N-Caffeoylphenalkylamide derivatives as bacterial efflux pump inhibitors

Article abstract of DOI:10.1016/j.bmcl.2006.12.059

As part of an ongoing project to identify plant natural products as efflux pump inhibitors (EPIs), bioassay-guided fractionation of the methanolic extract of Mirabilis jalapa Linn. (Nyctaginaceae) led to the isolation of an active polyphenolic amide: N-trans-feruloyl 4′-O-methyldopamine. This compound showed moderate activity as an EPI against multidrug-resistant (MDR) Staphylococcus aureus overexpressing the multidrug efflux transporter NorA, causing an 8-fold reduction of norfloxacin MIC at 292 μM (100 μg/mL). This prompted us to synthesize derivatives in order to provide structure-activity relationships and to access more potent inhibitors. Among the synthetic compounds, some were more active than the natural compound and N-trans-3,4-O-dimethylcaffeoyl tryptamine showed potentiation of norfloxacin in MDR S. aureus comparable to that of the standard reserpine.

Full text of DOI:10.1016/j.bmcl.2006.12.059

Bioorganic & Medicinal Chemistry Letters 17 (2007) 1755–1758
N-Caffeoylphenalkylamide derivatives as bacterial efflux
pump inhibitors
Serge Michalet,a,* Gilbert Cartier,a Bruno David,b Anne-Marie Mariotte,a
a
c
d
d
`
Marie-Genevieve Dijoux-franca, Glenn W. Kaatz, Michael Stavri and Simon Gibbons
aDPM UMR 5063 CNRS-UJF, Equipe de Pharmacognosie, UFR de Pharmacie de Grenoble, Domaine de La Merci,
38706 La Tronche cedex, France
bCentre de Recherche des Substances Naturelles, UMS CNRS/IRPF No. 2597, 3, rue des Satellites, 31432 Toulouse, France
cDepartment of Internal Medicine, Division of Infectious Diseases, School of Medicine,
Wayne State University and the John D. Dingell Department of Veteran Affairs Medical Center, Detroit, MI 48201, USA
dCentre for Pharmacognosy and Phytotherapy, School of Pharmacy, University of London, 29/39 Brunswick Square,
WC1N 1AX London, UK
Received 16 October 2006; revised 15 December 2006; accepted 16 December 2006
Available online 22 December 2006
Abstract—As part of an ongoing project to identify plant natural products as efflux pump inhibitors (EPIs), bioassay-guided frac-
tionation of the methanolic extract of Mirabilis jalapa Linn. (Nyctaginaceae) led to the isolation of an active polyphenolic amide:
N-trans-feruloyl 40-O-methyldopamine. This compound showed moderate activity as an EPI against multidrug-resistant (MDR)
Staphylococcus aureus overexpressing the multidrug efflux transporter NorA, causing an 8-fold reduction of norfloxacin MIC at
292 lM (100 lg/mL). This prompted us to synthesize derivatives in order to provide structure–activity relationships and to access
more potent inhibitors. Among the synthetic compounds, some were more active than the natural compound and N-trans-3,4-O-
dimethylcaffeoyl tryptamine showed potentiation of norfloxacin in MDR S. aureus comparable to that of the standard reserpine.
ꢀ 2007 Elsevier Ltd. All rights reserved.
In recent years bacterial resistance to antibiotics has be-
come a serious problem of public health that concerns
almost all antibacterial agents. Among the mechanisms
involved (target modification, enzymatic inactivation
or reduction of accumulation within the cell), active ef-
flux has received particular attention since it has been
recognised as one of the most important causes of intrin-
sic antibiotic resistance in bacteria.1
Of particular concern is the presence of multidrug resis-
tance (MDR) efflux pumps that extrude a wide range of
structurally unrelated compounds. The NorA protein of
S. aureus is a drug/proton antiporter belonging to the
major facilitator family (MFS) of transporters and is
responsible for the efflux of substances such as norflox-
acin, ciprofloxacin, ethidium bromide and acriflavin.5,6
It has been recognised as one of the major efflux pumps
protecting S. aureus from antibiotics.7
One species is particularly resistant to treatment, nota-
bly Staphylococcus aureus. This pathogen has evolved
from the MRSA phenotype (methicillin-resistant
S. aureus) to the VRSA phenotype (vancomycin-resis-
tant S. aureus), whereas vancomycin was the drug of
last-resort for treatment of MRSA.2 Resistance has also
been reported for newer agents such as linezolid and
daptomycin.3,4
One of the strategies employed to overcome bacterial
resistance is the use of EPIs that could restore antibiotic
activity in resistant strains. This approach is promising
as it would be a way to improve the efficacy and/or ex-
tend the clinical utility of existing antibiotics.8 The com-
bination of a resistance inhibitor with an antibiotic has
already proven its efficacy with the clavulanic acid/
amoxicillin association. Furthermore, this combination
can reduce the in vitro frequency of emergence of resis-
tant mutant strains.9 At present this combination thera-
py is taken into account by a number of pharmaceutical
companies and several strategies are being developed by
different groups.10
Keywords: EPI; Staphylococcus aureus; N-Hydroxycinnamic acid der-
ivatives; Mirabilis jalapa.
