Iron transport-mediated drug delivery: Practical syntheses and in vitro antibacterial studies of tris-catecholate siderophore-aminopenicillin conjugates reveals selectively potent antipseudomonal activity

Article abstract of DOI:10.1021/ja303446w

An artificial tris-catecolate siderophore with a tripodal backbone and its conjugates with ampicillin and amoxicillin were synthesized. Both conjugates exhibited significantly enhanced in vitro antibacterial activities against Gram-negative species compared to the parent drugs, especially against Pseudomonas aeruginosa. The conjugates appeared to be assimilated by an induced bacterial iron transport process as their activities were inversely related to iron concentration. The easily synthesized tris-catecolate siderophore has great potential for future development of various drug conjugates to target antibiotic-resistant Gram-negative bacteria.

Full text of DOI:10.1021/ja303446w

Communication
pubs.acs.org/JACS
Iron Transport-Mediated Drug Delivery: Practical Syntheses and In
Vitro Antibacterial Studies of Tris-Catecholate Siderophore−
Aminopenicillin Conjugates Reveals Selectively Potent
Antipseudomonal Activity
Cheng Ji, Patricia A. Miller, and Marvin J. Miller*
Department of Chemistry and Biochemistry, University of Notre Dame, 251 Nieuwland Science Hall, Notre Dame, Indiana 46556,
United States
S
* Supporting Information
since P. aeruginosa is intrinsically resistant to many of the
ABSTRACT: An artificial tris-catecolate siderophore with
a tripodal backbone and its conjugates with ampicillin and
amoxicillin were synthesized. Both conjugates exhibited
significantly enhanced in vitro antibacterial activities
against Gram-negative species compared to the parent
drugs, especially against Pseudomonas aeruginosa. The
conjugates appeared to be assimilated by an induced
bacterial iron transport process as their activities were
inversely related to iron concentration. The easily
synthesized tris-catecolate siderophore has great potential
for future development of various drug conjugates to target
antibiotic-resistant Gram-negative bacteria.
existing antibiotics, including β-lactams. Inadequate penetration
through the cell envelope is a significant factor in the resistance
of P. aeruginosa to β-lactam antibiotics, such as ampicillin and
amoxicillin.11,12 The use of drug conjugates of native side-
rophores of P. aeruginosa, namely pyoverdin and pyochelin, to
promote active uptake of antibiotics has been explored.13−17
Conjugates of aminopenicillin with artificial bis-, tris- and
tetrakis-catecholate siderophores based on polyamino carbox-
ylic acids have also been synthesized and found to effectively
inhibit growth of P. aeruginosa.18−21
Like many other types of bacteria, strains of P. aeruginosa
have developed receptors to recognize and transport Fe(III)−
siderophore complexes from other species (xenosiderophores)
in order to gain a competitive growth advantage.22 Enterobactin
(Figure 1) is a tris-catecholate siderophore primarily found in
he development of antimicrobial resistance among Gram-
T
negative pathogens poses a serious threat to global public
health.1,2 Among several mechanisms of antibiotic resistance in
Gram-negative bacteria, a major problem is the low
permeability of their outer membrane which serves as a barrier
to prevent antibiotic uptake by passive diffusion.3 Thus, the
development of methods to overcome this permeability-
mediated resistance is an important therapeutic goal. During
the course of infection, most microbes assimilate physiologically
essential iron by synthesizing and utilizing high-affinity ferric
ion chelators, called siderophores. The Fe(III)−siderophore
complexes are recognized, and active transport through the
bacterial cell membrane is intitiated via specific receptors.4,5
Attachment of antibiotics to siderophores produces potential
“Trojan Horse” conjugates that are anticipated to enter
pathogenic bacteria via their iron uptake system, thereby
circumventing the permeability-mediated drug resistance
problem. While studies of both natural and artificial side-
rophore−drug conjugates (sideromycins) have demonstrated
their potential for development of selective antimicrobial
agents,6,7 additional studies are needed. Herein, we report
that synthetic conjugates of a relatively simple siderophore
mimic and penicillins do, in fact, have selectively potent
antipseudomonal activity, while the parent antibiotics,
themselves, are inactive.
