Design, synthesis, and evaluation of efflux substrate-metal chelator conjugates as potential antimicrobial agents

Article abstract of DOI:10.1016/j.bmcl.2006.10.094

Maintaining a proper balance of metal concentrations is critical to the survival of bacteria. We have designed and synthesized a series of conjugates of metal chelators and efflux transporter substrates aimed at disrupting bacterial metal homeostasis to a

Full text of DOI:10.1016/j.bmcl.2006.10.094

Bioorganic & Medicinal Chemistry Letters 17 (2007) 707–711
Design, synthesis, and evaluation of efflux substrate–metal
chelator conjugates as potential antimicrobial agents
Yanling Zhang,a C. Eric Ballard,b Shi-Long Zheng,a Xingming Gao,a
Ko-Chun Ko,c Hsiuchin Yang,c Gary Brandt,a Xinhui Lou,b Phang C. Tai,c
Chung-Dar Luc and Binghe Wanga,*
aDepartment of Chemistry, Center for Biotechnology and Drug Design, Georgia State University, Atlanta, GA 30302-4098, USA
bDepartment of Chemistry, North Carolina State University, Raleigh, NC 27695, USA
cDepartment of Biology, Center for Biotechnology and Drug Design, Georgia State University, Atlanta, GA 30302-4098, USA
Received 1 October 2006; revised 26 October 2006; accepted 26 October 2006
Available online 2 November 2006
Abstract—Maintaining a proper balance of metal concentrations is critical to the survival of bacteria. We have designed and syn-
thesized a series of conjugates of metal chelators and efflux transporter substrates aimed at disrupting bacterial metal homeostasis to
achieve bacterial killing. Biological studies showed that two of the compounds had very significant antimicrobial effect with an MIC
value of 7.8 lg/mL against Gram-positive Bacillus subtilis.
Ó 2006 Elsevier Ltd. All rights reserved.
The emergence of drug-resistant bacteria and fungi pre-
sents a major problem in the medical field.1–4 Conse-
quently, there is an urgent need for the development
of novel types of antimicrobial agents targeting unique
mechanisms and pathways. Bacteria and fungi generally
develop drug resistance in three ways: producing metab-
olizing enzymes for the degradation of the drugs, mod-
ifying their targets to render the drugs ineffective, and
expressing a high level of efflux proteins that ‘pump’
the drug out in order to lower its concentration.1,2,4–11
The expression of efflux proteins is an especially com-
mon phenomenon even in non-drug-resistant strains
of bacteria. This is also the mechanism responsible for
multi-drug resistance (MDR).
First, efflux-mediated resistance is not the sole mecha-
nism through which bacteria or cancer develops resis-
tance. Second, efflux (MDR) proteins are a family of
proteins and therefore, inhibiting only one may not
achieve the desired effect of increasing drug concentra-
tion. Third, both bacterial and cancer cells have ways
to compensate for the inhibition of a particular efflux
pump. Considering all these factors, we designed an ap-
proach that takes advantage of the very existence of the
efflux transporters in microbes to develop new antimi-
crobial agents. This novel approach aims to turn the ef-
flux transporters into ‘suicide’ machines, which cause
disruption of metal homeostasis and consequently bac-
terial death. Aimed at testing the feasibility of this new
approach, we have designed, synthesized, and evaluated
a series of nine compounds. Two of these compounds (1
and 6) were shown to have very significant antimicrobial
effect with an minimum inhibition concentration (MIC)
of 7.8 lg/mL.
Conceivably, one can take advantage of the efflux pro-
teins in developing novel antimicrobial agents. Indeed,
the efflux transporters have been targeted for the devel-
opment of therapeutic agents aimed at reversing MDR.
This has been tested on both humans (cancer) and
microbes with only limited success.12–20 The less than
desirable effect of these efflux inhibitors in clinical trials
is most likely due to a combination of three factors.
Design. Again, the goal of the study is to turn MDR-
efflux transporters into ‘suicide’ machines to achieve
disruption of metal homeostasis and consequently bac-
terial death. MDR transporters in bacteria belong to
three major families of proteins. These include the major
facilitator superfamily (MFS) and the resistance-nodula-
tion-division family (RND)—both driven by the proton
motive force,21 and the ABC (ATP-binding cassette)
Keywords: Design; Synthesis; Efflux substrate–metal chelator conju-
gates; Antimicrobial agents.