*
Corresponding author. Tel.: +33 4 76 51 86 88; fax: +33 4 76 63 71
65; e-mail: sergemichalet@yahoo.fr
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.12.059

1756
S. Michalet et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1755–1758
R2
COOH
Plants use antimicrobials to protect themselves from
environmental stresses, but unlike fungi or bacteria
activities are often weak and with a narrow spectrum.
This poses the interesting question of how plants react
towards increasing antimicrobial resistance. Some
authors postulate that plants use an anti-MDR strategy
to potentiate their antimicrobials.11 This is the case for
the alkaloid berberine which has a weak antimicrobial
activity that is significantly enhanced by 50-meth-
oxyhydnocarpin (50-MHC), a flavonolignan produced
by the same Berberis species.12 50-MHC does not have
any antibacterial activity on its own but potentiates
the activity of berberine by inhibiting its efflux. A num-
ber of inhibitors produced by plants as secondary
metabolites and belonging to many chemical classes
have been identified but few are potent and have a broad
spectrum of action.13 This includes the alkaloid reser-
pine which is used as a standard but cannot be used in
therapy because of its toxicity. Searching for inhibitors
from natural sources is therefore an attractive strategy
to access a greater range of active compounds.
R1
NH2
+
R3
O
R4
R1
R2
R3
N
H
a
15-65%
R4
R2
H
N
R2=H, OH, OMe
R3=H, OH, OMe
R4=H, OH, OMe
R3
R4
R1=
or
Scheme 1. Reagents and condition: (a) BOP, Et3N, DMF, 20 h.
Compound 8 was prepared with 3,4-dimethoxy-dihydrocinnamic acid.
phonium hexafluorophosphate) as a coupling agent
(Scheme 1).19 Nine compounds were synthesized and
their related activities are presented in Table 1.20
MICs of the modulators were determined when good
potentiation was observed; a good resistance reverser
should actually not show any antibacterial effect at
the tested concentrations.
In this context, previous screening of ornamental and
exotic plants showed activity associated with extracts
of Mirabilis jalapa Linn. (Nyctaginaceae) in reversing
fluoroquinolone resistance in S. aureus overexpressing
the NorA MDR efflux pump (SA-1199B strain).14
This plant is used in folk medicine as an anti-infec-
tive.15,16 Bioassay-guided fractionation of the metha-
nolic extract from leaves and stems of M. jalapa led
to the isolation of an active phenolic compound,
namely N-trans-feruloyl 40-O-methyldopamine, which
has already been reported in the plant kingdom.17
This compound caused an 8-fold reduction of norflox-
acin MIC when tested at 100 lg/mL (292 lM). Synthe-
sis of derivatives was then undertaken as this small
molecule was a good starting point for structure–ac-
tivity relationships (SARs), and could be a good can-
didate for combination therapy.
Interpretation of the results led us to draw several
conclusions:
• For the cinnamic part: an hydroxyl substitution on
the aromaticring appears to be better than methoxy
or unsubstituted (compound 2 versus compounds 1
and 6); the double bond is essential for activity.
• For the amine part: trisubstitution on the aromatic
ring led to a decrease in activity but antibacterial
activity was observed (compound 5); methoxy substi-
tution gave better results than hydroxyl substitution
(compound 3 vs compound 1) which, in turn, were
better than no substitution (compound 1 vs com-
pound 7); tryptamine combinations showed the best
results (compound 9).
