Addressing the stability of C-capped dipeptide efflux pump inhibitors that potentiate the activity of levofloxacin in Pseudomonas aeruginosa

Article abstract of DOI:10.1016/S0960-894X(01)00033-6

Synthetic optimization of a biologically labile class of dipeptides that function as efflux pump inhibitors to potentiate the antibacterial agent levofloxacin in Pseudomonas aeruginosa has led to the discovery of a related series of compounds that are completely stable in a variety of biological matrices. Other than the stability profile, the in vitro profile of the new series is essentially identical to that observed with the original one. A prototypical compound from the new series demonstrates potentiation in an in vivo model of infection.

Full text of DOI:10.1016/S0960-894X(01)00033-6

Bioorganic & Medicinal Chemistry Letters 11 (2001) 663±667
Addressing the Stability of C-Capped Dipeptide E‚ux Pump
Inhibitors that Potentiate the Activity of Levo¯oxacin in
Pseudomonas aeruginosa
Thomas E. Renau,a,* Roger Leger,a Eric M. Flamme,a Miles W. She,a
Carla L. Gannon,a Kristina M. Mathias,a Olga Lomovskaya,a
Suzanne Chamberland,a Ving J. Lee,a Toshiharu Ohta,b
Kiyoshi Nakayamab and Yohei Ishidab
aMicrocide Pharmaceuticals, Inc., 850 Maude Ave., Mountain View, CA 94043, USA
bNew Product Research Laboratories I, Daiichi Pharmaceutical Co. Ltd., 1-16-13, Kita-Kasai, Edogawa, Tokyo 134-8630, Japan
Received 2 October 2000; accepted 3 January 2001
AbstractÐSynthetic optimization of a biologically labile class of dipeptides that function as e‚ux pump inhibitors to potentiate the
antibacterial agent levo¯oxacin in Pseudomonas aeruginosa has led to the discovery of a related series of compounds that are com-
pletely stable in a variety of biological matrices. Other than the stability pro®le, the in vitro pro®le of the new series is essentially
identical to that observed with the original one. A prototypical compound from the new series demonstrates potentiation in an in
vivo model of infection. # 2001 Elsevier Science Ltd. All rights reserved.
E‚ux of antibacterial agents is increasingly being
recognized as a major cause of microbial resistance.1À4
Studies of clinical isolates that overexpress e‚ux pumps
have demonstrated that nearly all antibacterials are
substrates for one or more of these pumps. Conse-
quently, several strategies to overcome e‚ux-mediated
resistance have been initiated. While one approach has
focused on identifying derivatives of known classes that
are poor substrates for the e‚ux pumps,5,6 another has
targeted the inhibition of these pumps so as to enhance
the activity of existing antibiotics.7,8
activity of LVFX 8-fold at concentrations as low as
2.5 mg/mL but were unstable in several biological
matrices such as mouse and rat human serum. These
results were understandable given that the compounds
contained natural (l) amino acids. Recognizing that
overcoming this problem was of critical importance for
the program, we devised a plan whereby we would con-
tinue our SAR studies using the original series as a
template and simultaneously initiate an e€ort to address
the instability of the entire series. The present report
describes our e€orts in this area. We show that follow-
ing iterative structural modi®cations a class of stable,
structurally related inhibitors was identi®ed that
We have pursued the latter approach and recently
reported the identi®cation of a class of C-capped
dipeptides that function as bacterial e‚ux pump inhibi-
tors (EPIs) to potentiate the activity of the ¯uoro-
quinolone levo¯oxacin (LVFX) in Pseudomonas
aeruginosa.9,10 The compounds, exempli®ed by l-phe-
nylalanine-l-ornithine-b-naphthylamine (l-Phe-l-Orn-
b-Na, 1a; Chart 1), were stable in growth media (e.g.,
Mueller Hinton Broth, MHB) and potentiated the
*Corresponding author. Fax: +1-650-428-3550; e-mail: tomrenau@
microcide.com
Chart 1.
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00033-6