*
Corresponding author. E-mail: wang@gsu.edu
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.10.094

708
Y. Zhang et al. / Bioorg. Med. Chem. Lett. 17 (2007) 707–711
superfamily driven by ATP as a source of energy.10,22,23
These proteins generally have broad substrate specificity
and can transport a wide variety of antibiotics and other
xenobiotics including tetracyclines, macrolides, b-lac-
tams, fluoroquinolones, isoniazid, rifampicin, ethambu-
tol, etc.21 In our approach of disrupting metal
homeostasis in bacteria, we plan to synthesize conju-
gates of metal chelators with efflux substrates as
potential antimicrobial agents. If the intracellular con-
centration of the target metal is higher than that in the
extracellular medium, one can envision a situation
where the chelator conjugate can passively diffuse into
bacteria in the free form and then bind to a metal in
the intracellular environment. The recognition of the
efflux substrate portion of the conjugate by the efflux
transporter should result in the outward movement of
the conjugate, which helps to ‘drag’ a metal ion with
it. The rapid passive diffusion of the free conjugates into
bacteria and their subsequent efflux after metal chelation
essentially generate an active membrane ‘pore’ for the
target metal and can be an efficient way of disrupting
the metal homeostasis of bacteria, which can lead to cell
death.
linkers, which are unlikely to affect the substrate proper-
ty of the phenothiazine unit, and yet not too long to
have other kind of unintended consequences. For these
reasons, we designed compounds 1–9.
Synthesis. The conjugation chemistry used is either ami-
dation or esterification. Phenothiazines with a carboxyl
group tethered through a 1- to 5-carbon linker were first
synthesized from phenothiazine or 2-trifluoromethyl-
phenothiazine by either a Michael addition34 followed
by hydrolysis leading to 12–13 or alkylation35–37 fol-
lowed by hydrolysis38 leading to 16–17 (Scheme 1). Cou-
pling 16 with 24 followed by deprotection of the Boc
group resulted in the formation of chelator 1. Efflux sub-
strate–metal chelator conjugates 2–9 were obtained by
either amidation or esterification between 12–13, 16–
17, and the appropriate chelators, 20, 21, and 26, respec-
tively (Fig. 1).
Chelators 20, 21, 24, and 26 were synthesized as de-
scribed in Scheme 2 following well-established literature
procedures. Thus, benzylation39 of 1,4,7,10-tetraoxa-13-
aza-cyclopentadecane and bispyridin-2-ylmethylamine
with p-cyanobenzylbromide gave compounds 18 and
19, respectively, which were hydrogenated with Raney-
nickel catalyst to afford chelators 20 and 21, respective-
ly. Chelator 24 was prepared from cyclam in three steps:
benzylation40,41 with p-cyanobenzylbromide, protection
with di-tert-butyl dicarbonate, and hydrogenation. Bro-
mination of 40-(4-methylphenyl)-[2,20:60,200]terpyridine
with NBS and AIBN gave chelator 26.
In designing the conjugates that can be used for the dis-
ruption of metal homeostasis in bacteria, one needs to
consider the substrate(s) to use, the chelator(s) to use,
and the linker position and length for conjugating the
chelator to the substrate. Many antibiotics are known
substrates for the efflux pumps in bacteria. However,
to avoid complication in interpreting the antimicrobial
results, we desire to choose something other than a
known antimicrobial agent as the ‘substrate’ moiety.
There have been recent reports indicating that phenothi-
azine can be recognized by the efflux systems of microor-
ganisms.24–33 Furthermore, the chemistry to conjugate
phenothiazine to a metal chelator is relatively straight-
forward. Therefore, we chose phenothiazine as the sub-
strate moiety. As for the chelators to use, in this
preliminary study we desire to ‘scan’ a few well-known
chelators to see which one would have the best effect.