Figure 1. Structures of enterobactin 1 and triscatecholate siderophore
2.
Gram-negative bacteria, such as Escherichia coli and Salmonella
typhimurium.23 Enterobactin can also promote iron uptake into
P. aeruginosa, and the uptake is specifically inducible by
enterobactin under iron-limiting conditions.24 Using enter-
obactin as a shuttle to deliver antibiotics into P. aeruginosa and
other producing bacteria is an attractive strategy to pursue.
However, it is also synthetically challenging as enterobactin has
no functionality or site suitable for drug conjugation.
Investigation of receptor binding and transport of Fe(III)-
enterobactin has revealed that an unsubstituted tris-catechol
iron center and the coordinated catechol amide groups are
Pseudomonas aeruginosa is an opportunistic Gram-negative
bacterium that endangers immunocompromised patients,
including those with cystic fibrosis (CF),8 cancer,9 or AIDS.10
Once the infection is established it is very difficult to eradicate
Received: April 10, 2012
Published: June 4, 2012
© 2012 American Chemical Society
9898
dx.doi.org/10.1021/ja303446w | J. Am. Chem. Soc. 2012, 134, 9898−9901

Journal of the American Chemical Society
Communication
a
Scheme 1. Syntheses of the Tris-Catecholate Siderophore and Its Aminopenicillin Conjugates 10 and 11
a
Reagents and conditions: (a) 1. BH3-THF, THF, reflux; 2. Boc2O, Et3N, MeOH, reflux, 83% for 2 steps; (b) NiCl2, NaBH4, MeOH, sonication, rt,
92%; (c) methyl succinyl chloride, Et3N, CH2Cl2, 0 °C to rt, 75%; (d) 6 N HCl, reflux, 100%; (e) 2,3-diacyloxybenzoyl chloride 8, aqueous
NaHCO3/THF, 0 °C to rt, 57%; and (f) 1. isobutyl chloroformate, N-methyl morpholine, THF, 0 °C; 2. ampicillin or amoxicillin, Et3N, THF/H2O,
0 °C to rt, 55% (for 10), 50% (for 11).
a
Table 1. Diameter of Growth Inhibition Zones (mm) in the Agar Diffusion Antibacterial Susceptibility Assay
b
10
11
ampicillin
amoxicillin
+Fe
−Fe
+Fe
−Fe
+Fe
−Fe
+Fe
−Fe
P. aeruginosa KW799/wt
P. aeruginosa KW799/61
P. aeruginosa PAO1
20
13P
0
24
21
16
18
0
25
13P
0
26
26
19
20
0
23p
27
42
0
18p
40
40
39
0
0
0
P. aeruginosa Pa4
0
0
0
h
0
h
P. aeruginosa Pa6
0
0
0
0
0
0
P. aeruginosa K648 (pvd−, pch−)
E. coli H1443 (fepA+, cir+, fiu+)
E. coli H1876 (fepA−, cir−, fiu−)
0
15
19
0
0
17
21
14p
0
19/23P
0
0
16/23p
0
0
0
19/23p
0
0
0
19/24P
0
0
0
a
p, partially clear inhibition zone/colonies in the inhibition zone; P, unclear inhibition zone/many colonies in the inhibition zone; h: indicates only a
hint of growth inhibition detectable; exactly 50 μL of a 0.2 mM solution of each compound dissolved in 1:9 DMSO/MeOH was added to 9 mm
wells in agar media (standard I nutrient agar, Serva or Mueller−Hinton II Agar, Becton, Dickinson and Company); inhibition zones were read after
b
incubation at 37 °C for 24 h. Compound 10 was tested at 0.1 mM.
essential for recognition as a siderophore, while the trilactone
backbone is not required and can be replaced with very
different molecular scaffolds.23 Building on this important
information, and with the goal of developing practical syntheses
of microbe-selective sideromycins, we anticipated that
simplified symmetrical siderophore analogs, like 2, with linkers
remote from the site of iron binding would be suitable for
conjugation to antibiotics.