664
T. E. Renau et al. / Bioorg. Med. Chem. Lett. 11 (2001) 663±667
maintains all of the favorable biological features of the
original series.
studies of 2 in Briton±Robinson bu€er over a pH range
of 4±8 demonstrated that the d-amino group of
ornithine readily cyclized to form a lactam, liberating b-
naphthylamine (Fig. 1). The lactam, like l-Orn-b-Na,
did not potentiate LVFX (data not shown).
Compound 1a, which exempli®es the original series,9
potentiates the activity of LVFX 8-fold at 5 mg/mL but
is unstable in the presence of rat serum (Table 1). The
initial product of serum degradation was identi®ed by
LC±MS as l-Orn-b-Na (data not shown). This ana-
logue had been prepared previously as part of our initial
SAR studies and was shown not to potentiate LVFX.9
To address the instability of the original series, we pre-
pared the N-methyl derivative 2 according to Scheme 1.
While compound 2 was as potent as 1a in vitro
(MPC8=5 mg/mL) it was considerably more stable than
1a in rat serum (Table 1); however, additional stability
The stability results with 2 led us to examine additional
structural motifs that might retain potentiation activity.
Replacing the ornithine was a logical ®rst step, but we
had previously demonstrated that although several
groups could substitute for ornithine, their introduction
typically led to a reduction in potency.9 Standard
approaches such as replacing the amide linkage of the
capping group with an ether or hydroxyethylene moiety
were considered. An alternative strategy involved
Table 1. Activity and stability pro®les of study compounds
Stability in rat seruma; % recovery at indicated time
3 min 5 min 10 min
80 64 36
Compd
1 min
2 h
MICb (mg/mL)
MPC8c (mg/mL)
1ad
2
3
4d
100
100
100
100
100
Ð
Ð
Ð
95
0
512
256
512
>512
>512
5
5
10
10
10
99
87
98
73
100
76
98
55
97
38
100
25
MC-207,110e
aStability of analogues was determined as described in the text. Values are the average of two or more experiments.
bMIC: minimum concentration (mg/mL) of e‚ux pump inhibitor required to inhibit the growth of PAM 1032, a laboratory strain of Pseudomonas
aeruginosa that overexpresses the MexAB-OprM e‚ux pump.12
cMPC8: minimum concentration (mg/mL) of e‚ux pump inhibitor required to reduce (potentiate) the MIC of levo¯oxacin 8-fold; analogues were
evaluated in PAM 1032.
dSee Chart 1. Analogue 1a is referred to as compound 2 in ref 9.
eMC-207,110 is referred to as compound 1 in ref 9.
Scheme 1. Synthesis of compound 2.
Figure 1. Lactam formation from 2.