As for the linker length, we chose some medium-length
Antimicrobial tests. The antimicrobial activities of these
conjugates were evaluated using two model systems: one
Gram-positive (Bacillus subtilis) and one Gram-negative
(Escherichia coli). The effects of these compounds were
very different on Gram-positive and Gram-negative
bacteria. Against Gram-positive B. subtilis, the tetraaza
macrocycle conjugate
1
showed very significant
activity with an MIC of 7.8 lg/mL (Table 1). The
bis(2-picolyl)amine compounds 5–7 also showed good
activities with MIC’s in the range of 7.8–30 lg/mL. It
CO2H
n
R'
n
H
R
N
S
R
N
R
N
S
ii
i
S
12 R = H, n = 2
13 R = CF3, n = 2
16 R = H, n = 1
17 R = H, n = 5
10 R = H, R' = CN, n = 2
R = H
R = CF3
11 R = CF3, R' = CN, n = 2
14 R = H, R' = CO2Et, n = 1
15 R = H, R' = CN, n = 5
iii
Boc
Boc
O
2 - 9, 25
N
N
N
N
N
H
iv
N
S
1
Boc
25
Scheme 1. Synthesis of efflux substrate-metal chelator conjugates. Reagents and conditions: (i)a—acrylonitrile, Bu4N+OHꢀ, 0 °C to rt (to give 10,
55% and 11, 59%); (i)b—ethyl bromoacetate, KOH, NaI, DMF, 50 °C (to give 14, 46%); or c—5-bromovaleronitrile, KOH, NaI, DMF, 50 °C
(to give 15, 65%); (ii) KOH, MeOH, reflux, 75–97%; (iii)a—20,21, or 24, EDCI, HOBT, DCM, rt (to give 2–7, 25, 40–77%); or b—26, NaHCO3,
DMF, rt (to give 8, 50% and 9, 60%); (iv) HCl, dioxane, rt, 67%.

Y. Zhang et al. / Bioorg. Med. Chem. Lett. 17 (2007) 707–711
709
O
O
O
HN
n
O
O
NH HN
N
S
N
N
H
O
N
S
N
HN
n = 2,
n = 5,
2
3
1
O
O
HN
O
n
N
N
N
O
HN
N
S
N
O
F3C
N
O
S
n = 2,
n = 5,
5
6
4
O
N
O
O
n
N
N
HN
N
S
F3C
N
N
N
S
N
7
n = 2,
n = 5,
8
9
Figure 1. Structure of compounds 1–9.
N
NC
O
N
NC
or
O
N
O
NH
O
N
i
O
or
N
HN
O
O
N
O
18
19
N
O
ii
N
H2N
H2N
N
or
O
O
N
O
20
21
Boc
Boc
X
CN
NH HN
NH HN
N
N
N
N
N
NH2
ii
iii
N
N
N
Boc
X
X
22, X = H
24
iv
23, X = Boc
N
N
v
N
N
Br
N
N
26
Scheme 2. Synthesis of chelators 20, 21, 24, and 26. Reagents and conditions: (i) p-cyanobenzyl bromide, K2CO3, DCM, rt, 48–50%; (ii) H2/Raney-
Ni, rt, 72–84%; (iii)a—B[NH(CH3)2]3, toluene, reflux; b—n-BuLi, p-cyanobenzyl bromide, THF, ꢀ30 °C to rt, 51%; (iv) (Boc)2O, NaHCO3, THF, rt,
97%; (v) NBS, AIBN, CCl4, ht, reflux, 60%.

710
Y. Zhang et al. / Bioorg. Med. Chem. Lett. 17 (2007) 707–711
Table 1. Antimicrobial effecta of efflux substrate–metal chelator
conjugates 1–9
#1a
#1b
Compound
ClogP
E. coli (MC4100)
B. subtilis
(mg/mL)
(mg/mL)
1
2.94
3.53
4.65
4.45
6.04
7.15
6.96
7.84
8.96
2
0.0078
2
Xb
Xb
Xb
Xb
Xb
Xb
Xb
Xb
#2a
2
3
0.06
0.5
4
#2b
#3
5
0.03
0.0078
0.016
Xb
6
7
8
#4
9
0.5
0.5
Xb
Phenothiazine
Chelator of 1
Chelator of 5
Ampicillin
Kanamycin
#5
#6
2
0.025
<0.003c
a Assay procedure: grow bacteria to lag phase; mix bacteria with soft
agar (0.7%) and spread in normal LB plates (LB+1.5% agar); dis-
solve 1–9 in DMSO and conduct 2-fold serial dilutions; spot the
solution in the bacterial plates.
b No inhibition zones were detected below 2 mg/mL.
c MIC was not obtained for Kanamycin, 0.003 mg/mL was the lowest
concentration examined. DMSO was used as negative control (data
not shown).