We chose couplings between cinnamic acid derivatives
and amines that occur in nature, as our project was to
identify active compounds from natural sources. Sev-
eral structure–activity relationship criteria were exam-
ined: substitution on the aromatic ring on each part,
methoxy or hydroxyl substitution influence, double
bond influence, aromatic ring nature on the amine
part (we chose here to test tryptamine combinations
as indolic compounds are known to be good inhibi-
tors).18 We chose N-trans-3,4-O-dimethylcaffeoyl dopa-
mine as lead compound as it showed the same activity
as the natural compound, and its asymetrical substitu-
tion allowed us to have a better understanding of the
influence of each substituant (i.e., hydroxyl or meth-
oxy) on each part of the molecule. Each time, one
part of the lead compound was conserved and the
other part was varied in order to provide structure–ac-
tivity relationships that could lead to further optimisa-
tion. Compounds were obtained by coupling of a
cinnamic acid (or dihydrocinnamic acid in the case
of compound 8) derivative with the corresponding
amine in DMF in the presence of triethylamine and
BOP (benzotriazol-1-yloxy-tris-(dimethylamino)-phos-
Among the 9 compounds tested, compound 9 showed
potentiation of norfloxacin comparable to that of the
reference alkaloid reserpine. In order to confirm the
inhibitory mechanism, we performed an efflux inhibition
assay using ethidium bromide (EtBr), a good substrate
of NorA, whose efflux can be measured by loss of fluo-
rescence within the cell (Fig. 1).21,22
Based on estimation of IC50, compound 9 was approxi-
mately 2-fold less active than reserpine at inhibiting
EtBr efflux (Fig. 1). However, this difference disap-
peared at a concentration of 30 lM for each compound.
Structure–activity relationships (SARs) showed that bet-
ter activities were obtained when the phenyl ring on the
cinnamic portion was substituted with 2 hydroxyls.
According to this result, we would expect N-trans-caf-
feoyl tryptamine (dihydroxylated equivalent of com-
pound 9) to be more active than compound 9. It

S. Michalet et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1755–1758
1757
Table 1. Potentiation of norfloxacin MIC on SA-1199B strain cells by synthetic compounds
Compound
Formula
[Concentration]/fold reduction of norfloxacin MIC
OH
OH
O
[292 lM]/8
[146 lM]/4
MeO
MeO
1 (lead)
N
H
OH
O
[317 lM]/8
[63 lM]/4
MIC P 406 lM
HO
HO
2
3
N
H
OH
OMe
O
[162 lM]/8
[54 lM]/2
MeO
MeO
N
H
OMe
OH
OH
O
[268 lM]/4
MeO
MeO
N
4
5
H
OH
OMe
OH
OH
[56 lM]/0
MIC = 356 lM
O
MeO
MeO
N
H
OH
OH
O
[353 lM]/16
[70 lM]/4
MIC > 452 lM
6
N
H
O
MeO
MeO
[322 lM]/4
[290 lM]/0
7
N
H
OH
OH
O
O
MeO
8
9
N
H
MeO
MeO
NH
[286 lM]/16
[57 lM]/8
[29 lM]/4
N
H
MIC > 366 lM
MeO
would also be interesting to evaluate whether this type
of compound would be a good EPI in Gram-negative
bacteria or in other models.
used to potentiate the bactericidal effect of antibiotics
such as norfloxacin in multidrug-resistant pathogenic
bacteria such as S. aureus. The ease of synthesis and
the small size of these compounds make them potential
candidates for SARs as EPIs in multidrug efflux
systems.