The syntheses of triscatecholate siderophore 2 and its
aminopenicillin conjugates (10 and 11) are summerized in
Scheme 1. As the backbone of the artificial siderophore,
tricarbamate 4 was synthesized from trinitrile 3 according to a
published procedure25 with minor modifications. In brief,
trinitrile 3 was reduced with borane in THF, and the resulting
triamine was protected with Boc2O to give tricarbamate 4.
Reduction of the nitro group in 4 was accomplished with
NaBH4 in the presence of a catalytic amount of NiCl2 to give
amine 5 in 92% yield. Treatment of amine 5 with methyl
succinyl chloride provided a succinate derivative 6, of which the
carboxyl terminus later served as the coupling site with the
aminopenicillins. Treatment of 6 with 6 N HCl effected
simultaneous removal of the Boc protecting groups and
hydrolysis of the methyl ester to give triamino acid 7 as its
tris HCl salt in quantitative yield. As demonstrated by
Mollmann et al.,21 using acylated catecholates as the side-
̈
rophore components has the benefit of not only facilitating
synthesis but also preventing pharmacological side effects of the
catechol groups. The acylated catecholates are expected to
serve as prodrugs to the required iron binding catechols while
circumventing potential methylation by a catechol O-methyl-
transferase (COMT) that would lead to a permanent loss of
effective iron binding ability necessary for siderophore
activity.26 Therefore, triamino acid 7 was acylated with 2,3-
diacyloxybenzoyl chloride 8 in aqueous sodium bicarbonate to
give tripodal siderophore 9, which can serve as a common
intermediate for attachment of various amino- or hydroxyl-
containing drugs via an amide or ester linkage. As a proof-of-
principle study, two aminopencillins, ampicillin and amoxicillin,
which by themselves are not active against wild-type strains of
P. aeruginosa, were chosen for the syntheses of conjugates.
Thus, siderophore 9 was coupled via its mixed anhydride with
ampicillin and amoxicillin to provide conjugates 10 and 11,
respectively.
Conjugates 10 and 11 were first evaluated for their
antibacterial ability against a collection of strains of P.
aeruginosa using the agar well diffusion test. The influence of
9899
dx.doi.org/10.1021/ja303446w | J. Am. Chem. Soc. 2012, 134, 9898−9901

Journal of the American Chemical Society
Communication
Table 2. In Vitro Antibacterial Activities of the Siderophore−β-lactam Conjugates 10 and 11 (MIC in μM)
10
11
ampicillin
amoxcillin
+Fe
−Fe
+Fe
−Fe
+Fe
−Fe
+Fe
−Fe
P. aeruginosa KW799/wt
P. aeruginosa KW799/61
P. aeruginosa PAO1
33
12.5
50
0.05
0.067
0.39
25
12.5
50
0.05
0.083
0.39
>200
0.52
>200
0.78
>200
0.46
>200
0.39
>200
>200
>200
16.7
>100
>100
>200
12.5
>200
>200
>200
4.17
>100
>100
>200
4.17
P. aeruginosa Pa4
25
0.39
25
0.21
P. aeruginosa Pa6
>200
150
>200
>200
1.56
>200
100
>200
>200
6.25
E. coli ATCC 25922
K. pneumoniae ATCC 8303 X68
>100
>100
>200
>100
100
>100
a
MIC90 values (μM) were determined using the broth microdilution method in Mueller−Hinton broth no. 2 (MHII) with visual end point analysis
according to the CLSI guidelines.31 Each compound was tested in triplicate.