Figure 2. Rows 1a,b: compound 1; rows 2a,b: compound 6; row 3:
compound 8; row 4: phenothiazine; row 5: chelator portion of 1; row 6:
chelator portion of 6. The starting concentrations for rows 1a, 2a, 3, 4,
5, and 6 were 2 mg/mL with serial dilutions of 5- and 4-fold for the first
two, and 2-fold for subsequent tests. Rows 1b and 2b show the results
of the subsequent 2-fold dilutions from rows 1a and 2a, respectively.
The MIC was 7.8 lg/mL for 1 and 6. Bacillus subtilis was the model
organism.
is interesting to note that the compound with the lon-
gest linker (6) in the bis(2-picolyl)amine series seemed
to have the best activity (7.8 lg/mL). The azacrown
ether compounds 2 and 4 and the tripyridine com-
pounds 8 and 9 showed poor or no activities with their
MIC’s generally over 0.5 mg/mL with the exception of
3, which has an MIC of 60 lg/mL. Again, the com-
pound (3) with the longest linker in the tripyridine series
also seems to be most active. One can think of partition
coefficient being a factor affecting the antimicrobial
activities of these compounds because of the increased
ClogP with increased linker length (Table 1), though
it is hard to draw conclusions given the limited number
of compounds studied.
above the usual cutoff point of 500–60042 (molecular
weight: 1, 558 and 6, 613).
It is interesting to note that the chelators for the two
most active compounds, 1 and 6, are both Zn2+ bind-
ers.43,44 Previous research using E. coli as a model
has shown that bacteria concentrate zinc and iron by
several orders of magnitude relative to the concentra-
tion in a typical growth medium until they reach a con-
centration of about 0.1 mM.45 Therefore, theoretically
Zn2+is a good target for developing the intended chela-
tor-based antimicrobial agents due to the expected
differential concentrations of this metal ion inside and
outside of bacteria.43,46,47 However, more work is
needed to elucidate the mechanism of action of these
compounds.
Figure 2 also shows part of the inhibition test results
with compounds 1 and 6 against Gram-positive B. sub-
tilis. Again, compounds 1 and 6 showed very significant
inhibitory activities with an MIC of 7.8 lg/mL (Fig. 2,
rows 1a,b and 2a,b). The three control compounds,
the phenothiazine portion and the chelator portions of
1 and 6, did not show any activities at comparable con-
centrations (Fig. 2, rows 4–6). Specifically, the phenothi-
azine portion (row 4) did not show any inhibitory
activity at or below 0.5 mg/mL and the chelator portions
(rows 5 and 6) of 6 and 1, respectively, did not show any
activity at the highest concentration tested, 2 mg/mL.
Row 3 shows the negative results for compound 8 as a
comparison. Kanamycin and ampicillin were used as
positive controls.
In conclusion, metal ions play critical roles in bacteria.
Consequently, intracellular metal concentrations are
tightly regulated by various cellular machineries. Minor
disruption of the metal-ion homeostasis can lead to a
detrimental effect resulting in cell death. Intuitively, it
is reasonable to think that disruption of the metal
homeostasis would lead to cell stress or death. In this
study, we have designed, synthesized, and evaluated a
series of metal chelator–efflux substrate conjugates.