T. E. Renau et al. / Bioorg. Med. Chem. Lett. 11 (2001) 663±667
665
switching the relative positions of the two amino acids
(Fig. 2). This approach furnished compounds such as
l-Orn-l-Phe-b-Na (3), the synthesis of which is shown
in Scheme 2.
original series counterpart, l-hPhe-l-Orn-3-NHQ (1b;
Chart 1). This was quite fortuitous since 4d, which
incorporated two unnatural (d) amino acids, was com-
pletely stable up to 24 h when incubated in mouse and
rat serum (Table 1). In addition, 4d, like 3, did not form
a lactam under basic conditions.
Compound 3 did potentiate LVFX, although it was less
potent than 1a and 2 (Table 1). Serum stability studies
of 3 demonstrated that it was unstable in rat serum
(recovery @ 10 min=38%, Table 1) but this result was
not surprising given that 3 still contained two amino
acids of the l-con®guration. Notably, in a more extensive
stability study compound 3 demonstrated no propensity
to form a lactam.
Compound 4d demonstrated little to no potentiating
activity (MPC8>4 0 mg/mL) against a strain of P. aeru-
ginosa lacking all three primary e‚ux pumps (PAM
1626), con®rming that it was acting by inhibition of the
pumps. Similar results were observed for 1a±1b.9,10
Further SAR studies supported the initial results
observed for 4d, demonstrating unequivocally that
compounds of the new series that incorporated both d
amino acids were stable under a variety of biological
conditions and had essentially the same potentiation
activity as compounds of the original series that con-
tained the natural (l) amino acids. These results were
extended in vivo where 4d, in combination with LVFX,
showed a signi®cant e€ect on the growth of P. aerugi-
nosa in a murine neutropenic thigh model. Both 4d and
LVFX alone resulted in growth similar to the untreated
controls. The potentiation activity in vivo was shown
not to be due to a pharmacokinetic interaction between
4d and LVFX resulting in elevated LVFX concentrations
(data not shown).
We capitalized on this observation by exploring the
SAR around 3. Interestingly, various compounds from
the new series were consistently 2- to 4-fold less potent
when compared to their original series counterpart
(such as 3 vs 1a). Within the new series, however, the
SAR pro®le of the compounds was essentially identical
to the original series.9 For instance, replacement of
phenylalanine (such as in 3) with homophenylalanine
(hPhe) consistently gave a 2-fold improvement in
potentiation activity. Similarly, replacement of b-naph-
thylamine with the 3-aminoquinoline (3-NHQ) moiety,
to furnish compounds such as l-Orn-l-hPhe-3-NHQ
(4a), gave derivatives having the best balance of potency
and intrinsic antibacterial activity.
The e€ect of stereochemistry on the potentiation activ-
ity was explored by comparing the activity of the four
isomers of l-Orn-l-hPhe-3-NHQ (4a). The results, dis-
played in Table 2, demonstrate that the activity of 4d
(MC-02,595, Chart 2) was essentially the same as its
In summary, we have identi®ed, via iterative synthetic
manipulations, a class of EPIs that is completely stable
to various biological matrices and that potentiate the
activity of LVFX both in vitro and in vivo in a manner
similar to that of the original series. The preparation
Figure 2. Switching the relative positions of the amino acids.
Scheme 2. Synthesis of compound 3.

666
T. E. Renau et al. / Bioorg. Med. Chem. Lett. 11 (2001) 663±667
Table 2. Role of stereochemistry on levo¯oxacin potentiation
MIC to levo¯oxacin (mg/mL in the presence of e‚ux pump inhibitora
E‚ux pump inhibitor concentration (mg/mL)
Compd
X
Y
0
5
10 20
40
MPC8b (mg/mL)
4a
4b
4c
4d
1bc
L
D
L
D
Ð
L
L
D
D
Ð
4
4
4
4
4
2
4
4
1
1
1
2
0.5
0.125
0.125
0.015
0.015
0.03
20
20
10
10
10
0.125
0.06
0.03
0.03
0.5
0.125
0.06
aCompounds were evaluated in PAM 1032, a laboratory strain of Pseudomonas aeruginosa that overexpresses the MexAB-OprM e‚ux pump.12
bMPC8: minimum concentration (mg/mL) of e‚ux pump inhibitor required to reduce (potentiate) the MIC of levo¯oxacin 8-fold.
cSee Chart 1. Analogue 1b is referred to as compound 12 in ref 9.
and evaluation of additional analogues of 4d, such as
various peptidomimetics, are underway. The results of
these studies will be the subject of future reports.
were subsequently removed and the samples ®ltered
using a 4 mm nylon ®lter with a 0.45 mm pore size. In a
parallel manner, test compounds were also incubated in
H2O. This control, in which all the compounds were
stable, ensured that the stability of the samples was not
a€ected by factors other than serum. Samples were
analyzed by HPLC11 and the results are recorded in
Table 1 as % recovery of unchanged material. The
initial lead,9 l-phenylalanine-l-arginine-b-naphthyl-
amine (MC-207,110), was used as a positive control. The
stability of 2 over a pH range of 4±8 was also measured
in Briton±Robinson bu€er at 35 ꢀC.
Chemicals
Standard peptide couplings were used in the prepara-
tion of the ®nal products. The preparation of 1a±1b is
described in ref 9. The N-methyl derivative 2 was
prepared using the synthesis outlined in Scheme 1.
Scheme 2 illustrates the route employed to synthesize
compounds 3,4a±d. All ®nal compounds were puri®ed
by reverse-phase MPLC and tested as their bis-tri-
¯uoroacetic acid (TFA) salts. The structural identity of
In Vitro Potentiation Assay
1
each compound was con®rmed by H NMR and MS.
LVFX was provided by Daiichi Pharmaceutical Co.,
Ltd. (Tokyo, Japan).
Compounds were assayed in the presence and absence
of LVFX against PAM 1032, a laboratory strain of
P. aeruginosa that overexpresses the MexAB-OprM
pump.12 The activity of the inhibitors was quanti®ed by
the term MPC8 which is the minimum concentration
(mg/mL) of inhibitor required to decrease (potentiate)
the MIC of LVFX 8-fold. The MIC (mg/mL) of each
compound was also determined to ensure that the
potentiation e€ect observed was not due to intrinsic
antibacterial activity of the putative EPI.
Stability Assays
The test compounds were incubated at 37 ꢀC in mouse
or rat serum for periods ranging from 1 min to 24 h. At
the appropriate time, 4 mL of trichloroacetic acid (TCA,
70%) and 200 mL CH3CN were added dropwise to each
sample. Samples were separated by centrifugation at
3600Âg for 6 min at room temperature. Supernatants
Acknowledgements
The authors thank Scott Hecker, William Watkins and
Mary Price for their review of the manuscript. We also
thankMary Price for analytical support.
References and Notes
1. Lawrence, L. E.; Barrett, J. F. Exp. Opin. Invest. Drugs
1998, 7, 199.
Chart 2.