b
iron concentration on antibacterial activity was probed as well
by conducting assays in iron-rich and iron-deficient media. As
shown in Table 1, the parent drugs, ampicillin and amoxicillin,
were inactive or only weakly active against wild-type strains of
P. aeruginosa (KW799/wt, PAO1, Pa4, and Pa6) due to low
membrane permeability. This was confirmed by studies of
inhibition of the P. aeruginosa permeable mutant K799/61 in
the assay, where both drugs are highly active as indicated by the
large zones of inhibition they induced. Thus, strains of P.
aeruginosa have the target of these classical antibiotics, but they
cannot normally be accessed by passive diffusion. Addition of
the siderophore portion to ampicillin or amoxicillin significantly
increased their antibiotic activity against wild-type P. aeruginosa
strains except Pa6, especially in iron-deficient media, which
reflects the situation in vivo in the infected host. As two control
samples, siderophore 9 and its conjugate with phenyl-
glycinamide (an analog of 10 without the β-lactam fragment,
see Supporting Information (SI)) did not show any inhibitory
effect in the agar well diffusion test, clearly indicating that the
observed activity of the conjugates 10 and 11 was totally due to
the β-lactam warhead.
It is known that the expression of genes that encode
siderophore transport systems is induced in bacteria by low-
iron availability but repressed when iron is sufficient.27
Therefore, the augmented activities of conjugates 10 and 11
most likely represent the increased expression of appropriate
siderophore receptors at the pathogen cell surface under iron-
limited conditions. This was better demonstrated in assays with
P. aeruginosa K648, a strain deficient for its native siderophore
pyoverdin and pyochelin biosynthesis (Table 1). While both
conjugates 10 and 11 were inactive against the K648 strain in
the iron-rich medium, the activities were largely elevated when
tested in the iron-deficient medium, indicating that uptake of
the conjugates was induced under iron-restricted conditions.
Interestingly, both conjugates 10 and 11 showed no activity
against a clinical isolate P. aeruginosa Pa6 even under iron-
deficient conditions. It is known that Pa6 as well as PAO1 and
Pa4 are different from each other in the type of pyoverdins they
produce and utilize.28 It will be of great interest to further
investigate the relationship between xenosiderophore usage and
native siderophore production in P. aeruginosa.
indicating that the conjugates use the enterobactin-mediated
iron transport system to penetrate the bacterial outer
membrane barrier.
Conjugates 10 and 11 were subjected to further assays to
determine their minimum inhibitory concentrations (MIC) in
both iron-rich and iron-deficient media (Table 2). Both
conjugates exhibited excellent antibacterial activity against the
various wild-type strains of P. aeruginosa in iron-deficient
medium, with MICs ranging from 0.05 to 0.39 μM, while
ampicillin and amoxicillin were generally inactive (>100 μM).
The only exception was Pa6, against which both conjugates
were inactive, consistent with the observation from the agar
diffusion assay. In iron-rich media, the inhibitory activities of
the conjugates were largely impaired, further demonstrating
that iron concentration of the media regulates the expression of
siderophore outer membrane receptors and is thus inversely
related to the activity of siderophore−drug conjugates.
Conjugates 10 and 11 were also tested against selected strains
of E. coli and Klebsiella pneumoniae, both of which are able to
synthesize and utilize enterobactin and its degraded product for
iron uptake under iron-limited conditions.29 When tested
against E. coli, conjugate 10 showed an 8-fold increase relative
to ampicillin, while conjugate 11 was almost as equipotent as
amoxicillin. In sharp contrast to the activity enhancement
observed in P. aeruginosa strains, both conjugates were found to
be inactive (>100 μM) against K. pneumoniae. It appears that P.
aeruginosa, E. coli, and K. pneumoniae, either induced or
inherently, have different abilities to use triscatecholate 2 as a
siderophore for iron uptake. Another possibility for the high
MICs from K. pneumoniae is that resistant mutants could
develop and proliferate in the time course of the assay. Similar
phenomena were observed in studies of E. coli exposed to
catechol-based siderophore−loracarbef conjugates30 and P.
aeruginosa exposed to pyoverdin−ampicillin conjugates.14
Attempts to isolate mutants and studies of bacterial growth
kinetics in the presence of the conjugates are currently under
investigation.