Such conjugates are designed to disrupt the metal
homeostasis of bacteria by turning MDR efflux trans-
porters of bacteria into ‘suicide’ machines, which help
to ‘drag’ metal ions through the cell membrane. Among
the nine compounds synthesized, two (1 and 6) showed
very significant antimicrobial activities, each with an
MIC of 7.8 lg/mL. Much more biological work is need-
ed to elucidate the mechanism of action. We also hope
that the initial success of this novel approach will stim-
ulate more work in this general area of targeting metal
None of the compounds tested inhibited Gram-negative
E. coli (Table 1). The lower activity against Gram-nega-
tive bacteria is consistent with the notion that the outer
membrane of Gram-negative bacteria might limit the
penetration of those compounds with molecular weight

Y. Zhang et al. / Bioorg. Med. Chem. Lett. 17 (2007) 707–711
711
18. Banerjee, S. K.; Bhatt, K.; Misra, P.; Chakraborti, P. K.
Mol. Gen. Genet. 2000, 262, 949.
19. Dantzig, A. H.; De Alwis, D. P.; Burgess, M. Adv. Drug
Delivery Rev. 2003, 55, 133.
homeostasis and efflux transporters for designing new
antimicrobial agents.
20. Desnottes, J. F. Trends Biotech. 1996, 14, 134.
21. Schweizer, H. P. Genet. Mol. Res. 2003, 2, 48.
22. Yu, E. W.; McDermott, G.; Zgurskaya, H. I.; Nikaido,
H.; Koshland, D. E. Science 2003, 300, 976.
23. Higgins, C. F.; Linton, K. J. Science 2001, 293, 1782.
24. Borges-Walmsley, M. I.; McKeegan, K. S.; Walmsley, A.
R. Biochem. J. 2003, 376, 313.
Acknowledgments
Financial support from the Georgia Cancer Coalition
through a Distinguished Cancer Scientist Award and
the Georgia Research Alliance through an Eminent
Scholar endowment and Eminent Scholar Challenge
grants is gratefully acknowledged.
25. Esposito, B. P.; Epsztejn, S.; Breuer, W.; Cabantchik, Z. I.
Anal. Biochem. 2002, 304, 1.
26. Gracio, A. M.; Gracio, A. J. D.; Viveiros, M.; Amaral, L.
Int. J. Antimicrob. Agents 2003, 22, 347.
Supplementary data
27. Kaatz, G. W.; Moudgal, V. V.; Seo, S. M.; Hansen, J. B.;
Kristiansen, J. E. Int. J. Antimicrob. Agents 2003, 22, 254.
28. Kaatz, G. W.; Moudgal, V. V.; Seo, S. M.; Kristiansen, J.
E. Antimicrob. Agent Chemother. 2003, 47, 719.
29. Kolaczkowski, M.; Michalak, K.; Motohashi, N. Int. J.
Antimicrob. Agents 2003, 22, 279.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2006.10.094.
30. Keshelava, N.; Tsao-Wei, D.; Reynolds, C. P. Clin. Cancer
Res. 2003, 9, 3492.
References and notes
31. Krishnegowda, G.; Thimmaiah, P.; Hegde, R.; Dass, C.;
Houghton, P. J.; Thimmaiah, K. N. Bioorg. Med. Chem.
2002, 10, 2367.
32. Molnar, J.; Foldeak, S.; Nakamura, M. J.; Rausch, H.;
Domonkos, K.; Szabo, M. APMIS 1992, 100, 24.
33. Molnar, J.; Hever, A.; Fakla, I.; Fischer, J.; Ocsovski, I.;
Aszalos, A. Anticancer Res. 1997, 17, 481.
1. Hogan, D.; Kolter, R. Curr. Opin. Microbiol. 2002, 5, 472.
2. Mitscher, L. A.; Pillai, S. P.; Gentry, E. J.; Shankel, D. M.
Med. Res. Rev. 1999, 19, 477.
3. Nikaido, H. J. Bacteriol. 1996, 178, 5853.
4. Nikaido, H.; Zgurskaya, H. I. Curr. Opin. Infect. Dis.
1999, 12, 529.
34. Tierney, M. T.; Grinstaff, M. W. Org. Lett. 2000, 2, 3413.
35. Fanni, S.; Keyes, T. E.; Campagna, S.; Vos, J. G. Inorg.
Chem. 1998, 37, 5933.