T. E. Renau et al. / Bioorg. Med. Chem. Lett. 11 (2001) 663±667
667
2. Van Bambeke, F.; Balzi, E.; Tulkens, P. M. Biochem.
Pharmacol. 2000, 60, 457.
9. Renau, T. E.; Leger, R.; Flamme, E. M.; Sangalang, J.;
She, M. W.; Yen, R.; Gannon, C. L.; Grith, D.; Chamber-
land, S.; Lomovskaya, O.; Hecker, S. J.; Lee, V. J.; Ohta, T.;
Nakayama, K. J. Med. Chem. 1999, 42, 4928.
10. Lomovskaya, O.; Warren, M. S.; Lee, A.; Galazzo, J.;
Fronko, R.; Lee, M.; Blais, J.; Cho, D.; Chamberland, S.;
Renau, T.; Leger, R.; Hecker, S.; Watkins, W.; Hoshino, K.;
Ishida, H.; Lee, V. Antimicrob. Agents Chemother. 2001, 45, 105.
11. HPLC conditions: Column: PLRP-S100A; 5 mm,
4.6 mmÂ15 cm. Flow rate: 1 mL/min. Mobile phase: 95%
0.1% TFA, 5% acetonitrile ramping to 70±73% bu€er/
organic over 20 min. Diode array UV detection wavelength:
225±245 and 240±280 nm.
3. Poole, K. Antimicrob. Agents Chemother. 2000, 44, 2233.
4. Levy, S. B. Antimicrob. Agents Chemother. 1992, 36, 695.
5. Testa, R. T.; Petersen, P. J.; Jacobus, N. V.; Sum, P. E.;
Lee, V. J.; Tally, F. P. Antimicrob. Agents Chemother. 1993,
37, 2270.
6. Brennan, L.; Duignan, J.; Petitpas, J.; Anderson, M.; Fu,
W.; Retsema, J.; Rainville, J.; Smyth, K.; Su, W.; Sutcli€e, J.
38th Interscience Conference on Antimicrobial Agents and
Chemotherapy, San Diego, CA, 24±27 September, 1998; Amer-
ican Society for Microbiology: Washington, DC, 1998; F-124.
7. Renau, T. E.; Hecker, S. J.; Lee, V. J. In Ann. Rep. Med.
Chem.; Bristol, J. A., Ed.; Academic: New York, 1998; Vol.
33, pp 121±130.
12. Lomovskaya, O.; Lee, A.; Hoshino, K.; Ishida, H.; Mis-
try, A.; Warren, M. S.; Boyer, E.; Chamberland, S.; Lee, V. J.
Antimicrob. Agents Chemother. 1999, 43, 1340.
8. Nelson, M. L.; Park, B. H.; Levy, S. B. J. Med. Chem.
1994, 37, 1355.