and Dow Corning 510. The tubes containing the oil mix and
cells were centrifuged at 14,000 rpm for 3 min. A 200 mL aliquot
of water was used to re-suspend the bacterial cells, then 4 mL
of scintillation fluid (Ultima Gold) was added. IC50 (con-
centration at which linezolid accumulation was potentiated by
50%) was calculated by determining the drug concentration at
which the accumulation of 14C-linezolid was half maximal of
that of the CCCP positive control. The per cent inhibition
(%I) was determined as follows: %I=100%Â([14C-linezolid
dpm drug treatedÀuntreated]/[14C-linezolid dpm CCCP
treatedÀuntreated]). (c) Heller, K. B.; Lin, E. C. C.; Wilson,
T. H. J. Bacteriol. 1980, 144, 274.
intracellular concentration of existing antibiotics. Sev-
eral analogues have been prepared aimed at addressing
two questions: first, the importance of the tether length
connecting the aryl ring and the piperidine ring, and
second, the importance of the 4-fluoro substituent in 1.
The results demonstrate that shortening the tether by
one carbon eliminates accumulation of linezolid, while
the three-carbon tether is significantly less potent than
the two-atom tether. Removal of the fluorine sub-
stituent had little effect on activity. These results have
allowed us to explore the importance of aryl substitu-
tion in a simpler system lacking the fluorine substituent
where we have found 3,4- and 2,5-dihalogen substitution
to be optimal.
References and Notes
15. (a) Caroon, J. M.; Clark, R. D.; Kluge, A. F.; Nelson,
J. T.; Strosberg, A. M.; Unger, S. H.; Michel, A. D.; Whiting,
R. L. J. Med. Chem. 1981, 24, 1320. (b) Bonjoch, J.; Linares,
A.; Guardia, M.; Bosch, J. Heterocycles 1987, 26, 2165.
16. Romero, A. G. Chem. Abstr. 1993, 199, 160138; EP
539209.
1. (a) Davies, J. Science 1994, 264, 375. (b) Nikaido, H.
Science 1994, 264, 382.
2. (a) Allen, N. E. In Prog. Med. Chem.; Ellis, G. P., Lus-
combe, D. K., Eds.; Elsevier Science: New York, 1995; Vol.
32, Chapter 4, pp 157–238. (b) Hayes, J. D.; Wolf, C. R. Bio-
chem. J. 1990, 272, 281.
17. Marvel, C. S.; Allen, R. E.; Overberger, C. G. J. Am.
Chem. Soc. 1946, 68, 1088.
3. Goldstein, E. J. Clin. Infect. Dis. 1996, 23, S25.
4. Barriere, J. C.; Berthaud, N.; Beyer, D.; D-Malen, S.; Paris,
J. M.; Desnottes, J. F. Curr. Pharm. Des. 1998, 4, 155.
5. Brickner, S. J.; Hutchinson, D. K.; Barbachyn, M. R.;
Manninen, P. R.; Ulanowicz, D. A.; Garmon, S. A.; Grega,
K. C.; Hendges, S. K.; Toops, D. S.; Ford, C. W.; Zurenko,
G. E. J. Med. Chem. 1996, 39, 673.
18. Rosen, T.; Taschner, M. J.; Heathcock, C. H. J. Org.
Chem. 1984, 49, 3994.
19. An accumulation assay only measures the direct increase
of intracellular concentration of radiolabel. The increased
concentration can be the result of pump inhibition or
improved import into the cell. At this stage, we are unable to
differentiate the two phenomena, but the described com-
pounds are not lytic (Stephen E. Buxser, unpublished results).
6. Nikaido, H. J. Bacteriol. 1996, 178, 5853.