36. Hanson, P.; Isham, W. J.; Lewis, R. J.; Stockburn, W. A.
J. Chem. Soc., Perkin Trans. 2 1981, 1492.
5. Chakraborti, P. K.; Bhatt, K.; Banerjee, S. K.; Misra, P.
Biosci. Rep. 1999, 19, 293.
6. Kawamura-Sato, K.; Shibayama, K.; Horii, T.; Iimuma,
Y.; Arakawa, Y.; Ohta, M. FEMS Microbiol. Lett. 1999,
179, 345.
37. Mehta, G.; Sambaiah, T.; Maiya, B. G.; Sirish, M.;
Dattagupta, A. . J. Chem. Soc., Perkin Trans. 1 1995, 295.
38. Godefroi, E. F.; Wittle, E. L. J. Org. Chem. 1956, 21,
1163.
39. Fan, X.; You, J.; Jiao, T.; Tan, G.; Yu, X. Org. Prep.
Proced. Int. 2000, 32, 284.
7. Kohler, T.; Kok, M.; MicheaHamzehpour, M.; Plesiat, P.;
Gotoh, N.; Nishino, T.; Curty, L. K.; Pechere, J. C.
Antimicrob. Agents Chemother. 1996, 40, 2288.
8. Levy, S. B. J. Appl. Microbiol. 2002, 92, 65S.
9. Loddenkemper, R.; Sagebiel, D.; Brendel, A. Eur. Resp. J.
2002, 20, 66S.
40. Bernard, H.; Yaouanc, J. J.; Clement, J. C.; Des Abbayes,
H.; Handel, H. Tetrahedron Lett. 1991, 32, 639.
41. Helps, I. M.; Parker, D.; Morphy, J. R.; Chapman, J.
Tetrahedron 1989, 45, 219.
10. Mizutani, T.; Kurahashi, K.; Murakami, T.; Matsumi, N.;
Ogoshi, H. J. Am. Chem. Soc. 1997, 8991.
11. Ouellette, M.; Kundig, C. Int. J. Antimicrob. Agents 1997,
8, 179.
42. Hancock, R. E. W.; Nikaido, H. J. Bacteriol. 1978, 136,
381.
12. van Zuylen, L.; Nooter, K.; Sparreboom, A.; Verweij, J.
Invest. New Drugs 2000, 18, 205.
43. De Silva, S. A.; Zavaleta, A.; Baron, D. E.; Allam, O.;
Isidor, E. V.; Kashimura, N.; Percarpio, J. M. Tetrahedron
Lett. 1997, 38, 2237.
44. Konig, B.; Pelka, M.; Klein, M.; Dix, I.; Jones, P. G.;
Lex, J. J. Inclusion Phenom. Macrocycl. Chem. 2000, 37,
39.
13. Kaplan, A.; Szabo, L. L. Clinical Chemistry: Interpretation
and Techniques; Lea & Febiger: Philadelphia, 1979.
14. Zubercova, O.; Babusikova, O. Neoplasma 1998, 45, 53.
15. Nacsa, J.; Nagy, L.; Sharples, D.; Hever, A.; Szabo, D.;
Ocsovszki, I.; Varga, A.; Konig, S.; Molnar, J. Anti-
Cancer Drug 1998, 71, 3093.
45. Finnery, L.; O’Halloran, T. Science 2003, 300, 931.
46. Hirano, T.; Kikuchi, K.; Urano, Y.; Higuchi, T.; Nagano,
T. Angew. Chem. Int. Ed. 2000, 39, 1052.
47. Hyun, H. J.; Sohn, J. H.; Ha, D. W.; Ahn, Y. H.; Koh,
J.; Yoon, Y. H. Invest. Ophthalmol. Visual Sci. 2001, 42,
460.
16. Lomovskaya, O.; Watkins, W. J. Mol. Microbiol. Bio-
technol. 2001, 3, 225.
17. Advani, R.; Saba, H. I.; Tallman, M.; Rowe, J. M.;
Wiernik, P. H.; Ramek, J.; Dugan, K.; Lum, B.; Villena,
J.; Davis, E.; Paietta, E.; Litchman, M.; Sikic, B. I.;
Greenberg, P. L. Blood 1991, 93, 787.