In summary, an artificial tris-catecolate siderophore with a
tripodal backbone and its conjugates with ampicillin and
amoxicillin were synthesized and studied. Relative to the parent
drugs, both conjugates exhibited significantly enhanced in vitro
antibacterial activities against Gram-negative species, especially
against P. aeruginosa. The conjugates use energy-dependent
active bacterial iron uptake systems to bypass the Gram-
negative outer membrane permeability barrier, which accounts
for their increased activities. The easily synthesized tris-
catecolate siderophore can and will be utilized in future
development of new drug conjugates that have different cellular
targets and modes of action against Gram-negative bacteria.
A further investigation of dependence of antibacterial activity
on siderophore receptors was conducted with wild-type E. coli
H1443 and mutant H1876, which has defects in its enter-
obactin-mediated iron transport system (Table 1). Compared
to the wild-type strain H1443, the activities of both conjugates
drastically decreased against the triple mutant H1876 (fepA−,
cir−, fiu−, genes that encode the receptors of Fe(III)-
enterobactin and its hydrolysis products in E. coli), clearly
9900
dx.doi.org/10.1021/ja303446w | J. Am. Chem. Soc. 2012, 134, 9898−9901

Journal of the American Chemical Society
Communication
(21) Mollmann, U.; Heinisch, L.; Bauernfeind, A.; Kohler, T.; Ankel-
Fuchs, D. Biometals 2009, 22, 615.
̈
̈
ASSOCIATED CONTENT
* Supporting Information
Experimental procedures, copies of spectral data, and character-
ization data. This material is available free of charge via the
Internet at http://pubs.acs.org.
■
S
(22) Poole, K.; McKay, G. A. Front. Biosci. 2003, 8, d661.
(23) Raymond, K. N.; Dertz, E. A.; Kim, S. S. Proc. Natl. Acad. Sci.
U.S.A. 2003, 100, 3584.
(24) Poole, K.; Young, L.; Neshat, S. J. Bacteriol. 1990, 172, 6991.
(25) Unciti-Broceta, A.; Holder, E.; Jones, L. J.; Stevenson, B.;
Turner, A. R.; Porteous, D. J.; Boyd, A. C.; Bradley, M. J. Med. Chem.
2008, 51, 4076.
AUTHOR INFORMATION
Corresponding Author
mmiller1@nd.edu
■
(26) Ohi, N.; Aoki, B.; Kuroki, T.; Matsumoto, M.; Kojima, K.;
Nehashi, T. J. Antibiot. 1987, 40, 22.
(27) Crosa, J. H.; Mey, A. R.; Payne, S. M. Iron Transport in Bacteria,
(ASM) Press:Washington, DC, 2004.
(28) Meyer, J. M.; Stintzi, A.; DeVos, D.; Cornelis, P.; Tappe, R.;
Taraz, K.; Budzikiewicz, H. Microbiology 1997, 143, 35.
(29) Fischbach, M. A.; Lin, H. N.; Liu, D. R.; Walsh, C. T. Nat. Chem.
Biol. 2006, 2, 132.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This research was partially supported by grant AI054193 from
the NIH. C.J. also acknowledges a fellowship from the Notre
Dame CBBI training grant (NIH T32 GM075762). We thank
(30) Ghosh, A.; Ghosh, M.; Niu, C.; Malouin, F.; Mollmann, U.;
̈
Miller, M. J. Chem. Biol. 1996, 3, 1011.
(31) Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria
that Grow Aerobically, 8th ed.;Clinical and Laboratory Standards
Institute (CLSI): Wayne, PA, 2009; approved standard document
M07-A7.