In summary, we have demonstrated that N-cinna-
moylphenalkylamides are a class of EPIs that could be

1758
S. Michalet et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1755–1758
7.16 (d, 1H, J = 2.0 Hz), d 7.13 (dd, 1H, J1 = 8.4 Hz;
J2 = 2.0 Hz), d 6.97 (d, 1H, J = 8.4 Hz), d 6.70 (d, 1H,
J = 8.0 Hz), d 6.69 (d, 1H, J = 2.0 Hz), d 6.57 (dd, 1H,
J1 = 8.0 Hz;J2 = 2.0 Hz),d6.46(d,1H,J = 15.6 Hz),3.87(s,
3H); 3.87 (s, 3H), d 3.48 (t, 2H, J = 7.6 Hz), d 2.68 (t, 2H,
J = 7.6 Hz),RMN13C(100 MHz,CD3OD):d167.6;d150.8;
d 149.3; d 144.9; d 143.4; d 140.2; d 130.7; d 128.0; d 121.8; d
119.6;d118.3;d115.5;d115.0;d111.3;d109.9;d55.0;d55.0;d
Å
41.2;d34.6.EIMS:m/z = 343[M+ ].ForCompound4:RMN
1H(400 MHz,CD3OD):d7.45(d,1H,J = 15.6 Hz),d6.87(s,
2H); d 6.71 (d, 1H, J = 8.0 Hz), d 6.69 (d, 1H, J = 2.0 Hz), d
6.57 (dd, 1H, J1 = 8.0 Hz; J2 = 2.0 Hz), d 6.52 (d, 1H,
J = 15.6 Hz),d3.87(s,3H),d3.87(s,3H),d3.80(s,3H),d3.48
(t, 2H, J = 6.8 Hz), d 2.62 (t, 2H, J = 6.8 Hz); RMN 13C
(100 MHz, CD3OD): d 167.2; d 155.4; d 144.6; d 143.5; d
142.3; d 140.2; d 130.9; d 130.8; d 120.0; d 119.6; d 117.6;
d115.5;d115.0;d114.9;d114.6;d104.9;d59.8;d55.3;d55.3;d
Å
41.1;d34.6.EIMS:m/z = 373[M+ ].ForCompound5:RMN
1H (400 MHz, CD3OD): d 7.39 (d, 1H, J = 15.6 Hz), d 7.08
(d,1H,J = 2.0 Hz),d7.05(dd,1H,J1 = 8.0 Hz;J2 = 2.0 Hz),
d 6.89 (d, 1 H, J = 8.4 Hz), d 6.39 (d, 1H, J = 15.6 Hz), d 6.62
(s,2H),d3.80(s,3H),3.79(s,3H),d3.39(t,2H,J = 7.2 Hz),d
2.57 (t, 2H, J = 7.2 Hz); RMN 13C (100 MHz, CD3OD): d
167.6; d 150.8; d 149.4; d 145.7; d 140.2; d 131.3; d 130.1; d
128.1; d 121.8; d 118.4; d 118.4; d 111.4; d 110.1; d 107.3;Åd
107.3; d 55.1; d 55.1; d 41.0; d 34.9. EIMS: m/z = 359 [M+ ].
For Compound 8: RMN 1H (400 MHz, CDCl3): d 6.70 (d,
1H, J = 8.4 Hz), d 6.68 (d, 1H, J = 8.0 Hz), d 6.60 (d, 1H,
J = 1.6 Hz), d 6.59 (dd, 1H, J1 = nd; J2 = 1,6 Hz), d 6.54 (d,
1H, J = 2.0 Hz), d 6.38 (dd, 1H, J1 = 8.0 Hz; J2 = 2.0 Hz), d
5.71(t,1H,J = 5.6 Hz),d3.75(s,3H),d3.72(s,3H),d3.31(m,
2H), d 2.78 (t, 2H, J = 7.2 Hz), d 2.49 (t, 2H, J = 7.2 Hz), d
2.35 (t, 2H, J = 7.6 Hz); RMN 13C (100 MHz, CDCl3):d
173.5; d 148.9; d 147.5; d 144.3; d 143.1; d 133.1; d 130.6; d
Figure 1. Ethidium efflux inhibition assay from SA-1199B strain cells:
(m) reserpine; (Ç) compound 9.
Acknowledgments
We are thankful to Institut Klorane (Laboratoires Pierre
Fabre) for a PhD studentship for S. Michalet. Region
ˆ
Rhone-Alpes is also acknowledged for the Eurodoc
scholarship program.
References and notes
1. Ryan, B. M.; Dougherty, T. J.; Beaulieu, D.; Chuang, J.;
Dougherty, B. A.; Barrett, J. F. Expert Opin. Investig.
Drugs 2001, 10, 1409.
2. Hiramatsu, K. Lancet Infect. Dis. 2001, 1, 147.
3. Tsiodras, S.; Gold, H.; Sakoulas, G.; Eliopoulos, G.;
Wennersten, C.; Venkataraman, L.; Moellering, R. C.;
Ferraro, M. J. Lancet 2001, 358, 207.
4. Hershberger, E.; Donadebian, S.; Konstantinou, K.;
Zervos, M. J. Clin. Infect. Dis. 2004, 38, 92.
5. Neyfakh, A. A.; Borsh, C. M.; Kaatz, G. W. Antimicrob.
Agents Chemother. 1993, 37, 128.
120.5;d120.4;d115.6;d115.4;d111.8;d111.5;d56.0;d55.9;d
Å
41.01; d 38.7; d 34.8; d 31.3. EIMS: m/z = 345 [M+ ]. For
Compound 9: RMN 1H (400 MHz, CDCl3): d 7.97 (s, 1H), d
7.60(d,1H,J = 8.0 Hz),d7.54(d,1H,J = 15.6 Hz),d7.37(d,
1H, J = 8.0 Hz), d 7.17 (td, 1H, J1 = 8.0 Hz; J2 = 7.2 Hz;
J3 = 1.2 Hz), d 7.10 (td, 1H, J1 = 8.0 Hz; J2 = 7.2 Hz;
J3 = 1.2 Hz), d 7.03 (s, 1H), d 7.01 (dd, 1H, J1 = 8.8 Hz,
J2 = 2.4 Hz), d 6.96 (d, 1H, J = 1.6 Hz), d 6.79 (d, 1H,
J = 8.4 Hz), d 6.24 (d, 1H, J = 15.6 Hz), d 6.12 (t, 1H,
J = 6.0 Hz),d3.87(s,3H),d3.84(s,3H),d3.72(m,2H),d3.02
(t, 2H, J = 6.8 Hz); RMN 13C (100 MHz, CDCl3): d 167.6; d
150.8; d 149.4; d 140.2; d 136.6; d 128.0; d 127.3; d 122.0; d
121.8; d 120.9; d 118.4; d 118.2; d 117.9; d 111.9; d 111.3; d
110.Å8; d 110.0; d 55.0; d 55.0; d 40.2; d 25.0. EIMS: m/z = 350
[M+ ].
6. Nakae, T.; Yoshihara, E.; Yoneyama, H. J. Infect.
Chemother. 1997, 3, 173.
7. Hsieh, P. C.; Siegel, S. A.; Rogers, B.; Davis, D.; Lewis, K.
Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 6602.
8. Lynch, A. S. Biochem. Pharmacol. 2006, 71, 949.
9. Schmitz, F.-J.; Fluit, A. C.; Luckefahr, M.; Engler, B.;
Hofmann, B.; Verhoef, J.; Heinz, H.-P.; Hadding, U.;
Jones, M. E. J. Antimicrob. Chemother. 1998, 42, 807.
10. Poole, K.;Lomovskaya, O. Drug Discovery Today 2006,3,145.
11. Stermitz, F. R.; Lorenz, P.; Tawara, J. N.; Zenewicz, L.
A.; Lewis, K. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 1433.
12. Stermitz, F. R.; Tawara-Matsuda, J. N.; Lorenz, P.;
Muller, P.; Zenewicz, L. A. J. Nat. Prod. 2000, 63, 1146.
13. Gibbons, S. Nat. Prod. Rep. 2004, 21, 263.
14. Cegiela-Carlioz, P. Ph.D. thesis, University of Grenoble I,
April 2004.
15. Siddiqui, S.; Siddiqui, B. S.; Adil, Q.; Begum, S. Fitoter-
apia 1990, 61, 471.
16. Kusamba, C.; Byamana, K.; Mbuyi, W. M. J. Ethnophar-
macol. 1991, 35, 197.
17. Cutillo, F.; d’Abrosca, B.; DellaGreca, M.; Di Marco, C.;
Golino, A.; Previtera, L.; Zarrelli, A. Phytochemistry
2003, 64, 1381.
20. Modulation assays were conducted according to National
Committee for Clinical Laboratory Standards, 1999.
Methods for dilution antimicrobial susceptibility tests for
bacteria that grow aerobically, 5th ed. Norfloxacin MIC
on SA-1199B cell is 100 lM (32 lg/mL), and the reference
alkaloid reserpine causes an 8-fold reduction of norflox-
acin MIC at 33 lM (20 lg/mL).
21. Ethidium bromide assay: EtBr is a substrate for many MDR
pumps, including NorA. The efficiency of efflux pumps for
which EtBr is a substrate can be assessed fluorometrically by
the loss of fluorescence over time from cells loaded with
EtBr. SA-1199B was loaded with EtBr as described previ-
ously and the effects of varying concentrations of compound
9 and reserpine were determined to generate dose–response
profiles. The total time course for the efflux assay was 5 min.
Assays were performed in duplicate and mean results were
expressed as the percentage reduction of total efflux
observed for test strains in the absence of inhibitors.
22. Kaatz, G. W.; Seo, S. M.; O’Brien, L.; Wahiduzzaman, M.;
Foster, T. J. Antimicrob. Agents Chemother. 2000, 44, 1404.
18. Samosorn, S.; Bremner, J. B.; Ball, A.; Lewis, K. Bioorg.
Med. Chem. 2006, 14, 857.
19. Analytical data for new compounds: for Compound 1: 1H
NMR (400 MHz, CD3OD): d 7.46 (d, 1H, J = 15.6 Hz), d