Dr. Ute Mollmann (Hans-Knoll-Institute, Jena, Germany) and
̈
̈
Prof. Iain Lamont (University of Otago, Otago, New Zealand)
for providing some strains of bacteria. We gratefully acknowl-
edge the use of the NMR facilities provided by the Lizzadro
Magnetic Resonance Research Center at the University of
Notre Dame (UND) under the direction of Dr. Jaroslav Zajicek
and the mass spectrometry services provided by The UND
Mass Spectrometry & Proteomics Facility (Mrs. N. Sevova, Dr.
W. Boggess, and Dr. M. V. Joyce; supported by the National
Science Foundation under CHE-0741793).
REFERENCES
■
(1) Ho, J.; Tambyah, P. A.; Paterson, D. L. Curr. Opin. Infect. Dis.
2010, 23, 546.
(2) Peleg, A. Y.; Hooper, D. C. N. Engl. J. Med. 2010, 362, 1804.
(3) Pages, J. M.; James, C. E.; Winterhalter, M. Nat. Rev. Microbiol.
2008, 6, 893.
(4) Hider, R. C.; Kong, X. L. Nat. Prod. Rep. 2010, 27, 637.
(5) Braun, V.; Hantke, K. Curr. Opin. Chem. Biol. 2011, 15, 328.
(6) Roosenberg, J. M.; Lin, Y. M.; Lu, Y.; Miller, M. J. Curr. Med.
Chem. 2000, 7, 159.
(7) Ji, C.; Juar
2012, 4, 297.
́ ́
ez-Hernandez, R. E.; Miller, M. J. Future Med. Chem.
(8) Irvin, R. T.; Govan, J. W.; Fyfe, J. A.; Costerton, J. W. Antimicrob.
Agents Chemother. 1981, 19, 1056.
(9) Rolston, K. V.; Bodey, G. P. Cancer Invest. 1992, 10, 43.
(10) Meynard, J. L.; Barbut, F.; Guiguet, M.; Batisse, D.; Lalande, V.;
Lesage, D.; Guiard-Schmid, J. B.; Petit, J. C.; Frottier, J.; Meyohas, M.
C. J. Infect. 1999, 38, 176.
(11) Nikaido, H.; Nikaido, K.; Harayama, S. J. Biol. Chem. 1991, 266,
770.
(12) Zimmermann, W. Antimicrob. Agents Chemother. 1980, 18, 94.
(13) Kinzel, O.; Budzikiewicz, H. J. Pept. Res. 1999, 53, 618.
(14) Kinzel, O.; Tappe, R.; Gerus, I.; Budzikiewicz, H. J. Antibiot.
1998, 51, 499.
(15) Hennard, C.; Truong, Q. C.; Desnottes, J. F.; Paris, J. M.;
Moreau, N. J.; Abdallah, M. A. J. Med. Chem. 2001, 44, 2139.
(16) Rivault, F.; Liebert, C.; Burger, A.; Hoegy, F.; Abdallah, M. A.;
Schalk, I. J.; Mislin, G. L. Bioorg. Med. Chem. Lett. 2007, 17, 640.
(17) Noel, S.; Gasser, V.; Pesset, B.; Hoegy, F.; Rognan, D.; Schalk, I.
J.; Mislin, G. L. Org. Biomol. Chem. 2011, 9, 8288.
(18) Heinisch, L.; Wittmann, S.; Stoiber, T.; Berg, A.; Ankel-Fuchs,
D.; Mollmann, U. J. Med. Chem. 2002, 45, 3032.
̈
(19) Heinisch, L.; Wittmann, S.; Stoiber, T.; Scherlitz-Hofmann, I.;
Ankel-Fuchs, D.; Mollmann, U. Arzneimittelforsch. Drug Res. 2003, 53,
̈
188.
(20) Wittmann, S.; Schnabelrauch, M.; Scherlitz-Hofmann, I.;
Mollmann, U.; Ankel-Fuchs, D.; Heinisch, L. Bioorg. Med. Chem.
̈
2002, 10, 1659.
9901
dx.doi.org/10.1021/ja303446w | J. Am. Chem. Soc. 2012, 134, 9898−9901